Split (1)

train · 521k rows

text stringlengths 254 1.16M
--- title: Inhibitory Effects of Extracellular Vesicles from iPS-Cell-Derived Mesenchymal Stem Cells on the Onset of Sialadenitis in Sjögren’s Syndrome Are Mediated by Immunomodulatory Splenocytes and Improved by Inhibiting miR-125b authors: - Qingguo Zhao - Eun-Hye Bae - Yu Zhang - Arash Shahsavari - Pranayvir Lotey - Ryang Hwa Lee - Fei Liu journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049013 doi: 10.3390/ijms24065258 license: CC BY 4.0 --- # Inhibitory Effects of Extracellular Vesicles from iPS-Cell-Derived Mesenchymal Stem Cells on the Onset of Sialadenitis in Sjögren’s Syndrome Are Mediated by Immunomodulatory Splenocytes and Improved by Inhibiting miR-125b ## Abstract Extracellular vesicles (EVs) from allogeneic-tissue-derived mesenchymal stem cells (MSCs) are promising to improve Sjögren’s syndrome (SS) treatment, but their application is hindered by high variations in and limited expandability of tissue MSCs. We derived standardized and scalable MSCs from iPS cells (iMSCs) and reported that EVs from young but not aging iMSCs (iEVs) inhibited sialadenitis onset in SS mouse models. Here, we aim to determine cellular mechanisms and optimization approaches of SS-inhibitory effects of iEVs. In NOD.B10.H2b mice at the pre-disease stage of SS, we examined the biodistribution and recipient cells of iEVs with imaging, flow cytometry, and qRT-PCR. Intravenously infused iEVs accumulated in the spleen but not salivary glands or cervical lymph nodes and were mainly taken up by macrophages. In the spleen, young but not aging iEVs increased M2 macrophages, decreased Th17 cells, and changed expression of related immunomodulatory molecules. Loading miR-125b inhibitors into aging iEVs significantly improved their effects on repressing sialadenitis onset and regulating immunomodulatory splenocytes. These data indicated that young but not aging iEVs suppress SS onset by regulating immunomodulatory splenocytes, and inhibiting miR-125b in aging iEVs restores such effects, which is promising to maximize production of effective iEVs from highly expanded iMSCs for future clinical application. ## 1. Introduction Sjögren’s syndrome (SS), a chronic inflammatory autoimmune disease, affects mainly salivary glands and lacrimal glands [1,2]. The consequent long-term dry mouth (xerostomia) exacerbates dental caries and periodontal disease and causes problems of taste, sleep, and speech, which severely impair quality of life. No therapy for SS has demonstrated to be really effective, and current therapeutic management is still based on the symptomatic treatment of sicca symptomatology and a variety of immunosuppressive agents for systemic features [3]. Mesenchymal stem cells (MSCs), multipotent stem cells isolated from bone marrow or various other tissues, can promote regeneration and modulate immune responses mainly through paracrine effects [4]. In preclinical studies and a few small clinical trials, allogeneic- but not autologous-tissue-derived MSCs alleviated xerostomia caused by SS after systemic infusion [5,6]. However, the clinical application of tissue-derived MSCs is hindered by their high functional variations (due to differences in donors and source tissues, methods of isolation and expansion), limited expandability, loss of therapeutic activities after prolonged expansion, safety concerns associated with live cell treatment, dynamic changes in vitro and in vivo, and high cost and infrastructure requirements [7,8,9]. To overcome limitations of tissue-derived MSCs, we derived standardized MSCs efficiently from transgene-free human induced pluripotent stem cells (iPSCs) with a theoretically limitless expandability [10]. The anti-inflammatory and pro-regenerative properties of our iPSC-derived MSCs (iMSCs) are superior or comparable to the best batches of bone marrow MSCs tested in our NIH-funded MSC distribution center and are consistent between different derivation batches [11,12,13]. Extracellular vesicles (EVs), including exosomes and microvesicles, carry bioactive molecules from their parent cells and facilitate the delivery of these molecules into recipient cells. EVs from tissue-derived MSCs show anti-inflammatory and pro-regenerative properties similar to MSCs and appear more feasible for clinical applications than live cells, but their application is still hindered by variations and the limited expandability of source MSCs [14,15]. Recently, we found that when infused intravenously at the pre-disease stage, EVs from early-passage iMSCs (young iEVs) but not from late-passage iMSCs (aging iEVs) can inhibit the onset of sialadenitis with efficiency comparable to young iMSCs and bone marrow MSCs in mouse SS models [12,16]. The loss of SS-inhibitory effects in aging iEVs is related to the decreased TGFβ1 and miR-21 and increased miR-125b levels compared with young iEVs [16]. Here, we report that splenic macrophages are a major type of recipient cells of intravenously infused iEVs, and young iEVs but not aging iEVs inhibit the onset of SS, likely through increasing M2 macrophages and decreasing T helper 17 (Th17) cells in the spleen, whereas inhibiting miR-125b in aging iMSCs make their EVs as competent as young iEVs in these activities. ## 2.1. Systemically Infused iEVs Were Mainly Taken up by Macrophages in Spleen To trace the biodistribution of iEVs in NOD.B10.H2b mice, a model of primary SS, PD15 iEVs (population doubling 15; early-passage iMSC-derived EVs; young iEVs), and PD45 iEVs (late-passage iMSC-derived EVs; aging iEVs) were labeled with a near-infrared fluorescent dye, DiR, as reported in [17], and IV infused into 4-month-old female NOD.B10.H2b mice. At 1, 3, and 24 h after injection, mice were imaged in vivo, and then major organs were collected for ex vivo DiR imaging. Imaging data indicated that DiR signals were mainly present in the upper abdomen regions, strongest in the liver and spleen, but not detected in submandibular glands (SMGs) or cervical lymph nodes (CLNs) collected at 24 h after injection of either young or aging iEVs (Figure 1A–C). Since autoreactive cells in the spleen drive disease manifestations of SS [18], the above data suggest that iEVs might suppress the autoimmune responses in NOD.B10.H2b mice by modulating splenocytes. To determine which types of cells in the spleen take up iEVs, we labeled both young and aging iEVs with a fluorescent dye, PKH26, as in reported [19], isolated splenocytes from 4-month-old female NOD.B10.H2b mice, and treated them with PKH26-labeled EVs for 3 h. These splenocytes were then subjected to flow cytometry analysis for PKH26 and markers for macrophages (F$\frac{4}{80}$), T cells (CD3), or B cells (CD19). In both young and aging iEV groups, most PKH26+ splenocytes were F$\frac{4}{80}$+ (>$70\%$), whereas much fewer PKH26+ splenocytes were CD3+ or CD19+ (Figure 1D,E). In these different types of splenocytes, the percentage of PKH26+ was also significantly higher in F$\frac{4}{80}$+ cells than in CD3+ or CD19+ cells (Figure 1F). These data indicated that macrophages are the major population uptaking iEVs in the spleen of NOD.B10. H2b mice. ## 2.2. Young iEVs but Not Aging iEVs Promoted M2 Polarization of Splenic Macrophages The pathogenesis of SS involves the polarization of macrophages into the pro-inflammatory M1 phenotype [20], whereas EVs from tissue-derived MSCs reprogram macrophages into the anti-inflammatory M2 phenotype in mouse models of several inflammation-related diseases [21,22]. Since the spleen is essential for the pathogenesis of SS [18] and the primary target organ of IV infused iEVs, we examined the polarization of splenic macrophages from NOD.B10.H2b mice at 2 weeks after last iEV treatment with flow cytometry. In F$\frac{4}{80}$+ splenic macrophages, IV infusion of PD15 iEVs but not PD45 iEVs significantly decreased percentages of CD38+ M1 macrophages, increased percentages of CD206+ M2 macrophages, and decreased the ratio of CD38+ vs. CD206+ cells (Figure 2A–E). Consistently, qRT-PCR assays indicated that compared with the PBS group, IV infusion of PD15 iEVs significantly decreased the mRNA expression of M1 macrophage markers iNos and Alox5ap and increased M2 macrophage markers Cd206, Alox15, and IL10 in the spleen, whereas PD45 iEVs only significantly affected Alox5ap and IL10 expression among these markers to a lesser extent (Figure 2F). Moreover, only PD15 but not PD45 iEVs significantly decreased the ratio of relative mRNA levels of iNos to Cd206 (Figure 2G). These data indicate that young but not aging iEVs shifted the polarization of splenic macrophages to the anti-inflammatory M2 phenotype. ## 2.3. Young iEVs but Not Aging iEVs Decreased Th17 Cells in Spleen T helper 17 (Th17) cells are important drivers of SS in various mouse models including the NOD.B10.H2b mice [23,24,25], whereas EVs from tissue-derived MSCs suppressed development of Th17 cells in mouse models of several other autoimmune diseases [26]. In CD4+ spleen Th cells from NOD.B10.H2b mice at 2 weeks after iEV treatment, as mentioned in Section 2.2, IV infusion of PD15 iEVs but not PD45 iEVs significantly decreased percentages of IL17+ cells (Figure 3A,B). Consistently, the qRT-PCR assay indicated that compared with the PBS group, PD15 iEVs but not PD45 iEVs significantly decreased the mRNA expression of Th17 markers including IL17a, IL21, and Rorc in the spleen (Figure 3C). Regulatory T (Treg) cells are also involved in the development of SS, whereas it is not the number but rather the function of Tregs that is the driving factor, as indicated in the NOD.B10.H2b mouse model [25]. In spleens of NOD.B10.H2b mice after iEV treatment, neither PD15 nor PD45 iEVs significantly affected percentages of Foxp3+ cells in splenic CD4+ T cells or the mRNA levels of Treg markers Foxp3, IL2ra/Cd25, and Tgfb1 (Supplementary Figure S1). Since IV-infused iEVs were predominantly taken up by macrophages in the spleen, the effects of these iEVs on Th17 differentiation are likely mediated by macrophages through IL1 signaling, as reported in [27]. Consistently, PD15 iEVs but not PD45 iEVs significantly increased the mRNA level of IL1 antagonist IL1rn in the spleen (Figure 3C). These data indicate that young iEVs but not aging iEVs inhibited Th17 differentiation in the spleen, likely through promoting M2 macrophage polarization. ## 2.4. Inhibition of miR-125b in Aging iMSCs Enabled Their EVs to Repress the Onset of Sialadenitis MicroRNAs are important mediators of the immune modulatory effects of MSC EVs [28]. We identified that miR-125b is highly enriched in aging iEVs compared with young iEVs, and the inhibition of miR-125b in aging iEVs increased their activities in suppressing Th17 responses in LPS-stimulated splenocytes [16]. Therefore, here we examined whether the inhibition of miR-125b can improve their inhibitory effects on SS onset in NOD.B10.H2b mice. Aging iEVs were isolated from late-passage (PD45) iMSCs transfected with control or miR-125b inhibitors, as we reported in [16], and termed Ctrl EVs or 125KD (knockdown) EVs. The mean sizes of Ctrl EVs and 125KD EVs were comparable, whereas the level of miR-125b detectable by qPCR was remarkably lower in 125KD EVs, as expected (Figure 4A,B). Interestingly, levels of miR-21 and the TGFβ1 protein, two immune suppressive molecules enriched in young iEVs, were significantly higher in 125KD EVs than in Ctrl EVs (Figure 4C,D), indicating that the change in 125KD iEVs is not limited to the decrease in miR-125b activity. These modified aging iEVs or PBS were infused into female 4-month-old NOD.B10.H2b mice at the pre-disease stage via IV twice a week for two weeks, as we reported in [16]. Two weeks after the last injection, we collected SMGs and serum samples for the following analyses. H&E staining of SMG sections indicated that the size of leukocyte infiltrates in SMGs significantly decreased by 125KD EVs but not Ctrl EVs compared with the PBS group (Figure 4E,F). Consistently, qRT-PCR assays indicated that the mRNA levels of markers for T cells (Cd3 and Cd4), B cells (Cd20 and Ighg3), and their activation (Cd40l) in SMGs were significantly decreased by 125KD EVs, whereas only one of these markers, Ighg3, was significantly downregulated by Ctrl EVs and the downregulation was much weaker than 125KD EVs (Figure 4G). The serum level of the anti-La autoantibody was also significantly decreased by 125KD EVs but not Ctrl EVs (Figure 4H). These data indicate that the transfection of miR-125b inhibitors into aging iMSCs can improve repressive effects of their EVs on the onset of sialadenitis in the NOD.B10.H2b mouse model of primary SS. ## 2.5. Inhibition of miR-125b in Aging iEVs Restored Their Activity to Promote M2 Polarization Previous studies showed that the overexpression of miR-125b induced the proinflammatory activation of macrophages and consequent T cell activation through inhibiting the expression of Irf4 [29,30]. Since our data above indicated that splenic macrophages are the major cells up taking iEVs and young iEVs but not aging iEVs promoted M2 polarization of splenic macrophages (Figure 1 and Figure 2), we further examined the effects of miR-125b inhibition in aging iEVs on the polarization of splenic macrophages in NOD.B10.H2b mice at 2 weeks after IV infusion of Ctrl or 125KD aging iEVs. In F$\frac{4}{80}$+ splenic macrophages, 125KD but not Ctrl iEVs significantly decreased percentages of CD38+ M1 macrophages, increased percentages of CD206+ M2 macrophages, and decreased the ratio of CD38+ vs. CD206+ cells (Figure 5A–E). As indicated by qRT-PCR assay, the mRNA expression of Irf4, an miR-125b target essential for M2 macrophage polarization [29,31], was significantly increased by 125KD iEVs compared to either the PBS or Ctrl iEV groups (Figure 5F). Consistently, compared with the PBS group, 125KD iEVs significantly decreased the mRNA expression of M1 macrophage markers iNos and Alox5ap and increased M2 macrophage markers Cd206, Alox15, and IL10 in the spleen, whereas Ctrl iEVs only significantly affected Alox5ap and IL10 expression among these markers to lesser extent (Figure 5F). Moreover, only 125KD but not Ctrl iEVs significantly decreased the ratio of relative mRNA levels of iNos to Cd206 (Figure 5G). These data indicate that the transfection of miR-125b inhibitors into aging iMSCs can restore the activity of their EVs to promote M2 polarization of splenic macrophages. ## 2.6. Inhibition of miR-125b in Aging iEVs Improved Their Activity to Decrease Splenic Th17 Cells Our data above showed that young iEVs but not aging iEVs decreased splenic Th17 cells likely through M2 macrophage polarization and consequent increase in the IL1 antagonist IL1rn (Figure 3). Notably, IL1rn is a putative target of miR-125b [32]. To determine effects of miR-125b inhibition in aging iEVs on splenic Th17 cells, we examined IL17+ Th cells from NOD.B10.H2b mice at 2 weeks after IV infusion of Ctrl or 125KD aging iEVs. In splenic CD4+ Th cells, 125KD but not Ctrl iEVs significantly decreased percentages of IL17+ cells (Figure 6A,B). Consistently, qRT-PCR assays indicated that compared with the PBS group, 125KD iEVs but not Ctrl iEVs significantly decreased the mRNA expression of Th17 markers including IL17a, IL21, and Rorc but increased that of Th17 inhibitor IL1rn in the spleen (Figure 6C). These data indicate that the transfection of miR-125b inhibitors into aging iMSCs can restore the activity of their EVs to inhibit Th17 differentiation in the spleen, likely through promoting M2 macrophage polarization. ## 3. Discussion The immunomodulatory effects of MSC EVs and underlying mechanisms have been mainly studied using tissue-derived MSCs with high variations and limited expandability, and are far from conclusive [14,26,33,34]. Our data here indicated that at the pre-disease stage IV-infused young iEVs but not aging iEVs inhibit SS progression by directly promoting macrophage polarization toward the anti-inflammatory M2 phenotype and the consequent decrease in Th17 cells in spleen. To confirm that splenic macrophages are the key mediator of iEVs, it is necessary to determine whether the adoptive transfer of splenic macrophages isolated from NOD.B10.H2b mice treated with iEVs into non-treated isogenic mice can inhibit SS progression in future studies. Several risk factors for primary SS have been identified and a family history of autoimmune disease showed the highest odds ratio of 5.93 [35]. As indicated by clinical studies in Norway and Sweden, serum autoantibodies were present for a median 4~5 years in 66–$81\%$ of patients before the diagnosis of primary SS with high positive predictive values [36,37]. More recently, several salivary protein markers for preclinical SS have also been identified [38]. Based on screening with these risk factors and predictive markers, the early intervention of preclinical SS appears possible to prevent the sicca symptoms. It remains unclear whether our iEVs can restore saliva secretion after the onset of primary SS, which needs be evaluated in SS mouse models at the disease stage in our future work. Both NOD and NOD.B10.H2b mice show gender differences in the exocrine gland manifestations of SS with far greater immune pathology in salivary glands of females and lacrimal glands in males [25,39]. Our previous and current studies focused on effects of iEVs on the salivary gland pathology in female NOD and NOD.B10.H2b mice [12,16]. Further studies in male NOD.B10.H2b mice will be necessary to confirm effects of iEVs on the lacrimal gland manifestations of SS. For the future clinical application of MSC EVs, MSCs need be expanded extensively to prepare sufficient amounts of EVs. Therefore, it is necessary to establish approaches to improve SS-inhibitory effects of aging MSC EVs. As the essential mediators of immune modulatory effects of MSC EVs [28], microRNAs are highly conserved between humans and mice [40] and can mediate cross-species communication [41]. miR-125b is highly enriched in aging iEVs compared with young iEVs [16] and reported to activate proinflammatory macrophages [29,30]. Our data confirmed that the transfection of miR-125b inhibitors into aging iMSCs restored activities of their EVs to promote M2 macrophage polarization and decrease Th17 cells in the spleen in vivo. While the direct overexpression of miR-125b in CD4+ T cells inhibited Th17 cell differentiation [42], we found that IV-infused iEVs were mainly taken up by splenic macrophages but not T cells, suggesting that young iEVs and 125KD aging iEVs decreased Th17 cells in the spleen indirectly through promoting M2 macrophage polarization. Moreover, the transfection of miR-125b inhibitors into aging iMSCs also increased levels of immune-suppressive miR-21 and the TGFβ1 protein in their EVs, which likely also contributed to the restoration of the SS-inhibitory effects of aging iEVs. In conclusion, our study indicates that inhibitory effects of iEVs on SS onset are related to the increase in M2 macrophages and decrease in Th17 cells in the spleen. To maximize the production of effective iEVs from highly expanded iMSCs for future clinical application, inhibiting miR-125b in aging iMSCs appears a promising approach. ## 4.1. iMSC Culture The human iMSCs established in our laboratory [10] were plated at a density of 500 cells per cm2 of growth area in complete culture medium (CCM; αMEM medium containing $17\%$ (v/v) heat-inactivated fetal bovine serum (FBS, Atlanta Biologicals, Flowery Branch, GA, USA), penicillin-streptomycin and l-glutamine) at 37 °C and $5\%$ CO2 and passaged at 70–$80\%$ confluence. To remove EVs introduced by FBS, PD 15 or PD45 iMSCs at 70–$80\%$ confluence were incubated with a serum-free and chemically defined medium (CD-CHO Medium, Invitrogen, Carlsbad, CA, USA) supplemented with HT supplements (10 mL/L; Invitrogen), 8 mM l-glutamine (Invitrogen), d-[+]-glucose (2 g/L; Sigma, St. Louis, MO, USA), 1× nonessential amino acids (Invitrogen), and 1X MEM vitamin solution (Invitrogen). After 6 h, the medium was replaced by fresh CD-CHO medium, and the conditioned medium was recovered at 48 h to isolate iMSC EVs. ## 4.2. Isolation of iMSC-EVs and Characterization For EV isolation, the conditioned medium was filtered at 0.22 μm to remove cellular debris, and then EVs were isolated from the supernatant by ultracentrifugation at 100,000× g for 16 h at 4 °C using Sorvall WX Floor Ultra Centrifuge with AH-629 36 mL swinging Bucket Rotor (Thermo Fisher Scientific, Waltham, MA, USA). Isolated EVs were resuspended with PBS at concentrations of 5 to 10 × 1010/mL. The particle size and number of EVs were analyzed using the NanoSight LM 10 Nanoparticle Tracking Analysis System (Malvern, Malvern, UK). For in vivo biodistribution assays, iEVs were labeled with a near-infrared fluorescent dye, DiR (ThermoFisher), as reported in [17,43]. To determine types of iEV recipient cells, iEVs were labeled with a fluorescent dye, PKH26 (Sigma), as reported in [44]. Splenocytes were isolated from NOD.B10.H2b mice, cultured with RMPI 1640 culture medium (Gibco, Billings, MT, USA) containing $5\%$ FBS, treated with 3 × 109 particles/mL PKH26-labeled iEVs, and then examined with flow cytometry for PKH26 signal and markers of macrophages, T cells, or B cells as detailed below. ## 4.3. Animal Studies All animal studies were approved by the Texas A&M University (TAMU) Institutional Animal Care and Use Committee (IACUC). NOD.B10.H2b mice were purchased from the Jackson Laboratory and kept in the specific pathogen-free environment maintained by TAMU Comparative Medicine Program following the NIH Guide with following room conditions: a 12 h light/12 h dark cycle, temperatures of 65–75 °F (~18–23 °C), and 40–$60\%$ humidity. Four-month-old female mice were randomly grouped for all treatments. For biodistribution assay, DiR-labeled EVs (1.5 × 1010 particles in 100 μL PBS) derived from young (PD15) or aging (PD45) iMSCs were injected into the tail vein. Mice were imaged with IVIS imaging system (PerkinElmer, Hopkinton, MA, USA) at 1, 3, and 24 h after iEV injection. For testing SS-inhibitory effects and mechanisms, PBS (100 μL) or EVs (1.5 × 1010 particles in 100 μL PBS) derived from PD15 or PD45 iMSCs were injected into the tail vein twice a week for two weeks, as we reported recently in [16]. Two weeks after the last injection, submandibular glands (SMGs) and serum were collected. The focus scores, numbers of inflammatory infiltrates of at least 50 cells per 4 mm2 area, were quantified from 3 H&E-stained non-consecutive sections cutting at 200 µm intervals, as recommended [45] from each of 5 SMGs per group. ## 4.4. Flow Cytometry Spleens from NOD.B10.H2b mice were minced and digested for 1 h with RPMI 1640 medium containing 1 mg/mL collagenase IV, 5 mM CaCl2, 50 mg DNase I, and $8\%$ fetal bovine serum with continuous shaking at room temperature to prepare single cells. These cells were stained with fluorescent-labeled antibodies against F$\frac{4}{80}$ (BD Pharmingen 123116, 123107, San Diego, CA, USA), CD3 (BioLegend 100236, San Diego, CA, USA), CD19 (BioLegend 152409), CD38 (BioLegend 102707), CD206 (BioLegend 141706), CD4 (BioLegend 100408), IL17a (130-112-009, Miltenyi Biotec, San Diego, CA, USA), or the corresponding isotype controls (BioLegend 400608, 400511, 400612, 400207, 1:100). Dead cells were excluded using a LIVE/DEAD Fixable Aqua Dead Cell Stain Kit (Invitrogen, Waltham, MA, USA). The stained cells were analyzed on a CytoFlex flow cytometer (Beckman Coulter, Brea, CA, USA). Data were analyzed using the FlowJo software (Version 10.8.1, FlowJo, Ashland, OR, USA). ## 4.5. Real-Time PCR Analysis of mRNAs and miRNAs RNAs were extracted from mouse SMGs with RNeasy Mini Kit (Qiagen, Germantown, MD, USA) and reverse transcribed with High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA, USA). The qPCR was performed with SYBR Green Master Mix (Bio-Rad, Hercules, CA, USA) on a CFX Connect PCR System (Bio-Rad). The primers were synthesized by Invitrogen with sequences retrieved from Primerbank (http://pga.mgh.harvard.edu/primerbank, accessed on 1 June 2021). The qPCR data were analyzed with Gapdh as the reference gene. Total RNA was isolated from ultracentrifuged EVs (1 × 1011 particles) with the EZNA Total RNA Kit (Omega Bio-Tek, Doraville, CA, USA). Levels of miR-21 and miR-125b were measured with corresponding TaqMan MicroRNA Assay kits and normalized by miR-143 that was consistently expressed in young and aging EVs [16]. $$n = 3$$ for PCR analyses of iEVs and 5 for mouse samples. ## 4.6. Cell Transfection Cells with $60\%$ confluence were transfected with 20 nM miRNA inhibitors or mimics for control or miR-125b (Invitrogen) using Lipofectamine RNAiMAX (Invitrogen) for 5 h. After transfection, cells were recovered with antibiotic-free CCM overnight for collecting EVs. ## 4.7. ELISA Human TGFβ1 in iEVs (1 × 1010 particle/mL) and Anti-La in serum from NOD.B10.H2b mice were measured by commercial ELISA kits (R&D Systems, Minneapolis, MN, USA; Signosis, Santa Clara, CA, USA) according to the manufacturer’s protocol. ## 4.8. Statistics Column data with one grouping variable were analyzed using one-way ANOVA, and grouped data with two grouping variables were analyzed using two-way ANOVA, both followed by Tukey’s multiple comparison tests. Statistical analysis and graphical generation of data were conducted with GraphPad Prism 9 software Version 9.4.1 (San Diego, CA, USA). ## References 1. Saraux A., Pers J.O., Devauchelle-Pensec V.. Treatment of primary Sjogren syndrome. Nat. Rev. Rheumatol. (2016) 12 456-471. DOI: 10.1038/nrrheum.2016.100 2. Brito-Zeron P., Baldini C., Bootsma H., Bowman S.J., Jonsson R., Mariette X., Sivils K., Theander E., Tzioufas A., Ramos-Casals M.. Sjogren syndrome. Nat. Rev. Dis. Prim. (2016) 2 16047. DOI: 10.1038/nrdp.2016.47 3. Vitali C., Minniti A., Pignataro F., Maglione W., Del Papa N.. Management of Sjogren’s Syndrome: Present Issues and Future Perspectives. Front. Med. (2021) 8 676885. DOI: 10.3389/fmed.2021.676885 4. Harrell C.R., Fellabaum C., Jovicic N., Djonov V., Arsenijevic N., Volarevic V.. Molecular Mechanisms Responsible for Therapeutic Potential of Mesenchymal Stem Cell-Derived Secretome. Cells (2019) 8. DOI: 10.3390/cells8050467 5. Xu J., Wang D., Liu D., Fan Z., Zhang H., Liu O., Ding G., Gao R., Zhang C., Ding Y.. Allogeneic mesenchymal stem cell treatment alleviates experimental and clinical Sjogren syndrome. Blood (2012) 120 3142-3151. DOI: 10.1182/blood-2011-11-391144 6. Yao G., Qi J., Liang J., Shi B., Chen W., Li W., Tang X., Wang D., Lu L., Chen W.. Mesenchymal stem cell transplantation alleviates experimental Sjogren’s syndrome through IFN-beta/IL-27 signaling axis. Theranostics (2019) 9 8253-8265. DOI: 10.7150/thno.37351 7. Prockop D.J.. The exciting prospects of new therapies with mesenchymal stromal cells. Cytotherapy (2017) 19 1-8. DOI: 10.1016/j.jcyt.2016.09.008 8. Lee R.H., Yu J.M., Foskett A.M., Peltier G., Reneau J.C., Bazhanov N., Oh J.Y., Prockop D.J.. TSG-6 as a biomarker to predict efficacy of human mesenchymal stem/progenitor cells (hMSCs) in modulating sterile inflammation in vivo. Proc. Natl. Acad. Sci. USA (2014) 111 16766-16771. DOI: 10.1073/pnas.1416121111 9. Galipeau J., Sensebe L.. Mesenchymal Stromal Cells: Clinical Challenges and Therapeutic Opportunities. Cell Stem Cell (2018) 22 824-833. DOI: 10.1016/j.stem.2018.05.004 10. Zhao Q., Gregory C.A., Lee R.H., Reger R.L., Qin L., Hai B., Park M.S., Yoon N., Clough B., McNeill E.. MSCs derived from iPSCs with a modified protocol are tumor-tropic but have much less potential to promote tumors than bone marrow MSCs. Proc. Natl. Acad. Sci. USA (2015) 112 530-535. DOI: 10.1073/pnas.1423008112 11. Yun Y.I., Park S.Y., Lee H.J., Ko J.H., Kim M.K., Wee W.R., Reger R.L., Gregory C.A., Choi H., Fulcher S.F.. Comparison of the anti-inflammatory effects of induced pluripotent stem cell-derived and bone marrow-derived mesenchymal stromal cells in a murine model of corneal injury. Cytotherapy (2017) 19 28-35. DOI: 10.1016/j.jcyt.2016.10.007 12. Hai B., Shigemoto-Kuroda T., Zhao Q., Lee R.H., Liu F.. Inhibitory Effects of iPSC-MSCs and Their Extracellular Vesicles on the Onset of Sialadenitis in a Mouse Model of Sjogren’s Syndrome. Stem Cells Int. (2018) 2018 2092315. DOI: 10.1155/2018/2092315 13. McNeill E.P., Zeitouni S., Pan S., Haskell A., Cesarek M., Tahan D., Clough B.H., Krause U., Dobson L.K., Garcia M.. Characterization of a pluripotent stem cell-derived matrix with powerful osteoregenerative capabilities. Nat. Commun. (2020) 11 3025. DOI: 10.1038/s41467-020-16646-2 14. Phinney D.G., Pittenger M.F.. Concise Review: MSC-Derived Exosomes for Cell-Free Therapy. Stem Cells (2017) 35 851-858. DOI: 10.1002/stem.2575 15. Zhao A.G., Shah K., Cromer B., Sumer H.. Mesenchymal Stem Cell-Derived Extracellular Vesicles and Their Therapeutic Potential. Stem Cells Int. (2020) 2020 8825771. DOI: 10.1155/2020/8825771 16. Kim H., Zhao Q., Barreda H., Kaur G., Hai B., Choi J.M., Jung S.Y., Liu F., Lee R.H.. Identification of Molecules Responsible for Therapeutic Effects of Extracellular Vesicles Produced from iPSC-Derived MSCs on Sjo gren’s Syndrome. Aging Dis. (2021) 12 1409-1422. DOI: 10.14336/AD.2021.0621 17. Wiklander O.P., Nordin J.Z., O’Loughlin A., Gustafsson Y., Corso G., Mager I., Vader P., Lee Y., Sork H., Seow Y.. Extracellular vesicle in vivo biodistribution is determined by cell source, route of administration and targeting. J. Extracell. Vesicles (2015) 4 26316. DOI: 10.3402/jev.v4.26316 18. Mahmoud T.I., Wang J., Karnell J.L., Wang Q., Wang S., Naiman B., Gross P., Brohawn P.Z., Morehouse C., Aoyama J.. Autoimmune manifestations in aged mice arise from early-life immune dysregulation. Sci. Transl. Med. (2016) 8 361ra137. DOI: 10.1126/scitranslmed.aag0367 19. Long Q., Upadhya D., Hattiangady B., Kim D.K., An S.Y., Shuai B., Prockop D.J., Shetty A.K.. Intranasal MSC-derived A1-exosomes ease inflammation, and prevent abnormal neurogenesis and memory dysfunction after status epilepticus. Proc. Natl. Acad. Sci. USA (2017) 114 E3536-E3545. DOI: 10.1073/pnas.1703920114 20. Gerasimova E.V., Popkova T.V., Gerasimova D.A., Kirichenko T.V.. Macrophage Dysfunction in Autoimmune Rheumatic Diseases and Atherosclerosis. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23094513 21. Wang C., Borger V., Sardari M., Murke F., Skuljec J., Pul R., Hagemann N., Dzyubenko E., Dittrich R., Gregorius J.. Mesenchymal Stromal Cell-Derived Small Extracellular Vesicles Induce Ischemic Neuroprotection by Modulating Leukocytes and Specifically Neutrophils. Stroke (2020) 51 1825-1834. DOI: 10.1161/STROKEAHA.119.028012 22. Tian J., Chen W., Xiong Y., Li Q., Kong S., Li M., Pang C., Qiu Y., Xu Z., Gong Q.. Small extracellular vesicles derived from hypoxic preconditioned dental pulp stem cells ameliorate inflammatory osteolysis by modulating macrophage polarization and osteoclastogenesis. Bioact. Mater. (2023) 22 326-342. DOI: 10.1016/j.bioactmat.2022.10.001 23. Lin X., Rui K., Deng J., Tian J., Wang X., Wang S., Ko K.H., Jiao Z., Chan V.S., Lau C.S.. Th17 cells play a critical role in the development of experimental Sjogren’s syndrome. Ann. Rheum. Dis. (2015) 74 1302-1310. DOI: 10.1136/annrheumdis-2013-204584 24. Verstappen G.M., Corneth O.B.J., Bootsma H., Kroese F.G.M.. Th17 cells in primary Sjogren’s syndrome: Pathogenicity and plasticity. J. Autoimmun. (2018) 87 16-25. DOI: 10.1016/j.jaut.2017.11.003 25. Coursey T.G., Bian F., Zaheer M., Pflugfelder S.C., Volpe E.A., de Paiva C.S.. Age-related spontaneous lacrimal keratoconjunctivitis is accompanied by dysfunctional T regulatory cells. Mucosal Immunol. (2017) 10 743-756. DOI: 10.1038/mi.2016.83 26. Shigemoto-Kuroda T., Oh J.Y., Kim D.K., Jeong H.J., Park S.Y., Lee H.J., Park J.W., Kim T.W., An S.Y., Prockop D.J.. MSC-derived Extracellular Vesicles Attenuate Immune Responses in Two Autoimmune Murine Models: Type 1 Diabetes and Uveoretinitis. Stem Cell Rep. (2017) 8 1214-1225. DOI: 10.1016/j.stemcr.2017.04.008 27. Arnold C.E., Gordon P., Barker R.N., Wilson H.M.. The activation status of human macrophages presenting antigen determines the efficiency of Th17 responses. Immunobiology (2015) 220 10-19. DOI: 10.1016/j.imbio.2014.09.022 28. Dabrowska S., Andrzejewska A., Janowski M., Lukomska B.. Immunomodulatory and Regenerative Effects of Mesenchymal Stem Cells and Extracellular Vesicles: Therapeutic Outlook for Inflammatory and Degenerative Diseases. Front. Immunol. (2020) 11 591065. DOI: 10.3389/fimmu.2020.591065 29. Chaudhuri A.A., So A.Y., Sinha N., Gibson W.S., Taganov K.D., O’Connell R.M., Baltimore D.. MicroRNA-125b potentiates macrophage activation. J. Immunol. (2011) 187 5062-5068. DOI: 10.4049/jimmunol.1102001 30. Parayath N.N., Gandham S.K., Leslie F., Amiji M.M.. Improved anti-tumor efficacy of paclitaxel in combination with MicroRNA-125b-based tumor-associated macrophage repolarization in epithelial ovarian cancer. Cancer Lett. (2019) 461 1-9. DOI: 10.1016/j.canlet.2019.07.002 31. Satoh T., Takeuchi O., Vandenbon A., Yasuda K., Tanaka Y., Kumagai Y., Miyake T., Matsushita K., Okazaki T., Saitoh T.. The Jmjd3-Irf4 axis regulates M2 macrophage polarization and host responses against helminth infection. Nat. Immunol. (2010) 11 936-944. DOI: 10.1038/ni.1920 32. Modi B.P., Washington S., Walsh S.W., Jackson-Cook C., Archer K.J., Strauss J.F.. Expression patterns of the chromosome 21 MicroRNA cluster (miR-99a, miR-125b and let-7c) in chorioamniotic membranes. Placenta (2017) 49 1-9. DOI: 10.1016/j.placenta.2016.11.002 33. Favaro E., Carpanetto A., Caorsi C., Giovarelli M., Angelini C., Cavallo-Perin P., Tetta C., Camussi G., Zanone M.M.. Human mesenchymal stem cells and derived extracellular vesicles induce regulatory dendritic cells in type 1 diabetic patients. Diabetologia (2016) 59 325-333. DOI: 10.1007/s00125-015-3808-0 34. Wen S., Dooner M., Cheng Y., Papa E., Del Tatto M., Pereira M., Deng Y., Goldberg L., Aliotta J., Chatterjee D.. Mesenchymal stromal cell-derived extracellular vesicles rescue radiation damage to murine marrow hematopoietic cells. Leukemia (2016) 30 2221-2231. DOI: 10.1038/leu.2016.107 35. Jin L., Dai M., Li C., Wang J., Wu B.. Risk factors for primary Sjogren’s Syndrome: A systematic review and meta-analysis. Clin. Rheumatol. (2023) 42 327-338. DOI: 10.1007/s10067-022-06474-8 36. Jonsson R., Theander E., Sjostrom B., Brokstad K., Henriksson G.. Autoantibodies present before symptom onset in primary Sjogren syndrome. JAMA (2013) 310 1854-1855. DOI: 10.1001/jama.2013.278448 37. Theander E., Jonsson R., Sjostrom B., Brokstad K., Olsson P., Henriksson G.. Prediction of Sjogren’s Syndrome Years Before Diagnosis and Identification of Patients With Early Onset and Severe Disease Course by Autoantibody Profiling. Arthritis Rheumatol. (2015) 67 2427-2436. DOI: 10.1002/art.39214 38. Di Giorgi N., Cecchettini A., Michelucci E., Signore G., Ceccherini E., Ferro F., Elefante E., Tani C., Baldini C., Rocchiccioli S.. Salivary Proteomics Markers for Preclinical Sjogren’s Syndrome: A Pilot Study. Biomolecules (2022) 12. DOI: 10.3390/biom12060738 39. Toda I., Sullivan B.D., Rocha E.M., Da Silveira L.A., Wickham L.A., Sullivan D.A.. Impact of gender on exocrine gland inflammation in mouse models of Sjogren’s syndrome. Exp. Eye Res. (1999) 69 355-366. DOI: 10.1006/exer.1999.0715 40. Lagos-Quintana M., Rauhut R., Meyer J., Borkhardt A., Tuschl T.. New microRNAs from mouse and human. RNA (2003) 9 175-179. DOI: 10.1261/rna.2146903 41. Liang H., Zen K., Zhang J., Zhang C.Y., Chen X.. New roles for microRNAs in cross-species communication. RNA Biol. (2013) 10 367-370. DOI: 10.4161/rna.23663 42. Fan Z.D., Cao Q., Huang N., Ma L., Ma H.H., Zhang Y.Y., Yu H.G., Zhou G.P.. MicroRNA-125b regulates Th17/Treg cell differentiation and is associated with juvenile idiopathic arthritis. World J. Pediatr. (2020) 16 99-110. DOI: 10.1007/s12519-019-00265-z 43. Zhao Q., Hai B., Zhang X., Xu J., Koehler B., Liu F.. Biomimetic nanovesicles made from iPS cell-derived mesenchymal stem cells for targeted therapy of triple-negative breast cancer. Nanomed. Nanotechnol. Biol. Med. (2020) 24 102146. DOI: 10.1016/j.nano.2019.102146 44. Franzen C.A., Simms P.E., Van Huis A.F., Foreman K.E., Kuo P.C., Gupta G.N.. Characterization of uptake and internalization of exosomes by bladder cancer cells. Biomed Res. Int. (2014) 2014 619829. DOI: 10.1155/2014/619829 45. Fisher B.A., Jonsson R., Daniels T., Bombardieri M., Brown R.M., Morgan P., Bombardieri S., Ng W.F., Tzioufas A.G., Vitali C.. Standardisation of labial salivary gland histopathology in clinical trials in primary Sjogren’s syndrome. Ann. Rheum. Dis. (2017) 76 1161-1168. DOI: 10.1136/annrheumdis-2016-210448
--- title: Could Vaspin Be a Potential Diagnostic Marker in Endometrial Cancer? authors: - Dominika Pietrzyk - Piotr Tkacz - Mateusz Kozłowski - Sebastian Kwiatkowski - Małgorzata Rychlicka - Ewa Pius-Sadowska - Bogusław Machaliński - Aneta Cymbaluk-Płoska journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049014 doi: 10.3390/ijerph20064999 license: CC BY 4.0 --- # Could Vaspin Be a Potential Diagnostic Marker in Endometrial Cancer? ## Abstract Obesity and being overweight are risk factors for many types of cancer, including endometrial cancer. Adipose tissue is thought to be an endocrine organ that produces various hormones, including one known as vaspin. Insulin resistance, metabolic syndrome and type 2 diabetes are all associated with higher vaspin levels. A total of 127 patients divided into study (endometrial cancer) and control groups (non-cancerous) participated in this research. Serum vaspin levels were measured for all patients. The analysis was performed while taking into account grading and staging. In order to assess the usefulness of the tested protein as a new diagnostic marker, we used the plotting of a curve (ROC) and the calculation of the AUC curve to characterize the sensitivity and specificity of the parameters tested. We concluded that there were significantly lower vaspin levels in patients with endometrial cancer compared to patients with benign endometrial lesions. Vaspin may be a useful diagnostic marker in separating benign lesions from endometrial cancer. ## 1. Introduction Endometrial cancer (EC) is the sixth most common malignancy for women worldwide [1]. Its incidence is increasing, particularly in postmenopausal women. Only $4\%$ of patients are under the age of 40 [2]. More than half of endometrial cancer cases are attributed to obesity, which has been identified as an independent risk factor for the disease [3]. Research confirms the link between obesity, hyperinsulinemia, type 2 diabetes and endometrial cancer [4,5,6,7]. The increase in cellular reactivity to insulin is connected with the activation of the MAPK/PI3K/AKT/mTOR signaling pathway, which is specific to cancer (EC) [8]. Often, this pathomechanism is further enhanced by the loss of the PTEN suppressor gene, which normally acts in opposition to the PI3K/AKT/mTOR pathway [9]. Bioavailable estrogens, especially when unopposed by progesterone, may increase the risk of EC through mitogenic effects in endometrial tissue [10]. Type I endometrial tumors usually express high levels of estrogen receptor (ER), and they are thought to be hormonally driven as opposed to type II endometrial cancer [11]. A low level of SHBG (sex hormone binding globulin) is induced by high body weight. Observations suggest a negative correlation between circulating SHBG levels and insulin resistance (IR). Decreased SHBG levels increase the bioavailability of androgens, which in turn leads to the progression of ovarian pathologies such as, among others, polycystic ovarian syndrome (PCOS) [12]. In addition, the effect of hormonal pathomechanisms can be a state of hyperprolactinemia, which further stimulates adrenal androgen production. Similarly, in the course of negative feedback in hypothyroidism, there is a stimulation of TSH (thyroid-stimulating hormone), which is also responsible for the state of hyperprolactinemia and affects proper sex hormone management [13,14]. As a result, the patient develops a state of relative hyperestrogenism. Therefore, obesity correlates with an increased risk of endometrial hyperplasia and, ultimately, EC [15]. Adipose tissue is seen as an endocrine organ, synthesizing so-called adipocytokines such as, among others, vaspin, which belongs to the serine protease inhibitor family [16,17]. After menopause, adipose tissue becomes the main location of estrogen synthesis and the source of aromatase, the enzyme responsible for converting androgens to estrogens. After binding to their receptors, estrogens can indirectly affect the transcription of such known proliferative factors as IGF1R and IGF1 [18]. It acts directly by stimulating endometrial proliferation through the MAPK and AKT signaling pathways. The use of single-ingredient hormonal contraception and hormone replacement therapy based solely on estradiol, although rarely practiced these days, increases the risk of the aforementioned pathomechanisms. A drug that can increase proliferation and, thereby, the risk of abnormal lesions is tamoxifen, which is used in the treatment of breast cancer [19]. It is well-established that a state of hyperestrogenism unbalanced by progesterone contributes to a significantly higher risk of endometrial cancer type 1 and its precursors [8,20]. By understanding the mechanisms of estrogens and progestogens in the endometrium, their undeniable proliferative and antiproliferative effects can be noted. Vaspin is an adipokine found in many tissues. Unlike the vast majority of cytokines, it has anti-inflammatory and antiproliferative effects by inhibiting inflammatory mediators such as NF-κB. Moreover, vaspin inhibits insulin degradation, thereby improving glucose tolerance. At the same time, it also has the ability to inhibit IRS-2 (Insulin Receptor Substrate 2) phosphorylation, a protective mechanism against the onset of tissue hyperinsulinemia [21]. Increased vaspin expression is observed in patients with type II diabetes, obesity and metabolic syndrome [7]. As low-grade inflammation and insulin resistance play an important role in the pathogenesis of endometrial cancer, we wondered whether there is a link between the occurrence of endometrial cancer and vaspin concentrations and if it is possible to use vaspin as a diagnostic marker in endometrial cancer. Despite ongoing research, no useful marker for endometrial cancer has yet been found. The purpose of this study was to determine the utility of vaspin in the diagnosis of endometrial cancer. In addition, we also investigated whether vaspin is useful in distinguishing grades and stages of endometrial cancer. ## 2.1. Participation in the Study A total of 127 patients with abnormal uterine bleeding/abnormal ultrasound images from the Department of Gynecological Surgery and Gynecological Oncology at the Pomeranian Medical University in Szczecin, Poland were included in the study. Lack of patient consent, endometrial hyperplasia diagnosed histopathologically, acute inflammation, other cancers, collagenosis, chronic kidney disease, cirrhosis, therapy with biological agents and immunotherapy were among the exclusion criteria. The material for the study was collected over a period of 24 months. The Pomeranian Medical University’s Ethical Committee approved the study (approvement no. ( KB-$\frac{0012}{148}$/2020), and each participating patient signed an informed consent form in order to take part. ## 2.2. Classification of Patients into Study and Control Groups Patients were divided into two groups according to the histological diagnosis obtained by endometrial biopsy, curettage or hysteroscopy. Group A, consisting of 62 patients who had benign endometrium lesions, was separated into two subgroups: A1 (endometrial polyps, $$n = 30$$) and A2 (uterine myomas, $$n = 32$$). Group B included 65 endometrial cancer patients. The group is described in Table 1. ## 2.3. Preparation of Pre-Laboratory Samples All patients’ serum vaspin levels were assessed before surgical treatment. Following surgical intervention in the study group, analysis was carried out while accounting for the tumor’s histopathological differentiation (grading) and clinical stage (staging). The patients were split into two groups following a histopathological examination. The characteristics of the group, taking into account grading and staging, are shown in Table 2. ## 2.4. Laboratory Analysis Following centrifugation of the blood samples and freezing of the resultant serum in Eppendorf-style containers maintained at −80 °C, biochemical analyses were carried out. Vaspin concentrations were measured in serum by using an immunoenzymatic ELISA-multiplex fluorescence assay (Luminex Corporation, Austin, TX, USA) and utilizing a commercial Bio Plex Pro RBM Human Metabolic Panel 2 (Biorad, Hercules, CA, USA). ## 2.5. Statistical Analysis The statistical evaluations were performed using Statistica version 10 PL software. The Shapiro–Wilk test was used to determine whether the study’s variables have a normal distribution. With the exception of the population-wide variable age, which has a normal distribution, none of the other variables have normal distributions. As a result, non-parametric methods (Spearman’s rank correlation coefficient) and non-parametric significance tests were used in the analysis to test their relationships. The assumption that the distributions of the two variables are representative of the same populations was confirmed using the Mann–Whitney U test of significance for independent samples. A non-parametric Kruskal–Wallis significance test was used to corroborate the hypothesis whether samples originated from the same distribution. The populations that varied were examined using post hoc tests. We used the plotting of a curve (ROC) and the calculation of the area under the curve (AUC) to characterize the sensitivity and specificity of the parameters tested in order to determine the usefulness of the tested protein as a new diagnostic marker. A value of $p \leq 0.05$ was considered an indicator of statistical significance. ## 3.1. Characteristics of the Study Group There were statistically significant differences between the rates of pre- and postmenopausal women in groups of patients with endometrial cancer and benign lesions. The group of patients with diagnosed EC included 13 premenopausal and 52 postmenopausal patients, whereas the control group contained 27 premenopausal and 35 postmenopausal patients. In addition, we divided the groups according to the presence of endometrial cancer risk factors such as hypertension, body mass and type II diabetes. The smallest group of EC patients was the normal-weight group, which was nearly equal in population to the overweight patient group ($$p \leq 0.031$$). Statistically significant differences were found among female patients with benign changes with BMI 18.9–24.9 (normal weight) compared to BMI 25–29.9 (pre-obesity). In the study group, the numbers of pre-obese and normal-weight patients were 25 and 14, respectively. In the control group, there were 24 pre-obese and 22 normal-weight patients. The results are described in Table 3. ## 3.2. Evaluation of Serum Vaspin Levels in Relation to Histopathological Diagnosis Median vaspin concentrations were significantly lower in patients with endometrial cancer in the study group compared to the median vaspin serum concentrations in the control group ($$p \leq 0.001$$). Statistically significant differences were observed in patients with EC compared to patients with endometrial polyps ($$p \leq 0.016$$) and uterine myomas ($$p \leq 0.028$$). However, statistically significant differences were not revealed between median concentrations of serum vaspin in the group of patients with endometrial polyps vs. patients with uterine myomas. The results are presented in Table 4 and Figure 1. ## 3.3. Assessing the Relationships of Serum Vaspin Levels in Patients with Endometrial Cancer at the Time of Collection Statistically significant differences were not revealed between the median concentration of vaspin in the group of patients at the time of diagnosis and 6–8 weeks later before surgical treatment. The results are presented in Table 5. ## 3.4. Evaluation of Vaspin as a New Diagnostic Marker— ROC Curve Analysis for Vaspin Protein Relative to Study and Control Group In order to evaluate the diagnostic values of vaspin, ROC curves were plotted, and the areas under the ROC curves (AUC) were calculated. For patients with endometrial cancer and benign endometrial lesions, the AUC was 0.88 (see Figure 2 and Table 6 below). The conclusion can be drawn that checking serum levels of vaspin before surgery can be a good diagnostic test to differentiate benign lesions from endometrial cancers. The data obtained from the appearance of ROC curves according to the hormonal status of patients were different. The AUC for premenopausal patients was 0.76, and for postmenopausal patients, it was 0.92. ## 3.5. Evaluation of Vaspin Protein as a Differential Test Relative to Grading—ROC Curve Analysis for Vaspin Protein Compared to Histopathological Differentiation Regarding the area under the AUC curve (0.38), we found that it was less than 0.5. Therefore, preoperative serum concentrations of vaspin cannot be considered for diagnostic use in the differential grading of endometrial cancer. The results are presented in Figure 3 and Table 7. ## 3.6. Evaluation of Vaspin Protein as a Differential Test Relative to Staging—ROC Curve Analysis for Vaspin Protein Depending on the Clinical Stage Regarding the area under the AUC curve (0.51), we found that it was greater than 0.5. Therefore, preoperative serum concentrations of vaspin can be considered for diagnostic use in the differential staging of endometrial cancer. The results are presented in Figure 4 and Table 8. ## 3.7. Evaluation of the Sensitivity and Specificity of Vaspin as a Diagnostic Factor in Endometrial Cancer Table 9 presents the percentages of sensitivity and specificity for vaspin as a diagnostic factor in the entire study group of patients, with distinction between the subgroups of premenopausal and menopausal patients. We found that sensitivity and specificity in the entire group were $86\%$ and $78\%$, respectively. For premenopausal patients, the results were $81\%$ sensitivity and $66\%$ specificity, and they were $88\%$ sensitivity and $72\%$ specificity for menopausal patients. ## 4. Discussion Overweight or obese people represent $60\%$ of the European population [22]. Obesity predisposes people to a number of metabolic disorders, such as insulin resistance, type 2 diabetes, hypertension and dyslipidemias. It is also a significant risk factor for cardiovascular disease and a number of cancers, particularly endometrial, ovarian, breast, pancreatic and colorectal cancers [23]. In our study, $61.9\%$ of patients with endometrial cancer were noted to be obese. In the control group, the percentage of obese patients was much lower at $38.1\%$. However, the increase in these risks is not linearly related to weight gain [24]. The presence of hyperestrogenism, inflammation and insulin resistance due to obesity with the associated metabolic syndromes increases the risk of oncogenesis [25]. Free fatty acids (FFAs), which are increased in obese individuals, stimulate TLR4 receptors, inducing the expression of metabolic pathways that promote inflammation [26]. Furthermore, studies show that inflammatory factors may induce IR (insulin resistance) through the NF-κB interaction [27]. A meta-analysis conducted by the Agency for Research on Cancer (IARC) in 2016 clearly found that the risk of endometrial cancer increases with BMI. The patients’ BMI-dependent relative risks were approximately 1.5 for pre-obese patients (BMI 25–29.9) compared to patients with diagnosed class III obesity (BMI > 40), for whom the relative risk (RR) was 6.25 [28]. Adipokines are polypeptide cytokines produced by the adipose tissue, and they are especially important in obesity-related cancers. Most of them have pro-inflammatory properties and are increased in cancers [29]. However, Li et al., point to the anti-inflammatory properties of one of the adipokines, called vaspin [30]. In our study, we focused our attention on vaspin, a part of the serpin (serine protease inhibitor) family, also called serpin A12 [31]. Pich et al. in a 2021 study reported its presence in many glands, such as the hypothalamus, pancreas, thyroid gland, ovaries, placenta and testes. Vaspin levels were proven to be elevated in type 2 diabetes, metabolic syndrome, obesity, coronary artery disease and insulin resistance [32,33]. In addition, it blocks the activation of NF-κB in endothelial and pancreatic cells, preventing the development of inflammation. By inhibiting Kallikrein 7, vaspin blocks insulin degradation, resulting in reduced insulin resistance and improved glucose tolerance [34]. In our study, we observed statistically significant differences in serum levels between patients with BMI levels indicating pre-obesity or obesity. As mentioned previously, insulin resistance and chronic low-grade inflammation induced by obesity are important risk factors for endometrial cancer. The multiple beneficial effects of vaspin as a factor in minimizing these conditions led us to investigate its use as a diagnostic marker for endometrial malignancies. The incidence of hypothyroidism in patients with endometrial cancer is significantly elevated, and pretreatment serum TSH (thyroid-stimulating hormone) levels are an independent risk factor for EC. Based on a review of available experimental and clinical data, hypothyroidism is closely associated with many EC risk factors, including metabolic syndrome, PCOS and infertility [35]. Seebacher et al. concluded that TSH levels were correlated with dyslipidemia [36,37]. Considering carbohydrate metabolism, our study shows that the concentrations of serum vaspin were higher in patients with diabetes, but the result was not statistically significant. Although the effect of vaspin in reducing insulin resistance is known, there are few reports that relate serum concentrations of vaspin in patients with heart disease, including those with hypertension. In our study, we found no differences in vaspin concentrations according to the presence of arterial hypertension. Other studies showed that vaspin can inhibit vascular endothelial apoptosis and cause weakness in blood vessels stimulated by high glucose levels [38,39]. Körner et al. achieved results different from our study. They demonstrated that serum vaspin levels correlated with blood pressure and may be associated with vascular endothelial damage [40]. In contrast to patients with benign endometrial lesions, patients with endometrial cancer have considerably decreased vaspin levels, according to our study. Significant reductions of vaspin levels in patients with endometrial cancer relative to patients with benign lesions were also found in earlier studies [21,41]. We found no statistically significant differences in vaspin concentrations between patients with endometrial polyps and uterine myomas. We also examined the dependence of vaspin concentrations on the time of sampling. There were no statistically significant differences in vaspin concentrations taken at the beginning of the diagnostic process and 6–8 weeks later before surgical treatment. To investigate the usefulness of vaspin as a new preoperative diagnostic marker, we used the ROC curve and calculated the AUC, which was 0.88. We also evaluated the use of vaspin as a differentiating factor for grading and staging endometrial cancer, and we obtained AUCs of 0.38 and 0.51, respectively. There is a connection between pathologies linked to hormonal imbalance and significantly increased vaspin in obese patients. Insulin resistance was directly correlated with vaspin mRNA expression but not with its circulating levels [42]. As a result, we speculated that elevated vaspin levels could be a compensatory mechanism in the beginnings of the early stages of insulin resistance. In addition, Tan et al. found that vaspin synthesis is stimulated by glucose in omental adipocytes [43]. According to epidemiological and clinical evidence, the development of EC is significantly influenced by insulin resistance and the accompanying hyperinsulinemia. It was also proven that the risk of developing EC rises quite quickly after the diagnosis of IR and diabetes, or nearly 6 months after their detection [44]. The effects of vaspin also extend to other physiological processes, including food intake and inflammation [45]. Undoubtedly, obesity is a significant factor that contributes to the progress of endometrial cancer [46]. Vaspin appears to be a promising indicator protein that can be used in patients with abnormal bleeding. More research on vaspin as a diagnostic factor in endometrial cancer remains to be considered. ## 5. Conclusions Vaspin may be a potential diagnostic marker to be used to differentiate endometrial cancer from benign lesions. It is not possible to distinguish grades of endometrial cancer using the tissue expression of vaspin. However, vaspin can be useful for distinguishing clinical stages of endometrial cancer. ## References 1. Sung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F.. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. (2021) 71 209-249. DOI: 10.3322/caac.21660 2. Giuseppe T., Claudiana O., Raffaele T., Gianluca R.D., Antonio P., Ettore C.. Conservative treatment in early stage endometrial cancer: A review. Acta Biomed. (2019) 90 405-410. DOI: 10.23750/abm.v90i4.7800 3. Onstad M.A., Schmandt R.E., Lu K.H.. Addressing the Role of Obesity in Endometrial Cancer Risk, Prevention, and Treatment. J. Clin. Oncol. (2016) 34 4225-4230. DOI: 10.1200/JCO.2016.69.4638 4. Vrachnis N., Iavazzo C., Iliodromiti Z., Sifakis S., Alexandrou A., Siristatidis C., Grigoriadis C., Botsis D., Creatsas G.. Diabetes mellitus and gynecologic cancer: Molecular mechanisms, epidemiological, clinical and prognostic perspectives. Arch. Gynecol. Obstet. (2016) 293 239-246. DOI: 10.1007/s00404-015-3858-z 5. Friberg E., Orsini N., Mantzoros C.S., Wolk A.. Diabetes mellitus and risk of endometrial cancer: A meta-analysis. Diabetologia (2007) 50 1365-1374. DOI: 10.1007/s00125-007-0681-5 6. Ryu T.Y., Park J., Scherer P.E.. Hyperglycemia as a Risk Factor for Cancer Progression. Diabetes Metab. J. (2014) 38 330-336. DOI: 10.4093/dmj.2014.38.5.330 7. Klöting N., Berndt J., Kralisch S., Kovacs P., Fasshauer M., Schön M.R., Stumvoll M., Blüher M.. Vaspin gene expression in human adipose tissue: Association with obesity and type 2 diabetes. Biochem. Biophys. Res. Commun. (2006) 339 430-436. DOI: 10.1016/j.bbrc.2005.11.039 8. McCampbell A.S., Walker C.L., Broaddus R.R., Cook J.D., Davies P.J.A.. Developmental reprogramming of IGF signaling and susceptibility to endometrial hyperplasia in the rat. Lab. Investig. (2008) 88 615-626. DOI: 10.1038/labinvest.2008.29 9. Guo F., Zhang H., Jia Z., Cui M., Tian J.. Chemoresistance and targeting of growth factors/cytokines signalling pathways: Towards the development of effective therapeutic strategy for endometrial cancer. Am. J. Cancer Res. (2018) 8 1317-1331. PMID: 30094104 10. Dashti S.G., English D.R., Simpson J.A., Karahalios A., Moreno-Betancur M., Biessy C., Rinaldi S., Ferrari P., Tjønneland A., Halkjær J.. Adiposity and Endometrial Cancer Risk in Postmenopausal Women: A Sequential Causal Mediation Analysis. Cancer Epidemiol. Biomark. Prev. (2021) 30 104-113. DOI: 10.1158/1055-9965.EPI-20-0965 11. Rodriguez A.C., Blanchard Z., Maurer K.A., Gertz J.. Estrogen Signaling in Endometrial Cancer: A Key Oncogenic Pathway with Several Open Questions. Horm. Cancer (2019) 10 51-63. DOI: 10.1007/s12672-019-0358-9 12. Qu X., Donnelly R.. Sex Hormone-Binding Globulin (SHBG) as an Early Biomarker and Therapeutic Target in Polycystic Ovary Syndrome. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21218191 13. Liu D., Jiang F., Shan Z., Wang B., Wang J., Lai Y., Chen Y., Li M., Liu H., Li C.. A cross-sectional survey of relationship between serum TSH level and blood pressure. J. Hum. Hypertens. (2010) 24 134-138. DOI: 10.1038/jhh.2009.44 14. Kang J.H., Kueck A.S., Stevens R., Curhan G., De Vivo I., Rosner B., Alexander E., Tworoger S.S.. A Large Cohort Study of Hypothyroidism and Hyperthyroidism in Relation to Gynecologic Cancers. Obstet. Gynecol. Int. (2013) 2013 743721. DOI: 10.1155/2013/743721 15. Schindler A.E.. Progestogen deficiency and endometrial cancer risk. Maturitas (2009) 62 334-337. DOI: 10.1016/j.maturitas.2008.12.018 16. Uchikova E., Uchikov P., Parahuleva P.. Obesity and Endometrial Carcinogenesis. Akush. Ginekol. (2015) 54 34-37 17. González C.R., Caminos J.E., Vázquez M.J., Garcés M.F., Cepeda L.A., Ángel A., González A.C., García-Rendueles M.E., Sangiao-Alvarellos S., López M.. Regulation of visceral adipose tissue-derived serine protease inhibitor by nutritional status, metformin, gender and pituitary factors in rat white adipose tissue. J. Physiol. (2009) 587 3741-3750. DOI: 10.1113/jphysiol.2009.172510 18. Vicennati V., Garelli S., Rinaldi E., Rosetti S., Zavatta G., Pagotto U., Pasquali R.. Obesity-related proliferative diseases: The interaction between adipose tissue and estrogens in post-menopausal women. Horm. Mol. Biol. Clin. Investig. (2015) 21 75-87. DOI: 10.1515/hmbci-2015-0002 19. Hanby A.M., Walker C.. Tavassoli FA, Devilee P: Pathology and Genetics: Tumours of the Breast and Female Genital Organs. WHO Classification of Tumours series—Volume IV. Lyon, France: IARC Press: 2003. 250pp. ISBN 92 832 2412 4. Breast Cancer Res. (2004) 6 133. DOI: 10.1186/bcr788 20. Lacey J.V., Ioffe O.B., Ronnett B.M., Rush B.B., Richesson D.A., Chatterjee N., Langholz B., Glass A.G., Sherman M.E.. Endometrial carcinoma risk among women diagnosed with endometrial hyperplasia: The 34-year experience in a large health plan. Br. J. Cancer (2008) 98 45-53. DOI: 10.1038/sj.bjc.6604102 21. Cymbaluk-Płoska A., Chudecka-Głaz A., Jagodzińska A., Pius-Sadowska E., Sompolska-Rzechuła A., Machalinski B., Menkiszak J.. Evaluation of biologically active substances promoting the development of or protecting against endometrial cancer. OncoTargets Ther. (2018) 11 1363-1372. DOI: 10.2147/OTT.S155942 22. Ortega F.B., Lavie C.J., Blair S.N.. Obesity and Cardiovascular Disease. Circ. Res. (2016) 118 1752-1770. DOI: 10.1161/CIRCRESAHA.115.306883 23. Avgerinos K.I., Spyrou N., Mantzoros C.S., Dalamaga M.. Obesity and cancer risk: Emerging biological mechanisms and perspectives. Metabolism (2019) 92 121-135. DOI: 10.1016/j.metabol.2018.11.001 24. Caballero B.. Humans against Obesity: Who Will Win?. Adv. Nutr. Int. Rev. J. (2019) 10 S4-S9. DOI: 10.1093/advances/nmy055 25. Ohashi K., Shibata R., Murohara T., Ouchi N.. Role of anti-inflammatory adipokines in obesity-related diseases. Trends Endocrinol. Metab. (2014) 25 348-355. DOI: 10.1016/j.tem.2014.03.009 26. Kawai T., Autieri M.V., Scalia R.. Adipose tissue inflammation and metabolic dysfunction in obesity. Am. J. Physiol. Physiol. (2021) 320 C375-C391. DOI: 10.1152/ajpcell.00379.2020 27. Liu S., Duan R., Wu Y., Du F., Zhang J., Li X., Guo S., Wang M., Zhang Q., Li Y.. Effects of Vaspin on Insulin Resistance in Rats and Underlying Mechanisms. Sci. Rep. (2018) 8 13542. DOI: 10.1038/s41598-018-31923-3 28. Lauby-Secretan B., Scoccianti C., Loomis D., Grosse Y., Bianchini F., Straif K.. Body Fatness and Cancer—Viewpoint of the IARC Working Group. N. Engl. J. Med. (2016) 375 794-798. DOI: 10.1056/NEJMsr1606602 29. Michalczyk K., Niklas N., Rychlicka M., Cymbaluk-Płoska A.. The Influence of Biologically Active Substances Secreted by the Adipose Tissue on Endometrial Cancer. Diagnostics (2021) 11. DOI: 10.3390/diagnostics11030494 30. Li Q., Chen R., Moriya J., Yamakawa J.-I., Sumino H., Kanda T., Takahashi T.. A Novel Adipocytokine, Visceral Adipose Tissue-derived Serine Protease Inhibitor (Vaspin), and Obesity. J. Int. Med. Res. (2008) 36 625-629. DOI: 10.1177/147323000803600402 31. Trojnar M., Patro-Małysza J., Kimber-Trojnar Ż., Czuba M., Mosiewicz J., Leszczyńska-Gorzelak B.. Vaspin in Serum and Urine of Post-Partum Women with Excessive Gestational Weight Gain. Medicina (2019) 55. DOI: 10.3390/medicina55030076 32. Kurowska P., Mlyczyńska E., Dawid M., Jurek M., Klimczyk D., Dupont J., Rak A.. Review: Vaspin (SERPINA12) Expression and Function in Endocrine Cells. Cells (2021) 10. DOI: 10.3390/cells10071710 33. von Loeffelholz C., Möhlig M., Arafat A.M., Isken F., Spranger J., Mai K., Randeva H.S., Pfeiffer A.F.H., Weickert M.O.. Circulating vaspin is unrelated to insulin sensitivity in a cohort of nondiabetic humans. Eur. J. Endocrinol. (2010) 162 507-513. DOI: 10.1530/EJE-09-0737 34. Wiśniewska-Ulfik D., Behrendt J., Nawrat A., Szymańska A., Stachurska-Klimczak A., Stojewska M.. Increased serum vaspin concentration in full-term neonates with early-onset infections. Pediatr. Polska (2019) 94 235-240. DOI: 10.5114/polp.2019.88040 35. Wang Y., Zhou R., Wang J.. Relationship between Hypothyroidism and Endometrial Cancer. Aging Dis. (2019) 10 190-196. DOI: 10.14336/AD.2018.0224 36. Du F.-M., Kuang H.-Y., Duan B.-H., Liu D.-N., Yu X.-Y.. Effects of thyroid hormone and depression on common components of central obesity. J. Int. Med. Res. (2019) 47 3040-3049. DOI: 10.1177/0300060519851624 37. Seebacher V., Bergmeister B., Grimm C., Koelbl H., Reinthaller A., Polterauer S.. The prognostic role of metformin in patients with endometrial cancer: A retrospective study. Eur. J. Obstet. Gynecol. Reprod. Biol. (2016) 203 291-296. DOI: 10.1016/j.ejogrb.2016.06.013 38. Kadoglou N.P., Gkontopoulos A., Kapelouzou A., Fotiadis G., Theofilogiannakos E.K., Kottas G., Lampropoulos S.. Serum levels of vaspin and visfatin in patients with coronary artery disease—Kozani study. Clin. Chim. Acta (2011) 412 48-52. DOI: 10.1016/j.cca.2010.09.012 39. Xu C., Yang X., Liu W., Yuan H., Yu C., Gao L., Zhao J.. Thyroid stimulating hormone, independent of thyroid hormone, can elevate the serum total cholesterol level in patients with coronary heart disease: A cross-sectional design. Nutr. Metab. (2012) 9 44. DOI: 10.1186/1743-7075-9-44 40. Körner A., Neef M., Friebe D., Erbs S., Kratzsch J., Dittrich K., Blüher S., Kapellen T.M., Kovacs P., Stumvoll M.. Vaspin is related to gender, puberty and deteriorating insulin sensitivity in children. Int. J. Obes. (2011) 35 578-586. DOI: 10.1038/ijo.2010.196 41. Erdogan S., Sezer S., Baser E., Gun-Eryilmaz O., Gungor T., Uysal S., Yilmaz F.M.. Evaluating vaspin and adiponectin in postmenopausal women with endometrial cancer. Endocr.-Relat. Cancer (2013) 20 669-675. DOI: 10.1530/ERC-13-0280 42. Franik G., Plinta R., Madej P., Owczarek A., Bozentowicz-Wikarek M., Chudek J., Skrzypulec-Plinta V., Olszanecka-Glinianowicz M.. Circulating vaspin levels and nutritional status and insulin resistance in polycystic ovary syndrome. Ginekol. Polska (2020) 91 251-255. DOI: 10.5603/GP.2020.0056 43. Gulcelik N.E., Karakaya J., Gedik A., Usman A., Gurlek A.. Serum vaspin levels in type 2 diabetic women in relation to microvascular complications. Eur. J. Endocrinol. (2009) 160 65-70. DOI: 10.1530/EJE-08-0723 44. Sidorkiewicz I., Jóźwik M., Niemira M., Krętowski A.. Insulin Resistance and Endometrial Cancer: Emerging Role for microRNA. Cancers (2020) 12. DOI: 10.3390/cancers12092559 45. Lu H., Wamba P.C.F., Lapointe M., Poirier P., Martin J., Bastien M., Cianflone K.. Increased Vaspin Levels Are Associated with Beneficial Metabolic Outcome Pre- and Post-Bariatric Surgery. PLoS ONE (2014) 9. DOI: 10.1371/journal.pone.0111002 46. Schmandt R.E., Iglesias D.A., Na Co N., Lu K.H.. Understanding obesity and endometrial cancer risk: Opportunities for prevention. Am. J. Obstet. Gynecol. (2011) 205 518-525. DOI: 10.1016/j.ajog.2011.05.042
--- title: 'Differences in Center for Epidemiologic Studies Depression Scale, Generalized Anxiety Disorder-7 and Kessler Screening Scale for Psychological Distress Scores between Smartphone Version versus Paper Version Administration: Evidence of Equivalence' authors: - Kazuki Hirao - Hyono Takahashi - Natsuki Kuroda - Hiroyuki Uchida - Kenji Tsuchiya - Senichiro Kikuchi journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049019 doi: 10.3390/ijerph20064773 license: CC BY 4.0 --- # Differences in Center for Epidemiologic Studies Depression Scale, Generalized Anxiety Disorder-7 and Kessler Screening Scale for Psychological Distress Scores between Smartphone Version versus Paper Version Administration: Evidence of Equivalence ## Abstract The use of electronic patient-reported outcomes has increased recently, and smartphones offer distinct advantages over other devices. However, previous systematic reviews have not investigated the reliability of the Center for Epidemiologic Studies Depression Scale (CES-D), Generalized Anxiety Disorder-7 (GAD-7), and Kessler Screening Scale for Psychological Distress (K6) when used with smartphones, and this has not been fully explored. This study aimed to evaluate the equivalence of the paper and smartphone versions of the CES-D, GAD-7, and K6, which were compared following a randomized crossover design method in 100 adults in Gunma, Japan. Participants responded to the paper and smartphone versions at 1-week intervals. The equivalence of paper and smartphone versions was evaluated using the intraclass correlation coefficient (ICCagreement). The mean participant age was 19.86 years (SD = 1.08, $23\%$ male). The ICCagreements for the paper and smartphone versions of the CES-D, GAD-7, and K6 were 0.76 ($95\%$ confidence interval [CI] 0.66–0.83), 0.68 ($95\%$ CI 0.59–0.77), and 0.83 ($95\%$ CI 0.75–0.88), respectively. Thus, the CES-D and K6 scales are appropriate for use in a smartphone version, which could be applied to clinical and research settings in which the paper or smartphone versions could be used as needed. ## 1. Introduction The use of patient-reported outcomes (PROs) is neccessary because of several advantages [1,2,3]. Previous studies have shown that the use of PROs to systematically monitor patient symptoms improves patient–physician communication, symptom oversight, and gaps in patient health, quality of life, and clinician perception of symptoms [1,2,3]. PROs are also widely used in the mental health field, and mental health clinicians suggest that the use of PROs in patient consultations can help in making treatment decisions and severity assessment [4,5]. Depressive symptoms, anxiety, and psychological distress are particularly common in the field of mental health, and it has been indicated that these symptoms may coexist and affect each other [6,7,8,9,10,11,12]. As a result, it is crucial to thoroughly evaluate utilizing PRO not just one symptom but also depressed symptoms, anxiety symptoms, and psychological distress. Currently, many PROs exist to measure depressive and anxiety symptoms and psychological distress. For example, the Center for Epidemiologic Studies Depression Scale (CES-D) [13,14], Generalized Anxiety Disorder-7 (GAD-7) [15,16], and Kessler Screening Scale for Psychological Distress (K6) [17,18] are widely used PROs to measure depressive and anxiety symptoms. The CES-D is a 20-item PRO developed to assess depressive symptoms in both clinical and nonclinical settings [13,14]. The GAD-7 is a 7-item PRO used in screening for generalized anxiety disorder and other anxiety disorders, such as panic disorder, social anxiety disorder, and post-traumatic stress disorder [15,16]. K6 is a 6-item PRO developed to measure psychological distress [17,18]. These PROs can be answered in a short time and are easy to grade [10,14,16,17,19,20,21]. In addition, they were translated in many languages, and their psychometric properties, including reliability and validity, have been reported [10,13,14,15,16,17,18,22,23,24,25,26,27,28,29]. The CES-D, GAD-7, and K6 have been translated into Japanese, and their reliability and validity have been examined in several studies [10,13,17,26,29,30]. Given these advantages, CES-D, GAD-7, and K6 are widely used in both clinical and epidemiological studies and diagnostic screening in the local general population [19,20,31,32]. Importantly, these PROs are also inevitably used as electronic patient-reported outcomes (ePROs) as increasingly more studies in the mental health field use the Internet [33,34,35,36,37,38]. Compared with paper-based PROs, ePROs minimize errors in score calculation and data entry and missing data, facilitating reliable analysis and reporting of PRO data [39,40,41,42]. Previous studies have also suggested that patients prefer ePROs to paper-based PROs, and by using ePROs, patients may disclose more sensitive information than paper-based PROs [42,43,44,45,46,47,48]. As a result, making the CES-D, GAD-7, and K6 available as ePROs, which have reliability and validity and are employed in many countries, will not only make these instruments easier to use for participants, researchers, and healthcare professionals but may also decrease administrative burden and avoid missing data. Smartphones are playing an increasingly important role in capitalizing on these potential benefits of ePRO use in clinical and research settings. Currently, many devices are being utilized for ePROs [42]. Among them, smartphones offer distinct advantages over other devices for the use of ePROs. Smartphone users are increasing worldwide, and most people carry their smartphones with them at all times [39,49]. In addition, more people are using smartphones than personal computers (PCs) to access the Internet [49]. Therefore, smartphones will enable PROs in a more real-time manner than PCs or tablets. Moreover, several studies point to the value of employing smartphones as ePRO devices [39,50,51]. However, to the best of our knowledge, no study has confirmed the equivalence of the electronic and paper versions of the K6 and GAD-7. However, several previous studies have confirmed the equivalence of the electronic and paper versions of the CES-D [52,53]. Contrarily, a previous systematic review did not verify the equivalence between the smartphone version of the CES-D and the paper version of the CES-D [52,53]. As a result, it is unlikely that the reliability of the CES-D, GAD-7, and K6 when applied to smartphones has been sufficiently researched. [ 38,46,54]. The International Society for Pharmacoeconomics and Outcomes Research (ISPOR) guidelines suggest that differences in how ePRO and original PRO questions are presented may adversely affect the reliability and validity [42]. In addition, and of particular importance, the reliability and validity of ePROs may be affected by the type of device used, e.g., PCs or tablets [42,46]. Because of these issues related to the transition from the original PRO to ePRO, the ISPOR guidelines need cognitive debriefing and usability testing to be conducted for minor changes (i.e., from circling the answer to touching the answer on the screen, etc.) when changing from PRO to ePRO. Moderate changes (i.e., need to scroll the screen, change font size, etc.) show the need to perform reliability measures (e.g., intraclass correlation coefficient), while large changes (e.g., concerning response choices or item wording) demonstrate the need for full psychometric testing [42]. Therefore, the ePRO and the original PRO should be compared on a device-by-device basis, and whether they are equivalent, if not superior, must be verified. Therefore, this study aimed to examine the measurement equivalence of the paper and smartphone versions of CES-D, GAD-7, and K6 based on ISPOR guidelines. ## 2.1. Study Design This study was conducted using a randomized crossover design to assess the format equivalence of the paper and smartphone versions of the CES-D, GAD-7, and K6. Figure 1 depicts the process of the randomized crossover design used in this investigation. The study was conducted in accordance with ISPOR guidelines [42] and was approved by the Ethical Review Board for Medical Research Involving Human Subjects of Gunma University (Approval no. HS2022-109). Written informed consent was obtained from each participant before study participation. ## 2.2. Participants and Procedure The study participants were recruited between October 2022 and December 2022 from Gunma University in Gunma, Japan. The recruitment was made by posting posters at Gunma University. Study participation was also encouraged via e-mail and social networking services. Individuals aged ≥18 years who were native Japanese speakers and had a smartphone were considered eligible for this study. Participants who met the eligibility criteria were asked to complete the CES-D, GAD-7, and K6 scales (paper and smartphone versions) after answering demographic information (age and sex) and lifestyle characteristics (i.e., drinking, exercise, and smoking habits). The order in which the PROs were filled out (paper version first or smartphone version first) was randomly determined. To reduce potential recall and carryover effects, the interval between the completion of the two questionnaires was 1 week. ## 2.3. Randomization Participants were randomly assigned in a 1:1 ratio to complete either the paper version first or the smartphone version first before answering the questionnaire (CES-D, GAD-7, and K6). The randomization list was generated by a permuted block method (block size 4) using a computer (Microsoft Excel) by a third party unrelated to the study. The randomization list was sent to the Central Registry Center at Kurashiki Heisei Hospital in Okayama Prefecture, Japan, for random assignment. ## 2.4. Sample Size The ISPOR guidelines report that 43 participants with no missing data are needed to declare an ICC of ≥0.7 at $80\%$ power and $95\%$ confidence level if the ICC observed in two measurements is expected to be 0.85, using the approximation used by Walter et al. [ 42,55]. Conversely, the Consensus-based Standards for the Selection of Health Measurement Instruments initiative suggests that a sample size of ≥100 is necessary to obtain statistical power when evaluating test–retest reliability [56]. Taken together, these findings suggest a target sample size of 100 study participants. ## 2.5.1. CES-D The CES-D is a 20-item self-report questionnaire used to measure depressive symptoms [13,14]. Each item has a 0–3 Likert scale (A = <1 day, $B = 1$–2 days, $C = 3$–4 days, and $D = 5$–7 days) with a total score of 0–60. Higher scores indicate high levels of depressive symptoms. Previous studies have reported the reliability and validity of the CES-D score [10,13,14,28,29,57]. ## 2.5.2. GAD-7 The GAD-7 is a 7-item self-report questionnaire used to measure generalized anxiety disorder, on a 0–3 Likert scale (0 = not at all sure, 1 = several days, 2 = over half the days, and 3 = nearly every day) [15,16]. The total scores range from 0 to 21, with higher scores indicating greater anxiety. Previous studies have reported the reliability and validity of the GAD-7 score [15,16,22,23,24,25]. ## 2.5.3. K6 The K6 is a 6-item self-report questionnaire used to measure psychological distress, using a 0–4 Likert scale (0 = none of the time, 1 = a little of the time, 2 = some of the time, 3 = most of the time, and 4 = all of the time) [17,18]. Total scores range from 0 to 24, with higher scores indicating greater psychological distress. Previous studies have reported the reliability and validity of the K6 score [10,17,18,26,27]. ## 2.6. Software Electronic versions of CES-D, GAD-7, and K6 were provided on participants’ smartphones using Google Forms. The questionnaires were presented in the order CES-D, GAD-7, and K6. The questions, answer choices, and order of questions in the electronic version are the same as those in the paper version of the three scales. Each questionnaire was presented on a separate page; however, all the questions for each questionnaire are displayed on the screen. Scrolling down the screen allows the user to move to the next answer. After answering all the questions in the questionnaire, the next questionnaire can be answered by pressing the “Next” button (specifically, the 20 questions in the CES-D are displayed on a single page, and after answering all of them, the “Next” button is pressed to move to the GAD-7 questionnaire page). Participants can select their answers by tapping the radio buttons on the screen. It is not possible to move to the next page without answering a question item or to select two answers to the same question. However, it is possible to change a previous answer by pressing the “Back” button. ## 2.7. Statistical Analysis In this study, the switch from the paper version to the smartphone version corresponds to the light to moderate adjustment suggested by the ISPOR guidelines [42]. As a result, to confirm the equivalence of each scale between the paper and smartphone versions, the intraclass correlation coefficient (ICCagreement) and its $95\%$ confidence interval were calculated based on the two-way random-effects model, one of the most commonly used statistical measures in equivalence studies of this kind [42,58]. Unlike the Pearson and Spearman correlation coefficients, the ICCagreement is more appropriate for assessing agreement because it considers not only chance errors but also systematic errors [56,59]. ICC is expressed as a value between 0 and 1, with values >0.70 indicating adequate reliability [56,58]. The internal consistency between the paper and smartphone versions of each questionnaire was calculated using Cronbach’s alpha and McDonald’s omega. Furthermore, $95\%$ confidence intervals (CIs) for these indices were calculated; values of Cronbach’s alpha and McDonald’s omega were denoted as 0–1. The alpha and omega values increase with the degree of correlation between the objects [60]. Good internal consistency is defined as Cronbach’s alpha and McDonald’s omega values of 0.7 or above [59,60]. In addition, linear mixed models (LMM) were used to confirm the carryover effect of each scale score [61]. In the LMM, the questionnaire administration format (paper or smartphone version), order of administration (paper or smartphone version first), and interaction between questionnaire administration format and order of administration are considered fixed-effect factors, whereas participants were considered random-effect factors. Statistical significance was set at $p \leq 0.05$ with a two-tailed test. All analyses were performed in R (version 4.0.2 for Windows; The R Project for Statistical Computing; Vienna, Austria). ## 3.1. Characteristics of the Study Participants Of the 100 participants who met eligibility, 100 completed the paper and smartphone versions of the questionnaire and provided complete data. In the paper-first group, 50 participants first completed a paper-version questionnaire. In the smartphone-first group, 50 participants first completed the smartphone version questionnaire. The mean age of the study participants was 19.86 years (SD = 1.08, $23\%$ male), 9 ($9\%$) had a drinking habit, 1 ($1\%$) had a smoking habit, and 37 ($37\%$) had an exercise habit (Table 1). ## 3.2. Mean and LMM Results The mean values for each group and the LMM results are shown in Table 2. The interaction of questionnaire format and order of administration on the CES-D score was not significant ($$p \leq 0.96$$; $95\%$ CI −1.71 to 1.79). The interaction of a questionnaire format and order of implementation on GAD-7 scores was not significant ($$p \leq 0.96$$; $95\%$ CI −0.82 to 0.78). The interaction of a questionnaire format and order of implementation on the K6 score was not significant ($$p \leq 0.17$$; $95\%$ CI −1.31 to 0.23). Based on these results, no carryover effects were observed. ## 3.3. Equivalence The ICCagreement values between the paper and smartphone versions of the CES-D, GAD-7, and K6 scores were 0.76 ($95\%$ CI 0.66–0.83), 0.68 ($95\%$ CI 0.59–0.77), and 0.83 ($95\%$ CI 0.75–0.88), respectively (Table 3). ## 3.4. Internal Consistency Cronbach’s alpha values for the CES-D score were 0.82 ($95\%$ CI 0.77–0.87) and 0.81 ($95\%$ CI 0.75–0.86) for the smartphone and paper versions, respectively. Cronbach’s alpha values for the GAD-7 score were 0.80 ($95\%$ CI 0.75–0.86) and 0.80 ($95\%$ CI 0.75–0.86) for the smartphone and paper versions, respectively. Cronbach’s alpha values for the K6 score were 0.88 ($95\%$ CI 0.84–0.92) and 0.82 ($95\%$ CI 0.77–0.88) for the smartphone and paper versions, respectively (Table 4). McDonald’s omega values for the CES-D score were 0.83 ($95\%$ CI 0.75–0.87) and 0.81 ($95\%$ CI 0.74–0.86) for the smartphone and paper versions, respectively. McDonald’s omega values for the GAD-7 score were 0.83 ($95\%$ CI 0.76–0.87) and 0.84 ($95\%$ CI 0.72–0.91) for the smartphone and paper versions, respectively. McDonald’s omega values for the K6 score were 0.87 ($95\%$ CI 0.84–0.92) and 0.83 ($95\%$ CI 0.76–0.88) for the smartphone and paper versions, respectively (Table 4). ## 4. Discussion This study evaluated the equivalence of the embodiments to the CES-D, GAD-7, and K6 evaluated in smartphone and paper versions. The results suggest that CES-D and K6 have good equivalence, with ICCagreements of 0.76 and 0.83, respectively. Cronbach’s alpha values of the smartphone versions of CES-D and K6 were 0.82 ($95\%$ CI 0.77–0.87) and 0.88 ($95\%$ CI 0.84–0.92), respectively, indicating that they not only have good internal consistency but also comparable internal consistency to the paper versions of CES-D (0.81; $95\%$ CI 0.75–0.86) and K6 (0.82; $95\%$ CI 0.77–0.88). McDonald’s omega values for the smartphone versions of CES-D and K6 were 0.83 ($95\%$ CI 0.75–0.87) and 0.87 ($95\%$ CI 0.84–0.92), respectively, indicating that they not only have good internal consistency but also comparable internal consistency to the paper versions of CES-D (0.81; $95\%$ CI 0.74–0.86) and K6 (0.83; $95\%$ CI 0.76–0.88). The results suggest that the smartphone versions of the CES-D and K6 produce comparable self-assessments as the paper versions of the CES-D and K6. Previous studies have suggested that both ICC and Cronbach’s alpha should be at least 0.7 for group-level use and 0.85–0.95 for individual-level use [42]. Considering the ICC and Cronbach’s alpha criteria, the smartphone versions of CES-D and K6 are at least considered suitable for use at the group level. In other words, the smartphone versions of the CES-D and K6 may not be suitable for use on an individual level. However, it is crucial to remember that the ICCagreement’s $95\%$ CI for K6 was 0.75–0.88 and for CES-D was 0.66–0.83. This $95\%$ CI indicates that, with a $95\%$ probability, the true value of ICCagreement for CES-D is 0.83 in the best case and 0.66 in the worst case [62]. Therefore, while the smartphone and paper versions of the CES-D reveal better agreement, they may also indicate lower agreement, below the threshold of 0.7, which is considered good. However, the ICC being below 0.7 may not necessarily be due to a low degree of agreement on the scale but also to issues of study design, such as low inter-subject variability sampled and sample size [63]. The low variability among sampled patients probably had an impact on the accuracy of the ICCagreement estimations because our study had a large enough sample size to assess the ICC suggested by the Consensus-based Guidelines for the Selection of Health Measuring Instruments initiative [63]. We were restricted to a relatively young population (18–22 years old) in our sample. As a result, further investigation in a broader age population is required to provide more accurate estimates of ICCagreement and its $95\%$ CI. The Cronbach’s alpha for the GAD-7 on smartphones was 0.80 ($95\%$ CI 0.75–0.86), indicating that it has the same internal consistency as the GAD-7 on paper (0.80; $95\%$ CI 0.75–0.86). McDonald’s omega values for the GAD-7 on a smartphone were also 0.83 ($95\%$ CI 0.76–0.88), and they were 0.83 ($95\%$ CI 0.76–0.88) for the GAD-7 on paper, indicating strong internal consistency. However, the ICCagreement for GAD-7 was 0.68 ($95\%$ CI 0.59–0.77), suggesting a low concordance between the smartphone and paper versions. This low ICCagreement could be attributed to the changes following the transition from the paper version to the smartphone version. In this study, participants scrolled the screen to answer the items in each of the smartphone versions of the questionnaire. In addition, the questions and their response items were displayed in different positions in the paper and smartphone versions. These changes are defined as a moderate level of modification in the ISPOR guidelines, which is the level of modification that requires equivalence assessment [42]. In GAD-7, these changes from paper to smartphone versions may not have been suitable. Future studies should create a smartphone version of the GAD-7 with a display format more similar to the paper version to evaluate equivalence. It is also essential to note that the $95\%$ CI for ICCagreement in GAD-7, as in CES-D, was 0.59–0.77. This $95\%$ CI means that the true value of ICCagreement for GAD-7 is 0.77 in the best case and 0.59 in the worst case, with a $95\%$ probability [62]. As a result, even while the GAD-7 on a smartphone or piece of paper would finally surpass the 0.7 criterion, they might still exhibit inferior agreement. However, even in the ICCagreement for the GAD-7, the effect of the low sample variability in this study cannot be ignored [63]. Hence, similar to the CES-D, more research in a larger age range is required to more precisely estimate the ICCagreement and its $95\%$ CI. As far as we could find, no studies have tested the equivalence of the electronic and paper versions of the K6 and GAD-7. However, previous studies have examined the equivalence of electronic and paper versions of the CES-D. A study of 2400 teachers in Taiwan, which tested the equivalence of the Internet-based CES-D and paper-based CES-D, found little difference in potential means and concluded that Internet-based CES-D is a promising alternative to paper-based CES-D [53]. In addition, the equivalence of the paper- and tablet-based methods was tested in 79 patients with low back pain, and the ICC was 0.75 (0.64–0.83), which is comparable to our results [52]. On the contrary, previous study have tested the equivalence of PC- and paper-based CES-Ds and suggested correlation coefficients ranging from 0.96 [64]. However, the Pearson and Spearman correlation coefficients are not extremely rigorous parameters for assessing equivalence because they do not account for systematic errors [42,59]. Considering the characteristics of the results of these previous studies and the potential advantages of smartphones (easy and ubiquitous accessibility), at least a smartphone version of CES-D may be a promising alternative strategy for PC- and tablet-based CES-D. This study has several limitations. First, the study participants were a relatively young population, aged 18–22 years. Therefore, the results of this study may not apply to other age groups. Second, the influence of the carryover effects cannot be ignored. In a crossover design, a carryover effect may occur if the interval between the first and second evaluations is short. We tried to reduce the carryover effect as much as possible by keeping the interval between the first and second evaluations to 1 week. In fact, no statistically significant differences in the carryover effects were found in this study. However, given the lack of consensus on the ideal implementation interval when testing the equivalence of PROs [65], the influence of carryover effects must be carefully considered. Third, the smartphone and paper versions of the PROs were administered in the same room under the supervision of the researcher. If participants responded to the smartphone version of the PRO without meeting the researcher face to face, they may have been more anonymous than in our study and could have responded in a more natural setting. Therefore, the presence or absence of a supervisor and the effect of locations such as the clinic or home setting, should be fully considered. On the contrary, responding in the same room with the researcher made it possible to prevent omissions in the paper version and control for test conditions that would reduce the general likelihood of noise, distraction, mood fatigue, etc. [ 66]. Fourth, due to the difficulty of the participant burden in completing the questions, this study did not examine cognitive debriefing or usability testing, which are classified as minor alterations. Future studies should incorporate cognitive debriefing and usability testing of the smartphone versions of the CES-D, GAD-7, and K6, as these characteristics may considerably alter their usefulness in research and clinical contexts. Fifth, the smartphone versions of the CES-D, GAD-7, and K6 employed in this study could not be completed until all items were answered. The equivalency results reached in this study may have been impacted if participants were made to complete tasks they could have skipped in the paper version. Consequently, the effect of forced responses in the smartphone version of this study should be properly considered. Future studies should explore the equivalence of the paper and smartphone versions of the CES-D, GAD-7, and K6 by including a “choose not to answer” or “skip question” option. Sixth, participants replied to the CES-D, GAD-7, and K6 in that order on both the paper and smartphone versions of the survey. Thus, it was impossible to rule out the impact of ordering effects. As a result, the effects of order effects should be taken into consideration while interpreting the findings of this study. ## 5. Conclusions This study demonstrates the equivalence of the paper and smartphone versions of the CES-D and K6. Accordingly, both the CES-D and K6 scales are appropriate for use in a smartphone version, which could be applied to clinical and research settings in which paper and smartphone versions could be selected as needed. However, the paper and smartphone versions of the GAD-7 should not be used interchangeably, as the paper and smartphone versions did not show equivalence because of low ICCagreement; thus, further research is needed. ## References 1. Snyder C.F., Aaronson N.K., Choucair A.K., Elliott T.E., Greenhalgh J., Halyard M.Y., Hess R., Miller D.M., Reeve B.B., Santana M.. Implementing patient-reported outcomes assessment in clinical practice: A review of the options and considerations. Qual. Life Res. (2012) 21 1305-1314. DOI: 10.1007/s11136-011-0054-x 2. Basch E., Barbera L., Kerrigan C.L., Velikova G.. Implementation of Patient-Reported Outcomes in Routine Medical Care. Am. Soc. Clin. Oncol. Educ. Book (2018) 38 122-134. DOI: 10.1200/EDBK_200383 3. Valderas J.M., Kotzeva A., Espallargues M., Guyatt G., Ferrans C.E., Halyard M.Y., Revicki D.A., Symonds T., Parada A., Alonso J.. The impact of measuring patient-reported outcomes in clinical practice: A systematic review of the literature. Qual. Life Res. (2008) 17 179-193. DOI: 10.1007/s11136-007-9295-0 4. Churruca K., Pomare C., Ellis L.A., Long J.C., Henderson S.B., Murphy L.E.D., Leahy C.J., Braithwaite J.. Patient-reported outcome measures (PROMs): A review of generic and condition-specific measures and a discussion of trends and issues. Health Expect. (2021) 24 1015-1024. DOI: 10.1111/hex.13254 5. Duffy F.F., Chung H., Trivedi M., Rae D.S., Regier D.A., Katzelnick D.J.. Systematic use of patient-rated depression severity monitoring: Is it helpful and feasible in clinical psychiatry?. Psychiatr. Serv. (2008) 59 1148-1154. DOI: 10.1176/ps.2008.59.10.1148 6. Kessler R.C., Berglund P., Demler O., Jin R., Merikangas K.R., Walters E.E.. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Arch. Gen. Psychiatry (2005) 62 593-602. DOI: 10.1001/archpsyc.62.6.593 7. Guest R., Tran Y., Gopinath B., Cameron I.D., Craig A.. Psychological distress following a motor vehicle crash: A systematic review of preventative interventions. Injury (2016) 47 2415-2423. DOI: 10.1016/j.injury.2016.09.006 8. Tiller J.W.. Depression and anxiety. Med. J. Aust. (2013) 199 S28-S31. DOI: 10.5694/mja12.10628 9. Bot M., Pouwer F., Ormel J., Slaets J.P., de Jonge P.. Predictors of incident major depression in diabetic outpatients with subthreshold depression. Diabet. Med. J. Br. Diabet. Assoc. (2010) 27 1295-1301. DOI: 10.1111/j.1464-5491.2010.03119.x 10. Sakurai K., Nishi A., Kondo K., Yanagida K., Kawakami N.. Screening performance of K6/K10 and other screening instruments for mood and anxiety disorders in Japan. Psychiatry Clin. Neurosci. (2011) 65 434-441. DOI: 10.1111/j.1440-1819.2011.02236.x 11. Nishi D., Imamura K., Watanabe K., Ishikawa H., Tachimori H., Takeshima T., Kawakami N.. Psychological distress with and without a history of depression: Results from the World Mental Health Japan 2nd Survey (WMHJ2). J. Affect. Disord. (2020) 265 545-551. DOI: 10.1016/j.jad.2019.11.089 12. Nishi D., Susukida R., Usuda K., Mojtabai R., Yamanouchi Y.. Trends in the prevalence of psychological distress and the use of mental health services from 2007 to 2016 in Japan. J. Affect. Disord. (2018) 239 208-213. DOI: 10.1016/j.jad.2018.07.016 13. Shima S., Shikano T., Kitamura T., Asai M.. New self-rating scale for depression. Seishin Igaku (Clin. Psychiatry) (1985) 27 717-723 14. Radloff L.S.. The CES-D scale: A self-report depression scale for research in the general population. J. Appl. Psychol. Meas. (1977) 1 385-401. DOI: 10.1177/014662167700100306 15. Muramatsu K., Miyaoka H., Kamijima K., Muramatsu Y., Fuse K., Yoshimine F., Hosaka M., Kutsumi R., Majima I., Katagiri A.. Examination of the validity and the utility of the Japanese version of the GAD-7. Jpn. J. Psychosom. Med. (2010) 50 166 16. Spitzer R.L., Kroenke K., Williams J.B., Löwe B.. A brief measure for assessing generalized anxiety disorder: The GAD-7. Arch. Intern. Med. (2006) 166 1092-1097. DOI: 10.1001/archinte.166.10.1092 17. Furukawa T.A., Kawakami N., Saitoh M., Ono Y., Nakane Y., Nakamura Y., Tachimori H., Iwata N., Uda H., Nakane H.. The performance of the Japanese version of the K6 and K10 in the World Mental Health Survey Japan. Int. J. Methods Psychiatr. Res. (2008) 17 152-158. DOI: 10.1002/mpr.257 18. Kessler R.C., Andrews G., Colpe L.J., Hiripi E., Mroczek D.K., Normand S.L., Walters E.E., Zaslavsky A.M.. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol. Med. (2002) 32 959-976. DOI: 10.1017/S0033291702006074 19. Mitchell C.M., Beals J.. The utility of the Kessler Screening Scale for Psychological Distress (K6) in two American Indian communities. Psychol. Assess (2011) 23 752-761. DOI: 10.1037/a0023288 20. Plummer F., Manea L., Trepel D., McMillan D.. Screening for anxiety disorders with the GAD-7 and GAD-2: A systematic review and diagnostic metaanalysis. Gen. Hosp. Psychiatry (2016) 39 24-31. DOI: 10.1016/j.genhosppsych.2015.11.005 21. Mossman S.A., Luft M.J., Schroeder H.K., Varney S.T., Fleck D.E., Barzman D.H., Gilman R., DelBello M.P., Strawn J.R.. The Generalized Anxiety Disorder 7-item scale in adolescents with generalized anxiety disorder: Signal detection and validation. Ann. Clin. Psychiatry (2017) 29 227-234a. PMID: 29069107 22. Sousa T.V., Viveiros V., Chai M.V., Vicente F.L., Jesus G., Carnot M.J., Gordo A.C., Ferreira P.L.. Reliability and validity of the Portuguese version of the Generalized Anxiety Disorder (GAD-7) scale. Health Qual. Life Outcomes (2015) 13 50. DOI: 10.1186/s12955-015-0244-2 23. Sawaya H., Atoui M., Hamadeh A., Zeinoun P., Nahas Z.. Adaptation and initial validation of the Patient Health Questionnaire-9 (PHQ-9) and the Generalized Anxiety Disorder-7 Questionnaire (GAD-7) in an Arabic speaking Lebanese psychiatric outpatient sample. Psychiatry Res. (2016) 239 245-252. DOI: 10.1016/j.psychres.2016.03.030 24. Löwe B., Decker O., Müller S., Brähler E., Schellberg D., Herzog W., Herzberg P.Y.. Validation and standardization of the Generalized Anxiety Disorder Screener (GAD-7) in the general population. Med. Care (2008) 46 266-274. DOI: 10.1097/MLR.0b013e318160d093 25. Kroenke K., Spitzer R.L., Williams J.B., Monahan P.O., Löwe B.. Anxiety disorders in primary care: Prevalence, impairment, comorbidity, and detection. Ann. Intern. Med. (2007) 146 317-325. DOI: 10.7326/0003-4819-146-5-200703060-00004 26. Nishi A., Noguchi H., Hashimoto H., Tamiya N.. Scale development of health status for secondary data analysis using a nationally representative survey. Environ. Health Prev. Med. (2012) 17 252-257. DOI: 10.1007/s12199-011-0240-z 27. Hajebi A., Motevalian A., Amin-Esmaeili M., Rahimi-Movaghar A., Sharifi V., Hoseini L., Shadloo B., Mojtabai R.. Adaptation and validation of short scales for assessment of psychological distress in Iran: The Persian K10 and K6. Int. J. Methods Psychiatr. Res. (2018) 27 e1726. DOI: 10.1002/mpr.1726 28. Thombs B.D., Hudson M., Schieir O., Taillefer S.S., Baron M.. Reliability and validity of the center for epidemiologic studies depression scale in patients with systemic sclerosis. Arthritis Rheum. (2008) 59 438-443. DOI: 10.1002/art.23329 29. Ohno S., Takahashi K., Inoue A., Takada K., Ishihara Y., Tanigawa M., Hirao K.. Smallest detectable change and test-retest reliability of a self-reported outcome measure: Results of the Center for Epidemiologic Studies Depression Scale, General Self-Efficacy Scale, and 12-item General Health Questionnaire. J. Eval. Clin. Pract. (2017) 23 1348-1354. DOI: 10.1111/jep.12795 30. Muramatsu K., Muramatsu Y., Miyaoka H., Fuse K., Yoshimine F., Hosaka M., Katagiri A., Kutsumi R.. Validation and utility of a Japanese version of the GAD-7. Panminerva Med. 20th World Congr. Psychosom. Med. Abstr. Book (2009) 51 31. Park S.H., Yu H.Y.. How useful is the center for epidemiologic studies depression scale in screening for depression in adults? An updated systematic review and meta-analysis. Psychiatry Res. (2021) 302 114037. DOI: 10.1016/j.psychres.2021.114037 32. Park S.H., Lee H.. Is the center for epidemiologic studies depression scale as useful as the geriatric depression scale in screening for late-life depression? A systematic review. J. Affect. Disord. (2021) 292 454-463. DOI: 10.1016/j.jad.2021.05.120 33. Donker T., van Straten A., Marks I., Cuijpers P.. Quick and easy self-rating of Generalized Anxiety Disorder: Validity of the Dutch web-based GAD-7, GAD-2 and GAD-SI. Psychiatry Res. (2011) 188 58-64. DOI: 10.1016/j.psychres.2011.01.016 34. Kim S.H., Park K., Yoon S., Choi Y., Lee S.H., Choi K.H.. A Brief Online and Offline (Paper-and-Pencil) Screening Tool for Generalized Anxiety Disorder: The Final Phase in the Development and Validation of the Mental Health Screening Tool for Anxiety Disorders (MHS: A). Front. Psychol. (2021) 12 639366. DOI: 10.3389/fpsyg.2021.639366 35. Sekizaki R., Nemoto T., Tsujino N., Takano C., Yoshida C., Yamaguchi T., Katagiri N., Ono Y., Mizuno M.. School mental healthcare services using internet-based cognitive behaviour therapy for young male athletes in Japan. Early Interv. Psychiatry (2019) 13 79-85. DOI: 10.1111/eip.12454 36. Nishi D., Imamura K., Watanabe K., Obikane E., Sasaki N., Yasuma N., Sekiya Y., Matsuyama Y., Kawakami N.. The preventive effect of internet-based cognitive behavioral therapy for prevention of depression during pregnancy and in the postpartum period (iPDP): A large scale randomized controlled trial. Psychiatry Clin. Neurosci. (2022) 76 570-578. DOI: 10.1111/pcn.13458 37. Kageyama K., Kato Y., Mesaki T., Uchida H., Takahashi K., Marume R., Sejima Y., Hirao K.. Effects of video viewing smartphone application intervention involving positive word stimulation in people with subthreshold depression: A pilot randomized controlled trial. J. Affect. Disord. (2021) 282 74-81. DOI: 10.1016/j.jad.2020.12.104 38. van Ballegooijen W., Riper H., Cuijpers P., van Oppen P., Smit J.H.. Validation of online psychometric instruments for common mental health disorders: A systematic review. BMC Psychiatry (2016) 16. DOI: 10.1186/s12888-016-0735-7 39. Aiyegbusi O.L.. Key methodological considerations for usability testing of electronic patient-reported outcome (ePRO) systems. Qual. Life Res. (2020) 29 325-333. DOI: 10.1007/s11136-019-02329-z 40. Marsh J.D., Bryant D.M., Macdonald S.J., Naudie D.D.. Patients respond similarly to paper and electronic versions of the WOMAC and SF-12 following total joint arthroplasty. J. Arthroplast. (2014) 29 670-673. DOI: 10.1016/j.arth.2013.07.008 41. Lee S.J., Kavanaugh A., Lenert L.. Electronic and computer-generated patient questionnaires in standard care. Best Pract. Res. Clin. Rheumatol. (2007) 21 637-647. DOI: 10.1016/j.berh.2007.02.001 42. Coons S.J., Gwaltney C.J., Hays R.D., Lundy J.J., Sloan J.A., Revicki D.A., Lenderking W.R., Cella D., Basch E.. Recommendations on evidence needed to support measurement equivalence between electronic and paper-based patient-reported outcome (PRO) measures: ISPOR ePRO Good Research Practices Task Force report. Value Health (2009) 12 419-429. DOI: 10.1111/j.1524-4733.2008.00470.x 43. Davis R.N.. Web-based administration of a personality questionnaire: Comparison with traditional methods. Behav. Res. Methods Instrum. Comput. (1999) 31 572-577. DOI: 10.3758/BF03200737 44. Buchanan T.. Online assessment: Desirable or dangerous?. Prof. Psychol. Res. Pract. (2002) 33 148. DOI: 10.1037/0735-7028.33.2.148 45. Ali F.M., Johns N., Finlay A.Y., Salek M.S., Piguet V.. Comparison of the paper-based and electronic versions of the Dermatology Life Quality Index: Evidence of equivalence. Br. J. Dermatol. (2017) 177 1306-1315. DOI: 10.1111/bjd.15314 46. Campbell N., Ali F., Finlay A.Y., Salek S.S.. Equivalence of electronic and paper-based patient-reported outcome measures. Qual. Life Res. (2015) 24 1949-1961. DOI: 10.1007/s11136-015-0937-3 47. Fritz F., Balhorn S., Riek M., Breil B., Dugas M.. Qualitative and quantitative evaluation of EHR-integrated mobile patient questionnaires regarding usability and cost-efficiency. Int. J. Med. Inform. (2012) 81 303-313. DOI: 10.1016/j.ijmedinf.2011.12.008 48. Bushnell D.M., Martin M.L., Parasuraman B.. Electronic versus paper questionnaires: A further comparison in persons with asthma. J. Asthma (2003) 40 751-762. DOI: 10.1081/JAS-120023501 49. Boettcher J., Magnusson K., Marklund A., Berglund E., Blomdahl R., Braun U., Delin L., Lundén C., Sjöblom K., Sommer D.. Adding a smartphone app to internet-based self-help for social anxiety: A randomized controlled trial. Comput. Hum. Behav. (2018) 87 98-108. DOI: 10.1016/j.chb.2018.04.052 50. Bush N.E., Skopp N., Smolenski D., Crumpton R., Fairall J.. Behavioral screening measures delivered with a smartphone app: Psychometric properties and user preference. J. Nerv. Ment. Dis. (2013) 201 991-995. DOI: 10.1097/NMD.0000000000000039 51. Zhen L., Wang G., Xu G., Xiao L., Feng L., Chen X., Liu M., Zhu X.. Evaluation of the Paper and Smartphone Versions of the Quick Inventory of Depressive Symptomatology-Self-Report (QIDS-SR16) and the Patient Health Questionnaire-9 (PHQ-9) in Depressed Patients in China. Neuropsychiatr. Dis. Treat. (2020) 16 993-1001. DOI: 10.2147/NDT.S241766 52. Azevedo B.R., Oliveira C.B., Araujo G.M.D., Silva F.G., Damato T.M., Pinto R.Z., Christofaro D.G.D.. Is There Equivalence Between the Electronic and Paper Version of the Questionnaires for Assessment of Patients with Chronic Low Back Pain?. Spine (2020) 45 E329-E335. DOI: 10.1097/BRS.0000000000003281 53. Yu S.C., Yu M.N.. Comparison of Internet-based and paper-based questionnaires in Taiwan using multisample invariance approach. Cyberpsychol. Behav. (2007) 10 501-507. DOI: 10.1089/cpb.2007.9998 54. Meirte J., Hellemans N., Anthonissen M., Denteneer L., Maertens K., Moortgat P., Van Daele U.. Benefits and Disadvantages of Electronic Patient-reported Outcome Measures: Systematic Review. JMIR Perioper Med. (2020) 3 e15588. DOI: 10.2196/15588 55. Walter S.D., Eliasziw M., Donner A.. Sample size and optimal designs for reliability studies. Stat. Med. (1998) 17 101-110. DOI: 10.1002/(SICI)1097-0258(19980115)17:1<101::AID-SIM727>3.0.CO;2-E 56. De Vet H.C., Terwee C.B., Mokkink L.B., Knol D.L.. Measurement in Medicine: A Practical Guide (2011) 57. Shima S., Kitagawa Y., Kitamura T., Fujinawa A., Watanabe Y.. Poststroke depression. Gen. Hosp. Psychiatry (1994) 16 286-289. DOI: 10.1016/0163-8343(94)90008-6 58. Storheim K., Brox J.I., Løchting I., Werner E.L., Grotle M.. Cross-cultural adaptation and validation of the Norwegian version of the Core Outcome Measures Index for low back pain. Eur. Spine J (2012) 21 2539-2549. DOI: 10.1007/s00586-012-2393-x 59. Terwee C.B., Bot S.D., de Boer M.R., van der Windt D.A., Knol D.L., Dekker J., Bouter L.M., de Vet H.C.. Quality criteria were proposed for measurement properties of health status questionnaires. J. Clin. Epidemiol. (2007) 60 34-42. DOI: 10.1016/j.jclinepi.2006.03.012 60. Dunn T.J., Baguley T., Brunsden V.. From alpha to omega: A practical solution to the pervasive problem of internal consistency estimation. Br. J. Psychol. (2014) 105 399-412. DOI: 10.1111/bjop.12046 61. Gueorguieva R., Krystal J.H.. Move over ANOVA: Progress in analyzing repeated-measures data and its reflection in papers published in the Archives of General Psychiatry. Arch. Gen. Psychiatry (2004) 61 310-317. DOI: 10.1001/archpsyc.61.3.310 62. Koo T.K., Li M.Y.. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. J. Chiropr. Med. (2016) 15 155-163. DOI: 10.1016/j.jcm.2016.02.012 63. Qin S., Nelson L., McLeod L., Eremenco S., Coons S.J.. Assessing test-retest reliability of patient-reported outcome measures using intraclass correlation coefficients: Recommendations for selecting and documenting the analytical formula. Qual. Life Res. (2019) 28 1029-1033. DOI: 10.1007/s11136-018-2076-0 64. Ogles B.M., France C.R., Lunnen K.M., Bell M.T., Goldfarb M.. Computerized depression screening and awareness. Community Ment. Health J. (1998) 34 27-38. DOI: 10.1023/A:1018760128239 65. Quadri N., Wild D., Skerritt B., Muehlhausen W., O’Donohoe P.. A literature review of the variance in interval length between administrations for assessment of test retest reliability and equivalence of pro measures. Value Health (2013) 16 A40-A41. DOI: 10.1016/j.jval.2013.03.230 66. Buchanan T., Smith J.L.. Using the Internet for psychological research: Personality testing on the World Wide Web. Br. J. Psychol. (1999) 90 125-144. DOI: 10.1348/000712699161189
--- title: Senescent Secretome of Blind Mole Rat Spalax Inhibits Malignant Behavior of Human Breast Cancer Cells Triggering Bystander Senescence and Targeting Inflammatory Response authors: - Amani Odeh - Hossam Eddini - Lujain Shawasha - Anastasia Chaban - Aaron Avivi - Imad Shams - Irena Manov journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049022 doi: 10.3390/ijms24065132 license: CC BY 4.0 --- # Senescent Secretome of Blind Mole Rat Spalax Inhibits Malignant Behavior of Human Breast Cancer Cells Triggering Bystander Senescence and Targeting Inflammatory Response ## Abstract Subterranean blind mole rat, Spalax, has developed strategies to withstand cancer by maintaining genome stability and suppressing the inflammatory response. Spalax cells undergo senescence without the acquisition of senescence-associated secretory phenotype (SASP) in its canonical form, namely, it lacks the main inflammatory mediators. Since senescence can propagate through paracrine factors, we hypothesize that conditioned medium (CM) from senescent Spalax fibroblasts can transmit the senescent phenotype to cancer cells without inducing an inflammatory response, thereby suppressing malignant behavior. To address this issue, we investigated the effect of CMs of Spalax senescent fibroblasts on the proliferation, migration, and secretory profile in MDA-MB-231 and MCF-7 human breast cancer cells. The results suggest that Spalax CM induced senescence in cancer cells, as evidenced by increased senescence-associated beta-galactosidase (SA-β-Gal) activity, growth suppression and overexpression of senescence-related p53/p21 genes. Contemporaneously, Spalax CM suppressed the secretion of the main inflammatory factors in cancer cells and decreased their migration. In contrast, human CM, while causing a slight increase in SA-β-Gal activity in MDA-MB-231 cells, did not decrease proliferation, inflammatory response, and cancer cell migration. Dysregulation of IL-1α under the influence of Spalax CM, especially the decrease in the level of membrane-bound IL1-α, plays an important role in suppressing inflammatory secretion in cancer cells, which in turn leads to inhibition of cancer cell migration. Overcoming of SASP in tumor cells in response to paracrine factors of senescent microenvironment or anti-cancer drugs represents a promising senotherapeutic strategy in cancer treatment. ## 1. Introduction Cellular senescence is a stress response program triggered by a variety of cellular stressors, including telomere shortening, DNA damaging agents, oxidative stress, and the activation of oncogenes. Such cells halt division, which blocks the transmission of mutations to the next generation of cells limiting the propagation of potential oncogenic cells [1,2]. Cellular senescence is accompanied by the secretion of various molecules such as cytokines, chemokines, growth factors, matrix metalloproteinases (MMPs), collectively referred to as the senescence-associated secretory phenotype (SASP) [3,4]. The SASP factors can affect the surrounding tissues, and this effect can either promote or suppress tumor development. The modulating effect of SASP on cancer progression depends on the composition of the aging secretome (context dependence), the origin of the tumor, the state of the surrounding tissue, and host factors (such as biological age, hereditary predisposition, metabolism, and immunity) [5,6,7]. Both cancer and senescent cells share many common features such as genomic instability, epigenetic changes, telomere attrition (but unlike senescent cells, cancer cells overcome cell cycle arrest by activating telomerase) [8]. The secretion of cancer cells, depending on the type of tumor, contains a set of factors that are similar to the SASP of senescent cells (IL6, IL1-ß, TGF-β, VEGF, MMPs and others), which, in turn, activate normal stromal cells (fibroblasts, macrophages, etc.), thereby enhancing positive feedback loop and favors malignancy. In addition, oxidative stress and hypoxia in the niche of cancer progression can cause senescence and SASP in noncancerous cells. Thus, the outcome of a tumor is determined by the interaction of cancer cells and stromal cells of the cancer microenvironment, as well as the mutual influence of their paracrine secretomes. Normal fibroblasts are the main stromal compartment in the niche of cancer, and therefore their physiological state determines to a large extent whether or not the host microenvironment supports the cancer progression. Our previous studies have demonstrated the unique ability of the stromal fibroblasts of long-lived, cancer-resistant underground rodents, Spalax, to inhibit growth of various cancer cells either via direct cell-to-cell contact or paracrine secretion [9]. Different molecular design strategies that evolved in Spalax to cope with its extreme habitat (hypoxia and hypercapnia) were previously investigated in our laboratory [10,11]. Along with adaptation to environmental stress, Spalax has also acquired the ability to resist cancer. Spalax’s efficient DNA repair mechanisms maintain genome stability, which allows rapid repair of DNA damage [12]. In addition, among molecular strategies against cancer we found reduced capacity of adipose-derived stem cells to penetrate tumor, which leads to suppression of intratumoral abnormal angiogenesis [13]. The latest report from our laboratory showed that the secretory phenotypes of senescent Spalax fibroblasts and senescent tissues are devoid of major inflammatory factors. Spalax fibroblasts undergo replicative and etoposide-induced senescence, showing arrest of proliferation, SA-β-Gal-positive staining, and increased expression of p21 and p53; however, the secretion of these cells lacked the expression of canonical inflammatory factors such as interleukins IL6, IL8, IL1α, SerpinB2, GROα, ICAM-1 [14]. Senescent cells communicate with their environment via transmitting senescence phenotype to neighboring cells either normal or malignant through paracrine signaling of SASP factors [15,16,17]. Therefore, SASP determines how the senescent cells influence the tissue microenvironment: [1] SASP can induce pro-inflammatory microenvironment that attracts immune cells which are responsible for the clearance of damaged and senescent cells (physiological response that occurs, for example, during wound healing); [2] In aging tissues, immune processes are weakened, therefore, senescent cells are accumulated in tissues and able to transmit senescence to adjacent cells, giving rise to the release of inflammatory mediators. Persistent chronic inflammation (“inflammaging”) predisposes the development of destructive diseases associated with aging and supports the proliferation and invasion of precancerous and malignant lesions. Since Spalax resists oncogenic stimuli and maintains a non-permissive tumor environment, while Spalax’s cellular senescence is not accompanied by the secretion of major inflammatory mediators, we hypothesized that such a non-canonical secretome could sensitize cancer cells to senesce without triggering inflammatory SASP. Thus, the paracrine network that is activated in recipient cancer cells (bystander senescent cells) can be equivalent to that in the donor cells of senescent fibroblasts of Spalax. The present study aims to answer the following questions: (i) whether Spalax senescent secretome can induce senescence in cancer cells, lacking the acquisition of pronounced inflammatory SASP? ( ii) whether non-canonical SASP of Spalax senescent fibroblasts can influence cancer cells throughout the reduction of their invasive and migratory behavior? ( iii) to explore the role of IL-1α/NF-κB pathway in the regulation of inflammatory secretion and malignant behavior in cancer cells exposed to senescent Spalax secretome. To address these issues, we investigated the effects of conditioned medium (CM) of Spalax and human senescent fibroblasts on the proliferation, migration, and inflammatory response of MDA-MB-231 and MCF-7 human breast cancer cells. MDA-MB-231 is an extremely aggressive cell line producing a large number of inflammatory mediators that support high growth rates and metastatic capacity [18]. Therefore, the search for mechanisms that ensure the suppression of the inflammatory response during therapy-induced senescence may be a promising strategy for the treatment of such aggressive tumors. ## 2.1. Spalax Senescent Secretome Decreases Proliferation and Induces Senescence-Associated Beta-Galactosidase (SA-β-Gal) Activity in MDA-MB-231 Cells To investigate whether CM collected from Spalax, and human senescent fibroblasts could affect growth and induce senescence in MDA-MB-231 cells, the population doubling rate, cell cycle analysis and SA-β-Gal staining were performed (Figure 1). As demonstrated, a marked reduction in the proliferation of cancer cells treated with Spalax CM was observed after 48 h. Following 96 h of exposure, the cancer cell growth decreased 1.82 times compared to untreated cells (p ≤ 0.05). Spalax CM-treated MDA-MB-231 cells multiplied 16-fold within 96 h, while untreated and human CM-treated cells multiplied 29 and 28- fold, respectively (Figure 1A). Cell cycle analysis demonstrates induction of S and G2/M phase arrest in MDA-MB-231cells when treated with Spalax CM. No changes vs. control were found in MDA-MB-231 cell cycle distribution upon treatment of human CM (Figure 1B). Next, we evaluated whether MDA-MB-231 exposed to the senescent Spalax secretome undergo senescence. The level of SA-β-Gal activity, a well-known senescence marker, was significantly increased in cancer cells treated with Spalax CM when compared to control and human CM (p ≤ 0.001; Figure 1C,D). Subsequently, we questioned whether the senescence phenotype preserved in MDA-MB-231 cells during long-term cultivation under the influence of Spalax senescent secretome. As shown in Figure S1, Spalax CM significantly inhibited cancer cell proliferation (Figure S1A), while the vast majority of cells retain the senescent phenotype after 10-day exposure (Figure S1B,C) ## 2.2. Spalax CM Increases the Expression of Senescence Related p21 and p53 Genes, Enhances Nuclear Accumulation of p53 and Reduces the Levels of Spontaneous Double Strand Breaks in MDA-MB-231 Cells The results presented in Figure 2A,B demonstrate that Spalax CM upregulates the expression of p21 and p53 mRNA levels in MDA-MB-231 cells. Human CM induces a moderate increase in the p53 mRNA level in MDA-MB-231 cells. Figure S2 shows a clear accumulation of p53 in the nucleus of cancer cells treated with Spalax senescent secretome. MDA-MB-231 develops spontaneous DNA double strand breaks (DSB) as evidenced by the phosphorylation of γH2AX, a well-known marker of DSBs. Since Spalax fibroblasts developed a high DNA repair capacity [12], it was important to check whether CM collected from Spalax senescent fibroblasts could influence DSBs levels in MDA-MB-231 cancer cells. As shown in Figure 2C,D, CM harvested from human senescent fibroblasts increased the level of γH2AX foci in MDA-MB-231 cells, the effect being manifested already after 24 h of treatment and persisting after 96 h. Treatment with Spalax CM for 24 h did not alter the DSB level in MDA-MB-231 cells, whereas it decreased significantly after 96 h exposure (p ≤ 0.01) compared to control. Conversely, treatment of MDA-MB-231 cells with CM of senescent human fibroblasts resulted in an increase in DSB levels in cancer cells, presumably as a consequence of the large amount of SASP factors that are normally present in the CM of senescent fibroblasts in most mammals, excluding of the previously described phenomenon of “non-canonical SASP” in Spalax, whose cellular senescence and body aging is not accompanied by the secretion of major inflammatory factors [14]. Figure 1: Influence of Spalax and human CM on proliferation rate and on the level of SA-β-Gal activity in MDA-MB-231 cells. (A) 25 × 103 cells were seeded in 12-well plates under either Spalax (Sp) or human (Hu) CM, and counted after 24, 48, 72- and 96- hours. Values represent the averages of three independent experiments (n = 3) in triplicates. p ≤ 0.05 differences between treated MDA-MB-231 with Spalax CM and control at 96 h (B) Flow cytometry analysis showing the percentage of cells in cell cycle stages. Values represent the average of two independent experiments in triplicate. (C) SA-β-Gal staining representative images: MDA-MB-231 cells were treated with Spalax and human CMs for 96 h. Bars, 100 μm (D) Percentage of SA-β-Gal-positive cells calculated from at least 300 cells in four independent fields for each biological repeat (n = 3) in triplicates (Spalax CMs were collected from senescent cells of three independent individuals). Human CMs were obtained from the same cells thawed at different times and at different passages (# 45–52; passages when human fibroblasts became senescent); *** p < 0.001 differences between control (untreated) and treated with Spalax CM.* Figure 2: Effects of Spalax and human CMs on p53/p21 senescence markers and on the spontaneous DSBs in MDA-MB231 cells. (A,B) p21 and p53 mRNA levels were quantified in MDA-MB231 under exposure to Spalax (Sp)/human (Hu) CMs by using relative qRT-PCR;. Experiments were performed in triplicates and repeated 3 times. Data are presented as mean ± SE p ≤ 0.05; * p ≤ 0.001, differences between treated MDA-MB-231 cells vs. untreated; “ns”, not significant p value. (C) MDA-MB231cells were treated with Spalax and human CM for 24- or 96 h; thereafter, cells were fixed and stained with anti- γ-H2AX antibody and counterstained with DAPI. Representative images of γ-H2AX (pSer13) foci in the nuclei of MDA-MB-231 untreated and treated for 96 h are shown. Bars, 10 μm. (D) The images were used for quantification using FociCounter software (minimum 250 nuclei per sample were analyzed). Data is presented as mean ± SD of three independent experiments (n = 3), p ≤ 0.01 differences between treated MDA-MB-231 with Spalax CM and control at 96 h. ** p ≤ 0.01 differences between treated MDA-MB-231 with human CM and control at 24 h. # means the number.* ## 2.3. Spalax Senescent Secretome Reduces MDA-MB-231 Cancer Cells Migration We were interested to test whether the senescent secretome of Spalax affects cancer cells migration. For these purposes, ‘scratch’ and Transwell® migration assays were performed (Figure 3). Spalax CM diminished the scratch wound closure of cancer cells in a time dependent manner compared with control (untreated cells), whereas human CM moderately increased wound closure at 12 h and resulted in nearly complete closure at 24 h thereby enhancing cancer cell migration, once more at the expense of inflammatory factors that are present in the human CM (Figure 3A,B). Next, the effects of Spalax/human CMs on cancer cell migration were examined using Transwell® migration chamber assay (Figure 3C,D). The results showed a significant decrease in the number of migrated cancer cells in the presence of Spalax CM compared to control cells, while human CM showed no effect on the ability of MDA-MB-231 cells to migrate. ## 2.4. CM of Spalax Senescent Fibroblasts Reduces the Levels of Inflammatory Factors in Secretome of MDA-MB-231 Cells Based on the data presented above, the senescent secretome of Spalax fibroblasts causes senescence in MDA-MB-231 cells and decreases migration of cancer cells. Then we thought whether the secretory phenotype of cancer cells, which provides a high growth rate and metastatic capacity in an autocrine manner, changes under exposure to Spalax senescent secretome. In theory, such cells, while they senesce, should produce SASP factors even more intensively, and this is the main obstacle to therapy-induced senescence. However, as demonstrated above, the paracrine secretion of senescent Spalax cells, which triggers the senescence program in cancer cells, simultaneously reduces the level of DNA damage that, in turn, could inhibit NF-κB activation and the inflammatory response in these cells. Indeed, exposure of MDA-MB-231 cells to Spalax CM reduced the expression of mRNA level of IL6, IL8, COX-2, ICAM-1 and GRO-α, while treatment with human secretome resulted in upregulation of mRNA levels of these inflammatory mediators (Figure 4A). Next, we investigated whether secretion of inflammatory proteins IL6 and IL8, which are highly expressed in MDA-MB-231 cells, may be affected by Spalax CM. As demonstrated (Figure 4B), Spalax senescent secretome reduced secretion of IL6 and IL8, while human CM increased or did not change secretion of IL6 and IL8, respectively, when compared with control. Next, we examined whether the inhibition of the inflammatory response in MDA-MB-231 cells persists after prolonged exposure to Spalax senescent secretome. Since cancer cells are known to be able to emerge from proliferation arrest [19], we first tested the expression levels of p53 and p21 in MDA-MB-231 cells after long-term exposure to Spalax CM. Figure S3A shows increased expression of p21/p53 mRNA in MDA-MB-231 cells treated with Spalax CM for 10 days. This finding is consistent with reduced cell proliferation and positive β-Gal staining shown in Figure S1 and confirms that cells remain in a senescent state. In parallel, the expression of SASP factors remains suppressed in these cells (Figure S3B). ## 2.5. Spalax CM Suppresses the Expression/Phosphorylation of a Large Number of Signaling Proteins Associated with Inflammation, Proliferation and Migration in MDA-MB-231 Cells The transcription factor NF-κB regulates a wide range of processes associated with inflammatory responses and cell proliferation. To reveal the NF-κB protein expression/phosphorylation profiles and status of related signaling molecules in cells treated with Spalax CM, NF-κB network antibody array was applied (Figure 5A,B). Of the 215 proteins analyzed, 123 were significantly suppressed, 84 did not undergo significant changes, and only six were enhanced in the treated cells compared to control. Among the downregulated proteins, NF-κB (p65; p$\frac{105}{50}$), IkB α/β, MAPKp38, CBP/p300 are widely represented (which are associated with the regulation and the secretion of inflammatory mediators). The level of expression and phosphorylation of phospholipase C-gamma 1 (PLCG1), which is involved in cell migration and metastasis of cancer cells, significantly decreased. In addition, protein kinase B (PKB or AKT) and protein kinase C (PKCA and PKCB), which are responsible for cell proliferation, migration and apoptosis, were also downregulated. Volcano plot (Figure 5B) displays the statistical significance of the differences (relative to the magnitude of difference) for every individual protein in the groups of MDA-MB-231 cells treated with Spalax CM versus untreated. Most NF-κB related proteins are significantly downregulated in treated cells versus control. Data points closer to “0” representing proteins that have similar expression levels (empty points). Western blots data confirm the suppression of phosphorylation of two of the most important players in the inflammatory response, namely p65 and p38, in MDA-MB-231 cells exposed to Spalax CM (Figure 6C,D and Figure S4). ## 2.6. The Role of IL1α in Suppressing the Inflammatory Response and Inhibition of MDA-MB-231 Cell Migration under the Influence of the Senescent Spalax Secretome Treatment with Spalax CM resulted in redistribution of IL1-α compared to control, where IL1-α was mainly concentrated in the cytoplasm of cells near the nuclei (Figure 6A). Treatment with human CM did not lead to visible changes in the distribution of IL1-α within cells (arrows indicate the presence of IL1-α on the membrane). Flow cytometry confirmed a decrease of surface membrane-bound IL1-α in MDA-MB-231 treated with Spalax CM compared with non-treated cells or treated with human CM (Figure 6B). It is noteworthy that the effect of Spalax CM was weakened by the simultaneous treatment with the recombinant human IL-1α protein (IL-1α agonist). Figure 5: (A) Heatmap representing differences in the levels of expression/phosphorylation of NF-κB—related proteins in MDA-MB-231 cells exposed to Spalax CM (Treated) compared to control; (B) Volcano plot representing the log2 FC (treatment/control) versus −log10 p value (cut-off range—p value ≤ 0.05). Dots represent individual proteins. We then examined whether a decrease in membrane-bound IL-1α affects NF-κB p65 and p38 phosphorylation in MDA-MB-231 cells exposed to Spalax CM. Western blot analysis shows that treatment by Spalax CM resulted in down-regulation of phosphorylation of both NF-κB-p65 and p38, but co-treatment with recombinant IL-1α agonist restores the level of NF-κB-p65 and p38 phosphorylation (Figure 6C,D and Figure S4). Treatment with human CM increased the level of phosphorylation of these proteins. Figure 6: Spalax CM suppresses membrane-bound IL1-α and inhibits phosphorylation of NF-κB and p38 in MDA-MB 231 cells. (A) Representative images showing MDA-MB231 untreated and treated with Spalax (Sp) and human (Hum) CMs. A magnified view of the enclosed regions is shown on a low panel, highlighting the differences in IL1-α distribution within the cells between treated with human and Spalax CMs for 24 h and untreated. Treated/untreated cells were stained with IL1α–FITC and NF-κB-p65 antibodies; nuclei were counterstained with DAPI. Arrows indicate the presence of IL1-α on the membrane (B) Representative images of flow cytometry showing differences in IL1α surface membrane in MDA-MB-231 untreated, treated with Spalax CM, human CM, with or without recombinant human IL-1α protein. (C) Representative western blot analysis demonstrating the phosphorylation of p65 (Ser536), p38 in MDA-MB231 treated with recombinant IL-1α, Spalax CM (SpCM), Spalax CM+IL-1α (SpCM+IL-1α), and human CM (Hu CM). (D) Densitometry quantification of western blots. Western blots and densitometries of additional biological repeats (MDA-MB-231 treated with Spalax CMs generated from fibroblasts of different Spalax individuals) are presented in Figure S4. ## 2.7. The IL1α Agonist Promotes Cancer Cell Migration Weakened by Spalax CM through SASP Activation We then analyzed how dysregulation of IL-1α induced by Spalax CM affects cancer cell migration. As demonstrated in Figure 3, Spalax CM suppressed wound healing and reduced the MDA-MB-231 migration. However, co-treatment of Spalax CM with recombinant IL-1α agonist diminished the effect of Spalax CM and restored MDA-MB-231 cancer cells migration (Figure 7A,B). Neither treatment by human CM nor human CM plus IL-1α agonist affected cancer cell migration. Since the IL-1α agonist activates NF-κB p65, a major regulator of the inflammatory response, we then examined the levels of SASP expression/secretion in MDA-MB-231 cells simultaneously exposed to Spalax CM and IL-1α. As shown (Figure 7C), when MDA-MB-231 cells exposed to Spalax CMs were simultaneously treated with IL1α agonist, the levels of IL6 mRNA expression and IL6 protein secretion were increased. ## 2.8. Spalax Senescent Secretome Induces Senescence of MCF-7 Breast Cancer Cells, Inhibits the Secretion of Inflammatory Mediators and Hold Back the Migration of Cancer Cells To prove that the observed anti-cancer effect is not limited to MDA-MB-231 cells, we also examined the acquisition of senescence in MCF-7 cells exposed to Spalax CM and wondered whether MCF-7 cell migration would be slowed down by the paracrine factors of Spalax senescent secretome. As demonstrated in Figure 8A, the level of SA-β-Gal activity was significantly increased in MCF-7 cancer cells treated with Spalax CM for 4 days when compared to control. In parallel, Spalax CM significantly reduced the migratory potential of MCF-7 cells, which resulted in slower wound closure by cancer cells exposed to Spalax CM compared to untreated cells (Figure 8B). As in the MDA-MB-231 study, Spalax CM reduced the mRNA expression level of a number of inflammatory mediators in MCF-7 cells (Figure 8C). ## 3. Discussion We demonstrated here that the senescent secretome of Spalax fibroblasts, which we previously reported, lacking the major inflammatory SASP factors [14], is capable of transmitting senescence phenotype to cancer cells. Having become senescent, breast cancer cells MDA-MB-231 and MCF-7 replicate the SASP pattern of senescent Spalax cells, namely, they demonstrate a decrease in the secretion of inflammatory mediators, which, in turn, leads to the reduction of cancer cell migration. Accumulating evidence suggests that the senescent phenotype can be transmitted from one cell to another via direct contact or via paracrine factors including extracellular vesicles [15,16,17,20]. Since cellular senescence is usually accompanied by an increased secretion of inflammatory factors, the recipient cell also becomes SASP positive. SASP, in its canonical pro-inflammatory and pro-growth design, is a barrier to therapy-induced senescence, which is a promising new cancer treatment strategy [21]: (i) cancer cells, in which senescence is induced, increase the production of their own inflammatory factors by triggering the SASP genes. Such cancer cells can transmit aging to neighboring cells, both neoplastic and non-malignant stromal cells; (ii) senescent stromal cells (the vast majority of which are fibroblasts, but also macrophages, immune cells, mesenchymal stem cells), in turn, release SASP factors. This reciprocal paracrine exchange gradually creates a tumor permissive microenvironment accelerating cancer cells proliferation, migration and stemness [22]. In this context, the tumor acquires resistance to anti-cancer drugs [23]. The deleterious effects of SASP can be abrogated by a senolytic strategy aimed at removing senescent cells. However, since senescent cells are involved in embryogenesis and wound healing, the widespread removal of senescent cells can disrupt these processes [21,24,25]. Unlike senolytics, the senomorphic strategy is designed to suppress SASP while maintaining cell cycle cessation in tumor cells [reviewed in [26]]. The composition of secreted molecules is extremely diverse and depends on the type of senescent cells, the type of stimulus that caused senescence, time passed since senescence initiation, the microenvironment, etc. [ 27,28]. However, a number of molecules are constant constituents among the various canonical SASPs, namely IL-1 α/β, IL-6, IL-8, GROα/β, MMP-1,3, MMP-10 ICAM-1, PAI-1 and IGFBPs [5]. Our recent research has shown that there are exceptions to this rule [14]. The secretory phenotype of replicative or etoposide-induced senescent fibroblasts of Spalax either did not contain many of the pro-inflammatory SASP factors (IL6, IL8, IL1α, GROα, SerpinB2, Cox2), or their expression reduced compared to young cells (ICAM-1). Decreased expression of pro-inflammatory SASP genes have also been found in senescent Spalax tissues. This SASP, devoid of major inflammatory factors, has been termed a “non-canonical SASP” [14]. In this report we demonstrated that Senescent secretome of Spalax induced senescence in MDA-MB-231 and MCF-7, as evidenced by increased SA- β-Gal activity, growth suppression and overexpression of senescence-related p53/p21 genes. In parallel, we demonstrated accumulation of p53 in the nuclei of MDA-MB-231 cells exposed to Spalax CM, which probably reflects the release of p53 from targeted degradation and, as a result, slowing down/stopping cell division [29]. In addition, Spalax CM suppressed the level of spontaneous DSB in MDA-MB-231. Cellular senescence is typically driven by a persistent DNA damage response [30]. Double strand breaks in cancer cells occur spontaneously, a phenomenon termed ‘self-inflicted DNA DSBs’ which sustain tumorigenesis [31] and maintain the expression/secretion of main inflammatory factors (IL6, IL8, GROα, ICAM-1, Cox-2). Thus, cancer cells acquired a senescent phenotype along with the non-canonical SASP of senescent Spalax fibroblasts via paracrine factors. Thus, if in the previous report we demonstrated that the cessation of division of senescent Spalax cells was not accompanied by the secretion of the main SASP factors [14], here we received clear evidence that this type of senescence can be transmitted to the recipient cancer cells, which leads to a decrease in their aggressive behavior. Importantly, long-term exposure of MDA-MB-231 cells to Spalax senescent secretome did not reverse senescence and did not restore inflammatory secretion in cancer cells. An important aspect in tumor progression is the migration of cancer cells from the primary location to secondary distant sites. Paracrine signaling of senescent stromal cells, which are abundant in benign tumors and are present in cancer [32,33], affects the malignant properties of tumor cells, mainly enhancing their proliferation and invasiveness [34,35,36]. To date, there is a growing body of research showing pro-tumorigenic role of SASP [37,38,39]. Conversely, information that the paracrine secretion of senescent cells can inhibit tumor development is very scarce and contradictory. In this context, studies of the senescence of hepatic stellate cells and its role in creating tumor suppressive microenvironment are of particular interest. Liver fibrosis that precedes the development of cirrhosis and liver cancer mainly arise due to activated stellate liver cells. However, when liver stellate cells undergo replicative or induced senescence, they exhibit a less fibrogenic secretory phenotype and are prone to spontaneous apoptosis [40,41]. The p53-expressing senescent stellate cells via SASP distort the polarization of macrophages toward the M1-state, inhibiting the tumor [42]. Interleukin-10 has been demonstrated to induce hepatic stellate cell senescence and alleviate liver fibrosis via STAT3-p53 pathway [43]. Lujambio et al. demonstrated that macrophages previously exposed to CM of senescent stellate cells did not affect the proliferation of precancerous hepatoblastoma progenitor cells in co-culture, while pre-exposure of macrophages with CM of proliferative stellate cells significantly increased the growth of precancerous cells [42]. Thus, there is a mechanism in the liver that prevents the development of fibrosis and cancer due to the induction of senescence in hepatic stellate cells, the secretion of which does not possess pro-oncogenic properties. Interestingly, analysis of genes driving cellular senescence and SASP in mammals showed a close overlap with anti-longevity genes rather than genes that promote longevity [44]. In this context, the evolution of longevity should include the selection of mechanisms that provide SASP inhibition along with a non-permissive cancer microenvironment. In our previous reports, we have shown that Spalax resists cancer and maintains a tumor protective microenvironment [9,13]. Suppression of the inflammatory response in senescent cells and tissues seems to be one of the main antitumor strategies of Spalax [14]. Targeting cancer cell migration is one of the main approaches to reducing tumor progression. Here, we demonstrate that senescent secretome of Spalax fibroblasts induced senescence in breast carcinoma cells MDA-MB-231 and MCF-7 while inhibiting migration of cancer cells. MDA-MB-231 cells are extremely aggressive metastatic malignant cells with a high level of secretion of their own inflammatory growth factors [18]. The induction of senescence in such cells, if accompanied by the establishment of the canonical SASP, should enhance the secretion of inflammatory factors. However, we received the opposite effect, namely the secretion of SASP was repressed. Exposure to Spalax CM significantly reduced IL-6 gene expression and IL-6 protein secretion in MDA-MB-231 cells. Another breast cancer cell line, MCF-7, also reduced malignancy under the influence of paracrine factors of senescent Spalax cells. Namely, MCF-7 cells acquired senescence and reduced the ability to migrate while reducing the expression of IL-6. A pleiotropic inflammatory cytokine IL-6 is considered a key factor in the growth of malignant neoplasms and metastases [45]. IL-6 facilitates epithelial to mesenchymal transition in lung cancer increasing vimentin expression [46], and in colorectal cancer via integrin β6 upregulation [45]. A growing body of reports demonstrates a strong correlation between IL6 levels and the invasiveness of breast cancer [45,47]. The small soluble protein IL8, which belongs to the CXC chemokine family, is overexpressed in breast cancer, and high IL8 levels are associated with poor prognosis and invasion [48]. GROα, ICAM 1 and Cox-2 are SASP factors that are also closely associated with the progression and metastasis of breast cancer [49,50,51]. All these factors belonging to SASP were suppressed in MDA-MB-231 by senescent secretome of Spalax. Since transcription factor NF-κB controls the secretion of inflammatory mediators, we analyzed the status of expression/phosphorylation of NF-κB-related proteins in MDA-MB-231 cancer cells treated with Spalax CM. The phospho-NF-κB antibody array showed suppression of about $60\%$ of proteins, including: NF-κB-p65, NFkB-p105/p50, IkB α/β, CBP/p300, p38, all strongly involved in SASP secretion. Among the proteins engaged in migration, the suppression of the expression and phosphorylation of phospholipase C-gamma 1(PLCγ1) is of particular interest. PLCγ1 phosphorylation status was shown to be a prognostic marker of metastatic risk in patients with breast cancer [52]. Antibody microarray demonstrated decreased phosphorylation of PLCγ1 in different sites of tyrosine (1253,783 and 771). Inhibition of phosphorylation of PLCγ1 (tyrosine 771) has been shown to reduce the risk of brain metastasis in experimental breast cancer [53]. IL1-α is a key protein among SASP molecules and, despite its minor secretion, is a regulator of NF-κB activation, its nuclear translocation, and subsequent expression of major inflammatory cytokines [34,54]. IL-1α expression has been reported to be significantly increased in many types of human malignant neoplasms and involved in cancer progression and metastasis [55,56,57]. IL1α stimulates an inflammatory response, mainly as a surface membrane-bound protein, which binds to the IL-1R receptor, initiating a signaling cascade that leads to NF-κB activation, translocation to the nuclei and subsequent production of inflammatory factors [58]. IL1α/NF-κB positive feedback loop is disrupted in Spalax senescent fibroblasts as we described [14] where we found no membrane-bound IL1α in senescent Spalax cells, while IL1α was mainly centered around the nuclei. Like Spalax, IL-1α of MDA-MB-231 cells in response to Spalax CM undergoes redistribution, namely, it disappears from the membrane, which obviously disrupts signal transmission from IL1R receptor to NF-κB. Lau et al. demonstrated that breakdown of the IL-1α signaling pathway effectively suppresses SASP but does not abolish cell cycle arrest in senescent cells [59]. These authors demonstrated that genetic ablation of IL-1α decreases pancreatic cancer progression. Our findings show that Spalax senescent secretome abrogated SASP and inhibit breast cancer cell migration by targeting IL-1α. Our observations are consistent with the data of Lau et al. [ 59], as they show that inhibition of IL-1α is responsible for suppressing inflammatory secretion in senescent cells without affecting cell cycle exit. Targeting the IL-1 signaling pathway to separate SASP from cell cycle exit is currently being considered as a useful new strategy for preventing aging-related inflammaging and slowing cancer progression [60]. How to induce aging of cancer cells while suppressing the harmful effects of SASP is a question that is currently being extensively explored. In this regard, the discovery of natural, well-tuned mechanisms of suppressing the inflammatory response in Spalax fibroblasts, as well as the mechanisms of transmission of paracrine senescence, not associated with the inflammatory component of SASP, into cancer cells could significantly accelerate the development of a complex of artificial molecular inhibitors for modulating SASP both in cancer cells and in the senescent stromal cells of the cancer microenvironment. Our previous evidence suggests that inhibition of SASP in Spalax is a strategy that supports healthy aging, free of inflammation-related diseases, including cancer. In this report, we obtained evidence that Spalax microenvironment not only does not support tumor lesions, but also ‘educates’ cancer cells to suppress the secretion of inflammatory factors in themselves, thereby maintaining a non-permissive cancer microenvironment. ## 4.1. Cell Culture and Generation of Conditioned Medium (CM) Spalax dermal fibroblasts were isolated from newborns as we described earlier [9] and stored in liquid nitrogen in our laboratory prior to use in the current work. Human foreskin fibroblasts and breast cancer cells MDA-MB-231 and MCF-7 were obtained from ATCC®, Manassas, VA, USA. Spalax and human fibroblasts were grown in DMEM-F12 medium, MDA-MB-231cancer cells were grown in DMEM high glucose medium (supplemented with $10\%$ FBS, L-glutamine (2 mM) and penicillin-streptomycin (100 u/mL, 0.1 mg/mL respectively) in standard CO2 incubator. Growth media and supplements were purchased from Biological Industries (Beit Haemeq, Israel). Spalax fibroblasts were serially passaged to achieve replicative senescence as we described [14]. At passage 5–7, most cells acquired an enlarged, flattened morphology, cell division was reduced below $25\%$ of the total population, and cells showed positive SA-β-Gal staining and increased expression of senescence-associated p53 and p21. Human foreskin fibroblasts reach senescence after about 40 passages [14]. Pre-senescent human fibroblasts (passages 30–35) were stored in liquid nitrogen and thawed to bring the cells to senescence and obtain senescence secretome. To create Spalax senescent secretomes (senescent CMs), Spalax senescent cells were incubated in DMEM-F12 medium supplemented with $10\%$ FBS for 8–10 days without changing the medium. The complete supernatants were collected and centrifuged at 120× g for 5 min at room temperature to remove any cell debris. ## 4.2. Treatments of Cancer Cells with CMs and Evaluation of Senescence Spalax and human CMs harvested from senescent fibroblasts were tested for anti-cancer activity by measuring cancer cell viability using the PrestoBlue® reagent (TermoFisher Scientific, Waltham, MA USA) [9]. Hep3B (HepG2) cells were used as reference cells due to their high sensitivity to factors secreted by Spalax, so the viability reduction effect could be assessed as early as 3–4 days. MDA-MB-231 and MCF-7 cells were exposed to CMs of senescent Spalax or human fibroblasts for various periods of time depending on the experiment. CM was added in a 1:1 ratio with fresh culture medium. With prolonged exposure (more than 5 days), CM was replaced by the same CM 1:1 with fresh medium. To investigate the role of IL-1α, recombinant human IL-1α protein (IL-1α agonist, Abcam, Cambridge, United Kingdom) was added to MDA-MB-231 (50 ng/mL) along with treatment with CMs. ## 4.2.1. Senescence-Associated β-Galactosidase (SA-β-Gal) Staining Cancer cells were seeded at 2.5 × 104 cells/well in six-well plates and treated by either Spalax or Human CMs. After 96 h (or after 10-day) of treatments, SA-β-Gal activity was determined. The X-Gal stock solution was prepared by dissolving 40 mg/mL X-Gal (Invitrogen, Carlsbad, CA, USA) in dimethylformamide immediately before staining. SA-β-Gal staining solution was prepared as follows: 1 mg/mL of X-Gal stock solution was dissolved in phosphate buffered saline containing 5 mM potassium ferrocyanide, 5 mM potassium ferricyanide, 2 mM MgCl2, adjusted pH to 6.0. Cells were fixed using $0.2\%$ glutaraldehyde in PBS for 15 min at RT, washed in PBS and incubated in fresh SA-β-Gal staining solution for overnight at 37 °C. The cells were checked for development of the blue color under a light microscope. Quantitative analysis was performed using Image J software in four independent fields (in triplicates). ## 4.2.2. Population Doubling Population doubling levels of MDA-MB-231 cells were assessed as follows: 2.5 × 104 cells were plated in six-well plates and treated with CMs for 24, 48, 72, and 96 h, cells were trypsinized and counted in triplicates. For long-term treatment (10 days), MDA-MB-231 cells were seeded in 12-well plates at a density of 0.5 × 104 cells per well. Spalax CMs were added in ratio 1:1 with fresh culture medium. The cell number was counted after 5, 7 and 10 days. CM was replaced after 5 days of treatment with the same CM in a ratio of 1:1 with fresh nutrient media. ## 4.2.3. Cell Cycle Analysis MDA-MB-231 cells were seeded in six-well plates, treated with CMs or untreated (5 × 10⁵ cells for each probe), washed twice with PBS, trypsinized, and transferred to 5- mL tubes, then washed three times with PBS and centrifuged at 850× g, thereafter hypotonic buffer (Sodium citrate $0.1\%$; Triton $0.1\%$) was added to the pellet of the cells followed by propidium iodide (PI) staining (final concentration 25 μg/mL); The PI fluorescence of individual nuclei was recorded by FACSaria (Becton Dickinson, NJ, USA). A total of 10,000 events were acquired and corrected for debris and aggregates. ## 4.3.1. Scratch Assay MDA-MB-231 and MCF-7 cells were seeded into 12-well plates. When the cells formed a confluent monolayer ($80\%$), the cell cultures were scratched with a sterile 200-μL tip to form a “wound” and incubated with or without CMs. The migration of cancer cell across the scratch wound was monitored (0, 12, and 24 h). The cells were photographed under phase-contrast microscopy. Scratch wound gaps were measured using Image J software. ## 4.3.2. Transwell® Migration Assay CMs of Spalax and human fibroblasts were added to the lower chamber (1:1 with fresh DMEM-F12 medium supplemented with $10\%$ FBS); suspension of MDA-MB231cells (1 × 105) in 500 μL of DMEM containing $10\%$ FBS were added to the upper chamber. After 24 h membranes of the inserts were fixed with $2.5\%$ glutaraldehyde solution for 10 min, washed with DDW, and stained with $0.5\%$ Toluidine Blue for 5 min. Migrated MDA-MB-231 (adhered to the lower surface of the transwell membranes with an 8 μm-pore size) were photographed. Migrated cancer cells were calculated using Image J software in four independent fields. ## 4.4. Immunoblotting Following the treatments, cells were washed with ice-cold PBS, lysed in RIPA/SDS buffer containing sodium orthovanadate phosphatase inhibitor and Complete Protease Inhibitor (Roche Diagnostics GmbH Roche Applied Science, Mannheim Germany). Samples were centrifuged, and supernatants were collected. Protein concentrations were determined with Bradford Assay (Bio-Rad Laboratories, Hercules, CA, USA); equal concentrations of proteins were then electrophoresed, blotted onto nitrocellulose membrane and incubated with primary antibodies over night at 4 °C, thereafter membranes were washed and incubated with secondary antibodies for 1 h at room temperature. Protein bands were visualized by a chemiluminescence detection kit for HRP EZ-ECL (Biological Industries, Beit Haemek, Israel) using MyECL Imager (Thermo Scientific, Wohlen, Switzerland) and quantified by Quantity One® 1-D analysis software (Bio-Rad Laboratories, Hercules, CA, USA). Information about antibodies used for the western blot analysis is presented in Table S1. ## 4.5. Immunofluorescence MDA-MB-231 cells were seeded in 6-well plate on glass coverslips at ~40 k cells/well and treated with Spalax and human CM or untreated. Thereafter, cells were fixed at room temperature with chilled methanol (−20 °C) for 5 min and washed twice with ice-cold PBS. After permeabilization for 10 min with PBS containing $0.5\%$ Triton X-100, $1\%$ tween, $0.1\%$ bovine serum albumin (BSA) cells were blocked with $1\%$ in PBS, cells were processed for immunostaining with primary/secondary antibodies. To evaluate the level of DSBs cells were stained with γ-H2AX antibodies and counterstained with DAPI as we described [12]. Cells were visualized under fluorescent microscope (Leica DMi8, equipped with Leica DFC365FX camera). The images were used for quantification of the foci using FociCounter software. At least 250 nuclei from several random fields were scored. For p53 visualization [1026-1] p53 RabMAb® (primary Ab) and anti-rabbit Alexa Flour 647 (secondary Ab), were used. The preparation of cells for staining with NF-κB-p65 and IL-1α was performed in the same manner as described above. Specification of antibodies and dilutions are shown in Table S1. ## 4.6. Preparation of RNA, cDNA Total RNA from freshly washed cells were extracted using RNeasy Mini Kit (QIAGEN) following the manufacturer’s instructions. cDNA samples were synthesized using iScript™ cDNA Synthesis Kit (Bio-Rad Laboratories Life Science Group, Hercules, CA, USA). ## 4.7. Quantitative Real-Time Polymerase Chain Reaction (RT-PCR) Species-specific primers were designed for each target by using Primer3 software (Applied BioSystems, San Francisco, CA, USA) based on the published sequences. Relative quantification of gene transcription was performed by using Fast SYBR Green (Applied BioSystems, San Francisco, CA, USA), and 1 μL of cDNA generated from 50 ng total RNA. Serial dilutions of the cDNA with the highest expression level for each target gene were used to build a relative standard curve and to test amplification efficiency for each experiment. Samples were tested in triplicates. The amplification parameters were as follows: 95 °C for 20 s, followed by 40 cycles of 95 °C for 3 s and 60 °C for 30 s. To verify a single product with fixed melting temperature, melting curve protocol was applied. The quantification relied on equal amounts of total RNA used in each sample, and the reliability of this method was tested and confirmed by housekeeping genes (Figure S9). Primers are presented in Table S2. ## 4.8. Antibody Array An antibody microarray for phospho-NF-κB PNK215 (Fullmoon, Biosystems, Sunnyvale, CA, USA) was performed using protein lysates from MDA-MB-231 cells, either untreated or treated with Spalax CM. This array consists of 215 highly specific antibodies designed to profile various proteins and specific phosphorylation sites involved in signaling pathways associated with NF-κB and inflammation. Cell preparation, protein determination, dye labeling, and embedding into slides containing antibodies were performed according to the manufacturer’s instruction. To obtain spots-images, we used GenePix® Microarray Scanner (Molecular Devices, San Jose, CA, USA). Data were analyzed using Protein Array Analyzer for Image J software. ## 4.9. Enzyme-Linked Immunosorbent Assay (ELISA) Human IL-6 and IL-8 ELISA kits (R&D systems, Minneapolis, USA) were used, according to the instructions of manufacturer. Briefly, cancer cells were treated with CMs for 4 days, then cells were washed twice with PBSx1 and incubated in DMEM serum-free media for 24 h. Complete supernatant was collected, (centrifuged at 120× g for 5 min at RT) and processed for ELISA. ## 4.10. Statistics and Reproducibility Experiments were repeated at least three times, unless mentioned otherwise. We used newborn fibroblasts from three different offspring of Spalax individuals caught at various times and in different territories of Israel. The data are presented as mean ± SD. The Mann–Whitney nonparametric test was applied to test the differences between groups; $p \leq 0.05$ was considered significant. ## References 1. Campisi J., Kim S.H., Lim C.S., Rubio M.. Cellular senescence, cancer and aging: The telomere connection. Exp. Gerontol. (2001) 36 1619-1637. DOI: 10.1016/S0531-5565(01)00160-7 2. Wright W.E., Shay J.W.. Cellular senescence as a tumor-protection mechanism: The essential role of counting. Curr. Opin. Genet. Dev. (2001) 11 98-103. DOI: 10.1016/S0959-437X(00)00163-5 3. Acosta J.C., Banito A., Wuestefeld T., Georgilis A., Janich P., Morton J.P., Athineos D., Kang T.W., Lasitschka F., Andrulis M.. A complex secretory program orchestrated by the inflammasome controls paracrine senescence. Nat. Cell Biol. (2013) 15 978-990. DOI: 10.1038/ncb2784 4. Coppe J.P., Patil C.K., Rodier F., Sun Y., Munoz D.P., Goldstein J., Nelson P.S., Desprez P.Y., Campisi J.. Senescence-associated secretory phenotypes reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor suppressor. PLoS Biol. (2008) 6 2853-2868. DOI: 10.1371/journal.pbio.0060301 5. Coppe J.P., Desprez P.Y., Krtolica A., Campisi J.. The senescence-associated secretory phenotype: The dark side of tumor suppression. Annu. Rev. Pathol. (2010) 5 99-118. DOI: 10.1146/annurev-pathol-121808-102144 6. Faget D.V., Ren Q., Stewart S.A.. Unmasking senescence: Context-dependent effects of SASP in cancer. Nat. Rev. Cancer (2019) 19 439-453. DOI: 10.1038/s41568-019-0156-2 7. Wu S., Zhu W., Thompson P., Hannun Y.A.. Evaluating intrinsic and non-intrinsic cancer risk factors. Nat. Commun. (2018) 9 3490. DOI: 10.1038/s41467-018-05467-z 8. Aunan J.R., Cho W.C., Soreide K.. The Biology of Aging and Cancer: A Brief Overview of Shared and Divergent Molecular Hallmarks. Aging Dis. (2017) 8 628-642. DOI: 10.14336/AD.2017.0103 9. Manov I., Hirsh M., Iancu T.C., Malik A., Sotnichenko N., Band M., Avivi A., Shams I.. Pronounced cancer resistance in a subterranean rodent, the blind mole-rat, Spalax: In vivo and in vitro evidence. BMC Biol. (2013) 11. DOI: 10.1186/1741-7007-11-91 10. Shams I., Avivi A., Nevo E.. Hypoxic stress tolerance of the blind subterranean mole rat: Expression of erythropoietin and hypoxia-inducible factor 1 alpha. Proc. Natl. Acad. Sci. USA (2004) 101 9698-9703. DOI: 10.1073/pnas.0403540101 11. Shams I., Nevo E., Avivi A.. Erythropoietin receptor spliced forms differentially expressed in blind subterranean mole rats. FASEB J. (2005) 19 1749-1751. DOI: 10.1096/fj.05-3975fje 12. Domankevich V., Eddini H., Odeh A., Shams I.. Resistance to DNA damage and enhanced DNA repair capacity in the hypoxia-tolerant blind mole rat Spalax carmeli. J. Exp. Biol. (2018) 221 jeb174540. DOI: 10.1242/jeb.174540 13. Mamchur A., Leman E., Salah S., Avivi A., Shams I., Manov I.. Adipose-Derived Stem Cells of Blind Mole Rat Spalax Exhibit Reduced Homing Ability: Molecular Mechanisms and Potential Role in Cancer Suppression. Stem. Cells (2018) 36 1630-1642. DOI: 10.1002/stem.2884 14. Odeh A., Dronina M., Domankevich V., Shams I., Manov I.. Downregulation of the inflammatory network in senescent fibroblasts and aging tissues of the long-lived and cancer-resistant subterranean wild rodent, Spalax. Aging Cell (2020) 19 e13045. DOI: 10.1111/acel.13045 15. Nelson G., Wordsworth J., Wang C., Jurk D., Lawless C., Martin-Ruiz C., von Zglinicki T.. A senescent cell bystander effect: Senescence-induced senescence. Aging Cell (2012) 11 345-349. DOI: 10.1111/j.1474-9726.2012.00795.x 16. Griukova A., Deryabin P., Shatrova A., Burova E., Severino V., Farina A., Nikolsky N., Borodkina A.. Molecular basis of senescence transmitting in the population of human endometrial stromal cells. Aging (Albany NY) (2019) 11 9912-9931. DOI: 10.18632/aging.102441 17. Sapega O., Mikyskova R., Bieblova J., Mrazkova B., Hodny Z., Reinis M.. Distinct phenotypes and “bystander” effects of senescent tumour cells induced by docetaxel or immunomodulatory cytokines. Int. J. Oncol. (2018) 53 1997-2009. DOI: 10.3892/ijo.2018.4553 18. Parekh A., Das D., Das S., Dhara S., Biswas K., Mandal M., Das S.. Bioimpedimetric analysis in conjunction with growth dynamics to differentiate aggressiveness of cancer cells. Sci. Rep. (2018) 8 783. DOI: 10.1038/s41598-017-18965-9 19. Saleh T., Tyutyunyk-Massey L., Murray G.F., Alotaibi M.R., Kawale A.S., Elsayed Z., Henderson S.C., Yakovlev V., Elmore L.W., Toor A.. Tumor cell escape from therapy-induced senescence. Biochem. Pharmacol. (2019) 162 202-212. DOI: 10.1016/j.bcp.2018.12.013 20. Steenbeek S.C., Pham T.V., de Ligt J., Zomer A., Knol J.C., Piersma S.R., Schelfhorst T., Huisjes R., Schiffelers R.M., Cuppen E.. Cancer cells copy migratory behavior and exchange signaling networks via extracellular vesicles. EMBO J. (2018) 37 e98357. DOI: 10.15252/embj.201798357 21. Myrianthopoulos V., Evangelou K., Vasileiou P.V.S., Cooks T., Vassilakopoulos T.P., Pangalis G.A., Kouloukoussa M., Kittas C., Georgakilas A.G., Gorgoulis V.G.. Senescence and senotherapeutics: A new field in cancer therapy. Pharmacol. Ther. (2019) 193 31-49. DOI: 10.1016/j.pharmthera.2018.08.006 22. Gonzalez-Meljem J.M., Apps J.R., Fraser H.C., Martinez-Barbera J.P.. Paracrine roles of cellular senescence in promoting tumourigenesis. Br. J. Cancer (2018) 118 1283-1288. DOI: 10.1038/s41416-018-0066-1 23. Schosserer M., Grillari J., Breitenbach M.. The Dual Role of Cellular Senescence in Developing Tumors and Their Response to Cancer Therapy. Front. Oncol. (2017) 7 278. DOI: 10.3389/fonc.2017.00278 24. Gal H., Lysenko M., Stroganov S., Vadai E., Youssef S.A., Tzadikevitch-Geffen K., Rotkopf R., Biron-Shental T., de Bruin A., Neeman M.. Molecular pathways of senescence regulate placental structure and function. EMBO J. (2019) 38 e100849. DOI: 10.15252/embj.2018100849 25. Demaria M., Ohtani N., Youssef S.A., Rodier F., Toussaint W., Mitchell J.R., Laberge R.M., Vijg J., Van Steeg H., Dolle M.E.. An essential role for senescent cells in optimal wound healing through secretion of PDGF-AA. Dev. Cell (2014) 31 722-733. DOI: 10.1016/j.devcel.2014.11.012 26. Cuollo L., Antonangeli F., Santoni A., Soriani A.. The Senescence-Associated Secretory Phenotype (SASP) in the Challenging Future of Cancer Therapy and Age-Related Diseases. Biology (2020) 9. DOI: 10.3390/biology9120485 27. Maciel-Baron L.A., Morales-Rosales S.L., Aquino-Cruz A.A., Triana-Martinez F., Galvan-Arzate S., Luna-Lopez A., Gonzalez-Puertos V.Y., Lopez-Diazguerrero N.E., Torres C., Konigsberg M.. Senescence associated secretory phenotype profile from primary lung mice fibroblasts depends on the senescence induction stimuli. Age (Dordr) (2016) 38 26. DOI: 10.1007/s11357-016-9886-1 28. Hoare M., Ito Y., Kang T.W., Weekes M.P., Matheson N.J., Patten D.A., Shetty S., Parry A.J., Menon S., Salama R.. NOTCH1 mediates a switch between two distinct secretomes during senescence. Nat. Cell Biol. (2016) 18 979-992. DOI: 10.1038/ncb3397 29. Solozobova V., Rolletschek A., Blattner C.. Nuclear accumulation and activation of p53 in embryonic stem cells after DNA damage. BMC Cell Biol. (2009) 10. DOI: 10.1186/1471-2121-10-46 30. Campisi J., d’Adda di Fagagna F.. Cellular senescence: When bad things happen to good cells. Nat. Rev. Mol. Cell Biol. (2007) 8 729-740. DOI: 10.1038/nrm2233 31. Liu X., Li F., Huang Q., Zhang Z., Zhou L., Deng Y., Zhou M., Fleenor D.E., Wang H., Kastan M.B.. Self-inflicted DNA double-strand breaks sustain tumorigenicity and stemness of cancer cells. Cell Res. (2017) 27 764-783. DOI: 10.1038/cr.2017.41 32. Haugstetter A.M., Loddenkemper C., Lenze D., Grone J., Standfuss C., Petersen I., Dorken B., Schmitt C.A.. Cellular senescence predicts treatment outcome in metastasised colorectal cancer. Br. J. Cancer (2010) 103 505-509. DOI: 10.1038/sj.bjc.6605784 33. Collado M., Gil J., Efeyan A., Guerra C., Schuhmacher A.J., Barradas M., Benguria A., Zaballos A., Flores J.M., Barbacid M.. Tumour biology: Senescence in premalignant tumours. Nature (2005) 436 642. DOI: 10.1038/436642a 34. Laberge R.M., Sun Y., Orjalo A.V., Patil C.K., Freund A., Zhou L., Curran S.C., Davalos A.R., Wilson-Edell K.A., Liu S.. MTOR regulates the pro-tumorigenic senescence-associated secretory phenotype by promoting IL1A translation. Nat. Cell Biol. (2015) 17 1049-1061. DOI: 10.1038/ncb3195 35. Bavik C., Coleman I., Dean J.P., Knudsen B., Plymate S., Nelson P.S.. The gene expression program of prostate fibroblast senescence modulates neoplastic epithelial cell proliferation through paracrine mechanisms. Cancer Res. (2006) 66 794-802. DOI: 10.1158/0008-5472.CAN-05-1716 36. Yoshimoto S., Loo T.M., Atarashi K., Kanda H., Sato S., Oyadomari S., Iwakura Y., Oshima K., Morita H., Hattori M.. Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome. Nature (2013) 499 97-101. DOI: 10.1038/nature12347 37. Farsam V., Basu A., Gatzka M., Treiber N., Schneider L.A., Mulaw M.A., Lucas T., Kochanek S., Dummer R., Levesque M.P.. Senescent fibroblast-derived Chemerin promotes squamous cell carcinoma migration. Oncotarget (2016) 7 83554-83569. DOI: 10.18632/oncotarget.13446 38. Kim Y.H., Choi Y.W., Lee J., Soh E.Y., Kim J.H., Park T.J.. Senescent tumor cells lead the collective invasion in thyroid cancer. Nat. Commun. (2017) 8 15208. DOI: 10.1038/ncomms15208 39. Yang J., Liu M., Hong D., Zeng M., Zhang X.. The Paradoxical Role of Cellular Senescence in Cancer. Front. Cell Dev. Biol. (2021) 9 722205. DOI: 10.3389/fcell.2021.722205 40. Krizhanovsky V., Yon M., Dickins R.A., Hearn S., Simon J., Miething C., Yee H., Zender L., Lowe S.W.. Senescence of activated stellate cells limits liver fibrosis. Cell (2008) 134 657-667. DOI: 10.1016/j.cell.2008.06.049 41. Kong X., Feng D., Wang H., Hong F., Bertola A., Wang F.S., Gao B.. Interleukin-22 induces hepatic stellate cell senescence and restricts liver fibrosis in mice. Hepatology (2012) 56 1150-1159. DOI: 10.1002/hep.25744 42. Lujambio A., Akkari L., Simon J., Grace D., Tschaharganeh D.F., Bolden J.E., Zhao Z., Thapar V., Joyce J.A., Krizhanovsky V.. Non-cell-autonomous tumor suppression by p53. Cell (2013) 153 449-460. DOI: 10.1016/j.cell.2013.03.020 43. Huang Y.H., Chen M.H., Guo Q.L., Chen Z.X., Chen Q.D., Wang X.Z.. Interleukin-10 induces senescence of activated hepatic stellate cells via STAT3-p53 pathway to attenuate liver fibrosis. Cell Signal (2020) 66 109445. DOI: 10.1016/j.cellsig.2019.109445 44. Avelar R.A., Ortega J.G., Tacutu R., Tyler E.J., Bennett D., Binetti P., Budovsky A., Chatsirisupachai K., Johnson E., Murray A.. A multidimensional systems biology analysis of cellular senescence in aging and disease. Genome Biol. (2020) 21 91. DOI: 10.1186/s13059-020-01990-9 45. Sun Q., Shang Y., Sun F., Dong X., Niu J., Li F.. Interleukin-6 Promotes Epithelial-Mesenchymal Transition and Cell Invasion through Integrin beta6 Upregulation in Colorectal Cancer. Oxid. Med. Cell Longev. (2020) 2020 8032187. DOI: 10.1155/2020/8032187 46. Jiang Y.N., Ni X.Y., Yan H.Q., Shi L., Lu N.N., Wang Y.N., Li Q., Gao F.G.. Interleukin 6-triggered ataxia-telangiectasia mutated kinase activation facilitates epithelial-to-mesenchymal transition in lung cancer by upregulating vimentin expression. Exp. Cell Res. (2019) 381 165-171. DOI: 10.1016/j.yexcr.2019.05.011 47. Hou L., Xie S., Li G., Xiong B., Gao Y., Zhao X., Hu J., Deng S., Jiang J.. IL-6 Triggers the Migration and Invasion of Oestrogen Receptor-Negative Breast Cancer Cells via Regulation of Hippo Pathways. Basic Clin. Pharmacol. Toxicol. (2018) 123 549-557. DOI: 10.1111/bcpt.13042 48. Lin Y., Huang R., Chen L., Li S., Shi Q., Jordan C., Huang R.P.. Identification of interleukin-8 as estrogen receptor-regulated factor involved in breast cancer invasion and angiogenesis by protein arrays. Int. J. Cancer (2004) 109 507-515. DOI: 10.1002/ijc.11724 49. Bhat K., Sarkissyan M., Wu Y., Vadgama J.V.. GROalpha overexpression drives cell migration and invasion in triple negative breast cancer cells. Oncol. Rep. (2017) 38 21-30. DOI: 10.3892/or.2017.5668 50. Rosette C., Roth R.B., Oeth P., Braun A., Kammerer S., Ekblom J., Denissenko M.F.. Role of ICAM1 in invasion of human breast cancer cells. Carcinogenesis (2005) 26 943-950. DOI: 10.1093/carcin/bgi070 51. Singh B., Berry J.A., Shoher A., Ramakrishnan V., Lucci A.. COX-2 overexpression increases motility and invasion of breast cancer cells. Int. J. Oncol. (2005) 26 1393-1399. DOI: 10.3892/ijo.26.5.1393 52. Lattanzio R., Iezzi M., Sala G., Tinari N., Falasca M., Alberti S., Buglioni S., Mottolese M., Perracchio L., Natali P.G.. PLC-gamma-1 phosphorylation status is prognostic of metastatic risk in patients with early-stage Luminal-A and -B breast cancer subtypes. BMC Cancer (2019) 19. DOI: 10.1186/s12885-019-5949-x 53. Gril B., Palmieri D., Bronder J.L., Herring J.M., Vega-Valle E., Feigenbaum L., Liewehr D.J., Steinberg S.M., Merino M.J., Rubin S.D.. Effect of lapatinib on the outgrowth of metastatic breast cancer cells to the brain. J. Natl. Cancer Inst. (2008) 100 1092-1103. DOI: 10.1093/jnci/djn216 54. Orjalo A.V., Bhaumik D., Gengler B.K., Scott G.K., Campisi J.. Cell surface-bound IL-1alpha is an upstream regulator of the senescence-associated IL-6/IL-8 cytokine network. Proc. Natl. Acad. Sci. USA (2009) 106 17031-17036. DOI: 10.1073/pnas.0905299106 55. Leon X., Bothe C., Garcia J., Parreno M., Alcolea S., Quer M., Vila L., Camacho M.. Expression of IL-1alpha correlates with distant metastasis in patients with head and neck squamous cell carcinoma. Oncotarget (2015) 6 37398-37409. DOI: 10.18632/oncotarget.6054 56. Litmanovich A., Khazim K., Cohen I.. The Role of Interleukin-1 in the Pathogenesis of Cancer and its Potential as a Therapeutic Target in Clinical Practice. Oncol. Ther. (2018) 6 109-127. DOI: 10.1007/s40487-018-0089-z 57. Baker K.J., Houston A., Brint E.. IL-1 Family Members in Cancer; Two Sides to Every Story. Front. Immunol. (2019) 10 1197. DOI: 10.3389/fimmu.2019.01197 58. Freund A., Orjalo A.V., Desprez P.Y., Campisi J.. Inflammatory networks during cellular senescence: Causes and consequences. Trends Mol. Med. (2010) 16 238-246. DOI: 10.1016/j.molmed.2010.03.003 59. Lau L., Porciuncula A., Yu A., Iwakura Y., David G.. Uncoupling the senescence-associated secretory phenotype from cell cycle exit via IL-1 inactivation unveils its pro-tumorigenic role. Mol. Cell Biol. (2019) 39 e00586-18. DOI: 10.1128/MCB.00586-18 60. Gottschlich A., Endres S., Kobold S.. Therapeutic Strategies for Targeting IL-1 in Cancer. Cancers (2021) 13. DOI: 10.3390/cancers13030477
--- title: Mortality Trends due to Falls in the Group of People in Early (65–74 Years) and Late (75+) Old Age in Poland in the Years 2000–2020 authors: - Monika Burzyńska - Tomasz Kopiec - Małgorzata Pikala journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049024 doi: 10.3390/ijerph20065073 license: CC BY 4.0 --- # Mortality Trends due to Falls in the Group of People in Early (65–74 Years) and Late (75+) Old Age in Poland in the Years 2000–2020 ## Abstract The aim of the study was to assess mortality trends due to falls in early (65–74 years) and late (75+) old age groups in Poland in 2000–2020. The study used a database of all deaths due to falls in two age groups. Per 100,000 men in early old age, the crude death rate (CDR) increased from 25.3 in 2000 to 25.9 in 2020. After 2012, a statistically significant decrease was observed (annual percentage change (APC) = −$2.3\%$). Similar trends were noted for standardized death rates (SDR). Among men 75 years and older, the CDR values between the years 2000 and 2005 decreased (APC = −$5.9\%$; $p \leq 0.05$), while after 2005, they increased ($1.3\%$; $p \leq 0.05$). The SDR value decreased from 160.6 in 2000 to 118.1 in 2020. Among women aged 65–74, the CDRs values between 2000–2020 decreased from 13.9 and 8.2 per 100,000 women. The SDR value decreased from 14.0 to 8.3, respectively (2000–2007: APC = −$7.2\%$; $p \leq 0.05$). Among women aged 75+, the CDR value decreased from 151.5 to 111.6 per 100,000 but after 2008, they began to increase (APC = $1.9\%$; $p \leq 0.05$). SDR decreased from 188.9 to 98.0 per 100,000 women. Further research on the mortality in falls is needed in order to implement preventive programs. ## 1. Introduction Progressive ageing of the global population accompanied by an increase in the proportion of older people determines the current structure of health needs and challenges that healthcare systems and social welfare systems, as well as society as a whole, have to face. In Poland, between 2000 and 2020, the life expectancy of women and men increased by 2.7 and 2.9 years, respectively, and reached values of 80.7 and 72.6 [1], thus shaping the pattern of morbidity and mortality of the entire population. Needs of elderly population are very diverse, and this age group itself is heterogeneous. Internal diseases are their main health problems. In addition, a characteristic feature of old age is progressive multimorbidity, i.e., coexistence of several chronic diseases [2]. For the purpose of assessing this phenomenon, the concept of the so-called “major geriatric problems”, defined as “chronic disorders that gradually lead to functional incapacity and thus negatively affect the quality of life of older patients”, was introduced into modern gerontology [3]. These diseases include, in particular, sphincter incontinence, visual and hearing impairment, dementia disorders with delirium, old-age depression, iatrogenic syndrome, impaired mobility, and falls. The World Health Organization (WHO) defines a fall as “an event which results in a person coming to rest inadvertently on the ground or floor or other lower level” [4]. Falls are a major global public health problem. It is estimated that 684,000 fatal falls occur each year, which makes them the second most common cause of death from unintentional injury after traffic accidents. All around the world, mortality from falls is highest among people over 65 years of age. In this age group, one person in three falls down at least once a year. In the group of people aged 70 years and older, the incidence of falls is 32–$42\%$, while by the age of 80, a fall occurs in every second person [5,6]. It is worth noting that 30–$60\%$ of long-term care residents aged 65 years and older and $20\%$ of hospitalized people in this age group experience falls at least once a year [7]. Risk factors of falls, according to the simplest classification, can be divided into intrinsic and extrinsic [8]. The first group includes nonmodifiable factors, such as age and gender and health-related factors, while the other group includes environmental factors such as home surroundings, government policy on environmental design, housing standards, public transport, neighborhood conditions, and social networks [9]. In the elderly population, an accumulation of risk factors can be observed in both groups. Intrinsic factors are exacerbated by the natural ageing process, developing medical conditions, and medication use. The most significant causes of falls in this group, which meet evidence-based medicine (EBM) criteria, include muscle weakness, history of falls, gait and balance disorders, use of gait-assist devices, visual impairment, joint inflammation, depression, memory disturbances and age over 80 years [10]. Individual predictive factors also include the so-called “post-fall syndrome”, which manifests with a fear of falling down again and results in reduced physical activity, which in turn is an important element of fall prevention [11,12,13]. External risk factors of falls include hazards in home environment, such as slippery surfaces, high doorsteps, too-bright or too-dim light, and cabinets which are fixed too high, as well as outside hazards, such as uneven surfaces or architectural barriers which prevent mobility [14]. Poor financial conditions, lack of social support, and limited availability of healthcare services are other causes of falls [15]. In 2020, the standardized death rate due to falls in people aged 65 years and older in Poland was 43.1 per 100,000 population in this age group, higher than the average value in European Union countries (40.9) and WHO European Region countries (33.9) [16]. In addition to being the most common cause of fatal injury among people over 65 years of age, falls are also a cause of nonfatal injuries leading to impaired functioning due to complications. Falls can cause hip and thigh injuries both in men and women. They are the most common reason for hip fracture hospital admissions (9 in 10 cases). Other injuries that result from falls include head injuries, wrist fractures, humerus, intracranial hematomas, and injuries to internal organs [17]. In the group of people aged 75 and older, $70\%$ of falls are fatal. About $5\%$ result in a fracture and 10–$20\%$ result in severe soft tissue injury [18]. Advanced age, frailty, and pre-existing medical conditions decrease the likelihood that older individuals will recover from fall-related injuries [19]. The Global Burden of Diseases (GBD), Injuries and Risk Factors Study 2017 shows that falls are ranked as the 18th leading cause of age-standardized disability-adjusted life year (DALY) rates [20]. Between the years 2000 and 2019, DALYs due to falls in the elderly increased globally by $60\%$ ($18\%$ per 100,000 population) [21], which confirms that this is an important and growing public health problem. Changes in mortality due to falls among people over 65 years old were analyzed. The strength of the study is the completeness of death registration and the long period of the analysis. The results of studies of mortality of the elderly usually define this group as 65+. However, it has been shown that this population is heterogeneous in terms of health problems. Therefore, our analysis was conducted in two subgroups—people in early and late old age. Moreover, the analysis of mortality trends using the joinpoint regression allowed to draw conclusions regarding changes in the fall-related mortality model which is an important contribution to the field and provides the novelty of the research. The results of the study can be used to set strategic directions for health policy regarding fall prevention. The justification of the study is the fact that falls of the elderly constitute a serious geriatric, psychiatric, social, and economic problem. These may cause significant morbidity and mortality. Falls can also threaten the independence of older people and may be responsible for an individual’s loss of independence. It is a growing socioeconomic problem which may add extra burden to the healthcare—especially with the ageing of populations of many countries, including Poland. This raises a need to conduct research in this area. The aim of this study was to assess mortality trends due to falls in the early (65–74 years) and late (75+) old age groups in Poland between the years 2000 and 2020. Specific objectives of the study included assessment of changes in mortality due to falls in two age groups distinguished by sex, using CDR and SDR, taking into account the rate and significance of changes in trend direction in the analyzed period. ## 2. Materials and Methods This study used a database of deaths of all Polish residents between 2000 and 2020, included in death certificates obtained from the Central Statistical Office in Poland. From this database, all deaths due to falls (according to the International Statistical Classification of Diseases and Health Related Problems—Tenth Revision—ICD-10, coded as W00–W19) and two age groups of the deceased were distinguished: early old age (65–74 years) and late old age (75+). The total number of deaths due to falls of people aged 65 years and older between the years 2000 and 2020 was 61,994 (10,560 in early old age and 51,434 in late old age). Crude death rates (CDRs) and standardized death rates (SDRs) were calculated according to the following formulas:CDR=kp×100,000 where k—number of deaths; p—population size. SDR=∑$i = 1$Nkipiwi∑$i = 1$Nwi where ki is the number of deaths in this i-age group, pi is population size of this i-age group, wi is the weight assigned to this i-age group, resulting from the distribution of the standard population, and N—number of the age groups. The standardization procedure was performed with the use of a direct method, in compliance with the European Standard Population, updated in 2012 [22]. The Revised European Standard Population is an unweighted average of individual populations of EU-27 and EFTA countries in each 5-year age band (with the exception of people under 5 years of age and 85 years or older). An analysis of time trends was carried out with joinpoint models and the Joinpoint Regression program, a statistical software package developed by the U.S. National Cancer Institute for the Surveillance, Epidemiology and End Results Program [23]. The joinpoint regression model is an advanced version of linear regression y = bx+a, where b is the slope coefficient, a is the y-intercept, y = ln(z), z is the measure evaluated in the study (SDR), and x is the calendar year. Time trends were determined with the use of segments joining in joinpoints, where trend values significantly changed ($p \leq 0.05$). To confirm whether the changes were statistically significant, the Monte Carlo permutation method was applied. In addition, the authors also calculated annual percentage change (APC) for each segment of broken lines and average annual percentage change (AAPC) for the whole study period with corresponding $95\%$ confidence intervals (CI). Annual percent change is one way to characterize trends in death rates over time, and it was calculated according to the following formula:APC=100×(expb−1) where b—the slope coefficient. Average annual percent change (AAPC) is a summary measure of the trend over a prespecified fixed interval. It allows us to use a single number to describe the average APCs over a period of multiple years. It is valid even if the joinpoint model indicates that there were changes in trends during those years. It is computed as a weighted average of the APCs from the joinpoint model, with weights equal to the length of the APC interval [24]. AAPC=exp∑wibi∑wi−1×100 where bi—the slope coefficient for each segment in the particular range of years, and wi—the length of each segment in the range of years. ## 3. Results The number of deaths due to falls in men in early old age (65–74 years) increased from 310 in 2000 to 501 in 2020 (Table 1). The crude death rate (CDR) in 2000 was 25.3 per 100,000 men in this age group. Between 2000 and 2009, CDRs remained stable (Table 1, Figure 1). Between the years 2009 and 2012, a statistically insignificant increase was observed, while after the year 2012, CDRs were significantly decreasing at an average annual rate (APC) of −$2.3\%$. As a result of these changes, the CDR value in 2020 was 25.9 per 100,000 men. Similar trends were observed for standardized death rates (SDRs). After a slight, statistically insignificant decrease in the SDR values between 2000 and 2009 and a statistically insignificant increase between 2009 and 2012, SDRs significantly decreased between 2012 and 2020 (APC = −$2.4\%$) (Table 2, Figure 2). As a consequence, the SDR value increased from 25.9 per 100,000 men in 2000 to 26.2 in 2020 (Table 1). Among men in late old age (75+), the number of deaths due to falls increased from 667 in 2000 to 1051 in 2020. Per 100,000 men, the CDR value in 2000 was 118.5 (Table 1). Between 2000 and 2005, CDR values decreased at an average annual rate of −$5.9\%$ ($p \leq 0.05$), while after 2005 they began to increase at a rate of $1.3\%$ ($p \leq 0.05$) and reached a value of 115.2 in 2020 (Table 2, Figure 1). SDR values due to falls in the late old age male group, after a rapid decline between 2000 and 2009 (APC = –$4.5\%$; $p \leq 0.05$), were characterized with statistically insignificant periodic increases (between 2009–2013 and 2016–2020) and decreases (between 2013–2016) (Table 2, Figure 2). As a result, the SDR value decreased from 160.6 in 2000 to 118.1 in 2020 (Table 1). There were 240 deaths due to falls in the early old age female group in 2000, and 204 deaths in 2020. The CDR values in 2000 and 2020 were 13.9 and 8.2 per 100,000 women aged 65–74 years, respectively (Table 3). Between 2000 and 2007, CDRs decreased at an average annual rate of −$6.8\%$ ($p \leq 0.05$). A very small and statistically insignificant increase of $0.2\%$ was noted after 2007 (Table 2, Figure 1). The SDR value in the early old age group of women decreased from 14.0 in 2000 to 6.4 in 2007 (APC = −$7.2\%$; $p \leq 0.05$) (Table 2, Table 3). After 2007, SDR values were slightly but insignificantly increasing at a rate of $0.8\%$, and in 2020 they reached the value of 8.3 per 100,000 (Table 3, Figure 2). There were 1831 deaths in women in late old age in 2000 and 1992 deaths in this age group in the year 2020 (Table 3). The CDR value in the year 2000 was 151.5 per 100,000 women aged 75 years and older. Between the years 2000 and 2008, CDRs decreased at an average annual rate of −$6.9\%$ ($p \leq 0.05$). After 2008, they began to increase at an average annual rate of $1.9\%$ ($p \leq 0.05$), reaching the value of 111.6 per 100,000 in 2020 (Table 3, Figure 1). A higher than twofold decrease in the SDR value in the group of women aged 75 years and older was observed between 2000 and 2009 (188.9 in 2000 and 89.8 in 2009) (Table 3, Figure 2). The average rate of decline between 2000 and 2009 was −$7.5\%$ ($p \leq 0.05$). There was a very slight, statistically insignificant decrease in the SDR value between the years 2009 and 2020 (APC = −$0.1\%$). In 2020, the SDR value was 98.0 per 100,000 women in the late old age group. ## 4. Discussion Falls are a serious medical, psychosocial, and economic problem. They can gradually reduce self-dependence, which in turn considerably worsen the quality of life [25,26]. The change in the age structure of the population, observed in most countries of the world, is the main factor of the increase in the absolute number of fatal falls. The number of people aged 65 years and older in Poland is steadily increasing at an alarming rate. In 2000, the number of men aged 65–74 years was 1,224,355, while the number of men aged over 75 years was 562,705. In 2020, these numbers were 1,932,697 and 912,463, respectively (an increase by approximately $58\%$ and $62\%$). In the year 2000, there were 1,730,079 women aged 65–74 years and 1,208,638 women aged 75 years or older. In 2020, these numbers increased by approximately $44\%$ and $48\%$, respectively. There were 2,489,826 and 1,784,999 women in early old age and in late old age. As the elderly population increases, the number of deaths from causes which are dominant in the older age group increases. These causes of death include falls. A total of $76\%$ of the total number of deaths due to falls, registered in Poland in 2020, occurred in people aged 65 years and older ($15\%$ in the 65–74 age group and $62\%$ in the 75 years and older age group) [27]. The increase in crude death rates in the late old age group observed in our study (since 2005 in the group of men and since 2008 in the group of women) is mainly related to the aforementioned changes in the age structure of the Polish population. Age-standardized death rates in the group of women aged 75 years and older remained relatively stable after 2009, while in the group of men aged 75 years and older they were subject to periodic, statistically insignificant increases and decreases. The lack of significant improvement in mortality rates due to falls in the Polish elderly population observed since 2009, despite a favorable trend in the first decade of the analysis, may be related to an increase in the dynamics of growth in the trend of the oldest people (i.e., those aged over 80), particularly noticeable in the last decade (from $6.4\%$ to $8.5\%$). Indeed, mortality rates due to falls increase exponentially with age in both sexes, reaching their highest values at the age of 85 and older. This can be attributed to age-related progressive degeneration of cognitive, sensory, and physical functions, as well as an increase in the number of comorbidities and the phenomenon of polyphagia [28]. With regards to the profile of people burdened with the highest risk of falling, considering all groups of factors, it can be concluded that elderly women, slim, with weakened muscles and gait disturbances, affected by multimorbidity, taking more than four medications per day and with a history of falls are at the highest risk of falling [29]. Women are three times more likely to experience a fall than men [30]. However, an analysis of mortality in the subpopulation of people experiencing falls reveals that the male population, rather than the female one, faces an increased risk of death due to this cause. This observation was confirmed in this study for both the early and late age groups. This is due to several factors. Firstly, men suffer from more comorbidities than women of the same age [31]. In addition, this is a result of different circumstances of the fall and kind of sustained injuries, which, as O’Neill points out in his study, is determined by gender [32]. Men are more likely to fall down outdoors, during acute episodes of the disease, hereby being more likely to sustain head and thoracic injuries, which increases mortality rates due to this cause [33]. Worldwide mortality trends due to falls are various in various areas. This can be explained by differences between countries both in terms of demographic structure (including race) and in terms of activity patterns of the older population [34]. In the year 2000, in which the authors of this study initiated their research, the average SDR values per 100,000 population in European Union countries, the WHO European Region, and in Poland were 50.9, 39.5, and 77.9, respectively. In all cases, a decrease in these values was observed by 2020—by $19.6\%$, $14.2\%$, and $44.7\%$, respectively [16]. In 2000, the highest mortality due to falls among European countries was observed in Hungary, i.e., 181.93 per 100,000 population. However, over a period of 20 years, a favorable trend was observed in this respect, and in the year 2020, the SDR value was almost three times lower (69.0). Over the analyzed period, the trends improved in most European countries. The greatest dynamics were noted in the first decade, which corresponds to the results of our analysis. In contrast, unfavorable mortality trends from this cause were particularly observed in the Netherlands (an increase from 7.9 in 2000 to 100.68 in 2020), Slovenia (an increase from 95.18 to 159.29), and Croatia (an increase from 91.78 to 111.96) [20]. Data obtained from the Center for Disease Control and Prevention revealed that in the year 2018, standardized death rates in the United States due to falls in people aged 65 years and older was 64 deaths per 100,000 elderly people, and has increased by approximately $30\%$ since 2009. The increase was observed in 30 states and in the District of Columbia. The rate was growing most rapidly in the population aged 85 years and older (approximately $4\%$ per year) [34,35,36]. In Poland, favorable mortality trends due to falls in the late old age group also started to decrease, and this negative trend began in the year 2009. Studies conducted in China between 2013 and 2020 also revealed unfavorable fall-related mortality trends in the population aged 65 years and older during the study period, particularly among women aged 85 years and older, where the rate of increase expressed as average percentage change (APC) was the highest [37]. Prognoses of the National Institute of Geriatrics, Rheumatology and Rehabilitation indicate that by 2050, the number of falls among the elderly in Poland will have almost doubled, which will result in nearly 2.3 million cases [38]. However, research shows that effective public health interventions prevent falls and their complications. From a public health perspective, concerted action should be taken to reduce the number of falls among the elderly in order to reduce fall-related injuries and complications. Multifaceted strategies aimed at reducing risk factors of the incidence of falls among people at high risk should be implemented. Specific interventions might include the development of policies to prevent falls in long-term care facilities and public places and education sessions on how to prevent falls. The key is to identify people at risk of falling and to refer them to local programs or resources. Prevention of falls must span the spectrum of ages and health states within the older population and address the diversity of causes of falls without unnecessarily compromising quality of life and independence [39]. Preventive actions usually involve improving safety in the home environment and the immediate environment of the elderly person, making a proper diagnosis and implementing an appropriate therapy of current diseases, monitoring pharmacotherapy, initiating exercises to improve balance and motor coordination, providing assistive equipment, and educating the patient and his/her relatives [40,41]. Taking comprehensive and personalized preventive measures can reduce the number of falls in the elderly by as much as 40 to $60\%$ [42,43], which confirms that there is a need to carry out in-depth research in this area. The above measures might also enable reducing the health, social, and economic impacts of this phenomenon in the subpopulation of elderly people as well as in the general population. ## 5. Conclusions The number of elderly people in Poland is increasing rapidly, which entails an increase in the number of fatal falls. However, falls are still not considered a public health problem, and this negligence is evidenced by the lack of comprehensive epidemiological data on detailed assessment of the situation and trends in this area. This raises a need to conduct in-depth research in this area, as assessment of mortality trends due to falls in different subpopulations can help to identify needs and implement appropriate prevention programs for specific target groups. In view of current demographic trends, fall prevention should be a priority in healthcare in Poland, and particular attention should be paid to tailoring interventions to cohorts of the oldest population and to those who are at the highest risk of consequences of falls. ## References 1. Life Expectancy in Poland 2. Khan M.R., Malik M.A., Akhtar S.N., Yadav S., Patel R.. Multimorbidity and its associated risk factors among older adults in India. BMC Public Health (2022.0) 22. DOI: 10.1186/s12889-022-13181-1 3. Cybulski M., Krajewska-Kułak E.. Wielkie Zespoły Geriatryczne (2021.0) 12-14 4. 4.WHO Global Report on Falls Prevention in Older Age: World Health OrganizationWHO PressGeneva, Switzerland2008. WHO Global Report on Falls Prevention in Older Age: World Health Organization (2008.0) 5. Falls 6. Grodzicki T., Kocemba J., Skalska A.. Geriatria z Elementami Gerontologii Ogólnej (2006.0) 18-27 7. Buczak-Stec E., Goryński P.. Hospitalizacja z powodu upadków osób starszych w Polsce w 2010 roku. Przegl. Epidemiol. (2013.0) 67 141-144 8. Iamtrakul P., Chayphong S., Jomnonkwao S., Ratanavaraha V.. The Association of Falls Risk in Older Adults and Their Living Environment: A Case Study of Rural Area, Thailand. Sustainability (2021.0) 13. DOI: 10.3390/su132413756 9. Magnuszewski L., Wojszel A., Kasiukiewicz A., Wojszel Z.B.. Falls at the Geriatric Hospital Ward in the Context of Risk Factors of Falling Detected in a Comprehensive Geriatric Assessment. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph191710789 10. Czerwiński E., Borowy P., Jasiak B.. Współczesne zasady zapobiegania upadkom z wykorzystaniem rehabilitacji. Ortop. Traumatol. Rehabil. (2006.0) 4 380-387 11. Bień B., Wojszel B., Kędziora-Kornatowska K., Muszalik M.. Kompleksowa ocena geriatryczna. Kompendium Pielęgnowania Pacjentów w Starszym Wieku (2007.0) 89-96 12. Bilikiewicz A.. Psychiatria (2003.0) 4-12 13. Scheffer A.C., Schuurmans M.J., Van Dijk N., Van Der Hooft T., De Rooij S.E.. Fear of falling: Measurement strategy, prevalence, risk factors and consequences among older persons. Age Ageing (2008.0) 37 19-24. DOI: 10.1093/ageing/afm169 14. Lee S.. Falls associated with indoor and outdoor environmental hazards among community-dwelling older adults between men and women. BMC Geriatr. (2021.0) 21. DOI: 10.1186/s12877-021-02499-x 15. Pin S., Spini D.. Impact of falling on social participation and social support trajectories in a middle-aged and elderly European sample. SSM Popul. Health (2016.0) 2 382-389. DOI: 10.1016/j.ssmph.2016.05.004 16. SDR(65+), Accidental Falls, per 100,000 17. Komisar V., Dojnov A., Yang Y., Shishov N., Chong H., Yu Y., Bercovitz I., Cusimano M.D., Becker C., Mackey D.C.. Injuries from falls by older adults in long-term care captured on video: Prevalence of impacts and injuries to body parts. BMC Geriatr. (2022.0) 22. DOI: 10.1186/s12877-022-03041-3 18. Krysińska M., Domosławska-Żylińska K., Fronk M.. Edukacja i Profilaktyka Urazów Wśród Osób Powyżej 60-go Roku Życia (2017.0) 12 19. Rönnquist S.S., Viberg B., Kristensen M.T., Palm H., Jensen J.-E.B., Madsen C.F., Åkesson K.E., Overgaard S., Rogmark C.. Frailty and osteoporosis in patients with hip fractures under the age of 60—a prospective cohort of 218 individuals. Osteoporos. Int. (2022.0) 33 1037-1055. DOI: 10.1007/s00198-021-06281-y 20. James S.L., Lucchesi L.R., Bisignano C., Castle C.D., Dingels Z.V., Fox J.T., Hamilton E.B., Henry N.J., Krohn K.J., Liu Z.. The global burden of falls: Global, regional and national estimates of morbidity and mortality from the Global Burden of Disease Study 2017. Inj. Prev. (2020.0) 26 i3-i11. DOI: 10.1136/injuryprev-2019-043286 21. Disease Burden, 2000–2019 22. Revision of the European Standard Population 23. Kim H., Fay M.P., Feuer E.J., Midthune D.N.. Permutation tests for joinpoint regression with applications to cancer rates. Stat. Med. (2000.0) 19 335-351. DOI: 10.1002/(SICI)1097-0258(20000215)19:3<335::AID-SIM336>3.0.CO;2-Z 24. Surveillance Research Programme. Average Annual Percent Change (AAPC) 25. Vaishya R., Vaish A.. Falls in Older Adults are Serious. Indian J. Orthop. (2020.0) 54 69-74. DOI: 10.1007/s43465-019-00037-x 26. Maresova P., Javanmardi E., Barakovic S., Husic J.B., Tomsone S., Krejcar O., Kuca K.. Consequences of chronic diseases and other limitations associated with old age—A scoping review. BMC Public Health (2019.0) 19. DOI: 10.1186/s12889-019-7762-5 27. Struktura ludności. GUS 28. Ambrose A.F., Paul G., Hausdorff J.M.. Risk factors for falls among older adults: A review of the literature. Maturitas (2013.0) 751 51-61. DOI: 10.1016/j.maturitas.2013.02.009 29. Huang E.S., Karter A.J., Danielson K.K., Warton E.M., Ahmed A.T.. The association between the number of prescription medications and incident falls in a multi-ethnic population of adult type-2 diabetes patients: The diabetes and aging study. J. Gen. Inter. Med. (2010.0) 25 141-146. DOI: 10.1007/s11606-009-1179-2 30. Gale C.R., Cooper C., Sayer A.A.. Prevalence and risk factors for falls in older men and women: The English Longitudinal Study of Ageing. Age Ageing (2016.0) 45 789-794. DOI: 10.1093/ageing/afw129 31. Kim S.H., Kim S., Cho G.C., Lee J.H., Park E.J., Lee D.H.. Characteristics of fall-related head injury versus non-head injury in the older adults. BMC Geriatr. (2021.0) 21. DOI: 10.1186/s12877-021-02139-4 32. O’Neill T.W., Varlow J., Silman A.J., Reeve J., Reid D.M., Todd C., Woolf A.D.. Age and sex influences on fall characteristics. Ann. Rheum. Dis. (1994.0) 53 773-775. DOI: 10.1136/ard.53.11.773 33. Greenberg M.R., Moore E.C., Nguyen M.C., Stello B., Goldberg A., Barraco R.D., Porter B.G., Kurt A., Dusza S.W., Kane B.G.. Perceived Fall Risk and Functional Decline: Gender Differences in Patient’s Willingness to Discuss Fall Risk, Fall History, or to Have a Home Safety Evaluation. Yale J. Biol. Med. (2016.0) 89 261-267. PMID: 27354852 34. Hartholt K.A., Lee R., Burns E.R., van Beeck E.F.. Mortality From Falls Among US Adults Aged 75 Years or Older, 2000–2016. JAMA (2019.0) 321 2131-2133. DOI: 10.1001/jama.2019.4185 35. Web-Based Injury Statistics Query and Reporting System (WISQARS) 36. Burns E.R., Kakara R.. Deaths from Falls Among Adults ≥65 Years—United States, 2007–2016. MMWR Morb. Mortal. Wkly. Rep. (2018.0) 67 509-514. DOI: 10.15585/mmwr.mm6718a1 37. Zhang K., Qi J., Zuo P., Yin P., Liu Y., Liu J., Wang L., Li L.. The mortality trends of falls among the elderly adults in the mainland of China, 2013–2020: A population-based study through the National Disease Surveillance Points system. Lancet Reg. Health West Pac. (2021.0) 19 100336. DOI: 10.1016/j.lanwpc.2021.100336 38. Jak Zapobiegać Upadkom Seniorów? 39. Natora A.H., Oxley J., Barclay L., Taylor K., Bolam B., Haines T.P.. Improving Policy for the Prevention of Falls Among Community-Dwelling Older People-A Scoping Review and Quality Assessment of International National and State Level Public Policies. Int. J. Public Health (2022.0) 27 1604604. DOI: 10.3389/ijph.2022.1604604 40. Hübner L., Voelcker-Rehage C.. Does physical activity benefit motor performance and learning of upper extremity tasks in older adults?—A systematic review. Eur. Rev. Aging Phys. Act. (2017.0) 14 15. DOI: 10.1186/s11556-017-0181-7 41. Tang S., Liu M., Yang T., Ye C., Gong Y., Yao L., Xu Y., Bai Y.. Association between falls in elderly and the number of chronic diseases and health-related behaviors based on CHARLS 2018: Health status as a mediating variable. BMC Geriatr. (2022.0) 22. DOI: 10.1186/s12877-022-03055-x 42. Gürler H., Bayraktar N.. The effectiveness of a recurrent fall prevention program applied to elderly people undergoing fracture treatment. Int. J. Orthop. Trauma Nurs. (2021.0) 40 100820. DOI: 10.1016/j.ijotn.2020.100820 43. Montero-Odasso L.M., van der Velde N., Martin F.C.. World guidelines for falls prevention and management for older adults: A global initiative. Age Ageing (2022.0) 51 9. DOI: 10.1093/ageing/afac205
--- title: Impact of COVID-19 on the Health-Related Quality of Life of Patients during Infection and after Recovery in Saudi Arabia authors: - Menyfah Q. Alanazi - Waleed Abdelgawwad - Thamer A. Almangour - Fatma Mostafa - Mona Almuheed journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049034 doi: 10.3390/ijerph20065026 license: CC BY 4.0 --- # Impact of COVID-19 on the Health-Related Quality of Life of Patients during Infection and after Recovery in Saudi Arabia ## Abstract This study evaluated the impact of COVID-19 and other factors on the health-related quality of life (HRQoL) of Saudi patients during infection and after recovery using the EQ-5D-5L and EQ-VAS instruments. An observational prospective study was conducted in November 2022, during which 389 COVID-19 patients were surveyed during their visit to a medical center. Two weeks after their recovery, they were contacted again to re-evaluate their HRQoL (192 patients either refused to participate or withdrew). The mean of the EQ-5D-5L index and EQ-VAS scores significantly increased from (0.69 ± 0.29 and 63.16 ± 24.9) during infection to (0.92 ± 0.14 and 86.96 ± 15.3) after recovery. Specifically, COVID-19 patients experienced improvement of several HRQoL dimensions post recovery, such as better mobility, enhanced self-care, returning to usual activities, less pain/discomfort, and alleviated anxiety/depression. Multiple linear regression analyses showed that having a normal weight, being employed, not being anemic, and previously taking the BCG vaccine were positively associated with a greater change in the HRQoL. An interaction between being asthmatic and taking the influenza vaccine positively predicted a lower change in the HRQoL. Having a normal weight positively predicted a greater change in the perceived health state after recovery. Increasing the consumption of natural supplements (honey and curcuma) did not improve the HRQoL or the perceived health state. Based on these findings, COVID-19 mildly impacted the HRQoL of Saudis with varying effects depending on some socio-demographic/clinical characteristics of the patients. ## 1. Introduction Coronavirus disease 2019 (COVID-19) is an infectious acute respiratory disease caused by the novel coronavirus SARS-CoV-2 [1]. Soon after the first case was reported in 2019, this highly infectious virus rapidly became a pandemic [1,2]. The severity of COVID-19 ranges from enduring mild to severe symptoms to pneumonia, acute respiratory distress syndrome (ARDS), multi-organ dysfunctions, and death [2,3,4]. Aside from the clinical consequences, the impact of COVID-19 on the quality of life extends further [5]. In the Arabian Gulf region, the Kingdom of Saudi Arabia was witnessing a surge of a MERS-CoV epidemic when the first COVID-19 case was confirmed on 2 March 2020. As of 2 November 2022, SARS-CoV-2 has infected more than 822 thousand Saudi Arabian residents and killed more than 9400 [6]. Previous studies mainly focused on the epidemiology and transmission of SARS-CoV-2, infection control and prevention, COVID-19 vaccination, disease burden, and treatment options. However, little effort was made to evaluate the impact of COVID-19 on the health-related quality of life (HRQoL) of Saudi Arabian residents infected by SARS-CoV-2 [7,8]. Quality of life (QoL) was conceptualized as a broad multidimensional subjective evaluation of an individual’s various life aspects [9]. The World Health Organization (WHO) referred to it as an individual’s perception of their position in life, an extension of the environment in which they live, and in relation to their goals, expectations, standards, and concerns [9,10]. Quality of life (QoL) is an important goal of treatment in chronic illnesses, and is also used to identify the range of problems that can affect patients [9,10]. HRQoL is a more specific measure that quantifies the physical and mental status of an individual or a group over time. This metric is generally used to represent the impact of an illness and its management on an individual’s ability to live a fulfilling life and their overall life satisfaction [11]. HRQoL is an essential indicator of the burden of a disease that helps clinicians and policy-makers optimize patient care and public health decisions [11,12]. Many instruments have been implemented to measure HRQoL, including the five-dimensional EuroQol (EQ-5D-5L). The EQ-5D-5L is one of the most frequently used, preference-based measures of HRQoL worldwide that evaluates the HRQoL based on the patient’s perspective of their health [13,14,15,16] EQ-5D-5L is a valid and reliable measure that has been applied to countless disease areas and is the most commonly used tool in cost–utility analyses and for appraising healthcare interventions [13,16,17,18,19,20]. Since its development, it has been validated across various settings and among different populations. However, in terms of COVID-19 and in conservative communities such as that in Saudi Arabia, the use of EQ-5D-5L as a measure of HRQoL was scarce [16,19]. Being a novel virus and a large-scale pandemic, it is important to evaluate HRQoL among individuals within the context of COVID-19 not only during the disease, but also after recovery. The individual’s experience with COVID-19 symptoms of various severities, quarantine, exposure to news on COVID-19 morbidity/mortality, and social stigmatization is unique. Exploring all the possible factors that positively impact the HRQoL during COVID-19 and proposing tailored interventions will lead to a better and an expedited recovery. In Saudi Arabia, there is a paucity of research on the impact of COVID-19 on HRQoL. Moreover, it is not clear what the possible factors are that would enhance HRQoL after recovery from COVID-19. Factors associated with HRQOL in chronic diseases such as diabetes, breast cancer, arthritis, and hypertension are commonly investigated [7,8]. For instance, improving certain lifestyle characteristics such as weight reduction, physical activity, and smoking cessation were associated with better HRQoL in such diseases. In terms of infectious diseases such as COVID-19, the experience might be different as other factors might play a role, such as previous vaccination (COVID-19, seasonal flu) and consumption of natural supplements such as drinking herbal drinks, consuming honey, or others which COVID-19 patients might perceive to be beneficial. Studies showed that herbal supplements and honey—which are commonly used in Saudi Arabia—have some antiviral activities [20,21,22,23]. In terms of vaccination, exploring the change in HRQoL during COVID-19 and after recovery can aid in persuading vaccine-hesitant groups to get vaccinated. In terms of consuming natural supplements, sharing personal experiences can inform researchers, policymakers, and other individuals at risk of contracting COVID-19. The aim of this study was to evaluate the impact of COVID-19 and other factors on the HRQoL of Saudi patients during infection and after recovery. We hypothesize that the mean difference in the HRQoL scores (during the disease and after recovery) might be associated with certain socio-demographics, health-related factors, vaccination status, and other self-reported complementary health interventions that participants perceive to be beneficial. ## 2.1. Study Design This was an observational prospective study, in which a cohort of confirmed COVID-19 patients who visited the influenza clinic at King Abdulaziz Medical City (KAMC), Riyadh, Saudi Arabia, were enrolled between November and December of 2022. For convenience, COVID-19-positive cases circumstantially present at the targeted setting during the study period were screened for eligibility then invited to participate. Baseline measures were obtained at the clinic and then repeated after two weeks of recovery. ## 2.2. Study Setting This study was conducted in the influenza clinic of KAMC, a 1505-bed university-affiliated tertiary care center, accredited by the Joint Commission International. ## 2.3. Study Population The participants were all adults between 18 and 59 years of age with a confirmed diagnosis of COVID-19, based on a rapid antigen test (RAT). Those who agreed to participate in this study provided informed consent. Participants who were severely ill and admitted to the hospital were excluded. Patients unable to read and write were also excluded from this study. ## 2.4. Data Collection A self-administered questionnaire was initially used in this study at baseline, followed by an online survey after two weeks of recovery. At baseline, the questionnaire was coupled with a letter of invitation and an informed consent, all in Arabic. The questionnaire was handed to participants by the research investigators who assisted them when needed. The study participants were provided with a Google Survey link to be accessed and filled in two weeks after of discharge. The research investigators were familiar with the targeted setting and population in terms of language and cultural norms. The questionnaire consisted of: ## 2.4.1. Study Exposures Socio-demographics included sex, age (years), level of education (higher education, middle education, low education), marital status (single, married, other), and employment status (employed, unemployed, student). Clinical or health-related characteristics included the previous medical history (having a chronic disease), previous vaccination history against seasonal influenza, Bacillus Calmette–Guérin (BCG), and COVID-19, smoking status (current smoker/never smoked/former smoker), body mass index (obesity: BMI ≥ 30 kg/m2), and any natural supplement (dietary) consumed during COVID-19 onset to overcome the discomfort of the disease. The symptoms and duration of COVID-19 infection were also recorded. ## 2.4.2. Study Outcomes After obtaining prior approval from the EuroQol Research Foundation, we utilized the EQ-5D-5L tool [16]. Several studies have confirmed the validity and reliability of the Arabic EQ-5D-5L in different Arabian populations, and the tool has proved to be a valid measure for HRQoL in Arabic speaking populations. The EQ-5D-5L instrument has two components: a description of the health state (EQ-5D-5L descriptive system) and a self-evaluation or perception of the individual’s health state using a visual analog scale (EQ-VAS). The EQ-5D-5L consists of five dimensions: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension has five levels of severity (no problems: 1, slight problems: 2, moderate problems: 3, severe problems: 4, extreme problems: 5. The responses are combined to produce a five-digit number describing the respondent’s health status (ranging from 11,111 to 55,555): 11,111 (having no problems in all dimensions, or full health) and 55,555 (having extreme problems in all dimensions, or worst health). The EQ-5D-5L scores are then standardized to become an index ranging from 0 (representing death) to 1 (representing full health), with negative values representing states worse than death [13,16,17]. The second part of the tool is a visual vertical scale ranging from “0” to “100” ranging from “Best imaginable health state” [100] to “Worst imaginable health state” [0] [13,16,17,18,19,20]. ## 2.5. Data Management and Analysis Statistical analyses were performed using SPSS (version 25; SPSS Inc., New York, NY, USA). Descriptive statistics of categorical variables were presented in frequency and percentage, while continuous variables were presented in mean ± SD, or median (interquartile range). Measures of internal consistency showed that the Cronbach’s alpha of EQ-5D-5L during COVID-19 and after recovery were 0.842 and 0.801, respectively. Same-group analyses (within subgroups) using the Wilcoxon signed ranks test was performed to compare the index scores of EQ-5D-5L dimensions and EQ-VAS during the disease and after recovery; given the distribution of EQ-5D-5L, data were ordinal and commonly skewed. The assumptions of analyses of covariance (ANCOVA) were violated due to the nature of skewed outcomes and the presence of logic outliers. Therefore, between-group analyses were performed using the Mann–Whitney test by comparing the differences in the EQ-5D-5L scores and EQ-VAS (recovery scores minus baseline scores). Higher differences in scores indicated a greater change in the HRQoL. Variables statistically significant at the bivariate level of analyses ($p \leq 0.05$) were counted as potential predictors of change in EQ-5D-5L and EQ-VAS scores. Accordingly, two multiple linear regression models were constructed. To meet the regression assumptions, the skewness in the EQ-5D-5L scores was corrected by normalization (scores minus the minimum value/range) to take a range from zero to one and then log10 transformed (skewness coefficient became −0.150). Homoscedasticity and normality of the data were confirmed after examining the P–P plot and scatterplot. An absence of multi-collinearity was confirmed by examining the variance inflation factor (VIF) values, which were all under 5, indicating that the assumption is met. In terms of EQ-5D-5L scores, no statistically significant interactions between various exposures were detected, except for being asthmatic and taking the seasonal influenza vaccine. The model was fit, and the adjusted R2 was $33.5\%$. Insignificant variables such as curcuma/onion/peppermint/wild thyme consumption, zinc deficiency, sleep duration, diabetes, and gender were dropped out, which increased the adjusted R2 to $34.2\%$. In terms of EQ-VAS scores, no significant interaction effect was present. The model was fit, and the adjusted R2 was $33.5\%$. Insignificant variables such as onion/peppermint/wild thyme consumption, smoking, sleep duration, asthma, anemia, gender, age, and BCG vaccination were dropped out, which increased the adjusted R2 to $38\%$. For all statistical tests, $p \leq 0.05$ was considered statistically significant. ## 2.6. Ethical Issues This study was approved by the Research Committee of King Abdullah International Medical Research Center (KAIMRC), King Saud Bin-Abdulaziz University for Health Sciences (IRB/$\frac{2666}{22}$). Patient confidentiality and privacy were secured by the principal investigator. ## 2.7. Sample Characteristics A total of 389 out of 590 Saudi patients were enrolled in this study and completed both questionnaires (response rate was $65.9\%$). Two thirds of the sample were women ($$n = 259$$). They had a mean of age 34.87 ± 8.36 years. The majority of the participants had a high level of educational ($82.5\%$). More than two thirds of the participants were employed ($67.4\%$), and married ($64.3\%$). The consumption of natural supplements during COVID-19 infection was reported by $55\%$ of the participants (curcuma $28.5\%$, peppermint $25.7\%$, and others). Increased consumption of honey was reported by $53\%$ of the participants, followed by an increased consumption of onion ($34.7\%$) and garlic ($33.2\%$) (Table 1). When asked about the number of times they had an episode of cold symptoms in the past six months, 352 ($90.5\%$) participants reported having one to three episodes, yet none were confirmed to be COVID-19 by a RAT test. Almost half of the participants ($55.8\%$) had at least one chronic disease. The most common disease reported by study participants was vitamin-D deficiency ($30.8\%$). Thirty percent of the patients were classified as obese. The self-reported symptoms of COVID-19 included fatigue ($66.1\%$), fever ($48\%$), cough ($35.5\%$), anxiety/depression ($34.4\%$), pain ($46\%$), numbness ($5.4\%$), and itching ($4.6\%$). The most common persisting symptoms after recovery were cough ($14.9\%$), diminished sense of taste ($13.6\%$), diminished sense of smell ($12.3\%$), fatigue ($9.8\%$), and auditory dysfunction ($2.4\%$) (Table 2). ## 3.1. Assessment of HRQoL and Perceived Health Status The EQ-5D-5L index scores significantly increased from 0.68 ± 0.31 during COVID-19 to 0.92 ± 0.13 after 2 weeks of recovery ($p \leq 0.001$ ). The severity of problems related to mobility, self-care, usual activities, pain/discomfort, and anxiety/depression (the five dimensions of EQ-5D-5L) all significantly decreased after recovery, which indicated that the overall quality of life of COVID-19 patients was enhanced due to recovery ($p \leq 0.001$ each). Similar improvements were observed in terms of EQ-VAS scores, as patients’ score increased from 63.16 ± 24.92 during infection to 86.96 ± 15.31 after recovery ($p \leq 0.001$) (Table 3). However, exploration was needed to determine which factors contributed to a significant enhancement in HRQoL and perceived health status. ## 3.2. Predictors of HRQoL and Health Status As shown in Table 4, within-group analyses showed that EQ-5D-5L index and EQ-VAS scores showed statistically significant improvement within all subgroups. However, the differences in scores (between recovery and COVID-19 infection) provided further insight into which subgroups are expected to show higher degrees of improvement compared to other groups. Between-group analyses showed that obesity, employment status, sleep duration, consumption of natural supplements (curcuma, peppermint, honey, and onion), previous medical history (asthma, anemia, zinc deficiency, and diabetes mellitus), and vaccination (influenza and BCG) were associated with changes in EQ-5D-5L index scores. For instance, the changes in EQ-5D-5L index scores were significantly greater among non-obese participants (0.37 ± 0.20), employed participants (0.27 ± 0.28), those who usually slept < 6 h (0.30 ± 0.29), and others shown in Table 4, indicating higher degrees of improvement. In terms of changes in EQ-VAS scores, males (28.11 ± 24.22), those with normal weight (45.74 ± 0.18), employed participants (27.71 ± 25.35), and others reported an improvement in their perceived health status (Table 4). Multiple linear regression analyses were performed to model the relationship between obesity, employment status, sleep duration, consumption of natural supplements, previous medical history, vaccination (influenza and BCG), and others with the changes in the ED-5D-5L index and EQ-VAS scores. In terms of the ED-5D-5L index, after adjusting for the effect of all variables, having a normal weight, being employed, not being anemic, and previously taking the BCG vaccine significantly contributed to a greater change in ED-5D-5L index scores. An interaction between being asthmatic and taking the influenza vaccine predicted lower changes in ED-5D-5L index scores. In terms of EQ-VAS scores, having a normal weight predicted a greater change in the EQ-VAS scores after recovery. Increasing the consumption of natural supplements (honey and curcuma) did not improve either of the study outcomes (Table 5). ## 4. Discussion COVID-19 infection is indisputably an unpleasant experience that impacted the HR-QoL of people in Saudi Arabia. Despite being an acute respiratory disease, some Saudi Arabians might not have fully recovered in many aspects such as mobility, self-care, daily activities, body pain, and anxiety/depression. In this study, we hypothesize that the recovery from COVID-19 will naturally enhance the HR-QoL among patients, yet some factors might have contributed to a better improvement of the HR-QoL. This study is one of the few that examined these factors using the EuroQol 5-dimension, 5-level (EQ-5D-5L) questionnaire and a visual analog scale (EQ-VAS) during and post infection. Recovery from COVID-19 enhanced the quality of life among Saudi patients, as the EQ-5D-5L index scores and EQ-VAS scores significantly increased after two weeks. This can be attributed to the absence of COVID-19 signs and symptoms. A previously published meta-analysis study stated that $58\%$ of COVID-19 patients reported having a poor quality of life after recovery [8]. Our observed EQ-5D-5L scores after recovery were similar to figures reported in China (0.949), yet higher than those reported in Morocco (0.86), the UK (0.714), Norway (0.690), and Belgium (0.620) [24,25,26,27,28]. Saudis were able to resume their normal physical activity and fully attend to their self-care needs. Pain and discomfort due to COVID-19 signs and symptoms resolved after the COVID-19 infection, despite the presence of persistent signs/symptoms in up to $27\%$ of our sample. Saudis were less anxious and less depressed after recovery, being able to resume their daily activities after quarantine [7,8,29,30]. Even in regard to EQ-VAS scores, Saudis reported higher scores compared to COVID-19-infected patients in Germany and Belgium [28,31]. This indicated that some variables might have played a role in the enhancement of HR-QoL besides the recovery from the disease. Sample characteristics were tested to identify which factors led to a significant improvement in the HRQoL. For instance, being obese was associated with a lower HR-QoL change in comparison to those with normal weight. According to the World Health Organization (WHO), the overall prevalence of obesity in Saudi Arabia is* estimated to be $33.7\%$ [32,33]. The severity of COVID-19 might have been higher among obese participants, which led to a delayed recovery. Previous studies showed that COVID-19 patients who are obese tend to have less improvement in their quality of life. This finding aligns with the conclusion of a large-scale systematic review that obesity is significantly associated with increased severity and higher mortality among COVID-19 patients. Obesity is also commonly associated with hypertension, dyslipidemia, cardiac problems, diabetes, and others, all of which impair the HRQoL in its various aspects [34]. Additionally, having a chronic disease such as asthma contributed to a delayed recovery or less improvement in the HRQoL. One study showed that among 562 asthma patients, $21\%$ were hospitalized, $3\%$ received mechanical ventilation, and one died. Anemia was also an independent risk factor associated with the severity of COVID-19 [35]. This entails that these patients in particular are expected to have a compromised quality of life after recovery and, thus, need more support. Previous vaccination status, especially BCG and seasonal influenza vaccination, was associated with the changes in the HRQoL after recovery. COVID-19 Saudi patients who were previously vaccinated against BCG and seasonal influenza were able to report higher changes in their HRQoL. It was reported that some BCG vaccine strains can be used as an additional defense in future pandemics [36]. The WHO and more than 20 European countries have already recommended the co-administration of influenza and COVID-19 vaccines [37]. Our findings can be useful to vaccine campaigns promoting the benefits of these vaccines, especially when targeting vaccine-hesitant groups. While questioning the Saudi study participants about consuming natural supplements during the onset of symptoms, they commonly reported an increased consumption of natural supplements such as honey, garlic/onion, and herbal drinks. The benefits of these natural supplements stem from deeply rooted cultural norms, despite the fact that their impact on HRQoL during COVID-19 infection was not rigorously tested in the past. A previous clinical trial enlisted some of honey’s antiviral and antibacterial properties, since its antioxidant content hugely impacts the co-morbidities associated with SARS-CoV-2 infection [23]. However, our study findings cannot be conclusive, as experimental study designs are required. ## 5. Limitations This study had some limitations. First, it was conducted in only one setting, which limits its generalizability to other hospitals. Two thirds of the participants were female, which might have produced some bias. Older participants had a lower response than the younger age groups. The measurement of health status using the EQ-5D-5L instrument may have resulted in over- or underestimation of QoL because some patients might have inflated their self-assessment of their quality of health. Finally, one of the challenges in this study was the lack of a baseline measure before the exposure to COVID-19 infection. Unfortunately, recruiting a healthy group, obtaining a baseline measure of HRQoL, then awaiting COVID-19 infection is a lengthy process and prone to a loss in follow-up. Questioning COVID-19 patients about their HRQoL prior to COVID-19 infection is also prone to recall bias. Despite these limitations, factors significantly associated with changes in the HRQoL and perceived health status in this study remain valid within the context of COVID-19 infections and the Saudi Arabian population. ## 6. Conclusions COVID-19 had a significant negative impact on Saudis’ HRQoL with varying degrees. The changes in the QoL two weeks after contracting COVID-19 were greater among individuals with normal weight, who were employed, non-anemic, and had previously taken the BCG vaccine. An interaction between being asthmatic and taking the influenza vaccine also contributed to greater changes. Further studies are needed to test the effectiveness of increased natural supplements on the changes in HR-QoL. Investigating the health status of Saudi patients with COVID-19 and identifying significantly associated factors with HRQoL will optimize future clinical approaches and inspire public health policies toward maximizing their effectiveness and efficiency. The implementation of such strategies should greatly improve COVID-19 patient outcomes and, of course, their quality of life. ## References 1. Szajkowski S., Pasek J., Dwornik M., Cieślar G.. The Impact of Coronavirus Infection on Health-Related Quality of Life in Amateur CrossFit Athletes. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph192416409 2. Zhu N., Zhang D., Wang W., Li X., Yang B., Song J., Zhao X., Huang B., Shi W., Lu R.. China Novel Coronavirus Investigating and Research Team. A novel coronavirus from patients with pneumonia in China, 2019. N. Engl. J. Med. (2020.0) 382 727-733. DOI: 10.1056/NEJMoa2001017 3. De Matos D.G., Aidar F.J., de Almeida-Neto P.F., Moreira O.C., de Souza R.F., Marçal A.C., Marcucci-Barbosa L.S., Martins Júnior F.d.A., Lobo L.F., dos Santos J.L.. The impact of measures recommended by the government to limit the spread of coronavirus (COVID-19) on physical activity levels, quality of life, and mental health of Brazilians. Sustainability (2020.0) 12. DOI: 10.3390/su12219072 4. Alanazi M.Q., AlQahtani H., Almangour T.A., Aleanizy F.S., Alqahtani F.Y.. Evaluation of the Clinical Outcome and Cost Analysis of Antibiotics in the Treatment of Acute Respiratory Tract Infections in the Emergency Department in Saudi Arabia. Antibiotics (2022.0) 11. DOI: 10.3390/antibiotics11111478 5. Lubetkin E.I., Long D., Haagsma J.A., Janssen M.F., Bonsel G.J.. Health inequities as measured by the EQ-5D-5L during COVID-19: Results from New York in healthy and diseased persons. PLoS ONE (2022.0) 17. DOI: 10.1371/journal.pone.0272252 6. WHO Coronavirus (COVID-19) Dashboard 7. Poudel A.N., Zhu S., Cooper N., Roderick P., Alwan N., Tarrant C., Ziauddeen N., Yao G.L.. Impact of COVID-19 on health-related quality of life of patients: A structured review. PLoS ONE (2021.0) 16. DOI: 10.1371/journal.pone.0259164 8. Malik P., Patel K., Pinto C., Jaiswal R., Tirupathi R., Pillai S., Patel U.. Post-acute COVID-19 syndrome (PCS) and health-related quality of life (HRQoL)—A systematic review and meta-analysis. J. Med. Virol. (2022.0) 94 253-262. DOI: 10.1002/jmv.27309 9. Group T.W.. The World Health Organization quality of life assessment (WHOQOL): Development and general psychometric properties. Soc. Sci. Med. (1998.0) 46 1569-1585. DOI: 10.1016/S0277-9536(98)00009-4 10. WHOQOL: Measuring Quality of Life: World Health Organization. (2019.0) 11. Haraldstad K., Wahl A., Andenæs R., Andersen J.R., Andersen M.H., Beisland E., Borge C.R., Engebretsen E., Eisemann M., Halvorsrud L.. A systematic review of quality of life research in medicine and health sciences. Qual. Life Res. (2019.0) 28 2641-2650. DOI: 10.1007/s11136-019-02214-9 12. Alanazi M.Q.. Evaluation of health-related quality of life in women with community-acquired urinary Tract infections using the EQ-5D-3L in Saudi Arabia. Patient Prefer. Adherence (2020.0) 14 2419-2426. DOI: 10.2147/PPA.S277367 13. Devlin N.J., Brooks R.. EQ-5D and the EuroQol group: Past, present and future. Appl. Health Econ. Health Policy (2017.0) 15 127-137. DOI: 10.1007/s40258-017-0310-5 14. Rupel V.P., Ogorevc M.. Crosswalk EQ-5D-5L value set for Slovenia. Slov. J. Public Health (2020.0) 59 189-194. DOI: 10.2478/sjph-2020-0024 15. Ludwig K., Graf von der Schulenburg J.-M., Greiner W.. German value set for the EQ-5D-5L. Pharmacoeconomics (2018.0) 36 663-674. DOI: 10.1007/s40273-018-0615-8 16. EQ-5D-5L User Guide. (2021.0) 17. Lins L., Carvalho F.M.. SF-36 total score as a single measure of health-related quality of life: Scoping review. SAGE Open Med. (2016.0) 4 2050312116671725. DOI: 10.1177/2050312116671725 18. Herdman M., Gudex C., Lloyd A., Janssen M., Kind P., Parkin D., Bonsel G., Badia X.. Development and preliminary testing of the new five-level version of EQ-5D (EQ-5D-5L). Qual. Life Res. (2011.0) 20 1727-1736. DOI: 10.1007/s11136-011-9903-x 19. van Hout B.A., Shaw J.W.. Mapping EQ-5D-3L to EQ-5D-5L. Value Health (2021.0) 24 1285-1293. DOI: 10.1016/j.jval.2021.03.009 20. Devlin N., Finch A.P., Parkin D.. Guidance to users of EQ-5D-5L value sets. Value Sets for EQ-5D-5L: A Compendium, Comparative Review & User Guide (2022.0) 213-233 21. Tahmasbi S.F., Revell M.A., Tahmasebi N.. Herbal medication to enhance or modulate viral infections. Nurs. Clin. (2021.0) 56 79-89. DOI: 10.1016/j.cnur.2020.10.007 22. Yitbarek K., Abraham G., Girma T., Tilahun T., Woldie M.. The effect of Bacillus Calmette–Guérin (BCG) vaccination in preventing severe infectious respiratory diseases other than TB: Implications for the COVID-19 pandemic. Vaccine (2020.0) 38 6374-6380. DOI: 10.1016/j.vaccine.2020.08.018 23. Soares S., Bornet M., Grosso C., Ramalhosa M.J., Gouvinhas I., Garcia J., Rodrigues F., Delerue-Matos C.. Honey as an Adjuvant in the Treatment of COVID-19 Infection: A Review. Appl. Sci. (2022.0) 12. DOI: 10.3390/app12157800 24. Ping W., Zheng J., Niu X., Guo C., Zhang J., Yang H., Shi Y.. Evaluation of health-related quality of life using EQ-5D in China during the COVID-19 pandemic. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0234850 25. Azizi A., Achak D., Saad E., Hilali A., Nejjari C., Khalis M., Youlyouz-Marfak I., Marfak A.. Health-Related Quality of Life of Moroccan COVID-19 Survivors: A Case-Control Study. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph19148804 26. Halpin S.J., McIvor C., Whyatt G., Adams A., Harvey O., McLean L., Walshaw C., Kemp S., Corrado J., Singh R.. Postdischarge symptoms and rehabilitation needs in survivors of COVID-19 infection: A cross-sectional evaluation. J. Med. Virol. (2021.0) 93 1013-1022. DOI: 10.1002/jmv.26368 27. Lerum T.V., Aaløkken T.M., Brønstad E., Aarli B., Ikdahl E., Lund K.M.A., Durheim M.T., Rodriguez J.R., Meltzer C., Tonby K.. Dyspnoea, lung function and CT findings 3 months after hospital admission for COVID-19. Eur. Respir. J. (2021.0) 57 2003448. DOI: 10.1183/13993003.03448-2020 28. Meys R., Delbressine J.M., Goërtz Y.M., Vaes A.W., Machado F.V., Van Herck M., Burtin C., Posthuma R., Spaetgens B., Franssen F.M.. Generic and respiratory-specific quality of life in non-hospitalized patients with COVID-19. J. Clin. Med. (2020.0) 9. DOI: 10.3390/jcm9123993 29. Carfì A., Bernabei R., Landi F.. Persistent Symptoms Patients after Acute COVID-19. JAMA (2020.0) 324 603-605. DOI: 10.1001/jama.2020.12603 30. Jacobs L.G., Gourna Paleoudis E., Lesky-Di Bari D., Nyirenda T., Friedman T., Gupta A., Rasouli L., Zetkulic M., Balani B., Ogedegbe C.. Persistence of symptoms and quality of life at 35 days after hospitalization for COVID-19 infection. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0243882 31. Daher A., Balfanz P., Cornelissen C.. Seguimiento de pacientes con enfermedad grave por coronavirus 2019 (COVID-19): Secuelas de enfermedad pulmonar y extrapulmonar. Respir. Med. (2020.0) 174 106197. DOI: 10.1016/j.rmed.2020.106197 32. Althumiri N.A., Basyouni M.H., AlMousa N., AlJuwaysim M.F., Almubark R.A., BinDhim N.F., Alkhamaali Z., Alqahtani S.A.. Obesity in Saudi Arabia in 2020: Prevalence, Distribution, and Its Current Association with Various Health Conditions. Healthcare (2021.0) 9. DOI: 10.3390/healthcare9030311 33. World Health Organization Diabetes Country Profiles. Saudi Arabia. (2016.0) 34. Singh R., Rathore S.S., Khan H., Karale S., Chawla Y., Iqbal K., Bhurwal A., Tekin A., Jain N., Mehra I.. Association of obesity with COVID-19 severity and mortality: An updated systemic review, meta-analysis, and meta-regression. Front. Endocrinol. (2022.0) 13 52. DOI: 10.3389/fendo.2022.780872 35. Robinson L.B., Wang L., Fu X., Wallace Z.S., Long A.A., Zhang Y., Camargo C.A., Blumenthal K.G.. COVID-19 severity in asthma patients: A multi-center matched cohort study. J. Asthma (2022.0) 59 442-450. DOI: 10.1080/02770903.2020.1857396 36. Borges K., da Costa A.C., de Souza Barbosa L.C., Ribeiro K.M., Dos Anjos L.R.B., Kipnis A., Junqueira-Kipnis A.P.. Tuberculosis, BCG Vaccination, and COVID-19: Are They Connected?. Mini Rev. Med. Chem. (2022.0) 22 1631-1647. DOI: 10.2174/1389557522666220104152634 37. Janssen C., Mosnier A., Gavazzi G., Combadière B., Crépey P., Gaillat J., Launay O., Botelho-Nevers E.. Coadministration of seasonal influenza and COVID-19 vaccines: A systematic review of clinical studies. Hum. Vaccines Immunother. (2022.0) 18 2131166. DOI: 10.1080/21645515.2022.2131166
--- title: Mental Health and Smoking-Related Determinants of Alcohol Drinking Experience in Korean Adolescents authors: - Sook Kyoung Park - Hae-Kyung Jo - Eunju Song journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049051 doi: 10.3390/ijerph20064724 license: CC BY 4.0 --- # Mental Health and Smoking-Related Determinants of Alcohol Drinking Experience in Korean Adolescents ## Abstract This study aimed to identify the determinants of mental health and smoking-related behaviors among Korean adolescents with drinking experience. Secondary data from the Korean Youth Risk Behavior Web-based Survey [2021] were analyzed. The final study sample consisted of data from 5905 adolescents with a history of smoking. Chi-square and multivariate logistic regression analyses were used to examine the factors related to drinking experience. The factors that influenced alcohol drinking were sex, school level, academic performance, self-reported depression, and smoking. The results of this study showed that there are numerous factors affecting adolescents’ drinking experience. Early education and interventions are needed to reduce alcohol consumption among adolescents. Integrated attention and support from society, school, and family are necessary so that they can cope with and adapt to stress in a healthy way. ## 1. Introduction Drinking that begins in adolescence is associated with long-term effects, such as decreased memory and attention [1]. Neurotoxic effects are one of the many types of adverse effects of consuming alcohol at a developmentally vulnerable age, and this can cause changes in brain functions, increased susceptibility to diseases in adulthood, difficulties with emotional regulation, increased risky sexual behavior, fighting, and physical aggression [2]. If this drinking behavior continues during adolescence, it can lead to difficulties in academic achievement and negatively affect every life. Owing to the 2019 Coronavirus Disease (COVID-19) pandemic, social restrictions and controls have been implemented in many countries. School closures and transitions to online classes have led to an increase in psychological difficulties among adolescents, such as depression and anxiety disorders, isolation, and interpersonal problems [3,4]. Adverse health behaviors have also been observed since the pandemic, such as smoking and alcohol use [5]. Subsequently, drinking alcohol is highly related to depression and behavioral problems, such as attention-deficit hyperactivity disorder (ADHD), which can occur when children or adolescents begin substance abuse early [6]. This also causes emotional problems, such as aggression, impulsivity, anxiety, and excitement, and is more likely to cause eating disorders than smoking [1,7]. Therefore, it is important to investigate adolescents’ mental health and substance use, such as smoking and drinking, during the pandemic. Substance use is strongly associated with attempted and completed suicide [8,9]. Recently, the use of conventional cigarettes and e-cigarettes by adolescents has risen, and adolescents who drink frequently are more likely to use e-cigarettes [10], indicating an urgent need for preventive measures against substance abuse. Many countries have conducted annual youth health behavior surveys among adolescents because of these risks. South Korea, specifically, has a higher prevalence of alcohol use than other countries and a culture of tolerance for drinking; greater access to alcohol incurs large social costs across the country [11,12]. Roughly half of all Korean adolescents have experienced alcohol consumption, and the prevalence of binge drinking among them has increased [10]. Accordingly, this study identifies the effects of general and mental health characteristics and smoking-related behaviors on adolescents’ drinking behaviors. This study utilized secondary data on youth health behaviors collected by the Korean government, with the aim of gaining insights into nursing interventions and alcohol abstinence education targeting adolescents. The specific research objectives were as follows:[1]Identify general characteristics, mental health characteristics, smoking behavior characteristics, and drinking experiences of adolescents.[2]Identify the factors that affect the drinking experience of adolescents. ## 2.1. Study Design Since 2005, the Korea Centers for Disease Control and Prevention (KCDC) has been conducting the Korean Youth Risk Behavior Web-based Survey every year using a self-administered questionnaire to identify health-related behaviors, including alcohol use, smoking, and mental health status among first- to third-year high school students in Korea [13]. Secondary data from the 2021 survey were analyzed. ## 2.2. Study Population The 17th Korean Youth Risk Behaviors Web-based Survey [2021] [13] data used in this study were divided into stratified sampling distributions and sampling stages of the population of middle and high school students nationwide as of April 2021, and then the sample was selected. A total of 54,848 students ($92.9\%$) from 796 of 400 middle and 400 high schools participated in the survey. Of the participants, 5905 were smokers who answered “yes” to the question “Have you ever smoked?”. In this data (a total of 54,848 students), 17,939 students answered “yes” to the question “Have you ever had more than one glass of alcohol?”. ## 2.3. Ethical Considerations The n Youth Risk Behaviors Web-based Survey was conducted after review and approval by the Medical Research Ethics Review Committee from the KCDC regarding the scales and investigation process used for the ethical consideration of the participants. Prior consent was obtained from all participants before data collection. In addition, due to the secondary nature of the data, prior to the research, the researcher applied for an exemption from the Institutional Review Board of the institution to which the researcher belonged and was exempted from deliberation (JBNU 2022-06-014). ## 2.4. Data Analysis Data analysis was performed using IBM SPSS Statistics for Windows v. 25.0 (IBM Corporation, Armonk, NY, USA). A chi-square test was conducted to assess the differences between the demographic, mental health, and smoking-related characteristics of Korean adolescents in the sample. Additionally, multivariate logistic regression analysis was performed using these variables. The weights for each variable in this study were already reflected in the original data provided by the KCDC, and the weights were calculated by multiplying the value obtained by multiplying the reciprocal of the extraction rate and the reciprocal of the response rate by the weight post-correction rate. Moreover, the calculation of biased results was controlled through the composite sample design, and the value was calculated using the weighted average. ## 2.4.1. General Characteristics Gender was indicated as “boy” or “girl”, school type was “middle school” or “high school”, and subjective academic performance over the previous 12 months was reclassified as “high”, “middle”, or “low.” Residence area was classified into “metropolitan cities”, “small cities”, or “rural area”, and the type of residence was reclassified into “living with family” or “living with someone other than family”. Economic status was reclassified as “high”, “middle”, and “low”. ## 2.4.2. Mental Health Characteristics Mental health characteristics included depression, stress perception, suicidal ideation, and perceived sleep satisfaction. Depression experience was classified as ‘yes’ or ‘no’ to the question “Have you ever felt so sad or hopeless that you stopped your daily life for 2 weeks in the past 12 months?” Perceived stress was assessed with the question “In general, how much stress do you usually feel?” Participants were asked to answer this question on a scale of “high”, “a little”, and “none”. Suicidal ideation was classified as “yes” or “no” to the question “Have you ever seriously thought of suicide in the past 12 months?” Perceived sleep satisfaction was reclassified into “sufficient”, “average”, and “insufficient” in response to “Do you think that you get adequate sleep to recover from fatigue on weekdays?”. ## 2.4.3. Smoking-Related Characteristics Characteristics of smoking behavior included the number of smoking days in the last 30 days and the time of first smoking. The number of smoking days in the last 30 days was classified as “non-smoking”, “1–2 days per month”, “3–5 days per month”, and “≥ 6 days per month”; the first smoking period was classified as “elementary school”, “middle school”, and “high school”. ## 3.1. Characteristics of the Participants The sociodemographic characteristics of this study were as follows: The gender of the sample showed a high proportion of boys, and most of the participants were high school students. In addition, $60.3\%$ of students reported living in small cities or rural areas, and $46.6\%$ of students reported belonging to middle-income households in terms of economic status. Overall, $40.4\%$ of all participants experienced depression, $45.9\%$ of students reported they were highly aware of stress, and $21.5\%$ of students reported that they had suicidal thoughts. More than half of the students had insufficient sleep. For smoking-related characteristics, only $28.8\%$ of participants smoked only once. The time of smoking initiation was the highest in middle school. Further details regarding the general characteristics of the study participants are shown in Table 1. ## 3.2. Differences in Drinking Experience According to the Characteristics of the Participants Table 2 shows the chi-square results of the general characteristics of this sample. For gender, it was found that there were more participants in the drinking experience group than those who had no drinking experience, which was statistically significant ($t = 12.22$, $p \leq 0.001$). In terms of school grade, there were more with drinking experience than those without drinking experience, which was statistically significant ($t = 147.03$, $p \leq 0.001$). In the case of academic achievement, those in the drinking experience group were higher than those in the group without drinking experience, and it was statistically significant (F§ = 4.83, $p \leq 0.001$). There was a statistically significant difference in the area of residence between the drinking and non-drinking groups ($t = 4.83$, $$p \leq 0.028$$). There was a statistically significant difference in economic status between the drinking and non-drinking groups (F§ = 5.85, $$p \leq 0.003$$). ## 3.3. The Effects of Adolescent Mental Health and Smoking Characteristics on Drinking Experience A multiple logistic regression analysis was conducted to identify the effects of the general characteristics, mental health, and smoking-related characteristics of adolescents on their drinking experience (Table 3). The following general characteristics were statistically significant: gender (boys) (OR = 1.98, CI: 1.77–2.16), middle school students (OR = 1.22, CI: 1.09–1.37), and low academic performance (OR = 2.18, CI: 1.94–2.45). Students with these characteristics showed an increased possibility of drinking. The p values were <0.001, <0.001, and 0.003, respectively. Statistically significant factors in mental health characteristics were as follows. Students who responded that they were depressed (OR = 1.17, CI: 0.87–1.57) showed an increased possibility of drinking ($p \leq 0.001$). In contrast, students who had sufficient sleep (OR = 0.57, CI: 0.41–0.84) and students who slept an average amount (OR = 0.74, CI: 0.55–0.99) showed statistically significantly lower rates of drinking behavior. The p-values were 0.004 and 0.040, respectively. Smoking 1–2 times a month (OR = 1.48, CI: 1.04–2.11), 3–5 times a month (OR = 1.54, CI: 0.91–2.60), and over 6 times a month (OR = 1.421, CI: 1.04–1.84) significantly predicted an increased possibility of drinking. The p-values were 0.029, 0.004, and < 0.001, respectively. In addition, starting smoking in middle school (OR = 2.55, CI: 1.73–3.75) and high school (OR = 2.67, CI: 1.65–4.31) significantly predicted drinking behavior ($p \leq 0.001$). ## 4. Discussion This study analyzed the factors influencing drinking experience based on data from the Korean Youth Risk Behavior Web-based Survey conducted by the KCDC in 2021. It was found that gender, school year, and academic performance were statistically significant predictors of drinking experience. Some studies have already reported that drinking behaviors were higher in male students [2,14]. This study found evidence consistent with this, possibly because female students had more positive attitudes than male students toward the early prevention of drinking or smoking [15]. However, in some recent studies, there has been little gender difference in drinking and smoking problems [16]. In 2020, blackout experiences of Korean adolescents due to binge drinking were higher among female than male students [17], and the prevalence of alcohol use among young Korean women was reportedly increasing [18]. Education on abstaining from alcohol consumption should be implemented without gender discrimination. Results also indicated that the proportion of high school students who had experienced drinking was high, supporting a previous study showing that the risk of drinking increases with age [2]. However, in Korea, the age of first-time drinkers is decreasing [12]. The possibility of drinking in middle school students was found to be high, which suggests that all periods of adolescence are exposed to the risk of drinking. As such, it appears that strong state sanctions, such as the lockdown or social distancing during the pandemic, did not have a significant impact on the decline in alcohol consumption among students. Furthermore, students with low grades drank more and were predicted to drink more. The highest stress among Korean adolescents was academic [19]. In Korea, which has a world-class educational zeal and college entrance rate, academic achievement can be a stressful factor at home and school. As this is a period of high interest in academic achievement, it is necessary to educate students about the effects of alcohol on the brain and its adverse effect on learning functions. The influencing factor of residential area was not significant, but the drinking rate among students in small cities and rural areas was high. This was similar to the findings that drinking and binge drinking were higher among young people living in rural areas than in urban areas [20,21]. Students from middle and lower economic statuses also had a high drinking rate, which was linked to the high prevalence of alcohol use disorder among Korean adults in the low-income group [18]. In contrast, some studies have reported that alcohol and tobacco consumption rank high among public health priorities for young adolescents in affluent countries [14]. The Korean economy is the world’s 10th largest in terms of scale and 18th largest in per capita income among the Organization for Economic Co-operation and Development (OECD) countries [22]. However, in this study, economic status was not a statistically significant factor influencing adolescents’ drinking behaviors. Among the mental health characteristics, depression was statistically significant in drinking experience and possibility, but suicidal ideation was not. Despite the results of this study, suicide attempts are higher when substance abuse begins at an early age [23]. In South Korea, suicide was reported as the number one cause of death among adolescents in 2021, during the pandemic, and suicide has been the number one cause of death among adolescents in the last nine years [24]. Therefore, early detection of depression can be achieved by assessing mental health during drinking interventions in adolescents. This will help prevent drinking, which is a risk factor for suicide, and provide adolescents with the tools to positively cope with risks to their mental well-being. Among the mental health characteristics, students who did not get enough sleep had a significantly higher drinking experience rate and a higher possibility of drinking experience. Insomnia rates among Korean adolescents are extremely high, mainly due to excessive academic pressure [25]. Sleep problems can continue to occur during adolescence when trying to keep up with schoolwork. Short sleep times and sleep dissatisfaction in adolescents are also associated with hypertension, obesity, depression, and suicidal ideation [26]. It has been reported that the main cause of relapse in patients with alcohol use disorder is the use of alcohol as a sleep inducer [27,28]. This suggests that it is necessary to closely reflect the relationship between sleep and alcohol in educational content from scientific and intellectual points of view. The longer the number of days smoked, the higher the rate of drinking experience, which became an influencing factor. In the pre-COVID-19 surveys, high school students had a high smoking rate [29,30], consistent with this study’s findings. In Korea, smoking prevention and anti-smoking projects for teenagers started in 1999 to lower the youth smoking rate, and in 2015, they expanded to schools nationwide, while the school smoking prevention project was being implemented for all students [31]. Nevertheless, in this study, there were cases in which elementary school students also smoked, and it is necessary to pay attention to this because the factors influencing drinking among middle school students who smoked were significant. In addition, the results of this study are related to the fact that the possibility of drinking among middle school students is higher than that among high school students due to “eighth-grade syndrome”, a South Korean term indicating the negative tendency of adolescents to be rebellious and aggressive, which is the crux of a cultural joke that North Korea cannot invade South Korea because of eighth-grade syndrome [32]. The present findings can be considered an indicator of the urgent need for preventive education for lower grades. ## 5. Conclusions Recently, alcohol consumption and smoking rates among adolescents have declined globally [1,14]. South Korea also showed a similar trend, with drinking and smoking rates in 2020 and 2021 decreasing by 3–$4\%$ compared to 2019 [13]. However, this trend in adolescents was interpreted as the effect of online classes and social distancing due to the outbreak of COVID-19, rather than improved health-related behavior. In South Korea, less than $20\%$ of drinking prevention education is conducted in schools [33]; thus, it is insufficient compared to mandatory smoking education. In addition, the pandemic prevented these programs from being actively conducted. This study found that strong control measures, such as social distancing cannot control substance use. It is impossible to predict how freedom from the endemic era will affect substance use among adolescents. However, integrated attention and support from society, school, and family are necessary for adolescents to cope with and adapt to stress in a healthy way. ## References 1. Fujii M., Kuwabara Y., Kinjo A., Imamoto A., Jike M., Otsuka Y., Itani O., Kaneita Y., Minobe R., Maesato H.. Trends in the co-use of alcohol and tobacco among Japanese adolescents: Periodical nationwide cross-sectional surveys 1996–2017. BMJ Open (2021) 11 e045063. DOI: 10.1136/bmjopen-2020-045063 2. Prieto-Ursúa M., Baena B.C., Caperos J.M., Falcón C.M., Olivares J.U.. Alcohol consumption in adolescents: The predictive role of drinking motives. Psicothema (2020) 2 189-196. DOI: 10.7334/psicothema2019.263 3. Qin H., Sanders C., Prasetyo Y., Syukron M., Prentice E.. Exploring the dynamic relationships between risk perception and behavior in response to the coronavirus disease 2019 (COVID-19) outbreak. Soc. Sci. Med. (2021) 285 114267. DOI: 10.1016/j.socscimed.2021.114267 4. Wang Y., Wang Y., Chen Y., Qin Q.. Unique epidemiological and clinical features of the emerging 2019 novel coronavirus pneumonia (COVID-19) implicate special control measures. J. Med. Virol. (2020) 92 568-576. DOI: 10.1002/jmv.25748 5. Stanton R., To Q.G., Khalesi S., Williams S.L., Alley S.J., Thwaite T.L., Fenning A.S., Vandelanotte C.. Depression, anxiety, and stress during COVID-19: Associations with changes in physical activity, sleep, tobacco, and alcohol use in Australian adults. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17114065 6. Frobel W., Grafe N., Meigen C., Vogel M., Hiemisch A., Kiess W., Poulain T.. Substance use in childhood and adolescence and its associations with quality of life and behavioral strengths and difficulties. BMC Public Health (2022) 22. DOI: 10.1186/s12889-022-12586-2 7. Berro J., Akel M., Hallit S., Obeid S.. Relationships between inappropriate eating habits and problematic alcohol use, cigarette and waterpipe dependence among male adolescents in Lebanon. BMC Public Health (2021) 21. DOI: 10.1186/s12889-021-10184-2 8. Agrawal A., Tillman R., Grucza R.A., Nelson E.C., McCutcheon V.V., Few L., Conner K.R., Lynskey M.T., Dick D.M., Edenberg H.J.. Reciprocal relationships between substance use and disorders and suicidal ideation and suicide attempts in the Collaborative Study of the Genetics of Alcoholism. J. Affect. Disord. (2017) 213 96-104. DOI: 10.1016/j.jad.2016.12.060 9. Esang M., Ahmed S.A.. Closer look at substance use and suicide. Am. J. Psychiatry Resid. J. (2018) 13 6-8. DOI: 10.1176/appi.ajp-rj.2018.130603 10. Lee Y., Lee K.S.. Association of alcohol and drug use with use of electronic cigarettes and heat-not-burn tobacco products among Korean adolescents. PLoS ONE (2019) 14. DOI: 10.1371/journal.pone.0220241 11. Lee T., Lee M., Shin M., Kim S.. Effects of mental health and drinking behavior on youth smoking: Using the 2019 youth health behavior survey. Korean J. Fam. Med. (2021) 11 127-134. DOI: 10.21215/kjfp.2021.11.2.127 12. Maeng S.J., Lee D.J., Kang J.H.. First drinking experiences during adolescence in South Korea: A qualitative study focusing on the internal and external factors. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18158200 13. The Statistics on Adolescent Health-Related Behavior in South Korea. (2021) 14. Strizek J., Uhl A., Schaub M., Malischnig D.. Alcohol and cigarette use among adolescents and young adults in Austria from 2004–2020: Patterns of change and associations with socioeconomic variables. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph182413080 15. Kjeld S.G., Glenstrup S., Bast L.S.. Gender and socioeconomic disparities in reasons for not smoking cigarettes among Danish adolescents. BMC Res. Notes (2021) 14. DOI: 10.1186/s13104-021-05454-6 16. Törrönen J., Roumeliotis F., Samuelsson E., Kraus L., Room R.. Why are young people drinking less than earlier? Identifying and specifying social mechanisms with a pragmatist approach. Int. J. Drug Policy (2019) 64 13-20. DOI: 10.1016/j.drugpo.2018.12.001 17. The Statistics on Adolescent Health-Related Behavior in South Korea. (2020) 18. Lee H.K.. Epidemiology of alcohol use disorders and alcohol policy. J. Korean Neuropsychiatr. Assoc. (2019) 58 152-158. DOI: 10.4306/jknpa.2019.58.3.152 19. Park S., Jang H., Lee E.S.. Major stressors among Korean adolescents according to gender, educational level, residential area, and socioeconomic status. Int. J. Environ. Res. Public Health (2018) 15. DOI: 10.3390/ijerph15102080 20. DeHaan L., Boljevac T.. Alcohol prevalence and attitudes among adults and adolescents: Their relation to early adolescent alcohol use in rural communities. J. Child Adolesc. Subst. Abus. (2010) 19 223-243. DOI: 10.1080/1067828X.2010.488960 21. Donath C., Grässel E., Baier D., Pfeiffer C., Karagülle D., Bleich S., Hillemacher T.. Alcohol consumption and binge drinking in adolescents: Comparison of different migration backgrounds and rural vs. urban residence—A representative study. BMC Public Health (2011) 11. DOI: 10.1186/1471-2458-11-84 22. National Assembly Budget Office Economic Overview of the Republic of Korea 2022. (2022) 23. Li L., Zhao Y., Shi M., Wang Y.. Relationship between the early initiation of substance use and attempted suicide among in-school adolescents in seven low-or middle-income African countries: An analysis of the global school-based student health survey data. Front. Psychol. (2021) 12 753824. DOI: 10.3389/fpsyg.2021.753824 24. Statistics on the Youth. (2022) 25. Nam Y., Park S., Jo M., Kim C.. Mental health status of Korean adolescents according to daily sleep time and subjective sleep satisfaction. Korean J. Psychosom. Med. (2018) 26 44-50. DOI: 10.22722/KJPM.2018.26.1.44 26. Anuradha R.K., Sathyavathi R.B., Reddy T.M., Hemalatha R., Sudhakar G., Geetha P., Reddy K.K.. Effect of social and environmental determinants on overweight and obesity prevalence among adolescent school children. Indian J. Endocrinol. Metab. (2015) 19 283-287. DOI: 10.4103/2230-8210.131765 27. Brooks A.T., Krumlauf M., Beck K.H., Fryer C.S., Yang L., Ramchandani V.A., Wallen G.R.. A mixed methods examination of sleep throughout the alcohol recovery process grounded in the social cognitive theory: The role of self-efficacy and craving. Health Educ. Behav. (2019) 46 126-136. DOI: 10.1177/1090198118757820 28. Manzar M.D., Salahuddin M., Maru T.T., Dadi T.L., Abiche M.G., Abateneh D.D., Pandi-Perumal S.R., Bahammam A.S.. Sleep correlates of substance use in community-dwelling Ethiopian adults. Sleep Breath (2017) 21 1005-1011. DOI: 10.1007/s11325-017-1567-5 29. The Statistics on Adolescent Health-Related Behavior in South Korea. (2018) 30. The Statistics on Adolescent Health-Related Behavior in South Korea. (2019) 31. Gunter R., Szeto E., Jeong S.H., Suh S., Waters A.J.. Cigarette smoking in South Korea: A narrative review. Korean J. Fam. Med. (2020) 41 3-13. DOI: 10.4082/kjfm.18.0015 32. Hwang M., Yeo T., Lim H., Kim C.. Image analysis of Korea’s eighth-grader syndrome using text mining: Focusing on Korea media articles in 2010, 2015, and 2019. Educ. Res. (2022) 83 81-111. DOI: 10.17253/swueri.2022.83.005 33. The 11th Korea Youth Risk Behavior Web-Based Survey 2015. (2015)
--- title: Cathepsin H Knockdown Reverses Radioresistance of Hepatocellular Carcinoma via Metabolic Switch Followed by Apoptosis authors: - Qiao Chen - Shugen Qu - Zhenzhen Liang - Yi Liu - Huajian Chen - Shumei Ma - Xiaodong Liu journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049059 doi: 10.3390/ijms24065257 license: CC BY 4.0 --- # Cathepsin H Knockdown Reverses Radioresistance of Hepatocellular Carcinoma via Metabolic Switch Followed by Apoptosis ## Abstract Despite the wide application of radiotherapy in HCC, radiotherapy efficacy is sometimes limited due to radioresistance. Although radioresistance is reported with high glycolysis, the underlying mechanism between radioresistance and cancer metabolism, as well as the role of cathepsin H (CTSH) within it, remain unclear. In this study, tumor-bearing models and HCC cell lines were used to observe the effect of CTSH on radioresistance. Proteome mass spectrometry, followed by enrichment analysis, were used to investigate the cascades and targets regulated by CTSH. Technologies such as immunofluorescence co-localization flow cytometry and Western blot were used for further detection and verification. Through these methods, we originally found CTSH knockdown (KD) perturbed aerobic glycolysis and enhanced aerobic respiration, and thus promoted apoptosis through up-regulation and the release of proapoptotic factors such as AIFM1, HTRA2, and DIABLO, consequently reducing radioresistance. We also found that CTSH, together with its regulatory targets (such as PFKL, HK2, LDH, and AIFM1), was correlated with tumorigenesis and poor prognosis. In summary, our study found that the cancer metabolic switch and apoptosis were regulated by CTSH signaling, leading to the occurrence of radioresistance in HCC cells and suggesting the potential value of HCC diagnosis and therapy. ## 1. Introduction Hepatocellular carcinoma (HCC) is one of the most common malignancies with a poor prognosis. Radiotherapy (RT) is widely applied in HCC patients, but its efficacy is limited due to inherent or acquired radioresistance [1,2,3]. Recently, evidence has shown that the cancer metabolic switch from aerobic respiration to glycolysis (with predominance in the Warburg effect) promotes radioresistance in several cancer types [4,5,6,7,8,9,10]. However, how the cancer metabolic switch regulates radiation reactivity in liver cancer remains unclear. Aberrant metabolism is a major hallmark of cancer [11]. In order to gain a survival advantage, cancer cells flexibly develop abnormal metabolic networks through different pathways [12]; a core component is the Warburg effect (aerobic glycolysis as the predominance). Although glycolysis is less efficient at producing ATP than aerobic respiration, intermediate glucose metabolism between aerobic respiration and aerobic glycolysis can shunt anabolic pathways such as lipid synthesis (adipogenesis), amino acid production, and nucleotide synthesis, thereby promoting tumor cells’ sustainable proliferation and progression and avoiding the occurrence of apoptosis [13,14]. Moreover, high-level aerobic glycolytic metabolism also helps cells avoid massive reactive oxygen species (ROS) accumulation without compromising the cancer cells’ energy demand and thus keeping them away from overloaded oxidative stress, which is another main cause of apoptosis [14,15,16]. Therefore, glycolysis is often used as the metabolic basis in tumors. Although several studies have reported radioresistance with high rates of glycolysis [17,18,19], the exact metabolic mechanism of ionizing radiation (IR) responsiveness remains unclear. IR triggers various forms of cell death, including apoptosis. After IR treatment, second messengers and damaged DNA can mediate apoptosis [20,21]. However, how cancer metabolites are involved in IR-induced apoptosis remains elusive. The family of cathepsins is involved in overall protein turnover and specific cellular processes, such as apoptosis, antigen presentation, and prohormone processing [22,23]. Despite being implicated in cancer development [24] and the processing of neurotransmitters [25], the role of CTSH, a ubiquitously expressed lysosomal cysteine protease in cancer apoptotic process, is rarely reported [26,27,28]. To solve the elusive questions above, here, we outline a comprehensive CTSH-involved metabolic mechanism to regulate apoptosis and radioresistance, suggesting novel therapeutic strategies in radiation sensitization via disrupting cancer metabolism. ## 2.1. Radiation-Inhibited Tumor Growth and Induced Apoptosis In Vivo First, IR treatment was applied to rats bearing an orthotopic tumor in the liver. Compared to the control group, tumor growth was significantly inhibited in IR treatment groups, and this effect was more obvious when the dose was added to 9 Gy × 3 fraction (Figure 1a,b; Supplementary Figure S1). Then, to further detect the mechanism of tumor inhibition induced by IR, transcriptome sequencing was conducted. The most significantly increased genes were collected. Biological process (BP) enrichment analysis showed that the up-regulated genes after IR were highly related with cell death, especially apoptosis (Figure 1c,d). In addition, with KEGG enrichment analysis, apoptosis and the TNF pathway seemed to be closely related to these genes (Figure 1e). Consistently, immunohistochemical (IHC) staining showed positive results of Caspase3 and the apoptosis-inducing factor (AIF) after IR, which confirmed the occurrence of apoptosis (Figure 1f). ## 2.2. CTSH Participated in Radioresistance Regulation of HCC Cells Due to the non-homology between human and rat models [29], to further confirm the change of genes in HCC cells, a PCR array was performed on MHCC97H and HepG2 cell lines focusing on collected genes from in vivo experiment (Figure 2a–c, Attachment File 1). As transcription always happens earlier than translation, we examined the cell lines at 1 and 6 h after IR, and a PCR array was conducted expression profiles were established. The genes expressed consistently with in vivo sequencing results were found (Figure 2b,c). Meanwhile, trypan blue staining flow cytometry verified cell death after IR. Inhibitors of apoptosis, ferroptosis, necrosis, and autophagy pretreatment were applied to HCC cells, and IR-induced cell death was found to be reversed by the pretreatment of ZVAD (commonly used as an apoptosis inhibitor). With the pretreatment of ZVAD, the value of the MHCC97H cell death signal peak was significantly reduced after IR (Figure 2d right), while in other groups, the rescue effect was not evident (Supplementary Figure S2A). Previous results had already informed us of the involved genes in vivo (Figure 1c–e). After filtering out genes expressed inconsistently with the in vivo results or those already well-reported [22,30,31,32], we mainly focused on CTSD, CTSH and FAS (marked with * in Figure 2b,c; Supplementary Figure S2B). Then, due to the low abundance of FAS and the inconsistency of CTSD expression (Figure 2e; Supplementary Figure S2C), we abandoned FAS, CTSD and selected CTSH for further study. To further understand the influence of CTSH, a CTSH knockdown model was established in HepG2 cells (Figure 2f), and the radio-sensitivity was examined using a colony formation assay. The results showed that knockdown (KD) of CTSH significantly increased the radio-sensitivity of HepG2 cells (Figure 2g,h). Consistently, MHCC97H cells showed significantly higher radio-sensitivity compared with HepG2 (Figure 2i,j), with a lower CTSH expression after IR stimulation (Figure 2e). Taking these results together, we realized that CTSH participated in the maintenance of radioresistance in HCC cells. ## 2.3. Restrained Glycolysis and Promoted Aerobic Respiration Inhibited Radioresistance of HepG2 Cells To further understand how CTSH performs in regulating HCC metabolism and radiation resistance, proteome mass spectrometry (MS) was performed. As the MS results showed, 360 proteins were up-regulated and 430 were down-regulated after CTSH knockdown (Figure 3a,b). Further enrichment analysis showed the up-regulated proteins mainly localized in the mitochondria (Figure 3c). In addition, these up-regulated proteins were related to mitochondrial function and aerobic respiration, while the down-regulated ones were associated with biosynthesis and metabolic processes (Figure 3d,e). As mitochondria are known to play an important role in metabolism and apoptosis [33] and the cancer metabolic switch is important for cancer cell fate [34], we wondered whether CTSH was affecting cell fate by regulating the metabolic switch. Remarkably, more detailed GSEA results confirmed our hypothesis; after CTSH knockdown, the glycolytic metabolism was inhibited, while the aerobic respiration was promoted (Figure 3f,g). Important molecules for glycolysis such as HK2, PFKL, PKM, and LDHA were all down-regulated. Furthermore, key factors of the tricarboxylic acid cycle (TCA), such as CS, OGDH, and IDH, and oxidative phosphorylation (OXPHOS)-promoting factors, such as AIFM1 and CYC1, were up-regulated (Figure 3h–j). At the same time, mitochondrion pyruvate carrier MPC1 increased, indicating an increase in mitochondria pyruvate intake. Thus, enhanced TCA and OXPHOS formed a strong up-regulation of the aerobic respiration cascade (Figure 3h–j). Considering the reported association of glycolysis with radioresistance, the above results indicate that CTSH knockdown inhibited the radioresistance of HepG2 cells by perturbing glycolysis and reversing the cancer metabolism to aerobic respiration (Figure 3k). ## 2.4. Knockdown of CTSH and Enhanced Aerobic Respiration Promoted Radiation-Induced Apoptosis via IAP Inhibition and AIF Signal As the involvement of CTSH in apoptosis was found by our bioinformatic analysis (Figure 1d,e), to directly confirm this, an Annexin V−PI staining flow cytometry was performed. The results showed that apoptosis was significantly promoted by CTSH knockdown in both IR and NC conditions (Figure 4a,b). Referring to our previous results (Figure 2d) and the up-regulation of AIFM1 (an apoptosis-inducing factor) (Figure 3j), to determine which signaling was involved during the proapoptotic process, we analyzed the MS results using GSEA. The results suggested a promotion of apoptosis; among these, many apoptotic-related genes were affected after CTSH knockdown (Figure 4c,d). Thus, to determine the target of CTSH apoptotic regulation through mitochondrial signaling, an intersection of apoptosis and mitochondrial-related genes was taken; a total of 14 genes, including 9 up-regulated and 5 down-regulated, were listed (Figure 4e,f). Among them, HTRA2 and DIABLO were up-regulated (Figure 4g,h), and they were both reported to promote the apoptotic process as inhibitors of apoptosis (IAPs) [35,36,37]. An immunofluorescence co-localization assay further verified our findings; compared to the blank-load transfection group (ShVec) and the control group (NC), HTRA2 and DIABLO increased in the CTSH knockdown group and inhibited the activity of the IAPs (XIAP and Survivin) in irradiated CTSH knockdown cells, and the inhibition of the IAPs occurred within both the cell nucleus and the cytoplasm (Figure 4i–k; Supplementary Figure S3). Accumulating data support that tumor suppression may be achieved by inhibiting glycolysis and promoting OXPHOS (part of the aerobic respiratory chain) [35]. As a killer protein, besides promoting OXPHOS, the up-regulated AIFM1 (Figure 4g) is also known to induce DNA fragmentation during the apoptotic process [38,39,40,41], suggesting a proapoptotic effect based on the up-regulated aerobic respiration cascade. Furthermore, the increased (activated) caspase family (Caspase9 and Caspase3 cleavage) expression finally confirmed these connections (Figure 4l). From these results, we realized that the proapoptotic effect of CTSH knockdown is carried out through promoted IAP inhibition and enhanced aerobic respiration, and this process is executed by HTRA2, DIABLO, and AIF signaling (Figure 4m). ## 2.5. CTSH Knockdown Changed Mitochondrial Membrane Permeability and Stability in Proapoptotic Signaling Apoptosis includes intrinsic (mitochondrial) and extrinsic (death receptor) mechanisms. As previous results had suggested the involvement of mitochondrial genes (Figure 3d,e and Figure 4c,d), we decided to focus on mitochondrial dysfunction and apoptosis for further study. Remarkably, the GSEA results showed an increased cascade of the transmembrane transport of mitochondria and cytochrome c release (Figure 5a). Among those, VDAC formed a channel through the mitochondrial outer membrane and allowed the diffusion of hydrophilic molecules. It opened at low or zero membrane potential and closed at potentials above 30–40 mV [42]. Fam162A, FIS1, and OPA1 were reported to be involved in proapoptotic factor release (such as cytochrome c), caspase activation (such as CASP9), and mitochondrial permeability transition induction [43,44,45]. Interestingly, after CTSH knockdown, all these genes were up-regulated (Figure 5b). Besides glycolysis, HK2 also plays a role in maintaining the integrity of the outer mitochondrial membrane and preventing the release of apoptogenic molecules from the intermembrane space and subsequent apoptosis [33]. In cancer cells, it binds to and inhibits VDAC to suppress mitochondrial function while stimulating glycolysis [46], and it has also been observed to be down-regulated (Figure 3j). All these results indicated an increase in the permeability and instability of the mitochondrial membrane. Mitochondrial membrane potential (MMP) examination after IR further confirmed the participation of these molecules (Figure 5c,d); a decrease in MMP (IR-induced) and membrane stability (CTSH KD−induced) made it easier for molecules to be released from the mitochondria after IR, thus promoting the apoptotic process. The above results suggested an explanation responsible for the increased apoptotic flux; CTSH knockdown promoted the release of some proapoptotic factors after IR through modulating the permeability and stability of the mitochondrial membrane (Figure 5e). ## 2.6. CTSH and Targets Were Correlated with Tumorigenesis and Poor Prognosis Then, to verify the potential of CTSH in clinical application, a series of bioinformatic analyses were performed using well-known databases. For the mRNA level, CTSH showed significantly higher expression in HCC than non-tumor samples (Figure 6a) (derived from GEPIA). In addition, in a clinical cohort of 370 HCC patients, using a validated mRNA signature, higher CTSH signaling was associated with poor prognosis (p-value = 0.025) (Figure 6b) (derived from Kaplan–Meier plotter). A survival Kaplan–Meier (KM) analysis of 365 patients showed similar results in the CTSH protein level (Figure 6c) (from The Human Protein Atlas). A multi-gene expression comparison of CTSH targets, especially those glycolysis-related genes, showed homogeneous trends related to tumorigenesis (Figure 6d) (from TNMplot). Furthermore, by combining CTSH expression and its downstream targets expressions (i.e., CTSH combined with PFKL, PKM, LDHA, HK2; and low IDH, MPC1, AIFM1, and HTRA2 combined with high CTSH expression), we found a remarkably significant facilitation of tumorigenesis and poor prognosis in HCC (Figure 6e) (Kaplan–Meier plotter). Then, to verify the universality of the above findings, profiles of CTSH expression were drawn out using two different databases (Figure 6f and Figure S4A) (TNMplot and GEPIA). Among them, human malignancies were selected because of their high CTSH tumor expression. Remarkably, CTSH was related with poor prognosis in all these cancers (Figure 6g). Furthermore, survival KM plotting showed consistent facilitations of poor prognosis in malignancies such as cervical squamous-cell carcinoma, esophageal carcinoma, and pancreatic adenocarcinoma (Figure 6h,i and Figure S4B,C) (Kaplan–Meier plotter). However, in other low-grade malignancies, this relationship seems insignificant or even reversed (Supplementary Figure S4D). Taken together, these findings implied significant therapeutic and diagnostic potential for clinical use. ## 3. Discussion RT is increasingly used in advanced HCC and has been reported to confer survival benefits [1,2,3]. Nevertheless, radioresistance has been a major hurdle. Several mechanisms of radioresistance encompassing different molecular pathways in HCC have been suggested [47,48,49,50,51], but most of the previous studies have mainly addressed one single signaling in mediating resistance and have not illustrated a general mechanism shared by different resistant backgrounds. To solve these problems, in the current study, we first attempted to characterize the radio-sensitivities of different HCC cells by attributing them to the differing CTSH-mediated metabolic styles, with the aim of better delineating the mechanism underlying radioresistance encountered in clinical practice. ## 3.1. CTSH-Modulated Metabolic Switch from Aerobic Respiration to Glycolysis Is Important for Radioresistance of HCC It has recently been reported that the cancer metabolic switch plays a role in resistance after treatment, such as through highly activated glycolysis, enhanced lipogenesis or fatty acid beta-oxidation, and increased nucleotide metabolism [4,5,6,7,8,9,10,47]. Due to the critical role of glycolysis in tumorigenesis [13], accumulating data support the idea that tumor suppression may be achieved by inhibiting aerobic glycolysis and promoting oxidative phosphorylation [52]. Although metabolic changes and cancer microenvironments have been the focus of recent attention, the internal mechanism after HCC radiotherapy remains unclear. Despite the report that CTSH increased the chemoresistance of bladder cancer [53], here, for the first time, we have described a CTSH-mediated survival mechanism after IR in HCC cells that substantially contributes to radioresistance. In this context, relatively radioresistant HCC cells were highly addicted to glycolytic metabolism, but when CTSH was knocked down, this cancer metabolic switch was significantly reversed (Figure 2g,h; Figure 3f,g). Here, glucose did not feed the Warburg mainstream aerobic glycolysis to meet the biosynthesis need for cell proliferation, while the cascade of the aerobic respiration cascade was up-regulated (pyruvate intake was also enhanced by MPC1 elevation), which was a strong destruction of the cellular immortality of HCC. With decreased biosynthesis supply and increased oxidative stress brought by aerobic respiration (OXPHOS), the radioresistance was reversed in HCC cells. Aberrant gene expression is a hallmark of cancer, and cancer cells are always under tremendous pressure to be selected for the pro-survival genetic phenotype [13]. Therefore, our discovery can explain the paradox that CTSH is highly expressed and associated with poor prognosis in certain types of human cancer, such as HCC, but has relatively low expression in other cancers. The high CTSH expression cancers are likely to be highly malignant. In this context, CTSH appears to be important to maintain cancer metabolism in cells with higher glycolytic activity but it is not so important for cancer cells that have less glycolytic activity. Furthermore, the up-regulation of the CTSH protein in HepG2 cells after IR gave us a definitive answer to explain the phenomenon that the HepG2 cell was more resistant to radiation than the MHCC97H cell, in which the CTSH expression did not increase as significantly as in the HepG2 cell (Figure 2e,i,j). From this perspective, we found a novel potential mechanism of CTSH to manipulate HCC metabolism and cell death. ## 3.2. CTSH Knockdown Promotes Apoptosis Following Reversed Metabolic Switch CTSH is a member of the cathepsin family and functions mostly as an endopeptidase important for the overall degradation of proteins in lysosomes. Despite previous reports that the cathepsin family promotes apoptosis [54,55,56], it has only been reported to protect insulin-secreting and immune β-cells against cytokine-induced apoptosis by perturbing the JNK pathway in type-1 diabetes [57,58]. However, whether CTSH is involved in the cell death of cancer still remained unclear. Here, firstly, we originally found that CTSH knockdown could facilitate the apoptotic process through mitochondrial signaling (such as AIFM1 signaling) in HCC cells. Besides the proapoptotic effect, AIFM1 is also known as an OXPHOS-promoting factor [59]. As OXPHOS promotes the accumulation of oxidative stress and thus the apoptotic cascade [14,15,16], the reversed cancer metabolic switch is related to apoptosis. Secondly, we discovered that CTSH knockdown could facilitate the apoptotic process by inhibiting the activity of the IAP family rather than directly activating the caspase family (although the latter is more widely recognized [60]). Thirdly, a decrease in MMP (IR-induced) and membrane stability (CTSH KD-induced) synergistically combined to promote apoptosis after IR release of the above molecules. In summary, this proapoptotic effect was conducted via not only a normal caspase-dependent pathway but also an independent pathway (AIF signaling) via aerobic respiration enhancement. ## 3.3. CTSH Together with Its Targets Show Potential Value in HCC Treatment Consistent with these laboratorial results, our clinical bioinformatic analysis also revealed that the CTSH-mediated cancer metabolic switch (Warburg effect) likely contributed to tumorigenesis and was correlated to poor prognosis in clinical HCC cohorts on both a transcriptome and a proteome level. In addition, this mechanism likely acted as a common feature among patients with higher CTSH levels in other cancers. As studies on inhibitors of the cathepsin family have been relatively well developed [61,62], and some of them are even Food and Drug Administration approved for other indications [63,64], they become possible options for clinical application. Therefore, these findings implicate the significant potential of translational medicine for diagnosis and therapy in HCC and other IR-resisting cancer types. In summary, our study describes an integration of HCC metabolism regulated by CTSH signaling, which mediates radioresistance through promoting glycolysis and inhibiting apoptosis due to the impairment of the release and expression of proapoptotic proteins in HCC cells. With the gradual blossoming of RT in HCC treatment, understanding such tumor metabolic vulnerability and elaborating the underlying mechanism may help to determine more effective combination therapeutic strategies for patients with HCC. ## Limitations of the Study In this study, we elucidated a comprehensive CTSH-involved cancer metabolism that inhibits apoptosis in driving radioresistance. However, the direct CTSH regulatory targets affecting the metabolic and apoptotic processes still need further study. More research will be accomplished in the future. In addition, more experiments in vivo will be conducted to further verify our findings. Prospectively, these works will be meaningful for therapeutic purposes in antagonizing HCC radioresistance. ## 4.1. Reagents and Antibodies An apoptosis inhibitor ZVAD (HY-16658) was obtained from MCE (Princeton, NJ, USA). Primary antibodies CTSH (sc-398527), FAS (sc-8009), Survivin (sc-17779), XIAP (sc-55550), DIABLO (sc-393118), and HTRA2 (sc-365594) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA); GAPDH (#5174S), AIF (#7495S), CASP9 (#9502), CASP3 (#9662), and TOM20 (42406S) were purchased from Cell Signaling Technology (Danvers, CO, USA); and anti-actin was purchased from Sigma Aldrich (Saint Louis, MO, USA, prod. no. A3853). Secondary antibodies goat antirabbit IgG- (H+L) HRP conjugate (cat. no. 170-6515) and goat anti-mouse IgG- (H+L) HRP conjugate (cat. no. 170-6516) were obtained from Bio-Rad Laboratories (Mississauga, ON, Canada). ## 4.2. Western Blot Cells were lysed in lysis buffer containing protease inhibitors (aprotinin, leupeptin, and PMSF) on ice, and centrifuged at 10,000× g for protein collection. Intestinal protein extracts were separated by $12\%$ SDS-PAGE and transferred to nitrocellulose membranes using the Protean Mini Cell (Bio-Rad). After completion of the transfer, membranes were blocked with $5\%$ nonfat milk in TBS/$0.1\%$ Tween 20 for 120 min. Incubation with the primary antibody (as indicated) was conducted overnight at 4 °C. Incubation with a peroxidase conjugated anti-mouse or anti-rabbit secondary antibody (1: 10,000) was performed for 120 min at room temperature. Finally, chemiluminescent analysis was performed. ## 4.3. Cell Lines and Cell Culture Human HCC cell lines HepG2, MHCC97H, and Huh-7 were obtained from the Chinese Academy of Sciences cell bank (Beijing, China). All cells were routinely cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Sigma, Saint Louis, MO, USA) containing $10\%$ fetal bovine serum (Solarbio, Beijing, China) at 37 °C in a humidified atmosphere of $5\%$ CO2. ## 4.4. Irradiation For experiments in vitro, an X-ray generator (X-RAD 320 ix, Precision X-ray Inc., North Branford, CT, USA) was utilized to deliver radiation at a dose rate of 300 cGy/min. The irradiation conditions were as follows: 20 kV, 12.5mA, filter 1, SSD 70 cm; the fractionated dose was 7Gy/9Gy/12Gy × 3 fraction and the single dose was 15 Gy in cells. For experiments in vivo, an X-ray accelerator (Clinac 23EX, Varian Medical Systems, Inc., Palo Alto, CA, USA) was utilized to deliver radiation at a dose rate of 400 MU/min. The irradiation conditions were as follows: 6MV, distance (SSD) 96.1 cm. For the imitation of the clinical SBRT condition, the dose was 7Gy/9Gy × 3 fraction. ## 4.5. Analysis of Mitochondrial Membrane Potential (MMP) Mitochondrial membrane potential was detected by a DIOC6 (D273, Invitrogen, USA) probe according to the manufacturer’s instructions. Cells were seeded in p35 plates with a density of 2.5 × 104/mL with 3 parallel wells in each group. After various treatments, cells were stained with a DIOC6 probe at a final concentration of 5 nM at 37 °C in the dark for 25 min. Then, cells were washed with PBS three times and were analyzed using a flow cytometer (ACEA NovoCyte 2040R, ACEA Biosciences, Inc., San Diego, CA, USA); the percentage of M2-2 was used to represent the membrane potential change and for analysis. ## 4.6. Analysis of Apoptosis by Cytometry Apoptosis was examined by Annexin V (BD, #556420, Franklin Lakes, NJ, USA) staining according to the manufacturer’s instructions. Cells were seeded in p35 plates with a density of 2.5 × 104/mL with 3 parallel wells in each group. After various treatments, cells were divided and stained with /PI at 37 °C in the dark for 20 min. Then, cells were washed with PBS three times, and FITC (Annexin V apoptosis signal) was detected and analyzed using a flow cytometer (ACEA NovoCyte 2040R, ACEA Biosciences, Inc., San Diego, CA, USA). Quadrants 4-2 and 4-4 were used to represent the apoptosis rate. ## 4.7. Lentiviral Production Lentiviral short hairpin RNA (shRNA) vector-targeting CTSH (pLKO.1-shCTSH) was constructed according to the Oxidative Medicine and Cellular Longevity protocol of pLKO.1-blasticidin vector (Addgene, Cambridge, MA, USA). Then, the forward oligo and reverse oligo were annealed and inserted into the pLKO.1-blasticidin vector: CTSH shRNA sequence: forward oligo: CCGG—GACGCAAAGATCACCAGCCAT—CTCGAG—ATGGCTGGTGATCTTTGCGTC—TTTTTG; reverse oligo: AATTCAAAAA—GACGCAAAGATCACCAGCCAT—CTCGAG—ATGGCTGGTGATCTTTGCGTC. ## 4.8. Transfection Target fragments were inserted into lentiviral vectors pLKO.1. All plasmids were verified using DNA sequencing. Together with pMD2G and psPAX2 plasmids, recombinant lentiviral plasmids were transfected into HEK293T cells, in which recombinant lentivirus was generated, and then virus-containing supernatant was collected 48 h after transfection. Target cells were incubated in lentivirus supernatant supplemented with 10 μg/mL polybrene (Sigma-Aldrich, H9268, Saint Louis, MO, USA), and then followed by drug selection with 7 μg/ mL Blasticidin for 7 to 14 days. After the efficiency of knockdown was confirmed via Western blot, surviving cells were used for further experiments. ## 4.9. Cell-Death Analysis by Trypan Blue Staining Trypan blue (Solarbio, Beijing, China) was used according to the manufacturer’s protocol. Cells were seeded in 6-well plates (3 × 103 cells/well) and collected, then analyzed after 15 min of trypan blue incubation using a flow cytometer (NovoCyte 2040R, ACEA Biosciences, Inc., San Diego, CA, USA). M2-2 was used to represent the cell-death portion and for analysis. ## 4.10. Immunofluorescence Co-Localization Assay A total of 48 h after transfection, cells were separated into plates, and slides were added before being divided. The cells were collected when they had completely adhered to the coverslip, fixed with $4\%$ paraformaldehyde for 20 min, and then subjected to $0.1\%$ TritonX-100 treatment for 15 min. After blocking the cells with $10\%$ normal non-immune goat serum for 1 h, the cells were treated with the appropriate primary antibodies, HTRA2, DIABLO, BIRC2, XIAP, and TOM20, overnight. The next day, cells were incubated with the corresponding secondary fluorescent antibodies for 1 h and then washed with PBS-tween for 5 min × 3 times. The slides were dried at room temperature avoiding light, then antifade solution was added. ## 4.11. Orthotopic Tumor Model Establishment A rat liver W256 carcinosarcoma tumor-bearing model was used in our preliminary experiment. Because its blood supply characteristics and growth behavior are similar to human liver cancer (mainly supplied by the hepatic artery), this model is widely used in laboratory therapy, imaging diagnosis, invasion and metastasis of liver cancer intervention, and other studies [29,65,66,67,68,69]. Cells collected from ascites were centrifugated and washed with 1 × PBS and then suspended to the required concentration (1 × 107 cells/mL). Six-week-old male Wistar rats weighing 160–180 g were used. After abdominal anesthesia with $5\%$ chloral-hydrate and iodophor disinfection, a 1–1.5 cm vertical incision was made in the left-upper-abdomen below the ribs. Then, the left liver lobe was squeezed out of the incision, and w256 cells were injected into the lobe slowly (about 1.5 × 106 cells in 0.125 mL, using a micro-syringe). Care was taken to ensure that the cells did not flow back out after the needle was pulled out, because this would cause lethal ascites). After resetting the liver and closing the abdomen, the incision was disinfected. Radiotherapy was initiated about 12 days after the establishment. ## 4.12. Bioinformatic Analyses GSEA was conducted using the GSEA official App. The bioinformatic database analyses conducted in this study are listed here: GEPIA (http://gepia.cancer-pku.cn/, accessed on 18 November 2022); Kaplan-Meier Plotter (https://kmplot.com/analysis/, accessed on 18 November 2022); Protein atlas (https://www.proteinatlas.org/, accessed on 18 November 2022); TNMplot (https://tnmplot.com/analysis/, accessed on 18 November 2022). ## 4.13. Statistical Analyses Statistical analyses between groups (e.g., Figure 2d,h,j, Figure 4b, and Figure 5c,d) were performed using the t test unless otherwise stated. p values of * $p \leq 0.1$, $p \leq 0.05$, * $p \leq 0.01$ were considered statistically significant. The unpaired t test was performed to compare between groups during Trypan blue cell death, MMP, and Annexin-V flow-cytometry. The paired t test was performed in colony-formation assay analysis. All statistical tests were performed in GraphPad Prism (version 9.0, Boston, MA, USA). The response variables were log-transformed to avoid skewness of the residuals, which also resulted in per-allele effects expressing percentagewise and not absolute changes of the response variable. ## 5. Conclusions In summary, our study describes the integration of HCC metabolism regulated by CTSH signaling, which mediates radioresistance by promoting glycolysis and inhibiting apoptosis due to the impairment of the release and expression of proapoptotic proteins in HCC cells. ## References 1. Rajyaguru D.J., Borgert A.J., Smith A.L., Thomes R.M., Conway P.D., Halfdanarson T.R., Truty M.J., Kurup A.N., Go R.S.. Radiofrequency Ablation Versus Stereotactic Body Radiotherapy for Localized Hepatocellular Carcinoma in Nonsurgically Managed Patients: Analysis of the National Cancer Database. J. Clin. Oncol. (2018) 36 600-608. DOI: 10.1200/JCO.2017.75.3228 2. Wahl D.R., Stenmark M.H., Tao Y., Pollom E.L., Caoili E.M., Lawrence T.S., Schipper M.J., Feng M.. Outcomes After Stereotactic Body Radiotherapy or Radiofrequency Ablation for Hepatocellular Carcinoma. J. Clin. Oncol. (2016) 34 452-459. DOI: 10.1200/JCO.2015.61.4925 3. Wei X., Jiang Y., Zhang X., Feng S., Zhou B., Ye X., Xing H., Xu Y., Shi J., Guo W.. Neoadjuvant Three-Dimensional Conformal Radiotherapy for Resectable Hepatocellular Carcinoma with Portal Vein Tumor Thrombus: A Randomized, Open-Label, Multicenter Controlled Study. J. Clin. Oncol. (2019) 37 2141-2151. DOI: 10.1200/JCO.18.02184 4. Oh J.M., Venters C.C., Di C., Pinto A.M., Wan L., Younis I., Cai Z., Arai C., So B.R., Duan J.. U1 snRNP regulates cancer cell migration and invasion in vitro. Nat. Commun. (2020) 11 1. DOI: 10.1038/s41467-019-13993-7 5. Tan Z., Xiao L., Tang M., Bai F., Li J., Li L., Shi F., Li N., Li Y., Du Q.. Targeting CPT1A-mediated fatty acid oxidation sensitizes nasopharyngeal carcinoma to radiation therapy. Theranostics (2018) 8 2329-2347. DOI: 10.7150/thno.21451 6. Rashmi R., Huang X., Floberg J.M., Elhammali A.E., McCormick M.L., Patti G.J., Spitz D.R., Schwarz J.K.. Radioresistant Cervical Cancers Are Sensitive to Inhibition of Glycolysis and Redox Metabolism. Cancer Res. (2018) 78 1392-1403. DOI: 10.1158/0008-5472.CAN-17-2367 7. Gunda V., Souchek J., Abrego J., Shukla S.K., Goode G.D., Vernucci E., Dasgupta A., Chaika N.V., King R.J., Li S.. MUC1-Mediated Metabolic Alterations Regulate Response to Radiotherapy in Pancreatic Cancer. Clin. Cancer Res. (2017) 23 5881-5891. DOI: 10.1158/1078-0432.CCR-17-1151 8. Zhou W., Yao Y., Scott A.J., Wilder-Romans K., Dresser J.J., Werner C.K., Sun H., Pratt D., Sajjakulnukit P., Zhao S.G.. Purine metabolism regulates DNA repair and therapy resistance in glioblastoma. Nat. Commun. (2020) 11 3811. DOI: 10.1038/s41467-020-17512-x 9. Souchek J.J., Baine M.J., Lin C., Rachagani S., Gupta S., Kaur S., Lester K., Zheng D., Chen S., Smith L.. Unbiased analysis of pancreatic cancer radiation resistance reveals cholesterol biosynthesis as a novel target for radiosensitisation. Br. J. Cancer (2014) 111 1139-1149. DOI: 10.1038/bjc.2014.385 10. Liu Y., Murray-Stewart T., Casero R.A., Kagiampakis I., Jin L., Zhang J., Wang H., Che Q., Tong H., Ke J.. Targeting hexokinase 2 inhibition promotes radiosensitization in HPV16 E7-induced cervical cancer and suppresses tumor growth. Int. J. Oncol. (2017) 50 2011-2023. DOI: 10.3892/ijo.2017.3979 11. Park J.H., Pyun W.Y., Park H.W.. Cancer Metabolism: Phenotype, Signaling and Therapeutic Targets. Cells (2020) 9. DOI: 10.3390/cells9102308 12. Pavlova N.N., Thompson C.B.. The Emerging Hallmarks of Cancer Metabolism. Cell Metab. (2016) 23 27-47. DOI: 10.1016/j.cmet.2015.12.006 13. Hanahan D., Weinberg R.A.. Hallmarks of cancer: The next generation. Cell (2011) 144 646-674. DOI: 10.1016/j.cell.2011.02.013 14. Vander Heiden M.G., Cantley L.C., Thompson C.B.. Understanding the Warburg effect: The metabolic requirements of cell proliferation. Science (2009) 324 1029-1033. DOI: 10.1126/science.1160809 15. Sosa V., Moliné T., Somoza R., Paciucci R., Kondoh H., ME L.L.. Oxidative stress and cancer: An overview. Ageing Res. Rev. (2013) 12 376-390. DOI: 10.1016/j.arr.2012.10.004 16. Hayes J.D., Dinkova-Kostova A.T., Tew K.D.. Oxidative Stress in Cancer. Cancer Cell (2020) 38 167-197. DOI: 10.1016/j.ccell.2020.06.001 17. Liu Y., He D., Xiao M., Zhu Y., Zhou J., Cao K.. Long noncoding RNA LINC00518 induces radioresistance by regulating glycolysis through an miR-33a-3p/HIF-1α negative feedback loop in melanoma. Cell Death Dis. (2021) 12 245. DOI: 10.1038/s41419-021-03523-z 18. Shimura T., Noma N., Sano Y., Ochiai Y., Oikawa T., Fukumoto M., Kunugita N.. AKT-mediated enhanced aerobic glycolysis causes acquired radioresistance by human tumor cells. Radiother. Oncol. (2014) 112 302-307. DOI: 10.1016/j.radonc.2014.07.015 19. Tomita K., Nagasawa T., Kuwahara Y., Torii S., Igarashi K., Roudkenar M.H., Roushandeh A.M., Kurimasa A., Sato T.. MiR-7-5p Is Involved in Ferroptosis Signaling and Radioresistance Thru the Generation of ROS in Radioresistant HeLa and SAS Cell Lines. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22158300 20. Kim B.M., Hong Y., Lee S., Liu P., Lim J.H., Lee Y.H., Lee T.H., Chang K.T., Hong Y.. Therapeutic Implications for Overcoming Radiation Resistance in Cancer Therapy. Int. J. Mol. Sci. (2015) 16 26880-26913. DOI: 10.3390/ijms161125991 21. Ferlini C., D’Amelio R., Scambia G.. Apoptosis induced by ionizing radiation. The biological basis of radiosensitivity. Subcell Biochem. (2002) 36 171-186. PMID: 12037980 22. Conus S., Simon H.U.. Cathepsins: Key modulators of cell death and inflammatory responses. Biochem. Pharmacol. (2008) 76 1374-1382. DOI: 10.1016/j.bcp.2008.07.041 23. Reiser J., Adair B., Reinheckel T.. Specialized roles for cysteine cathepsins in health and disease. J. Clin. Investig. (2010) 120 3421-3431. DOI: 10.1172/JCI42918 24. Gocheva V., Chen X., Peters C., Reinheckel T., Joyce J.A.. Deletion of cathepsin H perturbs angiogenic switching, vascularization and growth of tumors in a mouse model of pancreatic islet cell cancer. Biol. Chem. (2010) 391 937-945. DOI: 10.1515/bc.2010.080 25. Lu W.D., Funkelstein L., Toneff T., Reinheckel T., Peters C., Hook V.. Cathepsin H functions as an aminopeptidase in secretory vesicles for production of enkephalin and galanin peptide neurotransmitters. J. Neurochem. (2012) 122 512-522. PMID: 22582844 26. Cirman T., Oresić K., Mazovec G.D., Turk V., Reed J.C., Myers R.M., Salvesen G.S., Turk B.. Selective disruption of lysosomes in HeLa cells triggers apoptosis mediated by cleavage of Bid by multiple papain-like lysosomal cathepsins. J. Biol. Chem. (2004) 279 3578-3587. DOI: 10.1074/jbc.M308347200 27. Droga-Mazovec G., Bojic L., Petelin A., Ivanova S., Romih R., Repnik U., Salvesen G.S., Stoka V., Turk V., Turk B.. Cysteine cathepsins trigger caspase-dependent cell death through cleavage of bid and antiapoptotic Bcl-2 homologues. J. Biol. Chem. (2008) 283 19140-19150. DOI: 10.1074/jbc.M802513200 28. D’Angelo M.E., Bird P.I., Peters C., Reinheckel T., Trapani J.A., Sutton V.R.. Cathepsin H is an additional convertase of pro-granzyme B. J. Biol. Chem. (2010) 285 20514-20519. DOI: 10.1074/jbc.M109.094573 29. Zou Y., Yamagishi M., Horikoshi I., Ueno M., Gu X., Perez-Soler R.. Enhanced therapeutic effect against liver W256 carcinosarcoma with temperature-sensitive liposomal adriamycin administered into the hepatic artery. Cancer Res. (1993) 53 3046-3051. PMID: 8319211 30. Green D.R., Llambi F.. Cell Death Signaling. Cold Spring Harb. Perspect. Biol. (2015) 7 a006080. DOI: 10.1101/cshperspect.a006080 31. Xiao Y., Li H., Zhang J., Volk A., Zhang S., Wei W., Zhang S., Breslin P., Zhang J.. JTNF-α/Fas-RIP-1-induced cell death signaling separates murine hematopoietic stem cells/progenitors into 2 distinct populations. Blood (2011) 118 6057-6067. DOI: 10.1182/blood-2011-06-359448 32. Li H., Fan J., Zhao Y., Yang J., Xu H., Manthari R.K., Cheng X., Wang J., Wang J.. Calcium alleviates fluoride-induced kidney damage via FAS/FASL, TNFR/TNF, DR5/TRAIL pathways in rats. Ecotoxicol. Environ. Saf. (2021) 226 112851. DOI: 10.1016/j.ecoenv.2021.112851 33. Chiara F., Castellaro D., Marin O., Petronilli V., Brusilow W.S., Juhaszova M., Sollott S.J., Forte M., Bernardi P., Rasola A.. Hexokinase II detachment from mitochondria triggers apoptosis through the permeability transition pore independent of voltage-dependent anion channels. PLoS ONE (2008) 3. DOI: 10.1371/journal.pone.0001852 34. Kim J., Yu L., Chen W., Xu Y., Wu M., Todorova D., Tang Q., Feng B., Jiang L., He J.. Wild-Type p53 Promotes Cancer Metabolic Switch by Inducing PUMA-Dependent Suppression of Oxidative Phosphorylation. Cancer Cell (2019) 35 191-203.e8. DOI: 10.1016/j.ccell.2018.12.012 35. Bartke T., Pohl C., Pyrowolakis G., Jentsch S.. Dual role of BRUCE as an antiapoptotic IAP and a chimeric E2/E3 ubiquitin ligase. Mol. Cell (2004) 14 801-811. DOI: 10.1016/j.molcel.2004.05.018 36. Balakrishnan M.P., Cilenti L., Mashak Z., Popat P., Alnemri E.S., Zervos A.S.. THAP5 is a human cardiac-specific inhibitor of cell cycle that is cleaved by the proapoptotic Omi/HtrA2 protease during cell death. Am. J. Physiol. Heart Circ. Physiol. (2009) 297 H643-H653. DOI: 10.1152/ajpheart.00234.2009 37. Du C., Fang M., Li Y., Li L., Wang X.. Smac, a mitochondrial protein that promotes cytochrome c-dependent caspase activation by eliminating IAP inhibition. Cell (2000) 102 33-42. DOI: 10.1016/S0092-8674(00)00008-8 38. Sevrioukova I.F.. Structure/Function Relations in AIFM1 Variants Associated with Neurodegenerative Disorders. J. Mol. Biol. (2016) 428 3650-3665. DOI: 10.1016/j.jmb.2016.05.004 39. Ghezzi D., Sevrioukova I., Invernizzi F., Lamperti C., Mora M., D’Adamo P., Novara F., Zuffardi O., Uziel G., Zeviani M.. Severe X-linked mitochondrial encephalomyopathy associated with a mutation in apoptosis-inducing factor. Am. J. Hum. Genet. (2010) 86 639-649. DOI: 10.1016/j.ajhg.2010.03.002 40. Kalpage H.A., Bazylianska V., Recanati M.A., Fite A., Liu J., Wan J., Mantena N., Malek M.H., Podgorski I., Heath E.I.. Tissue-specific regulation of cytochrome c by post-translational modifications: Respiration, the mitochondrial membrane potential, ROS, and apoptosis. Faseb J. (2019) 33 1540-1553. DOI: 10.1096/fj.201801417R 41. Ow Y.P., Green D.R., Hao Z., Mak T.W.. Cytochrome c: Functions beyond respiration. Nat. Rev. Mol. Cell Biol. (2008) 9 532-542. DOI: 10.1038/nrm2434 42. Hiller S., Garces R.G., Malia T.J., Orekhov V.Y., Colombini M., Wagner G.. Solution structure of the integral human membrane protein VDAC-1 in detergent micelles. Science (2008) 321 1206-1210. DOI: 10.1126/science.1161302 43. Lee M.J., Kim J.Y., Suk K., Park J.H.. Identification of the hypoxia-inducible factor 1 alpha-responsive HGTD-P gene as a mediator in the mitochondrial apoptotic pathway. Mol. Cell Biol. (2004) 24 3918-3927. DOI: 10.1128/MCB.24.9.3918-3927.2004 44. Yu T., Fox R.J., Burwell L.S., Yoon Y.. Regulation of mitochondrial fission and apoptosis by the mitochondrial outer membrane protein hFis1. J. Cell Sci. (2005) 118 4141-4151. DOI: 10.1242/jcs.02537 45. Anand R., Wai T., Baker M.J., Kladt N., Schauss A.C., Rugarli E., Langer T.. The i-AAA protease YME1L and OMA1 cleave OPA1 to balance mitochondrial fusion and fission. J. Cell Biol. (2014) 204 919-929. DOI: 10.1083/jcb.201308006 46. Lemasters J.J., Holmuhamedov E.. Voltage-dependent anion channel (VDAC) as mitochondrial governator—Thinking outside the box. Biochim. Biophys. Acta (2006) 1762 181-190. DOI: 10.1016/j.bbadis.2005.10.006 47. Yu L., Sun Y., Li J., Wang Y., Zhu Y., Shi Y., Fan X., Zhou J., Bao Y., Xiao J.. Silencing the Girdin gene enhances radio-sensitivity of hepatocellular carcinoma via suppression of glycolytic metabolism. J. Exp. Clin. Cancer Res. (2017) 36 110. DOI: 10.1186/s13046-017-0580-7 48. Shao Y., Song X., Jiang W., Chen Y., Ning Z., Gu W., Jiang J.. MicroRNA-621 Acts as a Tumor Radiosensitizer by Directly Targeting SETDB1 in Hepatocellular Carcinoma. Mol. Ther. (2019) 27 355-364. DOI: 10.1016/j.ymthe.2018.11.005 49. Zhang Y., Zheng L., Ding Y., Li Q., Wang R., Liu T., Sun Q., Yang H., Peng S., Wang W.. MiR-20a Induces Cell Radioresistance by Activating the PTEN/PI3K/Akt Signaling Pathway in Hepatocellular Carcinoma. Int. J. Radiat. Oncol. Biol. Phys. (2015) 92 1132-1140. DOI: 10.1016/j.ijrobp.2015.04.007 50. Bamodu O.A., Chang H.-L., Ong J.-R., Lee W.-H., Yeh C.-T., Tsai J.-T.. Elevated PDK1 expression drives PI3K/AKT/MTOR signaling promotes radiation-resistant and dedifferentiated phenotype of hepatocellular carcinoma. Cells (2020) 9. DOI: 10.3390/cells9030746 51. Deng G., Mou T., He J., Chen D., Lv D., Liu H., Yu J., Wang S., Li G.. Circular RNA circRHOBTB3 acts as a sponge for miR-654-3p inhibiting gastric cancer growth. J. Exp. Clin. Cancer Res. (2020) 39 1. DOI: 10.1186/s13046-019-1487-2 52. Shen L., Sun X., Fu Z., Yang G., Li J., Yao L.. The fundamental role of the p53 pathway in tumor metabolism and its implication in tumor therapy. Clin. Cancer Res. (2012) 18 1561-1567. DOI: 10.1158/1078-0432.CCR-11-3040 53. Wang M., Chen X., Tan P., Wang Y., Pan X., Lin T., Jiang Y., Wang B., Xu H., Wang Y.. Acquired semi-squamatization during chemotherapy suggests differentiation as a therapeutic strategy for bladder cancer. Cancer Cell (2022) 40 1044-1059.e8. DOI: 10.1016/j.ccell.2022.08.010 54. Di Y.Q., Han X.L., Kang X.L., Wang D., Chen C.H., Wang J.X., Zhao X.F.. Autophagy triggers CTSD (cathepsin D) maturation and localization inside cells to promote apoptosis. Autophagy (2021) 17 1170-1192. DOI: 10.1080/15548627.2020.1752497 55. Chwieralski C.E., Welte T., Bühling F.. Cathepsin-regulated apoptosis. Apoptosis (2006) 11 143-149. DOI: 10.1007/s10495-006-3486-y 56. Faubion W.A., Gores G.J.. Death receptors in liver biology and pathobiology. Hepatology (1999) 29 1-4. DOI: 10.1002/hep.510290101 57. Fløyel T., Brorsson C., Nielsen L.B., Miani M., Bang-Berthelsen C.H., Friedrichsen M., Overgaard A.J., Berchtold L.A., Wiberg A., Poulsen P.. CTSH regulates β-cell function and disease progression in newly diagnosed type 1 diabetes patients. Proc. Natl. Acad. Sci. USA (2014) 111 10305-10310. DOI: 10.1073/pnas.1402571111 58. Fløyel T., Mirza A.H., Kaur S., Frørup C., Yarani R., Størling J., Pociot F.. The Rac2 GTPase contributes to cathepsin H-mediated protection against cytokine-induced apoptosis in insulin-secreting cells. Mol. Cell Endocrinol. (2020) 518 110993. DOI: 10.1016/j.mce.2020.110993 59. Pospisilik J.A., Knauf C., Joza N., Benit P., Orthofer M., Cani P.D., Ebersberger I., Nakashima T., Sarao R., Neely G.. Targeted deletion of AIF decreases mitochondrial oxidative phosphorylation and protects from obesity and diabetes. Cell (2007) 131 476-491. DOI: 10.1016/j.cell.2007.08.047 60. Kesavardhana S., Malireddi R.K.S., Kanneganti T.D.. Caspases in Cell Death, Inflammation, and Pyroptosis. Annu. Rev. Immunol. (2020) 38 567-595. DOI: 10.1146/annurev-immunol-073119-095439 61. Katz J.N., Arant K.R., Loeser R.F.. Diagnosis and Treatment of Hip and Knee Osteoarthritis: A Review. JAMA (2021) 325 568-578. DOI: 10.1001/jama.2020.22171 62. Han D., Liu J., Chen C., Dong L., Liu Y., Chang R., Huang X., Liu Y., Wang J., Dougherty U.. Anti-tumour immunity controlled through mRNA m(6)A methylation and YTHDF1 in dendritic cells. Nature (2019) 566 270-274. DOI: 10.1038/s41586-019-0916-x 63. Sharma C., Kang S.C.. Molecular dynamic simulation (MDS) and in vitro cathepsin-B inhibitory activity of decrusin angelate, ibuprofen, and thymol. Nat. Prod. Res. (2022) 36 1020-1025. DOI: 10.1080/14786419.2020.1843030 64. Liu T., Luo S., Libby P., Shi G.P.. Cathepsin L-selective inhibitors: A potentially promising treatment for COVID-19 patients. Pharmacol. Ther. (2020) 213 107587. DOI: 10.1016/j.pharmthera.2020.107587 65. Zhang Y., Gao Y., Gao Y.. Early changes in the urine proteome in a rat liver tumour model. Peer J. (2020) 8 e8462. DOI: 10.7717/peerj.8462 66. Camargo C.A., da Silva M.E., da Silva R.A., Justo G.Z., Gomes-Marcondes M.C., Aoyama H.. Inhibition of tumor growth by quercetin with increase of survival and prevention of cachexia in Walker 256 tumor-bearing rats. Biochem. Biophys. Res. Commun. (2011) 406 638-642. DOI: 10.1016/j.bbrc.2011.02.111 67. Zhang L., Liu L., Zhang D., Jin Q., Gao M., Wu T., Feng Y., Ni Y., Yin Z., Zhang J.. Synthesis and Evaluation of Diindole-Based MRI Contrast Agent for In Vivo Visualization of Necrosis. Mol. Imaging Biol. (2020) 22 593-601. DOI: 10.1007/s11307-019-01399-2 68. Camargo C.A., Gomes-Marcondes M.C., Wutzki N.C., Aoyama H.. Naringin inhibits tumor growth and reduces interleukin-6 and tumor necrosis factor α levels in rats with Walker 256 carcinosarcoma. Anticancer Res. (2012) 32 129-133. PMID: 22213297 69. Sun D.S., Chen J.H., Ling R., Yao Q., Wang L., Ma Z., Li Y.. Treatment of hepatoma with liposome-encapsulated adriamycin administered into hepatic artery of rats. World J. Gastroenterol. (2006) 12 4741-4744. DOI: 10.3748/wjg.v12.i29.4741
--- title: Prevalence of Anemia and Iron Deficiency in Women of Reproductive Age in Cuba and Associated Factors authors: - Gisela María Pita-Rodríguez - Beatriz Basabe-Tuero - María Elena Díaz-Sánchez - Karen Alfonso-Sagué - Ana María Gómez Álvarez - Minerva Montero-Díaz - Sonia Valdés-Perdomo - Cristina Chávez-Chong - Ernesto Rodríguez-Martinez - Yoandry Díaz-Fuentes - Elisa Llera-Abreu - Ahindris Calzadilla-Cámbara - Israel Ríos-Castillo journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049065 doi: 10.3390/ijerph20065110 license: CC BY 4.0 --- # Prevalence of Anemia and Iron Deficiency in Women of Reproductive Age in Cuba and Associated Factors ## Abstract This study aims to evaluate the prevalence of anemia and iron deficiency in women of reproductive age and the association with inflammation, global overweight, adiposity, and menorrhagia. A sample design of women of reproductive age from the Eastern, Central, and Havana Regions was carried out. Biochemical determinations of hemoglobin, serum ferritin, soluble transferrin receptors, leukocytes, C-reactive protein, alpha-1 acid glycoprotein, and homocysteine were performed. Serum ferritin was also adjusted by inflammation. Nutritional status was assessed, and menstrual characteristics were collected by survey. A total of 742 women were studied. The prevalence of anemia was $21.4\%$, iron storage deficiency at $16.0\%$, and erythropoietic dysfunction at $5.4\%$, with inflammation at $47.0\%$ and elevated homocysteine at $18.6\%$. Global overweight was $46.2\%$ and increased adiposity at $58.4\%$. Anemia is associated with iron deposition deficiency (OR = 3.023 (1.816–5.033)) and with erythropoietic deficiency (OR = 5.62 (3.03–10.39)), but not with inflammation, global overweight, and adiposity. Global overweight was found to be associated with inflammation (OR = 2.23 (1.41–3.53)). Anemia was associated with heavy menstrual bleeding (OR = 1.92 (1.34–2.76)). Homocysteine was associated with inflammation (OR = 2.05 (1.08–3.90)), but not with anemia. In conclusion, anemia in Cuba is classified as a moderate public health problem, but not iron deficiency. A high prevalence of overweight and obesity was found, associated with inflammation, but not with anemia or iron deficiency. Heavy menstrual bleeding is a factor associated with anemia. ## 1. Introduction Anemia in women of reproductive age (WRA) is a worldwide public health problem that remains a nutritional challenge [1], and there is a decrease of only four percentage points from 1995 ($33\%$) to 2011 ($29\%$) [2]. Factors associated with anemia have been reported as regular blood loss (due to menstrual bleeding), pregnancy-related complications, depletion of iron stores, and therefore, increased requirements [3]. The double burden of malnutrition consists of malnutrition due to micronutrient deficiency and obesity, which can be found in the individual as well as in the family and in populations [4]. Among the nutrient deficiencies, a relationship between iron deficiency and obesity has been observed [5]. In medical practice, inflammation is a common cause of anemia, but the magnitude of its attributable causation in developing countries is unknown [6]. Inflammation is a sign of the organism responding to infections with seasonal bacteria or viruses [7]. Thus, inflammation is an important trigger for chronic diseases and also for obesity [8]. During the inflammatory process, interleukins are secreted that stimulate the secretion of hepcidin, a known hormonal peptide released from the liver. Hepcidin acts on intestinal cells, preventing the absorption of iron from food, and on the phagocytic mononuclear system, thus sequestering circulating iron in the blood [9,10]. In these cases, iron deficiency is due to both low absorption caused by poor diet and to low use by the body. On the other hand, obesity is characterized by systemic, chronic, and low-grade inflammation, with an increase in C-reactive protein (CRP), Alpha 1-acid glycoprotein (AGP), and interleukin 6 (IL-6) [9,10,11]. Thus, malnutrition is related to iron deficiency and anemia through hepcidin secretion stimulated by high levels of IL-6 during inflammation [10]. In developing countries, the double burden of malnutrition and infections are increasing [4,12]. Therefore, the evaluation of the potential relationship of these factors with serum ferritin, as a protein for the evaluation of iron nutritional status, is important to explain the high prevalence of iron deficiency and its relationship with anemia [8]. Based on the knowledge of the metabolic pathway of homocysteine (Hcy) [13] and because of the difficulties in assessing serum or erythrocyte folate status, the use of Hcy metabolite as an indirect indicator of the nutritional status of these vitamins in the individual has been proposed [14]. Menstruation is a physiological process that can produce heavy menstrual bleeding (HMS), which interferes with a woman’s physical, social, emotional, and/or material quality of life, and which may or may not be accompanied by other symptoms [15]. Heavy menses is a common condition, affecting one in four WRA [16]. The group of WRA represents a very important link in the life cycle, but it has not been a population deeply studied in Cuba; their nutritional status is of great importance for a healthy pregnancy, a child with adequate birth weight, and adequate iron status. Therefore, it is necessary to know the factors that may be influencing anemia and iron deficiency in this group to propose actions. This study aimed to evaluate the prevalence of anemia and iron deficiency in women of reproductive age and its association with inflammation, global overweight, and volume of menses. ## 2. Materials and Methods The subject universe consisted of mothers of children aged 6–59 months studied by the national survey of anemia and iron deficiency in Cuban preschool children [17], conducted by the National Institute of Hygiene, Epidemiology, and Microbiology (INHEM acronym in Spanish) between February and April of each year during 2016–2018. Inclusion criteria were healthy WRA, aged 18–40 years, mothers of children from the Eastern Region (Santiago de Cuba and Holguin), Central Region (Sancti Spiritus and Cienfuegos), and Havana. Exclusion criteria were pregnant women, postpartum or with a delivery time of fewer than six months, and women with sickle cell disease or seen in consultations for hematological disorders. In addition, we excluded those women with acute illnesses, or those suffering from chronic diseases such as cancer, diabetes mellitus, arterial hypertension, hypothyroidism, hyperthyroidism, severe asthma, chronic obstructive pulmonary disease, or renal insufficiency. ## 2.1. Study Variables Table 1 shows the biochemical and anthropometric indicators with the cut-off points used. Anemia was defined as hemoglobin < 120 g/L, and iron deficiency was defined as ferritin concentration < 15 μg/L, the cut-off points recommended by WHO [9,10]. The nutritional status was assessed to evaluate overweight and obesity. Weight, height, and waist circumference were measured. In the analysis of the anthropometric results, the cases belonging to the “overweight” and “obese” groups were classified as “global overweight”. As risk factors for maternal anemia, the quality and quantity of menstruation were recorded with a questionnaire. In the menstrual survey, questions 5, 6, 8 and 9 of the “Survey of Symptoms during Your Monthly Period” [24] were adapted, and the practical clinical guidelines of the Society of Obstetricians and Gynecologists of Canada [23] were also used. The questions include description of the menstrual period (question 5); duration of the menstrual period (question 6); if the participant has seen a doctor because of vaginal bleeding (question 8); and if the participant has had to attend the emergency room for vaginal bleeding (question 9). The survey on menstrual bleeding or menorrhagia in women of childbearing age has been applied by other authors [25]. The survey includes questions about the amount of bleeding; bleeding frequency; duration of bleeding; regularity of the menstrual cycle; non-menstrual bleeding; and need for medical attention for bleeding. ## 2.2. Biochemical Data Blood was extracted by antecubital puncture after fasting for at least eight hours, before the questionnaire was administered, and after signing the consent form. Five mL of whole blood, 1 mL with $10\%$ EDTA for Hemoglobin (Hb) and leukocytes determination, and 4 mL for serum collection were extracted. On the day of the extraction, the samples were centrifuged, and the serum was stored at −40 °C until the time of analysis for ferritin, sTfR, Hcy, and inflammation indicators Hb and leukocytes were determined with an ABX Micros 60 Hematology Counter (Horiba, France). Iron deficiency was measured by ferritin and soluble transferrin receptor (sTfR) concentration, and inflammation by serum high-sensitivity C-reactive protein (CRP-hs) and α-1 acid glycoprotein (AGP). Ferritin and inflammation indicators were determined by the immunoturbidimetric method (CPM Scientifica Tecnologie Biomediche, Rome, Italy) using INLAB 240. Ferritin concentrations were adjusted using the quantile regression method [17,26,27]. The sTfR was performed by the ELISA method, Ramco Laboratories INC [28]. The sTfR was estimated in a subsample of Cienfuegos and Havana, and Hcy in a subsample of the eastern provinces and Havana. Hcy was determined by the enzymatic method (CPM Scientifica Tecnologie Biomediche, Rome, Italy) with INLAB 240. Homocysteine was used as an indirect indicator of folate deficiency. All cutoff points used are found in Table 1. The determinations were carried out at the INHEM’s Nutritional Anemia Laboratory. ## 2.3. Statistical Analysis Biochemical variables were described according to their distribution by percentiles (median, 25th percentile, and 75th percentile). For categorical variables, prevalence and $95\%$ confidence intervals ($95\%$ CI) were calculated. Prevalence calculations were performed taking into account the sample design. Serum ferritin was adjusted by inflammation using quantile regression, which is a natural extension of the standard regression model and allows separate regression models to be used for different parts of the dependent variable’s distribution. Quantile regression’s additional flexibility may broaden the description of inflammation’s effect on ferritin’s conditional distribution. An additional advantage to quantile regression is that it does not depend on normality assumptions or transformations. The quantile regression methodology has already been used in the correction of ferritin levels in preschool children in Cuba [17,26]. In the association analysis, the variables were categorized into two levels, and the odds ratio (OR) and $95\%$ CI were used. In this case, for the estimations, the sample design was also taken into account. For the analysis of the menstrual volumes of WRA, they were grouped into two groups, according to pad condition: Soaking through + Strongly wet/Moderate + Slightly wet. SPSS v20.0 statistical software was used for database preparation. The SAS 9.1 statistical package was used to take into account the complex sampling design of the sample for the estimation of the statistics and their standard errors as well as the statistical tests used. The SPSS program was used to evaluate the associations. A significance level of $p \leq 0.05$ was considered in all cases. ## 2.4. Ethical Aspects The project was approved by the ethics committee of the INHEM and authorized by the Maternal and Child Division of the Ministry of Public Health of Cuba. Written informed consent was obtained from the mothers of the children participating in the study. At each stage, the agreements of the World Medical Association in the Declaration of Helsinki on ethical principles for medical research in humans were taken into account, and compliance with the basic principles of all research with human beings was monitored [29]. ## 3. Results The final sample of WRA was 742 women, of which 729 were samples with usable Hb and leukocytes, and 711 with ferritin and inflammation. The cause of the missing data does not represent a risk of bias, since it responds to random causes such as coagulation of the whole blood sample or not enough serum for all determinations. A subsample of 249 was determined for sTfR and 381 for Hcy. ## 3.1. Anemia and Iron Nutritional Status Anemia was found in $21.4\%$ of the women studied (Table 2); the majority ($81.1\%$) was classified as mild, with $18.1\%$ moderate, and $0.8\%$ severe anemia. Iron storage deficiency was found in less than one-third of the cases and classified as a mild public health problem. The prevalence of erythropoietic dysfunction was low ($5.4\%$). Most ferritin values were in the reference range of 15–150 µmol/L $82.7\%$ (74.2–88.8), but seven cases were found with values above this range, indicative of risk of iron overload. Inflammation, both acute and chronic, was found in about one-third of all women and general inflammation in about half of the sample, but not leukocytosis, which was found in less than one-tenth. The prevalence of elevated Hcy was found in about one-fifth of the sample, resulting in a red flag in the study for this indicator (Table 2). ## 3.2. Nutritional Status In the nutritional evaluation of women, $42.8\%$ (39.5–46.1) had adequate nutritional status, $11\%$ (7.8–14.2) were undernourished, and $46.2\%$ were with global overweight (overweight $30.4\%$ (25.3–35.5), and obese $15.8\%$ (13.0–18.7). The degree of central adiposity was adequate in $58.4\%$ (52.9–63.9), with increased risk in $22.8\%$ (15.4–30.2), and very increased risk in $18.7\%$ (14.5–23.0). ## 3.3. Menstrual Analysis Of the WRA, $19.8\%$ (14.4–25.3) reported soaking through the sanitary pad, and $22.7\%$ (18.4–26.9) had very wet pads. The other group reported moderately wet, $40.2\%$ (31.2–49.2), and slightly wet, $17.2\%$ (9.3–25.1). Only one case reported that menstruation was variable in quantity. Most of the women—$87.7\%$ (85.5–89.9)—reported that the frequency of menses was normal (24–38 days), as was the duration (3–8 days) for $93.6\%$ (91.7–95.5). Acute bleeding was reported by $7.4\%$ (4.6–10.2), and $13.2\%$ (9.5–16.9) went to the doctor for this cause. ## 3.4. Association Analysis Anemia in women of reproductive age was associated with iron storage deficiency (OR = 3.02 (1.82–5.03)) and with erythropoietic deficiency (OR = 5.62 (3.03–10.39)), but not with inflammation (OR = 1.00 (0.65–1.54)), global overweight (OR = 0.80(0.57–1.12)), and central adiposity (OR = 0.80 (0.57–1.12)). Iron storage deficiency explained $29.8\%$ of the anemia found. Iron storage deficiency was not associated with global overweight (OR = 0.71 (0.48–1.06)), but it was associated with central adiposity (OR = 0.59 (0.38–0.91)) as a protection factor. Global overweight in women was found to be associated with inflammation (OR = 2.23 (1.41–3.53)), mainly with elevated CRP (OR = 3.06 (1.89–4.94)) rather than with AGP (OR = 1.80 (1.05–3.08)). Adiposity and inflammation behaved in a similar way, with (OR = 3.23 (2.32–4.51)), with higher values for CRP (OR = 3.76 (2.29–6.16)) than those for AGP (OR = 1.94 (1.19–3.15)). Anemia in women was explained by the higher volume of menstruation (Soaking through + Strongly wet/Moderate and Slightly) (OR = 1.92 (1.34–2.76)) and heavy menstrual bleeding, which explained $55.4\%$ of anemia in women, but did not explain iron storage deficiency as well (OR = 1.35 (0.80–2.27)). However, the volume of menstruation is not accurate, and it is subject to measurement failures. The analysis showed that Hcy was also not associated with anemia (OR = 0.94 (0.44–2.00)), but it was associated with inflammation (OR = 2.05 (1.08–3.90)), by elevated CRP (OR = 1.17 (0.85–1.60)) and mostly by elevated AGP (OR = 1.65 (1.07–2.55)). However, the association with global overweight and adiposity was not relevant (OR = 1.02 (0.73–1.42)) and (OR = 1.32 (0.74–2.33)), respectively. ## 4. Discussion Anemia in the studied WRA had a similar prevalence to that found in 2008 ($$n = 1802$$) in women from the eastern provinces ($19.9\%$) [30]. In our study, WRA had adequate ferritin values, and less than $20\%$ showed iron storage deficiencies that did not affect erythropoiesis. These data are striking because in the study performed on 391 WRA in Havana in 2017, the iron deficiency, adjusted for inflammation, was found to be $68\%$. Anemia was associated with iron deficiency and erythropoietic dysfunction with greater strength than that obtained in the Havana study [31], and no positive association was found with inflammation. Leukocytosis was low, unlike the results of the WRA group in Havana; however, indicators of subclinical inflammation in this study are higher than those found in 2014 (CRP $8.4\%$ and AGP $19.9\%$) [31]. Stevens et al. [ 32] performed a national, regional, and global estimation of the severity of anemia in 133 countries, in WRA aged 15–49 years. They reported an anemia prevalence of $30\%$ (27–$33\%$), with a decrease in severe and moderate anemia, but considered that this progress is insufficient to reach the proposed global goals for 2025 (to reduce anemia by $50\%$ according to each country’s baseline) [33]. The decrease in anemia by decade was mostly in the Latin American and Caribbean region. The reported prevalence was higher than those found in our study. Kinyoki et al. [ 34] conducted a geospatial study of anemia prevalence estimates in women of reproductive age (15–49 years) from 2000 to 2018 in 82 low- and middle-income countries. The results showed moderate improvements in the overall prevalence of anemia in the various countries, but only three countries (China, Iran and Thailand) could be identified as having a high likelihood of achieving the proposed targets at the national level. Wirth et al. [ 35] conducted a study in Somalia between 2018 and 2019 in 583 women of reproductive age, where they explored, in addition to anemia and iron deficiency, vitamin A deficiency, finding that the main risk factor for anemia was iron deficiency. No studies have been found for Cuba where folate concentrations in women of reproductive age have been estimated. There are only dietary studies indicating folate intake deficiency in preschool children and pregnant women, which were obtained by the nutritional surveillance system established in the country [36]. Rogers et al. [ 14] conducted an analysis of 45 relevant studies in 39 countries to estimate folate deficiencies in WRA between 2000 and 2014. The deficiency found was greater than $20\%$ in most low-income countries, in contrast to high-income countries, where it was less than $5\%$. In early stages of folate deficiency, homocysteine concentrations are elevated and appear even at elevated serum or plasma folate levels (10 or 14 nmol/L) [37]. Evaluation of homocysteine in WRA in this study was not useful for estimating folic acid deficiency [13]. However, it evidenced its association with inflammation; therefore, it is a factor to keep in mind in this population, and the elevated values in the subsample analyzed are not negligible. Rosabal Nieves [13] recognized several hereditary, pathological, nutritional, and pharmacological events capable of inducing hyper-homocysteinemia. Sex and age are the most important physiological causes of elevated plasma homocysteine. Serum tHcy levels increase with age, due in part to physiological decline in renal function. However, we did not evaluate the relationship between age and homocysteine. Therefore, it is necessary to continue investigating the relationship of anemia and iron deficiency with other biochemical markers, such as folate. Although the level of zinc was not evaluated in this study, the literature suggests that zinc can be related with anemia and iron deficiency. The recent study of Zn status conducted in a sample of 654 women of reproductive age evidenced that Zn deficiency was high, at $35.7\%$, and $25.0\%$ of Zn-deficient women had anemia. These results suggest an association between anemia and Zn deficiency not previously studied in Cuba [38]. Greffeuille et al. [ 39] evaluated Zn and Hb concentrations in 22,633 WRA from nationally representative studies from different countries and found a prevalence of Zn deficiency between 9.8 and $84.7\%$. The prevalence of Zn deficiency was greater than $20\%$ in most countries. Zn concentration had a positive association with Hb in about half of the countries, independent of iron status, and was significantly related to anemia in most countries, thus concluding that strategies to combat Zn deficiency may help reduce the prevalence of anemia. In humans, no direct relationship between Zn concentration and red cell production has been evidenced, but Zn is directly involved in erythropoietic differentiation and development [40]. Therefore, more studies are required to evaluate the relationship of anemia and iron deficiency with micronutrient deficiency, obesity, and NCDs. In the WRA of the present study, malnutrition had high figures, with a predominance of global overweight and a high percentage of adiposity, but neither of these was associated with anemia or iron deficiency. This result coincides with that of Havana, where excess body weight was not positively associated with anemia or iron deficiency, but central adiposity was a protective factor for both cases [31]. In a regional study in Latin America in 3254 women aged 15 to <45 years, Herrera-Cuenca et al. [ 41] found that $58.7\%$ were overweight, associated with adequate iron intake. Christian et al. [ 42] conducted a national study in Ghana in 1063 women aged 15–49 years and found that the prevalence of overall overweight was $39\%$, anemia $22\%$, and deficiency of a micronutrient $62\%$. The joint occurrence of overall overweight and anemia was $6.7\%$, and with deficiency of at least one micronutrient $23.6\%$. Kamruzzaman [43] conducted a study in Bangladesh, using nationally representative data on 5680 women aged 15–49 years, where he found that the probability of anemia was higher in the undernourished than in the overweight and obese. Although it is known that nutritional deficiencies are associated with anemia, the relationship between obesity, inflammation, and anemia has also been described in the literature [6,7]. Therefore, its study is important. Despite this, in this study, we did not observe a relationship between obesity and anemia or iron deficiency. These results are in agreement with those found in the current study, where anemia and iron deficiency were not associated with global overweight or adiposity, which, on the contrary, proved to be protective factors. The subjective assessment of women’s menstrual volume proved to be a rough indicator of how much anemia may be caused by menstrual bleeding, as more than half of the anemia can be explained by this cause. In a pilot study conducted in 44 adult African-American women by Bernardi et al. [ 44] in 2016, almost half reported heavy and very heavy menses; they found anemia in $18.2\%$ and iron deficiency in $69.2\%$, with significant association between anemia and perceived menstrual volume. Among those who reported high menstrual volume, $35\%$ had anemia. These figures are lower than those found in this work. Kiran et al. in 2018 [16] report that in England and Wales, an estimated 50,000 women with very heavy menses are transferred annually from primary care to secondary care services of gynecology in the National Health Service, and menstrual disorders occur in $20\%$ of women, affecting their quality of life. Ding et al. [ 45] in 2019 studied the prevalence and risk factors of 2356 WRA (18–50 years) in a Chinese community who experienced heavy menses and assessed its effect on daily life. They observed that $18.2\%$ of them reported heavy menses, and this was significantly associated with iron deficiency anemia (OR = 1.56 (1.06–2.29)). In the current study, heavy menstruation was associated with anemia, but not with iron deficiency. In contrast to our findings, using SF values adjusted for inflammation, Zazo et al. [ 46] studied the causes of iron deficiency anemia prevalence in 161,315 women of reproductive age 18–35 years, seen in Madrid from 2010–2015. They highlighted that $61.8\%$ of iron deficiency anemia, which had no known associated cause, was due solely to menstrual losses [46]. Anemia is a multifactorial disease, and there may be other associated factors that were not evaluated in this study. The limitations of the availability of iron for the erythroid progenitors during an infectious process are one of the central causes of anemia due to inflammation [7]. Mild iron deficiency is associated with protection from certain infections, thus preventing morbidity and mortality from infectious diseases, particularly in children. On the other hand, severe iron deficiency has a negative impact on the proliferation of immune cells and therefore weakens the immune response [47,48]. In places with high rates of infections, other factors for anemia are bleeding episodes of gastrointestinal origin, parasitism or urogenital losses, and menstruation disorder. These subjects develop absolute or relative iron deficiency. As a consequence, iron is not available for essential metabolic processes, including cellular respiration. Anemia in WRA is an important public health problem worldwide and in Cuba, mainly due to the increased risk of anemia during pregnancy and the negative effects on infants [33,36,38]. However, iron deficiency anemia is the last stage of nutritional iron deficiency [49]. Therefore, evaluating iron deficiency is crucial to avoid the consequences of anemia early on. Anemia prevention policies and programs require the generation of timely scientific evidence on the iron nutrition situation in the population. The present study on anemia and iron deficiency in women of childbearing age also allows us to identify the potential relationships that occur with malnutrition due to excess weight and menstrual disorders. Some limitations in our study are the design (cross-sectional), so that it was impossible to examine the temporality of the indicators with other factors; second, that it was not possible to study the WRA of the western provinces, which implies that the results may be influenced by the lack of these data. The third limitation is that it was not possible to study other important nutritional causes of anemia in the vulnerable population, such as folic acid deficiency. ## 5. Conclusions Anemia in WRA in Cuba is classified as a moderate public health problem, but iron deficiency is a mild public health problem. Women of reproductive age have a high prevalence of overweight and obesity that is not associated with anemia or iron deficiency but is associated with inflammation. Heavy menstrual bleeding is a factor associated with anemia. ## References 1. McLean E., Cogswell M., Egli I., Wojdyla D., De Benoist B.. Worldwide Prevalence of Anaemia, WHO Vitamin and Mineral Nutrition Information System, 1993–2005. Public Health Nutr. (2009.0) 12 444-454. DOI: 10.1017/S1368980008002401 2. Stevens G.A., Finucane M.M., De-Regil L.M., Paciorek C.J., Flaxman S.R., Branca F., Peña-Rosas J.P., Bhutta Z.A., Ezzati M.. Global, Regional, and National Trends in Haemoglobin Concentration and Prevalence of Total and Severe Anaemia in Children and Pregnant and Non-Pregnant Women for 1995–2011: A Systematic Analysis of Population-Representative Data. Lancet Glob. Health (2013.0) 1 e16-e25. DOI: 10.1016/S2214-109X(13)70001-9 3. Safiri S., Kolahi A.A., Noori M., Nejadghaderi S.A., Karamzad N., Bragazzi N.L., Sullman M.J.M., Abdollahi M., Collins G.S., Kaufman J.S.. Burden of Anemia and Its Underlying Causes in 204 Countries and Territories, 1990–2019: Results from the Global Burden of Disease Study 2019. J. Hematol. Oncol. (2021.0) 14 185. DOI: 10.1186/s13045-021-01202-2 4. Popkin B.M., Corvalan C., Grummer-Strawn L.M.. Dynamics of the Double Burden of Malnutrition and the Changing Nutrition Reality. Lancet (2020.0) 395 65-74. DOI: 10.1016/S0140-6736(19)32497-3 5. Cepeda-Lopez A.C., Baye K.. Obesity, Iron Deficiency and Anaemia: A Complex Relationship. Public Health Nutr. (2020.0) 23 1703-1704. DOI: 10.1017/S1368980019004981 6. Raiten D.J., Sakr Ashour F.A., Ross A.C., Meydani S.N., Dawson H.D., Stephensen C.B., Brabin B.J., Suchdev P.S., Van Ommen B., Ahluwalia N.. Inflammation and Nutritional Science for Programs/Policies and Interpretation of Research Evidence (INSPIRE). J. Nutr. (2015.0) 145 1039S-1108S. DOI: 10.3945/jn.114.194571 7. Nairz M., Haschka D., Demetz E., Weiss G.. Iron at the Interface of Immunity and Infection. Front. Pharmacol. (2014.0) 5 152. DOI: 10.3389/fphar.2014.00152 8. Abbaspour N., Hurrell R., Kelishadi R.. Review on Iron and Its Importance for Human Health. J. Res. Med. Sci. (2014.0) 19 164-174. PMID: 24778671 9. Wendt A.S., Jefferds M.E., Perrine C.G., Halleslevens P., Sullivan K.M.. Obese Women Less Likely to Have Low Serum Ferritin, Nicaragua. Public Health Nutr. (2015.0) 18 736-741. DOI: 10.1017/S1368980014000755 10. Becker C., Orozco M., Solomons N.W., Schümann K.. Iron Metabolism in Obesity: How Interaction between Homoeostatic Mechanisms Can Interfere with Their Original Purpose. Part I: Underlying Homoeostatic Mechanisms of Energy Storage and Iron Metabolisms and Their Interaction. J. Trace Elem. Med. Biol. (2015.0) 30 195-201. DOI: 10.1016/j.jtemb.2014.10.011 11. Mohamed-Ali V., Goodrick S., Rawesh A., Katz D.R., Miles J.M., Yudkin J.S., Klein S., Coppack S.W.. Subcutaneous Adipose Tissue Releases Interleukin-6, but Not Tumor Necrosis Factor-α, in Vivo. J. Clin. Endocrinol. Metab. (1997.0) 82 4196-4200. DOI: 10.1210/jc.82.12.4196 12. Dunachie S., Chamnan P.. The Double Burden of Diabetes and Global Infection in Low and Middle-Income Countries. Trans. R. Soc. Trop. Med. Hyg. (2019.0) 113 56-64. DOI: 10.1093/trstmh/try124 13. Rosabal Nieves E.L.. Sobre El Metabolismo de La Homocisteína. Implicaciones Para La Nutrición. Rev. Cuba. Aliment. Nutr. (2016.0) 26 157-171 14. Rogers L.M., Cordero A.M., Pfeiffer C.M., Hausman D.B., Tsang B.L., De-Regil L.M., Rosenthal J., Razzaghi H., Wong E.C., Weakland A.P.. Global Folate Status in Women of Reproductive Age: A Systematic Review with Emphasis on Methodological Issues. Ann. N. Y. Acad. Sci. (2018.0) 1431 35-57. DOI: 10.1111/nyas.13963 15. 15. National Institute for Clinical Excellence (NICE) Heavy Menstrual Bleeding: Assessment and Management—Clinical GuidelineNICELondon, UK2007. Heavy Menstrual Bleeding: Assessment and Management—Clinical Guideline (2007.0) 16. Kiran A., Geary R.S., Gurol-Urganci I., Cromwell D.A., Bansi-Matharu L., Shakespeare J., Mahmood T., Van Der Meulen J.. Sociodemographic Differences in Symptom Severity and Duration among Women Referred to Secondary Care for Menorrhagia in England and Wales: A Cohort Study from the National Heavy Menstrual Bleeding Audit. BMJ Open (2018.0) 8 e018444. DOI: 10.1136/bmjopen-2017-018444 17. Pita-Rodriguez G.M., Basabe-Tuero B., Diaz-Sanchez M.E., Alfonso-Sague K., Gomez-Alvarez A.M., Montero-Diaz M., Chavez-Chong C., Rodriguez-Martinez E., Diaz-Fuentes Y., Llera-Abreu E.. Anemia Is Related to Iron Deficiency and Inflammation, But No to Global Overweight in Cuban Preschool Children. Biomed. Sci. (2023.0) 9 1-8 18. Haemoglobin Concentrations for the Diagnosis of Anaemia and Assessment of Severity. Vitamin and Mineral Nutrition Information System (2011.0) 19. 19. WHO WHO Guideline on Use of Ferritin Concentrations to Assess Iron Status in Individuals and PopulationsWorld Health OrganizationGeneva, Switzerland2020. WHO Guideline on Use of Ferritin Concentrations to Assess Iron Status in Individuals and Populations (2020.0) 20. Díaz Sánchez M.E., Hernández Triana M., Porrata Maury C., Jiménez Acostay S., Rodríguez Suárez A., Valdés Fraga L., Ruiz Álvarez V., Macías Matos C., Pita Rodríguez G., Lanyau Domínguez Y., Carrillo Farnés O.. Indicadores Antropométricos de La Evaluación Nutricional. Alimentación, Nutrición y Salud (2011.0) 259 21. Shetty P.S., James W.P.. Body Mass Index. A Measure of Chronic Energy Deficiency in Adults. FAO Food Nutr. Pap. (1994.0) 56 1-57. PMID: 7925867 22. Han T.S., Van Leer E.M., Seidell J.C., Lean M.E.J.. Waist Circumference Action Levels in the Identification of Cardiovascular Risk Factors: Prevalence Study in a Random Sample. BMJ (1995.0) 311 1401. DOI: 10.1136/bmj.311.7017.1401 23. Singh S., Best C., Dunn S., Leyland N., Wolfman W.L., Wolfman W., Allaire C., Awadalla A., Heywood M., Lemyre M.. Abnormal Uterine Bleeding in Premenopausal Women. J. Obstet. Gynaecol. Can. (2013.0) 35 476-479. DOI: 10.1016/S1701-2163(15)30940-3 24. Your Period 25. Lumsden M.A., Gebbie A., Holland C.. Managing Unscheduled Bleeding in Non-Pregnant Premenopausal Women. BMJ (2013.0) 346 f3251. DOI: 10.1136/bmj.f3251 26. Montero-Díaz M., Chávez-Chong C., Rodríguez-Martínez E., Pita-Rodríguez G.M., Lambert-Lamazares B., Basabe-Tuero B., Alfonso-Sagué K.. Adjusting Iron Deficiency for Inflammation in Cuban Children Aged Under Five Years: New Approaches Using Quadratic and Quantile Regression. Medicc Rev. (2022.0) 24 36-45. DOI: 10.37757/MR2022.V24.N3-4.1 27. Rodríguez E., Pita G., Montero M., Chávez C., Basabe B.. Ajuste Por Inflamación Del Estado de Hierro de Las Madres de Niños Preescolares Cubanos Utilizando Modelos de Regresión Por Cuantiles. Rep. Investig. Inst. Cibernética Matemática Física (2022.0) 974 15 28. Flowers C.H., Skikne B.S., Covell A.M., Cook J.D.. The Clinical Measurement of Serum Transferrin Receptor. J. Lab. Clin. Med. (1989.0) 114 368-377. PMID: 2677198 29. World Medical Association Declaration of Helsinki. Ethical Principles for Medical Research Involving Human Subjects. J. Am. Med. Assoc. (2013.0) 310 2191-2194. DOI: 10.1001/jama.2013.281053 30. 30. Institute of Nutrition and Food Hygiene Informe Técnico Sobre Situación Nutricional de Poblaciones de RiesgoInstitute of Nutrition and Food HygieneHavana, Cuba2017. Informe Técnico Sobre Situación Nutricional de Poblaciones de Riesgo (2017.0) 31. Pita-Rodriguez G.M., Basabe-Tuero B., Diaz-Sanchez M.E., Gomez-Alvarez A.M., Campos-Hernandez D., Arocha-Oriol C., Diaz-Fuentes Y., Puentes-Marquez I., Herrera-Javier D.. Anemia and Iron Deficiency Related to Inflammation, Helicobacter Pylori Infection and Adiposity in Reproductive-Age Cuban Women. Medicc Rev. (2017.0) 19 10-17 32. Stevens G.A., Paciorek C.J., Flores-Urrutia M.C., Borghi E., Namaste S., Wirth J.P., Suchdev P.S., Ezzati M., Rohner F., Flaxman S.R.. National, Regional, and Global Estimates of Anaemia by Severity in Women and Children for 2000–19: A Pooled Analysis of Population-Representative Data. Lancet Glob. Health (2022.0) 10 e627-e639. DOI: 10.1016/S2214-109X(22)00084-5 33. 33. WHO Global Nutrition Targets 2025: Anaemia Policy BriefWorld Health OrganizationGeneva, Switzerland2014. Global Nutrition Targets 2025: Anaemia Policy Brief (2014.0) 34. Kinyoki D., Osgood-Zimmerman A.E., Bhattacharjee N.V., Schaeffer L.E., Lazzar-Atwood A., Lu D., Ewald S.B., Donkers K.M., Letourneau I.D., Collison M.. Anemia Prevalence in Women of Reproductive Age in Low- and Middle-Income Countries between 2000 and 2018. Nat. Med. (2021.0) 27 1761-1782. DOI: 10.1038/s41591-021-01498-0 35. Wirth J.P., Sesay F., Mbai J., Ali S.I., Donkor W.E.S., Woodruff B.A., Pilane Z., Mohamud K.M., Muse A., Yussuf H.O.. Risk Factors of Anaemia and Iron Deficiency in Somali Children and Women: Findings from the 2019 Somalia Micronutrient Survey. Matern. Child Nutr. (2022.0) 18 e13254. DOI: 10.1111/mcn.13254 36. Rodríguez Suárez A.. Análisis de Situación de País Para La Fortificación Del Arroz: Cuba (2016.0) 26 37. 37. WHO Serum and Red Blood Cell Folate Concentrations for Assessing Folate Status in PopulationsWorld Health OrganizationGeneva, Switzerland2015. Serum and Red Blood Cell Folate Concentrations for Assessing Folate Status in Populations (2015.0) 38. Gomez-Alvarez A., Pita-Rodriguez G., Garcia-Pino C., Mollineda-Trujillo A., Bacallao-Gallestey J., Cardella-Rosales L., Valdes-Collazo L.. Zinc Deficiency and Anemia in Women of Reproductive Age, Cuba. Biomed. Sci. (2022.0) 8 63-67 39. Greffeuille V., Fortin S., Gibson R., Rohner F., Williams A., Young M.F., Houghton L., Ou J., Dijkhuizen M.A., Wirth J.P.. Associations between Zinc and Hemoglobin Concentrations in Preschool Children and Women of Reproductive Age: An Analysis of Representative Survey Data from the Biomarkers Reflecting Inflammation and Nutritional Determinants of Anemia (BRINDA) Project. J. Nutr. (2021.0) 151 1277-1285. DOI: 10.1093/jn/nxaa444 40. Osawa M., Yamaguchi T., Nakamura Y., Kaneko S., Onodera M., Sawada K.I., Jegalian A., Wu H., Nakauchi H., Iwama A.. Erythroid Expansion Mediated by the Gfi-1B Zinc Finger Protein: Role in Normal Hematopoiesis. Blood (2002.0) 100 2769-2777. DOI: 10.1182/blood-2002-01-0182 41. Herrera-Cuenca M., Previdelli A.N., Koletzko B., Hernandez P., Landaeta-Jimenez M., Sifontes Y., Gómez G., Kovalskys I., García M.C.Y., Pareja R.. Childbearing Age Women Characteristics in Latin America. Building Evidence Bases for Early Prevention. Results from the Elans Study. Nutrients (2021.0) 13. DOI: 10.3390/nu13010045 42. Christian A.K., Steiner-Asiedu M., Bentil H.J., Rohner F., Wegmüller R., Petry N., Wirth J.P., Donkor W.E.S., Amoaful E.F., Adu-Afarwuah S.. Co-Occurrence of Overweight/Obesity, Anemia and Micronutrient Deficiencies among Non-Pregnant Women of Reproductive Age in Ghana: Results from a Nationally Representative Survey. Nutrients (2022.0) 14. DOI: 10.3390/nu14071427 43. Kamruzzaman M.. Is BMI Associated with Anemia and Hemoglobin Level of Women and Children in Bangladesh: A Study with Multiple Statistical Approaches. PLoS ONE (2021.0) 16. DOI: 10.1371/journal.pone.0259116 44. Bernardi L.A., Ghant M.S., Andrade C., Recht H., Marsh E.E.. The Association between Subjective Assessment of Menstrual Bleeding and Measures of Iron Deficiency Anemia in Premenopausal African-American Women: A Cross-Sectional Study. BMC Womens Health (2016.0) 16. DOI: 10.1186/s12905-016-0329-z 45. Ding C., Wang J., Cao Y., Pan Y., Lu X., Wang W., Zhuo L., Tian Q., Zhan S.. Heavy Menstrual Bleeding among Women Aged 18-50 Years Living in Beijing, China: Prevalence, Risk Factors, and Impact on Daily Life. BMC Womens Health (2019.0) 19. DOI: 10.1186/s12905-019-0726-1 46. Zazo P., Rubert M., Alberquilla A., de la Piedra C.. Anemia Ferropénica En La Premenopausia. Atención Primaria (2020.0) 52 511-513. DOI: 10.1016/j.aprim.2019.10.008 47. Vanarsa K., Ye Y., Han J., Xie C., Mohan C., Wu T.. Inflammation Associated Anemia and Ferritin as Disease Markers in SLE. Arthritis Res. Ther. (2012.0) 14 R182. DOI: 10.1186/ar4012 48. Ueda N., Takasawa K.. Impact of Inflammation on Ferritin, Hepcidin and the Management of Iron Deficiency Anemia in Chronic Kidney Disease. Nutrients (2018.0) 10. DOI: 10.3390/nu10091173 49. Cook J.D., Finch C.A., Smith N.J.. Evaluation of the Iron Status of a Population. Blood (1976.0) 48 449-455. DOI: 10.1182/blood.V48.3.449.449
--- title: Cross-Sectional Study of Location-Based Built Environments, Physical Activity, Dietary Intake, and Body Mass Index in Adult Twins authors: - Glen E. Duncan - Feiyang Sun - Ally R. Avery - Philip M. Hurvitz - Anne Vernez Moudon - Siny Tsang - Bethany D. Williams journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049069 doi: 10.3390/ijerph20064885 license: CC BY 4.0 --- # Cross-Sectional Study of Location-Based Built Environments, Physical Activity, Dietary Intake, and Body Mass Index in Adult Twins ## Abstract We examined relationships between walkability and health behaviors between and within identical twin pairs, considering both home (neighborhood) walkability and each twin’s measured activity space. Continuous activity and location data (via accelerometry and GPS) were obtained in 79 pairs over 2 weeks. Walkability was estimated using Walk Score® (WS); home WS refers to neighborhood walkability, and GPS WS refers to the mean of individual WSs matched to every GPS point collected by each participant. GPS WS was assessed within (WHN) and out of the neighborhood (OHN), using 1-mile Euclidean (air1mi) and network (net1mi) buffers. Outcomes included walking and moderate-to-vigorous physical activity (MVPA) bouts, dietary energy density (DED), and BMI. Home WS was associated with WHN GPS WS ($b = 0.71$, SE = 0.03, $p \leq 0.001$ for air1mi; $b = 0.79$, SE = 0.03, $p \leq 0.001$ for net1mi), and OHN GPS WS ($b = 0.18$, SE = 0.04, $p \leq 0.001$ for air1mi; $b = 0.22$, SE = 0.04, $p \leq 0.001$ for net1mi). Quasi-causal relationships (within-twin) were observed for home and GPS WS with walking (ps < 0.01), but not MVPA, DED, or BMI. Results support previous literature that neighborhood walkability has a positive influence on walking. ## 1. Introduction Diet and physical activity behaviors are influenced by multiple interacting factors ranging from biology to policy [1]. Aspects of built environments (BE) influence health behaviors, and, subsequently, obesity and associated chronic diseases [2,3,4]. Specifically, a dearth of health-promoting BEs, characterized by a lack of access to food retail outlets providing healthy foods, poor walkability, and fewer outdoor park and recreational facilities accessible in proximity to home, is inversely associated with healthful dietary behaviors [5,6] and physical activity levels [7,8] across all ages. Interventions that target these upstream community-level factors are proposed as cost-effective means to promote and improve health across populations [9]. For these reasons, research informing prevention efforts should shift focus from individual-level factors to macro-level factors such as the BE to guide policies that influence all people in their daily lives. The relationships between neighborhood BE and energy balance behaviors are nuanced and complex [10,11]. Self-selection bias, structural confounding, and measurement issues have hampered the neighborhood-effects literature [12]. Identical twins can be used as “quasi-experimental” controls of genetic and shared environmental factors (i.e., between-family confounds) that influence neighborhood selection, behaviors, and health [12,13,14,15]. As such, findings from twin analyses in genetically informed samples are considered “quasi-causal,” where the effect of the predictor on the outcome controls for any common genetic and familial environmental background they share and is, therefore, a partial approximation of any causal effect of the predictor [12,13,14,15]. Utilizing the twin study design to determine the relative influence of between-family factors, this study examined the between- and within-twin pair associations among the BE and health behaviors in identical twin pairs. For environmental exposures, the focus was on comparing the home neighborhood, which has been at the core of much of past research, and the total activity space experienced by individuals on a daily basis. We addressed the following research questions: [1] Do individuals who reside in more walkable neighborhoods spend more time in more walkable neighborhoods, including those not necessarily their own?; [ 2] Do individuals who reside in more walkable neighborhoods or spend more time in more walkable neighborhoods have lower dietary energy density and higher physical activity levels?; and [3] Do individuals who reside in more walkable neighborhoods or spend more time in more walkable neighborhoods have lower BMI? ## 2.1. Study Design and Participants This was a cross-sectional study of objective measures of physical activity amounts and locations among individuals living in the Puget Sound region around Seattle, WA, USA. Participants wore a Qstarz BT-Q1000XT GPS data logger (Qstarz International Co. Ltd., Taipei, Taiwan) and Actigraph GT3X+ accelerometer (Actigraph Inc., Pensacola, FL, USA) attached to an elastic belt worn around the waist to collect data under free-living conditions over a 2-week period. Validated surveys were used to collect information on demographic and health characteristics, as well as dietary intake. Data were collected during 2012–2015. To retain only high-quality data, the GPS data were censored to keep only those points that were recorded with Standard Positioning Service (SPS) or Differential GPS (DGPS), and that had a horizontal dilution of precision (HDOP) < 6. The initial sample included 187 identical twin participants from the community-based Washington State Twin Registry (WSTR). Recruitment procedures and details about the WSTR are reported elsewhere [16,17]. Recruited twins completed an in-person study visit followed by a 2-week remote data collection protocol. Inclusion criteria were living at the primary residence for at least 1 year and absence of physical conditions that limited mobility. The final sample consisted of 79 twin pairs ($$n = 158$$) who had complete data and met the minimum criteria for valid monitor “wear time”, defined as a minimum of 10 h per day of accelerometry wearing for 10 days [18] and any GPS measurement on each day. On average, participants wore both monitoring devices for 10.2 ± 3.1 days (median = 11, range = 2 to 14). Note that the lower end of the range indicates a participant who wore the accelerometer for the appropriate number of days but may have lacked GPS data on each of those days. ## 2.2. Exposures The primary exposure was neighborhood walkability, a measure of the BE estimated using Walk Score® (WS, Seattle, WA, USA) [19]. Each twin’s residential address and each GPS location coordinate were quantified using WS, which uses data from business listings, road networks, schools, and public transit to map walking distance to amenities in nine different categories weighted by importance [20]. Specific categories of points of interest include grocery stores, restaurants, coffee shops, banks, parks, schools, books (store or library), and various shopping and entertainment venues. The unpublished, proprietary algorithm then uses distances, counts, and weights to create a continuous score normalized on a scale of 0–100, with 0 representing the least and 100 being the most “walkable” neighborhoods [20]. This index is a valid measure of walkability [21]. In this study, WS values were used to classify home neighborhoods into five categories of neighborhood type: [1] Car dependent I (0–24), [2] Car dependent II (25–49), [3] Somewhat walkable (50–69), [4] Very walkable (70–89), and [5] Walker’s paradise (90–100). “Home WS” refers to WS corresponding to a participant’s home address, and “GPS WS” refers to WS values corresponding to logged GPS coordinates for that participant’s wear time over the two-week assessment. The GPS WS values were calculated for all coordinate points logged for that participant’s wear time (i.e., including walking and non-walking) and for walking bouts only (i.e., bouts with a median speed between 2 km/h to 6 km/h, see description of physical activity outcomes below). Further descriptions of GPS WS calculations follow. GPS WS was presented and analyzed as either a “within home” neighborhood (WHN) or an “out of home” neighborhood (OHN). The WHN GPS WS variables were assessed by averaging the walkability of GPS points that occurred within two defined neighborhood buffers surrounding the participant’s home address: [1] a 1-mile Euclidean (i.e., air1mi), and [2] a 1-mile network (i.e., net1mi) [22,23]. The OHN GPS WS variables were created by averaging the walkability of GPS points that occurred outside of those defined home neighborhood proximities. GPS WS was also presented and analyzed as “unweighted” or “weighted” values. The unweighted mean is the average WS of all GPS points. The weighted mean was weighted for the duration of time spent in each location and is the sum of WS values associated with each GPS point multiplied by the duration of the corresponding GPS point, divided by the total duration. The duration was defined as the average of the lead and lag time for each GPS point. ## 2.3. Outcomes Primary outcomes included diet quality (i.e., energy density), physical activity levels, and body mass index (BMI). The dietary energy density (DED) was assessed using a validated food-frequency questionnaire (FFQ), calculated by dividing the energy content of food items (in kilocalories) by the weight of food items (in grams) consumed. Two indicators of DED were analyzed: [1] only food, and [2] food and caloric beverages. The FFQ was developed by the Nutrition Assessment Shared Resource (NASR) of Fred Hutchinson Cancer Research Center (Seattle, WA, USA). Nutrient calculations were performed using the Nutrient Data System for Research (NDSR) software version 2017, developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN, USA. Physical activity levels were operationalized as walking bout minutes per week and moderate-to-vigorous physical activity (MVPA) bout minutes per week. Walking bouts were identified using a classification algorithm adapted from Kang et al. [ 24], described previously [25], and in brief below. MVPA bouts were identified as sustained intervals with 3D vector magnitude ≥ 2690 counts per minute (CPM) [26], using a modified 10-min bout definition that allows for up to two minutes outside the specified CPM threshold [18]. Accelerometry and GPS devices were configured to record data at 30-s intervals; data were combined into “LifeLogs” using common time stamps [11,27]. Light-to-moderate physical activity (LMPA) bouts used vector magnitude thresholds between 2000 and 6166 CPM. Walking bouts were identified as the subset of LMPA bouts that had [1] at least three records with GPS coordinates, [2] ≥$20\%$ of records with GPS coordinates, [3] median Doppler shift-based GPS speed between 2–6 kmh−1, and [4] spatial configuration. The spatial configuration criterion calculates the inter-point distance for all GPS coordinates in the bout and creates a minimum bounding circle (MBC) around the $95\%$ most tightly clustered points in the bout; bouts with MBC > 20 m that met all other criteria were flagged as walking. BMI was calculated from directly measured height and weight at a single in-person visit and expressed as kg/m2. ## 2.4. Covariates Participants’ age, sex, education level, and annual household income were used as covariates in the statistical analyses. Age was computed based on the reported date of birth. Sex was self-reported as male or female. Education level refers to the highest level of education, ranging from less than high school to a graduate/professional degree. This variable was recorded to categorize individuals as having a bachelor’s degree or above or not. Annual household income was self-reported in eight categories, ranging from “less than $20,000” to “$80,000 or more.” Household income was recoded into two categories for analysis: “less than $50,000,” and “$50,000 or more.” ## 2.5. Statistical Analysis All statistical analyses were performed in the statistical program R 4.0.2. Descriptive statistics are presented as means and standard deviations (for continuous variables) and counts and proportions (for categorical variables). Addressing research question 1, a series of linear mixed-effects models (LMMs) were used to examine the associations between home WS and GPS WS. For the LMMs, we first examined the association between an individual’s home WS and the average GPS WS. A random intercept was included to account for correlations between members of a twin pair, but not controlling for pair-level confounds. This is referred to as “phenotypic associations” (Model 1). Next, we included the mean home WS between twin pairs into the LMMs to estimate the average GPS WS (Model 2). A random intercept was included to control for within-pair correlations. These models control for the between-family confounds of the relationship between the home WS and the average GPS WS. The regression coefficient for the individual-level predictor represents the “quasi-causal” effect of the predictor on the outcome, controlling for between-pair confounds. The final sets of models further investigated the potential between-pair confounds, by including participants’ age, sex, education level (bachelor’s degree or above or not), and annual household income ($50,000 or more) into the LMMs (Model 3). Addressing research questions 2 and 3, similar sets of LMMs were performed to investigate the association between WS and DED, WS and physical activity outcomes (walking and MVPA), and WS and BMI. Physical activity was square root transformed, and BMI was log-transformed. Home and GPS WS were divided by 100 to allow variables to be on similar scales. Separate analyses were conducted for WHN versus OHN GPS WS, and unweighted versus duration-weighted GPS WS. For each analysis, the Bonferroni correction method was applied to control for Type I error rates. Adjusted p-values are indicated in the footnotes of each table. ## 3. Results Descriptive statistics of selected demographic characteristics and primary variables of interest are shown in Table 1. The age of the sample ranged from 23.4 to 75.3 years old ($M = 44.6$, SD = 12.4). Most participants self-identified as White ($85\%$) and non-Hispanic ($95.1\%$). In general, participants spent the highest proportion of their WHN time ($27\%$) and OHN time ($26\%$) in neighborhoods considered Very walkable (Supplementary Table S1). Supplementary Figure S1 illustrates the general study area and provides an example of a very walkable (left side) and car-dependent (right side) neighborhood. Addressing research question 1, Table 2 provides results from the LMMs estimating the associations between home WS and GPS WS. Results for Model 1 showed positive associations between home WS and WHN GPS WS ($b = 0.71$, SE = 0.03, $p \leq 0.001$ for air1mi; $b = 0.79$, SE = 0.03, $p \leq 0.001$ for net1mi), as well as OHN GPS WS ($b = 0.18$, SE = 0.04, $p \leq 0.001$ for air1mi; $b = 0.22$, SE = 0.04, $p \leq 0.001$ for net1mi). Thus, participants who reside in more walkable neighborhoods are more likely to spend more time in more walkable neighborhoods, both within and outside of their own home neighborhoods. Model 2 showed a quasi-causal effect of home WS on the average WHN GPS WS ($b = 0.63$, SE = 0.05, $p \leq 0.001$ for air1mi; $b = 0.76$, SE = 0.05, $p \leq 0.001$ for net1mi); thus, living in a more walkable neighborhood is associated with increased time spent in nearby walkable neighborhoods, after controlling for between-pair confounds. In contrast, results were attenuated and became non-significant in Model 2 for OHN GPS WS ($b = 0.08$, SE = 0.05, $$p \leq 0.140$$ for air1mi; $b = 0.10$, SE = 0.05, $$p \leq 0.063$$ for net1mi); thus, the association between home walkability and time spent in walkable areas outside of participants’ home neighborhoods is mediated by between-pair confounds. Results remained similar after accounting for additional confounding variables (Model 3). Comparable results were observed for weighted GPS WS. Addressing research question 2, Table 3, Table 4 and Table 5 present results of LMMs determining associations between WS with DED and physical activity. As shown in Table 3, there was no significant association between home WS and the two primary DED outcomes (Model 1). Results remained consistent when accounting for between-pair confounds (Model 2) and additional confounders (Model 3). Comparable results were observed when examining the associations between GPS WS and DED, across all buffer types, WHN versus OHN GPS WS, and unweighted versus duration-weighted GPS WS, across all Models (Table 4). Table 3 also shows results estimating the association between home WS and physical activity. There is a positive association between home WS with walking ($b = 4.89$, SE = −1.21, $p \leq 0.001$) and MVPA ($b = 3.03$, SE = 1.44, $$p \leq 0.037$$) (Model 1). Thus, individuals who reside in more walkable neighborhoods are more likely to have higher levels of physical activity. When between-pair confounds are included in Model 2, only the association between home WS and walking remained statistically significant ($b = 5.56$, SE = 1.72, $p \leq 0.001$), providing evidence for a quasi-causal effect of home WS on the amount of walking. However, there is no quasi-causal effect of home WS on the amount of MVPA ($b = 2.83$, SE = 1.96, $$p \leq 0.150$$). Results remained similar for both walking and MVPA after adjusting for additional confounders (Model 3). Comparable results were observed when examining the associations between unweighted GPS WS and physical activity (Table 5). Few significant associations were observed between duration-weighted WHN GPS WS and physical activity outcomes across all models. Quasi-causal associations were observed between duration-weighted OHN GPS WS with walking ($b = 11.09$, SE = 3.47, $$p \leq 0.002$$ for air1mi; $b = 10.17$, SE = 3.72, $$p \leq 0.008$$ for net1mi), but not MVPA when adjusting significance level for multiple comparisons ($b = 10.62$, SE = 3.86, $$p \leq 0.007$$ for air1mi; $b = 10.83$, SE = 4.14, $$p \leq 0.011$$ for net1mi) (Model 2). Results remained similar after accounting for additional confounding variables (Model 3). Addressing research question 3, Table 3 and Table 6 present results of LMMs determining associations between WS with BMI. In summary, there was no significant phenotypic association between home or GPS WS with BMI. ## 4. Discussion In this research, we investigated the effects of the BE on lifestyle behaviors in a community-based sample of adult identical twins who were reared together but now live apart. This unique sample allowed us to examine environmental influences on health-related outcomes, controlling for the between-family effects that might otherwise introduce selection biases into the choice of living environments. Specifically, the present study results provide novel insight into comparing the relationships between walkability measured at the location of individuals’ homes and at their actual daily locations with health behaviors, including physical activity and diet. Results suggest that individuals who reside in more walkable neighborhoods are more likely to spend more time in more walkable neighborhoods overall. Findings related to home walkability and spending time in walkable neighborhoods close to home remained significant in fully adjusted models (i.e., Model 3) and were thus quasi-causal, whereas the effect of time (i.e., weighted GPS) spent in walkable areas outside of the home neighborhood was confounded by between-family factors. Quasi-causal associations between both home and GPS WS with the amount of weekly walking were observed in fully adjusted models, but not with MVPA. Walkability surrounding participants’ homes and GPS locations were not associated with DED, a proxy measure of diet quality, or BMI. The current project offers innovative solutions to control confounding influences surrounding the association of BE with health. It builds on previous work by considering the BE over each twin’s objectively measured activity space—the environment in which they live, work, and spend time daily. Present findings indicate that those who reside in more walkable neighborhoods were more likely to spend time in more walkable neighborhoods, especially in locations near home. This novel finding objectively verifies a common assumption among many geographic and BE studies that the home BE affects behavior. Results also indicate that those who live and spend time in nearby walkable neighborhoods exhibit higher levels of walking. These findings support the larger body of literature identifying BE factors to be predictive of community physical activity. However, some studies have reported conflicting evidence, particularly in determining causal mechanisms [2,3,4]. For example, prior cross-sectional studies have shown that densities of community resources for a healthful diet and physical activity (e.g., food sources and greenspaces) were positively related to diet quality/composition [28,29] and physical activity measures [30,31]. In contrast, a review of neighborhood-based natural experiments reported a lack of longitudinal evidence that change in neighborhood BE was related to a change in residents’ diets and physical activity behaviors [32]. It should be noted that even minor changes to health behaviors attributable to the BE at the individual level have the potential to improve health outcomes at the population level [32,33]. In the current sample, neither participants’ homes nor GPS WS were related to levels of MVPA after adjusting for between family and demographic confounders. This aligns with previous research focused on BE influences on the location-based physical activity performed at higher intensities [34,35], including previous studies with the WSTR [13]. It expands on this work by introducing a location- and duration-based means to assess regular exposure to BE. Multiple study results indicated that associations between walkability features and MVPA were only significant when predictors and outcomes were located within the same geographic buffer [34,35]. Specifically, BE constructs surrounding the home were related to MVPA performed in the home neighborhood, but not total MVPA [34,35]. Importantly, MVPA was also reported to be performed more frequently outside of home and work neighborhoods compared to within [34]. Some studies suggest additional community-related factors drive the relationship between neighborhood environment and MVPA, such as perceived neighborhood safety [36] and greenspace [37]. This said, conflicting findings exist linking BE variables with location-based MVPA [36,37,38]. As is common in geospatial research, discrepancies in study findings may be attributed to nuanced differences in BE assessment, differing social constructs of geographic locations across studies, and differences in measures of health outcomes. Overall, the present study findings, in combination with previous literature, suggest the continued need for careful consideration of the spatial context in relation to health behaviors. Further qualitative and mixed-methods research may be needed to fully understand how individuals interact within their living and working environment, and location-based drivers of MVPA. In the current sample, the participant’s home and GPS WS were not related to DED or BMI. Previous studies have suggested that, especially in areas with a higher density of unhealthy food outlets, increased walkability may increase access to less healthful, energy-dense fast foods, and thus may have a negative impact on dietary outcomes [39,40] that could potentially nullify health benefits of either increased supermarket access or physical activity opportunities. There is substantial evidence to confirm that individuals residing in and traveling through neighborhoods with a high density of unhealthy food outlets consume more energy-dense foods and fewer fresh fruits and vegetables [5,41,42]. The current study utilized Walk Score® as a measure of the BE, which scored participants’ residences and actual locations primarily based on the surrounding density of businesses, but without necessarily considering the differential influence of specific business types (i.e., grocery versus parks versus entertainment). Given the previously confirmed associations between food access and physical activity behaviors with dietary intake, future studies could expand on this work by examining how standard BE features (i.e., street connectivity, population density, public transit, etc.) are spatially related to health-promoting food resources, and further investigate their independent and cumulative relationships with a variety of outcomes, including BMI. There are several strengths of the current study. First, the combined use of an accelerometer and GPS to determine physical activity allowed for precise and objective measurement of physical activity behaviors, including type, amount, and geospatial location. The assessment period being two weeks in duration is a study strength [34,43], and further included both weekdays and weekends, which is necessary to provide an accurate reflection of a person’s “usual” activity [44,45]. GPS WS were analyzed using two different buffer types (Euclidean and network), Within versus Out of home WS, and as both unweighted and duration weighted to allow for comparison and more precise interpretation of BE influences. Our scientific approach is genetically informed and integrates conceptual models from the behavioral and social sciences with biological, computational, and physical measures, and thus considers both between-family confounds and standard demographic control variables. Discussion of study weaknesses is warranted. It should be noted that although the twin design can effectively control for between-family confounds and is considered as “quasi-causal,” causality—especially reverse causality—cannot necessarily be inferred due to the cross-sectional design. The sample was primarily female ($72.2\%$), White ($85.0\%$), and educated at the bachelor’s level and above ($78.9\%$); thus, the generalizability of the study findings is limited. Dietary outcomes were self-reported, and as such, are subject to social desirability and other biases. The commercially available Walk Score® algorithm for estimating walkability is primarily an index of utilitarian destinations accessible by walking and does not equally include other measures of urban form that could influence physical activity; furthermore, the proprietary nature of the Walk Score® source data and algorithm reduce interpretability. Finally, while accelerometers record continuously, cold starts and signal impedance from urban canyons or other obstructions may result in incomplete data from GPS data loggers, leading to underestimation of the duration spent in various locations. ## 5. Conclusions In conclusion, the present study coupled innovative technology with advanced methods in geospatial data analysis and integrated spatial databases to determine relationships between the BE with health behaviors in a real-time and space continuum. Specifically, we investigated the effects of the BE on DED, walking, MVPA, and BMI in a community-based sample of adult identical twins to examine environmental influences on health-related outcomes independent of between-family confounders. A quasi-causal relationship was found between home neighborhood walkability with time spent in nearby walkable neighborhoods. Individuals who resided in and spent time in more walkable neighborhoods were more likely to exhibit higher levels of physical activity; relationships remained significant for walking, but not for MVPA, after adjusting for between-family confounds and demographic covariates. Walkability surrounding participants’ homes and GPS locations were not associated with DED or BMI. Urban policy and city planning should continue to promote neighborhood BE to improve walking and related community-level health outcomes. Future studies can build on this work by further investigating the importance of combined diet- and walk-related BE, and community-related drivers of physical activity modes performed at a higher intensity. ## References 1. 1. National Heart, Lung, and Blood Institute Predictors of Obesity, Weight Gain, Diet, and Physical Activity WorkshopNational Institutes of HealthBethesda, MD, USA2004. Predictors of Obesity, Weight Gain, Diet, and Physical Activity Workshop (2004.0) 2. Sallis J.F., Cerin E., Kerr J., Adams M.A., Sugiyama T., Christiansen L.B., Schipperijn J., Davey R., Salvo D., Frank L.D.A.. Built environment, physical activity, and obesity: Findings from the international physical activity and environment network (IPEN) adult study. Annu. Rev. Public Health (2020.0) 41 119-139. DOI: 10.1146/annurev-publhealth-040218-043657 3. Smith M., Hosking J., Woodward A., Witten K., MacMillan A., Field A., Baas P., Mackie H.. Systematic literature review of built environment effects on physical activity and active transport–an update and new findings on health equity. Int. J. Behav. Nutr. Phys. Act. (2017.0) 14 1-27. DOI: 10.1186/s12966-017-0613-9 4. Kärmeniemi M., Lankila T., Ikäheimo T., Koivumaa-Honkanen H., Korpelainen R.. The built environment as a determinant of physical activity: A systematic review of longitudinal studies and natural experiments. Ann. Behav. Med. (2018.0) 52 239-251. DOI: 10.1093/abm/kax043 5. Drewnowski A., Specter S.E.. Poverty and obesity: The role of energy density and energy costs. Am. J. Clin. Nutr. (2004.0) 79 6-16. DOI: 10.1093/ajcn/79.1.6 6. Engler-Stringer R., Le H., Gerrard A., Muhajarine N.. The community and consumer food environment and children′s diet: A systematic review. BMC Public Health (2014.0) 14. DOI: 10.1186/1471-2458-14-522 7. Sallis J.F., Cerin E., Conway T.L., Adams M.A., Frank L.D., Pratt M., Salvo D., Schipperijn J., Smith G., Cain K.L.. Physical activity in relation to urban environments in 14 cities worldwide: A cross-sectional study. Lancet (2016.0) 387 2207-2217. DOI: 10.1016/S0140-6736(15)01284-2 8. Roemmich J.N., Epstein L.H., Raja S., Yin L., Robinson J., Winiewicz D.. Association of access to parks and recreational facilities with the physical activity of young children. Prev. Med. (2006.0) 43 437-441. DOI: 10.1016/j.ypmed.2006.07.007 9. Wu S., Cohen D., Shi Y., Pearson M., Sturm R.. Economic analysis of physical activity interventions. Am. J. Prev. Med. (2011.0) 40 149-158. DOI: 10.1016/j.amepre.2010.10.029 10. Gordon-Larsen P.. Food availability/convenience and obesity. Adv. Nutr. (2014.0) 5 809-817. DOI: 10.3945/an.114.007070 11. Duncan G.E., Hurvitz P.M., Moudon A.V., Avery A.R., Tsang S.. Measurement of neighborhood-based physical activity bouts. Health Place (2021.0) 70 102595. DOI: 10.1016/j.healthplace.2021.102595 12. Duncan G.E., Mills B., Strachan E., Hurvitz P., Huang R., Moudon A.V., Turkheimer E.. Stepping towards causation in studies of neighborhood and environmental effects: How twin research can overcome problems of selection and reverse causation. Health Place (2014.0) 27 106-111. DOI: 10.1016/j.healthplace.2014.02.008 13. Duncan G.E., Cash S.W., Horn E.E., Turkheimer E.. Quasi-causal associations of physical activity and neighborhood walkability with body mass index: A twin study. Prev. Med. (2015.0) 70 90-95. DOI: 10.1016/j.ypmed.2014.11.024 14. Whitfield J.B., Zhu G., Heath A.C., Martin N.G.. Choice of residential location: Chance, family influences, or genes?. Twin Res. Hum. Genet. (2005.0) 8 22-26. DOI: 10.1375/twin.8.1.22 15. Willemsen G., Posthuma D., Boomsma D.I.. Environmental factors determine where the Dutch live: Results from the Netherlands twin register. Twin Res. Hum. Genet. (2005.0) 8 312-317. DOI: 10.1375/twin.8.4.312 16. Duncan G.E., Avery A.R., Strachan E., Turkheimer E., Tsang S.. The Washington State Twin Registry: 2019 Update. Twin Res. Hum. Genet. (2019.0) 22 788-793. DOI: 10.1017/thg.2019.36 17. Strachan E., Hunt C., Afari N., Duncan G., Noonan C., Schur E., Watson N., Goldberg J., Buchwald D.. University of Washington Twin Registry: Poised for the next generation of twin research. Twin Res. Hum. Genet. (2013.0) 16 455-462. DOI: 10.1017/thg.2012.124 18. Troiano R.P., Berrigan D., Dodd K.W., Masse L.C., Tilert T., McDowell M.. Physical activity in the United States measured by accelerometer. Med. Sci. Sport. Exerc. (2008.0) 40 181-188. DOI: 10.1249/mss.0b013e31815a51b3 19. Walk Score: Get Your Walk Score—A Walkability Score for any Address. (2013.0) 20. Walk Score: Walk Score Methodology 21. Carr L.J., Dunsiger S.I., Marcus B.H.. Validation of Walk Score for estimating access to walkable amenities. Br. J. Sport. Med. (2011.0) 45 1144-1148. DOI: 10.1136/bjsm.2009.069609 22. Hurvitz P.M., Moudon A.V.. Home versus nonhome neighborhood: Quantifying differences in exposure to the built environment. Am. J. Prev. Med. (2012.0) 42 411-417. DOI: 10.1016/j.amepre.2011.11.015 23. Hurvitz P.M., Moudon A.V., Kang B., Fesinmeyer M.D., Saelens B.E.. How far from home? The locations of physical activity in an urban U.S. Setting. Prev. Med. (2014.0) 69 181-186. DOI: 10.1016/j.ypmed.2014.08.034 24. Kang B., Moudon A.V., Hurvitz P.M., Reichley L., Saelens B.E.. Walking objectively measured: Classifying accelerometer data with GPS and travel diaries. Med. Sci. Sport. Exerc. (2013.0) 45 1419-1428. DOI: 10.1249/MSS.0b013e318285f202 25. Hwang L.D., Hurvitz P.M., Duncan G.E.. Cross Sectional Association between Spatially Measured Walking Bouts and Neighborhood Walkability. Int. J. Environ. Res. Public Health (2016.0) 13. DOI: 10.3390/ijerph13040412 26. Sasaki J.E., John D., Freedson P.S.. Validation and comparison of ActiGraph activity monitors. J. Sci. Med. Sport (2011.0) 14 411-416. DOI: 10.1016/j.jsams.2011.04.003 27. Hurvitz P.M., Moudon A.V., Kang B., Saelens B.E., Duncan G.E.. Emerging technologies for assessing physical activity behaviors in space and time. Front. Public Health (2014.0) 2 2. DOI: 10.3389/fpubh.2014.00002 28. Moore L.V., Diez Roux A.V., Nettleton J.A., Jacobs D.R.. Associations of the local food environment with diet quality—A comparison of assessments based on surveys and geographic information systems: The multi-ethnic study of atherosclerosis. Am. J. Epidemiol. (2008.0) 167 917-924. DOI: 10.1093/aje/kwm394 29. Zenk S.N., Lachance L.L., Schulz A.J., Mentz G., Kannan S., Ridella W.. Neighborhood Retail Food Environment and Fruit and Vegetable Intake in a Multiethnic Urban Population. Am. J. Health Promot. (2009.0) 23 255-264. DOI: 10.4278/ajhp.071204127 30. Kaczynski A.T., Besenyi G.M., Stanis S.A.W., Koohsari M.J., Oestman K.B., Bergstrom R., Potwarka L.R., Reis R.S.. Are park proximity and park features related to park use and park-based physical activity among adults? Variations by multiple socio-demographic characteristics. Int. J. Behav. Nutr. Phys. Act. (2014.0) 11 1-14. DOI: 10.1186/s12966-014-0146-4 31. Cohen D.A., Ashwood J.S., Scott M.M., Overton A., Evenson K.R., Staten L.K., Porter D., McKenzie T.L., Catellier D.. Public parks and physical activity among adolescent girls. Pediatrics (2006.0) 118 e1381-e1389. DOI: 10.1542/peds.2006-1226 32. MacMillan F., George E.S., Feng X., Merom D., Bennie A., Cook A., Sanders T., Dwyer G., Pang B., Guagliano J.M.. Do Natural Experiments of Changes in Neighborhood Built Environment Impact Physical Activity and Diet? A Systematic Review. Int. J. Environ. Res. Public Health (2018.0) 15. DOI: 10.3390/ijerph15020217 33. Rose G.. Sick individuals and sick populations. Int. J. Epidemiol. (2001.0) 30 427-432. DOI: 10.1093/ije/30.3.427 34. Troped P.J., Wilson J.S., Matthews C.E., Cromley E.K., Melly S.J.. The built environment and location-based physical activity. Am. J. Prev. Med. (2010.0) 38 429-438. DOI: 10.1016/j.amepre.2009.12.032 35. Rodríguez D.A., Brown A.L., Troped P.J.. Portable global positioning units to complement accelerometry-based physical activity monitors. Med. Sci. Sport. Exerc. (2005.0) 37 S572-S581. DOI: 10.1249/01.mss.0000185297.72328.ce 36. Moore J.B., Brinkley J., Crawford T.W., Evenson K.R., Brownson R.C.. Association of the built environment with physical activity and adiposity in rural and urban youth. Prev. Med. (2013.0) 56 145-148. DOI: 10.1016/j.ypmed.2012.11.019 37. Boakye K.A., Amram O., Schuna J.M., Duncan G.E., Hystad P.. GPS-based built environment measures associated with adult physical activity. Health Place (2021.0) 70 102602. DOI: 10.1016/j.healthplace.2021.102602 38. Frank L.D., Schmid T.L., Sallis J.F., Chapman J., Saelens B.E.. Linking objectively measured physical activity with objectively measured urban form: Findings from SMARTRAQ. Am. J. Prev. Med. (2005.0) 28 117-125. DOI: 10.1016/j.amepre.2004.11.001 39. Chandrabose M., Cerin E., Mavoa S., Dunstan D., Carver A., Turrell G., Owen N., Giles-Corti B., Sugiyama T.. Neighborhood walkability and 12-year changes in cardio-metabolic risk: The mediating role of physical activity. Int. J. Behav. Nutr. Phys. Act. (2019.0) 16 1-11. DOI: 10.1186/s12966-019-0849-7 40. Tabaei B.P., Rundle A.G., Wu W.Y., Horowitz C.R., Mayer V., Sheehan D.M., Chamany S.. Associations of residential socioeconomic, food, and built environments with glycemic control in persons with diabetes in New York City from 2007–2013. Am. J. Epidemiol. (2018.0) 187 736-745. DOI: 10.1093/aje/kwx300 41. Cutumisu N., Traore I., Paquette M.C., Cazale L., Camirand H., Lalonde B., Robitaille E.. Association between junk food consumption and fast-food outlet access near school among Quebec secondary-school children: Findings from the Quebec Health Survey of High School Students (QHSHSS) 2010–2011. Public Health Nutr. (2017.0) 20 927-937. DOI: 10.1017/S136898001600286X 42. Williams J., Scarborough P., Matthews A., Cowburn G., Foster C., Roberts N., Rayner M.. A systematic review of the influence of the retail food environment around schools on obesity-related outcomes. Obes Rev. (2014.0) 15 359-374. DOI: 10.1111/obr.12142 43. Tamura K., Wilson J.S., Goldfeld K., Puett R.C., Klenosky D.B., Harper W.A., Troped P.J.. Accelerometer and GPS data to analyze built environments and physical activity. Res. Q. Exerc. Sport (2019.0) 90 395-402. DOI: 10.1080/02701367.2019.1609649 44. Cain K.L., Geremia C.M.. Accelerometer Data Collection and Scoring Manual for Adult & Senior Studies (2011.0) 45. Tudor-Locke C., Camhi S.M., Troiano R.P.. Peer reviewed: A catalog of rules, variables, and definitions applied to accelerometer data in the National Health and Nutrition Examination Survey, 2003–2006. Prev. Chronic Dis. (2012.0) 9 E113. PMID: 22698174
--- title: Intelligent Fuzzy System to Predict the Wisconsin Breast Cancer Dataset authors: - Yamid Fabián Hernández-Julio - Leonardo Antonio Díaz-Pertuz - Martha Janeth Prieto-Guevara - Mauricio Andrés Barrios-Barrios - Wilson Nieto-Bernal journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049073 doi: 10.3390/ijerph20065103 license: CC BY 4.0 --- # Intelligent Fuzzy System to Predict the Wisconsin Breast Cancer Dataset ## Abstract Decision Support Systems (DSSs) are solutions that serve decision-makers in their decision-making process. For the development of these intelligent systems, two primary components are needed: the knowledge database and the knowledge rule base. The objective of this research work was to implement and validate diverse clinical decision support systems supported by Mamdani-type fuzzy set theory using clustering and dynamic tables. The outcomes were evaluated with other works obtained from the literature to validate the suggested fuzzy systems for categorizing the Wisconsin breast cancer dataset. The fuzzy Inference Systems worked with different input features, according to the studies obtained from the literature. The outcomes confirm that most performance’ metrics in several cases were greater than the achieved results from the literature for the output variable for the different Fuzzy Inference Systems—FIS, demonstrating superior precision. ## 1. Introduction Cancer is a group of diseases that cause cells in the body to change and spread out of control [1]. Breast cancer is considered the second most common cancer among women in the United States (some kinds of skin cancer are the most common). According to [2], among the signs and symptoms of breast cancer, we can find a lump or swelling in the breast, upper chest, or armpit; changes in the size or shape of the breast; a change in skin texture and color; rash, crusting, or modifications to the nipple. For the mentioned causes, it is critical to create simulations that help in the decision-making process for initial detection, proper therapy, and therapy [3] to achieve a rapid diagnosis. Fuzzy systems have been used for breast cancer classification [4,5], among other uses. Fuzzy set theory is known as the basis of all fuzzy logic methods [6]. Fuzzy set theory was proposed by Zadeh [7] as an extension of the classical set theory to model sets, whose elements have degrees of membership [8]. According to [7], a fuzzy set is a class of objects with a continuum of grades of membership. Such a set is characterized by a membership (characteristic) function which assigns each object a category of membership ranging between zero and one. A degree of one means that an object is a member of the set, a value of zero means it is not a member, and a value somewhere in-between shows a partial degree of membership [8]. This partial degree of membership is also known as the membership function. The notions of inclusion, union, intersection, complement, relation, convexity, etc., are extended to such sets, and various properties of these notions in the context of fuzzy sets are established [7]. The fuzzy set theory provides the tools to effectively represent linguistic concepts, variables, and rules, becoming a natural model to represent human expert knowledge [9]. According to [8], a linguistic value refers to a label for describing the experience that has meaning determined by its degree of the membership function. One of the most fruitful developments of fuzzy set theory is Fuzzy Rule-Base Systems—FRBs [8]. The Fuzzy Decision Support System (FDSS) was developed to convert knowledge from experts based on fuzzy rules to improve decision making [6]. For the development of this kind of Decision Support system, the Mamdani-type FIS is widely used [10,11]. Fuzzy Decision Support Systems are used in the knowledge field of Medicine [11,12,13,14,15]. For these reasons, the main goal of this research work was to create different intelligent fuzzy systems using clusters and dynamic tables for the classification of the Wisconsin breast cancer dataset. To validate the proposed models, the fuzzy inference systems—FIS—were conceived to classify the mentioned dataset and contrasted with other artificial intelligent technique models obtained from the literature. The originality of this work lies in its generation of membership functions. Some authors use different approaches for generation. We can find 2N + 1 regions, FCM, neural networks, GAs, etc. In our case, we proposed using clustering methods for this step. The main difference at this stage is that no fixed or random membership functions were generated, such as those caused by those works that used classical methods or were based on evolutionary algorithms, neural networks, or swarm intelligence techniques. Another difference between this study and the related works using neural networks, evolutionary or swarm algorithms is that we did not use random numbers, or any chromosome or particle scheme. Regarding the generation of the rule base for the system, some authors also used the same previously mentioned methods. The main difference with our work is that our approach uses pivot tables instead of other techniques. Other authors initialize with random weights and bias for each hidden neuron (neural networks), adjusting them through optimization functions such as gradient descendent and non-linear activation functions. Other methods use random schemes to generate the fuzzy rules, using the objectives function to adjust membership functions and the rule base, i.e., MSE. Our study did not propose using any objective function as a minimization problem. In addition, our study did not offer to employ or calculate any distances, attractiveness, or another parameter to generate the fuzzy rule base. The only component used for this task was pivot tables. Pivot tables did not use any calculation method or random or manual parameters (only sorting options). The main job of this technique is to eliminate redundant information. ## 2. Material and Methods To validate the framework proposed by [16,17]—(Figure 1), a case study was designed and implemented. Each of the stages suggested in the framework will be explained. ## 2.1. Identifying the Dataset The dataset used for this research was obtained from the UCI Machine Learning repository to evaluate the efficacy of the proposed framework using the Wisconsin Breast Cancer Dataset (WBCD) [18,19]. The dataset was compiled from the patients of the University of Wisconsin–Madison Hospitals. The instance of this dataset is a 699 data pair. The dataset contains missing values. In this case, the character “?” was changed to zero. This change was made because the values of the input variables are within the range of 1 to 10. In this case, the symbol “?” represents a lost value. It was decided not to apply statistical methods such as the mean of series, mean or median of nearby points, or linear interpolation because these variables are discrete variables. The result of the application of these methods is a continuous variable. When applying the mentioned methods, the results were always the same: 3.5. This makes the decision-making process difficult because it is not known to what value to assign it if it is three or four. In this case, the operation was manual, replacing the values of the symbols with the number zero. The number zero indicates that you do not have the value of that variable. The missing values belong to the variable Bare Nuclei (BN). Other changes were made in the dataset to adjust the number of the classification: two to one for benign instances and four to two for malignant instances. This change was made because if the system worked with these two values (2 and 4), some of the outputs obtained by the fuzzy inference system could be in the middle of the range of these two values; that is, number three could be obtained as an answer. In that case, this value could hinder the decision-making process, because this value cannot tell us if the selected instance is malignant or benign. This process was carried out manually, replacing the values of the output variable: two by one and four by two, comprising 458 benign cases and 241 malign cases. In this case, according to the data, the two classes are imbalanced. In this case, when an unbalanced dataset is reached, we usually obtain a high precision value in the Majority class (benign cases) and a low recall in the Minority class (malignant cases). However, according to the results obtained by the fuzzy inference system with better results (Table 1 and Table 2), these were excellent because the specificity value was 1.0, indicating that $100\%$ accuracy was obtained in the minority class (malignant instances). Because of this situation, the research team had no need to use strategies for handling unbalanced data, such as model parameter adjustments, modifying the dataset, using artificial or synthetic samples, or using balanced ensemble methods. The descriptive statistics of the dataset can be found in Onan [20]. There are nine input features and one output feature (Figure 2 and Figure 3). The attributes of the dataset are: Figure 2Input variables of the Wisconsin Breast Cancer dataset. Figure 3Output variable of the Wisconsin Breast Cancer dataset. ijerph-20-05103-t001_Table 1Table 1Confusion matrix for WCDB dataset. Specialists BenignMalignDDFDSSBenign4553Malign0241DDFDSS: Data-driven fuzzy decision-support system. Bold values represent accurate forecasts. ijerph-20-05103-t002_Table 2Table 2Performance metrics obtained with the proposed framework.[16]CVRSNum of variables: [2 4 5 6 8]K-MeansWardFCMK-Means WardFCM Num of Rules or Hidden neurons /technique248233190207208168Accuracy (%):$99.3\%$$99.4\%$$99.1\%$$99.0\%$$99.57\%$$98.43\%$Sensitivity:0.98570.98530.98510.99160.98770.9637Specificity:0.99690.9980.99390.98921.00000.9956F-Measure:0.98990.99070.98680.98540.99380.9775Area under curve:0.99330.99420.99030.98740.99670.986Kappa statistics:0.98450.98580.97980.97780.99050.9654CV: Cross-validation method. RS: random sampling. * Significant difference at $95\%$ of the Confidence Interval between them. ## 2.2. Data Preparation (Crisp Inputs) The first activity was to identify inputs and outputs features. In the experiment, all input and output features were chosen. In this phase, the data were processed because the dataset contains missing values. The symbol “?” was changed for zero (as mentioned above). In this stage, the preprocessed technique applied was clustering. This method is explained in Section 2.7 [21]. ## 2.3. Reviewing Existing Models In this phase, a search of the different related works about the problem was carried out. Several indexed databases, such as Scopus, Science Direct, among others, were used. The outcome of this stage is revealed in the discussion segment. ## 2.4. Evaluating the Optimal Number of Clusters In this stage, pivot tables were applied to determine the total number of rows for every input and output feature. For the experiment, the optimum number of clusters was 10. ## 2.5. Setting a Number of Clusters (Minimum and Maximum) According to the Previous Evaluation The minimum used value was two; the maximum quantity of clusters was ten for all input features. Two clusters were used for the output feature. The clusters’ optimal number was stablished as the maximum number of clusters to avoid greater fuzzy sets numbers in the output (input variables interaction). ## 2.6. Random Permutations For the dataset, the observed inputs and outputs values were randomized and commuted when used the suggested algorithms. ## 2.7. Cluster Analysis (Fuzzification Process) In this phase, three types of clusters (kmeans, Ward, and Fuzzy C-Means) were achieved and analyzed using the range of solutions created in the preceding stage. For the first two clustering algorithms, the Euclidean distance was selected as the default. For the fuzzy c-means clustering algorithm, the default value for the exponent for the fuzzy partition matrix U was 50, the maximum number of iterations was 100, the minimum improvement in the objective function was 1 × 10−5, and the objective function was displayed as false (zero). The selection criteria for clustering algorithm must correspond to the knowledge of the topic. The maximum number of clusters for each input and output feature for the dataset was the values of the optimum cluster (ten for the inputs variables and two for the output variable). ## 2.8. Sampling Datasets (Cross-Validation or Random Sampling) For the experiments, two methods of random data sampling were used. The first method used was random sampling, and the other one was the cross-validation method. For the random sampling, the user could select the percentage for every subset (training, validation and test), and the number of iterations. The default values for this kind of data partition method were as follows: training dataset: $70\%$; validation dataset: $30\%$; test dataset: $0\%$ [22]; and number of iterations: 3000. For the cross-validation process, the k-fold method was selected. For the case study, the used value of k by default was 10. This value was selected because it is one of the most used in this type of validation method of machine learning models. In this case, the dataset was divided into 10 equal parts, with an equal number of training and validations. For all the experiments, we used a personal computer (PC). The computer’s specification for the algorithm’s implementation was an AMD A12-9720P Radeon R7, 12 compute cores 4C + 8G 2.70 GHz with 16.00 GB RAM, Hard Disk Drive (HDD) of 1 Tera Byte (TB). ## 2.9. Pivot Tables For the experiment, the unique tables command was applied for the implementation of the subsequent sub-stages. ## 2.9.1. Combining Different Input Variable Clusters Datasets This phase comprises creating arrangements between input features and the sets of output features using dynamic tables. The permutations were carried out by applying the command “nchoosek” and “unique” for matrixes. The first command sends back a matrix comprising all possible permutations of the elements of vector v taken k at a time. The second command returns a copy of dataset A, which contains only the sorted unique observations [23]. ## 2.9.2. Stablishing the Fuzzy Rules This phase is established on the preceding one. The procedures carried out with the use of the dynamic tables one or several permutations can be used to make the rule bases for the FIS. To achieve this, we can use the unique command to avoid rules duplication. For all the experiments, the Center of Gravity method was chosen as the de-fuzzification process by default, contemplating all output options and converting the fuzzy set originated by inference into a numerical value, as proposed by [24,25]. Generally, software programs for the implementation of this type of model use the Centroid method for defuzzification. This method can be considered a weighted average, where the weights are represented by μA (xi), which indicates the degree of membership of the value xi with the concept modeled by the fuzzy output set A, and which, in its compound shape, is calculated by:[1]Z=µc(z)zδzµc(z)δz where Z is the consequent variable and µc(z) is the function of the composed shape. The result of the defuzzification process Z can be continuous or discrete [26]. ## 2.10. Elaborating the Decision Support System Based on Fuzzy Set Theory (Inference Engine) For the experiments, the FIS’ implementation was carried out in the MATLAB® R2017a software. In this stage, the aim is to join all components cited above in order. The first step was to generate a new FIS file. We put a name to the created FIS file. To define the defuzzification process, by default, we selected the centroid defuzzification method (a choice between “Centroid”, “som—small of maximum”, “mom—mean of maximum”, or “lom—large of maximum”). For all the experiments’ implementation, the fuzzy logic toolbox was not used because this tool does not work with a data-driven approach. This means that it is not an automated fuzzy inference systems developer. All fuzzy inference systems designed with this toolbox are developed manually. Instead, we used our algorithms (Pseudocodes available in the appendixes of Reference [16]). ## 2.11. Evaluating the Fuzzy System Performance (Defuzzification and Crisp Outputs) For the experiments, the system’s performance was measured through some of the following metrics: the classification accuracy (ACC), sensitivity, specificity, function of measure, area under the curve, and Kappa statistics. Additionally, we performed a statistical significance test called McNemar’s test. The aim is to examine whether the differences between the prediction performances of feature subsets are statistically significant or not [27]. This test was applied in those results that used random sampling as a data partition method only, because using a cross-validation, we have more than one confusion matrix; we have k-folds confusion matrixes; however, we calculate the test for all three clustering methods in all results. ## 3. Results and Discussion The following were the obtained outcomes for the cited dataset: The confusion matrix for the mentioned data-driven fuzzy clinical decision support system (DDFCDSS) are shown in Table 1. The performance metrics obtained with our proposed framework are shown in Table 2. The best results for a set of five features were obtained via the Ward clustering method. As can be seen, the DDFCDS had a specificity value of $100\%$, suggesting an outstanding performance predicting or classifying the true negatives cases of the WBCD. It means that all malignant cases were classified correctly. According to the confusion matrix, there are only three true positive values that are misclassified corresponding to a sensitivity value of 0.9877. In the following pages, we are going to compare the results obtained from the literature with our results. The results shown in the tables below correspond to the same characteristics noted by the researchers using the same dataset (WBCD). We used the same data partition method, the same features. According to the results, for the WBCD, the greatest performance belongs to Onan [20]. The author used a classification model based on the fuzzy–rough nearest neighbor algorithm, consistency-based feature selection, and fuzzy–rough instance selection for a medical diagnosis. He used a 10-fold cross-validation method as a data partition method. As can be seen in Table 3, the classification accuracy for his results was $99.72\%$, and the maximum value for classification accuracy of our results belongs to the k-means 10-fold cross-validation method. The sensitivity value for the author was $100\%$; however, his specificity value was 0.9947. Our results show the opposite. Our specificity value was 1.0, and the sensitivity value was 0.9703. The performance metric sensitivity indicates the true positive (TP) rate, and specificity means the true negative (TN) rate [28]. According to [28], in breast cancer, the TP signifies cases that are correctly categorized in the benign tumor, and the TN characterizes cases that are correctly categorized in the malignant tumor. This result shows that our model predicts $100\%$ of the true negative values. In this case, we can state that if a tumor is malignant, the fuzzy inference system is going to be classified as malignant with $100\%$ accuracy. Through the comparison of the three clustering methods results, we found that McNemar’s test indicated that none of them perform significantly better than the others, indicating that all the DDFCDSS have the same classification error rates. The test results were Ward vs. k-means: X12 = 0.0455; k-means vs. FCM: X12 = 0.0, and Ward vs. FCM: X12 = 0.12903, respectively. Ref. [ 29] proposed a Breast Cancer Computer Aid Diagnosis (BC-CAD) based on joint variable selection and a Constructive Deep Neural Network “ConstDeepNet”. A feature variable selection method was applied to decrease the number of inputs used to train a Deep Learning Neural Network. The authors used five-fold cross-validation as a partition data method. The classification accuracy for the set of features mentioned in Table 4 is $96.2\%$. Our results were higher than those obtained for these authors. Our classification accuracy using the cross-validation data partition method with $k = 5$ was $98.37\%$. For comparison of the three clustering methods, the McNemar’s test results are as follows: K-means vs. Ward: X12 = 0.3636; k-means vs. FCM: X12 = 1.8947, and Ward vs. FCM: X12 = 0.5625, indicating no significant differences between them. For the case of the second set of features used by the authors (Table 5), the classification accuracy obtained by the constructive deep neural network was $96.6\%$. Our results for the same set of features were higher than those obtained by the authors. Regarding the McNemar’s test results for the three clustering methods, they indicate that there is no significant difference among them. The test values are k-means vs. Ward: X12 = 0.3636; k-means vs. FCM: X12 = 1.8947; Ward vs. FCM: X12 = 0.5625. Another author who works with the same dataset was [30]. The authors introduced an automated medical data classification method using wavelet transformation (WT) and interval type-2 fuzzy logic system (IT2FLS. The authors used five-fold cross-validation as a data partition method. The classification accuracy for this set of features was $97.88\%$ (Table 6). The best performance of the three clustering methods was obtained for the Ward method, with $96.68\%$ showing a better performance between the models. Regarding the McNemar’s test results, the values were: K-means vs. Ward: X12 = 14.0192; K-means vs. FCM: X12 = 0.0294; Ward vs. FCM: X12 = 12.5000. The values higher than 3.84 can be interpreted as a significant difference. This means that we reject the null hypothesis and accept the alternative hypothesis indicating that the algorithms do not have the same classification error rate. In this case, the DDFCDSS using the k-means and FCM have the same classification error rates. Ref. [ 31] developed a manually Mamdani-type fuzzy inference system (FIS). The authors proposed a framework for the development of fuzzy inference systems using dynamic tables and clusters; however, the framework does not support a data-driven approach. The classification accuracy for the authors was $98.58\%$ (Table 7), showing a sensitivity of $100\%$; however, the specificity is lower than our results. The best performance for our DDFCDS was obtained by the k-means method using random sampling as a data partition method. McNemar’s test indicates that k-means vs. FCM has significant difference between them. The test results values are the following: k-means vs. Ward: X12 = 3.0625; k-means vs. FCM = X12 = 8.6538, and Ward vs. FCM: X12 = 2.2273. The other authors who had better results than our DDFCDSS were Abdel-Zaher and Eldeib [32]. Ref. [ 32] proposed an integration between Wavelet Transformation (WT) and Interval Type-2 Fuzzy Logic Systems (IT2FLS) to cope with both high-dimensional data challenge and uncertainty. The authors used all input variables and used random sampling (70–$30\%$) for data partition. The classification accuracy for this author was $99.68\%$, with a sensitivity of $100\%$ and a specificity of 0.9947 (Table 8). Our best performance using the same data partition was the k-means DDFCDSS, with a classification accuracy of 98.86. McNemar’s test showed that the highest performance of the DBN model, which uses nine variables, was significantly better than our Data-Driven Fuzzy CDSS, which has the highest performance. For the comparison among the three clustering methods, the test results suggest that they have no significant difference among them. The values for the test are K-means vs. FCM: X12 = 2.400; K-means vs. Ward: X12 = 1.250; Ward vs. FCM: X12 = 0. Ref. [ 28] proposed a fully connected layer first CNN (FCLF-CNN), in which the fully connected layers are embedded before the first convolutional layer. The authors used two data partition methods for the experiments. The authors used a five-fold cross-validation approach. The obtained results for this scheme are presented in Table 8. The authors also used two settings for the random sampling (train: $50\%$, test: $50\%$, and train: $75\%$, test: $25\%$). The results obtained from the random sampling were $98.57\%$ and $98.86\%$, respectively. As can be seen in Table 8, our proposed framework could obtain a better performance in the cross-validation method: The Ward method obtained a classification accuracy of $98.84\%$. Regarding the random sampling method, the k-means obtained the best performance with a classification accuracy of $98.86\%$, which was similar to the results obtained by [28] for the same dataset and random sampling configuration. The main differences and similarities between the mentioned related works with the proposed framework are as follows:[1]Like all the mentioned works, we identified all the input and output variables for the Wisconsin Breast cancer dataset classification problem, including the related works using the same datasets.[2]To generate the membership functions, the mentioned authors used different approaches, including logistic regression, support vector machine, random forest, fuzzy c-means, neural networks (MLP, DNN), K-nearest neighbor, genetic algorithms, etc. In our case, we proposed using clustering methods for this step. Among the clustering methods, we used k-means, the Ward method, and FCM. The main difference at this stage is that no fixed or random membership functions, such as those caused by those works that used classical methods or were based on evolutionary algorithms (GAs, FA, BBO), neural networks, or swarm intelligence techniques (PSO, ACO), were generated. Instead, the users can select the number of membership functions (number of clusters) they want to use for each input/output variable for each classification problem. Another difference between our framework and the related works using neural networks, evolutionary, or swarm algorithms is that we did not use random numbers, chromosomes, or particle schemes. Instead, our membership functions were obtained using well-known and recognized clustering methods. They indicate whether a sample belongs to a group, obtaining a vector with the values of one of the groups assigned to the input/output variable. Thus, the assignation of the number of groups is not random. In addition, we did not use any random population, random particles, random weights, or bias.[3]To generate the system’s rule base, the main difference between our work and the mentioned works is that our approach uses pivot tables instead of other techniques. As mentioned, every method for generating the intelligent systems’ rules or connections has its own characteristics. Some initialize with random weights and bias for each hidden neuron (neural networks), adjusting them through optimization functions such as gradient descendent and non-linear activation functions. Other methods use random schemes to generate the fuzzy rules, using the objectives function to change membership functions and the rule base, i.e., Mean Square Error (MSE). Our proposed framework did not offer the chance to use any objective function as a minimization problem. Additionally, our framework did not suggest using or calculating any distances, attractiveness, or parameter to generate the fuzzy rule base. The only component used for this task was pivot tables. Pivot tables did not use any calculation method or random or manual parameters (only sort options). The primary mission of this technique is to eliminate redundant information. Our framework’s main advantage is our algorithms’ simplicity using only primitive mathematical operators and clustering operations (Appendixes shown in Reference [16]). Our framework’s parameters are as follows: (a) To select inputs and outputs variables. ( b) To choose the clustering algorithm (k-means, Ward, FCM). ( c) To select the number of Membership Functions—MFs (number of clusters)—that the user wants. ( d) To adopt the data partition method (random sampling or cross-validation). ( e) To select the number of features the user wants to use (feature extraction). ( f) To set the parameters according to the selected data partition method. For example, if the user selects random sampling, they must determine the percentages for training, validation, and test datasets, and the number of iterations. Otherwise, the users must choose the cross-validation partition method (‘k’,’KFold’, ‘Holdout’,’LeaveOut’, or ‘Resubstitution’) and the iterations’ number. As can be read, among the parameters, there is nothing about a lower–upper bound, any random number, any inertia, momentum, distance, weight, bias, or population size to calculate or initialize. This means that the result of each iteration for every combination (Section 2.9.1. Combining different cluster datasets) is a fuzzy inference system because it is not necessary to adjust or optimize weights, bias, or any objective or fitness function. It should be noted that the only parameters configured internally were those used for the clustering methods, and they are mentioned in Section 2.7, Clusters analysis (Fuzzification process). These criteria have low computational requirements, offering precision, processing speed, and interpretability of the rules. ## 4. Conclusions The main objective of this research work was to implement and validate different decision support systems founded on Mamdani-type fuzzy set theory using clusters and dynamic tables. As could be demonstrated, in some cases, the proposed fuzzy models showed the best-performing indices related to this dataset, surpassing the outcomes obtained from advanced techniques (deep learning) such as Deep Neural Network and Convolutional Neural Networks. The obtained outcomes for the used performance metrics were nearer to one, indicating a robust fit between the predicted and the observed data. The area under the curve for this dataset ranged between 0.90 and 1.0, representing an excellent classification task [34]. The selected features shown in Table 2 for both data partition methods were: Uniformity of Cell Size (UCSi), Marginal Adhesion (MA), Single Epithelial Cell Size (SECS), Bare Nuclei (BN), and Normal Nucleoli (NN), indicating that it is not necessary to carry out the mitosis process accelerating diagnosis and a possible treatment [16,31]. According to the McNemar’s test results for the three clustering methods, the k-means have significant difference at $95\%$ of the confidence interval with the FCM clusters method (X12 = 5.7857), indicating that these two clusters methods have different error rate. For the other two clusters methods, the test evidenced that the clustering methods did not perform significantly differently. We can conclude that the current framework provides a real pattern for the development of data-driven Mamdani-type fuzzy decision-support systems for classification problems. Another conclusion is the computational performance of the algorithms has homogeneous behavior when running with similar datasets. Other main future work aims to implement this in other software development platform such as python, Scilab, and Octave, among others. ## References 1. 1. American Cancer Society Cancer Facts & Figures 2018American Cancer Society Inc.Atlanta, GA, USA201876. Cancer Facts & Figures 2018 (2018.0) 76 2. What are the Signs and Symptoms of Breast Cancer? 3. Hayat M.A., Hayat M.A.. Breast Cancer: An Introduction. Methods of Cancer Diagnosis, Therapy and Prognosis: Breast Carcinoma (2008.0) 1-3. DOI: 10.1007/978-1-4020-8369-3_1 4. Nilashi M., Ibrahim O., Ahmadi H., Shahmoradi L.. A knowledge-based system for breast cancer classification using fuzzy logic method. Telemat. Inform. (2017.0) 34 133-144. DOI: 10.1016/j.tele.2017.01.007 5. Gayathri B.M., Sumathi C.P.. Mamdani fuzzy inference system for breast cancer risk detection. Proceedings of the 2015 IEEE International Conference on Computational Intelligence and Computing Research (ICCIC) 1-6 6. Ahmadi H., Gholamzadeh M., Shahmoradi L., Nilashi M., Rashvand P.. Diseases diagnosis using fuzzy logic methods: A systematic and meta-analysis review. Comp. Methods Programs Biomed. (2018.0) 161 145-172. DOI: 10.1016/j.cmpb.2018.04.013 7. Zadeh L.A.. Fuzzy sets. Inform. Control (1965.0) 8 338-353. DOI: 10.1016/S0019-9958(65)90241-X 8. Riza L.S., Bergmeir C.N., Herrera F., Benítez Sánchez J.M.. Frbs: Fuzzy rule-based systems for classification and regression in R. J. Stat. Softw. (2015.0) 65. DOI: 10.18637/jss.v065.i06 9. Zadeh L.A.. The concept of a linguistic variable and its application to approximate reasoning—I. Inf. Sci. (1975.0) 8 199-249. DOI: 10.1016/0020-0255(75)90036-5 10. Hamam A., Georganas N.D.. A comparison of Mamdani and Sugeno fuzzy inference systems for evaluating the quality of experience of Hapto-Audio-Visual applications. Proceedings of the Haptic Audio Visual Environments and Games 87-92 11. Paul A.K., Shill P.C., Rabin M.R.I., Kundu A., Akhand M.A.H.. Fuzzy membership function generation using DMS-PSO for the diagnosis of heart disease. Proceedings of the 2015 18th International Conference on Computer and Information Technology (ICCIT) 456-461 12. Romero-Córdoba R., Olivas J.Á., Romero F.P., Alonso-Gómez F.. Clinical Decision Support System for the Diagnosis and Treatment of Fuzzy Diseases (2015.0) 128-138 13. d’Acierno A., Esposito M., De Pietro G.. An extensible six-step methodology to automatically generate fuzzy DSSs for diagnostic applications. BMC Bioinform. (2013.0) 14. DOI: 10.1186/1471-2105-14-S1-S4 14. Romero-Córdoba R., Olivas J.A., Romero F.P., Alonso-Gonzalez F., Serrano-Guerrero J.. An Application of Fuzzy Prototypes to the Diagnosis and Treatment of Fuzzy Diseases. Int. J. Intell. Syst. (2017.0) 32 194-210. DOI: 10.1002/int.21836 15. Nazari S., Fallah M., Kazemipoor H., Salehipour A.. A fuzzy inference-fuzzy analytic hierarchy process-based clinical decision support system for diagnosis of heart diseases. Expert Syst. Appl. (2018.0) 95 261-271. DOI: 10.1016/j.eswa.2017.11.001 16. Hernández-Julio Y.F., Prieto-Guevara M.J., Nieto-Bernal W., Meriño-Fuentes I., Guerrero-Avendaño A.. Framework for the development of data-driven Mamdani-type fuzzy clinical decision support systems. Diagnostics (2019.0) 9. DOI: 10.3390/diagnostics9020052 17. Hernández-Julio Y.F., Nieto-Bernal W., Muñoz-Hernández H.. Framework for the Development of Data-Driven Mamdani-Type Fuzzy Decision Support Systems Based on Fuzzy set Theory Using Clusters and Pivot Tables (2021.0) Volume 1 18. Bache K., Lichman M.. UCI Machine Learning Repository (2013.0) 19. Mangasarian O.L.. Cancer diagnosis via linear programming. SIAM News (1990.0) 23 1-18 20. Onan A.. A fuzzy-rough nearest neighbor classifier combined with consistency-based subset evaluation and instance selection for automated diagnosis of breast cancer. Expert Syst. Appl. (2015.0) 42 6844-6852. DOI: 10.1016/j.eswa.2015.05.006 21. Aghabozorgi S., Teh Y.W.. Stock market co-movement assessment using a three-phase clustering method. Expert Syst. Appl. (2014.0) 41 1301-1314. DOI: 10.1016/j.eswa.2013.08.028 22. Hernández-Julio Y.F., Yanagi T., de Fátima Ávila Pires M., Lopes M.A., Ribeiro de Lima R.. Models for Prediction of Physiological Responses of Holstein Dairy Cows. Appl. Art. Intell. (2014.0) 28 766-792. DOI: 10.1080/08839514.2014.952919 23. 23. The MathWorks Inc Unique Values in Array, 2017bThe MathWorks Inc.Natick, MA, USA2017. Unique Values in Array, 2017b (2017.0) 24. Tanaka K.. An Introduction to Fuzzy Logic for Practical Applications (1996.0) 25. Sivanandam S., Sumathi S., Deepa S.. Introduction to Fuzzy Logic Using MATLAB (2007.0) Volume 1 26. de Barros L.C., Bassanezi R.C.. Tópicos de Lógica Fuzzy e Biomatemática (2010.0) 27. Alpaydin E.. Introduction to Machine Learning (2009.0) 28. Liu K., Kang G., Zhang N., Hou B.. Breast Cancer Classification Based on Fully-Connected Layer First Convolutional Neural Networks. IEEE Access (2018.0) 6 23722-23732. DOI: 10.1109/ACCESS.2018.2817593 29. Zemouri R., Omri N., Devalland C., Arnould L., Morello B., Zerhouni N., Fnaiech F.. Breast cancer diagnosis based on joint variable selection and Constructive Deep Neural Network. Proceedings of the 2018 IEEE 4th Middle East Conference on Biomedical Engineering (MECBME) 159-164 30. Nguyen T., Khosravi A., Creighton D., Nahavandi S.. Medical data classification using interval type-2 fuzzy logic system and wavelets. Appl. Soft Comput. (2015.0) 30 812-822. DOI: 10.1016/j.asoc.2015.02.016 31. Hernández-Julio Y.F., Hernández H.M., Guzmán J.D.C., Nieto-Bernal W., Díaz R.R.G., Ferraz P.P., Rocha Á., Ferrás C., Paredes M.. Fuzzy Knowledge Discovery and Decision-Making Through Clustering and Dynamic Tables: Application in Medicine. Information Technology and Systems. ICITS 2019. Advances in Intelligent Systems and Computing (2019.0) Volume 918 122-130 32. Abdel-Zaher A.M., Eldeib A.M.. Breast cancer classification using deep belief networks. Expert Syst. Appl. (2016.0) 46 139-144. DOI: 10.1016/j.eswa.2015.10.015 33. Thungrut W., Wattanapongsakorn N., Unger H., Sodsee S., Meesad P.. Diabetes Classification with Fuzzy Genetic Algorithm. Recent Advances in Information and Communication Technology 2018. IC2IT 2018. Advances in Intelligent Systems and Computing (2018.0) Volume 769 107-114 34. Gorunescu F.. Data Mining: Concepts, Models and Techniques (2011.0) Volume 12 XII
--- title: Simulated Microgravity Alters P-Glycoprotein Efflux Function and Expression via the Wnt/β-Catenin Signaling Pathway in Rat Intestine and Brain authors: - Huayan Liu - Min Liang - Yulin Deng - Yujuan Li journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049079 doi: 10.3390/ijms24065438 license: CC BY 4.0 --- # Simulated Microgravity Alters P-Glycoprotein Efflux Function and Expression via the Wnt/β-Catenin Signaling Pathway in Rat Intestine and Brain ## Abstract The drug efflux transporter permeability glycoprotein (P-gp) plays an important role in oral drug absorption and distribution. Under microgravity (MG), the changes in P-gp efflux function may alter the efficacy of oral drugs or lead to unexpected effects. Oral drugs are currently used to protect and treat multisystem physiological damage caused by MG; whether P-gp efflux function changes under MG remains unclear. This study aimed to investigate the alteration of P-gp efflux function, expression, and potential signaling pathway in rats and cells under different simulated MG (SMG) duration. The altered P-gp efflux function was verified by the in vivo intestinal perfusion and the brain distribution of P-gp substrate drugs. Results showed that the efflux function of P-gp was inhibited in the 7 and 21 day SMG-treated rat intestine and brain and 72 h SMG-treated human colon adenocarcinoma cells and human cerebral microvascular endothelial cells. P-gp protein and gene expression levels were continually down-regulated in rat intestine and up-regulated in rat brain by SMG. P-gp expression was regulated by the Wnt/β-catenin signaling pathway under SMG, verified by a pathway-specific agonist and inhibitor. The elevated intestinal absorption and brain distribution of acetaminophen levels also confirmed the inhibited P-gp efflux function in rat intestine and brain under SMG. This study revealed that SMG alters the efflux function of P-gp and regulates the Wnt/β-catenin signaling pathway in the intestine and the brain. These findings may be helpful in guiding the use of P-gp substrate drugs during spaceflight. ## 1. Introduction During spaceflight, astronauts face extremely complex space environments, such as microgravity (MG), radiation, and noise [1]. MG is an adverse factor that is continuous and cannot be separated from spaceflight. Exposure to this condition could lead to cardiovascular dysfunction [2], muscle atrophy [3], bone loss [4], nausea and vomiting [5], and immune [6] and intestinal epithelial barrier function deficiency [7]. Oral drugs are used to prevent or treat the multisystem physiological damage induced by MG to ensure the health and safety of astronauts [8,9]. Several efflux transporters located in intestinal epithelial cells (IECs) and brain microvascular endothelial cells, such as permeability glycoprotein (P-gp), are involved in the absorption and distribution of oral drugs [10,11]. P-gp is a 170 kD transmembrane glycoprotein and a member of the ATP-binding cassette superfamily [12]. It is encoded by the MDR1 gene in humans and homologs mdr1a and mdr1b genes in rodents [13], and its expression is regulated by multiple signaling pathways, including the Wnt/β-catenin signaling pathway [14,15,16]. P-gp is abundantly found in the luminal membrane of the small intestine and blood-brain barrier (BBB) and in the apical membrane of excretory cells, such as hepatocytes and renal proximal tubular epithelium [17]. Owing to its strategic location, P-gp can functionally limit the cellular transport of drugs from the gastrointestinal lumen into enterocytes and from the blood circulation into the brain [18]. Therefore, P-gp plays an important role in the intestinal absorption and brain distribution of its substrate drugs. Several drugs used by astronauts during spaceflight are substrates of P-gp, such as acetaminophen (AP) [19], ciprofloxacin [20], and ibuprofen [21]. At present, astronauts use medications in accordance with terrestrial medical practices; however, whether the absorption or distribution of drugs during spaceflight is the same as on Earth remains unclear [8]. Our previous studies showed that the expression of P-gp decreased in rat ileum mucosa and increased in rat brain under simulated MG (SMG) [22,23]. However, the MG-induced change in the efflux function of P-gp and the potential mechanism have not been elucidated. If the efflux function of P-gp in the intestine and brain is altered under MG, these changes may affect the intestinal absorption and brain distribution of P-gp substrate drugs [24]. Any changes in absorption or distribution or both may lead to the change of efficacy or toxicity of the drug [25]. Therefore, studying the changes in P-gp efflux function and expression in the intestine and brain under MG is important to provide a new perspective for the rational use of P-gp substrate drugs during spaceflight. The Wnt signaling pathway is quite sensitive when in space, and 18 days of spaceflight significantly affected this pathway in melanoma cells [26]. Our previous study also found that the expression of β-catenin was altered in the rat intestinal mucosa and the rat brain under SMG [22,27]. Whether SMG regulates P-gp expression in rat small intestine and rat brain through the Wnt/β-catenin signaling pathway has not been reported. In this study, tail-suspended rat, random 3D rotary cultured human colon adenocarcinoma cell (Caco-2), and human cerebral microvascular endothelial cell (hCMEC/D3) models were used to simulate MG. The effect of different SMG duration on P-gp efflux function and expression in rat intestine and brain was elucidated. Whether SMG regulates P-gp expression in the intestine and brain through the Wnt/β-catenin signaling pathway was also explored and validated using pathway-specific agonist and inhibitor. Furthermore, the intestinal absorption and brain distribution of AP were determined to verify the change in P-gp efflux function. This study gained insights into the effects of SMG on P-gp efflux function and expression, explored its regulation mechanism, and provided scientific support as guidelines for the use of P-gp substrate drugs during spaceflight. ## 2.1.1. SMG Inhibited the Efflux Function and Expression of P-Gp in Rat Intestine The cyclosporine A (CsA) level in the rat plasma was used to reflect the P-gp efflux function in rat intestine [28]. The CsA level in rat plasma was increased by $16.1\%$ and $49.0\%$ after 7 and 21 days of SMG treatment, respectively (Figure 1A). This finding was similar to the results after the oral administration of Verapamil (Ver), a typical inhibitor of P-gp efflux function. These results indicated that 7 and 21 days of SMG treatment significantly inhibited the efflux function of P-gp in rat intestine. Western-Blot analysis indicated that the expression of P-gp decreased by $38.4\%$ and $56.8\%$ after 7 and 21 days of SMG treatment, respectively (Figure 1B–D). The relative expression levels of P-gp were expressed as the ratio of the gray value of the P-gp band to that of the total proteins in the same lane. The total protein gels are shown in Supplementary Figure S1: immunohistochemical (IHC) assay revealed that the P-gp staining area was significantly reduced after 7 and 21 days of SMG treatment (Figure 1E,F), and the expression of P-gp was down-regulated in rat intestine. Quantitative PCR (qPCR) showed that the expression levels of mdr1a and mdr1b in rat intestine were also remarkably down-regulated in 7 and 21 days of SMG-treated rats (Figure 1G). Therefore, 7 and 21 days of SMG treatment time dependently inhibited the efflux function and expression of P-gp in rat intestine. ## 2.1.2. 72 h SMG Treatment Inhibited the Efflux Function and Expression of P-Gp in Caco-2 Cells As reflected by the intracellular rhodamine 123 (Rho123) fluorescence intensity (Figure 2A), 72 h of SMG exposure significantly inhibited the function of P-gp in the Caco-2 cells. The protein expression of P-gp was consistently down-regulated after 12, 24, 48, and 72 h of SMG exposure (Figure 2B,C). The total protein gels are shown in Supplementary Figure S2. Immunofluorescence results showed that 12, 24, 48 and 72 h of SMG treatment time dependently down-regulated the expression of P-gp in the Caco-2 cells (Figure 2D,E). qPCR revealed that the expression of MDR1 was also consistently down-regulated by SMG treatment (Figure 2F). In summary, 72 h of SMG exposure significantly inhibited the efflux function and expression of P-gp in the Caco-2 cells. However, 12, 24, and 48 h of SMG exposure promoted the efflux function of P-gp while inhibiting its expression in Caco-2 cells. ## 2.2.1. SMG Inhibited the Efflux Function and Activated the Expression of P-Gp in Rat Brain The efflux function of P-gp in rat brain was reflected by the ratio of Rho123 concentration in rat brain and blood. After 7 and 21 days of SMG treatment, the ratio of Rho123 concentration in rat brain to that in the blood increased by $40.5\%$ and $173.8\%$, respectively, compared with that in the control (CON) group (Figure 3A). These results indicated that 7 and 21 days of SMG treatment significantly inhibited the efflux function of P-gp in rat brain. Western-Blot results showed that the expression of P-gp increased by $23.4\%$ and $24.2\%$ after 7 and 21 days of SMG treatment, respectively (Figure 3B–D). The total protein gels are shown in Supplementary Figure S1. IHC results also demonstrated the increase in P-gp expression after 7 and 21 days of SMG exposure (Figure 3E,F), and these findings were in accordance with the Western-Blot results. qPCR displayed that mdr1a gene expression increased by $28.5\%$ and $57.5\%$ in the brain of 7 and 21 days SMG-treated rats (Figure 3G). In conclusion, 7 and 21 days of SMG treatment could significantly inhibit the efflux function of P-gp in rat brain while promoting its expression in a time-dependent manner. ## 2.2.2. 48 and 72 h SMG Treatment Inhibited the Efflux Function and Activated the Expression of P-Gp in hCMEC/D3 Cells Figure 4A shows that 48 and 72 h of SMG treatment inhibited the efflux function of P-gp in hCMEC/D3 cells, resulting in a $13.4\%$ and $16.7\%$ increase in the fluorescence intensity of intracellular Rho123, respectively. Western-Blot analysis showed that P-gp expression was up-regulated by approximately $38.3\%$, $16.4\%$, and $81.9\%$ after 24, 48, and 72 h of SMG treatment, respectively (Figure 4B,C). The total protein gels are shown in Supplementary Figure S2. Immunofluorescence results also showed that P-gp expression increased after SMG treatment (Figure 2D,E). Finally, qPCR results showed that MDR1 gene expression level increased after 24 and 48 h of SMG treatment but decreased by $82.1\%$ after 72 h of SMG treatment. In summary, 48 and 72 h of SMG treatment significantly inhibited the efflux function and protein expression of P-gp in hCMEC/D3 cells. However, 24 h of SMG treatment promoted the efflux function of P-gp while inhibiting its protein and gene expression in hCMEC/D3 cells. ## 2.3. SMG Down-Regulated P-Gp Expression by Inhibiting the Wnt/β-Catenin Signaling Pathway in Rat Intestine and Caco-2 Cells The expression levels of Wnt3a, phosphorylated disheveled 2 (pho-Dvl2), glycogen synthase kinase-3β (GSK-3β), pho-GSK-3β, β-catenin, and Dickkopf 1 (DKK1) were determined using Western-Blot. Figure 5A–C shows that after 7 and 21 days of SMG treatment, the expression of Wnt3a decreased without significant difference, and that of pho-Dvl2, pho-GSK-3β/GSK-3β, and β-catenin was dramatically down-regulated. The total protein gels are shown in Supplementary Figure S3. This finding indicated that the Wnt/β-catenin pathway was inhibited by SMG. Meanwhile, the expression of DKK1, an inhibitor of the Wnt/β-catenin pathway, increased significantly. On the basis of the aforementioned results, both 7 and 21 days of SMG treatment could promote DKK1 expression and inhibit the Wnt/β-catenin signaling pathway in rat intestine. Figure 5D,E shows that the expression of Wnt3a and pho-GSK-3β/GSK-3β was down-regulated after 12, 24, 48, and 72 h of SMG treatment, but the difference was significant only after 72 h of SMG treatment. The expression of pho-Dvl2 and β-catenin was down-regulated after SMG treatment, and that of DKK1 showed an increasing trend after 12, 24, 48, and 72 h of SMG treatment. This finding indicated that SMG also promoted DKK1 expression and inhibited the Wnt/β-catenin signaling pathway in Caco-2 cells in a time-dependent manner. To explore whether the SMG-induced change in intestinal P-gp expression is regulated by the Wnt/β-catenin signaling pathway, Caco-2 cells were treated with 2 μM FzM1.8 (an agonist of the Wnt/β-catenin signaling pathway) [29] before 48 h of SMG treatment. Figure 5F,G shows that compared with those in the CON group, the Wnt/β-catenin signaling pathway was activated, and P-gp expression was increased in the Caco-2 cells after 48 h of FzM1.8 treatment. After 48 h of SMG treatment, the Wnt/β-catenin signaling pathway of Caco-2 cells was inhibited, and P-gp expression was significantly decreased. Meanwhile, FzM1.8 treatment could counteract the effect of SMG and up-regulate P-gp expression. In summary, 48 h of SMG treatment might down-regulate the expression of P-gp in Caco-2 cells via the Wnt/β-catenin signaling pathway. ## 2.4. SMG Up-Regulated P-Gp Expression by Activating the Wnt/β-Catenin Signaling Pathway in Rat Brain and hCMEC/D3 Cells The expression levels of Wnt3, pho-Dvl2, GSK-3β, pho-GSK-3β, and β-catenin in rat brain were determined using Western-Blot. The total protein gels are shown in Supplementary Figure S4. Figure 6A–C shows that the expression levels of pho-Dvl2 and pho-GSK-3β/GSK-3β tended to increase without significant difference, and those of Wnt3 and β-catenin were dramatically up-regulated after 7 days of SMG treatment. After 21 days of SMG treatment, the expression levels of Wnt3, pho-Dvl2, pho-GSK-3β/GSK-3β, and β-catenin were all significantly increased, indicating that the Wnt/β-catenin pathway in rat brain was activated by 21 days of SMG treatment. As shown in Figure 6D,E, the expression levels of Wnt3, pho-Dvl2, and β-catenin were significantly decreased after 24 h of SMG treatment. The ratio of pho-GSK-3β to GSK-3β also tended to be down-regulated by 24 h of SMG treatment. Although no obvious change was observed in the expression of Wnt3 under 48 h of SMG treatment, the expression levels of pho-Dvl2, pho-GSK-3β/GSK-3β, and β-catenin were all significantly increased. Furthermore, after 72 h of SMG treatment, the expression of Wnt/β-catenin signaling pathway proteins in hCMEC/D3 cells did not change significantly, except for the decrease in pho-Dvl2 expression. Under 48 h of SMG treatment, the change trend of P-gp expression was consistent with that of the Wnt/β-catenin signaling pathway in the hCMEC/D3 cells. Hence, the hCMEC/D3 cells were treated with 1 nM IWP-O1 (an inhibitor of the Wnt/β-catenin signaling pathway) [30] before 48 h of SMG treatment. Western-Blot assay results showed that compared with those in the CON group, the Wnt/β-catenin signaling pathway was inhibited, and P-gp expression was decreased in the hCMEC/D3 cells after IWP-O1 treatment for 48 h (Figure 6F,G). After 48 h of SMG treatment, the Wnt/β-catenin signaling pathway of hCMEC/D3 cells was activated, and P-gp expression significantly increased. Meanwhile, IWP-O1 treatment could reverse the effect of 48 h SMG treatment and down-regulate P-gp expression. In summary, 48 h of SMG treatment might up-regulate the expression of P-gp in hCMEC/D3 cells via the Wnt/β-catenin signaling pathway. ## 2.5. SMG Promoted the In Vivo Intestinal Absorption and Brain Distribution of AP The absorption of AP in the rat ileum was determined using single-pass intestinal perfusion (SPIP), and its penetration was evaluated according to the amount of luminal disappearance. The effective permeability (Peff) of absorbed solutes in rats correlates well with estimated Peff in humans, and the SPIP model can be used to predict in vivo absorption in humans and evaluate the specific contribution of drug transporters [31,32]. The contents of AP in importer and exporter perfusate were determined using fully validated high-performance liquid chromatography (HPLC)-UV. Retention times for AP and internal standard (IS; ferulic acid) were 7.9 and 10.5 min, and all peaks were well separated (Supplementary Figure S5). The absorption constant (Ka) and Peff of AP were calculated using Equations [1] and [2], and the results are shown in Table 1. After 7 and 21 days of SMG treatment, the Ka of AP in rat ileum increased by $10.3\%$ and $54.8\%$, respectively, and the Peff increased by $14.1\%$ and $74.2\%$, respectively. This finding was similar to the results after the oral administration of Ver. In summary, 7 and 21 days of SMG treatment promoted the absorption of AP in rat intestine. The increased in vivo intestinal absorption of AP also proved that P-gp efflux function in the intestine was inhibited under SMG. The rat brain distribution of AP was reflected by the ratio of the AP concentration in rat brain (μg/g) to that in rat plasma (μg/g) at 2 h post-oral administration of AP. Figure 7 shows that the brain-to-plasma concentration ratio (Kp brain) of AP exhibited an increasing trend after 7 days of SMG treatment and reached approximately $22.0\%$ higher than that in the CON group. Meanwhile, 21 days of SMG treatment significantly increased the brain-to-plasma concentration ratio of AP by $40.7\%$ compared with that in the CON group. The results of AP brain distribution also proved that the P-gp efflux function was inhibited in rat brain after 7 and 21 days of SMG treatment. In summary, the elevated intestinal absorption and brain distribution levels of AP confirmed the inhibited P-gp efflux function in rat intestine and brain under SMG. ## 3. Discussion During spaceflight, drugs are always used to reverse the physiologic insult induced by the complex space environment [33]. To date, these drugs are being administered under the assumption that they act as safely and efficaciously as they do on Earth. However, this assumption has not been systematically investigated [34]. If changes in the pharmacokinetics or pharmacodynamics of drugs taken during spaceflight are not fully considered, drug efficacy and safety will not be guaranteed [8]. The intestinal absorption of oral drugs could be mediated by several efflux transporters, including P-gp. The changes in P-gp efflux function in the IEC membrane might affect the amount of its substrate drugs absorbed via the intestine. The differences in efflux function or expression, or both, of intestinal P-gp potentially induce changes in drug bioavailability, efficacy, and safety [35,36,37]. This study first found that the efflux function and expression of P-gp were significantly inhibited in the 7 and 21 days of SMG-treated rat intestine and the 72 h of SMG-treated Caco-2 cells, implying that the intestinal absorption of P-gp substrate drugs might be promoted under SMG conditions. With prolonged SMG treatment, the efflux function and expression of P-gp in the intestine were significantly inhibited. The aforementioned results were validated by the SPIP analysis of AP, a substrate drug of P-gp [19]. Therefore, 7 and 21 days of SMG treatment could significantly promote AP intestinal absorption. Whether this effect alters the efficacy of AP or potentially causes toxicity warrants further investigation [38]. P-gp at the membranes of cerebral microvascular endothelial cells functions as an active efflux pump by extruding a wide range of substrates from the brain, including most drugs [39]. When central nervous system drugs cross the BBB, P-gp is an important factor limiting the drug delivery to the central nervous system and consequently reducing the efficacy of the drug [40]. Inhibiting the efflux function of P-gp on BBB could be beneficial to some neurological disease drugs to exert their efficacy [41,42,43]. For example, the coadministration of phenytoin and a P-gp inhibitor was significantly more effective in controlling seizures than phenytoin administration alone [44,45]. The efflux function of P-gp in the brain could be reflected by the Kp brain of P-gp substrate [46,47]. In this study, the effect of SMG on P-gp efflux function in rat brain was reflected by the Kp brain of Rho123. The results showed that P-gp efflux function was inhibited in 7 and 21 days of SMG-treated rat brain and 48 h and 72 h of SMG-treated hCMEC/D3 cells. The Kp brain of AP in 7 and 21 days of SMG-treated rats was higher than that in the CON group, revealing the inhibited P-gp efflux function in rat brain. Many central analgesic drugs, including AP, are substrates of P-gp. When P-gp efflux function is inhibited, the efficacy of these drugs may be enhanced [48]. Therefore, the inhibition of P-gp efflux function in the brain under SMG conditions might facilitate central nervous system drugs to cross the BBB and exert their efficacy. Hence, the adverse reactions possibly caused by excessive accumulation of drugs in the brain also warrant research [49,50]. During spaceflight, astronauts face various physiological changes, such as fluid shifts, changes in local blood flow, drug-binding plasma protein levels, and altered gastrointestinal motility. These changes potentially affect drug pharmacodynamics or pharmacokinetics [51]. Some oral drugs taken during spaceflight do not exhibit the expected effect [52,53,54,55]. Studies conducted in orbit have evaluated the salivary pharmacokinetics of AP [56]. The plasma pharmacokinetics of AP under SMG conditions have also been investigated. The results revealed that the absorption of AP in two astronauts increased on mission day 2 compared with that before the flight [57]. Additionally, the absorption of AP was promoted by 21 and 28 days of SMG treatment in rats and 80 days of SMG treatment in humans [38,58]. In the present study, SPIP analysis in rats showed that 7 and 21 days of SMG treatment could significantly promote AP intestinal absorption in vivo. The increased intestinal absorption of AP might be due to the inhibition of intestinal P-gp efflux function under SMG. Additionally, the absorption of P-gp oral drug substrates ibuprofen, Ver, propranolol, and promethazine in humans was promoted under SMG [59,60,61]. This finding may also result from the inhibited P-gp efflux function in the intestine under SMG. The present study suggested that the changes of P-gp efflux in the intestine and brain under a microgravity environment may lead to the difference in intestinal absorption and brain distribution of P-gp substrate oral drugs compared with that on the ground. This indicates that the dosing regimen should be carefully considered when these P-gp substrate drugs are taken by astronauts, including administered dose, frequency, administration routes, or potential drug-drug interactions. The optimization of drug administration thus requires further in-orbit validation. The Wnt/β-catenin signaling pathway can regulate the expression of P-gp [14,15,16]. The expression of REKEN cDNA 2210419D22 in melanoma cells in space was significantly different from that in Earth, and this phenomenon was related to the Wnt/β-catenin signaling pathway [26]. The canonical Wnt/β-catenin signaling pathway is also involved in the control of the response to SMG in nematode Caenorhabditis elegans [62]. Our previous study found that the expression of β-catenin was altered in rat intestinal mucosa and brain under SMG [22,27]. Therefore, the present study further investigated whether SMG regulates P-gp expression in the brain and intestine through the Wnt/β-catenin signaling pathway. The composition and regulation of the Wnt/β-catenin signaling pathway were as follows. In the absence of Wnt, β-catenin in the cytoplasm is phosphorylated by a degradation complex consisting of axis inhibitor (Axin), adenomatous polyposis coli protein (APC), casein kinase 1 (CK1), and GSK-3β, and is then further ubiquitinated and degraded by the ubiquitin–proteasome system [63]. Once Wnt binds to Frizzled (Fzd) and lipoprotein receptor-related proteins $\frac{5}{6}$ (LRP$\frac{5}{6}$), the cytoplasmic Dvl is phosphorylated [64], the β-catenin degradation complex is disassembled, and GSK-3β is phosphorylated and binds to LRP$\frac{5}{6}$ [65,66], hence impeding β-catenin degradation. β-Catenin accumulates in the cytoplasm, enters the nucleus, and binds to transcription factor (TCF) to promote the expression of target genes, including those encoding P-gp [67,68,69]. GSK-3β phosphorylation, can also activate the Wnt/β-catenin signaling pathway [70,71]. This study found that the levels of Wnt3a, pho-Dvl2, pho-GSK-3β/GSK-3β, β-catenin, and P-gp in rat ileum were down-regulated by 7 and 21 days of SMG treatment. The Wnt/β-catenin signaling pathway was also inhibited by SMG treatment in a time-dependent manner. For further investigation on whether SMG regulates the expression of P-gp in the intestine via the Wnt/β-catenin signaling pathway, 48 h of SMG treatment was selected, and the pathway-specific agonist FzM1.8 was added for verification. The results showed that the addition of FzM1.8 could counteract the inhibitory effect of 48 h of SMG treatment on the Wnt/β-catenin signaling pathway and increase the expression of P-gp. Therefore, SMG may down-regulate P-gp expression by inhibiting the Wnt/β-catenin signaling pathway. The Wnt/β-catenin signaling pathway was significantly activated in the brain of 7 and 21 days of SMG-treated rats and in 48 h of SMG-treated hCMEC/D3 cells. The addition of exclusive inhibitor IWP-O1 could attenuate the activation effect of 48 h of SMG treatment on the Wnt/β-catenin signaling pathway and decrease the expression of P-gp. This finding suggested that SMG might promote the Wnt/β-catenin signaling pathway to increase P-gp expression in the brain. However, after 24 and 72 h of SMG treatment, the change trend of P-gp expression in hCMEC/D3 cells was not consistent with that of the Wnt/β-catenin signaling pathway. In addition to the Wnt/β-catenin signaling pathway, various signaling pathways regulate P-gp expression, including protein kinase C [72], nuclear factor κB [73], mitogen-activated protein kinase [74], and PI3K/AKT signaling pathways [75]. Therefore, 24 and 72 h of SMG treatment may regulate the expression of P-gp in hCMEC/D3 cells through other pathways besides the Wnt/β-catenin signaling pathway; this finding warrants further research. ## 4.1. Animal Treatment and Sample Collection Sprague–Dawley male, specific pathogen-free rats (200 ± 20 g) were obtained from the Institute of Laboratory Animal Science (Beijing, China). All animal experiments were approved by the Beijing Institute of Technology Animal Care and Use Committee (Protocol No-SYXK-BIT-20200109002, Beijing, China) and complied with the Guide for the Care and Use of Laboratory Animals (NIH publication No. 85–23, revised in 1996). All rats were housed in a controlled environment with a maintained 24 °C ± 1 °C and $55\%$ ± $5\%$ relative humidity and had free access to food and water. After 7 days acclimatization, the rats were divided into four groups (the $$n = 6$$ per group): 7 days-CON, 7 days-SMG, 21 days-CON, and 21 days-SMG. The rats in SMG groups were subjected to tail suspension and 30° head-down for 7 days or 21 days according to the Morey–Holton model [76]. At the end of 7 days or 21 days of SMG treatment, all rats were anesthetized via intraperitoneal injection with chloral hydrate (350 mg/kg), and the ileum was removed. Approximately 1 cm ileum was fixed in $4\%$ paraformaldehyde (Solarbio, Beijing, China) over 24 h for IHC staining, and the remaining ileum was used to collect the mucosa. The rat brain was divided into two parts. The left brain was fixed in $4\%$ paraformaldehyde for IHC staining. The right brain and ileum mucosa were stored at −80 °C for further experiments. ## 4.2. Efflux Function Analysis of P-Gp in Rat Intestine and Brain After 7 days or 21 days of tail suspension, the CON rats were divided into the CON group and CON + Ver group ($$n = 6$$ per group). Ver (30 mg/kg; Solarbio, Beijing, China) was orally administered to the rats in the CON + Ver group, and CsA (15 mg/kg, TCI CO., LTD., Tokyo, Japan) was concurrently orally administered to all rats. After 3.5 h, all rats were anesthetized with chloral hydrate (350 mg/kg), and the plasma was sampled from the heart. The content of CsA in the rat plasma was determined using an ELISA kit following the kit instructions (Yuanmu Biotech Co., Ltd., Shanghai, China). For the analysis of P-gp efflux function in rat brain, the rats in the CON + Ver group were injected with Ver (1 mg/kg) through the tail vein after being tail suspended for 7 days or 21 days. After 45 min, the tail veins of all rats were injected with Rho123 (0.5 mg/kg, Solarbio, Beijing, China). At 15 min post-injection, all rats were anesthetized with chloral hydrate (350 mg/kg), and the blood was collected from the heart. After being subjected to transcardiac perfusion using the stroke-physiological saline solution from the heart, the brain samples of all rats were collected. In summary, 0.1 g of brain samples were homogenized with 0.4 mL of normal saline to obtain brain homogenate. Blank plasma and brain samples were collected from rats without a Rho123 injection and used to prepare Rho123 standard curve samples. The concentration of Rho123 in plasma and brain homogenate from all rats was detected by fluorescence intensity (λ excitation/λ emission = $\frac{485}{535}$) using a microplate reader (BioTek, Thorold, ON, Canada). The Kp brain of Rho123 was used to evaluate P-gp efflux function. ## 4.3. Immunohistochemical Staining The rat ileum segments and brain fixed in $4\%$ paraformaldehyde were embedded in paraffin and sectioned following routine procedures [77]. Slides were blocked in 1× Tris-buffered saline containing $3\%$ bovine serum albumin (BSA), $2\%$ serum, and $0.02\%$ Tween 20 at room temperature for 30 min. The sections were incubated with the primary antibody against P-gp (1:400 for rat ileum and 1:100 for rat brain, Abcam, Cambridge, UK) overnight at 4 °C, followed by incubation with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (ZSGB-Bio, Beijing, China) at room temperature for 2 h. The sections were then visualized using a diaminobenzidine solution. Between each of the aforementioned steps, the sections were washed using 1× Tris-buffered saline tween (TBST) three times for 5 min each. Finally, images were captured using a Nanozoomer S210 microscopic-resolution scanner equipped with Digital Pathology View 2.0 software (Hamamatsu Photonics, Shizuoka, Japan). Java-based image-processing and analysis software (Image-Pro Plus; Version 6.0.0.260; National Institutes of Health, Bethesda, MD, USA) was used to analyze the proportion of P-gp-stained area in the whole ileal or cerebral cortex section of each rat in each group, respectively. ## 4.4. Cell Culture and SMG Treatment Caco-2 cells were purchased from the Chinese Academy of Sciences (Shanghai, China) and maintained in MEM/EBSS medium (Hyclone, UT, USA) supplemented with $10\%$ fetal bovine serum (Sofar, Beijing, China), $1\%$ penicillin–streptomycin liquid (Solarbio, Beijing, China), and $1\%$ nonessential amino acids (Gibco, New York, NY, USA) in a humidified $5\%$ CO2 atmosphere at 37 °C. The Caco-2 cells were seeded in a T-12.5 flask (2 × 104 cells/cm2) and cultured for 21 days under normal gravity. The T-12.5 flask was filled with culture medium while preventing bubble formation and then exposed to SMG using random speed 3D clinostat (National Space Science Center, Beijing, China) [78] for 12, 24, 48, and 72 h. The Caco-2 cells in the CON group were cultured under normal gravity in the same CO2 incubator. The hCMEC/D3 cells were purchased from iCell Bioscience Inc. (Shanghai, China) and cultured in RPMI 1640 medium containing $1\%$ penicillin–streptomycin liquid (Solarbio, Beijing, China) and $10\%$ fetal bovine serum (Tianhang, Zhejiang, China) in the same environment as that of Caco-2 cells. The hCMEC/D3 cells were seeded at a density of 6 × 104 cells/cm2 and cultured for 2 days under normal gravity. The flasks were then filled with culture medium and exposed to SMG using random speed 3D clinostat for 24, 48, and 72 h. The hCMEC/D3 cells in the CON group were cultured under normal gravity in the same CO2 incubator. FzM1.8 and IWP-O1 (Wnt/β-catenin signaling pathway-specific agonist and inhibitor, respectively) were dissolved in dimethyl sulfoxide (DMSO) and diluted with culture medium to obtain the desired concentrations. The final concentration of DMSO was $0.1\%$. FzM1.8 (2 μM) was added to the Caco-2 cells in the CON + FzM1.8 and SMG + FzM1.8 groups. IWP-O1 (1 nM) was added to the hCMEC/D3 cells in the CON + IWPO-1 and SMG + IWPO-1 groups. The cells in the SMG, SMG + FzM1.8, or SMG + IWPO-1 group were then exposed to SMG for 48 h. ## 4.5. Efflux Function Analysis of P-Gp in Caco-2 and hCMEC/D3 Cells After 12, 24, 48, or 72 h of SMG treatment, the Caco-2 cells were treated with 2 mL of medium containing Rho123 (2 μM) for 1 h. The cells were then washed with phosphate buffer saline (PBS) solution five times (2 mL each time) and treated with 2 mL of $0.1\%$ Triton X-100 for 15 min. The cell solution was transferred to a black 96-well plate (Corning, NY, USA) with six wells in each group and 200 μL in each well after brief centrifugation. The fluorescence intensity of each well was detected using a microplate reader (BioTek, Thorold, ON, Canada). The excitation/emission wavelength of Rho123 was $\frac{485}{535}$ nm. After 24, 48, or 72 h of SMG treatment, the hCMEC/D3 cells were treated with 2 mL of medium containing Rho123 (2 μM) for 0.5 h. Subsequent operations were performed in the same way as for Caco-2 cells. ## 4.6. Western-Blot Total proteins in the rat ileum mucosa and brain and Caco-2 and hCMEC/D3 cells were extracted using radio immunoprecipitation assay buffer containing protease inhibitors and protein phosphatase inhibitors. The supernatant was collected after centrifugation (12,000× g, 4 °C, and 10 min). The protein concentration was tested using Coomassie brilliant blue staining. The supernatant was mixed with 4× protein loading buffer containing dithiothreitol (Solarbio, Beijing, China) and desaturated in a boiling water bath for 10 min. The total protein (60 μg for ileum and brain and 30 μg for Caco-2 and hCMEC/D3 cells) of the samples was loaded in each well and separated by $12\%$ sodium dodecyl sulfate–polyacrylamide gel electrophoresis using a TGX Stain-freeTM FastCastTM Acrylamide Kit (Bio-Rad, Hercules, CA, USA). The total protein on the gel was imaged under ChemiDoc XRS+ mode of Image Lab software (version 3.0; Bio-Rad, Hercules, CA, USA) and then transferred to polyvinylidene fluoride membranes (Millipore, Burlington, MA, USA). The membranes were blocked with $5\%$ non-fat milk or BSA for 2 h and incubated overnight at 4 °C with corresponding primary antibodies as follows: anti-P-gp (1:5000), anti-Wnt3a (1:1000), anti-pho-Dvl2 (1:5000), anti-β-catenin (1:5000), anti-DKK1 (1:500, Abcam, Cambridge, UK), anti-GSK-3β (1:2000), anti-pho-GSK-3β (1:2000, ZSGB-Bio, Beijing, China), and anti-Wnt3 (1:2000, Solarbio, Beijing, China). The membranes were then incubated with HRP-conjugated secondary antibody (ZSGB-Bio, Beijing, China) for 2 h at room temperature. Finally, the bands were visualized using an enhanced chemiluminescence reagent (Millipore, MA, USA). The gray value of each band was collected using Image LabTM Software (version 3.0, Bio-Rad, Hercules, CA, USA), and the relative expression levels of proteins were expressed as the ratio of the gray value of the target band to the total proteins in the same lane. ## 4.7. qPCR Total RNA in the rat ileum mucosa and brain and Caco-2 and hCMEC/D3 cells was extracted using a total RNA extraction kit (Solarbio, Beijing, China) following the manufacturer’s instructions. The concentration and purity of RNA were determined by comparing the absorbance at 260 and 280 nm. In summary, 5 μg of total RNA was reverse transcribed to cDNA using a universal RT-PCR kit (M-MLV; Solarbio, Beijing, China). The relative expression level of P-gp mRNA to that of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was determined by qPCR using TB Green™ Premix Ex Taq™ II (TaKaRa, Tokyo, Japan) in accordance with the manufacturer’s instructions. The forward and reverse primers are shown in Table 2. ## 4.8. Immunofluorescence The Caco-2 cells were cultured in a confocal plate for 10 days. The confocal plate was then filled with culture medium while ensuring no air bubbles had formed, and was then exposed to SMG using 3D clinostat for 12, 24, 48, and 72 h. The hCMEC/D3 cells were cultured on a four-well Lab-Tek II CC2 chamber slide overnight. The chamber slides were then filled with culture medium, sealed with parafilm while ensuring no air bubbles had formed, and exposed to SMG using 3D clinostat for 24, 48, or 72 h. After SMG treatment, the cells were washed with PBS three times for 5 min each, fixed with $4\%$ paraformaldehyde for 30 min, and washed with PBS again three times. The cells were permeabilized with $0.5\%$ Triton X-100 (Solarbio, Beijing, China) for 10 min, washed three times with PBS, blocked by $5\%$ BSA for 2 h at room temperature, and incubated with rabbit anti-P-gp antibody (1:200) overnight at 4 °C. After being washed with PBS, the cells were incubated with Rho-conjugated goat anti-rabbit IgG (ZSGB-Bio, Beijing, China) for 2 h at room temperature and washed again. 0.1 μg/mL 4′,6-diamidino-2-phenylindole (Solarbio, Beijing, China) was then used to stain the nuclei for 15 min. After repeated washes in the same manner, the confocal plate or chamber slide was blow-dried and covered with an antifade solution (Solarbio, Beijing, China). Images were captured using Nikon N-SIM Confocal (Nikon, Tokyo, Japan) and NIS element imaging software (version 4.50). The results were analyzed using ImageJ, and the average fluorescence intensity on random lines was used to define P-gp expression in the cells. ## 4.9. Rat Single-Pass Intestinal Perfusion of Acetaminophen Ver (30 mg/kg) was orally administered to the rats in the CON + Ver group. At 1 h post-oral administration, all rats were anesthetized with urethane (1.75 g/kg) and placed in a supine position. Approximately 10 cm of ileum was separated from the intestinal segments. Tubes were carefully inserted at both ends of this segment and ligated with a sterile surgical line. The segment was rinsed with 37 °C Krebs–Ringer’s (KR) solution, balanced with KR solution (containing 2.5 mg/mL AP) at a flow rate of 0.2 mL/min for 15 min, and continuously perfused with KR solution (containing 2.5 mg/mL AP). The flow rate was set as 0.2 mL/min. The outflow perfusate was collected in a pre-weighed tube, which was changed quickly every 15 min and weighed. Six tubes of this fluid were collected for each rat. The perfusion solution and outflow perfusate were diluted in methanol–water (1:9, v:v) and analyzed using HPLC. The absorption constant (Ka) and effective permeability (Peff) were calculated using the following equations:[1]Ka=Q1−CoutVout/CinVinπr2L [2]Peff=−QlnCoutVout/CinVin2πrL where Q is the perfusion flow rate; Cin and Cout are the inlet and outlet concentrations, respectively; Vin and Vout are the volumes of the importer and exporter perfusate, respectively; r is the internal diameter of the perfused intestinal segment ($r = 0.18$ cm) [79]; and L is the length of the perfused intestinal segment. ## 4.10. Rat Brain Distribution of Acetaminophen After 7 or 21 days of tail suspension, all rats were orally administered AP (1.2 g/kg). At 1 h post-oral administration, all rats were anesthetized via intraperitoneal injection with chloral hydrate (350 mg/kg), and rat blood was taken from the heart. Rat brains were rapidly removed after perfusion with prechilled saline from rat hearts. Then, 0.1 mL of plasma from each rat was used to precipitate proteins with 300 μL of methanol. Approximately 0.1 g of the brain tissue of each rat was homogenized and ultrasonicated in 1 mL of methanol. The supernatant was removed and dried via vacuum centrifugation and reconstituted with methanol for HPLC. The brain distribution of AP was evaluated as the ratio of AP content in brain tissue (μg/g) to that in plasma (μg/mL). ## 4.11. Acetaminophen Determination Using the HPLC-UV Method HPLC was performed on a Shimadzu LC-20AT HPLC system equipped with a UV detector (Shimadzu, Kyoto, Japan). Sample separation was conducted in a C18 column (5 μm, 4.6 × 150 mm). The mobile phases were methanol (A) and water (B) (1:9, v:v), and the flow rate was 1 mL/min. The volume of each sample injected into the HPLC system was 10 μL. The column temperature was maintained at 25 °C, and the detection wavelength was 275 nm. Ferulic acid (final concentration of 100 μg/mL) was used as an IS. ## 4.12. Statistical Analysis Statistical analysis was performed using SPSS 20.0 software (IBM, New York, NY, USA), and the results were expressed as mean ± SD. The difference between the two groups was determined by one-way ANOVA, and a p-value less than 0.05 was considered to have statistical significance. ## 5. Conclusions This study first elucidated that P-gp efflux function in rat intestine and brain was down-regulated by SMG. The efflux function of P-gp in Caco-2 and hCMEC/D3 cells was also inhibited by 72 h SMG treatment. Under SMG, P-gp expression decreased in rat intestine and Caco-2 cells and increased in rat brain and hCMEC/D3 cells. SMG might regulate P-gp expression in the intestine and brain via the Wnt/β-catenin signaling pathway as verified using the pathway-specific agonist in Caco-2 cells and inhibitor in hCMEC/D3 cells. The increased intestinal absorption and brain distribution levels of the P-gp substrate AP also confirmed the inhibition of P-gp function in the intestine and brain under SMG. These results might be helpful in understanding the effects and mechanism of SMG on P-gp function and expression in the intestine and brain, and provide scientific support as guidelines for the use of P-gp substrate drugs during spaceflight. ## References 1. Kandarpa K., Schneider V., Ganapathy K.. Human health during space travel: An overview. Neurol. India (2019) 67 S176-S181. PMID: 31134907 2. Shen M., Frishman W.H.. Effects of spaceflight on cardiovascular physiology and health. Cardiol. Rev. (2019) 27 122-126. DOI: 10.1097/CRD.0000000000000236 3. Zhang S., Yuan M., Cheng C., Xia D.H., Wu S.W.. Chinese herbal medicine effects on muscle atrophy induced by simulated microgravity. Aerosp. Med. Hum. Perform. (2018) 89 883-888. DOI: 10.3357/AMHP.5079.2018 4. Nagaraja M.P., Risin D.. The current state of bone loss research: Data from spaceflight and microgravity simulators. J. Cell. Biochem. (2013) 114 1001-1008. DOI: 10.1002/jcb.24454 5. Russomano T., Da Rosa M., Dos Santos M.A.. Space motion sickness: A common neurovestibular dysfunction in microgravity. Neurol. India (2019) 67 214. DOI: 10.4103/0028-3886.259127 6. Borchers A.T., Keen C.L., Gershwin M.E.. Microgravity and immune responsiveness: Implications for space travel. Nutrition (2002) 18 889-898. DOI: 10.1016/S0899-9007(02)00913-9 7. Alvarez R., Stork C.A., Sayoc-Becerra A., Marchelletta R.R., Prisk G.K., Mccole D.F.. A simulated microgravity environment causes a sustained defect in epithelial barrier function. Sci. Rep. (2019) 9 17531. DOI: 10.1038/s41598-019-53862-3 8. Kast J., Yu Y., Seubert C.N., Wotring V.E., Derendorf H.. Drugs in space: Pharmacokinetics and pharmacodynamics in astronauts. Eur. J. Pharm. Sci. (2017) 109 S2-S8. DOI: 10.1016/j.ejps.2017.05.025 9. Stepanek J., Blue R.S., Parazynski S.. Space medicine in the era of civilian spaceflight. N. Engl. J. Med. (2019) 380 1053-1060. DOI: 10.1056/NEJMra1609012 10. Choi Y.H., Yu A.M.. ABC transporters in multidrug resistance and pharmacokinetics, and strategies for drug development. Curr. Pharm. Des. (2014) 20 793-807. DOI: 10.2174/138161282005140214165212 11. Vilar S., Sobarzo-Sánchez E., Uriarte E.. In silico prediction of P-glycoprotein binding: Insights from molecular docking studies. Curr. Med. Chem. (2019) 26 1746-1760. DOI: 10.2174/0929867325666171129121924 12. Lavie Y., Fiuccii G., Liscovitch M.. Upregulation of caveolae in multidrug resistant cancer cells: Functional implications. Adv. Drug Deliv. Rev. (2001) 49 317-323. DOI: 10.1016/S0169-409X(01)00144-2 13. Fernandez C., Buyse M., German-Fattal M., Gimenez F.. Influence of the pro-inflammatory cytokines on P-glycoprotein expression and functionality. J. Pharm. Pharm. Sci. (2004) 7 359-371. PMID: 15576018 14. Shen D.Y., Zhang W., Zeng X., Liu C.Q.. Inhibition of Wnt/β-catenin signaling downregulates P-glycoprotein and reverses multi-drug resistance of cholangiocarcinoma. Cancer Sci. (2013) 104 1303-1308. DOI: 10.1111/cas.12223 15. Wijesuriya H.C., Bullock J.Y., Faull R.L., Hladky S.B., Barrand M.A.. ABC efflux transporters in brain vasculature of Alzheimer’s subjects. Brain Res. (2010) 1358 228-238. DOI: 10.1016/j.brainres.2010.08.034 16. Kim D.Y., Jung S.Y., Kim K., Kim C.J.. Treadmill exercise ameliorates Alzheimer disease-associated memory loss through the Wnt signaling pathway in the streptozotocin-induced diabetic rats. J. Exerc. Rehabil. (2016) 12 276-283. DOI: 10.12965/jer.1632678.339 17. Wessler J.D., Grip L.T., Mendell J., Giugliano R.P.. The P-glycoprotein transport system and cardiovascular drugs. J. Am. Coll. Cardiol. (2013) 61 2495-2502. DOI: 10.1016/j.jacc.2013.02.058 18. Lin J.H., Yamazaki M.. Role of P-glycoprotein in pharmacokinetics: Clinical implications. Clin. Pharmacokinet. (2003) 42 59-98. DOI: 10.2165/00003088-200342010-00003 19. Arellano C., Philibert C., Vachoux C., Woodley J., Houin G.. The metabolism of midazolam and comparison with other CYP enzyme substrates during intestinal absorption: In vitro studies with rat everted gut sacs. J. Pharm. Pharm. Sci. (2007) 10 26-36. PMID: 17498391 20. Zhang Y., Huang J., Liu Y., Guo T., Wang L.. Using the lentiviral vector system to stably express chicken P-gp and BCRP in MDCK cells for screening the substrates and studying the interplay of both transporters. Arch. Toxicol. (2018) 92 2027-2042. DOI: 10.1007/s00204-018-2209-9 21. Geng J., Luan L.B.. Transport characteristics of ibuprofen in Caco-2 cell monolayers. Chin. J. Pharm. (2010) 41 583-586 22. Li Y.J., Liu S., Liu H.Y., Cui Y.Y., Deng Y.L.. Dragon’s blood regulates Rac1-WAVE2-Arp2/3 signaling pathway to protect rat intestinal epithelial barrier dysfunction induced by simulated microgravity. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22052722 23. Li Y.J., Huang L.L., Iqbal J., Deng Y.L.. Investigation on P-glycoprotein function and its interacting proteins under simulated microgravity. Space Sci. Technol. (2021) 2021 9835728. DOI: 10.34133/2021/9835728 24. Lu S.K., Bai S., Javeri K., Brunner L.J.. Altered cytochrome P450 and P-glycoprotein levels in rats during simulated weightlessness. Aviat. Space Environ. Med. (2002) 73 112-118. PMID: 11846179 25. Wagner J., Abdel-Rahman S.M.. Pediatric pharmacokinetics. Pediatr. Rev. (2013) 34 258-269. DOI: 10.1542/pir.34.6.258 26. Xiang Q., Xiao C., Xu M., Xu B., Li H.Y., Geng C.Y., Fang Q., Tang J.T.. Effects of space environment on gene expression profile of melanoma cells. Chin. J. Aerospace. Med. (2006) 17 165-168 27. Yan R.R., Liu H.Y., Lv F., Deng Y.L., Li Y.J.. Rac1/Wave2/Arp3 Pathway Mediates Rat Blood-Brain Barrier Dysfunction under Simulated Microgravity Based on Proteomics Strategy. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22105165 28. Liu X.L., Tang J., Song J., He J., Xu P., Peng W.X.. The effect of tetramethylpyrazine on the pharmacokinetics of intragastrically administered cyclosporine A in rats. Acta Pharm. Sin. (2006) 41 882-887 29. Riccio G., Bottone S., La Regina G., Badolati N., Passacantilli S., Rossi G.B., Accardo A., Dentice M., Silvestri R., Novellino E.. A negative allosteric modulator of WNT receptor Frizzled 4 switches into an allosteric agonist. Biochemistry (2018) 57 839-851. DOI: 10.1021/acs.biochem.7b01087 30. You L., Zhang C., Yarravarapu N., Morlock L., Wang X., Zhang L., Williams N.S., Lum L., Chen C.. Development of a triazole class of highly potent Porcn inhibitors. Bioorg. Med. Chem. Lett. (2016) 26 5891-5895. DOI: 10.1016/j.bmcl.2016.11.012 31. Strindberg S., Plum J., Stie M.B., Christiansen M.L., Hagner Nielsen L., Rades T., Müllertz A.. Effect of supersaturation on absorption of indomethacin and tadalafil in a single pass intestinal perfusion rat model, in the absence and presence of a precipitation inhibitor. Eur. J. Pharm. Biopharm. (2020) 151 108-115. DOI: 10.1016/j.ejpb.2020.03.019 32. Stappaerts J., Brouwers J., Annaert P., Augustijns P.. In situ perfusion in rodents to explore intestinal drug absorption: Challenges and opportunities. Int. J. Pharm. (2015) 478 665-681. DOI: 10.1016/j.ijpharm.2014.11.035 33. Komorowski M., Fleming S., Kirkpatrick A.W.. Fundamentals of Anesthesiology for Spaceflight. J. Cardiothorac. Vasc. Anesth. (2016) 30 781-790. DOI: 10.1053/j.jvca.2016.01.007 34. Dello Russo C., Bandiera T., Monici M., Surdo L., Yip V.L.M., Wotring V.E., Morbidelli L.. Physiological adaptations affecting drug pharmacokinetics in space: What do we really know? A critical review of the literature. Br. J. Pharmacol. (2022) 179 2538-2557. DOI: 10.1111/bph.15822 35. Lown K.S., Mayo R.R., Leichtman A.B., Hsiao H.L., Turgeon D.K., Schmiedlin-Ren P., Brown M.B., Guo W., Rossi S.J., Benet L.Z.. Role of intestinal P-glycoprotein (mdr1) in interpatient variation in the oral bioavailability of cyclosporine. Clin. Pharmacol. Ther. (1997) 62 248-260. DOI: 10.1016/S0009-9236(97)90027-8 36. Hashida T., Masuda S., Uemoto S., Saito H., Tanaka K., Inui K.. Pharmacokinetic and prognostic significance of intestinal MDR1 expression in recipients of living-donor liver transplantation. Clin. Pharmacol. Ther. (2001) 69 308-316. DOI: 10.1067/mcp.2001.115142 37. Terada T., Hira D.. Intestinal and hepatic drug transporters: Pharmacokinetic, pathophysiological, and pharmacogenetic roles. J. Gastroenterol. (2015) 50 508-519. DOI: 10.1007/s00535-015-1061-4 38. Zhao J.. Study on the Pharmacokinetics of Vitamin B6 and Acetaminophen in Simulated Weightlessness SD Rats. Master’s Thesis (2018) 39. Bartels A.L., de Klerk O.L., Kortekaas R., de Vries J.J., Leenders K.L.. 11C-verapamil to assess P-gp function in human brain during aging, depression and neurodegenerative disease. Curr. Top. Med. Chem. (2010) 10 1775-1784. DOI: 10.2174/156802610792928059 40. Ding Y., Wang R., Zhang J., Zhao A., Lu H., Li W., Wang C., Yuan X.. Potential Regulation Mechanisms of P-gp in the Blood-Brain Barrier in Hypoxia. Curr. Pharm. Des. (2019) 25 1041-1051. DOI: 10.2174/1381612825666190610140153 41. Ejsing T.B., Linnet K.. Influence of P-glycoprotein inhibition on the distribution of the tricyclic antidepressant nortriptyline over the blood-brain barrier. Hum. Psychopharmacol. Clin. Exp. (2005) 20 149-153. DOI: 10.1002/hup.667 42. Marier J.F., Deschênes J.L., Hage A., Seliniotakis E., Gritsas A., Flarakos T., Beaudry F., Vachon P.. Enhancing the uptake of dextromethorphan in the CNS of rats by concomitant administration of the P-gp inhibitor verapamil. Life Sci. (2005) 77 2911-2926. DOI: 10.1016/j.lfs.2005.04.025 43. Liu X., Yang Z., Yang J., Yang H.. Increased P-glycoprotein expression and decreased phenobarbital distribution in the brain of pentylenetetrazole-kindled rats. Neuropharmacology (2007) 53 657-663. DOI: 10.1016/j.neuropharm.2007.07.012 44. van Vliet E.A., van Schaik R., Edelbroek P.M., Voskuyl R.A., Redeker S., Aronica E., Wadman W.J., Gorter J.A.. Region-specific overexpression of P-glycoprotein at the blood-brain barrier affects brain uptake of phenytoin in epileptic rats. J. Pharmacol. Exp. Ther. (2007) 322 141-147. DOI: 10.1124/jpet.107.121178 45. van Vliet E.A., van Schaik R., Edelbroek P.M., Redeker S., Aronica E., Wadman W.J., Marchi N., Vezzani A., Gorter J.A.. Inhibition of the multidrug transporter P-glycoprotein improves seizure control in phenytoin-treated chronic epileptic rats. Epilepsia (2006) 47 672-680. DOI: 10.1111/j.1528-1167.2006.00496.x 46. Watari R., Matsuda A., Ohnishi S., Hasegawa H.. Minimal contribution of P-gp on the low brain distribution of naldemedine, a peripherally acting μ-opioid receptor antagonist. Drug Metab. Pharmacokinet. (2019) 34 126-133. DOI: 10.1016/j.dmpk.2018.12.002 47. Uchida Y., Ohtsuki S., Kamiie J., Terasaki T.. Blood-brain barrier (BBB) pharmacoproteomics: Reconstruction of in vivo brain distribution of 11 P-glycoprotein substrates based on the BBB transporter protein concentration, in vitro intrinsic transport activity, and unbound fraction in plasma and brain in mice. J. Pharmacol. Exp. Ther. (2011) 339 579-588. PMID: 21828264 48. Shen Y., Li H.D.. P-glycoprotein mediated drug-drug interaction of central nervous system drugs. Chin. J. Clin. Pharmacol. (2008) 24 533-538 49. Tsujikawa K., Dan Y., Nogawa K., Sato H., Yamada Y., Murakami H., Ohtani H., Sawada Y., Iga T.. Potentiation of domperidone-induced catalepsy by a P-glycoprotein inhibitor, cyclosporin A. Biopharm. Drug Dispos. (2003) 24 105-114. DOI: 10.1002/bdd.343 50. Dan Y., Murakami H., Koyabu N., Ohtani H., Sawada Y.. Distribution of domperidone into the rat brain is increased by brain ischaemia or treatment with the P-glycoprotein inhibitor verapamil. J. Pharm. Pharmacol. (2002) 54 729-733. DOI: 10.1211/0022357021778880 51. Eyal S., Derendorf H.. Medications in Space: In Search of a Pharmacologist’s Guide to the Galaxy. Pharm. Res. (2019) 36 148. DOI: 10.1007/s11095-019-2679-3 52. Pool S.L., Nicogossian A.. Biomedical results of the Space Shuttle orbital flight test program. Aviat. Space Environ. Med. (1983) 54 S41-S49. PMID: 6661133 53. Caiani E.G., Martin-Yebra A., Landreani F., Bolea J., Laguna P., Vaïda P.. Weightlessness and cardiac rhythm disorders: Current knowledge from space flight and bed-rest studies. Front. Astron. Space Sci. (2016) 3 27. DOI: 10.3389/fspas.2016.00027 54. Wotring V.E.. Medication use by U.S. crewmembers on the International Space Station. FASEB J. (2015) 29 4417-4423. DOI: 10.1096/fj.14-264838 55. Wotring V.E., Smith L.K.. Dose Tracker Application for Collecting Medication Use Data from International Space Station Crew. Aerosp. Med. Hum. Perform. (2020) 91 41-45. DOI: 10.3357/AMHP.5392.2020 56. Eyal S.. How do the pharmacokinetics of drugs change in astronauts in space?. Expert. Opin. Drug Metab. Toxicol. (2020) 16 353-356. DOI: 10.1080/17425255.2020.1746763 57. Cintron N.M., Putcha L., Chen Y.M., Vanderploeg J.M.. Inflight salivary pharmacokinetics of acetaminophen in saliva. Results of the Life Science DSOs Conducted Aboard the Space Shuttle 1981–1986 (1987) 19-23 58. Gandia P., Bareille M.P., Saivin S., Le-Traon A.P., Lavit M., Guell A., Houin G.. Influence of simulated weightlessness on the oral pharmacokinetics of acetaminophen as a gastric emptying probe in man: A plasma and a saliva study. J. Clin. Pharmacol. (2003) 43 1235-1243. DOI: 10.1177/0091270003257229 59. Idkaidek N., Arafat T.. Effect of microgravity on the pharmacokinetics of Ibuprofen in humans. J. Clin. Pharmacol. (2011) 51 1685-1689. DOI: 10.1177/0091270010388652 60. Polyakov A.V., Svistunov A.A., Kondratenko S.N., Kovachevich I.V., Repenkova L.G., Savelyeva M.I., Kukes V.G.. Peculiarities of Pharmacokinetics and Bioavailability of Some Cardiovascular Drugs under Conditions of Antiorthostatic Hypokinesia. Bull. Exp. Biol. Med. (2020) 168 465-469. DOI: 10.1007/s10517-020-04732-w 61. Gandia P., Saivin S., Le-Traon A.P., Guell A., Houin G.. Influence of simulated weightlessness on the intramuscular and oral pharmacokinetics of promethazine in 12 human volunteers. J. Clin. Pharmacol. (2006) 46 1008-1016. DOI: 10.1177/0091270006291032 62. Zhao Y., Dong S., Kong Y., Rui Q., Wang D.. Molecular basis of intestinal canonical Wnt/β-catenin BAR-1 in response to simulated microgravity in Caenorhabditis elegans. Biochem. Biophys. Res. Commun. (2020) 522 198-204. DOI: 10.1016/j.bbrc.2019.11.082 63. Guo Y., Xiao L., Sun L., Liu F.. Wnt/beta-catenin signaling: A promising new target for fibrosis diseases. Physiol. Res. (2012) 61 337-346. DOI: 10.33549/physiolres.932289 64. Sebio A., Kahn M., Lenz H.J.. The potential of targeting Wnt/β-catenin in colon cancer. Expert Opin. Ther. Targets (2014) 18 611-615. DOI: 10.1517/14728222.2014.906580 65. Lorenzon A., Calore M., Poloni G., De Windt L.J., Braghetta P., Rampazzo A.. Wnt/β-catenin pathway in arrhythmogenic cardiomyopathy. Oncotarget (2017) 8 60640-60655. DOI: 10.18632/oncotarget.17457 66. Hussain M., Xu C., Lu M., Wu X., Tang L., Wu X.. Wnt/β-catenin signaling links embryonic lung development and asthmatic airway remodeling. Biochim. Biophys. Acta Mol. Basis Dis. (2017) 1863 3226-3242. DOI: 10.1016/j.bbadis.2017.08.031 67. Laksitorini M.D., Yathindranath V., Xiong W., Hombach-Klonisch S., Miller D.W.. Modulation of Wnt/β-catenin signaling promotes blood-brain barrier phenotype in cultured brain endothelial cells. Sci. Rep. (2019) 9 19718. DOI: 10.1038/s41598-019-56075-w 68. Kang Y., Zhang S., Cao W., Wan D., Sun L.. Knockdown of LncRNA CRNDE suppresses proliferation and P-glycoprotein-mediated multidrug resistance in acute myelocytic leukemia through the Wnt/β-catenin pathway. Biosci. Rep. (2020) 40 BSR20193450. DOI: 10.1042/BSR20193450 69. Xia Z., Guo M., Liu H., Jiang L., Li Q., Peng J., Li J.D., Shan B., Feng P., Ma H.. CBP-dependent Wnt/β-catenin signaling is crucial in regulation of MDR1 transcription. Curr. Cancer Drug Targets (2015) 15 519-532. DOI: 10.2174/1568009615666150506093643 70. Li W., Xue H., Li Y., Li P., Ma F., Liu M., Kong S.. ATDC promotes the growth and invasion of hepatocellular carcinoma cells by modulating GSK-3β/Wnt/β-catenin signalling. Clin. Exp. Pharmacol. Physiol. (2019) 46 845-853. DOI: 10.1111/1440-1681.13119 71. Yu X., Yan N., Li Z., Hua Y., Chen W.. FGF19 sustains the high proliferative ability of keratinocytes in psoriasis through the regulation of Wnt/GSK-3β/β-catenin signalling via FGFR4. Clin. Exp. Pharmacol. Physiol. (2019) 46 761-769. DOI: 10.1111/1440-1681.13103 72. Wang X., Hawkins B.T., Miller D.S.. Activating PKC-β1 at the blood-brain barrier reverses induction of P-glycoprotein activity by dioxin and restores drug delivery to the CNS. J. Cereb. Blood Flow Metab. (2011) 31 1371-1375. DOI: 10.1038/jcbfm.2011.44 73. Bentires-Alj M., Barbu V., Fillet M., Chariot A., Relic B., Jacobs N., Gielen J., Merville M.P., Bours V.. NF-kappaB transcription factor induces drug resistance through MDR1 expression in cancer cells. Oncogene (2003) 22 90-97. DOI: 10.1038/sj.onc.1206056 74. McCubrey J.A., Steelman L.S., Abrams S.L., Lee J.T., Chang F., Bertrand F.E., Navolanic P.M., Terrian D.M., Franklin R.A., D’Assoro A.B.. Roles of the RAF/MEK/ERK and PI3K/PTEN/AKT pathways in malignant transformation and drug resistance. Adv. Enzyme Regul. (2006) 46 249-279. DOI: 10.1016/j.advenzreg.2006.01.004 75. Choi B.H., Kim C.G., Lim Y., Shin S.Y., Lee Y.H.. Curcumin down-regulates the multidrug-resistance mdr1b gene by inhibiting the PI3K/Akt/NFkB pathway. Cancer Lett. (2008) 259 111-118. DOI: 10.1016/j.canlet.2007.10.003 76. Morey-Holton E., Wronski T.J.. Animal models for simulating weightlessness. Physiologist (1982) 24 45-48 77. Scanziani E.. Immunohistochemical staining of fixed tissues. Methods Mol. Biol. (1998) 104 133-140. PMID: 9711649 78. Ko Y.J., Zaharias R.S., Seabold D.A., Lafoon J., Schneider G.B.. Osteoblast differentiation is enhanced in rotary cell culture simulated microgravity environments. J. Prosthodont. (2007) 16 431-438. DOI: 10.1111/j.1532-849X.2007.00204.x 79. Kararli T.T.. Comparison of the gastrointestinal anatomy, physiology, and biochemistry of humans and commonly used laboratory animals. Biopharm. Drug Dispos. (1995) 16 351-380. DOI: 10.1002/bdd.2510160502
--- title: Effect of Combining Impact-Aerobic and Strength Exercise, and Dietary Habits on Body Composition in Breast Cancer Survivors Treated with Aromatase Inhibitors authors: - Marisol Garcia-Unciti - Natalia Palacios Samper - Sofía Méndez-Sandoval - Fernando Idoate - Javier Ibáñez-Santos journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049091 doi: 10.3390/ijerph20064872 license: CC BY 4.0 --- # Effect of Combining Impact-Aerobic and Strength Exercise, and Dietary Habits on Body Composition in Breast Cancer Survivors Treated with Aromatase Inhibitors ## Abstract This study examines both the effect of a twice-weekly combined exercise—1 h session of strength and 1 h session of impact-aerobic—on body composition and dietary habits after one year of treatment with aromatase inhibitors (AI) in breast cancer survivors. Overall, forty-three postmenopausal women with a BMI ≤ 35 kg/m2, breast cancer survivors treated with AI, were randomized into two groups: a control group (CG) ($$n = 22$$) and a training group (IG) ($$n = 21$$). Body composition, i.e., abdominal, visceral, and subcutaneous adipose tissue) was measured by magnetic resonance. In addition, some questionnaires were used to gather dietary data and to measure adherence to the Mediterranean diet. After one year, women in the IG showed a significant improvement in body composition, indicated by decreases in subcutaneous and visceral adipose tissue, and total fat tissue. Furthermore, the dietary habits were compatible with moderate adherence to the Mediterranean diet pattern and a low dietary intake of Ca, Zn, Folic Ac, and vitamins D, A, and E. A twice-weekly training program combining impact aerobic exercise and resistance exercise may be effective in improving the body composition for postmenopausal women who have breast cancer treated with AI, and the results suggest the need for nutritional counselling for this population. ## 1. Introduction Breast cancer (BC) is defined as a malignant tumor that affects different breast cells [1]. It is considered a hormone-dependent disease characterized by molecular mechanisms involving activation of human epidermal growth factor receptor 2 (HER2), hormone receptors (estrogen receptor and progesterone receptor) and/or BRCA mutations [2]. Most BCs (70–$80\%$) express a significant amount of estrogen receptors (ER) and/or progesterone receptors (PR), which are considered biomarkers of a favorable prognosis [3]. The disease can be classified according to the stage of the tumor and its localization, as well as according to the molecular subtype, determined by the analysis of the gene expression of HER2, and by quantitative hormone receptor (HR) [1]. Breast cancer is the most common tumor and is the first cause of death related to cancer among women [4]. According to estimates from Global Cancer Statistics, female breast cancer is the leading cause of global cancer incidence in 2020, representing $11.7\%$ of all cancer cases, and is the fifth leading cause of cancer mortality worldwide. Among women, breast cancer accounts for 1 in 4 cancer cases and for 1 in 6 cancer deaths, ranking first for incidence and mortality in the majority of countries. This increase has been related to a higher prevalence of reproductive and hormonal risk factors and lifestyle risk factors, such as excess body weight and physical inactivity, as well as increased detection through mammographic screening [4]. In addition, the increasing incidence is linked to estrogen receptor-positive cancer, due to the stronger and more consistent association of excess body weight with estrogen receptor-positive cancer and the impact of mammographic screening, which preferentially detects slow-growing estrogen receptor-positive cancers [5,6]. The treatment depends on the woman’s hormonal status, kind, and stage of tumor, resulting in surgery therapy, mastectomy, radiation, chemotherapy, or if carcinoma is estrogen receptor–positive, patients may also receive endocrine therapy [7,8]. Early invasive stages (I, IIa, IIb) and locally advanced stages (IIIa, IIIb, IIIc) have three treatment phases: preoperative phase, surgery, and postoperative phase. The preoperative phase uses systemic endocrine or immunotherapies when tumors express estrogen, progesterone, or ERBB2 receptors. Preoperative chemotherapy may also be used and is the only option when tumors have none of those three receptors, and the postoperative phase includes radiation, immunotherapy, chemotherapy and/or endocrine therapy. When endocrine therapy is needed, only tamoxifen should be used in premenopausal women [9], while aromatase inhibitors for years is the current treatment recommended in postmenopausal women 35 years or older who are at increased risk for breast cancer and low risk for adverse medication effects [10]. Treatment with AI reduces cancer recurrence and improves overall survival in women in the early stages of the disease. However, this treatment is associated with musculoskeletal symptoms such as arthralgias, myalgias, osteoporosis-related bone fractures and unfavorable changes in body composition—a reduction of bone mineral density and muscle mass, and an increase in fat tissue and body weight [7]. Both have been related to poorer survival and an increase in cardiovascular events [11,12,13,14]. Given the efficacy of AIs and the large proportion of women diagnosed with breast cancer using this treatment, it is important to have available interventions to improve those AI side effects as well as the quality of life and all-cause mortality. Therefore, all women should be advised to adopt a healthier lifestyle that promotes overall health. In this context, although the evidence regarding the benefits of non-pharmacological measures specific to breast cancer survivors is limited. Recently, the American Cancer Society [15], and other authors [16,17,18,19] have published recommendations about healthy lifestyles that can help to prevent possible recurrences or complications. These include increasing physical activity, following a healthy diet pattern, and maintaining a healthy weight. In addition, there is some evidence about the potential negative side effects from AIs being diminished through the implementation of regular physical activity [20] and weight-bearing exercise, which includes impact and resistance training, following a healthy diet with adequate daily calcium and vitamin D intake [21,22,23,24]. We hypothesized that a low-frequency and twice-weekly training program combining one session of impact aerobic exercise and one session of resistance exercise might improve the body composition—body weight, body fat tissue and muscle mass—of postmenopausal women breast cancer survivors after one year of treatment with AI. Therefore, taking into account the need for more information, both on the most suitable exercise for patients treated with Ais and on their nutritional status and dietary habits, the aim of this study was two-fold. First, to analyze the effect of a low-frequency and twice-weekly training program combining one session of impact aerobic exercise and one session of resistance exercise on body composition—body weight, body fat tissue and muscle mass—of postmenopausal women with breast cancer after one year of treatment with AI; and second, to examine the lifestyle habits—nutritional and physical activity habits—within this population after one year of treatment with AI. ## 2.1. Participants This was a single-blind study that was undertaken for 5 years, from January 2012 to September 2017. A group of 43 postmenopausal women, aged 55–70 years, with a BMI of 19–35 kg/m², who had surgery for hormone-dependent breast cancer and treatment with chemotherapy and/or radiotherapy, and were about to begin treatment with AI, were included voluntarily in the study. The participants were referred by the Service of Oncology of a hospital in Spain. Performing physical exercise regularly and the presence of osteoporosis, diabetes mellitus, hypothyroidism, liver or kidney dysfunction, high blood pressure, heart disease, asthma, chronic obstructive pulmonary disease (COPD), joint disorders, alcoholism, or any other drug addiction, were considered as exclusion criteria, resulting in a sample of 43 women. The randomization procedure used prevented investigators from influencing group allocation. All researcher staff remained blinded until the end of the study, except for the sports trainers. The volunteers only were informed about their group assignment after the randomization process was completed. The volunteers were randomized using numbered and sealed envelopes that contained a paper with an intervention or control group. One-to-one, and sequentially, participants chose freely the envelope they wanted, resulting in being assigned to a Control Group (CG, 22 patients) or Intervention Group (IG, 21 patients), in which, each volunteer was followed for one year. Participant flow through the study is presented in Figure 1. ## 2.2. Procedure The CG and IG followed the usual medical supervision by the Service of Oncology and similar pharmacological treatment with AI and calcium (1000 mg/d) and vitamin D (20 µg/d) supplementation. During the time of the study, the IG followed a supervised training with two sessions per week: one session of odd-impact training (multidirectional impact aerobic exercise) and one of strength training. Both groups maintained their regular lifestyle and filled out a diary in which the volunteers recorded their physical activity, the appearance of symptoms, and the medication taken for the year of study. Further, anthropometric measures and tests of lifestyle, diet and nutritional status were carried out at the beginning of the study, before starting treatment with AI, and also at 6 and 12 months of the treatment. ## 2.3. Training Program The training was divided into five training cycles in which the load was progressively increased: one cycle of 3 weeks of familiarization and four cycles of 9–14 weeks of duration. Each training session lasted between 60 and 90 min. There were two weeks of a break during the study, at weeks 31–32. The odd-impact training consisted of an aerobic choreography with multidirectional jumps. To increase the workload, the jump distance was increased from one-line jumps to 60 cm side squares jumps. The number of jumps per session varied from 96 during the familiarization period to 756 jumps at the end of the program. For strength training, different resistance machines were used (Technogym, Gambettola, Italy) for different exercises. The training workload began under $40\%$ of 1-RM during the familiarization cycle and progressively was increased to 40–$70\%$ of 1-RM in the next four cycles. Each session consisted of 3 sets of 4–12 repetitions for 5 exercises. Two training routines were alternated. Each one had 3 common exercises—leg press, chest press and hip extension—and 4 exercises that were alternately performed according to the routines: Routine 1, which consisted of knee extension and pulldown; and Routine 2, which consisted of knee-flexion and shoulder press. Every 4–8 weeks, the 1-RM test was made to recalculate weights. All the sessions included warm-up, main part, and cool-down. During cool-down, stabilizing work of the abdominal-pelvic area and general stretching of worked-out muscles were made. All participants were required $90\%$ to adhere to the training program. All the sessions were individually supervised by trained leaders. Besides transient musculoskeletal soreness, no major complications or injuries were reported. ## 2.4. Body Composition and Nutritional Status The study of body composition involved body weight, muscle mass and fat tissue through anthropometric measurement of body weight, height, waist and arm circumferences, and triceps skinfold (TSF). From these data, we calculated the Body Mass Index (BMI, kg/m2), total body fat (% BF) according to the CUN-BAE formula [25], the arm muscle circumference (AMC), and the arm muscle area (AMA) [26]. The measurements of these parameters were made by the same dietitian, following the protocol of the International Standards for Anthropometric Measurements of ISAK [27]. Patients were weighed in light clothing without shoes on SECA 714 weighing scale with a graduation of 0.1 kg. Patient height was determined with a height rod stadiometer (SECA 220) using a graduation of 1.0 mm. All measurements were obtained in duplicate. The measurements of waist, hip, and arm circumferences were made in duplicate or triplicate using a Cescorf flexible steel tape measure with a graduation of 1.0 mm. The triceps skinfold thickness was measured using a Harpenden skinfold calliper (0.2 mm). The participants were classified according to their BMI as described by World Health Organization [28]. Moreover, abdominal visceral adipose tissue (VAT) (cm2) and abdominal subcutaneous adipose tissue (SAT) (cm2) were measured at L3 discal level. The muscle volume was measured in the thigh. These three measurements were made by magnetic resonance (Magnetom Impact Expert; Siemens) by the same experienced operator using a body coil [29]. Intra-observer reliability for calculation of total VAT and SAT volumes was 0.99 with a coefficient of variation of 5–$8\%$. The Patient-Generated Subjective Global Assessment (PG-SGA) was applied during a face-to-face interview. According to the PG-SGA score, patients were classified from 0 to 8 with adequate nutritional status and ≥9 with undernutrition status [30]. ## 2.5. Level of Physical Activity The total level of physical activity was estimated in terms of metabolic equivalents (METs). We collected information about physical activity through a 17-item leisure time physical activity questionnaire, which provided valid and reproducible self-reported data [31,32,33]. METs were calculated according to the activity and the time spent on it per week, obtaining METs/h/week. In the IG, we added 3 additional METs. These correspond to the estimate of the energy expenditure of the strength training session of moderate intensity, carried out by the volunteers in this group [34,35]. ## 2.6. Dietary Habits The study of dietary habits was conducted by a trained nutritionist. It included the study of nutrient intake through non-consecutive 3-day food records [36]. To quantify food and beverage consumption at home, the volunteers used a scale (Leifheit), while to estimate the intake outside their homes, they used a photo album [37]. Additionally, the intake of vitamins and minerals supplements was recorded. The food weights were converted to nutrient intake estimates per day by the EasyDiet® program [38], using the Spanish food composition database. The study of the dietary pattern was assessed using a food frequency questionnaire validated in Spain [31,32,33] as well as the Mediterranean Diet Score (MDS) [39], which assesses adherence to the Mediterranean dietary pattern (MedDiet). For the present study, the score obtained in MDS was classified into three categories: with 0–2 points indicating low adherence, 3–6 points indicating moderate adherence, and 7–9 points indicating high adherence to the MedDiet [40]. Furthermore, diet quality was assessed according to the recommended food servings and the nutritional goals authorized by the Spanish Society for Community Nutrition (SENC) [41,42]. To carry out the analysis of the adequacy of intake to the recommendations, it was established that the volunteers met the recommendation if they reached the minimum indicated in the guidelines of the SENC. For products classified as “Occasional and moderate consumption”, the recommended weekly serving was established as one. ## 2.7. Statistical Analyses STATA statistical software package version 12.1 was used for the statistical analysis (StataCorp. [ 2011]. Stata Statistical Software: Release 12 College Station, TX: StataCorp LP). On the one hand, two-sided analysis and a level of $5\%$ using the t-test for independent samples and t-test for matched data for the parametric variables were used. On the other hand, the U of Mann-Whitney sign test or Wilcoxon rank test were used for non-parametric variables. For qualitative variables, frequency analysis was used. Data are presented as mean (SD) for quantitative variables and as a percentage of sample n (%) for qualitative variables. ## 3. Results At baseline, no significant differences were found between volunteers in the CG and IG. The only exception was the level of physical activity, which was significantly higher in the CG (Table 1). It should be noted that none of the participants were at risk of malnutrition; on the contrary, some level of overweight was observed (BMI > 24.9 kg/m²) in $63\%$ of the women (Table S1). Biochemical data are shown in supplementary documentation (Table S2). ## 3.1. Body Composition and Nutritional Status At the end of the study, the IG showed a decrease in weight, subcutaneous fat measured by the triceps skinfold (TSF), and waist circumference, associated with a decrease in abdominal fat, both subcutaneous (SAT) and visceral (VAT). Furthermore, the decrease of two points in PG-SGA is interpreted as an improvement in the nutritional status. In contrast, among the volunteers in the CG, there were no significant changes in the nutritional status or body composition, although a tendency towards an increase in total fat was observed ($$p \leq 0.07$$). When both groups were compared, significant differences were observed at the end of the study in weight, BMI, waist circumference, %Total BF, TSF, and VAT, in the expected direction of the intervention; more details are shown in Table 2, Figure 2 and Figure 3. ## 3.2. Physical Activity Level At baseline, the participants in the CG presented a higher level of physical activity than those in the IG. However, after one year of follow-up, we observed a clear tendency to a decrease in the total physical activity level of the CG (−11.8 METs/h/week; $$p \leq 0.07$$). Whereas, in the IG, the level of total physical activity was maintained, although the exercise program was equivalent to 10–15 METs/h/week. No differences between the two groups were observed in the total physical activity level at the end of the study (Table 2). ## 3.3. Dietary Habits According to SENC guidelines [41,42], both groups of volunteers had lower cereal and alcohol consumption, while the intake of fish was higher than the recommendations. For the IG, consumption of red meat, fast food, pastries, and sweetened soft drinks was also higher than the recommendations. The two groups showed a trend towards low fruit consumption (Table 3). Nevertheless, these results are in line with those obtained from the Mediterranean Diet Score (MDS). After 12 months of intervention, we observed only around $19\%$ of the total volunteers presenting a high adherence to the MedDiet pattern and $77\%$ of them having moderate adherence to the pattern (Table S3). After one year of follow-up, no changes in caloric intake and macronutrient intake were observed in any of the two groups of the study (Table S4). We only saw an increase in the intake of some micronutrients—total vitamin D in the CG (mean difference of 5.9 μg/d ± 13.1 μg/d; $p \leq 0.05$) and the iron (mean difference of 1.9 mg/d ± 3.8 mg/d; $p \leq 0.05$) and niacin (mean difference of 3.2 mg ± 5.0 mg; $p \leq 0.05$) intake in the IG (Table S5). ## 4.1. Body Composition and Nutritional Status The main finding of this study was that a twice-weekly training program combining one session of impact-aerobic exercise and one of resistance training in women treated with breast cancer with AI led to significant weight and fat weight loss, total fat loss, and decreased waist circumference, associated with a significant SAT and VAT loss. In contrast, the CG did not show any variation in body composition. When overweight, VAT and SAT volumes are positively related to cardiovascular risk, and it is known that cardiovascular disease is an important cause of death in patients with breast cancer [43,44]. Furthermore, although there is limited evidence that intentional weight loss after diagnosis may be beneficial for breast cancer-specific mortality, some studies suggest that weight loss over this time frame may be suggestive of a lower relative risk of breast cancer recurrence in estrogen receptor (ER)-positive breast cancer survivors [45]. Furthermore, in this context, it has been shown that therapy with AI in sedentary breast cancer patients is related to changes in fat distribution, with a relatively great VAT/SAT ratio, regardless of whether they gain or lose weight after therapy [46], and this pattern of fat distribution is associated with breast cancer recurrence [47]. Dimauro et al. [ 1] studied the effect of different physical exercise protocols as secondary and tertiary prevention among ER+ breast cancer survivors before and during pharmacological treatments, including aromatase inhibitors. The protocols included two or more training days per week, two or more days of resistance training per week, the inclusion of high-intensity (HIIT) training, or the combination of diet with the training exercise, with favorable results in body composition and decrease in the side effects of aromatase inhibitors in old BC patients. However, unlike our study, none of them included only one session of impact-aerobic exercise and one of resistance training without a concomitant hypocaloric diet. Thus, the improvements in body composition observed in this study show that this novel protocol, performed twice a week, could mitigate the risk of cardiovascular disease and breast cancer recurrence risk in postmenopausal women treated for breast cancer with AI [15]. However, our training program did not provide enough stimulus to achieve significant improvements in muscle mass. A study by Thomas et al. [ 48], which included three sessions of exercise per week, two of which combined strength and aerobic exercise and one session of 150 min of moderate-intensity exercise at home over 12 months, showed losses in fat mass as well as an increase in lean mass in cancer survivors treated with AI. The reason for this difference in our study may be related to the number of strength training sessions per week. Our training program consisted of only two days per week of scheduled activity, and only one of those days was dedicated to a strength workout. Unlike our study, the one carried out by Santos et al. [ 49] investigated the effects of resistance training, once a week for 8 weeks, on changes in body composition and muscular strength in breast cancer survivors, and although no changes in body composition were detected, they observed improved muscular strength [49]. These results suggest that although exercise programs equivalent to 10–15 METs/h/week may result in significant fat loss, it would be necessary to perform at least two strength-training sessions per week to achieve significant improvements in muscle mass in postmenopausal breast cancer survivors treated with AI. ## 4.2. Lifestyle: Physical Activity Level and Dietary Habits The 2018 American Physical Activity Guide showed that regular recreational physical activity is associated with a decrease in the risk of breast cancer death and mortality from any other cause [50]. At the beginning of our trial, the group of volunteers randomly assigned to the CG presented a level of physical activity 1.6 times higher than the IG, but the physical activity in both groups was considered moderate [51,52]. During the year of the study, the CG continued with their usual activities while the IG attended two weekly training sessions. However, at the end of the study, it was observed that the tendency in the CG was to decrease the level of physical activity. Meanwhile, for the IG, thanks to the programmed exercise, the physical activity was maintained, bringing both groups in line with each other. The behavior observed in the CG is aligned with that described in the Women Health Initiative study, in which results showed that the level of physical activity decreased over time, maintaining significantly low levels even 10 years after a breast cancer diagnosis. The tumor stage at diagnosis, the type of treatment received, the BMI, and the presence of other comorbidities were identified as possible causes of the decrease in physical activity [53]. In our study, both groups of volunteers reported an increase in arthralgias and myalgias associated with AI treatment, so this may be another factor to consider when explaining the decrease in the level of physical activity in the CG. To note, our training program resulted in a metabolic output within 10–15 METs/h/week, depending on the phase and intensity of training, which seems to be effective in counteracting the reduction in total physical activity level that spontaneously occurs in some women with cancer treated with AI. Some authors have found that the practice of physical activities equivalent to 10 METs/h/week or more, i.e., 2.5 h/week at moderate intensity, is associated with a 24–$27\%$ reduction in all-cause mortality and a $25\%$ reduction in breast cancer-specific mortality [54,55], improving overall survival after cancer diagnosis [56]. Recently, two systematic reviews and meta-analyses [57,58] compared women who had the highest levels of recreational or total physical activity with those who had the lowest levels and found reduced risks of $42\%$ all-cause and between $37\%$ and $42\%$ breast cancer-specific mortality in the most active versus least active categories. “ The protective effects of physical activity on breast cancer-specific mortality may include reduced exposure to estrogen and androgen, the effects of insulin and insulin-related factors, and reduced inflammation. Physical activity may affect these pathways directly or indirectly by its effects on reducing body weight. The lower risk for all-cause mortality may be linked to other benefits of physical activity through reduced cardiovascular risk (i.e., improved exercise capacity) and reduced risk for other comorbidities” [59]. Nevertheless, currently, there is insufficient information to support specific recommendations related to the domain, optimal dose and timing of activity that could contribute to the dose of 10–15 METs hours per week in postmenopausal women with breast cancer [15,57,60]. The evidence highlights the importance of breast cancer survivors engaging in any amount of physical activity they can, increasing their activity level when possible, and especially not decreasing physical activity after their diagnosis and treatment [15]. In this context, the training exercise protocol that we propose fits with the amount of physical activity recommended by these groups and leads us to consider that our program of intervention could be an essential part of the treatment. Dietary habits can also play a relevant role in body composition and the risk of cardiometabolic diseases [61,62,63,64,65,66,67,68], and evidence shows that high adherence to MedDiet reduces the risk of chronic disease (including cardiovascular disease), cancer mortality, and appears to have a preventive effect on the incidence of breast cancer [67,69,70,71,72,73]. An increase of only two points in the MDS has been associated with a $25\%$ reduction in total mortality [39]. In cancer survivors, there is little evidence indicating that dietary behaviors influence outcomes regarding recurrence and mortality. However, even though the evidence is limited, studies suggest that a high-quality diet such as MedDiet is beneficial for these patients [70,71] due to the presence of bioactive compounds that could act by different mechanisms, such as increasing antioxidant capacity, controlling body weight, and improving the glucose profile [65,68,70,73,74,75,76,77,78]. For this reason, knowing the eating habits of a population can provide more effective nutritional advice. In our study, although in both groups the consumption of most food groups was in line with the SENC recommendations about a healthy diet, the percentage of volunteers with high adherence to MedDiet decreased from $20.8\%$ to $18.6\%$ during the year of the study, and this was reflected in the frequency of consumption by food groups. In general, the consumption of red and processed meats, fast food, pastries, and soft drinks was higher than the recommendations, and this intake has been linked to an adverse effect on body weight control, cardiovascular risk, and cancer risk, due to their higher caloric intake and their content of saturated fat and cholesterol [62,63,64,65,66,68]. Moreover, high consumption of saturated fat, from red and processed meat, has been associated with an increased risk of hormone receptor-positive breast cancer and breast cancer-specific mortality [62,76,79]. Although these results are aligned with those observed in a recent study carried out in the north of Spain, as well as with the general pattern of consumption among the Spanish population [41,80], they differ from those obtained by other researchers, who showed dietary changes among breast cancer survivors in the first year after treatment with a lower intake of red meat consumption along with lower energy [81,82,83,84]. One aspect that appears to influence the acquisition of healthier lifestyle habits in patients may be professional advice. Recommendations by health professionals are associated with higher rates of behavior change among cancer survivors [85,86]. In this study, no nutritional intervention was carried out, which could justify the minimal changes observed in the dietary intake of our sample. Concerning the intake of macro and micronutrients, it was barely modified in the whole sample of volunteers during the year of the study, but it is also worth noting that the dietary intake of Calcium, Zinc, Folic acid and vitamins D, A and E did not cover the recommendations for this age group in the Spanish population [87]. The data obtained were similar to those reported by the ANIBES study [80,88] in Spanish women with a similar age range. However, the requirements of calcium and vitamin D were ensured by the prescription of pharmacological supplements since the beginning of the intervention. Previous studies have shown that low intakes of Calcium and vitamin D may contribute to low serum vitamin D levels, which are inversely related to breast cancer [89,90]. On the contrary, higher levels of these elements are related to increased survival, as well as an improvement in bone health [91]. This is especially relevant in patients treated with AI, where significant losses in bone mineral density have been reported. Although no evidence directly relates vitamins A and E intake with susceptibility or prevention of breast cancer, the antioxidant effect of both vitamins could act positively on cell damage and cell activity, aspects associated with the development and recurrence of cancer [91]. Concerning Folic Acid, the results of the EPIC cohort showed that folate status was not clearly associated with breast cancer risk [92,93]. In summary, according to the literature, lifestyle interventions that include an increased level of physical activity and nutritional counselling appear to be useful to improve therapeutic efficacy, limiting drug-induced side effects, including those related to AI [94], and improving overall survival after cancer diagnosis [56]. This leads us to consider, that these interventions should be an essential part of the therapeutic approach [76]. This suggests the need for professional counselling, so it would be important for health professionals to provide timely education to allow patients to perceive the benefits of healthier lifestyles throughout the treatment period [86,95]. ## 4.3. Limitations of the Study The main study limitation was the magnitude of the trial, which limited the ability to draw strong conclusions. Other limitations were the differences in physical activity between the groups before enrollment, but it was a description of the real physical activity of volunteers, not related to the practice of any programmed physical activity; and the use of food records to study dietary habits. However, dietary records are considered a reference method in validation studies. ## 5. Conclusions This study shows that a twice-weekly training program combining one session of impact-aerobic exercise and one session of resistance exercise improved the body composition in postmenopausal women treated for breast cancer with AIs. After one year, without a concomitant weight loss diet or dietary advice, women in the exercise group showed a decrease in total fat, SAT, VAT, and waist circumference. This low-frequency combined exercise program may be more acceptable, effective, and practical in optimizing aspects of physical fitness than programs that involve a higher frequency. It is suggested that the choice of this low-frequency regimen will be made when the ability of the individual to withstand a vigorous program or the willingness to devote a high weekly training frequency is a constraint. In addition, the dietary habits of our volunteers were characterized by dietary habits compatible with moderate adherence to the Mediterranean diet pattern, but with a low dietary intake of Ca, Zn, Folic Ac, Vit D, A and E, which suggests the need for nutritional counselling. All these issues should be considered when planning future advice programs for breast cancer survivors. ## References 1. Dimauro I., Grazioli E., Antinozzi C., Duranti G., Arminio A., Mancini A., Greco E.A., Caporossi D., Parisi A., Di Luigi L.. Estrogen-Receptor-Positive Breast Cancer in Postmenopausal Women: The Role of Body Composition and Physical Exercise. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18189834 2. Harbeck N., Penault-Llorca F., Cortes J., Gnant M., Houssami N., Poortmans P., Ruddy K., Tsang J., Cardoso F.. Breast cancer. Nat. Rev. Dis. Primers (2019) 5 66. DOI: 10.1038/s41572-019-0111-2 3. Harbeck N.. Advances in targeting HER2-positive breast cancer. Curr. Opin. Obstet. Gynecol. (2018) 30 55-59. DOI: 10.1097/GCO.0000000000000431 4. Sung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F.. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. (2021) 71 209-249. DOI: 10.3322/caac.21660 5. Pati S., Irfan W., Jameel A., Ahmed S., Shahid R.K.. Obesity and Cancer: A Current Overview of Epidemiology, Pathogenesis, Outcomes, and Management. Cancers (2023) 15. DOI: 10.3390/cancers15020485 6. Munsell M.F., Sprague B.L., Berry D.A., Chisholm G., Trentham-Dietz A.. Body Mass Index and Breast Cancer Risk According to Postmenopausal Estrogen-Progestin Use and Hormone Receptor Status. Epidemiol. Rev. (2014) 36 114-136. DOI: 10.1093/epirev/mxt010 7. Trayes K.P., Cokenakes S.E.H.. Breast Cancer Treatment. Am. Fam. Physician (2021) 104 171-178. PMID: 34383430 8. Fahad Ullah M.. Breast Cancer: Current Perspectives on the Disease Status. Advances in Experimental Medicine and Biology (2019) 51-64 9. Visvanathan K., Fabian C.J., Bantug E., Brewster A.M., Davidson N.E., DeCensi A., Floyd J.D., Garber J.E., Hofstatter E.W., Khan S.A.. Use of Endocrine Therapy for Breast Cancer Risk Reduction: ASCO Clinical Practice Guideline Update. J. Clin. Oncol. (2019) 37 3152-3165. DOI: 10.1200/JCO.19.01472 10. Force U.P.S.T., Owens D.K., Davidson K., Krist A.H., Barry M.J., Cabana M., Caughey A.B., Doubeni C.A., Epling J.W., Kubik M.. Medication Use to Reduce Risk of Breast Cancer: US Preventive Services Task Force Recommendation Statement. J. Am. Med. Assoc. (2019) 322 857-867. DOI: 10.1001/jama.2019.11885 11. Clines G.A., Choksi P., Van Poznak C.. Adjuvant Endocrine Therapy and Bone Health in Breast Cancer. Curr. Osteoporos. Rep. (2015) 13 263-273. DOI: 10.1007/s11914-015-0277-9 12. Goldvaser H., Barnes T.A., Šeruga B., Cescon D.W., Ocana A., Ribnikar D., Amir E.. Toxicity of Extended Adjuvant Therapy with Aromatase Inhibitors in Early Breast Cancer: A Systematic Review and Meta-analysis. Gynecol. Oncol. (2018) 110 31-39. DOI: 10.1093/jnci/djx141 13. Grossmann M., Ramchand S.K., Milat F., Vincent A., Lim E., Kotowicz M.A., Hicks J., Teede H.. Assessment and management of bone health in women with oestrogen receptor-positive breast cancer receiving endocrine therapy: Position statement of the Endocrine Society of Australia, the Australian and New Zealand Bone & Mineral Society, the Australasian. Clin. Endocrinol. (2018) 89 280-296. DOI: 10.1111/cen.13735 14. Roberts K., Rickett K., Greer R., Woodward N.. Management of aromatase inhibitor induced musculoskeletal symptoms in postmenopausal early Breast cancer: A systematic review and meta-analysis. Crit. Rev. Oncol. (2017) 111 66-80. DOI: 10.1016/j.critrevonc.2017.01.010 15. Rock C.L., Thomson C.A., Sullivan K.R., Howe C.L., Kushi L.H., Caan B.J., Neuhouser M.L., Bandera E.V., Wang Y., Robien K.. American Cancer Society nutrition and physical activity guideline for cancer survivors. CA: A Cancer J. Clin. (2022) 72 230-262. DOI: 10.3322/caac.21719 16. Campbell K.L., Winters-Stone K.M., Wiskemann J., May A.M., Schwartz A.L., Courneya K.S., Zucker D.S., Matthews C.E., Ligibel J.A., Gerber L.H.. Exercise Guidelines for Cancer Survivors: Consensus Statement from International Multidisciplinary Roundtable. Med. Sci. Sports Exerc. (2019) 51 2375-2390. DOI: 10.1249/MSS.0000000000002116 17. 17. WHO OMS, Serie de Informes Técnicos 916 Dieta, Nutrición Y Prevención de Enfermedades Crónicas Organización Mundial de la Salud GinebraWHOGeneva, Switzerland2018. OMS, Serie de Informes Técnicos 916 Dieta, Nutrición Y Prevención de Enfermedades Crónicas Organización Mundial de la Salud Ginebra (2018) 18. NCCN Guidelines for Patients. Survivorship Care for Healthy Living. (2020) 19. Clinton S.K., Giovannucci E.L., Hursting S.D.. The World Cancer Research Fund/American Institute for Cancer Research Third Expert Report on Diet, Nutrition, Physical Activity, and Cancer: Impact and Future Directions. J. Nutr. (2020) 150 663-671. DOI: 10.1093/jn/nxz268 20. Irwin M.L., Cartmel B., Gross C.P., Ercolano E., Li F., Yao X., Fiellin M., Capozza S., Rothbard M., Zhou Y.. Randomized Exercise Trial of Aromatase Inhibitor–Induced Arthralgia in Breast Cancer Survivors. J. Clin. Oncol. (2015) 33 1104-1111. DOI: 10.1200/JCO.2014.57.1547 21. Casla S., López-Tarruella S., Jerez Y., Marquez-Rodas I., Galvão D.A., Newton R.U., Cubedo R., Calvo I., Sampedro J., Barakat R.. Supervised physical exercise improves VO2max, quality of life, and health in early stage breast cancer patients: A randomized controlled trial. Breast Cancer Res. Treat. (2015) 153 371-382. DOI: 10.1007/s10549-015-3541-x 22. Dalla Via J., Daly R.M., Fraser S.F.. The effect of exercise on bone mineral density in adult cancer survivors: A systematic review and meta-analysis. Osteoporos. Int. (2018) 29 287-303. DOI: 10.1007/s00198-017-4237-3 23. Grossmann M., Ramchand S.K., Milat F., Vincent A., Lim E., Kotowicz M.A., Hicks J., Teede H.J.. Assessment and management of bone health in women with oestrogen receptor-positive breast cancer receiving endocrine therapy: Position statement summary. Med. J. Aust. (2019) 211 224-229. DOI: 10.5694/mja2.50280 24. Saarto T., Sievänen H., Kellokumpu-Lehtinen P., Nikander R., Vehmanen L., Huovinen R., Kautiainen H., Järvenpää S., Penttinen H.M., Utriainen M.. Effect of supervised and home exercise training on bone mineral density among breast cancer patients. A 12-month randomised controlled trial. Osteoporos. Int. (2012) 23 1601-1612. DOI: 10.1007/s00198-011-1761-4 25. Gómez-Ambrosi J., Silva C., Catalán V., Rodríguez A., Galofré J.C., Escalada J., Valentí V., Rotellar F., Romero S., Ramírez B.. Clinical Usefulness of a New Equation for Estimating Body Fat. Diabetes Care (2012) 35 383-388. DOI: 10.2337/dc11-1334 26. Heymsfield S.B., McManus C., Smith J., Stevens V., Nixon D.W.. Anthropometric measurement of muscle mass: Revised equations for calcu-lating bone-free arm muscle area. Am. J. Clin. Nutr. (1982) 36 680-690. DOI: 10.1093/ajcn/36.4.680 27. Stewart A., Markfell-Jones M., Olds T., Ridder H.. Protocolo Internacional para la valoracion antropométrica. Isak (2011) 82 1-117 28. 28. World Health Organization Waist Circumference and Waist-Hip Ratio Report of a WHO Expert ConsultationWorld Health OrganizationGeneva, Switzerland2008147. Waist Circumference and Waist-Hip Ratio Report of a WHO Expert Consultation (2008) 1-47 29. Idoate F., Ibañez J., Gorostiaga E.M., García-Unciti M., Martínez-Labari C., Izquierdo M.. Weight-loss diet alone or combined with resistance training induces different regional visceral fat changes in obese women. Int. J. Obes. (2011) 35 700-713. DOI: 10.1038/ijo.2010.190 30. Bauer J., Capra S., Ferguson M.. Use of the scored Patient-Generated Subjective Global Assessment (PG-SGA) as a nutrition assessment tool in patients with cancer. Eur. J. Clin. Nutr. (2002) 56 779-785. DOI: 10.1038/sj.ejcn.1601412 31. De La Fuente-Arrillaga C., Ruiz Z.V., Bes-Rastrollo M., Sampson L., Martinez-González M.A.. Reproducibility of an FFQ validated in Spain. Public Health Nutr. (2010) 13 1364-1372. DOI: 10.1017/S1368980009993065 32. Fernández-Ballart J.D., Piñol J.L., Zazpe I., Corella D., Carrasco P., Toledo E., Perez-Bauer M., Martínez-González M.Á., Salas-Salvadó J., Martín-Moreno J.M.. Relative validity of a semi-quantitative food-frequency questionnaire in an elderly Mediterranean population of Spain. Br. J. Nutr. (2010) 103 1808-1816. DOI: 10.1017/S0007114509993837 33. Martín-Moreno J.M., Boyle P., Gorgojo L., Maisonneuve P., Fernandez-Rodriguez J.C., Salvini S., Willett W.C.. Development and Validation of a Food Frequency Questionnaire in Spain. Int. J. Epidemiol. (1993) 22 512-519. DOI: 10.1093/ije/22.3.512 34. Ainsworth B.E., Haskell W.L., Whitt M.C., Irwin M.L., Swartz A.M., Strath S.J., O’Brien W.L., Bassett D.R., Schmitz K.H., Emplaincourt P.O.. Compendium of physical activities: An update of activity codes and MET intensities. Med. Sci. Sport. Exerc. (2000) 32 S498-S504. DOI: 10.1097/00005768-200009001-00009 35. Zanuso S., Bergamin M., Jimenez A., Pugliese G., D’Errico V., Nicolucci A., Ermolao A., Balducci S.. Determination of metabolic equivalents during low- and high-intensity resistance exercise in healthy young subjects and patients with type 2 diabetes. Biol. Sport (2016) 33 77-82. DOI: 10.5604/20831862.1194124 36. Ortega R.M., Perez-Rodrigo C., Lopez-Sobaler A.M.. Métodos de evaluación de la ingesta actual: Registro o diario dietético. Nutr. Hosp. (2015) 31 38-45. DOI: 10.3305/nh.2015.31.sup3.8749 37. Rusolillo G., Marques I.. Álbum Fotográfico de Porciones de Alimentos (2011) 38. 38. AEND Academia Española de Nutrición y DietéticaPrograma Informático Easydiet®Brisbane, Australia2008. Academia Española de Nutrición y Dietética (2008) 39. Trichopoulou A., Costacou T., Bamia C., Trichopoulos D.. Adherence to a Mediterranean Diet and Survival in a Greek Population. N. Engl. J. Med. (2003) 348 2599-2608. DOI: 10.1056/NEJMoa025039 40. Carlos S., De La Fuente-Arrillaga C., Bes-Rastrollo M., Razquin C., Rico-Campà A., Martínez-González M.A., Ruiz-Canela M.. Mediterranean Diet and Health Outcomes in the SUN Cohort. Nutrients (2018) 10. DOI: 10.3390/nu10040439 41. Bartrina J.A., Val V.A., Aldalur E.M., de Victoria Muñoz E.M., Anta R.M.O., Pérez-Rodrigo C., Izquierdo J.Q., Martín A.R., Viñas B.R., Castell G.S.. Dietary Guidelines for the Spanish population (SENC, diciembre 2016); the new graphic icon of healthy food. Nutr. Hosp. (2016) 33 1-48. DOI: 10.20960/nh.827 42. Aranceta-Bartrina J., Partearroyo T., López-Sobaler A.M., Ortega R.M., Varela-Moreiras G., Serra-Majem L., Pérez-Rodrigo C.. Updating the Food-Based Dietary Guidelines for the Spanish Population: The Spanish Society of Community Nutrition (SENC) Proposal. Nutrients (2019) 11. DOI: 10.3390/nu11112675 43. Clèries R., Ameijide A., Buxó M., Vilardell M., Martínez J.M., Font R., Marcos-Gragera R., Puigdemont M., Viñas G., Carulla M.. Ten-Year Probabilities of Death Due to Cancer and Cardiovascular Disease among Breast Cancer Patients Diagnosed in North-Eastern Spain. Int. J. Env. Res. Public Health (2023) 20. DOI: 10.3390/ijerph20010405 44. Nichols H.B., Trentham-Dietz A., Egan K.M., Titus-Ernstoff L., Holmes M.D., Bersch A.J., Holick C.N., Hampton J.M., Stampfer M.J., Willett W.C.. Body mass index before and after breast cancer diagnosis: Associations with all-cause, breast cancer, and cardiovascular disease mortality. Canc. Epidem. Preven. Biomark. (2009) 18 1403-1409. DOI: 10.1158/1055-9965.EPI-08-1094 45. Nechuta S., Chen W.Y., Cai H., Poole E.M., Kwan M.L., Flatt S.W., Patterson R.E., Pierce J.P., Caan B.J., Shu X.O.. A pooled analysis of post-diagnosis lifestyle factors in association with late estrogen-receptor-positive breast cancer prognosis. Int. J. Cancer J. Int. Du Cancer (2016) 138 2088-2097. DOI: 10.1002/ijc.29940 46. Battisti S., Guida F.M., Coppa F., Vaccaro D.M., Santini D., Tonini G., Zobel B.B., Semelka R.C.. Modification of Abdominal Fat Distribution After Aromatase Inhibitor Therapy in Breast Cancer Patients Visualized Using 3-D Computed Tomography Volumetry. Clin. Breast Cancer (2014) 14 365-370. DOI: 10.1016/j.clbc.2014.02.003 47. Zhu G., Zhang X., Wang Y., Xiong H., Zhao Y., Sun F.. Effects of exercise intervention in breast cancer survivors: A meta-analysis of 33 randomized controlled trails. OncoTargets Ther. (2016) 9 2153-2168. DOI: 10.2147/OTT.S97864 48. Thomas G.A., Cartmel B., Harrigan M., Fiellin M., Capozza S., Zhou Y., Ercolano E., Gross C.P., Hershman D., Ligibel J.. The effect of exercise on body composition and bone mineral density in breast cancer survivors taking aromatase inhibitors. Obesity (2017) 25 346-351. DOI: 10.1002/oby.21729 49. Dos Santos W.D.N., Vieira A., de Lira C., Mota J.F., Gentil P., Junior R.D.F., Battaglini C.L., Bottaro M., Vieira C.A.. Once a Week Resistance Training Improves Muscular Strength in Breast Cancer Survivors: A Randomized Controlled Trial. Integr. Cancer Ther. (2019) 18 1534735419879748. DOI: 10.1177/1534735419879748 50. 50. HHS Physical Activity Guidelines for Americans2nd ed.HHSWashington, DC, USA2018. Physical Activity Guidelines for Americans (2018) 51. Crespo-Salgado J.J., Delgado-Martín J.L., Blanco-Iglesias O., Aldecoa-Landesa S.. Guía básica de detección del sedentarismo y recomendaciones de actividad física en atención primaria. Aten. Prim. (2015) 47 175-183. DOI: 10.1016/j.aprim.2014.09.004 52. Garber C.E., Blissmer B., Deschenes M.R., Franklin B.A., LaMonte M.J., Lee I.-M., Nieman D.C., Swain D.P.. American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: Guidance for prescribing exercise. Med. Sci. Sport. Exerc. (2011) 43 1334-1359. DOI: 10.1249/MSS.0b013e318213fefb 53. Jones S.M.W., LaCroix A.Z., Li W., Zaslavsky O., Wassertheil-Smoller S., Weitlauf J., Brenes G.A., Nassir R., Ockene J.K., Caire-Juvera G.. Depression and quality of life before and after breast cancer diagnosis in older women from the Women’s Health Initiative. J. Cancer Surviv. (2015) 9 620-629. DOI: 10.1007/s11764-015-0438-y 54. Beasley J.M., Kwan M.L., Chen W.Y., Weltzien E.K., Kroenke C.H., Lu W., Nechuta S.J., Cadmus-Bertram L., Patterson R.E., Sternfeld B.. Meeting the physical activity guidelines and survival after breast cancer: Findings from the after breast cancer pooling project. Breast Cancer Res. Treat. (2012) 131 637-643. DOI: 10.1007/s10549-011-1770-1 55. Schmid D., Leitzmann M.F.. Association between physical activity and mortality among breast cancer and colorectal cancer survivors: A systematic review and meta-analysis. Ann. Oncol. (2014) 25 1293-1311. DOI: 10.1093/annonc/mdu012 56. George S.M., Irwin M.L., Smith A.W., Neuhouser M.L., Reedy J., McTiernan A., Alfano C.M., Bernstein L., Ulrich C.M., Baumgartner K.B.. Postdiagnosis diet quality, the combination of diet quality and recreational physical activity, and prognosis after early-stage breast cancer. Cancer Causes Control. (2011) 22 589-598. DOI: 10.1007/s10552-011-9732-9 57. Spei M.-E., Samoli E., Bravi F., La Vecchia C., Bamia C., Benetou V.. Physical activity in breast cancer survivors: A systematic review and meta-analysis on overall and breast cancer survival. Breast (2019) 44 144-152. DOI: 10.1016/j.breast.2019.02.001 58. Friedenreich C.M., Stone C.R., Cheung W.Y., Hayes S.C.. Physical Activity and Mortality in Cancer Survivors: A Systematic Review and Meta-Analysis. JNCI Cancer Spectr. (2019) 4 pkz080. DOI: 10.1093/jncics/pkz080 59. Rock C.L., Doyle C., Demark-Wahnefried W., Meyerhardt J., Courneya K.S., Schwartz A.L., Bandera E.V., Hamilton K.K., Grant B., McCullough M.. Nutrition and physical activity guidelines for cancer survivors. CA A Cancer J. Clin. (2012) 62 242-274. DOI: 10.3322/caac.21142 60. Fredriksen M., Halmøy A., Faraone S.V., Haavik J.. Long-term efficacy and safety of treatment with stimulants and atomoxetine in adult ADHD: A review of controlled and naturalistic studies. Eur. Neuropsychopharmacol. (2013) 23 508-527. DOI: 10.1016/j.euroneuro.2012.07.016 61. Nestel P.J., Mori T.A.. Dietary patterns, dietary nutrients and cardiovascular disease. Rev. Cardiovasc. Med. (2022) 23 17. DOI: 10.31083/j.rcm2301017 62. Brennan S.F., Woodside J.V., Lunny P.M., Cardwell C.R., Cantwell M.M.. Dietary fat and breast cancer mortality: A systematic review and meta-analysis. Crit. Rev. Food Sci. Nutr. (2017) 57 1999-2008. DOI: 10.1080/10408398.2012.724481 63. Chai W., Morimoto Y., Cooney R.V., Franke A.A., Shvetsov Y.B., Le Marchand L., Haiman C.A., Kolonel L.N., Goodman M.T., Maskarinec G.. Dietary Red and Processed Meat Intake and Markers of Adiposity and Inflammation: The Multiethnic Cohort Study. J. Am. Coll. Nutr. (2017) 36 378-385. DOI: 10.1080/07315724.2017.1318317 64. Givens D.I.. Review: Dairy foods, red meat and processed meat in the diet: Implications for health at key life stages. Animal (2018) 12 1709-1721. DOI: 10.1017/S1751731118000642 65. Mancini J.G., Filion K.B., Atallah R., Eisenberg M.J.. Systematic Review of the Mediterranean Diet for Long-Term Weight Loss. Am. J. Med. (2016) 129 407-415.e4. DOI: 10.1016/j.amjmed.2015.11.028 66. Sayon-Orea C., Martinez-Gonzalez M.A., Gea A., Alonso A., Pimenta A.M., Bes-Rastrollo M.. Baseline consumption and changes in sugar-sweetened beverage consumption and the incidence of hypertension: The SUN project. Clin. Nutr. (2015) 34 1133-1140. DOI: 10.1016/j.clnu.2014.11.010 67. Brandt P.A.V.D., Schulpen M.. Mediterranean diet adherence and risk of postmenopausal breast cancer: Results of a cohort study and meta-analysis. Int. J. Cancer (2017) 140 2220-2231. DOI: 10.1002/ijc.30654 68. Weigl J., Hauner H., Hauner D.. Can Nutrition Lower the Risk of Recurrence in Breast Cancer?. Breast Care (2018) 13 86-91. DOI: 10.1159/000488718 69. Buckland G., Travier N., Agudo A., Fonseca-Nunes A., Navarro C., Lagiou P., Demetriou C.A., Amiano P., Dorronsoro M., Chirlaque M.-D.. Olive oil intake and breast cancer risk in the Mediterranean countries of the European Prospective Investigation into Cancer and Nutrition study. Int. J. Cancer (2012) 131 2465-2469. DOI: 10.1002/ijc.27516 70. Castelló A., Boldo E., Pérez-Gómez B., Lope V., Altzibar J.M., Martín V., Castaño-Vinyals G., Guevara M., Dierssen-Sotos T., Tardón A.. Adherence to the Western, Prudent and Mediterranean dietary patterns and breast cancer risk: MCC-Spain study. Maturitas (2017) 103 8-15. DOI: 10.1016/j.maturitas.2017.06.020 71. Jochems S.H.J., Van Osch F.H.M., Bryan R., Wesselius A., Van Schooten F.J., Cheng K.K., Zeegers M.. Impact of dietary patterns and the main food groups on mortality and recurrence in cancer survivors: A systematic review of current epidemiological literature. BMJ Open (2017) 8 e014530. DOI: 10.1136/bmjopen-2016-014530 72. Ros E., Martínez-González M.A., Estruch R., Salas-Salvadó J., Fitó M., Martínez J.A., Corella D.. Mediterranean Diet and Cardiovascular Health: Teachings of the PREDIMED Study. Adv. Nutr. (2014) 5 330S-336S. DOI: 10.3945/an.113.005389 73. Toledo E., Salas-Salvadó J., Donat-Vargas C., Buil-Cosiales P., Estruch R., Ros E., Corella D., Fitó M., Hu F.B., Arós F.. Mediterranean Diet and Invasive Breast Cancer Risk Among Women at High Cardiovascular Risk in the PREDIMED Trial: A Randomized Clinical Trial. JAMA Intern. Med. (2015) 175 1752-1760. DOI: 10.1001/jamainternmed.2015.4838 74. Baena Ruiz R., Salinas Hernández P.. Cancer chemoprevention by dietary phytochemicals: Epidemiological evidence. Maturitas (2016) 94 13-19. DOI: 10.1016/j.maturitas.2016.08.004 75. Couto E., Sandin S., Löf M., Ursin G., Adami H.-O., Weiderpass E.. Mediterranean Dietary Pattern and Risk of Breast Cancer. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0055374 76. de Cicco P., Catani M.V., Gasperi V., Sibilano M., Quaglietta M., Savini I.. Nutrition and breast cancer: A literature review on prevention, treatment and recurrence. Nutrients (2019) 11. DOI: 10.3390/nu11071514 77. Psaltopoulou T., Kosti R.I., Haidopoulos D., Dimopoulos M., Panagiotakos D.B.. Olive oil intake is inversely related to cancer prevalence: A systematic review and a meta-analysis of 13800 patients and 23340 controls in 19 observational studies. Lipids Health Dis. (2011) 10 127. DOI: 10.1186/1476-511X-10-127 78. Savanelli M.C., Barrea L., Macchia P.E., Savastano S., Falco A., Renzullo A., Scarano E., Nettore I.C., Colao A., Di Somma C.. Preliminary results demonstrating the impact of Mediterranean diet on bone health. J. Transl. Med. (2017) 15 81. DOI: 10.1186/s12967-017-1184-x 79. Sieri S., Chiodini P., Agnoli C., Pala V., Berrino F., Trichopoulou A., Benetou V., Vasilopoulou E., Sánchez M.-J., Chirlaque M.-D.. Dietary Fat Intake and Development of Specific Breast Cancer Subtypes. Gynecol. Oncol. (2014) 106 dju068. DOI: 10.1093/jnci/dju068 80. Ruiz E., Ávila J.M., Valero T., Del Pozo S., Rodriguez P., Aranceta-Bartrina J., Gil Á., González-Gross M., Ortega R.M., Serra-Majem L.. Energy Intake, Profile, and Dietary Sources in the Spanish Population: Findings of the ANIBES Study. Nutrients (2015) 7 4739-4762. DOI: 10.3390/nu7064739 81. Fassier P., Zelek L., Lécuyer L., Bachmann P., Touillaud M., Druesne-Pecollo N., Galan P., Cohen P., Hoarau H., Latino-Martel P.. Modifications in dietary and alcohol intakes between before and after cancer diagnosis: Results from the prospective population-based NutriNet-Santé cohort. Int. J. Cancer (2017) 141 457-470. DOI: 10.1002/ijc.30704 82. Song D., Deng Y., Liu K., Zhou L., Li N., Zheng Y., Hao Q., Yang S., Wu Y., Zhai Z.. Vitamin D intake, blood vitamin D levels, and the risk of breast cancer: A dose-response meta-analysis of observational studies. Aging (2019) 11 12708-12732. DOI: 10.18632/aging.102597 83. Vance V., Campbell S., McCargar L., Mourtzakis M., Hanning R.. Dietary changes and food intake in the first year after breast cancer treatment. Appl. Physiol. Nutr. Metab. (2014) 39 707-714. DOI: 10.1139/apnm-2013-0400 84. Wayne S.J., Lopez S.T., Butler L.M., Baumgartner K.B., Baumgartner R.N., Ballard-Barbash R.. Changes in dietary intake after diagnosis of breast cancer. J. Am. Diet. Assoc. (2004) 104 1561-1568. DOI: 10.1016/j.jada.2004.07.028 85. Maróti-Nagy Á., Paulik E., Thurzó L.. Az egészségügyi személyzet életmód-tanácsadó tevékenységének szerepe daganatos betegséggel kezelt nők táplálkozási szokásainak megváltoztatásában. Magy. Onkol. (2010) 54 41-45. DOI: 10.1556/MOnkol.54.2010.1.6 86. Coa K.I., Smith K.C., Klassen A.C., Caulfield L.E., Helzlsouer K., Peairs K., Shockney L.. Capitalizing on the “teachable moment” to promote healthy dietary changes among cancer survivors: The perspectives of health care providers. Support. Care Cancer (2015) 23 679-686. DOI: 10.1007/s00520-014-2412-z 87. Moreiras O., Carbajal A., Cabrera L.C.C.. Ingestas recomendadas de energia y nutrientes para la poblacion espanola. Ediciones Piramide Grupo Anaya SE (ed) Tablas de Composicion de Alimentos (2017) 258-259 88. Ruiz E., Ávila J.M., Valero T., Del Pozo S., Rodriguez P., Aranceta-Bartrina J., Gil Á., González-Gross M., Ortega R.M., Serra-Majem L.. Macronutrient Distribution and Dietary Sources in the Spanish Population: Findings from the ANIBES Study. Nutrients (2016) 8. DOI: 10.3390/nu8030177 89. Chen P., Hu P., Xie D., Qin Y., Wang F., Wang H.. Meta-analysis of vitamin D, calcium and the prevention of breast cancer. Breast Cancer Res. Treat. (2010) 121 469-477. DOI: 10.1007/s10549-009-0593-9 90. Kim Y., Je Y.. Vitamin D intake, blood 25(OH)D levels, and breast cancer risk or mortality: A meta-analysis. Br. J. Cancer (2014) 110 2772-2784. DOI: 10.1038/bjc.2014.175 91. Tramm R., Mccarthy A.L., Yates P.. Dietary modification for women after breast cancer treatment: A narrative review. Eur. J. Cancer Care (2011) 20 294-304. DOI: 10.1111/j.1365-2354.2011.01238.x 92. Essén A., Santaolalla A., Garmo H., Hammar N., Walldius G., Jungner I., Malmström H., Holmberg L., Van Hemelrijck M.. Baseline serum folate, vitamin B12 and the risk of prostate and breast cancer using data from the Swedish AMORIS cohort. Cancer Causes Control. (2019) 30 603-615. DOI: 10.1007/s10552-019-01170-6 93. Matejcic M., de Batlle J., Ricci C., Biessy C., Perrier F., Huybrechts I., Weiderpass E., Boutron-Ruault M., Cadeau C., His M.. Biomarkers of folate and vitamin B12 and breast cancer risk: Report from the EPIC cohort. Int. J. Cancer (2017) 140 1246-1259. DOI: 10.1002/ijc.30536 94. Foglietta J., Inno A., de Iuliis F., Sini V., Duranti S., Turazza M., Tarantini L., Gori S.. Cardiotoxicity of Aromatase Inhibitors in Breast Cancer Patients. Clin. Breast Cancer (2017) 17 11-17. DOI: 10.1016/j.clbc.2016.07.003 95. Freisling H., Noh H., Slimani N., Chajès V., May A.M., Peeters P.H., Weiderpass E., Cross A.J., Skeie G., Jenab M.. Nut intake and 5-year changes in body weight and obesity risk in adults: Results from the EPIC-PANACEA study. Eur. J. Nutr. (2018) 57 2399-2408. DOI: 10.1007/s00394-017-1513-0
--- title: Outbred Mice with Streptozotocin-Induced Diabetes Show Sex Differences in Glucose Metabolism authors: - Boyoung Kim - Eun-Sun Park - Jong-Sun Lee - Jun-Gyo Suh journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049093 doi: 10.3390/ijms24065210 license: CC BY 4.0 --- # Outbred Mice with Streptozotocin-Induced Diabetes Show Sex Differences in Glucose Metabolism ## Abstract Outbred mice (ICR) with different genotypes and phenotypes have been reported to be more suitable for scientific testing than inbred mice because they are more similar to humans. To investigate whether the sex and genetic background of the mice are important factors in the development of hyperglycemia, we used ICR mice and divided them into male, female, and ovariectomized female (FOVX) groups and treated them with streptozotocin (STZ) for five consecutive days to induce diabetes. Our results show that fasting blood glucose and hemoglobin A1c (HbA1c) levels were significantly higher in diabetes-induced males (M-DM) and ovariectomized diabetes-induced females (FOVX-DM) than in diabetes-induced females (F-DM) at 3 and 6 weeks after STZ treatment. Furthermore, the M-DM group showed the most severe glucose tolerance, followed by the FOVX-DM and F-DM groups, suggesting that ovariectomy affects glucose tolerance in female mice. The size of pancreatic islets in the M-DM and FOVX-DM groups was significantly different from that of the F-DM group. The M-DM and FOVX-DM groups had pancreatic beta-cell dysfunction 6 weeks after STZ treatment. Urocortin 3 and somatostatin inhibited insulin secretion in the M-DM and FOVX-DM groups. Overall, our results suggest that glucose metabolism in mice is dependent on sex and/or genetic background. ## 1. Introduction Rodents, especially mice, are similar to humans at the genetic, anatomical, physiological, and pathophysiological levels, and can therefore replace humans for scientific testing [1]. Laboratory mice are the most commonly used experimental animals for understanding biological functions and translating them to humans in biomedical research [2]. Laboratory mice can be divided into inbred and outbred groups. For decades, inbred mouse strains have been more frequently used than outbred mouse strains because the first, which are genetically identical, have less phenotypic diversity than outbred mice (ICR) [3,4]. Therefore, inbred mice are usually selected for immunological, population genetic mapping, and molecular genetic studies because they are genetically stable, have almost the same phenotype, and have clean genetic background information [5]. Nevertheless, some studies suggest that ICR mice with different genotypes and phenotypes are more similar to humans than inbred mice. In research by Shan Wang et al., it was confirmed that inbred and outbred strains show different responses to metabolic enzymes and agonists [6]. In an inflammatory disease study, ICR mice were found to have greater variability in the phenotypic progression of immune cells compared with inbred mice [7]. These studies suggest that the common cause of the above results is the influence of the different genetic backgrounds of the two strains, which is a key factor to consider in biomedical studies. Outbred strains are more suitable for drug or therapeutic research in humans [8]. In a previous study, we induced diabetes with streptozotocin (STZ) in C57BL/6J mice, which are preferentially used in studies with inbred mouse strains [9]. We confirmed that there were phenotypic differences according to sex in the diabetic groups. For instance, fasting blood glucose (FBG) and hemoglobin A1c (HbA1c) levels were significantly higher in male diabetic mice than in female diabetic mice. These results suggest that glucose metabolism is more affected by STZ in male diabetic mice than in female diabetic mice. In this study, we further investigated whether the sex and genetic background of mice are important factors for the development of hyperglycemia. To test our hypothesis, ICR mice, which have high genetic diversity within the strain, were used to develop hyperglycemia by STZ injection over 5 days. FBG and HbA1c levels were higher in diabetic male mice and ovariectomized diabetic female mice than in diabetic female mice. In addition, pancreatic beta-cell dysfunction was more severe in male diabetic mice than in female diabetic mice. The expression of somatostatin increased in diabetic male and ovariectomized diabetic female mice through delta-cell-mediated regulation. These results demonstrate that glucose metabolism is dependent on sex and/or genetic background in mice. ## 2.1. STZ-Treated Male Mice Rapidly Developed Hyperglycemia and Diabetes FBG levels in STZ-treated diabetic male mice (M-DM) were significantly higher than in the M-control group from 1 to 6 weeks (except for weeks 4 and 5) after STZ treatment (Figure 1A). FBG levels were also significantly higher in the FOVX-DM group than in the FOVX-ctrl group at 3–6 weeks (Figure 1B). However, there was no significant difference in FBG levels between the F-DM and F-ctrl groups (Figure 1C). FBG levels increased sequentially in the M-DM, FOVX-DM, and F-DM groups. These results indicate that the FOVX-DM group was more sensitive to STZ treatment than the F-DM group. At weeks 3 and 6 after STZ treatment, HbA1c levels were significantly higher in the M-DM and FOVX-DM groups than in the control group, whereas the F-DM group had no significant difference from the control group (Figure 1D). In addition, HbA1c levels were significantly higher in the M-DM group than in the F-DM group. HbA1c levels in the FOVX-DM group were also significantly higher than in the F-DM group 6 weeks after STZ treatment. These results show that the FOVX-DM group is more similar to the M-DM than the F-DM group. ## 2.2. STZ-Treated Male Mice Showed the Strongest Glucose Tolerance The OGTT was performed 3 and 6 weeks after STZ treatment to confirm whether glucose tolerance was induced. At week 3, FBG levels in the M-DM, FOVX-DM, and F-DM groups were significantly higher than those in each control group 30, 60, and 120 min after glucose administration (Figure 2A). Similarly, AUC for blood glucose was significantly higher in all diabetic groups than in each control group at week 3. The AUC in the M-DM and FOVX-DM groups was significantly higher than that in the F-DM group. In addition, the AUC in the M-DM group was significantly higher than that in the FOVX-DM group 3 weeks after STZ treatment (Figure 2B). At week 6, glucose levels showed a similar pattern to that at week 3 (Figure 2C), and the AUC of the M-DM and FOVX-DM groups was significantly higher than that of the F-DM group at week 6. However, there was no significant difference between the M-DM and FOVX-DM groups (Figure 2D). Taken together, these results suggest that ovariectomy affects glucose tolerance in female mice. ## 2.3. The Size of Pancreatic Islets in M-DM and FOVX-DM Was Reduced by STZ Treatment Morphological evaluation was performed to confirm the size of the pancreatic islets after STZ treatment using hematoxylin–eosin staining. There was no significant difference between the DM and control groups 3 weeks after STZ treatment. In contrast, the size of the pancreatic islets was significantly decreased in the diabetic groups at week 6 compared with the control groups, except in the female group (Figure 3A,B). These results show that the pancreatic islets in the M-DM and FOVX-DM groups were severely damaged over time because they were more affected by STZ treatment. ## 2.4. The Cell Composition of Pancreatic Islets Was Altered in the M-DM and FOVX-DM Groups after STZ Treatment To identify the cell composition of the pancreatic islets, α- and β-cells were confirmed by immunofluorescence staining. The α-cells, which produce glucagon, are mainly located at the outer edge of the pancreatic islets. The β-cells, producing insulin, are located in the center of the pancreatic islets. The α-cells shifted from the outer border to the center of the pancreatic islets in diabetic mice 6 weeks after STZ treatment (Figure 4A). The number of α- and β-cells in the M-DM and FOVX-DM groups was significantly higher than that in the control group at week 6, whereas the number in the F-DM group was not significantly different from that in the F-ctrl group (Figure 4B). The glucagon/insulin ratios in the M-DM and FOVX-DM groups were significantly higher than in the control group (Figure 4C). These results reveal that the number of α-cells in pancreatic islets was increased by STZ treatment. ## 2.5. Groups M-DM and FOVX-DM Had Impaired Pancreatic Beta-Cell Function Pancreatic beta-cell function (%B) and insulin sensitivity (%S) were calculated from FBG and plasma insulin levels. The %B of the M-DM and FOVX-DM groups was significantly lower than that of the F-DM group at weeks 3 and 6. Moreover, the %B of the M-DM and FOVX-DM groups was significantly lower than that of the control group at 6 weeks (Figure 5A). The %S in the M-DM group was significantly lower than that in the F-DM group 3 weeks after STZ treatment. The %S of the F-DM group was significantly higher than that of the FOVX-DM group at 6 weeks (Figure 5B). The HOMA-IR in the M-DM and FOVX-DM groups tended to be increased compared with the control groups at weeks 3 and 6 (Figure 5C). Overall, %B and %S were more pronounced in the M-DM group than in the FOVX-DM group. ## 2.6. Insulin Secretion in the M-DM and FOVX-DM Groups Was Inhibited by Urocortin 3 and Somatostatin To confirm the damage to β-cells induced by STZ treatment, the expression of urocortin 3 (UCN3), a marker of mature β-cells, was measured. In the control groups (M-ctrl, FOVX-ctrl, and F-ctrl), 60–$80\%$ of β-cells expressed UCN3 at weeks 3 and 6 after STZ treatment (Figure 6A–C). This ratio was significantly lower in the STZ-treated groups than in the control groups at week 6 (Figure 6C). Somatostatin, produced by pancreatic delta cells, prevents the secretion of pancreatic hormones such as insulin and glucagon. The intensity of somatostatin was significantly increased in the M-DM and FOVX-DM groups com-pared with the control group at week 6. These results indicate that the secretion of insulin and glucagon was inhibited in the M-DM and FOVX-DM groups compared with the F-DM group (Figure 6D). ## 3. Discussion Type 2 diabetes is a disorder of glucose metabolism, with a metabolic phenotype that differs according to sex [10]. Our group reported that male inbred mice (C57BL/6J) had higher hyperglycemia than female mice after STZ injection [9]. In this study, we investigated whether the genetic background of the mice was important for the development of hyperglycemia and whether female hormones such as estrogen play a role in glucose metabolism after STZ injection. To test our hypothesis, ICR mice, which have high genetic diversity within the strain, were used to develop hyperglycemia by STZ injection over 5 days. FBG and HbA1c levels were significantly increased in the M-DM and FOVX-DM groups compared with the control groups, whereas the F-DM group showed no significant difference compared with the control group. Additionally, the levels of M-DM and FOVX-DM were significantly higher than those of F-DM (Figure 1). In this study, all DM groups developed glucose tolerance, regardless of sex. Notably, glucose tolerance in the M-DM group was significantly different from that in the FOVX-DM and F-DM groups 3 weeks after STZ treatment. At 6 weeks, glucose tolerance in the M-DM group was not significantly different from that in the FOVX-DM group, indicating that ovariectomy might affect glucose tolerance in female mice (Figure 2). These results indicate that female ICR mice did not exhibit hyperglycemia under the same conditions in which male mice developed diabetes. In addition, the phenotype of glucose metabolism in ovariectomized female mice was similar to that of male mice. Our results show that female hormones are a principal factor in the development of hyperglycemia after STZ injection. In contrast, Kim et al. [ 9] reported that inbred female mice (C57BL/6J) developed hyperglycemia under the same conditions. These different results could be due to genetic differences between the inbred and outbred mice. The size of the pancreatic islets was significantly smaller in the M-DM and FOVX-DM groups than in the control group (Figure 3). In the M-DM and FOVX-DM groups, the intensities of α- and β-cells were significantly increased compared with the intensity observed in each control group at week 6 after STZ treatment, whereas it was not significantly different in the F-DM group compared with the F-ctrl group. The glucagon/insulin ratio was significantly higher in the diabetic group than in the control group, except in the female group (Figure 4). The β-cell functions were significantly decreased in the M-DM and FOVX-DM groups compared with the control group at week 6 after STZ treatment. Insulin sensitivity was significantly lower in the FOVX-DM group than in the F-DM group (Figure 5). Remedi et al. [ 11] reported that pancreatic cells changed due to glucose toxicity when hyperglycemia was continuously maintained. Therefore, the size of pancreatic islets in M-DM and FOVX-DM groups decreased due to the death of β-cells, indicating that cell function deteriorated. Taken together, the morphological characteristics of pancreatic islets in M-DM, FOVX-DM, and F-DM were consistent with the results of diabetic parameters such as hyperglycemia and glucose tolerance. The β-cells expressing urocortin 3 (UCN3) were significantly decreased in the M-DM, FOVX-DM, and F-DM groups compared with the control groups at week 6 after STZ treatment. In addition, the intensity of somatostatin was significantly increased in the M-DM and FOVX-DM groups compared with the F-DM group at week 6, indicating that the insulin secretion was inhibited in the M-DM and FOVX-DM groups compared with control groups (Figure 6). UCN3 is expressed in murine β-cells, but not in α- and δ-cells [12]. UCN3 is the most appropriate marker for β-cell maturation [13,14,15]. In this study, the expression of UCN3 was significantly lower in the diabetic group than in the control group. This result suggests that the number of functionally mature β-cells was reduced by STZ injection. UCN3 stimulates somatostatin secretion in δ-cells [16] and regulates the negative feedback of somatostatin to decrease insulin secretion [17]. As the expression of the somatostatin was significantly decreased in the diabetic groups, the intensity of glucagon and insulin was significantly increased in the pancreatic islets of these groups, but not in the female diabetic group. Researchers have reported on how estrogen protects the beta cells of the pancreas from damage caused by STZ [18,19]. Overall, these results suggest that female hormones play an important role in the regulation of glucose metabolism in the STZ-induced diabetic mouse model. In conclusion, ICR mice showed different glucose metabolism characteristics compared with inbred mice (C57BL/6J) that, under the same conditions, developed diabetes after STZ injection. In addition, female hormones play an important role in regulating glucose metabolism, regardless of the genetic background of the mice. These results provide important information for the development of animal models and efficacy testing of drugs or functional foods for the treatment of diabetes. ## 4.1. Experimental Animals and Streptozotocin (STZ) Treatment Five-week-old ICR mice were obtained from Orientbio (Seoul, Republic of Korea) and had an adaptation period of 1 week. The mice were fed a chow diet (Purina, Seongnam-si, Republic of Korea) and water ad libitum. At 6 weeks of age, ovariectomy was performed to establish a postmenopausal female mouse model, followed by a recovery period of 2 weeks. Mice were divided into the following six groups: [1] male control (M-ctrl, $$n = 10$$), [2] STZ-treated diabetic males (M-DM, $$n = 12$$), [3] female control (F-ctrl, $$n = 10$$), [4] STZ-treated diabetic females (F-DM, $$n = 12$$), [5] ovariectomized female control (FOVX-ctrl, $$n = 15$$), and [6] ovariectomized STZ-treated diabetic females (FOVX-DM, $$n = 16$$). To induce type 2 diabetes, mice were continuously treated with STZ (40 mg/kg, Sigma, St. Louis, MO, USA) in 50 mM citrate buffer (pH 4.5) by intraperitoneal injection for 5 days at 8 weeks of age [9,20]. During STZ administration, mice were simultaneously administered a $10\%$ sucrose solution to prevent hypoglycemic shock. Mice were reared in an animal care facility at a temperature of 22 ± 2 °C, a humidity of 55 ± $10\%$, and a 12 h light–dark cycle (8 am–8 pm). ## 4.2. Measurement of FBG and HbA1c To measure FBG levels from 0 to 6 weeks after STZ treatment, mice were fasted for 6 h, and blood was collected from the retro-orbital plexus using a heparin capillary tube. FBG levels were measured using blood glucose meters (ACCU-CHEK; Roche, Indianapolis, IN, USA). Hemoglobin A1c levels, indicating the average blood glucose level over the past 2–3 months, were measured at 0, 3, and 6 weeks after STZ treatment. Whole blood collected in a heparin tube (BD Microtainer, Franklin Lakes, NJ, USA) was analyzed using an HbA1c analyzer (CareSURETM Analyzer 100, Wellsbio, Seoul, Republic of Korea) and dedicated cartridges (CareSURETM A1c cartridge, Wellsbio, Seoul, Republic of Korea). ## 4.3. Oral Glucose Tolerance Test (OGTT) and Area under the Curve (AUC) To confirm whether glucose tolerance was induced in the diabetic mice, an OGTT was performed at weeks 3 and 6. All mice were fasted for 6 h and received an oral gavage injection of 2 g/kg D-glucose (Sigma-Aldrich, St. Louis, MO, USA) in saline. Blood glucose was measured using blood glucose meters (ACCU-CHEK, Roche, Indianapolis, IN, USA) at 0, 30, 60, and 120 min after D-glucose treatment. Then, the AUC was calculated as the geometric mean value based on the area under the OGTT curve. ## 4.4. Histological Analysis of the Pancreas After the experiment, mice were anesthetized with isoflurane (flowmeter, 100–200 mL/min, vaporizer 2–$3\%$) and subjected to transcardiac perfusion with $4\%$ paraformaldehyde. The pancreas was dissected, embedded in paraffin blocks, and cut into 5 µm slices with a microtome (Reichert-Jung 2030 Biocut, Nußloch, Germany). To determine the size of the pancreatic islets, images were captured with an optical microscope (Diaphot 300, Nikon, Tokyo, Japan). The size was measured using Image J software version 1.8.0_172 (NIH, Stapleton, NY, USA). For immunofluorescence evaluation, slides were blocked with $10\%$ horse serum in Tris-buffered saline (TBS) and stained with anti-insulin (1:200, Invitrogen, Waltham, MA, USA), anti-glucagon (1:200, Cell Signaling, Denver, MA, USA), anti-somatostatin (1:100, Invitrogen, Waltham, MA, USA), and anti-urocortin III (1:200, Phoenix Pharmahandel, Mannheim, Germany) overnight at room temperature. On the second day, slides were washed with TBS and incubated with donkey anti-guinea pig Alexa Fluor 594 (1:200; Jackson ImmunoResearch, West Grove, PA, USA), donkey anti-rabbit Alexa Fluor 488 (1:100; Abcam, Cambridge, UK), and donkey anti-rat Alexa Fluor 647 (1:200; Abcam, Cambridge, UK) antibodies. After staining, islets were imaged using confocal microscopy (LSM 710, Carl Zeiss, Jena, Germany), and the intensities of insulin, glucagon, and somatostatin were measured using ZEISS Blue software (version 11). ## 4.5. Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) Insulin levels were measured in all plasma samples and analyzed using a microplate reader (VersaMaxTM, Molecular Devices, San Jose, CA, USA). A public HOMA2 calculator (https://www.dtu.ox.ac.uk/homacalculator/ accessed on 15 March 2022) was used to assess pancreatic beta-cell function (%B), insulin sensitivity (%S), and insulin resistance. ## 4.6. Statistical Analysis All data are expressed as mean ± standard error of the mean. Normally distributed data were tested for homogeneity of variance using Levene’s test. Data were analyzed using an independent two-sample t-test when the data followed an equal variance. Statistical significance was set at * $p \leq 0.05$, $p \leq 0.01$, * $p \leq 0.001$, and ** $p \leq 0.0001.$ All results were analyzed using IBM SPSS Statistics 25 (SPSS Inc., Chicago, IL, USA). ## References 1. Bryda E.C.. The Mighty Mouse: The impact of rodents on advances in biomedical research*. Mo. Med.* (2013) 110 207-211. PMID: 23829104 2. Taylor K., Gordon N., Langley G., Higgins W.. Estimates for worldwide laboratory animal use in 2005. Altern. Lab. Anim. (2008) 36 327-342. DOI: 10.1177/026119290803600310 3. Festing M.F.. Evidence should trump intuition by preferring inbred strains to outbred stocks in preclinical research. ILAR J. (2014) 55 399-404. DOI: 10.1093/ilar/ilu036 4. Festing M.F.. Phenotypic variability of inbred and outbred mice. Nature (1976) 263 230-232. DOI: 10.1038/263230a0 5. Tuttle A.H., Philip V.M., Chesler E.J., Mogil J.S.. Comparing phenotypic variation between inbred and outbred mice. Nat. Methods (2018) 15 994-996. DOI: 10.1038/s41592-018-0224-7 6. Wang S., Chen L., Wang Q., He Z., Chen S., Zhang H., Li H., Guo P., Li Q., Zhang R.. Strain differences between CD-1 and C57BL/6 mice in expression of metabolic enzymes and DNA methylation modifications of the primary hepatocytes. Toxicology (2019) 412 19-28. DOI: 10.1016/j.tox.2018.11.011 7. Martin M.D., Danahy D.B., Hartwig S.M., Harty J.T., Badovinac V.P.. Revealing the Complexity in CD8 T Cell Responses to Infection in Inbred C57B/6 versus Outbred Swiss Mice. Front. Immunol. (2017) 8 1527. DOI: 10.3389/fimmu.2017.01527 8. Enriquez J., Mims B.M.D., Trasti S., Furr K.L., Grisham M.B.. Genomic, microbial and environmental standardization in animal experimentation limiting immunological discovery. BMC Immunol. (2020) 21. DOI: 10.1186/s12865-020-00380-x 9. Kim B., Kim Y.Y., Nguyen P.T.-T., Nam H., Suh J.G.. Sex differences in glucose metabolism of streptozotocin-induced diabetes inbred mice (C57BL/6J). Appl. Biol. Chem. (2020) 63 59. DOI: 10.1186/s13765-020-00547-5 10. De Paoli M., Werstuck G.H.. Role of Estrogen in Type 1 and Type 2 Diabetes Mellitus: A Review of Clinical and Preclinical Data. Can. J. Diabetes (2020) 44 448-452. DOI: 10.1016/j.jcjd.2020.01.003 11. Remedi M.S., Emfinger C.. Pancreatic beta-cell identity in diabetes. Diabetes Obes. Metab. (2016) 18 110-116. DOI: 10.1111/dom.12727 12. Li C., Chen P., Vaughan J., Blount A., Chen A., Jamieson P.M., Rivier J., Smith M.S., Vale W.. Urocortin III is expressed in pancreatic beta-cells and stimulates insulin and glucagon secretion. Endocrinology (2003) 144 3216-3224. DOI: 10.1210/en.2002-0087 13. Van der Meulen T., Xie R., Kelly O.G., Vale W.W., Sander M., Huising M.O.. Urocortin 3 marks mature human primary and embryonic stem cell-derived pancreatic alpha and beta cells. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0052181 14. Van der Meulen T., Huising M.O.. Role of transcription factors in the transdifferentiation of pancreatic islet cells. J. Mol. Endocrinol. (2015) 54 R103-R117. DOI: 10.1530/JME-14-0290 15. Blum B., Roose A.N., Barrandon O., Maehr R., Arvanites A.C., Davidow L.S., Davis J.C., Peterson Q.P., Rubin L.L., Melton D.A.. Reversal of beta cell de-differentiation by a small molecule inhibitor of the TGFbeta pathway. Elife (2014) 3 e02809. DOI: 10.7554/eLife.02809 16. Li C., Chen P., Vaughan J., Lee K.F., Vale W.. Urocortin 3 regulates glucose-stimulated insulin secretion and energy homeostasis. Proc. Natl. Acad. Sci. USA (2007) 104 4206-4211. DOI: 10.1073/pnas.0611641104 17. Van der Meulen T., Donaldson C.J., Caceres E., Hunter A.E., Cowing-Zitron C., Pound L.D., Adams M.W., Zembrzycki A., Grove K.L., Huising M.O.. Urocortin3 mediates somatostatin-dependent negative feedback control of insulin secretion. Nat. Med. (2015) 21 769-776. DOI: 10.1038/nm.3872 18. May C.L., Chu K., Hu M., Ortega C.S., Simpson E.R., Korach K.S., Tsai M., Mauvais-Jarvis F.. Estrogens protect pancreatic beta-cells from apoptosis and prevent insulin-deficient diabetes mellitus in mice. Proc. Natl. Acad. Sci. USA (2006) 103 9232-9237. DOI: 10.1073/pnas.0602956103 19. Mauvais-Jarvis F., May C.L., Tiano J.P., Liu S., Kilic-Berkman G., Kim J.H.. The Role of Estrogens in Pancreatic Islet Physiopathology. Adv. Exp. Med. Biol. (2017) 1043 385-399. DOI: 10.1007/978-3-319-70178-3_18 20. Furman B.L.. Streptozotocin-Induced Diabetic Models in Mice and Rats. Curr. Protoc. Pharmacol. (2015) 70 e78. DOI: 10.1002/0471141755.ph0547s70
--- title: 'A Data-Mining Approach to Identify NF-kB-Responsive microRNAs in Tissues Involved in Inflammatory Processes: Potential Relevance in Age-Related Diseases' authors: - Luigina Micolucci - Giulia Matacchione - Maria Cristina Albertini - Massimo Marra - Deborah Ramini - Angelica Giuliani - Jacopo Sabbatinelli - Antonio Domenico Procopio - Fabiola Olivieri - Annalisa Marsico - Vladia Monsurrò journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049099 doi: 10.3390/ijms24065123 license: CC BY 4.0 --- # A Data-Mining Approach to Identify NF-kB-Responsive microRNAs in Tissues Involved in Inflammatory Processes: Potential Relevance in Age-Related Diseases ## Abstract The nuclear factor NF-kB is the master transcription factor in the inflammatory process by modulating the expression of pro-inflammatory genes. However, an additional level of complexity is the ability to promote the transcriptional activation of post-transcriptional modulators of gene expression as non-coding RNA (i.e., miRNAs). While NF-kB’s role in inflammation-associated gene expression has been extensively investigated, the interplay between NF-kB and genes coding for miRNAs still deserves investigation. To identify miRNAs with potential NF-kB binding sites in their transcription start site, we predicted miRNA promoters by an in silico analysis using the PROmiRNA software, which allowed us to score the genomic region’s propensity to be miRNA cis-regulatory elements. A list of 722 human miRNAs was generated, of which 399 were expressed in at least one tissue involved in the inflammatory processes. The selection of “high-confidence” hairpins in miRbase identified 68 mature miRNAs, most of them previously identified as inflammamiRs. The identification of targeted pathways/diseases highlighted their involvement in the most common age-related diseases. Overall, our results reinforce the hypothesis that persistent activation of NF-kB could unbalance the transcription of specific inflammamiRNAs. The identification of such miRNAs could be of diagnostic/prognostic/therapeutic relevance for the most common inflammatory-related and age-related diseases. ## 1. Introduction The nuclear factor (NF)-kB is a transcription factor (TF) activated by an evolutionarily conserved inflammatory signaling, induced by a wide range of external and internal danger signals [1,2,3]. The complex modulation of this signaling can be envisaged considering the different activation strategies, well known as “canonical” and “non-canonical” NF-kB activation signaling (reviewed in [4,5]). A fine-tuning activation of NF-kB promotes the expression of pro-inflammatory genes and participates in the regulation of survival, activation, and differentiation of innate immune cells and T cells [6]. On the contrary, a persistent activation of NF-kB signaling was described in conditions of cellular senescence and organismal aging, as well as in patients affected by the most common age-related degenerative diseases (ARDs) [7,8,9]. Many efforts have been made to understand which pathways are regulated by NF-kB and how the NF-kB pathway itself is modulated [10,11]. It has become clear that not only TFs but also a series of epigenetic factors, including non-coding microRNAs (miRNAs), are involved in the regulation of almost all the human transcriptional programs, both as inhibitors of mRNAs translation and as enhancers of mRNAs transcription [12,13,14]. Increasing evidence confirmed that these epigenetic factors play key roles in the development and progression of the most common human ARDs [15,16]. Regarding the canonical pathway of miRNA processing, that regulates gene expression at the post-transcriptional level, a primary transcript called pri-miRNA is cleaved to a precursor miRNA hairpin structure (pre-miRNA) in the nucleus by the Drosha/Pasha complex and transported into the cytoplasm, where the pre-miRNA is further processed into a miRNA:miRNA* duplex [17]. After being separated, the mature miRNA is loaded into the Argonaute 2 (Ago 2) containing RNA-induced silencing complexes (RISCs) and drives it to regulate its target mRNAs [17]. On one hand, a few miRNAs targeting mRNAs belonging to NF-kB pathway have already been identified, highlighting the activation of feedback loops aimed to restrain the inflammatory process triggered by NF-kB. Notably, some miRNAs involved in these feedback circuits were identified as deregulated in ARDs [5,18,19,20,21]. On the other hand, the full elucidation of miRNA biogenesis would be of paramount importance to identify their regulators and the role they might play in complex regulatory networks. Even if computational models were extensively applied to disentangle the complex effects of non-coding RNA in human diseases [22], for a long time, the difficulty of experimentally detecting miRNA promoters has limited the ability to identify the NF-kB binding sites in DNA sequences coding for miRNAs. However, the annotation of miRNA promoters, using high-throughput genomic data, allowed us to partially overcome this difficulty [23]. As important transcriptional regulators, miRNAs can upregulate or downregulate many target genes involved in the NF-kB signaling pathway via negative or positive feedback loops, and are involved in several human diseases, too, including the recent COVID-19 pandemic [24,25]. Since it is conceivable that age-related NF-kB activation could induce an overexpression of NF-kB responsive miRNAs, the identification of such miRNAs, and their targeted mRNAs and pathways, could contribute to clarifying the complex mechanisms that modulate healthy or unhealthy aging trajectories. In this work, we aimed to: (i) identify all human miRNAs potentially modulated by NF-kB, (ii) select and characterize those NF-kB-responsive miRNAs that are specifically expressed in healthy tissues involved in the modulation of the inflammatory processes (such as cells of the immune system, liver, blood, and bone marrow), (iii) discover their targeted mRNAs and relative pathways, and finally (iv) evaluate the involvement of such pathways in the development of human diseases, including ARDs. ## 2.1. Putative NF-kB Responsive miRNAs To select NF-kB responsive miRNAs, we analyzed the PROmiRNA database [23], FANTOM4 Libraries [26], “High confidence hairpins” in miRbase [27], and “Human expression dataset” [28], following the data-mining process highlighted in the data flow diagram in Figure 1. We analyzed primarily genome-wide PROmiRNA predictions, as well as TF-binding site predictions as reported in [23], to identify miRNAs with potential NF-kB binding sites in their promoter sequences. PROmiRNA is a miRNA promoter recognition method, based on a semi-supervised statistical model trained on multi-tissue deepCAGE FANTOM4 libraries and other sequence features. It is tailored to score the potential of CAGE-enriched genomic regions to be promoters of either intergenic or intragenic miRNAs, thereby modulating miRNA expression in a tissue-specific manner [23]. To identify the TFs that regulate specific miRNAs, for each predicted miRNA transcription start site (TSS), we retrieved the 1 kb centered on it and used the TRAP approach [29] to compute the affinity of TF binding sites for all predicted miRNA promoters using TF models stored in the JASPAR database [30]. NF-kB appears among the first 10 TFs with the highest affinity for the 1000 bp-long region surrounding the predicted TSSs for 722 miRNA hairpin precursors (Table S1). Since tissues show specific miRNA expression patterns, we aimed to highlight the list of putative NF-kB-responsive miRNAs expressed in tissues strictly involved in the modulation of the inflammatory processes, including inflammaging. To achieve this goal, we focused our subsequent research on those miRNAs transcribed in human tissues such as “T cells”, “T cells 2”, “monocytic-cells”, “immune system cells”, “bone marrow”, “blood”, and “liver”. Only the libraries relative to healthy tissues have been taken into consideration. This approach retrieved 399 miRNA hairpin precursors showing “expression at the promoter level” in at least one of these tissues (Table S2). In general, this is a good indication that the mature forms of these miRNAs are expressed in a specific tissue. However, each step from DNA–RNA transcription to mature miRNA expression can be modulated, thereby modifying or blocking the final expression. Moreover, FANTOM4 libraries are characterized by a certain level of “transcriptional noise”, so we should expect false positives in mature miRNA predictions [23]. Therefore, among these putative NF-kB responsive miRNAs, we selected the “high confidence” hairpins in miRbase [27], retrieving 73 pre-miRNAs (Table S3). A growing body of evidence suggests that mature sequences derived from both arms of the hairpin might be biologically functional and even that the dominant mature sequence can be processed from opposite arms [31,32]. Following the approach of selecting only the “high confidence” miRNA hairpins and filtering the dataset for “Human Expression dataset” [28], 68 “high confidence” expressed miRNAs were identified. This pool of miRNAs, reported in Table 1, constitutes our final set of putative NF-kB responsive miRNAs expressed in healthy tissues linked to inflammatory processes. ## 2.2. Genomic Features of Putative NF-kB Responsive miRNAs According to their genomic location, it is possible to distinguish two classes of miRNAs: “intergenic miRNAs” are those located in intergenic regions of the genome, whereas “intragenic miRNAs” are those embedded in introns or exons of annotated genes [23]. Among the latter, “intronic miRNAs” are those located inside the introns of other genes and can either be co-transcribed with their host gene [33] or have an independent promoter [34,35,36], whereas intergenic miRNAs can derive from a primary miRNA transcript (pri-miRNAs) located in independent gene units [23,37]. Parallelly, it is possible to distinguish different categories of miRNA promoters: “intergenic promoters” are promoters assigned to intergenic miRNAs; “intragenic promoters” are promoters assigned to intragenic miRNAs and include both “host gene promoters” and “intronic promoters”; finally, “hybrid promoters” are those promoters that fall into intergenic regions upstream of intragenic miRNAs and could not be assigned unambiguously to the miRNA [23]. As shown in Table 1, among the promoter locations of the 68 putative NF-kB responsive miRNAs, 19 are “intergenic”, 15 are “host gene”, and 28 “intronic”. Interestingly, miR-15a, miR-16, miR-103, miR-186, and miR-33b can be modulated by both “host gene” and “intronic” promoters, whereas miR-194 is regulated by both “host gene” and “intergenic promoters”. Growing evidence indicates that alternative promoters are a mechanism for creating diversity in miRNA transcriptional regulation, as ascertained for protein-coding genes [38]. Regarding the phylogenesis of the 68 putative NF-kB responsive miRNAs, we showed that 22 miRNAs are conserved up to the vertebrate lineage (v), 38 miRNAs are conserved up to the mammal lineage (m), miR-194 and miR-19b up to the mammal and vertebrate lineage, and, finally, only 6 miRNAs are conserved in the primate lineage (p). ## 2.3. Characterization of the Interplay Linking NF-kB, miRNAs, and Their Host Genes To better characterize miRNAs that share the promoters of the host gene and to determine whether those host genes are also known to be regulated by NF-kB, multiple assessments were conducted. Firstly, we retrieved available information regarding the host genes and their intragenic miRNAs, as reported in Table 2, whereas expression correlation plots between miRNAs and their host gene are shown in Figure S1. No experimental evidence was found regarding the host gene of hsa-mir-374a, hsa-mir-545, or hsa-mir-15a. All the others are intronic miRNAs of genes involved in various biological processes ranging from DNA replication to differentiation:NFYC (Nuclear transcription factor Y subunit gamma) is a component of the sequence-specific heterotrimeric TF (NF-Y) which specifically recognizes a 5′- CCAAT-3′ box motif found in the promoters of its target genes. NF-Y can function as both an activator and a repressor, depending on its interacting cofactors [39];ZRANB2 (Zinc finger Ran-binding domain-containing protein 2) is a splicing factor required for alternative splicing of TRA2B/SFRS10 transcripts. May interfere with constitutive 5′-splice site selection [40];IARS2 (Isoleucine-tRNA ligase, mitochondrial) is a nuclear gene encoding mitochondrial isoleucyl-tRNA synthetase on which depends the translation of mitochondrial-encoded proteins [41];SMC4 (Structural maintenance of chromosomes protein 4) is the central component of the condensin complex, a complex required for the conversion of interphase chromatin into mitotic-like condense chromosomes [42];MCM7 (DNA replication licensing factor MCM7) acts as a component of the MCM2-7 complex (MCM complex) which is the replicative helicase essential for “once per cell cycle” DNA replication initiation and elongation in eukaryotic cells. It is the core component of CDC45-MCM-GINS (CMG) helicase, the molecular machine that unwinds template DNA during replication, and around which the replisome is built [43,44,45,46,47,48];NR6A1 (Nuclear receptor subfamily 6 group A member 1) is an orphan nuclear receptor that binds to a response element containing the sequence 5′-TCAAGGTCA-3′. By similarity, it may be involved in the regulation of gene expression in germ cell development during gametogenesis. It is involved in regulating embryonic stem cell differentiation, reproduction, and neuronal differentiation [49];TENM4 (Teneurin-4) is involved in neural development, regulating the establishment of proper connectivity within the nervous system. It plays a role in the establishment of the anterior–posterior axis during gastrulation. Moreover, it regulates the differentiation and cellular process formation of oligodendrocytes and myelination of small-diameter axons in the central nervous system (CNS) [50];COPZ1 (Coatomer subunit zeta-1) is a cytosolic protein complex involved in intracellular trafficking, endosome maturation, lipid homeostasis, and autophagy [51,52]. It is associated with iron metabolism through the regulation of transferrin [53,54];DDIT3 (DNA damage-inducible transcript 3 protein) is a multifunctional TF in endoplasmic reticulum stress response. It plays an essential role in the response to a wide variety of cell stresses and induces cell cycle arrest and apoptosis [55,56,57];WWP2 (NEDD4-like E3 ubiquitin-protein ligase WWP2) plays an important role in protein ubiquitination and inhibits activation-induced T cell death by catalyzing EGR2 ubiquitination [58]. In human embryonic stem cells, WWP2 promotes the degradation of TF OCT4, which not only plays an essential role in maintaining the pluripotent and self-renewing state of embryonic stem cells but also acts as a cell fate determinant through a gene dosage effect [55];HOXB3 (Homeobox protein Hox-B3) is a sequence-specific TF that is part of a developmental regulatory system that provides cells with specific positional identities on the anterior–posterior axis. Therefore, it may regulate gene expression, morphogenesis, and differentiation [59];SREBF1 (Sterol regulatory element-binding protein 1) is a precursor of the TF form (Processed sterol regulatory element-binding protein 1), which is embedded in the endoplasmic reticulum membrane [60]. Its processed form is a key TF that regulates the expression of genes involved in cholesterol biosynthesis and lipid homeostasis [60,61,62];PANK2 (Pantothenate kinase 2) is the mitochondrial isoform that catalyzes the phosphorylation of pantothenate to generate 4′-phosphopantothenate in the first and rate-determining step of coenzyme A (CoA) synthesis [63,64,65,66]. It is required for angiogenic activity of the umbilical vein of endothelial cells (HUVEC) [67]. Notably, five genes out of thirteen are engaged in transcription regulation (NR6A1, DDIT3, HOXB3, SREBF1, and NFYC), and only three are considered housekeeping genes (NFYC, ZRANB2, and COPZ1). Experimentally validated interactions shared among the three groups of molecules, namely (i) the 21 NF-kB responsive miRNAs sharing the host gene promoter, (ii) their host genes, and (iii) the three TF members (NFKB1, REL, and RELA) are depicted in Figure 2. Important nodes can be identified on the basis of their node centrality measures, such as degree and betweenness. The degree of a node is the total number of connections to other nodes. High-degree nodes are considered important “hubs” in a network [70,71]. The betweenness measures the number of shortest paths going through a node, taking into consideration the global network structure. Nodes with higher betweenness are important “bottlenecks” in a network [70,71]. Nodes identified by NFKB1, REL, miR-16-5p, miR-103a-3p, and NR6A1 have high degree centrality values, whereas RELA, miR-10a-5p, and miR-30e-5p represent nodes that occur between two dense clusters and have a high betweenness centrality even if their degree centrality values are not high. Therefore, we performed an explorative evaluation of known and potential protein–protein interactions among REL, RELA, NFKB1, and miRNA-host genes (Figure 3) by querying the STRING Database [72,73,74]. The STRING network shows that almost all host gene proteins have some degree of interaction. Experimental and biochemical data confirm the functional association of NFKB1, REL, and RELA. On the other hand, the higher confidence interaction values suggest a functional link between DDIT3, NFYC, MCM7, and SREBF1, as well as between IARS2, SMC4, and WWP2. Of note, experimental evidence in Figure 2 indicated that NFBK1, REL, RELA, DDIT3, NFYC, MCM7, SREBF1, and SMC4 are all targets of miR-16-5p, but miR-103a-3p, in turn, regulates IARS2, MCM7, and WWP2. Finally, the significantly differentially expressed host genes in ARDs have been identified (Table 3). Worth a mention is the downregulation of DDIT3, SMC4, and TENM4 in replicative senescence of human fibroblasts; the upregulation of SMC4 and MCM7 after vitamin C treatment; the upregulation of HOXB3 and TENM4 in Alzheimer’s disease; and the deregulation of DDIT3 and SMC4 in COVID-19 disease. ## 2.4. Pathways Targeted by the 68 Putative NF-kB Responsive miRNAs By performing an Ingenuity Pathway Analysis (IPA) Target Filter Analysis, we identified mRNAs targeted by the putative NF-kB responsive miRNAs. A total of 18,095 mRNAs were retrieved, of which 9613 were experimentally observed or highly predicted. The significance was reported as p-value in Table S4. The let-7a-5p was the miRNAs with the highest associated number of mRNA targets (2014 targets). Then we performed a network analysis focusing on putative NF-kB responsive miRNAs targeting mRNAs coding for molecules belonging to the NF-kB pathways (Figure 4). Interestingly, the NF-kB responsive miRNAs do not directly target genes coding for the NF-kB different subunits, but most of them are able to target genes coding for molecules belonging to NF-kB activation pathways, such as TLR and MYD88. This result is very interesting, considering that the modulation of NF-kB biological activity is related to its activation, rather than to the modulation of NF-kB subunits expression. Further, to discover the main diseases and functions associated with the selected miRNAs dataset, we performed an IPA Core Analysis (Figure 5). The diseases and functions are shown by bar chart, sorted by their −log p-value (p-value from Fisher’s Exact test). Cancers, immunological diseases, neurological diseases, and metabolic diseases, all well-recognized as inflammatory-based diseases, are among the diseases associated with the highest probability with NF-kB responsive miRNAs. Focusing on metabolic diseases, the most affected diseases are the non-insulin dependent diabetes mellitus (−log p-value 11.955), Alzheimer disease (−log p-value 9.532), and diabetes mellitus (−log p-value 7.680). To better explain the association of identified NF-kB putative responsive miRNAs with these human diseases, we depicted miRNAs-diseases relationship in Figure 6. Figure 6A depicts NF-KB putative responsive miRNAs associated with metabolic diseases, whereas Figure 6B–D, show the association between identified NF-kB responsive miRNAs and cardiovascular diseases, neurological diseases, and cancer, respectively. ## 2.5. The 68 Putative NF-kB Responsive miRNAs and Previously Identified Inflammamirs To test whether the 68 putative NF-kB responsive miRNAs could have a biological value in the context of the previous evidence, we compared our results with those already present in the literature. Among these 68 miRNAs, 21 have been experimentally validated to be transcribed by NF-kB1: miR-16-2 [75], miR-10a [76], miR-140-3p, miR-140-5p [77], miR-148b [78], miR-15b [79], miR-186 [80], miR-146a, miR-155, miR-19b, miR-20a, miR-19a, miR-17, miR-221, miR-222, miR-18a, miR-92a, miR-101, miR-23a, miR-27a, and miR-30c [21]. In addition, we have chosen as a reference all available data on the miRNAs relevant to aging, inflammation, and immunity that can be referred as inflammamiRs [81]. A detailed comparison table has been provided in Table S5. Figure 7A shows the “word cloud” with the 68 “high confidence” expressed miRNAs. The more features a specific miRNA holds (such as: the number of promoter types, the number of miRNA precursors, if it is expressed in more than one tissue, and, finally, if it is known to target NF-kB), the bigger and bolder it appears in the figure. Figure 7B depicts a Venn diagram modified from [81], displaying the miRNAs related to inflammation, immunity, and aging based on their circulating shuttles. In the inner circles are grouped exosome-associated miRNAs, while, in the outer circles, the circulating miRNAs associated with Ago-2, HDL, or other microparticles are grouped. In this version, it is important to note that bold characters indicate miRNAs overlapping among the two groups. Most of the 68 high-confidence NF-kB responsive miRNAs (reported in panel A) were previously identified as circulating miRNAs associated with aging, immunological functions, and inflammation, i.e., inflammaging [81]. Only three miRNAs, such as miR-154, miR-377, and miR-885-5p, were not retrieved in previous analysis [81]. However, based on recent literature, all of them are related to NF-kB/inflammation pathways [82,83,84]. All of the 68 NF-kB responsive miRNAs are therefore included in the Venn diagram reported in panel B, highlighting that these miRNAs identified as tissues expressed miRNAs are also detectable in blood, and most of them were identified inside extracellular vesicles, i.e., exosomes (miRNAs depicted in inner circles Figure 7B). ## 2.6. mRNAs Targeted by the 68 Putative NF-kB Responsive miRNAs Belonging to Pathways Involved in Aging Process and/or Age-Related Diseases By further analyzing the IPA Target Filter Analysis results, we finally identified the mRNAs, either experimentally validated or highly predicted, to be targeted by the 68 putative NF-kB responsive miRNAs, belonging to pathways related to aging or to the most common ARDs. Among the 9613 mRNAs predicted to be targeted by such NF-kB responsive miRNAs, 189 mRNAs targeted by 46 out of 68 miRNAs were associated to “cellular senescence pathway” (Table S6). In addition, out of the 9613, 8599 mRNAs were related to diseases reported in Figure 6, such as metabolic diseases, cardiovascular diseases, neurological diseases, and cancer. All these conditions share an inflammatory etiopathogenesis and are prototypical ARD. ## 3. Discussion NF-kB is an ubiquitously and evolutionarily conserved TF activated by a plethora of external and internal proinflammatory stimuli [85,86,87]. The crucial role as a mediator of the inflammatory responses, together with the finding that the activation or inhibition of NF-kB can induce or reverse, respectively, the main features of aged organisms, has brought NF-kB under consideration as a key TF that drives the biological aging process [88]. In this framework, the identification of genes modulated by NF-kB can be considered a cutting-edge issue [89,90,91]. NF-kB-responsive genes were extensively investigated, whereas NF-kB-responsive genes for non-coding RNAs were only recently highlighted. Here, we demonstrated that, by applying a data-mining approach, it is possible to select the most reliable NF-kB responsive miRNAs. Most notably, the availability of data on TFs binding sites on human miRNAs sequences constituted a starting point and the foundation for studying all human miRNAs with potential NF-kB binding sites in their promoter regions. Some years ago, a general hypothesis was advanced that the aging process and the development of the most common ARDs could be fostered by a low-grade, chronic, systemic inflammatory process named “inflammaging” [92]. Inflammaging, which is principally sustained by the activation of the innate immune cells, is paralleled by the increased burden of senescent cells acquiring a senescence-associated secretory phenotype (SASP), which turns senescent cells into proinflammatory cells [86,93,94,95,96]. In immune cells and tissues obtained from patients affected by the most common ARDs, NF-kB is commonly constitutively activated [97]. Of note, NF-kB activation should be an inducible, but transient, event in physiological conditions. However, despite the presence of multiple checks and balances that control NF-kB activation, in cellular and organismal aging, as well as in many ARDs, NF-kB activation becomes persistent [98,99]. In this study, using PROmiRNA software and a data-mining approach, we provide a list of 73 putative “high confidence” pre-miRNAs sequences corresponding to 68 NF-kB responsive mature miRNAs sequences. Likewise, we highlighted the presence of distinct types of promoters that can regulate NF-kB responsive miRNAs. A total of 33 miRNAs of the 68 high confidence expressed miRNAs identified have an “intronic” promoter, and 5 of these have both an “intronic” and “host gene” promoter, whereas only one miRNA (miR-194) shares both “intergenic” and “host-gene” promoters. Alternative promoters are a common mechanism to create diversity in the transcriptional regulation of miRNA [38]. It has been demonstrated that “intronic” promoters convey an additional degree of freedom over intragenic miRNA transcriptional regulation by virtue of some peculiar characteristics, thus allowing the modulation of miRNA expression levels in a tissue- and condition-specific manner [23]. Besides the other features, in this context, it is important to stress that:“Intronic” promoters can explain cases of poor correlation between host gene and miRNA expression, functioning as a real alternative promoter [23]. As shown in Figure S1, the expression levels of NF-kB-miRNAs modulated by both “host gene” and “intronic” promoters (i.e., miR-16, miR-103, miR-186, and miR-33b) or by both “host gene” and “intergenic promoters” (i.e., miR-194) are not correlated with the expression levels of their host gene, whereas most of the miRNAs that share the host gene promoters are characterized by directly (e.g., miR-15b) or inversely (e.g., miR-30c, miR-616, and miR-93) correlated transcription levels. “Intronic” promoters are expressed in a tissue-specific manner, but “host gene” promoters are considered primarily for housekeeping gene regulation [23]. Housekeeping genes are required for the maintenance of essential functions of any cell type, so they are expected to be constitutively expressed in all cells and at any development stage [100]. Among the NF-kB-miRNA host genes, COPZ1, NFYC, and ZRANB2 have been cataloged as housekeeping genes (Table 2).“Intronic” promoters are mainly triggered by tissue-specific master regulator TFs, instead of TFs of “host gene” promoters, which broadly overlap with those of protein coding genes and can be considered mainly for housekeeping (“intergenic” promoters are regulated by a combination of intronic-specific and host-gene specific TFs). This suggests a different evolutionary mechanism [23]. In this study, the expression levels of the three housekeeping host genes (COPZ1, ZRANB2, and NFYC) and their miRNAs (miR-148b-3p, miR-186-5p, and, lastly, miR-30c-5p and miR-30e-5p, respectively) are mainly inversely correlated or not showing clear correlation trends (Figure S1).“Intronic” miRNA promoters are less evolutionarily conserved than either “intergenic” or “host gene” promoters [23].Conversely, evolutionarily conserved miRNAs are more likely to be regulated by an “intronic” promoter [23].Moreover, those intragenic miRNAs that share the promoters of the host gene interact with their own host genes (miR-16-2::MSC4; miR-106b::MCM7, miR-181b-2::NR6A1, miR-708::TENM4, miR-148b::COPZ1, and miR-10a::HOXB3), but also with the other functionally related host genes, creating a complex regulatory mechanism (Figure 2). NFKB1, REL, miR-16-5p, miR-103a-3p, and NR6A1 are the most important hub nodes in the network, whereas miR-10a-5p connects the hub nodes identified by NFKB1, NR6A1, and HOXB3, and miR-30e-5p connects REL, NR6A1, and ZRAMB2 hubs. Interestingly, in the network, it is possible to identify a clear TF-miRNA feed-forward loop involving DDIT3, miR-16-5p, and NFYC. In a TF-miRNA feed-forward loop, TF and miRNA co-regulate the target genes: in a “coherent” feed-forward loop, the TF and miRNA have the same effects on their common targets, whereas, in an “incoherent” feed-forward loop, the TF and miRNA carry out opposing effects, which precisely fine-tune gene expressions to minimize noise and maintain stability [68,101]. TF-miRNA feed-forward loops have a specific function in noise buffering effects, which can minimize the response of stochastic signaling noise to maintain steady-state target levels [102,103]. Disruption of feed-forward loops could lead to serious dysregulations at the origin of diseases and cancers, e.g., interference in the NF-kB/miR-19/CYLD loop can induce T cell leukemogenesis [103,104]. Therefore, investigating the regulatory motifs among DDIT3, 16-5p, and NFYC could provide valuable insights to dissect the molecular mechanisms underlying biological processes and diseases triggered by NF-kB constitutive activation. Protein–protein interaction analysis of protein-coding host genes revealed that most of them could be functionally related (Figure 3). Beyond the well-known functional association of NFKB1, REL (cREL), and RELA, some data have highlighted the association with endoplasmic reticulum stress, providing opportunities to fine-tune cellular stress responses [105]. In the framework of atherosclerosis, multiple links between NF-kB and ER stress were suggested. A disturbed flow can cause endoplasmic reticulum stress, leading to SREBF1 activation with nuclear localization and to DDIT3 expression triggered by endoplasmic reticulum stress response elements [106]. NFYC is a subunit of a trimeric complex (NFY) known to interact with several TFs to enable the synergistic activation of specific classes of promoters. NFY directly controls the expression of TF genes such as P53 (DNA-damage), XBP1, CHOP/DDIT3 (endoplasmic reticulum stress), and HSF1 (heat shock) [107,108]. Of note, experimental data have shown the upregulation of both SMC4 and MCM7 in mesenchymal stem cells after vitamin C treatment; the downregulation of DDIT3, SMC4, and TENM4 in replicative senescence of human fibroblasts; the upregulation of HOXB3 and TENM4 in Alzheimer’s disease; and, finally, the deregulation of DDIT3 and SMC4 in COVID-19 disease (Table 3). In this scenario, targeting NF-kB signaling is becoming a promising strategy for drug development and ARD treatments [91,109]. Almost all of the 68 miRNAs that we identified in our current analysis were previously associated with inflammaging processes and with the most common ARDs, such as metabolic diseases, cardiovascular diseases, neurodegenerative diseases, and cancers [110,111]. Out of the 9613 mRNAs targeted by the 68 NF-kB responsive miRNAs, 8599 mRNAs were related to such diseases. Of note, 189 mRNAs were associated with “cellular senescence pathway”, which is recognized as the main culprit of the aging process. Most of the NF-kB responsive miRNAs are involved in a negative feedback loop to restrain exacerbated inflammation [5,53,65,66,67,68,69,70]. Notably, the identified NF-kB responsive miRNAs are not able to directly modulate gene expression of NF-kB subunits but are able to target molecules belonging to NF-kB activation pathways (canonical and non-canonical pathway). Interestingly, among the NF-kB-responsive miRNAs genes identified with our approach, the most relevant examples of mRNAs that can target molecules belonging to the NF-kB canonical and non-canonical pathways, or related molecules, are miR-146a and miR-155. In fact, miR-146a and miR-155, control NF-kB activity during inflammation by a combinatory action without directly targeting NF-KB subunits [112]. miR-155 is rapidly upregulated by NF-kB during the early phase inflammatory response through a positive feedback loop necessary for signal amplification. miR-146a is rather gradually upregulated by NF-kB and forms a negative feedback loop attenuating NF-kB activity in the late phase of inflammation. The combined action of these two positive (NF-kB::miR-155) and negative (NF-kB::miR-146a) NF-kB-miRNA regulatory loops provides optimal NF-kB activity during inflammatory stimuli, and eventually lead to the resolution of the inflammatory response in physiological condition. Another example is miR-16 that targets the IKKα/β complex of the NF-kB canonical pathway polarizing macrophages toward an M2 phenotype [113].These results are in line with the known modulation of NF-kB biological activity, based on the activation and not on the expression of its subunits [5]. Interestingly, all the 68 NF-kB responsive miRNAs are detectable in blood, and most of them were identified inside extracellular vesicles, i.e., exosomes. Exosomes are currently considered to be a crucial intercellular cross-talk mechanism, acting at both the paracrine and systemic levels [114]. This result highlights the complexity of the feed-back loops between NF-kB activation in specific tissues, the expression of NF-kB responsive miRNAs, and their release in the bloodstream as a systemic intercellular communication mechanism. A further level of complexity can be envisaged considering that NF-kB is known to indirectly regulate miRNA expression through the modulation of other TFs. NF-kB can modulate AP-1 TF [115], which, in turn, is able to modulate different miRNAs genes, i.e., miR-21 [116]. Of note, among the 68 miRNAs, 21 were already experimentally identified as NF-kB responsive, reinforcing the reliability of our results. Our data also highlight the potential value of the 47 NF-kB putative responsive miRNAs (listed in Figure 7B) that are yet to be experimentally validated. Overall, our results are of interest in the framework of the research on the biomarkers/drugs of aging and inflammation related diseases. If NF-kB responsive miRNAs are hyper-transcribed in tissues involved in the modulation of inflammatory responses, the hypothesis that circulating miRNAs could be useful tools to track the trajectories of healthy or un-healthy aging is reinforced [117,118,119,120,121] and possible therapeutic strategies based on the inhibition of those miRNAs could be further tested. ## Limitation of the Study The data-mining process frequently encompasses a further phase involving the extraction of implicit relational patterns through traditional statistics or machine learning, but the particularity of the research question and the type of data available have been a hindrance to this kind of analysis. ## 4.1. Data Mining Process In the field of Knowledge Discovery in Databases (KDD), a data-mining approach is used to extract meaningful information and to develop significant relationships among variables stored in large data sets [122]. In this study, we have mined and integrated data from multiple databases to select NF-kB responsive miRNAs, and the process has been tailored based on the research question. Four main steps can be distinguished: ## 4.1.1. Database Selection The following data sources have been investigated to retrieve the data and develop the study: PROmiRNA [23], FANTOM4 libraries [26], “High confidence human hairpins” in miRBase [27], and “Human Expression” dataset (microrna.org) [28]. ## 4.1.2. Data Extraction and Integration This phase includes downloading, extracting, filtering, and combining the data from the databases previously identified. The integration of multiple datasets has been possible through the following steps. ## 4.1.3. Data Cleaning and Transformation Because the data originates from multiple sources, the integration often involves converting data formats, cleaning, removal of incorrect data, generating new variables, resolving redundancy, and checking against miRNA nomenclature consistency, both between miRNAs names originating in different miRBase versions and between the names of pri-miRNAs and the mature forms. This issue has been manually curated by comparing miRNA names in miRBase database version 21. ## 4.1.4. Assessment of the Results This is the final stage of a KDD process, involving the translation of aggregated data into comprehensible knowledge. The validity and reliability of the data were tested by comparing the results obtained in the data-mining process with those already published in the literature. The whole data-mining process is illustrated in the data flow diagram in Figure 1. Data obtained at each intermediate step are provided in Supplementary Tables S1–S3. The final miRNA-pool is reported in Table 1. ## 4.2.1. Evaluation of miRNA-Host Gene-Transcription Factor Interactions Host gene and intragenic miRNAs information (Table 2), as well as expression correlation data between miRNAs and their host genes (Figure S1), were retrieved from MiRIAD, a database integrating miRNA inter- and intragenic data (https://www.miriad-database.org/, accessed on 10 January 2023) [128]. In Table 2, host gene biological process were obtained from the UniProt database (Release 2022_05) (https://www.uniprot.org/, accessed on 10 January 2023) [129]; the Housekeeping and Reference Transcript Atlas (HRT Atlas v1.0) (https://housekeeping.unicamp.br/, accessed on 10 January 2023) [100] was investigated to discover those host genes cataloged as housekeeping genes. Experimentally validated interactions shared among NFKB1, REL, RELA, the NF-kB-responsive miRNAs sharing the host gene promoter, and their host genes, were identified (Figure 2) by querying: (i) DIANA-TarBase v8 (http://www.microrna.gr/tarbase, accessed on 10 January 2023), retrieving experimentally supported miRNA-gene interactions [130]; (ii) TRRUST v2 (www.grnpedia.org/trrust, accessed on 10 January 2023), a manually curated database of transcriptional regulatory interactions [131]; and (iii) STRING v10, to highlight the protein–protein interactions, with the constraint to retrieve only experimental evidences [132]. The whole process, including the final network creation and visualization, was handled using miRNet (version 2.0), a miRNA-centric network visual analytics platform (https://www.mirnet.ca/, accessed on 10 January 2023) [68,70,133]. STRING database Version 11.5 (https://string-db.org/, accessed on 10 January 2023) was used to discover known and potential interactions among REL, RELA, NFKB1, and miRNA-host gene proteins (Figure 3). STRING is a database of predicted and known protein–protein interactions. The interactions include direct (physical) and indirect (functional) associations; these stem from knowledge transfer between organisms, from interactions aggregated from other (primary) databases, and from computational prediction [72,73,74]. The network was created by setting a minimum required interaction score of 0.15. The RNA-seq datasets in Aging Atlas (https://ngdc.cncb.ac.cn/aging/index, accessed on 10 January 2023) were examined to explore age-related changes in host gene expression [134]. Table 3 shows differentially expressed host genes in strictly age-related conditions; only those genes showing \|log2FC\| > 1 and q-value < 0.005 (or p-value < 0.005 if q-value was not provided) have been reported. Data relative to particular experimental conditions (e.g., gene knockdown) have not been reported. All websites and online tools were accessed in the period between January and February 2023. ## 4.2.2. Ingenuity Pathway Analysis Bioinformatic evaluations (networks and disease analyses) were performed by the Ingenuity Pathway Analysis software (Qiagen, Hilden, Germany). The putative NF-kB responsive miRNAs identified through the data-mining process were analyzed to explore the experimentally observed or high predicted mRNA targets via the microRNA Target Filter Analysis. Furthermore, an IPA Core Analysis was performed to define the associated diseases and functions. Direct and indirect relationships from the Ingenuity Knowledge Base (gene only) datasets were considered. We filtered only molecules and/or relationships experimentally observed in any tissue from human, rat, or mouse. Across the observations, 51 miRNAs were ready to be analyzed (Table S4) [135]. All the networks, diseases, and biological functions were assessed using IPA software (Qiagen, Hilden, Germany). ## 5. Conclusions Here, we demonstrated that a well-settled data-mining approach may disclose the most reliable miRNAs having a key role in the regulation of specific pathways of interest. Deciphering the crosstalk between miRNAs and NF-kB is one of the major topics to be investigated to understand the complex derailment of several metabolic pathways in normal and pathological aging. Future studies are needed to confirm that the identification of such miRNAs is of diagnostic/prognostic/therapeutic relevance for the most common inflammatory- and age-related conditions. ## References 1. Salminen A., Kauppinen A., Suuronen T., Kaarniranta K.. SIRT1 longevity factor suppresses NF-kappaB-driven immune responses: Regulation of aging via NF-kappaB acetylation?. Bioessays (2008) 30 939-942. DOI: 10.1002/bies.20799 2. Salminen A., Huuskonen J., Ojala J., Kauppinen A., Kaarniranta K., Suuronen T.. Activation of innate immunity system during aging: NF-kB signaling is the molecular culprit of inflamm-aging. Ageing Res. Rev. (2008) 7 83-105. DOI: 10.1016/j.arr.2007.09.002 3. Liu P., Li Y., Wang W., Bai Y., Jia H., Yuan Z., Yang Z.. Role and mechanisms of the NF-kB signaling pathway in various developmental processes. Biomed. Pharm. (2022) 153 113513. DOI: 10.1016/j.biopha.2022.113513 4. Dorrington M.G., Fraser I.D.C.. NF-kappaB Signaling in Macrophages: Dynamics, Crosstalk, and Signal Integration. Front. Immunol. (2019) 10 705. DOI: 10.3389/fimmu.2019.00705 5. Olivieri F., Prattichizzo F., Giuliani A., Matacchione G., Rippo M.R., Sabbatinelli J., Bonafe M.. miR-21 and miR-146a: The microRNAs of inflammaging and age-related diseases. Ageing Res. Rev. (2021) 70 101374. DOI: 10.1016/j.arr.2021.101374 6. Liu T., Zhang L., Joo D., Sun S.C.. NF-kappaB signaling in inflammation. Signal Transduct. Target. Ther. (2017) 2 17023. DOI: 10.1038/sigtrans.2017.23 7. Adler A.S., Sinha S., Kawahara T.L., Zhang J.Y., Segal E., Chang H.Y.. Motif module map reveals enforcement of aging by continual NF-kappaB activity. Genes Dev. (2007) 21 3244-3257. DOI: 10.1101/gad.1588507 8. Tilstra J.S., Clauson C.L., Niedernhofer L.J., Robbins P.D.. NF-kappaB in Aging and Disease. Aging Dis. (2011) 2 449-465. PMID: 22396894 9. Bernal-Lopez M.R., Garrido-Sanchez L., Gomez-Carrillo V., Gallego-Perales J.L., Llorente-Cortes V., Calleja F., Gomez-Huelgas R., Badimon L., Tinahones F.J.. Antioxidized LDL antibodies are associated with different metabolic pathways in patients with atherosclerotic plaque and type 2 diabetes. Diabetes Care (2013) 36 1006-1011. DOI: 10.2337/dc12-1030 10. Bektas A., Zhang Y., Lehmann E., Wood W.H., Becker K.G., Madara K., Ferrucci L., Sen R.. Age-associated changes in basal NF-kappaB function in human CD4. Aging (2014) 6 957-974. DOI: 10.18632/aging.100705 11. Rothschild D.E., McDaniel D.K., Ringel-Scaia V.M., Allen I.C.. Modulating inflammation through the negative regulation of NF-kappaB signaling. J. Leukoc. Biol. (2018) 103 1131-1150. DOI: 10.1002/JLB.3MIR0817-346RRR 12. Arora S., Rana R., Chhabra A., Jaiswal A., Rani V.. miRNA-transcription factor interactions: A combinatorial regulation of gene expression. Mol. Genet. Genom. (2013) 288 77-87. DOI: 10.1007/s00438-013-0734-z 13. Iwama H.. Coordinated networks of microRNAs and transcription factors with evolutionary perspectives. Adv. Exp. Med. Biol. (2013) 774 169-187. DOI: 10.1007/978-94-007-5590-1_10 14. Elizarova A., Ozturk M., Guler R., Medvedeva Y.A.. MIREyA: A computational approach to detect miRNA-directed gene activation. F1000Research (2021) 10 249. DOI: 10.12688/f1000research.28142.2 15. Li Y., Shan G., Teng Z.Q., Wingo T.S.. Editorial: Non-Coding RNAs and Human Diseases. Front. Genet. (2020) 11 523. DOI: 10.3389/fgene.2020.00523 16. Peng Y.. Non-coding RNAs in human cancer. Semin. Cancer Biol. (2021) 75 1-2. DOI: 10.1016/j.semcancer.2021.04.010 17. Akhtar M.M., Micolucci L., Islam M.S., Olivieri F., Procopio A.D.. Bioinformatic tools for microRNA dissection. Nucleic Acids Res. (2016) 44 24-44. DOI: 10.1093/nar/gkv1221 18. Lukiw W.J.. NF-kappaB-regulated, proinflammatory miRNAs in Alzheimer’s disease. Alzheimers Res. Ther. (2012) 4 47. DOI: 10.1186/alzrt150 19. Wang L.L., Huang Y., Wang G., Chen S.D.. The potential role of microRNA-146 in Alzheimer’s disease: Biomarker or therapeutic target?. Med. Hypotheses (2012) 78 398-401. DOI: 10.1016/j.mehy.2011.11.019 20. Yang M., Ye L., Wang B., Gao J., Liu R., Hong J., Wang W., Gu W., Ning G.. Decreased miR-146 expression in peripheral blood mononuclear cells is correlated with ongoing islet autoimmunity in type 1 diabetes patients 1miR-146. J. Diabetes (2015) 7 158-165. DOI: 10.1111/1753-0407.12163 21. Markopoulos G.S., Roupakia E., Tokamani M., Alabasi G., Sandaltzopoulos R., Marcu K.B., Kolettas E.. Roles of NF-kappaB Signaling in the Regulation of miRNAs Impacting on Inflammation in Cancer. Biomedicines (2018) 6. DOI: 10.3390/biomedicines6020040 22. Chen X., Wang C.C., Guan N.N.. Computational Models in Non-Coding RNA and Human Disease. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21051557 23. Marsico A., Huska M.R., Lasserre J., Hu H., Vucicevic D., Musahl A., Orom U., Vingron M.. PROmiRNA: A new miRNA promoter recognition method uncovers the complex regulation of intronic miRNAs. Genome Biol. (2013) 14 R84. DOI: 10.1186/gb-2013-14-8-r84 24. Wu J., Ding J., Yang J., Guo X., Zheng Y.. MicroRNA Roles in the Nuclear Factor Kappa B Signaling Pathway in Cancer. Front. Immunol. (2018) 9 546. DOI: 10.3389/fimmu.2018.00546 25. Amini-Farsani Z., Yadollahi-Farsani M., Arab S., Forouzanfar F., Yadollahi M., Asgharzade S.. Prediction and analysis of microRNAs involved in COVID-19 inflammatory processes associated with the NF-kB and JAK/STAT signaling pathways. Int. Immunopharmacol. (2021) 100 108071. DOI: 10.1016/j.intimp.2021.108071 26. Kawaji H., Severin J., Lizio M., Waterhouse A., Katayama S., Irvine K.M., Hume D.A., Forrest A.R., Suzuki H., Carninci P.. The FANTOM web resource: From mammalian transcriptional landscape to its dynamic regulation. Genome Biol. (2009) 10 R40. DOI: 10.1186/gb-2009-10-4-r40 27. Kozomara A., Griffiths-Jones S.. miRBase: Annotating high confidence microRNAs using deep sequencing data. Nucleic Acids Res. (2014) 42 D68-D73. DOI: 10.1093/nar/gkt1181 28. Betel D., Wilson M., Gabow A., Marks D.S., Sander C.. The microRNA.org resource: Targets and expression. Nucleic Acids Res. (2008) 36 D149-D153. DOI: 10.1093/nar/gkm995 29. Thomas-Chollier M., Hufton A., Heinig M., O’Keeffe S., Masri N.E., Roider H.G., Manke T., Vingron M.. Transcription factor binding predictions using TRAP for the analysis of ChIP-seq data and regulatory SNPs. Nat. Protoc. (2011) 6 1860-1869. DOI: 10.1038/nprot.2011.409 30. Portales-Casamar E., Thongjuea S., Kwon A.T., Arenillas D., Zhao X., Valen E., Yusuf D., Lenhard B., Wasserman W.W., Sandelin A.. JASPAR 2010: The greatly expanded open-access database of transcription factor binding profiles. Nucleic Acids Res. (2010) 38 D105-D110. DOI: 10.1093/nar/gkp950 31. Griffiths-Jones S., Hui J.H., Marco A., Ronshaugen M.. MicroRNA evolution by arm switching. EMBO Rep. (2011) 12 172-177. DOI: 10.1038/embor.2010.191 32. Kozomara A., Birgaoanu M., Griffiths-Jones S.. miRBase: From microRNA sequences to function. Nucleic Acids Res. (2019) 47 D155-D162. DOI: 10.1093/nar/gky1141 33. Schanen B.C., Li X.. Transcriptional regulation of mammalian miRNA genes. Genomics (2011) 97 1-6. DOI: 10.1016/j.ygeno.2010.10.005 34. Davis B.N., Hata A.. Regulation of MicroRNA Biogenesis: A miRiad of mechanisms. Cell Commun. Signal. (2009) 7 18. DOI: 10.1186/1478-811X-7-18 35. Ozsolak F., Poling L.L., Wang Z., Liu H., Liu X.S., Roeder R.G., Zhang X., Song J.S., Fisher D.E.. Chromatin structure analyses identify miRNA promoters. Genes Dev. (2008) 22 3172-3183. DOI: 10.1101/gad.1706508 36. Monteys A.M., Spengler R.M., Wan J., Tecedor L., Lennox K.A., Xing Y., Davidson B.L.. Structure and activity of putative intronic miRNA promoters. RNA (2010) 16 495-505. DOI: 10.1261/rna.1731910 37. Krol J., Loedige I., Filipowicz W.. The widespread regulation of microRNA biogenesis, function and decay. Nat. Rev. Genet. (2010) 11 597-610. DOI: 10.1038/nrg2843 38. Carninci P., Sandelin A., Lenhard B., Katayama S., Shimokawa K., Ponjavic J., Semple C.A., Taylor M.S., Engstrom P.G., Frith M.C.. Genome-wide analysis of mammalian promoter architecture and evolution. Nat. Genet. (2006) 38 626-635. DOI: 10.1038/ng1789 39. Nakshatri H., Bhat-Nakshatri P., Currie R.A.. Subunit association and DNA binding activity of the heterotrimeric transcription factor NF-Y is regulated by cellular redox. J. Biol. Chem. (1996) 271 28784-28791. DOI: 10.1074/jbc.271.46.28784 40. Adams D.J., van der Weyden L., Mayeda A., Stamm S., Morris B.J., Rasko J.E.. ZNF265—A novel spliceosomal protein able to induce alternative splicing. J. Cell Biol. (2001) 154 25-32. DOI: 10.1083/jcb.200010059 41. Schwartzentruber J., Buhas D., Majewski J., Sasarman F., Papillon-Cavanagh S., Thiffault I., Sheldon K.M., Massicotte C., Patry L., Simon M.. Mutation in the nuclear-encoded mitochondrial isoleucyl-tRNA synthetase IARS2 in patients with cataracts, growth hormone deficiency with short stature, partial sensorineural deafness, and peripheral neuropathy or with Leigh syndrome. Hum. Mutat. (2014) 35 1285-1289. DOI: 10.1002/humu.22629 42. Kimura K., Cuvier O., Hirano T.. Chromosome condensation by a human condensin complex in Xenopus egg extracts. J. Biol. Chem. (2001) 276 5417-5420. DOI: 10.1074/jbc.C000873200 43. Rzechorzek N.J., Hardwick S.W., Jatikusumo V.A., Chirgadze D.Y., Pellegrini L.. CryoEM structures of human CMG-ATPgammaS-DNA and CMG-AND-1 complexes. Nucleic Acids Res. (2020) 48 6980-6995. DOI: 10.1093/nar/gkaa429 44. Jones M.L., Baris Y., Taylor M.R.G., Yeeles J.T.P.. Structure of a human replisome shows the organisation and interactions of a DNA replication machine. EMBO J. (2021) 40 e108819. DOI: 10.15252/embj.2021108819 45. Jenkyn-Bedford M., Jones M.L., Baris Y., Labib K.P.M., Cannone G., Yeeles J.T.P., Deegan T.D.. A conserved mechanism for regulating replisome disassembly in eukaryotes. Nature (2021) 600 743-747. DOI: 10.1038/s41586-021-04145-3 46. Baris Y., Taylor M.R.G., Aria V., Yeeles J.T.P.. Fast and efficient DNA replication with purified human proteins. Nature (2022) 606 204-210. DOI: 10.1038/s41586-022-04759-1 47. Ishimi Y., Irie D.. G364R mutation of MCM4 detected in human skin cancer cells affects DNA helicase activity of MCM4/6/7 complex. J. Biochem. (2015) 157 561-569. DOI: 10.1093/jb/mvv015 48. Ishimi Y.. A DNA helicase activity is associated with an MCM4, -6, and -7 protein complex. J. Biol. Chem. (1997) 272 24508-24513. DOI: 10.1074/jbc.272.39.24508 49. Wang Y., Zhang Y., Dai X., Liu Z., Yin P., Wang N., Zhang P.. NR6A1 couples with cAMP response element binding protein and regulates vascular smooth muscle cell migration. Int. J. Biochem. Cell Biol. (2015) 69 225-232. DOI: 10.1016/j.biocel.2015.10.026 50. Hor H., Francescatto L., Bartesaghi L., Ortega-Cubero S., Kousi M., Lorenzo-Betancor O., Jimenez-Jimenez F.J., Gironell A., Clarimon J., Drechsel O.. Missense mutations in TENM4, a regulator of axon guidance and central myelination, cause essential tremor. Hum. Mol. Genet. (2015) 24 5677-5686. DOI: 10.1093/hmg/ddv281 51. Beck R., Rawet M., Wieland F.T., Cassel D.. The COPI system: Molecular mechanisms and function. FEBS Lett. (2009) 583 2701-2709. DOI: 10.1016/j.febslet.2009.07.032 52. Razi M., Chan E.Y., Tooze S.A.. Early endosomes and endosomal coatomer are required for autophagy. J. Cell Biol. (2009) 185 305-321. DOI: 10.1083/jcb.200810098 53. Ward W.A., Tsai T.M., Breidenbach W.. Per Primam thumb replantation for all patients with traumatic amputations. Clin. Orthop. Relat. Res. (1991) 266 90-95. DOI: 10.1097/00003086-199105000-00015 54. Zhang Y., Kong Y., Ma Y., Ni S., Wikerholmen T., Xi K., Zhao F., Zhao Z., Wang J., Huang B.. Loss of COPZ1 induces NCOA4 mediated autophagy and ferroptosis in glioblastoma cell lines. Oncogene (2021) 40 1425-1439. DOI: 10.1038/s41388-020-01622-3 55. Yamaguchi H., Wang H.G.. CHOP is involved in endoplasmic reticulum stress-induced apoptosis by enhancing DR5 expression in human carcinoma cells. J. Biol. Chem. (2004) 279 45495-45502. DOI: 10.1074/jbc.M406933200 56. Ohoka N., Yoshii S., Hattori T., Onozaki K., Hayashi H.. TRB3, a novel ER stress-inducible gene, is induced via ATF4-CHOP pathway and is involved in cell death. EMBO J. (2005) 24 1243-1255. DOI: 10.1038/sj.emboj.7600596 57. Oliveira S.J., Pinto J.P., Picarote G., Costa V.M., Carvalho F., Rangel M., de Sousa M., de Almeida S.F.. ER stress-inducible factor CHOP affects the expression of hepcidin by modulating C/EBPalpha activity. PLoS ONE (2009) 4. DOI: 10.1371/journal.pone.0006618 58. Chen A., Gao B., Zhang J., McEwen T., Ye S.Q., Zhang D., Fang D.. The HECT-type E3 ubiquitin ligase AIP2 inhibits activation-induced T-cell death by catalyzing EGR2 ubiquitination. Mol. Cell Biol. (2009) 29 5348-5356. DOI: 10.1128/MCB.00407-09 59. Verma A., Arya R., Brahmachari V.. Identification of a polycomb responsive region in human HoxA cluster and its long-range interaction with polycomb enriched genomic regions. Gene (2022) 845 146832. DOI: 10.1016/j.gene.2022.146832 60. Xu D., Wang Z., Xia Y., Shao F., Xia W., Wei Y., Li X., Qian X., Lee J.H., Du L.. The gluconeogenic enzyme PCK1 phosphorylates INSIG1/2 for lipogenesis. Nature (2020) 580 530-535. DOI: 10.1038/s41586-020-2183-2 61. Yokoyama C., Wang X., Briggs M.R., Admon A., Wu J., Hua X., Goldstein J.L., Brown M.S.. SREBP-1, a basic-helix-loop-helix-leucine zipper protein that controls transcription of the low density lipoprotein receptor gene. Cell (1993) 75 187-197. DOI: 10.1016/S0092-8674(05)80095-9 62. Amemiya-Kudo M., Shimano H., Hasty A.H., Yahagi N., Yoshikawa T., Matsuzaka T., Okazaki H., Tamura Y., Iizuka Y., Ohashi K.. Transcriptional activities of nuclear SREBP-1a, -1c, and -2 to different target promoters of lipogenic and cholesterogenic genes. J. Lipid Res. (2002) 43 1220-1235. DOI: 10.1194/jlr.M100417-JLR200 63. Kotzbauer P.T., Truax A.C., Trojanowski J.Q., Lee V.M.. Altered neuronal mitochondrial coenzyme A synthesis in neurodegeneration with brain iron accumulation caused by abnormal processing, stability, and catalytic activity of mutant pantothenate kinase 2. J. Neurosci. (2005) 25 689-698. DOI: 10.1523/JNEUROSCI.4265-04.2005 64. Leonardi R., Zhang Y.M., Lykidis A., Rock C.O., Jackowski S.. Localization and regulation of mouse pantothenate kinase 2. FEBS Lett. (2007) 581 4639-4644. DOI: 10.1016/j.febslet.2007.08.056 65. Leonardi R., Rock C.O., Jackowski S., Zhang Y.M.. Activation of human mitochondrial pantothenate kinase 2 by palmitoylcarnitine. Proc. Natl. Acad. Sci. USA (2007) 104 1494-1499. DOI: 10.1073/pnas.0607621104 66. Zhang Y.M., Rock C.O., Jackowski S.. Biochemical properties of human pantothenate kinase 2 isoforms and mutations linked to pantothenate kinase-associated neurodegeneration. J. Biol. Chem. (2006) 281 107-114. DOI: 10.1074/jbc.M508825200 67. Pagani F., Trivedi A., Khatri D., Zizioli D., Garrafa E., Mitola S., Finazzi D.. Silencing of pantothenate kinase 2 reduces endothelial cell angiogenesis. Mol. Med. Rep. (2018) 18 4739-4746. DOI: 10.3892/mmr.2018.9480 68. Chang L., Zhou G., Soufan O., Xia J.. miRNet 2.0: Network-based visual analytics for miRNA functional analysis and systems biology. Nucleic Acids Res. (2020) 48 W244-W251. DOI: 10.1093/nar/gkaa467 69. Pavlopoulos G.A., Kontou P.I., Pavlopoulou A., Bouyioukos C., Markou E., Bagos P.G.. Bipartite graphs in systems biology and medicine: A survey of methods and applications. Gigascience (2018) 7 1-31. DOI: 10.1093/gigascience/giy014 70. Fan Y., Xia J.. miRNet-Functional Analysis and Visual Exploration of miRNA-Target Interactions in a Network Context. Methods Mol. Biol. (2018) 1819 215-233. DOI: 10.1007/978-1-4939-8618-7_10 71. Zhu X., Gerstein M., Snyder M.. Getting connected: Analysis and principles of biological networks. Genes Dev. (2007) 21 1010-1024. DOI: 10.1101/gad.1528707 72. Szklarczyk D., Gable A.L., Nastou K.C., Lyon D., Kirsch R., Pyysalo S., Doncheva N.T., Legeay M., Fang T., Bork P.. The STRING database in 2021: Customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Res. (2021) 49 D605-D612. DOI: 10.1093/nar/gkaa1074 73. Szklarczyk D., Gable A.L., Lyon D., Junge A., Wyder S., Huerta-Cepas J., Simonovic M., Doncheva N.T., Morris J.H., Bork P.. STRING v11: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. (2019) 47 D607-D613. DOI: 10.1093/nar/gky1131 74. Szklarczyk D., Morris J.H., Cook H., Kuhn M., Wyder S., Simonovic M., Santos A., Doncheva N.T., Roth A., Bork P.. The STRING database in 2017: Quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. (2017) 45 D362-D368. DOI: 10.1093/nar/gkw937 75. Shin V.Y., Jin H., Ng E.K., Cheng A.S., Chong W.W., Wong C.Y., Leung W.K., Sung J.J., Chu K.M.. NF-kappaB targets miR-16 and miR-21 in gastric cancer: Involvement of prostaglandin E receptors. Carcinogenesis (2011) 32 240-245. DOI: 10.1093/carcin/bgq240 76. Mu N., Gu J., Huang T., Zhang C., Shu Z., Li M., Hao Q., Li W., Zhang W., Zhao J.. A novel NF-kappaB/YY1/microRNA-10a regulatory circuit in fibroblast-like synoviocytes regulates inflammation in rheumatoid arthritis. Sci. Rep. (2016) 6 20059. DOI: 10.1038/srep20059 77. Toury L., Frankel D., Airault C., Magdinier F., Roll P., Kaspi E.. miR-140-5p and miR-140-3p: Key Actors in Aging-Related Diseases?. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms231911439 78. Jiang Z., Zhang J., Chen F., Sun Y.. MiR-148b suppressed non-small cell lung cancer progression via inhibiting ALCAM through the NF-kappaB signaling pathway. Thorac. Cancer (2020) 11 415-425. DOI: 10.1111/1759-7714.13285 79. Zhu B., Ye J., Ashraf U., Li Y., Chen H., Song Y., Cao S.. Transcriptional regulation of miR-15b by c-Rel and CREB in Japanese encephalitis virus infection. Sci. Rep. (2016) 6 22581. DOI: 10.1038/srep22581 80. Zhao X., Deng R., Wang Y., Zhang H., Dou J., Li L., Du Y., Chen R., Cheng J., Yu J.. Twist1/Dnmt3a and miR186 establish a regulatory circuit that controls inflammation-associated prostate cancer progression. Oncogenesis (2017) 6 e315. DOI: 10.1038/oncsis.2017.16 81. Prattichizzo F., Micolucci L., Cricca M., De Carolis S., Mensa E., Ceriello A., Procopio A.D., Bonafe M., Olivieri F.. Exosome-based immunomodulation during aging: A nano-perspective on inflamm-aging. Mech. Ageing Dev. (2017) 168 44-53. DOI: 10.1016/j.mad.2017.02.008 82. Kim M., Jo H., Kwon Y., Jeong M.S., Jung H.S., Kim Y., Jeoung D.. MiR-154-5p-MCP1 Axis Regulates Allergic Inflammation by Mediating Cellular Interactions. Front. Immunol. (2021) 12 663726. DOI: 10.3389/fimmu.2021.663726 83. Liu W.Y., Yang Z., Sun Q., Yang X., Hu Y., Xie H., Gao H.J., Guo L.M., Yi J.Y., Liu M.. miR-377-3p drives malignancy characteristics via upregulating GSK-3beta expression and activating NF-kappaB pathway in hCRC cells. J. Cell Biochem. (2018) 119 2124-2134. DOI: 10.1002/jcb.26374 84. Sole C., Domingo S., Penzo E., Moline T., Porres L., Aparicio G., Ferrer B., Cortes-Hernandez J.. Downregulation of miR-885-5p Promotes NF-kappaB Pathway Activation and Immune Recruitment in Cutaneous Lupus Erythematosus. J. Investig. Dermatol. (2023) 143 209-219.e213. DOI: 10.1016/j.jid.2022.08.036 85. Fukata M., Vamadevan A.S., Abreu M.T.. Toll-like receptors (TLRs) and Nod-like receptors (NLRs) in inflammatory disorders. Semin. Immunol. (2009) 21 242-253. DOI: 10.1016/j.smim.2009.06.005 86. Salminen A., Kauppinen A., Kaarniranta K.. Emerging role of NF-kappaB signaling in the induction of senescence-associated secretory phenotype (SASP). Cell Signal. (2012) 24 835-845. DOI: 10.1016/j.cellsig.2011.12.006 87. Hayden M.S., Ghosh S.. NF-kappaB, the first quarter-century: Remarkable progress and outstanding questions. Genes Dev. (2012) 26 203-234. DOI: 10.1101/gad.183434.111 88. Garcia-Garcia V.A., Alameda J.P., Page A., Casanova M.L.. Role of NF-kappaB in Ageing and Age-Related Diseases: Lessons from Genetically Modified Mouse Models. Cells (2021) 10. DOI: 10.3390/cells10081906 89. Bass V.L., Wong V.C., Bullock M.E., Gaudet S., Miller-Jensen K.. TNF stimulation primarily modulates transcriptional burst size of NF-kappaB-regulated genes. Mol. Syst. Biol. (2021) 17 e10127. DOI: 10.15252/msb.202010127 90. Kizilirmak C., Bianchi M.E., Zambrano S.. Insights on the NF-kappaB System Using Live Cell Imaging: Recent Developments and Future Perspectives. Front. Immunol. (2022) 13 886127. DOI: 10.3389/fimmu.2022.886127 91. Liu X., Shao Y., Zhou J., Qian G., Ma Z.. Nuclear Factor kappaB Signaling and Its Related Non-coding RNAs in Cancer Therapy. Mol. Ther. Nucleic Acids (2020) 19 208-217. DOI: 10.1016/j.omtn.2019.11.007 92. Franceschi C., Bonafe M., Valensin S., Olivieri F., De Luca M., Ottaviani E., De Benedictis G.. Inflamm-aging. An evolutionary perspective on immunosenescence. Ann. N. Y. Acad. Sci. (2000) 908 244-254. DOI: 10.1111/j.1749-6632.2000.tb06651.x 93. Rodier F., Coppe J.P., Patil C.K., Hoeijmakers W.A., Munoz D.P., Raza S.R., Freund A., Campeau E., Davalos A.R., Campisi J.. Persistent DNA damage signalling triggers senescence-associated inflammatory cytokine secretion. Nat. Cell Biol. (2009) 11 973-979. DOI: 10.1038/ncb1909 94. Laberge R.M., Zhou L., Sarantos M.R., Rodier F., Freund A., de Keizer P.L., Liu S., Demaria M., Cong Y.S., Kapahi P.. Glucocorticoids suppress selected components of the senescence-associated secretory phenotype. Aging Cell (2012) 11 569-578. DOI: 10.1111/j.1474-9726.2012.00818.x 95. Freund A., Orjalo A.V., Desprez P.Y., Campisi J.. Inflammatory networks during cellular senescence: Causes and consequences. Trends Mol. Med. (2010) 16 238-246. DOI: 10.1016/j.molmed.2010.03.003 96. Fulop T., Larbi A., Dupuis G., Le Page A., Frost E.H., Cohen A.A., Witkowski J.M., Franceschi C.. Immunosenescence and Inflamm-Aging As Two Sides of the Same Coin: Friends or Foes?. Front. Immunol. (2017) 8 1960. DOI: 10.3389/fimmu.2017.01960 97. Songkiatisak P., Rahman S.M.T., Aqdas M., Sung M.H.. NF-kappaB, a culprit of both inflamm-ageing and declining immunity?. Immun. Ageing (2022) 19 20. DOI: 10.1186/s12979-022-00277-w 98. Josephson A.M., Leclerc K., Remark L.H., Lopez E.M., Leucht P.. Systemic NF-kappaB-mediated inflammation promotes an aging phenotype in skeletal stem/progenitor cells. Aging (2021) 13 13421-13429. DOI: 10.18632/aging.203083 99. Haga M., Okada M.. Systems approaches to investigate the role of NF-kappaB signaling in aging. Biochem. J. (2022) 479 161-183. DOI: 10.1042/BCJ20210547 100. Hounkpe B.W., Chenou F., de Lima F., De Paula E.V.. HRT Atlas v1.0 database: Redefining human and mouse housekeeping genes and candidate reference transcripts by mining massive RNA-seq datasets. Nucleic Acids Res. (2021) 49 D947-D955. DOI: 10.1093/nar/gkaa609 101. Bracken C.P., Scott H.S., Goodall G.J.. A network-biology perspective of microRNA function and dysfunction in cancer. Nat. Rev. Genet. (2016) 17 719-732. DOI: 10.1038/nrg.2016.134 102. Inui M., Martello G., Piccolo S.. MicroRNA control of signal transduction. Nat. Rev. Mol. Cell Biol. (2010) 11 252-263. DOI: 10.1038/nrm2868 103. Xie G.Y., Xia M., Miao Y.R., Luo M., Zhang Q., Guo A.Y.. FFLtool: A web server for transcription factor and miRNA feed forward loop analysis in human. Bioinformatics (2020) 36 2605-2607. DOI: 10.1093/bioinformatics/btz929 104. Ye H., Liu X., Lv M., Wu Y., Kuang S., Gong J., Yuan P., Zhong Z., Li Q., Jia H.. MicroRNA and transcription factor co-regulatory network analysis reveals miR-19 inhibits CYLD in T-cell acute lymphoblastic leukemia. Nucleic Acids Res. (2012) 40 5201-5214. DOI: 10.1093/nar/gks175 105. Schmitz M.L., Shaban M.S., Albert B.V., Gokcen A., Kracht M.. The Crosstalk of Endoplasmic Reticulum (ER) Stress Pathways with NF-kappaB: Complex Mechanisms Relevant for Cancer, Inflammation and Infection. Biomedicines (2018) 6. DOI: 10.3390/biomedicines6020058 106. Le N.T., Sandhu U.G., Quintana-Quezada R.A., Hoang N.M., Fujiwara K., Abe J.I.. Flow signaling and atherosclerosis. Cell Mol. Life Sci. (2017) 74 1835-1858. DOI: 10.1007/s00018-016-2442-4 107. Dolfini D., Gatta R., Mantovani R.. NF-Y and the transcriptional activation of CCAAT promoters. Crit. Rev. Biochem. Mol. Biol. (2012) 47 29-49. DOI: 10.3109/10409238.2011.628970 108. Vaiman D., Calicchio R., Miralles F.. Landscape of transcriptional deregulations in the preeclamptic placenta. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0065498 109. Mato-Basalo R., Morente-Lopez M., Arntz O.J., van de Loo F.A.J., Fafian-Labora J., Arufe M.C.. Therapeutic Potential for Regulation of the Nuclear Factor Kappa-B Transcription Factor p65 to Prevent Cellular Senescence and Activation of Pro-Inflammatory in Mesenchymal Stem Cells. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22073367 110. Slota J.A., Booth S.A.. MicroRNAs in Neuroinflammation: Implications in Disease Pathogenesis, Biomarker Discovery and Therapeutic Applications. Noncoding RNA (2019) 5. DOI: 10.3390/ncrna5020035 111. Grants J.M., Wegrzyn J., Hui T., O’Neill K., Shadbolt M., Knapp D., Parker J., Deng Y., Gopal A., Docking T.R.. Altered microRNA expression links IL6 and TNF-induced inflammaging with myeloid malignancy in humans and mice. Blood (2020) 135 2235-2251. DOI: 10.1182/blood.2019003105 112. Mann M., Mehta A., Zhao J.L., Lee K., Marinov G.K., Garcia-Flores Y., Lu L.F., Rudensky A.Y., Baltimore D.. An NF-kappaB-microRNA regulatory network tunes macrophage inflammatory responses. Nat. Commun. (2017) 8 851. DOI: 10.1038/s41467-017-00972-z 113. Khalife J., Ghose J., Martella M., Viola D., Rocci A., Troadec E., Terrazas C., Satoskar A.R., Gunes E.G., Dona A.. MiR-16 regulates crosstalk in NF-kappaB tolerogenic inflammatory signaling between myeloma cells and bone marrow macrophages. JCI Insight (2019) 4. DOI: 10.1172/jci.insight.129348 114. Mensa E., Guescini M., Giuliani A., Bacalini M.G., Ramini D., Corleone G., Ferracin M., Fulgenzi G., Graciotti L., Prattichizzo F.. Small extracellular vesicles deliver miR-21 and miR-217 as pro-senescence effectors to endothelial cells. J. Extracell. Vesicles (2020) 9 1725285. DOI: 10.1080/20013078.2020.1725285 115. Fujioka S., Niu J., Schmidt C., Sclabas G.M., Peng B., Uwagawa T., Li Z., Evans D.B., Abbruzzese J.L., Chiao P.J.. NF-kappaB and AP-1 connection: Mechanism of NF-kappaB-dependent regulation of AP-1 activity. Mol. Cell Biol. (2004) 24 7806-7819. DOI: 10.1128/MCB.24.17.7806-7819.2004 116. Del Mar Diaz-Gonzalez S., Rodriguez-Aguilar E.D., Meneses-Acosta A., Valadez-Graham V., Deas J., Gomez-Ceron C., Tavira-Montalvan C.A., Arizmendi-Heras A., Ramirez-Bello J., Zurita-Ortega M.E.. Transregulation of microRNA miR-21 promoter by AP-1 transcription factor in cervical cancer cells. Cancer Cell Int. (2019) 19 214. DOI: 10.1186/s12935-019-0931-x 117. Olivieri F., Capri M., Bonafe M., Morsiani C., Jung H.J., Spazzafumo L., Vina J., Suh Y.. Circulating miRNAs and miRNA shuttles as biomarkers: Perspective trajectories of healthy and unhealthy aging. Mech. Ageing Dev. (2017) 165 162-170. DOI: 10.1016/j.mad.2016.12.004 118. Zhavoronkov A., Li R., Ma C., Mamoshina P.. Deep biomarkers of aging and longevity: From research to applications. Aging (2019) 11 10771-10780. DOI: 10.18632/aging.102475 119. Hamdan Y., Mazini L., Malka G.. Exosomes and Micro-RNAs in Aging Process. Biomedicines (2021) 9. DOI: 10.3390/biomedicines9080968 120. Kinser H.E., Pincus Z.. MicroRNAs as modulators of longevity and the aging process. Hum. Genet. (2020) 139 291-308. DOI: 10.1007/s00439-019-02046-0 121. Rovcanin Dragovic I., Popovic N., Zdralevic M., Radulovic L., Vukovic T., Marzano F., Tullo A., Radunovic M.. Inflammation-Related microRNAs-146a and -155 Are Upregulated in Mild Cognitive Impairment Subjects Among Older Age Population in Montenegro. J. Alzheimers Dis. (2022) 90 625-638. DOI: 10.3233/JAD-220676 122. Sahu H., Shrma S., Gondhalakar S.. A Brief Overview on Data Mining Survey. Int. J. Comput. Technol. Electron. Eng. (2011) 1 114-121 123. Kozomara A., Griffiths-Jones S.. miRBase: Integrating microRNA annotation and deep-sequencing data. Nucleic Acids Res. (2011) 39 D152-D157. DOI: 10.1093/nar/gkq1027 124. Griffiths-Jones S., Saini H.K., van Dongen S., Enright A.J.. miRBase: Tools for microRNA genomics. Nucleic Acids Res. (2008) 36 D154-D158. DOI: 10.1093/nar/gkm952 125. Griffiths-Jones S., Grocock R.J., van Dongen S., Bateman A., Enright A.J.. miRBase: microRNA sequences, targets and gene nomenclature. Nucleic Acids Res. (2006) 34 D140-D144. DOI: 10.1093/nar/gkj112 126. Griffiths-Jones S.. The microRNA Registry. Nucleic Acids Res. (2004) 32 D109-D111. DOI: 10.1093/nar/gkh023 127. Landgraf P., Rusu M., Sheridan R., Sewer A., Iovino N., Aravin A., Pfeffer S., Rice A., Kamphorst A.O., Landthaler M.. A mammalian microRNA expression atlas based on small RNA library sequencing. Cell (2007) 129 1401-1414. DOI: 10.1016/j.cell.2007.04.040 128. Hinske L.C., Franca G.S., Torres H.A., Ohara D.T., Lopes-Ramos C.M., Heyn J., Reis L.F., Ohno-Machado L., Kreth S., Galante P.A.. miRIAD-integrating microRNA inter- and intragenic data. Database (2014) 2014. DOI: 10.1093/database/bau099 129. UniProt C.. UniProt: The Universal Protein Knowledgebase in 2023. Nucleic Acids Res. (2023) 51 D523-D531. DOI: 10.1093/nar/gkac1052 130. Karagkouni D., Paraskevopoulou M.D., Chatzopoulos S., Vlachos I.S., Tastsoglou S., Kanellos I., Papadimitriou D., Kavakiotis I., Maniou S., Skoufos G.. DIANA-TarBase v8: A decade-long collection of experimentally supported miRNA-gene interactions. Nucleic Acids Res. (2018) 46 D239-D245. DOI: 10.1093/nar/gkx1141 131. Han H., Cho J.W., Lee S., Yun A., Kim H., Bae D., Yang S., Kim C.Y., Lee M., Kim E.. TRRUST v2: An expanded reference database of human and mouse transcriptional regulatory interactions. Nucleic Acids Res. (2018) 46 D380-D386. DOI: 10.1093/nar/gkx1013 132. Szklarczyk D., Franceschini A., Wyder S., Forslund K., Heller D., Huerta-Cepas J., Simonovic M., Roth A., Santos A., Tsafou K.P.. STRING v10: Protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. (2015) 43 D447-D452. DOI: 10.1093/nar/gku1003 133. Fan Y., Siklenka K., Arora S.K., Ribeiro P., Kimmins S., Xia J.. miRNet—Dissecting miRNA-target interactions and functional associations through network-based visual analysis. Nucleic Acids Res. (2016) 44 W135-W141. DOI: 10.1093/nar/gkw288 134. Aging Atlas C.. Aging Atlas: A multi-omics database for aging biology. Nucleic Acids Res. (2021) 49 D825-D830. DOI: 10.1093/nar/gkaa894 135. Kramer A., Green J., Pollard J., Tugendreich S.. Causal analysis approaches in Ingenuity Pathway Analysis. Bioinformatics (2014) 30 523-530. DOI: 10.1093/bioinformatics/btt703 136. Russo F., Di Bella S., Vannini F., Berti G., Scoyni F., Cook H.V., Santos A., Nigita G., Bonnici V., Lagana A.. miRandola 2017: A curated knowledge base of non-invasive biomarkers. Nucleic Acids Res. (2018) 46 D354-D359. DOI: 10.1093/nar/gkx854
--- title: 'Water from Nitrodi’s Spring Induces Dermal Fibroblast and Keratinocyte Activation, Thus Promoting Wound Repair in the Skin: An In Vitro Study' authors: - Filomena Napolitano - Loredana Postiglione - Ilaria Mormile - Valentina Barrella - Amato de Paulis - Nunzia Montuori - Francesca Wanda Rossi journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049109 doi: 10.3390/ijms24065357 license: CC BY 4.0 --- # Water from Nitrodi’s Spring Induces Dermal Fibroblast and Keratinocyte Activation, Thus Promoting Wound Repair in the Skin: An In Vitro Study ## Abstract The Romans knew of Nitrodi’s spring on the island of Ischia more than 2000 years ago. Although the health benefits attributed to Nitrodi’s water are numerous, the underlying mechanisms are still not understood. In this study, we aim to analyze the physicochemical properties and biological effects of Nitrodi’s water on human dermal fibroblasts to determine whether the water exerts in vitro effects that could be relevant to skin wound healing. The results obtained from the study indicate that Nitrodi’s water exerts strong promotional effects on dermal fibroblast viability and a significant stimulatory activity on cell migration. Nitrodi’s water induces alpha-SMA expression in dermal fibroblasts, thus promoting their transition to myofibroblast-protein ECM deposition. Furthermore, Nitrodi’s water reduces intracellular reactive oxygen species (ROS), which play an important role in human skin aging and dermal damage. Unsurprisingly, Nitrodi’s water has significant stimulatory effects on the cell proliferation of epidermal keratinocytes and inhibits the basal ROS production but enhances their response to the oxidative stress caused by external stimuli. Our results will contribute to the development of human clinical trials and further in vitro studies to identify inorganic and/or organic compounds responsible for pharmacological effects. ## 1. Introduction The skin is the largest organ and functions as a protective barrier against environmental insults such as pathogens, chemicals, physical agents, and solar UVR. Moreover, the skin performs physiological functions such as immune defense, thermoregulation, sensing, endocrine functions, as well as metabolic effects [1,2,3]. The skin consists of three distinct layers: the epidermis, the dermis, and the hypodermis or subcutaneous tissue. The epidermis, the outermost level, is mainly made up of keratinocytes and contains melanocytes, Langerhans cells, and Merkel cells. The dermis, the internal layer that provides structural integrity, elasticity, and nutrition, is a connective tissue enriched with collagen and elastic fibers of a high fibroblast density; it also contains blood and lymphatic vessels, sebaceous glands, sweat glands, nerve endings, and hair follicles invaginated from the epidermis. The hypodermis or subcutaneous tissue consists of well-vascularized, loose, areolar connective tissue and adipose tissue functioning as a fat storage [4,5]. The integrity of the skin, constantly challenged by a wide variety of external factors, plays a pivotal role in maintaining physiological homeostasis and is of utmost importance for the viability of the inner tissues [6]. If left untreated, improper wound healing may lead to major disability or even death. Therefore, the proper, fast, and complete healing of wounds is of high priority for the viability of internal organs and thus for the survival of the organism. It is important to underscore that the dermal fibroblasts represent the main regulators of skin homeostasis by interacting with the epidermis and other resident dermal cells, such as adipocytes, endothelial, neural, and inflammatory cells. Moreover, dermal fibroblasts are involved in various physio-pathological conditions, including wound healing, fibrosis, aging, and skin cancer [7,8]. Wound repair, a well-tuned biological process, occurs in three overlapping stages: inflammation, proliferation (including re-epithelialization, granulation tissue formation, and neovascularization), and remodeling [9]. Inflammation is the first stage of wound repair, which occurs immediately after tissue damage. The second stage—new tissue formation—is characterized by the activation of fibroblasts, some of which differentiate into myofibroblasts [10]. Myofibroblasts, characterized by the expression of alpha-smooth muscle actin (alpha-SMA), are contractile cells that, over time, start proliferating, migrate into the wound, and form the “granulation tissue”, which is rich in extracellular matrix (ECM) proteins. As fibroblasts begin to produce a new ECM, epithelialization is initiated. The epithelialization of wounds involves an orderly series of events in which keratinocytes migrate, proliferate, and differentiate to restore the epithelial barrier function [11]. Myofibroblasts support the growth of new blood vessels, reconstituting the wound bed and bringing the edges of a wound together [10]. Dysregulation in any phase of the wound healing cascade delays healing and may result in various skin pathologies, including non-healing or chronic ulceration [12,13]. The skin, as a critical protective barrier against the outside world, is also vulnerable to oxidative stress, potentially contributing to skin disorders ranging from functional impairments (skin cancer, dermatitis, and chronic and acute inflammatory diseases) to defects of aesthetic character, due to destruction of structural proteins and cellular changes, with the appearance of marks and lines of expressions and other signs typical of skin aging processes [14]. In particular, increased ROS levels determine DNA damage, inflammation, and the generation of matrix metalloproteinases (MMPs) that degrade collagen and elastin in the dermal skin layer [15,16]. Therefore, targeting oxidative stress may be an effective strategy for the anti-aging process and various skin disorders [17,18]. There is a growing body of evidence for the potential application of natural products (water, plants, and minerals), in addition to conventional treatment, to the enhancement of both acute and chronic wound healing [19,20,21,22]. We found it interesting that Nitrodi’s water affects the inflammatory microenvironment and enhances wound healing at the cellular level. The Nitrodi’s spring on the island of Ischia was known to Romans more than 2000 years ago as demonstrated by several marble votive reliefs found in this location, dated between the first century B.C. and the second century A.C. [23]. It has been found that in the spring, there was a school of medical hydrology attended by prominent physicians, such as Menippo, Aurelius Monnus, and Numerius Fabius [24]. Nitrodi’s water is classified as a mineral, hypothermal, sulfate alkaline and was extensively studied in the late 60s to evaluate the different modalities of exposition to the water. In particular, all studies showed that both internal and external use provided several benefits in the treatment of certain ailments [24], such as wound healing. The beneficial effects of Nitrodi’s water can be observed particularly in the treatment of dermatitis and in almost all skin-related illnesses. Contact with this water is particularly beneficial for people suffering from venous ulcers because it acts on the nutritional processes of the tissues; hence, it is recommended for the treatment of lesions, fistulas, furuncles, burns, and ulcerative colitis. However, despite its long history, the biological mechanisms underlying the therapeutic effects on wound healing have not yet been elucidated, nor have clinical trials been performed to demonstrate its beneficial effects on patients with rheumatic or dermatologic disorders. Recognizing the important role that traditional medicine continues to play, we have undertaken a preliminary in vitro study to clarify the biochemical and molecular mechanisms underlying the beneficial effects of Nitrodi’s spring water on skin inflammation and wound restoration. ## 2.1. Composition of Nitrodi’s Water Table 1 represents the main physical properties and mineral and organic composition of Nitrodi’s water. The pH evaluation showed that Nitrodi’s water was slightly acidic, with a mean value of 6.33 on three measurements. Concerning the chemical composition of analyzed samples, the results indicated a medium mineral content, according to the Marotta and Sica classification [25,26]. The main components found were chloride (Cl−) with a mean value of 93 mg/L, sulfates (SO42−) with a mean value of 204 mg/L, sodium (Na+) with a mean value of 174 mg/L, and calcium (Ca2+) with a mean value of 137 mg/L. These results obtained were comparable with those reported by Aversano, [27], which indicated an amount of 0.2200 g/L (220 mg/L), 0.1676 g/L (167.6 mg/L), and 0.1182 g/L (118.2 mg/L), for SO42−, Na+, and Ca2+, respectively. However, the HCO3− concentration obtained in this study (561 mg/L) was higher than that reported by Aversano (475.8 mg/L). Moreover, for bromide (Br−), the amount detected (0.18 mg/L) was in accordance with that reported by Inguaggiato (0.1–0.2 mg/L) [28]. On the other hand, the ammonium (NH4+) and total phosphorus (P) analyses showed values below LOQs, as well as for dissolved iron (Fe2+, Fe3+), and aluminum (Al3+). In addition to the parameters reported in Table 1, further verification analyses were carried out; the analyses included the evaluation of Polycyclic Aromatic Hydrocarbons (PAHs), Organochlorines (OCLs), Organophosphorus Pesticides (OPPs), Polychlorinated Biphenyls (PCBs), and Trihalomethanes (THMs), as determined by extraction and mass spectrometry analyses. The analyses showed for all analyte concentration values below LOQs. Therefore, according to the classification proposed by Nasermoaddeli and Kagamimori, based on its chemical composition and temperature, Nitrodi’s water was classified as mildly mineralized and hypothermal (20–30 °C) water [29]. In addition, as proposed by Marotta and Sica in the most widely accredited classification in Italy [25,26], as well as also by Mancioli [24], the results indicated that Nitrodi’s water falls into the category of medium-mineral waters of an essentially bicarbonate-sulfate-alkaline, alkaline-earthy, and hypothermal nature. ## 2.2. Nitrodi’s Water Sustains Dermal Fibroblast Viability via ERK Signaling Pathway Dermal fibroblast proliferation plays a central role in the skin-repair process, inducing re-epithelialization through the replacement of disorganized collagen and elastin structures and the reposition of the extracellular matrix (ECM) in aged skin [30]. We assumed that the beneficial effects of Nitrodi’s water were, at least in part, due to the enhancement of dermal fibroblast activity. To test this hypothesis, we used BJ cells from the skin of normal foreskin as a cell model of the human skin. As a preliminary step, we analyzed the effects of Nitrodi’s water by Western blot on the activation of extracellular signal-regulated protein kinases 1 and 2 (ERK $\frac{1}{2}$), which are members of the mitogen-activated protein kinase super-family that regulates growth and cell-cycle progression. Serum-starved BJ cells were treated with PBS alone (as a negative control), Nitrodi, or fMLF (as a positive control) for ERK $\frac{1}{2}$ activation [31,32,33]. Figure 1A shows that the incubation with Nitrodi for 5 min significantly increased the phosphorylation of ERK 2, but not of ERK 1, thus suggesting that Nitrodi’s water could act on BJ cell proliferation and cell-cycle progression. Subsequently, we performed proliferation assays to analyze the effects of Nitrodi’s water on the rate of dermal fibroblast growth. BJ cells were incubated with Nitrodi supplemented with $0.5\%$ BSA or $10\%$ FCS, PBS supplemented with $0.5\%$ BSA or $10\%$ FCS, and culture medium supplemented with $0.5\%$ BSA or $10\%$ FCS, as negative and positive controls, respectively. The cell viability was measured at 0, 24, 48, and 144 h. These different time points have been used to simulate in vitro a hypothetical in vivo treatment of patients affected by skin failure syndrome. As shown in Figure 1B, Nitrodi’s water is able to sustain BJ cell survival, both in starving and growing conditions. This result was in agreement with ERK 2 induction by Nitrodi’s water, as observed in the previous experiment. An in-depth analysis of the proliferative course throughout the 144 h showed that culture medium-treated cells with $10\%$ FCS produced an early peak at 24 h (O.D. 1.51 value) compared to T0 (O.D. 1.08 value), followed by a plateau of activity that persisted for 144 h. Nitrodi’s water-treated cells with $10\%$ FCS exhibited a slight proliferative increase at 24 h (O.D. 0.81 value) and a peak at 48 h (O.D. 0.95 value) compared to T0 (O.D. 0.71 value). The mineral and ionic composition of Nitrodi’s water, supplemented with FCS, is able to deliver the ions, the nutrients, and the soluble factors essential for the maintenance of cell adherence and vitality. In fact, a decrease of cell count in PBS $10\%$ FCS-treated cells was observed at the time points (at 48 h O.D. 0.601 value; at 144 h O.D. 0.56 value) subsequent to 24 h (O.D. 0.74 value) as expected as the lack of ions could induce cell detachment. To better understand the role of ERKs in the cell proliferation, we analyzed the effects of PD98059, a potent and selective cell-permeable inhibitor of ERKs, on BJ proliferation. Serum-starved BJ cells were pre-incubated with PD98059 (50 µM) and then exposed to PBS, Nitrodi, and culture medium supplemented both $0.5\%$ BSA and $10\%$ FCS. The cell viability was measured after 24 h. As reported in Figure 1C, the cells treated with Nitrodi’s water and culture medium exhibited a significant reduction of proliferation in the presence of PD98059, both in the starving condition and in the growing condition. The inhibition of cell proliferation by PD98059 in culture medium-treated cells showed that these cells in physiologic conditions need ERK activation to proliferate. The same effects were observed when BJ cells, pre-treated with PD98059, were exposed to Nitrodi’s water. This result demonstrates that Nitrodi’s water does not alter or modify intracellular signaling pathways that regulate BJ proliferation. All together, these preliminary data supported our hypothesis that Nitrodi’s water can offer several benefits for skin disorders by stimulating fibroblast survival via the MAPK/ERK signaling pathway. The decrease in cell count in the PBS negative control could be explained by the absence of ions, in particular, calcium and magnesium. This observation supports the hypothesis that the effects of Nitrodi’s water on fibroblast survival and viability are due to the mineral content, and further, future studies on the Nitrodi’s water microbiome will allow us to investigate its ability to stimulate skin cell proliferation. ## 2.3. Nitrodi’s Water Promoted In Vitro Wound Scratch Closure in Dermal Fibroblasts Wound healing is a complex process modulated by an intricate signaling network involving several cell types, numerous growth factors, cytokines, and chemokines [34]. Successful healing depends on the optimal functioning of many diverse processes that lead to the generation of new tissue. The failure of any of these steps results in chronic inflammation. Fibroblasts play a key role in wound healing and tissue repair because the proliferation and subsequent migration of fibroblasts are of paramount importance for wound repair and healing [35]. To analyze the effects of Nitrodi’s water on in vitro wound-scratch closure, we first tested its capability to induce directional migration (i.e., cell motility in one direction) after 24 h of treatment through a chemotaxis assay in BJ cells. The study of fibroblast migration was performed in a Boyden modified chamber using $5\%$ FCS as a generic chemoattractant. In these assays, the membrane coating with fibronectin also allowed us to investigate the capacity of fibroblasts to enter the wound matrix in response to Nitrodi’s water. Figure 2A shows that BJ cells treated with Nitrodi were readily able to migrate towards $5\%$ FCS and to invade through fibronectin-coated membrane as compared to cells treated with PBS alone. To clarify which signaling pathways are involved in directional fibroblast migration promoted by Nitrodi’s water, we further investigated the respective roles of ERKs and Rac1 as the main signaling mediators involved in cell migration. To this end, the pre-treatment of BJ cells with the specific inhibitors of ERKs (PD98059) and Rac1 (NSC23766) was carried out for 2 h at 37 °C. Thus, chemotaxis assay was performed. As shown in Figure 2B, when the cells were pre-treated with PD98059 (50 µM) and then exposed to Nitrodi’s water, the migration ability was obviously impaired. Conversely, the pretreatment with NSC23766 (25 µM), a widely used inhibitor of Rac1 activation, did not exert any effects on Nitrodi’s water-accelerated fibroblast migration. These data indicated that ERK pathway plays a crucial role in fibroblast chemotaxis promoted by Nitrodi’s water, as already demonstrated in the previous analysis of fibroblast proliferation. After the investigation of fibroblast unidirectional migration through chemotaxis assay, we examined the effects of Nitrodi’s water on wound-healing scratch assay, as a model of non-oriented cell migration. After scratching, BJ cells were incubated for 24 h with PBS, Nitrodi, Nitrodi’s water containing PD98059 (50 µM), and Nitrodi’s water containing NSC23766 (25 µM). Figure 2C shows the images acquired under the microscope at 4x magnification over time and the graph containing a data analysis of the images. In the graph, BJ cell migration was expressed as the percent of wound length over T0 (assumed as $100\%$). Nitrodi’s water alone enhanced BJ cell migration at 24 h as compared to cells treated with PBS, whereas the presence of ERK and Rac1 inhibitors impaired the migration/invasion of BJ cells into the wound. These results indicate that Nitrodi’s water promotes cell motility through both ERKs and Rac1 in the non-oriented migration assay, whereas ERK activation is preferentially required for unidirectional migration (chemotaxis assay) towards chemoattractant [36,37]. Finally, Nitrodi’s water not only exerts a strong chemotactic effect on BJ cells through the promotion of unidirectional migration but also induces a significant improvement in wound scratch closure. ## 2.4. Nitrodi’s Water Promoted Dermal Fibroblast Differentiation through Alpha-SMA Induction Tissue damage activates fibroblasts and differentiates them into myofibroblasts. These cells can contract and actively produce ECM proteins to enable wound closure [38]. Alpha smooth-muscle actin (alpha-SMA) expression is a marker of myofibroblast differentiation [39]. Thus, we investigated the impact of the Nitrodi’s water treatment on BJ cell transition to myofibroblasts. To this end, we used a Western blot analysis to evaluate the expression levels of alpha-SMA in BJ cells after incubation with PBS as a vehicle control and Nitrodi and fMLF (10−4 M) as a positive control for alpha-SMA induction [31,40]. As shown in Figure 2D, the alpha-SMA expression level increased significantly in BJ cells in response to Nitrodi compared to PBS-treated controls. Our data indicated that Nitrodi’s water induced the differentiation of BJ cells into myofibroblasts, which are necessary to optimize skin repair. ## 2.5. Nitrodi’s Water Promotes ECM Protein Deposition in Dermal Fibroblasts Upon injury, dermal fibroblasts migrate into wound granulation tissue and differentiate into myofibroblasts, which play a central role in the wound contraction and deposition of ECM proteins. To investigate the possibility that the synthesis of ECM could be promoted by Nitrodi’s water, we evaluated the deposition of vitronectin, fibronectin, and collagen type I by in situ ELISA in basal conditions (PBS-treated cells) and after treatment with Nitrodi’s water in BJ cells. As shown in Figure 3A, BJ cells exposed to Nitrodi’s water exhibited a significant increase in fibronectin deposition (7.06 ± 0.49 μg/mL) compared to the control (4.6 ± 1.58 μg/mL). No differences between the control (9.03 ± 0.80 μg/mL) and cells treated with Nitrodi’s water (8.6 ± 2.23 μg/mL) were detected in vitronectin deposition. A slightly increasing trend in cells treated with Nitrodi’s water (5.4 ± 1.15 μg/mL) was recorded in collagen deposition compared to control (3.7 ± 1.35 μg/mL). These data showed that Nitrodi’s water promoted the expression of fibronectin and, to a lesser extent, collagen, which are major proteins of ECM and are critical for both structural support and cell adhesion [41]. Instead, Nitrodi’s water did not exert any effect on the deposition of vitronectin, which is a provisional matrix component, mostly increased in patho-physiologic settings associated with acute inflammation [42]. Taken together, these results indicate that Nitrodi’s water contributes to the production of the main ECM components, particularly fibronectin, which plays a crucial role in ECM formation and in re-epithelialization during wound healing. ## 2.6. Nitrodi’s Water Exhibited Anti-Oxidant Properties in Dermal Fibroblasts Reactive oxygen species (ROS) play a central role in both chronological aging and photo-aging. Oxidative stress promotes tissue inflammation through the upregulation of genes that encode pro-inflammatory cytokines and the sustained activation of the NF-ΚB pathway. This low-grade chronic inflammation and the up-regulation of pro-inflammatory mediators are referred to as skin inflamm-aging [43,44,45,46]. Several natural compounds have been studied in order to evaluate their effects on skin inflamm-aging, but further investigations are needed. Here, we examined the effects of Nitrodi’s water on the ROS release from the BJ cell line, both in the absence and presence of oxidant stimuli. To this aim, BJ cells were loaded for 30 min with dichloro-dihydro-fluorescein diacetate (DCFH-DA), the most widely used probe for detecting intracellular oxidative stress, and then treated with PBS (as a negative control) or with Nitrodi for 5, 15, 30, and 60 min. Figure 3B shows that, at all different time points, BJ cells treated with Nitrodi’ water had significantly lower intracellular ROS levels compared to BJ cells treated with PBS, indicating that Nitrodi’s water can be considered as a promising source of natural antioxidants and is useful for the treatment of chronic inflammation and anti-aging strategies. Exogenous H2O2 is known to induce oxidative stress and cell death by oxidizing lipids and proteins [47]. Moreover, different cell types, such as fibroblasts, generate small amounts of ROS via the activation of NADPH oxidase in response to growth factors, cytokines, and G-protein-coupled receptor (GPCRs) agonists. Of importance, these endogenous ROS serve as second messengers to activate multiple intracellular signaling pathways essential to cell physiological responses, including the growth, migration, and modification of the ECM [48]. We have previously demonstrated that fMLF, a bacterial analog peptide that binds GPCRs, is able to induce ROS generation by NADPH oxidase complex activation, playing an important role in antimicrobial host defense and inflammation [31,40]. Thus, we performed an ROS detection assay in the presence of two different stimuli, H2O2, as an exogenous oxidative stress, and fMLF. As reported in Figure 3C, in H2O2-loaded BJ cells, Nitrodi’s water treatment exhibited a protective role against oxidative stress as compared to H2O2-loaded BJ cells in PBS solution as ROS levels were lower in Nitrodi-treated cells than in PBS-treated cells. Therefore, Nitrodi’s water could contribute to improving the oxidative response under stress conditions, thus, avoiding an uncontrolled ROS generation. Conversely, the presence of Nitrodi’s water ensured the ROS production in response to fMLF, demonstrating that Nitrodi’s water did not affect the maintenance of redox homeostasis in response to various cellular signaling pathways. ## 2.7. Nitrodi’s Water Promotes Multi-Oriented, but Not Directional, Epidermal Keratinocyte Migration Epidermal keratinocytes, the predominant cell type in the skin epidermis, are among the front line of skin defense. Keratinocytes contribute to ensuring efficient and harmonious wound healing through coordinated action with fibroblasts and immune cells. In particular, keratinocytes are the executors of the re-epithelialization phase, whereby keratinocytes migrate, proliferate, and differentiate to restore the epidermal tissue [49]. The initial phase of re-epithelialization results in keratinocyte migration from a surrounding tissue. Hence, we tested the capability of Nitrodi’s water to induce directional migration of HaCaT cells, as a model of human epidermal keratinocytes. Figure 4A shows that HaCaT cells treated with Nitrodi’s water exhibited a less vigorous migratory phenotype, as compared to control. Contrary to data obtained in BJ cells, it would seem that Nitrodi’s water exerts an inhibitory effect on HaCaT cell migration. Surprised by these results, we performed a wound-healing assay to analyze the effects of Nitrodi’s water on non-oriented migration of HaCaT cells. After scratching, HaCaT cells were incubated with PBS or Nitrodi for 24 and 72 h. Figure 4B shows the images acquired at different time points and related plots for wound-size quantification. Cell migration rate was expressed as a percentage of the length of wound size over T0 (assumed as $100\%$). Both cells treated with PBS and Nitrodi’s water exhibited the ability to invade the wound site, but our experiments illustrated how the two conditions, PBS/distilled water and PBS/Nitrodi’s water (equal in the solute but different in the solvent), influenced cell morphology and density differently. In light of these findings, we evaluated the effects of Nitrodi’s water and PBS on cell viability at 72 h through in situ Trypan blue staining. In Trypan blue assay, dead cells are stained because the dye cannot permeate the intact cell membrane. As shown in Figure 4C, Nitrodi’s water did not exhibit any cytotoxic effects on keratinocyte monolayer subjected to scratching, whereas the cells in PBS exhibited an intense trypan blue staining. Importantly, the results of these experiments showed that Nitrodi’s water did not alter the ability of epithelial cells to maintain close contact and continuity with each other and move forward in a collective way to restore the epithelial barrier at the wound site [50]. ## 2.8. Nitrodi’s Water Supports Epidermal Keratinocyte Viability and Survival Wound repair requires keratinocyte proliferation to restore the epithelial barrier [51]. Growth factors produced as a result of injury are released by several cell types to stimulate keratinocyte proliferation, and integrins on the keratinocyte surface enhance the accumulation of intracellular signaling mediators in order to enhance proliferation [52]. To evaluate whether Nitrodi’s water exerts any effect on keratinocyte proliferation, we performed in vitro analyses of cell proliferation in the HaCaT cell line. HaCaT cells were incubated with Nitrodi supplemented with $0.5\%$ BSA and $10\%$ FCS, or with PBS supplemented with $0.5\%$ BSA and $10\%$ FCS, and culture medium supplemented with $0.5\%$ BSA and $10\%$ FCS to reproduce the starving and growing condition, respectively. As shown in Figure 5, Nitrodi’s water had significant stimulatory effects on keratinocytes, as HaCaT cells treated with Nitrodi $0.5\%$ BSA showed a significant increase in cell proliferation as compared to BJ cells treated with PBS $0.5\%$ BSA. This significant proliferation increase was confirmed by a comparison of the treatment with Nitrodi $10\%$ FCS and HaCaT cells treated with PBS $10\%$ FCS. The increase in cell proliferation induced by Nitrodi’s water was observed when Nitrodi-treated cells were compared to the negative control (PBS), but not to the positive control (culture medium). It is important to underline that in the culture medium, there are vitamins, nutrients, and glucose that cells need to survive and proliferate, unlike in Nitrodi’s water. Despite this, during the 72 h curve, no significant decrease in cell number was observed in Nitrodi-treated cells as opposed to PBS. Therefore, a decrease in cell number was observed during the 72 h curve in both PBS $0.5\%$ BSA (0.38 O.D. vs. 0.13 O.D.) and PBS $10\%$ FCS (0.43 O.D. vs. 0.24 O.D.). On the contrary, cell survival was slightly affected in Nitrodi $0.5\%$ BSA (0.47 O.D. vs. 0.27 O.D.) and was not affected in Nitrodi $10\%$ FCS (0.59 O.D. vs. 0.53 O.D.). The conclusion would be that Nitrodi’s water supports cell survival in keratinocytes, and the beneficial effects are probably due to mineral content and soluble factors that are still unidentified. ## 2.9. Nitrodi’s Water Elicits an Acute Stress Response to Pro-Inflammatory Agents in Keratinocytes Cutaneous perturbations created by acute exposome factors (environmental and/or internal) induce responses to protect the organism and re-establish homeostasis [53]. Epidermal keratinocytes, which occupy the outermost layer of the skin, are always exposed to external stimuli, which constantly generate ROS in the cells [54]. Then, we analyzed the effects of Nitrodi’s water on oxidant activity in HaCaT cells. HaCaT cells were loaded for 30 min with DCFH-DA, and then treated with PBS (as a negative control) or with Nitrodi for 5, 15, 30, and 60 min. As shown in Figure 6A, in the absence of external stimuli, the measurement of intracellular ROS in Nitrodi-treated HaCaT cells revealed significantly decreased ROS levels compared to untreated cells, as already observed in BJ cells. Subsequently, we performed an ROS detection assay in the presence of H2O2 to cause oxidative damage stress, and we used a high concentration of fMLF (10−4 M) as a typical pro-inflammatory agent. HaCaT cells in PBS solution did not respond or responded poorly to H2O2 and fMLF, both at 5 and 15 min (Figure 6B). Instead, HaCaT cells treated with Nitrodi’s water early generated a significant amount of ROS in response to both H2O2 and fMLF, unlike what was observed in dermal fibroblasts (paragraph 2.6.). These data showed that Nitrodi’s water enhances the antioxidant defense-system potential under steady-state conditions, whereas it confers pro-oxidant phenotype in response to pro-inflammatory stimuli. ## 3. Discussion Bath therapy is an effective complementary approach to the management of several low-grade inflammation- and stress-related pathologies, particularly rheumatic and metabolic diseases [29]. However, despite the notable clinical benefits of these therapies, their role in modern medicine is still controversial, especially since the biological mechanisms underlying these benefits have not yet been completely clarified. Due to interesting peculiarities of water from Nitrodi’s spring, their clinical application is of great interest. Since Roman times, the therapeutic effectiveness of Nitrodi’s water has been known in various skin diseases, such as psoriasis, contact dermatitis, acne, and eczema [24]. This preliminary study aims to analyze the physicochemical composition and the biological effects of Nitrodi’s water at cellular level. To this end, Nitrodi’s water was taken from the historical thermal resort of Ischia, Italy. Firstly, the physicochemical analysis showed that Nitrodi’s water is a medium-mineral water, according to the Marotta and Sica [35,36]. The constituent mineral elements, such as silica, sulphates, sodium, and calcium, might be responsible for pharmacological actions. It is well-established that calcium plays a central role in wound healing and keratinocyte differentiation in the skin [55]. Nitrodi’s water contains conspicuous levels of calcium (137 mg/L), which could be responsible for benefits in the management of some skin diseases such as rosacea and psoriasis. Nitrodi’s water also contains silica (82 mg/L), which is involved in skin wound repair and plays a protective role on the connective tissue and cartilage. Silica might be also responsible for the suppression of pro-inflammatory cytokines [56]. Bicarbonate helps neutralize acid valences in cases of excessive muscle work; therefore, bicarbonate concentration [561 mg/L] in Nitrodi’s water could explain the clinical efficacy in the rehabilitation of the musculoskeletal. Moreover, the ability of Nitrodi’s water to improve skin regeneration could be due to calcium, magnesium, and bicarbonate, as previously established [57]. Sulfur exerts well-known anti-inflammatory and anti-oxidative activities in HaCaT cells [58]; therefore, the anti-inflammatory effect of Nitrodi’s water might be attributed to the sulfur content. In addition to an analysis of minerals dissolved in Nitrodi’s water, the evaluation of the organic fraction was performed, showing values below LOQs for PAHs, OCLs, OPPs, PCBs, and THMs. These data allowed us to assess the absence of toxicity risks since the absorption of these compounds into the skin, into the oral mucosa, or by inhalation is considerable [59]. The biochemical and molecular characteristics of the water from Nitrodi’s spring and its effects on skin inflammation and wound restoring were evaluated by using the BJ cell line as a model of human dermal fibroblasts. Moreover, a series of experiments were performed in HaCaT cells, as a model of skin keratinocytes, in order to have a complete picture of the effects of Nitrodi’s water in wound repair. In the research methods of this study, a PBS solution is used as a negative control. PBS reproduces the pH, the osmolarity, and the ion concentration of human body. Unlike water, PBS avoids cells rupturing due to osmosis, even though it was formulated without the addition of vitamins and amino acids. The absence of ions in PBS, such as calcium and magnesium, allowed us to investigate the role of Nitrodi’s water mineral content at cellular level. However, the beneficial effects of Nitrodi’s water on the skin are not only attributable to the mineral content, but it is conceivable that soluble factors or active compounds contained in these waters, in association with minerals, are effectively responsible for the multiple therapeutic effects. Therefore, the effects on fibroblasts and keratinocytes only partially reproduce the therapeutic potential of Nitrodi’s water. Skin wound healing is a complex process that can be divided into at least three continuous and overlapping processes: an inflammatory reaction, a proliferative process leading to tissue restoration, and tissue remodeling [9]. In the current study, we demonstrate that the water from Nitrodi’s spring supports BJ cell survival, and this activity is mediated by ERK signaling pathway. To investigate whether Nitrodi’s water exerts any effect on fibroblast migration, two different types of tests were performed, chemotaxis and wound healing. A chemotaxis assay was exploited to analyze the unidirectional movement of fibroblasts treated with Nitrodi’s water; whereas a wound healing assay was performed to study the ability of Nitrodi’s water to induce multi-oriented cell migration. Nitrodi’s water promotes both uni- and multi-directional migration in BJ cells. The ERK signaling pathway is required for chemotaxis, and both ERK signaling and Rac1 activation are required for fibroblast wound healing. Likely, ERK signaling moves fibroblasts between points, whereas Rac1 allows cells to explore their local environment and thus migrate into the wound. ERK signaling seems to coordinate the activity of Rac1 in fibroblast non-directional migration, as demonstrated by the inability of Nitrodi’s water to close the wound in the presence of both specific inhibitors of ERK and Rac1. Most importantly, Nitrodi’s water induced the expression of the alpha-SMA protein, thus stimulating the transition to myofibroblasts that are necessary for wound contraction. Therefore, when the fibroblasts migrate into the wound site from the surrounding tissue, they become activated, transform into myofibroblasts, and begin synthesizing extracellular matrix (ECM) proteins, mainly collagen. ECM deposition is essential for proper wound healing; in BJ cells, Nitrodi’s water is able to induce the synthesis of fibronectin, an adhesive molecule that plays a crucial role in wound healing, particularly in ECM formation and tissue regeneration. At baseline conditions, in vitro treatment of BJ cells with Nitrodi’s water reduced ROS production, suggesting that it keeps ROS levels low in the absence of stimuli, as compared to PBS. Moreover, the Nitrodi’s water treatment exhibited a protective role against oxidative stress caused by H2O2 in BJ cells, whereas in fMLF-stimulated cells it was able to preserve NADPH enzymatic complex activation in response to inflammatory stimuli. The tissue-repair process not only involves the generation of connective tissue via fibroblasts and the formation of new vessels via endothelial cells but also re-epithelialization via keratinocytes. Epithelialization is used as a defining parameter of a successful wound closure [60]. Our data show that Nitrodi’s water exerts paradoxical effects on keratinocyte migration. In chemotaxis towards a chemical signal, Nitrodi’s water inhibits cell migration; whereas, in the wound healing assay, the cells treated with Nitrodi’s water have the ability to respond effectively to mechanical stimuli generated by scraping off an area covered by cells. There may be several explanations for this phenomenon. It is conceivable that Nitrodi’s water strengthens the function of the physical, but not chemical, barrier of skin keratinocytes at the interface between the body and the environment; or it could happen that Nitrodi’s water contributes to wound closure by inducing keratinocyte proliferation rather than cell migration. In fact, further experiments show that Nitrodi’s water exerts significant stimulatory effects on cell survival in epidermal keratinocytes. Moreover, Nitrodi’s water inhibited the basal ROS production of epidermal keratinocytes, as already observed in dermal fibroblasts, but enhanced their response to the oxidative stress caused by external stimuli. Finally, the protective effects of Nitrodi’s water have not been previously described and may explain some of the positive effects of Nitrodi’s water in wound treatment. Further studies need to be conducted in order to identify which components of Nitrodi’s water aid in the wound healing process and ROS neutralization. Although this study demonstrated potential mechanisms of Nitrodi’s water in promoting wound healing, in vitro wound healing assays cannot mimic the complexity of the conditions that take place in vivo. Therefore, data obtained from in vitro assays should not be considered definitive and should be corroborated through in vivo models in order to exclude that Nitrodi’s water could induce fibroproliferative disorders, such as keloids and/or hypertrophic scarring. An investigation of which mineral component(s) in Nitrodi’s water are required for a therapeutic effect should be investigated in the future. Furthermore, the overall non-pathogenic bacteria populations of Nitrodi’s water, termed microbiota, may be responsible for its regenerative properties. These properties may be related to the production of so-far-unknown substances that promote regeneration, probably in synergy with macro- and micro-mineral elements of the spring water [61]. Importantly, the applied concentrations to experimental solutions do not match the concentration found in waters taken from the natural spring. Besides, the filtration of Nitrodi’s water alters the real composition of these waters, but sterile conditions have been required to perform this preliminary in vitro study. When searching for evidence in thermal medicine, the best strategy is to conduct human clinical trials, but in vitro evaluations of thermal waters are insufficient. In conclusion, the elucidation of the biological mechanisms underlying the benefits of Nitrodi’s water will contribute to the development of potential therapies for skin diseases. Nitrodi’s water could be a promising anti-inflammatory agent for the skin, as well as a potential wound-healing therapeutic agent. In addition, the antioxidant properties of Nitrodi’s water could be exploited to prevent symptoms related to photo-induced aging of the skin. ## 4.1. Peptides and Chemicals N-Formyl-methionyl-leucyl-phenylalanine (fMLF), PD98059 (1,4-diamino-2,3-dicyano-1,4-bis[phenylthio] butadiene), and NSC23766 were purchased from Calbiochem (La Jolla, CA, USA). Protein concentration was determined with a modified Bradford assay (Bio-Rad Laboratories). ECL Plus was obtained from GE Healthcare (Buckinghamshire, UK), and 29, 79-dichlorodihydrofluorescein diacetate (DCFH-DA) was obtained from Molecular Probes (Invitrogen, Paisley, UK). The protease and phosphatase inhibitors cocktails were obtained from Calbiochem (La Jolla, CA, USA). Mouse anti-phospho-ERK (catalog number sc-7383), rabbit anti-ERK 2 (catalog number sc-154), mouse anti-collagen type I (catalog number sc-166865), and mouse anti-fibronectin (catalog number sc-271098) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA); mouse anti-alpha-SMA (catalog number A 2547) was from Sigma-Aldrich (St. Louis, MO, USA); mouse anti-vitronectin (catalog number MAB1945) was from Chemicon International (Temecula, CA, USA); rabbit anti-actin (catalog number A2066) was from Sigma-Aldrich (St. Louis, MO, USA). Secondary anti-mouse and anti-rabbit Abs coupled to HRP were from Bio-Rad (Munchen, Germany). For the chemotaxis assay, 8-μm-pore polycarbonate membranes (Nucleopore, Pleasanton, CA, USA) coated with 10 μg/mL of fibronectin (Roche, Mannheim, Germany) were used. CellTiter 96 Aqueous One Solution Reagent was from Calbiochem (San Diego, CA, USA). Vitronectin was purchased from BD Biosciences (Bedford, MA, USA); collagen type I was purchased from Chemicon International (Temecula, CA, USA). Trypan blue solution was from Sigma-Aldrich (St. Louis, MO, USA). ## 4.2. Analysis of Nitrodi’s Water Composition The physicochemical characteristics were carried out using standard analytical methods (analyses were performed in accordance with Standard Methods for Examination of Water and Wastewater prepared and published jointly by APHA (American Public Health Association), AWWA (American Water Works Association), and AEF (American Environment Federation). The pH was evaluated using a Mettler Toledo-SevenExcellence pH/Cond meter S470-Std-K. Bicarbonate (HCO3−) and iodide (I−) were determined by the titrimetry method. Ammonium (NH4+), total phosphorus (p), and silica (SiO2) were evaluated by UV-Vis spectroscopy. For a determination of chlorides (Cl−) and sulfates (SO42−), a DIONEX Integrion HPICTM ICS1100 ion chromatography system (Thermofisher) was used [62]. The Dionex 1100 was equipped with a Dionex EGC 500 KOH RFICTM, a potassium hydroxide (KOH) eluent generator cartridge, and an IonPac AS27 RFICTM (4 × 250 mm) (Thermo Fisher Scientific, Waltham, MA, USA) analytical column. Deionized water (>18 MΩ) was used for generating the eluent. For the analysees of sodium (Na+), potassium (K+), calcium (Ca2+), magnesium (Mg2+), dissolved iron (Fe2+, Fe3+), stronzium (Sr2+), litium (Li+), and alluminium (Al3+), water samples were filtered using GF/F glass fiber filters (47 mm × 0.7 µm; Whatman, Maidstone, UK) and acidified with $1\%$ HNO3/HCl [63,64]. A Thermo ScientificTM ICAPTM RQ inductively-coupled plasma mass spectrometer (Q-ICP-MS), operating by QtegraTM Software (Version 2.7.2425.65), was used. The operating conditions of the Q-ICP-MS equipment were optimized using a tuning solution (Ba, Bi, Ce, Co, In, Li, U 1.00 µg/L, Thermo Scientific Waltham, MA, USA). The analyses were performed in KED (Kinetic Energy Discrimination) mode using Helium as collision gas. The solutions were prepared using deionized water, HNO3 ($65\%$ m/m), and HCl ($37\%$ m/m). The concentrations of the analytes were estimated using calibration lines (CertiPUR®, Merck, Darmstadt, Germany) (r2 > 0.98). For the evaluation of Polycyclic aromatic hydrocarbons (PAHs), the extraction was performed with solid-phase extraction (SPE) by Oasis HLB cartridge (6 mL, 500 mg; Waters, Milford, USA), according to the method proposed by Zhou et al. [ 65]. Before extraction, benzo[a]pyrene-d12, and indeno [1,2,3-cd]pyrene-d12 were added as surrogate solutions. Methylene chloride (5 mL), methanol (5 mL), and ultra-pure water (5 mL) were used to condition and wash the cartridge, and then the water sample was eluted with a flow rate of 10 mL min−1. Therefore, the cartridge was eluted with methylene chloride (10 mL), and the extract was concentrated to 500 µL in hexane for GC-MS analysis. Finally, chrysene-d12 was added to the sample as an internal standard. In order to perform the Organochlorine and Organophosphorus Pesticides (OCLs and OPPs) and the Polychlorinated biphenyls (PCBs) extraction, water samples were adjusted to pH~7 and preconcentrated using SPE Oasis HLB cartridges (6 mL, 500 mg; Waters, Milford, USA), previously preconditioned with 5 mL of ethyl acetate (EtOAc), 5 mL of methanol (MetOH) and 2.5 mL of deionized water [66]. The extracts were eluted with 6 mL of EtOAc, evaporated to dryness, and reconstituted in hexane for analysis by GC-MS. Semi-volatiles organic compounds (SVOCs) surrogate standard (2-fluorobiphenyl, nitrobenzene-d5, p-terphenyl-d14, 2-fluorophenol, phenol-d5, 2,4,6-tribromophenol) and SVOCs internal standard (acenaphtene-d10, crysene-d12, 1,4-dichlorobenzene-d4, naphtalene-d8, perylene-d12, phenantrene-d10) mixtures were used as surrogate and internal standards, respectively. For determining trihalomethanes (THMs) in water samples, purge-and-trap (PT) extraction followed by gas chromatography–mass spectrometric (GC–MS) analysis were used. For analysis of PAHs, OCLs, OPPs, PCBs, and THMs, a TRACETM 1310 Gas Chromatography coupled to ISQTM 7000 Single Quadrupole Mass Spectrometer (GC-MS, Thermo Scientific, USA) under selected-ion monitoring (SIM) mode was used to perform the analysis. A TG-5MS capillary column (30 mm length × 0.25 mm inner diameter × 0.25 μm film thickness) and helium as carrier gas (constant flow rate of 1 mL/min) were used. The MSD worked in the electron ionization (EI) mode, set at 70 eV. Splitless injection mode was adopted, and sample injection volume was 1 μL. To check the methods, all samples were analyzed in duplicate. Quality assurance and quality control were assessed using duplicates, method blanks, and standard reference materials. The accuracy of the instrumental methods was checked by triplication of the samples, as well as by using spiked samples, which was run after every single sample. The limit of detections (LODs) and limit of quantifications (LOQs) were estimated as 3 and 10 times, respectively, the signal/noise ratio for each analyte by five replicate analyses. The recoveries of detected analytes in spiked samples were in the range of 70–$130\%$, which met the quality control requirements. ## 4.3. Cell Cultures BJ cells (human healthy fibroblasts established from skin taken from normal foreskin of a neonatal male, ATCC CRL-2522) and HaCaT cells (normal human keratinocytes, ATCC CRL-2404) were maintained in Dulbecco’s modified essential medium (DMEM, Gibco, Thermo Fisher Scientific, Inc., Waltham, MA, USA) with L-glutamine, supplemented with $10\%$ FCS (fetal calf serum, Gibco), and $1\%$ antibiotic (Penicillin and Streptomicin, Gibco). All cultured cells were kept at 37 °C in a humidified atmosphere of $95\%$ air and $5\%$ of carbon dioxide (CO2). When the cells reached confluence, the culture medium was removed from the flask and cells were rinsed two times with sterile PBS (Phosphate-Buffered Saline, Gibco). The confluent layer was trypsinized using Trypsin/EDTA (Gibco) and then resuspended in fresh medium. ## 4.4. Preparation of Nitrodi’ Water Solution To analyze the biological effects of Nitrodi’s water in the cultured cells, the following salts that make up PBS were dissolved in 1.0 L of Nitrodi’s water: 80 g of NaCl; 2.0 g of KCl; 14.4 g of Na2HPO4; and 2.4 g of KH2PO4. After mixing to dissolve, the pH was adjusted to 7.2. This solution, indicated in the text and in the figures as Nitrodi, was diluted 1:10 and stored at room temperature. Cells treated with PBS solution (prepared in distilled water) were used as negative control. Both Nitrodi’s water and PBS solution were sterilized by filtration using 0.22 μm membrane filters. ## 4.5. Western Blot Analysis Immunoblotting experiments were performed according to standard procedures [67]. Briefly, cells were harvested in lysis RIPA buffer (RIPA Buffer: 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2 EDTA, 1 mM EGTA, $1\%$ NP-40, $1\%$ sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, and 1 µg/mL leupeptin) supplemented with a mixture of proteases and phosphatases inhibitors. Thirty micrograms of protein were electrophoresed on a $10\%$ SDS-PAGE and transferred onto a polyvinylidene fluoride membrane. The membrane was blocked with $5\%$ nonfat dry milk and probed with specific Abs: mouse anti-p-ERK $\frac{1}{2}$ (1 μg/mL), rabbit anti-ERK (1 μg/mL), mouse anti-alpha-SMA (2 μg/mL), and rabbit anti-β-actin (0.5 μg/mL). Finally, washed filters were incubated with HRP-conjugated anti-rabbit or anti-mouse antibodies. The immunoreactive bands were detected by a chemiluminescence kit and quantified by densitometry (ChemiDoc XRS, Bio-Rad). Quantifications of Western blot were performed using ImageJ software version 1.53 m (National Institute of Health, Bethesda, MD, USA). ## 4.6. Proliferation Assay Cultured cells and HaCaT cells were serum-starved overnight using DMEM $0.1\%$ BSA, plated at 5 × 103 cells/well in 96-well plates (Corning), and then treated with Nitrodi supplemented with $0.5\%$ BSA (starving condition) and $10\%$ FCS (growing condition). Cells treated with PBS supplemented with $0.5\%$ BSA and $10\%$ FCS were used as a negative control. Cells treated with culture medium supplemented with $0.5\%$ BSA and $10\%$ FCS were used as a positive control. In experiments on the role of ERKs in BJ cell proliferation, cultured cells were pre-treated with the ERK $\frac{1}{2}$ inhibitor PD98059 at 50 μM for 1 h at 37 °C and then exposed to PBS, Nitrodi, or culture medium. The cell proliferation was measured at different time points, as showed in the figures. At the end of the incubation, 20 μL/well CellTiter-96 was added. After incubation at 37 °C for 2 h, the absorbance was determined by an ELISA reader (Bio-Rad) at a wavelength of 490 nm according to the manufacturer’s instructions. ## 4.7. Chemotaxis Assay Chemotaxis assays were performed using a modified Boyden chamber technique [68]. Briefly, 25 μL of medium supplemented with $5\%$ FCS were placed in triplicate in the lower compartment of a microchemotaxis chamber (NeuroProbe, Cabin John, MD, USA). The lower compartments were covered with 8-μm-pore polycarbonate membranes coated with fibronectin (10 μg/mL). Fifty microliters of the cell suspension (5 × 104/well), resuspended in PBS (as a negative control) and Nitrodi, were loaded into the upper compartments. The chemotactic chamber was then incubated for 24 h at 37 °C in a humidified incubator with $5\%$ CO2. Then, the membrane was removed, the upper side was washed with PBS, and cells attached to the lower surface of the filter were fixed, stained with May-Grünwald-Giemsa, mounted on a microscope slide with Cytoseal (Stephens Scientific, Springfield, NJ, USA), and counted. In each experiment, 10 fields/triplicate filters were measured at ×40 magnification. ## 4.8. Wound Healing Assay BJ cells and HaCaT cells were seeded into a 12-well culture plate using DMEM containing $10\%$ FCS and incubated for 12 h at 37 °C to create a confluent monolayer. Cells were then scraped with a p200 pipette tip in a straight line to create a “scratch.” The debris was removed, and the edge of the scratch was smoothed by washing cells once with 2 mL growth medium. Cells were then incubated at 37 °C with PBS (as a negative control) or Nitrodi for different time points, at baseline ($T = 0$), 24, and 72 h. The data acquisition was conducted through microscopic-image capturing and gap measurement at each time point. Measurements of wound length were carried out manually using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Three measurements of length for each digital picture were performed in order to make an accurate measurement of the image. ## 4.9. In Situ Trypan Blue Staining To assess the viability of HaCaT cells subjected to wound healing assay (as described above), the cells were incubated with $0.4\%$ trypan blue solution. The cells were washed with PBS for three times, and then incubated with trypan blue for 10 min at room temperature. After trypan blue staining, cells were fixed with $4\%$ paraformaldehyde [69,70]. Stained and fixed cells were observed by 10× objective of inverted-phase contrast microscope (Olympus CKX41). ## 4.10. In Situ ELISA A quantitative assessment of extracellular matrix (ECM) proteins (vitronectin, fibronectin, and collagen type I) was performed by in situ ELISA. BJ cells were plated in 96-well plates (Corning) at a density of 5 × 104 cells per well and treated with PBS (as a negative control) and Nitrodi. In parallel, the plates were coated with purified ECM proteins (vitronectin, fibronectin, and collagen type I) at different concentrations (16, 8, 4, 2, 1, and 0 μg/mL), in order to provide a quantification of absorbed ECM proteins on the tissue culture plastic. The concentration of the test samples was determined by using their absorbance values and interpolating this from the calibration curve. After 24 h of incubation, proteins were fixed using acetone/methanol (v/v) for 10 min at 22 °C, incubated in $0.5\%$ PBS/BSA and $0.2\%$ Tween 20 for 30 min at 22 °C to minimize aspecific binding sites, and washed in PBS. Goat polyclonal anti-vitronectin (2 μg/mL) antibody, mouse monoclonal anti-fibronectin (2 μg/mL), and anti-collagen type 1 (2 μg/mL) antibodies were added for 1 h at 22 °C. After three washes in PBS, plates were incubated with HRP secondary antibodies for 30 min at 22 °C. After three washes in PBS, the substrate was added (1 mg/mL OFD, 0.1 mol/l citrate buffer [pH 5], and $0.006\%$ H2O2), and plates were incubated for 30 min at 37 °C in the dark. The reaction was then stopped by 1 M H2SO4, and the absorbance was read at 450 nm by a spectrophotometer. ## 4.11. Reactive Oxygen Species Detection Cells were plated overnight at 2 × 104 cells/well in 96-well plates using DMEM with $10\%$ FCS. Cells were incubated with DCFH-DA at a concentration of 5 µM for 30 min in the dark at 37 °C. The esterified form of DCFH-DA can permeate cell membranes before being deacetylated by intracellular esterases. The resulting compound, dichlorodihydrofluorescein, reacts with reactive oxygen species (ROS), producing an oxidized fluorescent compound, dichlorofluorescein (DCF), which can be detected using a multiplate reader. After incubation with DCFH-DA, cells were washed twice and treated with H2O2 (1 mM) or fMLF (10−4 M) in presence of PBS (as a negative control) and Nitrodi for 5, 15, 30, and 60 min at 37 °C in a humidified $5\%$ CO2 incubator. DCF was detected at a wavelength of 535 nm by a microplate reader (Tecan Trading AG, Switzerland). DCFH-DA-unloaded cells were examined in parallel and subtracted to DCFH-DA-loaded cell values. ## 4.12. Statistical Analysis All the experiments were performed at least in triplicate. The results were expressed as mean ± SEM. Values from groups were compared using a paired Student t-test [71]. Differences were considered significant when * $p \leq 0.05$ and $p \leq 0.001.$ ## References 1. Dąbrowska A., Spano F., Derler S., Adlhart C., Spencer N., Rossi R.. The relationship between skin function, barrier properties, and body-dependent factors*. Skin. Res. Technol.* (2018) 24 165-174. DOI: 10.1111/srt.12424 2. Proksch E., Brandner J.M., Jensen J.-M.. The skin: An indispensable barrier. Exp. Dermatol. (2008) 17 1063-1072. DOI: 10.1111/j.1600-0625.2008.00786.x 3. Lembo C., Raimondo A., de Paulis A., Mormile I., Rossi F.W., Lembo S., Balato A.. Clinical predictors of psoriatic arthritis and osteoclast differentiation. Exp. Dermatol. (2021) 30 1834-1837. DOI: 10.1111/exd.14424 4. Rognoni E., Watt F.M.. Skin Cell Heterogeneity in Development, Wound Healing, and Cancer. Trends Cell Biol. (2018) 28 709-722. DOI: 10.1016/j.tcb.2018.05.002 5. Watt F.M.. Mammalian skin cell biology: At the interface between laboratory and clinic. Science (2014) 346 937-940. DOI: 10.1126/science.1253734 6. Sotiropoulou P.A., Blanpain C.. Development and Homeostasis of the Skin Epidermis. Cold Spring Harb. Perspect. Biol. (2012) 4 a008383. DOI: 10.1101/cshperspect.a008383 7. Driskell R.R., Lichtenberger B.M., Hoste E., Kretzschmar K., Simons B.D., Charalambous M., Ferron S.R., Herault Y., Pavlovic G., Ferguson-Smith A.C.. Distinct fibroblast lineages determine dermal architecture in skin development and repair. Nature (2013) 504 277-281. DOI: 10.1038/nature12783 8. Mormile I., Russo R., Andolfo I., de Paulis A., Rossi F.W., Rendina D.. Rheumatoid arthritis and osteogenesis imperfecta: Is there a genetic causal association?. Osteoporos. Int. (2022) 33 2233-2235. DOI: 10.1007/s00198-022-06486-9 9. Lindley L.E., Stojadinovic O., Pastar I., Tomic-Canic M.. Biology and Biomarkers for Wound Healing. Plast. Reconstr. Surg. (2016) 138 18S-28S. DOI: 10.1097/PRS.0000000000002682 10. Gurtner G.C., Werner S., Barrandon Y., Longaker M.T.. Wound repair and regeneration. Nature (2008) 453 314-321. DOI: 10.1038/nature07039 11. Raja K., Sivamani M.S., Garcia R.R.. Wound re-epithelialization: Modulating keratinocyte migration in wound healing. Front. Biosci. (2007) 12 2849-2868. DOI: 10.2741/2277 12. Dekoninck S., Blanpain C.. Stem cell dynamics, migration and plasticity during wound healing. Nature (2019) 21 18-24. DOI: 10.1038/s41556-018-0237-6 13. Diegelmann R.F., Evans M.C.. Wound healing: An overview of acute, fibrotic and delayed healing. Front. Biosci. (2004) 9 283-289. DOI: 10.2741/1184 14. Rinnerthaler M., Bischof J., Streubel M.K., Trost A., Richter K.. Oxidative Stress in Aging Human Skin. Biomolecules (2015) 5 545-589. DOI: 10.3390/biom5020545 15. Kammeyer A., Luiten R.M.. Oxidation events and skin aging. Ageing Res. Rev. (2015) 21 16-29. DOI: 10.1016/j.arr.2015.01.001 16. Di Spigna G., Rossi F.W., Mormile I., Ladogana P., Buonavolontà L., Covelli B., Salzano S., Napolitano F., Giannini A., Postiglione L.. Serum Metalloprotease 3 (MMP-3) biomarker of therapeutic efficacy during treatment of rheumatoid arthritis. J. Biol. Regul. Homeost. Agents (2021) 35 1041-1045. DOI: 10.23812/21-86-L 17. Tan B.L., Norhaizan M.E., Liew W.-P.-P., Rahman H.S.. Antioxidant and Oxidative Stress: A Mutual Interplay in Age-Related Diseases. Front. Pharmacol. (2018) 9 1162. DOI: 10.3389/fphar.2018.01162 18. Dosoki H., Stegemann A., Taha M., Schnittler H., Luger T.A., Schröder K., Distler J.H.W., Kerkhoff C., Böhm M.. Targeting of NADPH oxidase in vitro and in vivo suppresses fibroblast activation and experimental skin fibrosis. Exp. Dermatol. (2016) 26 73-81. DOI: 10.1111/exd.13180 19. Davis S.C., Perez R.. Cosmeceuticals and natural products: Wound healing. Clin. Dermatol. (2009) 27 502-506. DOI: 10.1016/j.clindermatol.2009.05.015 20. Liang J., Kang D., Wang Y., Yu Y., Fan J., Takashi E.. Carbonate Ion-Enriched Hot Spring Water Promotes Skin Wound Healing in Nude Rats. PLoS ONE (2015) 10 e0117106. DOI: 10.1371/journal.pone.0117106 21. Rossi F.W., Rivellese F., Napolitano F., Mosella F., Selleri C., Montuori N., De Paulis A.. Effects of Polyurethane Foam Dressings as an Add-on Therapy in the Management of Digital Ulcers in Scleroderma Patients. Transl. Med. UniSa (2020) 22 10-14. PMID: 32523901 22. Shedoeva A., Leavesley D., Upton Z., Fan C.. Wound Healing and the Use of Medicinal Plants. Evid.-Based Complement. Altern. Med. (2019) 2019 1-30. DOI: 10.1155/2019/2684108 23. Forti L.. Rilievi Dedicati Alle Ninfe di Nitrodi. L’arte Tipograf: Napoli (1951) Volume XXVI 24. Mancioli M.. Le Proprietà Terapeutiche Delle Acque di Nitrodi e Olmitello (1984) 25. Sica M.A.. Classificazione Italiana Delle Acque Minerali (1929) Volume 19 26. Sica M.A.. Classificazione Italiana Delle Acque Minerali (1933) Volume 33 27. Aversano A., Rossi F.W., Cammarota F., De Paulis A., Izzo P., De Rosa M.. Nitrodi thermal water downregulates protein S-nitrosylation in RKO cells. Int. J. Mol. Med. (2020) 46 1359-1366. DOI: 10.3892/ijmm.2020.4676 28. Inguaggiato S., Pecoraino G., D’Amore F.. Chemical and isotopical characterisation of fluid manifestations of Ischia Island (Italy). J. Volcanol. Geotherm. Res. (2000) 99 151-178. DOI: 10.1016/S0377-0273(00)00158-X 29. Nasermoaddeli A., Kagamimori S.. Balneotherapy in medicine: A review. Environ. Health Prev. Med. (2005) 10 171-179. DOI: 10.1007/BF02897707 30. de Araújo R., Lôbo M., Trindade K., Silva D.F., Pereira N.. Fibroblast Growth Factors: A Controlling Mechanism of Skin Aging. Ski. Pharmacol. Physiol. (2019) 32 275-282. DOI: 10.1159/000501145 31. Napolitano F., Rossi F.W., Pesapane A., Varricchio S., Ilardi G., Mascolo M., Staibano S., Lavecchia A., Ragno P., Selleri C.. N-Formyl Peptide Receptors Induce Radical Oxygen Production in Fibroblasts Derived from Systemic Sclerosis by Interacting With a Cleaved Form of Urokinase Receptor. Front. Immunol. (2018) 9 574. DOI: 10.3389/fimmu.2018.00574 32. Rossi F.W., Prevete N., Rivellese F., Napolitano F., Montuori N., Postiglione L., Selleri C., De Paulis A.. The Urokinase/Urokinase Receptor System in Mast Cells: Effects of its Functional Interaction with fMLF Receptors. Transl. Med. UniSa (2016) 15 34-41. PMID: 27896225 33. Mormile I., Rossi F.W., Prevete N., Granata F., Pucino V., de Paulis A.. The N-Formyl Peptide Receptors and Rheumatoid Arthritis: A Dangerous Liaison or Confusing Relationship?. Front. Immunol. (2021) 12 685214. DOI: 10.3389/fimmu.2021.685214 34. Barrientos S., Stojadinovic O., Golinko M.S., Brem H., Tomic-Canic M.. Growth factors and cytokines in wound healing. Wound Repair Regen. (2008) 16 585-601. DOI: 10.1111/j.1524-475X.2008.00410.x 35. Bainbridge P.. Wound healing and the role of fibroblasts. J. Wound Care (2013) 22 407-411. PMID: 23924840 36. Fenteany G., Janmey P.A., Stossel T.P.. Signaling pathways and cell mechanics involved in wound closure by epithelial cell sheets. Curr. Biol. (2000) 10 831-838. DOI: 10.1016/S0960-9822(00)00579-0 37. Samson S.C., Khan A.M., Mendoza M.C.. ERK signaling for cell migration and invasion. Front. Mol. Biosci. (2022) 9 998475. DOI: 10.3389/fmolb.2022.998475 38. Li B., Wang J.H.-C.. Fibroblasts and myofibroblasts in wound healing: Force generation and measurement. J. Tissue Viability (2011) 20 108-120. DOI: 10.1016/j.jtv.2009.11.004 39. Hinz B., Celetta G., Tomasek J.J., Gabbiani G., Chaponnier C.. Alpha-Smooth Muscle Actin Expression Upregulates Fibroblast Contractile Activity. Mol. Biol. Cell (2001) 12 2730-2741. DOI: 10.1091/mbc.12.9.2730 40. Rossi F.W., Napolitano F., Pesapane A., Mascolo M., Staibano S., Matucci-Cerinic M., Guiducci S., Ragno P., di Spigna G., Postiglione L.. Upregulation of the. J. Immunol. (2015) 194 5161-5173. DOI: 10.4049/jimmunol.1402819 41. Hoffmann G.A., Smith M.L.. New Insights into Collagen and Fibronectin Reciprocity during Extracellular Matrix Formation. Chem (2019) 5 1930-1932. DOI: 10.1016/j.chempr.2019.07.012 42. Bae H.-B., Zmijewski J.W., Deshane J.S., Zhi D., Thompson L.C., Peterson C.B., Chaplin D.D., Abraham E.. Vitronectin Inhibits Neutrophil Apoptosis through Activation of Integrin-Associated Signaling Pathways. Am. J. Respir. Cell Mol. Biol. (2012) 46 790-796. DOI: 10.1165/rcmb.2011-0187OC 43. Davalli P., Mitic T., Caporali A., Lauriola A., D’Arca D.. ROS, Cell Senescence, and Novel Molecular Mechanisms in Aging and Age-Related Diseases. Oxidative Med. Cell. Longev. (2016) 2016 1-18. DOI: 10.1155/2016/3565127 44. Ganceviciene R., Liakou A.I., Theodoridis A., Makrantonaki E., Zouboulis C.C.. Skin anti-aging strategies. Derm.-Endocrinol. (2012) 4 308-319. DOI: 10.4161/derm.22804 45. Poljšak B., Dahmane R.G., Godić A.. Intrinsic skin aging: The role of oxidative stress. Acta Dermatovenerol. Alp. Panon. Adriat. (2012) 21 33-36 46. Xia S., Zhang X., Zheng S., Khanabdali R., Kalionis B., Wu J., Wan W., Tai X.. An Update on Inflamm-Aging: Mechanisms, Prevention, and Treatment. J. Immunol. Res. (2016) 2016 1-12. DOI: 10.1155/2016/8426874 47. Singh M., Sharma H., Singh N.. Hydrogen peroxide induces apoptosis in HeLa cells through mitochondrial pathway. Mitochondrion (2007) 7 367-373. DOI: 10.1016/j.mito.2007.07.003 48. Sauer H., Wartenberg M., Hescheler J.. Reactive Oxygen Species as Intracellular Messengers During Cell Growth and Differentiation. Cell. Physiol. Biochem. (2001) 11 173-186. DOI: 10.1159/000047804 49. Piipponen M., Li D., Landén N.X.. The Immune Functions of Keratinocytes in Skin Wound Healing. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21228790 50. Ascione F., Vasaturo A., Caserta S., D’Esposito V., Formisano P., Guido S.. Comparison between fibroblast wound healing and cell random migration assays in vitro. Exp. Cell Res. (2016) 347 123-132. DOI: 10.1016/j.yexcr.2016.07.015 51. McGee H.M., Schmidt B.A., Booth C.J., Yancopoulos G.D., Valenzuela D.M., Murphy A.J., Stevens S., Flavell R.A., Horsley V.. IL-22 Promotes Fibroblast-Mediated Wound Repair in the Skin. J. Investig. Dermatol. (2013) 133 1321-1329. DOI: 10.1038/jid.2012.463 52. Wang Y., Graves D.T.. Keratinocyte Function in Normal and Diabetic Wounds and Modulation by FOXO1. J. Diabetes Res. (2020) 2020 1-9. DOI: 10.1155/2020/3714704 53. Passeron T., Zouboulis C., Tan J., Andersen M., Katta R., Lyu X., Aguilar L., Kerob D., Morita A., Krutmann J.. Adult skin acute stress responses to short-term environmental and internal aggression from exposome factors. J. Eur. Acad. Dermatol. Venereol. (2021) 35 1963-1975. DOI: 10.1111/jdv.17432 54. Bito T., Nishigori C.. Impact of reactive oxygen species on keratinocyte signaling pathways. J. Dermatol. Sci. (2012) 68 3-8. DOI: 10.1016/j.jdermsci.2012.06.006 55. Elsholz F., Harteneck C., Muller W., Friedland K.. Calcium—A central regulator of keratinocyte differentiation in health and disease. Eur. J. Dermatol. (2014) 24 650-661. DOI: 10.1684/ejd.2014.2452 56. Yücesoy H., Dönmez A., Atmaca-Aydın E., Yentür S.P., Saruhan-Direskeneli G., Ankaralı H., Erdoğan N., Karagülle M.Z.. Effects of balneological outpatient treatment on clinical parameters and serum cytokine levels in patients with chronic low back pain: A single-blind randomized controlled trial. Int. J. Biometeorol. (2021) 65 1367-1376. DOI: 10.1007/s00484-021-02109-w 57. Nicoletti G., Saler M., Pellegatta T., Tresoldi M.M., Bonfanti V., Malovini A., Faga A., Riva F.. Ex vivo regenerative effects of a spring water. Biomed. Rep. (2017) 7 508-514. PMID: 29188053 58. Carbajo J.M., Maraver F.. Sulphurous Mineral Waters: New Applications for Health. Evidence-Based Complement. Altern. Med. (2017) 2017 1-11. DOI: 10.1155/2017/8034084 59. Szabó I., Varga C.. Finding possible pharmacological effects of identified organic compounds in medicinal waters (BTEX and phenolic compounds). Int. J. Biometeorol. (2019) 64 989-995. DOI: 10.1007/s00484-019-01808-9 60. Pastar I., Stojadinovic O., Yin N.C., Ramirez H., Nusbaum A.G., Sawaya A., Patel S.B., Khalid L., Isseroff R.R., Tomic-Canic M.. Epithelialization in Wound Healing: A Comprehensive Review. Adv. Wound Care (2014) 3 445-464. DOI: 10.1089/wound.2013.0473 61. Pellegatta T., Saler M., Bonfanti V., Nicoletti G., Faga A.. Novel perspectives on the role of the human microbiota in regenerative medicine and surgery. Biomed. Rep. (2016) 5 519-524. DOI: 10.3892/br.2016.778 62. Chapp A.D., Schum S., Behnke J.E., Hahka T., Huber M.J., Jiang E., Larson R.A., Shan Z., Chen Q.-H.. Measurement of cations, anions, and acetate in serum, urine, cerebrospinal fluid, and tissue by ion chromatography. Physiol. Rep. (2018) 6 e13666. DOI: 10.14814/phy2.13666 63. Chen M., Li F., Tao M., Hu L., Shi Y., Liu Y.. Distribution and ecological risks of heavy metals in river sediments and overlying water in typical mining areas of China. Mar. Pollut. Bull. (2019) 146 893-899. DOI: 10.1016/j.marpolbul.2019.07.029 64. Jia Y., Wang L., Qu Z., Yang Z.. Distribution, contamination and accumulation of heavy metals in water, sediments, and freshwater shellfish from Liuyang River, Southern China. Environ. Sci. Pollut. Res. (2017) 25 7012-7020. DOI: 10.1007/s11356-017-1068-x 65. Zhou J.L., Hong H., Zhang Z., Maskaoui K., Chen W.. Multi-phase distribution of organic micropollutants in Xiamen Harbour, China. Water Res. (2000) 34 2132-2150. DOI: 10.1016/S0043-1354(99)00360-7 66. Cruzeiro C., Pardal M., Rodrigues-Oliveira N., Castro L.F.C., Rocha E., Rocha M.J.. Multi-matrix quantification and risk assessment of pesticides in the longest river of the Iberian Peninsula. Sci. Total Environ. (2016) 572 263-272. DOI: 10.1016/j.scitotenv.2016.07.203 67. Bhattacharya A., Limone A., Napolitano F., Cerchia C., Parisi S., Minopoli G., Montuori N., Lavecchia A., Sarnataro D.. APP Maturation and Intracellular Localization Are Controlled by a Specific Inhibitor of 37/67 kDa Laminin-1 Receptor in Neuronal Cells. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21051738 68. Rossi F.W., Prevete N., Rivellese F., Lobasso A., Napolitano F., Granata F., Selleri C., De Paulis A.. HIV-1 Nef promotes migration and chemokine synthesis of human basophils and mast cells through the interaction with CXCR4. Clin. Mol. Allergy (2016) 14 1-9. DOI: 10.1186/s12948-016-0052-1 69. Perry S.W., Epstein L., Gelbard H.A.. In Situ Trypan Blue Staining of Monolayer Cell Cultures for Permanent Fixation and Mounting. Biotechniques (1997) 22 1020-1024. DOI: 10.2144/97226bm01 70. Perry S.W., Epstein L., Gelbard H.A.. Simultaneous In Situ Detection of Apoptosis and Necrosis in Monolayer Cultures by TUNEL and Trypan Blue Staining. Biotechniques (1997) 22 1102-1106. DOI: 10.2144/97226st01 71. Snedecor G.W.C.. Statistical Methods (1980)
--- title: The Interactive Relationship between Street Centrality and Land Use Intensity—A Case Study of Jinan, China authors: - Chengzhen Song - Qingfang Liu - Jinping Song - Ding Yang - Zhengyun Jiang - Wei Ma - Fuchang Niu - Jinmeng Song journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049117 doi: 10.3390/ijerph20065127 license: CC BY 4.0 --- # The Interactive Relationship between Street Centrality and Land Use Intensity—A Case Study of Jinan, China ## Abstract It is of great significance to study the interactive relationship between urban transportation and land use for promoting the healthy and sustainable development of cities. Taking Jinan, China, as an example, this study explored the interactive relationship between street centrality (SC) and land use intensity (LUI) in the main urban area of Jinan by using the spatial three-stage least squares method. The results showed that the closeness centrality showed an obvious “core-edge” pattern, which gradually decreased from the central urban area to the edge area. Both the betweenness centrality and the straightness centrality showed a multi-center structure. The commercial land intensity (CLUI) showed the characteristics of multi-core spatial distribution, while the residential land intensity (RLUI) and public service land intensity (PLUI) showed the characteristics of spatial distribution with the coexistence of large and small cores. There was an interactive relationship between SC and LUI. The closeness centrality and straightness centrality had positive effects on LUI, and LUI had a positive effect on closeness centrality and straightness centrality. The betweenness centrality had a negative impact on LUI, and LUI also had a negative impact on betweenness centrality. Moreover, good location factors and good traffic conditions were conducive to improving the closeness and straightness centrality of the regional traffic network. Good location factors, good traffic conditions and high population density were conducive to improving regional LUI. ## 1. Introduction Urban transportation and land use have always been popular research fields in urban geography, urban planning and urban economics. Relevant research can be traced back to the model of urban internal regional structure proposed by the “Chicago School” in the early 20th century (Park, 1915), such as the concentric circle model proposed by Burgess in 1925 [1], the fan model proposed by Hoyt in 1939 [2] and the multi-core model proposed by Harris and Ullman in 1945 [3]. All of these models mentioned the importance of transportation in shaping urban spatial structure. Moreover, the classic Western economic model also emphasized the importance of urban transportation in shaping urban spatial structure. For example, in the 1960s, Alonso [4], an American land economist, proposed the rent competition theory, and Mills [5] [1972] and Muth [6] [1969] proposed the single-center model. In China, since the reform and opening up of the country, the process of urbanization has been rapidly advancing. From 1978 to 2019, China’s urbanization rate increased rapidly from $17.9\%$ to $60.6\%$. In the context of rapid urbanization, China’s urban space has expanded rapidly, and many large- and medium-sized cities have emerged in the process of development in a “big pie” type of spatial disorder development mode, which has triggered a series of urban problems, such as overcrowding in urban central areas, a waste of land resources around the city and a decline in the quality of the urban environment [7]. In addition, in the process of the construction of new urban areas, problems such as excessive construction space of new urban areas and excessive motorization of roads generally exist, which are manifested in the excessive sizes of street space and the low density of road networks [8,9]. The establishment of a reasonable, complete and efficient urban transportation system is conducive to alleviating urban traffic congestion, facilitating citizens’ travel, optimizing the urban internal spatial structure and improving land use efficiency, which is of great significance to improving the quality of life for urban residents and promoting sustainable urban development. As an important part of urban infrastructure, the transportation system plays a key role in urban land use planning and development. Moreover, the spatial structure of urban land use also guides the planning and construction of urban transportation infrastructure. Therefore, urban traffic and land use interact with each other [10]. However, current scholars mainly carry out one-way impact studies and linear correlation studies on urban traffic and land use, and there are few studies on the interaction between urban traffic and land use. According to the 2017 Transportation Analysis Report of China’s Major Cities, Jinan once again ranked first in the list of China’s congested cities with a peak congestion delay index of 2.067. In 2018 and 2019, Jinan ranked among the top 10 most congested cities in China. Traffic congestion seriously restricts the healthy and sustainable development of cities. Different from the urban development mode of the United States, the central area of Chinese cities is the agglomeration area of production and the life of residents, as well as the concentration area of urban transportation and land development. How to coordinate the relationship between urban traffic and land use, further optimize the urban spatial structure and promote the healthy and high-quality development of the city, which are important parts of the future urban development of Jinan. This paper took Jinan, a typical city, as an example, and selected the spatial three-stage least squares method to explore the interactive relationship between SC and LUI. Moreover, this paper explored the factors that influence SC and LUI. Compared with existing studies, this paper has the following advantages: First, this study makes up for the deficiencies in the current quantitative research on the interactive relationship between urban transportation and land use. Second, this study took the concentrated area of urban transportation and land development in China (the main urban area of Jinan) as the research object, as it is representative and valuable. The research structure of this paper is as follows: Section 2 is the literature review. Section 3 comprises the study area overview, methods and data sources. Section 4 is the research results. Section 5 is the discussion of the research results. Section 6 is the main conclusions. ## 2. Literature Review At present, the academic community has accumulated some research experience in urban transportation and land use. Relevant research mainly focuses on the following aspects: ## 2.1. Exploring the Impact of Urban Traffic on Land Use European and American scholars have carried out earlier studies on the impact of urban traffic on land use, and they have improved urban land use patterns by optimizing transportation facilities. For example, Schaeffer explored the relationship between urban traffic development and urban spatial form, and analyzed the importance of transportation in shaping urban spatial form [11]. Giannopoulos systematically analyzed the important impact of transportation technology innovation on urban morphology from the perspective of technological innovation [12]. Moon explored the influence of Washington metro stations on commercial land and residential land, and found that traffic stations could promote the outward diffusion of the population in the central business district [13]. Baerwald found that traffic accessibility was the key factor to determining urban land for residential land development by exploring the impact of traffic accessibility on residential land development [14]. Rui et al. took Stockholm as the research object and found that road network centrality could affect the spatial distribution of land use types [15]. With the rapid development of urbanization in China, many Chinese scholars began to pay attention to the impact of urban traffic on land use. For example, taking Guangzhou as the research object, Mao Jiangxing explored the impact of the urban transportation system on land use price by using spatial analysis technology and the multivariate statistical analysis method [16]. Sun Jiuwen believed that the construction of urban rail transit could change the urban spatial development pattern, which was of great significance for alleviating urban traffic congestion and achieving high-quality urban development [17]. ## 2.2. Exploring the Impact of Land Use on Urban Traffic Some scholars found that urban land use structure could guide the rational planning and construction of urban transportation facilities. For example, Giuliaono found that urban land development could promote the development of urban transportation infrastructure after reaching a certain level [18]. Johnson explored the impact of land use on the demand for public transport infrastructure and concluded that improving the mix of urban land use could improve the demand for public transport infrastructure [19]. Based on the survey data of Boston and Hong Kong, Zhang Ming concluded through model empirical analysis that urban land use status had an important impact on citizens’ choice of travel mode [20]. Ding Chengri made a comparative study on the influence of the spatial structure of single-center and multi-center cities on the urban transportation system, and found that the single-center structure was conducive to the development of urban public transportation, while the multi-center urban structure did not necessarily reduce the demand for urban transportation, and the impact of mixed urban land use on urban transportation was uncertain [21]. ## 2.3. Exploring the Relationship between Urban Traffic and Land Use Since the 1960s, some developed areas, such as Europe and the United States, have begun to carry out some studies on urban development, land use and transportation planning by creating an integrated model of transportation planning and land use, such as the Lowry model, methods based on mathematical planning, input–output methods, and rent-competition function methods based on urban economics [22,23,24]. In recent years, some Chinese scholars have also begun to pay attention to research on urban transportation and land use. For example, Wang Shuai took Shenzhen as an example and explored the correlation between road network centrality and land use intensity using the Pearson correlation coefficient [25]. Taking the main urban area of Chongqing as the research object, Tian et al. measured the coordination degree of land use and urban traffic by using the DEA method [26]. Taking Hubei Province as an example, Yin Chaohui explored the relationship between the centrality of the road network and the landscape pattern of land use by using the Spearman correlation analysis method [27]. Yin Guanwen took Jinan as an example to explore the correlation and heterogeneity between the centrality of the urban road network and land use intensity [28]. ## 2.4. Exploring the Urban Traffic Network The rapid development of cyberspace science has greatly promoted the relevant research into the urban transportation network, which has provided convenient technical support for exploring the relationship between the urban transportation network and land use [29]. At present, there are two popular models for studying the traffic network: the spatial syntax model [30,31,32] and the multiple centrality assessment model [33,34,35]. In the 1980s, the British scholar Bill Hillier first proposed the spatial syntax model [36], which cannot only effectively measure the local spatial accessibility of a region, but also measure the overall spatial relevance of a region. It is often used to explore the correlation between urban economic activities and road network centrality [37]. In 2006, Cruciti proposed a multiple centrality assessment model based on the spatial syntax theory, which reflects the importance of road network nodes by measuring the centrality of road nodes [38,39]. Owing to the multiple centrality assessment model which measures the actual distance, the calculation results are more scientific and reliable [40]. In general, current academic circles mainly focus on one-way impact research on urban traffic and land use, such as the impact of urban traffic on land use or the impact of land use on urban traffic [41,42]. In the correlation study of urban traffic and land use, the Pearson correlation coefficient was mainly used to study the correlation between the two [43,44]. However, there is little research on the interactive relationship between urban traffic and land use, and the existing research mainly focuses on qualitative analysis. Moreover, in the research into the urban traffic network, the spatial syntax and multiple centrality assessment models are mainly used for quantitative evaluation. ## 3.1. Study Area Jinan, the capital of Shandong Province, is also a city famous for spring water. A large number of karst caves are distributed underground in Jinan, which causes great difficulties for urban subway construction and leads to the lagging development of local rail transit. In addition, restricted by the topographical conditions, the Jinan urban area presents a narrow and long belt space form from east to west and short from north to south [45]. This often leads to serious traffic congestion in Jinan. This paper took Jinan, China, as the research case, which was typical and representative. With reference to the Master Plan of Jinan City (2011–2020) and the distribution of the urban road network in Jinan, the study area determined by this study is shown in Figure 1. ## 3.2.1. Multiple Centrality Assessment Model The multiple centrality assessment model takes the urban road as the edge of the network and the intersection point of the road and the end point of the road as the network node, and calculates the centrality of the urban traffic network by calculating the distance between actual road nodes [46,47]. The model mainly evaluates the importance of the road network by calculating closeness centrality, betweenness centrality and straightness centrality. Closeness centrality (CC) represents the degree of closeness between a transport network node and all other nodes in the transport network, reflecting the reachability of this node in the network. The formula for calculating the closeness centrality of a road node i is [1]CiC=N−1∑$j = 1$;j≠iNdij where N is the number of nodes in the traffic network and dij is the shortest distance between node i and node j. Betweenness centrality (CB) reflects the transfer capacity of road nodes in the traffic network. The stronger the betweenness centrality of road nodes is, the stronger the hub role it plays in the traffic network is. The formula for calculating the betweenness centrality of the road node i is [2]CiB=1(N−1)(N−2)∑$j = 1$;$k = 1$;j≠k≠1Nnjk(i)njk where N is the number of nodes in the traffic network, njk is the number of shortest paths between node j and node k and njk(i) is the number of paths through node i in the shortest paths between node j and node k. Straightness centrality (CS) reflects the importance of road nodes. The stronger the straightness centrality, the stronger the traffic efficiency of the traffic network is. [ 3]CiS=1N−1∑$j = 1$;j≠iNdijEucldij where N is the number of nodes in the traffic network, dijEucl represents the Euclidean distance between node i and node j and dij is the shortest distance between node i and node j. ## 3.2.2. Kernel Density Estimation (KDE) KDE calculates the density of points distributed in a specific window, and takes the sum of the density of all points in the window as the kernel density value of the grid center [48,49]. In this paper, KDE was used to smooth the POI data of the main urban area of Jinan, and a continuous spatial distribution map was obtained to indirectly represent the LUI. [ 4]f^(x)=1nhd∑$i = 1$nK(x−xih) where K represents the kernel function, h is the threshold, n is the number of points within the threshold range and d represents the data dimension. In this study, after considering the smoothness of the data and the detail of the reflected data, the bandwidth selected was 1000 m. ## 3.2.3. Spatial Three-stage Least Squares Method The spatial three-stage least squares method constructs a model using simultaneous equations for spatial econometric analysis. This method not only considers the potential spatial correlation of endogenous variables, but also considers the correlation between the random error terms of each equation, thus making the results more scientific and effective. It cannot only avoid the traditional simultaneous equation model and ignore the possible spatial spillover effect between variables, but it can also solve the problem of variable endogeneity that may be generated by the spatial econometric model [50]. The spatial correlation test found that the global Moran index of SC and LUI in the main urban area of Jinan was higher than 0.8, and passed the significance test, indicating that there was significant spatial autocorrelation between the two. Therefore, spatial factors need to be taken into account in the analysis of influencing factors. In this study, the spatial three-stage least squares method was used to explore the interactive relationship between SC and LUI. The established spatial simultaneous equation was as follows:SCi=α0+α1∑j≠inWijSCj+α2∑j≠inWijLUIj+α3LUIj+αXi+εi [5]LUIi=β0+β1∑j≠inWijLUIj+β2∑j≠inWijSCj+β3SCj+βZi+ηi where i represents the sample region; SCi and LUIi represent SC and LUI, respectively; Wij represents the spatial weight matrix; Xi and Zi represent a group of control variables that affect the SC and the LUI; εi and ηi represent unobservable factors; α1 represents the spatial spillover estimation coefficient of the SC between neighboring regions; β1 represents the spatial spillover estimation coefficient of the LUI between neighboring regions; α2 and β2 represent the spatial interaction used to test the SC and the LUI and α3 and β3 are used to characterize the endogenous relationship between the SC and the LUI. Endogenous variables: Street centrality (SC) and land use intensity (LUI). Considering the availability, scientificity and completeness of data, and referring to the existing research results [51,52,53,54], the control variables Xi that affect the SC were selected. Road network density (ND): the control variable was represented by the road length within 1 km2. Economic factors (EC): the control variable was represented by the regional GDP within 1 km2, unit: 10,000 yuan/km2. Location factor (LOC): the control variable was characterized by the distance to the CBD. The control variables Zi that affect the LUI: Population factor (POP): the control variable was represented by the population within 1 km2, unit: person/km2. Economic factors (EC): the control variable was represented by the regional GDP within 1 km2, unit: 10,000 yuan/km2. Traffic factor (TRA): the control variable was measured by bus stop density, which was obtained by calculating the number of bus stops within a 1 km2 grid. Location factor (LOC): the control variable was characterized by the distance from the CBD. Considering the influence of variable regional units, this paper constructed spatial grid units for quantitative analysis. First, the fishnet creation tool in ArcGIS software was used to construct a square grid covering the study area with a side length of 1 km. Then, the grid was used as a statistical unit to calculate the SC, LUI and other control variable indexes through spatial correlation and other methods. First of all, in order to avoid the influence of heteroscedasticity and multicollinearity on the research results, the original data of variables were processed logically and tested using the variance inflation factor (VIF). The test results showed that the VIF of each variable was less than 7.5, so it could be considered that there was no multicollinearity between variables. Secondly, the test results of the validity and feasibility of the model showed that the R2 of the SC equation and of the LUI equation were both greater than 0.5, indicating that the fitting effect of the model results was good. Therefore, the model was valid and feasible. ## 3.3. Data Sources and Processing The traffic network data of Jinan were obtained from the China National Geographic Information Public Service Platform. POI data can record the attribute information and spatial location of buildings or geographical entities. POI data of Jinan in 2020 were collected based on Gaode API. The ArcGIS software was used to extract the highways, main roads, secondary roads and main branches of the Jinan main urban area. Secondly, disordered and disconnected routes were eliminated, and repeated routes were screened and combined with remote sensing satellite images. The traffic network data of the study area after processing are shown in Figure 2. The POI data can intuitively reflect the spatial distribution of various urban functional elements and facilities. This paper used the re-classified POI data kernel density smoothing results to indirectly represent the LUI. Firstly, data cleansing was performed on the obtained POI data. Then, the POI data of commercial facilities, residential facilities and public service facilities were selected for research (Table 1). ## 4.1.1. Spatial Distribution Characteristics of SC The closeness centrality showed an obvious “core-edge” pattern in space (Figure 3), with higher closeness centrality in the central area and lower closeness centrality in the marginal area. The closeness centrality gradually decreases from the urban central area to the marginal area. Specifically, high-value areas were mainly concentrated in the areas around Daming Lake, such as CBD, the commercial port area, etc. This area is the central area of the main urban area of Jinan, with a large number of commercial and tourist facilities, and the traffic network in this area is densely distributed. The shortest distance from road nodes to other nodes in the traffic network is small, and the traffic network accessibility is strong. The spatial distribution of betweenness centrality was different from that of closeness centrality. High-value areas were mainly distributed in the main urban traffic roads, such as Quancheng Road, Luoyuan Street, Jiefang Road, Beiyuan Street, Jingshi Road and Lishan Road. The commercial facilities, tourist facilities and public service facilities on both sides of these main traffic roads are densely distributed, and the daily traffic flow and vehicle flow are large, which bear most of the daily traffic flow in the main urban area of Jinan. The straightness centrality showed an obvious multi-center structure in space. There were three high-value clusters in the east, middle and west, such as the Lashan Business District in the west, the West Railway Station Business District in the middle and the High-Tech Development Zone in the east. These areas are far from the old urban area of Jinan, and the roads are mainly large traffic trunk roads, such as Gongye North Road, Gongye South Road, Jingshi East Road, Tourist Road and Jingshi West Road. The straightness centrality of road nodes in these regions was high, indicating that the path distance between two road nodes in this region is closer to the linear distance between them and that the regional traffic efficiency is high. ## 4.1.2. Spatial Distribution Characteristics of LUI In this study, reclassified POI data of commercial facilities, residential facilities and public service facilities were used to indirectly characterize the following three types of land use: commercial land, residential land and public service land, and LUI was obtained through kernel density smoothing (Figure 4). CLUI showed a multi-core distribution pattern in space. CBD was the largest agglomeration center, and several secondary agglomeration centers were distributed around it. High-value areas of CLUI were mainly distributed around Daming Lake and the CBD area within the second ring road. This area is the old urban area of Jinan, where many large commercial service facilities are distributed, such as Ginza Shopping Plaza, World Trade Plaza and China Merchants Building. RLUI showed the spatial distribution characteristics of the coexistence of large and small cores. High-value RLUI areas were mainly distributed in the eastern area of Daming Lake and the area around CBD, forming a large core. Residential communities in this area are densely distributed, such as Qinghou Community, Zhengjuesi Community, Shunyu Community and Jinjiling Villa District. There were two small cores in the east and southwest of the large core of RLUI. The distribution of PLUI was similar to that of RLUI, which also showed the spatial distribution characteristics of coexisting large and small cores. High-value areas of PLUI were mainly concentrated within the second ring road, forming a large core. A large number of public service facilities are distributed in the area, such as Shandong Museum, Shandong Radio and Television News, Qilu Hospital of Shandong University and Qianfo Mountain Hospital of Shandong Province. There were two small cores in the east and southwest of the large core of PLUI. ## 4.2.1. Interaction between SC and CLUI The estimated results of the closeness centrality equation (Model 1) are shown in Table 2. The coefficient of the lag term (W × lnLUI) of CLUI was positive and passed the significance test, indicating that the CLUI in the adjacent area could promote an improvement in the closeness centrality of the local area. The coefficient of CLUI (lnLUI) was positive and passed the significance test, indicating that the increase in CLUI could promote an improvement in the regional closeness centrality. The coefficient of the lag term (W × lnSC) of closeness centrality was positive and passed the significance test, indicating that improving the closeness centrality in the adjacent area could promote the local closeness centrality. The influence coefficient of the road network density (lnND) was positive and passed the significance test, indicating that the greater the road network density, the higher the closeness centrality of road nodes. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the smaller the closeness centrality of road nodes. The estimated results of the CLUI equation (Model 2) are shown in Table 2. The coefficient of the lag term (W × lnSC) of closeness centrality was positive and passed the significance test, indicating that an improvement in closeness centrality in the adjacent area could promote an improvement in local CLUI. The influence coefficient of closeness centrality (lnSC) was positive and passed the significance test, indicating that improving the closeness centrality was conducive to improving CLUI. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development was conducive to improving CLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the CLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the local population density, the higher the CLUI. The influence coefficient of the traffic factor (lnTRA) was positive and passed the significance test, indicating that the higher the density of public transport stations, the higher the CLUI. The estimated results of the betweenness centrality equation (Model 3) are shown in Table 2. The coefficient of the lag term (W × lnLUI) of CLUI was positive and passed the significance test, indicating that improving the CLUI in the adjacent areas could promote an improvement in the betweenness centrality of the local area. The coefficient of the lag term (W × lnSC) of betweenness centrality was negative and passed the significance test, indicating that improving the betweenness centrality in the adjacent region could lead to a decrease in the local betweenness centrality. The influence coefficient of the road network density (W × lnND) was negative and passed the significance test, indicating that the higher the road network density, the lower the betweenness centrality. The estimated results of the CLUI equation (Model 4) are shown in Table 2. The coefficient of the lag term (W × lnSC) of betweenness centrality was positive and passed the significance test, indicating that improving the betweenness centrality in the adjacent areas could promote an improvement in local CLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the CLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the CLUI. The influence coefficient of the traffic factor (lnTRA) was positive and passed the significance test, indicating that increasing the density of public transport stations was conducive to improving the CLUI. The estimated results of the straightness centrality equation (Model 5) are shown in Table 2. The coefficient of the lag term (W × lnLUI) of CLUI was positive and passed the significance test, indicating that the CLUI in the adjacent areas could promote an improvement in the straightness centrality of the local area. The influence coefficient of CLUI (lnLUI) was positive and passed the significance test, indicating that improving the CLUI could promote an improvement in the straightness centrality. The coefficient of the lag term (W × lnSC) of the straightness centrality was negative and passed the significance test, indicating that an increase in straightness centrality in the adjacent areas could lead to a decrease in local straightness centrality. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the straightness centrality of road nodes. The estimated results of the CLUI equation (Model 6) are shown in Table 2. The coefficient of the lag term (W × lnSC) of the straightness centrality was positive and passed the significance test, indicating that improving the straightness centrality in the adjacent areas could promote an improvement in local CLUI. The influence coefficient of the straightness centrality (lnSC) was positive and passed the significance test, indicating that an improvement in straightness centrality could promote an improvement in CLUI. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development could promote an improvement in CLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the CLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the CLUI. ## 4.2.2. Interaction between SC and RLUI The estimated results of the closeness centrality equation (Model 1) are shown in Table 3. The coefficient of the lag term (W × lnLUI) of RLUI was positive and passed the significance test, indicating that the RLUI in the adjacent area could promote an improvement in the closeness centrality of the local area. The coefficient of RLUI (lnLUI) was positive and passed the significance test, indicating that an increase in RLUI could promote an improvement in the closeness centrality. The influence coefficient of the road network density (lnND) was positive and passed the significance test, indicating that increasing the road network density could promote an improvement in closeness centrality. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development was conducive to improving the closeness centrality of road nodes. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the smaller the closeness centrality of road nodes. The estimated results of the RLUI equation (Model 2) are shown in Table 3. The coefficient of the lag term (W × lnSC) of closeness centrality was positive and passed the significance test, indicating that an improvement in closeness centrality in the adjacent area could promote an improvement in local RLUI. The influence coefficient of closeness centrality (lnSC) was positive and passed the significance test, indicating that improving closeness centrality was conducive to improving RLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the RLUI. The influence coefficient of the traffic factor (lnTRA) was positive and passed the significance test, indicating that the higher the density of public transport stations, the higher the RLUI. The estimated results of the betweenness centrality equation (Model 3) are shown in Table 3. The coefficient of the lag term (W × lnLUI) of RLUI was positive and passed the significance test, indicating that improving the RLUI in the adjacent areas could promote an improvement in the betweenness centrality of the local area. The influence coefficient of RLUI (lnLUI) was negative and passed the significance test, indicating that increasing the RLUI could lead to a decrease in betweenness centrality. The coefficient of the lag term (W × lnSC) of betweenness centrality was negative and passed the significance test, indicating that improving the betweenness centrality in the adjacent region could lead to a decrease in the local betweenness centrality. The influence coefficient of the road network density (W × lnND) was negative and passed the significance test, indicating that the higher the road network density, the lower the betweenness centrality. The estimated results of the RLUI equation (Model 4) are shown in Table 3. The influence coefficient of betweenness centrality (lnSC) was negative and passed the significance test, indicating that an increase in betweenness centrality could lead to a decrease in RLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the RLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the RLUI. The estimated results of the straightness centrality equation (Model 5) are shown in Table 3. The coefficient of the lag term (W × lnLUI) of RLUI was positive and passed the significance test, indicating that the RLUI in the adjacent areas could promote an improvement in straightness centrality of the local area. The influence coefficient of RLUI (lnLUI) was positive and passed the significance test, indicating that an improvement in RLUI was conducive to an improvement in straightness centrality. The coefficient of the lag term (W × lnSC) of the straightness centrality was negative and passed the significance test, indicating that an increase in straightness centrality in the adjacent areas could lead to a decrease in local straightness centrality. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the straightness centrality of road nodes. The estimated results of the RLUI equation (Model 6) are shown in Table 3. The influence coefficient of straightness centrality (lnSC) was positive and passed the significance test, indicating that an improvement in straightness centrality was conducive to an improvement in RLUI. The coefficient of the lag term (W × lnLUI) of the RLUI lag was positive and passed the significance test, indicating that improving RLUI in the adjacent areas could promote an improvement in local RLUI. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that an improvement in the economic development level was conducive to an improvement in RLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the RLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the RLUI. ## 4.2.3. Interaction between SC and PLUI The estimated results of the closeness centrality equation (Model 1) are shown in Table 4. The coefficient of the lag term (W × lnLUI) of PLUI was positive and passed the significance test, indicating that the PLUI in the adjacent area could promote an improvement in the closeness centrality of the local area. The coefficient of PLUI (lnLUI) was positive and passed the significance test, indicating that increasing the PLUI was conducive to promoting an improvement in closeness centrality. The influence coefficient of the road network density (lnND) was positive and passed the significance test, indicating that increasing road network density was conducive to promoting an improvement in closeness centrality. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test at the level of $10\%$, indicating that an improvement in the economic development level was conducive to an improvement in closeness centrality. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the smaller the closeness centrality. The estimated results of the PLUI equation (Model 2) are shown in Table 4. The coefficient of the lag term (W × lnSC) of closeness centrality was positive and passed the significance test, indicating that an improvement in closeness centrality in the adjacent area could promote an improvement in local PLUI. The influence coefficient of closeness centrality (lnSC) was positive and passed the significance test, indicating that an improvement in closeness centrality was conducive to an improvement in PLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the closer to the city center, the higher the PLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the local population density, the higher the PLUI. The estimated results of the betweenness centrality equation (Model 3) are shown in Table 4. The influence coefficient of PLUI (lnLUI) was negative and passed the significance test, indicating that an increase in PLUI could lead to a decrease in local betweenness centrality. The coefficient of the lag term (W × lnSC) of betweenness centrality was negative and passed the significance test, indicating that improving the betweenness centrality in the adjacent region could lead to a decrease in the local betweenness centrality. The influence coefficient of the road network density (W × lnND) was negative and passed the significance test, indicating that the higher the road network density, the lower the betweenness centrality. The estimated results of the PLUI equation (Model 4) are shown in Table 4. The influence coefficient of betweenness centrality (lnSC) was negative and passed the significance test, indicating that an increase in betweenness centrality could lead to a decrease in the PLUI. The coefficient of the lag term (W × lnLUI) of public service land was positive and passed the significance test, indicating that an improvement in the PLUI in the adjacent areas could promote an improvement in the local PLUI. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development could promote an improvement in the PLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the PLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the higher the population density, the higher the PLUI. The influence coefficient of the traffic factor (lnTRA) was positive and passed the significance test, indicating that the higher the density of public transport stations, the higher the PLUI. The estimated results of the straightness centrality equation (Model 5) are shown in Table 4. The influence coefficient of PLUI (lnLUI) was positive and passed the significance test, indicating that an improvement in PLUI could promote an improvement in straightness centrality. The coefficient of the lag term (W × lnSC) of the straightness centrality was negative and passed the significance test, indicating that an increase in straightness centrality in the adjacent areas could lead to a decrease in local straightness centrality. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development was conducive to promoting an improvement in straightness centrality. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the straightness centrality of road nodes. The estimated results of the PLUI equation (Model 6) are shown in Table 4. The coefficient of the lag term (W × lnSC) of the straightness centrality was negative and passed the significance test, indicating that the straightness centrality in adjacent areas had a negative effect on the local PLUI. The influence coefficient of the straightness centrality (lnSC) was positive and passed the significance test, indicating that an improvement in straightness centrality can promote an improvement in PLUI. The coefficient of the lag term (W × lnLUI) of the PLUI was positive and passed the significance test, indicating that improving the PLUI in the adjacent areas could promote an improvement in the local PLUI. The influence coefficient of the economic factors (lnEC) was positive and passed the significance test, indicating that improving the level of economic development could promote an improvement in the PLUI. The influence coefficient of the location factor (lnLOC) was negative and passed the significance test, indicating that the farther away from the city center, the lower the PLUI. The influence coefficient of population size (lnPOP) was positive and passed the significance test, indicating that the greater the population density, the higher the PLUI. ## 5.1. Impact of SC on LUI The statistical results of the significant influence coefficient of SC on LUI are shown in Table 5. The influence coefficients of closeness centrality on CLUI, RLUI and PLUI were all positive, indicating that improving the closeness centrality was conducive to improving regional LUI, which is consistent with the research results of Yin et al. [ 28]. The impact of closeness centrality on LUI was as follows: public service land > residential land > commercial land. This may be attributed to the wide distribution of public service land in the main urban area of Jinan, which was more strongly influenced by the accessibility of the transportation network. The influence coefficients of betweenness centrality on CLUI, RLUI and PLUI were all negative, indicating that increasing the closeness centrality of the transportation network could lead to a decrease in regional LUI. The impact of betweenness centrality on LUI was as follows: public service land > residential land > commercial land. This may be attributed to the wide distribution of public service land in the main urban area of Jinan, which was more strongly influenced by the betweenness centrality of the transportation network. The influence coefficients of straightness centrality on CLUI, RLUI and PLUI were all positive, indicating that improving the straightness centrality of the transport network was conducive to improving regional LUI. The impact of straightness centrality on LUI was as follows: public service land > commercial land > residential land. This may be attributed to the fact that public service land and commercial land were more affected by the straightness centrality of the transportation network. In general, the SC in the main urban area of Jinan had the strongest effect on the PLUI and the least impact on the CLUI, which is contrary to the research results of some European cities [55,56]. This may be attributed to the wider distribution of public service land in the main urban area of Jinan. For example, public service facilities such as hospitals and schools are highly concentrated in the main urban area of Jinan. The old urban area is the central area of the main urban area of Jinan, where commercial facilities are highly concentrated and land use intensity is high. This is different from the infrastructure distribution of major cities in the United States [57]. In the fringe area of the main urban area, owing to it being far away from the urban center, there are fewer commercial facilities, residential facilities and public services, and the intensity of land use is low. Therefore, we can improve the distribution pattern of transportation facilities and improve the closeness and straightness centrality of the edge areas, which are conducive to improving the LUI. ## 5.2. Impact of LUI on SC The statistical results of the significant influence coefficient of LUI on SC are shown in Table 6. The influence coefficients of CLUI on closeness centrality and straightness centrality were both positive, while the influence coefficient on betweenness centrality was negative, indicating that improving CLUI was conducive to improving the accessibility and traffic efficiency of the transport network, but could lead to a decline in the mediating effect of the transport network, which is consistent with the research results of Chen et al. [ 45]. The impact of CLUI on SC was as follows: straightness centrality > closeness centrality > betweenness centrality. This may be attributed to the fact that the traffic efficiency of the traffic network in the main urban area of Jinan was more strongly influenced by the commercial land use intensity, which is similar to the study of Liu et al. [ 58]. The influence coefficients of RLUI on closeness centrality and straightness centrality were both positive, while the influence coefficient on betweenness centrality was negative. This shows that improving RLUI was conducive to improving the accessibility and traffic efficiency of the transportation network, but could lead to a decline in the intermediary role of the transportation network. The impact of RLUI on the SC was as follows: straightness centrality > betweenness centrality > closeness centrality. This may be attributed to the fact that the traffic efficiency of the transportation network in the main urban area of Jinan was more strongly influenced by residential land use. The influence coefficients of PLUI on closeness centrality and straightness centrality were both positive, while the influence coefficient on betweenness centrality was negative, indicating that improving the PLUI was conducive to improving the accessibility and traffic efficiency of the transport network, but could lead to a decline in the mediating effect of the transport network. The impact of the PLUI on the SC was as follows: betweenness centrality > straightness centrality > closeness centrality. This may be attributed to the fact that the mediating effect of the transportation network in the main urban area of Jinan was more strongly influenced by the PLUI. In general, the SC was most strongly influenced by the PLUI, which is consistent with the wide distribution of public service land in the main urban area of Jinan [28]. LUI had the strongest influence on the straightness centrality of the transportation network, which also shows that the traffic efficiency of the transportation network was most affected by LUI [59]. The traffic accessibility and traffic efficiency of the marginal areas of the main urban area of Jinan were low, meaning that they were not conducive to the sustainable development of urban traffic. Therefore, local government departments can improve the infrastructure construction of the region, such as building residential facilities and public service facilities of a certain scale, improving the intensity of regional land use, and thus improving regional traffic accessibility and traffic efficiency. This study used the spatial three-stage least squares method to explore the two-way interaction between SC and LUI, which was innovative to a certain extent. However, this study also had the following shortcomings: First, the reclassified POI data were used to indirectly represent the urban functional land, but they were not scientific and accurate. In the analysis of influencing factors, owing to the difficulty in obtaining micro-data, this study selected fewer control variables, and did not consider the possible impacts of natural environmental factors and policy factors on LUI and SC. The follow-up study will further expand the scope of study to multiple cities through a series of studies, and summarize whether there are general rules in different types of cities. Moreover, to overcome the difficulty of data acquisition, we will try to measure the commercial land, residential land and public service land more accurately based on plot area ratio, and compare these with the results of this study. ## 6. Conclusions and Policy Suggestions This study explored the spatial distribution characteristics of SC and LUI in the main urban area of Jinan. Moreover, using the spatial three-stage least squares method, this study explored the two-way interaction between the SC and the LUI. The main research conclusions were as follows. The closeness centrality showed an obvious “core-edge” pattern. Betweenness centrality and straightness centrality showed a multicentric structure. CLUI showed a multi-core distribution pattern in space. CBD was the largest agglomeration center, and several secondary agglomeration centers were distributed around it. RLUI and PLUI showed the characteristics of spatial distribution with the coexistence of large and small cores. There was an interactive relationship between SC and LUI. Closeness centrality and straightness centrality had positive effects on LUI, and LUI had a positive effect on closeness centrality and straightness centrality. Betweenness centrality had a negative effect on LUI, and LUI had a negative effect on betweenness centrality. Moreover, the influence of the road network density and of the location factors on SC were the strongest. The greater the road network density and the closer to the city center, the higher the closeness centrality and straightness centrality, and the lower the betweenness centrality. Location factors, population density and traffic factors had the strongest effect on LUI. The closer the distance to the city center, the greater the population density, and the higher the density of bus stations, the higher the LUI. According to the research results, this paper put forward the following policy suggestions for the future urban development of Jinan. First of all, the local government can formulate reasonable urban development policies, optimize the land use structure of the old urban area of Jinan, and promote the transformation and upgrading of transportation facilities in the old urban area. Secondly, it can improve the infrastructure construction of the new urban area and enhance the land use intensity of the new urban area. Finally, it can coordinate the construction of new and old urban areas, which is conducive to promoting the coordinated development of urban transportation and land use. ## References 1. Burgess E.W., Park R.E., Burgess E.W., Mackenzie R.D.. The growth of the city: An introduction to a research project. The City (1925.0) 1-47 2. Hoyt H.. The Structure and Growth of Residential Neighborhoods in American Cities (1940.0) 3. Harris C.D., Ullman E.L.. The Nature of Cities. Ann. Am. Acad. Political Soc. Sci. (1945.0) 242 7-17. DOI: 10.1177/000271624524200103 4. Alonso W.. Location and Land Use (1964.0) 5. Mills E.S.. Studies in the Structure of the Urban Economy (1972.0) 6. Muth R.F.. Cities and Housing (1969.0) 7. Yu J., Zhou K., Yang S.. Land use efficiency and influencing factors of urban agglomerations in China. Land Use Policy (2019.0) 88 104143. DOI: 10.1016/j.landusepol.2019.104143 8. Frank L.D., Saelens B.E., Powell K.E., Chapman J.E.. Stepping towards causation: Do built environments or neighborhood and travel preferences explain physical activity, driving, and obesity?. Soc. Sci. Med. (2007.0) 65 1898-1914. DOI: 10.1016/j.socscimed.2007.05.053 9. Zhang W., Kockelman K.M.. Optimal policies in cities with congestion and agglomeration externalities: Congestion tolls, labor subsidies, and place-based strategies. J. Urban Econ. (2016.0) 95 64-86. DOI: 10.1016/j.jue.2016.08.003 10. Lee M., Barbosa H., Youn H., Holme P., Ghoshal G.. Morphology of travel routes and the organization of cities. Nat. Commun. (2017.0) 8 2229. DOI: 10.1038/s41467-017-02374-7 11. Schaeffer K.H., Sclar E.. Access for All: Transportation and Urban Growth (1975.0) 12. Giannopoulos G.A., Curdes G.. Innovations in Urban Transport and the Influence on Urban Form. An Historical Review. Transp. Rev. (1992.0) 12 15-32. DOI: 10.1080/01441649208716801 13. Moon H.. Land Use Around Suburban Transit Stations. Transportation (1990.0) 17 67-88. DOI: 10.1007/BF02125504 14. Baerwald T.J.. The Site Selection Process of Suburban Residential Builders. Urban Geogr. (1981.0) 2 339-357. DOI: 10.2747/0272-3638.2.4.339 15. Rui Y., Ban Y.. Exploring the relationship between street centrality and land use in Stockholm. Int. J. Geogr. Inf. Sci. (2014.0) 28 1425-1438. DOI: 10.1080/13658816.2014.893347 16. Mao J., Yan X.. Research on the impact of high-density development of urban transportation system on land use—Taking Guangzhou as an example. Econ. Geogr. (2005.0) 185-188 17. Sun J., Zhang A., Zhou Z.. Research on urban rail transit promoting urbanization process. China Soft Sci. (2020.0) 96-111 18. Giuliaono G., Naravan D.. Another Look at Travel Patterns and Urban Form: The US and Great Britain. Urban Stud. (2003.0) 40 2295-2312. DOI: 10.1080/0042098032000123303 19. Johnson A.. Bus Transit and Land Use: Illuminating the Interaction. J. Public Transp. (2003.0) 6 21-39. DOI: 10.5038/2375-0901.6.4.2 20. Zhang M.. The Role of Land Use in Travel Mode Choice: Evidence from Boston and Hong Kong. J. Am. Plan. Assoc. (2004.0) 70 344-361. DOI: 10.1080/01944360408976383 21. Ding C.. Impact of urban spatial structure and land use mode on urban traffic. Urban Transp. (2010.0) 5 28-35 22. Casetti E.. Spatial analysis: Perspectives and prospects. Urban Geogr. (1993.0) 14 526-537. DOI: 10.2747/0272-3638.14.6.526 23. Ghosh A., McLafferty S.. Location Strategies for Retail and Service Firms (1987.0) 24. Cutini V.. Configuration and centrality-some evidence from two Italian case studies. Proceedings of the Space Syntax 3rd International Symposium 25. Wang S., Xu G., Guo Q.. Street Centralities and Land Use Intensities Based on Points of Interest (POI) in Shenzhen, China. ISPRS Int. J. Geo-Inf. (2018.0) 7. DOI: 10.3390/ijgi7110425 26. Tian T., Gu D., Niu D.. Evaluation of the coordination relationship between urban land use and urban transportation—Taking the main urban area of Chongqing as an example. J. Southwest Univ. (Nat. Sci. Ed.) (2019.0) 41 96-104 27. Yin C., Xi Y., Liu Y.. Analysis of the relationship between road network centrality and land use landscape pattern in Hubei Province. Yangtze River Basin Resour. Environ. (2017.0) 26 1388-1396 28. Yin G., Liu T., Chen Y., Hou Y.. Disparity and Spatial Heterogeneity of the Correlation between Street Centrality and Land Use Intensity in Jinan, China. Int. J. Environ. Res. Public Health. (2022.0) 19. DOI: 10.3390/ijerph192315558 29. Batty M., Fisher P., Unwin D.. Network geography: Relations, interactions, scaling and spatial processes in GIS. Re-Presenting Geographical Information Systems (2003.0) 149-170 30. Kim H.K., Sohn D.W.. An analysis of the relationship between land use density of office buildings and urban street configuration: Case studies of two areas in Seoul by space syntax analysis. Cities (2002.0) 19 409-418. DOI: 10.1016/S0264-2751(02)00071-9 31. Scoppa M.D., Peponis J.. Distributed attraction: The effects of street network connectivity upon the distribution of retail frontage in the city of Buenos Aires. Environ. Plan. B-Plan. Des. (2015.0) 42 354-378. DOI: 10.1068/b130051p 32. Omer I., Goldblatt R.. Spatial patterns of retail activity and street network structure in new and traditional Israeli cities. Urban Geogr. (2016.0) 37 629-649. DOI: 10.1080/02723638.2015.1101258 33. Wang F., Chen C., Xiu C., Zhang P.. Location analysis of retail stores in Changchun, China: A street centrality perspective. Cities (2014.0) 41 54-63. DOI: 10.1016/j.cities.2014.05.005 34. Kang C.. Spatial access to pedestrians and retail sales in Seoul, Korea. Habitat Int. (2016.0) 57 110-120. DOI: 10.1016/j.habitatint.2016.07.006 35. Lin G., Chen X., Liang Y.. The location of retail stores and street centrality in Guangzhou, China. Appl. Geogr. (2018.0) 100 12-20. DOI: 10.1016/j.apgeog.2018.08.007 36. Hillier B., Hanson J.. The Social Logic of Space (1984.0) 37. Li Q., Zhou S., Wen P.. The relationship between centrality and land use patterns: Empirical evidence from five Chinese metropolises. Comput. Environ. Urban Syst. (2019.0) 78 101356. DOI: 10.1016/j.compenvurbsys.2019.101356 38. Crucitti P., Latora V., Porta S.. Centrality in networks of urban streets. Chaos (2006.0) 16 15113. DOI: 10.1063/1.2150162 39. Crucitti P., Latora V., Porta S.. Centrality measures in spatial networks of urban streets. Phys. Rev. E (2006.0) 73 36125. DOI: 10.1103/PhysRevE.73.036125 40. Al-Saaidy H.J.E., Alobaydi D.. Studying street centrality and human density in different urban forms in Baghdad, Iraq. Ain Shams Eng. J. (2021.0) 12 1111-1121. DOI: 10.1016/j.asej.2020.06.008 41. Shen T., Hong Y., Thompson M.M., Liu J., Huo X., Wu L.. How does parking availability interplay with the land use and affect traffic congestion in urban areas? The case study of Xi’an, China. Sustain. Cities Soc. (2020.0) 57 102126. DOI: 10.1016/j.scs.2020.102126 42. Zhou H., Gao H.. The impact of urban morphology on urban transportation mode: A case study of Tokyo. Case Stud. Transp. Policy (2018.0) 8. DOI: 10.1016/j.cstp.2018.07.005 43. Bao Z., Ou Y., Chen S., Wang T.. Land Use Impacts on Traffic Congestion Patterns: A Tale of a Northwestern Chinese City. Land (2022.0) 11. DOI: 10.3390/land11122295 44. Sarzynski A., Wolman H.L., Galster G., Hanson R.. Testing the conventional wisdom about land use and traffic congestion: The more we sprawl, the less we move?. Urban Stud. (2006.0) 43 601-626. DOI: 10.1080/00420980500452441 45. Chen Y., Yin G., Hou Y.. Street centrality and vitality of a healthy catering industry: A case study of Jinan, China. Front. Public Health (2022.0) 10 1032668. DOI: 10.3389/fpubh.2022.1032668 46. Wang F., Antipova A., Porta S.. Street centrality and land use intensity in Baton Rouge, Louisiana. J. Transp. Geogr. (2011.0) 19 285-293. DOI: 10.1016/j.jtrangeo.2010.01.004 47. Wang F.. Quantitative Methods and Socio-Economic Applications in GIS (2015.0) 48. Fotheringham A.S., Brunsdon C., Charlton M.. Quantitative Geography: Perspectives on Spatial Data Analysis (2000.0) 146-149 49. Downs J.A., Horner M.W.. Characterising linear point patterns. Proceedings of the GIS Research UK Annual Conference 50. Xing H., Meng Y.. Revealing deep semantic commercial patterns: Insights from urban landscape depiction. Comput. Environ. Urban Syst. (2020.0) 79 101404. DOI: 10.1016/j.compenvurbsys.2019.101404 51. Lau B.P.L., Ng B.K.K., Yuen C., Tuncer B., Chong K.H.. The Study of Urban Residential’s Public Space Activeness Using Space-Centric Approach. IEEE Internet Things J. (2021.0) 8 11503-11513. DOI: 10.1109/JIOT.2021.3051343 52. Tang J., He Y., Ta N.. Spatial distribution characteristics and influencing factors of Shanghai catering industry based on POI data. Trop. Geogr. (2020.0) 40 1015-1025 53. Zeng X., Cui H., Liu Z.. Distribution and evolution characteristics and influencing factors of restaurants in Guangzhou. Econ. Geogr. (2019.0) 39 143-151 54. Wei M., Haijuan Y., Yuying C., Jiawei M.. The implementation of land planning indicators and its influencing factors in the relocation of immigrants in southern Shaanxi—Taking Shangzhou District as an example. J. Northwest Univ. (Nat. Sci. Ed.) (2022.0) 52 43-51 55. Porta S., Latora V., Wang F., Strano E., Cardillo A., Scellato S., Iacoviello V., Messora R.. Street centrality and densities of retail and services in Bologna, Italy. Environ. Plan. B-Plan. Des. (2009.0) 36 450-465. DOI: 10.1068/b34098 56. Porta S., Latora V., Wang F., Rueda S., Strano E., Scellato S., Cardillo A., Belli E., Cardenas F., Cormenzana B.. Street centrality and the location of economic activities in barcelona. Urban Stud. (2012.0) 49 1471-1488. DOI: 10.1177/0042098011422570 57. Chang Y.S., Lee Y.J., Choi S.S.B.. Is there more traffic congestion in larger cities?—Scaling analysis of the 101 largest US urban centers. Transp. Policy (2017.0) 59 54-63 58. Liu Y., Wei X., Jiao L., Wang H.. Relationships between street centrality and land use intensity in Wuhan, China. J. Urban Plan. Dev. ASCE (2016.0) 142 5015001. DOI: 10.1061/(ASCE)UP.1943-5444.0000274 59. Zhao P., Hu H.. Geographical patterns of traffic congestion in growing megacities: Big data analytics from Beijing. Cities (2019.0) 92 164-174. DOI: 10.1016/j.cities.2019.03.022
--- title: 'Metformin May Alter the Metabolic Reprogramming in Cancer Cells by Disrupting the L-Arginine Metabolism: A Preliminary Computational Study' authors: - Bryan Alejandro Espinosa-Rodriguez - Daniela Treviño-Almaguer - Pilar Carranza-Rosales - Monica Azucena Ramirez-Cabrera - Karla Ramirez-Estrada - Eder Ubaldo Arredondo-Espinoza - Luis Fernando Mendez-Lopez - Isaias Balderas-Renteria journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049129 doi: 10.3390/ijms24065316 license: CC BY 4.0 --- # Metformin May Alter the Metabolic Reprogramming in Cancer Cells by Disrupting the L-Arginine Metabolism: A Preliminary Computational Study ## Abstract Metabolic reprogramming in cancer is considered to be one of the most important hallmarks to drive proliferation, angiogenesis, and invasion. AMP-activated protein kinase activation is one of the established mechanisms for metformin’s anti-cancer actions. However, it has been suggested that metformin may exert antitumoral effects by the modulation of other master regulators of cellular energy. Here, based on structural and physicochemical criteria, we tested the hypothesis that metformin may act as an antagonist of L-arginine metabolism and other related metabolic pathways. First, we created a database containing different L-arginine-related metabolites and biguanides. After that, comparisons of structural and physicochemical properties were performed employing different cheminformatic tools. Finally, we performed molecular docking simulations using AutoDock 4.2 to compare the affinities and binding modes of biguanides and L-arginine-related metabolites against their corresponding targets. Our results showed that biguanides, especially metformin and buformin, exhibited a moderate-to-high similarity to the metabolites belonging to the urea cycle, polyamine metabolism, and creatine biosynthesis. The predicted affinities and binding modes for biguanides displayed good concordance with those obtained for some L-arginine-related metabolites, including L-arginine and creatine. In conclusion, metabolic reprogramming in cancer cells by metformin and biguanides may be also driven by metabolic disruption of L-arginine and structurally related compounds. ## 1. Introduction Metabolic reprogramming is an adaptation mechanism implemented by cancer cells in response to their microenvironment, which alters their metabolism to sustain the energy request for growth and proliferation [1]. Usually, the tumor microenvironment is hypoxic, acidic, and low in nutrients; despite these conditions, the cells present an abnormal metabolism to maintain themselves and to keep growing, including alterations in the tricarboxylic acids (TCA) cycle, glycolysis, the urea cycle (UC), nitric oxide (NO) metabolism, polyamines biosynthesis, and protein, lipid, and nucleic acid biosynthesis [2]. Today, one of the most important emergent therapies with remarkable potential is to modulate the metabolic reprogramming of cancer cells [2]. L-Arginine plays a central axis due to its ability to be incorporated into the anabolic and catabolic pathways mentioned before. In non-cancer cells, L-arginine is usually derived from exogenous uptake from the diet and endogenous biosynthesis through UC intermediates [3]. Once it has been absorbed, L-arginine can be incorporated into protein synthesis, acting as a building block. Moreover, L-arginine is reported to positively regulate anabolism itself by upregulating mTORC1 activity through the modulation of CASTOR1 and SLC38A9, the arginine sensors of the cells [4]. In addition, L-arginine is the precursor of a plethora of substances needed for proliferation, immune system regulation, DNA repair, or regulation of gene expression, such as polyamines, NO, and creatine. L-Arginine is key in two polyamine-producing pathways: the first pathway is the decarboxylation of L-arginine to agmatine by arginine decarboxylase (ADC) and subsequent hydrolysis via agmatinase, culminating with putrescine production; the second pathway involves L-ornithine, which is derived from L-arginine hydrolysis via arginase (ARG) in the UC. In addition, L-arginine is also involved as a precursor in the nitric oxide synthase (NOS)-catalyzed NO production and in the protein arginine methyltransferase (PRMT)-dependent production of the dimethylated derivatives of L-arginine: asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA). There are three NOS isoforms: inducible (iNOS), endothelial (eNOS), and neural (nNOS); eNOS is involved in vascular tone regulation, but iNOS is normally overexpressed during inflammation. On the other hand, ADMA is considered to be a potent competitive inhibitor of NOS that leads to less NO production. In addition to this, in the kidneys, L-arginine serves as the precursor of guanidinoacetate (GAA) via arginine:glycine amidinotransferase (AGAT), which is transported to the liver and converted into creatine by guanidinoacetate N-methyltransferase (GAMT) [3]. Interconversion between creatine and phosphocreatine in the cells in the so-called phosphagen system can maintain the different ATP pools by phosphorylating ADP to ATP via different isoforms of creatine kinases (CK), such as the muscle-type and brain-type CK [5]. Cancer cells reprogram L-arginine metabolism to support the proliferation and progression of malignancy, suppressing the immune response, and regulating gene expression [2]. The absorption of L-arginine in cancer cells is achieved by the overexpression of cationic amino acid transporters [6]. Indeed, hyperactivation of mTORC1 by L-arginine sensors has been reported in several types of cancer and is associated with increased anabolic pathways, such as the biosynthesis of nucleotides, protein, and fatty acids, and with the suppression of autophagy [7]. In tumors, UC was reported to be disrupted at different points. For instance, ORNT1, OTC, ASS1, and ASL were commonly downregulated. Additionally, it was observed at a higher carbamoyl phosphate concentration and its consequent input in the pyrimidine biosynthesis by carbamoyl phosphate synthetase 2 (CPS2) [8]. These metabolic changes led to increases in cellular growth and survival. Additionally, increased polyamines concentrations are usually observed in cancer cells. Polyamines have been implicated in nucleic acids and protein synthesis, chromatin stabilization, the regulation of paracrine communication, and prevention against oxidative DNA damage [9]. It has been reported that cancer cells exhibit increased ARG expression and activity, leading to a high production of L-ornithine [10]. Some studies also reported the overexpression of ODC1 due to MYC activity with a subsequent elevation in the putrescine concentration [11,12]. These alterations contribute to cancer progression. Furthermore, cancer cells can restore energy by using the phosphagen system to allow increased proliferation and survival. Additionally, creatine plays a key role in immunity by modulating T cells and acting as an efficient energy buffering mechanism when the cells demand high levels of ATP due to the high phosphate transfer potential of phosphocreatine [13,14]. Although some studies have showed contradictory results about expression of CK isoforms, CK has been correlated with the cancer prognosis [14,15,16]. Regarding NO metabolism, it was reported that NO had a paradoxical dual activity, i.e., it promoted several hallmarks of cancer, such as apoptosis inhibition, epithelial-to-mesenchymal transition induction, and increased vascular infiltration and permeability, but at the same time, it was reported that NO counteracts the mechanisms mentioned before [17]. Further, decreased ADMA levels were reported in prostate cancer due to the upregulation of ADMA breakdown via dimethylarginine dimethylaminohydrolase (DDAH), leading to high NOS activity and angiogenesis [18]. Taking advantage of L-arginine dependency in some tumors, in arginine deprivation therapy (ADT), L-arginine levels are decreased by administering arginine-depleting agents, such as pegylated arginine deiminase or pegylated arginase 1 [19]. ADT has shown promising results due to the pivotal role of L-arginine and their related metabolites involved in cancer metabolism, e.g., NO, ADMA, polyamines, and creatine, although cancer sensitivity seems to be dependent on the urea cycle enzymes [19,20]. Hence, the modulation of metabolic reprogramming by deprivation therapies or other approaches seems a rational alternative to some specific cancer types, and more research is needed to improve its efficacy and application. In pharmacology, the disruption of specific metabolic pathways has been performed using structurally related molecules that resemble endogenous ligands. For instance, statins imitate the structure of β-hydroxy-β-methylglutaryl-CoA (HMG-CoA), leading to the inhibition of HMG-CoA reductase and blocking cholesterol biosynthesis [21]. In line with this, there are several reports about metformin altering different metabolites related to L-arginine, suggesting the potential to interfere with UC, creatine biosynthesis, NO production, and polyamines metabolism [22]. Metformin, buformin, and phenformin are biguanides with anti-diabetic properties but nowadays, metformin is the only biguanide on the market and the first-line treatment for management of type 2 diabetes mellitus. Beyond their anti-diabetic properties, anticancer effects have been described in both in vitro and in vivo assays [23]. Despite it being such a versatile drug, the mechanism (or mechanisms) of action behind these effects remain elusive. Currently, the most established anti-cancer mechanism of action is the activation of AMP-activated protein kinase (AMPK) due to the energy deficit induced by the metformin-dependent inhibition of mitochondrial complex 1. In turn, AMPK activation negatively regulates those energy-consuming metabolic pathways, e.g., gluconeogenesis, glycogen synthesis, and fatty acid synthesis, but upregulates energy-generating metabolic pathways, e.g., glycolysis, glycogen breakdown, fatty acid oxidation, and autophagy [24]. However, there are some inconsistencies about this previous description related to the simplicity of metformin’s chemical structure. With a small structure and the lack of directing groups compared to other complex drugs, it is virtually impossible that metformin could bind to only one target in the cell. Even the more complex drugs usually act through binding to different targets [25]. Molecular promiscuity appears to be a more logical approach to understand metformin’s anti-cancer mechanism of action. Given that the basic metformin’s pharmacophore is the guanidine group, it is reasonable to think that it may resemble guanidine-containing endogenous ligands inside the cell, e.g., L-arginine, creatine, ADMA, or some other related metabolites (Figure 1). Our hypothesis states that metformin (and biguanides in general) may be acting in the cancer cell as an antagonist of L-arginine and its related metabolites, leading to modulation of their corresponding targets, and, therefore, altering the metabolic reprogramming of cancer cells. Computational tools can be employed to generate, transform, interpret, predict, and visualize data of a chemical and biological nature. These tools are commonly applied in drug discovery and exploration of the biological effects of molecules, allowing the prediction of physicochemical properties and drug-likeness, or even the prediction of possible targets according to the ligand structure [26]. In bioinformatics, molecular dynamics and molecular docking are commonly used to model the binding of ligands to specific targets, and to predict the affinities of ligands to targets of interest. However, molecular docking is normally used as a quick and inexpensive exploratory tool to generate preliminary insights to design further experiments [27]. On the other hand, given the computational power needed to run molecular dynamics simulations and the long waiting times needed to perform the thermodynamic calculations, it has been used as a more reliable computational technique to further validate the interactions observed in molecular docking simulations or experimental evidence [26]. In addition, it is common to use machine learning algorithms to extract relevant and significant information from the data [28]. In general, computational tools have demonstrated a high potential to generate, develop, and sustain hypotheses, as well as, to gain insights about the molecular actions of substances, as a first step before experiment design and performance [26]. For instance, exploration of the possible mechanism of action behind the arrest of the cell cycle in the sub-G1 phase in cancer cells led to the identification of possible interactions of geranyl farnesol, sahandinone, and 4-dehydrosalvilimbinol, a group of terpenoids present in Salvia lachnocalyx, with DNA topoisomerase I through a computational methodology centered on molecular docking and molecular dynamics simulations, suggesting that these terpenoids could be good candidates for the design of new drugs [29]. On the other hand, it has been reported that six possible targets genes (e.g., AR, HSP90AA1, MMP9, PGR, PTGS2, and TNF) are responsible for the action of the pseudo phosphorous stem of Cremastra appendiculata by employing a reverse network pharmacology approach that can allow for the in vitro or in vivo targeted study of molecular actions in the future [30]. The computational methodology they followed included tools for predicting the physicochemical properties and biological targets, such as SwissADME and SwissTargetPrediction, respectively, as well as tools to predict the binding and affinity of ligands in different targets of interest, such as molecular docking [30]. To test this hypothesis, in the present work, we employed different cheminformatic tools to seek for structural and physicochemical similarities between biguanides and different L-arginine-related metabolites by employing tools, such as the PubChem Score Matrix Service, SwissADME, and SwissTargetPrediction. Finally, we employed molecular docking, a bioinformatic technique, to test if the chemical similarity observed for biguanides may be translated into similar binding modes and affinities to those of L-arginine-related metabolites in their corresponding targets. Given that the complete metformin’s mechanism of action is still unknown, the relevance behind our hypothesis and this study is the establishment and support of a new possible theoretical framework to progress in the understanding of metformin’s biological actions in cancer cells, although this knowledge may be extrapolated to improve the comprehension of its therapeutic effects in other diseases where L-arginine, or its related metabolites, play a pivotal role. ## 2.1. Database Creation Before performing the different comparisons, we created a database containing the candidate metabolites, whose simplified molecular-input line-entry system (SMILES) and their compound ID (CID) were collected from PubChem. There were 20 candidate molecules in the final version of the constructed database, including 3 biguanides (metformin, buformin, and phenformin), 7 members of the UC (L-ornithine, L-citrulline, carbamoyl phosphate, L-aspartic acid, L-argininosuccinic acid, fumaric acid, and urea), L-arginine and their endogenous methylated derivatives (ADMA and SDMA), 5 members from the polyamine metabolism (agmatine, putrescine, spermine, spermidine, and cadaverine), and 2 members from creatine biosynthesis (guanidinoacetic acid and creatine). ## 2.2. Structural Comparison between Biguanides and L-arginine-Related Metabolites After generating the database, we proceeded to assess the structural similarity between biguanides and the candidate metabolites by employing the PubChem Score Matrix Service to compare their chemical structures at two levels: the 2D level and 3D level. Selected results about the comparison of biguanides against candidate metabolites are shown in Table 1 for the 2D and 3D analyses, while the complete results can be observed in the Supplementary Material. Tanimoto coefficients (TC) were generated from 2D structural comparisons. They ranged from 0 to 100 depending on the grade of the 2D structural similarity. The higher the TC value is, the more structurally similar the compared molecules are. Given metformin, buformin, and phenformin share the same pharmacophore, we expected to see high TC values (TC ≥ 65) between them. However, as shown in Table 1, a moderate similarity was observed only between metformin and buformin with a TC of 63, while phenformin showed low similarities with metformin (TC of 23) and buformin (TC of 33). Compared to those of the candidate metabolites, there were moderate similarities observed for buformin against different polyamines, e.g., cadaverine, spermidine, spermine, and putrescine, with TC values ranging from 51 to 62. It is noteworthy that buformin showed a high similarity with agmatine (TC of 88), but a moderate similarity against guanidinoacetic acid (TC of 50), an intermediate of creatine biosynthesis. On the other hand, metformin showed moderate similarity only with agmatine (TC of 56), and phenformin did not even reach or surpass a TC of 50 with any candidate metabolite. Related to TC, Shape Tanimoto (ST) and Color Tanimoto (CT) were generated from 3D structural comparisons. They ranged from 0 to 100 depending on the grade of similarity in terms of shape or features (H-bonding donors or acceptors, rings, etc.), respectively. When ST and CT were summed up, combo T was generated, and it describes both the shape and features in only one parameter going from 0 to 200. Similar to the 2D structural comparisons, at the 3D level, the results suggested a moderate similarity between metformin and buformin with a combo T of 122, but phenformin still showed a low level of similarity with metformin (combo T of 89). Additionally, when phenformin was compared with buformin at the 3D level, its similarity was moderate (combo T of 111). Compared to the candidate metabolites, metformin showed moderate similarities against L-arginine, cadaverine, creatine, agmatine, and L-ornithine with combo T values going from 101 to 129. It is noteworthy that metformin showed higher similarities compared to those of guanidinoacetic acid and L-aspartic acid (combo T > 130). In the same way, buformin showed moderate similarities against L-arginine, agmatine, guanidinoacetic acid, creatine, and spermidine, with combo T values ranging from 102 to 120. Only phenformin showed lower similarities (combo T < 100) compared to those of all the candidate metabolites. Remarkably, the moderate-to-high similarity observed for metformin or buformin compared to those of some candidate metabolites was due principally to their high level of shape resemblance with ST values > 80. On the other hand, the CT values tended to be less than 50 because of the high diversity of functional features of tested metabolites. However, the presence of the common guanidine moiety in L-arginine, agmatine, creatine, and guanidinoacetic acid led to the highest CT values observed for biguanides. Finally, we performed another structural similarity assay employing the SwissSimilarity tool, aiming to reproduce the previous observations by seeking if there were a common scaffold between the biguanides and some candidate metabolites. As we anticipated in our hypothesis, the results showed that L-arginine, agmatine, creatine, spermidine, and cadaverine shared a common scaffold with metformin and buformin, but not phenformin (data not shown). Considering only structural criteria, these results suggested moderate-to-high structural similarities between the biguanides, especially metformin and buformin, and some candidate metabolites, such as agmatine, intermediates from UC and, especially, creatine biosynthesis. ## 2.3. Physicochemical Comparison between Biguanides and L-arginine-Related Metabolites The physicochemical comparisons between biguanides and the candidate metabolites began with the prediction of their physicochemical properties, employing the collected SMILES through the SwissADME tool [31]. The predictions were performed in ionized (pH 7.4) and non-ionized modalities. For each molecule, we collected seven physicochemical parameters, including molecular weight, consensus log P, topological polar surface area (TPSA), rotatable bonds, the number of hydrogen bond donors and acceptors, and the fraction of sp3 carbons. After the results were collected for each modality, the analysis of those parameters aiming to obtain the similarity under physicochemical criteria was performed. To identify physicochemical relationships between biguanides and the candidate metabolites, an unsupervised machine learning approach with the collected results was employed. First, we performed a principal component analysis (PCA), employing our variables as an exploratory analysis of the data. The PCA of ionized modality showed a close relationship of metformin and buformin with creatine, guanidinoacetate, and L-aspartate. In terms of the physicochemical properties, phenformin was isolated in the PCA. When it was performed in the non-ionized modality, the PCA showed that all the biguanides were similar to creatine, agmatine and L-ornithine in terms of physicochemical properties. After that, we proceeded to perform hierarchical clustering (HC) using the different physicochemical profiles. First, preliminary hierarchical clustering including only candidate metabolites was performed. After that, each biguanide was compared, one at a time, against the different candidate metabolites in HC. Without including any of the biguanides, the preliminary hierarchical clustering at both the ionized and non-ionized modalities generated a cluster for L-arginine, L-citrulline, ADMA, and SDMA, metabolites that belong to the urea cycle and NO metabolism. Polyamines were grouped in two clusters: one for low-molecular-weight polyamines (putrescine and cadaverine) and high-molecular-weight polyamines (spermidine and spermine). In the case of ionized modality, L-ornithine was grouped together with spermidine and spermine. With some differences, creatine, agmatine, guanidinoacetic acid, and L-aspartic acid were grouped together at both of the modalities, accounting for metabolites from the urea cycle, polyamines metabolism, and, mainly, creatine metabolism. L-Argininosuccinic acid alone and fumaric acid with urea (and carbamoyl phosphate in the ionized modality) are represented in the last two clusters. When metformin was included in the HC, this was grouped together with creatine, guanidinoacetic acid, and L-aspartic acid in both modalities, although in the non-ionized modality, L-ornithine and agmatine were also included (Figure 2). In the case of buformin, this was grouped together with agmatine, creatine, guanidinoacetic acid, and L-aspartic acid, although in the non-ionized modality, L-ornithine and carbamoyl phosphate were also included. Finally, phenformin was grouped with fumaric acid and urea, but carbamoyl phosphate was included in the ionized modality. As can be observed, the physicochemical comparison agreed with the previous structural relationships obtained for metformin and buformin against creatine, guanidinoacetate, agmatine, and L-aspartate. Although phenformin showed a physicochemical relationship with agmatine, creatine, and L-ornithine during the PCA in the non-ionized modality, physiologically, non-ionized relationships are of little relevance because all the candidate metabolites and biguanides are ionized inside the cell. This is observed when the PCA was performed in the ionized modality, where phenformin was alone. PCA analyses and all the hierarchical clustering are shown in the Supplementary Material. By limiting our analysis to both structural and physicochemical criteria, these results suggested a relationship between metformin and buformin, but not phenformin, with agmatine, L-aspartate, and especially, intermediates from creatine metabolism. ## 2.4. Structure-Based Target Prediction of Biguanides and L-arginine-Related Metabolites Once we observed structural and physicochemical similarities of biguanides with some arginine-related metabolites, we were interested in testing if this similarity could be translated into an affinity for targets whose main substrates or ligands belong to the candidate metabolites included in the database that we tested before. To prove this assumption, we performed a structure-based target prediction employing the SwissTargetPrediction tool [32]. For each predicted target, its probability score, and its known actives 3D/2D were collected. We observed that biguanides, especially phenformin, were predicted to target the three different isoforms of nitric oxide synthase, whose main substrate is L-arginine, and its main inhibitor is ADMA. However, only the neural isoform of NOS was predicted with a high probability score, although the known actives 3D/2D parameter suggested all biguanides could target the inducible, endothelial, and neural isoforms. Furthermore, the known actives 3D/2D suggested that biguanides could also bind to several isoforms of carbonic anhydrases, whose known ligands are polyamines. Indeed, these observations agree with our previous results of structural and physicochemical similarities that showed moderate-to-high similarities of biguanides to intermediates from polyamines metabolism, NO production, and UC. The complete predicted targets for biguanides can be found in the Supplementary Material. ## 2.5. Affinity Comparison between Biguanides and L-arginine-Related Metabolites Following the target prediction in the SwissTargetPrediction tool, we proceeded to perform a molecular docking simulation to test if the affinities were comparable to those of candidate metabolites in their corresponding targets. The protein files required for the simulation were obtained from Protein Data Bank [33]. All the sites employed in the simulation were based on information from UniProt and predictions from the DoGSiteScorer tool [34,35]. The results from the molecular docking simulations are shown in Figure 3. In general, there was a trend for metformin, buformin and phenformin, in that order, of being the biguanides with the highest affinities to each target. Additionally, according to binding energies and predicted inhibition constants, phenformin showed even more affinity than the endogenous ligands did during their binding to some targets. For instance, the binding energies for metformin (−8.09 kcal/mol), buformin (−9.39 kcal/mol), and phenformin (−9.72 kcal/mol) in ARG1 were comparable to that obtained for L-arginine (−9.56 kcal/mol). Another example was CASTOR1, one of the L-arginine sensors in the cell, in which the binding energy for L-arginine was −9.3 kcal/mol, but in the case of metformin, buformin, and phenformin, they were −7.21 kcal/mol, −7.92 kcal/mol, and −10.34 kcal/mol, respectively. In the case of GAMT, biguanides showed comparable or even higher affinities (−6.9 kcal/mol to −9.48 kcal/mol) than guanidinoacetate did (−5.59 kcal/mol), the precursor of creatine. Furthermore, the binding energies of biguanides (from −7.76 kcal/mol to −9.9 kcal/mol) were similar to that of putrescine (−8.44 kcal/mol) in spermidine synthase. Finally, for the three isoforms of NOS, the estimated binding energies of biguanides ranged from −4.37 kcal/mol to −6.56 kcal/mol, while for L-arginine and ADMA, they ranged from −5.01 kcal/mol to −5.51 kcal/mol and from −5.07 kcal/mol to −5.42 kcal/mol, respectively. Remarkably, these results agreed with phenformin and metformin being the most and least potent biguanides, respectively [36]. ## 2.6. Binding Comparison between Biguanides and L-arginine-Related Metabolites Despite the similarity between the predicted binding energies of biguanides and the candidate metabolites, we proceeded to assess the binding modes inside the tested targets to sustain our hypothesis about a structure-dependent cross mechanism of action. The different binding mode analyses were carried out in Chimera X 1.2.5. It is noteworthy that some biguanides’ binding modes were similar to those of candidate metabolites in their corresponding targets as shown in Figure 4 for ARG1, CASTOR1, and brain-type CK. Other interactions can be found in the Supplementary Material. We observed the establishment of hydrogen bonds between L-arginine and Asp 128, Asn 130, Ser 137, Thr 246, and Asp 232 in ARG1. Our simulation for the L-arginine binding mode in ARG1 agreed with the binding site reported in UniProt, which includes His 126, Thr 127, Asp 128, Ile 129, Asn 130, Ser 137, Gly 138, Asn 139, Asp 183, Thr 246, and Glu 277, indicating the good reproducibility of our methodology [34]. Although with some differences, biguanides established hydrogen bonds with amino acid residues present in the L-arginine binding site, including Asp 128, Thr 246, and Glu 277, among others. As well as for ARG1, in CASTOR1, our simulation for L-arginine showed good reproducibility compared to that of the reported interacting amino acid residues from UniProt [34]. In our methodology, L-arginine established hydrogen bonds with Val 112, Gly 274, Val 281, Thr 300, Phe 301, and Asp 304. With some differences, the biguanides shared interactions with different L-arginine-interacting amino acid residues, including Gly 274, Thr 300, Phe 301, and Asp 304. Finally, for the brain-type CK, the reported interacting amino acid residues in UniProt include Val 72, Glu 232, and Se 285 [34]. Our simulation for creatine predicted the binding of the guanidine moiety to Glu 231 and Glu 232 and the carboxylate moiety to Arg 132 and Arg 292. Such as creatine, the guanidine moiety of biguanides showed a tendency to establish hydrogen bonds with Glu 231 and Glu 232, except for buformin, which established two hydrogen bonds with Arg 236 instead of Glu 232. Remarkably, the orientations of the guanidine groups inside ARG1, CASTOR1, and brain-type CK for L-arginine and biguanides were strongly similar. As can be observed, these results are in line with our previous observations, where it should be expected to observe a good concordance of binding if the molecules are supposed to be similar. ## 3. Discussion Beyond its anti-diabetic properties, metformin has showed to exert anti-cancer effects in different types of cancer models and epidemiological studies [23]. Despite its clinical success and the progress made to elucidate its anticancer molecular actions, a complete picture of metformin’s anticancer biological effects is lacking. However, based only on structural and physicochemical characteristics, here, we proposed and sustained, with a computational methodology, our hypothesis that described another possible mechanism by which biguanides may resemble different endogenous L-arginine-related metabolites, leading to an antagonist effect in their corresponding targets. In pharmacology, the binding of ligands to a specific target is dependent on several factors such as shape, charges, hydrogen-bonding capacity, flexibility, planarity, polarity, and size, among others [37]. According to our hypothesis, if biguanides are supposed to bind targets from arginine-related metabolites, they must be comparable at both the structural and physicochemical levels. For this reason, the methodology followed here began with the creation of a database that contains different metabolites related to L-arginine metabolism and other related metabolic pathways including the urea cycle, nitric oxide, polyamines, and creatine metabolism. Subsequently, comparisons of the structural and physicochemical elements by employing cheminformatic tools were performed. In the final step to test our hypothesis, bioinformatic tools were carried out to prove if the comparable structural and physicochemical elements between biguanides and arginine-related metabolites could be translated into a comparable binding in their corresponding targets. If this was true, it is reasonable to think that biguanides may exert an anticancer effect by affecting the metabolic pathways where these metabolites are involved. In the following paragraphs, evidence that sustains or refutes our hypothesis is provided. However, the evidence is focused on metformin mainly because it is the most widely studied biguanide. Anabolism supports proliferation, survival, and invasion processes in cancer cells. mTORC1 is one of the most important master regulators of anabolic metabolism in normal and cancer cells [38]. The upregulation of mTORC1 has been reported to be dependent on two interrelated stimuli: proliferative signaling pathways by growth factors and nutrient sensing pathways by L-arginine, L-glutamine, and branched-chain amino acids [4]. In cancer cells, anabolism dependency on L-arginine via mTORC1 has been associated with the activation of the RAGULATOR-RAG complex in the lysosomal membrane by both SLC38A9 and CASTOR1, the two intracellular sensors of L-arginine. SLC38A9 is a transmembrane transporter of L-leucine and L-arginine located on the lysosome that activates the RAGULATOR-RAG complex in response to changes in amino acids concentrations [39]. The sensing of an amino acid via SLC38A9 has been reported to be needed in pancreatic cancer cells to form tumors [40]. On the other hand, the binding of L-arginine to CASTOR1 disrupts its suppressing interaction with GATOR2, leading to the GATOR2-dependent inhibition of GATOR1 and the activation of the RAGULATOR-RAG complex [41]. The CASTOR1-dependent inhibition of mTORC1 has shown a tumor suppressor role in lung adenocarcinoma leading to lower proliferation, migration, and invasion, and when it is downregulated, it is associated with a poor prognosis [42]. It has been reported that the inhibition of mTORC1 through competitive binding in CASTOR1 by analogs of L-arginine, including L-citrulline and L-ornithine, among others, avoids the disruption of CASTOR1–GATOR2 interaction [43]. Our hypothesis predicted that it may be possible the suppression of mTORC1 by the competitive binding of biguanides against L-arginine in CASTOR1 and SLC38A9, leading to decreased activity of the RAGULATOR-RAG complex and the consequent modulation of mTORC1. According to our results, metformin and buformin showed moderate structural similarities with L-arginine and L-ornithine, two reported ligands of CASTOR1. In addition, metformin and buformin are similar to L-ornithine in terms of the physicochemical properties. Additionally, the affinities and binding modes of all the biguanides were comparable or even higher than those obtained for L-arginine in CASTOR1 and SLC38A9. In agreement with our hypothesis and results, the inhibition of mTORC1 by a metformin treatment has been reported, independently of AMPK and TSC$\frac{1}{2}$, and in a RAG GTPase-dependent manner [44]. Recently, it has been reported that L-arginine exerts an epigenetic regulation over TEA-like domain 4 (TEAD4) in prostate cancer cells [45]. TEAD4 is a transcription factor that controls the expression of genes involved in oxidative phosphorylation. This suggests that global metabolism and epigenetics can be controlled by L-arginine in cancer cells via different sensing systems, such as CASTOR1, SLC38A9, and possibly TEAD4. As our hypothesis predicted, there was reported the metformin targeting of the YAP1–TEAD4 axis in bladder cancer cells [46]. The urea cycle is related to several metabolic pathways that allow anabolism. For instance, the urea cycle was reported to be linked to the TCA via fumarate and oxalacetate-derived L-aspartate [47]. Given this connection, the urea cycle disruptions were reported to modify TCA metabolites, leading to the metabolic and epigenetic changes needed for cancer cells. It is noteworthy that some essential building blocks and epimetabolites are directly derived from TCA, such as α-ketoglutarate [48]. For instance, it has been reported that the fumarate accumulation in cancer cells allowed epithelial-to-mesenchymal transition due to the inhibition of the α-ketoglutarate-dependent dioxygenases, proteins with histone demethylase activity [49]. Our results may indicate a possible reduction of urea cycle metabolites after the metformin treatment due to a cross structure-dependent inhibition of key enzymes. According to our analyses, metformin and buformin showed moderate-to-high similarities when they were compared at both the structural and physicochemical levels with some urea cycle metabolites, including L-aspartic acid, L-arginine, and L-ornithine. As expected, the structural and physicochemical similarities were translated in comparable affinities and binding modes of biguanides with urea cycle enzymes, at least in ARG1 and ARG2. Similar to this study, Detroja and Samson [2022] reported the possible inhibition of ARG1 by metformin based on molecular docking simulations and molecular dynamics. They found that the binding of metformin to the active sites of both ARG1 and ARG2 was stable for up to 50 ns in molecular dynamics simulations [50]. There is evidence of biguanides affecting the urea cycle. According to Zhang et al. [ 2021], metformin has been shown to negatively regulate the urea cycle intermediates including L-arginine and L-aspartate in an in vivo xenograft model of HCT116, a colorectal cancer cell line [51]. However, the authors reported that this reduction was due to the decreased expression in urea cycle enzymes such as CPS1, ARG1, and OTC. Additionally, it has been reported that the metformin treatment decreased the ARG1 activity of granulocytic myeloid-derived suppressor cells in a tumor-bearing mouse model of colon carcinoma [52]. Furthermore, the decreased arginase activity was maintained despite the use of compound C, an AMPK inhibitor, suggesting that the reduction of arginase activity was independent on AMPK. In the same way, Koroglu-Aydin et al. [ 2021] reported a decreased activity of arginase in kidney homogenates from rats with diabetes and prostate cancer treated with metformin compared to that of non-treated rats [53]. When they are altered, polyamines in cancer have been associated with increased proliferation, survival, and resistance to therapy, i.e., a poor cancer prognostic. In the case of polyamine metabolism, our results strongly supported the resemblance between biguanides and some specific intermediates. There was a remarkable structural similarity between buformin and metformin with agmatine and L-ornithine. These relationships were maintained during the physicochemical comparisons, although not in both modalities. As well as for the UC, metformin was capable of decreasing the concentrations of putrescine via decreased expression of ODC in an in vivo xenograft model of HCT116 [51]. Cell division is a high-energy-demanding process that needs to be supported by plenty of energy. In cancer cells, a reported strategy to maintain the cell cycle forward was the overexpression of creatine kinases isoforms, especially in those sites where a lot of energy is needed [54]. For these reasons, the phosphagen system represents one of the most important ways of storing, buffering, and transferring energy in tumors. Regarding creatine metabolism, our data indicated a competitive binding of biguanides against targets related to creatine biosynthesis or usage and the consequent disruption of cell bioenergetics. Our results showed moderate-to-high similarity at both the 2D and, especially, 3D levels. Additionally, the hierarchical clustering of biguanides with guanidinoacetic acid and creatine suggested a close relationship in terms of the physicochemical properties. On the other hand, the affinities and, for some isoforms of creatine kinases, the binding modes of biguanides were comparable to those obtained for creatine. It is noteworthy that metabolites belonging to creatine metabolism were the ones most related to biguanides, especially metformin and buformin, according to our results during structural, physicochemical, and binding comparisons. Recently, has authors have reported the competitive inhibition of AGAT by a metformin treatment (3 × 1500 mg daily for 6 weeks) in individuals with Becker muscular dystrophy, leading to reduced concentrations of guanidinoacetic acid in both serum and urine [55]. Remarkably, decreased renal and pancreatic AGAT activity, but not decreased mRNA expression, have been reported after creatine supplementation in rats, suggesting that creatine can be involved in a negative feedback regulation mechanism [56]. If biguanides were comparable to creatine, it may be possible that biguanides could become involved in that regulation, resembling creatine. Regarding the evidence supporting the action of biguanides on phosphocreatine biosynthesis, Garbati et al. [ 2017] reported the decrement of $40\%$ in brain-type creatine kinase activity by metformin treatment (10 mM) during enzymatic assays, and its lowering effect on the ATP/AMP ratio of the KM-H2, SHSY-5Y, and MDA-MB-468 cancer cell lines [57]. It is noteworthy that metformin was suggested to bind to a site different from that for creatine binding, acting in a non-competitive manner. It is possible that metformin may bind to the other isoforms of creatine kinases given the strong homology of this family of proteins [58]. It is noteworthy that our main limitation was that our results were generated through a computational methodology. Hence, the experimental demonstration is needed to sustain or refute the results presented here and to test our hypothesis. Another limitation of our computational methodology was the absence of molecular dynamics simulations to test the binding of biguanides and L-arginine-related metabolites to the targets of interest in a more reliable manner, since both the protein and the ligand are flexible [59]. However, the use of molecular dynamics simulations was beyond the scope of our original methodology, as our principal aim was to establish our hypothesis and to provide preliminary computational evidence about the possible structural, physicochemical, and binding similarities that biguanides share with some L-arginine-related metabolites, as a first approach. In the future, molecular dynamics simulations are intended to be included with an experimental methodology to provide stronger evidence for or against our hypothesis. Additionally, as mentioned above, authors have already reported the molecular dynamics simulations of metformin’s binding to ARG$\frac{1}{2}$, partially strengthening our molecular docking simulations results for metformin in the same targets [50]. Another opportunity for improvement in the future is that in the present study, we limited our computational analysis to certain L-arginine-related metabolites, but there are many metabolites that may also be related structurally and functionally to biguanides, such as N(G)-monomethyl-L-arginine, L-proline, L-glutamate, or L-homoarginine, as well as different biguanides, including galegine or even microbiome-derived metabolites of biguanides. ## 4.1. Database Creation The database was created including different candidate metabolites related to L-arginine metabolism in the cell, including compounds from the urea cycle, creatine biosynthesis, nitric oxide metabolism, and polyamines metabolism, among others. Given that buformin and phenformin are more potent drugs than metformin is, but belong to the same pharmacological class, we decided to include both drugs to enrich the analyses. For every molecule, its CID number and SMILES from PubChem were collected. ## 4.2. Structural Comparison between Biguanides and L-arginine-Related Metabolites Assessment of the 2D and 3D similarities between biguanides and candidate metabolites were carried out employing the PubChem Score Matrix Service [60]. In the case of the 3D similarity assay, both shape optimized and feature optimized analyses were carried out for 10 conformers per CID. Two-dimensional similarity assays generated a TC value that ranged from 0 to 100 depending on the grade of similarity, with 0 being a null similarity and 100 being an identical molecule. Three-dimensional similarity assays generated two scores: ST assessed the shape and CT assessed the features (e.g., hydrogen bond donors and acceptors, rings, etc.) of the compared molecules. Both scores ranged from 0 to 100 following the same logic as that used for TC. When these two scores were summed up, combo T was generated [61]. This score ranged from 0 to 200 and was used to assess shape and features together in only one parameter. Given that the TC cut offs ranged from 50 to 80 in the literature, for 2D similarity, the scores were “high” when TC ≥ 65. In the case of 3D similarity, it was “high” when ST ≥ 80 and CT ≥ 50, according to Bolton et al. [ 2011], or when combo T ≥ 130 [61]. For every pair of compared molecules, their TCs were collected from the 2D analysis, as well as their ST and CT from the 3D analysis. Additionally, the identification of a common scaffold between biguanides and the structurally related metabolites was performed in the SwissSimilarity tool (http://www.swisssimilarity.ch/ (accessed on 10 January 2023)). This tool is capable of performing high-throughput screening of molecules similar to an input molecule based on 2D and 3D molecular descriptors. The search can be performed in different classes of compounds, e.g., drugs and commercial or bioactive substances. For this methodology, we introduced the different SMILES in the menu and selected “Bioactive” as the class of compounds. Finally, we chose the Chemical Entities of Biological Interest (ChEBI) as the library and “Scaffold” as the screening method [62]. ## 4.3. Physicochemical Comparison between Biguanides and L-arginine-Related Metabolites The physicochemical properties of biguanides and candidate metabolites presented in the database were predicted by using the collected SMILES of every molecule with the SwissADME tool (http://www.swissadme.ch/ (accessed on 21 December 2022)) [31]. The physicochemical comparisons were performed in both ionized and non-ionized modalities. For the ionized modality, all the molecules in our database were ionized to physiological pH of 7.4 according to scientific reports and their corresponding new SMILES were collected. For every molecule, several physicochemical parameters were collected, including molecular weight, accounting for size, hydrogen bond donors and hydrogen bond acceptors, accounting for hydrogen-bonding capacity, consensus log P and TPSA, accounting for polarity, rotatable bonds, accounting for flexibility, and sp3-carbon fraction, accounting for planarity. After this, a PCA of the collected variables was performed as an exploratory analysis of data from physicochemical profiles. Finally, an HC was carried out aiming to test if an unsupervised machine learning algorithm was capable of identifying a relationship between biguanides and candidate metabolites. Each biguanide was compared, one at a time, against the candidate metabolites. PCA and HC were performed in the R software (version 4.2.2) by employing RStudio as our graphical user interface and some R-packages including tidyverse, factoextra, cluster, ggplot2, ggcorrplot, and readr [63,64,65,66,67,68,69]. Data analysis began with the loading of our database in RStudio. This database contained biguanides and our candidate metabolites and all their collected physicochemical parameters. However, these data are presented at different scales. For this reason, these magnitudes were scaled into the standard scale of the Z-score. PCA was performed by using the prcomp function based on singular value decomposition. Indeed, according to R documentation, a better numerical accuracy was reported for this method compared with that of the eigen decomposition. Additionally, it is not necessary to generate the covariance matrix in this method. On the other hand, the clusters of the HC were generated by calculating the Euclidean distances and employing the Ward’s linkage with the hclust function. All dendrograms were generated in RStudio using the packages mentioned above. ## 4.4. Structure-Based Target Prediction of Biguanides and L-arginine-Related Metabolites The biguanides were subjected to a structure-based target prediction by using the collected SMILES employing the SwissTargetPrediction Tool (www.swisstargetprediction.ch/ (accessed on 29 December 2022)) from the Swiss Institute of Bioinformatics [32]. In a similar manner to our hypothesis, the SwissTargetPrediction tool is based on the chemical similarity, i.e., the algorithm was trained with a big collection of molecules (almost 400,000), whose targets were identified experimentally. In simple terms, when a new ligand is presented to the algorithm, it searches for similar molecules at the 2D and the 3D levels, returning their associated macromolecules as possible targets for the new ligand. The output of a SwissTargetPrediction assay are two parameters: the probability score and the known actives (3D/2D). The probability score represents a combined score of the 2D and 3D similarity values between the input ligand and those molecules from the algorithm. The known actives (3D/2D) represent the list of molecules similar at the 2D and the 3D levels to those of the input ligand, whose interaction has been demonstrated experimentally. In this methodology, for every molecule, its predicted biological targets associated with candidate metabolites and their corresponding probability scores and known actives (3D/2D) were collected. After this, we confirmed if the interactions between the predicted targets and the tested ligands have been reported in scientific literature. ## 4.5. Affinity Comparison between Biguanides and L-arginine-Related Metabolites Once we predicted the targets for biguanides and candidate metabolites, a molecular docking simulation was carried out employing AutoDock 4.2 [70]. Other non-predicted targets whose ligand or substrate was reported to be one of the candidate metabolites included in our database were subjected to simulation. All the protein structures for the simulation were obtained from Protein Data Bank [33]. Our search was limited to PDB files with a resolution lower than 3Å and derived from Homo sapiens. In the case of SLC38A9, we used a PDB file from Danio rerio because the human files showed lower resolution. The water molecules and other ligands co-crystallized with the protein were removed from the PDB files. The tested candidate metabolites were constructed in Avogadro based on the different collected SMILES [71]. Additionally, hydrogen atoms were added to the ligands, simulating a pH of 7.4. Finally, the structures of ligands were optimized in Avogadro using both the force field MMFF94 and the algorithm steepest descent, and they were saved as.mol2 files. AutoDock Tools (ADT) was employed as the graphical user interface to perform the docking simulations. In ADT, polar-only hydrogens and Kollman charges were added to proteins, while ligands were subjected to additions of polar-only hydrogens and Gasteiger charges. The docking sites were chosen based on information from the UniProt database and based on the predictions obtained from the DoGSiteScorer tool [34,35]. The complete 3D coordinates used for docking simulations are shown in the Supplementary Material. During the simulations, the ligand was flexible, but the protein was maintained in a rigid modality. The affinity maps of simulations were computed using a grid spacing of 0.375Å. Additionally, we used a Lamarckian genetic algorithm for simulations employing a population of 150, with a rate of mutation of 0.2 and a maximum number of generations of 27,000. For every simulation, its predicted binding energy and inhibition constant were collected. Our molecular docking methodology was validated through redocking. The validation methodology used here is shown in the Supplementary Material. ## 4.6. Binding Comparison between Biguanides and L-arginine-Related Metabolites In order to analyze the different binding modes obtained from molecular docking simulations, Chimera X 1.2.5 was used to study the interactions, orientation, and conformations of biguanides and candidate metabolites in the tested targets [72]. For this analysis, we included the conformations that showed the most negative binding energies from the previous step. Additionally, we calculated the distances and found the established hydrogen bonds between the different ligands and the tested targets using the H-bonds tool from the Structure Analysis menu. According to the documentation of Chimera X 1.2.5, the H-bonds tool identifies the possible hydrogen bonds based on the atom types present in the macromolecule and the ligand and geometric criteria. The settings were established as follows: the radius at 0.075 Å, the tolerance distance at 0.4 Å, and the angle tolerance at 20°. Established interactions and conformations were compared with data from UniProt and scientific reports. All the figures shown here were generated using Chimera X 1.2.5. ## 5. Conclusions Nowadays, therapies targeting metabolic reprogramming are gaining relevance due to the growing field of metabolomics applied to cancer research and the new findings about metabolic vulnerabilities in some specific cancer types. Metformin has been reported to alter L-arginine metabolism and other related metabolic pathways in different biological models, including humans, such as ADT, suggesting a potential use for cancer treatment in combination with chemotherapy. However, the molecular actions by which metformin performed such biological effects are incomplete. Here, we demonstrated a possible relationship between biguanides, especially metformin and buformin, with some L-arginine-related metabolites, particularly those from creatine metabolism, using a computational methodology based on cheminformatic and bioinformatic tools. The results obtained here, and the evidence discussed may suggest a new possible mechanism of action in which biguanides may resemble L-arginine and its related metabolites, leading to the modulation of their corresponding targets. In the future, this structure-dependent cross mechanism of action must be confirmed with experimental evidence. Additionally, the elucidation of the complete metformin’s mechanism of action can contribute to establishing better therapy interventions through the rational design of chemotherapy combinations and the repurposing of metformin for other diseases. ## References 1. Weinberg R.A., Hanahan D.. Hallmarks of cancer: The next generation. Cell (2011.0) 144 646-674. DOI: 10.1016/j.cell.2011.02.013 2. Chen C.L., Hsu S.C., Ann D.K., Yen Y., Kung H.J.. Arginine Signaling and Cancer Metabolism. Cancers (2021.0) 13. DOI: 10.3390/cancers13143541 3. Morris S.M.. Arginine Metabolism Revisited. J. Nutr. (2016.0) 146 2579S-2586S. DOI: 10.3945/jn.115.226621 4. Saxton R.A., Sabatini D.M.. mTOR Signaling in Growth, Metabolism, and Disease. Cell (2017.0) 168 960-976. DOI: 10.1016/j.cell.2017.02.004 5. Bonilla D.A., Kreider R.B., Stout J.R., Forero D.A., Kerksick C.M., Roberts M.D., Rawson E.S.. Metabolic Basis of Creatine in Health and Disease: A Bioinformatics-Assisted Review. Nutrients (2021.0) 13. DOI: 10.3390/nu13041238 6. Werner A., Pieh D., Echchannaoui H., Rupp J., Rajalingam K., Theobald M., Closs E.I., Munder M.. Cationic Amino Acid Transporter-1-Mediated Arginine Uptake Is Essential for Chronic Lymphocytic Leukemia Cell Proliferation and Viability. Front. Oncol. (2019.0) 9 1268. DOI: 10.3389/fonc.2019.01268 7. Cheng C.T., Qi Y., Wang Y.C., Chi K.K., Chung Y., Ouyang C., Chen Y.R., Oh M.E., Sheng X., Tang Y.. Arginine starvation kills tumor cells through aspartate exhaustion and mitochondrial dysfunction. Commun. Biol. (2018.0) 1 178. DOI: 10.1038/s42003-018-0178-4 8. Lee J.S., Adler L., Karathia H., Carmel N., Rabinovich S., Auslander N., Keshet R., Stettner N., Silberman A., Agemy L.. Urea Cycle Dysregulation Generates Clinically Relevant Genomic and Biochemical Signatures. Cell (2018.0) 174 1559-1570. DOI: 10.1016/j.cell.2018.07.019 9. Casero R.A., Murray-Stewart T., Pegg A.E.. Polyamine metabolism and cancer: Treatments, challenges and opportunities. Nat. Rev. Cancer (2018.0) 18 681-695. DOI: 10.1038/s41568-018-0050-3 10. Roci I., Watrous J.D., Lagerborg K.A., Lafranchi L., Lindqvist A., Jain M., Nilsson R.. Mapping Metabolic Events in the Cancer Cell Cycle Reveals Arginine Catabolism in the Committed SG2M Phase. Cell Rep. (2019.0) 26 1691-1700. DOI: 10.1016/j.celrep.2019.01.059 11. Elitsur Y., Moshier J.A., Murthy R., Barbish A., Luk G.D.. Polyamine levels, ornithine decarboxylase (ODC) activity, and ODC-mRNA expression in normal and cancerous human colonocytes. Life Sci. (1992.0) 50 1417-1424. DOI: 10.1016/0024-3205(92)90260-V 12. Bello-Fernandez C., Packham G., Cleveland J.L.. The ornithine decarboxylase gene is a transcriptional target of c-Myc. Proc. Natl. Acad. Sci. USA (1993.0) 90 7804-7808. DOI: 10.1073/pnas.90.16.7804 13. Zhang Y., Li H., Wang X., Gao X., Liu X.. Regulation of T Cell Development and Activation by Creatine Kinase B. PLoS ONE (2009.0) 4. DOI: 10.1371/journal.pone.0005000 14. Fenouille N., Bassil C.F., Ben-Sahra I., Benajiba L., Alexe G., Ramos A., Pikman Y., Conway A.S., Burgess M.R., Li Q.. The creatine kinase pathway is a metabolic vulnerability in EVI1-positive acute myeloid leukemia. Nat. Med. (2017.0) 23 301-313. DOI: 10.1038/nm.4283 15. Li X.H., Chen X.J., Ou W.B., Zhang Q., Lv Z.R., Zhan Y., Ma L., Huang T., Yan Y.B., Zhou H.M.. Knockdown of creatine kinase B inhibits ovarian cancer progression by decreasing glycolysis. Int. J. Biochem. Cell Biol. (2013.0) 45 979-986. DOI: 10.1016/j.biocel.2013.02.003 16. Jiang Y., Su Z., Lin Y., Xiong Y., Li C., Li J., Wang R., Zhong R., Cheng B., He J.. Prognostic and predictive impact of creatine kinase level in non-small cell lung cancer treated with tyrosine kinase inhibitors. Transl. Lung Cancer Res. (2021.0) 10 3771-3781. DOI: 10.21037/tlcr-21-600 17. Keshet R., Erez A.. Arginine and the metabolic regulation of nitric oxide synthesis in cancer. Dis. Model Mech. (2018.0) 11 dmm033332. DOI: 10.1242/dmm.033332 18. Reddy K.R.K., Dasari C., Duscharla D., Supriya B., Ram N.S., Surekha M.V., Kumar J.M., Ummanni R.. Dimethylarginine dimethylaminohydrolase-1 (DDAH1) is frequently upregulated in prostate cancer, and its overexpression conveys tumor growth and angiogenesis by metabolizing asymmetric dimethylarginine (ADMA). Angiogenesis (2018.0) 21 79-94. DOI: 10.1007/s10456-017-9587-0 19. Alexandrou C., Al-Aqbi S.S., Higgins J.A., Boyle W., Karmokar A., Andreadi C., Luo J.L., Moore D.A., Viskaduraki M., Blades M.. Sensitivity of Colorectal Cancer to Arginine Deprivation Therapy is Shaped by Differential Expression of Urea Cycle Enzymes. Sci. Rep. (2018.0) 8 12096. DOI: 10.1038/s41598-018-30591-7 20. Hall P.E., Ready N., Johnston A., Bomalaski J.S., Venhaus R.R., Sheaff M., Krug L., Szlosarek P.W.. Phase II Study of Arginine Deprivation Therapy with Pegargiminase in Patients with Relapsed Sensitive or Refractory Small-cell Lung Cancer. Clin. Lung Cancer (2020.0) 21 527-533. DOI: 10.1016/j.cllc.2020.07.012 21. Istvan E.S., Deisenhofer J.. Structural Mechanism for Statin Inhibition of HMG-CoA Reductase. Science (2001.0) 292 1160-1164. DOI: 10.1126/science.1059344 22. Kim H.W.. Metabolomic Approaches to Investigate the Effect of Metformin: An Overview. Int. J. Mol. Sci. (2021.0) 22. DOI: 10.3390/ijms221910275 23. Aljofan M., Riethmacher D.. Anticancer activity of metformin: A systematic review of the literature. Future Sci. OA (2019.0) 5 FSO410. DOI: 10.2144/fsoa-2019-0053 24. Jeon S.M.. Regulation and function of AMPK in physiology and diseases. Exp. Mol. Med. (2016.0) 48 e245. DOI: 10.1038/emm.2016.81 25. Hu Y., Bajorath J.. Compound promiscuity: What can we learn from current data?. Drug Discov. Today (2013.0) 18 644-650. DOI: 10.1016/j.drudis.2013.03.002 26. Malathi K., Ramaiah S.. Bioinformatics approaches for new drug discovery: A review. Biotechnol. Genet. Eng. Rev. (2018.0) 34 243-260. DOI: 10.1080/02648725.2018.1502984 27. Torres P.H.M., Sodero A.C.R., Jofily P., Silva-Jr F.P.. Key topics in molecular docking for drug design. Int. J. Mol. Sci. (2019.0) 20. DOI: 10.3390/ijms20184574 28. Lo Y.C., Rensi S.E., Torng W., Altman R.B.. Machine learning in chemoinformatics and drug discovery. Drug Discov. Today (2018.0) 23 1538-1546. DOI: 10.1016/j.drudis.2018.05.010 29. Hadavand Mirzaei H., Jassbi A.R., Pirhadi S., Firuzi O.. Study of the mechanism of action, molecular docking, and dynamics of anticancer terpenoids from Salvia lachnocalyx. J. Recept. Signal Transduct. Res. (2020.0) 40 24-33. DOI: 10.1080/10799893.2019.1710847 30. Zhang L., Yang K., Wang M., Zeng L., Sun E., Zhang F., Cao Z., Zhang X., Zhang H., Guo Z.. Exploring the mechanism of Cremastra Appendiculata (SUANPANQI) against breast cancer by network pharmacology and molecular docking. Comput. Biol. Chem. (2021.0) 94 107396. DOI: 10.1016/j.compbiolchem.2020.107396 31. Daina A., Michielin O., Zoete V.. SwissADME: A free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. (2017.0) 7 42717. DOI: 10.1038/srep42717 32. Daina A., Michielin O., Zoete V.. SwissTargetPrediction: Updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Res. (2019.0) 47 W357-W364. DOI: 10.1093/nar/gkz382 33. Berman H.M., Westbrook J., Feng Z., Gilliland G., Bhat T.N., Weissig H., Shindyalov I.N., Bourne P.E.. The Protein Data Bank. Nucleic Acids Res. (2000.0) 28 235-242. DOI: 10.1093/nar/28.1.235 34. UniProt: A hub for protein information. Nucleic Acids Res. (2015.0) 43 D204-D212. DOI: 10.1093/nar/gku989 35. Volkamer A., Kuhn D., Rippmann F., Rarey M.. DoGSiteScorer: A web server for automatic binding site prediction, analysis and druggability assessment. Bioinformatics (2012.0) 28 2074-2075. DOI: 10.1093/bioinformatics/bts310 36. Bailey C.J.. Metformin: Historical overview. Diabetologia (2017.0) 60 1566-1576. DOI: 10.1007/s00125-017-4318-z 37. Andersson C.D., Chen B.Y., Linusson A.. Mapping of ligand-binding cavities in proteins. Proteins (2010.0) 78 1408-1422. DOI: 10.1002/prot.22655 38. Zou Z., Tao T., Li H., Zhu X.. mTOR signaling pathway and mTOR inhibitors in cancer: Progress and challenges. Cell Biosci. (2020.0) 10 31. DOI: 10.1186/s13578-020-00396-1 39. Wang S., Tsun Z.Y., Wolfson R.L., Shen K., Wyant G.A., Plovanich M.E., Yuan E.D., Jones T.D., Chantranupong L., Comb W.. Lysosomal amino acid transporter SLC38A9 signals arginine sufficiency to mTORC1. Science (2015.0) 347 188-194. DOI: 10.1126/science.1257132 40. Wyant G.A., Abu-Remaileh M., Wolfson R.L., Chen W.W., Freinkman E., Danai L.V., Vander-Heiden M.G., Sabatini D.M.. mTORC1 Activator SLC38A9 Is Required to Efflux Essential Amino Acids from Lysosomes and Use Protein as a Nutrient. Cell (2017.0) 171 642-654. DOI: 10.1016/j.cell.2017.09.046 41. Chantranupong L., Scaria S.M., Saxton R.A., Gygi M.P., Shen K., Wyant G.A., Wang T., Wade-Harper J., Gygi S.P., Sabatini D.M.. The CASTOR Proteins Are Arginine Sensors for the mTORC1 Pathway. Cell (2016.0) 165 153-164. DOI: 10.1016/j.cell.2016.02.035 42. Zhou X., Cheng Z., Chen H., Shi S., Wang X., Orang M., Zhao J.. CASTOR1 suppresses the progression of lung adenocarcinoma and predicts poor prognosis. J. Cell. Biochem. (2018.0) 119 10186-10194. DOI: 10.1002/jcb.27360 43. Saxton R.A., Chantranupong L., Knockenhauer K.E., Schwartz T.U., Sabatini D.M.. Mechanism of arginine sensing by CASTOR1 upstream of mTORC1. Nature (2016.0) 536 229-233. DOI: 10.1038/nature19079 44. Kalender A., Selvaraj A., Kim S.Y., Gulati P., Brule S., Viollet B., Kemp B.E., Bardeesy N., Dennis P., Schlager J.J.. Metformin, Independent of AMPK, Inhibits mTORC1 in a Rag GTPase-Dependent Manner. Cell Metab. (2010.0) 11 390-401. DOI: 10.1016/j.cmet.2010.03.014 45. Chen C.L., Hsu S.C., Chung T.Y., Chu C.Y., Wang H.J., Hsiao P.W., Yeh S.D., Ann D.K., Yen Y., Kung H.J.. Arginine is an epigenetic regulator targeting TEAD4 to modulate OXPHOS in prostate cancer cells. Nat. Commun. (2021.0) 12 2398. DOI: 10.1038/s41467-021-22652-9 46. Wu Y., Zhen Q., Li Y., Wang G., Gao S., Zhang X., Yan X., Zhang X., Xie J., Wang Y.. Metformin targets a YAP1-TEAD4 complex via AMPKα to regulate CCNE1/2 in bladder cancer cells. J. Exp. Clin. Cancer Res. (2019.0) 38 376. DOI: 10.1186/s13046-019-1346-1 47. Pesi R., Balestri F., Ipata P.L.. Metabolic interaction between urea cycle and citric acid cycle shunt: A guided approach. Biochem. Mol. Biol. Educ. (2018.0) 46 182-185. DOI: 10.1002/bmb.21099 48. Shahmirzadi A.A., Edgar D., Liao C.Y., Hsu Y.M., Lucanic M., Shahmirzadi A.A., Wiley C.D., Gan G., Kim D.E., Kasler H.G.. Alpha-Ketoglutarate, an Endogenous Metabolite, Extends Lifespan and Compresses Morbidity in Aging Mice. Cell Metab. (2020.0) 32 447-456. DOI: 10.1016/j.cmet.2020.08.004 49. Sciacovelli M., Goncalves E., Johnson T.I., Zecchini V.R., Henriques-da Costa A.S., Gaude E., Vercauteren-Drubbel A., Theobald S.J., Riekje-Abbo S., Binh-Tran M.G.. Fumarate is an epigenetic modifier that elicits epithelial-to-mesenchymal transition. Nature (2016.0) 537 544-547. DOI: 10.1038/nature19353 50. Detroja T.S., Samson A.O.. Virtual Screening for FDA-Approved Drugs That Selectively Inhibit Arginase Type 1 and 2. Molecules (2022.0) 27. DOI: 10.3390/molecules27165134 51. Zhang T., Hu L., Tang J.F., Xu H., Tian K., Wu M.N., Huang S.Y., Du Y.M., Zhou P., Lu R.J.. Metformin Inhibits the Urea Cycle and Reduces Putrescine Generation in Colorectal Cancer Cell Lines. Molecules (2021.0) 26. DOI: 10.3390/molecules26071990 52. Xu P., Yin K., Tang X., Tian J., Zhang Y., Ma J., Xu H., Xu Q., Wang S.. Metformin inhibits the function of granulocytic myeloid-derived suppressor cells in tumor-bearing mice. Biomed. Pharmacother. (2019.0) 120 109458. DOI: 10.1016/j.biopha.2019.109458 53. Koroglu-Aydın P., Bayrak B.B., Bugan I., Karabulut-Bulan O., Yanardag R.. Histological and biochemical investigation of the renoprotective effects of metformin in diabetic and prostate cancer model. Toxicol. Mech. Methods (2021.0) 31 489-500. DOI: 10.1080/15376516.2021.1919810 54. Yan Y.B.. Creatine kinase in cell cycle regulation and cancer. Amino Acids (2016.0) 48 1775-1784. DOI: 10.1007/s00726-016-2217-0 55. Tsikas D.. Determination of equilibria constants of arginine:glycine amidinotransferase (AGAT)-catalyzed reactions using concentrations of circulating amino acids. Amino Acids (2023.0) 55 203-213. DOI: 10.1007/s00726-022-03218-5 56. Da Silva R.P., Clow K., Brosnan J.T., Brosnan M.E.. Synthesis of guanidinoacetate and creatine from amino acids by rat pancreas. Br. J. Nutr. (2014.0) 111 571-577. DOI: 10.1017/S0007114513003012 57. Garbati P., Ravera S., Scarfi S., Salis A., Rosano C., Poggi A., Damonte G., Millo E., Balestrino M.. Effects on Energy Metabolism of Two Guanidine Molecules, (Boc). J. Cell. Biochem. (2017.0) 118 2700-2711. DOI: 10.1002/jcb.25914 58. Mühlebach S.M., Gross M., Wirz T., Wallimann T., Perriard J.C., Wyss M.. Sequence homology and structure predictions of the creatine kinase isoenzymes. Mol. Cell. Biochem. (1994.0) 133–134 245-262. DOI: 10.1007/BF01267958 59. De Vivo M., Masetti M., Bottegoni G., Cavalli A.. Role of Molecular Dynamics and Related Methods in Drug Discovery. J. Med. Chem. (2016.0) 59 4035-4061. DOI: 10.1021/acs.jmedchem.5b01684 60. Bolton E.E., Chen J., Kim S., Han L., He S., Shi W., Simonyan V., Sun Y., Thiessen P.A., Wang J.. PubChem3D: A new resource for scientists. J. Cheminform. (2011.0) 3 32. DOI: 10.1186/1758-2946-3-32 61. Kim S.. Exploring Chemical Information in PubChem. Curr. Protoc. (2021.0) 1 e217. DOI: 10.1002/cpz1.217 62. Bragina M.E., Daina A., Perez M.A.S., Michielin O., Zoete V.. The SwissSimilarity 2021 Web Tool: Novel Chemical Libraries and Additional Methods for an Enhanced Ligand-Based Virtual Screening Experience. Int. J. Mol. Sci. (2022.0) 23. DOI: 10.3390/ijms23020811 63. 63. R Core-Team R: A Language and Environment for Statistical ComputingR Foundation for Statistical ComputingVienna, Austria2021Available online: https://www.r-project.org(accessed on 23 January 2023). R: A Language and Environment for Statistical Computing (2021.0) 64. 64. RStudio-Team RStudio: Integrated Development Environment for RRStudio, PBCBoston, MA, USA2021Available online: http://www.rstudio.com/(accessed on 23 January 2023). RStudio: Integrated Development Environment for R (2021.0) 65. Wickham H., Averick M., Bryan J., Chang W., McGowan L.D., Francois R., Grolemund G., Hayes A., Henry L., Hester J.. Welcome to the Tidyverse. J. Open Source Softw. (2019.0) 4 1686. DOI: 10.21105/joss.01686 66. Kassambara A., Mundt F.. factoextra: Extract and Visualize the Results of Multivariate Data Analyses 67. Wickham H.. ggplot2: Elegant Graphics for Data Analysis (2016.0) 68. Kassambara A.. ggcorrplot: Visualization of a Correlation Matrix Using ggplot2 69. Wickham H., Hester J., Bryan J.. readr: Read Rectangular Text Data 70. Morris G.M., Huey R., Lindstrom W., Sanner M.F., Belew R.K., Goodsell D.S., Olson A.J.. AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. J. Comput. Chem. (2009.0) 30 2785-2791. DOI: 10.1002/jcc.21256 71. Hanwell M.D., Curtis D.E., Lonie D.C., Vandermeersch T., Zurek E., Hutchison G.J.. Avogadro: An advanced semantic chemical editor, visualization, and analysis platform. J. Cheminform. (2012.0) 4 17. DOI: 10.1186/1758-2946-4-17 72. Pettersen E.F., Goddard T.D., Huang C.C., Meng E.C., Couch G.S., Croll T.I., Morris J.H., Ferrin T.E.. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci. (2021.0) 30 70-82. DOI: 10.1002/pro.3943
--- title: Sugar-Sweetened Beverages Consumption in a Multi-Ethnic Population of Young Men and Association with Sociodemographic Characteristics and Obesity authors: - Jozaa Z. AlTamimi - Naseem M. Alshwaiyat - Hana Alkhalidy - Nora M. AlKehayez - Reham I. Alagal - Reem A. Alsaikan - Malak A. Alsemari - Mona N. BinMowyna - Nora A. AlFaris journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049135 doi: 10.3390/ijerph20064861 license: CC BY 4.0 --- # Sugar-Sweetened Beverages Consumption in a Multi-Ethnic Population of Young Men and Association with Sociodemographic Characteristics and Obesity ## Abstract Sugar-sweetened beverages are frequently consumed among adults and are linked with the incidence of obesity. We aimed to determine rates of weekly and daily sugar-sweetened beverage intake in a multi-ethnic population of young men and their association with sociodemographic characteristics and obesity. This cross-sectional study included 3600 young men who lived in Riyadh, KSA. Participants’ sociodemographic characteristics and frequency of sugar-sweetened beverage consumption were gathered through personal interviews. The outcome variables in this study are based on the weekly and daily consumption of sugar-sweetened beverages. Weight and height were measured following standard protocols. The rates of weekly and daily sugar-sweetened beverage intake by participants were $93.6\%$ and $40.8\%$, respectively. Nationality was a predictor of weekly and daily consumption of sugar-sweetened beverages. The highest rates of weekly ($99.5\%$) and daily ($63.9\%$) consumption were observed in subjects from the Philippines and Yemen, respectively, while Bangladeshi subjects had the lowest rates of weekly ($76.9\%$) and daily ($6.9\%$) consumption. Obesity was another predictor of sugar-sweetened beverage consumption. Obese participants had a significantly higher odds ratio of weekly sugar-sweetened beverage consumption than non-obese subjects (OR = 4.53, $$p \leq 0.037$$). In conclusion, sugar-sweetened beverage consumption was relatively high and our results support an association between the consumption of sugar-sweetened beverages and certain sociodemographic variables and obesity. ## 1. Introduction Sugar-sweetened beverages (SSBs) are considered among the leading energy-dense foods in terms of consumption globally [1]. They are defined as beverages prepared with added sugar [2]. Unfortunately, SSBs are a main dietary source of added sugar intake [3]. They are rich in calories and poor in nutrients which may negatively affect overall diet quality [4]. Surplus sugar consumption seems to have become a serious public health issue worldwide [5]. Therefore, the WHO advises keeping consumed added sugar lower than $10\%$ of daily calories [5]. Current evidence indicates that frequent SSB intake is linked to gaining weight and could cause obesity and obesity-related diseases such as type 2 diabetes and cardiovascular disease [6,7]. Nevertheless, declining SSB consumption may lower obesity incidence and many other interrelated chronic diseases [8]. Consumption of SSBs is highly prevalent among adults, especially young men [9,10]. Therefore, understanding patterns of SSB consumption and any associated sociodemographic characteristics is necessary for developing effective public health strategies to lower the consumption of foods with added sugars, especially SSBs [10]. Obesity is a serious health issue in the Kingdom of Saudi Arabia (KSA) and is concurrent with unhealthy lifestyles, including sedentary behaviors, low-quality diet, and high intake of SSBs [11,12,13]. In 2019, the Saudi Ministry of Health implemented a nationally representative health survey targeting adults aged fifteen years or older [14]. They found that the prevalence of those that were overweight and obese were $38\%$ and $20\%$, respectively. Males were more likely to be overweight ($43\%$) than females ($33\%$), whereas females were more likely to be obese ($21\%$) than males ($19\%$). Moreover, the prevalence of obesity rises with age from $10\%$ in the age group 18–29 years to $29\%$ in the age group 70–79 years, before dropping to $22\%$ in the respondents aged 80 years or over [14]. Abnormal waist circumference, defined as a waist circumference of more than 80 cm in women and more than 94 cm in men, was reported among $30\%$ of respondents ($34\%$ in females vs. $27\%$ in males). There was a steady increase in abnormal waist circumference with age, from $23\%$ in the age group 15–29 years to $58\%$ in the age group 80 years or over. The majority of respondents ($91\%$) have an abnormal waist–hip ratio defined as a waist–hip ratio equal to or more than 0.85 and 0.9 in women and men, respectively [14]. Hypertension was observed among $14\%$ of respondents ($12\%$ in females vs. $15\%$ in males). The prevalence of hypertension dramatically increases with age from $6\%$ in the age group 15–29 years to $56\%$ in the age group 80 years or over [14]. The prevalence of respondents with impaired glucose tolerance (random glucose levels ≥ 7.8 and <11.1 mmol/L) was $11\%$, while the prevalence of diabetes mellitus (random glucose levels ≥ 11.1 mmol/L) was $4\%$. The prevalence of diabetes mellitus increases with age from $4\%$ in the age group 15–29 years to $7\%$ in the age group 80 years or over [14]. Hypercholesterolemia (total cholesterol ≥ 5 mmol/L) was reported among $43\%$ of respondents. The rate of respondents with hypercholesterolemia increases dramatically with age from $39\%$ in the age group 15–29 years to $68\%$ in the age group 80 years or over [14]. Furthermore, the ten leading causes of mortality in 2019 in KSA were ischemic heart disease, road injuries, stroke, chronic kidney disease, lower respiratory infections, falls, cirrhosis, diabetes, other unintentional causes, and chronic obstructive pulmonary disease (COPD) [15]. The Kingdom leads the Middle East region in oil production and has a rapidly expanding economy. Not surprisingly, KSA attracts employees from various countries around the world. Around $50\%$ of the country’s workforce and about $90\%$ of jobs in the private business sector in KSA are occupied by expatriates [16]. Approximately, $30\%$ of the inhabitants in KSA are not Saudi citizens, and most of them are males [17]. Studying the disparities in diet patterns and their relationship to illness incidence in diverse populations is a fascinating area of research once the ethnic origins of this expatriate population are considered more closely. Therefore, the present study aimed to assess the prevalence of weekly and daily SSB intake in a multi-ethnic sample of young men living in KSA and examine the association between the consumption of SSBs and related sociodemographic factors and obesity. Our results will be valuable for policy-makers in the health care system in KSA, especially when developing strategies to reduce the consumption of SSBs for public and high-risk subgroups. ## 2.1. Research Design and Subjects Current data are from a cross-sectional research project called ROAD-KSA which was carried out from February to June 2019 and targeted young and middle-aged men [18,19,20,21,22,23,24,25]. Study subjects were randomly selected from public sites in Riyadh using a stratified clustered sampling method. The eligibility criteria included young men (20 to 35 years), who lived in Riyadh, were physically fit, and were citizens of one of the following countries: KSA, Egypt, Yemen, Syria, Jordan, Sudan, Turkey, Pakistan, Afghanistan, India, Bangladesh, and the Philippines. Per the Helsinki Declaration, participants signed informed consent forms. This work was ethically cleared by the research ethics committee/Princess Nourah bint Abdulrahman University. ## 2.2. Sociodemographic Characteristics Personal interviews were adopted to find the sociodemographic characteristics of subjects. Collected variables include nationality, age, residency period, household type, marital status, educational level, and monthly income. ## 2.3. Measurements Weight and height were measured by a qualified research team. Weight was measured to the nearest 0.1 kg by digital weight balance with the participants wearing light clothing and being barefoot. In the same way, height was measured to the nearest 0.1 cm by a portable stadiometer in a standing position and also barefoot. Body mass index (BMI) was derived by dividing weight in kilograms by height in squared meters. Obesity is defined as BMI ≥ 30 [26]. ## 2.4. Sugar-Sweetened Beverage Consumption A valid and reliable questionnaire was used to determine the frequency of SSB consumption. An impartial judgment from five nutrition research specialists examined the questionnaire’s face validity. The questionnaire’s reliability was measured using a test–retest pilot study with a two-week delay. Personal interviews were adopted to gather data. This study defines SSBs as manufactured drinks that contain added simple carbohydrates such as fructose and sucrose and does not include milk, tea, coffee, or alcohol [27]. Therefore, SSBs are categorized into the following products: regular soda, fruit drinks, and energy drinks. Regular soda is typically defined as a carbonated drink that is sweetened with simple carbohydrates. Fruit drinks involve fruit beverages that contain added simple carbohydrates but do not comprise $100\%$ natural fruit juices. Energy drinks refer to artificial drinks with a high level of a stimulant ingredient (usually caffeine) and added sugar. The frequency of SSB consumption was assessed by questioning the subjects about how many servings (about 12 fluid-ounce units) of each SSB product they consumed weekly or daily in the previous twelve months. The prevalence of weekly and daily intake of overall SSBs and individual products of SSBs were calculated according to the following definitions. Weekly and daily consumption were specified as consuming a minimum of one serving in a typical week or day, respectively [28]. ## 2.5. Statistical Analysis For data analysis, IBM SPSS Statistics for Windows (version 26. Armonk, New York, NY, USA, 2019) was used. Two binary outcome variables were used in this study: the weekly and daily consumption of SSBs. The analysis of categorical variables was handled using the Chi-square test, which was then reported as frequencies and percentages. To identify the variables connected to weekly and daily SSB consumption, a multivariate logistic regression analysis was run by adjusting for studied sociodemographic variables and obesity. All of the p-values were calculated using two-tailed testing. Statistical significance was regarded only when p-values were lower than 0.05. ## 3. Results This research involved 3600 participants. The prevalence of weekly SSB consumption among study participants stratified by sociodemographic characteristics and obesity is displayed in Table 1. For the complete study sample, the prevalence of weekly SSB consumption was $93.6\%$, whilst it was $86.9\%$ for regular soda, $60.0\%$ for fruit drinks, and $41.4\%$ for energy drinks. Stratifying subjects based on nationality shows significant differences (p ˂ 0.001) in rates of weekly SSB intake. The highest rate of SSB consumption per week was detected in participants from the Philippines ($99.5\%$), while the lowest prevalence was detected in Bangladeshi subjects ($76.9\%$). Participants who had resided for six years or more in KSA had a significantly higher prevalence of weekly SSB intake ($96.2\%$) than those with five years or less of a residency period ($91.9\%$). Interestingly, single participants had a significantly greater prevalence of weekly SSB intake ($95.5\%$) than married participants ($91.4\%$). Obese participants had a significantly higher rate of weekly SSB intake ($99.1\%$) than non-obese participants ($93.2\%$). The prevalence of daily SSB intake stratified by sociodemographic characteristics and obesity is presented in Table 2. For the complete study sample, the rate of daily SSB intake was $40.8\%$, whilst it was $14.9\%$ for regular soda, $3.9\%$ for fruit drinks, and $1.8\%$ for energy drinks. Stratifying participants based on nationality shows significant differences (p ˂ 0.001) in daily SSB consumption rates. Yemeni subjects had the highest rates of daily intake of SSBs ($63.9\%$), regular soda ($32.5\%$), fruit drinks ($11.6\%$), and energy drinks ($6.3\%$). Contrarily, Bangladeshi subjects had the lowermost rates of daily intake of SSBs ($6.9\%$), regular soda ($2.0\%$), and fruit drinks ($0.0\%$). Participants who lived within family households had a significantly greater daily SSB consumption rate ($45.7\%$) than those who lived away from their families ($39.6\%$). Single subjects had significantly greater daily intake rates of SSBs ($44.3\%$), regular soda ($17.1\%$), and fruit drinks ($4.7\%$) than married participants ($36.8\%$, $12.5\%$, and $2.9\%$, respectively). Surprisingly, participants with high education levels had a significantly higher daily SSB consumption rate ($48.8\%$) than those with low education levels ($36.2\%$). Obese subjects had higher daily consumption rates of SSBs, regular soda, fruit drinks, and energy drinks ($45.9\%$, $23.0\%$, $4.1\%$, and $2.7\%$, respectively) than non-obese subjects ($40.4\%$, $14.4\%$, $3.9\%$, and $1.7\%$, respectively). However, the difference was only significant in the case of regular soda (p ˂ 0.05). The likelihoods of weekly and daily SSB intake for sociodemographic characteristics and obesity are displayed in Table 3. Nationality was a predictor of SSB intake per week or day. Compared with Bangladeshi subjects, those from other countries had a significantly higher likelihood of weekly SSB intake (odds ratio [OR] ranged from 2.15 to 44.98, p ˂ 0.05). In the same way, compared with participants from Bangladesh, participants from other countries (except Sudan) had a significantly higher likelihood of daily SSB intake (OR ranging from 6.69 to 29.94, p ˂ 0.001). Advance in age was significantly associated with a lower likelihood of daily SSB consumption (OR = 0.96, $$p \leq 0.001$$). Moreover, subjects who resided for six years or more in KSA had a significantly higher likelihood of weekly SSB consumption than those with five years or less of a residency period (OR = 1.99, p ˂ 0.001). Married participants had a significantly lower likelihood of weekly SSB consumption (OR = 0.52, p ˂ 0.001) than single subjects. Participants with a high monthly income had a significantly lower likelihood of weekly (OR = 0.37, p ˂ 0.001) and daily (OR = 0.58, p ˂ 0.001) SSB consumption than those with a low monthly income. Lastly, obese participants had a significantly higher likelihood of weekly SSB consumption than non-obese participants (OR = 4.53, $$p \leq 0.037$$). ## 4. Discussion The current study looked at the prevalence of SSB consumption on a weekly and daily basis in a multi-ethnic sample of young men from KSA. Most participants ($93.6\%$) were weekly SSB consumers, and nearly two-fifths of the participants ($40.8\%$) were daily consumers of SSBs. Several studies have examined SSB consumption rates in adults. A recent nationally representative study from KSA found that $71.2\%$ of adults were weekly consumers of SSBs, while $35.5\%$ of adults were daily consumers. Furthermore, $80.2\%$ of young adults aged 25 to 34 years were weekly SSB consumers [28]. Another study reported that $60\%$ of young Jordanian adults were daily SSB consumers [29]. A representative study observed that $47.3\%$ and $13.6\%$ of Australian adults had weekly and daily SSB consumption, respectively. The prevalence of weekly and daily SSB consumption among young adults (18–30 years) was $67.1\%$ and 17.2, respectively [10]. Another report from the United Kingdom found that daily SSB consumption was observed in $20.4\%$ of adults [9]. According to a Scandinavian survey, $41\%$ of men in Norway regularly consume SSBs [30]. Data from five nationally representative surveys (NHANES 1999–2000 to 2007–2008) showed that the SSB consumption rate in young American adults (20–34 years) ranged between $73\%$ and $78\%$, whilst the prevalence of SSB heavy consumption among them ranged between $20\%$ and $29\%$ [31]. A Malaysian study reported that the rate of daily SSB intake in young adults was $89.3\%$ [32]. Our findings revealed significant variations in rates of SSB use among participants from various countries. This result is in harmony with results from several earlier studies that reported a significant variation in SSB intake among adults from different countries, geographic regions, or ethnicity [31,33,34]. A study comparing SSB consumption across 187 countries distributed into 21 world geographical regions found substantial variability in SSB consumption. The highest and lowest regional intake levels varied almost tenfold. The Caribbean had the greatest SSB intake, while East Asia had the lowest consumption [33]. Additionally, SSB and SSB subtype consumption by adults varied by region of residence in the USA. Rates of daily SSB consumption in the Northeast, South, West, and Midwest were $68.4\%$, $66.7\%$, $61.2\%$, and $58.8\%$, respectively. Compared with adults in the South region, adults in the Northeast region had a higher likelihood of daily SSB intake (OR = 1.13), but adults in the Midwest region (OR = 0.70) and the West region (OR = 0.78) had a lower likelihood of daily SSB intake [34]. In another report from the USA, SSB consumption differs by ethnicity. Black and Hispanic young adults had a higher likelihood of drinking SSBs than their White counterparts [31]. The reasons for differences in the rates of SSB intake based on nationality are not completely resolute. However, the possible elucidations could include variations in the environmental exposure to SSBs that the adults experience in their home countries in earlier life periods regarding SSB availability and accessibility [35,36]. Another reason could be the variations in their reception and response to SSB advertising in the host country due to variations in language and cultural norms [37]. In the current study, Bangladeshi and Sudanese participants had relatively the lowest rates of daily SSB intake. In KSA, most residents from Bangladesh and Sudan are less educated, come from rural communities, and have manual-type occupations in the services, construction, or agricultural sectors. This could make them less exposed to SSB consumption and be reflected in their dietary behavior. However, differences in SSB intake rates by nationality should be considered when interventions to reduce SSB intake are planned [38]. Discrepancies in SSB intake rates in adults by sociodemographic characteristics have been well recognized in the literature [39]. Our results were consistent with results from former studies that described a greater likelihood of SSB intake among subjects of a younger age [10,28,32,39,40,41]. The current study found significant associations between weekly consumption of SSBs and age, residency duration, marital status, and monthly income. We found that a longer residency duration in KSA was associated with a higher likelihood of weekly SSB consumption. The development experienced in this country and how a modified lifestyle has affected adults’ choices may be to blame for this [42]. As immigrants live in a new country for a longer time, their health will worsen. Health problems may be driven by cultural influences, social and economic changes, as well as adjustments to eating habits created by migration [43]. Another finding was that single subjects had a higher likelihood of SSB intake than their married counterparts. Single individuals frequently consume fast foods, which is linked to a higher SSB intake [28,41]. Our results agreed with past studies that noticed that a greater likelihood of SSB intake was associated with lower monthly income [9,31,32]. Many people do not have continuous access to healthy food due to their economic status. Budget-friendly food choices such as fast food and SSBs are usually low in essential nutrients and high in calories. People with low income are at higher risk of nutritional deficiencies and health problems such as obesity, diabetes, and cardiovascular disease. Therefore, addressing disparities in food access for different groups of the community, especially those with a low income, is a crucial element to achieve food justice and promote public health. Food justice aims to guarantee that everyone has access to nutritious, affordable, and culturally acceptable food [44]. Remarkably, our results confirmed the relationship between weekly SSB intake and obesity. Obesity is an expensive illness with elevated rates of morbidity and mortality that affect many adults [45,46]. Serval studies linked the consumption of SSBs to obesity and other obesity-associated diseases [47,48,49]. Several mechanisms were proposed to explain how SSB consumption can result in obesity. The calorie intake of adults is significantly increased by the consumption of SSBs, which shifts the energy balance toward positive [50]. According to a recent study, young Jordanian adults consumed an average of 481 kcal of total daily calories from SSBs [29]. Furthermore, when opposed to consuming calories from solid foods with high fiber content, the fluid calories from SSBs promote more hunger sense, which could cause an excessive intake of calories [51]. Another explanation is higher hepatic lipogenesis due to consuming high fructose from SSBs, which affects insulin secretion and leads to insulin resistance and excess fat deposition [52]. At the community level, SSB intake should be reduced through various environmental and regulatory changes [27]. It may be possible to reduce SSB consumption by limiting their availability in community areas and lower exposure to SSB advertising in mass media [10]. On the other hand, general attempts to limit SSB intake may be helped by increasing public awareness regarding the negative health effects of SSBs through public health education campaigns [53]. Governments have been urged to take action by international public health organizations due to the extensive and rapid rise in SSB consumption. One effective policy to limit SSB intake is forcing taxes on SSB prices [54,55]. Furthermore, strategies that promote using energy-free liquids like water in place of SSBs should be taken into account [55]. The health sector transformation program that is included in the Vision 2030 program for KSA focuses on boosting public health and disease prevention through programs that target different members of the community in the country, including citizens and foreign residents. The key aspect of this program includes legislation that limits access to unhealthy foods such as SSBs by applying taxes and educational programs implemented by various governmental bodies such as the Ministry of Health and universities to promote healthy eating and limit unhealthy foods such as SSBs [28,56]. A few limitations can be viewed in this study. Causality could not be inferred from the significant correlations due to the cross-sectional design. Another problem was relying on participants’ memories while collecting consumption data using a self-reported frequency technique. This method may yield an underestimated SSB intake compared to a 24-h recall dietary method. Because SSBs are defined differently in different studies, comparing our findings with those of earlier research might be difficult. Lastly, our study did not account for daily energy consumption to examine the link between SSB intake and obesity. Nevertheless, the current study offers insightful information on the prevalence of SSB intake and related factors. ## 5. Conclusions This study found that the prevalence of SSB intake in young men from KSA was comparatively high. The findings revealed that nationality was a predictor of SSB intake per week or day. The results also support a link between obesity and some sociodemographic characteristics and SSB consumption. ## References 1. Malik V.S., Pan A., Willett W.C., Hu F.B.. Sugar-sweetened beverages and weight gain in children and adults: A systematic review and meta-analysis. Am. J. Clin. Nutr. (2013.0) 98 1084-1102. DOI: 10.3945/ajcn.113.058362 2. 2. Centers for Disease Control and Prevention (CDC) The CDC Guide to Strategies for Reducing the Consumption of Sugar-Sweetened BeveragesCDCAtlanta, GA, USA2010. The CDC Guide to Strategies for Reducing the Consumption of Sugar-Sweetened Beverages (2010.0) 3. Luger M., Lafontan M., Bes-Rastrollo M., Winzer E., Yumuk V., Farpour-Lambert N.. Sugar-Sweetened Beverages and Weight Gain in Children and Adults: A Systematic Review from 2013 to 2015 and a Comparison with Previous Studies. Obes. Facts (2017.0) 10 674-693. DOI: 10.1159/000484566 4. Sharkey J.R., Johnson C.M., Dean W.. Less-healthy eating behaviors have a greater association with a high level of sugar-sweetened beverage consumption among rural adults than among urban adults. Food Nutr. Res. (2011.0) 55 5819. DOI: 10.3402/fnr.v55i0.5819 5. 5. World Health Organization (WHO) Guideline: Sugars Intake for Adults and Children (2015)World Health OrganizationGeneva, Switzerland2015Available online: https://www.who.int/publications/i/item/9789241549028(accessed on 24 February 2023). Guideline: Sugars Intake for Adults and Children (2015) (2015.0) 6. Hu F.B., Malik V.S.. Sugar-sweetened beverages and risk of obesity and type 2 diabetes: Epidemiologic evidence. Physiol. Behav. (2010.0) 100 47-54. DOI: 10.1016/j.physbeh.2010.01.036 7. Bermudez O.I., Gao X.. Greater consumption of sweetened beverages and added sugars is associated with obesity among US young adults. Ann. Nutr. Metab. (2010.0) 57 211-218. DOI: 10.1159/000321542 8. Hu F.B.. Resolved: There is sufficient scientific evidence that decreasing sugar-sweetened beverage consumption will reduce the prevalence of obesity and obesity-related diseases. Obes. Rev. (2013.0) 14 606-619. DOI: 10.1111/obr.12040 9. Barrett P., Imamura F., Brage S., Griffin S.J., Wareham N.J., Forouhi N.G.. Sociodemographic, lifestyle and behavioural factors associated with consumption of sweetened beverages among adults in Cambridgeshire, UK: The Fenland Study. Public Health Nutr. (2017.0) 20 2766-2777. DOI: 10.1017/S136898001700177X 10. Miller C., Ettridge K., Wakefield M., Pettigrew S., Coveney J., Roder D., Durkin S., Wittert G., Martin J., Dono J.. Consumption of Sugar-Sweetened Beverages, Juice, Artificially-Sweetened Soda and Bottled Water: An Australian Population Study. Nutrients (2020.0) 12. DOI: 10.3390/nu12030817 11. Al-Rethaiaa A.S., Fahmy A.-E.A., Al-Shwaiyat N.M.. Obesity and eating habits among college students in Saudi Arabia: A cross sectional study. Nutr. J. (2010.0) 9 39. DOI: 10.1186/1475-2891-9-39 12. Alfaris N.A., Al-Tamimi J.Z., Al-Jobair M.O., Al-Shwaiyat N.M.. Trends of fast food consumption among adolescent and young adult Saudi girls living in Riyadh. Food Nutr. Res. (2015.0) 59 26488. DOI: 10.3402/fnr.v59.26488 13. Benajiba N., Mahboub S.M.. Consumption of Sugar-Sweetened Soft Drinks Among Saudi Adults: Assessing Patterns and Identifying Influencing Factors Using Principal Component Analysis. Pak. J. Nutr. (2019.0) 18 401-407. DOI: 10.3923/pjn.2019.401.407 14. 14. Saudi Ministry of Health World Health Survey—Saudi Arabia (KSAWHS) 2019: Final ReportSaudi Ministry of HealthRiyadh, Saudi ArabiaAvailable online: https://www.moh.gov.sa/en/Ministry/Statistics/Population-Health-Indicators/Documents/World-Health-Survey-Saudi-Arabia.pdf.(accessed on 24 February 2023). World Health Survey—Saudi Arabia (KSAWHS) 2019: Final Report 15. Vos T., Lim S.S., Abbafati C., Abbas K.M., Abbasi M., Abbasifard M., Bhutta Z.A.. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: A systematic analysis for the Global Burden of Disease Study 2019. Lancet (2020.0) 396 1204-1222. DOI: 10.1016/S0140-6736(20)30925-9 16. De Bel-Air F.. Demography, Migration and Labour Market in Saudi Arabia. Gulf Labour Markets and Migration. European University Institute (EUI) and Gulf Research Center (GRC). GLMM-EN-No. 1/2014. (2014.0) 17. Population by Gender, Age Groups and Nationality (Saudi/Non-Saudi), The Fifth Saudi Census, 2010 18. Alagal R.I., AlFaris N.A., AlTamimi J.Z., Alshwaiyat N.M., Ahmad A., Alzaheb R.A., AlKehayez N.M.. Differences in Overweight and Obesity Prevalence among Young Men from Twelve Middle Eastern and Asian Countries Living in Saudi Arabia. Healthcare (2022.0) 10. DOI: 10.3390/healthcare10040690 19. AlFaris N.A., Alshwaiyat N.M., Alkhalidy H., Alagal R.I., AlTamimi J.Z., AlKehayez N.M.. Breakfast Skipping in a Multi-Ethnic Population of Middle-Aged Men and Relationship With Sociodemographic Variables and Weight Status. Front. Nutr. (2022.0) 8 761383. DOI: 10.3389/fnut.2021.761383 20. AlFaris N.A., Alshwaiyat N.M., AlTamimi J.Z., Alagal R.I., Al-Jamal H.A., AlKehayez N.M.. Physical Activity Levels of a Multi-Ethnic Population of Middle-Aged Men Living in Saudi Arabia and Factors Associated With Physical Inactivity. Int. J. Public Health (2022.0) 66 1604328. DOI: 10.3389/ijph.2021.1604328 21. AlFaris N.A., Alshwaiyat N.M., Alkhalidy H., ALTamimi J.Z., Alagal R.I., Alsaikan R.A., Alsemari M.A., BinMowyna M.N., AlKehayez N.M.. Sugar-Sweetened Beverages Consumption in a Multi-ethnic Population of Middle-Aged Men and Association with Sociodemographic Variables and Obesity. Front. Nutr. (2022.0) 9 987048. DOI: 10.3389/fnut.2022.987048 22. AlTamimi J.Z., Alagal R.I., AlKehayez N.M., Alshwaiyat N.M., Al-Jamal H.A., AlFaris N.A.. Physical Activity Levels of a Multi-Ethnic Population of Young Men Living in Saudi Arabia and Factors Associated with Physical Inactivity. Front. Public Health (2022.0) 9 734968. DOI: 10.3389/fpubh.2021.734968 23. AlTamimi J.Z., Alshwaiyat N.M., AlFaris N.A., AlKehayez N.M., Ahmad A., Alagal R.I.. Differences in Overweight and Obesity Prevalence in Middle-Aged Men from Twelve Middle Eastern and Asian Countries Living in Saudi Arabia. Int. J. Gen. Med. (2022.0) 15 3333-3343. DOI: 10.2147/IJGM.S359639 24. AlTamimi J.Z., Alshwaiyat N.M., Alkhalidy H., AlFaris N.A., AlKehayez N.M., Alsemari M.A., Alagal R.I.. Prevalence of Fast Food Intake among a Multi-Ethnic Population of Young Men and Its Connection with Sociodemographic Determinants and Obesity. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph192214933 25. AlTamimi J.Z., Alshwaiyat N.M., Alkhalidy H., AlFaris N.A., AlKehayez N.M., Alagal R.I.. Breakfast Skipping among a Multi-Ethnic Population of Young Men and Relationship with Sociodemographic Determinants and Weight Status. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph19052903 26. 26. World Health Organization (WHO) Obesity and OverweightWHOGeneva, Switzerland2017. Obesity and Overweight (2017.0) 27. Vercammen K.A., Moran A.J., Soto M.J., Kennedy-Shaffer L., Bleich S.N.. Decreasing Trends in Heavy Sugar-Sweetened Beverage Consumption in the United States, 2003 to 2016. J. Acad. Nutr. Diet. (2020.0) 120 1974-1985.e5. DOI: 10.1016/j.jand.2020.07.012 28. Al-Hanawi M.K., Ahmed M.U., Alshareef N., Qattan A.M.N., Pulok M.H.. Determinants of Sugar-Sweetened Beverage Consumption Among the Saudi Adults: Findings From a Nationally Representative Survey. Front. Nutr. (2022.0) 9 744116. DOI: 10.3389/fnut.2022.744116 29. Bawadi H., Khataybeh T., Obeidat B., Kerkadi A., Tayyem R., Banks A.D., Subih H.. Sugar-Sweetened Beverages Contribute Significantly to College Students’ Daily Caloric Intake in Jordan: Soft Drinks Are Not the Major Contributor. Nutrients (2019.0) 11. DOI: 10.3390/nu11051058 30. Paulsen M.M., Myhre J.B., Andersen L.F.. Beverage Consumption Patterns among Norwegian Adults. Nutrients (2016.0) 8. DOI: 10.3390/nu8090561 31. Han E., Powell L.M.. Consumption Patterns of Sugar-Sweetened Beverages in the United States. J. Acad. Nutr. Diet. (2013.0) 113 43-53. DOI: 10.1016/j.jand.2012.09.016 32. Norliza A., Muhammad Afiq M.Z., Nur Azilah A., Khor B.H., Halimatus Sakdiah M.. Prevalence and factors associated with sugar-sweetened beverage intake among undergraduate students in a public university in Malaysia. Pak. J. Nutr. (2019.0) 18 354-363 33. Singh G.M., Micha R., Khatibzadeh S., Shi P., Lim S., Andrews K.G.. Global, regional, and national consumption of sugar-sweetened beverages, fruit juices, and milk: A systematic assessment of beverage intake in 187 countries. PLoS ONE (2015.0) 10. DOI: 10.1371/journal.pone.0124845 34. Park S., McGuire L.C., Galuska D.A.. Regional Differences in Sugar-Sweetened Beverage Intake among US Adults. J. Acad. Nutr. Diet. (2015.0) 115 1996-2002. DOI: 10.1016/j.jand.2015.06.010 35. Elfassy T., Adjoian T., Lent M.. Sugary Drink Consumption Among NYC Children, Youth, and Adults: Disparities Persist Over Time, 2007–2015. J. Community Health (2019.0) 44 297-306. DOI: 10.1007/s10900-018-0587-9 36. Fiorito L.M., Marini M., Mitchell D.C., Smiciklas-Wright H., Birch L.L.. Girls’ Early Sweetened Carbonated Beverage Intake Predicts Different Patterns of Beverage and Nutrient Intake across Childhood and Adolescence. J. Am. Diet. Assoc. (2010.0) 110 543-550. DOI: 10.1016/j.jada.2009.12.027 37. Morley B.C., Niven P.H., Dixon H.G., Swanson M.G., McAleese A.B., Wakefield M.A.. Controlled cohort evaluation of the LiveLighter mass media campaign’s impact on adults’ reported consumption of sugar-sweetened beverages. BMJ Open (2018.0) 8 e019574. DOI: 10.1136/bmjopen-2017-019574 38. Levy D.T., Friend K.B., Wang Y.C.. A Review of the Literature on Policies Directed at the Youth Consumption of Sugar Sweetened Beverages. Adv. Nutr. Int. Rev. J. (2011.0) 2 182S-200S. DOI: 10.3945/an.111.000356 39. Rehm C.D., Matte T.D., Van Wye G., Young C., Frieden T.R.. Demographic and Behavioral Factors Associated with Daily Sugar-sweetened Soda Consumption in New York City Adults. J. Urban Health (2008.0) 85 375-385. DOI: 10.1007/s11524-008-9269-8 40. Mendy V.L., Vargas R., Payton M., Cannon-Smith G.. Association Between Consumption of Sugar-Sweetened Beverages and Sociodemographic Characteristics Among Mississippi Adults. Prev. Chronic Dis. (2017.0) 14 E137. DOI: 10.5888/pcd14.170268 41. Pollard C.M., Meng X., Hendrie G.A., Hendrie D., Sullivan D., Pratt I.S., Kerr D.A., Scott J.A.. Obesity, socio-demographic and attitudinal factors associated with sugar-sweetened beverage consumption: Australian evidence. Aust. New Zealand J. Public Health (2015.0) 40 71-77. DOI: 10.1111/1753-6405.12482 42. Alzeidan R.A., Rabiee F., Mandil A.A., Hersi A.S., Ullah A.A.. Changes in dietary habits and physical activity and status of metabolic syndrome among expatriates in Saudi Arabia. East. Mediterr. Health J. (2017.0) 23 836-844. DOI: 10.26719/2017.23.12.836 43. Lassetter J.H., Callister L.C.. The impact of migration on the health of voluntary migrants in western societies: A review of the literature. J. Transcult. Nurs. (2009.0) 20 93-104. DOI: 10.1177/1043659608325841 44. Murray S., Gale F., Adams D., Dalton L.. A scoping review of the conceptualisations of food justice. Public Health Nutr. (2023.0). DOI: 10.1017/S1368980023000101 45. Abdelaal M., le Roux C.W., Docherty N.G.. Morbidity and mortality associated with obesity. Ann. Transl. Med. (2017.0) 5 161. DOI: 10.21037/atm.2017.03.107 46. Alshwaiyat N.M., Ahmad A., Hassan W.M.R.W., Al-Jamal H.A.N.. Association between obesity and iron deficiency (Review). Exp. Ther. Med. (2021.0) 22 1268. DOI: 10.3892/etm.2021.10703 47. De Koning L., Malik V.S., Kellogg M.D., Rimm E.B., Willett W.C., Hu F.B.. Sweetened beverage consumption, incident coronary heart disease, and biomarkers of risk in men. Circulation (2012.0) 125 1735-1741. DOI: 10.1161/CIRCULATIONAHA.111.067017 48. de Koning L., Malik V.S., Rimm E.B., Willett W.C., Hu F.B.. Sugar-sweetened and artificially sweetened beverage consumption and risk of type 2 diabetes in men. Am. J. Clin. Nutr. (2011.0) 93 1321-1328. DOI: 10.3945/ajcn.110.007922 49. Malik V.S., Hu F.B.. The role of sugar-sweetened beverages in the global epidemics of obesity and chronic diseases. Nat. Rev. Endocrinol. (2022.0) 18 205-218. DOI: 10.1038/s41574-021-00627-6 50. Bachman C.M., Baranowski T., Nicklas T.A.. Is there an association between sweetened beverages and adiposity?. Nutr. Rev. (2006.0) 64 153-174. DOI: 10.1111/j.1753-4887.2006.tb00199.x 51. Brand-Miller J.C., Holt S.H., Pawlak D.B., McMillan J.. Glycemic index and obesity. Am. J. Clin. Nutr. (2002.0) 76 281S-285S. DOI: 10.1093/ajcn/76.1.281S 52. Ludwig D.S.. Examining the Health Effects of Fructose. JAMA (2013.0) 310 33-34. DOI: 10.1001/jama.2013.6562 53. Pomeranz J.L.. Advanced policy options to regulate sugar-sweetened beverages to support public health. J. Public Health Policy (2012.0) 33 75-88. DOI: 10.1057/jphp.2011.46 54. Lee M.M., Falbe J., Schillinger D., Basu S., McCulloch C.E., Madsen K.A.. Sugar-Sweetened Beverage Consumption 3 Years After the Berkeley, California, Sugar-Sweetened Beverage Tax. Am. J. Public Health (2019.0) 109 637-639. DOI: 10.2105/AJPH.2019.304971 55. Colchero M.A., Molina M., Guerrero-López C.M.. After Mexico implemented a tax, purchases of sugar-sweetened beverages decreased and water increased: Difference by place of residence, household composition, and income level. J. Nutr. (2017.0) 147 1552-1557. DOI: 10.3945/jn.117.251892 56. Jalloun R.A., Qurban M.A.. The impact of taxes on soft drinks on adult consumption and weight outcomes in Medina, Saudi Arabia. Hum. Nutr. Metab. (2022.0) 27 200139. DOI: 10.1016/j.hnm.2022.200139
--- title: Uridine Alleviates Sepsis-Induced Acute Lung Injury by Inhibiting Ferroptosis of Macrophage authors: - Kai Lai - Congkuan Song - Minglang Gao - Yu Deng - Zilong Lu - Ning Li - Qing Geng journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049139 doi: 10.3390/ijms24065093 license: CC BY 4.0 --- # Uridine Alleviates Sepsis-Induced Acute Lung Injury by Inhibiting Ferroptosis of Macrophage ## Abstract Uridine metabolism is extensively reported to be involved in combating oxidative stress. Redox-imbalance-mediated ferroptosis plays a pivotal role in sepsis-induced acute lung injury (ALI). This study aims to explore the role of uridine metabolism in sepsis-induced ALI and the regulatory mechanism of uridine in ferroptosis. The Gene Expression Omnibus (GEO) datasets including lung tissues in lipopolysaccharides (LPS) -induced ALI model or human blood sample of sepsis were collected. In vivo and vitro, LPS was injected into mice or administered to THP-1 cells to generate sepsis or inflammatory models. We identified that uridine phosphorylase 1 (UPP1) was upregulated in lung tissues and septic blood samples and uridine significantly alleviated lung injury, inflammation, tissue iron level and lipid peroxidation. Nonetheless, the expression of ferroptosis biomarkers, including SLC7A11, GPX4 and HO-1, were upregulated, while lipid synthesis gene (ACSL4) expression was greatly restricted by uridine supplementation. Moreover, pretreatment of ferroptosis inducer (Erastin or Era) weakened while inhibitor (Ferrostatin-1 or Fer-1) strengthened the protective effects of uridine. Mechanistically, uridine inhibited macrophage ferroptosis by activating Nrf2 signaling pathway. In conclusion, uridine metabolism dysregulation is a novel accelerator for sepsis-induced ALI and uridine supplementation may offer a potential avenue for ameliorating sepsis-induced ALI by suppressing ferroptosis. ## 1. Introduction Acute lung injury (ALI) is a severe respiratory disease characterized by diffused non-cardiogenic pulmonary edema due to alveolar damage in pathological manifestation and featured irreformable hypoxemia and respiratory distress in clinical performance [1]. For lack of effective treatment options currently, ALI often develops into its extreme form acute respiratory distress (ARDS), a clinical life-threatening syndrome with high mortality globally [2]. In pathogenesis, ALI can be caused by a variety of pathogenic factors through direct damage or indirect injury imposed by hyper inflammation [3]. Among these, sepsis is the primary risk factor for leading ALI, as the lung is extremely susceptible to the grave immune response stimulated by sepsis during the multiorgan dysfunction stage [4,5]. However, to date, the exact mechanism underpinning sepsis-induced ALI remains obscure and effective agents are exceedingly limited [6,7]. Therefore, it is imperative to dig out the pathogenesis of sepsis-induced ALI and try to discover effective interventions to improve outcomes for patients who are struggling with sepsis. Uridine, a pyrimidine nucleoside, is composed of uracil and ribose and employed to synthesize pyrimidine nucleotide [8]. Uridine metabolism, including anabolism and catabolism, are a series of fine controlled processes and intimately regulated by a spectrum of enzymes [9]. Increasing studies have shown that uridine, the most abundant nucleoside in human blood, affects multiple physiological processes, genetic material and glycogen synthesis [10]. Additionally, the disturbance of uridine metabolism has been implicated in the progression of a set of disease courses [11,12,13]. It is conceivable theoretically that the elaborately regulated processes of uridine metabolism are inseparable from its essential role in modulating pathophysiological processes. As a matter of fact, emerging researches have reported that uridine exerts anti-inflammation [14,15], anti-fibrosis [16], antioxidation [17,18,19] and anti-aging functions [20]. For example, Irina B. Krylova1 discovered that uridine treatment attenuated myocardial ischemic-reperfusion injury via antioxidation by activating ATP-dependent potassium (mitoKATP) channel [21]. Nevertheless, in a cross-species metabolomic study, higher concentration of uridine was identified to exist in blood of younger counterparts and supplementation of uridine was validated to recover regenerative capability of older tissues and promoted the proliferation of stem cells [22]. Moreover, uridine exhibited potent anti-inflammatory actions during endotoxemia in murine model [23]. However, the landscape of uridine metabolism in sepsis-induced ALI is largely unknown and the other mechanisms underpinning the protective effects of uridine deserve to be explored. Ferroptosis is a newly discovered cell death form, defined by iron-dependent lipid peroxidation leading to membrane destruction and cellular contents outflowing, eventually causing cell death [24]. Mounting evidence has demonstrated that ferroptosis is implicated in various pathophysiological processes and involved in various of human diseases, including ALI [25], acute myocardial infarction [26], tumorigenesis [27], acute kidney injury [28], etc. Nuclear factor erythroid 2-related factor 2 (NFE2L2, or Nrf2), a key modulator for combating cellular oxidative stress, is well reported to play a critical role in regulating ferroptosis [29]. Given the antioxidant property of uridine, it is conceivable and desirable to probe into the relationship between uridine metabolism and ferroptosis in sepsis-induced ALI. Herein, we discovered that uridine metabolism was disturbed in sepsis-induced ALI. In addition, uridine supplementation mitigated sepsis-induced ALI by inhibiting macrophage ferroptosis via activating Nrf2 pathway and inhibiting Acyl-CoA synthetase long-chain family member 4 (ACSL4) expression. Overall, uridine is identified to be an endogenous anti-inflammatory metabolite, and exogenous supplementation of uridine could be a promising therapeutic intervention for sepsis-induced ALI. ## 2.1. Uridine Metabolism Homeostasis Was Disturbed in Sepsis-Induced ALI To explore the role of uridine metabolism in sepsis-induced ALI, we collected uridine metabolism-related genes and downloaded microarray datasets of sepsis-induced ALI. The cohorts with same platform were integrated into a larger cohort by removing batch effects. As shown in Figure 1A,B, the general expression value in each integrated GEO cohort demonstrated that batch effects were removed. Then, differentially expressed analysis was performed for each integrated GEO cohort and the differentially expressed genes were intersected with uridine metabolism-related genes. Interestingly, there were both two differentially expressed uridine metabolism-related genes in GPL339 and GPL1261, and we found that uridine phosphorylase 1 (UPP1) was the only differentially expressed uridine metabolism-related gene in sepsis-induced lung tissue compared to control (Figure 1C–E). UPP1 is well known for functioning as key enzyme in catalyzing uridine to uracil and ribose-1-phosphate. To confirm the function of UPP1 further, we constructed the PPI network for UPP1 and discovered these genes were closely related to uridine metabolism (Figure 1F,G). Moreover, we detected the mRNA level of UPP1 in lung tissues of sepsis-induced ALI. Consistently, UPP1 was validated to be significantly upregulated in sepsis-induced ALI (Figure 1H). Taken together, these data suggested that uridine metabolism in lung tissues was interfered in sepsis-induced ALI. ## 2.2. Uridine Supplementation Alleviated Sepsis-Induced ALI and Inflammation of Lung Tissue In the first part, we discovered UPP1 was upregulated after LPS stimulation. Considering the function of UPP1 proposed to decrease uridine level, we hypothesized that maintaining uridine metabolism homeostasis through uridine supplementation could benefit sepsis-induced ALI. To address the assumption and investigate the function of uridine, sepsis-induced ALI model was established by LPS injection intraperitoneally and uridine treatment at the same time (Figure 2A). As expected, compared to the LPS group, uridine treatment attenuated sepsis-induced ALI pathological changes, as manifested by pulmonary hemorrhage, interstitial edema and thickening of the alveolar wall in HE staining (Figure 2B). Furthermore, LPS stimulation markedly increased the lung injury score, while uridine significantly reversed the damage (Figure 2C). In addition, the protein level in BALF and lung edema extent were also lightened in uridine treatment group compared with LPS stimulation alone group (Figure 2D,E). For inflammatory cell infiltration level, we detected the activity of MPO, a biomarker for neutrophils, and found that MPO activity was significantly higher in the LPS group than in the control group, however, uridine treatment greatly reversed this situation (Figure 2F). Similarly, uridine supplementation markedly prevented the inflammatory response in lung tissues, as reflected by the reduced protein and mRNA level of IL-6, TNF-α and IL-1β (Figure 2G–J). Altogether, these data unveiled that uridine supplementation could inhibit the extent of lung damage as well as inflammation response in sepsis-induced ALI in vivo. ## 2.3. Uridine Inhibited Ferroptosis in Lung Tissue Dependent on Nrf2 Activation and ACSL4 Inhibition Previous study has reported that uridine could mitigate lipid peroxidation in myocardial ischemic/reperfusion model [21]. Additionally, the connection between ferroptosis and uridine is enlightened by the discovery that the pyrimidine biosynthesis enzyme, dihydroorotate dehydrogenase (DHODH), is a ferroptosis defender [30]. However, the relationship between ferroptosis (iron-dependent lipid peroxidation) and uridine metabolism in sepsis-induced ALI is still largely unknown. To address the role of uridine in ferroptosis, the level of glutathione (GSH, a reductive agent for antioxidation), lipid peroxidation products MDA and tissue iron level were assessed. It was found that LPS stimulation significantly reduced the GSH level, while uridine treatment reversed this situation and increased GSH extent in lung tissue (Figure 3A). In contrast, the MDA level and tissue iron were decreased after uridine supplementation (LPS+uridine) compared with LPS group (Figure 3B,C). Moreover, SLC7A11, functioning to exchange intracellular glutamate for extracellular cystine, and glutathione-dependent antioxidant enzyme GPX4 are two well accepted markers of ferroptosis [31]. Our results demonstrated that the protein and mRNA of SLC7A11 and GPX4 were both downregulated after LPS stimulation, however, uridine treatment greatly upregulated their expression (Figure 3D,E,G). It is well known that Nrf2 is an antioxidant response element, which could promote HO1, GPX4 and SLC7A11 expression to inhibit ferroptosis. Thus, we detected the expression of Nrf2/HO1 axis. As expected, uridine augmented HO1 expression via activating Nrf2 (Figure 3F,G). Additionally, ACSL4 is a key enzyme that regulates lipid synthesis, thus, to promote ferroptosis. Our result unveiled that uridine obviously weakened the expression of ACSL4, which was markedly raised in LPS group (Figure 3G). Conclusively, our results revealed that uridine inhibited the ferroptosis via activating Nrf2 pathway and suppressing ACSL4 expression in sepsis-induced ALI in vivo. ## 2.4. Uridine Supplementation Restrained Inflammation and Oxidative Stress in THP-1 Cells To determine which cell may benefit uridine supplementation most, we evaluated the UPP1 expression after LPS stimulation on three cell types, which are representatives for playing a vital role in the pathogenesis of sepsis-induced ALI. As shown in Figure 4A, UPP1 mRNA expression was dramatically enhanced in THP-1 cells (on behalf of alveolar macrophage) than other two cells. Therefore, THP-1 cells were selected for further experiments. Next, to ascertain the optimal concentration for uridine treatment and anti-inflammatory properties of uridine in vitro, we measured the mRNA level of IL-1β, IL-6 and TNF-α after LPS stimulation supplemented with different concentration of uridine. The results showed that the anti-inflammatory effects of uridine were fortified with the increase of concentration (Figure 4B–D). Eventually, we selected 200 μM uridine to treat THP-1 cells. In addition, we validated the antioxidative capacity of uridine in vitro. The ROS fluorescence intensity was observably promoted by LPS stimulation, which was significantly bated in LPS+uridine group (Figure 4E,G). Likewise, the C11-BODIPY fluorescence probe was used to determine cellular lipids and the probe will show green fluorescence when bound to oxidized lipids. It is quite clear from the results that oxidized lipids were dramatically increased after LPS stimulation, while uridine supplementation prevented that situation (Figure 4F,H). As a further verification, we examined the GSH and MDA level in THP-1 cells. Consistent with the results in lung tissue, uridine treatment also increased GSH level and lowered the MDA content in vitro (Figure 4I,J). These data suggested that uridine treatment stabilized cellular oxidation and reduction balance in vitro, thus, to exhibit anti-inflammatory effects. ## 2.5. Uridine Suppressed Ferroptosis in THP-1 Cells via Activating Nrf2 and Inhibiting ACSL4 Expression To assess the relationship between uridine metabolism and ferroptosis in vitro, we treated uridine in THP-1 cells and examined the protein and mRNA expression of markers of ferroptosis. Similar results were observed in vitro just the same as in vivo. LPS administration significantly downregulated Nrf2, HO1, GPX4 and SLC7A11 expression in protein level, which were all greatly upregulated after uridine addition (Figure 5A). Furthermore, ACSL4 was decreased after uridine supplementation (Figure 5A). Meanwhile, we also discovered that the downregulated mRNA level of GPX4, SLC7A11 and HO1 by LPS were all restored after uridine addition (Figure 5B–D). Subsequently, the cell viability was examined by CCK8 for ferroptosis could lower the cell viability. As shown in Figure 5E, LPS stimulation markedly attenuated cell viability, while uridine supplementation abated the death effect of LPS. In addition, it was visualized that ferroptosis inhibitor ferrostatin-1 (Fer-1) application strengthened while ferroptosis inducer Erastin (Era) application weakened the protective effects of uridine (Figure 5E). Apart from cell viability, the inhibitory effect of uridine for regulating cellular ROS level was also adjusted by Fer-1 and Era (Figure 5F). Collectively, these data uncovered that uridine could inhibit ferroptosis in THP-1 cells. ## 2.6. Nrf2 Knockdown Abrogated the Ferroptosis Resistance Effect of Uridine To further validate the role of Nrf2 in the process of uridine combating ferroptosis, the Nrf2 was knockdown by siRNA and it successfully silenced Nrf2 expression in THP-1 cells (Figure 6A). At the same time, GPX4, SLC7A11 and HO1 expression were significantly decreased after Nrf2 silence, and the protective effect of uridine was subdued (Figure 6A). Otherwise, compared with LPS+uridine group, Nrf2 silence group had a lower cellular GSH level and higher MDA level (Figure 6B,C). Mitochondrial membrane potential decrease often indicates mitochondrial damage and mitochondrial oxidative stress occurrence, which is associated with ferroptosis. Thus, we measured the mitochondrial membrane potential by applying the JC-1 probe. When mitochondrial membrane is damaged, green fluorescence (monomer) accumulates. Consistently, uridine supplementation mitigated the mitochondrial damage, however, Nrf2 silence abrogated the protective effect (Figure 6D). These data further consolidated the role of Nrf2 in the ferroptosis resistant effect of uridine. ## 2.7. Dysregulated Uridine Metabolism Was Confirmed in Patients with Sepsis To confirm the role of uridine metabolism in patients with sepsis, we collected GEO cohorts including human blood samples from healthy individuals and sepsis patients. Differentially expressed analysis showed that uridine metabolism was indeed dysregulated in sepsis and upregulated UPP1 was also verified in other two GEO datasets (Figure 7A–C). Furthermore, we carried out immune infiltration analysis based on ssGSEA algorithm. It was obvious that, in contrast to the UPP1 highly expressed group, the lower neutrophils infiltration was seen in the UPP1 less expressed group (Figure 7D). In addition, as shown in Figure 7D, most immune cells possessed lower infiltration level in the UPP1 highly expressed group, which reflected that UPP1 may serve an inducer for the formation of sepsis-induced immunoparalysis state [32,33]. Next, to explore the ability of UPP1 to diagnose sepsis and predict survival of patients with sepsis, we drew receiver operating characteristic (ROC) curve and calculated area under curve (AUC). The results demonstrated that UPP1 possessed good diagnostic performance (AUC = 0.916) and prognostic prediction performance (AUC = 0.662) for patients with sepsis (Figure 7E,F). To sum up, these bioinformatic analysis further strengthened the notion that uridine metabolism is indeed dysregulated in sepsis and uridine supplementation may be a potent effective avenue for treating sepsis and sepsis-induced ALI. ## 3. Discussion In the circumstance of sepsis-induced ALI, metabolism in macrophage is extensively reprogramming to response to stress signal, thus, to produce inflammation and aggravate lung damage [34,35]. However, uridine metabolism is poorly investigated in such situations. In the present study, we demonstrated that uridine metabolism is disturbed in sepsis-induced ALI, as manifested by the upregulation of UPP1, which meant to degrade uridine. Subsequently, we discovered that uridine supplementation is sufficient to mitigate sepsis-induced ALI via suppressing ferroptosis of macrophage. Mechanistically, uridine treatment promoted Nrf2 expression and resultant increase of Nrf2-dependent antioxidative targeted genes, including SLC7A11, GPX4 and HO1. Moreover, ACSL4 was also suppressed by exogenous uridine supplementation, which further reinforced the protective effect of uridine in combating ferroptosis. Consistent with previous studies [17,36], we also uncovered that proinflammatory factors, such as TNF-α, IL-1β and IL6, were evidently reduced by uridine treatment. Based on our results, it is supposed that uridine could be a mighty endogenous anti-inflammatory metabolite and a promising preventive remedy for intervening sepsis-induced ALI (Figure 8). Uridine metabolism plays a pivotal role in cellular biochemical processes, as it is intimately connected to cellular other metabolic processes, such as glucoses, lipids and amino acids homeostasis [10]. Herein, we compared the expression of uridine metabolism-related genes in LPS-induced lung tissue and control to acquire the landscape of uridine metabolism in sepsis-induced ALI. Unexpectedly but reasonably, most uridine metabolism-related genes were unchanged but UPP1 was unraveled to raise in ALI group in both GEO datasets, which was confirmed by experimental ALI model. Further, we validated the results in human blood samples from patients with sepsis. As expected, UPP1 was significantly upregulated in samples from sepsis compared to healthy. These data revealed that uridine metabolism is dysregulated in sepsis-induced ALI and it provided rationale for uridine supplementation to treat sepsis-induced ALI. Actually, a similar situation existed in the context of osteoarthritis, in which UPP1 was upregulated and uridine was decreased, and uridine treatment attenuated the severity of osteoarthritis [37]. Moreover, restraining UPP1 expression to stable uridine homeostasis was reported to prevent progression of liver fibrosis [38,39]. Therefore, we speculated that restoring uridine homeostasis by exogenous supplementation might also be a promising intervention for sepsis-induced ALI. Subsequently, to test the hypothesis, we treated mice with uridine when sepsis-induced ALI was built. Reassuringly, uridine supplementation observably alleviated lung injury in vivo and refrained macrophage inflammation in vitro. This is consistent with previous studies that the systematic inflammation was strongly inhibited by uridine treatment and this further solidified the fact that uridine possesses anti-inflammatory effect [23]. Ferroptosis is a regulated cell death mediated by redox imbalance, which is intimately regulated by Nrf2 and its downstream targeted genes [29]. Several lines of study have revealed that ferroptosis serves a pathogenic role in sepsis-induced ALI [40,41,42]. Uridine was extensively reported to exhibit antioxidant effect and regarded as an antioxidative metabolite [18,43]. For example, it has been covered that uridine converted to uridine diphosphate (UDP) and subsequently activated mitoKATP to suppress ROS production in myocardial tissue [21]. Thus, we explored the relationship between uridine metabolism and ferroptosis under the context of ALI in vivo and vitro. It was found that uridine treatment upregulated Nrf2 and its targeted genes and increased GSH level and reduced lipid peroxidative product MDA in lung tissue and macrophage. Moreover, the iron level in lung tissue was markedly lowered. Furthermore, Nrf2 silence abrogated the protective effect of uridine. Our results suggested that uridine might suppress ferroptosis via activating antioxidant system through boosting Nrf2 expression. Of note, the lipid synthesis gene ACSL4 was found to be suppressed after uridine supplementation. It is well characterized that ACSL4 promotes polyunsaturated fatty acids synthesis and facilitates ferroptosis [44,45]. Our data prompted that LPS stimulation upregulated ACSL4 expression, while was prevented by uridine supplementation. It may be explained that uridine promotes cell energy excess [46], thus, to boost lipidolysis and inhibit lipid synthesis, which is still deserving of investigation. Finally, there are some limitations in this study. In our study, although we disclosed that uridine prevented sepsis-induced ALI via controlling ferroptosis, it should not be ignored that uridine exerted anti-inflammatory effects by means of other pathways, such as nuclear transcription factor-κB (NF-κB) pathway [47]. Moreover, the mechanism for uridine to activate Nrf2 is not explored in this study, which needs further study. Furthermore, we mainly investigated the relationship between uridine and ferroptosis in the perspective of lipid peroxidation, but it is also meaningful to explore the relationship between uridine metabolism and iron metabolism. ## 4.1. Chemical and Reagents Uridine (HY-B1449) was purchased from MedChem Express (Shanghai, China), purity > $99.99\%$. Lipopolysaccharides (LPS) and Escherichia coli O55:B5 (#L2880) was purchased from Sigma-Aldrich (St Louis, MO, USA). Primary antibodies against Beta Actin (66009-1-Ig), SLC7A11 (26864-1-AP), GPX4 (67763-1-Ig), HO-1 (10701-1-AP), Nrf2 (16396-1-AP) were purchased from Proteintech (Wuhan, China). Anti-ASCL4 (A20414) was purchased from Abclonal Technology (Wuhan, China). Secondary antibodies (GB23301, GB23303) were purchased from Servicebio (Wuhan, China). Malondialdehyde (MDA) assay kit (TBA method) (A003-1-2) and tissue iron assay kit (A039-2-1) were purchased from Jiancheng Bioengineering Institute (Nanjing, China). MPO activity assay kit was purchased from Abcam (Cambridge, UK). ELISA kits for Tumor necrosis factor-α (TNF-α), Interleukin-1beta (IL-1β) and Interleukin 6 (IL6) were acquired through Cloud-Clone (Wuhan, China). The DCFH-DA probe, CCK-8 assay kit, JC-1 probe and Glutathione Peroxidase (GSH-PX) assay kit were obtained from Beyotime Biotechnology (Shanghai, China). The C11-BODIPY $\frac{581}{591}$ was purchased from Invitrogen (Waltham, MA, USA). ## 4.2. Establishment of Sepsis-Induced Acute Lung Injury Model All the animal procedures conducted in the present study conformed to the guidelines of the Care and Use of Laboratory Animals and approved by the Animal Use Committees of Renmin Hospital of Wuhan University. Male C57BL/6J mice (6–8 weeks old and 18–21 g weight) were provided by Hubei Province Experimental Animal (Wuhan, Hubei, China) and fed in appropriate temperature and humidity, with free access to standard food and water, in a room with a 12 h light:12 h dark cycle. All animals were raised in a specific-pathogen free (SPF) environment. Mice were randomly divided into four groups: PBS group, uridine group, LPS group and LPS+uridine group. Unless stated otherwise, the number of animals in each group was 6. Intraperitoneal LPS injection (10 mg/kg) was employed to establish sepsis-induced ALI model according to previous study [48]. Uridine dissolved in PBS was intraperitoneally injected as soon as possible after LPS injection. LPS simulation and uridine administration was simultaneously proceeded. After stimulating with LPS 12 h, animals were sacrificed. Bronchoalveolar lavage fluid (BALF) and lung tissues were collected for further analysis. ## 4.3. BALF Collection, Processing and Determination of Protein Concentration For BALF collection, the cervical trachea was exposed, and 1 mL phosphate-buffered sterile saline was instilled into alveoli via the trachea by means of a syringe. After rinsing the lungs three times with the recovered lavage fluid, final BALF was collected. Subsequently, the BALF was centrifuged at 800 g for 10 min, and the protein level in the supernatant was determined by bicinchoninic acid (BCA) protein assay kit (Beyotime, Shanghai, China). ## 4.4. Lung Wet-to-Dry Weight Ratio Measurement To assess lung tissue edema, the right lungs were cleaned and weighed to obtain the wet weight. Then the lungs were kept at 70 °C for 48 h in an incubator. Lungs were then weighed to obtain the dry weight. The wet-to-dry (W/D) weight ratio of the lung was calculated. ## 4.5. H&E Staining and Lung Injury Score Evaluation Left lung tissues were fixed with $10\%$ buffered formalin, embedded in paraffin, and then cut into slices with thickness about 4 μm. Subsequently, the lung sections were stained with hematoxylin and eosin (H&E) (Servicebio, Wuhan, China) and observed under a microscope. The severity of lung injury was scored according to published criteria [49]. ## 4.6. Assay of MPO Activity Myeloperoxidase (MPO) activity assay kit (#ab105136, Abcam) was used to detect the MPO activity of lung tissue homogenate. The absorbance at 412 nm was determined through a microplate reader. ## 4.7. Quantification of Cytokines, Detection of Tissue Iron and Lipid Peroxidation Level TNF-α, IL-1β, and IL-6 were quantified by following the ELISA kit instructions. Tissue iron concentrations determination was implemented by following the specification. The level of lipid peroxidation was measured by MDA assay kit as instructed. ## 4.8. Cell Culture and Transfection The human monocytic leukemia cell line (THP-1), the human-type II cell alveolar epithelial cells (A549) and human umbilical vein endothelial cells (HUVEC) were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). THP-1 cells were cultured in RPMI 1640 culture medium (Servicebio, Wuhan, China). A549 cells were cultured in F12K medium and HUVEC was cultured in DMEM medium with high glucose. All the culture media were supplemented with $10\%$ fetal bovine serum (FBS; Gibco, Waltham, MA, USA) and $1\%$ penicillin/streptomycin (P/S) (Biosharp, Hefei, China). Cells were maintained at 37 °C, $5\%$ CO2 humidified incubator. For Nrf2 knockdown assay, the THP-1 cells were transfected with Nrf2 siRNA (5′-AUUGAUGUUUCUGAUCUAUCACUTT-3′) using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA) according to the instruction. ## 4.9. Macrophage Differentiation and Treatment The THP-1 monocytes were stimulated with 100 ng/mL 1 phorbol 12-myristate 13-acetate (PMA; MCE) for 48 h to acquire macrophage-like state. Then, cells were digested using trypsin and centrifuged and resuspended with culture medium and then seeded into 6-wells cell culture plates (Corning, Corning, NY, USA). After PMA induced, differentiated macrophage-like THP-1 cells were cultured in RPMI 1640 medium without PMA but containing $1\%$ PBS and $10\%$ FBS for 12 h. After that, cells were treated with uridine or LPS or both. The concentrations for uridine were 50 μM, 100 μM, 200 μM in the initial cell experiment for selecting appropriate concentration and eventually 200 μM was selected as the optimum. As for LPS, unless stated, 10 μg/mL was used to stimulate. In experiments, we treated uridine simultaneously after LPS stimulation. ## 4.10. GSH Measurement Glutathione Peroxidase (GSH-PX) assay kit was used to detect the total GSH in cells and tissue according to the instruction. ## 4.11. Cell Viability Assay Cell Counting Kit-8 (CCK-8) assay was used to detect the cell viability. In brief, 10 μL CCK-8 solution was added into the 96-well plates and incubated for an hour at incubator. The absorbance was examined at 450 nm according to CCK8 kit instruction. ## 4.12. RNA Extraction and Real-Time Polymerase Chain Reaction The total RNA of lung tissues and THP-1 cells were extracted using TRIpure Total RNA Extraction Reagent (Biosharp, China), and cDNA was synthesized using cDNA Synthesis Kit (Servicebio, Wuhan, China). Quantitative real-time PCR was performed using SYBR Green PCR SuperMix (Servicebio, China). The relative expression of target genes unified to Actin were calculated. All primers used in this study were listed in Table 1. ## 4.13. Western Blotting Protein extraction from cells or tissues were performed by the lysis of RIPA Lysis Buffer (Servicebio, Wuhan, China) contained with $1\%$ Phenylmethanesulfonyl fluoride (PMSF, Servicebio, Wuhan, China). Protein separation was proceeded on 10–$15\%$ SDS-polyacrylamide gradient gels and PVDF membranes were used for proteins transmembrane. Protein Free Rapid Blocking Buffer (5×) was used to block non-specific binding for 15 min, and membranes were incubated with primary antibodies in 4 °C for overnight. Then, membranes were washed three times with TBST, each time for 10 min. Subsequently, membranes were incubated with anti-rabbit-HRP (1:2000; Servicebio) or anti-mouse-HRP (1:2000; Servicebio, Wuhan, China) in 37 °C for 1 h. The internal reference protein was Actin. The protein bands were exposed by the enhanced chemiluminescence western blotting detection system (Bio-Rad, Hercules, CA, USA). ## 4.14. ROS, Lipid Peroxidation and Mitochondrial Potential Measurement The DCFH-DA probe was used to determine the reactive oxygen species (ROS) level of THP-1 cells. The C11 BODIPY was used to detect the lipid peroxidation level of THP-1 cells. The JC-1 probe was used to detect mitochondrial potential of THP-1 cells. All the operational processes were performed according to instructions and fluorescence intensity was detected by fluorescence microscopy. ## 4.15. Dataset Collection and Preprocessing Uridine metabolism-related genes were collected from previously published literatures. GEO cohorts (GSE130936, GSE2411, GSE15379, GSE1871, GSE185263, GSE211210, GSE137340) and related platform files were downloaded from GEO database (https://www.ncbi.nlm.nih.gov/geo/ (accessed on 15 September 2022)). The probes were transformed to gene symbols via corresponding platform annotation files. GSE130936 and GSE2411 were integrated into a new cohort using the ‘ComBat’ algorithm of ‘sva’ package to reduce the batch effects. GSE15379 and GSE1871 were also doing the same. The information about the GEO cohorts in the present study were listed in Table 2. ## 4.16. Bioinformatic Analysis and Visualization The limma package was used to identify the differentially expressed genes. The ggplot2 package was used to visualize the results from limma through drawing volcano plots. STRING website and Cytoscape software were used to construct protein–protein interaction (PPI) network. Gene ontology (GO) analysis was implemented in Metascape online website. ssGSEA algorithm of GSVA package was used to implement immune infiltration analysis. The Receiver Operating Characteristic (ROC) curves were drawn by pROC package and area under curve (AUC) were calculated. ## 4.17. Statistical Analysis All statistical analyses were performed with GraphPad Prism 9 software and R software. Data were expressed as mean ± standard deviation (SD). Two group comparison were performed using the Student’s t-test with Welch’s correction. While for multiple group comparisons, the one-way ANOVA with the post hoc Tukey test was performed. Unless stated otherwise, the data represent at least three independent experiments. $p \leq 0.05$ was regarded as statistically significant. ## 5. Conclusions In summary, we described the dysregulated uridine metabolism in sepsis-induced ALI for the first time and discovered that uridine supplementation could inhibit ferroptosis of macrophage via Nrf2 pathway and ACSL4 inhibition. It provides a novel insight and rationale for targeting UPP1 by specific inhibitor or uridine supplementation for clinical use to combat sepsis-induced ALI. ## References 1. Butt Y., Kurdowska A., Allen T.C.. Acute Lung Injury: A Clinical and Molecular Review. Arch. Pathol. Lab. Med. (2016) 140 345-350. DOI: 10.5858/arpa.2015-0519-RA 2. Pham T., Rubenfeld G.D.. Fifty Years of Research in ARDS. The Epidemiology of Acute Respiratory Distress Syndrome. A 50th Birthday Review. Am. J. Respir. Crit. Care Med. (2017) 195 860-870. DOI: 10.1164/rccm.201609-1773CP 3. De Freitas Caires N., Gaudet A., Portier L., Tsicopoulos A., Mathieu D., Lassalle P.. Endocan, sepsis, pneumonia, and acute respiratory distress syndrome. Crit. Care (2018) 22 280. DOI: 10.1186/s13054-018-2222-7 4. Sadowitz B., Roy S., Gatto L.A., Habashi N., Nieman G.. Lung injury induced by sepsis: Lessons learned from large animal models and future directions for treatment. Expert Rev. Anti Infect. Ther. (2011) 9 1169-1178. DOI: 10.1586/eri.11.141 5. Kumar V.. Pulmonary Innate Immune Response Determines the Outcome of Inflammation During Pneumonia and Sepsis-Associated Acute Lung Injury. Front. Immunol. (2020) 11 1722. DOI: 10.3389/fimmu.2020.01722 6. Wick K.D., McAuley D.F., Levitt J.E., Beitler J.R., Annane D., Riviello E.D., Calfee C.S., Matthay M.A.. Promises and challenges of personalized medicine to guide ARDS therapy. Crit. Care (2021) 25 404. DOI: 10.1186/s13054-021-03822-z 7. Patel V.J., Biswas Roy S., Mehta H.J., Joo M., Sadikot R.T.. Alternative and Natural Therapies for Acute Lung Injury and Acute Respiratory Distress Syndrome. Biomed. Res. Int. (2018) 2018 2476824. DOI: 10.1155/2018/2476824 8. Connolly G.P., Duley J.A.. Uridine and its nucleotides: Biological actions, therapeutic potentials. Trends. Pharmacol. Sci. (1999) 20 218-225. DOI: 10.1016/S0165-6147(99)01298-5 9. Anderson C.M., Parkinson F.E.. Potential signalling roles for UTP and UDP: Sources, regulation and release of uracil nucleotides. Trends Pharmacol. Sci. (1997) 18 387-392. DOI: 10.1016/S0165-6147(97)90667-2 10. Zhang Y., Guo S., Xie C., Fang J.. Uridine Metabolism and Its Role in Glucose, Lipid, and Amino Acid Homeostasis. Biomed. Res. Int. (2020) 2020 7091718. DOI: 10.1155/2020/7091718 11. Yamamoto T., Koyama H., Kurajoh M., Shoji T., Tsutsumi Z., Moriwaki Y.. Biochemistry of uridine in plasma. Clin. Chim. Acta (2011) 412 1712-1724. DOI: 10.1016/j.cca.2011.06.006 12. Russo R., Marra R., Andolfo I., Manna F., De Rosa G., Rosato B.E., Radhakrishnan K., Fahey M., Iolascon A.. Uridine treatment normalizes the congenital dyserythropoietic anemia type II-like hematological phenotype in a patient with homozygous mutation in the CAD gene. Am. J. Hematol. (2020) 95 1423-1426. DOI: 10.1002/ajh.25946 13. Zhang X., Zhang H., Gong J., Yu H., Wu D., Hou J., Li M., Sun X.. Aging affects the biological activity of fibroblast growth factor (FGF) in gastric epithelial cell, which is partially rescued by uridine. Bioengineered (2022) 13 3724-3738. DOI: 10.1080/21655979.2022.2029066 14. Chenna Narendra S., Chalise J.P., Magnusson M., Uppugunduri S.. Local but Not Systemic Administration of Uridine Prevents Development of Antigen-Induced Arthritis. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0141863 15. Jeengar M.K., Thummuri D., Magnusson M., Naidu V.G.M., Uppugunduri S.. Uridine Ameliorates Dextran Sulfate Sodium (DSS)-Induced Colitis in Mice. Sci. Rep. (2017) 7 3924. DOI: 10.1038/s41598-017-04041-9 16. Cicko S., Grimm M., Ayata K., Beckert J., Meyer A., Hossfeld M., Zissel G., Idzko M., Müller T.. Uridine supplementation exerts anti-inflammatory and anti-fibrotic effects in an animal model of pulmonary fibrosis. Respir. Res. (2015) 16 105. DOI: 10.1186/s12931-015-0264-9 17. Jiang N., Zhao Z.. Intestinal aging is alleviated by uridine via regulating inflammation and oxidative stress in vivo and in vitro. Cell Cycle (2022) 21 1519-1531. DOI: 10.1080/15384101.2022.2055252 18. Adant I., Bird M., Decru B., Windmolders P., Wallays M., de Witte P., Rymen D., Witters P., Vermeersch P., Cassiman D.. Pyruvate and uridine rescue the metabolic profile of OXPHOS dysfunction. Mol. Metab. (2022) 63 101537. DOI: 10.1016/j.molmet.2022.101537 19. Le T.T., Ziemba A., Urasaki Y., Hayes E., Brotman S., Pizzorno G.. Disruption of uridine homeostasis links liver pyrimidine metabolism to lipid accumulation. J. Lipid Res. (2013) 54 1044-1057. DOI: 10.1194/jlr.M034249 20. Ye J., Jin Z., Chen S., Guo W.. Uridine relieves MSCs and chondrocyte senescence in vitvo and exhibits the potential to treat osteoarthritis in vivo. Cell Cycle (2022) 21 33-48. DOI: 10.1080/15384101.2021.2010170 21. Krylova I.B., Selina E.N., Bulion V.V., Rodionova O.M., Evdokimova N.R., Belosludtseva N.V., Shigaeva M.I., Mironova G.D.. Uridine treatment prevents myocardial injury in rat models of acute ischemia and ischemia/reperfusion by activating the mitochondrial ATP-dependent potassium channel. Sci. Rep. (2021) 11 16999. DOI: 10.1038/s41598-021-96562-7 22. Liu Z., Li W., Geng L., Sun L., Wang Q., Yu Y., Yan P., Liang C., Ren J., Song M.. Cross-species metabolomic analysis identifies uridine as a potent regeneration promoting factor. Cell Discov. (2022) 8 6. DOI: 10.1038/s41421-021-00361-3 23. Mironova G.D., Khrenov M.O., Yu T.E., Glushkova O.V., Parfenyuk S.B., Novoselova T.V., Lunin S.M., Belosludtseva N.V., Novoselova E.G., Lemasters J.J.. The role of mitochondrial KATP channel in anti-inflammatory effects of uridine in endotoxemic mice. Arch. Biochem. Biophys. (2018) 654 70-76. DOI: 10.1016/j.abb.2018.07.006 24. Dixon S.J., Lemberg K.M., Lamprecht M.R., Skouta R., Zaitsev E.M., Gleason C.E., Patel D.N., Bauer A.J., Cantley A.M., Yang W.S.. Ferroptosis: An iron-dependent form of nonapoptotic cell death. Cell (2012) 149 1060-1072. DOI: 10.1016/j.cell.2012.03.042 25. Liu X., Zhang J., Xie W.. The role of ferroptosis in acute lung injury. Mol. Cell. Biochem. (2022) 477 1453-1461. DOI: 10.1007/s11010-021-04327-7 26. Pan Y., Wang X., Liu X., Shen L., Chen Q., Shu Q.. Targeting Ferroptosis as a Promising Therapeutic Strategy for Ischemia-Reperfusion Injury. Antioxidants (2022) 11. DOI: 10.3390/antiox11112196 27. Mou Y., Wang J., Wu J., He D., Zhang C., Duan C., Li B.. Ferroptosis, a new form of cell death: Opportunities and challenges in cancer. J. Hematol. Oncol. (2019) 12 34. DOI: 10.1186/s13045-019-0720-y 28. Li S., Wang R., Wang Y., Liu Y., Qiao Y., Li P., Chen J., Pan S., Feng Q., Liu Z.. Ferroptosis: A new insight for treatment of acute kidney injury. Front. Pharmacol. (2022) 13 1065867. DOI: 10.3389/fphar.2022.1065867 29. Dodson M., Castro-Portuguez R., Zhang D.D.. NRF2 plays a critical role in mitigating lipid peroxidation and ferroptosis. Redox Biol. (2019) 23 101107. DOI: 10.1016/j.redox.2019.101107 30. Mao C., Liu X., Zhang Y., Lei G., Yan Y., Lee H., Koppula P., Wu S., Zhuang L., Fang B.. DHODH-mediated ferroptosis defence is a targetable vulnerability in cancer. Nature (2021) 593 586-590. DOI: 10.1038/s41586-021-03539-7 31. Stockwell B.R.. Ferroptosis turns 10: Emerging mechanisms, physiological functions, and therapeutic applications. Cell (2022) 185 2401-2421. DOI: 10.1016/j.cell.2022.06.003 32. Martin M.D., Badovinac V.P., Griffith T.S.. CD4 T Cell Responses and the Sepsis-Induced Immunoparalysis State. Front. Immunol. (2020) 11 1364. DOI: 10.3389/fimmu.2020.01364 33. Jensen I.J., Sjaastad F.V., Griffith T.S., Badovinac V.P.. Sepsis-Induced T Cell Immunoparalysis: The Ins and Outs of Impaired T Cell Immunity. J. Immunol. (2018) 200 1543-1553. DOI: 10.4049/jimmunol.1701618 34. Li R., Li X., Zhao J., Meng F., Yao C., Bao E., Sun N., Chen X., Cheng W., Hua H.. Mitochondrial STAT3 exacerbates LPS-induced sepsis by driving CPT1a-mediated fatty acid oxidation. Theranostics (2022) 12 976-998. DOI: 10.7150/thno.63751 35. Yang L., Xie M., Yang M., Yu Y., Zhu S., Hou W., Kang R., Lotze M.T., Billiar T.R., Wang H.. PKM2 regulates the Warburg effect and promotes HMGB1 release in sepsis. Nat. Commun. (2014) 5 4436. DOI: 10.1038/ncomms5436 36. Zhang L., Li B., Zhang D., Wang Z., Zhao Y., Yu Q.. Uridine alleviates LPS-induced ARDS and improves insulin sensitivity by decreasing oxidative stress and inflammatory processes. Physiol. Int. (2022) 109 215-229. DOI: 10.1556/2060.2022.00169 37. Liu Q., Zhai L., Han M., Shi D., Sun Z., Peng S., Wang M., Zhang C., Gao J., Yan W.. SH2 Domain-Containing Phosphatase 2 Inhibition Attenuates Osteoarthritis by Maintaining Homeostasis of Cartilage Metabolism via the Docking Protein 1/Uridine Phosphorylase 1/Uridine Cascade. Arthritis Rheumatol. (2022) 74 462-474. DOI: 10.1002/art.41988 38. Zheng W.V., Li Y., Cheng X., Xu Y., Zhou T., Li D., Xiong Y., Wang S., Chen Z.. Uridine alleviates carbon tetrachloride-induced liver fibrosis by regulating the activity of liver-related cells. J. Cell. Mol. Med. (2022) 26 840-854. DOI: 10.1111/jcmm.17131 39. Gonçalves da Silva E.F., Costa B.P., Nassr M.T., de Souza Basso B., Bastos M.S., Antunes G.L., Reghelin C.K., Rosa Garcia M.C., Schneider Levorse V.G., Carlessi L.P.. Therapeutic effect of uridine phosphorylase 1 (UPP1) inhibitor on liver fibrosis in vitro and in vivo. Eur. J. Pharmacol. (2021) 890 173670. DOI: 10.1016/j.ejphar.2020.173670 40. Zhang J., Zheng Y., Wang Y., Wang J., Sang A., Song X., Li X.. YAP1 alleviates sepsis-induced acute lung injury via inhibiting ferritinophagy-mediated ferroptosis. Front. Immunol. (2022) 13 884362. DOI: 10.3389/fimmu.2022.884362 41. Wang Y., Chen D., Xie H., Jia M., Sun X., Peng F., Guo F., Tang D.. AUF1 protects against ferroptosis to alleviate sepsis-induced acute lung injury by regulating NRF2 and ATF3. Cell. Mol. Life Sci. (2022) 79 228. DOI: 10.1007/s00018-022-04248-8 42. He R., Liu B., Xiong R., Geng B., Meng H., Lin W., Hao B., Zhang L., Wang W., Jiang W.. Itaconate inhibits ferroptosis of macrophage via Nrf2 pathways against sepsis-induced acute lung injury. Cell Death Discov. (2022) 8 43. DOI: 10.1038/s41420-021-00807-3 43. Krylova I.B., Bulion V.V., Selina E.N., Mironova G.D., Sapronov N.S.. Effect of uridine on energy metabolism, LPO, and antioxidant system in the myocardium under conditions of acute coronary insufficiency. Bull. Exp. Biol. Med. (2012) 153 644-646. DOI: 10.1007/s10517-012-1787-4 44. Li Y., Feng D., Wang Z., Zhao Y., Sun R., Tian D., Liu D., Zhang F., Ning S., Yao J.. Ischemia-induced ACSL4 activation contributes to ferroptosis-mediated tissue injury in intestinal ischemia/reperfusion. Cell Death Differ. (2019) 26 2284-2299. DOI: 10.1038/s41418-019-0299-4 45. Doll S., Proneth B., Tyurina Y.Y., Panzilius E., Kobayashi S., Ingold I., Irmler M., Beckers J., Aichler M., Walch A.. ACSL4 dictates ferroptosis sensitivity by shaping cellular lipid composition. Nat. Chem. Biol. (2017) 13 91-98. DOI: 10.1038/nchembio.2239 46. Liu Y., Xie C., Zhai Z., Deng Z.Y., De Jonge H.R., Wu X., Ruan Z.. Uridine attenuates obesity, ameliorates hepatic lipid accumulation and modifies the gut microbiota composition in mice fed with a high-fat diet. Food Funct. (2021) 12 1829-1840. DOI: 10.1039/D0FO02533J 47. Luo Y., Chen H., Huang R., Wu Q., Li Y., He Y.. Guanosine and uridine alleviate airway inflammation via inhibition of the MAPK and NF-κB signals in OVA-induced asthmatic mice. Pulm. Pharmacol. Ther. (2021) 69 102049. DOI: 10.1016/j.pupt.2021.102049 48. Park I., Kim M., Choe K., Song E., Seo H., Hwang Y., Ahn J., Lee S.H., Lee J.H., Jo Y.H.. Neutrophils disturb pulmonary microcirculation in sepsis-induced acute lung injury. Eur. Respir. J. (2019) 53 1800786. DOI: 10.1183/13993003.00786-2018 49. Li Y., Cao Y., Xiao J., Shang J., Tan Q., Ping F., Huang W., Wu F., Zhang H., Zhang X.. Inhibitor of apoptosis-stimulating protein of p53 inhibits ferroptosis and alleviates intestinal ischemia/reperfusion-induced acute lung injury. Cell Death Differ. (2020) 27 2635-2650. DOI: 10.1038/s41418-020-0528-x
--- title: 'Work-Related Psychosocial Factors and Global Cognitive Function: Are Telomere Length and Low-Grade Inflammation Potential Mediators of This Association?' authors: - Caroline S. Duchaine - Chantal Brisson - Caroline Diorio - Denis Talbot - Elizabeth Maunsell - Pierre-Hugues Carmichael - Yves Giguère - Mahée Gilbert-Ouimet - Xavier Trudel - Ruth Ndjaboué - Michel Vézina - Alain Milot - Benoît Mâsse - Clermont E. Dionne - Danielle Laurin journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049148 doi: 10.3390/ijerph20064929 license: CC BY 4.0 --- # Work-Related Psychosocial Factors and Global Cognitive Function: Are Telomere Length and Low-Grade Inflammation Potential Mediators of This Association? ## Abstract The identification of modifiable factors that could maintain cognitive function is a public health priority. It is thought that some work-related psychosocial factors help developing cognitive reserve through high intellectual complexity. However, they also have well-known adverse health effects and are considered to be chronic psychosocial stressors. Indeed, these stressors could increase low-grade inflammation and promote oxidative stress associated with accelerated telomere shortening. Both low-grade inflammation and shorter telomeres have been associated with a cognitive decline. This study aimed to evaluate the total, direct, and indirect effects of work-related psychosocial factors on global cognitive function overall and by sex, through telomere length and an inflammatory index. A random sample of 2219 participants followed over 17 years was included in this study, with blood samples and data with cognitive function drawn from a longitudinal study of 9188 white-collar workers ($51\%$ female). Work-related psychosocial factors were evaluated according to the Demand–Control–Support and the Effort–Reward Imbalance (ERI) models. Global cognitive function was evaluated with the validated Montreal Cognitive Assessment (MoCA). Telomere length and inflammatory biomarkers were measured using standardised protocols. The direct and indirect effects were estimated using a novel mediation analysis method developed for multiple correlated mediators. Associations were observed between passive work or low job control, and shorter telomeres among females, and between low social support at work, ERI or iso-strain, and a higher inflammatory index among males. An association was observed with higher cognitive performance for longer telomeres, but not for the inflammatory index. Passive work overall, and low reward were associated with lower cognitive performance in males; whereas, high psychological demand in both males and females and high job strain in females were associated with a higher cognitive performance. However, none of these associations were mediated by telomere length or the inflammatory index. This study suggests that some work-related psychosocial factors could be associated with shorter telomeres and low-grade inflammation, but these associations do not explain the relationship between work-related psychosocial factors and global cognitive function. A better understanding of the biological pathways, by which these factors affect cognitive function, could guide future preventive strategies to maintain cognitive function and promote healthy aging. ## 1. Introduction Dementia is emerging as a public health priority due to the aging of the population [1]. In 2020, the Lancet Commission on Dementia Prevention, Intervention, and Care reiterated its position on the relevance of a life-course approach for the identification of modifiable risk factors, targeting those contributing favorably to cognitive reserve [2]. A cognitive reserve refers to the neurobiological maintenance and adaptability of the brain, enabling preservation of cognitive function in everyday activities, despite the presence of brain pathology [3]. Modifiable factors at midlife, such as occupational complexity, frequent social contact, and intellectual and social activities, have been suggested to benefit cognitive reserve [2,3]. The role of work-related psychosocial factors in affecting cognitive function has generated increasing interest in recent years. It has been suggested that these factors could help build cognitive reserve when work involves high intellectual complexity, decision making, creativity, skill discretion and development, and high social interaction with coworkers and supervisors. Systematic reviews and meta-analyses have reported associations between high complexity at work and better cognitive performance [4,5,6], and low risk of dementia [5,6]. However, work-related psychosocial factors that are perceived as stressful may have detrimental health effects [7,8,9]. Chronic psychosocial stressors at work could increase the synthesis of inflammatory biomarkers, including C-reactive protein (CRP) and interleukin-6 (IL-6), which, in turn, could induce low-grade inflammation [10]. This low-grade inflammation promotes the oxidative stress that is involved in the DNA degradation process and telomere shortening [11]. Telomeres are protective and stabilizing structures, made up of repeated non-coding DNA sequences, located at the end of chromosomes [12]. Telomere shortening is part of the natural cellular life cycle, but an increased rate of shortening is associated with premature cellular aging [11]. Both low-grade inflammation and shorter telomeres have been associated with a cognitive decline [13,14,15,16,17,18,19,20] and dementia [21,22,23,24] in longitudinal studies, but these findings have not consistently been replicated [25,26,27,28]. Untangling biological mechanisms by which work-related psychosocial factors affect cognitive function is important to distinguish those that have the potential to increase cognitive reserve from those that could precipitate cognitive decline. Work-related psychosocial factors have been conceptualized by two recognised theoretical models. According to the Demand–Control–Support (DCS) model [29,30], exposure to high psychological demands, combined with low job control, constitutes high job strain, increasing physiological and psychological stress responses, whereas exposure to low psychological demands, combined with high job control, constitutes a low job strain [29]. Exposure to high psychological demands, combined with high job control, defines active work, which provides high developmental and intellectual opportunities, whereas exposure to low psychological demands, combined with low job control, defines passive work [29]. Exposure to low social support at work from colleagues and supervisors, the third component of this model, can amplify the adverse effect of high job strain on health [30]. Based on the Effort–Reward Imbalance (ERI) model [31], exposure to an imbalance between high efforts spent at work and low economic, social, or organisational rewards obtained in return, is associated with a stress condition that has detrimental health effects. Exposures to high job strain and to ERI have both been associated with an increased risk of poorer health outcomes, such as hypertension [32], cardiovascular diseases [33,34], diabetes [35], and mental health problems [7,8,9]. Longitudinal studies provide evidence linking passive work [36,37,38] and low job control [36,37,38,39,40] to worse global cognitive function. The association of high job strain with lower global cognitive function has been documented [36,39], although inconsistently [37,38,40]. Some longitudinal studies have reported associations between work-related psychosocial factors from the DCS or ERI models and higher concentrations of inflammatory biomarkers [41,42,43]. We previously reported an association between exposure to high job strain, combined with low social support or exposure to ERI, and an inflammatory index, especially among males [44]. To our knowledge, no previous longitudinal study has examined the association of these work-related psychosocial factors with telomere length. Additionally, the mediating effects of telomere length or inflammatory biomarkers have not been examined in previous studies, regarding the relationship between work-related psychosocial factors and cognitive function. Furthermore, sex differences have not been thoroughly examined, even though several sex differences have been reported with respect to exposure to work-related psychosocial factors [45], inflammatory biomarkers [46], telomere length [47], cognitive function [48], decline over time [49], and the prevalence of dementia [50]. This study aimed to evaluate the interplay between work-related psychosocial exposures, global cognitive function, telomere length and inflammatory biomarkers, overall and by sex. Specifically, this study examined the total, direct, and indirect effects of work-related psychosocial factors on global cognitive function through telomere length and an inflammatory index, combining CRP and IL-6 within the PROspective Quebec (PROQ) study on work and health [51]. ## 2.1. Study Design and Data Collection The PROQ study is a longitudinal occupational cohort study, designed to evaluate the effects of work-related psychosocial factors on cardiovascular and mental health outcomes. The study population at baseline included 9188 white-collar workers (initial participation: $75\%$; $51\%$ females), recruited in 19 public and semi-public organizations from Quebec City, Canada, 1991–1993 (T1). Two follow-ups were carried out after 8 (T2, 1999–2000) and 24 years (T3, 2015–2018). Methodological details of the study have been described elsewhere [51]. In brief, at each measurement time point, data collection included a self-reported questionnaire, a face-to-face interview, and anthropometric and clinical evaluations. At T3, two new components were added to the main study: one on mental health and the other on biomarkers. Before recontacting participants for T3, one third of all participants at T1 were randomly selected for the study of biomarkers, with the exclusion of 207 participants who had either died between T1 and T2 ($$n = 117$$), did not want to be recontacted for the follow-up ($$n = 77$$) or were not traceable at T3 ($$n = 13$$) (Figure 1). When participants were recontacted at T3, most of them were retired ($77\%$). To ensure sufficient statistical power for the planned analysis, a second random selection was therefore performed, including $50\%$ of participants still working at T3. Finally, 3411 participants were selected for the study of biomarkers. Among those, 959 were not included in the present analysis, because they either died between T2 and T3 ($$n = 233$$), were lost to follow-up ($$n = 82$$), refused to participate in any part of the T3 data collection ($$n = 321$$), declined to participate in the face-to-face interview including the cognitive assessment but granted access to medical data only ($$n = 218$$), or completed the self-reported questionnaire only ($$n = 334$$). Finally, 4 participants were excluded because of missing data on cognitive function, leaving 2219 participants for the present analysis. All participants, or their respondents, provided written informed consent. The CHU de Québec-Université Laval Research Ethics Board reviewed and approved this study. ## 2.2. Work-Related Psychosocial Factors For this study, we used work-related psychosocial factors measured at T2 for baseline exposure, because data from both models were available. Unlike the DCS model, the ERI model was not yet published at T1. Psychological demands (9 items), job control (9 items), and social support at work (11 items) were measured using the validated French adaptation of the Karasek questionnaire [29,30,52,53,54]. Reward was measured using 9 of the 11 items from the validated French adaptation of the Siegrist ERI questionnaire [31,55,56]. Each item was measured on a four-point Likert scale, and scores were calculated for each factor by summing the items. The scores for psychological demands and job control were dichotomised according to the median of a representative sample of Quebec workers [57], and four categories were created: high job strain (high demands and low control), low job strain (low demands and high control), active work (high demands and high control), and passive work (low demands and low control) [29]. To evaluate exposure to high job strain, the reference group comprised the three other categories, and the same procedure was used to evaluate the exposure to passive work. The score of social support at work was dichotomized at the median of the study sample at T2. Exposure to iso-strain was defined by the exposure to high job strain, combined with the exposure to lower-than-median social support at work. The ERI ratio was created by dividing the score of psychological demands, as a proxy of the Siegrist effort scale, by the score of reward. The psychometric qualities of this ERI version have been demonstrated [58]. Participants with a ratio greater than 1 were categorized as exposed to ERI. ## 2.3. Cognitive Function Global cognitive function was evaluated with the validated Montreal Cognitive Assessment (MoCA) [59], during a standardised face-to-face interview at T3. The MoCA is a 30-point screening test for mild cognitive impairment or cognitive dysfunction, evaluating visuospatial abilities, executive function, short-term memory recall, attention, concentration, working memory, language and orientation in time and space. The sensitivity, specificity, and positive and negative predictive values of MoCA are about $90\%$ compared to a clinical diagnosis of mild cognitive impairment made by a neurologist [59]. The screening performance for mild cognitive impairment of MoCA is superior to that of the Mini-Mental State Examination and other tests for global cognitive function, justifying its use in cognitively healthy adults [59,60,61]. ## 2.4. Blood Samples Non-fasting blood samples were taken at T3 by a trained nurse, following a standardised protocol. Venous blood samples were centrifuged within 24 to 72 h, and aliquots of the serum, plasma and buffy coats were kept frozen at minus 80 °C until measurement. ## 2.5. Telomere Length Telomere length was measured as described by Ennour-Idrissi et al. [ 62], with some modifications. DNA was extracted from leucocytes using the Gentra PureGene Cell Kit (QIAGEN Inc., Canada), according to the manufacturer’s protocol. DNA quality and quantity were assessed using the NanoDrop12000c spectrophotometer (Thermo Scientific, Fisher Scientific Canada). Mean relative telomere length was measured using a quantitative polymerase chain reaction (qPCR) method, first described by Cawthon [63], with slight modifications. In short, both the genetic material of interest (telomere) and the single copy gene human beta-globin (Hbg) were amplified in quadruplicates, and the mean value was calculated. A negative control (no DNA template), reference DNA sample for normalisation between the experiments, and two cell line samples (one low-passage for longer telomeres and one high-passage for shorter telomeres) were run in each batch. Standard curves for telomere and Hbg amplifications were done using this same reference DNA sample: the efficiency was $99\%$ and $91\%$, respectively. The mean cycle threshold (Ct) values were calculated using the three closest values of quadruplicate samples, with exclusion of the fourth value when it fell outside two standard deviations (SD) from the mean [64]. The intra-assay coefficient of variation (CV) of the mean was $1.45\%$ and $1.78\%$, and the inter-assay CV was $2.80\%$ and $2.04\%$ for telomere and Hbg, respectively. The relative T/S ratio was calculated using this formula: relative T/S ratio = 2 − (mean telomere − mean Hbg) of sample − (mean telomere − mean Hbg) of the reference DNA [65]. All assays were performed blinded to the participants’ characteristics and clinical data. The T/S ratio was logarithmically transformed (base 2) for interpretation purposes, such that an increase of one unit of log2 (ratio T/S) can be interpreted as a two-fold longer telomere. ## 2.6. Inflammatory Biomarkers Serum CRP concentrations were obtained by a highly sensitive immunoturbidimetric assay on an automated Modular Roche Diagnostics platform (System ExP Modular Hitachi). The coefficient of variation (CV) for CRP was <$5\%$. Serum concentrations of IL-6 were obtained by ELISA (enzyme-linked immunosorbent assay), according to the maker’s protocol (R&D Systems, Minneapolis, USA), and the fluorescent signal was measured with an automated fluorimeter (Fluoros-Kan, LabSystems). Each sample was analysed in duplicate, and the inter- and intra-batch CVs were $12.7\%$ and $5.3\%$, respectively. CRP and IL-6 concentrations were log-transformed, because their distributions were right skewed. An inflammatory index was created by summing standardised log-concentrations of CRP and IL-6 [44]. A positive score indicated higher inflammation. ## 2.7. Covariates Based on the literature, covariates that might confound or modify the evaluated associations, were identified. Covariates collected at T1 were selected to respect the temporal sequence of events and minimize overadjustment, except for formal education, for which T3 data were used, as this variable was assessed in the numbers of years of schooling during a face-to-face interview. Age, sex, smoking status (never smoked regularly, former regular, current occasional or current regular smokers), alcohol intake (number of drinks per week), physical activity (frequency of 30 min leisure-time physical activity per week), diabetes, cardiovascular diseases (heart diseases, stroke, angina), and social support outside of work (number of confidants, number of helpers, relationship with spouse, and relationship with children) were evaluated using a self-report questionnaire [57]. Height, weight, hip and waist circumferences were measured in-person, using standardised protocols. Body mass index (BMI) (kg/m2) and waist-to-hip ratio were calculated and used as continuous variables. Blood pressure was measured following recognised guidelines [66,67]. Hypertension was defined by either systolic blood pressure >140 mmHg, diastolic blood pressure >90 mmHg [66,67], report of diagnosed hypertension, or use of antihypertensive medication. ## 2.8. Statistical Analysis Associations between each work-related psychosocial factor and telomere length or the inflammatory index, and between telomere length or the inflammatory index and cognitive performance were evaluated using linear regression models with a robust variance estimator, using the generalised estimating equations (GEE) method. The potential effect modification of the inflammatory index in the relationship between telomere length and cognitive performance was tested with an interaction term. The exploration of a potential quadratic interaction in this relationship was tested by adding the square of the inflammatory index and an interaction term between the square of the inflammatory index and telomere length. All models were fitted globally and stratified by sex, while adjusting for the covariates previously described. The same procedure was used to estimate the total effect of each work-related psychosocial factor with global cognitive function, i.e., the total effect of these factors on cognitive performance, including direct and indirect effects through any potential mediator. The Vansteelandt and Daniel method for multiple mediators was used to estimate direct, indirect and residual indirect effects [68]. This method accounts for multiple mediators, even when they are correlated with each other and when the structure of this correlation is unknown. The direct effect represents the association between each work-related psychosocial factor and cognitive performance that is not mediated by the telomere length or the inflammatory index. The indirect effect represents an association between each work-related factor and cognitive performance that is mediated by the telomere length or the inflammatory index, each considered separately. The residual indirect effect represents the association between each work-related factor and cognitive performance that is mediated through the interrelation between the telomere length and the inflammatory index, no matter the structure of this interrelation. This method was chosen given a priori evidence of a strong interrelation between the two mediators [11] that is generally a violation of the default assumptions for standard mediation analysis methods [68]. Standard errors for these estimated direct and indirect effects were obtained by bootstrap with 199 replications. Multiple imputation (MI), using chained equations [69], was conducted for: [1] missing data in T1 variables among 3618 participants (3411 participants randomly selected for the study of biomarkers, plus 207 participants not included in the random selection); [2] missing data in T2 variables among the 3180 participants that were selected for the study of biomarkers and were either present in T2 data collection ($$n = 3087$$), or not present in T2, but present in the in-person data collection at T3 with cognitive data ($$n = 93$$); and [3] missing data in T3 variables among the 2219 participants included in the present study. Sixty imputations were computed, because $60\%$ of the study population had at least one missing datum on a variable [69]. Inverse probability of censoring weights (IPCWs) were computed to correct for the differences in the characteristics between included participants and those lost to follow-up between each measurement time. IPCWs were calculated using predicted values obtained from logistic regressions of the probability of being censored between T1 and T2, and between T2 and T3, according to exposure and specific covariates at T1, and at T1 and T2, respectively [70]. Weights were recomputed for each bootstrap sample. Results from each imputed dataset were combined, producing average estimated effects and Wald-type confidence intervals. For more detailed information about the mediation analysis, MI and IPCW methods, their inclusion in the mediation analysis and the predictors used in each model, see Supplementary Methods for statistical analyses. The procedure to estimate direct and indirect effects was implemented in R version 4.1.3. All other analyses were performed using the SAS software, version 9.4. ## 3.1. Characteristics of the Study Population Among the 8981 eligible participants for follow-up (Figure 1), those randomly selected for the study of biomarkers ($$n = 3411$$) were similar to those not selected ($$n = 5570$$) (Supplementary Materials Table S1). Among the participants randomly selected ($$n = 3411$$), those included in the present analysis ($$n = 2219$$; $51.1\%$ females) were more likely to have a university degree and were less likely to be office workers, regular smokers, or to present with a diagnosis of diabetes or hypertension compared to those not included ($$n = 1192$$ (959 + 233 deceased), Figure 1 and Supplementary Materials Table S2). The mean follow-up time from T2 to T3 was 16.8 (standard deviation: 1.4) years and the mean age of participants at T2 (baseline for exposure) was 46.5 (7.9) years (Table 1). Compared to males, females were slightly younger, less educated and more often exposed to passive work, high job strain, low job control and iso-strain. The exposures to high psychological demands, low social support at work and ERI were similar for males and females. At T3, the mean values for education and the MoCA score were 15.3 (2.9) years completed and 25.6 (2.6), respectively. ## 3.2. Longitudinal Associations of Work-Related Psychosocial Factors with Telomere Length and Inflammatory Index Associations were observed between exposure to passive work or low job control and shorter telomeres (β = −0.04, $95\%$ CI −0.08 to −0.00 and β = −0.04, −0.07 to 0.00, respectively, Table 2). These same associations were stronger in females (β = −0.08, $95\%$ CI −0.13 to −0.03 and β = −0.08, −0.13 to −0.02, respectively). No statistically significant associations between other work-related psychosocial factors and telomere length were observed overall or in females, and no statistically significant associations were observed in males. Regarding the inflammatory index, statistically significant associations were observed between low social support at work, ERI or iso-strain, and a higher inflammatory index among males only (β = 0.17, $95\%$ CI 0.03 to 0.31, β = 0.14, 0.00 to 0.29, and β = 0.32, 0.10 to 0.55, respectively). No statistically significant associations with other work-related psychosocial factors and the inflammatory index were observed in males, and no statistically significant associations were observed overall or in females. ## 3.3. Cross-Sectional (T3) Associations of Telomere Length and Inflammatory Index with Cognitive Function An association was observed between longer telomeres and higher cognitive performance (β = 0.26, $95\%$ CI −0.00 to 0.52, $$p \leq 0.0527$$, Table 3). This association was stronger in males (β = 0.50, $95\%$ CI 0.15 to 0.86). No statistically significant associations were observed between the inflammatory index and cognitive function, overall or by sex. Exploration of the effect modification of the inflammatory index on the association between telomeres and cognitive function showed a quadratic interaction and a U-shaped modifying association (Supplementary Materials Figure S1). The association between telomere length and cognitive performance was positive for participants with a low and high inflammatory index, but null or negative for middle values of the inflammatory index. ## 3.4. Mediating Effects of Telomere Length and Inflammatory Index Overall, total and direct associations were observed between passive work and a lower cognitive performance (βtotal = −0.30, $95\%$ CI −0.55 to −0.06, and βdirect = −0.29, −0.54 to −0.04, Table 4), and between high psychological demands and higher cognitive performance (βtotal = 0.23, $95\%$ CI 0.03 to 0.44, and βdirect = 0.23, 0.02 to 0.43). In stratified analyses, these associations were driven by male participants (βtotal = −0.40, $95\%$ CI −0.78 to −0.02, and βdirect = −0.40, 0.78 to −0.03 for passive work; βtotal = 0.38, $95\%$ CI 0.09 to 0.67, and βdirect = 0.39, 0.09 to 0.68 for high demands). Total and direct associations were also observed between high job strain and higher cognitive performance in females (βtotal and direct = 0.34, $95\%$ CI 0.01 to 0.68), and between low reward and lower cognitive performance in males (βtotal = −0.35, $95\%$ CI −0.69 to −0.01 and βdirect = −0.32, −0.67 to 0.03). None of these associations were mediated by telomere length or the inflammatory index. No statistically significant total effect or direct effect were observed with low job control, low social support at work, ERI, or iso-strain. No statistically significant indirect or residual effects were detected overall, or by sex. Figure 2 illustrates an example of the total, direct, indirect, and residual effects through telomere length and inflammatory index, between passive work and cognitive function. ## 4.1. Interpretation of Results In this 17-year longitudinal study of more than 2000 white-collar workers, it was found that passive work in the study sample and low reward in males were associated with a poorer cognitive performance, whereas high psychological demands in the study sample and high job strain in females were associated with a better cognitive performance. However, none of these associations were mediated by telomere length or the inflammatory index. To our knowledge, this is the first study evaluating the indirect effects of work-related psychosocial factors on global cognitive function through biological mechanisms, including telomere length and low-grade inflammation. This is also the first longitudinal study evaluating the association between work-related psychosocial factors and telomere length. Only two studies [41,71] had previously evaluated the possible mediating effects of inflammatory biomarkers on the relationship between work-related psychosocial factors and other health outcomes associated with cognitive function [2,72]. The first study was conducted among a subsample of 2101 white-collar workers from the Whitehall II study, aged 50 years on average at the time of exposure measurement, and followed for 10 years [41]. This study reported a weak indirect effect for the association between low social support at work and diabetes among females though IL-6 plasma concentration. The second was a cross-sectional study conducted among a small sample of 204 young male adult workers from veterinary, agricultural, textile or poultry industries in Jordan [71]. This study found that the association between exposure to ERI and the metabolic syndrome was mediated by CRP serum concentrations. The fact that we found no indirect effects through these biomarkers suggests that other biological pathways may be involved in the association between work-related psychosocial factors and cognitive function. The associations found with exposure to passive work and poorer cognitive performance, or with exposure to high psychological demand and better cognitive performance, are aligned with the cognitive reserve theory [73]. A higher level of education, as well as activities with greater intellectual complexity, or employment that requires more complex skills, may promote the development and the efficiency of the brain, enhancing cognitive reserve and protecting against premature cognitive decline [74]. Passive work is characterized by work with a low cognitive stimulation, in terms of low psychological demands, repetitive work, and low skill utilization and development (i.e., low job control), and is thus less favorable for the preservation of cognitive reserve. A systematic review with meta-analysis reported that exposure to high complexity at work was associated with a lower risk of dementia compared to low complexity at work, exposure to passive work or job strain was associated with a faster cognitive decline compared to active work, and that exposure to high psychological demands was associated with a slower cognitive decline compared to low psychological demands [6]. Moreover, longitudinal studies that have evaluated the association between exposure to passive or active work and cognitive function or dementia, found similar associations as those observed in the present study. In these studies, associations were observed with exposure to passive work and lower global cognitive function [36], lower performance on neuropsychological tests [75,76], or higher risk of dementia [77,78]. Two longitudinal studies also found protective associations between active work and global cognitive function [37] or dementia [79]. Negative findings were reported in two longitudinal studies for the association between passive work and global cognitive function [39] or annual decline of neuropsychological scores [80]. For the most part, these results support the cognitive reserve theory as a plausible explanation for the effect of work-related psychosocial factors on long-term cognitive function. Although mediation effects were not supported in the current study, associations were observed between specific work-related psychosocial factors and shorter telomeres, especially among females, along with a higher inflammatory index, especially among males, suggesting that these factors can increase systemic low-grade inflammation and oxidation. Only two previous cross-sectional studies have examined the association between these factors and telomere length and their results are difficult to compare to ours, because neither examined potential sex differences [81,82]. The first study was conducted among a sample of 435 workers, aged 61 years on average, from the Multiethnic Study of Atherosclerosis (MESA) study [81]. No association was found between exposure to high job strain, active work, passive work, high psychological demands or low job control evaluated (with a validated questionnaire and a job exposure matrix) and telomere length measured with the T/S ratio. The second study was conducted among a sample of 141 employees working in geriatric care units in Germany, aged 44 years on average [82]. No association was found between exposure to high psychological demands, low job control or low social support at work (evaluated with a validated questionnaire) and telomere length (expressed as relative length compared to albumin DNA). In our study, no association was found between high job strain, high psychological demands, and low social support at work, and telomere length. However, exposure to passive work or low job control was associated with shorter telomeres, especially among females. The current findings suggest that some associations between exposures and outcomes varies according to biological sex. Females tend to be more exposed to psychosocial stressors at work [45] and have longer telomeres [47]. Moreover, sex differences have been suggested in the effect of chronic stress on the inflammatory and oxidative response to these exposures [46]. In another sample from the MESA study conducted among 1029 participants, sex differences were observed in the association between chronic stress exposure and telomere shortening, using the T/S ratio [83]. Similar to our findings here, the association between greater chronic stress and telomere shortening was stronger among females than males [83]. These results suggested that some differences between males and females could be present, but the body of knowledge is still insufficient to draw solid conclusions. This study is the first to observe a U-shaped modifying effect of inflammatory biomarkers in the association between telomere length and global cognitive function. One previous longitudinal study conducted among a sample of 497 young adult residents of Jerusalem evaluated the modifying effect of an inflammatory index, combining CRP, fibrinogen and a number of leukocytes on the association between telomere shortening over 5 to 10 years, and global cognitive function evaluated with a neurocognitive test battery [84]. The authors reported an association between faster telomere shortening and poorer cognitive function, but the association was not mediated by inflammation. Nevertheless, considering that telomere shortening and low-grade inflammation are biologically interrelated [11], this effect modification needs to be explored in future studies. ## 4.2. Strengths and Limitations This study has several strengths, including a large sample size, with similar number of males and females for the evaluation of sex differences, and a longitudinal design. Work-related psychosocial factor exposure was measured with validated instruments and well before the assessment of cognitive function and biomarkers. Rigorous and innovative statistical methods were employed to evaluate the indirect effects of telomeres and inflammation in the presence of correlation and interaction between these mediators, and to correct for the potential selection bias. Moreover, several potential confounders were measured before exposure assessment and were controlled for in statistical models, which minimized the potential of overadjustment by mediators. This study also has limitations. First, telomere length, inflammatory biomarkers and global cognitive function were all measured at the end of follow-up. Thus, causal inference about the associations observed between these variables cannot be made. Variation in telomere length and in inflammatory biomarkers can either be a cause or a consequence of the variation in global cognitive function. However, as several previous longitudinal studies have observed associations between inflammatory biomarkers or telomere length and cognitive function [13,14,15,16,17,18,19,20] or dementia [21,22,23,24], a causal hypothesis is supported by a priori evidence. Second, cognitive function was not evaluated at the beginning of the study, thus it was not possible to control for cognitive performance at baseline and change in cognition over time could not be determined. However, the fact that participants were relatively young, with only $4\%$ aged 65 years or older, and that $90\%$ were still working at T2, does not favor the presence of cognitive impairment at T2, although it cannot be completely ruled out. Third, participants may have changed jobs and working conditions during their career over the 17-year follow-up. We do not have this information and misclassification of the exposure over time is possible. However, due to the good working conditions generally offered to white-collar workers, our study population tended to maintain their occupational position over time [44]. Nevertheless, the possibility that exposed participants may have been more likely to quit or change their job cannot be excluded. This would result in a potential selection bias, that generally tends to underestimate the associations. Fourth, the healthy worker effect could be present, as healthier workers and those less exposed to work-related psychosocial factors tend to stay longer in the labour force [85,86], and if so, could lead to an underestimation of the observed associations [85,86,87]. While statistical analyses partially controlled for the healthy worker effect during the course of the follow-up, using IPCW for death and loss to follow-up, underestimation remains possible [87]. Fifth, multiple testing may have increased the risk of false discovery, although all comparisons were determined a priori and were scientifically justified. Finally, the study population was composed of white-collar workers and mostly Caucasians, and thus may not be representative of the entire workforce population. Caution should be exercised in generalizing results. However, participants held a wide variety of jobs, such as office workers, technicians, professionals, and managers, ensuring exposure diversity, while controlling for the physicochemical exposures present in blue-collar workers. White-collar workers constitute the largest segment of the Canadian workforce, a fact which contributes to increasing the scope of the results. ## 5. Conclusions This study suggests that some work-related psychosocial factors could be associated with shorter telomeres and low-grade inflammation, but these associations do not explain the relationship between work-related psychosocial factors and global cognitive function. Further research on the biological pathways by which these factors affect cognitive function could help identify which specific work-related psychosocial factors with positive effects could be targeted for enhancement, with the potential of increasing cognitive reserve. Similarly, factors with negative effects could be targeted for reduction, in order to prevent systemic low-grade inflammation, oxidation, and cellular aging. Interventions already exist to reduce occupational exposures. Work-related psychosocial factors, along with other modifiable lifestyle risk factors, including physical activity, diet, and vascular health, could be part of future prevention strategies aimed at maintaining cognitive function and promoting healthy aging. ## References 1. 1. Alzheimer’s Disease International From Plan to Impact IV: Progress towards Targets of the WHO Global Action Plan on DementiaAlzheimer’s Disease InternationalLondon, UK2021. From Plan to Impact IV: Progress towards Targets of the WHO Global Action Plan on Dementia (2021) 2. Livingston G., Huntley J., Sommerlad A., Ames D., Ballard C., Banerjee S., Brayne C., Burns A., Cohen-Mansfield J., Cooper C.. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet (2020) 396 413-446. DOI: 10.1016/S0140-6736(20)30367-6 3. Stern Y., Arenaza-Urquijo E.M., Bartrés-Faz D., Belleville S., Cantilon M., Chetelat G., Ewers M., Franzmeier N., Kempermann G., Kremen W.S.. Whitepaper: Defining and investigating cognitive reserve, brain reserve, and brain maintenance. Alzheimer’s Dement. J. Alzheimer’s Assoc. (2020) 16 1305-1311. DOI: 10.1016/j.jalz.2018.07.219 4. Nexo M.A., Meng A., Borg V.. Can psychosocial work conditions protect against age-related cognitive decline? Results from a systematic review. Occup. Environ. Med. (2016) 73 487-496. DOI: 10.1136/oemed-2016-103550 5. Then F.S., Luck T., Luppa M., Thinschmidt M., Deckert S., Nieuwenhuijsen K., Seidler A., Riedel-Heller S.G.. Systematic review of the effect of the psychosocial working environment on cognition and dementia. Occup. Environ. Med. (2014) 71 358-365. DOI: 10.1136/oemed-2013-101760 6. Huang L.Y., Hu H.Y., Wang Z.T., Ma Y.H., Dong Q., Tan L., Yu J.T.. Association of Occupational Factors and Dementia or Cognitive Impairment: A Systematic Review and Meta-Analysis. J. Alzheimers Dis. (2020) 78 217-227. DOI: 10.3233/JAD-200605 7. Rugulies R., Aust B., Madsen I.E.. Effort-reward imbalance at work and risk of depressive disorders. A systematic review and meta-analysis of prospective cohort studies. Scand. J. Work Environ. Health (2017) 43 294-306. DOI: 10.5271/sjweh.3632 8. Duchaine C.S., Aubé K., Gilbert-Ouimet M., Vézina M., Ndjaboué R., Massamba V., Talbot D., Lavigne-Robichaud M., Trudel X., Pena-Gralle A.B.. Psychosocial Stressors at Work and the Risk of Sickness Absence Due to a Diagnosed Mental Disorder: A Systematic Review and Meta-analysis. JAMA Psychiatry (2020) 77 842-851. DOI: 10.1001/jamapsychiatry.2020.0322 9. Madsen I.E.H., Nyberg S.T., Magnusson Hanson L.L., Ferrie J.E., Ahola K., Alfredsson L., Batty G.D., Bjorner J.B., Borritz M., Burr H.. Job strain as a risk factor for clinical depression: Systematic review and meta-analysis with additional individual participant data. Psychol. Med. (2017) 47 1342-1356. DOI: 10.1017/S003329171600355X 10. Straub R.H., Cutolo M.. Psychoneuroimmunology-developments in stress research. Wien. Med. Wochenschr. (2018) 168 76-84. DOI: 10.1007/s10354-017-0574-2 11. Zhang J., Rane G., Dai X., Shanmugam M.K., Arfuso F., Samy R.P., Lai M.K., Kappei D., Kumar A.P., Sethi G.. Ageing and the telomere connection: An intimate relationship with inflammation. Ageing Res. Rev. (2016) 25 55-69. DOI: 10.1016/j.arr.2015.11.006 12. Blackburn E.H.. Telomere states and cell fates. Nature (2000) 408 53-56. DOI: 10.1038/35040500 13. Devore E.E., Prescott J., De Vivo I., Grodstein F.. Relative telomere length and cognitive decline in the Nurses’ Health Study. Neurosci. Lett. (2011) 492 15-18. DOI: 10.1016/j.neulet.2011.01.041 14. Yaffe K., Lindquist K., Kluse M., Cawthon R., Harris T., Hsueh W.C., Simonsick E.M., Kuller L., Li R., Ayonayon H.N.. Telomere length and cognitive function in community-dwelling elders: Findings from the Health ABC Study. Neurobiol. Aging (2011) 32 2055-2060. DOI: 10.1016/j.neurobiolaging.2009.12.006 15. Pudas S., Josefsson M., Nordin Adolfsson A., Landfors M., Kauppi K., Veng-Taasti L.M., Hultdin M., Adolfsson R., Degerman S.. Short leukocyte telomeres, but not telomere attrition rates, predict memory decline in the 20-year longitudinal Betula study. J. Gerontol. A Biol. Sci. Med. Sci. (2020) 76 955-963. DOI: 10.1093/gerona/glaa322 16. West N.A., Kullo I.J., Morris M.C., Mosley T.H.. Sex-specific associations of inflammation markers with cognitive decline. Exp. Gerontol. (2020) 138 110986. DOI: 10.1016/j.exger.2020.110986 17. Lindbergh C.A., Casaletto K.B., Staffaroni A.M., Elahi F., Walters S.M., You M., Neuhaus J., Rivera Contreras W., Wang P., Karydas A.. Systemic Tumor Necrosis Factor-Alpha Trajectories Relate to Brain Health in Typically Aging Older Adults. J. Gerontol. A Biol. Sci. Med. Sci. (2020) 75 1558-1565. DOI: 10.1093/gerona/glz209 18. Walker K.A., Gottesman R.F., Wu A., Knopman D.S., Gross A.L., Mosley T.H., Selvin E., Windham B.G.. Systemic inflammation during midlife and cognitive change over 20 years: The ARIC Study. Neurology (2019) 92 e1256-e1267. PMID: 30760633 19. Zheng F., Xie W.. High-sensitivity C-reactive protein and cognitive decline: The English Longitudinal Study of Ageing. Psychol. Med. (2018) 48 1381-1389. DOI: 10.1017/S0033291717003130 20. Fiocco A.J., Krieger L., D’Amico D., Parrott M.D., Laurin D., Gaudreau P., Greenwood C., Ferland G.. A systematic review of existing peripheral biomarkers of cognitive aging: Is there enough evidence for biomarker proxies in behavioral modification interventions? An initiative in association with the nutrition, exercise and lifestyle team of the Canadian Consortium on Neurodegeneration in Aging. Ageing Res. Rev. (2019) 52 72-119. PMID: 31059801 21. Kuo H.K., Yen C.J., Chang C.H., Kuo C.K., Chen J.H., Sorond F.. Relation of C-reactive protein to stroke, cognitive disorders, and depression in the general population: Systematic review and meta-analysis. Lancet Neurol. (2005) 4 371-380. DOI: 10.1016/S1474-4422(05)70099-5 22. Koyama A., O’Brien J., Weuve J., Blacker D., Metti A.L., Yaffe K.. The role of peripheral inflammatory markers in dementia and Alzheimer’s disease: A meta-analysis. J. Gerontol. A Biol. Sci. Med. Sci. (2013) 68 433-440. DOI: 10.1093/gerona/gls187 23. Dziedzic T.. Systemic inflammatory markers and risk of dementia. Am. J. Alzheimers Dis. Other Demen. (2006) 21 258-262. DOI: 10.1177/1533317506289260 24. Fuchs T., Trollor J.N., Crawford J., Brown D.A., Baune B.T., Samaras K., Campbell L., Breit S.N., Brodaty H., Sachdev P.. Macrophage inhibitory cytokine-1 is associated with cognitive impairment and predicts cognitive decline–the Sydney Memory and Aging Study. Aging Cell (2013) 12 882-889. DOI: 10.1111/acel.12116 25. Zhan Y., Clements M.S., Roberts R.O., Vassilaki M., Druliner B.R., Boardman L.A., Petersen R.C., Reynolds C.A., Pedersen N.L., Hagg S.. Association of telomere length with general cognitive trajectories: A meta-analysis of four prospective cohort studies. Neurobiol. Aging (2018) 69 111-116. DOI: 10.1016/j.neurobiolaging.2018.05.004 26. Colicino E., Wilson A., Frisardi M.C., Prada D., Power M.C., Hoxha M., Dioni L., Spiro A., Vokonas P.S., Weisskopf M.G.. Telomere Length, Long-Term Black Carbon Exposure, and Cognitive Function in a Cohort of Older Men: The VA Normative Aging Study. Environ. Health Perspect. (2017) 125 76-81. DOI: 10.1289/EHP241 27. Harris S.E., Marioni R.E., Martin-Ruiz C., Pattie A., Gow A.J., Cox S.R., Corley J., von Zglinicki T., Starr J.M., Deary I.J.. Longitudinal telomere length shortening and cognitive and physical decline in later life: The Lothian Birth Cohorts 1936 and 1921. Mech. Ageing Dev. (2016) 154 43-48. DOI: 10.1016/j.mad.2016.02.004 28. Legdeur N., Heymans M.W., Comijs H.C., Huisman M., Maier A.B., Visser P.J.. Age dependency of risk factors for cognitive decline. BMC Geriatr. (2018) 18. DOI: 10.1186/s12877-018-0876-2 29. Karasek R.. Job-demands, job decision latitude, and mental strain: Implications for job redesign. Adm. Sci. Q. (1979) 24 285-308. DOI: 10.2307/2392498 30. Johnson J.V.. Theoretical developments in psychosocial work environment research. Int. J. Health Serv. (1989) 19 457-458. DOI: 10.2190/DHDH-WUA4-3P0B-XVT9 31. Siegrist J.. Adverse health effects of high-effort/low-reward conditions. J. Occup. Health Psychol. (1996) 1 27-41. DOI: 10.1037/1076-8998.1.1.27 32. Gilbert-Ouimet M., Trudel X., Brisson C., Milot A., Vézina M.. Adverse effects of psychosocial work factors on blood pressure: Systematic review of studies on demand-control-support and effort-reward imbalance models. Scand. J. Work Environ. Health (2014) 40 109-132. DOI: 10.5271/sjweh.3390 33. Sara J.D., Prasad M., Eleid M.F., Zhang M., Widmer R.J., Lerman A.. Association Between Work-Related Stress and Coronary Heart Disease: A Review of Prospective Studies Through the Job Strain, Effort-Reward Balance, and Organizational Justice Models. J. Am. Heart Assoc. (2018) 7 e008073. DOI: 10.1161/JAHA.117.008073 34. Huang Y., Xu S., Hua J., Zhu D., Liu C., Hu Y., Liu T., Xu D.. Association between job strain and risk of incident stroke: A meta-analysis. Neurology (2015) 85 1648-1654. DOI: 10.1212/WNL.0000000000002098 35. Pena-Gralle A.P.B., Talbot D., Duchaine C.S., Lavigne-Robichaud M., Trudel X., Aubé K., Gralle M., Gilbert-Ouimet M., Milot A., Brisson C.. Job strain and effort-reward imbalance as risk factors for type 2 diabetes mellitus: A systematic review and meta-analysis of prospective studies. Scand. J. Work Environ. Health (2022) 48 5-20. DOI: 10.5271/sjweh.3987 36. Pan K.Y., Xu W., Mangialasche F., Dekhtyar S., Fratiglioni L., Wang H.X.. Working Life Psychosocial Conditions in Relation to Late-Life Cognitive Decline: A Population-Based Cohort Study. J. Alzheimers Dis. (2019) 67 315-325. DOI: 10.3233/JAD-180870 37. Andel R., Crowe M., Kareholt I., Wastesson J., Parker M.G.. Indicators of job strain at midlife and cognitive functioning in advanced old age. J. Gerontol. B Psychol. Sci. Soc. Sci. (2011) 66 287-291. DOI: 10.1093/geronb/gbq105 38. Duchaine C.S., Brisson C., Talbot D., Gilbert-Ouimet M., Trudel X., Vézina M., Milot A., Diorio C., Ndjaboué R., Giguère Y.. Cumulative exposure to psychosocial stressors at work and global cognitive function: The PROspective Quebec Study on Work and Health. Occup. Environ. Med. (2021) 78 884-892. DOI: 10.1136/oemed-2021-107407 39. Dong L., Eaton W.W., Spira A.P., Agnew J., Surkan P.J., Mojtabai R.. Job strain and cognitive change: The Baltimore Epidemiologic Catchment Area follow-up study. Occup. Environ. Med. (2018) 75 856-862. DOI: 10.1136/oemed-2018-105213 40. Nilsen C., Andel R., Fors S., Meinow B., Darin Mattsson A., Kareholt I.. Associations between work-related stress in late midlife, educational attainment, and serious health problems in old age: A longitudinal study with over 20 years of follow-up. BMC Public Health (2014) 14. DOI: 10.1186/1471-2458-14-878 41. Magnusson Hanson L.L., Virtanen M., Rod N.H., Steptoe A., Head J., David Batty G., Kivimaki M., Westerlund H.. Does inflammation provide a link between psychosocial work characteristics and diabetes? Analysis of the role of interleukin-6 and C-reactive protein in the Whitehall II cohort study. Brain Behav. Immun. (2019) 78 153-160. DOI: 10.1016/j.bbi.2019.01.017 42. Shirom A., Toker S., Berliner S., Shapira I.. The Job Demand-Control-Support Model and stress-related low-grade inflammatory responses among healthy employees: A longitudinal study. Work Stress (2008) 22 138-152. DOI: 10.1080/02678370802180830 43. Coronado J.I.C., Chandola T., Steptoe A.. Allostatic Load and Effort-Reward Imbalance: Associations over the Working-Career. Int. J. Environ. Res. Public Health (2018) 15. DOI: 10.3390/ijerph15020191 44. Duchaine C.S., Brisson C., Talbot D., Gilbert-Ouimet M., Trudel X., Vézina M., Milot A., Diorio C., Ndjaboué R., Giguère Y.. Psychosocial stressors at work and inflammatory biomarkers: PROspective Quebec Study on Work and Health. Psychoneuroendocrinology (2021) 133 105400. DOI: 10.1016/j.psyneuen.2021.105400 45. Hoven H., Siegrist J.. Work characteristics, socioeconomic position and health: A systematic review of mediation and moderation effects in prospective studies. Occup. Environ. Med. (2013) 70 663-669. DOI: 10.1136/oemed-2012-101331 46. Bekhbat M., Neigh G.N.. Sex differences in the neuro-immune consequences of stress: Focus on depression and anxiety. Brain Behav. Immun. (2018) 67 1-12. DOI: 10.1016/j.bbi.2017.02.006 47. Gardner M., Bann D., Wiley L., Cooper R., Hardy R., Nitsch D., Martin-Ruiz C., Shiels P., Sayer A.A., Barbieri M.. Gender and telomere length: Systematic review and meta-analysis. Exp. Gerontol. (2014) 51 15-27. DOI: 10.1016/j.exger.2013.12.004 48. Li R., Singh M.. Sex differences in cognitive impairment and Alzheimer’s disease. Front. Neuroendocrinol. (2014) 35 385-403. DOI: 10.1016/j.yfrne.2014.01.002 49. Levine D.A., Gross A.L., Briceño E.M., Tilton N., Giordani B.J., Sussman J.B., Hayward R.A., Burke J.F., Hingtgen S., Elkind M.S.V.. Sex Differences in Cognitive Decline Among US Adults. JAMA Netw. Open (2021) 4 e210169. DOI: 10.1001/jamanetworkopen.2021.0169 50. 2022 Alzheimer’s disease facts and figures. Alzheimer’s Dement. J. Alzheimer’s Assoc. (2022) 18 700-789. DOI: 10.1002/alz.12638 51. Trudel X., Gilbert-Ouimet M., Milot A., Duchaine C.S., Vézina M., Laurin D., Sultan-Taieb H., Brisson C.. Cohort Profile: The PROspective Quebec (PROQ) Study on Work and Health. Int. J. Epidemiol. (2018) 47 693-693i. DOI: 10.1093/ije/dyy026 52. Brisson C., Blanchette C., Guimont C., Dion J., Moisan J., Vézina M.. Reliability and Validity of the French Version of the 18-item Karasek Job Content Questionnaire. Work Stress (1998) 12 322-336. DOI: 10.1080/02678379808256870 53. Karasek R., Brisson C., Kawakami N., Houtman I., Bongers P., Amick B.. The Job Content Questionnaire (JCQ): An instrument for internationally comparative assessments of psychosocial job characteristics. J. Occup. Health Psychol. (1998) 3 322-355. DOI: 10.1037/1076-8998.3.4.322 54. Niedhammer I.. Psychometric properties of the French version of the Karasek Job Content Questionnaire: A study of the scales of decision latitude, psychological demands, social support, and physical demands in the GAZEL cohort. Int. Arch. Occup. Environ. Health (2002) 75 129-144. DOI: 10.1007/s004200100270 55. Niedhammer I., Siegrist J., Landre M.F., Goldberg M., Leclerc A.. Psychometric properties of the French version of the Effort-Reward Imbalance model. Rev. Epidemiol. Sante Pub. (2000) 48 419-437 56. Siegrist J., Starke D., Chandola T., Godin I., Marmot M., Niedhammer I., Peter R.. The measurement of effort-reward imbalance at work: European comparisons. Soc. Sci. Med. (2004) 58 1483-1499. DOI: 10.1016/S0277-9536(03)00351-4 57. Et votre coeur, ça va?. Rapport de L’enquête Québécoise sur la Santé Cardiovasculaire (1990) 58. Aboa-Éboulé C., Brisson C., Blanchette C., Maunsell E., Bourbonnais R., Abdous B., Vézina M., Milot A., Dagenais G.R.. Effort-reward imbalance at work and psychological distress: A validation study of post-myocardial infarction patients. Psychosom Med. (2011) 73 448-455. DOI: 10.1097/01.PSY.0000399790.82499.d7 59. Nasreddine Z.S., Phillips N.A., Bedirian V., Charbonneau S., Whitehead V., Collin I., Cummings J.L., Chertkow H.. The Montreal Cognitive Assessment, MoCA: A brief screening tool for mild cognitive impairment. J. Am. Geriatr. Soc. (2005) 53 695-699. DOI: 10.1111/j.1532-5415.2005.53221.x 60. Dong Y., Lee W.Y., Basri N.A., Collinson S.L., Merchant R.A., Venketasubramanian N., Chen C.L.. The Montreal Cognitive Assessment is superior to the Mini-Mental State Examination in detecting patients at higher risk of dementia. Int. Psychogeriatr. (2012) 24 1749-1755. DOI: 10.1017/S1041610212001068 61. Pinto T.C.C., Machado L., Bulgacov T.M., Rodrigues-Junior A.L., Costa M.L.G., Ximenes R.C.C., Sougey E.B.. Is the Montreal Cognitive Assessment (MoCA) screening superior to the Mini-Mental State Examination (MMSE) in the detection of mild cognitive impairment (MCI) and Alzheimer’s Disease (AD) in the elderly?. Int. Psychogeriatr. (2018) 31 491-504. DOI: 10.1017/S1041610218001370 62. Ennour-Idrissi K., Tetu B., Maunsell E., Poirier B., Montoni A., Rochette P.J., Diorio C.. Association of Telomere Length with Breast Cancer Prognostic Factors. PLoS ONE (2016) 11. DOI: 10.1371/journal.pone.0161903 63. Cawthon R.M.. Telomere length measurement by a novel monochrome multiplex quantitative PCR method. Nucleic Acids Res. (2009) 37 e21. DOI: 10.1093/nar/gkn1027 64. Yang I.V., Chen E., Hasseman J.P., Liang W., Frank B.C., Wang S., Sharov V., Saeed A.I., White J., Li J.. Within the fold: Assessing differential expression measures and reproducibility in microarray assays. Genome Biol. (2002) 3 research0062. PMID: 12429061 65. Livak K.J., Schmittgen T.D.. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods (2001) 25 402-408. DOI: 10.1006/meth.2001.1262 66. The Joint National Committee on Prevention Detection Evaluation and Treatment of High Blood Pressure. The Sixth Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Arch. Intern. Med. (1997) 157 2413-2446. PMID: 9385294 67. Strugo V., Glew F.J., Davis J., Opie L.H.. Update: Recommendations for human blood pressure determination by sphygmomanometers. Hypertension (1990) 16 594. DOI: 10.1161/01.HYP.16.5.594 68. Vansteelandt S., Daniel R.M.. Interventional Effects for Mediation Analysis with Multiple Mediators. Epidemiology (2017) 28 258-265. DOI: 10.1097/EDE.0000000000000596 69. White I.R., Royston P., Wood A.M.. Multiple imputation using chained equations: Issues and guidance for practice. Stat. Med. (2011) 30 377-399. DOI: 10.1002/sim.4067 70. Hernán M.A., Robins J.M.. Causal Inference: What If. (2020) 71. Almadi T., Cathers I., Chow C.M.. Associations among work-related stress, cortisol, inflammation, and metabolic syndrome. Psychophysiology (2013) 50 821-830. DOI: 10.1111/psyp.12069 72. Misiak B., Leszek J., Kiejna A.. Metabolic syndrome, mild cognitive impairment and Alzheimer’s disease—The emerging role of systemic low-grade inflammation and adiposity. Brain Res. Bull. (2012) 89 144-149. DOI: 10.1016/j.brainresbull.2012.08.003 73. Stern Y.. What is cognitive reserve? Theory and research application of the reserve concept. J. Int. Neuropsychol. Soc. (2002) 8 448-460. DOI: 10.1017/S1355617702813248 74. Stern Y.. Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol. (2012) 11 1006-1012. DOI: 10.1016/S1474-4422(12)70191-6 75. Zhuo L.B., Pei J.J., Yan Z., Yao W., Hao C.F., Wang H.X.. Working life job strain status and cognitive aging in Europe: A 12-year follow-up study. J. Affect. Disord. (2021) 295 1177-1183. DOI: 10.1016/j.jad.2021.08.114 76. Sabbath E.L., Andel R., Zins M., Goldberg M., Berr C.. Domains of cognitive function in early old age: Which ones are predicted by pre-retirement psychosocial work characteristics?. Occup. Environ. Med. (2016) 73 640-647. DOI: 10.1136/oemed-2015-103352 77. Pan K.Y., Xu W., Mangialasche F., Grande G., Fratiglioni L., Wang H.X.. The role of Apolipoprotein E epsilon4 in the association between psychosocial working conditions and dementia. Aging (2020) 12 3730-3746. DOI: 10.18632/aging.102843 78. Wang H.X., Wahlberg M., Karp A., Winblad B., Fratiglioni L.. Psychosocial stress at work is associated with increased dementia risk in late life. Alzheimer’s Dement. J. Alzheimer’s Assoc. (2012) 8 114-120 79. Kivimäki M., Walker K.A., Pentti J., Nyberg S.T., Mars N., Vahtera J., Suominen S.B., Lallukka T., Rahkonen O., Pietiläinen O.. Cognitive stimulation in the workplace, plasma proteins, and risk of dementia: Three analyses of population cohort studies. BMJ (2021) 374 n1804. DOI: 10.1136/bmj.n1804 80. Agbenyikey W., Karasek R., Cifuentes M., Wolf P.A., Seshadri S., Taylor J.A., Beiser A.S., Au R.. Job strain and cognitive decline: A prospective study of the framingham offspring cohort. Int. J. Occup. Environ. Med. (2015) 6 79-94. DOI: 10.15171/ijoem.2015.534 81. Fujishiro K., Diez-Roux A.V., Landsbergis P.A., Jenny N.S., Seeman T.. Current employment status, occupational category, occupational hazard exposure and job stress in relation to telomere length: The Multiethnic Study of Atherosclerosis (MESA). Occup. Environ. Med. (2013) 70 552-560. DOI: 10.1136/oemed-2012-101296 82. Chmelar C., Jorres R.A., Kronseder A., Muller A., Nowak D., Weigl M.. Associations Between Age, Psychosocial Work Conditions, Occupational Well-Being, and Telomere Length in Geriatric Care Professionals: A Mixed-Methods Study. J. Occup. Environ. Med. (2017) 59 949-955. DOI: 10.1097/JOM.0000000000001102 83. Meier H.C.S., Hussein M., Needham B., Barber S., Lin J., Seeman T., Diez Roux A.. Cellular response to chronic psychosocial stress: Ten-year longitudinal changes in telomere length in the Multi-Ethnic Study of Atherosclerosis. Psychoneuroendocrinology (2019) 107 70-81. DOI: 10.1016/j.psyneuen.2019.04.018 84. Cohen-Manheim I., Doniger G.M., Sinnreich R., Simon E.S., Pinchas R., Aviv A., Kark J.D.. Increased attrition of leukocyte telomere length in young adults is associated with poorer cognitive function in midlife. Eur. J. Epidemiol. (2016) 31 147-157. DOI: 10.1007/s10654-015-0051-4 85. Massamba V.K., Talbot D., Milot A., Pearce N., Trudel X., Brisson C.. Assessment of the healthy worker survivor effect in the relationship between psychosocial work-related factors and hypertension. Occup. Environ. Med. (2019) 76 414-421. DOI: 10.1136/oemed-2018-105460 86. Pearce N., Checkoway H., Kriebel D.. Bias in occupational epidemiology studies. Occup. Environ. Med. (2007) 64 562-568. DOI: 10.1136/oem.2006.026690 87. Buckley J.P., Keil A.P., McGrath L.J., Edwards J.K.. Evolving methods for inference in the presence of healthy worker survivor bias. Epidemiology (2015) 26 204-212. DOI: 10.1097/EDE.0000000000000217
--- title: The Potential Chemopreventive Effect of Andrographis paniculata on 1,2-Dimethylhydrazine and High-Fat-Diet-Induced Colorectal Cancer in Sprague Dawley Rats authors: - Tharani Subarmaniam - Rusydatul Nabila Mahmad Rusli - Kokila Vani Perumal - Yoke Keong Yong - Siti Hadizah - Fezah Othman - Khaled Salem - Nurul Husna Shafie - Rosnani Hasham - Khoo Boon Yin - Khairul Kamilah Abdul Kadir - Hasnah Bahari - Zainul Amiruddin Zakaria journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049149 doi: 10.3390/ijms24065224 license: CC BY 4.0 --- # The Potential Chemopreventive Effect of Andrographis paniculata on 1,2-Dimethylhydrazine and High-Fat-Diet-Induced Colorectal Cancer in Sprague Dawley Rats ## Abstract Colorectal cancer (CRC) is responsible for a notable rise in the overall mortality rate. Obesity is found to be one of the main factors behind CRC development. Andrographis paniculata is a herbaceous plant famous for its medicinal properties, particularly in Southeast Asia for its anti-cancer properties. This study examines the chemopreventive impact of A. paniculata ethanolic extract (APEE) against a high-fat diet and 1,2-dimethylhydrazine-induced colon cancer in Sprague Dawley rats. Sprague Dawley rats were administered 1,2-dimethylhydrazine (40 mg/kg, i.p. once a week for 10 weeks) and a high-fat diet (HFD) for 20 weeks to induce colorectal cancer. APEE was administered at 125 mg/kg, 250 mg/kg, and 500 mg/kg for 20 weeks. At the end of the experiment, blood serum and organs were collected. DMH/HFD-induced rats had abnormal crypts and more aberrant crypt foci (ACF). APEE at a dose of 500 mg/kg improved the dysplastic state of the colon tissue and caused a $32\%$ reduction in the total ACF. HFD increased adipocyte cell size, while 500 mg/kg APEE reduced it. HFD and DMH/HFD rats had elevated serum insulin and leptin levels. Moreover, UHPLC-QTOF-MS analysis revealed that APEE was rich in anti-cancer phytochemicals. This finding suggests that APEE has anti-cancer potential against HFD/DMH-induced CRC and anti-adipogenic and anti-obesity properties. ## 1. Introduction In terms of overall mortality rates, colorectal cancer (CRC) ranks third in males and second in females globally [1]. According to the GLOBOCAN database, the mortality rate of CRC among females ($9.5\%$) is higher compared with males ($9.3\%$) in the year 2020 [2]. In Southeast Asia, Malaysia is ranked as having the 3rd highest incidence and mortality of CRC. Specifically, the mortality and incidence of CRC were found to be increasing more among males than females, particularly in those of Chinese ethnicity [3]. The global economic cost of CRC care approaches USD 100 billion, with medical spending alone expected to exceed USD 20 billion [4]. As a result, there is a pressing requirement for a better knowledge of the pathophysiology of colorectal cancer as well as the discovery of new therapeutic methods. There are several factors influencing CRC development such as age, sex, smoking, poor diet, genetics, and obesity [5]. The most prevalent risk factors for colorectal cancer (CRC) development are a high-fat diet (HFD). The diet, which has a high proportion of fat, has been linked to obesity because of its tendency to promote weight gain over time [6]. Recent epidemiological research has concluded that obesity leads to rising morbidity and mortality linked with colorectal cancer [7]. This is because a high-fat diet has been connected to pro-inflammatory potential, raising the risk of CRC. The pro-inflammatory factors that associate with obesity such as adipokines (secreted by adipocytes) and cytokines will cause low-grade inflammation, which provides a favorable environment for cancer tumor growth [8]. Likewise, a previous study showed the adipose tissue secretes high levels of adipokines and cytokines, which aid in inducing CRC in high-fat-diet-fed rats [9]. Surgery, chemotherapy, and radiation are the primary conventional therapies for CRC. These therapies can also be used in combination, depending on the location and course of the disease [10]. Fluorouracil (5-FU) is a fluoropyrimidine chemotherapeutic drug that has been extensively applied to treat a variety of malignant tumors, and particularly CRC, for over 50 years [11,12]. Even though 5-FU is one of the safest chemotherapy medicines, severe side and toxic effects occur in some CRC patients [13]. The problems of conventional chemotherapies are related to their properties as anti-metabolites that disturb the formation of vital proteins, cause subsequent cell degradation, and result in long-term adverse outcomes [14]. Therefore, the urge to develop the safest therapy has drawn researchers’ attention toward medicinal plants. Recently, significant research has been conducted on medicinal plants to understand their properties to cure certain acute diseases such as cancer [15]. Among them, Andrographis paniculata is one of the medicinal plants that has been broadly studied for its anti-cancer properties. Andrographis paniculata (A. paniculata), often identified as “King of Bitters”, is an Acanthaceae family plant [16]. It is grown extensively in southern Asia. Throughout the years, leaves and roots have been utilized in alternative medicine systems for a variety of therapeutic purposes, as a cure for a wide range of diseases, or as a health supplement [17]. Andrographis paniculata has been shown to be a promising potential cure for many diseases, especially cancer. A. paniculata has been found to contain a variety of phytoconstituents such as labdane diterpenoids, quinic acids, flavonoids, noriridoids, and xanthones. The major constituents of Andrographis paniculata are andrographolide, a labnade diterpenoid that has been studied for its chemopreventive properties [17]. For example, research has provided evidence that andrographolide that is isolated from A. paniculata helps to slow down the spread of CRC by inducing cell apoptosis [18]. Moreover, prior research investigated the potential chemopreventive activities of A. paniculata against colorectal cancer. However, there is little understanding concerning the effect of A. paniculata on CRC under HFD conditions. This study’s goal was to examine the anti-cancer properties of A. paniculata ethanolic extract on 1,2-dimethylhydrazine (DMH)-induced colon cancer in Sprague Dawley rats in high-fat diet conditions (Figure 1). ## 2.1. Phytochemical Screening and Identification of A. paniculata Ethanol Extract Table 1 shows a list of the main phytochemical compounds of A. paniculata ethanolic extract, and Figure 2 demonstrates the UHPLC-QTOF-MS base peak intensity (BPI) metabolic profile of the ethanol extract of A. paniculata. A total of six isolated compounds were mainly from diterpenoids, flavonoids, and quinic acid. The 19β-Glucosyl-14-deoxy-11,12-didehydroand-rographoside, 12S-Hydroxyandrographolide, 10-Hydroxyligustroside, and 19β-Glucosyl-14-deoxyandrographoside were from the terpenoids group. From flavonoids, we found the Genistein-7,4′-di-O-β-D-glucoside compound, while 3,4-O-Dicaffeoylquinic acid was from the quinic acid group. ## 2.2. Impact of A. paniculata on Food Intake, Body Weight, Retroperitoneal White Adipose Tissue (RpWAT), Colon Weight, and Colon Length Table 2 displays the food intake and changes in the body weight, colon weight, and colon length of the treated rats. The body weight and RpWAT weight of the DMH/high-fat diet consumed by the rats were significantly lower ($p \leq 0.05$) than the high-fat diet given to the rats. However, there was no notable change in their food intake. Three doses of A. paniculata ethanolic extract had no significant effect on the body weight, RpWAT weight, colon weight, and colon length of the treated rats. ## 2.3. Impact of A. paniculata on the Histopathological Finding of the Colon in Treated Rats The histopathological finding of colon tissue is shown in Figure 3 and Figure 4, which show the histology examination of the colon aberrant crypt foci (ACF) type using H&E staining. The histopathological changes in the colons of the rats consuming a standard and high-fat diet induced by DMH showed an abnormal crypt structure with elongated, enlarged, and stratified nuclei. A reduction in the number of goblet cells and mucin was also found in the colon tissue. In contrast, the colon morphology of the high-fat diet rats revealed a low grade of the abnormal crypt. The histology features of the colons in the treatment group HCAP125 showed slightly enlarged and elongated nuclei, while in the HCAP250 and HCAP500 groups, the colon was found to have a normal crypt with no crowding or stratification of nuclei and mucin depletion. ## 2.4. Impact of A. paniculata on the Overall Number of ACF in the Colon The impact of A. paniculata on DMH and the high-fat-diet-treated ACF development in the rats’ colons are summarized in Table 3. The findings demonstrated that the rats fed with a normal diet did not develop ACF in their colons. Nevertheless, the only high-fat-diet-fed group (H) has ACF development without DMH induction. The DMH/HFD-treated group had significantly higher ACF development compared with the normal diet/DMH-treated group ($p \leq 0.05$). The primary effect analysis revealed that giving A. paniculata ethanolic extract and Fluorouracil significantly decreased the overall number of ACF contrasted with the DMH/HFD-treated rats; 500 mg/kg A. paniculata ethanol extract lowered the overall amount of ACF by $32\%$. A decrease of $19\%$ of ACF was noticed in the treatment with 250 mg/kg A. paniculata ethanol extract, $16\%$ in the treatment with 125 mg/kg A. paniculata ethanol extract, and $13\%$ in the treatment with Fluorouracil. For the number of ACF consisting of one crypt, the treatment with 250 mg/kg A. paniculata ethanol extract significantly differed from the DMH/HFD-treated group. Contrasted with the DMH/HFD group, the treatment with 250 mg/kg A. paniculata ethanol extract generated significant results in ACF having four crypts, while treatment with 125 mg/kg A. paniculata produced substantial results in ACF having more than five crypts. ## 2.5. Impact of A. paniculata on the Retroperitoneal White Adipose Tissue (RpWAT) The H&E staining of the retroperitoneal white adipose tissue is shown in Figure 5. The high-fat diet intake developed an expansion of adipocytes in the HFD group rats compared with the normal-diet-fed rats. Interestingly, A. paniculata ethanolic extract had the same effect as 5FU on reducing the size of the adipocytes. The adipocyte size (area) is demonstrated in Figure 6. The mean area of the high-fat diet rats and DMH/HFD rats was significantly larger than the normal chow diet rats and DMH/normal diet rats. The A. paniculata ethanolic extract in the dose of 125 mg/kg, 500 mg/kg, and the 5-fluorouracil groups reduced the mean adipocyte cell area significantly compared with the HFD/DMH group rats. ## 2.6. Impact of A. paniculata on the Serum Leptin, Adiponectin, and Insulin Concentration Figure 7a shows the impact of A. paniculata on serum leptin levels. The serum leptin level of the high-fat-diet-treated rats was significantly greater than the normal chow diet rats ($p \leq 0.05$). The administration of 5-FU and 500 mg/kg A. paniculata ethanolic extract significantly lowered the serum leptin level compared with DMH/HFD-induced rats. Figure 7b presents the effect of A. paniculata ethanol extract on serum adiponectin levels. A. paniculata ethanol extract did not significantly alter the serum adiponectin concentration. Figure 7c illustrates the impact of A. paniculata ethanol extract on serum insulin levels. A high-fat diet elevated the serum insulin level compared with the normal diet. The A. paniculata ethanol extract of all three doses significantly lowered the insulin level in the serum of the DMH/HFD-treated rats. ## 3. Discussion Colorectal cancer (CRC) contributes to a higher mortality rate worldwide [19]. There are many risk factors for CRC. However, poor lifestyle and diet remain major risk factors for developing CRC. Recently, numerous studies on epidemiology have claimed that there is a solid link between high-fat diet intake and an elevated risk of CRC. Furthermore, many in vivo studies have illustrated that the chemically induced precancerous ACF formation increased due to high-fat diet consumption [20]. Therefore, in this experiment, the rats consumed a high-fat diet along with 1,2-dimethylhydrazine (DMH), which was induced to accelerate the CRC condition. The typical treatment for CRC is surgical, followed by another approach, such as chemotherapy and immunotherapy, according to the disease onset [21]. Many chemopreventive studies have been conducted in the past, especially on the therapeutical plants [22]. Therefore, the existing research was intended to look at the potential chemopreventive impact of A. paniculata against DMH-induced colon cancer in Sprague Dawley rats under a high-fat diet condition. A. paniculata is a herbaceous plant abundant in phytochemicals that help reduce the risk of getting cancer. These phytochemicals are proven to have antioxidant and chemopreventive agents [23,24]. The primary chemical components of A. paniculata are flavonoids, polyphenols, and terpenoids [25]. In this study, six main compounds were isolated from A. paniculata ethanol extract from terpenoids, flavonoids, and quinic acid groups. The primary diterpenoid compounds in A. paniculata are deoxyandrographolide, 14-deoxy-11,12-didehydroandrographide, isoandrographolide, and neoandrographolide [26]. However, the major flavonoids discovered in A. paniculata ethanol extraction are 5-hydroxy-7,8,2′,5′-tetramethoxyflavone, 5-hydroxy-7,8-dimethoxyflavone,5-hydrox-7,8,2′,3′tetramethoxyflavone, 2′-methyl ether, and 7-O-methylwogonin [27]. In a previous study, these terpenoids and flavanoids were exposed as stopping cancer proliferation by provoking apoptosis and cell cycle arrest [28,29,30]. This anti-cancer strategy occurs by activating the tumor suppressors p53 and p21, which inhibit the spread of cancer cells [31]. Moreover, these phytochemicals in ethanol extraction of the A. paniculata were discovered to slow down the oxidation of the cell due to their antioxidant properties. In a prior study, the potential of uptaking the free radical in A. paniculata extract was confirmed by NO, FRAP, and DPPH bioassays [32]. The anti-cancer potential of A. paniculata is directly attributed to the antioxidant activities that it possesses [33]. Weight loss is a common symptom faced by CRC patients. It happens due to cachexia caused by cancer. Cachexia is a condition described as losing body weight, primarily through losing weight in adipose tissue and skeletal muscle. Cachexia is brought on by a number of reasons, including decreased food intake, metabolic alterations, and inflammation [34]. This study observed body weight and RpWAT weight reduction in the DMH/HFD group despite no differences in total food intake. In the previous study, researchers faced the same body weight loss scenario in cancer-induced rats due to cachexia [35]. From the result, we speculate that the unexplainable weight loss of the rats might be due to cancer-associated cachexia. The 1,2-dimethylhydrazine (DMH) is a procarcinogen used to cause aberrant crypt foci (ACF) in rats for CRC studies [36]. Nonetheless, the crypt development is not exclusively a result of DMH injection-induced inflammation. A high-fat diet can lead to obesity, which causes the adipocyte to release more adipokines that cause inflammation [20]. In this study, the histopathological examination of DMH/HFD rats demonstrated abnormal development in the colon tissue matching with previous findings [37]. Furthermore, DMH-induced rats with a high-fat diet had a considerably greater number of ACF than DMH-induced rats with a normal-fat diet. This result ties well with previous studies wherein Guang Ying et al. claimed that DMH-induced rats with a high-fat diet revealed a greater number of ACF than the moderate-fat-diet-treated rats [38]. This result again proves that a high-fat diet has a high-level potential to aggravate existing and initiate colorectal cancer [39]. A. paniculata ethanolic extract significantly improved the dysplastic state of the colon tissue and lowered the total number of ACF, suggesting the anti-cancer effect of A. paniculata against DMH/HFD-inducing rats. A similar result was proposed in the previous study, in which A. paniculata extract reduced the total number ACF and improved colon morphology in AOM-induced CRC rats [14]. With the progress of obesity, adipose tissue will endure tissue restoration in which the adipocyte increases in size and number. Hypertrophy of the adipocyte will lead to overloaded lipids, resulting in fluctuations in hormone secretion. The excess lipid will start to deposit in the organs, such as the pancreas, muscle, and liver [35]. In a recent study, the HFD affects the total area of the adipocyte, and the histological changes were significant. The average sizes of the adipocytes in high-fat diet rats and DMH/HFD rats were higher than those of the normal diet group and DMH/normal diet rats, respectively. Even in the histological examination, the adipocyte size of the HFD group and DMH/HFD group was more significant compared with normal diet and DMH/normal diet group rats. This again demonstrates agreement the previous study that found that HFD increased the adipocyte size [40]. The adipocyte size was reduced in DMH/HFD rats upon administering A. paniculata ethanol extract. This result agrees with that of Ramgopal Mopuri et al. According to their findings, A. paniculata extract may have beneficial effects in the fight against obesity and adipogenesis [41]. Adipocyte tissue secretes the adipokines leptin and adiponectin. Insulin resistance, hyperleptinemia, and diminished adipose-derived adiponectin secretion are associated with adipocyte tissue expansion [42]. Leptin is one of the obesity adipokines that accelerate the proliferation of CRC and also elevates insulin concentration [43,44]. However, adiponectin is negatively correlated with insulin resistance and obese [45,46]. Adiponectin is a type of adipokine that also acts as an anti-inflammatory and insulin-sensitizing agent [47]. In addition, the serum/plasma adiponectin concentration is also inversely related to CRC threat [48]. Previous animal research has demonstrated that leptin and insulin levels were high in rats fed with a high-fat diet, while there were no major differences in the adiponectin level [49]. In this study, the same theory was implied where due to the expansion of adipocytes, the leptin and insulin levels were elevated in the high-fat diet group compared with those in the normal diet group. However, the adiponectin levels were similar between groups. Interestingly, the leptin and insulin levels were regulated by the A. paniculata ethanol extract matching the study by Ding et al. [ 50]. From this perspective, we can conclude that leptin and adiponectin levels directly influence insulin levels in rats. The phytochemicals present in the A. paniculata were again shown to have the potential to regulate the adipokines associated with obesity and CRC [49]. ## 4.1. Plant Material The A. paniculata plant was purchased from Ethno resources Sdn Bhd, Sungai Buloh, Malaysia, and studied at the Herbarium Biodiversity Unit (UBD) at the Institute of Bioscience, University Putra Malaysia. The A. paniculata whole plant was oven dried and ground into powder. The 100 g of A. paniculata dried powder was soaked for 24 h in 1 L of $95\%$ ethanol at normal temperature. The plant material underwent a maceration process three times for each batch to achieve the optimum yield for 3 days. The soaked extract was filtered using filter paper (Whatman, Maidstone, UK, 125 mm) to obtain the crude ethanolic extract. The solvent in the filtered ethanolic extract was evaporated under low pressure at 45 °C using a rotatory evaporator. The remaining solvent content was eliminated by oven drying (45 °C). The final ethanol crude extract was stored at −20 °C for further usage [14]. ## 4.2. High-Fat Diet Preparation The high-fat diet (HFD) had 414 calories per 100 g and consisted of $17\%$ protein, $40\%$ fat, and $43\%$ carbohydrates; the diet contained $6\%$ ghee (Crispo, CrispoTato (M) Sdn Bhd, Kuala Lumpur, Malaysia), $6\%$ corn oil (Vecorn, Yee Lee Corporation Berhad, Kuala Lumpur, Malaysia), $68\%$ standard chow pellet (Gold Coin Feedmills (M) Sdn Bhd, Selangor, Malaysia), and $20\%$ milk powder (Dutch Lady, Dutch Lady Milk Industries Berhad, Selangor, Malaysia). A normal chow pellet has 306.2 calories per 100 grams of protein, $3\%$ fat, and $48.8\%$ carbohydrates [51]. The high-fat diet was freshly made and refrigerated weekly. ## 4.3. Animal Study In the Animal House, Faculty of Medicine and Health Sciences, University Putra Malaysia, 48 healthy male Sprague Dawley (SD) rats (150–200 g) were placed in separate cages. The Animal Care and Use Committee of University Putra Malaysia approved the study. One week was spent acclimating these rats to the normal chow diet and water. After acclimation, the rats were split into eight groups. Each group had six rats. Group N: Standard chow diet + tap water. Group NC: Standard chow diet + DMH. Group H: High-fat diet + tap water. Group HC: High-fat diet + DMH. Group HCF: High-fat diet + DMH + Fluorouracil Group HCAP125: High-fat diet + DMH + 125 mg/kg A. paniculata ethanol extract. Group HCAP250: High-fat diet + DMH + 250 mg/kg A. paniculata ethanol extract. Group HCAP500: High-fat diet + DMH + 500 mg/kg A. paniculata ethanol extract. N indicates normal, C indicates cancer, H indicates high-fat diet, F indicates fluorouracil, and AP indicates A. paniculata. The rats were orally administered a standard chow and high-fat diet for 20 consecutive weeks; 40 mg/kg DMH was injected subcutaneously at the groin region once a week for 10 weeks [19], and 35 mg/kg of fluorouracil (5-FU) was given by intra-peritoneal injection two times a week for 20 weeks. Three different doses of A. paniculata ethanol extract were orally administrated daily for 20 weeks. Every week till the last week of the in vivo experiment, the food intake and body weight were measured [14]. ## 4.4. Chemicals The 1,2-dimethylhydrazine was diluted in 1 mM ethylenediaminetetraacetic acid (EDTA, Sigma Co., Ronkonkoma, NY, USA), and $10\%$ sodium hydroxide was used to change the pH to 6.5. To cause cancer, it was given subcutaneously at a dose of 40 mg/kg once every week for ten weeks [19]; 2 g of 5-Fluorouracil was diluted in 100 mL of $0.9\%$ normal saline to achieve a final concentration of approximately 2000 mg/100 mL (20 mg/mL) and then given intraperitoneally at a dosage of 35 mg/kg twice weekly for 20 weeks [14]. ## 4.5. Phytochemical Screening UHPLC was used to separate the chemical components (ACQUITY UPLC I-Class system from Waters). The separation was carried out with a 40 °C ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm × 1.8 m). Gradient elution of $0.1\%$ formic acid-containing water (A) and acetonitrile (B) was as follows: $0\%$ B; $0.5\%$ B; 16:00 B; 18:00 B; $20\%$ B. One liter was injected at 0.6 mL/min. The UHPLC system was linked to a Waters Vion IMS QTOF hybrid mass spectrometer. The source temperature was 120 °C, the desolvation gas temperature was 550 °C, the desolvation gas flow was 800 L/h, and the cone gas flow was 50 L/h. Nitrogen (>$99.5\%$) was used as the desolvation and cone gas. The HDMSE data were gathered at 0.1 s/scan from 50 to 1500 m/z. During the run, two scans with different collision energies (CE) were alternately acquired: a low-energy (LE) scan with a constant CE of 4 eV and a high-energy (HE) scan with a ramp from 10 to 40 eV. Collision-induced dissociation (CID) gas was utilized, which was $99.999\%$ pure argon [52]. ## 4.6. Biochemical Test The rats’ blood samples were stored in a plain serum tube by puncturing their hearts. For 10 min, the plain tubes were spun at 3000 rpm. For further investigation, the serum was frozen at −80 °C. ELISA kits from Elabscience were used to measure the concentrations of leptin, adiponectin, and insulin in the blood [53]. ## 4.7. Histopathological Examination During the rat sacrifice, $10\%$ formalin was used to fix the colon and adipose tissues for 24 h. After fixation, the colon and adipose tissue were placed in a cassette. The specimen cassettes underwent tissue processing. Next, the cassettes were fixed in paraffin wax and sectioned using a rotatory microtome with a 5 µm thickness. The sliced specimen was placed on slides. The specimen slides were stained with hematoxylin and eosin. The stained specimen slides were covered using slip slides [37]. The histopathology changes were examined by a pathologist. ## 4.8. ACF Counting The stained colon segments were placed on an electric light microscope, and the number of crypt foci was quantified manually on random observation fields at 4× and 10× magnification [54]. ## 4.9. Adipocyte Area Counting The stained adipocyte histology pictures were captured using an electric light microscope at 10× magnification. The image was then uploaded into ImageJ for area counting. A total of 100 individual adipocytes were calculated for each slide manually by using the ImageJ application [55]. ## 4.10. Statistical Analysis The data from the food intake, body weight, organ weight, biochemical test, and adipocyte count were collected. Each experiment was repeated a minimum of three times. Independent samples were evaluated with a one-way ANOVA and subsequently Tukey’s post hoc test using IBM SPSS statistics version 27. $p \leq 0.05$ was significant, and all the values were given as the mean ± SE [39]. ## 5. Conclusions In a nutshell, these findings indicate that HFD worsened the condition of ACF induced by DMH. This happened due to the imbalanced ratio between leptin and adiponectin. Therefore, A. paniculata played an essential role in this study as a potential intervention for colorectal cancer. The anti-cancer-rich phytochemicals such as 12S-Hydroxyandrographolide and Genistein-7,4′-di-O-β-D-glucoside present in the A. paniculata ethanol extract aided in altering the morphological identity of the crypt from abnormal to almost normal and reduced the total number of ACF. Moreover, A. paniculata extract too interfered in the expansion of the adipocyte, which helped regulate the serum levels of leptin and insulin. This could be recognized as the anti-adipogenic and anti-obesity properties of the A. paniculata extract. Overall, A. paniculata shows potential chemopreventive effects on CRC, and along with that, it helps to regulate the obesity factors such as adipokine that aggravate the proliferation of CRC. ## References 1. Bray F., Ferlay J., Soerjomataram I., Siegel R.L., Torre L.A., Jemal A.. Global Cancer Statistics 2018: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. (2018) 68 394-424. DOI: 10.3322/caac.21492 2. Sung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F.. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. (2021) 71 209-249. DOI: 10.3322/caac.21660 3. Hassan M.R.A., Ismail I., Suan M.A.M., Ahmad F., Khazim W.K.W., Othman Z., Said R.M., Tan W.L., Mohammed S.R.N.S., Soelar S.A.. Incidence and Mortality Rates of Colorectal Cancer in Malaysia. Epidemiol. Health (2016) 38 e2016007. DOI: 10.4178/epih.e2016007 4. Mariotto A.B., Robin Yabroff K., Shao Y., Feuer E.J., Brown M.L.. Projections of the Cost of Cancer Care in the United States: 2010–2020. JNCI J. Natl. Cancer Inst. (2011) 103 117. DOI: 10.1093/jnci/djq495 5. Sninsky J.A., Shore B.M., Lupu G.V., Crockett S.D.. Risk Factors for Colorectal Polyps and Cancer. Gastrointest. Endosc. Clin. N. Am. (2022) 32 195-213. DOI: 10.1016/j.giec.2021.12.008 6. Vandevijvere S., Chow C.C., Hall K.D., Umali E., Swinburn B.A.. Increased Food Energy Supply as a Major Driver of the Obesity Epidemic: A Global Analysis. Bull. World Health Organ. (2015) 93 446. DOI: 10.2471/BLT.14.150565 7. Moghaddam A.A., Woodward M., Huxley R.. Obesity and Risk of Colorectal Cancer: A Meta-Analysis of 31 Studies with 70,000 Events. Cancer Epidemiol. Biomark. Prev. (2007) 16 2533-2547. DOI: 10.1158/1055-9965.EPI-07-0708 8. Riondino S., Roselli M., Palmirotta R., Della-Morte D., Ferroni P., Guadagni F.. Obesity and Colorectal Cancer: Role of Adipokines in Tumor Initiation and Progression. World J. Gastroenterol. (2014) 20 5177-5190. DOI: 10.3748/wjg.v20.i18.5177 9. Zhang J., Guo S., Li J., Bao W., Zhang P., Huang Y., Ling P., Wang Y., Zhao Q.. Effects of High-Fat Diet-Induced Adipokines and Cytokines on Colorectal Cancer Development. FEBS Open Bio (2019) 9 2117-2125. DOI: 10.1002/2211-5463.12751 10. Schmoll H.J., Van Cutsem E., Stein A., Valentini V., Glimelius B., Haustermans K., Nordlinger B., Van de Velde C.J., Balmana J., Regula J.. ESMO Consensus Guidelines for Management of Patients with Colon and Rectal Cancer. A Personalized Approach to Clinical Decision Making. Ann. Oncol. (2012) 23 2479-2516. DOI: 10.1093/annonc/mds236 11. Vodenkova S., Buchler T., Cervena K., Veskrnova V., Vodicka P., Vymetalkova V.. 5-Fluorouracil and Other Fluoropyrimidines in Colorectal Cancer: Past, Present and Future. Pharmacol. Ther. (2020) 206 107-447. DOI: 10.1016/j.pharmthera.2019.107447 12. Garg M.B., Lincz L.F., Adler K., Scorgie F.E., Ackland S.P., Sakoff J.A.. Predicting 5-Fluorouracil Toxicity in Colorectal Cancer Patients from Peripheral Blood Cell Telomere Length: A Multivariate Analysis. Br. J. Cancer (2012) 107 1525-1533. DOI: 10.1038/bjc.2012.421 13. Wigmore P.M., Mustafa S., El-Beltagy M., Lyons L., Umka J., Bennett G.. Effects of 5-FU. Adv. Exp. Med. Biol. (2010) 678 157-164. DOI: 10.1007/978-1-4419-6306-2_20 14. Al-Henhena N., Ying R.P.Y., Ismail S., Najm W., Khalifa S.A.M., El-Seedi H., Abdulla M.A.. Chemopreventive Efficacy of Andrographis Paniculata on Azoxymethane-Induced Aberrant Colon Crypt Foci In Vivo. PLoS ONE (2014) 9. DOI: 10.1371/journal.pone.0111118 15. Aiello P., Sharghi M., Mansourkhani S.M., Ardekan A.P., Jouybari L., Daraei N., Peiro K., Mohamadian S., Rezaei M., Heidari M.. Medicinal Plants in the Prevention and Treatment of Colon Cancer. Oxid. Med. Cell. Longev. (2019) 2019 2075614. DOI: 10.1155/2019/2075614 16. Jayakumar T., Hsieh C.Y., Lee J.J., Sheu J.R.. Experimental and Clinical Pharmacology of Andrographis Paniculata and Its Major Bioactive Phytoconstituent Andrographolide. Evid. Based. Complement. Alternat. Med. (2013) 2013 846740. DOI: 10.1155/2013/846740 17. Malik Z., Parveen R., Parveen B., Zahiruddin S., Aasif Khan M., Khan A., Massey S., Ahmad S., Husain S.A.. Anticancer Potential of Andrographolide from Andrographis Paniculata (Burm.f.) Nees and Its Mechanisms of Action. J. Ethnopharmacol. (2021) 272 1-16. DOI: 10.1016/j.jep.2021.113936 18. Samy R.P., Thwin M.M., Gopalakrishnakone P., Ignacimuthu S.. Ethnobotanical Survey of Folk Plants for the Treatment of Snakebites in Southern Part of Tamilnadu, India. J. Ethnopharmacol. (2008) 115 302-312. DOI: 10.1016/j.jep.2007.10.006 19. Zhu Q.C., Gao R.Y., Wu W., Guo B.M., Peng J.Y., Qin H.L.. Effect of a High-Fat Diet in Development of Colonic Adenoma in an Animal Model. World J. Gastroenterol. (2014) 20 8119. DOI: 10.3748/wjg.v20.i25.8119 20. Li Q., Zhang J., Zhou Y., Qiao L.. Obesity and Gastric Cancer. Front. Biosci. (2012) 17 2383-2390. DOI: 10.2741/4059 21. Zhang Y., Chen Z., Li J.. The Current Status of Treatment for Colorectal Cancer in China: A Systematic Review. Medicine (2017) 96 e8242. DOI: 10.1097/MD.0000000000008242 22. Dai Y., Chen S.R., Chai L., Zhao J., Wang Y., Wang Y.. Overview of Pharmacological Activities of Andrographis Paniculata and Its Major Compound Andrographolide. Crit. Rev. Food Sci. Nutr. (2019) 59 S17-S29. DOI: 10.1080/10408398.2018.1501657 23. Shrivastava N., Varma A., Padh H.. Andrographolide: A New Plant-Derived Antineoplastic Entity on Horizon. Evid.-Based Complement. Altern. Med. (2011) 2011 815390. DOI: 10.1093/ECAM/NEP135 24. Adedapo A.A., Adeoye B.O., Sofidiya M.O., Oyagbemi A.A.. Antioxidant, Antinociceptive and Anti-Inflammatory Properties of the Aqueous and Ethanolic Leaf Extracts of Andrographis Paniculata in Some Laboratory Animals. J. Basic Clin. Physiol. Pharmacol. (2015) 26 327-334. DOI: 10.1515/jbcpp-2014-0051 25. Chao W.W., Lin B.F.. Isolation and Identification of Bioactive Compounds in Andrographis Paniculata (Chuanxinlian). Chin. Med. (2010) 5 17. DOI: 10.1186/1749-8546-5-17 26. Cheung H.Y., Cheung C.S., Kong C.K.. Determination of Bioactive Diterpenoids from Andrographis Paniculata by Micellar Electrokinetic Chromatography. J. Chromatogr. A (2001) 930 171-176. DOI: 10.1016/S0021-9673(01)01160-8 27. Reddy M.V.B., Kishore P.H., Rao C.V., Gunasekar D., Caux C., Bodo B.. New 2‘-Oxygenated Flavonoids from Andrographis Affinis. J. Nat. Prod. (2003) 66 295-297. DOI: 10.1021/np020331k 28. Suriyo T., Pholphana N., Rangkadilok N., Thiantanawat A., Watcharasit P., Satayavivad J.. Andrographis Paniculata Extracts and Major Constituent Diterpenoids Inhibit Growth of Intrahepatic Cholangiocarcinoma Cells by Inducing Cell Cycle Arrest and Apoptosis. Planta Med. (2014) 80 533-543. DOI: 10.1055/s-0034-1368399 29. Sharma V., Qayum A., Kaul S., Singh A., Kapoor K.K., Mukherjee D., Singh S.K., Dhar M.K.. Carbohydrate Modifications of Neoandrographolide for Improved Reactive Oxygen Species-Mediated Apoptosis through Mitochondrial Pathway in Colon Cancer. ACS Omega (2019) 4 20435-20442. DOI: 10.1021/acsomega.9b01249 30. Rajendran P., Rengarajan T., Nandakumar N., Palaniswami R., Nishigaki Y., Nishigaki I.. Kaempferol, a Potential Cytostatic and Cure for Inflammatory Disorders. Eur. J. Med. Chem. (2014) 86 103-112. DOI: 10.1016/j.ejmech.2014.08.011 31. Lim J.C.W., Chan T.K., Ng D.S., Sagineedu S.R., Stanslas J., Wong W.F.. Andrographolide and Its Analogues: Versatile Bioactive Molecules for Combating Inflammation and Cancer. Clin. Exp. Pharmacol. Physiol. (2012) 39 300-310. DOI: 10.1111/j.1440-1681.2011.05633.x 32. Low M., Khoo C.S., Münch G., Govindaraghavan S., Sucher N.J.. An in Vitro Study of Anti-Inflammatory Activity of Standardised Andrographis Paniculata Extracts and Pure Andrographolide. BMC Complement. Altern. Med. (2015) 15. DOI: 10.1186/s12906-015-0525-7 33. Chen L., Zhu H., Wang R., Zhou K., Jing Y., Qiu F.. Ent-Labdane Diterpenoid Lactone Stereoisomers from Andrographis Paniculata. J. Nat. Prod. (2008) 71 852-855. DOI: 10.1021/np0704452 34. Baracos V.E., Martin L., Korc M., Guttridge D.C., Fearon K.C.H.. Cancer-Associated Cachexia. Nat. Rev. Dis. Prim. (2018) 4 17105. DOI: 10.1038/nrdp.2017.105 35. Darband S.G., Sadighparvar S., Yousefi B., Kaviani M., Ghaderi-Pakdel F., Mihanfar A., Rahimi Y., Mobaraki K., Majidinia M.. Quercetin Attenuated Oxidative DNA Damage through NRF2 Signaling Pathway in Rats with DMH Induced Colon Carcinogenesis. Life Sci. (2020) 253 117584. DOI: 10.1016/j.lfs.2020.117584 36. Venkatachalam K., Vinayagam R., Anand M.A.V., Isa N.M., Ponnaiyan R.. Biochemical and Molecular Aspects of 1,2-Dimethylhydrazine (DMH)-Induced Colon Carcinogenesis: A Review. Toxicol. Res. (2020) 9 2-18. DOI: 10.1093/toxres/tfaa004 37. Sherif D.A., Makled M.N., Suddek G.M.. The HIV Reverse Transcriptase Inhibitor Tenofovir Suppressed DMH/HFD-Induced Colorectal Cancer in Wistar Rats. Fundam. Clin. Pharmacol. (2021) 35 940-954. DOI: 10.1111/fcp.12679 38. Qi G., Tang B., Zhou L., Jikihara H., Kiwata A., Sakamoto Y., Tang F., Xiao S., Wang Z., Wu Q.. Effects of High-Fat Diet on 1,2-Dimethylhydrazine-Induced Aberrant Crypt Foci and Colorectal Tumours in Rats. Biomed. Rep. (2015) 3 289. DOI: 10.3892/br.2015.433 39. Hassan H.F.H., Mansour A.M., Salama S.A., El-Sayed E.S.M.. The Chemopreventive Effect of Thymol against Dimethylhydrazine and/or High Fat Diet-Induced Colon Cancer in Rats: Relevance to NF-ΚB. Life Sci. (2021) 274 119335. DOI: 10.1016/j.lfs.2021.119335 40. Poret J.M., Souza-Smith F., Marcell S.J., Gaudet D.A., Tzeng T.H., Braymer H.D., Harrison-Bernard L.M., Primeaux S.D.. High Fat Diet Consumption Differentially Affects Adipose Tissue Inflammation and Adipocyte Size in Obesity-Prone and Obesity-Resistant Rats. Int. J. Obes. (2018) 42 535-541. DOI: 10.1038/ijo.2017.280 41. Mopuri R., Ganjayi M., Banavathy K.S., Parim B.N., Meriga B.. Evaluation of Anti-Obesity Activities of Ethanolic Extract of Terminalia Paniculata Bark on High Fat Diet-Induced Obese Rats. BMC Complement. Altern. Med. (2015) 15. DOI: 10.1186/s12906-015-0598-3 42. Sabaratnam R., Svenningsen P.. Adipocyte-Endothelium Crosstalk in Obesity. Front. Endocrinol. (2021) 12 681290. DOI: 10.3389/fendo.2021.681290 43. Kennedy A., Spiers J.P., Crowley V., Williams E., Lithander F.E.. Postprandial Adiponectin and Gelatinase Response to a High-Fat versus an Isoenergetic Low-Fat Meal in Lean, Healthy Men. Nutrition (2015) 31 863-870. DOI: 10.1016/j.nut.2015.01.009 44. Chen Y.-C., Chien C.-Y., Hsu C.-C., Lee C.-H., Chou Y.-T., Shiah S.-G., Liu S.-Y., Yen C.-Y., Hsieh A.C.-T., Wabitsch M.. Obesity-Associated Leptin Promotes Chemoresistance in Colorectal Cancer through YAP-Dependent AXL Upregulation. Am. J. Cancer Res. (2021) 11 4220-4240. PMID: 34659884 45. Arita Y., Kihara S., Ouchi N., Takahashi M., Maeda K., Miyagawa J.I., Hotta K., Shimomura I., Nakamura T., Miyaoka K.. Paradoxical Decrease of an Adipose-Specific Protein, Adiponectin, in Obesity. Biochem. Biophys. Res. Commun. (1999) 257 79-83. DOI: 10.1006/bbrc.1999.0255 46. Weyer C., Funahashi T., Tanaka S., Hotta K., Matsuzawa Y., Pratley R.E., Tataranni P.A.. Hypoadiponectinemia in Obesity and Type 2 Diabetes: Close Association with Insulin Resistance and Hyperinsulinemia. J. Clin. Endocrinol. Metab. (2001) 86 1930-1935. DOI: 10.1210/jcem.86.5.7463 47. Kizer J.R.. A Tangled Threesome: Adiponectin, Insulin Sensitivity, and Adiposity: Can Mendelian Randomization Sort Out Causality?. Diabetes (2013) 62 1007. DOI: 10.2337/db12-1673 48. Otake S., Takeda H., Fujishima S., Fukui T., Orii T., Sato T., Sasaki Y., Nishise S., Kawata S.. Decreased Levels of Plasma Adiponectin Associated with Increased Risk of Colorectal Cancer. World J. Gastroenterol. (2010) 16 1252. DOI: 10.3748/wjg.v16.i10.1252 49. Chen C.C., Lii C.K., Lin Y.H., Shie P.H., Yang Y.C., Huang C.S., Chen H.W.. Andrographis Paniculata Improves Insulin Resistance in High-Fat Diet-Induced Obese Mice and TNFα-Treated 3T3-L1 Adipocytes. Am. J. Chin. Med. (2020) 48 1073-1090. DOI: 10.1142/S0192415X20500524 50. Ding L., Li J., Song B., Xiao X., Huang W., Zhang B., Tang X., Qi M., Yang Q., Yang Q.. Andrographolide Prevents High-Fat Diet-Induced Obesity in C57BL/6 Mice by Suppressing the Sterol Regulatory Element-Binding Protein Pathway. J. Pharmacol. Exp. Ther. (2014) 351 474-483. DOI: 10.1124/jpet.114.217968 51. Bahari H., Abidin A.Z., Balan S.S., Perumal K.V., Rosli N.S., Lotafi A.H.A., Danabala S., Manimaran M., Shafie N.H., Abdullah M.A.. The Effects of. Pharmacogn. Mag. (2020) 16 706. DOI: 10.4103/pm.pm_142_20 52. Sharma S., Joshi R., Kumar D.. Quantitative Analysis of Flavonols, Flavonol Glycoside and Homoisoflavonoids in Polygonatum Verticillatum Using UHPLC-DAD-QTOF-IMS and Evaluation of Their Antioxidant Potential. Phytochem. Anal. (2020) 31 333-339. DOI: 10.1002/pca.2899 53. Moon H.S., Liu X., Nagel J.M., Chamberland J.P., Diakopoulos K.N., Brinkoetter M.T., Hatziapostolou M., Wu Y., Robson S.C., Iliopoulos D.. Salutary Effects of Adiponectin on Colon Cancer: In Vivo and in Vitro Studies in Mice. Gut (2013) 62 561-570. DOI: 10.1136/gutjnl-2012-302092 54. Fujii T., Sakai E., Takahashi H., Yamada E., Ohkubo H., Higurashi T., Nakajima A.. The Distribution of Human Rectal Aberrant Crypt Foci and Criteria for Defining the Counting Area. Oncol. Lett. (2017) 13 4501-4504. DOI: 10.3892/ol.2017.6003 55. Szymlek-Gay E.A., Gray A.R., Heath A.-L.M., Ferguson E.L., Skeaff C.M.. Red Meat Consumption and Serum Lipids and Fatty Acids in Toddlers: Secondary Outcomes of a Randomized Controlled Trial. J. Pediatr. Gastroenterol. Nutr. (2018) 67 395-400. DOI: 10.1097/MPG.0000000000002018
--- title: 'Understanding the Interplay between Air Pollution, Biological Variables, and Major Depressive Disorder: Rationale and Study Protocol of the DeprAir Study' authors: - Elisa Borroni - Angela Cecilia Pesatori - Guido Nosari - Paola Monti - Alessandro Ceresa - Luca Fedrizzi - Valentina Bollati - Massimiliano Buoli - Michele Carugno journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049152 doi: 10.3390/ijerph20065196 license: CC BY 4.0 --- # Understanding the Interplay between Air Pollution, Biological Variables, and Major Depressive Disorder: Rationale and Study Protocol of the DeprAir Study ## Abstract Major depressive disorder (MDD) is a serious and disabling condition, whose etiological mechanisms are not fully understood. The aim of the DeprAir study is to verify the hypothesis that air pollution exposure may exacerbate neuroinflammation with consequent alterations in DNA methylation of genes involved in circadian rhythms and hormonal dysregulation, resulting in the worsening of depressive symptoms. The study population consists of 420 depressed patients accessing the psychiatry unit of the Policlinico Hospital (Milan, Italy), from September 2020 to December 2022. Data collection is still ongoing for about 100 subjects. For each participant demographic and lifestyle information, depression history and characteristics, as well as blood samples, were collected. MDD severity was assessed through five rating scales commonly used in clinical practice to assess the severity of affective symptoms. Exposure to particulate and gaseous air pollutants is assigned to each subject using both air pollution monitoring station measurements and estimates derived from a chemical transport model. DeprAir is the first study investigating in a comprehensive picture whether air pollution exposure could be an important modifiable environmental factor associated with MDD severity and which biological mechanisms mediate the negative effect of air pollution on mental health. Its results will represent an opportunity for preventive strategies, thus entailing a tremendous impact on public health. ## 1. Background Major depressive disorder (MDD) is a serious and disabling condition that, if not promptly managed, is associated with poor quality of life and high social costs [1,2]. Lifetime prevalence of MDD ranges from 2 to $21\%$, with the highest rates detected in some European countries, also as a result of population ageing [3]. As a matter of fact, people over the age of seventy have a prevalence of this disorder greater than $6\%$ [4]. MDD is twice more common in females than males [5], but completed suicide is more frequent in these latter who have also a minor response to antidepressant treatments [6,7]. In 2019, depression was the second-leading cause of disability worldwide [8]. Despite the high social impact of this condition, the etiological mechanisms underlying MDD are not fully understood. The most accredited etiological model contemplates the contribution of environmental and psychosocial factors in individuals biologically predisposed to develop a mood disorder [9]. MDD is associated with abnormalities of several biological systems resulting in a dysregulation of neurotransmitters, especially of serotonin and noradrenalin [10]. In this regard, the genetic variants predisposing to MDD refer to the serotonin receptors or transporters [11], but also to genes that regulate systems outside the central nervous system (CNS), such as the skeletal one [12], thus explaining the systemic nature of the disorder and the vulnerability of depressed subjects to different medical illnesses [13]. In addition, single nucleotide polymorphisms in clock genes (i.e., those regulating sleep-wake cycles) confer a higher risk of MDD [14]. As MDD and associated medical conditions (e.g., being overweight) share alterations in circadian rhythms [15], recent research has focused on the epigenetic mechanisms underpinning appetite and sleep dysregulation in depressed patients [16]. Circadian rhythms are regulated by the endogenous cellular clock, located in the suprachiasmatic nucleus of the hypothalamus [17]. Moreover, this master clock is regulated at molecular level by complex mechanisms involving positive and negative transcriptional/translational feedback loops, which drive the circadian rhythmicity of “clock gene” transcripts [18]. The positive feedback usually acts during the daytime and the negative one during the night. In the positive one, CLOCK and BMAL1 genes (as a protein dimer complex) are the principal activators of transcription, initiating the molecular circadian cycle with the activation of PERIOD (PER 1-2) and CRYPTOCHROME (CRY 1-2) genes. In the negative feedback, when PER and CRY genes are translated into proteins in the cytoplasm, these enter the nucleus to inhibit their own transcription by binding to the CLOCK–BMAL complex [19]. In addition to these transcriptional mechanisms, it is known that the regulation of these genes is mainly due to DNA methylation, a molecular mechanism of gene expression regulation, which is able to react and be reprogrammed by environmental stimuli [20]. On the other hand, the clock genes regulate the transcription of glucocorticoid receptors [21]. This biological mechanism explains why patients affected by MDD show a paradoxical state of chronic systemic over-inflammation in presence of increased plasma cortisol levels as a result of resistance to the effects of this hormone [22]. Of note, different authors reported that plasma levels of cytokines belonging to innate and adaptive immunity are increased in depressed patients compared to healthy controls and that antidepressant treatment can normalize the levels of these inflammatory factors [23]. Over-inflammation, in turn, triggers oxidative stress [24] and the activation of the hypothalamus–pituitary–adrenal (HPA) axis and the related hypercortisolemia typical of subjects affected by MDD [25]. Different environmental and psychosocial factors were reported to contribute to the onset of MDD, including obstetric complications [26] and childhood trauma [27]. In the last decade, air pollution has been hypothesized as a potential contributor of the onset of MDD, also in the light of the increasing mental health social costs due to urbanization [28]. Convincing data show that several air pollutants may be implicated in the onset or the worsening of depressive symptoms [29]. Of note, air pollution may exacerbate depressive symptoms by different biological mechanisms: (i) increasing systemic inflammation that in turn modifies neurotransmitter release and alters circadian rhythms [30], (ii) overcoming the blood–brain barrier and having a direct toxic effect on the CNS [31], and (iii) stimulating brain microglia by changes in bone marrow of the skull activated by chronic peripheral damage (e.g., in the respiratory system) [32]. ## 2. Aims and Hypotheses Given the above summarized available evidence, the relationships between air pollution exposure and inflammation, clock gene methylation, and hormonal dysregulation appear a promising mechanism for explaining MDD development and worsening. Our hypothesis is that air pollution exposure may exacerbate neuroinflammation with consequent epigenetic and hormonal dysregulation, resulting in the worsening of depressive symptoms (Figure 1). To verify our hypothesis, we will follow a multi-step approach within a cross-sectional study aimed to We hereby present study design, field activities, management organization, and characteristics of study subjects for which data collection has been completed. ## 3.1. Study Design The DeprAir study is a cross-sectional study conducted in the Lombardy region, Italy, whose aim is to understand the interplay between air pollution, biological variables, and MDD. Lombardy is located in the northern part of Italy, covering an area of 23.864 km2 and with a resident population of about 10 million people (https://bit.ly/3kXL6NZ, accessed on 13 February 2023). It is composed of 12 provinces, among which Milan is the capital, with around 1.3 million residents (https://bit.ly/3HsVogO, accessed on 13 February 2023). ## 3.2. Study Population The study population consists of 420 depressed patients accessing the psychiatry unit of the Policlinico Hospital in Milan (Italy), from September 2020 to December 2022. Data collection is still ongoing for about 100 subjects, while it has been completed for 317 subjects ($75\%$ of the target population), recruited up to 1 July 2022. Participants are recruited by trained psychiatrists among hospitalized or day-hospital patients or outpatients, who have been accessing the hospital since 2003 for MDD. The physician contacts already known patients by phone or meets them in person, in case they are hospitalized or outpatients, describes the study aims, and asks for participation in the study. To be eligible, patients must fulfill the following criteria: being ≥18 years old at enrollment; having received a diagnosis of MDD and having signed the consent form. Patients are excluded when they: have a medical condition associated to behavioral disorders (e.g., unbalanced hypothyroidism or stroke); have abused of drugs in the last four weeks; have comorbidities related to other psychiatric disorders (except for personality disorders different from borderline personality disorder); have medical conditions which may alter inflammatory markers (e.g., autoimmune diseases); have known ongoing infections; are taking treatments which may influence biological markers of interest (e.g., corticosteroids or interferons); are pregnant; are <18 years old. The participation rate in the study period was $75\%$. ## 3.3. Epidemiological and Clinical Data Collection At recruitment, each enrolled subject is asked to sign a consent form to:extract personal information from medical records (if already known);answer two questionnaires administered by the psychiatrist to collect demographic and lifestyle information, as well as depression history and characteristics;donate 30 mL of blood (five EDTA tubes of 6 mL each). ## 3.4. Questionnaire on Sociodemographic and Lifestyle Characteristics Each patient is interviewed by the psychiatrist who fills in the questionnaire. The questionnaire includes information on sociodemographic data (date of birth, sex, height, weight, education, occupation status), recent residential history (current complete address, previous complete address if changed in the last year, traffic status in the residential area), smoking history, including passive smoking at home and at workplace (smoking status; duration of smoking; number of cigarettes smoked; age at starting; age at quitting if former smoker; number of smoking family members; number of smoking colleagues at work), current health status including information on history of selected diseases (hypertension, hypercholesterolemia, diabetes, cancer, heart disease, renal failure) and medication, physical activity levels and sedentary behavior, type of diet (eating everything, vegetarian, vegan), and drinking habits (how much tea, coffee, wine, beer, and spirits). Table 1 summarizes the main demographic and lifestyle characteristics of the study population. MDD patients, recruited up to 1 July 2022, have a mean age of 51.5 years and are primarily composed of females ($67.2\%$). The largest majority has a high school or a university degree ($78.6\%$), is employed ($42.9\%$), and has been recruited as either new ($38.2\%$) or already known outpatients ($27.8\%$). About $29\%$ of the entire population is currently smoking, while $13\%$ is represented by former smokers. ## 3.5. Questionnaire on History and Characteristics of Depression The anamnestic questionnaire collects information about depression history and characteristics, in particular: family psychiatric history [including the type(s) of psychiatric disorder(s)], age at onset, duration of untreated illness in months, total duration of illness in years, duration of the latest episode in months, number of depressive episodes, hospitalizations (no vs. yes + total number of hospitalizations), suicide attempts (no vs. yes + total number of suicide attempts), psychotic symptoms (no vs. yes), seasonality of depression (no vs. yes), subtype of depression (melancholic, psychotic, with strong symptoms of anxiety, atypical), history of lifetime substances abuse (never, single-abuse or multiple-abuse and, if the subject ever suffered of substances abuse, type(s) of abuse(s) from alcohol, cocaine, cannabis, heroin, LSD, amphetamines, drugs, and MDMA), antidepressant treatment (no vs. yes + type of antidepressant assumed, active principle, dose, number of active principles ever assumed, suspension, and other treatments). Table 2 summarizes MDD characteristics. Mean age at onset of MDD is 39.5 years, mean number of MDD episodes is 2.9, and mean total duration of illness is 11.3 years. About $48.9\%$ of participants has a family history of psychiatric disorders, while $36\%$ has a family history of MDD. Overall, $23\%$ has been hospitalized for MDD, and $15.1\%$ has committed at least one suicide attempt. The largest majority has a MDD with strong symptoms of anxiety ($37.2\%$) followed by the melancholic subtype ($36.9\%$). About one fifth of patients ($19\%$) has suffered from single or multiple substance(s) abuse with alcohol and cannabis being the most frequent addictions. Almost all ($88.3\%$) patients take an antidepressant treatment, with selective serotonin reuptake inhibitors ($62.9\%$) and serotonin and norepinephrine reuptake inhibitors ($16.4\%$) being the most frequent ones. ## 3.6. Diagnostic Criteria and Rating Scales Diagnoses of MDD are confirmed by using Structural Clinical Interview for DSM-5 (SCID—Italian version) [33]. The psychiatrist evaluates depression severity of recruited patients by administering them the following rating scales, which are commonly used in clinical practice to assess the severity of affective symptoms:Montgomery-Asberg Depression Rating Scale (MADRS): it is a tool used to assess core symptoms of MDD (e.g., anhedonia). It is composed of the following 10 items: apparent sadness, reported sadness, inner tension, reduced sleep, reduced appetite, concentration difficulties, lassitude, inability to feel, pessimistic thoughts, suicidal thoughts. Each item has a severity scale from 0 to 6, with higher scores reflecting more severe symptoms. Ratings can be summarized in an overall score (from 0 to 60), which allows to stratify severity of depression as: 0–6: no depression, 7–19: mild depression, 20–34: moderate depression, ≥35: severe depression [34];Hamilton Depression Rating Scale (HAM-D) 21-item: this tool is indicated to assess anxiety and somatization symptoms of MDD. It is composed of 21 questions on types of symptoms associated with depression such as anxiety, mood, insomnia, and somatic symptoms experienced within the past week. Each symptom is rated on a scale of 0–2, 0–3, or 0–4 with 0 being absent and 2, 3, or 4 being the most severe. To obtain the overall score of severity (from 0 to 67), ratings can be added, and the total score can be stratified as: 0–7: no depression, 8–16: mild depression, 17–23: moderate depression, ≥24: severe depression [35];Clinical Global Impression-severity of illness (CGI): this tool is used by the psychiatrist to evaluate the global severity of illness answering the following question: “Considering your total clinical experience with this particular population, how mentally ill is the patient at this moment?”. The answer is given following this seven-point rating scale: 1 = normal, not at all ill; 2 = borderline mentally ill; 3 = mildly ill; 4 = moderately ill; 5 = markedly ill; 6 = severely ill; 7 = among the most extremely ill patients [36];Sheehan Disability Scale (SDS): this scale is used to evaluate the social dysfunction associated with MDD. It consists of a self-reported assessment of functional impairment composed of five items. The first three are global rating scales which assess impairment in work, home, and family responsibilities. There are two additional questions which measure perceived stress and social support. The items are scored individually on 10-point numerical rating scales, except for the “social support” one that can be scored 0–100 [37];Global Assessment of Functioning (GAF): this tool is used to evaluate the overall impairment associated with MDD. In particular, it measures how much a person’s symptoms affect his/her day-to-day life on a scale of 0 to 100. This scale is broken into 10 sections, which are known as anchor points. The higher the score is, the better the patient is able to handle daily activities, suggesting that a lower score indicates a greater social disfunction associated with depression [38]. Summary statistics for severity of MDD scales are reported in Table 3. Scales mean scores are 27.1 for MADRS, 22.7 for HAM-D, and 59.6 for GAF. Based on CGI scores, most of the patients are mildly ($22.4\%$) or moderately ($30\%$) ill, while mean scores for SDS single domains are: 7.2 for work, 6.7 for relationships, 6.6 for family, 6.3 for stress, and 59.8 for social support. ## 3.7. Blood Sample Collection Specific laboratory standard operating procedures have been developed to ensure quality control of every step involved in biospecimen collection and storage. Blood drawing is performed directly by the psychiatrist. Each subject provides a 30 mL blood sample in five EDTA tubes, which are delivered to laboratory and processed within 4 h. One of the tubes is used for blood cell count, while the remaining ones are centrifuged and processed to obtain plasma and buffy coat fractions. Plasma and buffy coat samples are stored at −80 °C for subsequent quantification of inflammatory and hormonal markers and DNA methylation analysis, respectively. ## 3.8. Inflammatory Markers As mentioned above, markers of both innate and adaptive immunity have been widely associated with the severity of MDD. Considering the innate immunity, the following markers will be measured—IL-1, IL-6, and TNFα, while the following markers of adaptive immunity will be considered—IL-8, IL-12, and CCL1. In addition, the levels of malondialdehyde will be measured as a parameter of oxidative stress. All these markers will be evaluated on plasma by using ELISA (enzyme linked immunosorbent assay) kits. ## 3.9. DNA Methylation of Clock and Clock-Controlled Genes We have selected 10 target genes (CLOCK, BMAL1, PER1, PER2, OX1R, CRY1, CRY2, OXTR, FOXp3, HERV-W), which include clock genes and genes directly stimulated by clock pathways, to measure DNA methylation by pyrosequencing. Following genomic DNA extraction from buffy coat, we performed this using a Promega kit (Madison, WI, USA), 3 μg DNA (concentration 25 ng/μL), which will be bisulfite-treated using EZ DNA Methylation-Gold™ Kit (Zymo Research, Orange, CA, USA) according to the manufacturer’s protocol. Bisulfite-treated DNA will be stored at −20 °C and used shortly after treatment. For each reaction, a 50 μL PCR will be carried out by adding 10 μL of bisulfite-treated genomic DNA to 25 μL of GoTaq Green Master mix (Promega, Madison, WI, USA), 1 μL of forward primer (10 μM), 1 μL of reverse primer (10 μM), and water. One of the primers is biotin-labelled and is used to purify the final PCR product by Sepharose beads. The PCR product will be bound to Streptavidin Sepharose HP (Amersham Biosciences, Uppsala, Sweden), and the Sepharose beads containing the immobilized PCR product will be purified, washed, denatured using a 0.2 M NaOH solution, and washed again using the Pyrosequencing Vacuum Prep Tool (Pyrosequencing, Inc., Westborough, MA, USA), as recommended by the manufacturer. Then, 0.3 μΜ Pyrosequencing primer will be annealed to the purified single-stranded PCR product, and Pyrosequencing will be performed using the PyroMark Q96 MD Pyrosequencing System (QIAGEN). Methylation quantification will be performed using the provided software (Pyro Q-CpG software, version 1.0.9—Biotage, Uppsala, Sweden)). The degree of methylation will be expressed as percentage of 5-methylated cytosines (%5mC) over the sum of methylated and unmethylated cytosines. We will use built-in controls to verify bisulfite conversion efficiency. ## 3.10. Hormonal Markers Hormonal changes, as well as inflammation, can be correlated with the severity of MDD. The following hormones (including neuropeptides) will be measured: adrenal corticotropic hormone (ACTH), cortisol, neurophysin I (a good marker of oxytocin levels in the CNS), vasopressin, kisspeptin, orexin, and prolactin. Plasma samples of the recruited subjects are collected at similar time (around 11 a.m., as hormone levels change according to circadian rhythms) and measured using ELISA kits. ## 3.11. Exposure Assessment As air pollutants of interest, we consider particulate matter with diameter less than or equal to 10 (PM10) and 2.5 µm (PM2.5), nitrogen dioxide (NO2), sulfur dioxide (SO2), and ozone (O3). Each patient’s residential address is translated into spatial coordinates using the web tool GPS Visualizer (https://www.gpsvisualizer.com/, accessed on 17 January 2023) and geocoded using QGIS (QGIS Development Team, 2022. QGIS Geographic Information System. Open-Source Geospatial Foundation Project. http://qgis.osgeo.org, accessed on 17 January 2023). Air pollution levels are assigned to each patient following two approaches (Figure 2):(i)PM10, PM2.5, NO2, SO2, and O3 measurements retrieved from the air quality monitoring stations of the Regional Environmental Protection Agency (ARPA Lombardia). Daily means of pollutant levels measured at the station closest to the subject’s residential address are assigned to each subject. Missing values for each pollutant on a specific day and monitor are imputed by computing the average of measurements of that pollutant for the previous and the following seven days.(ii)PM10, PM2.5, NO2, and O3 daily mean estimates are derived from the Flexible Air quality Regional (FARM) model [39,40]. This type of Eulerian model takes into account the atmospheric chemistry, together with transport, dispersion, and deposition phenomena [41,42]. By integrating data measured from ARPA air quality and meteorological monitoring stations, emissions, concentrations at the beginning of the simulation period, and trend in adjacent areas, it estimates pollutants’ concentrations as daily/hourly means covering the whole Lombardy territory with a grid of 1 × 1 km cell. Each subject is assigned the daily average of pollutants’ exposure estimated inside the grid cell where his/her residential address fell. Given the finer spatial resolution of its estimates, we will give priority to the second approach. We will make use of data from monitoring stations, should the FARM model estimates not be available. Meteorological data (e.g., temperature, humidity) are retrieved from ARPA monitoring stations too. To take into account the potential effect of short- and medium-term exposures, we will investigate several time windows: (i) single daily lags obtained considering pollutants’ daily means from the day of recruitment (lag0) up to 30 days before (lag30), (ii) averaged daily lags obtained by averaging pollutants’ levels of the day of recruitment with the levels of the day before (lag01) and of each preceding day up to 30 days before (lag030). Exploratory analyses will also be performed to investigate time windows potentially representative of cumulative exposures (e.g., annual averages). ## 3.12. Ethical Issues The study design, research aims, and measurements have been approved by the local Institutional Review Board of the Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico (approval numbers 498_2020bis and 950_2020). Participants agreed to sign a written informed consent explaining the study in detail, as well a consent for biobanking of the blood samples for future research studies. In any case, new measurements will only be performed after approval of the Ethical Committee. ## 3.13. Statistical Analysis and Power Calculation We will use standard descriptive statistics to summarize data. Graphical inspection of the main variables of interest will be performed to examine their distribution and uncover the need for transformation. Associations between concentration levels of air pollutants and MDD severity, and between air pollution and biological variables, will be assessed by multiple linear regression models, adjusted for main confounders of interest (e.g., age, sex, socioeconomic status, smoking habit, season, temperature, etc.). Potential non-linearity of the associations will be verified flexibly modelling air pollution variables as splines with various knots depending on their distribution. Results will be expressed either as regression coefficients (β, which will take the unit of measure of the outcome variable, e.g., rating scale scores) or as percent change in the investigated outcome, with corresponding $95\%$ Confidence Intervals ($95\%$CI) for a given variation in air pollutant exposure levels. We will perform similar models with biological variables used as independent variables to assess the association between biological variables and MDD severity. This will allow us to evaluate a pool of biological markers, which might predict MDD severity. Sensitivity analyses will be also performed, e.g., considering duration of illness (both treated and untreated) and age at onset or stratifying by depression subtype to verify its potential for effect modification. To investigate potential pathways that could explain the observed associations between exposure to air pollutants and MDD severity, we will perform a mediation analysis, an approach which allows to examine how intermediate variables (i.e., the mediators) are related to the observed exposure-outcome relationship [43]. As our study will test several associations, a formal power calculation was not possible, and sample size was calculated on the basis of the primary hypothesis that short-term exposure to air pollution is associated with MDD severity. We based our sample size calculation on preliminary results on 195 women randomly selected from the SPHERE population [44], for which we observed a positive association between the Beck Depression Inventory (BDI) score and the average exposure level of PM10 in the third day preceding the day of recruitment. A sample size of 420 achieves $98\%$ power to detect a change in slope from 0.09 (under the null hypothesis) to 0.19 (under the alternative hypothesis) when the standard deviation of the PM10 variable distribution is 17 μg/m3, the standard deviation of the BDI variable distribution is 9, and the two-sided significance level is 0.05. These data represent an informative source for our study, as the BDI scale can be converted into the HAM-D scale (which we administer) and vice versa, with equipercentile linking [45]. Based on our previous experience on studies on health effects of air pollution exposure (e.g., [46,47]) and clinical practice on depressed patients, as well as on some of the characteristics of the participants enrolled so far, we expect that the distribution of the variables in our study participants will resemble those of our preliminary data. ## 3.14. Data Management and Privacy Protection To protect each patient’s privacy, all collected information and biological samples are anonymized from personal identifying information and each participant can be identified only through a five-digit randomly assigned barcode. The information linking each subject’s identity to his/her personal barcode is held in a secure database. Questionnaire data are collected in paper forms and are subsequently imputed in the database, regularly checking quality and completeness of information. Data processing is anonymous, and the highest level of confidentiality is assured for all personal information. Quality of collected data is routinely checked by comparing information from different sources (clinical records, questionnaires, biochemical exams), assessing variable range and distribution, and verifying database completeness through simple statistics. ## 3.15. Dissemination of Results A study website has been created (https://deprair.com/, accessed on 10 February 2023) and will be updated regularly. It contains relevant information about the study and related events. Results and dissemination material will be published on the website, at the time they will be produced. Furthermore, results will be converted into user-friendly materials and published in press releases, educational programs, as well as scientific conferences and journals. ## 3.16. Collaboration Opportunities The rich set of clinical information and molecular data of the DeprAir study makes it a good environment for collaboration opportunities. Proposals from outside the study team for research projects to test specific hypotheses on the DeprAir population will be reviewed by our research group and can be sent to info@deprair.com. ## 4. Discussion To the best of our knowledge, DeprAir is the first study investigating whether exposure to air pollution could be an important modifiable environmental factor associated with MDD severity and which biological mechanisms mediate the negative effect of air pollution on mental health. We had previously reported that particulate air pollution is associated with a higher risk of manic episodes with mixed features in hospitalized bipolar patients and showed that increased levels of PM move the manic episode towards the depressive pole of the bipolar disorder spectrum [46]. Our current study will thus allow us to further deepen our knowledge in the complex pathway that links environmental exposures to psychiatric disorders. In particular, DeprAir will clarify the role of exposure to environmental air pollution on MDD severity. This aim will be achieved by considering two methods for environmental exposure, thus maximizing the probability of assigning accurate exposure values to our population. Both approaches have proved their soundness in similar settings, as documented in previous studies on health effects of air pollution exposure [44,46,47,48]. In particular, the assignment of personal exposures based on residential address using daily pollutant levels at a 1 × 1 km resolution represents the best compromise between spatial and temporal resolution that we can currently achieve when considering our entire regional territory. Finally, the association of interest will be studied considering single-exposure models, as well as multiple-exposure techniques, to estimate the most accurate impact of air pollution on MDD. For the same purpose, roles of temperature and humidity in the association between air pollution exposure and MDD will also be evaluated. The present study will also shed light on the biological mechanisms underlying MDD worsening, assessing whether exposure to air pollution determines alterations in inflammatory, epigenetic, and hormonal variables, as well as identifying a pool of biological markers that might predict MDD severity. In addition, DeprAir will allow to disentangle the role that each element of the hypothesized causal pathway (environmental stressors and inflammatory, epigenetic, and hormonal variables) might play in determining MDD severity through mediation models. To apply these models, some basic conditions have to be met: the association between air pollution exposure and the outcome (MDD severity) is statistically significant; the hypothesized mediators have an effect on the outcome when the exposure is controlled for; air pollution has an effect on the mediators. If these criteria will be satisfied, the mediation analysis will allow to evaluate direct (DE) and indirect effects (IE). DE is the effect of air pollution on MDD severity adjusted for the mediators, while IE estimates the proportion of the effect of air pollution exposure on MDD severity that is mediated by the biological variables. Correlations among variables will be considered through the use of high dimensional mediation analysis. We are aware that one possible limitation of our study is its limited sample size, which, however, is an intrinsic constraint of studies examining preliminary hypothesis with collection of biological data. Should our hypotheses be confirmed, our findings will have to be validated in larger study populations. The results of DeprAir will represent an opportunity for preventive strategies, having the possibility to entail a tremendous impact on public health. In particular, the present study represents an opportunity to open new strategies in terms of prevention for MDD: if air pollution will be confirmed as an important environmental risk factor for the severity of depressive symptoms, reduction in the levels of air pollutants may represent an easy strategy to reduce the severity of depressive symptoms with a clear economic saving due for example to reduced psychiatric hospitalizations. Of note, MDD is not only associated with direct costs due to hospitalizations, but also to indirect costs, such as loss of days of work for patients and their caregivers. Caregivers suffer from the high social burden due to MDD, as they often look after their relatives, also in the light of the high risk of suicide associated with the depressive illness. As such, even a small improvement in the management of patients affected by depression could lead to a great advantage in terms of the patients’ quality of life, burden for caregivers, and social costs. In addition, the investigation of the biological factors that potentially mediate the toxic effects of air pollution on the brain may open to new pharmacological strategies that may be represented by anti-inflammatory drugs and regularization of circadian rhythms or of hormone-associated biological cascades. Furthermore, the topic of air pollution and its consequences on mental health can give a further drive towards the realization of towns that include green areas or other facilities aimed to air quality improvement. ## References 1. Kraus C., Kadriu B., Lanzenberger R., Zarate C.A., Kasper S.. Prognosis and Improved Outcomes in Major Depression: A Review. Transl. Psychiatry (2019) 9 17. DOI: 10.1038/s41398-019-0460-3 2. Malhi G.S., Mann J.J.. Depression. Lancet (2018) 392 2299-2312. DOI: 10.1016/S0140-6736(18)31948-2 3. Gutiérrez-Rojas L., Porras-Segovia A., Dunne H., Andrade-González N., Cervilla J.A.. Prevalence and Correlates of Major Depressive Disorder: A Systematic Review. Braz. J. Psychiatry (2020) 42 657. DOI: 10.1590/1516-4446-2020-0650 4. Smith M., Haedtke C., Shibley D.. Late Life Depression Detection: An Evidence-Based Guideline. J. Gerontol. Nurs. (2015) 41 18. DOI: 10.3928/00989134-20150115-01 5. Hasin D.S., Sarvet A.L., Meyers J.L., Saha T.D., Ruan W.J., Stohl M., Grant B.F.. Epidemiology of Adult DSM-5 Major Depressive Disorder and Its Specifiers in the United States. JAMA Psychiatry (2018) 75 336-346. DOI: 10.1001/jamapsychiatry.2017.4602 6. Ceresa A., Esposito C.M., Surace T., Legnani F., Cirella L., Cetti D., Di Paolo M., Nosari G., Zanelli Quarantini F., Serati M.. Gender Differences in Clinical and Biochemical Parameters of Patients Consecutively Hospitalized for Unipolar Depression. Psychiatry Res. (2022) 310 114476. DOI: 10.1016/j.psychres.2022.114476 7. Dumais A., Lesage A.D., Alda M., Rouleau G., Dumont M., Chawky N., Roy M., Mann J.J., Benkelfat C., Turecki G.. Risk Factors for Suicide Completion in Major Depression: A Case-Control Study of Impulsive and Aggressive Behaviors in Men. Am. J. Psychiatry (2005) 162 2116-2124. DOI: 10.1176/appi.ajp.162.11.2116 8. Ferrari A.J., Santomauro D.F., Mantilla Herrera A.M., Shadid J., Ashbaugh C., Erskine H.E., Charlson F.J., Degenhardt L., Scott J.G., McGrath J.J.. Global, Regional, and National Burden of 12 Mental Disorders in 204 Countries and Territories, 1990–2019: A Systematic Analysis for the Global Burden of Disease Study 2019. Lancet Psychiatry (2022) 9 137-150. DOI: 10.1016/S2215-0366(21)00395-3 9. Li Z., Ruan M., Chen J., Fang Y.. Major Depressive Disorder: Advances in Neuroscience Research and Translational Applications. Neurosci. Bull. (2021) 37 863. DOI: 10.1007/s12264-021-00638-3 10. Stapelberg N.J.C., Bui T.A., Mansour V., Johnson S., Branjerdporn G., Adhikary S., Ashton K., Taylor N., Headrick J.P.. The Pathophysiology of Major Depressive Disorder through the Lens of Systems Biology: Network Analysis of the Psycho-Immune-Neuroendocrine Physiome. J. Neuroimmunol. (2022) 372 577959. DOI: 10.1016/j.jneuroim.2022.577959 11. Fabbri C., Hosak L., Mössner R., Giegling I., Mandelli L., Bellivier F., Claes S., Collier D.A., Corrales A., Delisi L.E.. Consensus Paper of the WFSBP Task Force on Genetics: Genetics, Epigenetics and Gene Expression Markers of Major Depressive Disorder and Antidepressant Response. World J. Biol. Psychiatry (2016) 18 5-28. DOI: 10.1080/15622975.2016.1208843 12. Howard D.M., Adams M.J., Clarke T.-K., Hafferty J.D., Gibson J., Shirali M., Coleman J.R.I., Hagenaars S.P., Ward J., Wigmore E.M.. Genome-Wide Meta-Analysis of Depression Identifies 102 Independent Variants and Highlights the Importance of the Prefrontal Brain Regions. Nat. Neurosci. (2019) 13 21. DOI: 10.1038/s41593-018-0326-7 13. Cambria C., Ingegnoli F., Borzi E., Cantone L., Coletto L.A., Rizzuto A.S., De Lucia O., Briguglio S., Ruscica M., Caporali R.. Synovial Fluid-Derived Extracellular Vesicles of Patients with Arthritides Contribute to Hippocampal Synaptic Dysfunctions and Increase with Mood Disorders Severity in Humans. Cells (2022) 11. DOI: 10.3390/cells11152276 14. Charrier A., Olliac B., Roubertoux P., Tordjman S.. Clock Genes and Altered Sleep–Wake Rhythms: Their Role in the Development of Psychiatric Disorders. Int. J. Mol. Sci. (2017) 18. DOI: 10.3390/ijms18050938 15. Iodice S., Ceresa A., Esposito C.M., Mucci F., Conti D.M., Pergoli L., Tarantini L., Vigna L., Bollati V., Buoli M.. The Independent Role of Body Mass Index (BMI) and Severity of Depressive Symptoms on Biological Changes of Women Affected by Overweight/Obesity and Stand-Up Project Group. Public Health (2021) 18. DOI: 10.3390/ijerph18062923 16. Pitsillou E., Liang J., Hung A., Karagiannis T.C.. The Circadian Machinery Links Metabolic Disorders and Depression: A Review of Pathways, Proteins and Potential Pharmacological Interventions. Life Sci. (2021) 265 118809. DOI: 10.1016/j.lfs.2020.118809 17. Tahara Y., Shibata S.. Circadian Rhythms of Liver Physiology and Disease: Experimental and Clinical Evidence. Nat. Rev. Gastroenterol. Hepatol. (2016) 13 217-226. DOI: 10.1038/nrgastro.2016.8 18. Sato F., Kohsaka A., Bhawal U.K., Muragaki Y.. Potential Roles of Dec and Bmal1 Genes in Interconnecting Circadian Clock and Energy Metabolism. Int. J. Mol. Sci. (2018) 19. DOI: 10.3390/ijms19030781 19. Lopez-Minguez J., Gómez-Abellán P., Garaulet M.. Circadian Rhythms, Food Timing and Obesity. Proc. Nutr. Soc. (2016) 75 501-511. DOI: 10.1017/S0029665116000628 20. Alegría-Torres J.A., Baccarelli A., Bollati V.. Epigenetics and Lifestyle. Epigenomics (2011) 3 267-277. DOI: 10.2217/epi.11.22 21. Leliavski A., Dumbell R., Ott V., Oster H.. Adrenal Clocks and the Role of Adrenal Hormones in the Regulation of Circadian Physiology. J. Biol. Rhythm. (2015) 30 20-34. DOI: 10.1177/0748730414553971 22. Horowitz M.A., Zunszain P.A.. Neuroimmune and Neuroendocrine Abnormalities in Depression: Two Sides of the Same Coin. Ann. N. Y. Acad. Sci. (2015) 1351 68-79. DOI: 10.1111/nyas.12781 23. Köhler C.A., Freitas T.H., Stubbs B., Maes M., Solmi M., Veronese N., de Andrade N.Q., Morris G., Fernandes B.S., Brunoni A.R.. Peripheral Alterations in Cytokine and Chemokine Levels After Antidepressant Drug Treatment for Major Depressive Disorder: Systematic Review and Meta-Analysis. Mol. Neurobiol. (2018) 55 4195-4206. DOI: 10.1007/s12035-017-0632-1 24. Salim S.. Oxidative Stress and the Central Nervous System. J. Pharmacol. Exp. Ther. (2017) 360 201-205. DOI: 10.1124/jpet.116.237503 25. Pariante C.M.. Why Are Depressed Patients Inflamed? A Reflection on 20 Years of Research on Depression, Glucocorticoid Resistance and Inflammation. Eur. Neuropsychopharmacol. (2017) 27 554-559. DOI: 10.1016/j.euroneuro.2017.04.001 26. Serati M., Bertino V., Malerba M.R., Mucci F., Barkin J.L., Grassi S., Altamura A.C., Buoli M.. Obstetric Complications and Subsequent Risk of Mood Disorders for Offspring in Adulthood: A Comprehensive Overview. Nord. J. Psychiatry (2020) 74 470-478. DOI: 10.1080/08039488.2020.1751878 27. Zisook S., Planeta B., Hicks P.B., Chen P., Davis L.L., Villarreal G., Sapra M., Johnson G.R., Mohamed S.. Childhood Adversity and Adulthood Major Depressive Disorder. Gen. Hosp. Psychiatry (2022) 76 36-44. DOI: 10.1016/j.genhosppsych.2022.03.008 28. Gong P., Liang S., Carlton E.J., Jiang Q., Wu J., Wang L., Remais J.V.. Urbanisation and Health in China. Lancet (2012) 379 843. DOI: 10.1016/S0140-6736(11)61878-3 29. Borroni E., Pesatori A.C., Bollati V., Buoli M., Carugno M.. Air Pollution Exposure and Depression: A Comprehensive Updated Systematic Review and Meta-Analysis. Environ. Pollut. (2022) 292 118245. DOI: 10.1016/j.envpol.2021.118245 30. Hergenhan S., Holtkamp S., Scheiermann C.. Molecular Interactions Between Components of the Circadian Clock and the Immune System. J. Mol. Biol. (2020) 432 3700. DOI: 10.1016/j.jmb.2019.12.044 31. Kang Y.J., Tan H.Y., Lee C.Y., Cho H.. An Air Particulate Pollutant Induces Neuroinflammation and Neurodegeneration in Human Brain Models. Adv. Sci. (2021) 8 2101251. DOI: 10.1002/advs.202101251 32. Cugurra A., Mamuladze T., Rustenhoven J., Dykstra T., Beroshvili G., Greenberg Z.J., Baker W., Papadopoulos Z., Drieu A., Blackburn S.. Skull and Vertebral Bone Marrow Are Myeloid Cell Reservoirs for the Meninges and CNS Parenchyma. Science (2021) 373. DOI: 10.1126/science.abf7844 33. First M.B., Williams J.B.W., Kang R.S., Spitzer R.L., Fossati A., Borroni S.. SCID-5-CV—Intervista Clinica Strutturata per i Disturbi Del DSM-5® (2017) 34. Quilty L.C., Robinson J.J., Rolland J.-P., Fruyt F.D., Rouillon F., Bagby R.M.. The Structure of the Montgomery-Åsberg Depression Rating Scale over the Course of Treatment for Depression. Int. J. Methods Psychiatr. Res. (2013) 22 175-184. DOI: 10.1002/mpr.1388 35. Sharp R.. The Hamilton Rating Scale for Depression. Occup. Med. (2015) 65 340. DOI: 10.1093/occmed/kqv043 36. Busner J., Targum S.D.. The Clinical Global Impressions Scale: Applying a Research Tool in Clinical Practice. Psychiatry (2007) 4 28-37 37. Coles T., Coon C., DeMuro C., McLeod L., Gnanasakthy A.. Psychometric Evaluation of the Sheehan Disability Scale in Adult Patients with Attention-Deficit/Hyperactivity Disorder. Neuropsychiatr. Dis. Treat. (2014) 10 887-895. DOI: 10.2147/NDT.S55220 38. Aas I.M.. Guidelines for Rating Global Assessment of Functioning (GAF). Ann. Gen. Psychiatry (2011) 10 2. DOI: 10.1186/1744-859X-10-2 39. Silibello C., Calori G., Brusasca G., Giudici A., Angelino E., Fossati G., Peroni E., Buganza E.. Modelling of PM10 Concentrations over Milano Urban Area Using Two Aerosol Modules. Environ. Model. Softw. (2008) 23 333-343. DOI: 10.1016/j.envsoft.2007.04.002 40. Mircea M., Ciancarella L., Briganti G., Calori G., Cappelletti A., Cionni I., Costa M., Cremona G., D’Isidoro M., Finardi S.. Assessment of the AMS-MINNI System Capabilities to Simulate Air Quality over Italy for the Calendar Year 2005. Atmos. Environ. (2014) 84 178-188. DOI: 10.1016/j.atmosenv.2013.11.006 41. Bessagnet B., Pirovano G., Mircea M., Cuvelier C., Aulinger A., Calori G., Ciarelli G., Manders A., Stern R., Tsyro S.. Presentation of the EURODELTA III Intercomparison Exercise-Evaluation of the Chemistry Transport Models’ Performance on Criteria Pollutants and Joint Analysis with Meteorology. Atmos. Chem. Phys. (2016) 16 12667-12701. DOI: 10.5194/acp-16-12667-2016 42. Kukkonen J., Olsson T., Schultz D.M., Baklanov A., Klein T., Miranda A.I., Monteiro A., Hirtl M., Tarvainen V., Boy M.. A Review of Operational, Regional-Scale, Chemical Weather Forecasting Models in Europe. Atmos. Chem. Phys. (2012) 12 1-87. DOI: 10.5194/acp-12-1-2012 43. Valeri L., Vander Weele T.J.. Mediation Analysis Allowing for Exposure-Mediator Interactions and Causal Interpretation: Theoretical Assumptions and Implementation with SAS and SPSS Macros. Psychol. Methods (2013) 18 137-150. DOI: 10.1037/a0031034 44. Bollati V., Iodice S., Favero C., Angelici L., Albetti B., Cacace R., Cantone L., Carugno M., Cavalleri T., De Giorgio B.. Susceptibility to Particle Health Effects, MiRNA and Exosomes: Rationale and Study Protocol of the SPHERE Study. BMC Public Health (2014) 14. DOI: 10.1186/1471-2458-14-1137 45. Furukawa T.A., Reijnders M., Kishimoto S., Sakata M., Derubeis R.J., Dimidjian S., Dozois D.J.A., Hegerl U., Hollon S.D., Jarrett R.B.. Translating the BDI and BDI-II into the HAMD and Vice Versa with Equipercentile Linking. Epidemiol. Psychiatr. Sci. (2020) 29 e24. DOI: 10.1017/S2045796019000088 46. Carugno M., Palpella D., Ceresa A., Pesatori A.C., Buoli M.. Short-Term Air Pollution Exposure Is Associated with Lower Severity and Mixed Features of Manic Episodes in Hospitalized Bipolar Patients: A Cross-Sectional Study in Milan, Italy. Environ. Res. (2021) 196 110943. DOI: 10.1016/j.envres.2021.110943 47. Carugno M., Dentali F., Mathieu G., Fontanella A., Mariani J., Bordini L., Milani G.P., Consonni D., Bonzini M., Bollati V.. PM10 Exposure Is Associated with Increased Hospitalizations for Respiratory Syncytial Virus Bronchiolitis among Infants in Lombardy, Italy. Environ. Res. (2018) 166 452-457. DOI: 10.1016/j.envres.2018.06.016 48. Ferrari L., Borghi F., Iodice S., Catelan D., Rossi S., Giusti I., Grisotto L., Rovelli S., Spinazzè A., Alinovi R.. INSIDE Project: Individual Air Pollution Exposure, Extracellular Vesicles Signaling and Hypertensive Disorder Development in Pregnancy. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17239046
--- title: PPAR Alpha Activation by Clofibrate Alleviates Ischemia/Reperfusion Injury in Metabolic Syndrome Rats by Decreasing Cardiac Inflammation and Remodeling and by Regulating the Atrial Natriuretic Peptide Compensatory Response authors: - María Sánchez-Aguilar - Luz Ibarra-Lara - Agustina Cano-Martínez - Elizabeth Soria-Castro - Vicente Castrejón-Téllez - Natalia Pavón - Citlalli Osorio-Yáñez - Eulises Díaz-Díaz - María Esther Rubio-Ruíz journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049157 doi: 10.3390/ijms24065321 license: CC BY 4.0 --- # PPAR Alpha Activation by Clofibrate Alleviates Ischemia/Reperfusion Injury in Metabolic Syndrome Rats by Decreasing Cardiac Inflammation and Remodeling and by Regulating the Atrial Natriuretic Peptide Compensatory Response ## Abstract Metabolic syndrome (MetS) is a cluster of factors that increase the risk of developing diabetes, stroke, and heart failure. The pathophysiology of injury by ischemia/reperfusion (I/R) is highly complex and the inflammatory condition plays an important role by increasing matrix remodeling and cardiac apoptosis. Natriuretic peptides (NPs) are cardiac hormones with numerous beneficial effects mainly mediated by a cell surface receptor named atrial natriuretic peptide receptor (ANPr). Although NPs are powerful clinical markers of cardiac failure, their role in I/R is still controversial. Peroxisome proliferator-activated receptor α agonists exert cardiovascular therapeutic actions; however, their effect on the NPs’ signaling pathway has not been extensively studied. Our study provides important insight into the regulation of both ANP and ANPr in the hearts of MetS rats and their association with the inflammatory conditions caused by damage from I/R. Moreover, we show that pre-treatment with clofibrate was able to decrease the inflammatory response that, in turn, decreases myocardial fibrosis, the expression of metalloprotease 2 and apoptosis. Treatment with clofibrate is also associated with a decrease in ANP and ANPr expression. ## 1. Introduction Metabolic syndrome (MetS) is an entity characterized by various risk factors, such as hypertension, obesity, insulin resistance, and atherogenic dyslipidemia that includes reduced high density lipoprotein (HDL) cholesterol and increased triglycerides [1]. This pathology is associated with a sedentary lifestyle and increased caloric intake as well as genetic factors [2]. The association of MetS and its components with increased risk of adverse cardiovascular outcomes, morbidity, and mortality is well established. There are previous reports that MetS doubles the incidence of coronary artery disease, increases the progression of atheromatous plaque, and elevates the incidence of cardiac failure [3]. Heart failure remains the leading cause of death, morbidity, and medical expenses worldwide. Myocardial infarction and the pathophysiology of injury by ischemia/reperfusion (I/R) are highly complex. The production of several cytokines increases in myocardial damage by I/R, including tumor necrosis factor alpha (TNF-α), interleukin-6 (IL-6), interleukin-1 beta (IL-1β), and transforming growth factor-beta (TFG-β1) [4,5]. Furthermore, inflammation plays an important role in cardiac remodeling by regulating three processes: [1] increasing deposition of types I and III collagen and extracellular matrix crosslinking (fibrosis), [2] increasing the expression and activity of metalloproteinases (MMPs), such as MMP-2 and MMP-9, and [3] stimulating infiltration of leukocytes, which increases the inflammatory response. Altogether, these processes lead to matrix remodeling and cardiomyocyte apoptosis that may contribute to the development and progression of heart failure [6,7]. The myocardial injury that occurs following a period of I/R is not yet completely understood in its underlying pathophysiological mechanism. In response to myocardial damage, a cascade of compensatory events occurs that moderates the cardiac remodeling process including kinins, urocortins, adrenomedullin, incretins, and natriuretic peptides (NPs) [8]. NPs form part of a family of humoral components that are involved in cardiovascular homeostasis but also have effects at the endocrine level and participate in energy balance, mitochondrial biogenesis, respiration, and lipid oxidation [9]. NPs have been proposed as therapeutic strategies for obesity due to their production by other organs, such as adipose tissue, liver, and skeletal muscle, among others, and in diseases, such as MetS and type 2 diabetes [10]. There are three types of NPs: atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C natriuretic peptide (CNP), whose expression is regulated by several factors including the pro-inflammatory cytokines present in pathological conditions [11]. It is well known that ANP and N-terminal pro-BNP (NT-ProBNP) are powerful clinical markers of heart failure; however, it is unclear whether enhancement of ANP and BNP expression is protective or if it aggravates myocardial damage. The activities of NPs are mediated by natriuretic peptide receptors A, B, and C, which are cell surface, single-span transmembrane receptors linked to the activity of an intrinsic guanylate cyclase. The three NP receptors differ in structure, biological effects, and ligand specificity; type A and B receptors activate downstream signaling pathways through the synthesis of the intracellular second messenger, cGMP, while type C receptor is mainly involved in their clearance. Receptor type A has greater affinity for ANP but can also bind to BNP [12,13]. The expression of the NPs’ receptors is regulated by several factors, such as vitamin D, angiotensin II, and endothelin, and by transcription factors, such as cGMP response element-binding protein (CREBP). Patients with hypertension, obesity, or MetS have lower levels of ANP, NT-proBNP, and ANP receptor [11,14,15]. On the other hand, several therapeutic approaches have been proposed for the control of MetS due to its multifactorial nature. Specifically, drugs, such as statins or fibrates, have been extensively used to treat dyslipidemia [1]. Moreover, some reports have shown the pleiotropic effects of fibrates to protect the heart from damage through the activation of peroxisome-proliferator-activated receptors (PPAR). Three members of the PPAR family, PPARα, PPARγ, and PPARβ/δ, have been investigated widely. PPARα plays important roles in many processes including inflammation, remodeling, metabolism, mitochondria biogenesis, and apoptosis [16]. We and others have reported that PPARα agonists constitute an effective treatment to decrease myocardial damage in several experimental models. The anti-inflammatory effect of PPARα in myocardial injury is mainly through inhibition of pro-inflammatory signaling pathways and improvement of the lipid profile. Moreover, PPARα also modulates the myocardial renin–angiotensin pathway and resets the insulin signaling pathway [17,18]. Although PPARα signaling is known to be associated with a cardioprotective effect, there is little evidence regarding the effect of PPARα on the regulation of NPs and their receptors in myocardial injury. In this work, we hypothesized that a PPARα agonist, clofibrate, exerts a cardioprotective effect by decreasing myocardial inflammation, remodeling, and apoptosis in a MetS rat model. Therefore, the aim of this study was to evaluate the effect of clofibrate on the expression of NPs and ANPr in hearts from MetS rats subjected to I/R injury. ## 2. Results The body and serum biochemical parameters confirm the presence of MetS in our experimental rat models and are presented in Table 1. The MetS animals were hypertensive, dyslipidemic (high triglycerides and non-HDL cholesterol and low HDL cholesterol concentrations), and they had central obesity and insulin resistance. Clofibrate treatment decreased central obesity and body weight and improved lipid profile and insulin sensitivity in MetS rats. In contrast, only the levels of non-HDL-c were significantly decreased with the same treatment in control (Ct) rats (Table 1). Figure 1 shows concentrations of IL-1β, IL-6, and TNF-α in left ventricles from Ct and MetS rats. Our results show that MetS sham-operated rats had significantly higher levels (approximately three times more) of these pro-inflammatory interleukins when compared to Ct-Sh animals. I/R conditions promoted an increase in the concentrations of these variables from Ct and MetS rats and the 7-day clofibrate pre-treatment prevented this increase in the same proportion in the experimental groups (approximately $45\%$ for IL-1β, $30\%$ for IL-6, and $55\%$ for TNF-α). Dying cardiomyocytes secrete cytokines to recruit leukocytes into the infarcted myocardium, and this intense inflammatory reaction initiates a reparative response that includes synthesis of collagen. For this reason, we decided to perform hematoxylin–eosin stain (HE) and picrosirius red (PSR) staining to evaluate the number of infiltrated cells and ventricular fibrosis in all experimental groups (Figure 2A,B). The hearts from rats with MetS presented more infiltrated cells and collagen deposition when compared to hearts from Ct animals (Figure 2A,B). HE staining of tissues revealed that I/R injury promoted cell swelling and the disruption and irregular arrangement of myocardial fibers when compared to sham groups; these changes are more evident in hearts from MetS rats (Figure 2A). We also observed numerous inflammatory cells infiltrated in the left ventricles of MetS rats after I/R injury (yellow arrows). When the MetS animals were pre-treated with clofibrate, these showed minor morphological changes and a decrease in the number of infiltrated cells (Figure 2A), which is an indication that some pro-inflammatory process has been disactivated. Sirius Red staining showed a significant increase in the collagen volume fraction (CVF) as deposition of collagen I and III fibers in the left ventricle from MetS rats in Sh-operated and under I/R conditions. Treatment with clofibrate was associated with the decrease of the CVF. There were no significant differences among the groups in control rats (Figure 2B). Figure 3 shows the levels of protein expression of MMP-2 that participate in the cardiac remodeling process. Western blot analysis indicated that in sham-operated and under I/R conditions, hearts from MetS rats had a higher level of MMP-2 expression compared to the Ct group. However, clofibrate treatment was able to significantly decrease the MMP-2 expression in ventricles from both Ct and MetS animals compared to vehicle-treated corresponding groups. The immunoexpression of ANP and ANP receptor in our experimental groups was determined to assess myocardial NPs’ response in I/R injury. The expression levels of ANP and ANPr were similar in hearts from Ct and MetS animals. There was a marked increase in the ANP/ANPr signal in the ventricles of MetS-V-IR rats compared with sham-operated MetS animals (Figure 4 and Figure 5). In contrast, MetS rats pre-treated with clofibrate displayed a significant decreased in the density of ANP and its receptor. There was no significant difference in the levels of expression of ANP and ANPr in ventricles from Ct groups. The concentration of NT-proBNP was determined in the myocardia from Ct and MetS animals due to the fact that it is considered a biomarker for diagnosis and severity of heart failure. NT-proBNP concentrations were not statistically significantly different among groups (Figure 6). The presence of ANP and its receptor in mitochondria was investigated using immunogold labeling for electronic microscopy because cardiac NPs have beneficial effects at the level of this organelle. Figure 7 shows that ANP and ANPr were present in greater quantities in mitochondria and fibers from ventricles from MetS-Sh rats than in hearts from Ct-Sh rats. Myocardial I/R injury in MetS rats was associated with intense immunoreactivity that was concentrated on the fibers and mitochondria. The MetS group treated with clofibrate showed a decrease in the expression of ANP and ANPr (Figure 7A,B). In contrast, the gold marks in ventricles from Ct groups remained unchanged. The effect of clofibrate treatment on cell death was evaluated by the TUNEL assay due to the fact that apoptosis after I/R injury may aggravate myocardial remodeling and decrease cardiac function. Figure 8 shows a statistically significant increase in the number of apoptotic cells in MetS-Sh when compared to Ct-Sh animals. The number of dead cells did not show a significant change among the different groups under I/R conditions; however, a 7-day treatment with clofibrate prevented apoptosis in the MetS group without there being evident change in the Ct group. ## 3. Discussion MetS is a cluster of risk factors that leads to cardiovascular diseases. MetS doubles the incidence of coronary artery disease; it increases the progression of the atheromatous plaque and is associated with myocardial infarct. The treatment of choice for reducing myocardial ischemic injury is timely and effective myocardial reperfusion; however, reperfusion also adds a further component to myocardial injury [19]. Pro-inflammatory cytokines, growth factors, and several neurohumoral pathways participate in cardiac repair and remodeling after the damage of cardiac tissue evoked by I/R; however, the role of NPs in myocardial damage is still controversial [5]. NPs exert their cardioprotective functions not only as the circulating peptide hormones but also as local autocrine and paracrine factors. The transcriptional regulation of these hormones is less well understood; therefore, the identification of pathways or molecules that regulate NPs’ expression is of potential importance [20,21]. On the other hand, growing evidence suggests that PPARα agonists have a protective function. It has been shown that fibrates administration reverts some of the effects caused by myocardial damage by regulating several processes, such as energy metabolism, oxidative stress, inflammation, and cell differentiation [17,18]. However, as far as we know, the effect of these compounds on NPs’ signaling has not been completely explored. In this work, we show that 7 days of clofibrate pre-treatment is able to diminish the production of pro-inflammatory cytokines, collagen synthesis, and the MMP-2 activation associated with I/R injury. Our study also demonstrates that ANP and ANPr were present in myocardial fibers and mitochondria and were elevated at the same time as inflammation progressed. Here, we also establish a novel role for the PPARα agonist on the regulation of NPs and ANPr expression in hearts from MetS rats subjected to I/R damage. The results in Table 1 show that clofibrate administration had beneficial effects in the MetS model by reducing body weight and central adiposity. As expected, clofibrate pre-treatment had a triglyceride-lowering effect and increased insulin sensitivity without significantly affecting blood pressure, the concentration of glucose, or total cholesterol. Our data are in line with previous reports in which another PPARα agonist was used [19,22,23]. In this study, we used an I/R model because ischemic damage can be reversed by early reperfusion and because the restoration of blood flow can cause more damage by affecting the myocardial repair and remodeling processes. During myocardial injury due to I/R, inflammation, proliferation, and maturation of cells form part of the reparative response [24]. First, we measured pro-inflammatory cytokine concentrations in the damaged tissue. Figure 1 shows that cytokines were increased in hearts from MetS rats, that there was a larger increase in I/R conditions, and that clofibrate pre-treatment was able to decrease these inflammatory biomarkers. It has been reported that MetS is a pathological condition in which inflammation is present; moreover, TNF-α and IL-1β have been related to the presence of insulin resistance and to the promotion of the pro-inflammatory state by activating NF-κB pathway in a vicious circle [25,26,27]. Regarding the anti-inflammatory effect of clofibrate, our results agree with previous reports by Sun et al. [ 23] and by our group [18]. PPARα represses inflammation by inhibiting mediators of the NF-κB signaling pathway. It inhibits the production of IL-6 dependent on IL-1 in addition to prostaglandins, such as COX-2, in vascular smooth muscle cells [28]. In addition, it also favors the anti-inflammatory signaling pathway by increasing IκB [29]. Inflammatory cytokines and other growth factors recruit immune cells into the infarcted region in the damaged myocardium, which intensifies the inflammatory reaction and modulates downstream signaling cascades involved in remodeling and reparative responses [30]. Moreover, an imbalance in the equilibrium of synthesis and degradation of extracellular matrix components occurs after I/R injury; hence, we decided to evaluate the levels of fibrosis and the expression of MMP-2 in our model. Our results indicate that the number of infiltrated cells and the accumulation of collagen were increased in MetS and I/R conditions (Figure 2A,B). Clofibrate administration was able to prevent these effects. It is well known that neutrophils and/or macrophages show a strong affiliation to invade the damaged area due to the apoptotic cardiomyocytes and the need for their removal; however, in the present paper, we did not evaluate the cell subpopulations infiltrated [6,30]. Our data are in line with several studies that have demonstrated that MetS signs, such as obesity, insulin resistance, lipotoxicity, and inflammation, play an important role in myocardial fibrosis by activating several downstream signaling cascades [31,32,33]. In addition, scientific evidence has implicated angiotensin II in the inflammatory process as well as in the progression of myocardial fibrosis via binding to AT1 receptor [34]. Regarding this, we previously reported that PPARα agonist (fenofibrate) diminishes the angiotensin II concentrations and AT1 expression in hearts from MetS rats subjected to ischemic injury. Therefore, this might be an additional mechanism to attenuate myocardial damage in our model [18]. Myocardial remodeling is a dynamic process in which MMPs play an important role. MMPs directly activate cytokines, chemokines, cell surface receptors, growth factors, and other proteases, and they contribute to fundamental processes, such as cell proliferation, differentiation, adhesion, migration, and apoptosis. In I/R injury, expression of MMP-2 is of particular interest because MMP-2 degrades extracellular matrix substrates including Type IV collagen, laminin, elastin, interstitial fibrillar collagen, and sarcomeric proteins [35,36]. The expression of this zinc-dependent protease is regulated by mechanical signals, inflammatory factors, hormones, and NPs, among other factors; some authors even suggest that MMP-2 can be used as a possible pharmacological target in the treatment of heart failure [31,36,37,38]. Therefore, we analyzed the expression of MMP-2 in all experimental groups and found that the expression of this enzyme was higher in hearts from MetS rats in sham-operated conditions and under I/R damage without there being changes in the Ct group. Pre-treatment with clofibrate significantly decreased the expression of this protease in both Ct and MetS groups (Figure 3). This protective effect of the PPARα activation by decreasing the expression of MMPs has been previously shown in other experimental models of damage to the myocardium [17,39,40]. On the other hand, there is a strong association between MMP-2 levels with NPs in patients with heart failure [41]. The role of these peptides has been explored in several organs and in pathologies, such as MetS; however, their therapeutic potential and their receptors are just beginning to be expanded [11,42,43]. Hence, we decided to evaluate the effect of clofibrate on the expression of ANP and ANPr in our experimental groups. Results in Figure 4 and Figure 5 demonstrate that the peptide and its receptor were present in higher levels in ventricles from MetS rats subjected to I/R as a compensatory response to injury and that this increase was reduced by pre-treatment with the drug. Our results are in accordance with previous publications that reported that fenofibrate treatment, another PPAR agonist, blocked the increase in plasma levels of NPs in a pig model of heart failure [44]. Some reports have shown that angiotensin II can promote the production of ANP [45,46], and in a previous report, we showed that treatment with fenofibrate decreases the levels of this peptide [18]; therefore, this might be another synergic way by which clofibrate could decrease the ANP levels in our experimental model. Relatively little information exists regarding the regulation of the ANPr expression. Our findings demonstrate a positive association between the ANP and ANPr in ventricles from MetS rats subjected to I/R injury, which supports the results obtained by Pandey et al. [ 47] showing that NPs may regulate ANPr gene expression. Our work suggests that the activation of PPARα was able to regulate the expression of this cardiac receptor; however, further studies are needed to evaluate the binding of this transcriptional factor to the receptor gene. Moreover, several studies have shown that NPs may exhibit protective effects in I/R injury through mitochondria-mediated mechanisms. Some authors have reported that NPs provide myocardial protection by regulating mitochondrial biogenesis and swelling, the production of reactive oxygen species, the oxidation of fat, and the synthesis of ATP, leading to decreased cell death [48,49,50]. The representative electron micrographs presented in Figure 7 demonstrate that hearts damaged by I/R from MetS animals showed significantly higher levels of ANP and ANPr in myocardial fibers and mitochondria compared to the Ct group. This effect was abolished by treatment with clofibrate. However, additional studies are required to elucidate the pathophysiological significance of the presence of these factors in mitochondrial dynamics and function during myocardial damage in MetS rats. Currently, BNP and NT-proBNP are widely used as diagnostic biomarkers for heart failure and myocardial infarction in clinical medicine [51]. Cardiac myocytes constitute a major source of BNP-related peptides in response to cardiac wall stress of the left ventricle in acute myocardial infarction. BNP is synthesized first as a 108 amino acid prohormone (proBNP), and proBNP is then cleaved by furin into the biologically active 32 amino acid BNP (C-terminal fragment) and into the biologically inactive 76 amino acid N-terminal fragment (NT-proBNP). In our model, we decided to measure NT-proBNP rather than BNP because of its higher half-life (120 min vs. 20 min) [51,52]. We observed no statistically significant difference of NT-proBNP among groups (I/R or MetS conditions compared to Ct-Sh) (Figure 6). We hypothesized that we would have possibly observed differences in BNP among the different groups because BNP is biologically active and its biological actions counteract the fibrosis and inflammation associated with myocardial infarction in rats [52,53,54]. Additionally, BNP and NT-proBNP diagnostic utility in clinical studies has been researched by measuring both biomarkers in circulation (serum or plasma) [51]; thus, we can possibly observe differences in BNP peptides among groups when measuring circulating NT-proBNP and/or BNP. Therefore, we need to compare in situ and circulating levels of BNP and NT-proBNP concentrations among our different groups to test this hypothesis. On the other hand, a small quantity of BNP is stored in cardiomyocyte granules; instead, BNP is transcribed as needed in response to wall stress and once proBNP is released into circulation, it is cleaved to NT-proBNP and finally BNP [55,56]. Therefore, the lack of statistically significant differences in NT-proBNP in our groups might be also related to the small amounts of BNP stored in cardiomyocytes; however, additional studies are needed to test this hypothesis. Since ANP and BNP have similar affinities for ANPr [46] and in our model we observed that ANPr was upregulated in I/R groups, we hypothesized that BNP might act locally by ANPr binding to reduce myocardial fibrosis. We need to carry out ligand binding assays to verify this hypothesis; however, this is outside the objectives of this study. Factors including endocrine and paracrine modulation by other neuro-hormones and cytokines are also of importance. Our findings support the idea to perform assays of BNP and other biomarkers, such as creatine kinase, cardiac troponin I, heart-type fatty acid-binding protein, adrenomedullin, and osteoprotegerin, in combination with imaging analysis as a tool for differential diagnosis and therapy in heart diseases [57,58]. Finally, we evaluated the clofibrate effect on cell death due to the fact that an excessive inflammatory response after I/R injury induces myocardial apoptosis and that apoptosis may aggravate the formation of cardiac scars, leading to decreased cardiac function [59]. Figure 8 shows that MetS is associated with an increase in cardiac apoptosis when compared to the Ct group. This increase was higher when hearts were subjected to I/R conditions. The 7-day pre-treatment with clofibrate reverted this effect. Our data are in accordance with previous studies that showed the association of apoptosis and heart failure and the anti-apoptotic role of PPARα agonists [16,17,39]. ## 4.1. Experimental Animals All animal protocols and procedures were performed in compliance with the recommendations of the official normative for laboratory animal care protocols (SAGARPA, NOM-062-ZOO-1999, Mexico). Male Wistar rats, 25 days old, were randomly separated into two groups of 12 animals: group 1, control (Ct) rats that were given tap water for drinking, and group 2, MetS rats that received $30\%$ sucrose in their drinking water for 24 weeks. The animals were kept under 12 h light/obscurity cycles and environmental temperature ranging from 18 to 26 °C. They were fed commercial rodent pellets (PMI Nutrition International Inc., LabDiet 5008, Richmond, IN, USA) ad libitum [18]. Next, both Ct and MetS animals were randomly subdivided into two equal groups according to receive vehicle (Vh) or clofibrate (Clo, 100 mg/kg/day) by intraperitoneal injection every day for 7 days [16,17]. At the end of the treatment, the animals were weighed, and systolic arterial blood pressure was determined in conscious animals by the tail-cuff plethysmography technique previously performed [18]. The intra-abdominal white adipose tissue (retroperitoneal fat pad) was also carefully dissected with scissors after euthanasia, wet weight was determined, and then the tissue was discarded. ## 4.2. Ischemia Reperfusion Model Animals were anesthetized for ischemia/reperfusion (I/R) surgery with an intramuscular drug mixture of ketamine hydrochloride 80mg/kg (from Laboratorios Aranda, Queretaro, Qro. Mexico) and xylazine hydrochloride 10mg/kg (from PiSA Farmaceutica, Guadalajara, Jal., Mexico). The I/R procedure was performed as previously reported by Oidor Chan et al. [ 22] with a modification in the time of reperfusion. After inducing sedation and maintaining mechanical ventilation (70 breaths per minute; vol. 8–10 mL/kg), asepsis of the thoracic area was performed. The chest was opened by a lateral incision along the upper margin of the fourth rib to expose the heart. A PE-10 tube was inserted between the suture and anterior descending coronary artery (LAD) immediately before ligation. After 30 min of coronary artery ligation, we removed the PE-10 tubing to establish reperfusion for 60 min. In sham (Sh), control, and MetS groups, the procedure was identical, except the left anterior descending artery was not transiently ligated. ## 4.3. Serum Biochemical Variables We determined the basal serum biochemical parameters in animals that were fasted overnight. The blood samples were collected by vessel puncture from the vena cava. Serum was isolated by centrifugation and stored until needed. The baseline fasting values of glucose, total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides were measured with commercial enzymatic kits (RANDOX Laboratories Ltd., Crumlin, Country Antrim, UK); insulin was determined by a rat-specific insulin radioimmunoassay (Linco Research, Inc., Saint Charles, MO, USA), as previously reported [60]. ## 4.4. Cardiac Cytokines and NT-proBNP Quantification At the end of the surgery procedure, left ventricle samples were obtained and homogenized with lysis buffer (50 mM HEPES, pH 7.5; 150 mM NaCl, $1\%$ glycerol, $1\%$ Triton X-100, 1.5 mM MgCl2 and 5mM EGTA, 1mM PMSF) and protease inhibitors cocktail. Later, the homogenates were centrifuged at 10,000× g and the protein quantification was performed with bicinchoninic acid technique (Pierce BCA protein assay kit, Thermo Scientific, Rockford, IL, USA). The cytokine values were obtained by the ELISA sandwich technique as previously reported [17]. N-terminal pro-brain natriuretic peptide (NT-proBNP) concentrations were assayed in left ventricle homogenates using a Rat NT-proBNP ELISA kit (MyBioSource, San Diego, CA, USA) according to the manufacturer’s instructions. ## 4.5. Western Blot Frozen left ventricle from the different experimental groups was homogenized with lysis buffer pH 7.4 (250 mM Tris-HCl, 2.5 mM EDTA) and protease inhibitors cocktail (Complete® tablets, Roche Applied Science, Mannheim, Germany) using a tissue homogenizer (Fisher Scientific, Waltham, MA, USA) at 4 °C. The homogenate was centrifuged at 5000× g for 10 min at 4 °C and the supernatant was separated and stored at −70 °C until required. Total protein was determined by the Bradford method [61]. One hundred µg of protein from all experimental groups was separated in SDS-PAGE gel ($10\%$) and electro transferred to polyvinylidene difluoride (PVDF) membrane. The blots were blocked with $5\%$ non-fat dehydrated milk for three hours under continuous stirring at room temperature. Later, the membranes were incubated overnight with primary antibodies to MMP-2 from Santa Cruz Biotechnology (Santa Cruz, CA, USA) at 4 °C and subsequently with its corresponding secondary antibody for three hours at room temperature (Jackson ImmunoResearch, Suffolk, UK). All the blots were incubated with β-actin as a loading control. After incubation, plates were revealed with a peroxidase-chemiluminescent kit (Clarity western ECL substrate, Bio-Rad Laboratories, Inc. Hercules, CA, USA). The bands were quantified by densitometry employing the Quantity One software 4.6 (Bio-Rad Laboratories, Inc. Hercules, CA, USA) by using a GS-800 densitometer. The results are expressed as arbitrary units (AU) of the ratio between MMP-2 and β-actin [60]. ## 4.6. Histology A fragment of the left ventricle wall obtained from 3 rats from each group, previously preserved at −70 °C, was gradually thawed (consecutive changes to −20 °C, 4 °C, and room temperature). Then, the tissue was fixed in $4\%$ paraformaldehyde (PFA) for 48 h, with 5 washes (15 min each) with PBS [62]. Subsequently, cryoprotection was performed by placing the tissues from each group in $30\%$ sucrose for 48 h [63,64]. Frozen cross-sections at 10 μm were mounted on gelatinized or electrocharged slides. The slides were kept at 4 °C until required. ## 4.7. Hematoxylin–Eosin Stain (HE) and Picrosirius Red (PSR) HE and PSR staining were performed in the gelatinized slide sections. Slides containing cardiac tissue sections (2 blocks with 3 heart fragments, from 3 different rats from each group on each slide) were selected, dried at room temperature for one hour, and rehydrated for 30 min before use. HE staining (ab245880, Abcam PLC, Cambridge, UK) was performed to visualize the morphology of the tissue as well as the cellular infiltrates. First, most of the water from the cuts was removed and 300 μL of hematoxylin, Mayer’s (Lillie’s Modification) was applied to completely cover each tissue section and incubated for 5 min. The slides were rinsed in two changes of distilled water to remove excess stain; we applied adequate Bluing Reagent to completely cover each tissue section and incubated for 10–15 s; we rinsed the slides in two changes of distilled water and we immersed the slides in absolute alcohol to blot excess off. Then, we applied 300 μL Eosin Y Solution (Modified Alcoholic) to completely cover each tissue section to excess and incubated for 2–3 min; the slides were rinsed using absolute alcohol and dehydrated in three changes of absolute alcohol. Finally, the slides were mounted in synthetic resin. Myocardial collagen volume fraction (CVF) was measured by picrosirius red (PSR) staining (ab150681, Abcam PLC, Cambridge, UK) [65,66]. Slides containing cardiac tissue sections (2 blocks with 3 heart fragments, from 3 different rats from each group on each slide) were selected, dried at room temperature for one hour, and rehydrated for 30 min before use. The water from the slides was removed and 350 μL of PSR Solution was applied to completely cover the tissue section and incubated for 60 min; the slides were rinsed quickly in two changes of acetic acid solution. We rinsed the slides in absolute alcohol and dehydrated in two changes of absolute alcohol. Finally, the slides were mounted in synthetic resin. Visualization of HE and PSR staining were performed by light microscopy (Olympus BX51) and PSR was detected on a Floid Cell Imaging Station (Life Technologies, Carlsbad, CA, USA). The number of infiltrated cells in HE images was quantified and FVC values were calculated (ventricular collagen area/field area) [65] from the measurement of the corresponding areas in PSR images acquired at 20× employing Image-Pro Premier 9.0 (Media Cybernetics) software. ## 4.8. Immunodetection of Atrial Natriuretic Peptide (ANP) and Atrial Natriuretic Peptide Receptor (ANPr) Electrocharged slide sections containing cardiac tissue sections were chosen, dried at room temperature for one hour, passed through xylol–alcohol until rehydrated in PBS, and re-fixed with $4\%$ PFA (10 min). The fluorescence generated by aldehydes was quenched by incubation in a solution of glycine 0.1M/PBS pH 7.4 for 2 min. Antigen retrieval was performed with Tris (0.5M)/EDTA (0.1M) solution, pH 9 at 95 °C for 10 min. The sections were washed with PBS 5 times, depositing and removing sufficient volume on the sections with a micropipette. They were then permeabilized with TritonX-100 $0.2\%$/BSA $1.5\%$ in PBS. Each tissue block was surrounded with a hydrophobic pencil. After incubation with the blocking solution (BS) (BSA $3\%$/ 0.1 triton X-100, in PBS) at room temperature for 30 min in a humid chamber [67], the sections were incubated with the corresponding primary antibody for ANP (4.77 μg/mL (ab225844, Abcam PLC, Cambridge, UK)) or ANPr (2.5 μg/mL (ab14356, Abcam PLC, Cambridge, UK)) for 72 h at 4 °C in the dark and in a humid chamber [64]. Once this period was over, the sections were washed with PBS (4× 10 min), depositing and removing sufficient volume on the sections with a micropipette. Incubation with secondary antibody (10μg/mL (ab150079, Abcam PLC, Cambridge, UK)) in BS for 1.5 h at room temperature was performed. Staining with 4’,6-diamidino-2-Phenylindole (DAPI) (1.43 μM) was used to visualize the nuclei, and they were mounted with mounting medium for fluorescence. Nonspecific binding was verified by using negative controls, which were incubated with the blocking solution ($3\%$ BSA/0.1 triton X-100, in PBS) without the primary antibody. Immunofluorescence assays were performed using previously reported methods with small modifications [68,69]. ## 4.9. TUNEL Test Once the PNA immunoassay was completed, the reaction mixture containing the enzyme solution (TdT) and label solution (fluorescein-dUTP) corresponding to the TUNEL assay (In Situ Cell Death Detection Kit, Fluorescein (Roche Applied Science, Mannheim, Germany, 11684795910)) was applied [64,70,71]. The slides were incubated at 37 °C for 60 min in the dark inside a humid chamber. At the end of incubation, the sections were washed with PBS and the sections were mounted with mounting medium and analyzed by fluorescence microscopy. Tissues fixed, permeabilized, and treated with label solution (without terminal transferase) instead of TUNEL reaction mixture were used as negative controls; as positive controls, we used tissues incubated with DNase I recombinant (3000 U/mL-10 min-room temperature). TUNEL-positive cells were quantified and the percentage of positive cells (marked with DAPI) per total cell number per field was calculated. ## 4.10. Image Acquisition, Number of Cell Infiltrate, Positive TUNEL Cells, and ANP and ANPr Intensity Quantification Image acquisition for light microscopy para HE was performed with a Q-Imaging camera (Microplublisher 5.0 Real-Time Viewing (RTV) coupled to an Olympus BX51 microscope. Visualization and acquisition of fluorescent images for PSR, ANP, ANPr, and TUNEL were performed using Floid Cell Imaging Station equipment. The number of infiltrated cells (HE staining), positive TUNEL cells with respect to the total number of cells (DAPI) per field, and the quantification of fluorescence intensity per area unity (integrated optical density (IOD) (lum/pix^2)) to ANP and ANPr were carried out with the Image-Pro Premier 9 (Media Cybernetics). At least 4 fields (20X) of each animal (3 per group) were quantified, with a total of at least 12–24 determinations per condition. ## 4.11. Immune Colloidal Gold Technique The samples were processed as reported by Soria Castro et al. [ 72] with small modifications. The hearts were excised and divided for histological analyses while fresh; small pieces of rat heart left ventricle were fixed for 2 h in $4\%$ paraformaldehyde (cat 30525-894. Electron Microscope Science, Haffield, PA, USA) and $0.1\%$ glutaraldehyde (cat.111-30-8; Electron Microscope Science) in 0.1M PBS, pH 7.4. The samples were dehydrated in gradual alcohols and embedded in LR White resin (cat.14381; Electron Microscope Science, Haffield, PA, USA). Ultrathin sections (50 nm) were placed on carbon/formvar-coated nickel grids of 150 mesh. The grids were floated on droplets of PBS for 10 min and then whole goat serum for 1 h. Next, the grids were washed with droplets of PBS 10 min 3 times, after which each grid was floated on 30 μL of antibodies against the ANP (cat. ab225844 Abcam Cambridge, UK) and the ANP receptor ((ab14356), Abcam, Cambridge, UK) at a dilution of 1:20 overnight at 4 °C in a wet chamber. The grids were then washed with droplets of PBS for 10 min 3 times. Finally, the expressions were displayed with anti-rabbit IgG antibody conjugated to 15 nm gold particles (25112 Electron microscopy sciences, Hatfield, PA, USA). The grids were first stained with $2\%$ uranyl acetate and then lead citrate, washed with water, and air dried. The sections were observed in an electron microscope (Model JEM-1011) at 80 kV (JEOL Ltd., Tokyo, Japan) and analyzed with AMT-5.42.391 software. Ten random fields were taken over a total area of 62.5 microns at ×50,000 magnification. ## 4.12. Statistical Analysis The program GraphPad Prism version 9.4 (GraphPad software, La Jolla, CA, USA) was used to generate graphs and to perform statistical analyses. Results are expressed as mean ± standard error of the mean (SEM). Statistical significance was determined by one-way ANOVA followed by Tukey’s post-hoc test and Kruskal–Wallis if the data were normally distributed or not. The differences were considered when $p \leq 0.05.$ ## 5. Conclusions The PPARα agonist clofibrate has a beneficial cardiac effect on ischemia/reperfusion conditions by decreasing the inflammatory response that, in turn, decreases myocardial fibrosis, the MMP-2 expression, and apoptosis. As a consequence of these processes, the ANP/ANPr signaling is diminished as a compensatory response. Our study provides important insight into the regulation of both ANP and ANPr in the hearts of MetS rats by clofibrate. ## References 1. Shipman K.E., Strange R.C., Ramachandran S.. Use of fibrates in the metabolic syndrome: A review. World J. Diabetes (2016) 7 74-88. DOI: 10.4239/wjd.v7.i5.74 2. Rochlani Y., Pothineni N.V., Kovelamudi S., Mehta J.L.. Metabolic syndrome: Pathophysiology, management, and modulation by natural compounds. Ther. Adv. Cardiovasc. Dis. (2017) 11 215-225. DOI: 10.1177/1753944717711379 3. Lee E.Y., Han K., Kim D.H., Park Y.M., Kwon H.S., Yoon K.H., Kim M.K., Lee S.H.. Exposure-weighted scoring for metabolic syndrome and the risk of myocardial infarction and stroke: A nationwide population-based study. Cardiovasc. Diabetol. (2020) 19 153. DOI: 10.1186/s12933-020-01129-x 4. Lafuse W.P., Wozniak D.J., Rajaram M.V.S.. Role of cardiac macrophages on cardiac inflammation, fibrosis and tissue repair. Cells (2020) 10. DOI: 10.3390/cells10010051 5. Krylatov A.V., Tsibulnikov S.Y., Mukhomedzyanov A.V., Boshchenko A.A., Goldberg V.E., Jaggi A.S., Erben R.G., Maslov L.N.. The role of natriuretic peptides in the regulation of cardiac tolerance to ischemia/reperfusion and postinfarction heart remodeling. J. Cardiovasc. Pharmacol. Ther. (2021) 26 131-148. DOI: 10.1177/1074248420952243 6. Zhao W., Zhao J., Rong J.. Pharmacological modulation of cardiac remodeling after myocardial infarction. Oxid. Med. Cell Longev. (2020) 2020 8815349. DOI: 10.1155/2020/8815349 7. Ibrahim N.E., Januzzi J.L.. Established and emerging roles of biomarkers in heart failure. Circ. Res. (2018) 123 614-629. DOI: 10.1161/CIRCRESAHA.118.312706 8. Burley D.S., Hamid S.A., Baxter G.F.. Cardioprotective actions of peptide hormones in myocardial ischemia. Heart Fail. Rev. (2007) 12 279-291. DOI: 10.1007/s10741-007-9029-y 9. Verboven K., Hansen D., Jocken J.W.E., Blaak E.E.. Natriuretic peptides in the control of lipid metabolism and insulin sensitivity. Obes. Rev. (2017) 18 1243-1259. DOI: 10.1111/obr.12598 10. Schlueter N., de Sterke A., Willmes D.M., Spranger J., Jordan J., Birkenfeld A.L.. Metabolic actions of natriuretic peptides and therapeutic potential in the metabolic syndrome. Pharmacol. Ther. (2014) 144 12-27. DOI: 10.1016/j.pharmthera.2014.04.007 11. Santhekadur P.K., Kumar D.P., Seneshaw M., Mirshahi F., Sanyal A.J.. The multifaceted role of natriuretic peptides in metabolic syndrome. Biomed. Pharmacother. (2017) 92 826-835. DOI: 10.1016/j.biopha.2017.05.136 12. D’Souza S.P., Davis M., Baxter G.F.. Autocrine and paracrine actions of natriuretic peptides in the heart. Pharmacol. Ther. (2004) 101 113-129. DOI: 10.1016/j.pharmthera.2003.11.001 13. Nakagawa Y., Nishikimi T., Kuwahara K.. Atrial and brain natriuretic peptides: Hormones secreted from the heart. Peptides (2019) 111 18-25. DOI: 10.1016/j.peptides.2018.05.012 14. Rubattu S., Sciarretta S., Ciavarella G.M., Venturelli V., De Paolis P., Tocci G., De Biase L., Ferrucci A., Volpe M.. Reduced levels of N-terminal-proatrial natriuretic peptide in hypertensive patients with metabolic syndrome and their relationship with left ventricular mass. J. Hypertens. (2007) 25 833-839. DOI: 10.1097/HJH.0b013e32803cae3c 15. Ramos H.R., Birkenfeld A.L., de Bold A.J.. Cardiac natriuretic peptides and obesity: Perspectives from an endocrinologist and a cardiologist. Endocr. Connect. (2015) 4 R25-R36. DOI: 10.1530/EC-15-0018 16. Li S., Yang B., Du Y., Lin Y., Liu J., Huang S., Zhang A., Jia Z., Zhang Y.. Targeting PPARα for the treatment and understanding of cardiovascular diseases. Cell. Physiol. Biochem. (2018) 51 2760-2775. DOI: 10.1159/000495969 17. Ibarra-Lara L., Sánchez-Aguilar M., Soria-Castro E., Vargas-Barrón J., Roldán F.J., Pavón N., Torres-Narváez J.C., Cervantes-Pérez L.G., Pastelín-Hernández G., Sánchez-Mendoza A.. Clofibrate treatment decreases inflammation and reverses myocardial infarction-induced remodelation in a rodent experimental model. Molecules (2019) 24. DOI: 10.3390/molecules24020270 18. Ibarra-Lara L., Sánchez-Aguilar M., Sánchez-Mendoza A., Del Valle-Mondragón L., Soria-Castro E., Carreón-Torres E., Díaz-Díaz E., Vázquez-Meza H., Guarner-Lans V., Rubio-Ruiz M.E.. Fenofibrate therapy restores antioxidant protection and improves myocardial insulin resistance in a rat model of metabolic syndrome and myocardial ischemia: The role of Angiotensin II. Molecules (2016) 22. DOI: 10.3390/molecules22010031 19. Frank A., Bonney M., Bonney S., Weitzel L., Koeppen M., Eckle T.. Myocardial ischemia reperfusion injury: From basic science to clinical bedside. Semin. Cardiothorac. Vasc. Anesth. (2012) 16 123-132. DOI: 10.1177/1089253211436350 20. Rahmutula D., Cui J., Chen S., Gardner D.G.. Transcriptional regulation of type B human natriuretic Peptide receptor gene promoter: Dependence on Sp1. Hypertension (2004) 44 283-288. DOI: 10.1161/01.HYP.0000136908.60317.92 21. Gardner D.G., Chen S., Glenn D.J., Grigsby C.L.. Molecular biology of the natriuretic peptide system: Implications for physiology and hypertension. Hypertension (2007) 49 419-426. DOI: 10.1161/01.HYP.0000258532.07418.fa 22. Oidor-Chan V.H., Hong E., Pérez-Severiano F., Montes S., Torres-Narváez J.C., Del Valle-Mondragón L., Pastelín-Hernández G., Sánchez-Mendoza A.. Fenofibrate plus Metformin produces cardioprotection in a Type 2 diabetes and acute myocardial infarction model. PPAR Res. (2016) 2016 8237264. DOI: 10.1155/2016/8237264 23. Sun B., Xie Y., Jiang J., Wang Y., Xu X., Zhao C., Huang F.. Pleiotropic effects of fenofibrate therapy on rats with hypertriglycemia. Lipids. Health Dis. (2015) 14 27. DOI: 10.1186/s12944-015-0032-3 24. Lindsey M.L., Bolli R., Canty J.M., Du X.J., Frangogiannis N.G., Frantz S., Gourdie R.G., Holmes J.W., Jones S.P., Kloner R.A.. Guidelines for experimental models of myocardial ischemia and infarction. Am. J. Physiol. Heart Circ. Physiol. (2018) 314 H812-H838. DOI: 10.1152/ajpheart.00335.2017 25. Fernández-Sada E., Torres-Quintanilla A., Silva-Platas C., García N., Willis B.C., Rodríguez-Rodríguez C., De la Peña E., Bernal-Ramírez J., Treviño-Saldaña N., Oropeza-Almazán Y.. Proinflammatory cytokines are soluble mediators linked with ventricular arrhythmias and contractile dysfunction in a rat model of metabolic syndrome. Oxid. Med. Cell Longev. (2017) 2017 7682569. DOI: 10.1155/2017/7682569 26. Rubio-Ruiz M.E., Guarner-Lans V., Lopez-García C., Pérez-González P.. Inflammation and the use of anti-inflammatory agents in signs and cardiovascular consequences of metabolic syndrome. Handbook on Metabolic Syndrome. Classification, Risk Factors and Health Impact (2012) 169-188 27. Jiang N., Li Y., Shu T., Wang J.. Cytokines and inflammation in adipogenesis: An updated review. Front. Med. (2019) 13 314-329. DOI: 10.1007/s11684-018-0625-0 28. Brown J.D., Plutzky J.. Peroxisome proliferator-activated receptors as transcriptional nodal points and therapeutic targets. Circulation (2007) 115 518-533. DOI: 10.1161/CIRCULATIONAHA.104.475673 29. Berger J., Moller D.E.. The mechanisms of action of PPARs. Annu. Rev. Med. (2002) 53 409-435. DOI: 10.1146/annurev.med.53.082901.104018 30. Peterson E.A., Sun J., Wang J.. Leukocyte-mediated cardiac repair after myocardial infarction in Non-regenerative vs. regenerative systems. J. Cardiovasc. Dev. Dis. (2022) 9. DOI: 10.3390/jcdd9020063 31. Kruszewska J., Cudnoch-Jedrzejewska A., Czarzasta K.. Remodeling and fibrosis of the cardiac muscle in the course of obesity-pathogenesis and involvement of the extracellular matrix. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23084195 32. Nishida K., Otsu K.. Inflammation and metabolic cardiomyopathy. Cardiovasc. Res. (2017) 113 389-398. DOI: 10.1093/cvr/cvx012 33. Tuleta I., Frangogiannis N.G.. Fibrosis of the diabetic heart: Clinical significance, molecular mechanisms, and therapeutic opportunities. Adv. Drug Deliv. Rev. (2021) 176 113904. DOI: 10.1016/j.addr.2021.113904 34. Liu X., Shan X., Chen H., Li Z., Zhao P., Zhang C., Guo W., Xu M., Lu R.. Stachydrine ameliorates cardiac fibrosis through inhibition of Angiotensin II/Transformation Growth Factor 1 fibrogenic axis. Front. Pharmacol. (2019) 10 538. DOI: 10.3389/fphar.2019.00538 35. Wilson E.M., Spinale F.G.. Myocardial remodelling and matrix metalloproteinases in heart failure: Turmoil within the interstitium. Ann. Med. (2001) 33 623-634. DOI: 10.3109/07853890109002108 36. Cheung P.Y., Sawicki G., Wozniak M., Wang W., Radomski M.W., Schulz R.. Matrix metalloproteinase-2 contributes to ischemia-reperfusion injury in the heart. Circulation (2000) 101 1833-1839. DOI: 10.1161/01.CIR.101.15.1833 37. Gonçalves P.R., Nascimento L.D., Gerlach R.F., Rodrigues K.E., Prado A.F.. Matrix Metalloproteinase 2 as a pharmacological target in heart failure. Pharmaceuticals (2022) 15. DOI: 10.3390/ph15080920 38. Tsuruda T., Boerrigter G., Huntley B.K., Noser J.A., Cataliotti A., Costello-Boerrigter L.C., Chen H.H., Burnett J.C.. Brain natriuretic peptide is produced in cardiac fibroblasts and induces matrix metalloproteinases. Circ. Res. (2002) 91 1127-1134. DOI: 10.1161/01.RES.0000046234.73401.70 39. Yue T.L., Bao W., Jucker B.M., Gu J.L., Romanic A.M., Brown P.J., Cui J., Thudium D.T., Boyce R., Burns-Kurtis C.L.. Activation of peroxisome proliferator-activated receptor-alpha protects the heart from ischemia/reperfusion injury. Circulation (2003) 108 2393-2399. DOI: 10.1161/01.CIR.0000093187.42015.6C 40. Chen J., Tung C.H., Allport J.R.. Near-infrared fluorescent imaging of matrix metalloproteinase activity after myocardial infarction. Circulation (2005) 111 1800-1805. DOI: 10.1161/01.CIR.0000160936.91849.9F 41. Jian L., Fan Q., Ding T., Sun L., Liu S., Niu S.. Correlation between matrix metalloproteinases and brain natriuretic peptide in patients with chronic heart failure. Chin. J. Geriatr. (2017) 12 739-741 42. Fish-Trotter H., Ferguson J.F., Patel N., Arora P., Allen N.B., Bachmann K.N., Daniels L.B., Reilly M.P., Lima J.A.C., Wang T.J.. Inflammation and circulating natriuretic peptide levels. Circ. Heart Fail. (2020) 13 e006570. DOI: 10.1161/CIRCHEARTFAILURE.119.006570 43. Ramakrishnan V., Burnett J.C.. Natriuretic peptides, inflammation, and sounding the alarm. Circ. Heart Fail. (2020) 13 e007208. DOI: 10.1161/CIRCHEARTFAILURE.120.007208 44. Brigadeau F., Gelé P., Wibaux M., Marquié C., Martin-Nizard F., Torpier G., Fruchart J.C., Staels B., Duriez P., Lacroix D.. The PPARalpha activator fenofibrate slows down the progression of the left ventricular dysfunction in porcine tachycardia-induced cardiomyopathy. J. Cardiovasc. Pharmacol. (2007) 49 408-415. DOI: 10.1097/FJC.0b013e3180544540 45. Fu S., Ping P., Wang F., Luo L.. Synthesis, secretion, function, metabolism and application of natriuretic peptides in heart failure. J. Biol. Eng. (2018) 12 2. DOI: 10.1186/s13036-017-0093-0 46. Kuwahara K.. The natriuretic peptide system in heart failure: Diagnostic and therapeutic implications. Pharmacol. Ther. (2021) 227 107863. DOI: 10.1016/j.pharmthera.2021.107863 47. Pandey K.N., Nguyen H.T., Sharma G.D., Shi S.J., Kriegel A.M.. Ligand-regulated internalization, trafficking, and down-regulation of guanylyl cyclase/atrial natriuretic peptide receptor-A in human embryonic kidney 293 cells. J. Biol. Chem. (2002) 277 4618-4627. DOI: 10.1074/jbc.M106436200 48. Calamera G., Moltzau L.R., Levy F.O., Andressen K.W.. Phosphodiesterases and compartmentation of cAMP and cGMP signaling in regulation of cardiac contractility in normal and failing hearts. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23042145 49. Sun Y., Zhang Y., Yan M., Wu Y., Zheng X.. B-type natriuretic peptide-induced cardioprotection against reperfusion is associated with attenuation of mitochondrial permeability transition. Biol. Pharm. Bull. (2009) 32 1545-1551. DOI: 10.1248/bpb.32.1545 50. Domondon M., Nikiforova A.B., DeLeon-Pennell K.Y., Ilatovskaya D.V.. Regulation of mitochondria function by natriuretic peptides. Am. J. Physiol. Renal. Physiol. (2019) 317 F1164-F1168. DOI: 10.1152/ajprenal.00384.2019 51. Cao Z., Jia Y., Zhu B.. BNP and NT-proBNP as diagnostic biomarkers for cardiac dysfunction in both clinical and forensic medicine. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20081820 52. Hall C.. Essential biochemistry and physiology of (NT-pro)BNP. Eur. J. Heart Fail. (2004) 6 257-260. DOI: 10.1016/j.ejheart.2003.12.015 53. Sergeeva I.A., Christoffels V.M.. Regulation of expression of atrial and brain natriuretic peptide, biomarkers for heart development and disease. Biochim. Biophys. Acta. (2013) 1832 2403-2413. DOI: 10.1016/j.bbadis.2013.07.003 54. Cao Z.P., Zhang Y., Mi L., Luo X.Y., Tian M.H., Zhu B.L.. The expression of B-Type natriuretic peptide after CaCl2-induced arrhythmias in rats. Am. J. Forensic. Med. Pathol. (2016) 37 133-140. DOI: 10.1097/PAF.0000000000000193 55. Hasić S., Hadzović-Dzuvo A., Jadrić R., Kiseljakovic E.. B-type natriuretic peptide and adiponectin releases in rat model of myocardial damage induced by isoproterenol administration. Bosn. J. Basic. Med. Sci. (2013) 13 225-229. DOI: 10.17305/bjbms.2013.2329 56. Potter L.R., Yoder A.R., Flora D.R., Antos L.K., Dickey D.M.. Natriuretic peptides: Their structures, receptors, physiologic functions and therapeutic applications. Handb. Exp. Pharmacol. (2009) 191 341-366 57. Richards A.M.. New biomarkers in heart failure: Applications in diagnosis, prognosis and guidance of therapy. Rev. Esp. Cardiol. (2010) 63 635-639. DOI: 10.1016/S1885-5857(10)70137-7 58. Omran F., Kyrou I., Osman F., Lim V.G., Randeva H.S., Chatha K.. Cardiovascular biomarkers: Lessons of the past and prospects for the future. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23105680 59. Hayakawa K., Takemura G., Kanoh M., Li Y., Koda M., Kawase Y., Maruyama R., Okada H., Minatoguchi S., Fujiwara T.. Inhibition of granulation tissue cell apoptosis during the subacute stage of myocardial infarction improves cardiac remodeling and dysfunction at the chronic stage. Circulation (2003) 108 104-109. DOI: 10.1161/01.CIR.0000074225.62168.68 60. Sánchez-Aguilar M., Ibarra-Lara L., Del Valle-Mondragón L., Soria-Castro E., Torres-Narváez J.C., Carreón-Torres E., Sánchez-Mendoza A., Rubio-Ruíz M.E.. Nonclassical axis of the renin-angiotensin system and neprilysin: Key mediators that underlie the cardioprotective effect of PPAR-Alpha activation during myocardial ischemia in a metabolic syndrome model. PPAR Res. (2020) 2020 8894525. DOI: 10.1155/2020/8894525 61. Bradford M.M.. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. (1976) 72 248-254. DOI: 10.1016/0003-2697(76)90527-3 62. Elami A., Gavish Z., Korach A., Houminer E., Schneider A., Schwalb H., Arav A.. Successful restoration of function of frozen and thawed isolated rat hearts. J. Thorac. Cardiovasc. Surg. (2008) 135 666-672. DOI: 10.1016/j.jtcvs.2007.08.056 63. Marques L.S., Fossati A.A.N., Rodrigues R.B., Da Rosa H.T., Izaguirry A.P., Ramalho J.B., Moreira J.C.F., Santos F.W., Zhang T., Streit D.P.. Slow freezing versus vitrification for the cryopreservation of zebrafish (. Sci. Rep. (2019) 9 15353. DOI: 10.1038/s41598-019-51696-7 64. Rivas Leonel E.C., Lucci C.M., Amorim C.A.. Cryopreservation of human ovarian tissue: A review. Transfus. Med. Hemother. (2019) 46 173-181. DOI: 10.1159/000499054 65. Lu M., Qin Q., Yao J., Sun L., Qin X.. Induction of LOX by TGF-β1/Smad/AP-1 signaling aggravates rat myocardial fibrosis and heart failure. IUBMB Life (2019) 71 1729-1739. DOI: 10.1002/iub.2112 66. Wegner Kyle A., Keikhosravi A., Eliceiri K.W., Vezina C.M.. Fluorescence of picrosirius red multiplexed with immuno-histochemistry for the quantitative assessment of collagen in tissue sections. J Histochem Cytochem. (2017) 65 479-490. DOI: 10.1369/0022155417718541 67. Butler J.L., Barham B.J., Heidenreich B.A.. Comparison of indirect peroxidase and avidin-biotin-peroxidase complex (ABC) immunohistochemical staining procedures for c-fos in rat brain. J. Anat. (2019) 234 936-942. DOI: 10.1111/joa.12967 68. Honvo-Houéto E., Truchet S.. Indirect immunofluorescence on frozen sections of mouse mammary gland. J. Vis. Exp. (2015) 106 e53179. DOI: 10.3791/53179 69. Im K., Mareninov S., Diazm M.F.P., Yong W.H.. An introduction to performing immunofluorescence staining. Methods Mol. Biol. (2019) 1897 299-311. DOI: 10.1007/978-1-4939-8935-5_26 70. Oberhaus S.M.. TUNEL and immunofluorescence double-labeling assay for apoptotic cells with specific antigen(s). Methods Mol. Biol. (2003) 218 85-96. DOI: 10.1385/1-59259-356-9:85 71. Kyrylkova K., Kyryachenko S., Biehs B., Klein O., Kioussi C., Leid M.. BCL11B regulates epithelial proliferation and asymmetric development of the mouse mandibular incisor. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0037670 72. Soria-Castro E., Soto M.E., Guarner-Lans V., Rojas G., Perezpeña-Diazconti M., Críales-Vera S.A., Manzano Pech L., Pérez-Torres I.. The kidnapping of mitochondrial function associated with the SARS-CoV-2 infection. Histol. Histopathol. (2021) 36 947-965. DOI: 10.14670/HH-18-354
--- title: 'Pleurotus pulmonarius Strain: Arsenic(III)/Cadmium(II) Accumulation, Tolerance, and Simulation Application in Environmental Remediation' authors: - Yuhui Zhang - Xiaohong Chen - Ling Xie journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049176 doi: 10.3390/ijerph20065056 license: CC BY 4.0 --- # Pleurotus pulmonarius Strain: Arsenic(III)/Cadmium(II) Accumulation, Tolerance, and Simulation Application in Environmental Remediation ## Abstract The arsenic (As, III) and cadmium (Cd, II) accumulation and tolerance traits of a new strain Pleurotus pulmonarius MT were evaluated, and the utilization of the strain for repairing contaminated liquid and soil was explored. The hypha cultivated in potato dextrose agar (PDA) exhibited medium or high Cd accumulation (0 to 320 mg/L), medium Cd tolerance (maximum tolerated concentration, MTC ≥ 640 mg/L), medium As accumulation (0 to 80 mg/L), and high As tolerance (MTC > 1280 mg/L). The hypha has application potential in processes related to the removal of Cd and As in aqueous pollutants at concentrations of 80 mg/L Cd and 20 mg/L As. The trends obtained for the fruiting bodies of P. pulmonarius MT seemed to deviate from those of the hypha of this strain. The results show that the fruiting bodies featured medium As accumulation (0 to 40 mg/kg), medium As tolerance (MTC > 160 mg/kg), medium Cd accumulation (0 to 10 mg/kg), and high Cd tolerance (MTC > 1280 mg/kg). The fruiting bodies of P. pulmonarius MT were utilized in processes related to the recovery of Cd and As in substrates, that is, $12\%$ contaminated soil mixed with 50 mg/kg Cd and 200 mg/kg As; thus, the hypha and fruiting bodies of P. pulmonarius MT can be used for the decontamination of water and soil containing As(III) and Cd(II). ## 1. Introduction The issue of soil and water pollution has been a persistent concern for humans, leading to adverse and long-lasting impacts. More than 20 million hectares of soil have been contaminated worldwide by heavy metals (HMs) and other metalloid pollutants [1]. Water pollution accounts for deaths of more than 14,000 people daily and 1.6 million children each year worldwide [2,3]. Rapid global industrial development, together with the long-term use of fertilizers and pesticides have led to a significantly increased risk of environmental contamination with HMs [4,5]. Some non-essential HMs such as cadmium (Cd), mercury (Hg), and silver (Ag), which are toxic and useless to plants, can adversely affect the soil quality and reduce crop production [6]. These pollutants are also major causes of life-threatening degenerative diseases affecting humans such as cancer, liver injury, and immune-related and inflammatory diseases [7,8,9,10]. Among these above, Cd is the most widespread and dangerous for living organisms; it reduces plant growth or causes plant death [11] and causes human diseases such as kidney disease and several cancers [12]. As a metalloid, arsenic (As) also poses a significant toxicity threat to plants, humans, and animals, given that it is a food chain contaminant [13]. The removal of toxic trace elements (TTEs)from the environment has become a major issue, and removal technologies have received increasing attention. Biosorption is the ability of biological materials to remove TTEs through metabolic mediation or absorption. It has several inherent advantages, including a more complete removal, lower economic cost, higher feasibility, and higher safety, different from traditional methods, including chemical precipitation, membrane separation, and ultrafiltration [14]. Biosorption mainly consists of phytoremediation [15] and mycoremediation [16]. Compared with phytoremediation, a plant-based technology for the removal of TTEs, mycoremediation (bioremediation using fungi) is a more promising method because of its low cost, short remediation period, and high efficiency [17,18]. Mycoremediation also has limitations such as its sensitivity to high levels of humidity as well as extremely high or low temperatures. Several studies reported the outstanding accumulation of high levels of Cd, As, and other TTEs by fungi such as Agaricus brasiliensis [19], Pleurotus ostreatus [20,21], Lentinula edodes [22], Trichoderma [23], and Phanerochaete chrysosporium [24]. Some reports indicated that a microorganism strain normally featured capacities of high accumulation and low tolerance, low accumulation and high tolerance, or low accumulation and low tolerance [19,25]. It is a slight possibility that these microorganisms will be applied for TTEs removal until researchers find a microorganism strain that simultaneously has high accumulation and medium tolerance capacities or medium accumulation and high tolerance capacities. In most cases, two or more kinds of TTEs always exist in environmental pollutants. It is necessary to explore some efficient techniques or fungi for removing two or more kinds of TTEs. Furthermore, most studies on the association between fungi and HMs pollution only investigated a specific stage of fungal growth (mycelia or fruiting body stages), such as P. ostreatus [21] and L. edodes [22]. Consequently, the applicability of mycelial TTEs stress patterns to the fruiting bodies of the same fungal strain remains uncertain. This creates practical limitations, particularly in simulated soil experiments, due to the sensitivity of macro fungi to high soil content. While some reports on the use of P. leurotus sp. such as P. ostreatus [21] and P. eryngii [26] for micro-mediation exist, the potential of P. pulmonarius in this domain requires further investigation. Our preliminary experiments indicated that the P. pulmonarius MT strain we collected in the wild could thrive in the substrate containing significant amounts of soil, increasing the likelihood of success in our simulated soil experiments. Our current study deals with evaluation of accumulation and tolerance to Cd (II) and As (III) in hypha and fruiting bodies of P. pulmonarius MT, and mycoremediation feasibility for removal of Cd and As by simulated soil experiments. ## 2.1. Strain, Chemicals, Medium, and Substrates The P. pulmonarius MT strain (CCTCC M 2021011) obtained via tissue isolation from an abandoned cultivation base for edible mushrooms contaminated with HMs (Changsha, Hunan Province, China) was used in this study. This strain was stored in the China Center for Type Culture Collection, Wuhan University, Wuhan, China, and has been filed for Chinese patent application (application no. 202110040646.2, State Intellectual Property Office of P. R. China). The hypha was initially incubated at 25 °C for 2 weeks in potato dextrose agar (PDA) containing 12 g/L potato extract, 20 g/L glucose, and 20 g/L agar (Solarbio, China). Furthermore, $99.4\%$ wheat grains boiled in water for 0.5 h, $0.3\%$ calcium carbonate (CaCO3), and $0.3\%$ calcium sulfate (CaSO4) (w/w) were mixed as substrates for strain cultivation. Substrates were packed in glass bottles, and 6 mm hypha blocks were transplanted on the substrates. The cultivation strain was incubated at 25 °C until the grains were completely covered with hypha. Deionized water was used during all experiments. The As and Cd concentrations were controlled in PDA, potato dextrose broth (PDB), wheat grains, cottonseeds hull, corncob, and unpolluted soil (As ≤ 1 ppm; Cd ≤ 1 ppm; dry weight). ## 2.2. Preparation of HM Salts Solutions CdCl2 and NaAsO2 were pre-dried to achieve constant weight, and then 8.15 g and 8.67 g of CdCl2 and NaAsO2, respectively, were dissolved in 100 mL of deionized water. Then, 50 g/L of Cd (II) and As (III) solutions were prepared and stored at 4 °C. ## 2.3. Cd (II) or As (III) Stress Treatment for Hypha The hypha of P. pulmonarius MT under Cd stress was monitored on a PDA medium supplemented with different Cd concentrations (Cd2+: 0, 80, 160, 320, 640, 1280, and 2560 mg/L). Under As stress, there were other seven groups of different As concentrations (As3+: 0, 80, 160, 320, 640, 1280, and 2560 mg/L). Each treatment was conducted four times. The addition methods for As were the same as those for Cd. Hypha block (6 mm) was then inoculated on the PDA of every plate and incubated for 2 weeks at 25 °C. After 14 days, the colony diameter was measured, and the hypha inhibition rate (HIR) was calculated. The hypha was harvested from the PDA surface, dried at 60 °C for 24 h, and ground. Approximately 0.1 g of dry samples was collected and then digested with 7 mL HNO3 and 1.5 mL H2O2 via a microwave digestion system before inductively coupled plasma–mass spectrometry (ICP-MS, Agilent 7900, Waldbronn, Germany) analysis. The process for the solid medium after incubation was the same as that of the hypha above, with drying at 105 °C for 24–36 h, and 0.2 g of dry samples was collected. The As or Cd concentrations in all samples were determined via ICP-MS [27]. The calibration range was 0–100 μg/L for As and Cd. HIR (%) = (Wc − Wa)/Wc × $100\%$, where Wc (g) represents the weight of the fresh hyphae in the control (CK) group; and Wa (g) represents the weight of the fresh hyphae cultivated under TTEs stress. ## 2.4. Remediation Experiment of Contaminated Liquid with Cd and As Different concentrations of an artificially contaminated liquid medium with Cd and As (0 and 0, 80 and 20, 160 and 40, 20 and 80 mg/L) were separately obtained after PDB was mixed with As and Cd solutions. Each treatment was conducted for four repetitions in 250 mL flasks with 100 mL liquid. After four fungi blocks (6 mm) were added into the liquid, the mycelia of P. pulmonarius MT was monitored in PDB supplemented with different concentrations of Cd and As for 2 weeks at 25 °C and 120 r/min. After 14 days, the mycelia was filtered, collected, washed with deionized water three times, dried at 60 °C for 24 h, and weighed as biological yield. Then, 5 mL samples of the liquid medium before and after incubation were separately collected, vaporized to obtain dry samples, and then prepared as mycelia under Cd or As stress. The Cd and As amounts were analyzed via ICP-MS. ## 2.5. Cd or As Stress Treatment for Fruiting Bodies and Cultivation Management First, $85\%$ cotton seeds hull (dw), $12\%$ corncob (dw), and $3\%$ CaCO3 (w/w) were mixed as cultivation substrates for the fruiting bodies of P. pulmonarius MT. The fruiting bodies under Cd stress were monitored on the substrates above supplemented with different concentrations (Cd2+: 0, 5, 10, 20, 40, 80, 160, 320, 640, and 1280 mg/kg). Under As stress, there were 10 other groups of different concentrations (As3+: 0, 5, 10, 20, 40, 80, 160, 320, 640, and 1280 mg/kg). Each treatment was conducted for four repetitions. The detailed procedure is as follows: A total weight of 3 kg substrate ingredients from one treatment was placed in a porcelain container, mixed with about 300 mL diluents of Cd for pre-wetting until relative humidity of 55–$65\%$. The lack of liquid for pre-wetting was replaced with deionized water. The pH values of the substrates were adjusted to 7–7.5. The addition strategies for As were the same as that for Cd. The experimental site was the Hunan Engineering Research Center of Edible Fungi. Substrates were filled into polyethylene plastic bags (17 cm × 33 cm × 0.045 cm) and sterilized at 121 °C for 2 h. After the substrates were cooled, they were inoculated in each bag with 10 g hypha from bottles as described in the second paragraph of Section 2.1. Incubation was conducted at 25 °C and 80–$85\%$ (air humidity) for 20–30 days. At the fructification phase, cultivation was conducted at a temperature of 24–26 °C, relative air humidity of 90–$95\%$, and CO2 of 2000 ppm. Harvesting was usually performed early in the morning before spore ejection after another 20–30 day cultivation. As the first batch normally accounts for about $50\%$ of the total biological yield, the fruiting bodies of the first batch harvested were only weighed after drying at 60 °C for 24–48 h, and they were used for biological yield (g, dw) calculation and ICP-MS determination. Moreover, 0.2 g fruiting bodies, substrates before cultivation, and substrates after cultivation were separately prepared via the same approach as above before the Cd or As amount was evaluated via ICP-MS. All the samples were dry. ## 2.6. Remediation Experiment of Cd- and As-Contaminated Soil First, 5 g/L of Cd solution was obtained after 16 mL, 50 g/L of Cd solution was mixed with 144 mL deionized water. Then, 1.25 g/L of As solution was also obtained after 4 mL, 50 g/L of As solution was mixed with 156 mL deionized water. Then, 4 kg of artificially contaminated soil (pH = 6.7) with 50 mg/kg of Cd and 200 mg/kg of As was obtained after the soil was completely mixed with two diluted solutions as described in Section 2.2. After the contaminated soil was dried at 105 °C for 48–72 h, it was collected and stored in a dry environment. Except for the CK group, four groups of substrates were mixed with different proportions of the contaminated soil described above before being filled into plastic bags. The proportions of contaminated soil were $12\%$, $24\%$, $36\%$, and $48\%$. The cultivation process was the same as that for the fruiting bodies above. In this experiment, the whole cultivation time was about 90 d. The preparation process for all samples was the same as that for the fruiting bodies above. The Cd and As amounts were analyzed via ICP-MS. ## 2.7. Statistical Analysis Statistical analysis was performed via analysis of variance. The Statistical Package for Social Sciences 18.0 software package (USA) was used to perform statistical analysis, and differences with p values of ≤0.05 were considered statistically significant. ## 3.1. Effect of Cd Stress on the Hypha of P. pulmonarius MT The results of Figure 1(A1–A4) reveal the effect of Cd at the tested levels for the hypha of P. pulmonarius MT. When the concentration of Cd stress in PDA is 80 mg/L, the colony diameter of the hypha cultivated for 2 weeks is slightly lower than that for control group (CK) ($p \leq 0.05$). This shows that the presence of ≥160 mg/L Cd significantly inhibits the colony diameter on agar ($p \leq 0.05$). Under the maximum tolerated concentration (MTC), the hypha growth is completely inhibited, and the HIR is about $93\%$ (89–6 mm/89 mm). Therefore, the MTC of Cd for P. pulmonarius MT in PDA is ≥640 mg/L. Although no significant difference in hypha growth is observed between CK and 80 mg/L treatment, the Cd content in hypha rapidly rises from 5.78 to 5619.62 mg/kg. The colony diameter for 160 mg/L treatment is reduced to 55.75 mm ($p \leq 0.05$), while the Cd content in hypha is enhanced up to 5712.77 mg/kg. The 320 mg/L treatment gives the highest values (7354.61 mg/kg) for hypha, and the 80 mg/L treatment gives the highest bio-enrichment coefficients (BCF, 418–591). ## 3.2. Effect of As Stress on the Hypha of P. pulmonarius MT When the As concentration ranges from 80 to 640 mg/L in PDA, the colony diameter is slightly lower than that of the CK group ($p \leq 0.05$) (Figure 1(B2,B3)). The results illustrate that the presence of ≥1280 mg/L As significantly inhibits the hypha growth on agar plates ($p \leq 0.05$). Moreover, the MTC of As for P. pulmonarius MT in PDA is >1280 mg/L. The As content in the hypha increases from 9.69 to 402.87 mg/kg with increasing As concentration in the PDA (0–1280 mg/L) (Figure 1(B4)). For the same concentration of As treatment, the As content in the hypha is lower than that in the PDA after incubation, except for CK and 80 mg/L treatments. In other words, only P. pulmonarius MT under a low concentration of As stress (≤80 mg/L) can transfer more As ions from PDA into the hypha and shows a trait of low As accumulation (BCF, 1.4–1.5) or medium As accumulation (BCF, 66.9–125). ## 3.3. Remediation Effect of P. pulmonarius MT on Cd- and As-Contaminated Liquid Based on the traits of medium or high Cd accumulation (0 to 320 mg/L) and medium Cd tolerance (MTC ≥ 640 mg/L) in hypha cultivated in PDA and medium As accumulation (0 to 80 mg/L) and As high tolerance (MTC >1280 mg/L), the hypha of P. pulmonarius MT was used to decontaminate liquid polluted with As and Cd. With the increase in the As and Cd concentrations in PDB to 20 and 80 mg/L, the Cd removal rate is about $98.99\%$ (Figure 2(C3)); the As is partly removed by $15.43\%$; and the mycelia growth is not significantly inhibited ($p \leq 0.05$, Figure 2(C1,C3)). Meanwhile, the Cd content detected in PDB for treatment is 0.77 mg/L after incubation (Figure 2(C2)). Therefore, the Cd concentration in contaminated liquid (1–80 mg/L Cd) can reach the effluent standard after treatment using P. pulmonarius MT. Moreover, the Cd removal rate for treatment (ii) is about $75.71\%$, and As is partly removed (Figure 2(C3)), although the mycelia growth is significantly inhibited ($p \leq 0.05$). Also, compared to the CK treatment, the size of hypha pellets for treatments (ii) and (iii) (Figure 2(C1)) is reduced and rough selvedge disappears, which is related to the toxic effect of As and Cd. ## 3.4. Effect of Cd Stress on Fruiting Bodies of P. pulmonarius MT The rules applied for the hypha of P. pulmonarius MT under TTEs stress were not the same as those for the fruiting bodies of this strain. It seems that the As and Cd accumulation and tolerance in fruiting bodies differ from those in hypha in our study. As shown in Figure 3(D3), with Cd doses increased from 0 to 40 mg/kg in substrates, the BCF of Cd increases to 25.5–32.4 and then reduces to 1.4–1.8. Moreover, there is no significant difference in biological yield between any ≤320 mg kg−1 treatment and CK ($p \leq 0.05$, Figure 3(D1,D2)). Regarding the Cd tolerance capacity, Figure 3(D1,D2) illustrates that the presence of >1280 mg/kg Cd (MTC) completely inhibits the fruiting body growth. ## 3.5. Effect of As Stress on Fruiting Bodies of P. pulmonarius MT The MTC of Cd is >1280 mg/kg in the substrates of the fruiting bodies of P. pulmonarius MT, while the MTC of As is >160 mg/kg (Figure 3(E1,E2)). There is no significant difference in biological yield between any treatment and CK ($p \leq 0.05$) (0 to 80 mg/kg). When the As dose ranges from 0 to 160 mg/kg in the substrate, the BCF of As decreases from 81.5–106 to 2.7–3.2 (Figure 3(E3)), and the As accumulation capacity is medium (0 to 40 mg/kg) or low (40 to 160 mg/kg). ## 3.6. Remediation Effect of P. pulmonarius MT on Cd- and As-Contaminated Soil Based on the traits of medium As accumulation (0 to 40 mg/kg) and medium As tolerance (MTC > 160 mg/kg) in the fruiting bodies shown above and medium accumulation (0 to 10 mg/kg) and high tolerance to Cd (MTC > 1280 mg/kg), the fruiting bodies of P. pulmonarius MT were used to decontaminate As- and Cd-contaminated soil. When the As and Cd concentrations in contaminated substrates are 24 and 6 mg/kg, the As and Cd removal rates are $71.51\%$ and $90.79\%$, respectively (Figure 4(F3)). Moreover, the fruiting body growth is not significantly inhibited compared with that of CK ($p \leq 0.05$, Figure 4(F1,F2)). The strain for treatment II can take up about $15.58\%$ of Cd and $43.89\%$ of As of the substrates into the fruiting bodies, and the biological yield is not significantly different from that of CK ($p \leq 0.05$). Meanwhile, as shown in Figure 4(F1) II and III, the thickness of the pileus reduces, and the color changes from drab yellow to gray in fruiting bodies under treatment II (48 and 12 mg/kg for As and Cd, respectively) and treatment III (72 and 18 mg/kg for As and Cd, respectively), compared with those of CK treatment. This phenomenon is perhaps related to the toxic effect of the TTEs or nutritional deficiency. However, the results demonstrate that intoxication symptoms will not appear in the fruiting bodies under Cd of 10–20 mg/kg or As of 40–80 mg/kg (Figure 3(D1,E1)). We speculate that nutritional deficiency is the main cause of this phenomenon. ## 4. Discussion Mycoremediation is of considerable interest as a branch of low-cost and eco-friendly technology in contaminated aquatic systems and soil. In recent years, researchers have found some wild fungi resources with high- or hyper-accumulation capacity of TTEs, for example, Amanita strobiliformis and Suillus luteus [28,29]. Through biological techniques, researchers have also acclimated or bred some fungi strains in labs, including Pleurotus ostreatus HAU-2 [21] and Lentinula edodes W1 [30,31]. The non-MT Cd-binding protein from Lentinula edodes (LECBP) was investigated as a potential remediation tool for Cd biosorption in Escherichia coli. This research not only sheds light on the potential of LECBP for Cd bioremediation, but also contributes to a better understanding of the relationship between LECBP’s structure and functionality. Also, several studies focused on the influence factors and mechanisms of accumulation and tolerance [32,33,34]. It seems that most microorganism strains feature extreme accumulation and tolerance capacities to TTE (i.e., high accumulation and low tolerance, low accumulation and high tolerance, or low accumulation and low tolerance). Despite these advantages, the most important question is how to improve the accumulation and tolerance capacities of fungi, and obtaining a fungus strain featuring the two traits at reasonable degrees is important for environmental applications [35,36]. It is a slight possibility that these microorganisms will be applied for TTE removal until researchers find a microorganism strain that simultaneously has high accumulation and medium tolerance capacities or medium accumulation and high tolerance capacities. This paper aims to evaluate the TTEs tolerance and accumulation traits of a new strain P. pulmonarius MT to obtain a strain with TTEs high accumulation and medium tolerance or with TTEs medium accumulation and high tolerance. TTEs are among the major pollutants discharged by urban and agricultural runoffs, industrial effluents, mining, and other processes. Various TTEs such as Pb, Mn, As, Al, Cr, Cd, Co, Cu, Zn, Ni, and Fe have been detected in coal washery effluents [16]. In most cases, two or more kinds of TTEs always exist in environmental pollutants. It is necessary to explore some efficient techniques or fungi for removing two or more kinds of TTEs. A fungi strain that can simultaneously remove two or more kinds of TTEs in pollutants has rarely been reported. Laccaria bicolor has the potential to ameliorate the effects of non-essential Cd and essential Cu through different blocking strategies [37]. In one study, the fungus Pleurotus ostreatus HAU-2 could remove Cd and Cr in liquid culture through absorption [21]. Our study aims at obtaining an efficient fungus for decontaminating pollutants containing Cd and As. The application potential of Basidiomycota, including Pleurotus, an important edible mushroom, in TTEs accumulation has started to draw increasing attention. Pleurotus ostreatus has been reported to have promising applications for removing TTEs from washery effluent and absorbing Cd and Cr from soil [16,21]. Another study provided insights into the transcriptional response of Pleurotus eryngii to extremely high levels of HMs [26]. In Q. Li’s study, the upregulation of genes encoding putative oxidoreductases, dehydrogenases, reductases, transferases, and transcription factors following exogenous NO induction contributed to the increased tolerance of P. eryngii to high levels of HMs. This study sheds new light on the transcriptional response of P. eryngii to HMs and the role of NO in improving heavy metal tolerance. In our next study, we will attempt to delve into researching how P. pulmonarius MT responds to extremely high concentrations of TTEs stress by transcriptomics. One study evaluated the feasibility of removing Cu(II) and Zn(II) using ahybrid immobilized biosorbent of Pleurotus sajor-caju and Jasmine sambac [14]. Cihangir and Saglam observed that the Cd uptake rate by Pleurotus sajor-caju onto dry biomass was between $88.9\%$ and $91.8\%$ using aqueous media with concentrations ranging from 0.125 to 1.0 mM (about 14–112 mg/L), and this strain might successfully be utilized for the removal of TTEs in polluted water with 0.5 mM Cd2+ (about 56 mg/L) [38]. In this study, when the concentrations of Cd and As were 20 and 80 mg/L, respectively, the Cd removal rate was about $98.99\%$, and $15.43\%$ of As was removed; moreover, the hypha growth was not significantly inhibited. The trend of Cd accumulation in PDB in our research almost accords with that of Cihangir and Saglam’s report. The abilities of some filiform fungi, ectomycorrhiza, and yeast for TTEs removal from liquid and solid media have been confirmed [39,40,41]. It has been suggested that the hypha of P. pulmonarius MT might successfully be used in processes related to the removal and recovery of TTEs from aqueous pollutants. Edible mushrooms including P. pulmonarius MT have a striking advantage over other microorganisms as an efficient method of mycoremediation. This reflects a high biological efficiency of edible mushrooms. For example, the biological efficiency of P. pulmonarius MT is about $70\%$. In other words, the whole weight of 1 kg (dw) substrates used to cultivate P. pulmonarius MT resulted in a total biological yield of fruiting bodies of generally 0.70 kg (fw) after the harvest of three or four batches. This explains an unusual point in Figure 4(F3), why the removal rate of Cd ($90.79\%$) is higher than that of As ($71.51\%$) with As and Cd concentrations of 24 and 6 mg/kg in contaminated substrates, respectively; however, there is no large difference in the accumulation capacity between 20 mg/kg for As treatment (Figure 3(E1)) and 5 mg/kg for Cd treatment (Figure 3(D1)). ## 5. Conclusions The hypha of P. pulmonarius MT cultivated in PDA exhibit medium or high Cd accumulation (0 to 320 mg/L), medium Cd tolerance (MTC ≥ 640 mg/L), medium As accumulation (0 to 80 mg/L), and high As tolerance (MTC >1280 mg/L). Moreover, the hypha of P. pulmonarius MT was utilized in processes related to the removal of Cd and As present in aqueous pollutants at concentrations of 80 mg/L and 20 mg/L, respectively. The trends for the fruiting bodies of P. pulmonarius MT seem to deviate from those of the hypha of this strain. The results show that the fruiting bodies feature medium As accumulation (0 to 40 mg/kg), medium As tolerance (MTC > 160 mg/kg), medium Cd accumulation (0 to 10 mg/kg), and high Cd tolerance (MTC > 1280 mg/kg). The fruiting bodies were further utilized in processes related to the recovery of Cd and As in substrates, that is, $12\%$ contaminated soil mixed with 50 mg/kg Cd and 200 mg/kg As. According to the results, P. pulmonarius MT is a promising candidate for the remediation of As- and Cd-contaminated soil or water. Nevertheless, there are still some challenges in understanding the remediation mechanism of this strain, recycling the TTEs from contaminated hypha or fruiting bodies, and fully utilizing residuals. ## References 1. Thakare M., Sarma H., Datar S., Roy A., Pawar P., Gupta K., Pandit S., Prasad R.. Understanding the holistic approach to plant-microbe remediation technologies for removing heavy metals and radionuclides from soil. Curr. Res. Biotechnol. (2021) 3 84-98. DOI: 10.1016/j.crbiot.2021.02.004 2. Fernandez-Luqueno F., Lopez-Valdez F., Gamero-Melo P., Luna-Suarez S., Aguilera-Gonzalez E.N., Martínez A.I., García-Guillermo M., Hernandez-Martinez G., Herrera-Mendoza R., Álvarez-Garza M.A.. Heavy metal pollution in drinking water-a global risk for human health: A review. Afr. J. Environ. Sci. Technol. (2013) 7 567-584 3. Nachana’a Timothy E.T.W.. Environmental pollution by heavy metal: An overview. Chemistry (2019) 3 72-82 4. Vardhan K.H., Kumar P.S., Panda R.C.. A review on heavy metal pollution, toxicity and remedial measures: Current trends and future perspectives. J. Mol. Liq. (2019) 290 111197. DOI: 10.1016/j.molliq.2019.111197 5. Zwolak A., Sarzyńska M., Szpyrka E., Stawarczyk K.. Sources of soil pollution by heavy metals and their accumulation in vegetables: A review. Water Air Soil Pollut. (2019) 230 164. DOI: 10.1007/s11270-019-4221-y 6. Ojuederie O.B., Babalola O.O.. Microbial and plant-assisted bioremediation of heavy metal polluted environments: A review. Int. J. Environ. Res. Public Health (2017) 14. DOI: 10.3390/ijerph14121504 7. Andjelkovic M., Buha Djordjevic A., Antonijevic E., Antonijevic B., Stanic M., Kotur-Stevuljevic J., Spasojevic-Kalimanovska V., Jovanovic M., Boricic N., Wallace D.. Toxic effect of acute cadmium and lead exposure in rat blood, liver, and kidney. Int. J. Environ. Res. Public Health (2019) 16. DOI: 10.3390/ijerph16020274 8. Ganapathy S., Yu T., Makriyannis A., Chen C., Liu J., Xiong R.. Chronic low dose arsenic exposure preferentially perturbs mitotic phase of the cell cycle. Genes Cancer (2019) 10 39-51. DOI: 10.18632/genesandcancer.185 9. Siblerud R., Mutter J., Moore E., Naumann J., Walach H.. A hypothesis and evidence that mercury may be an etiological factor in Alzheimer’s disease. Int. J. Environ. Res. Public Health (2019) 16. DOI: 10.3390/ijerph16245152 10. Moradi F., Maleki V., Saleh-Ghadimi S., Kooshki F., Pourghassem Gargari B.. Potential roles of chromium on inflammatory biomarkers in diabetes: A Systematic. Clin. Exp. Pharmacol. Physiol. (2019) 46 975-983. DOI: 10.1111/1440-1681.13144 11. Huybrechts M., Cuypers A., Deckers J., Iven V., Vandionant S., Jozefczak M., Hendrix S.. Cadmium and plant development: An agony from seed to seed. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20163971 12. Olmedo P., Grau-Perez M., Fretts A., Tellez-Plaza M., Gil F., Yeh F., Umans J.G., Francesconi K.A., Goessler W., Franceschini N.. Dietary determinants of cadmium exposure in the Strong Heart Family Study. Food Chem. Toxicol. (2017) 100 239-246. DOI: 10.1016/j.fct.2016.12.015 13. Zhang X., Zhong T., Liu L., Ouyang X.. Impact of Soil Heavy Metal Pollution on Food Safety in China. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0135182 14. Ashraf A., Nadeem R., Sharif S., Ansari T.M., Munir H., Mahmood A.. Study of hybrid immobilized biomass of. Int. J. Environ. Sci. Technol. (2015) 12 717-724. DOI: 10.1007/s13762-013-0471-1 15. Pilon-Smits E.. Phytoremediation. Annu. Rev. Plant Biol. (2005) 56 15-39. DOI: 10.1146/annurev.arplant.56.032604.144214 16. Vaseem H., Singh V.K., Singh M.P.. Heavy metal pollution due to coal washery effluent and its decontamination using a macrofungus,. Ecotox. Environ. Safe (2017) 145 42-49. DOI: 10.1016/j.ecoenv.2017.07.001 17. Damodaran D., Balakrishnan R.M., Shetty V.K.. The uptake mechanism of Cd(II), Cr(VI), Cu(II), Pb(II), and Zn(II) by mycelia and fruiting bodies of. Biomed. Res. Int. (2013) 2013 149120. DOI: 10.1155/2013/149120 18. Liu S.H., Zeng G.M., Niu Q.Y., Liu Y., Zhou L., Jiang L.H., Tan X.F., Xu P., Zhang C., Cheng M.. Bioremediation mechanisms of combined pollution of PAHs and heavy metals by bacteria and fungi: A mini review. Bioresour. Technol. (2017) 224 25-33. DOI: 10.1016/j.biortech.2016.11.095 19. Liu P.H., Huang Z.X., Luo X.H., Chen H., Weng B.Q., Wang Y.X., Chen L.S.. Comparative transcriptome analysis reveals candidate genes related to cadmium accumulation and tolerance in two almond mushroom (. PLoS ONE (2020) 15. DOI: 10.1371/journal.pone.0239617 20. Alfaro M., Castanera R., Lavin J.L., Grigoriev I.V., Oguiza J.A., Ramirez L., Pisabarro A.G.. Comparative and transcriptional analysis of the predicted secretome in the lignocellulose-degrading basidiomycete fungus. Environ. Microbiol. (2016) 18 4710-4726. DOI: 10.1111/1462-2920.13360 21. Li X., Wang Y., Pan Y., Yu H., Zhang X., Shen Y., Jiao S., Wu K., La G., Yuan Y.. Mechanisms of Cd and Cr removal and tolerance by macrofungus. J. Hazard. Mater. (2017) 330 1-8. DOI: 10.1016/j.jhazmat.2017.01.047 22. Llorente-Mirandes T., Barbero M., Rubio R., López-Sánchez J.F.. Occurrence of inorganic arsenic in edible Shiitake (. Food Chem. (2014) 158 207-215. DOI: 10.1016/j.foodchem.2014.02.081 23. Tripathi P., Singh P.C., Mishra A., Tripathi R.D., Nautiyal C.S.. Trichoderma inoculation augments grain amino acids and mineral nutrients by modulating arsenic speciation and accumulation in chickpea (. Ecotoxicol. Environ. Saf. (2015) 117 72-80. DOI: 10.1016/j.ecoenv.2014.10.027 24. Xu P., Liang L., Zeng G., Huang D., Cui L., Zhao M., Chao H., Li N., Zhen W., Wu H.. Heavy metal-induced glutathione accumulation and its role in heavy metal detoxification in. Appl. Microbiol. Biotechnol. (2014) 98 6409. DOI: 10.1007/s00253-014-5667-x 25. Beneš V., Leonhardt T., Kaňa A., Sácký J., Kotrba P.. Heterologous expression of Zn-binding peptide RaZBP1 from. Folia Microbiol. (2019) 64 835-844. DOI: 10.1007/s12223-019-00696-1 26. Li Q., Huang W., Xiong C., Zhao J.. Transcriptome analysis reveals the role of nitric oxide in. Chemosphere (2018) 201 294-302. DOI: 10.1016/j.chemosphere.2018.03.011 27. Türkmen M., Budur D.. Heavy metal contaminations in edible wild mushroom species from Turkey’s Black Sea region. Food Chem. (2018) 254 256-259. DOI: 10.1016/j.foodchem.2018.02.010 28. Osobová M., Urban V., Jedelský P.L., Borovička J., Gryndler M., Ruml T., Kotrba P.. Three metallothionein isoforms and sequestration of intracellular silver in the hyperaccumulator. New Phytol. (2011) 190 916-926. DOI: 10.1111/j.1469-8137.2010.03634.x 29. Nguyen H., Rineau F., Vangronsveld J., Cuypers A., Colpaert J.V., Ruytinx J.. A novel, highly conserved metallothionein family in basidiomycete fungi and characterization of two representative SlMTa and SlMTb genes in the ectomycorrhizal fungus. Environ. Microbiol. (2017) 19 2577-2587. DOI: 10.1111/1462-2920.13729 30. Dong X.B., Huang W., Bian Y.B., Feng X., Ibrahim S.A., Shi D.F., Qiao X., Liu Y.. Remediation and Mechanisms of Cadmium Biosorption by a Cadmium-Binding Protein from. J. Agric. Food Chem. (2019) 67 11373-11379. DOI: 10.1021/acs.jafc.9b04741 31. Yan L., Xu R., Bian Y., Li H., Zhou Y.. Expression Profile of Laccase Gene Family in White-Rot Basidiomycete. Genes (2019) 10. DOI: 10.3390/genes10121045 32. Rahman M.A., Hassler C.. Is arsenic biotransformation a detoxification mechanism for microorganisms?. Aquat. Toxicol. (2014) 146 212-219. DOI: 10.1016/j.aquatox.2013.11.009 33. Reddy M.S., Prasanna L., Marmeisse R., Fraissinet-Tachet L.. Differential expression of metallothioneins in response to heavy metals and their involvement in metal tolerance in the symbiotic basidiomycete. Microbiology (2014) 160 2235-2242. DOI: 10.1099/mic.0.080218-0 34. Sácký J., Beneš V., Borovička J., Leonhardt T., Kotrba P.. Different cadmium tolerance of two isolates of. Fungal Biol. (2019) 123 247-254. DOI: 10.1016/j.funbio.2018.12.006 35. Arini A., Daffe G., Gonzalez P., Feurtet-Mazel A., Baudrimont M.. Detoxification and recovery capacities of. Environ. Pollut. (2014) 192 74-82. DOI: 10.1016/j.envpol.2014.04.012 36. Xu P., Leng Y., Zeng G., Huang D., Lai C., Zhao M., Wei Z., Li N., Huang C., Zhang C.. Cadmium induced oxalic acid secretion and its role in metal uptake and detoxification mechanisms in. Appl. Microbiol. Biotechnol. (2015) 99 435-443. DOI: 10.1007/s00253-014-5986-y 37. Chai L., Huang M., Cao X., Liu M., Huang Y.. Potential metal-binding ability of proteins in the extracellular slime of. Environ. Sci. Pollut. Res. Int. (2019) 26 20418-20427. DOI: 10.1007/s11356-019-05201-2 38. Cihangir N., Saglam N.. Removal of cadmium by. Acta Biotechnol. (1999) 19 171-177. DOI: 10.1002/abio.370190212 39. Hložková K., Matěnová M., Žáčková P., Strnad H., Hršelová H., Hroudová M., Kotrba P.. Characterization of three distinct metallothionein genes of the Ag-hyperaccumulating ectomycorrhizal fungus. Fungal Biol. (2016) 120 358-369. DOI: 10.1016/j.funbio.2015.11.007 40. Lim C.J., Jo H., Kim K.. A protective role of methionine-R-sulfoxide reductase against cadmium in. J. Microbiol. (2014) 52 976-981. DOI: 10.1007/s12275-014-3512-7 41. Mishra A., Malik A.. Simultaneous bioaccumulation of multiple metals from electroplating effluent using. Water Res. (2012) 46 4991-4998. DOI: 10.1016/j.watres.2012.06.035
--- title: 'School Problems and School Support for Children with Narcolepsy: Parent, Teacher, and Child Reports' authors: - Karin Janssens - Pauline Amesz - Yvonne Nuvelstijn - Claire Donjacour - Danielle Hendriks - Els Peeters - Laury Quaedackers - Nele Vandenbussche - Sigrid Pillen - Gert Jan Lammers journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049178 doi: 10.3390/ijerph20065175 license: CC BY 4.0 --- # School Problems and School Support for Children with Narcolepsy: Parent, Teacher, and Child Reports ## Abstract Objective: To assess problems faced by children with type 1 narcolepsy (NT1) at school and obtain insight into potential interventions for these problems. Methods: We recruited children and adolescents with NT1 from three Dutch sleep-wake centers. Children, parents, and teachers completed questionnaires about school functioning, interventions in the classroom, global functioning (DISABKIDS), and depressive symptoms (CDI). Results: Eighteen children (7–12 years) and thirty-seven adolescents (13–19 years) with NT1 were recruited. Teachers’ most frequently reported school problems were concentration problems and fatigue (reported by about $60\%$ in both children and adolescents). The most common arrangements at school were, for children, discussing school excursions ($68\%$) and taking a nap at school ($50\%$) and, for adolescents, a place to nap at school ($75\%$) and discussing school excursions ($71\%$). Regular naps at home on the weekend (children $71\%$ and adolescents $73\%$) were more common than regular naps at school (children $24\%$ and adolescents $59\%$). Only a minority of individuals used other interventions. School support by specialized school workers was associated with significantly more classroom interventions (3.5 versus 1.0 in children and 5.2 versus 4.1 in adolescents) and napping at school, but not with better global functioning, lower depressive symptom levels, or napping during the weekends. Conclusions: Children with NT1 have various problems at school, even after medical treatment. Interventions to help children with NT1 within the classroom do not seem to be fully implemented. School support was associated with the higher implementation of these interventions. Longitudinal studies are warranted to examine how interventions can be better implemented within the school. ## 1. Introduction Type 1 narcolepsy (NT1) is a chronic neurological sleep disorder characterized by excessive daytime sleepiness (EDS), which frequently induces problems with sustained attention. Emotion-triggered cataplexy, sleep paralysis, hypnagogic hallucinations, and disrupted nocturnal sleep are also common [1]. EDS is typically the first symptom to appear, followed by cataplexy up to several years later [1]. Onset is generally during adolescence but can also occur in childhood. Prevalence is 0.02–$0.05\%$ in western countries and an increase in annual incidence has been found among children [2]. NT1 is distinguished from Type 2 narcolepsy (NT2); in NT2, EDS is present, but cataplexy or low hypocretin levels are, in contrast to NT1, absent. Young people with NT1 often experience widespread impairment in their psychological well-being, education, and social relationships [3,4,5]. The importance of a biopsychosocial approach in the treatment of children with NT1 is stressed [6,7,8]. Despite the urge for such a biopsychosocial approach, not much is known about the problems and possible interventions for children with NT1 at school. Research that has been performed pointed out that youngsters with NT1 often face problems at school. Despite average IQ levels [9,10,11,12], young people with NT1 have more educational difficulties than youngsters with EDS alone or healthy controls [1,3,13]. Teachers have noticed several problems in children with NT1, including poor attention span, hyperactivity, distractibility, and cognitive underperformance. These school problems contribute significantly to a lower quality of life (QoL) in these individuals [3,14]. Adults with NT1 retrospectively evaluated their academic careers as more complex [13,15]. Moreover, prospective studies found that individuals with child- or adolescent-onset NT1 had lower educational levels, lower grading, and lower employment rate and income than healthy controls [16]. Treatment of NT1 in youngsters is focused on medication use [17] and behavioral approaches. Little research has been conducted on behavioral interventions for NT1. There is some evidence that daily napping is helpful in reducing EDS [18]. Another study that consisted of older adolescents and adults (please note that only $8.3\%$ were younger than 17) reported some symptom improvement with sleep hygiene (taking regular naps and keeping a nocturnal schedule), environmental changes (such as temperature regulation) and physical exercise [19]. Keeping nocturnal schedules, sleep hygiene, balanced diet, and physical activities, are also mostly recommended in clinical-based guidelines [20]. To the best of our knowledge specific studies on interventions to decrease attention problems of children or adolescents with NT1 in classroom are lacking. It is unknown to which extent these behavioral interventions are implemented in the school-life of youngsters suffering from NT1. Specialized school workers in the Netherlands (‘the LWOE’, https://www.lwoe.nl/, accessed on 9 March 2023) help youngsters with epilepsy implementing interventions at school. Such a team of school workers has also start helping children with NT1. It is unknown whether such a team is beneficial for implementing behavioral interventions for NT1 at school. This observational study aims to obtain more insight into specific problems encountered by children and adolescents with NT1 in school and into the implementation of potential interventions to increase school functioning. We also examined whether school support by a specialized team was associated with the higher implementation of these interventions within classrooms and better daytime functioning. ## 2.1. Participants One hundred twelve children and adolescents diagnosed with NT1, aged from 4 to 19 years, were approached for participation in the study at three specialized sleep centers in the Netherlands with expertise in narcolepsy. The inclusion criteria of the study were being diagnosed with NT1 according to the ICSD3-criteria and attending primary or secondary school. Diagnoses were made according to ICSD-3 criteria [21]. Medical records were used to determine whether participants received school support from a specialized team of school workers (‘the LWOE’). Fifty-five individuals ($49\%$) agreed to participate in the study. Young people were grouped according to school type; 18 participants were in primary school (‘children’), and 37 were in secondary school (‘adolescents’). Participants were recruited at different points after diagnosis. The median number of months since diagnosis was 33 for children and 29 for adolescents (Table 1). Most children and adolescents (>$75\%$) received their diagnosis more than 1 year ago. ## 2.2. Procedure Parents and adolescents gave informed consent after a full explanation of the study procedure. We reviewed medical charts to collect information regarding current medical treatment and comorbid conditions. Teachers were approached after parents and adolescents gave permission. They received paper-and-pencil questionnaires from the participating children and adolescents. They filled out the forms at home and returned them directly to the research institute. Adolescents, parents, and teachers completed questionnaires about problems experienced at school, school functioning, implementation of potential interventions at school, global functioning, and depressive symptom level. The study was conducted in accordance with the Declaration of Helsinki. The protocol was evaluated by the Medical Ethics Committee and judged as exempt from needing formal ethical approval due to its observational design. ## 2.3.1. Global Functioning in Children with Chronic Conditions (DISABKIDS) The Quality of Life Questionnaire for Children with Chronic Conditions (DISABKIDS) is a questionnaire used to assess global functioning in children and adolescents with chronic diseases. A parent-version (proxy) was used for children in primary education. Adolescents in secondary schools completed a self-report version. The DISABKIDS (both proxy and self-report) contains 37 items, and answers are scored on a 5-point Likert scale. A high total score indicates better global functioning. The questionnaire has six sub-scales: Independence, Emotions, Social Inclusion, Social Exclusion, Physical Limitation, and Impact of Treatment. Independence assesses how much the disease impairs a child/adolescent and whether he/she can live an independent life. Emotion describes to what extent the condition causes worry or concern for the child/adolescent. Social inclusion measures the closeness and positivity of friends and family, whereas social exclusion measures the child’s/adolescent’s feelings of stigma. Physical Limitation assesses to what degree the disease limits the child/adolescent. The Impact of Treatment subscale measures the child’s/adolescent’s negative feelings about taking medication. Raw total and sub-scale scores are transformed into t-scores for further interpretation. Normative scores are available from children/adolescents with various chronic conditions, among which idiopathic epilepsy [22]. The Cronbach’s alpha for the different scales in our study varied. Some were poor [physical limitation (0.47)/social inclusion (0.59)]; others acceptable [independence (0.74)/social exclusion (0.75)/total scale (0.79)], and others good [emotion (0.84) treatment (0.85)]. ## 2.3.2. Depressive Symptoms The Children’s Depression Inventory (CDI) is a self-reported questionnaire to measure depressive symptoms in children and adolescents. The questionnaire is validated for ages 7–17 [23,24]. Each item contains three possible answers and generates an item score ranging from 0 (=absence of symptom) to 2 (=symptom present). The sum of all twenty-seven item scores generates the raw total score. No age-specific reference values are available for the CDI. ## 2.3.3. School Functioning Children, adolescents, their parents, and teachers were asked to complete questionnaires to assess problems with NT1 encountered at school. A panel of professionals, consisting of neurologists, nurse practitioners, psychologists, and student counselors working with young people with NT1, developed the questionnaire. Additionally, teachers answered open-ended questions about the three most critical problems they had noticed in their students with NT1. Two researchers (KJ and DH) independently categorized these answers, e.g., the answers ‘tiredness,’ ‘feeling fatigued,’ ‘overtiredness,’ and ‘lack of energy’ were all assigned to the category “tiredness/lack of energy.” These categories were compared and discussed until a consensus was reached. Thus, an overview of the most common problems experienced in the classroom was generated. ## 2.3.4. School Interventions Adolescents (≥12 years), parents (of children/adolescents <12), and their teachers answered questions about whether arrangements were made between the young person and teacher about adjustments at school. Fourteen different adjustments were considered (Appendix A). Adolescents, parents, or teachers could answer whether they made agreements about these interventions by choosing one of the following answers: ‘yes’/’no’/’unknown.’ All questions that were answered with ‘yes’ were cumulated to calculate the total amount of arranged interventions for each child. Agreement between parents and teachers differed between poor and excellent for the separate interventions (Appendix A). The correlations between the total number of arranged interventions were adequate (Spearman rho parent/teacher: 0.44; adolescent/teacher: 0.68). ## 2.3.5. The LWOE The LWOE consists of a team of school counselors specialized in supporting young people with epilepsy and NT1 at school. They have regular appointments with young people with NT1, their parents, and teachers. They discuss problems young people with NT1 face at school, potential interventions for these problems, and evaluate the outcomes of these interventions. They help them to achieve more understanding of their problems from their teachers and classmates. Further, they discuss whether they receive appropriate education or whether further assistance or a specialized school is necessary to reach the full potential of the youngster. ## 2.4. Analyses Young people’s scores on the DISABKIDS were compared with normative scores of young people with epilepsy with one-sample t-tests. Data of the DISABKIDS in both samples were normally distributed. An overview of problems encountered by young people with NT1 reported by the teachers was calculated. Two researchers independently grouped the answers into different categories. Cases of disagreement were discussed. Further, an overview was generated of the number of interventions in the classroom that were discussed by students and teachers. For the primary school children, we used parent, and teacher questionnaires. For the secondary school children, self-reports and teacher questionnaires were used. We then compared the results between young people who received school support to those who did not receive school support on global functioning, depressive symptom level, and a number of arranged interventions, using independent sample t-tests. The difference between both groups in napping at school and during the weekends was calculated with Fisher’s exact tests. When a whole questionnaire was not completed by a child/parent or teacher, these participants were excluded from the analyses concerning this questionnaire. Single missing items were coded in a separate category, “unknown.” Percentages were calculated while considering this category. All analyses were performed with SPSS, version 23. A p-value of <0.05 was considered statistically significant. ## 3. Results In total 18 children and 37 adolescents with NT1 participated. Eight of 18 ($44\%$) children and 18 of 37 ($48\%$) adolescents received school support from specialized school workers. Characteristics of the groups are shown in Table 1. Young people with NT1 scored lower than those with epilepsy on the emotion, social exclusion, physical limitation scale, and total scores (Table 2). The mean depression score of children with NT1 was 11.0 (SD = 9.1); that of adolescents was 7.2 (SD = 4.7). The clinical cut-off of the CDI of 16 has been recommended as an indication of a depressive disorder in all age categories [23]. Four of the twelve children ($33\%$) and one of 36 adolescents ($3\%$) who completed this questionnaire scored above this cut-off. According to the parents, 4 out of 17 ($25\%$) children napped daily at school, whereas 12 out of 17 ($71\%$) napped daily during the weekends. According to parents, 23 of 36 ($64\%$) of the adolescents napped daily during school, and 29 of 36 ($81\%$) napped daily during the weekends. One parent did not answer these questions. According to the adolescents themselves, 22 out of 37 ($59\%$) napped daily during school, and 27 of 37 ($73\%$) napped daily during the weekends. School reports were obtained for 16 children ($89\%$) and 24 adolescents ($65\%$). The most frequent problems reported by the teachers of the 16 children with NT1 were tiredness/lack of energy, concentration problems, problems with social interaction, and living in their own world/autistic-like behavior (Table 3). The most frequent problems reported by the teachers of adolescents with NT1 were tiredness/lack of energy, concentration problems, and feeling anxious/insecure or auditory information processing problems. According to the teachers, half the children at primary school arranged to take naps during class, and about one-third had a place to sleep outside the classroom (Table 4). The most common was to discuss school excursions, and a minority arranged to obtain extra time for exams. Teachers reported a higher level of arrangement than parents did. Three-quarters of the adolescents at secondary school had a place to nap outside the classroom, and most arranged to take naps during class (Table 4). Extra time for taking a test, discussing school excursions, and adaptations during sports classes were common interventions for adolescents. ## School Support Children and adolescents who received school support from a specialized team reported significantly more interventions during class than children who did not receive school support (see Table 5). Children who received school support napped more frequently during school, i.e., 4 out of 7 who received school support napped, whereas 0 out of 9 who did not receive school report napped at school (Fisher’s exact test: $$p \leq 0.02$$). The same holds for adolescents, i.e., 14 out of 17 who received school support napped, whereas 7 out of 17 who did not receive school reports did (Fisher’s exact test: $$p \leq 0.03$$). Children who received school support did not differ in whether they napped on a regular basis at home during the weekends; 5 out of 8 who received school reports napped at home during the weekends, and 6 out of 9 who did not receive school support did (Fisher’s exact test: $$p \leq 0.63$$). The same was true for adolescents, 15 out of 17 who received school reports napped at home during the weekends, and 12 out of 17 who did not receive school support did (Fisher’s exact test: $$p \leq 0.39$$). Receiving school support was not associated with better global functioning or fewer depressive symptoms (Table 5). ## 4. Discussion This observational study suggests that children and adolescents with NT1 experience a diversity of issues at school. The most common problems at school, according to teachers, were concentration problems and feeling fatigued. Some interventions were common, whereas most were only used by a minority of the children with NT1. Receiving school support from a specialized team was associated with more interventions being implemented within the classroom and with napping on a regular basis at school. According to the teachers, the most commonly reported problems at school were concentration problems and feeling fatigued at school. These symptoms have also been found to be the most burdensome symptoms, according to adults with NT1 [19]. Adaptations within the classroom might be helpful to diminish these problems. Some school interventions were common, such as arranging school excursions and taking a nap at school. Other interventions, such as those aimed at increasing attention (e.g., drawing, walking and eating during instructions, and the use of audiobooks), were only used by a minority of the young people. The low rate of interventions to increase attention leaves room for improvement. Children with NT1 might, for example, benefit from organizational skills interventions that have been developed and found to be beneficial in children with attention-deficit/hyperactivity disorder [25]. Secondary school teachers reported more implemented interventions than primary school teachers did. For example, only $31\%$ of the primary school teachers reported that they made arrangements about a place to nap inside the school, whereas $75\%$ of the secondary school teachers did. It might be that some interventions have been less applicable to primary school children since school days for primary school children are shorter, they have less homework, and schools are often located closer to children’s homes. The difference between the number of children that were daily napping at school and daily napping at home during the weekends was also large. Twenty-five percent of children with NT1 napped during school days, whereas $71\%$ napped during the weekends. The reasons for this difference are unclear. Future research is needed to examine why behavioral interventions are not always implemented at school. For example, are facilities not available, is there a lack of understanding from classmates or teachers to implement these interventions, or do adolescents feel ashamed to make use of these interventions? We might learn from adherence research in adolescents with diabetes type 1 since these adolescents also have to actively manage their disease in the classroom [26,27]. The current study shows that young people who received school support from specialized teams of school workers made more arrangements for interventions than children who did not receive school support. Interestingly, young people who received school support napped more frequently at school but did not nap more at home during the weekends. In this observational study, however, receiving school support was not associated with better school functioning, better global functioning, or lower depressive symptom level. This might be related to the cross-sectional observational nature of this study. Whether or not the children received school support was not randomly determined but was determined by the clinician in accordance with the child and the parent. Children who experienced more school problems or problems with global functioning might therefore have been more prone to receive school support. This might have weakened potential associations between school interventions and global functioning. Moreover, since school support only focuses on school problems, a broader intervention program for children with NT1 might be necessary to have an effect on the global functioning level. Strong points of this study include the multi-informant design. Information was obtained from young people, their parents, and their teachers. We were, therefore, able to examine school functioning and interventions from different perspectives. Although perspectives were different, given the sometimes low correlations, the percentages of implemented interventions were largely comparable in different groups, which enlarges the robustness of our findings. Finally, we used extensive questionnaires and have a broad overview of youngsters functioning in and outside school. Some limitations of this study should also be mentioned. The number of participants was low—but NT1 is a relatively rare disease. The participation rate of almost $50\%$ was somewhat low but in line with dropping participation rates in general [28]. We used paper-and-pencil questionnaires. This might have reduced the number of received responses and diminished the generalizability of our findings. Additionally, participants were not prompted to choose an answer, which resulted in open fields or questions indicated as “unknown.” Future qualitative studies in which participants, parents, and teachers are interviewed about school functioning might help in getting additional information about the specific needs of children and adolescents with NT1 and ways to ease the implementation of interventions. Another limitation is the cross-sectional and observational study design. We were not able to study the effect of school support on functioning in young people with NT1. Longitudinal studies with assessment of school functioning before and after school support are necessary to obtain more insight into school support’s effects on young people with NT1. This study shows that there is room for improvement in the implementation of interventions for children with NT1 in the classroom, such as taking naps during school time. Although we should be careful because of the observational character of our study, our study suggests that a team of specialized school workers might be helpful in improving this implementation. We further believe that awareness of teachers for problems children and adolescents with NT1 experience at school, especially overtiredness and concentration problems are important. The teacher might think along with students and stimulate students to use these interventions during school time. Further research is necessary to develop a specific behavioral intervention to increase attention in youngsters with NT1. For example, it might be good to examine whether CBT-based interventions developed to decrease attention problems in children with attention deficit/hyperactivity disorder [29,30] are also helpful for youngsters with NT1. ## 5. Conclusions Children and adolescents with NT1 are at risk of experiencing problems inside the school. Feeling fatigued and experiencing concentration problems are common at school. Potential interventions were only partially implemented, especially since the number of primary school children that napped during school time ($25\%$) was low. School support by specialized teams in this observational study was associated with better implementation of these interventions and specifically more napping at school, but not with better school functioning or better quality of life. Longitudinal interventional research is necessary to get more insight into the benefit of behavioral interventions for narcolepsy at school and into ways to implement them. ## References 1. Rocca F.L., Pizza F., Ricci E., Plazzi G.. Narcolepsy during Childhood: An Update. Neuropediatrics (2015) 46 181-198. DOI: 10.1055/s-0035-1550152 2. Partinen M., Saarenpää-Heikkilä O., Ilveskoski I., Hublin C., Linna M., Olsén P., Nokelainen P., Alén R., Wallden T., Espo M.. Increased Incidence and Clinical Picture of Childhood Narcolepsy following the 2009 H1N1 Pandemic Vaccination Campaign in Finland. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0033723 3. Plazzi G., Clawges H.M., Owens J.A.. Clinical Characteristics and Burden of Illness in Pediatric Patients with Narcolepsy. Pediatr. Neurol. (2018) 85 21-32. DOI: 10.1016/j.pediatrneurol.2018.06.008 4. Prihodova I., Dudova I., Mohaplova M., Hrdlicka M., Nevsimalova S.. Childhood narcolepsy and autism spectrum disorders: Four case reports. Sleep Med. (2018) 51 167-170. DOI: 10.1016/j.sleep.2018.07.017 5. Quaedackers L., van Gilst M.M., van Mierlo P., Lammers G.-J., Dhondt K., Amesz P., Peeters E., Hendriks D., Vandenbussche N., Pillen S.. Impaired social functioning in children with narcolepsy. Sleep (2019) 42 228. DOI: 10.1093/sleep/zsy228 6. Graef D.M., Byars K.C., Simakajornboon N., Dye T.. Topical Review: A Biopsychosocial Framework for Pediatric Narcolepsy and Idiopathic Hypersomnia. J. Pediatr. Psychol. (2020) 45 34-39. DOI: 10.1093/jpepsy/jsz085 7. Xiao L., Chen A., Parmar A., Frankel L., Toulany A., Murray B.J., Narang I.. Narcolepsy Treatment: Voices of Adolescents. Behav. Sleep Med. (2021) 20 260-268. DOI: 10.1080/15402002.2021.1916496 8. Ingram D.G., Jesteadt L., Crisp S.C., Simon S.L.. Treatment and care delivery in pediatric narcolepsy: A survey of parents, youth, and sleep physicians. J. Clin. Sleep Med. (2021) 17 875-884. DOI: 10.5664/jcsm.9054 9. Szakács A., Hallböök T., Tideman P., Darin N., Wentz E.. Psychiatric Comorbidity and Cognitive Profile in Children with Narcolepsy with or without Association to the H1N1 Influenza Vaccination. Sleep (2015) 38 615-621. DOI: 10.5665/sleep.4582 10. Dorris L., Zuberi S.M., Scott N., Moffat C., McArthur I.. Psychosocial and intellectual functioning in childhood narcolepsy. Dev. Neurorehabilit. (2008) 11 187-194. DOI: 10.1080/17518420802011493 11. Posar A., Pizza F., Parmeggiani A., Plazzi G.. Neuropsychological Findings in Childhood Narcolepsy. J. Child Neurol. (2013) 29 1370-1376. DOI: 10.1177/0883073813508315 12. Janssens K.A.M., Quaedackers L., Lammers G.J., Amesz P., Van Mierlo P., Aarts L., Peeters E., Hendriks D., Vandenbussche N., Overeem S.. Effect of treatment on cognitive and attention problems in children with narcolepsy type 1. Sleep (2020) 43 114. DOI: 10.1093/sleep/zsaa114 13. Stores G., Montgomery P., Wiggs L., Willot S., Uhlen S., Michaud L., Briand G., Bonnevalle M., Sfeir R., Gottrand F.. The Psychosocial Problems of Children With Narcolepsy and Those With Excessive Daytime Sleepiness of Uncertain Origin. Pediatrics (2006) 118 e1116-e1123. DOI: 10.1542/peds.2006-0647 14. Szakács A., Chaplin J.E., Tideman P., Strömberg U., Nilsson J., Darin N., Hallböök T.. A population-based and case-controlled study of children and adolescents with narcolepsy: Health-related quality of life, adaptive behavior and parental stress. Eur. J. Paediatr. Neurol. (2019) 23 288-295. DOI: 10.1016/j.ejpn.2019.01.004 15. White M., Charbotel B., Fort E., Bastuji H., Franco P., Putois B., Mazza S., Peter-Derex L.. Academic and professional paths of narcoleptic patients: The Narcowork study. Sleep Med. (2020) 65 96-104. DOI: 10.1016/j.sleep.2019.07.020 16. Jennum P., Ibsen R., Kjellberg J.. Long-term health and socioeconomic consequences of childhood and adolescent-onset of narcolepsy. Sleep Med. (2020) 67 23-27. DOI: 10.1016/j.sleep.2019.10.013 17. Lammers G.J.. Drugs Used in Narcolepsy and Other Hypersomnias. Sleep Med. Clin. (2018) 13 183-189. DOI: 10.1016/j.jsmc.2018.02.009 18. Mullington J., Broughton R.. Scheduled Naps in the Management of Daytime Sleepiness in Narcolepsy-Cataplexy. Sleep (1993) 16 444-456. DOI: 10.1093/sleep/16.5.444 19. Maski K., Steinhart E., Williams D., Scammell T., Flygare J., McCleary K., Gow M.. Listening to the Patient Voice in Narcolepsy: Diagnostic Delay, Disease Burden, and Treatment Efficacy. J. Clin. Sleep Med. (2017) 13 419-425. DOI: 10.5664/jcsm.6494 20. Blackwell J.E., Kingshott R.N., Weighall A., Elphick H.E., Nash H.. Paediatric narcolepsy: A review of diagnosis and management. Arch. Dis. Child. (2022) 107 7-11. DOI: 10.1136/archdischild-2020-320671 21. Ruoff C., Rye D.. The ICSD-3 and DSM-5 guidelines for diagnosing narcolepsy: Clinical relevance and practicality. Curr. Med. Res. Opin. (2016) 32 1611-1622. DOI: 10.1080/03007995.2016.1208643 22. Schmidt S., Debensason D., Mühlan H., Petersen C., Power M., Simeoni M.C., Bullinger M.. The DISABKIDS generic quality of life instrument showed cross-cultural validity. J. Clin. Epidemiol. (2006) 59 587-598. DOI: 10.1016/j.jclinepi.2005.09.012 23. Roelofs J., Braet C., Rood L., Timbremont B., van Vlierberghe L., Goossens L., van Breukelen G.. Norms and screening utility of the Dutch version of the Children’s Depression Inventory in clinical and nonclinical youths. Psychol. Assess. (2010) 22 866-877. DOI: 10.1037/a0020593 24. Timbremont B., Braet C., Dreessen L.. Assessing Depression in Youth: Relation Between the Children’s Depression Inventory and a Structured Interview. J. Clin. Child Adolesc. Psychol. (2004) 33 149-157. DOI: 10.1207/S15374424JCCP3301_14 25. Bikic A., Reichow B., McCauley S.A., Ibrahim K., Sukhodolsky D.G.. Meta-analysis of organizational skills interventions for children and adolescents with Attention-Deficit/Hyperactivity Disorder. Clin. Psychol. Rev. (2017) 52 108-123. DOI: 10.1016/j.cpr.2016.12.004 26. Borus J.S., Laffel L.. Adherence challenges in the management of type 1 diabetes in adolescents: Prevention and intervention. Curr. Opin. Pediatr. (2010) 22 405-411. DOI: 10.1097/MOP.0b013e32833a46a7 27. Núñez-Baila M., Gómez-Aragón A., González-López J.. Social Support and Peer Group Integration of Adolescents with Diabetes. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18042064 28. Galea S., Tracy M.. Participation Rates in Epidemiologic Studies. Ann. Epidemiol. (2007) 17 643-653. DOI: 10.1016/j.annepidem.2007.03.013 29. Anastopoulos A.D., Langberg J.M., Eddy L.D., Silvia P.J., Labban J.D.. A randomized controlled trial examining CBT for college students with ADHD. J. Consult. Clin. Psychol. (2021) 89 21-33. DOI: 10.1037/ccp0000553 30. Solanto M.V., Scheres A.. Feasibility, Acceptability, and Effectiveness of a New Cognitive-Behavioral Intervention for College Students with ADHD. J. Atten. Disord. (2021) 25 2068-2082. DOI: 10.1177/1087054720951865
--- title: LOX-1 Activation by oxLDL Induces AR and AR-V7 Expression via NF-κB and STAT3 Signaling Pathways Reducing Enzalutamide Cytotoxic Effects authors: - Felix Duprat - Catalina Robles - María Paz Castillo - Yerko Rivas - Marcela Mondaca - Nery Jara - Francisco Roa - Romina Bertinat - Jorge Toledo - Cristian Paz - Iván González-Chavarría journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049196 doi: 10.3390/ijms24065082 license: CC BY 4.0 --- # LOX-1 Activation by oxLDL Induces AR and AR-V7 Expression via NF-κB and STAT3 Signaling Pathways Reducing Enzalutamide Cytotoxic Effects ## Abstract The oxidized low-density lipoprotein receptor 1 (LOX-1) is one of the most important receptors for modified LDLs, such as oxidated (oxLDL) and acetylated (acLDL) low-density lipoprotein. LOX-1 and oxLDL are fundamental in atherosclerosis, where oxLDL/LOX1 promotes ROS generation and NF-κB activation inducing the expression of IL-6, a STAT3 activator. Furthermore, LOX-1/oxLDL function has been associated with other diseases, such as obesity, hypertension, and cancer. In prostate cancer (CaP), LOX-1 overexpression is associated with advanced stages, and its activation by oxLDL induces an epithelial-mesenchymal transition, increasing angiogenesis and proliferation. Interestingly, enzalutamide-resistant CaP cells increase the uptake of acLDL. Enzalutamide is an androgen receptor (AR) antagonist for castration-resistant prostate cancer (CRPC) treatment, and a high percentage of patients develop a resistance to this drug. The decreased cytotoxicity is promoted in part by STAT3 and NF-κB activation that induces the secretion of the pro-inflammatory program and the expression of AR and its splicing variant AR-V7. Here, we demonstrate for the first time that oxLDL/LOX-1 increases ROS levels and activates NF-κB, inducing IL-6 secretion and the activation of STAT3 in CRPC cells. Furthermore, oxLDL/LOX1 increases AR and AR-V7 expression and decreases enzalutamide cytotoxicity in CRPC. Thus, our investigation suggests that new factors associated with cardiovascular pathologies, such as LOX-1/oxLDL, may also promote important signaling axes for the progression of CRPC and its resistance to drugs used for its treatment. ## 1. Introduction Lipoproteins are macromolecular complexes that transport lipids in a soluble form through blood circulation [1]. Lipoproteins are classified into five classes: chylomicrons (Quis), very low-density lipoproteins (VLDLs), intermediate-density lipoproteins (IDLs), low-density lipoproteins (LDLs), and high-density lipoproteins (HDLs) [2]. Human LDL is the primary lipoprotein for delivering exogenous cholesterol into the cells. High serum levels of LDLs are closely related to the development and progression of cardiovascular pathologies, such as endothelial dysfunction and atherosclerosis [3]. For years it was postulated that LDLs themselves promoted the development and progression of atherosclerotic plaques [4]. However, a simple observation changed this paradigm by showing that foam cells in the fatty streak of atherosclerotic plaque endocytose, were an oxidized form of LDLs (oxLDLs), not LDLs [5]. Then, other studies demonstrated that atherosclerotic plaques present high levels of oxLDLs, which undergo endocytosis via scavenger receptors that induce endothelial dysfunction and the formation of foam cells, two critical events in the progression of atherosclerosis [6]. The scavenger receptor LOX-1 is one of the most important receptors for oxLDLs, and its activation induces NADPH oxidase activity and reactive oxygen species (ROS) production, which can activate p38MAPK, ERK$\frac{1}{2}$, PI3K, NF-κB, and STAT3 signaling pathways that drive the expression of proinflammatory cytokines and the atherosclerosis progression [7,8,9,10]. Thus, it has been described that LOX-1 and oxLDLs are essential modulators of endothelial dysfunction, which is characterized by the increase in the secretion of pro-inflammatory cytokines, an increase in the ROS levels, the expression of pro-coagulants molecules, expression of adhesion molecules, such as I-CAM, V-CAM, E-selectin, and P-selectin, and an increase in endothelium permeability [11,12,13]. Furthermore, LOX-1 and oxLDLs promote macrophage activation and actively participate in foam cell formation. These cells are the main component of fatty streaks and are generated by oxLDL endocytosis through scavenger receptors, such as LOX-1, CD36, SR-B1, and SR-A1 expressed in the membrane of macrophages that transmigrate to the tunica intima of the arteries in response to the proinflammatory process. Finally, oxLDL/LOX-1 and other scavenger receptors participate in the generation of complicated atherosclerotic plaque through matrix metalloproteinases secretion (MMP2 and MMP9) mediated by activated macrophages present in atherosclerotic plaque [14,15,16,17]. However, the role of LOX-1 and oxLDLs is not only restricted to pathologies, such as atherosclerosis or endothelial dysfunction, but has also been associated with the progression of chronic pathologies, such as obesity [18,19], type II diabetes mellitus [20,21], and various types of cancer. The role of LOX-1 in cancer is related to the hallmarks of cancer, such as angiogenesis, invasion, metastasis, resisting cell death, and sustained proliferative signaling between others [22,23,24,25,26,27,28,29,30,31,32], and has been addressed in depth in several reviews [20,33,34,35,36]. Prostate cancer (CaP) is the most prevalent malignancy among adult men in the developed world [37]. Advanced CaP is treated with androgen deprivation therapy (ADT), which reduces the systemic androgen concentration, which is critical for the proliferation and survival of tumor CaP cells [38]. However, the patients in ADT will eventually develop castration-resistant prostate cancer (CRPC), which does not respond to ADT [39,40,41]. The androgen receptor (AR) antagonist enzalutamide is used for CRPC treatment, but a significant percentage of CRPC patients develop resistance to this drug in a short period of time [42]. The decrease in enzalutamide cytotoxic effects is associated with the activation of NF-κB and STAT3 signaling pathways, which promote the overexpression and activation of AR and AR-V7 and the secretion of cytokines, such as IL-6, that strengthen the activation of STAT3 signaling [43,44,45]. Thus, the overexpression of AR and AR-v7 decreased the cytotoxic and anti-proliferative effects of enzalutamide, inducing resistance against this drug. High concentrations of oxLDLs were observed in patients with advanced CaP, and the overexpression of LOX-1 has been associated with high Gleason scores and clinicopathological stages III and IV [31,32]. Notably, oxLDL/LOX-1 induce angiogenesis, proliferation, and EMT in CaP cell lines and increase their tumorigenic potential. Furthermore, LOX-1 expression is necessary for tumor growth of CaP cell xenografts [32]. Interestingly, enzalutamide resistant LNCaP cells increase the uptake of acetylated LDLs (acLDLs) [46]. The acLDLs and oxLDLs are not recognized by LDL receptors, but instead by scavenger receptors, such as LOX-1 [47]. Thus, although some functions of LOX-1 and oxLDLs have been reported in CaP, the role of LOX-1 and oxLDLs in the expression of AR and AR-v7, the activation of downstream signaling pathways, and their impact on AR-targeting drugs have not been studied. In the present study, we demonstrate that the activation of LOX-1 by oxLDLs increases ROS production and activates NF-κB. The activation of NF-κB induces IL-6 secretion and STAT3 activation, which in turn promotes the overexpression of AR and AR-V7, and, therefore, decreases enzalutamide cytotoxicity in CRPC cells lines. ## 2.1. LOX-1 Activation by oxLDLs Increases ROS Generation in CRPC Cells LOX-1 activation by oxLDLs increases ROS generation in different cellular models associated with atherosclerosis progression, and has been described as an important signal for NF-κB activation [48]. To evaluate whether LOX-1 activation by oxLDLs promotes ROS generation in CRPC, C4-2B and 22RV-1 cells were transfected with siRNA LOX-1 or siRNA control (Figure 1A) and treated with or without 25 μM Trolox for 48 h, and then the cells were incubated with 50 μg/mL oxLDLs for 3 h. The results showed that oxLDLs significantly increased ROS generation by 2.3- and 1.3-fold in siRNA control C4-2B and 22Rv1 cells, respectively. However, in LOX-1 knockdown C4-2B and 22Rv1 cells, the effect of oxLDLs was prevented to similar levels to those observed in Trolox-treated cells (Figure 1B) suggesting that LOX-1 activation by oxLDLs induces ROS regeneration in C4-2B and 22Rv1 cells. ## 2.2. OxLDL/LOX-1 Induces the Activation of NF-κB in C4-2B and 22RV1 Cells To determine the effect of LOX-1 activation by oxLDLs on the NF-κB signaling pathway, C4-2B and 22RV1 cells transfected with siRNA control or siRNA LOX-1 were treated with 50 μg/mL oxLDLs for 1 h, and the phosphorylation of p65 and IκB-α were analyzed by Western blot. The results showed that oxLDL induces p65 phosphorylation by 1.37- and 1.33-fold on C4-2B and 22RV1 cells (siRNA control), respectively. In IκB-α, the oxLDL induces the phosphorylation by 1.40- and 3.19-fold on C4-2B and 22RV1 (siRNA control) cells, respectively. In contrast, the oxLDL-induced phosphorylation of p65 and IκB-α was prevented in the LOX-1 knockdown C4-2B and 22RV1 cells. This suggests that activation of LOX-1 by oxLDLs induces the NF-κB signaling pathway activation (Figure 2A–C). To confirm these results, we evaluated the NF-κB activity using the reporter plasmid pHAGE NFKB-TA-LUC-UBC-dTomato-W on C4-2B or 22RV1 cells co-transfected with siRNA control or siRNA LOX-1 and treated with oxLDLs. The results showed that 50 μg/mL of oxLDLs increased the p65-luciferase reporter activity in siRNA control cells by 1.54- and 1.70-fold compared with C4-2B and 22RV1 (siRNA control) untreated cells. The LOX-1 knockdown C4-2B and 22RV1 cells exhibited low basal p65 promoter activity, and the effects of oxLDLs on p65 activity were prevented (Figure 2B–D). ## 2.3. Activation of LOX-1 by oxLDLs Induces the Secretion of IL-6 and STAT3 Activation in C4-2B and 22Rv1 Cells NF-κB activation significantly increases the expression and secretion of pro-inflammatory cytokines, such as IL-6, which has been described as an important cytokine in CRPC generation and enzalutamide resistance [49,50]. To determine the effect of LOX-1 activation on IL-6 secretion, C4-2B and 22RV1 cells transfected with siRNA control or siRNA LOX-1 were treated with 50 μg/mL oxLDLs for 24 h, and IL-6 levels were measured in the supernatants by ELISA. The results showed that oxLDL induces an increase in IL-6 concentration from 0.8 to 1.9 pg/m and 1.1 to 2.8 ng/mL in siRNA control C4-2B and 22Rv1 (Figure 3A). However, the effects of oxLDLs on the IL-6 secretion were prevented in the LOX-1 knockdown in C4-2B and 22RV1 cells (Figure 3A). Therefore, LOX1 activation by oxLDLs promotes the secretion of IL-6, probably through NF-κB activation in CRPC cell lines. IL-6-induced STAT3 activation partially promotes the resistance to enzalutamide in CRPC through AR and AR-V7 overexpression [51]. Given that LOX-1 activation by oxLDLs stimulated the secretion of IL-6 in our CRPC cell models, we analyzed the effect of oxLDLs on the STAT3 signaling pathway. C4-2B and 22RV1 cells were transfected with control siRNA or LOX-1 siRNA for 48 h, and then were treated with 50 μg/mL oxLDLs for 24 h, and the phosphorylation of STAT3 was determined by Western blot. The results showed that oxLDL increases STAT3 phosphorylation by 1.6- and 1.7-fold in siRNA control C4-2B and 22RV1 cells compared to untreated cells (Figure 4A). In LOX-1 knockdown in C4-2B and 22RV1 cells and the phosphorylation of STAT3 was prevented, suggesting that activation of LOX-1 by oxLDLs activates STAT3 signaling (Figure 3B). ## 2.4. LOX-1/oxLDL Induces AR and AR-V7 Overexpression in C4-2B and 22RV1 Cells Overexpression and increased activity of AR and its splicing variant AR-V7 have been widely associated with CRPC and enzalutamide resistance [52]. In the case of AR-V7, is one of the most relevant variants in CRPC, responsible for the constitutive activation of AR signaling, even in the absence of a ligand [53]. In contrast to the native AR, AR-V7 lacks the ligand (DHT) binding site at which enzalutamide acts in an antagonistic fashion [54]. The clinical relevance of AR-V7 arises from a comparison between AR-V7 negative and positive patients in which the latter showed a lower survival rate. Indeed, $39\%$ of patients with metastatic CRPC treated with enzalutamide express AR-V7 in the tumoral cells [52,55]. In turn, the presence of AR-V7 is associated with 9–$15\%$ of patients with a higher probability of generating enzalutamide resistance [52,55,56]. The mechanism associated with AR and AV-V7 overexpression has been correlated with NF-κB and STAT-3 activation, to this part, our data suggest that oxLDLs through LOX-1 induce ROS/NF-KB/IL-6 and STAT3 axis, which indicates that AR and AR-v7 could be overexpressed by LOX-1/oxLDL. To determine this, 22RV1 cells were transfected with siRNA control and siRNA LOX-1 for 48 h and then treated with 50 μg/mL of oxLDLs. The expression of AR and AR-V7 was analyzed by Western blot, the AR activity was analyzed through its translocation to the nucleus by immunocytochemistry assay, and the expression of prostate-specific antigen (PSA, gen under AR control) was analyzed by real-time PCR. The results showed that oxLDL induces by 1.65- and 1.7-fold the AR and AR-V7 expression in C4-2B (siRNA control) cells, respectively, and by 1.75- and 1.60-fold in 22RV1 (siRNA control) cells, respectively, compared with untreated cells. The effect of oxLDL-induced overexpression of AR and AR-V7 expression was prevented in LOX-1 knockdown C4-2B and 22RV1 cells (Figure 4A). Furthermore, oxLDL treatment in non-transfected cells induced the translocation of AR to the nucleus and increased PSA mRNA levels by 2.5- and 3.6-fold in C4-2B and 22RV1 cells, respectively (Figure 4B). Overall, these data suggest that oxLDLs not only induce the expression of AR and AR-V7 but also activate them in these CRPC cell lines, probably by ROS/NF-KB/IL-6 and STAT3 axis. To analyze whether the AR expression induced by oxLDLs is mediated by ROS, NF-κB, and STAT3 activation, we used a ROS inhibitor (Trolox), three NF-κB inhibitors: triptolide (TLP), bay 11-7082 (BAY), caffeic acid phenethyl ester (CAPE), and a chemical inhibitor of STAT3 activity Stattic. C4-2B and 22RV1 cells were co-treated with 25 μM Trolox, 0.750 ng/mL TPL, 10 μM BAY, 10 μM CAPE, or 5 μM Stattic and 50 μg/mL oxLDLs for 24 h, and the expression of AR was analyzed by Western blot. The results showed that Trolox, BAY, CAPE, and Stattic prevent oxLDL-induced AR overexpression in both CRPC cell lines (Figure 5A,B). Moreover, the results show that in C4-2B cells, STAT-3 inhibition generates a significant decrease in AR expression (Figure 5A). Similarly, inhibition of NF-κB by TPL in 22RV1 cells decreases the AR expression levels (Figure 5B). ## 2.5. LOX-1 Activation by oxLDLs Decreases the Cytotoxic Effects of Enzalutamide on C4-2B and 22RV1 Cells The activation of NF-κB and STAT3 signaling pathways and the overexpression of AR, AR-V7, PSA, and IL-6 have been recognized as markers and inducers or effectors (AR- AR-V7) of enzalutamide resistance in CRPC [49,57]. Our results indicate that LOX-1 activation by oxLDLs induces the axis ROS/NF-KB/IL-6/STAT3 with the consequent overexpression of AR and AR-v7. We hypothesized that the overexpression of AR and AR-V7 could decrease the cytotoxic effect of enzalutamide, an antagonist drug of AR. Our results showed that oxLDL significantly increases the surviving fraction and the IC50 for enzalutamide by 1.8- and 2.7-fold in C4-2B and 22Rv1 cells, respectively (Figure 6). Moreover, LOX-1 knockdown in C4-2B and 22Rv1 cells prevented the effects of oxLDL, even displaying increased sensitivity to enzalutamide and higher cytotoxic effect of co-treatment with oxLDL/enzalutamide (Figure 6A,B). In this regard, we observed that enzalutamide treatment for 24 h decreased AR and AR-V7 expression in C4-2B and 22RV-1; however, this effect was prevented by oxLDLs/enzalutamide co-treatments (Figure 6A,B), suggesting that the cytotoxicity-lowering effects of enzalutamide mediated by oxLDLs are mediated by the increase in the expression of classical markers of enzalutamide resistance, such as AR and AR-V7 in C4-2B and 22RV-1 cells. ## 3. Discussion LOX-1 receptor and oxLDLs are associated with the development and progression of several pathologies, such as atherosclerosis, obesity, type 2 diabetes, and different cancers [22,23,24,25,26,27,28,29,30,31,32]. Notably, in CaP, we have previously shown that LOX-1 activation by oxLDLs stimulates tumor angiogenesis [31], epithelial-mesenchymal transition, and tumorigenic potential, being determinant for tumor growth [32], which could explain why obese patients with CaP progress to more malignant stages in a shorter period of time compared to patients with average weight [32]. In atherosclerosis and endothelial dysfunction, activation of LOX-1 stimulates the NF-κB signaling pathway to promote the expression of proinflammatory cytokines, such as IL-1b, TNF-α, MCP1, IL-8, and IL-6 [58]. Among these cytokines, IL-6 is particularly relevant because it induces the activation of STAT3, a critical factor in the development of CRPC and enzalutamide resistance [49,50,59]. In the present work, we demonstrated that oxLDL/LOX-1 increases ROS levels and activates NF-κB signaling pathway, inducing the secretion of IL-6 and the activation of STAT3 in CRPC cell lines. These data indicate that atherosclerosis and CRPC, two very different diseases at first sight, share pathological mechanisms at the molecular level, at least regarding the oxLDL/LOX-1 axis. Furthermore, oxLDL/LOX-1 increases AR and AR-V7 protein levels, two important markers for enzalutamide resistance, which explains the increase in IC50 for enzalutamide in CRCP cell lines. The ROS generation induced by oxLDL/LOX-1 has been reported in many studies associated with atherosclerosis, endothelial dysfunction, and macrophage activation [60]. We extended these findings to human CRPC cell lines in the present study. In endothelial cells, high concentrations of oxLDL are cytotoxic and induce cell death [61]. In contrast, low concentrations of oxLDL induce oxidative stress through stimulation of NAPDH oxidase, increased ROS production, and activation of NF-κB, which promotes the secretion of pro-inflammatory cytokines [62]. In contrast, breast cancer or CaP cells do not show significant changes in cell viability at a high concentration of oxLDLs. Moreover, previous studies by our group showed that oxLDLs at concentrations up to 100 μg/mL increased cell proliferation [63,64]. In this sense, high concentrations of oxLDLs could play an important role in the activation of several processes associated with tumor progression. Several authors have described that the activation of NF-κB by ROS induces IκBα phosphorylation-dependent or independent of IKK [65,66,67]. NF-κB signaling pathway induces IL-6 expression, which plays an essential role in prostate cancer progression, the development of CRPC, and enzalutamide resistance [51,68]. The overexpression of IL-6 increases the resistance of prostate cancer cells mainly through the JAK/STAT3 axis [69]. Thus, it has been demonstrated that inhibitors of the IL-6/STAT3 axis, such as galiellalactone or Stattic [70] and antibodies against IL-6 [71], among others, are viable alternatives to decrease the AR activity in prostate tissue and CaP. The ectopic expression of IL-6 in IL-6 negative LNCaP cells has been shown to significantly increase clonogenic and proliferative capacity, accompanied by JAK/STAT3 activation [72]. In this regard, CRPC cells secrete IL-6, which through an autocrine mechanism, induces enzalutamide resistance via the constitutive activation of STAT3, whereas its inhibition prevents or reverts enzalutamide resistance [51]. Likewise, the activation of NF-κB by oxLDL/LOX-1-ROS could induce the secretion of IL-6 observed in our study, which in turn may act in an autocrine fashion to stimulate STAT3 signaling, as it has been observed in other types of cancer [73]. Furthermore, our results showed that oxLDL promotes the expression of AR and its variant AR-V7, and a ROS, NF-κB, and STAT3 inhibitor prevented these effects. Similarly, oxLDL/LOX-1 increased AR and AR-V7 expression, promoted AR nuclear translocation, and increased PSA levels in the CRPC cell models used in our study. In the context of CRPC, it has been described that NF-κB can mediate the expression of enzalutamide resistance-related proteins, such as AR and AR-V7 [74]. Although there is no evidence linking oxLDL/LOX-1 to AR or AR-V7 expression, however, several reports have demonstrated that NF-κB activation is involved in the expression of these effectors [49,74,75,76]. In a model of benign prostatic hyperplasia, Austin et al. identified that NF-κB is important in the progression of this pathology and can lead to prostate cancer development [75]. In this same study, NF-κB activation significantly increased AR-V7 mRNA levels. Moreover, in the androgen-dependent cell line LNCaP or the androgen-independent cell line 22Rv1, AR and AR-V7 overexpression increased NF-κB activation, suggesting that there may be crosstalk between these pathways. Liu et al. demonstrated that the inhibition of NF-κB with melatonin prevents AR and AR-V7 overexpression in CaP cells [49]. Furthermore, in castration models, in which dihydrotestosterone concentration was significantly decreased, the high NF-κB activation is evidenced by phosphorylation of the p65 subunit [39]. Moreover, Nadiminty et al. demonstrated that in the androgen-sensitive cell line LNCaP, the over-activation of the non-canonical p52 NF-κB pathway also promoted AR-V7 expression [77]. In 22Rv1 cells, Kiliccioglu et al. observed that AR-V7 and AR expression was decreased upon inhibition of NF-κB with BAY 11-7082, reinforcing the idea that NF-κB is important in the overexpression of AR and AR-V7, two enzalutamide resistance markers [78]. Our results demonstrated that oxLDL treatments decrease enzalutamide cytotoxicity, an effect that is prevented when LOX-1 is silenced. These results confirm that the expression of enzalutamide resistance markers, such as AR and AR-V7, are associated with oxLDL-mediated LOX-1 activation. Our analysis demonstrated that CRPC cells co-treated with oxLDLs, and enzalutamide maintain the expression of AR and AR-V7 at the level of cells without enzalutamide treatment, strongly supporting that oxLDL decreases enzalutamide cytotoxicity mediated by LOX-1 activation. Based on this evidence, it is reasonable to propose that oxLDL/LOX-1-mediated AR and AR-V7 expression in CRPC cells is a result of the NF-κB pathway activation. However, it is relevant to note that this effect could occur directly through oxLDL/LOX-1/ROS/NK-kB/AR axis activation or indirectly through the overexpression and secretion of proinflammatory cytokines, such as IL-6, which in turn act in an autocrine fashion to activate STAT3/AR axis. From this perspective, even if either of these events could justify the AR and AR-V7 overexpression, LOX-1 activation should induce an increase in the IC50 of AR-targeting drugs, such as enzalutamide. Thus, our results in conjunction with other investigations, demonstrate a relevant role of LOX-1 in cancer progression. These allow us to propose LOX-1 as a target in cancer therapy or, as in our study, as a target for the prevention of the resistance to enzalutamide in CRPC. In this regard, statins, an HMG-CoA reductase inhibitor, positively prevent cancer progression [79]. Likewise, it has been determined that they prevent LOX-1 expression and that statins, such as lovastatin would directly antagonize the binding of oxLDLs to LOX-1 [80]. Thus, FDA-approved drugs such as statins, could be recommended to prevent enzalutamide resistance in CRPC patients due to their antagonistic effect on LOX-1, a receptor that promotes antiandrogen resistance through the overexpression of AR and AR-v7. In conclusion, treating CRPC patients is a complex process involving using different drugs for its control. However, resistance to drugs used in CRPC, such as enzalutamide, is observed in a large percentage of patients. Our results showed for the first time that LOX-1 and oxLDLs, two critical factors associated with pathologies of lipid metabolism and CaP tumor progression, are also determinants in the resistance to enzalutamide in human CRPC cell lines through NF-κB or STAT3 signaling pathway and AR and AR-V7 overexpression. ## 4.1. Cell Culture Human prostate cancer cells C4-2B and 22RV1 were grown in RPMI 1640 medium, supplemented with 2 mM L-glutamine (Hyclone), $10\%$ fetal bovine serum, and $1\%$ penicillin-streptomycin (GIBCO). The cell lines were maintained at 37 °C with $5\%$ CO2. The C4-2B and 22RV1 cell lines are cell lines widely used in castration-resistant prostate cancer research. The C4-2B cell line is from LNCaP (androgen sensitive), co-inoculated in a nude mouse with human fibroblasts derived from osteosarcoma. The nude mice were castrated after eight weeks of incubation, and castration-resistant tumor cells (C4 cells) were isolated and characterized. The process was repeated to obtain the C4-2B cell line. The 22RV1 cell line is derived from xenograft that was serially propagated in mice after castration-induced regression and relapse of the parental, androgen-dependent CWR22 cell line xenograft. Both cell lines are androgen-independent but can respond to testosterone or DHT, and are able to secrete PSA. The cell lines express AR and its variant AR-V7. In this sense, it has been reported that 22RV1 expresses high levels of AR-V7 and other variants compared to C4-2B [81,82]. ## 4.2. OxLDL Preparation The LDL isolation and oxLDL generation were carried out, according to our previous reports [31,32]. The oxLDL was extensively dialyzed against PBS and stored at 4 °C. ## 4.3. siRNA Transfection siRNA against LOX-1 were acquired from ThermoFisher Scientific (S9842). For silencing the LOX-1 RNA expression, C4-2B and 22RV1 prostate cancer cell lines were transfected with 150 pmol of siRNA/LOX-1 or siRNA/control using Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, MA, USA). Two days post-transfection, LOX-1 silencing was verified by Western blot. ## 4.4. Immunodetection Western blot and immunostaining were performed using standard protocols [26]. Anti-LOX-1 (LOX19-22) sc-66155, anti-AR (N-20) sc-816, anti-GAPDH [0411] sc-47724, anti-NF-kB p65 C-20) sc-372 and anti-IκB-alfa (C-21) sc-371, and anti-STAT3 (c-20) sc-482 antibodies were obtained from Santa Cruz Biotechnology. Anti-p-p65 s536 7F1, anti-p-IκB-alfa S32/S36 5A5, androgen receptor (AR-V7 Specific), and anti-p-STAT3 (Y705) (D3A7) antibodies were obtained from Cell Signaling. For Western blot, the anti-mouse IgG-Alexa Fluor 680 and anti-rabbit IgG-Alexa Fluor 790 secondary antibodies were obtained from Jackson ImmunoResearch. Western blots were analyzed using LICOR-CLX. An anti-rabbit IgG/FITC secondary antibody from Jackson ImmunoResearch was used for immunofluorescences. Photographs were obtained by confocal microscopy (Olympus IX81, Japan). To determine IL-6 concentration from supernatants, the human ELISA Kit (D6050, R&D) was used according to manufacturer’s instructions. ## 4.5. Real Time PCR Real-time PCR for LOX-1 (Forward 5′- AGATCCAGACTGTGAAGGACCAGC-3′, reverse 5′- CAGGCACCACCATGGAGAGTAAAG -3′) and prostate-specific antigen (PSA) (Forward 5′- GCATTACCGGAAGTGGATCAAGGA-3′, reverse 5′- TTGAGTCTTGGCCTGGTCATTTCC-3′) were performed using KAPA one step SYBR® FAST qPCR KIT (KAPA Biosystems, Wilmington, MA, USA) and Agilent AriaMx real-time PCR system equipment. The results were analyzed as 2−ΔΔCT relative quantification. The comparative threshold cycle values were normalized for b-actin (Forward 5′-TGTACCCTGGCATTGCCGACAG-3′, Reverse 5′-ACGGAGTACTTGCGCTCAGGAG-3′. ## 4.6. ROS Determination C4-2B and 22RV1 cells were transfected (20.000 cells) with siRNA control or siRNA LOX-1. Then, after 48 h, cells were seeded in 96-well black plates for fluorescence analysis. Cells were pre-incubated with 5 µM CM-H2DCFDA general oxidative stress indicator (C6827, Invitrogen) for 15 min at 37 °C and then treated with 50 µg/mL of oxLDLs during 3 h at 37 °C. As a control, C4-2B or 22RV1 cells were treated with 25 µM Trolox for 2 h and then incubated with 50 µg/mL oxLDLs for 3 h. ## 4.7. Determination of NF-κB and STAT3 Activation Two experimental approaches were used to determine the effects of oxLDL/LOX-1 on NF-κB activation. On the one hand, C4-2B and 22RV1 cells transfected with siRNA control or siRNA LOX-1 for 48 h were treated with 50 µg/mL of oxLDLs for 1 h and the phosphorylation state of p65 and IκB-alpha subunit was analyzed by Western blot. On the other hand, transcriptional activity assays were performed using the luciferase reporter pHAGE NFKB-TA-LUC-UBC-dTomato-W (Catalog # 49335, Adggene) [61]. To this end, C4-2B and 22RV1 cells were co-transfected with a siRNA control or siRNA LOX-1 and the pHAGE NFKB-TA-LUC-UBC-dTomato-W plasmid. Luciferase activity was detected using the Synergy multi plate reader (BioTek) using the Luciferase assay system (E1500, Promega), normalizing against dTomato (excitation: 554 nm, emission: nm 581). For STAT3 activation, Tyr 705 phosphorylation was analyzed using Western blot. In this case, C4-2B and 22RV1 cells transfected with siRNA control or siRNA LOX-1 for 48 h were treated with 50 µg/mL of oxLDLs for 24 h. Furthermore, co-treatment with 50 µg/mL of oxLDLs and ROS inhibitor Trolox, three NF-κB inhibitors: triptolide (TLP) (Cayman Chemical # 11973), bay 11-7082 (BAY) (Cayman Chemical # 10010266), caffeic acid phenethyl ester (CAPE) (Cayman Chemical # 70750) and a chemical inhibitor of STAT3 activity Stattic (STAT3 inhibitor V, CAS 19983-44-9, Santa Cruz Biotechnology) were used in the ROS and pathways inhibition. ## 4.8. Cytotoxic Assay Using Enzalutamide Enzalutamide (CAS No. 915087-33-1, Cayman Chemical) was dissolved at 40 mM in DMSO and stored at −20 °C. The cytotoxic effects of enzalutamide on C4-2B and 22RV1 LOX-1 knockdown cells were determined by clonogenic assay. C4-2B and 22RV1 parental or LOX-1 knockdown cells were seeded on 6-well plates and pre-incubated with or without 50 µg/mL of oxLDL. Then, the cells were incubated with increasing concentrations of enzalutamide (0–50 mM) and cultured for 14 days at 37 °C. Colonies were stained with crystal violet, images were photo-documented using the LI-COR ODYSSEY CLX equipment, and total colony numbers were counted for each condition. ## References 1. Hoover-Plow J., Huang M.. Lipoprotein(a) metabolism: Potential sites for therapeutic targets. Metabolism (2013) 62 479-491. DOI: 10.1016/j.metabol.2012.07.024 2. Austin M.A., Hokanson J., Brunzell J.D.. Characterization of low-density lipoprotein subclasses: Methodologic approaches and clinical relevance. Curr. Opin. Lipidol. (1994) 5 395-403. DOI: 10.1097/00041433-199412000-00002 3. Elshourbagy N.A., Meyers H., Abdel-Meguid S.. Cholesterol: The good; the bad, and the ugly—Therapeutic targets for the treatment of dyslipidemia. Med. Princ. Pract. (2014) 23 99-111. DOI: 10.1159/000356856 4. Koba S., Dyslipidemia T.H., Rinsho N.. Dyslipidemia and Atherosclerosis. Nihon Rinsho (2011) 69 138-143 5. Linton M.R.F., Yancey P.G., Davies S.S., Jerome W.G., Linton E.F., Song W.L., Doran A.C., Vickers K.C., Feingold K.R., Anawalt B., Blackman M.R.. The Role of Lipids and Lipoproteins in Atherosclerosis. Endotext (2000) 6. Chistiakov D.A., Melnichenko A.A., Myasoedova V.A., Grechko A.V., Orekhov A.N.. Mechanisms of foam cell formation in atherosclerosis. J. Mol. Med. (2017) 95 1153-1165. DOI: 10.1007/s00109-017-1575-8 7. Tsai K.L., Chen L.H., Chiou S.H., Chiou G.Y., Chen Y.C., Chou H.Y., Chen L.K., Chen H.Y., Chiu T.H., Tsai C.S.. Coenzyme Q10 suppresses oxLDL-induced endothelial oxidative injuries by the modulation of LOX-1-mediated ROS generation via the AMPK/PKC/NADPH oxidase signaling pathway. Mol. Nutr. Food Res. (2011) 55 S227-S240. DOI: 10.1002/mnfr.201100147 8. Hermonat P.L., Zhu H., Cao M., Mehta J.L.. LOX-1 transcription. Cardiovasc. Drugs Ther. (2011) 25 393-400. DOI: 10.1007/s10557-011-6322-8 9. Kattoor A.J., Goel A., Mehta J.. LOX-1: Regulation, Signaling and Its Role in Atherosclerosis. Antioxidants (2019) 8. DOI: 10.3390/antiox8070218 10. Zingg J.M., Vlad A., Ricciarelli R.. Oxidized LDLs as Signaling Molecules. Antioxidants (2021) 10. DOI: 10.3390/antiox10081184 11. Akhmedov A., Sawamura T., Chen C.H., Kraler S., Vdovemko D., Lüscher T.F.. Lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1): A crucial driver of atherosclerotic cardiovascular disease. Eur. Heart J. (2021) 42 1797-1807. DOI: 10.1093/eurheartj/ehaa770 12. Barreto J., Karathanasis S.K., Remaley A., Sposito A.C.. Role of LOX-1 (Lectin-Like Oxidized Low-Density Lipoprotein Receptor 1) as a Cardiovascular Risk Predictor: Mechanistic Insight and Potential Clinical Use. Arterioscler. Thromb. Vasc. Biol. (2021) 41 153-166. DOI: 10.1161/ATVBAHA.120.315421 13. Jin P., Cong S.. LOX-1 and atherosclerotic-related diseases. Clin. Chim. Acta (2019) 491 24-29. DOI: 10.1016/j.cca.2019.01.006 14. Kattoor A.J., Kanuri S., Mehta J.. Role of Ox-LDL and LOX-1 in Atherogenesis. Curr. Med. Chem. (2019) 26 1693-1700. DOI: 10.2174/0929867325666180508100950 15. Pirillo A., Norata G., Catapano A.. LOX-1, OxLDL. Mediat. Inflamation (2013) 2013 152786 16. Sharma T., Romeo F., Mehta J.. LOX-1: Implications in atherosclerosis and myocardial ischemia. EXCLI J. (2022) 21 273-278. PMID: 35368459 17. Singh S., Gautam A.. Upregulated LOX-1 Receptor: Key Player of the Pathogenesis of Atherosclerosis. Curr. Atheroscler. Rep. (2019) 21 38. DOI: 10.1007/s11883-019-0801-y 18. Madrigal-Ruiz P.M., Navarro R.E., Ruíz S.L., Corona F.I., Vázquez M., Gómez E., Castro J., Sandoval F., Fores L.J., Martín B.T.. Low CD36 and LOX-1 Levels and CD36 Gene Subexpression Are Associated with Metabolic Dysregulation in Older Individuals with Abdominal Obesity. J. Diabetes Res. (2016) 2016 5678946. DOI: 10.1155/2016/5678946 19. Santiago-Fernandez C., Martín-Reyes F., Tome M., Gutierrez-Repiso C., Fernandez-Garcia D., Ocaña-Wilhelmi L., Rivas-Becerra J., Tatzber F., Pursch E., Tinahones F.J.. Oxidized LDL Increase the Proinflammatory Profile of Human Visceral Adipocytes Produced by Hypoxia. Biomedicines (2021) 9. DOI: 10.3390/biomedicines9111715 20. Balzan S., Lubrano V.. LOX-1 receptor: A potential link in atherosclerosis and cancer. Life Sci. (2018) 198 79-86. DOI: 10.1016/j.lfs.2018.02.024 21. Duan Q., Si H., Tian L., Zhang N., Qiu J., Yu J., Liu J., Zhang Q.. LOX-1 attenuates high glucose-induced autophagy via AMPK/HNF4alpha signaling in HLSECs. Heliyon (2022) 8 e12385. DOI: 10.1016/j.heliyon.2022.e12385 22. Deng C.F., Zhu N., Zhao T.J., Li H.F., Gu J., Liao D.F., Qin L.. Involvement of LDL and ox-LDL in Cancer Development and Its Therapeutical Potential. Front. Oncol. (2022) 12 803473. DOI: 10.3389/fonc.2022.803473 23. Hao F., Chen J., Wu J., Ge X., Lv X., Zhang D., Chen J.. Expression of Serum sLOX-1 in Patients with Non-Small-Cell Lung Cancer and Its Correlation with Lipid Metabolism. Can. Respir. J. (2022) 2022 6619331. DOI: 10.1155/2022/6619331 24. Kindt N., Journé F., Carlier S., Trelcat A., Scalia A., Saussez S.. Effect of Oxidized Low-Density Lipoprotein on Head and Neck Squamous Cell Carcinomas. Biomedicines (2021) 9. DOI: 10.3390/biomedicines9050513 25. Li C., Liu F., Yang X., Guo B., Li G., Yin J., He G., Yang C., Xu L., Li S.. Targeting lectin-like oxidized low-density lipoprotein receptor-1 triggers autophagic program in esophageal cancer. Cell Death Differ. (2022) 29 697-708. DOI: 10.1038/s41418-021-00884-y 26. Murdocca M., Torino F., Pucci S., Costantini M., Capuano R., Graggi C., Polidoro C., Somma G., Pasqualetti V., Ketchanji Y.. Urine LOX-1 and Volatilome as Promising Tools towards the Early Detection of Renal Cancer. Cancers (2021) 13. DOI: 10.3390/cancers13164213 27. Tsumita T., Maishi N., Annam D.A., Towfik M.A., Matsuda A., Onodera Y., Nam J.M., Hida Y., Hida K.. The oxidized-LDL/LOX-1 axis in tumor endothelial cells enhances metastasis by recruiting neutrophils and cancer cells. Int. J. Cancer (2022) 151 944-956. DOI: 10.1002/ijc.34134 28. Scalia A., Kindt N., Trelcat A., Nachtergael A., Duez P., Journé F., Carlier S.. Development of a Method for Producing oxLDL: Characterization of Their Effects on HPV-Positive Head and Neck Cancer Cells. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms232012552 29. Xu Q., Wu Q., Li H., Tian X., Zuo Y., Zhang Y., Zhang X., Lin Y., Wu Y., Wang Y.. Joint High Level of Oxidized Low-Density Lipoprotein and High-Sensitivity C-Reactive Protein are Associated with Recurrent Stroke and Poor Functional Outcome in Minor Stroke or Transient Ischemic Attack. J. Am. Heart Assoc. (2022) 11 e027665. DOI: 10.1161/JAHA.122.027665 30. Zheng S.M., Chen H., Sha W.H., Chen X.F., Yin J.B., Zhu X.B., Zheng Z.W., Ma J.. Oxidized low-density lipoprotein stimulates CD206 positive macrophages upregulating CD44 and CD133 expression in colorectal cancer with high-fat diet. World J. Gastroenterol. (2022) 28 4993-5006. DOI: 10.3748/wjg.v28.i34.4993 31. Gonzalez-Chavarria I., Cerro R.P., Parra N.P., Sandoval F.A., Zuñiga F., Omazábal V.A., Lamperti L.I., Jiménez S.P., Fernandez E.A., Gutiérrez N.A.. Lectin-like oxidized LDL receptor-1 is an enhancer of tumor angiogenesis in human prostate cancer cells. PLoS ONE (2014) 9. DOI: 10.1371/journal.pone.0106219 32. Gonzalez-Chavarria I., Fernandez E., Gutierrez N., González-Horta E.E., Sandoval F., Cifuentes P., Castillo C., Cerro C., Sanchez O., Toledo J.R.. LOX-1 activation by oxLDL triggers an epithelial mesenchymal transition and promotes tumorigenic potential in prostate cancer cells. Cancer Lett. (2018) 414 34-43. DOI: 10.1016/j.canlet.2017.10.035 33. Jiang J., Yan M., Metha J.L., Hu C.. Angiogenesis is a link between atherosclerosis and tumorigenesis: Role of LOX-1. Cardiovasc. Drugs Ther. (2011) 25 461-468. DOI: 10.1007/s10557-011-6343-3 34. Lu J., Mitra S., Wang X., Khaidakov M., Mehta J.L.. Oxidative stress and lectin-like ox-LDL-receptor LOX-1 in atherogenesis and tumorigenesis. Antioxid. Redox Signal. (2011) 15 2301-2333. DOI: 10.1089/ars.2010.3792 35. Murdocca M., De Masi C., Pucci S., Mango R., Novelli G., Di Natale C., Sangiuolo F.. LOX-1 and cancer: An indissoluble liaison. Cancer Gene Ther. (2021) 28 1088-1098. DOI: 10.1038/s41417-020-00279-0 36. Rizzacasa B., Morini E., Pucci S., Murdocca M., Novelli G., Amati. F.. LOX-1 and Its Splice Variants: A New Challenge for Atherosclerosis and Cancer-Targeted Therapies. Int. J. Mol. Sci. (2017) 18. DOI: 10.3390/ijms18020290 37. Rawla P.. Epidemiology of Prostate Cancer. World J. Oncol. (2019) 10 63-89. DOI: 10.14740/wjon1191 38. Karantanos T., Corn P., Thompson T.. Prostate cancer progression after androgen deprivation therapy: Mechanisms of castrate resistance and novel therapeutic approaches. Oncogene (2013) 32 5501-5511. DOI: 10.1038/onc.2013.206 39. Harris W.P., Mostaghel E.A., Nelson P.S., Montgomery B.. Androgen deprivation therapy: Progress in understanding mechanisms of resistance and optimizing androgen depletion. Nat. Clin. Pract. Urol. (2009) 6 76-85. DOI: 10.1038/ncpuro1296 40. Saad F., Hotte S.. Guidelines for the management of castrate-resistant prostate cancer. Can. Urol. Assoc. J. (2010) 4 380-384. DOI: 10.5489/cuaj.10167 41. Drake J.M., Graham N.A., Lee J.K., Stoyanova T., Faltermeier C.M., Sud S., Titz B., Huang J., Pienta K.J., Graeber T.G.. Metastatic castration-resistant prostate cancer reveals intrapatient similarity and interpatient heterogeneity of therapeutic kinase targets. Proc. Natl. Acad. Sci. USA (2013) 110 E4762-E4769. DOI: 10.1073/pnas.1319948110 42. Chen X., Lu J., Xia L., Li G.. Drug Resistance of Enzalutamide in CRPC. Curr. Drug Targets (2018) 19 613-620. DOI: 10.2174/1389450118666170417144250 43. Adelaiye-Ogala R., Gryder B.E., Nguyen Y.T.M., Alilin A.N., Grayson A.R., Bajwa W., Jansson K.H., Beshiri M.L., Agarwal S., Rodriguez-Nieves J.A.. Targeting the PI3K/AKT Pathway Overcomes Enzalutamide Resistance by Inhibiting Induction of the Glucocorticoid Receptor. Mol. Cancer Ther. (2020) 19 1436-1447. DOI: 10.1158/1535-7163.MCT-19-0936 44. Culig Z.. Response to Androgens and Androgen Receptor Antagonists in the Presence of Cytokines in Prostate Cancer. Cancers (2021) 13. DOI: 10.3390/cancers13122944 45. Wang C., Peng G., Huang H., Liu F., Kong D.P., Dong K.Q., Dai L.H., Zhou Z., Wang K.J., Yang J.. Blocking the Feedback Loop between Neuroendocrine Differentiation and Macrophages Improves the Therapeutic Effects of Enzalutamide (MDV3100) on Prostate Cancer. Clin. Cancer Res. (2018) 24 708-723. DOI: 10.1158/1078-0432.CCR-17-2446 46. Tousignant K.D., Rockstroh A., Poad B.L.J., Talebi A., Young R.S.E., Fard A.T., Gupta R., Zang T., Wang C., Lehman M.L.. Therapy-induced lipid uptake and remodeling underpin ferroptosis hypersensitivity in prostate cancer. Cancer Metab. (2020) 8 11. DOI: 10.1186/s40170-020-00217-6 47. Canton J., Neculai D., Grinstein S.. Scavenger receptors in homeostasis and immunity. Nat. Rev. Immunol. (2013) 13 621-634. DOI: 10.1038/nri3515 48. Burtenshaw D., Kitching M., Redmond E.M., Megson I.L., Cahill P.A.. Reactive Oxygen Species (ROS), Intimal Thickening, and Subclinical Atherosclerotic Disease. Front. Cardiovasc. Med. (2019) 6 89. DOI: 10.3389/fcvm.2019.00089 49. Liu V.W.S., Yau W.L., Tam C.W., Yao K.M., Shiu S.Y.W.. Melatonin Inhibits Androgen Receptor Splice Variant-7 (AR-V7)-Induced Nuclear Factor-Kappa B (NF-kappaB) Activation and NF-kappaB Activator-Induced AR-V7 Expression in Prostate Cancer Cells: Potential Implications for the Use of Melatonin in Castration-Resistant Prostate Cancer (CRPC) Therapy. Int. J. Mol. Sci. (2017) 18 1130. PMID: 28561752 50. Xu L., Chen X., Shen M., Dong-Rong Y., Fang L., Weng G., Tsai Y., Keng P.C., Chen Y., Lee S.O.. Inhibition of IL-6-JAK/Stat3 signaling in castration-resistant prostate cancer cells enhances the NK cell-mediated cytotoxicity via alteration of PD-L1/NKG2D ligand levels. Mol. Oncol. (2018) 12 269-286. DOI: 10.1002/1878-0261.12135 51. Liu C., Zhu Y., Lou W., Cui Y., Evans C.P., Gao A.C.. Inhibition of constitutively active Stat3 reverses enzalutamide resistance in LNCaP derivative prostate cancer cells. Prostate (2014) 74 201-209. DOI: 10.1002/pros.22741 52. Antonarakis E.S., Lu C., Wang H., Luber B., Nakazawa M., Roeser J.A., Chen Y., Mohammad T.A., Chen Y., Fedor H.L.. AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N. Engl. J. Med. (2014) 371 1028-1038. DOI: 10.1056/NEJMoa1315815 53. Sprenger C.. Androgen Receptor Variants and Castration-resistant Prostate Cancer: Looking Back and Looking Forward. Eur. Urol. (2018) 73 724-726. PMID: 29306511 54. Buttigliero C., Tucci M., Bertaglia F., Bironzo P., Di Maio M., Scagliotti G.V.. Understanding and overcoming the mechanisms of primary and acquired resistance to abiraterone and enzalutamide in castration resistant prostate cancer. Cancer Treat. Rev. (2015) 41 884-892. DOI: 10.1016/j.ctrv.2015.08.002 55. Zhu Y., Dalrymple S.L., Coleman I., Zheng S.L., Xu J., Hooper J.E., Antonarakis E.S., De Marzo A.M., Meeker A.K., Nelson P.S.. Role of androgen receptor splice variant-7 (AR-V7) in prostate cancer resistance to 2nd-generation androgen receptor signaling inhibitors. Oncogene (2020) 39 6935-6949. DOI: 10.1038/s41388-020-01479-6 56. Fujita K., Nonomura N.. Role of Androgen Receptor in Prostate Cancer: A Review. World J. Mens Health (2019) 37 288-295. DOI: 10.5534/wjmh.180040 57. Thaper D., Vahid S., Kaur R., Kumar S., Nouruzi S., Bishop J.L., Johansson M., Zoubeidi A.. Galiellalactone inhibits the STAT3/AR signaling axis and suppresses Enzalutamide-resistant Prostate Cancer. Sci. Rep. (2018) 8 17307. DOI: 10.1038/s41598-018-35612-z 58. Xu S., Ogura S., Chen J., Little P.J., Moss J., Liu P.. LOX-1 in atherosclerosis: Biological functions and pharmacological modifiers. Cell Mol. Life Sci. (2013) 70 2859-2872. DOI: 10.1007/s00018-012-1194-z 59. Don-Doncow N., Marginean F., Coleman I., Nelson P.S., Ehrnström R., Krzyzanowska A., Morrissey C., Hellsten R., Bjartell A.. Expression of STAT3 in Prostate Cancer Metastases. Eur. Urol. (2017) 71 313-316. DOI: 10.1016/j.eururo.2016.06.018 60. Lubrano V., Balzan S.. LOX-1; ROS, inseparable factors in the process of endothelial damage. Free Radic. Res. (2014) 48 841-848. DOI: 10.3109/10715762.2014.929122 61. Sata M., Walsh K.. Endothelial cell apoptosis induced by oxidized LDL is associated with the down-regulation of the cellular caspase inhibitor FLIP. J. Biol. Chem. (1998) 273 33103-33106. DOI: 10.1074/jbc.273.50.33103 62. Heinloth A., Heermeier K., Raff U., Wanner C., Galle J.A.N.. Stimulation of NADPH oxidase by oxidized low-density lipoprotein induces proliferation of human vascular endothelial cells. J. Am. Soc. Nephrol. (2000) 11 1819-1825. DOI: 10.1681/ASN.V11101819 63. Hirsch H.A., Iliopoulos D., Joshi A., Zhang Y., Jaeger S.A., Bulyk M., Tsichlis P.N., Liu S., Strulh K.. A transcriptional signature and common gene networks link cancer with lipid metabolism and diverse human diseases. Cancer Cell (2010) 17 348-361. DOI: 10.1016/j.ccr.2010.01.022 64. Khaidakov M., Mehta J.. Oxidized LDL triggers pro-oncogenic signaling in human breast mammary epithelial cells partly via stimulation of MiR-21. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0046973 65. Bubici C., Papa S., Dean K., Franzoso G.. Mutual cross-talk between reactive oxygen species and nuclear factor-kappa B: Molecular basis and biological significance. Oncogene (2006) 25 6731-6748. DOI: 10.1038/sj.onc.1209936 66. Lingappan K.. NF-kappaB in Oxidative Stress. Curr. Opin. Toxicol. (2018) 7 81-86. DOI: 10.1016/j.cotox.2017.11.002 67. Morgan M.J., Liu Z.. Crosstalk of reactive oxygen species and NF-kappaB signaling. Cell Res. (2011) 21 103-115. DOI: 10.1038/cr.2010.178 68. Chandrasekar T., Yang J.C., Gao A.C., Evans C.P.. Mechanisms of resistance in castration-resistant prostate cancer (CRPC). Transl. Androl. Urol. (2015) 4 365-380. PMID: 26814148 69. Johnson D.E., O’Keefe R.A., Grandis J.. Targeting the IL-6/JAK/STAT3 signalling axis in cancer. Nat. Rev. Clin. Oncol. (2018) 15 234-248. DOI: 10.1038/nrclinonc.2018.8 70. Schust J., Sperl B., Hollis A., Mayer T.U., Berg T.. Stattic: A small-molecule inhibitor of STAT3 activation and dimerization. Chem. Biol. (2006) 13 1235-1242. DOI: 10.1016/j.chembiol.2006.09.018 71. Culig Z., Puhr M.. Interleukin-6: A multifunctional targetable cytokine in human prostate cancer. Mol. Cell Endocrinol. (2012) 360 52-58. DOI: 10.1016/j.mce.2011.05.033 72. Lou W., Ni Z., Dyer K., Tweardy D.J., Gao A.C.. Interleukin-6 induces prostate cancer cell growth accompanied by activation of stat3 signaling pathway. Prostate (2000) 42 239-242. DOI: 10.1002/(SICI)1097-0045(20000215)42:3<239::AID-PROS10>3.0.CO;2-G 73. Sreenivasan L., Wang H., Yap S.Q., Leclair P., Tam A., Lim C.J.. Autocrine IL-6/STAT3 signaling aids development of acquired drug resistance in Group 3 medulloblastoma. Cell Death Dis. (2020) 11 1035. DOI: 10.1038/s41419-020-03241-y 74. Jin R., Yamashita H., Yu X., Wang J., Franco O.E., Hayward S.W., Matusik R.J.. Inhibition of NF-kappa B signaling restores responsiveness of castrate-resistant prostate cancer cells to anti-androgen treatment by decreasing androgen receptor-variant expression. Oncogene (2015) 34 3700-3710. DOI: 10.1038/onc.2014.302 75. Austin D.C., Strand D.W., Love H.L., Franco O.E., Jang A., Grabowska M.M., Miller N.L., Hameed O., Clark P.E., Fowke J.H.. NF-kappaB and androgen receptor variant expression correlate with human BPH progression. Prostate (2016) 76 491-511. DOI: 10.1002/pros.23140 76. Kiliccioglu I., Konac E., Dikmen A.U., Sozen S., Bilen C.Y.. Hsp-27 and NF-kappaB pathway is associated with AR/AR-V7 expression in prostate cancer cells. Gene (2019) 697 138-143. DOI: 10.1016/j.gene.2019.02.055 77. Nadiminty N., Tummala R., Liu C., Yang J., Lou W., Evans C.P., Gao A.C.. NF-kappaB2/p52 induces resistance to enzalutamide in prostate cancer: Role of androgen receptor and its variants. Mol. Cancer Ther. (2013) 12 1629-1637. DOI: 10.1158/1535-7163.MCT-13-0027 78. Kiliccioglu I., Bilen C.Y., Sozen S., Konac E.. Upregulation of potential regulatory signaling molecules correlate with androgen receptor splice variants AR-V7 and AR-V567es in prostate cancer metastasis. Gene (2021) 772 145377. DOI: 10.1016/j.gene.2020.145377 79. Gales L., Forsea L., Mitrea D., Stefanica I., Stanculescu I., Mitrica R., Georgescu M., Trifanescu O., Anghel R., Serbanescu L.. Antidiabetics, Anthelmintics, Statins, and Beta-Blockers as Co-Adjuvant Drugs in Cancer Therapy. Medicina (2022) 58. DOI: 10.3390/medicina58091239 80. Biocca S., Iacovelli F., Matarazzo S., Vindigni G., Oteri F., Desideri A., Falconi M.. Molecular mechanism of statin-mediated LOX-1 inhibition. Cell Cycle (2015) 14 1583-1595. DOI: 10.1080/15384101.2015.1026486 81. Wadosky K.M., Koochekpour S.. Androgen receptor splice variants and prostate cancer: From bench to bedside. Oncotarget (2017) 8 18550-18576. DOI: 10.18632/oncotarget.14537 82. Wu X., Gong S., Roy-Burman P., Lee P., Culig Z.. Current mouse and cell models in prostate cancer research. Endocr. Relat. Cancer (2013) 20 R155-R170. DOI: 10.1530/ERC-12-0285
--- title: The Impact of Temperature on 24-Hour Movement Behaviors among Chinese Freshmen Students authors: - Hongjun Yu - Yiling Song - Yangyang Wang - Xiaoxin Wang - Haoxuan Li - Xiaolu Feng - Miao Yu journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049201 doi: 10.3390/ijerph20064970 license: CC BY 4.0 --- # The Impact of Temperature on 24-Hour Movement Behaviors among Chinese Freshmen Students ## Abstract Background: Human populations worldwide have experienced substantial climate change issues. Gaps in scientific literature remain regarding the relationship between temperature and 24-hour movement behavior among people. The purpose of this study is to examine the impact of temperature on 24-hour movement behavior including physical activity (PA), sedentary behavior (SB) and sleep duration among university students living in Beijing, China. Methods: We conducted follow-up health surveys on 44,693 freshmen students enrolled at Tsinghua University from 2012 to 2018. PA and SB were measured by using the short version of the International Physical Activity Questionnaire (IPAQ-s); sleep duration was estimated by using The Pittsburgh Sleep Quality Index (CPSQI). Corresponding temperature data measured by the Beijing Meteorological Service were collected to include average daily temperature from the nearest weather station to Tsinghua university. The data were analyzed using linear individual fixed-effect regressions. Results: An increase in temperature (temperature range 2.29–28.73 °C) by 1 °C was associated with an increase in 0.66 weekly minutes of vigorous physical activity (VPA) ($95\%$ confidence interval [CI] = 0.49, 0.82), an increase in 0.56 weekly minutes of moderate physical activity (MPA)($95\%$ CI = 0.32, 0.79), an increase in 1.21 weekly minutes of moderate to vigorous physical activity (MVPA) ($95\%$ CI = 0.90, 1.53), an increase in 0.55 weekly minutes of walking ($95\%$ CI = 0.31, 0.78), an increase in 1.76 weekly minutes of total PA ($95\%$ CI = 1.35, 2.17), and a reduction in 1.60 weekly minutes of sleeping ($95\%$ CI = −2.09, −1.11). There was no significant correlation between temperature and sedentary behavior among participants. Conclusions: Temperature was significantly positively correlated with physical activity levels in the Chinese freshmen students, and significantly negatively correlated with sleep duration. Replication of this study is warranted among various populations within China. The evidence of this novel study focused on understanding the relationship between climate change and 24-hour movement behaviors among people for developing effective adaptation strategies to climate change to improve people’s health behavior. This study has important implications for future study, as knowledge of the impact of temperature on movement behavior may help in the interpretation of their results and translate into improving people’s health behavior. ## 1. Introduction Climate change poses a great threat to human lives and health in a variety of ways and has become a concern in global health [1]. Studies have shown that climate change has direct or indirect effects on human health [2]. For instance, changes in extreme temperatures can lead to increased morbidity and mortality [3], as well as exposure to toxic substances and degraded air quality [4], all of which affect human health. Even climate change can affect people’s mental health [5]. The health problems caused by climate change have become one of the most important public health problems [6]. At present, physical inactivity is a global public health problem and is the fourth leading cause of death globally [7]. Physical activity (PA) is any bodily movement produced by skeletal muscle contraction that causes energy expenditure to exceed the resting metabolic rate [8]. Regular PA is associated with a reduced risk of all-cause mortality, cardiovascular disease, type 2 diabetes, many cancers, as well as better mental health and quality of life [9,10,11]. Public health guidelines recommend that university students have at least 150 min of moderate physical activity (MPA) or 75 min of vigorous physical activity (VPA) per week [12,13,14]. Sedentary behavior (SB) is a conscious activity characterized by low energy expenditure and performed in a sitting or reclining position, such as watching TV, reading, or driving [15]. Moreover, there is evidence that SB is also a risk factor for health [16]. For example, too much SB may impair cognitive function in the brain [17]. In addition, sleep duration plays a vital role in social, physical, psychological and cognitive health, and it is also an important part of overall health and well-being [18]. PA, SB, and sleep duration are the main movement behaviors that occur in individuals within 24 hours, which are referred to as 24-hour movement behaviors [19]. Appropriate 24-hour movement behavior has important implications for individual health and disease prevention. Therefore, the 24-hour movement behavior of individuals deserves our attention. With global average temperatures likely to be 2 °C above pre-industrial levels by the end of the 21st century [20], understanding the impact of climate change on 24-hour movement behaviors is important. A recent review has reported that climate change affects individuals’ 24-hour movement behaviors [21]. Researchers need to further explore the impact of climate change in order to propose targeted countermeasures for the improvement of human health behavior. For instance, an individual’s physical activity behavior is affected by a variety of factors [22], such as social relationships, education, and climate change. Among the climate change factors that affect physical activity, temperature plays an important role [23,24,25,26]. The natural environment includes many factors that affect physical activity, such as seasons and weather [23]. Previous literature has reported that ambient temperature has a potential impact on physical activity [24]. A cross-sectional study from Norway showed that adolescents had higher odds of meeting physical activity recommendations in spring compared to winter, meaning that a higher temperature was associated with higher levels of physical activity in adolescents [27]. Moreover, one study found that adults were more likely to participate in walking for recreation in summer, autumn and spring compared to winter; more likely to participate in walking for transportation in autumn; and more likely to participate in MPA in summer. The likelihood of achieving adequate MPA was significantly greater in summer, autumn and spring [28]. Another study reported that physical activity volume and more moderate to vigorous physical activity (MVPA) of participants in summer were greater than those in winter, and noted that physical activity volume and MVPA of participants were positively correlated with temperature [29]. Furthermore, a study from Canada showed that Canadians had an average total daily energy expenditure of $31.0\%$ higher in summer than in winter, and were $86\%$ more likely to be active in recreational physical activity in summer than in winter [30]. Ho et al. [ 31] recently reported that high temperature (between 32.6 °C and 36.5 °C) was associated with decreases of 800–1500 daily steps compared to optimal temperature (between 16 °C and 19.3 °C) among Chinese adults. A recent systematic review and meta-analysis showed that, in children and adolescents, higher temperature was associated with MVPA, while lower temperature was associated with more SB [32]. In addition, a systematic review has shown that ambient temperature also affects an individual’s sleep duration and quality [33]. Research with American teenagers also showed that, when the temperature increased from −13.3 °C to 8.3 °C, the midpoint time of sleep moved to an earlier time, and then when the temperature increased from 8.3 °C to 31.7 °C, the midpoint time of sleep moved to a later time, which showed that the midpoint time of sleep will also be affected by temperature [34]. Despite the aforementioned work, three major gaps in the previous studies remain. First, although there are currently some studies on seasonal climate change and individual health behaviors, there is a lack of studies on the impact of temperature on individual 24-hour movement behavior. Second, there are only a few studies with large cohort samples. Third, studies that consider PA, SB and sleep duration at temperatures among different seasons are lacking. With continued greenhouse gas emissions and global warming due to urban development, we need to consider combining temperature management strategies with PA interventions to potentially promote comfortable and safe PA [35]. The freshmen are in a transitional phase from adolescence to adulthood, a critical time for developing healthy lifestyles and healthy behaviors. Meanwhile, compared to high school, college is a new living environment for freshmen, and their learning style and lifestyle have changed dramatically. Freshmen are experiencing this change and will have more autonomy in choosing their lives and studies, and freshmen are more susceptible to the influence of the outside environment during this period. Therefore, the purpose of this study is to evaluate the impact of temperature on 24-h movement behavior among Chinese freshmen students and to provide a theoretical basis for relevant departments to formulate environmental intervention policies and measures. For example, campus environmental interventions and educational guidance can be used to enable freshmen students to adapt to their new environment while developing healthy behavioral habits. We hypothesized that temperature was positively correlated with PA in Chinese freshmen students, and negatively correlated with SB and sleep duration. ## 2.1. Participants Data from this study were obtained from a paper–pencil-based follow-up health survey that all freshmen at Tsinghua University are asked to fill out. Survey participation was voluntary, and all participants signed a consent form. The study was approved by the Tsinghua University Institutional Review Board (IRB#2012534001). The survey is designed to evaluate the participants’ health behavior, health conditions, and the detailed descriptions of the questionnaire content, implementation procedures, and member cohorts have been published and available elsewhere [36,37]. Briefly, the faculty administered the survey in a class, and all of the freshmen completed the survey within a health education class. In the survey, participants reported their student identification number, which was used to link multiple survey questionnaires completed by the same respondent. Briefly, the freshmen participants enrolled in a health education class (every other day, Monday to Friday) and completed a faculty-administered paper-based questionnaire. Students handed in the questionnaires at the end of the class. Data on six cohort groups were administered at Tsinghua University: 2012–2013 cohort, 2013–2014 cohort, 2014–2015 cohort, 2015–2016 cohort, 2016–2017 cohort, and 2017–2018 cohort ($$n = 44$$,693). ( See Table 1). The inclusion criteria of this study were as follows: [1] free from diseases or any medical conditions; [2] obtained written informed consent to participate in the study; [3] only freshmen who completed the survey at least two times were included in the analysis. The exclusion criteria of this study were as follows: [1] missing data on physical activity, sedentary behavior or sleep; [2] the participant completed the survey only once. The final sample ($$n = 33$$,923) had non-missing values for the specific outcome and all covariates (See Figure 1). ## 2.2. Sleep Measurement We assessed participants’ sleep duration and used The Pittsburgh Sleep Quality Index (CPSQI), which has been validated in China, to measure sleep quality [38]. We modified the question to collect data from the Chinese version of The Pittsburgh Sleep Quality Index (CPSQI): “During the last week, on average, how many hours of sleep did you get per night?” [ 38,39,40]. ## 2.3. Physical Activity Measurement The short version of the International Physical Activity Questionnaire (IPAQ) was used to measure physical activity. The short version of the IPAQ includes 9 items and has been validated in China [41,42,43]. Data from the short version of the IPAQ provided information on the time spent walking and time spent doing moderate and vigorous intensity activities. Total minutes of VPA in the last week were calculated based on the answers to two questions adapted from the IPAQ: “During the last seven days, how many days did you do vigorous physical activities, such as aerobics, running, fast bicycling or fast swimming?” and “How much time did you usually spend doing vigorous physical activities on one of those days?”. Total minutes of VPA in the last week were calculated through multiplying daily average number of minutes spent on VPA by its corresponding number of days. Total minutes of moderate physical activity (MPA) in the last week were calculated based on the answers to two questions adapted from the IPAQ: “During the last seven days, how many days did you do moderate physical activities, such as carrying light loads, bicycling at a regular pace or playing doubles tennis?” and “How much time did you usually spend doing moderate physical activities on one of those days?”. Total minutes of MPA in the last week were calculated through multiplying daily average number of minutes spent on VPA by its corresponding number of days. Total minutes of walking—light physical activity (LPA)—in the last week were calculated based on the answers to two questions adapted from the IPAQ: “During the last seven days, on how many days did you walk for at least 10 min at a time?” and “How many minutes did you usually spend on one of those days’ walking?”. Total minutes of walking in the last week were calculated through multiplying daily average minutes spent on walking by its corresponding number of days. Total weekly minutes of MVPA were calculated through multiplying weekly average minutes spent on MPA by weekly average minutes spent doing VPA. ## 2.4. Sedentary Behavior Measurement The short version of the International Physical Activity Questionnaire (IPAQ), which has been validated in the Chinese population, was used to measure SB [42]. Total hours of SB in the last week were calculated based on the answers to the question adapted from (IPAQ): “During the last 7 days, how much time did you usually spend sitting on a day?” [ 44] Total hours of SB in the last week were calculated through multiplying daily average hours spent on SB by seven. ## 2.5. Environmental Measures Meteorological data were obtained on each day that data were collected via the publicly accessible information of the China Meteorological Administration. The nearest weather station to Tsinghua university was identified (Haidian Meteorological Bureau), which is around 4 KM from Tsinghua University. The obtained data were temperature (daily maximum and daily minimum in Celsius degrees), relative humidity (daily average in percentage), average wind speed (m/s), and percentage of rainy days in Beijing, China. These measures were taken seven days before the survey was given. Air pollution data (PM2.5) were provided by the Beijing Municipal Ecological Environment Bureau (from 6 December 2013), and the hourly ambient PM2.5 concentration data were provided by the Mission China Air Quality Monitoring Program, which was run by the U.S. Department of State (before 6 December 2013, for the 2012–2013 cohort). ## 2.6. Statistical Analyses Means, SD, and percentages were summarized and compared for characteristics of the overall sample. Continuous variables obtained from follow-up surveys were examined using the one-way repeated measures ANOVA test and t-test. Categorical variables were collated using chi-square tests. Linear individual fixed-effect regressions were performed based on the repeated-measure survey data from the four freshmen cohorts (i.e., 2012–2013, 2013–2014, 2014–2015, 2015–2016, 2016–2017 and 2017–2018). Weekly average minutes of VPA, MPA, and MVPA; daily average hours of sedentary behavior; and daily average hours of sleep in the past week were used for the study’s continuous outcome variable. The key independent variables of the study were average temperature in the past one week. Individual-level time-variant covariates and environmental measures (i.e., average wind speed and percentage of rainy days) and air pollution measure (PM2.5) are controlled for the study’s independent variables. An entire sample with both genders, male only, and female only used separate regressions that were conducted for each outcome variable. We examined the effects of temperature on individual-level physical activity, sedentary behavior and sleep duration outcomes by using linear individual fixed-effect regressions based on the repeated-measure survey data from the six freshmen cohorts (i.e., 2012–2018) at Tsinghua University. In all regressions, the key independent variable was the standardized, average temperature over the last seven days before the survey. All models were controlled for the individual-level time-variant covariates as well as the environmental variables, including average PM2.5 concentration, average wind speed, and percentage of rainy days during the last seven days. Our previous studies [36,37] showed that air pollution had different effects on sedentary behavior and physical activity by gender. Thus, in this study, we examined the effects of temperature on PA, SB and sleep by gender. Each outcome variable was then analyzed by using separate regression and stratified by gender. The use of fixed effects models is widely accepted for cohort or longitudinal data analysis. Using fixed-effects (FE) models means that we are only interested in analyzing the impact of PA, SB and sleep at varying temperatures. The FE model explores the relationship between temperature and PA, SB and sleep variables within all participants. Each participant has their own individual characteristics that may or may not influence PA, SB and sleep duration (for example, BMI, habits, and personal preferences). When using FE, we assume that something within the individual may impact or bias the PA, SB and sleep outcome variables and that is what we need to control. This is the rationale behind the assumption of the correlation between participants’ error term. The FE model treats individuals as their own control; all between-person time variation is conditioned out of the model and the model mainly analyzes within-person variation. Furthermore, FE remove the effect of those time-invariant characteristics so we can assess the net relationship between the temperature and PA, SB and sleep. Another important assumption of the FE model is that those time-invariant characteristics are unique to the individual and should not be correlated with other individual characteristics. Potential omitted variable bias due to differences in time-invariant individual characteristics such as genes, gender, ethnicity, habits, and personal preferences was removed through our use of individual fixed-effect regression. Individual fixed-effect regressions could only estimate the effect of a time-variant independent variable due to its exclusive dependence upon within-individual variations in an outcome measure. Therefore, study participant gender, ethnicity, and other time-invariant individual characteristics were not examined. ## 2.7. Individual-Level Covariates Individual-level time-variant covariates were controlled for in the regression analyses. Continuous variables for age in years; body mass index (BMI; kg/m2); self-rated physical health (1–10, poor-excellent); and self-rated mental health (1–10, poor-excellent) were included. Dichotomous variables for current smoking status (non-smokers as the reference group) and drinking status (non-drinkers as the reference group) were additionally measured. All statistical procedures were performed using Stata 17.0 SE version (StataCorp, College Station, TX, USA). The Eicker-Huber-White sandwich estimator addressed within-individual serial correlations and was used to estimate the standard errors of regression coefficients. ## 3. Results Table 1 presents the baseline characteristics of the survey participants. A majority of the sample was male participants ($67.53\%$). The mean BMI of participants was 21.39 kg/m2 (SD = 3.55). Moreover, only $0.49\%$ of participants were current smokers and $2.58\%$ were current drinkers. The self-rated physical health and self-rated mental health scores had mean values of 5.34 (SD = 2.19) and 6.23 (SD = 2.46), respectively. There were significant differences between males and females in all the above variables ($p \leq 0.001$). Table 2 shows the average physical activity, temperature, and other environmental variables in the last seven days before the survey. The average temperature increased from 6.29 °C (SD = 3.45) to 21.43 °C (SD = 5.16), and the weekly total minutes of PA increased from 348.53 min/week (SD = 305.60) to 419.29 min/week (SD = 282.63) within the 2014–2015 cohort. The temperature increased from 25.57 °C (SD = 2.44) to 27.00 °C (SD = 0.82), and the daily total hours of sleeping decreased from 7.34 h/day (SD = 1.01) to 7.03 hr/day (SD = 0.73) within the 2015–2016 cohort. The temperature increased from 5.57 °C (SD = 1.72) to 20.86 °C (SD = 5.18), and the daily total hours of sedentary behavior decreased from 9.45 h/day (SD = 2.86) to 9.22 h/day (SD = 2.90) within the 2016–2017 cohort. Table 3 reports the estimated effects of temperature on individual-level 24-hour movement behavior outcomes in the past week per day using linear individual fixed-effect regressions. Temperatures were found to be positively associated with total minutes of physical activity in the last week among survey participants. Specifically, an increase in temperature by 1 °C was associated with an increase in weekly minutes of VPA by 0.66 ($95\%$ confidence interval [CI] = 0.49, 0.82), an increase in weekly minutes of MPA by 0.56 ($95\%$ CI = 0.32, 0.79), an increase in weekly minutes of MVPA by 1.21 ($95\%$ CI = 0.90, 1.53), an increase in weekly minutes of walking by 0.55 ($95\%$ CI = 0.31, 0.78), and an increase in weekly minutes of total PA by 1.76 ($95\%$ CI = 1.35, 2.17). There was no significant correlation between temperature and SB among survey participants. However, temperature was found to be negatively associated with the total minutes of sleep duration in the last week among survey participants. Specifically, an increase in temperature by 1 °C was associated with a reduction in daily minutes of sleeping by 1.60 ($95\%$ CI = −2.09, −1.11). There were no significant differences between temperature and PA and sleep duration by gender. ## 4. Discussion The purpose of this study was to investigate the impact of temperature on 24-hour movement behaviors among Chinese freshmen students. Our study found that temperature had a significant positive correlation with PA behavior in the Chinese freshmen students and a significant negative correlation with sleep duration, which was consistent with some of our hypothesis. However, we did not find a significant negative correlation between temperature and SB. This study found a significant positive correlation between temperature and individual PA levels in a survey of the Chinese freshmen students, with higher temperature being associated with higher levels of PA, which is consistent with previous findings. A study of 1115 Auckland children showed that for boys, a 10 °C increase in average ambient temperature was associated with a small increase in steps on weekdays and a modest increase in steps on weekends, while the effects for girls were small and unclear [45]. The reasons for this gender-related difference are uncertain, but researchers suggested that outdoor activities, which may be affected by ambient temperature, are more popular among boys than among girls [45]. In addition, a study reported a positive correlation between ambient temperature and physical activity, showing that daily temperature increases strongly predicted increases in daily steps. Regardless of month, every 10 °C increase in temperature was associated with a $2.9\%$ increase in daily steps [46]. Furthermore, a study of 1293 Canadian teenagers showed that for every 10 °C rise in temperature, the time of PA among teenagers would increase by $1\%$ to $2\%$ every day. Moreover, teenagers had less physical activity in winter and more physical activity in warm months [47]. Based on previous studies reporting that levels of PA appear to vary seasonally, the subsequent effects of severe or extreme weather have been identified as barriers to participation in physical activity in different populations [23]. For example, a study from China showed that the average steps of walks for adults in five major Chinese cities at extremely high temperatures(between 32.6 °C and 36.5 °C) dropped by 800 to 1500 steps compared to optimal temperatures (between 16 °C and 19.3 °C) [31]. Obradovich et al. shows that both cold and extremely hot temperatures reduce PA [48]. This is because extreme temperatures may reduce PA due to thermal discomfort, and a cooler climate may have a positive effect on physical activity [35]. In addition, some studies believe that if a person performs PA in a high-temperature environment, it will cause pressure on the individual’s thermoregulatory system, and then the individual may change their behavior [49]. Moreover, a study from the United States [50] reported a non-linear effect of increasing temperature on healthy cycling. Data analysis showed that with increasing temperature, the number of daily cycling hours and distance traveled by individuals increased significantly, but then declined at temperatures above 26–28 °C [50]. Turrisi et al. [ 29] reported that the natural environment can affect health by promoting or inhibiting PA. Interventions in behavior have to take into account possible weather effects. Extreme weather brought on by climate change may short-circuit health-promoting physical activity and, over time, encourage migration of people in the quest of more hospitable climatic niches. However, there are also some studies that are inconsistent with the findings of this study. For example, a study of 2088 adults in the Arabian Gulf region showed that higher average temperatures and increased humidity were associated with a decrease in the number of steps taken per day [51]. Another study of 40 Chinese adults showed no significant relationship between temperature and PA levels [52]. Inconsistency regarding these finding may be explained by the social or cultural differences among the participants. Consistent with previous research that there was no significant correlation between temperature and SB among our survey participants [53,54,55,56,57,58], a study on a sample of 740 Hong Kong adolescents reported that there was no significant relationship between temperature and SB during weekdays [57]. Another study from 722 children aged 10–12 years in five countries found that temperature was not significantly associated with SB [58]. However, this study was inconsistent with previous studies in that there was a significant negative association between temperature and SB [59,60,61,62,63]. Katapally et al. reported that that Canadian children were less sedentary in Warm-Wet-Calm weather and more sedentary in Cold-Dry-Windy weather [61]. Only a negative correlation between temperature and sedentary behavior was shown in our study, but it was not statistically significant. A possible explanation for this difference could be that there were differences in the study’s geographic location, region, evaluation method, age, and sample size. Our study shows a significant negative correlation between temperature and individual sleep duration in Chinese freshmen students, which is consistent with previous studies by scholars. Previous research showed that higher temperatures and longer day lengths are associated with reduced nighttime sleep duration in adolescents [34]. A systematic review found that higher temperatures and extreme weather can affect the quality of sleep in individuals [33]. Moreover, a study from Japan showed that individuals’ sleep duration and sleep quality were affected by seasonal changes, which shows that the longest sleep time is in winter and the shortest is actually in summer [64]. In addition, a study of Danish children showed that compared to spring, children spent $2\%$ longer in sleep in winter [65]. Moreover, a report from the Netherlands showed that adults’ sleep time was extended by 31.8 min/day in winter. Furthermore, previous studies reported that the average sleep time of children in winter was 41 min longer than that in summer, 31 min longer in autumn, and 15 min longer in spring, which may be related to seasonal and temperature changes [66]. To our knowledge, this is the first study to use a longitudinal design to investigate the effect of temperature on 24-h movement behaviors among a large cohort sample size. The repeated measurements in one university ensured variation in temperature elements among Chinese freshmen. However, there are a few major limitations of this study. For example, it used a self-report method to assess PA, SB and sleep duration of the respondents. More precise measurements (e.g., pedometer, connected watches, etc.) are needed for better clarity and reproducibility. There was also a lack of investigation on high temperatures (above 29 °C) and low temperatures (<0 °C). In future studies, we suggest using more objective assessment tools to assess 24-h movement behaviors, such as accelerometers. Furthermore, this study is limited by an important methodological issue, namely the issue of uncertain geography [67], given that people move from one place to another on a daily basis. Therefore, meteorological and air conditions in a fixed geographic setting (in this case, a campus) are not representative of the true environmental exposures experienced by individuals on a daily basis and may further lead to biased scientific findings. We did not use spatial interpolation to get more accurate weather conditions in Tsinghua campus. Thus, the results of this study should be treated with caution. Moreover, all participants were recruited from a convenience sample. Freshmen from one university cannot represent all students in China. In addition, new students’ behaviors may not yet have adjusted to the new environment, and their behaviors such as physical activity and sleeping tend to retain more of the habits they have developed over the years from their previous environment, therefore limiting the generalizability of the study’s findings. Future studies are warranted to produce more generalized estimates. ## 5. Conclusions This study found that temperature was significantly positively correlated with physical activity levels in Chinese freshmen students, and was significantly negatively correlated with sleep duration. Replication of this study is warranted among various populations within China. This novel study evidence focused on understanding the relationship between climate change and 24-h movement behaviors among people for developing effective adaptation strategies to climate change to improve people’s health behavior. Knowledge of the impact of temperature on movement behavior is likely to be a variable influencing intraindividual 24-h movement-related behavior, which may help interpretation of their results and translate into improving people’s health behavior. ## References 1. Climate Change. (2021) 2. Banwell N., Rutherford S., Mackey B., Street R., Chu C.. Commonalities between Disaster and Climate Change Risks for Health: A Theoretical Framework. Int. J. Environ. Res. Public Health (2018) 15. DOI: 10.3390/ijerph15030538 3. Orru H., Andersson C., Ebi K.L., Langner J., Astrom C., Forsberg B.. Impact of climate change on ozone-related mortality and morbidity in Europe. Eur. Respir. J. (2013) 41 285-294. DOI: 10.1183/09031936.00210411 4. Hassan N.A., Hashim Z., Hashim J.H.. Impact of Climate Change on Air Quality and Public Health in Urban Areas. Asia Pac. J. Public Health (2016) 28 38S-48S. DOI: 10.1177/1010539515592951 5. Hayes K., Poland B.. Addressing Mental Health in a Changing Climate: Incorporating Mental Health Indicators into Climate Change and Health Vulnerability and Adaptation Assessments. Int. J. Environ. Res. Public Health (2018) 15. DOI: 10.3390/ijerph15091806 6. Hathaway J., Maibach E.W.. Health Implications of Climate Change: A Review of the Literature About the Perception of the Public and Health Professionals. Curr. Environ. Health Rep. (2018) 5 197-204. DOI: 10.1007/s40572-018-0190-3 7. Kohl H.W., Craig C.L., Lambert E.V., Inoue S., Alkandari J.R., Leetongin G., Kahlmeier S., Lancet Physical Activity Series Working G.. The pandemic of physical inactivity: Global action for public health. Lancet (2012) 380 294-305. DOI: 10.1016/S0140-6736(12)60898-8 8. Thivel D., Tremblay A., Genin P.M., Panahi S., Rivière D., Duclos M.. Physical Activity, Inactivity, and Sedentary Behaviors: Definitions and Implications in Occupational Health. Front. Public Health (2018) 6 288. DOI: 10.3389/fpubh.2018.00288 9. Piercy K.L., Troiano R.P.. Physical Activity Guidelines for Americans From the US Department of Health and Human Services. Circ. Cardiovasc. Qual. Outcomes (2018) 11 e005263. DOI: 10.1161/CIRCOUTCOMES.118.005263 10. Shiroma E.J., Lee I.M.. Physical activity and cardiovascular health: Lessons learned from epidemiological studies across age, gender, and race/ethnicity. Circulation (2010) 122 743-752. DOI: 10.1161/CIRCULATIONAHA.109.914721 11. Lee I.M., Shiroma E.J., Lobelo F., Puska P., Blair S.N., Katzmarzyk P.T., Lancet Physical Activity Series Working G.. Effect of physical inactivity on major non-communicable diseases worldwide: An analysis of burden of disease and life expectancy. Lancet (2012) 380 219-229. DOI: 10.1016/S0140-6736(12)61031-9 12. Sáez I., Solabarrieta J., Rubio I.. Reasons for Sports-Based Physical Activity Dropouts in University Students. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18115721 13. Garber C.E., Blissmer B., Deschenes M.R., Franklin B.A., Lamonte M.J., Lee I.-M., Nieman D.C., Swain D.P.. Quantity and Quality of Exercise for Developing and Maintaining Cardiorespiratory, Musculoskeletal, and Neuromotor Fitness in Apparently Healthy Adults: Guidance for Prescribing Exercise. Med. Sci. Sport. Exerc. (2011) 43 1334-1359. DOI: 10.1249/MSS.0b013e318213fefb 14. 14. World Health Organization The Global Action Plan on Physical Activity 2018–2030: More Active People for a Healthier WorldWorld Health OrganizationGeneva, Switzerland2018. The Global Action Plan on Physical Activity 2018–2030: More Active People for a Healthier World (2018) 15. Tremblay M.S., Aubert S., Barnes J.D., Saunders T.J., Carson V., Latimer-Cheung A.E., Chastin S.F.M., Altenburg T.M., Chinapaw M.J.M., Participants S.T.C.P.. Sedentary Behavior Research Network (SBRN)—Terminology Consensus Project process and outcome. Int. J. Behav. Nutr. Phys. Act. (2017) 14 75. DOI: 10.1186/s12966-017-0525-8 16. Owen N., Sparling P.B., Healy G.N., Dunstan D.W., Matthews C.E.. Sedentary behavior: Emerging evidence for a new health risk. Mayo Clin. Proc. (2010) 85 1138-1141. DOI: 10.4065/mcp.2010.0444 17. Hillman C.H., Erickson K.I., Kramer A.F.. Be Smart, Exercise Your Heart: Exercise Effects on Brain and Cognition. Nat. Rev. Neurosci. (2008) 9 58-65. DOI: 10.1038/nrn2298 18. Meo S.A., Alkhalifah J.M., Alshammari N.F., Alnufaie W.S., Algoblan A.F.. Impact of COVID-19 pandemic on sleep quality among medical and general science students: King Saud University Experience. Pak. J. Med. Sci. (2022) 38 639-644. DOI: 10.12669/pjms.38.3.5171 19. Tremblay M.S., Esliger D.W., Tremblay A., Colley R.. Incidental movement, lifestyle-embedded activity and sleep: New frontiers in physical activity assessment. Can. J. Public Health (2007) 980 S208-S217 20. Field C.B., Barros V.R., Mastrandrea M.D., Mach K.J., Abdrabo M.A.-K., Adger W.N., Anokhin Y.A., Anisimov O.A., Douglas J.B., Jonathon B.. IPCC, 2014: Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global and Sectoral Aspects (2014) 21. Zisis E., Hakimi S., Lee E.Y.. Climate change, 24-hour movement behaviors, and health: A mini umbrella review. Glob. Health Res. Policy (2021) 6 15. DOI: 10.1186/s41256-021-00198-z 22. Sallis J.F., Floyd M.F., Rodriguez D.A., Saelens B.E.. Role of built environments in physical activity, obesity, and cardiovascular disease. Circulation (2012) 125 729-737. DOI: 10.1161/CIRCULATIONAHA.110.969022 23. Tucker P., Gilliland J.. The effect of season and weather on physical activity: A systematic review. Public Health (2007) 121 909-922. DOI: 10.1016/j.puhe.2007.04.009 24. Bernard P., Chevance G., Kingsbury C., Baillot A., Romain A.J., Molinier V., Gadais T., Dancause K.N.. Climate Change, Physical Activity and Sport: A Systematic Review. Sport. Med. (2021) 51 1041-1059. DOI: 10.1007/s40279-021-01439-4 25. Saneinejad S., Roorda M.J., Kennedy C.. Modelling the impact of weather conditions on active transportation travel behaviour. Transp. Res. Part D Transp. Environ. (2012) 17 129-137. DOI: 10.1016/j.trd.2011.09.005 26. Garriga A., Sempere-Rubio N., Molina-Prados M.J., Faubel R.. Impact of Seasonality on Physical Activity: A Systematic Review. Int. J. Environ. Res. Public Health (2021) 19. DOI: 10.3390/ijerph19010002 27. Kolle E., Steene-Johannessen J., Andersen L.B., Anderssen S.A.. Seasonal variation in objectively assessed physical activity among children and adolescents in Norway: A cross-sectional study. Int. J. Behav. Nutr. Phys. Act. (2009) 6 9. DOI: 10.1186/1479-5868-6-36 28. Mccormack G.R., Friedenreich C., Shiell A., Giles-Corti B., Doyle-Baker P.K.. Sex- and age-specific seasonal variations in physical activity among adults. J. Epidemiol. Community Health (2009) 64 1010-1016. DOI: 10.1136/jech.2009.092841 29. Turrisi T.B., Bittel K.M., West A.B., Hojjatinia S., Hojjatinia S., Mama S.K., Lagoa C.M., Conroy D.E.. Seasons, weather, and device-measured movement behaviors: A scoping review from 2006 to 2020. Int. J. Behav. Nutr. Phys. Act. (2021) 18 24. DOI: 10.1186/s12966-021-01091-1 30. Merchant A.T., Akhtar-Danesh D.N.. Seasonal Variation in Leisuretime Physical Activity Among Canadians. Can. J. Public Health (2007) 98 203-208. DOI: 10.1007/BF03403713 31. Ho J.Y., Goggins W.B., Mo P.K.H., Chan E.Y.Y.. The effect of temperature on physical activity: An aggregated timeseries analysis of smartphone users in five major Chinese cities. Int. J. Behav. Nutr. Phys. Act. (2022) 19 68. DOI: 10.1186/s12966-022-01285-1 32. Zheng C., Feng J., Huang W., Wong S.H.-S.. Associations between weather conditions and physical activity and sedentary time in children and adolescents: A systematic review and meta-analysis. Health Place (2021) 69 102546. DOI: 10.1016/j.healthplace.2021.102546 33. Rifkin D.I., Long M.W., Perry M.J.. Climate change and sleep: A systematic review of the literature and conceptual framework. Sleep Med. Rev (2018) 42 3-9. DOI: 10.1016/j.smrv.2018.07.007 34. Quante M., Wang R., Weng J., Kaplan E.R., Rueschman M., Taveras E.M., Rifas-Shiman S.L., Gillman M.W., Redline S.. Seasonal and weather variation of sleep and physical activity in 12-14-year-old children. Behav. Sleep Med. (2019) 17 398-410. DOI: 10.1080/15402002.2017.1376206 35. Lanza K., Gohlke J., Wang S., Sheffield P.E., Wilhelmi O.. Climate change and physical activity: Ambient temperature and urban trail use in Texas. Int. J. Biometeorol. (2022) 66 1575-1588. DOI: 10.1007/s00484-022-02302-5 36. Yu H., Yu M., Gordon S.P., Zhang R.. The association between ambient fine particulate air pollution and physical activity: A cohort study of university students living in Beijing. Int. J. Behav. Nutr. Phys. Act. (2017) 14 136. DOI: 10.1186/s12966-017-0592-x 37. Yu H., Cheng J., Gordon S.P., An R., Yu M., Chen X., Yue Q., Qiu J.. Impact of Air Pollution on Sedentary Behavior: A Cohort Study of Freshmen at a University in Beijing, China. Int. J. Environ. Res. Public Health (2018) 15. DOI: 10.3390/ijerph15122811 38. Tsai P.S., Wang S.Y., Wang M.Y., Su C.T., Yang T.T., Huang C.J., Fang S.C.. Psychometric evaluation of the Chinese version of the Pittsburgh Sleep Quality Index (CPSQI) in primary insomnia and control subjects. Qual. Life Res. (2005) 14 1943-1952. DOI: 10.1007/s11136-005-4346-x 39. Buysse D.J., Reynolds C.F., Monk T.H., Berman S.R., Kupfer D.J.. The Pittsburgh Sleep Quality Index—A New Instrument for Psychiatric Practice and Research. Psychiatry Res. (1989) 28 193-213. DOI: 10.1016/0165-1781(89)90047-4 40. Sadeh A.. Sleep Assessment Methods. Monogr. Soc. Res. Child Dev. (2015) 80 33-48. DOI: 10.1111/mono.12143 41. Bassett D.R.. International physical activity questionnaire: 12-country reliability and validity. Med. Sci. Sport. Exerc. (2003) 35 1396. DOI: 10.1249/01.MSS.0000078923.96621.1D 42. Macfarlane D.J., Lee C.C., Ho E.Y., Chan K.L., Chan D.T.. Reliability and validity of the Chinese version of IPAQ (short, last 7 days). J. Sci. Med. Sport (2007) 10 45-51. DOI: 10.1016/j.jsams.2006.05.003 43. Lee P.H., Macfarlane D.J., Lam T.H., Stewart S.M.. Validity of the International Physical Activity Questionnaire Short Form (IPAQ-SF): A systematic review. Int. J. Behav. Nutr. Phys. Act. (2011) 8 115. DOI: 10.1186/1479-5868-8-115 44. Craig C.L., Marshall A.L., Sjostrom M., Bauman A.E., Booth M.L., Ainsworth B.E., Pratt M., Ekelund U., Yngve A., Sallis J.F.. International physical activity questionnaire: 12-country reliability and validity. Med. Sci. Sport. Exerc. (2003) 35 1381-1395. DOI: 10.1249/01.MSS.0000078924.61453.FB 45. Duncan J.S., Hopkins W.G., Schofield G., Duncan E.K.. Effects of weather on pedometer-determined physical activity in children. Med. Sci. Sport. Exerc. (2008) 40 1432-1438. DOI: 10.1249/MSS.0b013e31816e2b28 46. Chan C.B., Ryan D.A.J., Tudor-Locke C.. Relationship between objective measures of physical activity and weather: A longitudinal study. Int. J. Behav. Nutr. Phys. Act. (2006) 3 9. DOI: 10.1186/1479-5868-3-21 47. Belanger M., Gray-Donald K., O’Loughlin J., Paradis G., Hanley J.. Influence of weather conditions and season on physical activity in adolescents. Ann. Epidemiol. (2009) 19 180-186. DOI: 10.1016/j.annepidem.2008.12.008 48. Obradovich N., Fowler J.H.. Climate change may alter human physical activity patterns. Nat. Hum. Behav. (2017) 1 0097. DOI: 10.1038/s41562-017-0097 49. Cramer M.N., Jay O.. Biophysical aspects of human thermoregulation during heat stress. Auton. Neurosci. (2016) 196 3-13. DOI: 10.1016/j.autneu.2016.03.001 50. Heaney A.K., Carrion D., Burkart K., Lesk C., Jack D.. Climate Change and Physical Activity: Estimated Impacts of Ambient Temperatures on Bikeshare Usage in New York City. Environ. Health Perspect. (2019) 127 37002. DOI: 10.1289/EHP4039 51. Al-Mohannadi A.S., Farooq A., Burnett A., Mercia V.D.W., Al-Kuwari M.G.. Impact of Climatic Conditions on Physical Activity: A Two-Year Cohort Study in the Arabian Gulf Region. J. Phys. Act. Health (2016) 13 929-937. DOI: 10.1123/jpah.2015-0593 52. Wang G., Li B., Zhang X., Niu C., Li J., Li L., Speakman J.R.. No seasonal variation in physical activity of Han Chinese living in Beijing. Int. J. Behav. Nutr. Phys. Act. (2017) 14 48. DOI: 10.1186/s12966-017-0503-1 53. Cepeda M., Koolhaas C.M., van Rooij F.J.A., Tiemeier H., Guxens M., Franco O.H., Schoufour J.D.. Seasonality of physical activity, sedentary behavior, and sleep in a middle-aged and elderly population: The Rotterdam study. Maturitas (2018) 110 41-50. DOI: 10.1016/j.maturitas.2018.01.016 54. Lewis L.K., Maher C., Belanger K., Tremblay M., Chaput J.P., Olds T.. At the Mercy of the Gods: Associations Between Weather, Physical Activity, and Sedentary Time in Children. Pediatr. Exerc. Sci. (2016) 28 152-163. DOI: 10.1123/pes.2015-0076 55. Hunter S., Rosu A., Hesketh K.D., Rhodes R.E., Rinaldi C.M., Rodgers W., Spence J.C., Carson V.. Objectively Measured Environmental Correlates of Toddlers’ Physical Activity and Sedentary Behavior. Pediatr. Exerc. Sci. (2019) 31 480-487. DOI: 10.1123/pes.2018-0270 56. Kharlova I., Deng W.H., Mamen J., Mamen A., Fredriksen M.V., Fredriksen P.M.. The Weather Impact on Physical Activity of 6-12 Year Old Children: A Clustered Study of the Health Oriented Pedagogical Project (HOPP). Sports (2020) 8. DOI: 10.3390/sports8010009 57. Zheng C., Huang W.Y., Wong S.H.S.. Associations of weather conditions with adolescents’ daily physical activity, sedentary time, and sleep duration. Appl. Physiol. Nutr. Metab. (2019) 44 1339-1344. DOI: 10.1139/apnm-2019-0309 58. Yildirim M., Schoeni A., Singh A.S., Altenburg T.M., Brug J., De Bourdeaudhuij I., Kovacs E., Bringolf-Isler B., Manios Y., Chinapaw M.J.M.. Daily Variations in Weather and the Relationship With Physical Activity and Sedentary Time in European 10-to 12-Year-Olds: The ENERGY-Project. J. Phys. Act. Health (2014) 11 419-425. DOI: 10.1123/jpah.2012-0102 59. Wu Y.-T., Luben R., Wareham N., Griffin S., Jones A.P.. Weather, day length and physical activity in older adults: Cross-sectional results from the European Prospective Investigation into Cancer and Nutrition (EPIC) Norfolk Cohort. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0177767 60. Arnardottir N.Y., Oskarsdottir N.D., Brychta R.J., Koster A., Van Domelen D.R., Caserotti P., Eiriksdottir G., Sverrisdottir J.E., Johannsson E., Launer L.J.. Comparison of Summer and Winter Objectively Measured Physical Activity and Sedentary Behavior in Older Adults: Age, Gene/Environment Susceptibility Reykjavik Study. Int. J. Environ. Res. Public Health (2017) 14. DOI: 10.3390/ijerph14101268 61. Katapally T.R., Rainham D., Muhajarine N.. The Influence of Weather Variation, Urban Design and Built Environment on Objectively Measured Sedentary Behaviour in Children. AIMS Public Health (2016) 3 663-681. DOI: 10.3934/publichealth.2016.4.663 62. Harrison F., van Sluijs E.M.F., Corder K., Ekelund U., Jones A.. The changing relationship between rainfall and children’s physical activity in spring and summer: A longitudinal study. Int. J. Behav. Nutr. Phys. Act. (2015) 12 9. DOI: 10.1186/s12966-015-0202-8 63. Harrison F., Goodman A., van Sluijs E.M.F., Andersen L.B., Cardon G., Davey R., Janz K.F., Kriemler S., Molloy L., Page A.S.. Weather and children’s physical activity; how and why do relationships vary between countries?. Int. J. Behav. Nutr. Phys. Act. (2017) 14 13. DOI: 10.1186/s12966-017-0526-7 64. Suzuki M., Taniguchi T., Furihata R., Yoshita K., Arai Y., Yoshiike N., Uchiyama M.. Seasonal changes in sleep duration and sleep problems: A prospective study in Japanese community residents. PLoS ONE (2019) 14. DOI: 10.1371/journal.pone.0215345 65. Hjorth M.F., Chaput J.P., Michaelsen K., Astrup A., Tetens I., Sjödin A.. Seasonal variation in objectively measured physical activity, sedentary time, cardio-respiratory fitness and sleep duration among 8–11year-old Danish children: A repeated-measures study. BMC Public Health (2013) 13. DOI: 10.1186/1471-2458-13-808 66. Nixon G.M., Thompson J.M., Han D.Y., Becroft D.M., Clark P.M., Robinson E., Waldie K.E., Wild C.J., Black P.N., Mitchell E.A.. Short sleep duration in middle childhood: Risk factors and consequences. Sleep (2008) 31 71-78. DOI: 10.1093/sleep/31.1.71 67. Kwan M.-P.. The Uncertain Geographic Context Problem. Ann. Assoc. Am. Geogr. (2012) 102 958-968. DOI: 10.1080/00045608.2012.687349
--- title: A Case-Finding Protocol for High Cardiovascular Risk in a Primary Care Dental School—Model with Integrated Care authors: - Amazon Doble - Raul Bescos - Robert Witton - Shabir Shivji - Richard Ayres - Zoë Brookes journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049228 doi: 10.3390/ijerph20064959 license: CC BY 4.0 --- # A Case-Finding Protocol for High Cardiovascular Risk in a Primary Care Dental School—Model with Integrated Care ## Abstract Background: National Health Service (NHS) strategies in the United Kingdom (UK) have highlighted the need to maximise case-finding opportunities by improving coverage in non-traditional settings with the aim of reducing delayed diagnosis of non-communicable diseases. Primary care dental settings may also help to identify patients. Methods: Case-finding appointments took place in a primary care dental school. Measurements of blood pressure, body mass index (BMI), cholesterol, glucose and QRisk were taken along with a social/medical history. Participants with high cardiometabolic risk were referred to their primary care medical general practitioner (GP) and/or to local community health self-referral services, and followed up afterwards to record diagnosis outcome. Results: A total of 182 patients agreed to participate in the study over a 14-month period. Of these, 123 ($67.5\%$) attended their appointment and two participants were excluded for age. High blood pressure (hypertension) was detected in 33 participants, 22 of whom had not been previous diagnosed, and 11 of whom had uncontrolled hypertension. Of the hypertensive individuals with no previous history, four were confirmed by their GP. Regarding cholesterol, 16 participants were referred to their GP for hypercholesterolaemia: 15 for untreated hypercholesterolaemia and one for uncontrolled hypercholesterolaemia. Conclusions: Case-finding for hypertension and identifying cardiovascular risk factors has high acceptability in a primary dental care setting and supported by confirmational diagnoses by the GP. ## 1. Introduction National Health Service (NHS) health screening in the United Kingdom (UK), was introduced in 2009 to reduce cardiovascular disease (CVD) risks and adverse event occurrence [1]. Subsequently, national reviews have highlighted the need to maximise prevention by improving cardiovascular risk case-finding coverage and outputs, while making every patient contact count [2]. Since that time, other organisations such as pharmacies and community health programs have implemented preventative cardiovascular case-finding in non-clinical settings, in order to reduce the morbidity and mortality of CVD and ease pressure on the NHS system [3]. Cardiovascular case-finding has had varying degrees of success within the UK [2,4,5], and this may relate to the variability of measurements taken in non-clinical settings [6]. Early diagnosis and preventative treatment of high blood pressure (hypertension), including identifying undiagnosed patients, could avoid 9710 heart attacks and 14,500 strokes, potentially saving the NHS £274 million per year [7]. Similarly, undiagnosed high cholesterol levels (hypercholesterolemia), are recognised as a contributing factor to coronary artery disease and stroke, with nearly 8 million adults in the UK currently taking lipid-lowering drugs, such as statins, to reduce the risk of developing these complications after being diagnosed [8]. Additionally, undiagnosed and uncontrolled high blood glucose (hyperglycaemia) can lead to many severe complications, including but not limited to kidney, heart, peripheral vascular system and eye damage [9]. Thus, it is vital to manage hyperglycaemia effectively to prevent disease states and improve patient outcomes [10]. However, according to research conducted by the British Heart Foundation, $11.4\%$ of the current UK population is at high risk of a major cardiovascular event (e.g., stroke, heart attack) due to non-controlled hypertension, hypercholesterolemia and/or hyperglycaemia [11]. Hypertension is known as the ‘silent killer’ [12], and this emphasises the importance of case-finding patients as a preventive strategy to reduce the risk of major cardiovascular events (e.g., stroke). Dental practices are good settings in which to identify people with untreated hypertension because a substantial number of people, $36.9\%$ of the UK adult population, have attended an NHS dentist within the last 24 months [13], either for cleaning or treatment. The dental environment provides a unique site for obtaining accurate cardiometabolic health measurements, given the abundance of clinically trained staff and emergency medical supplies, whilst operating under strict patient confidentiality and data protection. They are also equipped for referral to other primary care providers, such as general practitioners (GPs), and other secondary care services, including cancer care. Being in a medical environment instils confidence in patients when receiving results and advice [14]. This is an advantage over some other non-medical environments which have been used previously, such as shopping malls [6]. There is also positive evidence demonstrating higher patient acceptance of participating in cardiovascular case-finding at the dentist [14,15]. A study by Creanor et al. [ 2014] assessed patients’ attitudes towards diabetes and cardiovascular screening within a dental school setting, concluding that $83\%$ of patients attending student dental clinics would be willing to participate in health assessments at the dentist [14]. Similarly, Sproat et al. [ 2009] measured the blood pressure of 114 individuals visiting the dentist in the UK, and found that $39\%$ of them had either a systolic reading greater than 140 mmHg, or a diastolic reading greater than 90 mmHg, or both [15]. This highlights the potential of undertaking health assessments in non-traditional settings in the UK. To the best of our knowledge, a dental school setting has not been previously used to case-find patients with cardiovascular risk factors including high blood pressure, cholesterol, and glucose, with the inclusion of GP referral and robust follow up investigations [16]. This study has thus developed an applied model of cardiometabolic health measurements within the University of Plymouth Dental School, in collaboration with Peninsula Dental Social Enterprise (PDSE) primary care clinics, serving an area of low socioeconomic status [17]. Collaboration with PDSE provides a good opportunity to recruit participants that may not be sufficiently health conscious to self-refer themselves for health screening at the GP [18,19,20]. Indeed, the area of Devonport that PDSE dental school serves is currently ranked the second most deprived ward in Plymouth (2nd out of 20), and is within the <$1\%$ most deprived neighbourhoods in the UK [17]. Thus, this protocol will also be facilitating patient access to care for some of the most at risk groups in the UK for management of cardiovascular risk factors including hypertension, a current NHS priority [21]. The main aims of this study were to use a dental school primary care setting, due to the large volume of patients routinely seen, to (i) establish case-finding clinics to identify people with high cardiometabolic risk factors; and (ii) establish a dental-GP referral network for systemic patient health. ## 2.1. Ethical Aspects This study was approved by the Human Ethics Committee of the University of Plymouth (reference number: 2684). The protocol and design comply with all ethical standards outlined by the responsible committee on human research experimentation. ## 2.2. Data Recording A unique code was given for each patient to maintain confidentiality within the data storage. Restricted and confidential clinical data were recorded and stored long term in the dental software system, on password-protected computers behind restricted swipe card access in PDSE. ## 2.3. Participants and Recruitment Participants were recruited into the study through the dental school onboarding routes (Figure 1), including:Triage sessions, in which people without a current dentist are assessed by dental school staff for their suitability as patients for dental school students. These sessions were set up and supported by clinical staff members of the Peninsula Dental School/PDSE.Student/Staff clinics, which invited participants already undergoing treatment at the dental school to book an appointment for our case-finding clinic. Phone recruitment, which recruited participants who were missed at triage clinics or on the waiting list for triage clinics to optimise participant uptake. This involved sending participants the information sheet via email or letter after discussing the study with them via phone. These pathways are outlined in Figure 1. Participants were booked into a case-finding clinic a minimum of one week later, and the patient’s routine dental care at the dental school continued separate to this appointment. Trial staff provided each participant with an information sheet and answered any questions participants had. All participants within the study were recruited from PDSE clinic facilities in the Plymouth area, including Devonport Dental Education Facility. Patients recruited to the trial were over 40 years of age, which is in agreement with NHS health screening initiatives [1] and given the increased risk of hypertension development in this age group [22]. All patients, irrespective of age and acceptance at dental triage, received routine dental care separate from the case-finding pilot. However, after recruitment, two patients were excluded as they were found to be <40 years of age. No other exclusion criteria were applied for participation in case-finding. However, full social and medical histories were taken to assess previous history of cardiovascular risk factors and social determinants. ## 2.4. Applied Model The applied model for cardiometabolic health measurements within a dental school environment is outlined in Figure 2; however, a further explanation of this can be seen below. Participants were taken to a separate clinic room within the dental school. Measurements were all taken by either dentists, dental nurses or research scientists trained on the same protocols for standardisation and overseen by a clinical lead. A medical, dental, and social history was also taken using a health questionnaire, including any previous diagnosis of hypertension, hypercholesterolaemia, stroke, heart attack, or diabetes, and a social history including their level of education and current occupation. Participants were then asked to lay down on a medical bed in a supine position for 10 min, before systolic (SBP) and diastolic blood pressure (DBP) were recorded in the left arm. Three measurements were taken, with an average of the lowest two values being taken. After this, body height and weight were measured using a stadiometer and mechanical scale, respectively, to calculate the body mass index (BMI). Finally, blood cholesterol and glucose were measured using a portable electronic device (Accutrend Plus, [23]) to assess cardiovascular risk using the Qrisk [24,25]. The Qrisk was used to support any referral for blood pressure, glucose, or cholesterol to their GP. We reviewed patient feedback after the first 100 patients, and broadened the case-finding initiative with the addition of blood glucose measurements. A glucose reading was only taken from participants who had arrived 3 h fasted, as per our information sheet. The case-finding clinic was conducted separately from dental appointments to minimise the impact of any dental anxiety when taking blood pressure readings [26]. ## 2.5. Referral and Follow Up On completion of the case-finding clinic, those participants under high cardiovascular risk, including those with a previous medical history or current medication use for the risk factor (high levels of hypertension, hyperglycaemia or hypercholesterolaemia (outlined in Section 2.6) were provided with a letter including their assessment data, and they were strongly encouraged to visit their GP for diagnostic purposes and management. Patients were also offered community self-referral services, such as Plymouth’s Livewell self-referral services [27], for management of diet, weight loss, anxiety, smoking cessation and alcohol use. Additionally, a letter was sent to the participants’ GPs from the clinic if they were found to have hypertension, hyperglycaemia or hypercholesterolaemia, in accordance with NICE guidelines [7] (Figure 3). Furthermore, a cholesterol-lowering plan leaflet from Heart UK, a cholesterol charity, was provided to participants presenting with hypercholesterolemia [28]. Patients were then followed up at 2.5-week intervals until such time as they had either visited their GP for confirmation of the potential hypertension/hypercholesterolaemia diagnosis, received any lifestyle or medication intervention outside the GP setting, or decided to take no action on our health findings. ## 2.6. Identification and Management of Variables Participants were classified into categories of cardiovascular health as outlined below. According to blood pressure levels, participants were classified hypertension if they presented at the hypertension case-finding clinic with an SBP over 140 mmHg and/or DBP over 90 mmHg, following current NICE hypertension guidelines [29], irrespective of whether they were using anti-hypertensive medication. Participants with an SBP of over 180 mmHg and/or DBP over 120 mmHg were asked to contact their GP on the same day, and if unable to do this, to attend an A&E unit, following NICE hypertension guidelines [29]. Using cholesterol measurements, participants were classified as having healthy cholesterol (<5 mmol/L), high cholesterol (5–6 mmol/L) or extremely high cholesterol (>6 mmol/L), as per NICE guidelines [30], irrespective of whether they are using statin medication. Cholesterol, blood pressure and health questionnaire answers were used to calculate a Qrisk3 for each participant, which provided the patient with information on their risk of developing a heart attack or stroke within the next 10 years [24]. Participants were classified as low risk (Qrisk3 < $10\%$) or high risk (Qrisk > $10\%$) as per NICE guidelines, using the website to input and calculate these scores contemporaneously [25]. With glucose measurements, participants within the range of 4–7 mmol/L were classified as healthy, but values were only considered if patients arrived at the appointment after 3 h of fasting, in order to conform to NHS health check procedures [31,32]. Glucose measurements above 7 mmol/L were classified as hyperglycaemic, as per NICE guidelines [33,34]. However, these measurements were supportive rather than diagnostic at this stage, due to lack of consistency with fasting. For BMI measurements, participants were classified as overweight (BMI between 25 and 29.9) or obese (BMI between 30 and 39.9) [35], with BMI calculated at the time using height and weight measurements input into the NHS BMI calculator [36]. ## 3. Results Participant recruitment and attendance is shown in Figure 4, and the population’s characteristics are displayed in Table 1. Two participants were excluded early in the trial, as the age range for the study was decreased for the reasons outlined in Section 2.3. These two participants, represented in Figure 4, did receive care and a case-finding appointment; however, their data have been excluded from our report. Of the 182 people that booked a hypertension case-finding appointment, 123 ($67.5\%$) attended their appointment; however, only 121 were included in our assessment due to age limitations. Hypertension was detected in thirty-three ($27.3\%$) attendees, and consequently, they were referred to their GP. Twelve participants with high blood pressure not diagnosed previously followed up with their GP; however, only ten undertook further monitoring under the GP’s direction. Four were diagnosed with hypertension. Furthermore, eleven ($9.1\%$) of the participants that were detected with high blood pressure were already using anti-hypertensive medication. Blood cholesterol was measured in fifty-three ($43.8\%$) participants. Sixteen of them ($30.0\%$) were referred to their GP for presenting with hypercholesterolemia, including one ($1.9\%$) participant on statin medication for high cholesterol. Five participants with suspected untreated hypercholesterolemia followed up with their GP; however, only three undertook further monitoring under the GP’s direction. Of those who followed up with their GP and undertook further monitoring, one was successfully diagnosed with hypercholesterolemia without a previous diagnosis. The Qrisk of the twenty-three participants ($43.4\%$) that underwent cholesterol measurements was over $10\%$. Furthermore, the BMI of forty-three participants ($35.5\%$) of this study was >25 (overweight), and forty-two participants ($34.7\%$) had a BMI >30 (obesity). Twenty-five ($20.7\%$) participants were current smokers and twelve ($9.9\%$) were former smokers. Of the 123 participants, only ten ($8.1\%$) underwent glucose case-finding due to equipment delays and technical difficulties. Of these ten, one ($10.0\%$) presented with suspected untreated hyperglycaemia. We are still awaiting communication from this participant regarding whether they followed up with their GP after referral. No participants were identified as having uncontrolled hyperglycaemia. ## 4. Discussion The current study successfully describes a functional and applied protocol within a UK dental school setting for measuring blood pressure, BMI, cholesterol, and glucose, combined with medical history taking. In turn, it has been able to successfully establish a hypertension case-finding clinic for CVD within a primary care dental setting, thus allowing the identification of people with increased cardiometabolic risk factors. The participants in this study would not otherwise have attended for health assessments, as they attended the dental clinics with medical histories perceived as free of cardiovascular disease or with CVD under control with medication. The proportion of patients with untreated hypertension was higher than the national average of $12\%$ [37]. This may be a result of operating within an area of lower socioeconomic status, as rates of hypertension are higher in this population [38]. Additionally, we successfully identified 11 ($9.1\%$) individuals as having uncontrolled hypertension, which is higher than the 2019 NHS national average of $5\%$ [37]. However, after communication regarding GP follow-up and further monitoring, four individuals with suspected untreated hypertension were confirmed with a diagnosis by their GP, representing $3.3\%$ of our overall population. We believe that the confirmed diagnosis would be higher if more than $54.5\%$ of those referred for untreated hypertension had followed up with their GP. We are the first case-finding study in a dental setting that we are aware of that has followed up with the GP in this way. Considering the extra care taken with accuracy in our clinical setting, the reduced incidence of referral demonstrates the importance of integrated care in any case-finding model, and may explain the lack of success previously reported with some hypertension health screening approaches [6,15,39]. Ideally, we would also undertake ambulatory blood pressure monitoring to further improve the accuracy of the measurement with which we refer to the GP, and this will form the basis of future studies. Additionally, a large proportion of our population were obese ($34.7\%$); this figure again was higher than the national average of $28\%$ [40]. This supports our assumption of the lack of health-conscious individuals in our clinic location. However, our protocol identified that $13.2\%$ of our participants had possible untreated and uncontrolled hypercholesterolemia, which is less than the national average of $43\%$ for hypercholesterolemia [37]. However, only $43.8\%$ of our population received a cholesterol reading, due to factors such as technical difficulties or blood/needle phobia on the day. Perhaps this result would be higher and in line with the national average if we had been able to measure blood cholesterol in the whole population of this study. Overall, $38\%$ of our population were referred to their GP due to cardiometabolic variables that may increase the risk of cardiovascular disease; this figure is higher than that quoted by the British Heart Foundation, who state that that $11.4\%$ of the current UK population is at high risk of a major cardiovascular event [11]. Given that $20.7\%$ of our population were current smokers, which is slightly higher than the national average of $16\%$ [37], and that a larger amount of our population were obese, there appears to be a greater need for further lifestyle counselling and education in a range of settings, which should sit alongside smoking cessation programs. To our knowledge, this study is the first to establish a cardiovascular risk case-finding environment within a UK dental school. Overall, $82.9\%$ of people approached for the study took up the offer of an appointment for cardiovascular risk case-finding; our findings remain consistent with the study by Creanor et al. [ 2014] which reported that almost $83\%$ of people attending student dental clinics were willing to participate in health assessments at the dentist [14]. Here, moving from perceived acceptance to actual attendance, $67.5\%$ of participants then attended their booked appointment, supporting known difficulties with attending the dentist in areas of lower socioeconomic status. Additionally, our study identified $27.3\%$ of participants with increased blood pressure, whereas Sproat et al. [ 2009]’s study within a dental practice was able to identify $39\%$. However, their study did utilise participants aged 18 and over, which may be something to consider for further studies [15]. The study by Sproat et al. [ 2009] also advised participants to contact their GP [15], whereas here, we actively referred the participant to their GP and actively confirmed diagnoses, with data to evidence this. These case-finding clinics have established a dental–GP referral network for systemic patient health within an area of social deprivation, carrying out health assessments within a population at higher risk of CVD. It was considered that working interprofessionally with GPs to ensure accurate case-findings was essential to the success of this program. The cardiovascular health assessments performed in the dental setting appeared to be accurate, as participants had their potential diagnoses confirmed by the GP, whereas other studies did not report on such referrals [15,41]. If case-finding is not performed as carefully and accurately as possible at the dentist, there is a risk of burdening of GPs with false-positive results, which this study has attempted to avoid by describing this referral process. This protocol also fostered the dental school’s collaboration with the Office of the Chief Dental Officer in the UK and the CORE20PLUS5 programme to influence national and local policy and improve links between oral and systemic health, including care pathways. As many dental schools are positioned close to areas of socio-economic deprivation [42], in the future, it may be that dental schools will be ideal settings for providing access to oral health services to people in deprived areas in which CVD is more common, reaching some of the most at-risk populations. These data have been fed into an existing hypertension case-finding program within the dental school to work towards the next step of commissioning, which is beyond the scope of this study. However, this study has some limitations which need to be considered. For example, it only included participants over 40 years old, in order to assess those most at risk of CVD; however, younger individuals may be less vulnerable, but nevertheless are also susceptible to cardiometabolic complications if left without assessment [22]. Future research trials would benefit from the inclusion of participants over the age of 18, not just over 40, to encourage individuals to manage their health from a younger age. Furthermore, participants were placed in the supine position when taking the blood pressure; our method, can therefore be viewed as having lower blood pressure readings compared to studies done on the sitting position, which could be something else to consider in future studies [43]. Additionally, this study continues to be run in conjunction with other research studies which have provided sufficient clinical and non-clinical trial staff to be available to run the clinics. If dental schools or primary care dental practices wanted to run these clinics outside of a trial environment, they would be required to find funding, equipment and human resources. This study began to test for fasting glucose levels; indeed, others have performed similar studies, measuring blood glucose at the dentist [44]. It was interesting that here, we found a patient with uncontrolled diabetes despite the small population, and that patients wanted this additional measurement to be performed. Our future studies will continue glucose testing and adapt the method further, with, for example, HbA1c monitoring and 3-month average glucose levels, to try and avoid false-positive and false-negative glucose results, similar to our aim with blood pressure [31,45]. Additionally, despite all precautions being undertaken within our protocol to remove white-coat syndrome, when taking blood pressure readings, there is the possibility that dental anxiety may still be present at opportunistic blood pressure assessments within a dental healthcare setting [26]. Furthermore, 24 h or 1-week ambulatory monitors could be given as part of the trial to reduce false positives and aid in confirming diagnoses. Lastly, given the pressure on NHS GPs within the UK currently [46], with our network referring at-risk participants to their GP, we must determine whether GP surgeries have the capacity needed to deliver appointments to these referred individuals. ## 5. Conclusions In conclusion, this study is the first to successfully establish a case-finding protocol for people with high cardiovascular risk within a UK dental school. This intervention can help to reduce the risk of major cardiovascular events and their cost to the NHS. ## References 1. Robson J., Dostal I., Sheikh A., Eldridge S., Madurasinghe V., Griffiths C., Coupland C., Hippisley-Cox J.. The NHS Health Check in England: An evaluation of the first 4 years. BMJ Open (2015.0) 6 e008840. DOI: 10.1136/bmjopen-2015-008840 2. Atkins L., Stefanidou C., Chadborn T., Thompson K., Michie S., Lorencatto F.. Influences on NHS Health Check behaviours: A systematic review. BMC Public Health (2020.0) 17. DOI: 10.1186/s12889-020-09365-2 3. Missed GP Appointments Costing NHS Millions; NHS England. (2019.0) 4. Fisher R.F., Croxson C.H., Ashdown H.F., Hobbs F.R.. GP views on strategies to cope with increasing workload: A qualitative interview study. Br. J. Gen. Pract. (2017.0) 67 e148-e156. DOI: 10.3399/bjgp17X688861 5. Kypridemos C., Allen K., Hickey G.L., Guzman-Castillo M., Bandosz P., Buchan I., Capewell S., O’Flaherty M.. Cardiovascular screening to reduce the burden from cardiovascular disease: Microsimulation study to quantify policy options. BMJ Open (2016.0) 353 i2793. DOI: 10.1136/bmj.i2793 6. Edwards L.A., Campbell P., Taylor D.J., Shah R., Edgar D.F., Crabb D.P.. Healthy shopper? Blood pressure testing in a shopping centre Pop-Up in England. BMC Public Health (2019.0) 19. DOI: 10.1186/s12889-018-6370-0 7. CVD Prevention: Detecting and Treating Hypertension 2022 8. 8. British Heart Foundation UK FactsheetBritish Heart Foundation2022Available online: https://www.bhf.org.uk/-/media/files/for-professionals/research/heart-statistics/bhf-cvd-statistics-uk-factsheet.pdf?rev=e771367bf0654a4dae85cbc9dbefae17&hash=76C0182379BB6EE118EC6F76FA35A158(accessed on 2 February 2023). UK Factsheet (2022.0) 9. Deshpande A., Harris-Hayes M., Schootman M.. Epidemiology of Diabetes and Diabetes-Related Complications. Phys. Ther. (2008.0) 88 1254-1264. DOI: 10.2522/ptj.20080020 10. Ohishi M.. Hypertension with diabetes mellitus: Physiology and pathology. Hypertens. Res. (2018.0) 41 389-393. DOI: 10.1038/s41440-018-0034-4 11. 11. British Heart Foundation CVD Statistics BHF 2022, Chapter 2 MorbidityBritish Heart Foundation2022Available online: https://www.bhf.org.uk/what-we-do/our-research/heart-statistics/heart-statistics-publications/cardiovascular-disease-statistics-2022(accessed on 2 February 2023). CVD Statistics BHF 2022, Chapter 2 Morbidity (2022.0) 12. Fatima S., Mahmood S.. Combatting a silent killer—The importance of self-screening of blood pressure from an early age. EXCLI J. Exp. Clin. Sci. (2021.0) 20 1326-1327 13. 13. NHS NHS Dental Statistics for England, 2021–2022, Annual Report. NHS Dental Statistics for England—2021–22: Annex 1 (Tables and Charts)—Table 4aNHS Dental Statistics2022Available online: https://digital.nhs.uk/data-and-information/publications/statistical/nhs-dental-statistics/2021-22-annual-report(accessed on 2 February 2023). NHS Dental Statistics for England, 2021–2022, Annual Report. NHS Dental Statistics for England—2021–22: Annex 1 (Tables and Charts)—Table 4a (2022.0) 14. Creanor S., Millward B.A., Demaine A., Price L., Smith W., Brown N., Creanor S.L.. Patients attitudes towards screening for diabeties and other medical conditions in the dental setting. Br. Dent. J. (2014.0) 216 E2. DOI: 10.1038/sj.bdj.2013.1247 15. Sproat C., Beheshti S., Harwood A.N., Crossbie D.. Should we screen for hypertension in general dental practice?. BDJ (2009.0) 207 275-277. DOI: 10.1038/sj.bdj.2009.815 16. Greenberg B.L., Glick M., Goodchild J., Duda P.W., Conte N.R., Conte M.. Screening for cardiovascular risk factors in a dental setting. J. Am. Dent. Assoc. (2007.0) 138 798-804. DOI: 10.14219/jada.archive.2007.0268 17. 17. Plymouth-City-Council Index of Multiple Deprivation (IMD) 2019—Plymouth Summary Analsysis Plymouth City CouncilPlymouth-City-Council2019Available online: https://www.plymouth.gov.uk/sites/default/files/IMD%202019%20report%20Final%200.1.pdf(accessed on 2 February 2023). Index of Multiple Deprivation (IMD) 2019—Plymouth Summary Analsysis Plymouth City Council (2019.0) 18. Kawabata Y., Ekuni D., Miyai H., Kataoka K., Yamane M., Mizutani S., Irie K., Azuma T., Tomofuji T., Iwasaki Y.. Relationship Between Prehypertension/Hypertension and Periodontal Disease: A Prospective Cohort Study. Am. J. Hypertens. (2016.0) 29 388-396. DOI: 10.1093/ajh/hpv117 19. Martin-Cabezas R., Seelam N., Petit C., Agossa K., Gaertner S., Tenenbaum H., Davideau J.L., Huck O.. Association between periodontitis and arterial hypertension: A systematic review and meta-analysis. Am. Heart J. (2016.0) 180 98-112. DOI: 10.1016/j.ahj.2016.07.018 20. Area Profile: Devonport Ward: Office of the Director of Public Health, Plymouth City Council. (2020.0) 21. The NHS Long Term Plan: NHS. (2019.0) 22. Worldwide trends in hypertension prevalence and progress in treatment and control from 1990 to 2019: A pooled analysis of 1201 population-representative studies with 104 million participants. Lancet (2021.0) 398 957-980. DOI: 10.1016/S0140-6736(21)01330-1 23. 23. Roche Diagnostics Accutrend Plus Users Manual2012Available online: https://beta-static.fishersci.com/content/dam/fishersci/en_US/documents/programs/healthcare/technical-documents/user-manuals/roche-accutrend-plus-users-manual.pdf(accessed on 2 February 2023). Accutrend Plus Users Manual (2012.0) 24. 24. Qrisk Qrisk3-2018ClinRisk Ltd.2018Available online: https://qrisk.org/(accessed on 2 February 2023). Qrisk3-2018 (2018.0) 25. 25. NICE Scenario: Management of People with an Estimated Risk of 10% or MoreNICE2023Available online: https://cks.nice.org.uk/topics/cvd-risk-assessment-management/management/cvd-risk-10percent-or-more/(accessed on 2 February 2023). Scenario: Management of People with an Estimated Risk of 10% or More (2023.0) 26. Andersson H., Hedstrom L., Berghm H.. White-coat hypertension detected during opportunistic blood pressure screening in a dental healthcare setting. Scand. J. Prim. Health Care (2021.0) 39 348-354. DOI: 10.1080/02813432.2021.1958496 27. Specialists in Community Health and Care: Livewell. (2022.0) 28. 28. Charity HUTC Ultimate Cholesterol Loweing PlanHeart UK The Cholesterol Charity2019Available online: https://www.heartuk.org.uk/downloads/uclp-consumer-booklet.pdf(accessed on 2 February 2023). Ultimate Cholesterol Loweing Plan (2019.0) 29. Hypertension in Adults: Diagnosis and Management NICE: NICE. (2022.0) 30. Understanding Your Cholesterol Test Results 2022 31. High Blood Sugar (Hyperglycaemia): NHS. (2022.0) 32. Can I Eat and Drink before Having a Blood Test? NHS. (2021.0) 33. 33. NICE Type 1 Diabetes in Adults: Diagnosis and managementNICE2015Available online: https://www.nice.org.uk/guidance/ng17/resources/type-1-diabetes-in-adults-diagnosis-and-management-1837276469701(accessed on 2 February 2023). Type 1 Diabetes in Adults: Diagnosis and management (2015.0) 34. Type 2 Diabetes in Adults: Management: NICE. (2022.0) 35. What Is the Body Mass Index (BMI)?. (2019.0) 36. 36. NHS NHS BMI CalculatorNHS BMI2022Available online: https://www.nhs.uk/live-well/healthy-weight/bmi-calculator/(accessed on 2 February 2023). NHS BMI Calculator (2022.0) 37. 37. NHS Health Survey for England, 2019: Data Tables—Health Survey for England, 2019: Adults Health Data TablesNHS DigitalLeeds, UK2020. Health Survey for England, 2019: Data Tables—Health Survey for England, 2019: Adults Health Data Tables (2020.0) 38. Blok S., Haggenburg S., Collard D., Van Der Linden E.L., Galenkamp H., Moll van Charante E.P., Agyemang C., Van Den Born B.H.. The association between socioeconomic status and prevalence, awareness, treatment and control of hypertension in different ethnic groups: The Healthy Life in an Urban Setting study. J. Hypertens. (2022.0) 40 897-907. DOI: 10.1097/HJH.0000000000003092 39. Kellogg S.D., Gobetti J.P.. Hypertension in a Dental School Patient Population. J. Dent. Educ. (2004.0) 68 956-964. DOI: 10.1002/j.0022-0337.2004.68.9.tb03844.x 40. 40. NHS Health Survey for England, 2019: Data Tables—Health Survey for England, 2019: Overweight and Obesity in Adults and Children Data TablesNHS Digital2020Available online: https://digital.nhs.uk/data-and-information/publications/statistical/health-survey-for-england/2019(accessed on 2 February 2023). Health Survey for England, 2019: Data Tables—Health Survey for England, 2019: Overweight and Obesity in Adults and Children Data Tables (2020.0) 41. Al-Zahrami M.. Prehypertension and undiagnosed hypertension in a sample of dental school female patients. Int. J. Dent. Hyg. (2009.0) 9 74-78. DOI: 10.1111/j.1601-5037.2009.00441.x 42. Ministry of Housing, UK. The English Indices of Deprivation 2019 (IoD2019). (2019.0) 43. Eşer I., Khorshid L., Güneş U.Y., Demir Y.. The effect of different body positions on blood pressure. J. Clin. Nurs. (2007.0) 16 1137-1140. DOI: 10.1111/j.1365-2702.2005.01494.x 44. Abdul Aziz A.F., Mohd-Dom T.N., Mustafa N., Said A.H., Ayob R., Mohamed Isa S., Hatah E., Wan Puteh S.E., Mohd Alwi M.F.F.. Screening for type 2 diabetes and periodontitis patients(CODAPT-My@). A multidiciplinary care approach. BMC Health Serv. Res. (2022.0) 22. DOI: 10.1186/s12913-022-08429-w 45. A1c Test MayoClinic Test Procedures: MayoClinic. (2022.0) 46. 46. British Medical Association Pressures in General Practice Data AnalysisBMA2023Available online: https://www.bma.org.uk/advice-and-support/nhs-delivery-and-workforce/pressures/pressures-in-general-practice-data-analysis(accessed on 2 February 2023). Pressures in General Practice Data Analysis (2023.0)
--- title: 'Practices of Care and Relationship-Building: A Qualitative Analysis of Urban Agriculture’s Impacts on Black People’s Agency and Wellbeing in Philadelphia' authors: - Ashley B. Gripper journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049229 doi: 10.3390/ijerph20064831 license: CC BY 4.0 --- # Practices of Care and Relationship-Building: A Qualitative Analysis of Urban Agriculture’s Impacts on Black People’s Agency and Wellbeing in Philadelphia ## Abstract Gardens and farms provide individuals and communities with access to affordable, nutritious, and culturally significant foods. There is a rich body of literature unpacking the connections between Black urban growing and agency, freedom, resistance, and care. However, spirituality remains one aspect of health and wellbeing that has not been studied extensively in relation to agriculture. The main goal of this study was to conduct focus groups with Philly-based growers to understand the self-determined impacts of urban agriculture on health, agency, and wellbeing. The secondary goal of this work was to determine if these impacts differ by race. I apply a collective agency and community resilience theoretical framework to this study. This framework offers a model to understand agriculture as a way for communities to become self-determined, self-reliant, and self-sustained. For this study exploring the impacts of urban agriculture on health, there were three eligibility criteria. Participants had to be at least 18 years old, identify as Black or White, and have grown food in a garden or farm in Philadelphia. I hosted six race-specific focus groups at Bartram’s Garden in Southwest Philadelphia. The audio recordings were transcribed, and the full transcripts were coded using open and axial coding methods and a “key concepts” framework. We also employed several methods of triangulation to help ensure the credibility and validity of the findings. Four major themes emerged from the data: growing as a demonstration of agency and power, growing as a facilitator of body–mind wellness, community care and relationship-building, and deepened spiritual connection and interdependence. There were both similarities and differences in the impacts of urban agriculture by race. Across the six focus groups, people talked about concepts related to community care and relationship-building as being major benefits of growing food. In both groups, people also brought up significant issues and barriers around land security. Mentions of spirituality appeared more frequently and more emphatically in the Black focus groups. Black focus groups were more likely to discuss the collective impacts of agriculture, while White participants were more likely to discuss the impacts on themselves as individuals. The findings of this focus group study point to some key domains through which agriculture impacts the health of farmers and growers in Philadelphia. ## 1. Introduction Inadequate access to nutrient-rich, affordable, and culturally significant foods is a form of slow violence [1]. Rob Nixon defines slow violence as “a violence that occurs gradually and out of sight, a violence of delayed destruction that is dispersed across time and space, an attritional violence that is typically not viewed as violence at all” [1]. Access to fresh, chemical-free produce is often harder to find in low-income Black neighborhoods than more affluent or White neighborhoods [2,3,4,5,6,7]. When Black communities have access to supermarkets or grocery stores, they generally offer lower quality produce or carry produce that is not affordable to Black communities with lower incomes. These barriers to nutrient-rich, nourishing food can be described as food apartheid. Food apartheid is different from a food desert. Deserts are naturally occurring landscapes that can be rich with life and food. People who are unfamiliar with deserts often do not know where to look for that food. Food apartheid, on the other hand, highlights the structural factors and actors that have contributed to an intentionally inequitable and inadequate food system for Black Americans and other groups. Food apartheid, food insecurity, and the extraction of resources from Black communities has led to disturbingly high rates of disease and mortality among Black Americans [2,6,8,9,10,11,12,13,14]. In Public Health research, Black people and Black communities are often characterized through a “damaged-centered” research lens and by the disparities and inequities they experience in comparison to White people [15]. While studying health inequities offers incredible value to naming the problems that people face (which is necessary to create helpful interventions), deficit framing should not be the only analysis used to characterize Black people and Black communities. In response to food apartheid, Black people across the country have engaged in acts of resistance to heal and care for themselves, especially when public health interventions have been ineffective at meeting individual and community needs [16,17,18,19,20,21,22]. Over the past several decades, the presence of urban agriculture has rapidly increased in cities across the United States [23,24]. Due to the fact of food shortages during the COVID-19 pandemic, many people are realizing the benefits of urban food-growing for themselves [25,26,27]. In the early days of the pandemic, people flocked to grocery stores to ensure they had enough supplies to get through lockdowns. This led to empty food shelves and limited nutritious options, especially for people in low-income areas. Particularly for Black Americans, the COVID-19 pandemic restrictions and protocols have allowed space and time for many people to participate in gardens, farms, and urban growing practices. Gardens and farms provide individuals and communities with access to affordable, nutritious, and culturally significant foods. They offer exposure to greenness, opportunities for physical activity, and potential benefits to the microbiome [28,29]. These spaces often facilitate spiritual connections to land. They bolster climate change mitigation and ecosystem services through stormwater retention and cooling effects for urban heat islands [30,31,32,33]. Community growing spaces facilitate information exchange and intergenerational knowledge sharing [17]. They are a means of strengthening social capital and support [18,34] and offer community members spaces to gather and celebrate without fear of crime or state-sanctioned violence. Agricultural spaces led by people of the global majority (PGM) [35] demonstrate the capacity of a community to engage in shared decision-making and community-based solutions to health inequities [17,18,19,36,37]. PGM refers to people of the global majority. This includes Black, Indigenous, and Asian people and other people of color. Originators say, “this wording points out the demographic inaccuracy of the euphemism ‘minority’ and can feel more empowering for some people”. Dyg et al. [ 2020] performed a thematic review of the literature on community gardens and health [38]. While community gardens can be a form of urban agriculture, not all types of urban agriculture are community gardens. This review identified a total 51 studies that investigated the relationship between community gardens and health. The studies were quantitative, qualitative, and employed mixed methods. Common designs included case studies, cross-sections, and interventions. Overall, this review found that community gardens are consistently positively associated with relationship-building and social connection. The associations between community gardens and other forms of health were mixed, although many studies showed positive impacts on physical health, food knowledge, and communities/neighborhoods [38]. Other urban agriculture and health literature has demonstrated positive associations between agriculture and mental health [34,39], specifically depression and anxiety [40,41]. However, there is an increasing body of literature that provides evidence that the leadership of food-growing space also matters [19,42,43]. There may be a difference in the impacts of urban, community food-growing depending on if a project is community-led or led by people from outside the community [42,43,44]. For instance, although White-led urban agriculture projects situated in communities of color tend to receive more funding than PGM-led projects [35,45], their programming has weaker impacts on the communities they serve [46]. In a literature review on the impacts of urban agriculture, Dr. Sheila Golden found that the “culture around local and healthy food has often been associated with those that have higher-educations and incomes,” and that this type of urban agriculture can be exclusionary to low-income Black communities, who are often the target of urban agriculture projects [42,43,44]. In addition to increasing access to nutritious foods, PGM-led spaces offer additional community benefits. These spaces foster community connection, agency, and resistance. For example, through two qualitative studies on farmers in Detroit, Dr. Monica White demonstrated how Black-led community food-growing projects do more than provide fresh fruits and vegetables [17,18]. They help foster community self-determination, self-reliance, and activism [47,48]. Other qualitative literature has also shown that urban, community food-growing positively impacts the physical health, community development, and social capital for those who engage in it [34,43,49]. There is also a rich body of literature unpacking the connections between Black urban growing and agency, freedom, resistance, and care. Scholars in anthropology and sociology have used case studies to show that agriculture fosters power and collective agency within Black communities [17,18,20,21,37]. A major way that Black communities, particularly agricultural communities, organize is through creating access to resources that have historically and socially been denied to them [36]. Spirituality is one aspect of health and wellbeing that has not been studied extensively in relation to agriculture. Afro-indigenous people have kept and expanded agricultural knowledge and techniques for tens of thousands of years. European colonizers kidnapped Africans from their continent and enslaved them in a foreign land to capitalize off their agricultural knowledge and wisdom [50]. The atrocities committed—genocide of people Indigenous to the Americas and the brutal enslavement of people Indigenous to Africa—were conducted so on this land and for the theft and commodification of this land. In his book Something Torn and New, Ngugi wa Thiong’o talks about the dismembering of the African continent and people [51]. He says, “The requirements of the slave plantation demanded the physical removal of human resources from the continent to work on land stolen from other peoples, mainly native Caribbeans and Native Americans. The result was an additional dismemberment of the diasporic African, who was now separated not only from his continent and his labor but also from his very sovereign being”. Dismembering serves a specific purpose. It seeks to separate people from their spiritual practices, from their foodways, and from their relatives both human and more-than-human [52]. The goal of dismembering is humiliation, and it is also to prevent re-membering [51]. When Black people forget who we are, where we come from, and our ways of knowing and loving and living, then it is easier to control what we eat and control us. I use “we” because I am a Black woman working to help other Black people reconnect and rebuild our relationships to our ancestral foodways, spiritual practices, land practices, and culture. Settler-colonizers attempted to dismember Africans from their cultures, rituals, spiritual practices, languages, and foodways while extracting their ancestral farming knowledge. Unbeknownst to them, abducted African women often braided seeds from home into their hair, carrying with them their stories and power across oceans. Though in many ways the goal was dehumanization, dismemberment, and disconnection, Africans preserved and maintained their spiritual practices and ways of relating to each other and to Earth. In “Growing Black food on sacred land: Using Black liberation theology to imagine an alternative Black agrarian future,” Dr. Priscilla McCutcheon uses Black liberation theology to understand the presence and role of “Spirit” in Black farming communities [53]. In her work, McCutcheon argues that spirituality should be an important factor of consideration for researchers studying Black farming and agrarian communities. To date, there has been little scholarship exploring the relationship between spirituality and agriculture [17,53,54]. There are many definitions for “urban agriculture”. I define urban agriculture as the process of growing food outdoors in a city or urban area. My work and this study explore the impacts of land-based urban agriculture on health, spirituality, and communities. I consider urban farmers and gardeners to be anyone who grows food on a farm or in a garden, and in a city, respectively. While the USDA has stricter guidelines concerning what qualifies someone to be a farmer [55], I choose to adopt an approach that includes and values the labor of all people who tend to and grow food on land of considerable size. This definition of urban agriculture and urban farmers is consistent with what others have outlined in sociological literature [36]. ## 1.1. Theoretical Framework: Collective Agency and Community Resilience (CACR) Collective agency and community resilience are widely practiced within Black agricultural communities [18,20,21,56]. In her book entitled Freedom Farmers: Agricultural Resistance and the Black Freedom Movement, Dr. Monica White describes the theoretical framework of collective agency and community resilience (CACR) as capturing “the activities communities enact as a means to be self-reliant and self-sufficient ” [36]. This framework offers three primary strategies through which agriculture helps Black communities realize their collective agency and resilience. The strategies are commons as praxis, prefigurative politics, and economic autonomy. Commons as praxis refers to the resource commons and includes collective decision making, collective action, and fair allocation of shared resources. Prefigurative politics involves a process of creating and building alternative democratic political systems, particularly when the dominant political system excludes Black people, Black ideas, and Black interests. Economic autonomy refers to “efforts to create an alternative system of resource exchange within the community, and these funds and resources have direct benefits for all of its members” [36]. Fostering economic autonomy can include the development of economic systems alternative to capitalism, such as socialism, or more cooperative economic practices. Bartering and labor exchange practices can also help increase economic autonomy for Black farmers and growers. As a framework, collective agency and community resilience highlights and unpacks the role of agriculture in helping Black communities become self-determined, self-sustained, and self-reliant. I use this framework as a guide to understanding how participation in urban agriculture primarily impacts Black growers in Philly and their communities. This study also looks at the impacts on White urban growers primarily as a comparator. ## 1.2. Collective Agency White describes collective agency as involving “social actors’ ability to create and enact behavioral options necessary to affect their political future” [36]. She goes on to argue “a community does not have consciousness in the same way an individual does, but when a group of people comes together and believes in their mutual success, this creates a separate type of consciousness that drives collective agency” [36]. Collective agency, as I apply it in this study, is not limited to the behaviors that affect political futures but also includes behaviors that affect holistic community healing and wellbeing. It refers to the ability of a community to work together to understand, realize, and act on a course of action to support their political power, wellbeing, and collective self-determination. ## 1.3. Constructive Resistance White’s theoretical framework on collective agency and community resilience also highlights the ways farmers engage in constructive resistance. She compares constructive resistance to Benedict Kerkvliet and James Scott’s concept of “everyday strategies of resistance” [57]. Everyday strategies of resistance describe how individuals engage in small, disruptive actions to highlight their opposition to injustice, while White’s theory focuses on the constructive actions that Black farming communities take to resist oppressive social, political, and economic conditions [36]. Often, these constructive forms of resistance result in improved psychological, physical, spiritual, and social health for communities oppressed by settler governments and regimes. The primary aim of this study was to conduct an exploratory focus group study with Philadelphia-based growers. The goal was to understand the self-determined impacts of urban agriculture on health, agency, and spiritual wellbeing. The secondary goal of this work was to determine if these impacts differed by race. Using a mixed-methods approach, the findings from this body of work will be used to inform the development of a scale that assesses how spirituality, wellbeing, agency, and power manifest in urban agriculture communities. ## 2.1. Research Design Overview Being deeply familiar with the study population, I recognized that richer data would come from having growers in the same room listening to, learning from, and in conversation with other growers. I have worked closely with the study host site, Bartram’s Garden, as well as Philadelphia-area urban agriculture organizations for several years helping to design, host, and implement community programs and events. These carefully cultivated relationships allowed me access to the stories and experiences of communities of farmers and growers in Philadelphia. I chose a double-layer design because there are two categories of experiences that this research aims to capture [58]: Black urban growers and White urban growers. Black growers are the primary population of interest and make up one of the largest demographics of growers in Philadelphia. Most urban growers in Philadelphia are Black, White, Latino/a/x, or Southeast Asian. While this study focuses only on the experiences of Black and White growers, future work will include the experiences of Black, Indigenous, and other growers of color in Philadelphia and other cities. To allow for greater expression of thought and experiences, focus groups were organized by racial group. I chose focus group methodology to allow for the natural conversation that would emerge between urban growers. Focus groups are generally the preferred method when the goal of research is to explore people’s thoughts, beliefs, and attitudes toward a particular issue or subject [58]. This method is useful when one wants to understand how perspectives may differ between different groups of people or across geographic regions. Focus groups are also advantageous when “the researchers need information to design a large-scale quantitative study” [58]. The information and insights that were gained through this focus group study were directly translated into a scale for related research. Data often emerge “through interaction within the group” [58,59]. Particularly among African Americans, oral and linguistic traditions, such as call-and-response and “co-signing”, are a big part of the culture [60]. Call-and-response, traditionally practiced in African American church, music, and theater, involves audience participation—responses—to the main actor’s prompt—call. Within these focus groups, I wanted to capture the ways that Black participants might engage in call-and-response by responding to each other’s prompts and co-signing each other’s statements. Considering that one of my goals is to understand collective and community processes of agency and resistance, then it is appropriate to use a conversational method, such as focus groups, which will allow those characteristics to emerge. ## 2.2. Recruitment Process For this study, there were three screening/eligibility criteria. Participants had to be at least 18 years old, identify as Black or White, and have grown food in a garden or farm in Philadelphia. For each category of participant, it is an acceptable practice to have at least three focus groups so that we reach the point of saturation in our data collection—that is, are we still learning new information from these groups, or, are the same key concepts and themes being repeated [58,61]? Since there were two categories of participants, we hosted six focus groups with five to eight participants in each group. Groups with fewer than five people have the potential to lack variation, while groups with more than eight people may be difficult to manage. Five to eight people was the compromise between variation, saturation, and feasibility. My target goal for enrollment was 30–48 people. ## 2.3. Participant Selection Participants were identified through snowball sampling. I leveraged key community partners and influencers in Philadelphia’s agriculture communities to help with recruitment (see table in appendix). Participants were also recruited through several social networks and urban farmer listservs. Since urban farmers and gardeners are a niche group in Philadelphia, the most appropriate recruitment strategy was to rely heavily on the networks of grassroots and nonprofit community partners. Recruitment goals were met two weeks after outreach began. All eligible participants were enrolled in the study ($$n = 42$$), but some had to withdrawal prior to the focus groups due to schedule conflicts and inclement weather ($$n = 10$$). Across six focus groups, we engaged 32 people. Twenty participants racially identified as Black, and twelve participants racially identified as White. All participants completed a screening questionnaire prior to enrollment, COVID screening, and demographic questionnaire. ## 2.4. Data Collection All data were collected through focus group interviews and questionnaires. I obtained demographic information through electronic and paper surveys. The assistant moderator, and I conducted six focus groups from August 2021 to September 2021 at Bartram’s Garden in Southwest Philadelphia. They were split evenly by race, so there were three White and three Black focus groups. The groups were staggered between evenings and weekends to accommodate a variety of work schedules. Each focus group ranged from 80 to 110 min, with an average interview time of 90 min. After brief introductions and an overview of the study, we asked permission from the participants before beginning the audio recording. Throughout the focus groups, the participants were asked six primary questions: [1] Why do you farm or grow food? What does it mean to you? [ 2] Do you think being a farmer or grower impacts your health? If so, how? [ 3] Do you think urban agriculture impacts your community? How? [ 4] What are the benefits and/or barriers to farming or growing? [ 5] What strategies have you used to respond to these barriers? [ 6] Is there anything else you would like to share about your experiences with urban agriculture? I asked probing questions to gain further clarity on some of the concepts, ideas, and experiences shared. Participants were encouraged to talk directly to others and respond to or build off statements made by other members of the focus group. During focus groups, I took notes on key themes and ideas that emerged, as well as my own thoughts and responses regarding the content of group conversations. The assistant moderator was responsible for managing the audio equipment and taking notes on body language throughout the focus groups. At the conclusion of each focus group, I asked participants to provide feedback on the moderation and flow of the focus group. If comfortable, they were given the opportunity to share verbally in a communal setting. The participants were also given electronic or paper follow-up surveys to gauge their feedback on the focus groups. Each participant who completed a focus group was compensated at a competitive rate of USD 100 as a symbol of appreciation and respect for the time and energy they had given to this research. In the first three interviews, I asked participants to write down important areas of consideration for further urban agriculture research at the end of the focus group. Realizing that it made more sense to have people write down their ideas in the beginning of the focus group, I shifted to asking participants to write down topic areas before the focus group dialogue began. In the early focus groups, I asked fewer probing questions and was much more “to the point”. As I became more comfortable and rejected the mythological notions of researcher objectivity and professionalism, I leaned into being more of my authentic self. This led to asking more probing, sometimes challenging, questions of participants. The conversation seemed much richer as a result. The response from participants was also affirming of this approach. ## 2.5. Analysis To understand the impacts of urban agriculture on the health of growers, this study used several methods of analysis. We conducted a transcription-based analysis, which means that entire focus group audio recordings were transcribed. I employed a “key concepts” framework. This framework was a good fit for this study, because it focuses on discovering core ideas and understanding how participants conceptualize and relate to a topic. A key task of this framework is to “identify a limited number of important ideas, experiences, or preferences that illuminate the study” [58]. The analyses of the transcripts were computer-based and performed in both NVivo and Taguette, while still employing many elements of the classic approach. I compiled transcripts by category of participant. One section included all of the transcripts from the focus groups with Black growers in Philadelphia and another section included all to the transcripts from White growers in Philadelphia. I compiled a list of preliminary codes (~34) based on the notes from the focus groups, memory, CACR theoretical framework, and my lived experience as a Black urban grower. Using the preliminary codes, the assistant moderator and I independently performed a combination of open and axial coding. The final codes represented a combination of preliminary codes and codes that emerged during the transcript-based coding process. After we completed the initial coding, we met to compare codes, highlight differences, and arrive at a consensus regarding the findings. Once the coding was complete, I organized the codes into major themes. I also employed several methods of triangulation to help ensure the credibility and validity of the findings. In addition to multiple coders, I coded and recoded transcripts multiple times, conducted focus group debriefs, engaged in continuous processes of self-reflection, performed “member checks,” and carefully reviewed the follow-up survey responses [59,62,63,64]. Overall, the analysis process was continuous and reflexive. The insights and findings for the impacts of agriculture on health are described in the next section. ## 3. Findings The findings show that urban growers in Philadelphia have a range of experiences when it comes to agriculture’s impacts on health and wellbeing. These impacts are both similar and different by race. Using an interpretivist approach—accepting that there is not only one but multiple ways and perceptions of how people’s health can be shaped by agriculture –we identified 89 total codes that I then organized into seven themes. The findings section will concentrate on the four most prominent themes: growing as a demonstration of agency and power, growing as a facilitator of body–mind wellness, community care and relationship-building, and deepened spiritual connection and interdependence. Some of these themes were more pronounced based on the racial makeup of a focus group. The other themes that emerged, to a much lesser degree, are environmental health knowledge, perceptions of self, and barriers and strategies to growing food. The findings in this section are organized by prominence of the theme. Within each thematic section, the data are accompanied by interpretations of quotes and, at times, organized by race. To protect the confidentiality of the participants, aliases are used for the focus group participants throughout the remainder of this article. ## 3.1. Community Care and Relationship-Building The concept of community care and relationship-building was the top theme that emerged across all six focus groups. I have been interrogating how this theme differs from the well-known concept of social capital and have several conclusions. Social capital, by name, arises from the concept of capital and capitalism [65]. Its first articulation in academia is commonly attributed to Pierre Bourdieu, James Coleman, and Robert Putnam [65,66]. It focuses on the resource—the what—that is being shared through a network [65]. The concept of “community care” and the value placed on “relationship-building” have roots in grassroots organizing and activism [67,68]. Civil rights movements and organizing practices prioritized relationships and care for one another along with their political goals [68]. Community care and relationship-building as a theme places emphasis on the people for whom resources are being exchanged. These concepts uplift the care and nurturing of one another. Community care and relationship-building are not about what is being shared but why it’s being shared. The focus here is on the people for whom resources and knowledge are being distributed. People in both the Black and White focus groups discussed community care and relationship-building in the context of preserving and restoring familial relationships and bonds. Julie, a participant in one of the White focus groups, shared that growing food allows her to connect with family members across social and political differences. She said, “It’s helped give me some common ground with my family who I do not always see eye to eye with”. For Megan, a participant in the White focus group, growing food helped her to connect with her family in a slightly different way: In her case, gardening may have been a more accessible and interesting topic to discuss for her family members. Her agricultural knowledge is valued and sought out by her family members and that leaves her feeling respected. For both Megan and Julie, agriculture helped them to connect with and restore familial relationships. The preservation and restoration of familial relationships had similar undertones in Black focus groups but seemed to be more focused on preserving and passing down agricultural knowledge and wisdom. Dr. Williams described the importance of preserving family foodways and reconnecting with relatives who live in the South. For her, growing food and preserving foodways provided a way for her to pass down her food and family’s traditions to her children. Sharing foodways and agricultural history has also provided Dr. Williams’ children with opportunities to build and deepen relationships with their family, especially after the death of their grandmother. Building relationships with and extending care to family is something Kisha also mentioned. She explained that her journey of growing food inspired her family members to be more intentional and active in caring for their mental and physical health: Her comment also highlights a consideration that Dr. Williams shared—future generations. Both women think of agriculture and their work as helping to preserve agricultural knowledge and pass it on to the next generation. Within the theme of community care and relationship-building, this subtheme is “intergenerational transfer of knowledge”. While both White and Black participants talked about strengthening familial relationships, the intergenerational aspect only appeared in conversations among Black participants. Another difference that emerged between Black participants and White participants was the way they described agriculture’s impacts on relationship-building. Within the White focus groups, participants often explained how growing food has helped increase their social connections. Social connection, as it appeared in the focus groups, was about the ways in which agriculture facilitated opportunities to meet, connect with, and learn about other people. Tom shared how growing food has helped him diversify his network. Allison and Mary echoed what Tom said: This idea of gardening helping one to diversify the people with whom they come into contact only appeared in White focus groups. There was no mention of racially diversifying one’s network or relationships in the Black focus groups. While some people in the Black focus groups did talk about the importance of social connection, people more frequently described relationships as a byproduct of community-building and community care. Community-building is about more than social connection. While community-building certainly encompasses social connection, it often consists of more than two people. It involves processes of community care, healing, and collective action. Building community requires intention and a level of reciprocity not necessary for social connection. Social connections can occur on the surface of interactions, whereas community-building, particularly among Black agricultural people, often involves considerable depth and substance to establish that trust within a group. Often, community-building is what enables the collective agency and collective action needed for organizing and activism. Ms. Crystal, an elder in the Black focus group, explains how the garden can be a space of relationship-building, organizing, and collective action: Ms. Crystal and others in the Black focus groups saw themselves not necessarily as individuals but as part of a larger whole. Tasha explicitly states how growing food has strengthened their identity: Seeing oneself as “we” instead of “I” is rooted in Indigenous African traditions and ways of being [69]. This ideology was consistent across the various Black focus groups. It was not as prevalent in the White focus groups. The concept of “othering” was more likely to appear in White focus groups. White participants often seemed to not see themselves as part of the community they were working in, in search of social connection, whereas Black participants tended to talk about themselves as a part of the community. Black focus groups also more frequently discussed the collective impacts of agriculture, while White participants were more likely to discuss the impacts on themselves as individuals. The theme of community care and relationship-building, as it appeared in the data, also seemed to be reflected in how participants interacted with and treated the environment. There were threads and codes related to climate change and sustainability that emerged. Participants often said caring for plants helped model how to care for and build relationships with people—bringing to life the saying, “as with the land, is with the people”. ## 3.2. Growing Food Facilitates Body–Mind Wellness Mental and physical wellness were strong themes across Black and White focus groups. Across all focus groups, participants talked about their mental wellbeing in relation to COVID-19 restrictions and protocols. For instance, Ms. Judy, an elder participant in the Black focus groups, talked about her experience quarantining in the garden: Allison, a participant in the White focus group, also shared that caring for her garden helped her to feel a sense of purpose during the pandemic: Gardening and farming seemed to offer Black and White participants opportunities to relieve stress, anxiety, and depression during the COVID-19 pandemic. However, grief processing and grief healing were topics that emerged more often in the Black focus groups than in the White focus groups. Sister Sahar shared how growing food and caring for gardens helped her process the passing of her mom: Other participants talked more generally about the impacts gardening or farming has on their mental health. Growing food and having your hands in the soil may impact both mental and physical wellbeing, in part, through soil microbes [28]. Quinn, a participant in one of the White focus groups, shared that working in the soil was the most impactful treatment for their mental health: Many focus group participants also described agriculture’s impacts on their physical health. People shared a range of reasons as to why they began growing. Many were related to sickness caused by eating processed or junk foods. Ms. Cheryl’s agricultural journey began with pursuing physical wellbeing, but she also shared how her personal journey led to physical wellness for her children as well: Body–Mind wellness seemed to have more similarities across races than differences. Many participants started gardening and/or farming because they wanted to have more control—agency—over their physical health. Especially during the pandemic, people sought being in the garden for the respite, peace, and stress relief that it offered. As articulated by several focus group participants, growing food also provides one with a heightened somatic sensory experience, helping them to realize the deep connections between mind and body. ## 3.3. Land-Based Spirituality and Interdependence Growing food and caring for the land have also helped people to nurture greater spiritual wellness and practices. Through being on the land, people have developed deeper connections to Earth, to more-than-human beings [52], to each other, and for some to a higher power and their ancestors. For Black and White participants, a primary way that agriculture has helped facilitate a deeper spirituality is through connection to Earth. Within White focus groups, this was most often through interdependence or Ubuntu. Ubuntu is a Zulu proverb that means “I am because we are” or “I am a person through you” [70]. Ubuntu is about the interdependence and interconnection of all humans. This quote from Sarah, a participant in one of the White focus groups, demonstrates a realization of interdependence: Sarah uses words and phrases such as “collective existence” and “collaboration” to illustrate how growing food has helped her to understand and think of the interdependence of all inhabitants of Earth. Ubuntu/interdependence was also a subtheme of land-based spirituality that appeared within the Black focus groups. Mostly, connections to Earth seemed to be more often described in these groups as talking with and hearing from plants. Kisha shared what the experience was like for her in her beginning days of farming: She uses words like “sacred” and “alchemy” to refer to the relationships that she was developing with plants. Other participants in the Black focus groups expressed similar sentiments to Kisha. Some described their relationship with plants as being ancestral. Tasha broke down the ways farming has helped them reconnect with their ancestors and reclaim their ancestral agricultural knowledge: Tasha offered valuable contributions to the conversation by naming the “symbiotic connection of just being on the land” and the “rituals of weeding and hoeing and prepping rows and planting”. Several of the Black growers talked about agriculture’s impacts on their spirituality through participating in rituals. Many of these rituals were actions, prayers, and ceremonies performed by their ancestors, who were also farmers and growers. In addition to rituals, some participants talked about the garden, and particularly soil, as a facilitator of deeper spiritual connections. Sister Sahar shares: Sometimes these experiences can feel unique to Black people. Kisha offered the following response to another focus group participant who shared an experience connecting to her husband when he became an ancestor. Kisha responded to this participant saying: Within the Black focus groups, spirituality was largely captured through connection. Connection was described in relation to God or a higher power, to ancestors, and to Earth. That connection was realized through ritualized actions, such as weeding, planting, and hoeing. These connections were also realized through simply being present in a garden or on a farm and being open to receiving communication from plants and all the life around. In the White focus groups, spirituality was a far less common topic of conversation; however, when it was, the participants described feeling connected to Earth and understanding the interdependence of everyone on this planet. ## 3.4. Growing Food as a Demonstration of Agency and Power Patricia Hill-Collins defines agency as “an individual or social group’s will to be self-defining and self-determining” [71]. I consider power to be the actualization of agency. In the context of urban agriculture, growing can be a demonstration of agency and power because of the ways it allows growers to be self-defining and self-determining. Across focus groups, but more prominently in the Black focus groups, agency and power were conveyed through conversations about economic alternatives to capitalism that enhance community wellbeing, as well as the importance of financial security. There were conversations about self-reliance and self-sufficiency, sovereignty, negative impacts of capitalism, and the process of unlearning. In many of these conversations, the participants named how agriculture has facilitated a process of re-educating themselves about food, economic systems, resource allocation, financial stability, and political systems. Many participants also talked about how they practice and actualize what they have learned from growing food with others. They spoke of unlearning incomplete truths that uphold the status quos of society. These incomplete truths were often in reference to societal norms, expectations, or lessons taught in school. The underlying theme that connected these threads is growing as a demonstration of agency and power. Truly knowing and understanding what you eat requires knowledge of where your food comes from and how it was produced. Instead of relying on someone else to tell you where food comes, you can rely on yourself because you grew the food. This is providing agency with respect to food. Being able to actualize this agency is power. Jada offered one example of how agency is demonstrated in the context of agriculture. In the Black focus group, she explained how growing food provides agency with respect to one’s health: Similarly, Ms. Valerie spoke to the importance of agency and how growing food can help people become more self-sufficient and not rely on city government to bring supermarkets and quality food to Black neighborhoods. One way to increase self-reliance is through growing food and raising animals. We can increase community self-reliance by passing on knowledge, wisdom, and skills to other people and generations. As an older woman, Ms. Valerie often highlighted her goal of transferring knowledge intergenerationally and how that facilitates agency and power. Ms. Crystal and Imani highlighted another aspect of agency and power: resistance. In “Sisters of the Soil: Urban Gardening as Resistance in Detroit,” Dr. Monica White characterizes resistance as being in response to “the social structures that have perpetuated inequality in terms of healthy food access,” and as the ability to “create outdoor, living, learning, and healing spaces for themselves and for members of the community” [17]. For Ms. Crystal and Imani, participating in urban agriculture provides them with the opportunity to organize and strategize on how best to obtain the resources their communities need to thrive. Here, Imani also speaks to the challenges of trying to operate within a White supremacist, or White violent, society. Her solution is to forge a path for her community to care for themselves through gardening and farming. Ms. Crystal also speaks of resistance by using phrases such as “join you in the fight”. She talks about resistance in relation to community organizing while also highlighting agriculture and the garden as a facilitator of community and relationship-building. This quote was referenced earlier to describe sentiments of community-building. It also speaks to agency and power: Participants in the White focus groups, at times, used similar language as participants in the Black focus groups to describe their experiences around agency, power, and resistance. Two examples of this are from Tom and Quinn: Mentions of resistance and justice were coded as examples of activism and resistance. “ Activism and resistance” was one of several subthemes that made up the larger theme, “growing food as a demonstration of agency and power”. An interesting thread that appeared across White focus groups was a desire to be a part of a “larger movement for justice”. Some participants talked about how agriculture helps them to feel connected to resistance and justice work for Black and other PGM communities, even though they do not necessarily belong to the communities experiencing multiple systems of oppression. Quinn, a nonbinary person in their early twenties, exemplifies an understanding of shared language around agriculture as resistance but comes from a very different set of experiences than many of the people in the Black focus groups. For instance, Quinn relocated to the Philadelphia area after college and worked as an urban grower while most of the Black participants were native to Philadelphia living in some of the city’s most rapidly gentrifying neighborhoods. They, Quinn, mentions how systems “try to keep you from” realizing your connection to the Earth and food. It raises the question of, to which systems is Quinn referring? When they talk of systems and systems of oppression, does this also include systems of gentrification, displacement, and neighborhood resource extraction? Resistance is not just about gardening or farming or knowing where one’s food comes from despite the trend we see among White people to claim gardening and farming as resistance. There is a difference between gardening as a hobby and growing food as resistance for yourself, your family, and your community. Growing as resistance, as collective agency, as community power, is about working against the intertwined systems and structures that seek to oppress communities, particularly Black and PGM communities in America. Growing food can be a model of constructive resistance and collective agency but not when it is focused on the individual benefits of growing food for a single person who already holds immense privilege; or when it involves a person of immense privilege entering Black communities hoping to help or save those communities. Growing food as constructive resistance means “the aggrieved actively build alternatives to existing political and economic relationships” [36]. It is how people living under oppression engage in radical acts of “building knowledge, skills, community, and economic independence” [36]. Some participants in the White focus groups were quicker than others to acknowledge the privilege that they have and how growing food is different for them than it may be for someone else. One example of this is from Megan. She says, “I feel like there’s a, I mean, a part of my own privilege where I’m not reliant on the food that I’m growing”. As participants in the White focus groups engaged with ideas of self-reliance and self-sufficiency, another concept that emerged was around future preparedness or “dooms-day prepping” [72]. This relates to the theme of growing food as a demonstration of agency and power because participants talked about growing food as a skill necessary for taking care of oneself and surviving in the future. Both Quinn and Danielle referenced this skill in relation to Octavia Butler novels: Another thread that was tied to agency and power is the recognition of some of the impacts of living under a capitalist economic system. Megan shared what that realization was like for her: Megan explored growing food as a potential way to decrease her engagement with capitalism. Because she is growing her own food, she does not have to rely as heavily on exchanging money for goods exploitatively produced for profit. Offering another scenario for this interaction between growing food and capitalism, Sarah shared how an economic system can lead to some negative physical impacts: The exploitation of migrant workers and other farmworkers is often tied to profit-driven agricultural businesses [73]. Many of these businesses focus more on yields than they do on the health, safety, and wellbeing of their workers. The disregard and exploitation of farmworkers in the United States has deep roots in racism and xenophobia [73]. It is also tied to a culture of production and productivity. Another participant in the White focus groups, Jennifer, acknowledged some of the subtle ways that a culture of productivity convinced her that she should be doing more and growing more: Amy offered the following experience in response to Jennifer’s comment: Overall, agency and power showed up in White focus groups as unlearning, critiques of capitalism, and future preparedness. White participants engaged with concepts of resistance, activism, and justice from social positions that differed from Black participants. While people across racial groups talked about topics related to agency and power, Black participants discussed them more frequently. There was also a difference in the emphasis people placed on specific topics. For instance, in one of the Black focus groups, participants dedicated most of the allotted time to talking about activism and resistance. While White participants also named resistance as an impact of urban agriculture, the conversations around resistance were shorter, less emphatic, and did not contain the same sense of urgency as in the Black focus groups. In general, Black focus groups included more co-signing of ideas and intergroup dialogue than the White focus groups. ## 4. Discussion Rich and interesting information emerged from these data. First, the concept of community care and relationship-building was the number one theme that emerged across all six focus groups. The data also showed that there are both similarities and differences in the perceived impacts of urban agriculture by race. Mentions of spirituality appeared more frequently and more emphatically in the Black focus groups. Both Black and White focus groups emphasized the mental health impacts of urban agriculture. They talked about concepts related to community care and relationship-building as being a primary benefit of growing food. In both groups, people brought up significant issues and barriers around land security. Many of the emerging themes of the data align with Dr. Monica White’s theoretical framework on collective agency and community resilience, especially within the Black focus groups. Her framework discusses Black agriculture in the context of economic autonomy, commons as praxis, and prefigurative politics. White’s work articulates the centrality of agriculture and Black farmers to Black organizing, self-determination, and liberation [36]. I believe a major strength of this study is that it builds on the quantitative findings of an earlier study that looks at the associations between neighborhood demographics and community food-growing spaces [22]. Through spatial and historical analyses of urban agriculture in Philadelphia, this study found patterning of community gardens and urban farms in Black neighborhoods, low-income neighborhoods, and neighborhoods with low food access. The authors suggested that the establishment of gardens and farms may be a response to the extraction experienced in those neighborhoods. Because that study was nontemporal and noncausal, authors suggest future studies unpack the ways that Black-led agriculture impacts community health. The focus group design employed in this study allowed me to better understand these impacts in a more in-depth way that may not be possible using quantitative methods. The impacts of urban agriculture on spirituality and collective agency have not been extensively studied. The findings of this focus group study point to some key domains through which agriculture impacts the health of farmers and growers in Philadelphia. Future studies could focus on quantitatively assessing the association between the themes outlined in this article and health and agency. Lastly, all of the outcome variables in this study are not directly observable or measurable. When studying how an exposure affects these kinds of latent concepts, I think the most appropriate way to gain insights and answers is to begin with asking the people impacted directly. Lastly, this work and the recruitment process highlight the absolute importance of thoroughly engaging with and supporting community partners. As I mentioned earlier, the recruitment goals were met in under two weeks. I attribute this to the deep relationships and connections I have with urban growers and the food justice community throughout the city of Philadelphia. In each of the focus groups, I spent some time asking participants to write down a few topic areas or questions related to urban agriculture that they thought were the most important to include in a survey. The data and findings that emerged from this study directly informed the development of a scale that was recently validated. The goal of this tool is to measure concepts such as spirituality, wellbeing, agency, and power among agriculturalists and agrarian communities. ## References 1. Nixon R.. Slow Violence and the Environmentalism of the Poor (2011.0). DOI: 10.4159/harvard.9780674061194 2. Larson N.I., Story M.T., Nelson M.C.. Neighborhood environments: Disparities in access to healthy foods in the U.S. Am. J. Prev. Med. (2009.0) 36 74-81. DOI: 10.1016/j.amepre.2008.09.025 3. Jeffery R.W., French S.A.. Epidemic obesity in the United States: Are fast foods and television viewing contributing?. Am. J. Public Health. (1998.0) 88 277-280. DOI: 10.2105/AJPH.88.2.277 4. Block J.P., Scribner R.A., DeSalvo K.B.. Fast food, race/ethnicity, and income: A geographic analysis. Am. J. Prev. Med. (2004.0) 27 211-217. DOI: 10.1016/j.amepre.2004.06.007 5. Berger N., Kaufman T.K., Bader M.D., Rundle A.G., Mooney S.J., Neckerman K.M., Lovasi G.S.. Disparities in trajectories of changes in the unhealthy food environment in New York City: A latent class growth analysis, 1990–2010. Soc. Sci. Med. (2019.0) 234 112362. DOI: 10.1016/j.socscimed.2019.112362 6. Zenk S.N., Schulz A.J., Israel B.A., James S.A., Bao S., Wilson M.L.. Neighborhood Racial Composition, Neighborhood Poverty, and the Spatial Accessibility of Supermarkets in Metropolitan Detroit. Am. J. Public Health. (2005.0) 95 660-667. DOI: 10.2105/AJPH.2004.042150 7. Morland K., Diez Roux A.V., Wing S.. Supermarkets, other food stores, and obesity: The atherosclerosis risk in communities study. Am. J. Prev. Med. (2006.0) 30 333-339. DOI: 10.1016/j.amepre.2005.11.003 8. Fang D., Thomsen M.R., Nayga R.M.. The association between food insecurity and mental health during the COVID-19 pandemic. BMC Public Health. (2021.0) 21. DOI: 10.1186/s12889-021-10631-0 9. Medical News Today 10. Diez-Roux A.V., Nieto F.J., Caulfield L., Tyroler H.A., Watson R.L., Szklo M.. Neighbourhood differences in diet: The Atherosclerosis Risk in Communities (ARIC) Study. J. Epidemiol. Community Health. (1999.0) 53 55-63. DOI: 10.1136/jech.53.1.55 11. Rundle A., Roux A.V.D., Freeman L.M., Miller D., Neckerman K.M., Weiss C.C.. The Urban Built Environment and Obesity in New York City: A Multilevel Analysis. Am. J. Health Promot. (2007.0) 21 326-334. DOI: 10.4278/0890-1171-21.4s.326 12. Nelson M.C., Gordon-Larsen P., Song Y., Popkin B.M.. Built and social environments associations with adolescent overweight and activity. Am. J. Prev. Med. (2006.0) 31 109-117. DOI: 10.1016/j.amepre.2006.03.026 13. Stimpson J.P., Nash A.C., Ju H., Eschbach K.. Neighborhood Deprivation is associated with lower levels of serum carotenoids among adults participating in the Third National Health and Nutrition Examination Survey. J. Am. Diet. Assoc. (2007.0) 107 1895-1902. DOI: 10.1016/j.jada.2007.08.016 14. Chang V.W.. Racial residential segregation and weight status among US adults. Soc. Sci. Med. 1982 (2006.0) 63 1289-1303. DOI: 10.1016/j.socscimed.2006.03.049 15. Tuck E.. Suspending Damage: A Letter to Communities. Harv. Educ. Rev. (2009.0) 79 409-428. DOI: 10.17763/haer.79.3.n0016675661t3n15 16. Gripper A.. We Don’t Farm Because It’s Trendy; We Farm as Resistance, for Healing and Sovereignty. Agents of Change in Environmental Health 17. White M.M.. Sisters of the Soil: Urban Gardening as Resistance in Detroit. Race/Ethn. Multidiscip. Glob. Context. (2011.0) 5 13-28. DOI: 10.2979/racethmulglocon.5.1.13 18. White M.M.. Environmental Reviews & Case Studies: D-Town Farm: African American Resistance to Food Insecurity and the Transformation of Detroit. Environ. Pract. (2011.0) 13 406-417. DOI: 10.1017/S1466046611000408 19. White M.M.. Shouldering Responsibility for the Delivery of Human Rights: A Case Study of the D-Town Farmers of Detroit. Race/Ethn. Multidiscip. Glob. Context. (2010.0) 3 189-211 20. Reese A.M.. “We will not perish; we’re going to keep flourishing”: Race, Food Access, and Geographies of Self-Reliance. Antipode (2018.0) 50 407-424. DOI: 10.1111/anti.12359 21. Smith B.J.. Building Emancipatory Food Power: Freedom Farms, Rocky Acres, and the Struggle for Food Justice. J. Agric. Food Syst. Community Dev. (2019.0) 8 33-43. DOI: 10.5304/jafscd.2019.084.009 22. Gripper A.B., Nethery R., Cowger T.L., White M., Kawachi I., Adamkiewicz G.. Community solutions to food apartheid: A spatial analysis of community food-growing spaces and neighborhood demographics in Philadelphia. Soc. Sci. Med. (2022.0) 310 115221. DOI: 10.1016/j.socscimed.2022.115221 23. Kaufman J., Bailkey M.. Farming Inside Cities: Entrepreneurial Urban Agriculture in the United States (2000.0) 24. Food and Agriculture Organization of the United Nations 25. Lal R.. Home gardening and urban agriculture for advancing food and nutritional security in response to the COVID-19 pandemic. Food Secur. (2020.0) 12 871-876. DOI: 10.1007/s12571-020-01058-3 26. Walljasper C., Polansek T.. Home Gardening Blooms around the World during Coronavirus Lockdowns. Reuters 27. Robinson-Jacobs K.. The Pandemic Pushes more Black Americans to Take up Urban Farming to Fight “Food Apartheid”. The GroundTruth Project. (2021.0) 28. Blum W.E.H., Zechmeister-Boltenstern S., Keiblinger K.M.. Does Soil Contribute to the Human Gut Microbiome?. Microorganisms (2019.0) 7. DOI: 10.3390/microorganisms7090287 29. Brown K.H., Jameton A.L.. Public Health Implications of Urban Agriculture. J. Public Health Policy (2000.0) 21 20-39. DOI: 10.2307/3343472 30. Gill S.E., Handley J.F., Ennos A.R., Pauleit S.. Adapting Cities for Climate Change: The Role of the Green Infrastructure. Built Environ. (2007.0) 33 115-133. DOI: 10.2148/benv.33.1.115 31. Solecki W.D., Rosenzweig C., Parshall L., Pope G., Clark M., Cox J., Wiencke M.. Mitigation of the heat island effect in urban New Jersey. Glob. Environ. Change Part B Environ. Hazards (2005.0) 6 39-49. DOI: 10.1016/j.hazards.2004.12.002 32. Oliveira S., Andrade H., Vaz T.. The cooling effect of green spaces as a contribution to the mitigation of urban heat: A case study in Lisbon. Build. Environ. (2011.0) 46 2186-2194. DOI: 10.1016/j.buildenv.2011.04.034 33. Prudencio L., Null S.E.. Stormwater management and ecosystem services: A review. Environ. Res. Lett. (2018.0) 13 033002. DOI: 10.1088/1748-9326/aaa81a 34. Alaimo K., Reischl T.M., Allen J.O.. Community gardening, neighborhood meetings, and social capital. J. Community Psychol. (2010.0) 38 497-514. DOI: 10.1002/jcop.20378 35. PGMONE: People of the Global Majority Outdoor in the Outdoors, Nature and Environment 36. White M.M.. Freedom Farmers: Agricultural Resistance and the Black Freedom Movement (2018.0) 37. White M.M.. Freedom’s Seeds: Collective Agency and Community Resilience: A Theoretical Framework to Understand Agricultural Resistance. J. Agric. Food Syst. Community Dev. (2017.0) 7 17-21. DOI: 10.5304/jafscd.2017.074.014 38. Malberg Dyg P., Christensen S., Peterson C.J.. Community gardens and wellbeing amongst vulnerable populations: A thematic review. Health Promot. Int. (2020.0) 35 790-803. DOI: 10.1093/heapro/daz067 39. George D.R.. Harvesting the Biopsychosocial Benefits of Community Gardens. Am. J. Public Health (2013.0) 103 e6. DOI: 10.2105/AJPH.2013.301435 40. Austin E.N., Johnston Y.A.M., Morgan L.L.. Community Gardening in a Senior Center: A Therapeutic Intervention to Improve the Health of Older Adults. Ther. Recreat. J. (2006.0) 40 41. Hartwig K.A., Mason M.. Community Gardens for Refugee and Immigrant Communities as a Means of Health Promotion. J. Community Health (2016.0) 41 1153-1159. DOI: 10.1007/s10900-016-0195-5 42. Golden S.. Urban Agriculture Impacts: Social, Health, and Economic: An Annotated Bibliography (2013.0) 22 43. Bradley K., Galt R.E.. Practicing food justice at Dig Deep Farms & Produce, East Bay Area, California: Self-determination as a guiding value and intersections with foodie logics. Local Environ. (2014.0) 19 172-186. DOI: 10.1080/13549839.2013.790350 44. McClintock N.. Radical, reformist, and garden-variety neoliberal: Coming to terms with urban agriculture’s contradictions. Local Environ. (2014.0) 19 147-171. DOI: 10.1080/13549839.2012.752797 45. Cohen N., Reynolds K.. Resource needs for a socially just and sustainable urban agriculture system: Lessons from New York City. Renew. Agric. Food Syst. (2015.0) 30 103-114. DOI: 10.1017/S1742170514000210 46. Ghose R., Pettygrove M.. Urban Community Gardens as Spaces of Citizenship. Antipode (2014.0) 46 1092-1112. DOI: 10.1111/anti.12077 47. Franzen S.. The value of farming: Multifaceted wealth generation through cooperative development. Econ. Anthropol. (2020.0) 7 279-292. DOI: 10.1002/sea2.12178 48. Safransky S.. Rethinking Land Struggle in the Postindustrial City. Antipode (2017.0) 49 1079-1100. DOI: 10.1111/anti.12225 49. Ober Allen J., Alaimo K.A., Elam D., Perry E.. Growing Vegetables and Values: Benefits of Neighborhood-Based Community Gardens for Youth Development and Nutrition. J. Hunger Environ. Nutr. (2008.0) 3 418-439. DOI: 10.1080/19320240802529169 50. Carney J.. In the Shadow of Slavery: Africa’s Botanical Legacy in the Atlantic World (2011.0) 51. wa Thiong’o N.. Something Torn and New: An African Renaissance (2009.0) 52. 52. Robin Wall Kimmerer Braiding Sweetgrass: Indigenous Wisdom, Scientific Knowledge and the Teachings of Plants1st ed.Milkweed EditionsMinneapolis, MN, USA2013Available online: http://ezp-prod1.hul.harvard.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&db=nlebk&AN=683745&site=ehost-live&scope=site(accessed on 28 March 2022). Braiding Sweetgrass: Indigenous Wisdom, Scientific Knowledge and the Teachings of Plants (2013.0) 53. McCutcheon P.. Growing Black food on sacred land: Using Black liberation theology to imagine an alternative Black agrarian future. Environ. Plan. Soc. Space (2021.0) 39 887-905. DOI: 10.1177/02637758211032975 54. McCutcheon P.. “Returning Home to Our Rightful Place”: The Nation of Islam and Muhammad Farms. Geoforum (2013.0) 49 61-70. DOI: 10.1016/j.geoforum.2013.05.001 55. USDA ERS-Glossary 56. Quisumbing King K., Wood S.D., Gilbert J., Sinkewicz M.. Black Agrarianism: The Significance of African American Landownership in the Rural South. Rural Sociol. (2018.0) 83 677-699. DOI: 10.1111/ruso.12208 57. Kerkvliet B.J.T., Scott J.C.. Everyday Forms of Peasant Res Cb: Everyday Forms Res Asia (2013.0). DOI: 10.4324/9781315827926 58. Krueger R.A., Casey M.A.. Focus Groups: A Practical Guide for Applied Research (2014.0) 59. Morgan D.L.. Focus Groups. Annu. Rev. Sociol. (1996.0) 22 129-152. DOI: 10.1146/annurev.soc.22.1.129 60. Sale M.. Call and Response as Critical Method: African-American Oral Traditions and Beloved. Afr. Am. Rev. (1992.0) 26 41-50. DOI: 10.2307/3042075 61. Fusch P.I., Ness L.R.. Are We There Yet? Data Saturation in Qualitative Research. Qual. Rep. (2015.0) 20 1408-1416. DOI: 10.46743/2160-3715/2015.2281 62. Denzin N.K., Lincoln Y.S.. The SAGE Handbook of Qualitative Research (2017.0) 63. Blaikie N.. Designing Social Research: The Logic of Anticipation (2009.0) 64. Scandura T., Williams E.A.. Research methodology in management: Current practices, trends, and implications for future research. Acad. Manag. J. (2000.0) 43 1248-1264. DOI: 10.2307/1556348 65. Fine B., Lapavitsas C.. Social Capital And Capitalist Economies. South-East Eur. J. Econ. (2004.0) 2 17-34 66. Social Capital–Infed.Org 67. 67.Lessons from Social MovementsNational Academies Press (US)Washington, DC, USA2014Available online: https://www.ncbi.nlm.nih.gov/books/NBK268722/(accessed on 28 March 2022). Lessons from Social Movements (2014.0) 68. Ganz M.. Leading change: Leadership, organization, and social movements. Handb. Lead. Theory Pract. (2010.0) 19 1-10 69. Asante P.A.H.. The African Concept of Community and Individual in the Context of the Market. Distant Markets, Distant Harms (2014.0). DOI: 10.1093/acprof:oso/9780199370993.003.0008 70. Foundation C.. The Spirit of Ubuntu. Medium. (2017.0) 71. Hill Collins P.. Black Feminist Thought: Knowledge, Consciousness, and the Politics of Empowerment (2000.0) 72. Smith N., Thomas S.J.. Doomsday Prepping During the COVID-19 Pandemic. Front. Psychol. (2021.0) 12 659925. DOI: 10.3389/fpsyg.2021.659925 73. Sbicca J., Minkoff-Zern L.A., Coopwood S.. Because they are connected”: Linking structural inequalities in farmworker organizing. Hum. Geogr. (2020.0) 13 263-276. DOI: 10.1177/1942778620962045
--- title: BGP-15 Protects against Doxorubicin-Induced Cell Toxicity via Enhanced Mitochondrial Function authors: - Alexandra Gyongyosi - Nikolett Csaki - Agota Peto - Kitti Szoke - Ferenc Fenyvesi - Ildiko Bacskay - Istvan Lekli journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049233 doi: 10.3390/ijms24065269 license: CC BY 4.0 --- # BGP-15 Protects against Doxorubicin-Induced Cell Toxicity via Enhanced Mitochondrial Function ## Abstract Doxorubicin (DOX) is an efficacious and commonly used chemotherapeutic agent. However, its clinical use is limited due to dose-dependent cardiotoxicity. Several mechanisms have been proposed to play a role in DOX-induced cardiotoxicity, such as free radical generation, oxidative stress, mitochondrial dysfunction, altered apoptosis, and autophagy dysregulation. BGP-15 has a wide range of cytoprotective effects, including mitochondrial protection, but up to now, there is no information about any of its beneficial effects on DOX-induced cardiotoxicity. In this study, we investigated whether the protective effects of BGP-15 pretreatment are predominantly via preserving mitochondrial function, reducing mitochondrial ROS production, and if it has an influence on autophagy processes. H9c2 cardiomyocytes were pretreated with 50 μM of BGP-15 prior to different concentrations (0.1; 1; 3 μM) of DOX exposure. We found that BGP-15 pretreatment significantly improved the cell viability after 12 and 24 h DOX exposure. BGP-15 ameliorated lactate dehydrogenase (LDH) release and cell apoptosis induced by DOX. Additionally, BGP-15 pretreatment attenuated the level of mitochondrial oxidative stress and the loss of mitochondrial membrane potential. Moreover, BGP-15 further slightly modulated the autophagic flux, which was measurably decreased by DOX treatment. Hence, our findings clearly revealed that BGP-15 might be a promising agent for alleviating the cardiotoxicity of DOX. This critical mechanism appears to be given by the protective effect of BGP-15 on mitochondria. ## 1. Introduction Doxorubicin (DOX) is a potent chemotherapeutic agent widely used to treat a variety of cancers [1]. The clinical use of DOX has been associated with cumulative, dose-dependent cardiotoxicity, while off-target drug toxicity is associated with oxidative stress that involves the development of heart failure [2]. The mechanisms of DOX-induced toxicity have not been clearly elucidated, but are known to involve, at least in part, mitochondrial dysfunction, leading to an increased generation of intracellular ROS, oxidative stress, and apoptosis [3,4]. Thus, DOX cardiotoxicity is closely associated with mitochondrial injury, which is characterized by iron overload and an early loss of mitochondrial membrane potential (MMP) followed by dysregulation of the mitochondrial quality control mechanism [5]. Additionally, DOX activated apoptosis, due to an imbalance in oxidant and anti-oxidant. Therefore, many pathways might be speculated to be responsible for apoptosis induction and there may be cross-talk between these various pathways, including the mitochondrial pathway through caspase-3 activation [6]. More recently, it has been suggested that dysregulation of autophagy may also play a contributing role in DOX-induced cardiotoxicity [7,8]. Autophagy has been shown to have dual functions. Autophagy can enhance cellular function and survival by degrading damaged or unwanted organelles and by inhibiting apoptosis. Alternatively, autophagy can also induce cell death [9]. Several studies have shown that DOX treatment affects autophagy in vitro and in vivo. Some have shown that DOX treatment increases autophagy and some have shown that DOX decreases autophagy [10]. Since autophagy has dual functions in the life and death of cardiomyocytes, several investigators have employed chemical means of manipulating autophagy to elucidate its role in DOX-induced cardiotoxicity. However, it is important to note that many of the most commonly used chemical modulators of autophagy have off-target effects to be considered when interpreting results [11]. With all of the above molecular mechanisms leading to DOX-induced cardiotoxicity, its clinical use is limited. In this present study, our attempt was to investigate the effect of BGP-15 on DOX-induced injury. BGP-15 (O-[3-piperidino-2-hydroxy-1-propyl]-nicotinic amidoxime) possesses a wide range of cytoprotective effects but lacks a clear intracellular molecular target [12]. BGP-15 protects the mitochondrial membrane system, decreases oxidative stress [13], inhibits the nuclear translocation of apoptosis-inducing factor (AIF) from mitochondria, and inhibits mitogen-activated protein kinase (MAPK) activation [12]. BGP-15 shows several beneficial cardiovascular effects and has increasingly raised scientific interest in a wide range of pathological conditions in several disease models [14,15]. Although several protective mechanisms of BGP-15 were identified, its effects on DOX-induced cardiotoxicity are not yet investigated. In the current study, we have tested the effect of BGP-15 treatment on DOX-induced injury in H9c2 cells. ## 2.1. Effects of BGP-15 Pretreatment on Cell Viability and LDH Release of DOX-Induced Cardiotoxicity In order to evaluate the potential cardioprotective effect of BGP-15 against DOX-induced toxicity, a cell viability assay was carried out. As shown in Figure 1, panels A and B, 12 or 24 h of DOX exposure at doses between 0.1 and 3 μM induced a significant dose-dependent decrease in cell viability in comparison with the control group ($p \leq 0.0001$). Of note, no cytotoxicity was observed in response to 50 μM of BGP-15 alone. Thus, we assessed the effect of DOX on cell viability in the presence of BGP-15. Our findings showed that BGP-15 pretreatment at 0.1, 1, and 3 μM DOX groups significantly ameliorated cell viability in comparison with only DOX-treated cardiomyocytes at the same concentration after 12 h and 24 h of DOX exposure, respectively. To further confirm the protective effect of BGP-15 treatment on DOX-induced toxicity, the LDH content of cell culture media was determined by a colorimetric assay. Our results showed (Figure 1C) that DOX increased the LDH release of the cells in a dose-dependent manner. In line with the MTT assay, BGP-15 pretreatment did not just improve the cell viability, but also significantly decreased the DOX-induced LDH release of the cardiomyocytes. We measured a significant decrement in LDH release in the presence of BGP-15 compared to DOX treatment alone (1 μM DOX: 22.88 ± $0.78\%$ vs. 17.04 ± $0.59\%$ and 3 μM DOX: 25.91 ± $1.14\%$ vs. 20.33 ± $0.85\%$). ## 2.2. BGP-15 Attenuates the DOX-Induced Generation of Mitochondrial ROS and Slightly Diminishes the Activation of Caspase-3 Apoptosis Marker in H9c2 Cells ROS are involved in DOX-induced cell death [3]. Several studies have suggested that cardiomyocyte mitochondria are important intracellular targets of excess ROS during DOX-induced cardiotoxicity. Superoxide is one of the major ROS generated after DOX treatment [16]. Thus, to study the role of ROS in the protection induced by BGP-15 treatment (Figure 1), cells were analyzed for mitochondrial superoxide anion generation by flow cytometry in the presence or absence of BGP-15 in cardiomyocytes challenged by DOX treatment (Figure 2A). Our results indicated that DOX increased the mitochondrial superoxide generation compared to the control cells in a dose-dependent manner. Quantitative measurements of the mean fluorescence intensities of the samples demonstrated that 1 and 3 μM DOX alone significantly increased the ROS level (696.58 ± 42.34 and 992.03 ± 143.17, respectively) in contrast to the control group (408.18 ± 9.75). Conversely, enhanced MitoSOX fluorescence intensity induced by the DOX treatment was lessened by pretreatment with BGP-15, which was significantly lower in the BGP-15 + DOX3 group in comparison with DOX3-treated cells, indicating that the level of mitochondrial superoxide generation decreased in H9c2 cells in the presence of BGP-15. Fluorescent microscopy was employed to visualize MitoSOX staining (Figure 2B). However, we observed a notable accumulation of DOX in the nucleus, which makes it difficult to quantify the fluorescence intensity of microscopic images. These results suggest that decreased ROS generation may play a role in the cytoprotective effect of BGP-15 in H9c2 cells against DOX-induced cell toxicity. In order to investigate the activation of apoptosis, we analyzed the ratio of cleaved-caspase-3 (17 kDa) /pro-caspase-3 (35 kDa) after the cardiomyocyte cells were exposed to 1 μM DOX for 24 h in the absence or presence of 50 μM BGP-15 pretreatment (Figure 2C). Our results showed that 1 μM DOX for 24 h significantly enhanced the ratio of cleaved-caspase-3/pro-caspase-3 (0.57 ± 0.08) in comparison with the control group (0.05 ± 0.01), indicating the activation of apoptosis. BGP-15 alone did not alter the ratio of cleaved-caspase-3/pro-caspase-3. Although the pretreatment of BGP-15 could slightly withhold the activation of apoptosis, unfortunately, the ratio of these abovementioned proteins was not statistically significant (0.43 ± 0.07) (p value = 0.26). ## 2.3. Effects of BGP-15 Pretreatment on Mitochondrial Depolarization of DOX-Exposed H9c2 Cells Mitochondria are the primary target organelles of DOX-induced cardiotoxicity [17]. Mitochondrial membrane potential (MPP) is necessary for the production of ATP, which is crucial in living cells. JC-1 was used to assess the Δψm in H9c2 cardiomyocytes. This dye can selectively enter the mitochondria where it reversibly changes color as membrane potentials increase (over values of about 80–100 mV). The monomeric form of JC-1 in the cytosol emits a green fluorescence, and aggregates of the dye in the mitochondria of normal cells emit a red fluorescence. To confirm our fluorescent intensity (ratio red/green) results, JC-1 staining was carried out. Samples were visualized by fluorescent microscopy, with healthy mitochondria in red and unhealthy mitochondria in green. As shown in Figure 3A, B, the Dox-induced depolarization was mitigated by BGP-15 pretreatment. The ratio of fluorescent intensity was 120 ± 5.58 in the BGP-15 alone group. Our results revealed that DOX induced significant MMP loss in the 1 and 3 μM DOX groups (62.76 ± 4.36 and 50.43 ± 5.01, respectively) versus the control group ($100\%$); however, MMP was recovered by the BGP-15 pretreatment (79.6 ± 5.75 and 63.94 ± 7.37, respectively), which was a significant improvement on BGP-15 + DOX 1 vs. DOX 1. ## 2.4. Effects of BGP-15 on Autophagy Flux in DOX-Induced Cytotoxicity To monitor autophagic flux, cells were treated with chloroquine, which is a known autophagic flux inhibitor. Protein expression levels of LC3B (Figure 4A) and p62 (Figure 4C) were measured with Western blot, and lysosome and LC3B or p62 colocalization (Figure 4B,D) were determined by fluorescent microscopy. Our results showed that chloroquine significantly increased the LC3B relative protein expression in control vs. control+ chloroquine (1 ± 0 vs. 1.98 ± 0.23) and BGP-15 vs. BGP-15+ chloroquine (1.22 ± 0.08 vs. 2.37 ± 0.22) groups. However, chloroquine enhanced only moderately the LC3B expression in the case of DOX 1 (1.05 ± 0.11 vs. 1.59 ± 0.23) and BGP-15 + DOX 1 (1.02 ± 0.15 vs. 1.40 ± 0.18) groups. In contrast, expression of p62 was significantly reduced in the DOX 1 (0.16 ± 0.02) and BGP-15 + DOX 1 (0.11 ± 0.03) groups compared to the control (1 ± 0) and BGP-15 (1.12 ± 0.09) groups. Western-blot results were supported by microscopic images. However, it appears that modulation of autophagic flux is not likely to play a direct role in the cytoprotective effects of BGP-15 in DOX-induced toxicity. ## 3. Discussion Pharmacological interventions that are able to enhance the resistance of myocardium against DOX-induced cardiac complications may offer a new perspective on the application of DOX in different tumors. In the current study, we found that BGP-15 mitigates DOX-induced cell death in H9c2 cells, evidenced by enhanced cell survival and decreased LDH release upon DOX treatment. Furthermore, BGP-15 decreased mitochondrial ROS production and mitochondrial depolarization in DOX-challenged cells. Earlier, BGP-15, a nicotinic acid derivative, has been shown to protect the myocardium against different injuries, including ischemia/reperfusion and heart failure with different triggers [14,15,18,19]. Mitochondrial dysfunction plays an important role in different cardiovascular diseases including DOX-induced cardiotoxicity. However, the mechanisms contributing to DOX-induced cardiotoxicity are not fully understood; the role of increased ROS production and enhanced oxidative stress appears to be one of the major factors. An enhanced amount of ROS impairs redox balance causing DNA damage, lipid peroxidation, mitochondrial dysfunction, and dysregulation of autophagy and apoptosis [20,21,22]. Ultimately, these alterations led to contractile dysfunctions, cardiomyopathy, and heart failure. DOX redox cycles on mitochondrial complex I, leading to ROS generation [23]. Moreover, an enhanced mitochondrial iron level upon DOX treatment also contributes to ROS generation [24]. Increased mitochondrial ROS leads to compromised mitochondrial integrity opening of the mitochondrial permeability transition pores, which causes modulation of Keap1/Nrf2 and alters regulation of the mitochondrial biogenesis [25,26]. BGP-15 has been shown to protect against oxidative stress and LPS-induced mitochondrial depolarization [27]. The authors suggested that BGP-15 inhibits mitochondrial Complex I and III, thereby suppressing ROS production and ultimately preventing the activation of ROS-dependent signaling pathways including MAPK and PARP and influencing cell death [27,28]. In line with the literature, we found decreased ROS production of cells treated with DOX in the presence of BGP-15, evidenced by the results of flow cytometry. Moreover, BGP-15 prevented DOX-induced mitochondrial depolarization. Our results show a slight decrement in the activation of caspase-3 in the presence of BGP-15 in DOX-treated H9c2 cells. However, it was not statistically significant. Impaired autophagy also plays a role in DOX-induced cell toxicity. Earlier, we have shown that DOX treatment impairs autophagic flux, which can be restored by metformin treatment. Metformin also targets mitochondrial complex I leading to a decreased ATP/AMP ratio, which activates AMPK and suppresses mTOR signaling leading to the activation of autophagy [29]. In the current study, we also found a weakened autophagy flux in the presence of DOX. Suppression of autophagic flux is mostly reported dose-dependently by DOX. It has been suggested that 1 μM DOX concentration does reflect the clinically relevant context [30], so we employed that concentration. However, BGP-15 did not restore it. Similar results were seen by Li et al. [ 31]. They observed that the treatment of NRVM with DOX (1 μM) resulted in a decrease in the autophagic flux within 6 h based on the measured LC3B-II and p62 levels. Moreover, by tracking lysosomes with Lysotracker Red, a fluorescence dye that labels acidic organelles, we also found that DOX decreased Lysotracker Red puncta. Although we did not quantify Lysotracker Red staining puncta, based on the microscopic results depicted in Figure 4, panels C and D, it is visible that upon DOX treatment the number of lysosomes decreased. It has been reported in some cell types that an increase in lysosome pH can impair the fusion of lysosome with autophagosomes [31,32]. Of note, based on our microscopic pictures, the fluorescent signals were slightly increased in BGP-15 + DOX1 + Q treated cells. Interestingly, the extent of autophagic flux perturbation correlated with the level of DOX-induced ROS production, leading further support to the notion that restoration of autophagic flux protects against DOX-induced cardiotoxicity. Taken together, based on our data, we cannot completely rule out that BGP-15 may influence autophagic flux; however, further studies need to be carried out to clarify the question. In conclusion, our results indicated that BGP-15 could prevent DOX-induced cell toxicity by decreasing mitochondrial ROS production and attenuating mitochondrial depolarization. ## 4.1. Materials Medium, serum, MTT, chloroquine, and LDH were purchased from Sigma (St. Louis, MO, USA). JC-1, MitoSOX and Lysotracker were bought from Life Technologies (Paisley, Scotland). Stain-Free gels and PVDF membrane were purchased from Bio-Rad Laboratories (Hercules, CA, USA). LC3B, p62, and Caspase-3 antibodies were obtained from Cell Signaling Technology (Boston, MA, USA). For fluorescent microscopy, p62 was purchased from Abcam (Cambridge, UK). ## 4.2. Cell Culture The H9c2 cells were obtained from ATCC, CRL-1446, LGC Standards GmbH Wesel, Germany. Cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with $10\%$ fetal bovine serum and $1\%$ penicillin-streptomycin at 37 °C in a humidified incubator consisting of $5\%$ CO2 and $95\%$ air. Cells were fed every 3 days, and cells were passaged by trypsinization when reaching 70–$80\%$ confluence. The passage number of the cells was between 8 and 26. The cells were pretreated with 50 μM BGP-15 for 24 h and then treated with DOX at indicated concentrations for 12 or 24 h. The stock solution of 2 mg/mL DOX was diluted weekly, and the BGP-15 solution was prepared before treatment in the medium with the composition mentioned above. ## 4.3. Cell Viability Assay by MTT In this assay, cells were seeded into 96-well culture plates with 3000 cells/well, pretreated with 50 μM BGP-15 for 24 h, and treated with different doses (0.1; 1; 3 μM) of DOX for 12 or 24 h. After treatment, MTT solution (final concentration of 0.5 mg/mL) was added to each well and incubated for 3 h at 37 °C. After that, the medium was replaced by isopropyl alcohol to dissolve the formazan product. Absorbance was measured with a Multiskan GO Microplate Spectrophotometer (Thermo Fisher Scientific Oy, Ratastie, Finland) at 570 and 690 nm. The values were calculated as follows: the resulting colored solution is quantified by measuring absorbance at 570 nm and subtracting background absorbance at 690 nm. These values were expressed relative to the control, which was represented as $100\%$ of viability. One percent H2O2 was used as the positive control. Absorbance values were averaged across 6 replicate wells and repeated 6–9 times. ## 4.4. Determination of Intracellular Reactive Oxygen Species Generation and Mitochondrial Function In this experiment, 2000 cells/well were seeded on round glass coverslips placed into 24-well plates. Cells were treated with different doses (0.1; 1; 3 μM) of DOX for 24 h in the presence or absence of BGP-15 pretreatment (50 μM). At the end of the treatment, the medium was removed and cells were washed 3 times with Hank’s Balanced Salt Solution (HBSS). MitoSOX™ Red was added for 10 min at 37 °C in the dark. The nucleus was stained by DAPI. Finally, cells were fixed with $4\%$ methanol-free formaldehyde and washed with HBSS, and coverslips were placed on a slide. Specimens were visualized using a fluorescence microscope. Images were captured by a Zeiss Axio Scope. A1 fluorescent microscope and analyzed with ZEN 2011 v.1.0.1.0. Software (Carl Zeiss Microscopy GmbH, München, Germany). The images were captured using the 63× oil immersion objective lens. For flow cytometry experiments, 20000 cells/well were seeded into 24-well plates, and the same protocol was carried out. Cells were trypsinized and fixed with $4\%$ methanol-free formaldehyde. Cellular fluorescence was analyzed by a Guava Easy Cyte 6HT-2L flow cytometer (Merck Ltd., Darmstadt, Germany). MitoSOX Red was analyzed by using 510 nm excitation and 580 nm emission wavelengths. Using flow cytometry of H9c2 stained cells with and without MitoSOX Red, we were able to separate the red fluorescence signal elicited by DOX. ## 4.5. Assessment of Mitochondrial Membrane Potential Mitochondrial membrane potential (MMP) was assessed using the fluorescent indicator 5,5’,6,6’-tetrachloro-1,1’3,3’-tetraethylbenzimidazolocarbo-cyanine iodide (JC-1; from Life Technologies (Paisley, Scotland)). Cells were seeded into black 96-well culture plates with 3000 cells/well and 24-well culture plates with coveslips (2000 cells/well), then non-pretreated/pretreated with 50 μM BGP-15, and treated with different doses (0.1; 1; 3 μM) of DOX for 24 h. After treatment, cells were incubated with 1 mg/mL JC-1 in Krebs–Henseleit buffer for 30 min at 37 °C. After incubation time, cells were washed once with Krebs–Henseleit buffer. Red and green fluorescence intensities of the samples were measured with a Multiskan GO Microplate Spectrophotometer (Thermo Fisher Scientific Oy, Ratastie, Finland) at 492 nm excitation and 520 and 590 nm emission wavelengths. DAPI as a nuclear stain was measured at 365 nm excitation and 445 emission wavelengths. The ratio of red and green fluorescence values was normalized to the blue fluorescence values. Using a spectrophotometer on H9c2-stained cells with and without JC-1 we were able to separate the red fluorescence signal elicited by DOX. Absorbance values were averaged across 4 replicate wells and repeated 5 times. The coverslips were placed on a slide and visualized using a fluorescence microscope. Images were captured by Zeiss Axio Scope. A1 fluorescent microscope using the 63× oil immersion objective lens and analyzed with ZEN 2011 v.1.0.1.0. Software (Carl Zeiss Microscopy GmbH, München, Germany). A shift from red to green fluorescence indicates a loss of MMP, which was assessed by obtaining multiple merged images. ## 4.6. LDH (Lactate Dehydrogenase) Release Assay LDH release was measured by the LDH-cytotoxicity assay kit (Sigma, St. Louis, MO, USA) according to the manufacturer’s instructions. Cells were seeded into 96-well culture plates with 5000 cells/well, pretreated with 50 μM BGP-15, and treated with different doses (0.1; 1; 3 μM) of DOX for 24 h. Absorbance was measured with a Multiskan GO Microplate Spectrophotometer (Thermo Fisher Scientific Oy, Ratastie, Finland) at 492 and 620 nm. The values were expressed relative to the positive control ($2\%$ TritonX-100 in assay medium), which was represented as maximal LDH release. Absorbance values were averaged across 8 replicate wells and repeated 5 times. ## 4.7. Autophagy Flux Determined by Fluorescent Microscopy For the analyses of autophagy flux, we used Lysotracker Red, LC3B, and p62 antibodies. For these experiments, 2000 cells/well were seeded on round glass coverslips placed into 24-well culture plates. The treatment protocol was the following: with or without pretreatment with 50 μM BGP-15 and treated with different doses (1 μM) of DOX for 24 h, Rapamycin (5 mM) was used as the positive control, and the autophagic process was inhibited by chloroquine (10 mM, for 18 h). After treatments, the medium was removed, and cells were washed 3 times with HBSS. Lysotracker Red was added for 30 min at 37 °C in the dark. Cells were fixed with $4\%$ methanol-free formaldehyde. Cells on coverlips were permeabilized and blocked with HBSS containing $5\%$ normal goat serum and $0.3\%$ TritonX-100 for 30 min. Thereafter, the cells were incubated with primary antibodies (LC3B or p62: 1:1000 with $1\%$ BSA and $0.3\%$ TritonX-100 in HBSS) for 2 h at 37 °C and incubated with a secondary antibody (Alexa Flour 488 goat anti-rabbit IgG (H + L) 1:500 with $0.2\%$ BSA in HBSS) for 1 h at 37 °C in dark. The nucleus was stained by DAPI. The cells were washed with HBSS after each step. The coverslips were placed on a slide and visualized using a fluorescence microscope. Images were captured by a Zeiss Axio Scope. A1 fluorescent microscope and analyzed with ZEN 2011 v.1.0.1.0. Software (Carl Zeiss Microscopy GmbH, München, Germany). The images were captured using the 63× oil immersion objective lens. ## 4.8. Protein Isolation After treatment, total protein fractions were extracted from the cultured H9C2 cells based on the previously described [33]. Afterward, the isolation protein concentration was determined using a BCA kit (Thermo Scientific, Rockford, IL, USA). ## 4.9. Western Blot Analysis A 25 μg protein sample was loaded and separated in 4–$20\%$ Mini-PROTEAN® TGX Stain-Free™ Protein gel. Then, gels were exposed to UV light, and thereby trihalo compounds contained in stain-free gels covalently bind to tryptophan residues in proteins, allowing total protein quantification, and were transferred onto PVDF membranes for 1 h at 100 V. Membranes were exposed by another brief irradiation, the resulting fluorescence signals were recorded, and the signal intensity was considered proportional to the total protein volume. After blocking with $5\%$ of non-fat dry milk in Tris Buffered Saline with Tween 20 (TBST), membranes were incubated with primary antibody solution (LC3B, p62 and Caspase-3: 1:1000 in TBST) at 4 °C overnight. The membranes were washed with TBST and incubated with HRP-conjugated secondary antibody solution (1:3000 in TBST). After washing, the membranes were incubated with Clarity Western ECL substrate (Bio-Rad Laboratories) for visualization by enhanced chemiluminescence bands according to the recommended procedure (ChemiDoc Touch, Bio-Rad Laboratories). The chemiluminescent bands and each total protein lane intensity were measured by Image Lab software (version 5.2.1) (Bio-Rad Laboratories). During quantification, protein density is measured directly on the membranes and reflected in total loaded proteins. Thus, this type of normalization eliminates the need to select housekeeping proteins. The software calculates the normalization factor, which is the total volume (intensity) of the stain-free reference lane/total lane stain-free (intensity) of each lane. The protein expression was quantified by normalized volume, which means normalization factor x volume (intensity) [34]. ## 4.10. Statistical Analysis The data were expressed as mean ± SEM. Statistical analyses were performed with GraphPad Prism version 5 (La Jolla, CA, USA). The one-way analysis of variance (ANOVA) test was followed by Tukey’s multiple comparison tests, which identified the significant difference between control and treated groups, and the Šidák method was used to compare the treated groups (MitoSOX assay). A probability value of $p \leq 0.05$ was used as the criterion for statistical significance. Significant ($p \leq 0.05$), , , , and represent $p \leq 0.05$, $p \leq 0.01$, $p \leq 0.001$, and $p \leq 0.0001$ in the Tukey’s post-test, respectively. ## References 1. Carvalho C., Santos R.X., Cardoso S., Correia S., Oliveira P.J., Santos M.S., Moreira P.I.. Doxorubicin: The good, the bad and the ugly effect*. Curr. Med. Chem.* (2009) 16 3267-3285. DOI: 10.2174/092986709788803312 2. Carvalho F.S., Burgeiro A., Garcia R., Moreno A.J., Carvalho R.A., Oliveira P.J.. Doxorubicin-induced cardiotoxicity: From bioenergetic failure and cell death to cardiomyopathy. Med. Res. Rev. (2014) 34 106-135. DOI: 10.1002/med.21280 3. Rawat P.S., Jaiswal A., Khurana A., Bhatti J.S., Navik U.. Doxorubicin-induced cardiotoxicity: An update on the molecular mechanism and novel therapeutic strategies for effective management. Biomed. Pharmacother. (2021) 139 111708. DOI: 10.1016/j.biopha.2021.111708 4. Wenningmann N., Knapp M., Ande A., Vaidya T.R., Ait-Oudhia S.. Insights into Doxorubicin-induced Cardiotoxicity: Molecular Mechanisms, Preventive Strategies, and Early Monitoring. Mol. Pharmacol. (2019) 96 219-232. DOI: 10.1124/mol.119.115725 5. Rabinovich-Nikitin I., Love M., Kirshenbaum L.A.. Inhibition of MMP prevents doxorubicin-induced cardiotoxicity by attenuating cardiac intracellular and extracellular matrix remodelling. Cardiovasc. Res. (2021) 117 11-12. DOI: 10.1093/cvr/cvaa198 6. Zhao L., Zhang B.. Doxorubicin induces cardiotoxicity through upregulation of death receptors mediated apoptosis in cardiomyocytes. Sci. Rep. (2017) 7 44735. DOI: 10.1038/srep44735 7. Lu L., Wu W., Yan J., Li X., Yu H., Yu X.. Adriamycin-induced autophagic cardiomyocyte death plays a pathogenic role in a rat model of heart failure. Int. J. Cardiol. (2009) 134 82-90. DOI: 10.1016/j.ijcard.2008.01.043 8. Dirks-Naylor A.J.. The role of autophagy in doxorubicin-induced cardiotoxicity. Life Sci (2013) 93 913-916. DOI: 10.1016/j.lfs.2013.10.013 9. Nikoletopoulou V., Markaki M., Palikaras K., Tavernarakis N.. Crosstalk between apoptosis, necrosis and autophagy. Biochim. Biophys. Acta (2013) 1833 3448-3459. DOI: 10.1016/j.bbamcr.2013.06.001 10. Christidi E., Brunham L.R.. Regulated cell death pathways in doxorubicin-induced cardiotoxicity. Cell Death Dis. (2021) 12 339. DOI: 10.1038/s41419-021-03614-x 11. Klionsky D.J., Abdelmohsen K., Abe A., Abedin M.J., Abeliovich H., Acevedo Arozena A., Adachi H., Adams C.M., Adams P.D., Adeli K.. Guidelines for the use and interpretation of assays for monitoring autophagy (3rd edition). Autophagy (2016) 12 1-222. DOI: 10.1080/15548627.2015.1100356 12. Peto A., Kosa D., Feher P., Ujhelyi Z., Sinka D., Vecsernyes M., Szilvassy Z., Juhasz B., Csanadi Z., Vigh L.. Pharmacological Overview of the BGP-15 Chemical Agent as a New Drug Candidate for the Treatment of Symptoms of Metabolic Syndrome. Molecules (2020) 25. DOI: 10.3390/molecules25020429 13. Nagy G., Szarka A., Lotz G., Doczi J., Wunderlich L., Kiss A., Jemnitz K., Veres Z., Banhegyi G., Schaff Z.. BGP-15 inhibits caspase-independent programmed cell death in acetaminophen-induced liver injury. Toxicol. Appl. Pharmacol. (2010) 243 96-103. DOI: 10.1016/j.taap.2009.11.017 14. Szabados E., Literati-Nagy P., Farkas B., Sumegi B.. BGP-15, a nicotinic amidoxime derivate protecting heart from ischemia reperfusion injury through modulation of poly(ADP-ribose) polymerase. Biochem. Pharmacol. (2000) 59 937-945. DOI: 10.1016/S0006-2952(99)00418-9 15. Bombicz M., Priksz D., Gesztelyi R., Kiss R., Hollos N., Varga B., Nemeth J., Toth A., Papp Z., Szilvassy Z.. The Drug Candidate BGP-15 Delays the Onset of Diastolic Dysfunction in the Goto-Kakizaki Rat Model of Diabetic Cardiomyopathy. Molecules (2019) 24. DOI: 10.3390/molecules24030586 16. Zorov D.B., Juhaszova M., Sollott S.J.. Mitochondrial reactive oxygen species (ROS) and ROS-induced ROS release. Physiol. Rev. (2014) 94 909-950. DOI: 10.1152/physrev.00026.2013 17. Wallace K.B., Sardao V.A., Oliveira P.J.. Mitochondrial Determinants of Doxorubicin-Induced Cardiomyopathy. Circ. Res. (2020) 126 926-941. DOI: 10.1161/CIRCRESAHA.119.314681 18. Sapra G., Tham Y.K., Cemerlang N., Matsumoto A., Kiriazis H., Bernardo B.C., Henstridge D.C., Ooi J.Y., Pretorius L., Boey E.J.. The small-molecule BGP-15 protects against heart failure and atrial fibrillation in mice. Nat. Commun. (2014) 5 5705. DOI: 10.1038/ncomms6705 19. Horvath O., Ordog K., Bruszt K., Deres L., Gallyas F., Sumegi B., Toth K., Halmosi R.. BGP-15 Protects against Heart Failure by Enhanced Mitochondrial Biogenesis and Decreased Fibrotic Remodelling in Spontaneously Hypertensive Rats. Oxidative Med. Cell. Longev. (2021) 2021 1250858. DOI: 10.1155/2021/1250858 20. Zamorano J.L., Lancellotti P., Rodriguez Munoz D., Aboyans V., Asteggiano R., Galderisi M., Habib G., Lenihan D.J., Lip G.Y.H., Lyon A.R.. 2016 ESC Position Paper on cancer treatments and cardiovascular toxicity developed under the auspices of the ESC Committee for Practice Guidelines: The Task Force for cancer treatments and cardiovascular toxicity of the European Society of Cardiology (ESC). Eur. Heart J. (2016) 37 2768-2801. DOI: 10.1093/eurheartj/ehw211 21. Varga Z.V., Ferdinandy P., Liaudet L., Pacher P.. Drug-induced mitochondrial dysfunction and cardiotoxicity. Am. J. Physiol.-Heart Circ. Physiol. (2015) 309 H1453-H1467. DOI: 10.1152/ajpheart.00554.2015 22. McGowan J.V., Chung R., Maulik A., Piotrowska I., Walker J.M., Yellon D.M.. Anthracycline Chemotherapy and Cardiotoxicity. Cardiovasc. Drugs Ther. (2017) 31 63-75. DOI: 10.1007/s10557-016-6711-0 23. Angsutararux P., Luanpitpong S., Issaragrisil S.. Chemotherapy-Induced Cardiotoxicity: Overview of the Roles of Oxidative Stress. Oxidative Med. Cell. Longev. (2015) 2015 795602. DOI: 10.1155/2015/795602 24. Ichikawa Y., Ghanefar M., Bayeva M., Wu R., Khechaduri A., Naga Prasad S.V., Mutharasan R.K., Naik T.J., Ardehali H.. Cardiotoxicity of doxorubicin is mediated through mitochondrial iron accumulation. J. Clin. Investig. (2014) 124 617-630. DOI: 10.1172/JCI72931 25. Montaigne D., Marechal X., Preau S., Baccouch R., Modine T., Fayad G., Lancel S., Neviere R.. Doxorubicin induces mitochondrial permeability transition and contractile dysfunction in the human myocardium. Mitochondrion (2011) 11 22-26. DOI: 10.1016/j.mito.2010.06.001 26. Nordgren K.K., Wallace K.B.. Keap1 redox-dependent regulation of doxorubicin-induced oxidative stress response in cardiac myoblasts. Toxicol. Appl. Pharmacol. (2014) 274 107-116. DOI: 10.1016/j.taap.2013.10.023 27. Sumegi K., Fekete K., Antus C., Debreceni B., Hocsak E., Gallyas F., Sumegi B., Szabo A.. BGP-15 Protects against Oxidative Stress- or Lipopolysaccharide-Induced Mitochondrial Destabilization and Reduces Mitochondrial Production of Reactive Oxygen Species. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0169372 28. Ray P.D., Huang B.W., Tsuji Y.. Reactive oxygen species (ROS) homeostasis and redox regulation in cellular signaling. Cell. Signal. (2012) 24 981-990. DOI: 10.1016/j.cellsig.2012.01.008 29. Zilinyi R., Czompa A., Czegledi A., Gajtko A., Pituk D., Lekli I., Tosaki A.. The Cardioprotective Effect of Metformin in Doxorubicin-Induced Cardiotoxicity: The Role of Autophagy. Molecules (2018) 23. DOI: 10.3390/molecules23051184 30. Minotti G., Menna P., Salvatorelli E., Cairo G., Gianni L.. Anthracyclines: Molecular advances and pharmacologic developments in antitumor activity and cardiotoxicity. Pharmacol. Rev. (2004) 56 185-229. DOI: 10.1124/pr.56.2.6 31. Li D.L., Wang Z.V., Ding G., Tan W., Luo X., Criollo A., Xie M., Jiang N., May H., Kyrychenko V.. Doxorubicin Blocks Cardiomyocyte Autophagic Flux by Inhibiting Lysosome Acidification. Circulation (2016) 133 1668-1687. DOI: 10.1161/CIRCULATIONAHA.115.017443 32. Kawai A., Uchiyama H., Takano S., Nakamura N., Ohkuma S.. Autophagosome-lysosome fusion depends on the pH in acidic compartments in CHO cells. Autophagy (2007) 3 154-157. DOI: 10.4161/auto.3634 33. Gyongyosi A., Szoke K., Fenyvesi F., Fejes Z., Debreceni I.B., Nagy B., Tosaki A., Lekli I.. Inhibited autophagy may contribute to heme toxicity in cardiomyoblast cells. Biochem. Biophys. Res. Commun. (2019) 511 732-738. DOI: 10.1016/j.bbrc.2019.02.140 34. Gurtler A., Kunz N., Gomolka M., Hornhardt S., Friedl A.A., McDonald K., Kohn J.E., Posch A.. Stain-Free technology as a normalization tool in Western blot analysis. Anal. Biochem. (2013) 433 105-111. DOI: 10.1016/j.ab.2012.10.010
--- title: 'BabyByte: Qualitative Research to Inform the Development of an App to Improve Responsive Feeding Practices in Parents of Infants and Toddlers' authors: - Amy R. Mobley - Danielle E. Jake-Schoffman - David A. Fedele - Elder Garcia Varela - Jamie Zeldman journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049234 doi: 10.3390/ijerph20064769 license: CC BY 4.0 --- # BabyByte: Qualitative Research to Inform the Development of an App to Improve Responsive Feeding Practices in Parents of Infants and Toddlers ## Abstract Responsive feeding is associated with a reduced risk of childhood obesity. The objective of this qualitative study was to determine parental preferences for mobile health (mHealth) app content and features designed to improve responsive feeding practices. Parents of 0–2-year-old children were interviewed individually. Interview questions were informed by the Technology Acceptance Model, and parents provided feedback on sample app content and features. Interviews were audio-recorded, transcribed verbatim, and coded by two researchers using thematic analysis; responses were compared by parent gender and income. Parents ($$n = 20$$ fathers, $$n = 20$$ mothers) were, on average, 33 years old, low-income ($50\%$), identified as non-white ($52.5\%$), and had a bachelor’s degree or higher ($62\%$). Overall, parents were most interested in feeding tips and recipe content, and app features that allowed tracking child growth and setting feeding goals. Fathers were most interested in content about first foods, choking hazards, and nutrition information, while mothers preferred content on breastfeeding, picky eating, and portion sizes. Parents with lower incomes were interested in nutrition guidelines, breastfeeding, and introducing solids. Non-low-income parents preferred information related to food allergies, portion sizes, and picky eating. The findings of this study provide considerations when developing mHealth apps to improve responsive feeding practices in parents. ## 1. Introduction The first 1000 days (conception to age 2) are critical for obesity prevention [1], yet, there are few effective, sustainable interventions designed for this time period [2]. In fact, one out of four children are already overweight or obese by five years of age [3]. These rates disproportionally affect children from lower-income families [4], continue to increase sharply into adulthood, and are associated with lifelong health-related comorbidities, including cardiovascular disease risk factors [5,6,7,8]. There is a pressing need for interventions that reach large numbers of at-risk children ages 0–2 years because parents [9] and healthcare providers [10] may fail to adequately address early obesity risk. Children’s food preferences and eating patterns are established at a young age [11], and these patterns can affect risk factors for obesity and chronic diseases later in life [12]. Unfortunately, many infants and toddlers are fed in ways that predispose them to obesity, such as eating inadequate fruit, vegetables, and iron-rich foods and receiving weaning foods before 4–6 months of age [13]. However, despite the public health need to prevent childhood obesity, there is minimal and varying outreach to parents regarding early optimal feeding practices, and the messages that are conveyed are often inconsistent (e.g., when to introduce first foods) [14]. Further, while mothers are often the main target for childhood obesity prevention interventions, it is recognized that fathers or other family members in the household may influence target behaviors related to early childhood obesity risk [15,16,17,18,19]. Further research is needed to determine the best methods for engaging fathers in family interventions focused on nutrition and health [20]. Therefore, both mothers and fathers are important targets for early childhood obesity prevention outreach. In addition, because obesity disproportionally affects children from families with lower incomes, it is important to consider the different preferences and needs of families with lower incomes compared to higher incomes. However, the likelihood of individuals with lower incomes using digital health interventions to improve or support health or dietary behaviors may be lower than individuals with higher incomes [21]. Alternatively, another reason why individuals with lower incomes may be less likely to use digital health technology may be attributed to a real or perceived “digital divide”, although more recent evidence indicates that $76\%$ of lower SES adults report owning a smartphone. However, this proportion is lower compared to middle- ($87\%$) or higher-income adults ($97\%$) [22]. Further inquiry is needed to determine interest in and preferences by parents with lower incomes in using digital technology, such as a child feeding app. The relationship between parental feeding practices, feeding styles, and children’s dietary intake and weight status has propelled interest in responsive feeding to prevent childhood obesity [23]. Responsive feeding is the process of recognizing a child’s hunger and fullness cues and responding appropriately. In turn, this encourages self-regulation and supports cognitive, emotional, and social development in young children [24]. Implementing responsive feeding from birth helps children maintain their ability to recognize and respond to their hunger and satiety cues [25], thereby allowing better regulation of dietary intake based on physiological needs and decreasing excessive caloric intake and the risk of obesity. Indeed, responsive feeding interventions for parents with infants can curb the development of early childhood obesity [26]. When considering intervention strategies, mobile health (mHealth) may be an effective way to promote behavior change related to nutrition and dietary practices using nutrition messaging [27]. To date, however, commercial digital technology tools for children’s healthy eating lack evidence-based content related to dietary and feeding guidance [28], and there is a dearth of tools specifically for feeding young children. Current conflicting messages about child feeding may contribute to confusion among parents and caregivers about what and when to feed their young children and, in turn, lead to inappropriate feeding practices increasing obesity risk [14]. Although there is limited research on mobile phone-delivered information to parents on topics related to child feeding or nutrition, the resources available contain information for breastfeeding or preschool-aged children and do not typically include feeding recommendations regarding the crucial developmental ages in the first 1000 days, including from 0–2 years [29]. Specifically, a systematic assessment of infant feeding apps revealed that although the quality of available apps has improved in the last decade, the credibility and accuracy of content have not improved, with almost half of the available apps containing poor quality, inaccurate, or inadequate information [30]. Innovative approaches to delivering comprehensive and evidence-based feeding messages to parents and caregivers of young children are critically needed. The objective of this qualitative study was to determine parents’ preferred content and features for a mHealth app designed to improve responsive feeding practices. Differences between mothers’ and fathers’ preferred content and features (e.g., goal setting) and income status (low-income vs. non-low-income) were also evaluated. ## 2.1. Participant Eligibility, Recruitment, and Data Collection The study was approved as exempt by the University of Florida (UF) Institutional Review Board for Human Subjects. Using study flyers and email, parents were recruited through local childcare centers, early Head Start, and community locations serving families with young children. We also partnered with UF HealthStreet, a community engagement program at UF designed to navigate community members to ongoing research projects. We reached out to families who agreed to be contacted for research studies via UF’s Consent2Share program, and we also posted advertisements on social media sites. Parents were eligible if they were at least 18 years of age, able to speak and read in English, had at least one child between 0–2 years old, were responsible for feeding their child at least $50\%$ of the time, and owned or recently owned a smartphone within the last 6 months. Purposeful recruitment was used to include a convenience sample with equal numbers of mothers and fathers and low-income and non-low-income parents. Interviews were conducted until data saturation was reached. Low income was defined as being eligible for certain state or federal nutrition assistance or related programs in the United States (e.g., Supplemental Nutrition Assistance Program) that have income eligibility criteria. After obtaining written informed consent, parents participated in one-on-one semi-structured interviews lasting 45–60 min with a trained researcher and completed a separate demographics and technology use questionnaire. A $25 gift card incentive was provided immediately after the interview. Interviews were guided by a trained researcher using a semi-structured interview script (File S1, Supplementary Materials) and sample mobile phone screenshots that were shown to the participant on a mobile phone that could be included in a future child feeding app. Questions were based on the Technology Acceptance Model, which draws upon constructs such as perceived usefulness and ease of use, attitudes, and behavioral intentions to predict the actual use of health-related technology [31]. Additional constructs derived from the Theory of Planned Behavior (TPB) and complementary to the TAM were also applied in developing questions within the interview script about the proposed app features and ideas in general to determine perceived usefulness, ease of use, social attitude/subjective norms and behavioral intention for using a child feeding app. ## 2.2. Child Feeding App Development BabyByte, the name of the proposed mHealth app, was developed using Marvel online prototyping software (marvelapp.com, London, UK, accessed on 5 March 2023) to display potential app features and content on a mobile phone screen during the interview. It was designed to appear as screens from an interactive app on iOS and Android operating systems but was not able to capture direct user input or data. Instead, it served as a visual guide to identify preferred topics and features by parents. The BabyByte sample app screens contained a main menu with proposed feature options: Goal Setting (allows parents to set goals related to child feeding, such as how long a mother may plan to breastfeed or when a parent plans to introduce solid foods); Tracking Baby’s Growth and Milestones (a tool to enter child measures to track growth and also track feeding milestones such as when a child is able to drink from a cup); Parent Discussion Board (an opportunity to connect with other parents with similar feeding questions or concerns); Ask the Expert (connection to a registered dietitian or health professional to answer questions related to child feeding); Surveys & Quizzes (a tool to reinforce content learned and reward parents for submitting and completing questions); Videos (reinforcement of content information and how-to on topics such as introducing or preparing foods, etc.); and More Information (additional electronic resources such as websites and local programs). The features in the app were designed to support behavior change and optimal responsive feeding practices through goal setting, monitoring, social support, reinforcement, and improved self-efficacy. In addition, a user account example screen and content sections containing feeding topics for three main age groups of children were included as 0–6 months, 6–12 months, and 12–24 months. Age groups were organized based on optimal developmental feeding milestones. Content for the app prototype was based on early childhood obesity prevention information and related features based on priority topics identified by the target population from prior research conducted by the corresponding author [32,33,34] and key strategies from the Robert Wood Johnson Foundation’s (RWJF) Feeding Guidelines for Infants and Young Toddlers: A Responsive Parenting Approach [24]. These topics included when and how to introduce solid foods, how to safely introduce allergenic foods, how to encourage autonomy and self-regulation, and how to increase the variety and healthy food options in a young child’s diet. ## 2.3. Data Analysis Interviews were audio-recorded and transcribed verbatim. Transcripts were coded by two research team members using standard word processing software (Microsoft Word) for data extraction and a thematic analysis approach [35]. The research team coded the transcripts independently and met to compare similarities and discrepancies in the coding to reach a consensus. Qualitative crosstab analysis was used to compare responses by gender and income status. Major themes for the overall group of parents and subthemes by parent gender and income were then summarized in descending order of how often mentioned. Verbatim sample quotes from fathers (FA) or mothers (MO) and low-income (L) or non-low-income parents (NL) were labeled accordingly. Descriptive statistics for demographics and responses to technology use questions were also summarized using IBM Corp. Released 2021. IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY, USA: IBM Corp. ## 3.1. Demographic and Technology Use Characteristics Parents ($$n = 20$$ fathers, $$n = 20$$ mothers) were, on average, 33 years old, with more than half who were non-white ($$n = 21$$, $52.5\%$), non-Hispanic ($$n = 39$$, $97.5\%$), had a bachelor’s degree or higher education ($$n = 25$$, $62\%$), were employed full-time ($$n = 24$$, $60\%$), and were married ($$n = 31$$, $77.5\%$) (Table 1). Half of the participants ($$n = 20$$, $50\%$) were considered low-income, as indicated by self-reported eligibility for federal or local nutrition assistance or related programs. When assessing technology use, most parents reported they own an Apple ($$n = 20$$, $50\%$) or Android ($$n = 19$$, $47.5\%$) smartphone, had an unlimited data plan ($$n = 25$$, $67.6\%$), used apps on their phone more than once per day ($$n = 29$$, $72.5\%$), and searched for health information on their phone at least a few times per week ($$n = 26$$, $65\%$) (Table 1). When considering the feasibility of using an app to support responsive feeding, the majority ($$n = 34$$, $85\%$) of parents indicated that they would be likely or very likely to use it (rating it a 4 or 5 on a scale of 1 to 5). ## 3.2. Qualitative Results Overarching themes resulting from the interviews with parents are summarized in Table 2, with subthemes outlined by parent gender and income comparisons. Most parents, irrespective of gender and income, expressed positive attitudes towards using a child feeding app like BabyByte and indicated it as useful and easy to use and that they would be likely or very likely to use it. Specifically, parents perceived app content options related to feeding tips, recipes, and food allergy guidance and the tracking growth and milestones and setting feeding goals features as most helpful or important to include in a child feeding app, with very few comments regarding features that were not helpful. As one mother shared, “The food tips, and recipes. Because that can help me include some things that I didn’t know, in my… food routine.” ( MO-28-L). Most parents also indicated that they were comfortable using all the menu options. When considering social attitudes or norms, parents indicated that the overall information provided in the app would be beneficial to others. The app’s convenience, reliability, and user-friendliness were also noted as benefits. For example, one mother shared, “To me it’s such wonderful information at your fingertips. It’s just right there, and it’s a good resource to be able to keep track of their eating and their health.” ( MO-13-NL) Additionally, one father noted, “Knowing that you have credible, good current information. It’s not just like a Google search. It’s more like a polished professional type of app.” ( FA-03-NL) The option to communicate with other parents was also a benefit mentioned by fathers, “I think getting connected to a network of other parents can be a huge benefit for those looking for others to communicate about what are you going through, how’s feeding your child doing, what kind of tips do you have.” ( FA-10-L). When comparing fathers’ and mothers’ preferences for app content, fathers indicated that they were most interested in information about first foods, choking hazards, nutrition information, and food allergies. For instance, one father shared that food allergies were a topic he was most interested in within the app, “Allergies that a baby can have from foods. Because they can get sick then you have reactions and stuff and you don’t want the baby have to go through that. So, the parents get informed about possible things that are big amongst their babies.” ( FA-26-L) In comparison, mothers preferred content on breastfeeding, picky eating, and portion sizes. One mother shared that she preferred the content about “…what you can do potentially early on to avoid things down the road of having more picky eating habits.” ( MO-18-NL) Another mother noted that she preferred information including “…how much to feed the baby, and all of that stuff. How much on a good day a child should eat, opposed to not overfeeding, and that kind of stuff to me, just matters.” ( MO-32-L). Both fathers and mothers favored the “Tracking Growth or Milestones” and “Setting Goals” features when queried about which feature of the app would be most useful or important to include in a child-feeding app. As one of the fathers shared, “Oh, the tracking my baby. That’s the most customizable thing. All the other stuff, if it’s a hassle, I can go and actually find those things. This is just for my baby.” ( FA-07-L) In addition, one of the mothers noted, “One of the things I was most excited about was… the tracking part of it and the milestones… that’s not something we spend a lot of time on.” ( MO-19-NL). When comparing parents’ preferences for app content by income status, parents with lower incomes were interested in information about nutrition guidelines, breastfeeding, and introducing solids. One low-income parent shared that she was interested in content related to “Definitely breastfeeding, I think should be at the highest because why not? Educate more people to know that breast milk, it’s been going on since forever. I think it’s very important.” ( MO-32-L) In comparison, parents who were not low-income tended to prefer information related to portion sizes and picky eating. “ Probably that all kids are going to be picky and that whole deal, pickiness and how to get your kid to eat something that is new. And then I like the portion video also talking about what they should be eating on a good day.” ( FA-40-NL). When considering the app features, parents with lower incomes perceived the feature “Setting Goals” as the most useful, while non-low-income parents favored “Eating Milestones.” One parent with a lower income noted that “Setting some goals is the most extremely helpful thing out of this whole app, because it allows you to basically set long-term… short-term goals. Things where you are hands on with your baby, where you’re constantly paying attention to everything that they’re doing, everything that in-taking, their growth, all this kind of stuff.” ( FA-09-L) While a non-low-income parent shared, “I like the eating milestones and having that in an app, I’m more inclined to do everything.” ( MO-23-NL). Overall, parent participants also indicated that they thought other parents would like the app, with one main caveat of the importance of ensuring that the information is complete and up to date. To encourage further or continued use of the app, parents indicated that the overall layout and design, convenient access, up-to-date content from a reputable and identifiable source, videos, and incentives for completing modules were important features or considerations. As one mother noted, “It definitely would probably be the only useful app. And if University of Florida made it, it’s going to be a really good app.” ( MO-22-L) Considering motivators to use the app, another mother added, “If there was a little incentive to use it, that’d be like, yes that’s my new favorite app.” ( MO-11-L). ## 4. Discussion Overall, parents expressed interest in using a child-feeding mobile phone app like BabyByte for accessing feeding-related guidance for children ages 0–2 years. With current parental confusion surrounding early childhood feeding [14,24] and reported nutrition education needs by parents [33,34] for early childhood obesity prevention, an app like BabyByte may be helpful as an intervention option. This is especially true for family members who may be more difficult to reach or less willing to commit to in-person meetings or sessions [36]. The app features most frequently mentioned as helpful or important by parents included “Goal Setting” and “Tracking Baby’s Growth or Milestones.” Not surprisingly, Goal Setting has been noted as one of the most commonly applied behavior change techniques of digital interventions, especially those targeting individuals with lower socioeconomic status [37]. In addition, if parents experience confusion related to appropriate child development or milestones, having further support and reassurance could be viewed as beneficial. The use of developmental milestones has been highlighted as an opportunity to reach parents for a range of behaviors [38]. It has been prior noted that almost one out of four parents worry that their child is behind on milestones [39]. While overall preferences for features and content within BabyByte were mostly consistent across parent type and income status, some minor differences in preferences emerged. Overall preferences for app features and content included topics such as breastfeeding, first foods, picky eating, and food allergies. Fathers were often interested in topics that related to safety, such as choking hazards and food allergies. This may reflect an instinctive role of a father to serve as a provider or “protector” of his child [40]. In comparison, mothers more frequently mentioned the importance of topics such as breastfeeding and picky eating, likely because it is a behavior in which they may be more engaged. However, emerging technologies and interventions target fathers on topics such as breastfeeding to encourage support for their partners [41]. There were some minor variations in content preferences between low-income and non-low-income parents, which is consistent with prior research assessing the nutrition education needs of parents in relation to child feeding [34]. In addition, there were minimal differences when comparing low-income and non-low-income parents’ preferences for the app features. While the lack of technology-based interventions and evaluation of acceptability for lower socioeconomic families has been acknowledged prior [42], these sentiments were not shared by parents with lower incomes in this study. When considering the use of technology by mothers and fathers related to infant feeding, researchers determined that more mothers used the internet when seeking child-feeding information compared to fathers [20]. In addition, while fewer fathers were using the internet to seek child health or feeding-related information compared to mothers [20], it is possible that online information is not currently targeted or tailored towards fathers and, therefore, less likely for fathers to seek or find this information [43]. It has been recommended to increase engagement with fathers in early obesity prevention, especially during the first 1000 days [15]. Some fathers in the study indicated that the app would be beneficial by providing the opportunity to connect or communicate with other parents. Other child development focus group reports have recommended this aspect of social support, especially for first-time parents and fathers who are often not the primary target for parent support groups [38]. For example, a prior pilot initiative established the feasibility of engaging fathers in a perinatal intervention providing an example of how digital technology can serve as a viable mechanism to target fathers [44]. The current study provides further evidence of interest, positive attitudes, and behavioral intentions from fathers of young children in using an app to improve responsive child-feeding practices. Overall, while prior studies have explored the use of technology for family interventions as it relates to child health, mHealth tools are less frequently used, if at all, and results have been mixed [36]. Thus, it is important to also evaluate and consider the different components of mHealth tools and how they support behavior change either individually or synergistically, such as through the use of the Multiphase Optimization Strategy (MOST) [45]. At a minimum, as it relates to child development, it has been suggested to provide clear and concise recommendations to parents without overwhelming them, offering “how-to” guidance, providing support as recommendations are implemented, and giving options and alternatives for diverse needs [38]. ## Limitations While the study has many strengths, such as the inclusion of both mothers and fathers and low-income parents, as well as providing qualitative considerations for using a child-feeding-related mobile phone app, there are some limitations. A convenience sample of parents recruited for the study was from one geographic region or state (Florida) of the USA and tended to have higher education levels than the general population; therefore, their perceptions may not be widely transferable, and the findings may not be generalizable. Data were self-reported, and questions were based on an initial app prototype that was not yet fully developed or functional. It is unknown how and if parents would actually use the app in their daily routine. ## 5. Conclusions The findings of this study are important when considering the development of future mHealth tools for parents of children ages 0–2 years to improve responsive feeding practices and prevent early childhood obesity. While parents of children ages 0–2 years were in favor of using an app to support child feeding, some nuanced differences emerged in terms of preferred content or features. Thus, having various features, topics, and personalization available could allow choices and tailoring when developing mHealth tools to promote and support optimal child-feeding behaviors for parents. Future research is needed to determine the usability and feasibility of a child feeding app by parents of children ages 0–2 years and to evaluate the impact on parental feeding behaviors and associated child health and nutrition outcomes. ## References 1. 1. World Health Organization Report of the Commission on Ending Childhood ObesityWHOGeneva, Switzerland2016. Report of the Commission on Ending Childhood Obesity (2016.0) 2. Salvy S.J., de la Haye K., Galama T., Goran M.I.. Home visitation programs: An untapped opportunity for the delivery of early childhood obesity prevention. Obes. Rev. (2017.0) 18 149-163. DOI: 10.1111/obr.12482 3. Skinner A., Ravanbakht S., Skelton J., Perrin E., Armstrong S.. Prevalence of Obesity and Severe Obesity in US Children, 1999–2016. Pediatrics (2018.0) 141 e20173459. DOI: 10.1542/peds.2017-3459 4. Pan L., Freedman D.S., Sharma A.J.. Trends in Obesity Among Participants Aged 2–4 Years in the Special Supplemental Nutrition Program for Women, Infants, and Children—United States, 2000–2014. Morb. Mortal. Wkly. Rep. (2016.0) 65 1256-1260. DOI: 10.15585/mmwr.mm6545a2 5. Freedman D.S., Khan L.K., Serdula M.K., Dietz W.H., Srinivasan S.R., Berenson G.S.. The relation of childhood BMI to adult adiposity: The Bogalusa Heart Study. Pediatrics (2005.0) 115 22-27. DOI: 10.1542/peds.2004-0220 6. Ward Z.J., Long M.W., Resch S.C., Giles C.M., Cradock A.L., Gortmaker S.L.. Simulation of Growth Trajectories of Childhood Obesity into Adulthood. N. Engl. J. Med. (2017.0) 377 2145-2153. DOI: 10.1056/NEJMoa1703860 7. Barlow S.E.. Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: Summary report. Pediatrics (2007.0) 120 S164-S192. DOI: 10.1542/peds.2007-2329C 8. Daniels S.R., Pratt C.A., Hayman L.L.. Reduction of risk for cardiovascular disease in children and adolescents. Circulation (2011.0) 124 1673-1686. DOI: 10.1161/CIRCULATIONAHA.110.016170 9. Zehle K., Wen L.M., Orr N., Rissel C.. “It’s not an issue at the moment”: A qualitative study of mothers about childhood obesity. MCN Am. J. Matern. Child Nurs. (2007.0) 32 36-41. PMID: 17308456 10. Camp N.L., Robert R.C., Dawes C.S., Kelly K.P.. Identification of Overweight and Obesity in Low-Income Minority Children by Pediatric Providers and Child Characteristics Associated With Underrecognition. J. Pediatr. Health Care (2019.0) 33 162-168. DOI: 10.1016/j.pedhc.2018.07.007 11. Mikkila V., Rasanen L., Raitakari O.T., Pietinen P., Viikari J.. Consistent dietary patterns identified from childhood to adulthood: The cardiovascular risk in Young Finns Study. Br. J. Nutr. (2005.0) 93 923-931. DOI: 10.1079/BJN20051418 12. Barker D.J., Osmond C., Kajantie E., Eriksson J.G.. Growth and chronic disease: Findings in the Helsinki Birth Cohort. Ann. Hum. Biol. (2009.0) 36 445-458. DOI: 10.1080/03014460902980295 13. Siega-Riz A.M., Deming D.M., Reidy K.C., Fox M.K., Condon E., Briefel R.R.. Food consumption patterns of infants and toddlers: Where are we now?. J. Am. Diet. Assoc. (2010.0) 110 S38-S51. DOI: 10.1016/j.jada.2010.09.001 14. Dwyer J.T., Butte N.F., Deming D.M., Siega-Riz A.M., Reidy K.C.. Feeding Infants and Toddlers Study 2008: Progress, continuing concerns, and implications. J. Am. Diet. Assoc. (2010.0) 110 S60-S67. DOI: 10.1016/j.jada.2010.09.003 15. Davison K.K., Gavarkovs A., McBride B., Kotelchuck M., Levy R., Taveras E.M.. Engaging Fathers in Early Obesity Prevention During the First 1000 Days: Policy, Systems, and Environmental Change Strategies. Obesity (2019.0) 27 525-533. DOI: 10.1002/oby.22395 16. Vollmer R.L., Adamsons K., Foster J.S., Mobley A.R.. Association of fathers’ feeding practices and feeding style on preschool age children’s diet quality, eating behavior and body mass index. Appetite (2015.0) 89 274-281. DOI: 10.1016/j.appet.2015.02.021 17. Vollmer R.L., Adamsons K., Foster J.S., Mobley A.R.. How Are Fathers’ Demographic Characteristics Related to Preschool-Age Children’s Weight and Obesity Risk Factors?. Ecol. Food Nutr. (2017.0) 56 381-392. DOI: 10.1080/03670244.2017.1343726 18. Vollmer R.L., Adamsons K., Gorin A., Foster J.S., Mobley A.R.. Investigating the Relationship of Body Mass Index, Diet Quality, and Physical Activity Level between Fathers and Their Preschool-Aged Children. J. Acad. Nutr. Diet. (2015.0) 115 919-926. DOI: 10.1016/j.jand.2014.12.003 19. Vollmer R.L., Mobley A.R.. Comparing Low-income Mothers’ and Fathers’ Concern for Young Children’s Weight. J. Pediatr. Nurs. (2017.0) 37 97-100. DOI: 10.1016/j.pedn.2017.08.017 20. Laws R., Walsh A.D., Hesketh K.D., Downing K.L., Kuswara K., Campbell K.J.. Differences Between Mothers and Fathers of Young Children in Their Use of the Internet to Support Healthy Family Lifestyle Behaviors: Cross-Sectional Study. J. Med. Internet Res. (2019.0) 21 e11454. DOI: 10.2196/11454 21. Latulippe K., Hamel C., Giroux D.. Social Health Inequalities and eHealth: A Literature Review With Qualitative Synthesis of Theoretical and Empirical Studies. J. Med. Internet Res. (2017.0) 19 e136. DOI: 10.2196/jmir.6731 22. Vogels E.A.. Digital Divide Persists Even as Americans with Lower Incomes Make Gains in Tech Adoption 23. Vollmer R.L., Mobley A.R.. Parenting styles, feeding styles, and their influence on child obesogenic behaviors and body weight. A review. Appetite (2013.0) 71 232-241. DOI: 10.1016/j.appet.2013.08.015 24. Perez-Escamilla R., Segura-Perez S., Lott M.. Feeding Guidelines for Infants and Young Toddlers: A Responsive Parenting Approach 25. Davison K.K., Birch L.L.. Childhood overweight: A contextual model and recommendations for future research. Obes. Rev. (2001.0) 2 159-171. DOI: 10.1046/j.1467-789x.2001.00036.x 26. Daniels L.A., Mallan K.M., Battistutta D., Nicholson J.M., Perry R., Magarey A.. Evaluation of an intervention to promote protective infant feeding practices to prevent childhood obesity: Outcomes of the NOURISH RCT at 14 months of age and 6 months post the first of two intervention modules. Int. J. Obes. (2012.0) 36 1292-1298. DOI: 10.1038/ijo.2012.96 27. West J.H., Belvedere L.M., Andreasen R., Frandsen C., Hall P.C., Crookston B.T.. Controlling Your “App”etite: How Diet and Nutrition-Related Mobile Apps Lead to Behavior Change. JMIR Mhealth Uhealth (2017.0) 5 e95. DOI: 10.2196/mhealth.7410 28. Schoffman D.E., Turner-McGrievy G., Jones S.J., Wilcox S.. Mobile apps for pediatric obesity prevention and treatment, healthy eating, and physical activity promotion: Just fun and games?. Transl. Behav. Med. (2013.0) 3 320-325. DOI: 10.1007/s13142-013-0206-3 29. Nystrom C.D., Sandin S., Henriksson P., Henriksson H., Trolle-Lagerros Y., Larsson C., Maddison R., Ortega F.B., Pomeroy J., Ruiz J.R.. Mobile-based intervention intended to stop obesity in preschool-aged children: The MINISTOP randomized controlled trial. Am. J. Clin. Nutr. (2017.0) 105 1327-1335. DOI: 10.3945/ajcn.116.150995 30. Cheng H., Tutt A., Llewellyn C., Size D., Jones J., Taki S., Rossiter C., Denney-Wilson E.. Content and Quality of Infant Feeding Smartphone Apps: Five-Year Update on a Systematic Search and Evaluation. JMIR Mhealth Uhealth (2020.0) 8 e17300. DOI: 10.2196/17300 31. Holden R.J., Karsh B.T.. The technology acceptance model: Its past and its future in health care. J. Biomed. Inform. (2010.0) 43 159-172. DOI: 10.1016/j.jbi.2009.07.002 32. Heller R.L., Mobley A.R.. Development and Qualitative Pretesting of Child Feeding and Obesity Prevention Messages for Parents of Infants and Toddlers. J. Acad. Nutr. Diet. (2021.0) 121 1528-1541. DOI: 10.1016/j.jand.2021.02.004 33. Heller R.L., Chiero J.D., Trout N., Mobley A.R.. A qualitative study of providers’ perceptions of parental feeding practices of infants and toddlers to prevent childhood obesity. BMC Public Health (2021.0) 21. DOI: 10.1186/s12889-021-11305-7 34. Heller R.L., Chiero J.D., Puglisi M., Mobley A.R.. Feeding Infants and Toddlers: A Qualitative Study to Determine Parental Education Needs. Child. Obes. (2019.0) 15 443-450. DOI: 10.1089/chi.2019.0024 35. Krueger R., Casey M.. Focus Groups: A Practical Guide for Applied Research (2000.0) 36. Doss B.D., Feinberg L.K., Rothman K., Roddy M.K., Comer J.S.. Using technology to enhance and expand interventions for couples and families: Conceptual and methodological considerations. J. Fam. Psychol. (2017.0) 31 983-993. DOI: 10.1037/fam0000349 37. Ronteltap A., Bukman A.J., Nagelhout G.E., Hermans R.C.J., Hosper K., Haveman-Nies A., Lupker R., Bolman C.A.W.. Digital health interventions to improve eating behaviour of people with a lower socioeconomic position: A scoping review of behaviour change techniques. BMC Nutr. (2022.0) 8. DOI: 10.1186/s40795-022-00635-3 38. Guzman L., Ramos-Olazagasti M., Miller E., Dym Bartlett J.. First-time Parents’Knowledge of Early Child Development: Focus Group Report (2018.0) 1-36 39. 39. C.S. Mott Children’s Hospital Michigan Medicine Milestones: How Parents Understand Child DevelopmentSusan Meister Child Health Evaluation and Research CenterAnn Arbor, MI, USA2021Volume 38. Milestones: How Parents Understand Child Development (2021.0) Volume 38 40. Bryan D.M.. To parent or provide? The effect of teh provider role on low-income men’s decisions about fatherhood and paternal engagement. Fathering (2013.0) 11 71-89. DOI: 10.3149/fth.1101.71 41. White B.K., Giglia R.C., Scott J.A., Burns S.K.. How New and Expecting Fathers Engage With an App-Based Online Forum: Qualitative Analysis. JMIR Mhealth Uhealth (2018.0) 6 e144. DOI: 10.2196/mhealth.9999 42. Hall C.M., Bierman K.L.. Technology-assisted Interventions for Parents of Young Children: Emerging Practices, Current Research, and Future Directions. Early Child. Res. Q. (2015.0) 33 21-32. DOI: 10.1016/j.ecresq.2015.05.003 43. He Y., Dworkin J.. Cyber dads: What are fathers doing online?. Fathering (2015.0) 13 219-230. DOI: 10.3149/fth.1303.219 44. Whooten R.C., Kwete G.M., Farrar-Muir H., Cournoyer R.N., Barth E.A., Kotelchuck M., Taveras E.M.. Engaging fathers in the first 1000 days to improve perinatal outcomes and prevent obesity: Rationale and design of the First Heroes randomized trial. Contemp. Clin. Trials (2021.0) 101 106253. DOI: 10.1016/j.cct.2020.106253 45. Collins L.M., Murphy S.A., Strecher V.. The multiphase optimization strategy (MOST) and the sequential multiple assignment randomized trial (SMART): New methods for more potent eHealth interventions. Am. J. Prev. Med. (2007.0) 32 S112-S118. DOI: 10.1016/j.amepre.2007.01.022
--- title: 'Correlation of Dengue and Meteorological Factors in Bangladesh: A Public Health Concern' authors: - Md. Aminul Islam - Mohammad Nayeem Hasan - Ananda Tiwari - Md. Abdul Wahid Raju - Fateha Jannat - Sarawut Sangkham - Mahaad Issa Shammas - Prabhakar Sharma - Prosun Bhattacharya - Manish Kumar journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049245 doi: 10.3390/ijerph20065152 license: CC BY 4.0 --- # Correlation of Dengue and Meteorological Factors in Bangladesh: A Public Health Concern ## Abstract Dengue virus (DENV) is an enveloped, single-stranded RNA virus, a member of the Flaviviridae family (which causes Dengue fever), and an arthropod-transmitted human viral infection. Bangladesh is well known for having some of Asia’s most vulnerable Dengue outbreaks, with climate change, its location, and it’s dense population serving as the main contributors. For speculation about DENV outbreak characteristics, it is crucial to determine how meteorological factors correlate with the number of cases. This study used five time series models to observe the trend and forecast Dengue cases. Current data-based research has also applied four statistical models to test the relationship between Dengue-positive cases and meteorological parameters. Datasets were used from NASA for meteorological parameters, and daily DENV cases were obtained from the Directorate General of Health Service (DGHS) open-access websites. During the study period, the mean of DENV cases was 882.26 ± 3993.18, ranging between a minimum of 0 to a maximum of 52,636 daily confirmed cases. The Spearman’s rank correlation coefficient between climatic variables and Dengue incidence indicated that no substantial relationship exists between daily Dengue cases and wind speed, temperature, and surface pressure (Spearman’s rho; r = −0.007, $p \leq 0.05$; $r = 0.085$, $p \leq 0.05$; and r = −0.086, $p \leq 0.05$, respectively). Still, a significant relationship exists between daily Dengue cases and dew point, relative humidity, and rainfall ($r = 0.158$, $p \leq 0.05$; $r = 0.175$, $p \leq 0.05$; and $r = 0.138$, $p \leq 0.05$, respectively). Using the ARIMAX and GA models, the relationship for Dengue cases with wind speed is −666.50 [$95\%$ CI: −1711.86 to 378.86] and −953.05 [−2403.46 to 497.36], respectively. A similar negative relation between Dengue cases and wind speed was also determined in the GLM model (IRR = 0.98). Dew point and surface pressure also represented a negative correlation in both ARIMAX and GA models, respectively, but the GLM model showed a positive association. Additionally, temperature and relative humidity showed a positive correlation with Dengue cases (105.71 and 57.39, respectively, in the ARIMAX, 633.86, and 200.03 in the GA model). In contrast, both temperature and relative humidity showed negative relation with Dengue cases in the GLM model. In the Poisson regression model, windspeed has a substantial significant negative connection with Dengue cases in all seasons. Temperature and rainfall are significantly and positively associated with Dengue cases in all seasons. The association between meteorological factors and recent outbreak data is the first study where we are aware of the use of maximum time series models in Bangladesh. Taking comprehensive measures against DENV outbreaks in the future can be possible through these findings, which can help fellow researchers and policymakers. ## 1. Introduction Dengue is caused by RNA Dengue virus (DENV), which contains ~11 kb RNA genome including an ORF (open reading frame), three structural proteins, namely capsid (C), pre-membrane or membrane (prM/M), an envelope (E) [1]. Aedes female mosquito species carrying one of the four serotypes, such as DEN-1, DEN-2, DEN-3, and DEN-4, are the carriers of the acute febrile viral illness Dengue fever. Over 390 million individuals worldwide suffer from this illness each year, and $50\%$ of the world’s population belongs to the risk group [2]. The gold standard RT-PCR can find viral titers that aid patient diagnosis and therapy planning [3,4,5]. The clinical manifestations of Dengue infection are symptomatic, asymptomatic, or mild illness (such as fever, headache, myalgia, decreased platelet counts, and leucopenia), similar to flu symptoms [3]. Dengue hemorrhagic fever (DHF), Dengue shock syndrome (DSS), and life-threatening scenarios are sometimes considered in the case of severe illnesses [6]. However, DHF hematomas are found when patients are characterized by thrombocytopenia or extremely low platelet counts [1]. From an analysis of previous data, it was obtained that the climatic factors, especially temperature, rainfall, and humidity, have abruptly changed in the past few years [7]. The Dengue outbreaks and positive cases also accrued with the discrepancy of the climate. Meteorological conditions are linked to Dengue disease because Aedes Aegypti growth and life cycle are affected by rainfall, temperature, humidity, and wind, either directly or indirectly [7]. It has been reported that rising temperatures and precipitation have aided in the rise of Dengue incidences [8]. Such settings are advantageous for breeding places that promote vector growth as a result of an increase in human and vector interaction that enhances viral transmission from an infected person to a new person. These environmental modifications may flare up the genetic mutation of the viruses [8]. According to Mutsuddy et al. [ 2019], climate alteration may affect some vectors for multiplication or extinction [9]. In $70\%$ of Asian nations, Dengue infection is a severe public health concern [4]. The geographic location, climatic factors, rapid urbanization, deforestation, water pollution, lack of wastewater management system, high population density, and ineffective vector control strategies are responsible for DENV outbreaks. Several previously published articles reported relationships between various meteorological factors and Dengue cases, but this relationship has not been tested in recent years. Using data from 2000 to 2023 and five significant time series models, we thoroughly examined the association between meteorological factors and Dengue cases in this study. This study can guide fellow researchers and policymakers in tackling the Dengue outbreak. ## 2.1. Dengue Cases and Meteorological Factors Confirmed positive cases were downloaded from the Directorate General of Health Services (DGHS)’s website for patients from all over Bangladesh taking eight divisions (Supplementary Data S1). In this study, the daily DENV new cases between 1 January 2000 and 31 January 2023 were used from the DGHS website (https://old.dghs.gov.bd/index.php/bd/home/5200-daily-dengue-status-report) (accessed on 10 January 2023). We also obtained meteorological information from the NASA website to perform a time series analysis (NASA, 2022) (Supplementary Data S1) (https://www.nasa.gov/) (accessed on 10 January 2023). In this study, we considered daily dew point (°C), daily temperatures (°C), precipitation (mm), relative humidity (%), surface pressure (kPa), and wind velocity (m/s) at a level of 2 m height above ground level (Supplementary Data S1). ## 2.2. Statistical Time Series Models The Simple Exponential Smoothing Model (SES), Auto-Regressive Integrated Moving Average Model (ARIMA), Seasonal Auto-Regressive Integrated Moving Average Model (SARIMA), and GA time series models were used in this work to forecast recent Dengue cases. Additionally, Auto-Regressive Integrated Moving Average with Explanatory Variables (ARIMAX), Generalized Additive Model (GA), and Generalized Linear Mixed Models (GLM) time series models were used in this work to evaluate the correlations between meteorological parameters and Dengue cases. The SES model is utilized as a baseline to assess the prediction accuracy of other models in this investigation. All models were used to forecast new DENV cases for 30 days [10,11,12]. ## 2.2.1. Simple Exponential Smoothing Model (SES) SES forecasting enables a short-term prediction to use a reasonably steady mean while assuming data changes [13,14,15] created the original version of this model, which is still a helpful observation technique, and its application has been rapidly expanding in recent years. SES was noted as one of the most well-liked and potent models ideal for investigations where there is no clear trend or weather pattern, according to Weller and Crone [2012]. The R package “fpp2” created the SES model, a univariate time series prediction. The following equation can present the SES model:Vt = V (t−1) + α(v (t−1) – V (t−1))[1] In Equation [1], vt denotes the actual value of the series at time t, Vt stipulates the forecast value of the series at the time as well as t, and α is a weighting factor that takes a value between 0 and 1. In Equation [1], vt stands for the series’ actual value at time t, Vt for its forecasted value at the desired time, and α is a weighting factor that can take a value between 0 and 1. ## 2.2.2. Auto-Regressive Integrated Moving Average Model (ARIMA) A statistical time series analysis technique called the auto-regressive integrated moving average model (ARIMA) uses time series data to analyze data sets and forecast trends. A time series model called the ARIMA auto-regressive integrated moving average is used in statistical analysis to analyze seasonal patterns and estimate results. This model has three steps: identification, factor prediction, and model diagnosis. Using autocorrelation (ACF) and partial autocorrelation functions, auto-regressive and shifting average constituents are chosen after seasonality and stationary recognition. The ARIMA examines opposite goals in the lack of seasonality in the time series data [16] where the “forecast R package” is used in this current study, and Equation [2] is presented below [17]:Vt = α + φ1V (t−1) + … +φpV (t−p) + θ1e (t−1) + … +θqe (t − q) + Wt[2] In this equation, α is the constant value, φ1 − φp means auto-regressive model parameters, θ1 − θq indicates moving-average model parameters, and Vt for its forecasted value at the desired time, where the wt is white noise. ## 2.2.3. Auto-Regressive Integrated Moving Average with Explanatory Variables (ARIMAX) Auto-Regressive Integrated Moving Average with Explanatory Variables (ARIMAX) is another significant and familiar time series model which allows external parameters, such as climatic factors, to upgrade the analysis and forecasting accuracy [18,19,20,21]. This model analyzes the association between the time series data for arranging genuine designs, exploiting the straight perceptions, and deleting high-frequency commotion. ## 2.2.4. Seasonal Auto-Regressive Integrated Moving Average Model (SARIMA) Seasonal-ARIMA, or SARIMA, is another time series model analogous to ARIMA but more precise and used for forecasting trends and seasonal alterations. These models are used for contrasting the time series data for seasonal frequency, although for non-seasonal cases [22,23,24,25]. ## 2.2.5. Generalized Additive Model (GA) The GAM, or generalized additive model, is applied for analyzing the interaction of climatic factors with Dengue cases already being used for various research purposes. This time series model is more eminent and user-friendly, which is preferable for analyzing national morbidity, mortality, and air pollution studies. GAM fits the generalized additive model for parametric and nonparametric regression and smoothing. It also augments the conventional GLM (generalized linear models) by changing linear predictors of the form η = Σj βjxj with η = Σjfj (xj). fj (xj) is used here for a nonparametric function R package for analyzing GAM in this study. ## 2.2.6. Generalized Linear Mixed Models (GLM) The GLMM, or Generalized Linear Mixed Model, is a statistical model that augments the Generalized Linear Model (GLM). This model is used to analyze the random effects from clustered categorical data. The ability of the GLMM model to distinguish the impact of nested data is a significant advantage. The season is the second category in the current study, while year characteristics are used as repeated observations. A standard Equation [3] of this model is presented below:y = Xβ + Zu + ε[3] This equation denotes anN × 1 as the outcome variable; X is anN × p matrix; β is a p × 1 column vector; N × q is the design matrix; u is a q × 1 vector of the random; and ε is anN × 1 column vector of the residuals. ## 2.3. Statistical Analysis This study applied five statistical time series models to see the trend and forecasting of Dengue cases for the recent future, along with forecasting from three time series models showing a correlation between Dengue cases and meteorological conditions. Additionally, Spearman’s rank correlation coefficients were employed in this investigation. For determining seasonal variation, we used the Poisson regression model in three major seasons in Bangladesh. Data can be characterized by a Poisson distribution when observations are counted in whole numbers and when event occurrences are independent (one event occurrence does not affect the chance of another event occurring), and when the specifics of the observed time interval are known and are the same for each participant. ## 3.1. Dengue Status in Bangladesh In 2022, the first Dengue case was confirmed in the month of January, and 126 Dengue-positive patients were recorded. From the DGSH data collection, it can be shown that Dengue cases rose from April 2022 to November [19,334] and December [5024], including May (163 cases), June [737], July [1571], August [3521], September [9911], and October [21,932] (Figure 1). The highest number of mortality was recorded in November [113] and October [86]. The Dengue outbreak was tremendously devastating, with 62,382 patients in 2022 compared with only 28,429 confirmed cases in the year 2021 (Figure 1). The mean for Dengue cases is 882.26, with a standard deviation (SD) of 3993.18, a minimum of 0, and a maximum of 52,636 (Table 1). Table 1 also points out that there are 52,636 maximum cases in Bangladesh. The average daily confirmed cases are approximately 882.26, while daily average values for temperature, dew point, relative humidity, precipitation, surface pressure, and wind speed are 25.38 °C, 19.80 °C, $74.78\%$, 5.81 mm/day, 100.68 kPa, and 1.93 m/s, respectively (Table 1). The descriptive analysis of this study stipulates that the lowest temperature is 15.39 °C, whereas the highest is 31.86 °C from 1 January 2000 to 31 January 2023 in Bangladesh. In addition, precipitation is found to be a minimum of 0 mm/day and a maximum of 28.39 mm/day. Moreover, the lowest surface pressure is recorded as 99.75 kPa, and the highest is 101.52 kPa. ## 3.2. Association between Daily Dengue Cases and Meteorological Variables Correlations between meteorological parameters and Dengue cases are presented in Figure 2. This graph suggests that none of the associations are statistically significant. However, Spearman’s rank correlation coefficients between meteorological variables and confirmed daily Dengue cases indicate a meaningful but weak relationship between Dengue cases and climatic parameters (Figure 2). Concerning daily Dengue cases, the mean temperature and wind speed show a weakly negative link (r = −0.007, $p \leq 0.05$ and $r = 0.085$, $p \leq 0.05$, respectively). However, there is a significant positive correlation among dew point, relative humidity, and rainfall with daily Dengue cases ($r = 0.158$, $p \leq 0.05$, $r = 0.175$, $p \leq 0.05$, and $r = 0.138.$ $p \leq 0.05$), respectively. On the contrary, surface pressure exhibits an insignificant negative correlation with daily Dengue cases (r = −0.086, $p \leq 0.05$) (Figure 2). ## 3.3. Time Series Model Results The forecasting results from the time series models are displayed in Figure 3, along with the confirmed and projected Dengue cases from January 2000 to January 2023. With AIC, AICc, and BIC of 6083.90, 6083.99, and 6094.77, respectively, we describe a consistent pattern between observed and predicted national Dengue cases in the SES model (Table 2 and Figure 3). Dengue with AIC, AICc, and BIC of 5232.86, 5233.01, and 5247.34 for the ARIMA model, compared with 5088.54, 5088.77, and 5106.44 for the SARIMA model, and 5238.35, 5239.18, and 5274.56 for the ARIMAX model, we detected a strong, growing trend between observed and predictive Dengue cases. However, the GA model performs weakly (i.e., AIC = 5358.541, AICc = 5358.870, and BIC = 5396.837). All models, except for SES, forecast a considerable increase in cases of Dengue over the following 30 days (Figure 3). In the ARIMAX and GA models, wind speed (−666.50 [$95\%$ CI: −1711.86 to 378.86] and −953.05 [−2403.46 to 497.36], respectively) and dew point (−102.00 [−1308.42 to 1104.43] and −616.31 [−2079.18 to 846.55], respectively) have a slight negative interaction with Dengue cases. Nevertheless, temperature and relative humidity are positively affiliated with Dengue cases (105.71 [−1034.44 to 1245.87] and 633.86 [−2079.18 to 846.55], respectively) in the ARIMAX model, and the GA model also exhibits identical results. In contrast, rainfall is negatively associated with the ARIMAX model (−69.52 [−178.61 to 39.58]) and positively associated (21.59 [−114.83 to 158.02]) in the GA model (Table 3). In the GLM model, temperature and dew point (0.55 [$95\%$ CI: 0.31 to 0.98] and 2.06 [1.11 to 3.82], respectively) have a significant association with Dengue cases. Temperature displayed a negative association, whereas dew point stipulates a positive association according to the GA model. However, wind speed, relative humidity, rainfall, and surface pressure are fruitlessly associated with Dengue cases, though wind speed and rainfall (0.98 [$95\%$ CI: 0.68 to 1.42] and 1.01 [0.98 to 1.04], respectively) have a positive association with Dengue cases. Contrarily, relative humidity (0.88 [$95\%$ CI: 0.75 to 1.04]) negatively affects Dengue cases (Table 3). In the Poisson regression model, temperature and relative humidity (1.30 [$95\%$ CI: 1.29 to 1.31] and 1.15 [1.14 to 1.15], respectively) have a substantial positive correspondence with Dengue cases in the winter season. In that season, wind speed and rainfall (0.01 [$95\%$ CI: 0.01 to 0.02] and 1.08 [1.07 to 1.09], respectively) have a substantial negative correspondence with Dengue cases. In the summer, temperature (IRR = 1.39), dew point (6.37), rainfall (1.08), and surface pressure (10.59) have a significant positive association with Dengue cases (Table 4). In that season, wind speed (0.65) and relative humidity (0.84) have a significant antagonistic association with Dengue cases. In the monsoon season, temperature (2.52), rainfall (1.08), and surface pressure (10.57) have an essential curvilinear correlation with Dengue cases. In that season, wind speed (0.84) and relative humidity (0.97) significantly negatively affect Dengue cases. ## 4. Discussion Bangladesh is still fighting against the most devastating COVID-19 outbreak, and the recent dreadful flood also struck it [26,27,28,29]. Although the SARS-CoV-2 pandemic raises awareness of zoonotic infections and the need for new vaccinations against the novel, emerging, or re-emerging viruses, other zoonotic viral diseases are also rising day by day without control [14,20,22,23]. This study hypothesizes that daily Dengue cases can be related to meteorological factors. This study has indicated that dew point, relative humidity, and rainfalls are three meteorological parameters that might be linked to daily Dengue cases. Everyday Dengue illness occurrences in Bangladesh were correlated with dew point, relative humidity, and precipitation, according to an analysis of the metrological parameters ($r = 0.158$, $p \leq 0.05$; $r = 0.175$, $p \leq 0.05$; and $r = 0.138$, $p \leq 0.05$, respectively). Other elements, such as stagnant water and sewage effluent, which also generated distinct habitats for mosquito reproduction, are associated with the rise of Dengue. Additionally, the utilization of metrological data series for more than 20 years demonstrated the connection between current, widespread sickness in Bangladesh and climate change. Another linkage could be established during the flooding period as a reduction in disease outbreaks because the flood could wash away the mosquito larvae from wetlands, marshlands, and floodplains. In this study, a one-degree increase in dew point might increase the risk of Dengue cases two times according to the GLM model, and in the summer, it is five times according to the Poisson regression model. The GLM and Poisson models’ results revealed a substantial correlation between them for dew point and Dengue incidences. Analogous observations of a significant relationship between dew point and Dengue incidence were observed in Brazil [30]. This study also reveals that the high humidity ameliorates the Dengue cases in the winter and amortizes in the summer and monsoon seasons. High humidity favors the elevated longevity of adult mosquitoes and the shortening of the viral incubation period, thereby allowing an increased transmission intensity [31]. Humidity also affects adult mosquitoes’ survival and biting frequency [32]. The results of a few other researchers have also been inconsistent and inconclusive. Relative humidity, with a 3–4 month lag period, was shown to be the most important predicting factor for Dengue outbreaks in Indonesia, according to a study conducted there [33]. According to another study, a Dengue outbreak typically occurs the following year when there is a low relative humidity level in September and October [34]. Rainfalls escalate the water stagnation and proliferation of mosquitoes so that it may intensify DENV. In summer and monsoon seasons, a 1 mm increase in rainfall can increase Dengue cases by $8\%$. Rainfall-induced increases in vector density and a corresponding decline in Dengue incidence have been documented. Despite Bangladesh having a year-round potential for Dengue transmission, there have been very few winter/dry season. Dengue cases reported over the past 20 years due to a lack of moisture needed to refill frequent mosquito breeding sites [35]. A long-term study of meteorological factors with the number of Dengue cases using Poisson regression showed that the seasonal meteorological factors were positively correlated with them. This is congruent with research on the correlation between Dengue cases and climatic variables in Jeddah, Saudi Arabia, including relative humidity and temperature [36]. The climate significantly influences the prevalence of infection with Dengue hemorrhagic fever. The climate significantly impacts the frequency of Dengue hemorrhagic fever infection in the Kolaka area, according to a study by Tosepu et al. [ 2018]. It was the country’s first study, as far as we know. However, other sources have also reported on a season-specific pattern of Dengue cases in Southeast Asia [37]. Our findings concur with other extensive research from the same geographic area, Myanmar and India [38], and the studies from the Gulf of Thailand and Puerto Rico [39]. Temperature and rainfall were discovered to be important contributing factors in earlier investigations [40]. Moreover, earlier studies presupposed that the impacts of climate factors are not seasonal. The effects of climate variables might, however, also vary throughout time. There are different ways that rainfall can affect the incidence of Dengue, according to several studies. Additionally, earlier research presupposed that the impacts of climatic factors are not seasonal. The impacts of climatic factors might, however, also vary throughout time. Numerous studies have shown that rainfall’s impact on Dengue occurrence can change throughout the year [41]. Since rain can damage potential mosquito habitats, the heavy rain that falls during monsoon season is likely to have a detrimental impact on the growth of the mosquito population. On the other hand, rains throughout the winter might leave stagnant pools of water ideal for mosquito breeding. Few human-made and natural occurrences in some regions alter precipitation, which affects mosquito dispersion and population size [42]. Forecasting the frequency and length of outbreaks may be performed using weather variables, including relative humidity, wind speed, lowest and peak temperatures, and average and highest values [36]. The overall number of hospitalizations in Bangladesh as of 12 October 2022, as reported by the DGSH, was 23,592 individuals, of whom 17,456 were hospitalized in Dhaka and 6136 elsewhere. According to the DGSH report on 12 October 2022, the total number of registered hospitalizations was 23,592 in Bangladesh, of which 17,456 were in Dhaka city, and 6136 were outside of Dhaka city. The Department of Health, GoB, reported that 12,875 DENV patients were admitted to different hospitals. At the same time, 10,017 were from the capital, and the second-highest numbers were found in the Chittagong division [2814] and the Khulna division [1082], while Barisal [787], Rajshahi [616], Mymensingh [267], Rangpur [36], and the Sylhet division [22] followed. In the year 2022, the total mortality was more than 70, where 45 people died in Dhaka city, 1 in Narayanganj district, 32 in Chattogram, and 5 in the Barishal division (Figure 1). The first 14 days of October saw at least 7500 patients, or about $32\%$ of all hospitalized patients, and September alone saw 9911 patients, the most of any month. In addition, 3571 patients were admitted to the hospital in August, followed by 1571 patients in July, 737 patients in June, 163 patients in May, 23 patients in April, 20 patients each in February and March, and 126 patients in January 2022. In the Rohingya refugee camps (Forcibly Displaced Myanmar Nationals (FDMN)) in Cox’s Bazar district, Bangladesh, around 7687 confirmed positive DENV cases, with six deaths disclosed (case fatality rate roughly $0.08\%$). The replication and multiplication of vectors for this disease were highly dependent on climatic factors [28,29,43]. The climate of Bangladesh, as a part of tropical Asian countries, consists of two monsoons with year-to-year time variation. Bangladesh has a tropical-moist climate with variable seasonal rainfall, a moderately warm climate, and significant relative humidity [44]. In this country, there are typically four distinct climatic seasons per year: The winter, which lasts from December to February, is when temperatures are lowest; The summer, from March to May, is when they are highest; The rainy season, from June to September, and the fourth number, the post-monsoon autumn, which lasts from October to November [45]. From the observations of the 61 years (1960–2021) of daily temperature data, we discovered an average temperature of 25.2 °C in July, a minimum temperature of 12.9 °C in January, and a maximum temperature of 33.5 °C in April in Bangladesh. We also observed a maximum precipitation of 496 mm in July and a minimum of 4 mm in January. It is counted that the rising world temperature may associate with escalating vector-borne diseases. According to these study findings, climatic factors are closely associated with ameliorating Dengue infection. Most positive cases in Bangladesh were recorded during the monsoon season when rainy weather promotes mosquito development and survival. The findings of this study also showed a correlation between Dengue cases and relative humidity and rainfall. The development of the Dengue virus in ambient temperature and other factors impact Aedes mosquitoes, which are daytime feeders that start to spread at about 29.3 °C. This suggested that mosquito growth is best at temperatures between 25 °C and 27 °C. Additionally, a higher temperature in the range of 27 °C to 31 °C is a good predictor for the growth of Aedes mosquitoes, having a preventative impact on the spread of Dengue. Another climatic factor, rainfall, is crucial for replicating Aedes mosquitoes, which helps lay the vector eggs [46,47,48,49]. First, appropriate preventive steps must be followed to solve the Dengue issue to avoid the affected mosquito contact. It should be mentioned that Aedes mosquitoes are more active and likely to bite during the day. As a result, people should avoid standing water and wastewater and use insecticides to kill them instead of coils and nets, long-sleeved clothing, or dresses (CDC, 2022). Vaccination and drug therapy should be developed in addition to taking different prevention steps [50,51]. Antipyretics (paracetamol) can be used for fever, analgesics, or painkillers for joint pain. Severe patients with DHF/DSS should be immediately hospitalized, where oral rehydration therapy can be followed for dehydration. Platelet transfusion is recommended when its level reaches 20,000 or below [52]. These climatic elements can be incorporated into early warning systems to improve Dengue forecasting, vector management, and control in the future [53]. Vector control should follow according to the WHO’s Integrated Vector Management (IVM) system to regulate the vector’s breeding areas, lessen the vector’s environment management, decrease the stagnant areas of water, and maximize the use of larvicides, fogging, fumigation in camps, and the treatment of waste and wastewater before releasing into the environment. Bangladesh also recently experienced the dreadful flood that commenced on 17 May 2022, and mainly captivated two districts of the northeastern division of Sylhet with 79 deaths, thousands of injuries, and enormous economic loss. Due to widespread infection, the ongoing COVID-19 infectious disease caused by SARS-CoV-2 has damaged global public health, enterprises, and economies [28,29,51]. It is not crystal clear until now why the Dengue infection is augmentation while government actions, environment, season, wastewater, and public health knowledge gap are prevalent. It can turn into another outbreak if it cannot be immediately controlled. Few environmental variables are interconnected with each other. For example, the dew point is affected by atmospheric humidity and air temperature. Humidity influences rainfall, and the quantity of rain influences water stagnation everywhere, such that it may increase the vector of DV (mosquito). Global warming, rising sea levels, and increased water stagnation in Bangladesh contribute to an increase in Dengue disease vectors. It is almost impossible to include all of these environmental factors in current mathematical models because many works are in parallel, and many are local factors. Investigating and determining the ecological factors’ role in this outbreak can simplify interpretation. However, we believe the current study can provide a new perspective for discussion. ## 5. Conclusions Dengue is a flavivirus infection spread by mosquitoes and is most common in tropical and subtropical areas. The significance of meteorological factors in Dengue disease is identified in many studies where the temperature was responsible for quick viral vector replication, the humidity boosted vector transmission, and precipitation also showed different effects on the spread of mosquito eggs as well as larvae. According to this study result, climatic parameters, temperature, and relative humidity showed positive correlations with Dengue cases using 23 years of data. SARIMA scored better than the other four time series models, suggesting that we were able to estimate events more precisely when we took into account seasonal influences. These results also validate our Poisson regression model. All three models reported less correlation with Dengue cases than by the Poisson regression model. Moreover, the Poisson model with seasonal segmented data showed a stronger correlation. This scenario is mainly attributed to inadequate education, knowledge gaps, a lack of appropriate legislative rules, and a poor sanitation system without proper wastewater management. Combining wastewater-based and serological-based hospital surveillance can detect hotspots and predict patient numbers. A multi-sectoral approach should be applied to amortize Dengue transmission, urbanization planning, environmental degradation, production of active therapeutic vaccines, and education to prevent and control the disease. Bangladesh has been attacked by Dengue many times, and one of the events recorded in 2019 is its highest case. A multifaceted approach is needed for vector control, rapid diagnosis at a low cost, a proper care system, and vaccines for four strains of Dengue. The necessity of collaborative research is crucial to gain a deeper understanding of this disease, immunity, and virus. ## References 1. Murugesan A., Manoharan M.. Dengue Virus. Emerging and Reemerging Viral Pathogens (2020.0) 281-359 2. Wang Y., Wei Y., Li K., Jiang X., Li C., Yue Q., Zee B.C.-Y., Chong K.C.. Impact of extreme weather on Dengue fever infection in four Asian countries: A modelling analysis. Environ. Int. (2022.0) 169 107518. DOI: 10.1016/j.envint.2022.107518 3. Perera R., Kuhn R.J.. Structural proteomics of Dengue virus. Curr. Opin. Microbiol. (2008.0) 11 369-377. DOI: 10.1016/j.mib.2008.06.004 4. Kakarla S.G., Kondeti P.K., Vavilala H.P., Boddeda G.S.B., Mopuri R., Kumaraswamy S., Kadiri M.R., Mutheneni S.R.. Weather integrated multiple machine learning models for prediction of Dengue prevalence in India. Int. J. Biometeorol. (2022.0) 67 285-297. DOI: 10.1007/s00484-022-02405-z 5. Herath J.M.M.K., Abeyasundara H.T.K., De Silva W.A.P.P., Weeraratne T.C., Karunaratne S.H.P.P.. Weather-Based Prediction Models for the Prevalence of Dengue Vectors Aedes aegypti and Ae. albopictus. Chowdhury, R., Ed. J. Trop. Med. (2022.0) 2022 4494660. DOI: 10.1155/2022/4494660 6. Li Z., Dong J.. Big Geospatial Data and Data-Driven Methods for Urban Dengue Risk Forecasting: A Review. Remote Sens. (2022.0) 14. DOI: 10.3390/rs14195052 7. Ochida N., Mangeas M., Dupont-Rouzeyrol M., Dutheil C., Forfait C., Peltier A., Descloux E., Menkes C.. Modeling present and future climate risk of Dengue outbreak, a case study in New Caledonia. Environ. Health (2022.0) 21 20. DOI: 10.1186/s12940-022-00829-z 8. Gómez Gómez R.E., Kim J., Hong K., Jang J.Y., Kisiju T., Kim S., Chun B.C.. Association between Climate Factors and Dengue Fever in Asuncion, Paraguay: A Generalized Additive Model. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph191912192 9. Mutsuddy P., TahminaJhora S., Shamsuzzaman A.K.M., Kaisar S.M.G., Khan M.N.A.. Dengue Situation in Bangladesh: An Epidemiological Shift in terms of Morbidity and Mortality. Can. J. Infect. Dis. Med. Microbiol. (2019.0) 2019 3516284. DOI: 10.1155/2019/3516284 10. Islam M.A., Hasan M.N., Ahammed T., Anjum A., Majumder A., Siddiqui M.N.-E., Mukharjee S.K., Sultana K.F., Sultana S.. Association of household fuel with acute respiratory infection (ARI) under-five years children in Bangladesh. Front. Public Health (2022.0) 10 2296-2565;. DOI: 10.3389/fpubh.2022.985445 11. Islam M.A., Ahammed T., Ahmed Noor S.T., Hasan M.N., Hoque M.N., Tiwari A., Harapan H., Dhama K., Islam T., Bhattacharya P.. An Estimation of Five-decade Long Monkeypox Case Fatality Rate: Systematic Review and Meta-analysis. J. Pure Appl. Microbiol. (2022.0) 16 3036-3047. DOI: 10.22207/JPAM.16.SPL1.16 12. Islam M.A., Hemo M.K., Chopra H., Amin M.R., Bhattacharya P., Dhama K.. Old Enemy with a New Face: Re-emerging Monkeypox Disease—An Update. J. Pure Appl. Microbiol. (2022.0) 16 2972-2988. DOI: 10.22207/JPAM.16.SPL1.18 13. Slam A., Sangkham S., Tiwari A., Vadiati M., Hasan M.N., Noor S.T.A., Mumin J., Bhattacharya P., Sherchan S.P.. Association between Global Monkeypox Cases and Meteorological Factors. Int. J. Environ. Res. Public Health (2022.0) 19 15638. PMID: 36497712 14. Dhama K., Chandran D., Chakraborty S., Yatoo M.I., Islam A., Bhattacharya M., Chakraborty C., Harapan H., Chaicumpa W.. Zoonotic concerns of Marburg virus: Current knowledge and counteracting strategies including One Health approach to limit animal-human interface: An update. Int. J. Surg. (2022.0) 106 106941. DOI: 10.1016/j.ijsu.2022.106941 15. Islam A., Rahman A., Jakariya N.M., Bahadur F., Hossen S.K., Mukharjee M.S., Hossain A., Tasneem A., Haque F.. A 30-day follow-up study on the prevalence of SARS-COV-2 genetic markers in wastewater from the residence of COVID-19 patient and comparison with clinical positivity. Sci. Total Environ. (2022.0) 858 159350. DOI: 10.1016/j.scitotenv.2022.159350 16. Aldaej A., Ahamed Ahanger T., Yousuf Uddin M., Ullah I.. Secure Dengue Epidemic Prediction System: Healthcare Perspective. Comput. Mater. Contin. (2022.0) 73 1723-1745. DOI: 10.32604/cmc.2022.027487 17. De Livera A.M., Hyndman R.J., Snyder R.D.. Forecasting Time Series with Complex Seasonal Patterns Using Exponential Smoothing. J. Am. Stat. Assoc. (2011.0) 106 1513-1527. DOI: 10.1198/jasa.2011.tm09771 18. Islam A., Hossen F., Rahman A., Sultana K.F., Hasan M.N., Haque A., Sosa-Hernández J.E., Oyervides-Muñoz M.A., Parra-Saldívar R., Ahmed T.. An opinion on Wastewater-Based Epidemiological Monitoring (WBEM) with Clinical Diagnostic Test (CDT) for detecting high-prevalence areas of community COVID-19 Infections. Curr. Opin. Environ. Sci. Health (2022.0) 31 100396. DOI: 10.1016/j.coesh.2022.100396 19. Chakraborty S., Chandran D., Mohapatra R.K., Islam A., Alagawany M., Bhattacharya M., Chakraborty C., Dhama K.. Langya virus, a newly identified Henipavirus in China—Zoonotic pathogen causing febrile illness in humans, and its health concerns: Current knowledge and counteracting strategies—Correspondence. Int. J. Surg. (2022.0) 105 106882. DOI: 10.1016/j.ijsu.2022.106882 20. Chakraborty S., Mohapatra R.K., Chandran D., Alagawany M., Sv P., Islam A., Chakraborty C., Dhama K.. Monkeypox vaccines and vaccination strategies: Current knowledge and advances. An update—Correspondence. Int. J. Surg. (2022.0) 105 106869. DOI: 10.1016/j.ijsu.2022.106869 21. Chakraborty C., Bhattacharya M., Sharma A.R., Roy S.S., Islam A., Chakraborty S., Nandi S.S., Dhama K.. Deep learning research should be encouraged for diagnosis and treatment of antibiotic resistance of microbial infections in treatment associated emergencies in hospitals. Int. J. Surg. (2022.0) 105 106857. DOI: 10.1016/j.ijsu.2022.106857 22. Chakraborty S., Chandran D., Mohapatra R.K., Yatoo M.I., Islam A., Sharma A.K., Dhama K.. Sciences Marburg Virus Disease—A Mini-Review. J. Exp. Biol. Agric. (2022.0) 10 689-696 23. Chandran D., Dhama K., M K M.A., Chakraborty S., Mohapatra R.K., Yatoo M.I., Isiam A., Alagawany M., Sharma A., Mohankumar P.. Monkeypox: An Update on Current Knowledge and Research Advances. J. Exp. Biol. Agric. Sci. (2022.0) 10 679-688. DOI: 10.18006/2022.10(4).679.688 24. Ahmed F., Islam M.A., Kumar M., Hossain M., Bhattacharya P., Islam M.T., Hossen F., Hossain S., Islam S., Uddin M.. First detection of SARS-CoV-2 genetic material in the vicinity of COVID-19 isolation Centre in Bangladesh: Variation along the sewer network. Sci. Total Environ. (2021.0) 776 145724. DOI: 10.1016/j.scitotenv.2021.145724 25. Haider N., Chang Y.-M., Rahman M., Zumla A., Kock R.A.. Dengue outbreaks in Bangladesh: Historic epidemic patterns suggest earlier mosquito control intervention in the transmission season could reduce the monthly growth factor and extent of epidemics. Curr. Res. Parasitol. Vector-Borne Dis. (2021.0) 1 100063. DOI: 10.1016/j.crpvbd.2021.100063 26. Islam A., Marzan A.A., Islam S., Sultana S., Parvej I., Hossain M.S., Amin M.T., Hossain F.E., Barek A., Hossen S.. Sex-specific epidemiological and clinical characteristics of COVID-19 patients in the southeast region of Bangladesh. medRxiv (2021.0). DOI: 10.1101/2021.07.05.21259933 27. Islam M.A., Haque M.A., Rahman M.A., Hossen F., Reza M., Barua A., Marzan A.A., Das T., Baral S.K., He C.. A Review on Measures to Rejuvenate Immune System: Natural Mode of Protection Against Coronavirus Infection. Front. Immunol. (2022.0) 13. DOI: 10.3389/fimmu.2022.837290 28. Sakib M.M.H., Nishat A.A., Islam M.T., Raihan Uddin M.A., Iqbal M.S., Bin Hossen F.F., Ahmed M.I., Bashir S., Hossain T., Tohura U.S.. Computational screening of 645 antiviral peptides against the receptor-binding domain of the spike protein in SARS-CoV-2. Comput. Biol. Med. (2021.0) 136 104759. DOI: 10.1016/j.compbiomed.2021.104759 29. Correia Filho W.L.F.. Influence of meteorological variables on Dengue incidence in the municipality of Arapiraca, Alagoas, Brazil. Rev. Soc. Bras. Med. Trop. (2017.0) 50 309-314. DOI: 10.1590/0037-8682-0432-2016 30. Mutheneni S.R., Morse A.P., Caminade C., Upadhyayula S.M.. Dengue burden in India: Recent trends and importance of climatic parameters. Emerg. Microbes Infect. (2017.0) 6 1-10. DOI: 10.1038/emi.2017.57 31. Banu S., Guo Y., Hu W., Dale P., Mackenzie J.S., Mengersen K., Tong S.. Impacts of El Niño Southern Oscillation and Indian Ocean Dipole on Dengue incidence in Bangladesh. Sci. Rep. (2015.0) 5 16105. DOI: 10.1038/srep16105 32. Bhatia S., Bansal D., Patil S., Pandya S., Ilyas Q.M., Imran S.. A Retrospective Study of Climate Change Affecting Dengue: Evidences, Challenges and Future Directions. Front. Public Health (2022.0) 10. DOI: 10.3389/fpubh.2022.884645 33. Halide H., Ridd P.. A predictive model for Dengue Hemorrhagic Fever epidemics. Int. J. Environ. Health Res. (2008.0) 18 253-265. DOI: 10.1080/09603120801966043 34. Paul K.K., Dhar-Chowdhury P., Haque C.E., Al-Amin H.M., Goswami D.R., Kafi M.A.H., Drebot M.A., Lindsay L.R., Ahsan G.U., Brooks W.A.. Risk factors for the presence of Dengue vector mosquitoes, and determinants of their prevalence and larval site selection in Dhaka, Bangladesh. Terenius, O., Ed. PLoS ONE (2018.0) 13. DOI: 10.1371/journal.pone.0199457 35. Alkhaldy I.. Modelling the association of Dengue fever cases with temperature and relative humidity in Jeddah, Saudi Arabia—A generalised linear model with break-point analysis. Acta Trop. (2017.0) 168 9-15. DOI: 10.1016/j.actatropica.2016.12.034 36. Schwartz E., Weld L.H., Wilder-Smith A., von Sonnenburg F., Keystone J.S., Kain K.C., Torresi J., Freedman D.O.. Seasonality, Annual Trends, and Characteristics of Dengue among Ill Returned Travelers, 1997–2006. Emerg. Infect. Dis. (2008.0) 14 1081-1088. DOI: 10.3201/eid1407.071412 37. Victor T.J., Malathi M., Asokan R., Padmanaban P.. Laboratory-based Dengue fever surveillance in Tamil Nadu, India. Indian J. Med. Res. (2007.0) 126 112-115. PMID: 17932434 38. Johansson M.A., Dominici F., Glass G.E.. Local and Global Effects of Climate on Dengue Transmission in Puerto Rico. Massad, E., Ed. PLoS Negl. Trop. Dis. (2009.0) 3. DOI: 10.1371/journal.pntd.0000382 39. Hossain M.P., Zhou W., Ren C., Marshall J., Yuan H.-Y.. Prediction of Dengue annual incidence using seasonal climate variability in Bangladesh between 2000 and 2018. Hamers, R.L., Ed. PLoS Glob. Public Health (2022.0) 2. DOI: 10.1371/journal.pgph.0000047 40. Yuan H.-Y., Liang J., Lin P.-S., Sucipto K., Tsegaye M.M., Wen T.-H., Pfeiffer S., Pfeiffer D.. The effects of seasonal climate variability on Dengue annual incidence in Hong Kong: A modelling study. Sci. Rep. (2020.0) 10 4297. DOI: 10.1038/s41598-020-60309-7 41. Langkulsen U., na Sakolnakhon K.P.. Identifying High-risk Areas of Dengue by Meteorological factors in Thailand. IOP Conf. Ser. Earth Environ. Sci. (2022.0) 987 012001. DOI: 10.1088/1755-1315/987/1/012001 42. Hossain F.E., Islam S., Islam M.A., Islam S., Ahmed F.. Detection of virulence genes of APEC (avian pathogenic. Bioresearch Commun. (2021.0) 7 967-972. DOI: 10.3329/brc.v7i1.54253 43. Kamruzzaman M., Mandal T., Rahman A.T.M.S., Abdul Khalek M., Alam G.M.M., Rahman M.S.. Climate Modeling, Drought Risk Assessment and Adaptation Strategies in the Western Part of Bangladesh (2021.0) 21-54 44. Shahid S., Khairulmaini O.S.. Spatio-temporal variability of rainfall over Bangladesh during the time period 1969–2003. Asia-Pac. J. Atmos. Sci. (2009.0) 43 375-389 45. Roy S., Ripon A.R., Begum R., Bhowmik D.R., Amin M.T., Islam A., Ahmed F., Hossain M.S.. Arachidonic acid supplementation attenuates adipocyte inflammation but not adiposity in high fat diet induced obese mice. Biochem. Biophys. Res. Commun. (2022.0) 608 90-95. DOI: 10.1016/j.bbrc.2022.03.089 46. Haque M.A., Wang F., Chen Y., Hossen F., Islam M.A., Hossain M.A., Siddique N., He C., Ahmed F.. Front. Microbiol. (2022.0) 12 783103. DOI: 10.3389/fmicb.2021.783103 47. Roy S., Bhowmik D.R., Begum R., Amin M.T., Islam A., Ahmed F., Hossain M.S.. Aspirin attenuates the expression of adhesion molecules, risk of obesity, and adipose tissue inflammation in high-fat diet-induced obese mice. Prostaglandins Other Lipid Mediat. (2022.0) 162 106664. DOI: 10.1016/j.prostaglandins.2022.106664 48. Jakariya M., Ahmed F., Islam M.A., Al Marzan A., Hasan M.N., Hossain M., Ahmed T., Hossain A., Reza H.M., Hossen F.. Wastewater-based epidemiological surveillance to monitor the prevalence of SARS-CoV-2 in developing countries with onsite sanitation facilitie. Environ. Pollut. (2022.0) 119679. DOI: 10.1016/j.envpol.2022.119679 49. CDC (Dengue) 50. Rahman F.I., Ether S.A., Islam M.R.. Upsurge of Dengue Prevalence During the Third Wave of COVID-19 Pandemic in Bangladesh: Pouring Gasoline to Fire. Clin. Pathol. (2022.0) 15 2632010X2210760. DOI: 10.1177/2632010X221076068 51. Nasir N.H., Mohamad M., Lum L.C.S., Ng C.J.. Effectiveness of a fluid chart in outpatient management of suspected Dengue fever: A pilot study. Huy, N.T., Ed. PLoS ONE (2017.0) 12. DOI: 10.1371/journal.pone.0183544 52. Iguchi J.A., Seposo X.T., Honda Y.. Meteorological factors affecting Dengue incidence in Davao, Philippines. BMC Public Health (2018.0) 18. DOI: 10.1186/s12889-018-5532-4 53. Rakib S.H., Masum S., Patwari M.R.I., Fahima R.A., Farhana A., Islam M.A.. Design and Development of a low cost Ultraviolet Disinfection system to reduce the cross infection of SARS-CoV-2 in ambulances. Proceedings of the 2021 International Conference on Electronics, Communications and Information Technology (ICECIT) IEEE 1-4
--- title: 'Functionality and Quality of Life with Parkinson’s Disease after Use of a Dynamic Upper Limb Orthosis: A Pilot Study' authors: - María Jiménez-Barrios - Jerónimo González-Bernal - Esther Cubo - José María Gabriel-Galán - Beatriz García-López - Anna Berardi - Marco Tofani - Giovanni Galeoto - Martin J. A. Matthews - Mirian Santamaría-Peláez - Josefa González-Santos journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049252 doi: 10.3390/ijerph20064995 license: CC BY 4.0 --- # Functionality and Quality of Life with Parkinson’s Disease after Use of a Dynamic Upper Limb Orthosis: A Pilot Study ## Abstract Parkinson’s disease (PD) is a chronic, neurodegenerative movement disorder, whose symptoms have a negative impact on quality of life and functionality. Although its main treatment is pharmacological, non-pharmacological aids such as the dynamic elastomeric fabric orthosis (DEFO) merit an evaluation. Our objective is to assess the DEFO in upper limb (UL) functional mobility and in the quality of life of PD patients. A total of 40 patients with PD participated in a randomized controlled crossover study, and were assigned to a control group (CG) and to an experimental group (EG). Both groups used the DEFO for two months, the experimental group the first two months of the study and the control group the last two. Motor variables were measured in the ON and OFF states at the baseline assessment and at two months. Differences from the baseline assessment were observed in some motor items of the Kinesia assessment, such as rest tremor, amplitude, rhythm or alternating movements in the ON and OFF states with and without orthosis. No differences were found in the unified Parkinson’s disease rating scale (UPDRS) or the PD quality-of-life questionnaire. The DEFO improves some motor aspects of the UL in PD patients but this does not translate to the amelioration of the standard of functional and quality-of-life scales. ## 1. Introduction In the Global Burden of Diseases, Injuries and Risk Factors (GBD) study conducted in 2016, it was estimated that between the years 1990 and 2016 the number of people affected by Parkinson’s disease (PD) doubled worldwide, with an incidence rate of 8 to 18 people per 100,000 per year [1]. PD is defined as a chronic, neurodegenerative movement disorder, whose most characteristic motor symptoms are resting tremor, rigidity, and bradykinesia. Resting tremor is characterized by a prominent involuntary, rhythmic muscle movement in the distal upper limb (UL) at a frequency of about 4 to 6 Hz. Rigidity is an increased resistance to passive movement. The third most characteristic symptom is bradykinesia, characterized by slow movement and difficulty planning, initiating, and carrying out a movement [2]. Other non-motor symptoms such as sleep problems, constipation, anxiety, depression and fatigue may also appear [3] The motor and non-motor symptoms of PD have negative repercussions on the quality of life and functionality of people with the disease. The burden of motor symptoms and impairment of some activities of daily living (ADLs), such as eating, hygiene and clothing, related to alterations in functional mobility, has been identified as one of the major predictors of quality of life with this disease [4,5,6]. Following the perspective of the International Classification of Functioning and Disability (ICF) of the World Health Organization (WHO), three interconnected levels of human functioning are differentiated: [1] Body functions and structures, physiological and psychological functions, and bodily and anatomical impairments; [2] Limitations in the performance of activities; and [3] Restrictions in participation in daily life [7,8]. The progression of PD leads to alterations in body function, limited performance of ADLs and increased dependence, while reducing quality of life [4,8,9,10]. As the disease progresses, the worsening of symptoms, such as tremor, rigidity and bradykinesia, leads to a deterioration of manual dexterity, which translates to a greater difficulty in performing some ADLs. The most commonly reported basic self-care activities affected by PD symptoms are bathing/showering, dressing, and grooming/personal hygiene. Other instrumental activities of daily living that are affected are driving, preparing food, shopping, and writing [11]. Therefore, the presence of these symptoms is closely related to a poorer quality of life [5,12]. The treatment of PD is mainly based on the administration of levodopa, whose efficacy decreases over time and can produce side effects such as motor fluctuations, dyskinesias and dopaminergic dysregulation syndrome. The onset of the disease and the variety of possible symptoms makes it difficult to design a therapeutic regimen for the treatment of the disease. So far, approved therapies have focused on compensatory approaches aimed at treating clinical symptoms. However, the current research is focused on delaying or halting disease progression and not only on temporary symptomatic relief. Currently, all therapies are directed toward ameliorating motor deficits by increasing dopamine, but unfortunately, this loses efficacy over time as dopaminergic neurodegeneration progresses, with symptoms worsening in the long-term. Therefore, new non-pharmacological therapies need to be assessed [13,14,15]. There are several non-pharmacological therapies design to reduce functional impairments of this disease and, although evidence of their efficacy is increasing, there is still a limited number of studies on them and on the necessary intervention doses [16,17]. New non-pharmacological therapies that can be easily implemented can complement pharmacological treatment in order to improve the patients’ functional mobility and quality of life. In this regard, the dynamic elastomeric fabric orthoses (DEFO, Figure 1) may be a suitable candidate for reducing motor symptoms and improving functional movement and quality of life in patients with PD. These types of devices were developed by dynamic movement orthoses®, led by clinical orthopedist and managing director Martin Matthews. They are custom-designed devices for the user’s limbs or other parts of his/her body. Through the application of traction forces, they bring the limb into a better biomechanical alignment, while allowing and guiding movement. The elastic fabric promotes the extension of fingers and wrist, the stability of the thumb and the supination or pronation of the forearm. In addition, due to the localized compression of the soft tissues and the stimulation of the dermal and proprioceptive receptors, it is possible to regulate motor activity, avoiding atrophy and muscle rigidity, improving the patient’s quality of life [18,19]. These orthoses, compared to other orthopedic devices, have demonstrated better tolerance and high user satisfaction [20,21,22]. This type of device has been effective in children with cerebral palsy (CP). In the study conducted by Pavão et al., the use, by children with CP, of a vest made of this material showed better postural stability when performing a manual reaching activity [23], and in another study, it showed improved balance, postural control, and manual dexterity [24]. On the other hand, wearing these devices on the foot and ankle improved balance and walking speed in multiple sclerosis [25,26], and pain and function in patients suffering with complex regional pain syndrome (CRPS) [16]. Stroke has been the condition in which the use of this UL orthosis has been most investigated, and several studies have shown positive effects on strength, manual dexterity, and UL functionality, which need to be confirmed in studies with larger sample sizes [27,28]. DEFOs have not yet been investigated in in a wide range of motor variables in PD. In the recent review, conducted by Son Nguyen, studies of different types of portable orthoses for UL tremor suppression were assessed, the majority being active orthoses ($45\%$), followed by semi-active orthoses ($35\%$), and passive orthoses ($20\%$). All orthoses have proven to be effective in suppressing tremors, but several had inconveniencies such as being heavy and bulky, had not been evaluated in laboratory settings or were not yet commercially available [29]. Although current orthoses have proven to be effective in suppressing tremor, their clinical or home use is still limited. This limited their clinical or home use for suppressing tremor. Given these former results and lack of studies in PD, our main objective was to analyze the efficacy of a lighter device for the UL, such as the DEFO, in motor variables, functional mobility and quality of life in PD. ## 2.1. Participants A longitudinal crossover study, with a control group and an experimental group, was carried out. Participants with PD were recruited by consecutive non-probability sampling from September to October 2021. The inclusion criteria were: male and female patients diagnosed with PD, who, during the recruitment period, were attending the Neurology Department of the Burgos University Hospital, in any of the stages of severity of the disease, who had tremor and rigidity as a consequence of the disease in at least one of the UL. Patients whose tremor was a consequence of another associated disease according to the neurologist’s judgment or/and those with scores less than or equal to 26 on the Montreal cognitive assessment (MoCA) were excluded [30]. The diagnosis of PD was established following the criteria established by the International Parkinson and Movement Disorder Society. The prerequisite for the application of these criteria is the presence of bradykinesia in combination with resting tremor, rigidity or both. In addition, at least two of the four supporting criteria had to be met: resting tremor, dramatic improvement with dopaminergic therapy, occurrence of dyskinesias as a consequence of levodopa or olfactory loss, or cardiac sympathetic denervation on myocardial scintigraphy [31,32]. Each participant signed a written informed consent approved by the Clinical Research Ethics Committee of the Health Area of Burgos and Soria (Spain) with reference number CEIM-$\frac{2119}{2019}$ before participating in the present study (ClinicalTrials.gov test number: NCT04815382). Likewise, the study was conducted in accordance with the ethical principles set forth in the Declaration of Helsinki [33]. ## 2.2. Procedures The calculation of the sample size was based on the tremor and rigidity improvement as the main variables of the study. Given alpha risk of 0.05 and a beta risk of 0.20, in bilateral contrast, it is estimated that 40 participants (20 for each group) were required to detect a minimum difference of 0.50 in the rigidity and tremor scores of the most affected UL using the unified Parkinson’s disease rating scale, motor subscale part III (UPDRS) [34]. Considering the $10\%$ dropout rate during follow-up, a total sample of 40 patients was deemed necessary. Using the Epidat 4.2 program, participants were randomly assigned to the experimental group (EG) or the control group (CG). The EG treatment protocol consisted of implementing the DEFO in the most affected UL for two months (intervention period), whereas subjects in the CG led life as usual during the first two months (control period). One month prior to the implementation of the DEFO, measurements of the size and posture of the UL were conducted for the customization of the orthosis in the participants of both groups. At the first visit, the sociodemographic and clinical data of the participants were collected, and their fulfilment of the inclusion criteria was ensured. The participants were instructed to maintain their prescribed dopaminergic medication regimen. The effects of the DEFO were evaluated during the ON state (under the benefit of levodopa) and during the OFF state (1 h before the next levodopa intake). Motor assessments were conducted in the EG, at the end of the DEFO implementation period. Then, the DEFO was withdrawn and a second assessment was conducted two months later to evaluate if a carry-over effect was maintained during that time (Figure 2). Several assessment tools were administered to evaluate the functional activity, quality of life, and manual dexterity of the subjects. To obtain the primary outcomes, the unified Parkinson’s disease rating scale subscale II (UPDRS) was administered to assess functional activity consisting of 13 items. The score for each item is from 0 (normal) to 4 (worst), with a maximum score of 52 points, where higher scores indicate worse functional activity [35,36,37]. To assess the quality of life of each participant, the 39-item Parkinson’s disease questionnaire (PDQ-39) was administered, which consists of 29 items grouped into 8 domains: mobility, activities of daily living, emotional well-being, stigma, social support, cognition, communication, and grief and distress. Participants have to answer the questions based on their experience in the last four weeks. Each item is scored from 0 (never) to 4 (always). The maximum possible score is 156, with higher scores corresponding to worse quality of life [38,39]. For the assessment of UL dexterity, different motor aspects were evaluated. Subscale III of the UPDRS was administered, consisting of 17 items with a score range from 0 to 4 (from normal symptomatology to the most severe impairment), with a maximum score of 68 [35,36,37]. The Kinesia ONE motor assessment was used to collect and quantify the severity of motor symptoms such as tremor, bradykinesia, and dyskinesia. It provides an objective monitoring of Subscale III of the UPDRS. It is an electronic device consisting of software and a motion sensor. This sensor is positioned on the second finger of the hand during the time the patient performs a protocol of 12 tasks. The software scores each item from 0 (no symptoms) to 4 (severe impairment) [40]. Finally, the Purdue board test (PPT), the Minnesota manual dexterity test (MMDT) and the squares test (ST) were used to assess manual dexterity. The PPT consists of a two-column board that includes of 25 holes each, together with pins, washers, and rings located in four semicircles at the top of the board. The test is composed of four subtests that must be performed a total of three times, so that the total score is the average score obtained from the three attempts at each subtest. Thus, higher scores indicate better manual dexterity [41,42]. T, the abbreviated version of the MMDT, contains a rectangular wooden board that includes 60 holes distributed in 15 columns and 4 rows, as well as 60 circular pieces with one black and one red side of the same dimension as the holes in the board. It consists of two subtests that are performed a total of 4 times, obtaining as the final score, the average of the four attempts of each test. The final score is the time spent in performing the test, so the longer a patient spends, the worse the patient’s manual dexterity [43]. Finally, the ST contains a sheet of paper with four grids printed with 6 mm long squares. In the practice test, the patient must draw as many squares as possible for 10 s, while for the real test, he/she will have 30 s. The score is obtained for each hand by adding the number of dots drawn inside the squares, without touching the edges. Thus, a higher number of dots drawn indicates a better manual dexterity [44] (Figure 3). ## 3.1. Baseline Characteristics of the Study Participants The study had a simple crossover design, a total sample of 40 people with PD, 20 assigned to the CG, and 20 to the EG. Table 1 summarizes the baseline socio-demographic characteristics of the participants according to the study group. Men represented $75\%$ of the participants ($$n = 30$$), aged between 48 and 89 years, with a mean age of 71.00 ± 9.20 years and with 5.38 ± 4.23 years of disease evolution. The majority of participants ($$n = 35$$, $87.5\%$) lived accompanied at home, a minority lived alone at home ($$n = 4$$, $10\%$), and one in a religious community. Of the participants, $62.5\%$ ($$n = 25$$) had greater involvement in the right UL, while 37,$5\%$ ($$n = 15$$) had greater involvement in the left UL. Most participants ($87.5\%$, $$n = 35$$) did not receive any type of non-pharmacological treatment and the rest, $12.5\%$ ($$n = 5$$), attended physiotherapy, speech therapy, and/or occupational therapy. Table 2 shows the Kinesia ONE® measurements of action, resting tremor, and rigidity of the CG and EG participants before starting the intervention. The only differences between the groups are in resting tremor of the left UL in the OFF state. ## 3.2. Functionality and Quality of Life Table 3 shows the differences observed in the baseline assessment with and without orthosis in the OFF state in the motor variables evaluated with Kinesia ONE®. Wearing the orthesis reduces “postural tremor” compared with not wearing it ($$p \leq 0.042$$), which can improve functionality and quality of life. The same effect may reduce the orthesis of “finger tapping amplitude” ($$p \leq 0.18$$) and of “speed in rapid alternating movements” with orthosis ($p \leq 0.001$). Table 4 shows the differences in the motor variables evaluated with Kinesia ONE® (Cleveland, OH, USA) in the baseline assessment with and without orthosis in the ON state. Wearing the orthosis reduces “resting tremor” compared with not wearing it ($$p \leq 0.009$$), which can improve functionality and quality of life. In the same way, the reduction with the orthosis of “finger tapping amplitude” ($$p \leq 0.027$$) and in the item “amplitude of rapid alternating movements” ($$p \leq 0.017$$) with orthosis can favor functionality and quality of life. When comparing the change scores obtained on the UPDRS-II according to the patients’ condition and group type, no statistically significant differences were observed between the initial assessment and after two months of orthosis implementation in either the patients’ ON or OFF state. This means that no improvement in the UPDRS-II score was obtained after DEFO (Table 5). Likewise, no differences ($$p \leq 0.933$$) were observed in the quality of life of the subjects after the implementation of the orthosis (Table 6). ## 4. Discussion The aim of this study was to analyze the efficacy of the use of a DEFO for the UL on the functionality and quality of life of people with PD. The main findings of the present study are an immediate improvement after the implementation of the orthosis in the OFF and ON states of motor variables in the postural tremor task; only in the OFF state in the speed of rapid alternating movements and only in the ON state in the rhythm of hand movements and amplitude of rapid alternating movements. No differences were observed after two months of orthosis use in the improvement of functionality or in the quality of life of the patient with PD. Neurological disorders, such as PD, are currently the leading source of disability in the world. The global burden of disease study estimated that the number of people with PD will double from about 7 million in 2015 to approximately 13 million in 2040. This estimation of the growth of the population with PD is worrying considering the amount of burden this disease carries for society [45]. The neurodegenerative effects of PD lead to a loss of functional mobility in balance, postural stability and gait, decreasing independence in the performance of activities, and compromising their participation both at home and in the community [6,46,47]. On the other hand, contextual factors such as age, the feeling of being a person with a disability, unemployment or perceived control are examples of personal and environmental factors that have a negative impact on the functional mobility and quality of life of the individual [6,7,8,47,48]. There have been many advances in the knowledge of the etiopathogenesis and in the symptomatic treatment of PD in recent years. However, there are no effective neuroprotective or disease-modifying therapies that slow disease progression and improve functionality and quality of life without producing side effects on the patient [5]. Due to the fact that pharmacological treatment loses its efficacy with the passage of time and produces side effects in the person and the lack of precise knowledge about the currently existing non-pharmacological therapies, it is necessary to implement new non-pharmacological therapies that allow an improvement in the functionality and quality of life of the patient [16,17]. All DEFOs are made in the same way, being able to be designed and adapted to the needs of the pathology and the user, so that they can be devices for UL, lower limbs or vest for the whole body. In diseases such as CP, different studies carried out with vests and meshes of these characteristics have demonstrated their efficacy on postural control, balance, walking speed, and manual dexterity [23,24,49]. In the study conducted by Yasukawa et al., in which DEFOs were implemented for UL in two cases with CP with hemiplegia and brachial plexus palsy, improved limb alignment and improved functionality of the affected UL were observed [18]. In the same way, they have also been effective in improving balance and walking speed in people with MS, as well as in improving pain and functionality of the lower limb in people with CRPS [20,25,26]. In a single case study conducted by Watson et al., the beneficial functional effects of a lycra orthosis in a multiple sclerosis patient were equivocal [50]. In another study of 16 patients with hemiparesis resulting from brain damage, the use of these devices showed a reduction in muscle tone and swelling, and improved wrist and finger movement [51]. Although some studies have shown that the use of these devices in people with brain damage improve strength, manual dexterity, and UL functionality, there is a need for studies with larger sample sizes [27,28]. These results coincide with those observed in the present study, in the sense that the implementation of the orthosis showed improvements on motor aspects of UL such as resting tremor, rhythm of hand movements or speed of rapid alternating movements, assessed with Kinesia, leading to an improvement in manual dexterity. No differences were observed in PDQ-39 and UPDRS-III scores on quality of life and functionality after orthosis implementation. Regarding the use of orthoses for tremor reduction, the review by Fromee et al. showed that the implementation of orthoses had a positive effect on the reduction in involuntary movement, being a complementary device to medical treatment. However, these orthoses turn out to be difficult to handle and unattractive, so they often lead to rejection by the patient. Therefore, there is a need to design orthosis that combines a tremor suppression mechanism with a soft, compact, and lightweight suppression system that increases patient acceptance [52]. Similarly in the review conducted by Mo et al., it was concluded that weight reduction in wearable orthosis for tremor reduction is an important research priority, as they have so far only been evaluated in patient cohorts or on the bench with simulated data and with very small samples, which may weaken the reliability of the data [53]. These findings should be considered within the context of their strengths and limitations; the results show new information about the efficacy of this type of orthosis in patients with PD. On the other hand, the evaluations have been carried out in the “ON” and “OFF” state of the disease, which gives us information on its effect in the different states of the disease. This device has proven to be a non-pharmacological treatment that is easy to implement, with high adherence to treatment and without any type of contraindication. With respect to the limitations, the nature of the intervention was such that the participants and investigators of the initial evaluation were not blinded, and it has not been possible to ascertain whether the results have been maintained in the long term due to the limited duration of the study. ## 5. Conclusions The DEFO is a lightweight and easy-to-implement device. As a non-pharmacological treatment, it can be complementary to medication for the improvement of the motor aspects of UL in PD. Non-pharmacological interventions show promise in PD and need further studies. ## References 1. Dorsey E.R., Sherer T., Okun M.S., Bloemd B.R.. The Emerging Evidence of the Parkinson Pandemic. J. Parkinsons. Dis. (2018.0) 8 S3-S8. DOI: 10.3233/JPD-181474 2. Sirajo M.U., Murtala K., Oyem J.C., Ishola A.O., Owolabi L.F.. Motor function test protocol for parkinsonian triad in rodent model of Parkinson’s disease. J. Neurobehav. Sci. (2022.0) 9 1 3. What Is Parkinson’s Disease? 4. Jankovic J., Tan E.K.. Parkinson’s disease: Etiopathogenesis and treatment. J. Neurol. Neurosurg. Psychiatry (2020.0) 91 795-808. DOI: 10.1136/jnnp-2019-322338 5. Kuhlman G.D., Flanigan J.L., Sperling S.A., Barrett M.J.. Predictors of health-related quality of life in Parkinson’s disease. Park. Relat. Disord. (2019.0) 65 86-90. DOI: 10.1016/j.parkreldis.2019.05.009 6. Forhan M., Gill S.V.. Obesity, functional mobility and quality of life. Best Pract. Res. Clin. Endocrinol. Metab. (2013.0) 27 129-137. DOI: 10.1016/j.beem.2013.01.003 7. Raggi A., Leonardi M., Covelli V., Sattin D., Scaratti C., Schiavolin S., Willems M., Meucci P.. The ICF as a framework to collect and interpret data on the extent and variety of disability in neurological conditions. NeuroRehabilitation (2015.0) 36 17-22. DOI: 10.3233/NRE-141186 8. McQuillen A.D., Licht M.H., Licht B.G.. Contributions of disease severity and perceptions of primary and secondary control to the prediction of psychosocial adjustment to Parkinson’s disease. Health Psychol. (2003.0) 22 504-512. DOI: 10.1037/0278-6133.22.5.504 9. Raggi A., Covelli V., Leonardi M., Meucci P., Scaratti C., Schiavolin S., Willems M., Sattin D.. Determinants of disability using count-based approaches to ICF-based definition of neurological disability. NeuroRehabilitation (2015.0) 36 23-29. DOI: 10.3233/NRE-141187 10. King L.A., Horak F.B.. Delaying mobility disability in people with Parkinson disease using a sensorimotor agility exercise program. Phys. Ther. (2009.0) 89 384-393. DOI: 10.2522/ptj.20080214 11. Deal L.S., Flood E., Myers D.E., Devine J., Gray D.L.. The Parkinson’s Disease Activities of Daily Living, Interference, and Dependence Instrument. Mov. Disord. Clin. Pract. (2019.0) 6 678. DOI: 10.1002/mdc3.12833 12. Shibley R., Griffin H.J., Quinn N.P., Jahanshahi M.. Quality of life in Parkinson’s disease: The relative importance of the symptoms. Mov. Dis. (2008.0) 23 1428-1434 13. Cervantes-Arriaga A., Rodríguez-Violante M., Yescas P., Alonso-Vilatela E.. Fluctuaciones motoras. Rev. Med. Inst. Mex. Seguro Soc. (2012.0) 50 141-146. PMID: 22882981 14. Oertel W., Schulz J.B.. Current and experimental treatments of Parkinson disease: A guide for neuroscientists. J. Neurochem. (2016.0) 139 325-337. DOI: 10.1111/jnc.13750 15. Ellis J.M., Fell M.J.. Current approaches to the treatment of Parkinson’s Disease. Bioorg. Med. Chem. Lett. (2017.0) 27 4247-4255. DOI: 10.1016/j.bmcl.2017.07.075 16. Van de Weijer S.C.F., Hommel A., Bloem B.R., Nonnekes J., De Vries N.M.. Promising non-pharmacological therapies in PD: Targeting late stage disease and the role of computer based cognitive training. Park. Relat. Disord. (2018.0) 46 S42-S46. DOI: 10.1016/j.parkreldis.2017.09.002 17. Ahn S., Chen Y., Bredow T., Cheung C., Yu F.. Effects of Non-Pharmacological Treatments on Quality of Life in Parkinson’s Disease: A Review. J. Park. Dis. Alzheimer′s Dis. (2017.0) 4 1-10. DOI: 10.1093/geroni/igx004.1184 18. Yasukawa A., Uronis J.. Effectiveness of the Dynamic Movement Orthosis Glove for a Child with Cerebral Palsy Hemiplegia and Obstetric Brachial Plexus Palsy. J. Prosthet. Orthot. (2014.0) 26 107-112. DOI: 10.1097/JPO.0000000000000022 19. Betts L.. Dynamic movement lycra orthosis in multiple sclerosis. Br. J. Neurosci. Nurs. (2015.0) 11 60-64. DOI: 10.12968/bjnn.2015.11.2.60 20. Powell J.E., Boehm J.O., Bicher J.H., Reece C.L., Davis S.A., Pasquina P.F.. The Utility of Dynamic Movement Orthoses in the Management of Complex Regional Pain Syndrome—A Case Series. Mil. Med. (2021.0). DOI: 10.1093/milmed/usab418 21. Duncan C.C., Edgley S.R.. Upper Limb Orthoses for the Stroke- and Brain-Injured Patient. Atlas Orthoses Assist. Devices (2019.0) 1 146-156 22. Miller L., Van Wijck F., Lamont L., Preston J., Hair M.. Sensory dynamic orthoses in mild to moderate upper limb tremor in multiple sclerosis: A mixed methods feasibility study. Clin. Rehabil. (2016.0) 30 1060-1073. DOI: 10.1177/0269215515605356 23. Pavão S.L., Visicato L.P., Da Costa C.S.N., De Campos A.C., Rocha N.A.C.F.. Effects of Suit-Orthosis on Postural Adjustments during Seated Reaching Task in Children with Cerebral Palsy. Pediatr. Phys. Ther. (2018.0) 30 231-237. DOI: 10.1097/PEP.0000000000000519 24. Giray E., Karadag-Saygi E., Ozsoy T., Gungor S., Kayhan O.. The effects of vest type dynamic elastomeric fabric orthosis on sitting balance and gross manual dexterity in children with cerebral palsy: A single-blinded randomised controlled study. Disabil. Rehabil.* (2018.0) 42 410-418. DOI: 10.1080/09638288.2018.1501098 25. Keyvani H.A., Aminian G., Azimian M., Bah-Ramizadeh M., Safaeepour Z., Biglarian A., Keivani M.. Design and Preliminary Evaluation of a New Ankle Foot Orthosis on Kinetics and Kinematics parameters for Multiple Sclerosis Patients. J. Biomed. Phys. Eng. (2020.0) 10 792. PMID: 33364217 26. Parian S.S., Fereshtenejad N., Hillier S., Sadeghi-Demneh E.. The Comparison of the Effects of Flexible vs Rigid Ankle-foot orthoses on Balance and Walking Performance in Individuals With Multiple Sclerosis: A Crossover Study. Iran. Rehabil. J. (2021.0) 19 199-206. DOI: 10.32598/irj.19.2.1421.2 27. González-Bernal J., de la Fuente-Anuncibay R., González-Santos M.J., Cuesta-Gómez J., Cubo E.. Efficacy of a dynamic orthosis on the upper limbs in the chronic phase of strokes. A longitudinal study. Rev. Neurol. (2017.0) 65 209-215. PMID: 28849862 28. Alexander J., Dawson J., Langhorne P.. Dynamic hand orthoses for the recovery of hand and arm function in adults after stroke: A systematic review and meta-analysis of randomised controlled trials. Top. Stroke Rehabil. (2022.0) 29 114-124. DOI: 10.1080/10749357.2021.1878669 29. Nguyen H.S., Luu T.P.. Tremor-Suppression Orthoses for the Upper Limb: Current Developments and Future Challenges. Front. Hum. Neurosci. (2021.0) 15 186. DOI: 10.3389/fnhum.2021.622535 30. Skorvanek M., Goldman J.G., Jahanshahi M., Marras C., Rektorova I., Schmand B., van Duijn E., Goetz C.G., Weintraub D., Stebbins G.T.. Global scales for cognitive screening in Parkinson’s disease: Critique and recommendations. Mov. Disord. (2018.0) 33 208-218. DOI: 10.1002/mds.27233 31. Postuma R.B., Berg D., Stern M., Poewe W., Olanow C.W., Oertel W., Obeso J., Marek K., Litvan I., Lang A.E.. MDS clinical diagnostic criteria for Parkinson’s disease. Mov. Disord. (2015.0) 30 1591-1601. DOI: 10.1002/mds.26424 32. Armstrong M.J., Okun M.S.. Diagnosis and Treatment of Parkinson Disease: A Review. JAMA (2020.0) 323 548-560. DOI: 10.1001/jama.2019.22360 33. Declaración de Helsinki de la AMM—Principios éticos para las Investigaciones Médicas en Seres Humanos. (2017.0) 34. Aman J.E., Elangovan N., Yeh I.L., Konczak J.. The effectiveness of proprioceptive training for improving motor function: A systematic review. Front. Hum. Neurosci. (2015.0) 8 1075. DOI: 10.3389/fnhum.2014.01075 35. Martinez-Martin P., Rodriguez-Blazquez C., Alvarez-Sanchez M., Arakak T., Bergareche-Yarza A., Chade A., Garretto N., Gershanik O., Kurtis M.M., Martinez-Castrillo J.C.. Expanded and independent validation of the Movement Disorder Society-Unified Parkinson’s Disease Rating Scale (MDS-UPDRS). J. Neurol. (2013.0) 260 228-236. DOI: 10.1007/s00415-012-6624-1 36. Goetz C.C.. The Unified Parkinson’s Disease Rating Scale (UPDRS): Status and recommendations. Mov. Disord. (2003.0) 18 738-750. PMID: 12815652 37. Steffen T., Seney M.. Test-retest reliability and minimal detectable change on balance and ambulation tests, the 36-item short-form health survey, and the unified Parkinson disease rating scale in people with parkinsonism. Phys. Ther. (2008.0) 88 733-746. DOI: 10.2522/ptj.20070214 38. Hagell P., Nygren C.. The 39 item Parkinson’s disease questionnaire (PDQ-39) revisited: Implications for evidence based medicine. J. Neurol. Neurosurg. Psychiatry (2007.0) 78 1191-1198. DOI: 10.1136/jnnp.2006.111161 39. Bushnell D.M., Martin M.L.. Quality of life and parkinson’s disease: Translation and validation of the US Parkinson’s Disease Questionnaire (PDQ-39). Qual. Life Res. (1999.0) 8 345-350. DOI: 10.1023/A:1008979705027 40. Isaacson S.H., Boroojerdi B., Waln O., McGraw M., Kreitzman D.L., Klos K., Revilla F.J., Heldman D., Phillips M., Terricabras D.. Effect of using a wearable device on clinical decision-making and motor symptoms in patients with Parkinson’s disease starting transdermal rotigotine patch: A pilot study. Park. Relat. Disord. (2019.0) 64 132-137. DOI: 10.1016/j.parkreldis.2019.01.025 41. Proud E.L., Miller K.J., Bilney B., Morris M.E., McGinley J.L.. Construct validity of the 9-Hole Peg Test and Purdue Pegboard Test in people with mild to moderately severe Parkinson’s disease. Physiotherapy (2020.0) 1 202-208. DOI: 10.1016/j.physio.2019.12.002 42. Lawson I.. Purdue Pegboard Test. Occup. Med. (Chic Ill) (2019.0) 69 376-377. DOI: 10.1093/occmed/kqz044 43. Tesio L., Simone A., Zebellin G., Rota V., Malfitano C., Perucca L.. Bimanual dexterity assessment: Validation of a revised form of the turning subtest from the Minnesota Dexterity Test. Int. J. Rehabil. Res. (2016.0) 39 62. DOI: 10.1097/MRR.0000000000000145 44. Soke F., Colakoglu B.D., Keskinoglu P., Genc A.. Reliability, validity and responsiveness of the squares test for manual dexterity in people with Parkinson’s disease. Clin. Neurol. Neurosurg. (2019.0) 186 105542. DOI: 10.1016/j.clineuro.2019.105542 45. Dorsey R.E., Elbaz A., Nichols E., Abd-Allah F., Abdelalim A., Adsuar J.C., Ansha M.G., Brayne C., Choi J.-Y.J., Collado-Mateo D.. Global, regional, and national burden of Parkinson’s disease, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. (2018.0) 17 139-153. DOI: 10.1016/S1474-4422(18)30295-3 46. Koerts J., König M., Tucha L., Tucha O.. Working capacity of patients with Parkinson’s disease—A systematic review. Park. Relat. Disord. (2016.0) 27 9-24. DOI: 10.1016/j.parkreldis.2016.03.017 47. Raggi A., Leonardi M., Ajovalasit D., Carella F., Soliveri P., Albanese A., Romito L.. Disability and profiles of functioning of patients with Parkinson’s disease described with ICF classification. Int. J. Rehabil. Res. (2011.0) 34 141-150. DOI: 10.1097/MRR.0b013e328344ae09 48. Zhao N., Yang Y., Zhang L., Zhang Q., Balbuena L., Ungvari G.S., Zang Y.-F., Xiang Y.-T.. Quality of life in Parkinson’s disease: A systematic review and meta-analysis of comparative studies. CNS Neurosci. Ther. (2020.0) 27 270-279. DOI: 10.1111/cns.13549 49. Matthews M.J., Watson M., Richardson B.. Effects of dynamic elastomeric fabric orthoses on children with cerebral palsy. Prosthet. Orthot. Int. (2009.0) 33 339-347. DOI: 10.3109/03093640903150287 50. Watson M.J., Crosby P., Matthews M.. An evaluation of the effects of a dynamic lycra. Brain Inj. (2009.0) 21 753-761. DOI: 10.1080/02699050701481613 51. Gracies J.M., Marosszeky J.E., Renton R., Sandanam J., Gandevia S.C., Burke D.. Short-term effects of dynamic lycra splints on upper limb in hemiplegic patients. Arch. Phys. Med. Rehabil. (2000.0) 81 1547-1555. DOI: 10.1053/apmr.2000.16346 52. Fromme N.P., Camenzind M., Riener R., Rossi R.M.. Need for mechanically and ergonomically enhanced tremor-suppression orthoses for the upper limb: A systematic review. J. NeuroEng. Rehabil. (2019.0) 16 1-15. DOI: 10.1186/s12984-019-0543-7 53. Mo J., Priefer R.. Medical Devices for Tremor Suppression: Current Status and Future Directions. Biosensors (2021.0) 11. DOI: 10.3390/bios11040099
--- title: Anoxia Rapidly Induces Changes in Expression of a Large and Diverse Set of Genes in Endothelial Cells authors: - Antonella Antonelli - Emanuele Salvatore Scarpa - Santina Bruzzone - Cecilia Astigiano - Francesco Piacente - Michela Bruschi - Alessandra Fraternale - Christian A. Di Buduo - Alessandra Balduini - Mauro Magnani journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049254 doi: 10.3390/ijms24065157 license: CC BY 4.0 --- # Anoxia Rapidly Induces Changes in Expression of a Large and Diverse Set of Genes in Endothelial Cells ## Abstract Sinusoidal endothelial cells are the predominant vascular surface of the bone marrow and constitute the functional hematopoietic niche where hematopoietic stem and progenitor cells receive cues for self-renewal, survival, and differentiation. In the bone marrow hematopoietic niche, the oxygen tension is usually very low, and this condition affects stem and progenitor cell proliferation and differentiation and other important functions of this region. Here, we have investigated in vitro the response of endothelial cells to a marked decrease in O2 partial pressure to understand how the basal gene expression of some relevant biological factors (i.e., chemokines and interleukins) that are fundamental for the intercellular communication could change in anoxic conditions. Interestingly, mRNA levels of CXCL3, CXCL5, and IL-34 genes are upregulated after anoxia exposure but become downmodulated by sirtuin 6 (SIRT6) overexpression. Indeed, the expression levels of some other genes (such as Leukemia Inhibitory Factor (LIF)) that were not significantly affected by 8 h anoxia exposure become upregulated in the presence of SIRT6. Therefore, SIRT6 mediates also the endothelial cellular response through the modulation of selected genes in an extreme hypoxic condition. ## 1. Introduction In the bone marrow (BM) hematopoietic niche, the oxygen tension is usually very low, and this condition affects stem and progenitor cell proliferation and differentiation. The partial pressure of oxygen (pO2) at a cellular level regulates hematopoietic growth factors, chemokines, and adhesion molecule expression that, in turn, affect the proliferation and maturation of other cellular components of the niche. Sinusoidal endothelial cells are the predominant vascular surface of the BM and constitute the functional hematopoietic niche. These cells comprise the platform where trafficking into and out of the BM occurs and where hematopoietic stem and progenitor cells harbor and receive cues for self-renewal, survival, and differentiation. Several authors have contributed in defining niches and mobilization pathways for hematopoietic stem and progenitor cells; this includes the identification of some of the cell types involved, such as osteoblasts, adventitial reticular cells, endothelial cells, monocytic cells, and granulocytic cells, as well as the main factors that anchor stem cells in the niche and/or induce their quiescence, such as vascular cell adhesion molecule (VCAM)-1, CD44, hematopoietic growth factors (e.g., stem cell factor (SCF)), and chemokines, including IL-12 and IL-8 [1]. While elucidating the basic mechanisms of intracellular cross-talk between the different cell components of the vascular niche, and considering the role of oxygen in the hematopoietic niche, we recently demonstrated, through an in vitro anoxic model using monocyte–macrophage-type cells representative of components of the hematopoietic niche, that modulation of pO2 has an effect on the modulation of the expression of some cytokines [2]. Reduced oxygen tension has been shown to enhance the production of erythroid, megakaryocytic, and granulocytic–monocytic progenitors in vitro. Due to the inaccessibility of the bone marrow to direct noninvasive oxygen measurements, some authors have used mathematical modeling of pO2 distributions in the bone marrow and speculated that stem cells are located at regions with very low pO2 levels (almost anoxic) because this prevents oxygen radicals from damaging these important cells [3]. The modulating effects of vascular endothelial growth factor (VEGF) are essentially limited to endothelial cells, the only cell type consistently shown to express VEGF receptors. It has already been demonstrated that the induction of VEGF mRNA in Human Umbilical Vein Endothelial Cells (HUVECs) was dependent on the degree of hypoxic stimulation [4,5]. Here, we report data collected during studies performed using an in vitro cell model of anoxia. HUVECs were chosen as a cell type that can mimic endothelial cells present around the sinusoidal vessels of the vascular hematopoietic niche that live at low O2 conditions. We investigated in vitro how an extreme reduction in O2 percentage of the HUVEC environment can alter the basal expression of some genes. Moreover, we determined the levels of chemokines, interleukins, and other molecules that are affected by pO2 changes in the culture medium. We started by investigating the cellular response to anoxic conditions following Vascular Endothelial Growth Factor (VEGF) gene expression given its role as a key regulator of angiogenesis and hematopoiesis [6,7]. The first studies determined the length of anoxic exposure required to obtain maximum levels of VEGFA mRNA expression (8 h of anoxia) in our HUVEC model; all subsequent scientific evaluations here reported were performed using this time point. As it is known that VEGFA is one of the genes upregulated by HIF-1α functionally induced in low-oxygen conditions, we analyzed HIF1A mRNA and HIF-1α protein levels in cell extracts of HUVECs after 8 h of anoxia. Under normoxic conditions, the α subunit of HIF-1 is hydroxylated by prolyl hydroxylases (PHDs), which causes it to be recognized by the protein product of the von-Hippel-Lindau (VHL) gene, ubiquitinated, and degraded by the proteasome [8]. When cells are at a low O2 concentration, the PHDs are not active; consequently, HIF-1α is not degraded but can translocate to the nucleus, where it can dimerize with the β subunit. The heterodimeric transcription factor induces the transcription of genes mediating cellular adaptation to a low oxygen environment. Moreover, it is known that SIRT6, a member of the NAD-dependent deac(et)ylases sirtuin family, is involved in the regulation of several different biological processes, including metabolism, DNA repair, and aging and can regulate HIF-1α in different ways. For example, it has been demonstrated that SIRT6 functions as a co-repressor of HIF-1α transcriptional activity, with metabolic effects such as inhibition of glucose uptake and glycolysis [9]. Recent research has also shown how SIRT6 under oxidative stress prevents the degradation of HIF-1α by deubiquitination at two specific K residues, thereby promoting angiogenesis [10]. Overall, the aim of this study was to investigate how a marked decrease in O2 affects gene expression in HUVECs, with a focus on the production of molecules such as chemokines and interleukins and on the ability of SIRT6 to modulate the response when endothelial cells are found at extremely low O2 levels. ## 2. Results and Discussion In our experimental cell model, we have taken the hypoxic condition to the extreme. By flushing an appropriate $95\%$ N2 and $5\%$ CO2 gas mixture in the Hypoxia Incubator Chamber, endothelial cells were exposed to very low pO2 levels (almost 0–$0.5\%$ oxygen or anoxia). Some authors have hypothesized that regions with very low pO2 levels (almost anoxic) exist in the bone marrow [11] where they prevent oxygen radicals from damaging stem cells [3]. We used an anoxic cell model to investigate how sinusoidal vessel endothelial cells living in the vascular hematopoietic niche could also be affected by very low pO2 levels leading to the up- or downregulation of specific gene targets. To test for anoxia-specific regulation, we followed VEGFA gene expression as a biomarker. The mRNA levels of VEGFA were analyzed by RT-q-PCR in cell extracts of HUVECs maintained for different times (2, 4, 6, 8, 21 and 24 h) under anoxic conditions obtained by using the Hypoxic incubator Chamber as reported in the Materials and Methods. Cells maintained in normoxia ($21\%$ O2) were used as control samples. Figure 1 shows a representative time-course experiment performed with the HUVEC cell model exposed to anoxia. The results indicate an upregulation effect on VEGF gene expression, whose mRNA levels were already increased after 2 h of treatment (1.28 ± 0.29 fold), reaching a maximum peak of expression from 6 to 8 h (1.92 ± 0.28 fold, $$p \leq 0.02$$) and returning to basal levels after 24 h. To validate the time-point showing the highest peak of VEGFA expression, we performed additional experiments to evaluate VEGFA and HIF1A mRNA levels at 8 h under anoxic conditions. Figure 2a shows the increased VEGF mRNA expression reaching a 2.35 ± 0.29-fold change, confirming the data obtained during the time course (Figure 1); interestingly, real-time quantitative PCR analysis performed in parallel in the same cell extracts derived from anoxic and normoxic cell samples to determine the gene expression of HIF1A (since hypoxia-inducible factor 1 α is the inducible subunit of the HIF-1 transcription factor) showed a decrease of HIF1A mRNA levels (0.23 ± 0.05-fold) in cell extracts of 8 h anoxic HUVEC samples with respect to control cells (Figure 2b). Moreover, HIF-1α protein was detected by western immunoblotting analysis; as expected, HIF-1α was accumulated in cell cytoplasm upon HUVEC challenged with anoxia (Figure 2c). Data support the involvement of this important transcriptional effector regulating the responses of HUVECs exposed to anoxic stimulus, evidencing an accumulation of a 120 kDa protein band in 8 h anoxia-treated samples in comparison with cells maintained for the same time in normoxic conditions. These results are in agreement with the literature; when oxygen is not present in adequate amounts, hydroxylation cannot occur, which means that HIF1-α can be stably expressed [12]. As a result, this protein can bind to HIF1β and facilitate gene transcription of many HIF target genes. Despite this increased amount of HIF 1α protein being clearly evident after an 8 h anoxia treatment, the levels of HIF1A mRNA do not appear to be upregulated; on the contrary, in comparison with basal levels of expression, a downregulation occurs. It is commonly accepted that under low oxygen (hypoxic) conditions, the hydroxylase activity of the PHD enzymes is inhibited; HIF therefore escapes hydroxylation and degradation to initiate a transcriptional program of cellular response and adaptation to hypoxia. However, it was reported in the literature that in some cases, negative regulatory feedback occurs [13]. Moreover, some other authors had already reported that the inhibition of HIF1A transcription is necessary to counteract the transcription-dependent degradation of HIF-1α which occurs during hypoxic/anoxic conditions. Under normoxia, HIF1-α is expressed at low levels and is degraded via the classic PHD/VHL pathway and therefore does not activate the feedback loop. Instead, under anoxia/hypoxia, HIF1-α accumulates and transcriptionally activates its own degradation that is independent from the PHD/VHL pathway. Thus, our data, indicating that in our cell model, HIF1A mRNA expression levels significantly decrease in 8 h anoxia-treated HUVECs (Figure 2b), further promoting HIF1-α accumulation in HUVEC cells (Figure 2c), are in line with previous reports [14]. ## 2.1. Transcriptional Analysis and Evaluation of Chemokine Release RNA-Seq analysis provided us the top 50 most variable genes found in the total RNA extracted from HUVECs treated or not in anoxia for 8 h as reported in Material and Methods section. A representation of a generated heat map evidencing up- or downregulated genes in anoxic HUVEC sample (S2) respect to normoxic HUVEC sample (S1) was reported in Figure 3. Fold changes in gene expression of the 50 most variable genes screened (Figure 4) are expressed as anoxic (S2) versus normoxic (S1) condition (Figure 4a). In addition to the 50 most variable genes derived from BMR Genomics data, we calculated the fold-changes of those genes that most seemed to be involved in the regulation of the hematopoietic niche environment, such as VEGFA, as well as those that mediate the cellular response to anoxia/hypoxia. An increase of at least 1.5-fold and a decrease of at least 0.6-fold in the mRNA levels with respect to basal levels in normoxic condition were considered. The expression of many mediators, such as interleukins including IL-2, IL-3, IL-5, IL-6, IL-8, and factors such as PDGFB (Platelet Derived Growth Factor Subunit B), UGCG (UDP-Glucose Ceramide Glucosyltransferase), P2RY11 (purinergic receptor), and VEGFA (Vascular Endothelial Growth Factor A) were found to be upregulated (>2-fold change), whereas a downregulation of some other genes (<0.5-fold change), such as CCL2 (C-C Motif Chemokine Ligand 2), KITLG (KIT Ligand) and CSF3 (Colony Stimulating Factor 3 Receptor), was evidenced (Figure 4b). The secretome profile of anoxia-exposed HUVECs was compared with that of HUVECs kept in normoxic conditions via a BIO-Rad 20-plex detection kit as reported in the Materials and Methods section. The detection of some secreted human interleukins/cytokines was reported in Table 1. VEGF was not considered since it is a supplement of the HUVEC-specific culture medium. The data highlighted a strong increase in protein release, particularly of the inducible protein 10 (IP10) (>50 fold), IL-6 (>7 fold), macrophage chemoattractant protein-1 (MCP-1) (>5 fold), and IL-5 (>2.2 fold), while IL-1b and IL-2 secretion as well as the platelet-derived growth factor (PDGF-BB) appeared to be unaffected by the anoxic environment, with their amounts in the respective cell supernatants being borderline detectable. Since some chemokines were not detected by the BIO-Plex kit, only data of the most reliable factors were reported in Table 1; moreover, a comparison between released proteins and their corresponding mRNA expression was not possible for all the target genes obtained with RNA-Seq reported in Figure 4. For example, the mRNA expression levels for IL34 and CXCL3 could not be correlated to the respective protein amounts released in cellular supernatants due to the absence of these factors in the BIO-Plex kit; conversely, for IP-10, the protein amount in the cell supernatant was assessed, but not the mRNA by RNA-Seq. It is evident that in our experimental cell model, the anoxia stimulus leads to a modulation of mRNA expression of key cytokines having a role in the control of cellular metabolism, energetics, and post-transcriptional gene regulation by O2. Thus, understanding the effect of low pO2 levels on endothelial cells could be important to identify which biological factors are involved in the anoxic response with the aim of investigating the possible cell activity of sinusoid endothelial cells of hematopoietic tissues living in extreme hypoxia conditions. Significantly increased CXCL2, CXCL8, CXCL3, CXCL5, PDGFB, and VEGFA mRNA levels (≥2-fold) detected by transcriptional analyses could be connected to the modulation of specific biological pathways in anoxic/hypoxic conditions since oxygen tension (pO2) is an important determinant of hematopoietic stem and progenitor cell proliferation and differentiation. For example, IL-3 was one of the earliest cytokines implicated in self-renewal, proliferation, and differentiation of primitive and mature hematopoietic stem cells. Although the role of IL-3 in the bone marrow has not been fully unraveled, it is a cytokine with multilineage potential, and it is known that its action shows functional redundancy with other cytokines such as the granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-5. IL-3 was shown to increase and decrease the expansion and/or self-renewal capacity of adult HSCs; however, it was also evidenced that IL-3 supports the growth of early hematopoietic progenitors and promotes their response to other, later-acting cytokines [15]. Moreover, IL-3 was reported to be able to induce the expression of interleukin 2 (IL-2) receptor (IL-2R) (CD25) on a subset of early myeloid cells in normal human bone marrow that had been first depleted of mature hematopoietic cells [16]. The role of IL-2 in HSC maintenance is unknown and some studies suggest that equilibrium between IL-2 and IFN-γ is critical for steady state hematopoiesis [17]. Abundant literature reports VEGF as a potent mitogen and permeability factor for endothelial cells playing a central role in angiogenesis, inflammation, and cancer. VEGF also mediates the homeostatic adaptation to hypoxic conditions by promoting an increase in vascular density to compensate for decreased oxygenation; therefore, the synthesis and secretion of VEGF is required for the maintenance of the integrity of the vascular system. Furthermore, CXCL8/IL8, a potent proangiogenic and inflammatory chemokine, upregulates VEGF mRNA and protein levels in endothelial cells by acting on its receptor, CXCR2 (receptor for IL-8), and this results in the autocrine activation of VEGFR2 (VEGF receptor 2) [18,19]. IL-34 mRNA levels increased by 2.38-fold in anoxic conditions (Figure 4b). Structurally, IL-34 belongs to the short-chain helical hematopoietic cytokine family but shows no apparent consensus structural domains, motifs, or sequence homology with the other 53 cytokines. IL-34 is a novel cytokine that was identified in 2008 in a comprehensive proteomic analysis as a tissue-specific ligand of the CSF-1 receptor (CSF-1R). This cytokine is involved in several processes; it promotes the proliferation, survival, and differentiation of monocytes and macrophages and also plays an important role in innate immunity and in inflammatory processes promoting the release of proinflammatory chemokines [20]. The pro-angiogenic potential of IL-34 was further confirmed experimentally in vitro, where IL-34 was able to recruit endothelial cells to form vascular structures. In the HUVEC endothelial cell line, IL-34 activates several kinases, including PI3K, Src, FAK, and ERK$\frac{1}{2}$, which importantly contribute to cell differentiation into vascular cords. These effects were mediated by the interaction between IL-34 and glycosaminoglycans at the cellular surface, indicating a CSF-1R-independent mechanism [21]. Additionally, IL-34 can exert additional pro-angiogenic functions via CSF-1R-dependent mechanisms by inducing the secretion of several factors that contribute to angiogenesis, such as IP-10, MCP-1, and IL-8 in PBMCs [22]. The interferon-gamma-induced protein 10 (IP-10), also classified as CXCL10, has been described as a chemokine produced by T cells, monocytes, endothelial cells and keratinocytes after stimulation with IFN-γ. In our setting, IP-10 was found to be significantly increased in cell supernatant after 8 h of anoxic stimulus (>50-fold with respect to control cells). Although IP-10 was named as a protein inducible by IFN-γ [23], other agents, including LPS, IL-1α, IL-6, TNF-α, IFN-α, and IFN-β, were also found to highly induce IP-10 expression in vitro in a variety of cells, including endothelial cells. Despite its well-described function of recruiting leukocytes to sites of inflammation, IP-10 also plays a role in the generation and function of effector cells [24]. Some authors have studied the potential molecular mechanism by which hypoxia/ischemia regulates expression of CXCL10 in endothelial cells, especially in the cardiac microvascular endothelial cells. CXCL10 secretion and mRNA expression were increased by hypoxia/ischemia treatment in a time-dependent manner [25]. Monocyte chemoattractant protein-1 (MCP-1/CCL2), that was also found to be secreted after 8 h anoxia in cell supernatants (5-fold increased with respect to control samples), is one of the key chemokines that regulates migration and infiltration of monocytes/macrophages. Both CCL2 and its receptor CCR2 have been demonstrated to be induced and involved in various diseases. Migration of monocytes from the blood stream across the vascular endothelium is required for routine immunological surveillance of tissues, as well as in response to inflammation. The effect of MCP-1 on the migration and proliferation of hematopoietic progenitor cells has been examined [26,27]. Some work reported that cardiac stem cells have been shown to play a protective role against hypoxia-induced injury in vivo via an MCP-1-dependent mechanism [28]. Moreover, an increased release of IL-6 (>7-fold versus control sample) was also found in the cell medium after 8 h under anoxic conditions. It was reported that in hypoxic conditions, increased levels of the IL-6 protein induce changes in endothelial permeability. In fact, IL-6 signaling mediates a vast array of effects in the vascular wall, including endothelial activation, vascular permeability, immune cell recruitment, endothelial dysfunction, as well as vascular hypertrophy and fibrosis [29]. IL-6 production can affect both stromal and hematopoietic cells, for example, stimulating megakaryocyte (Mk) growth and maturation in vitro as well as increasing Mk ploidy. Moreover, some authors have recently reported that abnormal levels of IL-6 might interfere with the stability of the bone marrow hematopoietic microenvironment [30,31]. It was already reported that IL-8 is also responsible for the enhanced proliferation and mobilization of HSCs in the bone marrow [32]. In our anoxic cell model, increased IL-8 mRNA and protein levels were found (Figure 4b and Table 1). The role of IL-8 as a mediator of angiogenesis and tumor growth has already been described. Increased IL-8 levels in human umbilical cord segments exposed to hypoxia (compared with normoxic controls) were reported. The incubation of human endothelial cells in hypoxia (PO2 ~14–18 mmHg) led to the time-dependent release of IL-8 antigens into the conditioned medium; this was accompanied by increased chemotactic activity for some polymorphonuclear leukocytes, which was blocked by antibodies to IL-8 [33,34]. It is clear that the most detectable chemokines and interleukins, released in culture medium from the anoxic cell model, are part of those biological molecules and growth factors which play pivotal roles in bone marrow vascular niche regulation; it is well known that the different cell components together all play an essential role in the maintenance of bone marrow vascular niche function by secreting large quantities of several molecules with hematopoietic activity, including VEGF, IL-6 and IL-8 [35,36]. After 8 h anoxia treatment, we also found increased expression levels of other types of molecules, such as the P2Y11 receptor (Figure 4b), a member of the purinergic receptor family. The P2Y11 receptor (P2RY11) has a wide distribution in all cell types relevant for cardiovascular pathology: cardiomyocytes, fibroblasts, endothelial and immune cells [37,38]. Some authors have investigated the role of P2Y11R signaling in vascular dysfunction, (where endothelial cell activation was included) reporting that activation of the P2Y11 receptor (P2Y11R) in human dendritic cells, cardiofibroblasts, and cardiomyocytes was protective against hypoxia/reoxygenation lesions. Thus, P2Y11R activation may also protect blood vessels from vascular injury induced from low pO2 levels. In particular, the P2Y11 receptor, activated by ATP, has anti-inflammatory actions which could be implicated in endothelial cell protection in cardiovascular diseases [36,37,38,39,40,41]. UGCG (UDP-glucose ceramide glucosyltransferase) mRNA levels were also increased (>3-fold change) in HUVECs following anoxia treatment. UGCG is a key enzyme in the sphingolipid metabolism as it generates glucosylceramide, the precursor for all glycosphingolipids, which are essential for proper cell function. Although UGCG has been associated with several cancer-related processes such as maintaining cancer stem cell properties or multidrug resistance induction, the precise mechanisms underlying these processes are unknown [42]. CXCL5 (C-X-C motif chemokine ligand 5) mRNA levels were also increased 3-fold after anoxic stimulus. It is known that anoxia/hypoxia also affects the CXC chemokine system, which leads to changes in the level of these chemoattractant cytokines in the cancer microenvironment; moreover, CXC chemokines differ in their effect on angiogenesis. CXCL5 elicits this effect by interacting with the cell surface chemokine receptor CXCR2. Moreover, this chemokine stimulates the chemotaxis of neutrophils possessing angiogenic properties; increasing evidence has indicated that CXCL5 is involved in the tumorigenesis of various malignancies [43]. Next, we investigated the expression of some genes involved in the metabolism of glutathione due to its relevant role in the maintenance of cellular redox homeostasis and the related susceptibility of endothelial cells to oxidant injury. A major initiator of endothelial injury is oxidative stress, which results from an imbalanced state of increased reactive oxygen species (ROS) generation and insufficient intracellular antioxidant activity. Table 2 shows mRNA level fold-changes calculated using 2log FC and the DEG table of BMR Genomics of some mediators such as those of the glutathione peroxidase (GPXs) family, one of the important antioxidant enzymes and an important reactive oxygen species (ROS) free-radical scavenger in an organism. In particular, Glutathione peroxidase-7 (GPX7), a newly discovered non-selenium-containing protein with glutathione peroxidase activity, showed increased mRNA level expression (fold-change > 2.68) in HUVECs after 8 h of anoxia treatment. Mammalian GPX7 is a non-selenocysteine containing phospholipid hydroperoxide glutathione peroxidase activity that plays an important role in maintaining the redox state. When cells are exposed to oxidative stress, the unique pressure-sensor function of GPX7 can effectively detect redox levels and transmit ROS signals to redox-sensitive cells and thiol-containing proteins, further promoting protein folding and releasing non-targeted short interfering RNAs related to stress [44]. Previous studies have shown that the loss of GPx7 resulted in systemic oxidative stress damage, increased carcinogenesis, and shortened life span in mice [45,46,47,48,49]. In Table 2, there is also a clearly evident increase in the ribonucleoside-diphosphate reductase subunit M1 (RRM1, 2.1-fold) and ribonucleoside-diphosphate reductase subunit M2 (RRM2, 3.6-fold) that appears to be strongly involved in the cellular response to anoxia stimulus. Ribonucleotide reductase (RR), comprising RRM1 and RRM2 subunits, is a unique enzyme that catalyzes the conversion of ribonucleotides (NDPs) to deoxyribonucleotides (dNDPs), which are the building blocks for DNA synthesis and thus essential for cell proliferation. This enzyme catalyzes a crucial step of de novo DNA synthesis by converting ribonucleoside diphosphates to deoxyribonucleoside diphosphates; tight control of the deoxyribonucleotide triphosphate (dNTP) pool is essential for cellular homeostasis [50]. Some papers have questioned the source of dNTPs in hypoxia and discussed probable mechanisms by which the RRM1/RRM2 enzymes are capable of retaining activity in hypoxia in order to preserve ongoing replication and avoid the accumulation of DNA damage [51]. Oxygen is an essential cofactor for mammalian ribonucleotide reductase to be capable of de novo synthesis of dNTPs and the upregulation of RRM1 and RRM2 found in our cell model is in accordance with the literature indicating that HUVEC cells are also able to sustain DNA replication in anoxic conditions. Regeneration of the ribonucleotide reductase is accomplished either by GSH-dependent glutaredoxin (Grx) or thioredoxin (Trx), which have been recognized as the primary electron donors to reduce RR [52]. Electrons for such a reduction come ultimately from NADPH via glutathione reductase (GR) and thioredoxin reductase (TR) [53]. Parallel studies demonstrated how GSH and Trx systems, which play a central role in the cellular redox state, are altered by O2 reduction [54] and are also involved in the control of reactive oxygen species (ROS), which will be investigated afterwards. ## 2.2. GSH Content in HUVECs after Anoxia The exposure to severe hypoxic conditions results in the rapid usage of glutathione, leading to a marked alteration in the redox status of the intracellular milieu [55]. Therefore, GSH levels in HUVEC were evaluated using HPLC at different time points (1, 4, and 8 h), as reported in [56]. GSH content decreased markedly after 1 h in anoxia and a similar profile was detected after 4 h. Only at 8 h were the cells able to restore their baseline levels, probably thanks to the activation of their antioxidant defense systems, as illustrated in Figure 5. GSH could be restored from GSSG via GSR, or through its de novo synthesis (GCL, GSS), although expression of these enzymes seems to be unaffected by anoxic conditions (Table 2). However, due to the complexity of the system, this aspect of modulating GSH synthesis and degradation by anoxia requires further investigation. Importantly, the anoxia-dependent upregulation of the RR and the Gpx7 together with the drop in GSH after 1 h and 4 h presumes a shift to a more oxidized environment within the cells, suggesting that the direct application of redox-active molecules or overexpression of SIRT6 could represent a novel strategy to maintain redox balance and physiological redox signaling. ## 2.3. Reactome Functional Interaction Network We considered it interesting to investigate, through the use of the Reactome Functional Interaction Network (https://reactome.org/PathwayBrowser/#TOOL=AT): URL (accessed on 21 June 2021) which possible pathways were implicated by these upregulated selected genes. The 50 most variable genes as well as ones such as VEGFA, whose increased fold-changes were calculated by RNA-Seq of the available BMR platform, were loaded onto the Reactome software tool for analysis of specific molecular pathways or sub-pathways. Supplementary Materials show how many of the genes we detected with an 8 h anoxia-related increase in expression were present in the 25 most relevant pathways sorted by p-value; for example, 3 of the detected genes were found to belong to the pool of genes involved in the following pathways: [1] cellular response to hypoxia, [2] signaling by VEGF, [3] platelet degranulation, and [4] response to elevated platelet cytosolic Ca2+ pathways. ## 2.4. Results on HUVEC-SIRT6 Transfected Cells after Anoxia Treatment Biologic factors and processes influencing the behavior of hematopoietic cells during mobilization and transplantation have still not been fully discovered and understood; however, it is known that in the complex milieu of components participating in HSC maintenance, survival, proliferation, and differentiation [57,58], different elements such as adhesive molecules, growth factors, cytokines, proteases, miRNAs, and sirtuins are involved [59,60,61]. Among these factors, sirtuins may play an important role in the mobilization of HSCs through their deacetylating activity, which regulates many important processes related to the fate of HSCs, their metabolism, stress response, differentiation, aging, and apoptosis [62,63]. Some recent studies revealed a significant increase in the expression of sirtuins, in particular, SIRT6, that may be associated with the efficacy of hematopoietic stem cell mobilization [64]. SIRT6, whose activity depends on the availability of intracellular NAD+ [65], is involved in the regulation of a wide variety of processes, including metabolism, cancer, and inflammation. Among other SIRT family members, SIRT6 is unique, being endowed with deacetylase activity, but also with de-fatty acylase and mono-ADP-ribosyltransferase activities [66,67,68]. Moreover, it is also known that some sirtuin enzymes, which depend on NAD+ binding, can mediate the transcriptional activation of IL-8 and regulate hypoxia-inducible factor-1 α (HIF-1α) stabilization, which is routinely used to screen for hypoxic conditions [69,70]. HUVECs were engineered to overexpress SIRT6 with the aim to investigate the role of SIRT6 in the release of chemokines or interleukins in this anoxic cell model. Firstly, to evaluate the effective overexpression of SIRT6 in HUVEC-SIRT6wt samples, immunoblotting analysis of the protein was performed on cell extracts (Figure 6). It is evident that Sirtuin 6 protein (42 kDa) was remarkably overexpressed in HUVEC-SIRT6wt samples (lane 3, Figure 6) whereas in control cell samples (HUVEC wt, lane 1; HUVEC-PBP, lane 2), the band protein is undetectable, demonstrating the efficiency of the transfection. SIRT6 mRNA levels in cell extracts of native HUVEC, HUVEC-PBP and HUVEC-SIRT6wt samples were also analyzed both in normoxic and 8 h anoxic conditions (Figure 7). The data showed that 8 h anoxia treatment significantly downregulates HIF1A mRNA levels (0.50 ± 0.08-fold change; $$p \leq 0.019$$) in HUVEC-SIRT6 cells (Figure 7E), whereas there was not a significant decrease (0.84 ± 0.06; $$p \leq 0.057$$) in HUVEC-PBP samples (Figure 7B). Moreover, anoxia exposure did not significantly alter SIRT6 mRNA levels either in HUVEC-SIRT6 samples (0.80-fold change, Figure 7D) or in HUVEC-PBP samples (1.01-fold change, Figure 7A) in comparison with SIRT6 mRNA levels of cells in normoxic conditions. The VEGFA mRNA levels in HUVEC-PBP and HUVEC-SIRT6 samples were also reported; after 8 h anoxia exposure, both cell samples presented a significant increase (3–4-fold change) in VEGFA expression levels (Figure 7C,F). ## 2.5. G6PDH, NADH and ROS Evaluation in HUVEC-SIRT6 Cells Exposed to Anoxic Conditions It was previously demonstrated that SIRT6 can prevent oxidative stress damage through upregulation of NADPH levels as a consequence of glucose-6-phosphate dehydrogenase (G6PD) activation. G6PD activity and NADPH levels were increased in SIRT6-overexpressing MCF-7 cells [71]. In agreement with these data, the overexpression of SIRT6 in HUVEC cells caused an increase in G6PD activity (HUVEC-SIRT6 versus HUVEC-PBP; Figure 8a) and in the NADPH/NADP+ ratio (Ctrl; Figure 8b). During the anoxia exposure ($0.5\%$ O2 for 8 h) of HUVEC-PBP samples (cells transduced with empty vector), the NADPH/NADP+ ratio decreased, consistent with NADPH utilization to reduce GSSG to GSH. The NADPH/NAD+ decline was significantly counteracted by SIRT6 overexpression (HUVEC-SIRT6 in 8 h anoxic conditions; Figure 8b). It is also known that the role of the redox homeostasis in the maintenance of stem cell self-renewal and differentiation of stem cells plays a critical role [72,73]. Previous studies have implicated reactive oxygen species (ROS) and cytokines in the regulation of endothelial permeability. Some authors reported that prolonged hypoxia alters HUVEC permeability to increasing ROS generation, which amplifies cytokine production. IL-6 and IL-8 secretion increased 4-fold over 24 h in a pattern corresponding to changes in HUVEC permeability as measured by transendothelial electrical resistance [74]. In hypoxic conditions, production of ROS, as detected by a fluorescent probe, was significantly reduced in cells overexpressing SIRT6 (Figure 9), suggesting that the increased NADPH availability in SIRT6-overexpressing cells counteracts ROS production. ## 2.6. Transcriptional Analysis of Cell Extracts from 8 h Anoxia-Treated HUVEC-SIRT6 Cells The heatmap generated from transcriptional profiling BMR analysis performed using the total RNA derived from HUVEC-SIRT6 cells exposed for 8 h to anoxic conditions (S6) versus HUVEC-SIRT6 cells in normoxic conditions (S5) is reported in Figure 10, which shows the 50 most variable genes detected. Moreover, we calculated the mRNA fold-change (by using 2log FC and DEG table of the available RNA-Seq BMR platform) of some other genes that, on the basis of the literature, could be involved in cell communication in the hematopoietic niche. We tried to understand which genes were effectively affected by the presence of SIRT6 and to relate this to results obtained with normoxic and anoxic HUVEC-PBP cells, for which the corresponding heatmap is reported in Figure 11. Figure 12 shows the expression levels of some of the genes that were modulated during the 8 h anoxia treatment by the presence of overexpressed SIRT6 in HUVEC cells. For example, we found a downregulation of CXCL3 (C-X-C motif chemokine ligand 3) and CXCL5 (C-X-C motif chemokine ligand 5), with both showing a 1.2-fold change vs. a 2.7-and 2.2-fold change, respectively, in the HUVEC-PBP control samples in anoxic conditions. In particular, IL-34 mRNA levels were remarkably downregulated in HUVEC-SIRT6wt cells (0.3-fold change) compared with the higher level found in control HUVEC-PBP samples (11.5-fold) in anoxic conditions. On the contrary, an upregulation of ST6GALNAC2 (ST6 N-Acetylgalactosamide Alpha-2,6-Sialytransferase 2) was evident in HUVEC-SIRT6 cells (1.82-fold change vs. 0.53-fold in HUVEC-PBP control cells). Another target gene is the Leukemia inhibitory factor (LIF), which was upregulated (2.3-fold) in the presence of overexpressing SIRT6 in comparison with HUVEC-PBP control cells (1.72-fold) after 8 h of anoxia exposure. Some works describe LIF as a member of the interleukin-6 cytokine superfamily; it is a pleiotropic protein expressed in multiple types of tissues and cells which regulates an array of important biological functions. For example, LIF maintains the pluripotency of embryonic stem cells, while it induces the differentiation of several myeloid leukemia cells and inhibits their growth. LIF is frequently overexpressed in a variety of solid tumors including colorectal, breast, and skin cancers. However, the mechanism for LIF overexpression in tumors is not well-understood. Hypoxia is a hallmark of solid tumors, including colorectal cancers. Some authors have studied the effect of hypoxia on LIF expression in colorectal cancer cells. Data clearly demonstrated that hypoxia induces LIF expression, mainly through HIF-2α, and this is an important underlying mechanism for LIF overexpression in human colorectal cancers [75,76]. Using the Reactome software tool, which analyses specific molecular pathways or sub-pathways, we were able to identify in which of the 25 most relevant pathways the modulated genes in anoxic HUVEC-SIRT6 samples were present (Supplementary Materials). At least 3 genes were found in the pool of genes involved in “Cellular response to hypoxia” and “Signaling by VEGF pathways”. The transcriptional analyses performed with the RNA samples derived from HUVEC-PBP and HUVEC-SIRT6 cells in normoxic conditions are reported in Figure 13. It is evident that the 50 most variable genes found are different from those detected in HUVEC-PBP and HUVEC-SIRT6 cells exposed to anoxic conditions. Notably, compared with control HUVEC-PBP samples in normoxic conditions, the genes modulated by the presence of an overexpressed SIRT6 in HUVEC-SIRT6 samples (HUVEC cells transfected with the SIRT6 cDNA construct) are involved in the regulation of pathways such as: [1] processes of DNA damage repair; [2] processing of DNA double strand break ends; [3] homology-directed repair (HDR); [4] HDR through homologous recombination or single-strand annealing; and [5] the Notch signaling pathway (Supplementary Materials). These indicate the role of the SIRT6 enzyme in the mechanisms of DNA damage repair in both physiological and oxidative stress conditions, in accordance with the literature. It is evident that these pathways are completely different from those specifically activated when HUVEC-SIRT6 samples are exposed to anoxic conditions (Supplementary Materials). ## 3.1. Cell Cultures HUVECs purchased from Lonza (Catalog #: C2517A; Human Umbilical Vein Endothelial Cells, Single Donor) were cultured in endothelial cell growth medium (EGM TM-2 BulletKit TM Medium, Lonza, Basel, Switzerland) in 25 cm2 flasks (Cell Star, Greiner bio-one, Kremsmünster, Austria) pre-coated with 25 µg/mL fibronectin (Sigma, St Louis, MO, USA) and maintained at 37 °C in a humidified $5\%$ CO2 atmosphere. Cells were used at passages 4–8. Approximately 2.5 × 106 cells were used for each experimental condition. HUVECs were exposed to anoxia using a Hypoxia Incubator Chamber (Chamber for generation of a hypoxic environment for tissue culture, StemCell Technologies, Vancouver, BC, Canada) flushed with the appropriate $95\%$ N2 and $5\%$ CO2 gas mixture for 15 min to reach anoxic conditions; normoxic cells were maintained in the incubator as control sample while the anoxic condition was maintained for all experimental time points reported in this work. For experiments involving HUVEC-pBABE-puro (PBP) and HUVEC-SIRT6 samples, cells were either engineered to stably overexpress Sirtuin 6 (SIRT6) or with the corresponding empty plasmid (PBP) by retroviral transduction as previously described. Positively infected HUVEC cells were selected with 1 μg/mL puromycin. ## 3.2. Western Blotting Analysis Cells were lysed in buffer containing urea 6M, tiourea 2M, DTT 100 mM, Tris-HCl 30 mM, pH 7.5, triton $1\%$, and glycerol $9\%$ supplemented with protease inhibitors (cComplete Mini; Roche, Basel, Switzerland); lysates were boiled for 7 min, sonicated twice at 100 Watt for 10 s and cleared by centrifugation at 16,000× g for 10 min, and the supernatants were recovered [2]. Appropriate amounts of proteins as determined by the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA, USA) were resolved by $8\%$ SDS polyacrylamide gel electrophoresis (SDS-PAGE) and afterward transferred onto nitrocellulose membranes (100 V, 70 min at 4 °C). The blots were probed with the following primary antibodies: anti-HIF-1α (#14179, monoclonal, recognizing amino acidic residues surrounding Lys460 that is codified by exon 10 of the HIF1A CDS1) and anti-SIRT6 (D8D12 Rabbit mAb #12486) from Cell Signaling Technology (Danvers, MA, USA); anti-β-actin (#VMA00048, monoclonal) from Bio-Rad (Hercules, CA, USA); and anti-Lamin A/C (#sc-376248, monoclonal) from Santa Cruz Biotechnology (Dallas, TX, USA). Immunoreactive bands were detected by horseradish peroxidase (HRP)-conjugated secondary antibodies (Bio-Rad, Hercules, CA, USA). Peroxidase activity was detected with the enhanced chemiluminescence detection method (WesternBright ECL, Advasta, Menlo Park, CA, USA) using the ChemiDoc MP Imaging System (Bio-Rad, Hercules, CA, USA). Quantification of the protein bands was performed using Image Lab analysis software version 5.2.1 (Bio-Rad, Hercules, CA, USA). ## 3.3. Real-Time Quantitative PCR Gene-specific expression analyses were performed as already reported [2]. Fluorescence intensity of each amplified sample was measured with an ABI PRISM 7500 sequence detection system (Applied Biosystems, Foster City, CA, USA). All measurements were performed at least in triplicate and reported as the average value ± standard deviation of the mean (mean ± SD). Target gene values were normalized to B2M mRNA measurements, and expression data were calculated according to the 2-ΔΔCt method. Primers were designed using Primer 3 Plus, and their sequences are: VEGF-F: 5′-TCACAGGTACAGGGATGAGGACAC-3′; VEGF-R: 5′-CAAAGCACAGCAATGTCCTGAAG-3′. B2M-F: 5′-GCCTGCCGTGTGAACCAT-3′; B2M-R: 5′-CATCTTCAAACCTCCATGATGCT-3′; HIF1A-F: 5′-TCTG GGTTGAAACTCAAGCAACTG-3′; and HIF1A-R: 5′-CAACCGGTTTAAGGACACATTCTG-3′. SIRT6-F: 5′-CCTCCTCCGCTTCCTGGTC-3′; SIRT6-R: 5′-GTCTTACACTTGGCACATTCTTCC-3′. ## 3.4. mRNA-Seq Analysis Transcriptomic analyses were performed by BMR Genomics S.r.l. ( Via Redipuglia, 22, 35131 Padova, Italy) using 3 μg of total RNA obtained from 2 × 106 control HUVECs or HUVECs maintained for 8 h in anoxic conditions. For total RNA isolation, cells were treated with the RNeasy Plus mini kit (Qiagen, Hilde, Germany) according to the manufacturer’s instructions as previously reported [2]. RNA-seq transcriptome analysis provided the heatmaps showing the 50 most variable genes found for normoxic and anoxic cell samples; these heatmaps allowed us to graphically identify those genes that are up- or downregulated after 8 h of anoxia treatment. In addition to this selected gene pool (the 50 most variable genes) found in normoxic or 8 h anoxia-treated native HUVEC samples (or HUVEC-PBP and HUVEC-SIRT6 samples), a differential gene expression analysis enabled us to research other genes of specific interest by using DEG Table (fold-changes for gene targets were calculated with the following formula: 2log FC). ## 3.5. Determination of Cytokine and Interleukin Release Cell supernatants derived from HUVECs in normoxia and anoxia were collected at 8 h (cells used for RNA Seq), spun down at 1000× g for 10 min at 4 °C to remove any cell debris, and stored after addition of protease inhibitors (cComplete Mini; Roche, Basel, Switzerland) at −80 °C until analyses. A Bio-Plex Assay plate (Bio-Plex Pro Hu Screening Panel 27plx XPL) was purchased from Bio-Rad (Hercules, CA, USA) to detect human cytokines, interleukins, and other mediators and used according to the manufacturer’s recommendations. This kit permitted the detection of the following cytokines: FGF, G-CSF, IL-1β, IL-1α, IL-3, IL-16, IL-2, IL-5, IL-6, IL-8, LIF, IL-12, IL-15, MCP-1, SCF, SCGF-β, PDGF-BB, VEGF, MIF, M-CSF. Plates were read using a BioPlex® 200 instrument (Bio-Rad, Hercules, CA, USA), and interleukin concentrations (expressed as picograms per milliliter) were calculated by use of a standard curve and software provided by the manufacturer (Bio-Plex manager software, v.6.1) and normalized by protein content defined by the Bradford assay. ## 3.6. NADP(H) Evaluation HUVEC cells were plated at a density of 2 × 105 cells/well in 12-well plates, cultured for 8 h in normoxic or anoxic conditions, harvested, and lysed in 0.1 mL 0.6 M perchloric acid (PCA) (for NADP+) or 0.1 M NaOH (for NADPH) at 4 °C. Cell extracts were centrifuged for 3 min at 16,000× g, and the supernatants were collected and neutralized: samples in PCA were neutralized by diluting the extracts in 100 mM sodium phosphate buffer (pH 8), whereas samples in NaOH were warmed at 72 °C for 10 min and neutralized in 10 mM Tris-HCl, pH 6. The intracellular NADP(H) content was assessed with a sensitive enzyme cyclic assay, which exploits the use of glucose-6-phosphate dehydrogenase (G6PD): 0.1 mL of the cycling reagent [10 mM glucose-6-phosphate, 0.02 U/mL G6PD, 20 μM resazurin, 5 μg/mL diaphorase, 10 μM FMN, 10 mM nicotinamide and 100 mM sodium phosphate, pH 8.0] were added to each well containing 0.1 mL of the diluted samples. The increase in resorufin fluorescence (544 nm excitation, 590 nm emission) was measured every minute over a 4 h period using a fluorescence plate reader (Fluostar Optima, BMG Labtechnologies GmbH, Offenburg, Germany). A NADP(H) standard curve was always run in parallel in each assay. NADP(H) levels in each sample were normalized to the protein content as determined by the Bradford assay. ## 3.7. Assay of Glucose-6-Phosphate Dehydrogenase Activity HUVECs were lysed in ice-cold buffer [25 mM Tris-HCl (pH 7.4), 1 mM EDTA, and protease inhibitors] by brief sonication. The lysates were centrifuged at 10,000× g for 10 min at 4 °C and the supernatants (100 μg protein) were assayed for glucose-6-phosphate dehydrogenase (G6PD) activity by measuring the reduction of NADP+ in the reaction buffer [100 mM Tris-HCl, pH 7.4, 0.5 mM EDTA, 10 mM MgCl2, 0.2 mM NADP+, and 0.6 mM glucose-6-phosphate] at 25 °C. NADPH production was monitored at 355 nm excitation, 460 nm emission, using a fluorescence plate reader (see above). ## 3.8. ROS Assay The ROS-ID® Hypoxia/Oxidative stress detection kit (Enzo Life Sciences, Farmingdale, NY, USA) was used according to the manufacturer’s protocols to evaluate the total ROS production induced in anoxic cells. HUVECs were seeded at a density of 0.01 × 104 cells/well in a 96-well plate. Cells were incubated for 8 h in a Clariostar plus microplate reader (BMG Labtech, Ortenberg, Germany). The level of ROS was determined with the microscope Evos M5000 (Thermo Fisher Scientific, Waltham, MA, USA) via a GFP filter (Ex./Em. $\frac{470}{525}$), whereas anoxic cells were acquired via the Texas red filter (Ex./Em. $\frac{585}{628}$). ## 3.9. Glutathione Detection HUVECs (0.5 × 106 cells/flask) were washed twice with PBS and lysed with 100 µL of lysis buffer ($0.1\%$ Triton X-100, 0.1 M Na2HPO4, 5 mM EDTA, pH 7.5) followed by 15 µL of 0.1 N HCl and 140 µL of precipitating solution (100 mL containing 1.67 g (w/v) of glacial metaphosphoric acid, 0.2 g (w/v) of disodium EDTA and 30 g (w/v) of NaCl). After centrifugation at 12,000× g for 10 min at 4 °C, pellets were resuspended in 100 µL 0.1 N NaOH for protein quantification via the Bradford assay (Bio-Rad, Hercules, CA, USA). Supernatants were collected and $25\%$ (v/v) Na2HPO4 0.3 M and $10\%$ (v/v) DTNB were added for thiol determination by HPLC. A 5 µm, 150 × 4.6 mm BDS HypersilTM C18 column (Thermo Scientific, Waltham, MA, USA) was used for these studies. The mobile phase consisted of 10 mM KH2PO4 solution, pH 6.0 (buffer A), and buffer A containing $60\%$ (v/v) of acetonitrile (buffer B). All solutions were filtered through a 0.22 µm Millipore filter. The elution conditions were: 10 min $100\%$ buffer A, and gradually up to $100\%$ buffer B for 20 min. Buffer B was hold for 5 min before returning the gradient to $100\%$ buffer A within 3 min. The initial conditions were restored in 4 min. The flow rate was 1 mL/min and detection was carried out at 330 nm. Analyses were performed at room temperature and quantitative measurements were obtained by injection of known concentration standards and normalized protein content. ## 3.10. Statistical Analysis Statistics and graphical representations were performed using GraphPad PrismTM 9 (Boston, MA, USA). Student’s t-test and ANOVA test performed with Past Software version 3 were used for statistical analysis of the data; differences between groups were considered statistically significant when $p \leq 0.05.$ Quantitative analyses of the images were accomplished with CellProfiler Software. ## 4. Conclusions Depending on the oxygen content in the stem cell microenvironment, a number of adaptive physiological responses occur that regulate metabolism, redox homeostasis and vascular remodeling. Hypoxia has been linked to stem cell quiescence, whereas normoxia is thought to be required for stem cell activation. Given the importance of understanding stem cell biology, the effect of microenvironment and oxygen tension is of interest. However, the regulation of the stem cell microenvironment by normoxic/hypoxic conditions has yet to be defined, and there are few works that report observations on the effect of anoxia in vitro on bone marrow cultures. We found that marked anoxia exposure of endothelial cells triggers the transcription of a large number of genes involved in a variety of cellular processes such as glycolysis, angiogenesis, and cell proliferation, similar to those studied in hypoxic conditions, that are aimed at minimizing the deleterious effects caused by oxygen insufficiency at the cellular level. In recent years, the sirtuin family (SIRT1—SIRT7) has emerged as key regulators of many important biological processes, depending on their enzymatic activity, subcellular localization, and target specificity [77,78]. Among the seven sirtuins, SIRT6 is a chromatin-binding protein with diverse roles in genome stability, glucose metabolism, tumor suppression, and the organismal lifespan [71]. Previous studies have shown that SIRT6 deacetylates H3K9ac or H3K56ac and acts as a transcriptional co-repressor of the NF-kB-, C-JUN-, MYC-, and HIF-1α-mediated pathways in a tissue- and context-dependent manner. Many of these pathways have been implicated in the regulation of adult stem cell function and maintenance. However, there is no clear evidence indicating whether SIRT6 is functional relevant to HSC biology. Interestingly, the CXCL3, CXCL5, and IL-34 genes, whose mRNA levels are upregulated in native HUVECs after anoxia exposure, become downregulated by SIRT6 overexpression in HUVEC-SIRT6 cells exposed to anoxia for the same amount of time. Moreover, while LIF mRNA expression was not significantly affected by 8 h anoxia exposure in native HUVECs, it becomes upregulated in the presence of SIRT6. Therefore, the ability of Sirtuin 6 to modulate specific genes is evident. It is known that CXCL5 encodes a protein that is a member of the CXC subfamily of chemokines which recruit and activate leukocytes and which are classified by function (inflammatory or homeostatic) or by structure. This protein is proposed to bind the G-protein-coupled receptor chemokine (C-X-C motif) receptor 2 to recruit neutrophils, to promote angiogenesis, and to remodel connective tissues. It is widely identified in different cells and organs, such as endothelial cells and brain [79]. This protein is thought to play a role in cancer cell proliferation, migration, and invasion. A recent paper [80] reports CXCL5 (that binds with CXCR1 and CXCR2 and activates the p38 MAP kinase signaling pathways) to be upregulated in ischemic stroke. The authors constructed an ischemic/hypoxic model in human brain microvascular endothelial cells (BMECs) and investigated the function of CXCL5 and its potential value as a therapeutic target for ischemia-induced brain disease. The results demonstrated that CXCL5 silencing attenuated the ischemic/hypoxic-induced injury in human BMECs. Furthermore, the p38 inhibitor SB203580 significantly abolished the function of CXCL5 in model cells. In our anoxic cell model, the presence of overexpressed Sirtuin 6 is able to downregulate the expression of CXCL5 which, in contrast, is upregulated in native HUVECs exposed for 8 h to anoxia conditions. Some authors reported a chemotaxis assay showing that CXCL5 induces the migration of hematopoietic stem cells, suggesting that the differential regulation of the chemokine CXCL5 between the endosteal osteoblast (OB) niche and endothelial cells is involved in HSC mobilization from the OB niche or bone marrow to peripheral blood [81]. ST6GalNac2, known to have a role in cancer, [82] also appears to be upregulated in the presence of overexpressed SIRT6 in HUVEC-SIRT6 samples. Concerning LIF, it has already been reported that it plays an important role in a wide array of biological processes, including stem cell self-renewal. Hypoxia plays a critical role in LIF overexpression in solid tumors. The expression of LIF is also regulated by many cytokines. In cultured normal human bone marrow stromal cells, IL-1α, IL-1β, TGF-β, and TNF-α can all increase the transcription of LIF mRNA [83]. In our anoxic model, SIRT6 induces an upregulation of this factor, which could in turn lead to the modulation of interleukins and chemokines. We have shown here that SIRT6 is able to maintain the NADPH/NADP+ ratio in an anoxic environment. Sirtuins are indispensable for the maintenance of cellular redox homeostasis as they serve as key regulators of oxidative stress mechanisms. Sirtuins can both directly and indirectly regulate a wide variety of targets associated with primary antioxidant responses, including redox metabolic enzymes (such as G6PD), DNA repair enzymes, transcription factors, and co-factors [84]. This result is in line with the previously reported role for SIRT6 in preventing oxidative stress damage induced by H2O2 or by doxorubicin in breast cancer cells; this occurs through upregulation of NADPH levels as a consequence of G6PD activation [71]. In addition to this mechanism, in human mesenchymal cells, SIRT6 can prevent oxidative stress by transactivating NRF2-regulated antioxidant genes, including heme oxygenase 1 [85]. Finally, the increased levels of NADPH observed in SIRT6-overexpressing cells is in line with a recent report demonstrating that SIRT6 inhibits NADPH oxidase expression and activity in endothelial cells [86]. Of note, our results show that, in normoxic conditions, the stable transfection of endothelial HUVEC cells with the construct for the overexpression of Sirtuin 6 induces a pool of genes involved in the activation of molecular pathways modulated by Notch. The evolutionarily conserved “Notch signaling pathway” functions as a major mediator of cell fate determination during development and regulates several cellular functions including differentiation, proliferation, cell survival, and hematopoiesis process [87]. Interestingly, our results show that after 8 h of anoxia treatment of HUVEC-SIRT6 cells, there was a significant increase in VEGFA mRNA levels as well as activation of the molecular pathways categorized as “Cellular response to hypoxia” and “Signaling by VEGF”, indicating the induction of those molecular mechanisms that can support the expansion of the hematopoietic stem cell population. ## References 1. Richter R., Forssmann W., Henschler R.. Current Developments in Mobilization of Hematopoietic Stem and Progenitor Cells and Their Interaction with Niches in Bone Marrow. Transfus. Med. Hemother. (2017) 44 151-164. DOI: 10.1159/000477262 2. Antonelli A., Scarpa E.S., Magnani M.. Human Red Blood Cells Modulate Cytokine Expression in Monocytes/Macrophages Under Anoxic Conditions. Front. Physiol. (2021) 12 e632682. DOI: 10.3389/fphys.2021.632682 3. Chow D.C., Wenning L.A., Miller W.M., Papoutsakis E.T.. Modeling pO2 Distributions in the Bone Marrow Hematopoietic Compartment. I. Krogh’s Model. Biophys. J. (2001) 81 675-684. DOI: 10.1016/S0006-3495(01)75732-3 4. Namiki A., Brogi E., Kearney M., Kim E.A., Wu T., Couffinhal T., Varticovski L., Isner J.M.. Hypoxia Induces Vascular Endothelial Growth Factor in Cultured Human Endothelial Cells. J. Biol. Chem. (1995) 270 31189-31195. DOI: 10.1074/jbc.270.52.31189 5. Shibuya M.. Vascular endothelial growth factor (VEGF) and its receptor (VEGFR) signaling in angiogenesis: A crucial target for anti- and pro-angiogenic therapies. Genes Cancer (2011) 2 1097-1105. DOI: 10.1177/1947601911423031 6. Rehn M., Olsson A., Reckzeh K., Diffner E., Carmeliet P., Landberg G., Cammenga J.. Hypoxic induction of vascular endothelial growth factor regulates murine hematopoietic stem cell function in the low-oxygenic niche. Blood (2011) 118 1534-1543. DOI: 10.1182/blood-2011-01-332890 7. Geindreau M., Bruchard M., Vegran F.. Role of Cytokines and Chemokines in Angiogenesis in a Tumor Context. Cancers (2022) 14. DOI: 10.3390/cancers14102446 8. Loboda A., Jozkowicz A., Dulak J.. IF-1 versus HIF-2 is one more important than the other?. Vasc. Pharmacol. (2012) 20156 245-251. DOI: 10.1016/j.vph.2012.02.006 9. Zhong L., D’Urso A., Toiber D., Sebastian C., Henry R.E., Vadysirisack D.D., Guimaraes A., Marinelli B., Wikstrom J.D., Nir T.. The Histone Deacetylase Sirt6 Regulates Glucose Homeostasis via Hif1α. Cell (2010) 140 280-293. DOI: 10.1016/j.cell.2009.12.041 10. Yang Z., Huang Y., Zhu L., Yang K., Liang K., Tan J., Yu B.. SIRT6 promotes angiogenesis and hemorrhage of carotid plaque via regulating HIF-1α and reactive oxygen species. Cell Death Dis. (2021) 12 1-15. DOI: 10.1038/s41419-020-03372-2 11. Suda T., Takubo K., Semenza G.L.. Metabolic Regulation of Hematopoietic Stem Cells in the Hypoxic Niche. Cell Stem Cell (2011) 9 298-310. DOI: 10.1016/j.stem.2011.09.010 12. Schofield C.J., Ratcliffe P.J.. Oxygen sensing by HIF hydroxylases. Nat. Rev. Mol. Cell Biol. (2004) 5 343-354. DOI: 10.1038/nrm1366 13. Ginouves A., Ilc K., Macias N., Pouyssegur J., Berra E.. PHDs overactivation during chronic hypoxia “desensitizes” HIFalpha and protects cells from necrosis. Proc. Natl. Acad. Sci. USA (2008) 105 4745-4750. DOI: 10.1073/pnas.0705680105 14. Demidenko Z., Rapisarda A., Garayoa M.. Accumulation of hypoxia-inducible factor-1α is limited by transcription-dependent depletion. Oncogene (2005) 24 4829-4838. DOI: 10.1038/sj.onc.1208636 15. Bryder D., Jacobsen S.. Interleukin-3 supports expansion of long-term multilineage repopulating activity after multiple stem cell divisions in vitro. Blood (2000) 96 1748-1755. DOI: 10.1182/blood.V96.5.1748 16. Gazzola M.V., Collins N.H., Tafuri A., Keever C.. Recombinant interleukin 3 induces interleukin 2 receptor expression on early myeloid cells in normal human bone marrow. Exp. Hematol. (1992) 20 201-208. PMID: 1371965 17. Giampaolo S., Wójcik G., Serfling E., Patra A.K.. Interleukin-2-regulatory T cell axis critically regulates maintenance of hematopoietic stem cells. Oncotarget (2017) 8 29625-29642. DOI: 10.18632/oncotarget.16377 18. Dela Paz N.G., Walshe T.E., Leach L.L., Saint-Geniez M., D′Amore P.A.. Role of shear-stress-induced VEGF expression in endothelial cell survival. J. Cell Sci. (2012) 125 831-843. DOI: 10.1242/jcs.084301 19. Martin D., Galisteo R., Silvio J.. Gutkind. CXCL8/IL8 Stimulates Vascular Endothelial Growth Factor (VEGF) Expression and the Autocrine Activation of VEGFR2 in Endothelial Cells by Activating NF B through the CBM (Carma3/Bcl10/Malt1) Complex. J. Biol. Chem. (2009) 284 6038-6042. DOI: 10.1074/jbc.C800207200 20. Ségaliny A.I., Mohamadi A., Dizier B., Lokajczyk A., Brion R., Lanel R., Amiaud J., Charrier C., Boisson-Vidal C., Heymann D.. Interleukin-34 promotes tumor progression and metastatic process in osteosarcoma through induction of angiogenesis and macrophage recruitment. Int. J. Cancer (2015) 137 73-85. DOI: 10.1002/ijc.29376 21. Baghdadi M., Umeyama Y., Hama N., Kobayashi T., Han N., Wada H., Seino K.I.. Interleukin-34, a comprehensive review. J. Leukoc. Biol. (2018) 104 931-951. DOI: 10.1002/JLB.MR1117-457R 22. Eda H., Zhang J., Keith R.H., Michener M., Beidler D.R., Monahan J.B.. Macrophage-colony stimulating factor and interleukin-34 induce chemokines in human whole blood. Cytokine (2010) 52 215-220. DOI: 10.1016/j.cyto.2010.08.005 23. Lee E., Seo M., Juhnn Y.-S., Kim J.Y., Hong Y.J., Lee Y.J., Song Y.W.. Potential role and mechanism of IFN-gamma inducible protein-10 on receptor activator of nuclear factor kappa-B ligand (RANKL) expression in rheumatoid arthritis. Thromb. Haemost. (2011) 13 R104-R108. DOI: 10.1186/ar3385 24. Dufour J.H., Dziejman M., Liu M.T., Leung J.H., Lane T.E., Luster A.D.. IFN-γ-Inducible Protein 10 (IP-10; CXCL10)-Deficient Mice Reveal a Role for IP-10 in Effector T Cell Generation and Trafficking. J. Immunol. (2002) 168 3195-3204. DOI: 10.4049/jimmunol.168.7.3195 25. Xia J.B., Liu G.H., Chen Z.Y., Cheng-Zhou Mao C.Z., Deng-Cheng Zhou D.C., Wu H.Y., Park K.S., Zhao H., Kim S.K., Cai D.Q.. Hypoxia/ischemia promotes CXCL10 expression in cardiac microvascular endothelial cells by NFkB activation. Cytokine (2016) 81 63-70. DOI: 10.1016/j.cyto.2016.02.007 26. Deshmane S.L., Kremlev S., Amini S., Sawaya B.E.. Monocyte Chemoattractant Protein-1 (MCP-1): An Overview. J. Interf. Cytokine Res. (2009) 29 313-326. DOI: 10.1089/jir.2008.0027 27. Ou X., Wen T., Ying J., He Q., Xuan A., Ruan D.. MCP 1/CCR2 axis inhibits the chondrogenic differentiation of human nucleus pulposus mesenchymal stem cells. Mol. Med. Rep. (2022) 26 277. DOI: 10.3892/mmr.2022.12793 28. Park C.-Y., Choi S.C., Kim J.H., Choi J.H., Joo H.J., Hong S.J., Lim D.S.. Cardiac Stem Cell Secretome Protects Cardiomyocytes from Hypoxic Injury Partly via Monocyte Chemotactic Protein-1-Dependent Mechanism. Int. J. Mol. Sci. (2016) 17. DOI: 10.3390/ijms17060800 29. Didion S.P.. Cellular and Oxidative Mechanisms Associated with Interleukin-6 Signaling in the Vasculature. Int. J. Mol. Sci. (2017) 18. DOI: 10.3390/ijms18122563 30. Noetzli L.J., French S.L., Machlus K.. New Insights into the Differentiation of Megakaryocytes from Hematopoietic Progenitors. Arter. Thromb. Vasc. Biol. (2019) 39 1288-1300. DOI: 10.1161/ATVBAHA.119.312129 31. Chen Y.F., Wu Z.M., Xie C., Bai S., Zhao L.D.. Expression Level of IL-6 Secreted by Bone Marrow Stromal Cells in Mice with Aplastic Anemia. ISRN Hematol. (2013) 2013 986219. DOI: 10.1155/2013/986219 32. Blaser B.W., Moore J., Li B., Tamplin O.J., Binder V., Zon L.I.. IL-8 and CXCR1 Remodel the Vascular Niche to Promote Hematopoietic Stem and Progenitor Cell Engraftment. Blood (2015) 126 783. DOI: 10.1182/blood.V126.23.783.783 33. Karakurum M., Shreeniwas R., Chen J., Pinsky D., Yan S.D., Anderson M., Sunouchi K., Major J., Hamilton T., Kuwabara K.. Hypoxic induction of interleukin-8 gene expression in human endothelial cells. J. Clin. Investig. (1994) 93 1564-1570. DOI: 10.1172/JCI117135 34. Florczyk U., Czauderna S., Stachurska A., Tertil M., Nowak W., Kozakowska M., Poellinger L., Jozkowicz A., Loboda A., Dulak J.. Opposite effects of HIF-1α and HIF-2α on the regulation of IL-8 expression in endothelial cells. Free. Radic. Biol. Med. (2011) 51 1882-1892. DOI: 10.1016/j.freeradbiomed.2011.08.023 35. Zhang P., Zhang C., Li J., Han J., Liu X., Yang H.. The physical microenvironment of hematopoietic stem cells and its emerging roles in engineering applications. Stem Cell Res. Ther. (2019) 10 1-13. DOI: 10.1186/s13287-019-1422-7 36. D′Atri L.P., Pozner R.G., Nahmod K.A., Landoni V.I., Isturiz M., Negrotto S.. Paracrine regulation of megakaryo/thrombopoiesis by macrophages. Exp. Hematol. (2011) 39 763-772. DOI: 10.1016/j.exphem.2011.03.009 37. Dănilă M.-D., Piollet M., Aburel O.-M., Angoulvant D., Lefort C., Chadet S., Roger S., Muntean M.-D., Ivanes F.. Modulation of P2Y11-related purinergic signaling in inflammation and cardio-metabolic diseases. Eur. J. Pharmacol. (2020) 876 173060. DOI: 10.1016/j.ejphar.2020.173060 38. Dreisig K., Kornum B.R.. A critical look at the function of the P2Y11 receptor. Purinergic Signal. (2016) 12 427-437. DOI: 10.1007/s11302-016-9514-7 39. Piollet M., Sturza A., Chadet S., Gabillard-Lefort C., Benoist L., Muntean D.-M., Aburel O.-M., Angoulvant D., Ivanes F.. P2Y11 Agonism Prevents Hypoxia/Reoxygenation- and Angiotensin II-Induced Vascular Dysfunction and Intimal Hyperplasia Development. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22020855 40. Lefort C., Benoist L., Chadet S., Piollet M., Heraud A., Babuty D., Baron C., Ivanes F., Angoulvant D.. Stimulation of P2Y11 receptor modulates cardiac fibroblasts secretome toward immunomodulatory and protective roles after Hypoxia/Reoxygenation injury. J. Mol. Cell. Cardiol. (2018) 121 212-222. DOI: 10.1016/j.yjmcc.2018.07.245 41. Vultaggio-Poma V., Di Virgilio F.. P2 Receptors: Novel Disease Markers and Metabolic Checkpoints in Immune Cells. Biomolecules (2022) 12. DOI: 10.3390/biom12070983 42. Schömel N., Hancock S.E., Gruber L., Olzomer E.M., Byrne F.L., Shah D., Hoehn K.L., Turner N., Grösch S., Geisslinger G.. UGCG influences glutamine metabolism of breast cancer cells. Sci. Rep. (2019) 9 15665. DOI: 10.1038/s41598-019-52169-7 43. Wegner M.S., Gruber L., Mattjus P.. The UDP-glucose ceramide glycosyltransferase (UGCG) and the link to multidrug resistance protein 1 (MDR1). BMC Cancer (2018) 18. DOI: 10.1186/s12885-018-4084-4 44. Korbecki J., Kojder K., Kapczuk P., Kupnicka P., Gawrońska-Szklarz B., Gutowska I., Chlubek D., Baranowska-Bosiacka I.. The Effect of Hypoxia on the Expression of CXC Chemokines and CXC Chemokine Receptors—A Review of Literature. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22020843 45. Chen Y.-I., Wei P.-C., Hsu J.-L., Su F.-Y., Lee W.-H.. NPGPx (GPx7): A novel oxidative stress sensor/transmitter with multiple roles in redox homeostasis. Am. J. Transl. Res. (2016) 8 1626-1640. PMID: 27186289 46. Peppelenbosch M.P., Spaander M.C.W., Bruno M.J.. Glutathione peroxidase 7 prevents cancer in the oesophagus. Gut (2013) 63 537-538. DOI: 10.1136/gutjnl-2013-304906 47. Chen Z., Hu T., Zhu S., Mukaisho K., El-Rifai W., Peng D.-F.. Glutathione peroxidase 7 suppresses cancer cell growth and is hypermethylated in gastric cancer. Oncotarget (2017) 8 54345-54356. DOI: 10.18632/oncotarget.17527 48. Guerriero E., Capone F., Accardo M., Sorice A., Costantini M., Colonna G., Castello G.. GPX4 and GPX7 over-expression in human hepatocellular carcinoma tissues. Eur. J. Histochem. (2015) 59 2540. DOI: 10.4081/ejh.2015.2540 49. Wei P.-C., Hsieh Y.-H., Su M.-I., Jiang X., Hsu P.-H., Lo W.-T., Weng J.-Y., Jeng Y.-M., Wang J.-M., Chen P.-L.. Loss of the Oxidative Stress Sensor NPGPx Compromises GRP78 Chaperone Activity and Induces Systemic Disease. Mol. Cell (2012) 48 747-759. DOI: 10.1016/j.molcel.2012.10.007 50. Shao J., Liu X., Zhu L., Yen Y.. Targeting ribonucleotide reductase for cancer therapy. Expert Opin. Ther. Targets (2013) 17 1423-1437. DOI: 10.1517/14728222.2013.840293 51. Foskolou I.P., Jorgensen C., Leszczynska K.B., Olcina M.M., Tarhonskaya H., Haisma B., D’Angiolella V., Myers W.K., Domene C., Flashman E.. Ribonucleotide Reductase Requires Subunit Switching in Hypoxia to Maintain DNA Replication. Mol. Cell (2017) 66 206-220.E9. DOI: 10.1016/j.molcel.2017.03.005 52. Haffo L., Lu J., Bykov V.J.N., Martin S.S., Ren X., Coppo L., Wiman K.G., Holmgren A.. Inhibition of the glutaredoxin and thioredoxin systems and ribonucleotide reductase by mutant p53-targeting compound APR-246. Sci. Rep. (2018) 8 12671. DOI: 10.1038/s41598-018-31048-7 53. Sengupta R., Coppo L., Mishra P., Holmgren A.. Glutathione-glutaredoxin is an efficient electron donor system for mammalian p53R2–R1-dependent ribonucleotide reductase. J. Biol. Chem. (2019) 294 12708-12716. DOI: 10.1074/jbc.RA119.008752 54. Bruschi M., Vanzolini T., Sahu N., Balduini A., Magnani M., Fraternale A.. Functionalized 3D scaffolds for engineering the hematopoietic niche. Front. Bioeng. Biotechnol. (2022) 10 968086. DOI: 10.3389/fbioe.2022.968086 55. Motterlini R., Foresti R., Bassi R., Calabrese V., Clark J.E., Green C.J.. Endothelial Heme Oxygenase-1 Induction by Hypoxia. J. Biol. Chem. (2000) 275 13613-13620. DOI: 10.1074/jbc.275.18.13613 56. Fraternale A., De Angelis M., De Santis R., Amatore D., Masini S., Monittola F., Menotta M., Biancucci F., Bartoccini F., Retini M.. Targeting SARS-CoV-2 by synthetic dual-acting thiol compounds that inhibit Spike/ACE2 interaction and viral protein production. FASEB J. (2023) 37 e22741. DOI: 10.1096/fj.202201157RR 57. Hopman R.K., DiPersio J.F.. Advances in stem cell mobilization. Blood Rev. (2014) 28 31-40. DOI: 10.1016/j.blre.2014.01.001 58. Mendelson A., Frenette P.S.. Hematopoietic stem cell niche maintenance during homeostasis and regeneration. Nat. Med. (2014) 20 833-846. DOI: 10.1038/nm.3647 59. Nowicki M., Szemraj J., Wierzbowska A., Misiewicz M., Małachowski R., Pluta A., Grzybowska-Izydorczyk O., Robak T., Szmigielska-Kapłon A.. miRNA-15a, miRNA-16, miRNA-126, miRNA-146a, and miRNA-223 expressions in autologous hematopoietic stem cell transplantation and their impact on engraftment. Eur. J. Haematol. (2018) 100 426-435. DOI: 10.1111/ejh.13036 60. Nowicki M., Szemraj J., Wierzbowska A., Pluta A., Grzybowska-Izydorczyk O., Nowicka A., Stelmach P., Czemerska M., Szmigielska-Kapłon A.. Alterations in microRNA Expression during Hematopoietic Stem Cell Mobilization. Biology (2021) 10. DOI: 10.3390/biology10070668 61. Gattazzo F., Urciuolo A., Bonaldo P.. Extracellular matrix: A dynamic microenvironment for stem cell niche. Biochim. Et Biophys. Acta (BBA)–Gen. Subj. (2014) 1840 2506-2519. DOI: 10.1016/j.bbagen.2014.01.010 62. Skokowa J., Lan D., Thakur B., Gupta K., Cario G., Brechlin A., Stanulla M., Welte K.. NAMPT is essential for the G-CSF-induced myeloid differentiation via a NAD+-sirtuin-1-dependent pathway. Nat. Med. (2009) 221 A64. DOI: 10.1055/s-0029-1222684 63. O′Callaghan C., Vassilopoulos A.. Sirtuins at the crossroads of stemness, aging, and cancer. Aging Cell (2017) 16 1208-1218. DOI: 10.1111/acel.12685 64. Liszt G., Ford E., Kurtev M., Guarente L.. Mouse Sir2 Homolog SIRT6 Is a Nuclear ADP-ribosyltransferase. J. Biol. Chem. (2005) 280 21313-21320. DOI: 10.1074/jbc.M413296200 65. Bruzzone S., Fruscione F., Morando S., Ferrando T., Poggi A., Garuti A., D’Urso A., Selmo M., Benvenuto F., Cea M.. Catastrophic NAD+ depletion in activated T lymphocytes through Nampt inhibition reduces demyelination and disability in EAE. PLoS ONE (2009) 4. DOI: 10.1371/journal.pone.0007897 66. Kugel S., Mostoslavsky R.. Chromatin and beyond: The multitasking roles for SIRT6. Trends Biochem. Sci. (2014) 39 72-81. DOI: 10.1016/j.tibs.2013.12.002 67. Mao Z., Hine C., Tian X., Van Meter M., Au M., Vaidya A., Seluanov A., Gorbunova V.. SIRT6 Promotes DNA Repair Under Stress by Activating PARP1. Science (2011) 332 1443-1446. DOI: 10.1126/science.1202723 68. Nowicki M., Wierzbowska A., Stec-Martyna E., Kulczycka-Wojdala D., Nowicki G., Szmigielska-Kapłon A.. SIRT1-SIRT7 Expression in Patients with Lymphoproliferative Disorders Undergoing Hematopoietic Stem Cell Mobilization. Cancers (2022) 14. DOI: 10.3390/cancers14051213 69. Edatt L., Aswini P., Grace R.R., Vishnu R., Sharath S.S., Sameer K.V.B.. Role of Sirtuins in Tumor Angiogenesis. Front. Oncol. (2020) 9 1516. DOI: 10.3389/fonc.2019.01516 70. Fiorentino F., Mai A., Rotili D.. Emerging Therapeutic Potential of SIRT6 Modulators. J. Med. Chem. (2021) 64 9732-9758. DOI: 10.1021/acs.jmedchem.1c00601 71. Sociali G., Grozio A., Caffa I., Schuster S., Becherini P., Damonte P., Sturla L., Fresia C., Passalacqua M., Mazzola F.. SIRT6 deacetylase activity regulates NAMPT activity and NAD(P)(H) pools in cancer cells. FASEB J. (2018) 33 3704-3717. DOI: 10.1096/fj.201800321R 72. Kobayashi C.I., Suda T.. Regulation of reactive oxygen species in stem cells and cancer stem cells. J. Cell. Physiol. (2011) 227 421-430. DOI: 10.1002/jcp.22764 73. Morris G., Gevezova M., Sarafian V., Maes M.. Redox regulation of the immune response. Cell. Mol. Immunol. (2022) 19 1079-1101. DOI: 10.1038/s41423-022-00902-0 74. Ali M.H., Schlidt S.A., Chandel N.S., Hynes K.L., Schumacker P.T., Gewertz B.L.. Endothelial permeability and IL-6 production during hypoxia: Role of ROS in signal transduction. Am. J. Physiol. (1999) 277 L1057-L1065. DOI: 10.1152/ajplung.1999.277.5.L1057 75. Lu X., Kang Y.. Hypoxia and Hypoxia-Inducible Factors: Master Regulators of Metastasis. Clin. Cancer Res. (2010) 16 5928-5935. DOI: 10.1158/1078-0432.CCR-10-1360 76. Wu L., Yu H., Zhao Y., Zhang C., Wang J., Yue X., Yang Q., Hu W.. HIF-2α mediates hypoxia-induced LIF expression in human colorectal cancer cells. Oncotarget (2015) 6 4406-4417. DOI: 10.18632/oncotarget.3017 77. Houtkooper R.H., Pirinen E., Auwerx J.. Sirtuins as regulators of metabolism and healthspan. Nat. Rev. Mol. Cell Biol. (2012) 13 225-238. DOI: 10.1038/nrm3293 78. Klein M.A., Denu J.M.. Biological and catalytic functions of sirtuin 6 as targets for small-molecule modulators. J. Biol. Chem. (2020) 295 11021-11041. DOI: 10.1074/jbc.REV120.011438 79. Krishna Mohan S., Krishna Mohan P., Krishna R.. Solution structure of CXCL5–a novel chemokine and adipokine implicated in inflammation and obesity. PLoS ONE (2014) 9. PMID: 24695525 80. Min Y., Xiaokun M., Dudu J., Lijing W., Qing Z., Jiangmin Z.. CXC chemokine ligand 5 (CXCL5) disrupted the permeability of human brain microvascular endothelial cells via regulating p38 signal. Microbiol. Immunol. (2021) 65 40-47. PMID: 33026667 81. Yoon K.-A., Cho H.-S., Shin H.-I., Cho J.-Y.. Differential Regulation of CXCL5 by FGF2 in Osteoblastic and Endothelial Niche Cells Supports Hematopoietic Stem Cell Migration. Stem Cells Dev. (2012) 21 3391-3402. DOI: 10.1089/scd.2012.0128 82. Murugaesu N., Iravani M., Van Weverwijk A., Ivetic A., Johnson D.A., Antonopoulos A.. An In Vivo Functional Screen Identifies ST6GalNAc2 Sialyltransferase as a Breast Cancer Metastasis Suppressor. Cancer Discov. (2014) 4 304-317. DOI: 10.1158/2159-8290.CD-13-0287 83. Wetzler M., Talpaz M., Lowe D.G., Baiocchi G., Gutterman J.U., Kurzrock R.. Constitutive expression of leukemia inhibitory factor RNA by human bone marrow stromal cells and modulation by IL-1, TNF-alpha, and TGF-beta. Exp. Hematol. (1991) 19 347-351. PMID: 1709108 84. Shahgaldi S., Kahmini F.R.. A comprehensive review of Sirtuins: With a major focus on redox homeostasis and metabolism. Life Sci. (2021) 282 119803. DOI: 10.1016/j.lfs.2021.119803 85. Pan H., Guang-Hui L., Liu X., Li J., Wang L., Wu J., Weizhou Z., Zhang W., Ren R., Zhang W.. SIRT6 safeguards human mesenchymal stem cells from oxidative stress by coactivating NRF2. Cell Res. (2016) 26 190-205. DOI: 10.1038/cr.2016.4 86. Greiten L.E., Zhang B., Roos C.M., Hagler M., Jahns F.-P., Miller J.D.. Sirtuin 6 Protects Against Oxidative Stress and Vascular Dysfunction in Mice. Front. Physiol. (2021) 12 753501. DOI: 10.3389/fphys.2021.753501 87. Butler J.M., Nolan D.J., Vertes E.L., Varnum-Finney B., Kobayashi H., Hooper A.T., Seandel M., Shido K., White I.A., Kobayashi M.. Endothelial Cells Are Essential for the Self-Renewal and Repopulation of Notch-Dependent Hematopoietic Stem Cells. Cell Stem Cell (2010) 6 251-264. DOI: 10.1016/j.stem.2010.02.001
--- title: Parasite and Pesticide Impacts on the Bumblebee (Bombus terrestris) Haemolymph Proteome authors: - Dalel Askri - Edward A. Straw - Karim Arafah - Sébastien N. Voisin - Michel Bocquet - Mark J. F. Brown - Philippe Bulet journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049270 doi: 10.3390/ijms24065384 license: CC BY 4.0 --- # Parasite and Pesticide Impacts on the Bumblebee (Bombus terrestris) Haemolymph Proteome ## Abstract Pesticides pose a potential threat to bee health, especially in combination with other stressors, such as parasites. However, pesticide risk assessment tests pesticides in isolation from other stresses, i.e., on otherwise healthy bees. Through molecular analysis, the specific impacts of a pesticide or its interaction with another stressor can be elucidated. Molecular mass profiling by MALDI BeeTyping® was used on bee haemolymph to explore the signature of pesticidal and parasitic stressor impacts. This approach was complemented by bottom-up proteomics to investigate the modulation of the haemoproteome. We tested acute oral doses of three pesticides—glyphosate, Amistar and sulfoxaflor—on the bumblebee Bombus terrestris, alongside the gut parasite Crithidia bombi. We found no impact of any pesticide on parasite intensity and no impact of sulfoxaflor or glyphosate on survival or weight change. Amistar caused weight loss and 19–$41\%$ mortality. Haemoproteome analysis showed various protein dysregulations. The major pathways dysregulated were those involved in insect defences and immune responses, with Amistar having the strongest impact on these dysregulated pathways. Our results show that even when no response can be seen at a whole organism level, MALDI BeeTyping® can detect effects. Mass spectrometry analysis of bee haemolymph provides a pertinent tool to evaluate stressor impacts on bee health, even at the level of individuals. ## 1. Introduction Pollinators are vital components of natural and managed ecosystems, contributing USD 235–577 billion a year to the global economy through the ecosystem service of pollination [1]. The majority of this economically important pollination is carried out by bees [1], but numerous studies indicate that bees are threatened and in decline [2,3], posing a threat to pollination services. A range of anthropogenic pressures are believed to threaten bee health, from land-use change [4] to pesticides [5,6,7], acting both individually and in combination [8]. Two stressors in particular—pesticides and parasites—have received significant research attention due to the large impacts they may have on bee health [5,7,9,10,11,12,13,14,15,16]. The impact of these stressors can be measured at a range of levels, from populations to molecules [15,17,18,19]. Studies at the molecular level can provide an understanding of the mechanistic interaction between stressors and bee health [15,20]. As such, they underpin and explain studies of lethal and sublethal effects in individuals, colonies, and populations [10,11,13,14,15,21,22]. In addition, changes in gene expression or protein production in response to stressors may provide potential biomarkers that can be used in health monitoring [11,12,17,23,24,25,26,27]. Previous molecular-level studies in bees exposed to pesticides and parasites have largely focused on gene expression. For example, Haas et al. [ 2022] analysed genomic data for 75 bee species and demonstrated by the recombinant expression of 26 CYP9Q3 putative functional orthologs that detoxification is an evolutionarily conserved mechanism across bee families [28]. They observed a conserved capacity to metabolise certain insecticides across all major bee families while identifying a limited number of bee species where this function may have been lost. Similarly, Al-Naggar and Baer [2019] studied the effects of short-term exposure to a sublethal dose of the flupyradifurone-based insecticide Sivanto early in life on survival and immunity in A. mellifera [12]. They selected five genes involved in detoxification, CYP305D1, CYP6AS14, CYP9Q3, GSTD, and SODH2 as representatives of antioxidant-enzyme families that are known to target pesticides and secondary metabolites [29,30], along with genes involved in the insect immune response. The defensin1 gene was down-regulated compared to controls, while the apismin, Lys-1, and chitinase genes were significantly up-regulated in pesticide-exposed bees compared to control bees. Glyphosate exposure can alter immune response pathways by down-regulating the gene expression coding for host-produced AMPs (abaecin, apidaecin, defensin, and hymenoptaecin) in A. mellifera [27]. Moreover, A. cerana cerana and A. mellifera ligustica are affected by glyphosate commercial formulation [31]. However, this exposure seems to increase rather than decrease the expression of many genes involved in immunity, agrochemical detoxification and resistance, such as antimicrobial peptides, cuticle proteins, and cytochrome P450 families. In other studies, immune/detoxifying gene expression was variable (up and down) [23,25]. No significant pesticide–parasite interactions were found for any of the genes investigated. Proteomic changes in bees exposed to parasites and/or pesticides are also an area of active research. By analysing haemolymph proteome, several physiological functions of honey bees, such as energy metabolism, detoxification, metamorphosis, and chemosensing, have been shown to be disrupted by Varroa [32]. Up-regulation of proteins involved in stress response, carbohydrate metabolism and energy synthesis, and protein folding/binding was observed in the head proteome of nurse honey bees [33]. Houdelet et al. [ 2021] found changes to the gut proteome following exposure of A. mellifera to Nosema spp. [ 22]. The team observed both up- and down-regulation of various proteins mainly involved in metabolism and response to stimuli. Previously, fipronil was also observed to induce important neuroproteomic changes in the brains of honey bees [34]. However, the majority of these studies have used the managed honey bee A. mellifera [35,36,37], which is not representative of the more than 20,000 species of bee [38], and have focused on insecticides rather than other types of pesticide [35,36,37]. Here we investigate proteomic responses of the bumble bee Bombus terrestris, an abundant and important wild and managed pollinator of crops and wildflowers [1,39,40], to individual and combined exposure to a range of agrochemicals and the highly prevalent parasite Crithidia bombi [41,42,43,44]. To address the broad range of agrochemicals, we use (i) sulfoxaflor, a sulfoximine insecticide that has been shown to detrimentally affect bumble bee health [45,46,47] and that was banned in the EU in 2022 for outside use (EC, 2022); and (ii) Amistar, a broad-spectrum fungicide product containing azoxystrobin, which saw high uptake in the early 2000s and is still used widely today. Amistar is the flagship formulation for the active ingredient azoxystrobin, although it has now moved out of patent, and 71 other azoxystrobin products are available in the UK alone (Straw and Stanley, In Review). An emulsifier/surfactant co-formulant in Amistar, alcohol ethoxylates, has been found to cause damage to bumblebee gut tissue, leading to food aversion, weight loss, and ultimately death [48]. Finally, (iii) glyphosate, a herbicide that is the world’s most used pesticide [49,50]. The impacts of glyphosate on bees are hotly contested [51,52], with the most rigorous evidence pointing to potential sublethal impacts on the microbiome [53], with conflicting evidence as to lethal effects [54,55,56]. To represent the most likely co-exposure situation in the field, we used the trypanosome gut parasite C. bombi, which is the most prevalent parasite of bumble bees across Europe [41,42,43,44]. C. bombi impacts physiology [57,58], learning [59,60,61], and colony fitness [58,62]. Using modified OECD risk assessment protocols and fully crossed experiments, combined with MALDI BeeTyping® and bottom-up proteomics by LC-ESI-MS/MS [22,63,64], we ask (i) how exposure to an insecticide, fungicide, and herbicide, individually or in combination with the parasite, impacts the haemolymph proteome profile; (ii) which proteins respond to these stressors; and (iii) how these responses map onto higher level effects of exposure on individual longevity, weight change, and parasite load. ## 2.1.1. Survival In the Amistar experiment, Amistar caused significant mortality ($41.4\%$), as did the Amistar + C. bombi treatment ($18.8\%$). No bees in the negative control or C. bombi-only treatments died, while all bees in the positive control died, confirming the test’s ability to detect lethal effects. The global chi-square test found a significant impact of treatment (X2 (2, $$n = 92$$) = 16.32, $p \leq 0.001$). Individually, the mortality impacts of Amistar and Amistar + C. bombi were significantly higher relative to the control (X2 (1, $$n = 49$$) = 13.55, $p \leq 0.001$) and (X2 (1, $$n = 63$$) = 4.43, $$p \leq 0.035$$). In the glyphosate and sulfoxaflor experiments, glyphosate did not cause any mortality to the bees, while sulfoxaflor exposure caused limited, but non-significant, mortality ($5.6\%$), similar to the combined sulfoxaflor + C. bombi exposure ($9.1\%$). No bees in the negative control, C. bombi only, glyphosate only or glyphosate + C. bombi treatments died, while all bees in the positive control (dimethoate) died, confirming the test’s ability to detect lethal effects. Due to low mortality, Fisher’s Exact tests were exclusively used for a treatment versus control comparison. There was no significant effect of either sulfoxaflor alone or the sulfoxaflor + C. bombi treatment on mortality relative to the control (Fisher’s exact test (two-sided) $$p \leq 0.190$$ and $$p \leq 0.053$$, respectively). ## 2.1.2. Weight Change In the Amistar experiment, bees in the positive control gained the most weight, while all other treatments gained less weight, or even lost weight. Bees in the positive control on average gained 20.9 mg, while bees in the C. bombi treatment gained 11.9 mg on average. Bees exposed to Amistar alone gained less weight, at just 6.2 mg, while Amistar + C. bombi-exposed bees lost an average of 4.8 mg. The weight change in the C. bombi alone treatment was not significantly different to the control (PE = −0.01, CI = −0.02 to 0.0). In contrast, the weight changes in the Amistar-only and Amistar + C. bombi treatments were significantly different relative to the control (Amistar only: PE = −0.02, CI = −0.03 to −0.0; Amistar + C. bombi: PE = −0.03, CI = −0.04 to −0.01). In the glyphosate and sulfoxaflor experiments, bees in the positive control lost some weight, while all other treatments made limited weight gains. Bees in the positive control on average lost −2.5 mg, while bees in the C. bombi treatment gained 8.6 mg on average. Glyphosate-only bees gained 1.9 mg, while glyphosate + C. bombi-exposed bees gained 0.5 mg. Sulfoxaflor-only bees gained the most weight at 11.3 mg, while sulfoxaflor + C. bombi-exposed bees gained 6.5 mg. However, relative to the control, no weight loss or gain was statistically significant (C. bombi: PE = 0.01, CI = −0.00 to 0.03; glyphosate only: PE = 0.00, CI = −0.01 to 0.02; glyphosate + C. bombi: PE = 0.0, CI = 0.01 to 0.02; sulfoxaflor only: PE = 0.01, CI = −0.00 to 0.03; sulfoxaflor + C. bombi: PE = 0.01, CI = −0.00 to 0.02). For the whole organism metrics, impacts varied by substance. Amistar caused significant mortality and weight loss (or lack of weight gain). This mirrors the effects found in Straw and Brown [2021] [48], which found that a co-formulant, alcohol ethoxylates, was responsible for these effects, while the active ingredient (Azoxystrobin) did not contribute to the mortality effects. The weight loss in the Amistar + C. bombi treatment, and significant lack of weight gain in the Amistar-only treatment was likely caused by melanisation of the gut tissue, reducing appetite and a bee’s ability to intake energy. Ultimately, this likely explains the mortality seen in these treatments. The reduced mortality in the Amistar + C. bombi treatment is likely stochastic, as there is little reason C. bombi would ameliorate the impacts of the pesticide. It is worth noting that this work pre-dates the experiments in Straw and Brown [2021] [52], so the mortality was unexpected, hence the sample size for the haemoproteome analysis is reduced as only living bees had haemolymph extracted. Neither glyphosate nor sulfoxaflor, nor their combination with C. bombi, caused any significant impacts on survival or change in weight. These findings confirm prior findings that acute exposure to glyphosate has little to no measured impact on these metrics [52]. Sulfoxaflor caused a non-significant amount of mortality, although 5–$9\%$ indicates that the 0.06 µg dose used was potentially beyond our intention of a fully non-lethal dose. That no impacts were noted with this high exposure gives confidence that sulfoxaflor does not impact these traits. No pesticides caused an impact on parasite intensity, indicating that they do not meaningfully interact over this timescale or with this exposure. While this experimental design is more parametrised to detect pesticidal effects, the lack of change in parasite intensity suggests that even with an experiment tailored to detect parasite-driven effects, none would be seen. For all whole organism metrics, there was no impact of C. bombi, even alongside pesticide exposure. This reaffirms prior results showing that in OECD 247 style acute toxicity tests, C. bombi does not contribute to mortality [52]. Additionally, it confirms previous findings, using different methods, that C. bombi does not meaningfully interact with pesticidal stressors [47,52]. ## 2.2. Molecular Mass Fingerprints (MFPs) Exposure to Amistar, either alone or in combination with the parasite C. bombi, impacted the haemoproteome when compared to control bees or bees exclusively infected with C. bombi, (Figure S1). No discrimination between C. bombi parasitised and control bees, nor Amistar treated versus Amistar infected with C. bombi, was observed. By comparing the PCAs of the control, Amistar, C. bombi, and Amistar + C. bombi experiments, there was a clear separation between the two groups exposed to Amistar versus the control and C. bombi (Figure 1A). A similar separation was observed in the PCAs of control versus glyphosate, control versus glyphosate + C. bombi and glyphosate versus glyphosate + C. bombi (Figure S2). In the case of sulfoxaflor, based on the PCAs, bees treated with sulfoxaflor alone were discriminated from the control bees and from bees infected with C. bombi alone. In our experimental conditions, an infection with C. bombi did not lead to discrimination between samples (Figures S2 and S3). This is concordant with what we observed in the pairwise analysis. For glyphosate, there was a tentative differentiation between glyphosate and glyphosate + C. bombi versus C. bombi and the control (Figure 1B). For sulfoxaflor, there was no discrimination as all groups overlapped (Figure 1C). ## 2.3. Modulated Molecular Ions (MMIs) Following Amistar, Sulfoxaflor, and Glyphosate Exposure and Co-Infection with C. bombi Supporting the PCA data, a high number of significantly modulated molecular ions (MMIs) were detected in the Amistar experiment (Amistar-exposed bees versus (i) control ($76.92\%$ MMIs), (ii) C. bombi alone ($76.24\%$), (iii) C. bombi with Amistar ($79.90\%$), and (iv) when we compared C. bombi to Amistar + C. bombi-treated bees ($81.55\%$)). Lower numbers of significant MMIs were observed in the glyphosate and sulfoxaflor experiments, as shown in Figure 2. Across all three experiments, we did not observe any significant MMIs following C. bombi infection alone. The details of total, stable and modulated ions for all pairwise comparisons are available in Table S1. ## 2.4. Variation in Three Bee Immune Peptides—Apidaecin, Abaecin, and Chymotrypsin Inhibitor—Following Pesticide Exposure To understand the differences in molecular ion levels, we analysed the generated peak lists and focused on peptides that are recognised as indicators of an activated bee immune response (namely apidaecin, abaecin, and chymotrypsin inhibitor) with average molecular-related ions identified by MALDI BeeTyping® as m/z 1978.6, 4396.5 and 5937.8, respectively. Apidaecin and abaecin peak values responded similarly across treatments, but chymotrypsin inhibitor responded differently (Table 1). The details of the percentage calculation are available in Table S2. Chymotrypsin inhibitor was reported to be impacted by bee stressors such as the Nosema parasite in A. mellifera [22], and could be a bee health response marker in B. terrestris. Under glyphosate and sulfoxaflor exposure, the average molecular-related ions of apidaecin and chymotrypsin inhibitor did not change ($p \leq 0.05$) (Table S2). Furthermore, no significant variation was noted for abaecin in any of the treatments. However, apidaecin varied significantly following Amistar exposure. Statistical analysis showed that only Amistar exposure led to significant changes in apidaecin (PWKW control versus Amistar < 0.000001, Amistar versus C. bombi 0.0000013, control versus Amistar + C. bombi, and C. bombi versus Amistar + C. bombi < 0.000001) and chymotrypsin inhibitor (PWKW control versus Amistar 0.000387, Amistar versus C. bombi 0.000407, control versus Amistar + C. bombi, and C. bombi versus Amistar + C. bombi < 0.000001). ## 2.5. Protein Quantity Variations Following Pesticide and Parasite Exposure Demonstrated by Differential Bottom-Up Proteomics Using LFQ, we were able to quantify a total of 621 proteins, including 369 unique proteins, across the experiments (Table S3). Among them, 65 unique proteins were differentially expressed (DEPs), reflecting an impact on the proteomes by a given experimental treatment (Table S4). The results of this section are reported by experiment, i.e., all the different treatment groups related to a pesticide. Interestingly, the highest percentage of DEPs was observed after Amistar exposure ($35.69\%$), followed by glyphosate ($13.81\%$) and sulfoxaflor ($5.95\%$) (Figure 3). The proteins that were dysregulated following C. bombi exposure compared to any of the other conditions were not parasite-specific, as they were also seen in the remaining comparisons. This is in concordance with the lack of an effect of the parasite on the whole-body metrics. For further analysis and interpretation, we focused on the DEPs and analysed their variation per pesticide, i.e., Amistar, glyphosate, and sulfoxaflor. A Venn diagram (Figure S4) was generated to identify proteins detected only in a specific treatment or proteins that were DEPs across the different exposures. Of the 65 dysregulated proteins, 46 unique proteins were found after Amistar exposure, 13 after glyphosate, and 4 after sulfoxaflor. Two proteins were differentially expressed under two pesticide treatments: peptidoglycan recognition protein SA, ATL64812.1 after Amistar or glyphosate exposure, and uncharacterised protein LOC107189219 (XP_015433190.1) after glyphosate or sulfoxaflor exposure. Functional annotation using Gene Ontology was performed for the three experiments (Figure 4). It showed that the most affected processes (Figure 4A) after Amistar exposure were carbohydrate metabolic process, lipid transport, and proteolysis. For molecular functions (Figure 4B), the most impacted were lipid transporter activity, chitin binding, protein binding, serine-type endopeptidase inhibitor, ATP binding, and zinc ion binding. For glyphosate, various molecular functions were identified for the DEPs, namely transferase activity, hydrolase activity, molecular function regulator activity, oxidoreductase activity, antioxidant activity, and catalytic activity. The biological processes found were cellular-modified amino acid metabolic process, cell differentiation, anatomical structure development, defence response to other organisms, reproductive process, carbohydrate derivative metabolic process, cell adhesion, establishment or maintenance of cell polarity, immune system process, and metal ion homeostasis. The lists of the identified proteins in each biological process are available in Table S5. Interestingly, from the processes listed above, the protein ATL64812.1 was found to be differentially expressed after both Amistar and glyphosate exposure. This peptidoglycan recognition protein is known to play an important role in the response of insects to bacteria, and according to our OmicsBox interrogation was found to be involved in defence responses to other organisms and immune system processes. This protein was up-regulated after glyphosate exposure (glyphosate + C. bombi versus C. bombi and glyphosate versus C. bombi) and down-regulated after exposure to C. bombi alone. For sulfoxaflor, only one biological process, lipid metabolic process, and two molecular functions, hydrolase activity and regulator activity, were identified. In this section, we focused on investigating the variation in proteins that could be markers of pesticide and/or pathogen exposure. Specifically, we examined proteins that were shown to play roles, or to be key in immune response, response to stimulus/stress, and response to oxidative stress. Indeed, we observed that following Amistar exposure, nearly all proteins involved in the processes mentioned above were up-regulated. When Amistar was compared to the control, 14 up- versus 3 down-regulated proteins were found. For Amistar + C. bombi versus the control, 19 up- versus 14 down-regulated proteins were highlighted, while for the pairwise samples Amistar + C. bombi versus C. bombi, 16 up- versus 6 down-regulated proteins were identified. Here we suggest that Amistar activates the processes of bee immunity in contrast to C. bombi. This was further supported when we compared C. bombi to the other conditions, as this time more proteins were down-regulated: C. bombi versus control, 6 down and 2 up; and C. bombi versus Amistar, 10 down and only one up. Furthermore, some proteins of interest have been shown to be involved in response to stimuli and defence mechanisms [65,66,67,68,69,70]. As examples, chitinase-like protein (XP_016769017.1 and XP_012237228.1), interferon-related developmental regulator 1-like (XP_017879492.1), heat-shock 70 kDa protein cognate 4 (KMQ87979.1), transferrin-like (XP_035740737.1), apolipophorins, and the two proteins ferritin (ABV68875.1) and vitellogenin (AUX13057.1) that were up-regulated only when Amistar was compared to another condition. We also observed that some of these proteins were up-regulated when associated with Amistar and down-regulated when associated with C. bombi. In addition, sugar metabolism appeared to be stimulated after bees were exposed to Amistar. Specifically, glucose dehydrogenase (XP_020718843.1) and pyruvate kinase (KYQ58406.1) were up-regulated 54.30 and 100 times in the Amistar + C. bombi versus C. bombi and Amistar + C. bombi versus control comparisons, respectively. However, glucose dehydrogenase was down-regulated when we compared C. bombi to Amistar (ratio 0.04). A total of 25 DEPs were identified following glyphosate exposure, with 10 being up- and 15 down-regulated. When C. bombi was present (alone or in combination with the treatment), almost all DEPs were down-regulated (Table S4). For example, if C. bombi was present, the proteins peptidoglycan recognition protein SA (ATL64812.1, ratio 0.08), the uncharacterised protein LOC107189219 (XP_015433190.1), arginine kinase isoform X (XP_039309898.1, ratio 1 0.03), titin-like (LOC100881637), transcript variant X4 (CAB0031481.1, ratio 0.07), and the peroxidase-like isoform X1 (XP_012141527.1, ratio 0.42) were down-regulated. However, arginine kinase isoform X1 (XP_032455210.1), peptidoglycan recognition protein SA (ATL64812.1), and the uncharacterised protein LOC107189219 (XP_015433190.1) were up-regulated when glyphosate was present (alone or combined with C. bombi). For sulfoxaflor, more than $50\%$ of the DEPs were up-regulated. Following sulfoxaflor exposure and compared to C. bombi, up-regulation of proteins involved in defence systems, namely chymotrypsin inhibitor-like (XP_003708656.1, ratio 80.98) and heat-shock protein beta-1 (KYQ52813.1, ratio 95.82), was seen, in addition to up-regulation of an uncharacterised protein LOC107189219 (XP_015433190.1, ratio 50.60). Similar proteins were observed to be up-regulated when we compared sulfoxaflor + C. bombi versus C. bombi. These proteins were observed to be down-regulated when C. bombi was present compared to the control. This seems to be a common response of the bees to the pesticides compared to C. bombi, as discussed above. However, no DEPs were detected when sulfoxaflor was compared to the control, even when combined with C. bombi ($p \leq 0.05$). Furthermore, we examined the most impacted molecular pathways following pesticide exposure. All pathways and proteins are available in Table S6. After Amistar exposure, 133 impacted pathways had at least one DEP involved. In contrast, there were 31 after glyphosate and 22 after sulfoxaflor exposure. We also analysed the overlap between them (Figure 5). The list of these pathways (common and specific) is available in Table S7. The top 15 most impacted pathways by exposure to Amistar, glyphosate, and sulfoxaflor are illustrated in Table 2. Interestingly, the pathway “Neutrophil degranulation_ R-DME-6798695” (Figure S5), which belongs to the innate immune system, was common to Amistar and glyphosate responses. It is involved in immune responses to bacterial infection [71,72,73]. In our study, the impact on protein abundance (Figure S5, Table S4) depended on the substance. Indeed, we found abundance changed either consistently up or down, or varied, depending on the treatment (Figure S5). The down-regulated proteins when the bees were exposed to glyphosate treatment were transferrin (XP_003486912.1, ratio 0.01 and $p \leq 0.01$), peroxidase-like isoform X1 (XP_012141527.1, ratio 0.417 and $p \leq 0.05$), and antichymotrypsin-2-like isoform X4 (XP_033189693.1, ratio 0.018 and $p \leq 0.05$). However, the protein transferrin (XP_003486912.1) was up-regulated when the bees were exposed to C. bombi compared to the control (ratio 100 and $p \leq 0.01$). After bee exposure to Amistar, the dysregulated proteins involved in neutrophil degranulation were up- and down-regulated depending on the treatment. Indeed, when Amistar was present (alone or combined with C. bombi) compared to other conditions (C. bombi or control), the highest number of proteins were up-regulated. Among them, the heat-shock 70 kDa protein cognate 4 (KMQ87979.1) was up-regulated following either Amistar treatment alone (ratio 7.12 and $p \leq 0.05$) or when combined with C. bombi (ratio 12.76 and $p \leq 0.01$) compared to control. Additionally, we explored the dynamics of the DEPs and pathways and how they could be connected together. Cytoscape networks (Figure S6) illustrated the most important proteins (forming clusters) and their associated pathways that are key in the response to the stressors investigated in this paper. For Amistar, we identified a protein–pathway network with 166 nodes and 207 edges; among them, a cluster was formed with 22 proteins showing the highest number of inter-connexions. For glyphosate, the network consists of 41 nodes and 38 edges, with only 5 connected proteins. Lastly, for sulfoxaflor, we identified fewer dynamics with 26 nodes and 22 edges without connection between the corresponding DEPs. The average number of neighbours was 2.67, 2, and 1.83 for Amistar, glyphosate, and sulfoxaflor, respectively. ## 3. Material and Methods The experimental work comprises two sections, the experimental treatment and whole organism metrics, undertaken at Royal Holloway University of London, and the haemoproteome work, performed at BioPark (Archamps, France). To cover all three pesticides, two experiments were conducted, one with just Amistar, and one with both glyphosate and sulfoxaflor. ## 3.1. Bees Ten Bombus terrestris audax colonies were ordered from Agralan Ltd., Swindon, UK, for the glyphosate and sulfoxaflor experiments and three from Koppert Biological Systems, Haverhill, UK, for the azoxystrobin experiments. They were fed ad libitum sucrose and honey-bee-collected pollen from Thorne, Windsor, UK, and Agralan Ltd., Swindon, UK, respectively. All colonies were queenright. All experiments were performed in a temperature-controlled room at 25 °C ± 2 °C and $60\%$ RH ± $10\%$ RH. The room was kept in either darkness or red light so as to minimise stress to the bees. Ten workers per colony were screened for micro-parasites [43], with no infections detected. Only workers were used in the experiment. Bees were not age controlled as we were following an OECD protocol (see below). ## 3.2. Pesticides Glyphosate, sulfoxaflor, and dimethoate were sourced as pure active ingredients; (Sigma-Aldrich, St. Louis, MO, USA) CAS-no: 1071-83-6 (Greyhound Chromatography and Allied Chemicals) CAS-no: 946578-00-3, and (Sigma-Aldrich) CAS-no: 60-51-5, respectively. Azoxystrobin is poorly soluble in most solvents viable for bee testing, so the highly stable formulation (Amistar) was used instead. Amistar was purchased online through Agrigem Ltd. (www.agrigem.co.uk, accessed on 2 September 2019) (formulation identifiers are UK MAPP: 18039, Syngenta ID: A12705B). ## 3.3. Parasites The details of the parasite exposure are identical to that in the modified ecotoxicological protocol OECD 247 in [52]. Briefly, bees in parasite treatments were orally fed an inoculum of 10,000 C. bombi cells, which is known to lead to a field-realistic infection level [74,75]. Infection was validated by dissection after exposure, and only three samples were found to have a failed infection. The infection was allowed to develop for a week, prior to pesticide exposure. ## 3.4. Exposure Bees were allocated to treatments so as to ensure an even allocation of bees per colony per treatment. Bees were acutely and orally exposed to the pesticides, adapted from OECD 247 [76]. The exposure methodology is documented in full in [52] under the section marked modified ecotoxicological protocol OECD 247. Briefly, bees were fed the doses detailed in Table 1 and Table 2 in a 40 µL droplet of sucrose after 2–4 h of starvation. Mortality was recorded until haemolymph extraction, 48 h after exposure. The 200 µg dose of glyphosate and azoxystrobin (as 0.8 µL of Amistar) was chosen as the regulatory standard dose for a limit test. The 0.06 µg dose of sulfoxaflor was chosen as a high, but non-lethal, dose so as to simulate a worst-case sublethal acute exposure. Preliminary data from Alberto Linguadoca (pers. comm.) were used to derive the 0.06 µg value. The glyphosate and glyphosate + C. bombi whole organism results (survival, weight change, and parasite intensity) are reported and presented in full in [52] without the sulfoxaflor and sulfoxaflor + C. bombi results, which are presented here. The proteomic work on the glyphosate and glyphosate + C. bombi experiments is presented here only. ## 3.5. Metrics Survival, weight change, and parasite intensity were all recorded as per Straw and Brown [2021] [52]. For mortality, model assumptions for mixed effects and Cox proportional hazards models were not met, so chi-squared testing was used. Initially a global test was conducted, followed by individual comparisons of each treatment to the control. For the sulfoxaflor and glyphosate experiment, mortality was too low for chi-square testing, so Fisher’s exact tests were used. In treatments with no mortality, no comparison to the control was performed. Weight change and parasite intensity were analysed using mixed effects linear models. The model used was (Metric~Treatment + (1\|Colony)). As all dimethoate-exposed bees died within four hours, they were excluded from analyses. The parasite intensity analysis excluded treatments that were not parasite inoculated. The two experiments were analysed separately. ## 3.6. Haemolymph Extraction At 48 h post-exposure, bees were moved onto ice until docile (52 min ± 20 min). The bees were weighed to allow for a measurement of weight change from the start of the experiment. Haemolymph was collected according to the method established by Arafah et al. [ 2019] [63] using a specific haemolymph collection kit. Once docile, bees were held in place using plastic tubing, and their abdomen was punctured using a pulled glass capillary (Sutter Instrument Corp, Model P-30, Novato, CA, USA). The glass capillary was inserted dorsally under the second tergum of the abdomen. A 1–5 μL volume of haemolymph was extracted with light suction. Where a sample was cloudy or brown, it was excluded. The collected haemolymph was transferred into a chilled Eppendorf® LoBind Protein microtube (Sigma-Aldrich, St. Louis, MO, USA) pre-coated with PTU and PMSF to prevent melanisation and proteolysis, respectively. The anaesthetised bee was moved into a standard 1.5 mL Eppendorf® tube. Both the bee and sample were stored on ice and moved to a −20 °C freezer regularly. Haemolymph samples were shipped to BioPark on dry ice. ## 3.7. Batches The azoxystrobin experiment (Table 3) was run in a single batch, while the combined sulfoxaflor, glyphosate, and Crithidia bombi experiment was split over two days as two batches (Table 4). All experimental conditions were matched between batches, with only a day’s stagger separating the batches as part of a 10-day experiment ($82\%$ overlap). ## 3.8. Haemolymph Analyses: Chemicals and Reagents For sample preparation and analysis, acetonitrile (ACN) and trifluoroacetic acid (TFA), methanol, and ethanol of LC-MS grade quality or higher were obtained from Carlo Erba Reagents (Val de Reuil, France). For MALDI mass spectrometer calibration, two calibration kits Peptide Standard Calibration II and Protein Standard Calibration I from Bruker Daltonics (Bremen, Germany) were used. Ammonium bicarbonate (ABC), 1,4-dithiothreitol (DTT), 4-vinyl-pyridin (4-VP), phenylthiourea as melanisation inhibitor (PTU), phenylmethylsulfonyl fluoride (PMSF) as protease inhibitor, and alpha-cyano-4-hydroxycinnamic acid (4-HCCA) matrix were from Sigma-Aldrich (St. Louis, MO, USA). Trypsin solution was purchased from Promega. RapiGest™ SF was purchased from Waters. Ultrapure water (MilliQ water; Millipore, Billerica, MA, USA) was used. ## 3.9. Haemolymph Preparation for MALDI Molecular Mass Fingerprint (MFP) To obtain MFPs by MALDI mass spectrometry (MALDI BeeTyping®), haemolymph samples were handled according to the protocol published by Arafah et al. [ 2019] with modifications to optimise sample analysis [63]. Each individual haemolymph sample was analysed with an AutoFlex III Smartbeam® MALDI-TOF mass spectrometer (Bruker Daltonics, Germany). MFPs were acquired following the Bruker BioTyper® recommendations (matrix, method of sample deposition, and detection) with minor adjustments. Briefly, the haemolymph samples were diluted 1:10 in water acidified with $1\%$ TFA. A volume of 1 µL from each diluted sample was spotted on a MALDI MTP 384 polished ground steel plate (Bruker Daltonics), dried under gentle vacuum, and then mixed with 1 μL of 4-HCCA. Following co-crystallisation of the haemolymph spots with the matrix droplet, MALDI MFPs were recorded in a linear positive mode and in an automatic data acquisition using FlexControl software v3.4 (Bruker Daltonics). The samples were manually spotted in triplicate, each of the three spots being read three times. ## 3.10. MALDI BeeTyping® Acquisition For MALDI-MS analysis, the following instrument settings were used: 1.5 kV of electric potential difference, dynamic range of detection of 600–18,000 m/z, a global attenuator offset of $70\%$ with 200 Hz laser frequency, and 1000 accumulated laser shots per spectrum of haemolymph. The linear detector gain was set at 1.762 kV with a suppression mass gate up to m/z 600 to prevent detector saturation by clusters of the 4-HCCA matrix. Calibration of the mass spectrometer was performed using a standard mixture of peptides and proteins (Peptide Standard Calibration II and Protein Standard Calibration I, Bruker Daltonics) to cover the dynamic range selected for analysis. ## 3.11. Data Processing and Statistical Analyses MALDI-MS datasets were submitted to ClinProTools™ 2.2 Software (Bruker Daltonics) for post-processing and statistical analyses. Baseline subtraction and spectral smoothing were applied for all the acquired spectra. The total averaged spectra were calculated based on a signal-to-noise ratio equal to 3 for peak-picking and area calculations. The irrelevant spectra that did not pass the required signal intensity and resolution were excluded from the analysis. A post-processing step involving spectral normalisation of all calculated peak areas was performed with ClinProTools™ software prior to the generation of the principal component analysis (PCA). For intensity comparisons, we used Wilcoxon–Kruskal–Wallis tests. To test normality, we used the p of the Anderson–Darling test PAD: if close to 1, the data follow a normal distribution; if close to 0, they do not. In the latter case, further analyses used non-parametric tests. The peak lists generated from the software detail the number of ions (peaks) that are significant (PWKW < 0.0083 ($\frac{0.05}{6}$)) and are discriminant between the pairwise comparisons. The peak lists are also used to analyse the percentage of significant peaks considered in the experimental condition comparisons. ## 3.12. Bottom-Up Proteomics-Based Nano LC-MS/MS Based on the MFPs profiles, individual bees were selected to form pools for label-free quantitative bottom-up proteomics analyses by liquid chromatography–electrospray ionisation tandem mass spectrometry (LC-ESI-MS/MS). The same control and C. bombi pools were used for both the sulfoxaflor and glyphosate batches. For each experimental condition, three pools composed of five individual haemolymph samples were prepared. The pools were dried by vacuum centrifugation (Labconco, Kansas City, MO) before bottom-up proteomics studies according to Masson et al. [ 2018] [77] and Houdelet et al. [ 2021] [22]. Briefly, 20 µL of $0.1\%$ RapiGest in 50 mM ABC buffer was added to the samples. After adding 2 μL of 280 mM DTT (disulfide bond reducing agent), tubes were incubated at 56 °C for 45 min in the dark, centrifuged briefly, and then allowed to cool down. A 4 µL volume of 4-VP (alkylating agent to block cysteine residues) was added, followed by a 30 min incubation in the dark at room temperature. A 2 µL volume of 0.2 µg/µL trypsin solution (Promega) was used for protein digestion. The samples were incubated overnight at 37 °C under gentle agitation. To stop the enzymatic reaction and inactivate RapiGest, samples were acidified by 5 µL of ACN 20–$10\%$ TFA and incubated for 45 min at 37 °C. The digested samples were centrifuged for 10 min at 15,000 g, and 10 µL of the samples was analysed by LC/ESI-MS/MS using an U3000 nano-HPLC connected to a high-resolution Q-Exactive Orbitrap (all instruments Thermo Scientific). The tryptic digests were separated by reverse-phase chromatography on an Acclaim PepMap 100 C18 nanocolumn (75 μm internal diameter, 150 mm length, 3 μm granulometry, and 100 Å porosity; Thermo Fisher Scientific, Bremen, Germany) on-line with a concentration micro-precolumn C18 PepMap 100 (300 μm internal diameter, 3 μm granulometry, and 100 Å porosity; Thermo Fisher Scientific). The flow rate was set to 300 nL min−1 using a diphasic linear gradient of $0.1\%$ formic acid in water (FA, v/v) as mobile phase A and ACN with $0.1\%$ FA as mobile phase B. A multistep gradient of 155 min started at $2\%$ B for 6 min, reaching $35\%$ B in 120 min, then from $35\%$ to $70\%$ B in 5 min, followed by a plateau for 5 min. The gradient ended with a return to the initial mobile phase condition ($2\%$ B) for 4 min and a column stabilisation for 15 min. NanoLC-MS/MS datasets were acquired in positive-ion and data-dependent modes of analysis. Oxidation of methionine and tryptophan residues was selected for dynamic modification and pyridylethyl on cysteine for static modification. The protein databases used to perform the identifications were downloaded from NCBI and contained sequences from Hymenoptera and the relevant parasites. ## 3.13. Label-Free Quantification (LFQ) The Proteome Discoverer 2.4 (Thermo Fisher Scientific) was used to perform the label-free quantification. Using a consensus method, the ion-based quantification relied on unique and razor peptides, and the peptide abundance calculation was based on intensity following a normalisation of the datasets made of all the peptides characterised in the LC-MSMS runs. The protein quantification was calculated using the summed abundance with subsequent ANOVA tests. The processing workflow was performed on the retention time frame between 20 min and 135 min, with a precursor mass tolerance value set to 20 ppm and a fragment mass tolerance of 0.5 Da. The minimum trace length value was set to 5, and the maximum retention time shift of isotope pattern was equal to 0.2 min. Proteins with a ratio <0.5 (down-regulation) and >2 (up-regulation) were considered as significant along with $p \leq 0.05.$ A post-hoc test (Bonferroni) was considered in order to compare protein abundance between the experimental conditions. ## 3.14. Functional Annotation: Gene Ontology and Pathways Analysis For functional annotation of the sequences generated from the LC-ESI-MS/MS analyses, the bioinformatic solution OmicBox software (v2.1.10, functional analysis module Blast2Go https://www.biobam.com, accessed on 10 May 2022) was used. To obtain the most complete annotation labels, the analyses were performed using the four cloud-powered algorithms (Blast, InterProScan, GO Mapping, GO slim). Separate lists of dysregulated proteins of the pairwise comparisons were loaded to investigate the biological pathways and the protein functions following bee exposure to sulfoxaflor, Amistar, or glyphosate alone or combined with C. bombi. Combined pathway analysis was performed on the annotated sequences (proteins) joining Reactome and KEGG to identify enriched pathways with expression profiles. Furthermore, protein–protein interaction and pathway networks were constructed using Cytoscape (v3.9.1 https://cytoscape.org/, accessed on 27 May 2022). The network was statistically analysed as an undirected graph. The complete workflow of the experiments is presented in Figure 6. ## 4. Conclusions Neither the high dose of glyphosate, nor the sublethal dose of sulfoxaflor caused an observable effect on the whole organism, while the high dose of Amistar caused considerable impacts. However, these whole organism metrics do not capture the totality of the impact of the pesticides, and the haemolymph analysis revealed that, at the doses used in this study, sulfoxaflor has less impact on the B. terrestris haemoproteome than glyphosate and Amistar. The latter showed a higher impact across an array of biological processes than either glyphosate or sulfoxaflor. This was observed on the MFPs of individual bees and at the level of the whole haemolymph proteome. However, the trypanosome C. bombi showed almost no impact on haemolymph composition. Additional proteomic studies should be carried out on the gut tissue which is the initial target of the parasite C. bombi. ## References 1. Potts S.G., Imperatriz-Fonseca V., Ngo H.T., Aizen M.A., Biesmeijer J.C., Breeze T.D., Dicks L.V., Garibaldi L.A., Hill R., Settele J.. Safeguarding Pollinators and Their Values to Human Well-Being. Nature (2016) 540 220-229. DOI: 10.1038/nature20588 2. 2. European Commission Directorate General for the Environment IUCN (International Union for Conservation of Nature) European Red List of BeesPublications OfficeLuxembourg, LU, USA2014. European Red List of Bees (2014) 3. 3. IPBES The Assessment Report of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services on Pollinators, Pollination and Food ProductionZenodoOnline2016. The Assessment Report of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services on Pollinators, Pollination and Food Production (2016) 4. Hemberger J., Crossley M.S., Gratton C.. Historical Decrease in Agricultural Landscape Diversity Is Associated with Shifts in Bumble Bee Species Occurrence. Ecol. Lett. (2021) 24 1800-1813. DOI: 10.1111/ele.13786 5. Rundlöf M., Andersson G.K.S., Bommarco R., Fries I., Hederström V., Herbertsson L., Jonsson O., Klatt B.K., Pedersen T.R., Yourstone J.. Seed Coating with a Neonicotinoid Insecticide Negatively Affects Wild Bees. Nature (2015) 521 77-80. DOI: 10.1038/nature14420 6. McArt S.H., Fersch A.A., Milano N.J., Truitt L.L., Böröczky K.. High Pesticide Risk to Honey Bees despite Low Focal Crop Pollen Collection during Pollination of a Mass Blooming Crop. Sci. Rep. (2017) 7 46554. DOI: 10.1038/srep46554 7. Woodcock B.A., Isaac N.J.B., Bullock J.M., Roy D.B., Garthwaite D.G., Crowe A., Pywell R.F.. Impacts of Neonicotinoid Use on Long-Term Population Changes in Wild Bees in England. Nat. Commun. (2016) 7 12459. DOI: 10.1038/ncomms12459 8. Straub L., Strobl V., Yañez O., Albrecht M., Brown M.J.F., Neumann P.. Do Pesticide and Pathogen Interactions Drive Wild Bee Declines?. Int. J. Parasitol. Parasites Wildl. (2022) 18 232-243. DOI: 10.1016/j.ijppaw.2022.06.001 9. Tsvetkov N., Samson-Robert O., Sood K., Patel H.S., Malena D.A., Gajiwala P.H., Maciukiewicz P., Fournier V., Zayed A.. Chronic Exposure to Neonicotinoids Reduces Honey Bee Health near Corn Crops. Science (2017) 356 1395-1397. DOI: 10.1126/science.aam7470 10. Tantillo G., Bottaro M., Di Pinto A., Martella V., Di Pinto P., Terio V.. Virus Infections of Honeybees. Ital. J. Food Saf. (2015) 4 5364. DOI: 10.4081/ijfs.2015.5364 11. Tesovnik T., Cizelj I., Zorc M., Citar M., Bozic J., Glavan G., Narat M.. Immune Related Gene Expression in Worker Honey Bee (. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0187079 12. Al Naggar Y., Baer B.. Consequences of a Short Time Exposure to a Sublethal Dose of Flupyradifurone (Sivanto) Pesticide Early in Life on Survival and Immunity in the Honeybee (. Sci. Rep. (2019) 9 19753. DOI: 10.1038/s41598-019-56224-1 13. Martin C.D., Fountain M.T., Brown M.J.F.. Varietal and Seasonal Differences in the Effects of Commercial Bumblebees on Fruit Quality in Strawberry Crops. Agric. Ecosyst. Environ. (2019) 281 124-133. DOI: 10.1016/j.agee.2019.04.007 14. Koziy R.V., Wood S.C., Kozii I.V., van Rensburg C.J., Moshynskyy I., Dvylyuk I., Simko E.. Deformed Wing Virus Infection in Honey Bees (. Vet. Pathol. (2019) 56 636-641. DOI: 10.1177/0300985819834617 15. El-Seedi H.R., Ahmed H.R., El-Wahed A.A.A., Saeed A., Algethami A.F., Attia N.F., Guo Z., Musharraf S.G., Khatib A., Alsharif S.M.. Bee Stressors from an Immunological Perspective and Strategies to Improve Bee Health. Vet. Sci. (2022) 9. DOI: 10.3390/vetsci9050199 16. Balakrishnan B., Wu H., Cao L., Zhang Y., Li W., Han R.. Immune Response and Hemolymph Microbiota of. J. Econ. Entomol. (2021) 114 1310-1320. DOI: 10.1093/jee/toab047 17. Le Conte Y., Navajas M.. Climate Change: Impact on Honey Bee Populations and Diseases. Rev. Sci. Tech. Int. Off. Epizoot. (2008) 27 485-510. DOI: 10.20506/rst.27.2.1819 18. Grassl J., Holt S., Cremen N., Peso M., Hahne D., Baer B.. Synergistic Effects of Pathogen and Pesticide Exposure on Honey Bee (. J. Invertebr. Pathol. (2018) 159 78-86. DOI: 10.1016/j.jip.2018.10.005 19. Havard T., Laurent M., Chauzat M.-P.. Impact of Stressors on Honey Bees (. Diversity (2019) 12. DOI: 10.3390/d12010007 20. Breda D., Frizzera D., Giordano G., Seffin E., Zanni V., Annoscia D., Topping C.J., Blanchini F., Nazzi F.. A Deeper Understanding of System Interactions Can Explain Contradictory Field Results on Pesticide Impact on Honey Bees. Nat. Commun. (2022) 13 5720. DOI: 10.1038/s41467-022-33405-7 21. Schöning C., Gisder S., Geiselhardt S., Kretschmann I., Bienefeld K., Hilker M., Genersch E.. Evidence for Damage-Dependent Hygienic Behaviour towards. J. Exp. Biol. (2012) 215 264-271. DOI: 10.1242/jeb.062562 22. Houdelet C., Sinpoo C., Chantaphanwattana T., Voisin S.N., Bocquet M., Chantawannakul P., Bulet P.. Proteomics of Anatomical Sections of the Gut of. J. Proteome Res. (2021) 20 804-817. DOI: 10.1021/acs.jproteome.0c00658 23. Vázquez D.E., Ilina N., Pagano E.A., Zavala J.A., Farina W.M.. Glyphosate Affects the Larval Development of Honey Bees Depending on the Susceptibility of Colonies. PLoS ONE (2018) 13. DOI: 10.1371/journal.pone.0205074 24. Nicewicz Ł., Nicewicz A.W., Kafel A., Nakonieczny M.. Set of Stress Biomarkers as a Practical Tool in the Assessment of Multistress Effect Using Honeybees from Urban and Rural Areas as a Model Organism: A Pilot Study. Environ. Sci. Pollut. Res. (2021) 28 9084-9096. DOI: 10.1007/s11356-020-11338-2 25. Vázquez D.E., Latorre-Estivalis J.M., Ons S., Farina W.M.. Chronic Exposure to Glyphosate Induces Transcriptional Changes in Honey Bee Larva: A Toxicogenomic Study. Environ. Pollut. (2020) 261 114148. DOI: 10.1016/j.envpol.2020.114148 26. Guo Y., Zhang Z., Zhuang M., Wang L., Li K., Yao J., Yang H., Huang J., Hao Y., Ying F.. Transcriptome Profiling Reveals a Novel Mechanism of Antiviral Immunity Upon Sacbrood Virus Infection in Honey Bee Larvae (. Front. Microbiol. (2021) 12 615893. DOI: 10.3389/fmicb.2021.615893 27. Motta E.V.S., Powell J.E., Moran N.A.. Glyphosate Induces Immune Dysregulation in Honey Bees. Anim. Microbiome (2022) 4 16. DOI: 10.1186/s42523-022-00165-0 28. Haas J., Hayward A., Buer B., Maiwald F., Nebelsiek B., Glaubitz J., Bass C., Nauen R.. Phylogenomic and Functional Characterization of an Evolutionary Conserved Cytochrome P450-Based Insecticide Detoxification Mechanism in Bees. Proc. Natl. Acad. Sci. USA (2022) 119 e2205850119. DOI: 10.1073/pnas.2205850119 29. Johnson R.M., Mao W., Pollock H.S., Niu G., Schuler M.A., Berenbaum M.R.. Ecologically Appropriate Xenobiotics Induce Cytochrome P450s in. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0031051 30. Hu Y.-T., Wu T.-C., Yang E.-C., Wu P.-C., Lin P.-T., Wu Y.-L.. Regulation of Genes Related to Immune Signaling and Detoxification in. Sci. Rep. (2017) 7 41255. DOI: 10.1038/srep41255 31. Zhao H., Li G., Guo D., Wang Y., Liu Q., Gao Z., Wang H., Liu Z., Guo X., Xu B.. Transcriptomic and Metabolomic Landscape of the Molecular Effects of Glyphosate Commercial Formulation on. Sci. Total Environ. (2020) 744 140819. DOI: 10.1016/j.scitotenv.2020.140819 32. Słowińska M., Nynca J., Bąk B., Wilde J., Siuda M., Ciereszko A.. 2D-DIGE Proteomic Analysis Reveals Changes in Haemolymph Proteome of 1-Day-Old Honey Bee (. Apidologie (2019) 50 632-656. DOI: 10.1007/s13592-019-00674-z 33. Zaluski R., Bittarello A.C., Vieira J.C.S., Braga C.P., Padilha P.d.M., Fernandes M.d.S., Bovi T.d.S., Orsi R.d.O.. Modification of the Head Proteome of Nurse Honeybees (. Sci. Rep. (2020) 10 2190. DOI: 10.1038/s41598-020-59070-8 34. Roat T.C., dos Santos-Pinto J.R.A., dos Santos L.D., Santos K.S., Malaspina O., Palma M.S.. Modification of the Brain Proteome of Africanized Honeybees (. Ecotoxicology (2014) 23 1659-1670. DOI: 10.1007/s10646-014-1305-8 35. Lundin O., Rundlöf M., Smith H.G., Fries I., Bommarco R.. Neonicotinoid Insecticides and Their Impacts on Bees: A Systematic Review of Research Approaches and Identification of Knowledge Gaps. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0136928 36. Yordanova M., Evison S.E.F., Gill R.J., Graystock P.. The Threat of Pesticide and Disease Co-Exposure to Managed and Wild Bee Larvae. Int. J. Parasitol. Parasites Wildl. (2022) 17 319-326. DOI: 10.1016/j.ijppaw.2022.03.001 37. Cullen M.G., Thompson L.J., Carolan J.C., Stout J.C., Stanley D.A.. Fungicides, Herbicides and Bees: A Systematic Review of Existing Research and Methods. PLoS ONE (2019) 14. DOI: 10.1371/journal.pone.0225743 38. Michener C.D.. The Bees of the World (2000) 39. Dafni A., Kevan P., Gross C.L., Goka K.. Appl. Entomol. Zool. (2010) 45 101-113. DOI: 10.1303/aez.2010.101 40. Garibaldi L.A., Steffan-Dewenter I., Winfree R., Aizen M.A., Bommarco R., Cunningham S.A., Kremen C., Carvalheiro L.G., Harder L.D., Afik O.. Wild Pollinators Enhance Fruit Set of Crops Regardless of Honey Bee Abundance. Science (2013) 339 1608-1611. DOI: 10.1126/science.1230200 41. Shykoff J.A., Schmid-Hempel P.. Incidence and Effects of Four Parasites in Natural Populations of Bumble Bees in Switzerland. Apidologie (1991) 22 117-125. DOI: 10.1051/apido:19910204 42. Korner P., Schmid-Hempel P.. Correlates of Parasite Load in Bumblebees in an Alpine Habitat. Entomol. Sci. (2005) 8 151-160. DOI: 10.1111/j.1479-8298.2005.00113.x 43. Rutrecht S.T., Brown M.J.. The Life-History Impact and Implications of Multiple Parasites for Bumble Bee Queens. Int. J. Parasitol. (2008) 38 799-808. DOI: 10.1016/j.ijpara.2007.11.004 44. Jones C.M., Brown M.J.F.. Parasites and Genetic Diversity in an Invasive Bumblebee. J. Anim. Ecol. (2014) 83 1428-1440. DOI: 10.1111/1365-2656.12235 45. Siviter H., Scott A., Pasquier G., Pull C.D., Brown M.J.F., Leadbeater E.. No Evidence for Negative Impacts of Acute Sulfoxaflor Exposure on Bee Olfactory Conditioning or Working Memory. PeerJ (2019) 7 e7208. DOI: 10.7717/peerj.7208 46. Siviter H., Brown M.J.F., Leadbeater E.. Sulfoxaflor Exposure Reduces Bumblebee Reproductive Success. Nature (2018) 561 109-112. DOI: 10.1038/s41586-018-0430-6 47. Vaughan O.P., Straw E.A., Linguadoca A., Brown M.J.F.. No Effect of Dual Exposure to Sulfoxaflor and a Trypanosome Parasite on Bumblebee Olfactory Learning. Sci. Rep. (2022) 12 8611. DOI: 10.1038/s41598-022-12714-3 48. Straw E.A., Brown M.J.F.. Co-Formulant in a Commercial Fungicide Product Causes Lethal and Sub-Lethal Effects in Bumble Bees. Sci. Rep. (2021) 11 21653. DOI: 10.1038/s41598-021-00919-x 49. Duke S.O., Powles S.B.. Glyphosate: A Once-in-a-Century Herbicide: Glyphosate: A Once-in-a-Century Herbicide. Pest Manag. Sci. (2008) 64 319-325. DOI: 10.1002/ps.1518 50. Benbrook C.M.. Trends in Glyphosate Herbicide Use in the United States and Globally. Environ. Sci. Eur. (2016) 28 3. DOI: 10.1186/s12302-016-0070-0 51. Battisti L., Potrich M., Sampaio A.R., de Castilhos Ghisi N., Costa-Maia F.M., Abati R., Dos Reis Martinez C.B., Sofia S.H.. Is Glyphosate Toxic to Bees? A Meta-Analytical Review. Sci. Total Environ. (2021) 767 145397. DOI: 10.1016/j.scitotenv.2021.145397 52. Straw E.A., Brown M.J.F.. No Evidence of Effects or Interaction between the Widely Used Herbicide, Glyphosate, and a Common Parasite in Bumble Bees. PeerJ (2021) 9 e12486. DOI: 10.7717/peerj.12486 53. Motta E.V.S., Raymann K., Moran N.A.. Glyphosate Perturbs the Gut Microbiota of Honey Bees. Proc. Natl. Acad. Sci. USA (2018) 115 10305-10310. DOI: 10.1073/pnas.1803880115 54. Thompson H.M., Levine S.L., Doering J., Norman S., Manson P., Sutton P., von Mérey G.. Evaluating Exposure and Potential Effects on Honeybee Brood (. Integr. Environ. Assess. Manag. (2014) 10 463-470. DOI: 10.1002/ieam.1529 55. Odemer R., Rosenkranz P.. Chronic Exposure to a Neonicotinoid Pesticide and a Synthetic Pyrethroid in Full-Sized Honey Bee Colonies. J. Apic. Res. (2020) 59 2-11. DOI: 10.1080/00218839.2019.1675337 56. Almasri H., Tavares D.A., Pioz M., Sené D., Tchamitchian S., Cousin M., Brunet J.-L., Belzunces L.P.. Mixtures of an Insecticide, a Fungicide and a Herbicide Induce High Toxicities and Systemic Physiological Disturbances in Winter. Ecotoxicol. Environ. Saf. (2020) 203 111013. DOI: 10.1016/j.ecoenv.2020.111013 57. Brown M.J.F., Loosli R., Schmid-Hempel P.. Condition-Dependent Expression of Virulence in a Trypanosome Infecting Bumblebees. Oikos (2000) 91 421-427. DOI: 10.1034/j.1600-0706.2000.910302.x 58. Brown M.J., Moret Y., Schmid-Hempel P.. Activation of Host Constitutive Immune Defence by an Intestinal Trypanosome Parasite of Bumble Bees. Parasitology (2003) 126 253-260. DOI: 10.1017/S0031182002002755 59. Gegear R.J., Otterstatter M.C., Thomson J.D.. Bumble-Bee Foragers Infected by a Gut Parasite Have an Impaired Ability to Utilize Floral Information. Proc. R. Soc. B Biol. Sci. (2006) 273 1073-1078. DOI: 10.1098/rspb.2005.3423 60. Otterstatter M.C., Thomson J.D.. Contact Networks and Transmission of an Intestinal Pathogen in Bumble Bee (. Oecologia (2007) 154 411-421. DOI: 10.1007/s00442-007-0834-8 61. Otterstatter M.C., Thomson J.D.. Within-Host Dynamics of an Intestinal Pathogen of Bumble Bees. Parasitology (2006) 133 749-761. DOI: 10.1017/S003118200600120X 62. Yourth C.P., Brown M.J.F., Schmid-Hempel P.. Effects of Natal and Novel. Insectes Sociaux (2008) 55 86-90. DOI: 10.1007/s00040-007-0974-1 63. Arafah K., Voisin S.N., Masson V., Alaux C., Le Conte Y., Bocquet M., Bulet P.. MALDI–MS Profiling to Address Honey Bee Health Status under Bacterial Challenge through Computational Modeling. Proteomics (2019) 19 1900268. DOI: 10.1002/pmic.201900268 64. Houdelet C., Bocquet M., Bulet P.. Matrix-assisted Laser Desorption/Ionization Mass Spectrometry Biotyping, an Approach for Deciphering and Assessing the Identity of the Honeybee Pathogen. Rapid Commun. Mass Spectrom. (2021) 35 e8980. DOI: 10.1002/rcm.8980 65. Ober C., Chupp G.L.. The Chitinase and Chitinase-like Proteins: A Review of Genetic and Functional Studies in Asthma and Immune-Mediated Diseases. Curr. Opin. Allergy Clin. Immunol. (2009) 9 401-408. DOI: 10.1097/ACI.0b013e3283306533 66. Pham D.Q.D., Winzerling J.J.. Insect Ferritins: Typical or Atypical?. Biochim. Biophys. Acta BBA-Gen. Subj. (2010) 1800 824-833. DOI: 10.1016/j.bbagen.2010.03.004 67. Wang Y.H., Hu Y., Xing L.S., Jiang H., Hu S.N., Raikhel A.S., Zou Z.. A Critical Role for CLSP2 in the Modulation of Antifungal Immune Response in Mosquitoes. PLoS Pathog. (2015) 11. DOI: 10.1371/journal.ppat.1004931 68. Park H.G., Lee K.S., Kim B.Y., Yoon H.J., Choi Y.S., Lee K.Y., Wan H., Li J., Jin B.R.. Honeybee (. Dev. Comp. Immunol. (2018) 85 51-60. DOI: 10.1016/j.dci.2018.04.001 69. Wang Z., Zhou W., Huang B., Gao M., Li Q., Tao Y., Wang Z.. Molecular and Functional Characterization of Peptidoglycan Recognition Proteins OfPGRP-A and OfPGRP-B in. Insects (2022) 13. DOI: 10.3390/insects13050417 70. Kodrík D., Ibrahim E., Gautam U.K., Frydrychová R.Č., Bednářová A., Krištůfek V., Jedlička P.. Changes in Vitellogenin Expression Caused by Nematodal and Fungal Infections in Insects. J. Exp. Biol. (2019) 222 jeb.202853. DOI: 10.1242/jeb.202853 71. Eichelberger K.R., Goldman W.E.. Manipulating Neutrophil Degranulation as a Bacterial Virulence Strategy. PLOS Pathog. (2020) 16. DOI: 10.1371/journal.ppat.1009054 72. El-Benna J., Hurtado-Nedelec M., Gougerot-Pocidalo M.-A., Dang P.M.-C.. Effects of Venoms on Neutrophil Respiratory Burst: A Major Inflammatory Function. J. Venom. Anim. Toxins Trop. Dis. (2021) 27 e20200179. DOI: 10.1590/1678-9199-jvatitd-2020-0179 73. Gierlikowska B., Stachura A., Gierlikowski W., Demkow U.. Phagocytosis, Degranulation and Extracellular Traps Release by Neutrophils—The Current Knowledge, Pharmacological Modulation and Future Prospects. Front. Pharmacol. (2021) 12 666732. DOI: 10.3389/fphar.2021.666732 74. Baer B., Schmid-Hempel P.. Experimental Variation in Polyandry Affects Parasite Loads and Fitness in a Bumble-Bee. Nature (1999) 397 151-154. DOI: 10.1038/16451 75. Ruiz-González M.X., Kelly M., Moret Y., Brown M.J.F.. Parasite Resistance and Immunity across Female Castes in a Social Insect. Behav. Ecol. Sociobiol. (2022) 76 56. DOI: 10.1007/s00265-022-03162-0 76. Test No. 247: Bumblebee, Acute Oral Toxicity Test. OECD Guidelines for the Testing of Chemicals, Section 2 (2017) 77. Masson V., Arafah K., Voisin S., Bulet P.. Comparative Proteomics Studies of Insect Cuticle by Tandem Mass Spectrometry: Application of a Novel Proteomics Approach to the Pea Aphid Cuticular Proteins. Proteomics (2018) 18 1700368. DOI: 10.1002/pmic.201700368
--- title: Particulate Matter Elevates Ocular Inflammation and Endoplasmic Reticulum Stress in Human Retinal Pigmented Epithelium Cells authors: - Sunyoung Jeong - Eui-Cheol Shin - Jong-Hwa Lee - Jung-Heun Ha journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049273 doi: 10.3390/ijerph20064766 license: CC BY 4.0 --- # Particulate Matter Elevates Ocular Inflammation and Endoplasmic Reticulum Stress in Human Retinal Pigmented Epithelium Cells ## Abstract Because of their exposure to air, eyes can come into contact with air pollutants such as particulate matter (PM), which may cause severe ocular pathologies. Prolonged ocular PM exposure may increase inflammation and endoplasmic reticulum stress in the retina. Herein, we investigated whether PM exposure induces ocular inflammation and endoplasmic reticulum (ER) stress-related cellular responses in human retinal epithelium-19 (ARPE-19) cells. To understand how PM promotes ocular inflammation, we monitored the activation of the mitogen-activated protein kinase (MAPK)/nuclear factor kappa beta (NFκB) axis and the expression of key inflammatory mRNAs. We also measured the upregulation of signature components for the ER-related unfolded protein response (UPR) pathways, as well as intracellular calcium ([Ca2+]i) levels, as readouts for ER stress induction following PM exposure. Ocular PM exposure significantly elevated the expression of multiple cytokine mRNAs and increased phosphorylation levels of NFκB-MAPK axis in a PM dose-dependent manner. Moreover, incubation with PM significantly increased [Ca2+]i levels and the expression of UPR-related proteins, which indicated ER stress resulting from cell hypoxia, and upregulation of hypoxic adaptation mechanisms such as the ER-associated UPR pathways. Our study demonstrated that ocular PM exposure increased inflammation in ARPE-19 cells, by activating the MAPK/NFκB axis and cytokine mRNA expression, while also inducing ER stress and stress adaptation responses. These findings may provide helpful insight into clinical and non-clinical research examining the role of PM exposure in ocular pathophysiology and delineating its underlying molecular mechanisms. ## 1. Introduction Prolonged exposure to air pollutants poses a threat to human health and the quality of life [1,2]. Numerous epidemiological studies have demonstrated that exposure to airborne particulate matter (PM) is associated with a significant increase in hospitalization and mortality rates [3,4]. In 2013, the International Agency for Research on Cancer classified PM as a Group 1 carcinogen to humans [5]. Moreover, approximately $99\%$ of the global population is exposed to PM concentrations above the safety limit standards for air quality recommended by the World Health Organization. The PM is a mixture of solid and liquid particles suspended in air and consists of various components, such as elemental carbon (soot), organic carbon (including polycyclic aromatic hydrocarbons [PAH], nitro-PAHs, and endotoxins), sulfate, nitrate, and minerals [6]. There are three common categorization groups for PM depending on the aerodynamic diameter of its particles: [1] coarse (PM10; ≤10 μm), [2] fine (PM2.5; ≤2.5 μm), and [3] ultrafine (PM0.1; ≤0.1 μm) particles [7]. Inhaled PM is primarily deposited in the upper respiratory tract, whereas fine particles are delivered to the lower respiratory tract [8,9]. Moreover, ultrafine PM can be delivered into the bloodstream [8]. Therefore, the distribution and toxicity of PM mainly depend on particle size [8]. Exposure to PM has been reported to adversely affect the respiratory [9], cardiovascular [10], renal [11,12], hepatic [13], dermal [14], and nervous systems [15]. PM can also come into contact and affect the visual system, either via the bloodstream, which is initiated by pulmonary inhalation [16] or through the ocular surface [17,18]. Ocular PM exposure causes various types of discomfort to the eye, such as redness, itching, foreign substances, and dryness [19,20]. The cornea and conjunctiva are directly exposed to external toxicants since they are the outermost protective layers of the eye, and are therefore more vulnerable to suspended air particles, compared to posterior regions, such as the retina. Because of these characteristics, dry eye syndrome (DES) and allergic conjunctivitis closely correlate with ocular PM exposure [21,22]. Recent epidemiological studies have supported that PM exposure closely correlates with an increase in retinal, as well as ocular surface diseases. Higher PM exposure has been associated with an increased prevalence of age-related macular degeneration (AMD) in South Korea [23], the UK [24], Canada [25], and Taiwan [26]. Moreover, patients with diabetes who are also chronically exposed to PM, present with a significantly increased risk of diabetic retinopathy (DR) in Taiwan [27] and China [28]. In a cohort study, participants exposed to higher PM and NO2 concentrations showed adverse retinal structure features in the UK [29]. Short-term (24 h exposure to PM significantly narrowed retinal vessels, as well as both arterioles and venules in healthy adults [30]. Toxicological studies suggest that acute respiratory PM exposure causes retinal edema, which may be mediated by hypoxic responses [12]. In another rodent-based study, PM exposure in the eye triggered retinal vascular permeability, with retinal edema and inflammation [31]. Long-term PM exposure in the rat eye markedly reduced the thickness of the total retinal layer, including an observed decrease in the thickness of the nerve fiber layer/ganglion cell layer in the retina [32]. At a molecular level, PM treatment increased c-Jun N-terminal kinase (JNK) phosphorylation and led to increased secretion of interleukin-6 (IL-6) and tumor necrosis factor alpha (TNFα) in human retinal pigment epithelium (RPE) cells [33]. Furthermore, ocular PM exposure may interrupt intact ocular function by inducing intracellular production of reactive oxygen species, mitochondrial dysfunction, and enhancing epithelial-mesenchymal transition, in spontaneously arising RPE 19 (ARPE-19) cells [34]. However, a more robust toxicological mechanism is required to understand how ocular PM exposure disrupts ocular homeostasis. Inflammation is considered one of the main effects of PM exposure in various organs including the eye [15,35,36,37]. In rodent studies, several lines of evidence have supported that topical PM administration on the eye induces symptoms similar to clinical DES [17,21,37,38] or conjunctivitis [18,37]. Currently, only limited publications demonstrate that PM exposure triggers retinal inflammation [31,33]. Topical PM administration (10 μg, four times daily, for two days) reportedly increased inflammatory mRNA expression of genes such as intercellular adhesion molecule 1 (Icam1), lymphocyte common antigen (CD45), and the nucleotide-binding oligomerization domain-like receptor family pyrin domain containing 3 (Nlrp3), in the rat retina [31]. In addition, PM exposure has also been linked to increased inflammatory protein expression and leukocyte infiltration into retinal vessels [31]. In another study, a localized PM eye drop (2 μg, twice daily, for 21 days) caused a significant degree of myopia, with ocular inflammation in Syrian hamsters, and drastically elevated inflammatory TNFα and IL-6 protein expression in the retina, cornea, and sclera [39]. Similarly to topical PM exposure, whole-body exposure (long-term; six months) can also reportedly induce retinal inflammatory cytokines release in mice, including TNFα and cleaved IL-1β, while concomitantly causing retinal thinning and apoptosis, and ultimately lead to poor response to light stimuli [40]. There is a growing literature that suggests ocular inflammation and endoplasmic reticulum (ER) stress are closely intertwined and are strong pathological triggers for DR [41,42,43,44], AMD [42,43,44,45], and retinitis pigmentosa [42,43,44]. To date, there is no direct association between PM exposure and ocular ER stress. However, PM exposure has been shown to increase unfolded protein responses (UPRs) in other organs. Prolonged PM inhalation induces ER stress with UPR activation in lung and liver tissues of murine models [46], and chronic PM inhalation (10 months) also triggers macrophage infiltration in mouse white adipose tissue in vivo [47]. In addition, whole-body exposure to PM for 3–6 months deteriorates renal function, with autophagy induction, UPRs, and apoptosis in Sprague Dawley rats [48]. Since ocular PM exposure triggers inflammation, we postulated whether it may also increase ocular ER stress and the associated adaptive UPRs. Here, we investigated whether PM exposure induced ocular inflammation and ER stress-related cellular responses in ARPE-19 cells. To examine how PM promotes ocular inflammation, we monitored the mitogen-activated protein kinase (MAPK)/nuclear factor kappa beta (NFκB) axis activation, as well as the expression of inflammatory cytokine mRNAs. Moreover, ocular UPR pathways induction and intracellular calcium ([Ca2+]i) levels were also monitored, to delineate the mechanism behind PM-dependent retinal ER stress. ## 2.1. PM Preparation PM (PM10-like; European reference material ERM-CZ120) was purchased from Sigma-Aldrich (St. Louis, MO, USA), and resuspended in normal phosphate buffer saline (PBS; 100 mg/mL or diluted as needed), aliquoted, and stored at −20 °C until future use. Suspensions were sonicated (Powersonic 420; Hwasin Technology Co., Seoul, Republic of Korea) for 20 min and vortexed for 1 min before each experiment to minimize aggregation. ## 2.2. Cell Culture The ARPE-19 cell line (CRL-2302) was purchased from the American Type Culture Collection (Manassas, VA, USA) and cultured in Dulbecco’s modified eagle medium (DMEM/F12; Gibco, Carlsbad, CA, USA) supplemented with $10\%$ fetal bovine serum (FBS; Biowest, Nuaillé, France), and $1\%$ gentamicin (Gibco), at a $5\%$ CO2, 37 °C incubator (Vision Scientific Co., Ltd., Daejeon, Republic of Korea). ## 2.3. PM Treatment The ARPE-19 cells were plated in 6-well plates (SPL Life Sciences Co., Ltd., Pocheon, Republic of Korea) at a density of 5.0 × 105 cells/well and cultured for 48 h. Our experiments were conducted using the following research designs:For inflammatory response, cells were exposed to varying doses of PM (0, 50, 100, 250, or 500 μg/mL) for 30 min (for western blot analysis) or 2 h (for qRT-PCR analysis). To understand whether mitogen-activated protein kinase kinase (MEK) is involved in PM-induced ocular inflammation, cells were pretreated for 30 min with a mixture of 20 μM U0126; a selective inhibitor of MEK$\frac{1}{2}$, and 500 μg/mL PM.For ER stress-inducing conditions, cells were treated with 2 mg/mL of PM for 0, 2, 4, 6, and 8 h, while $0.01\%$ Tween-80 was used as vehicle treatment. ## 2.4. Cell Viability Assay Cell viability was examined by the 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) assay. The ARPE-19 cells were seeded onto the bottom of a 96-well plate (SPL Life Sciences Co., Ltd.) at a concentration of 2.0 × 104 cells per well. The MTT assay was performed under two different conditions: [1] to determine the cytotoxicity of PM, cells were treated with various doses of PM (0, 25, 50,100, 250, 500, 1000, and 2000 μg/mL, respectively), or PBS as a control, for 2 h, and [2] to examine ER stress-related responses, $0.01\%$ Tween-80 and 2000 μg/mL of PM were co-treated for 4 or 8 h. The MTT solution (1 mg/mL) was then added to each well, followed by incubation at 37 °C for 3 h. The medium was removed carefully and formazan crystals were dissolved in 100 μL dimethyl sulfoxide (DMSO; Daejung Chemicals & Metals Co., Ltd., Siheung, Republic of Korea). Absorbance was measured at 540 nm using a microplate reader. Each condition was replicated across five wells on each plate. ## 2.5. Western Blot Analysis To extract the total protein, ARPE-19 cells were washed thrice with ice-cold PBS and lysed in radioimmunoprecipitation assay (RIPA) buffer (ATTO, Tokyo, Japan) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific, Waltham, MA, USA). The protein concentration was determined using the bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific, Waltham, MA, USA). Thirty micrograms of protein were loaded and separated on $10\%$ sodium dodecyl sulfate (SDS) polyacrylamide gels, and subsequently transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad Laboratories, Hercules, CA, USA) using a wet-transfer method. Next, blots were blocked with $5\%$ skim milk (BD Difco, Franklin Lakes, NJ, USA) for 1 h at room temperature, and were then incubated with primary antibodies overnight at 4 °C. After washing, the membranes were incubated with secondary horseradish peroxidase (HRP)-conjugated antibodies at room temperature for 1 h. Finally, the blots were developed using a chemiluminescent reagent (Thermo Fisher Scientific, Waltham, MA, USA). The specific bands were visualized on a Davinch Western™ Imaging system (Davinch-K, Seoul, Republic of Korea) and then quantified using ImageJ software version 1.53k (National Institutes of Health, Bethesda, MD, USA). Each band was normalized to the relevant total protein amount of the glyceraldehyde 3-phosphate dehydrogenase (GAPDH), which served as an internal loading control. The primary and secondary antibodies used for western blotting are summarized in Table 1. ## 2.6. Quantitative RT-PCR Analysis Total RNA was extracted using a NucleoSpin RNA Plus kit (Macherey-Nagel, Düren, Germany) according to the manufacturer’s instructions. Then, 1 μg of total RNA was reverse-transcribed to cDNA using ReverTra Ace™ qPCR RT Master Mix (Toyobo, Osaka, Japan). Real-time PCR was performed using a RealMOD™ Green W2 2x qPCR mix (Intron Biotechnology, Sungnam, Republic of Korea) according to the manufacturer’s protocol. The relative gene expression was normalized to that of 36B4 [49], which did not vary significantly with treatment. The PCR primers used are listed in Table 2. The ANKRD37 primers were purchased from Bio-Rad Laboratories (unique assay ID: qHsaCID0010971; Hercules, CA, USA). ## 2.7. Intracellular Calcium Release Cytosolic [Ca2+]i was measured using a Fluo-4 NW Calcium Assay kit (F36206; Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s protocol. The ARPE-19 cells were plated at a density of 2.0 × 104 cells/well in a dark-walled 96-well plate (Corning, Corning, NY, USA) and cultured overnight. The next day, cells were gently washed with HBSS, loaded with 100 μL of Fluo-4 NW dye and incubated at 37 °C for 30 min in the dark. Following the addition of PM, fluorescence was monitored using a microplate reader (Molecular Devices, San Jose, CA, USA). ## 2.8. Statistical Analysis All data are expressed as means ± standard deviations (SDs). Significant differences were statistically analyzed using one-way analysis of variance (ANOVA), followed by Tukey’s post hoc analysis. Differences were considered statistically significant at $p \leq 0.05$ and are further indicated by a filled asterisk above each bar. To test the efficacy of U0126, the 500 and 500+U0126 groups were compared by Student’s t-test. Statistically significant differences were defined as $p \leq 0.05$ and are indicated by a filled pound sign above the bar for the 500+U0126 group. All statistical analyses were performed using the GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA). ## 3.1. PM Exposure Triggers Ocular Inflammatory Responses in ARPE-19 Cells The ARPE-19 cells were treated with 0–2000 μg/mL of PM for two hours, and cellular viability was measured by the MTT assay. Our results indicated that PM did not significantly affect cell viability at any of the tested concentrations (Figure 1). To determine whether PM exposure induced MAPK/NFκB axis-dependent ocular inflammatory responses in ARPE-19 cells, we examined the phosphorylation status of p38 MAPK (p38), JNK, and extracellular signal-regulated kinase (ERK), as well as NFκB inhibitor alpha (IκBα) and NFκB, by western blot analysis. Exposure to PM significantly increased the phosphorylated (p) portion of all MAPKs tested, in a dose-dependent manner (Figure 2). The amount of p-p38 protein significantly increased compared to total (t) p38, after exposure at PM concentrations ≥100 μg/mL, in direct correlation with the PM amount present (Figure 2B). Similarly, phosphorylation of JNK and ERK was also notably enhanced by ocular PM exposure at concentrations ≥100 μg/mL (Figure 2C,D). Posttranslational phosphorylation of MAPKs is closely associated with IκBα degradation and the NFκB’s translocation to the nucleus. As expected, phosphorylation of IκBα and NFκB in human retinal epithelial cells was increased following exposure to PM at doses ≥100 μg/mL and 500 μg/mL, respectively (Figure 2E,F). To identify the potential mechanism of the MAPKs/NFκB axis activation by ocular PM exposure, an ERK inhibitor, U0126 was used prior to PM incubation. The U0126 pretreatment with PM significantly prevented the phosphorylation of p38, ERK, IκBα, and NFκB (Figure 2). This result indicates that activation of the MAPK/NFκB axis by PM exposure likely depends on the phospho-activation of ERK. Moreover, ocular inflammatory responses are regulated not only by posttranslational modifications but also by transcriptional mechanisms. Therefore, we analyzed the mRNA expression of ocular cytokines following PM exposure. As shown in Figure 3, PM treatment significantly elevated expression of the ocular cytokine mRNAs of TNFα ([PM] ≥ 250 μg/mL), IL-1β ([PM] ≥ 50 μg/mL), IL-6 ([PM] ≥ 250 μg/mL), and monocyte chemoattractant protein-1 (MCP-1; [PM] ≥ 50 μg/mL) (Figure 3). Similarly to our described results for MAPK/NFκB, ERK inhibitor treatment also significantly attenuated the PM-induced ocular cytokine mRNA expressions (Figure 3). These results indicate that PM promotes ocular inflammation via the activation of NFκB-related inflammatory pathways and upregulation of inflammatory cytokines. ## 3.2. PM Exposure Induces Ocular ER Stress in ARPE-19 Cells To increase the delivery efficacy of PM within cells, the PM was prepared with $0.01\%$ Tween-80 to prevent internal aggregation, and also increase the cellular absorption rate by maintaining a smaller size. Park et al. have reported that PM size is smaller when Tween-80 is used as a vehicle, compared to PM alone [50]. Therefore, Tween-80 as a vehicle control could serve as an efficient means of exacerbating cellular toxicity, by facilitating PM delivery. In our experimental setting, treatment with $0.01\%$ Tween-80 alone did not affect the viability of ARPE-19 cells at any of the time points monitored (4 and 8 h), as this was similar to the untreated control (Figure 4). However, when 2 mg/mL of PM was delivered together with $0.01\%$ Tween-80, retinal cell viability was significantly decreased to $87\%$ and $83\%$ after 4 and 8 h, respectively (Figure 4). Endoplasmic reticulum stress is a strong cellular pathogenic trigger of various metabolic complications. Neurodegenerative disorders such as Alzheimer’s or Parkinson’s disease, diabetes, atherosclerosis, liver disease, cancer [51], and infectious diseases caused by bacteria or viruses [52] have been associated with prolonged ER stress. Exacerbation of ER stress is also a key pathological signature of diverse ocular diseases such as DR [53,54], glaucoma [53], and AMD [53,55]. Therefore, modulation of ocular ER stress is crucial for mitigating or preventing ocular diseases. To investigate whether PM causes ocular ER stress in ARPE-19 cells, we examined the expression levels of the ER stress-related mRNAs; binding of immunoglobulin protein (BiP), CCAAT/enhancer-binding protein homologous protein (CHOP), and X-box binding protein 1 (XBP-1), by qRT-PCR. As seen in Figure 5, in line with our hypothesis, PM treatment significantly increased expression of ER stress-related mRNAs compared to our vehicle control ($0.01\%$ Tween-80). Expression of BiP significantly increased after 4 h of PM treatment, while CHOP mRNA expression peaked earlier, at 2 h following ocular PM exposure, in a similar manner to XBP-1, whose expression was also markedly elevated at ≥2 h post-PM treatment (Figure 5). Calcium is essential in maintaining the integrity of protein folding and posttranslational modification mechanisms [56], and the ER is a major site of intracellular calcium storage. Moreover, prolonged ocular ER stress is closely correlated with increased [Ca2+]i levels and the subsequent induction of hypoxic adaptation responses, including the hypoxia-related UPR. An imbalance in [Ca2+]i levels between cytosol and the ER in the eye may induce significant ocular ER stress. Pharmacological induction of ER stress by treatment with the non-competitive sarco/endoplasmic reticulum inhibitor of the Ca²⁺ ATPase, thapsigargin, has been shown to increase [Ca2+]i levels and expression of hypoxia-associated mRNAs such as vascular endothelial growth factor (VEGF), in ARPE-19 cells [49]. Since we observed that PM exposure remarkably elevates expression of ER stress-related transcripts (Figure 5), we posited whether PM may also promote hypoxia and activate hypoxic adaptation responses in retinal cells. In our experimental setting, ocular PM exposure decreased the zonula occludens-1 (ZO-1) mRNA expression in a time-dependent manner (Figure 6A). This observed reduction in ZO-1 mRNA levels may be related to retinal inflammation (Figure 2 and Figure 3), hypoxia (Figure 6B,C), and UPRs (Figure 5 and Figure 7) following PM exposure. As per our expectations, expression of hypoxia-related mRNAs, VEGFα and ankyrin repeat domain 37 (ANKRD37) was remarkably elevated following ocular PM exposure (Figure 6B,C). Both VEGFα and ANKRD37 are the direct target genes of the hypoxia-inducible factor 1 (HIF-1), therefore, their upregulation following PM exposure indicates that PM likely promotes hypoxic responses in retinal cells. Furthermore, induction of the ocular hypoxic response may also be closely related to the disruption of epithelial tight junction integrity [57]. Ocular epithelial tight junctions have been shown to be affected by various stress conditions, including increased inflammation [58], hypoxia [57], and unfolded protein responses [59]. PM treatment elevated cytosolic [Ca2+]i levels and lead to the upregulation of hypoxia-related UPRs (Figure 7). Exposure to PM leads to a significant increase in the protein levels of protein kinase R-like endoplasmic reticulum kinase (PERK), eukaryotic translation initiation factor 2 alpha (eIF2a), activating transcription factor 4 (ATF4), and CHOP, as well as of inositol-requiring enzyme 1 alpha (IRE1α), XBP-1, and the ER chaperone BiP (Figure 7A–H). Compared to the vehicle control, [Ca2+]i levels in ARPE-19 cells were remarkably elevated after exposure to PM doses between 50–1000 μg/mL (Figure 7I). Interestingly, [Ca2+]i levels closely mirrored the gradual increase in PM concentration in a dose-dependent manner up to 250 μg/mL [PM], but exhibited a rapid decrease at PM doses ≥500 μg/mL (Figure 7I). This trend in the fluctuation of cytosolic [Ca2+]i may be related to retinal apoptosis, however, further mechanistic studies are required. ## 4. Discussion The pathological progression of retinal diseases is closely intertwined with significant risk factors such as inflammation and ER stress, which may ultimately threaten vision. Increased ocular inflammation and ER stress have been associated with DR [41,54], and AMD [45]. Progression of ocular diseases cannot be easily identified or predicted in patients, thereby prevention through avoidance of pathological toxicants is highly recommended. Owing to industrial development, carbon-containing fuels (i.e., fossil fuels) are widely used, which emit significant amounts of by-products after combustion, including PM. According to current literature, PM can be a potent inducer of inflammatory [15,35,36] and ER stress responses [46,47,48] in tissues. In this study, we investigated whether ocular PM exposure induces inflammatory responses and ER stress in retinal cells. Our findings demonstrated that treatment with PM induced ocular inflammatory responses in ARPE-19 cells by activation of the MAPK/NFκB axis, with a concomitant upregulation in the expression of inflammatory mRNAs. In addition, ocular PM exposure increased [Ca2+]i levels and promoted hypoxia-related UPRs. The retina plays an important role in vision because it converts incoming light into electrical signals. We employed the ARPE-19 cell line for our experiments since RPE cells have a primary immune defense role by locating the outermost layer of the retina and forming a blood-retinal barrier, thereby contributing to ocular integrity through the maintenance of tight junctions [60]. The RPE also possesses a phagocytic function that eliminates the photoreceptor outer segments, rendering the light-sensing region sensitive to light exposure [60]. Furthermore, the RPE also transports numerous nutrients and eliminates metabolic by-products in the eye [61]. Therefore, abnormalities in RPE cells may cause retinal dysfunction and ultimately lead to the development of vision-related disorders [62]. The TNFα, IL-1β, and IL-6 cytokines are expressed as part of the initial response to inflammation [63]. Moreover, MCP-1 is also a chemokine involved in the inflammatory cascade that regulates the recruitment of monocytes/macrophages to local sites of inflammation [64]. In the present study, ocular PM exposure led to a significant increase in inflammatory responses in ARPE-19 cells, as it induced expression of the cytokine mRNAs TNFα, IL-1β, IL-6, and MCP-1 (Figure 3). The activation of NFκB-signaling promotes pro-inflammatory cytokine and chemokine production [65,66]. Along with NFκB, MAPK-signaling is mainly activated by cellular stress as well as inflammatory stimuli; therefore, the MAPK/NFκB axis plays an integral role in the overall inflammatory responses, such as antigen presentation, leukocyte infiltration, and cytokine production [67,68,69]. In our work, we observed that ocular PM exposure increased the phosphorylation of MAPKs (p38, JNK, and ERK) and subsequently elevated phosphorylation levels of IκBα and NFκB in a PM dose-dependent manner (Figure 2). Our findings clearly demonstrated that ocular PM exposure increased ocular inflammation in ARPE-19 cells by activating MAPK/NFκB signalling and cytokine mRNA expression. In addition, ocular PM exposure elevated not only inflammatory responses but also ER stress in ARPE-19 cells. Along with its role in protein and lipid synthesis, the ER is also a major quality control center for newly synthesized peptides, as it monitors and modifies folding accordingly or discards misfolded proteins. Moreover, the ER regulates [Ca2+]i levels. The ER stress response is triggered by various pathological stimuli such as calcium depletion, hypoxia, oxidative stress, or the production of non-functional proteins, resulting in the accumulation of misfolded or unfolded proteins in the ER lumen [70,71]. Under ER stress conditions, cells activate UPR as a protective signal to maintain homeostasis by three distinct mechanisms: [1] decrease of new protein synthesis by halting protein entry into the ER, [2] increase in protein folding activity, and [3] clearance of misfolded proteins by secretion into the cytosol for ubiquitination and subsequent degradation. However, prolonged ER stress may result in cell death [72,73]. The UPR is initiated with the dissociation of the ER chaperone BiP from three ER membrane sensors (PERK, IRE1α, and ATF6). Activated PERK phosphorylates eIF2α, which halts overall protein translation. Next, phosphorylated eIF2α selectively upregulates ATF4, which translocates to the nucleus and activates proapoptotic CHOP in the nucleus, among other UPR targets. Activated IRE1α splices XBP-1 and controls UPR target genes such as chaperones and ER-associated degradation (ERAD). The IRE1α also binds tumor necrosis factor receptor-associated factor 2 (TRAF2), leading to the recruitment of the IκB kinase (IKK), which phosphorylates IκBa and activates NFκB that subsequently promotes cytokine production [74]. The IRE1α-TRAF2 complex may also recruit apoptosis signal-regulating kinase 1 (ASK1), which upregulates JNK and p38 [75]. Activated ATF6 translocates to the Golgi where it is cleaved and then enters the nucleus to activate UPRs. In the present study, we observed the induction of two PM-inducible UPR pathways: [1] the PERK-eIF2a-ATF4-CHOP axis, and [2] the IRE1α-XBP-1 axis, with concomitant activation of BiP as well (Figure 7A–H). These data are consistent with the upregulation observed in the mRNA expression for BiP, CHOP, and XBP-1 (Figure 5). Moreover, ocular PM exposure significantly elevated [Ca2+]i levels (Figure 7I), which may trigger hypoxia and angiogenesis. In our experimental setting, the hypoxic markers ANKRD37 and VEGFα were increased significantly following retinal PM exposure (Figure 6B,C). ## 5. Conclusions In conclusion, our current work demonstrated that ocular PM exposure led to ocular inflammation and ER stress in ARPE-19 cells, which in turn promoted ER stress-related responses, such as UPR. Our findings could provide useful insight into clinical and non-clinical investigations that focus on the mechanisms by which ocular PM exposure may trigger inflammatory responses in the retina, and can lend further support in linking PM with the pathophysiology of the visual system. Taken together, our findings corroborated our initial hypothesis that ocular PM exposure induces ER stress and subsequent hypoxic adaptation responses, which also include upregulation of UPR-related pathway components. ## References 1. Zandalinas S.I., Fritschi F.B., Mittler R.. Global Warming, Climate Change, and Environmental Pollution: Recipe for a Multifactorial Stress Combination Disaster. Trends Plant Sci. (2021) 26 588-599. DOI: 10.1016/j.tplants.2021.02.011 2. Jung S.H., Bae C.H., Kim J.H., Park S.D., Shim J.J., Lee J.L.. Lactobacillus casei HY2782 and Pueraria lobata Root Extract Complex Ameliorates Particulate Matter-Induced Airway Inflammation in Mice by Inhibiting Th2 and Th17 Immune Responses. Prev. Nutr. Food Sci. (2022) 27 188-197. DOI: 10.3746/pnf.2022.27.2.188 3. Zyrek D., Krzeminska A., Zyrek N., Wajda A., Pabian W., Pacholski M., Sokolski M., Zymlinski R.. Effects of exposure to air pollution on acute cardiovascular and respiratory admissions to the hospital and early mortality at emergency department. Adv. Clin. Exp. Med. (2022) 31 1129-1138. DOI: 10.17219/acem/149400 4. Kaihara T., Yoneyama K., Nakai M., Higuma T., Sumita Y., Miyamoto Y., Watanabe M., Izumo M., Ishibashi Y., Tanabe Y.. Association of PM(2.5) exposure with hospitalization for cardiovascular disease in elderly individuals in Japan. Sci. Rep. (2021) 11 9897. DOI: 10.1038/s41598-021-89290-5 5. Loomis D., Grosse Y., Lauby-Secretan B., El Ghissassi F., Bouvard V., Benbrahim-Tallaa L., Guha N., Baan R., Mattock H., Straif K.. The carcinogenicity of outdoor air pollution. Lancet Oncol. (2013) 14 1262-1263. DOI: 10.1016/S1470-2045(13)70487-X 6. Adams K., Greenbaum D.S., Shaikh R., van Erp A.M., Russell A.G.. Particulate matter components, sources, and health: Systematic approaches to testing effects. J. Air Waste Manag. Assoc. (2015) 65 544-558. DOI: 10.1080/10962247.2014.1001884 7. 7. World Health Organization WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon MonoxideWorld Health OrganizationGeneva, Switzerland2021. WHO Global Air Quality Guidelines: Particulate Matter (PM2.5 and PM10), Ozone, Nitrogen Dioxide, Sulfur Dioxide and Carbon Monoxide (2021) 8. Schraufnagel D.E.. The health effects of ultrafine particles. Exp. Mol. Med. (2020) 52 311-317. DOI: 10.1038/s12276-020-0403-3 9. Kyung S.Y., Jeong S.H.. Particulate-Matter Related Respiratory Diseases. Tuberc. Respir. Dis. (2020) 83 116-121. DOI: 10.4046/trd.2019.0025 10. Fiordelisi A., Piscitelli P., Trimarco B., Coscioni E., Iaccarino G., Sorriento D.. The mechanisms of air pollution and particulate matter in cardiovascular diseases. Heart Fail. Rev. (2017) 22 337-347. DOI: 10.1007/s10741-017-9606-7 11. Nemmar A., Al-Salam S., Zia S., Yasin J., Al Husseni I., Ali B.H.. Diesel exhaust particles in the lung aggravate experimental acute renal failure. Toxicol. Sci. (2010) 113 267-277. DOI: 10.1093/toxsci/kfp222 12. Morsi A.A., Fouad H., Alasmari W.A., Faruk E.M.. The biomechanistic aspects of renal cortical injury induced by diesel exhaust particles in rats and the renoprotective contribution of quercetin pretreatment: Histological and biochemical study. Environ. Toxicol. (2022) 37 310-321. DOI: 10.1002/tox.23399 13. Ito Y., Yanagiba Y., Ramdhan D.H., Hayashi Y., Li Y., Suzuki A.K., Kamijima M., Nakajima T.. Nanoparticle-rich diesel exhaust-induced liver damage via inhibited transactivation of peroxisome proliferator-activated receptor alpha. Environ. Toxicol. (2016) 31 1985-1995. DOI: 10.1002/tox.22199 14. Kim B.E., Kim J., Goleva E., Berdyshev E., Lee J., Vang K.A., Lee U.H., Han S., Leung S., Hall C.F.. Particulate matter causes skin barrier dysfunction. JCI Insight (2021) 6 e145185. DOI: 10.1172/jci.insight.145185 15. Kang Y.J., Tan H.Y., Lee C.Y., Cho H.. An Air Particulate Pollutant Induces Neuroinflammation and Neurodegeneration in Human Brain Models. Adv. Sci. (2021) 8 e2101251. DOI: 10.1002/advs.202101251 16. Kim S., Park H., Park H., Joung B., Kim E.. The acute respiratory exposure by intratracheal instillation of Sprague-Dawley rats with diesel particulate matter induces retinal thickening. Cutan. Ocul. Toxicol. (2016) 35 257-259. DOI: 10.3109/15569527.2015.1104329 17. Tan G., Li J., Yang Q., Wu A., Qu D.Y., Wang Y., Ye L., Bao J., Shao Y.. Air pollutant particulate matter 2.5 induces dry eye syndrome in mice. Sci. Rep. (2018) 8 17828. DOI: 10.1038/s41598-018-36181-x 18. Tang Y.J., Chang H.H., Chiang C.Y., Lai C.Y., Hsu M.Y., Wang K.R., Han H.H., Chen L.Y., Lin D.P.. A Murine Model of Acute Allergic Conjunctivitis Induced by Continuous Exposure to Particulate Matter 2.5. Investig. Ophthalmol. Vis. Sci. (2019) 60 2118-2126. DOI: 10.1167/iovs.18-26214 19. Klopfer J.. Effects of environmental air pollution on the eye. J. Am. Optom. Assoc. (1989) 60 773-778. PMID: 2685084 20. Jung S.J., Mehta J.S., Tong L.. Effects of environment pollution on the ocular surface. Ocul. Surf. (2018) 16 198-205. DOI: 10.1016/j.jtos.2018.03.001 21. Yang Q., Li K., Li D., Zhang Y., Liu X., Wu K.. Effects of fine particulate matter on the ocular surface: An in vitro and in vivo study. Biomed. Pharmacother. (2019) 117 109177. DOI: 10.1016/j.biopha.2019.109177 22. Choi S.Y., Eom Y., Song J.S., Kim H.M.. Fine dust and eye health. J. Korean Med. Assoc. (2019) 62 486-494. DOI: 10.5124/jkma.2019.62.9.486 23. Ju M.J., Kim J., Park S.K., Kim D.H., Choi Y.H.. Long-term exposure to ambient air pollutants and age-related macular degeneration in middle-aged and older adults. Environ. Res. (2022) 204 111953. DOI: 10.1016/j.envres.2021.111953 24. Chua S.Y.L., Warwick A., Peto T., Balaskas K., Moore A.T., Reisman C., Desai P., Lotery A.J., Dhillon B., Khaw P.T.. Association of ambient air pollution with age-related macular degeneration and retinal thickness in UK Biobank. Br. J. Ophthalmol. (2022) 106 705-711. DOI: 10.1136/bjophthalmol-2020-316218 25. Grant A., Leung G., Aubin M.J., Kergoat M.J., Li G., Freeman E.E.. Fine Particulate Matter and Age-Related Eye Disease: The Canadian Longitudinal Study on Aging. Investig. Ophthalmol. Vis. Sci. (2021) 62 7. DOI: 10.1167/iovs.62.10.7 26. Liang C.L., Wang C.M., Jung C.R., Chang Y.C., Lin C.J., Lin Y.T., Hwang B.F.. Fine particulate matter measured by satellites predicts the risk of age-related macular degeneration in a longitudinal cohort study. Environ. Sci. Pollut. Res. Int. (2022) 29 51942-51950. DOI: 10.1007/s11356-022-19278-9 27. Pan S.C., Huang C.C., Chin W.S., Chen B.Y., Chan C.C., Guo Y.L.. Association between air pollution exposure and diabetic retinopathy among diabetics. Environ. Res. (2020) 181 108960. DOI: 10.1016/j.envres.2019.108960 28. Shan A., Chen X., Yang X., Yao B., Liang F., Yang Z., Liu F., Chen S., Yan X., Huang J.. Association between long-term exposure to fine particulate matter and diabetic retinopathy among diabetic patients: A national cross-sectional study in China. Environ. Int. (2021) 154 106568. DOI: 10.1016/j.envint.2021.106568 29. Chua S.Y.L., Khawaja A.P., Dick A.D., Morgan J., Dhillon B., Lotery A.J., Strouthidis N.G., Reisman C., Peto T., Khaw P.T.. Ambient Air Pollution Associations with Retinal Morphology in the UK Biobank. Investig. Ophthalmol. Vis. Sci. (2020) 61 32. DOI: 10.1167/iovs.61.5.32 30. Louwies T., Panis L.I., Kicinski M., De Boever P., Nawrot T.S.. Retinal microvascular responses to short-term changes in particulate air pollution in healthy adults. Environ. Health Perspect. (2013) 121 1011-1016. DOI: 10.1289/ehp.1205721 31. Gu Y., Hao S., Liu K., Gao M., Lu B., Sheng F., Zhang L., Xu Y., Wu D., Han Y.. Airborne fine particulate matter (PM(2.5)) damages the inner blood-retinal barrier by inducing inflammation and ferroptosis in retinal vascular endothelial cells. Sci. Total. Environ. (2022) 838 156563. DOI: 10.1016/j.scitotenv.2022.156563 32. Kang W.S., Choi H., Jang G., Lee K.H., Kim E., Kim K.J., Jeong G.Y., Kim J.S., Na C.S., Kim S.. Long-Term Exposure to Urban Particulate Matter on the Ocular Surface and the Incidence of Deleterious Changes in the Cornea, Conjunctiva and Retina in Rats. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21144976 33. Choi J.Y., Bae J.E., Kim J.B., Jo D.S., Park N.Y., Kim Y.H., Lee H.J., Kim S.H., Kim S.H., Jeon H.B.. 2-IPMA Ameliorates PM2.5-Induced Inflammation by Promoting Primary Ciliogenesis in RPE Cells. Molecules (2021) 26. DOI: 10.3390/molecules26175409 34. Lee H., Hwang-Bo H., Ji S.Y., Kim M.Y., Kim S.Y., Park C., Hong S.H., Kim G.Y., Song K.S., Hyun J.W.. Diesel particulate matter2.5 promotes epithelial-mesenchymal transition of human retinal pigment epithelial cells via generation of reactive oxygen species. Environ. Pollut. (2020) 262 114301. DOI: 10.1016/j.envpol.2020.114301 35. Jeong S., Lee J.H., Ha J.H., Kim J., Kim I., Bae S.. An Exploratory Study of the Relationships Between Diesel Engine Exhaust Particle Inhalation, Pulmonary Inflammation and Anxious Behavior. Int. J. Environ. Res. Public. Health (2021) 18. DOI: 10.3390/ijerph18031166 36. Yuan C.S., Lai C.S., Tseng Y.L., Hsu P.C., Lin C.M., Cheng F.J.. Repeated exposure to fine particulate matter constituents lead to liver inflammation and proliferative response in mice. Ecotoxicol. Environ. Saf. (2021) 224 112636. DOI: 10.1016/j.ecoenv.2021.112636 37. Lee H.S., Han S., Seo J.W., Jeon K.J.. Exposure to Traffic-Related Particulate Matter 2.5 Triggers Th2-Dominant Ocular Immune Response in a Murine Model. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17082965 38. Mu N., Wang H., Chen D., Wang F., Ji L., Zhang C., Li M., Lu P.. A Novel Rat Model of Dry Eye Induced by Aerosol Exposure of Particulate Matter. Investig. Ophthalmol. Vis. Sci. (2022) 63 39. DOI: 10.1167/iovs.63.1.39 39. Wei C.C., Lin H.J., Lim Y.P., Chen C.S., Chang C.Y., Lin C.J., Chen J.J., Tien P.T., Lin C.L., Wan L.. PM2.5 and NOx exposure promote myopia: Clinical evidence and experimental proof. Environ. Pollut. (2019) 254 113031. DOI: 10.1016/j.envpol.2019.113031 40. Li L., Song M., Zhou J., Sun X., Lei Y.. Ambient particulate matter exposure causes visual dysfunction and retinal neuronal degeneration. Ecotoxicol. Environ. Saf. (2022) 247 114231. DOI: 10.1016/j.ecoenv.2022.114231 41. Zhang S.X., Sanders E., Wang J.J.. Endoplasmic reticulum stress and inflammation: Mechanisms and implications in diabetic retinopathy. J. Ocul. Biol. Dis. Inform. (2011) 4 51-61. DOI: 10.1007/s12177-011-9075-5 42. Whitcup S.M., Nussenblatt R.B., Lightman S.L., Hollander D.A.. Inflammation in retinal disease. Int. J. Inflam. (2013) 2013 724648. DOI: 10.1155/2013/724648 43. McLaughlin T., Medina A., Perkins J., Yera M., Wang J.J., Zhang S.X.. Cellular stress signaling and the unfolded protein response in retinal degeneration: Mechanisms and therapeutic implications. Mol. Neurodegener. (2022) 17 25. DOI: 10.1186/s13024-022-00528-w 44. Kaur G., Singh N.K.. Inflammation and retinal degenerative diseases. Neural Regen. Res. (2023) 18 513-518. DOI: 10.4103/1673-5374.350192 45. Kheitan S., Minuchehr Z., Soheili Z.S.. Exploring the cross talk between ER stress and inflammation in age-related macular degeneration. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0181667 46. Laing S., Wang G., Briazova T., Zhang C., Wang A., Zheng Z., Gow A., Chen A.F., Rajagopalan S., Chen L.C.. Airborne particulate matter selectively activates endoplasmic reticulum stress response in the lung and liver tissues. Am. J. Physiol. Cell. Physiol. (2010) 299 C736-C749. DOI: 10.1152/ajpcell.00529.2009 47. Mendez R., Zheng Z., Fan Z., Rajagopalan S., Sun Q., Zhang K.. Exposure to fine airborne particulate matter induces macrophage infiltration, unfolded protein response, and lipid deposition in white adipose tissue. Am. J. Transl. Res. (2013) 5 224-234. PMID: 23573366 48. Hsu Y.H., Chuang H.C., Lee Y.H., Lin Y.F., Chen Y.J., Hsiao T.C., Wu M.Y., Chiu H.W.. Traffic-related particulate matter exposure induces nephrotoxicity in vitro and in vivo. Free. Radic. Biol. Med. (2019) 135 235-244. DOI: 10.1016/j.freeradbiomed.2019.03.008 49. Ha J.H., Shil P.K., Zhu P., Gu L., Li Q., Chung S.. Ocular inflammation and endoplasmic reticulum stress are attenuated by supplementation with grape polyphenols in human retinal pigmented epithelium cells and in C57BL/6 mice. J. Nutr. (2014) 144 799-806. DOI: 10.3945/jn.113.186957 50. Park S.Y., An K.S., Lee B., Kang J.H., Jung H.J., Kim M.W., Ryu H.Y., Shim K.S., Nam K.T., Yoon Y.S.. Establishment of particulate matter-induced lung injury model in mouse. Lab. Anim. Res. (2021) 37 20. DOI: 10.1186/s42826-021-00097-x 51. Ozcan L., Tabas I.. Role of endoplasmic reticulum stress in metabolic disease and other disorders. Annu. Rev. Med. (2012) 63 317-328. DOI: 10.1146/annurev-med-043010-144749 52. Choi J.A., Song C.H.. Insights Into the Role of Endoplasmic Reticulum Stress in Infectious Diseases. Front. Immunol. (2019) 10 3147. DOI: 10.3389/fimmu.2019.03147 53. Kroeger H., Chiang W.C., Felden J., Nguyen A., Lin J.H.. ER stress and unfolded protein response in ocular health and disease. FEBS J. (2019) 286 399-412. DOI: 10.1111/febs.14522 54. Li J., Wang J.J., Yu Q., Wang M., Zhang S.X.. Endoplasmic reticulum stress is implicated in retinal inflammation and diabetic retinopathy. FEBS Lett. (2009) 583 1521-1527. DOI: 10.1016/j.febslet.2009.04.007 55. Salminen A., Kauppinen A., Hyttinen J.M., Toropainen E., Kaarniranta K.. Endoplasmic reticulum stress in age-related macular degeneration: Trigger for neovascularization. Mol. Med. (2010) 16 535-542. DOI: 10.2119/molmed.2010.00070 56. Michalak M., Robert Parker J.M., Opas M.. Ca2+ signaling and calcium binding chaperones of the endoplasmic reticulum. Cell Calcium (2002) 32 269-278. DOI: 10.1016/S0143416002001884 57. Yanai R., Ko J.A., Morishige N., Chikama T., Ichijima H., Nishida T.. Disruption of zonula occludens-1 localization in the rabbit corneal epithelium by contact lens-induced hypoxia. Investig. Ophthalmol. Vis. Sci. (2009) 50 4605-4610. DOI: 10.1167/iovs.09-3407 58. Tien P.T., Lin H.J., Tsai Y.Y., Lim Y.P., Chen C.S., Chang C.Y., Lin C.J., Chen J.J., Wu S.M., Huang Y.J.. Perfluorooctanoic acid in indoor particulate matter triggers oxidative stress and inflammation in corneal and retinal cells. Sci. Rep. (2020) 10 15702. DOI: 10.1038/s41598-020-72600-8 59. Ma J.H., Wang J.J., Li J., Pfeffer B.A., Zhong Y., Zhang S.X.. The Role of IRE-XBP1 Pathway in Regulation of Retinal Pigment Epithelium Tight Junctions. Investig. Ophthalmol. Vis. Sci. (2016) 57 5244-5252. DOI: 10.1167/iovs.16-19232 60. Naylor A., Hopkins A., Hudson N., Campbell M.. Tight Junctions of the Outer Blood Retina Barrier. Int. J. Mol. Sci. (2019) 21. DOI: 10.3390/ijms21010211 61. Simo R., Villarroel M., Corraliza L., Hernandez C., Garcia-Ramirez M.. The retinal pigment epithelium: Something more than a constituent of the blood-retinal barrier--implications for the pathogenesis of diabetic retinopathy. J. Biomed. Biotechnol. (2010) 2010 190724. DOI: 10.1155/2010/190724 62. Yang S., Zhou J., Li D.. Functions and Diseases of the Retinal Pigment Epithelium. Front. Pharmacol. (2021) 12 727870. DOI: 10.3389/fphar.2021.727870 63. Ferraccioli G., Bracci-Laudiero L., Alivernini S., Gremese E., Tolusso B., De Benedetti F.. Interleukin-1beta and interleukin-6 in arthritis animal models: Roles in the early phase of transition from acute to chronic inflammation and relevance for human rheumatoid arthritis. Mol. Med. (2010) 16 552-557. DOI: 10.2119/molmed.2010.00067 64. Deshmane S.L., Kremlev S., Amini S., Sawaya B.E.. Monocyte chemoattractant protein-1 (MCP-1): An overview. J. Interferon Cytokine Res. (2009) 29 313-326. DOI: 10.1089/jir.2008.0027 65. Liu T., Zhang L., Joo D., Sun S.C.. NF-kappaB signaling in inflammation. Signal Transduct. Target. Ther. (2017) 2 17023. DOI: 10.1038/sigtrans.2017.23 66. Baker R.G., Hayden M.S., Ghosh S.. NF-kappaB, inflammation, and metabolic disease. Cell Metab. (2011) 13 11-22. DOI: 10.1016/j.cmet.2010.12.008 67. Kyriakis J.M., Avruch J.. Mammalian MAPK signal transduction pathways activated by stress and inflammation: A 10-year update. Physiol. Rev. (2012) 92 689-737. DOI: 10.1152/physrev.00028.2011 68. Newton K., Dixit V.M.. Signaling in innate immunity and inflammation. Cold Spring Harb. Perspect. Biol. (2012) 4 a006049. DOI: 10.1101/cshperspect.a006049 69. Omage S.O., Orhue N.E.J., Omage K.. Dennettia tripetala Combats Oxidative Stress, Protein and Lipid Dyshomeostasis, Inflammation, Hepatic Injury, and Glomerular Blockage in Rats. Prev. Nutr. Food Sci. (2021) 26 177-185. DOI: 10.3746/pnf.2021.26.2.177 70. Hetz C., Papa F.R.. The Unfolded Protein Response and Cell Fate Control. Mol. Cell (2018) 69 169-181. DOI: 10.1016/j.molcel.2017.06.017 71. Fujii J., Homma T., Kobayashi S., Seo H.G.. Mutual interaction between oxidative stress and endoplasmic reticulum stress in the pathogenesis of diseases specifically focusing on non-alcoholic fatty liver disease. World J. Biol. Chem. (2018) 9 1-15. DOI: 10.4331/wjbc.v9.i1.1 72. Kadowaki H., Nishitoh H.. Signaling pathways from the endoplasmic reticulum and their roles in disease. Genes (2013) 4 306-333. DOI: 10.3390/genes4030306 73. Siangcham T., Prathaphan P., Ruangtong J., Thongsepee N., Martviset P., Chantree P., Sornchuer P., Sangpairoj K.. Camboginol and Morelloflavone from Garcinia dulcis (Roxb.) Kurz Flower Extract Promote Autophagic Cell Death against Human Glioblastoma Cells through Endoplasmic Reticulum Stress. Prev. Nutr. Food Sci. (2022) 27 376-383. DOI: 10.3746/pnf.2022.27.4.376 74. Zhang K., Kaufman R.J.. From endoplasmic-reticulum stress to the inflammatory response. Nature (2008) 454 455-462. DOI: 10.1038/nature07203 75. Sisinni L., Pietrafesa M., Lepore S., Maddalena F., Condelli V., Esposito F., Landriscina M.. Endoplasmic Reticulum Stress and Unfolded Protein Response in Breast Cancer: The Balance between Apoptosis and Autophagy and Its Role in Drug Resistance. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20040857
--- title: Respiratory Health Effects of In Vivo Sub-Chronic Diesel and Biodiesel Exhaust Exposure authors: - Katherine R. Landwehr - Ryan Mead-Hunter - Rebecca A. O’Leary - Anthony Kicic - Benjamin J. Mullins - Alexander N. Larcombe journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049281 doi: 10.3390/ijms24065130 license: CC BY 4.0 --- # Respiratory Health Effects of In Vivo Sub-Chronic Diesel and Biodiesel Exhaust Exposure ## Abstract Biodiesel, which can be made from a variety of natural oils, is currently promoted as a sustainable, healthier replacement for commercial mineral diesel despite little experimental data supporting this. The aim of our research was to investigate the health impacts of exposure to exhaust generated by the combustion of diesel and two different biodiesels. Male BALB/c mice ($$n = 24$$ per group) were exposed for 2 h/day for 8 days to diluted exhaust from a diesel engine running on ultra-low sulfur diesel (ULSD) or Tallow or Canola biodiesel, with room air exposures used as control. A variety of respiratory-related end-point measurements were assessed, including lung function, responsiveness to methacholine, airway inflammation and cytokine response, and airway morphometry. Exposure to Tallow biodiesel exhaust resulted in the most significant health impacts compared to Air controls, including increased airway hyperresponsiveness and airway inflammation. In contrast, exposure to Canola biodiesel exhaust resulted in fewer negative health effects. Exposure to ULSD resulted in health impacts between those of the two biodiesels. The health effects of biodiesel exhaust exposure vary depending on the feedstock used to make the fuel. ## 1. Introduction Diesel exhaust exposure is known to lead to negative health impacts on multiple organ systems including, but not limited to, the respiratory, cardiovascular, nervous, endocrine, and urinary systems [1,2]. It has been implicated in lung [3,4], brain [5], and bladder cancer [6], increased blood pressure [7], altered neurological activity [8], increased thrombotic risk [9], increased stroke risk [10], increased risk of type 2 diabetes [11,12,13], and asthma [14]. Biodiesel exhaust shares many of the same physicochemical characteristics as diesel exhaust, such as oxides of nitrogen (NOx), carbon monoxide and dioxide, particulate matter consisting of mostly ultrafine particles [15,16,17], polycyclic aromatic hydrocarbons (PAHs), aldehydes, ketones, and heavy metals [18,19,20], and, thus, it is suspected that it will be associated with many of the same negative health impacts. Previous studies into engine and exhaust characteristics that compared diesel and biodiesel often show that biodiesel exhaust contains more NOx, PAHs, and ultrafine particles (<100 nm) but less overall particulate matter by weight [18,20,21,22,23,24] compared with mineral diesel exhaust. This is of concern as ultrafine particles, when compared to larger sizes, are more commonly linked to the negative health effects of air pollution [25,26]. Up to $90\%$ of diesel exhaust particles by number consist of nucleation mode particles, newly formed from combustion and chemical reactions and are under 30 nm in size [15,16]. Thus, a further increase in the proportion of ultrafine particles in biodiesel exhaust is of great concern, in part due to increased surface-area-to-volume ratios allowing for more dangerous chemicals to be ab/adsorbed onto them for a given particle mass [22]. Despite this, biodiesel fuel usage is increasing worldwide [27]. Previous studies, both in vitro and in vivo, looking into the health effects of biodiesel exhaust typically use less than optimal exposure models [28,29]. Many previous studies have focused solely on the health effects of the exhaust particles by collecting them on a filter and adding a set concentration directly to a flask of cells/bacteria or instilling it into the nose/trachea of rats/mice [28,30,31,32,33]. While it allows an accurate dosage of particles to be given [33], in return, it both ignores the effects of the exhaust gases, which have their own set of negative health impacts [34], and also removes the ultrafine particles, arguably one of the most toxic components of diesel exhaust, which readily agglomerate on filters to form larger-sized particles [35]. This generates an artificial particle size spectrum, with previous studies showing that over a 16-fold increase in particle concentration is needed to generate the similar health impacts as if whole exhaust was used [36]. In addition to this, the majority of previous biodiesel exhaust toxicology studies have used bacterial AMES tests to study mutagenic effects, or have exposed immortalized cell lines [23,24,37,38] that are not always human or even derived from the respiratory system, the first exposed and likely most affected tissue [37,39]. Previous studies that expose animals most often use instillation to expose the mice/rats to the particulate matter in solution, with few studies performing inhalation exposures. These few studies are often divided into several different publications, likely due to the difficulty in conducting them, which artificially inflates the actual number of inhalation studies performed [40,41,42,43,44,45,46,47,48,49,50]. Of the few studies that do use inhalation exposures, only half expose mice/rats to pure biodiesel exhausts and the remainder use biodiesel blended with diesel (at ratios of <$30\%$ biodiesel in diesel fuel). With the biodiesel concentration being below half of the total fuel content, there is a chance for biodiesel-exhaust-exposure-induced health effects to be masked by those of diesel. That said, blended fuels are highly relevant to what is being used today with biodiesel already being blended up to $20\%$ in some countries [51,52,53]. There is also a tendency in the literature to treat all biodiesels as the same, despite evidence that the feedstock used to make the biodiesel vastly affects the fuel and exhaust properties and, thus, the resulting health impacts of exhaust exposure [54]. Studies often use just one biodiesel type and make claims about biodiesel in general based on the results of that type [55]. Some studies do not even disclose the feedstock used to make their biodiesel [55,56]. Methodological differences inherent in different study designs—from the exact engine type, the use (or not) of exhaust after-treatment systems, differing exhaust concentrations, the use of whole exhaust compared to filter-extracted particles, and the wide range of health effects measured, including mutagenicity, cytotoxicity, and immunological effects [23,24,29,40,41,46,48,50,57]—make comparisons of fuel feedstocks between different studies difficult. Thus, the aim of this study was to compare the respiratory health effects of exposure to one of two different types of biodiesel exhaust, using air and ultra-low sulfur mineral diesel (ULSD) as controls, to evaluate the impacts of biodiesel exhaust exposure and how these impacts can change between different feedstock types. Tallow and Canola were chosen for study, both because they are commonly used today [58,59] and previous research found them to be at extreme ends of the health effects in both a submerged and air–liquid-interface cell-culture exposure model [54,60]. Mice were exposed for two hours each day to one of these four options for eight consecutive days. The main hypotheses were that (i) exposure to Tallow biodiesel exhaust would result in more severe and a greater range of negative health effects than ULSD exhaust exposure and (ii) that exposure to Canola biodiesel exhaust would result in less health impacts. ## 2.1. Exhaust Gas Characteristics The mean and standard deviation for each fuel and gas type are shown in Table 1, with the exception of CO that showed only the highest reading at 10 min for each of the repeated exposures due to the cold-start effect on the performance of the catalytic converter. Trends over time can be found in the Supplementary Materials (Figure S1). All fuels displayed similar trends over time with NOx (NO and NO2), CO2, and SO2 increasing rapidly in the first 30 min of the exposure, O2 decreasing rapidly in the first 20 min, and CO peaking in the first 10 min before decreasing rapidly to undetectable concentrations. Canola was found to be the most different of the tested fuels with significant changes in each of the measured combustion gases except for CO when compared to both Tallow biodiesel diesel exhaust ($p \leq 0.05$). In contrast, Tallow and ULSD exhaust were only different for O2, CO2, and SO2. ## 2.2. Exhaust Particle Characteristics Particle size spectra were obtained for all exhausts between the sizes of 3–340 nm; however, no differences were observed for any of the fuels for total particle number concentrations (Figure 1). Particle mass concentrations (Table 1) were highest in ULSD; however, the concentrations in the Canola and Tallow biodiesels were $78\%$ and $92\%$ of those measurements, respectively, showing little differences between fuels. ## 2.3. Mouse Weights Mice were weighed prior to any exposure, and again before lung function assessment, allowing for calculation of the weight changes between the exposure groups (Figure 2). On the day of study, Canola exhaust exposure weights were significantly less than those of ULSD and Tallow. No significant differences were found in % weight increase over the 8 days of exposure for any of the groups. ## 2.4. Lung Function at Functional Residual Capacity Thoracic gas volume (TGV) and lung mechanics (airway resistance (Raw), tissue damping (G), tissue elastance (H), and η (hysteresivity)) at functional residual capacity (FRC) were measured (Table 2). TGV was significantly higher in Canola mice compared with all other groups. Due to this, lung function parameters at FRC were normalized to TGV to generate specific lung function measurements. Canola-exhaust-exposed mice had significantly higher sRaw in comparison to ULSD and Tallow groups ($p \leq 0.01$). Specific G, H, and hysteresivity were significantly higher for Canola mice in comparison to all other groups ($p \leq 0.01$). ## 2.5. Volume Dependence of Lung Function Respiratory pressure–volume curves and volume-dependent Raw, G, and H (Figure 3) were measured throughout a slow, induced inflation–deflation maneuver for each mouse. Specific compliance was significantly lower for Canola-biodiesel-exhaust-exposed mice compared to all other groups ($p \leq 0.0001$). At a volume of 0.7 mL (chosen as it is the largest lung volume with data for all mice), Canola-exposed mice were the most different to every other group, with significantly lower Raw ($p \leq 0.01$), whereas ULSD-exposed mice showed significantly increased tissue damping compared to Air controls ($p \leq 0.05$). ## 2.6. Responsiveness to Methacholine Raw, G, and H were measured after exposure to increasing doses of methacholine (Figure 4). There were significant effects of treatment on the responsiveness to MCh with respect to airway resistance. Tallow mice were significantly more responsive than Air ($p \leq 0.018$) but Canola mice were significantly less responsive than all other groups ($p \leq 0.001$ in all cases). For G, Tallow mice were significantly more responsive than Air mice and Canola mice had significantly lower responses than all other groups ($p \leq 0.01$ in all cases). For H, all treatments were significantly different to each other with Tallow having the highest response and Canola the lowest ($p \leq 0.0001$ in all cases). This pattern was repeated in terms of sensitivity to MCh (evocative concentration needed to reach a $30\%$ increase in Raw, G, and H from saline; Figure 5). The dose of MCh required to elicit a $30\%$ increase in response was significantly lower in the Tallow mice for Raw and H, significantly lower in the ULSD mice for H, and significantly higher in the Canola mice for Raw, G, and H when compared to Air mice ($p \leq 0.05$ in all cases). ## 2.7. Bronchoalveolar Lavage Cells, Mediators, Protein, and Phospholipid Concentrations Total and differential cell counts were performed on bronchoalveolar lavage (Figure 6). Significantly more cells were found in the BAL of Tallow and ULSD mice compared to Air ($p \leq 0.05$). The Tallow group also had more cells than the Canola and ULSD mice ($p \leq 0.05$). This pattern was seen again in the macrophage cell counts, with the addition of ULSD having more macrophages in the BAL than Canola ($p \leq 0.05$). In contrast, the ULSD and Canola mice had significantly fewer neutrophils than the Air-exposed mice ($p \leq 0.05$). Tallow mice had more lymphocytes than the Air, Canola, and ULSD mice ($p \leq 0.05$). No other cell types were detected. In terms of BAL mediator concentrations (Table 3), the majority of significant differences was found between Tallow and Air, with 1 mediator being significantly increased and 3 mediators significantly decreased for Tallow-biodiesel-exhaust-exposed mice ($p \leq 0.05$). Total protein and phospholipid concentrations within the BAL were also measured (Figure 7). There were few effects of exposure on either of these parameters; however, Tallow mice had a significantly increased protein concentration in comparison with Air mice ($p \leq 0.05$). ## 2.8. Multivariate Analysis Fourteen key outcomes obtained from the responsiveness to methacholine analyses (maximum response in airway resistance, tissue damping, and tissue elastance) and bronchoalveolar lavage (total cellular inflammation, numbers of macrophages, neutrophils and lymphocytes, levels of protein, G-CSF, IL-6, IL-10, KC, MIP-1α, and TNF-α) were fitted to the ten exhaust variables (levels of gases O2, CO, CO2, NO, NO2, NOx, and SO2, in addition to particle size, number, and mass) using a redundancy analysis (RDA) (Figure 8) in order to assess how well the exhaust components explained the resulting biological results. Due to inherent limitations of RDA, only data with complete datasets for all mice in all exhaust-exposed groups were able to be included in analyses. To not overfit the model by overloading it with all cytokine data available, six cytokines were chosen from the 21 analyzed mediator levels. These were chosen based on either their importance in previous exhaust toxicology studies [1,22,28,54] (IL-6, IL-10, TNF- α) or for observed significant differences between exhaust groups and the air control (G-CSF, KC, MIP-1 α). The fitted RDA model showed that over half ($56.98\%$) of the biological variability observed in responsiveness to MCh and BAL outcomes was explained by the exhaust parameters. This type of analysis is a powerful tool in supporting the validity of our findings as it shows strong correlations between related parameters. For example, neutrophilic inflammation is strongly positively correlated with levels of KC (a neutrophilic chemoattractant mediator). Similarly, the responsiveness to MCh outcomes (Raw, G, and H) are all positively correlated with each other, as are the majority of exhaust gases. The oxygen concentration in exhaust is negatively correlated with all other exhaust components, which are strongly correlated with each other (as indicated by the red arrows in Figure 8). This is unsurprising due to fuel combustion requiring oxygen for the reaction. Additionally, neutrophilic inflammation is correlated with neither the total number of inflammatory cells in the bronchoalveolar lavage nor the numbers of other cell types (macrophages and lymphocytes); instead, it is highly negatively correlated with particle number and CO2. The total cell count is highly correlated with NO2 concentrations. Perhaps most importantly, many of the mediator and methacholine response measurements are not correlated with any one exhaust component, suggesting the more complex relationship between exposure and toxicological result that is not currently explainable by the inputted data. ## 2.9. Airway Morphometry Size-corrected total airway wall, airway smooth muscle mass, and airway epithelial thickness were measured (Table 4, representative images in Figure S2). Chord length, collagen, and total tissue % were also measured (Table 4). There was no effect of treatment on any airway morphometry parameter ($p \leq 0.05$ in all cases). Chord length was significantly higher in Tallow mice compared to Air mice, and Tallow mice had significantly more collagen than Canola and ULSD mice did, suggesting that exposure to Tallow biodiesel exhaust was causing measurable damage to the airways ($p \leq 0.05$). ## 3. Discussion The results of this study show that exposure to diluted diesel or biodiesel exhaust causes a range of negative health impacts in a murine exposure model. These include impacts on lung function, cellular inflammation, small changes to lung structure, and large impacts to the immune response. Of the two biodiesels tested, Tallow biodiesel exhaust exposure was associated with the widest range of negative health effects with a greater increase in responsiveness to methacholine, a greater than two-fold increase in inflammatory cell numbers in the lungs, a wider disruption in the local mediator release of the lungs, increased protein concentrations in the BAL and a small impact on lung structure with significantly increased chord length. In contrast, Canola biodiesel exhaust exposure only led to negative impacts on lung function at FRC, specific compliance, some decreases in mediator release, and decreased neutrophilic (but not total) inflammation. Mice exposed to Canola biodiesel exhaust were less responsive to MCh than Air-exposed controls, an interesting finding for which we do not have an explanation. The impacts of exposure to ULSD exhaust were generally between those of Canola and Tallow, with increased tissue damping in volume-dependent lung mechanics, several increases in methacholine responses, some decreases in mediator release, and increased immune cell numbers in the lungs of exposed mice. A concerning implication for this study is that negative health impacts (with implications for wide-reaching consequences) were identified, yet the exhaust used mostly met Safe Work Australia standards [62]. These standards are equivalent to the standards used in Europe and USA [63,64]. The Safe Work Australia standards for various exhaust components are time-weighted 8 h averages of 30 ppm of CO, 5000 ppm of CO2 (peak concentration not exceeding 30,000 ppm), 25 ppm of NO, 3 ppm of NO2 (with peak concentrations not exceeding 5 ppm), and 2 ppm of SO2 (peak concentration not exceeding 5 ppm). Oxygen levels below $19.5\%$ are considered “unsafe” [65]. Table 1 shows that almost all exhaust gases in this study (with the exception of a slightly too high NO2 and a slightly too low oxygen concentration) met these standards. The European Union has set a recent particulate matter occupational exposure limit of 50 ug/m3 of elemental carbon [66,67], whereas in America, the limit for a non-coal mining setting is set at 160 µg/m3 of total carbon [68], and in Australia, it is recommended that diesel exhaust not exceed 100 µg/m3 of elemental carbon [69]. In this study, particle mass concentrations were between 42.6 and 54.4 µg/m3, again showing that common exposure standards were not exceeded. In the current study, biodiesel exhaust did not contain higher levels of NOx or lower levels of PM compared with ULSD exhaust, as is commonly reported in the literature [18,19,20,22,70]. That said, other studies that measured NOx and PM have also reported either no differences between biodiesel and mineral diesel exhausts, or a decrease in the biodiesel exhaust [19,47,55]. Our previous studies have also found a wide variation in NOx and PM concentrations in biodiesel exhaust compared to ULSD, with differences dependent on the feedstock type used to make the biodiesel [22,54,71]. This suggests that differences in NOx and PM concentrations between diesel and biodiesel are subtle enough that the dilutions used in toxicology studies to make concentrations “real-world”-relevant can mask the changes [18,20,54] and/or that the differences are related to feedstock type. The overarching idea in the literature that biodiesel exhaust overall contains more NOx and less PM may be feedstock-specific and should, thus, be viewed critically [54]. This idea is further supported by the findings that Tallow biodiesel exhaust is no different to ULSD in terms of PM and NOx, but that Canola biodiesel exhaust contained significantly less NOx. Another potential explanation is that many previously published exhaust-only comparisons have been conducted using old technology engines not equipped with a diesel particulate filter and/or diesel oxidation catalyst [18,20,23] and, thus, increased NOx and decreased PM in biodiesel exhaust may only be applicable to older technology engines [19,47,55]. Regardless, the multivariate RDA results suggest a correlation between exhaust gas components and several toxicological responses, showing that it is important that any future experiments analyzing the toxicity of diesel or biodiesel exhaust should use whole exhaust exposure methods instead of focusing solely on the health effects of exhaust particles alone, as is too often conducted in the previous literature [1,28]. For example, we identified a strong correlation between NO2 and inflammation in the form of total cells and the number of macrophages present in the BAL (Figure 8). This correlation is not surprising based on the known inflammatory effects of NO2 [72]. A key finding of this study is that the Tallow-biodiesel-exhaust-exposed mice were hyperresponsive to MCh with respect to airway resistance, tissue damping, and tissue elastance. The mice in this study were exposed for only two hours per day for 8 days, to exhaust that largely met Safe Work Australia Standards, and yet responsiveness to methacholine increased significantly. The response measured is smaller compared to similar exposure studies in smoking, asthmatic, and respiratory viral infection mouse models [73,74,75]; however, comparisons between models that employ a variety of environmental exposures are difficult. Previous studies testing the response to methacholine in mice after intranasal instillation of black carbon or diesel exhaust found a greater hyperresponsiveness than was measured in our study, although differences in diesel exhaust exposure protocols and methacholine dosages make direct comparisons difficult [76,77]. Studies that co-exposed house dust mites and diesel exhaust also found increased responsiveness to methacholine, although only in the co-exposed group and not in the diesel-exhaust-alone-exposed group [78,79]. Studies testing the response to methacholine after diesel exhaust exposure in asthma and atopy also found increased hyperresponsiveness, although, once again, these studies cannot be directly compared, due to differences in subject type and measurements [80,81]. Our finding of increased airway hyperresponsiveness has concerning implications for those with asthma and allergies who are currently exposed for prolonged periods of time to diesel exhaust, as the increased responsiveness to methacholine suggests that a swap to using some biodiesel feedstocks for fuel may elicit worse responses. Additionally, diesel exhaust can act as a sensitizer to aeroallergens and our data suggest that Tallow-derived biodiesel may further enhance that effect [82]. The results of the RDA (Figure 8) suggest that the relationship between exhaust exposure and methacholine response is not straightforward, nor is it correlated with any one particular exhaust component, unlike the measurement of protein concentration present in the BAL, which is highly correlated with the number of particles present in the exhaust. The finding that mice exposed to Canola biodiesel exhaust for 2 h per day for 8 days were less responsive to methacholine than Air controls was unexpected. Despite being the least toxic in terms of methacholine response and pulmonary cellular inflammation, the Canola biodiesel exposure group displayed both positive and negative health impacts. While the lower 8-day methacholine responsiveness compared to Air, increased thoracic gas volume measurements (despite the Canola mice being significantly smaller than the other groups) and decreased airway resistance in the volume-dependent measurements could be interpreted as positive findings (i.e., “improvements” compared with Air controls), when combined with the negative indications of increased specific Raw, G, and H at FRC, it instead suggests that the complete picture is much more complicated. Diesel (and, thus, likely biodiesel) exhaust is a highly complex mixture made up of thousands of different chemicals [18,37,57,83,84,85,86] and it is possible (and indeed likely, from the results of this study) that exposure to such a mixture could lead to both “positive” and “negative” health impacts as seen for Canola biodiesel exhaust. Further experiments are needed to explore what makes the Canola biodiesel exhaust exposure group so unique. Such research could identify what is changing in the lungs of exposed mice, and also what component(s) of the Canola biodiesel exhaust are associated with the changes. We attempted to address this by examining surfactant levels in the lungs via measurement of BAL choline containing a phospholipid concentration. The surfactant is comprised of approximately $70\%$ of phosphatidylcholine, which, in turn, makes up approximately $80\%$ of phosphatidylcholine in the lungs [87,88]. It is both produced naturally and also used medicinally to improve breathing in preterm children and other children at risk of respiratory failure, as it acts to decrease surface tension at the air–liquid interface of the lung alveoli [88]. However, no difference in phospholipid concentrations was found. Thus, reasons for why the Canola mice responded as they did are difficult to elucidate and warrants further investigation. Another key finding of this study was the increased cell numbers in the BAL of Tallow biodiesel and ULSD-exhaust-exposed mice. This increase mostly consisted of an increase in macrophages, and an increase in lymphocytes in the Tallow-exposed group. A decrease in neutrophils was also observed in the Air group compared to both the Canola and ULSD groups, a finding that is further supported by the RDA showing negative correlations between total cell count and neutrophil count. This suggests that some immune dysregulation might be occurring in mice sub-chronically exposed to exhaust, a finding that is supported by the local (BAL) mediator response that shows significant decreases in the Tallow-exposed mice compared to their respective Air controls. Due to kinetics of immune mediator release after exhaust exposure, wherein the greatest immune responses in a previous study were found 3–6 h after exposure with decreases back to baseline levels observed by 24 h [89], a decrease after a single day of exposure could be explained as immune mediators being “used up”. However, with the depletion effect ongoing after 8 days of exhaust exposure, combined with the decrease in neutrophil numbers in all groups, even if this decrease was only statistically significant for Canola and ULSD, this instead suggests an inability for the mouse immune system to cope with ongoing exhaust exposure, which could have serious consequences for cancer and infection [90,91,92,93]. These findings have been mirrored in a diesel exhaust human exposure study of occupationally exposed workers [90], which found workers exposed to high amounts of exhaust for prolonged periods showed immune dysregulation and decreases in serum inflammatory mediators, such as IL-8 and MIP-1β. In addition, previous studies co-exposing mice to both a respiratory infection and diesel exhaust found that exposure increased infection susceptibility [93,94]. Studies have also been able to induce allergic airways disease using diesel exhaust [82] and human exposure studies on populations with allergic rhinitis found that diesel exhaust exacerbated allergic inflammation, likely by dysregulating the immune systems’ ability to remove eosinophils [95]. There were also minor changes in the lung structure of Tallow-biodiesel-exhaust-exposed mice in terms of a small but statistically significant increase in chord length. Chord length, also known as mean linear intercept, is a measure of the mean space between airway structures [96,97], and increased chord length has been linked to airway damage and disease such as emphysema [98], although it is not a direct measurement of airway size [96]. The Tallow-exposed mice also had increased protein content in the BAL, which is a marker of increased lung permeability and epithelial damage [76,99] and further supports our previous finding of increased epithelial cell damage and increased permeability in air–liquid-interface cell cultures [60]. Increased epithelial damage and increased chord length would indicate damage to the airways after Tallow biodiesel exhaust exposure [97,98], which is concerning after such a relatively short-duration exposure period. There were no other indications of changes to airway morphometry; however, very mild exposures were used in comparison to some previous studies [41,42,100]. ## 4.1. Animals Ninety-six seven-week-old male BALB/c mice were purchased from the Animal Resources Center (Murdoch, WA, Australia) and housed in individually vented cages (IVC Allentown XJ model, ECO FLO Air handling unit set at 22–23 °C with 30–$31\%$ humidity, 50 air changes per hour). They were left to acclimatize for one week before being weighed and randomly assigned into one of 4 different groups ($$n = 24$$ per group). These groups were exposed for 2 h per day for 8 days to Air or the diluted exhaust of an engine running on ULSD, Canola, or Tallow biodiesel (Figure 9). Twenty-four hours after the last exposure, mice were weighed and prepared for end exposure outcomes as previously described [101]. ## 4.2. Engine Configuration and Fuel Information Exhaust was generated by a single-cylinder, 435 cc design Yanmar L100V engine (Yanmar, Italy) coupled with an electric fuel pump with multistage filtration and a dynamometer and fitted with Euro V/VI after-treatment technology consisting of a diesel particulate filter and oxidation catalyst (Daimler, Germany) [71]. All exposures were run from cold start with a constant load of $40\%$ and a speed of 2000 rpm. Air exposures were conducted simultaneously alongside exhaust exposures. The diesel fuel used as the control was obtained from a local supplier (SHELL, Australia, <10 ppm sulfur). Both Canola and Tallow biodiesel were created following an established sodium methoxide transesterification process [102] using high-quality oils obtained from local suppliers (Campbells Wholesale Reseller, WA, Aus and Range Products, WA, Aus). Detailed FAME profiles have been previously published [54,103]. The diesel exhaust exposure consumed less fuel due than was required for both biodiesel exposures, likely due to the differences in fuel efficiency [18]. ## 4.3. Exposure Protocol The exposure methodology is based on a combination of previously published protocols [41,103]. To make exposures more realistic to occupational settings, mice were exposed for short time periods to exhaust diluted approximately $\frac{1}{10}$ with air with cold-start emissions included as part of the exposure. The exhaust was diluted inside a mixing chamber attached to the exhaust piping and pumped through an isokinetic sampling point at a rate of 5 L/min into a sealed incubator (Model 1535, Sheldon Manufacturing, OR, USA) maintained at 28 °C containing a 27 L exposure chamber with the mice inside divided into individual cubicles. The sealed incubator was used to dampen the sound of the engine and keep chamber temperatures constant. During exposures, exhaust was gently vacuumed out of the exposure chamber for physicochemical analysis of gas and particle properties (Figure 10). Simultaneously to the exhaust exposure, a second 4 L exposure chamber was also placed inside the incubator and attached to piping that allowed air to be pumped inside for the Air exposure controls. The difference of pumping air into and vacuuming exhaust out of the different chamber boxes created a pressure gradient that made certain of no chance for cross-exposure contamination, in case of any leakages within the sealed exposure chambers. Fewer Air mice were exposed at any one time (i) because of the smaller control exposure chamber and (ii) to ensure that there were control animals on each data acquisition day. All exposure chambers were thoroughly washed and dried between exposures. ## 4.4. Gas and Particle Analysis Exhaust exiting the exposure chamber was analyzed at the beginning of every exposure and then every 10 min until the end of exposure for concentrations of combustion gas products including O2, NOx (NO and NO2), CO, CO2, and SO2 using a combustion gas analyzer (TESTO 350, Testo, Lenzkirch, Germany). Similarly, exhaust was analyzed every 10 min for particle concentrations between the sizes of 3 nm and 340 nm using a Universal Scanning Mobility Particle Sizer (U-SMPS 1700 Palas, Karlsruhe, Germany). Particles less than 10 nm in size were excluded from further calculations due to the high variability of measurements at that size range. Median particle size count was calculated using the mean number of particles. Particle mass was calculated from particle spectra, assuming sphericity and using the $40\%$ load diesel exhaust particle density as previously described [104]. Particle number was further separated into two fractions: nucleation mode particles below 23 nm in size and solid particles above 23 nm [61]. Due to the high dilution and aftertreatment devices used in this experiment, there was not enough particulate matter collected on quartz filters to perform detailed chemical analysis on polycyclic aromantic hydrocarbon, aldehyde, or heavy metal concentrations present in the exhaust [60]. ## 4.5. Lung Function Measurements Measurements of thoracic gas volume (TGV) and lung mechanics were conducted as previously described [73,105,106]. In brief, mice were anesthetized via intraperitoneal injection of a solution containing ketamine (40 mg/mL; Troy Laboratories, New South Wales, Australia) and xylazine (2 mg/mL; Troy Laboratories, New South Wales, Australia) at a dose of 0.1 mL/10 g body weight, tracheostomized with a 10 mm long cannula with an internal diameter of 0.86 mm, and attached to a mechanical ventilator (HSE Harvard Minivent; Hugo Sachs Harvard Elektronik, March-Hugstetten, Germany). They were ventilated at a rate of 400 breaths/min with a tidal volume of 8 mL/kg and 2 cmH2O of positive-end expiratory pressure, which is sufficient to allow measurement of lung function parameters without either induction of paralysis or autonomous breathing. Plethysmography was used to measure TGV. At end expiration, the trachea was occluded and the intercostal muscles electrically stimulated (six 2 to 3 ms, 20 V pulses, model S44 electrical stimulator; Grass Instruments, Quincy, MA, USA) to induce inspiration with tracheal pressure and plethysmograph box pressure measured throughout. TGV was then calculated using Boyle’s law, after correction for thermal properties and impedance of the plethysmograph [106]. Respiratory system impedance (Zrs) was measured using a wave-tube system adapted for use in small animals [107,108] and a modification of the forced oscillation technique [108]. The constant-phase model was fit to Zrs to generate the parameters of airway resistance (Raw), tissue damping (G), and tissue elastance (H). Zrs was measured at functional respiratory capacity and also during a slow inflation–deflation maneuver from 0 to 20 cmH2O transrespiratory pressure (Prs), allowing for construction of absolute pressure–volume curves and assessment of the volume dependence of lung mechanics. Specific lung compliance was then calculated between Prs = 8 cm/H2O and 3 cm/H2O on the deflationary arm [109]. ## 4.6. Methacholine Challenge After measurement of TGV and lung mechanics, a randomized selection of half the mice from each group ($$n = 12$$) were transferred to a small animal ventilator (Legacy flexiVent; SCIREQ, Montreal, QC, Canada) for assessment of responsiveness to methacholine (MCh; acetyl β-methacholine chloride; Sigma-Aldrich, St. Louis, MO, USA) as previously described [110]. Briefly, 5x forced oscillation technique (FOT) measurements were taken at FRC/baseline (1 per minute), then after a 10 s saline aerosol and again after increasing doses of MCh from 0.1 to 30 mg/mL. Peak responses to MCh at each dose were used to construct dose–response curves. ## 4.7. Bronchoalveolar Lavage (BAL) Collection and Cell Measurement At the end of the methacholine challenge, BAL fluid was collected by washing 0.5 mL of chilled saline in and out of the lungs three times via the tracheal cannula ($$n = 12$$ per group). Lavage samples were processed as previously described for total cell counts [111] and differential cell counts were obtained using DiffQuik (Thermofisher Scientific, Waltham, MA, USA) staining as per the manufacturer’s protocol. In short, samples were centrifuged at 400× g for 4 min to pellet the cells and the supernatant was removed and stored at −80 °C for future mediator, protein, and phospholipid analysis. A total cell count was determined from the cell pellet by staining an aliquot of cells with trypan blue and counting cells with a hemocytometer. Remaining cells were cytospun, stained with DiffQuik, and scanned using a Panoramic MIDI® scanner (3DHISTECH Ltd., Budapest, Hungary) prior to visualization using ImageJ [112] to determine the proportion of cell types within a randomized count of 300 cells. ## 4.8. Lung Fixation, Airway Morphometry, and Histology Non-methacholine-challenged mice had lungs inflation fixed at Prs = 10 cmH2O using $10\%$ formalin [96] prior to removal en bloc for airway morphometry analysis ($$n = 12$$ for all groups). The left lung was embedded in paraffin, and 5 μm thick sections were taken from the proximal region, where the primary bronchi were first fully enclosed by tissue. Three sections from each mouse were stained using Masson’s trichome and the most intact sections were imaged using a Panoramic MIDI® scanner (3DHISTECH Ltd., Budapest, Hungary). Semiautomated assessment of chord length was performed [97,113] and collagen content was quantified as a percentage of total tissue in the cross-sectional area using ImageJ [112]. The cross-sectional area of the outside bronchi wall, airway smooth muscle, the gap between smooth muscle and the epithelium, and the airway epithelium were measured. The square root of all areas was normalized to the internal perimeter of the basement membrane to correct for differences in airway size [114]. ## 4.9. Mediator, Phospholipid, and Protein Analysis BAL was analyzed for mediators as per kit protocol using Bio-Rad Mouse Cytokine 23plx kits (Bio-Rad) and accompanying software (Bio-Plex Manager, v6.1.1, Bio-Rad, Tokyo, Japan). The protein concentration of BAL was used as an indirect measurement of airway epithelial damage and was assessed as per kit protocol using a Pierce™ BCA protein assay kit (Thermofisher Scientific). Phospholipid (choline containing) concentration within the BAL was analyzed to assess surfactant concentrations as per kit protocol using a Colorimetric Phospholipid Assay Kit (Abcam). Serum immunoglobulin concentration was analyzed as per kit protocol using Mouse Immunoglobulin Isotyping Magnetic Bead Panel (Milliplex, MERCK). ## 4.10. Statistical Analysis Data are presented as mean ± standard deviation. This study was performed as part of a larger study with both 1-day and 8-day exposure groups. As the study was initially designed to have exposure groups for both days, statistical analyses were performed on the whole dataset (including both timepoints). Due to the complexity of the data, this larger study has now been separated into two. The 8 days of exposure results are reported in this paper. All statistical analyses were performed using R statistical software (v3.4.3) [115] loaded with the packages “lme4”, “mgcv”, and “Vegan” [116]. p-values less than 0.05 were considered significant. All statistical analyses excluding gas concentration data were completed using multivariate general linear modeling methodologies with the families “Gamma(inverse/log)” and “gaussian(identity/log)” as best fits to the data, applying a backwards elimination approach to remove insignificant predictive variables. For combustion gas analysis, a separate General Additive Model (GAM) file was fitted to each gas measurement with concentration as the response variable and time as the predictor, thus allowing for non-parametric fits as caused by cold-start effects. Redundancy analysis (RDA) [116,117] was performed using the package “Vegan” and a previously published protocol, inputting data obtained only from the methacholine challenge and the bronchoalveolar lavage due to the limitations of the model’s ability to deal handle missing data values, meaning that only whole datasets could be appropriately analyzed. This excluded both the histology data and lung function measurements at FRC due to different sets of mice being used for different measurements, and the mice used for airway morphometry measurements were not the same group as those used for the remainder of the results. To not overfit the model with the highly correlated cytokine data, 6 cytokines were chosen for analysis: IL-6, IL-10, G-CSF, KC, MIP-1α, and TNF-α chosen for either their importance to the innate immune response or for the observed significant differences compared to air-exposed controls [88,89,90,91,92,93,94]. Before completing the RDA, data were standardized using the command “decostand” to help account for different units of measurement used for different values [116,117]. The RDA was then performed as recommended by the “Vegan” package guidelines. ## 5. Conclusions Exposure to diesel and/or biodiesel exhaust impacted lung function measured at FRC, volume-dependent lung function, methacholine responsiveness, inflammation, and airway morphometry in a mouse model. In line with our previous research [54,60]. Tallow biodiesel exhaust exposure resulted in the widest range of negative health impacts, followed by ULSD exhaust with Canola biodiesel exhaust causing the most limited impacts and arguably even having a positive effect on methacholine response. More research is needed to parse out reasons for this. ## References 1. Landwehr K.R., Larcombe A.N., Reid A., Mullins B.J.. Critical review of diesel exhaust exposure health impact research relevant to occupational settings: Are we controlling the wrong pollutants?. Expo. Health (2021.0) 13 141-171. DOI: 10.1007/s12403-020-00379-0 2. Long E., Carlsten C.. Controlled human exposure to diesel exhaust: Results illuminate health effects of traffic-related air pollution and inform future directions. Part. Fibre Toxicol. (2022.0) 19 11. DOI: 10.1186/s12989-022-00450-5 3. Silverman D.T., Samanic C.M., Lubin J.H., Blair A.E., Stewart P.A., Vermeulen R., Coble J.B., Rothman N., Schleiff P.L., Travis W.D.. The Diesel Exhaust in Miners study: A nested case-control study of lung cancer and diesel exhaust. J. Natl. Cancer Inst. (2012.0) 104 855-868. DOI: 10.1093/jnci/djs034 4. Attfield M.D., Schleiff P.L., Lubin J.H., Blair A., Stewart P.A., Vermeulen R., Coble J.B., Silverman D.T.. The Diesel Exhaust in Miners study: A cohort mortality study with emphasis on lung cancer. J. Natl. Cancer Inst. (2012.0) 104 869-883. DOI: 10.1093/jnci/djs035 5. Andersen Z.J., Pedersen M., Weinmayr G., Stafoggia M., Galassi C., Jørgensen J.T., Sommar J.N., Forsberg B., Olsson D., Oftedal B.. Long-term exposure to ambient air pollution and incidence of brain tumor: The European Study of Cohorts for Air Pollution Effects (ESCAPE). Neuro Oncol. (2018.0) 20 420-432. DOI: 10.1093/neuonc/nox163 6. Latifovic L., Villeneuve P.J., Parent M.-É., Johnson K.C., Kachuri L., Harris S.A.. Bladder cancer and occupational exposure to diesel and gasoline engine emissions among Canadian men. Cancer Med. (2015.0) 4 1948-1962. DOI: 10.1002/cam4.544 7. Cosselman K.E., Krishnan R., Oron A.P., Jansen K., Peretz A., Sullivan J.H., Larson T.V., Kaufman J.D.. Blood pressure response to controlled diesel exhaust exposure in human subjects. Hypertension (2012.0) 59 943-948. DOI: 10.1161/HYPERTENSIONAHA.111.186593 8. Crüts B., van Etten L., Törnqvist H., Blomberg A., Sandström T., Mills N.L., Borm P.J.. Exposure to diesel exhaust induces changes in EEG in human volunteers. Part. Fibre Toxicol. (2008.0) 5 4. DOI: 10.1186/1743-8977-5-4 9. Mills N.L., Törnqvist H., Gonzalez M.C., Vink E., Robinson S.D., Söderberg S., Boon N.A., Donaldson K., Sandström T., Blomberg A.. Ischemic and Thrombotic Effects of Dilute Diesel-Exhaust Inhalation in Men with Coronary Heart Disease. N. Engl. J. Med. (2007.0) 357 1075-1082. DOI: 10.1056/NEJMoa066314 10. Zhu N., Li H., Han M., Guo L., Chen L., Yun Y., Guo Z., Li G., Sang N.. Environmental nitrogen dioxide (NO. Toxicol. Lett. (2012.0) 214 120-130. DOI: 10.1016/j.toxlet.2012.08.021 11. Balti E.V., Echouffo-Tcheugui J.B., Yako Y.Y., Kengne A.P.. Air pollution and risk of type 2 diabetes mellitus: A systematic review and meta-analysis. Diabetes Res. Clin. Pract. (2014.0) 106 161-172. DOI: 10.1016/j.diabres.2014.08.010 12. Eze I.C., Hemkens L.G., Bucher H.C., Hoffmann B., Schindler C., Künzli N., Schikowski T., Probst-Hensch N.M.. Association between ambient air pollution and diabetes mellitus in Europe and North America: Systematic review and meta-analysis. Environ. Health Perspect. (2015.0) 123 381-389. DOI: 10.1289/ehp.1307823 13. Fleisch A.F., Kloog I., Luttmann-Gibson H., Gold D.R., Oken E., Schwartz J.D.. Air pollution exposure and gestational diabetes mellitus among pregnant women in Massachusetts: A cohort study. Environ. Health A Glob. Access Sci. Source (2016.0) 15 40. DOI: 10.1186/s12940-016-0121-4 14. Evans K.A., Halterman J.S., Hopke P.K., Fagnano M., Rich D.Q.. Increased ultrafine particles and carbon monoxide concentrations are associated with asthma exacerbation among urban children. Environ. Res. (2014.0) 129 11-19. DOI: 10.1016/j.envres.2013.12.001 15. Kittelson D., Watts W., Johnson J.. Diesel Aerosol Sampling Methodology (2002.0) 16. Ris C.U.S.. EPA Health Assessment for Diesel Engine Exhaust: A Review. Inhal. Toxicol. (2007.0) 19 229-239. DOI: 10.1080/08958370701497960 17. Yusuf A.A., Yusuf D.A., Jie Z., Bello T.Y., Tambaya M., Abdullahi B., Muhammed-Dabo I.A., Yahuza I., Dandakouta H.. Influence of waste oil-biodiesel on toxic pollutants from marine engine coupled with emission reduction measures at various loads. Atmos. Pollut. Res. (2022.0) 13 101258. DOI: 10.1016/j.apr.2021.101258 18. Fontaras G., Karavalakis G., Kousoulidou M., Tzamkiozis T., Ntziachristos L., Bakeas E., Stournas S., Samaras Z.. Effects of biodiesel on passenger car fuel consumption, regulated and non-regulated pollutant emissions over legislated and real-world driving cycles. Fuel (2009.0) 88 1608-1617. DOI: 10.1016/j.fuel.2009.02.011 19. Prokopowicz A., Zaciera M., Sobczak A., Bielaczyc P., Woodburn J.. The Effects of Neat Biodiesel and Biodiesel and HVO Blends in Diesel Fuel on Exhaust Emissions from a Light Duty Vehicle with a Diesel Engine. Environ. Sci. Technol. (2015.0) 49 7473-7482. DOI: 10.1021/acs.est.5b00648 20. Graver B.M., Frey H.C., Hu J.. Effect of Biodiesel Fuels on Real-World Emissions of Passenger Locomotives. Environ. Sci. Technol. (2016.0) 50 12030-12039. DOI: 10.1021/acs.est.6b03567 21. Giakoumis E.G., Rakopoulos C.D., Dimaratos A.M., Rakopoulos D.C.. Exhaust emissions of diesel engines operating under transient conditions with biodiesel fuel blends. Prog. Energy Combust. Sci. (2012.0) 38 691-715. DOI: 10.1016/j.pecs.2012.05.002 22. Mullins B.J., Kicic A., Ling K.-M., Mead-Hunter R., Larcombe A.N.. Biodiesel Exhaust–Induced Cytotoxicity and Proinflammatory Mediator Production in Human Airway Epithelial Cells. Environ. Toxicol. (2016.0) 31 44-57. DOI: 10.1002/tox.22020 23. Westphal G.A., Krahl J., Munack A., Rosenkranz N., Schröder O., Schaak J., Pabst C., Brüning T., Bünger J.. Combustion of Hydrotreated Vegetable Oil and Jatropha Methyl Ester in a Heavy Duty Engine: Emissions and Bacterial Mutagenicity. Environ. Sci. Technol. (2013.0) 47 6038-6046. DOI: 10.1021/es400518d 24. Gioda A., Rodríguez-Cotto R.I., Amaral B.S., Encarnación-Medina J., Ortiz-Martínez M.G., Jiménez-Vélez B.D.. Biodiesel from Soybean Promotes Cell Proliferation in Vitro. Toxicol. Vitr. (2016.0) 34 283-288. DOI: 10.1016/j.tiv.2016.05.004 25. Oberdörster G., Celein R.M., Ferin J., Weiss B.. Association of Particulate Air Pollution and Acute Mortality: Involvement of Ultrafine Particles?. Inhal. Toxicol. (1995.0) 7 111-124. DOI: 10.3109/08958379509014275 26. Breitner S., Liu L., Cyrys J., Brüske I., Franck U., Schlink U., Leitte A.M., Herbarth O., Wiedensohler A., Wehner B.. Sub-micrometer particulate air pollution and cardiovascular mortality in Beijing, China. Sci. Total Environ. (2011.0) 409 5196-5204. DOI: 10.1016/j.scitotenv.2011.08.023 27. International Energy Statistics 28. Larcombe A.N., Kicic A., Mullins B.J., Knothe G.. Biodiesel exhaust: The need for a systematic approach to health effects research. Respirology (2015.0) 20 1034-1045. DOI: 10.1111/resp.12587 29. André V., Barraud C., Capron D., Preterre D., Keravec V., Vendeville C., Cazier F., Pottier D., Morin J.P., Sichel F.. Comparative mutagenicity and genotoxicity of particles and aerosols emitted by the combustion of standard vs. rapeseed methyl ester supplemented bio-diesel fuels: Impact of after treatment devices: Oxidation catalyst and particulate filter. Mutat. Res./Genet. Toxicol. Environ. Mutagen. (2015.0) 777 33-42. DOI: 10.1016/j.mrgentox.2014.11.007 30. Møller P., Scholten R.H., Roursgaard M., Krais A.M.. Inflammation, oxidative stress and genotoxicity responses to biodiesel emissions in cultured mammalian cells and animals. Crit. Rev. Toxicol. (2020.0) 50 383-401. DOI: 10.1080/10408444.2020.1762541 31. Madden M.C.. A paler shade of green? The toxicology of biodiesel emissions: Recent findings from studies with this alternative fuel. Biochim. Biophys. Acta-Gen. Subj. (2016.0) 1860 2856-2862. DOI: 10.1016/j.bbagen.2016.05.035 32. Swanson K.J., Madden M.l.C., Ghio A.J.. Biodiesel Exhaust: The Need for Health Effects Research. Environ. Health Perspect. (2007.0) 115 496-499. DOI: 10.1289/ehp.9631 33. Yanamala N., Hatfield M.K., Farcas M.T., Schwegler-Berry D., Hummer J.A., Shurin M.R., Birch M.E., Gutkin D.W., Kisin E., Kagan V.E.. Biodiesel versus diesel exposure: Enhanced pulmonary inflammation, oxidative stress, and differential morphological changes in the mouse lung. Toxicol. Appl. Pharmacol. (2013.0) 272 373-383. DOI: 10.1016/j.taap.2013.07.006 34. Chen T.-M., Kuschner W.G., Gokhale J., Shofer S.. Outdoor Air Pollution: Nitrogen Dioxide, Sulfur Dioxide, and Carbon Monoxide Health Effects. Am. J. Med. Sci. (2007.0) 333 249-256. DOI: 10.1097/MAJ.0b013e31803b900f 35. Morin J.-P., Hasson V., Fall M., Papaioanou E., Preterre D., Gouriou F., Keravec V., Konstandopoulos A., Dionnet F.. Prevalidation of in vitro continuous flow exposure systems as alternatives to in vivo inhalation safety evaluation experimentations: Outcome from MAAPHRI-PCRD5 research program. Exp. Toxicol. Pathol. (2008.0) 60 195-205. DOI: 10.1016/j.etp.2008.01.007 36. Lichtveld K.M., Ebersviller S.M., Sexton K.G., Vizuete W., Jaspers I., Jeffries H.E.. In Vitro Exposures in Diesel Exhaust Atmospheres: Resuspension of PM from Filters versus Direct Deposition of PM from Air. Environ. Sci. Technol. (2012.0) 46 9062-9070. DOI: 10.1021/es301431s 37. Bünger J., Krahl J., Baum K., Schröder O., Müller M., Westphal G., Ruhnau P., Schulz T.G., Hallier E.. Cytotoxic and mutagenic effects, particle size and concentration analysis of diesel engine emissions using biodiesel and petrol diesel as fuel. Arch. Toxicol. (2000.0) 74 490-498. DOI: 10.1007/s002040000155 38. Cervena T., Rossnerova A., Sikorova J., Beranek V., Vojtisek-Lom M., Ciganek M., Topinka J., Rossner P.. DNA Damage Potential of Engine Emissions Measured In Vitro by Micronucleus Test in Human Bronchial Epithelial Cells. Basic Clin. Pharmacol. Toxicol. (2017.0) 121 102-108. DOI: 10.1111/bcpt.12693 39. Jalava P.I., Aakko-Saksa P., Murtonen T., Happo M.S., Markkanen A., Yli-Pirilä P., Hakulinen P., Hillamo R., Mäki-Paakkanen J., Salonen R.O.. Toxicological properties of emission particles from heavy duty engines powered by conventional and bio-based diesel fuels and compressed natural gas. Part. Fibre Toxicol. (2012.0) 9 37. DOI: 10.1186/1743-8977-9-37 40. Brito J.M., Belotti L., Toledo A.C., Antonangelo L., Silva F.S., Alvim D.S., Andre P.A., Saldiva P.H.N., Rivero D.H.R.F.. Acute Cardiovascular and Inflammatory Toxicity Induced by Inhalation of Diesel and Biodiesel Exhaust Particles. Toxicol. Sci. (2010.0) 116 67-78. DOI: 10.1093/toxsci/kfq107 41. De Brito J.M., Mauad T., Cavalheiro G.F., Yoshizaki K., de André P.A., Lichtenfels A.J.F.C., Guimarães E.T., Rivero D.H.R.F., Antonangelo L., Oliveira L.B.. Acute exposure to diesel and sewage biodiesel exhaust causes pulmonary and systemic inflammation in mice. Sci. Total Environ. (2018.0) 628–629 1223-1233. DOI: 10.1016/j.scitotenv.2018.02.019 42. Gavett S.H., Wood C.E., Williams M.A., Cyphert J.M., Boykin E.H., Daniels M.J., Copeland L.B., King C., Krantz T.Q., Richards J.H.. Soy biodiesel emissions have reduced inflammatory effects compared to diesel emissions in healthy and allergic mice. Inhal. Toxicol. (2015.0) 27 533-544. DOI: 10.3109/08958378.2015.1054966 43. Farraj A.K., Haykal-Coates N., Winsett D.W., Gilmour M.I., King C., Krantz Q.T., Richards J., Hazari M.S.. Comparative electrocardiographic, autonomic and systemic inflammatory responses to soy biodiesel and petroleum diesel emissions in rats. Inhal. Toxicol. (2015.0) 27 564-575. DOI: 10.3109/08958378.2015.1057884 44. Hazari M.S., Haykal-Coates N., Winsett D.W., King C., Krantz Q.T., Gilmour M.I., Farraj A.K.. The effects of B0, B20, and B100 soy biodiesel exhaust on aconitine-induced cardiac arrhythmia in spontaneously hypertensive rats. Inhal. Toxicol. (2015.0) 27 557-563. DOI: 10.3109/08958378.2015.1054967 45. Bass V.L., Schladweiler M.C., Nyska A., Thomas R.F., Miller D.B., Krantz T., King C., Gilmour M.I., Ledbetter A.D., Richards J.E.. Comparative cardiopulmonary toxicity of exhausts from soy-based biofuels and diesel in healthy and hypertensive rats. Inhal. Toxicol. (2015.0) 27 545-556. DOI: 10.3109/08958378.2015.1060279 46. Douki T., Corbière C., Preterre D., Martin P.J., Lecureur V., André V., Landkocz Y., Pottier I., Keravec V., Fardel O.. Comparative study of diesel and biodiesel exhausts on lung oxidative stress and genotoxicity in rats. Environ. Pollut. (2018.0) 235 514-524. DOI: 10.1016/j.envpol.2017.12.077 47. Valand R., Magnusson P., Dziendzikowska K., Gajewska M., Wilczak J., Oczkowski M., Kamola D., Królikowski T., Kruszewski M., Lankoff A.. Gene expression changes in rat brain regions after 7- and 28 days inhalation exposure to exhaust emissions from 1st and 2nd generation biodiesel fuels—The FuelHealth project. Inhal. Toxicol. (2018.0) 30 299-312. DOI: 10.1080/08958378.2018.1520370 48. Magnusson P., Dziendzikowska K., Oczkowski M., Øvrevik J., Eide D.M., Brunborg G., Gutzkow K.B., Instanes C., Gajewska M., Wilczak J.. Lung effects of 7- and 28-day inhalation exposure of rats to emissions from 1st and 2nd generation biodiesel fuels with and without particle filter—The FuelHealth project. Environ. Toxicol. Pharmacol. (2019.0) 67 8-20. DOI: 10.1016/j.etap.2019.01.005 49. Dziendzikowska K., Gajewska M., Wilczak J., Mruk R., Oczkowski M., Żyła E., Królikowski T., Stachoń M., Øvrevik J., Myhre O.. The effects of 1st and 2nd generation biodiesel exhaust exposure on hematological and biochemical blood indices of Fisher344 male rats—The FuelHealth project. Environ. Toxicol. Pharmacol. (2018.0) 63 34-47. DOI: 10.1016/j.etap.2018.08.001 50. Magnusson P., Oczkowski M., Øvrevik J., Gajewska M., Wilczak J., Biedrzycki J., Dziendzikowska K., Kamola D., Królikowski T., Kruszewski M.. No adverse lung effects of 7- and 28-day inhalation exposure of rats to emissions from petrodiesel fuel containing 20% rapeseed methyl esters (B20) with and without particulate filter—The FuelHealth project. Inhal. Toxicol. (2017.0) 29 206-218. DOI: 10.1080/08958378.2017.1339149 51. EERE, Energy Efficiency & Renewable Energy, Alternate Fuels Data Centre. US Department of Energy 2020 52. 52. ASTM International ASTM D7467-20a, Standard Specification for Diesel Fuel Oil, Biodiesel Blend (B6 to B20)ASTMWest Conshohocken, PA, USA202010.1520/D7467-20A. ASTM D7467-20a, Standard Specification for Diesel Fuel Oil, Biodiesel Blend (B6 to B20) (2020.0). DOI: 10.1520/D7467-20A 53. Hamje H.D., Hass H., Lonza L., Maas H., Reid A., Rose K., Venderbosch T.. EU renewable energy targets in 2020: Revised analysis of scenarios for transport fuels. Sci. Policy Rep. Jt. Res. Cent. Eur. Comm. (2014.0) 27. DOI: 10.2790/1725 54. Landwehr K.R., Hillas J., Mead-Hunter R., Brooks P., King A., O’Leary R.A., Kicic A., Mullins B.J., Larcombe A.N.. Fuel feedstock determines biodiesel exhaust toxicity in a human airway epithelial cell exposure model. J. Hazard. Mater. (2021.0) 420 126637. DOI: 10.1016/j.jhazmat.2021.126637 55. Hawley B., L’Orange C., Olsen D.B., Marchese A.J., Volckens J.. Oxidative stress and aromatic hydrocarbon response of human bronchial epithelial cells exposed to petro- or biodiesel exhaust treated with a diesel particulate filter. Toxicol. Sci. Off. J. Soc. Toxicol. (2014.0) 141 505-514. DOI: 10.1093/toxsci/kfu147 56. Ackland M.L., Zou L., Freestone D., Van De Waasenburg S., Michalczyk A.A.. Diesel exhaust particulate matter induces multinucleate cells and zinc transporter-dependent apoptosis in human airway cells. Immunol. Cell Biol. (2007.0) 85 617-622. DOI: 10.1038/sj.icb.7100109 57. Hemmingsen J.G., Møller P., Nøjgaard J.K., Roursgaard M., Loft S.. Oxidative Stress, Genotoxicity, And Vascular Cell Adhesion Molecule Expression in Cells Exposed to Particulate Matter from Combustion of Conventional Diesel and Methyl Ester Biodiesel Blends. Environ. Sci. Technol. (2011.0) 45 8545-8551. DOI: 10.1021/es200956p 58. 58. OECD/FAO OECD-FAO Agricultural Outlook 2020–2029FAORome, Italy2020. OECD-FAO Agricultural Outlook 2020–2029 (2020.0) 59. 59. ARENA (Australian Renewable Energy Agency), Clean Energy Finance Corporation Biofuels and Transport: An Australian OpportunityARENA (Australian Renewable Energy Agency)Canberra, Australia2018. Biofuels and Transport: An Australian Opportunity (2018.0) 60. Landwehr K.R., Hillas J., Mead-Hunter R., King A., O’Leary R.A., Kicic A., Mullins B.J., Larcombe A.N.. Toxicity of different biodiesel exhausts in primary human airway epithelial cells grown at air-liquid interface. Sci. Total Environ. (2022.0) 832 155016. DOI: 10.1016/j.scitotenv.2022.155016 61. Amanatidis S., Ntziachristos L., Giechaskiel B., Bergmann A., Samaras Z.. Impact of Selective Catalytic Reduction on Exhaust Particle Formation over Excess Ammonia Events. Environ. Sci. Technol. (2014.0) 48 11527-11534. DOI: 10.1021/es502895v 62. 62. SafeWork Australia Workplace Exposure Standards for Airborne ContaminantsSafeWork AustraliaCanberra, Australia2019. Workplace Exposure Standards for Airborne Contaminants (2019.0) 63. Workplace Exposure Standards for Airborne Contaminants 64. Permissible Exposure Limits–Annotated Tables 65. 65. SafeWork Australia Confined spaces Code of PracticeSafeWork AustraliaCanberra, Australia2018. Confined spaces Code of Practice (2018.0) 66. European Union Directive 2004/37/EC of the European Parliament and of the Council of 29 April 2004 on the protection of workers from the risks related to exposure to carcinogens or mutagens at work. Off. J. Eur. Union. (2004.0) 47 50-76 67. European Union Directive (EU) 2019/130 of the European Parliament and of the Council of 16 January 2019 amending Directive 2004/37/EC on the protection of workers from the risks related to exposure to carcinogens or mutagens at work (Text with EEA relevance). Off. J. Eur. Union (2019.0) 30 112 68. 68. Mine Safety and Health Administration Exposure of Underground Miners to Diesel ExhaustMine Safety and Health AdministrationArlington, VA, USA2016. Exposure of Underground Miners to Diesel Exhaust (2016.0) 69. 69. Australian Institute of Occupational Hygienists Diesel Particulate Matter and Occupational Health Issues-Position PaperAustralian Institute of Occupational HygienistsTullamarine, Australia2017. Diesel Particulate Matter and Occupational Health Issues-Position Paper (2017.0) 70. Unosson J., Kabéle M., Boman C., Nyström R., Sadiktsis I., Westerholm R., Mudway I.S., Purdie E., Raftis J., Miller M.R.. Acute cardiovascular effects of controlled exposure to dilute Petrodiesel and biodiesel exhaust in healthy volunteers: A crossover study. Part. Fibre Toxicol. (2021.0) 18 22. DOI: 10.1186/s12989-021-00412-3 71. Landwehr K.R., Hillas J., Mead-Hunter R., O’Leary R.A., Kicic A., Mullins B.J., Larcombe A.N.. Soy Biodiesel Exhaust is More Toxic than Mineral Diesel Exhaust in Primary Human Airway Epithelial Cells. Environ. Sci. Technol. (2019.0) 53 11437-11446. DOI: 10.1021/acs.est.9b01671 72. Hesterberg T.W., Bunn W.B., McClellan R.O., Hamade A.K., Long C.M., Valberg P.A.. Critical review of the human data on short-term nitrogen dioxide (NO2) exposures: Evidence for NO2 no-effect levels. Crit. Rev. Toxicol. (2009.0) 39 743-781. DOI: 10.3109/10408440903294945 73. Larcombe A.N., Janka M.A., Mullins B.J., Berry L.J., Bredin A., Franklin P.J.. The effects of electronic cigarette aerosol exposure on inflammation and lung function in mice. Am. J. Physiol.-Lung Cell. Mol. Physiol. (2017.0) 313 L67-L79. DOI: 10.1152/ajplung.00203.2016 74. Wang K.C., Cras T.D.L., Larcombe A.N., Zosky G.R., Elliot J.G., James A.L., Noble P.B.. Independent and combined effects of airway remodelling and allergy on airway responsiveness. Clin. Sci. (2018.0) 132 327-338. DOI: 10.1042/CS20171386 75. Collins R.A., Gualano R.C., Zosky G.R., Atkins C.L., Turner D.J., Colasurdo G.N., Sly P.D.. Hyperresponsiveness to inhaled but not intravenous methacholine during acute respiratory syncytial virus infection in mice. Respir. Res. (2005.0) 6 142. DOI: 10.1186/1465-9921-6-142 76. Lambert A.L., Mangum J.B., DeLorme M.P., Everitt J.I.. Ultrafine Carbon Black Particles Enhance Respiratory Syncytial Virus-Induced Airway Reactivity, Pulmonary Inflammation, and Chemokine Expression. Toxicol. Sci. (2003.0) 72 339-346. DOI: 10.1093/toxsci/kfg032 77. Nemmar A., Al-Salam S., Zia S., Marzouqi F., Al-Dhaheri A., Subramaniyan D., Dhanasekaran S., Yasin J., Ali B.H., Kazzam E.E.. Contrasting actions of diesel exhaust particles on the pulmonary and cardiovascular systems and the effects of thymoquinone. Br. J. Pharm. (2011.0) 164 1871-1882. DOI: 10.1111/j.1476-5381.2011.01442.x 78. Brandt E.B., Biagini Myers J.M., Acciani T.H., Ryan P.H., Sivaprasad U., Ruff B., LeMasters G.K., Bernstein D.I., Lockey J.E., LeCras T.D.. Exposure to allergen and diesel exhaust particles potentiates secondary allergen-specific memory responses, promoting asthma susceptibility. J. Allergy Clin. Immunol. (2015.0) 136 295-303.e297. DOI: 10.1016/j.jaci.2014.11.043 79. Brandt E.B., Kovacic M.B., Lee G.B., Gibson A.M., Acciani T.H., Le Cras T.D., Ryan P.H., Budelsky A.L., Khurana Hershey G.K.. Diesel exhaust particle induction of IL-17A contributes to severe asthma. J. Allergy Clin. Immunol. (2013.0) 132 1194-1204.e1192. DOI: 10.1016/j.jaci.2013.06.048 80. Nordenhäll C., Pourazar J., Ledin M.-C., Levin J.-O., Sandström T., Ädelroth E.. Diesel exhaust enhances airway responsiveness in asthmatic subjects. Eur. Respir. J. (2001.0) 17 909-915. DOI: 10.1183/09031936.01.17509090 81. Carlsten C., MacNutt M.J., Zhang Z., Sava F., Pui M.M.. Anti-Oxidant N-Acetylcysteine Diminishes Diesel Exhaust-Induced Increased Airway Responsiveness in Person with Airway Hyper-Reactivity. Toxicol. Sci. (2014.0) 139 479-487. DOI: 10.1093/toxsci/kfu040 82. Alvarez-Simón D., Muñoz X., Gómez-Ollés S., de Homdedeu M., Untoria M.-D., Cruz M.-J.. Effects of diesel exhaust particle exposure on a murine model of asthma due to soybean. PLoS ONE (2017.0) 12. DOI: 10.1371/journal.pone.0179569 83. Khalek I.A., Bougher T.L., Merritt P.M., Zielinska B.. Regulated and Unregulated Emissions from Highway Heavy-Duty Diesel Engines Complying with U.S. Environmental Protection Agency 2007 Emissions Standards. J. Air Waste Manag. Assoc. (2011.0) 61 427-442. DOI: 10.3155/1047-3289.61.4.427 84. Hesterberg T.W., Long C.M., Sax S.N., Lapin C.A., McClellan R.O., Bunn W.B., Valberg P.A.. Particulate Matter in New Technology Diesel Exhaust (NTDE) is Quantitatively and Qualitatively Very Different from that Found in Traditional Diesel Exhaust (TDE). J. Air Waste Manag. Assoc. (2011.0) 61 894-913. DOI: 10.1080/10473289.2011.599277 85. Karavalakis G., Stournas S., Bakeas E.. Light vehicle regulated and unregulated emissions from different biodiesels. Sci. Total Environ. (2009.0) 407 3338-3346. DOI: 10.1016/j.scitotenv.2008.12.063 86. Kisin E.R., Shi X.C., Keane M.J., Bugarski A.B., Shvedova A.A.. Mutagenicity of biodiesel or diesel exhaust particles and the effect of engine operating conditions. J. Environ. Eng. Ecol. Sci. (2013.0) 2. DOI: 10.7243/2050-1323-2-3 87. Bernhard W.. Lung surfactant: Function and composition in the context of development and respiratory physiology. Ann. Anat. = Anat. Anz. Off. Organ Anat. Ges. (2016.0) 208 146-150. DOI: 10.1016/j.aanat.2016.08.003 88. Chakraborty M., Kotecha S.. Pulmonary surfactant in newborn infants and children. Breathe (2013.0) 9 476-488. DOI: 10.1183/20734735.006513 89. Boylen C.E., Sly P.D., Zosky G.R., Larcombe A.N.. Physiological and inflammatory responses in an anthropomorphically relevant model of acute diesel exhaust particle exposure are sex and dose-dependent. Inhal. Toxicol. (2011.0) 23 906-917. DOI: 10.3109/08958378.2011.625454 90. Dai Y., Ren D., Bassig B.A., Vermeulen R., Hu W., Niu Y., Duan H., Ye M., Meng T., Xu J.. Occupational exposure to diesel engine exhaust and serum cytokine levels. Environ. Mol. Mutagen. (2018.0) 59 144-150. DOI: 10.1002/em.22142 91. Gowdy K.M., Krantz Q.T., King C., Boykin E., Jaspers I., Linak W.P., Gilmour M.I.. Role of oxidative stress on diesel-enhanced influenza infection in mice. Part. Fibre Toxicol. (2010.0) 7 34. DOI: 10.1186/1743-8977-7-34 92. Zarcone M.C., van Schadewijk A., Duistermaat E., Hiemstra P.S., Kooter I.M.. Diesel exhaust alters the response of cultured primary bronchial epithelial cells from patients with chronic obstructive pulmonary disease (COPD) to non-typeable Haemophilus influenzae. Respir. Res. (2017.0) 18 27. DOI: 10.1186/s12931-017-0510-4 93. Shears R.K., Jacques L.C., Naylor G., Miyashita L., Khandaker S., Lebre F., Lavelle E.C., Grigg J., French N., Neill D.R.. Exposure to diesel exhaust particles increases susceptibility to invasive pneumococcal disease. J. Allergy Clin. Immunol. (2020.0) 145 1272-1284.e1276. DOI: 10.1016/j.jaci.2019.11.039 94. Larcombe A.N., Foong R.E., Boylen C.E., Zosky G.R.. Acute diesel exhaust particle exposure increases viral titre and inflammation associated with existing influenza infection, but does not exacerbate deficits in lung function. Influenza Other Respir. Viruses (2012.0) 7 701-709. DOI: 10.1111/irv.12012 95. Pawlak E.A., Noah T.L., Zhou H., Chehrazi C., Robinette C., Diaz-Sanchez D., Müller L., Jaspers I.. Diesel exposure suppresses natural killer cell function and resolution of eosinophil inflammation: A randomized controlled trial of exposure in allergic rhinitics. Part. Fibre Toxicol. (2016.0) 13 24. DOI: 10.1186/s12989-016-0135-7 96. Hsia C.C., Hyde D.M., Ochs M., Weibel E.R.. An Official Research Policy Statement of the American Thoracic Society/European Respiratory Society: Standards for Quantitative Assessment of Lung Structure. Am. J. Respir. Crit. Care Med. (2010.0) 181 394-418. DOI: 10.1164/rccm.200809-1522ST 97. Crowley G., Kwon S., Caraher E.J., Haider S.H., Lam R., Batra P., Melles D., Liu M., Nolan A.. Quantitative lung morphology: Semi-automated measurement of mean linear intercept. BMC Pulm. Med. (2019.0) 19. DOI: 10.1186/s12890-019-0915-6 98. Mitzner W.. Use of mean airspace chord length to assess emphysema. J. Appl. Physiol. (2008.0) 105 1980-1981. DOI: 10.1152/japplphysiol.90968.2008 99. Zarcone M.C., Duistermaat E., Schadewijk A.V., Jedynska A., Hiemstra P.S., Kooter I.M.. Cellular response of mucociliary differentiated primary bronchial epithelial cells to diesel exhaust. Am. J. Physiol.-Lung Cell. Mol. Physiol. (2016.0) 311 L111-L123. DOI: 10.1152/ajplung.00064.2016 100. Larcombe A.N., Phan J.A., Kicic A., Perks K.L., Mead-Hunter R., Mullins B.J.. Route of exposure alters inflammation and lung function responses to diesel exhaust. Inhal. Toxicol. (2014.0) 26 409-418. DOI: 10.3109/08958378.2014.909910 101. Larcombe A.N., Zosky G.R., Bozanich E.M., Turner D.J., Hantos Z., Sly P.D.. Absence of cholinergic airway tone in normal BALB/c mice. Respir. Physiol. Neurobiol. (2008.0) 161 223-229. DOI: 10.1016/j.resp.2008.01.009 102. Knothe G., de Castro M.E.G., Razon L.F.. Methyl Esters (Biodiesel) from and Fatty Acid Profile of Gliricidia sepium Seed Oil. J. Am. Oil Chem. Soc. (2015.0) 92 769-775. DOI: 10.1007/s11746-015-2634-3 103. Landwehr K.R., Hillas J., Mead-Hunter R., Brooks P., King A., O’Leary R.A., Kicic A., Mullins B.J., Larcombe A.N.. In Vitro primary human airway epithelial whole exhaust exposure. MethodsX (2021.0) 8 101561. DOI: 10.1016/j.mex.2021.101561 104. Olfert J.S., Symonds J.P.R., Collings N.. The effective density and fractal dimension of particles emitted from a light-duty diesel vehicle with a diesel oxidation catalyst. J. Aerosol Sci. (2007.0) 38 69-82. DOI: 10.1016/j.jaerosci.2006.10.002 105. Larcombe A.N., Foong R.E., Berry L.J., Zosky G.R., Sly P.D.. In utero cigarette smoke exposure impairs somatic and lung growth in BALB/c mice. Eur. Respir. J. (2011.0) 38 932-938. DOI: 10.1183/09031936.00156910 106. Jánosi T.Z., Adamicza Á., Zosky G.R., Asztalos T., Sly P.D., Hantos Z.. Plethysmographic estimation of thoracic gas volume in apneic mice. J. Appl. Physiol. (2006.0) 101 454-459. DOI: 10.1152/japplphysiol.00011.2006 107. Peták F., Hantos Z., Adamicza Á., Asztalos T., Sly P.D.. Methacholine-induced bronchoconstriction in rats: Effects of intravenous vs. aerosol delivery. J. Appl. Physiol. (1997.0) 82 1479-1487. DOI: 10.1152/jappl.1997.82.5.1479 108. Sly P.D., Collins R.A., Thamrin C., Turner D.J., Hantos Z.. Volume dependence of airway and tissue impedances in mice. J. Appl. Physiol. (2003.0) 94 1460-1466. DOI: 10.1152/japplphysiol.00596.2002 109. Limjunyawong N., Fallica J., Horton M.R., Mitzner W.. Measurement of the Pressure-volume Curve in Mouse Lungs. JoVE (2015.0) 27 e52376. DOI: 10.3791/52376 110. Larcombe A.N., Foong R.E., Bozanich E.M., Berry L.J., Garratt L.W., Gualano R.C., Jones J.E., Dousha L.F., Zosky G.R., Sly P.D.. Sexual dimorphism in lung function responses to acute influenza A infection. Influenza Other Respir Viruses (2011.0) 5 334-342. DOI: 10.1111/j.1750-2659.2011.00236.x 111. Foong R.E., Sly P.D., Larcombe A.N., Zosky G.R.. No role for neutrophil elastase in influenza-induced cellular recruitment, cytokine production or airway hyperresponsiveness in mice. Respir. Physiol. Neurobiol. (2010.0) 173 164-170. DOI: 10.1016/j.resp.2010.08.003 112. Schneider C.A., Rasband W.S., Eliceiri K.W.. NIH Image to ImageJ: 25 years of image analysis. Nat. Methods (2012.0) 9 671-675. DOI: 10.1038/nmeth.2089 113. Larcombe A.N., Papini M.G., Chivers E.K., Berry L.J., Lucas R.M., Wyrwoll C.S.. Mouse Lung Structure and Function after Long-Term Exposure to an Atmospheric Carbon Dioxide Level Predicted by Climate Change Modeling. Environ. Health Perspect. (2021.0) 129 017001. DOI: 10.1289/EHP7305 114. James A.L., Hogg J.C., Dunn L.A., Paré P.D.. The use of the internal perimeter to compare airway size and to calculate smooth muscle shortening. Am. Rev. Respir. Dis. (1988.0) 138 136-139. DOI: 10.1164/ajrccm/138.1.136 115. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria 116. Oksanen J., Blanchet F.G., Friendly M., Kindt R., Legendre P., McGlinn D., Minchin P.R., O’Hara R.B., Simpson G.L., Solymos P.. The vegan package (versions 2.5-7). Community Ecol. Package (2020.0) 10 719 117. Cooper T.F., Berkelmans R., Ulstrup K.E., Weeks S., Radford B., Jones A.M., Doyle J., Canto M., O’Leary R.A., van Oppen M.J.. Environmental factors controlling the distribution of symbiodinium harboured by the coral Acropora millepora on the Great Barrier Reef. PLoS ONE (2011.0) 6. DOI: 10.1371/journal.pone.0025536
--- title: The Regulatory Roles of Ezh2 in Response to Lipopolysaccharide (LPS) in Macrophages and Mice with Conditional Ezh2 Deletion with LysM-Cre System authors: - Areerat Kunanopparat - Asada Leelahavanichkul - Peerapat Visitchanakun - Patipark Kueanjinda - Pornpimol Phuengmaung - Kritsanawan Sae-khow - Atsadang Boonmee - Salisa Benjaskulluecha - Tanapat Palaga - Nattiya Hirankarn journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049283 doi: 10.3390/ijms24065363 license: CC BY 4.0 --- # The Regulatory Roles of Ezh2 in Response to Lipopolysaccharide (LPS) in Macrophages and Mice with Conditional Ezh2 Deletion with LysM-Cre System ## Abstract The responses of macrophages to lipopolysaccharide (LPS) might determine the direction of clinical manifestations of sepsis, which is the immune response against severe infection. Meanwhile, the enhancer of zeste homologue 2 (Ezh2), a histone lysine methyltransferase of epigenetic regulation, might interfere with LPS response. Transcriptomic analysis on LPS-activated wild-type macrophages demonstrated an alteration of several epigenetic enzymes. Although the Ezh2-silencing macrophages (RAW264.7), using small interfering RNA (siRNA), indicated a non-different response to the control cells after a single LPS stimulation, the Ezh2-reducing cells demonstrated a less severe LPS tolerance, after two LPS stimulations, as determined by the higher supernatant TNF-α. With a single LPS stimulation, Ezh2 null (Ezh2flox/flox; LysM-Crecre/−) macrophages demonstrated lower supernatant TNF-α than Ezh2 control (Ezh2fl/fl; LysM-Cre−/−), perhaps due to an upregulation of Socs3, which is a suppressor of cytokine signaling 3, due to the loss of the Ezh2 gene. In LPS tolerance, Ezh2 null macrophages indicated higher supernatant TNF-α and IL-6 than the control, supporting an impact of the loss of the Ezh2 inhibitory gene. In parallel, Ezh2 null mice demonstrated lower serum TNF-α and IL-6 than the control mice after an LPS injection, indicating a less severe LPS-induced hyper-inflammation in Ezh2 null mice. On the other hand, there were similar serum cytokines after LPS tolerance and the non-reduction of serum cytokines after the second dose of LPS, indicating less severe LPS tolerance in Ezh2 null mice compared with control mice. In conclusion, an absence of Ezh2 in macrophages resulted in less severe LPS-induced inflammation, as indicated by low serum cytokines, with less severe LPS tolerance, as demonstrated by higher cytokine production, partly through the upregulated Socs3. ## 1. Introduction Characteristics of sepsis (a potentially life-threatening condition in response to severe infection) are surprisingly similar, regardless of the organismal causes of the infection, and sepsis from bacteria, viruses, fungi, or parasites can be fatal [1,2,3]. Sepsis-induced immune dysfunction is roughly divided into hyperinflammation and immune exhaustion (immune paralysis) [4,5], with some differences in the characteristics. While sepsis-induced hyperinflammation possibly leads to the dysfunction of several organs from hypercytokinemia, immune exhaustion causes secondary infection from inadequate inflammation to control the organism, resulting in another episode of severe infection-induced sepsis [6]. As such, direction of the clinical manifestations in sepsis in the individual patient may be determined by balance of the immune responses, which appear to occur simultaneously in the same patients [7]. A suitable immune regulation, such as an anti-inflammatory treatment during sepsis-hyperinflammation and an escalation of immune responsiveness during immune exhaustion, may be helpful [8,9,10,11,12,13,14,15]. Nevertheless, sepsis-induced immune exhaustion with secondary infection may become more common in patients with sepsis due to the improved supportive care for the maintenance of patients during hyper-inflammatory sepsis, which decreases mortality, especially in the early phase of sepsis [16]. Although several factors are mentioned as the cause of sepsis-induced immune exhaustion, including apoptotic death of several immune cells, myeloid-derived suppressor cells, and increased regulatory T cells, data on lipopolysaccharide (LPS) tolerance, which is defined as the decreased responses following secondary or prolonged LPS stimulation [17,18,19,20], are relatively more rare compared with other mechanisms. During sepsis, the translocation of LPS, which is a major molecule of Gram-negative bacteria (the most abundant gut organism), from the intestine into the blood circulation, referred to as “leaky gut”, is a common cause of endotoxemia [21,22,23]. Meanwhile, an adaptation to the prolonged LPS stimulations in sepsis may initiate LPS tolerance [24,25]. Then, LPS tolerance might, at least in part, be important in sepsis-induced immune exhaustion. Interestingly, underlying mechanisms of LPS tolerance, especially in monocytes or macrophages, are still unclear, and possibly consist of epigenetic modifications, chromatin remodeling, and interferences in cell energy status [26,27,28]. Among these, epigenetic alterations in LPS stimulation and LPS tolerance have been extensively studied [29,30]. Epigenetics, phenotypic alterations without changes in the DNA sequence, are the processes for the switch on and off of DNA transcription through DNA methylation (adding the methyl group into the DNA), histone modification, and noncoding RNA action (microRNA) [2]. DNA methylation and histone modification are processed through three groups of enzymes, including (i) the writers’ epigenetic regulation (methylation, acetylation, phosphorylation, ubiquitination), (ii) the epigenetic erasers for removing the modifications, and (iii) the epigenetic readers for the binding to different covalent modifications by the writers to mediate physiological outcomes [3]. Among all, the methylation at histone 3 lysine 27 (H3K27) is one of the most common histone changes during several cell activities, including activation by LPS [31]. After LPS stimulation, the attachment of methyl groups at lysine 27 on histone 3 (H3K27) by histone demethylase is induced in macrophages by the Polycomb repressor complex group 2 (PCR2), which belongs to the Chromobox family proteins that mediate gene silencing (a repressor of the transcriptional activity). Then, PCR2 might be associated with (i) the control of overwhelming production of cytokines and other proteins after LPS stimulation as counter anti-inflammation and/or (ii) the too-low productivity of cytokines and other molecules in LPS tolerance [32,33]. Thus, the methylation of histone (H3K27) by the PCR2 complex, which consists of several subunits, including Ezh2 (histone-lysine N-methyltransferase-2 or Enhancer of Zeste Homolog), might reduce cytokine production through the switch-off of DNA transcription [34,35]. Because Ezh is an important catalytic subunit for the methylation process, the over-expression of Ezh enhances PCR2 function, leading to anti-inflammatory properties [36,37] and the blockage of Ezh enhances pro-inflammatory responses [38]. Interestingly, the Ezh2 inhibitor has been used as an anti-cancer drug (Tazemetostat) to promote the tumoricidal immune response and is currently approved for the treatment of follicular lymphoma and epithelioid sarcoma [39]. Likewise, the downregulated Ezh2 enhanced the severity of a colitis mouse model through the facilitated nuclear factor kappa B (NF-κB) and tumor necrosis factor-alpha (TNF-α) [40]. However, there are some reports on reduced inflammatory responses by Ezh2 blockage through the upregulation of Suppressor of cytokine signaling 3 (Socs3) and the inhibition of Janus kinase/signal transducer and activator of transcription (JAK/STAT) pathway [41] which has been used to reduce the severity of atherosclerosis [42]. Hence, the impacts of Ezh2 on inflammation and sepsis are still inconclusive. Moreover, Ezh2 was one of the upregulated genes in macrophages with LPS tolerance found in our previous publication [43]. The screening of epigenetic inhibitors also demonstrated that the Ezh2 inhibitor enhances TNF-α expression in the LPS-tolerant macrophages [44]. Thus, the control of macrophage responses through the manipulations of epigenetics is interesting for controlling sepsis immune responses [13,45]. Here, the influence of Ezh2 on LPS was explored in both a single LPS activation (hyper-inflammatory responses) and after two stimulations (LPS tolerance), in vitro and in vivo, using conditional Ezh2 deletion mice with a LysM-Cre system, which selectively affected Ezh2 only in macrophages. ## 2.1. An Influence of Epigenetic Alteration in LPS-Activated Macrophages, the Transcriptomic Analysis, and siRNA Experiments A correlation between the epigenetic alteration and macrophage responses against LPS was evaluated through transcriptomic analysis. The differences between the bone-marrow-derived macrophages of wild-type (WT) mice with a single LPS stimulation and media control were demonstrated by the heat map graphic pattern (Figure 1A). The analysis of differentially expressed genes (DEG) indicated some genes with significant differences or a tendency of difference in epigenetic alteration between LPS-activated macrophages and the control group. First, there was a higher lysine deacetylase (epigenetic eraser) in LPS-stimulated macrophages, including Hdac1 (histone deacetylase), Kdm3a (lysine demethylase 3A), and Kdm6b (lysine demethylase 6B), with a tendency of the upregulated Sirt6 (sirtuin 6) (Figure 1B). Second, the lysine methyltransferase (epigenetic writer) was higher in LPS-stimulated cells with a prominent upregulated Kmt5a (lysine methyltransferase 5A) and an upregulated trend for Ezh1 (histone-lysine N-methyltransferase-1) and Ezh2 (histone-lysine N-methyltransferase-2) (Figure 1C). Third, there was downregulated serine-threonine/tyrosine protein kinase (epigenetic writer) after LPS stimulation with the decreased Aurkb (aurora kinase B) and a tendency for downregulation of Aurkc (aurora kinase C) (Figure 1D). Fourth, there was a trend of enhanced lysine ubiquitin ligase (epigenetic writer) and Rnf2 (ring finger protein 2) in LPS-activated cells (Figure 1E). These data supported a correlation between epigenetic alteration through these enzymes with the macrophage responses against LPS. ## 2.2. The Reduced TNF-a and an Elevation of TNF-α and IL-6 in Ezh2 Null Macrophages with a Single LPS Stimulation and LPS Tolerance, Respectively Due to the frequent alteration at H3K27 in activated macrophages which possibly be correlated with Ezh2 and PCR2 [31], the associations between Ezh2 with one and two LPS stimulations were initially explored through the silencing of Ezh2 gene by siRNA (Ezh2 siRNA) on the macrophage cell line (RAW264.7) (Figure 2A). Notably, the difference between cytokine responses between one and two LPS stimulations was determined 2 days after the incubation (Figure 2A) to control the duration of culture in both groups. There is a difference in the duration of culture in the comparison between cytokines after the first and the second LPS stimulations, which are 1- and 2-days incubation periods, respectively. In a single LPS stimulation (N/LPS), the silencing of Ezh2, an inhibitory gene against cytokine production, in macrophages did not significantly change supernatant cytokines, including TNF-α, IL-6, and IL-10, and macrophage polarization (Figure 2B–I). The M1 polarization-associated genes, including interleukin-1β (IL-1β) and inducible nitric oxide synthase (iNOS) (Figure 2B–D), and M2 polarization-associated markers, including resistin-like molecule-1 (Fizz-1), arginase-1 (Arg-1), and transforming growth factor-β (TGF-β) (Figure 2E–I), were evaluated. In LPS tolerance (LPS/LPS), there was a reduction in supernatant cytokines (TNF-α and IL-6 but not IL-10) in both Ezh siRNA macrophages and the control when compared with the N/LPS stimulation, which supported the main characteristics of LPS tolerance as previously described [24,46,47]. However, there were higher TNF-α and IL-6 and lower IL-10 levels in Ezh2-silencing macrophages with LPS tolerance when compared with LPS tolerance in the control cells (Figure 2B–D). These data support the impact of the removal of an inhibitory gene (Ezh2), with similar markers on macrophage polarization (Figure 2E–I). Notably, the impact of LPS on cytokine production disappeared 48 h after the incubation, as all supernatant cytokines (TNF-α, IL-6, and IL-10) in macrophages activated with LPS followed by media control (LPS/N) protocol were similar to the control level (Supplementary Figure S1B–D). #### LysM-Crecrecre/−) Mice after a Single LPS Stimulation and LPS Tolerance To investigate the impact of Ezh2 on LPS stimulation, macrophages from mice with conditional Ezh2 deletion using the LysM-Cre system were used (Figure 3A). Similar to the Ezh2 silencing with siRNA, the severity of LPS tolerance of macrophages from Ezh2 null mice (Ezhfl/fl; LysM-Crecre/−) was lower than LPS tolerance in the litter mate control (Ezhfl/fl; LysM-Cre−/−) (Ezh2 control) as indicated by the higher supernatant TNF-α and IL-6 (but not IL-10) in macrophages from Ezh2 null mice compared with the cells from Ezh2 control (Figure 3B–D). The characteristics of LPS tolerance, as indicated by a lower cytokine production in LPS tolerance (LPS/LPS) compared with N/LPS, in macrophages from Ezh2 null mice were demonstrated by reduced pro-inflammatory cytokines (TNF-α and IL-6) in the supernatant, while LPS tolerance in the cells from Ezh2 control was indicated only by the lower IL-6 level (Figure 3B,C). Surprisingly, in a single LPS stimulation (N/LPS), supernatant IL-10 (an anti-inflammatory cytokine) was the only elevated cytokine in macrophages from Ezh2 null mice (Figure 3D), supporting a possible less severe LPS tolerance from an absence of an inhibitory Ezh2 gene [48]. In contrast, the supernatant TNF-α in Ezh2 null macrophages with N/LPS protocol was lower than the cells from Ezh2 control (Figure 3B), perhaps due to other inhibitory factors against the cytokine production. However, supernatant IL-6 from N/LPS-stimulated Ezh2 null macrophages was not different from N/LPS-activated Ezh2 control macrophages (Figure 3C), implying a lower impact of Ezh2 siRNA in the inhibition of IL-6 compared with the reduction in TNF-α (Figure 3B). ## 2.4. A Possible Impact of Inhibitory Socs3 and Cell Energy Status in Ezh2 Null Macrophages with a Single LPS Stimulation and LPS Tolerance Among several genes, Ezh1, Ezh2, and suppressor of cytokine signaling 3 (Socs3) were explored due to the possible inhibitory effect on inflammatory responses [49,50]. As such, Ezh1 was downregulated in both N/LPS and LPS tolerance (LPS/LPS) in Ezh2 null and control macrophages (Figure 4A). In control macrophages (Ezh2 control), both N/LPS and LPS/LPS similarly upregulated Ezh2 and Socs3 when compared with the stimulation with media control (N/N) (Figure 4B,C). Meanwhile, in Ezh2 null macrophages, there was a higher Socs3 upregulation after both N/LPS and LPS tolerance compared with the activations in Ezh2 control cells (Figure 4B,C). Additionally, Socs3 upregulation in Ezh2 null macrophages with N/LPS was higher than the activation by LPS tolerance (Figure 4C), implying a possibly more potent inhibition by Socs3 after a single LPS stimulation, perhaps to neutralize LPS-induced hyper-inflammation from an absence of the Ezh2 gene. Indeed, the Ezh2 gene also controls Socs3 expression, as the presence of Ezh2 inhibits Socs3 expression, as previously mentioned [50]. Because of the association between the energy status in macrophages with several activities, especially inflammatory responses [23,51,52,53,54], the extracellular flux analysis between Ezh2 null macrophages and control cells was also explored. However, there was no statistical difference in the cell energy state (glycolysis and mitochondrial activities) among all groups of the experiments (Figure 4D,E). Additionally, the baseline of cell energy status between Ezh2 null macrophages and control cells was similar, with a tendency for lower mitochondrial activities after activation by both N/LPS and LPS tolerance (LPS/LPS) in both groups (Figure 4D). Thus, Ezh2 seems to have less impact on the cell energy status, despite several impacts on cytokine production. ## 2.5. A Less Pro-Inflammatory Response to LPS in Ezh2 Null Mice over the Control Mice, a Possible New Strategy against Sepsis-Induced Hyper-Inflammation by Ezh2 Interference Due to the impacts of Ezh2-manipulated macrophages toward activation by a single LPS stimulation (N/LPS) and LPS tolerance (LPS/LPS), we further explored this in Ezh2 null (Ezhfl/fl; LysM-Crecre/−) and Ezh2 control (Ezhfl/fl; LysM-Cre−/−) mice (Figure 5A). Serum TNF-α and IL-6, but not IL-10, was lower in Ezh2 null mice compared with the Ezh2 control group (Figure 5B–D), supporting impacts of the lower supernatant cytokines in LPS-activated Ezh2 null macrophages compared to the control cells (Figure 3B). In LPS tolerance, the characteristics of lower serum cytokines after the second dose of LPS were demonstrated by both TNF-α and IL-6 in the control mice, as demonstrated in the open circles and open square that represented a single LPS stimulation and LPS tolerance, respectively (Figure 5B,C). Meanwhile, the LPS tolerance-induced lower cytokine production was demonstrated only by decreased serum IL-6 (but not TNF-α) in the Ezh2 null mice as indicated by the blue circle (a single LPS stimulation) and red square (LPS tolerance) (Figure 5B,C). These data implied a less severe LPS tolerance in the Ezh2 null group than in the control mice. Notably, the characteristic of LPS tolerance could not be demonstrated by serum IL-10 in all mouse strains (Figure 5D). Thus, the absence of Ezh2 in mice was beneficial for anti-inflammation after LPS stimulation and in LPS tolerance when compared with the control mice. ## 3.1. Epigenetic Regulation in LPS Stimulation, an Interesting Strategy in Immune Response Manipulation for Sepsis The presence of lipopolysaccharide (LPS), a microbial molecule from Gram-negative bacteria, in blood has been clinically demonstrated as endotoxemia in several conditions, including sepsis [55,56,57], partly through gut barrier damage [1,22,24,58]. The activation of macrophages by LPS is possibly important in sepsis because macrophages are the innate immune cells responsible for the recognition of foreign molecules [51,59]. Among several alterations in LPS-stimulated macrophages, the epigenetic modifications, especially DNA methylation at the cytosine-phosphate-guanine (CpG) sites and the methylation at the N-terminal tails of histones, are critical regulators of chromatin structure that determine gene expression for responses, differentiation, and proliferation [60]. In epigenetic-induced histone modification, several key enzymes control chromatin accessibility by regulating methyl and acetyl marks at the tail of histone H3 by either installing (writers) or removing (erasers) these marks at histone. Although gene repression by the presence of trimethylation of H3K27 (H3K27me3) at promoter regions is well known in cancer [61], data on the influence of H3K27me3 in sepsis are still scarce [62]. After LPS stimulation, seven lysine methyltransferases, enzymes for the manipulation of methyl groups at lysine on histone protein as writers (adding the proteins) and erasers (removing these chemical tags), were increased, while only a few enzymes of other epigenetic processes were elevated. Perhaps, the methylation in LPS-activated macrophages might be more common than other epigenetic alteration processes that use lysine methylation (H3K27Me) to safeguard for the overwhelming production of inflammatory cytokines (immune hyper-responsiveness) during sepsis [33]. Between the key repressive and activating methyl mark at H3K27 by writers (Ezh1/Ezh2) and erasers (Kmt5a/Kmt5b) [35,63], respectively, Ezh1, Ezh2, and Kmt5a are in the lists of our data, highlighting the importance of lysine methylation in macrophage response to LPS. Indeed, the increased expression of Ezh2 and H3K27 is demonstrated in the circulation of patients with sepsis, which is correlated with increased mortality [64,65]. Although Ezh2 showed only a trend to be upregulated after LPS stimulation here, the clinical availability of Ezh2 inhibitors in the treatment of cancer [66] makes Ezh2 an interesting target with an easy clinical translation. Theoretically, the presence of Ezh2 induces gene repression (anti-inflammation), and the blockage of Ezh2 might enhance inflammation that is possibly beneficial for LPS tolerance but worsens LPS-induced hyper-inflammation. Notably, the reduced inflammation in LPS tolerance might be too low and is inadequate for the proper inflammation necessary for microbial control. The attenuation of LPS tolerance might reduce the secondary infection [67,68,69]. ## 3.2. Impact of Ezh2 in Sepsis-Hyper Inflammation and Immune Exhaustion Although Ezh2 blockage might theoretically enhance pro-inflammation, a few reports using Ezh2 inhibitors and the conditioning of Ezh2-deleted mice in pneumococcal sepsis demonstrate beneficial impacts as an anti-inflammatory molecule, partly through the upregulation of Suppressor of Cytokine Signaling 3 (Socs3) gene [64,70]. In Ezh2-deficient macrophages, the reduced suppressive H3K27Me3 marks at the (Socs3) transcriptional start site (and distal enhancer) result in upregulated Socs3, despite the possibly more effective transcription of cytokine-producing genes due to the loss of methylation blockage of DNA reading by H3K27Me (Figure 6). As such, the cytosolic Socs3 inhibits the TLR-induced MyD88–TRAF6–NF-κB signaling pathway, partly through the enhanced ubiquitination and proteasomal degradation that suppresses the NF-κB–dependent inflammatory genes [35,49]. Likewise, Ezh2 inhibitors have also attenuated sepsis-induced intestinal disorders, multiple sclerosis, and glucose-activated peritoneal fibrosis in previous reports [50,62,71]. In contrast, the suppression of Ezh2 worsens inflammatory bowel diseases and sepsis-induced muscle cell apoptosis [40,72,73,74], perhaps through the more effective pro-inflammatory cytokine production after the loss of DNA reading blockage by methylation (H3K27Me). Nevertheless, it is possible that H3K27Me is responsible for the control of some groups of cytokines more than other groups. For example, we demonstrated that LPS-activated Ezh2 null macrophage induced higher IL-10, and lower TNF-a compared with the LPS-stimulated WT cells (Figure 3C). Hence, the impacts of Ezh2 in sepsis are still inconclusive, and an exploration of the influence of Ezh2 with LPS tolerance has never been carried out. Here, we supported an enhanced Socs3 expression with the absence of Ezh2 in both a single LPS stimulation (LPS response; N/LPS) and double LPS activation (LPS tolerance; LPS/LPS) using macrophages from Ezh2 null mice (Ezhfl/fl; LysM-Crecre/−) versus the control mice (Ezhfl/fl; LysM-Cre−/−) (Figure 4C). The upregulated Socs3 was more prominent in the LPS tolerance of Ezh2 null macrophages than in the WT cells, possibly due to differences in the H3K27Me abundance between a single versus the repeat LPS stimulation (LPS tolerance). Notably, Ezh2 silencing by siRNA in RAW264.7 cells did not demonstrate anti-inflammation after N/LPS, which was different from the N/LPS in Ezh2 null macrophages, possibly due to the limitation of the gene silencing by siRNA with the non-complete silencing of the interested genes. In LPS tolerance, there was an increase in cytokines in both Ezh2 silencing by siRNA and Ezh2 null macrophages, highlighting an improved severity of LPS tolerance despite a possible non-complete Ezh2 silencing of the siRNA process. With an absence of the Ezh2 gene, there was an IL10 upregulation in single LPS response (N/LPS) but not in LPS tolerance (LPS/LPS), which might be correlated with the more prominent Socs3 expression in the N/LPS group compared with LPS tolerance. Indeed, Socs3 might be correlated with IL-10 because (i) Ezh2 inhibits Socs3 expression and blockage of Ezh2 upregulates Socs3 [50], (ii) IL10 directly upregulates Socs3 expression in LPS-stimulated macrophages [75], and (iii) there is more prominent Socs3 upregulation together with supernatant IL-10 in N/LPS than LPS tolerance (LPS/LPS) of Ezh2 null macrophages (Figure 3D and Figure 4C). Additionally, Socs3 activity might require IL10 to inhibit the inflammatory responses, as indicated by severe impairment of Socs3 and IL10 in macrophages with single LPS responses [76,77], upregulated IL-10 in LPS tolerance macrophages [78,79,80,81], and altered IL-10 in LPS-stimulated macrophages with an Ezh2 inhibitor (EPZ-6438) [82]. Moreover, reduced pro-inflammatory cytokines (TNF-α and IL-6) (Figure 3B,C) with an escalation in Socs3 and IL-10 in Ezh2 null N/LPS macrophages (Figure 3D and Figure 4C) also supported a linkage among Ezh2, Socs3, and inflammatory responses. However, Socs3 can be regulated by both anti-inflammatory and pro-inflammatory cytokines [83], which is possibly driven by specific cytokines [84]. More studies on this topic are needed. Hence, upregulated Socs3 might be responsible for the anti-inflammatory direction of Ezh2 null macrophages and mice after a single LPS stimulation. In LPS tolerance, the loss of inhibitory Ezh2 in Ezh2 null macrophages resulted in an elevation of both cytokine-producing genes, such as NF-κB, and Socs3. However, Socs3 upregulation in the tolerance macrophages was not as prominent as the single LPS-stimulated macrophages leading to a lower Sosc3-inhibitory effect on cytokine production and higher cytokines in Ezh2 null macrophages than control cells after LPS tolerance. Perhaps the balance between the pro-inflammatory NF-κB-dependent pathways and the counteracting Socs3-STAT3 anti-inflammation after TLR-4 activation [85] is the natural control of hyper-inflammation (Figure 6), and the absence of Ezh2 tips the balance of the signaling by inducing the higher effective Socs3 than NF-κB, leading to the prominent Sosc3-inhibitory effect, resulting in an anti-inflammatory stage. Because (i) COX-2 promotor is a CpG island, which is commonly found in DNA methylation, (ii) LPS enhanced COX-2 expression in both mRNA and protein levels [86,87], and (iii) Ezh2 inhibition suppresses COX2 via Sosc3/STAT3 in macrophages and microglial cells [88,89]; with the well-known Socs3 and STAT3 correlation [90], the absence of Ezh2 might convert the Socs3-STAT3 function from a pro- to an anti-inflammatory phase in LPS tolerance. In LPS-activated mice (N/LPS), the anti-inflammatory effect of Ezh2 depletion only in macrophages by the Cre-LoxP system [91] was demonstrated through the reduced serum TNF-α and IL-6 with relatively high serum IL-10 compared with the control mice. These data support that serum cytokine, in response to an LPS injection, was mainly produced from macrophages [92]. Due to lower levels of pro-inflammatory cytokines after LPS injection, the severity of LPS tolerance, as indicated by the difference between the first and second dose of LPS, in Ezh2 null mice was lower than in the control. While the lower serum TNF-α and IL-6 levels were very obvious in the control mice with LPS tolerance, the relatively high serum TNF-α in LPS tolerance-stimulated Ezh2 null mice was demonstrated through the non-different serum TNF-α between the first and second dose of LPS, indicating a less severe LPS tolerance. Despite the elevated supernatant IL-10 in Ezh2 null macrophages with LPS tolerance, serum IL-10 was not increased in Ezh2 null LPS tolerance mice, possibly due to an impact of LPS tolerance in other cells as IL-10 could be produced by innate immune cells, adaptive immune cells, and organ parenchymal cells [93,94,95,96]. Interestingly, the depletion of Ezh2 in macrophages not only protected the mice from too high pro-inflammatory septic shock but, on the other hand, also safeguarded the mice from prominent LPS tolerance (fewer pro-inflammatory cytokines) which possibly correlated with better control on the secondary infection through the prevention of the microbial spread by an appropriate inflammation [97]. ## 3.3. Clinical Aspect and Future Experiments Although the sepsis immune responses are theoretically crudely divided into hyper-inflammation and immune exhaustion, the identification of the tips of the balance between these responses by several sophisticated biomarkers might be the future of sepsis immune modulation. For example, high serum IL-6 and IL-1 might be biomarkers for sepsis hyper-inflammation [98,99], while downregulated HLA-DR and viral reactivation, such as cytomegalovirus, which is the common dormant virus in the human host, possibly indicate sepsis immune exhaustion [100,101]. However, the use of drugs that are beneficial in both sepsis responses (hyper- and anti-inflammation) will be more convenient. From our data on mice with conditional Ezh2 deletion, the blockage of Ezh2 might be one of the interesting drugs that are beneficial in both hyper-inflammation and immune exhaustion; however, reduced immune responses by Ezh2 inhibition possibly did not induce immune exhaustion. Hence, Ezh2 blockage can be conveniently administered in both sepsis immune responses and is also clinically available [66]. The extended indication of Ezh2 blockage in sepsis, in addition to cancer treatment, might be beneficial. Nevertheless, LPS tolerance is only a subset of the sepsis-induced immune exhaustion [20] and the evaluation of Ezh2 blockage in sepsis is still scarce. Further experiments on the influence of Ezh2 deletion and Ezh2 inhibitors in sepsis are warranted. Finally, there were several limitations in the current study. First, the results are from a limited number of mice in a proof-of-concept study and additional research is required to reach a solid conclusion. Second, the different ages and genders of the mice may affect the results and the experiments using female mice and/or different ages might result in a different conclusion. Third, additional information on the mechanisms behind the impact of Ezh2 is required to comprehend the influence on sepsis and LPS activation. Despite these various limitations, our data support the influence of Ezh2 in sepsis. ## 4.1. Macrophage Cell Line and Small Interfering RNA (siRNA) Murine macrophage-like cells (RAW264.7; TIB-71), purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA), were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; Cytiva HyClone) supplemented with $10\%$ Fetal Bovine Serum (FBS) in a humidified incubator at 37 °C with $5\%$ CO2. To initially explore an importance of Ezh2 in responses against LPS, the small interfering RNA (siRNA) on Ezh2 was evaluated. Briefly, RAW264.7 at 106 cells/mL was seeded into 6-well plates and incubated overnight at 37 °C with $5\%$ CO2. Then, the siRNA for Ezh2 (DharmaconTM AccellTM, Horizon Discovery, Watwebeach, England, UK) was prepared with siRNA buffer before adding into the cells after removal of the cell culture media (the final concentration at 1 μM siRNA per well) at 37 °C with $5\%$ CO2 for 48 h. The non-targeting pool siRNA (DharmaconTM AccellTM) was used as a control. Then, macrophages with Ezh2-siRNA and non-siRNA were activated by 3 different protocols, including (i) a single LPS stimulation, beginning with DMEM followed by LPS (100 ng/mL) 24 h later (N/LPS), or (ii) LPS tolerance, using the double stimulations by 100 ng/mL of LPS (LPS/LPS), or control (N/N) using the double DMEM incubation, before the sample collection (supernatant and cells). Notably, the supernatant of the stimulated cells in all groups was gently removed, and washed with DMEM, before the re-administration of LPS or DMEM as mentioned in previous publications [24,102,103]. Supernatant cytokines (TNF-α, IL-6, and IL-10) were evaluated by ELISA (Invitrogen, Carlsbad, CA, USA) and the gene expression was evaluated by quantitative real-time polymerase chain reaction (PCR) as previously described [104,105,106,107]. Briefly, the RNA was extracted from the cells with TRIzol Reagent (Invitrogen, Carlsbad, CA, USA) together with RNeasy Mini Kit (Qiagen, Hilden, Germany) as 1 mg of total RNA was used for cDNA synthesis with iScript reverse transcription supermix (Bio-Rad, Hercules, CA, USA). Quantitative real-time PCR was performed on a QuantStudio 5 real-time PCR system (Thermo Fisher Scientific, Waltham, MA, USA) using SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA). Expression values were normalized to Beta-actin (β-actin) as an endogenous housekeeping gene and the fold change was calculated by the ∆∆Ct method. The primers used in this study are listed in Table 1. ## 4.2. Animal and Animal Model The approved protocol (No. $\frac{017}{2562}$) by the Institutional Animal Care and Use Committee of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand according to the National Institutes of Health (NIH) criteria were used. Here, 8-week-old male wild-type (WT) C57BL/6 mice were purchased from Nomura Siam, Pathumwan, Bangkok, Thailand. In parallel, Ezh2flox/flox and LyM-CreCre/Cre mice were obtained from RIKEN BRC Experimental Animal Division (Ibaraki, Japan) and cross-bred until we reached an Ezh2 littermate control (Ezhfl/fl; LysM-Cre−/−) or Ezh2 null (Ezhfl/fl; LysM-Crecre/−) in the F3 generation of the breeding protocol. The Ezh2flox/flox mice have loxP sites upstream and downstream of the 2.7 kb SET domain, and were bred with LysM-CreCre/Cre mice, the mice with a cre recombinase under the control of lysozyme M to target Ezh2 for deletion in myeloid cells (macrophages and neutrophils). The offspring were either Ezh2flox/flox with no LysM-Cre (Ezhfl/fl; LysM-Cre−/−), referred to as “the littermate controls or Ezh2 control”, or were positive for the Cre driver (Ezh2 null or Ezhfl/fl; LysM-Crecre/−). For LPS activation, the conditional targeted Cre-positive mice (Ezh2 null) were gender- and age-matched with floxed/floxed male littermate controls (Ezh2 control) aged 8–10 weeks old. To genotype these mice on the loxP sites insertion, the following primers were used for Ezh2: reverse 1: 3′ of loxp: 5′-AGG GCA TCA GCC TGG CTGTA-3′; Forward 2: 5′ of loxp: 5′-TTA TTC ATA GAG CCA CCTGG-3′; Forward 3: left loxp: 5-ACG AAA CAG CTC CAG ATTCAG GG-3′ according to a previous publication [70]. The mice homozygous for the flox were selected and genotyped for the expression of LysM-Cre using the primers; Forward: 5′-CTTGGGCTGCCAGAATTCTC-3′; Reverse: 5′CCCAGAAATGCCA GATTACG-3′. Then, the mice were divided into 3 groups, including (i) LPS tolerance (LPS/LPS), intraperitoneal injection of 0.8 mg/kg LPS (Escherichia coli 026:B6) (Sigma-Aldrich, St. Louis, MO, USA) with another dose of 4 mg/kg LPS 48 h later, (ii) a single LPS stimulation, intraperitoneal phosphate buffer solution (PBS) injection followed by LPS (4 mg/kg) 48 h later, and (iii) control (N/N), double intraperitoneal PBS injection with 48 h duration between the dose. Notably, there was a lower dose of the first LPS administration (0.8 mg/kg) compared with the second dose (4 mg/kg) in the LPS tolerance protocol because the higher LPS in the first administration might result in the sustained elevated serum cytokines at 24 h of the first dose (before the 2nd dose LPS) which might interfere with the interpretation. After these protocols, blood was collected through (i) tail vein nicking at 1 and 3 h after the last injection and (ii) cardiac puncture under isoflurane anesthesia at 6 h of the protocol. Then, serum cytokines were evaluated by ELISA (Invitrogen, Carlsbad, CA, USA). ## 4.3. Bone-Marrow-Derived Macrophages and the Transcriptome Analysis The RNA sequencing analysis was performed to determine the influence of epigenetic alteration in LPS-activated macrophages. As such, bone-marrow-derived macrophages were prepared from the femurs of wild-type (WT) mice using supplemented Dulbecco’s Modified Eagle’s Medium (DMEM) with a $20\%$ conditioned medium of the L929 cells (ATCC CCL-1), which are fibroblasts used as a source of macrophage colony-stimulating factor as previously published [51,52,54,107]. Then, the macrophages at 5 × 104 cells/well in supplemented DMEM (Thermo Fisher Scientific) were incubated in $5\%$ carbon dioxide (CO2) at 37 °C for 24 h before being treated by LPS stimulation (100 mg/mL), and control DMEM for 24 h. For transcriptome analysis, the RNA from macrophages was extracted by RNeasy mini kit (Qiagen) and processing with the RNA sequencing of BGISEQ-50 platform based on triplicate macrophage samples as previously published [108]. The sequencing quality was determined using FastQC, and the raw sequencing reads were mapped and aligned against Mus musculus reference genome GRCm39 using STAR [109], followed by gene quantification against the reference mouse transcriptome using Kallisto [110]. Read counts were normalized and analyzed for differentially expressed genes (DEGs) using edgeR package [111] and limma-voom package [112,113]. Genes were considered significant expressions (p-value < 0.05) when the log2 value of fold change of expression was less than −2 or greater than 2, indicating down- or upregulation, respectively. Clustering of DEGs was performed based on Euclidean distance and the Ward. D2 method using the ComplexHeatmap package version 2.12.1 [114]. Expression levels in log2 (TPM; transcript count per million) of selected epigenetic-related genes [115] were compared between untreated and LPS-treated groups to determine statistical significance using the Wilcoxon test in ggpubr package [116], where a p-value less than 0.05 indicates statistical significance. ## 4.4. Bone-Marrow-Derived Macrophages and Extracellular Flux Analysis Bone-marrow-derived macrophages from the Ezh2 control (Ezhfl/fl; LysM-Cre−/−) or Ezh2 null (Ezhfl/fl; LysM-Crecre/−) mice were extracted from femurs before activation by N/LPS (single LPS stimulation), LPS/LPS (LPS tolerance), or N/N (control), similar to the protocol for siRNA macrophages, and we measured supernatant cytokines (TNF-α, IL-6, and IL-10) and gene expression for (Ezh1, Ezh2, and Socs3) by PCR as mentioned above. Additionally, the seahorse XFp Analyzers (Agilent, Santa Clara, CA, USA) were used to determine the cell energy status (extracellular flux analysis), with oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) representing mitochondrial function (respiration) and glycolysis activity, respectively, following previous publications [23,53,108,117,118]. Briefly, the macrophages (1 × 105 cells/well) at 24 h after the stimulations (N/N, N/LPS, and LPS/LPS) were incubated in Seahorse media (DMEM complemented with glucose, pyruvate, and L-glutamine) (Agilent, 103575–100) before activation by different metabolic interference compounds such as oligomycin, carbonyl cyanide-4-(trifluoromethoxy)-phenylhydrazone (FCCP), and rotenone/antimycin A for OCR evaluation. In parallel, glucose, oligomycin, and 2-Deoxy-d-glucose (2-DG) were used for ECAR measurement. The graphs of OCR and ECAR were demonstrated. ## 4.5. Statistical Analysis The results are shown in mean ± S.E.M. All data were analyzed with GraphPad Prism version 6. Student’s t-test or one-way analysis of variance (ANOVA) with Tukey’s comparison test was used for the analysis of experiments with two and more than two groups, respectively. For all data sets, a p-value less than 0.05 was considered significant. ## 5. Conclusions The transcriptomic analysis of LPS activation in wild-type macrophages demonstrated the correlation between several enzymes of epigenetic processes and the responses to sepsis. The reduced supernatant pro-inflammatory cytokines with a single LPS stimulation and the less severe LPS tolerance, as indicated by the lower differences in supernatant cytokines after the first and second dose of LPS, using a double LPS stimulation in Ezh2 null (Ezhfl/fl; LysM-Crecre/−) macrophages were compared with the control cell (Ezhfl/fl; LysM-Cre−/−). These data supported the possible benefits of Ezh2 blockage in both acute responses against LPS stimulation and LPS tolerance after several LPS activations in macrophages. Likewise, the less severe hyper-inflammation and LPS tolerance after a single and double LPS injection, respectively, in Ezh2 null mice over the control mice also support the benefits of interference of Ezh2 during sepsis. More studies in other animal models and/or in patients with sepsis using clinically available drugs are encouraged. ## References 1. Amornphimoltham P., Yuen P.S.T., Star R.A., Leelahavanichkul A.. Gut Leakage of Fungal-Derived Inflammatory Mediators: Part of a Gut-Liver-Kidney Axis in Bacterial Sepsis. Dig. Dis. Sci. (2019) 64 2416-2428. DOI: 10.1007/s10620-019-05581-y 2. Lacal I., Ventura R.. Epigenetic Inheritance: Concepts, Mechanisms and Perspectives. Front. Mol. Neurosci. (2018) 11 292. DOI: 10.3389/fnmol.2018.00292 3. Han M., Jia L., Lv W., Wang L., Cui W.. Epigenetic Enzyme Mutations: Role in Tumorigenesis and Molecular Inhibitors. Front. Oncol. (2019) 9 194. DOI: 10.3389/fonc.2019.00194 4. Hotchkiss R.S., Moldawer L.L., Opal S.M., Reinhart K., Turnbull I.R., Vincent J.-L.. Sepsis and septic shock. Nat. Rev. Dis. Prim. (2016) 2 16045. DOI: 10.1038/nrdp.2016.45 5. Liu D., Huang S.-Y., Sun J.-H., Zhang H.-C., Cai Q.-L., Gao C., Li L., Cao J., Xu F., Zhou Y.. Sepsis-induced immunosuppression: Mechanisms, diagnosis and current treatment options. Mil. Med. Res. (2022) 9 56. DOI: 10.1186/s40779-022-00422-y 6. Brady J., Horie S., Laffey J.G.. Role of the adaptive immune response in sepsis. Intensive Care Med. Exp. (2020) 8 20. DOI: 10.1186/s40635-020-00309-z 7. Delano M.J., Ward P.A.. Sepsis-induced immune dysfunction: Can immune therapies reduce mortality?. J. Clin. Investig. (2016) 126 23-31. DOI: 10.1172/JCI82224 8. Németh K., Leelahavanichkul A., Yuen P.S.T., Mayer B., Parmelee A., Doi K., Robey P.G., Leelahavanichkul K., Koller B.H., Brown J.M.. Bone marrow stromal cells attenuate sepsis via prostaglandin E(2)-dependent reprogramming of host macrophages to increase their interleukin-10 production. Nat. Med. (2009) 15 42-49. DOI: 10.1038/nm.1905 9. Leelahavanichkul A., Yasuda H., Doi X., Hu X., Zhou H., Yuen P.S.T., Star R.A.. Methyl-2-acetamidoacrylate, an ethyl pyruvate analog, decreases sepsis-induced acute kidney injury in mice. Am. J. Physiol. Renal. Physiol. (2008) 295 F1825-F1835. DOI: 10.1152/ajprenal.90442.2008 10. Taratummarat S., Sangphech N., Vu C.T.B., Palaga T., Ondee T., Surawut S., Sereemaspun A., Ritprajak P., Leelahavanichkul A.. Gold nanoparticles attenuates bacterial sepsis in cecal ligation and puncture mouse model through the induction of M2 macrophage polarization. BMC Microbiol. (2018) 18. DOI: 10.1186/s12866-018-1227-3 11. Panpetch W., Chancharoenthana W., Bootdee K., Nilgate S., Finkelman M., Tumwasorn S., Leelahavanichkul A.. Lactobacillus rhamnosus L34 Attenuates Gut Translocation-Induced Bacterial Sepsis in Murine Models of Leaky Gut. Infect. Immun. (2018) 86 e00700-17. DOI: 10.1128/IAI.00700-17 12. Issara-Amphorn J., Chancharoenthana W., Visitchanakun P., Leelahavanichkul A.. Syk Inhibitor Attenuates Polymicrobial Sepsis in FcgRIIb-Deficient Lupus Mouse Model, the Impact of Lupus Characteristics in Sepsis. J. Innate Immun. (2020) 12 461-479. DOI: 10.1159/000509111 13. Dang C.P., Leelahavanichkul A.. Over-expression of miR-223 induces M2 macrophage through glycolysis alteration and attenuates LPS-induced sepsis mouse model, the cell-based therapy in sepsis. PLoS ONE (2020) 15. DOI: 10.1371/journal.pone.0236038 14. Chancharoenthana W., Udompronpitak K., Manochantr Y., Kantagowit P., Kaewkanha P., Issara-Amphorn J., Leelahavanichkul A.. Repurposing of High-Dose Erythropoietin as a Potential Drug Attenuates Sepsis in Preconditioning Renal Injury. Cells (2021) 10. DOI: 10.3390/cells10113133 15. Dang C.P., Issara-Amphorn J., Charoensappakit A., Udompornpitak K., Bhunyakarnjanarat T., Saisorn W., Sae-Khow K., Leelahavanichkul A.. BAM15, a Mitochondrial Uncoupling Agent, Attenuates Inflammation in the LPS Injection Mouse Model: An Adjunctive Anti-Inflammation on Macrophages and Hepatocytes. J. Innate Immun. (2021) 13 359-375. DOI: 10.1159/000516348 16. Perner A., Rhodes A., Venkatesh B., Angus D.C., Martin-Loeches I., Preiser C.-J., Vincent J.-L., Marshall J., Reinhart K., Joannidis M.. Sepsis: Frontiers in supportive care, organisation and research. Intensive Care Med. (2017) 43 496-508. DOI: 10.1007/s00134-017-4677-4 17. Mithal L.B., Arshad M., Swigart L.R., Khanolkar A., Ahmed A., Coates B.M.. Mechanisms and modulation of sepsis-induced immune dysfunction in children. Pediatr. Res. (2022) 91 447-453. DOI: 10.1038/s41390-021-01879-8 18. Schrijver I.T., Théroude C., Roger T.. Myeloid-Derived Suppressor Cells in Sepsis. Front. Immunol. (2019) 10 327. DOI: 10.3389/fimmu.2019.00327 19. Cao C., Ma T., Chai Y.-F., Shou S.-t.. The role of regulatory T cells in immune dysfunction during sepsis. World J. Emerg. Med. (2015) 6 5-9. DOI: 10.5847/wjem.j.1920-8642.2015.01.001 20. Vergadi E., Vaporidi K., Tsatsanis C.. Regulation of Endotoxin Tolerance and Compensatory Anti-inflammatory Response Syndrome by Non-coding RNAs. Front. Immunol. (2018) 9 2705. DOI: 10.3389/fimmu.2018.02705 21. Chancharoenthana W., Sutnu N., Visitchanakun P., Sawaswong V., Chitcharoen S., Payungporn S., Schuetz A., Schultz M.J., Leelahavanichkul A.. Critical roles of sepsis-reshaped fecal virota in attenuating sepsis severity. Front. Immunol. (2022) 13 940935. DOI: 10.3389/fimmu.2022.940935 22. Chancharoenthana W., Kamolratanakul S., Ariyanon W., Thanachartwet V., Phumratanaprapin W., Wilairatana P., Leelahavanichkul A.. Abnormal Blood Bacteriome, Gut Dysbiosis, and Progression to Severe Dengue Disease. Front. Cell Infect. Microbiol. (2022) 12 890817. DOI: 10.3389/fcimb.2022.890817 23. Hiengrach P., Visitchanakun P., Tongchairawewat P., Tangsirisatian P., Jungteerapanich T., Ritprajak P., Wannigama D.L., Tangtanatakul P., Leelahavanichkul A.. Sepsis Encephalopathy Is Partly Mediated by miR370-3p-Induced Mitochondrial Injury but Attenuated by BAM15 in Cecal Ligation and Puncture Sepsis Male Mice. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23105445 24. Thim-Uam A., Makjaroen J., Issara-Amphorn J., Saisorn W., Wannigama D.L., Chancharoenthana W., Leelahavanichkul A.. Enhanced Bacteremia in Dextran Sulfate-Induced Colitis in Splenectomy Mice Correlates with Gut Dysbiosis and LPS Tolerance. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23031676 25. Ondee T., Surawut S., Taratummarat S., Hirankarn N., Palaga T., Pisitkun P., Pisitkun T., Leelahavanichkul A.. Fc Gamma Receptor IIB Deficient Mice: A Lupus Model with Increased Endotoxin Tolerance-Related Sepsis Susceptibility. Shock (2017) 47 743-752. DOI: 10.1097/SHK.0000000000000796 26. Seeley J.J., Ghosh S.. Molecular mechanisms of innate memory and tolerance to LPS. J. Leukoc. Biol. (2017) 101 107-119. DOI: 10.1189/jlb.3MR0316-118RR 27. Gillen J., Ondee T., Gurusamy D., Issara-Amphorn J., Manes N.P., Yoon S.H., Leelahavanichkul A., Nita-Lazar A.. LPS Tolerance Inhibits Cellular Respiration and Induces Global Changes in the Macrophage Secretome. Biomolecules (2021) 11. DOI: 10.3390/biom11020164 28. López-Collazo E., del Fresno C.. Pathophysiology of endotoxin tolerance: Mechanisms and clinical consequences. Crit. Care (2013) 17 242. DOI: 10.1186/cc13110 29. Naler L.B., Hsieh Y.-P., Geng S., Zhou Z., Li L., Lu C.. Epigenomic and transcriptomic analyses reveal differences between low-grade inflammation and severe exhaustion in LPS-challenged murine monocytes. Commun. Biol. (2022) 5 102. DOI: 10.1038/s42003-022-03035-2 30. Koos B., Moderegger E.L., Rump K., Nowak H., Willemsen K., Holtkamp C., Thon P., Adamzik M., Rahmel T.. LPS-Induced Endotoxemia Evokes Epigenetic Alterations in Mitochondrial DNA That Impacts Inflammatory Response. Cells (2020) 9. DOI: 10.3390/cells9102282 31. Falcao-Holanda R.B., Brunialti M.K.C., Jasiulionis M.G., Salomão R.. Epigenetic Regulation in Sepsis, Role in Pathophysiology and Therapeutic Perspective. Front. Med. (2021) 8 685333. DOI: 10.3389/fmed.2021.685333 32. De Santa F., Narang V., Yap Z.H., Tusi B.K., Burgold T., Austenaa L., Bucci G., Caganova M., Notarbartolo S., Casola S.. Jmjd3 contributes to the control of gene expression in LPS-activated macrophages. EMBO J. (2009) 28 3341-3352. DOI: 10.1038/emboj.2009.271 33. De Santa F., Totaro M.G., Prosperini E., Notarbartolo S., Testa G., Natoli G.. The histone H3 lysine-27 demethylase Jmjd3 links inflammation to inhibition of polycomb-mediated gene silencing. Cell (2007) 130 1083-1094. DOI: 10.1016/j.cell.2007.08.019 34. Kondo T., Ito S., Koseki H.. Polycomb in Transcriptional Phase Transition of Developmental Genes. Trends Biochem. Sci. (2016) 41 9-19. DOI: 10.1016/j.tibs.2015.11.005 35. Neele A.E., de Winther M.P.J.. Repressing the repressor: Ezh2 mediates macrophage activation. J. Exp. Med. (2018) 215 1269-1271. DOI: 10.1084/jem.20180479 36. Laugesen A., Hojfeldt J.W., Helin K.. Role of the Polycomb Repressive Complex 2 (PRC2) in Transcriptional Regulation and Cancer. Cold Spring Harb. Perspect. Med. (2016) 6 a026575. DOI: 10.1101/cshperspect.a026575 37. Tan J.Z., Yan Y., Wang X.-x., Jiang Y., Xu H.E.. EZH2: Biology, disease, and structure-based drug discovery. Acta Pharmacol. Sin. (2014) 35 161-174. DOI: 10.1038/aps.2013.161 38. Nakagawa M., Kitabayashi I.. Oncogenic roles of enhancer of zeste homolog 1/2 in hematological malignancies. Cancer Sci. (2018) 109 2342-2348. DOI: 10.1111/cas.13655 39. Garber K.. Histone-writer cancer drugs enter center stage. Nat. Biotechnol. (2020) 38 909-912. DOI: 10.1038/s41587-020-0621-1 40. Liu Y., Peng J., Sun T., Li N., Zhang L., Ren J., Yuan H., Kan S., Pan Q., Li X.. Epithelial EZH2 serves as an epigenetic determinant in experimental colitis by inhibiting TNFalpha-mediated inflammation and apoptosis. Proc. Natl. Acad. Sci. USA (2017) 114 E3796-E3805. PMID: 28439030 41. Qin H., Holdbrooks A.T., Liu Y., Reynolds S.L., Yanagisawa L.L., Benveniste E.N.. SOCS3 deficiency promotes M1 macrophage polarization and inflammation. J. Immunol. (2012) 189 3439-3448. DOI: 10.4049/jimmunol.1201168 42. Neele A.E., Chen H.-J., Gijbels M.J.J., Velden S.V.D., Hoeksema M.A., Boshuizen M.C.S., Bossche J.V.D., Tool A.T., Matlung H.L., Berg T.K.V.D.. Myeloid Ezh2 Deficiency Limits Atherosclerosis Development. Front. Immunol. (2020) 11 594603. DOI: 10.3389/fimmu.2020.594603 43. Ruenjaiman V., Butta P., Leu Y.-W., Pongpanich M., Leelahavanichkul A., Kueanjinda P., Palaga T.. Profile of Histone H3 Lysine 4 Trimethylation and the Effect of Lipopolysaccharide/Immune Complex-Activated Macrophages on Endotoxemia. Front. Immunol. (2019) 10 2956. DOI: 10.3389/fimmu.2019.02956 44. Benjaskulluecha S., Boonmee A., Pattarakankul T., Wongprom B., Klomsing J., Palaga T.. Screening of compounds to identify novel epigenetic regulatory factors that affect innate immune memory in macrophages. Sci. Rep. (2022) 12 1912. DOI: 10.1038/s41598-022-05929-x 45. Cheng Y., He C., Wang M., Ma X., Mo F., Yang S., Han J., Wei X.. Targeting epigenetic regulators for cancer therapy: Mechanisms and advances in clinical trials. Signal Transduct. Target Ther. (2019) 4 62. DOI: 10.1038/s41392-019-0095-0 46. Jaroonwitchawan T., Visitchanakun P., Dang P.C., Ritprajak P., Palaga T., Leelahavanichkul A.. Dysregulation of Lipid Metabolism in Macrophages Is Responsible for Severe Endotoxin Tolerance in FcgRIIB-Deficient Lupus Mice. Front. Immunol. (2020) 11 959. DOI: 10.3389/fimmu.2020.00959 47. Ondee T., Jaroonwitchawan T., Pisitkun T., Gillen J., Nita-Lazar A., Leelahavanichkul A., Somparn P.. Decreased Protein Kinase C-beta Type II Associated with the Prominent Endotoxin Exhaustion in the Macrophage of FcGRIIb-/- Lupus Prone Mice is Revealed by Phosphoproteomic Analysis. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20061354 48. Duan R., Du W., Guo W.. EZH2: A novel target for cancer treatment. J. Hematol. Oncol. (2020) 13 104. DOI: 10.1186/s13045-020-00937-8 49. Carow B., Rottenberg M.E.. SOCS3, a Major Regulator of Infection and Inflammation. Front. Immunol. (2014) 5 58. DOI: 10.3389/fimmu.2014.00058 50. Zhang X., Wang Y., Yuan J., Li N., Pei S., Xu J., Luo X., Mao C., Liu J., Yu T.. Macrophage/microglial Ezh2 facilitates autoimmune inflammation through inhibition of Socs3. J. Exp. Med. (2018) 215 1365-1382. DOI: 10.1084/jem.20171417 51. Binmama S., Dang C.P., Visitchanakun P., Hiengrach P., Somboonna N., Cheibchalard T., Pisitkun P., Chindamporn A., Leelahavanichkul A.. Beta-Glucan from. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms231810951 52. Hiengrach P., Visitchanakun P., Finkelman M.A., Chancharoenthana W., Leelahavanichkul A.. More Prominent Inflammatory Response to Pachyman than to Whole-Glucan Particle and Oat-beta-Glucans in Dextran Sulfate-Induced Mucositis Mice and Mouse Injection through Proinflammatory Macrophages. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23074026 53. Makjaroen J., Thim-Uam A., Dang C.P., Pisitkun T., Somparn P., Leelahavanichkul A.. A Comparison Between 1 Day versus 7 Days of Sepsis in Mice with the Experiments on LPS-Activated Macrophages Support the Use of Intravenous Immunoglobulin for Sepsis Attenuation. J. Inflamm. Res. (2021) 14 7243-7263. DOI: 10.2147/JIR.S338383 54. Udompornpitak K., Charoensappakit A., Sae-Khow K., Bhunyakarnjanarat T., Dang C.P., Saisorn W., Visitchanakun P., Phuengmaung P., Palaga T., Ritprajak P.. Obesity Exacerbates Lupus Activity in Fc Gamma Receptor IIb Deficient Lupus Mice Partly through Saturated Fatty Acid-Induced Gut Barrier Defect and Systemic Inflammation. J. Innate Immun. (2022) 15 240-261. DOI: 10.1159/000526206 55. Tungsanga S., Panpetch W., Bhunyakarnjanarat T., Udompornpitak K., Katavetin P., Chancharoenthana W., Chatthanathon P., Somboonna N., Tungsanga K., Tumwasorn S.. Uremia-Induced Gut Barrier Defect in 5/6 Nephrectomized Mice Is Worsened by Candida Administration through a Synergy of Uremic Toxin, Lipopolysaccharide, and (1→3)-β-D-Glucan, but Is Attenuated by Lacticaseibacillus rhamnosus L34. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23052511 56. Visitchanakun P., Panpetch W., Saisorn W., Chatthanathon P., Wannigama D.L., Thim-Uam A., Svasti S., Fucharoen S., Somboonna N., Leelahavanichkul A.. Increased susceptibility to dextran sulfate-induced mucositis of iron-overload β-thalassemia mice, another endogenous cause of septicemia in thalassemia. Clin. Sci. (2021) 135 1467-1486. DOI: 10.1042/CS20210328 57. Boonhai S., Bootdee K., Saisorn W., Takkavatakarn K., Sitticharoenchai P., Tungsanga S., Tiranathanagul K., Leelahavanichkul A.. TMAO reductase, a biomarker for gut permeability defect insduced inflammation, in mouse model of chronic kidney disease and dextran sulfate solution-induced mucositis. Asian Pac. J. Allergy Immunol. (2021). DOI: 10.12932/AP-100321-1084 58. Hiengrach P., Panpetch W., Chindamporn A., Leelahavanichkul A.. Macrophage depletion alters bacterial gut microbiota partly through fungal overgrowth in feces that worsens cecal ligation and puncture sepsis mice. Sci. Rep. (2022) 12 9345. DOI: 10.1038/s41598-022-13098-0 59. Charoensappakit A., Sae-Khow K., Leelahavanichkul A.. Gut Barrier Damage and Gut Translocation of Pathogen Molecules in Lupus, an Impact of Innate Immunity (Macrophages and Neutrophils) in Autoimmune Disease. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23158223 60. Ezponda T., Licht J.D.. Molecular pathways: Deregulation of histone h3 lysine 27 methylation in cancer-different paths, same destination. Clin. Cancer Res. (2014) 20 5001-5008. DOI: 10.1158/1078-0432.CCR-13-2499 61. Nichol J.N., Dupéré-Richer D., Ezponda T., Licht J.D., Miller Jr W.H.. H3K27 Methylation: A Focal Point of Epigenetic Deregulation in Cancer. Adv. Cancer Res. (2016) 131 59-95. PMID: 27451124 62. Yue D., Wang Z., Yang Y., Hu Z., Luo G., Wang F.. EZH2 inhibitor GSK343 inhibits sepsis-induced intestinal disorders. Exp. Ther. Med. (2021) 21 437. DOI: 10.3892/etm.2021.9854 63. Eliyahu A., Barel O., Greenbaum L., Hoffer G.Z., Goldberg Y., Raas-Rothschild A., Singer A., Bar-Joseph I., Kunik V., Javasky E.. Refining the Phenotypic Spectrum of KMT5B-Associated Developmental Delay. Front. Pediatr. (2022) 10 844845. DOI: 10.3389/fped.2022.844845 64. Zhang Q., Sun H., Zhuang S., Liu N., Bao X., Liu X., Ren H., Lv D., Li Z., Bai J.. Novel pharmacological inhibition of EZH2 attenuates septic shock by altering innate inflammatory responses to sepsis. Int. Immunopharmacol. (2019) 76 105899. DOI: 10.1016/j.intimp.2019.105899 65. Zhao D., Li Z., Liu X., Liu N., Bao X., Sun H., Meng Q., Ren H., Bai J., Zhou X.. Lymphocyte expression of EZH2 is associated with mortality and secondary infectious complications in sepsis. Int. Immunopharmacol. (2020) 89 107042. DOI: 10.1016/j.intimp.2020.107042 66. Adema V., Colla S.. EZH2 Inhibitors: The Unpacking Revolution. Cancer Res. (2022) 82 359-361. DOI: 10.1158/0008-5472.CAN-21-4311 67. Leelahavanichkul A., Somparn P., Bootprapan T., Tu H., Tangtanatakul P., Nuengjumnong R., Worasilchai N., Tiranathanagul K., Eiam-ong S., Levine M.. High-dose ascorbate with low-dose amphotericin B attenuates severity of disease in a model of the reappearance of candidemia during sepsis in the mouse. Am. J. Physiol. Regul. Integr. Comp. Physiol. (2015) 309 R223-R234. DOI: 10.1152/ajpregu.00238.2014 68. Vu C.T.B., Thammahong A., Leelahavanichkul A., Ritprajak P.. Alteration of macrophage immune phenotype in a murine sepsis model is associated with susceptibility to secondary fungal infection. Asian Pac. J. Allergy Immunol. (2022) 40 162-171. PMID: 31586490 69. Vu C.T.B., Thammahong A., Yagita H., Azuma M., Hirankarn N., Ritprajak P., Leelahavanichkul A.. Blockade Of PD-1 Attenuated Postsepsis Aspergillosis Via the Activation of IFN-gamma and the Dampening of IL-10. Shock (2020) 53 514-524. DOI: 10.1097/SHK.0000000000001392 70. Kitchen G.B., Hopwood T., Ramamoorthy T.G., Downton P., Begley N., Hussell T., Dockrell D.H., Gibbs J.E., Ray D.W., Loudon A.S.I.. The histone methyltransferase Ezh2 restrains macrophage inflammatory responses. FASEB J. (2021) 35 e21843. DOI: 10.1096/fj.202100044RRR 71. Wang Q., Xu L., Zhang X., Liu D., Wang R.. GSK343, an inhibitor of EZH2, mitigates fibrosis and inflammation mediated by HIF-1alpha in human peritoneal mesothelial cells treated with high glucose. Eur. J. Pharmacol. (2020) 880 173076. DOI: 10.1016/j.ejphar.2020.173076 72. Bamidele A.O., Svingen P.A., Sagstetter M.R., Sarmento O.F., Gonzalez M., Braga N.M.B., Kugathasan S., Lomberk G., Urrutia R.A., Faubion W.A.. Disruption of FOXP3-EZH2 Interaction Represents a Pathobiological Mechanism in Intestinal Inflammation. Cell Mol. Gastroenterol. Hepatol. (2019) 7 55-71. DOI: 10.1016/j.jcmgh.2018.08.009 73. He J., Song Y., Li G., Xiao P., Liu Y., Xue Y., Cao Q., Tu X., Pan T., Jiang Z.. Fbxw7 increases CCL2/7 in CX3CR1hi macrophages to promote intestinal inflammation. J. Clin. Investig. (2019) 129 3877-3893. DOI: 10.1172/JCI123374 74. Yong H., Wu G., Chen J., Liu X., Bai Y., Tang N., Liu L., Wei J.. lncRNA MALAT1 Accelerates Skeletal Muscle Cell Apoptosis and Inflammatory Response in Sepsis by Decreasing BRCA1 Expression by Recruiting EZH2. Mol. Ther. Nucleic. Acids (2020) 21 1120-1121. DOI: 10.1016/j.omtn.2020.07.033 75. Prele C.M., Keith-Magee A.L., Yerkovich S.T., Murcha M., Hart P.H.. Suppressor of cytokine signalling-3 at pathological levels does not regulate lipopolysaccharide or interleukin-10 control of tumour necrosis factor-alpha production by human monocytes. Immunology (2006) 119 8-17. DOI: 10.1111/j.1365-2567.2006.02383.x 76. Berlato C., Cassatella M.A., Kinjyo I., Gatto L., Yoshimura A., Bazzoni F.. Involvement of suppressor of cytokine signaling-3 as a mediator of the inhibitory effects of IL-10 on lipopolysaccharide-induced macrophage activation. J. Immunol. (2002) 168 6404-6411. DOI: 10.4049/jimmunol.168.12.6404 77. Qasimi P., Ming-Lum A., Ghanipour A., Ong C.J., Cox M.E., Ihle J., Cacalano N., Yoshimura A., Mui A.L.. Divergent mechanisms utilized by SOCS3 to mediate interleukin-10 inhibition of tumor necrosis factor alpha and nitric oxide production by macrophages. J. Biol. Chem. (2006) 281 6316-6324. DOI: 10.1074/jbc.M508608200 78. Frankenberger M., Pechumer H., Ziegler-Heitbrock H.W.. Interleukin-10 is upregulated in LPS tolerance. J. Inflamm. (1995) 45 56-63. PMID: 7583353 79. Godet C., Goujon J.M., Petit I., Lecron J.C., Hauet T., Mauco G., Carretier M., Robert R.. Endotoxin tolerance enhances interleukin-10 renal expression and decreases ischemia-reperfusion renal injury in rats. Shock (2006) 25 384-388. DOI: 10.1097/01.shk.0000209528.35743.54 80. Salkowski C.A., Detore G.R., Vogel S.N.. Lipopolysaccharide and monophosphoryl lipid A differentially regulate interleukin-12, gamma interferon, and interleukin-10 mRNA production in murine macrophages. Infect. Immun. (1997) 65 3239-3247. DOI: 10.1128/iai.65.8.3239-3247.1997 81. Shimauchi H., Ogawa T., Okuda K., Kusumoto Y., Okada H.. Autoregulatory effect of interleukin-10 on proinflammatory cytokine production by Porphyromonas gingivalis lipopolysaccharide-tolerant human monocytes. Infect. Immun. (1999) 67 2153-2159. DOI: 10.1128/IAI.67.5.2153-2159.1999 82. Mola S., Pinton G., Erreni M., Corazzari M., De Andrea M., Grolla A.A., Martini V., Moro L., Porta C.. Inhibition of the Histone Methyltransferase EZH2 Enhances Protumor Monocyte Recruitment in Human Mesothelioma Spheroids. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22094391 83. Alexander W.S.. Suppressors of cytokine signalling (SOCS) in the immune system. Nat. Rev. Immunol. (2002) 2 410-416. DOI: 10.1038/nri818 84. Cassatella M.A., Gasperini S., Bovolenta C., Calzetti F., Vollebregt M., Scapini P., Marchi M., Suzuki R., Suzuki A., Yoshimura A.. Interleukin-10 (IL-10) selectively enhances CIS3/SOCS3 mRNA expression in human neutrophils: Evidence for an IL-10-induced pathway that is independent of STAT protein activation. Blood (1999) 94 2880-2889. DOI: 10.1182/blood.V94.8.2880.420k31_2880_2889 85. Aziz M., Jacob A., Matsuda A., Wu R., Zhou M., Dong W., Yang W.L., Wang P.. Pre-treatment of recombinant mouse MFG-E8 downregulates LPS-induced TNF-alpha production in macrophages via STAT3-mediated SOCS3 activation. PLoS ONE (2011) 6. DOI: 10.1371/journal.pone.0027685 86. Eliopoulos A.G., Dumitru C.D., Wang C.C., Cho J., Tsichlis P.N.. Induction of COX-2 by LPS in macrophages is regulated by Tpl2-dependent CREB activation signals. EMBO J. (2002) 21 4831-4840. DOI: 10.1093/emboj/cdf478 87. Grkovich A., Johnson C.A., Buczynski M.W., Dennis E.A.. Lipopolysaccharide-induced cyclooxygenase-2 expression in human U937 macrophages is phosphatidic acid phosphohydrolase-1-dependent. J. Biol. Chem. (2006) 281 32978-32987. DOI: 10.1074/jbc.M605935200 88. Williams L., Bradley L., Smith A., Foxwell B.. Signal transducer and activator of transcription 3 is the dominant mediator of the anti-inflammatory effects of IL-10 in human macrophages. J. Immunol. (2004) 172 567-576. DOI: 10.4049/jimmunol.172.1.567 89. Zhu J., Li S., Zhang Y., Ding G., Zhu C., Huang S., Zhang A., Jia Z., Li M.. COX-2 contributes to LPS-induced Stat3 activation and IL-6 production in microglial cells. Am. J. Transl. Res. (2018) 10 966-974. PMID: 29636886 90. Lu Y., Fukuyama S., Yoshida R., Kobayashi T., Saeki K., Shiraishi H., Yoshimura A., Takaesu G.. Loss of SOCS3 gene expression converts STAT3 function from anti-apoptotic to pro-apoptotic. J. Biol. Chem. (2006) 281 36683-36690. DOI: 10.1074/jbc.M607374200 91. Kim H., Kim M., Im S.-K., Fang S.. Mouse Cre-LoxP system: General principles to determine tissue-specific roles of target genes. Lab. Anim. Res. (2018) 34 147-159. DOI: 10.5625/lar.2018.34.4.147 92. Arango Duque G., Descoteaux A.. Macrophage cytokines: Involvement in immunity and infectious diseases. Front. Immunol. (2014) 5 491. DOI: 10.3389/fimmu.2014.00491 93. Jiang H.R., Muckersie E., Robertson M., Xu H., Liversidge J., Forrester J.V.. Secretion of interleukin-10 or interleukin-12 by LPS-activated dendritic cells is critically dependent on time of stimulus relative to initiation of purified DC culture. J. Leukoc. Biol. (2002) 72 978-985. DOI: 10.1189/jlb.72.5.978 94. Mazer M., Unsinger J., Drewry A., Walton A., Osborne D., Blood T., Hotchkiss R., Remy K.E.. IL-10 Has Differential Effects on the Innate and Adaptive Immune Systems of Septic Patients. J. Immunol. (2019) 203 2088-2099. DOI: 10.4049/jimmunol.1900637 95. Saraiva M., O’Garra A.. The regulation of IL-10 production by immune cells. Nat. Rev. Immunol. (2010) 10 170-181. DOI: 10.1038/nri2711 96. Iyer S.S., Cheng G.. Role of interleukin 10 transcriptional regulation in inflammation and autoimmune disease. Crit. Rev. Immunol. (2012) 32 23-63. DOI: 10.1615/CritRevImmunol.v32.i1.30 97. Chen L., Deng H., Cui H., Fang J., Zuo Z., Deng J., Li Y., Wang X., Zhao L.. Inflammatory responses and inflammation-associated diseases in organs. Oncotarget (2018) 9 7204-7218. DOI: 10.18632/oncotarget.23208 98. Song J., Park D.W., Moon S., Cho H.-J., Park J.H., Seok H., Choi W.S.. Diagnostic and prognostic value of interleukin-6, pentraxin 3, and procalcitonin levels among sepsis and septic shock patients: A prospective controlled study according to the Sepsis-3 definitions. BMC Infect. Dis. (2019) 19. DOI: 10.1186/s12879-019-4618-7 99. Wang L., Zhao H., Wang D.. Inflammatory cytokine expression in patients with sepsis at an intensive care unit. Exp. Ther. Med. (2018) 16 2126-2131. DOI: 10.3892/etm.2018.6376 100. Greco M., Mazzei A., Suppressa S., Palumbo C., Verri T., Lobreglio G.. Human Leukocyte Antigen-DR Isotype Expression in Monocytes and T Cells Interferon-Gamma Release Assay in Septic Patients and Correlation with Clinical Outcome. J. Clin. Med. Res. (2021) 13 293-303. DOI: 10.14740/jocmr4474 101. Mansfield S., Grießl M., Gutknecht M., Cook C.H.. Sepsis and cytomegalovirus: Foes or conspirators?. Med. Microbiol. Immunol. (2015) 204 431-437. DOI: 10.1007/s00430-015-0407-0 102. Makjaroen J., Phuengmaung P., Saisorn W., Udomkarnjananun S., Pisitkun T., Leelahavanichkul A.. Lipopolysaccharide Tolerance Enhances Murine Norovirus Reactivation: An Impact of Macrophages Mainly Evaluated by Proteomic Analysis. Int. J. Mol. Sci. (2023) 24. DOI: 10.3390/ijms24031829 103. Ondee T., Gillen J., Visitchanakun P., Somparn P., Issara-Amphorn J., Dang P.C., Chancharoenthana W., Gurusamy D., Nita-Lazar A., Leelahavanichkul A.. Lipocalin-2 (Lcn-2) Attenuates Polymicrobial Sepsis with LPS Preconditioning (LPS Tolerance) in FcGRIIb Deficient Lupus Mice. Cells (2019) 8. DOI: 10.3390/cells8091064 104. Hiengrach P., Panpetch W., Chindamporn A., Leelahavanichkul A.. Helicobacter pylori, Protected from Antibiotics and Stresses Inside Candida albicans Vacuoles, Cause Gastritis in Mice. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23158568 105. Panpetch W., Visitchanakun P., Saisorn W., Sawatpanich A., Chatthanathon P., Somboonna N., Tumwasorn S., Leelahavanichkul A.. Lactobacillus rhamnosus attenuates Thai chili extracts induced gut inflammation and dysbiosis despite capsaicin bactericidal effect against the probiotics, a possible toxicity of high dose capsaicin. PLoS ONE (2021) 16. DOI: 10.1371/journal.pone.0261189 106. Phuengmaung P., Mekjaroen J., Saisorn W., Chatsuwan T., Somparn P., Leelahavanichkul A.. Rapid Synergistic Biofilm Production of Pseudomonas and Candida on the Pulmonary Cell Surface and in Mice, a Possible Cause of Chronic Mixed Organismal Lung Lesions. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23169202 107. Singkham-In U., Phuengmaung P., Makjaroen J., Saisorn W., Bhunyakarnjanarat T., Chatsuwan T., Chirathaworn C., Chancharoenthana W., Leelahavanichkul A.. Chlorhexidine Promotes Psl Expression in Pseudomonas aeruginosa That Enhances Cell Aggregation with Preserved Pathogenicity Demonstrates an Adaptation against Antiseptic. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23158308 108. Issara-Amphorn J., Dang C.P., Saisorn W., Limbutara K., Leelahavanichkul A.. Candida Administration in Bilateral Nephrectomy Mice Elevates Serum (1→3)-β-D-glucan That Enhances Systemic Inflammation Through Energy Augmentation in Macrophages. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22095031 109. Dobin A., Davis C.A., Schlesinger F., Drenkow J., Zaleski C., Jha S., Batut P., Chaisson M., Gingeras T.R.. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics (2013) 29 15-21. DOI: 10.1093/bioinformatics/bts635 110. Bray N.L., Pimentel H., Melsted P., Pachter L.. Near-optimal probabilistic RNA-seq quantification. Nat. Biotechnol. (2016) 34 525-527. DOI: 10.1038/nbt.3519 111. Robinson M.D., McCarthy D.J., Smyth G.K.. edgeR: A Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics (2010) 26 139-140. DOI: 10.1093/bioinformatics/btp616 112. Law C.W., Chen Y., Shi W., Smyth G.K.. voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. (2014) 15 R29. DOI: 10.1186/gb-2014-15-2-r29 113. Smyth G.K.. Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. (2004) 3 3. DOI: 10.2202/1544-6115.1027 114. Gu Z., Eils R., Schlesner M.. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics (2016) 32 2847-2849. DOI: 10.1093/bioinformatics/btw313 115. Shah S.G., Mandloi T., Kunte P., Natu A., Rashid M., Reddy D., Gadewal N., Gupta S.. HISTome2: A database of histone proteins, modifiers for multiple organisms and epidrugs. Epigenetics Chromatin (2020) 13 31. DOI: 10.1186/s13072-020-00354-8 116. Kassambara A.K.M., Biecek P., Fabian S.. Survminer: Drawing Survival Curves Using ‘ggplot2’ (2017) 1 117. Visitchanakun P., Kaewduangduen W., Chareonsappakit A., Susantitaphong P., Pisitkun P., Ritprajak P., Townamchai N., Leelahavanichkul A.. Interference on Cytosolic DNA Activation Attenuates Sepsis Severity: Experiments on Cyclic GMP–AMP Synthase (cGAS) Deficient Mice. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms222111450 118. Saisorn W., Saithong S., Phuengmaung P., Udompornpitak K., Bhunyakarnjanarat T., Visitchanakun P., Chareonsappakit A., Pisitkun P., Chiewchengchol D., Leelahavanichkul A.. Acute Kidney Injury Induced Lupus Exacerbation Through the Enhanced Neutrophil Extracellular Traps (and Apoptosis) in Fcgr2b Deficient Lupus Mice with Renal Ischemia Reperfusion Injury. Front. Immunol. (2021) 12 669162. DOI: 10.3389/fimmu.2021.669162
--- title: 'The Parental Stress Scale: Psychometric Properties in Pediatric Hospital Emergency Setting' authors: - Néstor Montoro-Pérez - Silvia Escribano - Miguel Richart-Martínez - María Isabel Mármol-López - Raimunda Montejano-Lozoya journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049284 doi: 10.3390/ijerph20064771 license: CC BY 4.0 --- # The Parental Stress Scale: Psychometric Properties in Pediatric Hospital Emergency Setting ## Abstract Parental psychological distress has been identified as a predisposing factor in attendance at and the inappropriate use of hospital pediatric emergency departments (PEDs). The aim of the study was to validate the Parental Stress Scale (PSS), a 12-item Spanish scale, in parents seeking care at PEDs. The study involved 270 participants with a mean age of 37.9 (SD = 6.76) years, of which $77.4\%$ were women. The properties of the PSS were analyzed. The scale showed adequate internal consistency for the different factors (0.80 for the “Stressors” factor and 0.78 for the “Baby’s Rewards” factor) and optimal model fit (chi-square = 107.686; df = 53; CFI = 0.99; TLI = 0.98; RMSEA = 0.028; $90\%$ CI = 0.00–0.05). The 12-item Spanish version of the PSS is a valid and reliable instrument for assessing the stress levels of parents seeking care in PEDs. ## 1. Introduction Bringing a child into the family and parenting is often rewarding for parents, but it can lead to feelings of stress. In fact, the transition and journey towards parenthood is considered by many parents to be one of the most difficult periods of their lives [1]. Parental stress is, therefore, taken to mean the stress that parents feel as a result of the daunting task of rearing and bringing up their children. In particular, parents have the subjective perception that they are unable to cope with the demands of parenting, leaving them feeling ineffective and powerless in fulfilling their parental role and culminating in stressful experiences [2,3]. A number of factors play a role in the onset of parental stress. On the one hand, these include the characteristics of the family, such as unemployment, financial difficulties, marital stress, divorce, single parenthood, and low perceived social support. On the other hand, are child characteristics, such as behavioral, health, and developmental problems [4,5,6]. It has in fact been confirmed that sustained high levels of parental stress over time can lead to poor emotional management of children. This ultimately affects parents’ capacity to handle the tasks and demands of parenting, leading to cognitive developmental problems, attention difficulties, inappropriate prosocial behavior, and insecure attachment [7,8]. This sets in motion a vicious circle, disrupting the quality of life and relationships of both [9]. Meanwhile, the scientific literature shows that psychological distress and high levels of parental stress are associated with elevated demand for care in hospital pediatric emergency departments (PEDs), which is not justified by the actual seriousness of the pathology [10,11,12]. Consequences of the inappropriate use of emergency departments include increased costs, service overload, poor care, staff burnout, user dissatisfaction, and at times even users walking out without having been examined, diagnosed, or treated by a doctor [13]. In this context, effective parental stress assessment tools are, therefore, needed to reduce the frequent and non-emergency visits to such services resulting from high levels of parental stress. Several tools are available for measuring parental stress, but one of the most widely used internationally is the Parental Stress Scale (PSS), developed by Berry and Jones [2,14,15]. The study for the creation and initial validation of the scale was carried out on a sample of clinical and non-clinical children. The final instrument, following exploratory factor analysis (EFA), consisted of 16 items and four factors: “Parental Stressors”, “Lack of Control”, “Parental Satisfaction”, and “Parental Rewards”. In the EFA, of the initial 18 items, items 2 [There is little or nothing I wouldn’t do for my child (ren) if it was necessary] and 3 [Caring for my child (ren) sometimes takes more time and energy than I have to give] failed to load on any factor, while item 16 (Having children has meant having too few choices and too little control over my life) shared factor loadings with the “Lack of Control” and “Parental Stressors” factors (0.43 and 0.54, respectively). Furthermore, item 18 [I find my child (ren) enjoyable] also shared factor loadings with the “Parental Rewards” and “Parental Satisfaction” factors (0.50 and 0.47, respectively). The final scale that was obtained following EFA had good psychometric properties: internal consistency (α = 0.83), test-retest reliability over a six-week period ($r = 0.81$; $p \leq 0.01$), and convergent validity with the Parenting Stress Index ($r = 0.75$; $p \leq 0.01$) and the Perceived Stress Scale ($r = 0.50$; $p \leq 0.01$) [2]. Several studies have since validated the scale in different languages (Bahasa Malaysia, Brazilian, Portuguese, Chinese, Danish, Hindi, Greek, and Spanish) and different population samples (normative population, parents of children with behavioral and/or developmental problems, sick children, and children with sleeping disorders), obtaining differing factorial structures (1, 2, 3, and 4 factors) and removing certain items because they failed to load substantially on any factor [14,15,16,17]. Mixão et al. [ 17] validated the instrument in Portuguese in a sample of 416 parents of children ranging in age from one month to 15 years old, who had sought care in the pediatric emergency department of a general hospital. The authors performed an EFA with the original 18 items yielding four factors. The internal consistency that was obtained for the various dimensions ranged from 0.57 to 0.78. In Spain, the only study available is by Oronoz et al. [ 18], in which the tool was adapted to Spanish with a sample of 411 first-time parents, producing a final scale after EFA of 12 items and two factors: “Baby’s Rewards” and “Stressors”. However, to the best of our knowledge, there is no valid instrument in Spanish for measuring parental stress in the hospital pediatric emergency setting, hence the importance of this research which aims to examine the psychometric properties of the 12-item PSS in Spanish that was developed by Oronoz et al. [ 18] in parents of children seeking care in the PED. ## 2.1. Design & Participants An instrumental study was developed, that was designed for adapting instruments to new contexts and analyzing their psychometric properties [19]. Participants were selected by non-probability convenience sampling at a referral hospital in Valencia (Spain). Selection took place between September 2021 and January 2022 according to the following selection criteria: [1] be a parent of a child seeking care in the hospital’s pediatric emergency department, and [2] speak Spanish fluently. The sample size was calculated using statistical analyses of optimal conditions, which gave a minimum sample size of 200 participants for a confirmatory factor analysis (CFA) [20,21]. The final sample, once dropouts had been eliminated, comprised of a total of 270 people. ## 2.2. Data Collection Tools For data collection purposes, three documents were merged into one. This included:-An ad hoc questionnaire: prepared specifically for this study and including sociodemographic variables such as: age (as a continuous variable); respondent (father, mother); nationality (Spanish, other); level of education (primary, secondary, higher); marital status (single, married/in a stable relationship, divorced/widowed); family unit (nuclear family, extended family, homoparental family, separated-parent family, blended family); socioeconomic status based on monthly household income (with nine response categories ranging from less than 600 euros to more than 6000 euros per month); age of the child in months (as a continuous variable); number of children in the family unit (as a continuous variable); whether or not the child was the first child (yes, no); and the perceived level of anxiety (with five response categories ranging from 1 = not anxious at all to 5 = very anxious).-Parental Stress Scale (PSS): original instrument developed by Berry and Jones [2]. We used the 12-item Spanish version that was developed by Oronoz et al. [ 18]. The items are answered using a five-point Likert-type scale (1 = strongly disagree to 5 = strongly agree). A higher score on the scale indicates higher levels of parental stress. As a result of the translation-retrotranslation process and in consensus with experts in the field of parenting, item 16 was eliminated from the scale on the grounds of ambiguity, leaving an initial scale of 17 items. The final scale after EFA consisted of 12 items and two factors (“Stressors” variance = $10.1\%$ and “Baby’s Rewards” variance = $23.4\%$). The instrument shows good internal consistency for both factors (α = 0.76 for “Stressors” and α = 0.77 for “Baby’s Rewards”). Convergent validity was determined using the State-Trait Anxiety Inventory (STAI) scale and the Beck Depression Inventory (BDI), obtaining statistically significant correlations in the hypothesized direction and magnitude ($r = 0.49$, $p \leq 0.001$ and $r = 0.51$, $p \leq 0.001$, respectively) [18].-State-Trait Anxiety Inventory (STAI E-7): the 7-item Spanish version that was developed by Perpiñá-Galvañ et al. [ 22]. This instrument consists of the items that were proposed by Chlan et al. plus item 1 of the original scale that was developed by Spielberger. The questionnaire measures state anxiety. The items are answered using a four-point Likert scale (0 = not at all to 3 = very much). A higher score on the scale indicates higher levels of anxiety. The scale has good internal consistency (α = 0.89) [22]. The ordinal alpha coefficient in our study was 0.91. ## 2.3. Procedure The parents’ initial contact with PED healthcare staff took place in the triage area. Here the study objective was explained and they were invited to participate. All those wishing to participate were provided with one document containing all the study variables with the PSS and STAI-E7 scales, and a separate document for signing the informed consent form. Once the paperwork was completed, it was placed in a sealed envelope and passed to the lead researcher. Preliminary tests were carried out to check whether the document was difficult for the interviewees to interpret, as well as to ascertain if there were any comprehension issues. After the first 10 participants, it was noted that the document did not present reading comprehension difficulties. ## 2.4. Data Analysis SPSS Statistics version 21.1 for OX [23] was used for descriptive, normative, and product-moment correlation analyses. To analyze the scale’s psychometric properties, the free software R (version 3.6.4) was used [24]. The assessment instrument’s performance was examined by calculating the skewness and kurtosis of the data and the floor and ceiling effects. The distribution of the variables is considered normal when the skewness and kurtosis coefficients lie between −1.5 and 1.5 [20,21]. According to the literature, floor or ceiling effects occur when more than $15\%$ of the participants’ responses fall in the lower or upper response category ranges, suggesting a reduced ability to differentiate between scores [25]. The data were considered ordinal as per the criteria of Rhemtulla et al. [ 26]. A CFA was performed. Estimates were obtained using the robust weighted least squares mean and variance adjusted (WLSMV) method, used with ordinal variables [25] from the Lavaan [27] package in R. The model fit to the data was analyzed using the comparative fit index (CFI), Tucker–Lewis Index (TLI), and root mean square error of approximation (RMSEA) [27], with index values > 0.90 and RMSEA < 0.06 [28,29] being considered adequate. Internal consistency was calculated by means of the ordinal alpha coefficient, which is more accurate for categorical response scales. The coefficient α ≥ 0.70 was accepted as an indicator of good reliability [30,31]. Convergent validity was assessed by product-moment correlation between the PSS, the STAI E-7 and the criterion variable of perceived anxiety. Significant positive correlations between 0.2 and 0.5 were expected [32], which would confirm the hypothesis that higher levels of parental stress lead to higher levels of anxiety. ## 2.5. Ethical Considerations All participants in the study agreed to take part voluntarily. The study was approved by the center’s research ethics committee (Registration No. 2020-486-1). With respect to data confidentiality, privacy and confidentiality were ensured in accordance with Regulation (EU) $\frac{2016}{679}$ of the European Parliament and of the Council of 27 April 2016 [33]. ## 3.1. Sample Sociodemographic Characteristics The sample characteristics can be seen in Table 1. The study involved 270 participants: the mean age of all the participants was 37.9 (SD = 6.76) years; $77.4\%$ were women; $85.2\%$ were of Spanish nationality; and $27.8\%$ were single, $63\%$ were married or in a stable relationship, and $9.2\%$ were separated, divorced, or widowed. The average household socioeconomic level was between 1501 and 3000 euros per month ($35.60\%$). The majority ($56.70\%$) had a higher level of education. Significant sex-based differences were found for most of the sociodemographic variables. ## 3.2. Characteristics Performance and Psychometric Properties of the Scale Table 2 shows the performance of the scale’s component items. Ceiling and floor effects, skewness, and kurtosis were observed for all the items, so the data were considered ordinal. The analysis was based on a congeneric model taking into account the scale structure that was obtained in the EFA by Oronoz et al. [ 18]. In our study, the CFA of the questionnaire showed an adequate fit of the data to the structure of the 12-item Spanish version (chi-square = 107.686; df = 53; CFI = 0.99; TLI = 0.98; RMSEA = 0.028, $90\%$ CI = 0.00–0.05). The estimated factor loadings ranged from 0.50 to 0.85 (Figure 1). The ordinal alpha coefficient by dimension was 0.80 for the “Stressors” factor and 0.78 for the “Baby’s Rewards” factor. In terms of convergent validity, scores on the PSS scale’s “Baby’s Rewards” factor correlate positively with the STAI E-7 ($r = 0.152$, $p \leq 0.05$). However, no statistical association was found with the “Stressors” factor ($r = 0.117$; $$p \leq 0.055$$). In the criterion variable of perceived anxiety, the PSS correlates positively for both the “Baby’s Rewards” factor ($r = 0.218$; $p \leq 0.01$) and the “Stressors” factor ($r = 0.197$; $p \leq 0.01$). ## 3.3. Descriptive Analysis of Parental Stress in the Paediatric Emergency Setting Table 3 shows the mean values of the instrument’s factors. A mean score of 6.52 (SD = 2.35) out of a maximum of 16 was obtained for the “Baby’s Rewards” factor and 17.14 (SD = 6.11) out of a maximum of 35 for the “Stressors” factor. In the analysis by sex, mothers showed higher levels of stress than fathers. ## 4. Discussion To our knowledge, this is the first study to confirm the 12-item structure of the PSS that was proposed by Oronoz et al. [ 18] in the PED setting. The CFA shows that the Spanish scale retains the structure determined in the EFA by the authors who validated it in 2007, and the fit indices that were obtained in this study are adequate. Although the original scale that was developed by Berry and Jones [2] consists of four factors (“Parental Stressors”, “Lack of Control”, “Parental Satisfaction”, and “Parental Rewards”), internationally there is inconsistency between the factors that were obtained in the process of adapting the instrument, with two main factors (“Parental Satisfaction” and “Parental Stressors”) being commonly identified [34,35,36]. Only three studies were found in which the original tetra-factorial structure was confirmed with good statistical fit [34,37,38]. This factorial inconsistency could be explained by Pontoppidan et al. [ 39] and Harding et al. [ 35], who argue that the tetra-factorial model appears to be a subdivision of two primary factors, whereby the “Parental Stressors” and “Lack of Control” factors make up the “Parental Stressors” factor, while the “Parental Satisfaction” and “Parental Rewards” factors make up the “Parental Satisfaction” factor. An analysis of the scientific literature suggests that the different settings and populations in which the scale has been validated affect the stability of the items and the presence of the different factors [15]. There is, however, no doubt that the available versions accurately measure the construct of parental stress, with those where the items are grouped around two main factors being easier to interpret. In particular, the version that was proposed by Oronoz et al. [ 18] may be regarded as less invasive and easier to use within the context of this research. In terms of the performance of the scale and item functioning, a ceiling and floor effect is found for all the instrument’s items, most likely indicating a lack of variability between the different scores, which could make it difficult to adequately discriminate between parents with varying levels of stress [25]. This phenomenon has also been observed in the study by Leung et al. [ 40] which suggested that the PSS may pose difficulties for parents with low levels of stress. In terms of internal consistency, the indices were optimal and in line with those that were described in the literature [30,31,32], with slightly higher values being identified than those that were obtained in other validation studies of the instrument [18,34,36,41,42,43]. From the perspective of convergent validity, the results show an association between higher levels of stress and anxiety. Our findings are corroborated by several authors who have found that parents with higher levels of stress also have higher levels of anxiety [44,45,46]. In terms of the instruments that are used for convergent validity, the STAI has been widely relied upon in the scientific literature to confirm this psychometric property [18,38,41]. Nevertheless, the values in our study are slightly lower, with one factor identified as having a low correlation and another with no statistical association. This may be explained by the fact that the STAI-E7 only measures momentary anxiety (state), whereas the PSS assesses parental stress over time (trait) [2,18,22]. Although the “Parental Stressors” factor does not correlate with the STAI-E7, the values do come close to being significant, which is confirmed by the criterion variable of perceived anxiety that is included in the study. With regard to the scores that were obtained for parental stress levels in the PED, our results cannot be compared, as there are no known scientific studies measuring stress levels using the version of the PSS that was proposed by Oronoz et al. [ 18] in parents seeking care in the PED. This notwithstanding, the data from this study are in line with the literature, with the common phenomenon of mothers having higher levels of stress than fathers across the different study settings [2,15,18]. Finally, the intrinsic limitations of this research must be taken into account when interpreting the results, in particular the ceiling and floor effect that was observed in most of the items, which may make discriminating between the results difficult. However, we have provided percentiles of the instrument in the context studied, thus providing future researchers with useful references to help with the correct interpretation of the results and establish the effectiveness of any future interventions. As far as further lines of research are concerned, there is a need for quantitative studies measuring parental stress in PEDs, given that it has been qualitatively affirmed that parental psychological distress, stress, and anxiety are predisposing factors in the demand for care falling outside the scope of such services [11,12,47]. ## 5. Conclusions The 12-item PSS in Spanish is a valid, reliable, brief, and minimally invasive assessment instrument that is capable of detecting parents with high levels of stress in PED. With knowledge of the stress levels that are experienced by parents, healthcare professionals will be able to develop strategies and interventions to eliminate parental stress, with the ultimate implication of reducing non-emergency visits to the PED caused by these high levels of parental stress. ## References 1. Raphael-Leff J., Trevarthen C., Klaus M.. Psychological Processes of Childbearing (2018.0). DOI: 10.4324/9780429482922 2. Berry J.O., Jones W.H.. The Parental Stress Scale: Initial Psychometric Evidence. JSPR (1995.0) 3 463-472. DOI: 10.1177/0265407595123009 3. Gavita O.A., David D., DiGiuseppe R.. You are such a bad child! Appraisals as mechanisms of parental negative and positive affect. J. Gen. Psychol. (2014.0) 2 113-129. DOI: 10.1080/00221309.2013.874971 4. Berenguer C., Rosello B., Miranda A.. Mothers’ stress and behavioral and emotional problems in children with ADHD: Mediation of coping strategies. Scand. J. Psychol. (2020.0) 2 141-149. DOI: 10.1111/sjop.12680 5. Bode A.A., George M.W., Weist M.D., Stephan S.H., Lever N., Youngstrom E.A.. The impact of parent empowerment in children’s mental health services on parenting stress. JCFS (2016.0) 10 3044-3055. DOI: 10.1007/s10826-016-0462-1 6. Yang X., Ke S., Gao L.L.. Social support, parental role competence and satisfaction among Chinese mothers and fathers in the early postpartum period: A cross-sectional study. Women Birth (2020.0) 3 e280-e285. DOI: 10.1016/j.wombi.2019.06.009 7. Berryhill M.B.. Mothers’ parenting stress and engagement: Mediating role of parental competence. Marriage Fam. Rev. (2015.0) 5 461-480. DOI: 10.1080/01494929.2015.1113600 8. Gordo L., Oliver-Roig A., Martínez-Pampliega A., Iriarte Elejalde L., Fernández-Alcantara M., Richart-Martínez M.. Parental perception of child vulnerability and parental competence: The role of postnatal depression and parental stress in fathers and Mothers. PLoS ONE (2018.0) 8. DOI: 10.1371/journal.pone.0202894 9. Huang Y.P., Chang M., Chi Y.L., Lai F.C.. Health-related quality of life in fathers of children with or without developmental disability: The mediating effect of parental stress. Qual. Life Res. (2013.0) 1 175-183. DOI: 10.1007/s11136-013-0469-7 10. Butun A., Linden M., Lynn F., McGaughey J.. Exploring parents’ reasons for attending the emergency department for children with minor illnesses: A mixed methods systematic review. Emerg. Nurs. (2018.0) 1 39-46. DOI: 10.1136/emermed-2017-207118 11. Barwise-Munro R., Morgan H., Turner S.. Physician and Parental Decision-Making Prior to Acute Medical Paediatric Admission. Healthcare (2018.0) 3. DOI: 10.3390/healthcare6030117 12. Butun A., Hemingway P.. A qualitative systematic review of the reasons for parental attendance at the emergency department with children presenting with minor illness. Int. Emerg. Nurs. (2017.0) 36 56-62. DOI: 10.1016/j.ienj.2017.07.002 13. Burokiene S., Raistenskis J., Burokaite E.. Factors determining parents’ decisions to bring their children to the pediatric emergency department for a minor illness. Med. Sci. Monit. (2017.0) 23 4141-4148. DOI: 10.12659/MSM.902639 14. Holly L.E., Fenley A.R., Kritikos T.K., Merson R.A., Abidin R.R., Langer D.A.. Evidence-base update for parenting stress measures in clinical samples. Clin. Child Adolesc. Psychol. (2019.0) 5 685-705. DOI: 10.1080/15374416.2019.1639515 15. Louie A.D., Cromer L.D., Berry J.O.. Assessing parenting stress. Fam. J. (2017.0) 4 359-367. DOI: 10.1177/1066480717731347 16. Nielsen T., Pontoppidan M., Rayce S.B.. The Parental Stress Scale Revisited: Rasch-based construct validity for Danish parents of children 2–18 years old with and without behavioral problems. Health Qual. Life Outcomes (2020.0) 1 281. DOI: 10.1186/s12955-020-01495-w 17. Mixão M., Leal I., Maroco J.. Parenting Stress Scale. Assessment in Sexuality and Parenting (2010.0) 18. Oronoz B., Alonso I., Balluerka M.N.. A Spanish adaptation of the parental stress scale. Psicothema (2007.0) 4 687-692 19. Carretero-Dios H., Pérez C.. Standards for the development and review of instrumental studies: Considerations about test selection in psychological research. Int. J. Clin. Health Psychol. (2007.0) 7 863-882 20. Lloret-Segura S., Ferreres-Traver A., Hernández-Baeza A., Tomás-Marco I.. El análisis factorial Exploratorio de los ítems: Una Guía Práctica, revisada y actualizada. An. Psicol. (2014.0) 3 1151-1169. DOI: 10.6018/analesps.30.3.199361 21. Ferrando P., Lorenzo-Seva U., Hernández-Dorado A., Muñiz J.. Decálogo para el Análisis Factorial de los Ítems de un Test. Psicothema (2022.0) 1 7-17. DOI: 10.7334/psicothema2021.456 22. Perpiñá-Galvañ J., Richart-Martínez M., José Cabañero-Martínez M.. Reliability and validity of a short version of the STAI anxiety measurement scale in respiratory patients. Arch. Bronconeumol. (2011.0) 4 184-189. DOI: 10.1016/S1579-2129(11)70044-1 23. 23. IBM IBM SPSS Statistics for Macintosh, version 21.0IBMArmonk, NY, USA2017. IBM SPSS Statistics for Macintosh, version 21.0 (2017.0) 24. 24. R Core Team R: A Language and Environment for Statistical ComputingR Foundation for Statistical ComputingVienna, Austria2019Available online: https://www.R-project.org/(accessed on 1 January 2023). R: A Language and Environment for Statistical Computing (2019.0) 25. Lim C.R., Harris K., Dawson J., Beard D.J., Fitzpatrick R., Price A.J.. Floor and ceiling effects in the OHS: An analysis of the NHS PROMs data set. BMJ Open (2015.0) 5 e007765. DOI: 10.1136/bmjopen-2015-007765 26. Rhemtulla M., Brosseau-Liard P.É., Savalei V.. When can categorical variables be treated as continuous? A comparison of robust continuous and categorical SEM estimation methods under suboptimal conditions. Psychol. Methods (2012.0) 17 354-373. DOI: 10.1037/a0029315 27. Rosseel Y.. Lavaan: Latent Variable Analysis. (2018.0) 28. Beauducel A., Herzberg P.Y.. On the performance of maximum likelihood versus means and variance adjusted weighted least squares estimation in CFA. Struct. Equ. Model. (2006.0) 13 186-203. DOI: 10.1207/s15328007sem1302_2 29. Hu L., Bentler P.M.. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct. Equ. Model. A Multidiscip. J. (1999.0) 6 1-55. DOI: 10.1080/10705519909540118 30. Zumbo B.D., Gadermann A.M., Zeisser C.. Ordinal versions of coefficients alpha and theta for likert rating scales. J. Mod. Appl. Stat. Methods (2007.0) 1 21-29. DOI: 10.22237/jmasm/1177992180 31. Viladrich C., Angulo-Brunet A., Doval E.. A journey around alpha and omega to estimate internal consistency reliability. An. Psicol. (2017.0) 33 755-782. DOI: 10.6018/analesps.33.3.268401 32. Luján-Tangarife J.A., Cardona-Arias J.A.. Construcción y validación de escalas de medición en salud: Revisión de propiedades psicométricas. Arch. Med. (2015.0) 11 1-11 33. Reglamento (UE) 2016/679 del Parlamento Europeo y del Consejo, de 27 de Abril de 2016, Relativo a la Protección de Las Personas Físicas en lo Que Respecta al Tratamiento de Datos Personales y a la Libre Circulación de Estos Datos y Por el Que Se deroga la Directiva 95/46/CE (BOE de 27 de Abril) 34. Kölch M., Lübke L., Müller S., Reis O., Brähler E., Lincke L.. Stress durch elternschaft: Psychometrische Evaluation der deutschen version der parental stress scale in der Allgemeinbevölkerung. Med. Psychol. (2022.0) 9 445-451. DOI: 10.1055/a-1778-5316 35. Harding L., Murray K., Shakespeare--Finch J., Frey R.. Understanding the parental stress scale with a foster carer cohort. Fam. Relat. (2020.0) 4 865-879. DOI: 10.1111/fare.12483 36. Nærde A., Sommer-Hukkelberg S.. An examination of validity and reliability of the parental stress scale in a population-based sample of Norwegian parents. PLoS ONE (2020.0) 12. DOI: 10.1371/journal.pone.0242735 37. Algarvio S., Leal I., Maroco J.. Parental stress scale: Validation study with a Portuguese population of parents of children from 3 to 10 years old. J. Child Health Care (2018.0) 4 563-576. DOI: 10.1177/1367493518764337 38. Zelman J.J., Ferro M.A.. The Parental Stress Scale: Psychometric properties in families of children with chronic health conditions. Family Relat. (2018.0) 2 240-252. DOI: 10.1111/fare.12306 39. Pontoppidan M., Nielsen T., Kristensen I.H.. Psychometric properties of the Danish Parental Stress Scale: Rasch Analysis in a sample of mothers with infants. PLoS ONE (2018.0) 11. DOI: 10.1371/journal.pone.0205662 40. Leung C., Tsang S.K.M.. The Chinese Parental Stress Scale: Psychometric evidence using Rasch Modeling on clinical and nonclinical samples. J. Pers. Assess. (2010.0) 1 26-34. DOI: 10.1080/00223890903379209 41. Park S.Y., Kim J.H., Jeong M.Y., Lee Y.S., Shin J.E., Lee Y.J.. Reliability and validity of the Korean version of the Parental Stress Scale for children with attention-deficit/hyperactivity disorder. Psychiatry Investig. (2021.0) 12 1188-1197. DOI: 10.30773/pi.2021.0116 42. Ekizoglou C., Vousoura E., Anagnostopoulos D., Papageorgiou C., Zervas I.M.. Validation of parental stress scale in a Greek sample of mothers with infants from 0 to 12 months. Infant Ment. Health J. (2022.0) 2 300-310. DOI: 10.1002/imhj.21971 43. Brito A., Faro A.. Diferenças por sexo, Adaptação e validação da escala de estresse parental. Avaliação Psicológica (2017.0) 1 38-47. DOI: 10.15689/ap.2017.1601.05 44. Bassi G., Mancinelli E., Di Riso D., Salcuni S.. Parental stress, anxiety and depression symptoms associated with self-efficacy in paediatric type 1 diabetes: A literature review. Int. J. Environ. Res. Public Health (2020.0) 1. DOI: 10.3390/ijerph18010152 45. Ren J., Li X., Chen S., Chen S., Nie Y.. The influence of factors such as parenting stress and social support on the state anxiety in parents of special needs children during the COVID-19 epidemic. Front. Psychol. (2020.0) 11 565393. DOI: 10.3389/fpsyg.2020.565393 46. Wilkie C.F., Ames E.W.. The relationship of infant crying to parental stress in the transition to parenthood. J. Marriage Fam. (1986.0) 3 545-550. DOI: 10.2307/352040 47. Canévet J.P., Bourgogne C., Querbès B., Jourdain M., Vrignaud B., Gras-Leguen C.. Recours inappropriés aux urgences pédiatriques par les parents: Approche qualitative d’un comportement multidimensionnel. Ann. Fr. Med. D’urgence (2018.0) 5 295-300. DOI: 10.3166/afmu-2018-0033
--- title: Developing and Implementing an Action Plan among the “Orang Asali” Minority in Southernmost Thailand for Equitable Accessibility to Public Health Care and Public Services Following the United Nations Sustainable Development Goals authors: - Praves Meedsen - Chutarat Sathirapanya journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049293 doi: 10.3390/ijerph20065018 license: CC BY 4.0 --- # Developing and Implementing an Action Plan among the “Orang Asali” Minority in Southernmost Thailand for Equitable Accessibility to Public Health Care and Public Services Following the United Nations Sustainable Development Goals ## Abstract Ending social inequality by 2030 is a goal of the United Nations’ endorsed sustainable development agenda. Minority or marginalized people are susceptible to social inequality. This action research qualitatively evaluated the requirements for and barriers to full access to public services of the Orang Asali (OA), a minority people living in the Narathiwas province in southernmost Thailand. With the cooperation of the staff of the Southern Border Provinces Administrative Center (SBPAC), we interviewed the OA, local governmental officers and Thai community leaders regarding the OA’s living conditions and health status. Then, an action plan was developed and implemented to raise their living standards with minimal disruption to their traditional cultural beliefs and lifestyle. For systematic follow-ups, a Thai nationality registration process was carried out before the assistance was provided. Living conditions and livelihood opportunities, health care and education were the main targets of the action plan. Universal health coverage (UHC), according to Thai health policy, was applied to OA for holistic health care. The OA were satisfied with the assistance provided to them. While filling the gap of social inequality for the OA is urgent, a balance between the modern and traditional living styles should be carefully considered. ## 1. Introduction For better well-being for all and at all ages by 2030, the “Sustainable Development Goals (SDGs)” agenda was endorsed by the United Nations in 2015. No poverty (Goal 1), good health and well-being (Goal 3) and quality education (Goal 4) were highlighted among the seventeen goals [1]. Poverty and illiteracy are common causes that limit individuals’ accessibility to public health care, leading to poor hygiene and illness. Additionally, poverty, illiteracy and illness have been considered factors of a vicious cycle of unhealthy lives. Therefore, providing equitable access to health care along with quality education and raising economic standards should be strategies for improving the well-being of individuals in a society. In Thailand, the national “Universal Health Coverage (UHC)” payment scheme for the health service was launched in 2002, so that all Thais could access the public health care system equitably from birth until death. The UHC is composed of various aspects that complete the health service loop, i.e., health promotion, prevention, treatment and rehabilitation, which aim at healthy lives for all in the country. Moreover, other public services necessary for daily living (i.e., electricity, piped water and sanitation), standard education, occupational training, employment assistance, etc., should be available equitably among the people of a nation. However, there are gaps in the nationwide coverage of the UHC program as well as other public facilities in Thailand, especially among minority groups or marginalized peoples, migrant workers, nomads, etc. Inaccessibility to the public health care system and a lack of health knowledge among these people to manage their own health cause them to be highly vulnerable to unhealthy living conditions and to acquire either communicable or non-communicable diseases (NCDs), most of which are preventable. Providing accessible health services and other facilities necessary for better living conditions to these special groups of people should be systematically organized and accepted by these target groups of people. The “Orang Asali” (OA), which means “the first people” or” original people”, is a group of indigenous people who have lived dispersedly throughout the Malay peninsula and southernmost Thailand, where it bordered with Malaysia, for 25,000 to 60,000 years [2,3]. According to anthropological information, the OA comprise three ethnically different groups, i.e., the Senoi, Porto-Malay (or Aborigine Malay) and Negrito [4]. The Senoi are the largest group of OA in Malaysia, while the OA in Thailand are mostly Negrito. In Thailand, the OA people live primarily in two mountain ranges, the Banthut mountain range in the Phattalung, Satun, and Trang provinces and the Sunkalakiri mountain range in the Yala and Narathiwas provinces. The latter group is close to the Thai–Malaysia border; therefore, their cultural beliefs and lifestyles are similar to those of the OA in Malaysia. The available research related to the OA in Thailand to date mainly involve anthropology, social status, and lifestyle rather than living conditions, health care services and education. According to a systematic review, the OA in Malaysia commonly suffered from malnutrition, lower growth rates in children, soil-transmitted helminths, pulmonary diseases and cardiometabolic diseases [5]. These acquired diseases also occur among the OA in Thailand due to similar living conditions and environments. To promote secure and healthy living conditions among the OA in Thailand, various interventions have been applied, such as providing these groups with Thai citizenship and accessibility to standard education, especially for school-aged children; health education and health promotion programs; and resettlement areas prepared for their permanent residence. According to the policy of the Ministry of the Interior of Thailand (MOI-T), OA in Thailand are considered Thai citizens, as other Thais are. Numerous operational plans have been deployed to improve their living conditions and increase their ability to access all public services and facilities. In the public health sector, the provision of equitable access to public health services is an aim of health care system management for the OA based on the respect for their human rights, as for other Thais, and in response to one of the SDGs. The UHC is a key measure to eliminate inequitable accessibility to health care service [6]. For some time, the OA in the study area have received basic living support, including medical care from the local government when the requested. We expect that the current registration program for the OA as Thais will solidify a sustainable life-supporting system for them. The registered OA would then be able to access Thai public services, as other Thai citizens are able to. Furthermore, the outcomes and progression for the improvement of their living conditions can be followed systemically. Herein, we describe a pilot program of actions to promote the OA’s voluntary adoption of public services and facilities, including public health care services among a group of OA living in the Chanae district of Narathiwas province. ## 2.1. Study Population and Setting This was a qualitative study using semi-structured interviews for data acquisition. The study participants were nine OA leaders and their representatives selected by the OA villagers from Toapaku and Biyis villages (OA); five local governmental personnel working in the offices of civil registration, education, agriculture and public health (GP); and six local Thai community leaders or associate leaders (TC), e.g., the heads of Thai local villages, the head of sub-district offices, religious leaders, etc. We selected two of five OA villages in the area in which the villagers had settled permanently as the study sites. Two weeks after the research information, objectives and process were clearly described to the OA villagers via translators, verbal consents were obtained voluntarily from the OA villagers because they are not able to understand spoken or written Thai. Additionally, the GP and TC groups were informed of the study process and written consents were obtained. The whole research study process was conducted in the OA settlement areas and the local governmental offices in Chanae District, Narathiwas province, with the support of the Southern Border Provinces Administrative Center (SBPAC), which acted as the coordinator for the related local governmental agencies. ## 2.2. Preparation for Data Collection After ethical approval and consent from the study participants was obtained, the research team started to gather preliminary information regarding the residential locations and the environments of the two study OA villages, their usual lifestyle and, significantly, their willingness to adopt Thai citizenship. On the official side, the relevant Thai laws or regulations and practical guidelines for verification of the OA as Thai people were reviewed. The preliminary information included an initial interview with the GP and TC groups regarding what had been carried out previously and the associated outcomes. Then, the in-depth interview questions were designed and tested for content validity by three experts in anthropology, public health and qualitative research. The questions were divided into three sections according to the research participants, i.e., the OA, GP and TC groups. We used translators who understood the OA spoken language as assistants in data collection. We visited and interviewed the OA participants in their homes, where we spent an average of two hours to complete each interview. We spent an average of one hour to interview the GP and TC participants at their offices. The interviewed content was recorded on audio recorders for later review and validation. The questions used for the interviews with the study participants were as follows: For the OA: ○After receiving the information from local Thai officers, are you willing to adopt Thai citizenship and why?○After receiving the information from local Thai officers, do you understand your rights and responsibilities after you become Thai.○In the past, how did you receive information about the registration process?○How do you contact other persons who are not OA like you?○What are the activities in your daily life?○Have there been any recent changes to your current living conditions, and if so, what are they?○How do you regularly manage your living areas, food supply and treatment for sick persons?○What is your perception of how your life and the life of your community might change if you accept Thai nationality?○Do you currently access any Thai official services, such as livelihood assistance, education and health care? For GP: ○How many groups of OA are currently living in the Chanae district?○How many families are there in each group, and how many people per family?○What are their houses (“Tub” in Thai) built of?○In what kinds of environmental conditions do they usually build their houses?○What means do Thai officials use to contact the OA and who is/are the mediator/s helping the contacts?○What will be the steps for providing Thai identification cards to the OA?○What actions or services will the Thai governmental agencies provide for the OA, e.g., livelihood assistance, education or health care, after they receive Thai citizenship?○How do you expect an OA to perceive the changes in living conditions or services that will be provided to them after receiving Thai citizenship? For TC: ○What kinds of settlements (permanent or migratory) do they (OA) have?○How long have they settled here?○What are their houses (“Tub” in Thai) built of?○In what kinds of environmental conditions do they usually build their houses?○What harvestable natural products do they use or consume, and how do they harvest them?○What were the previous living conditions of the OA who are going to be registered as Thai citizens?○What factors do you think will encourage an OA individual to accept the official offer to be registered as Thai? ## 2.3. Data Analysis Before we started the analysis, the GP and TC reviewed and validated the in-depth interview content themselves, while an independent translator was used to ensure the correctness of the translation of the OA’s interview content. We performed data analysis following the “Thematic Analysis” principle [7,8], which used six steps: [1] data familiarization and writing familiarization notes, [2] systematic data coding, [3] generating initial themes from coding and collated data, [4] developing and reviewing themes, [5] refining, defining and naming the themes, and [6] writing the report. After the analysis, an action plan was co-designed by the study team staff, GP and TC based on the results from the thematic analysis and the designed action plan was implemented among the OA study participants. We performed a short-term outcome evaluation, and long-term evaluation was also planned to be done in the future (Figure 1). ## 2.4. Ethical Considerations Ethical approval for the study was granted by the Ethics Committee of Public Policy Institute, Prince of Songkla University (EC code: $\frac{008}{64}$, date of approval $\frac{10}{06}$/2021). We strictly followed the 1964 Declaration of Helsinki, its amendments and related guidelines for the ethical conduct of research studies. All the participants’ identifiable information was completely anonymized. ## 3.1. Study Participants’ Characteristic We enrolled nine OA leaders and their representatives selected by the villagers of the Toapaku and Biyis villages, five local governmental personnel (GP) and six local Thai community leaders or associates (TC) of the Chanae District, Narathiwas province. The characteristics of the study participants are shown in Table 1. ## 3.2. Preliminary Information from the Initial Survey Initially, we interviewed the GP and TC groups on three topics for preliminary information before in-depth interviews and subsequent action planning were carried out. The aims of this preliminary interview were to evaluate the preparedness of the study participants from the Thai official sector, who would be involved in the process of planning further action, as well as the current Thai regulations. The topics of the preliminary interview included the following. ## 3.2.1. The Settlement Areas and the Lives of the OA in the Study Area Overall, there were five villages of OA in the study area, of which the villagers in three villages lived nomadically by hunting or harvesting natural forest products on the mountain, while the remaining two groups, i.e., the Toapaku and Biyis villages, had permanent settlements. The Toapaku village had six households with 32 members, while the Biyis village had five households with 27 members. The leaders of both villages were males whose leadership was derived from their ancestors. They spoke OA, Malay or, less frequently, the Thai language. The preliminary information obtained from the talks with the GP and TC groups in the area included the following: ## 3.2.2. Current Thai Law and Regulations and Previous Experience of Granting Thai Citizenship to OA in the Other Provinces The district head governor and his staff followed the relevant policies of the MOI-T and studied the laws and regulations applicable to this issue. They set up legally based portals for the OA to receive Thai citizenship. Initially, the OA were clearly informed about the steps required to obtain Thai citizenship, and their rights and responsibilities after becoming Thai citizens according to the Thai laws. Significantly, it was emphasized to the OA that their acceptance of Thai citizenship was voluntary if they fulfilled the required legal criteria for becoming Thai. The local governor’s teams also studied a previous successful project of verification and providing Thai citizenship to OA carried out in the Betong district of the Yala province as a model. ## 3.2.3. Evaluation of the OA’s Willingness to Adopt Thai Citizenship and the Barriers of the Verification Process for the OA as Thais Two months prior to our interview, the district governor’s team staff evaluated the OA’s willingness to adopt Thai nationality by informing them of their rights and responsibilities as Thais before they could decide freely. At the same time, they prepared the OA for our research team to perform the in-depth interviews for data collection. ## 3.3. In-Depth Interview Results, the Developed Action Plan and Actions Performed The research team, with the assistance of the local governor’s staff, visited the OA who lived permanently on the Dusongyor mountain to carry out in-depth interviews for data collection. We interviewed the OA regarding their living conditions and health care, education and other public services they required. We once again explained the rights and the responsibilities they would have after they were registered as Thai citizens before their willingness to accept the offer was confirmed. We first traced the evidence to confirm their longtime settlement and their relative links with other OA members living in this area. If the OA fulfilled these criteria according to the MOI-T’s policies and regulations, civil registration process as Thais was done; and an IDC and IDN were eventually given to them. The district government strictly followed the guidelines for the civil registration process issued by the MOI-T and strongly emphasized the preservation of the OA’s traditional ways of living despite their new nationality as Thai people. ## 3.3.1. Visiting the OA Living Areas to Ask Their Willingness We found that the villagers of the two villages were willing to be registered as Thais and to comply with Thai laws. They understood their rights as granted by Thai officials and the obligatory responsibilities to Thai society, whilst their traditional ways of living would be preserved. Their reasons for the adoption of Thai citizenship were that their children could attend school, they could receive health care and other public services, participate in health promotion programs, and receive an adequate food supply. The medical treatment among the OA depended on ancestral practices. For example, every pregnant woman went through childbirth naturally with assistance from a village midwife without a prenatal evaluation of maternal and fetal risks. No vaccinations for newborns or the aged were provided. Although they had experience in the medicinal properties of many natural products used as medicines, many complicated diseases were unable to be treated successfully. ## 3.3.2. Clearance of Legal Issues and Preparation for Providing the OA with Thai Nationality To save the expense and time travelling to the various district offices to complete the steps of verification and registration process by the OA themselves, the district governor’s team and associated local governmental agencies visited the OA settlements to complete the process, following which, the Thai IDCs and IDNs were given to them. ## 3.3.3. Theme Development After the completion of the study participant interviews, the themes were developed under the thematic analysis disciplines as follows: ThemesCodesA. Preparation for the registration process for the OA1. A plan for improvement of the overall well-being of the OA is required.2. Respect their equitable human rights as Thai people.3. Compliance with legal standard practices for the provision of Thai citizenship to the OA4. Learning from the previous practices of the other governance areas. Living areas and livelihoods of the OA Two of five OA villages were settled permanently, whilst the rest of remain nomadic. The harvestable forest products have progressively diminished over recent years, leading to inadequate resources to maintain adequate household consumption. Some OAs worked as laborers in the Chanae district, Narathiwas, to earn a living. OA children had malnourishment and illiteracy that affects their growth and development. The OA demand that they will be allowed to maintain their traditional ways of living with the forest after they are registered as Thai. Livelihood assistance, health care and all-level education suitable for each OA individual’s needs are necessary. The OA used ancestral methods to treat illnesses they experienced. They had no knowledge of health prevention or promotion. They were very anxious when they needed to visit a hospital due to their misunderstanding of current medical treatments. Studying current Thai laws, regulations and national policies to facilitate the process of the provision of Thai citizenship to the OA by local governmental staff. Because the OA are regarded as Thai people, the Ministry of the Interior, Thailand (MOI-T), instituted a policy for the registration of the OA as Thai citizens. Related acts and regulations, including guidelines for the verification of the Thai citizenship process will be reviewed and discussed among the governmental sectors at both the national and local levels. The local governmental agencies will collaborate in planning and carrying out the registration process. The successful registration process carried out in the Betong district of the Yala province was studied as a model. The OA are willing to adopt Thai nationality. Detailed information regarding the OA’s rights and responsibilities as Thai citizens was provided before their voluntary decision. Their ethnic identities and lifestyle will be preserved. Earning a living, health care access and basic or vocational education for children or adults, respectively, were considered essential for the OA. B. Registration process according to the developed action plan1. The legal process of verification and registration as Thai people.2. Registered OA obtain equitable rights and have the same responsibilities as the other Thais.3. Collaborative work of related local government agencies is a key for successful registration.4. Living conditions, health service and education are the three main targets for the development of the OA’s well-being.5. Health services under the UHC payment scheme is essential for the OA to access health care. The local governmental staff visited the OA living area on the mountain to ask about their willingness to be registered as Thai people. The OA would like to accept the conditions after registration as Thai. The registration to accept Thai citizenship is voluntary. Personal verification will be carried out by local government staff based on the MOI-T’s policies and regulations. Language was an information provision barrier. The registration practice successfully carried out in the Betong district, Yala province, is to be followed. Thai IDCs and IDNs will be provided to OA aged 7 years or over, and they will be listed in the Thai civil list after completing the verification process. A parcel of land for residence or earning a living will be provided. Health insurance under the UHC payment scheme of the Thai public health system will be provided to the OA.UHC will support payment for the OA to receive health services. OA children will be allowed to attend public schools to study Thai. C. Outcome evaluation1. Immediate after-action evaluation of the registration process outcomes focusing on improved living conditions, health services accessibility and children’s education. 2. Long term follow-ups and repeated evaluations in the future are planned. The OA’s homes were redesigned for hygienic living. Follow-ups will be carried out to ensure the OA have received equitable social support and access to welfare programs as Thai citizens. Evaluation of the understanding of and accessibility to public health services under the UHC payment scheme among the OA and their satisfaction with the services will be evaluated. Teachers and local Thai community leaders will encourage OA children to attend a primary school. The OA parents and their children’ satisfaction with the organized education system will be evaluated. ## 3.3.4. Official Provision of Thai Citizenship to the OA According to the Plan Nearly 3 months after the survey of the OA’s willingness to be registered as Thai and the preparation by the district governmental agencies, the Chanae district governor’s team and the associated agencies managing the legal issues and planning the process for registration and provision of Thai citizenship, provided Thai IDCs and IDNs to the OAs. ## 3.4. Post-Action Short-Term Evaluation After the registration process for the OA was completed, we carried out a follow-up visit a few weeks later to assess how frequently the OA accessed public support, as well as their satisfication. We found that the OA were satisfied with the help of the local governmental agencies to improve their well-being and quality of life. Since they received their Thai citizenships, every individual OA was able to own a parcel of land to plant crops or raise livestock, access the public health care or health promotion services, and their children were able to attend the local primary schools, etc. Moreover, the informal education and vocational training were provided for OA who were 15 years or older. Scheduled classes were regularly organized in the community-shared building in which 10–20 OA youths attended each class. To assist in the improvement of their living conditions, they were taught to crop vegetables and raise livestock to maintain their food security. Additionally, seeds, baby chickens, baby ducks, etc., were provided. Regarding public health services, thirteen and five OA individuals voluntarily received one and two doses of COVID-19 vaccine, respectively. They responded well to the COVID-19 vaccination campaign. They still preferred to use traditional herbal medicines for initial treatment, except for a complicated disease for which they were willing to receive modern medical treatment from the Thai community health volunteers, who regularly visited them, or in the district hospital when it was required. The number of OA utilizing modern medical services increased. ## 4. Discussion One principal item in the Thai constitution emphasizes equitable rights of all Thais to access and receive public services or support. The OA in this study, as well as other minority people in Thailand, have the rights to receive public support according to the rights outlined in the clause of Thai constitution. Hence, the verification and registration process for the OA in the Chanae district of Narathiwas province was undertaken. All aspects of the quality of life of the study OA are significantly affected due to their migratory living style, which depends on the quantity of natural products harvestable or wild animals caught for adequate consumption. Good livelihood for ending hunger and poverty (SDG 1 and 2), equitable access to health care (SDG 3) and quality education (SDG 4), the three of the seventeen SDGs endorsed by the UN, have been prioritized by Thai official agencies as the primary targets for upgrading the living conditions of all Thais, including the OA in this study. Additionally, reducing inequality (SDG 10) in accessing public support services with the aim for achieving the three targeted SDGs equitably was stressed in this project. The verification and registration process for the OA in the current study was the first and principal action which initiated the cooperation among the related local governmental agencies under the administration of the SBPAC. The actions performed were based on the thematic principle of establishing equitable accessibility to public support, as other Thai citizens are able, without the significant disruption of the OA’s identity and traditional living. Many previous projects in Thailand and Malaysia involving the resettlement of the OA to new living areas failed because their traditional living styles were abruptly changed due to the policies being implemented without receiving their agreement beforehand. The abrupt changes from traditional to modernized living conditions adversely affected the OA’s traditional ways of life and cultural practices. Most of the resettled OAs soon left the new housing provided by the officials and returned to their previous living sites in the forest. We learned from our previous experiences and were aware that it is necessary to balance the preservation of the identity of this ethnic group and their traditional living style with the officially supported modern living. If the changes were not familiar to the OA, they would reject the offers. Compulsory changes by the provision of certain support programs, despite seemingly useful from the provider’s perspective, commonly bring about conflict or project failure. For these reasons, the SBPAC first conducted integrative actions of local governmental agencies by surveying the OA’s requirements, living styles, cultural beliefs and practices and, especially, their willingness to adopt the registration and development programs provided for the improvement of their well-being. After the legal conditions were fulfilled and the OA’s consent to join the program was obtained, then the integrative actions were started. The program in this study prioritized the improvements of the OA’s livelihoods, health care and education as initial and urgent targets for receiving local governmental support, because these were considered powerful influencers that interactively affected individuals’ well-being. It was known that the OAs lived by harvesting natural forest products for their daily household consumption. They had no knowledge of how to plant vegetables or other plants or to raise livestock for their food reserves. Normally, they followed a nomadic lifestyle, migrating to a new location in the mountain range every 7–10 days on average, when the harvestable forest products in their current living area became inadequate for consumption. The increased Thai population and accompanying requirements of land for agriculture and forest industries led to ecological changes in the forests. Both natural and man-made ecological changes in the forest have had a negative impact on the amount of forest products harvestable by the OA, resulting in an insecure food supply among the OA. This is the reason why some of the OA were required to permanently settle in a single living location or come down from the hills for labor work in the commercial area of the district. To reduce poverty and lack of food security in response to the first and second SDG goals, our program encouraged and supported the OA to settle permanently in locations along the forest margins, as well as teaching them planting and livestock raising techniques. After the preliminary discussions with the OA and the program were undertaken, we believe that this method of providing social support is suitable for and satisfies the OA very much, in that their ancestral lifestyle and beliefs have not been seriously affected. The traditional health beliefs among the OA were based on their strong belief in supernatural powers rather than their own inner power or self-efficacy or control in managing their own health. According to the health locus of control concept, they had a lower belief in an internal health locus of control than in an external health locus of control. This kind of belief causes adverse effects to a person’s health [9,10,11,12,13]. Additionally, they lacked the conceptual thoughts or knowledge necessary to generate an appropriate health belief model (HBM) [14] to care for their own health. This concept was also recently used to explain the lack of compliance for receiving a COVID-19 vaccine during the COVID-19 pandemic and vaccination campaign [15,16]. From our interviews with the OA, we learned they were very anxious when discussing modern medical care. Their long-held perception was that attending a hospital led to a dreadful outcome or death. With the help and psychological support from the local public health volunteers, they felt more secure and relaxed. Apart from medical treatment, we believe that the health education from the UHC program will enable them to voluntarily follow health prevention and promotion advice. Previous studies have found that improved knowledge followed by attitudes and practices together influenced soil-transmitted helminth (STH) infection control among the OA in Malaysia [17,18]. The UHC payment scheme in Thailand includes all aspects health services, i.e., health promotion, disease prevention, treatment and rehabilitation, and is available to all Thais from birth to death. Local public health volunteers are the points of first contact in the system when accessing health services. The UHC payment scheme ensures that all Thais will receive the holistic health services equitably. After the OA in this study were successfully registered as Thai citizens, they had equal rights as Thai people to access UHC programs. A study showed that the OA in Malaysia were found to have shorter life expectancies than the Malay people, with overall life expectancies of 53 years (54 years for females and 53 years for males) [19]. Common diseases diagnosed in the OA in Malaysia were STH infections, pulmonary diseases, liver diseases and malnutrition, all of which were occasionally severe enough to cause death [5,20]. We expect that the UHC program will be of considerable benefit for the healthy lives of the OA in this study. Education is another target of action planned to be promoted in parallel with improving living conditions and health care. Quality education, either formal or informal, can help the OA children or youths to understand Thai or train them vocationally to offer them more choices of career in the future. The OA in this study were encouraged to allow their children of school age to attend formal education, while informal (non-school) as well as vocational education programs were available for adult OA. Because the OA have only their own spoken language but no written language, teaching them the Thai language requires teachers who understand the OA language well. We found that most of the OA parents and children appreciated this offer of educational opportunity. The small sample size is a limitation of this study. The difficulty of travelling to the OA living areas where they reside on the hilltops and their lifestyle of hunting or harvesting in the forest during daytime are the causes of this issue. However, we made an effort to include all available OA, including their leaders, in our interviews. Male OA of a comparable age range were predominantly included for the interview, since they were the main group of OA making their livings and leading their family members’ lives. ## 5. Conclusions “We will never leave anyone behind”, a theme of social equality compatible with the UN-endorsed SDGs, was the major concept of the current action plan implemented for the OA communities in this study. We found that the OA in this study were satisfied with the officially provided support. Herein, we suggest that the changing of any indigenous people’s living conditions while aiming for their better well-being by an official project should carefully consider their traditional beliefs and practices. Changing living conditions or implementing obligatory public services that markedly disrupt a minority’s ancestral beliefs and lives and, significantly, without their willingness to adopt these changes as if they are sharing the ownership of the project, commonly result in unfavorable outcomes. Finally, we suggest that the long-term follow-ups of the OA’s accessibility to official services or support programs will elucidate the sustainable benefits and satisfaction among the recipients of the support. ## References 1. Sustainable Development Goals 2. Bellwood P.. Prehistory of the Indo-Malaysian Archipelago (2007.0) 3. Jinam T.A., Hong L., Phipps M.E., Stoneking M., Ameen M., Edo J., Saitou N.. Evolutionary history of continental southeast Asians: ‘Early Train’ hypothesis based on genetic analysis of mitochondrial and autosomal DNA data. Mol. Biol. Evol. (2012.0) 29 3513-3527. DOI: 10.1093/molbev/mss169 4. Lim L.S., Ang K.C., Mahani M.C., Shahrom A.W., Md-Zain B.M.. Mitochondrial DNA polymorphism and phylogenetic relationships of proto-Malays in peninsular Malaysia. J. Biol. Sci. (2010.0) 10 71-83. DOI: 10.3923/jbs.2010.71.83 5. Mahmud M.H., Baharudin U.M., Md Isa Z.. Diseases among Orang Asli community in Malaysia: A systematic review. BMC Public Health (2022.0) 22. DOI: 10.1186/s12889-022-14449-2 6. Wong Y.S., Allotey P., Reidpath D.D.. Sustainable development goals, universal health coverage and equality in health systems: The Orang Asli commons approach. Glob. Health Epidemiol. Genom. (2016.0) 1 e12. DOI: 10.1017/gheg.2016.8 7. Braun V., Clarke V.. Using thematic analysis in psychology. Qual. Res. Psychol. (2006.0) 3 77-101. DOI: 10.1191/1478088706qp063oa 8. Braun V., Clarke V.. Thematic Analysis: A Practical Guide (2021.0) 9. Lindström M., Rosvall M.. Health locus of control and mortality: A population-based prospective cohort study. Public Health (2020.0) 185 209-211. DOI: 10.1016/j.puhe.2020.05.005 10. Abredari H., Bolourchifard F., Rassouli M., Nasiri N., Taher M., Abedi A.. Health locus of control and self-care behaviors in diabetic foot patients. Med. J. Islam. Repub. Iran (2015.0) 29 283. PMID: 26913246 11. Bennett B.L., Goldstein C.M., Gathright E.C., Hughes J.W., Latner J.D.. Internal health locus of control predicts willingness to track health behaviors online and with smartphone applications. Psychol. Health Med. (2017.0) 22 1224-1229. DOI: 10.1080/13548506.2017.1317354 12. Náfrádi L., Nakamoto K., Schulz P.J.. Is patient empowerment the key to promote adherence? A systematic review of the relationship between self-efficacy, health locus of control and medication adherence. PLoS ONE (2017.0) 12. DOI: 10.1371/journal.pone.0186458 13. Tahmasebi R., Noroozi A.. Is Health Locus of Control a Modifying Factor in the Health Belief Model for Prediction of Breast Self-Examination?. Asian Pac. J. Cancer Prev. (2016.0) 17 2229-2233. DOI: 10.7314/APJCP.2016.17.4.2229 14. Rosenstock I.M.. Why people use health services. Milbank Mem. Fund Q (1966.0) 44 94-127. DOI: 10.2307/3348967 15. Arceo E., Jurado J.E., Cortez L.A., Sibug N., Sarmiento G.L., Lawingco A.C., Carbungco C., Tiongco R.E.. Understanding COVID-19 preventive behavior: An application of the health belief model in the Philippine setting. J. Educ. Health Promot. (2021.0) 10 228. DOI: 10.4103/jehp.jehp_1441_20 16. Shmueli L.. Predicting intention to receive COVID-19 vaccine among the general population using the health belief model and the theory of planned behavior model. BMC Public Health (2021.0) 21. DOI: 10.1186/s12889-021-10816-7 17. Nasr N.A., Al-Mekhlafi H.M., Ahmed A., Roslan M.A., Bulgiba A.. Towards an effective control program of soil-transmitted helminth infections among Orang Asli in rural Malaysia. Part 2: Knowledge, attitude, and practices. Parasites Vectors (2013.0) 6 28. DOI: 10.1186/1756-3305-6-28 18. Al-Delaimy A.K., Al-Mekhlafi H.M., Lim Y.A., Nasr N.A., Sady H., Atroosh W.M., Mahmud R.. Developing and evaluating health education learning package (HELP) to control soil-transmitted helminth infections among Orang Asli children in Malaysia. Parasites Vectors (2014.0) 7 416. DOI: 10.1186/1756-3305-7-416 19. Tuan Abdul Aziz T.A., Teh L.K., Md Idris M.H., Bannur Z., Ashari L.S., Ismail A.I., Ahmad A., Isa K.M., Nor F.M., Rahman T.H.. Increased risks of cardiovascular diseases and insulin resistance among the Orang Asli in Peninsular Malaysia. BMC Public Health (2016.0) 16. DOI: 10.1186/s12889-016-2848-9 20. Nasr N.A., Al-Mekhlafi H.M., Ahmed A., Roslan M.A., Bulgiba A.. Towards an effective control programme of soil-transmitted helminth infections among Orang Asli in rural Malaysia. Part 1: Prevalence and associated key factors. Parasites Vectors (2013.0) 6 27. DOI: 10.1186/1756-3305-6-27
--- title: 'Gut-Derived Uremic Toxins in CKD: An Improved Approach for the Evaluation of Serum Indoxyl Sulfate in Clinical Practice' authors: - Gianvito Caggiano - Loredana Amodio - Alessandra Stasi - Nicola Antonio Colabufo - Santina Colangiulo - Francesco Pesce - Loreto Gesualdo journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049313 doi: 10.3390/ijms24065142 license: CC BY 4.0 --- # Gut-Derived Uremic Toxins in CKD: An Improved Approach for the Evaluation of Serum Indoxyl Sulfate in Clinical Practice ## Abstract In the past years, indoxyl sulfate has been strongly implicated in kidney disease progression and contributed to cardiovascular morbidity. Moreover, as a result of its elevated albumin affinity rate, indoxyl sulfate is not adequately cleared by extracorporeal therapies. Within this scenario, although LC-MS/MS represents the conventional approach for IS quantification, it requires dedicated equipment and expert skills and does not allow real-time analysis. In this pilot study, we implemented a fast and simple technology designed to determine serum indoxyl sulfate levels that can be integrated into clinical practice. Indoxyl sulfate was detected at the time of enrollment by Tandem MS from 25 HD patients and 20 healthy volunteers. Next, we used a derivatization reaction to transform the serum indoxyl sulfate into Indigo blue. Thanks to the spectral shift to blue, its quantity was measured by the colorimetric assay at a wavelength of 420–450 nm. The spectrophotometric analysis was able to discriminate the levels of IS between healthy subjects and HD patients corresponding to the LC-MS/MS. In addition, we found a strong linear relationship between indoxyl sulfate levels and Indigo levels between the two methods (Tandem MS and spectrophotometry). This innovative method in the assessment of gut-derived indoxyl sulfate could represent a valid tool for clinicians to monitor CKD progression and dialysis efficacy. ## 1. Introduction As a result of chronic kidney disease (CKD) progression, the gradual loss of the glomerular filtration rate represents the driving force for the accumulation of toxic solutes involved in uremic syndrome. The hefty retention of these compounds in the body contributes to detrimental consequences and is strongly associated with poor outcomes in CKD patients [1,2]. In the past years, a considerable number of harmful molecules have been identified and grouped in line with their size, structure, chemical properties and ability to bind plasma proteins [3]. According to this classification, the first cluster of molecules is represented by small toxins (e.g., creatinine and urea), which are successfully cleared by traditional dialysis methods. The second group comprises middle-weight solutes such as large peptides, proteins, and lipoproteins (e.g., b2-microglobulin, parathyroid hormone, cytokines, etc.). Finally, due to their high affinity with blood proteins, a further class is represented by protein-bound uremic toxins (PBUTs) [4]. Of note, PBUTs include gut-derived compounds that originate from the proteolytic metabolism of intestinal microbiota. In detail, as a result of the decrease in renal filtration rate, urea accumulates in the intestinal lumen [5]. This uremic milieu supports the overgrowth of harmful microbial communities such as tyrosinase- and tryptophanase-positive bacteria. Consequently, the large retention of indoles, cresols, and phenols drives intestinal permeability, resulting in the translocation of such toxins into the blood [6,7,8]. A growing amount of data demonstrates that the elevated plasma concentration of PBUTs causes multi-organ damage. Notably, several PBUTs, including indoxyl sulfate (IS) and p-Cresyl Sulfate (pCS), are of particular interest for their detrimental effects associated with the exacerbation of renal damage and CKD-associated complications such as renal inflammation and fibrosis, cardiovascular morbidity, atherosclerosis thrombosis, and mortality. For instance, several in vivo and in vitro studies indicated that IS and pCS were implicated in vascular damage by inducing oxidative stress, senescence, and apoptosis in endothelial cells [9,10,11,12,13]. Moreover, at the renal level, indoles and cresols play a pivotal role in triggering intrarenal inflammation and tubulo-interstitial fibrosis by modulating several intracellular signals, including p53, AhR, Nf-kB, and klotho [14,15,16,17]. On the other hand, due to their strong pro-oxidant effect, the protein-bound solutes were shown to drive metabolic dysfunction by inducing insulin resistance, sarcopenia, and adipocyte injury [18,19,20]. IS levels are associated with poor prognosis, especially during the late stages of kidney disease [21,22,23]. In particular, the plasma levels of IS progressively increase from stage I to stage V, reaching a maximum in end-stage kidney disease (ESKD), indicating that the clearance of IS gradually decreases with the advanced decline of kidney function [23,24]. In addition, hemodialysis (HD) is unable to purify the blood from PBUTs since the plasma–protein bond hinders their clearance [24,25]. The introduction of high-flux membranes (HF) and the use of hemodiafiltration (HDF) improved the middle molecules’ removal by relying on larger pore sizes and increasing the convection mechanism. However, potential albumin and nutrient loss might occur and should be taken into account with these methods; in addition, they have limited effect on removing PBUTs [4,26]. As a consequence of this accumulation, HD patients exhibit elevated serum concentrations of indoxyl sulfate, resulting in detrimental conditions, including cardiovascular damage and poor survival rates [27]. In this scenario, the quantification of IS could aid clinicians in monitoring both the CKD progression and the efficiency of the dialysis treatment. Moreover, the early quantification of IS could represent a good strategy for predicting CKD-related complications such as cardiovascular events. Conventionally, liquid chromatography–tandem mass spectrometry (LC-MS/MS) represents the main diagnostic system in the measurement of IS thanks to its elevated performance [28,29,30,31]. Nevertheless, it requires dedicated skills and costly equipment, and results are obtained after a long period. In recent years, several working groups have improved the LC-MS/MS method for the quantification of blood IS. However, most of these methods still require laborious sample preparation and various steps to be carried out [32,33]. In this study, we aimed to develop a novel approach for rapidly estimating the blood’s indoxyl sulfate levels in CKD patients that could be implemented in clinical practice. Importantly, compared to the current systems, Theremino represents the new frontier in the context of spectrophotometry-based devices that can be used in clinical practice. In detail, Theremino is a smart spectrophotometer thanks to its compatibility with tablets, PCs, and smartphones since it is equipped with a USB connection. Therefore, the data can be quickly transmitted in a dataset or to a control room. Moreover, thanks to its small size (40 × 10 cm) and light weight (500 g), it is easily transportable to where the test will take place (i.e., outpatient or inpatient rooms) and can acquire data within a selected UV-Vis spectrum. In contrast, currently used spectrophotometric instruments are expensive and cannot be used in a clinical setting since some are rather unwieldy and heavy and require specific skills or laborious protocols. On this basis, considering that Indigo formation occurs in a few minutes and its spectrum is promptly recorded, it is possible to consider Theremino as a first-line tool for detecting the levels of indoxyl sulfate in the CKD population. To the best of our knowledge, no study has implemented the detection of this uremic toxin without using mass spectrometry as of yet. In addition, there are no published methods for IS quantification using a spectrophotometric assay for clinical settings. ## 2.1. Indigo Derivatization from IS The derivatization of IS by FeCl3 generates the Indigo-blue chromophore that is able to absorb light at λ = 450 nm. The summary of the derivatization reaction is shown in Figure 1. Briefly, we added 300 µL of Acetonitrile to the serum. After centrifugation, 100 µL of sulfuric acid solution (1:1) and 500 µL of FeCl3 0.001 M (in HCl 1N) were added to the supernatant. The mixture was held for 15 min at 70 °C. Finally, Indigo was measured by both LC-MS/MS and Theremino. ## 2.2. Calibration and Linearity Linear correlation coefficients (R) were determined based on the correlation between absorbance spectral values obtained with known concentrations of Indigo. The best wavelengths for estimating the concentration of Indigo were found by Lambert–Beer’s law: A(λ)= Ɛ M l, (A(λ) = UV absorbance value at certain wavelength; Ɛ = molar absorptivity coefficient for Indigo; M = molarity (mol/L)). A calibration curve was constructed by using six points for different concentrations of analytes (7.5, 15, 30, 50, 75, and 95 µM) (Figure 2). The calibration equation was $A = 0.0039$C + 0.017 (A = analyte absorption; C = relative concentration of the analyte). We confirmed the formation of Indigo with the derivatization reaction of indoxyl sulfate by LC-MS/MS analysis. ## Linearity The correlation coefficient of R2 = 0.9986 (Figure 2) proved linearity over the concentration range. The equation of the calibration curve was $y = 67.07$x + 2.25 for IS and $y = 973.04$x + 539.94 for pCS, where y represents the peak area ratio of analytes to DHTC, and x represents the relative concentration of the analytes. ## Accuracy Accuracy was evaluated at two different concentration levels (7.5 and 50 mg/mL) and with 9-fold injection. The accuracy was 97.6 ± $0.2\%$ for 7.5 mg/mL and 99.2 ± $0.1\%$ for 50 mg/mL of the analytical solution. ## Precision The limit of detection and the limit of quantification were calculated by equations LOD = 3.3 SD/s and LOQ = 10 SD/s, respectively. In the equation, s represents the slope of the calibration curve, and SD represents the standard deviation of the peak area. The acceptance criterion was LOD = 0.067 and LOQ = 0.204 for IS and LOD = 0.072 and LOQ = 0.218 for pCS. The intraday precision of the assay method was evaluated by carrying out nine independent assays of an analytic solution prepared at two concentration levels (7.5 and 50 mg/mL). For interday precision at each concentration level, a single injection of the solution was assayed daily for three consecutive days. The % relative standard derivations of intra- and interday assays were $0.5\%$ for 7.5 mg/mL and $0.8\%$ for 50 mg/mL, as well as $9.5\%$ for 7.7 mg/mL and $8.7\%$ for 50 mg/mL of the analyte. ## 2.3. Method Comparison Serum from both HD and healthy individuals was collected in order to measure IS and Indigo levels by both colorimetric assay (using the Theremino Spectrophotometer) and LC-MS/MS. IS levels measured by tandem mass spectrometry are shown in Figure 3A. As expected, the amount of serum IS in HD patients increased when compared to healthy controls. Next, from each serum sample, we performed the derivatization of the IS into Indigo by FeCl3 (Figure 1). At the end of each reaction, the total amount of IS-derived Indigo was measured by both a Theremino spectrophotometer and LC-MS/MS (Figure 3B,C). Notably, the levels of Indigo measured by Theremino were higher in the HD group compared with heathy people. Of note, these data indicate that the total amount of IS in the serum samples of HD patients completely reacted in producing Indigo blue. Additionally, the same result was obtained when Indigo from the same sample was quantified by standard Tandem MS. Importantly, the levels of both Indigo and IS in each group were highly comparable when measured by the Theremino spectrophotometer and LC-MS/MS, respectively (Figure 3D). Therefore, altogether, these results demonstrate that the new method based on Theremino was able to discriminate the levels of IS (using Indigo) between healthy subjects and HD patients corresponding to the LC-MS/MS. In order to evaluate differences between the free and protein-bound fractions, the amount of free indoxyl sulfate was measured by LC-MS/MS and Theremino. Interestingly, the quantification of IS-bound fractions observed with LC-MS/MS (bound form, >97 ± $1.9\%$) was comparable to the quantification of Indigo with Theremino (<94 ± $1.3\%$). Consequently, the unbound fractions of IS and Indigo were both less than $10\%$. To investigate the difference between the two methods, we related the levels of IS measured by LC-MS/MS to the levels of Indigo quantified by the Theremino spectrophotometer (Figure 4A) using the serum samples of HD participants. A linear correlation was represented by the regression line $y = 0.9$x + 9.86 with a relationship (R2) of 0.951 ($p \leq 0.001$). Furthermore, there was a significant association between the concentration of Indigo detected by each method (Figure 4B) and the levels of IS and Indigo detected by LC-MS/MS alone (Figure 4C) ($p \leq 0.001$). Finally, a congruent agreement was found between the quantification of indoxyl sulfate by the standard approach and the levels of Indigo detected by Theremino (Figure 4D). ## 3. Discussion As a result of kidney failure, CKD patients developed a strong retention of different uremic compounds, which are implicated in the advancement of kidney injury. In this scenario, HD patients are characterized by the highest levels of these toxins, including IS and pCS, which are closely related to poor prognosis [34,35]. Moreover, a large number of clinical reports also suggest that the level of serum IS is able to predict the progression of CKD since its incremental accumulation is directly correlated to the gradual failure of renal function [36,37,38]. Moreover, animal and human studies observed that the elevated concentration of IS is tightly associated with all-cause mortality [39,40]. Finally, the total amount of gut-derived indoxyl sulfate was also able to predict the prevalence of CKD-related complications, including endothelial injury, cardiac failure, metabolic dysfunction, and bone disease [41,42,43]. Altogether, these data highlight the importance of IS as a novel potential biomarker in the context of CKD since the current strategies seem to be inadequate in predicting the CKD-associated comorbidities promoted by protein-bound uremic toxins. Therefore, the rapid measurement of IS concentrations through the CKD stages could represent a reliable index for clinicians in predicting adverse outcomes and also assessing adherence to nutritional therapy [1,40,44,45]. Additionally, the assessment of IS amounts in HD populations could be employed to better evaluate the efficiency of dialysis therapies. Based on this statement, the evaluation of IS levels in CKD and ESKD patients can be employed as a measurement-of-risk rate with respect to kidney injury progression, CVD, and mortality [46,47]. Interestingly, it has been shown that during urethral infections, several sulfatase- and phosphatase-producing bacteria can cause urine color change to blue-purple. Interestingly, this biological process involves the enzymatic transformation of indoxyl sulfate into Indigo, which is promoted by microbic phosphatases or sulfatases [48]. The foundation of our work is the transformation of IS into Indigo by a chemical reaction and its spectrophotometric measurement. Chiefly, tandem mass spectrometry represents a widely used diagnostic approach used for the Indole measurement thanks to its great performance in sensitivity [49,50,51]. This method allows for higher sensitivity and accuracy; however, it requires dedicated equipment and skills for analyses, and results are obtained after a long period. Hence, to counteract these issues, we developed a real-time approach based on a spectrophotometric assay that is exploitable in clinical settings. We implemented chemical derivatizations in order to transform the total amount of IS into Indigo blue in serum samples from HD patients and healthy donors. This method allowed a rapid quantification of IS-derived Indigo at λ = 420–450 nm by using the mini-spectrophotometer Theremino. Moreover, to address the method’s feasibility, we compared the levels of Indigo measured with Theremino to the levels of IS and Indigo measured with LC-MS/MS. Both the specificity and the repeatability of Indigo levels monitored by Theremino were strongly correlated to the standard method. Since numerous studies show that the serum IS levels of healthy subjects are <10 μM and >20 μM for HD patients, we considered that the linearity together with the LLQ was satisfactory for sample measurements. Furthermore, our method was not influenced by external agents or other uremic toxins such as pCS, demonstrating its high specificity for IS. In our analysis, the spectrophotometric method showed that Indigo’s intensity showed a 3-fold increase in the patient group compared to the baseline amount. Remarkably, this result was closely comparable to the LC-MS/MS analysis of IS and IS-derived Indigo before and after derivatization, respectively. According to these data, the new method based on the Theremino spectrophotometer is characterized by high sensitivity in discriminating the levels of IS (by Indigo) between healthy subjects and HD patients corresponding to the standard method. Moreover, we found that the serum amount of the total and free fraction of indoxyl sulfate was highly comparable between LC-MS/MS and Theremino. This result was in line with previous data showing a high albumin-binding ratio of $90\%$ [52]. Additionally, we observed a significant direct association between Indigo measured by the Theremino spectrophotometer and IS measured by LC-MS/MS in serum samples of HD patients (R2 = 0.9512, $p \leq 0.001$). Finally, the correlation between the two methods using the Bland–Altman Plot highlighted a profound agreement between the techniques. Furthermore, the reliability of this novel test was justified by high reproducibility and good accuracy and was not influenced by the matrix effect. We observed that the sensitivity of this assay was not affected by the quality of blood sample (e.g., cloudy serum, blood clots, or freeze/thaw cycles). This study has some limitations. In particular, our analysis was carried out on a relatively small cohort of subjects (25 HD patients and 20 healthy individuals), and given the sample’s size, this could be considered a pilot study. In addition, this method is unable to simultaneously measure other biological markers, such as creatinine, albumin, or other toxins. It should also be noted that this method is unable to simultaneously measure further biological markers or toxins implicated in kidney pathophysiology. The colorimetric evaluation of IS-derived Indigo is not able to discriminate between the free and bound fraction of the analyte. Future studies are needed to split the free fraction of PBUT from the protein-bound fraction by using a differential separation of proteins in serum samples. It can be argued that although tandem MS is a high-ranking technique in the measurement of several toxins as well as the markers of CKD, it requires dedicated equipment and expert skills and does not allow real-time analysis. Our method allows size reductions and is cost-cutting. Moreover, this approach is easier and can be performed without the need of expert laboratory professionals. Importantly, the Theremino spectrophotometer can provide data in situ to promptly detect the level of IS in CKD and HD patients. Based on these data, clinicians will be able to evaluate the effects of treatments aimed at reducing uremic toxin levels, such as a low-protein diet, microbiota replacement, and oral adsorbent [17,53,54]. ## 4. Materials and Methods A multiresidue analysis by tandem MS was carried out in the estimation of the concentration of IS in the blood of both HD and the control group. Accordingly, the serum IS was derivatized by a rapid reaction in a stable chromophore called Indigo blue. The results of the chemical conversion of IS into Indigo were monitored by LC-MS/MS, demonstrating that the total amount of serum IS was completely converted into Indigo after 15 min. Ultimately, the total concentration of IS-derived Indigo was evaluated by using a smart UV-Vis spectrometer. Furthermore, we developed a smart device named “Theremino Spectrophotometer”, which can convert the absorbance level of Indigo (defined by the range of Indigo quantification from 10 mM to 400 mM in UV-VIS) in a concentration value of IS that can be digitally acquired and managed. ## 4.1. Patients We enrolled a total of 20 healthy volunteers and 25 HD patients from the Nephrology Unit of the “Policlinico di Bari’’, Bari, Italy. The study has been carried out in accordance with the Helsinki Declaration and approved by the Ethical Committee of the “Azienda Ospedaliero Universitaria Consorziale Policlinico”, Bari, Italy (approval number: 6137). All participants provided written informed consent. We recruited 25 male patients with CKD stage V on standard hemodialysis, and they were all Caucasian and aged 18 years and older (median age: 56 years). Standard hemodialysis consisted of four-hour sessions for a total of three times per week (dialysis vintage ≥ 3 months). Exclusion criteria were as follows: acute kidney injury (AKI), history of heart failure, and previous peritoneal dialysis or renal transplantation. Serum was collected from total blood samples before the hemodialysis session and centrifugated at 3000× g for 15 min. ## 4.2. Chemicals and Reagents For Tandem MS, internal standards including indoxyl sulfate potassium salt and ammonium acetate ($99\%$ purity) were supplied by Sigma-Aldrich (product of Germany). Methanol (CH3OH) and acetonitrile (MeCN; CH3CN) were purchased from Merck (Darmstadt, Germany). For the synthesis of Indigo, we used iron (III) chloride anhydrous salt with purity > $97\%$ (Fisher Chemicals-Waltham, MA, USA; Lot. 1915377); hydrochloric acid solution (1N) (Honeywell, Fluka, Lot. K0900); ethyl acetate (Honeywell, Fluka-Seelze, Germany; Lot. I3180); and sulfuric acid, $97\%$ v/v (Sigma Aldrich-St. Louis, MO, USA; Lot. JO070). ## 4.3. Sample Pretreatment The total amount of indoxyl sulfate was obtained by the centrifugation of MeCN with the serum (1:1, v/v) at 5000× g for a total of 5 min. Next, the supernatant was blended with 10mM of ammonium acetate buffer (1:1, v/v) and DHTC (2 μg/mL). In order to quantify the amount of the unbound form of indoxyl sulfate, the serum was ultrafiltered by centrifugation at 10.000× g for 30 min using a 3K MWCO filter to separate proteins from the whole serum. After deproteinization, the ultrafiltered serum was treated with 10mM of ammonium acetate buffer (1:1) and DHTC (2 μg/mL). Finally, the total and the free fraction of indoxyl sulfate were quantified by both LC-MSMS and Theremino. ## 4.4. LC-MS/MS Conditions of Uremic Toxins and Indigo Uremic toxins and Indigo were analyzed using LCMS 8040 Triple Quadrupole LC/MS/MS (Shimadzu), as previously described [55]. Electrospray (ESI, negative mode at 4000 V) was performed for each analysis in LC-MS/MS. The mobile phase was composed of MeCN mixed with 2 mM of ammonium acetate. Nitrogen was employed for the nebulizer and desolvation gas and finally utilized as collision gas for molecule dissociation [28]. Notably, for Indigo determination, $10\%$ MeCN was used for equilibrating the column (Luna Omega, size 150 × 4.6 mm, Phenomenex). Next, MeCN was raised up to $50\%$ and finally scaled at $10\%$ of MeCN [56]. The fragmentation voltages and collision energies of analytes for the first and second methods were optimized, as reported in Table 1. The chromatograms and MRM (multiple reaction monitoring) spectrum of IS, pCS, DHTC, and Indigo are shown in Figure 5. ## 4.5. Spectrometer Apparatus The Theremino Spectrometer provides results based on the measurement of transmitted light by Indigo at λ = 410–570 nm. The Theremino System software calculates the light intensity derived from each pixel, resulting in the measurement of the amount of light emission. For Indigo, we measured the transmitted light in the range of λ = 420–450 nm. ## 4.6. Stability The sample’s freezing stability was determined by evaluating the concentration of analytes in each sample at +4° after 24 h and at −80° after 7 days and 30 days. IS quantification with Theremino is based on spectrophotometry, thus relying on the capacity of IS to originate Indigo. Based on our analyses, we observed that the sensitivity of the Theremino was not dependent on blood rheology (e.g., cloudy serum or blood clots) as there was no interference with IS concentrations or the derivatization reaction. Moreover, we observed that sample quality did not influence IS reactions, such as Indigo generation or its absorbance at wavelengths of 420–450 nm, and evaluated the ability of Theremino to perform quantifications under different conditions, such as freeze/thaw cycles. ## 4.7. Statistical Analysis GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA) and IBM SPSS Statistics v.25 (Apple MacOS) (IBM, Armonk, NY, USA) were used to perform correlations, regression analysis, and to obtain a Bland–Altman plot for method comparison. The statistical significance of each analysis was considered at a p-value < 0.05. ## 5. Conclusions IS represents a hallmark in CKD progression and CVD-related risks in HD patients. Our innovative approach can provide rapid information regarding the serum concentration of IS in CKD and HD populations. The routine measurement of this toxin using our novel tool in clinical practice can be useful to better manage these patients and evaluate the effectiveness of different dialysis treatments. In the past years, a large number of strategies aimed at lowering the level of serum IS, such as diet modification and the administration of charcoal and biotic supplements, have been suggested. Hence, the detection of this toxin may offer a screening method for the CKD population at high risk of CV events. Moreover, the short-time detection of gut-derived indoxyl sulfate would help clinicians assess the effects of several treatments aimed at lowering uremic toxins and predict the outcome of renal dysfunction. ## References 1. Wu I.-W., Hsu K.-H., Lee C.-C., Sun C.-Y., Hsu H.-J., Tsai C.-J., Tzen C.-Y., Wang Y.-C., Lin C.-Y., Wu M.-S.. P-Cresyl Sulphate and Indoxyl Sulphate Predict Progression of Chronic Kidney Disease. Nephrol. Dial. Transplant. (2011) 26 938-947. DOI: 10.1093/ndt/gfq580 2. Lim Y.J., Sidor N.A., Tonial N.C., Che A., Urquhart B.L.. Uremic Toxins in the Progression of Chronic Kidney Disease and Cardiovascular Disease: Mechanisms and Therapeutic Targets. Toxins (2021) 13. DOI: 10.3390/toxins13020142 3. Rosner M.H., Reis T., Husain-Syed F., Vanholder R., Hutchison C., Stenvinkel P., Blankestijn P.J., Cozzolino M., Juillard L., Kashani K.. Classification of Uremic Toxins and Their Role in Kidney Failure. Clin. J. Am. Soc. Nephrol. (2021) 16 1918-1928. DOI: 10.2215/CJN.02660221 4. Yamamoto S., Kazama J.J., Wakamatsu T., Takahashi Y., Kaneko Y., Goto S., Narita I.. Removal of Uremic Toxins by Renal Replacement Therapies: A Review of Current Progress and Future Perspectives. Ren. Replace. Ther. (2016) 2 43. DOI: 10.1186/s41100-016-0056-9 5. Caggiano G., Cosola C., di Leo V., Gesualdo M., Gesualdo L.. Microbiome Modulation to Correct Uremic Toxins and to Preserve Kidney Functions. Curr. Opin. Nephrol. Hypertens. (2020) 29 49-56. DOI: 10.1097/MNH.0000000000000565 6. Wong J., Piceno Y.M., DeSantis T.Z., Pahl M., Andersen G.L., Vaziri N.D.. Expansion of Urease- and Uricase-Containing, Indole- and p-Cresol-Forming and Contraction of Short-Chain Fatty Acid-Producing Intestinal Microbiota in ESRD. Am. J. Nephrol. (2014) 39 230-237. DOI: 10.1159/000360010 7. Vaziri N.D., Yuan J., Norris K.. Role of Urea in Intestinal Barrier Dysfunction and Disruption of Epithelial Tight Junction in Chronic Kidney Disease. Am. J. Nephrol. (2013) 37 1-6. DOI: 10.1159/000345969 8. Lisowska-Myjak B.. Uremic Toxins and Their Effects on Multiple Organ Systems. Nephron Clin. Pract. (2014) 128 303-311. DOI: 10.1159/000369817 9. Dou L., Jourde-Chiche N., Faure V., Cerini C., Berland Y., Dignat-George F., Brunet P.. The Uremic Solute Indoxyl Sulfate Induces Oxidative Stress in Endothelial Cells. J. Thromb. Haemost. (2007) 5 1302-1308. DOI: 10.1111/j.1538-7836.2007.02540.x 10. Rodrigues G.G.C., Dellê H., Brito R.B.O., Cardoso V.O., Fernandes K.P.S., Mesquita-Ferrari R.A., Cunha R.S., Stinghen A.E.M., Dalboni M.A., Barreto F.C.. Indoxyl Sulfate Contributes to Uremic Sarcopenia by Inducing Apoptosis in Myoblasts. Arch. Med. Res. (2020) 51 21-29. DOI: 10.1016/j.arcmed.2019.12.020 11. Hung S., Kuo K., Wu C., Tarng D.. Indoxyl Sulfate: A Novel Cardiovascular Risk Factor in Chronic Kidney Disease. J. Am. Heart Assoc. (2017) 6 e005022. DOI: 10.1161/JAHA.116.005022 12. Yisireyili M., Shimizu H., Saito S., Enomoto A., Nishijima F., Niwa T.. Indoxyl Sulfate Promotes Cardiac Fibrosis with Enhanced Oxidative Stress in Hypertensive Rats. Life Sci. (2013) 92 1180-1185. DOI: 10.1016/j.lfs.2013.05.008 13. Colombo G., Astori E., Landoni L., Garavaglia M.L., Altomare A., Lionetti M.C., Gagliano N., Giustarini D., Rossi R., Milzani A.. Effects of the Uremic Toxin Indoxyl Sulphate on Human Microvascular Endothelial Cells. J. Appl. Toxicol. (2022) 42 1948-1961. DOI: 10.1002/jat.4366 14. Sun C.-Y., Hsu H.-H., Wu M.-S.. P-Cresol Sulfate and Indoxyl Sulfate Induce Similar Cellular Inflammatory Gene Expressions in Cultured Proximal Renal Tubular Cells. Nephrol. Dial. Transplant. (2013) 28 70-78. DOI: 10.1093/ndt/gfs133 15. Edamatsu T., Fujieda A., Itoh Y.. Phenyl Sulfate, Indoxyl Sulfate and p-Cresyl Sulfate Decrease Glutathione Level to Render Cells Vulnerable to Oxidative Stress in Renal Tubular Cells. PLoS ONE (2018) 13. DOI: 10.1371/journal.pone.0193342 16. Shimizu H., Yisireyili M., Higashiyama Y., Nishijima F., Niwa T.. Indoxyl Sulfate Upregulates Renal Expression of ICAM-1 via Production of ROS and Activation of NF-ΚB and P53 in Proximal Tubular Cells. Life Sci. (2013) 92 143-148. DOI: 10.1016/j.lfs.2012.11.012 17. Barba C., Soulage C.O., Caggiano G., Glorieux G., Fouque D., Koppe L.. Effects of Fecal Microbiota Transplantation on Composition in Mice with CKD. Toxins (2020) 12. DOI: 10.3390/toxins12120741 18. Koppe L., Croze M.L., Monteiro E.B., Benoit B., Bres E., Guebre-Egziabher F., Daleprane J.B., Fouque D., Soulage C.O.. The Protein-Bound Uremic Toxin p-Cresyl-Sulfate Promotes Intracellular ROS Production and Lipid Peroxidation in 3T3-L1 Adipose Cells. Biochimie (2021) 189 137-143. DOI: 10.1016/j.biochi.2021.06.020 19. Koppe L., Pillon N.J., Vella R.E., Croze M.L., Pelletier C.C., Chambert S., Massy Z., Glorieux G., Vanholder R., Dugenet Y.. P-Cresyl Sulfate Promotes Insulin Resistance Associated with CKD. J. Am. Soc. Nephrol. (2013) 24 88-99. DOI: 10.1681/ASN.2012050503 20. Uchiyama K., Wakino S., Irie J., Miyamoto J., Matsui A., Tajima T., Itoh T., Oshima Y., Yoshifuji A., Kimura I.. Contribution of Uremic Dysbiosis to Insulin Resistance and Sarcopenia. Nephrol. Dial. Transplant. (2020) 35 1501-1517. DOI: 10.1093/ndt/gfaa076 21. Moradi H., Sica D.A., Kalantar-Zadeh K.. Cardiovascular Burden Associated with Uremic Toxins in Patients with Chronic Kidney Disease. Am. J. Nephrol. (2013) 38 136-148. DOI: 10.1159/000351758 22. Falconi C.A., da Cruz Junho C.V., Fogaça-Ruiz F., Vernier I.C.S., da Cunha R.S., Stinghen A.E.M., Carneiro-Ramos M.S.. Uremic Toxins: An Alarming Danger Concerning the Cardiovascular System. Front. Physiol. (2021) 12 686249. DOI: 10.3389/fphys.2021.686249 23. Duranton F., Cohen G., de Smet R., Rodriguez M., Jankowski J., Vanholder R., Argiles A.. Normal and Pathologic Concentrations of Uremic Toxins. J. Am. Soc. Nephrol. (2012) 23 1258-1270. DOI: 10.1681/ASN.2011121175 24. Clark W.R., Dehghani N.L., Narsimhan V., Ronco C.. Uremic Toxins and Their Relation to Dialysis Efficacy. Blood Purif. (2019) 48 299-314. DOI: 10.1159/000502331 25. Atoh K., Itoh H., Haneda M.. Serum Indoxyl Sulfate Levels in Patients with Diabetic Nephropathy: Relation to Renal Function. Diabetes Res. Clin. Pract. (2009) 83 220-226. DOI: 10.1016/j.diabres.2008.09.053 26. Krieter D.H., Hackl A., Rodriguez A., Chenine L., Moragues H.L., Lemke H.-D., Wanner C., Canaud B.. Protein-Bound Uraemic Toxin Removal in Haemodialysis and Post-Dilution Haemodiafiltration. Nephrol. Dial. Transplant. (2010) 25 212-218. DOI: 10.1093/ndt/gfp437 27. Itoh Y., Ezawa A., Kikuchi K., Tsuruta Y., Niwa T.. Correlation between Serum Levels of Protein-Bound Uremic Toxins in Hemodialysis Patients Measured by LC/MS/MS. Mass Spectrom. (2013) 2 S0017. DOI: 10.5702/massspectrometry.S0017 28. Shu C., Chen X., Xia T., Zhang F., Gao S., Chen W.. LC-MS/MS Method for Simultaneous Determination of Serum. Biomed. Chromatogr. (2016) 30 1782-1788. DOI: 10.1002/bmc.3753 29. Wang Z., Jiang H., Chen X., Song X., Xu F., Chen F., Mao Z., Gao S., Chen W.. A Rapid and Sensitive Method for Simultaneous Determination of Eight Protein-Bound Uremic Toxins in Human Serum by UHPLC-MS/MS: Application in Assessing Peritoneal Dialysis. J. Pharm. Biomed. Anal. (2020) 186 113312. DOI: 10.1016/j.jpba.2020.113312 30. André C., Bennis Y., Titeca-Beauport D., Caillard P., Cluet Y., Kamel S., Choukroun G., Maizel J., Liabeuf S., Bodeau S.. Two Rapid, Accurate Liquid Chromatography Tandem Mass Spectrometry Methods for the Quantification of Seven Uremic Toxins: An Application for Describing Their Accumulation Kinetic Profile in a Context of Acute Kidney Injury. J. Chromatogr. B (2020) 1152 122234. DOI: 10.1016/j.jchromb.2020.122234 31. Oda A., Suzuki Y., Sato B., Sato H., Tanaka R., Ono H., Ando T., Shin T., Mimata H., Itoh H.. Highly Sensitive Simultaneous Quantification of Indoxyl Sulfate and 3-carboxy-4-methyl-5-propyl-2-furanpropanoic Acid in Human Plasma Using Ultra-high-performance Liquid Chromatography Coupled with Tandem Mass Spectrometry. J. Sep. Sci. (2022) 45 1672-1682. DOI: 10.1002/jssc.202100950 32. Ahmed S., Sparidans R.W., Lu J., Mihaila S.M., Gerritsen K.G.F., Masereeuw R.. A Robust, Accurate, Sensitive LC–MS/MS Method to Measure Indoxyl Sulfate, Validated for Plasma and Kidney Cells. Biomed. Chromatogr. (2022) 36 e5307. DOI: 10.1002/bmc.5307 33. Al Za’abi M., Ali B., al Toubi M.. HPLC-Fluorescence Method for Measurement of the Uremic Toxin Indoxyl Sulfate in Plasma. J. Chromatogr. Sci. (2013) 51 40-43. DOI: 10.1093/chromsci/bms103 34. Rysz J., Franczyk B., Ławiński J., Olszewski R., Ciałkowska-Rysz A., Gluba-Brzózka A.. The Impact of CKD on Uremic Toxins and Gut Microbiota. Toxins (2021) 13. DOI: 10.3390/toxins13040252 35. Liu W.-C., Tomino Y., Lu K.-C.. Impacts of Indoxyl Sulfate and P-Cresol Sulfate on Chronic Kidney Disease and Mitigating Effects of AST-120. Toxins (2018) 10. DOI: 10.3390/toxins10090367 36. Corradi V., Martino F., Godi L., Mattarollo C., Barzon E., de Cal M., Caprara C., Giavarina D., Ronco C.. P0727P-Cresyl Sulfate and Indoxyl Sulfate: A Possible Role as Biomarkers in the Prediction of Renal Function and CKD Progression. Nephrol. Dial. Transplant. (2020) 35 P0727. DOI: 10.1093/ndt/gfaa142.P0727 37. Lin T.-J., Hsu B.-G., Wang J.-H., Lai Y.-H., Dongoran R.A., Liu C.-H.. Serum Indoxyl Sulfate as a Potential Biomarker of Aortic Arterial Stiffness in Coronary Artery Disease. Nutr. Metab. Cardiovasc. Dis. (2020) 30 2320-2327. DOI: 10.1016/j.numecd.2020.07.035 38. Tan X., Cao X., Zou J., Shen B., Zhang X., Liu Z., Lv W., Teng J., Ding X.. Indoxyl Sulfate, a Valuable Biomarker in Chronic Kidney Disease and Dialysis. Hemodial. Int. (2017) 21 161-167. DOI: 10.1111/hdi.12483 39. Li Q., Zhang S., Wu Q.-J., Xiao J., Wang Z.-H., Mu X.-W., Zhang Y., Wang X.-N., You L.-L., Wang S.-N.. Serum Total Indoxyl Sulfate Levels and All-Cause and Cardiovascular Mortality in Maintenance Hemodialysis Patients: A Prospective Cohort Study. BMC Nephrol. (2022) 23. DOI: 10.1186/s12882-022-02862-z 40. Chen C.N., Chou C.C., Tsai P.S.J., Lee Y.J.. Plasma Indoxyl Sulfate Concentration Predicts Progression of Chronic Kidney Disease in Dogs and Cats. Vet. J. (2018) 232 33-39. DOI: 10.1016/j.tvjl.2017.12.011 41. Opdebeeck B., Maudsley S., Azmi A., de Maré A., de Leger W., Meijers B., Verhulst A., Evenepoel P., D’Haese P.C., Neven E.. Indoxyl Sulfate and P-Cresyl Sulfate Promote Vascular Calcification and Associate with Glucose Intolerance. J. Am. Soc. Nephrol. (2019) 30 751-766. DOI: 10.1681/ASN.2018060609 42. Barreto F.C., Barreto D.V., Liabeuf S., Meert N., Glorieux G., Temmar M., Choukroun G., Vanholder R., Massy Z.A.. Serum Indoxyl Sulfate Is Associated with Vascular Disease and Mortality in Chronic Kidney Disease Patients. Clin. J. Am. Soc. Nephrol. (2009) 4 1551-1558. DOI: 10.2215/CJN.03980609 43. Mozar A., Louvet L., Godin C., Mentaverri R., Brazier M., Kamel S., Massy Z.A.. Indoxyl Sulphate Inhibits Osteoclast Differentiation and Function. Nephrol. Dial. Transplant. (2012) 27 2176-2181. DOI: 10.1093/ndt/gfr647 44. Fan P.-C., Chang J.C.-H., Lin C.-N., Lee C.-C., Chen Y.-T., Chu P.-H., Kou G., Lu Y.-A., Yang C.-W., Chen Y.-C.. Serum Indoxyl Sulfate Predicts Adverse Cardiovascular Events in Patients with Chronic Kidney Disease. J. Formos. Med. Assoc. (2019) 118 1099-1106. DOI: 10.1016/j.jfma.2019.03.005 45. Imazu M., Fukuda H., Kanzaki H., Amaki M., Hasegawa T., Takahama H., Hitsumoto T., Tsukamoto O., Morita T., Ito S.. Plasma Indoxyl Sulfate Levels Predict Cardiovascular Events in Patients with Mild Chronic Heart Failure. Sci. Rep. (2020) 10 16528. DOI: 10.1038/s41598-020-73633-9 46. Tao X., Thijssen S., Levin N., Kotanko P., Handelman G.. Enhanced Indoxyl Sulfate Dialyzer Clearance with the Use of Binding Competitors. Blood Purif. (2015) 39 323-330. DOI: 10.1159/000381008 47. Niwa T., Aoyama I., Takayama F., Tsukushi S., Miyazaki T., Owada A., Shiigai T.. Uremic Toxicity: Urinary Indoxyl Sulfate Is a Clinical Factor That Affects the Progression of Renal Failure. Miner. Electrolyte Metab. (1999) 25 118-122. DOI: 10.1159/000057433 48. Dealler S.F., Hawkey P.M., Millar M.R.. Enzymatic Degradation of Urinary Indoxyl Sulfate by Providencia Stuartii and Klebsiella Pneumoniae Causes the Purple Urine Bag Syndrome. J. Clin. Microbiol. (1988) 26 2152-2156. DOI: 10.1128/jcm.26.10.2152-2156.1988 49. Niwa T.. Update of Uremic Toxin Research by Mass Spectrometry. Mass Spectrom. Rev. (2011) 30 510-521. DOI: 10.1002/mas.20323 50. Niwa T.. Recent Progress in the Analysis of Uremic Toxins by Mass Spectrometry. J. Chromatogr. B (2009) 877 2600-2606. DOI: 10.1016/j.jchromb.2008.11.032 51. Vanholder R.C., de Smet R.V., Ringoir S.M.. Assessment of Urea and Other Uremic Markers for Quantification of Dialysis Efficacy. Clin. Chem. (1992) 38 1429-1436. DOI: 10.1093/clinchem/38.8.1429 52. Basile C., Libutti P., di Turo A.L., Casino F.G., Vernaglione L., Tundo S., Maselli P., de Nicolo E.V., Ceci E., Teutonico A.. Removal of Uraemic Retention Solutes in Standard Bicarbonate Haemodialysis and Long-Hour Slow-Flow Bicarbonate Haemodialysis. Nephrol. Dial. Transplant. (2011) 26 1296-1303. DOI: 10.1093/ndt/gfq543 53. Schulman G., Vanholder R., Niwa T.. AST-120 for the Management of Progression of Chronic Kidney Disease. Int. J. Nephrol. Renov. Dis. (2014) 2014 49-56. DOI: 10.2147/IJNRD.S41339 54. Stasi A., Cosola C., Caggiano G., Cimmarusti M.T., Palieri R., Acquaviva P.M., Rana G., Gesualdo L.. Obesity-Related Chronic Kidney Disease: Principal Mechanisms and New Approaches in Nutritional Management. Front. Nutr. (2022) 9 925619. DOI: 10.3389/fnut.2022.925619 55. Calaf R., Cerini C., Génovésio C., Verhaeghe P., Jourde-Chiche N., Bergé-Lefranc D., Gondouin B., Dou L., Morange S., Argilés A.. Determination of Uremic Solutes in Biological Fluids of Chronic Kidney Disease Patients by HPLC Assay. J. Chromatogr. B (2011) 879 2281-2286. DOI: 10.1016/j.jchromb.2011.06.014 56. Dei Cas M., Casagni E., Gambaro V., Cesari E., Roda G.. Determination of Daptomycin in Human Plasma and Breast Milk by UPLC/MS-MS. J. Chromatogr. B (2019) 1116 38-43. DOI: 10.1016/j.jchromb.2019.03.036
--- title: 'Physical Activity, Screen Time, and Academic Burden: A Cross-Sectional Analysis of Health among Chinese Adolescents' authors: - Yiting E - Jianke Yang - Yifei Shen - Xiaojuan Quan journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049325 doi: 10.3390/ijerph20064917 license: CC BY 4.0 --- # Physical Activity, Screen Time, and Academic Burden: A Cross-Sectional Analysis of Health among Chinese Adolescents ## Abstract This paper aims to analyze the effects of physical activity, screen time, and academic burden on adolescent health in China and compare their effects by using the nationally representative sample data from the CEPS (China Educational Panel Survey) cross-section data. This paper first uses regression analysis to examine the relationship between physical activity, screen time, academic burden and health among Chinese adolescents. Then, this paper uses the clustering analysis the influence of physical activity, screen time, and academic burden on the health of Chinese adolescents. The empirical results show that: [1] along with exercise, helping with the housework also has a clear health-promoting effect on adolescents; [2] the time spent surfing the Internet or playing video games, and heavy studying or homework off campus have a negative effect on adolescents’ self-rated health and mental health; [3] physical activity has the greatest impact on self-rated health, while screen time has the greatest impact on mental health, and academic burden is not the most important factor affecting adolescent health in China. ## 1. Introduction The decline in adolescent health is a challenge common to many countries. According to the data released by UNICEF in 2020, more than 13 percent of adolescents aged 10–19 worldwide will have a mental illness [1]. Therefore, how to promote adolescent health has become a major concern for policy makers, educators, and parents in many countries. Scholars have actively explored the decline in adolescent health [2,3], and reached a basic consensus that socio-environmental factors, rather than genetic (biological) factors, are at play. For example, some scholars discovered that socio-environmental factors are particularly important in a period of dramatic lifestyle changes and increasing risks to adolescent health levels [4]. Although a number of studies have examined the effects of physical activity, screen time, and academic burden on adolescent health [5,6,7], few studies have been conducted on Chinese samples. The present study aims to fill this research gap by simultaneously considering the relationship of physical activity, screen time, and academic burden on health, and further examining different influences of the above three on health among Chinese adolescents. ## 2.1.1. Physical Activity Physical activity is one important health behavior and is defined as any bodily movement that is coordinated by skeletal muscle, resulting in increased energy expenditure above that of basal metabolism [8,9]. There is a large amount of research focusing on the linkage between physical activity and health. The Harvard Alumni Health Study began to focus on the social group differences in physical activity and the internal relationship between physical activity time, intensity, and health after discovering the effect of physical activity on life expectancy and the risk of chronic disease [10]. These studies greatly changed the public perception of physical activity by revealing that physical activities could play an important role in disease prevention and health promotion [11]. However, a global review showed that $81\%$ of adults do not meet the WHO recommendation of at least 60 min of moderate to vigorous physical activity (MVPA) per day [12]. This may be explained by the fact that low and middle-income countries are now experiencing increased rates of inadequate physical activity levels. Physical activity of adolescents is a fundamental determinant of a positive outlook on life [13], and engaging in more organized sports activities can foster lifelong attitudes and behaviors [14]. However, the lack of physical activity has not been effectively addressed, and the situation of young people is even more serious. According to a World Health Organization-funded study with a sample size of 1.6 million students between the ages of 11 and 17 in 146 countries/regions, $77.6\%$ of boys and $84.7\%$ of girls do not meet the recommended levels of physical activity [12]. Only $15\%$ and $18.4\%$ of children and adolescents in Europe and the United States, respectively, meet the WHO-recommended standard of “moderate to high-intensity physical activity for 60 min a day” [15,16]. The lack of physical activity levels is not only observed in the developed economies, but also in the low and- middle income countries as well. In China, only $29.9\%$ of Chinese children and adolescents met the recommended criteria [17]. As a result, the risk of developing diseases such as obesity, hypertension, diabetes, cancer, etc., is greatly increased, and their present and future health is severely compromised [11]. Insufficient physical activity has become one of the most significant public health issues of this century, if not the most significant. ## 2.1.2. Screen Time As a modern health risk factor, sedentary behavior refers to sitting, reclining, or lying positions with energy expenditure ≤ 1.5 METs in the awake state [18]. Unlike physical activity, sedentary has caught attention in recent years [19,20]. However, it has similarities to physical inactivity in terms of prevalence and harmfulness. Sedentary behavior can be caused by screen time and academic burden in adolescence, although they have different origins. The former is mainly due to the addiction to the Internet and new media technologies, while the latter is caused by competitive education. It is generally acknowledged that physical activity declines while sedentary behavior increases from childhood to adolescence [21,22,23,24]. Additionally, with the rise of the mobile internet and new media technology, tablets and mobile phones have become a particularly integral part of adolescents’ daily lives. They frequently engage in sedentary screen-based activities, rather than outdoor exercise, sports, or play. They are strongly advised to limit their time spent sitting down, especially when engaging in recreational screen time [25]. Moreover, the sedentary behavior is device-specific. One cross-nation survey found the amount of time adolescents spent on computer and smart phone use increased dramatically while the time in watching television decreased slightly [26]. More studies show that screen time can deteriorate both physical and mental health. On the one hand, exposure to screens can increase adolescents’ risk of obesity [27]. On the other hand, excessive screen time might cause mental issues. Adolescents addicted to the Internet or video games and watch too much television are more likely to develop depression and various behavioral problems [28,29,30,31]. ## 2.1.3. Academic Burden Academic burden can reduce sleep time and, therefore, has a significant negative impact on adolescents’ physical and mental health [32]. In addition, urban middle-class parents typically have high expectations for their children’s education and invest a variety of high-quality resources in studying so that their children can participate in shadow education. According to a survey of Ho Chi Minh City schools, $49\%$ of parents believe that their children’s participation in shadow education adversely impacts their physical and mental health [33]. Such extracurricular tutoring can deprive adolescents of leisure time and social contact, thus increasing their tendency to develop internalized behavioral disorders such as anxiety and depression [34,35,36]. Chinese adolescents are exposed to intense academic competition and pressure and carry heavier academic burdens compared to those in other countries and regions. They frequently experience stress and anxiety regarding homework and after-school studies [37]. Although the government has designated the policy of reducing the academic burden on primary and secondary school students as one of the main responsibilities of compulsory education, it emphasizes that the government, schools, families, and society must cooperate in implementing the policy throughout the educational process and promote the healthy growth of students. However, the situation is still unfavorable. Chinese adolescents still carry a heavy academic burden [38]. ## 2.2. Current Study While exploring the social environmental factors affecting adolescent health in China and western countries, prior work has primarily examined the linkage between physical activity and health [39,40,41,42], or the effects of academic burden on physical health and mental health [43,44,45,46], and paid little attention to consider these three factors in an analytical framework. Our study of this issue not only provides a more comprehensive understanding of how these three factors are associated with adolescent health, but also helps develop targeted interventions and recommendations to parents and schools. The main purpose of this study is to explore the relationships between physical activity, screen time, academic burden, and adolescent health in the Chinese context. Based on the previous literature review, two main research questions are proposed here. First, do physical activity, screen time and academic burden have significant effects on adolescent health? Second, which has the strongest effects on adolescent health among these three factors? ## 3.1. Data This study utilizes data from the China Educational Panel Survey (CEPS) conducted by National Survey Research Center (NSRC) at Renmin University of China. Taking the 2013–2014 data as the baseline, the CEPS applies a stratified, multistage sampling design with probability proportional to size (PPS), randomly selecting a school-based, nationally representative sample of approximately 20,000 students in 438 classrooms of 112 schools in 28 county-level units in mainland China. In the academic year 2014–2015, a follow-up visit was conducted with the eighth-grade students (seventh graders at the baseline survey) who participated in the baseline survey. The research object consists of 9449 students who were successfully followed up. By deleting the samples with missing values for the key variables, 8213 effective cases were finally obtained. ## 3.2.1. Self-Rated Health and Mental Health The CEPS questionnaire uses a five-point Likert scale to measure self-rated health. Respondents are required to answer the question “How is your overall health now”, and the corresponding options were “1. very bad, 2. not good, 3. fair, 4. better, 5. very good”. Since the number of cases in option 1 was too small, we combined it with option 2 to form an ordered variable with four categories: “1. not good, 2. average, 3. better, 4. very good”. For mental health, we used a scale from the CEPS consisting of 10 questions measuring anxiety or depression. Each question corresponds to five options “1. never, 2. rarely, 3. sometimes, 4. often, 5. always”. The respondents were assigned a value of 1 for “3. sometimes, 4. often, 5. always” and a value of 0 for “1. never, 2. seldom.” The results of the new assignments for the 10 questions were then summed to obtain a count variable with values ranging from 0 to 10. The higher the number, the worse the mental health. ## 3.2.2. Physical Activity Time Physical activity time includes daily physical exercise time and housework time. The average daily physical exercise time was calculated by replacing the value of each case that answers “more than 4 h” with 5 h. After that, it was operationalized into three categories: 1. less than or equal to quarter an hour, 2. more than quarter an hour less than or equal to half an hour, and 3. more than half an hour. For housework time, this study replaced the case values over 4 h with 5 h and over 60 min with 60 min, thus calculating the average daily housework time, and then transformed it into three categories, “1. less than or equal to half an hour, 2. more than half an hour less than or equal to 1 h, 3. more than 1 h ”. ## 3.2.3. Screen Time Screen time was measured by the time spent in watching TV and surfing internet or playing game. The questionnaire firstly asked respondents how much time they watch TV per day, and the corresponding options were “1. none, 2. less than 1 h, 3. about 1–2 h, 4. about 2–3 h, 5. about 3–4 h, and 6. more than 4 h”. Next, the respondents were asked how much time they spent watching TV on weekends, and the corresponding options were “1. none, 2. less than 2 h, 3. about 2–4 h, 4. about 4–6 h, 5. about 6–8 h, and 6. more than 8 h”. We recoded values to each option with “1. 0 h, 2. 0.5 h, 3. 1.5 h, 4. 2.5 h, 5. 3.5 h, and 6. 4.5 h” and calculated the average daily TV watching time and converted it into three categories: 1. Less than or equal to 1 h, 2. More than 1 h less than or equal to 2 h, and 3. More than 2 h. Students were asked about the amount of time they spent online and playing games each day, and the corresponding options were set the same as those for “time spent watching TV each day. Following the above calculating way. ## 3.2.4. Academic Burden Time The homework time on campus and off campus per day was measured in the same way as the screen time, divided into “1. less than or equal to 2 h, 2. more than 2 h less than or equal to 4 h, 3. more than 4 h”. ## 3.2.5. Control Variable Social-economic characteristics include gender (female = 0; male = 1), household registration (agricultural household registration = 0, non-agricultural household registration = 1), years of parental education, family economic conditions (poor, medium, rich), whether they live with both parents (yes = 1, no = 0), only child status (yes = 1, no = 0), sleep time (1. less than and equal to 6 h; 2. more than 6 h and less than 8 h; 3. more than 8 h), and school residence status (yes = 1, no = 0). There were three survey questions applied to measure the extent of introversion. These three questions have four responses: 1. completely disagree, 2. not agree, 3. relatively agree, and 4. completely agree. By summing up option values for each question, a continuous variable with a range of 2 to 12 can be obtained. The greater the value, the more introverted the personality. ## 3.3. Data Analysis Statistical analysis was performed using Stata 17.0. In accordance with the purpose of the study, three statistical methods are used in this paper. The first method is correlation analysis, which examines correlated relationship between physical activity, screen time, and academic burden. The second is a regression analysis in which we examine the effects of physical activity, screen time, and academic burden on adolescent health in the same framework. There are two dependent variables in this study: self-rated health and mental health. For self-rated health, it is an ordered variable, so we use the logistic regression model; for mental health, it is a counting variable, and its variance is greater than the mean value, which is an over-discrete case. Therefore, we use the negative binominal regression model in our study. The third analysis is the coefficient clustering analysis, which aims to compare the effects of physical activity and sedentary behavior on adolescent health, as well as the effect intensities of various types of physical activity and sedentary behavior. The clustering coefficients can classify the variables in the regression into distinct groups and assume that distinct groups of variables influence the dependent variable through a latent variable. ## 4.1. Descriptive Statistic and Correlations The results of the descriptive statistical analysis of relevant variables are displayed in Table 1. Most students felt that their overall health was generally poor. Only $29.93\%$ of students believed their own health was good, $35.15\%$ felt unwell, and $34.29\%$ thought their health was generally good. The mean mental health score was 3.50, indicating that each student had an average of more than three types of anxiety or depression. The average daily exercise duration was 22 min, and nearly $50\%$ of students exercise for less than 15 min per day. The average amount of time spent on housework was 1.08 h, but only $33.08\%$ of students spent longer than 1 h. The average daily time spent watching television, browsing the Internet, and playing video games was 2.89 h, far exceeding the state-advocated 1 h limit. The average time spent on homework inside and outside of school was 2.95 h, which was about twice the national average (90 min). In the sample used for analysis, the average number of years of education for parents is 11.09, indicating a junior high school education. $68.15\%$ of the sampled population lived with both parents, while $44.69\%$ of the students were only children. The proportions of the poor, the medium, and the wealthy families were $28.43\%$, $67.16\%$, and $4.41\%$, respectively. The percentages of students who slept less than or equal to 6 h, more than 6 h and less than or equal to 8 h, and more than 8 h were $5.83\%$, $51.13\%$, and $43.04\%$, respectively, and $30.37\%$ of students lived in school from Monday to Thursday. The results of the correlation analysis between physical activity, screen time, and aca demic burden are displayed in Table 2. The students spent more time on TV are doing more housework ($r = 0.122$). The time spent in using the Internet and playing video games was positively correlated with the time spent in exercising ($r = 0.071$) and TV watching ($r = 0.395$). The students who spent more time in homework on campus are more likely to engage in exercise ($r = 0.064$) and housework ($r = 0.022$). The correlation coefficient between homework time off campus is positively correlated with exercise time ($r = 0.087$) and study time in school ($r = 0.907$). As expected, students with more time on homework spent less time on TV watching (r = −0.021). ## 4.2. Regression Analysis The results of the regression analysis of physical activity, screen time, academic burden, and adolescent health are presented in Table 3. Overall health is the dependent variable in models [1] and [2], whereas mental health is the dependent variable in models [3] and [4]. Models [1] and [3] only include control variables, while models [2] and [4] also consist of variables for physical activity, screen time, and academic burden. Household registration status or the number of siblings had no effects on self-rated health and mental health. The degree of students’ introverted personality has significantly lowered their self-rated health and increased their mental problems. Male students with higher family economic condition and longer sleep duration and those who lived with their parents showed a higher self-rated health score and were less likely to experienced mental health issues. Individual mental health has been significantly reduced by parents with a higher level of education, which has no bearing on the self-rated health status. Students who lived in the dorm had experienced more mental problems than those living off campus. Model [2] indicates that students’ self-rated health will increase significantly when they engage in physical activities, no matter whether they are having exercising or doing housework. The cumulative probability of being classified in a certain health category is $13.6\%$ and $45.8\%$ lower for those with exercise time of 0.25–0.5 h; and those with exercise time of more than 0.5 h than those with exercise time of less than or equal to 0.25 h. In addition, the cumulative probability of being in a particular health category was $8.5\%$ and $13.6\%$ lower for those who worked for 0.5–1 h and more than 1 h than for those who worked for less than 0.5 h. As for screen time, watching TV seems to be beneficial to their self-rated health, while browsing the Internet, or playing video games seems to be opposite. Regarding the academic burden, students’ overall self-rated health is not significantly affected by the amount of time spent on homework in school. However, the cumulative probability of being in a particular health category was $25\%$ higher for those who spent more than 4 h on schoolwork and homework than for those who spent less than or equal to 2 h. Model [4] shows that those with exercise periods of 0.25–0.5 h and more than 0.5 h had lower mental problems by $4.6\%$ and $8.5\%$ compared to those with exercise duration of less than 0.25 h. However, the duration of housework does not impact mental health. In addition, those who watched more than 2 h of television per day had $7.1\%$ more psychological problems than those who watched 1 h or less. Those who spent more time surfing the Internet and playing games had greater mental problems than those who spent less than 1 h. There was no significant effect on the mental health of the students who engaged in school study time every day. However, the mental health of the students who engaged in homework off campus for more than 4 h was significantly worse, which was $30.3\%$ higher than that of the students who spent on homework off campus for less than 2 h. ## 4.3. Coefficient Clustering Analysis The results of clustering analysis for the coefficients of physical activity, screen time, academic burden, and adolescent health are presented in Table 4. The comparison of self-rated health and mental health is calculated using models [2] and [4] from Table 4. As mentioned in the prior article, coefficient clustering analysis allows for the division of the variables in the regression into various groups and makes the assumption that these various groups of variables have an effect on the dependent variable through a latent variable. Here, we divide the independent variables into three groups: physical activity, screen time, and academic burden, and assume that these three groups have a coefficient-based effect on the dependent variable. After clustering, the coefficient (i.e., the total effect of a set of variables) is always a nonnegative number that represents only the magnitude of the influence and is independent of its direction. It is clear that physical activity, screen time, and academic burden have a significant impact on adolescents’ physical and mental health, but to varying degrees. Physical activity has a greater impact on self-rated health than screen time (χ2 = 22.84, $$p \leq 0.000$$) as well as academic burden (χ2 = 35.25, $$p \leq 0.000.$$ However, screen time has a greater impact on mental health than physical activity and academic burden (χ2 = 16.69, $$p \leq 0.000$$). There is no difference between the impact intensity of physical activity and academic burden (χ2 = 1.23, $$p \leq 0.268$$). ## 5. Discussion This study included physical activity, screen time, and academic burden in the same framework for comparative analysis and explored the independent effects of each factor on health. Previous research has provided a wealth of evidence for improving adolescent physical and mental health, suggesting that increased physical activity and reduced screen time contribute to adolescent health [47,48]. However, existing literature provides very limited insights into the academic burden, which is specific to China’s socio-cultural traditions and class mobility status. ## 5.1. Main Findings The first finding of this study is that the more the time spent exercising, the more the time spent on homework on campus and off campus. A strand of prior research has confirmed that academic burden was cited as the primary reason for not having sufficient physical activity during transition grades and high school years [49]; one possible explanation is that limited time necessitates the substitution of one behavior for another [50]. Therefore, homework or academic burden is the primary barrier for adolescents to engage in physical activity [51,52,53]. Regarding the social causes of the above results, this is attributed to differences in family social capital and parenting styles. The greater a family’s access to social resources, the greater the number of tasks children must master in and out of the classroom and the lengthier the learning period. For example, an empirical study in New Zealand found that while screen time type does appear to be implicated in academic achievement, the mechanism appears to be specific to higher socioeconomic status families [54]. Furthermore, the study showed that the time spent on the Internet and playing video games was positively correlated with the time spent in physical activity, and moderate screen time is good for physical and mental health [55]. Previous studies have revealed that the greater screen time will impairs cognition [56]. The research conducted by Zhang et al., supports our findings. He believes that screen-based activities can be properly implemented to meet cognitive development requirements. For young children, it appears essential to encourage physical activity and participation in organized and unorganized activities, as well as to adhere to screen time recommendations for cognitive development [42]. As the research results above, watching TV seems to be beneficial to adolescents’ self-rated health, while browsing the Internet, or playing video games seems to be opposite. One possible explanation is that Chinese students are believed to have high academic burden and pressure due to high expectations of their parents and fierce competitions with their peers [57], so they do not have much time to watch TV. Meanwhile, from a global perspective, adolescents’ internet addiction has become a worldwide problem, and it has high risks to physical and mental health [58,59,60]. However, with the extensive development of the Internet, the pooled Internet addiction detection rate of Chinese college students was $11\%$, which is higher than in some other countries [61]. Second, our study found that physical activity consisting of exercise and housework time has significant effects on adolescents’ self-rated health, while exercise time instead of housework time has negatively effects on mental health. In other words, the more exercise, the less likelihood of a mental problem. In terms of screen time, TV watching seemed to increase their self-rated health ranking. Compared to those watching less one hour per day, those watching TV more than two hours are more likely to have mental problem. As for those students spending more time in online surfing and game playing, they are more likely to low their self-rated health ranking and have more mental problems. During COVID-19, many people live a sedentary lifestyle, including lack of physical activity and long screen time, and have a poor psychological status, which may lead to considerable health risks [62,63]. We also found that time spent on schoolwork had no significant effects on self-rated health and mental health. However, compared to those students spending on study and homework off campus less 2 h, those who spent more than 4 h reduced their self-rate health ranking and increased their mental health problems. This result has important policy implications. It suggests that the school and family should work together and reduce the over burden of their students’ study on campus and homework off campus. In addition, schools should emphasize comprehensive wellness strategies to address multiple behaviors to maximize student health and academic success [64]. The final and most important finding is that the academic burden is not the most significant risk factor for Chinese adolescents’ health. In other words, physical activity has a greater effect on self-rated health than screen time and academic burden, while screen time has a greater impact on mental health than both physical activity and academic burden. One possible explanation could be that physical activity may improve functional status, and good function status has related to self-rated health status [65,66]. As for screen time, on the one hand, people who spend more time in front of the screen have more sleep problems, which may affect their mental state and lead to increased depression and anxiety [67,68]. On the other hand, adolescents’ internet addiction will cause loneliness and low self-esteem, which are harmful to mental health [69,70,71]. This indicates that a separate physical activity promotion policy does not necessarily lead to a reduction in screen time in practice. Nonetheless, school intervention is still required to increase students’ physical activity levels [72,73,74]. This finding provides new empirical evidence for managing screen time with an emphasis on increasing physical activity and reducing academic burden, as physical activity, screen time, and academic burden have different health effects. Furthermore, digital media have become an important way for individuals to access information, communication, and entertainment. In addition to positive government interventions, families and schools should give high priority to adolescents’ use of electronic devices, which not only contributes to children’s educational outcomes but also to their health status. The current cross-sectional study addresses this research gap by examining the complex associations among physical activity, screen time, and academic burden and health. With the rapid growth in economy, smartphones and tablets are gaining widespread use among children and youth and may already be key contributors to screen time. In addition to positive government interventions, families and schools should give high priority to adolescents’ use of electronic devices, and well-designed trials that are currently accruing participants are examining innovative and more comprehensive strategies for reducing screen time [75]. In Finland, a specific content on physical activity and screen time based on Health Action Process Approach model was integrated into routinely scheduled three health education lessons with the help of educational material in secondary schools [76]. ## 5.2. Study Limitations There are, however, several limitations in this paper. First, the depth and accuracy of the data analysis are limited since the data used were not collected through a special survey, which makes it difficult to measure all relevant variables. An example of this would be the lack of information regarding the type, intensity, and content of physical activity when playing games and surfing the internet. Second, despite the facts that the analysis samples are nationally representative, they are all eighth-graders, or junior high school students. The research on the health and related behaviors of adolescents of other ages requires additional investigation and analysis. Third, this study is the cross-sectional character of our dataset. It is a consensus that results derived from a cross-sectional study cannot address the issue of causality, thus longitudinal studies are expected in order to explore the causal relationships. ## 6. Conclusions On the basis of national survey data, this study examines the relationship among physical activity, screen time, and academic burden, and compares their effects on the physical and mental health of adolescents. The key findings are as follows: first, in addition to physical exercise, helping with housework also has a clear health-promoting effect on adolescents. Second, screen time and academic pressure will generally have a detrimental effect on adolescents’ health, particularly the time spent surfing the Internet, playing video games, excessive studying outside of school. Third, screen time has the greatest impact on mental health, while physical activity has the greatest impact on self-rated health. It appears that academic burden is not the greatest social risk factor for adolescents’ health in China. In addition, the current research results here are worthy of further study in the future to explore the influencing factors and mechanisms of adolescent health in China, in order to find a balance between all parties and provide more solid and reliable scientific basis for adolescent health promotion. This implies that policy makers should make different policy key points in term of how to improve adolescents’ self-rated health and mental health. ## References 1. Keeley B.. The State of the World’s Children 2021: On My Mind—Promoting, Protecting and Caring for Children’s Mental Health (2021) 2. Coady S.A., Jaquish C.E., Fabsitz R.R., Larson M.G., Cupples L.A., Myers R.H.. Genetic variability of adult body mass index: A longitudinal assessment in Framingham families. Obes. Res. (2002) 10 675-681. DOI: 10.1038/oby.2002.91 3. Dolgin E.. The myopia boom. Nature (2015) 519 276. DOI: 10.1038/519276a 4. Zhou M., Ding X.. Internet use, depression, and cognitive outcomes among Chinese adolescents. J. Community Psychol. (2021) 51 768-787. DOI: 10.1002/jcop.22779 5. Joseph D.M., Alivia R.N., Guy F., Diane G.J., Thomas L.S., Allison P.K.. The association of physical activity, sleep, and screen time with mental health in Canadian adolescents during the COVID-19 pandemic: A longitudinal isotemporal substitution analysis. Ment. Health Phys. Act. (2022) 23 100473. PMID: 36156917 6. Tomoka K., Takeo F., Satomi D., Yui Y.. Association between Hope for the Future and Academic Performance in Adolescents: Results from the K-CHILD Study. Int. J. Environ. Res. Public Health (2022) 19 11890. PMID: 36231192 7. Waldman D.Z., Kristie R.. Physical Activity, Sports Participation, and Psychosocial Health in Adolescents with Hearing Loss. J. Adolesc. Health (2022) 71 635-641. PMID: 35718651 8. Qin W., Guo Y.. Physical activity among older African Americans: The role of diabetes diagnosis and mastery. J. Ethn. Cult. Divers. Soc. Work (2023) 32 46-56. DOI: 10.1080/15313204.2020.1855495 9. Shakoor H., Platat C., Ali H.I., Ismail L.C., Al Dhaheri A.S., Bosevski M., Apostolopoulos V., Stojanovska L.. The benefits of physical activity in middle-aged individuals for cardiovascular disease outcomes. Maturitas (2023) 168 49-52. DOI: 10.1016/j.maturitas.2022.11.002 10. Raven P.B., Wasserman D.H., Squires W.G., Murray T.D.. Exercise Physiology (2012) 11. Blair S.N.. Physical inactivity: The biggest public health problem of the 21st century. Br. J. Sport. Med. (2009) 43 1-2 12. Guthold R., Stevens G.A., Riley L.M., Bull F.C.. Global trends in insufficient physical activity among adolescents: A pooled analysis of 298 population-based surveys with 1·6 million participants. Lancet Child Adolesc. Health (2020) 4 23-35. DOI: 10.1016/S2352-4642(19)30323-2 13. Malinowska-Cieślik M., Mazur J., Nałęcz H., Małkowska-Szkutnik A.. Social and behavioral predictors of adolescents’ positive attitude towards life and self. Int. J. Environ. Res. Public Health (2019) 16. DOI: 10.3390/ijerph16224404 14. Emm-Collison L.G., Lewis S., Reid T., Matthews J., Sebire S.J., Thompson J.L., Salway R., Jago R.. Striking a balance: Physical activity, screen-viewing and homework during the transition to secondary school. Int. J. Environ. Res. Public Health (2019) 16. DOI: 10.3390/ijerph16173174 15. Currie C., Zanotti C., Morgan A., Currie D., de Looze M., Roberts C., Samdal O., Smith O.R., Barnekow V.. Social Determinants of Health and Well-Being among Young People: Health Behaviour in School-Aged Children (HBSC) Study: International Report from the 2009/2010 Survey (2012) 16. Kann L., McManus T., Harris W.A., Shanklin S.L., Flint K.H., Queen B., Lowry R., Chyen D., Whittle L., Thornton J.. Youth risk behavior surveillance—United States, 2017. MMWR Surveill. Summ. (2018) 67 1-114. DOI: 10.15585/mmwr.ss6708a1 17. Fan X., Cao Z.-B.. Physical activity among Chinese school-aged children: National prevalence estimates from the 2016 Physical Activity and Fitness in China—The Youth Study. J. Sport Health Sci. (2017) 6 388-394. DOI: 10.1016/j.jshs.2017.09.006 18. Tremblay M.S., Aubert S., Barnes J.D., Saunders T.J., Carson V., Latimer-Cheung A.E., Chastin S.F., Altenburg T.M., Chinapaw M.J.. Sedentary behavior research network (SBRN)–terminology consensus project process and outcome. Int. J. Behav. Nutr. Phys. Act. (2017) 14 75. DOI: 10.1186/s12966-017-0525-8 19. Iannotti R.J., Wang J.. Trends in physical activity, sedentary behavior, diet, and BMI among US adolescents, 2001–2009. Pediatrics (2013) 132 606-614. DOI: 10.1542/peds.2013-1488 20. Matthews C.E., Chen K.Y., Freedson P.S., Buchowski M.S., Beech B.M., Pate R.R., Troiano R.P.. Amount of time spent in sedentary behaviors in the United States, 2003–2004. Am. J. Epidemiol. (2008) 167 875-881. DOI: 10.1093/aje/kwm390 21. Arundell L., Fletcher E., Salmon J., Veitch J., Hinkley T.. A systematic review of the prevalence of sedentary behavior during the after-school period among children aged 5–18 years. Int. J. Behav. Nutr. Phys. Act. (2016) 13 93. DOI: 10.1186/s12966-016-0419-1 22. Corder K., van Sluijs E.M., Ekelund U., Jones A.P., Griffin S.J.. Changes in children’s physical activity over 12 months: Longitudinal results from the SPEEDY study. Pediatrics (2010) 126 e926-e935. DOI: 10.1542/peds.2010-0048 23. Dumith S.C., Gigante D.P., Domingues M.R., Kohl III H.W.. Physical activity change during adolescence: A systematic review and a pooled analysis. Int. J. Epidemiol. (2011) 40 685-698. DOI: 10.1093/ije/dyq272 24. Marks J., Barnett L.M., Strugnell C., Allender S.. Changing from primary to secondary school highlights opportunities for school environment interventions aiming to increase physical activity and reduce sedentary behaviour: A longitudinal cohort study. Int. J. Behav. Nutr. Phys. Act. (2015) 12 59. DOI: 10.1186/s12966-015-0218-0 25. Bull F.C., Al-Ansari S.S., Biddle S., Borodulin K., Buman M.P., Cardon G., Carty C., Chaput J.-P., Chastin S., Chou R.. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br. J. Sport. Med. (2020) 54 1451-1462. DOI: 10.1136/bjsports-2020-102955 26. Bucksch J., Sigmundova D., Hamrik Z., Troped P.J., Melkevik O., Ahluwalia N., Borraccino A., Tynjälä J., Kalman M., Inchley J.. International trends in adolescent screen-time behaviors from 2002 to 2010. J. Adolesc. Health (2016) 58 417-425. DOI: 10.1016/j.jadohealth.2015.11.014 27. Rocka A., Jasielska F., Madras D., Krawiec P., Pac-Kożuchowska E.. The impact of digital screen time on dietary habits and physical activity in children and adolescents. Nutrients (2022) 14. DOI: 10.3390/nu14142985 28. Gentile D.A., Choo H., Liau A., Sim T., Li D., Fung D., Khoo A.. Pathological video game use among youths: A two-year longitudinal study. Pediatrics (2011) 127 e319-e329. DOI: 10.1542/peds.2010-1353 29. Hamer M., Yates T., Sherar L.B., Clemes S.A., Shankar A.. Association of after school sedentary behaviour in adolescence with mental wellbeing in adulthood. Prev. Med. (2016) 87 6-10. DOI: 10.1016/j.ypmed.2016.02.021 30. Kleppang A.L., Steigen A., Finbråten H.. The association between screen time and psychological distress in Norwegian adolescents. Eur. J. Public Health (2019) 29 ckz185-076. DOI: 10.1093/eurpub/ckz185.076 31. Robinson T.N., Banda J.A., Hale L., Lu A.S., Fleming-Milici F., Calvert S.L., Wartella E.. Screen media exposure and obesity in children and adolescents. Pediatrics (2017) 140 S97-S101. DOI: 10.1542/peds.2016-1758K 32. Lee M., Larson R.. The Korean ‘examination hell’: Long hours of studying, distress, and depression. J. Youth Adolesc. (2000) 29 249-271. DOI: 10.1023/A:1005160717081 33. Dang H.-A.. A bird’s-eye view of the private tutoring phenomenon in Vietnam. IIAS Newsl. (2011) 56 26-27 34. Kuan P.-Y.. Effects of cram schooling on academic achievement and mental health of junior high students in Taiwan. Chin. Sociol. Rev. (2018) 50 391-422. DOI: 10.1080/21620555.2018.1526069 35. Mahmud R.. Learning in the shadows: Parents’ investments, family burden, and students’ workload in Dhaka, Bangladesh. Asia Pac. Educ. Rev. (2021) 22 41-52. DOI: 10.1007/s12564-020-09655-9 36. Tan J.. Confronting the Shadow Education System: What Government Policies for What Private Tutoring? (2012) 37. Hong E., Mason E., Peng Y., Lee N.. Effects of homework motivation and worry anxiety on homework achievement in mathematics and English. Educ. Res. Eval. (2015) 21 491-514. DOI: 10.1080/13803611.2015.1131721 38. Zhu X., Haegele J.A., Tang Y., Wu X.. Physical Activity and Sedentary Behaviors of Urban Chinese Children: Grade Level Prevalence and Academic Burden Associations. BioMed Res. Int. (2017) 2017 7540147. DOI: 10.1155/2017/7540147 39. Balaji S., Karthik R., Durga R., Harinie S., Ezhilvanan M.. Intensity of physical activity among school going adolescents in Chennai, South India. Int. J. Community Med. Public Health (2018) 5 2094-2098 40. Marconnot R., Marín-Rojas A.L., Delfa-de-la-Morena J.M., Pérez-Corrales J., Gueita-Rodríguez J., Fernández-de-Las-Peñas C., Palacios-Ceña D.. Recognition of barriers to physical activity promotion in immigrant children in Spain: A qualitative case study. Int. J. Environ. Res. Public Health (2019) 16. DOI: 10.3390/ijerph16030431 41. Nguyen N.-M., Dibley M.J., Tang H.K., Alam A.. Perceptions and practices related to obesity in adolescent students and their programmatic implications: Qualitative evidence from Ho Chi Minh City, Vietnam. Matern. Child Health J. (2017) 21 2199-2208. DOI: 10.1007/s10995-017-2340-x 42. Zhang Z., Wiebe S.A., Rahman A.A., Carson V.. Longitudinal associations of subjectively-measured physical activity and screen time with cognitive development in young children. Ment. Health Phys. Act. (2022) 22 100447. DOI: 10.1016/j.mhpa.2022.100447 43. Kim K.M., Kim D., Chung U.S.. Investigation of the trend in adolescent mental health and its related social factors: A multi-year cross-sectional study for 13 years. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17155405 44. Kyeong L.W.. Self-compassion as a moderator of the relationship between academic burn-out and psychological health in Korean cyber university students. Personal. Individ. Differ. (2013) 54 899-902. DOI: 10.1016/j.paid.2013.01.001 45. Zhang D., Hong J., Chen S., Liu Y.. Associations of physical activity with academic achievement and academic burden in Chinese children and adolescents: Do gender and school grade matter?. BMC Public Health (2022) 22. DOI: 10.1186/s12889-022-13886-3 46. Zhu X., Haegele J.A., Liu H., Yu F.. Academic stress, physical activity, sleep, and mental health among Chinese adolescents. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18147257 47. Babic M., Colyvas K., Morgan P., Plotnikoff R., Lonsdale C., Lubans D.. Longitudinal associations between recreational screen-time and mental health in Australian adolescents: A cross-lagged panel analysis. J. Sci. Med. Sport (2017) 20 e4. DOI: 10.1016/j.jsams.2016.12.012 48. Hasan M.K.C., Abdullah F., Firdaus M.K.Z.H., Jamaludin F.I.C.. Does physical activity and body weight status determine musculoskeletal health among adolescents in Malaysia?. Enfermería Clínica (2021) 31 S247-S251. DOI: 10.1016/j.enfcli.2020.09.010 49. Pearson N., Braithwaite R., Biddle S.J., van Sluijs E.M., Atkin A.J.. Associations between sedentary behaviour and physical activity in children and adolescents: A meta-analysis. Obes. Rev. (2014) 15 666-675. DOI: 10.1111/obr.12188 50. Gilchrist J.D., Battista K., Patte K.A., Faulkner G., Carson V., Leatherdale S.T.. Effects of reallocating physical activity, sedentary behaviors, and sleep on mental health in adolescents. Ment. Health Phys. Act. (2021) 20 100380. DOI: 10.1016/j.mhpa.2020.100380 51. Knowles A.-M., Niven A., Fawkner S.. A qualitative examination of factors related to the decrease in physical activity behavior in adolescent girls during the transition from primary to secondary school. J. Phys. Act. Health (2011) 8 1084-1091. DOI: 10.1123/jpah.8.8.1084 52. Mulvihill C., Rivers K., Aggleton P.. Views of young people towards physical activity: Determinants and barriers to involvement. Health Educ. (2000) 100 190-199. DOI: 10.1108/09654280010343555 53. Trost S.G., Pate R.R., Saunders R., Ward D.S., Dowda M., Felton G.. A prospective study of the determinants of physical activity in rural fifth-grade children. Prev. Med. (1997) 26 257-263. DOI: 10.1006/pmed.1996.0137 54. Skvarc D.R., Penny A., Harries T., Wilson C., Joshua N., Byrne L.K.. Type of screen time and academic achievement in children from Australia and New Zealand: Interactions with socioeconomic status. J. Child. Media (2021) 15 509-525. DOI: 10.1080/17482798.2021.1878045 55. Scarabottolo C.C., Tebar W.R., Araújo Guerra P.H., Martins C.M.d.L., Ferrari G., Beretta V.S., Christofaro D.G.D.. Association between Different Domains of Sedentary Behavior and Health-Related Quality of Life in Adults: A Longitudinal Study. Int. J. Environ. Res. Public Health (2022) 19. DOI: 10.3390/ijerph192416389 56. Donald R.J., Enid S.R.M., Joan V.A., Esteban B.E.. The convergent effects of primary school physical activity, sleep, and recreational screen time on cognition and academic performance in grade 9. Front. Hum. Neurosci. (2022) 16 1017598. PMID: 36438639 57. Sun J., Dunne M.P., Hou X.-Y.. Academic stress among adolescents in China. Australas. Epidemiol. (2012) 19 9-12 58. Abebe Z.E., Tadesse T., Abebaw T.S., Mamuye A.M., Yideg Y.G., Tadele A.F., Walle A.G., Jemere A.T., Chekol A.E., Tilahun M.Z.. Internet Addiction and Its Associated Factors Among African High School and University Students: Systematic Review and Meta-Analysis. Front. Psychol. (2022) 13 1156 59. Gavurova B., Ivankova V., Rigelsky M., Mudarri T.. Internet addiction in socio-demographic, academic, and psychological profile of college students during the COVID-19 pandemic in the Czech Republic and Slovakia. Front. Public Health (2022) 10 944085. DOI: 10.3389/fpubh.2022.944085 60. Imen M., Arij N., Imen S., Ines B., Sana E.M.. Adverse childhood experiences and sleep disorders among Tunisian adolescents: The mediating role of internet addiction. Child Abus. Negl. (2023) 136 106028 61. Shao Y., Zheng T., Wang Y., Liu L., Chen Y., Yao Y.. Internet addiction detection rate among college students in the People’s Republic of China: A meta-analysis. Child Adolesc. Psychiatry Ment. Health (2018) 12 25. DOI: 10.1186/s13034-018-0231-6 62. Akulwar-Tajane I., Shah A.V., Naik P.H., Parmar K.K.. Rethinking Screen Time during COVID-19: Impact on Sleep and Academic Performance in Physiotherapy Students. Health Res. (2021) 5 29-47 63. Qin F., Song Y., Nassis G.P., Zhao L., Dong Y., Zhao C., Feng Y., Zhao J.. Physical Activity, Screen Time, and Emotional Well-Being during the 2019 Novel Coronavirus Outbreak in China. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17145170 64. Howie E.K., Joosten J., Harris C.J., Straker L.M.. Associations between meeting sleep, physical activity or screen time behaviour guidelines and academic performance in Australian school children. BMC Public Health (2020) 20. DOI: 10.1186/s12889-020-08620-w 65. Chen J.D.. Benefits of physical activity on nutrition and health status: Studies in China. Asia Pac. J. Clin. Nutr. (1995) 4 29-33. PMID: 24398241 66. Rami R.K.K., Krishna R.B., Papa R.A.. Interaction among body composition, self-rated health and functional status of the elderly in an Indian population. Asia Pac. J. Clin. Nutr. (2004) 13 78-85. PMID: 15003919 67. Cyl A., Rt B.. Sleep duration does not mediate the association between screen time and adolescent depression and anxiety: Findings from the 2018 National Survey of Children’s Health. Sleep Med. (2021) 81 227-234. PMID: 33721600 68. Panchali M., Jagmeet M.. Impact of screen time during COVID-19 on eating habits, physical activity, sleep, and depression symptoms: A cross-sectional study in Indian adolescents. PLoS ONE (2022) 17 e0264951. PMID: 35259203 69. Barton E.H., Dimitrios V., Ralph N., Akosua T., Kwame B.P., Charles O.. The Impact of Mental Health Predictors of Internet Addiction among Pre-Service Teachers in Ghana. Behav. Sci. (2022) 13 20. PMID: 36661592 70. Mohsen K., Masumeh K., Aida F.. Personality traits and college students’ internet addiction: The mediating roles of general health and self-esteem. Scand. J. Psychol. (2022) 63 689-697. PMID: 35703171 71. Mwakilama E.P., Jamu E.S., Senganimalunje L., Manda T.D.. Data from “Internet Addiction and Mental Health among College Students in Malawi”. J. Open Psychol. Data (2022) 10 19. DOI: 10.5334/jopd.72 72. Ahmed K.R., Kolbe-Alexander T., Khan A.. Effectiveness of a school-based intervention on physical activity and screen time among adolescents. J. Sci. Med. Sport (2022) 25 242-248. DOI: 10.1016/j.jsams.2021.10.007 73. Gråstén A.. School-based physical activity interventions for children and youth: Keys for success. J. Sport Health Sci. (2017) 6 290. DOI: 10.1016/j.jshs.2017.03.001 74. Watson A., Timperio A., Brown H., Best K., Hesketh K.D.. Effect of classroom-based physical activity interventions on academic and physical activity outcomes: A systematic review and meta-analysis. Int. J. Behav. Nutr. Phys. Act. (2017) 14 114. DOI: 10.1186/s12966-017-0569-9 75. Saunders T.J., Vallance J.K.. Screen Time and Health Indicators among Children and Youth: Current Evidence, Limitations and Future Directions. Appl. Health Econ. Health Policy (2017) 15 323-331. DOI: 10.1007/s40258-016-0289-3 76. Anne-Mari J., Tommi V., Olavi P., Harri S., Kari T., Henri V.-Y., Anna B., Minna A.. KIDS OUT! Protocol of a brief school-based intervention to promote physical activity and to reduce screen time in a sub-cohort of Finnish eighth graders. BMC Public Health (2015) 15. PMID: 26160058
--- title: Oxidative Potential Characterization of Different PM2.5 Sources and Components in Beijing and the Surrounding Region authors: - Wei Wen - Tongxin Hua - Lei Liu - Xiaoyu Liu - Xin Ma - Song Shen - Zifan Deng journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049326 doi: 10.3390/ijerph20065109 license: CC BY 4.0 --- # Oxidative Potential Characterization of Different PM2.5 Sources and Components in Beijing and the Surrounding Region ## Abstract With the implementation of air pollution control measures, the concentration of air pollutants in the North China Plain has exhibited a downward trend, but severe fine particulate matter (PM2.5) pollution remains. PM2.5 is harmful to human health, and the exploration of its source characteristics and potential hazards has become the key to mitigating PM2.5 pollution. In this study, PM2.5 samples were collected in Beijing and Gucheng during the summer of 2019. PM2.5 components, its oxidative potential (OP), and health risks were characterized. The average PM2.5 concentrations in Beijing and Gucheng during the sampling period were 34.0 ± 6.1 μg/m3 and 37.1 ± 6.9 μg/m3, respectively. The principal component analysis (PCA) results indicated that the main sources of PM2.5 in Beijing were vehicle exhaust and secondary components and that the main sources in Gucheng were industrial emissions, dust and biomass combustion. The OP values were 91.6 ± 42.1 and 82.2 ± 47.1 pmol/(min·m3), respectively, at these two sites. The correlation between the chemical components and the OP values varied with the PM2.5 sources at these two locations. The health risk assessment results demonstrated that Cr and As were potentially carcinogenic to all populations at both sites, and Cd posed a potential carcinogenic risk for adults in Gucheng. Regional cooperation regarding air pollution control must be strengthened to further reduce PM2.5 pollution and its adverse health effects. ## 1. Introduction Rapid urbanization has led to an increase in the urban population density, vehicle ownership, and urban construction, which in turn has exacerbated air pollution [1]. Fine particulate matter (PM2.5) pollution is one of the major air pollution problems in China [2]. PM2.5 is defined as particulate matter with an aerodynamic diameter smaller than 2.5 μm, and its components include sulfate, nitrate, ammonium, carbonaceous components, and metallic components [3,4]. Compared to the characteristics of coarse particulate matter, PM2.5 is smaller in size and easier to transport, and its harmful components are more likely to enter the human body through the respiratory tract, causing health hazards [5]. Atmospheric PM2.5 exposure has been identified as an important risk factor of the disease burden in countries worldwide [6]. This is due to the oxidative stress triggered by PM2.5-induced reactive oxygen species (ROS) [7], which threaten the body’s antioxidant system and cause inflammation [8]. This represents an important potential mechanism through which PM2.5 affects human health [7,9,10]. The ability of PM2.5 to induce ROS is referred to as the oxidative potential (OP) [11,12]. This quantity is an important indicator of the health risks of atmospheric PM2.5. Through epidemiological and toxicological studies, it has been found that the OP of PM2.5 is strongly linked to human metabolic system disorders [13], fetal growth and development [14], and cardiovascular and respiratory diseases [15], such as heart disease, asthma, and lung cancer. Therefore, PM2.5 pollution has become a serious threat to public health and an obstacle to the healthy development of the economy and society. As one of the most important cities in the North China Plain, Beijing is influenced by multiple pollution sources [16], and the impact of regional transport cannot be ignored [17]. In recent years, with the implementation of the Air Pollution Prevention and Control Action Plan and the Three-Year Action Plan for Winning the Blue Sky Defense War, the concentration of major pollutants in Beijing has exhibited a downward trend, and the annual average PM2.5 concentration has declined by $13.67\%$ [18]. According to statistics obtained from the Beijing Municipal Ecology and Environment Bureau, the average PM2.5 concentration in 2019 (42 μg/m3) dropped by $53.1\%$ from that in 2013 (85 μg/m3) [18], and the air quality was greatly improved. In September 2020, China proposed carbon peaking and carbon neutrality goals [19]. This was followed by a series of policy measures to reduce pollution and carbon emissions and mitigate the health impacts of air pollution [20]. However, previous studies can hardly support this goal. Previous studies regarding the OP of PM2.5 in Beijing have focused on localized areas, while the sources and composition of PM2.5 have considerably changed with development. Therefore, there is a need for a comparative analysis of simultaneous observation studies in Beijing and surrounding regions. Hence, understanding the contributions of different sources to the OP value is essential for implementing relevant emission reduction measures. The OP values for different components and sources must be further explored to provide support for pollution prevention and control. Within the context of the above carbon peaking and carbon neutrality goals, this study aimed to investigate the sources of PM2.5 in Beijing and the surrounding region and to study its effects on the OP value. Summer PM2.5 monitoring and sample collection were conducted at monitoring sites in the urban areas of Beijing and suburban Gucheng, Hebei. Pearson correlation analysis (PCA) was employed to analyze the sources and concentration characteristics of the PM2.5 fractions at these two locations. Based on source analysis, the OP of PM2.5 was analyzed via the ascorbic acid (AA) method to assess the health effects of atmospheric PM2.5 in Beijing and Gucheng and to investigate the contribution of each component to the OP value. Additionally, a health risk assessment of several major elements in PM2.5 was conducted to evaluate their carcinogenic and noncarcinogenic risks to different populations. This study provides a reference basis for the development of pollution prevention and control measures and policies. ## 2.1. Site Description The sampling sites in this study were the Chinese Academy of Meteorological Sciences (CAMS) station and the Gucheng Ecological and Agricultural Meteorological Experiment Station of the CAMS. As shown in Figure 1, one site is located on Zhongguancun South Street, Haidian District, Beijing (39.93° N, 116.32° E, 53 m), and the other site is located in Gucheng town, Dingxing County, Hebei Province (39.08° N, 115.40° E, 14.2 m). The Beijing station is located at the center of the city between the Second and Third Ring Roads. There are many buildings around the station, and the flow of vehicles is high. It is affected by residents’ emissions and motor vehicle exhaust. The Gucheng station is located in an important PM2.5 regional transmission channel in Beijing and can represent the contributions of surrounding areas to pollution at this site. There are farmlands, villages, and highways around the Gucheng site, and the terrain is flat and open. In this study, PM2.5 samples were collected from 1 July to 31 July 2019. The sampling duration was 23 h, from 9:00 a.m. to 8:00 a.m. on the next day. The Beijing station adopted a TH1000 high-flow particle sampler, and the Hebei station adopted a TH-150AII medium-flow particle sampler (Tianhong Instrument Company, Wuhan, China), with sampling flow rates of 1.05 m3/min and 100 L/min, respectively. The sampling filter used in the experiment was a quartz filter produced by the PALL Company. Before sample collection, the quartz filters were baked at 600 °C for 5 h. The sampled filters were wrapped with aluminum foil and stored at −18 °C for subsequent analysis. The quality control procedures included the collection of field blanks obtained by installing filters in the sampler without air flow. Field blanks were collected before and after the sampling campaign. Meteorological data during the sampling period were obtained from the Meteorological Information Combine Analysis Process System (MICAPS). The meteorological conditions in Beijing and Gucheng during the sampling period are shown in Figure 2. The meteorological conditions during the sampling period in both places can reflect the typical characteristics of the study area in summer. Meteoinfo was used to analyze the impact of regional transport on Beijing and its surrounding areas during the sampling period. The meteorological field data used for the study were obtained from the National Centers for Environmental Prediction (NCEP) Global Data Assimilation System (GDAS) (https://www.ready.noaa.gov/archives.php (accessed on 15 October 2022)). A 72 h backward trajectory at an altitude of 500 m was simulated at an interval of six hours. ## 2.2.1. Component Analysis Organic carbon (OC) and elemental carbon (EC) were analyzed by a multiband thermal/optical carbon analyzer (DRI Model 2015, Magee, CA, USA) with a modified-A temperature protocol [21] using a 0.495 cm2 filter sample. A 2.01 cm2 filter sample was cut, sonicated in 10 mL of ultrapure water for 1 h at room temperature, and then filtered through a 0.45 μm polyethersulfone syringe filter. The Cl−, NO3−, SO42−, NH4+, K+, Mg2+, Na+, and Ca2+ concentrations in the extracts were measured via ion chromatography (ICS3000, Thermo Fisher, Waltham, MA, USA). Anions were analyzed through an IonPac AS14 column and an AG14 guard column using a 4.5 mmol/L Na2CO3 solution and a 1.4 mmol/L NaHCO3 solution, respectively, at a flow rate of 1.2 mL min−1. Cations were analyzed through an IonPac CS12 column and a CG12 guard column using a 20 mmol/L methanesulfonic acid solution at a flow rate of 1.0 mL min−1. The method detection limits for Cl−, NO3−, SO42−, NH4+, K+, Mg2+, Na+, and Ca2+ were 0.016, 0.024, 0.067, 0.022, 0.016, 0.015, 0.021, and 0.039 µg m−3, respectively. A 2.01 cm2 filter sample was cut and sonicated in 4 mL of ultrapure water for 30 min. The sample solution was filtered through a 0.45 μm polyethersulfone syringe filter, and 50 μL of concentrated nitric acid was then added. Inductively coupled plasma-mass spectrometry (7900, Agilent Technologies, Santa Clara, CA, USA) was used to determine the Cr, Mn, Fe, Co, Ni, Cu, Cd, and Pb concentrations. The method detection limits for Cr, Mn, Fe, Co, Ni, Cu, Cd, and Pb were 0.021, 0.009, 0.275, 0.001, 0.004, 0.052, 0.005, and 0.003 ng m−3, respectively. One standard sample was analyzed for every ten samples to ensure recoveries ranging from 80 to $120\%$. The chemical component data were corrected by the field blank data. ## 2.2.2. Oxidative Potential Analysis In this study, the AA method was used to determine the OP of the PM2.5 samples. A 2.01 cm2 piece was cut from the filter, sonicated in 4 mL of ultrapure water for 30 min and filtered through a 0.45 μm polyethersulfone syringe filter. Then, 30 μL of 10 mmol/L AA solution was added to 3 mL of the filtrate and thoroughly shaken. The remaining AA was determined via high-performance liquid chromatography (EX1700s, Wufeng Scientific Instruments, Shanghai, China) at 265 nm 5 times at 30 min intervals. The rate of AA consumption followed a linear relationship of AA versus time with a correlation coefficient greater than 0.985. The OP is expressed with volume normalization. [ 1]OP(pmol min−1m−1)=DS(pmol min−1)−Db(pmol min−1)Vt(m−3)×AP(cm2)At(cm2)×Vr(mL)Ve(mL) Ds is the AA consumption rate of the samples, Db is the AA consumption rate of the blank filter membrane, and Vt is the total volume sampled under standard conditions. Ap and At are the intercepted membrane area and the total membrane area, respectively. Vr and Ve are the volumes of the participating and total extraction solutions, respectively [22]. ## 2.3. Source Analysis PCA is a statistical method for correlation and analysis of variance. Matrix eigenvalues of the variance and covariance of the variables are calculated based on the initial concentrations of the components. The complex variables are downscaled to summarize the independent factors that play a major role. Based on the calculated loadings of each independent factor combined with the nature of the pollution sources of the variables, the dominant principal components are identified inductively, and the pollutant sources are analyzed qualitatively. ## 2.4. Health Risk Assessment Methods Heavy metals contained in the composition of atmospheric PM can enter the human body through breathing, contact exposure, and other means, posing a threat to human health. Health risk assessment can be performed using the model recommended by the US Environmental Protection Agency. [ 2] ADD=(C×IR×EF×ED)BW×AT [3] LADD=(C×IR×EF×ED)BW×AT ADD represents the average daily dose (mg/(kg·d)), LADD represents the life average daily dose (mg/(kg·d)), C represents the concentration of heavy metals (mg/m3), IR represents the inhalation rate (m3/d), EF represents the exposure frequency (d/a), ED represents the exposure duration (d), BW represents the body weight (kg), and AT represents the average contact time (d). The carcinogenic risk index and noncarcinogenic risk index are calculated as follows:[4] HQ=ADDRfD [5] R=LADD×SF HQ represents the hazard quotient; when HQ < 1, the noncarcinogenic risk is low, and when HQ ≥ 1, the exposure dose exceeds the threshold, and a noncarcinogenic risk may occur. RfD represents the reference dose (mg/(kg·d)). R also represents the reference dose, and a value of R < 1 × 10−6 indicates the acceptable level of the population. When 1 × 10−6 < R < 1 × 10−4, there is a potential carcinogenic risk. A value of R > 1 × 10−4 indicates that pollution may pose a cancer risk to the population. SF represents the carcinogenic intensity coefficient of a carcinogenic pollutant (kg·day/mg). The values of these parameters are shown in Table 1 and Table 2 (US EPA and Ministry of Ecology and Environment of the People’s Republic of China) [23,24,25,26]. ## 3.1.1. PM2.5 Concentration Characteristics The ranges of PM2.5 concentration changes at the sites in Beijing and Gucheng during the sampling period were 7.1~79.0 μg/m3 and 11.6~68.5 μg/m3, respectively. The monthly average mass concentrations of PM2.5 in Beijing and Gucheng were 34.0 ± 6.1 μg/m3 and 37.1 ± 6.9 μg/m3, respectively. Figure 3 presents the variation in daily average PM2.5 concentrations in July 2019 at the sampling sites in Beijing and Gucheng. Beijing and Gucheng are far from each other, but the PM2.5 daily average concentration trends were similar. This is mainly because the PM2.5 pollution has regional characteristics. The average mass concentration of PM2.5 in Gucheng is 1.09 times higher than that in Beijing. The daily average concentration of PM2.5 in Gucheng during the sampling period met the ambient air quality standard level II of PM2.5 concentration (75 μg/m3), but the Beijing PM2.5 concentration exceeded the standard limit on 26 July (79 μg/m3). According to meteorological data (Figure 2) during the sampling period, the day featured high temperatures, relatively low wind speeds, and a gradual increase in humidity. The influence of meteorological conditions on air quality cannot be ignored because they influence the concentration and distribution of various atmospheric pollutants. Therefore, this study focuses on the temperature, wind speed, and relative humidity, which affect pollution generation [27]. Figure 4 shows the fitted plots of the correlation between meteorological elements (temperature, wind speed, relative humidity) and PM2.5 concentrations at the two locations during the sampling period. The figure shows that temperature and relative humidity exhibit a positive correlation with the PM2.5 concentration at the two sites, while wind speed exhibits a more obvious negative correlation. The higher the temperature and relative humidity are, the higher the PM2.5 concentration. In contrast, the higher the wind speed is, the lower the PM2.5 pollution. Higher temperatures in summer can promote the formation of secondary aerosols, which can lead to an increase in PM2.5 concentrations [28,29]. Higher relative humidity is conducive to the conversion of precursor gases (SO2 and NOx) to PM2.5. Wind speed and wind direction are important meteorological factors affecting the accumulation, diffusion, and regional transmission of atmospheric pollutants [30]. The two sampling sites are located in the North China Plain, and the effects of these meteorological parameters on PM2.5 concentrations are similar to those in a previous study conducted by Yang et al. [ 29]. PM2.5 concentrations are influenced not only by local sources but also by regional transport. Data obtained from the Beijing Municipal Bureau of Ecology and Environment, covering the third round of PM2.5 source analysis in Beijing, indicated that regional transmission accounts for approximately $40\%$ of the total concentration [31]. The trajectory clustering results are shown in Figure 5. Six clustering trajectories were obtained in Beijing (a), and five clustering trajectories were obtained in Gucheng (b). The trajectories of clustered trajectories in Beijing during the sampling period were divided into six clusters. Cluster 1 involves short-range transport from the east over the Bohai Sea, Hebei, and Tianjin, which occurs frequently and accounts for $37.5\%$ of the total trajectories. As indicated in Table 3, the PM2.5 concentration is 35.1 μg/m3, which is probably due to the transport of industrial source emissions from Hebei and Tianjin. Clusters 2 and 6 represent long-range transport from the west and northwest via Mongolia and Inner Mongolia, respectively, with relatively high PM2.5 concentrations of 31.0 and 27.0 μg/m3, respectively. This material may represent sand and dust from the arid Gobi regions of Mongolia and Inner Mongolia. Cluster 3 comes from the southeast via the Bohai Sea, Shandong, and Hebei with a PM2.5 concentration of 43.3 μg/m3. This cluster may be related to marine transport emissions. Cluster 4 involves short-range transport from southern Hebei with the highest PM2.5 concentration of 48.6 μg/m3. This trajectory passes over dense industrial cities in southern Hebei and is associated with industrial source emissions. Cluster 5 is the smallest and cleanest proportion of long-distance transport from northeastern Inner Mongolia. There are five clustered trajectories in Gucheng. Cluster 1 comes from the southeast via the Bohai Sea and Shandong, accounting for $34.82\%$ of the total trajectories. The high PM2.5 concentration may originate from marine transportation pollution emissions. Cluster 2 involves long-distance transport from Inner Mongolia. Cluster 3 involves short-distance transport from southern Hebei and Shandong, accounting for $32.14\%$ of the total trajectories. The dense industry in the Handan area in southern Hebei may lead to its high PM2.5 concentration. Cluster 4 is relatively clean and is from northeastern Inner Mongolia. Cluster 5 involves short-range transport from the southwest through Shanxi. The highest PM2.5 concentrations may originate from emissions from industrial activities such as coal mining and metallurgy in Shanxi. ## 3.1.2. Characteristics of the Chemical Components in PM2.5 The average concentrations of atmospheric PM2.5 carbonaceous components and water-soluble ions obtained for Beijing and Gucheng are shown in Figure 6. Figure 6a shows the daily carbon component (OC and EC) and water-soluble element component in Beijing in summer, (b) shows Gucheng, and (c) shows the average concentration of the two sites. Figure 7 shows that OC, EC, and secondary ions (SO42−, NO3−, and NH4+) exhibited high proportions in PM2.5 at both locations. The highest NO3− proportion was $23.23\%$ in Beijing, and the rest comprised SO42−, OC, NH4+, other components and EC in this order. The highest SO42− content was $25.09\%$ in Gucheng, and the rest comprised NO3−, NH4+, OC, other components and EC in this order. OC and EC are important components of PM2.5. Most of the EC comes from the inadequate combustion of carbon-containing fuels. OC comes from direct emissions of primary OC and secondary OC generated by photochemical reactions [32]. The average concentrations of OC and EC in Beijing PM2.5 are 6.34 and 1.96 μg/m3, respectively. As shown in Figure 7, OC and EC account for $18.67\%$ and $5.76\%$ of the total PM2.5 concentration in Beijing, respectively. However, OC and EC accounted for $14.74\%$ and $7.58\%$ in Gucheng, respectively. A previous study showed that the main contributor to the OC and EC concentrations in Beijing was gasoline vehicles [33,34]. The differences in the OC and EC percentages between the two sites may originate from their different sources of PM2.5. The higher EC percentage in Gucheng was from the inadequate combustion of industrial coal emissions. The SO42−, NO3−, and NH4+ concentrations were also high at both sites. These components are mainly influenced by gaseous precursors: SO2, NOx, and NH3. The average concentrations of most ions in Gucheng are higher than those in Beijing. According to the report on the State of the Ecology and Environment in China in 2019, Hebei’s SO2 emissions are much larger than Beijing’s [18]. The high SO2 emissions in Hebei lead to elevated levels of SO42−. The Hebei region has high levels of industrial coal-fired emissions, while Beijing is dominated by motor vehicle and domestic source emissions. From Figure 7, NH4+ accounted for $16.34\%$ and $17.52\%$ in Beijing and Gucheng, respectively. NH4+ is mostly converted from NH3 in the atmosphere, and NH3 mainly comes from the use of ammonia fertilizer in agricultural and livestock production and other emission sources, such as soil microbial ammonification processes [35]. The fertilization of farmland around Gucheng made the percentage of NH4+ slightly higher in Gucheng than in Beijing. The percentage of NO3− is similar between Beijing and Gucheng. In recent years, the per capita ownership of motor vehicles in both Beijing and Hebei has increased rapidly, and the emission of vehicle exhaust has released a large amount of NOx into the atmosphere. In addition, the implementation of the coal to gas transition policy in the Beijing–Tianjin–Hebei region has led to an increase in the concentration of NO in the atmosphere [36]. Figure 8 shows the concentrations of the water-soluble elements at the two sites. As shown in Figure 8, most element concentrations were slightly higher in Gucheng than in Beijing. The data statistics showed that the elements with higher concentrations in PM2.5 at both sites were Zn, Fe, Pb, and Mn. This occurred because these polluting elements are mainly influenced by the surrounding production and living processes. The Zn contents in Beijing and Hebei were the highest. According to the study, Zn pollution was mainly attributed to the wear of motor vehicle tires and brake pads. The average concentrations of these metal elements in Beijing were 124.96, 66.97, 13.06 and 12.55 ng/m3, respectively. In Gucheng, the concentrations were higher than those in Beijing, at 127.25, 81.96, 33.65 and 18.72 ng/m3, respectively. The Cu, As, Ni, Cr, Cd and Co concentrations were relatively low. The concentrations of these elements in Beijing were 5.52, 4.72, 1.02, 1.14, 0.48, and 0.08 ng/m3, respectively, and in Gucheng, the concentrations of these elements were 8.30, 10.84, 2.19, 2.79, 1.44, and 0.14 ng/m3, respectively. This may be due to the higher industrial activity around Gucheng. ## 3.2. PM2.5 Source Analysis To further study the pollution characteristics of PM2.5 in Beijing and the surrounding regions in the summer, source analysis of the PM2.5 fractions was conducted. In this study, PCA of the carbonaceous fraction and water-soluble ion concentrations (OC, EC, Na+, NH4+, K+, Mg2+, Ca2+, Cl−, NO3−, and SO42−) of PM2.5 at the monitoring stations in Beijing and Gucheng was performed using SPSS 26.0. The data measured at the Beijing and Gucheng monitoring sites were imported into SPSS for PCA to obtain the results presented in Table 4 and Table 5. Table 4 shows that ten principal components were extracted, but only principal component 1 and principal component 2 have eigenvalues greater than 1, namely, 5.526 and 1.985, respectively. The cumulative contribution is $72.4\%$. This indicates that these two principal components can explain most of the data and reflect the main pollution sources in Beijing. Similarly, Table 5 shows that 10 principal components were also extracted for Gucheng with only principal component 1 and principal component 2 having eigenvalues greater than 1, namely, 6.542 and 1.817, respectively. The cumulative contribution is $76.0\%$. This indicates that these two principal components can explain most of the data and reflect the main pollution sources around Gucheng. Table 6 shows that the eigenvalues of principal component 1 are high in Beijing and Gucheng. The correlation coefficients of NH4+, SO42−, NO3−, K+, and EC in Beijing’s atmospheric PM2.5 were high in principal component 1. The correlation coefficients were 0.959, 0.917, 0.857, 0.848, and 0.802: all above 0.8. OC, Mg2+, and Ca2+ had high correlation coefficients of 0.765, 0.677, and 0.673 in principal component 2, respectively. The highest correlation coefficients of OC, K+, NH4+, Mg2+, and Na+ were 0.967, 0.900, 0.894, 0.867, and 0.795, respectively, in principal component 1 of atmospheric PM2.5 in Gucheng. The correlation coefficients of Cl−, Ca2+ and EC in principal component 2 were 0.708, 0.671, and 0.389, respectively. These results show that there are good correlations between all the above ions and the principal components. The ions with high correlation coefficients with principal component 1 in Beijing in the summer are NH4+, SO42−, NO3−, K+, and EC, indicating that the main component of air pollution in the Beijing urban area is SNA. Therefore, principal component 1 represents the secondary component, biomass combustion, and vehicle exhaust emissions. The ions with higher correlation coefficients with principal component 2 are OC, Mg2+, and Ca2+, so principal component 2 represents dust and coal combustion. The ions in the atmosphere of Gucheng with high correlations with principal component 1 are OC, K+, NH4+, Mg2+, and Na+, so principal component 1 may represent industrial coal combustion, dust, and biomass fuel combustion. The ions with high correlation coefficients with principal component 2 are Cl−, Ca2+, and EC. Therefore, principal component 2 represents coal combustion and dust. In summary, the results of the PCA show that the air pollution in Beijing in summer mainly comes from vehicle exhaust emissions and secondary components. Wang et al. [ 37] used PCA to analyze carbonaceous aerosols in Beijing in the summer and found that the contribution of motor vehicle emissions reached more than $60\%$. The air pollution in Gucheng mainly comes from industrial coal combustion, dust, and biomass combustion. The two sites have a relatively clear multisource and multipathway composite pollution phenomenon. In general, our results are consistent with previous studies, and secondary sources, vehicle emissions, and regional transport were the main contributors to PM2.5 pollution in Beijing [38]. ## 3.3. Oxidative Potential Analysis The AA method was used to determine the OP of the PM2.5 samples collected from Beijing and Gucheng. The average OP values in Beijing and Gucheng were 91.6 ± 42.1 and 82.2 ± 47.1 pmol/(min·m3), respectively. Figure 9a,b compares the daily PM2.5 concentrations and OP values in Beijing and Gucheng, respectively. The trends of Beijing and Gucheng are roughly the same, but the average OP value is slightly higher in Beijing than in Gucheng. Considering the ambient air quality index of China and the PM2.5 concentration in these measurements, days with average PM2.5 values below 35 μg/m3 are defined as clean days, days with values above 35 μg/m3 and below 75 μg/m3 are defined as lightly polluted days, and days with values above 75 μg/m3 are defined as polluted days. The counting results showed that the average OP values on lightly polluted days in Beijing and Gucheng were 120.0 and 106.4 pmol/(min·m3), respectively. These values were 1.59 times and 1.5 times greater than the values on clean days (75.3 and 70.7 pmol/(min·m3)). As shown in Figure 10, the OP values in Beijing and Gucheng show a trend of increasing with increasing PM2.5 concentration. To further investigate the contributions of the different chemical components of PM2.5 to the OP in Beijing and Gucheng, Pearson correlation coefficients were used to analyze the correlations between the PM2.5 components and OP values using SPSS 26.0. The Pearson correlation coefficients between the chemical components and the OP of PM2.5 at the two locations are shown in Figure 11. In Beijing, the water-soluble ions Cl− and NO3− in PM2.5 were moderately correlated (0.4 ≤ r <0.6, $p \leq 0.01$) with the OP. This may be due to the significant contribution of motor vehicle emissions as a source of PM2.5 in Beijing. In Gucheng, K+ had a strong correlation ($r = 0.61$) with the OP. Na+, SO42−, and Mg2+ had a moderate correlation with the OP. The correlation of K+ may originate from the contribution of biomass burning to PM2.5 in Gucheng [39]. SO42− is influenced by its gaseous precursor SO2, which may be associated with the more frequent industrial activities around Gucheng. Mg2+ is mainly from dust generated by construction activities. The EC in Gucheng PM2.5 is also moderately correlated with the OP, which may originate from the fossil fuel consumption caused by industrial activities and vehicle exhaust in the area. The contribution of water-soluble elements to OP is important, and it can induce the production of ROS [11]. In Beijing, there is a strong correlation between Cu ($r = 0.63$) and the generation of OP. Co and Ni have a moderate correlation with the OP. These elements may come from vehicle exhaust and brake wear [40]. In Gucheng, Cu, Pb, Mn, and Sb had strong correlations ($r = 0.85$, $r = 0.67$, $r = 0.65$, $r = 0.62$, respectively) with the OP. Ba, Fe, Cd, Sn, Zn, and Cr show moderate correlations with the OP (r > 0.5). Previous studies have shown that Cu, Pb, Fe, and Zn come from vehicle exhaust and industrial activities [41]. Mn, Sb, Ba, Cd, Sn, and Cr may originate from industrial activities and coal combustion [40]. Hence, industrial activities, vehicle exhaust, and coal combustion may be the main contributing sources to the OP in Gucheng. The relationships between the chemical components and the OP of PM2.5 in Beijing and Gucheng are very different. This mainly occurs because the PM2.5 sources are different between the two regions, and the contributions of each component to the PM2.5 concentrations are different, resulting in different contributions to the production of OP. The OP of PM2.5 in Beijing is mainly affected by vehicle exhaust, while that in Gucheng is also affected by dust and combustion sources, such as industrial coal burning and biomass burning. ## 3.4. Health Risk Assessment Noncarcinogenic risk calculations were performed for Cu, Zn, Pb, and Mn in the samples collected in Beijing and Gucheng. The risk assessment value of each element was calculated according to Equations [2]–[5]. Carcinogenic risk assessment calculations were performed for Cr, Ni, Cd, and As. The results are shown in Figure 12 and Figure 13. Figure 12 shows that for all populations, the noncarcinogenic risk values of Pb and Mn were high, and in Beijing, the HQ values of Pb and Mn for children and adolescents, adult females, and adult males were (5.73 ± 3.37) × 10−3, (7.13 ± 4.18) × 10−3, (7.49 ± 4.39) × 10−3, (7.91 ± 3.57) × 10−3, (9.82 ± 4.43) × 10−3, and (1.03 ± 0.47) × 10−3, respectively. In Gucheng, the HQ values of Pb were higher than those in Beijing, at (1.48 ± 1.11) × 10−2, (1.84 ± 1.38) × 10−2, and (1.93 ± 1.45) × 10−2, respectively. The HQ values of Cu and Zn at both locations were less than 5 × 10−3. However, the HQ values of the noncarcinogenic risk factors are all less than 1. Therefore, the noncarcinogenic risks of Cu, Zn, Pb, and Mn are negligible. The dashed line in Figure 13 denotes $R = 10$−6, which shows that Cr and As exceed this value in Beijing and that Cr, Cd and As exceed this value in Gucheng. Please refer to the determination of the R value in Section 2.4. The carcinogenic risk factors for Ni and Cd in Beijing are less than 10−6 for children, adolescents and adults, so the risk is negligible. The cancer risk factor values for Cr and As are between 10−6 and 10−4, indicating a potential risk for children, adolescents, and adults. The cancer risk factors for adults are higher than those for children and adolescents. The risk assessment calculations in Gucheng are higher than those in Beijing. The risk factor for Ni is less than 10−6 for all populations, and therefore, the risk of Ni-induced cancer is negligible. The risk factor for Cd is less than 10−6 for children and adolescents. Adults are in the range of 10−6 and 10−4, representing a potential cancer risk. The carcinogenic risk factor values for Cr and As are between 10−6 and 10−4, indicating a potential risk for all populations, similar to the situation for Beijing. The risk factors for adults are higher than those for children and adolescents. In summary, Cr and As have potential carcinogenic risks for all populations at both sites, while Cd has a potential carcinogenic risk for adults in Gucheng. The experimental results are similar to the PM2.5 health risk assessment results of Li et al. for 29 provincial capitals in China [42], with Cr and As posing higher carcinogenic risks. ## 4. Conclusions In this study, we found that Beijing and Gucheng exhibited high concentrations of OC, EC, SO42−, NO3−, and NH4+ in PM2.5. The highest NO3− content was $23.23\%$ in Beijing, and the highest SO42− content was $25.09\%$ in Gucheng. Then, we employed PCA and determined that the main sources of PM2.5 in Beijing were motor vehicle emissions and secondary components, while industrial emissions dominated in Gucheng. Moreover, both areas were influenced by airflows from the eastern, southern and southwestern directions. The AA method was used to analyze the OP values at the two sites. The contributions of Cu, Cl−, and NO3− to the OP values were larger in Beijing, and the contributions of Cu, K+, Pb, and SO42− were notable. This is related to the main sources of PM2.5 at both locations. The health risk assessment results indicated potential carcinogenic risks of Cr and As for all populations at both sites and a potential carcinogenic risk of Cd for adults in Gucheng. In the future, when studying or developing pollution prevention and control measures in the future, more attention should be given to the health effects of motor vehicle emissions in Beijing and industrial emission control in Hebei, and the effects of regional transmission should be fully considered. In this paper, the source of secondary components is still unknown, to better explore the source of all components, chemical model combined with different seasons experiment results still needs to be further conducted. ## References 1. Qi G., Wang Z., Wei L., Wang Z.. Multidimensional effects of urbanization on PM2.5 concentration in China. Environ. Sci. Pollut. Res. Int. (2022.0) 29 77081-77096. DOI: 10.1007/s11356-022-21298-4 2. Maji K.J., Dikshit A.K., Arora M., Deshpande A.. Estimating premature mortality attributable to PM2.5 exposure and benefit of air pollution control policies in China for 2020. Sci. Total Environ. (2018.0) 612 683-693. DOI: 10.1016/j.scitotenv.2017.08.254 3. Chan C.K., Yao X.. Air pollution in mega cities in China. Atmos. Environ. (2008.0) 42 1-42. DOI: 10.1016/j.atmosenv.2007.09.003 4. Pui D.Y.H., Chen S.-C., Zuo Z.. PM 2.5 in China: Measurements, sources, visibility and health effects, and mitigation. Particuology (2014.0) 13 1-26. DOI: 10.1016/j.partic.2013.11.001 5. Xing Y.F., Xu Y.H., Shi M.H., Lian Y.X.. The impact of PM2.5 on the human respiratory system. J. Thorac. Dis. (2016.0) 8 E69-E74. DOI: 10.3978/j.issn.2072-1439.2016.01.19 6. Sharma S., Chandra M., Kota S.H.. Health Effects Associated with PM2.5: A Systematic Review. Curr. Pollut. Rep. (2020.0) 6 345-367. DOI: 10.1007/s40726-020-00155-3 7. Valavanidis A., Vlachogianni T., Fiotakis K., Loridas S.. Pulmonary oxidative stress, inflammation and cancer: Respirable particulate matter, fibrous dusts and ozone as major causes of lung carcinogenesis through reactive oxygen species mechanisms. Int. J. Environ. Res. Public Health (2013.0) 10 3886-3907. DOI: 10.3390/ijerph10093886 8. Feng S., Gao D., Liao F., Zhou F., Wang X.. The health effects of ambient PM2.5 and potential mechanisms. Ecotoxicol. Environ. Saf. (2016.0) 128 67-74. DOI: 10.1016/j.ecoenv.2016.01.030 9. Wang Y., Wang M., Li S., Sun H., Mu Z., Zhang L., Li Y., Chen Q.. Study on the oxidation potential of the water-soluble components of ambient PM2.5 over Xi’an, China: Pollution levels, source apportionment and transport pathways. Environ. Int. (2020.0) 136 105515. DOI: 10.1016/j.envint.2020.105515 10. Cao W., Wang X., Li J., Yan M., Chang C.H., Kim J., Jiang J., Liao Y.P., Tseng S., Kusumoputro S.. NLRP3 inflammasome activation determines the fibrogenic potential of PM2.5 air pollution particles in the lung. J. Environ. Sci. (2022.0) 111 429-441. DOI: 10.1016/j.jes.2021.04.021 11. Fang T., Verma V., Bates J.T., Abrams J., Klein M., Strickland M.J., Sarnat S.E., Chang H.H., Mulholland J.A., Tolbert P.E.. Oxidative potential of ambient water-soluble PM. Atmos. Chem. Phys. (2016.0) 16 3865-3879. DOI: 10.5194/acp-16-3865-2016 12. Liu Q., Lu Z., Xiong Y., Huang F., Zhou J., Schauer J.J.. Oxidative potential of ambient PM2.5 in Wuhan and its comparisons with eight areas of China. Sci. Total Environ. (2020.0) 701 134844. DOI: 10.1016/j.scitotenv.2019.134844 13. Lim C.C., Thurston G.D.. Air Pollution, Oxidative Stress, and Diabetes: A Life Course Epidemiologic Perspective. Curr. Diab. Rep. (2019.0) 19 58. DOI: 10.1007/s11892-019-1181-y 14. Li Z., Tang Y., Song X., Lazar L., Li Z., Zhao J.. Impact of ambient PM2.5 on adverse birth outcome and potential molecular mechanism. Ecotoxicol. Environ. Saf. (2019.0) 169 248-254. DOI: 10.1016/j.ecoenv.2018.10.109 15. Rich D.Q., Zhang W., Lin S., Squizzato S., Thurston S.W., van Wijngaarden E., Croft D., Masiol M., Hopke P.K.. Triggering of cardiovascular hospital admissions by source specific fine particle concentrations in urban centers of New York State. Environ. Int. (2019.0) 126 387-394. DOI: 10.1016/j.envint.2019.02.018 16. Gao J., Peng X., Chen G., Xu J., Shi G.L., Zhang Y.C., Feng Y.C.. Insights into the chemical characterization and sources of PM(2.5) in Beijing at a 1-h time resolution. Sci. Total Environ. (2016.0) 542 162-171. DOI: 10.1016/j.scitotenv.2015.10.082 17. Han L., Cheng S., Zhuang G., Ning H., Wang H., Wei W., Zhao X.. The changes and long-range transport of PM2.5 in Beijing in the past decade. Atmos. Environ. (2015.0) 110 186-195. DOI: 10.1016/j.atmosenv.2015.03.013 18. Ministry of Ecology and Enviroment of the People’s Republic of China 19. Shi X., Zhang J., Lu S., Wang T., Zhang X.. China Carbon Neutralization Research Status and Research Frontier Tracking. Front. Environ. Sci. (2022.0) 10 896524. DOI: 10.3389/fenvs.2022.896524 20. Guo B., Feng Y., Wang Y., Lin J., Zhang J., Wu S., Jia R., Zhang X., Sun H., Zhang W.. Influence of carbon emission trading policy on residents’ health in China. Front. Public Health (2022.0) 10 1003192. DOI: 10.3389/fpubh.2022.1003192 21. Cao J.. Characteristics of carbonaceous aerosol in Pearl River Delta Region, China during 2001 winter period. Atmos. Environ. (2003.0) 37 1451-1460. DOI: 10.1016/S1352-2310(02)01002-6 22. Liu L., Zhang L., Wen W., Jiao J., Cheng H., Ma X., Sun C.. Chemical composition, oxidative potential and identifying the sources of outdoor PM2.5 after the improvement of air quality in Beijing. Environ. Geochem. Health (2022.0). DOI: 10.1007/s10653-022-01275-z 23. Wang Z., Yan J., Zhang P., Li Z., Guo C., Wu K., Li X., Zhu X., Sun Z., Wei Y.. Chemical characterization, source apportionment, and health risk assessment of PM2.5 in a typical industrial region in North China. Environ. Sci. Pollut. Res. Int. (2022.0) 29 71696-71708. DOI: 10.1007/s11356-022-19843-2 24. Han Y., Wang Z., Zhou J., Che H., Tian M., Wang H., Shi G., Yang F., Zhang S., Chen Y.. PM2.5-Bound Heavy Metals in Southwestern China: Characterization, Sources, and Health Risks. Atmosphere (2021.0) 12. DOI: 10.3390/atmos12070929 25. Li F., Yan J., Wei Y., Zeng J., Wang X., Chen X., Zhang C., Li W., Chen M., Lü G.. PM2.5-bound heavy metals from the major cities in China: Spatiotemporal distribution, fuzzy exposure assessment and health risk management. J. Clean. Prod. (2021.0) 286 124967. DOI: 10.1016/j.jclepro.2020.124967 26. Chen P., Bi X., Zhang J., Wu J., Feng Y.. Assessment of heavy metal pollution characteristics and human health risk of exposure to ambient PM2.5 in Tianjin, China. Particuology (2015.0) 20 104-109. DOI: 10.1016/j.partic.2014.04.020 27. Chen Z., Chen D., Zhao C., Kwan M.P., Cai J., Zhuang Y., Zhao B., Wang X., Chen B., Yang J.. Influence of meteorological conditions on PM2.5 concentrations across China: A review of methodology and mechanism. Environ. Int. (2020.0) 139 105558. DOI: 10.1016/j.envint.2020.105558 28. Li J.J., Wang G.H., Wang X.M., Cao J.J., Sun T., Cheng C.L., Meng J.J., Hu T.F., Liu S.X.. Abundance, composition and source of atmospheric PM2.5 at a remote site in the Tibetan Plateau, China. Tellus B Chem. Phys. Meteorol. (2013.0) 65 20281. DOI: 10.3402/tellusb.v65i0.20281 29. Yang Q., Yuan Q., Li T., Shen H., Zhang L.. The Relationships between PM2.5 and Meteorological Factors in China: Seasonal and Regional Variations. Int. J. Environ. Res. Public Health (2017.0) 14. DOI: 10.3390/ijerph14121510 30. Ma S., Shao M., Zhang Y., Dai Q., Xie M.. Sensitivity of PM. Sci. Total Environ. (2021.0) 792 148474. DOI: 10.1016/j.scitotenv.2021.148474 31. Beijing Municipal Ecology and Enviroment Bureau 32. Wang Z., Zhang D., Liu B., Li Y., Chen T., Sun F., Yang D., Liang Y., Chang M., Yang L.. Analysis of chemical characteristics of PM2.5 in Beijing over a 1-year period. J. Atmos. Chem. (2016.0) 73 407-425. DOI: 10.1007/s10874-016-9334-8 33. Jiang J., Hou L., Yan X., Liu H., Sun S., Li S., Xiong K., Zhao W.. Simultaneous Measurements of Carbonaceous Aerosol in Three Major Cities in the Beijing-Tianjin-Hebei Region, China. Aerosol Air Qual. Res. (2022.0) 22 220056. DOI: 10.4209/aaqr.220056 34. Qi M., Jiang L., Liu Y., Xiong Q., Sun C., Li X., Zhao W., Yang X.. Analysis of the Characteristics and Sources of Carbonaceous Aerosols in PM2.5 in the Beijing, Tianjin, and Langfang Region, China. Int. J. Environ. Res. Public Health (2018.0) 15. DOI: 10.3390/ijerph15071483 35. Wang X., Wei W., Cheng S., Yao S., Zhang H., Zhang C.. Characteristics of PM2.5 and SNA components and meteorological factors impact on air pollution through 2013–2017 in Beijing, China. Atmos. Pollut. Res. (2019.0) 10 1976-1984. DOI: 10.1016/j.apr.2019.09.004 36. Zhao S., Hu B., Gao W., Li L., Huang W., Wang L., Yang Y., Liu J., Li J., Ji D.. Effect of the “coal to gas” project on atmospheric NOX during the heating period at a suburban site between Beijing and Tianjin. Atmos. Res. (2020.0) 241 104977. DOI: 10.1016/j.atmosres.2020.104977 37. Wang G., Cheng S., Li J., Lang J., Wen W., Yang X., Tian L.. Source apportionment and seasonal variation of PM2.5 carbonaceous aerosol in the Beijing-Tianjin-Hebei region of China. Environ. Monit. Assess. (2015.0) 187 143. DOI: 10.1007/s10661-015-4288-x 38. Lv L., Chen Y., Han Y., Cui M., Wei P., Zheng M., Hu J.. High-time-resolution PM2.5 source apportionment based on multi-model with organic tracers in Beijing during haze episodes. Sci. Total Environ. (2021.0) 772 144766. DOI: 10.1016/j.scitotenv.2020.144766 39. Rai P., Furger M., Slowik J.G., Zhong H., Tong Y., Wang L., Duan J., Gu Y., Qi L., Huang R.J.. Characteristics and sources of hourly elements in PM10 and PM2.5 during wintertime in Beijing. Environ. Pollut. (2021.0) 278 116865. DOI: 10.1016/j.envpol.2021.116865 40. Zheng M., Salmon L.G., Schauer J.J., Zeng L., Kiang C.S., Zhang Y., Cass G.R.. Seasonal trends in PM2.5 source contributions in Beijing, China. Atmos. Environ. (2005.0) 39 3967-3976. DOI: 10.1016/j.atmosenv.2005.03.036 41. Watson J.G., Chow J.C., Houck J.E.. PM2.5 chemical source profiles for vehicle exhaust, vegetative burning, geological material, and coal burning in Northwestern Colorado during 1995. Chemosphere (2001.0) 43 1141-1151. DOI: 10.1016/S0045-6535(00)00171-5 42. Li L., Meng R., Lei Y., Wu S., Jiang Y.. Human health risk assessment of heavy metals from PM2.5 in China’s 29 provincial capital cities. Environ. Sci. Pollut. Res. Int. (2022.0) 29 63028-63040. DOI: 10.1007/s11356-022-20229-7
--- title: '“I’m Hooked on e-cycling, I Can Finally Be Active Again”: Perceptions of e-cycling as a Physical Activity Intervention during Breast Cancer Treatment' authors: - Kirsty Mollie Way - Jessica Elizabeth Bourne - Miranda Elaine Glynis Armstrong journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049330 doi: 10.3390/ijerph20065197 license: CC BY 4.0 --- # “I’m Hooked on e-cycling, I Can Finally Be Active Again”: Perceptions of e-cycling as a Physical Activity Intervention during Breast Cancer Treatment ## Abstract Electrically-assisted bicycles (e-bikes) are a means through which to increase individual physical activity (PA) and overcome some commonly reported barriers to engaging in conventional cycling. Fatigue is a common side effect to breast cancer treatment, and the rate of PA engagement drops significantly following a breast cancer diagnosis. The aim of this qualitative study was to examine perceptions of e-cycling as a means of increasing PA in this population. Twenty-four participants (mean age = 57.88 (standard deviation 10.8), $100\%$ female) who have had a breast cancer diagnosis, completed two semi-structured interviews via Zoom. One interview was conducted prior to an e-bike taster session and a second, after the session. Taster sessions were conducted by certified cycling instructors in the community. Interviews were conducted between December 2021 and May 2022. Data were transcribed verbatim and analyzed thematically using NVivo 12 software. An inductive and deductive approach to analysis was adopted. Five themes were generated: [1] Perceived role of e-bikes during treatment, [2] The relationship between e-bikes and fatigue, [3] Cancer-specific considerations, [4] Is e-cycling ‘enough’?, and [5] Optimizing the intervention. Negative perceptions of e-bikes noted before the taster session were altered following riding an e-bike. The multiple levels of assistance made cycling manageable and less impacted by fatigue, thereby enabling individuals to re-establish previous cycling habits. E-cycling may be a suitable option to increase PA behavior amongst individuals being treated for breast cancer, with the potential to overcome many of the barriers of conventional cycling. Enabling this population to trial an e-bike elicits positive physical and psychological responses that may help to promote future engagement. ## 1. Introduction Cancer is one of the leading causes of mortality worldwide, totaling approximately 10 million deaths in 2020 [1,2]. Breast cancer is the most common cancer in the UK, accounting for $15\%$ of all new cancer cases and approximately 55 million diagnoses, yearly [3]. There are a multitude of risk factors which contribute to the disease’s incidence, mortality, and survival rates, including: lifestyle factors, hereditary contributions, socioeconomic status, environment, and culture [4,5,6]. Despite some men developing the condition ($1\%$ of UK breast cancer cases) [3], breast cancer prevalence is greatest in women aged 50 and over, with ageing being the second greatest risk factor for the disease behind sex [7]. Cancer places great economic burden on society, with the cost of breast cancer chemotherapy alone mounting to over GBP 248 million, yearly [8]. Yet the financial costs of breast cancer can stretch further than direct treatment, with additional funding required for short- and long-term work absence, informal care provision, and carer emotional well-being [8]. Due to the high prevalence of breast cancer and associated costs, there is a need to identify effective lifestyle modifications for the tertiary prevention of breast cancer and associated mortality. ## 1.1. Breast Cancer and Physical Activity Physical activity (PA) has been identified as a highly effective, non-pharmaceutical intervention known to complement other breast cancer treatments and modify recurrence rates [9]. The benefits of engaging in PA are obtainable even if the activity is delayed until 12 months post diagnosis [10]. PA guidelines for a breast cancer diagnosis do not differ from the conventional guidelines. As such, individuals should aim to partake in 150 min of moderate intensity activity per week, or 75 min of vigorous activity per week, with resistance training incorporated twice a week [11]. A prospective study of 1340 breast cancer patients found that individuals who met the PA guidelines both before and one year post diagnosis had a $41\%$ reduced risk of breast cancer recurrence, and $49\%$ reduced risk of mortality from breast cancer, when compared to physically inactive individuals; these associations were strengthened after 2 years ($65\%$ and $68\%$, respectively) [12]. In addition, in oncology, PA engagement is recognized to improve both psychological and physiological functions, including the potential to mitigate common side effects such as incidence of fatigue, impaired health-related quality of life (HRQL), mental health decline, and reductions in physical health [13]. It is important to note that an individual’s willingness and their physical and psychological responses to PA can vary depending on the type of treatment they are receiving. A cross-sectional study of 37 breast cancer patients revealed that patients were more likely to experience perceived difficulty engaging in PA if they had undergone more than three types of cancer treatment, in comparison to patients that experience no perceived difficulty (RR 2.14; $95\%$ CI 1.07 to 4.27) [14]. ## 1.2. Participation Rates and Barriers to Physical Activity Despite the known benefits of PA and the intention to increase their PA participation once diagnosed, the number of women meeting national PA guidelines is low [15]. It is deemed challenging to find the motivation and capability to engage in PA following treatment, which is described as a debilitating process, thus limiting physical ability [16]. In 2012, a cohort study revealed that only $48.3\%$ of breast cancer patients met the PA guidelines of 10 metabolic equivalent (MET) hours/week [17]. While these guidelines are outdated, they do suggest a decline in PA participation post diagnosis. A 10-year cohort study of 634 breast cancer patients identified that pre-diagnosis, $34.0\%$ of women met the PA guidelines; this percentage stayed consistent at 24 months post enrollment ($34.0\%$) but decreased to $21.4\%$ after 10 years [18]. There are several perceived barriers to PA engagement during breast cancer treatment, including physical, psychosocial, environmental, and organizational factors [19]. Physically, women report impairments that arise due to surgery, such as shoulder problems, that limit PA engagement [19,20,21]. However, the most commonly reported barriers to PA engagement are lack of energy and fatigue, despite evidence to suggest that PA is beneficial for improving fatigue and boosting energy levels [19,22,23]. As such, it is important to identify different modalities of PA that are appealing to this population to promote engagement [24]. Smith-Turchyn and colleagues’ [25] qualitative research details how healthcare professionals vary the PA guidance they provide to patients depending on the treatment the patient is undergoing. Despite research promoting the benefits of PA during adjuvant chemotherapy [26], women tend to show an unwillingness to participate due to side effects such as nausea, fatigue, and emotional shock. In contrast, a five year longitudinal study found women undergoing chemotherapy gradually increased their participation in PA during the first 18 months of treatment, followed by a steady decline thereafter [27]. Emery and colleagues’ [27] research found that women are more likely to undertake PA if they had a lumpectomy, not a mastectomy. Therefore, it is important to note that physical implications resulting from surgery, such as restricted arm movement, can impact an individual’s perceived ability to take part in PA, as well as their self-efficacy and desire to be physically active [28]. Limited research has been conducted regarding when it is best to implement a PA intervention during breast cancer treatment. ## 1.3. The Emergence of E-Bikes Electrically-assisted bicycles (e-bikes, also known as pedelecs) have been identified as a means through which to increase PA, providing at least moderate intensity PA [29]. Electrical assistance is administered only when the rider is pedaling via sensors and a motor [30]. With the additional electrical assistance, riders are motivated to cycle for longer periods of time and longer distances, thus increasing their overall PA engagement [31]. Specifically, survey results from over 10,000 individuals in seven European cities found that the average cycle duration was longer on an e-bike (35.0 min, $95\%$ CI 31.7 to 38.3) compared to a conventional bike (25.9, $95\%$ CI 25.4 to 26.5) [32]. In addition, individuals cycled further on an e-bike than on a conventional bike (9.4 km, $95\%$ CI 8.6 to 10.2 and 4.8 km, $95\%$ CI 4.7 to 4.9, respectively). For older adults, e-cycling is reported as being preferable over conventional cycling due to the reduced physical exertion required on an e-bike, leading to reduced feelings of fatigue following a cycling journey [33]. Use of an e-bike has been reported to increase the user’s PA. A cross-sectional study of 340 Norwegian residents reported that PA increased by 353.9 min per week due to e-bike use [34]. Associations have been identified between e-bike use and physiological parameters. A longitudinal study by Dons and colleagues [35] found that body mass index (BMI) was −0.010 kg/m2 ($95\%$ CI −0.020 to −0.0002) lower per additional day of e-cycling per month. Additionally, maximal power output increased for both untrained men and women (192.19 ± 28.7 watts and 145.9 ± 24.8 watts, respectively) following six weeks of active commuting on an e-bike [31]. Furthermore, among sedentary women, e-cycling has been associated with greater levels of enjoyment compared to conventional cycling, whilst still eliciting sufficient levels of energy expenditure and power output [36]. As such, e-cycling is proposed to be a successful method to increase PA amongst sedentary women. Based on these findings, it is possible that e-cycling could be an appropriate means through which to introduce breast cancer patients to PA while undergoing treatment. At present however, no research has examined the role of e-cycling amongst breast cancer patients. ## 1.4. Aims and Research Questions The aim of this research was to examine perceptions of e-cycling amongst individuals who have had a breast cancer diagnosis. Based on this aim, the following research questions were developed:What are the perceived barriers and facilitators to e-bike usage during breast cancer treatment?When, during cancer treatment, is perceived to be optimum for introducing e-cycling?Does the implementation of an e-bike taster session elicit changes in perceptions of e-cycling during breast cancer treatment? ## 2.1. Study Design and Protocol Due to the novel exploration of e-cycling during cancer treatment, a qualitative approach was identified as the most appropriate research method to identify initial perceptions and insights. Two one-to-one semi-structured interviews were conducted with each participant; one before an e-bike taster session (referred to as Interview One) to gather the participants’ initial thoughts of e-cycling, and one post the e-bike taster session (referred to as Interview Two) to identify whether e-cycling opinions had changed. Semi-structured, one-to-one interviews were used due to their ability to keep the interview focused whilst still providing flexibility for the researcher to explore pertinent ideas that may be raised during the conversation [37]. Interviews took place virtually, without the presence of non-participants. Interviews were recorded using the Zoom record option and transcribed verbatim using the Zoom auto-transcription option. Further editing of the transcriptions was required to ensure accuracy in the data. Transcripts were not returned to the participants for comment or correction as they were anonymized. Twenty-four interviews were conducted for Interview One. Following this, five participants were no longer contactable and did not complete the e-bike taster session or Interview Two. Four participants did not require a taster session due to already owning an e-bike. As such, fifteen participants completed the taster session and nineteen completed Interview Two. ## 2.2. Researcher Characteristics All data collection was conducted by a female researcher, KMW (MSc). KMW was completing their MSc in Nutrition, Physical Activity and Public Health at the University of Bristol at the time of the study. ## 2.3.1. Inclusion and Exclusion Criteria Participants were eligible to take part in the study if they have had a breast cancer diagnosis. Metastatic cancer is prominent amongst breast cancer patients, therefore individuals with secondary cancer were included [38]. Participants were excluded from the study if they did not consent to take part in an e-bike taster session, were unwilling to travel to Bristol for a taster session, or if they were unable to ride a conventional bicycle. For individuals who already owned an e-bike and were confident using it, it was not a requirement to attend the taster session. ## 2.3.2. Sampling A convenience sampling technique was used based on the aforementioned inclusion criteria. Participants were recruited via a Bristol-based charity, Penny Brohn UK. Advertisements were placed in two instalments of the monthly newsletters and emails were sent to a consenting mailing list. Additional social media adverts were promoted via Breast Cancer Now, on platforms including Facebook and Instagram. All advertisements contained the researcher’s contact details. The recruitment reach through Penny Brohn UK was 281 via email mailing list and 4800 via monthly newsletter. From this, 34 individuals expressed initial interest in the study (approximately $0.7\%$ of those reached), were sent the participant information sheet (int), and were presented with a link to an online demographic survey. The demographic survey collected data which allowed the researcher to screen individuals for eligibility; a consent form for the survey was also attached which participants were asked to sign and date electronically. In total, 29 out of 34 who were sent the PIS completed the demographic survey. Of these 29, one was not eligible for the study due to not having a breast cancer diagnosis, and four dropped out due to personal reasons unrelated to the study. Data collection and analysis were conducted in parallel, therefore participant recruitment continued until saturation was met [39]. Data saturation was met after 20 pre-taster interviews, as enough data were collected to generate relevant themes and codes that were repeated across participants and it was assumed no additional useful insights would be provided with additional participants. However, four extra participants were interviewed to account for drop-out rates and their data was included in analysis (Table 1). Twenty-four participants completed Interview One and nineteen completed Interview Two. ## 2.4. Interview Questions Separate interview guides were developed for use during Interview One and Two (Supplementary Materials File S1). The average duration of Interview One was 33 min, 45 s and 26 min, 55 s for Interview Two. No field notes were made during the interviews. Interviews were conducted by the primary researcher, KMW. Informal, ice-breaker questions commenced the interviews followed by open-ended questions with probing questions included ad hoc, depending on the participant’s response, with the intention to deepen or expand on a point if necessary [40]. Pilot interviews were conducted to ensure coherence of the interview guides [41]. Interview questions were focused on understanding how cancer treatment impacted individuals’ PA, the potential barriers and facilitators to e-cycling during this time, and individuals’ opinion of when is best to introduce e-cycling during cancer treatment. In addition, questions focused on how perceptions of e-cycling changed following the trialing of an e-bike. ## 2.5. Taster Session E-cycling is a recently popularized mode of PA, therefore it could not be assumed that participants had engaged in e-cycling previously [30]. To ensure responses to the interview questions were based on personal experience, participants were invited to take part in a free, one-hour e-cycling taster session following Interview One. Taster sessions were conducted by qualified instructors at Life Cycle, a Bristol-based charity. A full safety briefing was delivered and safety equipment, such as high-visibility jackets and helmets, were provided on request. Participants were given an introductory ride on the e-bike around a Bristol park with a cycle path, which included both flat and hilly routes to ensure maximum exploration of the e-bike. Participants were given the autonomy to switch between levels of assistance and were provided with advice and support on how best to use the e-bike. ## 2.6. Ethical Approval Ethical approval was obtained via the University of Bristol, School for Policy Studies Research Ethics Committee (Ethical Approval Number: EAN 055-21). ## 2.7. Data Analysis Descriptive statistics were extracted from the demographic survey and presented in Table 1 as means and standard deviation (SD); analyses were performed using IMB SPSS Statistics, Version 28. Interview recordings were held securely on the University of Bristol server until transcriptions were complete, after which they were deleted. All identifiable information was removed or anonymized during the transcription process and each participant was assigned an anonymous identification code. Thematic analysis was used and guided by Braun and Clarke’s [42] six steps of analysis. Analysis was conducted concurrently with data collection (as described above) to determine saturation and iteratively generate codes and themes. NVivo 12 (OSR International Pty Ltd., New Delhi, India, v12, 2018) was used to carry out data analysis. Transcripts were read repeatedly by KMW to ensure the researcher was familiar with the data. Following data familiarization, segments of text were highlighted, and initial codes were derived inductively and deductively (i.e., based on pre-specified research questions). KMW and a second coder (TJC) independently coded three transcripts. These three transcripts were selected by KMW to reflect diverse responses. The researchers met to discuss and refine the codes and a coding framework was developed. The remaining transcripts were coded by KMW, who revisited previously coded transcripts as required if new codes were identified. Codes were organized into categories based on their content through discussion between KMW and the second coder. From these categories sub-themes and higher order themes were generated. To adhere to the quality criteria for all qualitative research, credibility, transferability, dependability, and confirmability were all considered [43], therefore a COREQ checklist was followed extensively when producing this report (Supplementary Materials File S2). ## 3. Results Five overarching themes were generated from a combination of Interview One and Two: [1] Perceived role of e-bikes during treatment, [2] The relationship between e-bikes and fatigue, [3] Cancer-specific considerations, [4] Is e-cycling ‘enough’?, and [5] Optimizing the intervention (Figure 1). ## 3.1.1. Cycling Ability/Participation Participants explained that e-bikes could play a significant role in overcoming barriers to cycling that are present following a diagnosis. For some participants, despite having a history of regular cycling, they were no longer able to participate because of their treatment due to factors such as reduced energy, weakness, fatigue, and lacking motivation: Yet, the prospect of an e-bike was appealing, with the electric motor offering the required assistance to get participants back into their previous cycling habits: Participants discussed that an e-bike would facilitate an improvement in cycling performance, providing the ability to cycle further, faster, and with more confidence. Most prominently, the e-bike was perceived to overcome the inability to cycle up hills, due to strength reductions and deterioration in cardiovascular fitness that result from treatment: One participant compared up-hill e-cycling to feeling the same as cycling a conventional bike on a flat road, emphasizing that the electrical assistance allowed them to overcome their fear of hills: ## 3.1.2. Cycling with Friends and Partners Cycling was often described as a social sport, with participants enjoying a leisurely ride with their families, friends, or community cycling groups. However, this was often diminished once treatment started, having a negative impact on the participants, both socially and psychologically: Yet, the e-bike was positively described as a way to overcome this barrier, due to the electrical assistance allowing for increased duration, thereby boosting their confidence to keep up with their peers and enhancing enjoyment: ## 3.1.3. Positive Impact on Psychology There were several physical benefits of e-cycling described by participants both pre- and post-taster session (Table 2), such as improving fitness, strengthening muscles, and getting the blood pumping to improve the dispersion of treatment drugs; yet the most prominent benefits were described to be psychological. Specifically, the act of getting outside and into the fresh air was enough to promote e-cycling amongst participants: Many participants also described the benefit e-cycling will have to their treatment response. In particular, the e-bike was commonly associated with alleviating the stress of both a diagnosis and when undergoing treatment: For some, the taster session raised concerns that introducing e-cycling during treatment could lead to added stress, particularly amongst participants with limited cycling history and a fear of road cycling: Often, negative perceptions of e-bikes noted before the taster session were altered following a trial on the e-bike, with participants stating that cycling was “easier” than expected and a more “tolerable” mode of PA, thus making them more inclined to e-cycle in the future. ## 3.1.4. Loss of Identity Many participants reported a loss of identity or feeling like their cancer defines them. Yet, the prospect of an e-bike provided a sense of independence that they had been lacking since receiving their diagnosis: For some, their diagnosis meant they were no longer able to drive. As such, the e-bike would provide a mode of travel that is quicker and more practical than walking: ## 3.2.1. Fatigue as a Barrier to Cycling Table 2 highlights the key reported barriers and facilitators to e-cycling. The biggest barrier to conventional cycling during cancer treatment was said to be lacking energy and fatigue. This perception was therefore reflected onto e-cycling, whereby finding the energy to pedal was described as “unlikely” or “impossible” before trying an e-bike. This opinion largely diminished following the taster session: The fear of not getting home due to tiredness once setting off on a conventional bike ride was enough to deter people from cycling completely. However, the use of an e-bike was assumed to address this problem: Yet, anxieties were still raised post-taster session at the thought of the e-bike battery running low or stopping completely, thus leaving participants hesitant about cycling too far: ## 3.2.2. Conserving Energy Although some participants expressed a desire to e-cycle when fatigued, others stated it would be detrimental to their health if they exerted themselves too much: Emphasis was placed on the importance of conserving energy for treatments and therefore not participating in e-cycling on days of high fatigue: ## 3.3.1. Physical Impairments As a result of a lumpectomy or mastectomy, the restrictions in the arm and armpit were feared to limit the ability to hold the handlebars on the e-bike. However, the taster session appeared to eliminate this concern: For some, it was the strength in the arm that appeared most preventative: ## 3.3.2. E-Cycling Discomfort E-cycling was sometimes described as a “bumpy ride”, with potholes or uneven road surfaces often inducing high, uncomfortable impact onto the bike and the rider. Some participants expressed concerns regarding this impact when undergoing treatment: As a result, it was suggested that an individual’s stage of cancer should be considered when prescribing e-cycling as a mode of PA: Yet, participants recognized that the e-bike is no more prone to pummeling when compared to a conventional bicycle: ## 3.4.1. Manageability Following the taster session, a few participants reported that they expected the e-bike to be a strenuous form of PA that was potentially unmanageable, however, the taster session changed their perceptions: Others’ preconception was that e-cycling may not have provided any form of PA due to the assistance. However, following the taster session, e-cycling was perceived as being of sufficient intensity to provide a sufficient workout: Participants stated that the bike ride raised their heart rate sufficiently, worked their leg muscles, and induced muscle soreness the following day; all of which were claimed to be an indication of a good amount of physical exertion without the expense of fatigue: ## 3.4.2. Level of Assistance Multiple assistance levels allowed participants to alter their exertion depending on how they were feeling, which was reassuring when experiencing fluctuations in energy and motivation during treatment: Some participants were willing to cycle with little or no assistance to ensure they made the most of the workout: However, the weight of the e-bike was frequently presented as a barrier to e-cycling (Table 2), and therefore the reason many participants would always cycle with at least some electrical assistance. Some participants claimed the e-bike played a type of “psychological trick” on them and their willingness to cycle. Despite feeling tired and fatigued, participants suggested they would be more inclined to cycle an e-bike rather than a conventional bike, despite potentially exerting the same amount of energy: ## 3.5. Theme 5: Optimizing the Intervention The opinion of whether e-cycling was possible during specific treatments varied greatly between participants, with only one common conclusion: it is solely down to the individual: ## 3.5.1. Diagnosis Regardless of whether cycling throughout breast cancer treatment was seen as beneficial, most participants suggested not to promote an e-bike at the start of the cancer journey, following diagnosis: However, P.7 suggested using the e-bike as soon as possible following diagnosis to set the patient up best for the treatment that follows: ## 3.5.2. Surgery Regarding surgery, it was popular to assume that the e-bike would not be used until the individual had recovered physically from the surgery. The recovery time appeared to depend on the type of surgery administered: Typically, post-surgery was claimed to be an ideal time to present an e-bike. In particular, the e-bike was described to act as a distraction away from the disease and the upcoming treatments (depending on the individual’s order of treatments): ## 3.5.3. Chemotherapy E-cycling during chemotherapy presented the most controversial opinions. For some, despite chemotherapy being an exhausting time, they were confident they could continue using an e-bike depending on the week the chemotherapy was administered: However, the prospect of e-cycling at any stage during chemotherapy seemed impossible for some: ## 3.5.4. Radiotherapy The main concern for e-cycling during radiotherapy was due to time commitments, rather than the treatment itself. The time necessary to attend appointments every day appeared to frequently result in fatigue, which then acts as a barrier to using the e-bike. The other barrier to e-cycling during radiotherapy was the skin burns: ## 3.5.5. Recovery Presenting the e-bike post treatment as a method to aid recovery seemed optimal amongst patients: The e-bike was promoted as a method to return to “normal life”; a way of putting the cancer in the past. ## 4. Discussion This is the first qualitative study to explore perceptions of e-cycling amongst individuals who have had a breast cancer diagnosis. With the aid of a 1-h taster session, this research highlighted the intention and desire for PA engagement during breast cancer treatment, with the e-bikes providing a promising method to overcome perceived barriers to physical exertion. Participants frequently referred to the “ease” of e-cycling, and its ability to re-introduce previous cycling habits that were disenabled because of their diagnosis and treatment. Reservations and considerations were reported regarding specific treatment side effects and the appropriate timing to introduce an e-cycling intervention. The findings of this research can be used by healthcare professionals when prescribing sustainable and manageable physical activity interventions alongside treatment. The emergence of themes and their alignment with current literature are considered below to determine the relevance and significance of e-cycling during breast cancer treatment. ## 4.1. The Impact of Taster Sessions Table 2 shows an increase in the perceived barriers, facilitators, and benefits to e-cycling following the taster session. The importance of allowing participants to trial a piece of equipment to gather more detailed perceptions is demonstrated in research [44]. In the current study, it was common to have pre-conceived ideas about e-cycling, therefore the taster session was necessary to change attitudes and elicit more accurate perceptions of e-cycling in general, as well as its suitability during breast cancer treatment. Not only this, but trialing the equipment before use is important for user safety and confidence when initially using the e-bike. As highlighted in previous research, a pilot study which implemented a 12-week e-bike trial amongst breast cancer patients reported that, when asked if they were able to straightaway easily use the e-bike, participants answered 7 on a scale of 1 to 10 [45]. Although reasonably high, this demonstrates that e-bikes require practice, therefore the taster session is necessary to facilitate not only accurate perceptions of an e-bike, but also maximize the rider’s usability. During pre-taster session interviews, it was evident that most participants recognized an e-bike’s image and usability, though were yet to consider its benefits for use during breast cancer treatment. Nevertheless, the prospect of e-cycling was desirable in the present study, with many participants claiming the taster session solidified their expectations that e-cycling would be an easier alternative to conventional cycling, allowing for increased distance and speed due to the electric motor. This aligns with previous research findings amongst e-bike owners in the Sacramento, California area, whereby the notion of increased distance inspired their purchase, with many participants disposing of their previously used conventional bicycles [46]. It is possible that e-cycling is appealing for individuals with breast cancer as it allows them to engage in a longer duration of exercise, something they deemed themselves incapable of following their diagnosis. A lost sense of identity is frequently reported in the present study, therefore managing to achieve what was previously a mediocre task could become fulfilling for individuals, allowing them to reconnect with their pre-diagnosis identities. ## 4.2.1. Physical Benefits Participants expressed their perceived importance of PA during treatment through discussion of current exercise patterns. Following the e-cycling taster session, participants described the bodily fatigue and muscular soreness they experienced the next day, as well as the respiratory requirements of e-cycling. Previous research has confirmed the reduction in exertion from an e-bike, but also its suitability for individuals with physical limitations as a result [33]. Johnson and Rose [47] highlighted that $16.4\%$ of elderly e-bike owners purchased their e-bike due to injury, illness, or disability, placing emphasis on their ability to now return to their cycling habits after previously being forced to stop. Further research also supports that e-cycling reduces stress on the body in comparison to conventional cycling and therefore increases its accessibility to many riders, thus providing a suitable alternative during breast cancer treatment [33,48,49]. Furthermore, frequent reference was made to the ability to cycle up a hill because of the electric motor, something which previously deterred individuals from getting on a bike. This reinforces previous research that recognized hills as a key barrier to conventional cycling and a key facilitator of e-cycling [50,51]. A 2002 survey of 600 UK e-bike users reported $37\%$ of respondents stated the ease of use on hills as a main advantage of an e-bike over a conventional-bike [52]. It is possible that post-taster muscular soreness and difficulty with hill cycling reflects the fitness and deconditioning of the participants in this study, rather than the intensity of e-cycling itself. Many participants claimed they did not engage in cycling frequently prior to study engagement and would therefore not likely have the cycling-specific strength to cycle up hills with ease. In hilly terrains, it is important to consider the importance of hill assistance when expecting individuals undergoing breast cancer treatment to engage in cycling as a mode of PA. ## 4.2.2. Psychological Benefits Interestingly, participants discussed the desire to turn the electric motor off entirely when cycling on flat terrains. It was perceived that, during the taster sessions, participants felt the need to “push themselves, which consequently resulted in them working harder than anticipated due to the weight of the e-bike. This was described as “psychological trickery”, whereby the absence of the electric motor in fact promoted more intense PA and a greater sense of achievement. This is consistent with previous literature which explained that a perceived sense of achievement is a key facilitator to PA engagement during breast cancer treatment, in a study protocol where exercise instructors and continuous remarks of praise were used as a method to increase exercise self-efficacy [53]. With participants in the present study often expressing feelings of lacking control, it is possible they felt empowered when turning off the motor, demonstrating an ability to not only be in charge of the e-bike, but also their lives and the disease. Enhancing the psychological benefits is important for improving long-term adherence to e-cycling and PA overall [54]. Research explains that intrinsic motivation is a key facilitator of PA participation, particularly if the activity is self-directed (i.e., more than $50\%$ of the program is implemented without close supervision) [55], which an e-cycling intervention would be. As such, it is vital that individuals undergoing breast cancer treatment not only enjoy e-cycling, but find themselves motivated to take part. As described by the participants in this study, a heightened sense of achievement and “psychological trickery” is a promising reflection of sufficient intrinsic motivation. ## 4.3.1. Fatigue Most participants shared their relationship with fatigue during their treatment. Although fatigue was presented as a key barrier to conventional cycling, especially regarding hills and non-flat terrains, many suggested they would still engage in e-cycling when they would consider themselves too fatigued to engage in other, more strenuous types of PA. However, some days, such as during weeks of chemotherapy, e-cycling was deemed “impossible” or “unsuitable” due to the need to conserve energy. This supports previous research which identified a decrease in PA engagement when undergoing adjuvant chemotherapy, as well as significant decreases in energy expended through PA [56]. The greatest concern regarding fatigue was the fear of the e-bike battery running low and therefore cycling home with no assistance. Although advancements in technology mean that e-bike battery life is extending, this does not come without an increase in price [57]. Fears regarding battery life are common in research and, although not explicitly mentioned in the present study, there are concerns regarding remembering to charge the battery fully before use [46,58]. This is a consideration that must be accounted for when expecting breast cancer patients to cycle long distances as they already have an increased mental load due to their diagnosis [59]. ## 4.3.2. Accessibility Despite listing extensive benefits and facilitators to e-cycling, participants were quick to identify limitations and barriers to this mode of PA. As e-bikes are a large financial investment, with prices increasing depending on the make and model [60], participants were overly concerned about safely storing and locking an e-bike. Many participants reported they would not leave it unattended in public areas, even with the use of a secure lock. Previous research reported similar findings, whereby the cost is enough to repel potential buyers or delay their purchase [61]. Notably, Bristol bicycle theft rates were at an all-time high in 2020; therefore, theft concern in the present study may be biased towards its Bristol-based demographic [62]. Furthermore, the socioeconomic status of users will determine the prevalence of the theft barrier; with socioeconomic status not measured in this study, it is difficult to determine its effect on study findings. Prospectively, the initial investment for an e-bike would be eliminated if e-bikes were used within a future intervention amongst cancer patients, as the equipment costs would be covered by research funding. However, investment must be considered when looking to induce long-term e-cycling habits, when participation in research studies is complete. A survey study conducted in Norway found that higher perceived benefits and familiarity with e-bikes were positively related with the intention to buy an e-bike [63]. However, further research should be conducted within a UK sample, in addition to a demographic with a cancer diagnosis, to investigate whether individuals would be willing to invest in an e-bike during treatment. Particularly, there should be inclusion of a taster session, in line with Simsekoglu and Klöckner’s [63] findings that familiarity increases intentions to purchase an e-bike and will therefore promote long-term adherence. ## 4.3.3. Social Interaction A key reported facilitator and, consequently, benefit of e-cycling was the social opportunities that e-cycling provides. Frequently, individuals highlighted that an e-bike would allow them to cycle alongside peers and family, something they had not been capable of since their diagnosis. Rey-Barth and colleagues [45] found that group-based interventions involving e-bikes for breast cancer rehabilitation were beneficial for PA maintenance. The social aspect created environments for encouragement and increased enjoyment, with the homogenization of motor speed allowing for reduced extrinsic social comparisons. Interestingly, participants in the present study mentioned the appealing opportunity for group e-cycling, which aligns with Rey-Barth and colleagues’ [45] findings. It is possible that breast cancer patients experience social isolation following their diagnosis, which is particularly distressing when it results in exclusion from family and peer activities. The potential to re-engage in group cycling is likely ameliorating for participants, providing they eliminate any stigma or negative stereotypes that e-bikes often carry [64]. ## 4.4. Cancer Treatment Considerations Fears regarding physical limitations resulting from surgery, such as restricted use of the affected arm, are common in literature investigating the barriers to PA during breast cancer treatment. Participants in Sander and colleagues’ [65] study reported avoidance of resistance exercise, or modifications to workout routines, because of their joint pain and arm restrictions post surgery. Yet, this barrier was only expressed by a small number of participants in the present study, suggesting the demands of e-cycling may present less of a worry. When considering the optimal time throughout the cancer journey to introduce e-cycling, it was most frequently reported that immediately after diagnosis would be least convenient. Extensive research has supported the notion that a cancer diagnosis is a significant source of psychological stress and distress, with many patients struggling to cope with the initial shock of a diagnosis [66,67,68]. Specifically, research from Kang et al. [ 69] found that high cancer-specific stress was significantly correlated with high symptom perception amongst one hundred women with newly diagnosed breast cancer. When applying this finding to the present study, it could be argued that participants are less inclined to commence an e-cycling intervention due to higher perceived rates of symptoms, especially fatigue. Yet, research suggests that either continuing or commencing PA behavior immediately after diagnosis could be associated with up to $45\%$ lower risk of death in comparison to women who were inactive both before and after diagnosis, (HR 0.55; $95\%$ CI, 0.22 to 1.38) [70]. This suggests that the earlier the e-cycling intervention begins, the better. However, considerations regarding the participant’s preferences and willingness to adhere to exercise protocols if asked to engage when not ready must be recognized. Uncertainty was displayed when deciding if e-cycling would be manageable and maintainable during other breast cancer treatments, in particular the cyclic nature of chemotherapy doses. Parallel to previous research, exercise adherence was reported lower during chemotherapy in comparison to radiotherapy ($64\%$ vs. $25\%$, $$p \leq 0.02$$) in a study where 68 breast cancer patients undertook regular supervized, moderate-intensity, aerobic, and resistance exercise [71]. All individuals respond to treatments differently which is likely to explain the inconsistencies in opinions of when to commence e-cycling. Additionally, intra-personal perceptions will likely fluctuate depending on the individual’s emotional and physical response to treatment. An optimal time to intervene is, therefore, difficult to distinguish. ## 4.5. Strengths and Limitations To the author’s knowledge, this is the first qualitative study to explore the perceptions of e-cycling amongst individuals who have had a breast cancer diagnosis. The inclusion of e-bike taster sessions increased the validity of the participants’ perceptions, ensuring they were based on personal experiences rather than assumptions, and ensured that a change in perceptions pre- and post-taster session could be measured. The implementation of semi-structured interviews provided flexibility in the research questions, eliciting elaboration on points outside of the interview guide. Thematic analysis accurately followed Braun and Clarke’s [43] guidelines, and inductive reasoning was carried out to extensively analyze the data. Furthermore, the use of a second coder enhanced the credibility and confirmability of the research, and the quality of this review was maximized by following a COREQ checklist (Supplementary Material File S2). Nevertheless, the study’s limitations must be acknowledged. The expectation for participants to predict the feasibility and usability of an e-bike intervention during breast cancer treatment that they will not be taking part in could be deemed a difficult task, especially after receiving minimal information regarding the expected protocol. Furthermore, reporting and reflecting on previous feelings and emotions relies on accurate memory recall; importantly, the average time since diagnosis was more than a year (52.6 months; Table 1). Therefore, memories could be distorted, potentially impacting validity of research findings. The use of a single gatekeeper, Penny Brohn UK, to recruit participants had potential to implement bias towards a population that already engaged in healthy behaviors, thus limiting the study’s external validity. The requirement to travel to Bristol to attend the e-bike taster session likely limited the national reach of the study’s recruitment, therefore e-cycling perceptions could be biased towards the Bristol or Somerset landscape. As the participants’ residency location was not collected, it is difficult to measure the impact of this limitation on research findings. The lack of ethnic diversity in the study sample, as shown in Table 1, also limited the study’s generalizability to different cultures and may potentially contribute to ethnic disparities in healthcare research. Furthermore, the short timeframe of the study resulted in an absence of member-checking, a technique for exploring the credibility of qualitative results [72]. As a result, a lack of confirmability is acknowledged for the present study. ## 4.6. Implications for Practice and Research The findings suggest that e-cycling may be a suitable form of PA for individuals undergoing breast cancer treatment. Specifically, e-cycling is reported to elicit both physical and psychological benefits and is consistently reported as a manageable form of PA. This suggests that the promotion of e-cycling during cancer treatment is appropriate. Specifically, with the right funding and provision, e-cycling could be offered by doctors, oncologists, and nurses alongside traditional treatments (such as chemotherapy and radiotherapy) to reduce or eliminate negative treatment side effects (such as reduced quality of life, mental health decline, and reductions in physical health) [13] and recurrence rates [9]. The reported psychological boost that e-cycling elicits could help patients to process the negative emotional state that accompanies treatment, thus contributing to a better quality of life [73]. Future researchers can be confident that breast cancer patients support the prospect of e-cycling during treatment and would be keen to partake if the opportunity was presented to them when undergoing treatment. Researchers must recognize that introducing an e-bike immediately after a cancer diagnosis could be detrimental, yet an optimal time to intervene is highly dependent on the individual. Participants in this study valued the taster session to gain a better understanding of the e-bike usability and practicality. Therefore, the implementation of an e-bike taster session will be essential when looking to recruit participants for a future e-bike intervention. Barriers to e-cycling predominantly surrounded initial investment, therefore it is imperative to ensure external funding is secured to cover costs of the e-bike if recruiting for a future intervention. Additionally, knowledge of how to access an e-bike once study involvement has concluded would contribute to long-term e-cycling engagement amongst cancer patients. To address the limitations of the present study, future research should aim to gather perceptions from a more ethnically diverse population to reduce disparities in healthcare research. Additional further research should qualitatively assess the motivations, facilitators, and barriers behind buying an e-bike during a cancer diagnosis. Understanding these factors may assist with prolonging e-cycling engagement throughout cancer treatment and beyond. ## 5. Conclusions This qualitative study provides novel insights to the perceived usefulness and practicality of e-bikes during breast cancer treatment. It is presented that individuals who have had a breast cancer diagnosis recognize the importance of engaging in PA during treatment and believe e-cycling is a suitable mode to overcome common barriers to conventional cycling, such as fatigue and physical exertion. There are both physical benefits of e-cycling, such as its suitability for individuals with a cancer diagnosis, and psychological benefits, such as fresh air and the influence to work harder than anticipated. However, worries regarding battery life and theft were prominent, and the initial financial investment deterred some participants from potentially purchasing an e-bike. An optimal time to introduce an e-bike intervention was difficult to discern for participants, as cancer treatment is explained to be highly dependent on the individual. Overall, e-cycling was deemed an appropriate mode of PA to engage in during breast cancer treatment. It suggests promise for increasing PA behaviors in this population, with the potential to overcome many barriers posed by conventional cycling. The positive physical and psychological responses associated with providing e-bike taster sessions in this population may help to promote future engagement. ## References 1. Cancer 2. Cancer 3. Breast Cancer Incidence (Invasive) Statistics 4. Hortobagyi G.N., de la Garza Salazar J., Pritchard K., Amadori D., Haidinger R., Hudis C.A., Khaled H., Liu M.C., Martin M., Namer M.. The global breast cancer burden: Variations in epidemiology and survival. Clin. Breast Cancer (2005.0) 6 391-401. DOI: 10.3816/CBC.2005.n.043 5. Momenimovahed Z., Salehiniya H.. Epidemiological characteristics of and risk factors for breast cancer in the world. Breast Cancer Targets Ther. (2019.0) 11 151-164. DOI: 10.2147/BCTT.S176070 6. Martin A.M., Weber B.L.. Genetic and hormonal risk factors in breast cancer. J. Natl. Cancer Inst. (2000.0) 92 1126-1135. DOI: 10.1093/jnci/92.14.1126 7. Sun Y.S., Zhao Z., Yang Z.N., Xu F., Lu H.J., Zhu Z.Y., Shi W., Jiang J., Yao P.P., Zhu H.P.. Risk factors and preventions of breast cancer. Int. J. Biol. Sci. (2017.0) 13 1387-1397. DOI: 10.7150/ijbs.21635 8. Parsekar K., Howard Wilsher S., Sweeting A., Patel A., Fordham R.. Societal costs of chemotherapy in the UK: An incidence-based cost-of-illness model for early breast cancer. BMJ Open (2021.0) 11 e039412. DOI: 10.1136/bmjopen-2020-039412 9. Loprinzi P.D., Cardinal B.J., Winters-Stone K., Smit E., Loprinzi C.L.. Physical activity and the risk of breast cancer recurrence: A literature review. Oncol. Nurs. Forum (2012.0) 39 269-274. DOI: 10.1188/12.ONF.269-274 10. Spei M.E., Samoli E., Bravi F., La Vecchia C., Bamia C., Benetou V.. Physical activity in breast cancer survivors: A systematic review and meta-analysis on overall and breast cancer survival. Breast (2019.0) 44 144-152. DOI: 10.1016/j.breast.2019.02.001 11. UK Chief Medical Officers’ Physical Activity Guidelines 12. Cannioto R.A., Hutson A., Dighe S., McCann W., McCann S.E., Zirpoli G.R., Barlow W., Kelly K.M., DeNysschen C.A., Hershman D.L.. Physical activity before, during, and after chemotherapy for high-risk breast cancer: Relationships with survival. J. Natl. Cancer Inst. (2021.0) 113 54-63. DOI: 10.1093/jnci/djaa046 13. Loprinzi P.D., Cardinal B.J.. Effects of physical activity on common side effects of breast cancer treatment. Breast Cancer (2012.0) 19 4-10. DOI: 10.1007/s12282-011-0292-3 14. Gurgel AR B., Mingroni-Netto P., Farah J.C., de Brito CM M., Levin A.S., Brum P.C.. Determinants of health and physical activity levels among breast cancer survivors during the COVID-19 pandemic: A cross-sectional study. Front. Physiol. (2021.0) 12 624169. DOI: 10.3389/fphys.2021.624169 15. Pinto B.M., Trunzo J.J., Reiss P., Shiu S.Y.. Exercise participation after diagnosis of breast cancer: Trends and effects on mood and quality of life. Psycho-Oncology (2002.0) 11 389-400. DOI: 10.1002/pon.594 16. Wurz A., St-Aubin A., Brunet J.. Breast cancer survivors’ barriers and motives for participating in a group-based physical activity program offered in the community. Support. Cancer Ther. (2015.0) 23 2407-2416. DOI: 10.1007/s00520-014-2596-2 17. Beasley J.M., Kwan M.L., Chen W.Y., Weltzien E.K., Kroenke C.H., Lu W., Nechuta S.J., Cadmus-Bertram L., Patterson R.E., Sternfeld B.. Meeting the physical activity guidelines and survival after breast cancer: Findings from the after breast cancer pooling project. Breast Cancer Res. Treat. (2012.0) 131 637-643. DOI: 10.1007/s10549-011-1770-1 18. Mason C., Alfano C.M., Smith A.W., Wang C.Y., Neuhouser M.L., Duggan C., Bernstein L., Baumgartner K.B., Baumgartner R.N., Ballard-Barbash R.. Long-term physical activity trends in breast cancer survivors. Cancer Epidemiol. Biomark. Prev. (2013.0) 22 1153-1161. DOI: 10.1158/1055-9965.EPI-13-0141 19. Brunet J., Taran S., Burke S., Sabiston C.M.. A qualitative exploration of barriers and motivators to physical activity participation in women treated for breast cancer. Disabil. Rehabil. (2013.0) 35 2038-2045. DOI: 10.3109/09638288.2013.802378 20. Larsson I.L., Jönsson C., Olsson A.C., Gard G., Johansson K.. Women’s experience of physical activity following breast cancer treatment. Scand. J. Caring Sci. (2008.0) 22 422-429. DOI: 10.1111/j.1471-6712.2007.00546.x 21. Blackburn N.E., Mc Veigh J.G., Mc Caughan E., Wilson I.M.. The musculoskeletal consequences of breast reconstruction using the latissimus dorsi muscle for women following mastectomy for breast cancer: A critical review. Eur. J. Cancer Care (2018.0) 27 e12664. DOI: 10.1111/ecc.12664 22. Steindorf K., Schmidt M.E., Klassen O., Ulrich C.M., Oelmann J., Habermann N., Beckhove P., Owen R., Debus J., Wiskemann J.. Randomized, controlled trial of resistance training in breast cancer patients receiving adjuvant radiotherapy: Results on cancer-related fatigue and quality of life. Ann. Oncol. (2014.0) 25 2237-2243. DOI: 10.1093/annonc/mdu374 23. Wilson D.J.. Exercise for the patient after breast cancer surgery. Semin. Oncol. Nurs. (2017.0) 33 98-105. DOI: 10.1016/j.soncn.2016.11.010 24. Rogers L.Q., Markwell S.J., Verhulst S., McAuley E., Courneya K.S.. Rural breast cancer survivors: Exercise preferences and their determinants. Psycho-Oncology (2009.0) 18 412-421. DOI: 10.1002/pon.1497 25. Smith-Turchyn J., Richardson J., Tozer R., McNeely M., Thabane L.. Physical activity and breast cancer: A qualitative study on the barriers to and facilitators of exercise promotion from the perspective of health care professionals. Physiother. Can. (2016.0) 68 383-390. DOI: 10.3138/ptc.2015-84 26. Van Waart H., Stuiver M.M., Van Harten W.H., Geleijn E., Kieffer J.M., Buffart L.M., De Maaker-Berkhof M., Boven E., Schrama J., Geenen M.M.. Effect of low-intensity physical activity and moderate- to high-intensity physical exercise during adjuvant chemotherapy on physical fitness, fatigue, and chemotherapy completion rates: Results of the PACES randomized clinical trial. Clin. Oncol. (2015.0) 33 1918-1927. DOI: 10.1200/JCO.2014.59.1081 27. Emery C.A.. Injury prevention in paediatric sport-related injuries: A scientific approach. Br. J. Sport. Med. (2009.0) 44 64-69. DOI: 10.1136/bjsm.2009.068353 28. Cheville A.L., Tchou J.. Barriers to rehabilitation following surgery for primary breast cancer. J. Surg. Oncol. (2007.0) 95 409-418. DOI: 10.1002/jso.20782 29. Bourne J.E., Sauchelli S., Perry R., Page A., Leary S., England C.. Health benefits of electrically-assisted cycling: A systematic review. Int. J. Behav. Nutr. Phys. Act. (2018.0) 15 116. DOI: 10.1186/s12966-018-0751-8 30. Fishman E., Cherry C.. E-bikes in the mainstream: Reviewing a decade of research. Trans. Rev. (2016.0) 36 72-91. DOI: 10.1080/01441647.2015.1069907 31. de Geus B., Kempenaers F., Lataire P., Meeusen R.. Influence of electrically assisted cycling on physiological parameters in untrained subjects. Eur. J. Sport Sci. (2013.0) 13 290-294. DOI: 10.1080/17461391.2011.606845 32. Castro A., Gaupp-Berghausen M., Dons E., Standaert A., Laeremans M., Clark A., Anaya-Boig E., Cole-Hunter T., Avila-Palencia I., Rojas-Rueda D.. Physical activity of electric bicycle users compared to conventional bicycle users and non-cyclists: Insights based on health and transport data from an online survey in seven European cities. Transp. Res. Interdiscip. Perspect. (2019.0) 1 100017. DOI: 10.1016/j.trip.2019.100017 33. Jones T., Harms L., Heinen E.. Motives, perceptions and experiences of electric bicycle owners and implications for health, wellbeing and mobility. J. Transp. Geogr. (2016.0) 53 41-49. DOI: 10.1016/j.jtrangeo.2016.04.006 34. Sundfør T.M., Svendsen M., Tonstad S.. Effect of intermittent versus continuous energy restriction on weight loss, maintenance and cardiometabolic risk: A randomized 1-year trial. Nutr. Metab. Cardiovasc. Dis. (2018.0) 28 698-706. DOI: 10.1016/j.numecd.2018.03.009 35. Dons E., Rojas-Rueda D., Anaya-Boig E., Avila-Palencia I., Brand C., Cole-Hunter T., de Nazelle A., Eriksson U., Gaupp-Berghausen M., Gerike R.. Transport mode choice and body mass index: Cross-sectional and longitudinal evidence from a European-wide study. Environ. Int. (2018.0) 119 109-116. DOI: 10.1016/j.envint.2018.06.023 36. Sperlich B., Zinner C., Hebert-Losier K., Born D.P., Holmberg H.C.. Biomechanical, cardiorespiratory, metabolic and perceived responses to electrically assisted cycling. Eur. J. Appl. Physiol. (2012.0) 112 4015-4025. DOI: 10.1007/s00421-012-2382-0 37. Adeoye-Olatunde O.A., Olenik N.L.. Research and scholarly methods: Semi-structured interviews. J. Am. Coll. Clin. Pharm. (2021.0) 4 1358-1367. DOI: 10.1002/jac5.1441 38. Redig A.J., McAllister S.S.. Breast cancer as a systemic disease: A view of metastasis. J. Intern. Med. (2013.0) 274 112-126. DOI: 10.1111/joim.12084 39. Guest G., Namey E., Chen M.. A simple method to assess and report thematic saturation in qualitative research. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0232076 40. Price B.. Laddered questions and qualitative data research interviews. J. Adv. Nurs. (2002.0) 37 273-281. DOI: 10.1046/j.1365-2648.2002.02086.x 41. Majid MA A., Othman M., Mohamad S.F., Lim SA H., Yusof A.. Piloting for Interviews in Qualitative Research: Operationalization and Lessons Learnt. Int. J. Acad. Res. Bus. Soc. Sci. (2017.0) 7 1073-1080. DOI: 10.6007/IJARBSS/v7-i4/2916 42. Braun V., Clarke V.. Using Thematic Analysis in Psychology. Qual. Res. Psycol. (2006.0) 4 77-101. DOI: 10.1191/1478088706qp063oa 43. Guba E.G.. Criteria for assessing the trustworthiness of naturalistic inquiries. Educ. Commun. Technol. J. (1981.0) 29 75-91. DOI: 10.1007/BF02766777 44. Saballe C., Le H., Dirin A.. Experience changes the perception and feelings: A case study on mvr application in educational context. Edulearn Proc. (2018.0) 1 10204-10213. DOI: 10.21125/edulearn.2018.2481 45. Rey-Barth S., Pinsault N., Terrisse H., Eychenne C., Rolland C., Foote A., Guyot C., Bosson J.L.. A program centered on smart electrically assisted bicycle outings for rehabilitation after breast cancer: A pilot study. Med. Eng. Phys. (2022.0) 100 103758. DOI: 10.1016/j.medengphy.2022.103758 46. Popovich N., Gordon E., Shao Z., Xing Y., Wang Y., Handy S.. Experiences of electric bicycle users in the Sacramento, California Area. Travel Behav. Soc. (2014.0) 1 37-44. DOI: 10.1016/j.tbs.2013.10.006 47. Johnson M., Rose G.. Extending life on the bike: Electric bike use by older Australians. J. Transp. Health (2015.0) 2 276-283. DOI: 10.1016/j.jth.2015.03.001 48. Leger S.J., Dean J.L., Edge S., Casello J.M.. “If I had a regular bicycle, I wouldn’t be out riding anymore”: Perspectives on the potential of e-bikes to support active living and independent mobility among older adults in Waterloo, Canada. Transp. Res. A Policy Pract. (2019.0) 123 240-254. DOI: 10.1016/j.tra.2018.10.009 49. Marincek D., Rérat P.. From conventional to electrically-assisted cycling. A biographical approach to the adoption of the e-bike. Int. J. Sustain. Transp. (2021.0) 15 768-777. DOI: 10.1080/15568318.2020.1799119 50. Dill J., Rose G.. Electric bikes and transportation policy. Transp. Res. Rec. (2012.0) 2314 1-6. DOI: 10.3141/2314-01 51. Fernández-Heredia Á., Monzón A., Jara-Díaz S.. Understanding cyclists’ perceptions, keys for a successful bicycle promotion. Transp. Res. A Policy Pract. (2014.0) 63 1-11. DOI: 10.1016/j.tra.2014.02.013 52. Melia S., Bartle C.. Who uses e-bikes in the UK and why?. Int. J. Sustain. Transp. (2022.0) 16 965-977. DOI: 10.1080/15568318.2021.1956027 53. Browall M., Mijwel S., Rundqvist H., Wengström Y.. Physical activity during and after adjuvant treatment for breast cancer: An integrative review of women’s experiences. Integr. Cancer Ther. (2018.0) 17 16-30. DOI: 10.1177/1534735416683807 54. Pudkasam S., Feehan J., Talevski J., Vingrys K., Polman R., Chinlumprasert N., Stojanovska L., Apostolopoulos V.. Motivational strategies to improve adherence to physical activity in breast cancer survivors: A systematic review and meta-analysis. Maturitas (2021.0) 152 32-47. DOI: 10.1016/j.maturitas.2021.06.008 55. Byrne H., Caulfield B., De Vito G.. Effects of self-directed exercise programmes on individuals with type 2 diabetes mellitus: A systematic review evaluating their effect on HbA1c and other metabolic outcomes, physical characteristics, cardiorespiratory fitness and functional outcomes. Sports Med. (2017.0) 47 717-733. DOI: 10.1007/s40279-016-0593-y 56. Demark-Wahnefried W., Hars V., Conaway M.R., Havlin K., Rimer B.K., McElveen G., Winer E.P.. Reduced rates of metabolism and decreased physical activity in breast cancer patients receiving adjuvant chemotherapy. Am. J. Clin. Nutr. (1997.0) 65 1495-1501. DOI: 10.1093/ajcn/65.5.1495 57. Kiefer C., Behrendt F.. Smart e-bike monitoring system: Real-time open source and open hardware GPS assistance and sensor data for electrically-assisted bicycles. IET Intell. Transp. Syst. (2016.0) 10 79-88. DOI: 10.1049/iet-its.2014.0251 58. Rérat P.. The rise of the e-bike: Towards an extension of the practice of cycling?. Mobilities (2021.0) 16 423-439. DOI: 10.1080/17450101.2021.1897236 59. Jenkins V., Shilling V., Fallowfield L., Howell A., Hutton S.. Does hormone therapy for the treatment of breast cancer have a detrimental effect on memory and cognition? A pilot study. Psycho-Oncology (2004.0) 13 61-66. DOI: 10.1002/pon.709 60. Nocerino R., Colorni A., Lia F., Luè A.. E-bikes and e-scooters for smart logistics: Environmental and economic sustainability in pro-e-bike Italian pilots. Transp. Res. Proc. (2016.0) 14 2362-2371. DOI: 10.1016/j.trpro.2016.05.267 61. MacArthur J., Dill J., Person M.. Electric bikes in North America: Results of an online survey. Transp. Res. Rec. (2014.0) 2468 123-130. DOI: 10.3141/2468-14 62. Crime Levels Overview 63. Simsekoglu Ö., Klöckner C.. Factors related to the intention to buy an e-bike: A survey study from Norway. Transp. Res. F Traffic Psychol. Behav. (2019.0) 60 573-581. DOI: 10.1016/j.trf.2018.11.008 64. Ahlbom J., Andersson D.. E-Bike Users Are Lazy… and Healthy: A Study in Consumer Behaviour on the Symbolic Values of E-Bikes, Why Some Want E-Bikes and Others Avoid Them. (Dissertation) 65. Sander A.P., Wilson J., Izzo N., Mountford S.A., Hayes K.W.. Physical activity and exercise in survivors of breast cancer: A qualitative study. Phys. Ther. (2012.0) 92 525-536. DOI: 10.2522/ptj.20110115 66. Iwamitsu Y., Shimoda K., Abe H., Tani T., Okawa M., Buck R.. Anxiety, emotional suppression, and psychological distress before and after breast cancer diagnosis. Psychosomatics (2005.0) 46 19-24. DOI: 10.1176/appi.psy.46.1.19 67. Schumacher J.R., Palta M., Loconte N.K., Trentham-Dietz A., Witt W.P., Heidrich S.M., Smith M.A.. Characterizing the psychological distress response before and after a cancer diagnosis. J. Behav. Med. (2013.0) 36 591-600. DOI: 10.1007/s10865-012-9453-x 68. Schubart J.R., Emerich M., Farnan M., Stanley Smith J., Kauffman G.L., Kass R.B.. Screening for psychological distress in surgical breast cancer patients. Ann. Surg. Oncol. (2014.0) 21 3348-3353. DOI: 10.1245/s10434-014-3919-8 69. Kang D.-H., Park N.-J., McArdle T.. Cancer-Specific Stress and Mood Disturbance: Implications for Symptom Perception, Quality of Life, and Immune Response in Women Shortly after Diagnosis of Breast Cancer. ISRN Nurs. (2012.0) 2012 608039. DOI: 10.5402/2012/608039 70. Irwin M.L., Smith A.W., McTiernan A., Ballard-Barbash R., Cronin K., Gilliland F.D., Baumgartner R.N., Baumgartner K.B., Bernstein L.. Influence of pre- and postdiagnosis physical activity on mortality in breast cancer survivors: The health, eating, activity, and lifestyle study. J. Clin. Oncol. (2008.0) 26 3958-3964. DOI: 10.1200/JCO.2007.15.9822 71. Kirkham A.A., Bonsignore A., Bland K.A., McKenzie D.C., Gelmon K.A., Van Patten C.L., Campbell K.L.. Exercise prescription and adherence for breast cancer: One size does not FITT All. Med. Sci. Sports. Exerc. (2018.0) 50 177-186. DOI: 10.1249/MSS.0000000000001446 72. Birt L., Scott S., Cavers D., Campbell C., Walter F.. Member checking: A tool to enhance trustworthiness or merely a nod to validation?. Qual. Health Res. (2016.0) 26 1802-1811. DOI: 10.1177/1049732316654870 73. Taze S.S., Kanan N.. Experiences of women after breast cancer surgery. Florence Nightingale J. Nurs. (2020.0) 28 174-183. DOI: 10.5152/FNJN.2020.19012 74. Tong A., Sainsbury P., Craig J.. Consolidated criteria for reporting qualitative research (COREQ): A 32-item checklist for interviews and focus groups. Int. J. Qual. Health Care (2007.0) 19 349-357. DOI: 10.1093/intqhc/mzm042
--- title: 'Distress following the COVID-19 Pandemic among Schools’ Stakeholders: Psychosocial Aspects and Communication' authors: - Arielle Kaim - Shahar Lev-Ari - Bruria Adini journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049332 doi: 10.3390/ijerph20064837 license: CC BY 4.0 --- # Distress following the COVID-19 Pandemic among Schools’ Stakeholders: Psychosocial Aspects and Communication ## Abstract In response to the COVID-19 pandemic, many governments ordered school closures as a containment measure, with Israel being among over 100 countries to do so. This resulted in the abrupt shift to online and remote education for many students. Despite attempts to minimize the effects of disrupted education and create a dynamic virtual learning environment, the literature highlights various challenges including lack of communication with implications of distress faced by key stakeholders (students and their parents, teachers, and principals). In this cross-sectional study, we assess the perceived levels of communication and psychosocial aspects during both distance and frontal learning, as well as the long-term impacts (following over two and a half years of an ongoing pandemic) on distress among the key stakeholders of the Israeli education system— high school students, parents, teachers, and principals. The study findings demonstrate severe implications of distance learning on communication and psychosocial aspects, with lingering long-term impacts on distress, among all stakeholders (particularly among students). This reveals the need for tailored capacity building and resilience intervention programs to be integrated in the long-term response to the current ongoing pandemic to improve well-being and reduce distress among the various stakeholders, with particular attention to those that are most vulnerable and were hit the hardest. ## 1. Introduction As part of the coronavirus disease (COVID-19) pandemic containment measures, school closures were among the commonest governmental responses, with Israel joining over 100 countries in instituting this measure [1]. Governments had ordered institutions to cease face-to-face instruction for most of their students, requiring them to switch, almost overnight, to online teaching and remote education [2,3]. In the context of Israel, 15 March 2020, marked the official frontal closure of classrooms, where over 2.3 million pupils at all education levels were affected [4]. Since the initial closure of schools, the last (approaching) three years have seen a wax and wane trajectory in implemented measures as a response to the pandemic. Recurrent national or local lockdowns following additional waves of COVID-19 and the emergence of new and more infectious COVID-19 variants, resulted in additional school closures and distance learning. The estimated impact of these measures affects over 90 percent of the world’s school children (1.6 billion), according to UNESCO [2020]. By the end of 2022, the weeks of school closure in some countries had extended to over 80 weeks [5]. The World Bank estimated that a school shutdown of 5 months could be worth 10 trillion USD in terms of learning losses [6], where on average, learning losses amount to 0.17 of a standard deviation, equivalent to roughly a one-half year’s worth of learning [7]. Beyond the impact on pupil academics, the closure of schools and the movement to distance learning limited the opportunities for peer-to-peer and student–teacher socialization, which is typically enhanced by in-person interactions [8]. Intricately tied, as contributing factors, to a lack of socialization and loneliness, are poorer mental health and higher levels of distress for both children and adults alike [9,10,11,12,13]. Despite efforts to mitigate and address the impact of disrupted education and create an engaging learning environment through virtual platforms, the literature has discussed various challenges that were encountered, presenting the conclusion that not all needs were fully met through this modality [12,13,14,15]. The implementation of distance learning, however, had much broader consequences on society than just on pupils. Pivotal impacts were observed among school staff and administration (principals), teachers, and parents [16,17,18]. Teachers and administrators were unable to carry out their teaching and facilitation process as usual and had to adapt to an online setting [19]. Many previously had no pedagogical experience in online teaching, where modalities needed to be adapted to digital platforms. These changes resulted in teachers reporting significant increases in workload, being burdened by the work, and socially isolated from fellow colleagues and students, resulting in high levels of distress [20,21,22]. Similarly, principals of schools also reported significant increases in workload [23]. In the context of the family, schools often provide safeguarding and supervision of children, allowing parents to work. When schools are closed, parents in some cases had to remain at home, with inevitable economic consequences, or they had to leave children unsupervised, similarly with implications for elevated parental mental distress [24]. As a result of COVID-19, school closures have shifted education from the classroom to the home, where parents had to fill an educational role as well (at least partially) [25]. It has previously been widely discussed in the scientific literature that communication between various stakeholders of school systems promotes improved well-being for all [26,27,28,29]. The findings of Lv et al., { 2016} indicate that parent–teacher/ school communication plays an important moderating role in student emotional well-being and academic achievement, and Kuusimaki and Uusitalo-Malmivaara [2019] found that interactions between parents and teachers improve the well-being of teachers [26,27]. The global school closures due to COVID-19 have shown the fragility of education systems [30,31,32]. The uprooting and the reintegration of the global schooling system and its stakeholders because of the pandemic emphasize the need to ensure that school systems are resilient through all adversities. Functional resilience is defined as a system’s ability to resist, absorb, and respond to the shock of disturbances while maintaining its critical functions, and then recover to its original state or adapt to a new one [33,34]. Among the elements that functional resilience consists of are psychosocial aspects, communication between the varied stakeholders of the school system, digital literacy, pedagogic support, resources, and infrastructure (Kaim et al., under review). The ability to handle disruptions and maintain learning, teaching, and the support of the above activities is improved when all stakeholders are healthy (both physically and mentally). The capacities of education systems to respond to the crisis by delivering distance learning to support children, families, teachers, and administrators have been diverse and uneven [7]. The existing literature has predominantly focused on the adverse impacts of distance learning on the psychosocial aspects of all four key stakeholders, highlighting shared experiences of feelings such as isolation, disconnection, overwhelming, stress, and reduced motivation [8,9,10,11,12,13]. A lack of effective communication among these stakeholders has been a notable obstacle throughout the pandemic [12,13,14,15]. The long-term consequences of distance learning during the COVID-19 era remain largely uncertain, but there are indications that the negative effects may endure beyond the pandemic’s conclusion, including in the sphere of mental health [20,21,22,23,24,25]. The current study was targeted to cover this gap. The aim of this study is to assess the perceived levels of communication and psychosocial aspects during both distance and frontal learning, as well as the long-term impacts (following over two and a half years of an ongoing pandemic) on distress among the key stakeholders of the education system—high school students, parents, teachers, and principals. As such, the study strives to contribute in its practicality, with direct and tangible policy implications for where the need is most apparent for capacity building and implementation of interventions targeted to improve the functional resilience of schools. ## 2.1. Study Design Considering the importance of achieving an understanding of the long-term impact of COVID-19 on the stakeholders of the educational system, a cross-sectional study was conducted in October–November 2022, approximately two and half years after the initial closure of the frontal learning in the school system in Israel on 15 March 2022. A total sample of 1802 participants were recruited for this study, divided into the four key stakeholders of the education system: 10th–12th grade students ($$n = 1000$$), parents ($$n = 301$$), teachers ($$n = 449$$), and principals ($$n = 52$$) from 890 Israeli Jewish high schools. To partake in the study, the participants had to confirm their willingness to participate voluntarily in the study. The data was collected by the largest Israeli internet panel company which consists of over 140,000 panelists representing all demographic and geographic sectors (http://www.ipanel.co.il). This internet panel provides an online platform that adheres to the stringent standards of the European Society for Opinion and Marketing Research (ESOMAR). The data was collected anonymously, following approval of the Ethics Committee of the Tel Aviv University (number 0004549-1 from 13 February 2022) and the Ministry of Education (number 12379 from 28 April 2022). ## 2.2. The Study Tool The survey contained a brief introduction, which provided information on the background, objective, procedure, voluntary nature of participation, and declarations of anonymity and confidentiality. The four questionnaires were tailored to each of the stakeholders’ groups and consisted of six parts, based on items and indices that were developed specifically for this study, except for the perceived stress scale (PSS), which was based on a validated tool (PSS-4) [35] (See Appendix A). The number of items were not identical for the four tools, as to reflect the relevant relationships between stakeholders and the roles/functions that each stakeholder plays. The four questionnaires were validated by twenty content experts and pilot-tested on 25 individuals prior to their dissemination. The components of each of the questionnaires consisted of the following: Communication during distance learning, communication during frontal learning, psychosocial aspects during distance learning, psychosocial aspects in frontal learning, and PSS-4. ## 2.2.1. Communication during Distance Learning Communication during distance learning was measured by 3 items among students, 5 items among parents, and 7 items among teachers and principals. These components of the questionnaire encompassed attitudes of how well the communication was managed throughout the distance learning among all the stakeholders. Some of the questions were identical (though adapted to each specific population). The reliability of the scale was measured by Alpha Cronbach and results for each of the four stakeholders were (α = 0.833) among students, (α = 0.887) among parents (α = 0.861), among teachers and (α = 0.764), and among principals. ## 2.2.2. Communication during Frontal Learning Communication during frontal learning was measured by 1 item among students, 2 items among parents, and 4 items among teachers and principals. The components of this questionnaire encompassed attitudes of how well the communication was managed throughout the frontal learning among all the stakeholders, whereas all stakeholders were asked to what extent they agree or disagree with the following sentences relating to when the school teaching was conducted through frontal learning, on a 5-point Likert scale, ranging from 1 = Disagree to a very great extent, to 5 = Agree to a very great extent. The reliability of the scale was measured by Alpha Cronbach and results for parents were (α = 0.612), (α = 0.812) among teachers, and (α = 0.797) among principals. ## 2.2.3. Psychosocial Aspects during Distance Learning Psychosocial aspects during distance learning were measured by 3 items among students and parents, and 5 items among teachers and principals. The components of this questionnaire encompassed attitudes towards distance learning among all the stakeholders, whereas they were asked to what extent they agree or disagree with the following sentences relating to when the school teaching was conducted through distance learning, on a 5-point Likert scale, ranging from 1 = Disagree to a very great extent, to 5 =Agree to a very great extent. The reliability of the scale was measured by Alpha Cronbach and results for each of the four stakeholders were (α = 0.819) among students, (α = 0.612) among parents, (α = 0.746) among teachers, and (α = 0.615) among principals. ## 2.2.4. Psychosocial Aspects during Frontal Learning Psychosocial aspects during frontal learning were measured by 2 items among students and parents, and 5 items among teachers and principals. The components of this questionnaire encompassed characteristics of attitudes towards frontal learning among all the stakeholders where they were asked to what extent they agree or disagree with the following sentences relating to when the school teaching was conducted frontally, on a 5-point Likert scale, ranging from 1 = Disagree to a very great extent, to 5 =Agree to a very great extent. The reliability of the scale was measured by Alpha Cronbach as well as by Pearson correlation of items where only two items were used. The results for each of the four stakeholders are ($r = 0.339$) among students, ($r = 0.357$) among parents (α = 0.746), among teachers, and (α = 0.883) among principals. ## 2.2.5. Perceived Stress Scale (PSS-4) PSS-4 [35] was measured by four items among students, parents, teachers, and principals. The components of this questionnaire assess how often certain feelings and thoughts relate to the respondent in the past month for each stakeholder, on a 5-point Likert scale, ranging from 1 = never, to 5 =very often. The reliability of the scale was measured by Alpha Cronbach and results for each of the four stakeholders were (α = 0.668) among students, (α = 0.726) among parents, (α = 0.655) among teachers, and (α = 0.683) among principals. ## 2.2.6. Demographics Demographics for students were assessed by 11 items including gender, year of birth, place of residence, number of children living in the same home under the age of 18, number of dependents over the age of 18 living with you in the same home, religion, degree of religiosity, school type, school location, grade, and class size. For parents, teachers, and principals, demographics assessed included 12 items including gender, year of birth, place of residence, marital status, number of children under the age of 18 living with you at home, number of dependent adults living at home, education level, religion, level of religiosity, income level, school type, and location of the school. ## 2.3. Statistical Analysis Descriptive statistics were used to describe the participants’ demographic characteristics (frequency, mean, and standard deviation) of all four stakeholders. In addition, descriptive statistics were used to describe the characteristics of the schools sampled (%), and to determine the spread tendency and central tendency in the five indexes. A one-way ANOVA test was used to assess variability between stakeholders. A post hoc test (Bonferroni) was further conducted to enable the identification of the differences among the varied groups. Pearson correlation tests were conducted to analyze the associations between all variable indexes. All statistical analyses were performed using SPSS software version 28. p-values lower than 0.05 were considered to be statistically significant. ## 3.1. Respondents and School Sample Characteristics The sample of the study included 1802 participants, including 1000 students, 301 parents, 449 teachers, and 52 principals. Table 1 presents the demographic characteristics of all four stakeholders of the surveyed population. The average age of the students sampled was 16.7, with the majority ($52.5\%$) being male. The average age of parents sampled was 48.0, with the majority being female ($68.4\%$). The average age of teachers sampled was 41.6, with the majority being female ($80.8\%$). Lastly, the average age of principals was 46.7, with the majority being female ($55.8\%$). Among students, parents, and principals, the majority ($54.4\%$, $59.1\%$, and $53.8\%$, respectively) considered themselves secular. Among teachers, the largest mass of the sampled population ($43.0\%$) similarly considered themselves secular. The majority of parents, teachers, and principals ($85.4\%$, $71.7\%$, and $88.6\%$, respectively) were in a relationship, with children. In addition, the majority of parents, teachers, and principals had received a bachelor’s degree and above. A total of 890 Jewish schools within Israel were sampled, where 67.1 % were state schools, and $32.9\%$ were religious. The largest number of schools sampled geographically were from the central region of Israel ($29.6\%$). Table 2 presents the characteristics of the schools sampled in this study. ## 3.2. Mean Levels of Communication, Psychosocial Aspects and PSS The mean level for the communication index during distance learning among all the stakeholders was found to be 3.51 ± 0.88, as compared to 3.91 ± 0.87 during frontal learning. With regard to psychosocial aspects indices, during distance learning the mean level was 2.92 ± 1.04 as compared to frontal learning—3.72 ± 0.92. Furthermore, the mean PSS score among all participants was 3.46 ± 0.72. ( See Table 3). ## 3.3. Correlation between Variables A positive, significant correlation was found between communication during distance learning and communication during frontal learning ($r = 0.297$), PSS ($r = 0.264$), psychosocial aspects during distance learning ($r = 0.28$), as well as psychosocial aspects during frontal learning ($r = 0.211$). Communication during frontal learning was found to be positively and significantly correlated with PSS ($r = 0.285$) and psychosocial aspects during frontal learning ($r = 0.47$). Lastly, PSS was found to be positively and significantly correlated with psychosocial aspects during both distance ($r = 0.106$) and frontal learning ($r = 0.344$). ( See Table 4). ## 3.4. Differences by Stakeholder and Distance Versus Frontal Learning Differences by stakeholders (students, parents, teachers, and principals) with respect to communication levels both during distance learning (DL) and frontal learning (FL) are displayed in Figure 1. During DL, students on average, display the lowest levels of the communication index (3.3), as compared to the other stakeholders, while parents display the highest levels (4.06). With regard to communication during FL, parents display the lowest mean levels of the index, while teachers indicate the highest levels. With regard to communication DL, significant differences are observed between students and parents ($p \leq 0.001$), students and teachers ($p \leq 0.001$), parents and principals ($$p \leq 0.001$$), and parents and teachers ($p \leq 0.001$). With respect to communication FL, significant differences are observed between students and teachers ($p \leq 0.001$), parents and principals ($p \leq 0.01$), and parents and teachers ($p \leq 0.001$). The results also suggest that each stakeholder during distance learning perceived the communication to be less effective as compared to frontal learning (students—3.3 DL versus 3.83 FL; teachers—3.61 DL versus 4.18 FL; principals—3.57 DL versus 4.15 FL), with the exception of parents (4.06 DL versus 3.7 FL). The differences were all significant according to the repeated measures test ($F = 72.54$, $p \leq 0.001$). Differences by stakeholders (students, parents, teachers, and principals) with respect to psychosocial aspect levels, during both distance (DL) and frontal learning (FL) are displayed in Figure 2. During DL, parents on average display the lowest levels of the psychosocial aspects index (2.57 ± 1.04), as compared to the other stakeholders (students—2.87 ± 1.10, teachers—3.26 ± 0.83, principals—3.06 ± 0.64), with teachers displaying the highest levels. With regard to the psychosocial aspects during FL, students display the lowest mean levels of the index (3.41 ± 0.85), while principals indicate the highest levels (4.45 ± 0.1.14). With regard to psychosocial aspects during DL, significant differences are observed between students and parents ($p \leq 0.001$), students and teachers ($p \leq 0.001$), parents and teachers ($p \leq 0.001$), and principals and parents ($p \leq 0.01$). With respect to psychosocial aspects during FL, significant differences are observed between students and parents ($p \leq 0.001$), students and teachers ($p \leq 0.001$), students and principals ($p \leq 0.001$), parents and teachers ($p \leq 0.001$), parents and principals ($p \leq 0.01$), and principals and teachers ($p \leq 0.01$). The results also suggest that each stakeholder during distance learning perceived psychosocial aspects to be less effective as compared to frontal learning (students—2.87 DL versus 3.41 FL, parents—2.57 versus 3.64, teachers—3.26 DL versus 4.03 FL, principals—3.06 DL versus 4.45 FL). The differences were all significant according to the repeated measures test ($F = 29.91$, $p \leq 0.001$). Lastly, the mean PSS-4 scores were 3.35 ± 0.72 for students, 3.65 ± 0.68 for parents, 3.57 ± 0.70 among teachers, and 3.50 ± 0.73 among principals. Significant differences were observed between students and parents ($p \leq 0.001$), as well as between students and teachers ($p \leq 0.001$). The lower the score, the higher the level of perceived stress, whereas the highest level of stress is exhibited among students (See Figure 3). ## 4. Discussion The transition to distance learning for education systems posed a major challenge in the wake of the COVID-19 pandemic. With schools forced to close their doors, education leaders, teachers, students, and parents had to quickly pivot to a remote learning model in order to maintain educational continuity. The shift highlighted the need for functional resilience in the education system, specifically the system’s ability to adapt to unexpected disruptions and ensure the delivery of quality education. Through the prioritization of functional resilience, education systems and their stakeholders can not only overcome current adversities but be better prepared for future disruptions. Several of the key aspects identified as crucial to the functional resilience of a system are communication between relevant stakeholders and attitudes towards psychosocial aspects among stakeholders. The findings of this investigation demonstrate several interesting phenomena with regard to these two elements, as well as the long-term implications of the interaction between these aspects. First, students perceived levels of communication most critically during distance learning as compared to the other stakeholders, with parents expressing the highest levels. In contrast, when referring to perceived communication levels during frontal learning, parents indicate the lowest levels across all the stakeholders. This finding may be a reflection of the high home-based involvement of parents in supporting their children’s learning throughout the school closures, as compared to routine patterns of interaction in the home. The literature similarly suggests that parent involvement in children’s learning increased throughout distance learning, where the home-learning context improved relationships with teachers, with sentiments reciprocated by teachers [36,37]. The more critical attitudes of students toward distance learning match findings in the literature, where students have expressed decreased communication between them and their instructors and their increased feelings of isolation [38]. Teachers and principals report higher levels of communication during in-person learning, consistent with the challenges that education encountered while adapting to remote classes, with poor communication and low interaction as key hurdles faced [39,40]. Interestingly, psychosocial attitudes during distance learning were perceived lowest among parents regarding their children, indicating that parents’ perceptions of their children’s attitudes towards the learning experience during distance learning are worse than the children’s own perception. Parents consistently have been found in the context of health-related quality of life to rate their child’s health as worse than the child’s reporting, related to unresolved concerns regarding the influence of parental distress [41,42,43]. Furthermore, teachers expressed the best psychosocial attitudes toward distance learning (feeling more motivated and less burdened by the transition), as compared to the other stakeholders. Despite this, across all four stakeholders, the implications of distance learning on motivation, and attitudes towards the transition’s burdensome nature are clear, supporting previous findings [44,45,46]. The most pertinent implications of this study indicate the long-lasting impacts of distance learning, as reflected in the findings regarding the PSS-4 index, as the measurement assesses distress feelings and thoughts in the past month. The findings suggest that beyond the distance and frontal learning experiences, students exhibit the highest persisting levels of distress, potentially an indication of being hit hardest by frontal school disruptions, among the four stakeholder groups. Despite the fact that no previous assessment of PSS-4 was found in the literature on adolescents to assess differences with pre-pandemic PSS-4 scores, previous studies have discussed long-term lingering (extension into long-term even after the end of the pandemic) distress impacts among adults [47,48,49]. In the context of the current study, parents display among all the stakeholders, the lowest levels of distress according to the PSS-4 measurement, potentially indicating that this sample of the population, may be less vulnerable after their return to a work–balance routine and is thus able to better cope and bounce back following return to normalcy. Furthermore, a positive relationship between higher perceived levels of communication during distance learning, and higher perceived levels of psychosocial aspects was found in this study. In addition, a higher PSS index (in the context of this study, a lower level of stress), is correlated with higher perceived communication and psychosocial levels. The relationship between perceived levels of communication, psychosocial aspects, and perceived stress were found to be positively correlated in this study. Communication has previously been established as a protective factor from distress among school children, and improved well-being across the various stakeholders [26,29,50]. Interventions previously examined among high school students offer various solutions, including the integration of meditation apps which have been shown to improve perceived stress among adolescents, as well inquiry-based stress reduction among teachers [16,51]. Furthermore, the integration of tailored interventions in the context and in the midst of future scenarios among stakeholders should be considered in order to prevent lingering long-term repercussions and foster lasting resilience. At a broader level of functional resilience, the measures adapted by schools during the pandemic include the establishment and implementation of virtual learning infrastructures, provision of technical, pedagogic and psychosocial support, as well as provision of required resources [52]. The current study has shed light on the important role of psychosocial aspects and communication in the functional resilience of the education system. By examining the attitudes of four key stakeholders towards distance learning, the study has uncovered significant relationships between these attitudes and long-term distress. Additionally, this research work has pinpointed possible vulnerabilities that exist within the educational system, and where special dedicated attention may be warranted. To the best of our knowledge, few studies have been carried out to evaluate the impacts of COVID-19 distance learning on four key stakeholders, especially investigating the relationship between attitudes towards communications levels, psychosocial aspects, and long-term distress. ## Limitations The study has several limitations. First, the study utilized an online panel to collect responses. While this methodology ensured a rapid turnover of information and provided a large sample of the Israeli Jewish population, the study conclusions are limited to individuals who have access to a source of internet and high computing skills. It should also be noted that though a very large sample of schools were included in the study, there is variability among schools concerning their technological, psychosocial and educational resources. Moreover, the study’s cross-sectional design limits its findings on attitudes towards distance learning, as they were not assessed during a school closure, but rather after. Furthermore, the findings of this study must be carefully considered, as generalizability and transferability to other societies may be limited; to support or refute the findings of this study, it is recommended that similar studies be carried out in additional schools in varied countries. ## 5. Conclusions Our findings have valuable implications as we show the relationship between attitudes toward communications levels, psychosocial aspects, and long-term distress among four key stakeholders of the education system. The findings of the study reveal the need for tailored intervention programs to be integrated in the long-term response to the current ongoing pandemic to improve the well-being and reduce distress among the various stakeholders, with particular attention to those that are most vulnerable and were hit the hardest, (in the context of the current findings, the students). Future research should explore the longitudinal impacts of various interventions on changing long-term distress patterns among the various stakeholder participants. It is also recommended to expand the study beyond the Jewish schools context to include, for example, Arab or ultra-religious schools (in the Israeli context), as well as beyond Israeli borders. ## References 1. Paltiel O., Hochner H., Chinitz D., Clarfield A.M., Gileles-Hillel A., Lahad A., Manor O., Nir-Paz R., Paltiel A., Stein-Zamir C.. Academic activism on behalf of children during the COVID-19 pandemic in Israel; beyond public health advocacy. Isr. J. Health Policy Res. (2021) 10 48. DOI: 10.1186/s13584-021-00485-7 2. Stein-Zamir C., Abramson N., Shoob H., Libal E., Bitan M., Cardash T., Cayam R., Miskin I.. A large COVID-19 outbreak in a high school 10 days after schools’ reopening, Israel, May 2020. Eurosurveillance (2020) 25 2001352. DOI: 10.2807/1560-7917.ES.2020.25.29.2001352 3. Donitsa-Schmidt S., Ramot R.. Opportunities and challenges: Teacher education in Israel in the Covid-19 pandemic. J. Educ. Teach. (2020) 46 586-595. DOI: 10.1080/02607476.2020.1799708 4. UN Secretary-General Warns of Education Catastrophe, Pointing to UNESCO Estimate of 24 Million Learners at Risk of Dropping Out. (2020) 5. Buchholz K.. The Longest School Closures of the Pandemic (2022) 6. Psacharopoulos G., Collis V., Patrinos H.A., Vegas E.. The COVID-19 Cost of School Closures in Earnings and Income across the World. Comp. Educ. Rev. (2021) 65 271-287. DOI: 10.1086/713540 7. Patrinos H.A., Vegas E., Carter-Rau R.. An Analysis of COVID-19 Student Learning Loss (2022). DOI: 10.1596/1813-9450-10033 8. Alsubaie M.A.. Distance education and the social literacy of elementary school students during the COVID-19 pandemic. Heliyon (2022) 8 e09811. DOI: 10.1016/j.heliyon.2022.e09811 9. Fegert J.M., Vitiello B., Plener P.L., Clemens V.. Challenges and burden of the Coronavirus 2019 (COVID-19) pandemic for child and adolescent mental health: A narrative review to highlight clinical and research needs in the acute phase and the long return to normality. Child Adolesc. Psychiatry Ment. Health (2020) 14 20. DOI: 10.1186/s13034-020-00329-3 10. Ravens-Sieberer U., Kaman A., Otto C., Adedeji A., Devine J., Erhart M., Napp A.K., Becker M., Blanck-Stellmacher U., Löffler C.. Mental health and quality of life in children and adolescents during the COVID-19 pandemic—Results of the COPSY study. Dtsch. Ärzteblatt Int. (2020) 117 828 11. Lee C.M., Cadigan J.M., Rhew I.C.. Increases in loneliness among young adults during the COVID-19 pandemic and association with increases in mental health problems. J. Adolesc. Health (2020) 67 714-717. DOI: 10.1016/j.jadohealth.2020.08.009 12. Dabrowski A., Mitchell P.. Effects of remote learning on mental health and socialisation. Lit. Rev. (2022). DOI: 10.37517/978-1-74286-682-6 13. Barrot J.S., Llenares I.I., Del Rosario L.S.. Students’ online learning challenges during the pandemic and how they cope with them: The case of the Philippines. Educ. Inf. Technol. (2021) 26 7321-7338. DOI: 10.1007/s10639-021-10589-x 14. Singh S., Roy D., Sinha K., Parveen S., Sharma G., Joshi G.. Impact of COVID-19 and lockdown on mental health of children and adolescents: A narrative review with recommendations. Psychiatry Res. (2020) 293 113429. DOI: 10.1016/j.psychres.2020.113429 15. Constantini K., Markus I., Epel N., Jakobovich R., Gepner Y., Lev-Ari S.. Continued participation of Israeli adolescents in online sports programs during the COVID-19 pandemic is associated with higher resilience. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18084386 16. Zadok-Gurman T., Jakobovich R., Dvash E., Zafrani K., Rolnik B., Ganz A., Lev-Ari S.. Effect of inquiry-based stress reduction (IBSR) intervention on well-being, resilience and burnout of teachers during the COVID-19 pandemic. Int. J. Environ. Res. Public Health (2021) 18. DOI: 10.3390/ijerph18073689 17. Grooms A.A., Childs J.. “We need to do better by kids”: Changing routines in US schools in response to COVID-19 school closures. J. Educ. Stud. Placed Risk (JESPAR) (2021) 26 135-156. DOI: 10.1080/10824669.2021.1906251 18. Lugo D.. The Effects of Virtual Learning during the COVID-19 Pandemic on Teachers and Students in Grades K-12 within the United States: A Systematic Review. Ph.D. Thesis (2021) 19. Pollock K.. School leaders’ work during the COVID-19 pandemic: A two-pronged approach. Int. Stud. Educ. Adm. (2020) 48 38 20. Flack C.B., Walker L., Bickerstaff A., Earle H., Margetts C.. Educator Perspectives on the Impact of COVID-19 on Teaching and Learning in Australia and New Zealand (2020) 21. Kaden U.. COVID-19 school closure-related changes to the professional life of a K–12 teacher. Educ. Sci. (2020) 10. DOI: 10.3390/educsci10060165 22. Stapleton P., Garby S., Sabot D.. Psychological distress and coping styles in teachers: A preliminary study. Aust. J. Educ. (2020) 64 127-146. DOI: 10.1177/0004944120908960 23. Flack C.B., Walker L., Bickerstaff A., Earle H., Johnson C.L.. Principal Perspectives on the Impact of COVID-19: Pathways Toward Equity in Australian Schools (2021) 24. Davis C.R., Grooms J., Ortega A., Rubalcaba J.A.-A., Vargas E.. Distance learning and parental mental health during COVID-19. Educ. Res. (2021) 50 61-64. DOI: 10.3102/0013189X20978806 25. Doyle O.. COVID-19: Exacerbating educational inequalities. Public Policy (2020) 9 1-10 26. Lv B., Zhou H., Guo X., Liu C., Liu Z., Luo L.. The Relationship between Academic Achievement and the Emotional Well-Being of Elementary School Children in China: The Moderating Role of Parent-School Communication. Front. Psychol. (2016) 7 948. DOI: 10.3389/fpsyg.2016.00948 27. Kuusimäki A.-M., Uusitalo-Malmivaara L., Tirri K.. The Role of Digital School-Home Communication in Teacher Well-Being. Front. Psychol. (2019) 10 2257. DOI: 10.3389/fpsyg.2019.02257 28. Wang M.T., Sheikh-Khalil S.. Does parental involvement matter for student achievement and mental health in high school?. Child Dev. (2014) 85 610-625. DOI: 10.1111/cdev.12153 29. Al-Hail M., Al-Fagih L., Koç M.. Partnering for Sustainability: Parent-Teacher-School (PTS) Interactions in the Qatar Education System. Sustainability (2021) 13. DOI: 10.3390/su13126639 30. Leo S., Alsharari N.M., Abbas J., Alshurideh M.T.. From Offline to Online Learning: A Qualitative Study of Challenges and Opportunities as a Response to the COVID-19 Pandemic in the UAE Higher Education Context. The Effect of Coronavirus Disease (COVID-19) on Business Intelligence (2021) 203-217 31. Manivannan M., Jogalekar M.P., Kavitha M.S., Maran B.A.V., Gangadaran P.. A mini-review on the effects of COVID-19 on younger individuals. Exp. Biol. Med. (2021) 246 293-297. DOI: 10.1177/1535370220975118 32. Krishnamoorthy R., Keating K.. Education crisis, workforce preparedness, and COVID-19: Reflections and recommendations. Am. J. Econ. Sociol. (2021) 80 253-274. DOI: 10.1111/ajes.12376 33. Biggs R., Schlüter M., Biggs D., Bohensky E.L., BurnSilver S., Cundill G., Dakos V., Daw T.M., Evans L.S., Kotschy K.. Toward Principles for Enhancing the Resilience of Ecosystem Services. Annu. Rev. Environ. Resour. (2012) 37 421-448. DOI: 10.1146/annurev-environ-051211-123836 34. Zhong S., Clark M., Hou X.Y., Zang Y., FitzGerald G.. Development of key indicators of hospital resilience: A modified Delphi study. J. Health Serv. Res. Policy (2015) 20 74-82. DOI: 10.1177/1355819614561537 35. Warttig S.L., Forshaw M.J., South J., White A.K.. New, normative, English-sample data for the Short Form Perceived Stress Scale (PSS-4). J. Health Psychol. (2013) 18 1617-1628. DOI: 10.1177/1359105313508346 36. Bubb S., Jones M.A.. Learning from the COVID-19 home-schooling experience: Listening to pupils, parents/carers and teachers. Improv. Sch. (2020) 23 209-222. DOI: 10.1177/1365480220958797 37. Dong C., Cao S., Li H.. Young children’s online learning during COVID-19 pandemic: Chinese parents’ beliefs and attitudes. Child. Youth Serv. Rev. (2020) 118 105440. DOI: 10.1016/j.childyouth.2020.105440 38. Alawamleh M., Al-Twait L.M., Al-Saht G.R.. The effect of online learning on communication between instructors and students during Covid-19 pandemic. Asian Educ. Dev. Stud. (2020) 11 380-400. DOI: 10.1108/AEDS-06-2020-0131 39. Orhan G., Beyhan Ö.. Teachers’ perceptions and teaching experiences on distance education through synchronous video conferencing during Covid-19 pandemic. Soc. Sci. Educ. Res. Rev. (2020) 7 8-44 40. Pressley T., Ha C., Learn E.. Teacher stress and anxiety during COVID-19: An empirical study. Sch. Psychol. (2021) 36 367. DOI: 10.1037/spq0000468 41. Varni J.W., Limbers C.A., Burwinkle T.M.. How young can children reliably and validly self-report their health-related quality of life?: An analysis of 8,591 children across age subgroups with the PedsQL™ 4.0 Generic Core Scales. Health Qual. Life Outcomes (2007) 5 1. DOI: 10.1186/1477-7525-5-1 42. Eiser C., Morse R.. Can parents rate their child’s health-related quality of life? Results of a systematic review. Qual. Life Res. (2001) 10 347-357. DOI: 10.1023/A:1012253723272 43. Berg-Nielsen T.S., Vika A., Dahl A.A.. When adolescents disagree with their mothers: CBCL-YSR discrepancies related to maternal depression and adolescent self-esteem. Child Care Health Dev. (2003) 29 207-213. DOI: 10.1046/j.1365-2214.2003.00332.x 44. Duraku Z.H., Hoxha N., Duraku Z.H.. The impact of COVID-19, school closure, and social isolation on gifted students’ wellbeing and attitudes toward remote (online) learning. Impact of the COVID-19 Pandemic on Education and Wellbeing (2020) 130-169 45. Rasmitadila R., Aliyyah R.R., Rachmadtullah R., Samsudin A., Syaodih E., Nurtanto M., Tambunan A.R.S.. The Perceptions of Primary School Teachers of Online Learning during the COVID-19 Pandemic Period: A Case Study in Indonesia. J. Ethn. Cult. Stud. (2020) 7 90-109. DOI: 10.29333/ejecs/388 46. Lase D., Zega T.G.C., Daeli D.O., Zaluchu S.E.. Parents’ perceptions of distance learning during COVID-19 in rural Indonesia. J. Educ. Learn. (EduLearn) (2022) 16 103-113. DOI: 10.11591/edulearn.v16i1.20122 47. Borgatti A.C., Crockett K.B., Jacob A.E., Davis A.L., Dutton G.R.. Correlates of psychological distress among adults with obesity during the COVID-19 pandemic. Psychol. Health (2022) 37 1547-1564. DOI: 10.1080/08870446.2022.2038790 48. Dhaliwal A.. The Lingering Effects: Examining mental health in India during a period of COVID-19 regression. Asia Pac. J. Health Manag. (2022) 17 1-3. DOI: 10.24083/apjhm.v17i2.821 49. Ramírez Vera Y.E., Navarro Chaya H.Y., Ortiz Cáceres D.C.. Effects of COVID-19 on women’s mental health: A systematic review. Gac. Médica Caracas (2022) 130. DOI: 10.47307/GMC.2022.130.s3.21 50. Tang S., Xiang M., Cheung T., Xiang Y.-T.. Mental health and its correlates among children and adolescents during COVID-19 school closure: The importance of parent-child discussion. J. Affect. Disord. (2021) 279 353-360. DOI: 10.1016/j.jad.2020.10.016 51. Kaskons E.. Simple but not Easy: An Examination of the Test-Retest Reliability of the Perceived Stress Scale (PSS-10) and Exploration of a Meditation App for Stress Management with High School Students. Ph.D. Thesis (2022) 52. Zengulaaru J., Nyamekye E.. Ensuring Future Resilience beyond ICT and Online Teaching and Learning of Social Studies in Ghanaian Senior High Schools: Lessons from COVID-19 Pandemic. Soc. Educ. Res. (2022) 14 23-44. DOI: 10.37256/ser.3320221442
--- title: 'The Impact of a Health Empowerment Program on Self-Care Enablement and Mental Health among Low-Income Families: Evidence from a 5 Year Cohort Study in Hong Kong' authors: - Fangcao Lu - Carlos King Ho Wong - Emily Tsui Yee Tse - Amy Pui Pui Ng - Lanlan Li - Joyce Sau Mei Lam - Laura Bedford - Daniel Yee Tak Fong - Patrick Ip - Cindy Lo Kuen Lam journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049337 doi: 10.3390/ijerph20065168 license: CC BY 4.0 --- # The Impact of a Health Empowerment Program on Self-Care Enablement and Mental Health among Low-Income Families: Evidence from a 5 Year Cohort Study in Hong Kong ## Abstract Health empowerment can be an effective way to reduce health inequities. This prospective cohort study evaluated the 5 year impact of a health empowerment program (HEP) on health outcomes among adults from low-income families. The Patient Enablement Instrument version 2 (PEI-2), Depression, Anxiety and Stress Scale 21 (DASS-21), and 12 item Short-Form Health Survey version 2 (SF-12v2) were administered at baseline and follow-up for both intervention and comparison groups. A total of 289 participants ($$n = 162$$ for intervention group, $$n = 127$$ for comparison group) were included in the analysis. Most of the participants were female ($72.32\%$), and aged from 26 to 66 years old ($M = 41.63$, SD = 6.91). Linear regressions weighted by inverse probability weighting using the propensity score showed that, after follow-up of 5 years, the intervention group demonstrated significantly greater increases in all items and total scores for the PEI-2 (all B > 0.59, $p \leq 0.001$), greater decreases in the DASS depression score (B = −1.98 $$p \leq 0.001$$), and greater increases in the Mental Component Summary score of the SF-12v2 ($B = 2.99$, $$p \leq 0.027$$) than the comparison group. The HEP may be an effective intervention enabling adults from low-income families to manage their health-related issues and improve their mental health, as evidenced by our study. ## 1. Introduction Poverty is a global problem linked to poor health outcomes [1]. In addition to difficulties in accessing healthy food, clean water, and safe shelter, limited healthcare recourses are also a common problem for people from low-income families [2]. Indeed, a study found that people who received low incomes and/or lived in poverty had poorer health than the age–sex-matched individuals from the general population [3]. Moreover, healthcare expenses can further divert already limited disposable income from the educational, social, and other needs of families, hindering children’s development and resulting in trans-generational poverty [4,5,6]. Thus, the close link between poverty and poor health forms a vicious cycle [7]. There have been calls for the development of effective interventions to improve health among people from low-income families and break the cycle of poverty and poorer health [1]. Primary healthcare plays a key role in improving public health and reducing health inequities [8,9]. Starting from self-care, primary healthcare involves health promotion, disease prevention, and management of health conditions [10]. The central component of primary healthcare is health empowerment, a process through which people are motivated to take greater control over their lives and health-related decisions [11]. The concept of health empowerment entails working in partnership with individuals to enhance health literacy, desire, self-assurance, ways of action, and utilization of external resources in order to maintain good health, practice self-care, and increase appropriate use of healthcare services [12]. There is a growing body of literature on the benefits of health empowerment (HE) interventions. First, HE helps improve health-related abilities, such as health literacy [13] and utilization of health services [14]. Second, it can stimulate the adoption of healthy habits, such as increasing physical activity and reducing sedentary behavior [15]. Third, HE can modify health-related attitudes, including enhancing self-efficacy in self-care [16], self-determination [17], and self-efficacy in physical activity [18]. Last, HE has been found to improve physical and psychological health among a range of participant groups, including people with diabetes mellitus [12], adolescents [17], and homebound older adults [19]. The effectiveness of HE has been documented in both high-income (the U.S. [18,19], Italy [20], South Korea [15], Sweden [21], Taiwan [14,16]) and low-income regions (Thailand [13], India [17]). Nevertheless, evidence on the effectiveness of HE for people from relatively low-income families in high-income regions is limited. Most studies targeting this disadvantaged group were conducted in U.S., showing effectiveness in reducing depression in both adults [22,23,24] and children [25,26] while increasing perceived quality of life and positive affect [27]. The HE prevented hospitalizations [28] and pediatric emergency room and clinic visits [29]. HE can also promote healthier lifestyles, such as through assisting in smoking cessation [30], promoting healthy eating, increasing physical activity [31,32], assisting in the adoption of general environmental health precautions, increasing self-efficacy [33], improving problem-solving abilities [34], and improving intellectual academic achievements [35], among members of low-income families. The HE interventions reported in the literature tend to be highly controlled, unidimensional strategies delivered exclusively to either parents or children in low-income families using short-term outcomes [22,27,28,34], which have limitations in terms of generalizability and sustainability. Building on the existing evidence, we designed a long-term, complex, community-based health empowerment program (HEP) with intercalated components comprising annual health assessments, health talks, self-care enablement courses, and health ambassador training, which were available to both parents and children in their natural environment on a voluntary basis. We believed that such an HE intervention would be more feasible and sustainable for self-care enablement and health. We hope this will stimulate a new direction in HE interventions that may eventually lead to specific changes in health policy and services, reducing the health inequity among people from low-income families. Hong Kong has undergone rapid economic development since the late 20th century, becoming a high-income region [36]. However, it has one of the highest Gini coefficients in the world (0.54) and wide income inequalities, such that the top $25\%$ of families earn at least double the population median household income [37]. Based on the local definition of poverty (i.e., <$50\%$ of population median income), 1.65 million people, which equals over one fifth of the Hong Kong population, live in poverty [38]. They are eligible for limited financial subsidies (e.g., up to 9488 HKD/month for a family of three in 2012) [39]. Families with monthly household income between $50\%$ and $75\%$ of the population median do not receive much government assistance (e.g., up to 1515 HKD/month for a family of three in 2012) [39]. Tung Chung is a developing district on an outlying island of Hong Kong where around $40\%$ of residents live in poverty [40]. There was only one public primary care clinic in Tung Chung [41] serving 78,000 local residents [42] when this study began in 2012. In 2013, there was a public regional hospital established in Tung Chung. However, it only has primary care, psychiatric, emergency, internal medicine, and allied health services. The public healthcare services in Hong Kong do not provide regular health assessments, and residents have to self-finance for this in the private section. However, Hong Kong tops the world in terms of costs of living, including medical costs [43], and the combination of low-income and limited public healthcare services puts the health of Tung Chung residents at risk. In 2012, the Trekkers Family Enhancement Scheme (TFES) was initiated by a local philanthropic group, the Kerry Group Kuok Foundation (Hong Kong) Limited (KGKF). The TFES offer supports related to health, education, employment, and environmental harmony to 200 low-income families in Tung Chung. A health empowerment program (HEP) with intercalated annual health assessments, health literacy and self-care enablement courses, and health ambassador training was delivered regularly to support the health of the TFES families. This study aimed to evaluate the 5 year effectiveness of the HEP for the health of low-income families. We examined whether the HEP was associated with greater health enablement, better mental health, and higher health-related quality of life over a 5 year follow-up. ## 2.1. Study Subjects and Data Collection Procedure The present study was a prospective, comparative cohort study. It involved two groups of low-income families with young children studying in grades 1–3 (aged 7 to 11). All TFES families were invited to enroll in the HEP (herein referred to as “intervention families”), and low-income families who had not participated in the TFES were also recruited in Tung Chung and Kwai Chung as the comparison group. Kwai Chung is also a developing satellite residential district with similar sociodemographics and public healthcare facilities as Tung Chung [44]. Families were recruited between July 2012 and September 2015 if they satisfied all the inclusion criteria: [1] there was at least one family member working (full-time or part-time); [2] there was at least one dependent child in the family who studied in grades 1–3; [3] the monthly income of the family did not exceed $75\%$ of the Hong Kong population median household income; and [4] written consent was provided. Participants in both groups completed a comprehensive health assessment and a telephone questionnaire survey at baseline upon enrollment and at around 5 years after the baseline assessment. All adults and children aged 7–11 years old at the initiation of the study from each family were included in the study. There were 369 adults in total invited from July 2012 to September 2015 ($$n = 191$$ for the intervention group, $$n = 178$$ for the comparison group), and 357 adults ($$n = 190$$ for the intervention group, $$n = 167$$ for the comparison group) provided consent and completed the baseline assessment. There were 68 participants who did not complete the follow-up assessment at 5 years, representing a drop-out rate of $19\%$ ($$n = 28$$ for the intervention group, $$n = 40$$ for the comparison group). Overall, $78\%$ of participants ($$n = 289$$, $$n = 162$$ for the intervention group, $$n = 127$$ for the comparison group) completed the follow-up assessment. These participants were included in the analysis. The mean and median durations of these participants’ follow-ups were both five years. Most of participants were females ($72.32\%$), and their averaged age was 41.63 years old (SD = 6.91). The subject recruitment and follow-up flowchart is presented in Figure 1. ## 2.2. Study Intervention—The HEP The HEP consisted of intercalated annual health assessments, health talks, self-care enablement courses, and health ambassador training. The health assessment program included an annual telephone health and health service use survey, clinical health assessments, and a health hotline. Based on the telephone surveys and clinical health assessments, those with significant health risks or abnormalities were counseled by a nurse or doctor from the project team or referred to appropriate services for further management. Regular health talks and seminars targeted common problems identified in the health assessments, which included healthy eating, weight management, the health benefits of exercise, liver diseases, nutrition, stress management, psychosomatic illnesses in children, and child development. Self-care enablement courses included stress management, nutrition, dancing and exercise training, and hiking. The courses contained multiple sessions and emphasized participants’ participation. The health talks and enablement courses were all delivered by specialists. The details can be found in Appendix A and Appendix B. During the nutrition and exercise training courses, we encouraged some participants to become the group leaders of the classes and coordinate group practices after class. This group of adults became health ambassadors of their families and peers. ## 2.3. Outcome Measures and Study Instruments The primary outcome was self-care enablement as assessed by the Chinese version of the Patient Enablement Instrument version 2 (PEI-2), which has been found to be valid and reliable among the local Chinese people [45]. It includes six items on perceived abilities to cope with life, understand and manage illness, maintain health, and help oneself. An example item is, “In the past four weeks, how much have you felt able to cope with life”. Responses to all the items of the PEI were based on a five-point Likert scale, in which 1 meant “not at all” and 5 meant “extremely well”. The scores for each item of the PEI were summated to form the total score, which ranged from 6 to 30. A higher score indicated greater enablement. The secondary outcomes included negative emotional states and health-related quality of life (HRQOL). We used the Chinese version of the Depression, Anxiety and Stress Scale 21 (DASS-21) to capture participants’ negative emotional states. This measure has shown good reliability in the Chinese setting [46]. DASS-21 includes three seven-item subscales regarding several negative emotional states, including depression, anxiety, and stress. The responses to all the items were provided on a four-point Likert scale, in which 0 meant “did not apply” and 3 meant “very much or most of the time”. An example item is, “I found it hard to wind down”. We first added the scores of the seven items in a subscale and then multiplied the sum by 2 so that they could be compared to the DASS normative data and to other publications on DASS [47]. Each transformed subscale score ranged from 0 to 42. Higher scores indicated greater emotional disturbance. We utilized the Chinese 12 item Short-Form Health Survey Version 2 (SF-12v2) to measure HRQOL. The Chinese SF-12v2 has shown good validity and reliability among Chinese populations [48,49]. The measure consists of 12 items covering eight domains: physical functioning (PF), general health (GH), bodily pain (BP), physical role-functioning (RP), emotional role-functioning (RE), social functioning (SF), vitality (VT), and mental health (MH). An example item is, “During the past 4 weeks, how much did pain interfere with your normal work (including both work outside the home and housework)?”. These eight domains were weighted into two summary scores: a physical component summary (PCS) score and a mental component summary (MCS) score. The scores for each domain ranged from 0 to 100. The PCS and MCS scores for the SF-12v2 were norm-based. Specifically, the population mean for the two scores was 50 and the standard deviation for the two was 10. A higher score indicated better HRQOL. Socioeconomic status, general state of health, and physician factors were found to be associated with patient enablement [50]. Given that, confounding factors (covariates), including age, gender, highest education level obtained, household income (monthly), working status, marital status, smoking status, alcohol consumption, obesity status, chronic morbidity, reception of government assistance, and use of a regular family doctor, were collected with a structured questionnaire. The Chinese PEI-2, DASS-21, SF-12v2, and the covariates questionnaire were administered by a trained interviewer in person or by telephone. All outcome and covariate data were self-reported. ## 2.4. Data Analysis Data analyses were based on the complete-case analysis and only participants without missing values were included. We used STATA version 16.0 (StataCorp LP, College Station, TX, USA) to conduct all the statistical analyses. The presented tests of significance were two-tailed, with p values lower than 0.05 indicating statistical significance. Descriptive statistics were used to present participants’ baseline characteristics. Inverse probability weighting based on propensity scores was used to account for residual confounding bias and minimize differences in the characteristics of the two groups. We first used a logistic regression model to calculate each participant’s propensity score, with adjustment for the aforementioned baseline covariates. The balance of baseline covariates between the two groups before and after the inverse probability weighting was assessed according to the p value, with $p \leq 0.05$ indicating an optimal balance between two groups. Cronbach’s α coefficient was measured to test the reliability for internal consistency for each measure, with values larger than 0.7 representing good reliability. Linear regressions weighted by inverse probability weighting using propensity scores were applied to identify the independent effects of the HEP on the participants’ changes in PEI-2, DASS-21, and SF-12v2 scores from baseline to the 5 year follow-up. We first conducted preliminary tests to check the assumptions of the multiple linear regression (e.g., normality, homoscedasticity, and multicollinearity). We checked that there were no cases of outliers (i.e., Cook’s distance < 1 [51]). Although the assumption of normality was violated, it did not substantively affect the results because of the large sample size (i.e., the number of observations per variable was greater than 10 [52]) in the present study. For each model, the R2 and F-test of overall significance are reported; the unstandardized coefficients (B), $95\%$ confidence level, and p-value are reported to indicate the effects of the intervention on each dependent outcome variable; furthermore, a power analysis for a two-sample means test was applied to calculate the post hoc power. To assess the robustness of the results, we conducted a sensitivity analysis of the multiple linear regressions, adjusting for baseline covariates, without inverse probability weighting. ## 2.5. Ethical Approval The current study received ethical approval (UW 12–517) from the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster. ## 3. Results Table 1 presents the baseline characteristics of the subjects by group before and after weighting. Before weighting, the marital and working statuses of the two groups were significantly different. Nevertheless, all baseline characteristics were balanced after propensity score weighting. In the weighted sample, most subjects were between 35 and 50 years old ($72.79\%$), and they were predominantly women ($71.26\%$). The measures showed good reliability with the data from the current study, including the PEI-2 (Cronbach’s α = 0.90), DASS-21 (Cronbach’s α = 0.84 for depression, Cronbach’s α = 0.79 for anxiety, Cronbach’s α = 0.85 for stress), and SF-12v2 (Cronbach’s α = 0.73 for PCS, Cronbach’s α = 0.81 for MCS). The PEI-2, DASS-21, and SF-12v2 scores at the baseline and follow-up assessments before and after weighting are presented in Table 2 for both the intervention and comparison groups. Table 3 presents the results for the regression of the HEP’s impact on the 5 year changes in the PEI-2, DASS-12, and SF-12v2 scores after propensity score weighting. The HEP intervention group showed significantly greater increases in all the items and the total score for the PEI-2 (B ranged between 0.59 and 5.22, all $p \leq 0.001$). Additionally, the HEP was significantly associated with a greater decrease in the DASS depression score (B = −1.98, $$p \leq 0.001$$). Nevertheless, the differences in the changes in the DASS anxiety and stress scores between the two groups were not significant. There was a positive association between the HEP and increases in MCS scores ($B = 2.99$, $$p \leq 0.027$$), but there was no difference in the changes in PCS scores between groups. The results consistently showed significantly greater increases in all the items and the total score for the PEI-2 in the HEP group than in the comparison group in the sensitivity analysis using multiple linear regression to adjust for baseline covariates (Appendix C). ## 4. Discussion As far as we know, the current longitudinal study is the first to examine the long-term health effects of a complex HE intervention in low-income families. The results with inverse probability weighting showed that participants in the HEP demonstrated greater increases in self-care enablement (PEI-2 scores) and mental HRQOL (SF-12v2 MCS score) and greater decreases in depressive symptoms (DASS depression score) than those without involvement in the HEP intervention. Previous studies have demonstrated the effectiveness of various HEPs in improving participants’ health-related outcomes, but most of them were short-term studies lasting for only a few months with follow ups of less than one year [53]. In particular, studies have shown that involvement in a short-term HEP can modify health-related attitudes [13,14,15,16]. For example, an HEP with a combination of teaching sessions, discussions, role playing, field tours, and so forth enhanced self-efficacy in self-care for 94 community-dwelling older adults in Taiwan after 12 weeks [16]. A HEP with 5 month problem-based learning sessions increased the perceived health control and sense of capacity to take health action among 63 immigrant women in Taiwan [14]. Furthermore, HEPs can improve psychological health [12,19]. For instance, by delivering a series of education courses over four months, an HEP increased psychological well-being and health-related quality of life for 54 Italians with diabetes [54]. An HEP with six weekly home visits by a trained nurse was found to improve psychological well-being for 32 homebound older adults in the United States [17]. As indicated above, although most HEPs only intervene for a few months and focus on small-scale individual programs, the effectiveness of these HEPs has been demonstrated across various populations, including for different ages [15,16,17,19], genders [14], ethnic groups [18], immigration statuses [14], health statuses [13,15,55], and regions [16,56]. However, research on the long-term effectiveness of HEPs and whether a family-based complex intervention is feasible and applicable in low-income families is limited. The findings of the present study complement those of previous studies in affirming the long-term effectiveness of HEPs in improving participants’ self-care enablement and mental health. One explanation regarding the mechanism of how HEPs increase enablement and mental health is that regular health assessments engage participants and raise their health awareness [57]; the health talks, self-care enablement courses, and advice on appropriate management empower them to take control of their health and cope with health problems [58]. This sense of control has the potential to inhibit the triggering of negative emotions and help with their management [17], therefore increasing the MCS score and decreasing the DASS depression score. The results indicate the feasibility and applicability of HEPs with interrelated components for the improvement of health care enablement and self-reported mental health among parents in low-income families. In particular, Hong Kong is one of the regions that has the highest Gini coefficients and significant income inequity [37]. The effectiveness of HEPs in Hong Kong targeting low-income families suggests the possibility of using HEPs to reduce health equity. Additionally, previous studies have indicated that HEPs can stimulate the adoption of healthy habits. For example, a Korean study showed that utilizing 8 week lifestyle improvement education, group discussions, and exercise training enabled 27 hypertensive older adults to increase physical activity and reduce sedentary behavior [15]. In contrast, we did not find any significant association between the HEP and increases in physical HRQOL (SF-12v2 PCS score). In general, we observed a decrease in the PCS scores in both groups over the 5 year follow up, which coincided with the outbreak of the COVID-19 pandemic. During the peak of the COVID-19 outbreak, the Hong Kong government released mandates to close indoor sports facilities and limit outdoor activities for over five months in 2020 [59]. The outbreak of the COVID-19 pandemic and limited physical activities could have reduced physical fitness, leading to worsening of physical functioning and general health, which are the major determinants of the SF-12v2 PCS [60]. This suggests that the failure of the current HEP to improve SF-12v2 PCS may have been related to the outbreak of the COVID-19 pandemic and its influence on individuals’ physical activity and physical health. Additionally, during the pandemic, the formats of the intervention components in the current HEP changed because the face-to-face intervention was not feasible. We therefore delivered the intervention via real-time video meetings, distribution of videos, a health app, telephone consultations, and so forth. While the hybrid mode of the HEP was successfully in improving self-reported self-enablement and mental health, the nonsignificant association between the HEP and the SF-12v2 PCS score suggested that the effect of the hybrid mode on physical health may have been limited. The comparison between the current HEP and other HE interventions involving low-income families [24,25,26,27,28,29,30,31] showed two main differences in terms of intervention design. First, our HEP included four interrelated components that covered various aspects of health by providing health knowledge, encouraging regular exercise, promoting family activities, facilitating regular health assessments and cues to take health action, increasing social interaction, and offering health consultations. Second, the complex intervention involved whole families, including children and parents. The association between family members’ health was considered in the design of the intervention components to maximize the effectiveness of the HEP. The involvement of the whole family and various components likely increased participation and participants were seldom lost during follow up, which also indicate the success of the present intervention. The findings of the current intervention contribute to the current health empowerment literature in three key ways. First, health empowerment among low-income families is feasible and has sustainable effectiveness. Second, both adults and children should be involved in the regular health assessment and self-care enablement activities in light of the close links among family members’ health and the mutual influence on health-seeking behavior. Third, multi-dimensional but intercalated components can be utilized to enhance not only health promotion knowledge and practice but also accessible professional support to solve health problems. The various components provided insights into the design of future complex health interventions in real-world settings. There are four potential limitations worthy of further discussion. First, participants in the HEP were those who volunteered to join and may therefore have had a prior interest in receiving health information and resources. Additionally, the intervention was limited to people who were currently residing in one district in Hong Kong, so the sample cannot be viewed as representative. Non-randomization and district limitations may affect the generalizability of the findings and their external validity. Second, since it was not a randomized control trial, the observed and unobserved confounding variables could not be fully accounted for in the analysis. However, our main analysis adjusted for residual confounding by applying inverse probability weighting based on propensity score. Third, we used face-to-face or telephone interviews to collect data. Furthermore, data regarding the variables of interest were based on participants’ self-reports. The different data collection methods and the self-reported data could have had the potential to cause measurement errors, but this was equally applicable to both the intervention and comparison groups. Finally, the long recruitment period and the loss of participants during follow up could have led to bias in the results. Nevertheless, the drop-out rates in this cohort study in both groups were lower than the upper limit of the acceptable rate (i.e., $50\%$) [61], which supported the reliability and validity of the results. ## 5. Conclusions This study shows the effectiveness of a longitudinal HEP in enabling self-care and improving mental health among adults from low-income families. It offers insights into the feasibility and applicability of HEPs in real-world community settings. Given the increasingly wide income inequality and the close link between poverty and poor health, similar HEPs may help to break the vicious cycle. This opens a new research agenda regarding how HEP care models can be more widely implemented to enhance health equity. ## References 1. Marmot M.. The health gap: The challenge of an unequal world. Lancet (2015) 386 2442-2444. DOI: 10.1016/S0140-6736(15)00150-6 2. Alley D.E., Soldo B.J., Pagán J.A., McCabe J., Deblois M., Field S.H., Asch D.A., Cannuscio C.. Material Resources and Population Health: Disadvantages in Health Care, Housing, and Food Among Adults Over 50 Years of Age. Am. J. Public Health (2009) 99 S693-S701. DOI: 10.2105/AJPH.2009.161877 3. Lam C.L.K., Guo V.Y., Wong C.K.H., Yu E.Y.T., Fung C.S.C.. Poverty and health-related quality of life of people living in Hong Kong: Comparison of individuals from low-income families and the general population. J. Public Health (2016) 39 258-265. DOI: 10.1093/pubmed/fdw046 4. Spencer N.J.. Poverty and Child Health (2018) 5. Engle P.L., Black M.M.. The Effect of Poverty on Child Development and Educational Outcomes. Ann. N. Y. Acad. Sci. (2008) 1136 243-256. DOI: 10.1196/annals.1425.023 6. Larson K., Halfon N.. Family Income Gradients in the Health and Health Care Access of US Children. Matern. Child Health J. (2010) 14 332-342. DOI: 10.1007/s10995-009-0477-y 7. Whitehead M., Bird P.. Breaking the poor health–poverty link in the 21st Century: Do health systems help or hinder?. Ann. Trop. Med. Parasitol. (2006) 100 389-399. DOI: 10.1179/136485906X97363 8. Barkley S., Starfield B., Shi L., Macinko J.. The contribution of primary care to health systems and health. Fam. Med. Class. Pap. (2016) 10 9781315365305 9. Kruk M.E., Porignon D., Rockers P.C., Van Lerberghe W.. The contribution of primary care to health and health systems in low- and middle-income countries: A critical review of major primary care initiatives. Soc. Sci. Med. (2010) 70 904-911. DOI: 10.1016/j.socscimed.2009.11.025 10. Pedersen L.L., Wilkin D.. Primary health care: Definitions, users and uses. Health Care Anal. (1998) 6 341-351. DOI: 10.1007/BF02678373 11. Nutbeam D., Muscat D.M.. Health Promotion Glossary 2021. Health Promot. Int. (2021) 36 1578-1598. DOI: 10.1093/heapro/daaa157 12. Anderson R.M., Funnell M.M., Butler P.M., Arnold M.S., Fitzgerald J.T., Feste C.C.. Patient Empowerment: Results of a randomized controlled trial. Diabetes Care (1995) 18 943-949. DOI: 10.2337/diacare.18.7.943 13. Janchai N., Deoisres W., Chaimongkol N.. A improving health literacy using the health education and health empowerment program in Thai adults with uncontrolled hypertension: A randomized controlled trial. Pac. Rim Int. J. Nurs. Res. (2021) 25 600-613 14. Tsai T.I., Lee S.Y.D., Yu W.R.. Impact of a problem-based learning (PBL) health literacy program on immigrant Women’s health literacy, health empowerment, navigation efficacy, and health care utilization. J. Health Commun. (2018) 23 340-349. DOI: 10.1080/10810730.2018.1445798 15. Chang A.K., Fritschi C., Kim M.J.. Sedentary Behavior, Physical Activity, and Psychological Health of Korean Older Adults with Hypertension: Effect of an Empowerment Intervention. Res. Gerontol. Nurs. (2013) 6 81-88. DOI: 10.3928/19404921-20121219-01 16. Lin S.-C., Chen I.-J., Yu W.-R., Lee S.-Y.D., Tsai T.-I.. Effect of a community-based participatory health literacy program on health behaviors and health empowerment among community-dwelling older adults: A quasi-experimental study. Geriatr. Nurs. (2019) 40 494-501. DOI: 10.1016/j.gerinurse.2019.03.013 17. Sarkar K., Dasgupta A., Sinha M., Shahbabu B.. Effects of health empowerment intervention on resilience of adolescents in a tribal area: A study using the Solomon four-groups design. Soc. Sci. Med. (2017) 190 265-274. DOI: 10.1016/j.socscimed.2017.05.044 18. Stephens T., Braithwaite H., Johnson L., Harris C., Katkowsky S., Troutman A.. Cardiovascular risk reduction for African-American men through health empowerment and anger management. Health Educ. J. (2008) 67 208-218. DOI: 10.1177/0017896908094638 19. Shearer N.B.C., Fleury J.D., Belyea M.. Randomized control trial of the Health Empowerment Intervention: Feasibility and impact. Nurs. Res. (2010) 59 203-211. DOI: 10.1097/NNR.0b013e3181dbbd4a 20. Buonomo I., Santoro P.E., Benevene P., Borrelli I., Angelini G., Fiorilli C., Gualano M.R., Moscato U.. Buffering the Effects of Burnout on Healthcare Professionals’ Health—The Mediating Role of Compassionate Relationships at Work in the COVID Era. Int. J. Environ. Res. Public Health (2022) 19. DOI: 10.3390/ijerph19158966 21. Arneson H., Ekberg K.. Evaluation of empowerment processes in a workplace health promotion intervention based on learning in Sweden. Health Promot. Int. (2005) 20 351-359. DOI: 10.1093/heapro/dai023 22. Levy L.B., O’Hara M.W.. Psychotherapeutic interventions for depressed, low-income women: A review of the literature. Clin. Psychol. Rev. (2010) 30 934-950. DOI: 10.1016/j.cpr.2010.06.006 23. Beeber L.S., Cooper C., Van Noy B.E., Schwartz T.A., Blanchard H.C., Canuso R., Robb K., Laudenbacher C., Emory S.L.. Flying under the radar: Engagement and retention of depressed low-income mothers in a mental health intervention. Adv. Nurs. Sci. (2007) 30 221-234. DOI: 10.1097/01.ANS.0000286621.77139.f0 24. Kaltman S., de Mendoza A.H., Serrano A., Gonzales F.A.. A mental health intervention strategy for low-income, trauma-exposed Latina immigrants in primary care: A preliminary study. Am. J. Orthopsychiatry (2016) 86 345-354. DOI: 10.1037/ort0000157 25. Reynolds A.J., Temple J.A., Ou S.R., Robertson D.L., Mersky J.P., Topitzes J.W., Niles M.D.. Effects of a school-based, early childhood intervention on adult health and well-being: A 19-year follow-up of low-income families. Arch. Pediatr. Adolesc. Med. (2007) 161 730-739. DOI: 10.1001/archpedi.161.8.730 26. Reynolds A.J., Temple J.A., Robertson D.L., Mann E.A.. Long-term effects of an early childhood intervention on educational achievement and juvenile arrest: A 15-year follow-up of low-income children in public schools. JAMA (2001) 285 2339-2346. DOI: 10.1001/jama.285.18.2339 27. Everhart R.S., Heron K.E., Leibach G.G., Miadich S.A.. Developing a Mobile Health Intervention for Low-Income, Urban Caregivers of Children with Asthma: A Pilot Study. Pediatr. Allergy Immunol. Pulmonol. (2017) 30 252-256. DOI: 10.1089/ped.2017.0794 28. Celano M.P., Holsey C.N., Kobrynski L.J.. Home-based family intervention for low-income children with asthma: A randomized controlled pilot study. J. Fam. Psychol. (2012) 26 171-178. DOI: 10.1037/a0027218 29. Herman A., Jackson P.. Empowering Low-Income Parents with Skills to Reduce Excess Pediatric Emergency Room and Clinic Visits through a Tailored Low Literacy Training Intervention. J. Health Commun. (2010) 15 895-910. DOI: 10.1080/10810730.2010.522228 30. Curry S.J., Ludman E.J., Graham E., Stout J., Grothaus L., Lozano P.. Pediatric-based smoking cessation intervention for low-income women: A randomized trial. Arch. Pediatr. Adolesc. Med. (2003) 157 295-302. DOI: 10.1001/archpedi.157.3.295 31. Bull E.R., Dombrowski S.U., McCleary N., Johnston M.. Are interventions for low-income groups effective in changing healthy eating, physical activity and smoking behaviours? A systematic review and meta-analysis. BMJ Open (2014) 4 e006046. DOI: 10.1136/bmjopen-2014-006046 32. Taylor W.C., Baranowski T., Young D.R.. Physical activity interventions in low-income, ethnic minority, and populations with disability. Am. J. Prev. Med. (1998) 15 334-343. DOI: 10.1016/S0749-3797(98)00081-6 33. Butterfield P.G., Hill W., Postma J., Butterfield P.W., Odom-Maryon T.. Effectiveness of a household environmental health intervention delivered by rural public health nurses. Am. J. Public Health (2011) 101 S262-S270. DOI: 10.2105/AJPH.2011.300164 34. Wadsworth M., Santiago C.D., Einhorn L., Etter E.M., Rienks S., Markman H.. Preliminary Efficacy of an Intervention to Reduce Psychosocial Stress and Improve Coping in Low-Income Families. Am. J. Community Psychol. (2011) 48 257-271. DOI: 10.1007/s10464-010-9384-z 35. Campbell F.A., Ramey C.T.. Effects of early intervention on intellectual and academic achievement: A follow-up study of children from low-income families. Child Dev. (1994) 65 684-698. DOI: 10.2307/1131410 36. Carroll J.M.. A Concise History of Hong Kong (2007) 37. 37. Census and Statistics Department Hong Kong 2016 Population By-Census—Thematic Report: Household Income Distribution in Hong KongCensus and Statistics DepartmentHong Kong, China2016. Hong Kong 2016 Population By-Census—Thematic Report: Household Income Distribution in Hong Kong (2016) 38. 38. Census and Statistics Department Hong Kong Poverty Situation Report 2020Census and Statistics DepartmentHong Kong, China2021. Hong Kong Poverty Situation Report 2020 (2021) 39. Help the Working Poor Families Move out of Poverty-Exploration on the Low Income Supplement in Hong Kong. (2013) 40. 40. The Hong Kong Council of Social Service Low Income Household Population in Outlying Island of Hong Kong ReportThe Hong Kong Council of Social ServiceHong Kong, China2006. Low Income Household Population in Outlying Island of Hong Kong Report (2006) 41. 41. Hospital Authority Hospital Authority Statistical Report (2011–2012)The Hospital AuthorityHong Kong, China2013. Hospital Authority Statistical Report (2011–2012) (2013) 42. 42. Census and Statistics Department 2011 Population Census—Main Tables (New Town)Census and Statistics DepartmentHong Kong, China2011. 2011 Population Census—Main Tables (New Town) (2011) 43. Economist Intelligence Unit, Worldwide Cost of Living 2022. (2022) 44. Kwai Tsing Health City and Safe Community. (2004) 45. Bedford L.E., Yeung M.H.Y., Au C.H., Tse E.T.Y., Yim W.Y., Yu E.Y.T., Wong C.K.H., Lam C.L.K.. The validity, reliability, sensitivity and responsiveness of a modified Patient Enablement Instrument (PEI-2) as a tool for serial measurements of health enablement. Fam. Pract. (2020) 38 339-345. DOI: 10.1093/fampra/cmaa102 46. Zanon C., Brenner R.E., Baptista M.N., Vogel D.L., Rubin M., Al-Darmaki F.R., Gonçalves M., Heath P.J., Liao H.Y., Mackenzie C.S.. Examining the dimensionality, reliability, and invariance of the Depression, Anxiety, and Stress Scale–21 (DASS-21) across eight countries. Assessment (2021) 28 1531-1544. DOI: 10.1177/1073191119887449 47. Wang K., Shi H.-S., Geng F.-L., Zou L.-Q., Tan S.-P., Wang Y., Neumann D.L., Shum D.H.K., Chan R.C.K.. Cross-cultural validation of the Depression Anxiety Stress Scale–21 in China. Psychol. Assess. (2016) 28 e88-e100. DOI: 10.1037/pas0000207 48. Lam C.L., Tse E.Y., Gandek B.. Is the standard SF-12 health survey valid and equivalent for a Chinese population?. Qual. Life Res. (2005) 14 539-547. DOI: 10.1007/s11136-004-0704-3 49. Lam E.T., Lam C.L., Fong D.Y., Huang W.W.. Is the SF-12 version 2 health survey a valid and equivalent substitute for the SF-36 version 2 health survey for the Chinese?. J. Eval. Clin. Pract. (2013) 19 200-208. DOI: 10.1111/j.1365-2753.2011.01800.x 50. Frost J., Currie M.J., Cruickshank M.. An Integrative Review of Enablement in Primary Health Care. J. Prim. Care Community Health (2015) 6 264-278. DOI: 10.1177/2150131915598373 51. Cook R.D.. Influential observations in linear regression. J. Am. Stat. Assoc. (1979) 74 169-174. DOI: 10.1080/01621459.1979.10481634 52. Schmidt A.F., Finan C.. Linear regression and the normality assumption. J. Clin. Epidemiol. (2018) 98 146-151. DOI: 10.1016/j.jclinepi.2017.12.006 53. Lindacher V., Curbach J., Warrelmann B., Brandstetter S., Loss J.. Evaluation of Empowerment in Health Promotion Interventions: A Systematic Review. Evaluation Health Prof. (2017) 41 351-392. DOI: 10.1177/0163278716688065 54. Forlani G., Zannoni C., Tarrini G., Melchionda N., Marchesini G.. An empowerment-based educational program improves psychological well-being and health-related quality of life in Type 1 diabetes. J. Endocrinol. Investig. (2006) 29 405-412. DOI: 10.1007/BF03344123 55. Kravitz R.L., Tancredi D.J., Grennan T., Kalauokalani D., Street R.L., Slee C.K., Wun T., Oliver J.W., Lorig K., Franks P.. Cancer Health Empowerment for Living without Pain (Ca-HELP): Effects of a tailored education and coaching intervention on pain and impairment. Pain (2011) 152 1572-1582. DOI: 10.1016/j.pain.2011.02.047 56. Williams J., Tucker C.M., Wippold G., Roncoroni J., Moutinho M.E., Lee S., Desmond F., Ateyah W.. Impact of a Health Empowerment-focused Program on Adult Employee’s Motivators of and Barriers to Engaging in Healthy Behaviors. Am. J. Health Educ. (2020) 51 169-178. DOI: 10.1080/19325037.2020.1744490 57. Champion V.L., Skinner C.S.. The health belief model. Health Behav. Health Educ. Theory Res. Pract. (2008) 4 45-65 58. Shearer N.B.C.. Health Empowerment Theory as a Guide for Practice. Geriatr. Nurs. (2009) 30 4-10. DOI: 10.1016/j.gerinurse.2009.02.003 59. 59. Leisure and Cultural Services Department Leisure and Cultural Services Department (LCSD) Suspends Enrolment Arrangements of Recreation and Sports ProgrammesLeisure and Cultural Services DepartmentHong Kong, China2020. Leisure and Cultural Services Department (LCSD) Suspends Enrolment Arrangements of Recreation and Sports Programmes (2020) 60. Wu X.Y., Han L.H., Zhang J.H., Luo S., Hu J.W., Sun K.. The influence of physical activity, sedentary behavior on health-related quality of life among the general population of children and adolescents: A systematic review. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0187668 61. Babbie E.. Survey Methods Research (1973)
--- title: Limonene, a Monoterpene, Mitigates Rotenone-Induced Dopaminergic Neurodegeneration by Modulating Neuroinflammation, Hippo Signaling and Apoptosis in Rats authors: - Lujain Bader Eddin - Sheikh Azimullah - Niraj Kumar Jha - Mohamed Fizur Nagoor Meeran - Rami Beiram - Shreesh Ojha journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049348 doi: 10.3390/ijms24065222 license: CC BY 4.0 --- # Limonene, a Monoterpene, Mitigates Rotenone-Induced Dopaminergic Neurodegeneration by Modulating Neuroinflammation, Hippo Signaling and Apoptosis in Rats ## Abstract Rotenone (ROT) is a naturally derived pesticide and a well-known environmental neurotoxin associated with induction of Parkinson’s disease (PD). Limonene (LMN), a naturally occurring monoterpene, is found ubiquitously in citrus fruits and peels. There is enormous interest in finding novel therapeutic agents that can cure or halt the progressive degeneration in PD; therefore, the main aim of this study is to investigate the potential neuroprotective effects of LMN employing a rodent model of PD measuring parameters of oxidative stress, neuro-inflammation, and apoptosis to elucidate the underlying mechanisms. PD in experimental rats was induced by intraperitoneal injection of ROT (2.5 mg/kg) five days a week for a total of 28 days. The rats were treated with LMN (50 mg/kg, orally) along with intraperitoneal injection of ROT (2.5 mg/kg) for the same duration as in ROT-administered rats. ROT injections induced a significant loss of dopaminergic (DA) neurons in the substantia nigra pars compacta (SNpc) and DA striatal fibers following activation of glial cells (astrocytes and microglia). ROT treatment enhanced oxidative stress, altered NF-κB/MAPK signaling and motor dysfunction, and enhanced the levels/expressions of inflammatory mediators and proinflammatory cytokines in the brain. There was a concomitant mitochondrial dysfunction followed by the activation of the Hippo signaling and intrinsic pathway of apoptosis as well as altered mTOR signaling in the brain of ROT-injected rats. Oral treatment with LMN corrected the majority of the biochemical, pathological, and molecular parameters altered following ROT injections. Our study findings demonstrate the efficacy of LMN in providing protection against ROT-induced neurodegeneration. ## 1. Introduction Parkinson’s disease (PD) is one of the prevalent slow progressing disorders, pathologically featured by the degeneration of dopamine-producing neurons in substantia nigra pars compacta (SNpc), which culminates in disabling motor dysfunctions, such as bradykinesia, muscular rigidity, resting tremor, and gait impairment [1,2]. Compelling experimental evidence implicates oxidative damage, inflammation, mitochondrial dysfunction and apoptosis as major factors in the induction and progression of PD [3]. Neuroinflammation is an intrinsic player in PD progression that apparently reflects the role of microglia in PD [4]. Activated microglia produces proinflammatory cytokines causing neurotoxicity as well as chemokines that can recruit leukocytes to the central nervous system, exacerbating inflammatory response [5]. In recent years, substantial evidence on mitochondrial dysfunction and apoptosis in the etiopathogenesis of PD has been demonstrated [6]. Various studies demonstrate that impaired mitochondrial bioenergetics and dysfunction, especially perturbations in electron transport chain (ETC) complex-I, play an important role in the pathogenesis of PD [7]. Available evidence revealed a strong correlation between the levels of α-synuclein and disease severity. Malfunction of the proteolytic machineries responsible for α-synuclein degradation may result in the aggregation and accumulation of the protein, which characterize the disease [8]. Consequently, accumulated α-synuclein results in oxidative stress and neuro-inflammation which are the two substantial pathological mechanisms in PD [9]. Overexpressed α-synuclein, especially in its mutant form, is associated with an increased rate of neuronal cell death [10]. These pathological events finally result in dopamine depletion following the degeneration of dopaminergic neurons and the emergence of motor symptoms in PD. Hence, oxidative stress, neuroinflammation, and apoptotic cell death of dopaminergic neurons emerged as vital targets for therapeutic intervention in PD [11]. Currently, there are no available pharmacological agents to completely cure or delay or halt the progression of PD. Levodopa is still considered the most effective symptomatic reliever in controlling the severity of symptoms associated with PD. Therefore, there is a need for developing agents which can target complex interlinked cascades involved in PD [12]. Rotenone (ROT), a naturally occurring pesticide, is widely used for the induction of dopaminergic neurodegeneration in experimental animals. It is a specific inhibitor of mitochondrial complex I of the respiratory chain, producing free-radical-mediated oxidative stress [13]. The model of ROT-induced dopaminergic neurodegeneration that represents features of PD is more advantageous than other chemical-based experimental models of PD as it resembles the neurological and behavioral changes of PD reflecting the selective degeneration of dopaminergic neurons [14]. ROT-injected animals evidently mimic the pathological features of PD in patients, such as the progressive loss of dopaminergic neurodegeneration, α-synuclein aggregation in dopaminergic neurons, Lewy body formation, oxidative stress, mitochondrial dysfunction, microglial activation, and neuroinflammation [15]. Therefore, exposing rats to ROT is a relevant and interesting model that has gained popularity in studying PD pathogenesis and in pharmacological evaluations of agents [13]. Plant-derived bioactive agents, commonly known as phytochemicals, that are used either as nutraceuticals or dietary supplements or pharmaceuticals, have caught the attention of most recent research due to their high therapeutic potential and low toxicity. Phytochemicals appear to have an inevitable role in the prevention and treatment of chronic diseases, including neurodegeneration. Among plant-derived bioactive compounds, the triterpenoid class of compounds has received enormous attention in recent times. One of the popular triterpenoid compounds, limonene (LMN) has been found to be effective against experimental models of numerous diseases, including neurological disorders such as Alzheimer’s disease [16], stroke [17], and cerebral ischemia [18]. LMN exhibits a variety of biological properties, such as antioxidant, anti-inflammatory, anticancer, gastroprotective, neuroprotective [19] properties, and induction of autophagy [20], along with negligible toxicity [21,22]. LMN has been reported to be absorbed considerably and undergoes rapid biotransformation to active metabolites (perillic acid, dihydroperillic acid, and limonene-1,2-diol) in both humans and animal models [23,24,25]. LMN is excreted in the urine within 48 h in humans and the half-life ranges from 12 to 24 h [25]. Considering the neuroprotective potential and favorable pharmacokinetic properties and druggability of LMN, it is imperative to investigate the effects of LMN on dopaminergic neurodegeneration, a characteristic of PD. Thus, this present study aimed to evaluate the therapeutic potential of LMN in ROT-induced PD in rats. Further, the underlying mechanism was determined by evaluating markers of oxidative stress, neuroinflammation, and apoptotic signaling along with immunostaining of dopaminergic neurons. ## 2.1. Behavioral Assessment of the Impact of Limonene on ROT-Induced Neurodegeneration To evaluate the effect of LMN on ROT-induced dyskinesia, a rotarod test was performed to measure rats’ ability to maintain themselves on the rotating rod. Behavioral test results showed a gradual decrease in balance level and muscle strength in ROT-injected rats (Figure 1). Rats injected with ROT displayed a significant ($p \leq 0.05$) decline in rotarod performance compared to normal control rats. LMN treatment to the rats injected with ROT showed improved ability to balance as observed in the rotarod test (Figure 1). It decreased the latency time to fall, where it showed a significant ($p \leq 0.05$) increase in retention time on the rotarod as compared with ROT-injected rats. LMN treatment did not affect the performance of rats compared to the normal control group (Figure 1). ## 2.2. Limonene Preserves Dopaminergic Neurons against ROT-Induced Neurodegeneration To investigate the effect of LMN on ROT-induced alterations in the nigrostriatal system in rats, the loss of dopaminergic neurons was detected by staining the TH-ir neurons and their corresponding fibers in the striatum. Immunohistochemistry of TH assessed the presence of TH-ir dopaminergic neurons in the SN and TH-ir dopaminergic striatal fibers. ROT-injected rats showed a significant ($p \leq 0.05$) reduced number of TH-positive cells in the SN and a reduced density of striatal TH-ir fibers (Figure 2). LMN treatment to ROT-injected rats decreased the loss of TH-positive neurons, evidenced by the increased number of dopaminergic neurons in SN and TH-ir fibers in striatum. In the normal-control- and LMN-treated groups, the SN contained large numbers of TH-immunopositive neurons along with dopaminergic content in the striatum. This suggests that LMN treatment can reduce the loss of dopaminergic neurons caused by ROT (Figure 2). ## 2.3. Effect of Limonene on Brain-Derived Neurotrophic Factor (BDNF) and α-Synuclein Expression in the Striatum of Rats Inefficient neuronal supply of BDNF can lead to PD due to the defects in synaptic plasticity associated with BDNF loss. To investigate the effects of LMN on BDNF, the expressions of BDNF in striatum of different experimental groups were studied (Figure 3). A significant decrease ($p \leq 0.05$) in the BDNF protein expression was observed, whereas LMN treatment was observed to be associated with increased expression of BDNF, which reflects a prevention of the loss of BDNF. To further confirm that LMN treatment protects neurons, we measured the level of α-synuclein, which accounts for deleterious effects following abnormal accumulation in dopaminergic neurons. Western blot analysis reveals that ROT injections induced a significant increase ($p \leq 0.05$) in expression of α-synuclein in the striatum when compared to the normal control rats (Figure 3). However, LMN treatment reduced the expression of α-synuclein that could be indicative of a reduced accumulation of α-synuclein in comparison with the rats that received only ROT (Figure 3). ## 2.4. Limonene Attenuates Lipid Peroxidation and Enhances the Activities/Concentrations of Enzymatic and Non-Enzymatic Antioxidant Status in the Midbrain of ROT-Induced Neurodegeneration The antioxidant effect of LMN was evaluated by assessing the activity of antioxidant enzymes and the extent of peroxidized compounds formed in the midbrain. The MDA levels were considerably ($p \leq 0.05$) increased whereas the activities of SOD, catalase, and concentrations of GSH were remarkably ($p \leq 0.05$) reduced in ROT-injected rats compared to normal control rats (Figure 4). However, LMN treatment significantly ($p \leq 0.05$) reduced the MDA levels and significantly ($p \leq 0.05$) improved the activities/concentrations of SOD, catalase, and GSH in rats that received ROT compared to the rats only injected with ROT (Figure 4). ## 2.5. Limonene Attenuates Proinflammatory Cytokines in the Midbrain of ROT-Induced Neurodegeneration Neuroinflammation is considered one of the key pathogenic events that plays a role in the progression of neurodegeneration. Therefore, we investigated the effect of LMN on ROT-induced neuroinflammation by measuring the levels of the proinflammatory cytokines (TNF-α, IL-1β, and IL-6) in the midbrain of rats. ROT injection to the rats caused a considerable ($p \leq 0.05$) rise in the levels of TNF-α, IL-1β, and IL-6 in the midbrain compared to normal control rats (Figure 5), whereas treatment with LMN produced a significant ($p \leq 0.05$) reduction in the levels of these proinflammatory cytokines in the midbrain compared to only ROT-injected rats (Figure 5). ## 2.6. Limonene Attenuates Activation of Microglia and Astrocytes in ROT-Induced Neurodegeneration Activation of GFAP and Iba-1 are considered as markers of ROS production and inflammatory process. Immunofluorescence staining of GFAP and Iba-1 in the striatum shows the high number of activated GFAP-positive astrocytes (Figure 6a) and Iba-1-positive microglia (Figure 6b) in ROT-injected rats compared to normal control rats. This is conjectured by the increased number as well as size of astrocytes and microglia. However, LMN treatment to ROT-administered rats showed reduced activation of astrocytes and microglia when compared to ROT-injected rats. The quantification of activated astrocytes and microglia are presented in Figure 2. Consistently, there is activation of more astrocytes and microglia in ROT-injected rats in comparison with the control rats, while LMN treatment exerted a reduction in the number of activated astrocytes and microglia when compared with rats injected with ROT only. Normal-control-treated rats and those treated with LMN only did not display notable expression of astrocytes and microglia, which is suggestive of no adverse effects on astrocytes and microglia with the studied doses. ## 2.7. Limonene Treatment Inhibits Expression of Inflammatory Mediators and NF-κB/IκB Activation in the Striatum of ROT-Induced Neurodegeneration To further characterize the inhibitory effect of LMN on mediators of inflammation, we investigated the activation of NF-κB via Western blotting. ROT-injected rats showed a significant ($p \leq 0.05$) rise in the protein expressions of iNOS, COX-2, p-NF-κB, and p-IκB in the striatum of rats (Figure 7). However, LMN treatment to ROT-injected rats produced a significant ($p \leq 0.05$) reduction in the expressions of iNOS, COX-2, p-NF-κB, and p-IκB in the striatum. Rats injected with LMN alone did not exhibit any noticeable alteration in the expressions of all these proteins in the striatum (Figure 7). ## 2.8. Limonene Treatment Reduces Phosphorylation of MAPK Signaling Proteins in the Striatum of ROT-Induced Neurodegeneration To confirm whether LMN regulates the role of the P38 MAPK and mTOR signaling pathways in PD induction, we further investigated p-JNK, p-p38, and p-mTOR activation in different experimental groups. Phosphorylation of JNK and P38 was significantly ($p \leq 0.05$) increased in ROT-injected rats in comparison with normal control rats (Figure 8), whereas LMN treatment to ROT-injected rats produced a significant ($p \leq 0.05$) reduction in the phosphorylation of MAPK proteins when compared to only ROT injected rats. Phosphorylated mTOR was downregulated in ROT-injected rats, whereas LMN treatment for ROT-injected rats restored normal expression. Both normal-control-, and LMN-alone-treated rats had no notable change in the expression of these phosphorylated MAPK signaling proteins (Figure 8). ## 2.9. Limonene Inhibits ROT-Induced Mitochondrial Complex-I Inhibition in the Striatum ROT is well known as a complex-I inhibitor. Therefore, we evaluated complex-I to ascertain the effect of LMN in ROT-induced PD. The expression of mitochondrial complex-I was significantly decreased ($p \leq 0.05$) in ROT-injected rats compared to normal control rats, whereas LMN treatment produced a significant ($p \leq 0.05$) rise in the expressions of mitochondrial complex-I compared to only ROT injected rats (Figure 9). ## 2.10. Limonene Treatment Attenuates Apoptosis and Hippo Signaling in ROT-Injected Rats The intrinsic pathway of apoptosis is initiated by proteolytic activation of the initiator caspase-9 followed by caspase-3 cleavage and activation. We studied the expression of apoptotic proteins and the change induced following LMN treatment. ROT injections to rats caused a significant ($p \leq 0.05$) rise in the protein expressions of Bax, cleaved caspase-3, cleaved caspase-9, cytochrome-C, CHOP, and p-MST1 compared to normal control rats (Figure 10), whereas treatment with LMN showed a significant ($p \leq 0.05$) decrease in the expression of these apoptotic and Hippo signaling proteins compared to only ROT injected rats. In contrast, the antiapoptotic protein Bcl2 was decreased in ROT-injected rats, whereas it was increased following LMN treatment. Expressions of these proteins remained unaffected in normal-control- and LMN-alone-treated rats (Figure 10). ## 3. Discussion To our knowledge, this current study is the first report on the neuroprotective role of LMN in ROT-induced PD in rats. The neuroprotective effects and mechanisms observed in the present study have been summarized in Figure 11. PD is characterized by a functional impairment of voluntary movements which leads to slowness in fine motor function, including dyskinesia, hypokinesia, or akinesia [26]. PD patients frequently reported difficulties in hand dexterity movements [27]. In our study, we observed reduced time spent on the rotarod for rats injected with ROT, which is consistent with previous reports which concluded that ROT-administered rats remained on the rotating rod for less time compared with pre-ROT injections [28]. However, pretreatment with LMN significantly attenuated ROT-induced locomotor deficits; a sign of the protective ability of LMN against ROT-induced neurodegenerative changes. ROT as a potent inhibitor of complex-1 of the mitochondrial electron transport chain leads to a syndrome that replicates both the neuropathological and behavioral symptoms of PD. Administration of ROT in vivo has been shown to induce oxidative stress, neuroinflammation, and apoptosis subsequent to both the loss of tyrosine hydroxylase and the formation of fibrillary cytoplasmic inclusions containing α-synuclein. It also induces motor deficits such as bradykinesia and gait instability [27]. ROT has been used for remodeling PD in research since its ability to reproduce the hallmarks of PD was reported [29]. Tyrosine hydroxylase is the rate-limiting enzyme of catecholamine biosynthesis converting tyrosine to DOPA. PD affects specifically TH-containing catecholamine neurons. The most prominent neurodegeneration in PD patients is observed in the nigrostriatal DA neurons, which contain abundant TH. Therefore, TH has been speculated to have an essential role in the pathophysiology of PD [30]. In the current study, LMN attenuated the dopaminergic neurodegeneration induced by ROT administration by preventing oxidative stress, reinstating mitochondrial complex-I activity, and maintaining TH-ir neurons in SN. BDNF as a neurotrophin mediates neurogenesis to maintain neuronal integrity. BDNF signaling through tropomyosin-related kinase B receptor (TrkB) is reported to protect nigrostriatal dopaminergic neurons in aged brains [31]. Imbalanced levels of BDNF can affect the motor and cognitive performance in parkinsonian patients. Deteriorated parkinsonian symptoms are correlated with low BDNF levels in the SN and caudal-putamen nuclei of PD patients [32]. Our study indicated that ROT caused a significant decrease in BDNF expressions, whereas LMN treatment reversed the effect of ROT and enhanced BDNF expressions. A critical event in PD that has been supported by strong evidence is α-synuclein accumulation mediating the pathogenesis of PD and leading to other mechanisms representing the cornerstone in PD pathology [33]. α-synuclein aberrant soluble oligomeric conformations, termed protofibrils, are known to mediate disruption of cellular homeostasis and neuronal death. Our data indicate that LMN co-treatment reduced the content of α-synuclein in comparison with ROT only injected rats, suggesting that LMN mediated the inhibition of α-synuclein aggregations. Mitochondrial complex-I is a rate-limiting enzyme complex involved in oxidative phosphorylation and ATP synthesis for maintaining the mitochondrial bioenergetics [34]. Intoxication of mitochondrial complex-I reproduces the motor symptoms of PD in experimental models [35]. ROT has a high affinity to inhibit electron transport chain-1 (ETC-I). Persistent inhibition of ETC-I leads to a leakage of electrons, which combine with O2 results in excessive ROS production [36]. LMN treatment restored the activity of complex-I and maintained its function. Dopaminergic neurons in PD exist in a state of constant oxidative stress, partly due to the free radical generation [37]. Free-radical-induced lipid peroxidation and oxidative stress play a crucial role in PD pathogenesis [38]. The loss of antioxidant defenses leads to a buildup of ROS which is associated with deleterious effects on dopaminergic neurons in PD [39]. ROT-administered rats showed a decline in antioxidant defense by decreasing the activities/concentrations of SOD, catalase, and GSH in the midbrain. Decreased activities/concentrations of SOD, catalase, and GSH in ROT-injected rats resulted from the inactivation of these enzymatic and non-enzymatic antioxidants by H2O2. Previous studies have revealed that LMN is a potent scavenger of toxic free radicals, such as hydrogen peroxide (H2O2), superoxide, and hydroxyl free radicals, in vitro [40,41]. However, LMN treatment to ROT-injected rats showed near normalized activities/concentrations of SOD, catalase, and GSH ascribed to its potent antioxidant activity. Consistent with previous reports, our study demonstrated that the neuroprotective ability of LMN is attributed to its free radical scavenging properties and preserving the enzymatic and non-enzymatic antioxidants. Microglial activation and increased reactive astrocytes are aggravated by the mutated forms of α-synuclein aggregates that act as chemoattractants of microglial migration toward damaged neurons [42]. Activated microglia followed by IκB phosphorylation and its proteolytic degradation trigger NF-κB nuclear activation and translocation which upregulate the inflammatory mediators iNOS and COX-2 and release enormous amounts of proinflammatory cytokines, such as TNF-α, IL-6, and IL-1β, which exacerbate neurodegeneration [43]. Our results showed that ROT-injected rats displayed sustained activation of microglia cells in parallel with an accumulation of inflammatory mediators and cytokines in the midbrain. These changes were remarkably attenuated in rats treated with LMN. The ability of LMN in limiting neuroinflammation is supported by previous reports where a reduction in the expression of IL-1β and TNF-α in the hippocampus of mice exposed to stress following maternal separation has been demonstrated [44]. MAPK family proteins are well known to play a role in regulating the release of proinflammatory cytokine gene expressions. P38 MAPK has a crucial role in PD progression through aggravating neuroinflammation by enhancing proinflammatory cytokines through microglial activation [45]. ROT-induced PD results in P38 MAPK upregulation, which plays a primary role in inducing neuroinflammation and apoptosis [46], where it controls NF-κB and the downstream cytokines [47]. Several reports showed that JNK phosphorylation with subsequent phosphorylation of c-Jun increased the production of proinflammatory cytokines [48]. In ROT-induced neurotoxicity in rats, MAPK signaling proteins were activated which were obviously reversed by LMN revealing its anti-inflammatory effect. Accumulation of aberrant misfolded proteins within the dopaminergic neurons leads to cytotoxic oxidative stress which activates cell death in PD. Apoptosis is responsible for neuronal loss in PD which is evidenced by the elevated activity of pro-apoptotic proteins in the postmortem brain tissue of PD patients [49]. Mitochondria plays a key role in mediating apoptotic cell death. Bax induces the permeabilization of the outer mitochondrial membrane, causing the release of cytochrome-C from the mitochondrial intermembranous space. This eventually leads to the cleavage of caspase-9, activation of caspase-3, and apoptosis induction [49]. Concomitantly, there has been a rise in the expression of Bax, cleaved caspase-3 and cytochrome-C, with declined Bcl-2 expression in ROT-injected rats. LMN treatment showed an attenuation of apoptosis as reflected by decreased Bax, cleaved caspase-3 and cytochrome-C, and increased Bcl-2 expressions. This demonstrates that LMN possesses the ability to counter ROT-induced neuronal cell death via inhibiting oxidative stress and apoptosis. It has been observed that ROT induces caspase-dependent and independent apoptosis through the suppression of mTOR signaling [50]. The activation of mTOR signaling has been reported to protect the dopaminergic neurons against apoptotic death and degradation [51]. The present study results revealed a similar effect, wherein LMN treatment produced neuroprotective effects by restoration of mTOR expression suppressed in ROT-injected rats. The initiation of programmed cell death involves remarkable upregulation of the transcription factor C/EBP-homologous protein (CHOP) that has been implicated in the neuronal death in the context of oxidative and endoplasmic reticulum stress (ER stress). Additionally, the Hippo signaling pathway represented by CHOP and MST has recently had a major emphasis placed on its profound effect in neurodegeneration subsequent to neuroinflammatory changes by potentiating apoptosis, cellular growth inhibition, and tissue degeneration [52]. The Hippo signaling pathway plays a critical role in dopaminergic neuronal loss evidenced by MST1 activation along with activation of oxidative-stress-induced cell death. Activated Hippo/MST1 is coupled with caspase-3 activation that is aligned with a loss of dopaminergic neurons in PD brains [53]. Overexpression of CHOP, an ER stress mediator, has been reported in various models of PD [54] and mediates apoptosis by upregulating the BH3 only family proteins [55]. MST phosphorylation has also been correlated with oxidative-stress-induced apoptosis in cardiomyopathy [56]. Our study findings indicate that ROT triggers the activation of CHOP/p-MST which contribute to neurodegeneration by augmenting the oxidative response and apoptosis. However, the protective effect of LMN against the upregulation of Hippo signaling proteins in the striatum is suggestive of its antiapoptotic property. ## 4.1. Drugs and Chemicals Rotenone, dimethyl sulfoxide, miglyol, phosphate-buffered saline, paraformaldehyde, and D-limonene were procured from Sigma Chemicals (St. Louis, MO, USA). The chemicals used in the present study were of analytical grade. ## 4.2. Experimental Animals Healthy male albino Wistar rats (260–300 g) were bred in the Animal Research Facility, College of Medicine and Health Sciences (CMHS), United Arab Emirates University (UAEU). The animals were kept in cages and housed in standard laboratory conditions of relative humidity, temperature, and light/dark cycles with unlimited access to commercially available rat chow diet and water ad libitum. A maximum of 5 rats were placed per cage. Experiments were conducted according to the guidelines approved (Approval No. ERA_2017_5500) by the Animal Ethics Committee of the United Arab Emirates University. ## 4.3. Experimental Protocol and Study Groups LMN has been evaluated in previous studies and our laboratory studies in the oral dose range of 25 to 100 mg/kg and shown to be effective in mitigating oxidative stress, inflammatory mediators, and favorably modulated cell signaling pathways in experimental models of different diseases [57,58,59]. The oral dose of LMN 50 mg/kg has been found to be effective in experimental models of acute lung injury [59], gastric ulcer [57], ulcerative colitis [58], acute myocardial infarction [60], orofacial pain [61], diabetes [62], and depression [63]. The dose of 50 mg/kg has been chosen for the evaluation in ROT-induced rat model of PD representing dopaminergic neurodegeneration in agreement with the previous studies. For the induction of dopaminergic neurodegeneration in rats mimicking PD in humans, a stock solution of ROT was prepared by dissolving in dimethyl sulfoxide (DMSO). The stock solution was further diluted in mygliol to reach a final concentration of 2.5 mg/mL. The rats were randomly divided into four groups, each containing fifteen. The rats in group I assigned as the normal control (CON) group received vehicle (olive oil) in similar volume to other groups, 5 days a week for a duration of 28 days. The rats in group II assigned as the ROT group were administered ROT (2.5 mg/kg, i.p.), 5 days a week for a period of 28 days. The rats in group III assigned as the LMN group were orally treated with LMN (50 mg/kg) daily, 5 days a week for a period of 28 days. The rats in group IV assigned as the LMN + ROT group were pretreated with LMN (50 mg/kg) and followed by ROT (2.5 mg/kg, i.p.), 5 days a week for a period of 28 days. ## 4.4. Tissue Collection and Preparation At the end of the experimental period, rats were anesthetized using an intraperitoneal injection of pentobarbital sodium (40 mg/kg body weight) and cardiac perfusion was performed by 0.01 M phosphate-buffered solution (PH 7.4) to wash out blood. Midbrain and striatum tissues were snap-frozen in liquid nitrogen and stored at −80 °C until further analysis. Brains used for immunohistochemistry were post-fixed in $4\%$ paraformaldehyde solution for 2 days and immersed in sucrose solution ($30\%$) at 4 °C for 3 days. The sucrose solution was changed daily, and the tissues were sectioned only after noticing they had completely sunk down in the sucrose solution. ## 4.5. Biochemical Studies Midbrain tissues were homogenized using a tissue homogenizer in potassium chloride buffer mixed appropriately with protease and phosphatase inhibitors acquired from Thermo Fisher Scientific (Rockford, IL, USA). The tissue homogenates were centrifuged at 14,000× g for 20 min at 4 °C. Supernatant was collected, stored, and utilized for various biochemical estimations. ## 4.6. Rotarod Test Rotarod test is a well-accepted test for the assessment of neurological disorders in animals and it can be repeatedly applied to each rat to evaluate muscle strength, fore and hind limb motor coordination, and balance of rats [64]. In this test, time was recorded for each rat placed on the rotating rod completing ipsilateral and contralateral rotation. This procedure was performed before the beginning of the treatment, to show no variation among groups and that all rats were healthy. Rats were given three days for training and adaptation. They were allowed on the rotating rod to get familiarized with the test. On the fourth day, the time for which each rat could remain on the rod was recorded. At the end of the treatment, the process was repeated again. The rats were given one day for training and recording took place on the following day. The speed was constant throughout all trials (30 rotations per minute). The data are presented as the mean retaining time on the rotating bar for the rats. ## 4.7. Immunofluorescence Staining for Glial Fibrillary Acidic Protein (GFAP) and Ionized Calcium-Binding Adapter Molecule 1 (Iba-1) The GFAP and Iba1 activation was assessed by immunofluorescence staining of the striatum. Floating striatum sections of 25 μm thickness were cut by microtome and kept temporarily in phosphate-buffered solution (PBS) (Sigma Aldrich, St. Louis, MO, USA) with sodium azide to prevent contamination. Selected sections were rinsed twice using PBS and kept for the duration of one hour in blocking solution containing $10\%$ normal goat serum and $0.3\%$ Triton-X 100 (Sigma Aldrich, St. Louis, MO, USA) in PBS. The floating sections were incubated over night at 4 °C with polyclonal rabbit anti-GFAP (1:1000) (Abcam, Waltham, MA, USA) and anti-Iba-1 (1:1000) (Vako Chemicals, Richmond, VA, USA) primary antibodies. On the following day, the primary antibodies were rinsed off and the sections were incubated for one hour with fluorescently conjugated secondary antibody Alexa Fluor® 488 anti-rabbit. Thereafter, the sections were mounted with anti-fade mounting medium Fluoroshield™ (Sigma-Aldrich, St. Louis, MO, USA) and preserved by a cover slip. The slides were viewed and images were collected using fluorescence microscope, EVOS FL (Thermo Fisher Scientific, Waltham, MA, USA). ## 4.8. Assessment of Activated GFAP and Iba-1 Coronal striatal sections were assessed for ROT-induced activation of glial cells. Increased fluorescence intensity and prolonged glial processes indicated the activation of astrocytes and microglia. Different regions with similar areas were analyzed for activated GFAP and Iba-1 and quantified by Image J software (NIH, Bethesda, MD, USA). ## 4.9. Immunohistochemistry of Tyrosine Hydroxylase (TH) Striatum (0.3 mm of bregma) and substantia nigra (SN) regions (−5 mm from bregma) were sectioned for stereological TH immunostaining analysis. The coronal sections of 25 µm were cut by using a cryostat (Leica, Wetzlar, Germany). The sections were washed with PBS (0.01 M) and blocked with $10\%$ normal goat serum and $0.3\%$ Triton-X 100, for the duration of one hour. Afterwards, rabbit anti-TH primary antibody (1:500) (Millipore, Burlington, MA, USA) was added and kept overnight at 4 °C. On the following day, the sections were rinsed and incubated with biotinylated secondary anti-rabbit antibody (1:1000) for the duration of an hour at room temperature. Visualization of TH immunoreactivity was achieved by incubating sections with avidin–biotin complex (Vector Laboratories Ltd., Burlingame, CA, USA) and 3,3′ diaminobenzidine (DAB). Lastly, the sections stained were mounted on slides using DPX mounting medium and viewed under a light microscope (Olympus, Hamburg, Germany). ## 4.10. Determination of Tyrosine Hydroxylase-Immunoreactive (TH-ir) Dopaminergic Neurons and TH-ir Nerve Fibers Loss In the visualized sections of SNpc and striatum, the immunoreactive neurons of TH were counted and their percentage was calculated in reference to control. Three unified areas from each section of the striatum were measured for their optical density by Image J software. Counting was carried out by taking into account the stained neuronal nucleus. The optical density of overlapping cortex was measured as background and deducted from the optical density of striatum. Assessment of TH-ir neurons was analyzed by an investigator blind to experimental groups. ## 4.11. Western Blotting Analysis Using radioimmunoprecipitation assay buffer (RIPA) (Millipore, Burlington, MA, USA), the striatal tissues were homogenized along with protease and phosphatase inhibitors (Merck Millipore, Burlington, MA, USA) to maximize the protein yield. The homogenized tissue was centrifuged at 12,000 rpm for 20 min at 4 °C to obtain the supernatant. The obtained supernatant was further dissolved in 4× Laemmli sample buffer (Bio-Rad, Hercules, CA, USA) and 2-mercaptoethanol (Sigma Chemicals, St. Louis, MO, USA). The protein content was unified across the samples, separated by polyacrylamide gel electrophoresis, transferred to PVDF membrane (Thermo Fisher Scientific, Rockford, IL, USA), blocked in a blocking solution, and immersed at 4 °C overnight with the following primary antibodies: anti-iNOS (1:2000), anti-phospho-MST (1:500) (Sigma Chemicals, St. Louis, MO, USA), anti-COX-2 (1:1000), anti-cleaved caspase-3 (1:1000), anti-cleaved caspase-9 (1:1000), anti-mTOR (1:1000), anti-phospho-mTOR (1:500), anti-CHOP (1:1000), anti-VDAC (1:2000), anti-phospho-IκBα (1:500), (Cell Signaling Technology, Danvers, MA, USA), anti-α-Syn (1:1000), anti-P38 (1:1000), anti-Cytochrome C (1:2000), anti-mitochondrial complex-I (1:1000), anti-brain-derived neurotropic factor (BDNF) (1:1000), P38 MAPK (1:2000), anti-Bcl-2 (1:500), anti-phospho-P38 MAPK (1:1000), anti-NF-κB (1:1000), anti-phospho-NF-κB (1:500), anti-Bax (1:2500), anti-JNK (1:1000), anti-phospho-JNK (1:5000), anti-β-actin (1:5000) (Santa Cruz Biotechnology, Dallas, TA, USA). The corresponding Horseradish peroxidase-conjugated secondary antibody was added to the membranes and incubated for 1 h at room temperature. The chemiluminescence substrate (Thermo Fisher Scientific, Rockford, IL, USA) was added to facilitate the protein visualization. Densitometric analysis was performed using Image J software. ## 4.12. Protein Estimation The amount of protein in each sample was determined using commercially available Pierce™ BCA protein assay kit (Thermo Fisher Scientific, Rockford, IL, USA). ## 4.13. Assessment of Enzymatic and Non-Enzymatic Antioxidant Status Activities of superoxide dismutase (SOD), catalase, and glutathione (GSH) were measured in the midbrain using assay kits (Cayman Chemicals Co., Ann Arbor, MI, USA; Sigma-Aldrich, St. Louis, MO, USA). The activities of SOD, catalase, and GSH were calculated as U/mL, nmol/min/mL, and μM/mL, respectively. ## 4.14. Malondialdehyde (MDA) Assay Malondialdehyde (MDA) levels were estimated using MDA detection kits (Northwest Life Science, Vancouver, WA, USA). The values are represented as μM/mL. ## 4.15. Assessment of Proinflammatory Cytokines The levels of proinflammatory cytokines, such as tumor necrosis factor-alpha (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6), were measured using commercial ELISA kits (BioSource International, Camarillo, CA, USA). The values were represented as pg/mL. ## 4.16. Mitochondrial Extraction from the Striatum Mitochondrial fraction was isolated from the striatum using a commercially available kit following manufacturer’s protocol (Abcam, Waltham, MA, USA). ## 4.17. Statistical Analysis The data are represented as the mean ± standard error of the mean (SEM). The data were statistically analyzed using one-way analysis of variance (ANOVA) followed by Duncan’s multiple range test (DMRT) using SPSS (28.0 version). The criterion of statistical significance was set at p ≤ 0.05. ## 5. Conclusions The findings of the present study show that LMN prevented behavioral deficits, reduced α-synuclein expression, rescued loss of dopaminergic neurons, mitigated oxidative stress, restored complex-I activity, reduced lipid peroxidation, and downregulated proinflammatory cytokines along with a favorable modulation of apoptotic pathways, including MAPK, mTOR, and Hippo signaling, in ROT-induced PD in rats. Based on these findings, it can be concluded that LMN is capable of protecting dopaminergic neurons and maintaining neuronal integrity attributed to its antioxidant, anti-inflammatory, antiapoptotic, and neurogenesis-enhancing properties. The current study substantiates that LMN may be a promising neuroprotective agent against dopaminergic neurodegeneration, a prominent pathologic feature of PD. Thus, LMN being a molecule of natural origin means that it could be further suggested for nutraceutical as well as pharmaceutical development and subjected to regulatory toxicology and human studies. ## References 1. Meireles J., Massano J.. Cognitive impairment and dementia in Parkinson’s disease: Clinical features, diagnosis, and management. Front. Neurol. (2012) 3 88. DOI: 10.3389/fneur.2012.00088 2. Obeso J.A.. Modeling clinical features of neurodegeneration. Nat. Med. (2010) 16 1372. DOI: 10.1038/nm1210-1372b 3. Gibson G.E., Huang H.M.. Mitochondrial enzymes and endoplasmic reticulum calcium stores as targets of oxidative stress in neurodegenerative diseases. J. Bioenerg. Biomembr. (2004) 36 335-340. DOI: 10.1023/B:JOBB.0000041764.45552.f3 4. Liddelow S.A., Guttenplan K.A., Clarke L.E., Bennett F.C., Bohlen C.J., Schirmer L., Bennett M.L., Münch A.E., Chung W.S., Peterson T.C.. Neurotoxic reactive astrocytes are induced by activated microglia. Nature (2017) 541 481-487. DOI: 10.1038/nature21029 5. Ramesh G., MacLean A.G., Philipp M.T.. Cytokines and chemokines at the crossroads of neuroinflammation, neurodegeneration, and neuropathic pain. Mediat. Inflamm. (2013) 2013 480739. DOI: 10.1155/2013/480739 6. Luo Y., Hoffer A., Hoffer B., Qi X.. Mitochondria: A therapeutic target for Parkinson’s disease?. Int. J. Mol. Sci. (2015) 16 20704-20730. DOI: 10.3390/ijms160920704 7. Zhang X., Du L., Zhang W., Yang Y., Zhou Q., Du G.. Therapeutic effects of baicalein on rotenone-induced Parkinson’s disease through protecting mitochondrial function and biogenesis. Sci. Rep. (2017) 7 9968. DOI: 10.1038/s41598-017-07442-y 8. Xilouri M., Brekk O.R., Stefanis L.. Autophagy and Alpha-Synuclein: Relevance to Parkinson’s Disease and Related Synucleopathies. Mov. Disord. (2016) 31 178-192. DOI: 10.1002/mds.26477 9. He J., Zhu G., Wang G., Zhang F.. Oxidative Stress and Neuroinflammation Potentiate Each Other to Promote Progression of Dopamine Neurodegeneration. Oxid. Med. Cell Longev. (2020) 2020 6137521. DOI: 10.1155/2020/6137521 10. Cookson M.R.. alpha-Synuclein and neuronal cell death. Mol. Neurodegener. (2009) 4 9. DOI: 10.1186/1750-1326-4-9 11. Anusha C., Sumathi T., Joseph L.D.. Protective role of apigenin on rotenone induced rat model of Parkinson’s disease: Suppression of neuroinflammation and oxidative stress mediated apoptosis. Chem. Biol. Interact. (2017) 269 67-79. DOI: 10.1016/j.cbi.2017.03.016 12. Dauncey M.J.. Genomic and epigenomic insights into nutrition and brain disorders. Nutrients (2013) 5 887-914. DOI: 10.3390/nu5030887 13. Amaral de Brito A.P., Galvão de Melo I., El-Bachá R.S., Guedes R.C.A.. Valeriana officinalis Counteracts Rotenone Effects on Spreading Depression in the Rat Brain in vivo and Protects Against Rotenone Cytotoxicity Toward Rat Glioma C6 Cells in vitro. Front. Neurosci. (2020) 14 759. DOI: 10.3389/fnins.2020.00759 14. Johnson M.E., Bobrovskaya L.. An update on the rotenone models of Parkinson’s disease: Their ability to reproduce the features of clinical disease and model gene–environment interactions. Neurotoxicology (2015) 46 101-116. DOI: 10.1016/j.neuro.2014.12.002 15. Zhang N., Dou D., Ran X., Kang T.. Neuroprotective effect of arctigenin against neuroinflammation and oxidative stress induced by rotenone. RSC Adv. (2018) 8 2280-2292. DOI: 10.1039/C7RA10906G 16. Shin M., Liu Q.F., Choi B., Shin C., Lee B., Yuan C., Song Y.J., Yun H.S., Lee I.S., Koo B.S.. Neuroprotective Effects of Limonene (+) against Aβ42-Induced Neurotoxicity in a Drosophila Model of Alzheimer’s Disease. Biol. Pharm. Bull (2020) 43 409-417. DOI: 10.1248/bpb.b19-00495 17. Wang X., Li G., Shen W.. Protective effects of D-Limonene against transient cerebral ischemia in stroke-prone spontaneously hypertensive rats. Exp. Ther. Med. (2018) 15 699-706. DOI: 10.3892/etm.2017.5509 18. Sadeghimanesh A., Khalaji-Pirbalouty V., Lorigooini Z., Rafieian-Kopaei M., Torki A., Rabiei Z.. Phytochemical and neuroprotective evaluation of Citrus aurantium essential oil on cerebral ischemia and reperfusion. Bangladesh J. Pharmacol. (2018) 13 353-361. DOI: 10.3329/bjp.v13i4.37408 19. Eddin L.B., Jha N.K., Meeran M.F.N., Kesari K.K., Beiram R., Ojha S.. Neuroprotective Potential of Limonene and Limonene Containing Natural Products. Molecules (2021) 26. DOI: 10.3390/molecules26154535 20. Yu X., Lin H., Wang Y., Lv W., Zhang S., Qian Y., Deng X., Feng N., Yu H., Qian B.. d-limonene exhibits antitumor activity by inducing autophagy and apoptosis in lung cancer. Onco Targets (2018) 11 1833-1847. DOI: 10.2147/OTT.S155716 21. Kim Y.W., Kim M.J., Chung B.Y., Bang D.Y., Lim S.K., Choi S.M., Lim D.S., Cho M.C., Yoon K., Kim H.S.. Safety evaluation and risk assessment of d-Limonene. J. Toxicol. Env. Health B Crit. Rev. (2013) 16 17-38. DOI: 10.1080/10937404.2013.769418 22. Almeida A.A.C., Ferreira J.R.O., de Carvalho R.B.F., Rizzo M.D.S., Lopes L.D.S., Dittz D., Castro E.S.J.M., Ferreira P.M.P.. Non-clinical toxicity of (+)-limonene epoxide and its physio-pharmacological properties on neurological disorders. Naunyn Schmiedebergs Arch. Pharm. (2020) 393 2301-2314. DOI: 10.1007/s00210-020-01943-w 23. Chen H., Chan K.K., Budd T.. Pharmacokinetics of d-limonene in the rat by GC-MS assay. J. Pharm. Biomed Anal. (1998) 17 631-640. DOI: 10.1016/S0731-7085(97)00243-4 24. Igimi H., Nishimura M., Kodama R., Ide H.. Studies on the metabolism of d-limonene (p-mentha-1,8-diene). I. The absorption, distribution and excretion of d-limonene in rats. Xenobiotica (1974) 4 77-84. DOI: 10.3109/00498257409049347 25. Vigushin D.M., Poon G.K., Boddy A., English J., Halbert G.W., Pagonis C., Jarman M., Coombes R.C.. Phase I and pharmacokinetic study of D-limonene in patients with advanced cancer. Cancer Research Campaign Phase I/II Clinical Trials Committee. Cancer Chemother. Pharm. (1998) 42 111-117. DOI: 10.1007/s002800050793 26. Magrinelli F., Picelli A., Tocco P., Federico A., Roncari L., Smania N., Zanette G., Tamburin S.. Pathophysiology of Motor Dysfunction in Parkinson’s Disease as the Rationale for Drug Treatment and Rehabilitation. Park. Dis. (2016) 2016 9832839. DOI: 10.1155/2016/9832839 27. Nijkrake M.J., Keus S.H., Quist-Anholts G.W., Overeem S., De Roode M.H., Lindeboom R., Mulleners W., Bloem B.R., Munneke M.. Evaluation of a Patient-Specific Index as an outcome measure for physiotherapy in Parkinson’s disease. Eur. J. Phys. Rehabil. Med. (2009) 45 507-512. PMID: 20032909 28. von Wrangel C., Schwabe K., John N., Krauss J.K., Alam M.. The rotenone-induced rat model of Parkinson’s disease: Behavioral and electrophysiological findings. Behav. Brain Res. (2015) 279 52-61. DOI: 10.1016/j.bbr.2014.11.002 29. Betarbet R., Sherer T.B., MacKenzie G., Garcia-Osuna M., Panov A.V., Greenamyre J.T.. Chronic systemic pesticide exposure reproduces features of Parkinson’s disease. Nat. Neurosci. (2000) 3 1301-1306. DOI: 10.1038/81834 30. Nagatsu T., Nakashima A., Ichinose H., Kobayashi K.. Human tyrosine hydroxylase in Parkinson’s disease and in related disorders. J. Neural Transm (Vienna) (2019) 126 397-409. DOI: 10.1007/s00702-018-1903-3 31. Mohammadi A., Amooeian V.G., Rashidi E.. Dysfunction in Brain-Derived Neurotrophic Factor Signaling Pathway and Susceptibility to Schizophrenia, Parkinson’s and Alzheimer’s Diseases. Curr. Gene (2018) 18 45-63. DOI: 10.2174/1566523218666180302163029 32. Wang Y., Liu H., Zhang B.S., Soares J.C., Zhang X.Y.. Low BDNF is associated with cognitive impairments in patients with Parkinson’s disease. Park. Relat. Disord. (2016) 29 66-71. DOI: 10.1016/j.parkreldis.2016.05.023 33. Fink A.L.. The aggregation and fibrillation of alpha-synuclein. Acc. Chem. Res. (2006) 39 628-634. DOI: 10.1021/ar050073t 34. Sharma L.K., Lu J., Bai Y.. Mitochondrial respiratory complex I: Structure, function and implication in human diseases. Curr. Med. Chem. (2009) 16 1266-1277. DOI: 10.2174/092986709787846578 35. Higgins G.C., Beart P.M., Shin Y.S., Chen M.J., Cheung N.S., Nagley P.. Oxidative stress: Emerging mitochondrial and cellular themes and variations in neuronal injury. J. Alzheimers Dis. (2010) 20 S453-S473. DOI: 10.3233/JAD-2010-100321 36. Martinez T.N., Greenamyre J.T.. Toxin models of mitochondrial dysfunction in Parkinson’s disease. Antioxid. Redox Signal (2012) 16 920-934. DOI: 10.1089/ars.2011.4033 37. Jenner P., Olanow C.W.. Oxidative stress and the pathogenesis of Parkinson’s disease. Neurology (1996) 47 S161-S170. DOI: 10.1212/WNL.47.6_Suppl_3.161S 38. Litteljohn D., Mangano E., Clarke M., Bobyn J., Moloney K., Hayley S.. Inflammatory mechanisms of neurodegeneration in toxin-based models of Parkinson’s disease. Park. Dis. (2011) 2011 713517. DOI: 10.4061/2011/713517 39. Indo H.P., Yen H.C., Nakanishi I., Matsumoto K., Tamura M., Nagano Y., Matsui H., Gusev O., Cornette R., Okuda T.. A mitochondrial superoxide theory for oxidative stress diseases and aging. J. Clin. Biochem. Nutr. (2015) 56 1-7. DOI: 10.3164/jcbn.14-42 40. Roberto D., Micucci P., Sebastian T., Graciela F., Anesini C.. Antioxidant activity of limonene on normal murine lymphocytes: Relation to H2O2 modulation and cell proliferation. Basic Clin. Pharm. Toxicol. (2010) 106 38-44. DOI: 10.1111/j.1742-7843.2009.00467.x 41. Shah B., Mehta A.. In vitro evaluation of antioxidant activity of D-Limonene. Asian J. Pharm. Pharmacol. (2018) 4 883-887. DOI: 10.31024/ajpp.2018.4.6.25 42. Wang S., Chu C.H., Stewart T., Ginghina C., Wang Y., Nie H., Guo M., Wilson B., Hong J.S., Zhang J.. α-Synuclein, a chemoattractant, directs microglial migration via H2O2-dependent Lyn phosphorylation. Proc. Natl. Acad. Sci. USA (2015) 112 E1926-E1935. DOI: 10.1073/pnas.1417883112 43. Le W., Wu J., Tang Y.. Protective Microglia and Their Regulation in Parkinson’s Disease. Front. Mol. Neurosci. (2016) 9 89. DOI: 10.3389/fnmol.2016.00089 44. Lorigooini Z., Boroujeni S.N., Sayyadi-Shahraki M., Rahimi-Madiseh M., Bijad E., Amini-Khoei H.. Limonene through Attenuation of Neuroinflammation and Nitrite Level Exerts Antidepressant-Like Effect on Mouse Model of Maternal Separation Stress. Behav. Neurol. (2021) 2021 8817309. DOI: 10.1155/2021/8817309 45. He J., Zhong W., Zhang M., Zhang R., Hu W.. P38 Mitogen-activated Protein Kinase and Parkinson’s Disease. Transl. Neurosci. (2018) 9 147-153. DOI: 10.1515/tnsci-2018-0022 46. Corrêa S.A., Eales K.L.. The Role of p38 MAPK and Its Substrates in Neuronal Plasticity and Neurodegenerative Disease. J. Signal Transduct. (2012) 2012 649079. DOI: 10.1155/2012/649079 47. Karunakaran S., Ravindranath V.. Activation of p38 MAPK in the substantia nigra leads to nuclear translocation of NF-κB in MPTP-treated mice: Implication in Parkinson’s disease. J. Neurochem. (2009) 109 1791-1799. DOI: 10.1111/j.1471-4159.2009.06112.x 48. Tak P.P., Firestein G.S.. NF-κB: A key role in inflammatory diseases. J. Clin. Investig. (2001) 107 7-11. DOI: 10.1172/JCI11830 49. Erekat N.S., Stoker T.B., Greenland J.C.. Apoptosis and its Role in Parkinson’s Disease. Parkinson’s Disease: Pathogenesis and Clinical Aspects (2018) 50. Zhou Q., Liu C., Liu W., Zhang H., Zhang R., Liu J., Zhang J., Xu C., Liu L., Huang S.. Rotenone induction of hydrogen peroxide inhibits mTOR-mediated S6K1 and 4E-BP1/eIF4E pathways, leading to neuronal apoptosis. Toxicol. Sci. (2015) 143 81-96. DOI: 10.1093/toxsci/kfu211 51. Choi K.C., Kim S.H., Ha J.Y., Kim S.T., Son J.H.. A novel mTOR activating protein protects dopamine neurons against oxidative stress by repressing autophagy related cell death. J. Neurochem. (2010) 112 366-376. DOI: 10.1111/j.1471-4159.2009.06463.x 52. Mercado G., Castillo V., Soto P., Sidhu A.. ER stress and Parkinson’s disease: Pathological inputs that converge into the secretory pathway. Brain Res. (2016) 1648 626-632. DOI: 10.1016/j.brainres.2016.04.042 53. Ahn E.H., Kang S.S., Qi Q., Liu X., Ye K.. Netrin1 deficiency activates MST1 via UNC5B receptor, promoting dopaminergic apoptosis in Parkinson’s disease. Proc. Natl. Acad. Sci. USA (2020) 117 24503-24513. DOI: 10.1073/pnas.2004087117 54. Sahu M.R., Mondal A.C.. The emerging role of Hippo signaling in neurodegeneration. J. Neurosci. Res. (2020) 98 796-814. DOI: 10.1002/jnr.24551 55. Silva R.M., Ries V., Oo T.F., Yarygina O., Jackson-Lewis V., Ryu E.J., Lu P.D., Marciniak S.J., Ron D., Przedborski S.. CHOP/GADD153 is a mediator of apoptotic death in substantia nigra dopamine neurons in an in vivo neurotoxin model of parkinsonism. J. Neurochem. (2005) 95 974-986. DOI: 10.1111/j.1471-4159.2005.03428.x 56. Liu W., Wu J., Xiao L., Bai Y., Qu A., Zheng Z., Yuan Z.. Regulation of neuronal cell death by c-Abl-Hippo/MST2 signaling pathway. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0036562 57. de Souza M.C., Vieira A.J., Beserra F.P., Pellizzon C.H., Nóbrega R.H., Rozza A.L.. Gastroprotective effect of limonene in rats: Influence on oxidative stress, inflammation and gene expression. Phytomedicine (2019) 53 37-42. DOI: 10.1016/j.phymed.2018.09.027 58. Yu L., Yan J., Sun Z.. D-limonene exhibits anti-inflammatory and antioxidant properties in an ulcerative colitis rat model via regulation of iNOS, COX-2, PGE2 and ERK signaling pathways. Mol. Med. Rep. (2017) 15 2339-2346. DOI: 10.3892/mmr.2017.6241 59. Chi G., Wei M., Xie X., Soromou L.W., Liu F., Zhao S.. Suppression of MAPK and NF-κB pathways by limonene contributes to attenuation of lipopolysaccharide-induced inflammatory responses in acute lung injury. Inflammation (2013) 36 501-511. DOI: 10.1007/s10753-012-9571-1 60. Younis N.S.. D-Limonene mitigate myocardial injury in rats through MAPK/ERK/NF-κB pathway inhibition. Korean J. Physiol. Pharmacol. (2020) 24 259-266. DOI: 10.4196/kjpp.2020.24.3.259 61. Pereira E.W.M., Heimfarth L., Santos T.K., Passos F.R.S., Siqueira-Lima P., Scotti L., Scotti M.T., Almeida J., Campos A.R., Coutinho H.D.M.. Limonene, a citrus monoterpene, non-complexed and complexed with hydroxypropyl-β-cyclodextrin attenuates acute and chronic orofacial nociception in rodents: Evidence for involvement of the PKA and PKC pathway. Phytomedicine (2022) 96 153893. DOI: 10.1016/j.phymed.2021.153893 62. Bacanlı M., Anlar H.G., Aydın S., Çal T., Arı N., Ündeğer Bucurgat Ü., Başaran A.A., Başaran N.. d-limonene ameliorates diabetes and its complications in streptozotocin-induced diabetic rats. Food Chem. Toxicol. (2017) 110 434-442. DOI: 10.1016/j.fct.2017.09.020 63. Zhou W., Yoshioka M., Yokogoshi H.. Sub-Chronic Effects of <i>s</i>-Limonene on Brain Neurotransmitter Levels and Behavior of Rats. J. Nutr. Sci. Vitaminol. (2009) 55 367-373. DOI: 10.3177/jnsv.55.367 64. Urbach Y.K., Bode F.J., Nguyen H.P., Riess O., von Hörsten S.. Neurobehavioral tests in rat models of degenerative brain diseases. Methods Mol. Biol. (2010) 597 333-356. DOI: 10.1007/978-1-60327-389-3_24
--- title: Integration of Transcriptomics and Non-Targeted Metabolomics Reveals the Underlying Mechanism of Skeletal Muscle Development in Duck during Embryonic Stage authors: - Zhigang Hu - Xiaolin Liu journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049352 doi: 10.3390/ijms24065214 license: CC BY 4.0 --- # Integration of Transcriptomics and Non-Targeted Metabolomics Reveals the Underlying Mechanism of Skeletal Muscle Development in Duck during Embryonic Stage ## Abstract Skeletal muscle is an important economic trait in duck breeding; however, little is known about the molecular mechanisms of its embryonic development. Here, the transcriptomes and metabolomes of breast muscle of Pekin duck from 15 (E15_BM), 21 (E21_BM), and 27 (E27_BM) days of incubation were compared and analyzed. The metabolome results showed that the differentially accumulated metabolites (DAMs), including the up-regulated metabolites, l-glutamic acid, n-acetyl-1-aspartylglutamic acid, l-2-aminoadipic acid, 3-hydroxybutyric acid, bilirubin, and the significantly down-regulated metabolites, palmitic acid, 4-guanidinobutanoate, myristic acid, 3-dehydroxycarnitine, and s-adenosylmethioninamine, were mainly enriched in metabolic pathways, biosynthesis of secondary metabolites, biosynthesis of cofactors, protein digestion and absorption, and histidine metabolism, suggesting that these pathways may play important roles in the muscle development of duck during the embryonic stage. Moreover, a total of 2142 (1552 up-regulated and 590 down-regulated), 4873 (3810 up-regulated and 1063 down-regulated), and 2401 (1606 up-regulated and 795 down-regulated) DEGs were identified from E15_BM vs. E21_BM, E15_BM vs. E27_BM and E21_BM vs. E27_BM in the transcriptome, respectively. The significantly enriched GO terms from biological processes were positive regulation of cell proliferation, regulation of cell cycle, actin filament organization, and regulation of actin cytoskeleton organization, which were associated with muscle or cell growth and development. Seven significant pathways, highly enriched by FYN, PTK2, PXN, CRK, CRKL, PAK, RHOA, ROCK, INSR, PDPK1, and ARHGEF, were focal adhesion, regulation of actin cytoskeleton, wnt signaling pathway, insulin signaling pathway, extracellular matrix (ECM)-receptor interaction, cell cycle, and adherens junction, which participated in regulating the development of skeletal muscle in Pekin duck during the embryonic stage. KEGG pathway analysis of the integrated transcriptome and metabolome indicated that the pathways, including arginine and proline metabolism, protein digestion and absorption, and histidine metabolism, were involved in regulating skeletal muscle development in embryonic Pekin duck. These findings suggested that the candidate genes and metabolites involved in crucial biological pathways may regulate muscle development in the Pekin duck at the embryonic stage, and increased our understanding of the molecular mechanisms underlying the avian muscle development. ## 1. Introduction Skeletal muscle is mainly composed of muscle cells, which performs functions through contraction and relaxation. Muscle growth and development, regulated by transcription and post-transcription and pathways, is a complex process [1]. Skeletal muscle development can be divided into three stages. The first stage is the formation of muscle fibers in the embryonic stage, including [1] the proliferation and differentiation of myoblast precursors, [2] the fusion of myoblasts into multinucleated myotubes, and [3] the formation of mature muscle fibers. The second stage is the development of myofibers in the fetus, which is mainly the change in various fiber types. The third stage is muscle regeneration in adulthood, mainly the increase in myofiber diameter and length and the repair of damaged muscle fibers [2,3]. Among them, the first stage is very important as the result of the embryonic stage determines the number of muscle fibers [4]. Skeletal muscle synthesis in avian embryos can be divided into two processes: primary and secondary muscle fiber formation. A small number of embryonic myoblasts begin to fuse, and the first multinucleated myofiber in vertebrate embryos is formed. Once the scaffold of primary myofibers is established, secondary myofibers can be formed through the fusion between embryonic myoblasts or with primary muscle fibers. Generally, the primary muscle fibers are similar to the slow muscle fibers of adult animals, while the secondary muscle fibers show the characteristics of fast muscle fibers [5]. Some studies have shown that the primary muscle fibers of chicken form from about E6 (the 6th day of incubation), and the secondary myofibers begin to differentiate from E12 to E16 (the 12th to 16th day of incubation) [6]. The difference in myogenesis between mammalian and avian embryos is that the development of avian embryos does not depend on the maternal uterus, but only on the nutrition provided by the yolk. Therefore, it is easy and meaningful to study the differences in regulatory genes and metabolites of skeletal muscle development in birds during different incubation periods. Transcriptomics is a method to study gene expression and transcriptional regulation at the RNA level of specific cell types, tissues, or organisms [7]. In recent years, RNA-Seq has been widely used in the study of animal transcriptome [8,9]. Compared with other gene expression profiling methods, RNA-Seq has the advantages in detecting mRNA expression in different tissues or at different stages of development, helping to reveal new genes, splice variants, and regulatory pathways [10,11,12,13]. Various metabolites and related metabolic pathways in the complex regulatory network were identified by detecting endogenous low-molecular-weight metabolites (molecular weight within 1000 Da) in cells, tissues, organs, or organisms before and after treatment using liquid chromatography/mass spectrometry (LC-MS), gas chromatography/mass spectrometry (GC-MS), and nuclear magnetic resonance (NMR). Metabolomics data identify changes in bioactive compounds during animal development, while RNA-seq data identify genes that regulate metabolic changes. Spurred by a massive advance in technology, the molecular mechanisms of organisms (genome, transcriptome, proteome, and metabolome) were systematically characterized quantitatively by multi-omics analysis [14]. The metabonomics analysis results provide different metabolites, which can make up for the lack of data in transcriptomics analysis, and the life activities and molecular regulation mechanisms can be explored from a multi-level perspective through the combined analysis of transcriptomics and metabonomics [15]. To date, there have been few studies to explore the potential key factors involved in the skeletal muscle development of embryonic duck using transcriptomics and metabolomics techniques. Our previous research carried out transcriptome sequencing on the skeletal muscle of Pekin duck in the embryonic period. The differentially expressed genes such as MYL4 and IGF2BP1 and the regulatory pathways such as focal adhesion and ECM-receptor interaction play a crucial role in the development of duck muscle [16]. Liu et al. found that the metabolism of duck meat changes with age (27, 50, 170, and 500 days of age) by NMR, that is, lactate and anserine increased with age, fumarate, betaine, taurine, inosine, and alkyl-substituted free amino acids decreased [17]. Zhou et al. conducted a comparative analysis of multi-omics data (genome, transcriptome, and metabolome) from the breast meat of Lianchen white ducks (LW) and Mianyang Shellducks (MS) at 300 days of age. The results showed that LW ducks had unique breed-specific genetic characteristics, including many differentially expressed mutant genes related to amino acid metabolism and transport activities. Moreover, the concentration of L-arginine, L-ornithine, and L-lysine in LW duck meat was significantly higher than that in MS duck muscle. In addition, guanosine monophosphate (GMP) was significantly higher in LW muscle, while L-aspartic acid was significantly enriched in MS duck meat [18]. A combination of gene function and metabolites are useful for the comprehensive study of the developmental mechanism of Pekin duck skeletal muscle during the embryonic stage. In this study, the dynamic metabolomic and transcriptomic profiles from breast muscle of Pekin duck at embryonic days 15, 21, and 27 were comparatively investigated, and some potential metabolites and the corresponding differentially expressed genes at the molecular levels were identified. This study revealed crucial metabolic pathways and metabolites of muscle development in Pekin duck at embryonic stages, providing important insights into the mechanisms underlying the muscle development in avian. ## 2.1.1. Overview of the Metabolomic Profiling During LC-MS/MS analysis, one quality control (QC) sample was injected into every 10 samples to monitor the repeatability of the analysis process. Therefore, the repeatability of metabolite extraction and detection can be judged by overlapping analysis of total ion chromatogram (TIC) in different QC samples. The TIC results show that the retention time and peak intensities were consistent, which indicated that the signal was stable when the same sample was detected at different times (Figure 1). In this study, the sample metabolites were analyzed by PCA to evaluate the reliability of the data, and preliminarily understand the overall metabolic difference and the degree of variation between the samples in each group. PCA analysis of samples from each group showed an obvious separation and formed a cluster, indicating that there are metabolic markers that can be classified (Figure 2A,B). Moreover, the results of Pearson correlation coefficient confirmed the high repeatability of samples within the group and ensured the reliability of screening for differential metabolites (Figure 2C,D). The overall differences in metabolites in duck muscle samples at three time points were observed by using partial least squares discriminant analyses (PLS-DA), and the three comparison groups have better separation in positive and negative ion modes (Figure S1). The PLS-DA model of each comparison group was established and then evaluated. R2 (cum) represents the interpretation ability of the model and Q2 (cum) represents the prediction ability of the model. When R2Y (cum) > 0.50 and Q2 (cum) > 0.50, it indicates that the model has high stability, good interpretation and prediction ability. Then, a permutation test of the PLS-DA model was performed (200 test times), and Q2 < 0 indicated that the model did not have overfitting. In this study, Q2 was −0.08 and −0.02, respectively, indicating that differential metabolites can be screened and analyzed according to the model (Figure S2). ## 2.1.2. Identification of Differentially Accumulated Metabolites After qualitative and quantitative analysis of the detected metabolites, differentially accumulated metabolites were found in each group based on fold change ≥2 or ≤$\frac{1}{2}$, p value < 0.05 and VIP ≥ 1. Hierarchical cluster analysis was performed to evaluate the DAM accumulation patterns (Figure 3A–C). There were 599, 584, 388 and 399, 349, 299 DAMs were identified in positive ion mode (pos) and negative ion mode (neg) of three comparison groups, respectively (Figure 3D–F, Table S1). For E15_BM vs. E21_BM, 243 (pos) and 183 (neg) were up-regulated, and 356 (pos) and 216 (neg) were down-regulated. Of the DAMs identified between E15_BM and E27_BM, 219 (pos) and 135 (neg), and 365 (pos) and 214 (neg) metabolites were up-regulated and down-regulated, respectively. Of the number of metabolites differentially accumulated in E21_BM compared to E27_BM, 156 (pos) and 134 (neg) metabolites were upregulated, and 232 (pos) and 165 (neg) metabolites were downregulated, respectively (Figure 4). The top 10 up- and down-regulated DAMs from three comparison groups were shown in Table S2. Of all DAMs, “Organic acids and derivatives”, “Lipids and lipid-like molecules”, “Organoheterocyclic compound”, “Organic oxygen compounds”, “Nucleosides, nucleotides, and analogues”, and “Benzenoids” accounted for a large proportion based on the HMDB database (Figure 5). ## 2.1.3. KEGG Analysis of Differentially Accumulated Metabolites As a useful approach during the investigation of various integrated metabolic pathways, a KEGG pathway analysis of the DAMs among three comparison groups was performed; most of the DAMs were enriched in metabolic pathways, biosynthesis of secondary metabolites, biosynthesis of cofactors, protein digestion and absorption, and histidine metabolism, suggesting that these metabolic pathways may play important roles in the muscle development of duck during the embryonic stage (Figure 6). The statistics of pathway enrichment analysis for the differential metabolites was also carried out, and l-glutamic acid was the most significantly up-regulated metabolite, followed by n-acetyl-1-aspartylglutamic acid, l-2-aminoadipic acid, 3-hydroxybutyric acid, and bilirubin. In contrast, the most significantly down-regulated metabolites were palmitic acid, 4-guanidinobutanoate, myristic acid, 3-dehydroxycarnitine, and s-adenosylmethioninamine, respectively, indicating their important roles in the regulation of duck skeletal muscle development. ## 2.2.1. Overview of Transcriptome To identify potential candidate genes affecting skeletal muscle in ducks during embryonic development, the gene expression profiles of breast muscles from Pekin ducks were examined using RNA-Seq. RNA was prepared from three breast muscles at different developmental stages (E15, E21 and E27), and nine cDNA libraries were then constructed. After sequencing and filtering, there were over 62 million filtered reads; the total mapped rate was 87.79~$88.38\%$, and the uniquely mapped rate was 83.71~$84.92\%$. All Q20 values were > $99.9\%$, and the Q30 value was up to $97.3\%$ (Table 1). ## 2.2.2. Identification of Differentially Expressed Genes The correlation analysis based on the gene expression profiles revealed that the correlations among two samples per group both were greater than 0.94, which indicated that the biological replicates were reliable in this study (Figure 7A). In concert with this, the principal component analysis (PCA) showed a distinguishable distribution in each group, and gene expression clusters of one group were clearly departed from the other two groups (Figure 7B). Similar results were also found in the heat map, that is, the expression of genes in each sample group was different (Figure 7C). A total of 13,383 DEGs were found in the three comparison groups (E15_BM vs. E21_BM, E15_BM vs. E27_BM and E21_BM vs. E27_BM). Among 2142 DEGs in the breast muscle of Pekin duck from E15_BM vs. E21_BM group, 1552 were up-regulated genes and 590 were down-regulated genes, while among 4873 DEGs in E15 vs. E27 group, the number of up-regulated genes was 3810 and the down-regulated genes was 1063. In addition, there were 2401 DEGs from E21_BM vs. E27_BM, including 1606 up-regulated genes and 795 down-regulated genes (Figure 7D). In breast muscle of Pekin duck, 564 DEGs were co-expressed across three developmental time points during the embryonic stage (Figure 7E). The top ten up- and down-regulated genes in each comparison group were listed in Table 2. ## 2.2.3. Go Ontology and KEGG Pathway Analysis These DEGs were categorized into three main GO categories including biological process, cellular component, and molecular function. In total, there were 456, 508, and 435 significantly enriched GO terms ($p \leq 0.05$) identified in E15_BM vs. E21_BM, E15_BM vs. E27_BM, and E21_BM vs. E27_BM, respectively. The significantly enriched terms from biological process in the comparisons were associated with muscle or cell growth and development, such as positive regulation of cell proliferation, regulation of cell cycle, actin filament organization, and regulation of actin cytoskeleton organization (Figure 8, Table 3). One hundred and thirty DEGs among the three comparison groups were associated with these terms, with some of these genes having been reported to be associated with growth, such as MYOG, SMYD1, MYH9, FGF10, TNNT2, IGFBP3, MYOD1, and MEF2C. Then, the DEGs were annotated using KEGG to identify enriched pathways, and seven significantly enriched pathways related to growth and development of breast muscle were identified, including focal adhesion, regulation of actin cytoskeleton, wnt signaling pathway, insulin signaling pathway, extracellular matrix (ECM)-receptor interaction, cell cycle, and adherens junction (Table 4). Eleven genes, FYN, PTK2, PXN, CRK, CRKL, PAK, RHOA, ROCK, INSR, PDPK1, and ARHGEF, were highly enriched in GO terms and significantly up-regulated or down-regulated expressed in KEGG pathways to regulate the development of skeletal muscle in Pekin duck during the embryonic stage (Figure 9). ## 2.3. Integrated Analysis of Transcriptomics and Metabolomics The KEGG pathway of metabolomics and transcriptomics were integrated and analyzed. By interactively comparing the metabolomic and transcriptomic data, the potential metabolites and the corresponding differentially expressed genes at the molecular and biochemical levels were identified, and 25 pathways were found in 3 comparison groups (Table S3). It is noteworthy that pathways involved in arginine and proline metabolism, protein digestion and absorption, and histidine metabolism were enriched significantly in both transcriptomic and metabolomic data (Figure 10, Figure 11 and Figure 12), which will provide a theoretical basis for revealing the genetic regulation mechanism of duck muscle development. ## 2.4. RT-qPCR Validation of the RNA-Seq Data Twelve DEGs from the six significantly enriched pathways (related to growth and development of breast muscle) were randomly selected to validate the results of RNA-Seq. RT-qPCR was performed on the same RNA samples as used for RNA-Seq. The results showed a similar regulated trend in the expression of these genes, which confirmed the validity of the data from RNA-Seq (Figure 13). ## 3. Discussion Meat quality is the key factor to determine the economic value of modern animal husbandry, and it is also an important reference index for duck breeding. Skeletal muscle development is regulated by DEGs, and the concentration and proportion of endogenous metabolites change with the extension of development time, showing certain differences. The identification of key genes and metabolites for skeletal muscle development in embryonic duck is helpful to explain the molecular mechanism of poultry muscle development. Moreover, The DAMs and DEGs can be used as molecular markers and candidate genes to provide data support for breeding new duck varieties (strains) based on molecular breeding methods, and also provide a method reference for future research on the genetic mechanism of important economic traits in livestock and poultry. Due to the limited evidence provided by single omics and the existence of ambiguous data, the problem cannot be fully explained. Therefore, the combination analysis of multi-omics has been widely used [19,20]. Here, the transcriptomic and metabolomic changes of muscle development in Pekin duck during the embryonic stage were tracked, which can provide a unique opportunity for us to deeply understand the candidate genes and metabolites in the process of avian skeletal muscle development. It has been known that primary myofibers form about E6 of incubation, and secondary muscle fibers begin to differentiate between E12 and E16. Some studies have shown that myofiber size in breast muscle of duck embryos decreased by $55\%$ from E22 to hatching (d 28), which is the occurrence of muscle fiber atrophy [6]. Therefore, from the middle stage of incubation, we collected samples every six days until the day before hatching, that is, the breast muscles at 15, 21, and 27 d of embryonic stage were selected for sequencing analysis. E15 was chosen because it is the time of secondary muscle fiber formation and also the middle stage of duck embryo incubation. E21 was selected because it was the period when the muscle fiber atrophy did not occur in the duck embryo, while E27 was the last period of duck embryo muscle fiber atrophy. ## 3.1. Metabolome Analysis In the present study, the pathways which participate in metabolic pathways, biosynthesis of secondary metabolites, biosynthesis of cofactors, protein digestion and absorption, and histidine metabolism may contribute to the dynamic process of skeletal muscle development in embryonic duck. Specifically, metabolic pathway refers to a series of continuous metabolic reactions, leading to the synthesis or decomposition of certain metabolites, in which decomposition mainly completes the work of obtaining energy and the “raw material” required for body composition, while synthesis mainly accomplishes the utilization of energy storage and the “raw material” to construct the components of the organism [21]. Likewise, protein digestion and absorption are the basic organic matter that constitute cells. It is essential to renew or repair tissues and maintain the nutritional balance in the body [22]. Histidine is a semi-essential amino acid, which is particularly important for the growth of infants and animals, and histidine metabolism also plays an important role in biological metabolism during development [23]. Moreover, a secondary metabolite is a kind of non-essential small-molecule organic compound produced by secondary metabolism for cell life activities or normal biological growth and development [24], and cofactors refer to non-protein compounds that bind to enzymes and are necessary for catalyzing reactions [25]. The biosynthesis of the two is important for the cell and tissue development of organisms. Metabolites are the key regulators and markers of animal growth and development. In short, the metabonomic analysis results of this study strongly suggest that the accumulation of these important metabolites of glutamate, n-acetyl-1-aspartylglutamic acid, l-2-aminoadipic acid, 3-hydroxybutyric acid, and bilirubin may directly contribute to promoting the development of embryonic duck skeletal muscle. Glutamate plays a key role in all transamination reactions in the body and in many other metabolic pathways in different organs (including skeletal muscle) [26], and n-acetyl-1-aspartylglutamic acid also affects muscle development [27]. Sato et al. showed that l-2-aminoadipic acid regulates protein turnover of C2C12 myotube [28], and 3-hydroxybutyric acid possesses a key role in promoting muscle development and maintaining muscle protein balance [29]. Similarly, bilirubin has a cytoprotective effect on various oxidative damages, and it can reduce ectopic lipid deposition in skeletal muscle and liver cells, which is the main key factor in the pathogenesis of diabetes mellitus type 2 [30]. Beyond that, there are some metabolites down-regulated during the development of skeletal muscle in the duck embryo, which also play an important role. Palmitic acid regulates muscle development by negatively affecting myotube diameter, fusion, and metabolism [31]. 4-guanidinobutanoate participates in a variety of metabolic activities and has high physiological activity [32]. Myristic acid can not only improve the secretion of insulin, but also improve the sensitivity of tissues to insulin action [33]. In addition, researchers have shown that both 3-dehydroxycarnitine and s-adenosylmethioninamine have unique metabolic characteristics related to the muscle or cell [34,35]. Therefore, these metabolic pathways and metabolites identified in this study are crucial for the development of skeletal muscle in embryonic duck. ## 3.2. Transcriptome Profiles During muscle development, myoblasts go through the steps of migration, adhesion, elongation, intercellular recognition, alignment, and myoblastic membrane fusion, and finally form myotubes. With the development of high-throughput sequencing, screening genes related to muscle development has attracted increasing attention, and massive amounts of transcriptional data have been produced. In this study, KEGG analysis showed that the DEGs were assigned to more than 17 significant pathways, and the identified genes were mainly enriched in focal adhesion, regulation of actin cytoskeleton, wnt signaling pathway, insulin signaling pathway, ECM-receptor interaction, cell cycle, and adherens junction. The site where integrin and proteoglycan mediated adhesion connects with the actin cytoskeleton is called focal adhesion (FA), which is dynamic multi-protein complexes that connect ECM with the intracellular cytoskeleton [36,37]. The formation and maturation of FA is a key procedure during myoblast differentiation. FAK (focal adhesion kinase), a non-receptor tyrosine kinase, plays a key role in the reorganization of the sarcomere by acting as a scaffold for the recruitment of focal adhesion protein [38], and as part of the mechanism related to the load and fiber type of fully developed muscle tissue, FAK regulates the dynamic formation and turnover of FAs and the molecules that control myofibrillar protein synthesis [39]. Moreover, FAK can also regulate cell proliferation and migration by recruiting additional kinases and inducing complex signal cascades [40]. The interaction between cells and ECM is the key to regulating cell and tissue homeostasis. ECM consists of a complex mixture of structural and functional macromolecules, mainly including collagen, fibronectin, and laminin [41]. Multiprotein complexes, such as FAs and fibrous adhesions, mediate cell-matrix adhesion, connect the ECM to the cytoskeleton, and promote cell-mediated matrix remodeling. ECM not only provides substrates for cell adhesion, but also influences various cellular functions, including cell survival, proliferation, migration, and differentiation, through signaling proteins located in and near the adhesion complex [42,43]. Studies have shown that cell differentiation was directly or indirectly controlled by the interaction between cells and ECM proteins [44]. Actin is a well-known cytoskeletal protein, and the actin cytoskeleton plays a role in development and reproduction [45]. Studies have indicated that cytoskeleton formation participates in cell migration and differentiation [46,47], and actin is closely related to contraction and myoblast differentiation [48]. Many morphological changes in cells are driven at least in part by the remodeling of the actin cytoskeleton [49]. Repetition of the cell cycle orchestrates genome duplication and the subsequent segregation of each genome duplication into new daughter cells, resulting in the proliferation of cells [50]. Cell proliferation and differentiation are closely related to signal pathways that regulate the cell cycle-controlled gene expression during animal development [51]. The regulation of the cell cycle is important not only for cell differentiation during development, but also for morphogenesis [52]. Cells are connected together by cell–cell junctions (adherens junctions, tight junctions, and desmosomes), which are critical to the homeostasis of tissues, especially during embryonic development and tissue maintenance when cells are constantly squeezed and stretched [53,54]. Adherens junction is well-known cell–cell junction structure, and is the attachment site of cadherin adhesion receptor to link the actin cytoskeleton of adjacent cells [55]. It is the site of mechanosensing and signal transmission, and regulates the dynamics of actomyosin, which then generates the force driving morphogenesis [56]. Wnt pathway, also known as the β-catenin pathway, plays a crucial role (multiple developmental events) during embryogenesis in vertebrates, and regulates the homeostatic processes in adulthood [57]. It can enhance the proliferation and differentiation of embryonic cells and other metabolic functions, and is an important regulator of the total number of muscle cells and the ratio of fast muscle cells to slow muscle cells [58,59]. In addition, the pathway also regulates cytoskeletal rearrangement during embryonic development [60], and lipid metabolism, glucose homeostasis, and energy balance [61]. Lu et al. found that the polymorphism of Wnt signaling pathway genes (RHOA, Wnt3A, CHP, RAC1, Wnt1, Wnt9A, MAPK9) was significantly related to chicken carcass traits, indicating that the Wnt signaling pathway played a major role in regulating chicken production traits (carcass characteristics) [62]. Insulin facilitates the entry of glucose into fat and muscle, where it is stored as intracellular triglycerides and glycogen [63], and $75\%$ of insulin mediated glucose uptake and utilization is carried out by skeletal muscle, indicating that the development of skeletal muscle is closely related to insulin [64,65]. Rhoads et al. found that insulin played an important role in skeletal muscle growth by regulating muscle hypertrophy, protein accumulation, and cell activity [66]. The insulin signaling pathway is a mechanism that regulates growth rate in response to nutrient availability [67]. Ma et al. transplanted hSKM (human skeletal myoblasts) into mouse skeletal muscle, and the transcription of multiple genes related to the insulin signaling pathway, and mitochondrial biogenesis and function were altered [68]. Briefly, cytoskeleton formation and ECM-integrin receptor interactions participate in myoblast differentiation [69]. Specifically, myoblast migration depends on the dynamics of the cytoskeleton, which is mainly related to actin and regulatory factors, and elongation, adhesion, and intercellular recognition in myoblasts mainly rely on the interactions between integrins and cytoskeletal proteins. Moreover, the ECM was connected to the actin cytoskeleton by FA [70,71,72]. Likewise, wnt signaling pathway, insulin signaling pathway, cell cycle, and adherens junction have been suggested to regulate the progressions of skeletal muscle transcription, cell proliferation, and differentiation during embryonic development. These pathways found in this study, including FA, regulation of actin cytoskeleton, wnt signaling pathway, insulin signaling pathway, ECM-receptor interaction, cell cycle, and adherens junction, synergistically promote skeletal muscle development in embryonic Pekin ducks. In this study, 11 genes (FYN, PTK2, PXN, CRK, CRKL, PAK, RHOA, ROCK, INSR, PDPK1, and ARHGEF) were also identified, which were likely to be involved in duck muscle development and were either up- or down-regulated with more than several-fold changes. FYN is a member of the Src kinase family with diverse biological functions, including regulation of mitogenic signaling and proliferation and integrin-mediated interactions, as well as cellular growth, survival, adhesion, motility, T-cell receptor signaling, and cytoskeletal remodeling [73]. Yamada et al. have shown that Fyn/STAT3/Vps34 signaling pathway can regulate fiber-type specific macroautophagy and muscle atrophy of mouse skeletal muscle [74]. PTK2 (protein tyrosine kinase-2) encodes FAK, and controls its expression. There is an association between muscle specific force and PTK2 SNPs [75,76]. PXN (paxillin) has two homologues, PXN-1 and PXN-2. PXN-1 may play a regulatory role in the matrix of C. elegans, and PXN-2 is very important for the later steps of the formation of the basement membrane, especially when mechanical adhesion is required between the tissue and the ECM. Lee et al. suggested that PXN-1 may play a role in the attachment of tissues and the guidance of neurons during the development of C. elegans [77]. Gotenstein et al. showed that PXN-2 was crucial to the embryogenesis of the C. elegans and the inhibition of the regeneration of adult axons [78]. CRK, including two splicing variants Crk I and Crk II, regulates signal transduction processes involving growth regulation, cell transformation, cell migration, and cell adhesion [79]. Crk II can be localized with other FA related proteins, such as Src, FAK, and paxillin [80]. Likewise, CRKL (Crk-like gene) participates in many signaling pathways and controls cell morphology, cell movement, cell proliferation, and differentiation [81]. There is a compelling evidence that Crk and Crk-like (Crkl) physically interacting with Dock proteins are required for myoblast fusion in zebrafish [82]. PAK (P21-activaed kinase), a serine/theronine kinase, is important for participants in insulin signaling and glucose homeostasis in muscle, pancreas, liver, and other tissues [83]. It is involved in cell proliferation, apoptosis, metastasis, and cytoskeleton remodeling. Varshney and Dey et al. showed that PAK2 can regulate glucose uptake and insulin sensitivity of neuronal cells [84]. PDPK1 (3-phosphoinositol dependent protein kinase-1) is a member of AGC serine/threonine kinase family, and the PDPK1/AGC kinase signaling pathway is involved in the regulation of cell proliferation, growth, autophagy, and apoptosis related physiological processes [85], as well as promotion of muscle growth [86]. PDPK-1 is closely related to the insulin signaling pathway, which can stimulate the increased catalytic activity of PDPK-1 in a PI3K-dependent manner [87]. RHOA, one of the Rho subfamily members of small GTPases, regulates cell proliferation and motility, and is considered to be the key regulator of actin cytoskeleton dynamics and organization in most cell types [88]. Likewise, ROCK (Rho-associated kinase) can regulate the contraction of stress fibers by regulating the phosphorylation level of myosin light chain [89]. Wozniak et al. believed that the contraction of stress fibers mediated by Rho and ROCK could regulate the formation of FA in vivo, which may regulate downstream signaling pathways and cell behavior [38]. Studies have shown that RhoA and Rock play a unique and independent regulatory role in the process of myogenic differentiation [90]. Likewise, RhoA/ROCK signaling in skeletal muscle also plays an important role [91]. ARHGEF3, also called XPLN, a RhoA/B-specific GEF, can be used as the effect of RhoGEF on RhoA/B, or the inhibitor of mTORC2-Akt signaling, negatively regulating myoblastic differentiation [92]. Moreover, ARHGEF3 may also control skeletal muscle regeneration and strength through autophagy presenting in the ARHGEF3 RhoA/B-ROCK signaling pathway [93]. INSR (insulin receptor) is a central starting point of insulin signaling, and is involved in the glucose homeostasis mechanism, proliferation, and growth of skeletal muscle and fat cells [94]. Interestingly, muscle differentiation is blocked by RhoA/Rho kinase through serine phosphorylation of insulin receptor substrates-1 and -2 [95]. These pathways and genes play an important role and may be potential markers in duck skeletal muscle development, but their mechanisms need further experimental verification. ## 3.3. Integrated Analysis of Transcriptomics and Metabolomics Through the combined analysis of transcriptome and metabolome, 26 significantly enriched pathways were identified that may regulate the muscle development of Pekin duck during the embryonic stage. Among these pathways significantly enriched, three pathways including arginine and proline metabolism, protein digestion and absorption, and histidine metabolism were found to be commonly enriched with DEGs (CKB, AGMAT, SRM, ODC1, GATM, P4HA2, MYADML2, GAMT, NOS2, PYCR3, LAP3, SMOX, P4HA3, ALDH18A1, COL19A1 and UROC1) and DEMs (4-guanidinobutanoate, l-glutamic acid, histamine, l-isoleucine, l-aspartic acid, 4-imidazolone-5-propanoate, s-adenosylmethioninamine, indole, 2-methylbutyric acid, 1-(5-phosphoribosyl)imidazole-4-acetate, 4-(beta-acetylaminoethyl)imidazole and piperidine). It is well known that amino acids play a role in muscle growth; for example, arginine can promote skeletal muscle fiber type transformation from fast-twitch to slow-twitch via Sirt1/AMPK pathway [96], and proline can increase the rates of protein synthesis in the muscle [97]. Histidine is also required for skeletal muscle development [98]. Amino acid metabolism is mainly used to synthesize proteins, polypeptides, and other nitrogenous substances needed by the body, and can also be converted into sugars, lipids, or can re-synthesize some non-essential amino acids. It can also be oxidized into carbon dioxide and water through the circulation of tricarboxylic acid and release energy. Similarly, under the action of protease, protein is eventually decomposed into amino acids and finally absorbed by the intestine through metabolism. ## 4.1. Animals and Sample Collection A total of 120 eggs of Pekin duck were incubated in a standard incubator according to the conventional incubation procedure. Eighteen embryos were randomly picked out from day 15 (E15), day 21 (E21), and day 27 (E27) of the incubation period, and the breast muscles were collected and immediately frozen in liquid nitrogen for RNA and DNA extraction. DNA was extracted according to the phenol-chloroform protocol, and sex identification primers (gCHD, F: 5′TGCAGAAGCA ATATTACAAGT3′; R: 5′AATTCATTATCATCTGGTGG3′) [99] were used to determine the sex of embryos. Because the vast majority of duck farms are laying ducks, and the same gender can also avoid the error of sequencing data, female embryos in this study were selected as the research objects. Animal care, slaughter, and experimental procedures were approved by Institutional Animal Care and Institutional Ethic Committee of Northwest A&F University (ethic code: #$\frac{1201}{2021}$). ## 4.2.1. Metabolites Extraction Six breast muscle samples of female embryos at each embryonic stage were randomly selected for the extraction of metabolites. The 100 mg samples were ground with liquid nitrogen and extracted with 120 μL of precooled $50\%$ methanol. Then, the mixtures were vortexed and mixed well, and incubated at room temperature for 10 min. The extractions were stored overnight at −20 °C to precipitate the protein in the samples. After centrifugation at 4000× g for 20 min, the supernatants were transferred into new 96-well plates. Metabolic samples were stored at −80 °C prior to LC-MS analysis, and the quality control (QC) sample was prepared by mixing an equal aliquot of the supernatants (10 μL) from all of the samples. ## 4.2.2. LC-MS/MS Analysis All chromatographic separations were performed using a Thermo Scientific UltiMate 3000 HPLC (Thermo Scientific, Waltham, MA, USA), equipped with an ACQUITY UPLC BEH C18 column (100 × 2.1 mm, 1.8 µm, Waters, UK) for the reversed phase separation. The auto-sampler temperature was 4 °C, and the flow rate was 0.4 mL/min. The mobile phase consisted of solvent A (water, $0.1\%$ formic acid) and solvent B (Acetonitrile, $0.1\%$ formic acid). The analysis was carried out with elution gradient as follows: 0~0.5 min, $5\%$ B; 0.5~7 min, $5\%$ to $100\%$ B; 7~8 min, $100\%$ B; 8~8.1 min, $100\%$ to $5\%$ B; 8.1~10 min, $5\%$ B. The injection volume for each sample was 4 µL. The metabolites eluted form the column were detected by high-resolution tandem mass spectrometer Q-Exactive (Thermo Scientific, Waltham, MA, USA). The Q-Exactive was operated in both positive and negative ion modes. Precursor spectra (70~1050 m/z) were collected at 70,000 resolution to hit an AGC target of 3e6. The maximum inject time was set to 100 ms. A top 3 configuration to acquire data was set in DDA mode. Fragment spectra were collected at 17,500 resolution to hit an AGC target of 1e5 with a maximum inject time of 80 ms. In order to evaluate the stability of the LC-MS during the whole acquisition, a quality control sample (pool of all samples) was acquired after every 10 samples. ## 4.2.3. Processing and Analysis of Metabolome Data The raw data were transformed into a readable data format (mzXML) using MSConvert software (GUI) of Proteowizard. The peak extraction and quality control were performed using XCMS software and the extracted substances were annotated with ions using CAMERA software. The metabolite identification was carried out using metaX software (primary mass spectrometry information for database matching identification, and secondary mass spectrometry information for matching identification with in-house standard database). Multivariate statistical analysis was applied, including unsupervised principal component analysis (PCA) and supervised partial least square discriminant analysis (PLS-DA). Moreover, univariate analysis was performed, including student’s t-test and fold change analysis. The metabolites with a variable importance in projection (VIP) values ≥ 1 and p-value < 0.05 were used as criteria to discover differentially expressed metabolites (DEMs). Metabolites identifications were carried out based on the metabolites information public database, such as HMDB and mzCloud, and analyses of metabolic pathways were conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG, http://www.kegg.jp/kegg, accessed on 30 November 2022). ## 4.3.1. RNA Extraction and Library Preparation Total RNA was extracted from breast muscle tissues of 3 female embryos from each embryonic stage (randomly selected from the same samples used for metabolome sequencing) using Trizol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s protocol. The RNA purity and concentration were verified by agarose gel electrophoresis and Nanodrop 2000 (Thermo, Waltham, CA, USA). The sample integrities were determined based on Agilent 2100 Bioanalyzer (Agilent Technologies, San Jose, CA, USA) to ensure that the RNA integrity number (RIN) was above 8.0. Nine sequencing libraries were constructed, and their quality was tested. The paired-end sequencing was performed on an Illumina Hiseq 4000 (LC Bio, Hangzhou, China) following the vendor’s recommended protocol, and a paired end read of 150 bp was generated. ## 4.3.2. RNA Sequencing and Data Analysis After sequencing, clean reads were obtained by removing reads containing adapters or poly-N, more than $10\%$ of unknown nucleotides and low-quality reads containing more than $50\%$ of low-quality (Q-value ≤ 10) bases from raw reads. At the same time, quality parameters for filtered data including Q30, GC content, and sequence-duplication level were used for data filtering. All the downstream analyses were based on clean reads with high quality. These clean reads were then mapped to the Anas platyrhynchos genome sequence (https://www.ncbi.nlm.nih.gov/genome/?term=DUCK, accessed on 30 November 2022) and annotated transcripts (https://www.ncbi.nlm.nih.gov/assembly/GCF_003850225.1, accessed on 30 November 2022). Only data with perfect match reads, or one mismatch were further analyzed and annotated based on the reference genome. The Hisat2 tool software were used to map with the reference genome. To compare the expression profiles in different samples, the gene expression levels were normalized by fragments per kilobase per million fragments (FPKM). FPKM represents the number of sequencing fragments contained in each thousand transcriptional sequencing bases per million sequenced bases. Differential expression was analyzed using the edgeR software (v.3.20, accessed on 4 December 2022), and the false discovery rate (FDR) <0.01 and fold change ≥2 were set as the threshold for screening differentially expressed genes (DEGs). Gene Ontology (GO) enrichment analysis was implemented by the GOseq R software package (https://bioconductor.org/packages/release/bioc/html/goseq.html, accessed on 4 December 2022), and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis and functional annotation for DEGs were performed using KOBAS 3.0. ## 4.3.3. Validation of RNA-seq Data In order to confirm the reliability of data, eighteen DEGs were selected to validate the RNA-Seq results. The cDNAs were synthesized from the same RNA samples used for transcriptome sequencing according to the manual of reverse transcription kit (TaKaRa, Dalian, China). The β-actin was used as an internal control, and the list of primers were described in Table 5. The RT-qPCR were performed using TransStart Tip Green qPCR SuperMix (Transgen, Beijing, China) on a EcoRT48 (OSA, London, UK). Three technical replicates were carried out per sample. The relative gene expression levels were calculated using the 2−ΔΔCt method and the results were expressed as mean ± SD of at least three independent biological replicates. The difference was analyzed using one-way analysis of variance (ANOVA) followed by Dunnet’s t-test and Tukey’s test, and $p \leq 0.05$ was considered significant difference and $p \leq 0.01$ was considered extremely significant difference. ## 4.4. Integrative Analysis of Metabolomics and Transcriptomics For the integrated analyses of transcriptome and metabolome, all DEGs and DAMs were simultaneously mapped into the KEGG database to obtain their common pathway information, and clarify the main biochemical pathways and signal transduction pathways DEGs and DEMs are involved in. ## 5. Conclusions In summary, many candidate genes and metabolites involved in crucial biological pathways underlying muscle development in Pekin duck during the embryonic stage were successfully identified according to transcriptomic and metabolomic datasets in this study. DAMs in the metabolome were significantly enriched in metabolic pathways, biosynthesis of secondary metabolites, biosynthesis of cofactors, protein digestion and absorption, and histidine metabolism in three comparison groups. DEGs, including FYN, PTK2, PXN, CRK, CRKL, PAK, RHOA, ROCK, INSR, PDPK1, and ARHGEF, were mainly enriched in focal adhesion, regulation of actin cytoskeleton, wnt signaling pathway, insulin signaling pathway, ECM-receptor interaction, cell cycle, and adherens junction. Moreover, by interactively comparing metabolomic and transcriptomic data, DEGs (CKB, AGMAT, SRM, ODC1, GATM, P4HA2, MYADML2, GAMT, NOS2, PYCR3, LAP3, SMOX, P4HA3, ALDH18A1, COL19A1, and UROC1) were highly correlated with the corresponding metabolites (4-guanidinobutanoate, l-glutamic acid, histamine, l-isoleucine, l-aspartic acid, 4-imidazolone-5-propanoate, s-adenosylmethioninamine, indole, 2-methylbutyric acid, 1-(5-phosphoribosyl)imidazole-4-acetate, 4-(beta-acetylaminoethyl)imidazole, and piperidine) that were involved in arginine and proline metabolism, protein digestion and absorption, and histidine metabolism. These results will provide basic materials for further investigating the molecular mechanism of skeletal muscle development in duck. ## References 1. Te Pas M.F., de Wit A.A., Priem J., Cagnazzo M., Davoli R., Russo V., Pool M.H.. Transcriptome ex-pression profiles in prenatal pigs in relation to myogenesis. J. Muscle Res. Cell Motil. (2005) 26 157-165. DOI: 10.1007/s10974-005-7004-6 2. Marie M., Marcelle C., Gros J.. A common somitic origin for embryonic muscle progenitors. Med. Sci. (2005) 21 915-917 3. Gros J., Manceau M., Thomé V., Marcelle C.. A common somitic origin for embryonic muscle progenitors and satellite cells. Nature (2005) 435 954-958. DOI: 10.1038/nature03572 4. Wigmore P.M., Stickland N.C.. Muscle development in large and small pig fetuses. J. Anat. (1983) 137 235-245. PMID: 6630038 5. Biressi S., Tagliafico E., Lamorte G., Monteverde S., Tenedini E., Roncaglia E., Ferrari S., Ferrari S., Angelis M.G.C., Tajbakhsh S.. Intrinsic phenotypic diversity of embryonic and fetal myoblasts is revealed by genome-wide gene expression analysis on purified cells. Dev. Biol. (2007) 304 633-651. DOI: 10.1016/j.ydbio.2007.01.016 6. Chen W., Lv Y.T., Zhang H.X., Ruan D., Wang S., Lin Y.C.. Developmental specificity in skeletal muscle of late-term avian embryos and its potential manipulation. Poul. Sci. (2013) 92 2754-2764. DOI: 10.3382/ps.2013-03099 7. Qian X., Ba Y., Zhuang Q.F., Zhong G.F.. RNA-Seq technology and its application in fish transcriptomics. Omics (2014) 18 98-110. DOI: 10.1089/omi.2013.0110 8. Zhu C., Song W., Tao Z., Liu H., Xu W., Zhang S., Li H.. Deep RNA sequencing of pectoralis muscle transcriptomes during late-term embryonic to neonatal development in indigenous Chinese duck breeds. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0180403 9. Zhang Z., Du H., Yang C., Li Q., Qiu M., Song X., Hu C.. Comparative transcriptome analysis reveals regulators mediating breast muscle growth and development in three chicken breeds. Anim. Biotech. (2019) 30 233-241. DOI: 10.1080/10495398.2018.1476377 10. Trapnell C., Roberts A., Goff L., Pertea G., Kim D., Kelley D.R., Pimentel H., Salzberg S.L., Rinn J.L., Pachter L.. Differential gene and transcript expression analysis of RNA-seq experiments with Top Hat and Cufflinks. Nat. Protoc. (2012) 7 562-578. DOI: 10.1038/nprot.2012.016 11. Guo B., Greenwood P.L., Cafe L.M., Zhou G.H., Zhang W.G., Dalrymple B.P.. Transcriptome analysis of cattle muscle identifies potential markers for skeletal muscle growth rate and major cell types. BMC Genom. (2015) 16. DOI: 10.1186/s12864-015-1403-x 12. Kumar D., Bansal G., Narang A., Basak T., Abbas T., Dash D.. Integrating transcriptome and proteome profiling: Strategies and applications. Proteomics (2016) 16 2533-2544. DOI: 10.1002/pmic.201600140 13. Sun L.M., Bai M., Xiang L.J., Zhang G.S., Ma W., Jiang H.Z.. Comparative transcriptome profiling of longissimus muscle tissues from Qianhua Mutton Merino and Small Tail Han sheep. Sci. Rep. (2016) 6 33586. DOI: 10.1038/srep33586 14. Rinschen M.M., Ivanisevic J., Giera M., Siuzdak G.. Identification of bioactive metabolites using activity metabolomics. Nat. Rev. Mol. Cell Biol. (2019) 20 353-367. DOI: 10.1038/s41580-019-0108-4 15. Morton J.T., Aksenov A.A., Nothias L.F., Foulds J.R., Quinn R.A., Badri M.H., Swenson T.L., van Goethem M.W., Northen T.R., Vazquez-Baeza Y.. Learning representations of microbe-metabolite interactions. Nat. Methods (2019) 16 1306-1314. DOI: 10.1038/s41592-019-0616-3 16. Hu Z.G., Cao J.T., Ge L.Y., Zhang J.Q., Zhang H.L., Liu X.L.. Characterization and Comparative Transcriptomic Analysis of Skeletal Muscle in Pekin Duck at Different Growth Stages Using RNA-Seq. Animals (2021) 11. DOI: 10.3390/ani11030834 17. Liu C.L., Pan D.D., Ye Y.Y., Cao J.X.. ¹H NMR and multivariate data analysis of the relationship between the age and quality of duck meat. Food Chem. (2013) 141 1281-1286. DOI: 10.1016/j.foodchem.2013.03.102 18. Zhou H., Yang Y., Wang L.X., Ye S.Q., Liu J.J., Gong P., Qian Y.G., Zeng H.J., Chen X.. Integrated multi-omic data reveal the potential molecular mechanisms of the nutrition and flavor in Liancheng white duck meat. Front. Genet. (2022) 13 939585. DOI: 10.3389/fgene.2022.939585 19. Qiao N., Yang Y.Y., Liao J.Z., Zhang H., Yang F., Ma F.Y., Han Q.Y., Yu W.L., Li Y., Hu L.M.. Metabolomics and transcriptomics indicated the molecular targets of copper to the pig kidney. Ecotoxicol. Environ. Saf. (2021) 218 112284. DOI: 10.1016/j.ecoenv.2021.112284 20. Sun X.J., Tu K., Li L., Wu B., Wu L., Liu Z.H., Zhou L.Q., Tian J.T., Yang A.G.. Integrated transcriptome and metabolome analysis reveals molecular responses of the clams to acute hypoxia. Mar. Environ. Res. (2021) 168 105317. DOI: 10.1016/j.marenvres.2021.105317 21. Ryall J.G.. Metabolic reprogramming as a novel regulator of skeletal muscle development and regeneration. FEBS J. (2013) 280 4004-4013. DOI: 10.1111/febs.12189 22. Liebau F., Deane A.M., Rooyackers O.. Protein absorption and kinetics in critical illness. Curr. Opin. Clin. Nutr. (2021) 24 71-78. DOI: 10.1097/MCO.0000000000000707 23. Holeček M.. Histidine in Health and Disease: Metabolism, Physiological Importance, and Use as a Supplement. Nutrients (2020) 12. DOI: 10.3390/nu12030848 24. Zaynab M., Fatima M., Abbas M., Sharif Y., Umair M., Zafar M.H., Bahadar K.. Role of secondary metabolites in plant defense against pathogens. Microb. Pathog. (2018) 124 198-202. DOI: 10.1016/j.micpath.2018.08.034 25. Fischer J.D., Holliday G.L., Rahman S.A., Thornton J.M.. The Structures and Physicochemical Properties of Organic Cofactors in Biocatalysis. J. Mol. Biol. (2010) 403 803-824. DOI: 10.1016/j.jmb.2010.09.018 26. Rutten E.P.A., Engelen M.P.K.J., Schols A.M.W.J., Deutz N.E.P.. Skeletal muscle glutamate metabolism in health and disease: State of the art. Curr. Opin. Clin. Nutr. (2005) 8 41-51. DOI: 10.1097/00075197-200501000-00007 27. Malomuzh A.I., Naumenko N.V., Guseva D.S., Urazaev A.K.. Effect of Dipeptide N-Acetylaspartylglutamate on Denervation-Induced Changes in the Volume of Rat Skeletal Muscle Fibers. Bull. Exp. Biol. Med. (2006) 142 683-684. DOI: 10.1007/s10517-006-0450-3 28. Sato T., Ito Y., Nagasawa T.. Regulatory effects of the L-lysine metabolites, L-2-aminoadipic acid and L-pipecolic acid, on protein turnover in C2C12 myotubes. Biosc. Biotech. Bioch. (2016) 80 2168-2175. DOI: 10.1080/09168451.2016.1210499 29. Goossens C., Weckx R., Derde S., Perre S.V., Derese I., van Veldhoven P.P., Ghesquière B., van den Berghe G., Langouche L.. Altered cholesterol homeostasis in critical illness-induced muscle weakness: Efect of exogenous 3-hydroxybutyrate. Crit. Care (2021) 25 252. DOI: 10.1186/s13054-021-03688-1 30. Hana C.A., Klebermass E.M., Balber T.B., Mitterhauser M.M., Quint R., Hirtl Y., Klimpke A., Somloi S., Hutz J., Sperr E.. Inhibition of Lipid Accumulation in Skeletal Muscle and Liver Cells: A Protective Mechanism of Bilirubin Against Diabetes Mellitus Type 2. Front. Pharmacol. (2021) 11 636533. DOI: 10.3389/fphar.2020.636533 31. Da Paixão A.O., Bolin A.P., Silvestre J.G., Rodrigues A.C.. Palmitic Acid Impairs Myogenesis and Alters Temporal Expression of miR-133a and miR-206 in C2C12 Myoblasts. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22052748 32. Wang N., Ma H., Li J., Meng C.Y., Zou J., Wang H., Liu K., Liu M.D., Xiao X.Z., Zhang H.L.. HSF1 functions as a key defender against palmitic acid-induced ferroptosis in cardiomyocytes. J. Mol. Cell. Cardiol. (2021) 150 65-76. DOI: 10.1016/j.yjmcc.2020.10.010 33. Khalil A.S.M., Giribabu N., Yelumalai S., Shahzad H., Kilari E.K., Salleh N.. Myristic acid defends against testicular oxidative stress, inflammation, apoptosis: Restoration of spermatogenesis, steroidogenesis in diabetic rats. Life Sci. (2021) 278 119605. DOI: 10.1016/j.lfs.2021.119605 34. Goethals S., Rombouts C., Hemeryck L.Y., van Meulebroek L., van Hecke T., Vossen E., van Camp J., de Smet S., Vanhaecke L.. Untargeted Metabolomics to Reveal Red versus White Meat-Associated Gut Metabolites in a Prudent and Western Dietary Context. Mol. Nutr. Food Res. (2020) 64 e2000070. DOI: 10.1002/mnfr.202000070 35. Cui X.D., Li X.X., He Y.T., Yu J., Fu J., Song B., Zhao R.C.H.. Combined NOX/ROS/PKC Signaling Pathway and Metabolomic Analysis Reveals the Mechanism of TRAM34-Induced Endothelial Progenitor Cell Senescence. Stem Cells Dev. (2021) 30 671-682. DOI: 10.1089/scd.2021.0062 36. Rustad K.C., Wong V.W., Gurtner G.C.. The role of focal adhesion complexes in fibroblast mechanotransduction during scar formation. Differentiation (2013) 86 87-91. DOI: 10.1016/j.diff.2013.02.003 37. Mishra Y.G., Manavathi B.. Focal adhesion dynamics in cellular function and disease. Cell. Signal. (2021) 85 110046. DOI: 10.1016/j.cellsig.2021.110046 38. Wozniak M.A., Modzelewska K., Kwong L., Keely P.J.. Focal adhesion regulation of cell behavior. Biochim. Biophys. Acta Mol. Cell Res. (2004) 1692 103-119. DOI: 10.1016/j.bbamcr.2004.04.007 39. Graham Z.A., Gallagher P.M., Cardozo C.P.. Focal adhesion kinase and its role in skeletal muscle. J. Muscle Res. Cell Motil. (2015) 36 305-315. DOI: 10.1007/s10974-015-9415-3 40. Mitra S.K., Schlaepfer D.D.. Integrin-regulated FAK-Src signaling in normal and cancer cells. Curr. Opin. Cell Biol. (2006) 18 516-523. DOI: 10.1016/j.ceb.2006.08.011 41. Wang J., Zhang Q.B., Li S., Chen Z.H., Tan J.J., Yao J.T., Duan D.L.. Low molecular weight fucoidan alleviates diabetic nephropathy by binding fibronectin and inhibiting ECM-receptor interaction in human renal mesangial cells. Int. J. Biol. Macromol. (2020) 150 304-314. DOI: 10.1016/j.ijbiomac.2020.02.087 42. Lock J.G., Wehrle-Haller B., Strömblad S.. Cell-matrix adhesion complexes: Master control machinery of cell migration. Semin. Cancer Biol. (2008) 18 65-76. DOI: 10.1016/j.semcancer.2007.10.001 43. Hu C., Zuo Q.S., Jin K., Zhao Z.Y., Wu Y.H., Gao J.C., Wang C.Y., Wang Y.J., Zhan W.D., Zhou J.. Retinoic acid promotes formation of chicken (Gallus gallus) spermatogonial stem cells by regulating the ECM-receptor interaction signaling pathway. Gene (2022) 820 146227. DOI: 10.1016/j.gene.2022.146227 44. Berman A.E., Kozlova N.I., Morozevich G.E.. Integrins: Structure and signaling. Biochemistry (2003) 68 1284-1299. DOI: 10.1023/B:BIRY.0000011649.03634.74 45. Santella L., Limatola N., Vasilev F., Chun T.J.. Maturation and fertilization of echinoderm eggs: Role of actin cytoskeleton dynamics. Biochem. Bioph. Res. Commun. (2018) 506 361-371. DOI: 10.1016/j.bbrc.2018.09.084 46. Kepser L.J., Damar F., de Cicco T., Chaponnier C., Prószyński T.J., Pagenstecher A., Rust M.B.. CAP2 deficiency delays myofibril actin cytoskeleton differentiation and disturbs skeletal muscle architecture and function. Proc. Natl. Acad. Sci. USA (2019) 116 8397-8402. DOI: 10.1073/pnas.1813351116 47. Lazki-Hagenbach P., Klein O., Sagi-Eisenberg R.. The actin cytoskeleton and mast cell function. Curr. Opin. Immunol. (2021) 72 27-33. DOI: 10.1016/j.coi.2021.03.002 48. Sandbo N., Dulin N.. Actin cytoskeleton in myofibroblast differentiation: Ultrastructure defining form and driving function. Transl. Res. (2011) 158 181-196. DOI: 10.1016/j.trsl.2011.05.004 49. Larsson C.. Protein kinase C and the regulation of the actin cytoskeleton. Cell. Signal. (2006) 18 276-284. DOI: 10.1016/j.cellsig.2005.07.010 50. Sharma R., Kumar D., Jha N.K., Jha S.K., Ambasta R.K., Kumar P.. Re-expression of cell cycle markers in aged neurons and muscles: Whether cells should divide or die?. Biochim. Biophys. Acta Mol. Basis Dis. (2017) 1863 324-336. DOI: 10.1016/j.bbadis.2016.09.010 51. Seoane J., Le H.V., Shen L.J., Anderson S.A., Massagué J.. Integration of Smad and Forkhead pathways in the control of neuroepithelial and glioblastoma cell proliferation. Cell (2004) 117 211-223. DOI: 10.1016/S0092-8674(04)00298-3 52. Murakami M.S., Moody S.A., Daar I.O., Morrison D.K.. Morphogenesis during Xenopus gastrulation requires Wee1-mediated inhibition of cell proliferation. Development (2004) 131 571-580. DOI: 10.1242/dev.00971 53. Angulo-Urarte A., van der Wal T., Huveneers S.. Cell-cell junctions as sensors and transducers of mechanical forces. Biochim. Biophys. Acta Biomembranes (2020) 1862 183316. DOI: 10.1016/j.bbamem.2020.183316 54. Wang M.N., Niu Z.B., Qin H.Q., Ruan B.Z., Zheng Y., Ning X.K., Gu S.Z., Gao L.H., Chen Z.L., Wang X.N.. Mechanical Ring Interfaces between Adherens Junction and Contractile Actomyosin to Coordinate Entotic Cell-in-Cell Formation. Cell Rep. (2020) 32 108071. DOI: 10.1016/j.celrep.2020.108071 55. Nagy J.I., Lynn B.D.. Structural and Intermolecular Associations Between Connexin36 and Protein Components of the Adherens Junction-Neuronal Gap Junction Complex. Neuroscience (2018) 384 241-261. DOI: 10.1016/j.neuroscience.2018.05.026 56. Padmanabhan A., Rao M.V., Wu Y., Zaidel-Bar R.. Jack of all trades: Functional modularity in the adherens junction. Curr. Opin. Cell Biol. (2015) 36 32-40. DOI: 10.1016/j.ceb.2015.06.008 57. Hussain M., Xu C.Y., Lu M.P., Wu X.L., Tang L.F., Wu X.M.. Wnt/β-catenin signaling links embryonic lung development and asthmatic airway remodeling. Biochim. Biophys. Acta Mol. Basis Dis. (2017) 1863 3226-3242. DOI: 10.1016/j.bbadis.2017.08.031 58. Anakwe K., Robson L., Hadley J., Buxton P., Church V., Allen S., Hartmann C., Harfe B., Nohno T., Brown A.M.C.. Wnt signalling regulates myogenic differentiation in the developing avian wing. Development (2003) 130 3503-3514. DOI: 10.1242/dev.00538 59. Alicea-Delgado M., García-Arrarás J.E.. Wnt/β-catenin signaling pathway regulates cell proliferation but not muscle dedifferentiation nor apoptosis during sea cucumber intestinal regeneration. Dev. Biol. (2021) 480 105-113. DOI: 10.1016/j.ydbio.2021.08.011 60. Bovolenta P., Rodriguez J., Esteve P.. Frizzled/RYK mediated signalling in axon guidance. Development (2006) 133 4399-4408. DOI: 10.1242/dev.02592 61. Helfer G., Tups A.. Hypothalamic Wnt signalling and its role in energy balance regulation. J. Neuroendocrinol. (2016) 28 12368. DOI: 10.1111/jne.12368 62. Lu Y., Chen S.R., Liu W.B., Hou Z.C., Xu G.Y., Yang N.. Polymorphisms in Wnt signaling pathway genes are significantly associated with chicken carcass traits. Poul. Sci. (2012) 91 1299-1307. DOI: 10.3382/ps.2012-02157 63. Baudler S., Krone W., Brüning J.C.. Genetic manipulation of the insulin signalling cascade in mice-potential insight into the pathomechanism of type 2 diabetes. Best Pract. Res. Clin. Endocrinol. Metab. (2003) 17 431-443. DOI: 10.1016/S1521-690X(03)00039-3 64. Richter E.A., Hargreaves M.. Exercise, GLUT4, and skeletal muscle glucose uptake. Physiol. Rev. (2013) 93 993-1017. DOI: 10.1152/physrev.00038.2012 65. Mukaida S., Evans B.A., Bengtsson T., Hutchinson D.S., Sato M.. Adrenoceptors promote glucose uptake into adipocytes and muscle by an insulin-independent signaling pathway involving mechanistic target of rapamycin complex 2. Pharmacol. Res. (2017) 116 87-92. DOI: 10.1016/j.phrs.2016.12.022 66. Rhoads R.P., Baumgard L.H., El-Kadi S.W., Zhao L.D.. Physiology and Endocrinology Symposium: Roles for insulin-supported skeletal muscle growth. J. Anim. Sci. (2016) 94 1791-1802. DOI: 10.2527/jas.2015-0110 67. Britton J.S., Lockwood W.K., Li L., Cohen S.M., Edgar B.A.. Drosophila’s insulin/PI3-kinase pathway coordinates cellular metabolism with nutritional conditions. Dev. Cell (2002) 2 239-249. DOI: 10.1016/S1534-5807(02)00117-X 68. Ma J.H., Su L.P., Zhu J., Law P.K., Lee K.O., Ye L., Wang Z.Z.. Skeletal myoblast transplantation on gene expression profiles of insulin signaling pathway and mitochondrial biogenesis and function in skeletal muscle. Diabetes Res. Clin. Pract. (2013) 102 43-52. DOI: 10.1016/j.diabres.2013.08.006 69. Jia B.X., Yu S., Yu D.Z., Liu N., Zhang S., Wu A.B.. Mycotoxin deoxynivalenol affects myoblast differentiation via downregulating cytoskeleton and ECM-integrin-FAK-RAC-PAK signaling pathway. Ecotox. Environ. Saf. (2021) 226 112850. DOI: 10.1016/j.ecoenv.2021.112850 70. Wolfenson H., Henis Y.I., Geiger B., Bershadsky A.D.. The heel and toe of the cell’s foot: A multifaceted approach for understanding the structure and dynamics of focal adhesions. Cell Motil. Cytoskel. (2009) 66 1017-1029. DOI: 10.1002/cm.20410 71. Petrany M.J., Millay D.P.. Cell fusion: Merging membranes and making muscle. Trends Cell Biol. (2019) 29 964-973. DOI: 10.1016/j.tcb.2019.09.002 72. Lehka L., Rędowicz M.J.. Mechanisms regulating myoblast fusion: A multilevel interplay. Semin. Cell Dev. Biol. (2020) 104 81-92. DOI: 10.1016/j.semcdb.2020.02.004 73. Wang Y.C., Yamada E., Zong H.H., Pessin J.E.. Fyn activation of mTORC1 stimulates the IRE1alpha-JNK pathway, leading to cell death. J. Biol. Chem. (2015) 290 24772-24783. DOI: 10.1074/jbc.M115.687020 74. Yamada E., Bastie C.C., Koga H., Wang Y.C., Cuervo A.M., Pessin J.E.. Mouse skeletal muscle fiber-type-specific macroautophagy and muscle wasting are regulated by a Fyn/STAT3/Vps34 signaling pathway. Cell Rep. (2012) 1 557-569. DOI: 10.1016/j.celrep.2012.03.014 75. Erskine R.M., Williams A.G., Jones D.A., Stewart C.E., Degens H.. Do PTK2 gene polymorphisms contribute to the interindividual variability in muscle strength and the response to resistance training? A preliminary report. J. Appl. Physiol. (2012) 112 1329-1334. DOI: 10.1152/japplphysiol.01137.2011 76. Stebbings G.K., Williams A.G., Morse C.I., Day S.H.. Polymorphisms in PTK2 are associated with skeletal muscle specific force: An independent replication study. Eur. J. Appl. Physiol. (2017) 117 713-720. DOI: 10.1007/s00421-017-3567-3 77. Lee J., Bandyopadhyay J., Lee J.I., Cho I., Park D., Cho J.H.. A Role for Peroxidasin PXN-1 in Aspects of. Mol. Cells (2015) 38 51-57. PMID: 25475546 78. Gotenstein J.R., Swale R.E., Fukuda T., Wu Z.L., Giurumescu C.A., Goncharov A., Jin Y.S., Chisholm A.D.. The. Development (2010) 137 3603-3613. DOI: 10.1242/dev.049189 79. Shi L.J., Racedo S.E., Diacou A., Park T., Zhou B., Morrow B.E.. Crk and Crkl have shared functions in neural crest cells for cardiac outf low tract septation and vascular smooth muscle differentiation. Hum. Mol. Genet. (2022) 31 1197-1215. DOI: 10.1093/hmg/ddab313 80. Oakes P.W., Gardel M.L.. Stressing the limits of focal adhesion mechanosensitivity. Curr. Opin. Cell Biol. (2014) 30 68-73. DOI: 10.1016/j.ceb.2014.06.003 81. Park T.J., Boyd K., Curran T.. Cardiovascular and Craniofacial Defects in Crk-Null Mice. Mol. Cell. Biol. (2006) 26 6272-6282. DOI: 10.1128/MCB.00472-06 82. Moore C.A., Parkin C.A., Bidet Y., Ingham P.W.. A role for the Myoblast city homologues Dock1 and Dock5 and the adaptor proteins Crk and Crk-like in zebrafish myoblast fusion. Development (2007) 134 3145-3153. DOI: 10.1242/dev.001214 83. Rane C.K., Minden A.. P21 activated kinases: Structure, regulation, and functions. Small GTPases (2014) 5 e28003. DOI: 10.4161/sgtp.28003 84. Varshney P., Dey C.S.. P21-activated kinase 2 (PAK2) regulates glucose uptake and insulin sensitivity in neuronal cells. Mol. Cell. Endocrinol. (2016) 429 50-61. DOI: 10.1016/j.mce.2016.03.035 85. Feng Q.T., Di R.M., Tao F., Chang Z., Lu S.S., Fan W.J., Shan C.J., Li X.L., Yang Z.Z.. PDK1 regulates vascular remodeling and promotes epithelial-mesenchymal transition in cardiac development. Mol. Cell. Biol. (2010) 30 3711-3721. DOI: 10.1128/MCB.00420-10 86. Kim H., Cho S.C., Jeong H.J., Lee H.Y., Jeong M.H., Pyun J.H., Ryu D., Kim M.S., Lee Y.S., Kim M.S.. Indoprofen prevents muscle wasting in aged mice through activation of PDK1/AKT pathway. J. Cachexia Sarcopeni (2020) 11 1070-1088. DOI: 10.1002/jcsm.12558 87. Chen H., Nystrom F.H., Dong L.Q., Li Y., Song S., Liu F., Quon M.J.. Insulin Stimulates Increased Catalytic Activity of Phosphoinositide-Dependent Kinase-1 by a Phosphorylation-Dependent Mechanism. Biochemistry (2001) 40 11851-11859. DOI: 10.1021/bi010743c 88. Zhang W.W., Huang Y.L., Wu Y.D., Gunst S.J.. A novel role for RhoA GTPase in the regulation of airway smooth muscle contraction. Can. J. Physiol. Pharmacol. (2015) 93 129-136. DOI: 10.1139/cjpp-2014-0388 89. Robert A., Herrmann H., Davidson M.W., Gelfand V.I.. Microtubule-dependent transport of vimentin filament precursors is regulated by actin and by the concerted action of Rho- and p21-activated kinases. FASEB J. (2014) 28 2879-2890. DOI: 10.1096/fj.14-250019 90. Castellani L., Salvati E., Alemà S., Falcone G.. Fine Regulation of RhoA and Rock Is Required for Skeletal Muscle Differentiation. J. Biol. Chem. (2006) 281 15249-15257. DOI: 10.1074/jbc.M601390200 91. Muñoz V.R., Gaspar R.C., Severino M.B., Macêdo A.P.A., Simabuco F.M., Ropelle E.R., Cintra D.E., da Silva A.S., Kim Y.B., Pauli J.R.. Exercise Counterbalances Rho/ROCK2 Signaling Impairment in the Skeletal Muscle and Ameliorates Insulin Sensitivity in Obese Mice. Front. Immunol. (2021) 12 702025. DOI: 10.3389/fimmu.2021.702025 92. Khanna N., Fang Y.M., Yoon M.S., Chen J.. XPLN is an endogenous inhibitor of mTORC2. Proc. Natl. Acad. Sci. USA (2013) 110 15979-15984. DOI: 10.1073/pnas.1310434110 93. You J.S., Singh N., Reyes-Ordonez A., Khanna N., Bao Z.H., Zhao H.M., Chen J.. ARHGEF3 Regulates Skeletal Muscl e Regeneration and Strength through Autophagy. Cell Rep. (2021) 34 108594. DOI: 10.1016/j.celrep.2020.108594 94. Ramalingam L., Oh E., Thurmond D.C.. Novel roles for insulin receptor (IR) in adipocytes and skeletal muscle cells via new and unexpected substrates. Cell. Mol. Life Sci. (2013) 70 2815-2834. DOI: 10.1007/s00018-012-1176-1 95. Lim M.J., Choi K.J., Ding Y., Kim J.H., Kim B.S., Kim Y.H., Lee J., Choe W., Kang I., Ha J.. RhoA/Rho Kinase Blocks Muscle Differentiation via Serine Phosphorylation of Insulin Receptor Substrate-1 and -2. Mol. Endocrinol. (2007) 21 2282-2293. DOI: 10.1210/me.2007-0114 96. Chen X.L., Guo Y.F., Jia G., Liu G.M., Zhao H., Huang Z.Q.. Arginine promotes skeletal muscle fiber type transformation from fast-twitch to slow-twitch via Sirt1/AMPK pathway. J. Nutr. Biochem. (2018) 61 155-162. DOI: 10.1016/j.jnutbio.2018.08.007 97. Brunton J.A., Baldwin M.P., Hanna R.A., Bertolo R.F.. Proline supplementation to parenteral nutrition results in greater rates of protein synthesis in the muscle, skin, and small intestine in neonatal Yucatan miniature piglets. J. Nutr. (2012) 142 1004-1008. DOI: 10.3945/jn.111.154534 98. Holeček M.. Influence of Histidine Administration on Ammonia and Amino Acid Metabolism: A Review. Physiol. Res. (2020) 69 555-564. DOI: 10.33549/physiolres.934449 99. Hu Z.G., Cao J.T., Liu G.Y., Zhang H.L., Liu X.L.. Comparative Transcriptome Profiling of Skeletal Muscle from Black Muscovy Duck at Different Growth Stages Using RNA-seq. Genes (2020) 11. DOI: 10.3390/genes11101228
--- title: Anti-Inflammatory Effects and Photo- and Neuro-Protective Properties of BIO203, a New Amide Conjugate of Norbixin, in Development for the Treatment of Age-Related Macular Degeneration (AMD) authors: - Valérie Fontaine - Christine Balducci - Laurence Dinan - Elodie Monteiro - Thinhinane Boumedine - Mylène Fournié - Vincent Nguyen - Louis Guibout - Justine Clatot - Mathilde Latil - Stanislas Veillet - José-Alain Sahel - René Lafont - Pierre J. Dilda - Serge Camelo journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049354 doi: 10.3390/ijms24065296 license: CC BY 4.0 --- # Anti-Inflammatory Effects and Photo- and Neuro-Protective Properties of BIO203, a New Amide Conjugate of Norbixin, in Development for the Treatment of Age-Related Macular Degeneration (AMD) ## Abstract 9′-cis-norbixin (norbixin/BIO201) protects RPE cells against phototoxicity induced by blue light and N-retinylidene-N-retinylethanolamine (A2E) in vitro and preserves visual functions in animal models of age-related macular degeneration (AMD) in vivo. The purpose of this study was to examine the mode of action and the in vitro and in vivo effects of BIO203, a novel norbixin amide conjugate. Compared to norbixin, BIO203 displays improved stability at all temperatures tested for up to 18 months. In vitro, BIO203 and norbixin share a similar mode of action involving the inhibition of PPARs, NF-κB, and AP-1 transactivations. The two compounds also reduce IL-6, IL-8, and VEGF expression induced by A2E. In vivo, ocular maximal concentration and BIO203 plasma exposure are increased compared to those of norbixin. Moreover, BIO203 administered systemically protects visual functions and retinal structure in albino rats subjected to blue-light illumination and in the retinal degeneration model of Abca4−/− Rdh8−/− double knock-out mice following 6 months of oral complementation. In conclusion, we report here that BIO203 and norbixin share similar modes of action and protective effects in vitro and in vivo. BIO203, with its improved pharmacokinetic and stability properties, could be developed for the treatment of retinal degenerative diseases such as AMD. ## 1. Introduction Age-related macular degeneration (AMD) is the commonest cause of severe visual loss and blindness in developed countries among individuals aged 60 and older and remains a major medical need [1,2]. AMD slowly progresses from early AMD to intermediate AMD (iAMD), which further evolves to late-stage neovascular AMD and/or geographic atrophy (GA) [2]. It has been known for many years that carotenoids, such as lutein and zeaxanthin, are constitutive of the macula, and dietary apocarotenoids such as crocetin, and 9′-cis-norbixin (norbixin, pharmaceutical ingredient of BIO201) preserve the retinal architecture and visual functions [2]. Norbixin is a 6,6′-di-apo-carotenoid extracted from annatto (Bixa orellana) seeds [3]. We have previously demonstrated that norbixin protects porcine primary retinal pigmented epithelial (RPE) cells from phototoxicity induced by blue-light illumination coupled with A2E exposure in vitro [4]. Norbixin modulates inflammation, as demonstrated by its inhibitory effect on A2E-induced expression of IL-6 and IL-8 and by its effect on the transactivation of NF-κB and AP-1. In addition, our previous observations demonstrated that norbixin impairs VEGF expression induced by A2E [5]. Norbixin also reduces the accumulation of A2E by primary porcine RPE cells in vitro [4]. In vivo, norbixin possesses anti-inflammatory and antioxidant properties [6]. In agreement with the reported role of inflammation during retinal degeneration, we have shown that norbixin is beneficial in various in vivo animal models of AMD [4,7]. Indeed, norbixin administered systemically is neuroprotective against blue-light-induced retinal degeneration in rats and BALB/c mice [4,7]. Moreover, 5- to 6-months oral treatment by complementation of Abca4−/− Rdh8−/− double-knockout mice with norbixin-containing chow results in neuroprotection and partially preserves the function of both rods and cones in vivo. Oral complementation with norbixin also reduces A2E and lipofuscin accumulation in RPE cells in these animals [7]. The beneficial effects of norbixin are potentially associated with its interaction with several nuclear receptors: PPAR-γ, PPAR-α, PPAR-β/δ, and also RXR-α [5,8]. We proposed that norbixin behaves as a neutral antagonist of these nuclear receptors, thus inhibiting their transactivation induced by A2E [5]. Based on all these properties, norbixin or related carotenoids appear promising to treat the intermediate form of AMD, for which an effective drug is still lacking [2]. Despite all these interesting properties, we observed that norbixin stability at all temperatures tested and its ocular bioavailability were not acceptable for further industrial/clinical development (unpublished data). Therefore, we aimed to identify possible hemisynthetic compounds derived from norbixin retaining its biological effects on RPE cells and in vivo but with improved stability and bioavailability. We produced a set of norbixin amide conjugates by hemisynthesis from bixin. We prepared a range of primary or secondary amides, which were assayed by the same battery of tests: in vitro bioactivity, pharmacokinetic parameters in plasma, and ocular exposure, as well as stability. We found that most of the resulting compounds had a better pharmacokinetic profile and an improved tropism for the eye when compared with norbixin (patent Dinan et al., 2019; US $\frac{17}{788}$,534). In addition, some of these compounds showed a retinal pigment epithelial (RPE) photo-protective activity equivalent to or greater than that of norbixin. In the present study, we describe the effects of one of these conjugates (BIO203) in in vitro and in vivo models of AMD. BIO203 stability and ocular pharmacokinetics are presented here in comparison with those of the parent compound, norbixin (BIO201). ## 2.1. BIO203 Is More Stable Than Norbixin (BIO201) We evaluated the stability of BIO203 compared to norbixin at room temperature (RT) (Figure 1A), +4 °C (Figure 1B), and −20 °C (Figure 1C) for a period of up to 18 months. After 3 months at RT, we already observed a very significant decrease in norbixin stability, with only $41\%$ left of the initial quantity. The amount of norbixin kept decreasing at 6- and 12-month time points but stabilized between 12 and 18 months to $22.1\%$ of the initial quantity of norbixin (Figure 1A). Setting the temperature at 4 °C slightly improved the stability of norbixin. Nevertheless, at 3 months, there was already a loss of $10\%$ of the compound. This loss was still ongoing at 6 months (−$20.8\%$) and 12 months (−$66.4\%$) (Figure 1B). As for the stability at RT, no further decay was observed between 12 and 18 months at +4 °C (Figure 1B). At −20 °C, norbixin appeared stable for up to 6 months; however, there was a significant loss of compound after one year (−$18.1\%$) that did not worsen during the last 6-month period (between 12 and 18 months; Figure 1C). Therefore, we observed a natural loss of norbixin at all temperatures, but this was limited and delayed when negative temperatures were applied. By contrast, BIO203 was stable at all temperatures tested for at least 18 months (Figure 1A–C). ## 2.1.1. In Vitro, BIO203 and Norbixin Are Equally Effective in Protecting RPE Cells against Phototoxicity Induced by Blue-Light Exposure in the Presence of A2E We demonstrated previously that norbixin inhibits the phototoxicity of primary RPE cells induced by A2E in the presence of blue light [4]. Here, we confirmed our previously published results [4], showing that norbixin treatment at 20 μM promotes the survival of more than $80\%$ of RPE cells subjected to blue-light illumination in combination with A2E exposure (Figure 2). In addition, we show here that increasing doses of BIO203 also protect RPE cells illuminated in the presence of A2E (Figure 2). BIO203 protective effect was observed from 10 μM and increased at 15 and 20 μM. Interestingly, BIO203 reaches its maximum efficacy at 15 μM while the highest protection by norbixin is obtained at 20 μM. Since BIO203 contain an asymmetric carbon in the amide radical, BIO203 is a racemic compound composed of two enantiomers called BIO203-R and BIO203-S obtained by the addition of the 3-MP-R and 3-MP-S radicals, respectively. We tested whether the protective activity of BIO203 lay in one of these enantiomers or if it was shared by both. We observed that both enantiomers have the same protective activity as BIO203 at all concentrations from 10 to 20 μM (Figure 2). This indicates that there is no difference between enantiomers and BIO203 (racemic mix) activities, and it suggests that the effect of the racemic compound (BIO203) should be equally shared by both of its enantiomers represented in equal amounts. ## 2.1.2. BIO203 Inhibits the Transactivation of PPARs, NF-κB, and AP-1 but Not RXR Transactivation Induced by A2E We have shown previously that norbixin behaves as an inhibitor of A2E-induced PPAR transactivation [5]. In the present study, we tested whether BIO203 at 20 μM had the same effect on these nuclear receptors. Here we report that BIO203 at 20 μM entirely inhibits the PPAR transactivation induced by A2E at 20 μM (Figure 3A). We previously showed that norbixin significantly inhibited A2E-induced transactivation of RXR in porcine RPE cells in vitro [5]. Here, we confirm our previous observation that A2E induces the transactivation of endogenous RXR in porcine RPE cells (Figure 3B). However, BIO203 did not inhibit the transactivation of RXR induced by A2E (Figure 3B), indicating that BIO203 does not share a complete similarity in its mode of action with norbixin. Nevertheless, we showed previously that inhibiting the transactivation of either PPAR or RXR or both resulted in a reduction of inflammation and angiogenesis in our in vitro model [5]. Therefore, we wanted to determine whether BIO203 at 20 μM could also limit inflammation induced by A2E (20 μM) on RPE cells in vitro. Previously, we demonstrated that norbixin inhibits A2E-induced NF-κB and AP1 transactivation, two transcription factors involved in the regulation of inflammation [5]. In the present study, we showed that NF-κb transactivation induced by A2E was significantly downregulated by BIO203 at 20 μM (−$69.4\%$, $p \leq 0.05$) (Figure 3C) in a similar fashion as norbixin (−$73\%$) [5]. Similarly, A2E-induced AP1 transactivation was also inhibited by BIO203 at 20 μM (Figure 3D). Altogether these observations suggest that BIO203 and norbixin share the same anti-inflammatory effect and similar modulation of PPAR nuclear receptors. ## 2.1.3. BIO203 Inhibits the Expression of IL-6, IL-8, and VEGF, Induced by A2E Transactivation of NF-κb and AP-1 are pivotal in inflammatory responses regulating multiple aspects of innate and adaptative immune functions as well as angiogenesis [9,10]. Here, we aimed to assess the effects of BIO203 at 20 μM on A2E-induced expression of inflammatory cytokines and VEGF, a predominant proangiogenic factor. We have shown previously that norbixin at 20 mM significantly reduced the mRNA expression of IL-6, IL-8, and VEGF [5]. Similarly, BIO203 (20 mM) completely abrogated the mRNA expression of IL-6 to lower levels than baseline without A2E induction (−$104.8\%$, $p \leq 0.001$) (Figure 4A), IL-8 (−$86.8\%$, $p \leq 0.001$) (Figure 4B) and VEGF (−$84.8\%$, $p \leq 0.001$) (Figure 4C) in porcine RPE cells stimulated by 20 μM of A2E. Based on these observations, it could be hypothesized that the BIO203 inhibition of PPAR transactivation induced by A2E might be involved in the inhibition of inflammation and angiogenesis induced by A2E in RPE cells in vitro. ## 2.2. BIO203 Displays Improved Ocular Pharmacokinetics Compared with Norbixin (BIO201) Following Single and Multiple Intraperitoneal Administrations We performed comparative pharmacokinetic studies of BIO203 with norbixin (BIO201) in the plasma and eyes of rats administered intraperitoneally with similar doses of norbixin (10 mg/kg) and BIO203 (8.8 mg/kg). In this experiment, BIO203 displayed increased intraocular AUC and Cmax compared to norbixin (Figure 5A). BIO203 displayed a Cmax value of 133.1 ng/eye at 1 h post-administration, compared to a norbixin Cmax value of less than 8.5 ng/eye at 0.25 h (Figure 5A). In addition, AUC for BIO203 (704.1 ng·h/eye) was approximately 280 times higher than norbixin’s (AUC 2.5 ng·h/eye) (Figure 5A). The remaining intraocular concentration of BIO203 at 24 h was close to 2 ng/mL. In comparison, at this time point, norbixin was no longer detectable in ocular tissues (Figure 5A). We then performed a pharmacokinetic comparison of four intraperitoneal administrations of 10 mg/kg of BIO201 and 2.5 mg/kg of BIO203 in Wistar rats. Despite a higher quantity of norbixin administered, ocular exposure of BIO203 (144.9 ng·h/eye) was approximately 10 times higher than that of BIO201 (16.7 ng·h/eye) (Figure 5B). Altogether, following single and multiple intraperitoneal administrations in rats, BIO203 displays improved ocular pharmacokinetics compared to BIO201. ## 2.3. BIO203 Is Neuroprotective and Preserves the Visual Function of Albino Rats Exposed to Blue-Light-Induced Photoreceptor Degeneration We previously demonstrated that intraperitoneal injections of norbixin at 50 mg/kg was the lowest dose of BIO201 giving the maximal retinal neuroprotection and optimal preservation of visual functions in albino rats exposed to blue-light illumination [4]. To determine whether systemic administration of BIO203 was neuroprotective and could preserve visual function of photoreceptors in vivo, we used the same model of BLD in albino rats. We performed preliminary experiments comparing doses ranging from 0.125 mg/kg to 2.5 mg/kg of BIO203. The tested doses of BIO203 were chosen according to ocular PK data in rats following the same administration protocol used in the BLD studies presented above (Figure 5B). However, in this preliminary experiment, only a limited neuroprotective effect of BIO203 at 2.5 mg/mL compared to vehicle alone was observed. Therefore, we decided to increase the dose of BIO203 from 5 to 25 mg/kg. Increased doses of BIO203 (5, 10, and 25 mg/kg) were injected intraperitoneally at four-time points: 30 min prior to BLD and 1, 2.5, and 4 h after the beginning of exposure to blue- light. The BIO203 effects on visual function and retinal histology at these three doses were compared to the effect of similar regimen administrations of vehicle and positive controls (PBN at 50 mg/kg and the optimal dose of norbixin (BIO201) at 50 mg/kg). Six hours of blue-light exposure induced severe loss of retinal function in vehicle-dosed rats, as measured seven days after exposure by ERG A-wave (Figure 6A) and ERG B-wave (Figure 6B). As previously demonstrated, treatments with PBN and norbixin protected the visual functions [4]. While there was no significant effect in the group treated with 5 and 10 mg/kg of BIO203, four consecutive intraperitoneal administrations of BIO203 at 25 mg/kg provided a protective effect on ERG A-wave and ERG B wave ($p \leq 0.0001$). This protective effect was like those observed in PBN and norbixin-treated animals (Figure 6A,B). We then evaluated the neuroprotective effect of intraperitoneal injections of BIO203 at 5, 10, and 25 mg/kg. In agreement with the effects observed on visual function, treatment with PBN, norbixin (BIO201) at 50 mg/kg, and BIO203 at 25 mg/kg partially preserved the retinal structure of rats subjected to 6 h of blue-light exposure ($p \leq 0.0001$) (Figure 6C,D). By contrast, we did not observe neuroprotection following treatment with lower doses of BIO203 at 5 and 10 mg/kg and with vehicle (Figure 6C,D). Altogether we show here that intraperitoneal injections of BIO203 are neuroprotective and partially preserve visual function in vivo in a model of blue-light damage in rats. ## 2.4. Effect of 6 Months BIO203 Oral Treatment by Complementation in 11–12-Month-Old Abca4−/− Rdh8−/− Mice Eleven to twelve-month-old Abca4−/− Rdh8−/− mice were fed either with normal pellets (control) or with pellets containing BIO203 (LD-BIO203 [50 μg/g] and high dose HD-BIO203 [500 μg/g]) for 6 months (Figure 7A). Firstly, we determined whether BIO203 could be detected in the eyes and plasma of mice fed with BIO203 (Figure 7B). BIO203 was not detected in the eyes of mice fed during 6 months with LD or HD -BIO203-containing pellets; however, in the plasma, BIO203 was detected in the group treated with HD-BIO203 but not with LD-BIO203 (Figure 7B). After 6 months of treatment with HD-BIO203 pellets, the scotopic A wave ERG (at flash intensities: of 1 and 10 cd.s/m2) was significantly superior when compared with the ERG of animals fed with control pellets (Figure 7C; $p \leq 0.05$). Similarly, 6 months of complementation with HD-BIO203 reduced the loss of scotopic B wave ERG (at flash intensity: 30 cd.s/m2) when compared with the intensity of scotopic B wave of animals treated with control pellets (Figure 7D; $p \leq 0.05$). However, no significant difference in photopic B wave ERG was observed between the group of mice treated with HD-BIO203 compared to the group of mice fed with control pellets (Figure 7E). No effect on the visual function was observed in the group of mice treated with LD-BIO203 (Figure 7C–E). The lack of effect of LD-BIO203 is in agreement with the absence of BIO203 plasmatic exposure in this group of mice and indicates a dose-response effect. No difference in the thickness of the photoreceptor nuclear layer was noted between eyes of 17-month-old mice treated with HD-BIO203 compared to eyes of mice fed with LD-BIO203 or control pellets for the previous 6 months (Figure 7F). In addition, we previously reported that norbixin reduced A2E accumulation in vitro [4] and in vivo in 17-month-old mice treated the previous 6 months by norbixin [7]. Here, we measured the amount of A2E in the retina of mice treated with control or with LD- or HD-BIO203 for 6 months. The amounts of ocular A2E did not differ significantly between mice treated with control pellets and with LD- and HD-BIO203-containing pellets during the 6 months (Figure 7G). However, we observed a trend towards a reduction of A2E concentration in the retinas of animals treated with HD-BIO203 compared to mice that were fed with control pellets (Figure 7G), but this difference did not reach significance (−$21.7\%$; $$p \leq 0.407$$). ## 3. Discussion Age-related macular degeneration (AMD) is the most frequent cause of severe visual function loss and blindness in developed countries among individuals aged 60 and older [1]. AMD is still a major unmet medical need, and the development of new therapeutic strategies to fight this blinding disease are still required. The major risk factors remain aging [11,12] along with smoking [13]. In addition, oxidative stress, angiogenesis, and inflammation have been defined as critical factors for retinal degeneration leading to AMD pathogenesis [14,15,16]. N-retinylidene-N-retinylethanolamine (A2E) accumulation in the retina and retinal pigmented epithelium (RPE) cells has been associated with the initial stage of the disease [17,18,19,20]. An important role for nuclear receptors in the pathogenesis of AMD has been strongly suggested [21,22,23,24]. We have previously shown that A2E activates PPAR-α, PPAR-β/δ, PPAR-γ, and RXR, and we proposed that this could lead to the cascade of inflammation, angiogenesis, and retinal degeneration observed during AMD [5]. Indeed, we have shown that 9′-cis-norbixin, a 6,6′-di-apo-carotenoid extracted from annatto (Bixa orellana), behaves as an antagonist of PPARs and RXR transactivation induced by A2E and counteracts its proinflammatory and proangiogenic effects [5]. For instance, in vitro, norbixin leads to the photoprotection of primary porcine RPE cells from A2E and blue-light illumination in rats [4] and mice [7]. Moreover, norbixin regulates inflammation, as demonstrated by the inhibitory effect of BIO201 on A2E-induced expression of IL-6, IL-8, transactivation of NF-κB, and AP-1 [5]. In addition, our previous observations demonstrated that norbixin impairs VEGF expression induced by A2E [5]. We also demonstrated the neuroprotective effects and the preservation of visual function, especially of rod photoreceptors by BIO201 in various in vivo animal models of AMD [4,7] and as previously demonstrated for retinal carotenoids such as lutein, norbixin reduced the accumulation of A2E [4,7,25]. These promising observations suggested that BIO201 could be developed as a new treatment for the initial stages of AMD [2]. However, we noted that ocular accumulation of BIO201 administered systematically was limited [7] and that its poor stability was not compatible with its industrial development as an effective oral drug to treat AMD. We developed by hemisynthesis a second generation of molecules. Starting from bixin, we reacted the free carboxylic group with a set of primary or secondary amines. We found that some of the resulting amides had a better pharmacokinetic profile and a better tropism for the eye than those of norbixin (patent Dinan et al., 2019; US $\frac{17}{788}$,534). In the present study, we describe the properties of BIO203, one of these norbixin conjugates. Here, we report that BIO203 has improved stability and ocular pharmacokinetics, with both Cmax and AUC increased by more than 10- and 200-fold compared to BIO201, respectively. The exact mechanisms explaining these improved properties remain unknown at present but could be due to a better ocular uptake of BIO203 originating from the plasma. It is perhaps possible that the addition of the amine group modifies the polarity of BIO203, which becomes more lipophilic. Further experiments to decipher the mechanisms at play are currently underway. We also observed an equivalent or slightly improved photoprotective efficacy of BIO203 compared to norbixin in the phototoxicity test in RPE cells challenged with blue-light and A2E in vitro at low concentrations (Figure 2). Interestingly, in vitro, the effect on the nuclear receptors PPARs and RXR of norbixin and its conjugate BIO203 were similar but not identical. Indeed, both norbixin and BIO203 inhibited equally PPARs transactivation induced by A2E. We show here that, contrary to norbixin, BIO203 does not inhibit RXR transactivation. Nevertheless, and probably due to the permissive nature of PPARs that form heterodimers with RXR [26,27,28,29], disrupting the binding of A2E with PPARs or RXR alone is sufficient to inhibit the inflammatory and angiogenic effects of A2E [5]. In agreement with this observation, BIO203, which only impairs PPARs transactivation but not RXR transactivation induced by A2E, reproduces all the anti-inflammatory and anti-angiogenic properties observed with norbixin. Indeed, we show here that BIO203 inhibits NF-κB and AP-1 transactivation induced by A2E. Accordingly, BIO203 also reduced IL-6, IL-8, and VEGF expression induced by A2E in primary porcine RPE. In addition, BIO203 slightly inhibited IL-6 expression on its own, but this tendency was not statistically significant. In vivo, repeated intraperitoneal injections of BIO203 protected visual function in albino rats exposed to blue-light exposure as previously observed with norbixin [4]. Indeed, BIO203 had a protective effect on scotopic and photopic ERG amplitudes similar to norbixin. Moreover, BIO203 as norbixin was able to reduce retinal degeneration in this acute blue-light exposure model in rats. We then evaluated the effects of long-term (6-months) oral treatment by complementation of two doses of BIO203 in ABCA4−/− RDH8−/− mice accumulating A2E during aging. Here we showed that the highest dose of BIO203 partly but significantly preserved scotopic A wave and B wave amplitude. By contrast, BIO203 did not limit retinal degeneration in ABCA4−/− RDH8−/− mice. Since BIO203 demonstrated neuroprotective properties in the albino rat model, we hypothesize that the ABCA4−/− RDH8−/− mice used in the present study were probably too old (12–17 months) to observe an effect. Indeed, we previously demonstrated that the maximum retinal degeneration in this animal model occurs in animals aged between 9 and 15 months [7]. Accordingly, BIO203 did not show an effect on photopic vision and A2E accumulation in mice. However, we observed a non-significant reduction (−$21.7\%$; $$p \leq 0.407$$) of A2E levels in the eyes of mice treated with HD-BIO203. Future experiments with higher numbers of animals will be required to confirm this trend. Importantly, there was a clear dose-response effect in these experiments performed in mice. Indeed, we could not detect BIO203 in the plasma of mice treated with low LD-BIO203-containing pellets, in which we also did not observe an effect of BIO203 in vivo. In contrast, preservation of visual function was observed in mice treated with HD-BIO203 and in which BIO203 could be detected in the plasma. Therefore, it can be hypothesized that increasing the dose of BIO203 further would allow to reproduce all the effects on retinal and visual functions in mice previously observed in mice treated with norbixin. Although the effects of HD-BIO203 were limited to the preservation of scotopic ERG, it is important to remind that protection of scotopic vision is crucial for the cure of the early stages of AMD. Indeed, the first clinical sign of intermediate AMD is the loss of “night vision” mediated mainly by rod photoreceptors whose activity is measured by scotopic ERG [2,30,31]. We also observed a dose-response effect of BIO203 in the rat BLD-model as shown in our preliminary experiments ranging from 0.125 mg/kg to 2.5 mg/kg and the experiments presented here with doses ranging from 5 to 25 mg/kg. The efficacy dose of BIO203 and norbixin in the rat model of BLD were similar (25 and 50 mg/kg, respectively) despite their different pharmacokinetics. The reason for this apparent absence of correlation between improved ocular uptake into the eye and an expected reduced efficacious dose of BIO203 is not understood at the moment and will be the subject of further pharmacological and toxicological studies that are required before evaluating the safety and efficacy of BIO203 in clinical studies in humans. In conclusion, norbixin (BIO201) efficacy in an in vitro model of RPE dysfunction and in vivo models of AMD using several strains of mice (WT and genetically modified (ABCA4−/−RDH8−/−) double knock-out mice) and WT rats have been presented in our three previous articles: [4,5,7]. In the present study, we have demonstrated that BIO203, which displays improved ocular PK characteristics and stability when compared to norbixin, reproduces the effects previously observed with norbixin. These include photoprotection in vitro, inhibition of transactivation of NF-κB and AP-1, and the expression of VEGF, IL-8, and IL-6, involved respectively in angiogenesis, and inflammation, which are critical for AMD pathogenesis. All these effects may be at least partially linked to their inhibition of PPAR transactivation induced by A2E. We also report that in vivo, BIO203 administered systemically is neuroprotective and has limited visual function loss, as norbixin did, in two models of retinal degeneration. We previously reviewed the potential clinical interest of carotenoids in general for the treatment over long periods of time (years) in patients diagnosed with the intermediate form of AMD [2]. Our results of visual function protection in ABCA4−/−RDH8−/− mice fed with pellets containing HD-BIO203 for 6 months suggest that this new compound could be well suited for the treatment of the initial stages of the disease and particularly for patients with early and intermediate AMD. BIO203, with improved eye exposure and stability compared to norbixin, may be an even better drug candidate for the treatment of AMD. In consequence, Biophytis is actively engaged in the clinical development of BIO203 in this indication. ## 4.1. Ethics Statement All procedures were carried out according to the guidelines on the ethical use of animals from the European Community Council Directive ($\frac{86}{609}$/EEC) and were approved by the French Ministry of Agriculture (OGM agreement 6193) and by the Committee on the Ethics of Animal Experiments of the French Ministry of Research. All efforts were made to minimize suffering. ## 4.2. Reagents/Chemicals/HPLC Equipment All usual chemicals and primers were from Sigma (St. Louis, MO, USA). Reagents for cell culture, transfection, and quantitative RT-PCR were from Thermo Fisher Scientific (Waltham, MA, USA). RNA extraction NucleoSpin® RNA kit was from Macherey Nagel (Düren, Germany). Cignal Pathway Reporter Assay Kits were from QIAGEN (Frederick, MD, USA). The Dual-Luciferase Reporter Assay System was purchased from Promega (Madison, WI, USA). All HPLC analyses were performed on an Agilent 1260 equipped with a DAD and a triple quadrupole mass spectrometer (API6420, Applied Biosystems, Les Ulis, France). Analytical conditions are indicated for the different uses in the appropriate sections. ## 4.3. Preparation of BIO201 and BIO203 BIO201 (9′-cis-Norbixin) (Figure 8A) was prepared from 9′-cis-Bixin (AICABIX P, purity $92\%$) purchased from Aica-Color (Cusco, Peru) upon alkaline hydrolysis as previously described [4,32]. The obtained product showed an HPLC purity of $97\%$ (Figure 8A), and its 9′-cis structure was confirmed by NMR as previously described [33] (Table 1). BIO201 was stored as a red powder at −80 °C, and fresh solutions were prepared in DMSO. BIO203 (Figure 8B) was synthesized in two steps: Peptide coupling: Away from light and under a nitrogen atmosphere, bixin (7 g, 17.8 mmol) was suspended in anhydrous DMF (56 mL). Triethylamine (7.4 mL, 53.4 mmol) and carbonyldiimidazole (5.77 g, 35.6 mmol) were then added, and the mixture was stirred at room temperature for 2 h. 3-methoxypiperidine hydrochloride (8.10 g, 54.4 mmol) was added, and the mixture was stirred overnight at room temperature. Then 200 mL of water was added portion-wise, and the mixture was stirred for 1 h at room temperature after the end of the addition. The mixture was filtered onto a glass frit, and the purple paste was washed with water (3 × 40 mL), 1N HCl aqueous solution (3 × 40 mL), and di-isopropyl ether (3 × 40 mL). The resulting paste was dried on the glass frit and was used without further purification (kept at −20 °C). Hydrolysis step: Away from light, the paste (ester) (23.1 g, max 17.8 mmol) was dissolved in THF/MeOH (175 mL/90 mL) solution. 1N NaOH aqueous solution (90 mL) was added, and the mixture was stirred overnight at room temperature. 1N HCl aqueous solution (90 mL) was slowly added, and the mixture was stirred for 20 min at room temperature. The mixture was filtered through a glass frit, and the residue was washed with water (120 mL, then 2 × 65 mL). The red paste was solubilized in water/acetonitrile solution and then lyophilized to result in BIO203 as a red powder (6.96 g, $r = 82$%). The 9′-cis structure and purity of BIO203 at $98\%$ were assessed through HPLC (Figure 8B) and NMR analysis (Table 1). Given the racemic structure of the 3-methoxypiperidine used, the resulting BIO203 is also a racemate, and the R and S enantiomers can be resolved by HPLC on a chiral column (Figure 8C). The assignment of R or S structures was unambiguously established by the synthesis of pure enantiomers by using the same protocol with pure (R)- or (S)-3-methoxypiperidine. ## 4.4. Evaluation of Norbixin (BIO201) and BIO203 Powder Stability at Room Temperature, 4 °C, and −20 °C The potential breakdown of BIO203 compared to BIO201 was assessed every three months in samples stored for up to 18 months at room temperature, 4 °C, and −20 °C. On the day of analysis, weighted samples of BIO203 and norbixin powder were solubilized in DMSO at 10 mg/mL final concentration. DMSO solutions were further diluted 400-fold before performing LC-MS/MS analysis which was performed on an LC 1260 System coupled with Mass Spectrometer QQQ-6420 with DAD (Supplier Agilent Technologies). BIO201 and BIO203 were eluted from a reverse-phase column (2.1 mm × 50 mm; 5 μm particles; Fortis C18) with the following gradient of acetonitrile in water (both containing $0.1\%$ formic acid): 10 to $100\%$ in 10 min), (flow-rate: 0.3 mL/min), (UV detection A: 460 nm, B: 254 nm, DAD: 210–500 nm), (MS scan+: 50 to 800 Frag 120 CAV 5, MS scan-: 50 to 800 Frag 120 CAV 5). The AUC value was used to determine the stability of each compound. ## 4.5. Synthesis of A2E A2E (N-retinylidene-N-retinylethanolamine) was synthesized by Orga-link (Magny-Les- Hameaux, France) as described before [4]. ## 4.6. RPE Phototoxicity and Cytokines Expression In Vitro RPE cells were obtained from pig eyes, as previously described [4]. RPE cells were seeded on a 96-well plate at a density of 1.5 × 105 cells/cm2 in DMEM $2\%$ FCS. BIO201, BIO203, or its pure enantiomers were added in increasing concentrations to the culture medium 48 h before illumination. A2E was added to the medium at a final concentration of 30 μM, and 19 h later, blue-light illumination was performed for 50 min using a 96 blue-led device (Durand, St Clair de la Tour, France) emitting at 470 nm (1440 mcd, 8.6 mA). Just before illumination, the culture medium was replaced by a modified DMEM without any photosensitizer and with $2\%$ FCS. Twenty-four hours after blue-light irradiation, all cell nuclei were stained with Hoechst 33342, and nuclei of dead cells with ethidium homodimer 2, fixed with paraformaldehyde ($4\%$ in PBS, 10 min) and 9 pictures per well were captured using a fluorescence microscope (Nikon TiE) equipped with a CoolSNAP HQ2 camera and driven by Metamorph Premier On-Line program. Quantification of live cells was performed using Metamorph Premier Off-Line and a homemade program by subtraction of dead cells from all cells. For cytokines mRNA analysis, RPE cells were seeded in 24-well plates and left untreated or treated with A2E or BIO203 alone or A2E plus BIO203. After 48 h, cells were lysed using the lysis buffer from the NucleoSpin® RNA kit, and the samples were stored at −80 °C. ## 4.7. PPAR, RXR, AP-1, and NF-κB Transactivation Assays Transactivation assays were performed as previously described [4]. Briefly, RPE cells were plated in DMEM $2\%$ FCS at a density of 6 × 104 cells/cm2. The next day transfections using the Cignal Reporter Assay Kits for PPAR, RXR, AP-1, and NF-κB were performed with Lipofectamine and Plus Reagent in serum-free medium. Three hours after the transfection, A2E and/or BIO203, both at 20 μM, were added to the culture medium for 24 h. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System and with a luminometer (Infinite M1000 from Tecan, Mannedorf, Switzerland). At least 3 independent transfections were performed in triplicate for each condition. ## 4.8. Quantitative RT-PCR Total RNA was extracted using the NucleoSpin® RNA kit. Reverse Transcription of 500 ng of RNA was performed using the SuperScript III Reverse Transcriptase. Five ng of cDNA were amplified using the SYBR GREEN real-time PCR method. PCR primers for target genes and housekeeping gene GAPDH were designed using Primer3Plus Bioinformatic software (Table 2). RT-PCR conditions have been described previously [4]. All experimental conditions were processed in triplicate, and each experiment was done at least 3 times. ## 4.9. Acute Ocular PK Studies of Norbixin (BIO201) and BIO203 in Rats Following a Single IP Administration The compounds were administered to 8–9 months-old Male Sprague-Dawley rats ($$n = 3$$ per group and per collection time) by the intraperitoneal route in vehicle (DMSO/NaCl/NaOH for BIO203 and PBS/Tween 80 ($\frac{95}{5}$) for norbixin) with an administration volume of 10 mL/kg. The solutions of BIO203 and norbixin were centrifuged, and the clear supernatants were administered to rats. The final injected concentrations were 10 mg/kg and 8.8 mg/kg for norbixin and BIO203, respectively. Doses administered differ because BIO201 and BIO203 solubility are not equivalent in their respective vehicles, and the dose of BIO203 was reduced to avoid toxicity of DMSO present in the vehicle, but the same volumes were injected into animals. Eyes were collected post-administration according to the following timing: 0, 0.08 h, 0.25 h, 0.5 h, 1 h, 1.5 h, 2 h, 4 h, 8 h, and 24 h. After euthanasia, the two eyes of each rat were collected in a Precellys tube, frozen, and stored at −80 °C. ## 4.10. Ocular PK Studies of BIO201 and BIO203 in Rats Following Four Repeated IP Administrations The compounds were administered to 8–9 months old Male Sprague-Dawley rats ($$n = 3$$ per group and per collection time) by four successive intraperitoneal administrations in vehicle (DMSO/NaCl/NaOH for BIO203 and PBS/Tween 80 ($\frac{95}{5}$) for BIO201). The solutions of BIO203 and BIO201 were centrifuged, and the clear supernatants were administered to rats. Administered volume was 10 mL/kg, and the final injected concentrations were 10 mg/kg and 2.5 mg/kg for BIO201 and BIO203, respectively. Again, doses administered differ because BIO201 and BIO203 solubility are not equivalent in their respective vehicles. The dose of BIO203 was reduced to avoid toxicity of DMSO present in the vehicle, but the same volumes were injected into animals. The eyes were collected 1 h after administrations n°1 to 4 according to the following timing: 1 h, 3.5 h, 5.5 h, and 7.5 h. After euthanasia, the two eyes of each rat were collected in a Precellys tube, frozen, and stored at −80 °C. ## 4.11. Electroretinogram (ERG) in Rats and Mice The ERG in rats was performed using the electrophysiological system RETI-animal® from Roland Consult. ERG was recorded on both eyes from overnight dark-adapted animals. Rats were anesthetized by an intramuscular injection of a mix of Rompun® (xylazine)/Imalgene® (ketamine) before ERG measurement. Ten to fifteen minutes before measurement, one drop of Mydriaticum® ($0.5\%$ tropicamide) was instilled for pupillary dilatation. The A-wave and B-wave implicit times (ms) and amplitudes (µV) were measured for each ERG. The B-wave amplitude (value and %) was given as informative data. ERG parameters: Ganzfeld Q450C, Color: white maximum, Maximum intensity: 3 cd/m2 (0 dB); Duration 0.24 ms; number of flashes: 1, Filter: 50 Hz, Impedance threshold: 50 kΩ. For scotopic ERGs, the A-wave amplitudes were measured from the baseline to the peak of the negative potential, whereas B-wave amplitudes were measured from the trough of the A-wave to the peak of the positive potential. The recordings obtained from these analyses were tabulated and normalized (%) for each eye. Group means, standard deviation, and median values were calculated, considering each eye from one animal as a separate value. ERG recordings in mice were performed as previously described [7]. ## 4.12. Histology and Measurement of the Retina Outer Nuclear Layer (ONL) Thickness in Rats and Mice At the end of the measurement period on Day 7, rats were euthanized by an intraperitoneal injection of pentobarbital. Immediately after euthanasia, both eyeballs were sampled, fixed in Davidson’s solution, and processed for histology. Sections (5 to 7 µm thick) were performed along the vertical meridian and stained with Haematoxylin/Eosin stain. The vertical meridian included the optic nerve. ONL Thickness was first measured at 250 µm from the optic nerve and then every 500 µm (seven points in total) to the peripheral retina in each part of the retina (superior and inferior) using a standard microscope (Leica) on live (pictures were not saved). The thickness of the outer nuclear layer was measured on each point. The ONL areas were calculated by integrating the area under the curve of the retinal thickness from 3.25 mm superior and 3.25 mm inferior to the ON. Histology and photoreceptor counting in mice were performed as previously described [7]. ## 4.13. Blue-Light Damage (BLD) Study in Rats Sixty-four animals were selected based on good health and on scotopic electroretinogram performed before induction when the A-wave amplitude of both eyes was in the range of the global mean (all animals) of A-wave amplitude ± 1.96 SD. Animals were then randomized into the study groups using a macro function in Excel® software based on the A-wave amplitude of both eyes (mean value). On Day 0, BIO203, negative controls, and comparator (N-tert-Butyl-a-phenylnitrone (PBN)) were intraperitoneally injected using a 25 G needle in rats that have been dark-adapted for 36 h. Individually housed rats were then exposed for 6 h (±5 min) to a continuous blue, fluorescent light (400–540 nm) in clear plastic cages. After exposure and administration, the rats were placed in a dark room for 24 h (±1 h) and then were returned to rearing cyclic light conditions. ## 4.14. BIO203-Containing Pellets A custom rodent diet was formulated and irradiated (25 kGy) by Envigo RMS SARL (Gannat, France). Low dose (LD)-BIO203 (50 μg/g) and high dose (HD)-BIO203 (500 μg/g) were incorporated into Teklad Custom Research Diet 2016© pellets. The pellets were stored at −20 °C until use and were administered ad libitum as the standard Teklad Global Rodent Diet 2016© used for the control group. The concentration of BIO203 in the customized pellets at the end of each batch was determined by HPLC MS/MS. The mean concentration was 46 μg/g ± 3.5 μg/g of pellet for the LD and 464 μg/g ± 24.3 μg/g of pellet for the HD. We calculated that in the LD group, male mice weighing 31.63 g after 2 months of complementation consumed 4.16 g of pellets, and in the HD group, male mice weighing 34.11 g after 2 months of complementation consumed 4.11 g of pellets every day, which correspond to a daily dose of 6.04 ± 0.46 mg/kg for the LD-BIO203 pellets and 55.91 ± 2.93 mg/kg for the HD-BIO203 pellets. ## 4.15. In Vivo Oral BIO203 Treatment by Complementation To test the preventive/curative action of oral administration of BIO203 against retinal neurodegeneration, a total of 27 Pigmented Abca4−/− Rdh8−/− mice carrying the Rpe65-Leu450 mutation and the rd8 mutation in the Crb1 gene from Case Western Reserve University [34] of eleven to twelve months of age were used. Three groups of 9 males received control chow (Teklad Custom Research Diet 2016© pellets) or chow containing LD-BIO203 or HD-BIO203 orally for 6 months. After 6 months of complementation, ERG was measured in both eyes. In the control group, two mice died during the period of complementation and one mouse during ERG measurement. In the HD-BIO203 group, one mouse died before the end of complementation, and 3 mice died just after anesthesia. All the mice that received LD-BIO203 were alive until the end of the complementation. One of them died during the photopic ERG measurement. Blood was collected by submandibular puncture in all mice before being euthanized. In each group, half of the eyes were removed for A2E and BIO203 measurements, and half of the eyes were used for histological analyses and photoreceptor layer quantification. ## 4.16. Determination of Norbixin (BIO201) and BIO203 Concentrations in Mice Plasma and Eye Samples Analysis was performed on an LC 1260 System coupled with Mass Spectrometer QQQ-6420 with DAD (Agilent Technologies). Norbixin and BIO203 were eluted from a reverse-phase column (2.1 × 50 mm; 3 μm particles; Ace Excel C18) with the following gradient of acetonitrile in water (both containing $0.1\%$ formic acid): 55 to $100\%$ in 5 min), (flowrate: 0.3 mL/min). BIO203 and its isomers were monitored by a mass spectrometer, in positive MRM mode, with the following transitions, respectively, 381.1 -> 144.9 and 478.1 -> 362.9. For quantification of BIO201 and BIO203, calibration curves were prepared under the same conditions as the sample matrix, with various amounts of BIO201 or BIO203 (25 to 10,000 ng/mL in plasma and 5 to 500 ng/mL for eyes). For determination of norbixin or BIO203 concentrations in plasma, aliquots of plasma samples (25 μL) and internal standard (retinoic acid at 10 µg/mL) were distributed in a 96-well microtiter plate and precipitated with methanol (100 μL). After 10 min mixing, the microtiter plate was frozen at −20 °C for 30 min, thawed, and then centrifuged. The hydro-alcoholic phase was removed from each well and transferred into another microtiter plate for LC-MS/MS analysis. Under these conditions, with 5 μL injections, the limit of quantification (LOQ) was 25 ng/mL (0.33 pmol for BIO201 and 0.26 pmol for BIO203). For determination of concentrations of BIO201 or BIO203 in ocular tissues, each eye was homogenized in CHCl3/MeOH (1:1, v/v) (0.5 mL) with a homogenizer (Precellys-24) during 2 cycles (30 s) at 6500 rpm. The internal standard (retinoic acid at 10 µg/mL) was added, and the organic layer was extracted. The homogenate was then extracted two times with CHCl3/CH2Cl2 (0.5 mL). The combined organic extracts were dried in vacuo without heating (EZ2, Genevac Ltd., Ipswich, UK). Then they were dissolved in 100 μL DMSO/MeOH (1:1, v/v) and transferred to microtiter plates for LC-MS/MS analysis. ## 4.17. A2E Measurement in Mice Eyes A2E quantification in eyes was performed using the HPLC-MS/MS method, which has been described previously [7]. ## 4.18. Statistical Analyses Statistical significance was determined by applying an analysis of variance (one-way ANOVA for one parameter) or a Kruskal–Walli’s test to assess differences among groups (AUC (ONL)). After a significant ANOVA, comparisons between groups were made with a Dunnett’s, Dunn’s, or Tukey’s test according to the homogeneity of variance. A second analysis with a two-way ANOVA for two parameters (ONL thickness and optic nerve distance) was performed to assess differences among groups. After a significant ANOVA, comparisons between groups were made with Tukey’s test. The significant threshold was fixed at 0.05, i.e., the p-value had to be lower than 0.05 to be significant. Tests were performed using Prism 7 (GraphPad Software, La Jolla, CA, USA). ## References 1. Rein D.B., Wittenborn J.S., Zhang X., Honeycutt A.A., Lesesne S.B., Saaddine J.. Forecasting age-related macular degeneration through the year 2050: The potential impact of new treatments. Arch. Ophthalmol. (2009) 127 533-540. DOI: 10.1001/archophthalmol.2009.58 2. Camelo S., Latil M., Veillet S., Dilda P.J., Lafont R.. Beyond AREDS Formulations, What Is Next for Intermediate Age-Related Macular Degeneration (iAMD) Treatment? Potential Benefits of Antioxidant and Anti-inflammatory Apocarotenoids as Neuroprotectors. Oxid. Med. Cell. Longev. (2020) 2020 4984927. DOI: 10.1155/2020/4984927 3. Satyanarayana A., Prabhakara Rao P.G., Rao D.G.. Chemistry processing and toxicology of Anatto (Bixa orellana L). J. Food Sci. Technol. (2003) 40 131-141 4. Fontaine V., Monteiro E., Brazhnikova E., Lesage L., Balducci C., Guibout L., Feraille L., Elena P.P., Sahel J.A., Veillet S.. Norbixin Protects Retinal Pigmented Epithelium Cells and Photoreceptors against A2E-Mediated Phototoxicity In Vitro and In Vivo. PLoS ONE (2016) 11. DOI: 10.1371/journal.pone.0167793 5. Fontaine V., Fournie M., Monteiro E., Boumedine T., Balducci C., Guibout L., Latil M., Sahel J.A., Veillet S., Dilda P.J.. A2E-induced inflammation and angiogenesis in RPE cells in vitro are modulated by PPAR-alpha, -beta/delta, -gamma, and RXR antagonists and by norbixin. Aging (2021) 13 22040-22058. DOI: 10.18632/aging.203558 6. Roehrs M., Figueiredo C.G., Zanchi M.M., Bochi G.V., Moresco R.N., Quatrin A., Somacal S., Conte L., Emanuelli T.. Bixin and norbixin have opposite effects on glycemia, lipidemia, and oxidative stress in streptozotocin-induced diabetic rats. Int. J. Endocrinol. (2014) 2014 839095. DOI: 10.1155/2014/839095 7. Fontaine V., Monteiro E., Fournie M., Brazhnikova E., Boumedine T., Vidal C., Balducci C., Guibout L., Latil M., Dilda P.J.. Systemic administration of the di-apocarotenoid norbixin (BIO201) is neuroprotective, preserves photoreceptor function and inhibits A2E and lipofuscin accumulation in animal models of age-related macular degeneration and Stargardt disease. Aging (2020) 12 6151-6171. DOI: 10.18632/aging.103014 8. Rohini A., Agrawal N., Kumar H., Nath V., Kumar V.. Norbixin, an apocarotenoid derivative activates PPARgamma in cardiometabolic syndrome: Validation by in silico and in vivo experimental assessment. Life Sci. (2018) 209 69-77. DOI: 10.1016/j.lfs.2018.08.001 9. Yang P., McKay B.S., Allen J.B., Jaffe G.J.. Effect of NF-kappa B inhibition on TNF-alpha-induced apoptosis in human RPE cells. Investig. Ophthalmol. Vis. Sci. (2004) 45 2438-2446. DOI: 10.1167/iovs.03-0805 10. Kleiner J., Hollborn M., Wiedemann P., Bringmann A.. Activator protein-1 contributes to the NaCl-induced expression of VEGF and PlGF in RPE cells. Mol. Vis. (2018) 24 647-666. PMID: 30310263 11. Bonilha V.L.. Age and disease-related structural changes in the retinal pigment epithelium. Clin. Ophthalmol. (2008) 2 413-424. DOI: 10.2147/OPTH.S2151 12. Kinnunen K., Petrovski G., Moe M.C., Berta A., Kaarniranta K.. Molecular mechanisms of retinal pigment epithelium damage and development of age-related macular degeneration. Acta Ophthalmol. (2012) 90 299-309. DOI: 10.1111/j.1755-3768.2011.02179.x 13. Neuner B., Komm A., Wellmann J., Dietzel M., Pauleikhoff D., Walter J., Busch M., Hense H.W.. Smoking history and the incidence of age-related macular degeneration--results from the Muenster Aging and Retina Study (MARS) cohort and systematic review and meta-analysis of observational longitudinal studies. Addict. Behav. (2009) 34 938-947. DOI: 10.1016/j.addbeh.2009.05.015 14. Ng E.W., Adamis A.P.. Targeting angiogenesis, the underlying disorder in neovascular age-related macular degeneration. Can. J. Ophthalmol. (2005) 40 352-368. DOI: 10.1016/S0008-4182(05)80078-X 15. Hollyfield J.G., Bonilha V.L., Rayborn M.E., Yang X., Shadrach K.G., Lu L., Ufret R.L., Salomon R.G., Perez V.L.. Oxidative damage-induced inflammation initiates age-related macular degeneration. Nat. Med. (2008) 14 194-198. DOI: 10.1038/nm1709 16. Curcio C.A., Huisingh C.. Probing the Role of Inflammation in Age-Related Macular Degeneration. JAMA Ophthalmol. (2017) 135 843-844. DOI: 10.1001/jamaophthalmol.2017.2008 17. Sparrow J.R., Nakanishi K., Parish C.A.. The lipofuscin fluorophore A2E mediates blue light-induced damage to retinal pigmented epithelial cells. Investig. Ophthalmol. Vis. Sci. (2000) 41 1981-1989. PMID: 10845625 18. Zhou J., Jang Y.P., Kim S.R., Sparrow J.R.. Complement activation by photooxidation products of A2E, a lipofuscin constituent of the retinal pigment epithelium. Proc. Natl. Acad. Sci. USA (2006) 103 16182-16187. DOI: 10.1073/pnas.0604255103 19. Iriyama A., Inoue Y., Takahashi H., Tamaki Y., Jang W.D., Yanagi Y.. A2E, a component of lipofuscin, is pro-angiogenic in vivo. J. Cell. Physiol. (2009) 220 469-475. DOI: 10.1002/jcp.21792 20. Ma W., Coon S., Zhao L., Fariss R.N., Wong W.T.. A2E accumulation influences retinal microglial activation and complement regulation. Neurobiol. Aging (2013) 34 943-960. DOI: 10.1016/j.neurobiolaging.2012.06.010 21. Malek G.. Nuclear receptors as potential therapeutic targets for age-related macular degeneration. Adv. Exp. Med. Biol. (2014) 801 317-321. DOI: 10.1007/978-1-4614-3209-8_40 22. Malek G., Hu P., Wielgus A., Dwyer M., Cousins S.. PPAR nuclear receptors and altered RPE lipid metabolism in age-related macular degeneration. Adv. Exp. Med. Biol. (2010) 664 429-436. DOI: 10.1007/978-1-4419-1399-9_49 23. Malek G., Lad E.M.. Emerging roles for nuclear receptors in the pathogenesis of age-related macular degeneration. Cell. Mol. Life. Sci. (2014) 71 4617-4636. DOI: 10.1007/s00018-014-1709-x 24. Iriyama A., Fujiki R., Inoue Y., Takahashi H., Tamaki Y., Takezawa S., Takeyama K., Jang W.D., Kato S., Yanagi Y.. A2E, a pigment of the lipofuscin of retinal pigment epithelial cells, is an endogenous ligand for retinoic acid receptor. J. Biol. Chem. (2008) 283 11947-11953. DOI: 10.1074/jbc.M708989200 25. Bhosale P., Serban B., Bernstein P.S.. Retinal carotenoids can attenuate formation of A2E in the retinal pigment epithelium. Arch. Biochem. Biophys. (2009) 483 175-181. DOI: 10.1016/j.abb.2008.09.012 26. Moise A.R.. Pharmacology of Retinoid Receptors. Tocris Scientific Review Series. (2019) 27. Krezel W., Ruhl R., de Lera A.R.. Alternative retinoid X receptor (RXR) ligands. Mol. Cell. Endocrinol. (2019) 491 110436. DOI: 10.1016/j.mce.2019.04.016 28. Hiebl V., Ladurner A., Latkolik S., Dirsch V.M.. Natural products as modulators of the nuclear receptors and metabolic sensors LXR, FXR and RXR. Biotechnol. Adv. (2018) 36 1657-1698. DOI: 10.1016/j.biotechadv.2018.03.003 29. IJpenberg A., Tan N.S., Gelman L., Kersten S., Seydoux J., Xu J., Metzger D., Canaple L., Chambon P., Wahli W.. In vivo activation of PPAR target genes by RXR homodimers. EMBO J. (2004) 23 2083-2091. DOI: 10.1038/sj.emboj.7600209 30. Curcio C.A., Medeiros N.E., Millican C.L.. Photoreceptor loss in age-related macular degeneration. Investig. Ophthalmol. Vis. Sci. (1996) 37 1236-1249. PMID: 8641827 31. Curcio C.A., Owsley C., Jackson G.R.. Spare the rods, save the cones in aging and age-related maculopathy. Investig. Ophthalmol. Vis. Sci. (2000) 41 2015-2018. PMID: 10892836 32. Santos A.A.A., Silva M.V.G., Guerreiro L.T.A., Alves M.V., Cunha Bastos V.L.F., Cunha Bastos J., Kovary K.. Influence of norbixin on plasma cholesterol-associated lipoproteins, plasma arylesterase/paraoxonase activity and hepatic lipid peroxidation of swiss mice on a high fat diet. Food Chem. (2002) 77 393-399. DOI: 10.1016/S0308-8146(01)00363-6 33. Bertho G., Oumezziane I.E., Caradeuc C., Guibout L., Balducci C., Dinan L., Dilda P.J., Camelo S., Lafont R., Giraud N.. Structural analysis of unstable norbixin isomers guided by pure shift nuclear magnetic resonance. Magn. Reason. Chem. (2022) 60 504-514. DOI: 10.1002/mrc.5252 34. Maeda A., Maeda T., Golczak M., Palczewski K.. Retinopathy in mice induced by disrupted all-trans-retinal clearance. J. Biol. Chem. (2008) 283 26684-26693. DOI: 10.1074/jbc.M804505200
--- title: 'Role of Portion Size in the Context of a Healthy, Balanced Diet: A Case Study of European Countries' authors: - Michele O. Carruba - Maurizio Ragni - Chiara Ruocco - Sofia Aliverti - Marco Silano - Andrea Amico - Concetta M. Vaccaro - Franca Marangoni - Alessandra Valerio - Andrea Poli - Enzo Nisoli journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049364 doi: 10.3390/ijerph20065230 license: CC BY 4.0 --- # Role of Portion Size in the Context of a Healthy, Balanced Diet: A Case Study of European Countries ## Abstract Over the past decades, a generalised increase in food portion sizes has probably contributed to the growing global obesity epidemic. Increasing awareness of appropriate portion sizes could contribute to reversing this trend through better control of calorie intake. In this study, a comparison of standard portion sizes in European countries for various food categories shows a wide variability of their importance for food, nutrient, and energy consumption according to government and institutional websites. On the other hand, the overall averages appear to be largely in line with the values indicated by the Italian Society of Human Nutrition, which is the most comprehensive and detailed document among those evaluated. The exceptions are milk and yoghurt, for which the reference portions in Europe are generally higher, and vegetables and legumes, for which portions are smaller than those reported in the Italian document. Moreover, the portion sizes of staple foods (e.g., pasta and potatoes) vary according to different food traditions. It is reasonable to consider that the creation of harmonised standard reference portions common to the European countries, based on international guidelines and scientific evidence, would significantly contribute to consumers’ nutritional education and ability to make informed choices for a healthy diet. ## 1. Introduction A portion is generally defined as the amount of food people intend to consume on one eating occasion [1]. In its generic meaning as the “quantity or allowance of food allotted to, or enough for, one person” (Oxford English Dictionary, 2023), food portion may, therefore, not coincide (and be greater or smaller) with the so-called standard reference portion generally defined by experts or institutions at the national level [2]. For example, in Italy, a portion is defined as “the quantity of a food that is assumed to be a reference unit recognised and identifiable by both nutritional professionals and the general public” [3]; therefore, in keeping with dietary tradition, it should be of a reasonable size that follows consumer expectations and be expressed in units of measurement that refer to natural or commercial units or common household units. Standard portions are the references for the quantities of different foods to be recommended as part of diets for various age groups or groups with specific nutritional needs (e.g., pregnancy and breastfeeding) and to be indicated on the nutritional labelling of food products (such as those defined by the Academy of Nutrition and Dietetics in the United States). They are usually defined according to specific criteria, such as the history of use, product density, and intake data [4,5]. In recent years, the concept of portion size has been considered central by national (e.g., the Italian Consiglio per la Ricerca in Agricoltura e l’Analisi dell’Economia Agraria and the National Health Institute) and international organisations (e.g., World Health Organisation) [6,7,8], which include it among the determinants of dietary balance. Recently with co-funding from the European Union, the British Nutrition Foundation published a guide dedicated to portions, specifying how a healthy and balanced diet is determined by the choice of foods that make up such a diet and the appropriate quantities of the foods themselves [9]. The graphical representation of the Mediterranean Diet—a dietary pattern that a growing body of evidence confirms to be associated with a lower prevalence of chronic degenerative diseases and related risk factors [10]—is also based on the concept of portion size, as well as the frequency and variety of different food categories [11]. In particular, “moderation in portion size” is emphasised as a determinant for adjusting dietary intake to fit the specific needs of persons in modern societies, which are characterised by increasingly sedentary lifestyles, as well as for eliminating waste, thus favouring the sustainability of dietary patterns [12,13]. According to the of the Academy of Nutrition and Dietetics guidelines, no product is to be excluded if consumed in moderation and with adequate quantities as part of an overall balanced diet, thereby emphasising variety and moderation in the context of a healthy lifestyle to help reduce consumer confusion [14]. In fact, energy intake results from the size of the portions consumed and the frequency of consumption itself (i.e., the number of consumption occasions in a given period of time) [15]. An increase in one of these two factors, if not adequately compensated by a reduction in the other, can lead to an altered calorie intake. Extensive literature has focused on the association between portion size and overweight/obesity, from the first ecological studies in the 1970s to the present. In the United States, for example, a progressive increase in the portion sizes of specific foods distributed in fast food chains and restaurants over thirty years (1977 to 2006) is considered to be responsible for an increase in energy intake, with some cases exceeding 100 kcal per unit of sale [16,17]. This phenomenon has been associated with a concomitant increase in the prevalence of overweight and obesity in the US population over the same period [18,19]. Similar associations have also been described in British children [20] and European adolescents [21]. Increased awareness of portion sizes could, therefore, play an essential educational role in promoting healthier eating and improving the consumption of certain foods, such as fruits, vegetables, and pulses, that have emerged as the determinants of the health benefits derived from specific eating patterns (e.g., the Mediterranean diet)—as already evident in some observational and intervention studies [11,22,23,24]. Our present analysis aims to conduct a comparative examination of standard portion claims for different food categories in European countries, which is crucial for a critical assessment of the possibility of defining harmonised portions to be adopted in different areas, ranging from possible nutrition policies to front-of-pack labelling proposals and food education strategies. ## 2. Materials and Methods The information included in this analysis was obtained from documents published in various Member States of the WHO European Region by governmental bodies or scientific societies available online [25]. In particular, an extensive web interrogation was carried out to find institutional documents reporting information on portion sizes that apply to the adult population as the reference standards by using selected keywords translated from different European languages, such as “nutrition/food-based guidelines/recommendations”, “food policy”, “portion”, and “reference/standard portion”. We used the conversion table published by the Italian Society of Human Nutrition (available at https://sinu.it/wp-content/uploads/$\frac{2019}{07}$/20141111_LARN_Porzioni.pdf, accessed on 20 October 2022) to translate the various units of measurement (e.g., slices or cups) as they are indicated in the documents of the different countries analysed into a standard unit of measurement (i.e., grams, units, or millilitres). Since our study aimed to exclusively describe and highlight differences in portion sizes compared to the average, we did not consider assessing or evaluating their statistical significance necessary. This will be the aim of additional analysis of these data that will be elaborated on in a forthcoming manuscript. ## 3.1. Standard Portion Sizes in EU Countries: The General Situation Table 1 lists the 34 European countries for which relevant public documents (e.g., guidelines and government recommendations) were obtained as described. Critical issues found in the comparative evaluation include the reference for some countries pertains to the serving rather than to the portion (see, for example, Portugal) and, explicitly concerning grains (pasta and rice), the different use of grains in the preparation of meals (i.e., main course in some cases and accompaniment of dishes in others), which can lead to substantial differences in portion sizes. All available data for the different food categories are listed in Table 2a,b for plant- and animal-based foods, respectively. The values expressed for different units, such as slices, cups, and spoons, have been reported in grams or millilitres where possible. The average values calculated for all 34 countries and separately for the 24 EU and 10 non-EU countries are shown in Table 3. A comparison was also performed with the reference portion sizes, which have been defined by the Italian Society of Human Nutrition in the “Reference intake levels of Nutrients and energy for the Italian population” (LARNs) as a valuable tool for nutritional research and dietary planning [3]. The sometimes considerable difference between the minimum and maximum values recorded for the portion sizes in the countries reflects the variability in the portions within the same food category. However, a comparison of the median values calculated for the EU and non-EU countries confirms the existence of some homogeneity between the two groups of countries. A comparison with the standard reference portion sizes defined by the Italian Society of Human Nutrition also shows that, in most cases, the average values, both total and separately for the EU and non-EU countries, do not differ significantly from the corresponding Italian figures [3]. The exceptions are nuts, potatoes, vegetables, fish, legumes, and fresh cheese, for which the average portion sizes in the analysed countries are lower than the Italian reference portions, and milk and breakfast cereals, for which the standard portion sizes are generally higher in other countries than in Italy. Analysing the average portion sizes across the countries according to geographical location (Southern, Central, and Northern Europe) allows us to appreciate a certain homogeneity of the data (Figure 1). The Italian reference values are higher for vegetables (200 g vs. 150 g on average, regardless of geographical area), fish and legumes (150 g for the Italian portion for each food category vs. an average value of just over 100 g for each food category for the other countries), and much smaller for a single portion of milk, which corresponds to a small glass (125 mL) in Italy, in comparison to an average of more than 200 mL (almost a large cup) in the other countries (Figure 1). Supplementary Figures S1–S18 show the portion sizes measured for the different food categories in the diverse countries compared with the relative average values and the reference portion established by the Italian Society of Human Nutrition. ## 3.2. Differences and Similarities in Foods Portions in European Countries According to the available documents, the amount of bread corresponding to one portion (50 g in Italy and seven other countries, including France, Spain, Austria, Germany, and Finland) varies from 15 g in Slovenia to 100 g in Iceland, Serbia, and Switzerland (Supplementary Figure S1). For “Pasta, rice, maize, barley, spelt”, the generic portion size indicated is 80 g for Italy, as well as for Germany, Malta, Norway, Spain, and Sweden (Supplementary Figure S2). Larger portions (above 100 g) are suggested in Belgium and Turkey; similarly, Cyprus, Hungary, Lithuania, and Slovakia adopt a portion of 100 g. In other countries, the portions are smaller than the average: 50 g in Estonia, 35 g in Portugal, and 30–35 g for pasta and 20 g for rice in Croatia. Interestingly, for some countries, the portion also refers to the cooked product: 110 g for Portugal, 70 g for Estonia, 200 g for Hungary, 125 g for the Czech Republic, and 150 g for Lithuania. Average portion sizes for ready-to-eat breakfast cereals and potatoes are 39 g and 152 g, respectively (Supplementary Figures S3 and S4). For fresh fruits, a portion size of about 150 g prevails in 12 countries, as in Italy, which also coincides with the general average (Supplementary Figure S5). In the Czech Republic, Estonia, Hungary, and Norway, the reference portion is set at 100 g, slightly higher than the 80 g suggested in Malta, Slovenia, and the United Kingdom. In Belgium and Germany, the portions are considerably higher than the average (250 g). For seven countries, two reference portions prevail for vegetables and greens: 100 g and 200 g. The average size is nearly 160 g, ranging from 80 g in Malta, Poland, Slovenia, and the United Kingdom to 300 g in Armenia and Hungary (Supplementary Figure S7). Concerning leafy vegetables, the standard portions are generally lower, at 80 g in Italy and ranging from 50 g in Lithuania to 200 g in Croatia. The portion size for fresh or canned pulses (Supplementary Figure S8) is 150 g in Italy, as well as in Austria, Greece, and Ireland, compared to an overall average of just over 110 g and a peak of 200 g in Lithuania. The portion size of dried legumes, which in Italy is defined as 50 g, is also reported in Estonia (10 g), Portugal and Croatia (25 g), Poland (40–60 g), Spain (60–80 g), Malta (70 g), and finally Slovenia (4 tablespoons or 50 g) (Supplementary Figure S9). For red meat, the 100 g stated by the LARNs for the reference portion size is shared by Armenia, Hungary, Iceland, the Netherlands, North Macedonia, Switzerland, and Turkey and is very close to the overall average value, which ranges from 150 g in Norway, Poland, and Slovenia to 30 g in Portugal, Cyprus, and Croatia (Supplementary Figure S10). For white meat, the portion sizes remain the same as for red meat in Italy, Malta, Portugal, Hungary, Cyprus, Poland, the Czech Republic, and Croatia (100 g), which coincides with the overall average (Supplementary Figure S11). The portion size suggested for fish is sometimes larger than for meat, not only in Italy (150 g, as well as in Greece, Hungary, Austria, and Poland) but especially in Norway (175 g) and Spain (180 g) (Supplementary Figure S13). The overall average is about 110 g. The reference portion for eggs corresponds to one egg in most countries. A fair variability is observed for the reference portions of milk (Supplementary Figure S14): the smallest, corresponding to 125 mL, is the one defined in Italy (i.e., one glass), as well as in Belgium and Lithuania. In other countries, larger portion sizes prevail (one “cup” or similar, corresponding to 200–250 g); consequently, the overall average is around 210 mL. Additionally, for yogurt, the 125 g indicated by the LARNs in Italy corresponds to the smallest portion, with an overall average of around 180 g (Supplementary Figure S15). The portions of cheeses, separated in most cases into fresh and hard cheeses (soft and hard), show considerable differences, with the averages being below the LARN values in both cases (Supplementary Figures S16 and S17): 75 g against 100 g for the leanest cheeses and just over 35 g against 50 g for the most aged cheeses. Concerning vegetable oils, including olive oil, the standard portion size is 10 g, not only in Italy but also in Albania, France, Georgia, Germany, Portugal, Serbia, Spain, Sweden, and Switzerland (Supplementary Figure S18). In Italy, standard portions are also defined for sweet food products: “cakes, spoon sweets, and ice cream”, “snacks, crisps and chocolate bars”, “sugar”, and “honey and jam” (for the latter two categories, the portions are also defined in Hungary and Estonia). More scattered are the reference portion values for beverages. A single portion of water is 200 mL in Italy, Spain, and Portugal; between 200 and 250 mL in Hungary; and 250 mL in Malta, Greece, and Slovenia. For fruit juices, iced tea, and soft drinks, the standard portion size is 200 mL (330 mL if in a can) in Italy, 100 mL in Estonia, 250 mL in Hungary (200 mL if in a can), 150 mL in Spain, and 125 mL in Greece. Finally, the consumption unit for wine is 125 mL in Italy and Greece, 100 mL in Slovenia and Spain, and 80 mL in Malta. The reference portion rises to 330 mL for beer in Italy and Greece, 250 mL in Slovenia and Malta, and 200 mL in Spain and Portugal. For spirits, the portion size is smaller: 25 mL in Malta, 30 mL in Slovenia, 40 mL in Italy, and 40–45 mL in Greece. ## 4. Discussion The analysis of the documents available on government and institutional websites shows how attention to the concept of portion size, as a determining element of food, nutrient, and energy consumption, is heterogeneous in the countries considered in this study. On the other hand, the evaluation of the overall averages reveals a substantial overlap in most cases with the values indicated by the LARNs as defined by the Italian Society of Human Nutrition, which represents the most detailed and complete document among those examined [3]. The exceptions are the cases of milk and yoghurt, for which most of the reference portions in Europe are larger than in Italy, and the portions of vegetables and legumes are smaller than those reported in the Italian document. In general, portions of foods that are more likely to be portioned per se, such as fruits, potatoes, and fish, tend to be more homogeneous than liquids or foods that are more difficult to refer to in terms of consumption units, such as milk and cheese. Another general observation concerns portion sizes, which are larger overall for staple foods (such as pasta and potatoes) and, therefore, vary according to the food traditions of the different countries: the same portion of pasta has a higher weight if it is considered a staple food than in countries where it is simply an accompaniment to dishes. Previous research has reported the need for more documents focusing on reference portion sizes in other EU countries [59,60]. However, the need to convey to the population indications regarding actual food and drink intakes stems mainly from the observation that larger portions encourage food intake and, if this is in excess, contribute to an increase in calorie intake and, consequently, in the risk of the onset of overweight and obesity [61]. The possible role of portions in determining food consumption quantities has been confirmed by a Cochrane Review [62]. The authors proposed that the size of tableware used at home should also be reduced to help improve food choices and consumption in quantitative terms. Based on this review’s results, however, reducing portion sizes does not appear to be a uniformly effective strategy: the available data show that reducing portions at the larger end of the size range leads to a reduction in food intake but does not allow a recommendation regarding whether reducing portions at the smaller end of the size range would be equally effective. Interestingly, a clear and significant association between portion sizes in school canteens and the risk of excess weight was described in a study conducted on a population of Italian children: the obesity rate was higher among children exposed to larger portions that were further away from the standard recommended portion sizes for children of the same age [63]. Confirmation of the role of portion sizes in overall food intake comes from a meta-analysis of 58 studies (with a total of 6603 participants): the effect that portion size, packaging, single units, or household tableware can have on the food consumption levels of adults and children, although statistically small or moderate, is significant. This suggests that controlling consumption units and, thus, portions, both at home and when outside, could effectively limit average energy intake by a relatively large proportion (about $10\%$ on average, according to the authors) [62]. On the other hand, it has been hypothesised that the inability to recognise the amount of food consumed on a single occasion may be an obstacle to controlling food intake [64]. The effect of the failure to regulate food intake with prolonged exposure to larger portions has been described in preschool children, particularly those with higher body weights, challenging the assumption that self-regulatory behaviour would be sufficient to counteract perturbations in energy intake [65]. Several factors can modulate this portion size effect [66]. This effect is particularly evident in the conditions in which the size of the portions directly influences the quantity of food consumed, as in the case of the Italian school canteens discussed above. A similar effect, although probably less marked in terms of quantity, can also be hypothesised for industrial foods presented in a clearly portioned form. Sharing the concept of standardised portions with food companies could, therefore, expose consumers to more appropriate amounts of packaged foods by helping companies correctly recognise the quantities of food to be consumed on a single occasion. This alliance between institutions and food companies should be vigorously pursued. On the other hand, the size of the sales unit is one of many variables to be considered. More variable and subjective factors also play a role in determining this effect, which is linked to both the individual and his/her context: for example, in physiological conditions, people who eat more quickly or have difficulty perceiving a sense of satiety are inclined to consume larger portions [67]. Furthermore, eating in the company of others or alone can condition the portions of food or drink consumed. The aforementioned Italian study highlighted the role of exposure to adequate portion sizes for children, who were more likely to be of normal weight if they had access to school canteens that provided meals quantitatively in line with nutritional recommendations [63]. On the other hand, it is clear that when a consumer decides autonomously the quantity of food to consume and does not find “ready-made” portions as in the school canteens mentioned above, the information in his/her possession becomes decisive. It is, in fact, undeniable that knowledge of the actual portion size can contribute to making consumption choices that are more balanced and conscious. A survey of more than 13,000 people in 6 European countries (Germany, United Kingdom, Spain, France, Poland, and Sweden) showed that consumers with a higher focus on health-related topics were also those who found information about the amount of product per portion most relevant [68]. Furthermore, almost all respondents in the various countries agreed on the definition of portion as the amount of food that should be consumed; only the respondents in Sweden identified the amount of food they could eat as a portion. By using a sequence of ad hoc distance tests, this study established how the availability of nutritional information on product packaging per portion, as well as per 100 g or 100 mL, could help consumers make more informed food choices: in fact, the percentage of respondents who were able to identify the amount of nutrients and energy per portion increased significantly (for saturated fat, for example, from $15\%$ to $86\%$), and in a much shorter time, moving from the situation where the presence of nutritional composition was only referred to as per 100 g of product to one where the same information was reported per portion [69]. A large-scale communication and information campaign on the concept of portions could, therefore, greatly help improve the quantitative aspects of the general population’s food consumption. In this regard, the use of front-of-pack labelling, which is being studied in the European community, could also contribute, provided that it is based on the adequate portion to be consumed. Such an approach would also make it possible to integrate this quantitative information with information on the qualitative composition of foods, thus facilitating the choices necessary to follow a balanced and complete food pattern. ## Standard Portions vs. Consumption Units The potential importance of the definition of standard reference portions to be used on nutrition labels in the European Union was pointed out by the authors of the Food4Me study, a multi-center, web-based protocol to determine the effects of personalised versus conventional nutrition communication to the general population [60]. The analysis of the food choices of selected population groups in Germany, Greece, Ireland, the Netherlands, Poland, Spain, and the United Kingdom revealed an overall homogeneity of the portions consumed in each country when compared to the national averages, but also a difference between the average portions in use in the different countries, which was significant for $42\%$ of the 156 foods considered from various categories. The most significant differences were in the categories of “grains”, “potatoes”, “rice and pasta”, and “meat and fish”; less heterogeneous were the consumption unit sizes for “soups”, “sauces and condiments”, “fats and spreads”, and “fruit”. The average portion sizes consumed differed from the weighted average for the study population in all seven countries in only $15.7\%$ of the cases, suggesting that attention should, therefore, be paid to those products which portion sizes, in general, differed significantly from the weighted average, and to those countries (mainly Ireland among the countries in the study) where portion sizes differed significantly more than the average. Acknowledging the lack of standardised portion sizes common to European countries at the time of publication, the authors themselves admitted that the typical portion sizes in use in different countries might even differ significantly from the 100 g or 100 mL indicated by the European legislation, pointing out that the harmonisation of standard portions is of potentially great value for the definition of reference quantities based on which nutritional information can be provided, which is in agreement with the results of previous research [69,70]. This aspect appears to be crucial: information on the nutritional composition of individual portions or sales/consumption units of a product is recognised as valuable support for consumers, who are, thus, able to immediately understand the amount of energy and nutrients obtained with the consumption of that food (Food Standards Australia and New Zeland) [71]. On the other hand, consumer confusion regarding the concept of adequate portion sizes for different foods is clear from the available literature [2]. Indeed, the authors of a review of the results of five research studies emphasise the importance of providing accurate information on the size of different portion sizes in line with the recommendations for healthy eating that are fundamental for proper nutrition education [72]. The appropriateness of standardisation of reference portion sizes based on nutritional guidelines also emerges strongly from the findings of an analysis of food products for sale in the Australian market and a review of available literature aiming to create an effective strategy to promote the production of more appropriate sales/consumption units by companies [73], to provide consumers with the information they need to align intake levels with recommendations [11], and to improve the quality of nutritional information on packaged product labels [19,74]. The observation that has emerged from the available studies points out that the definition of standardised and harmonised portions in the European Union could advance the definition of nutritional recommendations and, therefore, promote educational programmes on healthy eating that are common to various countries, while respecting the different eating habits and traditions and facilitate the communication of nutritional information by referring to the quantity of product consumed (rather than to 100 g or 100 mL). This could have a twofold result: educating people to consume different foods in appropriate quantities that are compatible with a balanced diet and helping them understand the contribution of foods to the overall diet [1,75]. Finally, the definition of reference food portions appears to be a proper strategy to improve the quality of diet and, thus, consumer well-being and reduce food waste and its associated costs [76]. ## 5. Conclusions In conclusion, the dissemination of the concept of portion sizes for various foods and the definition, based on international guidelines and scientific evidence, of harmonised standard reference portions common to the countries of the EU could make a significant contribution to the improvement of nutritional information provided to consumers and, therefore, enhance their ability to make informed choices for an overall healthy diet. Portion sizes can differ significantly depending on the country or culture. Cultural attitudes towards food, cuisine, eating habits, and food availability and affordability can influence portion sizes. It is important to note that portion sizes can also vary within a country based on regional differences, socioeconomic status, and personal preferences. However, an analysis of the reference portions available in the European countries to date suggests that the creation of harmonised standard portions for the main categories of food is desirable and decidedly practicable, following the example of countries that have made progress in this direction, such as Italy. However, particular attention must be paid to selected food categories, such as vegetables, legumes, fish, milk, and derivatives, for which the available scientific evidence supports the promotion of consumption that can easily be achieved by indicating more consistent reference portions in line with needs. Based on these considerations, the indication of harmonised standard reference portions appears relevant, especially for the definition of simplified front-of-pack labelling systems to optimise the transmission of correct information to consumers to allow them to contextualise the consumption of foods with different nutritional characteristics in an overall balanced diet. ## References 1. Lewis H.B., Ahern A.L., Jebb S.A.. How much should I eat? A comparison of suggested portion sizes in the UK. Public Health Nutr. (2012.0) 15 2110-2117. DOI: 10.1017/S1368980012001097 2. Faulkner G.P., Pourshahidi L.K., Wallace J.M., Kerr M.A., McCrorie T.A., Livingstone M.B.. Serving size guidance for consumers: Is it effective?. Proc. Nutr. Soc. (2012.0) 71 610-621. DOI: 10.1017/S0029665112000766 3. SICS. Reference Intake Levels of Nutrients and Energy for the Italian Population—LARN (Livelli Di Assunzione Di Riferimento Di Nutrienti Ed Energia Per La Popolazione Italiana) (2014.0) 4. 4. NHS Rough Guide—Fruit & Vegetable Portion SizesNHSLondon, UK2022. Rough Guide—Fruit & Vegetable Portion Sizes (2022.0) 5. Food Fact Sheet 2019. (2019.0) 6. 6. World Health Organization Diet, Nutrition and the Prevention of Chronic Diseases: Report of a Joint WHO/FAO Expert ConsultationWHOGeneva, Switzerland2003Volume 916149. Diet, Nutrition and the Prevention of Chronic Diseases: Report of a Joint WHO/FAO Expert Consultation (2003.0) Volume 916 149 7. Just Enough for You. In. (2009.0) 8. Alimenti e nutrizione.. (2019.0) 9. 9. British Nutrition Foundation Find Your Balance Get Portion Wise!British Nutrition FoundationLondon, UK2019. Find Your Balance Get Portion Wise! (2019.0) 10. Sofi F., Macchi C., Abbate R., Gensini G.F., Casini A.. Mediterranean diet and health status: An updated meta-analysis and a proposal for a literature-based adherence score. Public Health Nutr. (2014.0) 17 2769-2782. DOI: 10.1017/S1368980013003169 11. Bracale R., Vaccaro C.M., Coletta V., Cricelli C., Gamaleri F.C., Parazzini F., Carruba M.. Nutrition behaviour and compliance with the Mediterranean diet pyramid recommendations: An Italian survey-based study. Eat. Weight Disord. (2020.0) 25 1789-1798. DOI: 10.1007/s40519-019-00807-4 12. Bach-Faig A., Berry E.M., Lairon D., Reguant J., Trichopoulou A., Dernini S., Medina F.X., Battino M., Belahsen R., Miranda G.. Mediterranean diet pyramid today. Science and cultural updates. Public Health Nutr. (2011.0) 14 2274-2284. DOI: 10.1017/S1368980011002515 13. Dernini S., Berry E.M., Serra-Majem L., La Vecchia C., Capone R., Medina F.X., Aranceta-Bartrina J., Belahsen R., Burlingame B., Calabrese G.. Med Diet 4.0: The Mediterranean diet with four sustainable benefits. Public Health Nutr. (2017.0) 20 1322-1330. DOI: 10.1017/S1368980016003177 14. Freeland-Graves J.H., Nitzke S.. Academy of N. Dietetics. Position of the academy of nutrition and dietetics: Total diet approach to healthy eating. J. Acad. Nutr. Diet. (2013.0) 113 307-317. DOI: 10.1016/j.jand.2012.12.013 15. Higgins K.A., Hudson J.L., Hayes A.M.R., Braun E., Cheon E., Couture S.C., Gunaratna N.S., Hill E.R., Hunter S.R., McGowan B.S.. Systematic Review and Meta-Analysis on the Effect of Portion Size and Ingestive Frequency on Energy Intake and Body Weight among Adults in Randomized Controlled Feeding Trials. Adv. Nutr. (2022.0) 13 248-268. DOI: 10.1093/advances/nmab112 16. Nielsen S.J., Popkin B.M.. Patterns and trends in food portion sizes, 1977–1998. JAMA (2003.0) 289 450-453. DOI: 10.1001/jama.289.4.450 17. Piernas C., Popkin B.M.. Food portion patterns and trends among U.S. children and the relationship to total eating occasion size, 1977–2006. J. Nutr. (2011.0) 141 1159-1164. DOI: 10.3945/jn.111.138727 18. Harnack L.J., Jeffery R.W., Boutelle K.N.. Temporal trends in energy intake in the United States: An ecologic perspective. Am. J. Clin. Nutr. (2000.0) 71 1478-1484. DOI: 10.1093/ajcn/71.6.1478 19. Young L.R., Nestle M.. The contribution of expanding portion sizes to the US obesity epidemic. Am. J. Public Health (2002.0) 92 246-249. DOI: 10.2105/AJPH.92.2.246 20. Howell Davies O., Suleiman S., Nicholas J., Bradbury J., Msebele S., Prior G., Hall L., Wreford S., Jarvis L., McGee A.. Food portion weights in primary and secondary school lunches in England. J. Hum. Nutr. Diet. (2008.0) 21 46-62. DOI: 10.1111/j.1365-277X.2007.00846.x 21. Flieh S.M., Miguel-Berges M.L., Gonzalez-Gil E.M., Gottrand F., Censi L., Widhalm K., Manios Y., Kafatos A., Molnar D., Dallongeville J.. The Association between Portion Sizes from High-Energy-Dense Foods and Body Composition in European Adolescents: The HELENA Study. Nutrients (2021.0) 13. DOI: 10.3390/nu13030954 22. Campbell M.K., Demark-Wahnefried W., Symons M., Kalsbeek W.D., Dodds J., Cowan A., Jackson B., Motsinger B., Hoben K., Lashley J.. Fruit and vegetable consumption and prevention of cancer: The Black Churches United for Better Health project. Am. J. Public Health (1999.0) 89 1390-1396. DOI: 10.2105/AJPH.89.9.1390 23. Rolls B.J., Roe L.S., Meengs J.S.. Portion size can be used strategically to increase vegetable consumption in adults. Am. J. Clin. Nutr. (2010.0) 91 913-922. DOI: 10.3945/ajcn.2009.28801 24. Haydon P., Tobi R., Bridge G., Gurung I., Scott C.. Peas Please: Making a pledge for more veg. Nutr. Bull. (2020.0) 45 483-494. DOI: 10.1111/nbu.12463 25. 25. World Health Organization Food-Based Dietary Guidelines in the WHO European RegionWHOCopenhagen, Denmark2003. Food-Based Dietary Guidelines in the WHO European Region (2003.0) 26. Recommendations on Healthy Nutrition in Albania. (2008.0) 27. Food Based Dietary Guidelines in the WHO European Region. World Health Organization. (2003.0) 28. Ministry of Health of Austria the Austrian Food Pyramid—7 Steps to Health. (2010.0) 29. Dietary Guidelines for the Belgian Adult Population. Publication of the Superior Health Council No. 9284. (2019.0) 30. Ministry of Health of Croatia Dietary Guidelines for Adults. (2002.0) 31. Cyprus Ministry of Health Nutrition and Exercise Guidelines. (2007.0) 32. Nutritional Recommendations for the Population of the Czech Republic 33. The Official Dietary Guidelines—Good for Health and Climate. (2020.0) 34. National Institute for Health Development Estonian Recommendations for Nutrition and Exercise 2015 35. Finnish Food Authority—Nutrition and Food Recommendations. (2023.0) 36. Actualisation des Repères du PNNS: Révision des Repères de Consommations Alimentaires Avis de l’Anses Rapport D’expertise Collective. (2016.0) 37. World Health Organization. Health Eating—The Main Key to Health. (2005.0) 38. Oberritter H., Schäbethal K., von Ruesten A., Boeing H.. The DGE-Nutrition Circle—Representation and Fundamentals of the Food-Based Recommendations of the German Nutrition Society. Ernaehrungs Umschau International 2013 39. Institute of Preventive Medicine, Environmental and Occupational Health, Prolepsis. National Dietary Guide for Adults. (2014.0) 40. Dietary Guidelines for the Adult Population in Hungary. (2004.0) 41. Dietary Recommendations for Adults and Children from Two Years of Age. (2016.0) 42. Food Safety Authority of Ireland Healthy Eating, Food Safety and Food Legislation—A Guide Supporting the Healthy Ireland Food Pyramid. (2019.0) 43. Nutrition Recommendations for Adults 2020 44. Recommendations for a Healthy and Sustainable Diet 2020 45. Dietary Guidelines for Maltese adults: Healthy Eating—The Mediterranean Way. (2015.0) 46. Dutch Dietary Guidelines 2015 47. Institute of Public Health of the Republic of Macedonia, Department of Physiology and Monitoring of Nutrition Dietary Guidelines 2013 48. Nordic Council of Ministers Nordic Nutrition Recommendations 2012 Integrating Nutrition and Physical Activity. (2014.0) 49. Nutritional Recommendation for the Polish Population and Their Application. (2020.0) 50. The New Food Wheel, a Guide to a Daily Food Choice. (2003.0) 51. Gurinović M., Milešević J., Novaković R., Kadvan A., Djekić-Ivanković M., Šatalić Z., Korošec M., Spiroski I., Ranić M., Dupouy E.. Improving nutrition surveillance and public health research in Central and Eastern Europe/Balkan Countries using the Balkan Food Platform and dietary tools. Food Chem. (2016.0) 193 173-180. DOI: 10.1016/j.foodchem.2015.03.103 52. Public Health Authority of the Slovak Republic Ten Rules of a Healthy Plate 53. Healthy Plate: Recommendations for Healthy Eating. (2009.0) 54. Guía de la Alimentación Saludable Guide to Healthy Eating. (2004.0) 55. The Swedish Dietary Guidelines. (2015.0) 56. The Swiss Food Pyramid. Dietary Recommendations for Adults, Combining Pleasure and Balance 57. Dietary Guidelines for Turkey. (2006.0) 58. NHS Easy to Eat Well 2017 59. Montagnese C., Santarpia L., Buonifacio M., Nardelli A., Caldara A.R., Silvestri E., Contaldo F., Pasanisi F.. European food-based dietary guidelines: A comparison and update. Nutrition (2015.0) 31 908-915. DOI: 10.1016/j.nut.2015.01.002 60. Kirwan L., Walsh M.C., Brennan L., Gibney E.R., Drevon C.A., Daniel H., Lovegrove J.A., Manios Y., Martinez J.A., Saris W.H.. Comparison of the portion size and frequency of consumption of 156 foods across seven European countries: Insights from the Food4ME study. Eur. J. Clin. Nutr. (2016.0) 70 642-644. DOI: 10.1038/ejcn.2015.227 61. Almiron-Roig E., Forde C.G., Hollands G.J., Vargas M.A., Brunstrom J.M.. A review of evidence supporting current strategies, challenges, and opportunities to reduce portion sizes. Nutr. Rev. (2020.0) 78 91-114. DOI: 10.1093/nutrit/nuz047 62. Hollands G.J., Shemilt I., Marteau T.M., Jebb S.A., Lewis H.B., Wei Y., Higgins J.P., Ogilvie D.. Portion, package or tableware size for changing selection and consumption of food, alcohol and tobacco. Cochrane Database Syst. Rev. (2015.0) 2015 CD011045. DOI: 10.1002/14651858.CD011045.pub2 63. Balzaretti C.M., Ventura V., Ratti S., Ferrazzi G., Spallina A., Carruba M.O., Castrica M.. Improving the overall sustainability of the school meal chain: The role of portion sizes. Eat. Weight Disord. (2020.0) 25 107-116. DOI: 10.1007/s40519-018-0524-z 64. Benton D.. Portion size: What we know and what we need to know. Crit. Rev. Food Sci. Nutr. (2015.0) 55 988-1004. DOI: 10.1080/10408398.2012.679980 65. Smethers A.D., Roe L.S., Sanchez C.E., Zuraikat F.M., Keller K.L., Kling S.M.R., Rolls B.J.. Portion size has sustained effects over 5 days in preschool children: A randomized trial. Am. J. Clin. Nutr. (2019.0) 109 1361-1372. DOI: 10.1093/ajcn/nqy383 66. Zuraikat F.M., Smethers A.D., Rolls B.J.. Potential moderators of the portion size effect. Physiol. Behav. (2019.0) 204 191-198. DOI: 10.1016/j.physbeh.2019.02.043 67. Wansink B., Chandon P.. Can “low-fat” nutrition labels lead to obesity?. J. Mark. Res. (2006.0) 43 605-617. DOI: 10.1509/jmkr.43.4.605 68. Raats M., Wills J., Kennedy J.. Consumer preference for and use of portion information on food and drink packaging: A Pan-European Study. Ann. Nutr. Metab. Eur. J. Nutr. Metab. Dis. Diet. (2011.0) 58 69. Raats M.M., Hieke S., Jola C., Hodgkins C., Kennedy J., Wills J.. Reference amounts utilised in front of package nutrition labelling; impact on product healthfulness evaluations. Eur. J. Clin. Nutr. (2015.0) 69 619-625. DOI: 10.1038/ejcn.2014.190 70. Kliemann N., Kraemer M.V.S., Scapin T., Rodrigues V.M., Fernandes A.C., Bernardo G.L., Uggioni P.L., Proenca R.P.C.. Serving Size and Nutrition Labelling: Implications for Nutrition Information and Nutrition Claims on Packaged Foods. Nutrients (2018.0) 10. DOI: 10.3390/nu10070891 71. Curran M.A.. Nutrition labelling: Perspectives of a bi-national agency for Australia and New Zealand. Asia Pac. J. Clin. Nutr. (2002.0) 11 S72-S76. DOI: 10.1046/j.1440-6047.2002.00002.x 72. Bucher T., Murawski B., Duncanson K., Labbe D., Van der Horst K.. The effect of the labelled serving size on consumption: A systematic review. Appetite (2018.0) 128 50-57. DOI: 10.1016/j.appet.2018.05.137 73. Haskelberg H., Neal B., Dunford E., Flood V., Rangan A., Thomas B., Cleanthous X., Trevena H., Zheng J.M., Louie J.C.. High variation in manufacturer-declared serving size of packaged discretionary foods in Australia. Br. J. Nutr. (2016.0) 115 1810-1818. DOI: 10.1017/S0007114516000799 74. Van der Horst K., Bucher T., Duncanson K., Murawski B., Labbe D.. Consumer Understanding, Perception and Interpretation of Serving Size Information on Food Labels: A Scoping Review. Nutrients (2019.0) 11. DOI: 10.3390/nu11092189 75. Hieke S., Palascha A., Jola C., Wills J., Raats M.M.. The pack size effect: Influence on consumer perceptions of portion sizes. Appetite (2016.0) 96 225-238. DOI: 10.1016/j.appet.2015.09.025 76. Schwartz J., Riis J., Elbel B., Ariely D.. Inviting consumers to downsize fast-food portions significantly reduces calorie consumption. Health Aff. (2012.0) 31 399-407. DOI: 10.1377/hlthaff.2011.0224
--- title: Possible Sources of Trace Metals in Obese Females Living in Informal Settlements near Industrial Sites around Gauteng, South Africa authors: - Gladness Nteboheng Lion - Joshua Oluwole Olowoyo journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049368 doi: 10.3390/ijerph20065133 license: CC BY 4.0 --- # Possible Sources of Trace Metals in Obese Females Living in Informal Settlements near Industrial Sites around Gauteng, South Africa ## Abstract Trace metals have been reported in the literature to be associated with obesity. Exposure to some trace metals such as Mn, Cr, Ni, Cd, and Pb may pose a serious health risk to individuals living around a polluted environment. The present study assessed the levels of trace metals in the blood of obese females living around industrial areas in Gauteng, South Africa. The study was carried out using a mixed method approach. Only females with a BMI ≥ 30.0 were considered. A total of 120 obese females participated in the study (site 1: 40—industrial area, site 2: 40—industrial area, and site 3: 40—residential area), aged 18–45 and not in menopause. Blood samples were analysed for trace metals content using inductively coupled plasma mass spectrometry (ICP-MS). The mean concentrations of trace metals were in the order Pb > Mn > Cr > Co > As > Cd (site 1), Pb > Mn > Co > As > Cd (site 2), and Mn > Cr > Co > As > Pb > Cd (site 3). The blood Mn from site 1 ranged from 6.79 µg/L–33.99 µg/L, and the mean differences obtained from the participants from different sites were significant ($p \leq 0.01$). The blood levels of Mn, Pb, Cr, Co, As, and Cd were above the recommended limits set by the WHO in some of the participants. The present study noted, among others, closeness to industrial areas, lifestyle decisions such as the use of tobacco products by their partners indoors, and the method used for cooking as factors that might have accounted for the blood levels of Mn, Pb, Cd and Co. The study showed that there is a need for constant monitoring of the levels of trace metals in the blood of those living in these areas. ## 1. Introduction Comorbidities of human exposure to pollutants such as trace metals have been linked to obesity with worse health consequences and more complex clinical management [1,2]. Individuals living with obesity are more susceptible to the harmful effects of environmental pollutants, including trace metals [3]. Obesity is a rapidly growing problem in the world and in South Africa [4]. In South Africa, a report from the Heart and Stroke Foundation South Africa [5] suggests that the percentage of obese individuals has increased from 2002 to present. A study conducted in 2015 by the University of North West School of Biokinetics, Recreation and Sport Science revealed that nearly two-thirds of the South African population is overweight. The study further revealed that 70 percent of females are overweight. Reports from the literature show that females are most at risk of developing diseases associated with obesity [6,7,8,9]. Trace metals are natural components of the ecosystem but are on the rise due to various developmental programs [10,11]. Natural sources of trace metals include weathering of rock material, soil erosion, volcanic activity, and dissolution of water-soluble salts [12], and they can be added to the environment through anthropogenic activities such as fertilizers, biosolids, irrigation water, coal combustion residues, mining, waste disposal, vehicular emissions, urban runoff, and emissions from industrial activities [13,14,15]. Exposure to trace metals such as Mn, Cr, Ni, Cd, and Pb as a result of industrial activities and vehicle emissions has become a great concern [16]. Trace metals contribute to a variety of diseases, aging, and neurological and behavioural disorders [17,18]. The negative impact of trace metals in human tissue has been reported in the literature, including neurotoxicity, haemolysis, cancer, and cardiomyopathy [19,20,21]. Human exposure to trace metals may occur in three different pathways, which include inhalation, ingestion and dermal exposure. Through the ingestion pathway, trace metals can be transferred into food chains, where they can be absorbed by plants, which are consumed by humans [22]. Street et al. [ 23] noted in South Africa that informal backyard foundries might increase the likelihood of exposure to toxic metals for workers, family members, and communities. The study by Lion and Olowoyo [24] showed that agricultural soils may contain high amounts of trace metals, which may be bioaccumulated by plants. Exposure to trace metals via inhalation or dermal contact has also been documented. In a separate study conducted by Lion and Olowoyo [25], exposure to trace metals such as Cr, Cd, and As were documented. The procedure followed the use of the Target Hazard Quotient, which was first used by Nriagu et al. [ 26]. The study showed that the population around the area under investigation may be exposed to these toxic trace metals through inhalation or dermal exposure. In a study that compared trace metals in the blood of occupationally exposed mine workers and non-occupationally exposed individuals living around a mining area, it was reported that the trace metals Mn, Cd, As, and Pb were higher in the occupationally exposed group [27]. Exposure to trace metals such as Mn, Fe, Al, and Pb has been linked to Alzheimer’s disease [28,29,30]. In another study conducted by Etsuyankpa et al. [ 31] on occupationally exposed workers, the results showed that urine samples had higher concentrations of Mn, Pb, and Ni, while in blood samples, there were high concentrations of Mn, Pb, Ni, and Cr. The concentration of Mn in the blood samples ranged from 14.22 to 22.63 µg/L, and this concentration ranged from 8.30 to 12.20 µg/L in the urine. The concentration of Pb in blood varied between 190.10 and 212.10 µg/L, while the concentration in urine varied from 141.90 to 215.00 µg/L. Additionally, prolonged human exposure to Pb has been reported to decrease performance in the nervous system; cause weakness in the wrists, fingers and ankles; and increase blood pressure and anemia [32]. In South Africa, Nkosi et al. [ 33] reported that those living around mining areas have high levels of blood pressure. The study of Zunpuski et al. [ 34] further showed that children and adults living around gold mine tailings may have detectable amounts of uranium with higher concentrations in children’s hair when compared to adults. Mining activities have been reported to have adverse effects on the environment and human health [35,36]. Kamunda et al. [ 37] showed that the concentrations of Cr and Ni in soils collected around a gold mine basin in the Witwatersrand were higher than the permissible limit and further indicated a carcinogenic health risk for humans living around the area. A study by Olowoyo and Lion [38] had previously reported that people living in areas exposed to environmental pollution, such as those around mining and coal-fired power stations, may suffer from respiratory diseases because of pollutants around these areas. A similar study conducted by Okereafor et al. [ 39] also showed that trace metals such as Mn, Cr, Pb, Ni, Fe, and Cd were major pollutants from mining areas. The challenge and great dilemma are that mining contributes a great portion to the economy of South Africa. Mining alone contributes an income of USD 588 billion and is responsible for the employment of more than 451,427 people [35,40]. However, some of these mines are open cast, and it is common to see heaps of uncovered excavated waste material from mining sites in close proximity to informal settlements [38]. The pollutants from underground end up on the surface of the ground and are carried by the wind as dust to the residents living around the mining site in the informal settlements. Apart from mining, activities such as coal-fired power stations have also been reported as contributors of trace metals in the environment. Coal-fired power stations may expose the surrounding communities to pollutants that may have a negative impact on their health [38]. The impact of human exposure to coal combustion varies according to the composition of the coal. The study by Olowoyo et al. [ 26] conducted around a coal-powered station in Pretoria, South Africa, showed that plants and soil collected around this area had elevated levels of trace metals both in the soil and plants. Previous studies noted that the burning of coal from Brazil releases high levels of Zn and As [41,42], while coal-fired power stations in South Africa release high levels of Pb in the environment [43]. Over the years, South Africa has been under great pressure to sustain the increasing need for energy in the country. The country’s electric power supplier relies on the combustion of coal for the generation of electricity to fuel the country [44]. Coal is a fossil fuel and non-renewable energy source that is combusted and used to generate electricity [45]. A coal-fired power plant is among the great generators of environmental pollution, releasing large quantities of pollutants which include aerosol. As the status of load shedding and incessant power failures experienced in South Africa continues, the constant burning of coal may also increase the level of pollution in the environment. Eskom has also resorted to the burning of diesel, a form of fuel that has been reported to increase high levels of trace metals such as Fe, Cu, Cr, Ti and Mn in the environment [46]. Recent reports showed an association between trace metals and health complications among obese individuals [47,48,49]. The impact of exposure to the trace metals As, Pb, Cr, and Zn on obesity development and cardiovascular disease in children and adolescents was investigated by Nasab et al. [ 50]. Their findings showed that exposure to these metals might be a risk factor for children and adolescents to develop cardiovascular disease and obesity markers. Furthermore, they also demonstrated that these metals were associated with individual weight status. They concluded that there was a significant association between trace metals (As, Pb, Cr, and Zn) and body mass index (BMI), waist circumference (WC), fasting blood sugar (FBS), and lipid profile, among other things. These metals have also been reported to have the ability to raise leptin levels in the body, which is a protein produced by adipose tissue. Studies relating to trace metals, female obesity, and its comorbidities have not been carried out in detail in populations exposed to mining, coal-fired power stations, and other industrial activities in South Africa. Human exposure to pollutants such as Mn, Cr, Pb, and As due to living around a mining site and coal-fired power station may have negative health implications [48,51]. However, when coupled with obesity, the consequences maybe be fatal to humans [50,52]. The current study investigated the levels of trace metals in obese female individuals living in areas impacted by different areas close to industries. Though most of these areas are informal settlements, there are people living in these areas for different reasons. As stated above, mining and coal-fired power stations have the ability to introduce different types of trace metals into the environment; however, a comparison of the impact of these activities on humans and with particular reference to obese individuals living in areas associated with these types of industries has not been documented, and the effect is not fully known. This current study investigated the impact of these activities on the concentrations of trace metals in the blood of these individuals. To the best of our knowledge, this is the first study to compare and provide information on the impacts of different industries in introducing trace metals in the environment and may therefore serve as baseline data for future research. Hence, the aim of this study was to determine the concentration of trace metals in the blood of obese females and compare the impact of living around industrial sites and an area that was not industrial but instead residential. ## 2.1. Study Design and Setting The study was carried out using a mixed methods approach. The study was carried out around two industrial areas in Pretoria, and participants were obese females living in an informal settlement in these areas. A purposive sampling method was used for the study. As per observation from the first visit to the industrial areas, site 1 had 82 housing units, and site 2 had only 65 housing units. It was then assumed that there was an adult female member living in each house. According to the report provided by [53], $70\%$ of women in SA are obese; hence it was envisaged that there were about 57 obese females from industrial site 1 and 46 obese individuals from industrial site 2. However, only 120 obese females participated in the study who were aged between ages 18–45 and had not reached menopause. The participants sampled were 40 obese females from site 1, 40 obese females living in site 2, and 40 obese females living far away from these two sites in a residential area. ## 2.2. Sampling and Study Population Ethical clearance was obtained prior to the commencement of this study. Approval was obtained from Sefako Makgatho University Research Ethics Committee (SMUREC), with reference number SMUREC/S/$\frac{39}{2018}$:PG. A written informed consent was issued to each willing participant. Owing to the sensitive nature of the study, the identity of human subjects was protected. No unique identifiers were collected, and the study was conducted with only self-selected participants. The study was part of a Ph.D. program of the leading author and, as such, a cohort study. The authors had initially examined the relationship between the trace metal levels and reproductive hormones from some participants in this study [54]. The study only included obese individuals living around the industrial sites and a group from a residential area from a different site not impacted by industrial activities. Females with a higher BMI (≥30.0) were used as obese participants in the study, and those with a lesser BMI (≤29.9) were excluded. Individuals that were taking mineral supplements and those diagnosed with diabetes were also excluded from the study. Diabetic individuals within these groups were excluded from the study due to the fact that their condition may eliminate trace metals through the urine, and therefore, the results may be compromised [51]. The blood samples from all obese female participants were drawn by puncturing a vein. The procedure was performed by a qualified professional under controlled conditions to minimize contaminations. Only 50 mL of blood samples were collected using the certified BD Vacutainer sterile tubes for trace metals. An anticoagulant of 143 USP units of sodium heparin was added to each tube and enclosed with a royal blue hemoguard cap. Samples were then stored at cool temperatures before being analyzed. ## 2.3. Trace Metal Determination The analysis of blood samples for trace metals determination was carried out at accredited laboratories. An inductively coupled plasma mass spectrometer (ICP–MS) Thermo (Bremen, Germany) X-Series 2, with a concentric glass nebulizer and Peltier-cooled glass spray chamber, were used to analyse trace metals from the blood samples collected from all sites at an accredited lab. From the samples, 100µL of blood sample was diluted 45-fold with milli-Q water (>18 megohm/cm resistivity) containing $1\%$ nitric acid (Optima grade, Fisher Scientific, Shanghai, China) and $0.01\%$ TritonX-100 and allowed to digest overnight at room temperature. Then, 500µL hydrogen peroxide ($30\%$ Suprapur grade, Sigma-Aldrich, St. Louis, MI, USA) was added to each digest and allowed to sit for at least one hour prior to analysis. Then, the digested solution was diluted with distilled water and brought up to 50 mL volume. The blanks were prepared by adding reagents to deionized water in place of the samples so as to monitor the background concentration of all the analytes. ## 2.4. Data Analysis SPSS 26.0 for windows was used to carry out the statistical analysis. Analysis of variance was used to test for significant differences in the concentrations of trace metals in the blood of obese female participants from all the sites. ## 2.5. Quality Control and Assurance For quality control, all the materials, including the pipettes, glassware, and stoppers, were washed with $10\%$ nitric acid and rinsed with distilled water. The certification for ICP-MS was obtained by calibrating the machine with certified reference material before introducing the blanks and the samples. For quality assurance, samples were analyzed in triplicates, and a single wash was performed after a complete analysis of each sample. ## 3. Results and Discussion The participants in this study comprised only of females. The participants from sites 1 and 2 were all living in an informal settlement just around the industrial areas. Most of the participants were staying with their partners, who were casual workers in some of these industries. The age range was 18–45. From the study, $33\%$ of the participants had been staying in site 1 for a period between 1 and ≥ 10 years, 33 % had been staying in site 2 for 1– ≥ 10 years, and the remaining group was the participants from a residential area. Within the residential area group, the information received later showed that some participants had resided in areas not far from industrial activities at site 1 for more than 10 years. Most of the participants mentioned that their partners were active smokers, and none of them were smokers. The mode of cooking at home involved the use of firewood and the burning of charcoals in most cases. These two activities, as mentioned by some of the participants, may increase the concentrations of trace metals such as Pb, Cd, and As in their blood. The study by Jung et al. [ 55] showed that passive smoking or exposure to secondhand smoke may increase blood Cd. El Mohr et al. [ 56] also reported that passive smoking, as noted in our study, may increase the levels not just of Cd but also Pb and As in the serum of passive smokers. The study by Olufunsho et al. [ 57] also noted that exposure to gaseous pollutants and dust, which may come in the form of ashes, may increase the levels of toxic trace metals in the blood. The above-mentioned phenomena were noticed and practiced on a daily basis by some of the participants who presented high levels of toxic trace metals in our study. The results of trace metal analysis from blood samples collected from obese females from all the participants are represented in Table 1. From the results obtained from some participants, the mean concentrations were observed to be higher than the recommended limit by the WHO for human exposure to trace metals such as Cr, Co, As, and Pb. Furthermore, Cd concentrations were above recommended values by the WHO in some individuals that participated in the study, even though the mean concentration of Cd was lower than the recommended limit. Overall, the recorded mean concentrations of all trace metals in the study were in the order Mn > Pb > Cr > Co > As > Cd (site 1), Pb > Mn > Co > As > Cd (site 2), and Mn > Cr > Co > As > Pb > Cd (site 3). The highest mean concentration of Mn (12.97 ± 6.04 µg/L) from obese females in site 1 was significantly higher than other sites ($p \leq 0.01$). The second highest concentrations of Mn in obese female participants were recorded from site 2, with a mean and standard deviation of 11.41 µg/L and 4.35 and a range of 5.50–24.40 µg/L. The values recorded for Mn in the blood of some participants from all the sites were above the WHO recommended limit of 12.60 µg/L. There were significant differences in the concentrations of Mn from participants residing in site 1, site 2, and site 3 ($p \leq 0.01$). The lowest mean Mn concentrations in the participant’s blood were recorded from site 3. Although some participants live around site 3 for employment and study purposes, they frequently visit their homes which are either in sites 1 and 2, and these areas are associated with mining activities and other industrial activities. Due to the location of site 3 and nature of work of some participants in site 3, some are usually exposed to vehicular emissions on a daily basis, and the findings from this study should assist in having a basis for further research into the source of the Mn that was high in the blood of the participants in this area. The length of stay of obese females living around site 3 shows that individuals who lived in the area for more than 10 years recorded lower levels of blood Mn than those who have stayed in the area for less than 10 years (Figure 1). This suggests that those individuals may have been exposed to high levels of Mn in their previous residential areas [58]. The values of Mn recorded from site 1, an industrial area with active activities taking place from the industry situated opposite the informal settlement, are similar to those reported by Dey et al. [ 59], where the blood Mn from samples collected around a mining site exceeded the recommended limit. Their findings showed that Mn is considered a systemic toxicant that can damage multiple organs of humans. High levels of Mn may have adverse effect on the neurobehavioral attitude, as reported in the study by Pesch et al. [ 60]. The reports from other studies, such as Aguera et al. [ 61], showed mean values of 2.7 µg/L in females and 1.0 µg/L in males that are working as farmers and exposed to pesticides. The mean concentration of Cr from site 3 was 6.16 ± 6.30 µg/l and in the range of 1.58–29.29 µg/L. From participants in site 1, Cr concentrations recorded a mean and standard deviation of 3.70 µg/L and 4.11 µg/L, respectively. The range for blood Cr was between 0.41 and 19.01 µg/L for the participants in the study from this site. Cr in blood from participants from site 2 was below the detection limit. The levels of Cr in the blood of participants may be as a result of emissions from industrial activities. Levels of blood Cr recorded for site 3 could be as previously indicated due to the participants’ previous area of occupation. Site 3 is characterized by an influx of individuals who previously lived in other areas, including places around industrial areas. The blood Cr could be historical traces of their previous and current living area since trace metals do not biodegrade but rather bio-accumulate (Figure 2). In the absence of any known exposure to induce the levels of blood Cr recorded from site 3, this study is therefore suggesting a source of significant Cr exposure in this site, which will require further research involving more participants. Increased levels of Cr may also be due to exposure to tobacco products, in this case through passive smoking [26,56]. Furthermore, the findings of this study were similar to that of Etsuyankpa et al. [ 30], where the concentration of Cr, Mn, Pb, and Cd in the blood exceeded the set recommended limit in both the occupationally exposed groups and non-exposed groups (control). High Cr levels have been associated with microcytic anaemia and mitochondrial and DNA damage in blood cells, which in turn induces carcinogenicity, occupational asthma, airway hypersensitivity, and nose, eye and skin irritation [62]. The concentration of Co from participants from site 3 was higher than the other two sites, with a mean and standard deviation of 2.40 ± 8.92 µg/L and with values ranging from 0.09 µg/L to 50.88 µg/L. From site 1, Co concentrations showed a range of 0.20–18.84 µg/L and a mean of 2.19 ± 4.43 µg/L. Although the mean Co concentration from site 2 was 1.08 ± 1.34 µg/L, it was recorded as the lowest when compared to sites 1 and 3. However, the mean concentrations for all the sites were higher than the recommended limit for human exposure as set by the WHO. No significant difference was observed in the blood concentrations of Co in participants from the sites in this study ($p \leq 0.05$). The consumption of beer may be a source of high levels of Co recorded from the blood of participants as they were observed to be drinking during sampling from both sites 1 and 2. Although participants from site 3 were not observed to be drinking beer during the sampling visits, they do consume other alcoholic beverages. Studies also showed that individuals who consume high volumes of alcohol, especially beer with added cobalt chloride (CoCl2) or cobalt sulphate (CoSO4), a foam stabilizer, may suffer from cardiomyopathy [60,61,62]. Reports show that Mn and Co concentrations in beer samples ranged from 25.29 to 228.60 μg/L and 0.16 to 0.56 μg/L, respectively [63,64]. Therefore, the consumption of alcoholic beverages, which include beer, may also contribute to the high levels of Co observed in all the sites in the study. Furthermore, the length of stay of obese females living around site 3 shows that individuals who lived in the area for more than 10 years recorded the lowest levels of blood Co compared to those who have stayed in the area for less than 10 years (Figure 3), therefore suggesting exposure from previous residential areas. Cameán et al. [ 65] have also reported that Co and Mn are critical metals that can be found in high concentrations in coal-derived sources, such as coal refuse, coal combustion products, and coal acid mine drainage. Eren et al. [ 66] reported that Co has a concentration of 25 ppm in the Earth’s crust, which in addition to coal burning through mining and industrial activities, has resulted in the high blood Co concentrations observed from all the sites in this study. A report by Talan et al. [ 67] shows that high levels of Co has been associated with autism in children. The blood As concentrations from participants residing in site 3 were the highest, with minimum and maximum values of 0.70 µg/L and 3.43 µg/L, respectively, and a mean of 1.37 ± 0.67 µg/L. From site 1, the mean concentrations of As in the blood were 0.84 ± 0.41 µg/L and ranged from 0.27 µg/L to 2.20 µg/L (Table 1). From the results, it was observed that participants from sites 1 and 2 recorded lower mean concentrations. However, the concentrations of As from some participants from both sites were higher than the WHO recommended limit of 1.00 µg/L. There was no significant difference observed in the concentrations of blood As in participants from the sites in this study ($p \leq 0.05$). The high blood As recorded from site 3 could be historical traces of their previous and current living area (Figure 4). Furthermore, the burning of coal has been associated with high levels As [42,43], and other sources of As in the environment are anthropogenic in nature, as well as due to lifestyle habits of active or secondary smoking [26,56,68]. Lion and Olowoyo [24] have also reported that humans living around mining areas can be exposed to high levels of As, Mn, Cr, and Cd through inhalation. Arsenic is carcinogenic and can cause cancer in different parts of the body and chronic diseases, such as diabetes and hypertension [32,69]. From site 2, the mean blood Pb concentration was higher than those recorded for participants from sites 1 and 3 with a mean, standard deviation and range of 19.41 ± 11.84 µg/L (6.70–60.20 µg/L). The concentration of Pb from site 1 was also higher than the recommended limit of 0.80 µg/L, with a mean of 6.70 ± 3.40 µg/L and a range of 5.00–18.40 µg/L. Not all the participants had blood Pb higher than the recommended limit. Although the mean blood Pb concentrations from some participants from site 3 (0.51 ± 0.05 range 0.50–0.78 µg/L) were lower than the recommended limit, but with a caution as trace metals could bioaccumulate. There was a significant difference observed in the concentrations of blood Pb in participants from sites 1, site 2, and site 3 in this study ($p \leq 0.01$). Figure 5 shows that the levels of blood Pb recorded from all the participants from different sites in the study may be due to exposure to lead and not necessarily associated with the areas used for this study. The ban on unleaded petrol only came into effect in South Africa in 2005, and previous studies have shown the ability of Pb to remain in the environment for a prolonged period of time due to its non-biodegradable nature. Some studies have also reported high levels of Pb [70,71]. The report showed that once Pb is absorbed in the blood, it has the ability to cause anemia by decreasing the number of red blood cells in the body. The results obtained from participants residing in site 2 were comparable with that of Shekar et al. [ 72], where the highest concentrations of blood Pb from the exposed group were 45.43 ± 6.93 μg/L, which exceeds the recommended limit of 08.0 μg/L for human blood. Blood Cd from this study recorded the lowest value of all the trace metals in the participants’ blood. The mean concentration of Cd in the blood was lower than the recommended limit of 1.12 µg/L from all the participants. However, higher than the recommended limit of Cd in blood were recorded from some females residing in site 3. The length of stay of obese females living around site 3 shows that individuals who lived in the area for more than 10 years recorded lower levels of blood Cd than those who had stayed in the area for less than 10 years (Figure 6). The length of stay shows that participants may have been exposed to high levels of Cd from other sources. The recorded concentrations for Cd were 0.38 ± 0.55 µg/L (0.01–2.28 µg/L) from site 3. There was no significant difference observed in the concentrations of blood Cd in participants from the sites in this study ($p \leq 0.05$). Though lower concentrations of As, Pb, and Cd (Table 1) were recorded in the blood of some participants, there is a need for close monitoring to reduce the effects of exposure on human health. High levels of Cd in the blood may affect the respiratory, renal, and skeletal systems. These high levels can lead to renal and tubular dysfunction and an increased risk of cardiovascular disease, peripheral arterial disease, heart failure, myocardial dysfunction, and stroke [73,74]. This study is similar to that conducted by [20], where they reported an increase in the concentrations of Pb, Mn, and Cd in the blood of participants exposed to trace metal pollutants. Furthermore, the study by Devoy et al. [ 75] also reported higher concentrations of mean blood Cd and Pb exceeding the permissible limits of 1.12 μg/L and 0.80 μg/L, set by WHO, respectively. Their findings further show that high levels of these metals can be attributed to several factors, such as smoking, place of residence, and intake of some food. The participants in our study reported frequent headaches and dizziness as some of the major health problems encountered. This phenomenon is similar to those reported in the study [59]. Despite the fact that there are limited studies on occupational exposures to Mn, the study of [76], a study of haemolysis of particulate matter (PM) on red blood cells from a coal-burning lung cancer epidemic area, the report showed that Mn had a significant positive correlation with the oxidative capacity of PM in coal-burning environments, and this is a usual practice from some of the participants in our study. Furthermore, the report by [77] shows that Mn concentrations in blood cells increase as the dose of external exposure increases. Generally, the high levels of Mn in South African soils have previously been reported in the literature [78]. We do not know clearly if some of the participants in our study practice geophagia; however, this is a common practice in South Africa, and this may be responsible for the increase in the value of Mn recorded in the blood of some of the participants [79]. ## 4. Conclusions To the best of our knowledge, this is the first study to explore exposure to trace metals from obese females living in informal settlements around industrial areas. This study can assist in forming a baseline for future research by monitoring and assessing human health risks in these areas. From this study, it was evident that mean Mn levels in blood from participants in site 1 were higher than the other two sites, while the mean blood Pb in site 2 was also higher than in other areas. However, it is important to note that some of the participants from site 3 (Industrial area) also showed high levels of blood Cr, As, Co, and Cd. The results of participants from site 3 also showed that these participants resided in some of the areas that were used in this study, which included industrial activities such as mining areas and coal-fired power stations. The results of the study showed that industrial activities may not be the only route for human exposure to trace metals; as noted in this study, lifestyle may also play a major role in the blood levels of these toxic trace metals. Levels of some trace metals recorded from some of the participants in this study may require further intervention due to reports received from the participants and from previous studies. An increase in the levels of these trace metals in the blood may pose a health risk to obese female participants. Future studies should be geared towards examining the sources of blood trace metals recorded from participants in this study, especially from site 3, in this study. ## References 1. Rivas-Arancibia S., Balderas-Miranda J., Belmont-Zúñiga L., Martínez-Jáquez M., Hernández-Orozco E., Cornejo-Trejo V., Reséndiz-Ramos C., Cruz-García I., Espinosa-Caleti I., Valdés-Fuentes M.. Oxidative stress, antioxidant defenses, COVID-19 and pollution. Med. Res. Arch. (2020) 8. DOI: 10.18103/mra.v8i10.2254 2. Li T., Yu L., Yang Z., Shen P., Lin H., Shui L., Tang M., Jin M., Chen K., Wang J.. Associations of Diet Quality and Heavy Metals with Obesity in Adults: A Cross-Sectional Study from National Health and Nutrition Examination Survey (NHANES). Nutrients (2022) 14. DOI: 10.3390/nu14194038 3. Limaye S., Salvi S.. Obesity and asthma: The role of environmental pollutants. Immunol. Allergy Clin. (2014) 34 839-855 4. 4. World Health Organization Human Biomonitoring: Facts and FiguresWHO Regional Office for EuropeCopenhagen, Denmark2015. Human Biomonitoring: Facts and Figures (2015) 5. 5. Heart and Stroke Foundation South Africa Cardiovascular Disease Statistics Reference DocumentHSFSACape Town, South Africa2016Volume 3. Cardiovascular Disease Statistics Reference Document (2016) Volume 3 6. Joubert J., Norman R., Bradshaw D., Goedecke J.H., Steyn N.P., Puoane T.. Estimating the burden of disease attributable to excess body weight in South Africa in 2000. S. Afr. Med. J. (2007) 97 683-690. PMID: 17952225 7. Steyn K., Damasceno A.. Lifestyle and related risk factors for chronic diseases. Dis. Mortal. Sub-Sahar. Afr. (2006) 2 247-265 8. Zimmet P., Alberti K.G.M.M., Shaw J.. Global and societal implications of the diabetes epidemic. Nature (2001) 414 782-787. DOI: 10.1038/414782a 9. Sampaio M.S., Reddy P.N., Kuo H.T., Poommipanit N., Cho Y.W., Shah T., Bunnapradist S.. Obesity was associated with inferior outcomes in simultaneous pancreas kidney transplant. Transplantation (2010) 89 1117-1125. DOI: 10.1097/TP.0b013e3181d2bfb2 10. Turan Ö., Özdemir H., Demir G.. Deposition of heavy metals on coniferous tree leaves and soils near heavy urban traffic. Front. Life Sci. Relat. Technol. (2020) 1 35-41 11. TürksoY R., Terzioğlu G., Yalçin İ.E., Türksoy Ö., Demir G.. Removal of heavy metals from textile industry wastewater. Front. Life Sci. Relat. Technol. (2021) 2 44-50. DOI: 10.51753/flsrt.958165 12. Imseng M., Wiggenhauser M., Müller M., Keller A., Frossard E., Wilcke W., Bigalke M.. The fate of Zn in agricultural soils: A stable isotope approach to anthropogenic impact, soil formation, and soil−plant cycling. Environ. Sci. Technol. (2019) 53 4140. DOI: 10.1021/acs.est.8b03675 13. Shi T.R., Ma J., Wu F.Y., Ju T.N., Gong Y.W., Zhang Y.Y., Wu X., Hou H., Zhao L., Shi H.D.. Mass balance-based inventory of heavy metals inputs to and outputs from agricultural soils in Zhejiang Province, China Sci. Total Environ. (2019) 649 1269-1280. DOI: 10.1016/j.scitotenv.2018.08.414 14. Wang Z., Xiao J., Wang L., Liang T., Guo Q., Guan Y., Rinklebe J.. Elucidating the differentiation of soil heavy metals under different land uses with geographically weighted regression and self-organizing map. Environ. Pollut. (2020) 260 114065. DOI: 10.1016/j.envpol.2020.114065 15. Cai P., Cai G., Yang J., Li X., Lin J., Li S., Zhao L.. Distribution, risk assessment, and quantitative source apportionment of heavy metals in surface sediments from the shelf of the northern South China Sea. Mar. Pollut. Bull. (2023) 187 114589. DOI: 10.1016/j.marpolbul.2023.114589 16. George A., Shen B., Kang D., Yang J., Luo J.. Emission control strategies of hazardous trace elements from coal-fired power plants in China. J. Environ. Sci. (2020) 93 66-90. DOI: 10.1016/j.jes.2020.02.025 17. Hassan M.A., Xu T., Tian Y., Zhong Y., Ali F.A.Z., Yang X., Lu B.. Health benefits and phenolic compounds of Moringa oleifera leaves: A comprehensive review. Phytomedicine (2019) 93 153771. DOI: 10.1016/j.phymed.2021.153771 18. Mohammed R.S., Ibrahim W., Sabry D., El-Jaafary S.I.. Occupational metals exposure and cognitive performance among foundry workers using tau protein as a biomarker. Neurotoxicology (2020) 76 10-16. DOI: 10.1016/j.neuro.2019.09.017 19. Gill M.. Heavy metal stress in plants: A review. Int. J. Adv. Res. (2014) 2 1043-1055 20. Bomberg M., Ahonen L.. Geomicrobes: Life in terrestrial deep subsurface. Front. Microbiol. (2017) 8 103. DOI: 10.3389/fmicb.2017.00103 21. Muhammad Shehata I., Masood W., Nemr N., Anderson A., Bhusal K., Edinoff A.N., Cornett E.M., Kaye A.M., Kaye A.D.. The possible application of ketamine in the treatment of depression in Alzheimer’s disease. Neurol. Int. (2022) 14 310-321. DOI: 10.3390/neurolint14020025 22. Vincēviča-Gaile Z.. Impact of Environmental Conditions on Micro-and Macroelement Content in Selected Food from Latvia. Ph.D. Thesis (2014) 1-182 23. Street R.A., Goessler W., Naidoo S., Shezi B., Cele N., Rieger J., Ettinger K., Reddy T., Mathee A.. Exposure to lead and other toxic metals from informal foundries producing cookware from scrap metal. Environ. Res. (2020) 191 109860. DOI: 10.1016/j.envres.2020.109860 24. Lion G.N., Olowoyo J.O.. Population health risk due to dietary intake of toxic heavy metals from Spinacia oleracea harvested from soils collected in and around Tshwane, South Africa. S. Afr. J. Bot. (2013) 88 178-182. DOI: 10.1016/j.sajb.2013.07.014 25. Lion G.N., Olowoyo J.O.. Health risk assessment of dust samples (indoor and outdoor) collected from houses around a mining area in Brits, South Africa. Curr. Top. Toxicol. (2021) 17 99-111 26. Nriagu J.O., Bhattacharya P., Mukherjee A.B., Bundschuh J., Zevenhoven R., Loeppert R.H.. Arsenic in soil and groundwater: An overview. Trace Met. Other Contam. Environ. (2007) 9 3-60 27. Olowoyo J.O., Magoloi Z.G., Lion G.N., Ogunbanjo G.. Toxic trace metals in blood of occupationally exposed casual mine workers living in shacks around mining area in Brits, South Africa. Trace Elem. Electrolytes (2019) 37 1-11. DOI: 10.5414/TEX01587 28. Tchounwou P.B., Yedjou C.G., Patlolla A.K., Sutton D.J.. Heavy metal toxicity and the environment. Molecular, clinical and environmental toxicology. Environ. Toxicol. (2012) 3 133-164 29. Nallagouni C., Reddy K.P.. Aluminum and fluoride impacts cortex and hippocampus structure in rats: Protective role of resveratrol. Int. J. Appl. Biol. Pharm. Technol. (2017) 8 89-97 30. Tajahmadi S., Shamloo A., Shojaei A., Sharifzadeh M.. Adsorption Behavior of a Gd-Based Metal–Organic Framework toward the Quercetin Drug: Effect of the Activation Condition. ACS Omega (2022) 7 41177-41188. DOI: 10.1021/acsomega.2c04800 31. Etsuyankpa M.B., Ahmad A.S., Mustapha S., Baba N.M., Ndamitso M.M., Elabor R.. Levels of selected heavy metals in blood and urine of workers of Alkali Kongo Plaza GSM Market within Lafia Metropolis, Nasarawa State, Nigeria. Environ. Chall. (2022) 9 100649. DOI: 10.1016/j.envc.2022.100649 32. Hu Y., Zhou J., Du B., Liu H., Zhang W., Liang J., Zhang W., You L., Zhou J.. Health risks to local residents from the exposure of heavy metals around the largest copper smelter in China. Ecotoxicol. Environ. Saf. (2022) 171 329-336. DOI: 10.1016/j.ecoenv.2018.12.073 33. Nkosi N.Z., Manzi M.S., Westgate M., Roberts D., Durrheim R.J., Ogasawara H., Ziegler M., Rickenbacher M., Liebenberg B., Onstott T.C.. Physical property studies to elucidate the source of seismic reflectivity within the ICDP DSeis seismogenic zone: Klerksdorp goldfield, South Africa. Int. J. Rock Mech. Min. Sci. (2022) 155 105082. DOI: 10.1016/j.ijrmms.2022.105082 34. Zupunski L., Street R., Ostroumova E., Winde F., Sachs S., Geipel G., Nkosi V., Bouaoun L., Haman T., Schüz J.. Environmental exposure to uranium in a population living in close proximity to gold mine tailings in South Africa. J. Trace Elem. Med. Biol. (2023) 77 127141. DOI: 10.1016/j.jtemb.2023.127141 35. Martin S., Griswold W.. Human health effects of heavy metals. Environ. Sci. Technol. Briefs Citiz. (2009) 15 1-6 36. Brunetto G., Miotto A., Ceretta C.A., Schmitt D.E., Heinzen J., de Moraes M.P., Canton L., Tiecher T.L., Comin J.J., Girotto E.. Mobility of copper and zinc fractions in fungicide-amended vineyard sandy soils. Arch. Agron. Soil Sci. (2014) 60 609-624. DOI: 10.1080/03650340.2013.826348 37. Kamunda C., Mathuthu M., Madhuku M.. Potential human risk of dissolved heavy metals in gold mine waters of Gauteng Province, South Africa. J. Toxicol. Environ. Health Sci. (2018) 10 56-63 38. Olowoyo J.O., Lion G.N.. Urban farming as a possible source of trace metals in human diets. S. Afr. J. Sci. (2016) 112 30-35. DOI: 10.17159/sajs.2016/20140444 39. Okereafor G., Makhatha M., Mekuto L., Mavumengwana V.. Evaluation of trace elemental levels as pollution indicators in an abandoned gold mine dump in Ekurhuleni Area, South Africa. Trace Metals in the Environment—New Approaches and Recent Advances (2019) 40. Molloy I., Chen H., Li T., Wang Q., Li N., Bertino E., Calo S., Lobo J.. Mining roles with multiple objectives. ACM Trans. Inf. Syst. Secur. (2010) 13 1-35. DOI: 10.1145/1880022.1880030 41. Africa Mining, IQ. Africa’s Mining Portal: Online Mining Information Service. (2022) 42. Hendryx M., Zullig K.J., Luo J.. Impacts of coal use on health. Annu. Rev. Public Health (2020) 41 397-415. DOI: 10.1146/annurev-publhealth-040119-094104 43. Ortega-Flores P.A., Serviere-Zaragoza E., De Anda-Montañez J.A., Freile-Pelegrín Y., Robledo D., Méndez-Rodríguez L.C.. Trace elements in pelagic Sargassum species in the Mexican Caribbean: Identification of key variables affecting arsenic accumulation in. Sci. Total Environ. (2022) 806 150657. DOI: 10.1016/j.scitotenv.2021.150657 44. Zhao L., Dai S., Finkelman R.B., French D., Graham I.T., Yang Y., Li J., Yang P.. Leaching behavior of trace elements from fly ashes of five Chinese coal power plants. Int. J. Coal Geol. (2020) 219 103381. DOI: 10.1016/j.coal.2019.103381 45. Flues M., Sato I.M., Scapin M.A., Cotrim M.E.B., Camargo I.M.C.. Toxic elements mobility in coal and ashes of Figueira coal power plant, Brazil. Fuel (2013) 103 430-436. DOI: 10.1016/j.fuel.2012.09.045 46. Silva L.F., Ward C.R., Hower J.C., Izquierdo M., Waanders F., Oliveira M.L., Li Z., Hatch R.S., Querol X.. Mineralogy and leaching characteristics of coal ash from a major Brazilian power plant. Coal Combust. Gasif. Prod. (2010) 2 51-65. DOI: 10.4177/CCGP-D-10-00005.1 47. Okedeyi O.O., Dube S., Awofolu O.R., Nindi M.M.. Assessing the enrichment of heavy metals in surface soil and plant (. Environ. Sci. Pollut. Res. (2014) 21 4686-4696. DOI: 10.1007/s11356-013-2432-0 48. Vorster T., Mthombeni J., TeWaterNaude J., Phillips J.I.. The Association between the Histological Subtypes of Mesothelioma and Asbestos Exposure Characteristics. Int. J. Environ. Res. Public Health (2022) 19. DOI: 10.3390/ijerph192114520 49. Gasparotto J., Martinello K.D.B.. Coal as an energy source and its impacts on human health. Energy Geosci. (2021) 2 113-120. DOI: 10.1016/j.engeos.2020.07.003 50. Nasab H., Rajabi S., Eghbalian M., Malakootian M., Hashemi M., Mahmoudi-Moghaddam H.. Association of As, Pb, Cr, and Zn urinary heavy metals levels with predictive indicators of cardiovascular disease and obesity in children and adolescents. Chemosphere (2022) 294 133664. DOI: 10.1016/j.chemosphere.2022.133664 51. Fan J.G., Kim S.U., Wong V.W.S.. New trends on obesity and NAFLD in Asia. J. Hepatol. (2017) 67 862-873. DOI: 10.1016/j.jhep.2017.06.003 52. Griffin W.M., Michalek J., Matthews H.S., Hassan M.N.A.. Availability of biomass residues for co-firing in Peninsular Malaysia: Implications for cost and GHG emissions in the electricity sector. Energies (2014) 7 804-823. DOI: 10.3390/en7020804 53. De Ridder J.H., Coetzee D.. Childhood obesity in South Africa: Are we sitting on a time bomb?. Glob. J. Health Phys. Educ. Pedagog. (2013) 2 239-249 54. Lion G.N., Ogunbanjo G.A., Olowoyo J.O.. Concentration of Trace Metals in Blood and the Relationship with Reproductive Hormones (Estradiol and Progesterone) of Obese Females Living Around a Mining Area in Brits, South Africa. Nat. Environ. Pollut. Technol. (2020) 19 1163-1171. DOI: 10.46488/NEPT.2020.v19i03.029 55. Jung K.W., Lee S.Y., Lee Y.J.. Facile one-pot hydrothermal synthesis of cubic spinel-type manganese ferrite/biochar composites for environmental remediation of heavy metals from aqueous solutions. Bioresour. Technol. (2018) 261 1-9. DOI: 10.1016/j.biortech.2018.04.003 56. ElMohr M., Faris M., Bakry S., Hozyen H., Elshaer F.M.. Effect of passive smoking on heavy metals concentration in blood and follicular fluid of patients on going ICSI. Indian J. Sci. Technol. (2020) 13 2035-2040. DOI: 10.17485/IJST/v13i29.721 57. Olufunsho O.O., Tope-Ajayi A.A., Alabi G.O.. Effect of wood smoke on selected blood antioxidant parameters, essential and heavy metal levels in women. Eur. Sci. J. (2015) 11 238-244 58. Raimi I.O., Komolafe B.F., Agboola O.O., Mugivhisa L.L., Olowoyo J.O.. Influence of wind direction on the level of trace metals in plants collected around a Quarry Site in South Africa. Pol. J. Environ. Stud. (2019) 28 3385-3393. DOI: 10.15244/pjoes/94846 59. Dey S., Tripathy B., Kumar M.S., Das A.P.. Ecotoxicological consequences of manganese mining pollutants and their biological remediation. Environ. Chem. Ecotoxicol. (2023) 5 55-61. DOI: 10.1016/j.enceco.2023.01.001 60. Pesch B., Weiss T., Kendzia B., Henry J., Lehnert M., Lotz A., Heinze E., Käfferlein H.U., Van Gelder R., Berges M.. Levels and predictors of airborne and internal exposure to manganese and iron among welders. J. Expo Sci. Environ. Epidemiol. (2012) 22 291-298. DOI: 10.1038/jes.2012.9 61. Aguera R.G., da Silva Freires C., de Oliveira L.O., Monteiro L.R., Lini R.S., Romoli J.C.Z., Freire B.M., Nerilo S.B., Junior M.M., Batista B.L.. Risk evaluation of occupational exposure of southern Brazilian flower farmers to pesticides potentially leading to cholinesterase inhibition and metals exposure. Environ. Toxicol. Pharmacol. (2022) 93 103874. DOI: 10.1016/j.etap.2022.103874 62. Alexander C.S.. Cobalt-beer cardiomyopathy: A clinical and pathologic study of twenty-eight cases. Am. J. Med. (1972) 53 395-417. DOI: 10.1016/0002-9343(72)90136-2 63. Mannino D.M., Homa D.M., Matte T., Hernandez-Avila M.. Active and passive smoking and blood lead levels in U.S. adults: Data from the Third National Health and Nutrition Examination Survey. Nicotine Tob. Res. (2005) 7 557-564. DOI: 10.1080/14622200500185264 64. Mohanty N., Patra B.N.. Polypyrrole-Sodium Alginate Nanocomposites for Enhanced Removal of Toxic Organic and Metal Pollutants from Wastewater. Mater. Today Commun. (2023) 34 105325. DOI: 10.1016/j.mtcomm.2023.105325 65. Cameán A., López-Artíguez M., Roca I., Herce-Pagliai C., Menéndez M., Repetto M.. Determination of cobalt, manganese, and alcohol content in beers. J. Food Prot. (1998) 61 129-131. DOI: 10.4315/0362-028X-61.1.129 66. Eren H., Uzun H., Andac M., Bilir S.. Potentiometric monitoring of cobalt in beer sample by solid contact ion selective electrode. J. Food Drug Anal. (2014) 22 413-417. DOI: 10.1016/j.jfda.2014.03.003 67. Talan D., Huang Q.. A review study of rare Earth, Cobalt, Lithium, and Manganese in Coal-based sources and process development for their recovery. Miner. Eng. (2022) 189 107897. DOI: 10.1016/j.mineng.2022.107897 68. Lin Y., Chen Z., Dai M., Fang S., Liao Y., Yu Z., Ma X.. Co-pyrolysis kinetics of sewage sludge and bagasse using multiple normal distributed activation energy model (M-DAEM). Bioresour. Technol. (2018) 259 173-180. DOI: 10.1016/j.biortech.2018.03.036 69. Al-Jumaili A.A., Younus M.M., Kannan Y.J., Nooruldeen Z.E., Al-Nuseirat A.. Pharmaceutical regulations in Iraq: From medicine approval to post-marketing. East. Mediterr. Health J. (2021) 27 1007-1015. DOI: 10.26719/emhj.21.025 70. 70. National Research Council (NRC) Arsenic in Drinking Water UpdateNational Academy PressWashington, DC, USA2001Volume 7557564. Arsenic in Drinking Water Update (2001) Volume 7 557-564 71. Tangahu B.V., Sheikh Abdullah S.R., Basri H., Idris M., Anuar N., Mukhlisin M.. A review on heavy metals (As, Pb, and Hg) uptake by plants through phytoremediation. Int. J. Chem. Eng. (2011) 2011 939161. DOI: 10.1155/2011/939161 72. Shekar P.R., Nagaraju D., Ganesh V., Rao K.K.. Microhardness studies on as-grown (111) faces of some alkaline earth nitrates. Cryst. Res. Technol. J. Exp. Ind. Crystallogr. (2009) 44 652-656. DOI: 10.1002/crat.200800516 73. Patrick L.. Lead Toxicity, a review of the literature. Part I: Exposure, Evaluation, and treatment. Altern. Med. Rev. (2006) 11 2-22. PMID: 16597190 74. Alli L.A.. Blood level of cadmium and lead in occupationally exposed persons in Gwagwalada, Abuja, Nigeria. Interdiscip. Toxicol. (2015) 8 146-150. DOI: 10.1515/intox-2015-0022 75. Devoy J., Géhin A., Müller S., Melczer M., Remy A., Antoine G., Sponne I.. Evaluation of chromium in red blood cells as an indicator of exposure to hexavalent chromium. Toxicol. Lett. (2016) 255 63-70. DOI: 10.1016/j.toxlet.2016.05.008 76. Sanchez T.. Effects of mercury, lead, arsenic and zinc to human renal oxidative stress and functions: A review. Arch. Med. (2018) 4. DOI: 10.21767/2473-6457.10027 77. Ge X., Wang F., Zhong Y., Lv Y., Jiang C., Zhou Y., Li D., Xia B., Su C., Cheng H.. Manganese in blood cells as anexposure biomarker in manganese-exposed workers healthy cohort. J. Trace Elem. Med. Biol. (2018) 45 41-47. DOI: 10.1016/j.jtemb.2017.09.016 78. Okonkwo S.I., Idakwo S.O., Ameh E.G.. Heavy metal contamination and ecological risk assessment of soils around the pegmatite mining sites at Olode area, Ibadan southwestern Nigeria. Environ. Nanotechnol. Monit. Manag. (2021) 15 100424. DOI: 10.1016/j.enmm.2020.100424 79. Phakoago M.V., Ekosse G.E., Odiyo J.O.. The prevalence of geophagic practices and causative reasons for geophagia in Sekhukhune area, Limpopo Province, South Africa. Trans. R. Soc. South Afr. (2019) 74 19-26. DOI: 10.1080/0035919X.2019.1572669
--- title: Assessing Physical Activity Levels among Chinese College Students by BMI, HR, and Multi-Sensor Activity Monitors authors: - Dansong Liu - Xiaojuan Li - Qi Han - Bo Zhang - Xin Wei - Shuang Li - Xuemei Sui - Qirong Wang journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049372 doi: 10.3390/ijerph20065184 license: CC BY 4.0 --- # Assessing Physical Activity Levels among Chinese College Students by BMI, HR, and Multi-Sensor Activity Monitors ## Abstract We investigated the use of multi-sensor physical activity monitors, body mass index (BMI), and heart rate (HR) to measure energy expenditure (EE) of various physical activity levels among Chinese collegiate students, compared with portable indirect calorimetry. Methods: In a laboratory experiment, 100 college students, 18–25 years old, wore the SenseWear Pro3 Armband™ (SWA; BodyMedia, Inc., Pittsburg, PA, USA) and performed 7 different physical activities. EE was measured by indirect calorimetry, while body motion and accelerations were measured with an SWA accelerometer. Special attention was paid to the analysis of unidirectional and three-directional accelerometer output. Results: Seven physical activities were recorded and distinguished by SWA, and different physical activities demonstrated different data features. The mean values of acceleration ACz (longitudinal accel point, axis Z) and VM (vector magnitude) were significantly different ($$p \leq 0.000$$, $p \leq 0.05$) for different physical activities, whereas no significant difference was found in one single physical activity with varied speeds ($$p \leq 0.9486$$, $p \leq 0.05$). When all physical activities were included in a correlation regression analysis, a strong linear correlation between the EE and accelerometer reporting value was found. According to the correlation analysis, sex, BMI, HR, ACz, and VM were independent variables, and the EE algorithm model demonstrated a high correlation coefficient R2 value of 0.7. Conclusions: The predictive energy consumption model of physical activity based on multi-sensor physical activity monitors, BMI, and HR demonstrated high accuracy and can be applied to daily physical activity monitoring among Chinese collegiate students. ## 1. Introduction The need for reliable methods to accurately calculate daily energy expenditure is important for public health supervision, especially for monitoring the epidemic of overweight and obesity among various populations. Physical activity is associated with a reduced risk of morbidity and mortality in many chronic diseases, including CVD, diabetes, obesity, and some tumors [1,2]. Physical activities have also been shown to promote body weight control with appropriate nutrition and energy intake. The ability to accurately assess energy expenditure (EE) in free-living individuals can enhance the knowledge related to the link between the dose of physical activity and health status, as well as improve the understanding of how energy expenditure impacts energy balance related to body weight control and chronic diseases such as diabetes mellitus [3]. Research on the energy expenditure of physical activity can identify the recommended amount of targeted daily physical activity, which can be specific and personalized, such as the recommended amount of physical activity for different ages, genders, and population cohorts (e.g., for obese and overweight people). In addition to these benefits, a significant number of residents in China do not participate in sufficient amounts of physical activity to achieve its health benefits. A large sample size is always preferred in a physical activity and health study. Traditional physical activity research is majorly based on questionnaires, which is less objective. One of the challenges in energy expenditure research is the ability to accurately assess physical activity in free-living individuals. Numerous methods are available, but each of the current methods to assess physical activity and energy expenditure has limitations [4]. In the last decades, scientists have developed direct calorimetry (DC), doubly labeled water (DLW), indirect calorimetry (IC), and motion sensors for measuring energy expenditure with different physical activities (PAEEs), from which, DC, IC, and DLW are considered to be the gold standard of measuring [5]. Meanwhile, direct measuring is expensive and restricts people to a small and limited space, and it is not feasible to conduct a large cohort study at the same time. Doubly labeled water (DLW) is considered to be the most accurate technique to assess energy expenditure. The DLW method calculates the energy consumption of the human body according to the elimination rate of 2H and 18O from the human body within a period of time (5–14 days). It has the advantages of high accuracy, convenient sampling, wide application, and long-term monitoring, but it is expensive and cannot measure energy consumption in a short time for specific physical activity; therefore, its application is limited [6]. The IC method is also known as the gas metabolism method. Its principle is to estimate energy consumption by measuring gas exchange in human breath. Its main instruments are represented by COSMED K4 series and CORTEX METAMAX3B. However, these devices can only carry out monitoring for a short period of time (such as 2–4 h) and need to be equipped with a computer terminal near the test site. The subjects are also prone to discomfort when wearing breathing masks, so the IC method is limited to the energy consumption test under the condition of small sample sizes and short-term exercise. Self-reporting physical activity questionnaires have the limitation of reporting accuracy, compared with objective techniques. Motion sensors mainly include accelerometers and pedometers, which can indirectly reflect the level of physical activity by recording the vibration of the body during physical activity. A motion sensor is easy to wear and has high accuracy, so it has a wide application prospect in physical activity monitoring. However, motion sensors cannot accurately monitor energy consumption in activities, such as cycling and static fitness, due to small body vibration, which is a major limitation of motion sensors [7]. Because of the cost and technical demands, these methods are limited to small, pilot, and validation studies. Methods with good validity tend to be very costly or complicated for outdoor activities; however, practical and feasible methods for large populations have the limitation of poor accuracy and/or reliability. Developing an objective and accurate method to assess energy expenditure associated with physical activity continues to be a need in the field. Accelerometers are commonly used because they are practical and effective. Many commercial accelerometers are currently available, and there is considerable confusion over the appropriate test index and method to convert accelerometer counts into estimates of physical activity or energy expenditure [8,9]. As the main tool of this physical activity research, most accelerometer data analysis methods and energy consumption prediction models are based on the daily physical activity of adults. Because of the different characteristics of accelerometers, determining the accuracy and efficiency of accelerometers has become the key factor in relevant research. CSA, RT3, and SWA accelerometers have been widely used in research in recent years, and a large number of experiments have been carried out to illustrate their accuracy and applicability for energy consumption prediction [10,11]. Their methods and experimental designs are advanced and comprehensive. With the increasing understanding of the correlation between physical activity and health, the monitoring of physical activity EE and exercise intervention methods have gradually become a key focus of physical activity research. Scholars have been seeking a convenient, accurate, and practical monitoring method to effectively calculate and evaluate the energy consumption generated by physical activity. This method should first identify the differences in physical activities, and then calculate energy consumption by an accurate algorithm method. Therefore, this study, using a population of Chinese college students, establishes an energy consumption model by monitoring seven kinds of physical activities with a triaxial acceleration sensor. Through the calculation and analysis of the data available, a graph is drawn to verify the triaxial acceleration sensor in distinguishing different types of body activity, and the accuracy of the energy consumption algorithm is improved by adding additional relevant parameters (BMI, HR, and sex). ## 2.1. Participants One hundred Chinese collegiate students (aged 18–25) were recruited to participate in this research (as shown in Table 1). Subjects who met the exclusion criteria were not included (exclusion criteria: major disease or illness, use of medications that would affect body weight or metabolism, a current smoker). Participants were aware of the procedures and purposes of the study before they signed the informed consent documents. Participants were required to eat normally one week before the test and not to eat within 2 h before testing. A total of 80 participants enrolled as EXPERIMENTAL GROUP and took part in the test, and the other 20 enrolled as CONTROL GROUP to verify the accuracy of the model. ## 2.2. Measurement of Descriptive Variables All participants were in light-weight clothing (shorts, t-shirt, and barefoot) before the exercise sessions. Each subject arrived at the laboratory one hour before the test started. Weight and height were determined without shoes. Body weight was measured to the nearest 0.1 kg using a JY-200D height/weight digital scale (Jingyi equipment Co., Ltd., Beijing, China), and body height was determined to the nearest 0.1 cm using a horizontal headboard with an attached wall-mounted metric rule. Body mass index (BMI, kg/m2) was computed from weight and height. ## 2.3. Exercise Protocols Seven separate activities were included in this project (rest, sit, treadmill walk, normal run, stair walk, cycle ergometer, push-ups). The specific protocols are shown in Table 2. The order of these exercise protocols was performed randomly. Walking was performed on a motorized treadmill. Stair walk was performed with a 20.3 cm (6-inch) bench stair in the following sequence: left leg up, right leg up, left leg down, right leg down. Cycle ergometer exercise was performed on a Monark 827E stationary cycle ergometer (Monark Exercise AB, Vansbro, Sweden). A metronome was used to set pace during stair stepping and performing push-ups. During each activity, energy expenditure was measured simultaneously. Exercise heart rate was used to assess exercise intensity at each minute using a Polar heart rate monitor [12,13]. The experiments were carried out in the Exercise Physiology Laboratory (Wuhan Sports University, Wuhan, China), and the room was maintained at a temperature of 20~25 °C and 50~$80\%$ humidity. Participants were asked to perform the test one after another. In order to strictly control the airflow, other subjects and staff were not allowed to enter the room without permission during the test. The research project was approved by the Bioethics Committee of the Academy of Physical Education of Wuhan Sports University (Wuhan, China) and was conducted in accordance with the Declaration of Helsinki. The participants were informed and fully aware of the procedures of this study and acknowledged all the risks and benefits before they were recruited. All participants conducted the physical activities as they were required (as shown in Table 2). ## 2.4. Motion Sensors The SenseWear Pro3 Armband™ (Body Media, Pittsburgh, PA, USA) is a commercially available, comfortable, non-invasive physical activity monitor that is worn on the upper arm over the triceps muscle and provides information about body position (lying or upright) by detecting accelerations. The armband was placed on each subject’s arm before entering the laboratory, and the subject remained in a seated position for a period of 5 min before data collection to allow for acclimation to skin temperature. Energy expenditure during exercise was captured at 60 s intervals, the real-time sampling frequency was checked every 0.1 s, and the peak acceleration and mean absolute difference (MAD) values were captured every 6 s. ## 2.5. Energy Expenditure Energy expenditure during various exercises was examined via open-circuit, indirect calorimetry with the Cortex MetaMax3BTM metabolic system (Cortex Biophysik GmbH, Leipzig, Germany). Cortex MetaMax3BTM system, which is a battery-operated, portable, wireless metabolic system measuring gas exchange breath-by-breath, has been reported to be a valid and reliable measure of oxygen uptake [12]. The face mask was connected to a flow sensor to detect airflow from the rotation of fans, which allowed the determination of ventilation. A sampling line connected both flow sensor unit and sensor box; oxygen (O2) and Carbon Dioxide (CO2) from expired air were analyzed using a micro-fuel cell and thermal conductivity, respectively. Cortex MetaMax3B software was used to compute energy expenditure, which included oxygen uptake in milliliters per minute (mL/min), milliliters per kilogram of body weight per minute (mL/kg/min), and kilocalories per minute (Kcal/min). Experiments were carried out in a regular temperature and humidity lab setting for all exercise tests, airflow calibration was performed using an automatic flow calibrator, and the gas analyzers were calibrated ($5\%$ O2, $16\%$ CO2, and $79\%$ N2). All experiments were performed for at least 5 min, and a minimum of 3 min of resting gas exchange assessment data were collected both before and after each of the tests. The purpose of the resting data assessment is to make sure for each individual to reach a steady state during the entire test, and the time it takes to reach their steady state depends on their current physical and physiological characteristics. Participants were appropriately warmed up before the testing, and they were repeatedly encouraged to complete the activity with their “regular” habits and pace. ## 2.6. HR Monitoring HR was measured simultaneously using a Polar T31 telemetric system (Polar Electro OY, Kempele, Finland). The participants’ HR was monitored minute-by-minute during each activity. Each participant underwent individual calibrations to establish the relationship between HR and energy expenditure in all of the tested activities. The calibration activities were carried out in sequence, with a break to allow recovery of resting HR. ## 2.7. Statistical Analysis Statistical analysis and algorithm construction were performed using STATA 13.0 (StataCorp LLC, College Station, TX, USA). Data were analyzed for each exercise procedure. This study evaluated the validation between estimated energy expenditure from Cortex MetaMax3B and triaxial acceleration data from an SWA sensor and built energy expenditure algorithms based on seven different activities. Energy expenditure across each activity protocol was analyzed with ANOVA to assess mean differences in different physical activities. Dependent t-tests were performed to compare triaxial differences in different activities and within the same physical activity under different speeds/workloads. Statistical significance was defined with p-value ≤ 0.05. The gender estimation of energy expenditure was significantly different; therefore, men and women were separated for all analyses. Graphical procedures were used to spot differences in activities with triaxial accelerometer. SenseWear Pro3 Armband™ measures acceleration and deceleration in the three dimensions of space according to ACx (forward accel point), ACy (transverse accel point), and ACz (longitudinal accel point). Additionally, VM (vector magnitude) is calculated as VM = (ACx2 + ACy2 + ACz2)$\frac{1}{2.}$ ## 3.1. Posture and Movement Classification This study analyzed the validity of the SenseWear Pro3 Armband™ to capture and recognize different modes of activities in a laboratory setting. When the generalized algorithm provided by the manufacturer was applied to the data, the three-axis accelerometer data significantly distinguished these seven activities with ACx, Acy, and ACz reads (see Figure 1). Tilt sensing is a basic function provided by accelerometers in response to gravity or constant acceleration. Therefore, human postures, such as sitting and lying, can be distinguished according to the signal magnitude of accelerations along sensitive axes [14,15]. ## 3.2. Construction and Analysis of Energy Consumption Model of Physical Activity The data obtained from the triaxial acceleration sensor were imported into STATA 13.0 to match with the data from the gas analyzer, and the non-conforming acceleration data were deleted to ensure that the acceleration value was completely corresponding to the MetaMax 3B data in time. Stepwise regression was used to construct the physical activity energy consumption model. According to relevant studies, the vertical acceleration value of the longitudinal accel (ACz) and VM values have a significant effect on distinguishing different physical activities ($$p \leq 0.000$$, $p \leq 0.05$) [16,17]. The value of seven activities was calculated (as seen in Table 3). No significant difference was found in treadmill walking between slow and fast speeds ($$p \leq 0.9486$$, $p \leq 0.05$). Pearson correlation analysis was used to analyze the relationship between the BMI value, ACz value, VM value, and energy consumption in the test, and the correlation significance between energy consumption and other test values was statistically calculated. The correlation coefficients between the ACz value, VM value, BMI value, and energy consumption W were 0.59 ($$p \leq 0.000$$), 0.76 ($$p \leq 0.000$$), and 0.29 ($$p \leq 0.011$$), respectively, and all p-values were less than 0.05. BMI, ACz, VM, and energy consumption values were significantly correlated. Therefore, it is feasible to construct the energy consumption model with these three variables (as shown in Table 4 and Table 5). The basic energy consumption algorithm equation is W/min = β0 + β1 × ACz, and can be constructed with additional factors, e.g., sex, BMI, and HR, and the correlation coefficients were 0.46, 0.47, and 0.7, respectively. The accuracy of the algorithm was gradually improved when all of these variables were added to the model. The comprehensive energy consumption algorithm (based on the ACz) was as follows:W/min = −9.173 + 0.004 × ACz + 1.097 × SEX + 0.227 × BMI + 0.057 × HR ($M = 1$, $F = 0$) The comprehensive energy consumption algorithm (based on the VM axis) was as follows:W/min = −12.27 + 0.008 × VM + 0.953 × SEX + 0.232 × BMI + 0.059 × HR ($M = 1$, $F = 0$) As a result of this study, we developed new proprietary exercise-specific algorithms for common physical activities of Chinese college students. When the exercise-specific algorithms were applied to the data from the SenseWear Pro3 Armband™, the estimate of energy expenditure appeared to be improved. The data were put into the models, and the correlation between the calculated value and the real value was verified by comparing it with the results from Cortex MetaMax3B™. According to the results, the correlation coefficients between the actual energy consumption measured by Cortex MetaMax3B™ and the results calculated by the model were 0.8674 (ACz) and 0.88 (VM), respectively, both of which were greater than $85\%$, indicating that the algorithms accurately predicted energy consumption for different activities. Different parameters can be used to build different types of algorithms, and the accuracy of different algorithms varies greatly. Progress has been made to solve these issues, but it is likely that the fundamental challenge is to build an accurate and feasible energy expenditure algorithm with multiple parameters. Direct calorimetry is considered the most accurate method to assess physical activity by measuring gas exchange and interpreting it into energy expenditure. As a result, an algorithm model was built to improve the estimation of energy expenditure when used in combination with parameters such as accelerometry, HR, sex, and BMI. ## 4. Discussion The current study was conducted to investigate the relationship between energy expenditure and body acceleration during different physical activities. The development of an accurate, reliable, and feasible model to calculate daily energy expenditure in free-living conditions is an important priority for public health researchers. Our study built a correlation regression model based on a triaxial acceleration monitor during seven physical activities, integrating the variables: BMI, sex, and HR. Our correlation coefficient, R, became greater when integrating all three variables than when only using one or two of them, which indicates achieving greater accuracy of the regression model. Biomechanical detection for the movement of the human body and the classification of motion using accelerometer and center of gravity (COG)-based methodologies have been adopted globally. Approaches to the classification of biomechanical movement can be made by threshold-based or statistical-based classification schemes. Threshold-based motion classification takes advantage of well-known knowledge and information about the movements to be classified. Statistical-based motion classification utilizes a supervised machine learning procedure, which associates observations (or features) of movement to possible classifications in terms of the probability of the observation (see Figure 1). Spatial sensing is a basic function of accelerometers, which can respond to the change in the center of gravity or constant acceleration. Therefore, by wearing an accelerometer on the torso or arm, human postures (such as sitting, walking, and running) can be distinguished according to the signal magnitude of accelerations along the measuring axes. By using the wearable triaxial acceleration sensors, different types of motion can be identified by the three-directional variable, which was obviously distinguished in seven different types of physical activities. In our research, the postures tested can be distinguished by observing different orientations of body segments and the changes in the spatial movement of the body. The sedentary activities (rest and sitting) demonstrated almost the same spatial acceleration image characteristics, whereas the other activities produced distinct spatial acceleration images. Triaxial acceleration signals and discrete wavelet transformations can determine activities in ambulatory movement. Above all these three-dimensional data, vertical acceleration signals are best for distinguishing between different types of motion, as they are characterized by vertical acceleration and frequency peak in the signal spectrum. Several other studies have investigated the accuracy of posture recognition by accelerometers. Mathie et al. [ 18] reported a general classification framework consisting of a hierarchical binary tree for classifying postures, e.g., falls, jumps, walks, and other movements, using signals from a wearable triaxial accelerometer. This modular structure also allows modifying the algorithms for each classification under certain conditions or particular purposes. Trunk tilt variation due to a sit–stand postural shift was reported to be measured by integrating the signal from a gyroscope attached to the chest of the examinee [19]. The sit–stand postural shift can be recognized according to the patterns of vertical acceleration from an accelerometer at the waist [20]. Although Yang et al. [ 21] used a simplified scheme with a tilt threshold to distinguish standing and sitting, a single-accelerometer approach has difficulty in distinguishing between standing and sitting as both are upright postures. Information from the sensors together with sex, BMI, and HR was integrated into proprietary algorithms to estimate energy expenditure (EE). The significant correlations between energy expenditure and accelerometer readings are found in the laboratory and under free-living conditions, and the relationship between these parameters using triaxial accelerometers varies between different types of physical activities. From the results of our research based on Chinese college students, we developed new proprietary exercise-specific algorithms for energy expenditure prediction. When the exercise-specific algorithms were applied to the data from the control group and energy cost from Cortex MetaMax3BTM, the estimate of energy expenditure appeared to be improved. To facilitate comparison, we classified the seven different activities into the walking and running group and the non-walking and running group (rest, sitting, cycle ergometer, pushup, and stair walking). An energy consumption correlation analysis was carried out for the two groups, as illustrated in Figure 2. There was a linear correlation between total energy expenditure examined via the Cortex MetaMax3BTM and total energy expenditure estimated using indirect calorimetry in this study. The distribution of the scatter plot shows that the values calculated by the energy consumption algorithm based on ACz and VM have a good fit with the actual measured data from Cortex MetaMax3BTM. As can be seen from the scatter plot, the energy consumption of the walking and running group is higher than that of the non-walking and running group. The energy consumption algorithm based on ACz shows a higher difference between the two groups, whereas the energy consumption algorithm based on VM shows a small difference between the two groups. The reason may be that the VM value is a composite value of three-dimensional acceleration, which can reduce the difference when calculating the energy consumption of different types of physical activities. Relevant studies have shown that there is a high correlation between the three-dimensional space axis and PAEE, and the linear algorithm is easy to calculate, so most researchers construct linear equations and use ACz/VM as the independent variable to predict the energy consumption of physical activity [22]. The results show that the correlation between the energy consumption algorithm and the IC method is between 0.50 and 0.90. The classical energy consumption prediction formula, the Freedson formula, is based on a treadmill with three different speeds: 4.8 km/h, 6.4 km/h, and 9.7 km/h, and the R2 of the energy consumption algorithm is 0.82 [23]. Other researchers also took walking and running as normal daily activities and established multiple energy consumption models with ACz or VM as independent variables, gradually improving the comprehensiveness of the test and the accuracy of the prediction [24]. Different accelerometers have different validity in adult physical activity studies, and the SenseWear Pro3 Armband™ and some other accelerometers have been used during exercise to assess energy expenditure. For example, the research found that the SenseWear Pro3 underestimated energy expenditure, particularly for monitoring high-intensity exercise. However, there is currently no cure for improving the underestimation; therefore, the underestimation of energy expenditure continues to be a problem of many accelerometer-based physical activity monitors that are currently available. In a study comparing a triaxial accelerometer to indirect calorimetry, Jakicic et al. [ 25] reported that the accelerometer under-evaluated energy expenditure by a total of 30–50 kcal for 30 min of walking, 87–89 kcal for 20 min of cycling, and 44–51 kcal for 20 min of stair stepping. In another study of Jakicic et al. [ 26], the difference between indirect calorimetry and the SenseWear Pro3 Armband™ was studied by using similar exercise protocols; the total energy expenditure was also underrated. Jakicic found that the SenseWear Pro3 Armband™ is able to give an accurate estimate of energy expenditure of exercises mainly involving the upper extremities [26]. These results suggest that the SenseWear Pro3Armband™ may provide a more accurate estimate of energy expenditure using exercise-specific algorithms. Hustvedt et al. [ 27] evaluated the accuracy of the ActiReg (a three-dimensional accelerometer) alone and in combination with an HR monitor. The mean TEE (total energy expenditure) measured by the ActiReg was not different from DLW (doubly labeled water) ($$p \leq 0.45$$). Bland–Altman plots showed that the ActiReg underestimated TEE at high-intensity exercise, and the underestimation of TEE was corrected by using an HR monitor. Plasqui et al. [ 28] evaluated another three-dimensional monitor called the Tracmor. The participants’ age, body mass, and height were shown to explain $64\%$ of the variation in DLW-measured TEE, and by adding Tracmor activity counts to the model, there was an increase in explained variation of $19\%$ (total R2 = 0.83). Our study suggests that when exercise-specific algorithms are used in combination with triaxial acceleration, sex, BMI, and HR, this results in providing a more accurate estimate of energy expenditure, indicating that the body shifts and other physiological sensors provide useful information to improve estimates of energy expenditure. Despite the promising results obtained from this study, there are limitations that need to be discussed. In the current study, the researchers focused on the accuracy of exercise-specific algorithms and many different factors, such as age, gender, and activity forms, which need to be considered to estimate energy expenditure. The exercise-specific algorithms from our study are based on a laboratory study, and may not be as accurate under free-living conditions. This study is based on Chinese college students, and the participants are relatively young adults with normal body weights. It is unclear whether our findings hold true for individuals of different ages, weights, or levels of physical fitness. ## 5. Conclusions In summary, by the use of a triaxial acceleration sensor, body postures of seven physical activities were distinguished, and the algorithms showed promise for accurately measuring energy expenditure. On the one hand, the accuracy of the algorithms was improved by the additional three parameters (sex, BMI, and HR). On the other hand, the manufacturers of some accelerometers have already written the algorithms for their products, but they still need to improve the accuracy of the original data collected during each of the exercises to estimate energy expenditure. In future research, the accuracy of the model can be continuously improved by increasing the number of subjects. Meanwhile, the physical activities for measuring energy consumption can be more diversified and include activities such as outdoor sports, various ball games, and competitive and confrontational sports, such as tennis, table tennis, badminton, swimming, basketball, and football. In the process of constructing the energy consumption algorithm, different analysis methods can be added, such as the piecewise model, neural network model, etc., to enhance the accuracy of the actual prediction of energy expenditure. ## References 1. Patrick K., Norman G.J., Calfas K.J., Sallis J.F., Zabinski M.F., Rupp J., Cella J.. Diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch. Pediatr. Adolesc. Med. (2004) 158 385-390. DOI: 10.1001/archpedi.158.4.385 2. Aballay L.R., Eynard A.R., Diaz Mdel P., Navarro A., Munoz S.E.. Overweight and obesity: A review of their relationship to metabolic syndrome, cardiovascular disease, and cancer in South America. Nutr. Rev. (2013) 71 168-179. DOI: 10.1111/j.1753-4887.2012.00533.x 3. Van Remoortel H., Giavedoni S., Raste Y., Burtin C., Louvaris Z., Gimeno-Santos E., Langer D., Glendenning A., Hopkinson N.S., Vogiatzis I.. Validity of activity monitors in health and chronic disease: A systematic review. Int. J. Behav. Nutr. Phys. Act. (2012) 9 84. DOI: 10.1186/1479-5868-9-84 4. Vanhees L., Lefevre J., Philippaerts R., Martens M., Huygens W., Troosters T., Beunen G.. How to assess physical activity. Eur. J. Cardiovasc. Prev. Rehabil. (2005) 12 102-114. DOI: 10.1097/01.hjr.0000161551.73095.9c 5. Valanou E.M., Bamia C., Trichopoulou A.. Methodology of physical-activity and energy-expenditure assessment: A review. J. Public Health (2006) 14 58-65. DOI: 10.1007/s10389-006-0021-0 6. White T., Westgate K., Hollidge S., Venables M., Olivier P., Wareham N., Brage S.. Estimating energy expenditure from wrist and thigh accelerometry in free-living adults: A doubly labelled water study. Int. J. Obes. (2019) 43 2333-2342. DOI: 10.1038/s41366-019-0352-x 7. Matthew C.E.. Calibration of accelerometer output for adults. Med. Sci. Sports Exerc. (2005) 37 S512-S522. DOI: 10.1249/01.mss.0000185659.11982.3d 8. Migueles J.H., Delisle Nystrom C., Henriksson P., Cadenas-Sanchez C., Ortega F.B., Lof M.. Accelerometer Data Processing and Energy Expenditure Estimation in Preschoolers. Med. Sci. Sports Exerc. (2019) 51 590-598. DOI: 10.1249/MSS.0000000000001797 9. Lettink A., Altenburg T.M., Arts J., van Hees V.T., Chinapaw M.J.M.. Systematic review of accelerometer-based methods for 24-h physical behavior assessment in young children (0–5 years old). Int. J. Behav. Nutr. Phys. Act. (2022) 19 116. DOI: 10.1186/s12966-022-01296-y 10. Delisle Nystrom C., Pomeroy J., Henriksson P., Forsum E., Ortega F.B., Maddison R., Migueles J.H., Lof M.. Evaluation of the wrist-worn ActiGraph wGT3x-BT for estimating activity energy expenditure in preschool children. Eur. J. Clin. Nutr. (2017) 71 1212-1217. DOI: 10.1038/ejcn.2017.114 11. Jeran S., Steinbrecher A., Pischon T.. Prediction of activity-related energy expenditure using accelerometer-derived physical activity under free-living conditions: A systematic review. Int. J. Obes. (2016) 40 1187-1197. DOI: 10.1038/ijo.2016.14 12. Schneider P.L., Crouter S.E., Lukajic O., Bassett D.R.. Accuracy and reliability of 10 pedometers for measuring steps over a 400-m walk. Med. Sci. Sports Exerc. (2003) 35 1779-1784. DOI: 10.1249/01.MSS.0000089342.96098.C4 13. Seo D.C., King M.H., Kim N., Sovinski D., Meade R., Lederer A.M.. Predictors for moderate- and vigorous-intensity physical activity during an 18-month coordinated school health intervention. Prev. Med. (2013) 57 466-470. DOI: 10.1016/j.ypmed.2013.06.024 14. Bornstein D.B., Beets M.W., Byun W., McIver K.. Accelerometer-derived physical activity levels of preschoolers: A meta-analysis. J. Sci. Med. Sport (2011) 14 504-511. DOI: 10.1016/j.jsams.2011.05.007 15. Godfrey A., Bourke A.K., Olaighin G.M., van de Ven P., Nelson J.. Activity classification using a single chest mounted tri-axial accelerometer. Med. Eng. Phys. (2011) 33 1127-1135. DOI: 10.1016/j.medengphy.2011.05.002 16. Vetrovsky T., Clark C.C.T., Bisi M.C., Siranec M., Linhart A., Tufano J.J., Duncan M.J., Belohlavek J.. Advances in accelerometry for cardiovascular patients: A systematic review with practical recommendations. ESC Heart Fail. (2020) 7 2021-2031. DOI: 10.1002/ehf2.12781 17. Sardinha L.B., Judice P.B.. Usefulness of motion sensors to estimate energy expenditure in children and adults: A narrative review of studies using DLW. Eur. J. Clin. Nutr. (2017) 71 331-339. DOI: 10.1038/ejcn.2017.2 18. Mathie M.J., Celler B.G., Lovell N.H., Coster A.C.. Classification of basic daily movements using a triaxial accelerometer. Med. Biol. Eng. Comput. (2004) 42 679-687. DOI: 10.1007/BF02347551 19. Lyons G.M., Culhane K.M., Hilton D., Grace P.A., Lyons D.. A description of an accelerometer-based mobility monitoring technique. Med. Eng. Phys. (2005) 27 497-504. DOI: 10.1016/j.medengphy.2004.11.006 20. Karantonis D.M., Narayanan M.R., Mathie M., Lovell N.H., Celler B.G.. Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring. IEEE Trans. Inf. Technol. Biomed. (2006) 10 156-167. DOI: 10.1109/TITB.2005.856864 21. Yang C.C., Hsu Y.L.. Development of a wearable motion detector for telemonitoring and real-time identification of physical activity. Telemed. J. E Health (2009) 15 62-72. DOI: 10.1089/tmj.2008.0060 22. van Hees V.T.. The importance of nuance in statements about methods for human energy expenditure estimation that use motion sensors. Eur. J. Clin. Nutr. (2017) 71 1136-1137. DOI: 10.1038/ejcn.2017.65 23. Freedson P.S., Melanson E., Sirard J.. Calibration of the Computer Science and Applications, Inc. accelerometer. Med. Sci. Sports Exerc. (1998) 30 777-781. DOI: 10.1097/00005768-199805000-00021 24. Leenders N.Y., Nelson T.E., Sherman W.M.. Ability of different physical activity monitors to detect movement during treadmill walking. Int. J. Sports Med. (2003) 24 43-50. DOI: 10.1055/s-2003-37196 25. Jakicic J.M., Winters C., Lagally K., Ho J., Robertson R.J., Wing R.R.. The accuracy of the TriTrac-R3D accelerometer to estimate energy expenditure. Med. Sci. Sports Exerc. (1999) 31 747-754. DOI: 10.1097/00005768-199905000-00020 26. Jakicic J.M., Marcus M., Gallagher K.I., Randall C., Thomas E., Goss F.L., Robertson R.J.. Evaluation of the SenseWear Pro Armband to assess energy expenditure during exercise. Med. Sci. Sports Exerc. (2004) 36 897-904. DOI: 10.1249/01.MSS.0000126805.32659.43 27. Hustvedt B.E., Christophersen A., Johnsen L.R., Tomten H., McNeill G., Haggarty P., Lovo A.. Description and validation of the ActiReg: A novel instrument to measure physical activity and energy expenditure. Br. J. Nutr. (2004) 92 1001-1008. DOI: 10.1079/BJN20041272 28. Plasqui G., Joosen A.M., Kester A.D., Goris A.H., Westerterp K.R.. Measuring free-living energy expenditure and physical activity with triaxial accelerometry. Obes. Res. (2005) 13 1363-1369. DOI: 10.1038/oby.2005.165
--- title: Analysis of Plantar Tactile Sensitivity in Older Women after Conventional Proprioceptive Training and Exergame authors: - Claudio Henrique Meira Mascarenhas - José Ailton Oliveira Carneiro - Thaiza Teixeira Xavier Nobre - Ludmila Schettino - Claudineia Matos de Araujo - Luciana Araújo dos Reis - Marcos Henrique Fernandes journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049375 doi: 10.3390/ijerph20065033 license: CC BY 4.0 --- # Analysis of Plantar Tactile Sensitivity in Older Women after Conventional Proprioceptive Training and Exergame ## Abstract Objective: To evaluate and compare the effects of conventional proprioceptive training and games with motion monitoring on plantar tactile sensitivity in older women. Methods: A randomized controlled clinical trial, with 50 older women randomized into three groups: conventional proprioception ($$n = 17$$), games with motion monitoring ($$n = 16$$), and the control ($$n = 17$$). They underwent 24 intervention sessions, three times a week, for eight weeks. The conventional proprioception group performed exercises involving gait, balance, and proprioception. The games performed by the motion monitoring group included exercises using the Xbox Kinect One video game from Microsoft®. The evaluation of tactile pressure sensitivity was performed using Semmes–Weinstein monofilaments. Intragroup comparisons between the two paired samples were performed using paired Student’s t-test or Wilcoxon test. Intergroup comparisons between the three independent samples were performed using the Kruskal–Wallis test and Dunn’s post hoc test, with p ≤ 0.05. Results: The older women submitted to conventional games with motion monitoring training and showed improvement in plantar tactile sensitivity in the right and left feet. When comparing the intergroup results, the two training modalities obtained an improvement in the plantar tactile sensitivity of the older women when compared to the control group. Conclusions: We conclude that both training modalities may favor the improvement of plantar tactile sensitivity in older women, with no significant differences between conventional and virtual training. ## 1. Introduction Aging is a natural, dynamic, and progressive process, which is accompanied by morphological, functional, and biochemical changes, and an increase in the risk of several diseases, especially those related to the feet (neuropathy, diabetes, peripheral neuropathy, plantar fasciitis, diabetic foot, and others) [1,2]. In this regard, for every three older women living in the community, at least one of them has foot problems, especially women, who have about twice as many foot problems classified as moderate and severe [3]. In the somatosensory system, the peripheral nervous system can undergo changes such as the loss of myelinated and unmyelinated fibers and a decrease in nerve conduction velocity, leading the elderly to have a deficit in sensory discrimination [4,5,6]. With aging, there is a loss of receptors and also a reduction in the number of sensory fibers that innervate the peripheral receptors, which may cause peripheral neuropathies that affect the proprioceptive system [7,8]. The changes that arise in the feet of the elderly can compromise the performance of activities of daily living, interfering negatively with posture and gait and contributing to the development of disabilities. Among these alterations is the plantar tactile sensitivity impairment, which can bring consequences such as postural instability, gait disorders such as a reduction in the balance phase, speed and step symmetry, discomfort, pain, deformities, and risk of falls, thus impairing the quality of life of the elderly [4,5,6]. Management strategies focus on prevention, early detection, and the appropriate treatment of aging-related morbidities, reducing the socioeconomic impact, not only for the individual but also for society [6,7,8]. Proprioceptive rehabilitation is a resource that is widely used in physiotherapy with the purpose of stimulating the sensorimotor system. There are many ways and tools to structure these exercise programs; among them, the training conventional proprioceptive method is widely disseminated in the literature and relies on the use of materials such as balls, sticks, balance boards, and mats, among other resources for this purpose [2,3,4,5]. Despite the relevance of the problem, there is a scarcity of studies available in the literature that is aimed at comparing the effects of different training modalities on plantar tactile sensitivity in the healthy elderly population. Some research with emphasis on proprioceptive stimulation, sensory stimulation, and active exercises have demonstrated important results in the improvement of tactile sensitivity, but these studies include older women with some clinical impairment [7,9]. From this perspective, the present study aimed to evaluate and compare the effects of conventional proprioceptive training and games with motion monitoring on the plantar tactile sensitivity of older women. ## 2.1. Characterization of the Study This is a randomized controlled clinical trial, which was developed according to the recommendations of CONSORT (consolidated standards of reporting trials) [10]. It was developed in the city of Jequié-Bahia, and the sample was composed of older women who participated in four senior citizenship groups. ## 2.2. Ethical Issues This study was approved by the Research Ethics Committee of the Universidade Estadual do Sudoeste da Bahia (UESB) under opinion number 2.627.047, CAAE: 46887315.1.0000.005. The study was registered in the Brazilian Registry of Clinical Trials (REBEC) database, registration number RBR-592yyp. ## 2.3. Population and Sample To include the participants in the study, the following criteria were used: (a) a maximum age of 79 years old; (b) no practice of any type of physical exercise in the last three months; (c) absence of cognitive deficit, diabetes mellitus, vestibulopathies, cardiovascular diseases, visual, or hearing impairment; (d) independent ambulation and locomotion without auxiliary devices. The study excluded older women who had attended another proprioceptive rehabilitation program during the training or in the last three months and those who participated in less than $75\%$ of the training program. The sample size was defined based on the results of a pilot study with 5 older women in each group and had as an outcome the difference (i.e., performance before training or control—performance after training or control) in the TUGT (Timed Up & Go test). For the sample calculation, α = 0.05 and test power (1-β) = 0.95 were considered, with 3 groups (control x conventional x exergame), which obtained a sample number of 36 individuals (i.e., 12 in each group). Considering the possibility of sample loss over the course of the 8-week intervention, the sample size was estimated with a $25\%$ loss margin in each group; therefore, a sample number of 15 older women per group was expected (i.e., a total sample of 45 elderly). The calculation of the sample size was performed using the GPower® software version 3.1. After screening the participants according to the established criteria, and including 50 older women remained in the sample, which was submitted to stratified randomization by age (60–$\frac{69}{70}$–79) and BMI (low—less than or equal to 22.0/high—greater than or equal to 27.0), thus seeking a greater homogeneity in the allocation of the older women among the groups. From this stratification, the participants were distributed into four groups: age (60–69) and low BMI, age (60–69) and high BMI, age (70–79) and low BMI, and age (70–79) and high BMI. ## 2.4. Procedures Subsequently, a code was created for each participant, and randomization was performed in blocks of three individuals for each stratum. The blocks were randomized using Microsoft Excel version 2013 software, and subsequently, the codes were distributed in three arms of the study (control group, conventional group, and exergame group). The entire process was performed by a researcher with no clinical involvement in the trial, thus ensuring allocation confidentiality. The control and conventional groups were composed of 17 participants, and the exergame group was composed of 16 participants; at the end of the study, each group ended with 15 participants. The losses were related to participation below $75\%$ of the training program (3 older women) and dropouts (2 older women), totaling 5 losses. The control group (GCT), during the intervention period, did not participate in any training modality; the conventional group (GCV) participated in conventional proprioceptive training; and the exergame group (GEX) participated in proprioceptive training based on virtual realities. ## 2.4.1. Conventional Proprioceptive Training (Conventional Group/GCV) The training was carried out three times a week, for 8 weeks, for a total of 24 sessions, with a duration of 50 min per session and a minimum interval of 48 h between each session. The training protocol was organized as follows: warm-up (10 min), proprioceptive training (30 min), and cool-down (10 min), with the monitoring of blood pressure and heart rate before and after the activities. The warm-up was performed with walking (4 min) and stretching exercises for the muscles of the upper and lower limbs and spine (6 min). The warm-up was conducted with breathing exercises (5 min) and stretching exercises (5 min). The participants were warned not to alter their activities of daily living during the intervention period, thus avoiding the possible influences of external factors on the outcomes of the research. The conventional proprioceptive training protocol involved gait, balance, and proprioception training and was spatially organized in the form of a circuit with different textures and obstacles, consisting of seven stations. The materials used were: 1 dense mattress of dimension 120 × 70 × 10 cm (station 1), 1 foam module—a mini beam of dimension 190 × 22 × 10 cm (station 2), 4 agility rings with 42 cm diameter (station 3), 1 proprioceptive lateral board of dimension 60 × 36 × 8 cm (station 4), 2 agility cones of the dimensions 23 × 14 cm (station 5), 1 proprioceptive disc of 40 cm in diameter (station 6), and 3 agility barriers of the dimensions 70 × 15/ 70 × 20/ 70 × 25 cm (station 7). The older women participated, in groups of two or three, in specific exercises at each station that were combined with sensory and motor stimulation, as follows:-Station 1: Lateral strides (right and left), forward and backward strides on an unstable surface (dense mattress), exercises in bipodal and unipodal support (right and left) with eyes open and closed, agility training with ball throwing.-Station 2: Forward, backward, and sideward march (right and left) with a narrow base on an unstable surface (mini foam board), march alternating between floor and mini-board, agility training with ball throw.-Station 3: Forward, backward, sideways, and cross-legged march between the agility rings.-Station 4: Latero-lateral and anteroposterior exercise on the lateral proprioceptive board with eyes open and closed, agility training with ball throw.-Station 5: Forward, backward, and sideward march between cones with a narrow base and circumferential path with full foot support, with heel support only, and with forefoot support only.-Station 6: Exercises on the proprioceptive disk with multidirectional shifts with eyes open and closed and agility training with ball throwing.-Station 7: Forward, backward, and sideward march over agility barriers and agility training with ball throw. Each participant remained for two minutes at each station, with a thirty-second break between stations. After going through all seven stations, the forward, sideways, and backward march was performed again through all the stations continuously without breaks, and only a thirty-second break at the end of each circuit, until the proposed time of 30 min was completed. The degree of difficulty was increased throughout the training through the speed and execution of the activities. In all sessions, each elderly woman was accompanied by a researcher, and the execution and physical capacity of each participant in relation to the execution of the activities were taken into consideration. The exercises of the conventional training protocol were based on the consulted literature [9,10]. ## 2.4.2. Proprioceptive Training Based on Virtual Realities (Exergame Group/GEX) The proprioceptive training based on virtual realities “exergames” was performed using the Xbox Kinect One videogame from Microsoft®. This console uses technology with motion sensors and the Kinect, which captures the movements of the players, i.e., they are sensitive to changes in direction, speed, and acceleration, thus allowing the games to be controlled with body movement without the need to use any manual control [11]. The game used was Kinect Sports Rivals, which simulates six sports activities: jet ski racing, climbing, soccer, bowling, tennis, and target shooting. The selection of the games was guided by the analysis of the motor demands offered by each one of them, which ranged from basic motor skills, such as squatting and lifting, jumping, turning, tilting the trunk, moving laterally and anteroposteriorly, and moving the arms in all directions to more complex motor skills that stimulated coordination, balance, stability, and proprioception, such as extending an arm and flexing the contralateral leg, which is associated with body thrust (climbing game). This also included performing lateral-lateral displacements associated with the flexion/extension and adduction/abduction movements of the upper limbs (a tennis game); performing kicks, displacements, and body rotation (a soccer game); performing hip and knee flexion, with trunk rotation and inclination (jet ski game); and performing hip, knee, and ankle flexion with lower limbs in an alternate position, associated with trunk inclination and shoulder flexion/extension movement (bowling game). The training with exergames was carried out in a room with no objects that could interfere with the performance of the older women, and in which the games were projected on the wall using an Epson Power Lite S8+ projector and a set of Multilaser 60 WRms Sp088 speakers. The participants were accompanied by researchers and performed the activities in pairs, barefoot, and positioned in front of the Kinect sensor at a distance of three meters. Each session consisted of training with three games previously selected by lottery, and the duration of each game was 10 min, for a total of 30 min. The order of the games in each session was also conducted by lottery; every six sessions, a new lottery was held, where one game was replaced by another, allowing the participants to have contact at the end of the training with all five selected games. ## 2.4.3. Instruments The study used a questionnaire with sociodemographic variables (age, marital status, education, and monthly family income) and health-related variables (body mass index/BMI, presence of diagnosed diseases, musculoskeletal pain in the last 7 days, musculoskeletal pain in the last 12 months, and medications), and the evaluation of plantar tactile sensitivity. The evaluation of tactile pressure sensitivity in the plantar region was performed through the Semmes-Weinstein monofilaments “esthesiometer” of the brand SORRI®, which are composed of six nylon filaments of equal length of different colors and various diameters that produce a standardized pressure on the skin surface. The monofilaments have the purpose of evaluating and quantifying the tactile perception threshold and sensation of deep foot pressure [11]. The classification of the filaments is based on their colors, as follows: green color (0.05 gf) and blue color (0.2 gf): normal sensitivity; violet color (2.0 gf): difficulty with shape and temperature discrimination; dark red color (4.0 gf): mild loss of protective sensation, vulnerable to injury; orange color (10.0 gf): mild loss of protective sensation; magenta color (300.0 gf): loss of protective sensation; no response: total loss of sensitivity. The monofilaments were applied to 10 different points on each foot, as predefined by Armstrong et al. [ 12], consisting of the plantar region (PR) of the 1st toe; PR of the 3rd toe; PR of the 5th toe; PR of the 1st metatarsal; PR of the 3rd metatarsal; PR of the 5th metatarsal; the medial region (MR) of the plantar surface of the foot; the medial-lateral region (MLR) of the plantar surface of the foot; calcaneus; and the interphalangeal region (IR) between the 1st and 2nd toe. The evaluation protocol followed the instructions in the user’s manual of the manufacturer of the product “SORRI® Esthesiometer”, as well as other studies [13,14] (Figure 1). Before starting the procedure, a test was performed with the monofilament, which was applied to an area of the participants’ arm with preserved sensitivity so that the correct understanding of the test could be verified. The participants were positioned on a stretcher in a supine position, eyes closed, and in a quiet environment. Each monofilament was applied perpendicularly for about 1–2 s at each point so as to curve over the area without sliding over the skin of the elderly woman. The tests started with the thinnest and lowest pressure monofilament (0.05 gf, green color), and in case of no response, a monofilament of larger diameter and pressure (0.2 gf, blue color) was used, and so on until the participant was able to detect the touch. Each monofilament was pressed onto the skin, and the participant was instructed to indicate the time and place when she felt the pressure of the filament. The application was repeated twice on the same site and alternated with at least one simulated application in which the monofilament was not applied. This way, three questions were asked per application site and were considered an absent sensation if two answers were incorrect in the three attempts. It is noteworthy that the elderly were strictly monitored over the 8 weeks, noting their non-participation in activities that could influence the study. The recording of the tests was made by marking at each established point the color corresponding to the first monofilament that the participant detected by touch. To allow a comparison between the situations, a numerical score was stipulated for each monofilament that ranged from 0 (zero) no perception to 6 (six) normal sensitivity; that is, the higher the score, the better the plantar tactile sensitivity. The sensitivity was determined by regions of the right and left feet: the forefoot (the sum of the points of seven regions), the midfoot (the sum of the points of two regions), the hindfoot (the score of one region), and the whole foot (the sum of all points assessed). The evaluations of the variables were carried out in two moments: before training (T0) and after training (T1), by researchers who did not participate in the allocation process of older women and had no contact with the treatment groups. For the CGT, the participants were evaluated and reassessed following the same period and place established for the intervention groups. ## 2.5. Data Analysis To evaluate the homogeneous behavior of quantitative variables (age and BMI) at the baseline in the three groups (control, conventional, and exergame), the analysis of variance (ANOVA) and Kruskal–Wallis tests were used after checking the normality of the data using the Shapiro–Wilk test. Pearson’s chi-square test and Fischer’s exact test were used to comparing the categorical variables (marital status, education, family income, presence of diseases, pain in the last 7 days and 12 months, and medications) between the groups at the beginning of the study. In the inferential analysis (parametric or nonparametric) for the comparisons of plantar tactile sensitivity variables, the Shapiro–Wilk test was initially used to test the normality of the data. Intragroup comparisons between the two paired samples were performed using paired Student’s t-test or Wilcoxon test. Intergroup comparisons between three independent samples were performed using the Kruskal–Wallis test, and, in the case of statistical difference, Dunn’s post hoc test was used. The effect size was calculated for between-group comparisons (i.e., comparisons of differences between T0 and T1) using the partial eta2 parameter (partia leta squared, η2 partial) as an effect size indicator. The interpretation of the effect size adopted a small effect size when η2 = 0.01, a medium effect size when η2 = 0.06, and a large effect size when η2 = 0.14. The significance level adopted in all analyses was $5\%$ (α = 0.05), and the data were analyzed in an IBM Statistical Package for the Social Sciences (SPSS) for Windows, version 21.0. ## 3. Results The analysis of plantar tactile sensitivity, specifically in the forefoot, midfoot, hindfoot, and right and left whole foot regions in the control, conventional, and exergame groups at T0 showed that the variables did not show significant differences between the groups, indicating that the three groups had similar characteristics at the baseline of the study (Table 1). The comparisons between T0 and T1 in the control group showed a significant difference in the tactile sensitivity of the hindfoot R, midfoot L, hindfoot L, and whole foot L, indicating that, at the end of the period evaluated, the older women in this group presented significantly lower values in these regions, which characterized a worsening of the plantar tactile sensitivity (Table 2). The comparisons between T0 and T1 in the conventional group showed a significant difference in all variables, except for the tactile sensitivity of the forefoot R, indicating that, at the end of the treatment, the older women in this group presented significantly higher values, characterizing an improvement in the plantar tactile sensitivity of both feet, especially in the left foot (Table 3). The comparisons between T0 and T1 in the exergame group showed a difference in the tactile sensitivity of the forefoot R, whole foot R, forefoot L, and whole foot L, indicating that, at the end of the treatment, the older women in this group presented significantly higher values in these regions, which characterized an improvement in the plantar tactile sensitivity of both feet (Table 4). The comparative analysis of the changes in the plantar tactile sensitivity variables showed significant differences between the groups. For the tactile sensitivity of the hindfoot R, whole foot R, forefoot L, hindfoot L, and whole foot L, the results showed a better effect of conventional and exergame training when compared to the control group. Regarding the sensitivity of the forefoot R, there was a better effect in the exergame training compared to the control group. As for the sensitivity of the midfoot R and L, there was a better effect in conventional training compared to the control group (Table 5). Among all the variables studied, no significant differences were observed between the conventional and exergame groups, indicating a similar effect of the two types of training on the plantar tactile sensitivity of both feet. Regarding the effect size, the results indicate an effect that is classified as large for all the variables studied (0.141–0.295) (Table 5). ## 4. Discussion The results of the present study demonstrated that the older women from the control group, i.e., those who did not undergo any kind of intervention during the studied period, presented a worsening of the plantar tactile sensitivity in both feet, with predominance in the left foot. The impairment of the plantar tactile sensitivity that was observed in the older women of the control group may have made them more susceptible to suffering falls or to present difficulties in locomotion on uneven surfaces so that, according to Cenci et al. [ 15], the decrease in plantar sensitivity was one of the main factors that collaborated to the decrease in afferences to the motor control system, thus generating a decrease in balance, impairment of gaits, such as smaller cadence, shorter steps, and less acceleration, slowness in the correction of motor errors and obstacle transposition. In relation to the older women submitted to proprioceptive conventional and games with motion monitoring training, an improvement in the plantar tactile sensitivity of the right and left feet was evidenced in both groups. The improvement of this sensitivity can be attributed to the multisensory stimuli provided by proprioceptive training since studies have shown that physical exercises that promote stimuli and vary in texture, weight, and shape, whether associated or not with sound and visual stimuli, improve the blood supply to the lower limbs, thus contributing to the reduction in endoneural hypoxia and an improvement in nerve conduction [8,14,15]. Santos et al. [ 16] investigated the effect of conventional proprioceptive training on plantar tactile sensitivity in sedentary women. The study used a conventional proprioceptive protocol that was similar to the one developed in the present study, which involved gait, balance, and proprioception training in order to provide sensory stimulation on the plantar surface. The results showed significant improvement in plantar sensitivity after 24 intervention sessions. However, it is worth mentioning that most of the participants were younger than 60 years old, ranging from 50 to 70 years old; and that the results were not compared to other types of training, which made it impossible to generalize and compare the effects with another intervention. Studies have concluded that orthostasis on textured surfaces with varying densities caused an increase in peripheral nerve activity for healthy individuals as a function of changes in the transmission of afferent signals from the sole of the foot [13,14,15,16]. In the present study, the conventional training program adopted used techniques and resources similar to those addressed by the aforementioned authors. This variety of techniques and resources may have contributed to the stimulation of different areas of the foot innervated by the deep peroneal, sural, saphenous, and tibial nerves as well as to the activation of a greater quantity of exteroceptors, which provided an improvement of tactile sensitivity in the distinct plantar regions of older women who underwent this type of training. Regarding training with games and motion monitoring (GMM), although no studies have been found on its effects on plantar tactile sensitivity, the present study showed that this training modality contributed to an improvement in sensitivity. This resource consists of the reproduction of tasks that can be performed by the individual in interaction with a multidimensional and multisensory environment created by a computer that can be explored in real-time, thus contributing to sensory stimulation [17,18]. One possible explanation for the improvement in plantar tactile sensitivity from VR is that this type of training is able to provide tactile feedback to the central nervous system. The additional sensory stimulation of plantar cutaneous receptors improves tactile sensitivity, favoring the performance of activities and preventing risks of accidents since the perception of movement is favored when tactile feedback is available [18]. According to some authors, training based on VR allows for a greater number of repetitions, high variability of movements, and auditory and visual feedback [18,19,20]. All this range of activities provided by training with VR ends up exerting a positive influence on plantar tactile sensitivity. However, unlike the conventional training, which was developed on unstable surfaces and with different textures and densities, the VR was performed only on the ground, that is, on a stable surface and without different textures and densities, which may not have favored the stimulation and, consequently, the improvement of tactile sensitivity in all specific regions of the feet of this group of older women. The results of the present study showed that when the three groups of older women were compared, the plantar tactile perception presented significant differences when analyzing the before and after interventions, in which the conventional and exergame groups obtained better effects when compared to the control group. However, no significant differences were observed in plantar tactile sensitivity between the conventional and exergame groups. Based on these results, it is possible to state that the improvement of plantar tactile sensitivity in older women could be achieved with a physiotherapeutic intervention of easy access and low cost, as in the case of conventional training. The proposal of training, with games and motion monitoring, must be considered a great advance in the health area, since in the globalized world, participation in the technological process is inevitable, and the use of these advances by the health sciences is undeniable. Although VR is a resource that has higher costs than conventional physical therapy, it has shown positive results in several clinical areas. Similar to conventional training, VR is a therapeutic modality that can be performed in physical therapy clinics, in long-stay institutions for the elderly, or even at home, facilitating the treatment of people who do not have access to a rehabilitation center. In view of all the findings, both types of training must complement each other, always seeking the physical and emotional well-being of the elderly population. It is important to emphasize the need for further studies that can describe more characteristics and factors that are related to the effects of different exercise modalities on the sensory and functional responses of older women since this was one of the limitations of the present study. Another limitation of the present study was that a follow-up was not carried out to verify how long the effects of the intervention lasted after its completion. Thus, further studies are suggested, given the importance of gaining a greater understanding of this subject for possible reference measures and an improvement in the quality of life of this population. ## 5. Conclusions Based on the results of the present study, it is suggested that the proposed proprioceptive training with conventional games and motion monitoring may be favorable for the improvement of plantar tactile sensitivity in older women. When comparing the intergroup results, there was a better effect of the intervention groups when compared to the control group; however, this was without significant differences between conventional and virtual training regarding the outcomes studied. ## References 1. Prieto L.M.G., Galindo L.F.H., Deses C.D.L.. Prevalencia de patología del pie en una población geriátrica y su impacto en la función, la marcha y el síndrome de caídas. Rev. Esp. Geriatr. Gerontol.* (2012.0) 47 19-22. DOI: 10.1016/j.regg.2011.05.004 2. Rosa W.G.N., Navarro R.L., Conti A.C.C.F., Almeida M.R., Oltramari-Navarro P.V.P.. Assessment of cephalometric characteristics in the elderly. Braz. Oral Res. (2015.0) 29 1-9. DOI: 10.1590/1807-3107BOR-2015.vol29.0040 3. Stolt M., Suhonen R., Puukk A.P., Viitanen M., Voutilainen P., Leino-Kilpi H.. Foot health and selfcare activities of older people in home care. J. Clin. Nurs. (2012.0) 21 3082-3095. DOI: 10.1111/j.1365-2702.2012.04223.x 4. Lynch E.A., Hillier S.L., Stiller K., Campanella R.R., Fisher P.H.. Sensory retraining of the lower limb after acute stroke: A randomized controlled pilot trial. Arch. Phys. Med. Rehabil. (2007.0) 88 1101-1107. DOI: 10.1016/j.apmr.2007.06.010 5. Bretan O., Pinheiro R.M., Corrente J.E.. Functional evaluation of balance and plantar cutaneous sensitivity in community-dwelling elderly. Braz. J. Otorhinolaryngol. (2010.0) 2 219-224. DOI: 10.1590/S1808-86942010000200012 6. Pinto M.J., Freitas E.V., Py L.. The feet of the elderly and their repercussions on quality of life. Tratado de Geriatria e Gerontologia 1169-1179 7. Silva R.E.G., Morais A.C.S., Godoi C.D., Garcia O.A.G., Neto W.N.R.. Motor sensory evaluation and its correlation with the quality of life in carriers of diabetes mellitus. S. Am. J. Basic Educ. Tech. Technol. (2017.0) 4 118-132 8. Gomes A.A., Sartor C.D., João S.M.A., Sacco I.C.N., Bernik M.M.S.. Effects of physical therapy intervention on sensory and functional responses in neuropathic diabetics. Fisioter. Pesqui. (2007.0) 14 14-21 9. Santos A.A., Bertato F.T., Montebelo M.I.L., Guirro E.C.O.. Effect of proprioceptive training among women. Rev. Bras. Fisioter. (2008.0) 12 183-187 10. Alfieri F.M.. Plantar pressure distribution in elderly after proprioceptive intervention. Rev. Bras. Cineantropometria Desempenho Hum. (2008.0) 10 137-142 11. Khoshelham K., Elberink S.O.. Accuracy and resolution of kinect depth data for indoor mapping applications. Sensors (2012.0) 12 1437-1454. DOI: 10.3390/s120201437 12. Armstrong D.G., Lavery L.A., Vela S.A., Quebedeaux T.L., Fleischli J.G.. Chosing a practical screening instrument to identify patients at risk for diabetic foot ulceration. Arch. Intern. Med. (1998.0) 158 289-292. DOI: 10.1001/archinte.158.3.289 13. Sales K.L.S., Souza L.A., Cardoso V.S.. Static balance in individuals with diabetic peripheral neuropathy. Fisioter. Pesq. (2012.0) 19 122-127. DOI: 10.1590/S1809-29502012000200006 14. Machado A.S., Silva C.B.P., Rocha E.S., Carpes F.P.. Effects of plantar sensitivity manipulation on upright posture control in young and elderly adults. Rev. Bras. Reumatol. (2017.0) 57 30-36. DOI: 10.1016/j.rbr.2015.11.005 15. Cenci D.R., da Silva M.D., Gomes É.B., Pinheiro H.A.. Analysis of balance in diabetic patients using the F-Scan system and the Berg Balance Scale. Fisioter. Mov. (2013.0) 26 55-61. DOI: 10.1590/S0103-51502013000100006 16. Santos A.A.S., Gontijo L.B., Oliveira F., Bastos V.H.V., Machado T.P.G., Santos A.P.. Effects of a sensory reeducation program in type 2 diabetics. Rev. Neurosci. (2015.0) 23 499-505. DOI: 10.4181/RNC.2015.23.04.1063.07p 17. Ricci N.A., Gazzola J.M., Coimbra I.B.. Sensory systems in body balance in the elderly. Arq. Bras. Ciên Saúde (2009.0) 34 94-100 18. Mendes F.A.S., Arduini L., Botelho A., Cruz M.B., Santos-Couto-Paz C.C., Pompeu S.M.A.A., Piemonte M.E.P., Pompeu J.E.. Patients with Parkinson’s disease are able to improve their performance in Xbox Kinect. Motricity (2015.0) 11 68-80. DOI: 10.6063/motricidade.3805 19. Torriani-Pasin C., de O. Mota E.P., Moreira Sales A.L., Ricci M., Nishida P., Marques L., Lima L.F., Nogueira M., Lopes de Souza V.M., Zancani de Lima R.. Effect of foot motor and sensory stimulation hemiparetic in stroke patients. Rev. Neurosci. (2008.0) 16 25-29 20. Pompeu J.E., Mendes F.A., Silva K.G., Lobo A.M., Oliveira T.P., Zomignani A.P., Piemonte M.E.P.. Effect of Nintendo Wii™-based motor and cognitive training on activities of daily living in patients with Parkinson’s disease: A randomised clinical trial. Physiotherapy (2012.0) 98 196-204. DOI: 10.1016/j.physio.2012.06.004
--- title: Establishing Co-Culture Blood–Brain Barrier Models for Different Neurodegeneration Conditions to Understand Its Effect on BBB Integrity authors: - Jun Sung Park - Kyonghwan Choe - Amjad Khan - Myeung Hoon Jo - Hyun Young Park - Min Hwa Kang - Tae Ju Park - Myeong Ok Kim journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049378 doi: 10.3390/ijms24065283 license: CC BY 4.0 --- # Establishing Co-Culture Blood–Brain Barrier Models for Different Neurodegeneration Conditions to Understand Its Effect on BBB Integrity ## Abstract The blood–brain barrier (BBB) is a functional interface that provides selective permeability, protection from toxic substances, transport of nutrients, and clearance of brain metabolites. Additionally, BBB disruption has been shown to play a role in many neurodegenerative conditions and diseases. Therefore, the aim of this study was to establish a functional, convenient, and efficient in vitro co-cultured BBB model that can be used for several physiological conditions related to BBB disruption. Mouse brain-derived endothelial (bEnd.3) and astrocyte (C8-D1A) cells were co-cultured on transwell membranes to establish an intact and functional in vitro model. The co-cultured model and its effects on different neurological diseases and stress conditions, including Alzheimer’s disease (AD), neuroinflammation, and obesity, have been examined by transendothelial electrical resistance (TEER), fluorescein isothiocyanate (FITC) dextran, and tight junction protein analyses. Scanning electron microscope images showed evidence of astrocyte end-feet processes passing through the membrane of the transwell. Moreover, the co-cultured model showed effective barrier properties in the TEER, FITC, and solvent persistence and leakage tests when compared to the mono-cultured model. Additionally, the immunoblot results showed that the expression of tight junction proteins such as zonula occludens-1 (ZO-1), claudin-5, and occludin-1 was enhanced in the co-culture. Lastly, under disease conditions, the BBB structural and functional integrity was decreased. The present study demonstrated that the co-cultured in vitro model mimicked the BBB’s structural and functional integrity and, under disease conditions, the co-cultured model showed similar BBB damages. Therefore, the present in vitro BBB model can be used as a convenient and efficient experimental tool to investigate a wide range of BBB-related pathological and physiological studies. ## 1. Introduction Homeostasis of the brain is maintained by a specialized structure called the blood–brain barrier (BBB) [1]. Along with the homeostasis of the brain, the BBB performs several essential and specialized functions, as acting as a cellular barrier and promoting simple diffusion in carrier-mediated transport, receptor-mediated transcytosis, active efflux through ATP-binding cassette (ABC) transporters, and the clearance of neurotoxic substances [2]. Recent evidence has suggested that BBB disruption is an early biomarker for cognitive decline and dementia-associated diseases [3]. The BBB is a complex three-dimensional structure comprised of brain endothelial cells, astrocytes, and pericytes [4]. Prolonged astrocyte structures called astrocytic end-feet cover endothelial cells to maintain the BBB [5]. The barrier nature of this complex is maintained by specialized brain endothelial cells, which form a compact junctional complex including tight junctions and adherens junctions [6]. Tight junction structures include transmembrane proteins (occludin, claudins and junctional adhesion molecules), actin filaments, and cytoplasmic scaffold proteins (zonula occluden (ZO)) [7]. Studies have shown that tight junction-related BBB dysfunction is associated with Alzheimer’s disease [8,9]. The expression of tight junction proteins during cell-to-cell contact leads to the formation of a compact barrier-like property. This results in enhanced transendothelial electrical resistance (TEER) and low transcellular and paracellular permeability, which restricts the lateral diffusion of membrane proteins, but allows transport through a specialized transport system [10,11,12]. Normal functioning of the BBB is essential for the proper and regulated functioning of the central nervous system (CNS). However, any disruption or dysregulation in the BBB can lead to severe neurological disorders. Several studies have shown that the disruption of the BBB is involved in the onset of Alzheimer’s disease (AD) [13,14,15]. Moreover, neuroinflammation has shown to damage the BBB and cause further development in AD [16]. Additionally, several peripheral metabolic disorders, such as hyperglycemia, diabetes, and cardiovascular disease, are known to be responsible for the disruption of the BBB. Peripheral metabolic disorders also challenge the integrity of the BBB [17]. A high-fat diet is a significant risk factor for the development of metabolic-related disorders such as hyperglycemia, diabetes, and cardiovascular disease and neuroinflammation [18,19,20]. Palmitic acid (PA) is the most common form of saturated fatty acid in nature, including the human body, and is rich in high-fat diets [21]. Additionally, these metabolic-related disorders and neuroinflammation are risk factors for developing dementia, including Alzheimer’s disease (AD) [22,23,24]. Furthermore, studies have also shown a direct association between PA and AD [25,26,27]. Due to the crucial role of the BBB in neurological disease and pharmacological research, a BBB model is an essential component. Currently, several BBB models have been developed, ranging from mono- and co-cultured models to sophisticated models, such as spheroid and chip models that use animal and human cell types [28,29,30,31,32]. Sophisticated models represent the BBB structure using the 3D organization of cells and mimic cerebral blood flow [31,32]. Despite these advantages, they are difficult to set up and, most importantly, expensive. Therefore, a co-cultured transwell system provides a simpler, cost-effective, and functional BBB model. Therefore, the present study describes a contact co-culture in vitro model that uses mouse brain endothelial and astrocyte cell lines in transwells to provide an efficient and functional platform for various neurological research scenarios. ## 2.1. Endothelial and Astrocyte Co-Culture Model of the BBB Shows Efficient Inter-Cellular Communication by End-Feet Processes The brain endothelial cell line bEnd.3 and the astrocyte cell line C8-D1A were co-cultured on the transwells. The brain endothelial cells were cultured on the apical side, while C8-D1A cells were seeded on the basolateral side of the transwell during contact co-culture (Figure 1A–C). Cells that adhered to opposite sides of the membrane showed intercellular communication via protruded end-feet processes through the micropores of the membrane. Scanning electron microscope images showed evidence of the astrocyte end-feet processes passing through the membrane of the transwell (Figure 1D,F). This property mimics the actual BBB, where the astrocyte end-feet comes in contact with the endothelial cells (Figure 1F). ## 2.2. Co-Cultured BBB Model Showed Effective Barrier Function and Property After establishing a co-cultured BBB model, we examined its barrier function and property (Figure 2). First, TEER values for both mono- and co-cultured C8-D1A and bEnd.3 cells were analyzed (Figure 2A and Figure S1A). The TEER values depict a functionally compact layer of cells that resist permeability. The co-cultured (C8-D1A and bEnd.3) transwell showed a greater TEER value than the individual cell cultured (C8-D1A or bEnd.3) transwells, which showed peak values from 5 to 8 days post-seeding. Additionally, the TEER value increased with time due to the confluency of cells, indicating that the co-culture transwell showed maximum resistance compared to the cells cultured individually. Hence, the present co-culture in vitro model replicates the quality of an intact BBB. Furthermore, it has been reported that the BBB is impermeable to molecules larger than 400 Da [33]. To check the integrity of the co-culture BBB model, an FITC-dextran 70 permeability assay was performed. The FITC fluorescence values showed that the co-cultured model resisted the penetration of FITC dextran better than the individual cell cultured models (Figure 2B and Figure S1B). Next, the structural aspect of the present contact co-culture model was analyzed by solvent persistence and solvent leakage tests. The results showed that the co-cultured transwells had the highest solvent persistence and lowest leakage values compared to the individual cell cultured models (Figure 2C,D and Figure S1C,D). ## 2.3. The Co-Cultured Model Showed Enhanced Cell–Cell Adhesions The expression of tight junction proteins such as ZO-1, occludin-1, and caludin-5 were analyzed in both mono- (without C8-D1A) and co- (with C8-D1A) cultured models to assess the cell–cell adhesions. The Western blot analysis showed that all the aforementioned tight junction protein expressions were 1.5–2-fold higher in the co-cultured model than in the mono-cultured model (Figure 3A). Additionally, the fluorescence immunohistochemistry analysis validated this finding by showing enhanced ZO-1 expression in the co-cultured model compared to the mono-cultured models (Figure 3B). ## 2.4. Exposure to Palmitic Acid Mimics Obesity and Distresses the BBB Tight Junction The in vitro obesity model was established by exposing PA-BSA in the co-cultured model for 24 h. The immunofluorescence results showed that there was a significant decline in ZO-1 in the PA-BSA-exposed co-culture model (Figure 4A). In the Western blot analysis, tight junction proteins, including ZO-1, occludin-1, and claudin-5, also showed a significant decline in the expression levels among PA-BSA-treated cells (Figure 4B). Moreover, the deteriorating effect of PA-BSA on the BBB’s integrity was analyzed by measuring the electrical resistance across the BBB. The TEER analysis further validated that PA-BSA exposure significantly deteriorated the electrical resistance of the BBB, thus indicating a weakened BBB structure (Figure 4C). ## 2.5. Neuroinflammation Affects the Decline in BBB Permeability Neuroinflammation includes the microglial activation responsible for the secretion of interleukins and cytokines. To mimic this condition, our co-cultured model was exposed to microglial conditioned media (MCM), which is enriched with pro-inflammatory interleukins and cytokines. After 24 h of exposure, the immunofluorescence analysis showed that ZO-1 expression declined and glial fibrillary acidic protein (GFAP), a marker of astroglial injury, expression significantly increased in the MCM-treated group, as compared to the control group (Figure 5A,B). We validated this finding through western blot analysis and showed that the expression of tight junction proteins after MCM exposure significantly declined, while GFAP expression increased compared to the non-exposed control group (Figure 5C,D). Lastly, BBB integrity was analyzed using TEER analysis and showed a rapid decline in barrier capability after exposure to MCM (Figure 5E). ## 2.6. Amyloid Beta Pathology Leads to BBB Disruption An in vitro model for AD-like pathology with three co-cultured cell lines was created with bEnd.3 cells that were seeded on the apical side of the transwell, C8-D1A cells that were seeded on the basolateral side of the same transwell, and HT22 cells that were seeded in the wells of the plate. HT22 cells were transfected with the amyloid precursor protein overexpression plasmid pCAX APPswe/ind or exposed to Aβ (1–42) peptides dissolved in culture media for TEER analysis to detect the role of Aβ toxicity. The ZO-1 expression level in the mono-cultured model showed a significant decrease in the number of cells exposed to Aβ compared to the control (Figure 6A). Furthermore, tight junction proteins in the mono-cultured model revealed that ZO-1, claudin-5, and occludin-1 protein expression levels were significantly down-regulated (Figure 6B). In the co-cultured model exposed to Aβ, the TEER analysis also showed a rapid decline in barrier capability (Figure 6C) and the fluorescent immunohistochemistry results demonstrated the same finding that the tight junction protein expression significantly declined compared to the control (Figure 6D). These findings were further confirmed by analyzing the in vivo AD model established with a stereotaxic injection of Aβ1–42 in C57BL/6N mice. The Western blot analysis of the frontal cortex and hippocampal region of Aβ-injected mice showed that the ZO-1 protein expression level was comparably lower than the control (Figure 6E). Aβ co-localized in specific brain regions, which confirmed Aβ’s toxicity, together with the tight junction protein expression levels observed in the Aβ-induced AD model. ## 3. Discussion The aim of the present study was to establish and analyze an in vitro BBB model and investigate its application in various BBB dysregulated disorders, such as diabetes, TBI, and AD. In this study, we developed a transwell co-culture model that combined brain endothelial cells (bEnd.3) and astrocytes (C8-D1A). Overall, our co-cultured model, compared to the endothelial-only cell model (mono-culture), showed improved BBB structural and functional integrity, as well as higher tight junction protein levels. Additionally, our co-cultured model demonstrated decreased integrity and tight junction protein levels once exposed to the aforementioned disease environment. ## 3.1. Co-Cultured In Vitro Model Mimics the Structural and Functional Integrity of the BBB TEER is one of the most accurate non-invasive measures of BBB integrity during various cell growth and differentiation stages. The measurement of electrical resistance represents a quantitative analysis method of barrier integrity and function. This value depends on various important factors, such as temperature, cell passage number, the composition of cell media, and the mechano-electronics used for TEER calculation [34,35]. The TEER values of our co-culture were similar to those reported in a previous study using the same cell line and transwell setup [36]. Additionally, the peak value persisted for 4–5 days before declining, which indicates that this period is optimal for mechanistic studies of the BBB and its role in different diseases. Moreover, this model can also provide an opportunity to detect the permeability of neuro-therapeutic drugs and related molecules. Moreover, molecules that are not lipid-soluble are not able to pass through the BBB without carrier-mediated transport assistance [6]. Therefore, FITC-dextran is a suitable assay to estimate the passive permeability efficiency of an in vitro BBB model [37]. Our results showed that the co-cultured model provided a more compact and intact barrier than the mono-culture model, which was also shown in another study [38]. This may be due to the fact that C8-D1A astrocyte type 1 cells provided enhanced support to bEnd.3 cells in their building barrier property. Studies have shown that astrocytic end-feet processes adhere to endothelial cells to provide additional physiological support to the BBB and help in cell-to-cell communications, provide structural support to the BBB and enhance the expression of tight junction proteins by intercellular communication [39,40]. Furthermore, a compact monolayer of brain endothelial cells is responsible for the barrier properties of the BBB [41]. Additionally, tight junction proteins are involved in the strong cell-to-cell adhesion in this monolayer and control the integrity and barrier function of the BBB [6,42,43]. However, based on our findings, the co-cultured models showed higher tight junction protein levels compared to the endothelial monolayer, which was also shown in another study [38]. Studies have shown that astrocytes helps to maintain BBB integrity, providing structural support and promote the expression of ZO-1 and occludin [44]. ZO-1 strengthens cell-to-cell adhesion, which controls transcellular and paracellular permeability. The decrease in ZO-1 expression compromises BBB integrity and function and causes leakage of the BBB, which allows the unchecked transport of molecules during disease conditions [45,46,47]. ## 3.2. Co-Cultured In Vitro Model Is Ideal to Investigate BBB-Related Disorders The BBB, as mentioned previously, prevents toxic substances in the blood from crossing into the CNS and filters toxic compounds from the brain [48]. Studies have shown that damages to the BBB play a role in the pathogenesis of several disorders, such as obesity, neuroinflammation, and AD [49,50,51]. In order to investigate the role of BBB integrity in obesity due to a high-fat diet, our co-cultured model was exposed to PA-BSA, which showed decreased tight junction protein levels and TEER resistance. In vivo studies have also shown similar findings in which animals fed a high-fat diet showed decreased tight junction proteins levels in the brain [52,53]. Previously, our group has shown that LPS-induced BV-2 cells show increased levels of allograft inflammatory factor 1 (Iba-1), a marker for microglia, and pro-inflammatory cytokines, including tumor necrosis factor alpha (TNFα), and interleukin 1 beta (IL-1β) [54]. Therefore, this explains the increase in GFAP and validates the presence of pro-inflammatory mediators in the MCM. Furthermore, the traumatic brain injury (TBI) inflammatory response in the lesion site is mainly due to the activation of cytokines and studies have shown that tight junction proteins such as claudin-5, occludin, and ZO-1 decrease after the first few days of TBI [55,56]. Lastly, amyloidosis, which is associated with intracellular and extracellular insoluble aggregates that lead to fiber or plaque aggregation, is a hallmark of many neurodegenerative pathologies, including AD, Huntington’s disease, Parkinson’s disease, etc. [ 57]. Thus, we mimicked the condition of the disease by treating the cells co-cultured in transwell membranes exposed to Aβ (1–42) or by the overexpression of pCAX APPswe/ind plasmids in neurons, both of which lead to Aβ toxicity [58]. Although it is important to note that we measured the total tight junction protein expression in the frontal cortex and hippocampus region, the findings were in line with several in vivo studies, in which the studies showed decreased tight junction protein levels [59,60,61,62]. Moreover, decreased tight junction protein levels have also been shown in postmortem studies of patients with AD [8,9,61]. Therefore, our co-cultured in vitro BBB model validated the structural and functional integrity BBB damage that occurs in the aforementioned disorders. Although microvascular endothelial cells line the cerebral capillaries in the BBB, astrocytes are essential cells in the CNS that secrete substances such as neurotransmitters, neuromodulators, hormones and peptides, and growth and inflammatory factors, which can enhance or deteriorate the endothelial cells [63]. As such, pathology-associated inducers affect astrocyte activity and influence the structure of the BBB. For example, a study has shown that in mice, a high-fat diet led to obesity, systemic insulin resistance, dysregulated lipid metabolism, and depressive-like behavior. Additionally, it increased the expression of GFAP, shortened the processes of GFAP+ cells, and downregulated the expression of astrocytic neuroplasticity-related proteins, GLAST, GLT-1, and connexin-43 in the hippocampus [64]. Moreover, in another study, mice fed a high-fat diet showed increased production of reactive oxygen species (ROS) and pro-inflammatory and endothelial markers, e.g., TNFα, IL-1β, and vascular cell adhesion molecular 1 (VCAM-1). Furthermore, the mice demonstrated a decrease in the levels of claudin-5 and collagen IV, a type of collagen in the basal lamina, compared to the control mice [65]. Lastly, our group has shown, in multiple studies, that intracerebroventricular (i.c.v) injections of Aβ1–42 increased GFAP, inflammatory markers, e.g., nuclear factor-κB (NFκB) TNFα, IL-1β, and oxidative stress markers, e.g., nuclear factor erythroid-related factor 2 (Nrf2) and heme oxygenase 1 (HO1), which indicated that AD-associated pathology also increase astrocyte activity and promote inflammation [66,67,68,69]. Therefore, this may explain the decreased integrity of the tight junction dynamics in our transwell co-cultured model. Lastly, in addition to the aforementioned simple, inexpensive, and functionality of transwell cultures, the advantages of our transwell co-cultured was its wide application in pathology-induced neurodegenerative disease models, as shown in this study. However, it is important to mention that human cells demonstrate differences in morphology and function compared to non-human cells. Additionally, despite a small pore size of 0.4 μm, uncovered pores could provide diffuse penetration; thus, further tests are required to examine its use for novel brain-targeted drug candidates. Nevertheless, our in vitro model could be used prior to investigating human cell BBB models for neurodegenerative diseases. ## 4.1. Cell Culture The mouse brain endothelial cell line (bEnd.3) was purchased from American Type Culture Collection (No. CRL-2299, ATCC, Manassas, VA, USA). After cell revival, the cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM; GIBCO, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with $10\%$ fetal bovine serum (FBS) and $1\%$ penicillin-streptomycin (10,000 U/mL) of the final volume. Astrocyte type I clone (C8-D1A; No. CRL-2541, ATCC) cells were grown in DMEM with an additional $10\%$ FBS and $1\%$ penicillin-streptomycin (10,000 U/mL) of the final volume. The cells were incubated at 37 °C and $5\%$ carbon dioxide. Microglial-conditioned media was produced according to the previously mentioned protocol with some modifications [70]. Briefly, the mouse microglial cell line BV-2 was cultured to above $75\%$ confluency and was treated with lipopolysaccharides from Escherichia coli O111:B4 (50 ng/mL; Sigma-Aldrich, L2630, St. Louis, MO, USA) dissolved in cell culture media. After 24 h, the media were aspirated and centrifuged to remove cells and debris. The supernatant was collected for further use. ## 4.2. Cell Seeding in Transwells To establish a co-cultured model, bEnd.3 cells were seeded on the apical side of the transwell and C8-D1A cells were seeded on the basolateral side of the transwell. The transwell was placed in an inverted orientation to bring the basolateral side of the transwell upward for astrocyte cell seeding. The cells were allowed to adhere to the lower surface of the transwell for almost 48 h (Figure 1A). The whole procedure was performed in an aseptic environment to avoid contamination during the cell culturing procedures. Lastly, AD-like pathology with three cell line co-cultured models was established in a similar way with the addition of HT22 cells to be seeded on the wells of the plate. For the Western blot and immunofluorescence analyses, the cells were seeded with some modifications compared to the cells used for the permeability and TEER analyses. In this case, a 100 mm dish with permeable transwell support (Costar Corning, Kennebunk, ME, USA) was used. bEnd.3 cells were seeded on the transwell, while C8-D1A cells were seeded in the dish supplemented with 25 mm cell culture-treated coverslips (SPL life Sciences, Republic of Korea) to detect protein expression by immunofluorescence. ## 4.3. Bovine Serum Albumin-Conjugated Palmitic Acid (PA-BSA) Preparation Bovine serum albumin-conjugated palmitic acid (PA-BSA) was prepared based on a previous study [25] with some modifications. Briefly, 500 mM PA was dissolved in 1 mL of absolute ethanol with constant shaking at 65–70 °C, until completely dissolved. While $10\%$ BSA solution was prepared by dissolving 1.5 g of FFA-Free BSA in serum-free DMEM media at 37 °C, both the solutions were syringe-filtered under aseptic conditions to remove any undissolved particles. In addition, 10 µL of PA was dissolved in 1 mL of BSA at 55 °C, followed by water sonication to form a soluble PA-BSA and stored at −20 °C. The PA-BSA was incubated at 37 °C in a water bath prior to use. ## 4.4. TEER Measurement TEER was measured periodically to monitor cell confluence and the development of tight junctions. For the measurement of TEER, voltage and current electrode wires were connected via an electrode adaptor (WPI) to an EVOM2 epithelial voltammeter (Millipore, Burlington, MA, USA). The EVOM2 was adjusted to a 10 mA AC current at 12.5 Hz while measuring resistance. Background resistances (RBLANK) were subtracted from the total calculated resistance RTOTAL at each time point and normalized for the area, providing TEER values in Ω.·cm2 as in the following equation [12]:RTISSUE (Ω) = RTOTAL − RBLANK TEERREPORTED = RTISSUE (Ω) × MAREA (cm2) Each group, including an empty transwell for background resistance (RBLANK), bEnd.3 monoculture, C8-D1A monoculture, or both bEnd.3 and C8-D1A co-cultures under disease conditions, was processed and analyzed in triplicate according to the experimental condition. ## 4.5. Permeability Assays Fluorescein isothiocyanate (FITC) permeability assay was performed using 1 mg/mL of FITC dextran 70 (FITC:Glucose = 1:250, average molecular weight 70,000; Sigma-Aldrich) dissolved in Hanks’ balanced salt solution (HBSS). A total of 800 µL of the FITC dextran solution was added to the apical side of the transwell, while the basolateral side was supplemented with HBSS buffer. Every 6 h, 100 µL of the sample solution from the basolateral side of the transwell was collected [34]. The amount of FITC that penetrated the transwell membrane in the presence or absence of the cells was measured by fluorescence detection using a GLOMAX multi-detection system (Promega, Madison, WI, USA). Fluorescence was measured at 485 nm absorbance and 510 nm emission. Solvent persistence and leakage were analyzed by adding 800 µL of the cell culture media on the apical side of the transwell. Prior to the experiment, the mono- and co-cultured models were incubated in the media and during the experiment, the transwell was moved to a new culture dish and incubated in the incubator to examine the solvent persistence and leakage. A lid covered the dish to maintain moisture inside the membrane throughout the experiment. The volume seeped through the membrane and collected in the basolateral part of the transwell, which was calculated and measured every 30 min. The persistence property of the BBB was analyzed by measuring the volume that remained on the apical side of the transwell after 3 h. The media volumes were calculated every half hour using a 1 or 10 μL pipette. ## 4.6. APPswe/ind Plasmid Transfection The pCAX APPswe/ind plasmid, a gift from Dennis Selkoe and Tracy Young-Pearse (Addgene plasmid # 30145), was transfected in $75\%$ confluent HT-22 cells in a 6-well cell culturing plate. Transfection was conducted using Lipofectamine 3000 transfection reagent (ThermoFisher, Waltham, MA, USA) according to the manufacturer’s protocol. ## 4.7. Hematoxylin and Eosin (H&E) Staining To detect the cells’ confluency and verify that the cells remained adherent on the basolateral side of the transwell membrane against gravity, the transwells, after cell seeding and culturing, were stained with hematoxylin and eosin (H&E) after fixation with paraformaldehyde solution. The cells were revived by washing with phosphate buffer saline twice. Then, the cells were dehydrated with $70\%$, $80\%$, and $90\%$ ethanol. The slides were then treated with hematoxylin for 5 min, and the excess stain was removed by washing with PBS. The cells were co-stained with eosin for 30–45 s. The cells were observed under a compound microscope after the dehydration and washing steps to remove excess stains. ## 4.8. Electron Microscopy Cells cultured on the transwell membrane were fixed and processed for electron microscopy. A JSM-7610F scanning electron microscope (JEOL USA, Inc., Peabody, MA, USA) was used to obtain the scanning electron micrographs. ## 4.9. Experimental Animals The experimental procedures were approved by the Research Ethics Committee of the Department of Applied Life Sciences at Gyeongsang National University, the Republic of Korea. All experiments were performed according to the guidelines and regulations of the research ethics committee. Male C57BL/6N mice (8 weeks old, $$n = 8$$, 4 mice per group) weighing between 24 and 30 g were purchased from Samtako Bio, Osan, South Korea. The animals were handled as per the recommendations of the Institutional Animal Care and Use Committee (IACUC) of the Division of Applied Life Science, Gyeongsang National University, South Korea. The mice were acclimatized for seven days in an animal care house (4–5 per cage) under standard environmental conditions (temperature, 20 ± 2 °C humidity $40\%$ ± $10\%$; 12 h light/dark cycle) and were provided with normal pellet food and water ad libitum. ## 4.10. Amyloid Beta (1–42) Preparation Formation of aggregate-free amyloid beta 1–42 (Aβ1–42) peptides was performed using hexafluoroisopropane (HFIP), as previously reported with some modifications [71]. Briefly, 2.217 mL of HFIP was added to 10 mg of peptides using a glass Hamilton syringe (Thomas Scientific, New Jersey, United States.) equipped with a Teflon plunger to reach the final concentration of 1 mM and incubated for 30 min until dissolved completely. HFIP-dissolved Aβ solution was aliquoted into non-silanized microcentrifugation tubes for evaporation to allow a dry thin film of peptide to form; then re-dissolved in DMSO; and diluted with phenol-free F-12 cell culture media (Thomas Scientific, Swedesboro, NJ, USA) to reach the final concentration of 100 µM. F-12 cell media-dissolved Aβ peptides were kept at 4 °C until use. ## 4.11. Intracerebroventricular (ICV) Injection of Aβ Peptides Aβ1–42 peptides were injected according to the previously mentioned protocol [72,73]. Briefly, 1 mg/mL of human Aβ1–42 peptide stock solution was prepared and incubated at 37 °C for four days. Aggregated Aβ1–42 peptides and the respective vehicle were injected at a concentration of 5 µL/mouse or $0.9\%$ NaCl, 5 µL/mouse, respectively, with the help of a Hamilton microsyringe (−0.2 mm anteroposterior (AP), 1 mm mediolateral (ML), and −2.4 mm dorsoventral (DV) into the bregma). As previously suggested, the animals were anesthetized with a combination of 0.05 mL/100 g body weight of Rompun (xylazine) and 0.1 mL/100 g body weight of Zolitil (ketamine) delivered at a rate of 1 ul/5 min, with the injector remaining intact for 5 min. ## 4.12. Protein Extraction and Western Blot Analysis Protein extraction and Western blot analysis were performed as previously described [74,75]. Briefly, all the mice were anesthetized and decapitated. Their brains were immediately removed and the cortex and hippocampus were separated carefully and frozen at −80 °C. The cortex and hippocampus were homogenized in a pro-prepTM protein extraction solution (iNtRON Biotechnology, Inc., Sungnam, South Korea). The brain samples were then centrifuged at 13,000 r.p.m. at 4 °C for 25 min. The supernatants were collected and stored at −80 °C. Next, the cell lysate protein concentration was quantified using the Bradford assay (Bio-Rad protein assay kit, Bio-Rad Laboratories, Hercules, CA, USA). Equal amounts of protein (20 µg) were electrophoresed under the same experimental conditions, using 8–$12\%$ SDS gels and 1x MES SDS running buffer (Novex, Life Technologies, Kiryat Shmona, Israel) with a broad-range prestained protein marker (Xpert prestained protein marker, GenDEPOT, Texas, United States) as a molecular-size control. Membranes were blocked in $5\%$ (w/v) skim milk to reduce non-specific binding and then incubated with primary antibodies (1:1000–1:10,000 dilutions) overnight at 4 °C. After reactions with a horseradish peroxidase-conjugated secondary antibody, the proteins were detected using an enhanced chemiluminescence (ECL) detection reagent according to the manufacturer’s instructions (Amersham Pharmacia Biotech, Uppsala, Sweden). The X-ray films were scanned, and the optical densities of the bands were analyzed via densitometry, using the computer-based ImageJ software (version 1.50, Wayne Rasband and contributors, National Institutes of Health, Bethesda, MD, USA). ## 4.13. Immunofluorescence Analysis Immunofluorescence staining was performed according to a previously described protocol with minor modifications [58,76]. Slides with tissue or cells were washed with 0,1 M PBS twice, followed by incubation with proteinase K for 5 min. For blocking, the slides were incubated for 1 h with $2\%$ normal goat serum and $0.1\%$ Triton X-100 in 0.1 M PBS. After washing with PBS, the slides were incubated with primary antibodies overnight at 4°C and with secondary antibodies, including fluorescein isothiocyanate (FITC) or tetramethylrhodamine (TRITC) (anti-mouse and anti-rabbit), at room temperature for 90 min. The slides were stained with 4′,6-diamino-2-phenylindole dihydrochloride (DAPI) for 8 min and covered with fluorescent mounting media (Dako, Santa Clara, CA, USA) and a glass cover slide before microscopy. Immunofluorescence slides were examined using a confocal laser-scanning microscope (Flouview FV 1000, Olympus, Japan) and the relative integrated density values were recorded. ## 4.14. Antibodies The antibodies used in the present study included anti-ZO-1 (Invitrogen, Thermo Fisher scientific), anti-Occludin-1 (Invitrogen, Thermo Fisher scientific), anti-Claudin-5 (Invitrogen, Thermo Fisher scientific), anti-Amyloid beta and anti-beta-actin (sc-47778) from Santa Cruz Biotechnology (Dallas, TX, USA). ## 4.15. Statistical Analysis Statistical analyses were conducted by Prism 8.0.2 (GraphPad, San Diego, CA, USA). A Shapiro–Wilk normality test was conducted before any statistical analysis. The data are presented as the mean ± standard error of the mean (SEM). For statistical analysis, two-tailed t-tests and ANOVA with Bonferroni correction were performed. p-values less than 0.05 was considered to be statistically significant. ## 5. Conclusions The present study established an in vitro model and included analyses of physical properties, permeability assays, TEER evaluations, and tight junction protein expression. Furthermore, the present established model has been used to mimic different brain conditions. Our analyses indicate that this model is an excellent in vitro BBB model that can be applied to multiple disease conditions related to BBB integrity. In the future, this model can be used as a convenient tool to study complex physiological conditions, as well as a mode of action of therapeutic agents in diverse neurological disorders, such as Alzheimer’s disease, Parkinson’s disease, ischemia, TBI, and brain metabolic disorders. ## References 1. Schoknecht K., David Y., Heinemann U.. The blood-brain barrier-gatekeeper to neuronal homeostasis: Clinical implications in the setting of stroke. Semin. Cell Dev. Biol. (2015) 38 35-42. DOI: 10.1016/j.semcdb.2014.10.004 2. Sweeney M.D., Sagare A.P., Zlokovic B.V.. Blood-brain barrier breakdown in Alzheimer disease and other neurodegenerative disorders. Nat. Rev. Neurol. (2018) 14 133-150. DOI: 10.1038/nrneurol.2017.188 3. Nation D.A., Sweeney M.D., Montagne A., Sagare A.P., D’Orazio L.M., Pachicano M., Sepehrband F., Nelson A.R., Buennagel D.P., Harrington M.G.. Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. Nat. Med. (2019) 25 270-276. DOI: 10.1038/s41591-018-0297-y 4. Amtul Z., Yang J., Nikolova S., Lee T.Y., Bartha R., Cechetto D.F.. The Dynamics of Impaired Blood-Brain Barrier Restoration in a Rat Model of Co-morbid Injury. Mol. Neurobiol. (2018) 55 8071-8083. DOI: 10.1007/s12035-018-0904-4 5. Liebner S., Dijkhuizen R.M., Reiss Y., Plate K.H., Agalliu D., Constantin G.. Functional morphology of the blood-brain barrier in health and disease. Acta Neuropathol. (2018) 135 311-336. DOI: 10.1007/s00401-018-1815-1 6. Yang Y., Kimura-Ohba S., Thompson J.F., Salayandia V.M., Cosse M., Raz L., Jalal F.Y., Rosenberg G.A.. Vascular tight junction disruption and angiogenesis in spontaneously hypertensive rat with neuroinflammatory white matter injury. Neurobiol. Dis. (2018) 114 95-110. DOI: 10.1016/j.nbd.2018.02.012 7. Otani T., Furuse M.. Tight Junction Structure and Function Revisited. Trends Cell Biol. (2020) 30 805-817. DOI: 10.1016/j.tcb.2020.08.004 8. Carrano A., Hoozemans J.J., van der Vies S.M., van Horssen J., de Vries H.E., Rozemuller A.J.. Neuroinflammation and blood-brain barrier changes in capillary amyloid angiopathy. Neurodegener. Dis. (2012) 10 329-331. DOI: 10.1159/000334916 9. Carrano A., Hoozemans J.J., van der Vies S.M., Rozemuller A.J., van Horssen J., de Vries H.E.. Amyloid Beta induces oxidative stress-mediated blood-brain barrier changes in capillary amyloid angiopathy. Antioxid. Redox Signal. (2011) 15 1167-1178. DOI: 10.1089/ars.2011.3895 10. Villacrez M., Hellman K., Ono T., Sugihara Y., Rezeli M., Ek F., Marko-Varga G., Olsson R.. Evaluation of Drug Exposure and Metabolism in Locust and Zebrafish Brains Using Mass Spectrometry Imaging. ACS Chem. Neurosci. (2018) 9 1994-2000. DOI: 10.1021/acschemneuro.7b00459 11. Ghersi-Egea J.F., Saudrais E., Strazielle N.. Barriers to Drug Distribution into the Perinatal and Postnatal Brain. Pharm. Res. (2018) 35 84. DOI: 10.1007/s11095-018-2375-8 12. Srinivasan B., Kolli A.R., Esch M.B., Abaci H.E., Shuler M.L., Hickman J.J.. TEER measurement techniques for in vitro barrier model systems. J. Lab. Autom. (2015) 20 107-126. DOI: 10.1177/2211068214561025 13. Montagne A., Zhao Z., Zlokovic B.V.. Alzheimer′s disease: A matter of blood-brain barrier dysfunction?. J. Exp. Med. (2017) 214 3151-3169. DOI: 10.1084/jem.20171406 14. Lin Z., Sur S., Liu P., Li Y., Jiang D., Hou X., Darrow J., Pillai J.J., Yasar S., Rosenberg P.. Blood–Brain Barrier Breakdown in Relationship to Alzheimer and Vascular Disease. Ann. Neurol. (2021) 90 227-238. DOI: 10.1002/ana.26134 15. Kurz C., Walker L., Rauchmann B.-S., Perneczky R.. Dysfunction of the blood–brain barrier in Alzheimer’s disease: Evidence from human studies. Neuropathol. Appl. Neurobiol. (2022) 48 e12782. DOI: 10.1111/nan.12782 16. Zenaro E., Piacentino G., Constantin G.. The blood-brain barrier in Alzheimer’s disease. Neurobiol. Dis. (2017) 107 41-56. DOI: 10.1016/j.nbd.2016.07.007 17. Salvador E., Burek M., Forster C.Y.. An In Vitro Model of Traumatic Brain Injury. Methods Mol. Biol. (2018) 1717 219-227. DOI: 10.1007/978-1-4939-7526-6_17 18. Duan Y., Zeng L., Zheng C., Song B., Li F., Kong X., Xu K.. Inflammatory Links Between High Fat Diets and Diseases. Front. Immunol. (2018) 9 2649. DOI: 10.3389/fimmu.2018.02649 19. Fatima S., Hu X., Huang C., Zhang W., Cai J., Huang M., Gong R.-H., Chen M., Ho A.H.M., Su T.. High-fat diet feeding and palmitic acid increase CRC growth in β2AR-dependent manner. Cell Death Dis. (2019) 10 711. DOI: 10.1038/s41419-019-1958-6 20. Rebai R., Jasmin L., Boudah A.. Agomelatine effects on fat-enriched diet induced neuroinflammation and depression-like behavior in rats. Biomed Pharm. (2021) 135 111246. DOI: 10.1016/j.biopha.2021.111246 21. Carta G., Murru E., Banni S., Manca C.. Palmitic Acid: Physiological Role, Metabolism and Nutritional Implications. Front. Physiol. (2017) 8 902. DOI: 10.3389/fphys.2017.00902 22. Seidel G.A., Giovannetti T., Libon D.J.. Cerebrovascular disease and cognition in older adults. Curr. Top. Behav. Neurosci. (2012) 10 213-241. DOI: 10.1007/7854_2011_140 23. Albai O., Frandes M., Timar R., Roman D., Timar B.. Risk factors for developing dementia in type 2 diabetes mellitus patients with mild cognitive impairment. Neuropsychiatr. Dis. Treat. (2019) 15 167-175. DOI: 10.2147/NDT.S189905 24. Ahmad M.A., Kareem O., Khushtar M., Akbar M., Haque M.R., Iqubal A., Haider M.F., Pottoo F.H., Abdulla F.S., Al-Haidar M.B.. Neuroinflammation: A Potential Risk for Dementia. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23020616 25. Kim J.Y., Lee H.J., Lee S.J., Jung Y.H., Yoo D.Y., Hwang I.K., Seong J.K., Ryu J.M., Han H.J.. Palmitic Acid-BSA enhances Amyloid-beta production through GPR40-mediated dual pathways in neuronal cells: Involvement of the Akt/mTOR/HIF-1alpha and Akt/NF-kappaB pathways. Sci. Rep. (2017) 7 4335. DOI: 10.1038/s41598-017-04175-w 26. Patil S., Chan C.. Palmitic and stearic fatty acids induce Alzheimer-like hyperphosphorylation of tau in primary rat cortical neurons. Neurosci. Lett. (2005) 384 288-293. DOI: 10.1016/j.neulet.2005.05.003 27. Patil S., Balu D., Melrose J., Chan C.. Brain region-specificity of palmitic acid-induced abnormalities associated with Alzheimer’s disease. BMC Res. Notes (2008) 1. DOI: 10.1186/1756-0500-1-20 28. Borges N., Shi F., Azevedo I., Audus K.L.. Changes in brain microvessel endothelial cell monolayer permeability induced by adrenergic drugs. Eur. J. Pharmacol. Mol. Pharmacol. (1994) 269 243-248. DOI: 10.1016/0922-4106(94)90092-2 29. Hartz A.M., Miller D.S., Bauer B.. Restoring blood-brain barrier P-glycoprotein reduces brain amyloid-β in a mouse model of Alzheimer’s disease. Mol. Pharmacol. (2010) 77 715-723. DOI: 10.1124/mol.109.061754 30. Wang Y., Wang N., Cai B., Wang G.-Y., Li J., Piao X.-X.. In vitro model of the blood-brain barrier established by co-culture of primary cerebral microvascular endothelial and astrocyte cells. Neural Regen. Res. (2015) 10 2011. DOI: 10.4103/1673-5374.172320 31. Wang Y.I., Abaci H.E., Shuler M.L.. Microfluidic blood–brain barrier model provides in vivo-like barrier properties for drug permeability screening. Biotechnol. Bioeng. (2017) 114 184-194. DOI: 10.1002/bit.26045 32. Nzou G., Wicks R., Wicks E., Seale S., Sane C., Chen A., Murphy S., Jackson J., Atala A.. Human cortex spheroid with a functional blood brain barrier for high-throughput neurotoxicity screening and disease modeling. Sci. Rep. (2018) 8 7413. DOI: 10.1038/s41598-018-25603-5 33. Pardridge W.M.. The blood-brain barrier: Bottleneck in brain drug development. NeuroRx (2005) 2 3-14. DOI: 10.1602/neurorx.2.1.3 34. Robinson B.D., Shaji C.A., Lomas A., Tharakan B.. Measurement of Microvascular Endothelial Barrier Dysfunction and Hyperpermeability In Vitro. Methods Mol. Biol. (2018) 1717 237-242. DOI: 10.1007/978-1-4939-7526-6_19 35. Petto C., Lesko S., Gabel G., Bottner M., Wedel T., Kacza J., Pfannkuche H.. Establishment and characterization of porcine colonic epithelial cells grown in primary culture. Cells Tissues Organs (2011) 194 457-468. DOI: 10.1159/000323916 36. Booth R., Kim H.. Characterization of a microfluidic in vitro model of the blood-brain barrier (muBBB). Lab Chip (2012) 12 1784-1792. DOI: 10.1039/c2lc40094d 37. Devraj K., Guerit S., Macas J., Reiss Y.. An In Vivo Blood-brain Barrier Permeability Assay in Mice Using Fluorescently Labeled Tracers. J. Vis. Exp. (2018) 132 e57038. DOI: 10.3791/57038 38. Yue Q., Zhou X., Zhang Z., Hoi M.P.M.. Murine Beta-Amyloid (1–42) Oligomers Disrupt Endothelial Barrier Integrity and VEGFR Signaling via Activating Astrocytes to Release Deleterious Soluble Factors. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23031878 39. Abbott N.J.. Astrocyte-endothelial interactions and blood-brain barrier permeability. J. Anat. (2002) 200 629-638. DOI: 10.1046/j.1469-7580.2002.00064.x 40. Takeshita Y., Obermeier B., Cotleur A., Sano Y., Kanda T., Ransohoff R.M.. An in vitro blood-brain barrier model combining shear stress and endothelial cell/astrocyte co-culture. J. Neurosci. Methods (2014) 232 165-172. DOI: 10.1016/j.jneumeth.2014.05.013 41. Michinaga S., Koyama Y.. Protection of the Blood-Brain Barrier as a Therapeutic Strategy for Brain Damage. Biol. Pharm. Bull. (2017) 40 569-575. DOI: 10.1248/bpb.b16-00991 42. Kim S., Kim G.H.. Roles of claudin-2, ZO-1 and occludin in leaky HK-2 cells. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0189221 43. Coronado-Velazquez D., Betanzos A., Serrano-Luna J., Shibayama M.. An in vitro Model of the Blood-Brain Barrier: Naegleria fowleri Affects the Tight Junction Proteins and Activates the Microvascular Endothelial Cells. J. Eukaryot. Microbiol. (2018) 65 804-819. DOI: 10.1111/jeu.12522 44. Arthur F.E., Shivers R.R., Bowman P.D.. Astrocyte-mediated induction of tight junctions in brain capillary endothelium: An efficient in vitro model. Brain Res. (1987) 433 155-159. DOI: 10.1016/0165-3806(87)90075-7 45. Cabezas R., Avila M., Gonzalez J., El-Bacha R.S., Baez E., Garcia-Segura L.M., Jurado Coronel J.C., Capani F., Cardona-Gomez G.P., Barreto G.E.. Astrocytic modulation of blood brain barrier: Perspectives on Parkinson’s disease. Front. Cell Neurosci. (2014) 8 211. DOI: 10.3389/fncel.2014.00211 46. Irudayanathan F.J., Wang N., Wang X., Nangia S.. Architecture of the paracellular channels formed by claudins of the blood-brain barrier tight junctions. Ann. N.Y. Acad. Sci. (2017) 1405 131-146. DOI: 10.1111/nyas.13378 47. Hashimoto Y., Shirakura K., Okada Y., Takeda H., Endo K., Tamura M., Watari A., Sadamura Y., Sawasaki T., Doi T.. Claudin-5-Binders Enhance Permeation of Solutes across the Blood-Brain Barrier in a Mammalian Model. J. Pharmacol. Exp. Ther. (2017) 363 275-283. DOI: 10.1124/jpet.117.243014 48. Persidsky Y., Ramirez S.H., Haorah J., Kanmogne G.D.. Blood-brain barrier: Structural components and function under physiologic and pathologic conditions. J. Neuroimmune Pharm. (2006) 1 223-236. DOI: 10.1007/s11481-006-9025-3 49. Rhea E.M., Salameh T.S., Logsdon A.F., Hanson A.J., Erickson M.A., Banks W.A.. Blood-Brain Barriers in Obesity. Aaps J. (2017) 19 921-930. DOI: 10.1208/s12248-017-0079-3 50. Chodobski A., Zink B.J., Szmydynger-Chodobska J.. Blood-brain barrier pathophysiology in traumatic brain injury. Transl. Stroke Res. (2011) 2 492-516. DOI: 10.1007/s12975-011-0125-x 51. Zipser B.D., Johanson C.E., Gonzalez L., Berzin T.M., Tavares R., Hulette C.M., Vitek M.P., Hovanesian V., Stopa E.G.. Microvascular injury and blood-brain barrier leakage in Alzheimer’s disease. Neurobiol. Aging (2007) 28 977-986. DOI: 10.1016/j.neurobiolaging.2006.05.016 52. Kanoski S.E., Zhang Y., Zheng W., Davidson T.L.. The effects of a high-energy diet on hippocampal function and blood-brain barrier integrity in the rat. J. Alzheimers Dis. (2010) 21 207-219. DOI: 10.3233/JAD-2010-091414 53. Ouyang S., Hsuchou H., Kastin A.J., Wang Y., Yu C., Pan W.. Diet-induced obesity suppresses expression of many proteins at the blood-brain barrier. J. Cereb. Blood Flow Metab. (2014) 34 43-51. DOI: 10.1038/jcbfm.2013.166 54. Kim M.W., Choe K., Park J.S., Lee H.J., Kang M.H., Ahmad R., Kim M.O.. Pharmacological Inhibition of Spleen Tyrosine Kinase Suppressed Neuroinflammation and Cognitive Dysfunction in LPS-Induced Neurodegeneration Model. Cells (2022) 11. DOI: 10.3390/cells11111777 55. Li Y., Fanning A.S., Anderson J.M., Lavie A.. Structure of the conserved cytoplasmic C-terminal domain of occludin: Identification of the ZO-1 binding surface. J. Mol. Biol. (2005) 352 151-164. DOI: 10.1016/j.jmb.2005.07.017 56. Wen J., Qian S., Yang Q., Deng L., Mo Y., Yu Y.. Overexpression of netrin-1 increases the expression of tight junction-associated proteins, claudin-5, occludin, and ZO-1, following traumatic brain injury in rats. Exp. Ther. Med. (2014) 8 881-886. DOI: 10.3892/etm.2014.1818 57. Illan-Gala I., Pegueroles J., Montal V., Alcolea D., Vilaplana E., Bejanin A., Borrego-Ecija S., Sampedro F., Subirana A., Sanchez-Saudinos M.B.. APP-derived peptides reflect neurodegeneration in frontotemporal dementia. Ann. Clin. Transl. Neurol. (2019) 6 2518-2530. DOI: 10.1002/acn3.50948 58. Abid N.B., Yoon G., Kim M.O.. Molecular Cloning and Expression of Osmotin in a Baculovirus-Insect System: Purified Osmotin Mitigates Amyloid-beta Deposition in Neuronal Cells. Sci. Rep. (2017) 7 8147. DOI: 10.1038/s41598-017-08396-x 59. Biron K.E., Dickstein D.L., Gopaul R., Jefferies W.A.. Amyloid triggers extensive cerebral angiogenesis causing blood brain barrier permeability and hypervascularity in Alzheimer’s disease. PLoS ONE (2011) 6. DOI: 10.1371/journal.pone.0023789 60. Hartz A.M., Bauer B., Soldner E.L., Wolf A., Boy S., Backhaus R., Mihaljevic I., Bogdahn U., Klünemann H.H., Schuierer G.. Amyloid-β contributes to blood-brain barrier leakage in transgenic human amyloid precursor protein mice and in humans with cerebral amyloid angiopathy. Stroke (2012) 43 514-523. DOI: 10.1161/STROKEAHA.111.627562 61. Keaney J., Walsh D.M., O’Malley T., Hudson N., Crosbie D.E., Loftus T., Sheehan F., McDaid J., Humphries M.M., Callanan J.J.. Autoregulated paracellular clearance of amyloid-β across the blood-brain barrier. Sci. Adv. (2015) 1 e1500472. DOI: 10.1126/sciadv.1500472 62. Winkler E.A., Nishida Y., Sagare A.P., Rege S.V., Bell R.D., Perlmutter D., Sengillo J.D., Hillman S., Kong P., Nelson A.R.. GLUT1 reductions exacerbate Alzheimer’s disease vasculo-neuronal dysfunction and degeneration. Nat. Neurosci. (2015) 18 521-530. DOI: 10.1038/nn.3966 63. Verkhratsky A., Matteoli M., Parpura V., Mothet J.P., Zorec R.. Astrocytes as secretory cells of the central nervous system: Idiosyncrasies of vesicular secretion. Embo J. (2016) 35 239-257. DOI: 10.15252/embj.201592705 64. Tsai S.F., Wu H.T., Chen P.C., Chen Y.W., Yu M., Wang T.F., Wu S.Y., Tzeng S.F., Kuo Y.M.. High-fat diet suppresses the astrocytic process arborization and downregulates the glial glutamate transporters in the hippocampus of mice. Brain Res. (2018) 1700 66-77. DOI: 10.1016/j.brainres.2018.07.017 65. De Aquino C.C., Leitão R.A., Oliveira Alves L.A., Coelho-Santos V., Guerrant R.L., Ribeiro C.F., Malva J.O., Silva A.P., Oriá R.B.. Effect of Hypoproteic and High-Fat Diets on Hippocampal Blood-Brain Barrier Permeability and Oxidative Stress. Front. Nutr. (2018) 5 131. DOI: 10.3389/fnut.2018.00131 66. Ikram M., Jo M.H., Choe K., Khan A., Ahmad S., Saeed K., Kim M.W., Kim M.O.. Cycloastragenol, a Triterpenoid Saponin, Regulates Oxidative Stress, Neurotrophic Dysfunctions, Neuroinflammation and Apoptotic Cell Death in Neurodegenerative Conditions. Cells (2021) 10. DOI: 10.3390/cells10102719 67. Ahmad S., Jo M.H., Ikram M., Khan A., Kim M.O.. Deciphering the Potential Neuroprotective Effects of Luteolin against Aβ(1)-(42)-Induced Alzheimer’s Disease. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22179583 68. Khan A., Park T.J., Ikram M., Ahmad S., Ahmad R., Jo M.G., Kim M.O.. Antioxidative and Anti-inflammatory Effects of Kojic Acid in Aβ-Induced Mouse Model of Alzheimer’s Disease. Mol. Neurobiol. (2021) 58 5127-5140. DOI: 10.1007/s12035-021-02460-4 69. Ikram M., Jo M.G., Park T.J., Kim M.W., Khan I., Jo M.H., Kim M.O.. Oral Administration of Gintonin Protects the Brains of Mice against Aβ-Induced Alzheimer Disease Pathology: Antioxidant and Anti-Inflammatory Effects. Oxid. Med. Cell Longev. (2021) 2021 6635552. DOI: 10.1155/2021/6635552 70. Henn A., Lund S., Hedtjarn M., Schrattenholz A., Porzgen P., Leist M.. The suitability of BV2 cells as alternative model system for primary microglia cultures or for animal experiments examining brain inflammation. ALTEX (2009) 26 83-94. DOI: 10.14573/altex.2009.2.83 71. Brouillette J., Caillierez R., Zommer N., Alves-Pires C., Benilova I., Blum D., De Strooper B., Buee L.. Neurotoxicity and memory deficits induced by soluble low-molecular-weight amyloid-beta1-42 oligomers are revealed in vivo by using a novel animal model. J. Neurosci. (2012) 32 7852-7861. DOI: 10.1523/JNEUROSCI.5901-11.2012 72. Amin F.U., Shah S.A., Badshah H., Khan M., Kim M.O.. Anthocyanins encapsulated by PLGA@PEG nanoparticles potentially improved its free radical scavenging capabilities via p38/JNK pathway against Abeta1-42-induced oxidative stress. J. Nanobiotechnol. (2017) 15 12. DOI: 10.1186/s12951-016-0227-4 73. Ali T., Kim M.O.. Melatonin ameliorates amyloid beta-induced memory deficits, tau hyperphosphorylation and neurodegeneration via PI3/Akt/GSk3beta pathway in the mouse hippocampus. J. Pineal Res. (2015) 59 47-59. DOI: 10.1111/jpi.12238 74. Ali T., Rehman S.U., Khan A., Badshah H., Abid N.B., Kim M.W., Jo M.H., Chung S.S., Lee H.G., Rutten B.P.F.. Adiponectin-mimetic novel nonapeptide rescues aberrant neuronal metabolic-associated memory deficits in Alzheimer’s disease. Mol. Neurodegener. (2021) 16 23. DOI: 10.1186/s13024-021-00445-4 75. Shah S.A., Yoon G.H., Chung S.S., Abid M.N., Kim T.H., Lee H.Y., Kim M.O.. Novel osmotin inhibits SREBP2 via the AdipoR1/AMPK/SIRT1 pathway to improve Alzheimer′s disease neuropathological deficits. Mol. Psychiatry (2017) 22 407-416. DOI: 10.1038/mp.2016.23 76. Ahmad A., Ali T., Kim M.W., Khan A., Jo M.H., Rehman S.U., Khan M.S., Abid N.B., Khan M., Ullah R.. Adiponectin homolog novel osmotin protects obesity/diabetes-induced NAFLD by upregulating AdipoRs/PPARalpha signaling in ob/ob and db/db transgenic mouse models. Metab. Clin. Exp. (2019) 90 31-43. DOI: 10.1016/j.metabol.2018.10.004
--- title: 'Management of Glycemia during Acute Aerobic and Resistance Training in Patients with Diabetes Type 1: A Croatian Pilot Study' authors: - Marul Ivandic - Maja Cigrovski Berkovic - Klara Ormanac - Dea Sabo - Tea Omanovic Kolaric - Lucija Kuna - Vjera Mihaljevic - Silvija Canecki Varzic - Martina Smolic - Ines Bilic-Curcic journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049388 doi: 10.3390/ijerph20064966 license: CC BY 4.0 --- # Management of Glycemia during Acute Aerobic and Resistance Training in Patients with Diabetes Type 1: A Croatian Pilot Study ## Abstract [1] Background: The increased risk of developing hypoglycemia and worsening of glycemic stability during exercise is a major cause of concern for patients with type 1 diabetes mellitus (T1DM). [ 2] Aim: This pilot study aimed to assess glycemic stability and hypoglycemic episodes during and after aerobic versus resistance exercises using a flash glucose monitoring system in patients with T1DM. [ 3] Participants and Methods: We conducted a randomized crossover prospective study including 14 adult patients with T1DM. Patients were randomized according to the type of exercise (aerobic vs. resistance) with a recovery period of three days between a change of groups. Glucose stability and hypoglycemic episodes were evaluated during and 24 h after the exercise. Growth hormone (GH), cortisol, and lactate levels were determined at rest, 0, 30, and 60 min post-exercise period. [ 4] Results: The median age of patients was 53 years, with a median HbA1c of $7.1\%$ and a duration of diabetes of 30 years. During both training sessions, there was a drop in glucose levels immediately after the exercise (0′), followed by an increase at 30′ and 60′, although the difference was not statistically significant. However, glucose levels significantly decreased from 60′ to 24 h in the post-exercise period ($$p \leq 0.001$$) for both types of exercise. Glycemic stability was comparable prior to and after exercise for both training sessions. No differences in the number of hypoglycemic episodes, duration of hypoglycemia, and average glucose level in 24 h post-exercise period were observed between groups. Time to hypoglycemia onset was prolonged after the resistance as opposed to aerobic training (13 vs. 8 h, p = NS). There were no nocturnal hypoglycemic episodes (between 0 and 6 a.m.) after the resistance compared to aerobic exercise (4 vs. 0, p = NS). GH and cortisol responses were similar between the two sessions, while lactate levels were significantly more increased after resistance training. [ 5] Conclusion: Both exercise regimes induced similar blood glucose responses during and immediately following acute exercise. ## 1. Introduction Patients with T1DM and uncontrolled glucose levels can develop cardiovascular complications, while glucose-lowering treatment strategies can prevent or postpone their appearance. In those patients, exercise is advised for reducing the risk of several chronic conditions [1,2]. Nevertheless, there are some associated risks with exercise, such as fluctuating blood glucose levels from hyperglycemia to hypoglycemia [3]. In the past couple of decades, much attention has been placed on the relationship between T1DM and exercise, which led to new ideas for the treatment of T1DM [4]. Although exercise is advocated by the American Diabetes Association (ADA) as a means of treatment for people with diabetes, around $63\%$ of individuals with T1DM are inactive [5]. The main reason is the fear of potential adverse events that may occur during exercise, the most important being hypoglycemia. However, insulin dosage can be reduced and glycemic stability improved, minimizing the risk of hypoglycemia [5,6,7]. Additionally, aerobic exercise improves the quantity and function of insulin receptors on skeletal muscles and fat and increases the overall number of cellular glucose transporters, which leads to increased glucose uptake and a corresponding faster decrease of blood glucose during the activity [8,9]. In healthy individuals, during aerobic exercise, insulin release decreases while glucagon secretion increases, resulting in relatively stable glycemia. However, in T1DM, due to a lack of beta cell function, the only insulin available is the exogenous one. Its concentrations can be higher than endogenously produced insulin, and therefore, blood glucose levels are largely dependent on timing, quantity, and type of insulin administered. Thus, drops in blood glucose levels occur frequently and rapidly. Moreover, the liver produces less glucose in T1DM individuals who exercise because of higher blood insulin concentrations. Therefore, ingesting carbohydrates prior to aerobic exercise as well as reducing/withholding insulin doses might be a necessary strategy to avoid hypoglycemia [10,11,12]. On the other hand, exaggerated reduction or even skipping of insulin dosage, as well as consuming excessive amounts of carbohydrates, causes hyperglycemia before and during the exercise, possibly leading to ketosis [13,14,15]. During resistance training, insulin and glucagon cease to be the key energy regulation source, leading to increased lactate production due to non-oxygen-dependent glucose decomposition (anaerobic metabolism). Catecholamine levels increase in response to anaerobic changes push glucose production and release from the liver, causing blood glucose levels to rise and thus minimizing the risk of blood glucose drop during exercise. Nevertheless, hypoglycemia may develop for an extended period after exercise [12,16,17]. Better maintenance of glucose stability is associated with resistance exercise compared to aerobic exercise, although it can induce rebound hyperglycemia, probably due to increased secretion of counter-regulatory hormones [17,18]. Conversely, short and vigorous anaerobic exercise (for example, short sprints) or high-intensity interval training (HIIT) decreases the decline in blood glucose levels and thus has protective effects [19]. Minimization of hypoglycemia risk is attributed only to the exercise session itself; however, the current data for post-exercise hypoglycemia after anaerobic exercise do not support a clear conclusion [15,18,20,21]. Defining the optimal type and time of exercise is extremely difficult, even in healthy individuals. Patients with T1DM, besides individual preferences for different exercise types, need to consider the additional factor during physical activity, which is the maintenance of target glucose levels, although utilizing technology tools such as intermittent or continuous glucose monitoring, insulin pumps, and closed loop systems significantly improved the management of T1DM. Therefore, a better understanding of the impact of aerobic exercise, resistance training, or HIIT on glycemic variability, hypoglycemia, hyperglycemia, and energy expenditure is important for creating optimal strategies for better management of diabetes and a positive impact on HbA1c. The aim of this study was to assess glycemic levels and hormonal changes after acute aerobic versus resistance training using intermittent continuous glucose monitoring in patients with T1DM. ## 2.1. Participants We conducted a randomized crossover study including 14 untrained adult patients with T1DM for at least one year, with HbA1c of less than $9\%$ and a stable insulin regimen for the preceding 3 months. Patients with frequent hypoglycemia and those with hypoglycemia unawareness, patients on corticosteroid therapy, and those with chronic renal impairment (eGFR < 90 mL/min), active diabetic retinopathy, foot ulcers, and liver disease were excluded. The Ethical Committee of the Faculty of Medicine, University of Osijek, and the Ethical Committee of the Clinical Hospital Center Osijek approved the study protocol (R2-$\frac{4787}{2019}$ and 16 April 2019 respectively), and the investigation was conducted according to the Declaration of Helsinki. All patients gave their written informed consent before inclusion. Patients were randomized according to the type of exercise (aerobic and resistance exercise) with a recovery period of three days between group changes. Data on physical activity, diabetes glycemic stability, and hypoglycemia history were assessed before inclusion. All participants had an intermittently scanned continuous glucose monitoring (isCGM) FreeStyle Libre sensor placed 90 days prior to the study entry and were instructed to perform capillary blood glucose tests when measuring hypoglycemia with isCGM and to measure ketones in cases of blood glucose of >15 mmol/L. The study protocol was performed 48 h post sensor change to avoid any errors due to lower accuracy in the first 24 h or at the end of their useful lifespan. Interstitial glucose levels were monitored during a 24 h post-exercise period. Participants withheld from exercise for 24 h before or during the washout period. The study design scheme is presented in Figure 1. IsCGM data were handled according to recommendations from the international consensus on time in range by Battelino et al. [ 17]. ## 2.2. Experimental Design Participants arrived at the clinical research facility between 17:00 and 18:00 h. The following protocol for maintaining normoglycemia during exercise was applied: $75\%$ dose reduction of fast-acting insulin administered prior to consuming a meal containing 1 g of carbohydrate/kg 60 min before physical activity; post-workout meal containing 1 g of the low glycemic index (GI) carbohydrate/kg 60 min after physical activity with $50\%$ dose reduction of fast-acting insulin; bedtime snack containing 0.3 g of carbohydrate/kg with low GI and omission of prandial insulin [8]. An exact meal plan based on patients’ weight was created by a certified nutritionist. After a standardized 5 min warm-up of the main muscle groups, participants undertook aerobic/resistance exercises. The threshold for hypoglycemia was ≤3.9 and for hyperglycemia, >10.0 mmol/L. A hypoglycemic episode registered with isCGM was confirmed with a standard blood glucose meter. Data on insulin doses, food intake, and levels of physical activity for 24 h prior to exercise were obtained, and insulin doses and food intake were copied between experimental sessions of different types of exercise. Blood plasma was sampled at rest and 0 (immediately post exercise), 30, and 60 min (recovery phase) after cessation of the exercise, and blood glucose, lactate, GH, and cortisol levels were determined. Plasma was analyzed immediately after sampling. Measurements of blood glucose, HbA1c, and lactates were performed by routine assays using an automatic analyzer Olympus AV 640 (Olympus, Beckman Coulter, Inc., Brea, CA, USA). GH was determined using Beckman Coulter Access Dxi Chemiluminescence Immunoassay (CLIA), and cortisol was determined using the Alinity Chemiluminescence Microparticle Immunoassay system (CMIA), Abbott Diagnostics. ## 2.3. Exercise Protocol All participants performed two workout sessions with a 72 h recovery period in between and no additional activity. The aerobic exercise session lasted 45 min and consisted of walking on a treadmill to achieve a heart rate of $70\%$ intensity as determined by the estimated HRmax. After a 5 min warm-up at a heart rate of 50 to $60\%$ of HRmax (calculated for males as (202 − (0.55 × age) × 0.50) to (202 − (0.55 × age) × 0.60) and females as (216 − (1.09 × age)) × 0.50) to (216 − (1.09 × age)) × 0.60), followed a workout session for 30 min at a heart rate of 70–$75\%$ of HRmax (calculated for males as (202 − (0.55 × age) × 0.70) to (202 − (0.55 × age) × 0.75) and females as (216 − (1.09 × age)) × 0.70) to (216 − (1.09 × age)) × 0.75). The last 10 min were the same as the warm-up. Resistance training was conducted for 45 min using bodyweight and bar exercises including 10 exercises for different muscle groups in 3 sets of 12, 12, and 10 repetitions with a 30 s break between sets and a 2 min break between exercises. Resistance exercises included floor chest presses, overhead presses, bar squats, standing bent-over rows, and lying triceps extensions. The perceived rate of exertion was 5 to 6 on a scale of 1 to 10, which was below the expected anaerobic threshold and in a zone where, if needed, conversation could still be carried with effort. The anaerobic exercise target zone was considered to begin when $80\%$ of the maximal heart rate was reached, which is considered an anaerobic threshold. The exercise protocol is shown in Figure 2. ## 2.4. Statistical Methods Categorical data were presented in absolute and relative frequencies. Differences in categorical variables were tested by the McNemar–Bowker test or the marginal homogeneity test. The normality of the distribution of continuous variables was tested by Shapiro–Wilkinson test. Numerical data are described by median and interquartile range. Differences in numerical variables were tested by the Wilcoxon test. The association score is given by the Spearman correlation coefficient. All p-values are two-sided. The significance level was set to alpha = 0.05. MedCalc Statistical Software version 19.1.7 (MedCalc Software Ltd., Ostend, Belgium; https://www.medcalc.org; accessed on 4 March 2020) was used for statistical analysis. ## 3. Results Fourteen participants performed exercises. The median age of patients was 53 years, with a median HbA1c of $7.1\%$, duration of diabetes of 30 years, and a BMI of 24.5 kg/m2 (Table 1). Differences in glucose levels at different time points during aerobic and resistance exercise are summarized in Table 2. During both training sessions, there was a significant drop in blood glucose levels when comparing the pre-exercise period with 12 h and 24 h time points ($$p \leq 0.001$$). There was a drop in glucose levels immediately after the exercise (0′), followed by an increase at 30′ and 60′, although the difference was not statistically significant. Glycemic levels were comparable prior and after exercise for both training sessions. There were no differences between the two exercise types (Table 3). No differences in the number of hypoglycemic episodes, duration of hypoglycemia, nor average glucose level in 24 h post-exercise period were observed when comparing aerobic vs. resistance exercise. Time to hypoglycemia onset after resistance exercise compared to aerobic did not reach statistical significance (Table 4). There were 14 hypoglycemic episodes after aerobic and resistance exercise ($p \leq 0.99$) and 4 nocturnal hypoglycemic episodes after aerobic training, but none after resistance training ($$p \leq 0.13$$) (Table 5). Plasma cortisol and GH responses were similar at four different time points following resistance and aerobic exercise session. Lactate levels were significantly higher after resistance compared to the aerobic training session at 0, 30, and 60 min ($p \leq 0$,05); data obtained from blood plasma (Table 6). There was no change in lactate levels during aerobic training; however, a significant increase was noted immediately after cessation of resistance training (0 min) with complete normalization after 60 min (Figure 3). Cortisol levels were at their lowest 60 min after training (123 nmol/L; $$p \leq 0.02$$) in relation to all three-time points in the aerobic session, both before and immediately after aerobic exercise. Following resistance training, a decrease in cortisol was observed, whereas the peak level was registered at rest (270 nmol/L) with a tendency to drop (230–198–185 nmol/L, $$p \leq 0.01$$) over 60 min (Figure 4). An increase in growth hormone levels was observed after both aerobic and resistance training, followed by a decrease during the recovery phase (both $p \leq 0.001$); data obtained from blood plasma (Figure 5). ## 4. Discussion Lately, overcoming the problem of hypoglycemia and maintaining normoglycemia during exercise in T1D seems quite attainable, not only due to recently available technological tools such as continuous/flash glucose monitoring systems, insulin pumps, and closed loop systems, but also specifically developed protocols regarding insulin dose adjustments and carbohydrate intake in order to prevent hypoglycemia [19,22]. Still, it is not clear whether one type of exercise is more suitable than another [23]. Resistance exercise causes a smaller drop in glucose levels during the activity itself, yet larger reductions of glycemic levels in the post-exercise period were observed compared to aerobic training [22,24]. On the other hand, in some studies, resistance exercise was also associated with a modest increase in glycemia [18], while intense anaerobic workouts led to an increase in glucose levels [20]. Our results demonstrated similar advantages of resistance and aerobic training regarding rebound hyperglycemia and maintenance of glycemic stability over a 24 h post-exercise period. This could be explained by the study design implementing a protocol including low-GI meals 60 min prior to and after each type of exercise with the administration of reduced boluses by $25\%$ and $50\%$ of calculated bolus, respectively. In a previously published study investigating the impact of low- versus high-GI food on post-treadmill exercise glucose levels, the low-GI meal better prevented post-exercise hyperglycemia compared to the high-GI meal. However, both the high- and low-GI meals protected all patients from early hypoglycemia, but the risk of nocturnal hypoglycemia remained [11]. In addition, Campbell et al. showed that rapid-acting insulin reduction of $25\%$ pre-treadmill exercise and $50\%$ post-treadmill exercise maintained glucose levels and protected against early- (<8 h) but not late-onset hypoglycemia [25]. In addition, our study showed an increased occurrence of post-exercise hypoglycemic episodes; however, duration, and the number of episodes were similar regardless of the type of exercise, confirming the results of a recently published trial comparing aerobic and resistance training using CGM [22]. However, the time to hypoglycemia onset was prolonged for at least 8 h, confirming the usefulness of the applied protocol for hypoglycemia avoidance. Furthermore, all our participants were treated with multiple daily injections (MDI) using ultra-long-acting insulin, with no adjustments of basal insulin doses as recommended in previous studies [26,27]. Reduction in basal insulin dose might promote hyperglycemia at multiple points during the day (especially for patients using long- or ultra-long-acting insulin). Because it still reduces hypoglycemia risk during and after exercise, it could be particularly recommended for patients performing prolonged and intense activity, which was not the case in our study [26,27]. Several studies demonstrated that performing aerobic exercise in the afternoon or evening increases the risk of hypoglycemic episodes, especially nocturnal ones [26,28,29,30]. In our study, time to hypoglycemia onset was prolonged after the resistance as opposed to aerobic training (13 vs. 8 h), explaining the lack of nocturnal hypoglycemia after the resistance compared to aerobic exercise (4 vs. 0). This difference was not statistically significant but could point to the potential benefits of resistance versus aerobic exercise, especially if it is performed in the afternoon. Our results are further substantiated by the recent meta-analysis focusing on the delayed effects of different exercise modalities. Additionally, Valli et al. demonstrated that there is a reduced risk of hypoglycemia if exercise is performed in the morning rather than in the afternoon with a $50\%$ rapid-acting insulin reduction, but no definite benefits of resistance exercise were determined [30,31]. Based on the current recommendations, exercise should ideally be performed in the morning or at noon. However, many employed patients prefer exercising in the afternoon or evening; therefore, in this subset of patients, resistance training could be preferable to aerobic exercise [28]. IsCGM has been proven to be clinically valuable, reducing the risks of hypoglycemia, hyperglycemia, and glycemic variability (GV) and improving patient quality of life for a wide range of patient populations and clinical indications [32,33,34,35,36]. Still, isCGM also comes with a set of drawbacks. There is 10–15 min of lag time, because glucose is measured in interstitial fluid, as shown in previous studies [26,37]. If the glucose values are changing rapidly (e.g., intensive exercise, eating, etc.) the lag time is longer. In our participants, the maintenance of relatively good glycemic stability after both types of training could be partially attributed to the isCGM usage or close glucose monitoring, including trend arrows through which patients could easily follow glucose fluctuations over time, allowing timely detection and appropriate action if inadequate glucose values were observed [26,37,38]. As was expected, the GH response to both types of exercise were similar, with a significant increase immediately after the cessation, followed by a decrease in the recovery phase. The increase of GH levels facilitated by all types of exercise is well-documented [39,40,41]. GH response to stimuli depends on several factors, such as duration and intensity of physical activity, training, and fitness state [41]. It seems that in our study, the intensity of both types of exercise was comparable, as well as the training state of participants, leading to similar GH response. This is further substantiated by findings in several studies showing that during typical resistance training (RT) programs, such as the one used in this study, adults spend the majority of time at moderate intensity, regardless of BMI or age [42]. In our study, a decrease in cortisol concentrations during both types of training was observed, whereas previous studies reported the opposite effect, with higher cortisol levels during training [18,22]. Variations in circadian rhythm could justify a lack of response or even a decline of cortisol levels to both types of exercise even when they are performed in the afternoon [43]. Therefore, the lack of difference between glycemic stability in both exercise types could be explained by similar responses of counter-regulatory hormone in both exercise modes given that meals, basal insulin, and short-acting insulin reduction were standardized. Lactate levels were significantly increased following resistance training, while they remained completely unaffected after the aerobic session. This finding confirms a significant anaerobic component to RE reported in other studies [44,45]. Apparently, a rise in catecholamine secretion during RE induces muscle glycogenolysis enhancing glycogen decomposition mediated through phosphorylase α, leading to increased lactate production [44,46]. High lactate levels slow down glycogen utilization in the muscles, which could, in turn, help prevent early post-exercise hypoglycemia in patients with T1DM, as was the case in our study. The limitation of the current study is the small sample size. On the other hand, randomization to the cross-over design and detailed study protocol controlling both insulin doses and carbohydrate intake (low GI), add to the study’s strength. Moreover, measuring cortisol, GH, and lactate levels helped in the interpretation of glycemia levels after the aerobic and anaerobic exercise, suggesting that the higher lactate levels seen after resistance training potentially attenuate the decline in blood glucose levels, especially in the early post-exercise period, via stimulation of gluconeogenesis. ## 5. Conclusions Both resistance and aerobic exercise induced similar blood glucose responses during and immediately following acute exercise. Resistance training seems to be a more favorable choice in patients exercising in the afternoon or evening considering the prolonged time to hypoglycemia onset; however, this finding needs to be further investigated using a larger sample size. In addition, low GI meals and bolus dose reduction in combination with isCGM usage successfully prevented early post-exercise hyperglycemia in both types of exercise. ## References 1. Varkevisser R.D.M., Birnie E., Vollenbrock C.E., Mul D., van Dijk P.R., van der Klauw M.M., Veeze H., Wolffenbuttel B.H.R., Aanstoot H.J.. Cardiovascular risk management in people with type 1 diabetes: Performance using three guidelines. BMJ Open Diabetes. Res. Care (2022) 10 e002765. DOI: 10.1136/bmjdrc-2022-002765 2. Wu N., Bredin S.S.D., Guan Y., Dickinson K., Kim D.D., Chua Z., Kaufman K., Warburton D.E.R.. Cardiovascular Health Benefits of Exercise Training in Persons Living with Type 1 Diabetes: A Systematic Review and Meta-Analysis. J. Clin. Med. (2019) 8. DOI: 10.3390/jcm8020253 3. Fitzpatrick R., Davison G., Wilson J.J., McMahon G., McClean C.. Exercise, type 1 diabetes mellitus and blood glucose: The implications of exercise timing. Front. Endocrinol. (2022) 13 1021800. DOI: 10.3389/fendo.2022.1021800 4. Lu X., Zhao C.. Exercise and Type 1 Diabetes. Adv. Exp. Med. Biol. (2020) 1228 107-121. DOI: 10.1007/978-981-15-1792-1_7 5. Alkhateeb H., El Fathi A., Ghanbari M., Haidar A.. Modelling glucose dynamics during moderate exercise in individuals with type 1 diabetes. PLoS ONE (2021) 16. DOI: 10.1371/journal.pone.0248280 6. Litchfield I., Andrews R.C., Narendran P., Greenfield S.. Patient and Healthcare Professionals Perspectives on the Delivery of Exercise Education for Patients with Type 1 Diabetes. Front. Endocrinol. (2019) 10 76. DOI: 10.3389/fendo.2019.00076 7. Cai Y., Xie K.L., Zheng F., Liu S.X.. Aerobic Exercise Prevents Insulin Resistance Through the Regulation of miR-492/Resistin Axis in Aortic Endothelium. J. Cardiovasc. Transl. Res. (2018) 11 450-458. DOI: 10.1007/s12265-018-9828-7 8. Muñoz V.R., Gaspar R.C., Kuga G.K., da Rocha A.L., Crisol B.M., Botezelli J.D., Baptista I.L., Mekary R.A., da Silva A.S.R., Cintra D.E.. Exercise increases Rho-kinase activity and insulin signaling in skeletal muscle. J. Cell. Physiol. (2018) 233 4791-4800. DOI: 10.1002/jcp.26278 9. Deichmann J., Bachmann S., Burckhardt M.A., Szinnai G., Kaltenbach H.M.. Simulation-Based Evaluation of Treatment Adjustment to Exercise in Type 1 Diabetes. Front. Endocrinol. (2021) 12 723812. DOI: 10.3389/fendo.2021.723812 10. Bracken R.M., Page R., Gray B., Kilduff L.P., West D.J., Stephens J.W., Bain S.C.. Isomaltulose improves glycemia and maintains run performance in type 1 diabetes. Med. Sci. Sports Exerc. (2012) 44 800-808. DOI: 10.1249/MSS.0b013e31823f6557 11. Campbell M.D., Walker M., Trenell M.I., Stevenson E.J., Turner D., Bracken R.M., Shaw J.A., West D.J.. A low-glycemic index meal and bedtime snack prevents postprandial hyperglycemia and associated rises in inflammatory markers, providing protection from early but not late nocturnal hypoglycemia following evening exercise in type 1 diabetes. Diabetes Care (2014) 37 1845-1853. DOI: 10.2337/dc14-0186 12. Jung A.R., Kim H., Kim H.S., Kim C., Choi W.S.. Exercise Strategies to Prevent Hypoglycemia in Patients with Diabetes. Korean J. Fam. Med. (2021) 42 91-95. DOI: 10.4082/kjfm.19.0043 13. Bracken R.M., West D.J., Stephens J.W., Kilduff L.P., Luzio S., Bain S.C.. Impact of pre-exercise rapid-acting insulin reductions on ketogenesis following running in Type 1 diabetes. Diabet. Med. (2011) 28 218-222. DOI: 10.1111/j.1464-5491.2010.03162.x 14. West D.J., Stephens J.W., Bain S.C., Kilduff L.P., Luzio S., Still R., Bracken R.M.. A combined insulin reduction and carbohydrate feeding strategy 30 min before running best preserves blood glucose concentration after exercise through improved fuel oxidation in type 1 diabetes mellitus. J. Sports Sci. (2011) 29 279-289. DOI: 10.1080/02640414.2010.531753 15. Cockcroft E.J., Narendran P., Andrews R.C.. Exercise-induced hypoglycaemia in type 1 diabetes. Exp. Physiol. (2020) 105 590-599. DOI: 10.1113/EP088219 16. Farinha J.B., Krause M., Rodrigues-Krause J., Reischak-Oliveira A.. Exercise for type 1 diabetes mellitus management: General considerations and new directions. Med. Hypotheses (2017) 104 147-153. DOI: 10.1016/j.mehy.2017.05.033 17. Bally L., Zueger T., Buehler T., Dokumaci A.S., Speck C., Pasi N., Ciller C., Paganini D., Feller K., Loher H.. Metabolic and hormonal response to intermittent high-intensity and continuous moderate intensity exercise in individuals with type 1 diabetes: A randomised crossover study. Diabetologia (2016) 59 776-784. DOI: 10.1007/s00125-015-3854-7 18. Turner D., Luzio S., Gray B.J., Dunseath G., Rees E.D., Kilduff L.P., Campbell M.D., West D.J., Bain S.C., Bracken R.M.. Impact of single and multiple sets of resistance exercise in type 1 diabetes. Scand. J. Med. Sci. Sports (2015) 25 e99-e109. DOI: 10.1111/sms.12202 19. Pongrac Barlovic D., Harjutsalo V., Groop P.H.. Exercise and nutrition in type 1 diabetes: Insights from the FinnDiane cohort. Front. Endocrinol. (2022) 13 1064185. DOI: 10.3389/fendo.2022.1064185 20. Davey R.J., Paramalingam N., Retterath A.J., Lim E.M., Davis E.A., Jones T.W., Fournier P.A.. Antecedent hypoglycaemia does not diminish the glycaemia-increasing effect and glucoregulatory responses of a 10 s sprint in people with type 1 diabetes. Diabetologia (2014) 57 1111-1118. DOI: 10.1007/s00125-014-3218-8 21. Harmer A.R., Chisholm D.J., McKenna M.J., Morris N.R., Thom J.M., Bennett G., Flack J.R.. High-intensity training improves plasma glucose and acid-base regulation during intermittent maximal exercise in type 1 diabetes. Diabetes Care (2007) 30 1269-1271. DOI: 10.2337/dc06-1790 22. Reddy R., Wittenberg A., Castle J.R., El Youssef J., Winters-Stone K., Gillingham M., Jacobs P.G.. Effect of Aerobic and Resistance Exercise on Glycemic Control in Adults With Type 1 Diabetes. Can. J. Diabetes (2019) 43 406-414.e401. DOI: 10.1016/j.jcjd.2018.08.193 23. Petschnig R., Wagner T., Robubi A., Baron R.. Effect of Strength Training on Glycemic Control and Adiponectin in Diabetic Children. Med. Sci. Sports Exerc. (2020) 52 2172-2178. DOI: 10.1249/MSS.0000000000002356 24. de Abreu de Lima V., de Menezes F.J., da Rocha Celli L., França S.N., Cordeiro G.R., Mascarenhas L.P.G., Leite N.. Effects of resistance training on the glycemic control of people with type 1 diabetes: A systematic review and meta-analysis. Arch Endocrinol. Metab. (2022) 66 533-540. DOI: 10.20945/2359-3997000000487 25. Campbell M.D., Walker M., Trenell M.I., Jakovljevic D.G., Stevenson E.J., Bracken R.M., Bain S.C., West D.J.. Large pre- and postexercise rapid-acting insulin reductions preserve glycemia and prevent early- but not late-onset hypoglycemia in patients with type 1 diabetes. Diabetes Care (2013) 36 2217-2224. DOI: 10.2337/dc12-2467 26. Moser O., Riddell M.C., Eckstein M.L., Adolfsson P., Rabasa-Lhoret R., van den Boom L., Gillard P., Nørgaard K., Oliver N.S., Zaharieva D.P.. Glucose management for exercise using continuous glucose monitoring (CGM) and intermittently scanned CGM (isCGM) systems in type 1 diabetes: Position statement of the European Association for the Study of Diabetes (EASD) and of the International Society for Pediatric and Adolescent Diabetes (ISPAD) endorsed by JDRF and supported by the American Diabetes Association (ADA). Diabetologia (2020) 63 2501-2520. DOI: 10.1007/s00125-020-05263-9 27. Schubert-Olesen O., Kröger J., Siegmund T., Thurm U., Halle M.. Continuous Glucose Monitoring and Physical Activity. Int. J. Environ. Res. Public Health (2022) 19. DOI: 10.3390/ijerph191912296 28. Gomez A.M., Gomez C., Aschner P., Veloza A., Muñoz O., Rubio C., Vallejo S.. Effects of performing morning versus afternoon exercise on glycemic control and hypoglycemia frequency in type 1 diabetes patients on sensor-augmented insulin pump therapy. J. Diabetes Sci. Technol. (2015) 9 619-624. DOI: 10.1177/1932296814566233 29. Davey R.J., Howe W., Paramalingam N., Ferreira L.D., Davis E.A., Fournier P.A., Jones T.W.. The effect of midday moderate-intensity exercise on postexercise hypoglycemia risk in individuals with type 1 diabetes. J. Clin. Endocrinol. Metab. (2013) 98 2908-2914. DOI: 10.1210/jc.2013-1169 30. Toghi-Eshghi S.R., Yardley J.E.. Morning (Fasting) vs Afternoon Resistance Exercise in Individuals with Type 1 Diabetes: A Randomized Crossover Study. J. Clin. Endocrinol. Metab. (2019) 104 5217-5224. DOI: 10.1210/jc.2018-02384 31. Valli G., Minnock D., Tarantino G., Neville R.D.. Delayed effect of different exercise modalities on glycaemic control in type 1 diabetes mellitus: A systematic review and meta-analysis. Nutr. Metab. Cardiovasc. Dis. (2021) 31 705-716. DOI: 10.1016/j.numecd.2020.12.006 32. Bolinder J., Antuna R., Geelhoed-Duijvestijn P., Kröger J., Weitgasser R.. Novel glucose-sensing technology and hypoglycaemia in type 1 diabetes: A multicentre, non-masked, randomised controlled trial. Lancet (2016) 388 2254-2263. DOI: 10.1016/S0140-6736(16)31535-5 33. Haak T., Hanaire H., Ajjan R., Hermanns N., Riveline J.P., Rayman G.. Use of Flash Glucose-Sensing Technology for 12 months as a Replacement for Blood Glucose Monitoring in Insulin-treated Type 2 Diabetes. Diabetes Ther. (2017) 8 573-586. DOI: 10.1007/s13300-017-0255-6 34. Reddy M., Jugnee N., El Laboudi A., Spanudakis E., Anantharaja S., Oliver N.. A randomized controlled pilot study of continuous glucose monitoring and flash glucose monitoring in people with Type 1 diabetes and impaired awareness of hypoglycaemia. Diabet. Med. (2018) 35 483-490. DOI: 10.1111/dme.13561 35. Yaron M., Roitman E., Aharon-Hananel G., Landau Z., Ganz T., Yanuv I., Rozenberg A., Karp M., Ish-Shalom M., Singer J.. Effect of Flash Glucose Monitoring Technology on Glycemic Control and Treatment Satisfaction in Patients With Type 2 Diabetes. Diabetes Care (2019) 42 1178-1184. DOI: 10.2337/dc18-0166 36. Da Prato G., Pasquini S., Rinaldi E., Lucianer T., Donà S., Santi L., Negri C., Bonora E., Moghetti P., Trombetta M.. Accuracy of CGM Systems During Continuous and Interval Exercise in Adults with Type 1 Diabetes. J. Diabetes Sci. Technol. (2022) 16 1436-1443. DOI: 10.1177/19322968211023522 37. Muñoz Fabra E., Díez J.L., Bondia J., Laguna Sanz A.J.. A Comprehensive Review of Continuous Glucose Monitoring Accuracy during Exercise Periods. Sensors (2021) 21. DOI: 10.3390/s21020479 38. Fokkert M.J., van Dijk P.R., Edens M.A., Abbes S., de Jong D., Slingerland R.J., Bilo H.J.. Performance of the FreeStyle Libre Flash glucose monitoring system in patients with type 1 and 2 diabetes mellitus. BMJ Open Diabetes Res. Care (2017) 5 e000320. DOI: 10.1136/bmjdrc-2016-000320 39. Kanaley J.A., Weltman J.Y., Veldhuis J.D., Rogol A.D., Hartman M.L., Weltman A.. Human growth hormone response to repeated bouts of aerobic exercise. J. Appl. Physiol. (1997) 83 1756-1761. DOI: 10.1152/jappl.1997.83.5.1756 40. Weltman A., Weltman J.Y., Schurrer R., Evans W.S., Veldhuis J.D., Rogol A.D.. Endurance training amplifies the pulsatile release of growth hormone: Effects of training intensity. J. Appl. Physiol. (1992) 72 2188-2196. DOI: 10.1152/jappl.1992.72.6.2188 41. Sabag A., Chang D., Johnson N.A.. Growth Hormone as a Potential Mediator of Aerobic Exercise-Induced Reductions in Visceral Adipose Tissue. Front. Physiol. (2021) 12 623570. DOI: 10.3389/fphys.2021.623570 42. Bharti N., Hrubeniuk T., Mayo A., Sénéchal M., Bouchard D.R.. Resistance Training Contribute to the Aerobic Components of an Exercise Session in Adults but not as Much in Older Adults. Int. J. Exerc. Sci. (2017) 10 406-416. PMID: 28515837 43. Kanaley J.A., Weltman J.Y., Pieper K.S., Weltman A., Hartman M.L.. Cortisol and growth hormone responses to exercise at different times of day. J. Clin. Endocrinol. Metab. (2001) 86 2881-2889. DOI: 10.1210/jc.86.6.2881 44. King A., Helms E., Zinn C., Jukic I.. The Ergogenic Effects of Acute Carbohydrate Feeding on Resistance Exercise Performance: A Systematic Review and Meta-analysis. Sports Med. (2022) 52 2691-2712. DOI: 10.1007/s40279-022-01716-w 45. Cerretelli P., Samaja M.. Acid-base balance at exercise in normoxia and in chronic hypoxia. Revisiting the "lactate paradox". Eur. J. Appl. Physiol. (2003) 90 431-448. DOI: 10.1007/s00421-003-0928-x 46. Podolin D.A., Munger P.A., Mazzeo R.S.. Plasma catecholamine and lactate response during graded exercise with varied glycogen conditions. J. Appl. Physiol. (1991) 71 1427-1433. DOI: 10.1152/jappl.1991.71.4.1427
--- title: Role of Klotho and AGE/RAGE-Wnt/β-Catenin Signalling Pathway on the Development of Cardiac and Renal Fibrosis in Diabetes authors: - Beatriz Martín-Carro - Julia Martín-Vírgala - Sara Fernández-Villabrille - Alejandra Fernández-Fernández - Marcos Pérez-Basterrechea - Juan F. Navarro-González - Javier Donate-Correa - Carmen Mora-Fernández - Adriana S. Dusso - Natalia Carrillo-López - Sara Panizo - Manuel Naves-Díaz - José L. Fernández-Martín - Jorge B. Cannata-Andía - Cristina Alonso-Montes journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049403 doi: 10.3390/ijms24065241 license: CC BY 4.0 --- # Role of Klotho and AGE/RAGE-Wnt/β-Catenin Signalling Pathway on the Development of Cardiac and Renal Fibrosis in Diabetes ## Abstract Fibrosis plays an important role in the pathogenesis of long-term diabetic complications and contributes to the development of cardiac and renal dysfunction. The aim of this experimental study, performed in a long-term rat model, which resembles type 1 diabetes mellitus, was to investigate the role of soluble Klotho (sKlotho), advanced glycation end products (AGEs)/receptor for AGEs (RAGE), fibrotic Wnt/β-catenin pathway, and pro-fibrotic pathways in kidney and heart. Diabetes was induced by streptozotocin. Glycaemia was maintained by insulin administration for 24 weeks. Serum and urine sKlotho, AGEs, soluble RAGE (sRAGE) and biochemical markers were studied. The levels of Klotho, RAGEs, ADAM10, markers of fibrosis (collagen deposition, fibronectin, TGF-β1, and Wnt/β-catenin pathway), hypertrophy of the kidney and/or heart were analysed. At the end of study, diabetic rats showed higher levels of urinary sKlotho, AGEs and sRAGE and lower serum sKlotho compared with controls without differences in the renal Klotho expression. A significant positive correlation was found between urinary sKlotho and AGEs and urinary albumin/creatinine ratio (uACR). Fibrosis and RAGE levels were significantly higher in the heart without differences in the kidney of diabetic rats compared to controls. The results also suggest the increase in sKlotho and sRAGE excretion may be due to polyuria in the diabetic rats. ## 1. Introduction The global prevalence of diabetes has tripled in the last 20 years, and it is projected to increase by ∼$50\%$ by the year 2045 [1]. Diabetes-related complications, especially those related to the cardiovascular and renal system, compromise the quality of life of these patients and increase the premature mortality rate [2,3]. Fibrosis plays an important role in the pathogenesis of diabetic complications and contributes to the development of cardiac and renal dysfunction [4,5]. However, diabetes-associated fibrosis has not been fully elucidated regardless of its clinical relevance. In diabetes, the underlying signalling pathways involved in fibrosis are highly complex, with a wide range of functional drivers. Among them, increased advanced glycation end products (AGEs) and their receptor RAGE play an important role in renal and cardiac fibrosis by the activation of several overlapping fibrotic pathways [6,7]. Among the different mediators involved in the diabetic cardiomyopathy and nephropathy, several experimental studies have linked the Wnt/β-catenin signalling pathway with the development and progression of cardiac and renal fibrosis [8,9]. The Wnt/β-catenin is an evolutionarily conserved signalling pathway involved in heart and kidney development; however, it is functionally silent in adult tissues in which its activation is linked to the ageing- and injury-related fibrosis [10]. In addition, in diabetes the activity of the Wnt pathway could be modulated by the Klotho levels. Klotho is an anti-ageing protein predominantly expressed in renal tubular epithelial cells [11], which can be presented as transmembrane protein and a soluble form (sKlotho), that can be found in blood and urine, which plays a critical role in phosphorus homeostasis, acting as a humoral factor with pleiotropic effects on multiple organs [12,13]. Previous experimental studies have shown that Klotho exerts a protective effect on both the heart and kidneys [14]. This protection has been related with the inhibition of the Wnt pathway through protein–protein interactions between sKlotho and the Wnt pathway activators [15,16]. Furthermore, Klotho has been postulated as a biomarker of progression of heart and kidney damage [17,18,19]. The studies of Klotho in diabetes, mainly in type 1 diabetes mellitus (T1DM), are scarce, showing controversial results [20,21,22,23]. In addition, the relationship between Klotho and the AGE/RAGE pathway has not been studied in animal models of diabetes. Thus, the aim of this study was to investigate: (a) The Klotho levels (soluble and transmembrane protein in kidney) and its relationship to fibrosis; and (b) If Wnt/β-catenin and AGE/RAGE pathways are involved in heart and kidney fibrosis in a long-term (24 weeks) rat model that resembles the T1DM. ## 2.1. General, Glycaemic, Renal, and Cardiac Biochemical Parameters Twenty-four weeks after receiving streptozotocin (STZ), the rats showed significant lower body weight (BW) and higher urine volume compared to controls (Table 1). Diabetic rats showed higher levels of serum and urinary glucose, glycosylated haemoglobin (Hb1Ac), plasma and urinary AGEs and urinary soluble RAGE (sRAGE), but significantly lower levels of plasma sRAGE (Table 1). The urinary AGEs were significantly higher than in Week 9 (Figure S1A). Linear regression analysis showed a positive correlation between urinary AGEs and serum glucose on Week 24 ($r = 0.839$; $p \leq 0.001$), which persisted after adjustment for diabetes ($r = 0.859$; $p \leq 0.001$) (Figure S1C). The kidney analysis showed that the ratio kidney BW (left and right) was higher in the diabetic rats (Table 1). Rats that received STZ showed significantly higher levels of serum creatinine and urea, as well as urinary albumin/creatinine ratio (uACR) without changes in creatinine clearance, while serum levels of protein and albumin were significantly lower (Table 1). The uACR was significantly higher from Week 9 to Week 24 (Figure S1B). Compared with control, diabetic rats also showed a seven-fold higher level of urinary calcium without changes in serum calcium (Table 1). The urinary calcium was significantly higher on Week 9 (Figure S1D). The heart study showed a significative increase in the heart BW ratio and in the plasma levels of N-terminal pro-B-type natriuretic peptide (NT-proBNP) in diabetic rats compared to controls (Table 1). Serum sKlotho was significantly lower and urinary sKlotho significantly higher in diabetic rats (Table 1), the differences in urinary sKlotho increased at the end of the study (Figure 1A). A significant positive correlation was found between urinary sKlotho and urinary AGEs and uACR on Week 24 after adjusting for diabetes (Figure 1B,C). No differences were found in the renal Klotho expression, neither in mRNA nor in protein levels (Figure 2). ## 2.2.1. Histological Analysis The kidneys showed that the median values of the diameters of the proximal tubules were higher in diabetic rats (47.19 [45.87–49.50] vs. 54.89 [53.02–56.36] µm; $p \leq 0.001$) (Figure 3A). No changes in collagen deposition stained by Picrosirius Red were detected (Figure 3B). The electron microscopy examination showed a filtration barrier with thickened glomerular basement membrane and denuded foot processes of the podocytes in the glomeruli of the kidneys from diabetic rats, and pyknotic nuclei were also found (Figure S2). ## 2.2.2. Molecular Markers of Fibrosis, Wnt/β-Catenin, RAGE and ADAM10 Despite having no differences in renal collagen deposition between the two groups, higher mRNA levels of fibronectin (0.89 [0.62–1.39] vs. 1.32 [1.15–1.45] R.U.; $p \leq 0.05$) (Figure S3A) and no changes in Tgf-β1 were observed in diabetic rats compared to controls. In the kidney of diabetic rats, the Dkk1 gene (0.85 [0.68–1.11] vs. 1.80 [1.48–3.62] R.U.; $p \leq 0.001$) and protein (100.17 [82.24–104.58] vs. 114.02 [104.62–136.81] %; $p \leq 0.05$) expression were higher compared with controls (Figure S3B,C), without differences in the gene expression of Sost and Sfrp4; meanwhile Sfrp2 was significantly lower (0.72 [0.58–1.19] vs. 0.30 [0.20–0.61] R.U.; $p \leq 0.01$) (Figure S3D). No differences were found in the protein levels of active β-catenin. Although there were no differences in the Rage gene expression between the two groups, the protein levels (100.73 [73.72–116.78] vs. 119.62 [110.71–145.93] %; $p \leq 0.05$) were increased in the diabetic group. Additionally, the gene expression of Adam10 (0.95 [0.76–1.18] vs. 1.34 [0.94–1.45] R.U.; $p \leq 0.05$) was increased in the kidney of diabetic rats compared with controls (Figure S3E,F). ## 2.3.1. Histological Analysis The cardiomyocyte size was significantly higher (22.06 [20.66–23.38] vs. 25.29 [24.51–25.95] µm; $p \leq 0.001$) (Figure 4A) and deposition of collagen increased (1.96-fold), (2.61 [1.75–3.23] vs. 5.19 [4.49–6.49]%; $p \leq 0.001$) in diabetic rats (Figure 4B). ## 2.3.2. Molecular Markers of Fibrosis, Wnt/β-Catenin RAGE and ADAM10 The gene expression of fibrosis markers, such as fibronectin (0.99 [0.90–1.07] vs. 1.80 [1.32–2.27] R.U.; $p \leq 0.01$) and Tgf-β1 (0.99 [0.87–1.16] vs. 1.18 [1.09–1.30]), were significantly higher in diabetic rats (Figure S4A,B). High glucose levels maintained for 24 weeks resulted in a higher gene expression of Wnt pathway inhibitors Dkk1 (0.81 [0.56–1.01] vs. 2.11 [1.46–3.24] R.U.; $p \leq 0.05$) and Sfrp2 (0.51 [0.33–1.30] vs. 1.75 [0.85–2.51] R.U.; $p \leq 0.01$) compared to controls (Figure S4C,E), without changes in Sost and Sfrp4 expression. The diabetic rats also showed significantly higher protein levels of DKK1 (95.04 [84.57–108.49] vs. 119.17 [105.20–145.45] %; $p \leq 0.05$) and active β-catenin (95.34 [81.33–114.87] vs. 126.04 [103.51–177.54] %; $p \leq 0.05$) (Figure S4D,F). The mRNA (0.8776 [0.7504–1.2155 vs. 1.353 [1.227–1.575] R.U.; $p \leq 0.001$) and protein (99.59 [95.58–104.45] vs. 110.1 [105.9–119.6] %; $p \leq 0.01$) levels of the Rage and Adam10 (1.048 [0.87–1.09] vs. 1.72 [1.50–2.13] R.U.; $p \leq 0.001$) gene expression levels were significantly higher in diabetic rats compared with controls (Figure S4G–I). ## 3. Discussion The present study used a novel long-term (24 weeks) rat model of STZ-induced diabetes with insulin administration, which successfully resembles chronic T1DM characterized by weight loss, polyuria, hyperglycaemia, hyperglycosuria, and higher HbA1c. At the end of study, diabetic rats showed higher levels of urinary sKlotho, urinary AGEs and uACR, while serum sKlotho was lower compared with controls. There were no differences in the renal expression of Klotho, suggesting that the lower serum sKlotho found in the diabetic rats could be due to urinary loss as consequence of polyuria (19-fold higher urinary volume), rather than to a lower kidney expression. The lower serum levels of sKlotho due to its higher excretion could be related to the increase AGE/RAGE levels and the activation of pro-fibrotic Wnt/β-catenin pathway, observed mainly in the heart. Diabetes is associated with a greater risk of cardiovascular complications and kidney damage that leads to chronic heart and kidney failure, resulting in a bidirectional disorder known as cardiorenal syndrome [2,3]. Fibrosis, besides being a marker of injury progression, has been proposed as a driver of the pathophysiology of the cardiorenal syndrome, but the timing of its progression throughout the course of diabetes, as well as its association with other mediators, such as sKlotho and other important molecules measured in this study, are critically important for future preventive strategies. Few studies have analysed the changes of both, serum and urinary sKlotho in T1DM. Opposite to the results of the present work, a previous study using a similar rat model but only during 14 days of diabetes, found higher serum sKlotho and lower urine sKlotho and renal Klotho expression [24]. The difference between both studies could be due to the time and degree of renal damage induced by the diabetes. The mentioned study showed that albuminuria was much higher than in the present study (2.57 ± 0.67 mg/24 h vs. 0.67 ± 0.27 mg/24 h), with similar polyuria, suggesting a greater short-term renal damage. Meanwhile, in our long-term study, the renal damage could be considered “mild-moderate” and classified as Class I diabetic nephropathy, according to the Renal Pathology Society [25]. The rat model used in the present study produced a very mild renal function decline, and it could mimic what occurs in humans in the early phases of diabetes. In fact, and according to our results, a decrease in serum sKlotho has been reported in diabetic patients in the early stages of chronic kidney disease (CKD), while increased levels are observed thereafter [22]. The decline in sKlotho in the advanced stages of CKD has been related to the downregulation of Klotho kidney production [24,26]. However, this is not the case in our study, in which the diabetic rats showed almost no decrease in the glomerular filtration rate (GFR), as it occurred in the early stages of CKD, where the kidney involvement is only suspected by the appearance of albuminuria but not due to relevant reductions of the GFR. No association was found between serum and urinary sKlotho levels, suggesting that urinary sKlotho is the result of a complex process [27]. Physiologically, circulating sKlotho passes from the basal to the apical side of the proximal tubular cells through transcytosis process and is eliminated in the urine, but it can also be secreted from tubules cells by sheddases [28]. As it has been previously described in an experimental model of acute kidney damage [29], tubular epithelial cells can lose their brush border membrane leading to an increase in Klotho shedding to the lumen. In the present study, it cannot be established whether hyperglycaemia and the consequent hyperfiltration enhanced Klotho transcytosis; however, the tubular histological alterations and the higher renal gene expression of Adam10 suggest that the elevated urinary sKlotho could be a marker of the diabetes-induced inflammation and tubular damage. Cross-sectional studies in diabetic patients, mainly in type 2 diabetes mellitus (T2DM), have shown that serum sKlotho may play a relevant role in albumin homeostasis; however, only a few of them have found a correlation between serum sKlotho and uACR [20]; in fact, little is known about the possible relationship between urinary sKlotho and the excretion of albumin. In the present study, the diabetic status induced increase in both, urinary sKlotho levels (Figure 1A) and uACR (Figure S1B), throughout the whole duration of the study. This relationship was consistent, furthermore, at the end of the study, a positive significant correlation between these two parameters that persisted after adjustment for diabetes was found (Figure 1C). These results suggest that in the early phases of the diabetic kidney disease, with preserved GFR, urine sKlotho may be an earlier marker of kidney damage than serum sKlotho. However, the role of sKlotho in CKD is very complex and its possible use as an urinary marker of renal function has been poorly studied [23]. Previous studies in T1 and T2DM found a negative correlation between serum sKlotho levels and HbA1c level, one of the most studied glycated proteins and a good marker of diabetes [20,30], suggesting a close relationship between sKlotho and hyperglycaemia. In line with these findings, to the best of our knowledge, our results showed for the first time, a significant positive correlation between urinary sKlotho and urinary fluorescent AGEs, providing additional evidence for the relationship between sKlotho and diabetes. The strength of the correlation between urinary sKlotho and fluorescent AGEs (Figure 1B) suggest that the latter, an easier and quicker method to estimate sKlotho, could be used at least in the early stages of diabetes. The decrease in sKlotho levels and the increase in circulating AGEs due to the hyperglycaemia could play a role in the diabetic complications. It is known that AGEs formation includes many heterogeneous chemical structures which are increased in diabetes and can play a relevant pathophysiological role acting directly or via a receptor-mediated (RAGE) signalling [31]. In addition, it has been shown that serum sRAGE was lower in diabetic rats, despite a higher gene expression of Adam10, also involved in the cleavage of membrane-bound RAGE [32]. Previous studies found increased levels of serum sRAGE in diabetic patients [33,34]; in contrast, other studies found a decreased levels of circulating sRAGE in patients with T1 and T2DM [35,36]. In some cases, this discrepancy could be explained as a consequence of differences in renal function [34]. Analyses of serum and urinary sRAGE levels in the early stages of diabetic nephropathy are scarce. In the present study, serum levels of sRAGE were lower in diabetic rats that could be explained by the increase in sRAGE excretion, similar to that observed with sKlotho. The function of sRAGE is also controversial, it has been postulated that increased sRAGE production could be related to sustained inflammation [37] but there is a broad agreement that it plays a protective role acting as a decoy receptor for AGEs [32,35]. Thus, in diabetic rats, the decrease in sRAGE and the increase in serum AGEs could increase the AGE binding to RAGE, stimulating several fibrotic pathways [6,7]. In the present study, AGE/RAGE pathway activation occurred mainly in the heart, where the expression of RAGE was significantly higher measured at both mRNA and protein levels. In the kidney of diabetic rats, structural changes in tubules and glomerulus were associated with hyperfiltration without changes in fibrosis. By contrast, the hearts of the diabetic rats showed hypertrophy of the cardiomyocytes and increased fibrosis. These observed morphological changes are supported by the high expression of Tgf-β1, fibronectin and the active involvement of the Wnt/β-catenin pathway, which can trigger inhibitory and compensatory signals such as, Dkk1 and Sfrp2 [38,39]. In fact, although Sfrp2 has traditionally been considered a Wnt inhibitor, its activation has been associated with the increase in the activity of β-catenin pathways, which in the heart of diabetic rats, could have triggered the activation of β-catenin, leading to increased myocardial fibrosis [39]. The structural alterations observed in the heart of the diabetic rats could be also associated with the increase in NT-pro-BNP, a biomarker of the myocardial function predictor of cardiovascular events in patients with T2DM [40]. One limitation of the study is that although the diabetic rat models used could resemble important aspects of the human diabetic disease type 1, the fibrotic changes observed in the rodent models of diabetes are mild compared to the extensive and widespread fibrosis found in patients with long-standing diabetes. In summary, the difference in the fibrosis observed in the kidney and heart of the diabetic rats could be related to the Klotho and AGE/RAGE pathway levels. In the kidney—the main source of Klotho—Klotho expression was maintained at the same level as in the control rats and it could have played its known antifibrotic action, repressing the activation and translocation of β-catenin to the nucleus [41], though the increase in the RAGE expression levels was not clear. In the heart, the decrease in serum sKlotho—known to have a negative impact on cardiac function [42,43]—and the increase in RAGE expression, possibly related to the activation of the Wnt/β-catenin signalling pathway, may have promote the fibrotic process and the progression of the cardiac alterations observed in the diabetic rats. ## 4.1. Experimental Model Four month-old male Wistar rats weighing 425 ± 43 g were kept under conventional conditions in the Animal Facility of the University of Oviedo with free access to water and standard food. Experimental procedures were approved by the Ethics Committee for laboratory animals of the Oviedo University. Type 1 Diabetic Model Thirty-four animals, housed three per cage, were randomly divided into two groups receiving either a single intraperitoneal injection of 55 mg/kg BW of freshly prepared STZ (Sigma-Aldrich, St. Louis, MO, USA) in 0.1M citrate buffer pH = 4.5 (Diabetic rats; $$n = 17$$), or citrate buffer alone (Control rats; $$n = 17$$) after 6 h of fasting under light anaesthesia with isoflurane $2\%$. Drinking water was supplemented with $10\%$ sucrose solution for 24 h to prevent hypoglycaemic shock. Tail vein blood glucose and ketone bodies were tested after 24 h of STZ administration, and daily during the six days, using a Freestyle Optimum Neo device (Abbot Diabetes Care, Witney, UK). Subcutaneous long-acting biosynthetic human insulin (Lantus®, Aventis Pharma, Bad Soden, Germany) (1–2 IU) was administered when blood glucose > 500 mg/dL and/or ketone bodies > 3 mmol/L. After one week, the rats who had a blood glucose of >350 mg/dL for three consecutive days were considered “diabetic rats” and included in the study. Glycaemia and BW were monitored twice weekly. Those rats with blood glucose higher than 500 mg/dL (the upper limit of quantification of the glucometer) and significant body weight loss (>$10\%$) received subcutaneous injections of 1–2 IU long-acting biosynthetic human insulin or subcutaneous insulin pellets (0.5 IU/24 h slow-release) (Linshin, Toronto, ON, Canada). From Week 9 to 24, the rats were placed in metabolic cages for 24-h urine collection every 3 weeks and the day before the sacrifice. At the end of the study, the rats were anesthetized with isoflurane and sacrificed by exsanguination. Sections of the hearts and kidneys were weighed and fixed in $4\%$ formaldehyde for histological analyses or stored at −80 °C for mRNA and protein extraction. ## 4.2. Biochemical Analyses The blood was collected in tubes with or without EDTA and centrifuged at 3000 rpm for 15 min at 4 °C to obtain plasma or serum, respectively, and stored at −80 °C until analysis. The urine volume was measured and centrifuged for 5 min at 2500 rpm and the clear supernatant stored at −80 °C. Serum and urinary glucose, calcium, creatinine, total proteins, albumin, and serum urea were measured using a multi-channel auto-analyser (Hitachi 717; Boehringer Mannheim, Mannheim, Germany). Commercial enzyme-linked immunosorbent assay (ELISA) kits were performed to analysed the levels of AGEs (CSB-E09413r, Cusabio, Houston, TX, USA), sRAGE (MBS029347, MyBioSource, San Diego, TX, USA), and NT-proBNP (CSB-E08752r, Cusabio, Houston, TX, USA) in serum/plasma and/or urine, following manufacturer’s protocols. Fluorescent urine AGEs were also measured by a fluorometric measurement method, following a previously described procedure [44]. Serum sKlotho was measured by an immunoprecipitation–immunoblot assay at the O’Brien Kidney Research Centre in UT Southwestern, and urinary sKlotho was determined at our laboratory following the protocol of the O’Brien Kidney Research Center [45], briefly summarized below in the immunoblotting section. Urinary biomarkers were showed as a ratio of the urinary creatinine. Creatinine clearance and uACR were calculated using the following formulas: creatinine clearance (mL/min) = urinary creatinine (mg/dL) × urinary volume (mL)/serum creatinine (mg/dL) × 1440 (min); uACR (mg/g) = urinary albumin (mg/L)/urinary creatinine (g/L). Blood Hb1Ac was measured at sacrifice using A1Cnow+® (PTS Diagnostics, Indianapolis, IN, USA). ## 4.3. Histological Analyses The hearts and kidneys were formalin-fixed and paraffin-embedded. Tissue sections of 5 µm were stained with Picrosirius Red and periodic acid–Schiff following standard protocols to assess the degree of fibrosis, as well as cardiomyocyte and proximal tubules sizes. For this purpose, 5 and 10 random images were acquired under a light microscope (DMRXA2, Leica Microsystems, Wetzlar, Germany) equipped with a Leica DFC7000 T camera (Leica Microsystems). ImageJ software was used for treatment and analyses of acquired images. Kidney fibrosis was assessed by measuring the percentage of tissue area stained by Picrosirius Red divided by total tissue area at a magnification of 10× excluding perivascular fibrosis and kidney glomeruli. The length of the short axis of proximal tubules in the glomerulus was also measured at a magnification of 20×. Cardiomyocyte hypertrophy was assessed by measuring the cross-sectional widths of cardiomyocytes drawing a line that crossed the center of the nucleus at a magnification of 40×. A minimum of 150 proximal tubules and cardiomyocytes were measured by tissue section. For electron microscopy analysis, 7 kidney samples (3 controls and 4 diabetic rats) were randomly selected, the pieces (3 mm diameter) were immediately fixed in glutaraldehyde at room temperature for 3 h. After processing and embedding, the semithin (1 μM) sections were stained with toluidine blue for glomerular location. Ultrathin (200 Å) sections were collected, stained with uranyl acetate and lead citrate, and examined in a JEOL 1011 transmission electron microscope. ## 4.4. RNA Extraction and Quantitative Real-Time PCR Total RNA was extracted from the kidneys and hearts using TRI-Reagent (Sigma-Aldrich, St. Louis, MO, USA). cDNA was then synthesized from 1 µg of total RNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Waltham, MA, USA). To measure Dickkopf-related protein 1 (Dkk1) (Rn01501537, Thermo Fisher Scientific, Waltham, MA, USA), sclerostin (Sost) (Rn00577971), secreted frizzled-related protein 2 (Sfrp2) (Rn01458837), Sfrp4 (Rn00585549), fibronectin (Rn00569575), transforming growth factor beta 1 (TGF-β1) (Rn00572010), receptor for advanced glycation end products (Rage) (Rn00584249), a disintegrin and metalloproteinase domain-containing protein 10 (Adam10) (Rn01530753), and Klotho (Rn00580123) mRNA levels quantitative real-time PCR was used. TaqMan Universal PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA) was used for the amplification of target genes according to the manufacturer’s protocol in a QuantStudio 3 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) (Rn99999916) was used for normalization. The ∆∆CT method was used to quantify the relative expression of each gene [46]. ## 4.5. Immunoblotting Kidney and heart tissues were homogenized in a RIPA buffer, and the total protein content was measured by the DC protein assay reagents (Bio-Rad, Hercules, CA, USA). Proteins were separated by sodium dodecyl sulphate–polyacrylamide gel ($8\%$) electrophoresis (SDS-PAGE). Urine samples were loaded into commercial NuPAGE 4–$12\%$ Bis-Tris gels (Thermo Fisher Scientific, Waltham, MA, USA). In both cases, the proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (Amersham Hybond, Amersham Biosciences, Amersham, UK). Blotting efficiency was checked by Ponceau red dyeing (Sigma-Aldrich, St. Louis, MO, USA). The membranes were incubated with specific antibodies following the manufacturer’s instructions: Non-phospho (Active) β-Catenin (#8814, dilution 1:1000; Cell Signaling, Danvers, MA, USA), DKK1 (MAB1765, dilution 1:500; R&D Systems, Minneapolis, MN, USA), RAGE (PA5-78736, 0.5 µg/mL; Thermo Fisher Scientific, Waltham, MA, USA) and Klotho (#KO603, dilution 1:1250; Trans Genic Inc., Chuo-ku, Japan). Anti-rabbit, anti-rat, anti-goat, or anti-mouse (Santa Cruz, Dallas, TX, USA) were used as secondary antibodies detected with the ECL Western Blotting Detection Kit (Bio-Rad) and the ChemiDoc Gel Imaging System Model XRS (Bio-Rad), and were quantified using Quantity One 1-D Analysis Software Version 4 (Bio-Rad). All blots were rehybridized with glyceraldehyde-3-phosphate dehydrogenase (GAPDH; dilution 1:3000; Santa Cruz Biotechnology) for normalization. ## 4.6. Statistical Analysis Results were expressed as median and interquartile ranges. The differences between groups were assessed using the non-parametric Wilcoxon rank sum test. Linear regression was used to assess the correlation between continuous variables. Statistically significant differences were considered when $p \leq 0.05.$ The statistical analysis was carried out using R software for windows (Version 4.1.2). ## References 1. International Diabetes Federation IDF Diabetes Atlas 2. Birkeland K.I., Bodegard J., Eriksson J.W., Norhammar A., Haller H., Linssen G.C.M., Banerjee A., Thuresson M., Okami S., Garal-Pantaler E.. Heart failure and chronic kidney disease manifestation and mortality risk associations in type 2 diabetes: A large multinational cohort study. Diabetes Obes. Metab. (2020.0) 22 1607-1618. DOI: 10.1111/dom.14074 3. Kristofi R., Bodegard J., Norhammar A., Thuresson M., Nathanson D., Nystrom T., Birkeland K.I., Eriksson J.W.. Cardiovascular and Renal Disease Burden in Type 1 Compared With Type 2 Diabetes: A Two-Country Nationwide Observational Study. Diabetes Care (2021.0) 44 1211-1218. DOI: 10.2337/dc20-2839 4. Salvatore T., Pafundi P.C., Galiero R., Albanese G., Di Martino A., Caturano A., Vetrano E., Rinaldi L., Sasso F.C.. The Diabetic Cardiomyopathy: The Contributing Pathophysiological Mechanisms. Front. Med. (2021.0) 8 695792. DOI: 10.3389/fmed.2021.695792 5. Zeng L.F., Xiao Y., Sun L.. A Glimpse of the Mechanisms Related to Renal Fibrosis in Diabetic Nephropathy. Adv. Exp. Med. Biol. (2019.0) 1165 49-79. DOI: 10.1007/978-981-13-8871-2_4 6. Zhao J., Randive R., Stewart J.A.. Molecular mechanisms of AGE/RAGE-mediated fibrosis in the diabetic heart. World J. Diabetes (2014.0) 5 860-867. DOI: 10.4239/wjd.v5.i6.860 7. Matsui T., Higashimoto Y., Nishino Y., Nakamura N., Fukami K., Yamagishi S.I.. RAGE-Aptamer Blocks the Development and Progression of Experimental Diabetic Nephropathy. Diabetes (2017.0) 66 1683-1695. DOI: 10.2337/db16-1281 8. Yang R., Jia Q., Ma S.F., Wang Y., Mehmood S., Chen Y.. Exogenous H2S mitigates myocardial fibrosis in diabetic rats through suppression of the canonical Wnt pathway. Int. J. Mol. Med. (2019.0) 44 549-558. DOI: 10.3892/ijmm.2019.4237 9. Rooney B., O′Donovan H., Gaffney A., Browne M., Faherty N., Curran S.P., Sadlier D., Godson C., Brazil D.P., Crean J.. CTGF/CCN2 activates canonical Wnt signalling in mesangial cells through LRP6: Implications for the pathogenesis of diabetic nephropathy. FEBS Lett. (2011.0) 585 531-538. DOI: 10.1016/j.febslet.2011.01.004 10. Hu H.H., Cao G., Wu X.Q., Vaziri N.D., Zhao Y.Y.. Wnt signaling pathway in aging-related tissue fibrosis and therapies. Ageing Res. Rev. (2020.0) 60 101063. DOI: 10.1016/j.arr.2020.101063 11. Lindberg K., Amin R., Moe O.W., Hu M.C., Erben R.G., Ostman Wernerson A., Lanske B., Olauson H., Larsson T.E.. The kidney is the principal organ mediating klotho effects. J. Am. Soc. Nephrol. JASN (2014.0) 25 2169-2175. DOI: 10.1681/ASN.2013111209 12. Kuro O.M.. The Klotho proteins in health and disease. Nat. Rev. Nephrol. (2019.0) 15 27-44. DOI: 10.1038/s41581-018-0078-3 13. Hu M.C., Kuro-o M., Moe O.W.. Renal and extrarenal actions of Klotho. Semin. Nephrol. (2013.0) 33 118-129. DOI: 10.1016/j.semnephrol.2012.12.013 14. Hu M.C., Shi M., Gillings N., Flores B., Takahashi M., Kuro O.M., Moe O.W.. Recombinant alpha-Klotho may be prophylactic and therapeutic for acute to chronic kidney disease progression and uremic cardiomyopathy. Kidney Int. (2017.0) 91 1104-1114. DOI: 10.1016/j.kint.2016.10.034 15. Liu H., Fergusson M.M., Castilho R.M., Liu J., Cao L., Chen J., Malide D., Rovira I.I., Schimel D., Kuo C.J.. Augmented Wnt signaling in a mammalian model of accelerated aging. Science (2007.0) 317 803-806. DOI: 10.1126/science.1143578 16. Zhao Y., Wang C., Hong X., Miao J., Liao Y., Hou F.F., Zhou L., Liu Y.. Wnt/beta-catenin signaling mediates both heart and kidney injury in type 2 cardiorenal syndrome. Kidney Int. (2019.0) 95 815-829. DOI: 10.1016/j.kint.2018.11.021 17. Donate-Correa J., Martin-Nunez E., Mora-Fernandez C., Muros-de-Fuentes M., Perez-Delgado N., Navarro-Gonzalez J.F.. Klotho in cardiovascular disease: Current and future perspectives. World J. Biol. Chem. (2015.0) 6 351-357. DOI: 10.4331/wjbc.v6.i4.351 18. Bi X., Yang K., Zhang B., Zhao J.. The Protective Role of Klotho in CKD-Associated Cardiovascular Disease. Kidney Dis. (2020.0) 6 395-406. DOI: 10.1159/000509369 19. Oh H.J., Nam B.Y., Wu M., Kim S., Park J., Kang S., Park J.T., Yoo T.H., Kang S.W., Han S.H.. Klotho plays a protective role against glomerular hypertrophy in a cell cycle-dependent manner in diabetic nephropathy. Am. J. Physiol. Ren. Physiol. (2018.0) 315 F791-F805. DOI: 10.1152/ajprenal.00462.2017 20. Wu C., Wang Q., Lv C., Qin N., Lei S., Yuan Q., Wang G.. The changes of serum sKlotho and NGAL levels and their correlation in type 2 diabetes mellitus patients with different stages of urinary albumin. Diabetes Res. Clin. Pr. (2014.0) 106 343-350. DOI: 10.1016/j.diabres.2014.08.026 21. Inci A., Sari F., Coban M., Olmaz R., Dolu S., Sarikaya M., Yilmaz N.. Soluble Klotho and fibroblast growth factor 23 levels in diabetic nephropathy with different stages of albuminuria. J. Investig. Med. (2016.0) 64 1128-1133. DOI: 10.1136/jim-2016-000142 22. Kacso I.M., Bondor C.I., Kacso G.. Soluble serum Klotho in diabetic nephropathy: Relationship to VEGF-A. Clin. Biochem. (2012.0) 45 1415-1420. DOI: 10.1016/j.clinbiochem.2012.07.098 23. Yu L.X., Li S.S., Sha M.Y., Kong J.W., Ye J.M., Liu Q.F.. The controversy of klotho as a potential biomarker in chronic kidney disease. Front. Pharmacol. (2022.0) 13 931746. DOI: 10.3389/fphar.2022.931746 24. Typiak M., Kulesza T., Rachubik P., Rogacka D., Audzeyenka I., Angielski S., Saleem M.A., Piwkowska A.. Role of Klotho in Hyperglycemia: Its Levels and Effects on Fibroblast Growth Factor Receptors, Glycolysis, and Glomerular Filtration. Int. J. Mol. Sci. (2021.0) 22. DOI: 10.3390/ijms22157867 25. Tervaert T.W., Mooyaart A.L., Amann K., Cohen A.H., Cook H.T., Drachenberg C.B., Ferrario F., Fogo A.B., Haas M., de Heer E.. Pathologic classification of diabetic nephropathy. J. Am. Soc. Nephrol. JASN (2010.0) 21 556-563. DOI: 10.1681/ASN.2010010010 26. Sakan H., Nakatani K., Asai O., Imura A., Tanaka T., Yoshimoto S., Iwamoto N., Kurumatani N., Iwano M., Nabeshima Y.. Reduced renal alpha-Klotho expression in CKD patients and its effect on renal phosphate handling and vitamin D metabolism. PLoS ONE (2014.0) 9. DOI: 10.1371/journal.pone.0086301 27. Navarro-Gonzalez J.F., Sanchez-Nino M.D., Donate-Correa J., Martin-Nunez E., Ferri C., Perez-Delgado N., Gorriz J.L., Martinez-Castelao A., Ortiz A., Mora-Fernandez C.. Effects of Pentoxifylline on Soluble Klotho Concentrations and Renal Tubular Cell Expression in Diabetic Kidney Disease. Diabetes Care (2018.0) 41 1817-1820. DOI: 10.2337/dc18-0078 28. Hu M.C., Shi M., Zhang J., Addo T., Cho H.J., Barker S.L., Ravikumar P., Gillings N., Bian A., Sidhu S.S.. Renal Production, Uptake, and Handling of Circulating alphaKlotho. J. Am. Soc. Nephrol. JASN (2016.0) 27 79-90. DOI: 10.1681/ASN.2014101030 29. Qian Y., Guo X., Che L., Guan X., Wu B., Lu R., Zhu M., Pang H., Yan Y., Ni Z.. Klotho Reduces Necroptosis by Targeting Oxidative Stress Involved in Renal Ischemic-Reperfusion Injury. Cell. Physiol. Biochem. (2018.0) 45 2268-2282. DOI: 10.1159/000488172 30. Zubkiewicz-Kucharska A., Wikiera B., Noczynska A.. Soluble Klotho Is Decreased in Children With Type 1 Diabetes and Correlated With Metabolic Control. Front. Endocrinol. (Lausanne) (2021.0) 12 709564. DOI: 10.3389/fendo.2021.709564 31. Mengstie M.A., Chekol Abebe E., Behaile Teklemariam A., Tilahun Mulu A., Agidew M.M., Teshome Azezew M., Zewde E.A., Agegnehu Teshome A.. Endogenous advanced glycation end products in the pathogenesis of chronic diabetic complications. Front. Mol. Biosci. (2022.0) 9 1002710. DOI: 10.3389/fmolb.2022.1002710 32. Erusalimsky J.D.. The use of the soluble receptor for advanced glycation-end products (sRAGE) as a potential biomarker of disease risk and adverse outcomes. Redox Biol. (2021.0) 42 101958. DOI: 10.1016/j.redox.2021.101958 33. Prasad K.. Is there any evidence that AGE/sRAGE is a universal biomarker/risk marker for diseases?. Mol. Cell. Biochem. (2019.0) 451 139-144. DOI: 10.1007/s11010-018-3400-2 34. Tan K.C., Shiu S.W., Chow W.S., Leng L., Bucala R., Betteridge D.J.. Association between serum levels of soluble receptor for advanced glycation end products and circulating advanced glycation end products in type 2 diabetes. Diabetologia (2006.0) 49 2756-2762. DOI: 10.1007/s00125-006-0394-1 35. Tsoporis J.N., Hatziagelaki E., Gupta S., Izhar S., Salpeas V., Tsiavou A., Rigopoulos A.G., Triantafyllis A.S., Marshall J.C., Parker T.G.. Circulating Ligands of the Receptor for Advanced Glycation End Products and the Soluble Form of the Receptor Modulate Cardiovascular Cell Apoptosis in Diabetes. Molecules (2020.0) 25. DOI: 10.3390/molecules25225235 36. Katakami N., Matsuhisa M., Kaneto H., Matsuoka T.A., Sakamoto K., Nakatani Y., Ohtoshi K., Hayaishi-Okano R., Kosugi K., Hori M.. Decreased endogenous secretory advanced glycation end product receptor in type 1 diabetic patients: Its possible association with diabetic vascular complications. Diabetes Care (2005.0) 28 2716-2721. DOI: 10.2337/diacare.28.11.2716 37. Antonelli A., Di Maggio S., Rejman J., Sanvito F., Rossi A., Catucci A., Gorzanelli A., Bragonzi A., Bianchi M.E., Raucci A.. The shedding-derived soluble receptor for advanced glycation endproducts sustains inflammation during acute Pseudomonas aeruginosa lung infection. Biochim. Biophys. Acta–Gen. Subj. (2017.0) 1861 354-364. DOI: 10.1016/j.bbagen.2016.11.040 38. Akhmetshina A., Palumbo K., Dees C., Bergmann C., Venalis P., Zerr P., Horn A., Kireva T., Beyer C., Zwerina J.. Activation of canonical Wnt signalling is required for TGF-beta-mediated fibrosis. Nat. Commun. (2012.0) 3 735. DOI: 10.1038/ncomms1734 39. Lin H., Angeli M., Chung K.J., Ejimadu C., Rosa A.R., Lee T.. sFRP2 activates Wnt/beta-catenin signaling in cardiac fibroblasts: Differential roles in cell growth, energy metabolism, and extracellular matrix remodeling. Am. J. Physiol. Cell Physiol. (2016.0) 311 C710-C719. DOI: 10.1152/ajpcell.00137.2016 40. Malachias M.V.B., Jhund P.S., Claggett B.L., Wijkman M.O., Bentley-Lewis R., Chaturvedi N., Desai A.S., Haffner S.M., Parving H.H., Prescott M.F.. NT-proBNP by Itself Predicts Death and Cardiovascular Events in High-Risk Patients With Type 2 Diabetes Mellitus. J. Am. Heart Assoc. (2020.0) 9 e017462. DOI: 10.1161/JAHA.120.017462 41. Zhou L., Li Y., Zhou D., Tan R.J., Liu Y.. Loss of Klotho contributes to kidney injury by derepression of Wnt/beta-catenin signaling. J. Am. Soc. Nephrol. JASN (2013.0) 24 771-785. DOI: 10.1681/ASN.2012080865 42. Chen K., Wang S., Sun Q.W., Zhang B., Ullah M., Sun Z.. Klotho Deficiency Causes Heart Aging via Impairing the Nrf2-GR Pathway. Circ. Res. (2021.0) 128 492-507. DOI: 10.1161/CIRCRESAHA.120.317348 43. Memmos E., Sarafidis P., Pateinakis P., Tsiantoulas A., Faitatzidou D., Giamalis P., Vasilikos V., Papagianni A.. Soluble Klotho is associated with mortality and cardiovascular events in hemodialysis. BMC Nephrol. (2019.0) 20. DOI: 10.1186/s12882-019-1391-1 44. Perrone A., Giovino A., Benny J., Martinelli F.. Advanced Glycation End Products (AGEs): Biochemistry, Signaling, Analytical Methods, and Epigenetic Effects. Oxidative Med. Cell. Longev. (2020.0) 2020 3818196. DOI: 10.1155/2020/3818196 45. Barker S.L., Pastor J., Carranza D., Quinones H., Griffith C., Goetz R., Mohammadi M., Ye J., Zhang J., Hu M.C.. The demonstration of alphaKlotho deficiency in human chronic kidney disease with a novel synthetic antibody. Nephrol. Dial. Transplant. Off. Publ. Eur. Dial. Transpl. Assoc. Eur. Ren. Assoc. (2015.0) 30 223-233. DOI: 10.1093/ndt/gfu291 46. Livak K.J., Schmittgen T.D.. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods (2001.0) 25 402-408. DOI: 10.1006/meth.2001.1262
--- title: 'Prevalence of Dementia among Patients Hospitalized with Type 2 Diabetes Mellitus in Spain, 2011–2020: Sex-Related Disparities and Impact of the COVID-19 Pandemic' authors: - Ana Lopez-de-Andres - Rodrigo Jimenez-Garcia - Jose J. Zamorano-Leon - Ricardo Omaña-Palanco - David Carabantes-Alarcon - Valentin Hernández-Barrera - Javier De Miguel-Diez - Natividad Cuadrado-Corrales journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049429 doi: 10.3390/ijerph20064923 license: CC BY 4.0 --- # Prevalence of Dementia among Patients Hospitalized with Type 2 Diabetes Mellitus in Spain, 2011–2020: Sex-Related Disparities and Impact of the COVID-19 Pandemic ## Abstract [1] Background: To assess changes in the prevalence of dementia among patients hospitalized with type 2 diabetes (T2DM), to analyze the effects of dementia on in-hospital mortality (IHM) in this population, to evaluate sex differences, and to determine the impact of the COVID-19 pandemic on these parameters. [ 2] Methods: We used a nationwide discharge database to select all patients with T2DM aged 60 years or over admitted to Spanish hospitals from 2011 to 2020. We identified those with all-cause dementia, Alzheimer’s disease (AD), and vascular dementia (VaD). The effect of sex, age, comorbidity, and COVID-19 on the prevalence of dementia subtypes and on IHM was assessed using multivariable logistic regression. [ 3] Results: We identified 5,250,810 hospitalizations with T2DM. All-cause dementia was detected in $8.31\%$, AD in $3.00\%$, and VaD in $1.55\%$. The prevalence of all subtypes of dementia increased significantly over time. After multivariable adjustment, higher values were observed in women for all-cause dementia (OR 1.34; $95\%$ CI 1.33–1.35), AD (OR 1.6; $95\%$ CI 1.58–1.62), and VaD (OR 1.12; $95\%$ CI 1.11–1.14). However, female sex was a protective factor for IHM in patients with all-cause dementia (OR 0.90; $95\%$ CI 0.89–0.91), AD (OR 0.89; $95\%$ CI 0.86–0.91), and VaD (OR 0.95; $95\%$ CI 0.91–0.99). IHM among patients with dementia remained stable over time, until 2020, when it increased significantly. Higher age, greater comorbidity, and COVID-19 were associated with IHM in all dementia subtypes. [ 4] Conclusions: The prevalence of dementia (all-cause, AD, and VaD) in men and women with T2DM increased over time; however, the IHM remained stable until 2020, when it increased significantly, probably because of the COVID-19 pandemic. The prevalence of dementia is higher in women than in men, although female sex is a protective factor for IHM. ## 1. Introduction Recent years have seen an increase in the incidence of neurodegenerative diseases such as dementia, partially owing to the increase in life expectancy and in the prevalence of type 2 diabetes mellitus (T2DM) [1,2]. Dementia is a highly prevalent progressive disorder, and according to the World Health Organization, it is estimated that almost 10 million new cases will appear each year [3]. Compared with persons who do not have diabetes, those with T2DM are 1.5 to 2.5 times more likely to develop all-cause dementia, including its two main subtypes, Alzheimer’s disease (AD) and vascular dementia (VaD) [4,5]. This increased risk in patients with T2DM translates into an average earlier onset of dementia of 2.5 years compared with patients without T2DM [6]. The complex pathophysiology underlying this relationship may involve hyperglycemia, insulin resistance, neuroinflammation, and altered energy homeostasis [4,5]. Given the association between the aging global population and the rising prevalence of dementia, the link between T2DM and dementia constitutes a global public health concern [1,2,7]. Diabetes is a major comorbidity among patients hospitalized with COVID-19 [8]. Dementia can increase the risk of a poorer COVID-19 outcome. In the US, a study using a nationwide database found that dementia was associated with COVID-19-related hospitalization in patients with T2DM (OR 2.07; $95\%$ CI 1.79–2.39). However, the authors did not find dementia to be associated with in-hospital mortality (IHM) during COVID-19-related hospitalization in patients with T2DM (OR 0.85; $95\%$ CI 0.71–1.02) [9]. Sex differences may play a critical role in the incidence and outcomes of hospitalizations in patients with T2DM and dementia. Women with T2DM have a higher excess risk of cognitive decline and vascular dementia than men with T2DM, although the extent of these differences depends on the characteristics of the study populations and the methods used [10]. In any case, it remains unclear why men and women with T2DM are affected differently by dementia [11]. Therefore, the objectives of the study were to assess changes in the prevalence of dementia (all-cause, AD, and VaD) in hospitalized patients with T2DM in Spain between 2011 and 2020. We evaluated sex differences in the prevalence, clinical characteristics, and IHM of dementia between men and women with T2DM. We also analyzed which variables were associated with IHM among patients with T2DM and dementia. Finally, we assessed whether the COVID-19 pandemic affected the prevalence of dementia and IHM in patients hospitalized with T2DM in the year 2020. ## 2.1. Design and Data Source A retrospective, population-based observational study was conducted using the Spanish National Hospital Discharge Database (RAE-CMBD, Registro de Actividad de Atención Especializada-Conjunto Mínimo Básico de Datos [Register of Specialized Care–Basic Minimum Database]). A description of the RAE-CMBD methodology is available online [12]. The study period ran from 1 January 2011 to 31 December 2020. The RAE-CMBD collects age, sex, dates of admission and discharge, discharge destination (home, deceased, social institution, or voluntary discharge), up to 20 diagnoses, and 20 procedures conducted during hospitalization in either public and private hospitals. The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) was used for coding between 2011 and 2015, and the International Classification of Disease, Tenth Revision (ICD10) has been used since 2016. ## 2.2. Study Population and Study Variables The study population included patients aged ≥60 years with a T2DM code in any di-agnostic position (see ICD codes in Table S1). Patients with T1DM and data missing for sex, age, dates of admission and discharge, or discharge destination were excluded. To respond to the objectives of the study, the study population was stratified according to the presence of ICD codes for dementia (all-cause dementia, AD, and VaD) in any diagnostic position in the RAE-CMBD (Table S1). All analyses were subsequently stratified according to sex. The presence of comorbidity was assessed using the Charlson Comorbidity Index (CCI), excluding diabetes and dementia, and the ICD codes described by Sundararajan et al. [ 13] and Quan et al. [ 14]. Likewise, regardless of the diagnostic position, the presence of COVID-19 was evaluated (see ICD10 codes in Table S1) in the year 2020. Regarding hospital outcomes, we analyzed IHM, which was defined as the number of patients who died in hospital each year divided by the total number of hospitalizations that year. ## 2.3. Statistical Analysis We calculated the total prevalence of all-cause dementia, AD, and VaD in patients with T2DM according to year, sex, and age group. The results of the descriptive statistical analysis are expressed as total frequencies with percentages for categorical variables and means with standard deviations for continuous variables. The trend was analyzed using the Cochran–Mantel–Haenszel statistic or Cochran–Armitage test in the case of categorical variables and a linear regression t test or Jonckheere–Terpstra test in the case of continuous variables. Categorical variables were compared using the Fisher exact test. Continuous variables were compared using the t test. We used multivariable logistic regression to analyze factors associated with the presence of all-cause dementia, AD, and VaD, considering the effect of sex and the other study covariates. We also identified the variables associated with IHM in men and women with T2DM and all-cause dementia, AD, and VaD. We used the “enter modelling” method for logistic regression. This included five consecutive steps. First, a bivariate analysis of each variable. Second, the selection of variables for the multivariable analysis, with those with a p value of <0.10 being considered. Third, the contribution to the model of each variable was verified using the Wald statistic Forth, as variables were progressively included, the new model generated was compared to the previous one using the likelihood-ratio test. Finally, once the final model was obtained, we checked for linearity and possible interactions between variables. The results of these models are shown with the odds ratio (OR) and $95\%$ confidence intervals (CI). We used Stata version 14 to perform the statistical analysis (Stata, College Station, TX, USA). Statistical significance was set at $p \leq 0.05$ (2-tailed). ## 2.4. Ethics Statement To carry out this study, it was not necessary to request the informed consent of the patients or approval by an ethics committee, since the RAE-CMBD is an administrative database, and all personal data are anonymized. Any investigator can freely request RAE-CMBD data from the Spanish Ministry of Health [15]. ## 3. Results Between 2011 and 2020 in Spain, there were 5,250,810 hospitalizations of patients aged ≥60 years presenting a diagnosis code corresponding to T2DM. Of these, $8.31\%$ ($$n = 436$$,533) had an all-cause dementia code, $3.00\%$ ($$n = 157$$,674) had an AD code, and $1.55\%$ ($$n = 81$$,146) had a VaD code. ## 3.1. Time Trends in the Prevalence of Dementia in Patients with T2DM The prevalence of all-cause dementia among hospitalized patients with T2DM in Spain increased significantly between 2011 and 2020 ($6.9\%$ vs. $10.6\%$; $p \leq 0.001$). Likewise, the prevalence of AD and of VaD rose significantly throughout the study period ($2.71\%$ and $1.55\%$ in 2011 vs. $3.28\%$ and $1.63\%$ in 2020; all $p \leq 0.001$, respectively). As can be seen in Table 1, in patients presenting with all-cause dementia, AD, and VaD, the proportion of men increased over time, while that of women decreased (all $p \leq 0.001$). The mean age of patients with dementia (all-cause, AD, and VaD), as well as an associated comorbidity (CCI), increased significantly over the study period. In the three diseases under study, IHM remained stable between 2011 and 2019. However, in 2020, the IHM increased by around three percentage points, reaching $18.72\%$ for all-cause dementia, $18.85\%$ for AD, and $16.91\%$ for VaD (Table 1). ## 3.2. Sex Differences in the Prevalence and Characteristics of Dementia among Patients with T2DM As can be seen in Table 2, the prevalence of all-cause dementia was higher in women than in men with T2DM for all the study years, increasing significantly between 2011 and 2020 in both groups ($4.91\%$ to $8.4\%$ in men; $9.19\%$ to $19.97\%$ in women: $p \leq 0.001$). Women with all-cause dementia were older than men, although they had a lower CCI (Table 2). Age and concomitant comorbidity increased significantly between 2011 and 2020 in both men and women. IHM in men and women with all-cause dementia increased significantly over the study period ($16.03\%$ and $14.76\%$ in 2011 to $19.97\%$ and $17.67\%$ in 2020, respectively). As shown in Table 3, the prevalence of AD increased significantly between 2011 and 2020 in both men ($1.74\%$ to $2.07\%$; $p \leq 0.001$) and women ($3.82\%$ to $4.92\%$; $p \leq 0.001$), although it was higher in women for all the years studied. The distribution by age and comorbidity shows the same trend as in T2DM patients admitted with all-cause dementia. Between 2011 and 2020, the IHM in T2DM patients with AD increased significantly (from $15.6\%$ to $20.35\%$ in men and from $13.88\%$ to $18\%$ in women) (Table 3). The prevalence of VaD increased throughout the study period (Table 4) and was higher in women than in men. The distribution by age, comorbidity (expressed as the mean of the CCI), and IHM shows the same pattern as that described in the previous types of dementia. ## 3.3. Variables Associated with Sex Differences in the Prevalence and Characteristics of Dementia among Patients with T2DM Table 5 shows the results of the multivariable analysis to identify the factors associated with the presence of all-cause dementia and with IHM in men and women with T2DM and all-cause dementia. The presence of more comorbid conditions and the year of hospital admission (specifically the years 2013, 2014, and 2015 [reference year 2011]) were associated with a lower probability of presenting a code for all-cause dementia. However, the presence of all-cause dementia increased significantly between 2016 and 2020 in both men and women with T2DM. Older age and the presence of COVID-19 were also associated with a higher probability of presenting a code for all-cause dementia in both sexes. After adjusting for covariates, in the entire T2DM population, women were 1.34-fold more likely to have a code for all-cause dementia in their discharge report than men (OR 1.34; $95\%$ CI 1.33–1.35) (Table 5). Regarding IHM, older age, greater comorbidity, and COVID-19 increased the risk of dying during hospitalization in men and women with T2DM and all-cause dementia (Table 5). In the entire study population, being a woman was associated with lower IHM (OR 0.90; $95\%$ CI 0.89–0.91). Table S2 and Table S3 show the results for trends in the presence of and factors associated with IHM for patients with AD and VaD. The prevalence of AD and VaD increased significantly with older age and COVID-19. However, the presence of comorbid conditions was associated with a lower presence of AD (OR 1.6; $95\%$ CI 1.58–1.62) and with a higher presence of VaD (OR 0.66; $95\%$ CI 0.65–0.66). Using the year 2011 as a reference, the presence of codes for AD increased significantly between the years 2014 and 2020, and the prevalence of VaD decreased between the years 2013 and 2018 and in the year 2020. Furthermore, women were 1.6-fold and 1.12-fold more likely to have a code for AD and VaD than men (OR 1.6; $95\%$ CI 1.58–1.62 and OR 1.12; $95\%$ CI 1.11–1.14, respectively). The factors associated with all-cause dementia were also associated with an increased risk of IHM in patients hospitalized with T2DM and AD and VaD. These included older age, higher CCI, and COVID-19. As found for all-cause dementia, female sex was associated with a lower IHM for AD (OR 0.89; $95\%$ CI 0.86–0.91) and VaD (OR 0.95; $95\%$ CI 0.91–0.99). ## 4. Discussion The results obtained in this nationwide retrospective study of over 5 million patients with T2DM aged ≥60 years admitted to Spanish hospitals between 2011 and 2020 revealed several key findings. First, an increase in the prevalence of dementia (all-cause, AD, and VaD) was observed in men and women with T2DM between 2011 and 2020. Second, the prevalence of all-cause dementia was 1.34 times higher in women than in men with T2DM. Third, the presence of COVID-19 increased the risk of IHM in men and women with T2DM and any dementia subtype. Finally, we found that women with T2DM and all-cause dementia, AD, and VaD had a lower risk of dying in hospital than T2DM men. Several epidemiological studies report an increase in the prevalence of dementia in individuals with T2DM over time [16,17,18]. Our results confirm this trend, which has been reported elsewhere, both in AD and in VaD [17,18]. In the United Kingdom general population, using data from over 13 million individuals aged ≥18 years receiving primary care and recorded in the Health Improvement Network database, the overall prevalence of dementia among people with diabetes increased from $0.42\%$ in year 2000 to 2.51 % in 2016 ($p \leq 0.001$). The prevalence of dementia in women patients with diabetes was approximately 1.5 times higher compared to men patients with diabetes [16]. A study of persons aged 90 or over with dementia in Finland showed that the prevalence of diabetes doubled between 2000 and 2018 ($p \leq 0.001$) [19]. The increase in the prevalence of dementia in people with diabetes and vice versa over time, can be explained by greater survival, life expectancy, and improvements in the diagnoses of both pathologies [16,17,18,19]. As expected, we found that between 2011 and 2020, both patient age and the proportion of men with TDM2 and dementia increased. In addition, patients presented greater comorbidity over time. Different studies have described an increase in the prevalence of the main age-related chronic conditions in the general elderly population [20,21,22,23,24]. Furthermore, a recent population-based registry-based study on comorbidity trends during the last years of life in Finnish patients with dementia aged 70 years or older found an increase in comorbidities between 2001 and 2013 [25]. Another cohort study of 245,483 participants showed that older adults with multiple comorbid conditions had a higher risk of dementia [26]. Other factors reported as being relevant include aging, polypharmacy, and a heavier treatment burden, all of which might affect the brain and cause neural injuries [27,28]. The prevalence of dementia is higher in women with diabetes than in men, in terms of all-cause dementia, AD, and VaD [10,11]. A recent meta-analysis found that women with diabetes had a $19\%$ higher risk for VaD than men with diabetes (RR 1.19; $95\%$ CI 1.08–1.30) [11]. Various factors seem to contribute to the difference between men and women regarding dementia. In the case of diabetes, it has been reported that women with T2DM achieve glycemic and cardiovascular targets less frequently and are screened less frequently for the complications of diabetes. In the case of dementia, women are generally referred later than men and experience delays in receiving adequate supportive care when they have cognitive impairment [29]. Alternative explanations include hormonal aspects. Exposure to endogenous estradiol in females has been reported to increase the risk of dementia, especially in the presence of diabetes [30]. Moreover, since the female patients in this study were post-menopausal (age ≥60 years), alterations in sex hormones could play a role [31]. Gong et al. reported that mental health symptoms (depression and anxiety) and higher waist circumference have been found associated with a greater risk of dementia in women with T2DM in comparison with men [32]. A plausible explanation for this finding is that women are more likely to be prescribed with pharmacological treatments for depression, and the use of antidepressants has been linked to a greater risk of dementia. Whether the different body composition and fat distribution observed in women and men with diabetes, partially driven by the influence of sex hormones on visceral obesity, can explain the sex differences in obesity and dementia, requires further investigation [32]. Studies conducted in the general population have suggested that factors such as blood pressure, physical activity, longer education, and former alcohol use have a different effect in men and women on the risk of developing dementia [33,34,35,36]. It is possible that combinations or patterns of risk factors explain sex differences in cognitive decline and the subsequent risk of dementia. Therefore, a multi-domain approach to understanding and analyzing risk factors is arguably the best approach for investigating sex differences in dementia risk. Broad domains of risk factors have been previously classified as biomarkers, demographic variables, lifestyle factors, medical conditions and medications, and environmental factors. A broader understanding of overall patterns of risk factors for cardiometabolic disease and neurodegeneration is needed to inform tailored (potentially sex-specific) interventions [37]. Future investigations should use methods such as Bayesian Mindsponge Framework analytics, to provide a more in-depth analysis of sex differences in the association between diabetes and dementia [38]. Previous studies have reported an increased risk of IHM in patients admitted to hospital with a diagnosis of diabetes and dementia, because this population has survived ischemic heart disease and stroke [39]. Our study showed that among men and women hospitalized with TD2M who had any type of dementia, IHM remained stable over time until the year 2020, when it showed a significant increase, probably related to SARS-CoV-2 infection. As we expected, among hospitalized women and men with T2DM and all-cause dementia, AD, and VaD, the presence of COVID-19 was associated with IHM, as were advanced age and associated comorbidity. In a previous study conducted by our group, dementia was associated with IHM in men and women with T2DM hospitalized with COVID-19 in Spain in 2020 (OR in men, 2.42 [$95\%$ CI 2.14–2.74]; and OR in women, 1.64 [$95\%$ CI 1.46–1.84]) [40]. Previous evidence has shown that patients with dementia, especially those with comorbidities such as diabetes, are particularly vulnerable to SARS-CoV-2 infection and are more likely to develop severe illness [41]. In our study, women with diabetes hospitalized with all-cause dementia, VaD, and AD, had a lower risk of IHM than men with diabetes. This finding is consistent with previous research, which found that male sex predicts mortality in patients with dementia [42,43,44]. Connors et al. [ 44] found female sex to be a protective factor for mortality in patients with dementia (HR, hazard ratio 0.57; $95\%$ CI 0.43–0.74). To our knowledge no previous investigation has analyzed the possible factors that explain the higher IHM among men than women with diabetes and concomitant dementia. The observed increment in the prevalence of dementia among people hospitalized with diabetes found in our investigation has several practical implications. It is expected that the improvements in rates of chronic complications and longevity in diabetes patients will lead to more people with diabetes surviving into old age and developing dementia, therefore making it necessary that effective interventions are implemented [45,46]. According to the latest version of the Lancet Commission on Dementia Prevention, based on results from large cohort studies, up to $40\%$ of dementia cases could be prevented by modifying twelve risk factors: low education, midlife hearing loss, obesity, hypertension, late-life depression, smoking, physical inactivity, diabetes, social isolation, excessive alcohol consumption, traumatic brain injury, and air pollution [47]. Population-based approaches are likely to be the most impactful, cost-effective, and meaningful to reduce the global burden of dementia. Public health campaigns are needed to raise awareness about the link between diabetes, and the other mentioned risk factors, with dementia. Campaigns must encourage lifestyle changes linked to diet, exercise, and weight loss, which could reduce the incidence of diabetes and mitigate dementia risk [47,48]. Specific interventions that could reduce the risk or slow down the progression of dementia in people with diabetes include optimizing the treatment of cardiovascular risk factors, promoting the use of statins and oral hypoglycemic agents, and a tight control of blood glucose levels with an HbA1 under $7\%$ [49,50]. Our results provide policy makers with objective data on the burden that dementia causes among people with diabetes and very especially among women. Soto-Gordoa et al. have predicted that in Spain the number of cases of dementia will triple by 2050 unless effective interventions are implemented [51]. These is a very serious threat to the Spanish social and health care systems, as the associated economic burden will become barely sustainable. According to these authors, an intervention leading to a $20\%$ change in risk and protective factors would reduce dementia by $9\%$, prevent over 100,000 cases, and save nearly EUR 4900 million in 2050 [51]. In Germany, Fink et al. estimated that a relative reduction of diabetes incidence by $1\%$ annually would decrease dementia cases by around 30,000 [52]. As commented before, the reduction of the burden of dementia can only be achieved with multiple interventions that, taken separately, would yield only modest results. This evidence supports the need to include primary prevention in the form of reducing risk factors for both dementia and diabetes as a top priority of health policies [47,48,51,52]. The strengths of our study are the use of a national population database (RAE-CMBD), over a 10-year period, with a methodology that has been reported elsewhere [17,18]. However, our study is also subject to a series of limitations. While the RAE-CMBD collects practically all hospitalizations in Spain, it is an administrative database and does not collect all the variables included in the clinical history. Therefore, we have no data on disease severity, glycemic control, disease duration, or medication for diabetes or dementia. In addition, it was only possible to assess IHM, since we did not have information on the patients once they were discharged. However, the use of hospital discharge records and administrative databases for the diagnosis of psychiatric illnesses, including dementia, has been shown to be sufficiently sensitive and specific for epidemiological investigations [53,54]. Finally, the validity of diabetes ICD codes in health administrative databases, compared to clinical records, has been evaluated previously, concluding that it is reliable and can be used to address important research questions [55,56,57,58]. ## 5. Conclusions In conclusion, the prevalence of dementia, including the prevalence of VaD and AD, in men and women with T2DM increased between 2011 and 2020. Our data highlight important sex differences, indicating that the prevalence of all-cause dementia is 1.34 times higher in women than in men, with similar values for VaD and AD. However, female sex is a protective factor for IHM in hospitalized patients with all-cause dementia, VaD, and AD. A diagnosis of COVID-19, associated comorbidity, older age, and having been hospitalized in 2020 were predictors of IHM in men and women with T2DM and dementia (all-cause dementia, VaD, and AD). Clinicians should pay attention to the relationship between T2DM and dementia in order to avoid worse outcomes and reduce the burden of both diseases. ## References 1. 1. International Diabetes Federation IDF Diabetes Atlas10th ed.International Diabetes FederationBrussels, Belgium2021Available online: http://www.diabetesatlas.org(accessed on 19 January 2023). IDF Diabetes Atlas (2021.0) 2. World Population Prospects: The 2022 Revision, Key Findings and Advance Tables 3. Global Status Report on the Public Health Response to Dementia 2021 4. Thomassen J.Q., Tolstrup J.S., Benn M., Frikke-Schmidt R.. Type-2 diabetes and risk of dementia: Observational and Mendelian randomisation studies in 1 million individuals. Epidemiol. Psychiatr. Sci. (2020.0) 29 e118. DOI: 10.1017/S2045796020000347 5. Prince M., Albanese E., Guerchet M., Prina M.. World Alzheimer Report 2014: Dementia and Risk Reduction: An Analysis of Protective and Modifiable Risk Factors 6. Launer L.J., Andersen K., Dewey M.E., Letenneur L., Ott A., Amaducci L.A., Brayne C., Copeland J.R., Dartigues J.F., Kragh-Sorensen P.. Rates and risk factors for dementia and Alzheimer’s disease: Results from EURODEM pooled analyses. EURODEM Incidence Research Group and Work Groups. European Studies of Dementia. Neurology (1999.0) 52 78-84. DOI: 10.1212/WNL.52.1.78 7. Dolan C., Glynn R., Lawlor B.. A Systematic Review and Delphi Study to Ascertain Common Risk Factors for Type 2 Diabetes Mellitus and Dementia and Brain-Related Complications of Diabetes in Adults. Can. J. Diabetes (2020.0) 44 628-635. DOI: 10.1016/j.jcjd.2020.01.004 8. Richardson S., Hirsch J.S., Narasimhan M., Crawford J.M., McGinn T., Davidson K.W., Barnaby D.P., Becker L.B., Chelico J.D.. Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area. JAMA (2020.0) 323 2052-2059. DOI: 10.1001/jama.2020.6775 9. Boye K.S., Tokar Erdemir E., Zimmerman N., Reddy A., Benneyworth B.D., Dabora M.C., Hankosky E.R., Bethel M.A., Clark C., Lensing C.J.. Risk Factors Associated with COVID-19 Hospitalization and Mortality: A Large Claims-Based Analysis Among People with Type 2 Diabetes Mellitus in the United States. Diabetes Ther. (2021.0) 12 2223-2239. DOI: 10.1007/s13300-021-01110-1 10. Verhagen C., Janssen J., Biessels G.J., Johansen O.E., Exalto L.G.. Females with type 2 diabetes are at higher risk for accelerated cognitive decline than males: CAROLINA-COGNITION study. Nutr. Metab. Cardiovasc. Dis. (2022.0) 32 355-364. DOI: 10.1016/j.numecd.2021.10.013 11. Chatterjee S., Peters S.A., Woodward M., Mejia Arango S., Batty G.D., Beckett N., Beiser A., Borenstein A.R., Crane P.K., Haan M.. Type 2 Diabetes as a Risk Factor for Dementia in Women Compared with Men: A Pooled Analysis of 2.3 Million People Comprising More Than 100,000 Cases of Dementia. Diabetes Care (2016.0) 39 300-307. DOI: 10.2337/dc15-1588 12. Real Decreto 69/2015, de 6 de febrero, por el que se regula el Registro de Actividad de Atención Sanitaria Especializada. (Spanish National Hospital Discharge Database). BOE (2015.0) 35 10789-10809 13. Quan H., Sundararajan V., Halfon P., Fong A., Burnand B., Luthi J.C., Saunders L.D., Beck C.A., Feasby T.E., Ghali W.A.. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med. Care (2005.0) 43 1130-1139. DOI: 10.1097/01.mlr.0000182534.19832.83 14. Sundararajan V., Henderson T., Perry C., Muggivan A., Quan H., Ghali W.A.. New ICD-10 version of the Charlson comorbidity index predicted in-hospital mortality. J. Clin. Epidemiol. (2004.0) 57 1288-1294. DOI: 10.1016/j.jclinepi.2004.03.012 15. Solicitud de Extracción de Datos—Extraction Request (Spanish National Hospital Discharge Database) 16. Alsharif A.A., Wei L., Ma T., Man K.K.C., Lau W.C.Y., Brauer R., Almetwazi M., Howard R., Wong I.C.K.. Prevalence and Incidence of Dementia in People with Diabetes Mellitus. J. Alzheimers Dis. (2020.0) 75 607-615. DOI: 10.3233/JAD-191115 17. de Miguel-Yanes J.M., Jiménez-García R., Hernández-Barrera V., Méndez-Bailón M., de Miguel-Díez J., Muñoz-Rivas N., Ezpeleta D., López-de-Andrés A.. Hospital Admissions in People with Alzheimer’s Disease or Senile Dementia According to Type 2 Diabetes Status: An Observational 10-Year Study. Am. J. Alzheimer’s Dis. Other Demen. (2018.0) 33 12-19. DOI: 10.1177/1533317517726389 18. Muñoz-Rivas N., Méndez-Bailón M., Miguel-Yanes J.M., Hernández-Barrera V., Miguel-Díez J., Jimenez-Garcia R., López-de-Andrés A.. Observational study of vascular dementia in the Spanish elderly population according to type 2 diabetes status: Trends in incidence, characteristics and outcomes (2004–2013). BMJ Open (2017.0) 7 e016390. DOI: 10.1136/bmjopen-2017-016390 19. Halonen P., Enroth L., Jämsen E., Vargese S., Jylhä M.. Dementia and Related Comorbidities in the Population Aged 90 and Over in the Vitality 90+ Study, Finland: Patterns and Trends From 2001 to 2018. J. Aging Health (2022.0) 8982643221123451. DOI: 10.1177/08982643221123451 20. Xue M., Xu W., Ou Y.N., Cao X.P., Tan M.S., Tan L., Yu J.T.. Diabetes mellitus and risks of cognitive impairment and dementia: A systematic review and meta-analysis of 144 prospective studies. Ageing Res. Rev. (2019.0) 55 100944. DOI: 10.1016/j.arr.2019.100944 21. Zhang J., Chen C., Hua S., Liao H., Wang M., Xiong Y., Cao F.. An Updated Meta-Analysis of Cohort Studies: Diabetes and Risk of Alzheimer’s Disease. Diabetes Res. Clin. Pract. (2017.0) 124 41-47. DOI: 10.1016/j.diabres.2016.10.024 22. Andreu-Reinón M.E., Huerta J.M., Gavrila D., Amiano P., Mar J., Tainta M., Ardanaz E., Larumbe R., Navarro C., Colorado-Yohar S.M.. Incidence of Dementia and Associated Factors in the EPIC-Spain Dementia Cohort. J. Alzheimer’s Dis. (2020.0) 78 543-555. DOI: 10.3233/JAD-200774 23. Christensen K., Doblhammer G., Rau R., Vaupel J.W.. Ageing populations: The challenges ahead. Lancet (2009.0) 374 1196-1208. DOI: 10.1016/S0140-6736(09)61460-4 24. Crimmins E.M., Zhang Y.S., Kim J.K., Levine M.E.. Changing Disease Prevalence, Incidence, and Mortality Among Older Cohorts: The Health and Retirement Study. J. Gerontol. A Biol. Sci. Med. Sci. (2019.0) 74 S21-S26. DOI: 10.1093/gerona/glz075 25. Vargese S.S., Halonen P., Raitanen J., Forma L., Jylhä M., Aaltonen M.. Comorbidities in dementia during the last years of life: A register study of patterns and time differences in Finland. Aging Clin. Exp. Res. (2021.0) 33 3285-3292. DOI: 10.1007/s40520-021-01867-2 26. Hu H.Y., Zhang Y.R., Aerqin Q., Ou Y.N., Wang Z.T., Cheng W., Feng J.F., Tan L., Yu J.T.. Association between multimorbidity status and incident dementia: A prospective cohort study of 245,483 participants. Transl. Psychiatry (2022.0) 12 505. DOI: 10.1038/s41398-022-02268-3 27. Leelakanok N., D’Cunha R.R.. Association between polypharmacy and dementia—A systematic review and metaanalysis. Aging Ment. Health (2019.0) 23 932-941. DOI: 10.1080/13607863.2018.1468411 28. Pham A.N.Q., Lindeman C., Voaklander D., Wagg A., Drummond N.. Risk factors for incidence of dementia in primary care practice: A retrospective cohort study in older adults. Fam. Pract. (2022.0) 39 406-412. DOI: 10.1093/fampra/cmab168 29. Thomas E.G., Rhodius-Meester H., Exalto L., Peters S.A.E., van Bloemendaal L., Ponds R., Muller M.. Sex-Specific Associations of Diabetes With Brain Structure and Function in a Geriatric Population. Front. Aging Neurosci. (2022.0) 14 885787. DOI: 10.3389/fnagi.2022.885787 30. Carcaillon L., Brailly-Tabard S., Ancelin M.L., Rouaud O., Dartigues J.F., Guiochon-Mantel A., Scarabin P.Y.. High plasma estradiol interacts with diabetes on risk of dementia in older postmenopausal women. Neurology (2014.0) 82 504-511. DOI: 10.1212/WNL.0000000000000107 31. Matyi J.M., Rattinger G.B., Schwartz S., Buhusi M., Tschanz J.T.. Lifetime estrogen exposure and cognition in late life: The Cache County Study. Menopause (2019.0) 26 1366-1374. DOI: 10.1097/GME.0000000000001405 32. Gong J., Harris K., Hackett M., Peters S.A.E., Brodaty H., Cooper M., Hamet P., Harrap S., Mancia G., MacMahon S.. Sex differences in risk factors for cognitive decline and dementia, including death as a competing risk, in individuals with diabetes: Results from the ADVANCE trial. Diabetes Obes. Metab. (2021.0) 23 1775-1785. DOI: 10.1111/dom.14391 33. Gong J., Harris K., Peters S.A.E., Woodward M.. Sex differences in the association between major cardiovascular risk factors in midlife and dementia: A cohort study using data from the UK Biobank. BMC Med. (2021.0) 19. DOI: 10.1186/s12916-021-01980-z 34. Gong J., Harris K., Lipnicki D.M., Castro-Costa E., Lima-Costa M.F., Diniz B.S., Xiao S., Lipton R.B., Katz M.J., Wang C.. Sex differences in dementia risk and risk factors: Individual-participant data analysis using 21 cohorts across six continents from the COSMIC consortium. Alzheimer’s Dement. (2023.0). DOI: 10.1002/alz.12962 35. Mewton L., Visontay R., Hoy N., Lipnicki D.M., Sunderland M., Lipton R.B., Guerchet M., Ritchie K., Najar J., Scarmeas N.. The relationship between alcohol use and dementia in adults aged more than 60 years: A combined analysis of prospective, individual-participant data from 15 international studies. Addiction (2023.0) 118 412-424. DOI: 10.1111/add.16035 36. Geraets A.F.J., Leist A.K.. Sex/gender and socioeconomic differences in modifiable risk factors for dementia. Sci. Rep. (2023.0) 13 80. DOI: 10.1038/s41598-022-27368-4 37. Anstey K.J., Eramudugolla R., Hosking D.E., Lautenschlager N.T., Dixon R.A.. Bridging the translation gap: From dementia risk assessment to advice on risk reduction. J. Prev. Alzheimer’s Dis. (2015.0) 2 189-198. DOI: 10.14283/jpad.2015.75 38. Vuong Q.H., Nguyen M.H., La V.P.. The Mindsponge and BMF Analytics for Innovative Thinking in Social Sciences and Humanities (2022.0). DOI: 10.2478/9788367405119 39. Pearson-Stuttard J., Bennett J., Cheng Y.J., Vamos E.P., Cross A.J., Ezzati M., Gregg E.W.. Trends in predominant causes of death in individuals with and without diabetes in England from 2001 to 2018: An epidemiological analysis of linked primary care records. Lancet Diabetes Endocrinol. (2021.0) 9 165-173. DOI: 10.1016/S2213-8587(20)30431-9 40. de Miguel-Yanes J.M., Jimenez-Garcia R., de Miguel-Diez J., Hernández-Barrera V., Carabantes-Alarcon D., Zamorano-Leon J.J., Omaña-Palanco R., Lopez-de-Andres A.. Impact of Type 2 Diabetes Mellitus on the Incidence and Outcomes of COVID-19 Needing Hospital Admission According to Sex: Retrospective Cohort Study Using Hospital Discharge Data in Spain, Year 2020. J. Clin. Med. (2022.0) 11. DOI: 10.3390/jcm11092654 41. Chen F., Chen Y., Wang Y., Ke Q., Cui L.. The COVID-19 pandemic and Alzheimer’s disease: Mutual risks and mechanisms. Transl. Neurodegener. (2022.0) 11 40. DOI: 10.1186/s40035-022-00316-y 42. Garcia-Ptacek S., Farahmand B., Kåreholt I., Religa D., Cuadrado M.L., Eriksdotter M.. Mortality risk after dementia diagnosis by dementia type and underlying factors: A cohort of 15,209 patients based on the Swedish Dementia Registry. J. Alzheimer’s Dis. (2014.0) 41 467-477. DOI: 10.3233/JAD-131856 43. Xie J., Brayne C., Matthews F.E.. Survival times in people with dementia: Analysis from population based cohort study with 14 year follow-up. BMJ (2008.0) 336 258-262. DOI: 10.1136/bmj.39433.616678.25 44. Connors M.H., Ames D., Boundy K., Clarnette R., Kurrle S., Mander A., Ward J., Woodward M., Brodaty H.. Predictors of Mortality in Dementia: The PRIME Study. J. Alzheimer’s Dis. (2016.0) 52 967-974. DOI: 10.3233/JAD-150946 45. Gregg E.W., Li Y., Wang J., Burrows N.R., Ali M.K., Rolka D., Williams D.E., Geiss L.. Changes in diabetes-related complications in the United States, 1990–2010. N. Engl. J. Med. (2014.0) 370 1514-1523. DOI: 10.1056/NEJMoa1310799 46. Wu Y.T., Fratiglioni L., Matthews F.E., Lobo A., Breteler M.M., Skoog I., Brayne C.. Dementia in western Europe: Epidemiological evidence and implications for policy making. Lancet Neurol. (2016.0) 15 116-124. DOI: 10.1016/S1474-4422(15)00092-7 47. Livingston G., Huntley J., Sommerlad A., Ames D., Ballard C., Banerjee S., Brayne C., Burns A., Cohen-Mansfield J., Cooper C.. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet (2020.0) 396 413-446. DOI: 10.1016/S0140-6736(20)30367-6 48. Wallace L., Brayne C.. The need for a better global dementia response. Lancet Neurol. (2022.0) 21 115-116. DOI: 10.1016/S1474-4422(21)00459-2 49. Pal K., Mukadam N., Petersen I., Cooper C.. Mild cognitive impairment and progression to dementia in people with diabetes, prediabetes and metabolic syndrome: A systematic review and meta-analysis. Soc. Psychiatry Psychiatr. Epidemiol. (2018.0) 53 1149-1160. DOI: 10.1007/s00127-018-1581-3 50. Frison E., Dufouil C., Helmer C., Berr C., Auriacombe S., Chêne G.. Diabetes-Associated Dementia Risk and Competing Risk of Death in the Three-City Study. J. Alzheimer’s Dis. (2019.0) 71 1339-1350. DOI: 10.3233/JAD-190427 51. Soto-Gordoa M., Arrospide A., Moreno-Izco F., Martínez-Lage P., Castilla I., Mar J.. Projecting Burden of Dementia in Spain, 2010–2050: Impact of Modifying Risk Factors. J. Alzheimer’s Dis. (2015.0) 48 721-730. DOI: 10.3233/JAD-150233 52. Fink A., Doerre A., Demuth I., Doblhammer G.. Potential of prevention strategies for the modifiable risk factor type 2 diabetes with relation to the future number of dementia patients in Germany—A multi-state projection through 2040. BMC Neurol. (2022.0) 22. DOI: 10.1186/s12883-022-02682-6 53. Castillo E.G., Olfson M., Pincus H.A., Vawdrey D., Stroup T.S.. Electronic health records in mental health research: A framework for developing valid research methods. Psychiatr. Serv. (2015.0) 66 193-196. DOI: 10.1176/appi.ps.201400200 54. Spiranovic C., Matthews A., Scanlan J., Kirkby K.C.. Increasing knowledge of mental illness through secondary research of electronic health records: Opportunities and challenges. Adv. Ment. Health (2016.0) 14 14-25. DOI: 10.1080/18387357.2015.1063635 55. Leong A., Dasgupta K., Bernatsky S., Lacaille D., Avina-Zubieta A., Rahme E.. Systematic review and meta-analysis of validation studies on a diabetes case definition from health administrative records. PLoS ONE (2013.0) 8. DOI: 10.1371/journal.pone.0075256 56. Khokhar B., Jette N., Metcalfe A., Cunningham C.T., Quan H., Kaplan G.G., Butalia S., Rabi D.. Systematic review of validated case definitions for diabetes in ICD-9-coded and ICD-10-coded data in adult populations. BMJ Open (2016.0) 6 e009952. DOI: 10.1136/bmjopen-2015-009952 57. Fuentes S., Cosson E., Mandereau-Bruno L., Fagot-Campagna A., Bernillon P., Goldberg M., Fosse-Edorh S.. Identifying diabetes cases in health administrative databases: A validation study based on a large French cohort. Int. J. Public Health (2019.0) 64 441-450. DOI: 10.1007/s00038-018-1186-3 58. Wild S., Fischbacher C., McKnight J.. Using Large Diabetes Databases for Research. J. Diabetes Sci. Technol. (2016.0) 10 1073-1078. DOI: 10.1177/1932296816645120
--- title: Fisetin, a Natural Polyphenol, Ameliorates Endometriosis Modulating Mast Cells Derived NLRP-3 Inflammasome Pathway and Oxidative Stress authors: - Alessia Arangia - Ylenia Marino - Roberta Fusco - Rosalba Siracusa - Marika Cordaro - Ramona D’Amico - Francesco Macrì - Emanuela Raffone - Daniela Impellizzeri - Salvatore Cuzzocrea - Rosanna Di Paola journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049430 doi: 10.3390/ijms24065076 license: CC BY 4.0 --- # Fisetin, a Natural Polyphenol, Ameliorates Endometriosis Modulating Mast Cells Derived NLRP-3 Inflammasome Pathway and Oxidative Stress ## Abstract A chronic, painful, and inflammatory condition known as endometriosis is defined by the extra-uterine development of endometrial tissue. The aim of this study was to evaluate the beneficial effects of fisetin, a naturally occurring polyphenol that is frequently present in a variety of fruits and vegetables. Uterine fragments were injected intraperitoneally to cause endometriosis, and fisetin was given orally every day. At 14 days of treatment, laparotomy was performed, and the endometrial implants and peritoneal fluids were collected for histological, biochemical, and molecular analyses. Rats subjected to endometriosis presented important macroscopic and microscopic changes, increased mast cell (MC) infiltration, and fibrosis. Fisetin treatment reduced endometriotic implant area, diameter, and volumes, as well as histological alterations, neutrophil infiltration, cytokines release, the number of MCs together with the expression of chymase and tryptase, and diminished α smooth muscle actin (α-sma) and transforming growth factor beta (TGF β) expressions. In addition, fisetin was able to reduce markers of oxidative stress as well as nitrotyrosine and Poly ADP ribose expressions and increase apoptosis in endometrial lesions. In conclusion, fisetin could represent a new therapeutic strategy to control endometriosis perhaps by targeting the MC-derived NOD-like receptor family pyrin domain containing 3 (NLRP3) inflammasome pathway and oxidative stress. ## 1. Introduction Endometrium-like tissue that is present outside of the uterine cavity, particularly on the pelvic peritoneum and ovaries, is a symptom of the inflammatory disease called endometriosis, which is estrogen-dependent [1,2,3,4]. About $10\%$ of women in their reproductive years experience it, and it is linked to pelvic pain and infertility [5]. Retrograde menstruation is generally accepted to be the source of ectopic endometrial tissue, and the peritoneal immunological milieu is critical for the progression of endometriosis. Inflammation in the peritoneal cavity is brought on by elements of refluxed blood, such as apoptotic endometrial tissue, desquamated menstrual cells, lysed erythrocytes, and released iron. Reactive oxygen species (ROS) released by triggered macrophages cause oxidative stress [5]. Mast cells (MCs) are crucial immune cell types whose function is disturbed in the peritoneal milieu of endometriosis [6]. Supplementary studies report that MCs play crucial roles in the development of endometriosis [7]. More and more evidence points to the involvement of MCs in inflammatory/autoimmune disorders as well by enhancing vascular permeability, facilitating immunological responses, and controlling fibrosis [8]. There are more activated MCs in endometriotic lesions, which may emit inflammatory mediators such as histamine and tumor necrosis factor (TNF)-α [9,10]. Additionally, after being exposed to stimuli, MCs can produce endogenous ROS [11]. An intracellular receptor called NLRP3 (NOD-like receptor family pyrin domain containing 3) detects both exogenous and endogenous danger cues [12] and interacts with apoptosis-associated speck-like protein (ASC), caspase-1 to construct the NLRP3 inflammasome complex. Interleukin (IL)-1β and other pro-inflammatory cytokines are released when the complex is activated, and it also stimulates immunological response and pyroptosis [13]. Prior research has also identified the special function of NLRP3 inflammasome in the activation of MCs during autoinflammatory conditions [14]. ROS is becoming a crucial regulator of inflammasome NLRP3 activation, which is connected to a number of illnesses [15]. Endometriosis progression is correlated to a pro-oxidative and immune mechanism and overproduction of ROS is associated with malignancy diffusion and increased proliferation rate [16]. Increased oxidative stress markers have been found in samples from women affected by this disease [1]. Oral contraceptives or nonsteroidal anti-inflammatory medicines are recommended as the first-line therapy for endometriosis, although many patients continue to exhibit pelvic and lesion size enlargement [3]. Interestingly, clinical studies supported that the administering antioxidants to women with endometriosis decreases their persistent pelvic pain and inflammatory indicators in peritoneal fluids [17]. Thus, the evaluation of therapeutic approaches targeting oxidative imbalance or related ROS molecular pathways could help to prevent endometriosis. Flavonoids are a large group of minor metabolites in plants containing over 4000 composites. Their principal subclasses are oxoflavonoids (flavonols and flavones), flavan-3-ol derivatives (such as tannins and catechin anthocyanins), and isoflavones [18]. Among flavonoids, the flavonol fisetin (3,3′,4′,7-tetrahydroxyflavone), is frequently present in fruits and vegetables. The largest concentration of fisetin was discovered in strawberries (160 μg/g) followed by apples (26.9 μg g/g) and persimmon (10.5 μg g/g) [19,20]. Fisetin is largely characterized by a diphenylpropane structure and contains two aromatic rings connected through three carbons oxygenated heterocyclic rings and united with one oxo and four hydroxyl groups as substituents. In particular, the quantity and location of the hydroxyl groups in this structure are closely related to its biological activities [21]. The biological effects of fisetin, such as anticancer [22,23] antidiabetic [24], anti-inflammatory [25], antioxidant [26], and neuroprotective effects [15,27] were investigated both in vitro and in vivo models [28]. Among these properties, fisetin looked to apply a solid cytotoxic action versus several varieties of cancers such as colon, lung, ovarian, liver, and breast [22,23]. In vivo tests on fisetin toxicity reported that the rats do not exhibit symptoms like decreased body weight, restlessness, respiratory discomfort, diarrhea, contractions, or coma [29,30]. Fisetin is considered a drug and nutritive supplement and is also recognized as a medical food, which stimulates more preclinical studies that could be translate into human clinical trials. One current clinical trial revealed important enhancement in anti-inflammatory markers when fisetin was injected into subjects with colorectal cancer [31,32]. A clinical study in stroke patients showed that fisetin was able to improve the treatment results, particularly in people with delayed onset-of-treatment [33]. Emerging findings indicated the possible health benefits of fisetin in aging as a senolytic through the lessening of senescent markers and age-related pathologies [34,35]. In that regard, a pilot study is also underway to examine the effectiveness of fisetin in decreasing inflammatory factors in the blood and in reducing fragility and markers of inflammation, insulin resistance, and bone resorption in elderly adults. ClinicalTrials.gov Identifier: NCT03675724. Mechanistically, the anti-inflammatory activity of fisetin is often correlated with free radical scavenging and antioxidant properties, via regulating nuclear factor erythroid 2–related factor 2 Nrf-2/heme oxygenase-1 HO-1 and nuclear factor NF-κB, [26]. The previous study also reported that fisetin ameliorated inflammation and oxidative stress in lipopolysaccharide-induced endometritis [28]. Therefore, this study aimed to explore the antioxidant effect of fisetin on endometriosis considering the contribution of mast cells and the NLRP3 inflammasome pathway. ## 2.1. Fisetin Ameliorates Endometriotic Lesions At 14 days of induction—the end of the experiment—all animals from the vehicle and fisetin groups displayed endometriosis lesions (Figure 1A,B), while sham animals did not show any implants. Pelvic ultrasound evaluated the presence of endometriomas (Figure 1(A1,B1)). The evaluation included both anterior and posterior pelvic compartments to evaluate the different endometriosis locations (Figure 1(A1,B1)). The lesions from the fisetin group appeared smaller and more superficially attached to the peritoneal cavity compared to the vehicle group (Figure 1(B1)). These results were confirmed by macroscopic observation that showed that the cyst from the vehicle group was more evident than the cyst from the fisetin group (Figure 1A,B). In particular, both groups showed no distinctive number of cysts (Figure 1C). However, cysts diameter (Figure 1D), area (Figure 1E), and volume (Figure 1H) were lesser in the fisetin-treated group compared to the vehicle animals. Histologically, endometriotic lesions from vehicle-treated rats showed abundant stromal structure and endometrial-type glands (Figure 1(F,F1,I)). Fisetin administration reduced the histopathological marks of endometriosis (Figure 1(G,G1,I)). ## 2.2. Fisetin Reduces Mast Cell Activation on Endometriotic Lesions Several papers described the key role of inflammatory cell recruitment at the lesion site during endometriosis [36]. To better evaluate the role of MCs on endometriotic lesion, we performed toluidine blue staining. Explants from vehicle rats revealed increased MC recruitment (Figure 2A,G), while fisetin animals showed reduced MC infiltration (Figure 2B,G). In addition, to confirm the activity of MCs and their activation, we evaluated the chymase and tryptase expressions by immunohistochemical analysis. In particular, vehicle-treated animals showed increased positive staining for chymase and tryptase while the fisetin group showed reduced positivity (Figure 2C–I). Additionally, MPO a marker of neutrophil infiltration was also evaluated. Fisetin reduced MPO activity compared to the vehicle rats (Figure 2J). ## 2.3. Fisetin Reduces Fibrotic Process on Endometriotic Lesions Fisetin administration also showed important anti-fibrotic effects. Masson trichrome staining showed a reduction in collagen fibers in lesions from fisetin rats (Figure 3B,G), compared to the vehicle (Figure 3A,G). In accordance with the staining, α-sma and TGF-β expressions were also evaluated. In particular, positive immunoreactivity was markedly increased in the vehicle group while the fisetin group showed reduced expression for α-sma and TGF-β (Figure 3C–I). ## 2.4. Fisetin Reduces Oxidative Stress on Endometriotic Lesions Several works demonstrated the role of ROS on endometriosis development [37]. Superoxide anion (O2•−) and nitric oxide (NO) can react with each other contributing to the formation of peroxynitrite (ONOO−). This molecule can then act on proteins, leading to the nitration of protein tyrosines by the formation of nitrotyrosine. In addition, upon DNA damage, poly (ADP-ribose) polymerase-1 (PARP-1) is activated and catalyzes the formation of poly(ADP-ribose) (PAR) chains by transferring (ADP-ribose) from NAD+ onto itself and nuclear acceptor proteins. In that regard, increased expression of nitrotyrosine (Figure 4C,F) and PAR (Figure 4A,E) were observed in the vehicle group compared to animals treated with fisetin (Figure 4B,D–F). MDA levels were also evaluated as markers of lipid peroxidation. MDA levels were increased in the vehicle group, while these levels were reduced by fisetin treatment (Figure 4G). ## 2.5. Fisetin Reduces Inflammasome Pathway and NF-κB Expression To better investigate whether fisetin could act by inhibiting the inflammasome pathway, we performed Western blots for NLRP-3, ASC, and cleaved caspase-1. The expression of the NLRP-3, ASC, and cleaved caspase-1 was significantly upregulated in endometriotic lesions tissues of vehicle rats (Figure 5A–C). Treatment with fisetin notably inhibited the NLRP3, ASC, and cleaved caspase-1 expression in endometriotic tissues of vehicle rats subjected to endometriosis (Figure 5A–C). In addition, increased nuclear NF-κB expression was observed in the vehicle rats subjected to endometriosis (Figure 5D). Fisetin significantly reduced the level of nuclear NF-κB compared to the vehicle group (Figure 5D). ## 2.6. Fisetin Reduces Proinflammatory Cytokines Several cytokines including IL-1β and TNF-α were reported to be increased in the PF of women with endometriosis [38]. In addition, the significance of inflammasome and successive excretion of IL-1 family members is supposed to be intricated in the pathogenesis of various diseases [39]. In that regard, increased levels of IL-1β and TNF-α were found in animals subjected to endometriosis and treated with the vehicle compared to the sham (Figure 6A,B). The oral administration of fisetin diminished these proinflammatory cytokines levels (Figure 6A,B). ## 2.7. Fisetin Increases Apoptotic Process The disparity between endometriotic cell growth and apoptosis characterized endometriosis. To research the effect of fisetin administration on apoptosis, TUNEL assay and Western blot analysis were conducted. To quantify cells undergoing apoptosis, the TUNEL assay was used. Tissue samples from animals displayed a reduced number of apoptotic cells (Figure 7A,C), while fisetin was able to increase the number of apoptotic cells (Figure 7B,C). The results were also confirmed by Western blot. Fisetin was able to increase apoptosis by reducing Bcl-2 and increasing Bax and caspase 3 levels (Figure 7D–F) with respect to vehicle rats. ## 3. Discussion Women of reproductive age are susceptible to the gynecological and excruciating ailment known as endometriosis. It is characterized by non-uterine implants that resemble diseased endometrium. Although there is still debate on the pathophysiology of endometriosis, it is well-recognized that the inflammatory response is dangerous to this process. In this study, we provided mechanistic data on how fisetin, a natural flavonol, could have a positive action on estrogen-dependent endometriosis modulating MC-derived NLRP3 inflammasome. Further studies report that MC activation by estrogen plays a crucial role in the development of endometriosis [7], through NLRP3 inflammasome activation [40,41]. Previous research suggested that estrogen can encourage the evolution of endometriotic lesions and have a role in the pathogenesis of endometriosis by stimulating MCs, which may increase the release of TNF-α and quicken the disease’s progression [42]. Targeted inhibition of NLRP3 considerably restrained lesion development and fibrogenesis in a mouse model of endometriosis [9]. Because erythrocytes, macrophages, and apoptotic endometrial tissues are well-known oxidative stress inducers, it is possible that the peritoneal generation of ROS contributes to endometriosis [37]. Various pieces of evidence support the role of oxidative stress in the progression of endometriosis [43,44]. This finding may pave the way for the assessment of therapeutic strategies targeting oxidative imbalance. Fisetin is one nutritive polyphenol that has been meticulously investigated [35,45]. It is extensively existed in a variety of vegetables and fruits, such as cucumbers, onions, strawberries, and apples. Animal studies have indicated that fisetin administration has favorable effects against diverse diseases [20]. It has been stated that fisetin has numerous biological activities, such as antioxidative, anti-inflammatory, and antitumor effects [46]. In vivo experimentations showed that fisetin lessened inflammatory damage and MPO activity in LPS-stimulated mouse endometritis [28]. Interestingly, the NF-κB inhibition or Nrf2 activation of fisetin has been determined in both rats and cells [47,48]. In addition, recent studies also confirmed that fisetin ameliorated several diseases by inhibiting inflammatory reactions via NLRP3, which is consistent with the results of the present study [49,50]. Based on these findings, in this study, we demonstrated that fisetin was able to decrease cyst diameter, area, and volume, histological alterations, neutrophil infiltration, the number of MCs, as well as chymase and tryptase expressions. Extensive adhesions and fibrosis present in progressive endometriotic lesions are linked to pelvic morbidities like chronic pelvic discomfort and infertility. Fisetin administration reduced collagen deposition, α-sma, and TGF-β expressions showing reduced fibrosis. Oxidative stress induces ovarian injuries. In fact, endometriosis patients’ granulosa cells exhibit greater evidence of oxidative DNA damage. Granulosa cells of females with endometriosis display a higher prevalence of apoptotic bodies and nitrotyrosine than controls [51]. MDA has been valued as an index of lipid peroxides. Nasiri et al. also observed higher levels of MDA in the serum of subjects with endometriosis [52]. In that regard, here we evaluated markers of oxidative and nitrosative stress, in particular, we found high expression of nitrotyrosine and PAR expression as well as high levels of MDA in endometriotic lesions of vehicle rats while fisetin was able to decrease nitrotyrosine, PAR, and MDA levels. Additionally, it has been demonstrated that ROS play a significant role in triggering NLRP3 inflammasome [15]. NLRP3 inflammasome was initially designated by Tschopp et al. in 2002 [53] and is supposed to be indispensable in natural immunity [53]. The levels of NLRP3, caspase-1, apoptosis-associated speck-like protein, and IL-1 expression are considerably high during pathological conditions, causing an inflammatory response in the body, which, consequently leads to disproportionate production of several types of cell necrosis and programmed cell death [54]. Later, NLRP3 triggers a series of downstream signaling pathways, which determine the cleavage of inactive IL-1β and IL-18 precursors into mature and active IL-1β and IL-18, which are successively released into the extracellular compartment, instigating inflammatory responses and oxidative stress [49]. It has been suggested that NLRP3 inflammasome, which leads to the activation of IL-1β, contributes to the progression of endometriosis [55]. The activation of nuclear factor NF-κB, which increases the production of pro-inflammatory cytokines, is one of the detrimental effects of excessive ROS. NF-κB can incite the NLRP3 inflammasome and the maturation of pro-inflammatory cytokines [56]. Studies conducted in vivo and in vitro have shown its inflammatory activation in endometriotic cells [43]. ROS generation causes peritoneal macrophages to produce more NF-κB, which in turn causes endometriosis patients to produce proinflammatory, growth, and angiogenic factors [43]. Here, we demonstrated that endometriosis in rats caused the activation of NLRP3 and NF-κB pathways as well as the increase in levels of IL-1β and TNF-α, while fisetin inhibited the activation of NLRP3/NF-κB as well as diminished the levels of IL-1β and TNF-α. These results are in agreement with several works that showed that fisetin treatment was able to reduce the NF-κB/NLRP3 pathway as well as reduce the release of proinflammatory cytokines [15,57,58]. ROS generation is linked to enhanced proliferation rates in both tumor and endometriotic cells [59]. Apoptosis, which occurs during the menstrual cycle, keeps cells in a state of equilibrium by removing extra or malfunctioning cells. Apoptotic activity of endometriotic cells is regulated by diverse factors including both anti-apoptotic Bcl-2 and Bcl-xL and pro-apoptotic Bax and caspase-3. Several studies pointed out that the endometrium of endometriotic females is less sensitive to apoptosis than that of healthy controls [60,61]. In that regard, here we showed that rats subjected to endometriosis presented a reduced number of TUNEL-positive cells and reduced expression of proapoptotic proteins Bax and caspase 3 as well as increased Bcl-2. On the contrary, the administration of fisetin was able to increase the apoptotic process by increasing Bax and caspase 3 and reducing antiapoptotic Bcl-2 expressions. In conclusion, in this study, we reported the relevance of MCs and the inflammasome pathway together with oxidative stress in the development of endometriosis. Fisetin administration was able to ameliorate oxidative damage induced by endometriosis in rats by reducing the activation of MCs and targeting the NF-κB/inflammasome pathway. ## 4.1. Animals Sprague Dawley rats (Female, 250 gr) (Envigo, Milan, Italy) were housed in steel cages in a controlled room and were fed and watered regularly. The research was authorized by the University of Messina’s Animal Care Review Board. All animal tests complied with both Italian (D.Lgs $\frac{2014}{26}$) and EU (EU Directive $\frac{2010}{63}$) legislation. ## 4.2. Experimental Protocol Endometriosis was established as already described [2]. Animals were randomly divided into two groups, donor or recipient. To stimulate similar estrogen levels, donor rats were intraperitoneally injected with 10 IU of pregnant mare serum gonadotropin and euthanized later. The uterus from donors was removed and minced with scissors. The recipient animals were injected intraperitoneally with the equivalent of tissue from one uterus in 500 uL of PBS (Sigma Aldrich, St. Louis, MO, USA) along the midventral line. It took seven days for endometriosis to clinically manifest with the endometrial cyst formation that was confirmed by abdominal high-frequency ultrasound analysis. ## 4.3. Experimental Groups At 7 days after induction, rats were randomized and assigned to the following groups $$n = 12$$ for each group: [1] Vehicle group: rats were subjected to the experimental procedure as described above, and vehicle ($1\%$ dimethyl sulfoxide solvent (DMSO, Sigma Aldrich, St. Louis, MO, USA) was orally administered by gavage, on the 7th day and for the next 7 days. [2] Fisetin group: rats were subjected to an experimental procedure as described above, and fisetin (40 mg/Kg) was orally administered by gavage, on the 7th day and for the next 7 days. [3] Sham group: rats were injected intraperitoneally with 500 uL of PBS without endometrial tissue, and vehicle ($1\%$ DMSO) was orally administered by gavage, on the 7th day and for the next 7 days. The dose and route of administration of fisetin were chosen based on a previous study [15]. To evaluate the effect of fisetin on the endometriotic lesions, rats were sacrificed at 14 days after endometriosis induction. Laparotomy was performed to collect the endometriotic implants and peritoneal fluids and further processed for molecular analysis. ## 4.4. Histological Evaluation Histological sections were stained with hematoxylin and eosin (H&E) and evaluated using a Leica DM6 microscope (Leica Microsystems SpA, Milan, Italy) associated with Leica LAS X Navigator software 3.4.2. Histopathologic scores were evaluated as described previously [4]. Additionally, lesion volume was calculated according to the formula: V = (length × width2) × 0.5 [62]. Mast cell evaluation was performed by toluidine blue staining [4]. ## 4.5. Abdominal High-Frequency Ultrasound At the 7th and 14th days post-induction, ultrasonographic examinations were performed using an Esaote MYLAB OMEGA VET (Esaote Italia, Milan, Italy) on sedated rats positioned in dorsal recumbency. Measurements were performed offline by a reader blinded to the condition of the rat as previously described [2,63]. ## 4.6. Analysis of Myeloperoxidase (MPO) Activity MPO activity was measured in endometriosis lesions as already described [43,44] and measured at 450 nm [64,65]. ## 4.7. Cytokines Measurements (TNF)-α and IL-1β and were determined using ELISA kits (BioLegend, San Diego, CA, USA; R&D Systems, Milan, Italy) in peritoneal fluids [66,67,68,69]. ## 4.8. Immunohistochemical Analysis Immunohistochemical analyses were performed in endometriosis lesions as already described [1,70,71,72,73]. The following primary antibodies were used: anti-α-sma antibody (Santa Cruz Biotechnology SCB,1:100 Biogenerica srl CT, Italy) and TGF-β (SCB, sc-130348 1:100) anti-chymase (1:100, SCB #sc59586 Biogenerica srl CT, Italy), and anti-tryptase (1:100, SCB, #sc59587 Biogenerica srl CT, Italy), anti-poly ADP ribose PAR (H-250: sc-7150, 1:500, SCB, Biogenerica srl CT, Italy), anti-nitrotyrosine antibody (Millipore, 06-284, Biogenerica srl CT, Italy), All sections were washed with PBS and then treated as previously reported [74]. Stained sections were observed using a Leica DM6 microscope (Leica Microsystems SpA, Milan, Italy). The histogram profile is related to the positive pixel intensity value obtained. ## 4.9. Apoptosis Tunel Assay Apoptosis was analyzed by a TUNEL assay using a kit (DBA Italia, Milan, Italy) [65,75,76,77,78]. ## 4.10. Western Blot Analysis Cyst samples were homogenized, and Western blots were performed as already described [3,79,80,81,82,83,84]. The following primary antibodies were used: anti-nuclear factor NF-κB (SCB; 1:500 #sc8008, DBA Italia, Milan, Italy), anti-Bcl-2 (SCB, sc-7382, DBA Italia, Milan, Italy), anti-Bax (SCB, sc-7480, DBA Italia, Milan, Italy), anti-NRLP3 (SCB, sc-66846, DBA Italia, Milan, Italy), or anti-ASC antibody (SCB, N-15: sc-22514-R, DBA Italia, Milan, Italy), or anti-cleaved caspase 3 (sc-271028 SCB, DBA Italia, Milan, Italy), anti-Caspase-1 p20 (SCB, G-19: sc-1597 DBA Italia, Milan, Italy) in 1x PBS, $5\%$ (w/v) non-fat dried milk, $0.1\%$ Tween-20 at 4 °C overnight. Membranes were incubated with peroxidase-conjugated bovine anti-mouse IgG secondary antibody or peroxidase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch, West Grove, PA, USA; 1:2000, DBA Italia, Milan, Italy) for 1 h room temperature. Anti β actin or anti-lamin A/C (SCB, 1:5000, DBA Italia, Milan, Italy) antibodies were used as controls. The expression of protein bands was detected by a procedure previously described [3]. ## 4.11. Materials All chemicals were analytical grade or higher. Fisetin (Fustel) was purchased from Selleckchem, Biogenerica CT, Italy. ## 4.12. Statistical Evaluation SEM = the mean standard error of the mean of N observations; N = the number of animals. The photos for histology/immunohistochemistry were from at least three different experiments. The results were analyzed by t-test when comparing two groups and one-way ANOVA followed by a Bonferroni post hoc test for multiple comparisons. The 0.05 p-value was taken as significant. ## References 1. Cordaro M., Trovato Salinaro A., Siracusa R., D’Amico R., Impellizzeri D., Scuto M., Ontario M.L., Interdonato L., Crea R., Fusco R.. Hidrox((R)) and Endometriosis: Biochemical Evaluation of Oxidative Stress and Pain. Antioxidants (2021) 10. DOI: 10.3390/antiox10050720 2. Genovese T., Cordaro M., Siracusa R., Impellizzeri D., Caudullo S., Raffone E., Macri F., Interdonato L., Gugliandolo E., Interlandi C.. Molecular and Biochemical Mechanism of Cannabidiol in the Management of the Inflammatory and Oxidative Processes Associated with Endometriosis. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23105427 3. Siracusa R., D’Amico R., Cordaro M., Peritore A.F., Genovese T., Gugliandolo E., Crupi R., Impellizzeri D., Cuzzocrea S., Fusco R.. The Methyl Ester of 2-Cyano-3,12-Dioxooleana-1,9-Dien-28-Oic Acid Reduces Endometrial Lesions Development by Modulating the NFkB and Nrf2 Pathways. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22083991 4. Siracusa R., D’Amico R., Impellizzeri D., Cordaro M., Peritore A.F., Gugliandolo E., Crupi R., Salinaro A.T., Raffone E., Genovese T.. Autophagy and Mitophagy Promotion in a Rat Model of Endometriosis. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22105074 5. Mate G., Bernstein L.R., Torok A.L.. Endometriosis Is a Cause of Infertility. Does Reactive Oxygen Damage to Gametes and Embryos Play a Key Role in the Pathogenesis of Infertility Caused by Endometriosis?. Front. Endocrinol. (2018) 9 725. DOI: 10.3389/fendo.2018.00725 6. Borelli V., Martinelli M., Luppi S., Vita F., Romano F., Fanfani F., Trevisan E., Celsi F., Zabucchi G., Zanconati F.. Mast Cells in Peritoneal Fluid from Women With Endometriosis and Their Possible Role in Modulating Sperm Function. Front. Physiol. (2019) 10 1543. DOI: 10.3389/fphys.2019.01543 7. Zhu T.H., Zou G., Ding S.J., Li T.T., Zhu L.B., Wang J.Z., Yao Y.X., Zhang X.M.. Mast cell stabilizer ketotifen reduces hyperalgesia in a rodent model of surgically induced endometriosis. J. Pain Res. (2019) 12 1359-1369. DOI: 10.2147/JPR.S195909 8. Fehervari Z.. Mast cells in autoimmune disease. Nat. Immunol. (2018) 19 316. DOI: 10.1038/s41590-018-0080-8 9. Guo X., Xu X., Li T., Yu Q., Wang J., Chen Y., Ding S., Zhu L., Zou G., Zhang X.. NLRP3 Inflammasome Activation of Mast Cells by Estrogen via the Nuclear-Initiated Signaling Pathway Contributes to the Development of Endometriosis. Front. Immunol. (2021) 12 749979. DOI: 10.3389/fimmu.2021.749979 10. Elieh Ali Komi D., Wohrl S., Bielory L.. Mast Cell Biology at Molecular Level: A Comprehensive Review. Clin. Rev. Allergy Immunol. (2020) 58 342-365. DOI: 10.1007/s12016-019-08769-2 11. Swindle E.J., Metcalfe D.D.. The role of reactive oxygen species and nitric oxide in mast cell-dependent inflammatory processes. Immunol. Rev. (2007) 217 186-205. DOI: 10.1111/j.1600-065X.2007.00513.x 12. Kelley N., Jeltema D., Duan Y., He Y.. The NLRP3 Inflammasome: An Overview of Mechanisms of Activation and Regulation. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20133328 13. Swanson K.V., Deng M., Ting J.P.. The NLRP3 inflammasome: Molecular activation and regulation to therapeutics. Nat. Rev. Immunol. (2019) 19 477-489. DOI: 10.1038/s41577-019-0165-0 14. Nakamura Y., Kambe N.. Linkage of bacterial colonization of skin and the urticaria-like rash of NLRP3-mediated autoinflammatory syndromes through mast cell-derived TNF-alpha. J. Dermatol. Sci. (2013) 71 83-88. DOI: 10.1016/j.jdermsci.2013.04.009 15. Cordaro M., D’Amico R., Fusco R., Peritore A.F., Genovese T., Interdonato L., Franco G., Arangia A., Gugliandolo E., Crupi R.. Discovering the Effects of Fisetin on NF-kappaB/NLRP-3/NRF-2 Molecular Pathways in a Mouse Model of Vascular Dementia Induced by Repeated Bilateral Carotid Occlusion. Biomedicines (2022) 10. DOI: 10.3390/biomedicines10061448 16. Ngo C., Chereau C., Nicco C., Weill B., Chapron C., Batteux F.. Reactive oxygen species controls endometriosis progression. Am. J. Pathol. (2009) 175 225-234. DOI: 10.2353/ajpath.2009.080804 17. Santanam N., Kavtaradze N., Murphy A., Dominguez C., Parthasarathy S.. Antioxidant supplementation reduces endometriosis-related pelvic pain in humans. Transl. Res. (2013) 161 189-195. DOI: 10.1016/j.trsl.2012.05.001 18. Yang L., Gao Y., Bajpai V.K., El-Kammar H.A., Simal-Gandara J., Cao H., Cheng K.W., Wang M., Arroo R.R.J., Zou L.. Advance toward isolation, extraction, metabolism and health benefits of kaempferol, a major dietary flavonoid with future perspectives. Crit. Rev. Food Sci. Nutr. (2021) 23 1-17. DOI: 10.1080/10408398.2021.1980762 19. Ding H., Li Y., Chen S., Wen Y., Zhang S., Luo E., Li X., Zhong W., Zeng H.. Fisetin ameliorates cognitive impairment by activating mitophagy and suppressing neuroinflammation in rats with sepsis-associated encephalopathy. CNS Neurosci. Ther. (2022) 28 247-258. DOI: 10.1111/cns.13765 20. Khan N., Syed D.N., Ahmad N., Mukhtar H.. Fisetin: A dietary antioxidant for health promotion. Antioxid. Redox Signal. (2013) 19 151-162. DOI: 10.1089/ars.2012.4901 21. Kashyap D., Sharma A., Sak K., Tuli H.S., Buttar H.S., Bishayee A.. Fisetin: A bioactive phytochemical with potential for cancer prevention and pharmacotherapy. Life Sci. (2018) 194 75-87. DOI: 10.1016/j.lfs.2017.12.005 22. Park B.S., Choi N.E., Lee J.H., Kang H.M., Yu S.B., Kim H.J., Kang H.K., Kim I.R.. Crosstalk between Fisetin-induced Apoptosis and Autophagy in Human Oral Squamous Cell Carcinoma. J. Cancer (2019) 10 138-146. DOI: 10.7150/jca.28500 23. Sundarraj K., Raghunath A., Panneerselvam L., Perumal E.. Fisetin Inhibits Autophagy in HepG2 Cells via PI3K/Akt/mTOR and AMPK Pathway. Nutr. Cancer (2021) 73 2502-2514. DOI: 10.1080/01635581.2020.1836241 24. Zhang S., Xue R., Geng Y., Wang H., Li W.. Fisetin Prevents HT22 Cells from High Glucose-Induced Neurotoxicity via PI3K/Akt/CREB Signaling Pathway. Front. Neurosci. (2020) 14 241. DOI: 10.3389/fnins.2020.00241 25. Ren Q., Guo F., Tao S., Huang R., Ma L., Fu P.. Flavonoid fisetin alleviates kidney inflammation and apoptosis via inhibiting Src-mediated NF-kappaB p65 and MAPK signaling pathways in septic AKI mice. Biomed. Pharmacother. (2020) 122 109772. DOI: 10.1016/j.biopha.2019.109772 26. Chenxu G., Xianling D., Qin K., Linfeng H., Yan S., Mingxin X., Jun T., Minxuan X.. Fisetin protects against high fat diet-induced nephropathy by inhibiting inflammation and oxidative stress via the blockage of iRhom2/NF-kappaB signaling. Int. Immunopharmacol. (2021) 92 107353. DOI: 10.1016/j.intimp.2020.107353 27. Kumar R., Kumar R., Khurana N., Singh S.K., Khurana S., Verma S., Sharma N., Kapoor B., Vyas M., Khursheed R.. Enhanced oral bioavailability and neuroprotective effect of fisetin through its SNEDDS against rotenone-induced Parkinson’s disease rat model. Food Chem. Toxicol. (2020) 144 111590. DOI: 10.1016/j.fct.2020.111590 28. Jiang K., Yang J., Xue G., Dai A., Wu H.. Fisetin Ameliorates the Inflammation and Oxidative Stress in Lipopolysaccharide-Induced Endometritis. J. Inflamm. Res. (2021) 14 2963-2978. DOI: 10.2147/JIR.S314130 29. Prasath G.S., Subramanian S.P.. Modulatory effects of fisetin, a bioflavonoid, on hyperglycemia by attenuating the key enzymes of carbohydrate metabolism in hepatic and renal tissues in streptozotocin-induced diabetic rats. Eur. J. Pharmacol. (2011) 668 492-496. DOI: 10.1016/j.ejphar.2011.07.021 30. Touil Y.S., Auzeil N., Boulinguez F., Saighi H., Regazzetti A., Scherman D., Chabot G.G.. Fisetin disposition and metabolism in mice: Identification of geraldol as an active metabolite. Biochem. Pharmacol. (2011) 82 1731-1739. DOI: 10.1016/j.bcp.2011.07.097 31. Krishnakumar I.M., Jaja-Chimedza A., Joseph A., Balakrishnan A., Maliakel B., Swick A.. Enhanced bioavailability and pharmacokinetics of a novel hybrid-hydrogel formulation of fisetin orally administered in healthy individuals: A randomised double-blinded comparative crossover study. J. Nutr. Sci. (2022) 11 e74. DOI: 10.1017/jns.2022.72 32. Farsad-Naeimi A., Alizadeh M., Esfahani A., Darvish Aminabad E.. Effect of fisetin supplementation on inflammatory factors and matrix metalloproteinase enzymes in colorectal cancer patients. Food Funct. (2018) 9 2025-2031. DOI: 10.1039/C7FO01898C 33. Wang L., Cao D., Wu H., Jia H., Yang C., Zhang L.. Fisetin Prolongs Therapy Window of Brain Ischemic Stroke Using Tissue Plasminogen Activator: A Double-Blind Randomized Placebo-Controlled Clinical Trial. Clin. Appl. Thromb. Hemost. (2019) 25 1076029619871359. DOI: 10.1177/1076029619871359 34. Zhu Y., Doornebal E.J., Pirtskhalava T., Giorgadze N., Wentworth M., Fuhrmann-Stroissnigg H., Niedernhofer L.J., Robbins P.D., Tchkonia T., Kirkland J.L.. New agents that target senescent cells: The flavone, fisetin, and the BCL-X. Aging (2017) 9 955-963. DOI: 10.18632/aging.101202 35. Yousefzadeh M.J., Zhu Y., McGowan S.J., Angelini L., Fuhrmann-Stroissnigg H., Xu M., Ling Y.Y., Melos K.I., Pirtskhalava T., Inman C.L.. Fisetin is a senotherapeutic that extends health and lifespan. EBioMedicine (2018) 36 18-28. DOI: 10.1016/j.ebiom.2018.09.015 36. Binda M.M., Donnez J., Dolmans M.M.. Targeting mast cells: A new way to treat endometriosis. Expert Opin. Ther. Targets (2017) 21 67-75. DOI: 10.1080/14728222.2017.1260548 37. Scutiero G., Iannone P., Bernardi G., Bonaccorsi G., Spadaro S., Volta C.A., Greco P., Nappi L.. Oxidative Stress and Endometriosis: A Systematic Review of the Literature. Oxid. Med. Cell. Longev. (2017) 2017 7265238. DOI: 10.1155/2017/7265238 38. Wu M.Y., Ho H.N.. The role of cytokines in endometriosis. Am. J. Reprod. Immunol. (2003) 49 285-296. DOI: 10.1034/j.1600-0897.2003.01207.x 39. Tseng H.W., Samuel S.G., Schroder K., Levesque J.P., Alexander K.A.. Inflammasomes and the IL-1 Family in Bone Homeostasis and Disease. Curr. Osteoporos. Rep. (2022) 20 170-185. DOI: 10.1007/s11914-022-00729-8 40. Yu Y., Blokhuis B.R., Garssen J., Redegeld F.A.. Non-IgE mediated mast cell activation. Eur. J. Pharmacol. (2016) 778 33-43. DOI: 10.1016/j.ejphar.2015.07.017 41. Zaitsu M., Narita S., Lambert K.C., Grady J.J., Estes D.M., Curran E.M., Brooks E.G., Watson C.S., Goldblum R.M., Midoro-Horiuti T.. Estradiol activates mast cells via a non-genomic estrogen receptor-alpha and calcium influx. Mol. Immunol. (2007) 44 1977-1985. DOI: 10.1016/j.molimm.2006.09.030 42. Lin K.Q., Zhu L.B., Zhang X.M., Lin J.. Role of mast cells in estrogen-mediated experimental endometriosis in rats. Zhejiang Da Xue Xue Bao Yi Xue Ban (2015) 44 269-277. DOI: 10.3785/j.issn.1008-9292.2015.05.06 43. Lousse J.C., Van Langendonckt A., Defrere S., Ramos R.G., Colette S., Donnez J.. Peritoneal endometriosis is an inflammatory disease. Front. Biosci. (2012) 4 23-40. DOI: 10.2741/e358 44. Van Langendonckt A., Casanas-Roux F., Donnez J.. Oxidative stress and peritoneal endometriosis. Fertil. Steril. (2002) 77 861-870. DOI: 10.1016/S0015-0282(02)02959-X 45. Currais A., Prior M., Dargusch R., Armando A., Ehren J., Schubert D., Quehenberger O., Maher P.. Modulation of p25 and inflammatory pathways by fisetin maintains cognitive function in Alzheimer’s disease transgenic mice. Aging Cell (2014) 13 379-390. DOI: 10.1111/acel.12185 46. Shi Y.S., Li C.B., Li X.Y., Wu J., Li Y., Fu X., Zhang Y., Hu W.Z.. Fisetin Attenuates Metabolic Dysfunction in Mice Challenged with a High-Fructose Diet. J. Agric. Food Chem. (2018) 66 8291-8298. DOI: 10.1021/acs.jafc.8b02140 47. Subramanian P., Jayakumar M., Jayapalan J.J., Hashim O.H.. Chronotherapeutic effect of fisetin on expression of urea cycle enzymes and inflammatory markers in hyperammonaemic rats. Pharmacol. Rep. (2014) 66 1037-1042. DOI: 10.1016/j.pharep.2014.06.018 48. Yen J.H., Wu P.S., Chen S.F., Wu M.J.. Fisetin Protects PC12 Cells from Tunicamycin-Mediated Cell Death via Reactive Oxygen Species Scavenging and Modulation of Nrf2-Driven Gene Expression, SIRT1 and MAPK Signaling in PC12 Cells. Int. J. Mol. Sci. (2017) 18. DOI: 10.3390/ijms18040852 49. Dong W., Jia C., Li J., Zhou Y., Luo Y., Liu J., Zhao Z., Zhang J., Lin S., Chen Y.. Fisetin Attenuates Diabetic Nephropathy-Induced Podocyte Injury by Inhibiting NLRP3 Inflammasome. Front. Pharmacol. (2022) 13 783706. DOI: 10.3389/fphar.2022.783706 50. Huang X., Shen H., Liu Y., Qiu S., Guo Y.. Fisetin attenuates periodontitis through FGFR1/TLR4/NLRP3 inflammasome pathway. Int. Immunopharmacol. (2021) 95 107505. DOI: 10.1016/j.intimp.2021.107505 51. Goud P.T., Goud A.P., Joshi N., Puscheck E., Diamond M.P., Abu-Soud H.M.. Dynamics of nitric oxide, altered follicular microenvironment, and oocyte quality in women with endometriosis. Fertil. Steril. (2014) 102 151-159.e5. DOI: 10.1016/j.fertnstert.2014.03.053 52. Nasiri N., Moini A., Eftekhari-Yazdi P., Karimian L., Salman-Yazdi R., Arabipoor A.. Oxidative Stress Statues in Serum and Follicular Fluid of Women with Endometriosis. Cell J. (2017) 18 582-587. DOI: 10.22074/cellj.2016.4724 53. Martinon F., Burns K., Tschopp J.. The inflammasome: A molecular platform triggering activation of inflammatory caspases and processing of proIL-beta. Mol. Cell (2002) 10 417-426. DOI: 10.1016/S1097-2765(02)00599-3 54. Jo E.K., Kim J.K., Shin D.M., Sasakawa C.. Molecular mechanisms regulating NLRP3 inflammasome activation. Cell. Mol. Immunol. (2016) 13 148-159. DOI: 10.1038/cmi.2015.95 55. Murakami M., Osuka S., Muraoka A., Hayashi S., Bayasula Y., Kasahara R., Sonehara Y., Hariyama K., Shinjo H.. Effectiveness of NLRP3 Inhibitor as a Non-Hormonal Treatment for ovarian endometriosis. Reprod. Biol. Endocrinol. (2022) 20 58. DOI: 10.1186/s12958-022-00924-3 56. Abd El-Twab S.M., Hussein O.E., Hozayen W.G., Bin-Jumah M., Mahmoud A.M.. Chicoric acid prevents methotrexate-induced kidney injury by suppressing NF-kappaB/NLRP3 inflammasome activation and up-regulating Nrf2/ARE/HO-1 signaling. Inflamm. Res. (2019) 68 511-523. DOI: 10.1007/s00011-019-01241-z 57. Molagoda I.M.N., Athapaththu A., Choi Y.H., Park C., Jin C.Y., Kang C.H., Lee M.H., Kim G.Y.. Fisetin Inhibits NLRP3 Inflammasome by Suppressing TLR4/MD2-Mediated Mitochondrial ROS Production. Antioxidants (2021) 10. DOI: 10.3390/antiox10081215 58. Hassan S.S.U., Samanta S., Dash R., Karpinski T.M., Habibi E., Sadiq A., Ahmadi A., Bunagu S.. The neuroprotective effects of fisetin, a natural flavonoid in neurodegenerative diseases: Focus on the role of oxidative stress. Front. Pharmacol. (2022) 13 1015835. DOI: 10.3389/fphar.2022.1015835 59. Di Nisio V., Rossi G., Di Luigi G., Palumbo P., D’Alfonso A., Iorio R., Cecconi S.. Increased levels of proapoptotic markers in normal ovarian cortex surrounding small endometriotic cysts. Reprod. Biol. (2019) 19 225-229. DOI: 10.1016/j.repbio.2019.08.002 60. Delbandi A.A., Mahmoudi M., Shervin A., Heidari S., Kolahdouz-Mohammadi R., Zarnani A.H.. Evaluation of apoptosis and angiogenesis in ectopic and eutopic stromal cells of patients with endometriosis compared to non-endometriotic controls. BMC Women’s Health (2020) 20. DOI: 10.1186/s12905-019-0865-4 61. Dmowski W.P., Ding J., Shen J., Rana N., Fernandez B.B., Braun D.P.. Apoptosis in endometrial glandular and stromal cells in women with and without endometriosis. Hum. Reprod. (2001) 16 1802-1808. DOI: 10.1093/humrep/16.9.1802 62. Sun Y., Che X., Zhu L., Zhao M., Fu G., Huang X., Xu H., Hu F., Zhang X.. Pigment epithelium derived factor inhibits the growth of human endometrial implants in nude mice and of ovarian endometriotic stromal cells in vitro. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0045223 63. Abbate J.M., Macri F., Arfuso F., Iaria C., Capparucci F., Anfuso C., Ieni A., Cicero L., Briguglio G., Lanteri G.. Anti-Atherogenic Effect of 10% Supplementation of Anchovy (. Nutrients (2021) 13. DOI: 10.3390/nu13072137 64. Mullane K.M., Kraemer R., Smith B.. Myeloperoxidase activity as a quantitative assessment of neutrophil infiltration into ischemic myocardium. J. Pharmacol. Methods (1985) 14 157-167. DOI: 10.1016/0160-5402(85)90029-4 65. Fusco R., D’Amico R., Cordaro M., Gugliandolo E., Siracusa R., Peritore A.F., Crupi R., Impellizzeri D., Cuzzocrea S., Di Paola R.. Absence of formyl peptide receptor 1 causes endometriotic lesion regression in a mouse model of surgically-induced endometriosis. Oncotarget (2018) 9 31355-31366. DOI: 10.18632/oncotarget.25823 66. Gugliandolo E., Peritore A.F., D’Amico R., Licata P., Crupi R.. Evaluation of Neuroprotective Effects of Quercetin against Aflatoxin B1-Intoxicated Mice. Animals (2020) 10. DOI: 10.3390/ani10050898 67. Menegazzi M., Di Paola R., Mazzon E., Genovese T., Crisafulli C., Dal Bosco M., Zou Z., Suzuki H., Cuzzocrea S.. Glycyrrhizin attenuates the development of carrageenan-induced lung injury in mice. Pharmacol. Res. (2008) 58 22-31. DOI: 10.1016/j.phrs.2008.05.012 68. Cuzzocrea S., Nocentini G., Di Paola R., Agostini M., Mazzon E., Ronchetti S., Crisafulli C., Esposito E., Caputi A.P., Riccardi C.. Proinflammatory role of glucocorticoid-induced TNF receptor-related gene in acute lung inflammation. J. Immunol. (2006) 177 631-641. DOI: 10.4049/jimmunol.177.1.631 69. Petrosino S., Campolo M., Impellizzeri D., Paterniti I., Allara M., Gugliandolo E., D’Amico R., Siracusa R., Cordaro M., Esposito E.. 2-Pentadecyl-2-Oxazoline, the Oxazoline of Pea, Modulates Carrageenan-Induced Acute Inflammation. Front. Pharmacol. (2017) 8 308. DOI: 10.3389/fphar.2017.00308 70. Peritore A.F., D’Amico R., Siracusa R., Cordaro M., Fusco R., Gugliandolo E., Genovese T., Crupi R., Di Paola R., Cuzzocrea S.. Management of Acute Lung Injury: Palmitoylethanolamide as a New Approach. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22115533 71. Impellizzeri D., Siracusa R., Cordaro M., Peritore A.F., Gugliandolo E., D’Amico R., Fusco R., Crupi R., Rizzarelli E., Cuzzocrea S.. Protective effect of a new hyaluronic acid -carnosine conjugate on the modulation of the inflammatory response in mice subjected to collagen-induced arthritis. Biomed. Pharmacother. (2020) 125 110023. DOI: 10.1016/j.biopha.2020.110023 72. Impellizzeri D., Siracusa R., Cordaro M., Crupi R., Peritore A.F., Gugliandolo E., D’Amico R., Petrosino S., Evangelista M., Di Paola R.. N-Palmitoylethanolamine-oxazoline (PEA-OXA): A new therapeutic strategy to reduce neuroinflammation, oxidative stress associated to vascular dementia in an experimental model of repeated bilateral common carotid arteries occlusion. Neurobiol. Dis. (2019) 125 77-91. DOI: 10.1016/j.nbd.2019.01.007 73. Impellizzeri D., Siracusa R., Cordaro M., Peritore A.F., Gugliandolo E., Mancuso G., Midiri A., Di Paola R., Cuzzocrea S.. Therapeutic potential of dinitrobenzene sulfonic acid (DNBS)-induced colitis in mice by targeting IL-1beta and IL-18. Biochem. Pharmacol. (2018) 155 150-161. DOI: 10.1016/j.bcp.2018.06.029 74. D’Amico R., Cordaro M., Fusco R., Peritore A.F., Genovese T., Gugliandolo E., Crupi R., Mandalari G., Caccamo D., Cuzzocrea S.. Consumption of Cashew (. Nutrients (2022) 14. DOI: 10.3390/nu14071474 75. Impellizzeri D., Peritore A.F., Cordaro M., Gugliandolo E., Siracusa R., Crupi R., D’Amico R., Fusco R., Evangelista M., Cuzzocrea S.. The neuroprotective effects of micronized PEA (PEA-m) formulation on diabetic peripheral neuropathy in mice. FASEB J. (2019) 33 11364-11380. DOI: 10.1096/fj.201900538R 76. D’Amico R., Fusco R., Cordaro M., Siracusa R., Peritore A.F., Gugliandolo E., Crupi R., Scuto M., Cuzzocrea S., Di Paola R.. Modulation of NLRP3 Inflammasome through Formyl Peptide Receptor 1 (Fpr-1) Pathway as a New Therapeutic Target in Bronchiolitis Obliterans Syndrome. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21062144 77. Di Paola R., Impellizzeri D., Salinaro A.T., Mazzon E., Bellia F., Cavallaro M., Cornelius C., Vecchio G., Calabrese V., Rizzarelli E.. Administration of carnosine in the treatment of acute spinal cord injury. Biochem. Pharmacol. (2011) 82 1478-1489. DOI: 10.1016/j.bcp.2011.07.074 78. Cordaro M., Impellizzeri D., Siracusa R., Gugliandolo E., Fusco R., Inferrera A., Esposito E., Di Paola R., Cuzzocrea S.. Effects of a co-micronized composite containing palmitoylethanolamide and polydatin in an experimental model of benign prostatic hyperplasia. Toxicol. Appl. Pharmacol. (2017) 329 231-240. DOI: 10.1016/j.taap.2017.06.005 79. Fusco R., Salinaro A.T., Siracusa R., D’Amico R., Impellizzeri D., Scuto M., Ontario M.L., Crea R., Cordaro M., Cuzzocrea S.. Hidrox((R)) Counteracts Cyclophosphamide-Induced Male Infertility through NRF2 Pathways in a Mouse Model. Antioxidants (2021) 10. DOI: 10.3390/antiox10050778 80. Cordaro M., Paterniti I., Siracusa R., Impellizzeri D., Esposito E., Cuzzocrea S.. KU0063794, a Dual mTORC1 and mTORC2 Inhibitor, Reduces Neural Tissue Damage and Locomotor Impairment After Spinal Cord Injury in Mice. Mol. Neurobiol. (2017) 54 2415-2427. DOI: 10.1007/s12035-016-9827-0 81. Petrosino S., Cordaro M., Verde R., Schiano Moriello A., Marcolongo G., Schievano C., Siracusa R., Piscitelli F., Peritore A.F., Crupi R.. Oral Ultramicronized Palmitoylethanolamide: Plasma and Tissue Levels and Spinal Anti-hyperalgesic Effect. Front. Pharmacol. (2018) 9 249. DOI: 10.3389/fphar.2018.00249 82. Di Paola R., Fusco R., Impellizzeri D., Cordaro M., Britti D., Morittu V.M., Evangelista M., Cuzzocrea S.. Adelmidrol, in combination with hyaluronic acid, displays increased anti-inflammatory and analgesic effects against monosodium iodoacetate-induced osteoarthritis in rats. Arthritis Res. Ther. (2016) 18 291. DOI: 10.1186/s13075-016-1189-5 83. Cordaro M., Siracusa R., Fusco R., D’Amico R., Peritore A.F., Gugliandolo E., Genovese T., Scuto M., Crupi R., Mandalari G.. Cashew (. Antioxidants (2020) 9. DOI: 10.3390/antiox9080660 84. Genovese T., Mazzon E., Muia C., Bramanti P., De Sarro A., Cuzzocrea S.. Attenuation in the evolution of experimental spinal cord trauma by treatment with melatonin. J. Pineal Res. (2005) 38 198-208. DOI: 10.1111/j.1600-079X.2004.00194.x
--- title: The Evaluation of Energy Availability and Dietary Nutrient Intake of Sport Climbers at Different Climbing Levels authors: - Anna Chmielewska - Bożena Regulska-Ilow journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049433 doi: 10.3390/ijerph20065176 license: CC BY 4.0 --- # The Evaluation of Energy Availability and Dietary Nutrient Intake of Sport Climbers at Different Climbing Levels ## Abstract Proper nutrition is the basis for athletes’ performances when competing or training. The increasing training volume accompanying the increase in advancement should go hand in hand with the appropriate supply of energy as well as macro and micronutrients. The diet of climbing representatives due to the desire to achieve a low body weight may be deficient in energy and micronutrients. Our study aimed to evaluate the differences in energy availability and nutrient intake of female and male sport climbers at different climbing levels. Anthropometric parameters and the resting metabolic rate were measured, the questionnaire about climbing grade and training hours was filled, and a 3-day food diary was fulfilled by 106 sport climbers. Based on the collected data, the energy availability as well as the macro- and micronutrient intake was calculated. Low energy availability (EA) was observed among both genders of sport climbing representatives. A significant difference between EA in various levels of advancement was found in the male group ($p \leq 0.001$). Differences in carbohydrate intake (g/kg/BW) between sexes were observed ($$p \leq 0.01$$). Differences in nutrients intake between climbing grade were found in both the female and male groups. In the group of female elite athletes, the adequate supply of most of the micronutrients can imply a high-quality diet despite the low calorie content. It is necessary to educate sport climbing representatives about the importance of proper nutrition as well as the consequences of insufficient energy intake. ## 1. Introduction Sport climbing is rapidly growing in popularity, and according to the International Sport Climbing Federation, approximately 25 million people practice it globally. In the United States alone, as many as 1000–1500 new people try their hand at the sport every day. Furthermore, it is estimated that approximately 100,000 people practice the sport regularly in over 150 indoor climbing facilities in Poland [1]. Sport climbing was included in the Olympics for the first time in 2021 and comprised three disciplines, namely bouldering, speed climbing, and lead climbing [2]. Bouldering and lead climbing are the most frequently practiced disciplines [3]. Bouldering is performed rope-less with spotters and crash mats for protection, and indoor walls typically do not exceed four meters in height [4]. Lead climbing is a discipline based on leading long routes (usually 20–40 m) using a rope and quickdraws clipped to bolts placed in the rock or an artificial wall [5]. Characterized by various dynamic movements under conditions of compensated fatigue and changes in work intensity, sport climbing is considered a complex sport [6]. Climbing requires a significant proportion of whole-body aerobic capacity, with the anaerobic power more important for challenging routes with steeper angles [7]. All types of climbing require a high level of technique, strength, and endurance. However, during longer climbing routes, stamina is emphasized, whereas bouldering relies on power [8]. Sport climbers present with specific anthropometric characteristics similar to the profiles of ballet dancers and long-distance runners. Climbers are typically shorter, leaner, and lighter than non-climbing athletes, and elite climbers are short in stature and have low body mass and body fat [3]. A high power-to-mass ratio (SMR) is considered advantageous in rock climbing, and losing weight to tackle more challenging routes is reported among sports climbers [3]. However, consuming a diet too low in energy can place climbers at risk of the inadequate intake of nutrients such as carbohydrates (CHOs), protein, calcium, and iron. It can also cause fatigue and a weakened immune system [6]. As such, these behaviors put climbers at risk of restrictive eating and increase the potential for eating disorders (ED), disordered eating (DE), laxative/diuretic use, and relative energy deficiency in sport (RED-S) [3]. Proper nutrition is the basis for an athlete’s performance when competing or training. Indeed, a well-balanced diet provides energy when athletes perform various physical activities and is also necessary for post-match recovery [9]. An increasing number of studies have investigated the aspects of factors that contribute to a successful climbing performance, such as anthropometric parameters, bio-mechanics, and physiological and psychological factors [10]. However, the number of studies evaluating energy intake (EI) and assessing the dietary intake of sports climbers is still limited. Some studies involving small groups of climbers from different countries analyzed the intake of energy, macronutrients, and micronutrients [11,12,13], diet composition [14,15,16], and antioxidant intake [17]. Some considerations have been given to the nutritional requirements for Olympic-style climbing [3] and bouldering [4]. Thus far, research has found inadequate energy availability (EA) [7], unbalanced food quality, and poor nutrient timing in the diets of climbers of different ages and levels of advancement [18]. Furthermore, several studies have investigated supplement use in sports climbers [12,19,20] and analyzed supplement protocol use [21,22]. However, most of those studies involved a small group of top-level climbers, competitive representatives, training in one sport object. Our study aimed to involve a greater number of climbers at different levels of advancement, training on numerous sport facilities, which gives a bigger possibility to observe the differences in the general population of climbers of various grades. This study aimed to assess the dietary nutritional intake of sports climbers, investigate whether it matches the recommendations, and understand the main dietary deficiencies that occur. Along with the level of advancement of sports climbing and the growing number of training methods, the focus should shift to different factors that may influence the performance, including proper nutrition. It was assumed that differences would exist between the dietary nutrient intake of climbers of different levels, with a tendency for elite climbers to have a better nutrition pattern that included higher vitamin and mineral components. On the other hand, it was assumed, as in previous studies, that sport climbers tend to lower their EI to obtain a low body mass. This can be a reason for malnutrition and insufficient EA. This study aimed to evaluate the differences in the EA and the macronutrient and micronutrient intake of sports climbers, females and males, at various climbing levels. ## 2.1. Design Study information was spread through social media and directly through the trainers of the five climbing gyms active at the time. The inclusion criterion for the study was at least one year of regular sports climbing. The exclusion criterion was recreational climbing with practice less than once a week. All climbers who were willing to take part for the duration of the study and who fulfilled the criteria were accepted. Data collection took place between June 2019 and August 2021 with the obligatory break during the coronavirus pandemic between January 2020 and February 2021. As the level of advancement of participants was set based on the climbing routes crossing from the previous 6 months, the study performance was refrained for a longer period after the opening of sports facilities and the end of the strict quarantine, so the data were reliable. The Institutional Ethics Committee of Wroclaw Medical University approved the study (number KB-$\frac{45}{2019}$), which was conducted in accordance with the Declaration of Helsinki. ## 2.2. Procedure The procedures of the indirect calorimetry test, anthropometric measurements, and filling the questionnaire were carried out in a special design area in the Department of Dietetics, with the assistance of a trained nutritionist. It was requested that the study participants attend after fasting, an overnight rest, and having avoided of intense physical activity in the evening before the measurements were taken. The study participants were asked to come during the morning hours. Resting heart rate, systolic blood pressure, and diastolic blood pressure were measured using a sphygmomanometer to exclude the performance of indirect calorimetry in a state of anxiety. Anthropometric measurements, including body weight (BW), body fat, and fat-free mass (FFM), were performed with the X-CONTACT 356 analyzer (Jawon Medical Co., Seoul, Korea), and height was measured using a TANITA HR-001 stadiometer (TANITA, Tokyo, Japan). The energy needs of the study participants were determined by measuring their resting metabolic rate (RMR) with indirect calorimetry (IC) using the Fitmate WM (Cosmed, Rome, Italy) device. It was requested that the study participants attend after fasting, an overnight rest, and the avoidance of intense physical activity in the evening before the measurements were taken. The test took place in a darkened and soundproof room and lasted for 12 min. Participants’ breathing patterns were matched to the measurement requirements in the first two minutes, with the RMR measurement taking ten minutes. Contraction force measurement used a MAP 80K1 hand grip dynamometer (KERN & SOHN GmbH, Balingen, Germany). The contraction force of each hand was measured twice, and the mean value of two measurements was calculated. After performing the measurements, participants filled out the questionnaire. Each participant’s climbing grade was subjectively determined by assessing three different routes established over the preceding six months. Participants were asked to indicate the highest grade, lead, or boulder that they had managed to redpoint on three different routes/problems on either an artificial wall or a rock. Participants also declared their frequency and time of training in the week’s perspective. Next, each participant was asked to deliver a three-day food diary using the Fitatu mobile application [23]. Participants were instructed to enter each meal in the application, as it gives the possibility to convert the house measures of food products for grams. The application allows to scan products based on barcodes and contains information about suggested product portions, which makes it easier to fill in the diary. Photographs of the meals consumed supplemented the data and provided accurate information on portion sizes. In case of any doubts according to product and meal types and sizes, the participants were contacted to specify the information. ## 2.3. Participants One hundred and fourteen regular sports climbers, who were willing to take a part in the study and fulfilled the criteria, which were from Wroclaw in Poland, took part in the study. Of the study participants, 106 (40 females and 66 males) delivered dietary reports suitable for analysis. ## 2.4. Outcome Measures Based on the indirect calorimetry, blood pressure, and anthropometric and contraction force measures, the characteristics of the group were established. BMI value was calculated using body mass measured with the bioimpedance balance and height measured with a stadiometer mentioned in the design part. For the grip strength-to-body mass ratio (SMR), the mean value of all contraction force measures was used. The grade declared in the questionnaire was standardized according to the International Rock Climbing Research Association (IRCRA) scale [24]. For female climbers, an Intermediate score was between 10 and 14, an advanced score was between 15 and 20, and an elite score was between 21 and 26. For male climbers, an intermediate score was between 10 and 17, an advanced score was between 18 and 23, and an elite score was between 24 and 27. To assess the energy availability, the measurement of the energy exercise expenditure (EEE) was calculated based on the training time declared in the questionnaire and the activity logs using the metabolic equivalent of task (MET) [25]. Based on the obtained calculations and fat-free mass (FFM) value obtained with the measurements with the bioimpedance scale, the EA was calculated. EA was calculated using the following equation [26]:EA =EI – EEEFFM To assess the macro- and micronutrients intake, the food records were transferred to ESHA’s Food Processor® Nutrition Analysis software 11.7.217 database 11.7.1 (ESHA Research, Salem, OR, USA), which contains the Nutrient Tables of Foods database [27]. For the macronutrient intake analysis, the dietary macronutrient intake was converted to g/kg/BW as the general recommendations for athletes are expressed as g/kg/BW. Dietary intake was compared to Norms for the Polish Population from 2020 [28] for the female and male groups separately, as the nutritional recommendations are gender-specific. ## Statistical Analysis Considering the small sample size, and therefore the small power of a normality test, the non-parametric tests were used. The differences in the dietary intake of macronutrients and micronutrients and EA between the three climber groups (intermediate, advanced, and elite) were assessed using the Kruskal–Wallis test, and the differences between genders were assessed using the Mann–Whitney U test. Differences were recognized as statistically significant when $p \leq 0.05.$ For statistically significant differences, a post hoc Dunn test with Bonferroni correction assessed the differences between the study groups. Descriptive statistics were generated for the anthropometric measurements of the study group and the dietary intake of macronutrients and micronutrients. Data analysis was performed with STATISTICA 13.1 (StatSoft Inc., Tulsa, OK, USA). ## 3. Results The characteristics of the study groups based on the gender and climbing level are presented in Table 1. Significant differences were observed in systolic ($$p \leq 0.03$$ for elite vs. intermediate; $$p \leq 0.01$$ for elite vs. advanced) and diastolic ($$p \leq 0.03$$ for elite vs. intermediate) blood pressure in female climbing level groups. In males, differences in SMR ($p \leq 0.01$ for elite vs. intermediate; $$p \leq 0.03$$ for elite vs. advanced), contraction force of the left hand ($$p \leq 0.03$$ for elite vs. intermediate), and body mass index (BMI) ($$p \leq 0.04$$ for elite vs. intermediate) existed between climbing levels. Table 2 presents EA in the groups of male and female climbers of different climbing levels. EA decreased with a higher climbing level in both groups, which was significant in males but non-significant in females. Among the males, there were significant differences between the post hoc tests of the elite vs. intermediate groups ($$p \leq 0.04$$; $Z = 2.47$) and advanced vs. intermediate groups ($$p \leq 0.01$$; $Z = 2.86$). Moreover, EA was significantly different between the genders. Comparing the genders in each climbing level revealed significant differences in the advanced group ($$p \leq 0.04$$; $Z = 2.04$) with no significant differences between the intermediate ($$p \leq 0.12$$; $Z = 1.54$) and elite ($$p \leq 0.8$$; $Z = 0.21$) groups. Suboptimal EA (>30 kcal/kg FFM/day) was reached in $73\%$ of the intermediate, $45\%$ of the advanced, and $40\%$ of the elite female group. In male groups, only $47\%$ of intermediate participants and $21\%$ of the advanced group reached a suboptimal EA level. None of the elite male climbers exceeded the value of 30 kcal/kg/FFM. In the intermediate and advanced level, median EA values in male groups were lower than in female groups. The dietary intake of macronutrients in male and female groups on different climbing levels is presented in Table 3. No significant differences were found between climbing levels, although differences were found in the CHO intake between the male and female cohort. However, when measuring differences in the CHO intake between genders in each climbing level, marginally significant differences were found in the intermediate ($$p \leq 0.06$$; Z = −1.88) and the elite ($$p \leq 0.06$$; Z = −1.88) group, with no significance in the advanced group ($$p \leq 0.13$$; Z = −1.50) The dietary intake of macro and micronutrients of male and female climbers on different climbing levels is presented in Table 4. In females, the differences were observed in fiber ($$p \leq 0.007$$; $Z = 3.06$), iron ($$p \leq 0.003$$; $Z = 3.33$), magnesium ($$p \leq 0.008$$; $Z = 2.99$), phosphorus ($$p \leq 0.04$$; $Z = 2.46$), and potassium ($$p \leq 0.05$$; $Z = 2.43$) intake between the elite and advanced groups. Difference in iodine ($$p \leq 0.01$$; $Z = 2.89$) intake was found between advanced and intermediate groups. According to zinc intake in the post hoc test, no significant differences were confirmed. Males in the advanced and intermediate groups differed significantly in vitamin B6 ($$p \leq 0.4$$; $Z = 2.50$), phosphorus ($$p \leq 0.02$$; $Z = 2.72$), iron ($$p \leq 0.3$$; $Z = 2.60$), and zinc ($$p \leq 0.002$$; $Z = 3.42$) intake. According in the post hoc test, no significant differences in potassium intake were confirmed. Compared to the estimated average requirement (EAR) and adequate intake (AI) values for vitamin D, vitamin A, sodium, and potassium for the Polish population, the proper intake was reported for most micronutrients in all study groups. However, all groups reported an insufficient intake of vitamin D and iodine. In the advanced and intermediate female group, insufficient amounts of calcium and potassium were also reported. The elite group of female climbers presented the lowest median EI, with the median value intake of most of the micronutrients fulfilling the recommendations. Elite and advanced males reported inadequate intakes of potassium. ## 4. Discussion The study results partly confirmed the hypothesis on the differences between diet quality between the participants with different levels of advancement. Indeed, elite female climbers consumed significantly higher amounts of several nutrients than lower-level climbers. However, the male climber groups did not represent such a characteristic. None of the elite male group reached the suboptimal EA. However, the lowest median value of EA was reported in the elite female group. The anthropometric measurements of body mass and height for both sexes were similar to the participants of other studies involving sports climbers [11,12,15,16]. However, the body fat percentage was higher in the male groups [12,16]. Climbers from the elite groups had the lowest BMI values, which were 20.1 in females and 21.2 in males. In a study by Sas-Nowosielski et al. [ 11], the BMI of climbers significantly predicted their climbing abilities on the most challenging route climbed in the so-called redpoint style. Kemmler et al. [ 16] demonstrated a moderate negative effect of low BMI value on bone mineral density (BMD). Nonetheless, climbers had higher BDM values in most studied regions compared to non-climbing controls with the same BMI. Studies investigating EI in adolescents [29] and adult sports climbers [11,16,18] to date have reported the caloric value of the assessed diets to be too low for the predicted energy expenditure. Furthermore, few studies have focused on whether the dietary caloric intake provided enough EA in relation to the training volume. EA is the dietary EI minus the energy expended during exercise, with EA being an input to the body’s physiological systems that remain after exercise training [18]. Low EA occurs when either the dietary EI is too low or the energy expanded through exercise is too high, and the energy needed for the maintenance of physiological functions such as bone health, the menstrual cycle, and metabolic and immune function becomes insufficient. There is no optimal EA for high-performance athletes. To date, in studies of sedentary normal-weight females, 45 kcal/kg FFM/ day was defined as a value for achieving optimal energy balance. Meanwhile, a study on men reported that 40 kcal/kg FFM/ day was enough to support energy balance. An EA of 30–45 kcal/kg FFM/ day is already considered a reduced EA, and athletes should only stay within this value for a short period, such as when aiming to reduce BW [30]. The analysis of the entire cohort in the current study showed that $36\%$ presented with suboptimal EA (<45 kcal/kg FFM/day) and $59\%$ had low EA (LEA; <30 kcal/kg FFM/day). LEA was demonstrated by $45\%$ of female participants and $68\%$ of male climbers. Median EA was the lowest in the female elite group (13.4). In male groups, there was a significant difference in EA between the elite and intermediate groups and between the advanced and intermediate groups. In the study by Monedero et al. [ 13], suboptimal EA and LEA were evident in $88\%$ and $28\%$ of climbers, respectively. The prevalence of suboptimal EA was $93\%$ in male subjects and $82\%$ in female subjects, and the prevalence of LEA was $29\%$ in the male subjects and $27\%$ in the female subjects. In a study by Simič et al. assessing EA in adolescent climbers [7], the mean EA (27.5 ± 9.8 kcal/kg FFM/day) was below the recommended level, with no participant meeting the target of 45 kcal/kg FFM/day and $63\%$ being in the range of LEA. Furthermore, $26\%$ of the climbers failed to meet their predicted basal metabolic rate, and a significant difference in climbing levels was reported between the groups with suboptimal and low EA. Moreover, Michael et al. [ 29] demonstrated that $82\%$ of adolescent climbers did not meet the recommended EI. Monedero et al. [ 13] observed energy intakes that were significantly lower than daily requirements, but only in females. In a study by Gibson-Smith [12], the average EA was 41.4 ± 9 kcal/kg FFM/day, with a significantly higher EA in females than in males (45.6 ± 7 kcal/kg FFM/day vs. 37.2 ± 9 kcal/kg FFM/day, respectively). Furthermore, $78\%$ of the elite adult climbers failed to meet the predicted energy required to support a moderate level of physical activity of 12 h of training per week, while $18\%$ failed to meet the predicted RMR values [12]. In the current study, only the female intermediate group had a suboptimal EA, with a median EA <30 kcal/kg/FFM/day in the other groups. The median EA values were lowest in the elite groups of males and females. The median EIs in the elite female group was lower compared to the elite male group. Those climbing at the highest levels may be involved in competition or in maintaining/acquiring sponsorships. These factors may lead to additional pressure to achieve a lower BW or leanness, which could negatively affect eating patterns [29]. Furthermore, a study by Modaberi et al. [ 31] reported a positive correlation between anxiety, emotional eating, and external eating behavior in top rope climbers. Previous studies analyzing DE reported higher prevalence rates among elite athletes [32]. However, in the studies involving a range of disciplines [33] and those focusing on adolescents [29] and adults [30], females were considered to be at the greatest risk of DE. The results of our study show that the male groups presented a lower median of EAs in intermediate and advanced groups, however, the median of EAs in elite groups was smaller in female group. This is consistent with the results of previous research, which indicated that female groups presented with low EA and DE more often. Indeed, Joubert et al. [ 30] reported a DE prevalence of $6.3\%$ in males and $16.5\%$ in female climbers. Almost half of the females from the elite and high elite groups presented with DE ($42.9\%$), and DE was only significantly associated with climbing ability in the female group. In Sas-Nowosielski et al., relatively more females (8 out of 10) than males (5 out of 13) were dissatisfied with their body mass and felt a need to slim down [11]. Female athletes under-eat for reasons unrelated to sport, with Wardle et al. [ 34] reporting that there are approximately about twice as many young women than men at every decile of BMI perceive themselves to be overweight. Furthermore, $25\%$ of male athletes from esthetic, leanness-focused, or weight-sensitive sports show disordered dietary patterns, which was also closely associated with higher body fat percentages and body dissatisfaction [35]. In the context of exercise-related health risks, LEA has been extensively described in female athletes. However, the International Olympic Committee recently expanded the concept of the sportsmen triad to include the term relative energy deficiency in sports to describe the consequences of LEA on health and performance in both males and females [36]. It is suggested that an EA below 30 kcal/kg FFM/ day can lead to an abrupt decline in bone mineralization, with similar reductions in insulin-like growth factor-1 and tri-iodothyronine concentrations. A low EA is suspected of suppressing Type 1 immunity [29], and males may respond to LEA by developing an exercise hypogonadal male condition (EHMC), which affects reproductive function. During EHMC, the hypothalamic–pituitary–gonadal axis is disturbed, along with reduced serum testosterone levels (TES). Although TES values remain at the low end of the normal clinical range, symptoms of hypogonadism, such as fatigue, sexual dysfunction, and reduced BMD, are present. The syndrome was first diagnosed in endurance-trained males, although it is also seen in athletes of other disciplines, such as bodybuilding and power disciplines. LEA following restricted dietary EI or EEE can also induce metabolic adaptions to conserve energy, including a decline in the basal metabolic rate (BMR), non-exercise activity thermogenesis (NEAT), and the thermic effect of food (TEF) if the caloric intake is restricted [36]. In this study, the elite male group presented the highest median protein intake among the male groups (1.45 (g/kg/BW). Among the female groups, the highest median value was found in the advanced group (1.42 g/kg/BW). The lowest mean intakes among the female and male groups were found for the intermediate female group and the advanced male group, respectively, which were 1.34 g/kg/BW and 1.25 g/kg/BW. The current recommendation for sport climbers is a daily protein intake of 1.3–1.7 g/kg/BW [3] or as much as 2 g/kg/BW for bouldering [4], which suggests that, considering the median values, with the exception of the male advanced group, other groups have fulfilled the minimum protein intake necessary for the performance. A similar protein intake was observed among Polish advanced climbers by Sas-Nowosielski et al. [ 11], with a mean intake of 1.48 ± 0.34 g/kg/BW and a higher intake in males (1.59 ± 0.40 g/kg/BW) than females (1.34 ± 0.17). These values are consistent with the previous studies of Monedero et al. [ 13] of 1.5 ± 0.4 g/kg/BW and Gibson-Smith et al. [ 12] (1.6 ± 0.5 g/kg/d). Higher intakes were also reported in our previous study in advanced males (1.74 ± 0.6 g/kg/BW), which were lower in the advanced females (1.3 ± 0.42). Furthermore, the protein percentage in the diet ranged from $13\%$ of EI [16] to 17.0 ± $4.2\%$ [14]. Michael et al. [ 28] demonstrated that $77\%$ of adolescent climbers met or exceeded dietary protein needs, whereas Simič et al. [ 7] showed that $44\%$ of participants were above the target for protein, with a mean intake of 1.3 ± 0.4 g/kg/BW [7] Although a higher dietary protein intake may be considered necessary during periods of strength training, no studies have evaluated the protein needs of sports climbers [3]. The recommended CHO intake ranges from 3 to 7 g/kg/BW [3]. In this study, the median CHO intake was the highest in the intermediate group (4.58 g/kg/BW). Similar median values were reported in advanced and elite groups (4.16; 4.22). A lowering trend occurred in the male groups, with the mean CHO intake decreasing with each climbing level. As such, the elite male group had the lowest CHO intake (3.08 g/kg/BW). Monedero et al. [ 13] (3.6 ± 1.0 g/kg/BW) and Gibson-Smith et al. [ 12] (3.7 ± 0.9 g/kg/BW) found similar values, and a higher CHO intake was reported in our previous study [15] in the male (4.7 ± 1.4 g/kg/BW) and female (4.6 ±1.3 g/kg/BW) advanced groups. Sas-Nowosielski et al. [ 11] reported a mean CHO intake of approximately 4 g/kg/BW and indicated that female participants were more likely to cut out CHOs. In Kemmler et al. [ 16], $54\%$ of the EI was covered by CHOs. Meanwhile, Michael et al. [ 3] showed that the majority of adolescent climbers ($86\%$) failed to meet the CHO dietary target. In Simič et al. [ 7], a mean daily CHO intake of 4.3 ± 1.3 g/kg/BW placed $75\%$ of participants below the target value [7]. Some works suggest that a low CHO intake may be connected to the rising popularity of ketogenic diets due to the suggestion that they benefit the reduction in body mass [13]. In addition, a short-duration low-CHO diet, followed by a high-CHO diet, is reported to be a successful nutritional tool for increasing endurance in elite climbers [37]. However, a permanently low CHO intake may be a limiting factor in intensive training [4]. There is no specific recommendation for fat, although it is suggested that the intake should follow the recommendation of the general population and not exceed $35\%$ of the total calorie intake [3]. The fat intake in the diets of participants of this study shifted between median intakes of 1.10 and 1.26 g/kg/BW. The lowest median intake was found in the advanced male group (1.10 g/kg/BW) and the highest intake was found in the elite male group (1.26/kg/BW). Other studies involving sport climbers at different ability levels reported a fat intake of 1.5 ± 0.5 g/kg/BW [13], 1.4 ± 0.4 g/kg/BW [12], and 1.22 ± 0.36 g/kg/BW [11], and $29\%$ of EI [16]. Michael et al. [ 28] found that $73\%$ of adolescent climbers failed to meet the fat requirements. Furthermore, the mean daily fat intake was 31.8 ± $4.7\%$ of EI, but $37\%$ of participants did not meet the target of $30\%$ EI [7]. In the current study, the elite female group consumed adequate amounts of most micronutrients, except vitamin D and iodine. Meanwhile, the elite male group took in inadequate vitamin D, iodine, and potassium. It is difficult to meet the recommended 15 mcg of vitamin D intake from the diet, and vitamin D supplementation is recommended for the Polish population during all seasons because of insufficient sun exposure [38]. However, no more than $33\%$ of climbers declared using vitamin D supplements [20]. Nonetheless, the available data show that, although the correction of vitamin D deficiency is necessary, additional supplementation may not have added health benefits [39]. The use of salt was not noted in the food diaries. In Poland, regulation dictates that salt producers add iodine [40], meaning that the use of salt in everyday food preparation can significantly increase its intake. Low potassium intake in the male group could be due to low fruit and vegetable intake, as these are its’ greatest sources [24]. Zapf et al. [ 18] also reported a low vegetable intake in sports climbers. Monedero et al. [ 13] demonstrated inadequate calcium, magnesium, and vitamin D intakes in female climbers. Female subjects in the current study had lower than recommended intakes of protein and iron. In comparison to Monedero et al. [ 13], the males and females in this study consumed more calcium, magnesium, iron, and vitamin C, and similar amounts of vitamin D and B12. The elite female climbers had the highest iron, vitamin C, potassium, magnesium, and zinc intakes. However, these findings contrast with previous work showing deficiencies in the diets of elite female sports climbers [12]. Compared to the study by Kemmler et al. [ 16], male climbers presented lower intakes of calcium, magnesium, and potassium, similar intakes of iron, vitamins B, and E, and higher intakes of vitamins A and D. In this study, among the male groups, the elite climbers presented the lowest intakes of micronutrients compared to the other climbing levels. However, their diets still fulfilled most of the EAR recommendations. The observation that the elite female climbers had the highest intakes of most micronutrients, despite the lowest EA value, suggests that this group puts great focus on the quality of their diet. In the elite male group, taking into account their low EI, the micronutrient intake was satisfactory. However, as a higher training volume would increase their need for many vitamins and minerals, more emphasis should be placed on choosing foods of sufficient quality. ## 5. Study Limitations Despite the best efforts to ensure that the examination was of the best quality, there were some limitations that should be mentioned. Exercise energy expenditure was set based on the declared hours of training using the MET logs for the activity type. The exact time during which training was performed should be measured using more specialized tools, allowing the collection of more reliable data in terms of time and the specifics of the activity, which differ in energy requirements, such as warm ups, different training types, and stretching. However, some of the previous studies mentioned in the discussion also used METs, based upon which our results are comparable. Another limitation which should be considered is that our study involved climbers of different climbing types, namely bouldering and rope climbing, with each of these types having slightly different needs in terms of the energy substrates used for the training efforts, which influences the general macronutrient intake of participants representing each climbing type. However, Olympic-style climbing includes both of these types, and many climbers do not focus on one type of climbing, which is why such a specification in the study group significantly limits the amount of participants. ## 6. Conclusions The significant differences in EAs between the male groups, with both the male and female elite groups presenting the lowest values, suggest that the athletes of this discipline are prone to energy undernutrition regardless of gender. However, the differences between the dietary intake of micronutrients between various levels of advancement were more visible in the female groups, with the elite group having the highest intake of most micronutrients. In the male groups, the consumption of micronutrients was low among all ability levels, but it still reached the recommended values. In the elite group, a high-quality diet was observed despite very low energy availability. Despite the appropriate supply of micronutrients, a long-term insufficient intake of energy can lead not only to a limitation in training opportunities but also to hormonal disorders in both women and men. The low supply of carbohydrates and high supply of fat and protein in the group of elite men may suggest the use of fashionable low-carbohydrate diets, the effectiveness of which has not been confirmed in the case of athletes. It is necessary to educate sports climbers about the importance of proper nutrition and the consequences of insufficient EI. ## References 1. Winiarski J., Biela M., Chrostek R., Popowicz E., Ostromęcka U., Pawlas K.. Wpływ stosowania magnezji białej na stan zdrowia użytkowników zamkniętych obiektów wspinaczkowych. Med. Srod. (2018.0) 21 50-54. DOI: 10.19243/2018107 2. IOC Approves Five New Sports for Olympic Games Tokyo 2020. 3. Michael M.K., Witard O.C., Joubert L.. Physiological demands and nutritional considerations for Olympic-style competitiverock climbing. Cogent Med. (2019.0) 6 1667199. DOI: 10.1080/2331205X.2019.1667199 4. Smith E.J., Storey R., Ranchordas M.K.. Nutritional Considerations for Bouldering. IJSNEM (2017.0) 27 314-324. DOI: 10.1123/ijsnem.2017-0043 5. Ginszt M., Goniewicz M., Ginszt A.. Analiza przyczyn i następstw urazów u dzieci i młodzieży uprawiających wspinaczkę sportową. Hyg. Public Health (2012.0) 47 23-27 6. Stanković D., Ignjatović M., Puletić M., Raković A., Mladenović Ćirić I.. The nutrition of rock climbers. Proceedings of the XVI International Scientific Conference: FIS Communications 2013 in Physical Education, Sport and Recreation and I International Scientific Conference 7. Simič V., Jevšnik Š., Mohorko N.. Low energy availability and carbohydrate intake in competitive adolescent sport climbers. Kinesiolog (2022.0) 54 268-277. DOI: 10.26582/k.54.2.8 8. Strand M.. Attitudes towards disordered eating in the rock climbing community: A digital ethnography. J. Eat. Disord. (2022.0) 10 96. DOI: 10.1186/s40337-022-00619-5 9. Mashuri H., Mappaompo M.A., Purwanto D.. Analysis of energy requirements and nutritional needs of rock climbing athletes. J. Sport Area (2022.0) 7 437-445. DOI: 10.25299/sportarea.2022.vol7(3).10886 10. Saul D., Steinmetz G., Lehmann W., Schilling A.F.. Determinants for success in climbing: A systematic review. J. Exerc. Sci. Fit. (2019.0) 17 91-100. DOI: 10.1016/j.jesf.2019.04.002 11. Sas-Nowosielski K., Wycislik J.. Energy and macronutrient intake of advanced polish sport. JPES (2019.0) 19 829-832. DOI: 10.7752/jpes.2019.s3119 12. Gibson-Smith E., Storey R., Ranchordas M.. Dietary intake, body composition and iron status in experienced and elite climbers. Front Nutr. (2020.0) 7 122. DOI: 10.3389/fnut.2020.00122 13. Monedero J., Duff Ch Brendan E.. Dietary Intakes and the Risk of Low Energy Availability in Male and Female Advanced and Elite Rock Climbers. J. Strength Cond. Res. (2022.0) 37 e8-e15. DOI: 10.1519/JSC.0000000000004317 14. Úbeda N., Lorenzo-Carvacho C., García-González Á.. Energy and nutritional inadequacies in a group of rec-reational adult Spanish climbers. Arch. Med. Deporte (2021.0) 38 237-244. DOI: 10.18176/archmeddeporte.00047 15. Przeliorz-Pyszczek A., Gołąbek K., Regulska-Ilow B.. Evaluation of the relationship of the climbing level of sport climbers with selected anthropometric indicators and diet composition. CEJSSM (2019.0) 28 15-26. DOI: 10.18276/cej.2019.4-02 16. Kemmler W., Roloff I., Baumann H., Schöffl V., Weineck J., Kalender W., Engelke K.. Effect of exercise, body composition, and nutritional intake on bone parameters in male elite rock climbers. Int. J. Sports Med. (2006.0) 27 653-659. DOI: 10.1055/s-2005-872828 17. Przeliorz A., Regulska-Ilow B.. Relationship between the dietary intake of sport climbers according to climbing grading scales and the dietary supply of antioxidants. JPES (2022.0) 22 795-802. DOI: 10.7752/jpes.2022.03101 18. Zapf J., Fichtl B., Wielgoss S., Schmidt W.. Macronutrient intake and eating habits in elite rock climbers. MSSE (2001.0) 33 72. DOI: 10.1097/00005768-200105001-00407 19. Peoples G.E., Scott Parker R.A., Craddock J.. Rock climbers’ self-reported dietary practices and supplement use in the context of supporting climbing performance. JSES (2021.0) 12 130-138. DOI: 10.36905/jses.2021.02.06 20. Chmielewska A., Regulska-Ilow B.. Evaluation of Supplement Use in Sport Climbers at Different Climbing Levels. Nutrients (2022.0) 15. DOI: 10.3390/nu15010100 21. Sas-Nowosielski K., Wyciślik J., Kaczka P.. Beta-alanine supplementation and sport climbing performance. Int. J. Environ. Res. Public Health (2021.0) 18. DOI: 10.3390/ijerph18105370 22. Sas-Nowosielski K., Kaczka P.. A pilot study to examine the effects of beta-alanine and sodium citrate (CarnoRushTM, Olimp®) supplementation on climbing-specific performance. Baltic J. Health Phys. Act. (2022.0) 14 4. DOI: 10.29359/BJHPA.14.1.04 23. Fitatu Application 24. Draper N., Giles D., Schöffl V., Fuss F.K., Watts P.B., Wolf P., Baláš J., Romero V.E., Gonzales G.B., Fryer S.. Comparative grading scales, statistical analyses, climber descriptors and ability grouping: International Rock Climbing Research Association position statement. Sports Technol. (2015.0) 8 88-94. DOI: 10.1080/19346182.2015.1107081 25. Ainsworth B.E., Haskell W.L., Herrmann S.D., Meckes N., Bassett D.R., Tudor-Locke C., Greer J.L., Vezina J., Whitt-Glover M.C., Leon A.S.. Compendium of Physical Activities: A second update of codes and MET values. MSSE (2011.0) 43 1575-1581. DOI: 10.1249/MSS.0b013e31821ece12 26. Loucks A.B., Kiens B., Wright H.H.. Energy availability in athletes. J. Sports Sci. (2011.0) 29 7-15. DOI: 10.1080/02640414.2011.588958 27. Kunachowicz K., Nadolna I., Przygoda B., Iwanow K.. Tabele Składu i Wartości Odżywczej Żywności (2017.0) 28. Jarosz M., Rychlik E., Stoś K., Charzewska J.. Normy Żywienia dla Populacji Polski i ich Zastosowanie (2020.0) 29. Michael M.K., Joubert L., Witard O.C.. Assessment of Dietary Intake and Eating Attitudes in Rec-reational and Competitive Adolescent Rock Climbers: A Pilot Study. Front. Nutr. (2019.0) 6 64. DOI: 10.3389/fnut.2019.00064 30. Joubert L.M., Gonzalez G.B., Larson A.J.. Prevalence of Disordered Eating Among International Sport Lead Rock Climbers. Front. Sports Act. Living (2020.0) 2 86. DOI: 10.3389/fspor.2020.00086 31. Modaberi S., van Andel S., Saemi E., Joubert L.M., Taheri M.. Differences between boulderers and top rope climbers in the rela-tionship between anxiety and disordered eating. Sport Sci. Health 2022. DOI: 10.1007/s11332-022-00960-x 32. Sundgot-Borgen J., Torstveit M.K.. Prevalence of Eating Disorders in Elite Athletes Is Higher Than in the General Population. CJSM (2004.0) 14 25-32. DOI: 10.1097/00042752-200401000-00005 33. Sundgot-Borgen J., Garthe I.. Elite athletes in aesthetic and Olympic weight-class sports and the challenge of body weight and body compositions. J. Sports Sci. (2011.0) 29 101-114. DOI: 10.1080/02640414.2011.565783 34. Wardle J., Haase A.M., Steptoe A.. Body image and weight control in young adults: International comparisons in university students from 22 countries. IJO (2006.0) 30 644-651. DOI: 10.1038/sj.ijo.0803050 35. Goltz F.R., Stenzel L.M., Schneider C.D.. Disordered eating behaviors and body image in male athletes. Rev. Bras. Psiquiatr. (2013.0) 35 237-242. DOI: 10.1590/1516-4446-2012-0840 36. Mountjoy M., Sundgot-Borgen J.K., Burke L.M., Ackerman K.E., Blauwet C., Constantini N., Leburn C., Lundy B., Melin A.K., Meyer N.L.. IOC consensus statement on relative energy deficiency in sport (RED-S): 2018 update. Br. J. Sports Med. (2018.0) 52 687-697. DOI: 10.1136/bjsports-2018-099193 37. Galysheva S.M., Legchinova S.N.. Low carb-high carb diet plan to increase special endurance of elite climbers. Theory Pract. Phys. Cult. (2022.0) 1 41-43 38. Rusińska A., Płudowski P., Walczak M., Borszewska-Kornacka M.K., Bossowski A., Chlebna-Sokół D., Czech-Kowalska J., Dobrzańska A., Franek E., Helwich E.. Vitamin D Supplementation Guidelines for General Population and Groups at Risk of Vitamin D Deficiency in Poland—Recommendations of the Polish Society of Pediatric Endocrinology and Diabetes and the Expert Panel With Participation of National Specialist Consultants and Representatives of Scientific Societies—2018 Update. Front. Endocrinol. (2018.0) 9 246. DOI: 10.3389/fendo.2018.00246 39. Bouillon R., Manousaki D., Rosen C., Trajanoska K., Rivadeneira F., Richards J.B.. The health effects of vitamin D supplementation: Evidence from human studies. Nat. Rev. Endocrinol. (2022.0) 18 96-110. DOI: 10.1038/s41574-021-00593-z 40. Lewinski A., Zygmunt A.. Lodine prophylaxis in Poland-new, old challanges. AAEM (2014.0) 21 1-4. PMID: 24720252
--- title: 'Overweight in Older Adults: A Follow-Up of Fifteen Years of the SABE Survey' authors: - Tânia Aparecida de Araujo - Isabela Martins Oliveira - Tarsila Guimarães Vieira da Silva - Vanderlei Carneiro da Silva - Yeda Aparecida de Oliveira Duarte journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049442 doi: 10.3390/ijerph20065098 license: CC BY 4.0 --- # Overweight in Older Adults: A Follow-Up of Fifteen Years of the SABE Survey ## Abstract Despite extensive research on overweight and obesity, there are few studies that present longitudinal statistical analyses among non-institutionalized older adults, particularly in low- and middle-income countries. This study aimed to assess the prevalence and factors associated with excess weight in older adults from the same cohort over a period of fifteen years. A total of 264 subjects aged (≥60 years) from the SABE survey (Health, Wellbeing and Aging) in the years 2000, 2006, 2010, and 2015 in the city of São Paulo, Brazil, were evaluated. Overweight was assessed by a BMI of ≥28 kg/m2. Multinomial logistic regression models adjusted for sociodemographic and health data were used to assess factors associated with excess weight. After normal weight, overweight was the most prevalent nutritional status in all evaluated periods: $34.02\%$ in 2000 ($95\%$CI: 28.29–40.26); $34.86\%$ in 2006 ($95\%$CI: 28.77–$41.49\%$); $41.38\%$ in 2010 ($95\%$CI: 35.25–47.79); $33.75\%$ in 2015 ($95\%$CI: 28.02–40.01). Being male was negatively associated with being overweight in all years (OR: 0.34 in 2000; OR: 0.36 in 2006; OR: 0.27 in 2010; and OR: 0.43 in 2015). A greater number of chronic diseases and worse functionality were the main factors associated with overweight, regardless of gender, age, marital status, education, physical activity, and alcohol or tobacco consumption. Older adults with overweight and obesity, a greater number of chronic diseases, and difficulties in carrying out daily tasks required a greater commitment to healthcare. Health services must be prepared to accommodate this rapidly growing population in low- and middle-income countries. ## 1. Introduction Aging populations are poised to become the next global public health challenge [1]. It is estimated that by 2050, one in five people worldwide will be 60 years old or older, totaling 2 billion people. [ 2]. In Brazil, older adults represent the fastest growing age group [3]. It is uncertain, however, whether increased life expectancy is also associated with a better health status [4]. It is known that in addition to socioeconomic changes, aging is accompanied by progressive physiological changes that can impact functionality, autonomy, and nutritional status [5]. Weight loss can be caused by various factors such as changes in smell and taste [6], as well as reductions in lean mass, bone tissue, and muscle strength [7]. Additionally, changes in body composition are related to increased body weight, such as re-distribution of and increase in fat mass (especially in the abdominal region) and reductions in lean mass [8], which may lead to decreased energy expenditure rates. There is still much debate about the optimal weight in old age. The modified BMI classifications [9] suggest that a higher body weight is expected among older adults. Although cross-sectional studies have shown an increase in body weight in this population [10,11], the most common recommendation among specialists is the prevention of low weight [9]. Thus, while the world is facing alarming rates of overweight and obesity (especially among young people and adults), overweight does not seem to be a major concern among older adults. Longitudinal studies conducted in developed countries [12,13] have shown an increase in overweight, obesity, and abdominal obesity even among older adults. However, in developing countries, there is a scarcity of results from longitudinal studies with this population. For example, in an extensive systematic review of longitudinal studies conducted from 2012 to 2020 that identified the association between physical activity and the onset of obesity, only one study from a developing country, Mexico, was identified among the fourteen countries evaluated [14]. Kawai et al. [ 15] draw attention to the immense difficulty of conducting longitudinal studies, especially among older adults. According to the authors, in addition to the higher mortality in this population, physical limitations, low education, and economic situation, for example, make it difficult for older adults to participate in these studies. In developing countries, where investment in research is very limited, the difficulties in carrying out these studies can be even greater. As developing countries, including those in Latin America, experience rapid population aging, they face significant challenges in providing adequate healthcare for their growing older adult populations. Conducting longitudinal studies in these locations could help to shed light on the unique aging patterns in these populations, which could in turn provide valuable insights for improving the health and well-being of older adults in these regions. The objective of this study is to evaluate the prevalence and factors associated with excess weight in older adults, followed in a fifteen-year follow-up. ## 2.1. Sample and Study Population The SABE study—Health, Well-Being and Aging—began in 2000 as a population-based survey to assess the living conditions and health of older adults residing in seven urban centers in Latin America and the Caribbean. In Brazil, the center chosen was the city of São Paulo, which, under the coordination of members of the Epidemiology department of the School of Public Health at the University of São Paulo, became a multi-cohort longitudinal study. Thus, the older adults assessed in 2000 were revisited in 2006, 2010, and 2015, comprising cohort A. This study used panel data from the SABE study spanning a 15-year period (2000 to 2015). The sample for the SABE study was selected using a two-stage cluster sampling method, with proportional allocation based on size, at the census sector and household levels. In the first wave of data collection (cohort A—2000), due to the low population density, the sample sizes for the 75 years and older age group were increased. Additionally, the male samples were adjusted to match the female samples to account for the higher mortality rate in the male population. For the expanded groups, the sample composition was determined freely [16]. For this study, data were used only from individuals from cohort A, which started in 2000 and of which participants remained in the study until the last wave of collections in 2015 ($$n = 356$$). Those who did not have complete anthropometric data were excluded ($$n = 92$$), making a final sample of $$n = 264$.$ Trained interviewers collected data in household interviews, using a structured questionnaire that addressed socioeconomic variables, general health status, living conditions, and anthropometric measurements. In addition, to control the quality of the interviews and reduce information bias, the questionnaires returned by the interviewers went through a critical analysis of completion and initial consistency. Details about the methodology are described in another publication [17]. ## 2.2.1. Dependent Variable The dependent variable was overweight (≥28 kg/m2) assessed by BMI (kg/m2). To calculate BMI, body mass (in kilograms) was divided by the square of height (in meters). Body mass was measured using a calibrated portable scale (SECA® brand) with a capacity of 150 kg and sensitivity of ½ kg, and height was measured using a stadiometer fixed to a 2 m wall with a sensitivity of 1 mm. All anthropometric variables were measured in triplicate and the mean value was used for the analyses, according to the Frisancho [18] standardization. For the classification of BMI, we used the cutoff points adopted by the Pan American Health Organization (PAHO) for the SABE [19] study: normal weight (>23 and <28 kg/m2), underweight (≤23 kg/m2), and overweight (≥28 kg/m2). ## 2.2.2. Independent Variables Sociodemographic characteristics included gender, age (both continuous and categorized as 60–70 years, 71–80 years, and >80 years), years of education (categorized as illiterate, 1–6 years, and ≥7 years), and marital status (categorized as married, widowed, and single/divorced). Clinical characteristics: The respondents were asked whether a doctor or nurse had ever informed them of the diagnosis of any of the following diseases: hypertension, diabetes, cardiovascular diseases, and arthropathies. Number of chronic diseases was categorized as 0 or 1 and ≥2 diseases, and self-assessment of health was categorized as very good/good and regular/bad/very bad. Functional characteristics: The Katz scale [20] was used to assess the difficulty in performing ≥1 basic activity of daily living (ADL) and ≥1 instrumental activity of daily living (IADL). The ADLs considered were getting dressed alone, crossing a room, eating, bathing, using the bathroom, and getting up from a bed or chair. The IADLs considered were preparing meals, managing one’s own money, using the telephone, using transportation, taking medications, shopping, and performing light and heavy household chores. Behavioral characteristics: We assessed smoking status (never smoked, ever smoked and no longer smokes, and currently smokes); alcohol consumption (we asked how many days per week the older adults had consumed some alcoholic beverage and categorized as low consumption (does not consume or <1×/week), moderate (1 to 3×/week), and high (4×/week to every day)) [21]; and for physical activity, we asked whether the respondent performed physical activity and how many days per week (categorized as yes (≥3×/week) and no (<3×/week)). Anthropometric characteristics included waist circumference (WC) and calf circumference (CC). WC was measured using an inelastic tape measure at the midpoint between the last rib and the iliac crest, with the abdomen relaxed at the end of expiration and the area measured free of clothing. If it was not possible to measure the midpoint, the "natural waist" measurement was used, also with a relaxed abdomen. CC was measured using an inelastic tape measure at the midpoint of the dominant leg (as indicated by the older adults) with the older adult in a sitting position, forming a 90º angle with the knee, following the standardization of Lohman et al. [ 22]. ## 2.3. Statistical Analysis Considering the complex research design, survey weights were used to estimate the prevalence of overweight and other nutritional statuses (underweight and normal weight). To ensure the reliability of the evaluated data, and considering the significant losses that occurred during the follow-up, we compared the “lost” subjects to the interviewees in terms of age, BMI, WC, and CC. As all of these variables were continuous, we checked their normality using the Shapiro–Wilk test and histograms. For those variables that met the normality criteria, we used Student’s t-test to compare means. For continuous variables without a normal distribution, we used the Wilcoxon test. Categorical variables were analyzed using weighted proportions. Adjusted cross-sectional multinomial logistic regression models were used to estimate the odds ratios (ORs) and their $95\%$ confidence intervals (CI) for the outcome of overweight with demographic factors (sex) and clinical factors, while controlling for age, education, marital status, alcohol consumption, smoking, and physical activity. Differences between β values were estimated using the Wald test, comparing overweight individuals with normal weight individuals as the reference category. Tests with a p-value < 0.05 were considered statistically significant. All analyses were performed using STATA 14.2. The study was approved by the Ethics Committee of the Faculty of Public Health, University of São Paulo, under control numbers 315 [2000], 83 [2006], 2044 [2010], and 3,600,782 [2015]. Participants were asked to read and sign a consent letter before the assessments and interviews began. ## 3. Results The final sample of this study consisted of 264 participants who were assessed in 2000, 2006, 2010, and 2015. During the evaluation period, $65.3\%$ of these losses occurred due to deaths and $34.7\%$ due to other reasons such as institutionalization, refusals, change of address, and being unable to locate the participants. The flow chart of losses is presented in Figure 1. The data for individuals lost due to death or other reasons are presented in Table 1. In comparison to the evaluated older adults, those who died during the study follow-up had lower BMI and CC values. However, participants who were lost due to other reasons did not show differences in terms of age or anthropometric values from those who were assessed. Descriptive data from the baseline (year 2000) are presented according to nutritional status in Table 2. Since this is a sample of individuals who were followed for fifteen years from 2000, the majority of them were under 70 years of age at baseline. In the same year, however, almost half of the sample had already been diagnosed with hypertension. Among hypertensive individuals, overweight was more prevalent among women. Older adults classified as having a normal weight reported engaging in more physical activity (more than 3 times a week). The weighted prevalence of underweight, normal weight, and overweight were compared in 2000, 2006, 2010, and 2015. Except for the fourth wave in 2015, the prevalence of overweight followed an increasing trend in all evaluated waves, reaching its lowest percentage difference between normal weight and overweight in 2010. See Figure 2. The multinomial logistic regression model yielded raw and adjusted values of factors associated with excess weight in each wave of the study. The results showed that, regardless of age, education, marital status, alcohol consumption, smoking, and physical activity, men had a lower risk of being overweight compared to women (OR: 0.34–2000; OR: 0.36–2006; OR: 0.27–2010; and OR: 0.43–2015). On the other hand, for the entire sample, factors that increased the risk of being overweight were worse functionality (greater difficulty in performing ADLs and IADLs), diagnosis of hypertension, arthropathies, or having two or more non-communicable chronic diseases together. Table 3 shows the details. ## 4. Discussion Overweight and obesity are known to predispose individuals to disability and reduced physical functioning [8]. This study, which monitored the same population of older adults over a period of fifteen years, found an increase in the prevalence of overweight and obesity in the first three follow-up waves. Only in the last wave, with an average age of 77 among those evaluated, did the prevalence of overweight decrease. Previous longitudinal studies indicate an increase in BMI up to the age of 65 [23], stabilizing and starting to decrease after the age of 80 [24]. In a recent 15-year follow-up study, the authors reported that body weight stabilized up to the age of 65 and started to decrease after the age of 66 [25]. In another study carried out with the first three waves of the SABE study, which had a 10-year follow-up, a decrease in body weight was observed after the age of 70 [26]. The global prevalence of overweight and obesity appears to be higher among women than men [23], especially after the age of 50 [27]. The results found in this study confirm these findings, as men were less likely to be overweight than women in all evaluated waves. These results may be due to biological differences, such as the effect of menopause and sex hormones, for example, or to cultural and social factors that result in lower levels of education and a more sedentary lifestyle among women [27]. The increasing trend in the prevalence of overweight and obesity may be associated with economic development, urbanization, and the globalization of food systems [23,27]. These factors stimulate the production and availability of more processed and ultra-processed foods [28], which promote excessive consumption of energy-rich and nutrient-poor foods [27]. These types of foods may be particularly attractive to older adults, who often face challenges in purchasing and preparing food [29]. Furthermore, the modernization of lifestyles results in reduced physical activity, further contributing to the increase in body weight [26]. On the other hand, although children, adolescents, adults, and older adults may share some factors that contribute to overweight, the effects of overweight and obesity in old age are still uncertain [30]. For example, a meta-analysis that assessed mortality risks in older adults (aged 65 years or older) demonstrated that the increased risk of mortality occurred for individuals with a BMI <23 kg/m2 [31], while being overweight was not associated with an increased risk of mortality for older populations [32]. Contradicting these results, another study which included nearly 900,000 adults by the Prospective Studies Collaboration found a $30\%$ increase in mortality risk for every 5-unit increase in BMI above 22.5–25 kg/m2 [33]. Results from this work demonstrated that the mean BMI of those participants who died during the study was also lower than those who were evaluated. According to Cetin [30], it is possible that individuals vulnerable to the negative impacts of obesity have a lower life expectancy, as weight gain is not necessarily associated with increased survival. In contrast, those who survive into old age may have greater resistance to the harmful effects of obesity. Even though the risks of overweight and obesity among older adults are still divergent in the literature, our data revealed that those with overweight and obesity had a greater number of chronic non-communicable diseases and greater difficulty in basic and instrumental activities of daily living. This fact, per se, would indicate a worse health status. Overweight and obesity are among the main preventable risk factors for diseases such as type 2 diabetes mellitus, fatty liver disease, hypertension, myocardial infarction, stroke, dementia, osteoarthritis, obstructive sleep apnea, and various types of cancer [33]. These conditions also predispose individuals to greater disability and decreased physical functioning. Data from the Health, Aging and Body Composition study, for example, revealed that adults and older adults classified as overweight or obese had a hazard ratio of 2.38 for incident disability over a 7-year follow-up period [34]. According to Silva [35], excess weight is one of the factors associated with declines in strength, mobility, and flexibility in older adults, thereby altering their ability to perform daily activities [36]. In 2000, being overweight was also associated with a worse self-assessment of health. Although this association is not well explored in the literature, Borim [37] reported that the prevalence of good/excellent self-assessment of health was significantly lower in older individuals with BMI ≥ 30 kg/m2 in the city of Campinas, Brazil. A greater number of associated diseases and disabilities may explain this association. Batsis [8] states that although there are challenges in the diagnostic accuracy of obesity, regardless of the body composition or anthropometric measurement used, in older adults living in the community, obesity is associated with a worse prognosis for physical function. According to data from the Cardiovascular Health Study 1989–2015, an otherwise healthy lifestyle, including physical activity, diet, and weight control, can compress the number of years of disability. The strengths of this study include long-term monitoring of the community’s older adults for a period of 15 years. The complexity of the sampling process and careful execution of the research adds to its important internal validity. Weight and height values were directly measured rather than self-reported, increasing confidence in the findings. Additionally, using specific BMI cut-off points helped avoid overestimation of the prevalence of overweight and obesity. As some authors have noted, the WHO healthy weight range for adults may not be appropriate for older adults [31]. This manuscript also has some important limitations that are worth mentioning. Firstly, the data were analyzed cross-sectionally, which makes it difficult to establish causal relationships between the variables. However, the consideration of sample weight and adjusted regression models supports the reliability of the demonstrated results. Secondly, changes in weight and nutritional status may have occurred outside the study collection periods. Despite rigorous quality control, the diagnostic data for chronic diseases were obtained through self-reported accounts from the participants. Lastly, in the final wave of the study in 2015, a significant portion of those assessed ($24.2\%$) were excluded from the anthropometric measurement as they were unable to stand on their feet due to physical limitations. Additionally, nearly half of the population ($48.2\%$) was over 80 years old. ## 5. Conclusions “While the clinical approach to overweight and obesity in older adults is still controversial, prevention remains a safe measure. Knowledge of the evolution of nutritional status in older adults, allowed by longitudinal follow-ups like this study, can be useful for decision-makers. Even at older ages, this study demonstrated a high prevalence of overweight, which is associated with worse health indicators such as the presence of chronic diseases and lower functionality. Including older adults in programs to prevent and treat overweight is a necessary measure that public health managers should consider. Future research may indicate the effectiveness of measures such as diet, physical activity, and the promotion of healthy environments in reducing the prevalence of overweight and obesity in old age. ## References 1. Suzman R., Beard J.R., Boerma T., Chatterji S.. Health in an ageing world—What do we know?. Lancet (2015.0) 385 484-486. DOI: 10.1016/S0140-6736(14)61597-X 2. 2. WHO Global Strategy and Action Plan on Ageing and HealthWorld Health OrganizationGeneva, Switzerland2017. Global Strategy and Action Plan on Ageing and Health (2017.0) 3. Camarano A.A.. Novo Regime Demográfico: Uma Nova Relação Entre População e Desenvolvimento? (2014.0) 4. Chang A.Y., Skirbekk V.F., Tyrovolas S., Kassebaum N.J., Dieleman J.L.. Measuring population ageing: An analysis of the Global Burden of Disease Study 2017. Lancet Public Health (2019.0) 4 e159-e167. DOI: 10.1016/S2468-2667(19)30019-2 5. JafariNasabian P., Inglis J.E., Reilly W., Kelly O.J., Ilich J.Z.. Aging human body: Changes in bone, muscle and body fat with consequent changes in nutrient intake. J. Endocrinol. (2017.0) 234 R37-R51. DOI: 10.1530/JOE-16-0603 6. Van der Zanden L.D., Van Kleef E., Wijk R.A., Van Trijp H.C.. Understanding heterogeneity among elderly consumers: An evaluation of segmentation approaches in the functional food market. Nutr. Res. Rev. (2014.0) 27 159-171. DOI: 10.1017/S0954422414000092 7. Ilich J.Z., Kelly O.J., Inglis J.E., Panton L.B., Duque G., Ormsbee M.J.. Interrelationship among muscle, fat, and bone: Connecting the dots on cellular, hormonal, and whole body levels. Ageing Res. Rev. (2014.0) 15 51-60. DOI: 10.1016/j.arr.2014.02.007 8. Batsis J.A., Zagaria A.B.. Addressing Obesity in Aging Patients. Med. Clin. N. Am. (2018.0) 102 65-85. DOI: 10.1016/j.mcna.2017.08.007 9. Barkoukis H.. Nutrition Recommendations in Elderly and Aging. Med. Clin. N. Am. (2016.0) 100 1237-1250. DOI: 10.1016/j.mcna.2016.06.006 10. Andrade F.B., Caldas Junior A.F., Kitoko P.M., Batista J.E.M., Andrade T.B.. Prevalence of overweight and obesity in elderly people from Vitória-ES, Brazil. Ciência Saúde Coletiva (2012.0) 17 749-756. DOI: 10.1590/S1413-81232012000300022 11. Silveira E.A., Kac G., Barbosa L.S.. Prevalência e fatores associados à obesidade em idosos residentes em Pelotas, Rio Grande do Sul, Brasil: Classificação da obesidade segundo dois pontos de corte do índice de massa corporal. Cad. Saúde Pública (2009.0) 25 1569-1577. DOI: 10.1590/S0102-311X2009000700015 12. Fernandes S.B.J., Santos G.C., Vilton C.J., Ahmed T., Zunzunegui M.V., Curcio C.L., Gomez F., Oliveira Guerra R.. Abdominal Obesity and Mobility Disability in Older Adults: A 4-Year Follow-Up the International Mobility in Aging Study. J. Nutr. Health Aging (2018.0) 22 1228-1237. DOI: 10.1007/s12603-018-1100-y 13. Barbour K.E., Helmick C.G., Boring M., Qin J., Pan L., Hootman J.M.. Obesity Trends Among US Adults With Doctor-Diagnosed Arthritis 2009–2014. Arthritis Care Res. (2017.0) 69 376-383. DOI: 10.1002/acr.22958 14. Cleven L., Krell-Roesch J., Nigg C.R., Woll A.. The association between physical activity with incident obesity, coronary heart disease, diabetes and hypertension in adults: A systematic review of longitudinal studies published after 2012. BMC Public Health (2020.0) 20. DOI: 10.1186/s12889-020-08715-4 15. Kawai H., Ejiri M., Tsuruta H., Masui Y., Watanabe Y., Hirano H., Fujiwara Y., Ihara K., Tanaka M., Obuchi S.. Factors associated with follow-up difficulty in longitudinal studies involving community-dwelling older adults. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0237166 16. Lartey S.T., Magnussen C.G., Si L., Boateng G.O., Graaff B., Biritwum R.B., Minicuci N., Kowal P., Blizzard L., Palmer A.J.. Rapidly increasing prevalence of overweight and obesity in older Ghanaian adults from 2007–2015: Evidence from WHO-SAGE Waves 1 & 2. PLoS ONE (2019.0) 14. PMID: 31425568 17. Lebrão M.L., Duarte Y.A.O.. O Projeto SABE no Município de São Paulo: Uma Abordagem Inicial. Brasília: Organização Pan-Americana de Saúde. (2003.0) 18. Frisancho A.R.. New standards of weight and body composition by frame size and height for assessment of nutritional status of adults and the elderly. Am. J. Clin. Nutr. (1984.0) 40 808-819. DOI: 10.1093/ajcn/40.4.808 19. Multicéntrica E.. Salud Bienestar y envejecimiento (SABE) en América Latina y el Caribe. Organización Panamericana de la Salud (2001.0) 20. Lino V.T.S., Pereira S.R.M., Camacho L.A.B., Ribeiro Filho S.T., Buksman S.. Adaptação transcultural da Escala de Independência em Atividades da Vida Diária (Escala de Katz). Cad. Saúde Pública. (2008.0) 24 103-112. DOI: 10.1590/S0102-311X2008000100010 21. 21. U.S. Department of Health and Human Services National Institutes of Health What is a Standard Drink? 2018 [National Institute on Alcohol Abuse and Alcoholism]Available online: https://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/what-standard-drink(accessed on 22 April 2021) 22. Lohman T.G., Roche A.F., Martorel L.R.. Anthropometric Standardization Reference Manual (1988.0) 23. Chooi Y.C., Ding C., Magkos F.. The epidemiology of obesity. Metabolism (2019.0) 92 6-10. DOI: 10.1016/j.metabol.2018.09.005 24. Dahl A.K., Reynolds C.A., Fall T., Magnusson P.K.E., Pedersen N.L.. Multifactorial analysis of changes in body mass index across the adult life course: A study with 65 years of follow-up. Int. J. Obes. (2014.0) 38 1133-1141. DOI: 10.1038/ijo.2013.204 25. Guo J., Shang Y., Fratiglioni L., Johnell K., Welmer A.K., Marseglia A., Xu W.. Individual changes in anthropometric measures after age 60 years: A 15-year longitudinal population-based study. Age Ageing (2021.0) 50 1666-1674. DOI: 10.1093/ageing/afab045 26. Araujo T.A., Corona L.P., Andrade F.C.D., Roediger M.A., Duarte Y.A.O.. Factors associated with body mass index changes among older adults: A ten-year follow-up. Cad. De Saúde Pública (2021.0) 37 e00081320. DOI: 10.1590/0102-311x00081320 27. Wong M.C.S., Huang J., Wang J., Chan P.S.F., Lok V., Chen X., Leung C., Wang H.H.X., Lao X.Q., Zheng Z.J.. Global, regional and time-trend prevalence of central obesity: A systematic review and meta-analysis of 13.2 million subjects. Eur. J. Epidemiol. (2020.0) 35 673-683. DOI: 10.1007/s10654-020-00650-3 28. Monteiro C., Cannon G., Levy R., Moubarac J., Louzada M., Rauber F., Khandpur N., Cediel G., Neri D., Martinez-Steele E.. Ultra-processed foods: What they are and how to identify them. Public Health Nutr. (2019.0) 22 936-941. DOI: 10.1017/S1368980018003762 29. Ishikawa M., Yokoyama T., Nakaya T., Fukuda Y., Takemi Y., Kusama K., Yoshiike N., Nozue M., Yoshiba K., Murayama N.. Food accessibility and perceptions of shopping difficulty among elderly people living alone in Japan. J. Nutr. Health Aging (2016.0) 20 904-911. DOI: 10.1007/s12603-015-0694-6 30. Cetin D.C., Nasr G.. Obesity in the elderly: More complicated than you think. Cleve Clin. J. Med. (2014.0) 81 51-61. DOI: 10.3949/ccjm.81a.12165 31. Winter J.E., MacInnis R.J., Wattanapenpaiboon N., Nowson C.A.. BMI and all-cause mortality in older adults: A meta-analysis. Am. J. Clin. Nutr. (2014.0) 99 875-890. DOI: 10.3945/ajcn.113.068122 32. Whitlock G., Lewington S., Sherliker P., Clarke R., Emberson J., Halsey J., Qizilbash N., Collins R., Peto R.. Body-mass index and cause-specific mortality in 900,000 adults: Collaborative analyses of 57 prospective studies. Lancet (2009.0) 373 1083-1096. PMID: 19299006 33. Blüher M.. Obesity: Global epidemiology and pathogenesis. Nat. Rev. Endocrinol. (2019.0) 15 288-298. DOI: 10.1038/s41574-019-0176-8 34. Houston D.K., Ding J., Nicklas B.J., Harris T.B., Lee J.S., Nevitt M.C., Rubin S.M., Tylavsky F.A., Kritchevsky S.B.. Overweight and obesity over the adult life course and incident mobility limitation in older adults: The health, aging and body composition study. Am. J. Epidemiol. (2009.0) 169 927-936. DOI: 10.1093/aje/kwp007 35. Silva N.A., Menezes T.N., Melo R.L.P., Pedraza D.F.. Força de preensão manual e flexibilidade e suas relações com variáveis antropométricas em idosos. Rev. Assoc. Médica Bras. (2013.0) 59 128-135. DOI: 10.1016/j.ramb.2012.10.002 36. Jacob M.E., Yee L.M., Diehr P.H., Arnold A.M., Thielke S.M., Chaves P.H., Del Gobbo L., Hirsch C., Siscovick D., Newman A.B.. Can a Healthy Lifestyle Compress the Disabled Period in Older Adults?. J. Am. Geriatr. Soc. (2016.0) 64 1952-1961. DOI: 10.1111/jgs.14314 37. Borim F.S.A., Barros M.B.d.A., Neri A.L.. Autoavaliação da saúde em idosos: Pesquisa de base populacional no Município de Campinas, São Paulo, Brasil. Cad. Saúde Pública (2012.0) 28 4. DOI: 10.1590/S0102-311X2012000400016
--- title: 'The Mediation Effect of Pain on the Relationship between Kinesiophobia and Lumbar Joint Position Sense in Chronic Low Back Pain Individuals: A Cross-Sectional Study' authors: - Mohammad A. ALMohiza - Ravi Shankar Reddy - Faisal Asiri - Adel Alshahrani - Jaya Shanker Tedla - Snehil Dixit - Kumar Gular - Venkata Nagaraj Kakaraparthi journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049448 doi: 10.3390/ijerph20065193 license: CC BY 4.0 --- # The Mediation Effect of Pain on the Relationship between Kinesiophobia and Lumbar Joint Position Sense in Chronic Low Back Pain Individuals: A Cross-Sectional Study ## Abstract [1] Background: Fear of movement (kinesiophobia) and impaired lumbar joint position sense (LJPS) play a vital role in developing and maintaining non-specific chronic low back pain (CLBP). However, how kinesiophobia impacts LJPS is still being determined. The aims of this study are to [1] assess the correlation between kinesiophobia and LJPS in individuals with chronic low back pain; [2] compare LJPS between individuals with CLBP and those who are asymptomatic; and [3] evaluate if pain can mediate the relationship between kinesiophobia and LJPS in CLBP individuals. [ 2] Methods: Eighty-three individuals (mean age = 48.9 ± 7.5 years) with a diagnosis of CLBP and 95 asymptomatic individuals (mean age = 49.4 ± 7.0 years) were recruited into this cross-sectional study. Fear of movement in CLBP individuals was assessed using the Tampa Scale for Kinesiophobia (TSK). LJPS was determined using the active target repositioning technique using a dual-digital inclinometer. LJPS was evaluated in lumbar flexion, extension, and side-bending left and right directions, and the repositioning accuracy was determined in degrees using a dual digital inclinometer. [ 3] Results: Kinesiophobia showed a significant ($p \leq 0.001$) moderate positive correlation with LJPS (flexion: $r = 0.51$, extension: $r = 0.41$, side-bending left: $r = 0.37$ and side-bending right: $r = 0.34$). LJPS errors were larger in CLBP individuals compared to asymptomatic individuals ($p \leq 0.05$). Mediation analyses showed that pain significantly mediated the relationship between kinesiophobia and LJPS ($p \leq 0.05$) in CLBP individuals. [ 4] Conclusions: Kinesiophobia and LJPS were positively associated. LJPS is impaired in CLBP individuals compared to asymptomatic individuals. Pain may mediate adverse effects on LJPS. These factors must be taken into account when assessing and developing treatment plans for those with CLBP. ## 1. Introduction Low back pain is a highly prevalent and significant musculoskeletal disorder affecting the general population, leading to high healthcare costs [1,2]. Low back pain affects up to $20\%$ of the general population, and one-year prevalence is 36 to $38\%$ [3]. In addition, the duration of symptoms is directly tied to the condition’s prognosis; the longer low back pain persists, the worse the prognosis [4]. Individuals with chronic low back pain (CLBP) experience pain, functional disability, fear, and anxiety, which can significantly impact their quality of life [5,6]. Fear of movement or activity (kinesiophobia) is a psychological factor associated with the severity and persistence of pain, which has garnered much attention in individuals with chronic pain [7,8]. Kinesiophobia is a term that describes excessive fear of performing a bodily movement due to the expectation of injury or re-injury [9,10]. The prevalence of kinesiophobia is over $50\%$ in chronic pain individuals [11]. The precise way of assessing kinesiophobia is to measure with the Tampa Scale for Kinesiophobia (TSK) [12]. Higher TSK scores are associated with increased pain and disability [13,14]. The etiology of CLBP is complex and multifactorial, making it challenging to identify a single specific element that contributes to and maintains chronic pain [15]. Joint position sense is a critical factor that significantly contributes to the stability of the lumbar spine [16]. Decreases in muscle strength, endurance, and impaired lumbar joint position sense (LJPS) are documented in CLBP individuals, and these factors may significantly maintain the symptoms for a longer duration [17,18]. Additionally, catastrophic behavior driven by fear of injury might prolong the duration of acute pain, converting into chronic pain and causing significant disability and reduced quality of life [19,20]. Kinesiophobia may alter somatosensory changes affecting musculotendinous and capsule-ligamentous structures and affecting afferent proprioceptive input to the higher centers and impairing lumbar proprioception [21]. LJPS is estimated based on the ability of an individual to actively reposition the lumbar spine to a target position, and this reposition accuracy is measured in degrees [22,23]. Previous studies have shown a strong relationship between LJPS, kinesiophobia, pain intensity, and functional disability [24,25,26]. However, the evidence regarding how kinesiophobia impacts LJPS in CLBP is limited and has yet to be explored. To better understand and explain these relationships in patients with CLBP, synthesizing the evidence of the connection between kinesiophobia and LJPS would enable better clinical decision making and formulate effective treatment strategies. The bio-psycho-social framework explains that functional impairment is brought on by a confluence of elements, including pain severity and bio-psychological issues [27]. In this clinical arena, the fear-avoidance model (FAM) explains the relationship between pain and disability and their contribution to developing chronic pain via the psychological process [28]. Fear of movement and catastrophizing thoughts further deteriorate the functional progression of the CLBP individual by increased disability and decreased quality of life [29]. Persistent pain often occurs among CLBP individuals [30]. Different authors have shown a significant relationship between pain severity, proprioceptive impairment, frequency of falls, and balance impairments [31]. Increased pain is associated with increased proprioceptive errors and decreased balance and functional mobility [32]. Previous studies have shown that pain is a significant factor that can increase fear of movement and impair motor control [33,34,35,36,37]. However, it is unknown how pain influences kinesiophobia and its relationship with LJPS in CLBP individuals. We employed mediation analysis using multiple regression to understand the relationship [27]. In line with the statement of the problem, the objectives of this study are [1] to evaluate the correlation between kinesiophobia and LJPS in CLBP subjects; [2] to compare LJPS between CLBP individuals and asymptomatic individuals; and [3] to assess if pain mediates the relationship between kinesiophobia and LJPS in CLBP individuals. ## 2.1. Study Design and Settings This cross-sectional observational study was carried out in medical rehabilitation clinics at the Al-Qara campus, King Khalid University, Kingdom of Saudi Arabia, from May 2021 to December 2021. The local university ethics committee approved the study (ECM #2022-6012), and the study followed the Declaration of Helsinki principles. ## 2.2. Subjects Eighty-three patients from the university hospital were referred to the physical therapy clinic after being diagnosed with non-specific CLBP following lumbar spine radiography and a complete screening process by an orthopedic doctor. A physical therapist, who was a blind evaluator and experienced in musculoskeletal assessments, evaluated kinesiophobia, LJPS, and postural control measures. Consecutive presentations of people with a referral for CLBP treatment to the physiotherapy department were screened. Participants who were clinically diagnosed with chronic LBP were recruited and further assessed for eligibility. The subjects were included in this study if they met the following criteria: [1] males and females aged between 20 and 60; [2] suffering from non-specific CLBP for more than 12 weeks; [3] having a pain intensity of 3 or higher on a 0 to 10 visual analog scale; and [4] willing to participate in the study. The subjects were excluded if they were the following: [1] subjects with specific back pain (fracture, osteoporosis or degenerative changes, prolapse intervertebral disc, bone disorders, arthritis, tumor); [2] subjects with neurological involvement (radiculopathy, myelopathy); [3] subjects with previous spinal surgery; [4] subjects with spinal infections; [5] subjects with a severe psychiatric disorder; and [6] subjects unable to understand or follow examiner commands. Asymptomatic subjects were included if they were over 18 years, healthy, and able to follow the commands of the examiner. If they were using pain medication or had a history of back pain during the previous six months, they were excluded. All the recruited individuals provided written consent before the commencement of the study. ## 2.3.1. Pain Intensity The subjects’ current level of CLBP intensity was assessed using a visual analog scale (VAS). The VAS is a 0 to 10 mm line that denotes “0” as no pain and “10” as the worst pain. The VAS is the most frequently used and reliable instrument in assessing the severity of pain in LBP individuals [38]. ## 2.3.2. Kinesiophobia The Tampa Scale of Kinesiophobia (TSK) scale was first developed to assess fear of movement in individuals with CLBP [39]. It is a 17-item self-report questionnaire with a score range between 17 and 68. A score of 17 indicates the absence of kinesiophobia, and 68 represents the highest fear of movement [40]. A cut-off TSK score of ≥37 indicates fear of movement [41]. The TSK has presented a valid and reliable tool to indicate the presence of kinesiophobia in chronic non-specific CLBP individuals [42]. ## 2.3.3. Lumbar Joint Position Sense LJPS was determined using a dual digital inclinometer (Dualer IQ—Midvale, UT, USA). The LJPS test, which assesses the ability to recognize and reproduce lumbar spine positions, frequently necessitates the time-consuming task of analyzing images using particular software and/or high-tech tools, such as an isokinetic dynamometer, inertial sensors, an electro goniometer, or photo analysis (high-resolution camera), to monitor proprioceptive deficits in individuals with CLBP [22,43]. These measurement tools are trustworthy and accurate, but they are not portable, and their installation takes time. In order to rapidly and efficiently obtain proprioceptive errors, some authors have created and tested digital inclinometers. On the other hand, when compared to pricey and sophisticated equipment, the digital inclinometer is an economical, easy-to-use gadget that takes up less space, is operated by a single rater, and takes measurements quickly. A digital inclinometer is a reliable instrument ($r = 0.98$) that allows physical therapists to accurately assess the range of motion of joint position errors [22,44]. LJPS errors were measured in degrees in directions of lumbar flexion, extension, and side-bending left and right as an estimate of LJPS. All the tests were performed in a calm and well-ventilated lab. All the subjects were blindfolded during the LJPS testing procedure. We adopted the target lumbar positioning testing protocol by Reddy et al. [ 22]. The individual repositions his lumbar spine to a target position from the neutral position. To measure LJPS in lumbar flexion and extension, the primary inclinometer was positioned over the lateral aspect of the chest at the T12 level, and the secondary Inclinometer was placed over the hemipelvis at the S1 level (Figure 1). To measure LJPS in side-bending left and right, the primary inclinometer was placed on the upper back (over the T12 spinous process) and the secondary inclinometer over the mid and central aspect of the sacrum [22]. All subjects’ full range of motion (ROM) was measured (flexion, extension, and side-bending left and right), and $50\%$ of the available ROM was selected as their target to be repositioned during the repositioning task [22]. To start the testing, the subjects were asked to stand straight and were asked to determine their self-selected neutral spine position. The examiner gently guided the participants to the target position ($50\%$ of the available ROM), and they were maintained in this position for a period of five seconds and asked to memorize [22]. Following this, the examiner guided the individual back to the starting position. Next, the examiner asked the subjects to move actively and reposition their lumbar spine to the target position; once the participants reached the target position, they intimated the test by saying, “YES”, and the reposition errors were computed as LJPS in degrees [22]. LJPS was evaluated in lumbar flexion, extension, and side-bending left and right bending directions. Each test was repeated three times, and the mean value was used as the reposition accuracy value. ## 2.4. Sample Size Gpower statistical software estimated the sample size using the following formula: one study group vs. population and continuous variables [45]. A study group mean of 4.8, known population means of 3.6 (SD = 2.2), 1-β (statistical power) of 0.80, and an α (significance level) of 0.05 were used. The estimated sample was 80 in each group. ## 2.5. Statistical Analysis The Shapiro–Wilk test was used to assess the normality of the study variables, and the data followed a normal distribution. The relationship between the TSK score (kinesiophobia) and LJPS was estimated using Pearson’s correlation coefficient. ANOVA was used to compare LJPS between the CLBP group and the asymptomatic group. Mediation analysis was computed to assess the impact of pain on the relationship between kinesiophobia and LJPS in CLBP individuals. The mediation analysis included a three-step process (Figure 2). Bivariate regression assessed the total effect between TSK scores (kinesiophobia) and LJPS (step 1). The direct effect between pain and TSK score (pathway A) was assessed using bivariate regression (step 2). Multiple regression was used to assess the direct effect between TSK score and LJPS (Pathway C) and between TSK score and pain (Pathway B) as step 3. SPSS ver. 24 (SPSS Inc., Chicago, IL, USA) was used to conduct statistical analysis and the significance level was determined at $p \leq 0.05.$ ## 3. Results Eighty-three subjects with CLBP (mean age of 48.9 ± 7.5 years) and 105 asymptomatic individuals (mean age of 49.4 ± 7.0 years) participated in this study. The demographic and physical characteristics of the study subjects are displayed in Table 1. The individuals were overweight, as shown by BMI (26.4 ± 3.3 kg/m2). The CLBP individuals had a mean TSK score of 41.2 ± 3.2. LJPS error was larger in the CLBP group compared to the asymptomatic group. The proprioception was impaired significantly in the CLBP group ($p \leq 0.001$) in all the directions tested (Table 1). Kinesiophobia showed significant positive correlations with LJPS, as summarized in Table 2. TSK scores showed moderate correlations with JPS in flexion ($r = 0.51$; $p \leq 0.001$), extension (0.41; $p \leq 0.001$), and side-bending left ($r = 0.37$; $$p \leq 0.001$$), and side-bending right ($r = 0.34$; $$p \leq 0.002$$) directions. The results of the mediation analysis are summarized in Table 3 and Table 4. As shown in Figure 2 in this mediation model, the total effect was the observed effect of kinesiophobia on LJPS (pathway C). Kinesiophobia was significantly associated with LJPS (flexion: $B = 0.22$, p ≤ 0.001, extension: $B = 0.23$, p ≤ 0.001, side-bending left: $B = 0.18$, $$p \leq 0.001$$, side-bending right: $B = 0.16$, $$p \leq 0.002$$). The total effect also decomposed into the direct effect of kinesiophobia on LJPS (pathway C′) and the indirect effects of kinesiophobia on LJPS through pain (mediated: pathway A + B). The indirect effect was statistically significant (Sobel test), in which pain explained the association between kinesiophobia and LJPS ($p \leq 0.05$). ## 4. Discussion The present study investigates the relationship between kinesiophobia and LJPS. Furthermore, it assesses if pain could mediate the relationship between Kinesiophobia and LJPS in CLBP individuals. The individuals in this study demonstrated significant correlations (moderately positive) between kinesiophobia and LJPS. Pain mediated the relationship between kinesiophobia and LJPS in CLBP individuals. In this study, kinesiophobia significantly correlated with LJPS, indicating that fear of movement can be a coping strategy for impaired proprioceptive acuity. This relationship suggests that fear of movement significantly affects the lumbar motor control [46]. Increased fear of movement and catastrophic behavior in CLBP patients can reduce muscle strength, endurance, and functional capacity [47,48]. Previous studies have demonstrated a significant positive correlation between reduced muscle strength, endurance, and functional performance to impaired JPS [48,49,50]. This mechanism can explain the relationship between kinesiophobia and LJPS. Limited studies have evaluated the impact of kinesiophobia on LJPS in CLBP individuals. Similar to our study methods, Kandakurti et al. [ 26] assessed the impact of kinesiophobia on lumbar extensor endurance and position sense in patients with CLBP, and this study showed that individuals with increased TSK scores had decreased lumbar extensor endurance and high LJPS. Our study results are in accordance with the study conducted by Alshahrani et al. [ 33], in which kinesiophobia had a moderate association with knee joint position sense ($r = 0.38$ to 0.5, $p \leq 0.05$) in knee osteoarthritis individuals. Furthermore, the associations were significant in different target joint position sense angles (15°, 30°, and 60° of knee flexion) tested ($p \leq 0.05$). This implies that agonists and antagonists contract rhythmically to contribute to effective motor control and force-generating capabilities. It also implies that kinesiophobia can influence these components and alter the afferent proprioceptive input leading to impaired joint position sense [51]. A study by Pakzad and colleagues [52] discovered that fear of movement changed muscle activation patterns and neuromuscular control during walking. This suggests that kinesiophobia is a factor that can disrupt proprioception because it changes motor control and has an adverse impact on afferent proprioceptive input, impairing JPS. A study by Asiri et al. [ 34] demonstrated that cervical proprioception is impacted by fear of movement and showed a moderate positive association between TSK scores and cervical joint position sense errors in extension and rotation directions (r range between 0.31 and 0.48, $p \leq 0.05$). Furthermore, in the study conducted by Asiri et al., kinesiophobia significantly predicted cervical proprioceptive acuity [34]. The results of these studies confirm that kinesiophobia significantly impairs JPS. Contrary to our study results, the study by Aydo˘gdu et al. [ 53] in subjects with ACL reconstruction did not show an association between fear of movement and knee joint position sense. The explanation for the difference is that the mean TSK score for kinesiophobia in our study is higher (41.2 3.2) than that found by Aydo˘gdu et al. ( 36.54 4.22). LJPS may have been affected by participants’ greater TSK levels and chronic back pain in this study. Fear of movement for an extended period and increased pain intensity can lead to impaired lumbar proprioception [54]. A vicious loop exists of pain, muscle fatigue, atrophy, and impaired proprioception in subjects with CLBP [54,55]. Karayannis et al. [ 56] showed that increased pain intensity and chronicity, negative thoughts, and fear-avoidance behavior could significantly impact position sense and bodily balance [56]. We anticipated that pain might mediate between kinesiophobia and LJPS. With many musculoskeletal disorders, the fear of pain prolongs the acute pain course and helps it progress into a chronic condition [57]. In addition, proprioception may be compromised by decreased flexibility, deconditioning, and loss of muscle tone brought on by increased pain, chronicity, and disuse [58]. Pain has the ability to influence numerous aspects of the nervous system, including the sensitivity of muscle spindles and the way the central nervous system modulates proprioceptive afferent signals [59]. Asiri et al. [ 37] conducted a study investigating the mediation effect of pain on the relationship between kinesiophobia and postural control in fibromyalgia syndrome, and the results demonstrated that pain significantly mediated their relationship to produce altered motor control, hence impacting balancing ability. Similarly in our study, pain may have impacted the relationship between kinesiophobia and LJPS. Furthermore, fear of movement is significantly influenced by the presence of pain and is intensified by the chronicity of pain. Higher TSK scores (which indicate higher levels of kinesiophobia) were seen in individuals with more pain, as measured by their VAS scores. These individuals exhibit impaired motor control affecting LJPS. Although our investigation revealed the mediating influence of pain on kinesiophobia and LJPS, no studies have studied the pathophysiology behind this effect. Additional research is warranted to determine whether pain-relieving interventions can decrease the fear of movement and improve position sense in CLBP individuals. ## 4.1. Practical Clinical Implications This study revealed that patients with CLBP had decreased proprioception compared to asymptomatic individuals, and earlier research has demonstrated that this population falls more frequently [31]. These results corroborate other studies that have suggested individuals with CLBP fell more frequently and have shown that kinesiophobia can make patients with CLBP more susceptible to balance issues [60,61]. For individuals with CLBP who are undergoing rehabilitation, the study’s findings have clinical ramifications. Moreover, kinesiophobia played a major role in impaired motor control associated with CLBP, and therapeutic approaches for patients with CLBP may take this into account. ## 4.2. Future Research Implications We noticed that people with CLBP had higher TSK scores and more balance issues as measured by LJPS. A higher incidence of falls is directly correlated with decreased proprioception [62,63]. Future research should examine the direct connection between the number of falls and TSK scores. Additionally, evaluating degrees of kinesiophobia and their association with the frequency of falls across age groups and genders would yield crucial data for comprehending and treating individuals with CLBP. Cognitive behavioral therapy (CBT) is a psychological approach that aims to eradicate negative feelings and behaviors, while also modifying patients’ thoughts, beliefs, and behaviors in order to correct poor cognition [64]. It is appropriate for people without mental illnesses and is characterized by integrity, initiative, enthusiasm, and a brief course of treatment [64]. According to studies, CBT not only helps patients with kinesiophobia by identifying and treating errors in automatic thought and poor cognitive behavior, but it also boosts patients’ self-efficacy, decreases anxiety and sadness, increases physical activity, and enhances quality of life [65,66]. Further research should be conducted to determine the effectiveness of CBT in lowering kinesiophobia in CLBP participants and how it affects LJPS. ## 4.3. Limitations of the Study This study has a few limitations. Even though we tested LJPS with a digital inclinometer, a motion analysis system is the gold standard for measuring LJPS, but it is expensive, difficult to operate in a clinical environment, and cannot be carried out for testing on the field. On the other hand, when compared to expensive and sophisticated equipment, the digital inclinometer is an affordable, easy-to-use gadget that takes up less space, is operated by a single rater, and takes measurements quickly. This survey reflected the limited population of Saudi Arabia (recruited from a university clinic) suffering from CLBP. We did not evaluate the associations between functional outcome assessments and kinesiophobia. Future studies should assess these factors to understand their relationship and plan rehabilitation strategies for CLBP subjects. ## 5. Conclusions This study demonstrated that individuals with CLBP had moderate kinesiophobia and were positively correlated with LJPS. LJPS is impaired in CLBP individuals compared to asymptomatic individuals. Nonetheless, the findings are expected to contribute to the scientific base of experts involved in the physical examination and rehabilitation of individuals with CLBP experiencing kinesiophobia. Furthermore, our results suggest that pain mediated the relationship between kinesiophobia and LJPS. This should be considered by therapists or clinicians when developing preventive and rehabilitative programs for CLBP patients. CBT may reduce kinesiophobia, and referral to a health counselor to help patients with CLBP and future studies should assess the efficacy of CBT in reducing kinesiophobia and its effect on LJPS. ## References 1. Kamper S.J., Logan G., Copsey B., Thompson J., Machado G.C., Abdel-Shaheed C., Williams C.M., Maher C.G., Hall A.M.. What is usual care for low back pain? A systematic review of health care provided to patients with low back pain in family practice and emergency departments. Pain* (2020) 161 694-702. DOI: 10.1097/j.pain.0000000000001751 2. Carregaro R.L., Tottoli C.R., Rodrigues D.d.S., Bosmans J.E., da Silva E.N., van Tulder M.. Low back pain should be considered a health and research priority in Brazil: Lost productivity and healthcare costs between 2012 to 2016. PLoS ONE (2020) 15. DOI: 10.1371/journal.pone.0230902 3. Fatoye F., Gebrye T., Odeyemi I.. Real-world incidence and prevalence of low back pain using routinely collected data. Rheumatol. Int. (2019) 39 619-626. DOI: 10.1007/s00296-019-04273-0 4. Oliveira C.B., Pinheiro M.B., Teixeira R.J., Franco M.R., Silva F.G., Hisamatsu T.M., Ferreira P.H., Pinto R.Z.. Physical activity as a prognostic factor of pain intensity and disability in patients with low back pain: A systematic review. Eur. J. Pain (2019) 23 1251-1263. DOI: 10.1002/ejp.1395 5. Urits I., Burshtein A., Sharma M., Testa L., Gold P.A., Orhurhu V., Viswanath O., Jones M.R., Sidransky M.A., Spektor B.. Low back pain, a comprehensive review: Pathophysiology, diagnosis, and treatment. Curr. Pain Headache Rep. (2019) 23 23. DOI: 10.1007/s11916-019-0757-1 6. Ranger T.A., Cicuttini F.M., Jensen T.S., Manniche C., Heritier S., Urquhart D.M.. Catastrophization, fear of movement, anxiety, and depression are associated with persistent, severe low back pain and disability. Spine J. (2020) 20 857-865. DOI: 10.1016/j.spinee.2020.02.002 7. Luque-Suarez A., Martinez-Calderon J., Navarro-Ledesma S., Morales-Asencio J.M., Meeus M., Struyf F.. Kinesiophobia is associated with pain intensity and disability in chronic shoulder pain: A cross-sectional study. J. Manip. Physiol. Ther. (2020) 43 791-798. DOI: 10.1016/j.jmpt.2019.12.009 8. Pazzinatto M.F., Rio E.K., Crossley K.M., Coburn S.L., Johnston R., Jones D.M., Kemp J.L.. The relationship between kinesiophobia and self-reported outcomes and physical function differs between women and men with femoroacetabular impingement syndrome. Braz. J. Phys. Ther. (2022) 26 100396. DOI: 10.1016/j.bjpt.2022.100396 9. Bilgin S., Cetin H., Karakaya J., Kose N.. Multivariate analysis of risk factors predisposing to kinesiophobia in persons with chronic low back and neck pain. J. Manip. Physiol. Ther. (2019) 42 565-571. DOI: 10.1016/j.jmpt.2019.02.009 10. Lüning Bergsten C., Lundberg M., Lindberg P., Elfving B.. Change in kinesiophobia and its relation to activity limitation after multidisciplinary rehabilitation in patients with chronic back pain. Disabil. Rehabil. (2012) 34 852-858. DOI: 10.3109/09638288.2011.624247 11. Silva M.C.d.M., Tottoli C.R., Mascarenhas K., Marques Y.A., Toledo A.M., Carregaro R.L.. Is kinesiophobia associated with disturbances in dynamic balance in individuals with chronic non-specific low back pain?. BrJP (2022) 5 47-51 12. Palacios-Ceña D., Albaladejo-Vicente R., Hernández-Barrera V., Lima-Florencio L., Fernández-de-Las-Peñas C., Jimenez-Garcia R., López-de-Andrés A., de Miguel-Diez J., Perez-Farinos N.. Female Gender Is Associated with a Higher Prevalence of Chronic Neck Pain, Chronic Low Back Pain, and Migraine: Results of the Spanish National Health Survey, 2017. Pain Med. (2020) 22 382-395. DOI: 10.1093/pm/pnaa368 13. Peterson G.E., Ludvigsson M.H.L., O’Leary S.P., Dedering Å.M., Wallman T., Jönsson M.I., Peolsson A.L.. The effect of 3 different exercise approaches on neck muscle endurance, kinesiophobia, exercise compliance, and patient satisfaction in chronic whiplash. J. Manip. Physiol. Ther. (2015) 38 465-476.e464. DOI: 10.1016/j.jmpt.2015.06.011 14. Palomo-López P., Becerro-de-Bengoa-Vallejo R., Losa-Iglesias M.E., López-López D., Rodríguez-Sanz D., Romero-Morales C., Calvo-Lobo C., Mazoteras-Pardo V.. Kinesiophobia and pain intensity are increased by a greater hallux valgus deformity degree-kinesiophobia and pain intensity in hallux valgus. Int. J. Environ. Res. Public Health (2020) 17. DOI: 10.3390/ijerph17020626 15. Ouchi K., Watanabe M., Tomiyama C., Nikaido T., Oh Z., Hirano T., Akazawa K., Mandai N.. Emotional effects on factors associated with chronic low back pain. J. Pain Res. (2019) 12 3343. DOI: 10.2147/JPR.S223190 16. Azadinia F., Kingma I., Mazaheri M.. Effect of external lumbar supports on joint position sense, postural control, and postural adjustment: A systematic review. Disabil. Rehabil. (2022) 45 753-771. DOI: 10.1080/09638288.2022.2043464 17. Alsufiany M.B., Lohman E.B., Daher N.S., Gang G.R., Shallan A.I., Jaber H.M.. Non-specific chronic low back pain and physical activity: A comparison of postural control and hip muscle isometric strength: A cross-sectional study. Medicine (2020) 99 e18544. DOI: 10.1097/MD.0000000000018544 18. Borujeni B.G., Yalfani A.. Reduction of postural sway in athletes with chronic low back pain through eight weeks of inspiratory muscle training: A randomized controlled trial. Clin. Biomech. (2019) 69 215-220. DOI: 10.1016/j.clinbiomech.2019.09.006 19. Linton S.J.. A review of psychological risk factors in back and neck pain. Spine (2000) 25 1148-1156. DOI: 10.1097/00007632-200005010-00017 20. Alahmari K.A., Reddy R.S., Silvian P.S., Ahmad I., Kakaraparthi V.N., Alam M.M.. Association of age on cervical joint position error. J. Adv. Res. (2017) 8 201-207. DOI: 10.1016/j.jare.2017.01.001 21. Hodges P.W.. Pain and motor control: From the laboratory to rehabilitation. J. Electromyogr. Kinesiol. (2011) 21 220-228. DOI: 10.1016/j.jelekin.2011.01.002 22. Reddy R.S., Alahmari K.A., Samuel P.S., Tedla J.S., Kakaraparthi V.N., Rengaramanujam K.. Intra-rater and inter-rater reliability of neutral and target lumbar positioning tests in subjects with and without non-specific lower back pain. J. Back Musculoskelet. Rehabil. (2021) 34 289-299. DOI: 10.3233/BMR-200010 23. Ghamkhar L., Kahlaee A.H.. Pain and pain-related disability associated with proprioceptive impairment in chronic low back pain patients: A systematic review. J. Manip. Physiol. Ther. (2019) 42 210-217. DOI: 10.1016/j.jmpt.2018.10.004 24. Comachio J., Magalhães M.O., Silva A.P.d.M.C.C., Marques A.P.. A cross-sectional study of associations between kinesiophobia, pain, disability, and quality of life in patients with chronic low back pain. Adv. Rheumatol. (2019) 58 8. DOI: 10.1186/s42358-018-0011-2 25. Varallo G., Giusti E.M., Scarpina F., Cattivelli R., Capodaglio P., Castelnuovo G.. The association of kinesiophobia and pain catastrophizing with pain-related disability and pain intensity in obesity and chronic lower-back pain. Brain Sci. (2020) 11. DOI: 10.3390/brainsci11010011 26. Kandakurti P.K., Arulsingh W., S Patil S.. Influence of kinesiophobia on pain intensity, disability, muscle endurance, and position sense in patients with chronic low back pain—A case-control study. Trials (2022) 23 469. DOI: 10.1186/s13063-022-06406-6 27. Varallo G., Scarpina F., Giusti E.M., Cattivelli R., Guerrini Usubini A., Capodaglio P., Castelnuovo G.. Does kinesiophobia mediate the relationship between pain intensity and disability in individuals with chronic low-back pain and obesity?. Brain Sci. (2021) 11. DOI: 10.3390/brainsci11060684 28. Leeuw M., Goossens M.E., Linton S.J., Crombez G., Boersma K., Vlaeyen J.W.. The fear-avoidance model of musculoskeletal pain: Current state of scientific evidence. J. Behav. Med. (2007) 30 77-94. DOI: 10.1007/s10865-006-9085-0 29. Larsson C., Ekvall Hansson E., Sundquist K., Jakobsson U.. Kinesiophobia and its relation to pain characteristics and cognitive affective variables in older adults with chronic pain. BMC Geriatr. (2016) 16. DOI: 10.1186/s12877-016-0302-6 30. Pagé M.G., Tousignant-Laflamme Y., Dorais M., Beaudry H., Fernet M.. Characteristics of Persons Seeking Care for Moderate to Severe Pain Due to Chronic Low Back Pain and Osteoarthritis: A Cross-Sectional Study. J. Pain Res. (2022) 15 1125. DOI: 10.2147/JPR.S360314 31. Wong C.K., Mak R.Y., Kwok T.S., Tsang J.S., Leung M.Y., Funabashi M., Macedo L.G., Dennett L., Wong A.Y.. Prevalence, incidence, and factors associated with non-specific chronic low back pain in community-dwelling older adults aged 60 years and older: A systematic review and meta-analysis. J. Pain (2022) 23 509-534. DOI: 10.1016/j.jpain.2021.07.012 32. Peng X., Bao X., Xie Y., Zhang X., Huang J., Liu Y., Cheng M., Liu N., Wang P.. The mediating effect of pain on the association between multimorbidity and disability and impaired physical performance among community-dwelling older adults in southern China. Aging Clin. Exp. Res. (2020) 32 1327-1334. DOI: 10.1007/s40520-019-01324-1 33. Alshahrani M.S., Reddy R.S., Tedla J.S., Asiri F., Alshahrani A.. Association between kinesiophobia and knee pain intensity, joint position sense, and functional performance in individuals with bilateral knee osteoarthritis. Healthcare (2022) 10. DOI: 10.3390/healthcare10010120 34. Asiri F., Reddy R.S., Tedla J.S., ALMohiza M.A., Alshahrani M.S., Govindappa S.C., Sangadala D.R.. Kinesiophobia and its correlations with pain, proprioception, and functional performance among individuals with chronic neck pain. PLoS ONE (2021) 16. DOI: 10.1371/journal.pone.0254262 35. Alahmari K.A., Rengaramanujam K., Reddy R.S., Samuel P.S., Tedla J.S., Kakaraparthi V.N., Ahmad I.. The immediate and short-term effects of dynamic taping on pain, endurance, disability, mobility and kinesiophobia in individuals with chronic non-specific low back pain: A randomized controlled trial. PLoS ONE (2020) 15. DOI: 10.1371/journal.pone.0239505 36. Alshahrani M.S., Reddy R.S.. Relationship between Kinesiophobia and Ankle Joint Position Sense and Postural Control in Individuals with Chronic Ankle Instability—A Cross-Sectional Study. Int. J. Environ. Res. Public Health (2022) 19. DOI: 10.3390/ijerph19052792 37. Asiri F., Reddy R.S., Alshahrani M.S., Tedla J.S., Dixit S., Alshahrani A., Gular K., Raizah A.. Mediation Effect of Pain on the Relationship between Kinesiophobia and Postural Control: Comparison and Correlations in Individuals with Fibromyalgia Syndrome and Asymptomatic Individuals—A Cross-Sectional Study. Life (2023) 13. DOI: 10.3390/life13010175 38. Chiarotto A., Maxwell L.J., Ostelo R.W., Boers M., Tugwell P., Terwee C.B.. Measurement properties of visual analogue scale, numeric rating scale, and pain severity subscale of the brief pain inventory in patients with low back pain: A systematic review. J. Pain (2019) 20 245-263. DOI: 10.1016/j.jpain.2018.07.009 39. Vlaeyen J.W., Kole-Snijders A.M., Boeren R.G., Van Eek H.. Fear of movement/(re) injury in chronic low back pain and its relation to behavioral performance. Pain (1995) 62 363-372. DOI: 10.1016/0304-3959(94)00279-N 40. Rosenbloom B.N., Pagé M.G., Isaac L., Campbell F., Stinson J.N., Cribbie R., Katz J.. Fear of movement in children and adolescents undergoing major surgery: A psychometric evaluation of the Tampa Scale for Kinesiophobia. Eur. J. Pain (2020) 24 1999-2014. DOI: 10.1002/ejp.1643 41. Lundberg M., Larsson M., Ostlund H., Styf J.. Kinesiophobia among patients with musculoskeletal pain in primary healthcare. J. Rehabil. Med. (2006) 38 37-43. DOI: 10.1080/16501970510041253 42. Cruz-Díaz D., Romeu M., Velasco-González C., Martínez-Amat A., Hita-Contreras F.. The effectiveness of 12 weeks of Pilates intervention on disability, pain and kinesiophobia in patients with chronic low back pain: A randomized controlled trial. Clin. Rehabil. (2018) 32 1249-1257. DOI: 10.1177/0269215518768393 43. Georgy E.E.. Lumbar repositioning accuracy as a measure of proprioception in patients with back dysfunction and healthy controls. Asian Spine J. (2011) 5 201-207. DOI: 10.4184/asj.2011.5.4.201 44. Romero-Franco N., Montaño-Munuera J.A., Fernández-Domínguez J.C., Jiménez-Reyes P.. Validity and reliability of a digital inclinometer to assess knee joint position sense in an open kinetic chain. J. Sport Rehabil. (2019) 28 332-338. DOI: 10.1123/jsr.2017-0221 45. Kang H.. *Sample size determination and power analysis using the G Power software*. J. Educ. Eval. Health Prof.* (2021) 18 17. DOI: 10.3352/jeehp.2021.18.17 46. Osumi M., Sumitani M., Otake Y., Morioka S.. Fear of movement modulates the feedforward motor control of the affected limb in complex regional pain syndrome (CRPS): A single-case study. Med. Hypotheses (2018) 110 114-119. DOI: 10.1016/j.mehy.2017.12.002 47. Demoulin C., Huijnen I.P., Somville P.-R., Grosdent S., Salamun I., Crielaard J.-M., Vanderthommen M., Volders S.. Relationship between different measures of pain-related fear and physical capacity of the spine in patients with chronic low back pain. Spine J. (2013) 13 1039-1047. DOI: 10.1016/j.spinee.2013.02.037 48. Larivière C., Bilodeau M., Forget R., Vadeboncoeur R., Mecheri H.. Poor back muscle endurance is related to pain catastrophizing in patients with chronic low back pain. Spine (2010) 35 E1178-E1186. DOI: 10.1097/BRS.0b013e3181e53334 49. Kandakurti P.K., Reddy R.S., Kakarparthy V.N., Rengaramanujam K., Tedla J.S., Dixit S., Gautam A.P., Silvian P., Gular K., Eapen C.. Comparison and Association of Neck Extensor Muscles’ Endurance and Postural Function in Subjects with and without Chronic Neck Pain—A Cross-Sectional Study. Phys. Med. Rehabil. Kurortmed. (2021) 31 295-301. DOI: 10.1055/a-1395-1050 50. Saeed Alshahrani M., Reddy R.S., Asiri F., Tedla J.S., Alshahrani A., Kandakurti P.K., Kakaraparthi V.N.. Correlation and comparison of quadriceps endurance and knee joint position sense in individuals with and without unilateral knee osteoarthritis. BMC Musculoskelet. Disord. (2022) 23. DOI: 10.1186/s12891-022-05403-9 51. Criss C.R.. Central Nervous System Contributions to Subjective and Objective Measures of Function after Anterior Cruciate Ligament Reconstruction (2021) 52. Pakzad M., Fung J., Preuss R.. Pain catastrophizing and trunk muscle activation during walking in patients with chronic low back pain. Gait Posture (2016) 49 73-77. DOI: 10.1016/j.gaitpost.2016.06.025 53. Aydoğdu O., Zübeyir S.. The association between kinesiophobia and proprioception, postural stability, activity level, knee function, and quality of life following anterior cruciate ligament reconstruction. J. Exerc. Ther. Rehabil. (2020) 7 247-252 54. Muthukrishnan R., Shenoy S.D., Jaspal S.S., Nellikunja S., Fernandes S.. The differential effects of core stabilization exercise regime and conventional physiotherapy regime on postural control parameters during perturbation in patients with movement and control impairment chronic low back pain. BMC Sport. Sci. Med. Rehabil. (2010) 2. DOI: 10.1186/1758-2555-2-13 55. Reddy R.S., Meziat-Filho N., Ferreira A.S., Tedla J.S., Kandakurti P.K., Kakaraparthi V.N.. Comparison of neck extensor muscle endurance and cervical proprioception between asymptomatic individuals and patients with chronic neck pain. J. Bodyw. Mov. Ther. (2021) 26 180-186. DOI: 10.1016/j.jbmt.2020.12.040 56. Karayannis N.V., Smeets R.J., van den Hoorn W., Hodges P.W.. Fear of movement is related to trunk stiffness in low back pain. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0067779 57. Bonanni R., Cariati I., Tancredi V., Iundusi R., Gasbarra E., Tarantino U.. Chronic pain in musculoskeletal diseases: Do you know your enemy?. J. Clin. Med. (2022) 11. DOI: 10.3390/jcm11092609 58. Hasenbring M.I., Andrews N.E., Ebenbichler G.. Overactivity in chronic pain, the role of pain-related endurance and neuromuscular activity: An interdisciplinary, narrative review. Clin. J. Pain (2020) 36 162-171. DOI: 10.1097/AJP.0000000000000785 59. Meier M.L., Vrana A., Schweinhardt P.. Low back pain: The potential contribution of supraspinal motor control and proprioception. Neuroscientist (2019) 25 583-596. DOI: 10.1177/1073858418809074 60. Allah E., AbdAlla M.G.A., Abdel-Aziz H.R.. Risk of fall and its Contributing Factors among Older adults with Chronic Low Back Pain. Ann. Rom. Soc. Cell Biol. (2021) 25 18895-18905 61. Abit Kocaman A., Aydoğan Arslan S.. Comparison of gait speed, dynamic balance, and dual-task balance performance according to kinesiophobia level in older adults. Somatosens. Mot. Res. (2023) 1-7. DOI: 10.1080/08990220.2023.2165056 62. Ito T., Sakai Y., Nishio R., Ito Y., Yamazaki K., Morita Y.. Relationship between postural stability and fall risk in elderly people with lumbar spondylosis during local vibratory stimulation for proprioception: A retrospective study. Somatosens. Mot. Res. (2020) 37 133-137. DOI: 10.1080/08990220.2020.1756243 63. Song Q., Zhang X., Mao M., Sun W., Zhang C., Chen Y., Li L.. Relationship of proprioception, cutaneous sensitivity, and muscle strength with the balance control among older adults. J. Sport Health Sci. (2021) 10 585-593. DOI: 10.1016/j.jshs.2021.07.005 64. An H., He R.-H., Zheng Y.-R., Tao R.. Cognitive-behavioral therapy. Subst. Non-Subst. Addict. (2017) 491 321-329 65. Murphy R., Calugi S., Cooper Z., Dalle Grave R.. Challenges and opportunities for enhanced cognitive behaviour therapy (CBT-E) in light of COVID-19. Cogn. Behav. Ther. (2020) 13 e14. DOI: 10.1017/S1754470X20000161 66. Hajihasani A., Rouhani M., Salavati M., Hedayati R., Kahlaee A.H.. The influence of cognitive behavioral therapy on pain, quality of life, and depression in patients receiving physical therapy for chronic low back pain: A systematic review. Pmr (2019) 11 167-176. DOI: 10.1016/j.pmrj.2018.09.029
--- title: Adherence to the Mediterranean Diet Related to the Health Related and Well-Being Outcomes of European Mature Adults and Elderly, with an Additional Reference to Croatia authors: - Manuela Maltarić - Paula Ruščić - Mirela Kolak - Darija Vranešić Bender - Branko Kolarić - Tanja Ćorić - Peter Sousa Hoejskov - Jasna Bošnir - Jasenka Gajdoš Kljusurić journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049455 doi: 10.3390/ijerph20064893 license: CC BY 4.0 --- # Adherence to the Mediterranean Diet Related to the Health Related and Well-Being Outcomes of European Mature Adults and Elderly, with an Additional Reference to Croatia ## Abstract With the increase in life expectancy, expectation of a longer healthy life is also increasing. Importance of consumption of certain foods is confirmed to have a strong effect on quality of life. One of the healthiest dietary patterns consistently associated with a range of beneficial health outcomes is the Mediterranean diet (MD). The aim of this study was to assess MD adherence in the population over 50 years of age, in Europe, with special reference to Croatia, and to assess regional differences and investigate the association with health-related indicators (disease incidence, body mass index (BMI), grip strength measure, control, autonomy, self-realization, and pleasure scale (CASP-12)). This research uses data from the SHARE project for the population over 50 years of age. The frequency of individual responses was analyzed (frequencies, cross tables, and appropriate tests of significance, depending on the data set), and logistic regression was used to connect adherence to the Mediterranean diet with health indicators. The results of the study indicate a positive correlation between adherence to the principles of the Mediterranean diet with CASP and self-perception of health, which the followers of the MD pattern predominantly rate as “very good” or “excellent” ($37.05\%$) what is significantly different ($p \leq 0.05$) from individuals which do not follow the patterns of MD ($21.55\%$). The regression models indicate significant changes in the measure of maximum grip strength also among MD followers (ORMEDIUM = 1.449; ORHIGH = 1.293). Data for EU countries are also classified by regions (Central and Eastern; Northern, Southern and Western Europe), additionally allocating Croatia, and the trends in meat, fish and egg consumption showed the greatest differences for Croatian participants ($39.6\%$ twice a week) versus participants from four European regions. Data for Croatia deviates from the European average also in terms of the proportion of overweight and obese persons in all observed age groups, of which the largest proportion is in the 50–64 age group (normal BMI: only $30.3\%$). This study extended the currently available literature covering 27 European countries, placing the findings in a wider geographical context. The Mediterranean diet has once again proven to be an important factor related to health-related behavior. The presented results are extremely important for public health services, indicating possible critical factors in preserving the health of the population over 50 years old. ## 1. Introduction One of the critical public health issues in the elderly population is considered to be health-related behavior (HRB) in which diet is a central health-related behavior that needs to be investigated along with factors such as physical activity [1]. With the increase in the development of European countries, life expectancy is also increasing, as well as the expectation of a longer healthy life [2] for men 82.9 years and for women 85.1 years. The share of people over 65 years of age in the EU population was $20.6\%$ (in 2020), and it is expected that the share of this population group will increase to $40.6\%$ by 2050 [3]. Recently conducted research [1,4] indicates an increase in the number of sick people in general, and especially in the group of the elderly population, which among the many causes singles out the reduced healthy behavior of individuals, and thus the weaker general well-being of the individual. In 2019, there were $42.28\%$ of people over 50 years old in the Republic of Croatia [5]. In 2021, the aging of the population continued, and the average age of the total population of the Republic of Croatia was 44.3 years (men 42.5 years, women 45.9 years), which ranks it among the oldest nations in Europe [6]. The highest share of the population aged 65 and over, in relation to the total population, was in Šibenik-Knin County ($27.3\%$) and in Lika-Senj County ($26.2\%$), and the lowest in the city of Zagreb ($20.6\%$) and in Međimurje County ($20.5\%$). According to Eurostat’s Nomenclature of territorial units for statistics (NUTS), the mentioned counties are two different regions, so the first two counties are classified as Adriatic Croatia, and the last two as continental Croatia [7]. NUTS-2 regions between 2012 and 2020 were Adriatic and Continental Croatia, and since 2021, the Continental Croatia was additionally divided to three regions: Pannonian Croatia, Adriatic Croatia and Northern Croatia [8]. At the age over 50, and especially over 65, the rate of getting sick from various diseases increases [9]. Based on the Eurostat data, among the most common diseases are ischemic heart disease, malignant diseases of the respiratory system, diabetes, cirrhosis of the liver, cerebrovascular insult, malignant cancers of the colon and rectum, COPD, dementia, various kidney diseases and hypertensive heart disease. As chronological age increases, various changes occur at the molecular level that affect the entire organism in various ways [3]. However, in all population groups, including the population over 50 years of age, the proportion of overweight and obese people has increased. Due to increased food intake, hypertriglyceridemia and hypercholesterolemia occur, which favors the development of atherosclerosis of blood vessels and increased frequency of myocardial infarction and as a result of compensatory hyperinsulinemia, which occurs due to a decrease in the sensitivity of insulin receptors, diabetes occurs [10]. It is also known that in certain age increases the share of malnourished population. Malnutrition most often occurs either as a primary disease due to increased loss of nutrients or decreased consumption of foods or as a result of other diseases where is of particular concern the cachexia–state of extreme malnutrition and physical and metabolic exhaustion due to a serious illness [11]. According to all previously mentioned, it is clear that the increased involvement of public health services is due to the extension of life expectancy, setting as the main goal of the services to increase the number of years of healthy life and improve the quality of life in those years. In addition to public health service, prevention strategies may also be needed including the establishment of an enabling physical and institutional environment for a healthy living. Importance of consumption of certain food groups is confirmed to have a strong effect on quality of life [4,12] where consumption of fruits and vegetables is singled out to be positively associated with subjective health and quality of life, along with physical activity and good sleep where the middle-age adults is the better sleep more relevant while in the older population it is the consumption of fruits and vegetables [12]. Among the dietary patterns in which the emphasis is on the intake of fruits and vegetables, the Mediterranean diet (MD) stands out [13] and features of this diet are everyday consumption of fruits, vegetables, olive oil, wholegrain breads, cereals, herbs, spices and frequent intake of eggs, beans or legumes, fish and seafood. Studies show that MD adherence is associated with a lower risk of a number of infectious diseases and better health outcomes [14,15], which is particularly important for the population group of mature and elderly people. However, in the population over 50, there is an increased risk of impaired health, which increases proportionally with age. Vall Castello and Tubianosa [13] used the method of Mediterranean diet adherence (MD) in Europe, concluding on the population of adults aged 50 years and above that following the Mediterranean diet was negatively correlated with the incidence of chronic illnesses, as well as with levels of depressive symptoms, conducted on data collected during 2011–2017. The main objectives of this study are to investigate the Mediterranean adherence and its relation with the health-related indicators (disease incidence, maximum of grip strength measure, CASP-12 (control, autonomy, self-realization, and pleasure scale) and body mass index)) in the middle-aged and elderly population in Europe with a special emphasis on participants from Croatia (from the SHARE base). Results for Croatia, from the database SHARE are being presented for the first time. The objectives are focused to test and verify: (i) if relationships exist between adhering to the Mediterranean diet, or not (food intake and the frequency of consumption of certain food groups) and variations in health outcomes of individuals (as the incidence of chronic diseases), their levels of body mass index (BMI) and the levels of self-assessed well-being; (ii) if correlations can also be observed between physically active and healthy lifestyle with the same health outcomes. All objectives will be tested on a large database that includes 27 European countries and if the mentioned connections prove to be significant, they will have exceptional value for all participants in food production and processing chain up to the food on the table, intended for the population included in this study. For the first time, data for the Republic of Croatia were additionally separated for this population group. ## 2.1. Study Design This study uses data from the SHARE project (Survey of Health, Ageing and Retirement in Europe), wave 8 [2019], release 8.0.0 [16]. SHARE is a European multidisciplinary and cross-national panel database of microdata on health, socioeconomic status and social- and family networks. Croatia participates in SHARE since 2004 [17], and SHARE data set contains representative samples of the 50+ population in each European country, plus Israel. Data is collected in face-to-face interviews using the computer-assisted personal interviewing (CAPI) method. Proxy interviews are allowed when respondents are unable to participate in the survey, such as health reasons. For further methodological details of the SHARE project, please see [18]. In wave 8 participated 27 European countries (Austria, Germany, Sweden, Netherlands, Spain, Italy, France, Denmark, Greece, Switzerland, Belgium, Israel, Czech Republic, Poland, Luxembourg, Hungary, Slovenia, Estonia, Croatia, Lithuania, Bulgaria, Cyprus, Finland, Latvia, Malta, Romania and Slovakia) and Israel participated in the SHARE project. From the total sample size ($$n = 114$$,199) were excluded data for participants under 50 years ($0.5\%$, $$n = 236$$). The baseline indicators of the included population group are presented in Table 1. ## 2.2. Mediterranean Diet This survey contains questions that are intended to assess the quality of the dietary patterns of the participants. The quality of the diet was assessed using these questions:How often do you eat a serving of dairy products which means a cup of yogurt, cheese, a glass of milk, or a can of high protein supplements weekly?How often do you eat a serving of eggs, beans, or legumes weekly?How often do you eat a serving of poultry, meat, or fish weekly?How often do you eat a serving of fruits and vegetables weekly? The participants could choose following answers:Less than once a week;Once a week;Twice a week;3–6 times a week andEvery day. In the study of Alves and Perelman [14], it was indicated that with this data, it is possible to assess whether the participants followed the principles of the Mediterranean diet which is recognized as a recommended diet pattern. Although the Mediterranean diet is accompanied by the consumption of whole grain products, olive oil/olives, moderate wine consumption, according to previously published works that used data from the SHARE database, a recommendation was made based on 4 researched food groups [13,14,15]. Mediterranean diet is based on the everyday consumption of veggies and fruits and frequent intake of eggs, beans, legumes, meat, fish, or poultry (three-six times weekly [13,14]). From the collected data of the frequency of the before mentioned 3 food groups (not including the dairy) we have created a binary index identifying which of the participants may be classified as subscribing to a Mediterranean diet (value = 1 if following the diet; 0 if not following the diet). Limitation of this database is that it does not separate red meat from the other types of meat [15] but results of this study will be compared to those studies that had pointed out the same limitation. A summary of the frequency of food consumption from four food groups is shown in Table A1 (note—the consumption of dairy products is not included in the calculation of MD indicators). The main limitation of the above-mentioned estimation of MD adherence is the partial analysis due to the imprecise characterization of this dietary pattern. By such assessing of MD adherence, some key foods such as seeds and oilseeds, the consumption of olive oil, red wine and seafood [17], and the separation of the type of meat (red meat, cured meat products) are not included. ## 2.3. Health Related Indicator Two parameters were used as an indicator of population health, the first of which is related to nutrition–the body mass index, and the second is related to the most common cardiovascular and metabolic diseases (CMDs). The disease indicator was generated according to the study by [13] and a chronic disease variable was generated that included diseases related to cardiovascular diseases (myocardial infarction and other heart problems, hypertension, high blood cholesterol, stroke) and type 2 diabetes. Possible outcomes for this variable are 0 (absence of any of the mentioned diseases) and a maximum of 5, which indicates that the respondent stated that he suffers from all the mentioned diseases. Incidence of Cardio Metabolic Diseases for the population included in this study is shown in Table A2. The body mass index was calculated from the data on the body weight and body height of the subjects, and the final values were grouped into the groups malnourished, normally nourished and overweight and obese depending on age. For the age group up to 65 years, values greater than 25 indicate excess body mass and obesity [19], while according to the ESPEN guidelines for people older than 65, the limit has been moved and the range of normal nutrition is 21–27.5 kg/m2 [20]. The classification itself and the basic overview of the population group is given in Table A3 and Table A4. ## 2.4. Maximum of Grip Strength Measure Hand grip strength (HGS) is the maximum static force applied by the hand and it is measured by a dynamometer. It is simple and inexpensive to evaluate. Furthermore, this measurement is widely applied not only to assess hand function after injury and the outcome of hand surgery, but also to reflect general physical health and disability [21,22,23]. Lower HGS as concluded in a longitudinal study was connected to increased mortality risk from cardiovascular diseases and cancer, even after adjusting for body composition, multiple chronic diseases, and multi-morbidity, in both sexes [21,24]. ## 2.5. CASP-12 Two parameters were used as an indicator of population health, the first of which is CASP-12 (control, autonomy, self-realization, and pleasure scale) is one of the most common internationally used measures for quality of life in older adults, although its structure is not clearly established [25]. This scale is theoretically driven by the ‘needs satisfaction’ approach to measure QoL in early old age. It is based on Maslow’s Hierarchy of Needs [26]. The CASP-12 scale is a modification of the original CASP-19 [27]. Authors from study [28] conducted analyses on all countries in SHARE Wave 4 and concluded that the theoretical four-factor structure of the CASP did not meet the data properly. Another author [29], employing Item Response Theory (IRT) models, found a bifactor model with a strong global factor of QoL to better represent CASP-12 scores, with data used from SHARE Wave 6 [25]. Nevertheless, other studies have found good fit for the four-factor theoretical structure in the CASP-12 [25]. ## 2.6. Statistical Analyses Before the actual data processing, the missing data analysis was performed. For this cross-sectional study, the first step was to analyze missing data by country, because in the study by [30]. pointed out that data analysis will be credible if the proportion of missing values is less than $5\%$, which is not the case for some variables (e.g., economic and health indicators). Therefore, the procedure of a study [31] applied in this paper as well, and it suggests multiple imputation of missing values to increase the number of observations. After including these imputed variables, missing data remained (less than $1\%$, per country). In the SHARE documents are available coded data with nominal and scale variables. Based on the data type, different tests were applied to assess the differences in the observed characteristics of the examined population, statistical tests were performed to compare the two groups (t-test (t) and chi-square test (χ2)). Individuals following the Mediterranean diet (MD) were compared with those individuals who did not follow it. Data analysis was performed with covariates which included; age (ranged from 51–104), gender (female/male), marital status (all married or living with the partner were categorized as “living with partner” and the widowed, divorced, etc. are classified in the group “Not living with partner”), education level (primary (9 years of education); secondary (maximal 12 years of education) and tertiary (more than 12 years of education), economic status (household able to meet ends meet (with difficulty is defined as poor; fairly easily as fair and easily as good), employment (retired = retired while other categories were defined as not retired), self-perceived health (poor and fair= Poor/Fair, good = Good and very good and excellent is grouped as Very good/Excellent), the CASP index for quality of life and well-being (12–24 = Low; 25–36 = Medium; >36 = High), Sport or activities that are vigorous (Never or rare, At least once a week summarized all activities including at least activity once a week), CMD (0 = none, 1–5 = one or more diagnoses), BMI (those who fall into the category of normally nourished = normal, depending on the age group (18.5–25 kg/m2 for the aged under 65, and 21–27.5 for those over 65 years) and the max. of grip strength measure (1–25 = Low; 26–50 = Medium and >50 = High). Logistic regression was applied to explore variable changes related to adherence of Mediterranean diet (following vs not following). Crude regression model (CRM) observed difference based on the MD, while the bivariate-adjusted model included age, gender, and marital status (M1). To remove their influence on the adherence of MD, the next bivariate-adjusted model (M2) included Model 1 observations and education level, economic status, and employment. Odds ratios (ORs) with corresponding $95\%$ confidence intervals (CIs) were estimated for all models (Crude, Model 1, and Model 2). OR > 1 indicates a greater likelihood of an association with the exposure and outcome or an increased incidence of the investigated event. The $95\%$ confidence interval for the OR is given in square brackets, and statistical significance was tested with the chi-square test. Data analysis software SPSS v. 19 was used. ## 3. Results Before the conducted regression analysis, the baseline characteristics of the population group were determined and presented in Table 1 where the results are presented as frequencies (%), for each observed variable. On Figure 1A can be seen that the mean values of maximum grip strength for the participants who followed MD mostly show lower values than the participants who were not following MD, with a few peaks in the other direction. Contrary to this phenomenon, on Figure 1B is present an almost inverted situation meaning that the participants who were not following MD show lower mean CASP index for quality life and well-being than the participants who were following MD. In the study of [13] is already confirmed that the incidence of CMDs is slightly lower in those who follow the MD, so in Figure 1 are presented the similarities/differences in the max. of grip strength measure and index for quality of life and well-being for those participants who follow (or not follow) the patterns of the Mediterranean diet. Those results clearly indicate that participants following MD in general have higher CASP. There are few observations that can be spotted from Table 1. First, we can see that the distribution of “Not following MD” and “Following MD” among age and sex of the participants is almost equal, as well as the marital status, educational level and employment. The only variable that stands out from these “input” data is economic status, which is greater for the participants who were not following MD. This is quite self-explanatory because the majority of people cannot afford MD since it more expensive This statement is a result of a study published by Rubini and coworkers [32] which included 2833 subjects, concluding that the degree of adherence to the MD eating pattern was positively correlated with the monthly cost, emphasizing that the required economic effort to afford the MD is higher [32]. Self-perceived health is the highest for the participants following MD and lowest for participants not following MD which means that the participants are aware of the benefits the MD offers. Sports activities and CMD consequently followed by body mass index also presents higher percentage at the “Following MD” participants, while maximum of grip strength measure show slightly bigger values among “Not following MD” participants. This also matches the data observed from Figure 1. Croatian participants included in the study present $2.5\%$ valid samples used for the analyses ($$n = 1197$$). The characteristics based on observed significant variables for the Croatians are presented in Table 2. The data obtained from Table 2 show almost equal input parameters as Table 1, but with the segregation of the participants among the Adriatic and Continental parts of the country. Additionally, the characteristics of the population as age, gender and marital status, are not equally distributed among groups, so the conclusion drawn out from this table can only by partially taken. At this age, the mortality of the male population is higher [28] and the majority of participants who stated that they did not live with a partner were widowed [14,25]. The most notable observations can be spotted in the CASP parameter: no participant from Adriatic part following MD declared low values, while the highest values were observed from both Adriatic areas (“Following” and “Not following MD”). Maximum of grip strength measure was noted 0 at the Following MD Adriatic participants. The data from Table 3 tell us that the most participants in these conducted experiments fit in the overweight BMI category with almost $40\%$, while most of them aging between 65 and 74 years. The lowest frequency with less than $1\%$ can be seen at the underweight category. Means value of the BMI was 26.25 kg/m2 with a standard deviation of ±6.85 kg/m2 and in Table 4 are presented the shares of the population with the normal BMI according to the corresponding age group, per countries. The most notable observation from Table 4 is that the countries with the highest BMI from 50 to 64 years of age are Switzerland, France and Netherlands with values 48.4, 42.2 and 51.8, respectively. BMI from 65 to 74 years of age are participants from Sweden, Italy and Switzerland with 61.2, 59.7 and 57.3, respectively. BMI for the participants aging between 75 and 84 years are obtained for Sweden, Denmark and again Switzerland, with 62.4, 59.8 and 58.5, respectively. Lastly, the highest BMI values for the participants older than 85 years are obtained for Netherlands, Germany and Austria, with values 63.7, 61.8 and 60.7, respectively. The lowest values of BMI for all age groups are spotted for Malta with 14.6, 23.8, 30.2 and 24.4, respectively. Figure 2 shows trends in nutrition of the European population by age and additionally by gender in the Republic of Croatia. There is a clear trend of a higher proportion of overweight in the male population, regardless of age group, and the proportion of obese people is more pronounced in the female population. An increase in the proportion of malnourished people with increasing age is also evident. The Croatian population, which is in an unenviable, leading position in terms of the proportion of overweight and obese in the population, was additionally analyzed with the aim of determining the age groups in which additional education/intervention is necessary (Figure 2B). Finally, the logistic regressions were used to investigate the Mediterranean diet adherence (Following MD) in relation of variables that are associated to the health and well-being of the investigated participants (self-perceived health (SPH), index for quality of life and well-being (CASP), vigorous sport activities, CMDs, body mass index (BMI) and the maximum of grip strength measures). The models were adjusted for age, gender, marital status, education level, employment and economic status and the final result is presented in Table 5 Adjusted odd ratios indicated that with a lower SPH assessment, the adherence to the principles of the Mediterranean diet is lower (OR = 0.921), in contrast to those who rated it as good (OR = 0.927) and especially those with a rating of very good and excellent (OR = 1.541). The CASP index for quality of life and well-being showed a significant increase in following the MD (ORLow = 1.095 vs. ORHigh = 1.195, $p \leq 0.01$). For respondents who regularly engage in vigorous sport activities the relation with the MD adherence is significant (ORregular = 0.910 vs. ORNever = 0.942, $p \leq 0.01$). The cardiovascular and metabolic diseases variable indicated higher MD adherence if the participants have at least one of the diagnoses (ORnone = 0.981 vs. ORNever = 1.019). Regardless of which age group it is, the proportion of those with a normal BMI (up to 65 divisions according to standard tables, after 65 according to ESPEN guidelines) was singled out, the connection with following the guidelines of the Mediterranean diet is evident, indicating that in most of those who following the principles of the Mediterranean diet, a normal BMI is expected (OR = 1.054 with a CI = 0.993 to 1.119). The maximum of grip strength measure showed significant increase for low to high outcomes (ORLow = 0.774 vs. ORHigh = 1.293). Table A4 shows the frequency of food consumption from specific groups for four European regions as well as for two Croatian regions, while Table A5 shows the systematized incidence of diseases in the population over 50 years of age, also presented according regions [7,8,32]. Southern Europe countries dominate in the every day intake of dairy products, legumes, beans, eggs and meat, fish and poultry, while for the last food group, fruits and vegetable consumption on the everyday basis, the first place is shared with the countries of Western Europe ($74.5\%$). Statistically significant differences were indicated in the frequency of consumption of fruits and vegetables in the Northern European countries ($61.7\%$ $p \leq 0.05$) what is in accordance with the different nutritional habits in EU countries [33]. Incidence of observed diseases which were available in the SHARE database were systematically presented according four European regions with use of heatmap principles and their indication of high or low values by use of colors. Southern European countries have the lowest proportions of patients with $50\%$ of all twenty observed health issues (marked in dark green). Western countries have the highest proportion of population for whom depression was ever diagnosed/currently having ($43\%$ vs. 33.6 % in Southern European countries), as well as the Osteoarthritis/other rheumatism where the incidence is significantly higher ($25.9\%$ vs. $19.4\%$ in Central and Eastern countries; $18.8\%$ in Northern countries and $15.2\%$ in Southern European countries, respectively). Croatia follows the trend of Central and Eastern Europe related to the number of people suffering from the mentioned diseases (Table A5). ## 4. Discussion Presented results confirmed the relationship of the adherence MD and the health and well-being outcomes. The results firstly indicate the parameter as maximum of grip strength measure did not indicate significant relation with the adherence MD (Figure 1A) while the logistic regression models indicatted there was a significant relationship (Table 5). This is an outcome confirmed by Tak and coworkers [21] which investigated the relation of the grip outcome and diet. Different studies investigated the relationship between the CASP and adherence of MD [9,13]. Our findings confirm the proportional relation of the CASP outcomes to MD diet adherence, so if a person has a higher CASP, it is more likely that they will also follow the principles of the Mediterranean diet. Characteristics of the population related to following the MD indicated statistically significant differences based on gender (in favor of female population), living with the partner, having fair economic status and high CASP index, as well as for those participants who self-perceived their health as very good/excellent. Similar findings were confirmed with the difference that CMDs were in negative relation to MD followers [13]. Croatia specific geographic location (HR) is included in the Nomenclature of Territorial Units for Statistics (NUTS) of the European Union. The two defined NUTS2 regions since (2012 to 2021) was Adriatic and Continental Croatia. Since 2021, Croatia is divided to 4 (non-administrative) regions where Adriatic Croatia remained the same region, and continental Croatia was divided into three ones: Pannonian and Northern Croatia and the city of Zagreb) [7,8]. This is the reason why the characteristics of Croatian population is presented based on two regions (Table 1), where statistically significant differences on the regional level were detected for MD followers (in the largest age group (65–74 years old), the employment status, CMDs (absence of any) and those with the lowest outcome for the maximal of grip strength measure. Relating the regions with food preferences of the inhabitants, it is certainly interesting to study the similarities and/or differences in the two regions; Adriatic and Continental, because the first should definitely have a Mediterranean character, as well as to compare them with other countries in the region. In the world population the prevalence of overweight and obesity is increasing in all age groups, including the elderly [34,35]. Therefore, the concern for this population group is growing because it is known that obesity is associated with serious medical complications and in the elderly can accelerate age-related decline in physical function, while the underweight is related with some health hazards, such as reduced bone density and fractures [36]. Normal BMI of elderly is associated with significant improvements in quality of life, physical function and health [37], therefore Table 3 and Table 4 and the Figure 2A,B are dedicated to the review of the normally nourished in all countries with separate data for Croatia and the review of the share of the normally nourished and other nutrition groups, according to gender and age groups. The share of obese is higher in the female population, while the overweight is dominant for the male population, and the concerning nourishment group that raises with the age is the underweighted. Studies show association of the lower household income and national income as well, with population obesity prevalence [38,39] and it seems that in Croatia this is one of the factors of the extremely high proportion of overweight and obese people in the observed population group. The share of underweighted in EU countries is under the share indicated for Africa where nearly one among five older people was undernourished [35] and about $30\%$ of African elderly population were overweight or obese [35] while in the EU population, it is over $50\%$ for all age groups. In the Croatian population were not detected any significant differences in average values for different gender, while it is obvious that the average values fall after the age of 85 years, but not significantly ($p \leq 0.05$). Finally, in order to confirm or dismiss the positive effect, of following the Mediterranean diet, on health indicators as well as on well-being indicators, logistic regression models were used, which were adjusted considering additional indicators of the observed group such as age, gender, marital status, education level, employment and economic status. Previous investigations [13,17] confirmed the influence of education, retirement and economical status on the outputs as health and well-being, therefore those variables were used in the model adjustment. So, the MD adherence is related to the “et least self-perceived” health as good, CASP at least as Medium, with no regular vigorous sport activities. At least one or more CDMs will make the Mediterranean way of eating more attractive because Mediterranean diet is associated with better cardiovascular health outcomes, including clinically meaningful reductions in rates of coronary heart disease, ischemic stroke, and total cardiovascular disease [40]. The final parameter hand grip strength showed increased significantly relations with the MD adherence what is in accordance with studies investigating the association of adherence to a Mediterranean diet with BMI, muscle strength of adults with diabetes [41], older adult women [42] or in the Asian population [43,44]. Hence, the transition to a healthier diet, such as the Mediterranean diet, is related to the well-being and health of the population, therefore this study points to important factors that should be changed, as a whole, because failure to correct (increase) the mentioned indicators of health and well-being will result additional deterioration of the health of the mature and elderly population. Although the frequency of consumption of the food group fruit and vegetables does not differ significantly for three different European regions, it is important to note the published results of Eurostat in which income plays an important role in the consumption of this group of foods [34]. Consumption of foods from the group’s Dairy products and Fruits and vegetables, of Croatian elderly population (Adriatic and Continental Croatia) is in accordance with the trend of Southern, Western and Central and Eastern Europe countries (Table A4). However, in both Croatian regions, the regular consumption of foods from the group of Legumes, beans, eggs and Meat, fish, poultry is significantly different from the European trend. Research conducted in Croatia, which included socially organized nutrition and the relation with the Mediterranean diet pattern, confirmed that menus in Adriatic region showed the Mediterranean pattern, while Continental ones did not [45]. However, in the characterization of the Mediterranean diet of the elderly population in Croatia (Table 2), an almost negligible share of people from the continental part of Croatia, who follow the Mediterranean diet, was singled out. Although the continental part of Croatia was more represented in the number of participants (919 vs. 297 participants), it should also be noted that precisely in the continental part, especially in the city of Zagreb, there is a significantly larger number of institutions that offer organized nutrition and accommodation for the elderly, and the Mediterranean diet pattern is thus represented in the group of those people who use such accommodation. In two southern European countries, as the traditional diet of northern Portugal and northwestern Spain, the South European Atlantic Diet (SEAD) is present, and adherence to the SEAD has been shown to be associated with a lower risk of death from all causes among the elderly in Spain [46]. Eurostat data [34,47] confirm the findings of regional diversity, which is also reflected in the frequency of certain diseases. Data published in 2019 showed that more that $83\%$ of all deaths in Europe occurred among the elderly (people aged 65 years and over) and the standardized death rate from circulatory diseases was almost five times higher in Bulgaria (Central and Eastern country) than in Spain (Southern country) [43]. Cancer as cause of death is most common in Hungary (327.7 per 100,000 inhabitants) and Croatia (310.9 per 100,000 inhabitants) as a part of the region (Central and Eastern countries) with the highest death rate of 285.5 per 100,000 inhabitants), followed by the Northern countries (259.8 per 100,000 inhabitants), Western countries (237.5 per 100,000 inhabitants) and Southern countries (222.3 per 100,000 inhabitants) [47]. Mediterranean diet pattern rebounds by positive contribution to health status, representing the best nutritional strategy for obtaining a great combination of nutrients, antioxidants and other beneficial molecules able to promote the healthy aging process [48]. Independently of the above, we conducted an additional analysis of the Mediterranean diet and all observed diseases, and we determined the dominance of the Mediterranean diet in those diseases that are most prevalent in the European population as well in the Croatian. Namely participants who were following MD pointed out as their health problem (i) long-term illness ($63.7\%$), (ii) high blood pressure or hypertension ($57.7\%$) and (iii) diabetes ($49.6\%$). Bearing in mind that the Mediterranean diet has positioned itself in a leading position in diet therapy [49,50], the presence of the Mediterranean dietary pattern is clearer in the continental part of Croatia as well, and in general among the population that wants to maintain their health status through diet. ## 5. Conclusions Our findings confirmed the association of the studied indicators of health and general well-being with MD adherence, for people over 50 years of age. Our study extended the currently available literature covering 27 European countries, placing the findings in a wider geographical setting, however, with the same aspiration to achieve the ultimate goal of caring for the middle-aged and elderly population, their health and general well-being. The Mediterranean diet has once again proven to be an excellent aid in maintaining health. The results of the study also established a positive correlation between adherence to the Mediterranean diet with CASP and self-perceived health, which is predominantly assessed as “very good” or “excellent” by people following the MD regimen. A higher measure of grip strength was found precisely in the MD followers. Data for EU countries, including the data for the Republic of Croatia, have the same trends, but unfortunately, the Republic of Croatia deviates from the average in terms of the proportion of overweight and obese people. All the above results are extremely important for public health services, indicating possible critical factors in maintaining the health of this population. ## References 1. Rhodes R.E., Janssen I., Bredin S.S.D., Warburton D.E.R., Bauman A.. Physical activity: Health impact, prevalence, correlates and interventions. Psychol. Health (2017.0) 32 942-975. DOI: 10.1080/08870446.2017.1325486 2. Gadó K., Szabo A., Markovics D., Virág A.. Most common cardiovascular diseases of the elderly—A review article. Dev. Health Sci. (2022.0) 4 27-32. DOI: 10.1556/2066.2021.00048 3. Prevalence of the Main Longstanding Health Problems by Sex and Age. (2021.0) 4. Zhang Z., He Z., Chen W.. The relationship between physical activity intensity and subjective well-being in college students. J. Am. Coll. Health (2022.0) 70 1241-1246. DOI: 10.1080/07448481.2020.1790575 5. Population Structure and Ageing. European Union: European Commission. (2022.0) 6. Procjena Stanovništva Republike Hrvatske u 2021. (Estimates of the Population of the Republic of Croatia In 2021). (2022.0) 7. “Nacionalna Klasifikacija Prostornih Jedinica za Statistiku (NKPJS)”. Narodne Novine (in Croatian) (35/07). (2007.0) 8. “Nacionalna Klasifikacija Statističkih Regija 2021. (HR_NUTS 2021)”. Narodne Novine (in Croatian) (125/2019). (2019.0) 9. Baena C., Joarder T., Ahmed N.U., Chowdhury R.. Aging and the COVID-19 pandemic: The inter-related roles of biology, physical wellbeing, social norms and global health systems. Maturitas (2022.0) 167 99-104. DOI: 10.1016/j.maturitas.2022.07.008 10. Chu D.-T., Nguyet N.T.M., Dinh T.C., Lien N.V.T., Nguyen K.H., Ngoc V.T.N., Tao Y., Son L.H., Le D.-H., Nga V.B.. An update on physical health and economic consequences of overweight and obesity. Diabetes Metab. Syndr. (2018.0) 12 1095-1100. DOI: 10.1016/j.dsx.2018.05.004 11. Evans C.. Malnutrition in the Elderly: A Multifactorial Failure to Thrive. Perm. J. (2005.0) 9 38-41. DOI: 10.7812/TPP/05-056 12. Tan S.L., Storm V., Reinwand D.A., Wienert J., De Vries H., Lippke S.. Understanding the Positive Associations of Sleep, Physical Activity, Fruit and Vegetable Intake as Predictors of Quality of Life and Subjective Health Across Age Groups: A Theory Based, Cross-Sectional Web-Based Study. Front. Psychol. (2018.0) 9 977. DOI: 10.3389/fpsyg.2018.00977 13. Castelló J.V., Tubianosa C.. Linking Mediterranean Diet and Lifestyle with Cardio Metabolic Disease and Depressive Symptoms: A Study on the Elderly in Europe. Int. J. Environ. Res. Public Health (2020.0) 17. DOI: 10.3390/ijerph17197053 14. Alves R., Perelman J.. European mature adults and elderly are moving closer to the Mediterranean diet—A longitudinal study, 2013–2019. Eur. J. Public Health (2022.0) 32 600-605. DOI: 10.1093/eurpub/ckac070 15. Alves R.. Mediterranean Diet in Europe: How are mature adults and elderly moving closer to this diet pattern?. Eur. J. Public Health (2021.0) 31 ckab164.447. DOI: 10.1093/eurpub/ckab164.447 16. Börsch-Supan A.. Survey of Health, Ageing and Retirement in Europe (SHARE) Wave 8. 2022; Release Version: 8.0.0. SHARE-ERIC. Data Set (2022.0). DOI: 10.6103/SHARE.w8.800 17. SHARE Hrvatska. (2022.0) 18. Börsch-Supan A., Brandt M., Hunkler C., Kneip T., Korbmacher J., Malter F., Schaan B., Stuck S., Zuber S.. Data Resource Profile: The Survey of Health, Ageing and Retirement in Europe (SHARE). Int. J. Epidemiol. (2013.0) 42 992-1001. DOI: 10.1093/ije/dyt088 19. Schmerling R.H.. How Useful if the Body Mass Index (BMI). Harvard Health Blog, Harvard Health Publishing, Harvard Medical School 20. Volkert D., Beck A.M., Cederholm T., Cruz-Jentoft A., Goisser S., Hooper L., Kiesswetter E., Maggio M., Raynaud-Simon A., Sieber C.C.. ESPEN guideline on clinical nutrition and hydration in geriatrics. Clin. Nutr. (2019.0) 38 10-47. DOI: 10.1016/j.clnu.2018.05.024 21. Tak Y.J., Lee J.G., Yi Y.H., Kim Y.J., Lee S., Cho B.M., Cho Y.H.. Association of Handgrip Strength with Dietary Intake in the Korean Population: Findings Based on the Seventh Korea National Health and Nutrition Examination Survey (KNHANES VII-1). Nutrients (2016.0) 10. DOI: 10.3390/nu10091180 22. Bohannon R.W.. Muscle strength: Clinical and prognostic value of hand-grip dynamometry. Curr. Opin. Clin. Nutr. Metab. Care (2015.0) 18 465-470. DOI: 10.1097/MCO.0000000000000202 23. Scott D., Park M.S., Kim T.N., Ryu J.Y., Hong H.C., Yoo H.J., Baik S.H., Jones G., Choi K.M.. Associations of low muscle mass and the metabolic syndrome in Caucasian and Asian middle-aged and older adults. J. Nutr. Health Aging (2016.0) 20 248-255. DOI: 10.1007/s12603-015-0559-z 24. Rantanen T., Volpato S., Ferrucci L., Heikkinen E., Fried L.P., Guralnik J.M.. Handgrip Strength and Cause-Specific and Total Mortality in Older Disabled Women: Exploring the Mechanism. J. Am. Geriatr. Soc. (2003.0) 51 636-641. DOI: 10.1034/j.1600-0579.2003.00207.x 25. Oliver A., Sentandreu-Mañó T., Tomás J., Fernández I., Sancho P.. Quality of Life in European Older Adults of SHARE Wave 7: Comparing the Old and the Oldest-Old. J. Clin. Med. (2021.0) 10. DOI: 10.3390/jcm10132850 26. Maslow A.H.. Toward a Psychology of Being (1968.0) 27. Hyde M., Wiggins R.D., Higgs P., Blane D.B.. A measure of quality of life in early old age: The theory, development and properties of a needs satisfaction model (CASP-19). Aging Ment. Health (2003.0) 7 186-194. DOI: 10.1080/1360786031000101157 28. Borrat-Besson C., Ryser V.A., Gonçalves J.. An Evaluation of the CASP-12 Scale Used in the Survey of Ageing and Retirement in Europe (SHARE) to Measure Quality of Life Among People Aged 50+ (2015.0) 29. Kerry M.J.. Bifactor model of the CASP-12’s general factor for measuring quality of life in older patients. J. Patient-Rep. Outcomes (2018.0) 2 57. DOI: 10.1186/s41687-018-0078-x 30. Matos A.D., Barbosa F., Cunha C., Voss G., Correia F.. Social isolation, physical inactivity and inadequate diet among European middle-aged and older adults. BMC Public Health (2021.0) 21. DOI: 10.1186/s12889-021-10956-w 31. Jakobsen J.C., Gluud C., Wetterslev J., Winkel P., Jakobsen J.C., Gluud C., Wetterslev J., Winkel P.. When and how should multiple imputation be used for handling missing data in randomised clinical trials—A practical guide with flowcharts. BMC Med. Res. Methodol. (2017.0) 17. DOI: 10.1186/s12874-017-0442-1 32. Rubini A., Vilaplana-Prieto C., Flor-Alemany M., Yeguas-Rosa L., Hernández-González M., Félix-García F.J., Félix-Redondo F.J., Fernández-Bergés D.. Assessment of the Cost of the Mediterranean Diet in a Low-Income Region: Adherence and Relationship with Available Incomes. BMC Public Health (2022.0) 22. DOI: 10.1186/s12889-021-12433-w 33. EuroVoc. Publications Office of the EU. Retrieved 29 Jan 2023—Via Europa 34. Nutritional Habits Statistics. (2023.0) 35. Mabiama G., Millimono T., Adiogo D., Boumediene F., Preux P.-M., Desport J.-C., Fayemendy P., Jésus P.. Undernutrition, overweight and obesity prevalences among community-dwelling elderly in Africa-a systematic review. Clin. Nutr. (2022.0) 45 42. DOI: 10.1016/j.nutos.2022.08.004 36. Volkert D., Beck A.M., Cederholm T., Cereda E., Cruz-Jentoft A., Goisser S., de Groot L., Großhauser F., Kiesswetter E., Norman K.. Management of Malnutrition in Older Patients—Current Approaches, Evidence and Open Questions. J. Clin. Med. (2019.0) 8. DOI: 10.3390/jcm8070974 37. Page M.J., McKenzie J.E., Bossuyt P.M., Boutron I., Hoffmann T.C., Mulrow C.D., Shamseer L., Tetzlaff J.M., Akl E.A., Brennan S.E.. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. PLoS Med. (2019.0) 18. DOI: 10.1371/journal.pmed.1003583 38. Kim T.J., Knesebeck O.V.D.. Income and obesity: What is the direction of the relationship? A systematic review and meta-analysis. BMJ Open (2018.0) 8 e019862 39. Talukdar D., Seenivasan S., Cameron A.J., Sacks G.. The association between national income and adult obesity prevalence: Empirical insights into temporal patterns and moderators of the association using 40 years of data across 147 countries. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0232236 40. Martínez-González M.A., Gea A., Ruiz-Canela M.. The Mediterranean Diet and Cardiovascular Health. Circ. Res. (2019.0) 124 779-798. DOI: 10.1161/CIRCRESAHA.118.313348 41. Buchanan A., Villani A.. Association of Adherence to a Mediterranean Diet with Excess Body Mass, Muscle Strength and Physical Performance in Overweight or Obese Adults with or without Type 2 Diabetes: Two Cross-Sectional Studies. Healthcare (2021.0) 9. DOI: 10.3390/healthcare9101255 42. Barrea L., Muscogiuri G., Di Somma C., Tramontano G., De Luca V., Illario M., Colao A., Savastano S.. Association between Mediterranean diet and hand grip strength in older adult women. Clin. Nutr. (2019.0) 38 721-729. DOI: 10.1016/j.clnu.2018.03.012 43. Shimizu A., Okada K., Tomata Y., Uno C., Kawase F., Momosaki R.. Association of Japanese and Mediterranean Dietary Patterns with Muscle Weakness in Japanese Community-Dwelling Middle-Aged and Older Adults:. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph191912636 44. Kim I., Son K., Jeong S.J., Lim H.. Sex and Diet-Related Disparities in Low Handgrip Strength among Young and Middle-Aged Koreans: Findings Based on the Korea National Health and Nutrition Examination Survey (KNHANES) from 2014 to 2017. Nutrients (2022.0) 14. DOI: 10.3390/nu14183816 45. Kljusurić J.G., Bosanac V., Šanko K., Barić I.C.. Establishing energy-nutritional variety of boarding school daily menus as a result of regional differences by use of multivariate analysis. J. Food Compos. Anal. (2016.0) 51 61-68. DOI: 10.1016/j.jfca.2016.06.010 46. Carballo-Casla A., Ortolá R., García-Esquinas E., Oliveira A., Sotos-Prieto M., Lopes C., Lopez-Garcia E., Rodríguez-Artalejo F.. The Southern European Atlantic Diet and all-cause mortality in older adults. BMC Med. (2021.0) 19. DOI: 10.1186/s12916-021-01911-y 47. Causes of Death Statistics. (2019.0) 48. Mazza E., Ferro Y., Pujia R., Mare R., Maurotti S., Montalcini T., Pujia A.. Mediterranean Diet In Healthy Aging. J. Nutr. Health Aging (2021.0) 25 1076-1083. DOI: 10.1007/s12603-021-1675-6 49. Guasch-Ferré M., Willett W.C.. The Mediterranean diet and health: A comprehensive overview. J. Intern. Med. (2021.0) 290 549-566. DOI: 10.1111/joim.13333 50. van den Brink A.C., Brouwer-Brolsma E.M., Berendsen A.A.M., Van De Rest O.. The Mediterranean, Dietary Approaches to Stop Hypertension (DASH), and Mediterranean-DASH Intervention for Neurodegenerative Delay (MIND) Diets Are Associated with Less Cognitive Decline and a Lower Risk of Alzheimer’s Disease—A Review. Adv. Nutr. (2019.0) 10 1040-1065. DOI: 10.1093/advances/nmz054
--- title: 'Muscle Cell Insulin Resistance Is Attenuated by Rosmarinic Acid: Elucidating the Mechanisms Involved' authors: - Danja J. Den Hartogh - Filip Vlavcheski - Evangelia Tsiani journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049470 doi: 10.3390/ijms24065094 license: CC BY 4.0 --- # Muscle Cell Insulin Resistance Is Attenuated by Rosmarinic Acid: Elucidating the Mechanisms Involved ## Abstract Obesity and elevated blood free fatty acid (FFA) levels lead to impaired insulin action causing insulin resistance in skeletal muscle, and contributing to the development of type 2 diabetes mellitus (T2DM). Mechanistically, insulin resistance is associated with increased serine phosphorylation of the insulin receptor substrate (IRS) mediated by serine/threonine kinases including mTOR and p70S6K. Evidence demonstrated that activation of the energy sensor AMP-activated protein kinase (AMPK) may be an attractive target to counteract insulin resistance. We reported previously that rosemary extract (RE) and the RE polyphenol carnosic acid (CA) activated AMPK and counteracted the FFA-induced insulin resistance in muscle cells. The effect of rosmarinic acid (RA), another polyphenolic constituent of RE, on FFA-induced muscle insulin resistance has never been examined and is the focus of the current study. Muscle cell (L6) exposure to FFA palmitate resulted in increased serine phosphorylation of IRS-1 and reduced insulin-mediated (i) Akt activation, (ii) GLUT4 glucose transporter translocation, and (iii) glucose uptake. Notably, RA treatment abolished these effects, and restored the insulin-stimulated glucose uptake. Palmitate treatment increased the phosphorylation/activation of mTOR and p70S6K, kinases known to be involved in insulin resistance and RA significantly reduced these effects. RA increased the phosphorylation of AMPK, even in the presence of palmitate. Our data indicate that RA has the potential to counteract the palmitate-induced insulin resistance in muscle cells, and further studies are required to explore its antidiabetic properties. ## 1. Introduction Blood glucose homeostasis is tightly regulated by insulin. After postprandial glucose levels are increased, pancreatic β cells respond by releasing insulin into the bloodstream, where it is delivered to its target tissues. Specifically, insulin promotes the transport, utilization, and storage of glucose in skeletal muscle and adipose tissue [1,2], while inhibiting the endogenous production of glucose by the liver. These insulin actions are key to maintaining plasma glucose levels within a physiological range of 4–7 millimolar (mM). Insulin initiates its action by binding to its receptor. This initiates tyrosine phosphorylation of the receptor and insulin receptor substrate (IRS-1), and activation of the lipid kinase phosphatidylinositol-3 kinase (PI3K) and the serine threonine kinase protein kinase B/Akt, resulting in increased translocation of the glucose transporter (GLUT4) from an intracellular pool to the plasma membrane and increased glucose uptake [3,4]. Impairments in the PI3K-Akt cascade can lead to the development of insulin resistance and type 2 diabetes mellitus (T2DM) [1,2,5]. Skeletal muscle accounts for roughly 70–$80\%$ of postprandial glucose uptake and is quantitatively the most important insulin target tissue. Therefore, skeletal muscle insulin resistance is a major contributor to decreased glucose tolerance and T2DM. Insulin resistance is strongly associated with obesity and increased plasma lipid levels. In vitro studies have shown that exposure of muscle cells to the free fatty acid (FFA) palmitate results in insulin resistance [6]. Additionally, evidence from in vivo animal studies demonstrated that lipid infusion [7,8] or increased plasma lipid levels due to a high-fat diet results in muscle insulin resistance [7,9]. Studies have shown that serine phosphorylation of IRS-1 (Ser$\frac{636}{639}$ and Ser307) results in an impaired insulin/PI3K/Akt signaling pathway and increased insulin resistance [6,10,11]. Signaling molecules including, mammalian target of rapamycin (mTOR) [12,13], ribosomal protein S6 kinase (p70S6K) [14,15], c-Jun N-terminal kinase (JNK) [16], and protein kinase C (PKCs) [17] act to increase the serine phosphorylation of IRS-1 [18]. The World Health Organization and the International Diabetes Federation (IDF) estimate that T2DM is a disease on the rise [19] and presents a huge economic burden on global health care systems. Different strategies to prevent and treat insulin resistance and T2DM do currently exist; however, they are lacking in efficacy and, therefore, there is a need for new preventative measures and targeted therapies. Adenosine monophosphate (AMP)-activated protein kinase (AMPK) is a serine/threonine kinase that acts as a cellular energy sensor. AMPK is activated by an increased AMP/ATP ratio and/or via phosphorylation of its α-catalytic subunit by its upstream kinases, liver kinase B1 (LKB1), calmodulin-dependent protein kinase (CaMKKs), and transforming growth factor-β (TGF-β)-activated kinase 1 (TAK1) [20,21]. AMPK activity is driven by muscle contraction/exercise [21] and stimulated by several compounds, including the currently used antidiabetic drugs metformin [22] and thiazolidineones [23], and by various polyphenols/flavanols found in tea [24], red wine, citrus fruit, and cocoa [25], including resveratrol [26], naringenin [27], and cocoa flavanol [28], resulting in increased skeletal muscle glucose uptake. The utilization of AMPK activators has gained increasing attention as a novel pharmacological intervention for the prevention/treatment of T2DM and insulin resistance [21,29,30,31]. Chemicals found in plants/herbs that activate AMPK have attracted attention as potential diabetes treatment options. Rosemary (Rosmarinus officinalis Lamiaceae) is an aromatic evergreen plant reported to have antioxidant [32,33], anticancer [18,20], and antidiabetic properties [34,35,36,37,38]. Rosemary extract (RE) contains different classes of polyphenols, including phenolic acids (rosmarinic acid; RA), flavonoids, and phenolic diterpenes (carnosic acid; CA and carnosol; COH) [39]. Previous studies by our group showed a significant increase in muscle glucose uptake and AMPK activation by RE [40], CA [41], RA [42], and COH [43] treatment. More importantly, treatment of L6 muscle cells with RE [44] and CA [45] attenuated the palmitate-induced insulin resistance. Limited evidence from in vitro and in vivo models of insulin resistance indicate that RA may potentially be used to improve insulin sensitivity [46,47,48]. However, the mechanistic effects of RA on palmitate-induced insulin-resistant muscle cells remain to be elucidated. In the present study, we examined the potential of RA to counteract palmitate-induced insulin resistance in muscle cells. ## 2.1. The Palmitate-Induced Serine Phosphorylation of Insulin Receptor Substrate-1 (IRS-1) Is Prevented by Rosmarinic Acid Treatment An increase in the serine phosphorylation of IRS-1 at residues Ser307 and Ser$\frac{636}{639}$ is linked to impaired PI3K/Akt signaling and insulin resistance, and for this reason we first examined the effect of RA on IRS-1. Exposure of L6 muscle cells to 0.2 mM palmitate for 16 h significantly increased IRS-1 phosphorylation at residues Ser307 and Ser$\frac{636}{639}$ (P: 134.7 ± $9.2\%$ and 140.1 ± $7.2\%$ of control, $p \leq 0.05$ and $p \leq 0.01$, respectively, Figure 1A,B). The palmitate-induced Ser307 and Ser$\frac{636}{639}$ phosphorylation of IRS-1 was abolished with RA treatment (RA + P: 58.1 ± $10.5\%$ and 105.0 ± $7.8\%$ of control, $p \leq 0.01$ and $p \leq 0.05$, respectively, Figure 1A,B). Treatment with RA alone reduced Ser307 phosphorylation of IRS-1 (RA: 60.8 ± $10.7\%$ of control, $p \leq 0.05$, Figure 1A,B), but had no effect on basal Ser$\frac{636}{639}$ phosphorylation of IRS-1 (RA: 98.3 ± $11.3\%$ of control, Figure 1A,B). Moreover, the total levels of IRS-1 were unaffected by any treatment (P: 108.5 ± $2.5\%$, RA: 99.7 ± $8.1\%$, RA + P: 129.0 ± $12.1\%$ of control, Figure 1A,B). ## 2.2. The Insulin-Stimulated Akt Phosphorylation in Palmitate-Treated Myotubes Is Restored with Rosmarinic Acid Akt phosphorylation/activation is a key step in the insulin signaling cascade, leading to increased glucose uptake by muscle cells, and is impaired in insulin resistance [49]. Therefore, we investigated the effect of RA on Akt. Treatment of L6 myotubes with 100 nM insulin for 30 min resulted in a significant increase in Akt Ser473 phosphorylation, an indicator of activation (I: 816.5 ± $109.87\%$ of control, $p \leq 0.01$, Figure 2A,B). Exposure of the cells to palmitate impaired the insulin-stimulated phosphorylation of Akt (P + I: 159.7 ± $49.4\%$ of control, $$p \leq 0.009$$, Figure 2A,B). However, in the presence of RA, insulin-stimulated Akt phosphorylation was restored (RA + P + I: 507.2 ± $67.17\%$ of control, $p \leq 0.01$, Figure 2A,B). Palmitate alone had no effect on basal Akt phosphorylation (P: 71.0 ± $5.04\%$ of control, Figure 2A,B). The total levels of Akt were not significantly changed by any of the treatments (I: 110.3 ± $31.5\%$, P: 106.9 ± $12.8\%$, P + I: 118.8 ± $29.6\%$, RA + P + I: 109.8 ± $33.4\%$ of control, Figure 2A,B). ## 2.3. Rosmarinic Acid Restores Insulin-Stimulated GLUT4 Translocation to Plasma Membrane in Palmitate-Treated Myotubes Skeletal muscle glucose uptake in response to insulin is driven by glucose transporter GLUT4 translocation from an intracellular pool to the plasma membrane and is mediated by upstream activation of the PI3K/Akt cascade. We examined the effects of our treatment on GLUT4 transporter translocation to the plasma membrane using L6 cells that overexpress an myc-labelled GLUT4 glucose transporter [50]. Acute stimulation of GLUT4myc overexpressing L6 myotubes with 100 nM insulin for 30 min resulted in a significant increase in GLUT4 plasma membrane levels (I: 193.0 ± $6.42\%$ of control, $p \leq 0.001$, Figure 3). Palmitate impaired the insulin-stimulated GLUT4 plasma membrane levels (P + I: 131.4 ± $5.48\%$ of control, Figure 3) while RA restored the insulin-stimulated GLUT4 plasma membrane levels (RA + P + I: 175.1 ± $9.26\%$ of control, $p \leq 0.01$, Figure 3). ## 2.4. Rosmarinic Acid Restores the Insulin-Stimulated Glucose Uptake in Palmitate-Treated Myotubes Next, we examined the effects of RA on muscle cell glucose uptake. Stimulation of L6 myotubes with 100 nM insulin for 30 min significantly increased glucose uptake (201 ± $1.21\%$ of control, $p \leq 0.0001$, Figure 4). Exposure of the cells to 0.2 mM palmitate for 16 h almost abolished the insulin-stimulated glucose uptake (P + I: 119 ± $13.2\%$ of control), indicating impaired insulin action. Most importantly, palmitate-treated cells exposed to 5 µM RA had significantly increased insulin-stimulated glucose uptake (RA + P + I: 184 ± $15.5\%$ of control $p \leq 0.001$, Figure 4). Treatment with RA in the presence of palmitate did not have a significant effect on basal glucose uptake (RA + P: 124 ± $5.8\%$ of control). ## 2.5. The Palmitate-Induced Phosphorylation/Activation of mTOR and p70S6K Is Prevented in the Presence of Rosmarinic Acid mTOR and p70S6K are kinases implicated in serine phosphorylation of IRS-1 and are specifically known to phosphorylate Ser307 and Ser$\frac{636}{639}$, and therefore, we examined the effects of palmitate on mTOR and p70S6K phosphorylation/activation and expression. Exposure of the cells to 0.2 mM palmitate for 16 h significantly increased mTOR Ser2448 and p70S6K Thr389 phosphorylation (P: 174.6 ± $15.6\%$ and 572.7 ± $57.8\%$ of control, respectively, $p \leq 0.01$, Figure 5A–D). Treatment with RA alone did not affect the basal mTOR (RA: 91.8 ± $8.3\%$ of control, Figure 5A–D) or p70S6K (RA: 203.9 ± $60.9\%$ of control, Figure 5A–D) phosphorylation levels. However, RA treatment significantly prevented the palmitate-induced phosphorylation of mTOR and p70S6K (RA + P: 105.8 ± $4.35\%$ and 247.2 ± $54.02\%$ of control, respectively, $p \leq 0.05$, Figure 5A–D). The total levels of mTOR and p70S6K were not significantly changed by any treatment. ## 2.6. Rosmarinic Acid Increases the Phosphorylation of AMPK, ACC, and Raptor Previous studies by our group showed that RE and RE polyphenols increased muscle cell glucose uptake via activation of AMPK [40,41,42,43]. Here, we investigated the effect of RA on AMPK and its downstream targets ACC and Raptor. The phosphorylation of AMPK at Thr172 was significantly increased in cells treated with 5 µM RA (RA: 252.8 ± $36.8\%$ of control, $p \leq 0.05$, Figure 6A,B). We also examined phosphorylation of ACC, a direct target of activated AMPK, which has been established as a marker of AMPK activation. RA increased the phosphorylation of ACC (RA: 170.6 ± $18.6\%$ of control, $p \leq 0.01$, Figure 6C,D). Most importantly, RA increased the phosphorylation of AMPK and ACC even in the presence of 0.2 mM palmitate (RA + P: 229.4 ± $34.3\%$ and 178.9 ± $25.3\%$ of control, $p \leq 0.05$ and $p \leq 0.01$, respectively, Figure 6A–D). Treatment with palmitate alone had no significant effect on phosphorylated AMPK and ACC levels (P: $87\%$ and $74\%$ of control, respectively, Figure 6A–D). Furthermore, the total levels of AMPK (P: 121 ± $38\%$, RA: 103 ± $22\%$, RA + P: 108 ± $24\%$ of control, Figure 6A,B), and ACC (P: 104 ± $14\%$, RA: 93 ± $8\%$, RA + P: 99 ± $12\%$ of control, Figure 6C,D) were not affected by any treatment. The activity of mTOR is influenced by the regulatory-associated protein of the mammalian target of rapamycin (Raptor), and the phosphorylation of Raptor on Ser792 results in inhibition of mTOR [51,52,53]. AMPK activation leads to the direct phosphorylation of Raptor. Treatment with 5 μM RA increased the phosphorylation of Raptor (RA: 153 ± $5.7\%$ of control, $p \leq 0.01$, Figure 6E,F). Most importantly, RA increased the phosphorylation of Raptor even in the presence of 0.2 mM palmitate (RA + P: 155 ± $7.9\%$ of control, $p \leq 0.05$, Figure 6E,F). Treatment with palmitate alone had no effect on the phosphorylation of Raptor (P: 102 ± $8.1\%$ of control). Furthermore, the total levels of Raptor were not affected by any treatment (P: 96 ± $4\%$, RA: 94 ± $2\%$, RA + P: 93 ± $4\%$ of control, Figure 6E,F). ## 2.7. AMPK Inhibition Reverses the Effects of Rosmarinic Acid on Palmitate-Induced Phosphorylation of Raptor, mTOR, and p70S6K In an attempt to elucidate the mechanism of action of RA, we performed additional experiments utilizing compound C (CC), a specific AMPK inhibitor. The phosphorylation of Raptor at Ser792 was significantly increased in cells treated with 5 µM RA in the presence of 0.2 mM palmitate (RA + P: 193.63 ± $23.7\%$ of control, $p \leq 0.05$, Figure 7A,B), and importantly, pretreatment of the cells with CC abolished this response (RA + P + CC: 87.2 ± $16.04\%$ control, $p \leq 0.05$, Figure 7A,B). Furthermore, in the presence of CC, the effect of RA on suppressing the palmitate-induced mTOR and p70S6Kphosphorylation/activation (P: 244.2 ± $23.3\%$ and 259.3 ± $34.8\%$ of control, $p \leq 0.001$ and $p \leq 0.05$, respectively, Figure 7C–F), (RA + P: 121.6 ± $12.8\%$ and 114.1 ± $14.7\%$ of control, $p \leq 0.01$ and $p \leq 0.05$, respectively, Figure 7C–F) was abolished (RA + P + CC: 241.5 ± $30.1\%$ and 272.2 ± $14.9\%$ control, $p \leq 0.05$ and $p \leq 0.01$, respectively, Figure 7C–F). ## 3. Discussion Obesity and elevated FFA levels are strong indicators of insulin resistance and are major risk factors for T2DM [18], a disease affecting millions of people globally. Current drugs used to treat insulin resistance and T2DM manifest negative side effects, driving the search for natural, plant-derived compounds with the potential to counteract insulin resistance. In the current study, we examined the potential of the plant-derived compound RA to counteract the palmitate-induced insulin resistance in muscle cells. Exposure of L6 muscle cells to palmitate, a saturated FFA, to simulate the in vivo scenario of elevated FFA plasma levels often seen in obesity, dramatically reduced the plasma membrane levels of GLUT4 and insulin-stimulated glucose uptake, indicating insulin resistance (Figure 7). These findings are in agreement with previous data from our lab [44,45,54] and others [6,47,55,56]. Importantly, in the presence of RA, the insulin-stimulated plasma membrane GLUT4 levels and glucose uptake were restored to levels comparable to those achieved with insulin stimulation alone. Only one other study examined the effects of RA on L6 muscle cells. Jayanthy et al. found that treatment of L6 muscle cells with RA (20 µM) and palmitate (0.3 mM) for 24 h increased GLUT4 plasma membrane levels and glucose uptake [47]. Unfortunately, the study by Jayanthy et al. did not examine the effects of RA on insulin-induced responses. Although, in our study, we found that RA restored insulin responsiveness, we did not find any significant effect of RA on basal GLUT4 and glucose uptake. The differences between our study and that by Jayanthy et al. may be due to the different RA (5 vs. 20 µM) and/or palmitate (0.2 vs. 0.3 mM) concentrations used. Additionally, our study found that exposure of L6 muscle cells to palmitate markedly reduced the insulin-stimulated Akt phosphorylation and is in agreement with other in vitro studies using L6 [57] and C2C12 [58] muscle cells as well as in vivo studies showing reduced levels of Akt phosphorylation in soleus muscle harvested from mice fed a high fat diet (HFD) [59]. Remarkably, treatment with RA restored the insulin-stimulated phosphorylation of Akt, indicating that RA, similarly to metformin [60], counteracts the harmful effects of palmitate. In addition, our study found that treatment of L6 cells with palmitate increased the serine phosphorylation of IRS-1. This finding is in agreement with other studies indicating increased Ser307 and Ser$\frac{636}{639}$ phosphorylation of IRS-1 in the presence of palmitate in L6 [56] and C2C12 [61] cells. Increased phosphorylation of the serine residues of IRS-1 results in reduced downstream PI3K-Akt signaling and glucose uptake [62,63]. Importantly, our studies show that RA blocked the palmitate-induced serine phosphorylation of IRS-1, an effect that is similar to that of metformin [64]. These data are in agreement with the study by Jayanthy et al. that showed a reduction in the palmitate-induced Ser307 phosphorylation of IRS-1 via RA treatment of L6 muscle cells [47]. Furthermore, exposure of L6 myotubes to palmitate considerably augmented the phosphorylation of mTOR and its downstream effector p70S6K, and most importantly, treatment with RA attenuated these effects of palmitate (Figure 8). While it has already been established by other studies that palmitate treatment results in increased mTOR phosphorylation in L6 [65] and C2C12 cells [66], and in muscle tissue obtained from animals fed an HFD [65,67], this study is the first to report that RA has the potential to block these effects and act in a similar fashion as the mTOR inhibitor rapamycin [68] and metformin [69]. Importantly, studies have shown that the Ser307 and Ser$\frac{636}{639}$ phosphorylation of IRS-1 is mediated by mTOR and its downstream target p706SK to reduce PI3K/Akt signaling and glucose uptake [62,70]. The role of mTOR/p706SK signaling has been confirmed by many other groups as a critical mechanism involved in the induction of insulin resistance in insulin-sensitive tissues (muscle, fat and liver) [71,72,73,74]. Several studies have shown that activated mTOR causes phosphorylation of the growth factor receptor-binding protein 10 (Grb10) at Ser476, which in turn binds to the phosphorylated tyrosine residues of the insulin receptor and inhibits its tyrosine kinase activity, resulting in reduced downstream PI3K-Akt signaling [75,75,76]. Additionally, co-immunoprecipitation studies revealed that Grb10 was found to bind to the regulatory p85 subunit of PI3K, indicating that Grb10 directly associates with PI3K and reduces the PI3K catalytic activity, resulting in impaired insulin action in L6 myotubes [75]. Furthermore, overexpression of Grb10 inhibits the interaction of the insulin receptor with PI3K, thus reducing insulin signaling and causing insulin resistance [76,77]. Although we did not examine the effects of RA on Grb10, it is possible that RA treatment reduces Grb10 levels. Furthermore, our present study shows that RA markedly increased the phosphorylation/activation of AMPK even in the presence of palmitate, indicating that the effects of RA are similar to those of metformin, which also activates AMPK in the presence of palmitate in L6 and C2C12 myotubes [60,69]. Activation of AMPK directly inhibits mTOR activity by phosphorylating Raptor at Ser$\frac{722}{792}$ [78,79]. Our study indicates an activation/phosphorylation of Raptor with RA treatment, and that the RA-induced activation of AMPK may be the reason for the inhibition of mTOR and its downstream target p706SK (Figure 8). Indeed, pretreatment of the cells with the specific AMPK inhibitor compound C (CC) abolished the effects of RA on palmitate-induced responses. These data indicate that activation of AMPK plays a major role in the RA-induced effects. A small number of studies have investigated the antidiabetic effects of RA in vivo. In HFD-induced diabetic rats, the intraperitoneal administration of RA (200 mg/kg/day for 28 days) dose-dependently ameliorated hyperglycemia and increased insulin sensitivity assessed by the oral glucose tolerance test (OGTT). These effects were associated with reduced hepatic PEPCK protein expression and increased skeletal muscle GLUT4 protein levels [46]. Additionally, treatment of STZ-induced diabetic rats with RA resulted in amelioration of hyperglycemia [46]. Another study found that RA administration through oral gavage (100 mg/kg/day for 30 days) improved glucose homeostasis and significantly increased AMPK phosphorylation/activation and mitochondrial biogenesis/activity in the skeletal muscles of STZ-HFD-induced insulin-resistant rats [47]. These studies demonstrate that RA exhibits anti-hyperglycemic and antidiabetic properties in vivo. However, there are currently no studies that elucidate the mechanisms involved in the effects of RA. The present study found that RA prevented the palmitate-induced phosphorylation/activation of mTOR and p70S6K and restored insulin-stimulated Akt phosphorylation, GLUT4 glucose transporter translocation to the plasma membrane, and glucose uptake (Figure 8). ## 4.1. Materials The following materials were purchased from Sigma Life Sciences (St. Louis, MO, USA): bovine serum albumin (BSA), compound C (CC), cytochalasin B, dimethyl sulfoxide (DMSO), fetal bovine serum (FBS), palmitic acid, and rosmarinic acid. Trypan blue solution $0.4\%$ and material necessary for cell culture were purchased from GIBCO Life Technologies (Burlington, ON, Canada). [ 3H]-2-deoxy-D-glucose was purchased from PerkinElmer (Boston, MA, USA). Antibodies including phospho and total ACC (CAT 3661 and 3662, respectively), Akt (CAT 9271 and 9272, respectively), AMPK (CAT 2531 and 2532, respectively), HRP-conjugated anti-rabbit (CAT 7074), IRS-1 (CAT 2381, 2388, and 2382, respectively), mTOR (CAT 2971 and 2972, respectively), and p70S6K (CAT 9205 and 9202, respectively) were purchased from New England BioLabs (NEB) (Missisauga, ON, Canada). Insulin (Humulin R) was obtained from Eli Lilly (Indianapolis, IN, USA). Materials for Western blotting and Bradford protein assay reagent were purchased from BioRad (Hercules, CA, USA). ## 4.2. Preparation of Palmitate Stock Solution Stock palmitate solution was prepared by conjugating palmitic acid with fatty-acid-free BSA as previously reported by us [44,45,54] and others [6]. In short, palmitic acid was dissolved in 0.1 N NaOH and diluted in (45–50 °C) prewarmed $9.7\%$ (w/v) BSA solution. A stock solution of 8 mM palmitate with a final molar ratio of free palmitate/BSA of 6:1 was prepared and kept at −80 °C. ## 4.3. Cell Culture, Treatment, and Glucose Uptake All experiments utilized L6 rat skeletal muscle cells. Undifferentiated myoblasts were grown in α-Minimum Essential Medium (MEM) media containing $10\%$ v/v FBS until $80\%$ confluency and differentiated into myotubes in α-MEM media containing $2\%$ v/v FBS, as previously established [5,26]. Fully differentiated myotubes were achieved approximately 6 to 7 days after seeding. All treatments were performed using serum-free α-MEM media containing $0\%$ v/v FBS, followed by exposure to palmitate (0.2 mM) in the absence or presence of RA (5 µM) for 16 h followed by treatment without or with insulin (100 nM) for 0.5 h. Following treatment, the cells were rinsed with HEPES-buffered saline (HBS) and exposed to [3H]-2-deoxy-D-glucose (10 µM) for 10 min to measure cellular glucose uptake, as previously described [26,80]. Non-specific glucose uptake was measured in the presence of cytochalasin B (10 µM) and was subtracted from the total to obtain the specific glucose uptake. At the end of the glucose uptake assay, the cells were rinsed with $0.9\%$ NaCl and lysed using a 0.05 N NaOH solution. A liquid scintillation counter was used to measure the radioactivity, and the Bradford assay was used to determine the protein content. ## 4.4. GLUT4myc Translocation Assay Fully differentiated L6 GLUT4myc-overexpressing myotubes, grown in 24-well plates, were treated and fixed using $3\%$ paraformaldehyde dissolved in PBS for 10 min at 4 °C. The fixed cells were then rinsed and incubated with PBS containing $1\%$ glycine for 10 min, followed by blocking with $10\%$ goat serum containing PBS for 15 min. The cells were then exposed to blocking buffer containing primary anti-myc antibody (1 h, 1:500), followed by incubation with blocking buffer containing HRP-conjugated donkey anti-mouse antibodies (45 min, 1:1000) at 4 °C. The cells were washed extensively with PBS and were incubated with O-phenylenediamine dihydrochloride (OPD) reagent at room temperature and protected from light for 30 min. The reaction was stopped using 3 N HCL solution, and the supernatant was collected and visualized at 492 nm. The OPD reagent is a substrate for HRP and produces a yellow product that can be visualized using an absorbance plate reader (Synergy HT, BioTek Instruments, Winooski, VT, USA). The intensity of the color produced is proportionate to the levels of GLUT4myc detected in the plasma membrane. ## 4.5. Immunoprecipitation Whole-cell lysates (150 µg) were incubated with IRS1 antibody (1:50 volume ratio) conjugated to SureBeadsTM Protein G Magnetic beads (Biorad; Hercules, CA, USA) for 1 h at room temperature. The lysates–IRS-1–beads complex was collected by microcentrifugation and washed three times with PBS + $0.1\%$ Tween-20. Protein was eluted with glycine (20 mM, pH 2.0) solution for 5 min at room temperature and neutralized with PBS (1 M, pH 7.4) at $10\%$ eluent volume. A 3× sodium dodecyl sulfate (SDS) sample buffer was added to eluted protein and boiled for 5 min. ## 4.6. Immunoblotting After treatment, L6 myotubes were rinsed twice with pre-chilled (4 °C) PBS solution and the cells were lysed using ice-cold lysis buffer containing ethylene glycol-bis β-aminoethyl ether/egtazic acid (EGTA), 1 mM ethylenediaminetetraacetic acid (EDTA), 1mM sodium orthovanadate (Na3VO4), 1 mM p-glycerolphosphate, 20 mM Tris (pH 7.5), $1\%$ Triton X-100, 1 mM, 150 mM NaCI, 1 µg/mL leupeptin, 2.5 mM sodium pyrophosphate and 1 mM phenylmethylsulfonyl fluoride (PMSF) and were stored at −20 °C. A $5\%$ B-mercaptoethanol containing 3x SDS buffer was added and the samples were boiled for 5 min. The proteins were separated using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a polyvinylidene fluoride (PVDF) membrane followed by blocking with Tris-buffered saline containing $5\%$ (w/v) dry milk powder and incubation with the primary antibody overnight at (4 °C). To detect the primary antibody HRP-conjugated anti-rabbit secondary antibodies were used followed by exposure to LumGLOW reagent. The blots were visualized using ChemiDoc, imaging system (BioRad, Hercules, CA, USA). ## 4.7. Statistical Analysis Statistical analysis was completed using GraphPad Prism software 5.3 manufactured by Graphpad Software Inc. (La Jolla, CA, USA). The data from several experiments were pooled and presented as mean ± standard error (SE). The means of all the groups were obtained and compared to the control group using one-way analysis of variance (ANOVA), which was followed by Tukey’s post hoc test for multiple comparisons. ## 5. Conclusions The prevalence of T2DM is constantly increasing, and according to the International Diabetes Federation, it is expected to affect 420 million people worldwide by the year 2040 [19]. In addition, insulin resistance and T2DM are highly correlated with the development of other pathological states, including cardiovascular disease and cancer [18]. As a result, new strategies to aid in the prevention and management of T2DM will provide huge benefits to our society. As previously indicated, increased levels of FFA and obesity mediates insulin resistance in muscle cells. The present study has shown that the exposure of muscle cells to the FFA palmitate, as a way to mimic the elevated FFA levels seen in obesity, induced insulin resistance. Palmitate exposure to L6 muscle cells increased the phosphorylation of mTOR and p70S6K, while insulin-stimulated Akt phosphorylation and the insulin-stimulated glucose uptake and GLUT4 translocation were significantly reduced. Importantly, these effects of palmitate were attenuated by RA treatment, and insulin-stimulated glucose uptake was restored. In addition, RA increased the phosphorylation/activation of the energy sensor AMPK, an attractive target to counteract insulin resistance and T2DM. Our study is the first to show that RA has the potential to counteract palmitate-induced muscle cell insulin resistance, and further studies are required to explore its antidiabetic properties and to elucidate the exact cellular mechanisms involved. ## References 1. Kahn B.B., Flier J.S.. Obesity and Insulin Resistance. J. Clin. Investig. (2000) 106 473-481. DOI: 10.1172/JCI10842 2. Guo S.. Molecular Basis of Insulin Resistance: The Role of IRS and Foxo1 in the Control of Diabetes Mellitus and Its Complications. Drug Discov. Today Dis. Mech. (2013) 10 e27-e33. DOI: 10.1016/j.ddmec.2013.06.003 3. Taniguchi C.M., Emanuelli B., Kahn C.R.. Critical Nodes in Signalling Pathways: Insights into Insulin Action. Nat. Rev. Mol. Cell Biol. (2006) 7 85-96. DOI: 10.1038/nrm1837 4. Manning B.D., Cantley L.C.. AKT/PKB Signaling: Navigating Downstream. Cell (2007) 129 1261-1274. DOI: 10.1016/j.cell.2007.06.009 5. Tripathy D., Chavez A.O.. Defects in Insulin Secretion and Action in the Pathogenesis of Type 2 Diabetes Mellitus. Curr. Diab. Rep. (2010) 10 184-191. DOI: 10.1007/s11892-010-0115-5 6. Sinha S., Perdomo G., Brown N.F., O’Doherty R.M.. Fatty Acid-Induced Insulin Resistance in L6 Myotubes Is Prevented by Inhibition of Activation and Nuclear Localization of Nuclear Factor Kappa B. J. Biol. Chem. (2004) 279 41294-41301. DOI: 10.1074/jbc.M406514200 7. Samuel V.T., Petersen K.F., Shulman G.I.. Lipid-Induced Insulin Resistance: Unravelling the Mechanism. Lancet (2010) 375 2267-2277. DOI: 10.1016/S0140-6736(10)60408-4 8. Pereira S., Park E., Moore J., Faubert B., Breen D.M., Oprescu A.I., Nahle A., Kwan D., Giacca A., Tsiani E.. Resveratrol Prevents Insulin Resistance Caused by Short-Term Elevation of Free Fatty Acids In Vivo. Appl. Physiol. Nutr. Metab. (2015) 40 1129-1136. DOI: 10.1139/apnm-2015-0075 9. Hancock C.R., Han D.-H., Chen M., Terada S., Yasuda T., Wright D.C., Holloszy J.O.. High-Fat Diets Cause Insulin Resistance despite an Increase in Muscle Mitochondria. Proc. Natl. Acad. Sci. USA (2008) 105 7815-7820. DOI: 10.1073/pnas.0802057105 10. Feinstein R., Kanety H., Papa M.Z., Lunenfeld B., Karasik A.. Tumor Necrosis Factor-Alpha Suppresses Insulin-Induced Tyrosine Phosphorylation of Insulin Receptor and Its Substrates. J. Biol. Chem. (1993) 268 26055-26058. DOI: 10.1016/S0021-9258(19)74276-8 11. Ueno M., Carvalheira J.B.C., Tambascia R.C., Bezerra R.M.N., Amaral M.E., Carneiro E.M., Folli F., Franchini K.G., Saad M.J.A.. Regulation of Insulin Signalling by Hyperinsulinaemia: Role of IRS-1/2 Serine Phosphorylation and the MTOR/P70 S6K Pathway. Diabetologia (2005) 48 506-518. DOI: 10.1007/s00125-004-1662-6 12. Carlson C.J., White M.F., Rondinone C.M.. Mammalian Target of Rapamycin Regulates IRS-1 Serine 307 Phosphorylation. Biochem. Biophys. Res. Commun. (2004) 316 533-539. DOI: 10.1016/j.bbrc.2004.02.082 13. Mordier S., Iynedjian P.B.. Activation of Mammalian Target of Rapamycin Complex 1 and Insulin Resistance Induced by Palmitate in Hepatocytes. Biochem. Biophys. Res. Commun. (2007) 362 206-211. DOI: 10.1016/j.bbrc.2007.08.004 14. Manning B.D.. Balancing Akt with S6K: Implications for Both Metabolic Diseases and Tumorigenesis. J. Cell Biol. (2004) 167 399-403. DOI: 10.1083/jcb.200408161 15. Um S.H., Frigerio F., Watanabe M., Picard F., Joaquin M., Sticker M., Fumagalli S., Allegrini P.R., Kozma S.C., Auwerx J.. Absence of S6K1 Protects against Age- and Diet-Induced Obesity While Enhancing Insulin Sensitivity. Nature (2004) 431 200-205. DOI: 10.1038/nature02866 16. Hirosumi J., Tuncman G., Chang L., Görgün C.Z., Uysal K.T., Maeda K., Karin M., Hotamisligil G.S.. A Central Role for JNK in Obesity and Insulin Resistance. Nature (2002) 420 333-336. DOI: 10.1038/nature01137 17. Li Y., Soos T.J., Li X., Wu J., Degennaro M., Sun X., Littman D.R., Birnbaum M.J., Polakiewicz R.D.. Protein Kinase C Theta Inhibits Insulin Signaling by Phosphorylating IRS1 at Ser(1101). J. Biol. Chem. (2004) 279 45304-45307. DOI: 10.1074/jbc.C400186200 18. Hulver M.W., Dohm G.L.. The Molecular Mechanism Linking Muscle Fat Accumulation to Insulin Resistance. Proc. Nutr. Soc. (2004) 63 375-380. DOI: 10.1079/PNS2004351 19. Cho N.H., Shaw J.E., Karuranga S., Huang Y., da Rocha Fernandes J.D., Ohlrogge A.W., Malanda B.. IDF Diabetes Atlas: Global Estimates of Diabetes Prevalence for 2017 and Projections for 2045. Diabetes Res. Clin. Pract. (2018) 138 271-281. DOI: 10.1016/j.diabres.2018.02.023 20. Xie M., Zhang D., Dyck J.R.B., Li Y., Zhang H., Morishima M., Mann D.L., Taffet G.E., Baldini A., Khoury D.S.. A Pivotal Role for Endogenous TGF-Beta-Activated Kinase-1 in the LKB1/AMP-Activated Protein Kinase Energy-Sensor Pathway. Proc. Natl. Acad. Sci. USA (2006) 103 17378-17383. DOI: 10.1073/pnas.0604708103 21. Towler M.C., Hardie D.G.. AMP-Activated Protein Kinase in Metabolic Control and Insulin Signaling. Circ. Res. (2007) 100 328-341. DOI: 10.1161/01.RES.0000256090.42690.05 22. Zhou G., Myers R., Li Y., Chen Y., Shen X., Fenyk-Melody J., Wu M., Ventre J., Doebber T., Fujii N.. Role of AMP-Activated Protein Kinase in Mechanism of Metformin Action. J. Clin. Investig. (2001) 108 1167-1174. DOI: 10.1172/JCI13505 23. Fryer L.G.D., Parbu-Patel A., Carling D.. The Anti-Diabetic Drugs Rosiglitazone and Metformin Stimulate AMP-Activated Protein Kinase through Distinct Signaling Pathways. J. Biol. Chem. (2002) 277 25226-25232. DOI: 10.1074/jbc.M202489200 24. Chen R., Lai X., Xiang L., Li Q., Sun L., Lai Z., Li Z., Zhang W., Wen S., Cao J.. Aged Green Tea Reduces High-Fat Diet-Induced Fat Accumulation and Inflammation via Activating the AMP-Activated Protein Kinase Signaling Pathway. Food Nutr. Res. (2022) 66 7923. DOI: 10.29219/fnr.v66.7923 25. Martin M.A., Goya L., Ramos S.. Protective Effects of Tea, Red Wine and Cocoa in Diabetes. Evidences from Human Studies. Food Chem. Toxicol. (2017) 109 302-314. DOI: 10.1016/j.fct.2017.09.015 26. Breen D.M., Sanli T., Giacca A., Tsiani E.. Stimulation of Muscle Cell Glucose Uptake by Resveratrol through Sirtuins and AMPK. Biochem. Biophys. Res. Commun. (2008) 374 117-122. DOI: 10.1016/j.bbrc.2008.06.104 27. Zygmunt K., Faubert B., MacNeil J., Tsiani E.. Naringenin, a Citrus Flavonoid, Increases Muscle Cell Glucose Uptake via AMPK. Biochem. Biophys. Res. Commun. (2010) 398 178-183. DOI: 10.1016/j.bbrc.2010.06.048 28. Martin M.Á., Goya L., Ramos S.. Antidiabetic Actions of Cocoa Flavanols. Mol. Nutr. Food Res. (2016) 60 1756-1769. DOI: 10.1002/mnfr.201500961 29. Hardie D.G.. AMP-Activated Protein Kinase—An Energy Sensor That Regulates All Aspects of Cell Function. Genes Dev. (2011) 25 1895-1908. DOI: 10.1101/gad.17420111 30. Hardie D.G., Ross F.A., Hawley S.A.. AMPK: A Nutrient and Energy Sensor That Maintains Energy Homeostasis. Nat. Rev. Mol. Cell Biol. (2012) 13 251-262. DOI: 10.1038/nrm3311 31. Gasparrini M., Giampieri F., Alvarez Suarez J., Mazzoni L., Y Forbes Hernandez T., Quiles J.L., Bullon P., Battino M.. AMPK as a New Attractive Therapeutic Target for Disease Prevention: The Role of Dietary Compounds AMPK and Disease Prevention. Curr. Drug Targets (2016) 17 865-889. DOI: 10.2174/1573399811666150615150235 32. Cheung S., Tai J.. Anti-Proliferative and Antioxidant Properties of Rosemary Rosmarinus Officinalis. Oncol. Rep. (2007) 17 1525-1531. DOI: 10.3892/or.17.6.1525 33. Moore J., Yousef M., Tsiani E.. Anticancer Effects of Rosemary (. Nutrients (2016) 8. DOI: 10.3390/nu8110731 34. Bakirel T., Bakirel U., Keleş O.U., Ulgen S.G., Yardibi H.. In Vivo Assessment of Antidiabetic and Antioxidant Activities of Rosemary (Rosmarinus Officinalis) in Alloxan-Diabetic Rabbits. J. Ethnopharmacol. (2008) 116 64-73. DOI: 10.1016/j.jep.2007.10.039 35. Emam M.. Comparative Evaluation of Antidiabetic Activity of. ABJNA (2012) 3 247-252. DOI: 10.5251/abjna.2012.3.6.247.252 36. Ramadan K.S., Khalil O.A., Danial E.N., Alnahdi H.S., Ayaz N.O.. Hypoglycemic and Hepatoprotective Activity of Rosmarinus Officinalis Extract in Diabetic Rats. J. Physiol. Biochem. (2013) 69 779-783. DOI: 10.1007/s13105-013-0253-8 37. Romo Vaquero M., Yáñez-Gascón M.-J., García Villalba R., Larrosa M., Fromentin E., Ibarra A., Roller M., Tomás-Barberán F., Espín de Gea J.C., García-Conesa M.-T.. Inhibition of Gastric Lipase as a Mechanism for Body Weight and Plasma Lipids Reduction in Zucker Rats Fed a Rosemary Extract Rich in Carnosic Acid. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0039773 38. Naimi M., Vlavcheski F., Shamshoum H., Tsiani E.. Rosemary Extract as a Potential Anti-Hyperglycemic Agent: Current Evidence and Future Perspectives. Nutrients (2017) 9. DOI: 10.3390/nu9090968 39. Mena P., Cirlini M., Tassotti M., Herrlinger K.A., Dall’Asta C., Del Rio D.. Phytochemical Profiling of Flavonoids, Phenolic Acids, Terpenoids, and Volatile Fraction of a Rosemary (. Molecules (2016) 21. DOI: 10.3390/molecules21111576 40. Naimi M., Tsakiridis T., Stamatatos T.C., Alexandropoulos D.I., Tsiani E.. Increased Skeletal Muscle Glucose Uptake by Rosemary Extract through AMPK Activation. Appl. Physiol. Nutr. Metab. (2015) 40 407-413. DOI: 10.1139/apnm-2014-0430 41. Naimi M., Vlavcheski F., Murphy B., Hudlicky T., Tsiani E.. Carnosic Acid as a Component of Rosemary Extract Stimulates Skeletal Muscle Cell Glucose Uptake via AMPK Activation. Clin. Exp. Pharmacol. Physiol. (2017) 44 94-102. DOI: 10.1111/1440-1681.12674 42. Vlavcheski F., Naimi M., Murphy B., Hudlicky T., Tsiani E.. Rosmarinic Acid, a Rosemary Extract Polyphenol, Increases Skeletal Muscle Cell Glucose Uptake and Activates AMPK. Molecules (2017) 22. DOI: 10.3390/molecules22101669 43. Vlavcheski F., Baron D., Vlachogiannis I.A., MacPherson R.E.K., Tsiani E.. Carnosol Increases Skeletal Muscle Cell Glucose Uptake via AMPK-Dependent GLUT4 Glucose Transporter Translocation. Int. J. Mol. Sci. (2018) 19. DOI: 10.3390/ijms19051321 44. Vlavcheski F., Tsiani E.. Attenuation of Free Fatty Acid-Induced Muscle Insulin Resistance by Rosemary Extract. Nutrients (2018) 10. DOI: 10.3390/nu10111623 45. Den Hartogh D.J., Vlavcheski F., Giacca A., MacPherson R.E.K., Tsiani E.. Carnosic Acid Attenuates the Free Fatty Acid-Induced Insulin Resistance in Muscle Cells and Adipocytes. Cells (2022) 11. DOI: 10.3390/cells11010167 46. Runtuwene J., Cheng K.-C., Asakawa A., Amitani H., Amitani M., Morinaga A., Takimoto Y., Kairupan B.H.R., Inui A.. Rosmarinic Acid Ameliorates Hyperglycemia and Insulin Sensitivity in Diabetic Rats, Potentially by Modulating the Expression of PEPCK and GLUT4. Drug Des. Dev. Ther. (2016) 10 2193-2202. DOI: 10.2147/DDDT.S108539 47. Jayanthy G., Roshana Devi V., Ilango K., Subramanian S.P.. Rosmarinic Acid Mediates Mitochondrial Biogenesis in Insulin Resistant Skeletal Muscle through Activation of AMPK. J. Cell. Biochem. (2017) 118 1839-1848. DOI: 10.1002/jcb.25869 48. Bao T.-Q., Li Y., Qu C., Zheng Z.-G., Yang H., Li P.. Antidiabetic Effects and Mechanisms of Rosemary (. Am. J. Chin. Med. (2020) 48 1353-1368. DOI: 10.1142/S0192415X20500664 49. Petersen M.C., Shulman G.I.. Mechanisms of Insulin Action and Insulin Resistance. Physiol. Rev. (2018) 98 2133-2223. DOI: 10.1152/physrev.00063.2017 50. Randhawa V.K., Bilan P.J., Khayat Z.A., Daneman N., Liu Z., Ramlal T., Volchuk A., Peng X.R., Coppola T., Regazzi R.. VAMP2, but Not VAMP3/Cellubrevin, Mediates Insulin-Dependent Incorporation of GLUT4 into the Plasma Membrane of L6 Myoblasts. Mol. Biol. Cell (2000) 11 2403-2417. DOI: 10.1091/mbc.11.7.2403 51. Gwinn D.M., Shackelford D.B., Egan D.F., Mihaylova M.M., Mery A., Vasquez D.S., Turk B.E., Shaw R.J.. AMPK Phosphorylation of Raptor Mediates a Metabolic Checkpoint. Mol. Cell (2008) 30 214-226. DOI: 10.1016/j.molcel.2008.03.003 52. Rosner M., Siegel N., Valli A., Fuchs C., Hengstschläger M.. MTOR Phosphorylated at S2448 Binds to Raptor and Rictor. Amino Acids (2010) 38 223-228. DOI: 10.1007/s00726-008-0230-7 53. Watanabe R., Wei L., Huang J.. MTOR Signaling, Function, Novel Inhibitors, and Therapeutic Targets. J. Nucl. Med. (2011) 52 497-500. DOI: 10.2967/jnumed.111.089623 54. Den Hartogh D.J., Vlavcheski F., Giacca A., Tsiani E.. Attenuation of Free Fatty Acid (FFA)-Induced Skeletal Muscle Cell Insulin Resistance by Resveratrol Is Linked to Activation of AMPK and Inhibition of MTOR and P70 S6K. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21144900 55. Perdomo G., Commerford S.R., Richard A.-M.T., Adams S.H., Corkey B.E., O’Doherty R.M., Brown N.F.. Increased Beta-Oxidation in Muscle Cells Enhances Insulin-Stimulated Glucose Metabolism and Protects against Fatty Acid-Induced Insulin Resistance Despite Intramyocellular Lipid Accumulation. J. Biol. Chem. (2004) 279 27177-27186. DOI: 10.1074/jbc.M403566200 56. Dimopoulos N., Watson M., Sakamoto K., Hundal H.S.. Differential Effects of Palmitate and Palmitoleate on Insulin Action and Glucose Utilization in Rat L6 Skeletal Muscle Cells. Biochem. J. (2006) 399 473-481. DOI: 10.1042/BJ20060244 57. Powell D.J., Turban S., Gray A., Hajduch E., Hundal H.S.. Intracellular Ceramide Synthesis and Protein Kinase Czeta Activation Play an Essential Role in Palmitate-Induced Insulin Resistance in Rat L6 Skeletal Muscle Cells. Biochem. J. (2004) 382 619-629. DOI: 10.1042/BJ20040139 58. Capel F., Cheraiti N., Acquaviva C., Hénique C., Bertrand-Michel J., Vianey-Saban C., Prip-Buus C., Morio B.. Oleate Dose-Dependently Regulates Palmitate Metabolism and Insulin Signaling in C2C12 Myotubes. Biochim. Biophys. Acta (2016) 1861 2000-2010. DOI: 10.1016/j.bbalip.2016.10.002 59. Jung T.W., Kim H.-C., Abd El-Aty A.M., Jeong J.H.. Protectin DX Ameliorates Palmitate- or High-Fat Diet-Induced Insulin Resistance and Inflammation through an AMPK-PPARα-Dependent Pathway in Mice. Sci. Rep. (2017) 7 1397. DOI: 10.1038/s41598-017-01603-9 60. Wu W., Tang S., Shi J., Yin W., Cao S., Bu R., Zhu D., Bi Y.. Metformin Attenuates Palmitic Acid-Induced Insulin Resistance in L6 Cells through the AMP-Activated Protein Kinase/Sterol Regulatory Element-Binding Protein-1c Pathway. Int. J. Mol. Med. (2015) 35 1734-1740. DOI: 10.3892/ijmm.2015.2187 61. Deng Y.-T., Chang T.-W., Lee M.-S., Lin J.-K.. Suppression of Free Fatty Acid-Induced Insulin Resistance by Phytopolyphenols in C2C12 Mouse Skeletal Muscle Cells. J. Agric. Food Chem. (2012) 60 1059-1066. DOI: 10.1021/jf204496f 62. Gual P., Le Marchand-Brustel Y., Tanti J.-F.. Positive and Negative Regulation of Insulin Signaling through IRS-1 Phosphorylation. Biochimie (2005) 87 99-109. DOI: 10.1016/j.biochi.2004.10.019 63. Copps K.D., White M.F.. Regulation of Insulin Sensitivity by Serine/Threonine Phosphorylation of Insulin Receptor Substrate Proteins IRS1 and IRS2. Diabetologia (2012) 55 2565-2582. DOI: 10.1007/s00125-012-2644-8 64. Bogachus L.D., Turcotte L.P.. Genetic Downregulation of AMPK-Alpha Isoforms Uncovers the Mechanism by Which Metformin Decreases FA Uptake and Oxidation in Skeletal Muscle Cells. Am. J. Physiol. Cell Physiol. (2010) 299 C1549-C1561. DOI: 10.1152/ajpcell.00279.2010 65. Rivas D.A., Yaspelkis B.B., Hawley J.A., Lessard S.J.. Lipid-Induced MTOR Activation in Rat Skeletal Muscle Reversed by Exercise and 5’-Aminoimidazole-4-Carboxamide-1-Beta-D-Ribofuranoside. J. Endocrinol. (2009) 202 441-451. DOI: 10.1677/JOE-09-0202 66. Wang X., Yu W., Nawaz A., Guan F., Sun S., Wang C.. Palmitate Induced Insulin Resistance by PKCtheta-Dependent Activation of MTOR/S6K Pathway in C2C12 Myotubes. Exp. Clin. Endocrinol. Diabetes (2010) 118 657-661. DOI: 10.1055/s-0030-1252069 67. Woo J.H., Shin K.O., Lee Y.H., Jang K.S., Bae J.Y., Roh H.T.. Effects of Treadmill Exercise on Skeletal Muscle MTOR Signaling Pathway in High-Fat Diet-Induced Obese Mice. J. Phys. Ther. Sci. (2016) 28 1260-1265. DOI: 10.1589/jpts.28.1260 68. Le Bacquer O., Petroulakis E., Paglialunga S., Poulin F., Richard D., Cianflone K., Sonenberg N.. Elevated Sensitivity to Diet-Induced Obesity and Insulin Resistance in Mice Lacking 4E-BP1 and 4E-BP2. J. Clin. Investig. (2007) 117 387-396. DOI: 10.1172/JCI29528 69. Kwon B., Querfurth H.W.. Palmitate Activates MTOR/P70S6K through AMPK Inhibition and Hypophosphorylation of Raptor in Skeletal Muscle Cells: Reversal by Oleate Is Similar to Metformin. Biochimie (2015) 118 141-150. DOI: 10.1016/j.biochi.2015.09.006 70. Tzatsos A.. Raptor Binds the SAIN (Shc and IRS-1 NPXY Binding) Domain of Insulin Receptor Substrate-1 (IRS-1) and Regulates the Phosphorylation of IRS-1 at Ser-636/639 by MTOR. J. Biol. Chem. (2009) 284 22525-22534. DOI: 10.1074/jbc.M109.027748 71. Shah O.J., Wang Z., Hunter T.. Inappropriate Activation of the TSC/Rheb/MTOR/S6K Cassette Induces IRS1/2 Depletion, Insulin Resistance, and Cell Survival Deficiencies. Curr. Biol. (2004) 14 1650-1656. DOI: 10.1016/j.cub.2004.08.026 72. Tremblay F., Marette A.. Amino Acid and Insulin Signaling via the MTOR/P70 S6 Kinase Pathway. A Negative Feedback Mechanism Leading to Insulin Resistance in Skeletal Muscle Cells. J. Biol. Chem. (2001) 276 38052-38060. DOI: 10.1074/jbc.M106703200 73. Harrington L.S., Findlay G.M., Gray A., Tolkacheva T., Wigfield S., Rebholz H., Barnett J., Leslie N.R., Cheng S., Shepherd P.R.. The TSC1-2 Tumor Suppressor Controls Insulin-PI3K Signaling via Regulation of IRS Proteins. J. Cell Biol. (2004) 166 213-223. DOI: 10.1083/jcb.200403069 74. Guillén C., Benito M.. MTORC1 Overactivation as a Key Aging Factor in the Progression to Type 2 Diabetes Mellitus. Front. Endocrinol. (2018) 9 621. DOI: 10.3389/fendo.2018.00621 75. Deng Y., Bhattacharya S., Swamy O.R., Tandon R., Wang Y., Janda R., Riedel H.. Growth Factor Receptor-Binding Protein 10 (Grb10) as a Partner of Phosphatidylinositol 3-Kinase in Metabolic Insulin Action. J. Biol. Chem. (2003) 278 39311-39322. DOI: 10.1074/jbc.M304599200 76. Yu Y., Yoon S.-O., Poulogiannis G., Yang Q., Ma X.M., Villén J., Kubica N., Hoffman G.R., Cantley L.C., Gygi S.P.. Quantitative Phosphoproteomic Analysis Identifies the Adaptor Protein Grb10 as an MTORC1 Substrate That Negatively Regulates Insulin Signaling. Science (2011) 332 1322-1326. DOI: 10.1126/science.1199484 77. Hsu P.P., Kang S.A., Rameseder J., Zhang Y., Ottina K.A., Lim D., Peterson T.R., Choi Y., Gray N.S., Yaffe M.B.. The MTOR-Regulated Phosphoproteome Reveals a Mechanism of MTORC1-Mediated Inhibition of Growth Factor Signaling. Science (2011) 332 1317-1322. DOI: 10.1126/science.1199498 78. Cantó C., Auwerx J.. AMP-Activated Protein Kinase and Its Downstream Transcriptional Pathways. Cell. Mol. Life Sci. (2010) 67 3407-3423. DOI: 10.1007/s00018-010-0454-z 79. Mihaylova M.M., Shaw R.J.. The AMPK Signalling Pathway Coordinates Cell Growth, Autophagy and Metabolism. Nat. Cell Biol. (2011) 13 1016-1023. DOI: 10.1038/ncb2329 80. Johnson J.J.. Carnosol: A Promising Anti-Cancer and Anti-Inflammatory Agent. Cancer Lett. (2011) 305 1-7. DOI: 10.1016/j.canlet.2011.02.005
--- title: 'Comparison of the Effect of Different Conditioning Media on the Angiogenic Potential of Hypoxia Preconditioned Blood-Derived Secretomes: Towards Engineering Next-Generation Autologous Growth Factor Cocktails' authors: - Philipp Moog - Jessica Hughes - Jun Jiang - Lynn Röper - Ulf Dornseifer - Arndt F. Schilling - Hans-Günther Machens - Ektoras Hadjipanayi journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049474 doi: 10.3390/ijms24065485 license: CC BY 4.0 --- # Comparison of the Effect of Different Conditioning Media on the Angiogenic Potential of Hypoxia Preconditioned Blood-Derived Secretomes: Towards Engineering Next-Generation Autologous Growth Factor Cocktails ## Abstract Hypoxia Preconditioned Plasma (HPP) and Serum (HPS) are regenerative blood-derived growth factor compositions that have been extensively examined for their angiogenic and lymphangiogenic activity towards wound healing and tissue repair. Optimization of these secretomes’ growth factor profile, through adjustments of the conditioning parameters, is a key step towards clinical application. In this study, the autologous liquid components (plasma/serum) of HPP and HPS were replaced with various conditioning media (NaCl, PBS, Glucose $5\%$, AIM V medium) and were analyzed in terms of key pro- (VEGF-A, EGF) and anti-angiogenic (TSP-1, PF-4) protein factors, as well as their ability to promote microvessel formation in vitro. We found that media substitution resulted in changes in the concentration of the aforementioned growth factors, and also influenced their ability to induce angiogenesis. While NaCl and PBS led to a lower concentration of all growth factors examined, and consequently an inferior tube formation response, replacement with Glucose $5\%$ resulted in increased growth factor concentrations in anticoagulated blood-derived secretomes, likely due to stimulation of platelet factor release. Medium substitution with Glucose $5\%$ and specialized peripheral blood cell-culture AIM V medium generated comparable tube formation to HPP and HPS controls. Altogether, our data suggest that medium replacement of plasma and serum may significantly influence the growth factor profile of hypoxia-preconditioned blood-derived secretomes and, therefore, their potential application as tools for promoting therapeutic angiogenesis. ## 1. Introduction Complete restoration of physiological tissue architecture is the ultimate goal of regenerative medicine research [1]. Taking into account that wounds naturally heal via a set of complex and interactive processes, including hemostasis, inflammation, proliferation, and remodeling [2,3], gives an idea of the intricate nature of tissue repair. Research on regenerative therapies commonly focuses on stimulating angiogenesis (formation of new blood vessels) and improving tissue perfusion in order to provide an adequate supply of oxygen/nutrients to the wound bed, which presents an absolute prerequisite for optimal cellular function. Employing the body’s own resources towards this goal, via, for example, the utilization of autologous blood-derived products, is becoming a more favorable approach, since it overcomes the limitations imposed by our incomplete understanding of these complex mechanisms, while also harnessing the physiological response that is necessary for avoiding unwanted side effects [4]. In our previous work, we showed that angiogenesis and lymphangiogenesis have symbiotic roles in wound healing, since they both appear to rely on overlapping growth factor mechanisms [5,6,7,8]. There is now strong evidence indicating that hypoxia preconditioned blood-derived secretomes could constitute a new generation of autologous, bioactive compositions that can supply the necessary biochemical signals for stimulating angiogenesis/lymphangiogenesis, thus driving wound healing to completion [5,6,8,9,10,11,12,13,14]. These growth factor compositions can be obtained through the method of hypoxia-adjusted in vitro preconditioning of peripheral blood cells (PBCs), first proposed by Hadjipanayi and Schilling [5,10,11,12]. Conditioning PBCs under the very same conditions encountered within a wound microenvironment, i.e., physiological temperature and hypoxia, offers a means for optimizing the angiogenic potential of Hypoxia Preconditioned Plasma (HPP) and Hypoxia Preconditioned Serum (HPS), which can be differentially prepared by adjusting blood coagulation prior to hypoxic conditioning [5,6,9,10,11,12,15]. More specifically, we have shown that the angiogenic potential of blood-derived secretomes is defined by the complex stoichiometry of their component pro- and anti-angiogenic factor proteins, rather than the concentration of one or more individual growth factors [6,8,12,15]. The angiogenic potency of hypoxia preconditioned secretomes is further highlighted by the fact that they maintain their pro-angiogenic activity in vitro, even when they are prepared from peripheral blood that has been obtained from patients who receive oral anticoagulation due to underlying vascular pathology or who suffer from diabetes mellitus [16]. The ability to control the growth factor composition of these blood-derived secretomes is a powerful tool for optimizing their angiogenic potency and, thus, clinical value as a wound healing therapy. Beyond controlling the incubation temperature and oxygen tension, which are key parameters during PBC conditioning, it may be possible that the proteomic profile is dependent on the nature of the nutritional medium used during this process. It is indeed known that optimal wound healing is dependent on nutritional status, as shown, for example, by a correlation between low serum albumin and the development of pressure ulcers [17]. There may also be direct effects of the supply of nutrients to the wound bed, as demonstrated, for example, by the ability of honey (which comprises a wide variety of active compounds, including flavonoids, phenolic acid, organic acids, enzymes, and vitamins) to improve the wound healing process [18]. Furthermore, some micronutrients, such as vitamins A, C, and E, may deactivate free radicals and potentially accelerate wound healing [19,20]. Studies have also shown that vitamin A functions as a hormone, altering the activity of epithelial cells, melanocytes, fibroblasts, and endothelial cells through its action on the family of retinoic acid receptors [21], while vitamin C promotes neutrophil and fibroblast activity and is required for optimal angiogenesis [19]. Proteins, on the other hand, are vital in keeping a positive nitrogen balance for all stages of the wound healing cascade, including fibroblast proliferation, collagen synthesis, angiogenesis, and immune response [20]. Beyond optimal cellular function, the tissue repair signaling itself is dependent on protein support, since growth factor production requires amino acid availability [22]. On the negative side of things, hyperglycemia correlates with stiffer blood vessels, which causes slower circulation and, consequently, reduced tissue oxygenation [23,24,25]. Indeed, chronic and acute hyperglycemia can trigger platelet activation [26,27], while in diabetic patients, the production of several growth factors involved in initiating and sustaining the healing process is compromised [25]; for example, vascular endothelial growth factor (VEGF) and transforming growth factor-beta (TGF-beta) protein expression is reduced in diabetic dermal wounds [28,29]. Recently, plasma lactate has emerged as an early indicator of aberrant metabolism, specifically, development of insulin resistance and diabetes mellitus [30]. In the context of wound healing, lactate accumulates as a consequence of both aerobic and anaerobic glycolysis following microcirculation disruption, immune activation, and increased cell proliferation [31]. Studies have repeatedly demonstrated its important contribution in tissue repair by promoting angiogenesis and collagen production [31,32,33,34]. Based on the established correlation between nutritional status and the wound healing response, we sought to examine whether PBC growth factor production, and consequently the angiogenic activity of hypoxia preconditioned secretomes, is also dependent on the type of nutritional medium used during PBC hypoxic conditioning. This was tested by substituting the autologous liquid components in HPP and HPS (i.e., plasma and serum, respectively) with various media (NaCl, phosphate buffered saline (PBS), Glucose $5\%$ (G$5\%$), and AIM V serum-free culture medium) in order to investigate their influence on the concentration of key pro- (VEGF-A, epidermal growth factor (EGF)) and anti-angiogenic protein factors (platelet factor-4 (PF-4), thrombospondin-1 (TSP-1)), before analyzing their ability to promote microvessel formation in vitro. Our findings suggest that it is indeed possible to influence the bioactivity of hypoxia preconditioned secretomes through medium substitution, potentially opening a new avenue for developing next-generation autologous growth factor cocktails for tissue repair and regeneration. ## 2.1. Analysis of Lactate Concentration in Serum and HPS Depending on Exercise Level We hypothesized that an increased lactate concentration would develop in hypoxia preconditioned serum (HPS), compared to baseline fresh serum, as a result of the exposure of PBCs to persistent hypoxia. Furthermore, we sought to identify a possible effect of regular exercise on the lactate concentration in both fresh serum and HPS. As expected, the lactate concentration increased significantly by approx. 12-fold in both non-exercising (1.10 ± 0.09 vs. 12.26 ± 0.84 mmol/L, $p \leq 0.0001$) and exercising (1.47 ± 0.23 vs. 12.33 ± 1.72 mmol/L, $p \leq 0.0001$) subjects over the 4-day incubation period (Figure 1), which indirectly confirmed the development of a hypoxic microenvironment during blood conditioning at 37 °C. Interestingly, there was no difference in lactate levels in either serum or HPS between the non-exercising and exercising group. ## 2.2. Analysis of Pro-Angiogenic Growth Factor Concentration (VEGF-A, EGF) in Various PBC Conditioning Media To establish a growth factor concentration baseline, we first quantitatively analyzed via ELISA the concentration of angiogenesis-promoting growth factors (VEGF-A, EGF) in fresh plasma and serum and compared them to the hypoxia preconditioned plasma/serum (HPP/HPS) levels. We then analyzed the concentration of these protein factors in the different PBC conditioning media tested here as plasma/serum substitutes, i.e., hypoxia preconditioned normal saline (HPP/HPS-NaCl), hypoxia preconditioned phosphate buffered saline (HPP/HPS-PBS), hypoxia preconditioned Glucose $5\%$ (HPP/HPS-G$5\%$), and hypoxia preconditioned AIM V medium (HPP/HPS-AIM). The concentration of VEGF-A in HPP and HPS showed approx. a 5-fold increase after 4 days of incubation compared to the baseline level in fresh plasma (plasma vs. HPP: 142.63 vs. 757.00 pg/mL, $p \leq 0.05$) and serum (serum vs. HPS: 412.25 vs. 2350.33 pg/mL, $p \leq 0.05$) (Figure 2A,B). This similar relative increase implied that VEGF-A upregulation through PBC hypoxic conditioning was independent of platelet activation. Nonetheless, we found significantly higher levels of VEGF-A after conditioning without anticoagulants (4 day HPS vs. HPP: 2350.33 vs. 757.00 pg/mL, $p \leq 0.05$), while VEGF-A levels in fresh plasma and serum were not significantly different (142.63 vs. 412.25 pg/mL, $$p \leq 0.14$$), suggesting a moderate contribution from purely platelet-derived VEGF-A. Between day 4 and 8 of incubation, there was no further increase in the VEGF-A level in either HPP or HPS. In the next step of testing, we demonstrated an increase of the VEGF-A concentration from 2 to 4 and 8 days incubation in hypoxia preconditioned plasma substitutes (derived from anti-coagulated blood samples), with the highest level being achieved with HPP-AIM, which was twice as high compared to the HPP control on incubation day 8 (2079.25 vs. 974.50 pg/mL, $p \leq 0.05$) (Figure 2A). The observed increase in VEGF-A concentration, compared to baseline plasma levels, was least pronounced in HPP-PBS at all incubation days, with VEGF-A levels being significantly lower than the HPP control on incubation day 4 (348.5 vs. 757.00 pg/mL, $p \leq 0.05$) and day 8 (393.98 vs. 974.50 pg/mL, $p \leq 0.05$). A comparison of the VEGF-A concentration in hypoxia preconditioned serum substitutes (derived from coagulated blood samples) similarly indicated an increase from 2 to 4 and 8 days of incubation (Figure 2B). In contrast to all other media substitutes, the VEGF-A concentration of HPS-AIM samples exceeded the VEGF-A concentration of HPS control on each incubation day, although this difference was not statistically significant ($p \leq 0.05$). However, there were significant differences between HPS-AIM and other hypoxia preconditioned media VEGF-A levels, especially on incubation day 4, where a 7-fold increase was observed compared to HPS-NaCl (3637.00 vs. 517.00 pg/mL, $p \leq 0.05$) and HPS-G$5\%$ (3637.00 vs. 527.67 pg/mL, $p \leq 0.05$). This higher VEGF-A concentration in HPS-AIM persisted on incubation day 8, but was no longer significant. VEGF-A levels in all other hypoxia preconditioned serum substitutes (HPS-NaCl, HPS-PBS, and HPS-G$5\%$) were below (day 2 and 4) or approx. at the same level (day 8) as the HPS control and hardly differed from one another (Figure 2B). Similarly, the concentration of EGF in hypoxia preconditioned plasma (HPP) showed a 5-fold increase after 2 and 4 days of incubation compared to fresh plasma (101.90 vs. 27.50 pg/mL and 150.95 vs. 27.50 pg/mL, both $p \leq 0.01$), with no further significant difference between 4 and 8 days of incubation (Figure 2C). In contrast, a larger 40-fold increase in EGF concentration was observed in HPS compared to the baseline level in fresh serum (27.59 vs. 1133.13 pg/mL, $p \leq 0.05$) after 4 days of blood incubation. Importantly, the EGF concentration in HPS (range = 800–1200 pg/mL) was significantly higher than in HPP (range = 100–150 pg/mL) at all incubation time points, indicating the generation of a significant amount of platelet-derived EGF during conditioning. With regards to hypoxia preconditioned plasma substitutes, the EGF concentration of HPP-G$5\%$ exceeded that of the HPP control on all incubation days (day 2: 363.65 vs. 101.9 pg/mL; day 4: 327.10 vs. 150.95 pg/mL; day 8: 294.80 vs. 153.30 pg/mL, each $p \leq 0.05$). In contrast, the EGF concentration of HPP-NaCl and HPP-PBS was approx. half that of the HPP control on all incubation days ($p \leq 0.05$). HPP-AIM had a similar EGF level as the HPP control at all time points. A comparison of hypoxia preconditioned serum substitutes paradoxically showed that HPS-G$5\%$ had the lowest EGF concentration on all incubation days (Figure 2D), being significantly lower than the HPS control at 2 and 4 days of incubation (day 2: 235.63 vs. 742.5 pg/mL; day 4: 243.95 vs. 1133.13 pg/mL, both $p \leq 0.05$). HPS-NaCl, HPS-PBS, and HPS-AIM had a similar EGF concentration, which was consistently below that of the HPS control, albeit non-significantly ($p \leq 0.05$). ## 2.3. Analysis of Anti-Angiogenic Growth Factor Concentration (TSP-1, PF-4) in Various PBC Conditioning Media As a next step, we sought to analyze the concentration of the platelet-derived angiogenic inhibitors TSP-1 and PF-4 in fresh plasma and serum and compared them to their hypoxia preconditioned counterparts (HPP/HPS), as well as to the different conditioning media substitutes previously tested. As shown in Figure 3A, the TSP-1 concentration of hypoxia preconditioned plasma (HPP) was comparable to that of fresh plasma and remained relatively stable over the entire incubation period of 8 days. This finding confirmed minimal platelet activation as a result of anticoagulated blood conditioning. In terms of the various hypoxia preconditioned plasma substitutes (HPP-NaCl, HPP-PBS, HPP-G$5\%$, and HPP-AIM), the highest TSP-1 concentrations were achieved in HPP-G$5\%$ and HPP-AIM, with a peak level observed on incubation day 4 that was significantly higher than the HPP control (4490.00 vs. 1755.00 ng/mL, 3735.00 vs. 1755.00 ng/mL, both $p \leq 0.05$). HPP-NaCl and HPP-PBS showed the lowest TSP-1 concentration, with significant differences from the HPP control, on all incubation days ($p \leq 0.05$). A comparison of the TSP-1 concentration in fresh serum and hypoxia preconditioned serum (HPS) of 2, 4, and 8 days incubation showed no significant differences (Figure 3B, $p \leq 0.05$), although levels here were approx. 10-fold higher than those achieved in plasma and HPP. Furthermore, the TSP-1 concentration in the HPS control was overall greater than in the various hypoxia preconditioned serum substitutes (HPS-NaCl, HPS-PBS, HPS-G$5\%$, and HPS-AIM), and these differences were occasionally significant ($p \leq 0.05$). The lowest TSP-1 concentration at 2 and 4 days of incubation was seen in HPS-G$5\%$, which starkly contrasted with the relatively high level seen in the anticoagulated state of HPP-G$5\%$. Indeed, at these time points, the TSP-1 concentration in HPS-G$5\%$ was 5- to 10-fold lower than in the HPS control ($p \leq 0.05$) and 2- to 5-fold lower than in HPS-NaCl, HPS-PBS, and HPS-AIM ($p \leq 0.05$). As expected from the TSP-1 data, the PF-4 concentration of plasma and HPP was comparable at all time points (Figure 3C), consistent with minimal platelet activation in these secretomes. In a similar pattern, the PF-4 concentration of HPP-G$5\%$ was approx. 5-fold higher than in the other anticoagulated blood-derived conditioning media (HPP-NaCl, HPP-PBS, and HPP-AIM) at 2 and 4 days of incubation ($p \leq 0.05$). Interestingly, this difference appeared to be reversed on incubation day 8, when the PF-4 concentration of HPP-G$5\%$ was significantly lower than in HPP-NaCl, HPP-PBS, and HPP-AIM ($p \leq 0.05$). The approx. 10-fold higher PF-4 concentration observed in fresh serum (2397.00 ng/mL) compared to fresh plasma (270.10 ng/mL) confirmed the platelet activation as a result of blood coagulation (Figure 3C,D). An even larger increase (12-fold) in the PF-4 level was recorded with HPS following 2 days of blood incubation compared to fresh serum (26,376.75 vs. 2397.00 ng/mL, $p \leq 0.05$). This difference persisted on incubation days 4 and 8. The PF-4 concentration in the various coagulated blood-derived hypoxia preconditioned media (HP-NaCl, HP-PBS, HP-G$5\%$, and HP-AIM) was comparable to fresh serum and generally significantly lower than in HPS at all time points ($p \leq 0.05$). Notably, and in agreement with the TSP-1 results, the lowest PF-4 concentration was seen in HPS-G$5\%$ (Figure 3D). ## 2.4. Effect of Various Conditioning Media on Microvessel Formation In Vitro Following an analysis of key pro- and anti-angiogenic protein factors, we moved on to investigate the ability of the blood-derived hypoxia preconditioned media to induce microvessel formation in human umbilical vein endothelial cell (HUVEC) in vitro cultures. With regards to anti-coagulated blood-derived secretomes, there was a 4-fold increase in the mean number of tubules observed in HPP cultures compared to fresh plasma, which was significant for 2-day HPP incubation (47.75 vs. 11.16, $p \leq 0.001$), while longer preconditioning periods (4 and 8 days) did not produce significant changes ($p \leq 0.05$) (Figure 4), indicating that an early plateau was reached in terms of the angiogenic response in this setting. At 2 days incubation, the tube formation results from HPP-G$5\%$ were comparable to the HPP control ($p \leq 0.05$), while HPP-NaCl, HPP-PBS, and HPP-AIM showed lower tube formation, with one-third to one-fourth the number of tubes observed in the HPP control and HPP-G$5\%$ cultures ($p \leq 0.05$). Nonetheless, even in these HPP-NaCl and HPP-PBS cultures, microvessel formation showed a 5-fold increase between 2 and 4 days of incubation (9.92 vs. 57.41; 3.67 vs. 20.50, both $p \leq 0.01$), which indicated that the duration of the conditioning period positively influenced the angiogenic activity of these secretomes to a somewhat greater extent than with the glucose or AIM medium substitution. Despite the aforementioned differences in the protein quantification assays, there were no significant differences in terms of microvessel formation between the fresh serum- and HPS-incubated HUVEC cultures, regardless of the duration of the HPS preconditioning (Figure 5A,B). Furthermore, medium substitution with NaCl, PBS, G$5\%$, or AIM did not appear to offer an advantage at 2 and 4 days of incubation when these secretomes were derived from coagulated blood. While the mean number of tubules generated by the 2- and 4-day preconditioned media (HPS-NaCl, HPS-PBS, HPS-G$5\%$, and HPS-AIM) was comparable, there was a dramatic drop in the mean tube number observed in the cultures of 8-day preconditioned media, which was significant compared to the 4-day HPS-NaCl (all $p \leq 0.05$) and HPS-G$5\%$ (all $p \leq 0.05$) cultures. This suggested that, in contrast to the anticoagulated blood preconditioning, the length of the incubation period negatively affected the angiogenic activity of the coagulated blood-derived media. ## 3. Discussion Hypoxia preconditioned blood-derived secretomes represent a new generation of autologous growth factor preparations that can be produced through extracorporeal conditioning of peripheral blood cells (PBCs) under wound-simulating conditions, namely, physiological temperature and hypoxia [5,9,10,11,12]. We had previously demonstrated that hypoxia preconditioned plasma (HPP) and serum (HPS) supply angiogenesis- and lymphangiogenesis-specific signaling, similar to that naturally produced within the wound microenvironment [5,6,7,12]. HPP and HPS organically differ with respect to their protein factor composition, since they correlate with distinct wound healing phases, the former having a direct correlation with the hypoxia-induced, angiogenesis-driven proliferative phase, while the latter also incorporates the platelet-derived hemostatic phase [6,10,12]. Despite their differences, the clinical utility of these secretomes harnesses their angiogenic activity, since they can both provide a useful tool for stimulating microvessel sprouting and lymphatic vessel formation on demand [5,6,7,8,12]. As such, they could play an important role in a modern therapeutic strategy that aims to improve local tissue perfusion and accelerate tissue healing where this is a delay or stagnation [9,10,12,13,14]. We hypothesized that medium substitution, which effectively gets rid of lactate already present in serum, as well as that which accumulates during incubation (through continuous substitution), may offer a means for optimizing the conditioning microenvironment for improved PBC function and growth factor production. Here, we showed a 12-fold increase in the lactate concentration of HPS compared to fresh serum as a result of blood conditioning (Figure 1). We also hypothesized that the PBCs of subjects who regularly exercise may produce less lactate as a result of cellular adaptation [35,36]. However, the lactate concentrations in both serum and HPS were comparable between the exercising and non-exercising groups. Hunt et al. had postulated that an increased concentration of lactate in wounds presented a major signal for collagen synthesis and wound repair [31,32]. Indeed, lactate actively participates in the healing process through the activation of several molecular pathways that collectively promote angiogenesis via endothelial cell migration [31,33,37,38,39] and tube formation in vitro [31,37,38,40], as well as the recruitment of circulating vascular progenitor cells in vivo [31,34,41]. These results indicate that the by-product “lactate”, which is inadvertently generated via the exposure of PBCs to hypoxia, may actually support the wound healing activity of these secretomes; however, further experiments with high versus low lactate conditions, while controlling for other angiogenic factors, are needed to verify its angiogenic effect in the various conditioning media. The main thesis of this work is based on the notion that the net bioactivity of a growth factor-based regenerative therapy is effected through the balance of stimulatory and inhibitory signals. More specifically, it is known that a range of angiogenic inhibitors are of platelet origin and are thus released into coagulated blood-derived secretomes, e.g., HPS, as a result of platelet activation and degranulation [6,7,8,12,15]. In order to examine these effects, plasma and serum were substituted with different conditioning media, namely, hypoxia preconditioned normal saline (HPP/HPS-NaCl), phosphate buffered saline (HPP/HPS- PBS), Glucose $5\%$ (HPP/HPS-G$5\%$), and AIM V medium (HPP/HPS-AIM). These ‘novel’ compositions were subsequently analyzed in terms of the pro- and anti-angiogenic growth factor concentration and were tested for their ability to induce angiogenesis via tube formation assay. Quantitative analysis of pro-angiogenic growth factors (VEGF-A, EGF) in native hypoxia preconditioned secretomes (HPP/HPS) demonstrated a significant increase in VEGF-A concentration as a result of hypoxic conditioning, which became more evident after 4 days of incubation (Figure 2). However, none of the conditioning media substitutes tested appeared to offer a clear advantage in terms of this response, indicating that hypoxia regulation of VEGF-A expression may be the predominant factor determining its availability [42,43]. While there was a tendency for AIM V medium to induce more VEGF-A production, this effect only became significant after 8 days of incubation, in relation to the HPP control and other conditioning media, except for HPP-G$5\%$. The glucose-containing medium HPP-G$5\%$ appeared to induce more platelet-derived VEGF-A and EGF secretion, as the increase in the concentration of these factors (especially EGF) was only detectable when platelet activation was kept at a minimum through blood anticoagulation (Figure 2A,C). EGF levels in HPP-G$5\%$ were significantly increased already after 2 days of incubation compared to the HPP control, while this level was comparable to the corresponding HPS-G$5\%$ EGF value (Figure 2C,D). In this regard, the VEGF-A and EGF levels in HPP-G$5\%$ and HPS-G$5\%$ were comparable at all incubation time points, indicating no further factor release through platelet activation as a result of clotting, while in the absence of glucose, the effect of clotting-induced platelet activation was apparent in the 3-fold increase in the VEGF-A level and the 5-fold increase in the EGF level in the 4-day incubated HPS control compared to the HPP control (Figure 2A,B). These findings highlight the catalytic role that glucose plays in platelet-mediated factor secretion, in agreement with the literature [44,45,46]. Clinically, there is also abundant evidence to support that hyperglycemia in diabetic patients is associated with increased plasma VEGF-A, which in turn may cause hypertension and several vascular complications in diabetic patients [47,48,49,50]. HPP/HPS-NaCl and HPP/HPS-PBS showed the lowest level of VEGF-A and EGF production for all incubation periods tested, regardless of platelet activation. This may be due to the lack of nutritional components that support adequate cell viability and protein synthesis. Another reason could be the absence of important electrolytes, such as magnesium and, more importantly, calcium, which are both known to promote platelet activation [51,52]. With regards to anti-angiogenic growth factors, our data showed that the TSP-1 and PF-4 levels in HPP were comparable to fresh plasma at all incubation time points, and were significantly lower than in HPS, indicating that hypoxic conditioning itself did not promote TSP-1/PF-4 expression, but rather platelet activation was the main source of these angiogenic inhibitors in these secretomes (Figure 3A,B). This is consistent with the higher TSP-1 and PF-4 levels observed in anticoagulated blood-derived HPP-G$5\%$. The presence of glucose in the medium appeared to exert a strong stimulation of platelet activation, even in the background of previous heparin-mediated blood anticoagulation. However, similar to EGF, the levels of TSP-1 and PF-4 were significantly lower in HPS-G$5\%$ when compared to the HPS control at both 2 and 4 days of incubation (Figure 2 and Figure 3), suggesting that excess glucose interfered with platelet activation induced by clotting. This effect has been verified in the clinical setting, in which blood coagulation measurements via ROTEM (rotational thromboelastometry) are heavily interfered with if the blood sample is in a hyperglycemic state, resulting in an impaired clotting process, as evidenced by prolonged coagulation time measurements [53]. Altogether, these findings indicate that the elevated glucose concentration in the medium may significantly influence its final growth factor profile. Further investigation is required in order to decipher the relative contributions of glucose to PBC hypoxia-induced signaling and coagulation-mediated platelet factor secretion. The peak angiogenic response, analyzed here via in vitro tube formation assay, was observed with 4-day incubated secretomes and was generally stronger when these were derived from anti-coagulated rather than coagulated blood, likely due to the significantly higher concentration of angiogenic inhibitors in the latter (Figure 4 and Figure 5). Surprisingly, media substitution did not confer a significant improvement in terms of angiogenic response, compared to the HPP and HPS controls (Figure 4 and Figure 5). HPP/HPS-PBS consistently showed lower tube formation, in agreement with an overall weaker angiogenic growth factor response; however, even in these secretomes, there was a measurable increase in microvessels from 2 to 4 days, consistent with an increase in VEGF-A concentration (Figure 2A and Figure 4). Interestingly, HPP-G$5\%$ derived from anticoagulated blood induced a similar angiogenic response to the HPP control, despite the higher levels of the angiogenic inhibitors TSP-1 and PF-4, suggesting that this drawback was counterbalanced via an also higher concentration of pro-angiogenic factors, e.g., EGF, in HPP-G$5\%$. This is consistent with our previous results, which demonstrated that the presence of type 1 and type 2 diabetes mellitus does not appear to significantly impact the angiogenicity of HPP and HPS in vitro (as compared to secretomes obtained from healthy subjects) [16]. These findings suggest that the application of hypoxia preconditioning may serve as a tool for improving the angiogenic potency of blood-derived secretomes, thus repairing the angiogenic dysfunction that is a direct consequence of the disease state of diabetes and hyperglycemia. It is also important to acknowledge, however, that an oversupply of VEGF-A (and other pro-angiogenic factors) or blood glucose can lead to vascular leakage by disruption of cell–cell adherence, as well as tight-gap junction molecular signaling [54,55]. Thus, the characterization of endothelial identity and the analysis of the functionality of induced vasculature (e.g., the presence of immature endothelial microvessels and increased vascular permeability), after stimulation through conditioning media, are key points for investigation that should be examined in future studies, employing, for example, immunohistochemical staining for vascular leakage markers (e.g., VE-Cadherin) or more complex experiments for detecting leakage of proteins/cells in an in vivo vascular permeability assay [56,57]. ## 4.1. Ethical Approval All blood donors provided written informed consent, as directed by the ethics committee of the Technical University Munich, Germany, which approved this study (File Nr.: $\frac{497}{16}$S; amendment 2.0, date of approval: 18 September 2017). ## 4.2. Analysis of Lactate Concentration during Hypoxic Preconditioning Depending on Fitness Level Subjects were recruited in our clinic in 2021. We included 12 participants with an age of 34.5 ± 13.9 years (see demographics, Table 1). All participants were asked about their fitness level and number of exercising hours per week. Participants assigned to the “exercise group” were young healthy adults, without any medication and comorbidities, who exercised more than two hours per week. Participants who exercised less than two hours per week were assigned to the “no exercise” group. The age distribution was equal in both groups ($$p \leq 0.24$$). Smokers were defined as those who had smoked more than one cigarette in the past three months. Blood-derived secretomes from the participants were prepared as described in Section 2.3 and Section 2.4, and lactate levels were analyzed by blood gas analysis (Siemens Healthineers, Rapid Point 500, Erlangen, Germany). ## 4.3. Preparation of Blood Plasma/Serum and Hypoxia Preconditioned Plasma (HPP)/Serum (HPS) Samples Peripheral venous blood (20 mL) was collected from young, healthy, non-smoking subjects ($$n = 4$$), who were not taking any medication and were without known comorbidities, in a 30 mL polypropylene syringe (Omnifix®, Braun AG, Melsungen, Germany) that contained no additive for normal serum and HPS preparation or was prefilled with 1 mL heparin (Medunasal®, Heparin 500 I.U. 5 mL ampoules, Sintetica®, Münster, Germany) for normal plasma and HPP preparation, under sterile and standardized conditions (Blood Collection Set 0.8 × 19 mm × 178 mm; Safety-Lok, CE 0050, BD Vacutainer, Franklin Lakes, New Jersey, USA). In the next step, the blood was centrifuged at 2000 rpm for 15 min at room temperature (22 °C). After centrifugation, the blood was separated into three layers, from bottom to top: red blood cell component (RBCs), buffy coat/clot, and plasma/serum, so that the top layer (plasma or serum) could be filtered into a new syringe. For HPP/HPS preparation, the protocol described by Hadjipanayi et al. was used [10,12]. Briefly, following blood sampling, a 0.2 µm pore filter was attached to the syringe (Sterifix®, CE 0123, Braun AG, Melsungen, Germany), and by pulling the plunger, 5 mL of air was drawn into the syringe through the filter. Subsequently, the filter was removed, and the capped syringe was placed upright in an incubator (37 °C/$5\%$ CO2) and incubated for 2, 4, or 8 days (blood incubation time) without prior centrifugation. Pericellular local hypoxia (~$1\%$ O2) was induced in situ through cell-mediated O2 consumption by controlling the blood volume per unit area (BVUA > 1 mL/cm2) and, consequently, the PBC seeding density in the syringe [10,15]. After the predefined incubation time, the blood was passively separated into three layers, from top to bottom: hypoxia preconditioned plasma (HPP)/hypoxia preconditioned serum (HPS), buffy coat/clot, and red blood cell (RBC) component, so that the top layer comprising HPP or HPS could be filtered (0.2 µm pore filter, Sterifix®, Braun AG, Melsungen, Germany) into a new syringe, removing cells/cellular debris. ## 4.4. Preparation of Hypoxia Preconditioned Media Samples through Plasma and Serum Substitution Plasma/serum were prepared as described above, from the same four subjects. After centrifugation, plasma/serum was removed (note: special care was taken not to disturb the buffy coat and clot in plasma and serum samples, respectively) under sterile conditions and the volume measured (Figure 6). The same volume was then substituted in each case by saline (NaCl $0.9\%$, B. Braun AG, Melsungen, Germany), Phosphate buffered saline (PBS) formulated without calcium and magnesium (Gibco, Thermo Fisher Science, Waltham, MA, USA), Glucose $5\%$ solution containing glucose monohydrate (G-$5\%$, B. Braun AG, Melsungen, Germany), or AIM V medium (AIM V serum-free medium containing L-glutamine, 50 µg/mL streptomycin sulfate, and 10 µg/mL gentamicin sulfate; Gibco, Thermo Fisher Science, Waltham, MA, USA). Then, a 0.2 µm pore filter was attached to the syringe (Sterifix®, CE 0123, Braun AG, Melsungen, Germany), and by pulling the plunger, 5 mL of air was drawn into the syringe through the 0.2 µm pore filter (Sterifix®, B Braun AG, Melsungen, Germany). Subsequently, the filter was removed, and the capped syringe was placed upright in an incubator (37 °C/$5\%$ CO2) and incubated for 2, 4, or 8 days (blood incubation time). Pericellular local hypoxia (~$1\%$ O2) was induced in situ through cell-mediated O2 consumption by controlling the blood volume per unit area (BVUA > 1 mL/cm2) and, consequently, the PBC seeding density in the syringe [10,15]. After the predefined incubation time, hypoxia preconditioned media (HPP/HPS-NaCl/-PBS/-G-$5\%$/-AIM) were filtered (Sterifix®, Braun AG, Melsungen, Germany) into a new syringe, removing cells/cellular debris (Figure 6). ## 4.5. Quantitative Analysis of VEGF-A, EGF, TSP-1, PF-4 Concentration in Hypoxia Preconditioned Media Blood-derived secretomes—fresh plasma and fresh serum, hypoxia preconditioned plasma (HPP) and serum (HPS), and hypoxia preconditioned media (HPP/HPS -NaCl/-PBS/-G-$5\%$/-AIM) were sampled and analyzed by ELISA for vascular endothelial growth factor (VEGF-A), epidermal growth factor (EGF), thrombospondin-1 (TSP-1), and platelet-factor-4 (PF-4) (DY293B for VEGF-A, DY236 for EGF, DY3074 for TSP-1, DY795 for PF-4, Duoset, R&D Systems, Inc., Minneapolis, MN, USA), according to the manufacturer’s instructions. Factor concentrations in blood-derived secretomes/media were measured immediately after the predefined hypoxic incubation period (2, 4, or 8 days). All conditions were tested in triplicate per blood donor, and a total of four donors was taken for final evaluation. ## 4.6. Analysis of the Effect of Hypoxia Preconditioned Media on Microvessel Formation In Vitro The angiogenic potential of blood-derived secretomes was tested in an in vitro angiogenesis assay by assessing their ability to induce microvessel formation in human umbilical vein endothelial cells (HUVECs, CellSystems, Troisdorf, Germany) seeded on factor-reduced Matrigel (BD, Heidelberg, Germany). HUVECs were seeded at a density of 10 × 103/well, with 50 μL of test or control media added per well (μ-Slide Angiogenesis, Ibidi, Gräfelfing, Germany), and cultured in a $5\%$ CO$\frac{2}{37}$ °C incubator for 12 h. Cells were then stained with Calcein AM (PromoKine, Heidelberg, Germany), and endothelial cell tube formation was observed with fluorescence and phase contrast microscopy. Assessment of the extent of capillary-like network formation was carried out by counting the number of tubes per field. Plasma and serum controls were tested immediately (day 0), while hypoxia preconditioned plasma/serum (HPP/HPS) and hypoxia preconditioned media (HPP/HPS -NaCl/-PBS/-G-$5\%$/-AIM) were tested after the predefined incubation period (2, 4, or 8 days). All conditions were tested in triplicate per blood donor, and a total of four donors were taken for final evaluation. ## 4.7. Statistical Analysis For statistical analysis, we used the GraphPad Prism 9 software. Data sets were analyzed by two-way analysis of variance (ANOVA), with subsequent comparisons using Tukey’s post hoc analysis. All values are expressed as means ± standard deviation (SD). A value of $p \leq 0.05$ was considered statistically significant. ## 5. Conclusions Our data suggest that the presence of lactate in hypoxia preconditioned plasma and serum does not limit their bioactivity, which makes the need to remove it through medium substitution redundant. In addition to the availability of important electrolytes, e.g., calcium, the adjustment of blood glucose seems to have an influence on the pro- and anti-angiogenic growth factor profile of hypoxia preconditioned secretomes. Based on the results shown in this work, it appears that medium substitution of plasma/serum does not offer a clear benefit in terms of angiogenic potency, at least in vitro. Nonetheless, it may be possible to influence the secretomes’ pro-angiogenic factor (e.g., VEGF) levels via targeted medium substitution, especially with prolonged incubation periods. Medium substitution may, therefore, offer a useful tool for tailoring growth factor cocktails, based on peripheral blood hypoxic preconditioning, to specific applications. ## 6. Patents Device-based methods for localized delivery of cell-free carriers with stress-induced cellular factors. ( AU2013214187 (B2); 9 February 2017): Schilling Arndt, Hadjipanayi Ektoras, Machens Hans-Günther. ## References 1. Owlarn S., Klenner F., Schmidt D., Rabert F., Tomasso A., Reuter H., Mulaw M.A., Moritz S., Gentile L., Weidinger G.. Generic wound signals initiate regeneration in missing-tissue contexts. Nat. Commun. (2017) 8 2282. DOI: 10.1038/s41467-017-02338-x 2. Rodrigues M., Kosaric N., Bonham C.A., Gurtner G.C.. Wound Healing: A Cellular Perspective. Physiol. Rev. (2018) 99 665-706. DOI: 10.1152/physrev.00067.2017 3. Gurtner G.C., Werner S., Barrandon Y., Longaker M.T.. Wound repair and regeneration. Nature (2008) 453 314-321. DOI: 10.1038/nature07039 4. Laschke M.W., Menger M.D.. The simpler, the better: Tissue vascularization using the body’s own resources. Trends Biotechnol. (2022) 40 281-290. DOI: 10.1016/j.tibtech.2021.07.002 5. Hadjipanayi E., Bauer A.T., Moog P., Salgin B., Kuekrek H., Fersch B., Hopfner U., Meissner T., Schluter A., Ninkovic M.. Cell-free carrier system for localized delivery of peripheral blood cell-derived engineered factor signaling: Towards development of a one-step device for autologous angiogenic therapy. J. Control. Release (2013) 169 91-102. DOI: 10.1016/j.jconrel.2013.04.008 6. Moog P., Kirchhoff K., Bekeran S., Bauer A.T., Isenburg S., Dornseifer U., Machens HG S.A.H.E.. Comparative Evaluation of the Angiogenic Potential of Hypoxia Preconditioned Blood-Derived Secretomes and Platelet-Rich Plasma: An in vitro Analysis. Biomedicines (2020) 8. DOI: 10.3390/biomedicines8010016 7. Moog P., Schams R., Schneidinger A., Schilling A.F., Machens H.G., Hadjipanayi E., Dornseifer U.. Effect of Hypoxia Preconditioned Secretomes on Lymphangiogenic and Angiogenic Sprouting: An in Vitro Analysis. Biomedicines (2020) 8. DOI: 10.3390/biomedicines8090365 8. Jiang J., Cong X., Alageel S., Dornseifer U., Schilling A.F., Hadjipanayi E., Machens H.-G., Moog P.. In Vitro Comparison of Lymphangiogenic Potential of Hypoxia Preconditioned Serum (HPS) and Platelet-Rich Plasma (PRP). Int. J. Mol. Sci. (2023) 24. DOI: 10.3390/ijms24031961 9. Hadjipanayi E., Schilling A.. Hypoxia- based strategies for angiogenic induction. Organog. Landes Biosci. (2013) 9 261-272. DOI: 10.4161/org.25970 10. Hadjipanayi E., Schilling A.. Regeneration through autologous hypoxia preconditioned plasma. Organogenesis (2014) 10 164-169. DOI: 10.4161/org.29208 11. Hadjipanayi H., Bekeran S., Moog P.. Extracorporeal Wound Simulation as a Foundation for Tissue Repair und Regeneration Therapies. Int. J. Transpl. Plast. Surg. (2018) 2 1-10 12. Hadjipanayi E., Moog P., Bekeran S., Kirchhoff K., Berezhnoi A., Aguirre J., Bauer A.T., Kükrek H., Schmauss D., Hopfner U.. In Vitro Characterization of Hypoxia Preconditioned Serum (HPS)-Fibrin Hydrogels: Basis for an Injectable Biomimetic Tissue Regeneration Therapy. J. Funct. Biomater. (2019) 10. DOI: 10.3390/jfb10020022 13. Jiang J., Kraneburg U., Dornseifer U., Schilling A.F., Hadjipanayi E., Machens H.G., Moog P.. Hypoxia Preconditioned Serum (HPS)-Hydrogel Can Accelerate Dermal Wound Healing in Mice-An In Vivo Pilot Study. Biomedicines (2022) 10. DOI: 10.3390/biomedicines10010176 14. Jiang J., Röper L., Alageel S., Dornseifer U., Schilling A.F., Hadjipanayi E., Machens H.-G., Moog P.. Hypoxia Preconditioned Serum (HPS) Promotes Osteoblast Proliferation, Migration and Matrix Deposition. Biomedicines (2022) 10. DOI: 10.3390/biomedicines10071631 15. Hadjipanayi E., Kuhn P., Moog P., Bauer A.T., Kuekrek H., Mirzoyan L., Hummel A., Kirchhoff K., Salgin B., Isenburg S.. The Fibrin Matrix regulates Angiogenic Responses within the Hemostatic Microenvironment through Biochemical Control. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0135618 16. Moog P., Jensch M., Hughes J., Salgin B., Dornseifer U., Machens H.G., Schilling A.F., Hadjipanayi E.. Use of Oral Anticoagulation and Diabetes Do Not Inhibit the Angiogenic Potential of Hypoxia Preconditioned Blood-Derived Secretomes. Biomedicines (2020) 8. DOI: 10.3390/biomedicines8080283 17. Russell L.. The importance of patients’ nutritional status in wound healing. Br. J. Nurs. (2001) 10 S42-S49. DOI: 10.12968/bjon.2001.10.Sup1.5336 18. Tashkandi H.. Honey in wound healing: An updated review. Open Life Sci. (2021) 16 1091-1100. DOI: 10.1515/biol-2021-0084 19. Todorovic V.. Food and wounds: Nutritional factors in wound formation and healing. Br. J. Community Nurs. (2002) 7 43-54. DOI: 10.12968/bjcn.2002.7.Sup2.12981 20. Saghaleini S.H., Dehghan K., Shadvar K., Sanaie S., Mahmoodpoor A., Ostadi Z.. Pressure Ulcer and Nutrition. Indian J. Crit. Care Med. (2018) 22 283-289. DOI: 10.4103/ijccm.IJCCM_277_17 21. Quain A.M., Khardori N.M.. Nutrition in Wound Care Management: A Comprehensive Overview. Wounds (2015) 27 327-335. PMID: 27447105 22. Nishi H., Uchida K., Saito M., Yamanaka D., Nagata H., Tomoshige H., Miyata I., Ito K., Toyoshima Y., Takahashi S.-I.. Essential Amino Acid Intake Is Required for Sustaining Serum Insulin-like Growth Factor-I Levels but Is Not Necessarily Needed for Body Growth. Cells (2022) 11. DOI: 10.3390/cells11091523 23. Dinh T., Elder S., Veves A.. Delayed wound healing in diabetes: Considering future treatments. Diabetes Manag. (2011) 1 509-519. DOI: 10.2217/dmt.11.44 24. Tuttolomondo A., Maida C., Pinto A.. Diabetic foot syndrome: Immune-inflammatory features as possible cardiovascular markers in diabetes. World J. Orthop. (2015) 6 62-76. DOI: 10.5312/wjo.v6.i1.62 25. Spampinato S.F., Caruso G.I., De Pasquale R., Sortino M.A., Merlo S.. The Treatment of Impaired Wound Healing in Diabetes: Looking among Old Drugs. Pharmaceuticals (2020) 13. DOI: 10.3390/ph13040060 26. Assert R., Scherk G., Bumbure A., Pirags V., Schatz H., Pfeiffer A.F.. Regulation of protein kinase C by short term hyperglycaemia in human platelets in vivo and in vitro. Diabetologia (2001) 44 188-195. DOI: 10.1007/s001250051598 27. Kaur R., Kaur M., Singh J.. Endothelial dysfunction and platelet hyperactivity in type 2 diabetes mellitus: Molecular insights and therapeutic strategies. Cardiovasc. Diabetol. (2018) 17 121. DOI: 10.1186/s12933-018-0763-3 28. Kämpfer H., Pfeilschifter J., Frank S.. Expressional regulation of angiopoietin-1 and -2 and the tie-1 and -2 receptor tyrosine kinases during cutaneous wound healing: A comparative study of normal and impaired repair. Lab. Investig. (2001) 81 361-373. DOI: 10.1038/labinvest.3780244 29. Bitar M.S., Labbad Z.N.. Transforming growth factor-beta and insulin-like growth factor-I in relation to diabetes-induced impairment of wound healing. J. Surg. Res. (1996) 61 113-119. DOI: 10.1006/jsre.1996.0090 30. Hu J.R., Wu Y., Sacks F.M., Appel L.J., Miller Iii E.R., Young J.H., Juraschek S.P.. Effects of carbohydrate quality and amount on plasma lactate: Results from the OmniCarb trial. BMJ Open Diabetes Res. Care (2020) 8 e001457. DOI: 10.1136/bmjdrc-2020-001457 31. Porporato P.E., Payen V.L., De Saedeleer C.J., Préat V., Thissen J.P., Feron O., Sonveaux P.. Lactate stimulates angiogenesis and accelerates the healing of superficial and ischemic wounds in mice. Angiogenesis (2012) 15 581-592. DOI: 10.1007/s10456-012-9282-0 32. Hunt T.K., Conolly W., Aronson S.B., Goldstein P.. Anaerobic metabolism and wound healing: An hypothesis for the initiation and cessation of collagen synthesis in wounds. Am. J. Surg. (1978) 135 328-332. DOI: 10.1016/0002-9610(78)90061-2 33. Murray B., Wilson D.. A study of metabolites as intermediate effectors in angiogenesis. Angiogenesis (2001) 4 71-77. DOI: 10.1023/A:1016792319207 34. Hunt T.K., Aslam R.S., Beckert S., Wagner S., Ghani Q.P., Hussain M.Z., Roy S.T., Sen C.K.. Aerobically derived lactate stimulates revascularization and tissue repair via redox mechanisms. Antioxid. Redox Signal. (2007) 9 1115-1124. DOI: 10.1089/ars.2007.1674 35. Beneke R., Leithäuser R.M., Ochentel O.. Blood lactate diagnostics in exercise testing and training. Int. J. Sports Physiol. Perform. (2011) 6 8-24. DOI: 10.1123/ijspp.6.1.8 36. Favier R.J., Constable S.H., Chen M., Holloszy J.O.. Endurance exercise training reduces lactate production. J. Appl. Physiol. 1985 (1986) 61 885-889. DOI: 10.1152/jappl.1986.61.3.885 37. Vegran F., Boidot R., Michiels C., Sonveaux P., Feron O.. Lactate influx through the endothelial cell monocarboxylate transporter MCT1 supports an NF-kappaB/IL-8 pathway that drives tumor angiogenesis. Cancer Res. (2011) 71 2550-2560. DOI: 10.1158/0008-5472.CAN-10-2828 38. Sonveaux P., Copetti T., De Saedeleer C.J., Végran F., Verrax J., Kennedy K.M., Moon E.J., Dhup S.. Targeting the lactate transporter MCT1 in endothelial cells inhibits lactate-induced HIF-1 activation and tumor angiogenesis. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0033418 39. Beckert S., Farrahi F., Aslam R.S., Scheuenstuhl H., Königsrainer A., Hussain M.Z., Hunt T.K.. Lactate stimulates endothelial cell migration. Wound Repair Regen. (2006) 14 321-324. DOI: 10.1111/j.1743-6109.2006.00127.x 40. Burns P.A., Wilson D.J.. Angiogenesis mediated by metabolites is dependent on vascular endothelial growth factor (VEGF). Angiogenesis (2003) 6 73-77. DOI: 10.1023/A:1025862731653 41. Milovanova T.N., Bhopale V., Sorokina E.M., Moore J.S., Hunt T.K., Hauer-Jensen M., Velazquez O.C., Thom S.R.. Lactate stimulates vasculogenic stem cells via the thioredoxin system and engages an autocrine activation loop involving hypoxia-inducible factor 1. Mol. Cell Biol. (2008) 28 6248-6261. DOI: 10.1128/MCB.00795-08 42. Ramakrishnan S., Anand V., Roy S.. Vascular endothelial growth factor signaling in hypoxia and inflammation. J. Neuroimmune Pharmacol. (2014) 9 142-160. DOI: 10.1007/s11481-014-9531-7 43. Minchenko A., Bauer T., Salceda S., Caro J.. Hypoxic stimulation of vascular endothelial growth factor expression In Vitro and In Vivo. Lab. Investig. (1994) 71 374-379. PMID: 7933988 44. Sudic D., Razmara M., Forslund M., Ji Q., Hjemdahl P., Li N.. High glucose levels enhance platelet activation: Involvement of multiple mechanisms. Br. J. Haematol. (2006) 133 315-322. DOI: 10.1111/j.1365-2141.2006.06012.x 45. Fidler T.P., Marti A., Gerth K., Middleton E.A., Campbell R.A., Rondina M.T., Weyrich A.S., Abel E.D.. Glucose Metabolism Is Required for Platelet Hyperactivation in a Murine Model of Type 1 Diabetes. Diabetes (2019) 68 932-938. DOI: 10.2337/db18-0981 46. Shechter M., Merz C.N., Paul-Labrador M.J., Kaul S.. Blood glucose and platelet-dependent thrombosis in patients with coronary artery disease. J. Am. Coll. Cardiol. (2000) 35 300-307. DOI: 10.1016/S0735-1097(99)00545-8 47. Mazidi M., Rezaie P., Kengne A.P., Stathopoulou M.G., Azimi-Nezhad M., Siest S.. VEGF, the underlying factor for metabolic syndrome; fact or fiction?. Diabetes Metab. Syndr. (2017) 1 S61-S64. DOI: 10.1016/j.dsx.2016.12.004 48. Zhang Q., Fang W., Ma L., Wang Z.-D., Yang Y.-M., Lu Y.-Q.. VEGF levels in plasma in relation to metabolic control, inflammation, and microvascular complications in type-2 diabetes: A cohort study. Medicine (2018) 97 e0415. DOI: 10.1097/MD.0000000000010415 49. Mahdy R.A., Nada W.M., Hadhoud K.M., El-Tarhony S.A.. The role of vascular endothelial growth factor in the progression of diabetic vascular complications. Eye (2010) 24 1576-1584. DOI: 10.1038/eye.2010.86 50. Zafar M.I., Mills K., Ye X., Blakely B., Min J., Kong W., Zhang N., Gou L., Regmi A., Hu S.Q.. Association between the expression of vascular endothelial growth factors and metabolic syndrome or its components: A systematic review and meta-analysis. Diabetol. Metab. Syndr. (2018) 10 62. DOI: 10.1186/s13098-018-0363-0 51. Campbell J.E., Johnson K.P., Rodriguez A., Cap A.P.. Magnesium Augments Platelet Activation and Coagulation in Citrated Blood Product. Blood (2012) 120 3438. DOI: 10.1182/blood.V120.21.3438.3438 52. Davlouros P., Xanthopoulou I., Mparampoutis N., Giannopoulos G., Deftereos S., Alexopoulos D.. Role of Calcium in Platelet Activation: Novel Insights and Pharmacological Implications. Med. Chem. (2016) 12 131-138. DOI: 10.2174/157340641202160208195923 53. Shin H.J., Na H.S., Lee S., Lee G.W., Do S.H.. The effect of hyperglycemia on blood coagulation: In vitro, observational healthy-volunteer study using rotational thromboelastometry (ROTEM). Medicine (2016) 95 e4703. DOI: 10.1097/MD.0000000000004703 54. Wautier J.L., Wautier M.P.. Vascular Permeability in Diseases. Int. J. Mol. Sci. (2022) 23. DOI: 10.3390/ijms23073645 55. Weis S.M., Cheresh D.A.. Pathophysiological consequences of VEGF-induced vascular permeability. Nature (2005) 437 497-504. DOI: 10.1038/nature03987 56. Goncharov N.V., Nadeev A.D., Jenkins R.O., Avdonin P.V.. Markers and Biomarkers of Endothelium: When Something Is Rotten in the State. Oxid. Med. Cell Longev. (2017) 2017 9759735. DOI: 10.1155/2017/9759735 57. Claesson-Welsh L., Dejana E., McDonald D.M.. Permeability of the Endothelial Barrier: Identifying and Reconciling Controversies. Trends Mol. Med. (2021) 27 314-331. DOI: 10.1016/j.molmed.2020.11.006
--- title: A Comprehensive and Integrative Approach to MeCP2 Disease Transcriptomics authors: - Alexander J. Trostle - Lucian Li - Seon-Young Kim - Jiasheng Wang - Rami Al-Ouran - Hari Krishna Yalamanchili - Zhandong Liu - Ying-Wooi Wan journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049497 doi: 10.3390/ijms24065122 license: CC BY 4.0 --- # A Comprehensive and Integrative Approach to MeCP2 Disease Transcriptomics ## Abstract Mutations in MeCP2 result in a crippling neurological disease, but we lack a lucid picture of MeCP2′s molecular role. Individual transcriptomic studies yield inconsistent differentially expressed genes. To overcome these issues, we demonstrate a methodology to analyze all modern public data. We obtained relevant raw public transcriptomic data from GEO and ENA, then homogeneously processed it (QC, alignment to reference, differential expression analysis). We present a web portal to interactively access the mouse data, and we discovered a commonly perturbed core set of genes that transcends the limitations of any individual study. We then found functionally distinct, consistently up- and downregulated subsets within these genes and some bias to their location. We present this common core of genes as well as focused cores for up, down, cell fraction models, and some tissues. We observed enrichment for this mouse core in other species MeCP2 models and observed overlap with ASD models. By integrating and examining transcriptomic data at scale, we have uncovered the true picture of this dysregulation. The vast scale of these data enables us to analyze signal-to-noise, evaluate a molecular signature in an unbiased manner, and demonstrate a framework for future disease focused informatics work. ## 1. Introduction Experimental reproducibility is a key issue in the life sciences. Big data integration and analyses can uncover valuable insights otherwise missed in individual studies [1], but curating, processing, and analyzing data at scale is challenging. Researchers could aggregate publicly available processed results, but doing so will inevitably yield inconsistencies between datasets. Handling a meaningful quantity of raw high-throughput data requires extensive time, experience, and computational resources, and it would be wasteful for every researcher to do this themselves. Databases with abundant biological data exist but either do not focus on transcriptome perturbation (GTEx) [2] or require significant time and energy to extract and format data specific to a particular disease (ARCHS4) [3]. Similarly, many databases focus on specific model organisms and allow filtering by disease, such as Flybase [4] and the Rat Genome Database [5], but do not offer substantial disease-focused analysis. While there are some molecular-focused databases for well-studied diseases such as cancer (TCGA) [6], cBio Portal [7], and Alzheimer’s disease (AMP-AD) [8], there are no analogous databases for less-common diseases. The massive success of TCGA, in particular, makes the utility of disease databases clear. Rett syndrome (RTT) is a severe neurodevelopmental disorder in girls caused by mutations in the X-linked gene for methyl-CpG binding protein 2 (MECP2). Development proceeds normally until 6–18 months, at which point it stalls and then regresses [9]. Symptoms and progression can vary substantially between individuals, and despite recent advances, we still do not completely understand MeCP2′s molecular role. MECP2 duplication syndrome (MDS) is an overexpression in the same gene, and patients have substantial overlap in phenotype with RTT [10]. The pool of sequencing data for these diseases will only grow with time, and there is currently no centralized resource for it. Existing MeCP2-related disease databases are either primarily patient registries (Rett Database Network) [11], or they focus on mutation information, such as the IRSA North American Database [12] and RettBASE [13]. To fill this need, we created MECP2pedia, a database for molecular MeCP2. MECP2pedia is a uniformly processed and expansive collection of MeCP2 transcriptomic data, with readily accessible processed mouse data (expression, quality information, genomic tracks, and differential expression) that researchers can compare across any set of studies or data characteristics. In this study, we demonstrate a comprehensive approach by curating a vast resource of transcriptomic data and then unbiasedly analyzing and interpreting the results to understand expression dysregulations. We derived a consensus common core of misregulated genes, which we delineated into consistently up- and downregulated. We validated this separation through unbiased clustering and discovered distinct functional characteristics between the up and the down cores. We further confirmed this robust core with enrichment across species and comparisons to other disease models. Finally, power analysis showed how high of a false-negative rate the average individual transcriptomic profile incurs through a low sample number, and the presence of strong batch effects in these data demonstrates another problematic hurdle for researchers. Data integration is worthwhile but non-trivial. Batch effects, lack of power, reproducibility, and robustness are major hurdles for research. Big data helps mitigate these issues. Our approach to transcriptomic disease research yields results that are better and more complete than those attainable by conducting an individual experiment. Our demonstrated methodology can be applied broadly to other biological questions. ## 2.1. Comprehensive Resource of MeCP2 Transcriptomes The MECP2pedia portal can be accessed at http://www.mecp2pedia.org/. To generate this resource, we queried the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) [14] on 7 August 2019 for “MeCP2” and then filtered the results for “expression profiling by high throughput sequencing”. This search resulted in a preliminary list of 47 GEO entries. We filtered this list for entries dated 2015 or later and then further filtered for entries with at least two RNA-Seq wild-type samples and a treatment labeled either “knockout”, “Rett”, “point mutation”, “transgenic”, “overexpression”, or “MeCP2 duplication”. We retained 27 GEO entries, which we downloaded and processed with a uniform, streamlined Python pipeline (Figure 1A). In total, our processing yielded 546 sequence read archive (SRA) files, 753 FASTQ files, and 493 BAM (alignment) files. The total disk space used for processing was about 6 TB, and processing took about 2400 computing hours. The number of mouse samples per study is detailed in Figure S1 (see Supplementary Materials). For each of the processed samples, we generated raw read quality, alignment quality, and track information. For each study, we collected meta information and data characteristics (Figure 1B), and then, we aggregated all samples with matching characteristics into “contrasts” (a comparison of the expression between two groups) for differential gene analysis. Mouse studies typically contained one to two contrasts, with the exception of four studies that, respectively, contained 3, 4, 6, and 10 contrasts (bar width in Figure 1B). The 27 GEO entries provided a total of 58 contrasts, with 43 from mouse studies, 10 from humans, and 5 from other species. As this work grew, to both stay up to date with data and diversify our sources, both ArrayExpress and European Nucleotide Archive (ENA) were queried on 13 September 2022 for “MeCP2”. Identical inclusion criteria were applied, and 6 mouse studies comprising 11 contrasts were retained, all from ENA. All analysis, core, and figures were made with only the initial GEO data. A list of mouse contrasts and their metadata is provided in Table S1. Portal users can quickly and easily compare across studies the expression of specific genes of interest (Figure S7). Queries can be carried out individually or as a multi-gene search. Bar and scatter plots are available for each gene to show significance and fold change, and these results can be filtered across the uniformly annotated data characteristics. TPM (Transcripts Per Kilobase Million) is shown to allow the comparison of expression between contrasts [15]. Users can browse genome tracks by individual study, and studies can be compared to one another. Metadata for each contrast, raw read quality, and alignment quality are all available for each study. A “significant genes” tab allows users to filter genes using an FDR (false-discovery rate) and log2 fold change thresholds for each contrast, and this information is also downloadable. ## 2.2. MeCP2 Transcriptomics in Mice Reveal a Common Core of Misregulated Genes Our collected data are noteworthy for comprehensiveness and heterogeneity (Figure S2A). Tissue, mutation, and cell fraction are all highly variable across the breadth of collected RNA-Seq, giving users the most complete picture possible of MeCP2′s transcriptomic role. To understand these changes per study, we examined fold change and FDR from the differential gene expression (DEG) analysis, and as shown in Figure 2A, the ratio of significantly upregulated to downregulated genes is generally similar across the 43 contrasts. This finding is consistent with MeCP2′s reported role in both gene activation and repression [16]. Furthermore, changes in the majority of the dysregulated genes are less than two-fold. This magnitude of change is low when compared to other mouse disease models, which have substantial quantities of genes with changes greater than two-fold [17]. In a traditional RNA-Seq study with one contrast (a comparison of the expression between two groups), a gene is considered a DEG based on FDR thresholding and/or fold change cutoffs. However, there are many contrasts in our data, and a truly biologically important DEG should be observed consistently in several contrasts. We examined the common FDR thresholds of $10\%$ (0.1), $5\%$ (0.05), and $1\%$ (0.01), and due to MeCP2′s low magnitude of gene dysregulation, no fold change cutoff was used. We then examined number of contrast cutoffs of 1 (~$2.5\%$ of total contrasts), 4 (~$10\%$), 12 (~$30\%$), and 20 (~$50\%$). When examining the FDR thresholds for a specific contrast cutoff (Figure 2B and Figure S2B, per row), DEGs did not differ much. However, when comparing different numbers of contrasts for a specific FDR threshold (Figure 2B, per column), we observed large differences in DEG numbers. Moreover, when a gene is a DEG in many contrasts, it is likely to be either consistently upregulated or consistently downregulated. This indicates that the number of contrast threshold is a critical filter in establishing a robust set of DEGs. Thus, going forward, we set a strict FDR of $1\%$ and a number of contrasts cutoff ($10\%$) for common core DEGs, resulting in 2971 genes. Using average fold change, the common core DEGs were sorted into either core up [1666] or core down [1305] for further analysis. Mouse core DEG lists are in Table S2. We next investigated properties unique to the common core DEGs. First, we examined the annotations of these genes in relation to all the genes in the genome [18] and to all expressed genes. We observed that the common core DEGs have a notably high proportion of protein-coding genes. Specifically, while protein-coding genes constitute only $40.8\%$ (21,922 out of 53,661) of all the genes and $57.6\%$ (18,748 out of 32,551) of the expressed genes, they spike to $94\%$ (2794 out of 2971) in the common core DEGs (Figure 2C). Next, due to MeCP2′s role in both chromatin structure [19] and various epigenetic features [20], we examined the common core DEGs for positional bias across the mouse genome. Within each chromosome, there are differences in positional distribution between the up and down core DEGs (Figure 2D and Figure S2C). One striking example is on chromosome 8, where circular binary segmentation (CBS) [21] identified a stretch at the beginning of the chromosome (4,375,343–49,522,639) with many upregulated genes. To further examine MeCP2′s regulatory role on these common core DEGs in an unbiased manner, we carried out unsupervised Leiden clustering [22]. From the nine clusters obtained, the two largest clusters (clusters 0 and 1) consisted mainly of the consensus core up and core down genes, respectively (Figure 3A). The clear directional separation of these clusters validates our core DEG selection methodology. Subsequent gene ontology (GO) analysis of these two clusters showed an enrichment in RNA Polymerase II (Pol2) and other transcription-related terms for the upregulated cluster (cluster 0) and an enrichment in neuronal and general nervous system-related terms for the downregulated cluster (cluster 1) (Figure 3B). Both up- and downregulated clusters displayed significant enrichment for cell differentiation, signal transduction, and general developmental terms. Our computational approach therefore provides some evidence not only the roles of MeCP2 as both activator and repressor but also established the core genes and functions involved. We further examined the common core DEGs’ expression changes using a heatmap with annotation of contrast with cell fraction (Figure 3C). The unsupervised clustering shown on the heatmap categorizes the contrasts into three groups: [1] a mixture of all three types of cell fractions, [2] mainly with nucleus, and [3] mainly with whole cell. The common core DEGs are concordantly changed in about half of the contrasts, which fall into the first group of mixed-cell fractions. The two largest gene clusters identified from Leiden clustering are strongly up- and downregulated in this concordant set. Genes have lower expression changes in the second (nucleus) group. Notably, contrasts from the whole cell are categorized into two separate groups. The expression changes of common core DEGs are stronger and concordant in the first group and then weaker but still concordantly changed in the 17 contrasts of the third group. This is consistent with our findings in Figure S3, in which we observed solid overlap between common core DEGs and the common cores redefined separately by their sequenced cell fraction. Although the bulk of our data is comprised of whole cell, this picture of the transcriptome is not drastically different from MeCP2-dependent expression in the chromatin or the nucleus. ## 2.3. Cross-Species and Cross-Disease Comparisons of MeCP2′s Transcriptomic Signature As we seek to understand the broad and complex role of MeCP2, mouse data alone re insufficient. Accordingly, we uniformly processed three human datasets, yielding seven contrasts. Human data yield fewer DEGs on average than mouse data, but the DEGs have similar ranges in fold change (Figure 4A) and a similar proportion of upregulation versus downregulation. The overlap between DEGs from these ten contrasts is low (Figure S4A), suggesting limited homogeneity in the molecular signature of human MeCP2 dysregulation. This heterogeneity may reflect the fact that seven (GSE51607_1-4) out of the ten contrasts are from cell models, whereas the other three are from postmortem brain tissues. This separation is also seen in the heatmap of expression changes of human data on the mouse common core DEGs using unsupervised hierarchical clustering (Figure 4B). This heatmap shows limited qualitative correlation between the direction of human and mouse common core DEG dysregulation. We also found transcriptome data for rat (Rattus norvegicus), macaque (Macaca fascicularis), and zebrafish (Danio rerio) MeCP2 models, which we processed and compared to the mouse common core. We found two rat studies with one contrast each, one monkey study with two contrasts, and one zebrafish study with one contrast. These data yield more DEGs than the human data, with roughly even proportions of up and down gene dysregulation. After associating DEGs to mouse orthologs, we found some overlap between these cross-species models and the mouse common core, with more overlap seen between mouse and rat than with other species (Figure S4B). Figure 4B qualitatively shows a correlation between the mouse core and the rat data as well as some of the monkey data. To provide a quantitative correlation between these data sets, we performed gene set enrichment analysis (GSEA). Pre-ranked analysis was run with our up and down mouse cores as gene sets (Figure 4C). Ranked lists from the 15 non-mouse contrasts were checked for overrepresentation in both up and down cores. We expect contrasts to be positively enriched in the up core and negatively enriched in the down core. The rat model has the best enrichment concordance, with both contrasts enriched as expected. Overall, these contrasts were more significantly negatively enriched in the down core than significantly positively enriched in the up core. RTT and autism spectrum disorder (ASD) share a range of similar symptoms including loss of social, cognitive, and language skills. Altered MeCP2 expression is also commonly detected in autism brain samples [23]. Therefore, we hypothesized that our MeCP2 common core would display significant overlap with perturbed genes of other established ASD models. To generate an autism common core for comparison to our MeCP2 common core, we explored the expression changes in eight ASD models selected from the Simons Foundation Autism Research Initiative (SFARI) [24] (Figure 4D, Table S3). All experiments involving knockdown or modification of the SFARI mouse model genes were retrieved from the ARCHS4 database. Across the eight target model genes, we processed 223 samples in 18 studies from 15 authors, from which we generated 28 contrasts. In our initial attempts to generate an overall ASD core, we found that expression changes were not generally concordant across contrasts (Figure S4C), which was expected, as the contrasts contained a wide range of model genes and experimental procedures. We thus analyzed the contrasts individually and performed Fisher’s exact test to determine the significance of overlap between each contrast and the MeCP2 core. We observed significant overlap in five contrasts (Figure 4E), representing five studies and four model genes (ADNP, ARID1B, CHD8, and SHANK3). We plotted the fold change of MeCP2 core genes in these five contrasts (Figure S4E). We focused further on three of these contrasts (two CHD8 and one ADNP) with the largest DEG counts and most significant overlap and found that the significant overlap persisted even when considering only genes perturbed in the same direction in the contrast and the MeCP2 core (Figure 4F). We then carried out GO analysis on the significantly overlapping gene sets. GO analysis of genes upregulated in both the MeCP2 core and a CHD8 contrast reveals significant enrichment for Pol2-arelated terms, while downregulated genes are enriched for nervous system development terms (Figure S4D). This is consistent with our observations for up and down genes in the MeCP2 core. Even with the much smaller set of genes (~400 up and ~500 down in the overlap set compared to ~1600 up and ~1400 down in the MeCP2 core), we observed a similar gene ontology signal. ## 2.4. Sample Size Has a Major Impact on DEG Detection MeCP2 interacts with other genes in an expansive manner [25], which lends our transcriptome data a low signal-to-noise ratio (SNR). This contributes to the limited number of clear consensus expression targets as well as the high degree of discordance in many individual data sets. To increase data detection sensitivity with low SNR, the number of samples therefore plays an important role. Yet, published MeCP2 studies often fail to meet to the conventional recommendation of at least six biological replicates [26]. To learn whether this problem limited the DEGs delineated from MeCP2 studies, we performed differential gene analysis on replicate down-sampled subsets of the data set with the highest number of replicates (GSE128178 Contrast 1). We found that the number of replicates has a negative correlation with the number of DEGs detected (Figure 5A). With no fold change cutoff, we could not saturate the number of detected DEGs with as many as ten replicates, which is far more sequencing than found in most published MeCP2 datasets. With a mild fold change cutoff ($10\%$ changed), the number of additionally detected DEGs in higher sample counts still did not appear close to saturation. Some saturation and flattening of the power curve began to occur with a $20\%$ fold change cutoff. To confirm our findings, we repeated the analysis using an RNA-Seq dataset in a psoriatic skin disease model (GSE63979) [27]. We chose this dataset for its high sample size and to understand how different transcriptomic SNRs affect the optimal number of replicates. We found relatively similar patterns (Figure 5B) but a reduced DEG loss effect from fold change thresholding. This finding confirms that the impact of replicate number on detected DEG is not unique to MeCP2 or to disease models with low SNRs. Trends are also verified across common FDR thresholds, and Rand index is computed between full and down-sampled gene sets to understand how much the down-sampled results differ from the full results (Figure S5A,B). Figure S5C shows the differences between SNR in MeCP2 and psoriatic skin disease. The low power in MeCP2 data also supports our choice to require our common core DEGs to appear in just $10\%$ of the contrasts. To examine bias in and created by undetected DEGs in smaller n (sample number) studies, we first found the supersets of genes comprising each DEG cutoff number. For each DEG, we then computed average absolute log2 fold change (across contrasts). The higher n analyses detected DEGs at lower fold changes than analyses with lower sample numbers (Figure 5, last column), demonstrating that DEG sets from different sample sizes are affected differently by fold change cutoffs. A cutoff that removes many genes from a high sample size experiment may remove very few genes from a low sample size experiment. Moreover, the DEGs missed because they have too few biological replicates that are those with subtle perturbation, which is especially problematic for disease models with a low SNR. Researchers should be aware of this phenomenon when choosing fold change cutoffs and evaluating results. ## 2.5. Batch and Technical Variation Must Be Overcome in Order to Integrate and Understand Data Relevant non-biological factors, commonly called batch factors, are often present in a given researcher’s data. These batch factors could reflect the use of particular tissues, mouse litters, sequencing platforms, or other variables. Therefore, it is important to integrate and analyze transcriptomic data across years of work with dozens of meta-characteristics. We saw extreme batch effects on the raw count values for all samples included in MECP2pedia in that the samples were initially segregated by study (Figure 6A). After batch correction and normalizing the raw counts, the segregation was reduced, but samples remained grouped by study. This finding indicates that batch correction and normalization failed to fully resolve this batch effect. When we examined all available meta-characteristics, we observed that the clustering weakly overlapped with the studies’ prominent meta-characteristics, such as cell fraction, tissue, and gender (Figure S6A). As a basis for comparison, we performed the same analysis on a set of 316 samples from nine neurological degeneration studies analyzed in Wan et al. [ 28]. We observed a similar outcome: an extreme batch effect on raw data and an inability of batch effect correction and normalization algorithms to fully remove this batch effect (Figure 6A and Figure S6B). To better understand the concordance between sexes in MeCP2 models, we compared their molecular signatures. RTT occurs almost exclusively in females, but most MeCP2 studies are carried out in male mice due to their relative ease of use and availability [29]. Hence, we had only one fair comparison between male and female models of similar age and tissue (Figure 6B). We found that the DEGs from the male mouse model had no overlap with the female model at one time point and minimal overlap at a second time point. Furthermore, the direction of dysregulated genes did not show a concordant trend. More data are needed to understand these differences, but researchers should be mindful of sex in their experimental design, especially in RTT. ## 3. Discussion With the low cost and high quality of modern sequencing, the scope of publicly available data is rapidly expanding. However, to make the leap from big data to big insights, curation and automation are essential. In addition to gleaning insights from the portal’s data, researchers can compare their own novel data to this convenient aggregation of publicly available transcriptome profiles. We plan to add new datasets to the portal and also add a feature allowing users to upload their own processed data for comparison. MeCP2 transcriptomics from published studies often seem contradictory, perhaps due to low SNR and disparities in experimental design. However, by bringing a robust approach to the integration of big data, we uncovered a common core of MeCP2 DEGs with high concordance across studies, suggesting that MeCP2′s core function is universal across the examined breadth of tissues, cell fractions, mutations, and mouse strains. The positional bias in common core distribution demonstrates MeCP2′s importance to the epigenome, while unbiased clustering further underscores the concordance in core genes, providing insight into their regulatory relationship. When the clusters were enriched to GO terms associated with Pol2 activation and neuronal function, respectively, the two main clusters from the unsupervised clustering correspond to up- and downregulated genes. Exploration of the smaller mixed clusters may similarly reveal insights into other proposed mechanisms of MeCP2 action [30,31]. Examination of diverse MeCP2 models provides quantitative comparisons of MeCP2′s dysregulatory molecular signature across species. Robust enrichment across species in the downregulated DEGs supports the core we have derived, and this finding can be further explored in the context of our delineation on up versus down core genes. Comparison across disease models revealed links between MeCP2 disorders and common ASD models. Specifically, we observed highly significant concordant overlap between genes perturbed in the MeCP2 core and two ASD models: ADNP and CHD8. The RNA Pol2 function in the upregulated genes and neuronal development in the downregulated genes we observed in the MeCP2 core are also enriched genes common to MeCP2, ADNP, and CHD8 models. These overlaps could provide the basis for deeper exploration of the relationship between MeCP2 disorders and other autistic spectrum disorders. Since small sample size leads to low statistical power, the validity, specificity, and robustness of the DEGs delineated from a single study may be unreliable [26]. We validated this concern in our analysis of multiple datasets, and the problem is exacerbated when the SNR is low. Researchers who are interested in perturbations with a small effect should therefore aim to generate large datasets, with 10 or more samples per condition. If resources are limited, six samples per condition would be a good compromise. However, most studies failed to meet even this lower threshold, which may explain the lack of consensus conclusions across independent studies and their failure to capture the complete picture of transcriptomic perturbation. Another limitation of this resource is the low availability of female data, especially given Rett’s primary impact on girls. Our expansive study sheds light on the high variability in transcriptomic profiles of a disease model across different tissues, ages, sexes, species, and other biological and technical artifacts. The specific experimental conditions of a single study therefore cannot capture the complete picture of transcriptomic changes. Individual researchers will always be limited in the data that they can personally generate. Our big data integration platform solves this problem, making it invaluable for scientists studying complex diseases such as RTT and MDS. ## 4.1. Data Collection To generate this resource, we queried NCBI GEO (https://www.ncbi.nlm.nih.gov/gds) on 7 August 2019 for “MeCP2” and then filtered the results for “expression profiling by high throughput sequencing”. This search resulted in a preliminary list of 47 GEO entries. We filtered this list for entries dated 2015 or later (with an exception made for GSE51607 due to sparsity of human data) and then further filtered for entries with at least two RNA-Seq wild-type samples each and a treatment labeled either “Knock out”, “Rett”, “Point Mutation”, “Transgenic”, “Overexpression”, or “MeCP2 Duplication”. We retained 26 GEO entries. Datasets with no associated publications were included. We subsequently added three more studies, namely GSE123941, GSE128178, and GSE123372, based on the scope and relevance of their data. All preprocessing was carried out with a uniform, streamlined Python pipeline. SRA files were downloaded with prefetch from SRAtoolkit.2.9.6-1-centos_linux64 [32] and then converted to fastq with fasterq-dump version 2.3.5, using the --split-files option. Fastq files were then checked for quality with FastQC version 0.11.7 [33]. Both ArrayExpress and European Nucleotide Archive (ENA) were queried on 13 September 2022 for “MeCP2”. Identical inclusion criteria were applied, and six mouse studies comprising 11 contrasts were retained, all from ENA. These were downloaded with wget and processed identically to the GEO studies. ## 4.2. Mouse Data Processing Mouse samples were aligned to GENCODE GRCm38p6 primary assembly, version 18 (https://www.gencodegenes.org/mouse/release_M18.html, accessed 18 July 2019), with STAR version 2.6.0a [34] at default parameters. STAR gene quantifications were used (--quantMode GeneCounts). The assembly also contained an appended copy of human MeCP2 from hg38. BigWig files were generated with bamCoverage version 3.3.1 from deepTools [35]. We assessed alignment quality with RSeQC geneBody_coverage and read_distribution, both version 3.0.0. Overall quality per study was examined with MultiQC v1.7 [36]. DEG analysis was performed in R version 3.5.2 (Eggshell Igloo) with DESeq2 version 1.24.0 [37] after loose expression filtering (per contrast, a gene must have a sum of 10 counts in at least half the samples). Data were not trimmed except for samples SRR3679844, SRR3679845, SRR3679848, SRR3679849, SRR3679852, and SRR3679853 from GSE83474 due to slight anomalies in their raw sequences. The trim was performed with Trimmomatic-0.36 [38] using the following parameters: PE ILLUMINACLIP:TruSeq3-PE.fa:2:30:10 HEADCROP:8 ## 4.3. Data Annotation We downloaded SRA run tables for each GEO entry. Data characteristics of interest were: genotype, organism, experiment, run, sample name, cellular fraction, strain, age, cell line, cell type, tissue, sex, mutation, and disease. Incomplete run tables were filled in from the contents of their publications, if available. After processing, we annotated samples for sequencing depth and contrasts for number of DEG at FDR < 0.01 with no fold change cutoff. ## 4.4. Data Visualization Unless otherwise specified, plots were made in R with ggplot2 version 3.2.1 [39]. Box plot elements are as follows: minimum, first quartile, median, third quartile, and maximum. ## 4.5. Portal Development Mouse data are in the portal, while human and cross species are not. Python pipeline analysis results and GEO sample information were parsed and saved using the MongoDB NoSQL database. The web server was written in JavaScript and serves an API that gives access to the data and the web portal application. Data visualization uses the D3.js library and IGV.js [40] for genomics tracks. ENA studies are in the portal but not included in the analysis and common core. ## 4.6. Core Gene Identification and Clustering First, fold changes with FDR > 0.01 were set to zero. Contrasts with all non-significant fold changes were removed. To generate our set of core genes, we kept only genes with non-zero fold change in four or more contrasts ($10\%$). This resulted in a set of 2971 genes for further analysis. For consistency in the analysis and the visualization of gene regulation direction, we inverted the direction of fold change for the four contrasts of the MDS model (GSE123372_3, GSE66870_2, GSE71235_1, GSE71235_2). We also identified alternative core genes based on contrast metadata characteristics. For each type of cell fraction (chromatin, nucleus, whole cell) and the cortex and forebrain tissue types, we performed the same core gene identification filtering as we did for the main core. We considered only the contrasts with the metadata feature of interest and kept only genes with significant and non-zero fold change in at least $10\%$ of the contrasts. We included an upset plot generated with the R package UpSetR [41] as Figure S3. After identifying the set of core genes, we assigned unsupervised clusters using the Scanpy [42] implementation of the Leiden algorithm with the parameters: number of neighbors = 45 and resolution = 0.5. Then, we generated UMAP coordinates with the parameters: number of neighbors = 45, minimum distance = 0.1, and spread = 10. We then used UMAP’s Scanpy implementation to generate plots of the unsupervised clusters as well as up- and downregulation. ## 4.7. Core Gene Characteristics and Location We annotated mouse genes with GRCm38p6 primary, version 18 from GENCODE. Genes were sorted into eight super-categories to show broad function. Expressed genes [32,539] is a superset of the genes that pass the expression filter in any contrast. Core genes were plotted by their TSS (Transcription start site). Chromosome 8 was plotted with an equivalent number of randomly drawn non-core genes [150] to show the strength of its regional core up DEG enrichment. We validated this trend with the CBS algorithm implemented with R Package PSCBS version 0.65.0. The core up, core down, and non-core genes were, respectively, assigned values of 6, 0, and 3 for segmentation detection and plotting. ## 4.8. GO Analysis GO analysis was performed using the Python GOAtools package [43]. We performed an enrichment analysis on each MeCP2 Leiden cluster, using all NCBI protein coding mouse genes as the background set. For each Leiden cluster, we retained the top six biological process GO terms by frequency. We used the same methodology to perform GO enrichment on overlapping MeCP2 and ASD core genes, also retaining the top six biological process GO terms by frequency. ## 4.9. Human Data Processing and Comparative Analysis Human data were aligned to GRCh38p12 primary assembly, version 28 from GENCODE with STAR. BigWig generation, assembly quality metrics, and DEG analysis were performed identically on mouse data. Human genes were then queried for their orthology to mouse genes with DIOPT 8.0 [44] using the “return only best match” option. Upset plots were made with function upset from Package UpSetR, version 1.4.0. Human and other metadata are available in Table S5. ## 4.10. Other Model Data Processing and Comparative Analysis Rat data were aligned to the Rnor_6.0 toplevel assembly and annotated with Rnor_6.0.99 from Ensembl [45]. Zebrafish data were aligned to Danio_rerio. GRCz11 toplevel assembly and annotated with Danio_rerio. GRCz11.100 from Ensembl. The orthology tables for rat and zebrafish genes were retrieved from DIOPT, as was done for human data. Macaque data were aligned to Macaca_fascicularis_5.0 and annotated with Macaca_fascicularis_5.0.100 from Ensembl. Macaque gene orthology data were retrieved with the function getBM from R package biomaRt, version 2.38.0 [46]. The only data trimmed were GSE57974, using Trimmomatic-0.36 with the following parameters: LEADING:3 TRAILING:20 MINLEN:50. Genotype labels for GSE87855 were inferred based on MeCP2 level. Heatmaps were made in R with pheatmap package version 1.0.12 [47] using log2 fold change, no clustering on rows, and clustering_distance_cols = “euclidean”. For consistency in the analysis and the visualization of gene regulation direction, we inverted the direction of log2 fold change for the MDS contrast (GSE57974_1) ## 4.11. GSEA GSEA [48,49] version 4.1.0 Pre-Ranked was run with default parameters besides set_max of 100,000 and set_min of 1. Ranking values were computed per gene as –log10(adjusted p value) * log2 fold change. For consistency in the analysis and the visualization of gene regulation direction, we inverted the direction of normalized enrichment score fold change for the MDS contrast (GSE57974_1). GSEA results are available as Table S6. ## 4.12. ASD Model Comparison All experiments involving knockdown or modification of the SFARI mouse model genes were retrieved from the ARCHS4 database. There were 18 such studies, which we processed using DESeq2 to generate DEGs across 28 contrasts. We retained only DEGs with FDR < 0.01. With the DEGs for each ASD contrast, we used the hypergeometric and Fisher’s exact tests to determine the significance of overlap with the set of all the MeCP2 core genes and also both up- and downregulated subsets. Computed ASD contrast fold change is available as Table S4. We also used the previously described process to perform GO analysis on genes in the intersection of the ASD contrasts and MeCP2 cores. ## 4.13. Down Sampling Analysis MeCP2 data are from GSE128178. All 10 samples of wild-type and knockout whole-cell data were randomly selected to create 100 random drawings at each different sample number. ( For instance, sample 9 was one random sample removed from each genotype, and so on.) Once selected, samples were normalized with each other and analyzed with the same DEG methodology detailed above. The Rand index was then computed using the rand.index function in R (fossil, version 0.4.0) [50]. Vectors indicating if each gene was a DEG for a particular run were compared to the vectors of DEGs for the complete contrast of 10 wild-type and 10 knockout samples, respectively. Psoriatic skin data are from GSE63979 (SRP050971). This study contains the total RNA-Seq data of nine normal skin samples, nine lesional psoriatic samples, and twenty-seven uninvolved psoriatic samples. In order to conduct the downsampling analysis between two groups with the same sample number, only normal skin samples and lesional psoriatic samples were chosen. For each phenotype, all nine samples were randomly selected to create 100 random drawings at each different sample number. For instance, sample 8 was one random sample removed from each phenotype, and so on. The DEG analysis and Rand index comparison on psoriatic skin data was the same used for MeCP2 data. ## 4.14. Technical Variation/Batch Effect Analysis All raw MeCP2 mouse expression value data were dimensionally reduced using UMAP version 0.2.6.0 in R and plotted with color for contrast of origin. ComBat_seq [51] from R package sva, version 3.36.0, was then run with contrast as batch to deconvolute the data. ComBat_seq-normalized data were then size factor-normalized with DESeq2 and plotted again. To provide an alternate dataset to evaluate batch effect, we used a set of Alzheimer’s disease data stored on Synapse from Wan et al. [ 28]. We retrieved raw count data and plotted the samples using UMAP. Count data were normalized with DESeq2, and then, we used the ComBat_seq function to attempt to control for batch effect. ## 4.15. Sex Comparison The compared contrasts are GSE90736_1, GSE90736_2, and GSE66211_1. Genes were considered DEGs if they passed FDR < 0.01. The plot was made with R package VennDiagram version 1.6.20. ## References 1. Costa F.F.. Big data in biomedicine. Drug Discov. Today (2014.0) 19 433-440. DOI: 10.1016/j.drudis.2013.10.012 2. Lonsdale J., Thomas J., Salvatore M., Phillips R., Lo E., Shad S., Hasz R., Walters G., Garcia F., Young N.. The Genotype-Tissue Expression (GTEx) project. Nat. Genet. (2013.0) 45 580-585. DOI: 10.1038/ng.2653 3. Lachmann A., Torre D., Keenan A.B., Jagodnik K.M., Lee H.J., Wang L., Silverstein M.C., Ma’ayan A.. Massive mining of publicly available RNA-seq data from human and mouse. Nat. Commun. (2018.0) 9 1366. DOI: 10.1038/s41467-018-03751-6 4. Drysdale R.A., Crosby M.A.. FlyBase Consortium. FlyBase: Genes and gene models. Nucleic Acids Res. (2005.0) 33 D390-D395. DOI: 10.1093/nar/gki046 5. Smith J.R., Hayman G.T., Wang S.J., Laulederkind S., Hoffman M.J., Kaldunski M.L., Tutaj M., Thota J., Nalabolu H.S., Ellanki S.. The Year of the Rat: The Rat Genome Database at 20: A multi-species knowledgebase and analysis platform. Nucleic Acids Res. (2020.0) 48 D731-D742. DOI: 10.1093/nar/gkz1041 6. Tomczak K., Czerwińska P., Wiznerowicz M.. The Cancer Genome Atlas (TCGA): An immeasurable source of knowledge. Contemp. Oncol. (2015.0) 19 A68-A77. DOI: 10.5114/wo.2014.47136 7. Cerami E., Gao J., Dogrusoz U., Gross B.E., Sumer S.O., Aksoy B.A., Jacobsen A., Byrne C.J., Heuer M.L., Larsson E.. The cBio cancer genomics portal: An open platform for exploring multidimensional cancer genomics data. Cancer Discov. (2012.0) 2 401-404. DOI: 10.1158/2159-8290.CD-12-0095 8. Hodes R.J., Buckholtz N.. Accelerating medicines partnership: Alzheimer’s disease (AMP-AD) knowledge portal aids Alzheimer’s drug discovery through open data sharing. Expert Opin. Ther. Targets (2016.0) 20 389-391. DOI: 10.1517/14728222.2016.1135132 9. Amir R.E., Van den Veyver I.B., Wan M., Tran C.Q., Francke U., Zoghbi H.Y.. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat. Genet. (1999.0) 23 185-188. DOI: 10.1038/13810 10. Sandweiss A.J., Brandt V.L., Zoghbi H.Y.. Advances in understanding of Rett syndrome and MECP2 duplication syndrome: Prospects for future therapies. Lancet. Neurol. (2020.0) 19 689-698. DOI: 10.1016/S1474-4422(20)30217-9 11. Sampieri K., Meloni I., Scala E., Ariani F., Caselli R., Pescucci C., Longo I., Artuso R., Bruttini M., Mencarelli M.A.. Italian Rett database and biobank. Hum. Mutat. (2007.0) 28 329-335. DOI: 10.1002/humu.20453 12. Percy A.K., Lane J.B., Childers J., Skinner S., Annese F., Barrish J., Caeg E., Glaze D.G., MacLeod P.. Rett syndrome: North American database. J. Child Neurol. (2007.0) 22 1338-1341. DOI: 10.1177/0883073807308715 13. Krishnaraj R., Ho G., Christodoulou J.. RettBASE: Rett syndrome database update. Hum. Mutat. (2017.0) 38 922-931. DOI: 10.1002/humu.23263 14. Edgar R., Domrachev M., Lash A.E.. Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. (2002.0) 30 207-210. DOI: 10.1093/nar/30.1.207 15. RPKM, FPKM and TPM, Clearly Explained: RNA-Seq Blog. Rna-Seqblog. (2015.0) 16. Chahrour M., Jung S.Y., Shaw C., Zhou X., Wong S.T., Qin J., Zoghbi H.Y.. MeCP2, a key contributor to neurological disease, activates and represses transcription. Science (2008.0) 320 1224-1229. DOI: 10.1126/science.1153252 17. Al-Ouran R., Wan Y.W., Mangleburg C.G., Lee T.V., Allison K., Shulman J.M., Liu Z.. A Portal to Visualize Transcriptome Profiles in Mouse Models of Neurological Disorders. Genes (2019.0) 10. DOI: 10.3390/genes10100759 18. Frankish A., Diekhans M., Ferreira A.M., Johnson R., Jungreis I., Loveland J., Mudge J.M., Sisu C., Wright J., Armstrong J.. GENCODE reference annotation for the human and mouse genomes. Nucleic Acids Res. (2019.0) 47 D766-D773. DOI: 10.1093/nar/gky955 19. Gulmez Karaca K., Brito D., Oliveira A.. MeCP2: A Critical Regulator of Chromatin in Neurodevelopment and Adult Brain Function. Int. J. Mol. Sci. (2019.0) 20. DOI: 10.3390/ijms20184577 20. Picard N., Fagiolini M.. MeCP2: An epigenetic regulator of critical periods. Curr. Opin. Neurobiol. (2019.0) 59 95-101. DOI: 10.1016/j.conb.2019.04.004 21. Olshen A.B., Venkatraman E.S., Lucito R., Wigler M.. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics (2004.0) 5 557-572. DOI: 10.1093/biostatistics/kxh008 22. Traag V.A., Waltman L., van Eck N.J.. From Louvain to Leiden: Guaranteeing well-connected communities. Sci. Rep. (2019.0) 9 5233. DOI: 10.1038/s41598-019-41695-z 23. Nagarajan R., Hogart A., Gwye Y., Martin M.R., LaSalle J.M.. Reduced MeCP2 expression is frequent in autism frontal cortex and correlates with aberrant MECP2 promoter methylation. Epigenetics (2006.0) 1 172-182. DOI: 10.4161/epi.1.4.3514 24. Siegel M., Smith K.A., Mazefsky C., Gabriels R.L., Erickson C., Kaplan D., Morrow E.M., Wink L., Santangelo S.L.. The autism inpatient collection: Methods and preliminary sample description. Mol. Autism (2015.0) 6 61. DOI: 10.1186/s13229-015-0054-8 25. Tillotson R., Bird A.. The Molecular Basis of MeCP2 Function in the Brain. J. Mol. Biol. (2019.0) S0022-2836(19)30595-9. DOI: 10.1016/j.jmb.2019.10.004 26. Schurch N.J., Schofield P., Gierliński M., Cole C., Sherstnev A., Singh V., Wrobel N., Gharbi K., Simpson G.G., Owen-Hughes T.. How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use?. RNA (2016.0) 22 839-851. DOI: 10.1261/rna.053959.115 27. Greenberg E.N., Marshall M.E., Jin S., Venkatesh S., Dragan M., Tsoi L.C., Gudjonsson J.E., Nie Q., Takahashi J.S., Andersen B.. Circadian control of interferon-sensitive gene expression in murine skin. Proc. Natl. Acad. Sci. USA (2020.0) 117 5761-5771. DOI: 10.1073/pnas.1915773117 28. Wan Y.W., Al-Ouran R., Mangleburg C.G., Perumal T.M., Lee T.V., Allison K., Swarup V., Funk C.C., Gaiteri C., Allen M.. Meta-Analysis of the Alzheimer’s Disease Human Brain Transcriptome and Functional Dissection in Mouse Models. Cell Rep. (2020.0) 32 107908. DOI: 10.1016/j.celrep.2020.107908 29. Ribeiro M.C., MacDonald J.L.. Sex differences in Mecp2-mutant Rett syndrome model mice and the impact of cellular mosaicism in phenotype development. Brain Res. (2020.0) 1729 146644. DOI: 10.1016/j.brainres.2019.146644 30. Lyst M.J., Bird A.. Rett syndrome: A complex disorder with simple roots. Nat. Rev. Genet. (2015.0) 16 261-275. DOI: 10.1038/nrg3897 31. Ip J., Mellios N., Sur M.. Rett syndrome: Insights into genetic, molecular and circuit mechanisms. Nat. Rev. Neurosci. (2018.0) 19 368-382. DOI: 10.1038/s41583-018-0006-3 32. Sherry S., Xiao C., Durbrow K., Kimelman M., Rodarmer K., Shumway M., Yaschenko E.. NCBI sra toolkit technology for next generation sequence data. Proceedings of the Plant and Animal Genome XX Conference 33. Andrews S.. FastQC: A Quality Control Tool for High Throughput Sequence Data. (2010.0) 34. Dobin A., Davis C.A., Schlesinger F., Drenkow J., Zaleski C., Jha S., Batut P., Chaisson M., Gingeras T.R.. STAR: Ultrafast universal RNA-seq aligner. Bioinformatics (2013.0) 29 15-21. DOI: 10.1093/bioinformatics/bts635 35. Ramírez F., Ryan D.P., Grüning B., Bhardwaj V., Kilpert F., Richter A.S., Heyne S., Dündar F., Manke T.. deepTools2: A next generation web server for deep-sequencing data analysis. Nucleic Acids Res. (2016.0) 44 W160-W165. DOI: 10.1093/nar/gkw257 36. Ewels P., Magnusson M., Lundin S., Käller M.. MultiQC: Summarize analysis results for multiple tools and samples in a single report. Bioinformatics (2016.0) 32 3047-3048. DOI: 10.1093/bioinformatics/btw354 37. Love M.I., Huber W., Anders S.. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. (2014.0) 15 550. DOI: 10.1186/s13059-014-0550-8 38. Bolger A.M., Lohse M., Usadel B.. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics (2014.0) 30 2114-2120. DOI: 10.1093/bioinformatics/btu170 39. Wickham H.. ggplot2: Elegant Graphics for Data Analysis (2016.0) 40. Robinson J.T., Thorvaldsdóttir H., Turner D., Mesirov J.P.. igv.js: An embeddable JavaScript implementation of the Integrative Genomics Viewer (IGV). Bioinformatics (2023.0) 39 btac830. DOI: 10.1093/bioinformatics/btac830 41. Conway J.R., Lex A., Gehlenborg N.. UpSetR: An R Package for the Visualization of Intersecting Sets and their Properties. Bioinformatics (2017.0) 33 2938-2940. DOI: 10.1093/bioinformatics/btx364 42. Wolf F.A., Angerer P., Theis F.J.. SCANPY: Large-scale single-cell gene expression data analysis. Genome Biol. (2018.0) 19 15. DOI: 10.1186/s13059-017-1382-0 43. Klopfenstein D.V., Zhang L., Pedersen B.S., Ramírez F., Warwick Vesztrocy A., Naldi A., Mungall C.J., Yunes J.M., Botvinnik O., Weigel M.. GOATOOLS: A Python library for Gene Ontology analyses. Sci. Rep. (2018.0) 8 10872. DOI: 10.1038/s41598-018-28948-z 44. Hu Y., Flockhart I., Vinayagam A., Bergwitz C., Berger B., Perrimon N., Mohr S.E.. An integrative approach to ortholog prediction for disease-focused and other functional studies. BMC Bioinform. (2011.0) 12. DOI: 10.1186/1471-2105-12-357 45. Yates A.D., Achuthan P., Akanni W., Allen J., Allen J., Alvarez-Jarreta J., Amode M.R., Armean I.M., Azov A.G., Bennett R.. Ensembl 2020. Nucleic Acids Res. (2020.0) 48 D682-D688. DOI: 10.1093/nar/gkz966 46. Durinck S., Spellman P.T., Birney E., Huber W.. Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat. Protoc. (2009.0) 4 1184-1191. DOI: 10.1038/nprot.2009.97 47. Kolde R., Kolde M.R.. Package ‘pheatmap’. R Package (2015.0) Volume 1 790 48. Subramanian A., Tamayo P., Mootha V.K., Mukherjee S., Ebert B.L., Gillette M.A., Paulovich A., Pomeroy S.L., Golub T.R., Lander E.S.. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA (2005.0) 102 15545-15550. DOI: 10.1073/pnas.0506580102 49. Mootha V., Lindgren C., Eriksson K.-F., Subramanian A., Sihag S., Lehar J., Puigserver P., Carlsson E., Ridderstråle M., Laurila E.. PGC-1α-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet (2003.0) 34 267-273. DOI: 10.1038/ng1180 50. Vavrek M.J.. Fossil: Palaeoecological and Palaeogeographical Analysis Tools. Palaeontologia Electronica. R package Version 0.4.0. (2011.0) 14 1T 51. Zhang Y., Parmigiani G., Johnson W.E.. NAR Genom. Bioinform. (2020.0) 2 lqaa078. DOI: 10.1093/nargab/lqaa078
--- title: Dietary Supplementation of Microbial Dextran and Inulin Exerts Hypocholesterolemic Effects and Modulates Gut Microbiota in BALB/c Mice Models authors: - Iqra Jawad - Husam Bin Tawseen - Muhammad Irfan - Waqar Ahmad - Mujtaba Hassan - Fazal Sattar - Fazli Rabbi Awan - Shazia Khaliq - Nasrin Akhtar - Kalsoom Akhtar - Munir Ahmad Anwar - Nayla Munawar journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049499 doi: 10.3390/ijms24065314 license: CC BY 4.0 --- # Dietary Supplementation of Microbial Dextran and Inulin Exerts Hypocholesterolemic Effects and Modulates Gut Microbiota in BALB/c Mice Models ## Abstract Microbial exopolysaccharides (EPSs), having great structural diversity, have gained tremendous interest for their prebiotic effects. In the present study, mice models were used to investigate if microbial dextran and inulin-type EPSs could also play role in the modulation of microbiomics and metabolomics by improving certain biochemical parameters, such as blood cholesterol and glucose levels and weight gain. Feeding the mice for 21 days on EPS-supplemented feed resulted in only 7.6 ± $0.8\%$ weight gain in the inulin-fed mice group, while the dextran-fed group also showed a low weight gain trend as compared to the control group. Blood glucose levels of the dextran- and inulin-fed groups did not change significantly in comparison with the control where it increased by 22 ± $5\%$. Moreover, the dextran and inulin exerted pronounced hypocholesterolemic effects by reducing the serum cholesterol levels by $23\%$ and $13\%$, respectively. The control group was found to be mainly populated with Enterococcus faecalis, Staphylococcus gallinarum, Mammaliicoccus lentus and Klebsiella aerogenes. The colonization of E. faecalis was inhibited by 59–$65\%$ while the intestinal release of Escherichia fergusonii was increased by 85–$95\%$ in the EPS-supplemented groups, respectively, along with the complete inhibition of growth of other enteropathogens. Additionally, higher populations of lactic acid bacteria were detected in the intestine of EPS-fed mice as compared to controls. ## 1. Introduction Gastrointestinal microbiota is recognized as a “functional organism” as it provides various immunological, defensive, structural, and metabolic benefits to the host as a result of host-microbial interaction by creating an efficient ecosystem and ensuring body homeostasis [1,2]. The pattern of the gut microbiome in different individuals is influenced by age, genetics, medication, infection, personal hygiene, allergens contact, and most importantly diet and the use of probiotics and prebiotics [3,4,5]. Probiotics and prebiotics collectively fall in the category of functional foods. Food is considered functional if along with providing nutritional value, it proves beneficial for the well-being and health of an individual and reduces the risks of diseases [6]. The functionality of functional foods is constituted by certain bioactive components present in them. Among these components, microbial exopolysaccharides (EPSs) such as fructans and α-glucans have gained much research focus in terms of determining their role in improving host health. Fructans and α-glucans are homopolysaccharides comprising of single type of monomer, i.e., fructose or glucose, respectively. The diversity in these polysaccharides can be attributed to the fact that the constituent monomeric sugar units are linked through different types of glycosidic bonds in the backbone structure with varied branching patterns. Based on glycosidic linkage types, fructans are categorized into two major groups. These include levan and inulin, having β-(2→6) and β-(2→1) linkages among their main chain glycosidic units, respectively, with varying degrees of branching. Similarly, α-glucans are classified as dextran having α (1→6) linkages in the main chain with α-(1→3) linked branches, mutan having α-(1→3) linkages with α-(1→6) linked branches, reuteran having α-(1→4) linkages and alternan with α-(1→6)/α-(1→3) linkages between the D-glucose monomers. Fructans are present in bacteria, fungi, archaea and plants [7,8] while α-glucans are exclusively produced by lactic acid bacteria (LAB) [9]. Prebiotic exopolysaccharides (EPSs) are not hydrolyzed by digestive enzymes of humans and non-ruminant animals lacking carbohydrate active enzymes (CAZymes), but undergo bacterial fermentation in the distal part of gut by the CAZymes of the gut microbiome [10,11]. They interact with mucosa and microflora to regulate “gut health”. The enrichment of the population of advantageous bacteria by fermenting prebiotic exopolysaccharides not only enhances their ability to compete with the potential pathogens for the available nutrients but also produce short chain fatty acids that inhibit pathogen proliferation by lowering the pH of the lumen leading towards the competitive exclusion of pathogens by preventing their attachment to epithelial cells [12]. Hence prebiotics enhance the growth of beneficial bacteria that exhibit quorum sensing, competition for resources, adherence properties and possession of different metabolic pathways, leading to the eradication of the less desirable members of the intestinal microflora [13,14]. Besides the above-mentioned beneficial aspects of polysaccharides on the gut microbiome, a few studies have reported their impact on health metabolomics indicators and biochemical parameters, such as blood glucose and/or cholesterol levels. Notably, some investigators reported a significant reduction in serum cholesterol concentration in inulin-fed mice [15,16] while others observed no significant difference in serum cholesterol levels between inulin-fed model organisms (rats or mice) and controls [17]. It may be noted that mostly inulin extracted from plants has been used in these studies. A few investigators have used microbial levan in their studies and reported its cholesterol lowering effects on rats [18,19]. Thus, paucity of data on health beneficial effects of microbial EPSs and contradictory results accentuates further experimentation in this direction. In this paper, we have compared the effects of two different types of bacterial EPS, i.e., inulin and dextran, on serum glucose and cholesterol levels of mice. Further, we have also investigated their effects on the populations of culturable lactic acid bacteria and some common enteric pathogens in the mice gut. To the best of our knowledge, this is the first in vivo study on microbially produced EPSs to elucidate their metabolomic and microbiomic impacts in the gastrointestinal tract using mice models. ## 2.1. Microbial Synthesis of Exopolysaccharides Both polymers were synthesized under already optimized conditions to acquire the maximum yield. In physical appearance, inulin is non-sticky and has a white to pale tint whereas dextran has a white color and a sticky texture. Both of these polymers were obtained in fine powder form after freeze drying. In order to avoid the presence of any residual sucrose, the purity status of the synthesized polymers was confirmed by thin layer chromatography. After developing the TLC plates with urea developing solution (for inulin) and methanol developing solution (for dextran), no degradation or additional spots were observed on the TLC plate except that of polymers, confirming that polymers were in pure intact form and free of other substances such as residual sucrose and glucose (Figure 1). ## 2.2. Body Weight Analysis As already mentioned, the body weights of all the mice of three groups were determined on Day 0 and Day 21 of the experiment. Figure 2 shows the weight gain as percentage of their initial body weights during the course of this experiment on supplementation with EPS ($5\%$ w/w) in their feed. The initial weights among the groups were not significantly different. After 21 days of feeding, 19 ± $2.3\%$ gain in total body weight was observed in the control group. On the contrary, only 7.6 ± $0.8\%$ weight gain was observed in the inulin fed mice group. Though weight gain in the dextran group was not considered statistically significant ($p \leq 0.05$), it showed a lower (15.7 ± $1.8\%$) trend in comparison with the control group. ## 2.3. Blood Glucose Analysis Blood glucose levels of the control group of mice showed a significant increase (22 ± $5\%$) over a period of 21 days (Figure 3). Contrarily, no significant change in the blood glucose levels of the experimental groups was observed after 21 days of feeding on dextran- or inulin-supplemented feed, as compared to those on Day 0. Further, the blood glucose levels of both experimental groups were found to be significantly lower than those of the control group at the end of the experiment (day 21). ## 2.4. Serum Cholesterol Analysis Serum cholesterol levels of mice from all the diet groups were measured on the day of euthanasia (i.e., day 21). It was observed that after feeding the mice with inulin- or dextran-supplemented diet, their cholesterol level was reduced as compared to the control group. Dextran exhibited a more pronounced hypocholesterolemic effect as it reduced the serum cholesterol level by $23\%$ while inulin reduced it by $13\%$, compared to the control group (Table 1). ## 2.5. Fecal Lipid Analysis Analysis of total lipids in the fecal samples collected during the last 48 h of the experiments showed that feeding the mice with dextran had no significant effect on the amount of excreted lipids as compared to the control group (Table 1). However, feeding with inulin-supplemented feed resulted in a profound increase (about $78\%$) in the concentration of total lipids excreted in feces, in comparison with the samples from the control group. ## 2.6. Organ Weight Analysis After euthanasia of the mice on day 21 of the experiment, weights of their different body organs were determined. These organs included liver, kidneys, pancreas, lungs, heart, stomach and spleen. Comparison of the weights of these organs from inulin- and dextran-fed groups with the control group showed no significant difference (Table 2). ## 2.7. Analysis of Intestinal Microbiota of Mice Multi-colored colonies with various morphological appearances were obtained after a 16-h incubation period at 37 °C, when the intestinal fluid of mice from different treatment groups was spread on CHROMagar plates. The CHROMagar color orientation provided by the manufacturer was used for preliminary identification of the colonies. CHROMagar colony data showed that the microbiota in the intestinal fluid of the control group of mice fed with normal diet were mainly populated by Enterococcus (turquoise blue colored colonies) along with a significant population of Klebsiella (metallic blue colored colonies), Staphylococcus (cream-colored colonies) and Mammaliicoccus (sky blue colored colonies) species. A few pink colonies that belonged to Escherichia sp. were also seen in the intestinal microbiota of control mice. In contrast, microbiota in the intestinal mucosa (scraped material) exhibited a significantly higher population of Escherichia sp. along with all other bacteria found in the samples of the same control animals, raising the likelihood of the attachment of Escherichia sp. to intestinal epithelial cells. Interesting differences were found in the microbiota of dextran and inulin treatment groups of mice. The CHROMagar plates inoculated with intestinal fluid of dextran and inulin fed mice were dominantly populated with Escherichia sp. along with significantly lower Enterococcus population as compared to intestinal microbiota of the control group. No additional colored colonies were seen in the microbiota of both experimental groups of mice, indicating that feeding on the inulin- or dextran-supplemented feed had inhibited the growth of Klebsiella, Mammalicoccus and Staphylococcus species compared to the microbiota of control group (Figure S1). The role of microbial dextran and inulin in the eradication of Escherichia sp. was further confirmed by the results of MacConkey agar analysis which showed the similar microbial pattern as exhibited by CHROMagar analysis. It was observed that the plates dispersed with intestinal fluid of the control group were predominantly occupied by red and pale-pink colored colonies indicating the presence of Enterococcus and Staphylococcus, respectively. In contrast, the plates spread with intestinal fluid of dextran and inulin treatment mice groups were predominantly occupied by non-mucoid pink colored colonies indicating the presence of Escherichia species (Figure S1). These results strengthened the evidence that microbial dextran and inulin could play a significant role in the detachment of Escherichia sp. from intestinal epithelial cells and enhance the eradication of intestinal pathogenic bacteria. In addition to the eradication of pathogens, our quest was to investigate the effect of microbial dextran and inulin supplementation on the proliferation of beneficial microbiota in the gastrointestinal tract of mice. For this purpose, the intestinal fluids of all groups of mice were spread on MRS-agar plates supplemented with $2\%$ glucose or $20\%$ sucrose. After overnight incubation, a smaller number of colonies appeared on the plates of the control group but a significant population was observed on the plates spread with intestinal fluid of dextran- or inulin-fed groups of mice. Some results were surprisingly different in the intestinal fluid of the experimental group of mice analyzed on MRS-glucose plates as we observed large milky white colonies on these plates. These colonies were further grown and purified in YPD (yeast extract peptone dextrose) medium supplemented with kanamycin. When observed under the microscope, these white colored colonies appeared to be composed of yeast cells (Figure S2). These results indicated that the exopolysaccharides used in this study could have a prebiotic role exhibiting strong potential to increase the population of beneficial microbiota by reducing the pathogenic bacterial species. ## 2.8. Molecular Identification and Phylogenetic Analysis of Intestinal Microbial Isolates All morphologically distinct colonies from each plate were selected, purified and subjected to DNA extraction followed by 16S rRNA gene sequencing. Each bacterial colony’s 16S rRNA gene was PCR amplified using FD1 and RP1 primers, producing a 1500 bp-sized product. The Blastn program, offered by NCBI, was used to analyze the sequencing data of the PCR-amplified genes and reveal their percentage identity with other closely related strains. According to these results, the turquoise blue colony (labelled as IJ4) from the control group exhibited $100\%$ identity with Enterococcus faecalis strain NBRC 100481 whereas the golden opaque (labelled as IJ5), light blue (Labelled as IJ9) and metallic blue (labelled as IJ11) colored colonies from the control group showed $99.8\%$, $99.9\%$ and $99.68\%$ identity with Staphylococcus gallinarum strain VIII1, Mammaliicoccus lentus strain MAFF 911385 and Klebsiella aerogenes strain NBRC 13534, respectively. The Blastn of the 16S rRNA gene for pink colored colonies from experimental groups showed that these colonies were $99.49\%$ identical with Escherichia fergusonii strain NBRC 102419. This homology was further confirmed through phylogenetic analysis by phylogenetic tree that was constructed using the MEGA11 program employing neighbor joining algorithm. All of the isolates share their cluster with most identical strains as depicted by Blastn results (Figure 4). Using 16S rRNA gene sequence homology, isolates from MRS agar plates inoculated with intestinal fluids from dextran and Inulin-fed groups were also identified. According to Blastn results, the intestinal fluid of these mice contained three different types of bacteria in a dominant proportion labeled as IJ6, IJ7 and IQ4 having $98.11\%$, $99.80\%$ and $98.64\%$ homology with Lactococcus garvieae strain NIZO2415T, Bacillus subtilis strain BCRC 10255 and Bacillus licheniformis strain ATCC 14580, respectively. The phylogenetic analysis revealed that these microbes cluster closely with the above said respective closely related bacterial species (Figure 5). All the 16S rRNA gene sequences obtained in this study were submitted to NCBI and their accession numbers were obtained as given in Table 3. ## 2.9. Comparative Analysis of Intestinal Microbial Populations among Different Treatment Groups The comparative analysis of culturable microbial populations among different treatment groups showed that the population of Enterococcus was significantly inhibited down to $59\%$ and $65\%$ in intestinal fluids of dextran and inulin-fed mice, respectively, as compared to the control. Similarly, the release of intestinal E. fergusonii population in dextran and inulin-fed mice enhanced their levels up to $89\%$ and $95\%$, respectively, as compared to the control. The occurrence of a significant population of E. fergusonii in intestinal mucosa samples of the control also showed that both polymers trigger the detachment and release of E. fergusonii from the intestinal epithelial cells. The populations of K. aerogens, S. gallinarium and M. lentus were completely inhibited in the experimental groups demonstrating the effectiveness of microbially derived exopolysaccharides in maintaining gastrointestinal health by inhibiting the growth of enteric pathogens (Figure 6). The statistical analysis also revealed a significant difference in the number of colony forming units (CFU) of Enterococcus and E. fergusonii in dextran and inulin-fed groups as compared to the control (Table 4). ## 3. Discussion In spite of considerable information available on the synthesis and characterization of microbial EPSs, their experimental usage, especially in regard to biomedical applications, is limited only to plant-based inulin, fructo-oligosaccharide and some β-glucans. Recently, there has emerged a great deal of interest in the ability of EPSs of microbial origin to confer potential health benefits in animals and humans. In this context, levan type polysaccharide of bacterial origin has been shown to exert hypocholesterolemic effects in experimental mice [19]. In the present study, we have used two different types of microbial EPSs, i.e., dextran and inulin, in experimental mice models to investigate their hypoglycemic and hypocholesterolemic potentials. The experimental mice were fed with $5\%$ (w/w) EPSs, as this concentration has been found to be optimum for a noticeable hypocholesterolemic effect in the previous study by Yamamoto et al. [ 19]. Considerably less weight gain was observed in the mice fed on the EPS-supplemented diets in comparison to the control group mice. These results are contradictory to those reported by Yamamoto et al. [ 19], where levan was used as diet supplement for rats. In that study, no significant differences in the body weight gain were found between levan-fed rats and the control rats over the experimental period. Our results support the fact that upon consumption of prebiotic EPSs in their diet, the body weight gain of mice was controlled. This control in body weight could be explained on the basis of some previously reported mechanisms. One such mechanism is the potential reduction in the amount of calories obtained via fermentation of these EPS and the consequent absorption and metabolism of their fermented products, i.e., short chain fatty acids (SCFAs) [20]. The calories obtained are reported to be of less than half of the caloric value gained from direct metabolism of a similar amount of carbohydrates [21]. Some studies suggest a more complex phenomenon for this reduced caloric value by highlighting the variation of gut microbiota and its influence on harvesting energy from EPS [22]. It is also a well-known concept that differences in dietary patterns could lead to variation of the gut microbiota [23]. Thus, it can also be assumed that in the present study, mice fed with EPS developed variations in their microbiota composition compared to the control group and the resulting gut microbial diversity could be responsible for influencing dietary intake (satiety) and energy expenditure. Another explanation for this appetite regulation is the interaction of SCFAs with free fatty acid receptors FFA2 and FF3 which leads to the varied expression of hormones and peptides, influencing the energy metabolism [24]. Many natural polysaccharides are known to be potential anti-diabetic agents [25]. A majority of polysaccharides with anti-diabetic activity have yet not been validated scientifically due to the complexities in their structure and their undefined mechanisms. Further research on the potential anti-diabetic mechanisms of polysaccharides is the need of the hour. In the present studies, both the inulin and dextran were found to exert significant hypoglycemic effects on mice. In these studies, no significant change in the blood glucose level of inulin- and dextran-fed mice was observed over the experimental time span. One possible explanation for this hypoglycemic effect could be that the EPSs shorten gastric emptying time and reduce the transit time of the small intestine which in turn prevents postprandial elevations of blood glucose [26]. Another possible mechanism of controlling blood glucose level could be the influence of SCFAs on host metabolism as stated earlier. For instance, propionate has been found to upregulate hepatic glycolysis and downregulate hepatic gluconeogenesis. Similarly, butyrate in the colonic lumen enhances the expression of an anti-hyperglycemic hormone called glucagon-like peptide-1 (GLP-1) which is believed to promote pancreatic beta cell proliferation, thus increasing the amount of insulin secretion [27]. Further broad investigations on human subjects are required for the validation of these studies and to determine the exact mechanism of action. Lower cholesterol levels, as compared to the controls, were observed in the blood sera of the mice fed with inulin and dextran supplemented feed. Between these, dextran exhibited a more pronounced effect on lowering the serum cholesterol level as compared to inulin. These results indicate that inulin and dextran have the potential to exert hypocholesterolemic effects. It has been documented that inulin and other fermentable soluble dietary fibers exert hypocholesterolemic effects by two mechanisms, i.e., selective fermentation by intestinal microbiota leading to the production of SCFAs, and enhanced cholesterol excretion through feces due to decreased cholesterol absorption [28]. It has also been proposed that the fermentation products, especially propionate and other SCFAs might regulate cholesterol metabolism. Reabsorption of bile acid is reduced and colonic pH is reduced because of the production of SCFAs [29]. Moreover, propionate may also affect liver cholesterol synthesis by reaching the liver [30]. Not all investigators agree to such roles of SCFAs. Hence, there is no concrete evidence of SCFAs role in lipid metabolism, particularly for propionate [31]. However, there could be a link between serum cholesterol level and lipid content excreted through feces. Indeed, the binding and entrapping properties of dietary fibers have been found to result in lower levels of steroid re-absorption [32]. Similarly, in the previous study by Yamamoto et al. [ 1999] [19], levan-fed rats have been found to have remarkably higher amount of lipids and total sterols in their fecal excretions, as compared to controls. These results suggested that levan was responsible for disturbing re-absorption by binding and entrapping the steroids in the intestine. In the present study also, inulin was found to have a very pronounced effect on the fecal lipid concentration, where it increased to $78\%$ in comparison to the control. However, dextran had no effect on the concentration of lipids excreted through the fecal matter, though feeding the mice on a dextran-supplemented diet drastically reduced their serum cholesterol level. Inulin might have exerted this hypocholesterolemic effect via the same mechanism of binding and entrapping lipids as reported in the previous studies with other dietary fibers, whereas some different mechanism operates behind the hypocholesterolemic effect of dextran, which needs further exploration. Another aspect of the present study was determining the changes in certain intestinal microbial populations induced in mice models fed on dextran and inulin supplemented diets. The balance between the commensal microflora, the diet and the mucosa are vital to maintain the gut health while any imbalance or dysbiosis could lead to adverse health effects including inflammatory bowel disease (IBD), multiple sclerosis, Alzheimer’s disease, diarrhea, colorectal cancer, etc. [ 33]. In this context, studies on the impact of prebiotic exopolysaccharides on a variety of health-promoting, pathogen-inhibiting, and microbiota-modulating activities are being undertaken, both in vitro and in vivo. The key findings of our study highlighted that both of the tested microbial exopolysaccharides i.e., dextran and inulin, cause a shift in the gastrointestinal microbiota through drastic reduction in the proliferation of Enterococcus faecalis. These polymers were also found highly effective against many other common enteric pathogens including Enterococcus, Klebsiella, Staphylococcus and Mammaliicoccus species that were found as major inhabitants in the gastrointestinal tract of the control group of mice fed with a normal diet. In previous investigations, the enteric pathogen inhibitory potential of exopolysaccharides was mostly investigated in vitro. In contrast, our study is among very few in vivo investigations using mice models to evaluate the inhibition of enteric pathogens in response to the dietary supplementation of microbial dextran and inulin type EPSs. The available literature shows that EPSs exhibit strong enteropathogen inhibitory potential. For instance, the EPS released from Lactobacillus acidophilus A4 was found to decrease the biofilms of E. coli (EHEC) by $87\%$ along with the inhibition of Gram negative and Gram positive pathogens [34]. In another study, it was found by Osamu Kanuchi et al. [ 2002] that the administration of prebiotic germinated barley foodstuff (GBF) enhanced the population of bifidobacteria and Eubacterium limosum while inhibiting or reducing Bacteroides, leading to improvement in the ulcerative colitis in patients [35]. Previously, Xiaoqing Xu et al. [ 2020] found that Lacticaseibacillus casei NA-2 derived EPS possesses strong growth and biofilm inhibitory potential against Staphylococcus aureus, Bacillus cereus, E. coli O157:H7 and Salmonella typhimurium exhibiting strong inhibition ratios of $30.2\%$ ± $3.3\%$, $95.5\%$ ± $0.1\%$, $16.9\%$ ± $5.4\%$, and $14.3\%$ ± $0.6\%$, respectively, which promotes the antibacterial activity of L. casei NA-2 [36]. Our results strengthened these previous findings that prebiotic EPSs play key role in the modulation of gastrointestinal microbiota by eliminating or inhibiting enteropathogens. There may be a number of causes for this inhibition. It has also been reported that the fermentation of prebiotics leads to the enrichment of the population of advantageous bacteria which competitively exclude pathogens by competing them for available nutrients. They are reported to produce short chain fatty acids (acetate, propionate, butyrate, etc.) that inhibit pathogen proliferation by lowering the pH of the lumen leading towards the competitive exclusion of pathogens less tolerant to low pH [12,37]. Prebiotics either directly bind to pathogens or raise the intestinal lumen’s osmotic pressure. In addition, they produce different antagonistic agents such as organic acids and anti-microbial peptides for direct inhibition of pathogens [38]. All these considerations are relevant to this present study which describes the significant inhibition of common enteric pathogens in response to prebiotics supplementation. Another interesting finding of the present study is the increase in the release of intestinal E. fergusonii by preventing its attachment to epithelial cells. Among nine species of Escherichia, E. fergusonii is a relatively new species discovered for the first time in 1985 from clinical blood samples, with a close resemblance to E. coli [39]. Its major health complications include urinary tract infections or bacteremia and open wounds infection [40]. E. coli is extensively used as a model organism for adhesion related investigations due to its distinctive adhesion manner. Various oligosaccharide-binding proteins also called adhesins are expressed in E. coli through which it binds with the oligosaccharide receptors on the surface of host cells. Since it represents the first stage of infection, the initial non-intimate adherence is a crucial feature of EPEC pathogenesis [41]. By preventing this early adhesion, the infectious process may eventually be stopped. In the past, few in vitro studies have explored the anti-adherence potential of prebiotics against E. coli. According to a study conducted by Zhengqi Liu et al. [ 2017], the administration of Lactiplantibacillus plantarum WLPL04 derived EPS significantly inhibited the adhesion of Escherichia coli HT-29 cells in competition, replacement, and inhibition assays at a dose of 1.0 mg/mL along with inhibition against biofilm formation of Pseudomonas aeruginosa CMCC10104, E. coli O157:H7, Salmonella typhimurium ATCC13311, and Staphylococcus aureus CMCC26003 [42]. In another in vitro study, galacto-oligosaccharides at an optimum dosage of 16 mg/mL were reported to inhibit the adhesion of enteropathogenic E. coli to the epithelial cells such as HEp-2 and Caco-2 cells by 65 and $70\%$, respectively, due to their structural homology with the pathogen binding sites [41]. All of these previous studies were performed in vitro using tissue cultures. In contrast, our study includes detailed in vivo analysis and molecular characterization to confirm the attachment inhibition of enteropathogens. Moreover, to the best of our knowledge, this is the first in vivo study which was performed by accessing the microbiota of mice model to describe the anti-adhesive properties of dextran and inulin against E. fergusonii, which is the closest neighbor to E. coli. The results illustrated that the intestinal fluid of control mice did not harbor significant populations of E. fergusonii while the intestinal mucosa exhibited their presence as a prevalent population; this confirms the hypothesis that E. fergusonii are attached with the intestinal epithelium and get detached when the intestine of same control animal is gently scraped with the sterile spatula. N contrast, in the intestinal fluid of dextran and inulin fed mice, the E. fergusonii was present as a major population, indicating that prebiotic EPSs not only significantly reduce the prevalence of enteropathogens but are also involved in promoting the eradication of E. fergusonii from the intestinal tract by preventing its adhesion to epithelial cells. The reason behind this may be that EPSs structurally mimic the pathogen binding sites that coat the surface of gastrointestinal epithelial cells, preventing pathogenesis and leading to the eradication or flushing out of the pathogens from the GI tract [41]. Besides the pathogenic inhibitory and anti-adhesive potential of dextran and inulin, our study was also focused on the increase in the population of beneficial microbes or lactic acid bacteria that might have been involved in the competitive exclusion of pathogens. Our results depict that the intestinal fluid of dextran- and inulin-fed mice harbor significantly higher populations of beneficial microorganisms including Lactococcus garvieae, Bacillus subtilis and Bacillus licheniformis when cultured on MRS-agar plates. In some previous studies, fructo-oligosaccharides (FOS) and inulin have been reported to inhibit the growth of E. coli and Salmonella in the intestine along with enhancing the host defense mechanism by boosting the metabolic activity of the lactobacillaceae family and bifidobacteria resulting in the inhibition of enteropathogens, a process called “resistance of colonization” [43]. The microbial glucans are also proved to increase stress tolerance and probiotic potential by promoting the growth of lactobacilli in the colon [44]. In addition to strengthening these previous findings, our study highlighted an interesting fact: prebiotic EPSs are also involved in enhancing the proliferation of yeast, which is a relatively new aspect of this study. The overall gut health is influenced by the diet, mucosa and commensal flora as they ensure efficient functioning of the digestive system by interacting with host epithelial cells forming a delicate and dynamic equilibrium within the alimentary tract [45]. The intestinal epithelium is composed of diverse cells exhibiting tight junctions between them to maintain a barrier [46] that is breached by the enteric pathogens due to release of their toxins; the resulting leakage of luminal content into the lamina affects residing immune cells and instigates inflammatory responses that ultimately lead to serious health challenges [47,48]. Pathogenesis is initiated with the adherence of pathogenic bacteria to the mucosal epithelial cells which initiates ‘crosstalk’ between the microbial and epithelial cells before colonization of the colon, with continuing flow of intestinal chyme, triggering a signaling cascade to stimulate the intense inflammatory immune response. The adherence is assisted by the binding of carbohydrate binding protein receptors of pathogens with 3–5 monosaccharides long oligosaccharides receptor sites on surface of epithelial cells. Any failure to adhere leads to rapid elimination of enteric pathogens from the gut [43,49]. Besides stimulating the growth of probiotics, it is reported that structural resemblance is present between the prebiotics and these oligosaccharides receptor sites [14,49,50]. Prebiotics incorporate similar non-digestible oligosaccharides which act as blocking factors by mimicking the ligands for protein receptors, preventing the pathogen adhesion to mucosal cells [51,52]. Certain terminal sugars on these oligosaccharides, e.g., oligofructose, interfere with the receptors by binding to the bacteria and preventing attachment to the same sugar on microvillus glycoconjugates. Due to this anti-adhesive property, prebiotics work as decoy molecules for the pathogenic microorganisms leading to their displacement or flushing from the gastrointestinal (GI) tract by preventing their adherence, causing a decrease in their pathogenic potential [53,54]. In addition, prebiotics have a major advantage over antibiotics whose extensive use may result in the persistence of resistant pathogenic bacteria which may pose threats to human or animal health. Instead of using antibiotics to protect the host against the colonization of pathogenic bacteria, the most viable approach is to use dietary components that can enhance the colonization resistance against entero-pathogens [45,55]. Being safe, affordable, cheap, and accessible, prebiotics are best candidate for this purpose [56]. ## 4.1. Production and Purification of Exopolysaccharides Two different kinds of microbially produced exopolysaccharides, i.e., dextran and inulin, were used in this study. Dextran was produced by a high yielding Weisella ciberia strain M3b. This strain has been previously isolated from fermented malt grains and identified by 16S rRNA gene sequence analysis that has been submitted to NCBI under the GenBank accession number MH084846 (data to be published separately). This isolate has been deposited to NIBGE Biotech Resource Centre (NBRC) (http://nibge.org/Division.aspx?page=Plant%20Microbiology&div=EBT), Pakistan, under the accession number NBRC-528 with an open access to other researchers on payment of shipment charges. Inulin was produced by Lactobacillus gasseri DSM 20604 strain. For EPS production, the cultures of both of these bacteria were grown in De Man, Rogosa and Sharpe agar (MRS) medium having the composition: Peptone (Rapid Labs, Cholchester, Essex, UK) 10 g/L, beef extract (Bio Basic Inc., Konrad Crescent Markham, ON, Canada) 8 g/L, yeast extract (Biochem, Za Conse Sur Ioire, France) 4 g/L, K2HPO4 (MP Biomedicals Inc., Solon, OH, USA) 2 g/L, sodium acetate (Sigma Aldrich, Saint Louis, USA) 3 g/L, diammonium citrate (Uni Chem, Mumbai, India) 2 g/L, magnesium sulphate (Riedel-de Haen AG Seelze, Hannover, Germany) 0.2 g/L, manganese sulphate (Sigma Aldrich, Saint Louis, MO, USA) 0.5 g/L, Tween 80 (Bio Basic Inc., Markham, ON, Canada) 1 mL/L, pH 6.2–6.5) [57] supplemented with $20\%$ sucrose (PhytoTechnology Laboratories, Shawnee Mission, KS, USA). The W. cibaria and L. gasseri cultures were incubated at 28 °C under aerobic conditions and 37 °C under anaerobic conditions, respectively, for 96 h to get the maximum possible EPS yields. Presence of exopolysaccharides in the cultures and purity of the purified polysaccharide products were determined through thin layer chromatography (TLC) analysis. For this purpose, 2 μL of the sample was run on TLC plates (Silica gel 60 F254; Merck, Darmstadt, Germany) for about 6 h in a mobile phase comprising butanol/ethanol/water (5:5:3). To detect the fructose containing carbohydrates (e.g., inulin), TLC plate was air-dried and sprayed with urea developing solution (100 mL water saturated butanol, 3.0 g urea, 5.9 mL phosphoric acid, 5 mL ethanol) followed by heating at 120 °C for 15 min [58]. Other carbohydrate spots (e.g., dextran) were visualized by developing with a solution containing $5\%$ sulfuric acid in methanol. In order to purify EPSs, the W. cibaria and L. gasseri cultures were centrifuged at 4824× g for 20 min to remove the bacterial cells. The supernatants containing the EPSs were collected and the cell pellets were discarded. The proteins present in the cultures were removed by treatment with tricholoroacetic acid (TCA) following the method described by Abid et al. [ 2019] [59]. The dextran polymer was precipitated from the TCA-treated culture supernatant by adding one volume of ice-cold absolute ethanol, while two volumes of ethanol were used for the precipitation of inulin. The precipitated polymers were then washed by dissolving in distilled water and centrifuging again at 4824× g for 10 min. The pellets were preserved while the supernatants were discarded. This step was repeated 2–3 times until the purified polymers were obtained. Both the polymers were then dissolved in distilled water and analyzed by thin layer chromatography (TLC) to confirm the absence of any residual sucrose and glucose molecules. Finally, the polymers were suspended in water and freeze-dried to obtain in fine powder form. ## 4.2. Experimental Design Total 18 male BALB/c mice aged five weeks were purchased from Government College University Faisalabad, Pakistan, and equally divided into three groups i.e., control group, dextran-treated group, and inulin-treated group. Before starting the experiment, all the mice were given normal control diet in order to acclimatize them to the new environment. All three groups of mice were kept in their respective cages placed in a room where temperature was maintained at 27 ± 1 °C. Each group of mice was placed in a separate cage. The control group was only fed with a normal diet with no prebiotic polymer added while the other two experimental groups were fed with the diet mixed with the inulin and dextran polymers ($5\%$ w/w). Mice were kept under observation for 21 days. Daily feedings were given in accordance with their needs and the animals were given free access to pure water and meals. The diet was prepared by mixing purified powdered forms of prebiotic exopolysaccharides with commercially available mice feed. The feed was sterilized in an autoclave at 121 °C for 10 min to ensure the absence of any exogenous bacteria prior to the addition of the polymers. The feed was prepared on daily basis to avoid any contamination and decay. ## 4.3. Animal Euthanization and Sample Collection Feed was removed about 20 h before the end of the experimental period [19], after which the mice were anesthetized with pentobarbital using standard ethical procedures. After waiting for 5 to 10 min until mice were completely unconscious, they were euthanized. Intestines were separated and squeezed to collect the intestinal fluid sample of all groups of mice. In addition, the intestine of the mice was also scraped gently to collect the intestinal mucosa from the walls of the intestine. The fluid samples were suspended in 10 mL of sterilized normal saline ($0.9\%$ w/v NaCl (Biochem, Za Conse Sur Ioire, France) and mixed gently to homogenize it by vortexing at low speed. Samples were labelled and stored at 4 °C for further analysis. All animal-related procedures were according to the guidelines of the Animal House Committee after approval from the institutional biomedical ethical committee. International Guiding Principles for Biomedical Research Involving Animals as issued by the Council for the International Organizations of Medical Sciences were followed in these experiments. ## 4.4. Determination of Body and Organ Weight Overall body weights of mice of all three groups of this study were determined at the beginning (Day 0) and the end of experiment (Day 21) a using top load balance. The mice having similar initial body weights were selected for use in these experiments. The weights of different body organs were also measured at the end of the experiment using an analytical balance. ## 4.5. Analysis of Fecal Lipids Beddings in all the mice cages were replaced with fresh material, 48 h before the end of experiments. At the last day of the experiment, feces were collected from the cages for the determination of total lipids excretion. The fecal samples were lyophilized, ground and stored at −20 °C for further analysis. Total lipids were extracted from 1.0 g of fecal sample from each experiment in triplicate following the protocol described by Kraus et al. [ 2015] [60] and measured gravimetrically. ## 4.6. Blood and Tissue Collection At the end of the experimental period, the animals were anesthetized with pentobarbital and euthanized. The blood was collected by heart puncture and instantly transferred to Gel & Clot Activator tubes (Xinle, China) and kept on ice. The tubes were centrifuged at the speed of 3000 rpm for 15 min at 0 °C. Serum was separated and preserved in separate tubes at −20 °C. After blood sampling, the heart, kidneys, liver, lungs, pancreas, spleen and stomach were detached and weighed after drying with paper tissue. ## 4.7. Glucose and Cholesterol Analysis At the beginning of the experiment (day 0) and at the end of the experiment (day 21), blood drops were collected from each mouse by puncturing the vein in the tail with a sterile syringe needle. The concentration of blood glucose was determined using On Call Extra blood glucose monitoring system (ACON Laboratories, Hannover, Germany). Cholesterol levels in the preserved blood sera (from Section 4.6) were measured by Microlab 300 analyser (Merck, Darmstadt, Germany). ## 4.8. Analysis of Intestinal Microbiota In order to examine the presence and diversity of common enteric pathogens, intestinal microbiota of mice were examined using CHROMagar (Kanto Chemical Company Inc., Tokyo, Japan) and MacConkey agar media. Following the manufacturer’s instructions, the media were prepared by adding 33 g and 49.53 g of CHROMagar and MacConkey agar powder respectively to 1 L of distilled water followed by sterilization at 121 °C. Using sterilized normal saline, all samples containing intestinal fluid and mucosa from control and experimental animals were serially diluted 100–1000 times. 100 µL of each dilution was dispersed on both media agar plates and overnight incubation was provided at 37 °C to allow different microorganism to grow. To study the presence and abundance of lactobacilli in the colon of the control and experimental group of mice, all samples were similarly dispersed on MRS agar plates supplemented with $20\%$ sucrose and others with $2\%$ glucose. After overnight incubation, multicolored colonies were enumerated and compared in order to determine their relative abundance in control and experimental groups. ## 4.9. Isolation and Molecular Characterization of Various Intestinal Isolates Different colored colonies which appeared on the CHROM agar and MRS agar plates were isolated, purified and primarily characterized by phase contrast microscopy for initial confirmation of the morphology and the presence of similar microorganisms indicated in CHROMagar orientation. These isolates were further allowed to grow in LB media for genomic DNA extraction whereas the isolates from the MRS agar plates were grown in MRS broth. Some large white color colonies were grown on YEPD (yeast extract peptone dextrose) medium (Peptone 20 g/L, yeast extract 10 g/L, dextrose 20 g/L) supplemented with kanamycin. Genomic DNA of all these isolates was extracted by using Thermo-scientific Gene-jet Genomic DNA extraction kit. For molecular identification, genomic DNA of all the isolated microbes was used as template for their respective 16S rRNA gene amplification by using universal primers FD1 (forward primer) and RP1 (reverse primer) as described previously [61]. PCR products were purified through GeneJet PCR purification kit by following user provided instructions and commercially sequenced by Macrogen, Republic of Korea, using Sanger Sequencing method. Homology of the sequences was determined through the NCBI blastn program. The sequences of closest neighbors were retrieved from GenBank and phylogenetic trees were constructed by the neighbor joining method using MEGA11 (version 11.0.10) software. ## 4.10. Statistical Analysis Graph-Pad Prism (Version 5) was used to analyze the statistical significance of data. One-way analysis of variance (ANOVA) was performed and the mean differences among the different treatment groups were evaluated. Tukey’s multiple comparison test was applied to analyze the statistical differences among means ($p \leq 0.05$). ## 5. Conclusions In view of all the above-mentioned findings, it can be concluded that prebiotic EPSs are highly effective ingredients not only for improving metabolomics but also for the modulation of gastrointestinal microbiota in a beneficial manner. Besides exhibiting strong Enterococcus inhibitory potential, this study demonstrated the anti-adhesive properties of prebiotics against E. fergusonii to prevent its pathogenesis and strengthened the evidence that prebiotic EPSs promote the growth of beneficial bacteria and yeast in the GI tract. All these findings declare these EPSs as suitable candidates for use as prebiotics. ## References 1. Watt E., Gemmell M.R., Berry S., Glaire M., Farquharson F., Louis P., Murray G.I., El-Omar E., Hold G.L.. Extending colonic mucosal microbiome analysis—Assessment of colonic lavage as a proxy for endoscopic colonic biopsies. Microbiome (2016) 4 61. DOI: 10.1186/s40168-016-0207-9 2. Scarpellini E., Rinninella E., Basilico M., Colomier E., Rasetti C., Larussa T., Santori P., Abenavoli L.. From pre- and probiotics to post-biotics: A narrative review. Int. J. Environ. Res. Public Health (2021) 19. DOI: 10.3390/ijerph19010037 3. Qin J., Li R., Raes J., Arumugam M., Burgdorf K.S., Manichanh C., Nielsen T., Pons N., Levenez F., Yamada T.. A human gut microbial gene catalogue established by metagenomic sequencing. Nature (2010) 464 59-65. DOI: 10.1038/nature08821 4. Rajilić-Stojanović M., De Vos W.M.. The first 1000 cultured species of the human gastrointestinal microbiota. FEMS Microbiol. Rev. (2014) 38 996-1047. DOI: 10.1111/1574-6976.12075 5. Dieterich W., Schink M., Zopf Y.. Microbiota in the gastrointestinal tract. Med. Sci. (2018) 6. DOI: 10.3390/medsci6040116 6. Roberfroid M.. Functional food concept and its application to prebiotics. Dig. Liver. Dis. (2002) 34 S105-S110. DOI: 10.1016/S1590-8658(02)80176-1 7. Versluys M., Kirtel O., Toksoy Öner E., Van den Ende W.. The fructan syndrome: Evolutionary aspects and common themes among plants and microbes. Plant Cell Environ. (2018) 41 16-38. DOI: 10.1111/pce.13070 8. Kırtel O., Lescrinier E., Van den Ende W., Öner E.T.. Discovery of fructans in Archaea. Carbohydr. Polym. (2019) 220 149-156. DOI: 10.1016/j.carbpol.2019.05.064 9. van Hijum S.A., Kralj S., Ozimek L.K., Dijkhuizen L., van Geel-Schutten I.G.. Structure-function relationships of glucansucrase and fructansucrase enzymes from lactic acid bacteria. Microbiol. Mol. Biol. Rev. (2006) 70 157-176. DOI: 10.1128/MMBR.70.1.157-176.2006 10. Chen G., Xie M., Wan P., Chen D., Ye H., Chen L., Zeng X., Liu Z.. Digestion under saliva, simulated gastric and small intestinal conditions and fermentation in vitro by human intestinal microbiota of polysaccharides from Fuzhuan brick tea. Food Chem. (2018) 244 331-339. DOI: 10.1016/j.foodchem.2017.10.074 11. Gibson G.R., Fuller R.. Aspects of in vitro and in vivo research approaches directed toward identifying probiotics and prebiotics for human use. J. Nutr. (2000) 130 391S-395S. DOI: 10.1093/jn/130.2.391S 12. Celebioglu H.U., Svensson B.. Dietary nutrients, proteomes, and adhesion of probiotic lactobacilli to mucin and host epithelial cells. Microorganisms (2018) 6. DOI: 10.3390/microorganisms6030090 13. Quigley E.M.. Prebiotics and probiotics; modifying and mining the microbiota. Pharmacol. Res. (2010) 61 213-218. DOI: 10.1016/j.phrs.2010.01.004 14. Gibson G., McCartney A., Rastall R.. Prebiotics and resistance to gastrointestinal infections. Br. J. Nutr. (2005) 93 S31-S34. DOI: 10.1079/BJN20041343 15. Mortensen A., Poulsen M., Frandsen H.. Effect of a long-chained fructan Raftiline HP on blood lipids and spontaneous atherosclerosis in low density receptor knockout mice. Nutr. Res. (2002) 22 473-480. DOI: 10.1016/S0271-5317(02)00358-5 16. Zhang X., Hu P., Zhang X., Li X.. Chemical structure elucidation of an inulin-type fructan isolated from Lobelia chinensis lour with anti-obesity activity on diet-induced mice. Carbohydr. Polym. (2020) 240 116357. DOI: 10.1016/j.carbpol.2020.116357 17. Kim M., Shin H.K.. The water-soluble extract of chicory influences serum and liver lipid concentrations, cecal short-chain fatty acid concentrations and fecal lipid excretion in rats. J. Nutr. (1998) 128 1731-1736. DOI: 10.1093/jn/128.10.1731 18. Belghith K.S., Dahech I., Hamden K., Feki A., Mejdoub H., Belghith H.. Hypolipidemic effect of diet supplementation with bacterial levan in cholesterol-fed rats. Int. J. Biol. Macromol. (2012) 50 1070-1074. DOI: 10.1016/j.ijbiomac.2012.02.024 19. Yamamoto Y., Takahashi Y., Kawano M., Iizuka M., Matsumoto T., Saeki S., Yamaguchi H.. In vitro digestibility and fermentability of levan and its hypocholesterolemic effects in rats. J. Nutr. Biochem. (1999) 10 13-18. DOI: 10.1016/S0955-2863(98)00077-1 20. Flint H.J., Scott K.P., Duncan S.H., Louis P., Forano E.. Microbial degradation of complex carbohydrates in the gut. Gut Microbes (2012) 3 289-306. DOI: 10.4161/gmic.19897 21. Roberfroid M.B.. Caloric value of inulin and oligofructose. J. Nutr. (1999) 129 1436S-1437S. DOI: 10.1093/jn/129.7.1436S 22. Turnbaugh P.J., Ley R.E., Mahowald M.A., Magrini V., Mardis E.R., Gordon J.I.. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature (2006) 444 1027-1031. DOI: 10.1038/nature05414 23. Jumpertz R., Le D.S., Turnbaugh P.J., Trinidad C., Bogardus C., Gordon J.I., Krakoff J.. Energy-balance studies reveal associations between gut microbes, caloric load, and nutrient absorption in humans. Am. J. Clin. Nutr. (2011) 94 58-65. DOI: 10.3945/ajcn.110.010132 24. Sleeth M.L., Thompson E.L., Ford H.E., Zac-Varghese S.E., Frost G.. Free fatty acid receptor 2 and nutrient sensing: A proposed role for fibre, fermentable carbohydrates and short-chain fatty acids in appetite regulation. Nutr. Res. Rev. (2010) 23 135-145. DOI: 10.1017/S0954422410000089 25. Wu J., Shi S., Wang H., Wang S.. Mechanisms underlying the effect of polysaccharides in the treatment of type 2 diabetes: A review. Carbohydr. Polym. (2016) 144 474-494. DOI: 10.1016/j.carbpol.2016.02.040 26. Swennen K., Courtin C.M., Delcour J.A.. Non-digestible oligosaccharides with prebiotic properties. Crit. Rev. Food Sci. Nutr. (2006) 46 459-471. DOI: 10.1080/10408390500215746 27. Brubaker P., Drucker D.. Minireview: Glucagon-like peptides regulate cell proliferation and apoptosis in the pancreas, gut, and central nervous system. Endocrinology (2004) 145 2653-2659. DOI: 10.1210/en.2004-0015 28. Ooi L.-G., Liong M.-T.. Cholesterol-lowering effects of probiotics and prebiotics: A review of in vivo and in vitro findings. Int. J. Mol. Sci. (2010) 11 2499-2522. DOI: 10.3390/ijms11062499 29. Remesy C., Levrat M., Gamet L., Demigné C.. Cecal fermentations in rats fed oligosaccharides (inulin) are modulated by dietary calcium level. Am. J. Physiol.-Gastrointest. Liver Physiol. (1993) 264 G855-G862. DOI: 10.1152/ajpgi.1993.264.5.G855 30. Illman R.J., Topping D.L., Mclntosh G.H., Trimble R.P., Storer G.B., Taylor M.N., Cheng B.-Q.. Hypocholesterolaemic effects of dietary propionate: Studies in whole animals and perfused rat liver. Ann. Nutr. Metab. (1988) 32 97-107. DOI: 10.1159/000177414 31. Nishina P.M., Freedland R.A.. Effects of propionate on lipid biosynthesis in isolated rat hepatocytes. J. Nutr. (1990) 120 668-673. DOI: 10.1093/jn/120.7.668 32. Gallaher D., Schneeman B.O.. Intestinal interaction of bile acids, phospholipids, dietary fibers, and cholestyramine. Am. J. Physiol.-Gastrointest. Liver Physiol. (1986) 250 G420-G426. DOI: 10.1152/ajpgi.1986.250.4.G420 33. Marchesi J.R., Adams D.H., Fava F., Hermes G.D., Hirschfield G.M., Hold G., Quraishi M.N., Kinross J., Smidt H., Tuohy K.M.. The gut microbiota and host health: A new clinical frontier. Gut (2016) 65 330-339. DOI: 10.1136/gutjnl-2015-309990 34. Kim Y., Kim S.H.. Released exopolysaccharide (r-EPS) produced from probiotic bacteria reduce biofilm formation of enterohemorrhagic Escherichia coli O157: H7. Biochem. Biophys. Res. Commun. (2009) 379 324-329. DOI: 10.1016/j.bbrc.2008.12.053 35. Kanauchi O., Suga T., Tochihara M., Hibi T., Naganuma M., Homma T., Asakura H., Nakano H., Takahama K., Fujiyama Y.. Treatment of ulcerative colitis by feeding with germinated barley foodstuff: First report of a multicenter open control trial. J. Gastroenterol. (2002) 37 67-72. DOI: 10.1007/BF03326417 36. Xu X., Peng Q., Zhang Y., Tian D., Zhang P., Huang Y., Ma L., Dia V.P., Qiao Y., Shi B.. Antibacterial potential of a novel Lactobacillus casei strain isolated from Chinese northeast sauerkraut and the antibiofilm activity of its exopolysaccharides. Food Funct. (2020) 11 4697-4706. DOI: 10.1039/D0FO00905A 37. Adhikari P.A., Kim W.K.. Overview of prebiotics and probiotics: Focus on performance, gut health and immunity—A review. Ann. Anim. Sci. (2017) 17 949-966. DOI: 10.1515/aoas-2016-0092 38. Tannock G.W.. Probiotics and Prebiotics: Where Are We Going? (2002) 39. Farmer J., Fanning G., Davis B., O’hara C., Riddle C., Hickman-Brenner F., Asbury M., Lowery V., Brenner D.. J. Clin. Microbiol. (1985) 21 77-81. DOI: 10.1128/jcm.21.1.77-81.1985 40. Maheux A., Boudreau D., Bergeron M., Rodriguez M.. Characterization of. J. Appl. Microbiol. (2014) 117 597-609. DOI: 10.1111/jam.12551 41. Shoaf K., Mulvey G.L., Armstrong G.D., Hutkins R.W.. Prebiotic galactooligosaccharides reduce adherence of enteropathogenic Escherichia coli to tissue culture cells. Infect. Immun. (2006) 74 6920-6928. DOI: 10.1128/IAI.01030-06 42. Liu Z., Zhang Z., Qiu L., Zhang F., Xu X., Wei H., Tao X.. Characterization and bioactivities of the exopolysaccharide from a probiotic strain of Lactobacillus plantarum WLPL04. J. Dairy Sci. (2017) 100 6895-6905. DOI: 10.3168/jds.2016-11944 43. Bosscher D., Van Loo J., Franck A.. Inulin and oligofructose as prebiotics in the prevention of intestinal infections and diseases. Nutr. Res. Rev. (2006) 19 216-226. DOI: 10.1017/S0954422407249686 44. Russo P., López P., Capozzi V., De Palencia P.F., Dueñas M.T., Spano G., Fiocco D.. Beta-glucans improve growth, viability and colonization of probiotic microorganisms. Int. J. Mol. Sci. (2012) 13 6026-6039. DOI: 10.3390/ijms13056026 45. Montagne L., Pluske J., Hampson D.. A review of interactions between dietary fibre and the intestinal mucosa, and their consequences on digestive health in young non-ruminant animals. Anim. Feed Sci. Technol. (2003) 108 95-117. DOI: 10.1016/S0377-8401(03)00163-9 46. Marchiando A.M., Graham W.V., Turner J.R.. Epithelial barriers in homeostasis and disease. Ann. Rev. Pathol. (2010) 5 119-144. DOI: 10.1146/annurev.pathol.4.110807.092135 47. Su L., Shen L., Clayburgh D.R., Nalle S.C., Sullivan E.A., Meddings J.B., Abraham C., Turner J.R.. Targeted epithelial tight junction dysfunction causes immune activation and contributes to development of experimental colitis. Gastroenterology (2009) 136 551-563. DOI: 10.1053/j.gastro.2008.10.081 48. Kolling G., Wu M., Guerrant R.L.. Enteric pathogens through life stages. Front. Cell. Infect. Microbiol. (2012) 2 114. DOI: 10.3389/fcimb.2012.00114 49. Kelly D., Campbell J.I., King T.P., Grant G., Jansson E.A., Coutts A.G., Pettersson S., Conway S.. Commensal anaerobic gut bacteria attenuate inflammation by regulating nuclear-cytoplasmic shuttling of PPAR-γ and RelA. Nat. Immunol. (2004) 5 104-112. DOI: 10.1038/ni1018 50. Kunz C., Rudloff S., Baier W., Klein N., Strobel S.. Oligosaccharides in human milk: Structural, functional, and metabolic aspects. Annu. Rev. Nutr. (2000) 20 699. DOI: 10.1146/annurev.nutr.20.1.699 51. Ofek I., Doyle R.J.. Bacterial Adhesion to Cells and Tissues (2012) 52. Dai D., Nanthkumar N.N., Newburg D.S., Walker W.A.. Role of oligosaccharides and glycoconjugates in intestinal host defense. J. Pediatr. Gastroenterol. Nutr. (2000) 30 S23-S33. DOI: 10.1097/00005176-200003002-00005 53. Forchielli M.L., Walker W.A.. The role of gut-associated lymphoid tissues and mucosal defence. Br. J. Nutr. (2005) 93 S41-S48. DOI: 10.1079/BJN20041356 54. Steer T., Carpenter H., Tuohy K., Gibson G.R.. Perspectives on the role of the human gut microbiota and its modulation by pro- and prebiotics. Nutr. Res. Rev. (2000) 13 229-254. DOI: 10.1079/095442200108729089 55. Bonfiglio G., Neroni B., Radocchia G., Marazzato M., Pantanella F., Schippa S.. Insight into the Possible Use of the Predator Bdellovibrio bacteriovorus as a Probiotic. Nutrients (2020) 12. DOI: 10.3390/nu12082252 56. Sanders M.E., Merenstein D.J., Reid G., Gibson G.R., Rastall R.A.. Probiotics and prebiotics in intestinal health and disease: From biology to the clinic. Nat. Rev. Gastroenterol. Hepatol. (2019) 16 605-616. DOI: 10.1038/s41575-019-0173-3 57. De Man J., Rogosa d., Sharpe M.E.. A medium for the cultivation of lactobacilli. J. Appl. Bacteriol. (1960) 23 130-135. DOI: 10.1111/j.1365-2672.1960.tb00188.x 58. Trujillo Toledo L.E., Gómez Riera R., Banguela Castillo A., Soto Romero M., Arrieta Sosa J.G., Hernández García L.. Catalytical properties of N-glycosylated Gluconacetobacter diazotrophicus levansucrase produced in yeast. Electron. J. Biotechnol. (2004) 7 115-123. DOI: 10.2225/vol7-issue2-fulltext-4 59. Abid Y., Azabou S., Casillo A., Gharsallah H., Jemil N., Lanzetta R., Attia H., Corsaro M.M.. Isolation and structural characterization of levan produced by probiotic Bacillus tequilensis-GM from Tunisian fermented goat milk. Int. J. Biol. Macromol. (2019) 133 786-794. DOI: 10.1016/j.ijbiomac.2019.04.130 60. Kraus D., Yang Q., Kahn B.. Lipid Extraction from Mouse Feces. Bio-Protocol (2015) 5 e1375. DOI: 10.21769/BioProtoc.1375 61. Akhtar N., Ghauri M.A., Iqbal A., Anwar M.A., Akhtar K.. Biodiversity and phylogenetic analysis of culturable bacteria indigenous to Khewra salt mine of Pakistan and their industrial importance. Braz. J. Microbiol. (2008) 39 143-150. DOI: 10.1590/S1517-83822008000100029
--- title: Genetic Background of Metabolically Healthy and Unhealthy Obesity Phenotypes in Hungarian Adult Sample Population authors: - Peter Piko - Erand Llanaj - Karoly Nagy - Roza Adany journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049500 doi: 10.3390/ijms24065209 license: CC BY 4.0 --- # Genetic Background of Metabolically Healthy and Unhealthy Obesity Phenotypes in Hungarian Adult Sample Population ## Abstract A specific phenotypic variant of obesity is metabolically healthy (MHO), which is characterized by normal blood pressure and lipid and glucose profiles, in contrast to the metabolically unhealthy variant (MUO). The genetic causes underlying the differences between these phenotypes are not yet clear. This study aims to explore the differences between MHO and MUO and the contribution of genetic factors (single nucleotide polymorphisms—SNPs) in 398 Hungarian adults (81 MHO and 317 MUO). For this investigation, an optimized genetic risk score (oGRS) was calculated using 67 SNPs (related to obesity and to lipid and glucose metabolism). Nineteen SNPs were identified whose combined effect was strongly associated with an increased risk of MUO (OR = 1.77, $p \leq 0.001$). Four of them (rs10838687 in MADD, rs693 in APOB, rs1111875 in HHEX, and rs2000813 in LIPG) significantly increased the risk of MUO (OR = 1.76, $p \leq 0.001$). Genetic risk groups based on oGRS were significantly associated with the risk of developing MUO at a younger age. We have identified a cluster of SNPs that contribute to the development of the metabolically unhealthy phenotype among Hungarian adults suffering from obesity. Our findings emphasize the significance of considering the combined effect(s) of multiple genes and SNPs in ascertaining cardiometabolic risk in obesity in future genetic screening programs. ## 1. Introduction Obesity is one of the alarmingly increasing global health challenges which presently affects more than a billion people worldwide [1], and it is disproportionately prevalent among vulnerable socioeconomic groups and ethnic minorities [2,3,4]. Its direct and indirect influence on non-communicable diseases (NCDs) are well-documented, with a special emphasis on the risk of type two diabetes (T2D), cardiovascular diseases (CVDs), and several types of cancer attributed to increased adiposity [5,6,7,8]. In addition to the fact that obesity increases the risk of certain diseases, the World Obesity Federation has also identified obesity itself as a progressive, chronic, relapsing disease [9]. This finding is supported by the fact that the pathophysiology of obesity is influenced by the interaction of environmental/lifestyle factors (such as an energy-dense diet and sedentary lifestyle) and genetic predisposition, which leads to increased body weight [9] and a positive energy balance. Data from independent studies show that in a subgroup of obese individuals, no positive association between body mass index (BMI) and cardiometabolic risk was observed, and these individuals may be protected against obesity-related cardiometabolic diseases (CMD), or at least have a significantly lower risk of developing CMD [10] than expected. This is considered a metabolically healthy obesity (MHO) sub-phenotype, and despite the lack of a universal definition [11], such a phenotype refers to those being obese and yet having no cardiometabolic complications. Although there is no standardized definition of MHO, the following criteria have been proposed in addition to the diagnosis of obesity (BMI ≥ 30 kg/m2): fasting levels of serum glucose ≤ 6.1 mM (≤100 mg/dL); triacylglycerols (TAG) ≤ 1.7 mM (≤150 mg/dL); high-density lipoprotein (HDL) cholesterol > 1.0 mM (>40 mg/dL) in men or > 1.3 mM (>50 mg/dL) in women; systolic blood pressure (SBP) ≤130 mmHg; diastolic blood pressure (DBP) ≤ 85 mmHg; no drug treatment for dyslipidemia, diabetes, or hypertension; and no cardiovascular disease manifestation [12,13]. Accumulating evidence suggests that although people with metabolically healthy obesity could have a higher risk of all-cause mortality and cardiovascular events compared to metabolically healthy non-obese people, their risks are substantially lower compared to individuals with metabolically unhealthy obesity (MUO) [14,15]. The prevalence of metabolically healthy obesity is estimated to be somewhere between $10\%$ and $51\%$ depending on age and sex [12,16]. The proportion of MHO was higher in women and decreased with age, and slightly higher in groups of Asian origin compared to non-Asian origin [16]. Obesity is a complex trait influenced by genetic (polymorphism, epigenetic, and metagenomic factors) and non-genetic (environmental and lifestyle) factors and their interactions [17]. The elevated risk of obesity has been linked to individual single nucleotide polymorphisms (SNPs) in human genes encoding proteins such as adipocyte-, C1q-, and collagen domain-containing (ADIPOQ) fat mass- and obesity-associated gene (FTO), leptin (LEP), leptin receptor (LEPR), insulin-induced gene two (INSIG2), melanocortin four receptor (MC4R), proprotein convertase subtilisin/kexin type one (PCSK1), and peroxisome proliferator-activated receptor gamma (PPARG) [18,19]. Even though obesity is a hot topic in modern medicine and requires a multidisciplinary approach to its investigation, the genetic causes underlying the difference between metabolically healthy and unhealthy phenotypes are currently not well understood [20]. Presently, limited data exist with regard to a genetic predisposition to MHO/MUO on human samples. A publication by Berezina and colleagues in which 503 abnormally obese patients without cardiovascular disease were studied found that MHO and MUO patients did not differ in the distribution of the LEP gene genotypes G19G, G19A, and A19A and the Adiponectin (AN) gene genotypes G276G, G276T, and T276T [21]. Li and colleagues identified two SNPs (rs2237897 and rs2237892) in the Potassium Voltage-Gated Channel Subfamily Q Member one (KCNQ1) gene in Chinese children aged 6 to 18 years that influenced susceptibility to MHO in interaction with environmental and lifestyle factors. Sedaghati-khayat and colleagues identified four SNPs (rs1421085, rs1121980, rs1558902, and rs8050136) in the FTO gene that are significantly associated with MUO [22]. The impact of obesity on health [23,24] and the economy [25] is a growing burden that requires comprehensive knowledge to reduce. Since obesity (and its associated metabolic pathways) is significantly heritable and our knowledge of the reasons behind the difference between the MHO and MUO phenotypes is limited, it is worth conducting further research to identify the underlying genetic causes. Hence, the present study aims to investigate the genetic background of the MHO and MUO phenotypes using obese samples from the Hungarian adult population, and to establish a genetic score that can be used to estimate the risk of MUO at an individual or a population level. ## 2.1. Characteristics of the Obese Study Samples After the exclusion of subjects with incomplete geno- and/or phenotype data, a total of 398 obese individuals (317 MUO and 81 MHO) remained in the database for the current analyses (Table 1). Apart from expected differences in biomarkers’ values and the prevalence of medication use, MHO participants were younger compared to MUO, with no sex and education-specific differences. The statistical analyses were corrected for the age of the participants, thus avoiding any effects due to age differences. ## 2.2. The Best-Fitting Genetic Models by SNPs Twenty-three SNPs showed the strongest association with MUO for the recessive, 12 for the codominant, and 32 for the dominant inheritance model (Supplementary Table S1). ## 2.3. Optimization of Genetic Risk Score and the Association of Optimized Genetic Risk Score with MUO and Related Parameters The optimization process of genetic risk score (GRS) was performed based on SNPs that were shown to strengthen the association between GRS and MUO by logistic regression analysis, starting with the SNP with the strongest association (rs10838687: odds ratio (OR) = 1.92, $95\%$ confidence intervals ($95\%$ CI): 1.26–2.93; $$p \leq 0.002$$) and moving step by step in decreasing order to the weakest (rs659366: OR = 1.03, $95\%$ CI: 0.79–1.34; $$p \leq 0.844$$). During the process, 19 SNPs (for more details, see Supplementary Table S1) were selected, i.e., included in the present study. The mean value of optimized GRS (oGRS) was 21.4 ($95\%$ CI: 20.9–21.9) in the MHO group and 23.7 ($95\%$ CI: 23.5–23.9) in the MUO group. The distribution of the oGRS showed a significant difference ($p \leq 0.001$) between the two groups, and significantly higher oGRS values were observed in the MUO group compared to the MHO group (Figure 1). The 19 identified SNPs are located in 15 genes, mainly in three main clusters (first cluster: ADIPOQ, APOB, CETP, LIPC, LIPG and LPL; second cluster: PPARG; and third cluster: C2CD4B, CDKN2B, GIPR, HHEX, SLC2A2 and SLC30A8). The genes in the first cluster, with the exception of ADIPOQ, show a strong association with lipid metabolism, while genes in the second and third clusters are associated with type two and gestational diabetes. KCTD10 and MADD genes could not be classified in either cluster (Figure 2). None of the MUO-associated parameters tested (BMI, waist circumference (WC), SBP, DBP, fasting TAG, HDL-C, glucose, insulin, and homeostasis model assessment of insulin resistance (HOMA-IR)) showed any significant correlation with oGRS (after the test correction) (Supplementary Table S2). ## 2.4. The Discriminatory Power of MUO-Associated Genetic and Non-Genetic Risk Factors Based on ROC Curve Analyses Age showed the highest discriminatory power (area under the receiver operating characteristic (ROC) curve (AUC)age = 0.71, $95\%$ CI: 0.63–0.78) among the conventional risk factors not considered among the MHO/MHO differential diagnostic criteria by Wildman et al. [ 26] and Meigs et al. [ 27] (sex, age, BMI, and education). Among the physical and laboratory parameters used by them to define MUO (systolic and diastolic blood pressure, WC, fasting TAG, HDL-C, glucose, C-reactive protein (CRP), and HOMA-IR), TAG level showed the highest discriminatory power (AUCTAG = 0.77, $95\%$ CI: 0.72–0.82). For oGRS, the power of discrimination was found to be AUCoGRS = 0.77, $95\%$ CI: 0.71–0.83. Based on a statistical comparison of the AUC curves, the discriminatory power of the oGRS was calculated by using the four most highly MUO-related SNPs (rs10838687 in MAP Kinase Activating Death Domain—MADD, rs693 in Apolipoprotein B—APOB, rs1111875 in Hematopoietically Expressed Homeobox—HHEX and rs2000813 in Endothelial lipase—LIPG-oGRS4) and was not significantly different from the oGRS calculated by involving the full panel of 19 SNPs included in the study (AUCoGRS = 0.77 vs. AUCoGRS4 = 0.69, ΔAUC = 0.082, $$p \leq 0.014$$; Bonferroni corrected p-value < 0.0026). A combination of risk factors defined as conventional ones showed an overall discriminatory power of AUCconv. = 0.73 (95CI: 0.67–0.80). Conventional factors and oGRS4 together were as follows: AUCconv.+oGRS4 = 0.79 ($95\%$ CI: 0.74–0.85). Meanwhile, conventional and oGRS together were as follows: AUCconv.+oGRS = 0.85 ($95\%$ CI: 0.80–0.90). The combination of both oGRS4 (AUCconv. = 0.73 vs. AUCGRS4 = 0.79, ΔAUC = 0.059; $$p \leq 0.007$$) and oGRS (AUCconv. = 0.73 vs. AUCoGRS = 0.85, ΔAUC = 0.112; $p \leq 0.001$) significantly improved the discrimination index compared to the one that only took into account the conventional risk factors (Figure 3). The distribution of oGRS4 showed a significant difference between the MHO and MUO groups (MHO = 4.47, $95\%$ CI: 4.17–4.77 vs. MUO = 5.38, $95\%$ CI: 5.24–5.53; $p \leq 0.001$; Figure 4) and a significant correlation with MUO risk according to adjusted logistic regression analysis (OR = 1.76, $95\%$ CI: 1.43–2.17; $$p \leq 8.77$$ × 10−8). ## 2.5. Association of oGRS with the Prevalence of Metabolically Unhealthy Obesity and with the Age of Individuals Affected Based on oGRS, three genetic risk categories were formed (low, medium, and high), for which a significant trend was observed between higher genetic risk and a higher proportion of MUO individuals (prevalence of MUO in oGRSlow: $53.8\%$, $95\%$ CI: 43.6–63.7; oGRSmedium: $85.2\%$, $95\%$ CI: 80.3–89.3; and oGRShigh: $95.6\%$, $95\%$ CI: 88.7–98.7; p for trend <0.001). In the absence of knowledge of the exact time of MUO onset, Cox regression analyses were performed using the individuals’ age at the time that the questionnaire was recorded. Cox regression analysis showed that oGRS as a continuous variable was significantly associated with an increased risk of developing MUO earlier (hazard ratio (HR) = 1.10, $95\%$ CI: 1.05–1.15; $p \leq 0.001$). Among the genetic risk categories based on the oGRS, both in the medium (HR = 1.62, $95\%$ CI: 1.18–2.21; $$p \leq 0.002$$) and the high (HR = 1.83, $95\%$ CI: 1.26–2.65; $$p \leq 0.001$$) risk groups the risk of developing MHO at a younger age was significantly higher compared to the low-risk group (Figure 5). ## 2.6. Results of Trend and Multivariate Logistic Regression Analyses on the Association of oGRSs with the Metabolic Status The results of the trend analyses show a significant increasing tendency in the average oGRS and oGRS4 values of the BMI subgroups in the metabolically unhealthy individuals ($p \leq 0.001$). Among the metabolically healthy individuals, only the average value of oGRS showed a significant result after p-value adjustment (Table 2). The results of multivariate logistic regression analyses showed that oGRS, both separately (OR = 1.10, $$p \leq 0.00257$$) and in combination with BMI (OR = 1.07, $p \leq 0.001$), significantly increased the risk of metabolically unhealthy status in the total population (obese and non-obese together). oGRS4 separately did not show a significant association (OR = 1.15, $$p \leq 0.012$$) with metabolically unhealthy status, only in combination with BMI (OR = 1.01, $p \leq 0.001$) (Table 3). ## 3. Discussion Our study is the first one to assess the genetic background variations that differentiate the MHO phenotype from the MUO one in a sample of Hungarian adults, based on genetic risk models involving SNPs associated with glucose homeostasis, lipid metabolism, and adiposity. In the present study, we found a very high prevalence of MUO in the Hungarian obese sample population ($79.6\%$). It is known that genetic factors, which affect metabolic pathways involved in adipogenesis, fat distribution, insulin signaling, and insulin resistance, can modulate the predisposition of developing obesity-related complications and lead to MUO [28]. In our oGRS models, we involved SNPs associated exactly with these CM traits to elucidate the links between the genetic background and the transition of MHO to MUO among individuals suffering from obesity. The combined use of 19 SNPs (in 15 genes) that we examined showed a strong association of genetic factors with the MUO phenotype. In a GWAS-based study involving nearly 50,000 Koreans, it was found that polymorphisms in the LPL, APOA5, and CETP genes are associated with a higher risk of the metabolically unhealthy phenotype in the obese [29]. Furthermore, it was also shown that polymorphisms in the CDKN2B gene are also associated with the metabolically unhealthy phenotype even in normal-weight subjects. These results are in harmony with our findings. Based on the gene–gene interaction analysis, the fifteen genes identified can be grouped into three clusters. The first cluster contains five genes related to lipid metabolism (CETP, LIPG, APOB, LPL, and LIPC) and one gene related to glucose metabolism (ADIPOQ). The LPL and APIPOQ genes of this cluster are associated with obesity induced by the consumption of high-fat foods [30] and show a strong association with the PPARG gene, which forms the second cluster. The results of an experiment in rodents suggest that the expression pattern of the PPARG gene is associated with high fat intake, adipocyte development, and insulin resistance [31]. The direct effects of the ADIPOQ and PPARG genes on plasma lipid profile and adiponectin concentration, as well as their interaction with diet, have been demonstrated in humans [32]. Dietary habits influence the association of six genes (C2CD4B, CDKN2B, GIPR, HHEX, SLC2A2, and SLC30A8) forming the third cluster with diabetes, adipogenesis, and cardiovascular risk [33,34,35,36]. Based on these, it is possible to conclude that the direct effects on bio-mechanisms of the gene clusters we identified are likely to be influenced by dietary factors as well. This assumption is further supported by the results of our multivariable logistic regression analyses, which demonstrate that genetic risk (defined as oGRS or oGRS4) in combination with an increase in BMI strongly contributes to the development of metabolically unhealthy status. In the present study, we successfully identified a combination of four (rs10838687 in MADD, rs693 in APOB, rs1111875 in HHEX, and rs2000813 in LIPG) out of the 19 SNPs that significantly influence the risk of MUO developing. These sets of SNPs have been shown in previous studies to have significant effects on lipid and carbohydrate metabolisms. The rs10838687 in the MADD gene was found to be associated with a defect in the enzymatic conversion of proinsulin to insulin, resulting in increased fasting glucose levels, and with the development of T2D [37]. Based on our previous results, rs7944584 in linkage disequilibrium with rs10838687 is strongly associated with the early onset of insulin resistance in the Hungarian general and Roma populations [38]. The rs693 in the APOB gene increases cardiovascular risk [39,40] by raising the levels of APOB, TAG, TC, and LDL-C and reducing HDL-C [41] levels. The rs1111875 is located in the HHEX gene, which could be identified as a candidate gene for T2D using a genome-wide association approach. The association between HHEX and T2D has been reported in different ethnic groups [42]. The rs2000813 in the LIPG gene was found to be associated with lipid parameters and cardiovascular risk [43]. Based on our results, the CM markers most strongly associated with the risk of being MUO were BP, fasting glucose, HOMA-IR, and TAG, some of which are components of the metabolic syndrome (MetS). This is in line with a recent study, involving ten different cohorts from seven countries ($$n = 163$$,517 participants), which showed that BP, fasting glucose, and TAG were among the most frequent MetS components seen among MUO participants [44]. In this study, the most frequent MetS component among Finnish subjects suffering from obesity was elevated BP. In two other studies (i.e., one in Iran and one in Spain), dyslipidemia was found as the most frequent MetS component among obese individuals [45,46]. These parameters may be the most important indicators to predict the risk of metabolic deterioration to the MUO phenotype or the preservation of MHO status in the course of time. A meta-analysis including eight longitudinal studies showed that MHO individuals are at increased risk for all-cause mortality in the long term (≥10 years), which indicates that MHO might be an intermediate stage of MUO [47] and people with MHO tend to develop metabolic dysregulation over time and have increased long-term CVD risk [48,49]. A pan-European cohort study (EPIC-CVD) showed that obese individuals without metabolic syndrome were at a higher risk of coronary heart disease than metabolically healthy individuals of normal weight (risk ratio (RR) = 1.28, $95\%$ CI: 1.03–1.58, $$p \leq 0.001$$). Individuals with MHO have also a substantially higher risk for T2D than metabolically healthy individuals with normal weight (RR = 4.03, $95\%$ CI: 2.66–6.09, $p \leq 0.001$). The age difference of nearly 10 years between the MHO and MUO groups in our present study also supports the theory that the MHO is a dynamic condition and can transform into MUO over time [50,51,52], within 5.5 to 10.3 years of follow-up [51,52]. Our study model showed that a moderate to high genetic risk category was significantly associated with a lower mean age of participants with the MUO phenotype, strongly suggesting a link between genetic susceptibility to excessive fat adiposity and elevated CM disease risks in the early onset of obesity. Therefore, defining the variables that may predict the transition from metabolically healthy to unhealthy obesity in a specific population can help identify those who can benefit from it the most. Findings highlight the utility of potential interventions among MHO subjects with a higher susceptibility to MUO as a valid interim target, particularly in Hungary, one of the most obese countries in Europe [53]. In addition to the genetic background, other factors that may contribute to the transition of MHO to MUO have been studied in other populations. A prospective study conducted in a Spanish cohort ($$n = 3$$,052) found that any increase in BMI, waist size, or waist-to-hip ratio contributed to the transition from MHO to MUO, whereas adhering to a healthy dietary pattern, high levels of physical activity, and not smoking contributed to preventing this transition [50]. Future studies in Hungary should focus on refining and ascertaining specific factors that influence susceptibility to the transition to MUO, beyond our model. All individuals suffering from obesity should aim for metabolic health and normal weight. Given our findings, it may be reasonable to consider genetic-based screening for obese or susceptible individuals to slow or even prevent the development of MUO. Early detection can help to avoid or at least mitigate the development of subsequent obesity-related complications (such as diabetes and cardiovascular disease). This approach is supported by the results of a study that examined the efficacy and safety of weight-loss drugs to prevent the development of T2D [54]. Participants were classified into three CM risk groups (low, medium, and high) based on their Cardiometabolic Disease Staging Score and it was found that although the preventive phentermine/topiramate medication reduced the risk of developing diabetes in all groups compared to the placebo (lifestyle intervention only) group, the reduction was significantly greater in the high-risk group compared to the medium and low groups. Therefore, targeting patients at high risk might improve the cost-to-benefit ratio of interventions. It is important for professionals in the field of public health, healthcare research, and clinical practice, as well as patients, to acknowledge the complexity of factors and their interactions that contribute to the manifestation of obesity. This includes not only genetic (monogenic and polygenic), epigenetic, and developmental influences but also a multitude of interactions [55,56,57,58]. Currently, there is a renewed interest in defining models to explain the origins and development of obesity, leading to renewed debate. One proposed model, known as the Energy Balance Model (EBM), views overeating (consuming more calories than expended) as the primary cause of obesity. This model places emphasis on the role of unconscious signaling by the endocrine, metabolic, and nervous systems that control food intake [59], while highlighting the contribution of inexpensive, convenient, high in fat and sugar, “ultra-processed” (go through multiple processes) foods to the development of obesity. On the other hand, the Carbohydrate–Insulin Model (CIM) suggests that the hormonal response to highly processed carbohydrates plays a role in the partitioning of energy in the body, leading to increased deposition of fat in adipose tissue and reducing the calories available for the body’s metabolic needs [60]. This, in turn, can result in overeating to compensate for the sequestered calories. There have been efforts to reconcile these two models and create an integrated “push–pull” model of obesity pathogenesis [61]. Although the debate continues, public health action does not need to wait for a resolution, as both models identify major drivers of obesity and reflect the interactions of genes and the obesogenic environment. The study has some limitations, which should be considered when interpreting our findings. Our results were not replicated and since the current study was performed in a European population, findings may not apply to non-European populations. Replication studies, including other populations, are necessary to confirm our findings and determine their applicability to ethnically diverse groups. Finally, although genetics do not change over time and it is possible to use this information prospectively, our study design remains cross-sectional, requiring future prospective studies to replicate and validate our results. Despite these stated limitations, we believe our study provides valuable information on the genetic characteristics of MUO and MHO phenotypes at a population level. In conclusion, this study provides the first assessment on the genetic background of MHO and MUO phenotypes in a sample of Hungarian adults. Findings support the notion of early identification of individuals at high metabolic risk in populations suffering from obesity and show that in addition to environmental and lifestyle factors, one’s genetic background also has an important role in the development of MUO. Further, prospective study designs are warranted aiming at using genetic risk models not only to stratify the risk of impaired metabolic health among people suffering from obesity but also in normal-weight and overweight people. In summary, our study shows that obesity varies in its impact on metabolic health and renders unfavorable effects, offering a window of opportunity for early targeted public health interventions. ## 4.1. Sample Population and Relevant Parameters Defined The sample was derived from a population-based disease-monitoring program, the General Practitioners’ Morbidity Sentinel Stations Program (GPMSSP) in Hungary [62]. Detailed methods of sampling and the data collection process are thoroughly described in the Hungarian Metabolic Syndrome Survey (HMSS) [63]. In brief, in the present study, 59 GPs from eight counties representing diverse socio-economic regions within Hungary were invited to participate. Medical and socio-demographic characteristics were recorded and physical examinations (weight, height, waist circumference, and blood pressure measurements) were carried out. Blood samples were collected for laboratory tests (including routine diagnostic tests for fasting glucose, insulin, C-reactive protein, HDL-cholesterol, and triacylglycerols) and DNA isolation. HOMA-IR was calculated according to the following formula: fasting insulin (microU/L) x fasting glucose (nM)/22.5. Medications for hypertension, diabetes, and lipid disturbances were also recorded. Initially, data from 1819 participants representing $91\%$ of invited individuals were collected. In the present study, those with complete geno-/phenotype data ($$n = 1282$$) were included. The total population was divided into three subgroups based on BMI: normal weight (BMI < 25; $$n = 440$$), overweight (BMI: 25–< 30; $$n = 444$$), and obese (BMI ≥ 30; $$n = 398$$) (Figure 6). ## 4.2. Defining Metabolically Healthy and Unhealthy Obesity There is no universally accepted definition for the MHO and MUO phenotypes; therefore, MHO and MUO subjects in this analysis were identified by using a combination of classifying criteria established by Wildman et al. [ 26] and Meigs et al. [ 27] (Table 4). This was achieved by considering (a) the robustness of these criteria and (b) the availability of variables in our database. ## 4.3. DNA Isolation, SNP Selection and Genotyping DNA was isolated using a MagNA Pure LC system (Roche Diagnostics, Basel, Switzerland) with a MagNA Pure LC DNA Isolation Kit—Large Volume according to prespecified instructions of the manufacturer. Extracted DNA was eluted in 200 μL MagNA Pure LC DNA Isolation Kit—Large Volume elution buffer. SNPs strongly associated with obesity, lipid metabolism, and glucose homeostasis were identified by screening PubMed, HuGE Navigator, and Ensembl databases. As a result, a total of 67 SNPs (in 44 genes) were considered (Supplementary Table S1), of which (a) 23 were most strongly associated with obesity [64], (b) 22 with lipid metabolism [65], and (c) 22 with glucose homeostasis [66]. Concerning the effects of SNPs regarding the metabolic traits, overlaps are possible. Genotyping was performed using the MassARRAY platform (Sequenom Inc., San Diego, CA, USA) with iPLEX Gold chemistry by the Mutation Analysis Core Facility at the Karolinska University Hospital, Sweden. Validation, concordance analysis, and quality control were conducted by the Facility according to their protocols. ## 4.4. Identification and Coding of the Genetic Model Best Associated with HOMA-IR by SNPs For each SNP, three widely used genetic inheritance models (i.e., codominant, dominant, and recessive) were examined to determine which model had the strongest association with MUO as a binary outcome (i.e., MUO vs. MHO). Multivariable logistic analysis (controlled for age, sex, and education) was conducted to test each SNP’s association with MUO. Cox and Snell R2 (the higher the better) and p-values (the lower the better) guided the selection process of the best-fitted heritability model [67]. We considered the most suitable heritability model associated with MUO for each SNP used in the optimized genetic risk score (oGRS). Coding for each SNP was based on the following genetic model of inheritance criteria:a)Codominant genetic model: homozygote genotype with risk allele was labelled as 2, whereas heterozygote gene labelled as 1 and 0 was coded for no risk allele.b)Dominant genetic model: 2 was coded for the presence of one or two risk alleles and 0 was coded for the absence of a risk allele.c)Recessive genetic model: 2 was counted for the presence of two risk alleles, while 0 was counted for the homozygote gene with the absence of a risk allele and for the heterozygote gene. ## 4.5. Calculation and Optimization of the Genetic Risk Score The oGRS was calculated using the following equation:oGRS=∑$i = 1$IGi where Gi is the risk score according to the chosen heritability model (see the previous subsection). The genetic risk model optimization procedure selected SNPs with the strongest association with MUO (as a binary outcome variable). GRS optimization was performed using multivariable logistic analyses. The SNPs were tested in ascending order of p-value and each SNP was inserted into the model successively, starting from the SNP with the strongest association (lowest p-value), and the association between oGRS and MUO was examined after each succession. SNPs were selected and used for the final optimized GRS only if they increased the strength of association of oGRS with MUO. SNPs that did not affect or weakened the association were excluded from further analyses. Based on the oGRS, individuals were classified into three genetic risk groups based on tertiles. Individuals in the lowest tertile were assigned to the low-risk group, the individuals in the second tertile were assigned to the moderate-risk group, and the individuals in the third tertile were assigned to the high-risk group. ## 4.6. Statistical Analysis The statistical procedures used to develop the genetic risk were tested and developed on the obesity sample population ($$n = 398$$), while its interaction with and separately from BMI was tested on the total population (obese and non-obese subpopulation, $$n = 1282$$). Chi-square (χ2) was used to test the Hardy–Weinberg equilibrium (HWE) of genotyped SNPs and compare differences between non-quantitative variables within the study population. The Shapiro–Wilk test was used to examine whether the quantitative variables are normally distributed, and, if necessary, Templeton’s two-step method [68] was considered to transform the non-normal variables into normal ones. The Mann–Whitney U test was used to assess the distribution of non-normally distributed data between the study groups. Multivariable logistic analyses were used to determine the association between individual SNPs, the aggregate of them (oGRS), and MUO. Cox regression analysis was used to examine the association of oGRS with age at the onset of MUO. In these analyses, the age of the individual at the time the questionnaire was collected was used as the outcome variable. All regression analyses were carried out under an adjusted model. The online software Search Tool for the Retrieval of Interacting genes (STRING-version 11.5; https://string-db.org/; accessed on 10 January 2023) was used for interaction and cluster analysis and visualization of genes and proteins [69]. A minimum interaction score of 0.400 was used and Markov Cluster Algorithm was applied for determining clusters. The association of risk groups (low, medium, and high) based on oGRS scores with age-at-event (age at identifying MUO in the survey) was examined using multivariable logistic regression (i.e., adjusted for age, sex, BMI, and education). A statistically significant trend between the proportion of individuals with MUO and oGRS risk categories was tested with the Jonckheere-Terpstra test [70]. Receiver operating characteristic (ROC) analysis was employed to evaluate the discriminatory ability of the oGRS and the area under the curve (AUC) was used as an indicator of diagnostic accuracy. In addition, the minimum number of SNPs for which the discrimination accuracy is not significantly different compared to the oGRS was also determined on the bases of the ROC curves’ analyses and the effect of them was also examined. Multivariate logistic regression analysis was used to investigate the association between the genetic risk (defined by oGRS or oGRS4) and its interaction with BMI and the risk of developing metabolically unhealthy conditions for the total study population. In the statistical analysis of interaction variables, adjustments were made for age, sex, education, BMI, and oGRSs. Statistical tests were carried out using IBM SPSS version 26 statistics for Windows (Armonk, NY, United States). Bonferroni-corrected p-value was established for the case where several dependent or independent statistical tests were performed simultaneously on a single data set. ## References 1. Swinburn B.A., Kraak V.I., Allender S., Atkins V.J., Baker P.I., Bogard J.R., Brinsden H., Calvillo A., De Schutter O., Devarajan R.. The Global Syndemic of Obesity, Undernutrition, and Climate Change: The Lancet Commission report. Lancet (2019) 393 791-846. DOI: 10.1016/S0140-6736(18)32822-8 2. Wang L., Southerland J., Wang K., Bailey B.A., Alamian A., Stevens M.A., Wang Y.. Ethnic Differences in Risk Factors for Obesity among Adults in California, the United States. J. Obes. (2017) 2017 2427483. DOI: 10.1155/2017/2427483 3. Shai I., Jiang R., Manson J.E., Stampfer M.J., Willett W.C., Colditz G.A., Hu F.B.. Ethnicity, obesity, and risk of type 2 diabetes in women: A 20-year follow-up study. Diabetes Care (2006) 29 1585-1590. DOI: 10.2337/dc06-0057 4. Wen C.P., David Cheng T.Y., Tsai S.P., Chan H.T., Hsu H.L., Hsu C.C., Eriksen M.P.. Are Asians at greater mortality risks for being overweight than Caucasians? Redefining obesity for Asians. Public Health Nutr. (2009) 12 497-506. DOI: 10.1017/S1368980008002802 5. Pischon T., Boeing H., Hoffmann K., Bergmann M., Schulze M.B., Overvad K., van der Schouw Y.T., Spencer E., Moons K.G., Tjonneland A.. General and abdominal adiposity and risk of death in Europe. N. Engl. J. Med. (2008) 359 2105-2120. DOI: 10.1056/NEJMoa0801891 6. Malik V.S., Willett W.C., Hu F.B.. Global obesity: Trends, risk factors and policy implications. Nat. Rev. Endocrinol. (2013) 9 13-27. DOI: 10.1038/nrendo.2012.199 7. Stefan N., Haring H.U., Hu F.B., Schulze M.B.. Metabolically healthy obesity: Epidemiology, mechanisms, and clinical implications. Lancet Diabetes Endocrinol. (2013) 1 152-162. DOI: 10.1016/S2213-8587(13)70062-7 8. Finucane M.M., Stevens G.A., Cowan M.J., Danaei G., Lin J.K., Paciorek C.J., Singh G.M., Gutierrez H.R., Lu Y., Bahalim A.N.. National, regional, and global trends in body-mass index since 1980: Systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9.1 million participants. Lancet (2011) 377 557-567. DOI: 10.1016/S0140-6736(10)62037-5 9. Bray G.A., Kim K.K., Wilding J.P.H., World Obesity F.. Obesity: A chronic relapsing progressive disease process. A position statement of the World Obesity Federation. Obes. Rev. (2017) 18 715-723. DOI: 10.1111/obr.12551 10. Sims E.A.. Are there persons who are obese, but metabolically healthy?. Metabolism (2001) 50 1499-1504. DOI: 10.1053/meta.2001.27213 11. Tsatsoulis A., Paschou S.A.. Metabolically Healthy Obesity: Criteria, Epidemiology, Controversies, and Consequences. Curr. Obes. Rep. (2020) 9 109-120. DOI: 10.1007/s13679-020-00375-0 12. Bluher M.. Metabolically Healthy Obesity. Endocr. Rev. (2020) 41 bnaa004. DOI: 10.1210/endrev/bnaa004 13. Eckel N., Meidtner K., Kalle-Uhlmann T., Stefan N., Schulze M.B.. Metabolically healthy obesity and cardiovascular events: A systematic review and meta-analysis. Eur. J. Prev. Cardiol (2016) 23 956-966. DOI: 10.1177/2047487315623884 14. Stefan N., Haring H.U., Schulze M.B.. Metabolically healthy obesity: The low-hanging fruit in obesity treatment?. Lancet Diabetes Endocrinol. (2018) 6 249-258. DOI: 10.1016/S2213-8587(17)30292-9 15. Zembic A., Eckel N., Stefan N., Baudry J., Schulze M.B.. An Empirically Derived Definition of Metabolically Healthy Obesity Based on Risk of Cardiovascular and Total Mortality. JAMA Netw. Open (2021) 4 e218505. DOI: 10.1001/jamanetworkopen.2021.8505 16. Rey-Lopez J.P., de Rezende L.F., Pastor-Valero M., Tess B.H.. The prevalence of metabolically healthy obesity: A systematic review and critical evaluation of the definitions used. Obes. Rev. (2014) 15 781-790. DOI: 10.1111/obr.12198 17. Albuquerque D., Nobrega C., Manco L., Padez C.. The contribution of genetics and environment to obesity. Br. Med. Bull. (2017) 123 159-173. DOI: 10.1093/bmb/ldx022 18. Herrera B.M., Lindgren C.M.. The genetics of obesity. Curr. Diab Rep. (2010) 10 498-505. DOI: 10.1007/s11892-010-0153-z 19. Goodarzi M.O.. Genetics of obesity: What genetic association studies have taught us about the biology of obesity and its complications. Lancet Diabetes Endocrinol. (2018) 6 223-236. DOI: 10.1016/S2213-8587(17)30200-0 20. Navarro E., Funtikova A.N., Fito M., Schroder H.. Can metabolically healthy obesity be explained by diet, genetics, and inflammation?. Mol. Nutr. Food Res. (2015) 59 75-93. DOI: 10.1002/mnfr.201400521 21. Berezina A., Belyaeva O., Berkovich O., Baranova E., Karonova T., Bazhenova E., Brovin D., Grineva E., Shlyakhto E.. Prevalence, Risk Factors, and Genetic Traits in Metabolically Healthy and Unhealthy Obese Individuals. Biomed. Res. Int. (2015) 2015 548734. DOI: 10.1155/2015/548734 22. Sedaghati-Khayat B., Barzin M., Akbarzadeh M., Guity K., Fallah M.S., Pourhassan H., Azizi F., Daneshpour M.S.. Lack of association between FTO gene variations and metabolic healthy obese (MHO) phenotype: Tehran Cardio-metabolic Genetic Study (TCGS). Eat Weight Disord. (2020) 25 25-35. DOI: 10.1007/s40519-018-0493-2 23. Dixon J.B.. The effect of obesity on health outcomes. Mol. Cell Endocrinol. (2010) 316 104-108. DOI: 10.1016/j.mce.2009.07.008 24. Zhang Z.Y., Wang M.W.. Obesity, a health burden of a global nature. Acta Pharmacol. Sin. (2012) 33 145-147. DOI: 10.1038/aps.2011.185 25. Okunogbe A., Nugent R., Spencer G., Ralston J., Wilding J.. Economic impacts of overweight and obesity: Current and future estimates for eight countries. BMJ Glob. Health (2021) 6 e006351. DOI: 10.1136/bmjgh-2021-006351 26. Wildman R.P., Muntner P., Reynolds K., McGinn A.P., Rajpathak S., Wylie-Rosett J., Sowers M.R.. The obese without cardiometabolic risk factor clustering and the normal weight with cardiometabolic risk factor clustering: Prevalence and correlates of 2 phenotypes among the US population (NHANES 1999–2004). Arch Intern. Med. (2008) 168 1617-1624. DOI: 10.1001/archinte.168.15.1617 27. Meigs J.B., Wilson P.W., Fox C.S., Vasan R.S., Nathan D.M., Sullivan L.M., D’Agostino R.B.. Body mass index, metabolic syndrome, and risk of type 2 diabetes or cardiovascular disease. J. Clin. Endocrinol. Metab. (2006) 91 2906-2912. DOI: 10.1210/jc.2006-0594 28. Huang L.O., Loos R.J.F., Kilpelainen T.O.. Evidence of genetic predisposition for metabolically healthy obesity and metabolically obese normal weight. Physiol. Genom. (2018) 50 169-178. DOI: 10.1152/physiolgenomics.00044.2017 29. Park J.M., Park D.H., Song Y., Kim J.O., Choi J.E., Kwon Y.J., Kim S.J., Lee J.W., Hong K.W.. Understanding the genetic architecture of the metabolically unhealthy normal weight and metabolically healthy obese phenotypes in a Korean population. Sci. Rep. (2021) 11 2279. DOI: 10.1038/s41598-021-81940-y 30. Walton R.G., Zhu B.B., Unal R., Spencer M., Sunkara M., Morris A.J., Charnigo R., Katz W.S., Daugherty A., Howatt D.A.. Increasing Adipocyte Lipoprotein Lipase Improves Glucose Metabolism in High Fat Diet-induced Obesity. J. Biol. Chem. (2015) 290 11547-11556. DOI: 10.1074/jbc.M114.628487 31. VidalPuig A., JimenezLinan M., Lowell B.B., Hamann A., Hu E., Spiegelman B., Flier J.S., Moller D.E.. Regulation of PPAR gamma gene expression by nutrition and obesity in rodents. J. Clin. Investig. (1996) 97 2553-2561. DOI: 10.1172/JCI118703 32. AlSaleh A., Sanders T.A.B., O’Dell S.D.. Postgraduate Symposium Effect of interaction between PPARG, PPARA and ADIPOQ gene variants and dietary fatty acids on plasma lipid profile and adiponectin concentration in a large intervention study. Proc. Nutr. Soc. (2012) 71 141-153. DOI: 10.1017/S0029665111003181 33. Evseeva M.N., Dyikanov D.T., Karagyaur M.N., Prikazchikova T.A., Sheptulina A.F., Balashova M.S., Zatsepin T.S., Rubtsov Y.P., Kulebyakin K.Y.. Hematopoietically-expressed homeobox protein HHEX regulates adipogenesis in preadipocytes. Biochimie (2021) 185 68-77. DOI: 10.1016/j.biochi.2021.02.011 34. Narasimhan A., Chinnaiyan M., Karundevi B.. Ferulic acid regulates hepatic GLUT2 gene expression in high fat and fructose-induced type-2 diabetic adult male rat. Eur. J. Pharmacol. (2015) 761 391-397. DOI: 10.1016/j.ejphar.2015.04.043 35. Mehramiz M., Ghasemi F., Esmaily H., Tayefi M., Hassanian S.M., Sadeghzade M., Sadabadi F., Moohebati M., Azarpazhooh M.R., Parizadeh S.M.R.. Interaction between a variant of CDKN2A/B-gene with lifestyle factors in determining dyslipidemia and estimated cardiovascular risk: A step toward personalized nutrition. Clin. Nutr. (2018) 37 254-261. DOI: 10.1016/j.clnu.2016.12.018 36. Qi Q.B., Bray G.A., Hu F.B., Sacks F.M., Qi L.. Weight-loss diets modify glucose-dependent insulinotropic polypeptide receptor rs2287019 genotype effects on changes in body weight, fasting glucose, and insulin resistance: The Preventing Overweight Using Novel Dietary Strategies trial. Am. J. Clin. Nutr. (2012) 95 506-513. DOI: 10.3945/ajcn.111.025270 37. Strawbridge R.J., Dupuis J., Prokopenko I., Barker A., Ahlqvist E., Rybin D., Petrie J.R., Travers M.E., Bouatia-Naji N., Dimas A.S.. Genome-wide association identifies nine common variants associated with fasting proinsulin levels and provides new insights into the pathophysiology of type 2 diabetes. Diabetes (2011) 60 2624-2634. DOI: 10.2337/db11-0415 38. Piko P., Werissa N.A., Fiatal S., Sandor J., Adany R.. Impact of Genetic Factors on the Age of Onset for Type 2 Diabetes Mellitus in Addition to the Conventional Risk Factors. J. Pers. Med. (2020) 11. DOI: 10.3390/jpm11010006 39. Park M.H., Kim N., Lee J.Y., Park H.Y.. Genetic loci associated with lipid concentrations and cardiovascular risk factors in the Korean population. J. Med. Genet. (2011) 48 10-15. DOI: 10.1136/jmg.2010.081000 40. Nikolajevic Starcevic J., Santl Letonja M., Praznikar Z.J., Makuc J., Vujkovac A.C., Petrovic D.. Polymorphisms XbaI (rs693) and EcoRI (rs1042031) of the ApoB gene are associated with carotid plaques but not with carotid intima-media thickness in patients with diabetes mellitus type 2. Vasa (2014) 43 171-180. DOI: 10.1024/0301-1526/a000346 41. Niu C., Luo Z., Yu L., Yang Y., Chen Y., Luo X., Lai F., Song Y.. Associations of the APOB rs693 and rs17240441 polymorphisms with plasma APOB and lipid levels: A meta-analysis. Lipids Health Dis. (2017) 16 166. DOI: 10.1186/s12944-017-0558-7 42. Wang Y., Qiao W., Zhao X., Tao M.. Quantitative assessment of the influence of hematopoietically expressed homeobox variant (rs1111875) on type 2 diabetes risk. Mol. Genet. Metab. (2011) 102 194-199. DOI: 10.1016/j.ymgme.2010.09.013 43. Abudureyimu S., Abulaiti P., Li H., Xing Z., Liu S., Li W., Gao Y.. Roles of endothelial lipase gene related single nucleotide polymorphisms in patients with coronary artery disease. Gene (2021) 788 145669. DOI: 10.1016/j.gene.2021.145669 44. Van Vliet-Ostaptchouk J.V., Nuotio M.L., Slagter S.N., Doiron D., Fischer K., Foco L., Gaye A., Gogele M., Heier M., Hiekkalinna T.. The prevalence of metabolic syndrome and metabolically healthy obesity in Europe: A collaborative analysis of ten large cohort studies. BMC Endocr. Disord. (2014) 14. DOI: 10.1186/1472-6823-14-9 45. Tabatabaei-Malazy O., Saeedi Moghaddam S., Masinaei M., Rezaei N., Mohammadi Fateh S., Dilmaghani-Marand A., Abdolhamidi E., Razi F., Khashayar P., Mahdavihezaveh A.. Association between being metabolically healthy/unhealthy and metabolic syndrome in Iranian adults. PLoS ONE (2022) 17. DOI: 10.1371/journal.pone.0262246 46. Gutierrez-Repiso C., Linares-Pineda T.M., Gonzalez-Jimenez A., Aguilar-Lineros F., Valdes S., Soriguer F., Rojo-Martinez G., Tinahones F.J., Morcillo S.. Epigenetic Biomarkers of Transition from Metabolically Healthy Obesity to Metabolically Unhealthy Obesity Phenotype: A Prospective Study. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms221910417 47. Kramer C.K., Zinman B., Retnakaran R.. Are metabolically healthy overweight and obesity benign conditions?: A systematic review and meta-analysis. Ann. Intern. Med. (2013) 159 758-769. DOI: 10.7326/0003-4819-159-11-201312030-00008 48. Mongraw-Chaffin M., Foster M.C., Anderson C.A.M., Burke G.L., Haq N., Kalyani R.R., Ouyang P., Sibley C.T., Tracy R., Woodward M.. Metabolically Healthy Obesity, Transition to Metabolic Syndrome, and Cardiovascular Risk. J. Am. Coll. Cardiol. (2018) 71 1857-1865. DOI: 10.1016/j.jacc.2018.02.055 49. Deedwania P., Lavie C.J.. Dangers and Long-Term Outcomes in Metabolically Healthy Obesity: The Impact of the Missing Fitness Component. J. Am. Coll. Cardiol. (2018) 71 1866-1868. DOI: 10.1016/j.jacc.2018.02.057 50. Schroder H., Ramos R., Baena-Diez J.M., Mendez M.A., Canal D.J., Fito M., Sala J., Elosua R.. Determinants of the transition from a cardiometabolic normal to abnormal overweight/obese phenotype in a Spanish population. Eur. J. Nutr. (2014) 53 1345-1353. DOI: 10.1007/s00394-013-0635-2 51. Soriguer F., Gutierrez-Repiso C., Rubio-Martin E., Garcia-Fuentes E., Almaraz M.C., Colomo N., Esteva de Antonio I., de Adana M.S., Chaves F.J., Morcillo S.. Metabolically healthy but obese, a matter of time? Findings from the prospective Pizarra study. J. Clin. Endocrinol. Metab. (2013) 98 2318-2325. DOI: 10.1210/jc.2012-4253 52. Appleton S.L., Seaborn C.J., Visvanathan R., Hill C.L., Gill T.K., Taylor A.W., Adams R.J., North West Adelaide Health Study T.. Diabetes and cardiovascular disease outcomes in the metabolically healthy obese phenotype: A cohort study. Diabetes Care (2013) 36 2388-2394. DOI: 10.2337/dc12-1971 53. 53. WHO WHO European Regional Obesity Report 2022WHO Regional Office for EuropeCopenhagen, Denmark2022. WHO European Regional Obesity Report 2022 (2022) 54. Guo F., Garvey W.T.. Cardiometabolic Disease Staging Predicts Effectiveness of Weight-Loss Therapy to Prevent Type 2 Diabetes: Pooled Results From Phase III Clinical Trials Assessing Phentermine/Topiramate Extended Release. Diabetes Care (2017) 40 856-862. DOI: 10.2337/dc17-0088 55. Walley A.J., Asher J.E., Froguel P.. The genetic contribution to non-syndromic human obesity. Nat. Rev. Genet. (2009) 10 431-442. DOI: 10.1038/nrg2594 56. Choquet H., Meyre D.. Genetics of Obesity: What have we Learned?. Curr. Genom. (2011) 12 169-179. DOI: 10.2174/138920211795677895 57. Herrera B.M., Keildson S., Lindgren C.M.. Genetics and epigenetics of obesity. Maturitas (2011) 69 41-49. DOI: 10.1016/j.maturitas.2011.02.018 58. Thompson A.L.. Developmental origins of obesity: Early feeding environments, infant growth, and the intestinal microbiome. Am. J. Hum. Biol. (2012) 24 350-360. DOI: 10.1002/ajhb.22254 59. Hall K.D., Farooqi I.S., Friedman J.M., Klein S., Loos R.J.F., Mangelsdorf D.J., O’Rahilly S., Ravussin E., Redman L.M., Ryan D.H.. The energy balance model of obesity: Beyond calories in, calories out. Am. J. Clin. Nutr. (2022) 115 1243-1254. DOI: 10.1093/ajcn/nqac031 60. Ludwig D.S., Aronne L.J., Astrup A., de Cabo R., Cantley L.C., Friedman M.I., Heymsfield S.B., Johnson J.D., King J.C., Krauss R.M.. The carbohydrate-insulin model: A physiological perspective on the obesity pandemic. Am. J. Clin. Nutr. (2021) 114 1873-1885. DOI: 10.1093/ajcn/nqab270 61. Ludwig D.S., Sorensen T.I.A.. An integrated model of obesity pathogenesis that revisits causal direction. Nat. Rev. Endocrinol. (2022) 18 261-262. DOI: 10.1038/s41574-022-00635-0 62. Szeles G., Voko Z., Jenei T., Kardos L., Pocsai Z., Bajtay A., Papp E., Pasti G., Kosa Z., Molnar I.. A preliminary evaluation of a health monitoring programme in Hungary. Eur. J. Public Health (2005) 15 26-32. DOI: 10.1093/eurpub/cki107 63. Szigethy E., Szeles G., Horvath A., Hidvegi T., Jermendy G., Paragh G., Blasko G., Adany R., Voko Z.. Epidemiology of the metabolic syndrome in Hungary. Public Health (2012) 126 143-149. DOI: 10.1016/j.puhe.2011.11.003 64. Nagy K., Fiatal S., Sandor J., Adany R.. Distinct Penetrance of Obesity-Associated Susceptibility Alleles in the Hungarian General and Roma Populations. Obes. Facts (2017) 10 444-457. DOI: 10.1159/000478094 65. Piko P., Fiatal S., Kosa Z., Sandor J., Adany R.. Genetic factors exist behind the high prevalence of reduced high-density lipoprotein cholesterol levels in the Roma population. Atherosclerosis (2017) 263 119-126. DOI: 10.1016/j.atherosclerosis.2017.05.028 66. Werissa N.A., Piko P., Fiatal S., Kosa Z., Sandor J., Adany R.. SNP-Based Genetic Risk Score Modeling Suggests No Increased Genetic Susceptibility of the Roma Population to Type 2 Diabetes Mellitus. Genes (2019) 10. DOI: 10.3390/genes10110942 67. Salanti G., Southam L., Altshuler D., Ardlie K., Barroso I., Boehnke M., Cornelis M.C., Frayling T.M., Grallert H., Grarup N.. Underlying genetic models of inheritance in established type 2 diabetes associations. Am. J. Epidemiol. (2009) 170 537-545. DOI: 10.1093/aje/kwp145 68. Templeton G.F., Pope M.B., Burney L.L.. The Usefulness of the Two-Step Normality Transformation in Retesting Existing Theories: Evidence on the Productivity Paradox. Data Base Adv. Inf. Syst. (2021) 52 53-64. DOI: 10.1145/3447934.3447939 69. Szklarczyk D., Gable A.L., Lyon D., Junge A., Wyder S., Huerta-Cepas J., Simonovic M., Doncheva N.T., Morris J.H., Bork P.. STRING v11: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets. Nucleic Acids Res. (2019) 47 D607-D613. DOI: 10.1093/nar/gky1131 70. Jonckheere A.R.. A Distribution-free k-sample test against ordered alternatives. Biometrika (1954) 41 133-145. DOI: 10.1093/biomet/41.1-2.133
--- title: Effects of Rowing on Rheological Properties of Blood authors: - Mateusz Mardyła - Aneta Teległów - Bartłomiej Ptaszek - Małgorzata Jekiełek - Grzegorz Mańko - Jakub Marchewka journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049505 doi: 10.3390/ijerph20065159 license: CC BY 4.0 --- # Effects of Rowing on Rheological Properties of Blood ## Abstract The aim of this study was to analyze the selected hematological and rheological indices in female rowers during the competitive season. The study included 10 female rowers (aged 21.2 ± 2.6) and the control group consisted of 10 woman of corresponding age (non-athletes). The examination of athletes took place two times: at the beginning of the season during high endurance low intensity training period in January (baseline) and at the end of the competitive season in October (after). Blood samples taken from all woman were analyzed for hematological and rheological parameters. The training period of rowers during the 10 months resulted in decrease in red blood cell count and RBC deformability, in contrast to an improvement in some rheological functions such a decrease in fibrinogen concentration, plasma viscosity and aggregation index. The training program practice in rowing modulated some hematological and rheological indices. Some of them positively influenced the cardiovascular system and reduced potential risks connected with hard training and dehydration, but others may have followed from overtraining or not enough relaxation time between training units. ## 1. Introduction Based on actual knowledge, there is no doubt that the role of physical activity on the human body is non-substantiable and many works discuss benefits in the prevention of many current diseases [1,2]. Nowadays, much attention concentrates on the protective impact of regular activity on cardiovascular disease as well as cardiovascular adaptation [3]. Due to physical effort, there is increased demand for oxygen, glucose and nutrients. This, in the long run, improves metabolic status connected with the transformation of lipids and carbohydrates [4]. Additional benefits from regular exercise alternate with improvements in blood rheology properties in some clinical situations [5,6]. As well as in medical and sports training, the objective is to improve muscle strength, increase the aerobic capacity and metabolic profile [7]; however, the goals of an athlete’s training program are to increase their sporting abilities. Thus, the efficiency of muscle, circulatory and hematopoiesis systems play an important role in determining the possibilities to achieve high results in sport competitions. Nonetheless, studies present that strenuous exercise might result in cardiovascular incidents in professional athletes [8,9,10]. One of these dangers could be some periodic negative changes related to hemorheology. Due to physical exercise, the blood drop hemoconcentration phenomenon can occur resulting in a loss of extracellular water [11,12]. Red blood cells are almost the half amount of the whole blood volume, and they affect the circulatory fluid dynamics. Blood, regarding its mechanical relevance, is a non-Newtonian liquid, which means that an increase in shear rate (depends on vessel radius and speed of blood flow) causes a decrease in viscosity [11]. It has direct meaning regarding the speed of blood flow. The hemorheology effects of training are seen to be dependent on the time spent training, and relate mainly to short, medium and long term exercise [4]. Erythrocytes (RBC) are characterized by their deformability and aggregation properties. RBC have the ability to shape and reshape erythrocytes and, thanks to this, they can pass the capillaries with a smaller diameter than theirs [13]. Single physical efforts lead to increased heart rate, cardiac output and accordingly this accelerates blood flow to the respiratory and exercising muscles. During enhanced flow in the cardiovascular system, erythrocytes are subject to many forces, mainly in the arteries. Even an increase in hematocrit to $60\%$ will increase blood viscosity, but erythrocytes carry off the capillary system. Red blood cells’ deformability depends on membrane flexibility, cell geometry and internal cytoplasmic viscosity [14]. One of the deformability measurement procedures abuses a laser diffraction method as a fast and simple way to perform an elongation index of red cells [15]. In normal conditions, it is expected to decrease erythrocyte aggregation with reduced deformability, especially in low fibrinogen concentrations [11]. Erythrocytes could bind with each other and form aggregates, called rouleaux formation, and also 3D networks. This process is reversible. When higher shear forces occur in blood vessels, negative surface charge erythrocytes could disaggregate on single cells. RBC aggregation, which is strictly linked with blood flow especially in smaller venous compartments, could be the main agent of vessel resistance. Higher erythrocyte aggregation could increase vascular flow resistance and indirectly induce stronger work of the heart muscle, leading to the development of hypertension [16]. This can lead to cardiovascular consequences as well as some thrombotic diseases as a result of dehydration, increased blood coagulate, high physical work or contraceptive medications [17,18]. Nonetheless, regular physical exercise leads to an improvement in nitric oxide (NO) production by the endothelium and release from red blood cells [18]. As a result, this can develop protective mechanisms such as the regulation of arteriolar vasodilation. To our knowledge, the training effects of the whole rowing season on rheological blood properties was not studied before. Only the work of Teległów et al. [ 19] described the effects of intermittent hypoxic training among professional male rowers on performance, rheological properties of blood and metabolic activity of erythrocytes. It is worth emphasizing that hematological research is usually conducted on male athletes because of the many factors that affect making accurate conclusions among women. Thus, the present study aimed to investigate red blood cell deformability, aggregation, plasma viscosity and concentration of fibrinogen in professional female rowers, in different periods of the season. ## 2.1. Participants The study involved 11 women who trained as professional rowers at the University Sport Krakow Club. The women were aged 18–25 years (mean = 21). In the main season, when the most championships were held, they trained for a total of about 8–10 training units per week, which consisted of main training on-water, gym exercise, ergometer and running. The sportswomen trained all year, except for 3 weeks in August and a few days at the end of December. In the period of the most important sports events (from May to September), they started about every other weekend. To assess the potential changes in blood morphology and rheology, the authors decided to take the samples two times (in January and October) from each competitor. Blood was collected from the ulnar vein, on an empty stomach in the morning hours. In addition, the control group consisted of 10 healthy, non-training women. The research was approved by the Bioethics Committee at the Regional Medical Chamber in Cracow. The participants were also informed about the aim of the research, and they approved their participation. The samples were taken at the Laboratory of Blood Physiology at the University of Physical Education in Cracow. A qualified nurse collected blood into 2 types of Vacuette test tubes: with potassium EDTA and with clot activator for serum. Blood samples were tested for morphological and rheological properties. Blood was used for examination one hour after collecting samples. The measurements of the basic hematological indicators were performed on the HORIBA ABX Micros 60 (USA). ## 2.2. Measurement of Erythrocyte Elongation and Aggregation To study deformability and aggregation indicators, blood samples were analyzed in the LORRCA device (Laser-Optical Rotational Cell Analyzer (R&R Mechatronics, Hoorn, The Netherlands). For the determination of erythrocyte elongation, they have been put into 5 mL of standardized viscous solution Polyvinylpyrrolidone (PVP), $M = 360$,000, osmolality ~300 mOsm/kg. The temperature in Lorca was configured at 37 °C. Then, a sample was injected into the Lorca measuring system and was subject automatically to increasing shear stresses (from 0.3 Pa to 60.30 Pa). The system analyzed the diffraction of light on the blood cells and calculated the elongation index from the special formula:EI=a-ba+b a—is length of the red blood cell; b—is width of the red blood cell. To study aggregation, the same device (LORRCA) was used. A total of 1 mL of blood was previously subject to oxygenation for 10 min. Then, blood was injected into a rotating cylinder in LORCA. The computer was adjusted at 37 °C and it started moving the cylinder with a shear rate of >400 s−1. After 10 s, the cylinder stopped and the aggregation of red blood cells came out. Changes in the intensity of light diffraction and time when they occurred are presented on the syllectogram. The following parameters determining erythrocyte aggregation kinetics were assessed:AI=AA+B×$100\%$ AI [%]—aggregation index; A—area above syllectogram curve; B—area below syllectogram curve. Also analyzed were AMP [au]—total extent of aggregation; T½ [s]—half time kinetics of aggregation. ## 2.3. Determination of Fibrinogen Concentration To indicate plasma fibrin concentration, the Chrom 7-coagulometer (Bio-Ksel, Grudziądz, Poland) was used. To 50 µL of plasma, we added 100 µL of Bio-Ksel PT reagent containing thromboplastin with calcium chloride. This resulted in plasma coagulation and clot formation. After conversion of fibrinogen to fibrin, the light detector analyzed the concentration of fibrinogen. ## 2.4. Blood Plasma Viscosity (BPV) To examine plasma viscosity, we used the Myrenne Roetgen viscometer (D-52159, Myrenne GMBH, Roetgen, Germany). Before measuring the plasma viscosity, calibration was performed using two standard solutions Myrenne NP1 and NP2 for the lower and upper ranges of standard measurements, accordingly, for 1.10 mPas and 1.90 mPas. Then, 0.5 mL of plasma was placed into the measurement capillary. The device detected the time at which plasma passed through the distance from barrier L3 to L4 spectrophotometer, with constant pressure and temperature (37 °C). The normal range for human plasma viscosity is 1.10–1.90 mPas. ## 2.5. Statistical Analysis Continuous variables are presented as mean ± standard deviation (SD) or median and interquartile range (IQR), depending on the normality of distribution. The normality of distribution was tested using the Shapiro–Wilk test. For intergroup comparisons, we used ANOVA or, in the case of not meeting its assumptions, Kruskal–Wallis test followed by post hoc tests: Tukey or Dunn, respectively. Dunnett’s Multiple Comparison was additionally used for testing with respect to the control group. SS$\frac{1}{2}$ and EImax were calculated by fitting SS versus EI to equation, representing Lineweaver–Burke model, using a non-linear, curve-fitting algorithm available in a commercial statistical package (Prism 7.02, GraphPad Software Inc., La Jolla, CA, USA). The methodology was described in detail by Baskurt et al. [ 2013]. Calculations were performed using Statistica 12 (StatSoft®, Tulsa, OK, USA) software. Statistical significance was defined as p ≤ 0.053. ## 3. Results In the study group, we found a significance decrease in red blood cell count (RBC) by $7.82\%$. Additionally, there was a higher erythrocyte level reported in the control group. Mean corpuscular volume (MCV) was higher in the athlete group compared to untrained individuals. Eight months of training resulted in an increase in average mass corpuscular hemoglobin (MCH) by $4.82\%$. The two groups did not differ significantly in the levels of hematocrit (Hct) and hemoglobin (Hgb). The results of the basic hematological measurements are listed in Table 1. Comparing rowers to the control group, we have found statistically significant changes: average RBC count decreased by $7.23\%$ in rowers; average value of MCV (fl) increased by $6.19\%$ in rowers at the first examination (baseline) versus controls; average value of MCV (fl) increased by 3.70 % in rowers at the second examination (after) versus controls; average mass of corpuscular hemoglobin MCH (fmol) increased by 4.82 % in rowers at the second examination (after) versus at first examination (baseline). In hemoglobin and fibrinogen concentration as well as WBC and PLT count, there were not any significant changes in rowers compared to the control group. Blood plasma viscosity The viscosity in the athlete group was meaningfully lower at both examinations compared to the control group ($p \leq 0.05$). Compared to the first examination, there was a notable decrease in blood plasma viscosity of 13.85 % in rowers (baseline) compared to the control group and of $17.70\%$ in rowers (after) compared to the control group (see Table 2 below). Erythrocyte deformability The eight-month professional training was reflected by a significantly lower Elongation Index (EI) in the athlete group at a shear stress level between 0.30 Pa and 15.98 Pa; however, at higher levels (31.03 and 60.3 Pa), there were no differences observed. In a group of non-training women (control), there was a higher erythrocyte deformability noted at all shear stress levels compared to the athlete group (see Table 3 below). Aggregation parameters The degree of total aggregation (AMP) was found to be significantly higher in athletes compared to control groups. We did not observe any significant changes in terms of Aggregation Index (AI) and half time of total aggregation (t$\frac{1}{2}$) in athletes and controls. ## 4. Discussion The main objective of this study was to examine the effects of the changes in the resting morphological and rheological properties of blood in response to a long-term training period. There are enormous data about the impact of different exercises on basic blood morphological parameters and the effects of different energetic character, duration and other factors existing in specific sport disciplines. In rowing, applying high loads can lead to a disturbance of the morphotic elements in the bloodstream. At the beginning of the season, in rowers, the main objective of the implemented training is to enhance general performance and therefore aerobic capacity. Therefore, this time serves to enhance aerobic capacity, among other things, by the increase in red blood cells coming into the circulation. The training process in rowing is divided into classical periods proposed by Matvejev, such as preparatory, starting and transitional [20]. The samples were collected at the beginning of the preparatory period (baseline), where a significant part of the training provided aerobic schemes with cross country skiing, gym exercise and rowing indoors (60–$70\%$ of time). The second collection took place after the last national competition, at a moment of significant decrease in intensity and volume. This is the time when in some athletes the overtraining symptoms can occur. The total number of circulating red blood cells and total hemoglobin concentration are mostly responsible for oxygen transport to muscles via the cardiovascular system [12]. These factors are particularly important in professional athletes competing mainly in endurance sports such as cycling, cross-country skiing, long distance running, rowing and many others. However, RBCs rheological properties such as deformability and aggregation could change as a result of intensive physical training periods. Physical exercise affects the increase in demand and consumption of oxygen by the working tissues (mainly muscles and neurons). The best athletes should have high oxygen uptake to tolerate exhaustive efforts, without any risks for their health. RBC homeostasis is related to changes in hematopoiesis systems, including RBC production, RBC hemolysis, bone marrow activity and iron resources [21]. Basic hematological indicators such as hemoglobin, red blood cell count and hematocrit are strictly connected with the maximal oxygen uptake, which is one of the determinants of aerobic capacity [22]. In general, it is proven that endurance training results in a positive increase in red blood cell count, both in men and women [23]. Nonetheless, it is known that in women, the monthly menstruation cycle can disturb blood circulation, levels of erythrocytes or hemoglobin, and some rheological properties [24,25]. Additionally, hemoglobin concentration in women is strictly linked to the inhibitory impact of estrogens (estradiol) on bone marrow activity [26]. Although the examinations were placed in the intermenstrual cycle, it cannot be excluded for sure that the changes were related to monthly blood loss through menstruation, as well as amenorrhea or oligomenorrhea which is frequent in athletes. It is also worth underlining that regular endurance training seems to accelerate the selective elimination of rigid and old RBCs via the reticuloendothelial macrophage system [27]. The level of reduced hematological indices such as HGB, HCT and iron also could be influenced by the dietary intake, both with sport-related issues, such as the type of sport (endurance/power), and the amount of training [28,29]. The most recent data suggest that supplementation with iron preparations could fill the iron stores (serum ferritin), contributing to simultaneously increasing the energetic efficiency status in female rowers [30]. It is known that highly-trained athletes have changed parameters related to the erythrocytic system, such as higher mean corpuscular volume (MCV), higher hemoglobin concentration, higher 2,3 2,3-bisphosphoglycerate and adenosine triphosphate (ATP) in erythrocytes compared to non-trained subjects [31]. The distinct values enumerate red cell indices with a quicker turnover in hematopoietic systems. Under the regular physical training, there was an observed drop in mean corpuscular hemoglobin concentration (MCHC), which is strictly related to an increase in red blood cell elasticity. Diminished erythrocyte deformability can occur as the action of reactive oxygen species (ROS) intensifies during hard physical exercise or during long exposure to hypoxic conditions [32,33]. Certain researchers have underlined the effects of the adaptation of trained subjects (humans and mice), and no instance of erythrocyte damage has been found as a result of oxidative stress [34,35]. Decreased levels of HGB and PLT, as well as PV, could indicate that there is more plasma (as an effect of training). Interesting findings were provided by the Dunch research team [36], who compared female and male rowers with blood donor candidates. The retrospective analysis performed by comparing the hemoglobin levels in healthy people and athletes confirmed the higher concentration in both male and female athletes than in donors. Hemoglobin concentration measured photometrically, in $10.4\%$ of male rowers and $8.3\%$ of female rowers, was above the recommended level for the competition [36]; whereas, only $3.9\%$ male and $1.9\%$ female blood donors exceeded these values. A hemoglobin concentration of 10.5 mM equals a hematocrit of $51\%$ in males and 9.7 mM equals $47\%$ in females. Boyadijev and Taralov [37] found significant differences in mean red blood cell count, with it being lower in highly trained athletes generally compared to untrained subjects. The same research evidence on lower hematological indices such as hemoglobin and hematocrit and higher mean corpuscular volume was found in female highly trained athletes compared to untrained people. In agreement with this, the authors came to a similar conclusion. In the study of Dellavalle et al. [ 29], considering iron status, the hematological variables including Hb, Hct and mean cell volume were close to those reported in our work. The research of Skarpańska-Stejnborn et al. [ 38] demonstrates that under the influence of intensive exercise in rowers, there are disturbances in iron resources as a result of increased escape through macrophages. Some explorers discovered that also in other groups of athletes, such as footballers, there was a depletion of iron, independent of the training system, which can occur with lower concentration of hemoglobin. In our research, we could not detect these findings because there was not a measured level of iron in the serum. In the literature, we can find different procedures of measurement deformability index. In the present work, we used a method based on a laser diffractometer through the Lorrca device. This could be the reason for little changes in our results in comparison to other researchers who used alternative methods. The results of this paper indicated that regular training performed by professional rowers led to significant alterations in blood plasma viscosity, connected likely with the shift in blood compounds and plasma volume. At both time points, in rowers, there were significant lower plasma viscosity levels compared to untrained individuals. Many research conclusions approved the thesis that under the impact of regular endurance training, there was an enhanced volume of plasma. The reason for that is an increase in albumin concentration, followed by shift from the extracellular space [39]. What is interesting is that four females were tested shortly after 3 weeks of mountain training, which could be another proof of an activity that increases plasma volume and decreases its viscosity. This reaction is strictly connected with the renin–angiotensin–aldosterone axis. Although in the discussed article the authors did not measure the whole plasma volume, it may have corresponded with plasma viscosity. The earlier works concerning this issue [40,41] conclude that plasma viscosity is lower in athletes than in untrained subjects, which depends mainly on fibrinogen concentration. Other measured indices such as plasma total protein, globulin and albumin were also lower in runners, but their impact was probably not strong enough to modulate plasma viscosity. Additionally, plasma renin activity was significantly lower in runners, which corresponds with a lower heart rate and adrenergic tone than was previously described by the other authors [39]. Athletes who train excessively could be exposed to overtraining syndromes (OTS). The main symptoms associated with hemorheology could be the feeling of heavy legs, occurring with concomitant modifications in some hematological parameters such as decrease in red blood cells or hemoglobin. A study by Varlet-Marie et al. showed that overtraining syndromes could be connected with hemorheology disturbances such as mild hyperaggregation and mild hyperviscosity [42]; however, in this study, only a higher amplitude of aggregation was noted in female rowers at the end of the season, without increasing plasma viscosity, suggesting that athletes did not experience OTS. It could be related to a loss of water, increase in hematocrit, increase in sodium and potassium ions in the serum, nitrate in urine and transaminase, as a result of hard training (catabolic intensification) [43]. In natural conditions, the increase in hematocrit correlates with decreased performance; however, it is not the rule in every condition. In athletes, the better indicator of OTS experience is increased plasma viscosity. In our research, we noted a significant decrease in plasma viscosity in the experimental group. There was also a relationship during physical exercise between deformability of red blood cells and deficit of oxygen which is used in energetic processes. This leads to the formation of free radicals and may disturb the flexibility of the RBC membrane [4]. Despite antioxidant mechanisms (RBC have about 30 different enzymes), they are not free from oxidative damage. Additionally, the close neighborhood of leukocytes, activated through exercise, can lead to the structural alteration of erythrocytes [44]. We acknowledge some potential limitations in our study. The limitation of our study is that we did not examine hemorheology changes at different time points during all seasons in athletes. We think that a comparison with similarly practicing rowing men athletes would also provide valuable information on blood rheology and differences/convergence between genders. Moreover, we did not analyze performance indicators (such as cardio-respiratory gas ergospirometry), which could be helpful in the evaluation of training processes and their relationship to blood morphology. ## 5. Conclusions In our results, we did not find any significant changes in the comprehensive overview of blood hemorheology, but we could see some single modifications such as in plasma viscosity, red cell volume and aggregation. In the examined group, ensuring consistency between the time of blood collection, age of athletes and training experience led to us obtaining stable results with marginal deviations between athletes. ## References 1. Paffenbarger R.S., Kampert J.B., Lee I.M., Hyde R.T., Leung R.W., Wing A.L.. Changes in physical activity and other lifeway patterns influencing longevity. Med. Sci. Sports Exerc. (1994) 26 857-865. DOI: 10.1249/00005768-199407000-00008 2. Kojda G., Hambrecht R.. Molecular mechanisms of vascular adaptations to exercise. Physical activity as an effective antiox-idant therapy?. Cardiovasc. Res. (2005) 67 187-197. DOI: 10.1016/j.cardiores.2005.04.032 3. Brun J.F., Khaled S., Rayunaud E., Bouix D., Micallef J.P., Orsetti A.. The triphasic effects of exercise on blood rheology: Which relevance to physiology and pathophysiology?. Clin. Hemorheol. Microcirc. (1998) 19 89-104. PMID: 9849922 4. Brun J.F.. Exercise hemorheology as a three acts play with metabolic actors: Is it of clinical relevance?. Clin. Hemorheol. Microcirc. (2002) 26 155-174. PMID: 12082247 5. Dumortier M., Perez-Martin A., Pierrisnard E., Mercier J., Brun J.F.. Regular exercise (3 × 45 min/wk) decreases plasma viscosity in sedentary obese, insulin resistant patients parallel to an improvement in fitness and a shift in substrate oxidation balance. Clin. Hemo. Microcirc. (2002) 26 219-229 6. Sharman J.E., Brown J., Holland D.J., Macdonald G., Kostner K., Marwick T.H.. Influence of Altered Blood Rheology on Ventricular-Vascular Response to Exercise. Hypertension (2009) 54 1092-1098. DOI: 10.1161/HYPERTENSIONAHA.109.135525 7. Dâmaso A.R., Campos R.M.D.S., Caranti D.A., De Piano A., Fisberg M., Foschini D., Sanches P.D.L., Tock L., Lederman H.M., Tufik S.. Aerobic plus resistance training was more effective in improving the visceral adiposity, metabolic profile and inflammatory markers than aerobic training in obese adolescents. J. Sports Sci. (2014) 32 1435-1445. DOI: 10.1080/02640414.2014.900692 8. Weisel J.W., Litvinov R.I.. Red blood cells: The forgotten player in hemostasis and thrombosis. J. Thromb. Haemost. (2019) 17 271-282. DOI: 10.1111/jth.14360 9. Thompson P.D., Franklin B.A., Balady G.J., Blair S.N., Corrado D., EstesIII N.A.M., Fulton J.E., Gordon N.F., Haskell W.L.. American Heart Association Exercise and Acute Cardiovascular Events: Placing the risks into perspective. Med. Sci. Sports Exerc. (2007) 39 886-897. DOI: 10.1249/mss.0b013e3180574e0e 10. Corrado D., Drezner J., Basso C., Pelliccia A., Thiene G.. Strategies for the prevention of sudden cardiac death during sports. Eur. J. Prev. Cardiol. (2011) 18 197-208. DOI: 10.1177/1741826710389924 11. Baskurt O.K., Meiselman H.J.. Cellular determinants of low-shear blood viscosity. Biorheology (1997) 34 235-247. DOI: 10.3233/BIR-1997-34307 12. Brun J.-F., Connes P., Varlet-Marie E.. Alterations of blood rheology during and after exercise are both consequences and modifiers of body’s adaptation to muscular activity. Sci. Sports (2007) 22 251-266. DOI: 10.1016/j.scispo.2007.09.010 13. Tomaiuolo G.. Biomechanical properties of red blood cells in health and disease towards microfluidics. Biomicrofluidics (2014) 8 051501. DOI: 10.1063/1.4895755 14. Huisjes R., Bogdanova A., van Solinge W.W., Schiffelers R.M., Kaestner L., van Wijk R.. Squeezing for Life—Properties of Red Blood Cell Deformability. Front. Physiol. (2018) 9 656. DOI: 10.3389/fphys.2018.00656 15. Kim J., Lee H., Shin S.. Advances in the measurement of red blood cell deformability: A brief review. J. Cell. Biotechnol. (2015) 1 63-79. DOI: 10.3233/JCB-15007 16. Meiselman H.J.. Hemorheologic alterations in hypertension: Chicken or egg?. Clin. Hemorheol. Microcirc. (1999) 21 195-200. PMID: 10711743 17. Bishop M., Astolfi M., Padegimas E., DeLuca P., Hammoud S.. Venous Thromboembolism Within Professional American Sport Leagues. Orthop. J. Sports Med. (2017) 5 2325967117745530. DOI: 10.1177/2325967117745530 18. Grau M., Cremer J.M., Schmeichel S., Kunkel M., Bloch W.. Comparisons of Blood Parameters, Red Blood Cell Deformability and Circulating Nitric Oxide Between Males and Females Considering Hormonal Contraception: A Longitudinal Gender Study. Front. Physiol. (2018) 9 1835. DOI: 10.3389/fphys.2018.01835 19. Teległów A., Mardyła M., Myszka M., Pałka T., Maciejczyk M., Bujas P., Mucha D., Ptaszek B., Marchewka J.. Effect of Intermittent Hypoxic Training on Selected Biochemical Indicators, Blood Rheological Properties, and Metabolic Activity of Erythrocytes in Rowers. Biology (2022) 11. DOI: 10.3390/biology11101513 20. Bompa T.O., Haff G.G.. Periodization: Theory and Methodology of Training (2009) 21. Smith J.A.. Exercise, Training and Red Blood Cell Turnover. Sports Med. (1995) 19 9-31. DOI: 10.2165/00007256-199519010-00002 22. Astrand I.. Aerobic work capacity in men and women with special reference to age. Acta Physiol. Scand. Suppl. (1960) 49 1-92 23. Hu M., Lin W.. Effects of Exercise Training on Red Blood Cell Production: Implications for Anemia. Acta Haematol. (2012) 127 156-164. DOI: 10.1159/000335620 24. Kameneva M.V., Watach M.J., Borovetz H.S.. Gender difference in rheologic properties of blood and risk of cardiovascular diseases. Clin. Hemorheol. Microcirc. (1999) 21 357-363. PMID: 10711771 25. Murphy W.G.. The sex difference in haemoglobin levels in adults—Mechanisms, causes, and consequences. Blood Rev. (2014) 28 41-47. DOI: 10.1016/j.blre.2013.12.003 26. Jelkmann W.. Regulation of erythropoietin production. J. Physiol. (2011) 589 1251-1258. DOI: 10.1113/jphysiol.2010.195057 27. Brinkmann C., Bizjak D., Bischof S., Latsch J., Brixius K., Bloch W., Grau M.. Endurance training alters enzymatic and rheological properties of red blood cells (RBC) in type 2 diabetic men during in vivo RBC aging. Clin. Hemorheol. Microcirc. (2016) 63 173-184. DOI: 10.3233/CH-151957 28. Haymes E.M., Spillman D.M.. Iron Status of Women Distance Runners, Sprinters, and Control Women. Int. J. Sports Med. (1989) 10 430-433. DOI: 10.1055/s-2007-1024938 29. DellaValle D.M., Haas J.D.. Impact of Iron Depletion without Anemia on Performance in Trained Endurance Athletes at the Beginning of a Training Season: A Study of Female Collegiate Rowers. Int. J. Sport Nutr. Exerc. Metab. (2011) 21 501-506. DOI: 10.1123/ijsnem.21.6.501 30. Dellavalle D.M., Haas J.. Iron supplementation improves energetic efficiency in iron-depleted female rowers. Med. Sci. Sports Exerc. (2014) 46 1204-1215. DOI: 10.1249/MSS.0000000000000208 31. Brodthagen U.A., Hansen K., Knudsen J.B., Jordal R., Kristensen O., Pauley P.E.. Red cell 2,3-DPG, ATP, and mean cell volume in highly trained athletes—Effect of long-term submaximal exercise. Eur. J. Appl. Physiol. Occup. Physiol. (1985) 53 334-338. DOI: 10.1007/BF00422849 32. Hakim T.S., Macek A.S.. Effect of hypoxia on erythrocyte deformability in different species. Biorheology (1988) 25 857-868. DOI: 10.3233/BIR-1988-25606 33. Yalcin O., Bor-Kucukatay M., Senturk U.K., Baskurt O.K.. Effects of swimming exercise on red blood cell rheology in trained and untrained rats. J. Appl. Physiol. (2000) 88 2074-2080. DOI: 10.1152/jappl.2000.88.6.2074 34. Senturk U.K., Genturk Ǘ.K., Gunduz F., Kuru O., Aktekin M.R., Kipmen D., Yalcin O., Bor-Kucukatay M., Yesilkaya A., Baskurt O.K.. Exercise-induced oxidative stress affects erythrocytes in sedentary rats but not exercise-trained rats. J. Appl. Physiol. (2001) 91 1999-2004. DOI: 10.1152/jappl.2001.91.5.1999 35. Dane S., Tavsi S., Gul M., Akcay F., Gunal A.. Acute exercise induced oxidative stress is prevented in erythrocytes of male long distance athletes. Biol. Sport. (2008) 25 115 36. Johansson P.I., Ullum H., Jensen K., Secher N.H.. A retrospective cohort study of blood hemoglobin levels in blood donors and competitive rowers. Scand. J. Med. Sci. Sports (2009) 19 92-95. DOI: 10.1111/j.1600-0838.2008.00771.x 37. Birkeland K.I., Stray-Gundersen J., Hemmersbach P., Hallen J., Haug E., Bahr R.. Effect of rhEPO administration on serum levels of sTfR and cycling performance. Med. Sci. Sports Exerc. (2000) 32 1238-1243. DOI: 10.1097/00005768-200007000-00009 38. Skarpańska-Stejnborn A., Basta P., Trzeciak J., Szczęśniak-Pilaczyńska Ł.. Effect of intense physical exercise on hepcidin levels and selected parameters of iron metabolism in rowing athletes. Eur. J. Appl. Physiol. (2015) 115 345-351. DOI: 10.1007/s00421-014-3018-3 39. Convertino V.A., Brock P.J., Keil L.C., Bernauer E.M., Greenleaf J.E.. Exercise training-induced hypervolemia: Role of plasma albumin, renin, and vasopressin. J. Appl. Physiol. (1980) 48 665-669. DOI: 10.1152/jappl.1980.48.4.665 40. Letcher R.L., Pickering T.G., Chien S., Laragh H.. Effects of Exercise on Plasma Viscosity in Athletes and Sedentary Normal Subjects. Clin. Cardiol. (1981) 4 172-179. DOI: 10.1002/clc.4960040405 41. Charm S.E., Paz H., Kurland G.S.. Reduced plasma viscosity among joggers compared with non-joggers. Biorheology (1979) 15 185-191. DOI: 10.3233/BIR-1979-16307 42. Varlet-Marie E.A., Gaudard J., Mercier F., Bressolle J.F.. Brun. Is the Feeling of Heavy Legs in Overtrained Athletes Related to Impaired Hemorheology?. Clin. Hemorheol. Microcirc. (2003) 28 151-159. PMID: 12775897 43. Hong C.Z., Lien I.N.. Metabolic effects of exhaustive training of athletes. Arch. Phys. Med. Rehabil. (1984) 65 362-365. PMID: 6742992 44. Temiz A., Yalcin O., Resmi H., Baskurt O.K.. Can white blood cell activation be one of the major factors that affect hemorhe-ological parameters during and after exercise?. Clin. Hemorheol. Microcirc. (2002) 26 189-193. PMID: 12082250
--- title: The Antimicrobial Peptide AMP-IBP5 Suppresses Dermatitis-like Lesions in a Mouse Model of Atopic Dermatitis through the Low-Density Lipoprotein Receptor-Related Protein-1 Receptor authors: - Hai Le Thanh Nguyen - Ge Peng - Juan Valentin Trujillo-Paez - Hainan Yue - Risa Ikutama - Miho Takahashi - Yoshie Umehara - Ko Okumura - Hideoki Ogawa - Shigaku Ikeda - François Niyonsaba journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049508 doi: 10.3390/ijms24065200 license: CC BY 4.0 --- # The Antimicrobial Peptide AMP-IBP5 Suppresses Dermatitis-like Lesions in a Mouse Model of Atopic Dermatitis through the Low-Density Lipoprotein Receptor-Related Protein-1 Receptor ## Abstract The antimicrobial peptide derived from insulin-like growth factor-binding protein 5 (AMP-IBP5) exhibits antimicrobial activities and immunomodulatory functions in keratinocytes and fibroblasts. However, its role in regulating skin barrier function remains unclear. Here, we investigated the effects of AMP-IBP5 on the skin barrier and its role in the pathogenesis of atopic dermatitis (AD). 2,4-Dinitrochlorobenzene was used to induce AD-like skin inflammation. Transepithelial electrical resistance and permeability assays were used to investigate tight junction (TJ) barrier function in normal human epidermal keratinocytes and mice. AMP-IBP5 increased the expression of TJ-related proteins and their distribution along the intercellular borders. AMP-IBP5 also improved TJ barrier function through activation of the atypical protein kinase C and Rac1 pathways. In AD mice, AMP-IBP5 ameliorated dermatitis-like symptoms restored the expression of TJ-related proteins, suppressed the expression of inflammatory and pruritic cytokines, and improved skin barrier function. Interestingly, the ability of AMP-IBP5 to alleviate inflammation and improve skin barrier function in AD mice was abolished in mice treated with an antagonist of the low-density lipoprotein receptor-related protein-1 (LRP1) receptor. Collectively, these findings indicate that AMP-IBP5 may ameliorate AD-like inflammation and enhance skin barrier function through LRP1, suggesting a possible role for AMP-IBP5 in the treatment of AD. ## 1. Introduction Atopic dermatitis (AD) is a common chronic inflammatory skin disease with a complex and multifactorial etiology involving genetic, immunological, and environmental factors that elicit immune dysfunction and skin barrier disruption [1,2]. Cumulative evidence has shown that, similar to immune dysfunction, disruption of the skin barrier plays a central role in the pathogenesis of AD. Impairment of tight junctions (TJs) has consistently been reported to be a key player in AD pathogenesis, as TJs interconnect with other components of the skin barrier [3,4,5]. TJs are intercellular junctions located within the stratum granulosum (SG) and comprise the claudin, occludin, and zonula occludens (ZO) protein families. TJs are thought to shape a paracellular permeability barrier regulating the passage of water, ions, and solutes [6]. De Benedetto et al. reported a significant reduction in the expression of claudin-1, an important TJ-related protein in the epidermal barrier, in subjects with AD, suggesting the deficiency of TJs in AD patients [7]. Claudin-1 deficiency was demonstrated to cause increases in transepidermal water loss (TEWL) levels in mice, accompanied by notable decreases in skin barrier function [8]. In addition, TJ barrier defects were indicated to inhibit the formation of the stratum corneum (SC) by altering polar lipid and profilaggrin processing, suggesting that TJ failure might be correlated with dysfunction of the SC barrier in AD pathogenesis [9,10]. Antimicrobial peptides (AMPs) constitute the first-line of defense in the skin against pathogens. In addition to exhibiting wide-ranging killing activity against pathogenic microorganisms, AMPs also exhibit immunomodulatory activity by inducing cell proliferation, migration, and differentiation [11]. They regulate cytokine/chemokine production, enhance angiogenesis and wound healing, and maintain skin barrier function [11,12,13,14]. The antimicrobial peptide derived from insulin-like growth factor-binding protein 5 (AMP-IBP5) is a proteolytic cleavage product of insulin-like growth factor-binding protein (IGFBP)-5, which belongs to the IGFBP family. This family consists of six proteins, namely, IGFBP-1 to IGFBP-6 [15,16]. IGFBP-5 is expressed in various cells, including keratinocytes and fibroblasts [16,17]. AMP-IBP5 has been found to accelerate diabetic wound healing through its protective effects against glucotoxicity and promoting effects on angiogenesis. These functions reveal a promising role of AMP-IBP5 in the management of chronic wounds [18]. Additionally, AMP-IBP5 was reported to induce the proliferation and migration of both keratinocytes and fibroblasts through the receptor low-density lipoprotein receptor-related protein-1 (LRP1) [19]. LRP1 is a member of the low-density lipoprotein (LDL) receptor family. It comprises a membrane-anchoring 85-kDa subunit and a 515-kDa extracellular chain with four ligand-binding clusters [20,21]. LRP1 is expressed in several organs, including liver, lung, brain, and skin [22]. In the skin, LRP1 expression is detected mainly in the SG of the epidermis and in dermal fibroblasts [23]. Various AMPs, such as LL-37, human β defensins (hBDs), and S100A7, have been demonstrated to be involved in the mechanism regulating TJ barrier function in keratinocytes in vitro [24,25,26,27]. Because skin barrier dysfunction plays a central role in the pathogenesis of AD [28,29,30], AMPs might have therapeutic promise in the management of AD. Here, we aimed to clarify the effect of AMP-IBP5 on skin barrier function in the context of AD and to identify its underlying mechanism in order to determine a possible therapeutic approach for AD. ## 2.1. AMP-IBP5 Improves TJ Barrier Function We first investigated whether AMP-IBP5 can affect the expression of TJ-related proteins in human primary keratinocytes. Western blot analysis revealed that AMP-IBP5 markedly upregulated claudin-1, -4, and -7, occludin and ZO-1 expression (Figure 1). Because the intercellular distribution of TJ-related proteins is indispensable for the formation of functional TJs, the effect of AMP-IBP5 on the intercellular localization of TJ-related proteins was examined. AMP-IBP5 induced a broad distribution of claudin-1, -4, and -7, occludin and ZO-1 along intercellular borders (Figure S1A). In addition, to examine whether the AMP-IBP5-mediated intercellular localization of TJ-related proteins is indeed associated with TJ barrier function, we assessed the effect of AMP-IBP5 on transepithelial electrical resistance (TER), a sensitive measure of TJ barrier function [31,32]. Upon stimulation of keratinocyte monolayers with AMP-IBP5, the TER values dose- and time-dependently increased and peaked at 48 h, before gradually declining (Figure S1B). Collectively, our results indicate that AMP-IBP5 enhances the formation and function of the TJ barrier. ## 2.2. AMP-IBP5 Enhances TJ Barrier Function through Activation of Atypical Protein Kinase C (aPKC) ζ and Ras-Related C3 Botulinum Toxin Substrate 1 (Rac1) The two isoforms of the aPKC, namely, aPKCζ and aPKCi/λ, have been indicated to be associated with TJ barrier function [33,34]. Furthermore, activation of guanosine triphosphate (GTP)-bound Rac1 (GTP-Rac1) is fundamental for the formation of the TJ barrier and the maturation of epidermal keratinocytes [35]. Additionally, GTP-Rac1 activation is essential for the activation of aPKCζ [36]. Thus, we investigated whether AMP-IBP5 regulates TJ barrier function through activation of aPKC and GTP-Rac1. Treatment of human keratinocytes with AMP-IBP5 markedly increased the phosphorylation of aPKCζ, which peaked at 30 min. Phosphorylation of Rac1 (at Ser71) was observed 1 and 2 h post-stimulation (Figure 2A). Activation of both aPKC and Rac1 was necessary for AMP-IBP5-mediated regulation of the TJ barrier, as demonstrated by the inhibitory effects of GF 109203X, a pan-PKC inhibitor [37], and NSC23766, a specific inhibitor of Rac1 [38]. Indeed, the presence of GF 109203X notably suppressed the protein expression of claudin-1, occludin, and ZO-1 (Figure S2A) and impeded the intercellular localization of these TJ-related proteins (Figure 2B, upper panels). Moreover, treatment of keratinocyte monolayers with GF 109203X significantly suppressed the AMP-IBP5-induced increase in the TER (Figure 2C, left panel). Similarly, NSC23766 noticeably suppressed the expression of claudin-1, occludin, and ZO-1 (Figure S2B), impeded their intercellular localization (Figure 2B, lower panels), and decreased the TER (Figure 2C, right panel) in AMP-IBP5-treated keratinocytes. These observations suggest that activation of aPKCζ and Rac1 is needed for the AMP-IBP5-induced enhancement of TJ barrier function. ## 2.3. AMP-IBP5 Regulates TJ Barrier Function via the LRP1 Receptor AMP-IBP5 was reported to induce the proliferation and migration of keratinocytes and fibroblasts through LRP1 [19]; thus, we hypothesized that AMP-IBP5 might regulate TJ barrier function through LRP1. To confirm our hypothesis, we pretreated keratinocytes with LDL receptor-related protein-associated protein 1 (LRPAP), also named RAP, an antagonist of LRP1 [39]. The presence of RAP significantly impeded the intercellular localization of the TJ-related proteins claudin-1, occludin, and ZO-1 (Figure 3A), and decreased the protein expression of these TJ-related proteins (Figure S3). Additionally, pretreatment of keratinocyte monolayers with RAP suppressed the AMP-IBP5-induced increase in the TER (Figure 3B) and the phosphorylation of aPKCζ and Rac1 (Figure 3C). These findings suggest that AMP-IBP5 regulates TJ barrier function through LRP1 and that aPKCζ and Rac1 function downstream of this receptor. ## 2.4. Treatment with AMP-IBP5 Rescues IL-4/IL-13-Driven TJ Barrier Dysfunction TJ barrier failure, which was indicated to be involved in the pathogenesis of AD, not only disrupts the epidermal barrier but also impacts the immune response [3,4,5,40]. Interestingly, our analysis of microarray data from lesional and non-lesional skin of patients with AD revealed that the gene expression of IGFBP-5, the parent protein of AMP-IBP5, is downregulated in AD skin lesions (Figure 4A). Based on this observation, we examined the expression of IGFBP-5 in a mouse model of 2,4-dinitrochlorobenzene (DNCB)-induced AD-like pathology and found that the mRNA expression of IGFBP-5 was decreased in AD mice compared to normal mice (Figure 4B). This finding suggests that AMP-IBP5 deficiency may play a crucial role in the pathogenesis of AD. To test this hypothesis, we treated human primary keratinocytes with IL-4 and IL-13 to establish an in vitro model that mimics AD features. IL-4/IL-13 treatment of keratinocytes impaired the spontaneous intercellular localization and inhibited the protein expression of claudin-1, occludin, and ZO-1 (Figure 5A,B) and decreased the TER (Figure 5C). Interestingly, the presence of AMP-IBP5 enhanced the intercellular localization and increased the protein expression of TJ-related proteins (Figure 5A,B) and increased the TER of keratinocyte monolayers pretreated with IL-4/IL-13 (Figure 5C). These results indicate that the addition of AMP-IBP5 might rescue IL-4/IL-13-driven TJ barrier dysfunction. ## 2.5. AMP-IBP5 Ameliorates AD Symptoms in Mice with DNCB-Induced AD-like Pathology To further verify whether AMP-IBP5 improves skin barrier function in vivo, we established a model of DNCB-induced AD in BALB/c mice. The lesional ear skin of AD mice was treated subcutaneously with AMP-IBP5. AD model mice exhibited significant increases in dermatitis scores, ear thickness, scratching frequency, and TEWL. Although significant differences remained between normal mice and AMP-IBP5-treated AD mice, these features of AD were markedly improved following AMP-IBP5 treatment (Figure 6A–C). Moreover, the expression of claudin-1, the most important TJ-related protein, was significantly decreased in AD mice, and treatment with AMP-IBP5 noticeably restored claudin-1 expression (Figure 6D and Figure S4). Intriguingly, no differences in the alleviation of dermatitis-like symptoms in the AD mouse model, including dermatitis scores, ear thickness, scratching frequency, TEWL (Figure S5A,B), and the expression of claudin-1 (Figure S5C), were observed between mice treated with AMP-IBP5 via topical application and those treated via injection. To confirm whether the AMP-IBP5-mediated enhancement of TJ protein expression is associated with a functional TJ barrier in AD mice, we performed a TJ permeability assay using an NHS-LC-biotin tracer, as described previously [41]. In normal mice with a functional TJ barrier, the tracer does not pass through the outermost layer of the epidermis, whereas in mice with a disrupted TJ barrier the tracer does so readily [8,42]. As shown in Figure 6E, the penetration of the tracer (red) into the epidermis of normal mice was blocked (arrowheads) by the functional TJ barrier, which is represented by claudin-1 (green dots). However, in AD mice, the tracer was able to pass through the epidermis. As expected, after AMP-IBP5 treatment, tracer penetration was again blocked by the TJ barrier, indicating that AMP-IBP5 treatment restored TJ barrier function in AD mice. Interestingly, while the expression levels of type 2 cytokines IL-4, IL-13, and IL-33 and pruritic cytokines, such as IL-31 and thymic stromal lymphopoietin (TSLP), were increased in AD skin lesions, administration of AMP-IBP5 decreased the expression of type 2 and pruritic cytokines (Figure 7A). Injection of AMP-IBP5 into the skin lesions of AD mice also noticeably reduced the numbers of CD4+ T cells (Figure 7B) and mast cells (Figure 7C) and the total serum immunoglobulin (Ig) E level (Figure 7D). ## 2.6. LRP1 Is Required for the AMP-IBP5-Mediated Amelioration of AD To elucidate the role of LRP1 in the AMP-IBP5-mediated amelioration of AD symptoms, we subcutaneously administered mouse RAP, an antagonist of LRP1, to AD mice to inhibit LRP1. In our model, co-treatment with AMP-IBP5 and RAP did not result in improvements in the dermatitis score, ear thickness, scratching frequency, or TEWL in AD mice (Figure S6 and Figure 8A,B). Furthermore, in the presence of RAP, AMP-IBP5 failed to increase the epidermal expression of claudin-1 (Figure 8C) or enhance skin barrier function, as evaluated by a permeability assay (Figure 8D), in AD mice. Additionally, inhibition of LRP1 in AD mice blocked the effects of AMP-IBP5 on reducing the expression of type 2 and pruritic cytokines and the total IgE serum level (Figure 9A). AMP-IBP5 also failed to reduce the numbers of CD4+ T cells (Figure 9B) and mast cells (Figure 9C) in the presence of RAP in AD mice. Collectively, these results indicate that AMP-IBP5 ameliorates AD through LRP1. ## 3. Discussion Skin barrier dysfunction contributes significantly to the pathogenesis of AD. The TJ barrier consistently plays a crucial role, as it interconnects with other components of the skin barrier [8,40,43]. Interestingly, AMPs such as hBDs, LL-37, and S100A7 have been reported to improve the TJ barrier in vitro in keratinocyte monolayers [24,25,26,27]. Although accumulated evidence implies a promising role for AMPs in the treatment of AD [11,40,44], the precise role of AMPs in the immunopathogenesis of AD remains elusive. In this study, our microarray analysis revealed that the gene expression of IGFBP-5, the parent protein of AMP-IBP5, is lower in skin lesions of patients with AD than in non-lesional skin, suggesting that AMP-IBP5 might be involved in AD pathogenesis. Moreover, we showed that AMP-IBP5 increased the expression and enhanced the intercellular distribution of various TJ-related proteins and improved skin barrier function in both in vitro and in vivo AD models. The crucial role of the SC in the permeability barrier has been extensively demonstrated. However, the importance of the TJ barrier has been a recent research focus, particularly in skin diseases characterized by skin barrier defects, such as AD and psoriasis. In fact, knockout of claudin-1 in mice leads to an increase in TEWL and mortality with marked functional impairment of the skin barrier, indicating the importance of the epidermal barrier and TJ proteins [8]. Claudin-1 expression was found to be significantly decreased in skin lesions from patients with AD compared with healthy skin from nonatopic individuals [45]. Moreover, Tokumasu et al. indicated that claudin-1 orchestrates the features of AD and has a potential role in the pathogenesis, severity, and natural course of AD [46]. Herein, we demonstrated that AMP-IBP5 restores both claudin-1 expression and barrier function in AD mice and ameliorates AD symptoms in these mice, further indicating the crucial role of the TJ barrier in AD. There was no significant difference in alleviation of dermatitis-like symptoms between subcutaneous injection and topical application of AMP-IBP5. According to the 500 Dalton rule for the skin penetration of chemical compounds [47], AMP-IBP5 with a molecular weight of 2655 Dalton hardly penetrates the skin. We speculated that AMP-IBP5 succeeded in penetrating the skin because it was dissolved in acetic acid, which is known as a skin penetration enhancer [48,49]. Moreover, skin barrier impairment in AD may also facilitate the penetration of AMP-IBP5 into the skin. Further studies are required to clarify the topical effect of AMP-IBP5 in AD pathogenesis. LRP1 plays an important role not only in the regulation of lipoprotein metabolism but also in numerous aspects of cell signaling and function [22,50]. In the skin, LRP1 is expressed in various cell types, including keratinocytes, fibroblasts, dendritic cells, and endothelial cells, and is involved in the regulation of skin homeostasis [23,51]. Moreover, LRP1 plays an important role in regulating the host immune response to invading pathogens. In fact, LRP1 is involved in the innate recognition of microbial components and the inflammatory response in macrophages [52]. Here, our findings indicated that LRP1 might participate in regulating the TJ barrier in human keratinocytes and in mice with DNCB-induced dermatitis-like lesions, because AMP-IBP5 failed to ameliorate AD in RAP-treated keratinocytes and AD mice. These data reveal a critical role of LRP1 in the barrier-based pathogenesis of AD. LRP1 functions as a major regulator of the activation of mitogen-activated protein kinases, Rac1, and aPKCζ in several types of cells. For instance, LRP1 participates in the activation of mitogen-activated protein kinase cascades induced by AMP-IBP5, contributing to the proliferation and migration of keratinocytes and fibroblasts [19]. LRP1 was also demonstrated to be involved in Rac1-regulated murine embryonic cell migration [53] and in aPKCζ-regulated chondrocyte differentiation [54]. Therefore, our observation that AMP-IBP5 regulates the TJ barrier in human keratinocytes through activation of aPKCζ and Rac1 is consistent with previous studies. To further clarify the role of LRP1 in skin barrier regulation, we investigated whether AMP-IBP5 administration is effective in AD mice with LRP1 inhibition. AMP-IBP5 administration exhibited no therapeutic effect in AD mice with LRP1 inhibition, confirming that LRP1 is required for the AMP-IBP5-mediated amelioration of AD. To our knowledge, this is the first report about the involvement of LRP1 in the pathogenesis of AD. Further studies are required to investigate the mechanism by which LRP1 is involved in the pathogenesis of AD. Recent studies have revealed the involvement of AMPs in the pathogenesis of inflammatory skin diseases, in which overproduction of AMPs may exacerbate the inflammatory component of the disease. Indeed, hBDs activate T cells and mast cells to produce IL-4, IL-13, and IL-31, characteristic inflammatory cytokines in AD [55,56]. Activated T cells upregulate the production of Th2 cytokines, including IL-31, interferon-γ, and IL-22, in the presence of LL-37, indicating that LL-37 promotes the inflammatory environment in individuals with T-cell-related skin diseases [57]. hBDs and LL-37 were also demonstrated to promote the production of inflammatory cytokines in mast cells [56,58]. In addition to exhibiting proinflammatory activities, AMPs might also exert anti-inflammatory effects. For instance, LL-37 exerts an antagonistic effect on interferon-γ, tumor necrosis factor-α, IL-4, and IL-12 responses in several cell types [59,60,61]. Intriguingly, we demonstrated that AMP-IBP5 administration suppressed the expression of IL-4, IL-13, IL-31, IL-33, and TSLP in mice with DNCB-induced AD. Moreover, AMP-IBP5 treatment reduced the numbers of CD4+ T cells and mast cells and the total serum IgE level, suggesting a possible anti-inflammatory effect of AMP-IBP5 in the AD mouse model. AD is characterized by chronic eczematous lesions, skin dryness, and intense pruritus [62]. Scratching due to chronic pruritus in AD further exacerbates skin barrier dysfunction [63], causes sleep loss, and critically decreases the quality of life of AD patients [64,65]. Notably, hBDs and LL-37 stimulate mast cells to produce IL-31, which is strongly involved in the regulation of itch sensation in patients with AD [56,66]. Although hBDs and LL-37 have also been shown to be involved in the production of Th2 inflammatory and pruritic cytokines [12,67,68,69,70], they also inhibit the production of pruritic cytokines. Indeed, LL-37 promotes the production of semaphorin 3A, a chemorepulsive factor of the epidermal nerves that is downregulated in AD [71], and suppresses TSLP production in keratinocytes [61]. These functions indicate a possible role of LL-37 in itch inhibition in patients with AD. In this study, AMP-IBP5 suppressed the expression of IL-31 and TSLP and ameliorated pruritus in AD mice, suggesting its potential involvement in the pathogenesis of pruritic AD symptoms. However, the exact mechanism by which AMP-IBP5 regulates itch sensation in the context of AD, especially the role of LRP1, needs to be clarified. In summary, we demonstrated that AMP-IBP5 enhanced TJ barrier function in human keratinocytes and ameliorated dermatitis symptoms and restored skin barrier function in a mouse model of AD through LRP1. These results suggest a potential therapeutic approach for AD using AMP-IBP5. ## 4.1. Reagents AMP-IBP5 (AVYLPNCDRKGFYKRKQCKPSR-NH2; molecular weight: 2655.1) was obtained from the Peptide Institute (Osaka, Japan) and was dissolved in acetic acid ($0.01\%$). IgG isotype control and antibodies specific for phosphorylated and unphosphorylated aPKCζ and Rac1 were purchased from Cell Signaling Technology (Beverly, MA, USA). A rabbit monoclonal antibody specific for claudin-1 was purchased from Cell Signaling Technology. Mouse monoclonal antibodies specific for claudin-4, occludin, and ZO-1 were obtained from Invitrogen (Carlsbad, CA, USA). A rabbit polyclonal antibody specific for claudin-7 was purchased from Invitrogen. The aPKCζ inhibitor GF 109203X was obtained from Enzo Life Sciences (Farmingdale, NY, USA). The Rac1 inhibitor NSC23766 was obtained from Calbiochem (La Jolla, CA, USA). Enzyme-linked immunosorbent assay (ELISA) kits were obtained from R&D Systems (Minneapolis, MN, USA). The EZ-linkTM Sulfo-NHS-LC-Biotin tracer was purchased from Thermo Scientific (Waltham, MA, USA). Recombinant human and mouse LRPAP were obtained from R&D Systems. Toluidine blue solution ($0.05\%$, pH 4.1) was purchased from Muto Pure Chemicals Co. Ltd. (Tokyo, Japan). Recombinant human IL-4 and IL-13 were obtained from BioLegend (San Diego, CA, USA). A mouse monoclonal antibody specific for CD4 (PE rat anti-mouse CD4) was purchased from BD Bioscience (San Jose, CA, USA). ## 4.2. Keratinocyte Culture and Stimulation Primary human epidermal keratinocytes were purchased from Kurabo Industries (Osaka, Japan) and were cultured in serum-free HuMedia-KG2 keratinocyte growth medium, as described in a previous study [72]. An increase in the Ca2+ concentration in cultured keratinocytes leads to the formation of TJs and enhancement of skin barrier function [32]. Thus, keratinocytes were cultured in high-Ca2+ (1.8 mM) medium to generate the TJ-forming keratinocytes of the second layer of the SG [73]. To establish the in vitro model with dysfunction of the TJ barrier, keratinocytes were stimulated by the addition of 100 ng/mL recombinant IL-4 and IL-13 in combination with the culture medium, as reported previously [74,75]. ## 4.3. Animals To induce AD-like inflammation in BALB/c mice, a strategy of cutaneous DNCB sensitization and challenge was used. Six-week-old female BALB/c mice were obtained from Japan SLC, Inc. (Tokyo, Japan) and housed under specific pathogen-free controlled conditions. DNBC was applied topically to the same area of skin on each mouse to induce AD-like skin lesions as described in a previous study [76]. In brief, $1\%$ DNCB was applied to the ear skin of mice on Day 4. Four days later, the mice were challenged with $0.4\%$ DNCB on the same area of skin 3 times weekly for 3 weeks (Days 1–19). The total dermatitis score (maximum score 12), representing the clinical severity, was defined as the sum of the individual scores (0 (none), 1 (mild), 2 (moderate), and 3 (severe)) assigned for each of 4 symptoms (erythema/hemorrhage, scaling/dryness, edema, and excoriation/erosion). To reduce variability, clinical symptoms were evaluated by two examiners blinded to the treatment conditions of the groups. The ear thickness of the mice was evaluated by measuring the entire thickness of the ear at the widest point from the outermost layer of one SC to the outermost layer of another SC on the opposite side of the ear using a micrometer (Mitutoyo, Kawasaki, Japan). TEWL was determined by a Tewameter TM Nano measurement probe (Courage + Khazaka electronic GmbH, Köln, Germany). ## 4.4. Treatment of Mice The ear skin of the AD mice was subcutaneously injected with 25 μL of 25 μM AMP-IBP5 on Days 15, 17, and 18. In other experiments, AD mice were subcutaneously co-injected with AMP-IBP5 and 1 μg/mL recombinant mouse LRPAP on Days 15, 17, and 18. To compare the effects of injection and topical treatment, 25 μL of 25 μM AMP-IBP5 was topically applied to the ear skin of AD mice on Days 15, 17, and 18 (Figure S5). On Day 19, serum and skin biopsies were collected and analyzed. ## 4.5. Histological Analysis Mouse ear tissues were fixed with $20\%$ neutral buffered formalin solution, embedded in paraffin, and sectioned with a microtome. The slides were stained with hematoxylin and eosin (H&E) for histopathologic analysis. Mast cells were stained with $0.05\%$ toluidine blue. Images were acquired using a Zeiss microscope (Carl Zeiss, Jena, Germany) and were analyzed with ImageJ software (version 1.52a, National Institutes of Health (NIH), Bethesda, MD, USA). ## 4.6. Western Blot Analysis Samples of human keratinocytes and mouse skin tissues were harvested and were then lysed with RIPA lysis buffer (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM Na2 EDTA, 1 mM EGTA, $1\%$ NP-40, $1\%$ sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, and 1 μg/mL leupeptin). Protein concentrations were measured using Precision Red Advanced Protein Assay reagent, and equal amounts of total protein were separated by electrophoresis on 8–$15\%$ SDS–PAGE gels and then transferred to polyvinylidene fluoride membranes (Merck Millipore, Burlington, MA, USA). ImmunoBlock buffer was used to block the membranes for 1 h at room temperature. After blocking, the membranes were incubated at 4 °C with primary antibodies specific for claudin-1, -4, -7, occludin, and ZO-1. The primary antibodies were then detected using horseradish peroxidase-conjugated sheep anti-rabbit or sheep anti-mouse secondary antibodies. Membranes were developed with Luminata Forte Western horseradish peroxidase substrate (Merck Millipore, Burlington, MA, USA) and were then imaged using Fujifilm LAS-4000 Plus (Fujifilm, Tokyo, Japan). Band densities in the images were quantified using ImageJ. ## 4.7. Total RNA Extraction and Real-Time PCR Total RNA was extracted from keratinocytes and skin tissues using the RNeasy Plus Micro Kit (QIAGEN, Hilden, Germany) and RNeasy Plus Universal Mini Kit (QIAGEN), respectively. Reverse transcription of 1 μg of total RNA to first-strand cDNA was performed using ReverTra Ace qPCR RT Master Mix (Toyobo, Osaka, Japan) or ReverTra Ace qPCR RT Master Mix with gDNA Remover (Toyobo) according to the manufacturer’s instructions. Real-time PCR was performed using the QuantiTect SYBR Green PCR Kit (QIAGEN). The StepOnePlus Real-Time PCR System (Life Technologies, Carlsbad, CA, USA) was used to quantify the mRNA levels in the samples. The primer sequences used in this study are listed in Table 1. All primers were obtained from Thermo Fisher Scientific (Waltham, MA, USA). The target RNA levels were normalized to the endogenous Rps18 reference level, and changes in mRNA expression are reported as fold increases relative to vehicle. ## 4.8. Immunostaining Analysis Keratinocytes were cultured on collagen I-coated chamber slides (BD Biosciences, Bedford, MA, USA) and subsequently fixed with methanol. Protein Block Serum-Free containing $0.2\%$ Tween 20 was used for blocking prior to incubation overnight with the appropriate primary antibodies in $1\%$ bovine serum albumin (BSA) and phosphate-buffered saline (PBS) with $0.2\%$ Tween 20, followed by incubation with specific secondary antibodies conjugated to Alexa 594 (Invitrogen, Carlsbad, CA, USA). Confocal laser scanning microscopy (Carl Zeiss, Jena, Germany) was used for image acquisition. Tissues harvested from mice were directly embedded in an optimal cutting temperature compound, and frozen sections were fixed with preheated $4\%$ paraformaldehyde in PBS for 10 min. The sections were then permeabilized with $0.01\%$ Triton X-100 in PBS for 10 min, blocked with ImmunoBlock for 30 min, and incubated overnight at 4 °C with the appropriate primary antibodies. After incubation with secondary antibodies, the samples were mounted using antifade mountant with NucBlue stain (Invitrogen, Carlsbad, CA, USA). Images were acquired using a Zeiss LSM 780 system with ZEN 2011 software (Carl Zeiss, Jena, Germany). ImageJ software was used to quantify the immunofluorescence intensities in the samples. ## 4.9. ELISA To determine the level of total IgE in mouse serum, mouse serum was collected, and the total IgE level was measured as follows. Ninety-six-well plates were coated with 2 μg/mL purified rat anti-mouse IgE overnight at 4 °C and blocked with $20\%$ ImmunoBlock at 37 °C for 90 min. The samples and purified mouse IgE used as the standard were added to assay wells and incubated at 37 °C for 80 min. After incubation with horseradish peroxidase-conjugated anti-mouse IgE, TMB substrate was added to the assay wells for 20 min. Then, stop solution containing 1 M sulfuric acid was added. Finally, the optical density of each assay well was measured at 450 nm. ## 4.10. Microarray Analysis The microarray datasets GSE27887, GSE32924, GSE36842, GSE58558, GSE59294, GSE95759, GSE99802, GSE107361, GSE120899, GSE130588, GSE133385, GSE133477, and GSE140684 were obtained from the Gene Expression Omnibus database. The data were analyzed with Transcriptome Analysis Console software 4.0 (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA). ## 4.11. TJ Permeability Assay The surface biotinylation technique was used to perform the TJ permeability assay, as described in a previous study [41]. In brief, EZ-linkTM Sulfo-NHS-LC-Biotin in PBS containing 1 M CaCl2 was injected dermally into the lesional or non-lesional skin of mice. After 30 min of incubation, the skin was harvested and immediately embedded in optimal cutting temperature compound. Frozen sections (5 μm) were fixed with $4\%$ paraformaldehyde in PBS for 10 min and permeabilized with $0.01\%$ Triton X-100 in PBS for 10 min. The samples were then blocked with ImmunoBlock for 30 min and incubated overnight at 4 °C with primary antibodies specific for claudin-1. After washing, the sections were incubated with a mixture of an Alexa Fluor 488-conjugated goat anti-mouse antibody and streptavidin Alexa Fluor 594-conjugated antibody for 1 h. The sections were visualized as described above using a Zeiss LSM 780 system. ## 4.12. TER Assay For the TER assay, 3.6 × 105 keratinocytes were cultured on 0.6 cm2 Transwell filters and were then transferred into 1.8 mM high-Ca2+ medium. Then, 10 μM AMP-IBP5 was added to both the apical and basal compartments in the absence or presence of various inhibitors. Forty-eight hours after stimulation, the TER values of the keratinocyte monolayers were determined using a CellZscope system (NanoAnalytics, Münster, Germany). ## 4.13. Statistical Analysis GraphPad Prism 9 software (GraphPad Software Company, version 9.0.0, San Diego, CA) was used for all statistical analyses. Student’s t-test was utilized for comparisons between two groups, and one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons test was utilized for comparisons among multiple groups. $p \leq 0.05$ was considered to indicate a statistically significant difference. ## References 1. Fujii M.. Current understanding of pathophysiological mechanisms of atopic dermatitis: Interactions among skin barrier dysfunction, immune abnormalities and pruritus. Biol. Pharm. Bull. (2020) 43 12-19. DOI: 10.1248/bpb.b19-00088 2. Yang G., Seok J.K., Kang H.C., Cho Y.-Y., Lee H.S., Lee J.Y.. Skin Barrier abnormalities and immune dysfunction in atopic dermatitis. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21082867 3. Bäsler K., Bergmann S., Heisig M., Naegel A., Zorn-Kruppa M., Brandner J.M.. The role of tight junctions in skin barrier function and dermal absorption. J. Control. Release (2016) 242 105-118. DOI: 10.1016/j.jconrel.2016.08.007 4. Brandner J.M.. Importance of Tight junctions in relation to skin barrier function. Curr. Probl. Dermatol. (2016) 49 27-37. DOI: 10.1159/000441541 5. Tsakok T., Woolf R., Smith C.H., Weidinger S., Flohr C.. Atopic dermatitis: The skin barrier and beyond. Br. J. Dermatol. (2019) 180 464-474. DOI: 10.1111/bjd.16934 6. Kirschner N., Rosenthal R., Furuse M., Moll I., Fromm M., Brandner J.M.. Contribution of tight junction proteins to ion, macromolecule, and water barrier in keratinocytes. J. Investig. Dermatol. (2013) 133 1161-1169. DOI: 10.1038/jid.2012.507 7. De Benedetto A., Georas S.N., Cheadle C., Berger A.E.. Tight junction defects in atopic dermatitis. J. Allergy Clin. Immunol. (2011) 127 773-786. DOI: 10.1016/j.jaci.2010.10.018 8. Furuse M., Hata M., Furuse K., Yoshida Y., Haratake A., Sugitani Y., Noda T., Kubo A., Tsukita S.. Claudin-based tight junctions are crucial for the mammalian epidermal barrier: A lesson from claudin-1-deficient mice. J. Cell Biol. (2002) 156 1099-1111. DOI: 10.1083/jcb.200110122 9. Yuki T., Tobiishi M., Kusaka-Kikushima A., Ota Y., Tokura Y.. Impaired tight junctions in atopic dermatitis skin and in a skin-equivalent model treated with interleukin-17. PLoS ONE (2016) 11. DOI: 10.1371/journal.pone.0161759 10. Inoue S., Komiya A., Sugiyama Y., Yuki T., Kuze T., Kusaka A.. Impaired tight junctions obstruct stratum corneum formation by altering polar lipid and profilaggrin processing. J. Dermatol. Sci. (2012) 69 148-158. DOI: 10.1016/j.jdermsci.2012.11.595 11. Niyonsaba F., Kiatsurayanon C., Chieosilapatham P., Ogawa H.. Friends or Foes? Host defense (antimicrobial) peptides and proteins in human skin diseases. Exp. Dermatol. (2017) 26 989-998. DOI: 10.1111/exd.13314 12. Niyonsaba F., Nagaoka I., Ogawa H.. Human defensins and cathelicidins in the skin: Beyond direct antimicrobial properties. Crit. Rev. Immunol. (2006) 26 545-576. DOI: 10.1615/CritRevImmunol.v26.i6.60 13. Niyonsaba F., Nagaoka I., Ogawa H., Okumura K.. Multifunctional antimicrobial proteins and peptides: Natural activators of immune systems. Curr. Pharm. Des. (2009) 15 2393-2413. DOI: 10.2174/138161209788682271 14. Niyonsaba F., Kiatsurayanon C., Ogawa H.. The role of human β-defensins in allergic diseases. Clin. Exp. Allergy J. Br. Soc. Allergy Clin. Immunol. (2016) 46 1522-1530. DOI: 10.1111/cea.12843 15. Osaki T., Sasaki K., Minamino N.. Peptidomics-based discovery of an antimicrobial peptide derived from insulin-like growth factor-binding protein 5. J. Proteome Res. (2011) 10 1870-1880. DOI: 10.1021/pr101114a 16. Beattie J., Allan G.J., Lochrie J.D., Flint D.J.. Insulin-like growth factor-binding protein-5 (IGFBP-5): A critical member of the IGF axis. Biochem. J. (2006) 395 1-19. DOI: 10.1042/BJ20060086 17. Lin S.-C., Wang C.-P., Chen Y.-M., Lu S.-Y., Fann M.-J., Liu C.-J., Kao S.-Y., Chang K.-W.. Regulation of IGFBP-5 expression during tumourigenesis and differentiation of oral keratinocytes. J. Pathol. (2002) 198 317-325. DOI: 10.1002/path.1220 18. Yue H., Song P., Sutthammikorn N., Umehara Y., Trujillo-Paez J.V., Nguyen H.L.T., Takahashi M., Peng G., Ikutama R., Okumura K.. Antimicrobial peptide derived from insulin-like growth factor-binding protein 5 improves diabetic wound healing. Wound Repair Regen. (2022) 30 232-244. DOI: 10.1111/wrr.12997 19. Chieosilapatham P., Yue H., Ikeda S., Ogawa H., Niyonsaba F.. Involvement of the lipoprotein receptor LRP1 in AMP-IBP5-mediated migration and proliferation of human keratinocytes and fibroblasts. J. Dermatol. Sci. (2020) 99 158-167. DOI: 10.1016/j.jdermsci.2020.07.003 20. Gettins P.G.W., Dolmer K.. A proximal pair of positive charges provides the dominant ligand-binding contribution to complement-like domains from the LRP (low-density lipoprotein receptor-related protein). Biochem. J. (2012) 443 65-73. DOI: 10.1042/BJ20111867 21. Lin L., Hu K.. LRP-1: Functions, signaling and implications in kidney and other diseases. Int. J. Mol. Sci. (2014) 15 22887-22901. DOI: 10.3390/ijms151222887 22. Lillis A.P., Van Duyn L.B., Murphy-Ullrich J.E., Strickland D.K.. LDL receptor-related protein 1: Unique tissue-specific functions revealed by selective gene knockout studies. Physiol. Rev. (2008) 88 887-918. DOI: 10.1152/physrev.00033.2007 23. Galliano M.-F., Toulza E., Jonca N., Gonias S.L., Serre G., Guerrin M.. Binding of alpha2ML1 to the low density lipoprotein receptor-related protein 1 (LRP1) reveals a new role for LRP1 in the human epidermis. PLoS ONE (2008) 3. DOI: 10.1371/journal.pone.0002729 24. Akiyama T., Niyonsaba F., Kiatsurayanon C., Nguyen T.T., Ushio H., Fujimura T., Ueno T., Okumura K., Ogawa H., Ikeda S.. The human cathelicidin LL-37 host defense peptide upregulates tight junction-related proteins and increases human epidermal keratinocyte barrier function. J. Innate Immun. (2014) 6 739-753. DOI: 10.1159/000362789 25. Goto H., Hongo M., Ohshima H., Kurasawa M., Hirakawa S., Kitajima Y.. Human beta defensin-1 regulates the development of tight junctions in cultured human epidermal keratinocytes. J. Dermatol. Sci. (2013) 71 145-148. DOI: 10.1016/j.jdermsci.2013.04.017 26. Kiatsurayanon C., Niyonsaba F., Smithrithee R., Akiyama T., Ushio H., Hara M., Okumura K., Ikeda S., Ogawa H.. Host defense (antimicrobial) peptide, human β-defensin-3, improves the function of the epithelial tight-junction barrier in human keratinocytes. J. Investig. Dermatol. (2014) 134 2163-2173. DOI: 10.1038/jid.2014.143 27. Hattori F., Kiatsurayanon C., Okumura K., Ogawa H., Ikeda S., Okamoto K., Niyonsaba F.. The antimicrobial protein S100A7/psoriasin enhances the expression of keratinocyte differentiation markers and strengthens the skin’s tight junction barrier. Br. J. Dermatol. (2014) 171 742-753. DOI: 10.1111/bjd.13125 28. Elias P.M.. Primary role of barrier dysfunction in the pathogenesis of atopic dermatitis. Exp. Dermatol. (2018) 27 847-851. DOI: 10.1111/exd.13693 29. Elias P.M., Wakefield J.S., Man M.-Q.. Moisturizers versus current and next-generation barrier repair therapy for the management of atopic dermatitis. Skin Pharmacol. Physiol. (2019) 32 1-7. DOI: 10.1159/000493641 30. Kraft M.T., Prince B.T.. Atopic dermatitis is a barrier issue, not an allergy issue. Immunol. Allergy Clin. N. Am. (2019) 39 507-519. DOI: 10.1016/j.iac.2019.07.005 31. Matter K., Balda M.S.. Functional analysis of tight junctions. Methods (2003) 30 228-234. DOI: 10.1016/S1046-2023(03)00029-X 32. Yuki T., Haratake A., Koishikawa H., Morita K., Miyachi Y., Inoue S.. Tight junction proteins in keratinocytes: Localization and contribution to barrier function. Exp. Dermatol. (2007) 16 324-330. DOI: 10.1111/j.1600-0625.2006.00539.x 33. Helfrich I., Schmitz A., Zigrino P., Michels C., Haase I., le Bivic A., Leitges M., Niessen C.M.. Role of aPKC isoforms and their binding partners Par3 and Par6 in epidermal barrier formation. J. Investig. Dermatol. (2007) 127 782-791. DOI: 10.1038/sj.jid.5700621 34. Denning M.F.. Tightening the epidermal barrier with atypical PKCs. J. Investig. Dermatol. (2007) 127 742-744. DOI: 10.1038/sj.jid.5700694 35. González-Mariscal L., Tapia R., Chamorro D.. Crosstalk of tight junction components with signaling pathways. Biochim. Biophys. Acta (2008) 1778 729-756. DOI: 10.1016/j.bbamem.2007.08.018 36. Chen X., Macara I.G.. Par-3 controls tight junction assembly through the Rac exchange factor Tiam1. Nat. Cell Biol. (2005) 7 262-269. DOI: 10.1038/ncb1226 37. Turner J.R., Angle J.M., Black E.D., Joyal J.L., Sacks D.B., Madara J.L.. PKC-dependent regulation of transepithelial resistance: Roles of MLC and MLC kinase. Am. J. Physiol. (1999) 277 C554-C562. DOI: 10.1152/ajpcell.1999.277.3.C554 38. Baumer Y., Spindler V., Werthmann R.C., Bünemann M., Waschke J.. Role of Rac 1 and cAMP in endothelial barrier stabilization and thrombin-induced barrier breakdown. J. Cell Physiol. (2009) 220 716-726. DOI: 10.1002/jcp.21819 39. Emonard H., Théret L., Bennasroune A.H., Dedieu S.. Regulation of LRP-1 expression: Make the point. Pathol. Biol. (2014) 62 84-90. DOI: 10.1016/j.patbio.2014.02.002 40. Nguyen H.L.T., Trujillo-Paez J.V., Umehara Y., Yue H., Peng G., Kiatsurayanon C., Chieosilapatham P., Song P., Okumura K., Ogawa H.. Role of antimicrobial peptides in skin barrier repair in individuals with atopic dermatitis. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21207607 41. Chen Y., Merzdorf C., Paul D.L., Goodenough D.A.. COOH terminus of occludin is required for tight junction barrier function in early Xenopus embryos. J. Cell Biol. (1997) 138 891-899. DOI: 10.1083/jcb.138.4.891 42. Bergmann S., von Buenau B., Vidal-y-Sy S., Haftek M., Wladykowski E., Houdek P., Lezius S., Duplan H., Bäsler K., Dähnhardt-Pfeiffer S.. Claudin-1 decrease impacts epidermal barrier function in atopic dermatitis lesions dose-dependently. Sci. Rep. (2020) 10 1-12. DOI: 10.1038/s41598-020-58718-9 43. Egawa G., Kabashima K.. Multifactorial skin barrier deficiency and atopic dermatitis: Essential topics to prevent the atopic march. J. Allergy Clin. Immunol. (2016) 138 350-358.e351. DOI: 10.1016/j.jaci.2016.06.002 44. Suwanchote S., Waitayangkoon P., Chancheewa B., Inthanachai T., Niwetbowornchai N., Edwards S.W., Virakul S., Thammahong A., Kiatsurayanon C., Rerknimitr P.. Role of antimicrobial peptides in atopic dermatitis. Int. J. Dermatol. (2021) 61 532-540. DOI: 10.1111/ijd.15814 45. Batista D.I.S., Perez L., Orfali R.L., Zaniboni M.C., Samorano L.P., Pereira N.V., Sotto M.N., Ishizaki A.S., Oliveira L.M.S., Sato M.N.. Profile of skin barrier proteins (filaggrin, claudins 1 and 4) and Th1/Th2/Th17 cytokines in adults with atopic dermatitis. J. Eur. Acad. Dermatol. Venereol. (2015) 29 1091-1095. DOI: 10.1111/jdv.12753 46. Tokumasu R., Yamaga K., Yamazaki Y., Murota H., Suzuki K., Tamura A., Bando K., Furuta Y., Katayama I., Tsukita S.. Dose-dependent role of claudin-1 in vivo in orchestrating features of atopic dermatitis. Proc. Natl. Acad. Sci. USA (2016) 113 E4061-E4068. DOI: 10.1073/pnas.1525474113 47. Bos J.D., Meinardi M.M.. The 500 Dalton rule for the skin penetration of chemical compounds and drugs. Exp. Dermatol. (2000) 9 165-169. DOI: 10.1034/j.1600-0625.2000.009003165.x 48. Wang M.Y., Yang Y.Y., Heng P.W.. Role of solvent in interactions between fatty acids-based formulations and lipids in porcine stratum corneum. J. Control. Release (2004) 94 207-216. DOI: 10.1016/j.jconrel.2003.10.016 49. Thong H.Y., Zhai H., Maibach H.I.. Percutaneous penetration enhancers: An overview. Skin Pharmacol. Physiol. (2007) 20 272-282. DOI: 10.1159/000107575 50. Potere N., Del Buono M.G., Mauro A.G., Abbate A., Toldo S.. Low density lipoprotein receptor-related protein-1 in cardiac inflammation and infarct healing. Front. Cardiovasc. Med. (2019) 6 51. DOI: 10.3389/fcvm.2019.00051 51. Cheng C.-F., Fan J., Fedesco M., Guan S., Li Y., Bandyopadhyay B., Bright A.M., Yerushalmi D., Liang M., Chen M.. Transforming growth factor alpha (TGFalpha)-stimulated secretion of HSP90alpha: Using the receptor LRP-1/CD91 to promote human skin cell migration against a TGFbeta-rich environment during wound healing. Mol. Cell Biol. (2008) 28 3344-3358. DOI: 10.1128/MCB.01287-07 52. Luo L., Wall A.A., Tong S.J., Hung Y., Xiao Z., Tarique A.A., Sly P.D., Fantino E., Marzolo M.-P., Stow J.L.. TLR crosstalk activates LRP1 to recruit Rab8a and PI3Kγ for suppression of inflammatory responses. Cell Rep. (2018) 24 3033-3044. DOI: 10.1016/j.celrep.2018.08.028 53. Ma Z., Thomas K.S., Webb D.J., Moravec R., Salicioni A.M., Mars W.M., Gonias S.L.. Regulation of Rac1 activation by the low density lipoprotein receptor-related protein. J. Cell Biol. (2002) 159 1061-1070. DOI: 10.1083/jcb.200207070 54. Kawata K., Kubota S., Eguchi T., Moritani N.H., Shimo T., Kondo S., Nishida T., Minagi S., Takigawa M.. Role of the low-density lipoprotein receptor-related protein-1 in regulation of chondrocyte differentiation. J. Cell Physiol. (2010) 222 138-148. DOI: 10.1002/jcp.21930 55. Kanda N., Watanabe S.. Increased serum human β-defensin-2 levels in atopic dermatitis: Relationship to IL-22 and oncostatin M. Immunobiology (2012) 217 436-445. DOI: 10.1016/j.imbio.2011.10.010 56. Niyonsaba F., Ushio H., Hara M., Yokoi H., Tominaga M., Takamori K., Kajiwara N., Saito H., Nagaoka I., Ogawa H.. Antimicrobial peptides human β-defensins and cathelicidin LL-37 induce the secretion of a pruritogenic cytokine IL-31 by human mast cells. J. Immunol. (2010) 184 3526-3534. DOI: 10.4049/jimmunol.0900712 57. Kanda N., Hau C.S., Tada Y., Sato S., Watanabe S.. Decreased serum LL-37 and vitamin D3 levels in atopic dermatitis: Relationship between IL-31 and oncostatin M. Allergy Eur. J. Allergy Clin. Immunol. (2012) 67 804-812. DOI: 10.1111/j.1398-9995.2012.02824.x 58. Chen X., Niyonsaba F., Ushio H., Hara M., Yokoi H., Matsumoto K., Saito H., Nagaoka I., Ikeda S., Okumura K.. Antimicrobial peptides human beta-defensin (hBD)-3 and hBD-4 activate mast cells and increase skin vascular permeability. Eur. J. Immunol. (2007) 37 434-444. DOI: 10.1002/eji.200636379 59. Barlow P.G., Li Y., Wilkinson T.S., Bowdish D.M.E., Lau Y.E., Cosseau C., Haslett C., Simpson A.J., Hancock R.E.W., Davidson D.J.. The human cationic host defense peptide LL-37 mediates contrasting effects on apoptotic pathways in different primary cells of the innate immune system. J. Leukoc. Biol. (2006) 80 509-520. DOI: 10.1189/jlb.1005560 60. Yu J., Mookherjee N., Wee K., Bowdish D.M.E., Pistolic J., Li Y., Rehaume L., Hancock R.E.W.. Host defense peptide LL-37, in synergy with inflammatory mediator IL-1beta, augments immune responses by multiple pathways. J. Immunol. (2007) 179 7684-7691. DOI: 10.4049/jimmunol.179.11.7684 61. Chen X., Takai T., Xie Y., Niyonsaba F., Okumura K., Ogawa H.. Human antimicrobial peptide LL-37 modulates proinflammatory responses induced by cytokine milieus and double-stranded RNA in human keratinocytes. Biochem. Biophys. Res. Commun. (2013) 433 532-537. DOI: 10.1016/j.bbrc.2013.03.024 62. Weidinger S., Novak N.. Atopic dermatitis. Lancet (2016) 387 1109-1122. DOI: 10.1016/S0140-6736(15)00149-X 63. Furue M., Chiba T., Tsuji G., Ulzii D., Kido-Nakahara M., Nakahara T., Kadono T.. Atopic dermatitis: Immune deviation, barrier dysfunction, IgE autoreactivity and new therapies. Allergol. Int. (2017) 66 398-403. DOI: 10.1016/j.alit.2016.12.002 64. Furue M., Kadono T.. New therapies for controlling atopic itch. J. Dermatol. (2015) 42 847-850. DOI: 10.1111/1346-8138.13060 65. Chamlin S.L., Mattson C.L., Frieden I.J., Williams M.L., Mancini A.J., Cella D., Chren M.-M.. The price of pruritus: Sleep disturbance and cosleeping in atopic dermatitis. Arch. Pediatr. Adolesc. Med. (2005) 159 745-750. DOI: 10.1001/archpedi.159.8.745 66. Furue M., Yamamura K., Kido-Nakahara M., Nakahara T., Fukui Y.. Emerging role of interleukin-31 and interleukin-31 receptor in pruritus in atopic dermatitis. Allergy (2018) 73 29-36. DOI: 10.1111/all.13239 67. Niyonsaba F., Ushio H., Nakano N., Ng W., Sayama K., Hashimoto K., Nagaoka I., Okumura K., Ogawa H.. Antimicrobial peptides human beta-defensins stimulate epidermal keratinocyte migration, proliferation and production of proinflammatory cytokines and chemokines. J. Investig. Dermatol. (2007) 127 594-604. DOI: 10.1038/sj.jid.5700599 68. Smithrithee R., Niyonsaba F., Kiatsurayanon C., Ushio H., Ikeda S., Okumura K., Ogawa H.. Human β-defensin-3 increases the expression of interleukin-37 through CCR6 in human keratinocytes. J. Dermatol. Sci. (2015) 77 46-53. DOI: 10.1016/j.jdermsci.2014.12.001 69. Nguyen T.T., Niyonsaba F., Ushio H., Akiyama T., Kiatsurayanon C., Smithrithee R., Ikeda S., Okumura K., Ogawa H.. Interleukin-36 cytokines enhance the production of host defense peptides psoriasin and LL-37 by human keratinocytes through activation of MAPKs and NF-κB. J. Dermatol. Sci. (2012) 68 63-66. DOI: 10.1016/j.jdermsci.2012.07.010 70. Niyonsaba F., Someya A., Hirata M., Ogawa H., Nagaoka I.. Evaluation of the effects of peptide antibiotics human beta-defensins-1/-2 and LL-37 on histamine release and prostaglandin D(2) production from mast cells. Eur. J. Immunol. (2001) 31 1066-1075. DOI: 10.1002/1521-4141(200104)31:4<1066::AID-IMMU1066>3.0.CO;2-# 71. Umehara Y., Kamata Y., Tominaga M., Niyonsaba F., Ogawa H., Takamori K.. Cathelicidin LL-37 induces semaphorin 3A expression in human epidermal keratinocytes: Implications for possible application to pruritus. J. Investig. Dermatol. (2015) 135 2887-2890. DOI: 10.1038/jid.2015.243 72. Niyonsaba F., Ushio H., Nagaoka I., Okumura K., Ogawa H.. The human beta-defensins (-1, -2, -3, -4) and cathelicidin LL-37 induce IL-18 secretion through p38 and ERK MAPK activation in primary human keratinocytes. J. Immunol. (2005) 175 1776-1784. DOI: 10.4049/jimmunol.175.3.1776 73. Yoshida K., Yokouchi M., Nagao K., Ishii K., Amagai M., Kubo A.. Functional tight junction barrier localizes in the second layer of the stratum granulosum of human epidermis. J. Dermatol. Sci. (2013) 71 89-99. DOI: 10.1016/j.jdermsci.2013.04.021 74. Howell M.D., Kim B.E., Gao P., Grant A.V., Boguniewicz M., DeBenedetto A., Schneider L., Beck L.A., Barnes K.C., Leung D.Y.M.. Cytokine modulation of atopic dermatitis filaggrin skin expression. J. Allergy Clin. Immunol. (2009) 124 R7-R12. DOI: 10.1016/j.jaci.2009.07.012 75. Omori-Miyake M., Yamashita M., Tsunemi Y., Kawashima M., Yagi J.. In vitro assessment of IL-4- or IL-13-mediated changes in the structural components of keratinocytes in mice and humans. J. Investig. Dermatol. (2014) 134 1342-1350. DOI: 10.1038/jid.2013.503 76. Peng G., Mu Z., Cui L., Liu P., Wang Y., Wu W., Han X.. Anti-IL-33 antibody has a therapeutic effect in an atopic dermatitis murine model induced by 2, 4-dinitrochlorobenzene. Inflammation (2018) 41 154-163. DOI: 10.1007/s10753-017-0673-7
--- title: 'Children’s Influence on Their Parents’ Satisfaction with Physical Activity in Nature: An Exploratory Study' authors: - Jorge Rojo-Ramos - Antonio Castillo-Paredes - María Mendoza-Muñoz - José Carmelo Adsuar - Irene Polo-Campos - Santiago Gomez-Paniagua - Carmen Galán-Arroyo journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049514 doi: 10.3390/ijerph20065093 license: CC BY 4.0 --- # Children’s Influence on Their Parents’ Satisfaction with Physical Activity in Nature: An Exploratory Study ## Abstract Physical activity in nature has several benefits as it is important for good health, offering physical, social, psychological or even ecological benefits. Nevertheless, in order to maintain adherence to this practice, high levels of satisfaction with the practice are necessary. The objective of this study is to explore whether children’s characteristics influence parental satisfaction with physical activity in nature, analyzing possible differences according to the gender and age of their children. Two hundred and eighty parents responded to two sociodemographic questions in addition to the Physical Activity Enjoyment Scale (PACES), which consists of 16 items. The normality of the data was determined using the Kolmogorov-Smirnov test. Subsequently, nonparametric tests were used to analyze the variables of gender and age in the items, dimensions and total scores of the questionnaire. Statistical differences were found in some of the positive items, which varied according to the age of the children. However, no significant differences were found in the items with respect to the children’s gender or when examining the dimensions or total score of the questionnaire based on both variables. Likewise, age did not show significant correlations with the dimensions and the total score of the questionnaire. Consequently, this study indicates that a child’s age may influence parents’ positive perceptions of the enjoyment of physical activity in nature. Similarly, the gender of the child does not seem to influence these perceptions. ## 1. Introduction Sedentary habits and low levels of physical activity (PA) have been shown to have negative effects on human health [1]. According to the World Health Organization (WHO), “health” is a condition of whole physical, mental, and social well-being and not merely the absence of sickness or disability [2]. The WHO lists pediatric physical inactivity as one of its top priorities. Over the past 20 years, physical inactivity has become widespread [3]. In this context, different studies showed that increased sedentary lifestyles, linked to chronic diseases such as childhood obesity, may, for the first time, put today’s children at risk of living shorter lives than their parents [4]. Regular PA in outdoor settings has been suggested as a useful method for contrasting noncommunicable diseases and chronic pathologies all over the world [5]. Outdoor play has been more popular recently as a way to boost PA and fend off risk factors like obesity, hypertension, and dyslipidemia [6]. Exercise in a natural setting, as opposed to indoor or outdoor built surroundings, improves concentration, decreases negative emotions, and increases energy and a sense of renewal; participants have a greater desire to return to the activity in the future and express greater happiness and satisfaction with outdoor exercise [7]. Since this type of activity is linked with enjoyment, curiosity, effort, the desire to participate and the intention to continue practicing an activity, it has positive effects on young people [8]. This activity also leads to a more self-determined or intrinsic motivation [9]. In both the short (immediately after exposure) [10] and long-term (during the following four weeks) [11], green exercise has been linked to improved mental health. Another finding that corroborated the aforementioned data was the beneficial relationship between overall health and PA in nature [12]. In this context, an increasing number of studies demonstrate the importance of public and green areas planned for PA in cities [13] because they can reduce social problems and offer ecological, physical, psychological, social, health, and economic benefits [14,15]. Moreover, research by White et al. [ 16] showed a weak but substantial correlation between life satisfaction and proportions of green space in towns of no more than 1500 inhabitants. Motivation may also be a significant component. People frequently engage in leisure-time PA because they find it enjoyable or believe it will benefit them [17]; therefore, the key to autonomous motivation is making the decision to participate for enjoyment [18]. Following the self-determination theory, actions made out of autonomous motivation are more likely to be connected to the satisfaction of psychological needs (such as autonomy, competence, and relatedness), and when those needs are met, well-being is increased [19]. Consequently, the higher the level of enjoyment a person achieves with a given activity, the greater will be his or her satisfaction and, therefore, his or her adherence to it. By understanding this mechanism, it is anticipated that actions conducted out of autonomous motivation will result in more favorable psychological outcomes than actions taken out of controlled motivation [20]. In this sense, a person’s level of satisfaction can be understood as the subjective cognitive evaluation of their life and all of its facets, taking into account their standards of living, expectations and aspirations, and the goals they have attained; this evaluation is based on their own standards and is made in a positive way [21]. Additionally, research demonstrates that exercise with a choice is more likely to boost pleasant emotions and satisfaction than exercise without an option [22]. Conversely, the transition to parenthood is challenging since it requires learning new habits and parenting skills as well as dealing with worry about becoming a parent [23]. In addition, the responsibilities of any professional or domestic roles must be modified, and there are effects on mental and physical well-being [24]. Parents of small children have the risk of becoming sedentary [25], as the activities performed before parenthood often change afterward; therefore, leisure activities are often replaced by domestic and childcare activities [26]. Moreover, parents often experience feelings of guilt related to spending time away from their children for exercise, given that the time they have with their children is already restricted, further worsening people’s perception of how little time they have [27]. Research reports that $55.2\%$ of women quit working out before becoming pregnant [28], even though the negative effects of inactivity could be a serious issue for pregnant women [29]. For example, pregnancy-related health advantages of exercise include a reduction in gestational diabetes and gestational hypertension [30]. The fact that infants of physically active mothers have been reported to be more interested in exploring their surroundings is one of the observed fetal health benefits of maternal exercise [31]. For people who are already physically active and belong to more disadvantaged groups, living in green residential areas has been shown to reduce depressive symptoms during pregnancy [32]. However, work-life balance issues have a negative impact on people’s health, with working mothers with young children facing significant problems [33]. Reducing the amount of time spent sleeping or exercising can frequently help people overcome their difficulty balancing work and family life [34]. On the other hand, studies have shown that the more physical exercise parents do during their free time, the more their children do as well, thus demonstrating similarities in the PA behavior of adults and children [35]. In addition, parental warmth has been found to be positively associated with child PA [36]. In this context, the literature indicates that there is a new trend in terms of lack of time for PA, affecting both parents equally, in contrast to early research that indicated that mothers were the ones who were primarily affected [37]. Generally, the literature is predominated by studies that focus on how parents are influencing their children rather than how children are influencing their parents [38]. So, once we know in general terms the importance of PA in the natural environment, its multiple benefits in the short and long term, and the need to continue doing PA before and after becoming parents, the objective of this study is to find out if the satisfaction experienced by parents when doing PA in nature is influenced by the characteristics of their children. ## 2.1. Participants Participants were chosen using a non-probabilistic sample technique based on convenience sampling [39]. The study included 280 parents in total. There were $63.9\%$ women and $36.1\%$ men overall. ## 2.2. Instruments and Measures The questionnaire that was administered to parents through the Google Forms tool obtained the sociodemographic data of the participants. In order to assess the degree of enjoyment with physical exercise, the Spanish version of the Physical Activity Enjoyment Scale (PACES) [40] questionnaire was also used. Each of the 16 components that make up this instrument is preceded by the phrase “when I am active in nature (doing physical activity, physical exercise, or playing a sport…)”. A Likert-type scale is used, with values ranging from 1 to 5, where 1 is used to indicate “complete disagreement” and 5 is used to indicate “full agreement.” Nine of the sixteen questions are about favorably accepting PA, while seven are about adversely rejecting PA. Since the scale’s application yields a score based on the sum of all the things, with 16 serving as the minimum value for a low level of enjoyment of PA and 80 acting as the highest value for that activity, the negative elements were reversed. The authors of the original study obtained Cronbach’s Alpha reliability values of 0.89 for the scale adapted to the Spanish language [40]. ## 2.3. Procedures The Google Forms software was used to design the sociodemographic and PACES data surveys. It made it simpler to reduce expenses, distribute questionnaires to participants, and keep their responses in a single database [41]. Between September and December of 2022, the data were collected. The database of public schools in the Autonomous Community of Extremadura (Spain) maintained by the Department of Education and Employment of the Regional Government of Extremadura was utilized to access the sample (available at: http://estadisticaeducativa.educarex.es/?centros/ensenanzas/&curso=17&ensenanza_centro=101200001, accessed on 30 September 2022). The centers providing the second stage of early childhood education were picked, and contact information was chosen (3 to 6 years). The early childhood education teachers were then emailed with study information and requested to participate. The schools interested in participating were given the informed consent form, which required the participants’ (parents’) signatures. The PACES questionnaire, which comprised the sociodemographic questions, was sent to the parents in electronic form by a URL once they had consented to participate and signed the informed consent form. The decision was made to phone the center and send an email again informing them of the study and how to participate in it after the first month’s response rate was deemed insufficient. The sample was gradually expanded in order to collect the necessary data. ## 2.4. Statistical Analysis The information gathered from the surveys was examined using the Statistical Package for Social Sciences (SPSS) version 23.0 for MAC. The Kolmogorov-Smirnov test was used to determine whether the data were normally distributed. Nonparametric tests were utilized since the results of this test showed that the assumption was false. The median (Me) and interquartile range (IQR) are used to present descriptive data. To analyze the possible significant differences in the items of the questionnaire, their dimensions and the total scale, according to the sociodemographic variables of the children, the Mann-Whitney U test was used to explore gender and the Kruskal-Wallis test to evaluate the three different ages. In addition, Spearman’s Rho correlation test was used to check the relationship between each dimension and the age of the children. Finally, the reliability of the instrument was evaluated using Cronbach’s Alpha. Bernstein [42] claims that dependability levels between 0.60 and 0.70 are acceptable, whereas values between 0.70 and 0.90 are satisfactory. The level of statistical significance was set at $p \leq 0.05.$ ## 3. Results The distribution of frequencies according to the gender and age of the parents can be seen in Table 1. Table 2 provides information about the characteristics of the children. It takes into account three variables: gender, the grade of the preschool course in which they are enrolled and age. Table 3 presents descriptive data based on the median (Me) and interquartile range (IQR) for each gender and age for each question on the PACES questionnaire. The Mann-Whitney U test was used to analyze differences between gender, and the Kruskal-Wallis test was used for differences in ages. Overall, no significant differences were found in terms of gender and age of the children in the influence they have on their parents’ satisfaction with PA in nature. However, in four positive items (6, 10, 14 and 15), significant differences were found for the age of their children ($$p \leq 0.003$$, $$p \leq 0.000$$, $$p \leq 0.001$$ and $$p \leq 0.047$$). According to gender and age, Table 4 displays the total score and median Likert score obtained on the PACES. Regarding the dimensions and total scores of the questionnaire, no significant differences were observed in the responses when compared to the two grouping variables. The Spearman’s Rho test was used to examine the relationship between the median score on the questionnaire and the age variable (Table 5). The PACES scale dimensions and score do not seem to be related to age, as no significant differences were observed in Spearman’s Rho test. Regarding the reliability of the questionnaire when applied in the present study, the values obtained can be defined as satisfactory (Total = 0.89; Positive = 0.81; Negative = 0.83). It should also be noted that the responses obtained were minimally distributed, with most scores showing a ceiling effect. ## 4. Discussion The project was born from the need to know whether the sociodemographic characteristics of children can influence parents’ satisfaction with PA in the natural environment. The results showed significant differences in four items when comparing their scores among the three different age groups of the children. However, no statistically significant differences in item responses were observed according to age. In addition, dimension and total scale ratings do not appear to be influenced by age. Moreover, age showed no relationship with the dimensions and total score of the scales when evaluated. Finally, the ceiling effect observed in questionnaire responses has also been highlighted in previous studies, such as that of Kamnardsiri et al. [ 43], in assessing the enjoyment of an interactive game in older adults. Likewise, Ryuh et al. [ 44] observed this effect when promoting PA through exergaming in people with intellectual disabilities. On the other hand, Soylu et al. [ 45] did not find this ceiling effect when administering PACES after performing different aerobic training methods in adults. The decision to become a parent has been recognized as a significant life event with considerable time demands that appear to have a significant impact on PA and the decision to leave sports, particularly for women [46]. In addition, the different perceptions that children have of natural areas can modify the frequency of visits and their parents’ opinion of them [47]. A relatively recent study indicates that having children at home is positively associated with a preference for PA in nature [5]. Similarly, many studies have shown associations between parental stress related to reports of child behavioral and emotional problems, adjustment difficulties, and internalizing or externalizing problems [48], which can be reduced by exposure to nature by both parties [10]. Sleddens et al. [ 49] pointed out that the relationship of influence on PA of parents and children is bidirectional, despite the fact that the vast majority of the literature only explores the influence of parents on their children. Yang and coworkers found that positive affection on the part of sons predicted greater PA practice on the part of their mothers [50]. Song et al. [ 51] also confirmed these findings, pointing out that the association is much greater when the child is between 3 and 6 years of age. Regarding the positive items (6: “it gives me energy, 10: I get something extra, 14: It gives me strong feelings and 15: I feel good”), it can be observed that significant differences are found in the satisfaction that parents have with respect to PA in relation to their children’s age. It is well known that parents/carers want to pass on this tradition of PA to their daughters and sons, either due to the health benefits it may have or just for the satisfaction it brings to those who engage in PA [52]. Children can already create other worlds at the age of three [53] because of the development of their fine motor abilities, which enable them to manipulate their environment in a play area with sand, dirt, water, and loose pieces [54], which may explain the influence of age. As children are encouraged to explore and learn, chance also facilitates spontaneous exploration, which links physical and cognitive growth [55]. It should be mentioned that including children in activities has a favorable impact on parenting and other habits, such as engaging in fitness-related activities [56]. Moreover, a greater role of parents may be expected in preschool-aged children’s daily activity choices due to children’s early developmental stages [57]. However, parents of young children report less enjoyment with open areas than those without young children, according to a study [58]. There are no differences in gender, but in other studies [59], there is evidence that fathers have a greater influence on sons and mothers on daughters. According to Kirsten et al. [ 60], fathers were more inclined to utilize their personal conduct to encourage physical exercise in their children, whereas mothers affected their children’s PA by offering logistical support. This suggests that parental encouragement of their children’s PA affects both sexes differentially, which is consistent with Trost et al. [ 61]. On the other hand, the negative items do not seem to be influenced by the age of the children or by gender. However, if fathers are sedentary, they influence both genders, although to a greater extent in daughters, while mothers influence but not as much as fathers [62]. In general, there is no difference between parents’ satisfaction with PA and the gender of their children, as highlighted by another study, since this association is relative [57]. ## Limitations and Future Lines of Research As with other studies, this one has a number of limitations. Firstly, the sample selected was only of parents from the Community of Extremadura whose children were enrolled in the second cycle of Infant Education (3–5 years); therefore, there are sociocultural variables that may affect the responses obtained, such as the ceiling effect observed in the questionnaire scores. Second, because the sample was chosen at random, care should be taken while presenting the results. Finally, it is critical to draw attention to the dearth of prior research examining these issues. However, this study breaks with the previous trend of analysis by assessing how children affect parents’ satisfaction with PA rather than the other way around, so this research provides valuable preliminary information. Some potential areas of research include expanding the sample to a national level in all educational stages in order to understand the satisfaction that parents have when they engage in PA with their children and checking whether gender and age have an influence. To do this, a deal may be made with more researchers from the many autonomous communities to gather all required data. Future research should consider parents’ geographic backgrounds (rural vs. urban), how much time they spend engaging in nature activities with their children, their motivations for engaging in PA, and the most frequent obstacles they face. ## 5. Conclusions This study shows that parents’ satisfaction with PA in nature may vary depending on the age of their children, although this variable does not correlate with the dimensions and the total score of the questionnaire. On the other hand, the gender of the child does not seem to be a determining factor in parents’ satisfaction with PA. It is also necessary to highlight that parents exercise less due to work and family responsibilities [26]; however, those who continue to practice this activity in their leisure time enjoy it more, and it reminds them of when they were infants or when they started doing sports before they had more work and/or family burdens [63]. It is important to transmit a healthy lifestyle throughout life since if parents have motivations and habits of doing PA in the natural environment, it is more likely that their children will also have them. It is also essential to involve all administrations and the media to achieve this, as more support and information are needed. ## References 1. Lear S.A., Hu W., Rangarajan S., Gasevic D., Leong D., Iqbal R., Casanova A., Swaminathan S., Anjana R.M., Kumar R.. The Effect of Physical Activity on Mortality and Cardiovascular Disease in 130,000 People from 17 High-Income, Middle-Income, and Low-Income Countries: The PURE Study. Lancet (2017) 390 2643-2654. DOI: 10.1016/S0140-6736(17)31634-3 2. Constitution of the World Health Organization. Am. J. Public Health Nations Health (1946) 36 1315-1323. DOI: 10.2105/AJPH.36.11.1315 3. 3. World Health Organization Physical Activity Strategy for the WHO European Region. 2016–2025World Health Organisation Regional Office for EuropeCopenhagen, Denmark2016978-92-890-5147-7. Physical Activity Strategy for the WHO European Region. 2016–2025 (2016) 4. Olshansky S.J., Passaro D.J., Hershow R.C., Layden J., Carnes B.A., Brody J., Hayflick L., Butler R.N., Allison D.B., Ludwig D.S.. A Potential Decline in Life Expectancy in the United States in the 21st Century. Obstet. Gynecol. Surv. (2005) 60 450-452. DOI: 10.1097/01.ogx.0000167407.83915.e7 5. Calogiuri G., Elliott L.. Why Do People Exercise in Natural Environments? Norwegian Adults’ Motives for Nature-, Gym-, and Sports-Based Exercise. Int. J. Environ. Res. Public Health (2017) 14. DOI: 10.3390/ijerph14040377 6. Janssen I., Leblanc A.G.. Systematic Review of the Health Benefits of Physical Activity and Fitness in School-Aged Children and Youth. Int. J. Behav. Nutr. Phys. Act. (2010) 7 40. DOI: 10.1186/1479-5868-7-40 7. Thompson Coon J., Boddy K., Stein K., Whear R., Barton J., Depledge M.H.. Does Participating in Physical Activity in Outdoor Natural Environments Have a Greater Effect on Physical and Mental Wellbeing than Physical Activity Indoors? A Systematic Review. Environ. Sci. Technol. (2011) 45 1761-1772. DOI: 10.1021/es102947t 8. Jõesaar H., Hein V., Hagger M.S.. Youth Athletes’ Perception of Autonomy Support from the Coach, Peer Motivational Climate and Intrinsic Motivation in Sport Setting: One-Year Effects. Psychol. Sport Exerc. (2012) 13 257-262. DOI: 10.1016/j.psychsport.2011.12.001 9. Dysvik A., Kuvaas B., Gagné M.. An Investigation of the Unique, Synergistic and Balanced Relationships between Basic Psychological Needs and Intrinsic Motivation: Basic Psychological Needs and Intrinsic Motivation. J. Appl. Soc. Psychol. (2013) 43 1050-1064. DOI: 10.1111/jasp.12068 10. Bowler D.E., Buyung-Ali L.M., Knight T.M., Pullin A.S.. A Systematic Review of Evidence for the Added Benefits to Health of Exposure to Natural Environments. BMC Public Health (2010) 10. DOI: 10.1186/1471-2458-10-456 11. de Vries S., van Dillen S.M.E., Groenewegen P.P., Spreeuwenberg P.. Streetscape Greenery and Health: Stress, Social Cohesion and Physical Activity as Mediators. Soc. Sci. Med. (2013) 94 26-33. DOI: 10.1016/j.socscimed.2013.06.030 12. van Dillen S.M.E., de Vries S., Groenewegen P.P., Spreeuwenberg P.. Greenspace in Urban Neighbourhoods and Residents’ Health: Adding Quality to Quantity. J. Epidemiol. Community Health (2012) 66 e8. DOI: 10.1136/jech.2009.104695 13. Harvey C., Rodriguez D.A.. What Makes an Active Public Realm? Opportunities and Challenges for Research. Prev. Med. (2017) 103 S5-S6. DOI: 10.1016/j.ypmed.2017.06.017 14. Aspinall P., Mavros P., Coyne R., Roe J.. The Urban Brain: Analysing Outdoor Physical Activity with Mobile EEG. Br. J. Sports Med. (2015) 49 272-276. DOI: 10.1136/bjsports-2012-091877 15. Shanahan D.F., Franco L., Lin B.B., Gaston K.J., Fuller R.A.. The Benefits of Natural Environments for Physical Activity. Sports Med. (2016) 46 989-995. DOI: 10.1007/s40279-016-0502-4 16. White M.P., Alcock I., Wheeler B.W., Depledge M.H.. Coastal Proximity, Health and Well-Being: Results from a Longitudinal Panel Survey. Health Place (2013) 23 97-103. DOI: 10.1016/j.healthplace.2013.05.006 17. McCarthy P.J., Jones M.V., Clark-Carter D.. Understanding Enjoyment in Youth Sport: A Developmental Perspective. Psychol. Sport Exerc. (2008) 9 142-156. DOI: 10.1016/j.psychsport.2007.01.005 18. Ryan R.M., Deci E.L.. Intrinsic and Extrinsic Motivations: Classic Definitions and New Directions. Contemp. Educ. Psychol. (2000) 25 54-67. DOI: 10.1006/ceps.1999.1020 19. Weinstein N., Ryan R.M.. When Helping Helps: Autonomous Motivation for Prosocial Behavior and Its Influence on Well-Being for the Helper and Recipient. J. Personal. Soc. Psychol. (2010) 98 222-244. DOI: 10.1037/a0016984 20. Deci E.L., Ryan R.M.. Facilitating Optimal Motivation and Psychological Well-Being across Life’s Domains. Can. Psychol./Psychol. Can. (2008) 49 14-23. DOI: 10.1037/0708-5591.49.1.14 21. Carrión A.C., Mañes R.J.M., Sala F.G.. Estudio de La Satisfacción Personal Según La Edad de Las Personas. An. De Psicol./Ann. Psychol. (2000) 16 189-198 22. Lubans D.R., Smith J.J., Morgan P.J., Beauchamp M.R., Miller A., Lonsdale C., Parker P., Dally K.. Mediators of Psychological Well-Being in Adolescent Boys. J. Adolesc. Health (2016) 58 230-236. DOI: 10.1016/j.jadohealth.2015.10.010 23. Ball K., Timperio A., Salmon J., Giles-Corti B., Roberts R., Crawford D.. Personal, Social and Environmental Determinants of Educational Inequalities in Walking: A Multilevel Study. J. Epidemiol. Community Health (2007) 61 108-114. DOI: 10.1136/jech.2006.048520 24. Deave T., Johnson D.. The Transition to Parenthood: What Does It Mean for Fathers?. J. Adv. Nurs. (2008) 63 626-633. DOI: 10.1111/j.1365-2648.2008.04748.x 25. Bellows-Riecken K.H., Rhodes R.E.. A Birth of Inactivity? A Review of Physical Activity and Parenthood. Prev. Med. (2008) 46 99-110. DOI: 10.1016/j.ypmed.2007.08.003 26. Hamilton K., White K.M.. Understanding Parental Physical Activity: Meanings, Habits, and Social Role Influence. Psychol. Sport Exerc. (2010) 11 275-285. DOI: 10.1016/j.psychsport.2010.02.006 27. Dixon M.A.. From Their Perspective: A Qualitative Examination of Physical Activity and Sport Programming for Working Mothers. Sport Manag. Rev. (2009) 12 34-48. DOI: 10.1016/j.smr.2008.09.002 28. Lira Nascimento S., Garanhani Surita F., Godoy A.C., Tamy Kasawara K., Siani Morais S.. Physical Activity Patterns and Factors Related to Exercise during Pregnancy: A Cross Sectional Study. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0128953 29. Darroch F.E., Giles A.R.. A Postcolonial Feminist Discourse Analysis of Urban Aboriginal Women’s Description of Pregnancy-Related Weight Gain and Physical Activity. Women Birth (2016) 29 e23-e32. DOI: 10.1016/j.wombi.2015.08.003 30. Artal R.. Exercise: The Alternative Therapeutic Intervention for Gestational Diabetes. Clin. Obstet. Gynecol. (2003) 46 479-487. DOI: 10.1097/00003081-200306000-00026 31. Clapp J.F., Lopez B., Harcar-Sevcik R.. Neonatal Behavioral Profile of the Offspring of Women Who Continued to Exercise Regularly throughout Pregnancy. Am. J. Obstet. Gynecol. (1999) 180 91-94. DOI: 10.1016/S0002-9378(99)70155-9 32. McEachan R.R.C., Prady S.L., Smith G., Fairley L., Cabieses B., Gidlow C., Wright J., Dadvand P., van Gent D., Nieuwenhuijsen M.J.. The Association between Green Space and Depressive Symptoms in Pregnant Women: Moderating Roles of Socioeconomic Status and Physical Activity. J. Epidemiol. Community Health (2016) 70 253-259. DOI: 10.1136/jech-2015-205954 33. Fraguela Vale R., Lorenzo Castiñeiras J., Varela Garrote L.. Conciliación y actividad física de ocio en familias con hijos en Educación Primaria. Implicaciones para la infancia. Rev. Investig. Educ. (2011) 9 162-173 34. Martínez del Castillo J., Martin Rodriguez M., del Hierro Pinés D., Jiménez-Beatty Navarro J.E., González Rivera M.D.. Barreras a la Actividad Física en las Mujeres Adultas y Alternativas de Conciliación. Actividad Física Deporte Ciencia Profesión (2008) 9 12-24 35. Schoeppe S., Röbl M., Liersch S., Krauth C., Walter U.. Mothers and Fathers Both Matter: The Positive Influence of Parental Physical Activity Modeling on Children’s Leisure-Time Physical Activity. Pediatr. Exerc. Sci. (2016) 28 466-472. DOI: 10.1123/pes.2015-0236 36. Xu H., Wen L.M., Rissel C.. Associations of Parental Influences with Physical Activity and Screen Time among Young Children: A Systematic Review. J. Obes. (2015) 2015 546925. DOI: 10.1155/2015/546925 37. Bianchi S.M.. Family Change and Time Allocation in American Families. ANNALS Am. Acad. Political Soc. Sci. (2011) 638 21-44. DOI: 10.1177/0002716211413731 38. Patrick H., Hennessy E., McSpadden K., Oh A.. Parenting Styles and Practices in Children’s Obesogenic Behaviors: Scientific Gaps and Future Research Directions. Child. Obes. (2013) 9 S-73-S-86. DOI: 10.1089/chi.2013.0039 39. Salkind N.J., Escalona R.L., Valdés Salmerón V.. Métodos de Investigación (1999) 40. Moreno J.-A., González-Cutre D., Martínez C., Alonso N., López M.. Psychometric Properties of the Physical Activity Enjoyment Scale (PACES) in the Spanish Context. Stud. Psychol. (2008) 29 173-180. DOI: 10.1174/021093908784485093 41. Anderson T., Kanuka H.. E-Research: Methods, Strategies, and Issues (2003) 42. Nunnally J.C., Bernstein I.H.. Psychometric Theory (1994) 43. Kamnardsiri T., Phirom K., Boripuntakul S., Sungkarat S.. An Interactive Physical-Cognitive Game-Based Training System Using Kinect for Older Adults: Development and Usability Study. JMIR Serious Games (2021) 9 e27848. DOI: 10.2196/27848 44. Ryuh Y.J., Chen C.-C., Pan Z., Gadke D.L., Elmore-Staton L., Pan C.-Y., Cosgriff A.. Promoting Physical Activity through Exergaming in Young Adults with Intellectual Disabilities: A Pilot Study. Int. J. Dev. Disabil. (2022) 68 227-233. DOI: 10.1080/20473869.2019.1605771 45. Soylu Y., Arslan E., Sogut M., Kilit B., Clemente F.. Effects of Self-Paced High-Intensity Interval Training and Moderate-Intensity Continuous Training on the Physical Performance and Psychophysiological Responses in Recreationally Active Young Adults. Biol. Sport (2020) 38 555-562. DOI: 10.5114/biolsport.2021.100359 46. Mailey E.L., Huberty J., Dinkel D., McAuley E.. Physical Activity Barriers and Facilitators among Working Mothers and Fathers. BMC Public Health (2014) 14. DOI: 10.1186/1471-2458-14-657 47. Veitch J., Bagley S., Ball K., Salmon J.. Where Do Children Usually Play? A Qualitative Study of Parents’ Perceptions of Influences on Children’s Active Free-Play. Health Place (2006) 12 383-393. DOI: 10.1016/j.healthplace.2005.02.009 48. Ashford J., Smit F., van Lier P.A.C., Cuijpers P., Koot H.M.. Early Risk Indicators of Internalizing Problems in Late Childhood: A 9-Year Longitudinal Study. J. Child Psychol. Psychiatry (2008) 49 774-780. DOI: 10.1111/j.1469-7610.2008.01889.x 49. Sleddens E.F.C., Gubbels J.S., Kremers S.P.J., van der Plas E., Thijs C.. Bidirectional Associations between Activity-Related Parenting Practices, and Child Physical Activity, Sedentary Screen-Based Behavior and Body Mass Index: A Longitudinal Analysis. Int. J. Behav. Nutr. Phys. Act. (2017) 14 89. DOI: 10.1186/s12966-017-0544-5 50. Yang C.-H., Huh J., Mason T.B., Belcher B.R., Kanning M., Dunton G.F.. Mother-Child Dyadic Influences of Affect on Everyday Movement Behaviors: Evidence from an Ecological Momentary Assessment Study. Int. J. Behav. Nutr. Phys. Act. (2020) 17 56. DOI: 10.1186/s12966-020-00951-6 51. Song M., Dieckmann N.F., Stoyles S., Kim Y., Lumeng J.C.. Associations between Mother’s and Children’s Moderate-to-Vigorous Physical Activity and Sedentary Time in the Family Context. Prev. Med. Rep. (2017) 8 197-203. DOI: 10.1016/j.pmedr.2017.10.012 52. Costa F., Marques A.. Promoting Active and Healthy Lifestyle at School: Views of Students, Teachers, and Parents in Portugal (2011) 249-268 53. Allen K.E., Marotz L.R.. By the Ages: Behavior & Development of Children Pre-Birth through Eight (2000) 54. Dyment J.E., Bell A.C.. Grounds for Movement: Green School Grounds as Sites for Promoting Physical Activity. Health Educ. Res. (2007) 23 952-962. DOI: 10.1093/her/cym059 55. Frost J.L., Klein B.L., Thorson H., Pollard M., Schuetze U., Muller M., Weimer-Stuckmann G.. Children’s Play and Playgrounds (1979) 56. Weiss M.R., Hayashi C.T.. All in the Family: Parent-Child Influences in Competitive Youth Gymnastics. Pediatr. Exerc. Sci. (1995) 7 36-48. DOI: 10.1123/pes.7.1.36 57. Yao C.A., Rhodes R.E.. Parental Correlates in Child and Adolescent Physical Activity: A Meta-Analysis. Int. J. Behav. Nutr. Phys. Act. (2015) 12 10. DOI: 10.1186/s12966-015-0163-y 58. Sharghi A., Maulan S.B., Salleh I.B., Salim A.S.B.S., Sedghpour B.S.. Evaluation of Parental Satisfaction of Children’s Spaces Within High Rise Apartment Environments. J. Landsc. Ecol. (2014) 7 59-72. DOI: 10.2478/jlecol-2014-0016 59. Mitchell J., Skouteris H., McCabe M., Ricciardelli L.A., Milgrom J., Baur L.A., Fuller-Tyszkiewicz M., Dwyer G.. Physical Activity in Young Children: A Systematic Review of Parental Influences. Early Child Dev. Care (2012) 182 1411-1437. DOI: 10.1080/03004430.2011.619658 60. Kirsten Krahnstoever D., Cutting T.M., Birch L.L.. Parents’ Activity-Related Parenting Practices Predict Girls’ Physical Activity. Med. Sci. Sport. Exerc. (2003) 35 1589-1595. DOI: 10.1249/01.MSS.0000084524.19408.0C 61. Trost S., Kerr L., Ward D., Pate R.. Physical Activity and Determinants of Physical Activity in Obese and Non-Obese Children. Int. J. Obes. (2001) 25 822-829. DOI: 10.1038/sj.ijo.0801621 62. Ruiz-Juan F., Piéron M., Baena-Extremera A.. Socialización de la actividad físico-deportiva en adultos: Relación con familia, pareja y amigos. Rev. Iberoam. Diagnóstico Evaluación-Avaliação Psicológica (2012) 2 35-59 63. Brown P.J.. Benefits of Outdoor Recreation and Some Ideas for Valuing Recreation Opportunities (2019) 209-220
--- title: Eicosapentaenoic and Docosahexaenoic Acid Supplementation Increases HDL Content in n-3 Fatty Acids and Improves Endothelial Function in Hypertriglyceridemic Patients authors: - Paola Peña-de-la-Sancha - Adolfo Muñoz-García - Nilda Espínola-Zavaleta - Rocío Bautista-Pérez - Ana María Mejía - María Luna-Luna - Victoria López-Olmos - José-Manuel Rodríguez-Pérez - José-Manuel Fragoso - Elizabeth Carreón-Torres - Óscar Pérez-Méndez journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049536 doi: 10.3390/ijms24065390 license: CC BY 4.0 --- # Eicosapentaenoic and Docosahexaenoic Acid Supplementation Increases HDL Content in n-3 Fatty Acids and Improves Endothelial Function in Hypertriglyceridemic Patients ## Abstract High-density lipoproteins (HDLs) are known to enhance vascular function through different mechanisms, including the delivery of functional lipids to endothelial cells. Therefore, we hypothesized that omega-3 (n-3) eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) content of HDLs would improve the beneficial vascular effects of these lipoproteins. To explore this hypothesis, we performed a placebo-controlled crossover clinical trial in 18 hypertriglyceridemic patients without clinical symptoms of coronary heart disease who received highly purified EPA 460 mg and DHA 380 mg, twice a day for 5 weeks or placebo. After 5 weeks of treatment, patients followed a 4-week washout period before crossover. HDLs were isolated using sequential ultracentrifugation for characterization and determination of fatty acid content. Our results showed that n-3 supplementation induced a significant decrease in body mass index, waist circumference as well as triglycerides and HDL-triglyceride plasma concentrations, whilst HDL-cholesterol and HDL-phospholipids significantly increased. On the other hand, HDL, EPA, and DHA content increased by $131\%$ and $62\%$, respectively, whereas 3 omega-6 fatty acids significantly decreased in HDL structures. In addition, the EPA-to-arachidonic acid (AA) ratio increased more than twice within HDLs suggesting an improvement in their anti-inflammatory properties. All HDL-fatty acid modifications did not affect the size distribution or the stability of these lipoproteins and were concomitant with a significant increase in endothelial function assessed using a flow-mediated dilatation test (FMD) after n-3 supplementation. However, endothelial function was not improved in vitro using a model of rat aortic rings co-incubated with HDLs before or after treatment with n-3. These results suggest a beneficial effect of n-3 on endothelial function through a mechanism independent of HDL composition. In conclusion, we demonstrated that EPA and DHA supplementation for 5 weeks improved vascular function in hypertriglyceridemic patients, and induced enrichment of HDLs with EPA and DHA to the detriment of some n-6 fatty acids. The significant increase in the EPA-to-AA ratio in HDLs is indicative of a more anti-inflammatory profile of these lipoproteins. ## 1. Introduction Long-term prospective cohort studies have consistently demonstrated an association between higher intakes of omega-3 (n-3) fatty acids, particularly eicosapentaenoic acid (EPA, 20:5 n-3) and docosahexaenoic acid (DHA, 22:6 n-3), and a lower risk of developing coronary artery disease (CAD) [1,2,3,4,5,6]. EPA and DHA modulate some of the most important known risk factors of CAD, such as blood lipids, blood pressure, heart rate, heart rate variability, platelet aggregation, inflammation, and endothelial function [6]. Of particular interest, endothelial dysfunction is one of the earliest events in the pathological development of atherosclerotic diseases [6,7]. Accordingly, prospective studies have shown that flow-mediated dilation (FMD), the gold standard for measuring endothelial dysfunction in vivo, is an independent predictor of cardiovascular events, such as heart attack or stroke [6,7,8,9,10,11]. Therefore, therapies targeted to improve endothelial dysfunction, such as EPA and DHA supplementation, may report important benefits to subjects at risk of CAD. In this context, high-density lipoproteins (HDLs) have been described as lipoproteins with beneficial effects on endothelial function [12,13,14], probably associated with their EPA and DHA content [15]. HDLs are heterogeneous complexes of proteins and lipids that have been proposed to protect against cardiovascular disease through different mechanisms, including the reverse transport of cholesterol (RTC) [16]. It has also been described that HDLs induce endothelial nitric oxide synthase (eNOS) stability and phosphorylation, increasing its halftime and abundance [17,18]. Besides cholesterol efflux, we have recently demonstrated that HDLs also promote the influx of sphingomyelin and cholesterol, which improves endothelial cells in vitro [18]. Such observations suggest that HDLs are vectors that drive lipids to peripheral cells for structuring their membranes and other cellular functions [16,18]. Following this idea, we postulate that HDLs may carry n-3 fatty acids to endothelial cells in vivo, thus enhancing the functionality of this tissue. We, therefore, isolated HDLs from patients treated with EPA and DHA and determined whether these lipoproteins effectively become enriched in n-3 fatty acids during treatment. We further explored whether EPA and DHA induce an improvement of endothelial function in vivo and the potential contribution of HDLs to this beneficial effect. ## 2.1. Patients Included in the Study We evaluated for eligibility 105 individuals who were initially candidates for the study. Eighty-three subjects were excluded; 13 because of glucose above 125 mg/dL, 55 subjects reduced their triglycerides below 200 mg/dL, and 15 individuals abandoned the study during the dietary intervention. Then, 22 hypertriglyceridemic patients were enrolled. During the following stages of the trial, 4 patients withdrew from the study. Consequently, we reported the results of 18 patients who concluded all the phases of the protocol (Figure 1). ## 2.2. Anthropometric and Biochemical Characteristics during EPA and DHA Supplementation Anthropometric and biochemical data of patients before and after n-3 fatty acids or placebo are presented in Table 1. Slight decreases in body mass index (BMI) were observed in either, n-3 fatty acids, or placebo, but only the former reached statistical significance. Concomitantly, body fat mass diminished during treatment with EPA and DHA. Importantly, energy intake did not change during placebo or n-3 supplementation (Supplementary Table S1). Other anthropometric measurements remained unchanged after interventions. As expected, the median of triglycerides decreased, whereas HDL-cholesterol increased with n-3 supplementation (Table 1). ## 2.3.1. HDL Subclasses and Lipid Content In addition to the increase in HDL-cholesterol plasma concentrations after n-3 supplementation, HDL-phospholipids were significantly augmented by $27\%$, whereas HDL-triglycerides diminished by about $35\%$ (Table 1). These results suggest that n-3 supplementation is associated with structural modifications of HDL particles. Therefore, we determined the relative proportion of HDL subclasses and their lipid content. The relative proportion of HDL protein remained unchanged along the subclasses with either placebo or n-3 supplementation (Table 2). Cholesterol tended to increase in all HDL subclasses, but reached statistical significance only for small particles HDL3b and HDL3c, indicating that n-3 supplementation did not affect size distribution and had a mild effect on lipid content in HDLs (Supplementary Table S2). In contrast, the quality of fatty acids was modified; as expected, HDL content of EPA and DHA increased $131.0\%$ and $61.6\%$, respectively, after 5 weeks of n-3 supplementation. Besides the increase in EPA and DHA content in HDLs, three n-6 fatty acids significantly decreased, particularly cis-5,8,11-eicosatrienoic acid (20:3 n-6) precursor of arachidonic acid (AA, 20:4 n-6), and two fatty acids derived from AA, cis-7,10,13,16-docosatetraenoic acid (22:4 n-6), and cis-4,7,10,13,16-docosapentaenoic acid (22:5 n-6) (Table 3). Consequently, the EPA-to-AA ratio in HDLs increased more than twice the basal level after n-3 supplementation (from 0.0503 ± 0.033 to 0.128 ± 0.091, $$p \leq 0.001$$). Moreover, the DHA-to-AA ratio in HDLs drastically increased, from 0.074 ± 0.017 in basal conditions to 0.400 ± 0.153 after n-3 supplementation. ## 2.3.2. HDL Stability Polyunsaturated fatty acids are more fluid than mono or saturated fatty acids and consequently, the increase in n-3 content in HDLs may compromise the stability of the particle. Therefore, we analyzed HDL stability by estimating the proportion of apolipoproteins that abandoned the lipid environment of lipoprotein at 8 h in the presence of urea (percentage of denaturation), as previously described [19]. The median of the percentage of HDL denaturation in basal conditions was $69.5\%$ [33.2–$90.2\%$] and remained similar after the n-3 supplementation ($75.7\%$ [38.8–$91.4\%$], $p \leq 0.05$), indicating that the stability of HDLs was not affected by the increased proportion of n-3 fatty acids described above. ## 2.4.1. Flow-Mediated Vasodilation In Vivo We determined the endothelial function in vivo by flow-mediated vasodilation (FMD); results are shown in Table 4. The internal diameter of the brachial artery after 1 min of hyperemia increased significantly in hypertriglyceridemic patients during supplementation with n-3. In contrast, FMD remained similar before and after the placebo (Table 4). ## 2.4.2. Endothelial-Mediated Vasodilation In Vitro To explore whether HDLs contribute to FMD enhancement observed in vivo, we determined the endothelial-dependent vasorelaxation using a model of rat aorta rings incubated with HDL isolated from patients’ plasma; these results are shown in Figure 2. As observed, the vasorelaxation induced by increasing doses of acetylcholine is similar in the presence of HDLs obtained after n-3 supplementation or placebo, before or after intervention. ## 3. Discussion In this study, we explored whether the beneficial effects of EPA and DHA supplementation on vascular function might be linked to HDL structure. Both HDLs and-3 fatty acids have been shown to improve endothelial function [6,17,18,20], and such a function may be interrelated. In this context, previous reports have demonstrated the capacity of HDLs to induce Ser1177 phosphorylation of eNOS (indispensable to enzyme activation) related to the lipid content of these lipoproteins [18]. In addition, the capacity of HDLs to regulate ICAM-1 expression in cultured endothelial cells is also dependent on their lipid content [18]. These observations strongly suggest that phospholipids and triglycerides structured with EPA or DHA may be driven by HDLs into endothelial cells, thus facilitating the beneficial effects on the vascular function of these polyunsaturated fatty acids. Our results showed that 2 g/d of EPA and DHA effectively improved endothelial function assessed using FMD. This beneficial effect of n-3 on vascular function is still controversial [20,21,22]. Treatment with omega 3 failed to improve FMD in a group of 38 patients with T2DM, HbA1c < $7\%$, high cardiovascular risk, and optimal medical therapy for hypertension and plasma lipids [21]. In that study, patients were about 20 years older than individuals included in our trial. Moreover, we focused on individuals who did not take any medication and with glucose levels below 125 mg/dL. Consequently, our data support the idea that n3 supplementation is useful for improving FMD in patients with a vascular function relatively preserved, as observed in this and other studies [20,23]. Conversely, when advanced atherosclerotic disease is already established, EPA and DHA do not seem to report a benefit on vascular function [21]. In this context, n-3 supplementation is likely useful to prevent endothelial dysfunction rather than to treat wounded vessels. Long-term prospective studies are needed to address this hypothesis. Once we demonstrated the beneficial effect of EPA and DHA on vascular function, the next step was to explore whether HDLs effectively carried both n-3 fatty acids. Our results showed a significant increase in EPA and DHA as components of lipids carried by HDLs. Such increments were similar to that observed in the plasma of patients with high cardiovascular risk treated with 1800 mg/d of n-3 as fish oil, and who reached $220.3\%$ and $68.3\%$ of EPA and DHA increase in plasma, respectively [15]; this study also demonstrated a modification of HDL subclasses induced by n-3 supplementation. Conversely, we did not observe any change in HDL size distribution or in lipids of HDL subclasses, probably because our study group was more homogeneous in terms of risk factors than the previous report [15]. Besides EPA and DHA, three n-6 fatty acids significantly decreased during n-3 supplementation, the precursor of arachidonic acid, and two of its metabolites as mentioned in the results section. This observation suggests that HDLs, EPA, and DHA compete with other long-chain fatty acids for key enzymes related to inflammatory processes. Of particular interest, is the increase in the EPA-to-AA ratio of about twice the basal value. EPA and AA may compete for the active site of cyclooxygenases (COX’s) since both fatty acids are substrates of these enzymes [24]. However, EPA-derived prostaglandins (PG) are minimally inflammatory or may even have anti-inflammatory properties, such as PGE3 which may selectively promote M2a polarization, while inhibiting M1 [25]. Therefore, the increased proportion of the EPA-to-AA ratio is likely to enhance the anti-inflammatory properties of HDLs during n-3 supplementation. This evidence is consistent with the early use of n-3 fatty acids to lower interleukin-1 and tumor necrosis factor [26] and also agrees with the opposite statistical association of EPA and AA with the risk of CAD in young Chinese patients [27]; AA was associated with an increased risk of CAD, whereas EPA seemed to play a protective role against the disease. Consequently, the ratio of EPA-to-AA can increase the predictive value for diagnosing CAD more than EPA or AA alone [27]. Based on this evidence, the huge increase in EPA proportion with respect to AA in HDLs further supports the idea that n-3 supplementation enhances the atheroprotective properties of these lipoproteins. Moreover, DHA may contribute to the increase the anti-inflammatory properties of HDLs since this fatty acid, as well as EPA, are precursors of resolvins (inflammation-resolving mediators) via the lipoxygenase pathway [24]. Finally, it cannot be discarded that the reduction in some other n-6 fatty acids in HDL may also contribute to the beneficial effects of n-3 supplementation [15]. This issue needs to be addressed in further studies. Polyunsaturated fatty acids are more fluid than monounsaturated or saturated fatty acids; consequently, the increase in n-3 content in HDLs may compromise the stability of the particle via the modification of its surface tension, which is fundamental for the metabolism and function of these lipoproteins [28,29]. This relevant issue had to be considered since the stability of HDLs could limit their potential beneficial effects on vascular function. Then, to establish whether the increased proportion of n-3 in HDLs may affect the stability of the particle, we analyzed their capacity of apolipoprotein to remain associated with HDL-lipids in the presence of urea. Our results clearly showed that the increased proportions of n-3 fatty acids after supplementation did not affect the stability of HDLs. Of notice, EPA and DHA increased, but concomitantly, other polyunsaturated fatty acids, such as 20:3 n-6, 22:4 n-6 and 22:5 n-6, significantly decreased, compensating somehow for the fluidity of the particle. This speculative explanation needs to be explored in future studies. Once we confirmed that HDLs were enriched with n-3 fatty acids, we further explored the possibility that EPA and DHA contribute to endothelial function improvement via these lipoproteins. For this, we pre-incubated rat aorta rings with HDLs from patients and then contracted them with epinephrine. The addition of increasing doses of acetylcholine induces the endothelium-dependent relaxation of vascular smooth muscle [30]. Using this model, we were not able to demonstrate that HDL enriched with EPA and DHA enhanced endothelial function. It should be considered that vascular tissue was incubated with HDL for only 1 h; even if the exchange of HDL-lipids with cells occurs within the first 60 min [18], the intracellular signaling of n-3 [6] is likely to be a process of longer duration. Therefore, we cannot still discard the possibility that the beneficial effects of HDLs on endothelial function are linked to their EPA and DHA content, but in a long-term manner. Other models of endothelial function suitable for longer incubation times should be developed to address this issue. Finally, long-term effects of EPA and DHA that have been observed in some tissues [31] may have persisted through the wash-out period in patients who initiated in the n-3 supplementation arm. Moreover, an isocaloric diet may also provide some beneficial effects during the study. Consequently, some parameters, such as HDL-triglycerides and FMD, tended to be different before the placebo with respect to the pre-n-3 supplementation values, and the positive effects tended to increase by the end of the crossover period. Even if such differences did not reach statistical significance, they are inherent to the design of the trial and should be recognized as a weakness of this study. ## 4.1. Patients and Study Design One hundred and five non-smoking consecutive male volunteers from the blood bank of the Instituto Nacional de Cardiologia “Ignacio Chavez” were screened for this study. Subjects with triglycerides plasma levels > 200 mg/dL, who were not taking any anti-dyslipidemic drug, and did not have dysthyroidism, liver disease, autoimmune or congenital heart disease as determined using biochemical analyses or medical history, were invited to the next step of the study. Volunteers completed a survey regarding lifestyle factors, medication, and diet habits, which included a 24 h questionnaire. Then, subjects were instructed to follow an isocaloric diet personally designed to reach a balanced intake of 50, 30, and $20\%$ of total calories from carbohydrates, lipids, and proteins, respectively. Subjects whose fasting triglyceride plasma levels decreased below 200 mg/dL after one week of an isocaloric diet were excluded. Eighteen subjects were finally included to complete the trial. The Isocaloric diet was continued during the study (Figure 1). We performed a single-blind randomized crossover study that was placebo-controlled. Hypertriglyceridemic patients received EPA 460 mg and DHA 380 mg twice a day for 5 weeks or placebo, followed by a washout period of 4 weeks before crossover. Anthropometric measures, blood pressure, flow-mediated vasodilation, biochemical profile and HDL characterization were performed before and after placebo or EPA + DHA supplementation. The study was performed in accordance with the appropriate version of the Declaration of Helsinki and approved by the Ethics Committee from the Instituto Nacional de Cardiología “Ignacio Chávez” with registration numbers 17-996 and 22-1338. All the patients gave their written informed consent prior to the recruitment. ## 4.2. Laboratory Analysis After 12 h of overnight fasting, blood samples were drawn in EDTA or dry tubes, and centrifuged for 15 min at 1300× g within 15 min after collection. Plasma and serum were separated into 500 µL aliquots and then either immediately analyzed or frozen at −80 °C until analysis. Glucose, cholesterol, and triglycerides plasma concentrations were determined by commercially available enzymatic/colorimetric assays (Randox Labs. Ltd., Crumlin, Co., Antrim, Northern Ireland, UK). The phosphotungstic acid-Mg + 2 method was used to precipitate apo B-containing lipoproteins before quantifying HDL-cholesterol (HDL-C) and HDL-triglycerides (HDL-Tg). HDL-phospholipids (HDL-Pho) concentrations were quantified using a phospholipase C method (Wako Chemicals, Richmond, VA, USA). ## 4.3. HDL Subclasses Composition Assessment HDLs were isolated using ultracentrifugation and analyzed as previously described [32]. Briefly, after recovery, HDLs were further separated by their hydrodynamic diameter; 25 µg of HDL protein were run in a non-denaturing 3–$30\%$ gradient polyacrylamide gel electrophoresis. Gels were enzymatically stained for cholesterol or triglycerides and scanned to obtain the densitograms for these HDL components. Thereafter, gels were re-stained with Coomassie blue R-250 to detect proteins. HDL subclasses and size intervals were established using globular proteins as diameter references (high-molecular-weight calibration kit, Amersham Pharmacia Biotech, Buckinghamshire, UK). Each subclass was measured using VisionWorks 8.20 version and expressed as the percentage of total HDL-protein or lipid area of the densitogram (7.94 to 13.59 nm) [32,33] Total fatty acids in HDL were determined using gas chromatography after the hydrolysis of lipids [34] using an external service provider (OmegaQuant HQ, Sioux Falls, SD). For this, 50 μg of HDL protein was placed and dried in a paper support containing an antioxidant preservative following the service provider’s instructions. Data were reported as the percentage of the total fatty acids. ## 4.4. HDL Stability The stability of HDLs was determined using the Trp fluorescence red-shift during the denaturation of isolated lipoproteins at a final concentration of 30 μg/mL, in PBS-8M urea, pH 7.4 as previously described [19]. Briefly, Trp intrinsic fluorescence emission was measured at 330, 344 and 365 nm (Emλ30, Emλ344, and Emλ365, respectively) using an LS55 Perkin-Elmer fluorescence spectrophotometer (Walthman, MA, USA) at room temperature, with an excitation wavelength (λex) of 280 nm and bandwidth of 5.5 nm. Then, the ratio of fluorescence intensities (rfi) was calculated as:rfi=(Emλ344+Emλ365)Emλ330 For every sample of isolated HDL, rfi was determined at the initial time ($t = 0$ h), at 8 h, and finally at 26 h. The rfi at 0 h was considered as $0\%$ of denaturation, whereas the rfi value at 26 h was set as $100\%$ of the measurable denaturation of HDLs [19]. The proportional percentage of denaturation at 8 h represents the stability of the particle, i.e., the lower the percentage at 8 h, the higher the stability of the particle. ## 4.5. Flow-Mediated Vasodilation Endothelial function determined using flow-mediated vasodilation (FMD) was determined using the ultrasound method as previously described [35,36]. The brachial artery diameter was measured in the longitudinal plane, 5 cm above the antecubital fossa after at least 30 min of resting position. The site of measurement was marked on the skin, and the arm was held in the same position during the entire study. Reactive hyperemia was induced by inflating a blood pressure cuff to 200 mmHg for 5 min. The internal diameter of the brachial artery was measured 3 times and averages were taken at rest and in the first minute after deflating the cuff. FMD was defined as the change in the internal diameter of the brachial artery during reactive hyperemia compared to the basal diameter and expressed as a percentage [35,36]. ## 4.6. Vascular Reactivity of Aorta Rings The potential effect of HDLs on endothelial function was explored using aorta rings from Wistar rats in an organ bath, as previously described [30,33]. Aorta rings were incubated with isolated HDL to a final concentration of 50 mg/dL of cholesterol in Krebs solution and $95\%$ O2/$5\%$ CO2 continuous flow at 37 °C, 1 h previous to the test. After an equilibrium period, aorta rings were pre-contracted with phenylephrine 3 × 10−4 M. The endothelium-mediated relaxation (vasodilation) was evaluated by the increasing addition of acetylcholine from 5 × 10−9 to 8 × 10−7 M of final concentration. Vascular contractions were measured by means of an FT–03 Grass Force Displacement Transducer and recorded on a Grass Polygraph (Model 7D, Grass Medical Instruments, Quincy, MA, USA). All experiments were run in duplicate, and results were expressed as the percentage of relaxation with respect to pre-contraction with phenylephrine [30,33]. ## 5. Conclusions In this study, we demonstrated that EPA 460 mg and DHA 380 mg twice a day for 5 weeks improved vascular function assessed using flow-mediated dilation in hypertriglyceridemic patients without clinical symptoms of coronary heart disease and who were not receiving other drugs. Vascular function improvement was concomitant with an enrichment of HDL with EPA and DHA that did not affect the stability of these lipoproteins. The significant increase in the EPA-to-AA ratio in HDLs may contribute to the enhancement of the anti-inflammatory effects of these lipoproteins. ## References 1. Wiest E.F., Walsh-Wilcox M.T., Walker M.K.. Omega-3 Polyunsaturated Fatty Acids Protect Against Cigarette Smoke-Induced Oxidative Stress and Vascular Dysfunction. Toxicol. Sci. (2017) 156 300-310. DOI: 10.1093/toxsci/kfw255 2. Goodfellow J., Bellamy M.F., Ramsey M.W., Jones C.J., Lewis M.J.. Dietary supplementation with marine omega-3 fatty acids improve systemic large artery endothelial function in subjects with hypercholesterolemia. J. Am. Coll. Cardiol. (2000) 35 265-270. DOI: 10.1016/S0735-1097(99)00548-3 3. Siasos G., Tousoulis D., Oikonomou E., Zaromitidou M., Verveniotis A., Plastiras A., Kioufis S., Maniatis K., Miliou A., Siasou Z.. Effects of Ω-3 fatty acids on endothelial function, arterial wall properties, inflammatory and fibrinolytic status in smokers: A cross over study. Int. J. Cardiol. (2013) 166 340-346. DOI: 10.1016/j.ijcard.2011.10.081 4. Hoshi T., Wissuwa B., Tian Y., Tajima N., Xu R., Bauer M., Heinemann S.H., Hou S.. Omega-3 fatty acids lower blood pressure by directly activating large-conductance Ca. Proc. Natl. Acad. Sci. USA (2013) 110 4816-4821. DOI: 10.1073/pnas.1221997110 5. Miller P.E., Van Elswyk M., Alexander D.D.. Long-chain omega-3 fatty acids eicosapentaenoic acid and docosahexaenoic acid and blood pressure: A meta-analysis of randomized controlled trials. Am. J. Hypertens. (2014) 27 885-896. DOI: 10.1093/ajh/hpu024 6. Zehr K.R., Walker M.K.. Omega-3 polyunsaturated fatty acids improve endothelial function in humans at risk for atherosclerosis: A review. Prostaglandins Other Lipid Mediat. (2018) 134 131-140. DOI: 10.1016/j.prostaglandins.2017.07.005 7. Bockus L., Kim F.. Coronary endothelial dysfunction: From pathogenesis to clinical implications. Open Heart (2022) 9 e002200. DOI: 10.1136/openhrt-2022-002200 8. Rossi R., Nuzzo A., Origliani G., Modena M.G.. Prognostic role of flow-mediated dilation and cardiac risk factors in post-menopausal women. J. Am. Coll. Cardiol. (2008) 51 997-1002. DOI: 10.1016/j.jacc.2007.11.044 9. Shechter M., Issachar A., Marai I., Koren-Morag N., Freinark D., Shahar Y., Shechter A., Feinberg M.S.. Long-term association of brachial artery flow-mediated vasodilation and cardiovascular events in middle-aged subjects with no apparent heart disease. Int. J. Cardiol. (2009) 134 52-58. DOI: 10.1016/j.ijcard.2008.01.021 10. Witte D.R., Westerink J., de Koning E.J., van der Graaf Y., Grobbee D.E., Bots M.L.. Is the association between flow-mediated dilation and cardiovascular risk limited to low-risk populations?. J. Am. Coll. Cardiol. (2005) 45 1987-1993. DOI: 10.1016/j.jacc.2005.02.073 11. Modena M.G., Bonetti L., Coppi F., Bursi F., Rossi R.. Prognostic role of reversible endothelial dysfunction in hypertensive postmenopausal women. J. Am. Coll. Cardiol. (2002) 40 505-510. DOI: 10.1016/S0735-1097(02)01976-9 12. Del Gaudio I., Rubinelli L., Sasset L., Wadsack C., Hla T., Di Lorenzo A.. Endothelial Spns2 and ApoM Regulation of Vascular Tone and Hypertension Via Sphingosine-1-Phosphate. J. Am. Heart Assoc. (2021) 10 e021261. DOI: 10.1161/JAHA.121.021261 13. Takaeko Y., Matsui S., Kajikawa M., Maruhashi T., Yamaji T., Harada T., Han Y., Hashimoto H., Kihara Y., Hida E.. Relationship between high-density lipoprotein cholesterol levels and endothelial function in women: A cross-sectional study. BMJ Open (2020) 10 e038121. DOI: 10.1136/bmjopen-2020-038121 14. Acikgoz N., Kurtoğlu E., Yagmur J., Kapicioglu Y., Cansel M., Ermis N.. Elevated Monocyte to High-Density Lipoprotein Cholesterol Ratio and Endothelial Dysfunction in Behçet Disease. Angiology (2018) 69 65-70. DOI: 10.1177/0003319717704748 15. Cartolano F.C., Dias G.D., Miyamoto S., Damasceno N.R.T.. Omega-3 Fatty Acids Improve Functionality of High-Density Lipoprotein in Individuals With High Cardiovascular Risk: A Randomized, Parallel, Controlled and Double-Blind Clinical Trial. Front. Nutr. (2022) 8 767535. DOI: 10.3389/fnut.2021.767535 16. Luna-Luna M., Niesor E., Pérez-Méndez Ó.. HDL as Bidirectional Lipid Vectors: Time for New Paradigms. Biomedicines (2022) 10. DOI: 10.3390/biomedicines10051180 17. Kratzer A., Giral H., Landmesser U.. High-density lipoproteins as modulators of endothelial cell functions: Alterations in patients with coronary artery disease. Cardiovasc. Res. (2014) 103 350-361. DOI: 10.1093/cvr/cvu139 18. Muñoz-Vega M., Massó F., Páez A., Vargas-Alarcón G., Coral-Vázquez R., Mas-Oliva J., Carreón-Torres E., Pérez-Méndez Ó.. HDL-Mediated Lipid Influx to Endothelial Cells Contributes to Regulating Intercellular Adhesion Molecule (ICAM)-1 Expression and eNOS Phosphorylation. Int. J. Mol. Sci. (2018) 19. DOI: 10.3390/ijms19113394 19. López-Olmos V., Luna-Luna M., Carreón-Torres E., González-Pacheco H., Bautista-Pérez R., Posadas-Sánchez R., Fragoso J.M., Vargas-Alarcón G., Pérez-Méndez Ó.. Trp Fluorescence Redshift during HDL Apolipoprotein Denaturation Is Increased in Patients with Coronary Syndrome in Acute Phase: A New Assay to Evaluate HDL Stability. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22157819 20. Tousoulis D., Plastiras A., Siasos G., Oikonomou E., Verveniotis A., Kokkou E., Maniatis K., Gouliopoulos N., Miliou A., Paraskevopoulos T.. Omega-3 PUFAs improved endothelial function and arterial stiffness with a parallel antiinflammatory effect in adults with metabolic syndrome. Atherosclerosis (2014) 232 10-16. DOI: 10.1016/j.atherosclerosis.2013.10.014 21. Siniarski A., Haberka M., Mostowik M., Gołębiowska-Wiatrak R., Poręba M., Malinowski K.P., Gąsior Z., Konduracka E., Nessler J., Gajos G.. Treatment with omega-3 polyunsaturated fatty acids does not improve endothelial function in patients with type 2 diabetes and very high cardiovascular risk: A randomized, double-blind, placebo-controlled study (Omega-FMD). Atherosclerosis (2018) 271 148-155. DOI: 10.1016/j.atherosclerosis.2018.02.030 22. Sherratt S.C.R., Libby P., Budoff M.J., Bhatt D.L., Mason R.P.. Role of Omega-3 Fatty Acids in Cardiovascular Disease: The Debate Continues. Curr. Atheroscler. Rep. (2023) 25 1-17. DOI: 10.1007/s11883-022-01075-x 23. Khorshidi M., Sayyari A., Aryaeian N., Olang B., Alaei M., Khalili M., Hosseini A., Salehi M.. Effects of omega-3 supplementation on endothelial function, vascular structure, and metabolic parameters in adolescents with type 1 diabetes mellitus: A randomized clinical trial. Front. Nutr. (2022) 9 962773. DOI: 10.3389/fnut.2022.962773 24. Fabian C.J., Kimler B.F., Hursting S.D.. Omega-3 fatty acids for breast cancer prevention and survivorship. Breast Cancer Res. (2015) 17 62. DOI: 10.1186/s13058-015-0571-6 25. Cui J., Shan K., Yang Q., Qi Y., Qu H., Li J., Wang R., Jia L., Chen W., Feng N.. Prostaglandin E3 attenuates macrophage-associated inflammation and prostate tumour growth by modulating polarization. J. Cell Mol. Med. (2021) 25 5586-5601. DOI: 10.1111/jcmm.16570 26. Endres S., Ghorbani R., Kelley V.E., Georgilis K., Lonnemann G., van der Meer J.W., Cannon J.G., Rogers T.S., Klempner M.S., Weber P.C.. The effect of dietary supplementation with n-3 polyunsaturated fatty acids on the synthesis of interleukin-1 and tumor necrosis factor by mononuclear cells. N. Engl. J. Med. (1989) 320 265-271. DOI: 10.1056/NEJM198902023200501 27. Liu X., Sun L., Wen W., Qiu M., Luo J., Li W., Hao S., He M., Wu J., Hu Y.. Association between the ratio of serum eicosapentaenoic acid to arachidonic acid and risk of coronary artery disease in young Chinese patients. Front. Nutr. (2022) 9 1019058. DOI: 10.3389/fnut.2022.1019058 28. Koivuniemi A., Sysi-Aho M., Orešič M., Ollila S.. Interfacial properties of high-density lipoprotein-like lipid droplets with different lipid and apolipoprotein A-I compositions. Biophys. J. (2013) 104 2193-2201. DOI: 10.1016/j.bpj.2013.02.058 29. Mason R.P., Jacob R.F., Shrivastava S., Sherratt S.C.R., Chattopadhyay A.. Eicosapentaenoic acid reduces membrane fluidity, inhibits cholesterol domain formation, and normalizes bilayer width in atherosclerotic-like model membranes. Biochim. Biophys. Acta (2016) 1858 3131-3140. DOI: 10.1016/j.bbamem.2016.10.002 30. Bautista R., Carreón-Torres E., Luna-Luna M., Komera-Arenas Y., Franco M., Fragoso J.M., López-Olmos V., Cruz-Robles D., Vargas-Barrón J., Vargas-Alarcón G.. Early endothelial nitrosylation and increased abdominal adiposity in Wistar rats after long-term consumption of food fried in canola oil. Nutrition (2014) 30 1055-1060. DOI: 10.1016/j.nut.2014.01.010 31. James G., Bohannan W., Adewunmi E., Schmidt K., Park H.G., Shchepinov M.S., Agbaga M.P., Brenna J.T.. Pharmacokinetics and metabolism in mouse retina of bis-allylic deuterated docosahexaenoic acid (D-DHA), a new dry AMD drug candidate. Exp. Eye Res. (2022) 222 109193. DOI: 10.1016/j.exer.2022.109193 32. Quintanilla-Cantú A., Peña-de-la-Sancha P., Flores-Castillo C., Mejía-Domínguez A.M., Posadas-Sánchez R., Pérez-Hernández N., Bautista-Pérez R., Enriquez-Calderón R.E., Juárez-Oropeza M.A., Fragoso J.M.. Small HDL subclasses become cholesterol-poor during postprandial period after a fat diet intake in subjects with high triglyceridemia increases. Clin. Chim. Acta (2017) 464 98-105. DOI: 10.1016/j.cca.2016.11.018 33. Estrada-Luna D., Martínez-Hinojosa E., Cancino-Diaz J.C., Belefant-Miller H., López-Rodríguez G., Betanzos-Cabrera G.. Daily supplementation with fresh pomegranate juice increases paraoxonase 1 expression and activity in mice fed a high-fat diet. Eur. J. Nutr. (2018) 57 383-389. DOI: 10.1007/s00394-017-1394-2 34. Johnston D.T., Deuster P.A., Harris W.S., Macrae H., Dretsch M.N.. Red blood cell omega-3 fatty acid levels and neurocognitive performance in deployed U.S. Servicemembers. Nutr. Neurosci. (2013) 16 30-38. DOI: 10.1179/1476830512Y.0000000025 35. Corretti M.C., Anderson T.J., Benjamin E.J., Celermajer D., Charbonneau F., Creager M.A., Deanfield J., Drexler H., Gerhard-Herman M., Herrington D.. Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery: A report of the International Brachial Artery Reactivity Task Force. J. Am. Coll. Cardiol. (2002) 39 257-265. DOI: 10.1016/S0735-1097(01)01746-6 36. Meléndez-Ramírez G., Pérez-Méndez O., López-Osorio C., Kuri-Alfaro J., Espinola-Zavaleta N.. Effect of the treatment with allopurinol on the endothelial function in patients with hyperuricemia. Endocr. Res. (2012) 37 1-6. DOI: 10.3109/07435800.2011.566235
--- title: Cathepsin S Knockdown Suppresses Endothelial Inflammation, Angiogenesis, and Complement Protein Activity under Hyperglycemic Conditions In Vitro by Inhibiting NF-κB Signaling authors: - Shithima Sayed - Omar Faruq - Umma Hafsa Preya - Jee Taek Kim journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049538 doi: 10.3390/ijms24065428 license: CC BY 4.0 --- # Cathepsin S Knockdown Suppresses Endothelial Inflammation, Angiogenesis, and Complement Protein Activity under Hyperglycemic Conditions In Vitro by Inhibiting NF-κB Signaling ## Abstracts Hyperglycemia plays a key role in the development of microvascular complications, endothelial dysfunction (ED), and inflammation. It has been demonstrated that cathepsin S (CTSS) is activated in hyperglycemia and is involved in inducing the release of inflammatory cytokines. We hypothesized that blocking CTSS might alleviate the inflammatory responses and reduce the microvascular complications and angiogenesis in hyperglycemic conditions. In this study, we treated human umbilical vein endothelial cells (HUVECs) with high glucose (HG; 30 mM) to induce hyperglycemia and measured the expression of inflammatory cytokines. When treated with glucose, hyperosmolarity could be linked to cathepsin S expression; however, many have mentioned the high expression of CTSS. Thus, we made an effort to concentrate on the immunomodulatory role of the CTSS knockdown in high glucose conditions. We validated that the HG treatment upregulated the expression of inflammatory cytokines and CTSS in HUVEC. Further, siRNA treatment significantly downregulated CTSS expression along with inflammatory marker levels by inhibiting the nuclear factor-kappa B (NF-κB) mediated signaling pathway. In addition, CTSS silencing led to the decreased expression of vascular endothelial markers and downregulated angiogenic activity in HUVECs, which was confirmed by a tube formation experiment. Concurrently, siRNA treatment reduced the activation of complement proteins C3a and C5a in HUVECs under hyperglycemic conditions. These findings show that CTSS silencing significantly reduces hyperglycemia-induced vascular inflammation. Hence, CTSS may be a novel target for preventing diabetes-induced microvascular complications. ## 1. Introduction Hyperglycemia and hyperglycemia-induced vascular complications are associated with increased risks of high blood pressure, coronary heart disease, and strokes [1,2]. Hyperglycemia has direct impacts on vascular systems, and vascular complications are a major cause of morbidity and mortality in diabetic patients. High levels of glucose elevate the flux of sugar through the polyol pathway, leading to intracellular sorbitol accumulation that increases osmotic stress [3]. The development of diabetic complications has been linked to a number of pathophysiologic pathways, including the polyol pathway, non-enzymatic glycation, oxidative stress, and activation of protein kinase C. These mechanisms are mostly associated with excessive glucose transport into retinal cells [4]. Furthermore, inflammation plays a key role in the development of diabetic complications [5]. Micro- and macrovascular diabetic complications are largely due to long-term hyperglycemia, in addition to other major complications [2]. Retinopathy, nephropathy, and diabetic vascular diseases are due to the effects of high blood sugar on retinal, endothelial, and mesangial cells [6]. Microvascular endothelial cells can potentially be damaged in hyperglycemia since they are unable to downregulate the rate of glucose transport, resulting in elevated concentrations of glucose within the cells [7]. This leads to microvascular complications, including increased leukocyte adhesion, permeability, and procoagualant activity [7,8]. Furthermore, hyperglycemia-induced disruptions of vascular homeostasis in endothelial cells promote the release of inflammatory cytokines as well as the production of free radicals and reactive oxygen species (ROS) [9]. As a result, cellular damage and inflammation often progress in hyperglycemic states. Moreover, vascular remodeling via extracellular matrix (ECM) degradation is explicitly responsible for microvasculature alterations and neovascularization. In the early stages of diabetic retinopathy, microaneurysms have been observed, whereas, in the final stages, angiogenesis and fibrovascular proliferation are commonly seen [10]. Cathepsin S (CTSS) is a lysosomal protease and a unique member of the cysteine cathepsin protease family. It is involved in a variety of pathological processes, including cancer, cardiovascular disease, and arthritis. It has been demonstrated that CTSS expression is significantly higher in diabetes patients, and it serves as a robust biomarker for diabetes and atherosclerosis [11,12]. Further, it has a critical role in vascular remodeling via ECM degradation as an elastolytic protease and is related to severe microvascular complications [13,14]. Moreover, CTSS is associated with plasma leakages, which contribute to vascular wall degeneration and microvasculature damage [15]. Previous studies have reported that CTSS is involved in the inflammatory processes of immune diseases such as atopic dermatitis, psoriasis, bronchial asthma, and rheumatoid arthritis [16,17], as well as in autoantigen presentation in the immune system [18]. In recent research, CTSS has been shown to play an important role in invasion, metastasis, and tumorigenesis [19]. Furthermore, it has been linked to the development of type 2 diabetes, where it is elevated at both the mRNA and protein levels [20]. CTSS causes chronic inflammation during diabetic complications that is responsible for neurovascular degeneration [21,22] and promotes microvascular complications. CTSS activities were associated with the macrophage subpopulation in NOD mice and in human type 1 diabetic samples [23]. Cathepsin silencing reduces hyperglycemia-mediated complications and delays the progression of diabetes [24]. However, few studies have observed the relationship between silencing CTSS and vascular complications under hyperglycemia. Recent research has focused on understanding and determining the effects of hyperglycemia on specific types of endothelial cells (ECs) or vasculature components. ECs are found in various organs and vessels, with different functions depending on their location [25,26,27]. Hyperglycemic endothelial dysfunction (ED) presents differently in ECs from different parts of the vasculature. ECs in the aorta and lung exhibit differential dipeptidyl peptidase-4 expression, as well as different activity levels under hyperglycemia [28]. In addition, Karabach et al. observed significantly different cell viability and free radical formation in EA.hy.926 and human umbilical vein endothelial cells (HUVECs) under hyperglycemia [29]. Since HUVECs are a promising candidate for the in vitro study of hyperglycemia-induced vascular complications, we investigated the effects of a CTSS knockout in HUVECs in a hyperglycemic environment. The findings of this study may shed light on the importance of CTSS silencing in hyperglycemia and establish effective guidelines for the development of novel therapeutics to treat hyperglycemic-related vascular complications. ## 2.1. Effects of Different Glucose Concentrations on HUVECs’ Viability HUVEC viability was evaluated after 24 h of treatment with various concentrations of glucose. Figure 1A shows that after 24 h of treatment with 15 mM and 30 mM glucose, a cell viability of over $80\%$ was observed. In contrast, less than $80\%$ of cells were viable after 24 h of treatment with 60 mM glucose. These results confirm that treatment with 15 mM glucose achieves a cell viability of over $95\%$ and, hence, exerts only a minor hyperglycemic effect on the cells (Figure 1A). Therefore, to induce hyperglycemia in HUVECs, we selected a 30 mM glucose concentration for 24 h for further experiments. ## 2.2. Expressions of CTSS and Other Pro-Inflammatory Cytokines Are Induced by High Glucose (HG) in HUVECs To study whether CTSS expression was affected by hyperglycemic conditions, HUVECs were treated with the 30 mM glucose treatment for 24 h. The expressions of pro-inflammatory cytokines and CTSS were increased in HG-treated samples compared to those in the controls (Figure 1B), as evaluated by quantitative real-time polymerase chain reaction (qRT-PCR). Western blot (WB) analysis confirmed that CTSS, VEGFA, and other inflammatory markers (Figure 1C) were overexpressed in HG-treated samples than in the controls. Inflammatory markers TNF-α, IL-1β, IL-6, NF-κB, MCP-1, COX-2, VEGFA, and iNOS, and adhesion markers VCAM-1 and ICAM-1 showed higher expression after HG treatment, as revealed by qRT-PCR and WB. These results suggest that treatment with 30 mM glucose successfully mimicked hyperglycemia conditions in HUVECs. ## 2.3. Effects of CTSS Knockdown on HUVECs The effect of the CTSS knockdown was evaluated using qRT-PCR, WB, and immunostaining in HUVECs post-siRNA treatment. CTSS expression was downregulated after the siRNA treatment (Figure 1D,E). Immunostained images indicate the expression of CTSS in HG-treated HUVECs with and without CTSS siRNA transfection (Figure 1F). CTSS was highly expressed after HG treatment and significantly decreased after CTSS knockdown. After siRNA transfection of HUVECs with and without HG treatment, the cytotoxic effects of CTSS knockdown were examined (Figure 1G). Over $80\%$ cell viability was observed after 24 h of transfection. Fluorescence micrographs of different HUVEC samples showed similar cell proliferation. These data confirmed that siRNA transfection had no lethal effects on the cells. ## 2.4. CTSS Knockdown Inhibits the Expression of Pro-Inflammatory Markers To investigate whether silencing CTSS could reduce the expression of pro-inflammatory cytokines in HUVECs, their expressions at the gene (Figure 2A–G) and protein (Figure 2H) level were examined by qRT-PCR and WB. Pro-inflammatory cytokine expression was upregulated in HG-treated samples (siCON + HG) compared to that in the control samples (siCON) and significantly downregulated in the CTSS siRNA-transfected samples (siCTSS). Samples treated with both HG and CTSS siRNA (siCTSS + HG) also showed significant downregulation of the expression of all the studied gene markers. The localization of a major inflammatory cytokine, TNF-α, was visualized by immunostaining and fluorescence microscopy (Figure 2I). TNF-α was expressed at higher levels in the HG-treated sample, while its expression sharply decreased after CTSS siRNA treatment, which is in agreement with the qRT-PCR data (Figure 2I). ## 2.5. CTSS Knockdown Inhibits Vascular Endothelial Growth Factor and Vascular Adhesion Marker Expression The gene expressions of vascular endothelial growth marker (VEGFA) and vascular adhesion markers (VCAM-1 and ICAM-1) were investigated by qRT-PCR. CTSS knockdown (siCTSS + HG) caused significant downregulation of these markers compared to that in the controls (siCON + HG) (Figure 3A–C). The results of WB protein analysis largely agreed with the gene expression data (Figure 3D). These results (Figure 2 and Figure 3A–D) suggest that HG induces inflammatory cytokine expression while CTSS silencing decreases their expression. CTSS knockdown also significantly reduced the expressions of transcription factors: nuclear factor-kappa B (NF-κB) and inducible NO synthase (iNOS), along with those of VEGFA, ICAM-1, and VCAM-1. In contrast, higher expression was only observed in HG-treated samples (siCON + HG). ## 2.6. CTSS Knockdown Inhibits Angiogenesis To determine the effects of CTSS siRNA on angiogenesis and vascular remodeling, pretreated HUVECs were seeded on Matrigel to observe tube formation. Hyperglycemia induced tube formation in siCON + HG samples (Figure 3E–I). The CTSS knockdown inhibited tube formation even under hyperglycemic conditions. The inhibitory effects of CTSS siRNA treatment on angiogenesis in SiCTSS and SiCTSS + HG cells can be seen in Figure 3E–I. ## 2.7. CTSS Knockdown Downregulates Complement Activation in HUVECs To study the associations between CTSS and complement activation in hyperglycemia, the expression patterns of C3, C5, C3a, and C5a were analyzed. C3 and C5 mRNA levels were elevated in HG-treated samples (siCON + HG) but downregulated in samples that underwent siRNA treatment (siCTSS and siCTSS + HG) (Figure 4A,B). C3a and C5a protein levels were also increased in HG-treated samples (siCON + HG) and decreased in samples with CTSS silencing (siCTSS and siCTSS + HG) (Figure 4C). Immunocytochemical staining of C5a (Figure 4D) revealed trends similar to those observed in the WB and qRT-PCR data. ## 3. Discussion Hyperglycemia is responsible for microvascular damage, inducing a pro-inflammatory state that is the main feature of vasculopathy [30,31]. HUVECs were treated with HG to mimic hyperglycemic conditions in the current study, and hyperglycemia-associated changes in gene expression were assessed. The in vitro HUVEC model is a popular method for studying hyperglycemia-induced vascular complications and identifying new targets. Our study shows that pro-inflammatory cytokines, along with VEGFA and adhesion molecules, were sharply upregulated under hyperglycemic conditions (Figure 1B,C), which is in agreement with previous studies [32,33]. ECs can act as regulators of inflammatory processes and function as innate immune cells under many pathophysiologic conditions [33]. In patients with hyperglycemia, researchers have detected chronic low-grade, subclinical inflammation mediated by TNF-α, COX-2, IL-1β, IL-6, and MCP-1, which is responsible for vascular complications [32,34]. Chronic inflammation contributes to neurodegeneration, pericyte depletion, EC loss leading to ED, and disruptions of the blood-retinal barrier (BRB) [22,32]. Further, the toxic effects of HG are mediated by increased levels of oxidative stress [31] which induce lysosomal leakage through increased lysosomal permeabilization, releasing CTSS into the cytoplasm [35]. Cathepsin expression is influenced by metabolic status and biochemical requirements [15,18,36]. Moreover, CTSS induces protease-activated receptor (PAR)-2 on ECs and leads to ED [37]. Since CTSS is upregulated under high blood glucose levels in vivo and in diabetic rat models [24,38], overexpression of CTSS can be considered a pathological factor in the development of diabetes. We examined CTSS activity in HUVECs under HG conditions before and after CTSS silencing. Our study shows that CTSS siRNA effectively inhibited CTSS expression under HG conditions (Figure 1D–F), leading to reductions in inflammation, angiogenesis, and vascular complications. It has been demonstrated that hyperglycemia induces inflammation that is characterized by the activation of transcription factors, such as NF-κB, as well as increased chemokine and cytokine expression [39]. Higher CTSS levels are linked to inflammatory cytokine release, which leads to the onset of type 2 diabetes. In this investigation, CTSS knockdown significantly reduced HG-induced expression of pro-inflammatory cytokines, including MCP-1, COX-2, iNOS, IL-1β, IL-6, and NF-κB (Figure 2). NF-κB is a pivotal inflammatory mediator that serves as both an innate and an adaptive immune response regulator [40]. In vascular ECs, genes activated by NF-κB include those encoding for various inflammatory chemokines, cytokines, and adhesion molecules [40,41]. Further, this study observed that HG causes the upregulation of inducible enzymes such as iNOS and COX-2, which are also regulated by NF-κB [40,41]. COX-2 and iNOS are known to facilitate HUVEC apoptosis and angiogenesis via the overproduction of nitric oxide [42]. Our results demonstrate that CTSS knockdown inhibited COX-2 and iNOS expression, which subsequently prevented apoptosis and improved cell viability. TNF-α acts as a master regulator of inflammatory cytokine and ROS production; thus, it is necessary for both cell survival and death [43]. TNF-α protein levels in HG-treated samples were significantly higher than in the controls (Figure 2I); they were associated with increased ROS generation and cell apoptosis. Previous studies suggested ROS has a significant role in the synthesis of pro-inflammatory cytokines, such as TNF-α, IL-1β, etc. [ 44]. ROS causes the expression of pro-inflammatory cytokines by activating NF-κB pathways [45,46]. Besides the role of CTSS in inflammatory cytokine regulation, this study determined the effect of CTSS silencing on adhesion molecules such as VCAM-1 and ICAM-1. Similar to other studies, our study showed that hyperglycemia increased the expression of VCAM-1, ICAM-1, and VEGF-1 in HUVEC cells [29,47]. Hyperglycemia has been shown to increase the permeability of HUVEC monolayers, which is an early pathological mechanism in the progression of diabetic vascular complications [48]. MCP-1, VCAM-1, and ICAM-1 pro-inflammatory cytokines are expressed during ROS production and NF-κB translocation, which facilitate monocyte attachment to the vascular endothelium. Our findings showed that HG increased VEGFA expression in HUVECs while CTSS silencing significantly decreased VCAM-1, ICAM-1, and VEGFA expression (Figure 3A–D). VEGFA increases the adherence of leukocytes to vessel walls by upregulating the expression of ICAM-1 and VCAM-1. In addition, VEGFA is responsible for ED, angiogenesis, vascular permeability, and macular edema [49] by stimulating ECs through a protein kinase C (PKC)-dependent mechanism that subsequently enhances retinal and glomerular permeability [49,50]. In the current study, we found that HG levels increased VEGFA expression in HUVECs while CTSS silencing significantly decreased VCAM-1, ICAM-1, and VEGFA expression (Figure 3A–D). A close association has been observed between angiogenesis and inflammation under pathological conditions [51]. In addition to angiogenesis, VEGFA regulates HG-mediated increases in the expression of inflammatory cytokines IL-6 and MCP-1, as well as the adhesion molecule ICAM-1, via the NF-κB pathway activation. These factors contribute to angiogenesis and leukocyte recruitment [52]. In this study, significant tube formation was observed in HG-treated cells (Figure 3E–I), indicating that HG triggered angiogenesis and neovascularization in HUVECs. On the other hand, CTSS siRNA treatment under HG conditions suppressed tube formation as well as decreasing vascularization. The complement (C) system is known to be an adaptive and innate immunity effector that is also responsible for neovascularization and angiogenesis [53,54]. Hyperglycemia induces C3 and C5 upregulation through NF-κB signaling, which releases pro-inflammatory cytokines and induces angiogenesis [55,56]. C3a and C5a activation can also cause macrophage-mediated angiogenesis via the release of VEGFA, IL-6, and TNF-α [53]. However, the relationship between CTSS and complement activation has not been fully elucidated. We demonstrated that CTSS knockdown suppressed the expression of C3 and C5 (Figure 4A,B) as well as the activation of C3a and C5a (Figure 4C,D) under hyperglycemic conditions. Overall, hyperglycemia-induced cytokine release and vascular complications in HUVEC cells are promising tools for the study of diabetic complications. This study’s limitation include that we did not check whether the changes in CTSS caused by HG treatment were associated with hyperosmolarity or not. Further study of the relationship between high expressions of CTSS and hyperosmolarity would provide more robust data. Since the change in CTSS was already known for diabetic conditions, our study tried to reveal the immunomodulatory effect of CTSS silencing in an HG condition. More research is needed to understand the pathophysiological role of CTSS in hyperglycemia, but the findings of the current study provide strong evidence that CTSS plays a role in hyperglycemia-induced inflammation, angiogenesis, and vasculogenesis. CTSS silencing reduced the expression of pro-inflammatory cytokines, chemokines, and complement factors under hyperglycemic conditions (illustrated in Figure 5). These results suggest that CTSS may be considered a therapeutic target for the control and prevention of the development of hyperglycemic complications. ## 4.1. Antibodies and Chemical Reagents Fetal bovine serum, Dulbecco’s phosphate-buffered saline, and 1X Trypsin-EDTA solution were obtained from Welgene (Gyeongsan-si, Gyeongsangbuk-do, Republic of Korea). Penicillin–streptomycin solution was acquired from HyClone Laboratories Inc. (South Logan, NY, USA). Dimethyl sulfoxide (DMSO) and 3-[4,5-dimethyl-2-thiazolyl]-2,5-diphenyltetrazoliumbromide (MTT) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Paraformaldehyde ($4\%$) was obtained from Biosesang (Seongnam-si, Gyeonggi-do, Republic of Korea), and Matrigel was purchased from BD Biosciences (Bedford, MA, USA). HUVECs (C2517A), endothelial basal medium-2 (EBMTM-2, CC-3156), and an endothelial cell growth medium-2 (EGMTM-2) BulletKit (CC-3162) were obtained from Lonza (Walkersville, MD, USA). Halt™ Protease and Phosphatase Inhibitor Cocktail, a Pierce™ RIPA Buffer (RIPA: radio-immunoprecipitation assay), and a Pierce™ BCA Protein Assay Kit were acquired from Thermo Fisher Scientific (Rockford, IL, USA). Opti-MEM™ reduced serum medium was also purchased from Thermo Fisher Scientific (Grand Island, NY, USA). Phenylmethylsulphonyl fluoride was purchased from Roche (Mannheim, Germany). Mounting medium with 4′,6-diamidino-2-phenylindole (DAPI)-Aqueous Fluoroshield and antibodies against VEGFA and COX-2 were acquired from Abcam (Cambridge, MA, USA). Antibodies against TNF-α were obtained from R&D Systems (Minneapolis, MN, USA). CTSS siRNA (h), antibodies against CTSS, ICAM-1, IL-6, and β-actin, and mouse anti-goat IgG-HRP, mouse anti-rabbit IgG-HRP, and goat anti-mouse IgG- conjugated with HRP were purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Lipofectamine RNAiMAX reagent, calcein AM fluorescent dye, wheat germ agglutinin-Alexa Fluor™ 488 conjugate, Alexa Fluor 594 donkey anti-goat IgG (H&L), Alexa Fluor 594 goat anti-mouse IgG (H&L), and antibodies against MCP-1, NFκB p65, and iNOS were purchased from Invitrogen (Carlsbad, CA, USA). TB Green® Premix Ex Taq™ was purchased from Takara Bio Inc. (Kusatsu, Shiga, Japan). The protein loading buffer and iScript cDNA synthesis kit were obtained from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). The Ribospin II RNA extraction kit was purchased from GeneAll Biotechnology co., Ltd. (Songpa-gu, Seoul, Republic of Korea). ## 4.2. Cell Culture HUVECs (Lonza, Walkersville, MD, USA) were cultured in EGM-2 growth kit-supplemented EBM-2 medium. A trypan blue exclusion assay was used to quantify mononuclear cells in small aliquots before seeding them into a 150 mm culture dish at a density of 1.0 × 106 mL−1. This cell seeding density was used for subsequent experiments. HUVEC stocks were kept in a 37 °C incubator with $5\%$ CO2. ## 4.3. Treatment of HUVECs with High Glucose Levels HUVECs were seeded onto six-well tissue culture plates and incubated with EBM-2, supplemented with $2\%$ FBS and $1\%$ gentamicin, and maintained at 37 °C in a humidified incubator. For the HG treatments, final glucose concentrations of 15 mM, 30 mM, and 60 mM in EBM-2 were prepared and applied after $80\%$ confluence was reached at 24 h and 48 h. The highest nontoxic dose and exposure time were selected for further experiments. ## 4.4. Small Interfering RNA (siRNA) Transfection HUVECs were seeded with antibiotic-free culture medium containing EBM-2 (from the EGM-2 BulletKit) onto six-well tissue culture plates. After reaching 50–$60\%$ confluence, the cells were transfected with CTSS siRNA (h) (pool of 3 target-specific 19–25 nt siRNAs from Santa Cruz Biotechnology) and control siRNA at a final concentration of 10 nM using Lipofectamine RNAiMAX transfection reagent according to the manufacturer’s protocol. After transfection for 24 h, HUVECs were treated with 30 mM glucose for 24 h and then processed for further experiments. ## 4.5. Cell Viability and Proliferation Assay In vitro cell viability under HG and after siRNA transfection was assessed via the MTT assay using a standard testing protocol [ISO10993-5:2009I]. The MTT assay was performed after 24 h and 48 h of treatment under different concentrations of HG as well as after 24 h of siRNA transfection. The cell viabilities (determined by measuring the optical density) were determined by following the previous protocol [57]. In brief, MTT solution (5 mg/mL in PBS) was added at a ratio of 1:9 to the cell culture media, followed by incubation for 4 h at 37 °C to allow for formazan crystal formation. Subsequently, dissolution was carried out in DMSO for 1 h to extract the formazan crystals. Finally, the absorbance was measured using an ELISA reader (EL, 312, Biokinetics reader; Bio-Tek Instruments, Winooski, VT, USA) at 595 nm. All samples were tested in triplicate. In addition, according to the previous protocol [58], the growth and proliferation of siRNA-transfected HUVECs were observed using a fluorescence microscope (Leica DMi8, Wetzlar, Germany) after immunostaining the cells with wheat germ agglutinin-Alexa Fluor 488 conjugate for 10 min. ## 4.6. Quantitative RT-PCR Total RNA was prepared after 24 h of HG treatment using a Ribospin II RNA extraction kit (GeneAll Biotechnology). Next, a NanoDrop-One spectrophotometer (Thermo Fisher Scientific, Cambridge, MA, USA) was used to measure the RNA concentrations, and the RNA was reverse-transcribed into cDNA using an iScript cDNA synthesis kit (Bio-Rad Laboratories, Inc. Hercules, CA, USA) in a SimpliAmp Thermal Cycler (Applied Biosystems, Singapore). All cDNA samples were kept at −20 °C for subsequent gene expression analysis. Using TB Green® Premix Ex Taq™ (Takara Bio Inc., Kusatsu, Shiga, Japan) and a CFX96™ RT-PCR detection system, qRT-PCR was performed (Bio-Rad, Singapore). All real-time PCR experiments were performed in triplicate. The data were normalized against GAPDH levels, and the relative mRNA expression level was calculated using the 2−△△Ct method. The nucleotide sequences of the primers used are provided in Table 1. ## 4.7. WB Analysis Total protein was collected using a RIPA lysis buffer containing a 1X protease and phosphatase inhibitor cocktail. A BCA protein assay kit (Thermo Scientific, Rockford, IL, USA) was used to measure the total protein concentration. Protein samples in equal amounts were separated through $8\%$ and $12\%$ SDS-PAGE and transferred onto polyvinylidene difluoride membranes. Next, the membranes were incubated with specific primary antibodies overnight and subsequently incubated with specific secondary antibodies (HRP-conjugated). Chemiluminescent ECL reagents were used to detect the bands, and densitometric analyses were performed using a ChemiDoc™ XRS+ system (Bio-Rad, Hercules, CA, USA). As a loading control, β-actin was used. ## 4.8. Immunocytochemical Analysis CTSS siRNA-transfected HUVECs were fixed in $4\%$ paraformaldehyde solution after 24 h of HG treatment. The cells were permeabilized with $0.25\%$ Triton X-100 and blocked with $2.5\%$ bovine serum albumin (BSA). The cells were then immunostained overnight at 4 °C with antibodies against mouse CTSS (1:50), mouse C5a (1:50), and goat TNF-α (1:50). After incubation with the primary antibodies, the cells were incubated with the secondary antibody (Alexa Fluor 594, 1:1000). F-actin was stained with wheat germ agglutinin-Alexa Fluor 488 conjugate for 10 min. The cells were rinsed and mounted aqueously with Fluoroshield-DAPI. A fluorescence microscope (Leica DMi8) with Leica LAS AF software was used to visualize the images. ## 4.9. Tube Formation Assay Pre-cooled 48-well plates were coated with Matrigel (phenol red free). The coating procedure was carried out on ice. siRNA-transfected and HG-treated HUVECs were harvested and diluted (1 × 107 cells/mL) in EBM-2 medium containing low levels of serum. The treated HUVECs were then seeded on Matrigel-coated 48-well plates in triplicate and kept in a 37 °C incubator. Tube formation was observed from 4 h to 6 h. Images of tube formation were captured. To obtain fluorescent images of tube formation, 2 μM calcein AM solution was added, and the samples were then incubated at 37 °C for 20 min. Tube formation was observed, and a fluorescence microscope (Leica DMi8) was used to capture the phase-contrast and fluorescence images. ## 4.10. Statistical Analyses All data are expressed as the mean ± SD (standard deviation), and all experiments were repeated at least three times. Significant differences were determined using Bonferroni’s multiple comparison test with a one-way analysis of variance (ANOVA) between the control and other groups. Statistically significant differences based on the multiple comparisons were defined and marked as follows: * $p \leq 0.05$, $p \leq 0.01$, * $p \leq 0.001$ versus control samples; # $p \leq 0.05$, ## $p \leq 0.01$, ### $p \leq 0.001$ versus HG-treated samples. GraphPad Prism 8 was used to perform all statistical analyses. ## References 1. Paneni F., Beckman J.A., Creager M.A., Cosentino F.. Diabetes and vascular disease: Pathophysiology, clinical consequences, and medical therapy: Part I. Eur. Heart J. (2013) 34 2436-2443. DOI: 10.1093/eurheartj/eht149 2. DeFronzo R.A., Ferrannini E.. Insulin resistance: A multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care (1991) 14 173-194. DOI: 10.2337/diacare.14.3.173 3. Tang W.H., Martin K., Hwa J.. Aldose reductase, oxidative stress, and diabetic mellitus. Front. Pharmacol. (2012) 3 87. DOI: 10.3389/fphar.2012.00087 4. Cai J., Boulton M.. The pathogenesis of diabetic retinopathy: Old concepts and new questions. Eye (2002) 16 242-260. DOI: 10.1038/sj.eye.6700133 5. Joussen A.M., Poulaki V., Le M.L., Koizumi K., Esser C., Janicki H., Schraermeyer U., Kociok N., Fauser S., Kirchhof B.. A central role for inflammation in the pathogenesis of diabetic retinopathy. FASEB J. (2004) 18 1450-1452. DOI: 10.1096/fj.03-1476fje 6. Naudi A., Jove M., Ayala V., Cassanye A., Serrano J., Gonzalo H., Boada J., Prat J., Portero-Otin M., Pamplona R.. Cellular dysfunction in diabetes as maladaptive response to mitochondrial oxidative stress. Exp. Diabetes Res. (2012) 2012 696215. DOI: 10.1155/2012/696215 7. Beckman J.A., Creager M.. Vascular Complications of Diabetes. Circ. Res. (2016) 118 1771-1785. DOI: 10.1161/CIRCRESAHA.115.306884 8. Cheung N., Mitchell P., Wong T.. Diabetic retinopathy. Lancet (2010) 376 124-136. DOI: 10.1016/S0140-6736(09)62124-3 9. Giacco F., Brownlee M.. Oxidative stress and diabetic complications. Circ. Res. (2010) 107 1058-1070. DOI: 10.1161/CIRCRESAHA.110.223545 10. Rezzola S., Loda A., Corsini M., Semeraro F., Annese T., Presta M., Ribatti D.. Angiogenesis-Inflammation Cross Talk in Diabetic Retinopathy: Novel Insights From the Chick Embryo Chorioallantoic Membrane/Human Vitreous Platform. Front. Immunol. (2020) 11 581288. DOI: 10.3389/fimmu.2020.581288 11. Liu J., Ma L., Yang J., Ren A., Sun Z., Yan G., Sun J., Fu H., Xu W., Hu C.. Increased serum cathepsin S in patients with atherosclerosis and diabetes. Atherosclerosis (2006) 186 411-419. DOI: 10.1016/j.atherosclerosis.2005.08.001 12. Smyth P., Sasiwachirangkul J., Williams R., Scott C.J.. Cathepsin S (CTSS) activity in health and disease-A treasure trove of untapped clinical potential. Mol. Aspects Med. (2022) 88 101106. DOI: 10.1016/j.mam.2022.101106 13. Shi G.P., Sukhova G.K., Kuzuya M., Ye Q., Du J., Zhang Y., Pan J.H., Lu M.L., Cheng X.W., Iguchi A.. Deficiency of the cysteine protease cathepsin S impairs microvessel growth. Circ. Res. (2003) 92 493-500. DOI: 10.1161/01.RES.0000060485.20318.96 14. Shi G.P., Villadangos J.A., Dranoff G., Small C., Gu L., Haley K.J., Riese R., Ploegh H.L., Chapman H.A.. Cathepsin S required for normal MHC class II peptide loading and germinal center development. Immunity (1999) 10 197-206. DOI: 10.1016/S1074-7613(00)80020-5 15. Jobs E., Riserus U., Ingelsson E., Sundstrom J., Jobs M., Nerpin E., Iggman D., Basu S., Larsson A., Lind L.. Serum cathepsin S is associated with decreased insulin sensitivity and the development of type 2 diabetes in a community-based cohort of elderly men. Diabetes Care (2013) 36 163-165. DOI: 10.2337/dc12-0494 16. Ainscough J.S., Macleod T., McGonagle D., Brakefield R., Baron J.M., Alase A., Wittmann M., Stacey M.. Cathepsin S is the major activator of the psoriasis-associated proinflammatory cytokine IL-36gamma. Proc. Natl. Acad Sci. USA (2017) 114 E2748-E2757. DOI: 10.1073/pnas.1620954114 17. Kim N., Bae K.B., Kim M.O., Yu D.H., Kim H.J., Yuh H.S., Ji Y.R., Park S.J., Kim S., Son K.H.. Overexpression of cathepsin S induces chronic atopic dermatitis in mice. J. Investig. Dermatol. (2012) 132 1169-1176. DOI: 10.1038/jid.2011.404 18. Reiser J., Adair B., Reinheckel T.. Specialized roles for cysteine cathepsins in health and disease. J. Clin. Investig. (2010) 120 3421-3431. DOI: 10.1172/JCI42918 19. Fan Q., Wang X., Zhang H., Li C., Fan J., Xu J.. Silencing cathepsin S gene expression inhibits growth, invasion and angiogenesis of human hepatocellular carcinoma in vitro. Biochem. Biophys. Res. Commun. (2012) 425 703-710. DOI: 10.1016/j.bbrc.2012.07.013 20. Li X., Wu K., Edman M., Schenke-Layland K., MacVeigh-Aloni M., Janga S.R., Schulz B., Hamm-Alvarez S.F.. Increased expression of cathepsins and obesity-induced proinflammatory cytokines in lacrimal glands of male NOD mouse. Investig. Ophthalmol. Vis. Sci. (2010) 51 5019-5029. DOI: 10.1167/iovs.09-4523 21. Abcouwer S.F.. Muller Cell-Microglia Cross Talk Drives Neuroinflammation in Diabetic Retinopathy. Diabetes (2017) 66 261-263. DOI: 10.2337/dbi16-0047 22. Spencer B.G., Estevez J.J., Liu E., Craig J.E., Finnie J.W.. Pericytes, inflammation, and diabetic retinopathy. Inflammopharmacology (2020) 28 697-709. DOI: 10.1007/s10787-019-00647-9 23. Korpos E., Kadri N., Kappelhoff R., Wegner J., Overall C.M., Weber E., Holmberg D., Cardell S., Sorokin L.. The peri-islet basement membrane, a barrier to infiltrating leukocytes in type 1 diabetes in mouse and human. Diabetes (2013) 62 531-542. DOI: 10.2337/db12-0432 24. Lafarge J.C., Pini M., Pelloux V., Orasanu G., Hartmann G., Venteclef N., Sulpice T., Shi G.P., Clement K., Guerre-Millo M.. Cathepsin S inhibition lowers blood glucose levels in mice. Diabetologia (2014) 57 1674-1683. DOI: 10.1007/s00125-014-3280-2 25. Geenen I.L., Molin D.G., van den Akker N.M., Jeukens F., Spronk H.M., Schurink G.W., Post M.J.. Endothelial cells (ECs) for vascular tissue engineering: Venous ECs are less thrombogenic than arterial ECs. J. Tissue Eng. Regen. Med. (2015) 9 564-576. DOI: 10.1002/term.1642 26. Salcedo R., Resau J.H., Halverson D., Hudson E.A., Dambach M., Powell D., Wasserman K., Oppenheim J.J.. Differential expression and responsiveness of chemokine receptors (CXCR1-3) by human microvascular endothelial cells and umbilical vein endothelial cells. FASEB J. (2000) 14 2055-2064. DOI: 10.1096/fj.99-0963com 27. Zamora D.O., Riviere M., Choi D., Pan Y., Planck S.R., Rosenbaum J.T., David L.L., Smith J.R.. Proteomic profiling of human retinal and choroidal endothelial cells reveals molecular heterogeneity related to tissue of origin. Mol. Vis. (2007) 13 2058-2065. PMID: 18079679 28. Pala L., Pezzatini A., Dicembrini I., Ciani S., Gelmini S., Vannelli B.G., Cresci B., Mannucci E., Rotella C.M.. Different modulation of dipeptidyl peptidase-4 activity between microvascular and macrovascular human endothelial cells. Acta Diabetol. (2012) 49 S59-S63. DOI: 10.1007/s00592-010-0195-3 29. Karbach S., Jansen T., Horke S., Heeren T., Scholz A., Coldewey M., Karpi A., Hausding M., Kroller-Schon S., Oelze M.. Hyperglycemia and oxidative stress in cultured endothelial cells--a comparison of primary endothelial cells with an immortalized endothelial cell line. J. Diabetes Compl. (2012) 26 155-162. DOI: 10.1016/j.jdiacomp.2012.03.011 30. Frank R.N.. Diabetic retinopathy. N. Engl. J. Med. (2004) 350 48-58. DOI: 10.1056/NEJMra021678 31. Kowluru R.A., Chan P.. Oxidative stress and diabetic retinopathy. Exp. Diabetes Res. (2007) 2007 43603. DOI: 10.1155/2007/43603 32. Adamis A.P.. Is diabetic retinopathy an inflammatory disease?. Br. J. Ophthalmol. (2002) 86 363-365. DOI: 10.1136/bjo.86.4.363 33. Shao Y., Saredy J., Yang W.Y., Sun Y., Lu Y., Saaoud F., Drummer C.t., Johnson C., Xu K., Jiang X.. Vascular Endothelial Cells and Innate Immunity. Arterioscler. Thromb. Vasc. Biol. (2020) 40 e138-e152. DOI: 10.1161/ATVBAHA.120.314330 34. Creager M.A., Luscher T.F., Cosentino F., Beckman J.A.. Diabetes and vascular disease: Pathophysiology, clinical consequences, and medical therapy: Part I. Circulation (2003) 108 1527-1532. DOI: 10.1161/01.CIR.0000091257.27563.32 35. Chwieralski C.E., Welte T., Buhling F.. Cathepsin-regulated apoptosis. Apoptosis (2006) 11 143-149. DOI: 10.1007/s10495-006-3486-y 36. Hou W.S., Li W., Keyszer G., Weber E., Levy R., Klein M.J., Gravallese E.M., Goldring S.R., Bromme D.. Comparison of cathepsins K and S expression within the rheumatoid and osteoarthritic synovium. Arthritis Rheum (2002) 46 663-674. DOI: 10.1002/art.10114 37. Kumar Vr S., Anders H.. Cathepsins are potential therapeutic targets in kidney disease. Kidney Int. (2016) 90 933-935. DOI: 10.1016/j.kint.2016.07.034 38. Xu L., Feng B., Wang H., Li X.. Multiple statistical methods for assessing differential gene expression in microarray data of diabetic model rats to predict the molecular mechanism of atorvastatin on anti-atherogenesis. Exp. Clin. Endocrinol. Diabetes (2013) 121 272-279. DOI: 10.1055/s-0033-1334955 39. Suryavanshi S.V., Kulkarni Y.. NF-kappabeta: A Potential Target in the Management of Vascular Complications of Diabetes. Front. Pharmacol. (2017) 8 798. DOI: 10.3389/fphar.2017.00798 40. Liu T., Zhang L., Joo D., Sun S.C.. NF-kappaB signaling in inflammation. Signal Transduct. Target. Ther. (2017) 2 17023. DOI: 10.1038/sigtrans.2017.23 41. Yamamoto Y., Gaynor R.. IkappaB kinases: Key regulators of the NF-kappaB pathway. Trends Biochem. Sci. (2004) 29 72-79. DOI: 10.1016/j.tibs.2003.12.003 42. Sheu M.L., Ho F.M., Yang R.S., Chao K.F., Lin W.W., Lin-Shiau S.Y., Liu S.H.. High glucose induces human endothelial cell apoptosis through a phosphoinositide 3-kinase-regulated cyclooxygenase-2 pathway. Arterioscler. Thromb. Vasc. Biol. (2005) 25 539-545. DOI: 10.1161/01.ATV.0000155462.24263.e4 43. Kim J.J., Lee S.B., Park J.K., Yoo Y.D.. TNF-alpha-induced ROS production triggering apoptosis is directly linked to Romo1 and Bcl-X(L). Cell Death Differ. (2010) 17 1420-1434. DOI: 10.1038/cdd.2010.19 44. Padgett L.E., Broniowska K.A., Hansen P.A., Corbett J.A., Tse H.M.. The role of reactive oxygen species and proinflammatory cytokines in type 1 diabetes pathogenesis. Ann. N. Y. Acad Sci. (2013) 1281 16-35. DOI: 10.1111/j.1749-6632.2012.06826.x 45. Tse H.M., Milton M., Piganelli J.. Mechanistic analysis of the immunomodulatory effects of a catalytic antioxidant on antigen-presenting cells: Implication for their use in targeting oxidation-reduction reactions in innate immunity. Free Radic. Biol. Med. (2004) 36 233-247. DOI: 10.1016/j.freeradbiomed.2003.10.029 46. Tse H.M., Milton M.J., Schreiner S., Profozich J.L., Trucco M., Piganelli J.D.. Disruption of innate-mediated proinflammatory cytokine and reactive oxygen species third signal leads to antigen-specific hyporesponsiveness. J. Immunol. (2007) 178 908-917. DOI: 10.4049/jimmunol.178.2.908 47. Altannavch T.S., Roubalova K., Kucera P., Andel M.. Effect of high glucose concentrations on expression of ELAM-1, VCAM-1 and ICAM-1 in HUVEC with and without cytokine activation. Physiol. Res. (2004) 53 77-82. DOI: 10.33549/physiolres.930343 48. Dang L., Seale J., Qu X.. High glucose-induced human umbilical vein endothelial cell hyperpermeability is dependent on protein kinase C activation and independent of the Ca2+-nitric oxide signalling pathway. Clin. Exp. Pharmacol. Physiol. (2005) 32 771-776. DOI: 10.1111/j.1440-1681.2005.04266.x 49. Suarez S., McCollum G.W., Bretz C.A., Yang R., Capozzi M.E., Penn J.S.. Modulation of VEGF-induced retinal vascular permeability by peroxisome proliferator-activated receptor-beta/delta. Investig. Ophthalmol. Vis. Sci. (2014) 55 8232-8240. DOI: 10.1167/iovs.14-14217 50. Aiello L.P., Bursell S.E., Clermont A., Duh E., Ishii H., Takagi C., Mori F., Ciulla T.A., Ways K., Jirousek M.. Vascular endothelial growth factor-induced retinal permeability is mediated by protein kinase C in vivo and suppressed by an orally effective beta-isoform-selective inhibitor. Diabetes (1997) 46 1473-1480. DOI: 10.2337/diab.46.9.1473 51. Carmeliet P., Jain R.. Angiogenesis in cancer and other diseases. Nature (2000) 407 249-257. DOI: 10.1038/35025220 52. Wang J., Xu X., Elliott M.H., Zhu M., Le Y.Z.. Muller cell-derived VEGF is essential for diabetes-induced retinal inflammation and vascular leakage. Diabetes (2010) 59 2297-2305. DOI: 10.2337/db09-1420 53. Khan M.A., Assiri A., Broering D.. Complement and macrophage crosstalk during process of angiogenesis in tumor progression. J. Biomed. Sci. (2015) 22 58. DOI: 10.1186/s12929-015-0151-1 54. Tan X., Fujiu K., Manabe I., Nishida J., Yamagishi R., Nagai R., Yanagi Y.. Choroidal neovascularization is inhibited via an intraocular decrease of inflammatory cells in mice lacking complement component C3. Sci. Rep. (2015) 5 15702. DOI: 10.1038/srep15702 55. Li L., Wei T., Liu S., Wang C., Zhao M., Feng Y., Ma L., Lu Y., Fu P., Liu J.. Complement C5 activation promotes type 2 diabetic kidney disease via activating STAT3 pathway and disrupting the gut-kidney axis. J. Cell. Mol. Med. (2021) 25 960-974. DOI: 10.1111/jcmm.16157 56. Zhao Y., Luo C., Chen J., Sun Y., Pu D., Lv A., Zhu S., Wu J., Wang M., Zhou J.. High glucose-induced complement component 3 up-regulation via RAGE-p38MAPK-NF-kappaB signalling in astrocytes: In vivo and in vitro studies. J. Cell. Mol. Med. (2018) 22 6087-6098. DOI: 10.1111/jcmm.13884 57. Sayed S., Faruq O., Hossain M., Im S.B., Kim Y.S., Lee B.T.. Thermal cycling effect on osteogenic differentiation of MC3T3-E1 cells loaded on 3D-porous Biphasic Calcium Phosphate (BCP) scaffolds for early osteogenesis. Mater. Sci. Eng. C Mater. Biol. Appl. (2019) 105 110027. DOI: 10.1016/j.msec.2019.110027 58. Faruq O., Sayed S., Kim B., Im S.B., Lee B.T.. A biphasic calcium phosphate ceramic scaffold loaded with oxidized cellulose nanofiber-gelatin hydrogel with immobilized simvastatin drug for osteogenic differentiation. J. Biomed. Mater. Res. B Appl. Biomater. (2020) 108 1229-1238. DOI: 10.1002/jbm.b.34471
--- title: 'Effects of l-Arginine Plus Vitamin C Supplementation on l-Arginine Metabolism in Adults with Long COVID: Secondary Analysis of a Randomized Clinical Trial' authors: - Riccardo Calvani - Jacopo Gervasoni - Anna Picca - Francesca Ciciarello - Vincenzo Galluzzo - Hélio José Coelho-Júnior - Clara Di Mario - Elisa Gremese - Sara Lomuscio - Anna Maria Paglionico - Lavinia Santucci - Barbara Tolusso - Andrea Urbani - Federico Marini - Emanuele Marzetti - Francesco Landi - Matteo Tosato journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049539 doi: 10.3390/ijms24065078 license: CC BY 4.0 --- # Effects of l-Arginine Plus Vitamin C Supplementation on l-Arginine Metabolism in Adults with Long COVID: Secondary Analysis of a Randomized Clinical Trial ## Abstract Altered l-arginine metabolism has been described in patients with COVID-19 and has been associated with immune and vascular dysfunction. In the present investigation, we determined the serum concentrations of l-arginine, citrulline, ornithine, monomethyl-l-arginine (MMA), and symmetric and asymmetric dimethylarginine (SDMA, ADMA) in adults with long COVID at baseline and after 28-days of l-arginine plus vitamin C or placebo supplementation enrolled in a randomized clinical trial, compared with a group of adults without previous history of SARS-CoV-2-infection. l-arginine-derived markers of nitric oxide (NO) bioavailability (i.e., l-arginine/ADMA, l-arginine/citrulline+ornithine, and l-arginine/ornithine) were also assayed. Partial least squares discriminant analysis (PLS–DA) models were built to characterize systemic l-arginine metabolism and assess the effects of the supplementation. PLS–DA allowed discrimination of participants with long COVID from healthy controls with 80.2 ± $3.0\%$ accuracy. Lower markers of NO bioavailability were found in participants with long COVID. After 28 days of l-arginine plus vitamin C supplementation, serum l-arginine concentrations and l-arginine/ADMA increased significantly compared with placebo. This supplement may therefore be proposed as a remedy to increase NO bioavailability in people with long COVID. ## 1. Introduction l-arginine metabolism is involved in the regulation of several biological processes, including immune and vascular function [1,2]. Two main metabolic pathways are associated with the pleiotropic activities of l-arginine: its conversion to nitric oxide (NO) by NO synthase (NOS) or l-arginine catabolism to ornithine by arginase [3,4]. NOS and arginase entertain reciprocal regulatory interactions that impact NO bioavailability [5]. NO is a master regulator of cardiovascular function, metabolism, neurotransmission, and immunity [6]. The flux of l-arginine towards NO synthesis is associated with beneficial effects on immune and vascular health [5]. On the other hand, upregulation of arginase inhibits NO production and promotes immune and endothelial dysfunction [4]. Several clinical conditions, including hypertension, diabetes, and inflammatory diseases, are characterized by the rewiring of l-arginine metabolism towards increased arginase activity [7,8,9]. In addition to the NOS/arginase dyad, some l-arginine derivatives can modulate NO bioavailability. Methylarginine moieties, including monomethyl-l-arginine (MMA) and symmetric and asymmetric dimethylarginine (SDMA and ADMA), are released following proteolysis of arginine-methylated proteins [10]. In vitro and in vivo data have shown that MMA, SDMA, and ADMA may inhibit NOS [11]. Indeed, elevated circulating levels of ADMA and SDMA have been identified as independent risk factors for cardiovascular events and all-cause mortality across different populations [12,13,14]. Perturbations in l-arginine metabolism have been described in patients with COVID-19 across all disease stages [5,15,16]. During acute COVID-19, higher arginase activity shifts l-arginine away from NO synthesis to induce immune dysregulation and endothelial dysfunction, which both increase the risk of thrombosis, arterial stiffening, and vascular occlusion [5]. A low l-arginine-to-ornithine ratio, indicative of upregulated arginase activity, has been found in patients with COVID-19 and children with multisystem inflammatory syndrome (MIS-C), and has been associated with reduced circulating levels of l-arginine compared with healthy controls [17]. Arginine shortage and enhanced arginase activity have also been associated with expansion of myeloid-derived suppressor cells, lymphopenia, and lymphocyte dysfunction in COVID-19 patients with severe acute respiratory distress syndrome (ARDS) [18,19]. Moreover, elevated serum levels of ADMA and SDMA at hospital admission were associated with disease severity [20] and predicted in-hospital mortality in patients with COVID-19 [21]. Eight months after acute infection, young and middle-aged COVID-19 survivors showed reduced serum levels of l-arginine compared with controls without evidence of previous SARS-CoV-2 infection [16]. Based on these observations, interventions targeting l-arginine metabolism have been proposed to increase NO bioavailability and contrast immune and vascular complications of COVID-19 [1,5,22]. In vitro, l-arginine supplementation restored the proliferative capacity of T-cells obtained from COVID-19 patients with ARDS [19]. Oral l-arginine supplementation reduced the need for oxygen therapy and the length of hospital stay in patients with severe COVID-19 [23]. The combination of l-arginine plus vitamin C, which may support NOS activity in endothelial cells [24], showed a synergistic antiviral action on SARS-CoV-2 in vitro by inhibiting its main protease Mpro [25]. A supplement containing l-arginine and vitamin C relieved the burden of persistent symptoms and improved perceived exertion in a large cohort of adults with long COVID [26]. The latter condition is diagnosed in individuals who, at three months of an acute symptomatic COVID-19 episode, have persistence of symptoms for at least two months, which cannot be explained by any other possible diagnosis [27]. Finally, we recently showed that a 28-day oral supplementation with l-arginine plus vitamin C restored circulating l-arginine levels and improved walking performance, muscle strength, endothelial function, and fatigue in adults with long COVID [16]. In the present investigation, we conducted secondary analyses of a randomized clinical trial that involved 28 days of supplementation with l-arginine plus vitamin C or placebo. Data from a group of adults without previous history of SARS-CoV-2-infection were also analyzed. For this study, we assayed a comprehensive panel of metabolites pertaining to l-arginine metabolism to obtain further insights into systemic arginine metabolism and NO bioavailability in adults with long COVID. We also assessed the effects of l-arginine plus vitamin C supplementation on l-arginine metabolites. ## 2.1. Characteristics of Study Population and Concentrations of l-Arginine Metabolites at Baseline Fifty-seven participants were included in the present investigation: 46 adults with long COVID (47.8 ± 9.4 years; $65.2\%$ women) enrolled in a randomized clinical trial [16] and a group of 11 age- and sex-matched controls without evidence of previous SARS-CoV-2 infection (48.8 ± 11.1 years; $55.5\%$ women). As previously reported [16], approximately half of those with long COVID needed hospitalization during the acute COVID-19 episode, and four ($8.7\%$) were admitted to an intensive care unit. The average time elapsed from COVID-19 diagnosis to the inclusion in the study was 252.6 ± 113.7 days. Participants with long COVID had been randomized to receive either the l-arginine plus vitamin C ($$n = 23$$) or the placebo ($$n = 23$$) intervention for 28 days [16]. Demographic and anthropometric characteristics of study participants, standard blood biochemistry, as well as concentrations of l-arginine metabolites at baseline are reported in Table 1. ## 2.2. l-Arginine Metabolism in Participants with Long COVID and Controls To gain insights into possible differences in l-arginine metabolism between participants with long COVID and healthy controls, baseline serum concentrations of l-arginine-related analytes were processed through partial least squares discriminant analysis (PLS–DA). To obtain an unbiased validation of the results and estimate the confidence intervals (CIs) of the main figures of merit, repeated double cross-validation (rDCV) was conducted with 10 and eight cancelation groups in the outer and inner loops, respectively, and permutation tests with 1000 randomizations. The optimal model complexity was found to be 3 ± 1 latent variables and yielded an average classification accuracy of 80.2 ± $3.0\%$, corresponding to 78.1 ± $3.2\%$ and 88.9 ± $5.2\%$ correct classification rate for participants with long COVID and controls, respectively. The non-parametric estimation of the distribution of these figures of merit under the null hypothesis by permutation testing indicated that they were statistically significant ($p \leq 0.001$). A graphical representation of the discriminant ability of the model is depicted in Figure 1A. The figure shows the rDCV outer loop sample scores and the variable weights along the only canonical variate of the model. Inspection of the contribution of individual l-arginine metabolites to the classification model indicated that five out of the six measured analytes contributed significantly to the discrimination: ADMA, MMA, SDMA, citrulline (on average higher in participants with long COVID), and l-arginine (on average higher in controls) (Figure 1B). According to reference values proposed by major clinical medical laboratories [28,29,30], serum levels of ADMA and SDMA in participants with long COVID may be suggestive of endothelial dysfunction and increased cardiovascular risk. In participants with long COVID, lower circulating levels of l-arginine led to reduced l-arginine/ADMA (mean difference: −152.7, $95\%$ CI: −196.58 to −106.66; effect size = 0.81; $p \leq 0.0001$), lower global arginine bioavailability ratio (GABR) (mean difference: −0.73, $95\%$ CI: −1.03 to −0.45; effect size = 0.82; $p \leq 0.0001$), and lower l-arginine-to-ornithine ratio (mean difference: −0.99, $95\%$ CI: −1.45 to −0.57; effect size = 0.76; $p \leq 0.0001$) than healthy controls. These findings indicate that l-arginine bioavailability is reduced and, therein, NO biosynthetic capacity may be impaired in participants with long COVID compared with those without evidence of previous SARS-CoV-2 infection. ## 2.3. Effects of l-Arginine Plus Vitamin C Supplementation on l-Arginine Metabolism in Participants with Long COVID A PLS–DA model was built to explore the effects of 28-day supplementation with l-arginine plus vitamin C or placebo on l-arginine metabolism in participants with long COVID. To account for the repeated measures design of the study, the classification model was built using, for each participant, the difference between values at 28 days and those at baseline (Table 2). The optimal model complexity was found to be 4 ± 1 latent variables. The model had low discrimination power, with an average classification accuracy of 58.5 ± $5.8\%$, corresponding to 54.3 ± $6.3\%$ and 62.6 ± $8.6\%$ of correct classification in participants who received l-arginine plus vitamin C and those in the placebo group, respectively (Figure 2A). As previously reported [16], l-arginine plus vitamin C supplementation induced a significantly greater increase in circulating l-arginine levels compared with placebo (mean difference: 62.4 µM, $95\%$ CI: 11.1 to 113.7 µM; effect size = 0.72; $$p \leq 0.02$$). This finding was confirmed by the inspection of the variable weights plot of the PLS–DA model (Figure 2B). Although the increase in serum l-arginine concentrations observed in the active treatment group did not result in significant changes in GABR or l-arginine-to-ornithine ratio, the difference in l-arginine/ADMA between intervention groups was close to statistical significance ($$p \leq 0.05$$; Table 2). To assess whether l-arginine plus vitamin C supplementation induced changes in l-arginine metabolism towards healthy reference values, two PLS–DA models were built, comparing data at 28 days from the active treatment group and placebo versus healthy controls (Figure 3). Both PLS–DA models had a high mean classification accuracy (85.2 ± $2.6\%$ and 84.5 ± $3.1\%$, respectively) and indicated persistent disruption of l-arginine metabolism in participants with long COVID, regardless of treatment allocation (Figure 3A,B). However, following l-arginine plus vitamin C supplementation, l-arginine and citrulline no longer contributed to the discrimination between participants with long COVID and healthy controls (Figure 3C,D). In addition, after 28 days of l-arginine plus vitamin C supplementation, mean l-arginine/ADMA values were significantly different from both those in placebo-treated participants ($$p \leq 0.03$$) and healthy controls ($$p \leq 0.01$$), with a shift towards healthy reference values (Figure 4). ## 3. Discussion In the present investigation, we showed that l-arginine metabolism was altered in a group of adults with long COVID eight months after the diagnosis of COVID-19 compared with age- and sex-matched adults without history of SARS-CoV-2 infection. l-arginine metabolism was still disrupted after 28-day supplementation with l-arginine plus vitamin C. However, in those who received the active intervention, both serum l-arginine concentrations and l-arginine/ADMA, a marker of NO biosynthetic capacity, significantly shifted towards healthy reference values compared with participants who were allocated to placebo. Accumulating evidence indicates that arginine metabolism is altered in COVID-19 patients [17,19,31,32,33]. In particular, low circulating l-arginine levels and upregulated arginase activity have been associated with reduced NO bioavailability and immune and vascular dysfunction in acute COVID-19 [15]. Low arginine-to-ornithine ratio and low GABR, as well as two-fold increase in circulating levels of ADMA, were found in severely ill COVID-19 patients [34]. This suggests that SARS-CoV-2 infection may induce endothelial dysfunction and a pro-thrombotic vascular phenotype acting both on NOS substrate availability and enzyme activity. Our findings show that similar perturbations in l-arginine metabolism may be found in adults with long COVID several months after the acute episode. Notably, low GABR was associated with the development of coronary artery disease and increased risk of major adverse cardiovascular events over a 3-year follow-up in a cohort of 1010 patients undergoing elective cardiac catheterization [35]. In addition, reduced GABR and l-arginine-to-ornithine ratio were associated with markers of endothelial dysfunction and increased risk of cardiovascular mortality in patients referred for coronary angiography [36]. Increased circulating levels of ADMA were found in conditions characterized by impaired NO synthesis and endothelial dysfunction, such as hypertension, diabetes, atherosclerosis, and cerebrovascular diseases [37,38,39,40,41,42,43]. Elevated ADMA levels increase the risk of recurrent cardiovascular events or death in patients with a history of acute coronary disease [44,45], unstable angina [46], or diabetes [47]. Low l-arginine/ADMA is an independent risk factor for atherosclerosis [48] and microangiopathy-related cerebral damage [49], and has shown to be a better predictor of all-cause mortality than ADMA alone [48,50]. In this scenario, the reduced indices of l-arginine/NO bioavailability found in adults with long COVID-19 in the present investigation suggest a role for altered l-arginine metabolism in increasing the risk of endotheliopathy and long-term cardiovascular events [51,52,53,54]. l-arginine plus vitamin C supplementation increased circulating levels of l-arginine and shifted l-arginine/ADMA values towards healthy reference. Owing to its arginine-like structure, ADMA may directly compete with l-arginine both for its transport into the cell via the cationic amino acid transporter and NOS binding [55,56]. It follows that NO bioavailability may be influenced by the balance between l-arginine and ADMA [1]. Low l-arginine/ADMA results in a net inhibition of NO production [57]. Oral l-arginine supplementation may re-equilibrate l-arginine/ADMA, increase NO synthesis, and improve endothelial function [1,58]. Our previous findings corroborate this hypothesis, since in adults with long COVID supplemented with l-arginine plus vitamin C, the increase in circulating l-arginine concentrations and l-arginine/ADMA was associated with a significant improvement in flow-mediated dilation (FMD), a measure of NO-dependent endothelial reactivity, compared with placebo [16]. These results are in line with those from a meta-analysis of randomized clinical trials showing that short-term oral l-arginine supplementation improved endothelial function in individuals with reduced FMD [59]. In this context, l-arginine supplementation may be particularly suited for people with ascertained endothelial dysfunction and low l-arginine/ADMA, such as those with long COVID, since l-arginine supplementation in individuals with high FMD and low ADMA levels (or normal l-arginine/ADMA ratio) failed to improve either NO bioavailability or endothelial function [59,60,61]. Some limitations should be considered in the interpretation of the study results. Due to the small number of participants and the single-center nature of the study, our results should be considered preliminary. Further investigation with larger populations, conducted in multiple centers, and using different study methodologies (e.g., longer intervention, crossover design) is warranted to confirm our findings. The levels of physical activity as well as dietary habits of study participants may have influenced the concentration of l-arginine metabolites and the effects of interventions. However, participants were requested to refrain from exercising, limit the ingestion of foods rich in arginine, and taking substances with vasoactive properties for at least 12 h before study visits. Due to the heterogeneity of data on vaccination status (e.g., timing, types of vaccine, number of doses, refusal to disclose vaccination status), this information was not accounted for in the analyses. The panel of metabolites assessed in the present investigation provided relevant information on differences in l-arginine metabolism between adults with long COVID and healthy controls and allowed for evaluating the effectiveness of the tested intervention. However, we cannot exclude that a more comprehensive evaluation of NO metabolism (e.g., measurement of circulating levels of nitrite, nitrate, and NO derivatives), as well as the assessment of inflammatory, vascular, or neurological markers may provide further insights into the mechanisms by which l-arginine plus vitamin C supplementation affects the outcomes of interest. Vitamin C levels were not quantified; thus, the relationship between circulating vitamin C concentrations and study outcomes could not be explored. Because l-arginine levels were not measured days after the end of the intervention, it was not possible to appreciate the duration of the beneficial effects of l-arginine plus vitamin C supplementation on the parameters of interest. Finally, it cannot be ruled out that the co-administration of other nutraceuticals may convey additional beneficial effects on l-arginine metabolism and long COVID symptoms [22,62,63]. For instance, vitamin D may have positive effects on both NO synthesis and endothelial function [64]. Vitamin D deficiency is frequent in COVID-19 survivors and is associated with poor physical performance [65]. The combined use of l-arginine, coenzyme Q10, and vitamin D was found to reduce oxidative stress and stimulate NO synthesis in cardiac and endothelial cells to a greater extent than any of those compounds alone [66]. The combination has therefore been proposed as a cardiovascular protective remedy [66]. Further studies are needed to assess whether supplementation with different combinations of nutrients may be proposed as a remedy to restore l-arginine metabolism and limit post-acute COVID-19 sequelae. ## 4.1. Study Design and Participants Participants involved in the present investigation were adults with long COVID who were enrolled in a placebo-controlled randomized clinical trial that tested the effects of a combination of l-arginine plus vitamin C on physical performance, endothelial function, and persistent fatigue (NCT04947488) [16]. Eleven age- and sex-matched blood donors without evidence of previous SARS-CoV-2 infection were recruited and analyzed as a “healthy” reference. Trial operations were conducted at the post-acute COVID-19 outpatient clinic of the Fondazione Policlinico A. Gemelli IRCCS (Rome, Italy) from 1 July 2021 to 30 April 2022 [67]. Details on the clinical trial protocol and inclusion/exclusion criteria have been reported elsewhere [16]. Briefly, trial participants were men and women aged 20 to 60 years with a previous confirmed SARS-CoV-2 infection (certified by a positive RT–PCR molecular swab test), a negative COVID-19 test at least four weeks prior to enrolment, long COVID diagnosis according to national and international criteria [27,68], and persistent fatigue, defined as the response “most or all the time” to item seven of the Center for Epidemiological Studies Depression Scale (“I felt that everything I did was an effort”) [69]. The main exclusion criteria were: intolerance to preparations containing l-arginine or vitamin C, conditions and/or treatments that might affect trial outcomes or procedures (e.g., pregnancy or nursing, diabetes, and use of antihypertensive medications, corticosteroids, or non-steroidal anti-inflammatory drugs, immunosuppressants, nitrates), and participation in other long COVID intervention trials. Eligible participants were randomized to receive twice-daily an oral supplementation with either a combination of 1.66 g l-arginine plus 500 mg liposomal vitamin C (Bioarginina® C, Farmaceutici Damor, Naples, Italy) or placebo for 28 days. The dose was selected based on previous evidence of the beneficial effects of l-arginine supplementation during acute COVID-19 [23], and on the trial methodology followed by Rizzo et al. [ 26] to assess effectiveness of l-arginine plus vitamin C on improving long COVID symptoms. All trial procedures were conducted in accordance with the guidelines of the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use Good Clinical Practice and the principles of the Declaration of Helsinki. All participants provided written informed consent prior to enrolment. Trial participants and healthy controls were asked to refrain from exercising and consuming any product with vasoactive properties (e.g., tobacco, caffeinated drinks) for at least 12 h before blood drawing. ## 4.2. l-Arginine Metabolism Assessment Blood samples were collected after overnight fasting using standard collection tubes. Samples were left at room temperature for 30 min and were then centrifuged at 1000× g for 10 min at 4 °C. Serum aliquots were stored at −80 °C until analysis. Serum samples from participants with long COVID were collected at baseline and after 28 days of intervention. Blood samples from healthy controls were collected and processed according to the same protocol. The concentrations of l-arginine, citrulline, ornithine, ADMA, MMA, and SDMA were measured using an in-house validated liquid chromatography with tandem mass spectrometry method [70]. The chromatographic separation was performed with an ACQUITY UPLC I-Class System (Waters, Milford, MA, USA) using a HILIC column. Analyte detection was performed using a triple quadrupole Xevo-TQs Micro (Waters) equipped with an electrospray ion source operating in positive ion mode. A multiple reaction monitoring experiment was optimized for the detection and quantification of l-arginine and its metabolites. l-arginine-derived indexes associated with endothelial and immune dysfunction, such as l-arginine/ADMA, GABR (l-arginine/ornithine+citrulline), and l-arginine-to-ornithine ratio, were assessed in all study participants across the different timepoints [17,35,48]. ## 4.3. Statistical Analysis Personal characteristics of study participants are reported as mean ± standard deviation or median (interquartile range) for continuous variables, and as absolute values (percentages) for categorical variables. Normal distribution of data was assessed via the Shapiro–Wilk test. Changes from baseline for continuous variables are expressed as deltas (i.e., values at 28 days minus values at baseline) and differences between groups were evaluated using Student’s t-test for normally distributed variables or Mann–Whitney U test for skewed variables. Mean differences and effect size values (Cohen’s d for Student’s t-test and rank biserial correlation for Mann–Whitney U) were reported. One-way analysis of variance and post hoc tests were used to compare mean concentration values of l-arginine metabolites between participants with long COVID and healthy controls. All tests were two-sided with statistical significance set at $p \leq 0.05.$ All analyses were performed using Jamovi freeware version 2.0.0.0 (The Jamovi project, 2021; https://www.jamovi.org, accessed on 27 February 2023). Multivariate classification models, based on PLS–DA [71], were built to gain a more comprehensive insight into l-arginine metabolism during post-acute COVID-19 and to assess the effects of l-arginine plus vitamin C supplementation in adults with long COVID. PLS–DA is a classification method that exploits the advantages of the PLS algorithm for dealing with correlated variables. The PLS algorithm was originally developed for regression problems and relies on the projection of the predictor matrix, X, onto a reduced space of orthogonal latent variables, yielding a matrix of scores, T (coordinates of the samples onto the latent variables subspace):T = XR [1] R being a matrix of weights determining the projection. A regression model is then established between the scores T and the response, y, to be predicted: y = Tq [2] q being the regression coefficients. The same approach can be used for classification by using a dummy binary y coding for class belonging: the elements of y can be either 1 if the sample belongs to a category or 0 if it belongs to the other category. First, classification models were built to evaluate differences in l-arginine metabolism between long COVID participants ($y = 1$) and healthy controls ($y = 0$). Then, the effects of l-arginine plus vitamin C supplementation on systemic l-arginine metabolism were tested building models to discriminate between active treatment ($y = 1$) and placebo ($y = 0$). Finally, PLS–DA models were built to explore whether l-arginine supplementation could revert l-arginine metabolic profiles of long COVID participants towards the healthy reference status (using either l-arginine plus vitamin C or placebo-treated participants as one category ($y = 1$) and healthy controls as the other ($y = 0$)). Model validation was achieved through rDCV [72]. rDCV consists of two loops of cross-validation nested into one another: the outer loop mimics an external test set, while the inner loop is used for model selection (i.e., choosing the optimal number of latent variables). The procedure was repeated 50 times, changing the distribution of samples in the different cancelation groups, which allowed CIs to be calculated for all model parameters and figures of merit. To account for the repeated-measure design of the intervention study, the classification models were built using, for each participant, the difference between values at 28 days and values at baseline. Analyses were performed using in-house routines running under MATLAB R2015b environment (The MathWorks, Natick, MA, USA). ## 5. Conclusions In the present investigation, we showed that perturbations in l-arginine metabolism indicative of reduced l-arginine/NO bioavailability were found in adults with long COVID at eight months from acute disease compared with controls without previous history of SARS-CoV-2 infection. After 28-day supplementation with l-arginine plus vitamin C, serum l-arginine levels and l-arginine/ADMA ratio, a marker of NO biosynthetic capacity, increased relative to placebo. Given the preliminary nature of our findings, further studies are needed to conclusively establish whether l-arginine plus vitamin C supplementation restores l-arginine metabolism in adults with long COVID. ## References 1. Gambardella J., Khondkar W., Morelli M.B., Wang X., Santulli G., Trimarco V.. Arginine and endothelial function. Biomedicines (2020.0) 8. DOI: 10.3390/biomedicines8080277 2. Bronte V., Zanovello P.. Regulation of immune responses by L-arginine metabolism. Nat. Rev. Immunol. (2005.0) 5 641-654. DOI: 10.1038/nri1668 3. Wu G., Morris S.M.. Arginine metabolism: Nitric oxide and beyond. Biochem. J. (1998.0) 336 1-17. DOI: 10.1042/bj3360001 4. Durante W., Johnson F.K., Johnson R.A.. Arginase: A critical regulator of nitric oxide synthesis and vascular function. Clin. Exp. Pharmacol. Physiol. (2007.0) 34 906-911. DOI: 10.1111/j.1440-1681.2007.04638.x 5. Durante W.. Targeting Arginine in COVID-19-induced immunopathology and vasculopathy. Metabolites (2022.0) 12. DOI: 10.3390/metabo12030240 6. Lundberg J.O., Weitzberg E.. Nitric oxide signaling in health and disease. Cell (2022.0) 185 2853-2878. DOI: 10.1016/j.cell.2022.06.010 7. Zhang C., Hein T.W., Wang W., Miller M.W., Fossum T.W., McDonald M.M., Humphrey J.D., Kuo L.. Upregulation of vascular arginase in hypertension decreases nitric oxide-mediated dilation of coronary arterioles. Hypertension (2004.0) 44 935-943. DOI: 10.1161/01.HYP.0000146907.82869.f2 8. Pernow J., Jung C.. The Emerging role of arginase in endothelial dysfunction in diabetes. Curr. Vasc. Pharmacol. (2016.0) 14 155-162. DOI: 10.2174/1570161114666151202205617 9. Rouzaut A., Subirá M.L., De Miguel C., Domingo-De-Miguel E., González A., Santiago E., López-Moratalla N.. Co-expression of inducible nitric oxide synthase and arginases in different human monocyte subsets. Apoptosis regulated by endogenous NO. Biochim. Biophys. Acta (1999.0) 1451 319-333. DOI: 10.1016/S0167-4889(99)00106-8 10. Blanc R.S., Richard S.. Arginine methylation: The coming of age. Mol. Cell (2017.0) 65 8-24. DOI: 10.1016/j.molcel.2016.11.003 11. Tsikas D., Bollenbach A., Hanff E., Kayacelebi A.A.. Asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA) and homoarginine (hArg): The ADMA, SDMA and hArg paradoxes. Cardiovasc. Diabetol. (2018.0) 17 1. DOI: 10.1186/s12933-017-0656-x 12. Schlesinger S., Sonntag S.R., Lieb W., Maas R.. Asymmetric and symmetric dimethylarginine as risk markers for total mortality and cardiovascular outcomes: A systematic review and meta-analysis of prospective studies. PLoS ONE (2016.0) 11. DOI: 10.1371/journal.pone.0165811 13. Willeit P., Freitag D.F., Laukkanen J.A., Chowdhury S., Gobin R., Mayr M., Di Angelantonio E., Chowdhury R.. Asymmetric dimethylarginine and cardiovascular risk: Systematic review and meta-analysis of 22 prospective studies. J. Am. Heart Assoc. (2015.0) 4 e001833. DOI: 10.1161/JAHA.115.001833 14. Zhou S., Zhu Q., Li X., Chen C., Liu J., Ye Y., Ruan Y., Hei Z.. Asymmetric dimethylarginine and all-cause mortality: A systematic review and meta-analysis. Sci. Rep. (2017.0) 7 44692. DOI: 10.1038/srep44692 15. Adebayo A., Varzideh F., Wilson S., Gambardella J., Eacobacci M., Jankauskas S.S., Donkor K., Kansakar U., Trimarco V., Mone P.. Nutrients (2021.0) 13. DOI: 10.3390/nu13113951 16. Tosato M., Calvani R., Picca A., Ciciarello F., Galluzzo V., Coelho-Júnior H.J., Di Giorgio A., Di Mario C., Gervasoni J., Gremese E.. Effects of. Nutrients (2022.0) 14. DOI: 10.3390/nu14234984 17. Rees C.A., Rostad C.A., Mantus G., Anderson E.J., Chahroudi A., Jaggi P., Wrammert J., Ochoa J.B., Ochoa A., Basu R.K.. Altered amino acid profile in patients with SARS-CoV-2 infection. Proc. Natl. Acad. Sci. USA (2021.0) 118 e2101708118. DOI: 10.1073/pnas.2101708118 18. Sacchi A., Grassi G., Notari S., Gili S., Bordoni V., Tartaglia E., Casetti R., Cimini E., Mariotti D., Garotto G.. Expansion of myeloid derived suppressor cells contributes to platelet activation by L-arginine deprivation during SARS-CoV-2 infection. Cells (2021.0) 10. DOI: 10.3390/cells10082111 19. Reizine F., Lesouhaitier M., Gregoire M., Pinceaux K., Gacouin A., Maamar A., Painvin B., Camus C., Le Tulzo Y., Tattevin P.. SARS-CoV-2-induced ARDS associates with MDSC expansion, lymphocyte dysfunction, and arginine shortage. J. Clin. Immunol. (2021.0) 41 515-525. DOI: 10.1007/s10875-020-00920-5 20. Karu N., Kindt A., van Gammeren A.J., Ermens A.A.M., Harms A.C., Portengen L., Vermeulen R.C.H., Dik W.A., Langerak A.W., van der Velden V.H.J.. Severe COVID-19 is characterised by perturbations in plasma amines correlated with immune response markers, and linked to inflammation and oxidative stress. Metabolites (2022.0) 12. DOI: 10.3390/metabo12070618 21. Hannemann J., Balfanz P., Schwedhelm E., Hartmann B., Ule J., Müller-Wieland D., Dahl E., Dreher M., Marx N., Böger R.. Elevated serum SDMA and ADMA at hospital admission predict in-hospital mortality of COVID-19 patients. Sci. Rep. (2021.0) 11 9895. DOI: 10.1038/s41598-021-89180-w 22. Tosato M., Ciciarello F., Zazzara M.B., Pais C., Savera G., Picca A., Galluzzo V., Coelho-Júnior H.J., Calvani R., Marzetti E.. Nutraceuticals and dietary supplements for older adults with long COVID-19. Clin. Geriatr. Med. (2022.0) 38 565-591. DOI: 10.1016/j.cger.2022.04.004 23. Fiorentino G., Coppola A., Izzo R., Annunziata A., Bernardo M., Lombardi A., Trimarco V., Santulli G., Trimarco B.. Effects of adding L-arginine orally to standard therapy in patients with COVID-19: A randomized, double-blind, placebo-controlled, parallel-group trial. Results of the first interim analysis. EClinicalMedicine (2021.0) 40 101125. DOI: 10.1016/j.eclinm.2021.101125 24. Morelli M.B., Gambardella J., Castellanos V., Trimarco V., Santulli G.. Vitamin C and cardiovascular disease: An update. Antioxidants (2020.0) 9. DOI: 10.3390/antiox9121227 25. Đukić I., Kaličanin N., Sencanski M., Pajovic S.B., Milicevic J., Prljic J., Paessler S., Prodanović R., Glisic S.. Inhibition of SARS-CoV-2 Mpro with vitamin C, L-arginine and a aitamin C/L-arginine combination. Front. Biosci. Landmark Ed. (2023.0) 28 8. DOI: 10.31083/j.fbl2801008 26. Izzo R., Trimarco V., Mone P., Aloè T., Capra Marzani M., Diana A., Fazio G., Mallardo M., Maniscalco M., Marazzi G.. Combining L-arginine with vitamin C improves long-COVID symptoms: The LINCOLN Survey. Pharmacol. Res. (2022.0) 183 106360. DOI: 10.1016/j.phrs.2022.106360 27. A Clinical Case Definition of Post COVID-19 Condition by a Delphi Consensus. (2021.0) 28. ADMA—Overview: Asymmetric Dimethylarginine, Plasma 29. ADMA/SDMA\|Test Detail\|Quest Diagnostics 30. ADMA/SDMA 31. Dean M.J., Ochoa J.B., Sanchez-Pino M.D., Zabaleta J., Garai J., Del Valle L., Wyczechowska D., Baiamonte L.B., Philbrook P., Majumder R.. Severe COVID-19 is characterized by an impaired type I interferon response and elevated levels of arginase producing granulocytic myeloid derived suppressor cells. Front. Immunol. (2021.0) 12 695972. DOI: 10.3389/fimmu.2021.695972 32. Wu P., Chen D., Ding W., Wu P., Hou H., Bai Y., Zhou Y., Li K., Xiang S., Liu P.. The trans-omics landscape of COVID-19. Nat. Commun. (2021.0) 12 4543. DOI: 10.1038/s41467-021-24482-1 33. Al-Kuraishy H.M., Al-Gareeb A.I., Alexiou A., Batiha G.E.-S.. COVID-19 and L-arginine supplementations: Yet to find the missed key. Curr. Protein Pept. Sci. (2022.0) 23 166-169. DOI: 10.2174/1389203723666220512104039 34. Canzano P., Brambilla M., Porro B., Cosentino N., Tortorici E., Vicini S., Poggio P., Cascella A., Pengo M.F., Veglia F.. Platelet and endothelial activation as potential mechanisms behind the thrombotic complications of COVID-19 patients. JACC Basic Transl. Sci. (2021.0) 6 202-218. DOI: 10.1016/j.jacbts.2020.12.009 35. Tang W.H.W., Wang Z., Cho L., Brennan D.M., Hazen S.L.. Diminished global arginine bioavailability and increased arginine catabolism as metabolic profile of increased cardiovascular risk. J. Am. Coll. Cardiol. (2009.0) 53 2061-2067. DOI: 10.1016/j.jacc.2009.02.036 36. Sourij H., Meinitzer A., Pilz S., Grammer T.B., Winkelmann B.R., Boehm B.O., März W.. Arginine bioavailability ratios are associated with cardiovascular mortality in patients referred to coronary angiography. Atherosclerosis (2011.0) 218 220-225. DOI: 10.1016/j.atherosclerosis.2011.04.041 37. Surdacki A., Nowicki M., Sandmann J., Tsikas D., Boeger R.H., Bode-Boeger S.M., Kruszelnicka-Kwiatkowska O., Kokot F., Dubiel J.S., Froelich J.C.. Reduced urinary excretion of nitric oxide metabolites and increased plasma levels of asymmetric dimethylarginine in men with essential hypertension. J. Cardiovasc. Pharmacol. (1999.0) 33 652-658. DOI: 10.1097/00005344-199904000-00020 38. Gervasoni J., Bonelli F., Zuppi C., Zappacosta B., Mordente A., Calvani R., Persichilli S.. Determination of asymmetric dimethyl arginine in human serum by liquid chromatography-tandem mass spectrometry: Clinical application in hypertensive subjects. Clin. Chem. Lab. Med. (2011.0) 49 2109-2115. DOI: 10.1515/CCLM.2011.691 39. Krzyzanowska K., Mittermayer F., Wolzt M., Schernthaner G.. Asymmetric dimethylarginine predicts cardiovascular events in patients with type 2 diabetes. Diabetes Care (2007.0) 30 1834-1839. DOI: 10.2337/dc07-0019 40. Chirinos J.A., David R., Bralley J.A., Zea-Díaz H., Muñoz-Atahualpa E., Corrales-Medina F., Cuba-Bustinza C., Chirinos-Pacheco J., Medina-Lezama J.. Endogenous nitric oxide synthase inhibitors, arterial hemodynamics, and subclinical vascular disease: The PREVENCION Study. Hypertension (2008.0) 52 1051-1059. DOI: 10.1161/HYPERTENSIONAHA.108.120352 41. Maas R., Xanthakis V., Polak J.F., Schwedhelm E., Sullivan L.M., Benndorf R., Schulze F., Vasan R.S., Wolf P.A., Böger R.H.. Association of the endogenous nitric oxide synthase inhibitor ADMA with carotid artery intimal media thickness in the Framingham Heart Study offspring cohort. Stroke (2009.0) 40 2715-2719. DOI: 10.1161/STROKEAHA.109.552539 42. Janes F., Cifù A., Pessa M.E., Domenis R., Gigli G.L., Sanvilli N., Nilo A., Garbo R., Curcio F., Giacomello R.. ADMA as a possible marker of endothelial damage. A study in young asymptomatic patients with cerebral small vessel disease. Sci. Rep. (2019.0) 9 14207. DOI: 10.1038/s41598-019-50778-w 43. Grosse G.M., Schwedhelm E., Worthmann H., Choe C.U.. Arginine derivatives in cerebrovascular diseases: Mechanisms and clinical implications. Int. J. Mol. Sci. (2020.0) 21. DOI: 10.3390/ijms21051798 44. Cavusoglu E., Ruwende C., Chopra V., Yanamadala S., Eng C., Pinsky D.J., Marmur J.D.. Relationship of baseline plasma ADMA levels to cardiovascular outcomes at 2 years in men with acute coronary syndrome referred for coronary angiography. Coron. Artery Dis. (2009.0) 20 112-117. DOI: 10.1097/MCA.0b013e328323982f 45. Valkonen V.P., Päivä H., Salonen J.T., Lakka T.A., Lehtimäki T., Laakso J., Laaksonen R.. Risk of acute coronary events and serum concentration of asymmetrical dimethylarginine. Lancet (2001.0) 358 2127-2128. DOI: 10.1016/S0140-6736(01)07184-7 46. Krempl T.K., Maas R., Sydow K., Meinertz T., Böger R.H., Kähler J.. Elevation of asymmetric dimethylarginine in patients with unstable angina and recurrent cardiovascular events. Eur. Heart J. (2005.0) 26 1846-1851. DOI: 10.1093/eurheartj/ehi287 47. Cavusoglu E., Ruwende C., Chopra V., Poludasu S., Yanamadala S., Frishman W.H., Eng C., Pinsky D.J., Marmur J.D.. Relation of baseline plasma ADMA levels to cardiovascular morbidity and mortality at two years in men with diabetes mellitus referred for coronary angiography. Atherosclerosis (2010.0) 210 226-231. DOI: 10.1016/j.atherosclerosis.2009.10.034 48. Notsu Y., Yano S., Shibata H., Nagai A., Nabika T.. Plasma arginine/ADMA ratio as a sensitive risk marker for atherosclerosis: Shimane CoHRE study. Atherosclerosis (2015.0) 239 61-66. DOI: 10.1016/j.atherosclerosis.2014.12.030 49. Notsu Y., Nabika T., Bokura H., Suyama Y., Kobayashi S., Yamaguchi S., Masuda J.. Evaluation of asymmetric dimethylarginine and homocysteine in microangiopathy-related cerebral damage. Am. J. Hypertens. (2009.0) 22 257-262. DOI: 10.1038/ajh.2008.346 50. Böger R.H., Endres H.G., Schwedhelm E., Darius H., Atzler D., Lüneburg N., von Stritzky B., Maas R., Thiem U., Benndorf R.A.. Asymmetric dimethylarginine as an independent risk marker for mortality in ambulatory patients with peripheral arterial disease. J. Intern. Med. (2010.0) 269 349-361. DOI: 10.1111/j.1365-2796.2010.02322.x 51. Ahamed J., Laurence J.. Long COVID endotheliopathy: Hypothesized mechanisms and potential therapeutic approaches. J. Clin. Investig. (2022.0) 132 e161167. DOI: 10.1172/JCI161167 52. Wang W., Wang C.Y., Wang S.I., Wei J.C.C.. Long-term cardiovascular outcomes in COVID-19 survivors among non-vaccinated population: A retrospective cohort study from the TriNetX US collaborative networks. EClinicalMedicine (2022.0) 53 101619. DOI: 10.1016/j.eclinm.2022.101619 53. Xie Y., Xu E., Bowe B., Al-Aly Z.. Long-term cardiovascular outcomes of COVID-19. Nat. Med. (2022.0) 28 583-590. DOI: 10.1038/s41591-022-01689-3 54. Davis H.E., McCorkell L., Vogel J.M., Topol E.J.. Long COVID: Major findings, mechanisms and recommendations. Nat. Rev. Microbiol. (2023.0) 21 133-146. DOI: 10.1038/s41579-022-00846-2 55. Strobel J., Müller F., Zolk O., Endreß B., König J., Fromm M.F., Maas R.. Transport of asymmetric dimethylarginine (ADMA) by cationic amino acid transporter 2 (CAT2), organic cation transporter 2 (OCT2) and multidrug and toxin extrusion protein 1 (MATE1). Amino Acids (2013.0) 45 989-1002. DOI: 10.1007/s00726-013-1556-3 56. Vallance P., Leone A., Calver A., Collier J., Moncada S.. Endogenous dimethylarginine as an inhibitor of nitric oxide synthesis. J. Cardiovasc. Pharmacol. (1992.0) 20 S60-S62. DOI: 10.1097/00005344-199204002-00018 57. Hsu C.P., Zhao J.F., Lin S.J., Shyue S.K., Guo B.C., Lu T.M., Lee T.S.. Asymmetric dimethylarginine limits the efficacy of simvastatin activating endothelial nitric oxide synthase. J. Am. Heart Assoc. (2016.0) 5 e003327. DOI: 10.1161/JAHA.116.003327 58. Bode-Böger S.M., Scalera F., Ignarro L.J.. The L-arginine paradox: Importance of the L-arginine/asymmetrical dimethylarginine ratio. Pharmacol. Ther. (2007.0) 114 295-306. DOI: 10.1016/j.pharmthera.2007.03.002 59. Bai Y., Sun L., Yang T., Sun K., Chen J., Hui R.. Increase in fasting vascular endothelial function after short-term oral L-arginine is effective when baseline flow-mediated dilation is low: A meta-analysis of randomized controlled trials. Am. J. Clin. Nutr. (2009.0) 89 77-84. DOI: 10.3945/ajcn.2008.26544 60. Böger G.I., Rudolph T.K., Maas R., Schwedhelm E., Dumbadze E., Bierend A., Benndorf R.A., Böger R.H.. Asymmetric dimethylarginine determines the improvement of endothelium-dependent vasodilation by simvastatin: Effect of combination with oral L-arginine. J. Am. Coll. Cardiol. (2007.0) 49 2274-2282. DOI: 10.1016/j.jacc.2007.02.051 61. Walker H.A., McGing E., Fisher I., Böger R.H., Bode-Böger S.M., Jackson G., Ritter J.M., Chowienczyk P.J.. Endothelium-dependent vasodilation is independent of the plasma L-arginine/ADMA ratio in men with stable angina: Lack of effect of oral L-arginine on endothelial function, oxidative stress and exercise performance. J. Am. Coll. Cardiol. (2001.0) 38 499-505. DOI: 10.1016/S0735-1097(01)01380-8 62. Speakman L.L., Michienzi S.M., Badowski M.E.. Vitamins, supplements and COVID-19: A review of currently available evidence. Drugs Context (2021.0) 10 2021-6-2. DOI: 10.7573/dic.2021-6-2 63. Veljković M., Pavlović D.R., Stojanović N.M., Džopalić T., Popović Dragonjić L.. Behavioral and dietary habits that could influence both COVID-19 and non-communicable civilization disease prevention-what have we learned up to now?. Medicina (2022.0) 58. DOI: 10.3390/medicina58111686 64. Kim D.H., Meza C.A., Clarke H., Kim J.S., Hickner R.C.. Vitamin D and endothelial function. Nutrients (2020.0) 12. DOI: 10.3390/nu12020575 65. Galluzzo V., Ciciarello F., Tosato M., Zazzara M.B., Pais C., Savera G., Calvani R., Picca A., Marzetti E., Landi F.. Association between vitamin D status and physical performance in COVID-19 survivors: Results from the Gemelli against COVID-19 Post-Acute Care project. Mech. Ageing Dev. (2022.0) 205 111684. DOI: 10.1016/j.mad.2022.111684 66. Molinari C., Morsanuto V., Polli S., Uberti F.. Cooperative effects of Q10, vitamin D3, and L-arginine on cardiac and endothelial cells. J. Vasc. Res. (2018.0) 55 47-60. DOI: 10.1159/000484928 67. Landi F., Gremese E., Bernabei R., Fantoni M., Gasbarrini A., Settanni C.R., Benvenuto F., Bramato G., Carfì A., Ciciarello F.. Post-COVID-19 global health strategies: The need for an interdisciplinary approach. Aging Clin. Exp. Res. (2020.0) 32 1613-1620. DOI: 10.1007/s40520-020-01616-x 68. Indicazioni ad Interim sui Principi di Gestione del Long-COVID, Versione del 1° Luglio 2021 69. Radloff L.S.. The CES-D scale: A self-report depression scale for research in the general population. Appl. Psychol. Meas. (1977.0) 1 385-401. DOI: 10.1177/014662167700100306 70. Santucci L., Lomuscio S., Canu F., Primiano A., Persichilli S., Urbani A., Gervasoni J.. A rapid method for determination of underivatized arginine-related metabolites in human plasma using LC−MS/MS. Proceedings of the 54° National Conference of Società Italiana di Biochimica Clinica e Biologia Molecolare Clinica (SIBioC) Volume 46 71. Ståhle L., Wold S.. Partial least squares analysis with cross-validation for the two-class problem: A Monte Carlo study. J. Chemom. (1987.0) 1 185-196. DOI: 10.1002/cem.1180010306 72. Szymańska E., Saccenti E., Smilde A.K., Westerhuis J.A.. Double-check: Validation of diagnostic statistics for PLS-DA models in metabolomics studies. Metabolomics (2012.0) 8 3-16. DOI: 10.1007/s11306-011-0330-3
--- title: 'A Decision Aid for Postpartum Adolescent Family Planning: A Quasi-Experimental Study in Tanzania' authors: - Stella E. Mushy - Shigeko Horiuchi - Eri Shishido journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049540 doi: 10.3390/ijerph20064904 license: CC BY 4.0 --- # A Decision Aid for Postpartum Adolescent Family Planning: A Quasi-Experimental Study in Tanzania ## Abstract Background: We evaluated the effects of our postpartum Green Star family planning decision aid on the decisional conflict, knowledge, satisfaction, and uptake of long-acting reversible contraception among pregnant adolescents in Tanzania. Methods: We used a facility-based pre–post quasi-experimental design. The intervention arm received routine family planning counseling and the decision aid. The control received only routine family planning counseling. The primary outcome was the change in decisional conflict measured using the validated decision conflict scale (DCS). The secondary outcomes were knowledge, satisfaction, and contraception uptake. Results: We recruited 66 pregnant adolescents, and 62 completed this study. The intervention group had a lower mean score difference in the DCS than in the control (intervention: −24.7 vs. control: −11.6, $p \leq 0.001$). The mean score difference in knowledge was significantly higher in the intervention than in the control (intervention: 4.53 vs. control: 2.0, $p \leq 0.001$). The mean score of satisfaction was significantly higher in the intervention than in the control (intervention: 100 vs. control: 55.8, $p \leq 0.001$). Contraceptive uptake was significantly higher in the intervention [29 ($45.3\%$)] than in the control [13 ($20.3\%$)] ($p \leq 0.001$). Conclusion: The decision aid demonstrated positive applicability and affordability for pregnant adolescents in Tanzania. ## 1. Introduction Adolescent pregnancies remain a global public health concern with the highest rate occurring in developing countries. Most of these pregnancies are unintended [1,2], that is, unplanned or unwanted pregnancies [3]. The effective use of family planning is one of the most important and fundamental methods for preventing high-risk pregnancies, which typically happen too early and frequently [4]. A high-impact intervention that lowers the risks of adolescent pregnancy is the use of long-acting reversible contraception (i.e., intrauterine copper devices and implants) right away following childbirth [5]. Studies reported that adolescents who started long-acting reversible contraception following childbirth had a lower chance of recurring unwanted pregnancies than adolescents who started short-acting reversible contraception or who did not use any family planning techniques [6]. Despite substantial improvements in the education of healthcare professionals, the provision of quality care, and the distribution of family planning supplies, postpartum contraceptive usage among adolescents in Tanzania continues to be woefully underused. Only $12.2\%$ of adolescent mothers use postpartum contraception within three months of childbirth, with injectables being the most popular method followed by pills [7]. Pregnant adolescents require thorough information on all the contraceptive methods that are available for use immediately following childbirth in their country in order to make an informed choice about which method suits them. Intrauterine copper devices and implants are the only long-acting reversible contraceptive options available immediately after childbirth in Tanzania. Both intrauterine copper devices and implants are quite effective at preventing pregnancy, and they all have a long lifespan and are simple to use [5,8]. So, clarifying values, beliefs, and priorities is necessary in the case of pregnant adolescents to help them make the decision to use the method that suits them. Decision aids have been beneficial, as they inform and educate patients about the available treatment options, which helps reduce decisional conflicts [9]. Patient decision aids are the resources that make it easier for patients to participate in decision-making by outlining the choice that must be made, outlining the possibilities and potential results, and defining personal values [10]. Recently, decision aids are utilized to inform patients and the general public about health issues [11]. Research has demonstrated the effectiveness of decision aids in family planning counseling [12,13]. To our knowledge, however, there has been limited research on decision-support tools that emphasize long-acting reversible contraceptive techniques to increase family planning uptake by pregnant adolescents right after childbirth. This study designed [14] and tested the efficacy of the postpartum “Green Star” family planning decision aid on the decisional conflict, knowledge, satisfaction, and uptake of contraception among pregnant adolescents in Tanzania. We hypothesized that pregnant adolescents using the postpartum “Green Star” family planning decision aid will have a lower decision conflict scale (DCS) score than pregnant adolescents who are not using the decision aid (control group) and that knowledge and satisfaction scores will be higher in pregnant adolescents using the decision aid than in pregnant adolescents who are not using the decision aid (control group). ## 2.1. Study Design, Setting, and Participants This study used a pre–post quasi-experimental design with concurrent control. The study area was Pwani region, which is among the regions in Tanzania with the highest childbearing rates. Thirty percent ($30\%$) of pregnant women in the Pwani region are women aged 15–19 years [7]. Mkuranga and Kisarawe are among the 6 districts of the Pwani region in Tanzania. The selected district hospitals are government-run public health facilities that also serve low-income populations. The inclusion criteria were pregnant adolescents who were between 15–19 years (considered adults in Tanzania) in their 28 gestation weeks planning to deliver at the hospital where they are attending antenatal care services and who were willing and provided consent to participate (Supplementary File S1). The exclusion criteria were pregnant adolescents who were receiving family planning education from other programs. If the participants were minors (10–14 years), informed consent that included the signature of their parents/guardian was required (Supplementary File S2). Participants could also withdraw from this study at any time (Supplementary File S3). The legal age at which a young person in Tanzania may obtain contraceptive services is undefined, and parental consent is not required (Supplementary File S6 [page 5]). Thus, there is no clear law that prevents the use of contraceptives at any age. In addition, marriage law in Tanzania allows a girl who has attained an apparent age of 15 years to get married based on the Law of Marriage Act in Tanzania. Adolescents were selected as the study participants because they have a high risk of rapid repeat pregnancy (RRP), and they critically underutilize different types of modern contraception compared with women above 20 years [15]. ## 2.2. Sampling, Training, and Sample Size Consecutive sampling was used to select the study participants. Enrollment of this study’s participants in both groups ran concurrently, and the study sites were in different districts, which reduced data contamination risk. Two RAs (AM and CM) who were certified midwives recruited the participants and conducted this study in close coordination with the lead researcher (SM). The criteria for selecting the RAs were that they were experienced in family planning counseling, able to conduct intervention studies, and working in an antenatal clinic. Each RA received two days of training separately to avoid data contamination. The sample size was determined based on a systematic review by Stacey et al. [ 16] The data included all published studies that used a randomized controlled trial design evaluating patient decision aids. The data showed a mean difference in the decision conflict scores of −7.22 [−9.12, −5.31]; the estimated sample size needed in each group was 29 participants. In the present study, we also considered a dropout rate of about $10\%$ and thus set the number of participants to 32 in each group for a total of 64 participants recruited (effect size 0.5; power 0.8; significance level $5\%$). ## 2.3. Ethical Considerations The Research Ethics Boards of St. Luke’s International University (Supplementary File S4; Approval number: 20-A91; Approval date: 19 March 2021), Muhimbili University of Health and Allied Sciences (Supplementary File S5; Approval number: MUHAS-REC-1-2020-076; Approval date: 4 March 2020) and National Institute of Medical Research approved this study. This study was registered in the Clinical Trials Registry of University Hospital Information Network in Japan (UMIN000028471). ## 2.4.1. Development of the Postpartum “Green Star” Family Planning Decision Aid We developed the postpartum “Green Star” family planning decision aid by initially identifying the research gap, target population, and study objectives to address the research problem on adolescent pregnancy. We carefully focused and examined previously published studies when determining the study objectives [11,13,17,18]. Thereafter, we identified the individual needs of the participating pregnant adolescents by reviewing a previous study that looked into barriers to the utilization of family planning among female youths in Dar es Salaam, Tanzania [19]. The content, design, and arrangement of the developed prototype decision aid were based on the Ottawa Patient Decision Aid Development eTraining [10], International Patient Decision Aid Standards Collaboration Checklist [20], theory of planned behavior [21], health belief model [22], social cognitive theory [23], current clinical guides for family planning counseling for providers [1,8], and findings from previous studies on the benefits and side effects of the different options, satisfaction and continuation rates, and fertility return [24,25,26,27,28,29,30]. The prototype decision aid has four components based on the Ottawa Patient Decision Aid Development eTraining guide: (i) know how to make a decision with conviction; (ii) understand the characteristics of the decision; (iii) clarify what is important to you; and (iv) make the decision. We shared the prototype decision aid with three experts, namely, our research supervisor and two midwives, all with extensive years of experience in maternal and child health and in developing decision aids. The aim of sharing the prototype decision aid was to receive comments on the comprehensibility and usefulness of the prototype decision aid, which we incorporated in the modified and improved postpartum “Green Star” family planning decision aid. We then carried out a feasibility study of the prototype decision aid to assessing its practicality, usefulness, and acceptability as perceived by pregnant adolescents and healthcare providers [14]. Finally, we developed the third and final version of the decision aid and named it postpartum “Green Star” family planning decision aid. This decision aid was then assessed for its effects on decisional conflict, knowledge, satisfaction, and uptake of long-acting reversible contraception (Supplementary Files S7 and S8). ## 2.4.2. Intervention Group Each participant in the intervention group first received the routine family planning counseling offered by a healthcare provider on duty that day. Then, this was followed by individual face-to-face family planning counseling education using the contents of the postpartum “Green Star” family planning decision aid. It took about 30–40 min for the research assistant (RA) to present all the contents to each participant. Every participant received the 10-page postpartum “Green Star” family planning decision aid and brought it home for further reading and reference when needed. The participants received three education sessions at different times before giving birth (Figure 1). Three education sessions were considered sufficient for the participants to understand the methods and address all the barriers preventing them from using family planning. ## 2.4.3. Control Group The participants in the control group only received routine family planning counseling offered at each antenatal clinic visit. Each participant received three education sessions as in the intervention group (Figure 1). ## 2.5.1. Primary Outcome Decisional conflict was assessed at Time 1 (28 gestational weeks) and at Time 4 (within 2 days after childbirth) using the DCS. The DCS is a 10-item self-report questionnaire that measures a patient’s uncertainty about what treatment option to choose and the factors associated with the uncertainty (e.g., lack of information, myths and misconception, and lack of support) [31]. The DCS has 4 subscales: informed, clarity, support, and uncertainty). Informed was conceptualized as having clear information about the available long-active reversible contraception, including how they work to prevent pregnancy, their benefits, and side effects. Clarity was defined as the quality of someone being clear and easy to understand about personal values for benefits and risks/side effects of long-acting reversible contraception. Support was hypothesized as the feeling a participant had regarding the assistance she receives from healthcare providers and her significant others in deciding to use long-acting reversible contraception. Uncertainty was defined as the feeling of not being completely confident or sure of the best choice of long-acting reversible contraception. Items were answered using a 3-point Likert scale (0 = “Yes”; 2 = “unsure”; 4 = “no”) with scores ranging from 0 to 100. A higher score meant a higher decisional conflict and uncertainty and vice versa [31]. As we adapted items in the DCS, we conducted Cronbach’s Alpha test to check for internal consistency of the DCS items, and we obtained 0.848, which was higher than the commonly recommended value of 0.6. The results indicated that a set of items in the DCS were reliable (i.e., closely related as a group). ## 2.5.2. Secondary Outcomes The secondary outcomes were knowledge, satisfaction with decision-making, and uptake of long-acting reversible contraception. Knowledge was assessed at Time 1 and Time 3 (36–38 gestational weeks). We created a knowledge questionnaire to test the participants’ knowledge of long-acting reversible contraception. We adapted questions from different reports [8,19,24,25,28,29,32], which included advantages, disadvantages, side effects, myths, and misconceptions commonly existing in the community. The knowledge questionnaire had 7 questions, each with a score of 1 for a total of 7 scores. The questions required a “yes” or “no” response. A score of 7 indicated that a participant is knowledgeable about long-acting reversible contraception and vice versa. Satisfaction was only assessed at Time 4. We assessed satisfaction using an effective decision-making subscale of the DCS that had 4 items with 3 responses (4 = “yes”, 2 = “unsure”, 0 = “no”) with scores ranging from 0 to 100. A higher score indicated a higher satisfaction with decision-making and vice versa. The decision on which option to use following childbirth was only assessed at Time 4 using 1 question that asked, “Which option do you prefer?”. ## 2.5.3. Demographic Data The information collected as part of the questionnaire included age, parity, highest education level, occupation, and marital status. ## 2.6. Data Analysis Data were descriptively analyzed using IBM SPSS Statistics version 24.0. Descriptive analysis was used to analyze the demographic information of the participants to determine frequencies and percentages of their distribution within groups. The chi-square test was conducted to analyze data and to observe the distribution within the group for the ordinal and between groups for categorical data. The analysis was based on calculating the mean score differences of the selected variables at Time 1, Time 2, Time 3, and Time 4 between groups. The independent sample t-test was performed to compare the mean scores of DCS, knowledge, and satisfaction between groups. Multiple linear regression to predict the DCS score and logistic regression analysis for long-acting reversible contraception uptake were performed. The statistical tests were performed with a two-sided $5\%$ level of significance. ## 3.1. Flow of This Study Data were collected for 7 months from early March to the end of September 2021. The participant flow diagram is shown in Figure 1. A total of 80 pregnant adolescents were eligible, but only 66 gave their written informed consent and were included in this study with 33 participants in each group. Two participants, one from the intervention and one from the control group were lost to follow-up and were excluded from the analysis. Therefore, a total of 64 participants were included for analysis, each group with a total of 32 participants. ## 3.2. Sociodemographic Characteristics of the Study Participants Details on the sociodemographic information of the study participants are shown in Table 1. The mean age of the study participants in the intervention group was 17.5 (1.29) years, whereas that in the control group was 18.0 years (SD 0.71) [age range, 15–19 years]. Marital status and occupation showed a significant difference between the two groups ($p \leq 0.001$). The ratio of single pregnant adolescents was higher in the intervention group [23 ($71.9\%$)] than in the control group [7 ($21.9\%$)]. Similarly, the ratio of employed pregnant adolescents was higher in the intervention group [27 ($84.4\%$)] than in the control group [8 ($25\%$)]. ## 3.3. Decisional Conflict The total DCS mean score at Time 1 was significantly lower in the intervention group than in the control group (intervention group: 65.00 [SD 19.1] vs. control group: 77.80 [SD 18.4], t = −2.72, $$p \leq 0.008$$) (Table 2). The observed mean scores in the DCS subscales “Support” ($$p \leq 0.723$$) and “Uncertainty” ($$p \leq 0.548$$) were not significantly different between the two groups at Time 1. However, at Time 4, there was a significant difference in the total DCS mean score between the intervention group and the control group (intervention group: 3.13 [SD 4.7] vs. control group: 48.5 [SD 29.6], t = −8.55, $p \leq 0.001$). Additionally, the mean scores of all four subscales (i.e., informed, clarity, support, and uncertainty) in the DCS between the two groups at Time 4 showed a significant difference ($p \leq 0.001$). The mean score difference in the DCS (Time 4 minus Time 1) was significantly lower in the intervention group than in the control group (intervention group: −24.7 [SD 7.99] vs. control group: −11.6 [SD 10.9], t = −5.53, $p \leq 0.001$). The mean score difference of all four subscales in the DCS was significantly lower in the intervention group than in the control group. These results support the hypothesis that the decision-making conflict score will be lower in the intervention group than in the control group. ## 3.4. Multiple Linear Regression for Predicting DCS Score Multiple linear regression was performed to predict the DCS score based on age, occupation, and marital status at Time 1, Time 4, and at the time differences (Time 4 minus Time 1). Age was the only variable that showed a significant relationship in the control group both at Time 1 (β = 0.455, $$p \leq 0.015$$) and at Time 4 (β = 0.506, $$p \leq 0.006$$). However, age did not show a significant relationship at the time differences (Time 4 minus Time 1). As for the intervention group, none of the variables showed a significant relationship with the DCS scores (Table 3). ## 3.5. Contraceptive Knowledge, Satisfaction, and Uptake At Time 1, the results showed no significant difference in the knowledge mean score between the two groups (intervention group: 1.84 [SD 1.98] vs. control group: 2.34 [SD 1.61], t = −1.1, $$p \leq 0.274$$). At Time 3, the knowledge mean score was significantly higher in the intervention group than in the control group (intervention group: 6.38 [1.60] vs. control group: 4.34 [1.82], $t = 4.733$, $p \leq 0.001$). The mean score difference in knowledge (Time 3 minus Time 1) in the intervention group was larger than that in the control group (intervention group: 4.53 [2.54] vs. control group: 2.00 [1.45], $t = 4.88$, $p \leq 0.001$). These findings support the hypothesis that the level of knowledge of long-acting reversible contraception will be higher in the intervention group than in the control group. The mean score of satisfaction was significantly higher in the intervention group than in the control group (intervention group: 100.00 [SD 0.0] vs. control group: 55.8 [SD 30.7], $t = 8.112$, $p \leq 0.001$). The proportion of “yes” responses to each item about satisfaction was greater in the intervention group ($100\%$) than in the control group (< $40\%$). The chi-square test result of each item between the two groups was found to be significant in all items ($p \leq 0.001$). These results support the hypothesis that the state levels of satisfaction with decision-making will be higher in the intervention group than in the control group. The proportion of participants who decided to use long-acting reversible contraception showed significant differences between the two groups. The proportions of participants who “did not decide to use any option” were 3 ($4.7\%$) in the intervention group and 19 (29.7) in the control group. However, the proportions of participants who “decided to use implant” were 29 ($45.3\%$) in the intervention group and 13 ($20.3\%$) in the control group (x2 = 17.73, $p \leq 0.001$). These findings support the hypothesis that the proportion of long-acting reversible contraception uptake will be higher in the intervention group than in the control group. ## 3.6. Logistic Regression Analysis for Long-Acting Reversible Contraception Uptake Age, marital status, and occupation showed no significant relationship with long-acting reversible contraception uptake in the intervention group. However, age alone showed a significant relationship with long-acting reversible contraception uptake in the control group with a negative β value (Table 4). ## 4.1. Decisional Conflict for Pregnant Adolescents This study assessed the effect of the developed postpartum family planning decision aid on the decisional conflict, knowledge, satisfaction, and uptake of long-acting reversible contraception among pregnant adolescents. The study results showed that participants in the intervention group had a lower decisional conflict, higher knowledge, higher satisfaction, and higher contraceptive uptake prevalence than participants in the control group. The present study is the first study in Tanzania using decision aids related to adolescent use of long-acting reversible contraception. Although decision aids have been used in a range of other study areas, some findings concurred with the present study’s findings, whereas others did not. The present results concur with the results of a systematic review that involved 105 studies on decision aids from different areas, namely, cancer screenings, prenatal complication diagnosis, immunizations, and diabetic treatments [9]. These studies found that the use of decision aids more markedly reduced the mean difference in the DCS score in the intervention group (with decision aids) than in the control group (without decision aids) (MD −9.28, $95\%$ CI: −12.2 to −6.36). The effect of decision aids on medication choice for diabetes mellitus was also assessed in three randomized clinical trials [33,34]. The findings of these trials showed a significantly lower mean difference in the DCS score in the intervention group than in the control group. Two clinical trials involving pregnant women have also been conducted [17,35]. The first clinical trial assessed the effect of decision aids on the choice of pregnant women whether to have epidural anesthesia or not during labor [17]. The second randomized clinical trial evaluated the effect of decision aids on women with breech presentation at term [35]. The findings of the two previous clinical trials were similar to the findings of the present study in that the mean difference in the DCS score was significantly lower in the intervention group which received a decision aid than in the control group which received only standardized routine care. On the other hand, the present findings are inconsistent with the findings from a previous study that evaluated the effect of a decision aid on decision-making for the treatment of pelvic organ prolapse [36]. The previous study found that the DCS score of patients who received a decision aid and standard counseling was not significantly lower than the DCS score of patients who received only standard counseling ($$p \leq 0.566$$). The probable reasons include the already available pelvic organ prolapse decision aid in the setting and the regular review of information by the patients together with the clinician at the initial encounter. In the present study, a significant mean difference was observed in the intervention group because there are no family planning decision aids in antenatal clinics to help patients decide on the family planning option that they would take. A study on the experiences of diabetic patients and healthcare providers on shared decision-making conducted in Tanzania found that neither the patients nor the healthcare providers had been using decision-making aids; the patients reported that only health education tools are being used for educating them [37]. The multiple regression analysis of the DCS score based on age, marital status, and occupation showed age as the only variable that had a significant relationship with the DCS score in the control group at Time 1 and Time 4 but not in the intervention group. The findings further show that as age increases, the decision conflict score also increases and vice versa. These findings indicate that if younger adolescents receive the correct information about long-acting reversible contraception at the right time, this will improve their chances of utilizing family planning methods. Regarding the lower DCS score in the intervention group, the present findings suggest that the decision aid played an important role in imparting knowledge and correcting the held myths and misconceptions of the younger adolescents. ## 4.2. Knowledge, Satisfaction, and Uptake of Long-Acting Reversible Contraception for Pregnant Adolescents The mean score difference of knowledge, satisfaction, and uptake of long-acting reversible contraception was significantly higher in the intervention group than in the control group. In a systematic review [9] of 52 studies, 4 randomized control trials [33,34,35,38] and 1 survey [39] found a significant increase in the scores of knowledge, satisfaction with decision making, and choice for the treatment options in the intervention group which used a decision aid compared with the control group which used standardized routine care. Although the present study showed an increase in contraception uptake, none of the long-acting reversible contraception users chose an intrauterine copper device, as all the participants chose only implants. The main reasons for not using an intrauterine copper device might be related to individual perception factors, such as myths, misconceptions, and discomfort from postpartum pain [19]. This nonusage was related to the fear of expulsion and the risk of infection. If an intrauterine copper device is inserted immediately after childbirth and infection prevention control measures are adhered to, then the risk of expulsion and infection would be minimal. However, the postpartum “Green Star” family planning decision aid did not have any information regarding the timely insertion of an intrauterine copper device. In future studies, this information will be included in the decision aid to improve intrauterine copper device uptake. Findings from the logistic regression analysis showed that the marital status and occupation variables did not have a significant relationship with knowledge and satisfaction in both groups. Only the age variable showed a significant relationship with the uptake of long-acting reversible contraception in the control group. The study participants in the control group were $21\%$ (odds = 0.21) less likely to use long-acting reversible contraception than the study participants in the intervention group. Younger adolescents in the control group were more likely to utilize LARC than older adolescents. These findings inform that if family planning programs put great efforts into pregnant adolescents by ensuring that they get the right information at the right time about postpartum family planning, uptake will be improved following childbirth. In addition, we will plan to conduct a qualitative study using interviews focusing on the social–cultural context of adolescent pregnant mothers related to the decision aid of long-acting reversal contraception. ## 4.3. Strengths and Limitations To our knowledge, this is the first study in Tanzania that used a postpartum family planning decision aid to assist in the decision-making of pregnant adolescents regarding which long-acting reversible contraception they should use following childbirth. This means the open practicality of using a decision aid for healthcare even though the subjects were adolescents. To avoid data contamination between groups, the hospital in the intervention group was located in a different district from the hospital in the control group, that is, 2–3 h of driving using private transport. On the other hand, the present findings cannot be generalized unless a randomized controlled study is conducted. ## 5. Conclusions To our knowledge, this is the first quasi-experimental study with a control that evaluated the effects of our recently developed postpartum “Green Star” family planning decision aid on pregnant adolescents’ choice of using long-acting reversible contraception. The decision aid significantly lowered decision-making conflict, improved knowledge and satisfaction with decision-making, and enhanced the uptake of available long-acting reversible contraception. The overall findings indicate the usefulness of the postpartum “Green Star” family planning decision aid, as it supplemented and supported patient–provider communications during family planning counseling in antenatal clinics. ## References 1. Adolescent Pregnancy: Key Facts. (2018.0) 2. Neal S., Matthews Z., Frost M., Fogstad H., Camacho A.V., Laski L.. Childbearing in adolescents aged 12–15 years in low resource countries: A neglected issue. New estimates from demographic and household surveys in 42 countries. Acta Obstet. Gynecol. Scand. (2012.0) 91 1114-1118. DOI: 10.1111/j.1600-0412.2012.01467.x 3. Unintended Pregnancy in the United States: Fact sheet 4. 4. World Health Organization Primary Health Care: Report of the International Conference on Primary Health Care. Alma-Ata, USSR, 6–12 September 1978World Health Organization, United Nations Children’s FundGeneva, SwitzerlandSan Francisco, CA, USA1978Available online: https://www.who.int/publications/i/item/9241800011(accessed on 7 March 2023). Primary Health Care: Report of the International Conference on Primary Health Care. Alma-Ata, USSR, 6–12 September 1978 (1978.0) 5. Long-acting reversible contraception: Implants and intrauterine devices. Obstet Gynecol. (2017.0) 130 251-269 6. Baldwin M.K., Edelman A.B.. The Effect of Long-Acting Reversible Contraception on Rapid Repeat Pregnancy in Adolescents: A Review. J. Adolesc. Health (2013.0) 52 S47-S53. DOI: 10.1016/j.jadohealth.2012.10.278 7. Tanzania Demographic and Health Survey and Malaria Indicator Survey (TDHS-MIS) 2015–2016 8. Decision-Making Tool for Family Planning Clients and Providers. (2015.0) 9. Stacey D., Légaré F., Lewis K., Barry M.J., Bennett C.L., Eden K.B., Holmes-Rovner M., Llewellyn-Thomas H., Lyddiatt A., Thomson R.. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst. Rev. (2017.0) 2017 CD001431. DOI: 10.1002/14651858.CD001431.pub5 10. Patient Decision Aids 11. Bennett K.F., von Wagner C., Robb K.A.. Supplementing factual information with patient narratives in the cancer screening context: A qualitative study of acceptability and preferences. Health Expect. (2015.0) 18 2032-2041. DOI: 10.1111/hex.12357 12. Wu J.P., Damschroder L.J., Fetters M.D., Zikmund-Fisher B.J., Crabtree B.F., Hudson S.V., Iv M.T.R., Fucinari J., Kang M., Taichman L.S.. A Web-Based Decision Tool to Improve Contraceptive Counseling for Women with Chronic Medical Conditions: Protocol for a Mixed Methods Implementation Study. JMIR Res. Protoc. (2018.0) 7 e107. DOI: 10.2196/resprot.9249 13. Kim Y.M., Davila C., Tellez C., Kols A.. Evaluation of the World Health Organization′s family planning decision-making tool: Improving health communication in Nicaragua. Patient Educ. Couns. (2007.0) 66 235. DOI: 10.1016/j.pec.2006.12.007 14. Mushy S.E., Shishido E., Leshabari S., Horiuchi S.. Postpartum Green Star family planning decision aid for pregnant adolescents in Tanzania: A qualitative feasibility study. BMC Reprod. Health (2021.0) 18. DOI: 10.1186/s12978-021-01216-6 15. Cleland J., Conde-Agudelo A., Peterson H., Ross J., Tsui A.. Family planning needs during the first two years postpartum in Tanzania. Contraception and health. Lancet (2012.0) 380 149-156. DOI: 10.1016/S0140-6736(12)60609-6 16. Stacey D., Légaré F., Lewis K., Barry M.J., Bennett C.L., Eden K.B., Trevena L.. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst. Rev. (2017.0) 4. DOI: 10.1002/14651858.CD001431.pub5 17. Shishido E., Osaka W., Henna A., Motomura Y., Horiuchi S.. Effect of a decision aid on the choice of pregnant women whether to have epidural anesthesia or not during labor. PLoS ONE (2020.0) 15. DOI: 10.1371/journal.pone.0242351 18. Osaka W., Nakayama K.. Effect of a decision aid with patient narratives in reducing decisional conflict in choice for surgery among early-stage breast cancer patients: A three-arm randomized controlled trial. Patient Educ. Couns. (2017.0) 100 550-562. DOI: 10.1016/j.pec.2016.09.011 19. Mushy S.E., Tarimo E.A., Massae A.F., Horiuchi S.. Barriers to the uptake of modern family planning methods among female youth of Temeke District in Dar es Salaam, Tanzania: A qualitative study. Sex. Reprod. Health (2020.0) 24 100499. DOI: 10.1016/j.srhc.2020.100499 20. IPDAS Criteria for Judging the Quality of Patients Decision Aids. (2005.0) 21. Ajzen I.. The Theory of Planned Behavior. Organ. Behav. Hum. Decis. Process. (1991.0) 50 179-211. DOI: 10.1016/0749-5978(91)90020-T 22. Green E.C., Murphy E.. Health Belief Model. Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society (2014.0) 766-769 23. Bandura A., Van Lange P.A.M., Kruglanski A.W., Higgins E.T.. Social Cognitive Theory. Handbook of Theories of Social Psychology (2012.0) 349-374. DOI: 10.4135/9781446249215.n18 24. Peipert J.F., Zhao Q., Allsworth J., Petrosky E., Madden T., Eisenberg D., Secura G.. Continuation and Satisfaction of Reversible Contraception. Obstet. Gynecol. (2011.0) 117 1105-1113. DOI: 10.1097/AOG.0b013e31821188ad 25. Yisa S.B., A Okenwa A., Husemeyer R.P.. Treatment of pelvic endometriosis with etonogestrel subdermal implant (Implanon. J. Fam. Plan. Reprod. Health Care (2005.0) 31 67-69. DOI: 10.1783/0000000052972799 26. Funk S., Miller M.M., Mishell D.R., Archer D.F., Poindexter A., Schmidt J., Zampaglione E.. Safety and efficacy of Implanon™, a single-rod implantable contraceptive containing etonogestrel. Contraception (2005.0) 71 319-326. DOI: 10.1016/j.contraception.2004.11.007 27. Hubacher D., Grimes D.A.. Noncontraceptive Health Benefits of Intrauterine Devices: A Systematic Review. Obstet. Gynecol. Surv. (2002.0) 57 120-128. DOI: 10.1097/00006254-200202000-00024 28. Zheng S.-R., Zheng H.-M., Qian S.-Z., Sang G.-W., Kaper R.. A randomized multicenter study comparing the efficacy and bleeding pattern of a single-rod (Implanon. Contraception (1999.0) 60 1-8. DOI: 10.1016/S0010-7824(99)00053-0 29. Soeprono R.. Return to fertility after discontinuation of copper IUD use: A study of 55 pregnancies involving Multiload Cu-250 users among private patients in Indonesia. Adv. Contracept. (1988.0) 4 95-107. DOI: 10.1007/BF01849510 30. Buckshee K., Chatterjee P., Dhall G., Hazra M., Kodkany B., Lalitha K., Logambal A., Manchanda P., Nanda U., RaiChoudhury G.. Return of fertility following discontinuation of norplantR-II subdermal implants: ICMR task force on hormonal contraception. Contraception (1995.0) 51 237-242. DOI: 10.1016/0010-7824(95)00039-D 31. O’Connor A.M.. User Manual-Decisional Conflict Scale (10 Item Questions Format) (2010.0) 32. Lopez L.M., Grey T.W., Chen M., Hiller J.E.. Strategies for improving postpartum contraceptive use: Evidence from non-randomized studies. Cochrane Database Syst Rev. (2014.0) 11 CD011298. DOI: 10.1002/14651858.CD011298.pub22 33. Mathers N., Ng C.J., Campbell M.J., Colwell B., Brown I., Bradley A.. Clinical effectiveness of a patient decision aid to improve decision quality and glycaemic control in people with diabetes making treatment choices: A cluster randomised controlled trial (PANDAs) in general practice. BMJ Open (2012.0) 2 e001469. DOI: 10.1136/bmjopen-2012-001469 34. Mullan R.J., Montori V., Shah N.D., Christianson T.J.H., Bryant S.C., Guyatt G.H., Perestelo-Perez L.I., Stroebel R.J., Yawn B.P., Yapuncich V.. The Diabetes Mellitus Medication Choice Decision Aid. Arch. Intern. Med. (2009.0) 169 1560-1568. DOI: 10.1001/archinternmed.2009.293 35. Nassar N., Roberts C.L., Raynes-Greenow C.H., Barratt A., Peat B.. Evaluation of a decision aid for women with breech presentation at term: A randomised controlled trial [ISRCTN14570598]. BJOG Int. J. Obstet. Gynaecol. (2007.0) 114 325-333. DOI: 10.1111/j.1471-0528.2006.01206.x 36. Brazell H.D., O’Sullivan D.M., Forrest A., Greene J.F.. Effect of a Decision Aid on Decision Making for the Treatment of Pelvic Organ Prolapse. Female Pelvic Med. Reconstr. Surg. (2015.0) 21 231-235. DOI: 10.1097/SPV.0000000000000149 37. Vedasto O., Morris B., Furia F.F.. Shared decision-making between health care providers and patients at a tertiary hospital diabetic Clinic in Tanzania. BMC Health Serv. Res. (2021.0) 21. DOI: 10.1186/s12913-020-06041-4 38. Nagle C., Gunn J., Bell R., Lewis S., Meiser B., Metcalfe S., Ukoumunne O.C., Halliday C.. Use of a decision aid for prenatal testing of fetal abnormalities to improve women′s informed decision making: A cluster randomized controlled trial [ISRCTN22532458]. BJOG (2008.0) 115 339-347. DOI: 10.1111/j.1471-0528.2007.01576.x 39. Chewning B., Mosena P., Wilson D., Erdman H., Potthoff S., Murphy A., Kuhnen K.K.. Evaluation of a computerized contraceptive decision aid for adolescent patients. Patient Educ. Couns. (1999.0) 38 227-239. DOI: 10.1016/S0738-3991(99)00014-2
--- title: Surface-Modified Inhaled Microparticle-Encapsulated Celastrol for Enhanced Efficacy in Malignant Pleural Mesothelioma authors: - Xuechun Wang - Gautam Chauhan - Alison R. L. Tacderas - Aaron Muth - Vivek Gupta journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049545 doi: 10.3390/ijms24065204 license: CC BY 4.0 --- # Surface-Modified Inhaled Microparticle-Encapsulated Celastrol for Enhanced Efficacy in Malignant Pleural Mesothelioma ## Abstract Malignant pleural mesothelioma (MPM) is a rare and aggressive cancer affecting the pleural lining of the lungs. Celastrol (Cela), a pentacyclic triterpenoid, has demonstrated promising therapeutic potential as an antioxidant, anti-inflammatory, neuroprotective agent, and anti-cancer agent. In this study, we developed inhaled surface-modified Cela-loaded poly(lactic-co-glycolic) acid (PLGA) microparticles (Cela MPs) for the treatment of MPM using a double emulsion solvent evaporation method. The optimized Cela MPs exhibited high entrapment efficiency (72.8 ± $6.1\%$) and possessed a wrinkled surface with a mean geometric diameter of ~2 µm and an aerodynamic diameter of 4.5 ± 0.1 µm, suggesting them to be suitable for pulmonary delivery. A subsequent release study showed an initial burst release up to 59.9 ± $2.9\%$, followed by sustained release. The therapeutic efficacy of Cela MPs was evaluated against four mesothelioma cell lines, where Cela MP exhibited significant reduction in IC50 values, and blank MPs produced no toxicity to normal cells. Additionally, a 3D-spheroid study was performed where a single dose of Cela MP at 1.0 µM significantly inhibited spheroid growth. Cela MP was also able to retain the antioxidant activity of Cela only while mechanistic studies revealed triggered autophagy and an induction of apoptosis. Therefore, these studies highlight the anti-mesothelioma activity of Cela and demonstrate that Cela MPs are a promising inhalable medicine for MPM treatment. ## 1. Introduction Malignant mesothelioma (MM) is a rare, incurable, and malignant type of cancer arising from the mesothelial or sub-mesothelial cells of the pleura (lung lining), peritoneum (abdominal lining), pericardium (heart sac), or testes [1]. MM is classified as a rare disease with ~30,000 new cases diagnosed worldwide annually [2], with a subtype, malignant pleural mesothelioma (MPM), accounting for 80–$90\%$ of all MM cases. Occupational exposure to asbestos fibers has been identified as the major cause of MPM in patients, leading to various symptoms, such as difficulty breathing, chest pain, fluid buildup, and others [3,4]. Due to a latency period in tumor development of 20–50 years after exposure to asbestos, the incidences of MPM have increased slightly in the last decade, especially in industrialized countries, such as the UK, Australia, and Belgium [1,3]. Additionally, roughly 3000 new cases are diagnosed each year in the United States [5]. Current therapies include surgical resection or combinations of chemotherapy, radiation, and immunotherapy; however, these approaches have not shown great success with patients experiencing multiple adverse side effects and only give a ~$10\%$ 5-year survival rate [6]. Since MPM has been studied less extensively than other primary neoplasms of the chest, there is an increased urgency to explore and develop more efficacious treatment options while minimizing side effects. Celastrol (Cela) is an extensively studied phytochemical derived from the root extract of Tripterygium wilfordii Hook F (Thunder God Vine), which has demonstrated interesting and promising bioactivity [7]. Numerous studies have demonstrated Cela’s intriguing pharmacological properties, including anti-inflammation [8,9], neuroprotection [10,11], and anti-cancer [12,13]. Additionally, Cela has been shown to induce apoptosis in cancer cells by activating apoptotic proteins such as caspase-$\frac{3}{8}$/9 [14,15]. Cela has also been reported to induce autophagy and cause tumor growth inhibition [16,17]. Despite encouraging in vitro and preclinical efficacy, the clinical applications of Cela remain challenging due to its extremely low aqueous solubility, low bioavailability, narrow therapeutic window, and potential toxicity [7,12]. Therefore, the use of an effective carrier may help to improve these unfavorable physiochemical and pharmacokinetic properties and possibly enhance its treatment potential while reducing off-target cytotoxicity. Pulmonary drug delivery presents various advantages over oral and parenteral routes for localized treatment of chronic lung disorders. This route of administration provides targeted lung delivery, local accumulation, rapid onset, reduced systemic exposure, extensive surface area for drug absorption, and reduced side effects, and requires lower doses for similar lung deposition of therapies [18,19]. The aerodynamic diameter of inhaled particles should range from 2–5 µm for efficient deep lung targeting [20]; however, the entry of foreign materials into the respiratory tract may activate defense mechanisms which may remove them (e.g., mucociliary clearance and alveolar macrophage phagocytosis) [21]. Recently, polymer-based porous microparticles have been identified as effective carriers for pulmonary delivery due to their large geometric diameters and low densities, which not only provide protection from clearance by macrophages, but are also capable of being deposited in the deep lung [19,22,23,24]. A study by Hu et al. developed inhalable curcumin-loaded poly(lactic-co-glycolic) acid (PLGA) porous microparticles with a mean geometric diameter >10 µm and an aerodynamic diameter of only 3.12 µm, which effectively avoided uptake by alveolar macrophages [23]. PLGA has been widely used to develop porous microparticles for delivering small molecules, proteins, and plasmid DNA to the lungs [25,26,27,28,29]. This is the first report demonstrating Cela’s efficacy for MPM treatment. In addition, the present study explores surface-modified microparticles as effective inhaled carriers possessing deep lung deposition properties. Cela-loaded PLGA microparticles (Cela MPs) were prepared using pore-forming agent sodium chloride (NaCl) to create corrugated surface morphology that provides favorable aerodynamic properties while efficiently avoiding macrophage uptake. Optimized Cela MPs were analyzed for their aerodynamic performance as well as their 2D anti-proliferative properties and 3D anti-tumor efficacy against mesothelioma cell lines. Finally, multiple anti-cancer mechanisms of action of Cela MPs were elucidated. ## 2.1. Formulation of Celastrol Microparticles (Cela MPs) Cela-loaded microparticles were successfully prepared using a double w/o/w emulsion solvent evaporation method (Table 1 and Table S1). The F1 formulation resulted in a good %EE of $63.7\%$ with a negative of −30.3 mV. To increase zeta potential to potentially improve cell internalization, we introduced $1\%$ PEI in the inner aqueous phase and obtained an F2 formulation with a 39.5 mV zeta potential and ≈$60\%$ drug entrapment. Formulations F3 and F4 consisted of the same parameters as F1 and F2, respectively, with the exception that the outer aqueous phase was increased to $2\%$ PVA. The resulting %EEs (F3: $67.1\%$ and F4: $52.0\%$) and zeta potential (F3: −32.9 mV and F4: 35.7 mV) were found to be similar to that of the F1 and F2 formulations. Therefore, $1\%$ PVA was selected for the latter formulations. Comparatively, the addition of PEI in the composition (F2 and F4) reduced the %EE; therefore, PEI was not included in further formulations. In this project, we aimed to create large porous particles to achieve the required aerodynamic size for delivery and to reduce the risk of phagocytosis [30]. Two pore-forming agents, or porogens (NaHCO3 and NaCl), were examined. Incorporating NaHCO3 (F5) in the inner aqueous phase resulted in a reduced %EE of $34.8\%$, particle size of 2341 nm, and zeta potential of −21.6 mV, along with a very high PDI of 0.791. However, the addition of NaCl (F6) as a porogen resulted in a relatively higher %EE of 72.8 ± $6.1\%$, lower PDI of 0.3 ± 0.2, and a stable zeta potential of −36.6 ± 3.8 mV, along with a particle size of 2076 ± 390 nm geometric size and a PDI of 0.3 ± 0.2. F6 formulation was selected for further studies. ## 2.2. Morphological Analysis The morphology of Cela MPs was examined by SEM. A representative image is presented in Figure 1A, which reveals a wrinkled pattern that covers the surface of the MPs and is ~2 µm in diameter. The fold-like patterns on the surface are due to the addition of NaCl during the preparation process that is responsible for the disruption of the PLGA matrix as the MPs are formed. ## 2.3. In Vitro Release Study The release profile of Cela MPs in $1\%$ SLS in SLF at 37 °C is shown in Figure 1B. As can be seen, after 30 min, 59.9 ± $2.9\%$ of Cela was released, suggesting an initial burst release from the MP formulation, followed by sustained release up to 81.0 ± $1.0\%$ by 72 h. This demonstrates the ability of MPs to release a majority of the entrapped drug soon after reaching the lungs, followed by maintenance doses being released over a period of time. Several reports have suggested the elimination half-life of Cela being 7.5–10 h in rats, which outweighs the need to develop extended-release dosage forms [31]. ## 2.4. Differential Scanning Calorimetry (DSC) Studies The DSC thermogram of Cela (Figure 1C) shows a sharp endothermic peak at 155 °C and an exothermic peak at 211 °C. Both physical mixture samples displayed endothermic and exothermic peaks for Cela at 156 °C and 210 °C, respectively, with an additional endothermic peak originating from the amorphous PLGA polymer at ~50 °C [32]. However, these characteristic peaks were absent in the thermograms of the blank MP and Cela MP, which indicated the successful encapsulation of Cela in the microparticles. ## 2.5. X-ray Diffraction Studies (XRD) Diffractometry studies revealed multiple peaks representing the crystalline nature of Cela, shown in Figure 1D. The diffractograms of the physical mixtures comprising of Cela + PLGA + NaCl and Cela + Blank MP both clearly demonstrate the presence of Cela crystalline peaks. Upon analyzing the Blank MP and Cela MP, Cela peaks were absent over the entire range of 10–80° 2θ values. The results obtained from the XRD studies also indicate successful encapsulation of Cela in the microparticles. ## 2.6. In Vitro Stability The stability of optimized Cela MP was determined in terms of particle size, zeta potential, and percent drug entrapment as a function of storage time and conditions. As seen in Figure 2, there was no significant change in all three parameters after the Cela MP suspension was stored for 28 days at 4 °C. The particle size at day 28 was 2162.7 ± 376.1 nm, compared to 2358.3 ± 359.6 nm at day 0 (Figure 2A). The zeta potential at day 0 was −24.4 ± 2.5 mV, and at day 28 it was −25.2 ± 3.3 mV (Figure 2B). The %EE at day 0 was 84.7 ± $2.1\%$, and on day 28 it was 90.1 ± 5.7 mV (Figure 2C). The in vitro stability studies performed for Cela MP suspension found it to be stable at 4 °C for a period of 28 days. ## 2.7. Aerosolization of Cela MPs In vitro lung deposition studies were performed using a Next Generation Impactor (NGI). Figure 3A shows the deposition of microparticles at various stages of the NGI. As can be seen, most of the emitted particles were deposited in stage 3 and below, representing the bronchi-alveolar region. Figure 3B demonstrates the percent cumulative deposition of Cela MPs as a function of effective cut-off diameter. As can be seen, over $60\%$ of MPs possessed a cut-off diameter of 3.3 µm, which is equivalent to stage 4 or below on the NGI (Figure 3B). The mass median aerodynamic diameter (MMAD) of Cela MP (4.5 ± 0.1 µm) was within the range of a required aerodynamic diameter for sufficient delivery to the respirable region of the lungs, i.e., 2–5 µm (Figure 3C). The GSD was calculated to be 2.1 ± 0.1 µm, indicating the aerodynamic size range of MPs from the MMAD value. The FPF, or respirable particles with aerodynamic diameter from stage 3 (5.39 µm) of the NGI and below, was 85.3 ± $1.5\%$, and the amount of residual formulation left in the cup after nebulization period (percent device) was 51.8 ± $6.4\%$, as shown in Figure 3C. This aerosolization data suggests that the Cela MP can be efficiently delivered to the lungs. Additionally, %EE, PS, PDI, and zeta potential were evaluated for Cela MPs after nebulization. As presented in Table S2, all four parameters showed no significant difference from the data presented for the optimized formulation in Table 1. Therefore, Cela MPs maintained their stability after nebulization. ## 2.8. Cytotoxicity Studies The cytotoxicity of Cela MP was compared to Cela only against four mesothelioma cell lines using an MTT assay (Figure 4). As can be observed, Cela MP significantly improved the cytotoxic activity of Cela against all four cell lines (Figure 4A–D), with significantly more robust cytotoxic activity in H28 and H2452 cells. The IC50 values of Cela and Cela MPs were calculated using GraphPad Prism®, and it was observed that microparticle encapsulation reduced the IC50 of Cela ~1.3-fold against MTSO-211H cells (Cela: 3.6 ± 0.4 µM vs. Cela MP: 2.7 ± 0.3 µM; $p \leq 0.05$) (Figure 4A; Table 2). Similarly, IC50 values were reduced by ~3.4, ~5, and ~1.5-folds against H28, H2452, and ROB cells, respectively (Figure 4B–D and Table 2). To determine whether blank MPs imparted any notable toxicities, blank MPs were evaluated against MSTO-211H cells and normal human lung fibroblast (NHLF) cells. As shown in Figure 5, blank MPs did not induce any appreciable toxicity, and the cell viability remained ~$100\%$ for all concentrations evaluated, in both MSTO-211H mesothelioma cell lines (Figure 5A) and NHLF normal primary lung fibroblasts (Figure 5B). This highlights the improvement in Cela efficacy being due to microparticle encapsulation and not on other formulation components. ## 2.9.1. DPPH Assay Cela is a known antioxidant that may boost its anti-inflammatory and anti-cancer properties. To test the ability of Cela and Cela MPs in scavenging free radicals, a DPPH assay was performed. As shown in Figure 6A,B, the DPPH radical scavenging activity of Cela and Cela MPs dose-dependently increased using 0.5, 1, 5, and 10 µM concentrations. After 30 min, the antioxidant activity of Cela only and Cela MPs was very similar (Figure 6A). After 24 h, the antioxidant activity of Cela MPs surpassed that of the Cela solution (Figure 6B). For example, 5 µM Cela exhibited scavenging activity of 15.7 ± $1.9\%$, while Cela MP resulted in scavenging activity of 30.2 ± $4.1\%$. The antioxidant effects were statistically significant at higher concentrations (5 and 10 µM, $p \leq 0.0001$). These results suggest that the antioxidant activity of Cela was either retained or enhanced upon encapsulation into the PLGA microparticles. ## 2.9.2. Caspase-3 Assay Caspase-3 activity was then monitored to examine the effects of Cela and Cela MP treatment on apoptotic pathways. Figure 6C displays the percent fluorescence intensity of caspase-3 after MSTO-211H cells were treated for 6 h. As can be seen, compared to the control ($100\%$), lower concentration (0.5 µM) treatments resulted in no significant difference in caspase-3 levels (Cela: 102.1 ± $6.2\%$; Cela MP: 218.1 ± $99.5\%$; ns). Conversely, treatment of 1.0 µM Cela MP (503.1 ± $93.1\%$) resulted in a ~5-fold increase in caspase-3 levels compared to the control ($p \leq 0.01$) and ~3.5-fold increase compared to 1.0 µM Cela (143 ± $19.7\%$) ($p \leq 0.01$) (Figure 6C). Therefore, a significant apoptotic induction was observed with Cela MP-treated cells as compared to the drug only-treated cells, suggesting induction of apoptosis as a mechanism for its anti-cancer properties. ## 2.9.3. Effect of Cela and Cela MP on Cellular Autophagy Cela has also been reported to trigger autophagic pathways in which promote cell death in cancer [33,34,35]. In this study, an in vitro CYTO-ID® Autophagy Detection Kit was used to determine any effect on LC3 levels, where increased levels indicated induced autophagy. The effects of Cela and Cela MP were investigated in the MSTO-211H cell line. As shown in Figure 6D, the LC3 fluorescent signal upon treatment with 5 µM Cela MP increased significantly, ~1.3-fold relative to the control cells ($p \leq 0.01$) and ~1.5-fold relative to Cela ($p \leq 0.001$). No significant difference was found between the control and Cela only treatments. This signifies that Cela MP induced autophagosome formation in mesothelioma cells at a concentration of just 1.0 µM. ## 2.10.1. Tumor Volume Reduction Studies An in vitro tumor simulation model was utilized to mimic in vivo tumor conditions and better predict the clinical effectiveness of Cela MP. The single-dose study involved drug treatment once throughout the 15-day experimental period. Representative tumor spheroid images from single dose treatment are shown in Figure 7A. As can be seen, treatment with both 0.5 and 1.0 µM Cela MPs visibly inhibited spheroid growth as compared to the control and Cela only at the same concentrations (Figure 7A). A detailed analysis of the spheroid volume (normalized to day 0 volume), as represented in Figure 7B, indicated that after 15 days of treatment, a lower concentration of Cela MP (0.5 µM) significantly suppressed spheroid growth by ~1.6-fold ($p \leq 0.01$) as compared to the control. Additionally, a higher concentration of Cela MP (1.0 µM) suppressed growth by ~13-fold as compared to the control ($p \leq 0.0001$) and ~10-fold as compared to 1.0 µM Cela only ($p \leq 0.0001$). A multiple-dose study was then performed to mimic the physiological conditions a treatment may undergo (e.g., drug metabolism and clearance) [36,37]. Similarly, Cela MP in the multiple-dose studies showed superior anti-cancer activity as compared to the drug only (representative spheroid images from each group shown in Figure 8A). In the multiple-dose studies, all four treatment groups displayed significant suppression of spheroid growth as compared to the control (Figure 8B) (0.5 µM Cela: ~1.5-fold; 0.5 µM Cela MP: ~7.6-fold; 1.0 µM Cela: ~3.5-fold; 1.0 µM Cela MP: ~11.7-fold). However, at a concentration of just 0.5 µM, Cela MP-treated spheroids were significantly reduced in size as compared to those treated with Cela only (~5-fold reduction) ($p \leq 0.0001$). Therefore, the 3D spheroid volume data concludes that enhanced anti-tumorigenic efficacy is achieved for our formulation against MPM. ## 2.10.2. Cell Viability Assay A CellTiter Glo assay was also performed to determine the cytotoxic potential of the previously described spheroid treatments. Figure 7C represents the percent cell viability of single-dosed spheroids relative to the control. As can be seen, only 1.0 µM Cela MP treatment showed a significant reduction in cell viability as compared to the control (~9-fold reduction) and to 1.0 µM Cela only (~8-fold reduction) ($p \leq 0.0001$). The multiple-dose treatment regimen proved that all four treatment groups significantly reduced cell viability as compared to the control ($p \leq 0.0001$) and no statistical difference was found between the treatment groups (Figure 8C). The data obtained from the CellTiter Glo assay demonstrated that a single dose of 1.0 µM Cela MP can effectively penetrate the 3D spheroid and reduce cell viability. ## 2.10.3. Live/Dead Assay A Live/Dead assay helps visualize the viable and necrotic cells that are present in the treated spheroids. As shown in Figure 9A, green fluorescence (GFP) represents live cells, whereas red fluorescence (RFP) represents dead cells. As can be seen, a single dose of 1.0 µM Cela MP increased RFP intensity as compared to the control and other treatments. Multiple-dose spheroid images showed all four treatments reduced GFP intensity as compared to the control. Figure 9B,C represent quantitative comparisons of RFP intensity relative to the control spheroids. Quantification of the single dose study indicated that 1.0 µM Cela MP elevated RPF intensity by ~2.5-fold as compared to the control and 1.0 µM drug only. The multiple-dose data showed no significant difference in RFP intensity as compared to the control. This may be due to the decreased size of the treated spheroids as compared to the control (as mentioned in Section 2.9), thereby also resulting in reduced RFP intensity. ## 3. Discussion Celastrol (Cela), a natural product from Tripterygium wilfordii Hook F (Thunder God Vine), is known to be efficacious against a variety of cancers, including liver cancer, breast cancer, prostate cancer, multiple myeloma, glioma, etc. [ 38]. This project presents the first report of Cela’s efficacy for MPM treatment. Studies suggest that the anti-cancer properties of Cela can be attributed to: (i) induced apoptosis and autophagy, (ii) cell cycle arrest, (iii) anti-angiogenic activity, (iv) anti-inflammatory activity, and (v) antioxidant properties [38,39,40]. However, the clinical use of Cela has been limited by multiple factors: poor water solubility, low bioavailability, narrow therapeutic window, and undesired side effects [41,42]. To address these issues, we aimed to develop a microparticle carrier system for Cela with high aerosolization efficiency to reduce the effective dose and enhance the therapeutic efficacy of the drug for MPM treatment. The optimized Cela MP (F6 formulation) resulted in high drug entrapment (72.8 ± $6.1\%$), which is equivalent to 1.5 ± 0.12 mg (or 3.2 ± 0.3 mM), which represents a significant improvement in Celastrol solubilization and encapsulation of about ≈0.8 mM, as shown by Shukla et al. using cyclodextrin complexation [12]. The zeta potential of the MPs was found to be below −30 mV, which indicated stable nanofluids outside of the +30 and –30 mV range [43]. The geometric size of the MPs was analyzed to be 2076 ± 390 nm, which is within the range of 1–5 µm which is desired for the best delivery efficiency [44]. However, this particle size range was also previously shown to be ideal for phagocytosis, as per a study by Champion et al. [ 45]. Therefore, to circumvent this limitation, we developed microparticles with irregular surfaces using NaCl as the porogen. According to Yang et al., highly porous PLGA microparticles were able to avoid phagocytosis by macrophages, while non-porous small particles were quickly taken up by macrophages [46]. As observed from the SEM image, the MP is ~2 µm in diameter and possesses a wrinkled surface. Sodium chloride was used to create an osmotic gradient capable of driving water into the internal droplets, thus causing them to swell in size. During the evaporation process, the swollen inner droplets are immobilized within the PLGA matrix and are eventually evaporated, ultimately creating pores [47]. However, our formulation resulted in wrinkled surfaces instead of a porous structure. Based on mechanics described by Li et al., we theorize that the use of a very small amount of NaCl ($1\%$ of inner aqueous phase) caused the inner droplets to not swell enough as to adequately form pores in the PLGA matrix [47,48]. Additionally, during the evaporation process, the particles shrank as the inner droplets were removed. As shrinkage reaches a critical point, the particle surface bifurcates into fold-like structures to release the circumferential compression within the shell or elastic strain energy [48]. Therefore, the particle size of the MPs was not observed to be above 5 µm. Furthermore, studies have reported wrinkled particles prevent friction, interlocking forces, and water bridge formations due to smaller distances between the particles [49], as well as promote cell attachment without any chemical processing as compared to spherical particles [50]. Release of the drug from the microparticles is important to predict the in vivo performance of the formulation. The developed formulation displayed a biphasic release pattern characterized by an initial burst release due to the drug being present near the particle surface [51], followed by a sustained release that may be attributed to gradual degradation of the PLGA matrix and diffusion of Cela through the matrix [52,53]. This release profile proved to be highly effective at increasing the cytotoxic effect of Cela MP against MPM cells. In addition, Cela has shown to exhibit a long elimination half-life (more than 7 h) in rats; therefore, a sustained-release formulation may not be necessary to obtained the desired therapeutic efficacy of Cela [31]. DSC studies showed that Cela had a characteristic endothermic peak at 155 °C, corresponding to its melting point, and an exothermic peak at 211 °C, likely due to degradation of Cela [54]. Both DSC and XRD studies observed the absence of characteristic Cela peaks in the Cela MP sample, suggesting successful inclusion of Cela in the microparticle formulation. To design a successful delivery system, it was necessary to ensure a stable microparticle formulation. Since it is expected that the MP formulation will be stored in the refrigerator prior to use, the stability study was carried out at 4 °C over 28 days. Subsequently, it was demonstrated that Cela MPs maintained insignificant changes in particle size, zeta potential, and %EE, proving the polymers maintained the structural integrity of the MPs during long-term storage. Conventional routes of administration, such as oral and intravenous, are not ideal for Cela delivery. For instance, a pharmacokinetic study done on rats demonstrated the mean oral absolute bioavailability after oral administration (3 mg/kg) was very low at $3.14\%$ [55]. Even though bioavailability is not a concern in intravenous administration, Cela is associated with many side effects, including infertility, cardiotoxicity, hepatotoxicity, hematopoietic system toxicity, and nephrotoxicity [38]. Thus, the pulmonary route of administration is inarguably the best choice to efficiently deliver therapeutics locally to the lungs to improve bioavailability and avoid adverse effects. In a recent report, inhaled nintedanib (given at 1:120 of the oral dose) was found to deliver an oral-equivalent lung Cmax with lower systemic AUC, and was well-tolerated and effective at reducing bleomycin-induced pulmonary fibrosis [56]. However, simple pulmonary delivery of free drug suspension is insufficient for deep lung deposition due to the large crystalline drug particles and variability in particle sizes. Therefore, effective pulmonary drug delivery can be achieved by optimizing the physical properties of the particle formulation, including size, charge, density, shape, solubility, and lipophilicity [57]. NGI is a popular and competent tool to evaluate the aerodynamic properties of particles based on their deposition profile throughout the stages, and was used to evaluate the wrinkled-surface Cela MPs [52,58,59]. A MMAD of 4.5 ± 0.1 µm suggested that Cela MPs are within the ideal range of 1–5 µm for the greatest pulmonary delivery efficiency [60]. Multiple studies have demonstrated that wrinkle morphology enhances the aerodynamics of aerosol in inhalation purposes [61,62,63]. Therefore, Cela MP is a promising strategy to ensure efficient delivery to the respirable regions of the lung. Additional characterization studies performed on Cela MPs after nebulization demonstrated that the MPs maintained their original %EE, PS, PDI, and zeta potential. Thus, Cela MPs are suitable for the pulmonary route of administration. The in vitro cytotoxicity studies demonstrated that microparticles can be used as prospective carriers for MPM treatment by improving therapeutic efficacy at reduced doses. This may be explained by better cell attachment that may cause enhanced cellular uptake and subsequent cytotoxicity. Particles with wrinkled surfaces are widely found in nature, such as plant pollens and microorganisms (i.e., neutrophils). These wrinkles with their significantly enlarged surface areas provide enhanced survival tools, including pollen adhesion and hydration and cell signaling [64,65]. Inspired by these irregular morphologies, a study by Li et al. prepared wrinkled particles that cells readily attached, climbed, and conformed onto, which was not observed on the smooth particles [50]. Cell attachment to the surface of the particles was observed with actin networks appearing at the particle edges [50]. These findings provide explanation for our observations of Cela MPs performing significantly better than free Cela. Furthermore, the toxicity study of the blank MPs was evaluated for MSTO-211H cells and NHLF. The blank MPs displayed non-significant cell toxicity, suggesting the cytotoxic effects of Cela MPs are due to Cela encapsulation into the carrier system, not from other formulation components. The antioxidant activity of Cela has been reported in various diseases [66,67,68]. Previous studies have demonstrated that Cela enhanced the activity of enzymatic antioxidants (superoxide dismutase (SOD), catalase (CAT), glutathione peroxidases (GPx), and glutathione reductase (GR)) in bleomycin-induced pulmonary fibrotic rats [69]. Another study by Wang et al. found that Cela markedly increased the activities of SOD and GPx while also decreasing levels of reactive oxygen species and MDA in obese rat models [70]. In order to evaluate the antioxidant activity of Cela MP, a DPPH radical assay was performed to determine the capacity of the drug/formulation to react with free radicals [71]. After 30 min, Cela MP showed a similar radical scavenging effect to Cela only. However, after 24 h, the antioxidant activity of Cela MP was significantly greater than Cela. This can be explained by the ability of the microparticles to encapsulate and protect Cela from the DPPH solution, thereby keeping the drug stable and sustaining its release from 30 min to 24 h while continually scavenging DPPH free radicals. Caspase-3 is a cysteine protease that plays a vital role in the terminal course of programmed cell death (apoptosis). As a proteolytic enzyme, caspase-3 is responsible for the cleavage of DEVD peptide, poly(ADP-ribose) polymerase (PARP), DNA-dependent protein kinase, etc. [ 72,73]. Failure to activate apoptotic pathways in response to drug treatment may lead to drug resistance in tumor cells. In this study, 1.0 µM Cela MP was capable of inducing high levels of Caspase-3, indicating the induction of apoptosis in cancer cells. Conversely, autophagy is a cellular degradation process that occurs under stressful conditions in adaptation to starvation, development, cell death, and tumor suppression [74,75]. Autophagy is mediated by the formation of autophagosomes that collect degraded components and then fuse with lysosomes to be recycled [76]. The modulation of autophagy plays a role in both tumor suppression and promotion of many cancers, including MPM [4,76,77,78]. This study found that 1.0 µM Cela MP triggered autophagy by observing elevated fluorescent levels of LC3; however, plain Cela did not show any significant difference as compared to the control. A previous study by Liu et al. confirmed these findings, where Cela increased the formation of autophagosomes and accumulation of the LC3B-2 protein in glioma cells [79]. Li et al. also found that Cela increased levels of LC3B-2 in human osteosarcoma cells, and further observed that autophagy mediated by Cela promoted a pro-death function in cancer cells [35]. While the in vitro assays performed provided great insight into the efficacy of Cela MPs, they were likely inadequate in fully predicting the preclinical behavior of MPs. Tumors in the human body exist as a solid mass of cells that proliferate uncontrollably and grow exponentially into a three-dimensional structure. The in vitro monolayer studies lack the behavioral properties of a solid tumor, including cell-cell interactions, cellular heterogeneity, spatial architecture, and establishment of unique gene expression patterns [52,80]. Therefore, the efficacy of Cela MP was evaluated using a 3D spheroid model to mimic in vivo conditions. This study acts to bridge the gap between in vitro and in vivo studies. Spheroid volume analysis results found that Cela MP exhibited superior anti-tumor activity as compared to Cela only. Interestingly, a single-dose of Cela MP at 1.0 µM was sufficient to entirely inhibit spheroid growth after the 15-day period. To corroborate these results, we performed a cell viability assay and a live/dead cell fluorescence assay at the end of treatment period to further understand how Cela MPs suppressed spheroid growth. The results from both assays confirmed that a single dose of 1.0 µM Cela MP was effective in suppressing spheroid growth as compared to free Cela. In conclusion, results from the current study illustrate the great potential of Cela MP for the treatment of MPM. ## 4.1. Materials & Cell Lines Celastrol (Cela) was purchased from Adooq Bioscience (Irvine, CA, USA); Resomer® RG 502H (Poly(D,L-lactide-co-glycolide 50:50) (PLGA; MW 7000–17,000 Da, acid terminated) was purchased from Evonik Corporation (Piscataway, NJ, USA). Branched polyethyleneimine (PEI; average MW ~25,000 Da), Poly(vinyl alcohol) (PVA) (Mw 13,000–23,000 Da, 87–$89\%$ hydrolyzed), and 2,2-Diphenyl-1-picrylhydrazyl (DPPH) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Sodium chloride (NaCl), sodium bicarbonate (NaHCO₃), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dichloromethane (DCM), dimethyl sulfoxide (DMSO), acetonitrile (ACN), other UPLC grade chemicals, and LC-MS grade water were purchased from Fisher Scientific (Hampton, NH, USA). Molecular biology kits and supplies were acquired from other commercial vendors as listed at suitable places throughout the manuscript. Three immortalized malignant pleural mesothelioma cell lines, MSTO-211H, H28, and H2452 were procured from American Type Culture Collection (ATCC; Manassas, VA, USA). One primary patient-derived mesothelioma cell line, ROB, was obtained from Dr. Raffit Hassan at the National Cancer Institute (Bethesda, MD, USA) under an executed material transfer agreement and approved IRB protocol # 0818-034 from St. John’s University. This primary cell line was isolated from neoplastic effusions of patients undergoing therapeutic paracenteses, originally provided by The Stehlin Foundation (Houston, TX, USA) to the NIH, and were established as ROB in a study by Li et al. [ 81]. Normal human lung fibroblasts (NHLF) were obtained from Lonza (Basel, Switzerland). All mesothelioma cell lines were cultured in RPMI-1640 medium supplemented with $10\%$ fetal bovine serum (FBS) (R&D Systems, Minneapolis, MN, USA), $1\%$ sodium pyruvate, and $1\%$ penicillin/streptomycin (Corning, NY, USA). NHLF cells were cultured in FBMTM basal medium supplemented with fibroblast growth medium (FGMTM)-2 SingleQuotsTM supplements from Lonza. All cell lines were incubated at 37 °C/$5\%$ CO2 and 90–$100\%$ relative humidity. ## 4.2. UPLC Method for Celastrol (Cela) Content A reverse-phase liquid chromatography technique was established for measuring Cela content using a Waters Acquity series UPLC (Waters, Milford, MA, USA). The column used was an XBridge® BEH Shield RP18 2.5µm (3.0 × 100 mm) (Waters, Milford, MA, USA). The mobile phase was $0.1\%$ orthophosphoric acid (OPA) in HPLC grade water: acetonitrile (ACN) at a 10:90 ratio with a flow rate of 0.8 mL/min. The wavelength used for detection was 425 nm and data were collected and processed using EMPOWER 3.0 software (Waters, Milford, MA, USA). A representative UPLC chromatogram of Cela is shown in Figure S1. ## 4.3. Formulation of Celastrol Microparticles (Cela MPs) Cela MPs were fabricated using a double emulsion solvent evaporation method previously published with some modifications [71]. Specifically, $1\%$ NaCl (porogen) was dissolved in 0.5 mL of milli-Q water as the inner aqueous phase. The organic phase consisted of 2 mg of Cela and 60 mg of PLGA dissolved in 20 µL of DMSO and 3 mL of DCM. The primary emulsion was formed by probe sonicating (QSONICA-Q500, QSonica, Newtown, CT, USA) the inner aqueous phase with the organic phase for 1 min at $20\%$ amplitude and 10 s on-off cycles. The primary emulsion was added to 10 mL of a $1\%$ w/v PVA solution and homogenized (Homogenizer 850, Fisher Scientific, Hampton, NH, USA) for 1 min at 8000 rpm to obtain the double emulsion (w/o/w), followed by overnight stirring to remove DCM. The next day, smaller particles remaining in the supernatant were removed by centrifugation at 3000 rpm for 3 min. The remaining pellet containing the MPs were reconstituted in milli-Q water and washed three times by centrifugation at 21,000× g for 15 min. The final formulation was reconstituted in 1 mL of water. F1-F6 formulations were synthesized with minor alterations, including various polymers included in the inner aqueous phase (cationic polymer PEI and porogen NaHCO₃) and stabilizer PVA concentration. A summary of the formulation details is provided in Table 1. ## 4.4.1. Physiochemical Characterization: Particle Size (PS), Polydispersity Index (PDI), and zeta Potential A total of 20 µL of the formulation samples were separately diluted by 150× in HPLC water. Particle size (PS), polydispersity index (PDI), and zeta potential were measured using a Malvern Zetasizer (Malvern Panalytical Instruments Ltd., Malvern, UK). ## 4.4.2. Drug Content The percent encapsulation efficiency (%EE) and percent drug loading (%DL) of Cela in MPs were determined using a MP lysis method. Briefly, 20 µL of the formulation was lysed using a mixture of 10 µL DMSO, 20 µL DCM, and 1950 µL ACN, followed by centrifugation for 45 min at 21,191× g (4 °C). The supernatant containing free Cela was injected into the UPLC for analysis using the method defined in Section 2.2. The %EE and %DL were calculated based on Equations [1] and [2] respectively:[1]%EE=Drug entrapped in the NPsInitial drug added×$100\%$ [2]%DL=Drug entrapped in the NPsAmount of PLGA+ Initial Drug added×$100\%$ ## 4.4.3. Morphological Analysis The morphology of Cela MPs was examined using Helios Nano Lab 660 (FEI, Hillsboro, Oregon, USA). The microparticle suspension was dried on the SEM pin stub (Ted Pella, Inc., Redding, CA, USA) and then sputter coated with gold for 90 s and imaged at 5 kV. ## 4.4.4. In Vitro Release Study Samples for an in vitro release study were prepared by diluting 50 µL Cela MPs with 950 µL of $1\%$ sodium lauryl sulfate (SLS) in Simulated Lung Fluid (SLF), which was prepared according to the components in Gamble’s Solution [82]. The samples were shaken and incubated at 200 rpm and 37 °C with an incubating orbital shaker. At each time point between 0.5–72 h, one sample was removed and centrifuged at 17,500× g for 15 min. The supernatant was then analyzed by UPLC, as described in Section 2.2. ## 4.4.5. Differential Scanning Calorimetry (DSC) Studies Various samples were first lyophilized into dry powders using a Labconco FreeZone® freeze dryer system. DSC analyses were performed on a DSC 6000 (PerkinElmer, Inc; Waltham, MA, USA), using samples of 2–5 mg inside sealed pans. The samples were heated in sealed aluminum pans at a rate of 10 °C/min from 25–250 °C under a nitrogen flow rate of 50 mL/min. A blank sealed aluminum pan was used as the reference. ## 4.4.6. X-ray Diffraction Studies (XRD) XRD analysis was performed using XRD-6000 (Shimadzu, Kyoto, Japan). The diffraction patterns were measured with a voltage of 40 kV and a current of 30 mA. The freeze-dried samples were uniformly spread on a glass sample holder and analyzed over a range of 2θ or 10–60° at a scan speed of 2 °/min. ## 4.5. In Vitro Stability Study Triplicate samples of Cela MP suspensions were stored at 4 °C for a 4-week period. A total of 20 µL aliquots of each sample were withdrawn each week to determine the PS, PDI, and zeta potential by a Malvern Zetasizer as previously described. The %EE was analyzed by UPLC. ## 4.6. Aerosolization, Aerodynamic Properties, and Inhalability of Cela MPs The in vitro aerodynamic assessment of Cela MPs was performed using the Copley® 170 Next Generation ImpactorTM (NGI, MSP Corporation, Shoreview, MN, USA), following a previous report [58]. Briefly, the NGI plates were refrigerated for 2 h at 4 °C to prevent thermal transfer within the cascade impactor [52]. A total of 2 mL of Cela MP suspension was loaded into the nebulizer cup, which was attached to a PARI® LC plus nebulizer (PARI Respiratory Equipment, Midlothian, VA, USA). The vacuum pump (MSP Corp, Shoreview, MN, USA) was set at a flow rate of 15 L/min for 4 min as the sample travelled into the NGI. After the run, ACN was used to wash and collect the samples from each stage plate, followed by centrifugation at 21,191× g for 45 min to lyse the MPs. The resulting supernatant was analyzed using the UPLC method as mentioned in Section 2.2. PS, PDI, and zeta potential of Cela MPs were evaluated after nebulization to test whether the MPs maintained stability. The Fine Particle Fraction (FPF, %) was calculated as the ratio of fine particle dose (dae < 5.39 μm or amount of drug deposited from stage 3–8) to the total emitted dose (ED) (amount of drug emitted from mouthpiece to stage 8) deposited in the NGI. The mass median aerodynamic diameter (MMAD, D$50\%$) was determined by acquiring the diameter corresponding to $50\%$ of the cumulative deposition. The geometric standard deviation (GSD) was calculated using Equation [3] as shown below:[3]GSD=D$84.1\%$D$15.9\%$ ## 4.7. In Vitro Cell Culture Studies for the Determination of Anti-Cancer Efficacy The present work was carried out using four malignant pleural mesothelioma cell lines, MSTO-211H, H28, H2452, and primary ROB cells [4,83]. These cells were chosen as ideal models for mesothelioma, as they were derived directly from patient tumors, and therefore accurately represented in vivo tumor cells. ## Cytotoxicity Studies Cell viability studies for Cela and Cela MP were evaluated against MSTO-211H, H28, and H2452 cells, in addition to using a primary patient-derived cell line ROB using an MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) assay, following the methods published in our earlier studies [84,85]. Detailed methods are provided in Supplementary Information. IC50 values were calculated using the non-linear fitting module available in GraphPad Prism software (Version 6.0 for Windows, GraphPad Software, CA, USA). To further evaluate the toxicity of blank MPs, toxicity studies were performed in MSTO-211H and NHLF cell lines at corresponding celastrol concentrations (0.10–6.25 µM) after a 48-h incubation period. ## 4.8.1. DPPH Antioxidant Assay The potent antioxidant and anti-inflammatory activity of Cela has been shown to be useful to treat a multitude of diseases such as cancers (e.g., glioma, ovarian cancer, and colorectal cancer) [79,86,87,88,89]. 2,2-Diphenyl-1-picrylhydrazyl (DPPH) is a free radical that produces a violet solution once dissolved in methanol, and in the presence of an antioxidant, the free radical reduces to produce a colorless solution. Therefore, the DPPH assay was selected to evaluate the antioxidant capacity, or free radical scavenging ability, of Cela and Cela MP based on a previously described method [71]. Detailed methods are provided in Supplementary Information. ## 4.8.2. Caspase-3 Assay To evaluate the apoptotic properties of Cela and Cela MPs, the apoptosis mediator (caspase-3) activity was determined using an EnzChek Caspase-3 assay kit (Thermo Fisher Scientific, Waltham, MA, USA). MSTO-211H cells were seeded at a density of 5 × 105 cells/TC dish (100 mm diameter) (Thermo Scientific, Rochester, NY, USA). Treatments of Cela and Cela MP (0.5 and 1.0 μM) were incubated for 6 h, after which cells were harvested and washed with sterile PBS, followed by resuspending in 1X cell-lysis buffer and centrifugation. A total of 50 μL of the supernatant and 50 μL of a 2X substrate working solution (10 mM Z-DEVD-AMC substrate + 2X reaction buffer) were added to a 96-well plate while using 50 μL of lysis buffer, and a substrate working solution was added as a no enzyme control. The fluorescence was measured after incubating for 20 min at excitation/emission of $\frac{360}{460}$ nm. ## 4.8.3. Effect of Cela and Cela MP on Cellular Autophagy The CYTO-ID® Autophagy Detection Kit (Enzo Life Sciences, Farmingdale, NY, USA) was used to determine the effect of Cela and Cela MP on autophagy in MPM cells by following a previously published protocol [4]. MSTO-211H cells were used for this study. Detailed methods are provided in Supplementary Information. ## 4.9. Determination of Cela MP Efficacy in 3D Tumor Spheroid Model The anti-cancer efficacy of Cela MPs was further evaluated by determining the penetrability and efficacy of nanoparticles into solid tumors. As previous reports have indicated, 3D spheroid cell culture studies have been used in an attempt to bridge the gap between in vitro and in vivo systems by partially representing the tumor structure and microenvironment [90,91,92]. Our earlier study has suggested the feasibility of MSTO-211H 3D spheroids in testing efficacy of various therapeutics [93]. Briefly, MSTO-211H cells were seeded in a Corning® ultra-low attachment spheroid 96-well plate (Corning, NY, USA) at a density of 2.0 × 103 cells/well and incubated at 37 °C/$5\%$ CO2 for 3 days to allow spheroids to grow into a solid tumor mass, following which treatments were started (Day 0 for treatment). Totals of 0.5 and 1.0 µM celastrol concentrations were selected for treatment. For a single dose study, half the media (100 µL) was replaced with fresh media on each imaging day (except for Day 0). For the multiple dose study, half of the media was replaced with respective concentrations of Cela or Cela MP or fresh media (control). Images of the spheroids were taken using an inverted microscope (10× magnification, Laxco LMI-6000, Laxco Inc., Mill Creek, WA, USA) on days 0, 3, 6, 9, 12, and 15. NIH ImageJ software (Version 1.44) was used to measure the diameter of the spheroids, which was then used to calculate the spheroid volumes. ## 4.9.1. Cell Titer-Glo Cell Viability Study At the end of Day 15, cell viability within the spheroid core was assessed using Cell Titer-Glo® kit (Promega, Madison, WI, USA). Briefly, 100 µL of treatment was removed and replaced with 100 µL of CellTiter-Glo® reagent in each well. The contents were mixed for a few minutes, followed by incubation at room temperature for 30 min. Luminescence was measured using a Spark 10M plate reader (Tecan, Männedorf, Switzerland). ## 4.9.2. Live/Dead Cell Assay A Live/Dead cell assay was performed using a fluorescent Viability/Cytotoxicity assay kit (Biotium, Fremont, CA, USA) to visualize the live and dead cells on the surface of the spheroids. At day 15, the treatments were completely removed from each well and replaced with 100 µL of 2 µM calcein AM/4 µM EthD-III staining solution. The plate was then incubated in the dark at room temperature for 30 min and subsequently imaged on an EVOS-FL Cell Imaging fluorescence microscopy at 4X magnification (Thermo Fisher Scientific, Waltham, MA, USA). Green fluorescent protein (GFP, imaging calcein AM) is representative of viable cells, whereas red fluorescent protein (RFP, imaging EthD-III) indicates necrotic/dead cells. Fluorescent intensity was analyzed by NIH ImageJ software (Version 1.53c). ## 4.10. Statistical Analyses All data were addressed as mean ± SD or SEM, with $$n = 3$$ unless otherwise mentioned. At least three trials of cytotoxicity studies were performed for each control or treatment with $$n = 6$$ for each trial. All data were evaluated by an unpaired student’s t-test or one-way ANOVA followed by Tukey’s multiple comparisons test using GraphPad Prism software (Version 6.0 for Windows, GraphPad Inc., San Diego, CA, USA). A p value of <0.05 was considered statistically significant and was presented in data figures as a single asterisk (). However, some studies demonstrated a smaller p-value of 0.01 or less, which is indicated at respective places. ## 5. Conclusions In this study, surface-wrinkled Cela-loaded microparticles were developed with a suitable aerodynamic diameter for efficient deep lung delivery. Cela MP was able to retain the antioxidant activity of the drug, trigger autophagy, and induce apoptosis. These mechanisms, in combination with an efficient delivery system, resulted in significantly enhanced efficacy against MPM in both 2D cell culture and a 3D tumor spheroid model. This study lays the groundwork to pursue the development of Celastrol and Celastrol-encapsulated delivery systems in preclinical and potentially clinical settings. These findings also suggest that the successful encapsulation of Cela in polymeric microparticles addresses certain limitations of Cela, such as low aqueous solubility and potential toxicity at higher doses. Going forward, we plan to study the benefits of Cela MPs in vivo and evaluate their long-term safety and stability profiles. In conclusion, inhaled Cela MPs are a promising anti-cancer standalone therapy for MPM treatment. ## References 1. Berzenji L., Van Schil P.. Multimodality treatment of malignant pleural mesothelioma. F1000Research* (2018.0) 7 1681. DOI: 10.12688/f1000research.15796.1 2. Zhai Z., Ruan J., Zheng Y., Xiang D., Li N., Hu J., Shen J., Deng Y., Yao J., Zhao P.. Assessment of Global Trends in the Diagnosis of Mesothelioma From 1990 to 2007. JAMA Netw. Open (2021.0) 4 e2120360. DOI: 10.1001/jamanetworkopen.2021.20360 3. Malignant Mesothelioma Cancer\|Stages, Prognosis, Treatment [Internet] 4. Kulkarni N., Vaidya B., Parvathaneni V., Bhanja D., Gupta V.. Repurposing Quinacrine for Treatment of Malignant Mesothelioma: In-Vitro Therapeutic and Mechanistic Evaluation. Int. J. Mol. Sci. (2020.0) 21. DOI: 10.3390/ijms21176306 5. Key Statistics About Malignant Mesothelioma [Internet] 6. Survival Rates for Mesothelioma [Internet] 7. Lim H.Y., Ong P.S., Wang L., Goel A., Ding L., Wong A.L.-A., Ho P.C.-L., Sethi G., Xiang X., Goh B.C.. Celastrol in cancer therapy: Recent developments, challenges and prospects. Cancer Lett. (2021.0) 521 252-267. DOI: 10.1016/j.canlet.2021.08.030 8. Wang Y., Li C., Gu J., Chen C., Duanmu J., Miao J., Yao W., Tao J., Tu M., Xiong B.. Celastrol exerts anti-inflammatory effect in liver fibrosis via activation of AMPK-SIRT3 signalling. J. Cell. Mol. Med. (2019.0) 24 941-953. DOI: 10.1111/jcmm.14805 9. Pace S., Zhang K., Jordan P.M., Bilancia R., Wang W., Börner F., Hofstetter R.K., Potenza M., Kretzer C., Gerstmeier J.. Anti-inflammatory celastrol promotes a switch from leukotriene biosynthesis to formation of specialized pro-resolving lipid mediators. Pharmacol. Res. (2021.0) 167 105556. DOI: 10.1016/j.phrs.2021.105556 10. Zhang C., Zhao M., Wang B., Su Z., Guo B., Qin L., Zhang W., Zheng R.. The Nrf2-NLRP3-caspase-1 axis mediates the neuroprotective effects of Celastrol in Parkinson’s disease. Redox Biol. (2021.0) 47 102134. DOI: 10.1016/j.redox.2021.102134 11. Lin M.-W., Lin C.C., Chen Y.-H., Yang H.-B., Hung S.-Y.. Celastrol Inhibits Dopaminergic Neuronal Death of Parkinson’s Disease through Activating Mitophagy. Antioxidants (2019.0) 9. DOI: 10.3390/antiox9010037 12. Shukla S.K., Chan A., Parvathaneni V., Kanabar D.D., Patel K., Ayehunie S., Gupta V.. Enhanced solubility, stability, permeation and anti-cancer efficacy of Celastrol-β-cyclodextrin inclusion complex. J. Mol. Liq. (2020.0) 318 113936. DOI: 10.1016/j.molliq.2020.113936 13. Chen X., Zhao Y., Luo W., Chen S., Lin F., Zhang X., Fan S., Shen X., Wang Y., Liang G.. Celastrol induces ROS-mediated apoptosis via directly targeting peroxiredoxin-2 in gastric cancer cells. Theranostics (2020.0) 10 10290-10308. DOI: 10.7150/thno.46728 14. Shen B., Chen H.-B., Zhou H.-G., Wu M.-H.. Celastrol induces caspase-dependent apoptosis of hepatocellular carcinoma cells by suppression of mammalian target of rapamycin. J. Tradit. Chin. Med. (2021.0) 41 381-389. DOI: 10.19852/j.cnki.jtcm.2021.03.006 15. Zhang J., Wang R., Cheng L., Xu H.. Celastrol inhibit the proliferation, invasion and migration of human cervical HeLa cancer cells through down-regulation of MMP-2 and MMP-9. J. Cell Mol. Med. (2021.0) 25 5335-5338. DOI: 10.1111/jcmm.16488 16. Zhang W., Wu Z., Qi H., Chen L., Wang T., Mao X., Shi H., Chen H., Zhong M., Shi X.. Celastrol upregulated ATG7 triggers autophagy via targeting Nur77 in colorectal cancer. Phytomedicine (2022.0) 104. DOI: 10.1016/j.phymed.2022.154280 17. Cai Z., Qian B., Pang J., Tan Z.-B., Zhao K., Lei T.. Celastrol Induces Apoptosis and Autophagy via the AKT/mTOR Signaling Pathway in the Pituitary ACTH-secreting Adenoma Cells. Curr. Med Sci. (2022.0) 42 387-396. DOI: 10.1007/s11596-022-2568-6 18. Wang X., Wan W., Lu J., Zhang Y., Quan G., Pan X., Liu P.. Inhalable cryptotanshinone spray-dried swellable microparticles for pulmonary fibrosis therapy by regulating TGF-β1/Smad3, STAT3 and SIRT3 pathways. Eur. J. Pharm. Biopharm. (2022.0) 172 177-1792. DOI: 10.1016/j.ejpb.2022.02.012 19. Alipour S., Montaseri H., Tafaghodi M.. Inhalable, large porous PLGA microparticles loaded with paclitaxel: Preparation,. J. Microencapsul. (2015.0) 32 661-668. DOI: 10.3109/02652048.2014.944949 20. Sheth P., Stein S.W., Myrdal P.B.. Factors Influencing Aerodynamic Particle Size Distribution of Suspension Pressurized Metered Dose Inhalers. AAPS PharmSciTech (2014.0) 16 192-201. DOI: 10.1208/s12249-014-0210-z 21. Ngan C.L., Asmawi A.A.. Lipid-based pulmonary delivery system: A review and future considerations of formulation strategies and limitations. Drug Deliv. Transl. Res. (2018.0) 8 1527-1544. DOI: 10.1007/s13346-018-0550-4 22. Wang C., Yang J., Han H., Chen J., Wang Y., Li Q., Wang Y.. Disulfiram-loaded porous PLGA microparticle for inhibiting the proliferation and migration of non-small-cell lung cancer. Int. J. Nanomed. (2017.0) 12 827-837. DOI: 10.2147/IJN.S121948 23. Hu Y., Li M., Zhang M., Jin Y.. Inhalation treatment of idiopathic pulmonary fibrosis with curcumin large porous microparticles. Int. J. Pharm. (2018.0) 551 212-222. DOI: 10.1016/j.ijpharm.2018.09.031 24. Zhang X., Qin L., Su J., Sun Y., Zhang L., Li J., Beck-Broichsitter M., Muenster U., Chen L., Mao S.. Engineering large porous microparticles with tailored porosity and sustained drug release behavior for inhalation. Eur. J. Pharm. Biopharm. (2020.0) 155 139-146. DOI: 10.1016/j.ejpb.2020.08.021 25. Ungaro F., Bianca R.D.D.V., Giovino C., Miro A., Sorrentino R., Quaglia F., La Rotonda M.I.. Insulin-loaded PLGA/cyclodextrin large porous particles with improved aerosolization properties: In vivo deposition and hypoglycaemic activity after delivery to rat lungs. J. Control. Release (2009.0) 135 25-34. DOI: 10.1016/j.jconrel.2008.12.011 26. Oh Y.J., Lee J., Seo J.Y., Rhim T., Kim S.-H., Yoon H.J., Lee K.Y.. Preparation of budesonide-loaded porous PLGA microparticles and their therapeutic efficacy in a murine asthma model. J. Control. Release (2011.0) 150 56-62. DOI: 10.1016/j.jconrel.2010.11.001 27. Kim I., Byeon H.J., Kim T.H., Lee E.S., Oh K.T., Shin B.S., Lee K.C., Youn Y.S.. Doxorubicin-loaded highly porous large PLGA microparticles as a sustained- release inhalation system for the treatment of metastatic lung cancer. Biomaterials (2012.0) 33 5574-5583. DOI: 10.1016/j.biomaterials.2012.04.018 28. Shi X., Li C., Gao S., Zhang L., Han H., Zhang J., Shi W., Li Q.. Combination of doxorubicin-based chemotherapy and polyethylenimine/p53 gene therapy for the treatment of lung cancer using porous PLGA microparticles. Colloids Surf. B Biointerfaces (2014.0) 122 498-504. DOI: 10.1016/j.colsurfb.2014.07.020 29. Wei Y., Wang Y., Zhang H., Zhou W., Ma G.. A novel strategy for the preparation of porous microspheres and its application in peptide drug loading. J. Colloid Interface Sci. (2016.0) 478 46-53. DOI: 10.1016/j.jcis.2016.05.045 30. Gallo L., Ramírez-Rigo M.V., Bucalá V.. Development of porous spray-dried inhalable particles using an organic solvent-free technique. Powder Technol. (2018.0) 342 642-652. DOI: 10.1016/j.powtec.2018.10.041 31. Li P., Zhou X., Qu D., Guo M., Fan C., Zhou T., Ling Y.. Preliminary study on fabrication, characterization and synergistic anti-lung cancer effects of self-assembled micelles of covalently conjugated celastrol–polyethylene glycol–ginsenoside Rh2. Drug Deliv. (2017.0) 24 834-845. DOI: 10.1080/10717544.2017.1326540 32. Ayyoob M., Kim Y.J.. Effect of Chemical Composition Variant and Oxygen Plasma Treatments on the Wettability of PLGA Thin Films, Synthesized by Direct Copolycondensation. Polymers (2018.0) 10. DOI: 10.3390/polym10101132 33. Deng Y.-N., Shi J., Liu J., Qu Q.-M.. Celastrol protects human neuroblastoma SH-SY5Y cells from rotenone-induced injury through induction of autophagy. Neurochem. Int. (2013.0) 63 1-9. DOI: 10.1016/j.neuint.2013.04.005 34. Lee H.-W., Jang K.S.B., Choi H.J., Jo A., Cheong J.-H., Chun K.-H.. Celastrol inhibits gastric cancer growth by induction of apoptosis and autophagy. BMB Rep. (2014.0) 47 697-702. DOI: 10.5483/BMBRep.2014.47.12.069 35. Li H.-Y., Zhang J., Sun L.-L., Li B.-H., Gao H.-L., Xie T., Zhang N., Ye Z.-M.. Celastrol induces apoptosis and autophagy via the ROS/JNK signaling pathway in human osteosarcoma cells: An in vitro and in vivo study. Cell Death Dis. (2015.0) 6 e1604. DOI: 10.1038/cddis.2014.543 36. Kulkarni N.S., Parvathaneni V., Shukla S.K., Barasa L., Perron J.C., Yoganathan S., Muth A., Gupta V.. Tyrosine kinase inhibitor conjugated quantum dots for non-small cell lung cancer (NSCLC) treatment. Eur. J. Pharm. Sci. (2019.0) 133 145-159. DOI: 10.1016/j.ejps.2019.03.026 37. Shukla S.K., Kulkarni N.S., Chan A., Parvathaneni V., Farrales P., Muth A., Gupta V.. Metformin-Encapsulated Liposome Delivery System: An Effective Treatment Approach against Breast Cancer. Pharmaceutics (2019.0) 11. DOI: 10.3390/pharmaceutics11110559 38. Shi J., Li J., Xu Z., Chen L., Luo R., Zhang C., Gao F., Zhang J., Fu C.. Celastrol: A Review of Useful Strategies Overcoming its Limitation in Anticancer Application. Front. Pharmacol. (2020.0) 11 558741. DOI: 10.3389/fphar.2020.558741 39. Cascão R., Fonseca J.E., Moita L.F.. Celastrol: A Spectrum of Treatment Opportunities in Chronic Diseases. Front. Med. (2017.0) 4 69. DOI: 10.3389/fmed.2017.00069 40. Kashyap D., Sharma A., Tuli H.S., Sak K., Mukherjee T., Bishayee A.. Molecular targets of celastrol in cancer: Recent trends and advancements. Crit. Rev. Oncol. (2018.0) 128 70-81. DOI: 10.1016/j.critrevonc.2018.05.019 41. Cascão R., Carvalho T., Goncalves J., Moita L., Fonseca J.. AB0096 Efficacy and safety of oral administration of pure celastrol in aia rats. Ann. Rheum. Dis. (2017.0) 76 1080. DOI: 10.1136/annrheumdis-2017-eular.2868 42. Hou W., Liu B., Xu H.. Celastrol: Progresses in structure-modifications, structure-activity relationships, pharmacology and toxicology. Eur. J. Med. Chem. (2020.0) 189 112081. DOI: 10.1016/j.ejmech.2020.112081 43. Thakur P., Sonawane S.S., Sonawane S.H., Bhanvase B.A.. Nanofluids-based delivery system, encapsulation of nanoparticles for stability to make stable nanofluids. Encapsulation of Active Molecules and Their Delivery System [Internet] (2020.0) 141 44. Heijerman H., Westerman E., Conway S., Touw D.. Inhaled medication and inhalation devices for lung disease in patients with cystic fibrosis: A European consensus. J. Cyst. Fibros. (2009.0) 8 295-315. DOI: 10.1016/j.jcf.2009.04.005 45. Champion J.A., Walker A., Mitragotri S.. Role of Particle Size in Phagocytosis of Polymeric Microspheres. Pharm. Res. (2008.0) 25 1815-1821. DOI: 10.1007/s11095-008-9562-y 46. Yang Y., Bajaj N., Xu P., Ohn K., Tsifansky M.D., Yeo Y.. Development of highly porous large PLGA microparticles for pulmonary drug delivery. Biomaterials (2009.0) 30 1947-1953. DOI: 10.1016/j.biomaterials.2008.12.044 47. Gupta M., Aina A., Boukari Y., Doughty S., Morris A., Billa N.. Effect of volume of porogens on the porosity of PLGA scaffolds in pH-controlled environment. Pharm. Dev. Technol. (2018.0) 23 207-210. DOI: 10.1080/10837450.2017.1304415 48. Li B., Cao Y.-P., Feng X.-Q., Gao H.. Mechanics of morphological instabilities and surface wrinkling in soft materials: A review. Soft Matter (2012.0) 8 5728-5745. DOI: 10.1039/c2sm00011c 49. Lechanteur A., Evrard B.. Influence of Composition and Spray-Drying Process Parameters on Carrier-Free DPI Properties and Behaviors in the Lung: A review. Pharmaceutics (2020.0) 12. DOI: 10.3390/pharmaceutics12010055 50. Li M., Joung D., Hughes B., Waldman S.D., Kozinski J.A., Hwang D.K.. Wrinkling Non-Spherical Particles and Its Application in Cell Attachment Promotion. Sci. Rep. (2016.0) 6 30463. DOI: 10.1038/srep30463 51. Niu Z., Tedesco E., Benetti F., Mabondzo A., Montagner I.M., Marigo I., Gonzalez-Touceda D., Tovar S., Diéguez C., Santander-Ortega M.J.. Rational design of polyarginine nanocapsules intended to help peptides overcoming intestinal barriers. J. Control. Release (2017.0) 263 4-17. DOI: 10.1016/j.jconrel.2017.02.024 52. Shukla S.K., Kulkarni N.S., Farrales P., Kanabar D.D., Parvathaneni V., Kunda N.K., Muth A., Gupta V.. Sorafenib Loaded Inhalable Polymeric Nanocarriers against Non-Small Cell Lung Cancer. Pharm. Res. (2020.0) 37 1-19. DOI: 10.1007/s11095-020-02790-3 53. Kamaly N., Yameen B., Wu J., Farokhzad O.C.. Degradable Controlled-Release Polymers and Polymeric Nanoparticles: Mechanisms of Controlling Drug Release. Chem. Rev. (2016.0) 116 2602-2663. DOI: 10.1021/acs.chemrev.5b00346 54. Sanna V., Chamcheu J.C., Pala N., Mukhtar H., Sechi M., Siddiqui I.A.. Nanoencapsulation of natural triterpenoid celastrol for prostate cancer treatment. Int. J. Nanomed. (2015.0) ume 10 6835-6846. DOI: 10.2147/IJN.S93752 55. Zhan S., Paik A., Onyeabor F., Ding B., Prabhu S., Wang J.. Oral Bioavailability Evaluation of Celastrol-Encapsulated Silk Fibroin Nanoparticles Using an Optimized LC-MS/MS Method. Molecules (2020.0) 25. DOI: 10.3390/molecules25153422 56. Surber M.W., Beck S., Pham S., Marsden A.T., Gandi S.K., Baily J., McElroy M.C.. Inhaled nintedanib is well-tolerated and delivers key pharmacokinetic parameters required to treat bleomycin-induced pulmonary fibrosis. Pulm. Pharmacol. Ther. (2020.0) 63 101938. DOI: 10.1016/j.pupt.2020.101938 57. Parvathaneni V., Kulkarni N.S., Shukla S.K., Farrales P.T., Kunda N.K., Muth A., Gupta V.. Systematic Development and Optimization of Inhalable Pirfenidone Liposomes for Non-Small Cell Lung Cancer Treatment. Pharmaceutics (2020.0) 12. DOI: 10.3390/pharmaceutics12030206 58. Wang X., Parvathaneni V., Shukla S.K., Kanabar D.D., Muth A., Gupta V.. Cyclodextrin Complexation for Enhanced Stability and Non-invasive Pulmonary Delivery of Resveratrol—Applications in Non-small Cell Lung Cancer Treatment. AAPS PharmSciTech (2020.0) 21 1-14. DOI: 10.1208/s12249-020-01724-x 59. Elbatanony R.S., Parvathaneni V., Kulkarni N.S., Shukla S.K., Chauhan G., Kunda N.K., Gupta V.. Afatinib-loaded inhalable PLGA nanoparticles for localized therapy of non-small cell lung cancer (NSCLC)—development and in-vitro efficacy. Drug Deliv. Transl. Res. (2020.0) 11 927-943. DOI: 10.1007/s13346-020-00802-8 60. Pulivendala G., Bale S., Godugu C.. Inhalation of sustained release microparticles for the targeted treatment of respiratory diseases. Drug Deliv. Transl. Res. (2019.0) 10 339-353. DOI: 10.1007/s13346-019-00690-7 61. Chew N.Y.K., Chan H.-K.. Use of Solid Corrugated Particles to Enhance Powder Aerosol Performance. Pharm. Res. (2001.0) 18 1570-1577. DOI: 10.1023/A:1013082531394 62. Chan H.K.. What is the role of particle morphology in pharmaceutical powder aerosols?. Expert Opin. Drug. Deliv. (2008.0) 5 909-914. DOI: 10.1517/17425247.5.8.909 63. Chew N.Y., Chan H.-K.. The Role of Particle Properties in Pharmaceutical Powder Inhalation Formulations. J. Aerosol Med. (2002.0) 15 325-330. DOI: 10.1089/089426802760292672 64. Brasen J.C., Olsen L.F., Hallett M.B.. Cell surface topology creates high Ca. Cell Calcium (2010.0) 47 339-349. DOI: 10.1016/j.ceca.2010.01.005 65. Edlund A.F., Swanson R., Preuss D.. Pollen and Stigma Structure and Function: The Role of Diversity in Pollination. Plant Cell (2004.0) 16 S84-S97. DOI: 10.1105/tpc.015800 66. Jannuzzi A.T., Kara M., Alpertunga B.. Celastrol ameliorates acetaminophen-induced oxidative stress and cytotoxicity in HepG2 cells. Hum. Exp. Toxicol. (2017.0) 37 742-751. DOI: 10.1177/0960327117734622 67. Kurosawa R., Satoh K., Nakata T., Shindo T., Kikuchi N., Satoh T., Shimokawa H.. Identification of Celastrol as a Novel Therapeutic Agent for Pulmonary Arterial Hypertension and Right Ventricular Failure Through Suppression of Bsg (Basigin)/CyPA (Cyclophilin A). Arterioscler. Thromb. Vasc. Biol. (2021.0) 41 1205-1217. DOI: 10.1161/ATVBAHA.120.315731 68. Zhang B., Zhong Q., Chen X., Wu X., Sha R., Song G., Zhang C., Chen X.. Neuroprotective Effects of Celastrol on Transient Global Cerebral Ischemia Rats via Regulating HMGB1/NF-κB Signaling Pathway. Front. Neurosci. (2020.0) 14 847. DOI: 10.3389/fnins.2020.00847 69. Divya T., Dineshbabu V., Soumyakrishnan S., Sureshkumar A., Sudhandiran G.. Celastrol enhances Nrf2 mediated antioxidant enzymes and exhibits anti-fibrotic effect through regulation of collagen production against bleomycin-induced pulmonary fibrosis. Chem. Interactions (2016.0) 246 52-62. DOI: 10.1016/j.cbi.2016.01.006 70. Wang C., Shi C., Yang X., Yang M., Sun H., Wang C.. Celastrol suppresses obesity process via increasing antioxidant capacity and improving lipid metabolism. Eur. J. Pharmacol. (2014.0) 744 52-58. DOI: 10.1016/j.ejphar.2014.09.043 71. Wang X., Parvathaneni V., Shukla S.K., Kulkarni N.S., Muth A., Kunda N.K., Gupta V.. Inhalable resveratrol-cyclodextrin complex loaded biodegradable nanoparticles for enhanced efficacy against non-small cell lung cancer. Int. J. Biol. Macromol. (2020.0) 164 638-650. DOI: 10.1016/j.ijbiomac.2020.07.124 72. Los M.J., Mozoluk M., Ferrari D., Stepczynska A., Stroh C., Renz A., Herceg Z., Wang Z.-Q., Schulze-Osthoff K.. Activation and Caspase-mediated Inhibition of PARP: A Molecular Switch between Fibroblast Necrosis and Apoptosis in Death Receptor Signaling. Mol. Biol. Cell (2002.0) 13 978-988. DOI: 10.1091/mbc.01-05-0272 73. Zhang J., Wang X., Cui W., Wang W., Zhang H., Liu L., Zhang Z., Li Z., Ying G., Zhang N.. Visualization of caspase-3-like activity in cells using a genetically encoded fluorescent biosensor activated by protein cleavage. Nat. Commun. (2013.0) 4 2157. DOI: 10.1038/ncomms3157 74. Mizushima N.. Autophagy: Process and function. Genes Dev. (2007.0) 21 2861-2873. DOI: 10.1101/gad.1599207 75. Mizushima N.. The pleiotropic role of autophagy: From protein metabolism to bactericide. Cell Death Differ. (2005.0) 12 1535-1541. DOI: 10.1038/sj.cdd.4401728 76. Yun C.W., Lee S.H.. The Roles of Autophagy in Cancer. Int. J. Mol. Sci. (2018.0) 19. DOI: 10.3390/ijms19113466 77. Hwang K.-E., Kim Y.-S., Jung J.-W., Kwon S.-J., Park D.-S., Cha B.-K., Oh S.-H., Yoon K.-H., Jeong E.-T., Kim H.-R.. Inhibition of autophagy potentiates pemetrexed and simvastatin-induced apoptotic cell death in malignant mesothelioma and non-small cell lung cancer cells. Oncotarget (2015.0) 6 29482-29496. DOI: 10.18632/oncotarget.5022 78. Rosenfeldt M.T., Ryan K.M.. The multiple roles of autophagy in cancer. Carcinog. (2011.0) 32 955-963. DOI: 10.1093/carcin/bgr031 79. Liu X., Zhao P., Wang X., Wang L., Zhu Y., Song Y., Gao W.. Celastrol mediates autophagy and apoptosis via the ROS/JNK and Akt/mTOR signaling pathways in glioma cells. J. Exp. Clin. Cancer Res. CR (2019.0) 38 184. DOI: 10.1186/s13046-019-1173-4 80. Ferreira L., Gaspar V., Mano J.. Design of spherically structured 3D in vitro tumor models -Advances and prospects. Acta Biomater. (2018.0) 75 11-34. DOI: 10.1016/j.actbio.2018.05.034 81. Li Q., Verschraegen C.F., Mendoza J., Hassan R.. Cytotoxic activity of the recombinant anti-mesothelin immunotoxin, SS1(dsFv)PE38, towards tumor cell lines established from ascites of patients with peritoneal mesotheliomas. Anticancer. Res. (2004.0) 24 1327-1336. PMID: 15274292 82. Marques M.R.C., Loebenberg R., Almukainzi M.. Simulated Biological Fluids with Possible Application in Dissolution Testing. Dissolution Technol. (2011.0) 18 15-28. DOI: 10.14227/DT180311P15 83. Shukla S.K., Chan A., Parvathaneni V., Gupta V.. Metformin-loaded chitosomes for treatment of malignant pleural mesothelioma–A rare thoracic cancer. Int. J. Biol. Macromol. (2020.0) 160 128-141. DOI: 10.1016/j.ijbiomac.2020.05.146 84. Vaidya B., Kulkarni N.S., Shukla S.K., Parvathaneni V., Chauhan G., Damon J.K., Sarode A., Garcia J.V., Kunda N., Mitragotri S.. Development of inhalable quinacrine loaded bovine serum albumin modified cationic nanoparticles: Repurposing quinacrine for lung cancer therapeutics. Int. J. Pharm. (2020.0) 577 118995. DOI: 10.1016/j.ijpharm.2019.118995 85. Kulkarni N.S., Chauhan G., Goyal M., Sarvepalli S., Gupta V.. Development of gelatin methacrylate (GelMa) hydrogels for versatile intracavitary applications. Biomater. Sci. (2022.0) 10 4492-4507. DOI: 10.1039/D2BM00022A 86. Allison A.C., Cacabelos R., Lombardi V.R., Álvarez X.A., Vigo C.. Celastrol, a potent antioxidant and anti-inflammatory drug, as a possible treatment for Alzheimer’s disease. Prog. Neuro-Psychopharmacol. Biol. Psychiatry (2001.0) 25 1341-1357. DOI: 10.1016/S0278-5846(01)00192-0 87. You D., Jeong Y., Yoon S.Y., A Kim S., Kim S.W., Nam S.J., Kim S.. Celastrol attenuates the inflammatory response by inhibiting IL-1β expression in triple-negative breast cancer cells. Oncol. Rep. (2021.0) 45 89. DOI: 10.3892/or.2021.8040 88. Bufu T., Di X., Yilin Z., Gege L., Xi C., Ling W.. Celastrol inhibits colorectal cancer cell proliferation and migration through suppression of MMP3 and MMP7 by the PI3K/AKT signaling pathway. Anti-Cancer Drugs (2018.0) 29 530-538. DOI: 10.1097/CAD.0000000000000621 89. Yan X., Zhao N., Xu L., Ye P., Nan X., Zhou H., Shi Z.. Celastrol inhibits the growth of ovarian cancer cells in vitro and In Vivo. Gynecol. Oncol. (2019.0) 154 101. DOI: 10.1016/j.ygyno.2019.04.241 90. Brüningk S.C., Rivens I., Box C., Oelfke U., ter Haar G.. 3D tumour spheroids for the prediction of the effects of radiation and hyperthermia treatments. Sci. Rep. (2020.0) 10 1653. DOI: 10.1038/s41598-020-58569-4 91. Pinto B., Henriques A.C., Silva P.M.A., Bousbaa H.. Three-Dimensional Spheroids as In Vitro Preclinical Models for Cancer Research. Pharmaceutics (2020.0) 12. DOI: 10.3390/pharmaceutics12121186 92. Roper S.J., Coyle B.. Establishing an In Vitro 3D Spheroid Model to Study Medulloblastoma Drug Response and Tumor Dissemination. Curr. Protoc. (2022.0) 2 e357. DOI: 10.1002/cpz1.357 93. Parvathaneni V., Chilamakuri R., Kulkarni N.S., Wang X., Agarwal S., Gupta V.. Repurposing clofazimine for malignant pleural mesothelioma treatment — In-vitro assessment of efficacy and mechanism of action. Life Sci. (2022.0) 306 120843. DOI: 10.1016/j.lfs.2022.120843
--- title: Pilot Study on the Impact of Polymorphisms Linked to Multi-Kinase Inhibitor Metabolism on Lenvatinib Side Effects in Patients with Advanced Thyroid Cancer authors: - Silvia Cantara - Cristina Dalmiglio - Carlotta Marzocchi - Alfonso Sagnella - Lucia Brilli - Andrea Trimarchi - Fabio Maino - Laura Valerio - Maria Grazia Castagna journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049548 doi: 10.3390/ijms24065496 license: CC BY 4.0 --- # Pilot Study on the Impact of Polymorphisms Linked to Multi-Kinase Inhibitor Metabolism on Lenvatinib Side Effects in Patients with Advanced Thyroid Cancer ## Abstract Multi-kinase inhibitors (MKIs) represent the best therapeutic option in advanced thyroid cancer patients. The therapeutic efficacy and toxicity of MKIs are very heterogeneous and are difficult to predict before starting treatment. Moreover, due to the development of severe adverse events, it is necessary to interrupt the therapy some patients. Using a pharmacogenetic approach, we evaluated polymorphisms in genes coding for proteins involved with the absorption and elimination of the drug in 18 advanced thyroid cancer patients treated with lenvatinib, and correlated the genetic background with [1] diarrhea, nausea, vomiting and epigastric pain; [2] oral mucositis and xerostomia; [3] hypertension and proteinuria; [4] asthenia; [5] anorexia and weight loss; [6] hand foot syndrome. Analyzed variants belong to cytochrome P450 (CYP3A4 rs2242480 and rs2687116 and CYP3A5 rs776746) genes and to ATP-binding cassette transporters (ABCB1 rs1045642, rs2032582 and rs2235048 and ABCG2 rs2231142). Our results suggest that the GG genotype for rs2242480 in CYP3A4 and CC genotype in rs776746 for CYP3A5 were both associated with the presence of hypertension. Being heterozygous for SNPs in the ABCB1 gene (rs1045642 and 2235048) implicated a higher grade of weight loss. The ABCG2 rs2231142 statistically correlated with a higher extent of mucositis and xerostomia (CC genotype). Heterozygous and rare homozygous genotypes for rs2242480 in CYP3A4 and for rs776746 for CYP3A5 were found to be statistically linked to a worse outcome. Evaluating the genetic profile before starting lenvatinib treatment may help to predict the occurrence and grade of some side effects, and may contribute to improving patient management. ## 1. Introduction Targeted therapy for the treatment of locally advanced or metastatic thyroid cancer has been introduced in recent decades. Nowadays, multi-kinase inhibitors (MKIs) represent the best therapeutic option in advanced thyroid cancer patients [1]. MKIs are a class of drugs that inhibit mitogenic signals via the phosphorylation/dephosphorylation of several intracellular proteins involved in the MAPK pathway. Moreover, MKIs target growth factor receptors such as the vascular endothelial growth factor receptor (VEGFR) and the fibroblast growth factor receptor (FGFR), blocking angiogenesis and cell proliferation [2]. To date, several MKIs have been approved by the Food and Drug Administration (FDA) and European Medical Agency (EMA) for radioiodine (RAI) refractory differentiated thyroid cancer (sorafenib and lenvatinib) and medullary thyroid cancer (vandetanib and cabozantinib), while dabrafenib/trametinib combination has obtained regulatory approval from the FDA for anaplastic thyroid cancer with a BRAF V600 mutation [3,4,5,6,7]. Others kinase inhibitors, selpercatinib and pralsetinib, have recently been approved by the FDA for locally advanced or metastatic tumors with a rearranged during transfection (RET) gene fusion [8,9]. In these studies, it has been demonstrated that MKIs significantly improve the progression-free survival (PFS) of patients with advanced disease. In the SELECT trial, patients with progressive locally advanced or metastatic differentiated thyroid cancer treated with lenvatinib had a significantly longer PFS than those who received placebo (18.3 vs. 3.6 months, respectively). Several in vitro studies evaluated the pharmacokinetics of MKIs that are mainly metabolized in the liver by cytochrome P450 (CYP). These studies showed that CYP3A4, stimulated by cytochrome b5, is mainly involved in MKI metabolism and oxidation, but other cytochromes (i.e., CYP1A1, CYP1B1, CYP3A5, CYP2D6) also seem to be involved in this process [10,11,12,13]. Recent studies demonstrated that there is large inter-individual pharmacokinetic variability among patients treated with MKIs and that MKIs’ steady state is associated with the genetic polymorphisms in the genes coding for these proteins, which influence the absorption and elimination of the drug. In addition, the rate of development of adverse events during the treatment as well as the drug efficacy are significantly associated with these genetic polymorphisms in different human tumors [14,15,16]. A Japanese study evaluated the effect of CYP3A4-5 and ATP-binding cassette transporter (ABCC2) polymorphisms on lenvatinib pharmacokinetics in patients with thyroid cancer. A correlation between lenvatinib steady-state concentration and these polymorphisms was found, but the incidence of some adverse events (AE) (i.e., hypertension, hand foot syndrome, proteinuria), were not related to the presence of cytochrome polymorphism(s) [17]. Conclusive data are still missing in the literature. In this light, we retrospectively evaluated seven variants belonging to cytochrome P450 (CYP3A4 rs2242480 and rs2687116 and CYP3A5 rs776746) genes and to ATP-binding cassette transporter (ABCB1 rs1045642, rs2032582 and rs2235048 and ABCG2 rs2231142) genes in a small group of advanced thyroid cancer patients treated with lenvatinib to study the possible correlation between polymorphisms (SNPs), outcome and the appearance of drug-related AEs. ## 2.1. Patients In this retrospective study, we enrolled 18 advanced thyroid cancer patients (F:11) treated with lenvatinib. The mean age at the time of MKI treatment was 67.5 ± 13.8 years (median 69 years, 30–96 years). At histology, eight patients had differentiated thyroid cancer (DTC), eight patients had a poorly differentiated thyroid cancer (PDTC) and two medullary thyroid cancer (MTC). These last patients were included because in real-life clinical experience, lenvatinib showed interesting results as salvage therapy in patients with advanced progressive metastatic MTC [20]. The majority of patients ($\frac{13}{18}$, $72.3\%$) presented with distant metastasis in different sites before starting treatment (ranging from one to five different organs involved) and among them, $\frac{6}{13}$ had bone lesions. In total, $\frac{5}{18}$ ($27.7\%$) had locally advanced disease. Initial daily lenvatinib dose was 24 mg for $\frac{14}{18}$ ($77.7\%$) patients and 14 mg for $\frac{4}{18}$ ($22.3\%$) patients, due to older age and comorbidities. The dose was adjusted according to clinical conditions or the occurrence of AEs during the follow-up. The mean best-tolerated dose was 19.4 ± 5 mg (median 20 mg, 10–24). Lenvatinib was the first line of treatment for $\frac{14}{18}$ ($77.8\%$) patients and the second/third line of treatment for $\frac{4}{18}$ ($22.2\%$) patients (Table 1). Only patients without chronic use of concomitant medications that are known to be strong inducers or inhibitors of cytochrome P450 or substrates of the ATP-binding cassette transporters were included. We also excluded patients with severe comorbidities (i.e., gastrointestinal disease) to avoid a possible bias in terms of the evaluation of the adverse events. Considering the already described nine groups of adverse events (AEs), we divided patients according to the occurrence of less than $\frac{5}{9}$ AEs or more than $\frac{5}{9}$ of AEs, and we did not find any significant difference according to age ($$p \leq 0.53$$) or gender ($$p \leq 0.58$$). Even evaluating the higher grade of each adverse event and grouping patients according to the occurrence of mainly G1-AEs or G2/G3-AEs, we did not find any significant difference according to age ($$p \leq 0.93$$) or gender ($$p \leq 0.26$$). We report the observed drug-related adverse events in Table 2. The mean time of treatment was 23.0 months (range 2.7–52.9 months). During the first 12 months of treatment, $72.3\%$ of patients needed to stop lenvatinib therapy for a mean of 13.5 days (range 5–51) due to the severity of the side effects. During this period, the mean daily dosage in the cohort of patients was 17 ± 5 mg (not assessable in one patient due to lack of data). Overall, the rates of patients who suspended lenvatinib administration and had a subsequent dose reduction after the occurrence of each adverse events were: $50\%$ for diarrhea; $22\%$ for nausea, vomiting and epigastric pain; $0\%$ for oral mucositis and xerostomia; $10\%$ for hypertension; $40\%$ for proteinuria; $60\%$ for asthenia; $33\%$ for anorexia; $0\%$ for hand foot syndrome and weight loss. The median duration of dose reduction was 12 days for diarrhea; 10 days for nausea, vomiting and epigastric pain; 9 days for hypertension; 22.5 days for proteinuria; 8.66 days for asthenia and 15 days for anorexia. The reduction rate of starting dose was $24\%$ for diarrhea; $22\%$ for nausea, vomiting and epigastric pain; $16\%$ for hypertension; $41.6\%$ for proteinuria; $24.8\%$ for asthenia and $28.9\%$ for anorexia. In total, $\frac{2}{6}$ patients who experienced asthenia, $\frac{1}{5}$ patients who experienced diarrhea and $\frac{1}{3}$ patients who experienced anorexia needed a further $30\%$ reduction in drug dosage. ## 2.2. Correlation between SNPs and Side Effects Table 3 shows selected SNPs with indications of the most common allele in the European population according to the Ensembl database (https://www.ensembl.org/index.html (accessed on 27 January 2023)) and the observed allele frequency/genotypes in our population. Distribution for all SNPs was consistent with Hardy–Weinberg’s law at the level of significance of 0.05. For each variant, we evaluated the correlation with specific side effects in terms of presence/absence, grade and extent both at genotype and allele level. ## 2.2.1. Diarrhea Out of 18 analyzed patients, 14 presented with diarrhea ($77.8\%$). We found that both at genotype and allele level, none of the analyzed SNPs influenced the presence/absence of the side effect (Table 4). Patients were then stratified according to the grade of diarrhea into grade 1 (low, $$n = 9$$), grade 2 (moderate, $$n = 3$$) and grade 3 (severe, $$n = 2$$). At genotype level (Table 4), we found a statistically significant ($$p \leq 0.03$$) correlation for CYP3A4 rs2242480 and CYP3A5 rs776746. For CYP3A4 rs2242480, the GG genotype and for CYP3A5 rs776746, the CC genotype were associated with grade 1 ($64.5\%$ of the patients for both SNPs). At allele level, these data were confirmed (Table 4) with the G or the C allele mostly associated with the mildest manifestation of the side effect. Accordingly, for CYP3A4 rs2242480, the GA or AA genotypes were associated with grade 3 and 2 in $100\%$ of the cases. Similar results were observed for CYP3A5 rs776746. The TT genotype was associated with grade 2 in all patients and the CT genotype with grade 3. We then wondered whether a specific genetic profile was correlated with days of persistence of the maximum grade of the side effect. We stratified our patients accordingly and obtained three groups: less than 1 month (<1), between 1 and 4 months (1–4) and more than 4 months (>4). At genotype level, we did not find any association. Again, for CYP3A4 rs2242480, the G allele ($$p \leq 0.015$$) and for CYP3A5 rs776746, the C allele ($$p \leq 0.018$$) were statistically associated with >4 months (Table 5). ## 2.2.2. Nausea, Vomiting and Epigastric Pain Sixty-six percent ($\frac{12}{18}$; $66.7\%$) of patients suffered from nausea, vomiting and epigastric pain. In accordance with what was observed for diarrhea, none of the SNPs were associated with the presence/absence of the symptoms (considered together), both at genotype and allele level (Table 6). The majority of patients ($75\%$) who presented with the side effects were grade 1. Only three patients ($25\%$) were grade 2 and none were grade 3. At genotype level (Table 6), although not significant, we found the CC (rs1045642) or TT (rs2235048) genotypes in the ABCB1 gene to be slightly correlated with grade 1 ($$p \leq 0.08$$ and $$p \leq 0.06$$, respectively). The T allele in rs2235048 was statistically ($$p \leq 0.019$$) more present in patients with grade 1, as indicated in Table 6. Since these SNPs were close to significance, the haplotype was considered, but even combined together, they were not correlated with the grade of side effect ($$p \leq 0.13$$). No correlation was found between SNPs and days of persistence of the higher grade of the AE effect, both at genotype and allele levels. ## 2.2.3. Oral Mucositis and Xerostomia Several patients ($\frac{15}{18}$; $83.4\%$) presented with mucositis and xerostomia: $\frac{11}{15}$ ($73.4\%$) were grade 1, $\frac{3}{15}$ ($20\%$) were grade 2 and $\frac{1}{15}$ ($6.6\%$) was grade 3. None of the patients had the side effects for less than one month; for 5 patients ($33.4\%$), the duration was 1–4 months, and for $\frac{10}{15}$ ($66.6\%$), the extent was more than 4 months. We only found a statistically significant ($$p \leq 0.039$$) correlation between rs2231142 in the ABCG2 gene and the duration of the maximum grade at genotype level (Table 7). For this variant, the CC genotype was almost exclusively represented in the “>4 months” group ($\frac{9}{11}$ patients) and the CA genotype in the “1–4 months” ($\frac{3}{4}$ patients). No associations were found at allele level. ## 2.2.4. Hypertension and Proteinuria In total, $\frac{13}{18}$ patients ($72.3\%$) had hypertension. At genotype level, the rs2242480 in CYP3A4 (GG) and the rs776746 in CYP3A5 (CC) were associated ($$p \leq 0.02$$) with the presence of the side effect (Table 8). These data were confirmed at allele level (Table 8), with the G and C allele found in $75.8\%$ of the patients with hypertension ($$p \leq 0.001$$ and $$p \leq 0.0013$$, respectively). In total, $\frac{2}{13}$ ($15.3\%$) had grade 1, $\frac{7}{13}$ ($53.8\%$) had grade 2 and the remaining $30.9\%$ had grade 3. None of the SNPs were significantly associated with the severity and the duration of the side effect (Table 8). No correlation was found between variants and proteinuria at any levels. ## 2.2.5. Asthenia When evaluated for asthenia, we found that $53.8\%$ of the patients presented with grade 1, $30.8\%$ with grade 2 and the remaining $15.4\%$ with grade 3. The CC genotype and, consequently, the C allele for rs2231142 in ABCG2 were statistically associated ($$p \leq 0.017$$ and $$p \leq 0.028$$, respectively) with grade 1 (Table 9). At allele level, the G allele for rs2242480 (CYP3A4, $$p \leq 0.02$$) and the C allele for rs776746 (CYP3A5, $$p \leq 0.022$$) were also associated with grade 1 of asthenia. These SNPs both alone and combined together were slightly (but not significantly) linked to the severity of asthenia ($$p \leq 0.08$$) at genotype level. For the majority of patients ($59\%$), the side effect lasted more than 4 months. However, no statistically significant association was found between SNPs and asthenia duration. ## 2.2.6. Anorexia and Weight Loss Approximately all patients ($78\%$) experienced anorexia. This side effect was not linked to genotype at any level. Before treatment, mean body weight was 73.7 ± 20 kg and mean BMI was 27.2 ± 7 kg/m2; $\frac{5}{18}$ ($27.8\%$) patients were obese (BMI ≥ 30 kg/m2), $\frac{4}{18}$ ($22.2\%$) were overweight (BMI 25–29.9 kg/m2) and the remaining $50\%$ were normal weight (BMI 18.5–24.9 kg/m2). During the first year of treatment, the minimum mean weight reached was 66.9 ± 15 kg, ranging from 43.7 to 108.5 kg. Weight loss was evident in $83.4\%$ of patients, and it was ≥$5\%$ in $\frac{11}{18}$ ($61.1\%$) patients. In our cohort of patients, weight loss was not always present at the same time as anorexia, as it could be associated with diarrhea, nausea or vomiting. In our patients, weight loss was grade 1 in $36.4\%$, grade 2 in $45.4\%$ and grade 3 in $18.2\%$ of patients. The mean time needed to reach the minimum weight was 10.3 months (2.7–12 months). Both at genotype and allele levels, the ABCB1 gene was associated with weight loss (Table 10). Specifically, we found a correlation between the heterozygous CT genotype for both rs1045642 and rs2235048 ($$p \leq 0.003$$) corresponding to the C ($$p \leq 0.014$$) and the T ($$p \leq 0.02$$) allele, respectively. The analyzed SNPs were not associated with the severity of weight loss and duration of adverse events at either genotype or allele level. ## 2.2.7. Hand Foot Syndrome Fifty per cent of the patients ($\frac{9}{18}$) developed hand foot syndrome (HFS); $\frac{4}{9}$ ($44.4\%$) were grade 1 and $\frac{5}{9}$ ($55.6\%$) were grade 2. None of the patients reached grade 3. For $66.7\%$ of the subjects, HFS duration was more than 4 months (mean time 8.3 months), and for the remaining $33.3\%$ of patients, HFS duration was 1 to 4 months (mean 1.48 months). We did not find any association between SNPs and HFS at genotype or allele level. ## 2.2.8. Correlation between SNPs and Response to Lenvatinib Treatment As mentioned above, the mean time of treatment was 23.0 months (range 2.7–52.9 months). During the first 12 months of treatment, $72.3\%$ of patients needed to stop lenvatinib consumption for a mean of 13.5 days (range 5–51) due to the severity of the side effects. During this period, the mean daily dosage in the cohort of patients was 17 ± 5 mg (not assessable in one patient due to lack of data). None of the SNPs were significantly associated with the length of therapy interruption at genotype or allele levels. In terms of best response (calculated from the beginning of the treatment to the last CT scan), we observed $50\%$ stable response (SD), $44.5\%$ partial response (PR) and $5.5\%$ progressive disease (PD). Again, no association between SNPs and best response was observed (Table 11). A mean time of 8.83 months (range 2.767–21.233) was necessary to reach the best response. By grouping this interval into four categories (<5; 5–10; 10–20; >20 months) and excluding patients with a PD, we found an association ($$p \leq 0.013$$) with SNPs rs2242480 (CYP3A4) and rs776746 (CYP3A5). For both variants, heterozygous and rare homozygous patients fall exclusively into the category >20 months (Table 11) at genotype level. This result was even more evident when SNPs were analyzed at allele level (Table 11). Indeed, in both cases, the rare alleles (A for rs2242480, $8\%$ in the general population; T for rs776746, $6\%$ in the general population) significantly ($p \leq 0.0001$) correlated with the worst time interval. In order to explore if these results might be due to the lower dose of lenvatinib rather than the presence of SNPS, a comparison between patients with SNPS and patients carrying the most common genotypes was made. A higher dose of lenvatinib was administrated in patients with these SNPs (22.5 mg/die) compared to other patients (16.4 mg/die). In support of this hypothesis, a positive correlation was found between the mean dose of lenvatinib and the time necessary to reach the best response ($$p \leq 0.009$$, $R = 0.60$). Indeed, only $33.3\%$ of patients treated with more than 20 mg/die reached the BR during the first year of treatment compared to $81.8\%$ of patients taking a dosage less than 20 mg/die. ## 3. Discussion From a chemical point of view, lenvatinib is a member of the class of quinolines and is the carboxamide of 4-{3-chloro-4-[(cyclopropylcarbamoyl) amino] phenoxy}-7-methoxyquinoline-6-carboxylic acid, which works as a multi-kinase inhibitor and antineoplastic agent by targeting vascular endothelial growth factor receptors (VEGFR1–3), fibroblast growth factor receptors (FGFR1–4), platelet-derived growth factor receptor-alpha (PDGFRα), mast/stem cell growth factor receptor (KIT) and rearranged during transfection receptor (RET) proto-oncogenes. Lenvatinib is orally administered once daily at a maximum dosage of 24 mg in RAI-refractory DTC or at lower dosages according to patient clinical conditions. After oral administration and absorption, lenvatinib binds to plasma proteins (98–$99\%$), mainly albumin, with an elimination t$\frac{1}{2}$ of 28 h [12]. It is then mainly metabolized in the liver by cytochrome P450, mostly through CYP3A4 (>$80\%$), and also by CYP3A5, which has similar catalytic specificities [17,18]. Lenvatinib metabolites have low pharmacological activity and are excreted mainly via the biliary route [19,20,21]. Drug bioavailability is also determined by the P-glycoprotein encoded by the ABCB1 gene and the breast cancer resistance protein (BCRP) encoded by the ABCG2 gene, which are expressed in the small intestine, liver, kidney and blood–brain barrier, and are associated with drug availability, functioning to regulate the absorption and elimination of substrate drugs [22]. Inducers or inhibitors of these proteins can modify drug bioavailability, increasing or decreasing plasma concentration, respectively. For example, the co-administration of ketoconazole, which is a potent CYP3A, P-glycoprotein, and BCRP inhibitor, has been reported to increase the maximum plasma concentration of lenvatinib compared with placebo [23,24]. Patients treated with lenvatinib develop several adverse events such as hypertension, weight loss, anorexia, fatigue and gastrointestinal side effects [6]. Although these AEs are not life-threatening, they can be so severe as to dramatically impair patients’ quality of life with the consequence of a permanent drug withdrawal in a significant number of cases [1,6]. A recent study [17] tried to demonstrate a possible correlation between polymorphisms in the CYP3A4, CYP3A5, ABCB1, ABCC2 and ABCG2 genes and the incidence of adverse events during lenvatinib treatment. The authors showed that patients with CYP3A4 20230G>A polymorphism exhibited significantly lower dose-adjusted C0 values for lenvatinib. Nevertheless, the incidence of adverse events following lenvatinib treatment appeared not to be related to the investigated polymorphisms [17]. In our pilot study, we aimed to correlate the most common adverse reactions induced by lenvatinib with specific SNPs. We observed that at genotype level, the rs2242480 in CYP3A4 (GG) and the rs776746 in CYP3A5 (CC) were significantly associated with the presence of hypertension ($$p \leq 0.02$$). The most common genotypes for rs2242480 (CYP3A4: GG) and rs776746 (CYP3A5: CC) were associated with the mildest grade of diarrhea ($64.5\%$ of the patients for both SNPs). For all these SNPs, heterozygous or rare homozygous genotypes were associated with grade 2 or 3 of the side effects in $100\%$ of the cases. Regarding diarrhea, the same genetic profile also correlated with days of persistence of the maximum grade of the side effect (>4 months). A link with weight loss was also observed for heterozygous patients for the ABCB1 gene, both for rs1045642 and rs2235048. The ABCG2 gene was significantly correlated with the duration of grade maximum for mucositis and xerostomia. In this case, the heterozygous CA genotype was linked to shorter duration, whereas the CC genotype linked to a longer extent (>4 months). Polymorphisms in both cytochrome P450 genes and ATP-binding cassette transporters appear to be involved with asthenia. From this analysis, even though the group of patients was small, it seems that a particular genetic make-up can predict the occurrence and severity of common side effects that are MKI-related (Figure 1). The most implicated genes were CYP3A4 and CYP3A5. Our results showed that heterozygous and rare homozygous patients for these genes (rs2242480 and rs776746, respectively), took more than 20 months to reach the best response to treatment, compared to patients carrying the most common genotypes. In our cohort, this effect does not seem to be linked to lower drug dose, given that a higher dose of lenvatinib was administrated in patients with these SNPs (22.5 mg/die) compared to other patients (16.4 mg/die). In support of this hypothesis, a positive correlation was found between the mean dose of lenvatinib and the time necessary to reach the best response ($$p \leq 0.009$$, $R = 0.60$). Indeed only $33.3\%$ of patients treated with more than 20 mg/die reached the BR during the first year of treatment, compared to $81.8\%$ of patients taking a dosage less than 20 mg/die. We are aware that the limited number of patients strongly reduces the power of statistical analysis, and more patients are needed to confirm these data. In addition, since this is a retrospective study, the measurement of lenvatinib plasma concentrations was not available. However, in the era of precision medicine, the results obtained for some SNPs prompt us to suggest a genetic evaluation before starting MKI treatment. In our cohort, for example, patients who experienced diarrhea and asthenia required a suspension or a significant reduction in the dosage up to the half-life of the starting dose. With this in mind, predicting the occurrence and grade of some side effects will contribute to improving clinical management during lenvatinib treatment in patients with advanced thyroid carcinoma. ## 4.1. Patients This study retrospectively analyzed 18 patients with advanced thyroid cancer treated with lenvatinib, followed at the Unit of Endocrinology of Siena’s University Hospital (Italy), between November 2004 and May 2022. Informed consent has been signed by each patient enrolled in the study, and the study was approved by our local Ethical Committee (Ethics Committee of Region Toscana, Area Vasta Sud Est, AOUS. Protocol ID: 10167). The data collected included age at diagnosis, gender, anthropometric parameters, histological findings, stage at diagnosis, numbers of anatomical sites involved, tumor response and data on last follow-up/death. At baseline and during periodic follow-up, information regarding lenvatinib treatment was also collected: time lapse between diagnosis and treatment start, mean daily dosage, days of treatment suspension/reduction, appearance, grade and duration of AEs and duration of treatment. We evaluated the occurrence of the more common lenvatinib AEs during the first year of treatment or at last follow-up (if treatment lasted less than one year), grouped as follows: [1] diarrhea; [2] nausea, vomiting and epigastric pain; [3] oral mucositis and xerostomia; [4] hypertension and proteinuria; [5] asthenia; [6] anorexia and weight loss; [7] hand foot syndrome. For each group of AEs, we considered the overall grade and duration of the AE with the highest grade and duration. The severity of AEs was defined according to the NCI Common Terminology Criteria for Adverse Events v5.0 [https://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm (accessed on March 2022)]. Radiological evaluation was performed at baseline (before starting lenvatinib) and periodically (on average, at 3 and 6 months and thereafter annually) with contrast-enhanced computed tomography (CT) scanning. Treatment response was classified according to the Response Evaluation Criteria in Solid Tumors (RECIST) v.1.1 [18]. ## 4.2. DNA Extraction and SNPs Analysis Genomic DNA was extracted from peripheral blood leukocytes using the salting out procedure, obtained by fasting venous blood samples collected during lenvatinib treatment. DNA concentration and purity were assessed with a Nanodrop One (Thermo Scientific, Milan, Italy). The analysis of SNPs (Table 1) was carried out through PCR amplification. Specific primers were designed using the Primer3 Input program and purchased from Eurofins Genomics (Ebersberg, Germany). Primer sequences and PCR conditions are reported in the Supplementary Table S1. PCR products were subjected to Sanger sequencing using the BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems, Milan, Italy) and BigDye Xterminator Purification Kit on an automated DNA capillary sequencer (Applied Biosystems 3130xl Genetic Analyzer). Electropherograms were visualized using Chromas (http://technelysium.com.au/wp/chromas/ (accessed on March 2022)) and aligned with reference sequences using the Expasy SIM-Alignment Tool (https://web.expasy.org/sim/ (accessed on March 2022)). ## 4.3. Statistical Analysis All statistical analyses were carried out by using the software Statview version 5.0.1 for Windows (SAS Institute Inc., Cary, NC, USA). A p-value < 0.05 was considered statistically significant. The Hardy–Weinberg equilibrium was evaluated using the online calculator Gene Calc (https://gene-calc.pl/hardy-weinberg-page (accessed on 1 April 2022)). Haplotype frequencies and association statistics for the polymorphisms were evaluated using Haploview [19]. Association analyses were carried out using χ2 or Fischer’s exact tests at genotype, allele and haplotype levels. At allele level, the additive model was considered. ## References 1. Cabanillas M.E., Ryder M., Jimenez C.. Targeted Therapy for Advanced Thyroid Cancer: Kinase Inhibitors and Beyond. Endocr. Rev. (2019) 40 1573-1604. DOI: 10.1210/er.2019-00007 2. Matrone A., Valerio L., Pieruzzi L., Giani C., Cappagli V., Lorusso L., Agate L., Puleo L., Viola D., Bottici V.. Protein kinase inhibitors for the treatment of advanced and progressive radiorefractory thyroid tumors: From the clinical trials to the real life. Best Pract. Res. Clin. Endocrinol. Metab. (2017) 31 319-334. DOI: 10.1016/j.beem.2017.06.001 3. Wells S.A., Robinson B.G., Gagel R.F., Dralle H., Fagin J.A., Santoro M., Baudin E., Elisei R., Jarzab B., Vasselli J.R.. Vandetanib in patients with locally advanced or metastatic medullary thyroid cancer: A randomized, double-blind phase III trial. J. Clin. Oncol. (2012) 30 134-141. DOI: 10.1200/JCO.2011.35.5040 4. Elisei R., Schlumberger M.J., Müller S.P., Schöffski P., Brose M.S., Shah M.H., Licitra L., Jarząb B., Medvedev V., Kreissl M.C.. Cabozantinib in progressive medullary thyroid cancer. J. Clin. Oncol. (2013) 31 3639-3646. DOI: 10.1200/JCO.2012.48.4659 5. Brose M.S., Nutting C.M., Jarzab B., Elisei R., Siena S., Bastholt L., de la Fouchardiere C., Pacini F., Paschke R., Shong Y.K.. Sorafenib in radioactive iodine-refractory, locally advanced or metastatic differentiated thyroid cancer: A randomised, double-blind, phase 3 trial. Lancet (2014) 384 319-328. DOI: 10.1016/S0140-6736(14)60421-9 6. Schlumberger M., Tahara M., Wirth L.J., Robinson B., Brose M.S., Elisei R., Habra M.A., Newbold K., Shah M.H., Hoff A.O.. Lenvatinib versus placebo in radioiodine-refractory thyroid cancer. N. Engl. J. Med. (2015) 372 621-630. DOI: 10.1056/NEJMoa1406470 7. Subbiah V., Kreitman R.J., Wainberg Z.A., Cho J.Y., Schellens J.H.M., Soria J.C., Wen P.Y., Zielinski C., Cabanillas M.E., Urbanowitz G.. Dabrafenib and Trametinib Treatment in Patients with Locally Advanced or Metastatic BRAF V600-Mutant Anaplastic Thyroid Cancer. J. Clin. Oncol. (2018) 36 7-13. DOI: 10.1200/JCO.2017.73.6785 8. Bradford D., Larkins E., Mushti S.L., Rodriguez L., Skinner A.M., Helms W.S., Price L.S.L., Zirkelbach J.F., Li Y., Liu J.. FDA Approval Summary: Selpercatinib for the Treatment of Lung and Thyroid Cancers with RET Gene Mutations or Fusions. Clin. Cancer Res. (2021) 27 2130-2135. DOI: 10.1158/1078-0432.CCR-20-3558 9. Kim J., Bradford D., Larkins E., Pai-Scherf L.H., Chatterjee S., Mishra-Kalyani P.S., Wearne E., Helms W.S., Ayyoub A., Bi Y.. FDA Approval Summary: Pralsetinib for the Treatment of Lung and Thyroid Cancers with RET Gene Mutations or Fusions. Clin. Cancer Res. (2021) 27 5452-5456. DOI: 10.1158/1078-0432.CCR-21-0967 10. Indra R., Vavrová K., Pompach P., Heger Z., Hodek P.. Identification of Enzymes Oxidizing the Tyrosine Kinase Inhibitor Cabozantinib: Cabozantinib Is Predominantly Oxidized by CYP3A4 and Its Oxidation Is Stimulated by cyt b5 Activity. Biomedicines (2020) 8. DOI: 10.3390/biomedicines8120547 11. Indra R., Pompach P., Martínek V., Takácsová P., Vavrová K., Heger Z., Adam V., Eckschlager T., Kopečková K., Arlt V.M.. Identification of Human Enzymes Oxidizing the Anti-Thyroid-Cancer Drug Vandetanib and Explanation of the High Efficiency of Cytochrome P450 3A4 in its Oxidation. Int. J. Mol. Sci. (2019) 20 3392. PMID: 31295928 12. Fogli S., Gianfilippo G., Cucchiara F., Del Re M., Valerio L., Elisei R., Danesi R.. Clinical pharmacology and drug-drug interactions of lenvatinib in thyroid cancer. Crit Rev. Oncol. Hematol. (2021) 163 103366. DOI: 10.1016/j.critrevonc.2021.103366 13. Vavrová K., Indra R., Pompach P., Heger Z., Hodek P.. The impact of individual human cytochrome P450 enzymes on oxidative metabolism of anticancer drug lenvatinib. Biomed. Pharmacother. (2022) 145 112391. DOI: 10.1016/j.biopha.2021.112391 14. Abumiya M., Mita A., Takahashi S., Yoshioka T., Kameoka Y., Takahashi N., Miura M.. Effects of polymorphisms in NR1I2, CYP3A4, and ABC transporters on the steady-state plasma trough concentrations of bosutinib in Japanese patient with chronic myeloid leukemia. Med. Oncol. (2018) 35 90. DOI: 10.1007/s12032-018-1146-z 15. Hayashi H., Iihara H., Hirose C., Fukuda Y., Kitahora M., Kaito D., Yanase K., Endo J., Ohno Y., Suzuki A.. Effects of pharmacokinetics-related genetic polymorphisms on the side effect profile of afatinib in patients with non-small cell lung cancer. Lung Cancer (2019) 134 1-6. DOI: 10.1016/j.lungcan.2019.05.013 16. Kato H., Sassa N., Miyazaki M., Takeuchi M., Asai M., Iwai A., Noda Y., Gotoh M., Yamada K.. Association of axitinib plasma exposure and genetic polymorphisms of ABC transporters with axitinib-induced toxicities in patients with renal cell carcinoma. Cancer Chemother. Pharmacol. (2016) 78 855-862. DOI: 10.1007/s00280-016-3145-0 17. Ozeki T., Nagahama M., Fujita K., Suzuki A., Sugino K., Ito K., Miura M.. Influence of CYP3A4/5 and ABC transporter polymorphisms on lenvatinib plasma trough concentrations in Japanese patients with thyroid cancer. Sci. Rep. (2019) 9 5404. DOI: 10.1038/s41598-019-41820-y 18. Eisenhauer E.A., Therasse P., Bogaerts J., Schwartz L.H., Sargent D., Ford R., Dancey J., Arbuck S., Gwyther S., Mooney M.. New response evaluation criteria in solid tumours: Revised RECIST guideline (version 1.1). Eur. J. Cancer (2009) 45 228-247. DOI: 10.1016/j.ejca.2008.10.026 19. Barrett J.C., Fry B., Maller J., Daly M.J.. Haploview: Analysis and visualization of LD and haplotype maps. Bioinformatics (2005) 21 263-265. DOI: 10.1093/bioinformatics/bth457 20. Matrone A., Prete A., Nervo A., Ragni A., Agate L., Molinaro E., Giani C., Valerio L., Minaldi E., Piovesan A.. Lenvatinib as a salvage therapy for advanced metastatic medullary thyroid cancer. J. Endocrinol. Investig. (2021) 44 2139-2151. DOI: 10.1007/s40618-020-01491-3 21. Dubbelman A.-C., Rosing H., Nijenhuis C., Huitema A.D.R., Mergui-Roelvink M., Gupta A., Verbel D., Thompson G., Shumaker R., Schellens J.H.M.. Pharmacokinetics and excretion of (14)C-lenvatinib in patients with advanced solid tumors or lymphomas. Investig. New Drugs (2015) 33 233-240. DOI: 10.1007/s10637-014-0181-7 22. Boss D.S., Glen H., Beijnen J.H., Keesen M., Morrison R., Tait B., Copalu W., Mazur A., Wanders J., O’brien J.P.. A phase I study of E7080, a multitargeted tyrosine kinase inhibitor, in patients with advanced solid tumours. Br. J. Cancer (2012) 106 1598-1604. DOI: 10.1038/bjc.2012.154 23. Hu B., Zou T., Qin W., Shen X., Su Y., Li J., Chen Y., Zhang Z., Sun H., Zheng Y.. Inhibition of EGFR Overcomes Acquired Lenvatinib Resistance Driven by STAT3-ABCB1 Signaling in Hepatocellular Carcinoma. Cancer Res. (2022) 82 3845-3857. DOI: 10.1158/0008-5472.CAN-21-4140 24. Shumaker R., Aluri J., Fan J., Martinez G., Thompson G.A., Ren M.. Effects of Ketoconazole on the Pharmacokinetics of Lenvatinib (E7080) in Healthy Participants. Clin. Pharmacol. Drug Dev. (2015) 4 155-160. DOI: 10.1002/cpdd.140
--- title: The Role of Olfactomedin 2 in the Adipose Tissue–Liver Axis and Its Implication in Obesity-Associated Nonalcoholic Fatty Liver Disease authors: - Andrea Barrientos-Riosalido - Laia Bertran - Mercè Vilaró-Blay - Carmen Aguilar - Salomé Martínez - Marta Paris - Fàtima Sabench - David Riesco - Jessica Binetti - Daniel Del Castillo - Cristóbal Richart - Teresa Auguet journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049551 doi: 10.3390/ijms24065221 license: CC BY 4.0 --- # The Role of Olfactomedin 2 in the Adipose Tissue–Liver Axis and Its Implication in Obesity-Associated Nonalcoholic Fatty Liver Disease ## Abstract This study’s objective was to assess the involvement of olfactomedin 2 (OLFM2), a secreted glycoprotein related to lipid metabolism regulation, in nonalcoholic fatty liver disease (NAFLD) mediated by the adipose-tissue–liver axis. OLFM2 mRNA expression was analyzed in subcutaneous (SAT) and visceral (VAT) adipose tissue by RT–qPCR. The cohort included women with normal weight ($$n = 16$$) or morbid obesity (MO, $$n = 60$$) who were subclassified into normal liver ($$n = 20$$), simple steatosis ($$n = 21$$), and nonalcoholic steatohepatitis (NASH, $$n = 19$$) groups. The results showed that OLFM2 expression in SAT was enhanced in MO individuals and in the presence of NAFLD. Specifically, OLFM2 expression in SAT was increased in mild and moderate degrees of steatosis in comparison to the absence of it. Moreover, OLFM2 expression in SAT was negatively correlated with interleukin-6 levels. On the other hand, OLFM2 expression in VAT decreased in the presence of NASH and exhibited a positive correlation with adiponectin levels. In conclusion, OLFM2 in SAT seems to be implicated in hepatic lipid accumulation. Additionally, since we previously suggested the possible implication of hepatic OLFM2 in NAFLD progression, now we propose a possible interaction between the liver and SAT, reinforcing the potential implication of this tissue in NAFLD development. ## 1. Introduction In recent years, nonalcoholic fatty liver disease (NAFLD) has grown to be a major public health problem, and its incidence has increased drastically, with a prevalence of approximately $25\%$ in the global adult population [1]. NAFLD is a liver disorder characterized by the abnormal accumulation of intrahepatic fat in the absence of other etiologies. It is considered a multisystemic disease and a hepatic manifestation of metabolic syndrome (MetS). NAFLD is associated with type 2 diabetes mellitus (T2DM), obesity, and insulin resistance (IR) [2]. Its natural history begins with simple steatosis (SS), which occurs when there is more than $5\%$ fat in hepatocytes without evidence of hepatic dysfunction. However, this condition can develop into nonalcoholic steatohepatitis (NASH), a more serious disorder characterized by severe steatosis, hepatocellular ballooning, and lobular and/or portal inflammation with or without hepatic fibrosis. This severe condition can evolve into cirrhosis or hepatocellular carcinoma if it is not treated [2]. The pathogenesis of NAFLD is described by the multiple hit hypothesis, which postulates that there are many factors simultaneously impacting the liver that cause the development of NAFLD [3]. In addition, crosstalk with other organs, such as adipose tissue, can also affect the liver when the tissue is dysfunctional, as in obese or diabetic patients, resulting in a flow of free fatty acids (FFAs) and inflammatory adipokines to the liver and inducing hepatic fatty accumulation, liver inflammation, and hepatic damage (Figure 1). In recent years, the importance of the interaction between adipose tissue and the liver in the progression of NAFLD has been noted [4]. There are two types of white adipose tissue: subcutaneous adipose tissue (SAT), which is located under the skin and exhibits increased expression of proinflammatory genes in patients with morbid obesity (MO) [5,6]; and visceral adipose tissue (VAT), which covers internal organs and sends an increased flow of adipokines and FFAs to the liver through the portal circulation [5,6,7]. Conditions of excessive fat accumulation in VAT have been shown to be associated with the development of metabolic disorders and a predisposition for NAFLD [8,9]. In addition, epidemiological research has shown that SAT expansion induces insulin sensitivity and lowers the incidence of T2DM, while VAT build-up increases metabolic risk and total mortality [5]. In addition, the pathogenesis of NAFLD involves many metabolic pathways that have not yet been fully elucidated. For this reason and because this disease has a very invasive gold-standard diagnostic method, although it is very accurate, and the lack of specific pharmacological treatments accepted by regulatory agencies [10,11,12], more studies are necessary on the molecular mechanisms involved in NAFLD to find new therapeutic targets. Olfactomedin 2 (OLFM2) is a secreted glycoprotein [13] related to the regulation of lipid metabolism, insulin resistance, and obesity [14] that has been implicated in different diseases, such as ocular glaucoma [15,16] and hepatocellular carcinoma [17]. The Human Protein Atlas (HPA) reported that OLFM2 was mostly expressed in the brain, but high expression was also found in the liver and adipose tissue [18]. Therefore, in our previous study, we evaluated the potential role of hepatic OLFM2 in obesity-associated NAFLD. We conducted a study using liver biopsies, and we found increasing relative OLFM2 mRNA expression as hepatic disease became more severe, suggesting a potential key role of OLFM2 in the progression of NAFLD [19]. However, this protein has not been well studied, especially in terms of liver or metabolic diseases, making it difficult to draw conclusions. Since we first described the association between OLFM2 and NAFLD and increased expression of OLFM2 in adipose tissue was reported in the HPA [18], we wanted to investigate the relative mRNA expression of OLFM2 in the current research in SAT and VAT and study its role in adipose-tissue–liver-axis-mediated NAFLD in a well-established cohort of women with normal weight (NW) or MO who had different degrees of NAFLD involvement. ## 2.1. Patients’ Baseline Characteristics The clinical and biochemical values of the patients are shown in Table 1. The cohort included women who were separated into two groups based on their body mass index (BMI): those who were normal weight (NW, BMI < 25 kg/m2, $$n = 16$$) and those than showed MO (BMI ≥ 40 kg/m2, $$n = 60$$). Subsequently, MO patients were subdivided into three groups based on liver histology: normal liver (NL, $$n = 20$$), SS ($$n = 21$$), and NASH ($$n = 19$$). The cohort members were comparable in terms of age, diastolic blood pressure (DBP), systolic blood pressure (SBP), and low-density lipoprotein cholesterol (LDL-C). Apart from exhibiting higher weight and BMI in NL and SS groups than the NW group, we found enhanced values of the homeostatic model assessment method for insulin resistance (HOMA1-IR), glucose, insulin, alanine aminotransferase (ALT), and gamma-glutamyltransferase (GGT) in the SS and NASH groups compared to the NW cohort. In addition, we observed an increase in waist, glycosylated hemoglobin (HbA1c), and triglyceride (TG) levels in the MO group, which was the opposite of what was observed with high-density lipoprotein cholesterol (HDL-C), which was increased in the NW group. There were lower levels of cholesterol in the SS group compared to the NW group and an increase in aspartate aminotransferase (AST) levels in NASH subjects. Compared with NL patients, we reported an increase in glucose and alkaline phosphatase (ALP) levels in SS and NASH patients, while in the SS group, ALP was increased compared to that in the NASH group. In addition, we analyzed the liver histology of the cohort and some proinflammatory cytokines in women with MO, such as interleukin (IL)-8 and (IL)-6; and some anti-inflammatory mediators, such as adiponectin and IL-10. However, these cytokines did not exhibit significant differences between the groups. ## 2.2. Evaluation of the Relative mRNA Expression of OLFM2 in the Adipose Tissue of Patients with NW or MO First, we determined the relative mRNA expression of OLFM2 in VAT and SAT in the presence of obesity (BMI < 25 kg/m2 or BMI ≥ 40 kg/m2). We observed only a significant increase in OLFM2 mRNA expression in the SAT of MO subjects, as shown in Figure 2A. However, in VAT (Figure 2B), we did not find significant differences between the groups. ## 2.3. Evaluation of the Relative mRNA Expression of OLFM2 in Adipose Tissue Based on Hepatic Histology Since we had previously reported that hepatic OLFM2 could play a role in the progression of NAFLD and higher expression of OLFM2 mRNA was observed in the SAT of MO subjects, we analyzed the mRNA expression of OLFM2 according to liver involvement. First, we divided the patients according to whether they had NAFLD. In this regard, we observed an enhancement in relative OLFM2 mRNA abundance in the SAT of NAFLD patients compared to subjects with healthy livers (Figure 3A). However, we did not find significant differences in VAT (Figure 3B). Second, we evaluated the relative mRNA expression of OLFM2 in adipose tissues based on the different histopathological grades of NAFLD (NL, SS, and NASH). SAT exhibited a significant increase in OLFM2 mRNA expression in the SS and NASH groups compared to the NL and NW control groups, as shown in Figure 3C. However, in VAT, we observed nonsignificant differences between the groups (Figure 3D). ## 2.4. Evaluation of Relative OLFM2 mRNA Expression in Adipose Tissues Based on Liver Steatosis Degree Then, to further examine the connection between OLFM2 mRNA expression in adipose tissue and NAFLD, we analyzed the relative mRNA expression of OLFM2 in adipose tissues according to the different degrees of liver steatosis (absent, mild, moderate, and severe). In SAT, we observed enhanced expression of OLFM2 in mild and moderate steatosis stages in comparison with the absence of liver steatosis, as shown in Figure 4A. However, we did not find significant differences in the expression of OLFM2 depending on the different degrees of steatosis in VAT (Figure 4B). ## 2.5. Evaluation of Relative OLFM2 mRNA Expression in Adipose Tissue in the Presence of NASH and in Relation to NASH-Related Parameters Subsequently, we further examined the relationship between the expression of OLFM2 in adipose tissue and NASH in depth, and we analyzed the relative abundance of OLFM2 in the adipose tissue samples that was classified according to the presence or absence of NASH. Meanwhile, in SAT, we reported nonsignificant differences (Figure 5A), but in VAT, we reported lower OLFM2 mRNA expression in NASH patients than in non-NASH patients (Figure 5B). Given this significant difference, we evaluated the expression of OLFM2 in relation to different parameters related to NASH, such as portal or lobular inflammation and hepatocyte ballooning. Regarding lobular inflammation, we did not find significant differences in SAT (Figure 5C), while in VAT, we observed lower OLFM2 mRNA in patients with lobular inflammation than in subjects without lobular inflammation (Figure 5D). Regarding portal inflammation, in SAT, we did not find significant differences between the groups (Figure 5E), while in VAT, we reported lower expression of OLFM2 in the presence of portal inflammation than in its absence, as shown in Figure 5F. However, when analyzing the relative mRNA abundance of OLFM2 in terms of the presence of ballooning, we reported nonsignificant differences in adipose tissue among the groups. ## 2.6. Correlation between Relative OLFM2 mRNA Expression in SAT and VAT and Clinical and Biochemical NAFLD-Related Parameters To test our previous hypothesis, we analyzed correlations between OLFM2 in adipose tissue and different clinical and biochemical parameters related to NAFLD. We observed a negative association between the relative abundance of OLFM2 mRNA in SAT and circulating levels of IL-6, as shown in Figure 6A. On the other hand, in VAT, we found a positive association between the mRNA expression of OLFM2 and adiponectin levels (Figure 6B). ## 3. Discussion The most interesting findings of our study lie in the fact that we propose a possible interaction between OLFM2 in adipose tissue and NAFLD pathogenesis. Since we previously reported an association between hepatic OLFM2 expression and NAFLD progression [19], and given that this protein, which is highly expressed in adipose tissue [18], is associated with the regulation of lipid metabolism [14,16], we believe that it is both interesting and novel to study the expression of OLFM2 in adipose tissue (SAT and VAT) and its implication in obesity-associated NAFLD [4]. Our main findings were that the relative mRNA expression of OLFM2 in SAT was increased in patients with MO compared to NW subjects. Furthermore, OLFM2 expression was enhanced in the presence of NAFLD. Moreover, OLFM2 expression in SAT was increased in mild and moderate degrees of steatosis in comparison to the absence of it. Regarding OLFM2 in VAT, we reported a decrease in expression in the presence of NASH. Finally, we reported a negative association between OLFM2 mRNA expression in SAT and circulating IL-6 levels. Additionally, we found a positive correlation between OLFM2 expression in VAT and adiponectin levels. Our first finding shows that, in SAT, the relative mRNA expression of OLFM2 was increased in patients with MO compared to NW subjects. However, in VAT, we did not find significant differences between the groups. Similarly, Gonzalez-Garcia et al. carried out a study comparing OLFM2-null mice and wild-type mice and reported that OLFM2-null mice exhibited lower body weights [14]. In addition, mice with OLFM2 expression in the hypothalamus exhibited the opposite phenotype and had increased weight. This research showed that OLFM2 was involved in the regulation of energy metabolism [14]. Therefore, our results reinforce the findings of Gonzalez-Garcia et al., since our MO patients exhibited increased expression of OLFM2 in SAT, suggesting a potential role of OLFM2 in the regulation of energy metabolism in this tissue. In this sense, some studies have described the involvement of SAT in human metabolism, as well as an important role in lipotoxicity or insulin resistance [20,21]. Unfortunately, we did not find increased OLFM2 expression in VAT, which has been reported to be related to metabolic disruption [22]. Additionally, the expression of various genes in SAT and VAT in obesity was previously shown to be different [5,23]. This phenomenon could also occur in NAFLD and could explain our results, suggesting a potential role of OLFM2 in SAT in lipid metabolism or in inflammatory processes [24]. We reported increased expression of OLFM2 mRNA in the SAT of patients with NAFLD compared to patients without NAFLD. When the cohort was classified according to the degree of NAFLD, we observed an increase in OLFM2 mRNA abundance in patients with SS and NASH compared to controls. These results were similar to our previous study in which hepatic OLFM2 mRNA expression was increased as hepatic involvement became more severe [19]. Some studies reported that SAT exhibits increased expression of proinflammatory genes in patients with MO [5,6]; similarly, Plessis et al. demonstrated the significance of the SAT gene set in both the early stage of NAFLD and its progression to NASH [25], which may support the hypothesis that OLFM2 in SAT could play a role in NAFLD progression in the same way that OLFM2 affects the liver. There have been few reports on OLFM2, and it is difficult to understand the specific mechanisms and their implication in NAFLD. Because the expression of OLFM2 in adipose tissue mirrors its expression in the liver and increases as liver conditions worsen, and given the regulatory role that OLFM2 may play in lipid metabolism [14], we decided to evaluate the relative mRNA expression of OLFM2 in terms of the degree of steatosis in NAFLD patients. We found that OLFM2 expression was higher in subjects with mild and moderate steatosis than in those without it. This finding is consistent with the previous results and reinforces our hypothesis that OLFM2 may be related to the development of NAFLD and contribute to the regulation of lipid metabolism, as mentioned in Gonzalez-Garcia et al. [ 14]. However, additional studies are needed to confirm the effect of OLFM2 in adipose tissue on NAFLD. Concerning VAT, we reported a decrease in OLFM2 mRNA expression in NASH and in the presence of portal and lobular inflammation compared to the absence of these factors. It has been postulated that VAT plays a fundamental role in the development of NAFLD and metabolic diseases [26] due to its connection through the portal vein, directly exposing the liver to the flow of FFAs and proinflammatory factors [27]. This result may seem contradictory, since, in a previous study, we reported an increase in the expression of hepatic OLFM2 mRNA in patients with NASH and in subjects with lobular inflammation [19]. However, OLFM2 expression had not previously been evaluated in VAT. In addition, since most of our results related OLFM2 in SAT to NAFLD, we concluded that the key role of OLFM2 in VAT in inflammation related to NAFLD could be different or even contrary to that in the liver [20,21], as well as in mitochondrial respiration, which is a decreased in VAT but not in SAT in obese individuals with NAFLD [28]. We examined the correlations between the expression of OLFM2 and different biochemical and clinical variables that are related to NAFLD. We found a negative association between OLFM2 mRNA expression in SAT and circulating IL-6 levels. In this way, it is important to note that IL-6 is a proinflammatory adipokine that is usually increased in NAFLD [29]. However, our patients did not exhibit significant differences in IL-6 levels between the groups, probably due to the low-grade chronic inflammation that is caused by MO, which alters inflammatory mediator levels [30,31], as well as SAT metabolism [8]. Although obesity is closely related to NAFLD, excess storage of visceral fat is considered equally or more important and can mask the NAFLD condition; therefore, visceral-fat reduction is necessary to promote metabolic changes in NAFLD [32]. Additionally, we found a positive correlation between OLFM2 mRNA expression in VAT and circulating adiponectin levels. In this regard, adiponectin is an anti-inflammatory cytokine that is typically decreased in NAFLD [33]. Thus, this correlation was consistent with the association with OLFM2 mRNA expression in VAT, and we found decreased expression of this molecule in VAT in NASH, which would decrease adiponectin levels; however, in our subjects, we did not find significant differences in adiponectin, probably due to the abovementioned low-grade chronic inflammation [34]. Thanks to this study, we were able to evaluate the potential role of OLFM2 in adipose tissue in obesity-associated NAFLD. Meanwhile, we studied a cohort composed of only women, and subjects with liver biopsy exhibited MO. We found a potential role of OLFM2 in SAT in NAFLD, but these results are preliminary and cannot be extrapolated to other populations. Therefore, further research is needed in this field. ## 4.1. Patients The Institut Investigació Sanitària Pere Virgili (IISPV, Tarragona, Spain) approved this study through the institutional review committee (CEIm; 23c/2015; 11 May 2015). The cohort consisted of 76 Caucasian women with NW (BMI > 25 kg/m2, $$n = 16$$) and MO (BMI ≥ 40 kg/m2, $$n = 60$$). Written informed consent was obtained from all participants; SAT, VAT, and, during planned laparoscopic bariatric surgery, liver biopsies were collected. These were the conditions for exclusion: [1] an ethanol intake greater than 10 g/day or other toxins; [2] menopausal women or women taking contraceptives to prevent interference from hormones that can bias glucose and lipid metabolism, as well as cytokine determinations; [3] patients with infectious disease, neoplastic disease, or acute or chronic liver disease other than NAFLD; and [4] patients receiving fibrates because this medication can interfere with the metabolism of some metabolites derived from the microbiota studied in this work. ## 4.2. Sample Size To realize our objective, sample size was calculated using a GRANMO calculator (IMIM, Barcelona, Spain). Accepting an alpha risk of 0.05 and a beta risk of less than 0.2 in a two-sided test, a minimum of 22 controls and 67 cases are needed to recognize as statistically significant an odds ratio greater than or equal to 0.13. A proportion of the exposed subjects in the control group was estimated to be $0.25\%$, using the POISSON test. ## 4.3. Liver Pathology An experienced hematopathologist used the method of Kleiner and Brunt [35,36] to stratify the liver samples, using hematoxylin and eosin and Masson’s trichrome stains (see Figure 7 for visual information of the histology). Women with MO were classified according to their hepatic histopathology into NL ($$n = 20$$) and NAFLD ($$n = 40$$). Subjects with NAFLD were subclassified into SS (micro/macrovesicular steatosis without inflammation or fibrosis, $$n = 21$$) and NASH (Brunt grades 1–2, $$n = 19$$). In our group, there was no liver fibrosis porto-portal in any of the NASH patients. ## 4.4. Biochemical Analyses Prior to bariatric surgery and having fasted the night before, specialized nurses collected blood samples into tubes with or without ethylenediaminetetraacetic acid through a BD Vacutainer® system. These samples were then separated into aliquots of serum and plasma centrifugation (3500 rpm, 4 °C, and 15 min). A conventional automatic analyzer was used to analyze the biochemical parameters, and the IR was estimated through the homeostatic model for IR (HOMA1-IR). Cytokines, such as interleukin IL-6, IL-8, IL-10, TNF-α, and adiponectin, were determined in all samples of the cohort by multiplex sandwich immunoassays and MILLIPLEX MAP Human Adipokine Magnetic Bead Panel 1 (HADK1MAG-61K, Millipore, Billerica, MA, USA), the MILLIPLEX MAP High Sensitivity Human T-Cell Panel (HSTCMAG28SK, Millipore, Billerica, MA, USA), and the Bio-Plex 200 instrument, according to manufacturer’s instructions. All of these analyses were evaluated at the Omic Sciences Center (Eurecat, Reus, Spain), and the physical, anthropometric, and biochemical evaluations were performed on the entire cohort. ## 4.5. Gene Expression in Liver The hepatic and adipose tissue samples were collected during bariatric and conserved in tubes with RNAlater (Qiagen, Hilden, Germany) at 4 °C. Then samples were processed and stored at −80 °C. RNeasy mini kit (Qiagen, Barcelona, Spain) was used to extract total RNA from adipose tissue. Reverse transcription to cDNA was performed with the High-Capacity RNA-to-cDNA Kit (Applied Biosystems, Madrid, Spain). Real time quantitative polymerase chain reaction (PCR) was carried out with the TaqMan Assay predesigned by Applied Biosystems for the detection of OLMF2 (Hs01017934_m1). The expression of each gene was calculated and standardized to the expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Hs02786624_g1); after, they were normalized using the control group (NW) as a reference. All reactions were duplicated in 96-well plates, using the QuantStudio™ 7 Pro Real-Time PCR System (Applied Biosystem, Foster City, CA, USA). Determination of OLFM2 mRNA expression was analyzed in all adipose tissue samples obtained from the cohort ($$n = 76$$ (NW, $$n = 16$$; NL, $$n = 20$$; SS, $$n = 21$$; NASH, $$n = 19$$)); the raw data of the mRNA relative expression of OLFM2 in adipose tissue (CT/CQ values) are found in the Supplementary Materials Table S1. ## 4.6. Statistical Analysis The data were analyzed using the SPSS/PC+ for Windows statistical package (version 27.0; SPSS, Chicago, IL, USA). The distribution of variables was obtained using the Kolmogorov–Smirnov test, and the different comparative analyses were assessed using Mann–Whitney U test to compare groups. Using Spearmen’s method, the coefficient of correlation (rho) between variables was calculated. All results were expressed as the median and the interquartile range (25th–75th). The p-values < 0.05 were statistically significant. Graphics were elaborated using GraphPad Prism software (version 7.0; GraphPad, San Diego, CA, USA). ## 5. Conclusions In this study, we found that OLFM2 in SAT could regulate lipid metabolism involved in the progression of NAFLD. Since we previously suggested a possible implication of hepatic OLFM2 in NAFLD progression, we now propose a possible interaction between the OLFM2 of SAT and the liver, reinforcing the fact that a SAT–liver axis may be implicated in NAFLD development. ## References 1. Cotter T.G., Rinella M.. Nonalcoholic Fatty Liver Disease 2020: The State of the Disease. Gastroenterology (2020.0) 158 1851-1864. DOI: 10.1053/j.gastro.2020.01.052 2. Cobbina E., Akhlaghi F.. Non-Alcoholic Fatty Liver Disease (NAFLD)—Pathogenesis, Classification, and Effect on Drug Metabolizing Enzymes and Transporters. Drug Metab. Rev. (2017.0) 49 197-211. DOI: 10.1080/03602532.2017.1293683 3. Tilg H., Moschen A.R.. Evolution of Inflammation in Nonalcoholic Fatty Liver Disease: The Multiple Parallel Hits Hypothesis. Hepatology (2010.0) 52 1836-1846. DOI: 10.1002/hep.24001 4. Tilg H., Adolph T.E., Moschen A.R.. Multiple Parallel Hits Hypothesis in Nonalcoholic Fatty Liver Disease: Revisited After a Decade. Hepatology (2021.0) 73 833-842. DOI: 10.1002/hep.31518 5. Mittal B.. Subcutaneous Adipose Tissue & Visceral Adipose Tissue. Indian J. Med. Res. (2019.0) 149 571-573. DOI: 10.4103/ijmr.IJMR_1910_18 6. Alligier M., Gabert L., Meugnier E., Lambert-Porcheron S., Chanseaume E., Pilleul F., Debard C., Sauvinet V., Morio B., Vidal-Puig A.. Visceral Fat Accumulation During Lipid Overfeeding Is Related to Subcutaneous Adipose Tissue Characteristics in Healthy Men. J. Clin. Endocrinol. Metab. (2013.0) 98 802-810. DOI: 10.1210/jc.2012-3289 7. Tresserra-Rimbau A., Rimm E.B., Medina-Remón A., Martínez-González M.A., de la Torre R., Corella D., Salas-Salvadó J., Gómez-Gracia E., Lapetra J., Arós F.. Inverse Association between Habitual Polyphenol Intake and Incidence of Cardiovascular Events in the PREDIMED Study. Nutr. Metab. Cardiovasc. Dis. (2014.0) 24 639-647. DOI: 10.1016/j.numecd.2013.12.014 8. Spoto B., Di Betta E., Mattace-Raso F., Sijbrands E., Vilardi A., Parlongo R.M., Pizzini P., Pisano A., Vermi W., Testa A.. Pro- and Anti-Inflammatory Cytokine Gene Expression in Subcutaneous and Visceral Fat in Severe Obesity. Nutr. Metab. Cardiovasc. Dis. (2014.0) 24 1137-1143. DOI: 10.1016/j.numecd.2014.04.017 9. Smith C.J., Perfetti T.A., Hayes A.W., Berry S.C.. Obesity as a Source of Endogenous Compounds Associated With Chronic Disease: A Review. Toxicol. Sci. (2020.0) 175 149-155. DOI: 10.1093/toxsci/kfaa042 10. Dufour J.-F., Anstee Q.M., Bugianesi E., Harrison S., Loomba R., Paradis V., Tilg H., Wong V.W.-S., Zelber-sagi S.. Current Therapies and New Developments in NASH. Gut (2022.0) 71 2123-2134. DOI: 10.1136/gutjnl-2021-326874 11. Younossi Z.M., Koenig A.B., Abdelatif D., Fazel Y., Henry L., Wymer M.. Global Epidemiology of Nonalcoholic Fatty Liver Disease-Meta-Analytic Assessment of Prevalence, Incidence, and Outcomes. Hepatology (2016.0) 64 73-84. DOI: 10.1002/hep.28431 12. 12.Non-Alcoholic Fatty Liver Disease: Assessment and ManagementNational Institute for Health and Care ExcellenceLondon, UK2016978-1-4731-1996-3. Non-Alcoholic Fatty Liver Disease: Assessment and Management (2016.0) 13. Tomarev S.I., Nakaya N.. Olfactomedin Domain-Containing Proteins: Possible Mechanisms of Action and Functions in Normal Development and Pathology. Mol. Neurobiol. (2009.0) 40 122-138. DOI: 10.1007/s12035-009-8076-x 14. González-García I., Freire-Agulleiro Ó., Nakaya N., Ortega F.J., Garrido-Gil P., Liñares-Pose L., Fernø J., Labandeira-Garcia J.L., Diéguez C., Sultana A.. Olfactomedin 2 Deficiency Protects against Diet-Induced Obesity. Metabolism (2022.0) 129 155122. DOI: 10.1016/j.metabol.2021.155122 15. Holt R., Ugur Iseri S.A., Wyatt A.W., Bax D.A., Gold Diaz D., Santos C., Broadgate S., Dunn R., Bruty J., Wallis Y.. Identification and Functional Characterisation of Genetic Variants in OLFM2 in Children with Developmental Eye Disorders. Hum. Genet. (2017.0) 136 119-127. DOI: 10.1007/s00439-016-1745-8 16. Yang X., Sun N.-N., Zhao Z.-N., He S.-X., Zhang M., Zhang D.-D., Yu X.-W., Zhang J.-M., Fan Z.-G.. Coinheritance of. WJCC (2021.0) 9 697-706. DOI: 10.12998/wjcc.v9.i3.697 17. Zhang R., Ye J., Huang H., Du X.. Mining Featured Biomarkers Associated with Vascular Invasion in HCC by Bioinformatics Analysis with TCGA RNA Sequencing Data. Biomed. Pharmacother. (2019.0) 118 109274. DOI: 10.1016/j.biopha.2019.109274 18. OLFM2 Protein Expression Summary—The Human Protein Atlas 19. Bertran L., Jorba-Martin R., Barrientos-Riosalido A., Portillo-Carrasquer M., Aguilar C., Riesco D., Martínez S., Vives M., Sabench F., Castillo D.D.. New Insights of OLFM2 and OLFM4 in Gut-Liver Axis and Their Potential Involvement in Nonalcoholic Fatty Liver Disease. Int. J. Mol. Sci. (2022.0) 23. DOI: 10.3390/ijms23137442 20. Miles J.M., Jensen M.D.. Counterpoint: Visceral Adiposity Is Not Causally Related to Insulin Resistance. Diabetes Care (2005.0) 28 2326-2328. DOI: 10.2337/diacare.28.9.2326 21. Armstrong M.J., Hazlehurst J.M., Hull D., Guo K., Borrows S., Yu J., Gough S.C., Newsome P.N., Tomlinson J.W.. Abdominal Subcutaneous Adipose Tissue Insulin Resistance and Lipolysis in Patients with Non-Alcoholic Steatohepatitis. Diabetes Obes. Metab (2014.0) 16 651-660. DOI: 10.1111/dom.12272 22. Freedland E.S.. Role of a Critical Visceral Adipose Tissue Threshold (CVATT) in Metabolic Syndrome: Implications for Controlling Dietary Carbohydrates: A Review. Nutr. Metab. (2004.0) 1 12. DOI: 10.1186/1743-7075-1-12 23. Ronquillo M.D., Mellnyk A., Cárdenas-Rodríguez N., Martínez E., Comoto D.A., Carmona-Aparicio L., Herrera N.E., Lara E., Pereyra A., Floriano-Sánchez E.. Different Gene Expression Profiles in Subcutaneous & Visceral Adipose Tissues from Mexican Patients with Obesity. Indian J. Med. Res. (2019.0) 149 616-626. DOI: 10.4103/ijmr.IJMR_1165_17 24. Pandžić Jakšić V., Grizelj D.. Under the Surface of Subcutaneous Adipose Tissue Biology. Acta Derm. Croat. (2016.0) 24 250-260 25. du Plessis J., van Pelt J., Korf H., Mathieu C., van der Schueren B., Lannoo M., Oyen T., Topal B., Fetter G., Nayler S.. Association of Adipose Tissue Inflammation With Histologic Severity of Nonalcoholic Fatty Liver Disease. Gastroenterology (2015.0) 149 635-648.e14. DOI: 10.1053/j.gastro.2015.05.044 26. Chiyanika C., Wong V.W.-S., Wong G.L.-H., Chan H.L.-Y., Hui S.C.N., Yeung D.K.W., Chu W.C.W.. Implications of Abdominal Adipose Tissue Distribution on Nonalcoholic Fatty Liver Disease and Metabolic Syndrome: A Chinese General Population Study. Clin. Transl. Gastroenterol. (2021.0) 12 e00300. DOI: 10.14309/ctg.0000000000000300 27. Item F., Konrad D.. Visceral Fat and Metabolic Inflammation: The Portal Theory Revisited. Obes. Rev. (2012.0) 13 30-39. DOI: 10.1111/j.1467-789X.2012.01035.x 28. Pafili K., Kahl S., Mastrototaro L., Strassburger K., Pesta D., Herder C., Pützer J., Dewidar B., Hendlinger M., Granata C.. Mitochondrial Respiration Is Decreased in Visceral but Not Subcutaneous Adipose Tissue in Obese Individuals with Fatty Liver Disease. J. Hepatol. (2022.0) 77 1504-1514. DOI: 10.1016/j.jhep.2022.08.010 29. Uciechowski P., Dempke W.C.M.. Interleukin-6: A Masterplayer in the Cytokine Network. Oncology (2020.0) 98 131-137. DOI: 10.1159/000505099 30. Pararasa C., Bailey C.J., Griffiths H.R.. Ageing, Adipose Tissue, Fatty Acids and Inflammation. Biogerontology (2015.0) 16 235-248. DOI: 10.1007/s10522-014-9536-x 31. Kawai T., Autieri M.V., Scalia R.. Adipose Tissue Inflammation and Metabolic Dysfunction in Obesity. Am. J. Physiol. Cell Physiol. (2021.0) 320 C375-C391. DOI: 10.1152/ajpcell.00379.2020 32. Finelli C., Tarantino G.. Is Visceral Fat Reduction Necessary to Favour Metabolic Changes in the Liver?. J. Gastrointestin. Liver Dis. (2012.0) 21 205-208. PMID: 22720311 33. Polyzos S.A., Kountouras J., Zavos C., Tsiaousi E.. The Role of Adiponectin in the Pathogenesis and Treatment of Non-Alcoholic Fatty Liver Disease. Diabetes Obes. Metab. (2010.0) 12 365-383. DOI: 10.1111/j.1463-1326.2009.01176.x 34. Choi H.M., Doss H.M., Kim K.S.. Multifaceted Physiological Roles of Adiponectin in Inflammation and Diseases. Int. J. Mol. Sci. (2020.0) 21. DOI: 10.3390/ijms21041219 35. Kleiner D.E., Brunt E.M., Van Natta M., Behling C., Contos M.J., Cummings O.W., Ferrell L.D., Liu Y.-C., Torbenson M.S., Unalp-Arida A.. Design and Validation of a Histological Scoring System for Nonalcoholic Fatty Liver Disease. Hepatology (2005.0) 41 1313-1321. DOI: 10.1002/hep.20701 36. Brunt E.M., Janney C.G., Di Bisceglie A.M., Neuschwander-Tetri B.A., Bacon B.R.. Nonalcoholic Steatohepatitis: A Proposal for Grading and Staging The Histological Lesions. Am. J. Gastroenterol. (1999.0) 94 2467-2474. DOI: 10.1111/j.1572-0241.1999.01377.x
--- title: Distinct Plasma Immune Profile in ALS Implicates sTNFR-II in pAMPK/Leptin Homeostasis authors: - Vincent Picher-Martel - Hejer Boutej - Alexandre Vézina - Pierre Cordeau - Hannah Kaneb - Jean-Pierre Julien - Angela Genge - Nicolas Dupré - Jasna Kriz journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049559 doi: 10.3390/ijms24065065 license: CC BY 4.0 --- # Distinct Plasma Immune Profile in ALS Implicates sTNFR-II in pAMPK/Leptin Homeostasis ## Abstract Amyotrophic lateral sclerosis (ALS) is a clinically highly heterogeneous disease with a survival rate ranging from months to decades. Evidence suggests that a systemic deregulation of immune response may play a role and affect disease progression. Here, we measured 62 different immune/metabolic mediators in plasma of sporadic ALS (sALS) patients. We show that, at the protein level, the majority of immune mediators including a metabolic sensor, leptin, were significantly decreased in the plasma of sALS patients and in two animal models of the disease. Next, we found that a subset of patients with rapidly progressing ALS develop a distinct plasma assess immune–metabolic molecular signature characterized by a differential increase in soluble tumor necrosis factor receptor II (sTNF-RII) and chemokine (C-C motif) ligand 16 (CCL16) and further decrease in the levels of leptin, mostly dysregulated in male patients. Consistent with in vivo findings, exposure of human adipocytes to sALS plasma and/or sTNF-RII alone, induced a significant deregulation in leptin production/homeostasis and was associated with a robust increase in AMP-activated protein kinase (AMPK) phosphorylation. Conversely, treatment with an AMPK inhibitor restored leptin production in human adipocytes. Together, this study provides evidence of a distinct plasma immune profile in sALS which affects adipocyte function and leptin signaling. Furthermore, our results suggest that targeting the sTNF-RII/AMPK/leptin pathway in adipocytes may help restore assess immune–metabolic homeostasis in ALS. ## 1. Introduction Amyotrophic lateral sclerosis (ALS) is an adult onset neurodegenerative disease characterized by a progressive loss of upper and lower motor neurons [1,2]. Most ALS cases are classified as sporadic (sALS ± $90\%$), while the remaining ±$10\%$ are familial (fALS) [3]. Clinical presentations and evolution of the disease are highly heterogeneous between patients [4], even between those carrying the same genetic profile [5]. Although ALS is invariably lethal, approximately $20\%$ of the patients survive >5 years and $10\%$ may survive >10 years after the onset of the first symptoms [6]. On the other side of the spectrum, it has been observed that 10–$20\%$ of patients develop a rapidly progressing disease leading to death in less than a year following initial diagnosis [7]. The underlying cause for this high variability in the clinical course of disease remains unclear. The combination of genetic and/or environmental modifying factors may influence the rate of disease progression [8,9]. For instance, juvenile onset or younger age at diagnosis, upper motor neuron-predominant and some form of hereditary ALS are associated with longer survival [6,10,11], while, bulbar onset, low forced vital capacity (FVC < $80\%$) and lower revised ALS functional Rating Scale (ALSFRS-R) at the first visit are associated with shorter survival [11]. At present, growing evidence suggests that a chronic dysregulation of immunity may significantly affect the course of disease as well as metabolic homeostasis in ALS. It is noteworthy that, one of the conditions associated with a faster rate of functional decline and shorter survival is development of a hypermetabolic state in ALS [12,13]. In the current study, we took advantage of an unbiased multiplex metabolic/immune array approach and measured, at the protein level, sixty-two different cytokines/chemokines/adipokines in the plasma samples collected from sporadic ALS patients (sALS). The proof-of concept sample analyses revealed that some of the key immune mediators were either downregulated and/or not regulated. To our surprise, the metabolic sensor leptin was identified as the most consistently deregulated protein in the plasma of sALS samples. Of note, the observed deregulation was more pronounced in male patients. Similar observations were made in different mouse models of ALS, notably the SOD1G93A mice and the double transgenic TDP-43G348C; UBQLN2P497H mice. Normally implicated in regulation of food intake, growing evidence suggests a role for leptin in inflammation, neuroprotection, as well as learning and cognitive functions [14,15,16,17,18]. Prior reports have suggested a potential role of leptin in ALS pathogenesis, as its levels are inversely associated with the risk of developing disease [19,20]. Furthermore, studies in the TDP-43A315T mouse model revealed a reduction in leptin levels at the end-stage of disease [21], while treatment with the recombinant protein improved motor performance and delayed weight loss in this mouse line [22]. Next, we identified a unique immune–metabolic plasma molecular signature associated with fast progressing ALS, characterized by: (i) elevated levels of soluble TNF receptor II (sTNF-RII) and C-C motif chemokine ligand 16 (CCL16), (ii) further decrease in plasma levels of leptin when compared to slow sALS. In a search for underlying mechanisms, a series of in vitro experiments revealed that human adipocytes exposed to plasma of fast sALS and/or recombinant sTNF-RII reproduced a similar immune–metabolic response resulting in a marked downregulation of leptin secretion. The observed reduction in leptin correlated with an increase in AMP-activated protein kinase (AMPK) phosphorylation levels in adipocytes, a principal sensor implicated in leptin production. Conversely, treatment with AMPK inhibitors restored leptin production in human adipocytes. Together, this study provides in vitro and in vivo evidence of a unique immune/metabolic profile in plasma of sALS, and in particular fast progressing ALS patients, which alters adipocyte function and leptin homeostasis. ## 2. Results Growing evidence suggests that a chronic deregulation of immunity may represent one of the key elements in the pathobiology of neurodegenerative disorders such as ALS and it may contribute to the observed heterogeneity in the rate of disease progression and/or regulation of metabolic homeostasis [23,24,25]. To assess immune–metabolic profiles in ALS, we recruited 51 sALS patients and 38 age-matched controls. Patients were eligible for inclusion if they had a definite or probable diagnosis of ALS based on El-Escorial criteria, were aged at least 25 years, and had no familial history or genetic cause of ALS. Demographics were very similar between patients and controls for weight, body mass index (BMI) and age (Table 1). As presented in Table 2, the patients were categorized as slow and fast progressors based on difference in ALSFRS-R score between two-time points. Patients were classified as fast progressors if they lost more than 4 points/12 weeks in ALSFRS-R [16]. Using these criteria, 11 patients were classified as fast progressors with a mean loss of 7.884 points/12 weeks as compared to a mean loss of 1.683 points/12 weeks in slow progressors ($p \leq 0.0001$). ## 2.1. Reduced Levels of Immune Mediators and the Metabolic Sensor Leptin in Plasma of Sporadic ALS Patients To assess the effects of disease on immune and metabolic signaling in sporadic ALS, we took advantage of an unbiased screening array and measured plasma levels of 62 different cytokines and adipokines. To our surprise, quantitative analysis revealed that the protein levels of the majority of immune mediators including different cytokines and chemokines were decreased in plasma of sALS patients as compared to controls (Figure 1a). We detected a significant reduction in levels of leukemia inhibitory factor (LIF, −$8.64\%$, $p \leq 0.0001$) (Figure 1c), tissue inhibitor of metalloproteinase 1 (TIMP-1, −$7.70\%$, $$p \leq 0.0002$$) (Figure 1d), serum amyloid A (SAA, −$10.03\%$, $$p \leq 0.0006$$) (Figure 1e), chemokine (c-c motif) ligands 4 (CCL4, −$8.57\%$, $$p \leq 0.0005$$) (Figure 1f), TIMP-2 (−$9.90\%$, $$p \leq 0.0058$$) (Figure 1g), interferon-gamma (IFN-γ, −$6.24\%$, $$p \leq 0.0020$$) (Figure 1h), tumor necrosis factor-alpha (TNF-α, −$8.01\%$, $$p \leq 0.0007$$) (Figure 1i), CCL2 (−$6.16\%$, $$p \leq 0.0138$$) (Figure 1k) and metabolic sensor leptin (−$17.95\%$, $$p \leq 0.0006$$) (Figure 1b). This is in accordance with previous reports suggesting lower leptin levels in ALS patients [19]. Of note, correlation has been found between leptin levels and BMI (R2 = 0.08669, $$p \leq 0.0422$$) (Figure S1A). Importantly, leptin levels did not change in patients treated with riluzole (Figure S1D). ## 2.2. Differential Increase in Plasma Levels of sTNF-RII and CCL16 Is Associated with Fast Progressing Disease Next, we searched for potential differences in immune/metabolic profiles between distinct clinical subgroups of sALS patients classified as slow (<4 points/12 weeks lost) and fast progressors (>4 points/12 weeks lost), slow and fast ALS, respectively. Comparative analyses revealed a significant difference in the plasma levels of distinct immune mediators. Levels of CCL16 (+$9.28\%$, $$p \leq 0.0069$$) (Figure 2c) and soluble TNF receptor 2 (sTNF-RII, +$8.78\%$, $$p \leq 0.0425$$) (Figure 2d) were significantly increased in the plasma of fast ALS when compared to slow ALS and/or controls. In addition, we observed a marked tendency for lower plasma levels of leptin in fast ALS (Figure 2b). These findings were further confirmed by ELISA assay. As shown in Figure 3a, quantitative analyses revealed a significant increase in reduction in levels of leptin in sALS patients (12.53 ng/mL) as compared to age-matched controls (17.83 ng/mL) ($$p \leq 0.0253$$) (Figure 3a). Furthermore, leptin levels were significantly reduced in plasma of fast (4.787 ng/mL) as compared to slow ALS (14.91 ng/mL) ($$p \leq 0.0051$$) (Figure 3b). Next, in accordance with the results obtained in protein array analyses, CCL16 and sTNF-RII levels were significantly increased in plasma of fast ALS (13.01 ng/mL) as compared to slow ALS (9.994 ng/mL) ($$p \leq 0.0204$$) (Figure 3c) and (3.093 ng/mL vs. 2.563 ng/mL) ($$p \leq 0.0464$$) (Figure 3d), respectively. No correlation was found between CCL16 or sTNF-RII levels and BMI or riluzole treatment (Figure S1B,C,E,F). Further analyses revealed significant sex-specific differences in the plasma concentration of leptin in sALS. The levels of leptin in plasma of female sALS patients were highly variable and not significantly changed when compared to age-matched female controls. In contrast, plasma levels of leptin where significantly decreased in all male sALS patients as compared to age-matched male controls ($$p \leq 0.0058$$), age-matched female controls ($p \leq 0.0001$) and female sALS patients ($p \leq 0.0001$) (Figure 3e). Furthermore, a comparative analysis of leptin levels in the plasma of fast vs. slow ALS male patients revealed a significantly lower concentration of leptin in male patients suffering from the fast progressing disease (Figure 3f). Of note, although we observed an important tendency towards lower plasma leptin levels in fast ALS female patients, the number of fast progressing female sALS patients was too low to achieve statistical significance. Importantly, we observed no sex-specific differences in plasma levels of CCL16 and sTNF-RII. ## 2.3. Early Deregulation of Leptin in Plasma of SOD1G93A Mice We next investigated whether changes in metabolic and immune profiles observed in human sporadic disease were also replicated in the model of inherited disease, the SOD1G93A mouse [5,26]. By using a mouse model with a predictable disease onset at approximately 100 days of age, we investigated whether certain changes in immune and/or metabolic profile, such as deregulation of leptin occur early in disease, i.e., prior to the onset of clinical symptoms. The assessed immune–metabolic profiles were analyzed at three different time points of disease. As revealed in Figure 4a, very early in disease and before the onset of clinical symptoms (pre-onset), the plasma levels of the majority of cytokines/chemokines were either reduced and/or unchanged (Figure 4a). The levels of leptin in plasma of the SOD1G93A mice were significantly decreased as compared to age-matched non-transgenic controls ($$p \leq 0.0311$$) (Figure 4b). We also observed a decrease in the plasma levels of IL-1β ($$p \leq 0.0140$$), IL-2 ($$p \leq 0.0067$$), IL-3 ($$p \leq 0.0443$$) and IL-9 ($$p \leq 0.0237$$) (Figure 4f–i), while the levels of sTNF-RII, CCL1, CCL2, IFN-γ, TIMP-1, TIMP-2 and TNF-α remained unchanged (Figure 4d,e,j–m). As disease progressed, at the time of the clinical onset of disease (mild stage) as well as in advanced disease, the plasma leptin levels remained significantly reduced in SOD1G93A mice as compared to non-transgenic mice ($$p \leq 0.0027$$), Figure 4o, and ($$p \leq 0.0006$$), Figure 4n, respectively. See Figures S2 and S3 for other variable changes in cytokine levels associated with disease evolution. We next analyzed plasma levels of leptin in two additional experimental models of ALS carrying different disease causing mutations, including TDP-43G348C mice and the double transgenic TDP-43G348C; UBQLN2P497H mice [27]. Leptin levels were significantly reduced in the double transgenic UBQLN2P497H; TDP-43G348C mice at the mild symptomatic stage (8 months of age) ($$p \leq 0.0311$$), but not in the single transgenic TDP-43G348C mice (Figure S4A). Of note, the double transgenic mice exhibit motor deficits with associated neuronal loss, whereas single transgenic TDP-43G348C mice develop a frontotemporal dementia-like phenotype characterized by age-dependent cognitive decline and do not exhibit motor neuron death. Finally, we analyzed and compared leptin levels in plasma of male and female ALS mice. Interestingly, in early disease, the plasma levels of leptin were lower in male SOD1G93A mice as compared to female age-matched mice (Figure S4B). This tendency is not observed in more advanced stages of the disease, since both sexes exhibit a significant disease associated decrease in the plasma leptin levels (Figure S4C,D). ## 2.4. Downregulation of Leptin Correlates with Hyperactivation of the AMPK Pathway in Adipose Tissue of the SOD1G93Amice Leptin production in adipocytes is in part controlled by the metabolic sensor AMPK and mammalian target of rapamycin (mTOR) pathways [28]. However, to what extent changes in AMPK activation patterns and phosphorylation at the periphery (i.e., in adipose tissues and/or adipocytes) are associated with a marked reduction in plasma levels of leptin observed in human and mouse disease remains elusive. First, we analyzed levels of leptin in adipocyte extracts by ELISA in pre-onset, mild stage, and advanced stage SOD1G93A mice. As shown in Figure 5a,b, early in disease we observed a small but significant decrease in leptin levels in male SOD1G93A adipocytes (Figure 5a,b) and this was peaking at the time of disease onset (Figure 5e,f). To our surprise, in advanced disease, we did not detect a reduction in leptin production by adipocytes (Figure 5i,j), suggesting that it may represent early disease pathogenesis. As further demonstrated in Figure 5g,h, Western blot analysis of adipose tissue homogenates revealed an increase in phosphorylated-AMPK adipocyte extracts in both, male and female mice (Figure 5c,d,h). To our surprise, in advanced disease, we did not detect a reduction in leptin production by adipocytes and observed an overall increase in leptin levels in both non-transgenic and SOD1G93A mice compared to younger mice (Figure 5i,j). The discrepancy between the observed downregulation of plasma leptin in SOD1G93A mice and the similar adipocyte leptin secretion may potentially be explained by the noticeable fat atrophy, which leads to a total reduction in plasma leptin. Residual fat tissue may compensate by increasing its production of leptin. Indeed, a previous report has shown that end-stage TDP-43A315T mice exhibit an increase in adipocyte leptin mRNA levels associated with a downregulation of leptin plasma levels [21]. The observed increase in leptin production in non-transgenic mice may be explained by the significant fat gain in C57Bl/6 at this age. Together these findings suggest that adipocytes’ metabolism, leptin homeostasis and AMPK activation/function may be implicated in early pathogenic mechanisms in ALS. ## 2.5. Production of Leptin in Human Adipocytes Is Regulated by AMPK To further investigate the mechanisms involved in (de)-regulation of leptin secretion in ALS, we created a controlled humanized in vitro model-system. As shown in Figure 6a, the human adipocytes were differentiated and matured (contained lipid droplets) as revealed by positive Oil Red O coloration. The adipocytes were then conditioned for 12 h with pooled plasma samples from 12 healthy controls, 12 slow and 6 fast ALS. Importantly, mimicking the findings obtained in human disease in vivo, the adipocytes treated with fast ALS plasma samples had reduced production of leptin as compared to healthy controls and slow ALS (Figure 6b). In addition, the adipocytes exposed to plasma of the fast ALS expressed significantly higher levels of pAMPK as compared to slow progressors and controls (Figure 6d). As AMPK acts as the principal sensor implicated in leptin production, next we tested whether reduction in leptin production/secretion could be abrogated by the inhibition of AMPK. The adipocytes were pretreated for 1 h with an AMPK inhibitor (compound C, 10 μM) or a mTOR inhibitor (PP242, 1 μM) and then exposed for 12 h to different plasma samples. The mTOR inhibitors were used as a negative control since the mTOR pathway would normally be inhibited in the context of AMPK hyperphosphorylation. As shown in Figure 6d, AMPK inhibitors were efficient in reducing AMPK phosphorylation. Further analysis revealed that the pre-treatment with AMPK inhibitor restored leptin levels in human adipocytes conditioned with plasma from slow ($p \leq 0.0001$) and fast ALS ($$p \leq 0.0169$$) (Figure 6c). As expected, the pre-treatment with mTOR inhibitors had no impact on leptin secretion. ## 2.6. sTNF-RII Reduces Leptin Production in Human Adipocytes Our results so far suggest that the deregulation of immune profile observed in fast ALS, may have an additional impact on leptin production by adipocytes. Given that sTNF-RII and/or CCL16 are differentially increased in plasma of the fast progressing patients, we investigated their impact on leptin production in cultured adipocytes. To maximize leptin production and release, human adipocytes were treated with insulin which activates leptin release via PI3K/AKT activation in humans [29,30]. Next, the cells were exposed to different concentrations of recombinant proteins sTNF-RII and/or CCL16. As shown in Figure 6e, sTNF-RII blocked the insulin-driven production of leptin at 1000 ng/mL while treatment with CCL16 did not have a significant impact on leptin production (Figure 6f). Furthermore, exposure of adipocytes to sTNF-RII significantly increased the levels of pAMPK in a dose-dependent manner (p ≤ 0.0002) (Figure 6g,h). ## 3. Discussion Chronic deregulation of immunity is a hallmark of many neurodegenerative disorders including ALS. Here, we provide in vivo evidence of a chronic deregulation of plasma immune profile and leptin homeostasis in human disease, as well as in an experimental ALS model, notably the SOD1G93A mouse. We report a reduction, at the protein level, of several immune mediators (LIF, TIMP-1, TIMP-2, SAA, MIP-1β, IFN-γ, TNF-α and MCP-1) together with a marked and consistent decrease in the levels of the metabolic sensor leptin. A similar molecular profile was observed in plasma of the SOD1G93A mice. Importantly, both human and mice data suggested a shutdown of peripheral immune response/signaling, and disease evolution associated with a marked deregulation of leptin homeostasis. To date, the role of immunity in ALS remains controversial. Immune markers have been mainly studied in CSF and much evidence suggests increased levels of several cytokines/chemokines, including IL-2, IL-6, IL-8,IL-10, MCP-1, IL-18 IL-15, MIP-1β, MIP-1α and IFN-γ [31,32,33,34]. However, several conflicting results were obtained following analyses of immune markers in the blood of ALS patients. For example, IFN-γ was found to be increased in some studies [35,36,37], while Lu et al., and Polverino et al., found decreased levels of IFN-γ at the periphery [38,39]. The same discrepancy applies to TNF-α [37,39,40], IL-6 [41,42,43] and others [44]. Furthermore, while cell specific analysis of the immune profiles of peripheral blood monocytes from patients with ALS revealed a marked proinflammatory phenotype at the RNA level [45], some of the observed findings did not correlate with measured proteins, suggesting a certain dissociation of immune profiles at the RNA and proteins levels. The conflicting results in the literature could be in part explained by the variable approaches used for measurement, including individual ELISA, cell population studies, RT-qPCR or multiplex assays of several cytokines [44]. Our study may have had several technical advantages. First, we used an unbiased approach measuring numerous cytokines and metabolic markers at the protein level instead of targeting a unique cytokine. Second, the analysis was always performed in less than 1-week-old samples, reducing the risk of protein degradation and variability in both groups. Of importance, in patients with fast progressing disease we observed a differential increase in sTNF-RII and CCL16 plasma levels and further decrease in the plasma levels of leptin, i.e., significantly decreased leptin when compared to plasma levels observed in slow progressing ALS. Interestingly, we observed marked sex differences in the plasma levels of leptin in both human patients and SOD1G93A mice. Leptin levels were significantly decreased in the plasma of male, but not in female sALS. It has been established that at any age, leptin levels in females are generally $40\%$ higher than in male [46,47]. At present, the underlying causes of this dimorphism remain unclear, but it is thought to be associated with the fat metabolism associated with reproductive function and steroid levels but doesn’t seem to be associated with body mass index nor total body fat [48]. While it is possible that the observed sex differences may in part explain variability in the plasma leptin levels detected in sALS patients, nevertheless, our data strongly suggests that deregulation of leptin homeostasis remains a pathogenic factor in the subset of sALS patients with fast progressing disease. Indeed, leptin levels were reported to be inversely correlated with disease onset and progression in ALS, suggesting a protective role of leptin in the disease [19]. At present, it is unclear how chronic deregulation of leptin homeostasis in sALS may contribute to disease pathogenesis. Leptin acts on the hypothalamus to regulate energy balance and food intake, but evidence suggests its involvement in some central nervous system pathways. Leptin has been related to neuroprotection after spinal cord injury or stroke [15,49]. It reduces amyloid load and tau phosphorylation in Alzheimer’s disease [50,51,52], as well as dopaminergic cell death in Parkinson’s disease [53], and favors neurogenesis and synaptogenesis [54]. Finally, in accordance with previous work, the results of our study revealed that lower plasma leptin levels may represent a risk factor associated with the faster rate of disease progression [19]. While here we aimed to explore the interactions between immune response and the leptin/pAMPK signaling at the periphery, previous work suggests that leptin may protect against ALS by its direct action on motor neurons and/or its modulation on glial cells’ activity [55], or potentially by its impact on the hypothalamic secretion of different bioactive peptides [56]. The important question here is whether and/or to what extent alterations in the peripheral immunity contribute to the observed deregulation of leptin homeostasis in ALS, in particular, in patients with rapidly progressing disease. Here, it is noteworthy that, in our hands, elevated plasma levels of sTNF-RII were detected in the subset of patients with more rapidly progressing disease. Although sTNFR-II has been previously shown to be elevated in ALS, to date, there are no comprehensive studies analyzing its role on the rate of disease progression [57]. Previous work suggests that under pathological conditions associated with chronic inflammation such as multiple sclerosis, type 2 diabetes and/or cardiovascular disease, TNF-RII is detached from the cell surface through activation of the tumor necrosis factor-alpha converting enzyme (TACE), thus promoting the aberrant immune and noxious response of mononuclear cells [58,59,60,61,62,63]. Furthermore, results from our study strongly suggest that higher plasma levels of sTNFR-II, together with the concurrent hyperactivation of AMPK signaling contribute to reduced production of leptin in ALS-affected adipocytes. Many lines of evidence support this conclusion: (i) cultured human adipocytes released less leptin when conditioned with plasma from fast ALS; (ii) inhibition of AMPK signaling in human adipocytes exposed to plasma from fast ALS restored leptin production; (iii) treatment of human adipocytes with recombinant sTNF-RII induced overexpression of pAMPK and AMPK signaling and blocked leptin production and (iv) the lower levels of leptin correlated with the levels of pAMPK in SOD1G93A mouse adipocytes. While in ALS models, the role of AMPK has not been extensively studied at the periphery, previous work has suggested increased AMPK activity in SOD1 spinal cord culture and a motor neuron cell line [64,65,66]. Indeed, AMPK is hyperphosphorylated in SOD1G93A spinal cord lysates and in C. elegans SOD1G85R motor neurons [64,67]. AMPK hyperactivation was also detected in motor neurons of ALS patients [66]. At present, the pharmacological regulation of AMPK has retrieved conflicting results and it remains unclear whether AMPK activation is deleterious or beneficial in ALS. Indeed, activation of AMPK in SOD1G93A accelerated disease onset and progression in female mice [68]. Other data suggested that phospho-AMPK is reduced in mesenchymal stem cells from ALS patients and that AMPK activation may restore neuronal differentiation potency [69]. Taken together, our results illustrate that AMPK may have a distinct pathological profile/function in adipocytes as compared to neurons and other tissues. The specific inhibition of phospho-AMPK in adipocytes may restore leptin production and thus increase levels of leptin, which appear to be related to disease progression. In summary, we described a unique immune/metabolic profile in ALS patients and SOD1G93A mice. Using an unbiased approach we identified leptin as the most dysregulated assessed immune–metabolic mediator in plasma of sporadic sALS patients, in particular in men with rapidly progressing disease. Next, we showed that exposure to plasma from the fast progressing patients may have a direct impact on human adipocytes’ metabolism and leptin production/secretion via sTNFRII/AMPK signaling. Together, our results suggest that targeting the sTNF-RII/AMPK/leptin pathway in adipocytes may help restore metabolic homeostasis and potentially reduce the rate of decline in ALS patients with rapidly progressing disease. ## 4.1. Recruitment and Samples Preparation Patients were eligible for inclusion if they had a definite or probable diagnosis of ALS based on El-Escorial criteria, were aged at least 25 years, and had no familial history or genetic cause of ALS. Controls were generally the husband/wife of the patients, when willing to participate. Samples were collected at patients’ homes using EDTA collecting tubes. Samples were centrifuged at 10,000 RPM for 10 min. The supernatant was collected and snap-frozen in liquid nitrogen. This study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board (or Ethics Committee) of “CHU de Québec” (ethic code: 2021-5742, protocol renewal approved on 27 May 2022). ## 4.2. Human and Mouse Cytokines Array The Human Obesity Array C1 (RayBiotech, Peachtree Corners, GA, USA, #AAH-ADI-1-8) was performed on human plasma samples, less than one week from sample collection. The array was conducted according to the manufacturer’s protocol. Briefly, the samples were diluted in blocking buffers (1:10). After blocking, membranes were incubated overnight with biotinylated antibody cocktails followed by two hours with HRP-streptavidin. After washing, membranes were developed using high-resolution films. Cytokine intensities were measured using ImageLab Touch Software; Version 2.4.0.03. Each membrane contains six positive internal controls used for data normalization. The RayBiotech analysis tool employs positive controls in one membrane and normalizes to the positive controls in every membrane and therefore, normalizes the cytokine levels, ensuring consistency between every array performed and between both groups. Here, normalization without background suppression was used. In all experiments using human samples we used one membrane per sample. The same protocol was used for comparative analysis of mouse plasma samples, Cytokine Array C3 (RayBiotech, Peachtree Corners, GA, USA, #AAM-INF-1-8). ## 4.3. Experimental Animals Experiments were performed on wild-type non-transgenic (C57Bl/6), pre-symptomatic, mild and advanced symptomatic SOD1G93A, as well as mild symptomatic 8 months of age TDP-43G348C and double transgenic UBQLN2P497H; TDP-43G348C mice. Many of the experiments were performed on SOD1G93A mice since this model develops most of the disease characteristics. SOD1G93A mice (B6SJL-TgN (SOD1G93A)1Gur/j) were acquired from the Jackson Laboratory (Bar Harbor, ME, USA) and genotyped as suggested by Jackson Laboratory protocols. TDP-43G348C and UBQLN2P497H; TDP-43G348C transgenic mice were generated and genotyped as described in [27], respectively. TDP-43G348C develops cognitive deficits without motor impairment but double transgenic mice develop both cognitive and motor impairment from 8 months of age. Both males and females were used for experiments. All the experimental procedures were approved by the Laval University Animal Care Ethics Committee and are in accordance with the Guide to the Care and Use of Experimental Animals of the Canadian Council on Animal Care. ## 4.4. Blood and Tissue Collection, Protein Extraction, and Immunoblotting Mice were sacrificed at 50, 100 or 150 days of age to collect samples for plasma and tissue analysis. Mice were anesthetized via an intraperitoneal injection of ketamine/xylazine (100–10 mg/kg). Blood was removed from the mice by direct sampling from the heart and mice were then slowly perfused with saline. Blood samples were centrifuged at 10,000 RPM for 10 min and supernatant was kept for analysis. Abdominal fat was removed and snap-frozen in liquid nitrogen for protein extraction. Abdominal fat was homogenized in buffer (20 mM tris pH 7.8, 137 mM NaCl, 2.7 mM KCl, 1 mM MgCl2, $1\%$ triton X-100, $10\%$ glycerol, 1mM EDTA, 1mM dithiothreitol and 1X proteases and phosphatase inhibitor cocktails). The lysate was sonicated and incubated on ice for 30 min and centrifuged at 13,000 rpm for 30 min and supernatant kept for analysis. The top lipid layer after centrifugation was not collected with the supernatant. Antibodies used for immunoblotting were phospho-AMPKα (Cell signaling, Danvers, MA, USA, # 2531), AMPKα (Cell signaling, Danvers, MA, USA, # 2532) and GAPDH (Santa-Cruz, Dallas, TX, USA, Sc-32233). The immunoblots were developed using the ChemiDoc MP Imaging System (Bio-Rad, Hercules, CA, USA) and the ImageLab Touch Software; Version 2.4.0.03. ## 4.5. Pre-Adipocyte Culture and Treatment Human primary subcutaneous pre-adipocytes ATCC, (PCS-210-010) were cultured in T75 flasks with fibroblast basal medium containing growth kit-low serum (ATCC, Manassas, VA, USA, # PCS-201-041) (5 ng/mL rh FGFb, 7.5 mM L-glutamine, 50 mg/mL ascorbic acid, 1 mg/mL hydrocortisone/hemisuccinate, 5 mg/mL rh insulin and $2\%$ fetal bovine serum). Maintenance was performed by changing medium every 48 h until cells reached $80\%$ confluence and were ready for sub culturing. Cells were washed using 5 mL D-PBS, trypsinized and split into multiple T75 flasks for amplification. When pre-adipocytes reached $80\%$ confluence, they were trypsinized again and split into 6-well plates with approximately 170,000 cells/well with 2 mL of fibroblast basal medium. After 48 h, we began the initiation phase of the adipocyte differentiation procedure by removing old media and adding 2 mL of adipocyte differentiation initiation medium (ATCC, Manassas, VA, USA, #PCS-500-050) (15 mL adipocyte basal medium (grow kit-low serum and 1 mL AD supplement). A total of 1 mL of media was removed and replaced with 2 mL of fresh adipocyte differentiation initiation medium. From this step forward, adipocytes were never exposed to air to ensure that lipid vesicles did not burst, as suggested by the manufacturer. After 48 h, 2 mL of media was replaced with adipocyte differentiation maintenance medium (ATCC, Manassas, VA, USA, #PCS-500-050) (85 mL basal medium with 5 mL ADM supplement). This step was repeated every 72 h for a total of 15 days from initiation phase until adipocytes reached full maturity. To examine the impact of sALS plasma on leptin production, mature human adipocytes were treated with 1:100 diluted pooled plasma (in differentiation maintenance medium) from either 12 healthy controls, 12 slow ALS or 6 fast ALS. Previous reports have shown the feasibility of this approach on neuronal cells using different dilutions ranging from $1\%$ to $50\%$ [70,71]. Cells were exposed to plasma for 12 h, to optimize leptin secretion [72]. The mature adipocytes were then treated with 1:100 diluted pooled plasma samples from either 12 healthy controls, 12 slow ALS, or 6 fast ALS (in differentiation maintenance medium). Cells were exposed to plasma for 12 h. When pre-treatment was conducted, adipocytes were treated with 10 μm of compound C (Sigma-Aldrich, St-Louis, MO, USA, #171261) or 1 μm of PP242 (Sigma-Aldrich, St-Louis, MO, USA, # 475988) for 1 h before addition of plasma samples. Cells were also treated with different concentrations of recombinant CCL16 (R&D systems, Minneapolis, MN, USA, # 802-HC-025) or recombinant sTNF-RII (MyBioSource, MBS343136, London, UK). After treatment, media was collected, and cells were harvested for future analysis. ## Oil-Red O Coloration The adipocytes were also cultured on a 10 mm coverslip to perform Oil-red O coloration and assure full maturity. Cells were fixed using $4\%$ paraformaldehyde (PFA) for 30 min and washed four times with PBS. The cells were rinsed with isopropanol and stained with Oil Red O/ isopropanol solution for 15 min. Finally, cells were rinsed with isopropanol and distilled water. ## 4.6. ELISA All ELISAs were performed as suggested by the manufacturer’s protocol. Leptin plasma levels were measured using a human leptin ELISA kit (Invitrogen, Waltham, MA, USA, #KAC2281) with samples diluted 1:100. Each patient’s samples were processed in duplicate. The same kit was used to measure leptin levels in cultures of human adipocytes’ undiluted supernatant. CCL16 plasma levels were measured using a CCL16 ELISA kit (Invitrogen, Waltham, MA, #EHCCL16) with 1:10 dilution. sTNF-RII plasma levels were measured using human sTNF-RII quantikine ELISA kit (R&D system, Minneapolis, MN, USA, #DRT200) with 1:10 dilution. Finally, a mouse leptin ELISA kit (Invitrogen, Waltham, MA, USA, #KMCC2281) was used to measure leptin levels in mouse adipocyte extracts (dilution 1:20). ## 4.7. Statistics We used student’s unpaired t-test for cytokines array in human and mouse plasma samples (Figure 1, Figure 2 and Figure 4) and for ELISA analysis in Figure 3 and Figure 5. Two-way ANOVA was used for analysis (Figure 6). ## References 1. Al-Chalabi A., Hardiman O., Kiernan M.C., Chiò A., Rix-Brooks B., van den Berg L.H.. Amyotrophic lateral sclerosis: Moving towards a new classification system. Lancet Neurol.* (2016) 15 1182-1194. DOI: 10.1016/S1474-4422(16)30199-5 2. Mejzini R., Flynn L.L., Pitout I.L., Fletcher S., Wilton S.D., Akkari P.A.. ALS Genetics, Mechanisms, and Therapeutics: Where Are We Now?. Front Neurosci. (2019) 13 1310. DOI: 10.3389/fnins.2019.01310 3. Chia R., Chiò A., Traynor B.J.. Novel genes associated with amyotrophic lateral sclerosis: Diagnostic and clinical implications. Lancet Neurol. (2018) 17 94-102. DOI: 10.1016/S1474-4422(17)30401-5 4. Fournier C., Glass J.D.. Modeling the course of amyotrophic lateral sclerosis. Nat. Biotechnol. (2015) 33 45-47. DOI: 10.1038/nbt.3118 5. Picher-Martel V., Valdmanis P.N., Gould P.V., Julien J.-P., Dupré N.. From animal models to human disease: A genetic approach for personalized medicine in ALS. Acta Neuropathol. Commun. (2016) 4 1-29. DOI: 10.1186/s40478-016-0340-5 6. Pupillo E., Messina P., Logroscino G., Beghi E.. Long-term survival in amyotrophic lateral sclerosis: A population-based study. Ann. Neurol. (2014) 75 287-297. DOI: 10.1002/ana.24096 7. Swinnen B., Robberecht W.. The phenotypic variability of amyotrophic lateral sclerosis. Nat. Rev. Neurol. (2014) 10 661-670. DOI: 10.1038/nrneurol.2014.184 8. Logroscino G.. Classifying change and heterogeneity in amyotrophic lateral sclerosis. Lancet Neurol. (2016) 15 1111-1112. DOI: 10.1016/S1474-4422(16)30206-X 9. Sabatelli M., Conte A., Zollino M.. Clinical and genetic heterogeneity of amyotrophic lateral sclerosis. Clin. Genet. (2013) 83 408-416. DOI: 10.1111/cge.12117 10. Zoccolella S., Beghi E., Palagano G., Fraddosio A., Guerra V., Samarelli V., Lepore V., Simone I.L., Lamberti P., Serlenga L.. Analysis of survival and prognostic factors in amyotrophic lateral sclerosis: A population based study. J. Neurol. Neurosurg. Psychiatry (2008) 79 33-37. DOI: 10.1136/jnnp.2007.118018 11. Kollewe K., Mauss U., Krampfl K., Petri S., Dengler R., Mohammadi B.. ALSFRS-R score and its ratio: A useful predictor for ALS-progression. J. Neurol. Sci. (2008) 275 69-73. DOI: 10.1016/j.jns.2008.07.016 12. Steyn F.J., Ioannides Z.A., Van Eijk R.P.A., Heggie S., Thorpe K.A., Ceslis A., Heshmat S., Henders A.K., Wray N.R., Berg L.H.V.D.. Hypermetabolism in ALS is associated with greater functional decline and shorter survival. J. Neurol. Neurosurg. Psychiatry (2018) 89 1016-1023. DOI: 10.1136/jnnp-2017-317887 13. Kasarskis E.J., Mendiondo M.S., Matthews D.E., Mitsumoto H., Tandan R., Simmons Z., Bromberg M.B., Kryscio R.J.. Estimating daily energy expenditure in individuals with amyotrophic lateral sclerosis. Am. J. Clin. Nutr. (2014) 99 792-803. DOI: 10.3945/ajcn.113.069997 14. Fernandez-Martos C.M., Atkinson R.A.K., Chuah M.I., King A.E., Vickers J.C.. Combination treatment with leptin and pioglitazone in a mouse model of Alzheimer’s disease. Alzheimer’s Dement. (2017) 3 92-106. DOI: 10.1016/j.trci.2016.11.002 15. Fernández-Martos C.M., González P., Rodriguez F.J.. Acute leptin treatment enhances functional recovery after spinal cord injury. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0035594 16. Farr S.A., Banks W.A., Morley J.E.. Effects of leptin on memory processing. Peptides (2006) 27 1420-1425. DOI: 10.1016/j.peptides.2005.10.006 17. Santos-Alvarez J., Goberna R., Sánchez-Margalet V.. Human leptin stimulates proliferation and activation of human circulating monocytes. Cell. Immunol. (1999) 194 6-11. DOI: 10.1006/cimm.1999.1490 18. Gabay C., Dreyer M.G., Pellegrinelli N., Chicheportiche R., Meier C.A.. Leptin directly induces the secretion of interleukin 1 receptor antagonist in human monocytes. J. Clin. Endocrinol. Metab. (2001) 86 783-791. DOI: 10.1210/jc.86.2.783 19. Nagel G., Peter R.S., Rosenbohm A., Koenig W., Dupuis L., Rothenbacher D., Ludolph A.C.. Adipokines, C-reactive protein and Amyotrophic Lateral Sclerosis—Results from a population-based ALS registry in Germany. Sci. Rep. (2017) 7 4374. DOI: 10.1038/s41598-017-04706-5 20. Ngo S.T., Steyn F.J., Huang L., Mantovani S., Pfluger C.M.M., Woodruff T.M., O’Sullivan J., Henderson R., McCombe P.. Altered expression of metabolic proteins and adipokines in patients with amyotrophic lateral sclerosis. J. Neurol. Sci. (2015) 357 22-27. DOI: 10.1016/j.jns.2015.06.053 21. Ferrer-Donato A., Contreras A., Frago L.M., Chowen J.A., Fernandez-Martos C.M.. Alterations in Leptin Signaling in Amyotrophic Lateral Sclerosis (ALS). Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms221910305 22. Ferrer-Donato A., Contreras A., Fernandez P., Fernandez-Martos C.M.. The potential benefit of leptin therapy against amyotrophic lateral sclerosis (ALS). Brain Behav. (2022) 12 e2465. DOI: 10.1002/brb3.2465 23. Béland L.-C., Markovinovic A., Jakovac H., De Marchi F., Bilic E., Mazzini L., Kriz J., Munitic I.. Immunity in amyotrophic lateral sclerosis: Blurred lines between excessive inflammation and inefficient immune responses. Brain Commun. (2020) 2 fcaa124. DOI: 10.1093/braincomms/fcaa124 24. Beers D.R., Appel S.H.. Immune dysregulation in amyotrophic lateral sclerosis: Mechanisms and emerging therapies. Lancet Neurol. (2019) 18 211-220. DOI: 10.1016/S1474-4422(18)30394-6 25. Abe K., Aoki M., Tsuji S., Itoyama Y., Sobue G., Togo M., Hamada C., Tanaka M., Akimoto M., Nakamura K.. Safety and efficacy of edaravone in well defined patients with amyotrophic lateral sclerosis: A randomised, double-blind, placebo-controlled trial. Lancet Neurol. (2017) 16 505-512. DOI: 10.1016/S1474-4422(17)30115-1 26. Gurney M.E., Pu H., Chiu A.Y., Dal Canto M.C., Polchow C.Y., Alexander D.D., Caliendo J., Hentati A., Kwon Y.W., Deng H.X.. Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science (1994) 264 1772-1775. DOI: 10.1126/science.8209258 27. Picher-Martel V., Renaud L., Bareil C., Julien J.-P.. Neuronal Expression of UBQLN2. Mol. Neurobiol. (2019) 56 4680-4696. DOI: 10.1007/s12035-018-1411-3 28. Inoki K., Kim J., Guan K.-L.. AMPK and mTOR in Cellular Energy Homeostasis and Drug Targets. Annu. Rev. Pharmacol. Toxicol. (2012) 52 381-400. DOI: 10.1146/annurev-pharmtox-010611-134537 29. Lee M.J., Wang Y., Ricci M.R., Sullivan S., Russell C.D., Fried S.K.. Acute and chronic regulation of leptin synthesis, storage, and secretion by insulin and dexamethasone in human adipose tissue. Am. J. Physiol. Endocrinol. Metab. (2007) 292 E858-E864. DOI: 10.1152/ajpendo.00439.2006 30. Marques-Oliveira G.H., Silva T.M., Lima W.G., Valadares H.M.S., Chaves V.E.. Insulin as a hormone regulator of the synthesis and release of leptin by white adipose tissue. Peptides (2018) 106 49-58. DOI: 10.1016/j.peptides.2018.06.007 31. Mitchell R.M., Freeman W.M., Randazzo W.T., Stephens H.E., Beard J.L., Simmons Z., Connor J.R.. A CSF biomarker panel for identification of patients with amyotrophic lateral sclerosis. Neurology (2009) 72 14-19. PMID: 18987350 32. Guo J., Yang X., Gao L., Zang D.. Evaluating the levels of CSF and serum factors in ALS. Brain Behav. (2017) 7 e00637. DOI: 10.1002/brb3.637 33. Kuhle J., Lindberg R.L.P., Regeniter A., Mehling M., Steck A.J., Kappos L., Czaplinski A.. Increased levels of inflammatory chemokines in amyotrophic lateral sclerosis. Eur. J. Neurol. (2009) 16 771-774. PMID: 19236470 34. Italiani P., Carlesi C., Giungato P., Puxeddu I., Borroni B., Bossù P., Migliorini P., Siciliano G., Boraschi D.. Evaluating the levels of interleukin-1 family cytokines in sporadic amyotrophic lateral sclerosis. J. Neuroinflamm. (2014) 11 94. DOI: 10.1186/1742-2094-11-94 35. Liu J., Gao L., Zang D.. Elevated levels of IFN-γ in CSF and serum of patients with amyotrophic lateral sclerosis. PLoS ONE (2015) 10. DOI: 10.1371/journal.pone.0136937 36. Babu G.N., Kumar A., Chandra R., Puri S.K., Kalita J., Misra U.K.. Elevated inflammatory markers in a group of amyotrophic lateral sclerosis patients from northern India. Neurochem. Res. (2008) 33 1145-1149. DOI: 10.1007/s11064-007-9564-x 37. Tortelli R., Zecca C., Piccininni M., Benmahamed S., Dell’Abate M.T., Barulli M.R., Capozzo R., Battista P., Logroscino G.. Plasma Inflammatory Cytokines Are Elevated in ALS. Front. Neurol. (2020) 11 552295. DOI: 10.3389/fneur.2020.552295 38. Lu C.-H., Allen K., Oei F., Leoni E., Kuhle J., Tree T., Fratta P., Sharma N., Sidle K., Howard R.. Systemic inflammatory response and neuromuscular involvement in amyotrophic lateral sclerosis. Neurol. Neuroimmunol. Neuroinflamm. (2016) 3 e244. DOI: 10.1212/NXI.0000000000000244 39. Polverino A., Rucco R., Stillitano I., Bonavita S., Grimaldi M., Minino R., Pesoli M., Trojsi F., D’Ursi A.M., Sorrentino G.. In Amyotrophic Lateral Sclerosis Blood Cytokines Are Altered, but Do Not Correlate with Changes in Brain Topology. Brain Connect. (2020) 10 411-421. PMID: 32731760 40. Jin M., Günther R., Akgün K., Hermann A., Ziemssen T.. Peripheral proinflammatory Th1/Th17 immune cell shift is linked to disease severity in amyotrophic lateral sclerosis. Sci. Rep. (2020) 10 5941. PMID: 32246039 41. Pronto-Laborinho A., Pinto S., Gromicho M., Pereira M., Swash M., de Carvalho M.. Interleukin-6 and amyotrophic lateral sclerosis. J. Neurol. Sci. (2019) 398 50-53. DOI: 10.1016/j.jns.2019.01.026 42. Prado L.d.G.R., Rocha N.P., de Souza L.C., Bicalho I.C.S., Gomez R.S., Vidigal-Lopes M., Braz N.F.T., Vieira É.L.M., Teixeira A.L.. Longitudinal assessment of clinical and inflammatory markers in patients with amyotrophic lateral sclerosis. J. Neurol. Sci. (2018) 394 69-74. DOI: 10.1016/j.jns.2018.08.033 43. Andrés-Benito P., Moreno J., Domínguez R., Aso E., Povedano M., Ferrer I.. Inflammatory Gene Expression in Whole Peripheral Blood at Early Stages of Sporadic Amyotrophic Lateral Sclerosis. Front. Neurol. (2017) 8 546. DOI: 10.3389/fneur.2017.00546 44. Staats K.A., Borchelt D.R., Tansey M.G., Wymer J.. Blood-based biomarkers of inflammation in amyotrophic lateral sclerosis. Mol. Neurodegener. (2022) 17 11. DOI: 10.1186/s13024-022-00515-1 45. Zhao W., Beers D.R., Hooten K.G., Sieglaff D.H., Zhang A., Kalyana-Sundaram S., Traini C.M., Halsey W.S., Hughes A.M., Sathe G.M.. Characterization of Gene Expression Phenotype in Amyotrophic Lateral Sclerosis Monocytes. JAMA Neurol. (2017) 74 677-685. DOI: 10.1001/jamaneurol.2017.0357 46. Saad M.F., Damani S., Gingerich R.L., Riad-Gabriel M.G., Khan A., Boyadjian R., Jinagouda S.D., El-Tawil K., Rude R.K., Kamdar V.. Sexual dimorphism in plasma leptin concentration. J. Clin. Endocrinol. Metab. (1997) 82 579-584. DOI: 10.1210/jc.82.2.579 47. Vitery R.J., Ortega G., Salazar C.B.C.. Leptin sexual dimorphism, insulin resistance, and body composition in normal weight prepubescent. Rev. Chil. Pediatr. (2020) 91 924-929. PMID: 33861829 48. Martin L.J., Mahaney M.C., Almasy L., MacCluer J.W., Blangero J., Jaquish C.E., Comuzzie A.G.. Leptin’s sexual dimorphism results from genotype by sex interactions mediated by testosterone. Obes. Res. (2002) 10 14-21. DOI: 10.1038/oby.2002.3 49. Zhang F., Wang S., Signore A.P., Chen J.. Neuroprotective effects of leptin against ischemic injury induced by oxygen-glucose deprivation and transient cerebral ischemia. Stroke (2007) 38 2329-2336. DOI: 10.1161/STROKEAHA.107.482786 50. Fewlass D.C., Noboa K., Pi-Sunyer F.X., Johnston J.M., Yan S.D., Tezapsidis N.. Obesity-related leptin regulates Alzheimer’s Abeta. FASEB J. (2004) 18 1870-1878. DOI: 10.1096/fj.04-2572com 51. Greco S.J., Sarkar S., Casadesus G., Zhu X., Smith M.A., Ashford J.W., Johnston J.M., Tezapsidis N.. Leptin inhibits glycogen synthase kinase-3beta to prevent tau phosphorylation in neuronal cells. Neurosci. Lett. (2009) 455 191-194. DOI: 10.1016/j.neulet.2009.03.066 52. Greco S.J., Sarkar S., Johnston J.M., Tezapsidis N.. Leptin regulates tau phosphorylation and amyloid through AMPK in neuronal cells. Biochem. Biophys. Res. Commun. (2009) 380 98-104. DOI: 10.1016/j.bbrc.2009.01.041 53. Weng Z., Signore A.P., Gao Y., Wang S., Zhang F., Hastings T., Yin X.-M., Chen J.. Leptin protects against 6-hydroxydopamine-induced dopaminergic cell death via mitogen-activated protein kinase signaling. J. Biol. Chem. (2007) 282 34479-34491. DOI: 10.1074/jbc.M705426200 54. Bouret S.G.. Neurodevelopmental actions of leptin. Brain Res. (2010) 1350 2-9. DOI: 10.1016/j.brainres.2010.04.011 55. Fujita Y., Yamashita T.. The Effects of Leptin on Glial Cells in Neurological Diseases. Front. Neurosci. (2019) 13 828. DOI: 10.3389/fnins.2019.00828 56. Ahmed R.M., Phan K., Highton-Williamson E., Strikwerda-Brown C., Caga J., Ramsey E., Zoing M., Devenney E., Kim W.S., Hodges J.R.. Eating peptides: Biomarkers of neurodegeneration in amyotrophic lateral sclerosis and frontotemporal dementia. Ann. Clin. Transl. Neurol. (2019) 6 486-495. DOI: 10.1002/acn3.721 57. Cereda C., Baiocchi C., Bongioanni P., Cova E., Guareschi S., Metelli M.R., Rossi B., Sbalsi I., Cuccia M.C., Ceroni M.. TNF and sTNFR1/2 plasma levels in ALS patients. J. Neuroimmunol. (2008) 194 123-131. DOI: 10.1016/j.jneuroim.2007.10.028 58. Ronin E., Pouchy C., Khosravi M., Hilaire M., Grégoire S., Casrouge A., Kassem S., Sleurs D., Martin G.H., Chanson N.. Tissue-restricted control of established central nervous system autoimmunity by TNF receptor 2-expressing Treg cells. Proc. Natl. Acad. Sci. USA (2021) 118 e2014043118. PMID: 33766913 59. Probert L.. TNF and its receptors in the CNS: The essential, the desirable and the deleterious effects. Neuroscience (2015) 302 2-22. DOI: 10.1016/j.neuroscience.2015.06.038 60. Magliozzi R., Pezzini F., Pucci M., Rossi S., Facchiano F., Marastoni D., Montagnana M., Lippi G., Reynolds R., Calabrese M.. Changes in Cerebrospinal Fluid Balance of TNF and TNF Receptors in Naïve Multiple Sclerosis Patients: Early Involvement in Compartmentalised Intrathecal Inflammation. Cells (2021) 10. DOI: 10.3390/cells10071712 61. Mendelson M.M., Johannes R., Liu C., Huan T., Yao C., Miao X., Murabito J.M., Dupuis J., Levy D., Benjamin E.J.. Epigenome-Wide Association Study of Soluble Tumor Necrosis Factor Receptor 2 Levels in the Framingham Heart Study. Front. Pharmacol. (2018) 9 207. DOI: 10.3389/fphar.2018.00207 62. Shai I., Schulze M.B., Manson J.E., Rexrode K.M., Stampfer M.J., Mantzoros C., Hu F.B.. A prospective study of soluble tumor necrosis factor-alpha receptor II (sTNF-RII) and risk of coronary heart disease among women with type 2 diabetes. Diabetes Care (2005) 28 1376-1382. DOI: 10.2337/diacare.28.6.1376 63. Medler J., Wajant H.. Tumor necrosis factor receptor-2 (TNFR2): An overview of an emerging drug target. Expert Opin. Ther. Targets (2019) 23 295-307. DOI: 10.1080/14728222.2019.1586886 64. Lim M.A., Selak M.A., Xiang Z., Krainc D., Neve R.L., Kraemer B.C., Watts J.L., Kalb R.G.. Reduced activity of AMP-activated protein kinase protects against genetic models of motor neuron disease. J. Neurosci. (2012) 32 1123-1141. DOI: 10.1523/JNEUROSCI.6554-10.2012 65. Perera N.D., Turner B.J.. AMPK Signalling and Defective Energy Metabolism in Amyotrophic Lateral Sclerosis. Neurochem. Res. (2016) 41 544-553. DOI: 10.1007/s11064-015-1665-3 66. Liu Y.J., Lee L.M., Lai H.L., Chern Y.. Aberrant activation of AMP-activated protein kinase contributes to the abnormal distribution of HuR in amyotrophic lateral sclerosis. FEBS Lett. (2015) 589 432-439. DOI: 10.1016/j.febslet.2014.12.029 67. Perera N.D., Sheean R.K., Scott J.W., Kemp B.E., Horne M.K., Turner B.J.. Mutant TDP-43 deregulates AMPK activation by PP2A in ALS models. PLoS ONE (2014) 9. DOI: 10.1371/journal.pone.0090449 68. Kaneb H.M., Sharp P.S., Rahmani-Kondori N., Wells D.J.. Metformin treatment has no beneficial effect in a dose-response survival study in the SOD1 G93A mouse model of ALS and is harmful in female mice. PLoS ONE (2011) 6. DOI: 10.1371/journal.pone.0024189 69. Yun Y.C., Jeong S.G., Kim S.H., Cho G.W.. Reduced sirtuin 1/adenosine monophosphate-activated protein kinase in amyotrophic lateral sclerosis patient-derived mesenchymal stem cells can be restored by resveratrol. J. Tissue Eng. Regen. Med. (2019) 13 110-115. DOI: 10.1002/term.2776 70. Ruiz-Perera L.M., Höving A.L., Schmidt K.E., Cenan S., Wohllebe M., Greiner J.F.W., Kaltschmidt C., Simon M., Knabbe C., Kaltschmidt B.. Neuroprotection Mediated by Human Blood Plasma in Mouse Hippocampal Slice Cultures and in Oxidatively Stressed Human Neurons. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22179567 71. Wang P., Li X., Li X., Yang W., Yu S.. Blood Plasma of Patients with Parkinson’s Disease Increases Alpha-Synuclein Aggregation and Neurotoxicity. Park. Dis. (2016) 2016 7596482. DOI: 10.1155/2016/7596482 72. Tsubai T., Noda Y., Ito K., Nakao M., Seino Y., Oiso Y., Hamada Y.. Insulin elevates leptin secretion and mRNA levels via cyclic AMP in 3T3-L1 adipocytes deprived of glucose. Heliyon (2016) 2 e00194. DOI: 10.1016/j.heliyon.2016.e00194
--- title: Identification of the Constituents of Ethyl Acetate Fraction from Smilax china L. and Determination of Xanthine Oxidase Inhibitory Properties authors: - Xin Li - Shanshan Liu - Weili Jin - Wenkai Zhang - Guodong Zheng journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049564 doi: 10.3390/ijms24065158 license: CC BY 4.0 --- # Identification of the Constituents of Ethyl Acetate Fraction from Smilax china L. and Determination of Xanthine Oxidase Inhibitory Properties ## Abstract The aim of this work was to investigate the xanthine oxidase (XO)-inhibitory activity of ethanol extracts from Smilax china L. and to identify the active compounds in the ethyl acetate (EtOAc) fraction. Extraction of ethanol extracts from Smilax china L. and then ethanol extracts were concentrated, and the polyphenolic compounds were extracted with petroleum ether (PE), chloroform, EtOAc, n-butanol (n-BuOH), and residual ethanol fractions. Their effects on XO activity were then compared separately. The polyphenolic components of the EtOAc fraction were identified by HPLC and HPLC–mass spectrometry (HPLC-MS) analysis. Kinetic analysis demonstrated that all these extracts showed XO-inhibitory properties, and among them the EtOAc fraction had the strongest inhibitory effect (IC50 = 101.04 μg/mL). The inhibitory constant (Ki) of the EtOAc fraction on XO activity was 65.20 μg/mL, showing excellent inhibition on XO in the competitive mode. Sixteen compounds were identified from the EtOAc fraction. The study demonstrates that the EtOAc fraction of Smilax china L. may be a potential functional food to inhibit XO activity. ## 1. Introduction The pathological feature of gout is the precipitation of monosodium urate crystals in tendons, kidneys, joints, and surrounding tissues, resulting in painful acute gouty arthritis [1]. Hyperuricemia is usually related to the overproduction of uric acid, leading to an enhanced serum level of uric acid, which is due to elevated xanthine oxidase (XO) activity. XO is a molybdenum (Mo)-containing enzyme that catalyzes the oxidation of purine bases to uric acid and produces reactive oxygen and nitrogen species [2,3,4]. Excessive amounts of reactive oxygen species in the body can increase oxidative stress, causing a variety of pathological processes, such as inflammation, atherosclerosis, and cancer [5,6]. Therefore, XO inhibitors are regarded as the main drugs for treating these diseases. Although the therapeutic agent (i.e., allopurinol) is the primary inhibitor in the treatment of gout, it exhibits side effects including hepatotoxicity, Stevens Jones syndrome, and nephropathy [7,8,9]. It has been demonstrated that numerous food and medicinal herbs that suppress XO can be used to prevent or treat gout [10]. Furthermore, phenolics have been shown to be strong XO inhibitors [11]. The therapeutic effects of numerous plants have been ascribed to polyphenol because of its antioxidant and enzyme-inhibitory properties [12,13]. For example, apigenin is a strong XO inhibitor, with an inhibition constant (Ki) value of 0.61 ± 0.31 μM [14]. Myricetin was reported to inhibit uric acid formation in a mixed-type manner with an IC50 of (8.66 ± 0.03) ×10 −6 mol/L [15]. Dong et al. found that pinobanksin and galangin inhibited XO activity with IC50 values of 1.37 × 10−4 and 1.63 × 10−4 mol/L, respectively [16]. Moreover, kaempferol has been found to reversibly inhibit the activity of XO in a competitive manner with a Ki of 6.77 ± 1.02 μM [17]. Smilax china L., also called “Jingangteng (JGT)” and “Baqia”, is primarily derived from Liliaceae plants. It is cultivated throughout East China, Central and South China, and Southwest China. This plant is listed in the Chinese pharmacopoeia and generally used to treat pelvic inflammatory disease, diuresis, rheumatoid arthritis, detoxication, tumors, and other diseases; it is commonly used in traditional Chinese medicine, functional soup, and dessert food in China [18,19]. Extensive previous studies on Smilax china L. have revealed that it has anti-inflammatory [20], anti-obesity [21,22], anti-bacterial [23], and other activities. Phytochemical investigations have shown that steroid saponins and polyphenols are the major chemical constituents of Smilax china L. which show beneficial pharmacological effects [24,25]. This study aimed to explore the XO-inhibitory effect of ethyl acetate (EtOAc) of Smilax china L. and its chemical composition. In this study, the ethanol extract was partitioned successively by petroleum ether (PE), chloroform, EtOAc, and n-butanol (n-BuOH) according to solvent polarity. These extracts’ XO-inhibitory effects were investigated. Among them, the EtOAc extract exhibited the strongest XO-inhibitory activity, and its components were further identified by HPLC-MS. ## 2.1. The Contents of Total Phenolic and Total Flavonoid All extracts/fractions of Smilax china L. were measured for their total contents of phenolics and flavonoids, as shown in Figure 1. Among these extracts, the ethanol extract (337.612 ± 2.13 mg/g) contained highest number of phenolic compounds, followed by the EtOAc (331.47 ± 6.835 mg/g), n-BuOH (295.832 ± 1.11 mg/g), PE (249.301 ± 2.738 mg/g), chloroform (205.282 ± 4.356 mg/g), and residual ethanol (180.113 ± 1.846) fractions. In addition, the total flavonoid content was found to be highest in the EtOAc fraction (425.90 ± 7.315 mg/g). These data suggested that different solvent partitions successfully distinguished different-polarity polyphenols and flavonoids. ## 2.2. Effect of Smilax china L. Extracts on XO Activity As presented in Figure 2A, the XO-inhibitory effects of the ethanol extract (EtOH) and its extracted fractions, including PE, chloroform, EtOAc, n-BuOH, and residual ethanol solution, were compared. The XO-inhibitory effect of the extracts was reduced in the following order: EtOAc fraction > ethanol extract > PE fraction > n-BuOH fraction > residual ethanol > chloroform fraction, indicating that the extracts were potential XO inhibitors. Allopurinol is used as a XO-inhibitory drug, and its inhibitory activity against XO is stable around $90\%$ at a concentration of 20–100 μg/mL. The EtOAc fraction showed the best XO-inhibitory effect, and further linear fitting was carried out. The equation between inhibition and concentration is $Y = 54.00565$ − 3.52195X + 0.18884X2 − 0.00411X3 + (4.3291 × 10−5)X4 – (2.18682 × 10−7)X5 + (4.23862 × 10−10)X6 (R2 = 0.9914), where $Y = 50$ and $X = 101.04004.$ Thus, the IC50 value of the EtOAc fraction is 101.04004 μg/mL. Among them, the EtOAc extract had the best inhibitory activity on XO, demonstrating that the XO-inhibitory activity was efficiently enriched in the EtOAc fraction Additionally, the inhibitory effects of the EtOAc fraction and allopurinol on XO were compared (Figure 2B). The data indicated that a new, potent inhibitor could be explored in the EtOAc fraction. ## 2.3. The Type of Inhibition by Lineweaver–Burk Plot Analysis To measure whether the inhibitory activity of the EtOAc fraction on XO is reversible, plots of ν versus [XO] at various concentrations were constructed in Figure 3A. As the concentration of the EtOAc fraction increased, all the lines passed through the origin, and the slopes decreased. The results indicated that the inhibitory effect of the EtOAc fraction on XO was reversible and that there was non-covalent intermolecular interaction with XO. Furthermore, the kinetic mechanism of the EtOAc fraction was investigated using a double-reciprocal Lineweaver–Burk plot according to Equation [1]. As illustrated in Figure 3B and Table 1, the horizontal axis intercept (−1/Km) gradually increased with increasing EtOAc fraction concentration, while the vertical axis intercept (1/Vmax) remained constant, indicating that the EtOAc fraction was a competitive XO inhibitor. This was possible because the EtOAc fraction was connected directly to the active position of XO, blocking entry to the enzyme active site and resulting in an obvious reduction in the XO-inhibitory activity. Furthermore, the secretion of uric acid is decreased, but not sufficiently to render XO inactive. Thus, dialysis, ultrafiltration, and other methods are still used to restore enzyme activity [26]. According to Equation [2], the Ki value was 65.20 μg/mL, and the concentration reached 160 μg/mL. The lower the Ki value, the tighter the binding to XO, and consequently the greater the inhibitory effect on XO [27]. ## 2.4. Identification of Active Compounds Smilax china L. extracts were effective against XO, and the EtOAc fraction was the most effective, which might be related to the active components, such as polyphenols. Thus, it is very necessary to identify the components of the EtOAc fraction. Therefore, the composition of the EtOAc fraction was determined by applying the HPLC chromatography method (Figure 4). Quercetin, quercetin-3-O-rhamnoside, engeletin, rutin, chlorogenic acid, gallic acid, kaempferol, kaempferitrin, isoquercetin, and astragalin were used as standards. As shown in Figure 5 and Table 2, the polyphenols were identified based on the retention time and the chromatographic peaks of standards. The main components were identified as isoquercetin ($0.950\%$), quercetin-3-O-rhamnoside ($6.574\%$), and engeletin ($2.461\%$). In vivo studies have also evidenced that polyphenol compounds exhibit a significant effect on XO activity [13]. Huang et al. [ 28] discovered that puerarin, myricetin, morin, apigenin, kaempferol, and quercetin could significantly lower the serum uric acid level in hyperuricemic rats, and some of them also inhibited liver XO activity. Therefore, the results highlighted the potential of polyphenols to treat gout, illustrating that the EtOAc fraction, which was rich in XO inhibitors, is an active ingredient that deserves further investigation. ## 2.5. Chemical Constituent Analysis by HPLC-MS Compounds were tentatively identified according to precursor ions, fragment ions, the corresponding reference compound, and the literature and databases at http://www.massbank.jp/ (accessed on 8 May 2019), http://www.chemspider.com/ (accessed on 11 May 2019), and https://scifinder.cas.org/ (accessed on 13 May 2019). The retention times (Rt), observed and calculated mass er. rors (ppm), molecular formulas, MS/MS product ions, and proposed compounds are summarized in Table 3. The total ion chromatogram (TIC) is depicted in Figure 6. Thirteen compounds were identified and tentatively characterized. By comparing the retention time, MS, and MS/MS fragmentation pattern with the related literature, peak 1 was detected; the precursor ion [M−H]− was located at m/z 353.085 (C16H18O9), the major fragment ion was located at m/z 191.055, and referring to the literature, it was identified as chlorogenic acid (quinic acid) [27]. Research has shown that chlorogenic acid protects against tissue damage caused by ischemia/reperfusion, enhances antioxidant activity, and scavenges superoxide free radicals. In addition, chlorogenic acid interacts with enzymes, altering their structure and biological activity [29]. The production of reactive oxygen species (ROS) occurs during the process of XO catalyzing purine to generate uric acid. When XO activity is elevated, it will result in a large amount of ROS generation, causing oxidative damage in the body, which is related to the onset of ischemia–reperfusion injury [23]. Reactive oxygen species (ROS) are the main cause of organ damage after the blood supply to the ischemic tissue is restored. Therefore, the results indicated that the inhibitory effect of chlorogenic acid on XO needs further study. Peaks 2 and 5 revealed [M−H]− ions located at m/z 329.086 and 329.086, respectively. The MS2 spectrum showed that the characteristic vanillic acid molecular ions were at m/z 167.035 and 167.033. According to literature data from [27], peaks 2 and 5 were vanillic acid hexoside. Peak 3 was (−)-epicatechin or catechin-deprotonated pseudo-molecular ion; MS was m/z 289, and MS2 was m/z 121.026 [30,31]. Peaks 4 and 6 with an identical molecular formula (C46H28O10) were tentatively proposed as catechin-O-digalloyl-C-rhamnoside and its isomers based on the presence of a fragment ion at m/z 740.168, resulting in an [M−H]− ion at m/z 739.161. Thus, the peaks were characterized as catechin-O-digalloyl-C-rhamnoside (peak 4), and peak 6 should be identified as an isomer of catechin-O-digalloyl-C-rhamnoside by comparison with the literature [31]. Peaks 7 and 8 in the spectrum had the same molecular ion [M−H]−, which had an m/z value of 353. A major fragment at 450.116 and 450.110, respectively, was also observed for both peaks. Thus, the peaks were characterized as eriodictyol-C-hexose (peak 7) and eriodictyol-C-hexose isomers (peak 8), which are isomers. According to the literature, it was tentatively identified as eriodictyol-C-hexose [32]. Peaks 9 and 11 gave the same precursor ion at [M−H]− m/z 451.099 and MS2 fragments at m/z 341.063 with the molecular formula C24H20O9. These structural isomers exhibited different retention times to enable the identification of cinchonain I (peak 9, Rt = 6.26 min) and cinchonain I isomers (peak 11, Rt = 8.37 min) [33]. Another compound (peak 12) had a molecular ion [M−H]− at m/z 613.128 and was therefore tentatively identified as cinchonain I hexose, but it presented a different fragment from peaks 9 and 11. Peak 10 with [M−H]− in the negative mode at m/z 243.064 and an MS2 fragment ion at 175.073 was identified as uridine. Peak 13 with [M−H]− at m/z 227.068 and an MS2 ion at 143.048 was tentatively identified as [2]-2-(aspartylamino)-4-pentynoic acid by comparison with the reference spectral fragments. ## 3. Discussion Gout, also referred to “the disease of kings” or “rich man’s disease”, is a classical disease that is now recognized as being lifestyle-related. The incidence of gout and related hyperuricemia is increasing yearly [34]. Gout is caused by an increase in serum uric acid levels, which is the final metabolite of purine catabolism in humans. Inhibiting XO is one of the effective methods to lower the level of serum uric acid [2,3]. It has been shown that naturally occurring phytochemicals such as polyphenols possess XO-inhibitory properties, are effective XO inhibitors that lower the production of uric acid, and are considered potential therapeutic agents to prevent and treat gout. Smilax china L. polyphenols exhibit anti-obesity, anti-inflammatory, and anti-bacterial effects [20,21,22,23]. In this study, Smilax china L. had a good inhibitory effect on XO, with the ethyl acetate-soluble parts in particular having the best effect, which might be related to polyphenols and other active components. It has been confirmed that the ethyl acetate-soluble parts have the highest contents of polyphenols (Figure 1). In our current study, the three compounds identified by HPLC were flavonoids and glycosides. Quercetin-3-O-rhamnoside, known as quercitrin, is a well-known flavonoid compound that exerts anti-inflammatory and antioxidant activities [35]. Engeletin is a flavonoid glucoside substance. Studies have shown that engeletin has anti-inflammatory effects and improves the inflammatory symptoms related to endometritis by inhibiting the activation of NF-κB [36]. Wei et al. showed that engeletin improves diabetes complications and inhibits LPS-induced inflammation [37]. Likewise, isoquercetin (IC50 value of 0.185 mmol/L) was reported to demonstrate great α-glucosidase-inhibitory effects, inhibiting α-glucosidase activity in a non-competitive manner [38]. Recently, an in vitro and in silico study reported that isoquercetin from red onion (Allium cepa L.) solid waste exhibited potent xanthine oxidase enzyme-inhibitory activity [39]. Thirteen compounds in the EtOAc fraction were detected by HPLC-MS. Liquid phase and mass spectrometry is one of the most widely used methods for composition identification, particularly for identifying the active components of natural products. Researchers have reported that polyphenol has been shown to be more effective in inhibiting XO than allopurinol, with an IC50 value of 0.274 μM for XO and one of 4.784 μM for allopurinol [40]. Chlorogenic acid is an effective XO inhibitor; it binds to sites other than the active site through mixed-type inhibition [41]. According to the results reported by Zhou et al. [ 42], chlorogenic acid ameliorates hyperuricemia in vivo. In this study, the inhibition of the EtOAc fraction on XO was reversible and competitive, indicating that the EtOAc fraction reduced the enzyme activity by competing with the active center of XO, thus reducing the production of uric acid. Nevertheless, in this work, we did not use monomers to examine their efficacy. As a result, we are unable to distinguish the actual specific effect of each single compound. It is still necessary to conduct relevant experiments in vitro and in vivo to determine whether the Smilax china L. polyphenols are suitable candidates to treat gout. ## 4.1. Materials Xanthine oxidase, xanthine, and standard substances (≥$98\%$) including chlorogenic acid, gallic acid, quercetin-3-O-rhamnoside, kaempferol, quercetin, kaempferitrin, engeletin, iso-quercetin, astragalin, and rutin were bought from Beijing Solarbio Technology Co., Ltd. (Beijing, China). Allopurinol was supplied by Shanghai Xinyi Wanxiang Pharmaceutical Co., Ltd. (Shanghai, China). A laboratory mill was purchased from Zhejiang Rhodiola Industry and Trade Co., Ltd (DE-1000gA, Jinhua, China). The qualitative analysis of compounds was completed by HPLC analysis on an Agilent 1260 HPLC system (Agilent Technologies, Santa Clara, CA, USA) equipped with a C18 reverse-phase column. ## 4.2. Preparation of the Extracts Smilax china L. root was brought from a Simcere drugstore in Nanjing, China. The dried root of Smilax china L. was ground into a fine powder with a laboratory mill and ultrasonically extracted with $95\%$ ethanol (1:20) for 40 min. After concentration, the crude ethanol extract was obtained with a yield of $9.2\%$. The ethanol extract was further fractionated according to solvent polarity. The ethanol extract was then resuspended in petroleum ether (PE), chloroform, EtOAc, and n-butanol (n-BuOH) in order of increasing polarity at room temperature with occasional shaking using a separating funnel. After that, these fractions were centrifuged, filtered, and concentrated to obtain PE, chloroform, EtOAc, n-BuOH, and residual ethanol solution fractions. The extraction yields of the PE, chloroform, EtOAc, n-BuOH, and residual ethanol fractions were 0.64 ± $0.04\%$, 1.18 ± $0.07\%$, 5.02 ± $0.23\%$, 2.82 ± $0.15\%$, and 0.54 ± $0.04\%$ (w/w, on a dry-weight basis), respectively. Finally, these extracts were concentrated, freeze-dried, and then stored in a desiccator until use. The total phenolic content was measured using the Folin–Ciocalteu method, and the total flavonoid content was determined by a colorimetric method [43]. ## 4.3. XO-Inhibitory Activity The XO inhibition assay was carried out by the methods used by Zhou et al. [ 44] with minor modifications. In brief, in a 200 μL enzyme reaction system, 60 μL of sodium phosphate buffer (pH = 7.5), 30 μL of XO solution (0.1 U/mL), and various concentrations of ethanol extracts (50 μL) were added to a 96-well plate, and the mixture was incubated for 15 min at 25 °C. Then, the reaction was started by adding 60 μL of xanthine solution (150 μM) to the mixture as a substrate. After incubating, the absorbance was measured at 295 nm. Allopurinol was used as a positive control. The inhibition ratio (%) was calculated as follows: Inhibition ratio (%) = (1 − A/B) × $100\%$ where A is the reaction rate of the system containing the inhibitor; B is the reaction rate of the system without the inhibitor. The extent of inhibition was expressed as IC50 (half of the maximum effective concentration). ## 4.4. Lineweaver–Burk Plots In the reaction system between the inhibitor and enzyme, the concentration of fixed substrate was 150 μM, and the reversibility of enzyme inhibition was measured by varying concentrations of XO and different amounts of inhibitor. The curves of the reaction rate (v) vs. [XO] were depicted to assess the reversibility. ## 4.5. Inhibition Type of the EtOAc Fraction on XO by Kinetic Analysis Different concentrations of EtOAc fraction solutions were prepared (0, 80, and 160 μg/mL), the reaction conditions remained constant, and the concentration of XO was fixed at 0.1 U/mL. The relationship between the reaction rate and substrate concentration was determined under different inhibitor concentrations. The inhibition type was analyzed according to the Lineweaver–Burk plot. For competitive inhibition, the Lineweaver–Burk equation can be written as [29]:[1]1v=KmVmax1+IKi1S+1Vmax [2]Kmapp=Km[I]Ki+Km where Ki and Km denote the inhibition constant and Michaelis–Menten constant, respectively, and v is the enzyme reaction rate. [ I] and [S] are the concentrations of the inhibitor and substrate, respectively. Kmapp represents the apparent Michaelis–Menten constant. ## 4.6. HPLC–MS Analysis The EtOAc extracts were subjected to HPLC apparatus separation on an Agilent 1260 HPLC system (Agilent Technologies, Santa Clara, CA, USA) equipped with a series of SHODEX KS-804 and KS-802 columns (8 mm × 300 mm) and a refractive index detector. The mobile phase was acetonitrile (solvent A) and $0.1\%$ formic acid aqueous solution (solvent B). The linear gradient solvent system was as follows: 0–10 min, 15–$25\%$ A; 10–22 min, 25–$31\%$ A; 22–30 min, 31–$45\%$ A; 30–31 min, 45–$100\%$ A; and, finally, isocratic elution with $100\%$ phase A until 35 min. The column temperature was 20 °C, the injection volume was 10 μL, the detection wavelength was 280 nm, and the flow rate was maintained at 0.8 mL/min. The MS analysis was performed using an Agilent Technologies 6538 OHD accurate-mass quadrupole time-of-flight (Q-TOF) mass spectrometer to analyze and identify the EtOAc fraction’s active compounds. The Q-TOF mass spectrometer was equipped with an electrospray ionization source (ESI) in the negative mode. The operating parameters were set as follows: spray voltage of 4000 KV, gas collision at He, nebulizer pressure at 50 psi, collision voltage of 250 V, and drying temperature of 350 °C. The ion scanning range was 100–1000 m/z. Data from Agilent’s MassHunter 8.0 software were analyzed. ## 4.7. Statistical Analysis The experimental results were expressed as mean ± standard deviation (SD) and analyzed by one-way ANOVA using the Origin 8.5 software; $p \leq 0.05$ was considered significant. ## 5. Conclusions In this study, Smilax china L. ethanol extracts exhibited an inhibitory effect on XO, especially the EtOAc fraction. The EtOAc fraction showed excellent XO-inhibitory performance in the competitive mode. Sixteen compounds in the EtOAc fraction were identified by HPLC and HPLC-MS. Thus, the ethanol extract of Smilax china L. has great potential for inhibiting XO activity and preventing hyperuricemia. Future studies should focus on the pharmacological testing of these compounds in vivo. ## References 1. Tang H., Zhao D.. Investigation of the interaction between salvianolic acid C and xanthine oxidase: Insights from experimental studies merging with molecular docking methods. Bioorganic Chem. (2019) 88 102981. DOI: 10.1016/j.bioorg.2019.102981 2. Kumar R., Joshi G., Kler H., Kalra S., Kaur M., Arya R.. Toward an Understanding of Structural Insights of Xanthine and Aldehyde Oxidases: An Overview of their Inhibitors and Role in Various Diseases. Med. Res. Rev. (2018) 38 1073-1125. DOI: 10.1002/med.21457 3. Dong C., Montes M., Al-Sawai W.M.. Xanthine oxidoreductase inhibition—A review of computational aspect. J. Comput. Chem. (2020) 11 2040008. DOI: 10.1142/S0219633620400088 4. Luna G., Dolzhenko A.V., Mancera R.L.. Inhibitors of Xanthine Oxidase: Scaffold Diversity and Structure-Based Drug Design. Chem. Med. Chem. (2019) 14 714-743. DOI: 10.1002/cmdc.201900034 5. Saito H., Tanaka K., Iwasaki T., Oda A., Watanabe S., Kanno M., Kimura H., Shimabukuro M., Asahi K., Watanabe T.. Xanthine oxidase inhibitors are associated with reduced risk of cardiovascular disease. Sci. Rep. (2021) 11 1380. DOI: 10.1038/s41598-020-80835-8 6. Rodriguez S.A., Murray A.P., Leiro J.M.. Xanthine Oxidase Inhibition by Aqueous Extract of Limonium brasiliense (Plumbaginaceae). Chem. Proc. (2021) 3 123 7. Perez-Ruiz F., Dalbeth N., Bardin T.. A review of uric acid, crystal deposition disease, and gout. Adv. Ther. (2015) 32 31-41. DOI: 10.1007/s12325-014-0175-z 8. Liu C.W., Chang W.C., Lee C.C., Shau W.Y., Hsu F.S., Wang M.L., Chen T.C., Lo C., Hwang J.J.. The net clinical benefits of febuxostat versus allopurinol in patients with gout or asymptomatic hyperuricemia-A systematic review and meta-analysis. Nutr. Metab. Cardiovasc. Dis. (2019) 29 1011-1022. DOI: 10.1016/j.numecd.2019.06.016 9. El-Sheikh A.A., Abdelzaher W.Y., Gad A.A., Abdel-Gaber S.A.. Purine versus non-purine xanthine oxidase inhibitors against cyclophosphamide-induced cardiac and bone marrow toxicity in rats. Hum. Exp. Toxicol. (2020) 39 249-261. DOI: 10.1177/0960327119883412 10. Orhan I.E., Deniz F.S.S.. Natural Products and Extracts as Xantine Oxidase Inhibitors—A Hope for Gout Disease?. Curr. Pharm. Des. (2021) 27 143-158. DOI: 10.2174/1381612826666200728144605 11. Mehmood A., Li J., Rehman A.U., Kobun R., Llah I.U., Khan I., Althobaiti F., Albogami S., Usman M., Alharthi F.. Xanthine oxidase inhibitory study of eight structurally diverse phenolic compounds. Front. Nutr. (2022) 9 966557. DOI: 10.3389/fnut.2022.966557 12. Mohos V., Fliszár-Nyúl E., Poór M.. Inhibition of Xanthine Oxidase-Catalyzed Xanthine and 6-Mercaptopurine Oxidation by Flavonoid Aglycones and Some of Their Conjugates. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21093256 13. Chen Y., Zhao Z., Li Y., Yang Y., Li L., Jiang Y., Lin C., Cao Y., Zhou P., Tian Y.. Baicalein alleviates hyp ruricemia by promoting uric acid excretion and inhibiting xanthine oxidase. Phytomedicine (2021) 80 153374. DOI: 10.1016/j.phymed.2020.153374 14. Lin C.M., Chen C.S., Chen C.T., Liang Y.C., Lin J.K.. Molecular modeling of flavonoids that inhibits xanthine oxidase. Biochem. Biophys. Res. Commun. (2002) 294 167-172. DOI: 10.1016/S0006-291X(02)00442-4 15. Zhang C., Zhang G., Liao Y., Gong D.. Myricetin inhibits the generation of superoxide anion by reduced form of xanthine oxidase. Food Chem. (2017) 221 1569-1577. DOI: 10.1016/j.foodchem.2016.10.136 16. Dong Y., Huang H., Zhao M., Sun-Waterhouse D., Lin L., Xiao C.. Mechanisms underlying the xanthine oxidase inhibitory effects of dietary flavonoids galangin and pinobanksin. J. Funct. Foods (2016) 24 26-36. DOI: 10.1016/j.jff.2016.03.021 17. Wang Y.J., Zhang G.W., Pan J.H., Gong D.M.. Novel insights into the inhibitory mechanism of kaempferol on xanthine oxidase. J. Agric. Food Chem. (2015) 63 526-534. DOI: 10.1021/jf505584m 18. Zhang Y., Zhao Z., Chen H., Fu Y., Wang W., Li Q., Wang X., Fan G., Zhang Y.. The Underlying Molecular Mechanisms Involved in Traditional Chinese Medicine. Evid.-Based Complement. Altern. Med. (2021) 2021 5552532. DOI: 10.1155/2021/5552532 19. Wu L.S., Wang X.J., Wang H., Yang H.W., Jia A.Q., Ding Q.. Cytotoxic polyphenols against breast tumor cell in. J. Ethnopharmacol. (2010) 130 460-464. DOI: 10.1016/j.jep.2010.05.032 20. Li X., Chu L., Liu S., Zhang W., Lin L., Zheng G.. Phytomedicine (2022) 95 153728. DOI: 10.1016/j.phymed.2021.153728 21. Li X., Yang L., Xu M., Qiao G., Li C., Zheng G.. J. Funct. Foods (2021) 77 104332. DOI: 10.1016/j.jff.2020.104332 22. Li X., Yang L., Li J., Lin L., Zheng G.. A flavonoid-rich. Food Funct. (2021) 12 5862-5875. DOI: 10.1039/D1FO00282A 23. Xu M., Xue H., Li X., Zhao Y., Lin L., Yang L., Zheng G.. Chemical composition, antibacterial properties, and mechanism of. Appl. Microbiol. Biotechnol. (2019) 103 9013-9022. DOI: 10.1007/s00253-019-10100-0 24. Huang Z.H., Hao Q., Rong-Tao L.I., Hai-Zhou L.I.. Chemical constituents from rhizomes of smilax China linn. J. Kunming Univ. Sci. Technol. (2014) 1 16 25. Yang C., Tang Q., Liu J., Zhang Z., Liu W.. Preparative isolation and purification of phenolic acids from smilax china by high-speed counter-current chromatography. Sep. Purif. Technol. (2008) 61 474-478. DOI: 10.1016/j.seppur.2008.02.017 26. Hemmateenejad B., Shamsipur M., Samari F., Khayamian T., Ebrahimi M., Rezaei Z.. Combined fluorescence spectroscopy and molecular modeling studies on the interaction between harmalol and human serum albumin. J. Pharm. Biomed. Anal. (2012) 67–68 201-208. DOI: 10.1016/j.jpba.2012.04.012 27. Lin L., Yang Q., Zhao K., Zhao M.. Identification of the free phenolic profile of Adlay bran by UPLC-QTOF-MS/MS and inhibitory mechanisms of phenolic acids against xanthine oxidase. Food Chem. (2018) 253 108-118. DOI: 10.1016/j.foodchem.2018.01.139 28. Mo S.F., Zhou F., Lv Y.Z., Hu Q.H., Zhang D.M., Kong L.D.. Hypouricemic action of selected flavonoids in mice: Structure-activity relationships. Biol. Pharm. Bull. (2007) 30 1551-1556. DOI: 10.1248/bpb.30.1551 29. Si Y.X., Wang Z.J., Park D., Chung H.Y., Wang S.F., Yan L., Yang J.M., Qian G.Y., Yin S.J., Park Y.D.. Effect of hesperetin on tyrosinase: Inhibition kinetics integrated computational simulation study. Int. J. Biol. Macromol. (2012) 50 257-262. DOI: 10.1016/j.ijbiomac.2011.11.001 30. Abu-Reidah I.M., Arraez-Roman D., Quirantes-Pine R., Fernandez-Arroyo S., Segura-Carretero A., Fernandez-Gutierrez A.. HPLC–ESI-Q-TOF-MS for a comprehensive characterization of bioactive phenolic compounds in cucumber whole fruit extract. Food Res. Int. (2012) 46 108-117. DOI: 10.1016/j.foodres.2011.11.026 31. Gomez-Romero M., Zurek G., Schneider B., Baessmann C., Segura-Carretero A., Fernandez-Gutierrez A.. Automated identification of phenolics in plant-derived foods by using library search approach. Food Chem. (2011) 124 379-386. DOI: 10.1016/j.foodchem.2010.06.032 32. Pacifico S., Di Maro A., Petriccione M., Galasso S., Piccolella S., Di Giuseppe A.M.A., Scortichini M., Monaco P.. Chemical composition, nutritional value and antioxidant properties of autochthonous Prunus avium cultivars from Campania Region. Food Res. Int. (2014) 64 188-199. DOI: 10.1016/j.foodres.2014.06.020 33. Ming D.S., López A., Hillhouse B.J., French C.J., Hudson J.B., Towers G.H.. Bioactive constituents from Iryanthera megistophylla. J. Nat. Prod. (2002) 65 1412-1416. DOI: 10.1021/np020169l 34. Doherty M.. New insights into the epidemiology of gout. Rheumatology (2009) 48 ii2-ii8. DOI: 10.1093/rheumatology/kep086 35. Cui J., Zong W., Zhao N., Yuan R.. Burdock (Arctium lappa L.) leaf flavonoids rich in morin and quercetin 3-O-rhamnoside ameliorate lipopolysaccharide-induced inflammation and oxidative stress in RAW264.7 cells. Food Sci. Nutr. (2022) 10 2718-2726. DOI: 10.1002/fsn3.2875 36. Wu H., Zhao G., Jiang K., Li C., Qiu C., Deng G.. Engeletin Alleviates Lipopolysaccharide-Induced Endometritis in Mice by Inhibiting TLR4-mediated NF-κB Activation. J. Agric. Food Chem. (2016) 64 6171-6178. DOI: 10.1021/acs.jafc.6b02304 37. Ye W., Chen R., Sun W., Huang C., Lin X., Dong Y., Wen C., Wang X.. Determination and pharmacokinetics of engeletin in rat plasma by ultra-high performance liquid chromatography with tandem mass spectrometry. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. (2017) 1060 144-149. DOI: 10.1016/j.jchromb.2017.06.018 38. Li Y.Q., Zhou F.C., Gao F., Bian J.S., Shan F.. Comparative evaluation of quercetin, isoquercetin and rutin as inhibitors of alpha-glucosidase. J. Agric. Food Chem. (2009) 57 11463-11468. DOI: 10.1021/jf903083h 39. Nile A., Gansukh E., Park G.S., Kim D.H., Hariram Nile S.. Novel insights on the multi-functional properties of flavonol glucosides from red onion (Allium cepa L) solid waste—In vitro and in silico approach. Food Chem. (2021) 335 127650. DOI: 10.1016/j.foodchem.2020.127650 40. Tung Y.T., Chang S.T.. Inhibition of xanthine oxidase by Acacia confusa extracts and their phytochemicals. J. Agric. Food Chem. (2010) 58 781-786. DOI: 10.1021/jf901498q 41. Wang S.H., Chen C.S., Huang S.H., Yu S.H., Lai Z.Y., Huang S.T., Lin C.M.. Hydrophilic ester-bearing chlorogenic acid binds to a novel domain to inhibit xanthine oxidase. Planta Med. (2009) 75 1237-1240. DOI: 10.1055/s-0029-1185521 42. Zhou X., Zhang B., Zhao X., Lin Y., Wang J., Wang X., Hu N., Wang S.. Chlorogenic acid supplementation ameliorates hyperuricemia, relieves renal inflammation, and modulates intestinal homeostasis. Food Funct. (2021) 12 5637-5649. DOI: 10.1039/D0FO03199B 43. Bakar M.F.A., Mohamed M., Rahmat A., Fry J.. Phytochemicals and antioxidant activity of different parts of bambangan (Mangifera pajang) and tarap (Artocarpus odoratissimus). Food Chem. (2009) 113 479-483. DOI: 10.1016/j.foodchem.2008.07.081 44. Zhao J., Huang L., Sun C., Zhao D., Tang H.. Studies on the structure-activity relationship and interaction mechanism of flavonoids and xanthine oxidase through enzyme kinetics, spectroscopy methods and molecular simulations. Food Chem. (2020) 323 126807. DOI: 10.1016/j.foodchem.2020.126807
--- title: Predicting Key Genes and Therapeutic Molecular Modelling to Explain the Association between Porphyromonas gingivalis (P. gingivalis) and Alzheimer’s Disease (AD) authors: - Ahmed Hamarsha - Kumarendran Balachandran - Ahmad Tarmidi Sailan - Nurrul Shaqinah Nasruddin journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049565 doi: 10.3390/ijms24065432 license: CC BY 4.0 --- # Predicting Key Genes and Therapeutic Molecular Modelling to Explain the Association between Porphyromonas gingivalis (P. gingivalis) and Alzheimer’s Disease (AD) ## Abstract The association between Porphyromonas gingivalis (P. gingivalis) and Alzheimer’s disease (AD) remains unclear. The major aim of this study was to elucidate the role of genes and molecular targets in P. gingivalis-associated AD. Two Gene Expression Omnibus (GEO) datasets, GSE5281 for AD ($$n = 84$$ Alzheimer’s, $$n = 74$$ control) and GSE9723 ($$n = 4$$ P. gingivalis, $$n = 4$$ control), were downloaded from the GEO database. Differentially expressed genes (DEGs) were obtained, and genes common to both diseases were drawn. Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analysis was performed from the top 100 genes (50 upregulated and 50 downregulated genes). We then proceeded with CMap analysis to screen for possible small drug molecules targeting these genes. Subsequently, we performed molecular dynamics simulations. A total of 10 common genes (CALD1, HES1, ID3, PLK2, PPP2R2D, RASGRF1, SUN1, VPS33B, WTH3DI/RAB6A, and ZFP36L1) were identified with a p-value < 0.05. The PPI network of the top 100 genes showed UCHL1, SST, CHGB, CALY, and INA to be common in the MCC, DMNC, and MNC domains. Out of the 10 common genes identified, only 1 was mapped in CMap. We found three candidate small drug molecules to be a fit for PLK2, namely PubChem ID: 24971422, 11364421, and 49792852. We then performed molecular docking of PLK2 with PubChem ID: 24971422, 11364421, and 49792852. The best target, 11364421, was used to conduct the molecular dynamics simulations. The results of this study unravel novel genes to P. gingivalis-associated AD that warrant further validation. ## 1. Introduction Alzheimer’s disease (AD) is characterised as a progressive neurological disorder that impacts thinking, behaviour, and memory [1]. The etiology of AD is poorly understood [2]. Many studies have found that AD is linked with genetics. Increasingly, much evidence has shown that AD can also be caused by environmental risk. Many bacteria, including Streptococcus mutans and Actinomyces viscosus, have been associated with an increased risk of AD development via modifying gene expression [3]. Studies found that infection with the herpes simplex virus 1 (HSV-1) could lead to changes in gene expression in human brain cells that were similar to those observed in the brains of Alzheimer’s patients [4]. The researchers suggested that these changes could contribute to the development of AD. However, it is not yet clear exactly how microbes might be capable of altering genes in the brain, but one possibility is that they may be able to trigger changes in the epigenome, which is the set of chemical modifications to DNA and associated proteins that regulate gene expression [5]. Some studies have suggested that infection with certain microbes may lead to changes in the epigenome that could contribute to the development of AD [6]. While the exact mechanisms by which microbes might alter genes in the brain are not yet fully understood, there is growing evidence to suggest that this could be one way in which these microbes contribute to the development and progression of AD [7]. Additionally, microbes may induce changes in the epigenome through a process called DNA methylation. DNA methylation is the addition of a methyl group to DNA, which can alter gene expression [8]. Some studies have suggested that infection with certain microbes may lead to changes in DNA methylation patterns in the brain, which could contribute to the development of AD [9]. More recently, studies have found that Porphyromonas gingivalis (P. gingivalis), a Gram-negative bacterium, is localised in the brain of humans with AD [10]. According to recent research, P. gingivalis can lead to systemic inflammation and is associated with other chronic inflammatory disorders such as rheumatoid arthritis and cardiovascular disease [11]. P. gingivalis has also been connected to systemic diseases such as endocarditis and sepsis [12]. Gene expression profiling using microarray and sequencing technologies has become a common tool in research and has the potential to provide a more comprehensive understanding of the molecular mechanisms underlying these diseases [13]. Multiple gene expression studies on AD have been conducted by different researchers, resulting in the availability of a large number of gene expression datasets [14]. By integrating these datasets, it is possible to identify key genes that are involved in the development and prognosis of P. gingivalis-associated AD [15]. Therefore, in this study, gene expressions of AD and P. gingivalis were downloaded from Gene Expression Omnibus (GEO) database. We identified differentially expressed genes (DEGs) by bioinformatic tools and identified DEGs related to AD and P. gingivalis. Additionally, we utilised clustering tools to identify hub genes using Gene Ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG), and protein–protein network interaction (PPI) analysis. Hub genes identified from this study provide insights into the molecular mechanism to elucidate the association between AD and P. gingivalis. ## 2.1. Data Normalisation R 4.2.1 software was used to analyse the original data. Firstly, quality was evaluated, data were normalised, and the expression density is shown in Figures S1 and S2 (Supplementary Data). We found 2540 upregulated and 1776 downregulated in GSE 5281. A total of 39 were upregulated and 11 downregulated in GSE9723, as shown in Figure 1. These DEGs are shown by volcano plots in Figure 2. The volcano plots shown identify genes that are $p \leq 0.05$ and are used as criteria for significance. The heat maps (Figure 2) show the hierarchical clustering of the DEGs. ## 2.2. DEGs Common to AD and P. gingivalis Using R, we found the common genes associated with datasets GSE5281 and GSE9723. The 10 genes were CALD1, HES1, ID3, PLK2, PPP2R2D, RASGRF1, SUN1, VPS33B, WTH3DI/RAB6A, and ZFP36L1. We then built a heat map to show the relativity of these found genes. Both of these findings are depicted in Figure 3. ## 2.3. GO and KEGG Analysis To elucidate the role of these genes, a Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway was constructed. Most of the functions of the common genes were involved in calcium-mediated signalling, actin binding, filopodium, and some involvement in pathogenic E. coli infection. The summary of this finding is presented in Figure 4. ## 2.4. PPI Network and Hub Gene Selection The PPI network from 10 identified common genes showed limited interaction. We then selected the top 100 genes (top 50 upregulated and top 50 downregulated) for our PPI network. The nodes indicate proteins, and the edges indicate their interactions, as shown in Figure 5. The specific functions of the genes and their indications are provided in Supplementary File S1. ## 2.5. CMap and Molecular Docking Our CMap analysis only yielded one result of the ten common genes screened. PLK2 was then chosen for further analysis. A total of three drug targets were identified to exert biological changes to protein PLK2. They are summarised in Table 1, and the results of molecular docking are shown in Figure 6. ## 2.6. Molecular Dynamic Simulations The best docking complex was chosen (PubChem ID: 11364421). The mobility characteristics of docked proteins are determined by deformability and B-factor. The deformability and B-factors of the PLK2 and 11364421 complexes show the peaks corresponding to deformable regions in the proteins, with the greatest peaks representing high deformability regions (Figure 7A). Figure 7B shows the eigenvalue and variance graphs of the PLK2 and 11364421 complexes. The variation graph of 11364421 with the target PLK2 shows individual variance with purple-shaded bars and cumulative variance with green-shaded bars. The complex covariance matrix depicts the correlations between residues in a complex. The red colour in Figure 7C in the matrix represents a good correlation between residues, while the white colour represents uncorrelated motion. Furthermore, the blue tint indicates anticorrelations. The higher the correlation, the more complicated the system. The docked proteins’ elastic maps (Figure 7D) show the atoms’ connections, with darker grey areas indicating stiffer regions. ## 3.1. Common Genes Found in P.gingivalis and AD CALD1 (Calcium-binding protein 1) is a protein that is encoded by the CALD1 gene in humans. It is a member of the calcyphosine family of calcium-binding proteins and is expressed in various tissues, including the brain. Some research has suggested that CALD1 may be involved in the development of AD, although more research is needed to fully understand its role in this condition [16]. In a particular study, it was found that CALD1 expression was significantly increased in the brains of people with AD compared to those without the condition and that CALD1 may be involved in the production of amyloid beta, a protein that is believed to be a key contributor to the development of AD [17,18]. Some research has suggested that CALD1 may be involved in the development of periodontitis, a type of gum disease characterised by inflammation and loss of the tissue and bone that support the teeth [19,20]. It was found that CALD1 expression was significantly increased in the gingiva of people with periodontitis compared to those without the condition and that CALD1 may be involved in the immune response to periodontitis. The HES1 protein is a member of the family of transcription factors known as basic helix–loop–helix (bHLH) factors, and it functions as a transcriptional repressor for genes whose transcription is dependent on the bHLH protein. The protein attaches to the N-box promoter region rather than the typical enhancer box because it has a particular type of basic domain with a helix-interrupting protein (E-box). According to a study, HSE1 Melatonin’s protective effect on soluble A1-42-induced memory impairment, astrogliosis, and synaptic dysfunction in the rat hippocampus through the Musashi1/Notch1/Hes1 signalling pathway [21]. According to certain studies, HES1 may contribute to the onset of oral cancer cells sustainably infected with P. gingivalis exhibit resistance to Taxol and have a higher metastatic potential [22]. The human ID3 gene produces the DNA-binding protein inhibitor ID-3 protein [23]. Helix–loop–helix (HLH) proteins belonging to the ID family lack a fundamental DNA-binding domain and suppress transcription by forming dimers that are ineffective at binding to DNA. Regarding histone H3K9me3-based epigenome signatures, research has indicated that ID3 is connected with synaptic impairment in AD [24]. In one study conducted, gene expression changes in the various functional categories related to periodontitis in adults and aged animals the ID3 were decreased with periodontitis [25]. Protein kinase serine/threonine, the PLK2 gene in humans, codes for the enzyme PLK2. The ‘polo’ family of serine/threonine protein kinases, which includes serum-inducible kinase, is involved in healthy cell division. They discovered that PLK2 activity inhibition alters APP and tau pathology and enhances synaptic content in a sex-dependent manner in Alzheimer’s dementia, suggesting that it may contribute to the aetiology of the illness [26]. A study has shown that PLK2’s significance in cancer is somewhat debatable; evidence points to both an oncogenic and a tumour suppressor role in a variety of malignancies [27]. The PPP2R2D gene in humans encodes the protein known as PP2A subunit B isoform delta, often referred to as serine/threonine-protein phosphatase 2A 55 kDa regulatory subunit B delta isoform. There is no research on this gene’s relationship to P. gingivalis or AD. A nuclear envelope protein with an UNC84 (SUN) domain is encoded by the protein-coding gene SUN1 (Sad1 and UNC84 domain containing 1). This gene belongs to the unc-84 homolog family. The protein aids in nuclear migration and anchoring. Spliced transcript variations have also been reported as an alternative (provided by RefSeq, January 2019). A previous study found that the accumulation of the inner nuclear envelope protein Sun1 is pathogenic in progeroid and dystrophic laminopathies, which results in AD [28]. No study has shown SUN1-gene-associated P. gingivalis. VPS33B (vacuolar protein sorting-associated protein 33B) is a protein that, in humans, is encoded by the VPS33B gene. The translational profile of striatopallidal neurons is preferentially altered by deep brain stimulation of the subthalamic nucleus in an animal model of Parkinson’s disease, and one of the genes they identified was VPS33B [29]. By integrating transcriptome analysis, the gene VPS33B is implicated in the different gene expression traits in the interactions between epithelial cells and P. gingivalis [30]. WTH3DI/RAB6A is a type of protein coding; members of the small GTPase superfamily’s RAB family are encoded by this gene. The targeting and fusing of transport carriers to acceptor compartments are regulated by the binding of GTPases of the RAB family to different effectors. This protein is found at the Golgi apparatus, which controls both retrograde and forward trafficking provided by HUGO Gene Nomenclature. No study was found related to this gene-associated AD and P. gingivalis. The ZFP36L1 gene belongs to the TIS11 family of early response genes, which are activated by a variety of agonists, including the polypeptide mitogen EGF and the phorbol ester TPA. This gene is highly conserved between species and features motifs found in other early-response genes in its promoter. A distinctive putative zinc finger domain with a recurring cys-his pattern can be found in the encoded protein. Most likely, this putative nuclear transcription factor controls how the body reacts to growth stimuli. This gene has been associated with a variety of alternatively spliced transcript variants that encode distinct isoforms (provided by RefSeq, September 2011). Characterisation of ZFP36L1 in the context of multiple sclerosis and functional immunological effects connected to the susceptibility to the disease, according to a study [31]. APN impairs the ability of macrophages, which play a significant role in periodontitis, to function. Through dependent signalling pathways, APN first stimulates the synthesis of TNF-α, which increases the expression of IL-10 and, as a result, reduces the inflammatory response of macrophages exposed to LPS [32]. Moreover, by promoting macrophage autophagy, APN can also reduce the expression of inflammatory mediators brought on by LPS. APN increases the production of ZFP36L1, which inhibits the interaction between Bcl-2 and Beclin-1 and, in turn, stimulates Beclin-1-activated autophagy in macrophages by destabilizing the mRNA of Bcl-2 [33]. A summary of the functions of the genes and possible roles is provided in Figure 8. ## 3.2. PPI Network Selected Hub Genes UCHL1, SST, CHGB, CALY, and INA Ubiquitin carboxy-terminal hydrolase L1 (UCHL1) belongs to a gene family whose members hydrolyze short C-terminal ubiquitin adducts to produce the ubiquitin monomer [34]. Highly specialised neurons, diffuse neuroendocrine system cells, and their tumours express UCHL1 [35]. It is known to play a role in AD [36]. It is thought that UCHL1 may be involved in the clearance of proteins that accumulate in the presence of AD, such as amyloid-beta [37]. UCHL1 has been identified as a potential marker for periodontitis [38]. The presence of UCHL1 in periodontitis may be associated with increased inflammation and tissue destruction [39]. UCHL1 may be a potential target for the treatment of AD and periodontitis [40]. Somatostatin is present throughout the body and binds to high-affinity G-protein-coupled somatostatin receptors to prevent the production of multiple secondary hormones. Through its interactions with thyroid stimulating hormone and pituitary growth hormone, this hormone plays a key role in the regulation of the endocrine system (provided by RefSeq, July 2008). It is also believed to play a role in the progression of AD [41]. Studies have shown that somatostatin levels are significantly lower in individuals with AD and that this decrease is associated with increased levels of the amyloid beta peptide, which is known to be a major cause of AD [42]. Somatostatin has been found to play a role in periodontitis, or inflammation of the gums; it is believed that somatostatin may be involved in the inflammatory process that leads to periodontitis [43]. Secretogranin-1, also known as Chromogranin B, is a protein that the CHGB gene in humans codes for; it is a member of the grain protein family. Chromogranin B is a gene that has been associated with AD. Studies have shown that individuals with AD have higher levels of CHGB in their brains than those without the disease. It is believed that CHGB may be involved in the accumulation of tau proteins, which are associated with AD, and the formation of amyloid plaques, which are also associated with the disease [44]. Currently, there is no evidence to suggest that CHGB, or Chromogranin B, is involved in periodontitis. However, research into this gene and its role in other inflammatory diseases may provide insight into its potential role in periodontitis. Neuron-specific vesicular protein is a type II single transmembrane protein that is expressed by the CALY gene in humans. It is necessary for the maximum accelerated calcium release upon stimulation of purinergic or muscarinic receptors [45]. CALY has been associated with AD. Studies have shown that levels of Calcyon are significantly lower in individuals with AD than in those without the disease [46]. It is believed that Calcyon may be able to regulate the activity of certain proteins that are involved in the progression of AD [47]. Currently, there is no evidence to suggest that neuron-specific vesicular protein Calcyon is involved in periodontitis. Alpha-internexin is a type of class IV intermediate filament with a mass of 66 kDa. The rat spinal cord and optic nerve were originally used to purify the protein [48]. A comparable central rod domain found in alpha-internexin includes about 310 amino acid residues and forms a highly conserved alpha-helical region. An area called the amino-terminal head, and a region called the carboxy-terminal tail, surround the core rod domain, which is in charge of the coiled-coil structure [49]. INA encodes neuronal intermediate filament protein found mostly in the neurons of the nervous system during early development [50]. No study has shown a relationship between the INA gene and periodontitis. ## 3.3. Molecular Docking of PLK2 with 11364421 (C28H39N703) We used iMODS to evaluate and define the flexibility of PLK2 with 11364421. We hypothesize that based on the conceivable interactions of the identified proteins with PLK2, they can serve as prospective therapeutic candidates and targets to attenuate the pathological process in P. gingivalis-associated AD. ## 4.1. Pipeline of Research The flowchart of the conduct of this study is summarised in Figure 9. ## 4.2. Data Acquisition The National Centre for Biotechnology Information (NCBI) produced and maintains the gene expression database known as the GEO database. It includes high throughput gene expression data provided by international research institutions [51]. To determine the molecular mechanism of the occurrence and development of P. gingivalis-associated AD, two microarray datasets from the GEO database, namely, GSE5281 and GSE9723, were downloaded. The former was a study conducted on Alzheimer’s patients, and the latter was using cells infected with P. gingivalis. The data’s executive summary is presented in Table 2 below. ## 4.3. DEGs Screening Using the limma package in R, we identified the differentially expressed genes (DEGs) between AD and P. gingivalis. The criteria of $p \leq 0.05$ and log2FC > 1 were set before entering to VENN tool to visualise the intersecting genes. ## 4.4. Functional Analysis of Common Genes The Database for Annotation, Visualization, and Integrated Discovery (DAVID) web tool is used to carry out GO enrichment analysis and KEGG pathway analysis [55]. The three categories of GO analysis, which include biological processes (BPs), cellular components (CCs), and molecular functions (MFs), were performed [56]. A p-value < 0.05 is considered significant. ## 4.5. Construction of PPI Network A protein–protein interaction (PPI) analysis was carried out using the STRING database to investigate the relationships between common genes (https://string-db.org/) (accessed on 26 December 2022). All known and anticipated protein–protein interactions, including both physical and details functional relationships, are listed in the STRING database [57]. The minimum interaction score was set to 0.400 as a criterion for statistical significance. The PPI is illustrated by the nodes as proteins and lines as interactions. ## 4.6. CMap Analysis, Molecular Docking, and Simulation We will be predicting small drug molecules using CMap in accordance with specific gene expression signatures offered by the database. A small drug molecule with a negative mean score may be able to reverse biological effects and so have potential therapeutic utility ($p \leq 0.05$ is considered significant). The effectiveness and binding potential were assessed using molecular docking. Firstly, crystal structures are obtained from RCSB Protein Data Bank (PDB, http://www.rcsb.org) (accessed on 26 December 2022), and the mol2 file formats of the compounds were retrieved from the PubChem database. PyMOL 2.3.1 was used to dehydrate target proteins and remove ligands. AutoDock Tool 1.5.6 software was used to hydrogenate, calculate its charge, and store it in PDBQT format. PyRx and Autodock Vina v1.2.0 software was used to visualize the findings of molecular docking. The iMOD server iMOD server (iMODS) (http://imods.chaconlab.org) (accessed on 26 December 2022) was used to run molecular dynamics simulations to calculate the stability and molecular mobility of the bound protein-113644421 complexes. In addition to calculating molecular mobility, iMODS was employed to analyse the structural dynamics of the docking complexes. The elastic network, deformability, B-factor, eigenvalues, variance, and covariance map were all used to show how stable the protein-11364421 complexes were. All parameters for the docked PDB files used as input were left at default when they were uploaded to the iMODS server. ## 5. Conclusions In this study, 10 crossover genes were identified, indicating a potential association between AD and P. gingivalis as the key pathogen in periodontitis. Hence, one potential therapeutic target was identified for P. gingivalis-induced-AD treatment. However, further in vivo and in vitro studies are needed to confirm these findings. ## References 1. Ball S.L., Holland A.J., Watson P.C., Huppert F.A.. Theoretical Exploration of the Neural Bases of Behavioural Disinhibition, Apathy and Executive Dysfunction in Preclinical Alzheimer’s Disease in People with Down’s Syndrome: Potential Involvement of Multiple Frontal-Subcortical Neuronal Circuits. J. Intellect. Disabil. Res. (2010.0) 54 320-336. DOI: 10.1111/j.1365-2788.2010.01261.x 2. Jellinger K.A.. Alzheimer’s Disease. Neurobiol. Dis. (2007.0) 69-82. DOI: 10.1016/B978-012088592-3/50008-6 3. Balolong M.P., Mendoza M.A.F., Balolong M.P., Mendoza M.A.F.. Understanding Oral Diseases: Exploring Opportunities from Filipino Oral Microbiome Research. Dent. Caries (2021.0). DOI: 10.5772/INTECHOPEN.94751 4. Viral Role in Alzheimer’s Disease Discovered\|University of Oxford 5. Sharma V.K., Mehta V., Singh T.G.. Alzheimer’s Disorder: Epigenetic Connection and Associated Risk Factors. Curr. Neuropharmacol. (2020.0) 18 740. DOI: 10.2174/1570159X18666200128125641 6. Lunnon K., Mill J.. Epigenetic Studies in Alzheimer’s Disease: Current Findings, Caveats and Considerations for Future Studies. Am. J. Med. Genet. Part B Neuropsychiatr. Genet. Off. Publ. Int. Soc. Psychiatry Genet. (2013.0) 162 789. DOI: 10.1002/ajmg.b.32201 7. Fülöp T., Itzhaki R.F., Balin B.J., Miklossy J., Barron A.E.. Role of Microbes in the Development of Alzheimer’s Disease: State of the Art—An International Symposium Presented at the 2017 IAGG Congress in San Francisco. Front. Genet. (2018.0) 9 362. DOI: 10.3389/fgene.2018.00362 8. Jackson J.P., Lindroth A.M., Cao X., Jacobsen S.E.. Control of CpNpG DNA Methylation by the KRYPTONITE Histone H3 Methyltransferase. Nature (2002.0) 416 556-560. DOI: 10.1038/nature731 9. Balazs R.. Epigenetic Mechanisms in Alzheimer’s Disease. Degener. Neurol. Neuromuscul. Dis. (2014.0) 4 85. DOI: 10.2147/DNND.S37341 10. Elwishahy A., Antia K., Bhusari S., Ilechukwu N.C., Horstick O., Winkler V.. Porphyromonas Gingivalis as a Risk Factor to Alzheimer’s Disease: A Systematic Review. J. Alzheimers Dis. Rep. (2021.0) 5 721. DOI: 10.3233/ADR-200237 11. Mei F., Xie M., Huang X., Long Y., Lu X., Wang X., Chen L.. Pathogens (2020.0) 9. DOI: 10.3390/pathogens9110944 12. Li X., Kolltveit K.M., Tronstad L., Olsen I.. Systemic Diseases Caused by Oral Infection. Clin. Microbiol. Rev. (2000.0) 13 547. DOI: 10.1128/CMR.13.4.547 13. O’Brien M.A., Costin B.N., Miles M.F.. Using Genome-Wide Expression Profiling to Define Gene Networks Relevant to the Study of Complex Traits: From RNA Integrity to Network Topology. Int. Rev. Neurobiol. (2012.0) 104 91. DOI: 10.1016/B978-0-12-398323-7.00005-7 14. Kim K., Zakharkin S.O., Allison D.B.. Expectations, Validity, and Reality in Gene Expression Profiling. J. Clin. Epidemiol. (2010.0) 63 950. DOI: 10.1016/j.jclinepi.2010.02.018 15. Olsen I., Singhrao S.K.. Interaction between Genetic Factors, Porphyromonas Gingivalis and Microglia to Promote Alzheimer’s Disease. J. Oral Microbiol. (2020.0) 12 1820834. DOI: 10.1080/20002297.2020.1820834 16. What Causes Alzheimer’s Disease?\|National Institute on Aging 17. Kumari A., Rahaman A., Zeng X.-A., Farooq M.A., Huang Y., Yao R., Ali M., Ishrat R., Ali R.. Temporal Cortex Microarray Analysis Revealed Impaired Ribosomal Biogenesis and Hyperactivity of the Glutamatergic System: An Early Signature of Asymptomatic Alzheimer’s Disease. Front. Neurosci. (2022.0) 16 966877. DOI: 10.3389/fnins.2022.966877 18. Pereira A.C., Gray J.D., Kogan J.F., Davidson R.L., Rubin T.G., Okamoto M., Morrison J.H., McEwen B.S.. Age and Alzheimer’s Disease Gene Expression Profiles Reversed by the Glutamate Modulator Riluzole. Mol. Psychiatry (2017.0) 22 296. DOI: 10.1038/mp.2016.33 19. Zhang W., Wu Q., Su J., Wang W., Zhao X., Cao P.. Circulating MiRNAs as Epigenetic Mediators of Periodontitis and Preeclampsia Association. Dis. Markers (2022.0) 2022 2771492. DOI: 10.1155/2022/2771492 20. Gonzalez O.A., Kirakodu S., Nguyen L.M., Orraca L., Novak M.J., Gonzalez-Martinez J., Ebersole J.L.. Comparative Analysis of Gene Expression Patterns for Oral Epithelial Cell Functions in Periodontitis. Front. Oral Health (2022.0) 3 863231. DOI: 10.3389/froh.2022.863231 21. Zhang S., Wang P., Ren L., Hu C., Bi J.. Protective Effect of Melatonin on Soluble Aβ. Alzheimer’s Res. Ther. (2016.0) 8 40. DOI: 10.1186/s13195-016-0206-x 22. Woo B.H., Kim D.J., Choi J.I., Kim S.J., Park B.S., Song J.M., Lee J.H., Park H.R.. Oral Cancer Cells Sustainedly Infected with Porphyromonas Gingivalis Exhibit Resistance to Taxol and Have Higher Metastatic Potential. Oncotarget (2017.0) 8 46981. DOI: 10.18632/oncotarget.16550 23. Ellmeier W., Aguzzi A., Kleiner E., Kurzbauer R., Weith A.. Mutually Exclusive Expression of a Helix-Loop-Helix Gene and N-Myc in Human Neuroblastomas and in Normal Development. EMBO J. (1992.0) 11 2563-2571. DOI: 10.1002/j.1460-2075.1992.tb05321.x 24. Lee J., Roberts J.A.S., Atanasova K.R., Chowdhury N., Han K., Yilmaz Ö.. Human Primary Epithelial Cells Acquire an Epithelial-Mesenchymal-Transition Phenotype during Long-Term Infection by the Oral Opportunistic Pathogen, Porphyromonas Gingivalis. Front. Cell. Infect. Microbiol. (2017.0) 7 493. DOI: 10.3389/fcimb.2017.00493 25. Ebersole J.L., Al-Sabbagh M., Dawson D.R.. Heterogeneity of Human Serum Antibody Responses to. Immunol. Lett. (2020.0) 218 11. DOI: 10.1016/j.imlet.2019.12.004 26. Martínez-Drudis L., Sheta R., Pellegrinato R., Loukili S., Tremblay C., Calon F., Rivest S., Oueslati A.. Inhibition of PLK2 Activity Affects APP and Tau Pathology and Improves Synaptic Content in a Sex-Dependent Manner in a 3xTg Mouse Model of Alzheimer’s Disease. Neurobiol. Dis. (2022.0) 172 105833. DOI: 10.1016/j.nbd.2022.105833 27. de Cárcer G.. The Mitotic Cancer Target Polo-Like Kinase 1: Oncogene or Tumor Suppressor?. Genes (2019.0) 10. DOI: 10.3390/genes10030208 28. Chen C.Y., Chi Y.H., Mutalif R.A., Starost M.F., Myers T.G., Anderson S.A., Stewart C.L., Jeang K.T.. Accumulation of the Inner Nuclear Envelope Protein Sun1 Is Pathogenic in Progeric and Dystrophic Laminopathies. Cell (2012.0) 149 565-577. DOI: 10.1016/j.cell.2012.01.059 29. Visanji N.P., Sarvestani I.K., Creed M.C., Shoaei Z.S., Nobrega J.N., Hamani C., Hazrati L.N.. Deep Brain Stimulation of the Subthalamic Nucleus Preferentially Alters the Translational Profile of Striatopallidal Neurons in an Animal Model of Parkinson’s Disease. Front. Cell. Neurosci. (2015.0) 9 221. DOI: 10.3389/fncel.2015.00221 30. Zhang D., Hou J., Wu Y., Liu Y., Li R., Xu T., Liu J., Pan Y.. Distinct Gene Expression Characteristics in Epithelial Cell-Porphyromonas Gingivalis Interactions by Integrating Transcriptome Analyses. Int. J. Med. Sci. (2019.0) 16 1320. DOI: 10.7150/ijms.33728 31. Ottoboni L., Radassi K., De Jager P., Comi G., Farina C., Martino G.. Characterization of ZFP36L1 in the Context of Multiple Sclerosis and Functional Immunological Consequences Associated with the Susceptibility to the Disease. J. Neuroimmunol. (2014.0) 275 52. DOI: 10.1016/j.jneuroim.2014.08.137 32. Park P.H., McMullen M.R., Huang H., Thakur V., Nagy L.E.. Short-Term Treatment of RAW264.7 Macrophages with Adiponectin Increases Tumor Necrosis Factor-α (TNF-α) Expression via ERK1/2 Activation and Egr-1 Expression: Role of TNF-α in adiponectin-stimulated interleukin-10 production. J. Biol. Chem. (2007.0) 282 21695-21703. DOI: 10.1074/jbc.M701419200 33. Shrestha A., Pun N.T., Park P.H.. ZFP36L1 and AUF1 Induction Contribute to the Suppression of Inflammatory Mediators Expression by Globular Adiponectin via Autophagy Induction in Macrophages. Biomol. Ther. (2018.0) 26 446. DOI: 10.4062/biomolther.2018.078 34. Doran J.F., Jackson P., Kynoch P.A.M., Thompson R.J.. Isolation of PGP 9.5, a New Human Neurone-Specific Protein Detected by High-Resolution Two-Dimensional Electrophoresis. J. Neurochem. (1983.0) 40 1542-1547. DOI: 10.1111/j.1471-4159.1983.tb08124.x 35. Das C., Hoang Q.Q., Kreinbring C.A., Luchansky S.J., Meray R.K., Ray S.S., Lansbury P.T., Ringe D., Petsko G.A.. Structural Basis for Conformational Plasticity of the Parkinson’s Disease-Associated Ubiquitin Hydrolase UCH-L1. Proc. Natl. Acad. Sci. USA (2006.0) 103 4675-4680. DOI: 10.1073/pnas.0510403103 36. Zhao Z.B., Wu L., Xiong R., Wang L.L., Zhang B., Wang C., Li H., Liang L., Chen S.D.. MicroRNA-922 Promotes Tau Phosphorylation by Downregulating Ubiquitin Carboxy-Terminal Hydrolase L1 (UCHL1) Expression in the Pathogenesis of Alzheimer’s Disease. Neuroscience (2014.0) 275 232-237. DOI: 10.1016/j.neuroscience.2014.06.013 37. Oddo S.. The Ubiquitin-Proteasome System in Alzheimer’s Disease. J. Cell. Mol. Med. (2008.0) 12 363. DOI: 10.1111/j.1582-4934.2008.00276.x 38. Lin L., Li S., Hu S., Yu W., Jiang B., Mao C., Li G., Yang R., Miao X., Jin M.. UCHL1 Impairs Periodontal Ligament Stem Cell Osteogenesis in Periodontitis. J. Dent. Res. (2022.0) 102 00220345221116031. DOI: 10.1177/00220345221116031 39. Cobb C.M., Singla O., Feil P.H., Theiesen F.C., Schultz R.E.. Comparison of NK-Cell (Leu-7+ and Leu-11b+) Populations in Clinically Healthy Gingiva, Chronic Gingivitis and Chronic Adult Periodontitis. J. Periodontal Res. (1989.0) 24 1-7. DOI: 10.1111/j.1600-0765.1989.tb00851.x 40. Jiang Z., Shi Y., Zhao W., Zhou L., Zhang B., Xie Y., Zhang Y., Tan G., Wang Z.. Association between Chronic Periodontitis and the Risk of Alzheimer’s Disease: Combination of Text Mining and GEO Dataset. BMC Oral Health (2021.0) 21. DOI: 10.1186/s12903-021-01827-2 41. Solarski M., Wang H., Wille H., Schmitt-Ulms G.. Somatostatin in Alzheimer’s Disease: A New Role for an Old Player. Prion (2018.0) 12 1-8. DOI: 10.1080/19336896.2017.1405207 42. Saido T.C.. Metabolism of Amyloid β Peptide and Pathogenesis of Alzheimer’s Disease. Proc. Jpn. Acad. Ser. B Phys. Biol. Sci. (2013.0) 89 321. DOI: 10.2183/pjab.89.321 43. Memmert S., Damanaki A., Nokhbehsaim M., Nogueira A.V.B., Eick S., Cirelli J.A., Jäger A., Deschner J.. Regulation of Somatostatin Receptor 2 by Proinflammatory, Microbial and Obesity-Related Signals in Periodontal Cells and Tissues. Head Face Med. (2019.0) 15 2. DOI: 10.1186/s13005-018-0185-1 44. Lechner T., Adlassnig C., Humpel C., Kaufmann W.A., Maier H., Reinstadler-Kramer K., Hinterhölzl J., Mahata S.K., Jellinger K.A., Marksteiner J.. Chromogranin Peptides in Alzheimer’s Disease. Exp. Gerontol. (2004.0) 39 101-113. DOI: 10.1016/j.exger.2003.09.018 45. Heijtz R.D., Alexeyenko A., Xavier F.X.. Calcyon MRNA Expression in the Frontal-Striatal Circuitry and Its Relationship to Vesicular Processes and ADHD. Behav. Brain Funct. BBF (2007.0) 3 33. DOI: 10.1186/1744-9081-3-33 46. Muthusamy N., Chen Y.J., Yin D.M., Mei L., Bergson C.. Complementary Roles of the Neuron-Enriched Endosomal Proteins NEEP21 and Calcyon in Neuronal Vesicle Trafficking. J. Neurochem. (2015.0) 132 20-31. DOI: 10.1111/jnc.12989 47. Vazdarjanova A., Bunting K., Muthusamy N., Bergson C.. Calcyon Up-Regulation in Adolescence Impairs Response Inhibition and Working Memory in Adulthood. Mol. Psychiatry (2011.0) 16 672. DOI: 10.1038/mp.2011.14 48. Levavasseur F., Zhu Q., Julien J.P.. No Requirement of α-Internexin for Nervous System Development and for Radial Growth of Axons. Mol. Brain Res. (1999.0) 69 104-112. DOI: 10.1016/S0169-328X(99)00104-7 49. Lariviere R.C., Julien J.P.. Functions of Intermediate Filaments in Neuronal Development and Disease. J. Neurobiol. (2004.0) 58 131-148. DOI: 10.1002/neu.10270 50. Wang P., Wang S.M., Hsieh C.J., Chien C.L.. Neural Expression of α-Internexin Promoter In Vitro and In Vivo. J. Cell. Biochem. (2006.0) 97 275-287. DOI: 10.1002/jcb.20643 51. Edgar R., Domrachev M., Lash A.E.. Gene Expression Omnibus: NCBI Gene Expression and Hybridization Array Data Repository. Nucleic Acids Res. (2002.0) 30 207-210. DOI: 10.1093/nar/30.1.207 52. Readhead B., Haure-Mirande J.V., Funk C.C., Richards M.A., Shannon P., Haroutunian V., Sano M., Liang W.S., Beckmann N.D., Price N.D.. Multiscale Analysis of Independent Alzheimer’s Cohorts Finds Disruption of Molecular, Genetic, and Clinical Networks by Human Herpesvirus. Neuron (2018.0) 99 64-82.e7. DOI: 10.1016/j.neuron.2018.05.023 53. Liang W.S., Dunckley T., Beach T.G., Grover A., Mastroeni D., Walker D.G., Caselli R.J., Kukull W.A., McKeel D., Morris J.C.. Gene Expression Profiles in Anatomically and Functionally Distinct Regions of the Normal Aged Human Brain. Physiol. Genom. (2007.0) 28 311-322. DOI: 10.1152/physiolgenomics.00208.2006 54. Handfield M., Mans J.J., Zheng G., Lopez M.C., Mao S., Progulske-Fox A., Narasimhan G., Baker H.V., Lamont R.J.. Distinct Transcriptional Profiles Characterize Oral Epithelium-Microbiota Interactions. Cell. Microbiol. (2005.0) 7 811-823. DOI: 10.1111/j.1462-5822.2005.00513.x 55. Huang D.W., Sherman B.T., Lempicki R.A.. Systematic and Integrative Analysis of Large Gene Lists Using DAVID Bioinformatics Resources. Nat. Protoc. (2009.0) 4 44-57. DOI: 10.1038/nprot.2008.211 56. Blake J.A., Christie K.R., Dolan M.E., Drabkin H.J., Hill D.P., Ni L., Sitnikov D., Burgess S., Buza T., Gresham C.. Gene Ontology Consortium: Going Forward. Nucleic Acids Res. (2015.0) 43 D1049-D1056. DOI: 10.1093/NAR/GKU1179 57. Szklarczyk D., Gable A.L., Nastou K.C., Lyon D., Kirsch R., Pyysalo S., Doncheva N.T., Legeay M., Fang T., Bork P.. The STRING Database in 2021: Customizable Protein–Protein Networks, and Functional Characterization of User-Uploaded Gene/Measurement Sets. Nucleic Acids Res. (2021.0) 49 D605-D612. DOI: 10.1093/nar/gkaa1074
--- title: 'Hepatitis Risk in Diabetes Compared to Non-Diabetes and Relevant Factors: A Cross-Sectional Study with National Health and Nutrition Examination Survey (NHANES), 2013–2018' authors: - Ja-Young Han - Jae-Hee Kwon - Sun-Hwa Kim - Heeyoung Lee journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049568 doi: 10.3390/ijerph20064962 license: CC BY 4.0 --- # Hepatitis Risk in Diabetes Compared to Non-Diabetes and Relevant Factors: A Cross-Sectional Study with National Health and Nutrition Examination Survey (NHANES), 2013–2018 ## Abstract This study aimed to identify the development of hepatitis B or C infection in diabetes patients compared to those without and to elucidate factors associated with the prevalence of hepatitis B or C infection in diabetes. We conducted a cross-sectional study using data from the National Health and Nutrition Examination Survey (NHANES) 2013–2018. As evaluation factors, we included variables such as age, race, illicit drug use, and poverty. The diabetic group had a significantly higher prevalence of hepatitis B or C infection than the non-diabetic group (odds ratio (OR) = 1.73; $95\%$ confidence interval (CI), 1.36–2.21, $p \leq 0.01$). In multivariate Cox regression, non-poverty and non-illicit drug use were lower risk factors contributing to hepatitis development in diabetes (hazard ratio (HR) = 0.50; $95\%$ CI, 0.32–0.79, $p \leq 0.01$, and HR = 0.05; $95\%$ CI, 0.03–0.08, $p \leq 0.01$, respectively). Logistic regression also showed that these factors were significant contributors to hepatitis development in the diabetic group ($p \leq 0.01$). In patients with diabetes, the development of hepatitis was higher than that in those without, and hepatitis development was influenced by poverty and illicit drug use. This may provide supporting evidence of response strategies for diabetes to care for hepatitis development in advance. ## 1. Introduction Diabetes and hepatitis are among the most prevalent diseases worldwide [1]. The concomitant existence of diabetes and hepatitis potentially leads to a life-threatening status, increasing mortality by approximately $17\%$ among diabetic patients without comparing to other patients, such as non-virus-infected diabetic patients [2]. With a considerably high prevalence, 865 outbreaks of patients infected with hepatitis B virus (HBV) through population-based surveillance for infectious diseases at eight Emerging Infections Program (EIP) sites, which are part of a network of 10 state health departments in the United States [3], have been reported among adults diagnosed with diabetes [4]. Regarding the increased prevalence of HBV in patients with diabetes and severe clinical outcomes, the Advisory Committee on Immunization Practices (ACIP) of the Centers for Disease Control and Prevention (CDC) recommends appropriate methods for preventing HBV infection in patients with diabetes based on the risk of an inadequate immune response [5]. In addition, for hepatitis C virus (HCV) infection, one of the major sources of morbidity and mortality in people, the CDC indicated that injection drug use (IDU) is a primary risk factor for infection [6]. Considering IDU use in diabetic patients, according to the guidelines for diabetes, diabetic patients are given insulin as the primary treatment and use self-monitoring of blood glucose (SMBG) to manage their blood glucose, so frequent exposure to needles increases the risk of HCV transmission [7,8]. Regarding the frequent exposure to IDU along with needles through multiple blood sampling and unsafe injection practices, patients with diabetes are more vulnerable to HCV infection than those without diabetes [9]. A previous cross-sectional study which analyzed the prevalence of HCV among diabetic patients reported that insulin users were 3.2 times more likely to have HCV infection than non-insulin users [10]. Furthermore, hepatitis and diabetes seem to be negatively associated with the development of each other. Schillie et al. [ 11] reported a higher prevalence rate of HBV infection among persons with diabetes than those without diabetes (odds ratio (OR) = 1.60; $95\%$ confidence interval (CI), 1.30–1.90, $p \leq 0.05$). However, there is still controversy regarding the association between the development of diabetes and hepatic viral infection, as shown in a study in a tertiary hospital reporting low development of hepatitis C viral seropositivity among patients with type 2 diabetes mellitus [12]. To provide more substantial evidence explaining the association between diabetes and hepatitis, including HBV or HCV, large-scale epidemiological studies evaluating national data, such as the National Health and Nutrition Examination Survey (NHANES), are needed. Currently, a limited number of studies have analyzed large-scale data, such as the NHANES, which might provide a clear explanation of the association between diabetes and hepatitis. Thus, to provide sufficient evidence to evaluate the impact of diabetes on HBV or HCV development compared to the population without diabetes, the current study examined the association between hepatitis infection and diabetes and aimed to characterize the factors related to the prevalence of hepatitis in the diabetic population using the NHANES database. ## 2.1. Study Design We analyzed the 2013–2018 NHANES database. The NHANES is a cross-sectional monitoring program designed to assess the health and nutritional status of adults and children in the United States [13]. Multiple datasets were collected in this survey, including demographics, dietary, questionnaire, physical examinations, and laboratory testing of biologic samples in the U.S. populations [13]. The NHANES data have been collected in 2-year cycles without a break between cycles since 1999 [14]. The methodology for all databases is described on the NHANES website [14]. Since the participants of the NHANES are de-identified and assigned unique sequence numbers, details of the respondents and duplicate participation of people are unknown [15]. The NHANES is administered using a stratified multistage clustered probability sampling strategy to provide a nationally representative sample [15]. ## 2.2. Definition of Diabetes and Hepatitis Information on diabetes, HBV, and HCV were retrieved from the NHANES. Diabetes was defined according to the following criteria: HbA1c ≥ $6.5\%$, FPG ≥ 126 mg/dL, or participants being informed that they had diabetes by their doctor or other health professionals [16]. Both patients with type 1 and type 2 diabetes were included in our study. HBV was defined as positive for hepatitis B surface antigen (HBsAg) or if participants answered yes to the question, “Has a doctor or other health professional ever told you that you have hepatitis B?” [ 17]. HCV was defined as a positive for Hepatitis C virus ribonucleic acid (HCV-RNA) or if participants replied yes to the question “Has a doctor or other health professional ever told you that you have Hepatitis C?” [ 18]. HBV-infected and HCV-infected participants were assigned to the corresponding hepatitis B and C groups. ## 2.3. Covariates We selected significant variables through logistic regression for variable selection to be included in the analysis dataset [19]. In our analysis, the confounding variables included age, sex, race, body mass index (BMI), the ratio of family income to poverty, and the use of illicit drugs. Demographic covariates included age (years) [20], sex (male or female), race (Mexican-American, other Hispanic, non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, or other race) [21], and family income to poverty ratio (continuous from 0 to 4.99) [18]. We categorized age into four groups as follows: 0–19, 20–44, 45–64, and ≥65 years [20]. The family income to poverty ratio was divided into < or ≥1 [19]. A ratio < 1 indicated poverty, while a ratio ≥ 1 indicated without poverty [18]. BMI was collected from the laboratory dataset from the NHANES database and grouped as underweight (<18.5 kg/m²), normal weight (18.5~24.9 kg/m²), and overweight (≥25 kg/m²) [22]. Questionnaire covariates included the use of illicit drugs (yes or no) [23]. ## 2.4. Statistical Analysis Data are presented as frequency and percentage for categorical variables. Box plots consisted of quartiles for the prevalence over 2-year cycles of patients infected with hepatitis regarding diabetes status [24]. The OR with $95\%$ CI was calculated using univariate and multivariate logistic regression models to elucidate the factors affecting the development of hepatitis in patients with diabetes. Multivariate logistic regression analysis was performed after adjusting for covariates of race, poverty, and illicit drug use. The Cox proportional hazards model determined the hazard rate to be roughly constant at all time points, so our study used this statistical method [25,26]. The Cox proportional hazards model was used to calculate the hazard ratio (HR) and $95\%$ CI of the risk associated with hepatitis development in patients with diabetes. The univariate Cox proportional hazards model was used for all variables, and a multivariate model was utilized to adjust the race, poverty, and use of illicit drugs. All statistical analyses were performed using R software (version 4.2.1), with a p-value < 0.05 as statistically significant. ## 3.1. General Characteristics of Participants This study included 29,400 participants from the NHANES (2013–2018) database. Overall, 1160 participants with missing diabetes data were excluded. Then, 15,305 participants with missing data on illicit drug use, 28 participants with missing data on HBV, 3 participants with missing data on HCV, 1215 participants with missing data on the ratio of family income to poverty, and 88 participants with missing data on BMI were excluded in the present analysis. The additional analysis to clarify the variance between excluded data, including data related to illegal drug use, showed that the number of participants differed among age groups (Table S1). However, the distributions of excluded data for non-responding questions related to illegal drugs were not significantly different while the number of participants distributed among age groups differed. Ultimately, 11,601 participants were included in the final analysis (Figure 1 and Table 1). Of the 11,601 participants, 1709 were diagnosed with diabetes, whereas 9892 were classified in the non-diabetic group. The prevalence of HBV, HCV, and hepatitis B or C with diabetes was $2.3\%$, $3.2\%$, and $5.1\%$, respectively, whereas the prevalence in the non-diabetic group was $1.3\%$, $1.9\%$, and $3.0\%$, respectively (Table 2). ## 3.2. The Prevalence of Hepatitis According to Diabetes Status The prevalence of HBV in participants without diabetes was $1.3\%$, whereas that in participants with diabetes was $2.3\%$, which indicated HBV development was 1.7721 odds more than in patients with diabetes, with a significant difference (OR = 1.77; $95\%$ CI, 1.24–2.53, $p \leq 0.05$). The prevalence of HCV was $3.2\%$ among patients with diabetes and $1.9\%$ among those without diabetes. The development of HCV increased in patients with diabetes by 1.6751 odds compared to patients without diabetes, with a significant difference (OR = 1.68; $95\%$ CI, 1.23–2.28, $p \leq 0.05$). Moreover, the prevalence of $5.1\%$ for hepatitis B or C in patients with diabetes was significantly 1.7298 odds higher than that reported among non-diabetic subjects (OR = 1.73; $95\%$ CI, 1.36–2.21, $p \leq 0.0001$) (Figure 2, Table 2 and Table S2). Furthermore, the prevalence of hepatitis infection was not significantly different based on the FPG or HbA1c levels (Figure S1 and Table S3). ## 3.3. Characteristics of Hepatitis Development from 2013 to 2018 Analysis of the characteristics of hepatitis prevalence according to diabetes development at 2-year intervals (2013–2018) revealed that the prevalence of HBV in patients with diabetes increased from $1.5\%$ to $3.0\%$ in two cycles (2013–2016), and the non-diabetic group also showed an increased tendency from $1.3\%$ to $1.7\%$. However, compared with the 2013–2016 period, the HBV prevalence in 2017–2018 decreased to $2.4\%$ in patients with diabetes and $1.1\%$ in the non-diabetic group. The prevalence of HCV in the diabetic population decreased over time (2013–2018), though the non-diabetic group showed similar HCV development for two cycles (2013–2016) and indicated an increasing tendency towards $2.3\%$ from 2017 to 2018. In the diabetic group, the hepatitis B or C prevalence increased from $5.0\%$ to $5.7\%$ during 2013–2016, while in 2017–2018, a decreasing trend in the rate was observed (Table 2 and Figure S2). ## 3.4. Risk of Hepatitis Development in Diabetes Patients In univariate and multivariate Cox regression analysis results, the non-poverty group showed a tendency to have a lower risk of hepatitis B or C prevalence than the poverty group (univariate: HR = 0.48; $95\%$ CI, 0.31–0.74, $p \leq 0.01$; multivariate: HR = 0.50; $95\%$ CI, 0.32–0.79, $p \leq 0.01$, respectively) (Table 3). In addition, diabetes patients who did not use illicit drugs had a lower risk of hepatitis B or C infection than illicit drug users (univariate: HR = 0.05; $95\%$ CI, 0.03–0.08, $p \leq 0.01$; multivariate: HR = 0.05; $95\%$ CI, 0.03–0.08, $p \leq 0.01$, respectively). In multivariate analysis, non-Hispanic Asian particularly showed a greater association in hepatitis B or C prevalence than other races among the diabetic group. ( HR = 3.33; $95\%$ CI, 1.05–10.60, $p \leq 0.05$). ## 3.5. Factors of Developing Hepatitis in Diabetes Patients The results of the univariate and multivariate logistic regression analyses are summarized in Table 4. In the diabetic group, the univariate and multivariate logistic analyses showed that the non-poverty group was associated with a lower risk of hepatitis B or C development than the poverty group (univariate: OR 0.96; $95\%$ CI, 0.94–0.99, $p \leq 0.01$; multivariate: OR 0.96; $95\%$ CI, 0.94–0.99, $p \leq 0.01$, respectively). The diabetic group of non-users of illicit drugs was associated with reduced odds of hepatitis B or C infection compared with illicit drug users (univariate: OR 0.59; $95\%$ CI, 0.55–0.62, $p \leq 0.01$; multivariate: OR 0.59; $95\%$ CI, 0.56–0.62, $p \leq 0.01$, respectively). The non-Hispanic Asian group showed a tendency of higher correlation with hepatitis B or C infection than other races (univariate: OR 1.03; $95\%$ CI, 0.97–1.09, $$p \leq 0.40$$; multivariate: OR 1.05; $95\%$ CI, 0.99–1.11, $$p \leq 0.10$$, respectively). ## 4. Discussion The current study evaluated the prevalence of HBV and HCV infections in patients with diabetes and relevant factors. Our results showed that patients with diabetes had a higher prevalence of HBV and HCV infections than those without. Similarly, a previous cohort study conducted in the United Kingdom observed a higher prevalence of HBV in patients with diabetes than in those without diabetes [27]. Another previous study from Pakistan Hospital also indicated that patients with diabetes were more associated with HCV infection than the non-diabetic group (OR = 3.03; $95\%$ CI, 2.64–3.48, $p \leq 0.01$) [28]. Furthermore, due to their weakened immune system [29], patients with diabetes may be more susceptible to HBV or HCV infections compared to the general population. This susceptibility may be exacerbated by frequent exposure to needles during blood glucose monitoring, which can contribute to virus transmission [30]. Our findings demonstrated that poverty was a potential contributor that might influence HBV and HCV infection development. Patients with diabetes living in poverty were more associated with a higher development of hepatitis than those without poverty. Consistent with these findings, a previous cross-sectional study in Canada [31], which reported the association between socioeconomic income and the prevalence of diabetes and related conditions, also showed that those with a low household income had a higher prevalence of diabetes than the population with a higher income. Greene et al. [ 32] analyzed surveillance data in New York City and reported that chronic hepatitis C was included in diseases related to severe poverty with low income. Furthermore, based on the American Association for the Study of Liver Disease (AASLD), HBV and HCV are the leading infectious diseases that are closely related to poverty [33]. Although various factors contributed to poverty, such as age or education levels, and are closely related to infectious disease spread [34], considering the low self-awareness of their condition among impoverished people, the poverty group could show a progressive increase in the risk of disease such as hepatitis [35]. In support of this, in Brazil, a previous study also showed hepatitis B susceptibility rates ranging from approximately $32\%$ among individuals living with low income [36]. Therefore, poverty in people with complications such as DM is the cause of increasing HBV or HCV and is also likely to be an important factor in the incidence rate because there is a cost burden [37]. In our results, patients with diabetes without illicit drugs seem to have a lower risk of hepatitis infection than illicit drug users. Diabetic patients who are at risk of experiencing high stress and immune dysfunction may also be more susceptible to illegal drug use and HCV transmission [38,39]. For the association between HCV transmission and illicit drug use, Benjamin et al. [ 40] already demonstrated in a study of young American users of illegal drugs that 343 out of 714 participants were infected with HCV. Furthermore, the increase in the distribution of hepatitis caused by illegal drug use seems to be a global burden [41]. This is evidenced by the population-attributable fraction of hepatitis caused by illegal drug use in 2013, which was $10\%$ for HBV in North America and $1\%$ in Latin America and $81\%$ for HCV in North America and $31\%$ in Latin America [41]. This percentage has further increased since 1990, with HBV accounting for $6\%$ in North America and $1\%$ in Latin America and HCV accounting for $60\%$ in North America and $19\%$ in Latin America [41]. However, given the tendency for illicit drug users to frequently disregard physician advice, hepatitis infections in this population may become more severe or even go undetected [42]. Therefore, these findings can offer crucial insights to enhance screening protocols and identify a broader population of illegal drug users at high risk of acquiring HCV and HBV infections. This will aid in the early detection and treatment of the infections [43]. Among races, non-Hispanic Asians race seemed to significantly contribute to hepatitis in the diabetic group. Consistent with these findings, a previous study showed that non-Hispanic Asians were more associated with the prevalence of HBV infection than other races (OR = 3.85; $95\%$ CI, 2.97–4.97, $p \leq 0.05$) [44]. In addition, according to 2011–2014 NHANES data, non-Hispanic Asian adults showed a higher prevalence of HBV infection ($22.6\%$) than non-Hispanic White ($2.6\%$), non-Hispanic Black ($10.2\%$), and Hispanic ($3.6\%$) adults [45]. Furthermore, in our study, the non-Hispanic Black race seemed to contribute to hepatitis infection without significance. Pathologically, according to Thomas et al., interleukin 28 B (IL28B), which plays a significant primary role in the resolution of HCV, is less likely to be present in the Black population [46]. Because of the lack of IL28B, the non-Hispanic Black group may have a higher risk of HCV infection [46]. However, our study cannot be confidential because race affects the risk of hepatitis; therefore, further studies are needed. The current study had some limitations. First, our study did not distinguish between type 1 and type 2 diabetes mellitus. As type 2 diabetes accounts for more than $90\%$ of diagnosed diabetes mellitus in the United States [47], our findings largely reflect the risk factors of hepatitis in patients with type 2 diabetes mellitus. Thus, we expect that more future studies will be conducted to distinguish the types of diabetes mellitus to identify the impact of diabetes and hepatitis. Second, we did not evaluate the economic costs of treating HCV infection and diabetes mellitus. Therefore, further studies are required to evaluate the economic impact of HBV or HCV infection in patients with diabetes. Third, we were unable to conduct further analysis on the correlation between hepatitis infection and CVD or diabetic comorbidities in the current study. This was because the important cardiovascular health metrics data, which could aid analysis of the correlation of hepatitis infection with complications such as cardiovascular disease or diabetic complications, were scarce in the NHANES [48], or the guidelines for evaluating criteria for high blood pressure and cholesterol risks were changed [49,50]. Thus, we hope that future studies can address these limitations. Forth, the NHANES data were limited to the United States. This may not provide global evidence for an association between diabetes and hepatitis. Thus, we expect further studies using merged data from various countries. Finally, as this study used a small sample size to identify illegal injection drug users, the number might have been underestimated [51]. It is difficult to obtain accurate data about illegal drug users. Thus, there may be inevitable non-responsive biases with illicit drug use, and our findings should be interpreted with caution. ## 5. Conclusions The current study found that the risk of hepatitis B or C infection was higher in patients with diabetes than in those without. Therefore, the present study would help increase awareness regarding hepatitis prevention in patients with diabetes. Additionally, this study suggests that more attention should be paid to impoverished or illicit drug users among patients with diabetes regarding the threat of hepatitis infection. Finally, in addition to NHANES data, we expect that more global evidence will be provided through corresponding data from various countries. ## References 1. Ziaee M., Saljoughi M., Fardinfar F., Madarshahian F., Ebrahimzadeh A.. Comparison of Frequency Distribution of Diabetes in Hepatitis B and C Patients and Other People in Birjand. Mod. Care J. (2021.0) 18 e111024. DOI: 10.5812/modernc.111024 2. Singh K.K., Panda S.K., Shalimar S.K.. Patients with Diabetes Mellitus are Prone to Develop Severe Hepatitis and Liver Failure due to Hepatitis Virus Infection. J. Clin. Exp. Hepatol. (2013.0) 3 275-280. DOI: 10.1016/j.jceh.2013.11.003 3. Division of Preparedness and Emerging Infections (DPEI). (2021.0) 4. Use of Hepatitis B Vaccination for Adults with Diabetes Mellitus: Recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Morb. Mortal. Wkly. Rep. (2011.0) 60 1709-1711. PMID: 22189894 5. Kirkman M.S., Schaffner W.. Another shot to protect people with diabetes: Add hepatitis B vaccination to the checklist. Diabetes Care (2012.0) 35 941-942. DOI: 10.2337/dc12-0164 6. Schillie S., Wester C., Osborne M., Wesolowski L., Ryerson A.B.. CDC Recommendations for Hepatitis C Screening among Adults—United States, 2020. MMWR Recomm. Rep. (2020.0) 69 1-17. DOI: 10.15585/mmwr.rr6902a1 7. Silver B., Ramaiya K., Andrew S.B., Fredrick O., Bajaj S., Kalra S., Charlotte B.M., Claudine K., Makhoba A.. EADSG Guidelines: Insulin Therapy in Diabetes. Diabetes Ther. (2018.0) 9 449-492. DOI: 10.1007/s13300-018-0384-6 8. Self-monitoring of blood glucose among adults with diabetes—United States, 1997–2006. MMWR. Morb. Mortal. Wkly. Rep. (2007.0) 56 1133-1137. PMID: 17975526 9. Guo X., Jin M., Yang M., Liu K., Li J.W.. Type 2 diabetes mellitus and the risk of hepatitis C virus infection: A systematic review. Sci Rep. (2013.0) 3 2981. DOI: 10.1038/srep02981 10. Ba-Essa E.M., Mobarak E.I., Al-Daghri N.M.. Hepatitis C virus infection among patients with diabetes mellitus in Dammam, Saudi Arabia. BMC Health Serv. Res. (2016.0) 16. DOI: 10.1186/s12913-016-1578-0 11. Schillie S.F., Xing J., Murphy T.V., Hu D.J.. Prevalence of hepatitis B virus infection among persons with diagnosed diabetes mellitus in the United States, 1999–2010. J. Viral Hepat. (2012.0) 19 674-676. DOI: 10.1111/j.1365-2893.2012.01616.x 12. Balogun W.O., Adeleye J.O., Akinlade K.S., Kuti M., Otegbayo J.A.. Low prevalence of hepatitis-C viral seropositivity among patients with type-2 diabetes mellitus in a tertiary hospital. J. Natl. Med. Assoc. (2006.0) 98 1805-1808. PMID: 17128691 13. Neves J.S., Leitão L., Magriço R., Bigotte Vieira M., Viegas Dias C., Oliveira A., Carvalho D., Claggett B.. Caffeine Consumption and Mortality in Diabetes: An Analysis of NHANES 1999–2010. Front. Endocrinol. (2018.0) 9 547. DOI: 10.3389/fendo.2018.00547 14. National Health and Nutrition Examination Survey 15. Zipf G., Chiappa M., Porter K.S., Ostchega Y., Lewis B.G., Dostal J.. National Health and Nutrition Examination Survey: Plan and Operations, 1999–2010 (2013.0) 1-37 16. Huang Y.Q., Liu L., Huang C., Yu Y.L., Lo K., Huang J.Y., Chen C.L., Zhou Y.L., Feng Y.Q.. Impacts of Pre-Diabetes or Prehypertension on Subsequent Occurrence of Cardiovascular and All-Cause Mortality among Population without Cardiovascular Diseases. Diabetes Metab. Syndr. Obes. (2020.0) 13 1743-1752. DOI: 10.2147/DMSO.S255842 17. Zeng L.Y., Lian J.S., Chen J.Y., Jia H.Y., Zhang Y.M., Xiang D.R., Yu L., Hu J.H., Lu Y.F., Zheng L.. Hepatitis B surface antigen levels during natural history of chronic hepatitis B: A Chinese perspective study. World J. Gastroenterol. (2014.0) 20 9178-9184. DOI: 10.3748/wjg.v20.i27.9178 18. Moore K.J., Gauri A., Koru-Sengul T.. Prevalence and sociodemographic disparities of Hepatitis C in Baby Boomers and the US adult population. J. Infect. Public Health (2019.0) 12 32-36. DOI: 10.1016/j.jiph.2018.08.003 19. Chowdhury M.Z.I., Turin T.C.. Variable selection strategies and its importance in clinical prediction modelling. Fam. Med. Community Health (2020.0) 8 e000262. DOI: 10.1136/fmch-2019-000262 20. Fang J., Zhang Z., Ayala C., Thompson-Paul A.M., Loustalot F.. Cardiovascular Health among Non-Hispanic Asian Americans: NHANES, 2011–2016. J. Am. Heart Assoc. (2019.0) 8 e011324. DOI: 10.1161/JAHA.118.011324 21. Ansu Baidoo V.Y., Zee P.C., Knutson K.L.. Racial and Ethnic Differences in Eating Duration and Meal Timing: Findings from NHANES 2011–2018. Nutrients (2022.0) 14. DOI: 10.3390/nu14122428 22. Ellison-Barnes A., Johnson S., Gudzune K.. Trends in Obesity Prevalence among Adults Aged 18 through 25 Years, 1976–2018. JAMA (2021.0) 326 2073-2074. DOI: 10.1001/jama.2021.16685 23. Shing J.Z., Ly K.N., Xing J., Teshale E.H., Jiles R.B.. Prevalence of Hepatitis B Virus Infection Among US Adults Aged 20–59 Years with a History of Injection Drug Use: National Health and Nutrition Examination Survey, 2001–2016. Clin. Infect. Dis. (2020.0) 70 2619-2627. DOI: 10.1093/cid/ciz669 24. Soares L., Latta S.C., Ricklefs R.E.. Neotropical migratory and resident birds occurring in sympatry during winter have distinct haemosporidian parasite assemblages. J. Biogeogr. (2019.0) 47 748-759. DOI: 10.1111/jbi.13760 25. Barros A.J., Hirakata V.N.. Alternatives for logistic regression in cross-sectional studies: An empirical comparison of models that directly estimate the prevalence ratio. BMC Med. Res. Methodol. (2003.0) 3. DOI: 10.1186/1471-2288-3-21 26. Mirel L.B., El Bural Félix S., Zhang C., Golden C., Cox C.S.. Comparative Analysis of the National Health Interview Survey Public-Use and Restricted-Use Linked Mortality Files (2020.0) 1-32 27. Ferreira G.L.C., Marano C., De Moerlooze L., Guignard A., Feng Y., El Hahi Y., van Staa T.. Incidence and prevalence of hepatitis B in patients with diabetes mellitus in the UK: A population-based cohort study using the UK Clinical Practice Research Datalink. J. Viral Hepat. (2018.0) 25 571-580. DOI: 10.1111/jvh.12841 28. Jadoon N.A., Shahzad M.A., Yaqoob R., Hussain M., Ali N.. Seroprevalence of hepatitis C in type 2 diabetes: Evidence for a positive association. Virol. J. (2010.0) 7 304. DOI: 10.1186/1743-422X-7-304 29. Muller L.M., Gorter K.J., Hak E., Goudzwaard W.L., Schellevis F.G., Hoepelman A.I., Rutten G.E.. Increased risk of common infections in patients with type 1 and type 2 diabetes mellitus. Clin. Infect. Dis. (2005.0) 41 281-288. DOI: 10.1086/431587 30. Villar L.M., Geloneze B., Vasques A.C.J., Pires M.L.E., Miguel J.C., da Silva E.F., Marques V.A., Scalioni L.P., Lampe E.. Prevalence of hepatitis B and hepatitis C among diabetes mellitus type 2 individuals. PLoS ONE (2019.0) 14. DOI: 10.1371/journal.pone.0211193 31. Bird Y., Lemstra M., Rogers M., Moraros J.. The relationship between socioeconomic status/income and prevalence of diabetes and associated conditions: A cross-sectional population-based study in Saskatchewan, Canada. Int. J. Equity Health (2015.0) 14 93. DOI: 10.1186/s12939-015-0237-0 32. Greene S.K., Levin-Rector A., Hadler J.L., Fine A.D.. Disparities in Reportable Communicable Disease Incidence by Census Tract-Level Poverty, New York City, 2006–2013. Am. J. Public Health (2015.0) 105 e27-e34. DOI: 10.2105/AJPH.2015.302741 33. Ghany M.G., Morgan T.R.. Hepatitis C Guidance 2019 Update: American Association for the Study of Liver Diseases-Infectious Diseases Society of America Recommendations for Testing, Managing, and Treating Hepatitis C Virus Infection. Hepatology (2020.0) 71 686-721. DOI: 10.1002/hep.31060 34. Janjua P.Z., Kamal U.A.. The Role of Education and Income in Poverty Alleviation: A Cross-Country Analysis. Lahore J. Econ. (2011.0) 16 143-172. DOI: 10.35536/lje.2011.v16.i1.a6 35. Kim H.S., Rotundo L., Yang J.D., Kim D., Kothari N., Feurdean M., Ruhl C., Unalp-Arida A.. Racial/ethnic disparities in the prevalence and awareness of Hepatitis B virus infection and immunity in the United States. J. Viral Hepat. (2017.0) 24 1052-1066. DOI: 10.1111/jvh.12735 36. Scarponi C.F.O., Zolnikov T.R., Mol M.P.G.. Are waste pickers at risk for hepatitis B and C infections because of poverty or environmental exposures?. Rev. Soc. Bras. Med. Trop (2019.0) 52 e20190123. DOI: 10.1590/0037-8682-0123-2019 37. Shi L., Fonseca V., Childs B.. Economic burden of diabetes-related hypoglycemia on patients, payors, and employers. J. Diabetes Its Complicat. (2021.0) 35 107916. DOI: 10.1016/j.jdiacomp.2021.107916 38. Berbudi A., Rahmadika N., Tjahjadi A.I., Ruslami R.. Type 2 Diabetes and its Impact on the Immune System. Curr. Diabetes Rev. (2020.0) 16 442-449. DOI: 10.2174/1573399815666191024085838 39. Aregbesola A., Onyeka I.N., Olubamwo O., Ronkainen K., Tiihonen J., Föhr J., Kauhanen J.. Diabetes hospitalizations and deaths in a cohort of treatment-seeking illicit drug users. SAGE Open Med. (2018.0) 6 2050312118768164. DOI: 10.1177/2050312118768164 40. Eckhardt B., Winkelstein E.R., Shu M.A., Carden M.R., McKnight C., Des Jarlais D.C., Glesby M.J., Marks K., Edlin B.R.. Risk factors for hepatitis C seropositivity among young people who inject drugs in New York City: Implications for prevention. PLoS ONE (2017.0) 12. DOI: 10.1371/journal.pone.0177341 41. Degenhardt L., Charlson F., Stanaway J., Larney S., Alexander L.T., Hickman M., Cowie B., Hall W.D., Strang J., Whiteford H.. Estimating the burden of disease attributable to injecting drug use as a risk factor for HIV, hepatitis C, and hepatitis B: Findings from the Global Burden of Disease Study 2013. Lancet Infect. Dis. (2016.0) 16 1385-1398. DOI: 10.1016/S1473-3099(16)30325-5 42. Edlin B.R., Seal K.H., Lorvick J., Kral A.H., Ciccarone D.H., Moore L.D., Lo B.. Is it justifiable to withhold treatment for hepatitis C from illicit-drug users?. N. Engl. J. Med. (2001.0) 345 211-215. DOI: 10.1056/NEJM200107193450311 43. Edlin B.R.. Prevention and treatment of hepatitis C in injection drug users. Hepatology (2002.0) 36 S210-S219. DOI: 10.1053/jhep.2002.36809 44. Le M.H., Yeo Y.H., Cheung R., Henry L., Lok A.S., Nguyen M.H.. Chronic Hepatitis B Prevalence among Foreign-Born and U.S.-Born Adults in the United States, 1999–2016. Hepatology (2020.0) 71 431-443. DOI: 10.1002/hep.30831 45. Kruszon-Moran D., Paulose-Ram R., Denniston M., McQuillan G.. Viral Hepatitis among Non-Hispanic Asian Adults in the United States, 2011–2014 (2015.0) 1-8 46. Thomas D.L., Thio C.L., Martin M.P., Qi Y., Ge D., O’Huigin C., Kidd J., Kidd K., Khakoo S.I., Alexander G.. Genetic variation in IL28B and spontaneous clearance of hepatitis C virus. Nature (2009.0) 461 798-801. DOI: 10.1038/nature08463 47. Bullard K.M., Cowie C.C., Lessem S.E., Saydah S.H., Menke A., Geiss L.S., Orchard T.J., Rolka D.B., Imperatore G.. Prevalence of Diagnosed Diabetes in Adults by Diabetes Type—United States, 2016. MMWR. Morb. Mortal. Wkly. Rep. (2018.0) 67 359-361. DOI: 10.15585/mmwr.mm6712a2 48. Lopez-Neyman S.M., Davis K., Zohoori N., Broughton K.S., Moore C.E., Miketinas D.. Racial disparities and prevalence of cardiovascular disease risk factors, cardiometabolic risk factors, and cardiovascular health metrics among US adults: NHANES 2011–2018. Sci. Rep. (2022.0) 12 19475. DOI: 10.1038/s41598-022-21878-x 49. Whelton P.K., Carey R.M., Aronow W.S., Casey D.E., Collins K.J., Dennison Himmelfarb C., DePalma S.M., Gidding S., Jamerson K.A., Jones D.W.. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the Prevention, Detection, Evaluation, and Management of High Blood Pressure in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J. Am. Coll. Cardiol. (2018.0) 71 e127-e248. DOI: 10.1016/j.jacc.2017.11.006 50. Grundy S.M., Stone N.J., Bailey A.L., Beam C., Birtcher K.K., Blumenthal R.S., Braun L.T., de Ferranti S., Faiella-Tommasino J., Forman D.E.. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J. Am. Coll. Cardiol. (2019.0) 73 3168-3209. DOI: 10.1016/j.jacc.2018.11.002 51. Lee S.O., Lee J.E., Lee S., Lee S.H., Kang J.S.. Prevalence and patterns of illicit drug use in people with human immunodeficiency virus infection in Korea. PLoS ONE (2021.0) 16. DOI: 10.1371/journal.pone.0249361
--- title: 'Effectiveness of Health Coaching in Smoking Cessation and Promoting the Use of Oral Smoking Cessation Drugs in Patients with Type 2 Diabetes: A Randomized Controlled Trial' authors: - Li-Chi Huang - Yao-Tsung Chang - Ching-Ling Lin - Ruey-Yu Chen - Chyi-Huey Bai journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049574 doi: 10.3390/ijerph20064994 license: CC BY 4.0 --- # Effectiveness of Health Coaching in Smoking Cessation and Promoting the Use of Oral Smoking Cessation Drugs in Patients with Type 2 Diabetes: A Randomized Controlled Trial ## Abstract Introduction: This study looked into the effectiveness of a 6 month health coaching intervention in smoking cessation and smoking reduction for patients with type 2 diabetes. Methods: The study was carried out via a two-armed, double-blind, randomized-controlled trial with 68 participants at a medical center in Taiwan. The intervention group received health coaching for 6 months, while the control group only received usual smoking cessation services; some patients in both groups participated in a pharmacotherapy plan. The health coaching intervention is a patient-centered approach to disease management which focuses on changing their actual behaviors. By targeting on achieving effective adult learning cycles, health coaching aims to help patients to establish new behavior patterns and habits. Results: In this study, the intervention group had significantly more participants who reduced their level of cigarette smoking by at least $50\%$ than the control group ($$p \leq 0.030$$). Moreover, patients participating in the pharmacotherapy plan in the coaching intervention group had a significant effect on smoking cessation ($$p \leq 0.011$$), but it was insignificant in the control group. Conclusions: Health coaching can be an effective approach to assisting patients with type 2 diabetes participating in a pharmacotherapy plan to reduce smoking and may help those who participate in pharmacotherapy plan to quit smoking more effectively. Further studies with higher-quality evidence on the effectiveness of health coaching in smoking cessation and the use of oral smoking cessation drugs in patients with type 2 diabetes are needed. ## 1. Introduction Diabetes is a chronic disease which inflicts heavy burdens on medical systems and social systems, as well as causes numerous deaths, and the number of people throughout the world suffering from diabetes is increasing. The World Health Organization (WHO) notes that the global prevalence rate of diabetes among adults now tops $8.5\%$ [1]. In Taiwan, the prevalence rate of diabetes among the population aged 18 and older stands at around $9.82\%$ [2], ranking second place in health insurance outlay [3]. The prevention and treatment of diabetes involve lifestyle modification, among which smoking cessation is an important and challenging task to address. Smoking can cause insulin resistance, thereby increasing the risk of diabetes and accelerating the rate of diabetic deterioration [4]. Although nicotine withdrawal may increase the risk of obesity due to an increased appetite in the short term, it gradually decreases as the time from quitting smoking increases [5]. In the long run, quitting smoking has many benefits for people with diabetes. However, the association between smoking and diabetes is often overlooked by patients. The smoking rate in Taiwan in 2020 was about $13.1\%$, of which men aged 31–50 had the highest prevalence of smoking, reaching a high of $30\%$. Since 2012, Taiwan has promoted “The second-generation smoking cessation program”, which has greatly extended drug subsidy and increased participation flexibility [6]. According to the Health Promotion Administration (HPA) Taiwan, the 6 month smoking cessation clinical services, which combine the smoking cessation drugs and counseling services, can achieve a $27.1\%$ smoking abstinence rate, and the free online smoking cessation consultation hotline also has a $39.5\%$ smoking abstinence rate [6]. There are more studies conducted on groups with higher smoking rates, such as males and disadvantaged groups, or groups of patients who are more easily affected by smoking behaviors, such as pregnant women, hypertensive persons, and lung cancer patients, with relatively few studies focusing on people with diabetes [7,8,9]. Although there are some studies indicating the benefits of smoking cessation to patients with diabetes [10], in Taiwan, the current evidence of smoking cessation in diabetic patients appears to be insufficient. There are quite a few studies on smoking cessation interventions using health education or behavioral counseling for more than 20 years, and motivational interviewing (MI) is one of the techniques used to aid quitting [7,11]. It involves a series of guided questions specifically aimed at patients’ motivation to change, especially focusing on inducing ambivalence to increase patients’ motivation. However, in recent years, health coaching has emerged as an increasingly popular approach to behavioral change; however, since there seems to be only a very small number of smoking cessation-related health coaching studies, the behavioral counseling for smoking cessation is still dominated by MI. Health coaching is a patient-centered approach to disease management which focuses on patients’ decisions and their actual behaviors [12]; it is based on positive psychology and humanistic psychology, integrating many behavioral counseling theories and techniques such as MI, appreciation inquiry, design thinking, solution-focused therapy, and adult learning theory, and it focuses on achieving effective adult learning cycles to enable patients to make changes and establish new behavior patterns and habits. Compared with MI, health coaching is a more comprehensive approach, which makes up for the limitations of using MI alone, as it utilizes various skills and is more focused on adult learning and empathy. In recent years, in addition to traditional coach training institutions (such as the International Coach Federation, ICF), there are some institutions such as the National Board for Health and Wellness Coaching (NBHWC) that specialize in certifying and training health coaches and have a clear definition and specification for the coach [13,14]. Therefore, health coaching has a more complete framework than counseling using MI or any other technique alone, and there are international training, certification, and monitoring bodies that clearly define coaching and its practice [14]. Although no studies are available to compare whether implementing integrative method such as health coaching is significantly better than using a single technique such as MI, health coaching has become a mainstream in behavior modification for patients with chronic disease such as diabetes mellitus in recent years. However, because many health coaching studies often lack sufficient information on coach training, certification, and supervision, or do not clearly describe the methods used by coaches, it is impossible to effectively confirm whether coaches truly meet the definition of coaching by the aforementioned institutions, and many existing coaching studies may still be health education or counseling [12,15,16]. Therefore, the aim of this study was to test the effectiveness of a health coaching intervention to help diabetic patients quit smoking and/or participate in a pharmacotherapy plan under the premise of providing sufficient intervener information. ## 2. Materials and Methods The study involved a 6 month coaching intervention in a two-armed, double-blind, randomized controlled trial, approved by the Institutional Review Board (IRB) of Cathay General Hospital (Taipei, Taiwan). All relevant ethical safeguards were met in relation to participant protection, and ethical standards were in accordance with the Declaration of Helsinki. Two groups of subjects participated: one with a monthly coaching intervention and the other with usual care only. ## 2.1. Study Procedures and Randomization Settings Participants were recruited among patients with diabetes treated at the Department of Endocrinology and Metabolism at Cathay General Hospital (Taipei, Taiwan) from August 2020 and May 2021; however, this study had to be discontinued in May 2021 due to the coronavirus disease 2019 (COVID-19) pandemic. In addition, unfortunately, due to the risk from nitrosamine impurity in varenicline, it was requested to be taken off the shelves by the health authorities in Taiwan in July 2021 [17], and all the smoking cessation projects in Taiwan that year were suspended until the following year. The suspension of cessation programs along with the drug recall halted patients’ participation in smoking cessation drug treatment, resulting in the absence of post-test data collection. A diabetes health educator screened and tagged patients with type 2 diabetes and smoking habit from the hospital’s database, followed by an independent researcher who randomly assigned them to the intervention group and the control group using computer-generated random numbers by the PASW 22.0 software for windows (SPSS, Chicago, IL, USA). Then, two physicians recruited them individually. Patients in the intervention group were informed of the coaching program by a health coach, whereas the control group was provided with the usual smoking cessation plan (Figure 1). Therefore, the physicians, patients, and analysts were blinded, the participants in the control group did not know that this was an interventional trial, and the coach was not blinded since he was aware of the patients in the intervention group. The inclusion criteria were as follows: age 20 to 75 years with type 2 diabetes for at least 1 year, Mandarin Chinese or Taiwanese as the spoken language, smoking one or more cigarettes per day, and not taking cessation medication or joining any cessation plan. The exclusion criteria were signs of clinical depression or cognitive function impairment. Between August 2020 and May 2021, prior to the COVID-19 restrictions in May 2021 in Taiwan, 92 potential subjects were screened from database but only 74 subjects were enrolled in the study. In total, outcome measures were available for 35 participants in the intervention group and 33 in the control group, while six participants withdrew from the follow-up due to the drug recall (Figure 1). The recruitment rate was $80.4\%$, and the main reason for participants’ refusal was the unwillingness to quit smoking at that stage. ## 2.2. Pharmacotherapy Plan Both groups could freely choose to participate in a 2 month varenicline treatment plan. In Taiwan, the National Health Insurance (NHI) provides a maximum of two varenicline treatment subsidies a year, each for a period of 2 months, with three-stage drug treatment [18]. Only those assessed with the Fagerström Test for Nicotine Dependence (FTND) with a score of >3 could participate in the program. Participants had to return to the clinic in the third and seventh weeks to receive medicine, and their nicotine addiction was assessed again. The degree of addiction was tracked in the 12th and 24th weeks. Participants of the intervention group were referred to the case manager after participating in the pharmacotherapy plan. During the 2 month varenicline treatment, they also received coaching from the coach when returning to the hospital; afterward, coaching intervention resumed until the end of the 6 month intervention. Therefore, participants in the intervention group were tracked by the coach after the pharmacotherapy plan, while those in the control group were tracked by the case manager only. In total, four case managers were responsible for managing and tracking all the patients in the pharmacotherapy plan, and those four were nurses who are qualified diabetes health educators. ## 2.3. Intervention Patients in the intervention group received in-person coaching at baseline and were followed up by a monthly telephone call for a total of 6 months, or they could participate in the pharmacotherapy plan and have a coaching intervention the other 4 months. Coaching was provided on a one-on-one basis by a professional health coach who has a master’s degree in public health and received over 120 h of coach training before being certified as an International Coach Federation (ICF)’s Associated Certified Coach (ACC). To be specific, the coach used techniques such as MIs, appreciation inquiry, and value exploration to arouse participants’ motivation to quit smoking and allowed participants to weigh the pros and cons of quitting smoking due to diabetes and the needs of their personal work or family life. The coach used the patient-centered principle, self-determination theory, and transtheoretical model to guide the patients to think via techniques such as active listening and open-ended questions, instead of using traditional preaching methods of health education, thereby strengthening the psychological support and motivation felt by the patients. For some patients who have irrational beliefs about smoking behavior, such as the belief that smoking is a necessary means to establish working partnerships or carry out their creative work, the coach tried to use rational emotive cognitive behavior coaching (RECBC) skill for coaching. In the first session, the coach asked each participant to establish their 6 month smoking cessation goals, and then the coach assisted them in formulating a plan to gradually reduce smoking. Participants needed to design “SMART” (i.e., specific, measurable, attainable, realistic, and timely) goals for their action plans, and they could design a smoking reduction plan on a weekly or monthly basis, or just stop smoking straight away. In addition, the coach also asked the participants if they felt the need to use smoking cessation drugs, and some decided to participate in the pharmacotherapy plan during or after the initial coaching session. Most coaching calls were made while the participants were at home or during lunch break. During follow-up telephone coaching sessions, the coach continued to enhance each patient’s motivation to quit smoking, reviewed the goal and action plan with the patient, and discussed possible obstacles to quitting and the solutions together. Through coaching, each patient could rethink the necessity of smoking and its health effects; however, even if the patient chose to just reduce the number of cigarettes instead of quitting smoking entirely due to work or family needs, the coach still respected the patient’s choice. In general, it took more time, about 30–45 min, in the first coaching session than in follow-up telephone coaching sessions. Both intervention and control group patients received the usual smoking cessation education and tracking. Currently, the smoking cessation health education adopted in Taiwan includes the five As (ask, advise, assess interest, assist, and arrange) and five Rs (relevance, risks, rewards, roadblocks, and repetition) strategy of smoking cessation promoted by the WHO to strengthen the effectiveness of smoking cessation [19]. Health educator also provides health education relating to smoking cessation such as the disadvantages of smoking, withdrawal syndromes, and medication information according to the standardized smoking cessation consulting training designed by the Health Promotion Administration, Ministry of Health and Welfare Taiwan [20]. Patients in the control group did not receive any additional intervention and only had follow-up calls from the case manager in the 4th, 8th, 12th, and 24th weeks, where they were asked if they had quit smoking and/or required any health education or treatment needs. Therefore, patients in the control group had fewer follow-up times than those in the intervention group. ## 2.4. Sample Size Considering that smoking cessation behaviors may be affected by the current policy environment and culture in Taiwan, when calculating the sample size, we referred to the sample size calculation of medical or behavior intervention studies conducted in Taiwan and published in international peer-reviewed journals in recent years [21]. It seems that if effect size is used as a sample size estimation, a moderate effect size (ES = 0.5) may be an appropriate estimation criterion. Hence, to detect a 0.50 effect size with a probability of a type I error of 0.05 and a power of $80\%$, each group required at least 64 participants. ## 2.5. Outcome Measures The main outcome variables of this study were smoking abstinence and smoking reduction. The smoking abstinence was assessed by Fagerström Test for Nicotine Dependence (FTND), with zero points indicating success in quitting smoking. Although there are many smoking cessation studies using carbon monoxide measurement as the smoking cessation assessment indicator to determine whether a patient has smoked during the assessment period [22], it was not used in this study due to the COVID-19 epidemic prevention considerations; instead, the FTND was used. This self-assessment tool has six items with an overall score ranging 0–10 [23], and it is also the main smoking abstinence criterion for the second-generation smoking cessation service in Taiwan. The smoking reduction was assessed by participants’ self-reported number of cigarettes smoked per day, and a significant reduction was defined as they reduced their daily smoking by at least $50\%$ from baseline [24]. The independent variables included the coaching intervention, baseline number of daily cigarettes, length of smoking (years), the FTND score, pharmacotherapy plan participation, experience with smoking cessation, and sociodemographic characteristics. Sociodemographic characteristics included gender, age, educational level, and job position. ## 2.6. Statistical Analysis A chi-square test or t-test was employed to assess differences in sociodemographic variables and baseline smoking behaviors. The chi-square test was used to assess the difference in smoking cessation and reduction in smoking at the follow-up, and the paired t-test was used to assess the reduction in the number of cigarettes. The effect size of intervention was also calculated. Lastly, we used multivariate logistic regression to test whether coach intervention has higher odds of quitting smoking than the usual smoking cessation plan in Taiwan and adjusted the baseline value of daily cigarettes. Since there were no significant differences in demographic characteristics between the two groups, we included only the daily number of cigarettes at baseline as an adjustment variable. All tests were analyzed at a $95\%$ significance level ($p \leq 0.05$). Intention-to-treat analysis was not used since the ethical policy states that noncompliers who refuse to continue to participate are to be excluded from the analysis. Analyses were conducted using PASW 22.0 software for windows (SPSS, Chicago, IL, USA). ## 3. Results Table 1 shows the demographic characteristics of the 68 participants: $82.4\%$ were male, the mean age was 56.0 (standard deviation (SD) = 10.62) years, $39.7\%$ had a bachelor’s degree or higher, the mean length of smoking was 34.5 (SD = 9.70) years, mean addiction to nicotine was 5.3 (FTND score, SD = 2.46) points, $82.4\%$ had never tried to quit smoking, and about half of the participants were willing to use varenicline to quit smoking. There were no significant demographic differences between the two groups at baseline. With the 6 month coaching intervention or the 2 month varenicline use plus 4 month coaching intervention, the intervention group and the control group had $48.6\%$ and $36.4\%$ of participants, respectively, who quit smoking for a nonsignificant difference in the smoking cessation rate between the two groups ($$p \leq 0.309$$, Table 2). However, the intervention group had significantly more participants who reduced their smoking than the control group ($$p \leq 0.030$$), and there was a significantly greater reduction in the number of cigarettes than the control group ($$p \leq 0.032$$) with Cohen’s d of 0.53. In the intervention group, $54.3\%$ of participants decreased their smoking with a significant ($p \leq 0.001$) decrease of 12.88 (SD = 9.28) cigarettes per day. In the control group, $21.2\%$ of participants decreased their smoking with a significant ($p \leq 0.001$) decrease of 7.74 (SD = 10.03) cigarettes per day. Health coaching intervention seemed to improve the effectiveness of smoking cessation among varenicline users and smoking reduction for the participants without receiving varenicline treatment (Table 3). With the use of varenicline, it seems that the intervention group had more participants quit smoking than the control group ($$p \leq 0.082$$). In contrast, there were more participants in the intervention group reducing the number of cigarettes smoked when receiving coaching only rather than with the use of varenicline ($$p \leq 0.014$$). As to the smoking cessation rate, the multivariate logistic regression revealed that only varenicline use (odds ratio (OR) = 3.67, $95\%$ confidence interval (CI) = 1.27–10.60) significantly predicted successful smoking cessation (Nagelkerke R2 = 0.160, Table 4). When we added the interaction between coach intervention and the use of varenicline, participants in the intervention group who used varenicline had a higher odds ratio of smoking cessation (OR = 9.51, CI = 1.78–50.73). As to smoking reduction, the logistic regression revealed that only the coach intervention (OR = 2.87, $95\%$ CI = 1.06–7.80) significantly predicted successful smoking reduction (Nagelkerke R2 = 0.093). When we added the interaction between coach intervention and the use of varenicline, participants in the intervention group who used varenicline had a significantly higher odds ratio of smoking reduction (OR = 6.86, CI = 1.35–34.79). ## 4. Discussion Although this study was suspended from May 2021 due to COVID-19 restrictions for epidemic prevention and drug recall in Taiwan, which impeded us from recruiting the expected number of subjects and caused a smaller sample size, this study still found that the health coaching intervention seemed to enhance the effectiveness of smoking cessation in diabetic patients, especially those who participated in the pharmacotherapy plan. Patients in the intervention group showed a significant smoking reduction, and, even if they did not participate in the pharmacotherapy plan, health coaching seemed to help the patients to reduce smoking effectively; this seems to be consistent with the current rather limited study evidence [22]. In the multivariate analysis, it appears that coaching intervention did increase the effectiveness of using varenicline for smoking cessation. Although the $95\%$ CI was larger due to the small sample size, it seems that exploring the use of both coaching to assist in smoking cessation and varenicline is an interesting research topic. In the past, although there were no studies on the use of health coaches to enhance varenicline or nicotine replacement therapy (NRT), some studies tried to use MI for testing [25,26], and it seems that MI can indeed be effectively paired with NRT or varenicline treatment. However, these studies are also limited by the number of samples or the study design, as in this study, which makes it difficult to directly compare research evidence. In addition to the research design, the nausea caused by the side-effects of varenicline may also be the reason why patients are unwilling to participate in smoking cessation or stop using drugs [25,27], and this was one of the main reasons why the patients in this study stopped taking medication and failed to quit smoking. In addition, as the subjects in this study were limited to patients with type 2 diabetes and smoking, it not only confined this study to a smaller population, but also made it unsuitable for direct comparison with most smoking cessation health coaching studies, because most of these studies are currently aimed at patients with COPD [28,29]. However, we still believe that it is indeed worthwhile to further investigate the effectiveness of using health coaching intervention to enhance the efficacy of NRT or varenicline treatment with larger trials and a more diverse patient population in the future. The smoking cessation rate in the intervention group was about $48.6\%$ in this study, which is about twice the reported smoking cessation rate of smoking cessation services in Taiwan ($27.1\%$ in 2020) [18]; furthermore, in total, $52.9\%$ of the participants reduced their daily number of cigarettes, more than $50\%$ from the baseline. The smoking cessation rates seemed to vary considerably between different studies, and, because of the high heterogeneity of the studies, direct comparisons are not feasible [7,8]. Until now, there are still no conclusions as to which elements of behavioral counseling can effectively increase the smoking cessation rate. For example, one study reported that MI’s decision-making balance method strengthens smokers’ perception of the benefits of smoking [30], while another study suggested that MI may be more suitable for patients with higher education levels, while those with a lower education level are more suitable for five Rs health education [31]. Considering that the five Rs already represent Taiwan’s basic element of smoking cessation education, it was impossible for us to compare the effectiveness of the five Rs with MI or health coaching, such as this study. In addition, as MI and health coaching are not exactly the same methods, we can only conservatively assume that health coaching might be an effective method of smoking cessation. Therefore, more high-quality research is still needed. There are some hospitals in Taiwan trying to provide MI element health education or conduct related studies [32]; however, until now, MI has not been officially included in Taiwan’s smoking cessation professional training. In addition, since currently Taiwan does not have rigorously certified and supervised MI trainers, and even the use of the five Rs strategy lacks fidelity checks, we think this may indeed affect the effectiveness of smoking cessation in Taiwan. In fact, the fidelity of such behavioral counseling will greatly affect the effectiveness of smoking cessation [33], and this has still been the main limitation of most MI and health coaching studies so far. In contrast, due to the fact that our health coach paid special attention to patient-centered communication methods, adopted positive psychology techniques, and applied his rigorous behavioral coaching training to use MI and other skills proficiently, it allowed our patients to develop deep trust in the medical–patient relationship that should be established by physicians to a certain extent. Indeed, the degree of mutual trust between doctors and patients and the communication skills of medical staff seem to be crucial factors contributing to smoking cessation [34]. This is relatively difficult in the medical environment of Taiwan, because Taiwan’s medical system usually allots less than 5 min of a physician’s time to each patient during a consultation, which is completely insufficient for adequate, in-depth communication and discussion [35]. In particular, application of positive psychology communication skills seems to significantly improve a patient’s intrinsic motivation [36]; however, these skills have not been emphasized enough. Our study did not find significant impacts of demographic characteristics, smoking experience, or smoking cessation experience on smoking cessation rates, but we found that the number of daily cigarettes did significantly affect the success rate of smoking cessation, which aligns with some study evidence [37]. If there is a lower amount of smoking and a lower degree of addiction, it seems that it is indeed easier to quit smoking successfully. Race may also be an important factor affecting the success rate of smoking cessation [38], but this is not applicable in Taiwan. Different types of jobs and specific industries may also be contributing factors influencing smoking rates, but it is also relatively hard to find significant differences with a small sample size. One study conducted a long time ago saw a significantly higher smoking rate among workers of material-moving occupations, construction laborers, and vehicle mechanics and repairers [39]. In this study, our health educators and the coach also found such a tendency. The reason why many participants were unable to completely quit smoking seemed to be because almost everyone in their workplace smoked, which made it difficult for them to resist the temptation of cigarettes, and the pressure of rapport or human relationship also made it more difficult to quit smoking. Such effects and the addictive nature of nicotine may make it difficult for pure behavioral counseling to work. When these effects are reduced due to restrictions caused by the COVID-19 pandemic, the smoking cessation rate would naturally increase [40]. We decided not to continue the original research plan in view of the fact that we could not effectively judge the impact of the COVID restrictions on the smoking cessation rate. In summary, determining how to help patients reduce the impact of the workplace environment and human pressure on smoking cessation is still a very important research topic in the future. On the basis of our findings, health coaching intervention in smoking cessation seemed to have significantly higher quit rates when combined with a pharmacotherapy plan. Given that smoking cessation is a complex behavior involving lifestyle changes, addiction symptoms, and side-effects, we do not consider health coaching to be ineffective per se, but rather an effective element of a smoking cessation program that can bolster an original medication program. Hence, the study proposes some suggestions for future studies and our medical system. Firstly, more full-time professional behavior coaching services need to be introduced in Taiwan. The current medical system greatly limits the time and effectiveness of communication between doctors and patients, and there is also a lack of adequate smoking cessation behavior counselors with rigorous training and regular fidelity supervision. Secondly, more studies on health coaching with high-quality evidence are needed. Although this study seemed to have acceptable results, the number of samples did not meet the recruiting expectations, and the research was interrupted by the restrictions which could have affected the effectiveness of the evidence in this study. Although this study was not completed as scheduled, it still had some strengths. Firstly, the original randomized controlled design allowed this study to retain a certain degree of evidence, even if it was affected by some problems as described above, and smoking cessation studies specifically for diabetic patients are also quite rare. Secondly, the use of rigorously trained and certified coaches to perform the smoking cessation counseling was also one of the strengths in this research, as it may have greatly improved the quality of coaching. The training and practice fidelity of the coaches remain important factors influencing research in such behavioral interventions [41,42]. The small sample size was the main limitation of this study. The main reason might be that this study targeted patients with type 2 diabetes, whereby only about $10\%$ of T2DM patients smoked, which limited the sample characteristics of this study. Secondly, another important limitation is that biochemical tests such as carbon monoxide measurements or urine cotinine tests were not used; instead, only FTND scores were used considering COVID-19 prevention during the pandemic, which might have overestimated the effect of smoking cessation. The reason for this design is mainly because that the current referral standard for smoking cessation plans in Taiwan is based on FTND scores; however, given that some studies have found that tests such as CO testing may increase the motivation of patients to quit smoking, this may be one of the standardized procedures that the Taiwan HPA can incorporate into routine smoking cessation plans in the future [22]. It is suggested that similar studies with larger sample sizes, longer follow-up times, and more objective indicators of smoking cessation should be carried out to confirm the validity of the study’s findings in the post-COVID-19 era. ## 5. Conclusions The study found that health coaching may improve the effectiveness of tobacco control project in Taiwan with a reduction in the number of cigarettes smoked and increase the cessation rate of type 2 diabetic patients who participate in the pharmacotherapy plan. More prospective health coaching studies related to smoking cessation are needed before implementing health coaching services into smoking cessation plans in Taiwan. ## References 1. Sarwar N., Gao P., Seshasai S.R., Gobin R., Kaptoge S., Di Angelantonio E., Ingelsson E., Lawlor D.A., Selvin E., Stampfer M.. Diabetes mellitus, fasting blood glucose concentration, and risk of vascular disease: A collaborative meta-analysis of 102 prospective studies. Lancet (2010.0) 375 2215-2222. PMID: 20609967 2. The Prevalance of High Blood Presure, Blood Suger and Blood Lipid in Taiwan. (2021.0) 33 3. National Health Insurance Expenditure Statistics Report. (2018.0) 4. Bergman B.C., Perreault L., Hunerdosse D., Kerege A., Playdon M., Samek A.M., Eckel R.H.. Novel and reversible mechanisms of smoking-induced insulin resistance in humans. Diabetes (2012.0) 61 3156-3166. DOI: 10.2337/db12-0418 5. Bajaj M.. Nicotine and Insulin Resistance: When the Smoke Clears. Diabetes (2012.0) 61 3078-3080. DOI: 10.2337/db12-1100 6. Taiwan Tobacco Control Annual Report. (2020.0) 7. Hettema J.E., Hendricks P.S.. Motivational interviewing for smoking cessation: A meta-analytic review. J. Consult. Clin. Psychol. (2010.0) 78 868-884. DOI: 10.1037/a0021498 8. Heckman C.J., Egleston B.L., Hofmann M.T.. Efficacy of motivational interviewing for smoking cessation: A systematic review and meta-analysis. Tob. Control (2010.0) 19 410-416. DOI: 10.1136/tc.2009.033175 9. Lancaster T., Stead L.F.. Individual behavioural counselling for smoking cessation. Cochrane Database Syst. Rev. (2017.0) 3 CD001292. DOI: 10.1002/14651858.CD001292.pub3 10. Alshahrani A.. The Effects of Smoking Cessation on Diabetes Mellitus Patients. Curr. Diabetes Rev. (2020.0) 16 137-142. DOI: 10.2174/1573399815666190729111041 11. Lindson N., Thompson T.P., Ferrey A., Lambert J.D., Aveyard P.. Motivational interviewing for smoking cessation. Cochrane Database Syst. Rev. (2019.0) 7 CD006936. DOI: 10.1002/14651858.CD006936.pub4 12. Wolever R.Q., Simmons L.A., Sforzo G.A., Dill D., Kaye M., Bechard E.M., Southard M.E., Kennedy M., Vosloo J., Yang N.. A Systematic Review of the Literature on Health and Wellness Coaching: Defining a Key Behavioral intervention in Healthcare. Glob. Adv. Health Med. (2013.0) 2 38-57. DOI: 10.7453/gahmj.2013.042 13. What is a Health Coach? 10 September 14. Code of Ethics—Key Definition of Coaching 15. Schwellnus H., King G., Thompson L.. Client-centred coaching in the paediatric health professions: A critical scoping review. Disabil. Rehabil. (2015.0) 37 1305-1315. DOI: 10.3109/09638288.2014.962105 16. Singh H., Kennedy G.A., Stupans I.. Does the Modality Used in Health Coaching Matter? A Systematic Review of Health Coaching Outcomes. Patient Prefer. Adherence (2020.0) 14 1477-1492. DOI: 10.2147/PPA.S265958 17. Food and Drug Administration Taiwan Varenicline Recall Notice 18. Multiple Smoking-cessation Counseling Service 19. 19.Toolkit for Delivering the 5A’s and 5R’s Brief Tobacco Interventions to TB Patients in Primary CareWHOGeneva, Switzerland2014. Toolkit for Delivering the 5A’s and 5R’s Brief Tobacco Interventions to TB Patients in Primary Care (2014.0) 20. Smoking Cessation Guidline. (2021.0) 1-60 21. Guo J.L., Liao J.Y., Chang L.C., Wu H.L., Huang C.M.. The effectiveness of an integrated multicomponent program for adolescent smoking cessation in Taiwan. Addict. Behav. (2014.0) 39 1491-1499. DOI: 10.1016/j.addbeh.2014.05.009 22. Marler J.D., Fujii C.A., Utley D.S., Tesfamariam L.J., Galanko J.A., Patrick H.. Initial Assessment of a Comprehensive Digital Smoking Cessation Program That Incorporates a Mobile App, Breath Sensor, and Coaching: Cohort Study. JMIR mHealth uHealt (2019.0) 7 e12609. DOI: 10.2196/12609 23. Pérez-Ríos M., Santiago-Pérez M.I., Alonso B., Malvar A., Hervada X., de Leon J.. Fagerstrom test for nicotine dependence vs. heavy smoking index in a general population survey. BMC Public Health (2009.0) 9. DOI: 10.1186/1471-2458-9-493 24. Lam T.H., Cheung Y.T.D., Leung D.Y.P., Abdullah A.S.M., Chan S.S.C.. Effectiveness of smoking reduction intervention for hardcore smokers. Tob. Induc. Dis. (2015.0) 13 9. DOI: 10.1186/s12971-015-0034-y 25. Lim G., Park I., Park S., Song S., Kim H., Kim S.. Effectiveness of smoking cessation using motivational interviewing in patients consulting a pulmonologist. Tuberc. Respir. Dis. (2014.0) 76 276-283. DOI: 10.4046/trd.2014.76.6.276 26. Skelton E., Lum A., Cooper L.E., Barnett E., Smith J., Everson A., Machart J., Baker A.L., Halpin S., Nielssen O.. Addressing smoking in sheltered homelessness with intensive smoking treatment (ASSIST project): A pilot feasibility study of varenicline, combination nicotine replacement therapy and motivational interviewing. Addict. Behav. (2022.0) 124 107074. DOI: 10.1016/j.addbeh.2021.107074 27. Aubin H.-J., Bobak A., Britton J.R., Oncken C., Billing C.B., Gong J., Williams K.E., Reeves K.R.. Varenicline versus transdermal nicotine patch for smoking cessation: Results from a randomised open-label trial. Thorax (2008.0) 63 717-724. DOI: 10.1136/thx.2007.090647 28. Rehman H., Karpman C., Vickers Douglas K., Benzo R.P.. Effect of a Motivational Interviewing-Based Health Coaching on Quality of Life in Subjects With COPD. Respir. Care (2017.0) 62 1043-1048. DOI: 10.4187/respcare.04984 29. Jolly K., Sidhu M.S., Hewitt C.A., Coventry P.A., Daley A., Jordan R., Heneghan C., Singh S., Ives N., Adab P.. Self management of patients with mild COPD in primary care: Randomised controlled trial. BMJ (2018.0) 361 k2241. DOI: 10.1136/bmj.k2241 30. Krigel S.W., Grobe J.E., Goggin K., Harris K.J., Moreno J.L., Catley D.. Motivational interviewing and the decisional balance procedure for cessation induction in smokers not intending to quit. Addict. Behav. (2017.0) 64 171-178. DOI: 10.1016/j.addbeh.2016.08.036 31. Catley D., Goggin K., Harris K.J., Richter K.P., Williams K., Patten C., Resnicow K., Ellerbeck E.F., Bradley-Ewing A., Lee H.S.. A Randomized Trial of Motivational Interviewing: Cessation Induction Among Smokers With Low Desire to Quit. Am. J. Prev. Med. (2016.0) 50 573-583. DOI: 10.1016/j.amepre.2015.10.013 32. Chen Y.H., Wang P.C., Ko Y.L., Wang H.L.. Effects of Motivational Interview and Mobile Social Network Support on Smoking Cessation in Male Patients With Coronary Heart Disease. J. Nurs. Res. (2021.0) 68 53-64 33. Efraimsson E.O., Fossum B., Ehrenberg A., Larsson K., Klang B.. Use of motivational interviewing in smoking cessation at nurse-led chronic obstructive pulmonary disease clinics. J. Adv. Nurs. (2012.0) 68 767-782. DOI: 10.1111/j.1365-2648.2011.05766.x 34. Benton J.Z., Lodh A., Watson A.M., Tingen M.S., Terris M.K., Wallis C.J.D., Klaassen Z.. The association between physician trust and smoking cessation: Implications for motivational interviewing. Prev. Med. (2020.0) 135 106075. DOI: 10.1016/j.ypmed.2020.106075 35. Wu T.-Y., Majeed A., Kuo K.N.. An overview of the healthcare system in Taiwan. Lond. J. Prim. Care (2010.0) 3 115-119. DOI: 10.1080/17571472.2010.11493315 36. Lee E.J.. The Effect of Positive Group Psychotherapy and Motivational Interviewing on Smoking Cessation: A Qualitative Descriptive Study. J. Addict. Nurs. (2017.0) 28 88-95. DOI: 10.1097/JAN.0000000000000169 37. Marler J.D., Fujii C.A., Galanko J.A., Balbierz D.J., Utley D.S.. Durability of Abstinence After Completing a Comprehensive Digital Smoking Cessation Program Incorporating a Mobile App, Breath Sensor, and Coaching: Cohort Study. J. Med. Internet Res. (2021.0) 23 e25578. DOI: 10.2196/25578 38. Grobe J.E., Goggin K., Harris K.J., Richter K.P., Resnicow K., Catley D.. Race moderates the effects of Motivational Interviewing on smoking cessation induction. Patient Educ. Couns. (2020.0) 103 350-358. DOI: 10.1016/j.pec.2019.08.023 39. Bang K.M., Kim J.H.. Prevalence of cigarette smoking by occupation and industry in the United States. Am. J. Ind. Med. (2001.0) 40 233-239. DOI: 10.1002/ajim.1094 40. Jackson S.E., Beard E., Angus C., Field M., Brown J.. Moderators of changes in smoking, drinking and quitting behaviour associated with the first COVID-19 lockdown in England. Addiction (2022.0) 117 772-783. DOI: 10.1111/add.15656 41. Thepwongsa I., Muthukumar R., Kessomboon P.. Motivational interviewing by general practitioners for Type 2 diabetes patients: A systematic review. Fam. Pract. (2017.0) 34 376-383. DOI: 10.1093/fampra/cmx045 42. Ekong G., Kavookjian J.. Motivational interviewing and outcomes in adults with type 2 diabetes: A systematic review. Patient Educ. Couns. (2016.0) 99 944-952. DOI: 10.1016/j.pec.2015.11.022
--- title: Retinal Microvasculature and Neural Changes and Dietary Patterns in an Older Population in Southern Italy authors: - Rossella Tatoli - Luisa Lampignano - Rossella Donghia - Alfredo Niro - Fabio Castellana - Ilaria Bortone - Roberta Zupo - Sarah Tirelli - Madia Lozupone - Francesco Panza - Giovanni Alessio - Francesco Boscia - Giancarlo Sborgia - Rodolfo Sardone journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049576 doi: 10.3390/ijerph20065108 license: CC BY 4.0 --- # Retinal Microvasculature and Neural Changes and Dietary Patterns in an Older Population in Southern Italy ## Abstract Background: Like other parts of the body, the retina and its neurovascular system are also affected by age-related changes. The rising age of populations worldwide makes it important to study the pathologies related to age and their potential risk factors, such as diet and eating habits. The aim of this study was to investigate the predictive power of food groups versus retinal features among noninstitutionalized older adults from Southern Italy using a machine learning approach. Methods: We recruited 530 subjects, with a mean age of 74 years, who were drawn from the large population of the Salus in Apulia Study. In the present cross-sectional study, eating habits were assessed with a validated food frequency questionnaire. For the visual assessment, a complete ophthalmic examination and optical coherence tomography-angiography analyses were performed. Results: The analyses identified 13 out of the 28 food groups as predictors of all our retinal variables: grains, legumes, olives-vegetable oil, fruiting vegetables, other vegetables, fruits, sweets, fish, dairy, low-fat dairy, red meat, white meat, and processed meat. Conclusions: Eating habits and food consumption may be important risk factors for age-related retinal changes. A diet that provides the optimal intake of specific nutrients with antioxidant and anti-inflammatory powers, including carotenoids and omega-3 fatty acids, could have beneficial effects. ## 1. Introduction In our aging society, age-related diseases are an issue of great public interest. Age is a real risk factor for several chronic inflammatory diseases, altering the structure and/or function of organs and tissues [1]. Worldwide, the increasing number of subjects over 60 years of age has led to progressively growing interest in age-related clinical disorders and diseases. It is now essential to investigate their principal modifiable risk factors to identify new and possible strategies to prevent or regress retinal damage. The retina and its cellular and microvascular components are among the districts affected by these age-related changes that can impair their function, leading to gradual or sudden visual loss [2]. The retinal tissue plays a specific role in processing visual information. It has a complex structure composed of pigment epithelium, photoreceptor cells (rods and cones), and more than 60 distinct types of neuronal cells and their fibers [2]. The pigment epithelium is rich in melanin cells, highly pigmented, and is called retinal pigment epithelium (RPE). This type of cell plays numerous crucial roles in the health and normal functioning of the retina, including the delivery of nutrients to the photoreceptor cells and the transportation of ions, metabolites, and fluids to the blood-retinal barrier. At the center of the retina is the macula, a highly pigmented area containing the central fovea, affecting the clarity of vision [3]. The correct functioning of the RPE, the macula, and the central fovea are significant for the health of the retina, and any degeneration can have serious consequences on the visual process. Aging physiologically determines some cellular changes, such as reactive oxygen species (ROS) production, oxidative stress, metabolic dysfunction, and inflammatory responses that can potentially disrupt tissue homeostasis [4,5,6]. This leads to the activation of an adaptation mechanism known as cellular senescence, a cellular response induced by different types of stress that accompanies normal aging [5]. Disproportionate changes induced by cellular senescence, or a persistent accumulation of senescent cells due to chronic stress, exacerbate the adverse effects of aging and can lead to some clinical illnesses and different chronic disorders, including age-related macular degeneration (AMD) [7,8,9]. AMD is a complex medical condition that damages a part of the retina and the macula and compromises the central vision and the ability to see fine details. It is a socially debilitating condition that reduces the quality of life because it leads to loss of balance, mobility and independence issues, depression, and social isolation [10]. AMD is an age-related disease that is currently the first cause of visual impairment and central blindness in the Western population over 65 years of age. Globally, it affects 30–50 million individuals and is expected to increase ten-fold by 2040 [11]. In Italy, AMD affects one million people, and it is estimated that this number will increase with further population aging [12]. The National Eye Institute sponsored two clinical trials, The Age-Related Eye Disease Study (AREDS) and AREDS2, to learn more about AMD, its natural history, and risk factors [13]. The characteristic “drusen” of AMD is caused by hyaline deposits accumulating between the RPE and Bruch’s membrane. The size of the drusen is different in the various disease stages. In the early stages, a small drusen (63 to 124 μm in diameter) is present with hyper/hypopigmentations of the retinal epithelium [14]. In this stage, mild symptoms may or may not be present, but if they are, they will include blurred vision and impaired dark adaptation [15]. In the advanced forms, the drusen size increases (by 125 μm in diameter), and moderate to severe visual loss is present [13,14]. Late AMD can be distinguished as the nonexudative dry AMD (Geographic Atrophy, GA) and the exudative wet form (Neovascular AMD) [16]. Moreover, findings in the retina provide a noninvasive way to investigate the systemic health of the human body. Particularly, subjects with an increased risk of liver fibrosis [17] or hypertension [18] had thinner neuroretinal layers, according to high-resolution retinal scan images taken with optical coherence tomography (OCT) technology. OCT-angiography (OCT-A) allows for the study of retinal health and its microvascular network; this innovative version uses OCT technology to obtain a noninvasive depth-resolved visualization of the retinal microvasculature [19,20]. The main risk factors related to the development and progression of age-related retinal changes include gender, race, genetics, environmental, lifestyle, and dietary factors. For example, the Caucasian female population over the age of 65 is at greater risk [21,22]; environmental and lifestyle factors include light exposure and smoking [23,24]. Chakravarthy and colleagues reported that among older Europeans, the risk of AMD in smokers was 5 and 2.5 times higher for the dry and neurovascular forms, respectively [25]. Cigarette smoking doubles the risk of developing this condition and causes retinal damage through pro-oxidative and proinflammatory processes [25]. In these processes, a clear role could be played by inflammation and oxidative stress [26]. In fact, another risk factor is a diet with a poor content of vitamins A, C, and E and zinc, lutein, zeaxanthin, and omega-3 fatty acids such as DHA [27]. Currently, no curative treatment has been identified for AMD. Antivascular-endothelial growth factor (VEGF) drugs are used to treat the wet form, while no therapy is available to slow the progression of dry AMD. In the last decades, several studies have been focused on researching preventive measures and strategies to slow the degeneration related to this condition. AREDS 1 and 2 mainly evaluated the role of some nutrients and nutritional supplementation in preventing and reversing these eye diseases. A diet based on vegetables, fruits, unrefined products, fish, and extra virgin olive oil could be a valuable tool due to its protective properties against various noncommunicable diseases [28]. Studies demonstrated how adopting proper eating habits also reduces the risk of running into the vascular complications of diabetes, including diabetic retinopathy (DR) [29]. Oxidative stress and inflammation have also been shown to play a key role in DR pathophysiology [30] and antioxidants in its prevention and treatment. The therapeutic role of polyphenols and polyunsaturated fatty acids (PUFA) and the preventive role of vitamin C, vitamin E, lutein, and zeaxanthin have been hypothesized [31,32,33]. More attention should be paid to the dietary and nutritional aspects regarding the prevention and/or treatment of age-related visual degeneration. The present study aimed to investigate the relationship between retinal features and eating habits among noninstitutionalized older adults from Southern Italy using a machine learning approach. This is a new approach that has not been widely used in previous studies on the retina and its components [34]. ## 2.1. Study Design and Population This cross-sectional population-based study involved 530 subjects aged over 64. This was a subsample drawn from the “Salus in Apulia Study”, a public health initiative promoted by the Italian Ministry of Health and Apulia Regional Government and conducted at IRCCS “S. de Bellis” Research Hospital in Castellana Grotte, Southern Italy. In this study, 4537 individuals were enrolled from 2014 to 2019 in Castellana Grotte. This selection of study participants allowed us to utilize past individual data generated from other investigations. The sample is representative of the entire population of older people (age > 65 years) from Castellana Grotte in 2014, as described elsewhere [35]. In this study, we analyzed the data from those subjects who had both undergone nutritional and ophthalmological assessments. The IRB of the head institution, the National Institute of Gastroenterology and Research Hospital “S. de Bellis” in Castellana Grotte, Italy, approved the study. The study was conducted according to the Helsinki Declaration of 1975 and adhered to the “Standards for Reporting Diagnostic Accuracy Studies” (STARD) guidelines (http://www.stard-statement.org/ accessed on 22 December 2022). The manuscript was organized according to the “Strengthening the Reporting of Observational Studies in Epidemiology-Nutritional Epidemiology” (STROBE-nut) guidelines (https://www.strobe-nut.org/ accessed on 22 December 2022). For the present study, the participants were subject to nutritional and visual assessments. All of them signed an informed consent form before examination. ## 2.2. Clinical and Lifestyle Assessment The anthropometric parameters of height and weight were measured by a Seca 220 stadiometer and a Seca 711 scale. Body mass index (BMI) was calculated as weight measured in kg and height indicated by m2. Smoking habit was evaluated by asking the question: “Are you currently a smoker?”. ## 2.3. Dietary Assessment Diet and eating habits were evaluated with a validated food frequency questionnaire (FFQ) used in previous studies [36]. This selfadministered questionnaire was checked by a registered dietitian during an interview at the study center. It investigated the frequency intake of a predefined portion over the last year but not the differences in portion sizes. Each portion weight was expressed in grams. Originally, the FFQ was structured into 11 sections, representing foods with similar characteristics: grains, meat, fish, milk and dairy products, vegetables, legumes, fruits, miscellaneous foods, water and alcoholic beverages, olive oil and other edible fats, coffee/sugar, and salt. Then it was validated against the dietary records and adapted to our population [37]. In the final FFQ version, 85 food items were identified as typical local foods that reflect the local diet (Figure S1). These food items, together with some questions about the use of edible fats, have been regrouped into 28 food groups for statistical analyses [35]. The food group “edible cooking fats” could not be quantified and was not used in the present study. ## 2.4. Visual Assessment Each participant underwent a complete ophthalmic examination (described in detail elsewhere) [17]. Briefly, the examination included best-corrected visual acuity (BCVA) measurement, slit-lamp biomicroscopy, intraocular pressure (IOP) measurement, and funduscopy. Then, we used the Optovue RTVue XR 100 AVANTI, made by Optovue, Inc., to perform OCT and OCT-A. After detecting and segmenting several retinal layers using the AngioVue module of the Optovue RTVue AVANTI program, OCT-A analyses of the retinal vasculature (version 2015.100.0.35, Optovue, Inc., Fremont, CA, USA) were performed. Both the Angio Disc mode (4.54.5 mm2) and the Angio Retina mode (33 mm2) were used. The vessel density (VD, %) was defined as the proportion of vessel area with blood flow over the total area automatically measured by the OCT software. The OCT angiograms centered on the fovea automatically defined the superficial and deep vascular plexus. The VD at each plexus of the RPE (the superficial VD (SVD), and deep VD (DVD)) were determined for the whole 3 mm circle area centered on the fovea (whole retina) (Figure S2). The thickness (µm) of the ganglion cell complex (GCC), consisting of the thickness of the retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), and inner plexiform layer (IPL) (Figure S2) at the macular area, and, separately, of the RNFL, was measured at the same time using the same OCT. The device measures GCC and RNFL thickness within an automatically rendered 7 mm2 area, centered 1 mm temporally to the fovea [18]. Each retinal feature shown in Figure S2 is explained in Table S1. Ocular exclusion criteria for all study participants included an IOP > 22 mmHg, a history of glaucoma, optic neuropathies, demyelinating disorders, retinal diseases, including macular degeneration, diabetic or hypertensive retinopathy, epiretinal membrane, retinal detachment, an obvious media opacity reducing visual acuity below 1 LogMar and interfering with the OCT and OCT-A analysis, a refractive error of 6 diopters or more, an intraocular surgery performed in the previous 6 months, or ocular trauma. ## 2.5. Statistical Analysis Subject characteristics are reported as Median and Interquartile Range (IQR) for continuous variables and as frequencies and percentages (%) for categorical variables. To test the nonnormal distribution of variables Kolmogorov–Smirnov test for equality was used. To analyze the difference between two subgroups of age classes, the median value calculated on our cohort was chosen to cut the distribution of age variables perfectly in two parts. To select the predictors of the OCT variable, random forest (RF) was used. RF was computed by an ensemble of binary decision trees, which could be used to select the most important variables linked with the outcomes. Variable predictiveness could be assessed using variable importance measures for both single and grouped variables [38]. The random forest (RF) method is a machine-supervised learning algorithm based on a randomized decisional tree for ranking the prediction power of a set of variables regarding the outcome of interest. The “forest” it builds is an ensemble of decision trees, usually trained with the “bagging” method. The general idea of the bagging method is that a combination of learning models increases the overall result. We examined which variables best predict variance in the intervention effects by ranking the covariates in order of importance. The ranking is calculated as the sum of how often a given covariate is split at each depth of the forest. The sum is weighted so that early splits (low forest depth) are more important than late splits. Variables are considered “more important” if the variable is more frequently used for the first splits across all decision trees that are grown in the random forest. The parameter used for ranking was the importance score variable, calculated by adding up the improvement in the objective function given by the splitting criterion over all internal nodes of a tree and across all trees in the forest (separately for each predictor variable). The importance score variable was normalized by dividing all scores over the maximum score ($100\%$). Variables with high importance were the drivers of the outcome, and their score values have a significant impact on the outcome [36]. We used another statistical methodology to evaluate variable importance; in fact, predictor importance was estimated based on the minimal depth of the maximal subtree. The “Depth” was how many nodes down in the tree, starting the numbering at 0 for the root node, while “Minimal depth” was the minimal depth value for the first instance of a given splitting variable. “ Mean minimal depth” was the minima depth for a variable averaged across all trees in the forest. If a predictor was influential in a prediction, then the variable is likely to occur nearer to the root rather than the leaf nodes [39]. Depth is indicated by a vertical bar with the mean value. The smaller the mean minimal depth, the more important the variable and the higher up the y-axis the variable will be. The color gradient reveals the min and max minimal depth for each variable. The range of the x-axis is from zero to the maximum number of trees for the feature. We randomly split the data into training and testing subgroups to predict visual outcomes. The training data included $75\%$ of the sample ($$n = 397$$), while the remaining data (the test data) accounted for $25\%$ ($$n = 133$$) and were used to test the model and minimize the heterogeneity of the obtained subsamples for a continuous outcome. All statistical computations were made using StataCorp. 2021. Stata Statistical Software: Release 17. College Station, TX, USA: StataCorp LLC. And RStudio software (“Prairie Trillium” Release). ## 3. Results The present study was conducted on a sample of 530 subjects, with a median (IQR) age of 72.00 (68.00–78.00) years, drawn from the “Salus in Apulia Study” population. The female sex was slightly predominant, accounting for $55.7\%$. Only $7.2\%$ of subjects studied were smokers, while excess weight, expressed by median (IQR) BMI of 28.81 (25.98–32.04), was more common. Table 1 shows the sociodemographic and OCT variables of the sample. Table 2 reports the average consumption of the food groups. There are differences in food consumption by gender and age groups. As regards gender, the difference is statistically significant for dairy (p: 0.005), eggs (p: 0.05), white meat (p: 0.009), red meat (p: <0.0001), processed meat (p: <0.0001), fish (p: 0.04), seafood/shellfish (p: 0.001), root vegetables (p: 0.005), grains (p: 0.02), high-calorie drinks (p: 0.04), ready to eat dish (p: 0.02), wine (p: <0.0001), beer wine (p: <0.0001), and spirits (p: <0.0001). As regards the age classes, the difference is statistically significant for processed meat (p: 0.009), nuts (p: 0.001), olives and vegetable oil (p: 0.006), high-calorie drinks (p: 0.006), ready-to-eat dishes (p: <0.0001), coffee (p^: <0.0001), beer (p: 0.003), and spirits (p: 0.05). ## 3.1. Random Forest Food Group and Retinal Nerve Fiber Layer (RNFL) These analyses identified 15 of the 28 food groups as predictors of RNFL: grains (minimal depth: 3.96), sweets (minimal depth: 4.32), processed meat (minimal depth: 4.84), legumes (minimal depth: 5.01), fish (minimal depth: 5.27), olives-vegetable oil (minimal depth: 5.34), fruiting vegetables (minimal depth: 5.58), white meat (minimal depth: 5.64), leafy vegetables (minimal depth: 5.78), dairy (minimal depth: 5.83), fruits (minimal depth: 5.93), low-fat dairy (minimal depth: 6.06), red meat (minimal depth: 6.13), other vegetables (minimal depth: 6.30), and high-calorie drinks (minimal depth: 6.61) (Figure 1). ## 3.2. Random Forest Food Group and Ganglion Cell Layer (GCL) Thickness These analyses identified 15 of the 28 food groups as predictors of GCC thickness: grains (minimal depth: 4.07), legumes (minimal depth: 4.27), red meat (minimal depth: 4.39), low-fat dairy (minimal depth: 4.89), fruiting vegetables (minimal depth: 5.00), white meat (minimal depth: 5.21), processed meat (minimal depth: 5.46), fish (minimal depth: 5.51), fruits (minimal depth: 5.72), other vegetables (minimal depth: 5.88), sweets (minimal depth: 6.00), eggs (minimal depth: 6.11), seafood-shellfish (minimal depth: 6.24), olives-vegetable oil (minimal depth: 6.31), and dairy (minimal depth: 6.32) (Figure 2). ## 3.3. Random Forest Food Group and Whole Retina Superficial Vessel Density (SVD) These analyses identified 15 of the 28 food groups as predictors of the whole retina SVD: low-fat dairy (minimal depth: 4.16), grains (minimal depth: 4.41), leafy vegetables (minimal depth: 4.82), other vegetables (minimal depth: 4.96), fruits (minimal depth: 4.97), fish (minimal depth: 4.99), white meat (minimal depth: 4.99), dairy (minimal depth: 5.08), sweets (minimal depth: 5.30), olives-vegetable oil (minimal depth: 5.41), fruiting vegetables (minimal depth: 5.53), legumes (minimal depth: 5.63), high-calorie drinks (minimal depth: 5.64), red meat (minimal depth: 5.70)and processed meat (minimal depth: 5.85) (Figure 3). ## 3.4. Random Forest Food Group and Whole Retina Deep Vessel Density (DVD) These analyses identified 15 of the 28 food groups as predictors of the whole retina DVD: red meat (minimal depth: 3.87), grains (minimal depth: 4.56), dairy (minimal depth: 4.57), olives-vegetable oil (minimal depth: 4.90), legumes (minimal depth: 4.91), processed meat (minimal depth: 4.95), fruiting vegetables (minimal depth: 5.18), fruits (minimal depth: 5.25), white meat (minimal depth: 5.29), leafy vegetables (minimal depth: 5.34), sweets (minimal depth: 5.41), fish (minimal depth: 5.52), low-fat dairy (minimal depth: 5.62), seafood-shellfish (minimal depth: 5.77), and other vegetables (minimal depth: 5.80) (Figure 4). Table 3 summarizes the results from the random forest analyses for each variable investigated in the visual assessment. A total of 13 out of the 15 food groups were found to be common among the representative rankings of the four retinal variables selected for this study. ## 4. Discussion The present study, which was conducted on 530 subjects aged over 64 years in Castellana Grotte, sought to identify a dietary pattern that was predictive of the variables studied, emphasizing a link between the consumption of some food groups and specific neurovascular retinal features. The foods with a high prediction power include grains, legumes, fruiting vegetables, other vegetables, fruits, olives-vegetable oil, fish, dairy, low-fat dairy, red meat, processed meat, and sweets. The present findings are in line with previous studies that refer to specific foods that owe their effects to the composition of micro and macronutrients. In fact, growing attention is being paid to the role of nutrition in the health of the retina and visual impairment. Several studies suggested antioxidant and anti-inflammatory effects of some specific foods [40,41]. The AREDS1 and AREDS2 pieces of research focused on specific nutritional intake and dietary supplements as strategies for preventing and slowing down AMD development. AREDS1 studied the impact of some antioxidant nutrients, including vitamin C, vitamin E, beta-carotene, and zinc, while for AREDS2, lutein, zeaxanthin, docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA) were added and beta-carotene was removed [13,42]. The results were encouraging, showing a $25\%$ reduction in the probability of developing advanced forms for subjects at high risk of AMD. There seems to be no benefit in the early stages of the disease. This different food effect on the development of AMD led to a different mechanism behind the two forms being hypothesized. Early AMD may be caused by a parainflammatory response to relatively low levels of tissue stress, while late AMD may be caused by a chronic inflammatory response to local and systemic stress [43]. ## 4.1. Fruits and Vegetables Fruits and vegetables are good sources of micronutrients (folate, potassium, magnesium, vitamins A, C, E, and K) and phytochemicals [44], which are responsible for several health benefits [45,46,47]. In fact, nowadays, increasing the consumption of fruits and vegetables is recommended due to the preventive effects of some phytochemicals against several diseases, including cardiovascular disease (CVD) [48]. In order to benefit from the protective and preventive effect of fruit and vegetables against several chronic diseases, such as CVD and different types of cancer, the World Health Organization (WHO) recommends a minimum consumption of 400 g or five portions of 80 g each of fruits, greens and/or vegetables per day [49]. In addition, fruits and vegetables, in particular leafy green vegetables, are the primary dietary sources of carotenoids. Previous studies in the scientific literature have suggested that the consumption of carotenoids, specifically lutein and zeaxanthin, tends to enhance the health of the retina, protecting against the development of retinal changes [42]. Lutein and zeaxanthin are carotenoids belonging to the family of xanthophylls. The most significant sources include kale (48.0–114.7 μg/g), parsley (64–106.5 μg/g), spinach (59.3–79.0 μg/g), lettuce (10.0–47.8 μg/g), and broccoli (7.1–33.0 μg/g) [50]. The effects of these types of carotenoids on visual performance have been the object of a wide variety of research [3]. Many of these have demonstrated a positive association between the consumption of lutein and zeaxanthin, retina health, and the prevention of some retinal disorders, such as AMD. In fact, the risk of retinal changes is significantly lower in individuals that consume more dietary lutein and zeaxanthin [51,52,53]. These results were not found for the supplementation of these nutrients due to their different bioavailability. The beneficial role is due to the presence of lutein and zeaxanthin in the center of the retina, which is needed to form the macular pigment [54]. It is essential because it reduces the penetration of harmful blue light into the retinal tissues and is responsible for the antioxidant abilities of the retina. ## 4.2. Fish Like lutein and zeaxanthin, EPA and DHA also seem to be required for adequate retinal function [55]. They contribute to the prevention of cell apoptosis and oxidative damage, the development and maintenance of photoreceptor membranes and neurotransmitters, rhodopsin activation, and rod and cone development [56]. DHA is also the major structural component of retinal membranes, so it is essential for the development of the visual system [57]. EPA and DHA are omega-3 essential PUFAs that the human body cannot synthesize autonomously and so must obtain from a diet [58]. Changes in their levels of concentration or inadequate dietary intake can promote some retinal diseases, including AMD. Liu and colleagues reported significantly lower DHA levels in individuals with AMD than in those without [56]. The role of omega-3 fatty acids in prevention and treatment is still not completely clear. However, supplementation of EPA and DHA is known to not produce the same benefits as the dietary omega-3 fatty acids on the retina. Souied and colleagues studied the effects of DHA supplementation in the prevention and delay of the progression of exudative AMD. No significant differences were identified between the DHA and placebo groups [59]. Fish is an excellent source of EPA and DHA. In particular, there are high concentrations in dark-meat fish such as salmon, mackerel, sardines, herring, anchovies, and fish oils. Chong and colleagues showed a statistically significant association between a high intake of fish and a reduction in the development of late and intermediate AMD [60]. In the Mediterranean diet (MedDiet), regular fish consumption can ensure an adequate intake of omega-3 fatty acids. The MedDiet is also characterized by a balance between PUFA omega-6 and omega-3. This is one of the elements that makes the MedDiet the best dietary pattern in AMD risk management. Omega-6 has a proinflammatory function, while omega-3 is anti-inflammatory [55,61,62]. It has been demonstrated that an alteration in the balance of dietary omega 3- and omega 6, with a high food intake of omega 6, increases the risk of developing AMD [63]. The MedDiet has been defined as a longevity determinant thanks to the properties of its components [40,41]. Considerable scientific evidence supports the protective and preventive role of this food pattern against chronic noncommunicable diseases [64]. ## 4.3. Grains Grains are an important source of complex carbohydrates, which are then broken down into glucose by the digestive processes. Carbohydrate-source foods can be classified on the basis of their glycemic index, according to the effect on blood sugar levels over a period of two hours. Pure glucose is assigned a glycemic index (GI) value of 100. High-GI foods (>70) include white bread, potatoes, white rice, cereals, honey, and refined sugar [65]. Low-GI foods (<55) include whole fruit and vegetables, whole wheat bread, pasta, oats, bran, legumes, milk, and yogurt. Previous studies compared the effect of a low-GI diet and a high-GI diet on retinal alteration risk [66,67]. The risk was found to be higher in the high-GI diet group, probably due to inflammatory processes. However, the present study did not distinguish between whole and refined grains. Further studies are, therefore, needed to support this hypothesis. ## 4.4. Olives and Vegetable Olive Oil Several studies have demonstrated the beneficial effects of olive oil on health status [68,69]. In particular, extra virgin olive oil, a key food in the MedDiet, has beneficial effects on blood pressure, glycaemic control in diabetics, endothelial functioning, oxidative stress, and lipid profiles; in addition, it reduces the susceptibility of LDL to oxidation as well as concentrations of inflammatory markers, such as C-reactive protein and interleukin-6 [70]. Its nutritional and healthy value can be attributed to the bioactive components of olive oil, including monounsaturated fatty acids (MUFAs) and PUFAs, tocopherols, and polyphenols [71]. Many epidemiological studies, including randomized controlled trials, show that the intake of olive oil improves cardiovascular health [72]. Therefore, it is possible to assume that these beneficial effects also affect the retinal microvascular system; further studies are needed to confirm this hypothesis. ## 4.5. Dairy The predictive power of milk and dairy towards retinal changes needs to be clarified by further studies. Some scientific studies refer to the proinflammatory effect of dairy, with others to an anti-inflammatory effect [73]. Gopinath and colleagues found an increased odds ratio for developing AMD in the case of a low intake of dairy and calcium [74]. However, these results need to be further investigated. ## 4.6. Red and Processed Meat Many observational studies and meta-analyses demonstrated the association between a high intake of red and/or processed meat and chronic diseases such as obesity, type 2 diabetes, CVD, and a variety of cancers. The high consumption of these foods is also associated with an increased risk of total, cardiovascular, and cancer mortality [75]. In accordance with these results, red and processed meat are predictive factors for retinal changes. Previous studies identified red and processed meat as potential risk factors for retinal diseases, such as AMD [76,77]. The mechanism underlying this hypothesis needs further exploration. Some studies focused on the high-fat content of these foods, in particular, processed meat, such as sausage or salami [78]. Furthermore, the role of other components has been studied, including heme iron, nitrites, nitrous, and advanced glycated end products (AGEs). It seems that heme iron can increase the levels of N-nitrose components, causing damage to the retina [79]. However, AGEs increase oxidative stress and inflammation and alternate normal cellular function [80]. ## 4.7. Strengths and Limitations The strengths of the present study include its large population-based sample size and the generalizability of the results to southern, older Mediterranean populations. Moreover, this study is the first to evaluate the more specific quantitative parameters of retinal vasculature when compared to previous studies, in which the morphologic parameters were evaluated using only direct fundoscopy, which could be influenced by interexaminer variability. Another strength was the identification of a dietary pattern associated with neurovascular retinal features through a strong machine-learning approach, even though it cannot explain any biological reasons for the associations detected. In fact, the dietary pattern associated with neural and vascular retinal features is to be understood as an association ranking of certain food groups with a certain outcome. However, some limitations must also be taken into account. One of these is the nature of the study, which was cross-sectional and did not allow for the clear directionality of an association to be discerned. A further limitation is the FFQ as a dietary assessment method since its memory-based nature and consequent measurement errors make it particularly difficult to analyze small diet–disease associations. ## 5. Conclusions The aging of populations worldwide is leading to an increase in age-related pathological conditions, including retinal changes. These reduce a person’s quality of life, increasing the risk of depression, isolation, and falls. This makes it essential to focus research on new strategies that prevent this condition, slow its evolution, or mitigate its debilitating symptoms. More attention to diet and food intake should be encouraged. Starting from the data in the scientific literature on the protective role of some nutrients against retinal alterations, such as AMD, the present study analyzed the prediction power of a certain dietary pattern on retinal neurovascular features. We found a link between specific retinal variables and some food groups, such as fruiting vegetables, other vegetables, fruits, fish, olives–vegetable oil, which are sources of carotenoids and omega-3 fatty acids that are essential nutrients for the health of the retina. Thanks to its anti-inflammatory and antioxidant characteristics, the MedDiet could be a powerful tool in this sense. More specific studies on the effects of this dietary pattern on the risk of retinal alterations are warranted. ## References 1. Dillin A., Gottschling D.E., Nyström T.. The Good and the Bad of Being Connected: The Integrons of Aging. Curr. Opin. Cell Biol. (2014.0) 26 107-112. DOI: 10.1016/j.ceb.2013.12.003 2. Masland R.H.. The Neuronal Organization of the Retina. Neuron (2012.0) 76 266-280. DOI: 10.1016/j.neuron.2012.10.002 3. Walchuk C., Suh M.. Nutrition and the Aging Retina: A Comprehensive Review of the Relationship between Nutrients and Their Role in Age-Related Macular Degeneration and Retina Disease Prevention. Adv. Food Nutr. Res. (2020.0) 93 293-332. DOI: 10.1016/bs.afnr.2020.04.003 4. Kaarniranta K., Sinha D., Blasiak J., Kauppinen A., Veréb Z., Salminen A., Boulton M.E., Petrovski G.. Autophagy and Heterophagy Dysregulation Leads to Retinal Pigment Epithelium Dysfunction and Development of Age-Related Macular Degeneration. Autophagy (2013.0) 9 973-984. DOI: 10.4161/auto.24546 5. Blasiak J., Piechota M., Pawlowska E., Szatkowska M., Sikora E., Kaarniranta K.. Cellular Senescence in Age-Related Macular Degeneration: Can Autophagy and DNA Damage Response Play a Role?. Oxidative Med. Cell. Longev. (2017.0) 2017 5293258. DOI: 10.1155/2017/5293258 6. Zhang M., Jiang N., Chu Y., Postnikova O., Varghese R., Horvath A., Cheema A.K., Golestaneh N.. Dysregulated Metabolic Pathways in Age-Related Macular Degeneration. Sci. Rep. (2020.0) 10 2464. DOI: 10.1038/s41598-020-59244-4 7. Weng X., Zhang H., Kan M., Ye J., Liu F., Wang T., Deng J., Tan Y., He L., Liu Y.. Leukocyte Telomere Length Is Associated with Advanced Age-Related Macular Degeneration in the Han Chinese Population. Exp. Gerontol. (2015.0) 69 36-40. DOI: 10.1016/j.exger.2015.06.004 8. Hyttinen J.M.T., Kannan R., Felszeghy S., Niittykoski M., Salminen A., Kaarniranta K.. The Regulation of NFE2L2 (NRF2) Signalling and Epithelial-to-Mesenchymal Transition in Age-Related Macular Degeneration Pathology. Int. J. Mol. Sci. (2019.0) 20. DOI: 10.3390/ijms20225800 9. Oubaha M., Miloudi K., Dejda A., Guber V., Mawambo G., Germain M.-A., Bourdel G., Popovic N., Rezende F.A., Kaufman R.J.. Senescence-Associated Secretory Phenotype Contributes to Pathological Angiogenesis in Retinopathy. Sci. Transl. Med. (2016.0) 8 362ra144. DOI: 10.1126/scitranslmed.aaf9440 10. Senra H., Macedo A.F., Nunes N., Balaskas K., Aslam T., Costa E.. Psychological and Psychosocial Interventions for Depression and Anxiety in Patients with Age-Related Macular Degeneration: A Systematic Review. Am. J. Geriatr. Psychiatry (2019.0) 27 755-773. DOI: 10.1016/j.jagp.2019.03.001 11. Wong W.L., Su X., Li X., Cheung C.M.G., Klein R., Cheng C.-Y., Wong T.Y.. Global Prevalence of Age-Related Macular Degeneration and Disease Burden Projection for 2020 and 2040: A Systematic Review and Meta-Analysis. Lancet Glob. Health (2014.0) 2 e106-e116. DOI: 10.1016/S2214-109X(13)70145-1 12. Ministero Della Salute 2020 13. A Randomized, Placebo-Controlled, Clinical Trial of High-Dose Supplementation with Vitamins C and E, Beta Carotene, and Zinc for Age-Related Macular Degeneration and Vision Loss: AREDS Report No. 8. Arch. Ophthal. (2001.0) 119 1417-1436. DOI: 10.1001/archopht.119.10.1417 14. Weikel K.A., Chiu C.-J., Taylor A.. Nutritional Modulation of Age-Related Macular Degeneration. Mol. Asp. Med. (2012.0) 33 318-375. DOI: 10.1016/j.mam.2012.03.005 15. Cheung L.K., Eaton A.. Age-related Macular Degeneration. Pharmacother. J. Hum. (2013.0) 33 838-855. DOI: 10.1002/phar.1264 16. The Age-Related Eye Disease Study System for Classifying Age-Related Macular Degeneration from Stereoscopic Color Fundus Photographs: The Age-Related Eye Disease Study Report Number 6. Am. J. Ophthalmol. (2001.0) 132 668-681. DOI: 10.1016/S0002-9394(01)01218-1 17. Lampignano L., Niro A., Castellana F., Bortone I., Zupo R., Tirelli S., Tatoli R., Griseta C., De Nucci S., Sila A.. Liver Fibrosis and Retinal Features in an Older Mediterranean Population: Results from the Salus in Apulia Study. Front. Neurosci. (2022.0) 16 1048375. DOI: 10.3389/fnins.2022.1048375 18. Niro A., Sborgia G., Lampignano L., Giuliani G., Castellana F., Zupo R., Bortone I., Puzo P., Pascale A., Pastore V.. Association of Neuroretinal Thinning and Microvascular Changes with Hypertension in an Older Population in Southern Italy. J. Clin. Med. Res. (2022.0) 11. DOI: 10.3390/jcm11041098 19. Spaide R.F., Klancnik J.M., Cooney M.J.. Retinal Vascular Layers Imaged by Fluorescein Angiography and Optical Coherence Tomography Angiography. JAMA Ophthalmol. (2015.0) 133 45-50. DOI: 10.1001/jamaophthalmol.2014.3616 20. Kashani A.H., Chen C.-L., Gahm J.K., Zheng F., Richter G.M., Rosenfeld P.J., Shi Y., Wang R.K.. Optical Coherence Tomography Angiography: A Comprehensive Review of Current Methods and Clinical Applications. Prog. Retin. Eye Res. (2017.0) 60 66-100. DOI: 10.1016/j.preteyeres.2017.07.002 21. Coleman H.R., Chan C.-C., Ferris F.L., Chew E.Y.. Age-Related Macular Degeneration. Lancet (2008.0) 372 1835-1845. DOI: 10.1016/S0140-6736(08)61759-6 22. National Eye Institute Age-Related Macular Degeneration: What You Should Know. PsycEXTRA Dataset 23. Margrain T.. Do Blue Light Filters Confer Protection against Age-Related Macular Degeneration?. Prog. Retin. Eye Res. (2004.0) 23 523-531. DOI: 10.1016/j.preteyeres.2004.05.001 24. Vander J.F.. Cigarette Smoking, Fish Consumption, Omega-3 Fatty Acid Intake, and Associations with Age-Related Macular Degeneration: The US Twin Study of Age-Related Macular Degeneration. Yearb. Ophthalmol. (2007.0) 2007 128-129. DOI: 10.1016/S0084-392X(08)70101-1 25. Chakravarthy U., Augood C., Bentham G.C., de Jong P.T.V.M., Rahu M., Seland J., Soubrane G., Tomazzoli L., Topouzis F., Vingerling J.R.. Cigarette Smoking and Age-Related Macular Degeneration in the EUREYE Study. Ophthalmology (2007.0) 114 1157-1163. DOI: 10.1016/j.ophtha.2006.09.022 26. Rinninella E., Mele M.C., Merendino N., Cintoni M., Anselmi G., Caporossi A., Gasbarrini A., Minnella A.M.. The Role of Diet, Micronutrients and the Gut Microbiota in Age-Related Macular Degeneration: New Perspectives from the Gut. Nutrients (2018.0) 10. DOI: 10.3390/nu10111677 27. Coulston A.M., Boushey C.J., Ferruzzi M., Delahanty L.. Nutrition in the Prevention and Treatment of Disease (2017.0) 28. Bosy-Westphal A., Müller M.J.. Diet and Nutrition in the Prevention of Non-Communicable Diseases (NCD). Dtsch. Med. Wochenschr. (2021.0) 146 389-397. DOI: 10.1055/a-1248-5539 29. Cicero A.F.G., Fogacci F., Tocci G., Ventura F., Presta V., Grandi E., Rizzoli E., D’Addato S., Borghi C., Cicero A.F.G.. Awareness of Major Cardiovascular Risk Factors and Its Relationship with Markers of Vascular Aging: Data from the Brisighella Heart Study. Nutr. Metab. Cardiovasc. Dis. (2020.0) 30 907-914. DOI: 10.1016/j.numecd.2020.03.005 30. Bryl A., Mrugacz M., Falkowski M., Zorena K.. The Effect of Diet and Lifestyle on the Course of Diabetic Retinopathy—A Review of the Literature. Nutrients (2022.0) 14. DOI: 10.3390/nu14061252 31. Garcia-Medina J.J., Rubio-Velazquez E., Foulquie-Moreno E., Casaroli-Marano R.P., Pinazo-Duran M.D., Zanon-Moreno V., Del-Rio-Vellosillo M.. Update on the Effects of Antioxidants on Diabetic Retinopathy: In Vitro Experiments, Animal Studies and Clinical Trials. Antioxidants (2020.0) 9. DOI: 10.3390/antiox9060561 32. Brazionis L., Rowley K., Itsiopoulos C., O’Dea K.. Plasma Carotenoids and Diabetic Retinopathy. Br. J. Nutr. (2009.0) 101 270-277. DOI: 10.1017/S0007114508006545 33. Tikhonenko M., Lydic T.A., Opreanu M., Calzi S.L., Bozack S., McSorley K.M., Sochacki A.L., Faber M.S., Hazra S., Duclos S.. N-3 Polyunsaturated Fatty Acids Prevent Diabetic Retinopathy by Inhibition of Retinal Vascular Damage and Enhanced Endothelial Progenitor Cell Reparative Function. PLoS ONE (2013.0) 8. DOI: 10.1371/journal.pone.0055177 34. Grzybowski A.. Artificial Intelligence in Ophthalmology (2021.0) 35. Castellana F., Zupo R., Bortone I., Giannelli G., Donghia R., Lampignano L., Griseta C., De Pergola G., Boeing H., Cisternino A.M.. Traditional Old Dietary Pattern of Castellana Grotte (Apulia) Is Associated with Healthy Outcomes. Nutrients (2020.0) 12. DOI: 10.3390/nu12103097 36. Tatoli R., Lampignano L., Donghia R., Castellana F., Zupo R., Bortone I., De Nucci S., Campanile G., Lofù D., Vimercati L.. Dietary Customs and Social Deprivation in an Aging Population From Southern Italy: A Machine Learning Approach. Front. Nutr. (2022.0) 9 811076. DOI: 10.3389/fnut.2022.811076 37. Leoci C., Centonze S., Guerra V., Cisternino A.M., Misciagna G.. Reliability and Validity of a Semiquantitative Food Frequency Questionnaire. G. Ital. Nutr. Clin. Prev. (1933.0) 2 58-59 38. Donghia R., Guerra V., Pesole L.P., Liso M.. Contribution of Macro-and Micronutrients Intake to Gastrointestinal Cancer Mortality in the ONCONUT Cohort. Classical versus Modern Approaches. Front. Nutr. (2023.0) 10 56. DOI: 10.3389/fnut.2023.1066749 39. Ishwaran H., Kogalur U.B., Gorodeski E.Z., Minn A.J., Lauer M.S.. High-Dimensional Variable Selection for Survival Data. J. Am. Stat. Assoc. (2010.0) 105 205-217. DOI: 10.1198/jasa.2009.tm08622 40. Hess J.M., Stephensen C.B., Kratz M., Bolling B.W.. Exploring the Links between Diet and Inflammation: Dairy Foods as Case Studies. Adv. Nutr. (2021.0) 12 1S-13S. DOI: 10.1093/advances/nmab108 41. dos Santos J.L., dos Santos J.L., de Quadros A.S., Weschenfelder C., Garofallo S.B., Marcadenti A.. Oxidative Stress Biomarkers, Nut-Related Antioxidants, and Cardiovascular Disease. Nutrients (2020.0) 12. DOI: 10.3390/nu12030682 42. Lutein + Zeaxanthin and Omega-3 Fatty Acids for Age-Related Macular Degeneration: The Age-Related Eye Disease Study 2 (AREDS2) Randomized Clinical Trial. JAMA (2013.0) 309 2005-2015. DOI: 10.1001/jama.2013.4997 43. Chen M., Xu H.. Parainflammation, Chronic Inflammation, and Age-Related Macular Degeneration. J. Leukoc. Biol. (2015.0) 98 713-725. DOI: 10.1189/jlb.3RI0615-239R 44. Tosh S.M., Bordenave N.. Emerging Science on Benefits of Whole Grain Oat and Barley and Their Soluble Dietary Fibers for Heart Health, Glycemic Response, and Gut Microbiota. Nutr. Rev. (2020.0) 78 13-20. DOI: 10.1093/nutrit/nuz085 45. Champagne C.M.. Dietary Interventions on Blood Pressure: The Dietary Approaches to Stop Hypertension (DASH) Trials. Nutr. Rev. (2006.0) 64 S53-S56. DOI: 10.1111/j.1753-4887.2006.tb00234.x 46. Eichholzer M., Lüthy J., Gutzwiller F., Stähelin H.B.. The Role of Folate, Antioxidant Vitamins and Other Constituents in Fruit and Vegetables in the Prevention of Cardiovascular Disease: The Epidemiological Evidence. Int. J. Vitam. Nutr. Res. (2001.0) 71 5-17. DOI: 10.1024/0300-9831.71.1.5 47. Hamer M., Chida Y.. Intake of Fruit, Vegetables, and Antioxidants and Risk of Type 2 Diabetes: Systematic Review and Meta-Analysis. J. Hypertens. (2007.0) 25 2361-2369. DOI: 10.1097/HJH.0b013e3282efc214 48. Bacchetti T., Turco I., Urbano A., Morresi C., Ferretti G.. Relationship of Fruit and Vegetable Intake to Dietary Antioxidant Capacity and Markers of Oxidative Stress: A Sex-Related Study. Nutrition (2019.0) 61 164-172. DOI: 10.1016/j.nut.2018.10.034 49. Hawkesworth S., Dangour A.D., Johnston D., Lock K., Poole N., Rushton J., Uauy R., Waage J.. Feeding the World Healthily: The Challenge of Measuring the Effects of Agriculture on Health. Philos. Trans. R. Soc. Lond. B Biol. Sci. (2010.0) 365 3083-3097. DOI: 10.1098/rstb.2010.0122 50. Abdel-Aal E.-S.M., Akhtar H., Zaheer K., Ali R.. Dietary Sources of Lutein and Zeaxanthin Carotenoids and Their Role in Eye Health. Nutrients (2013.0) 5 1169-1185. DOI: 10.3390/nu5041169 51. Delcourt C., Carrière I., Delage M., Barberger-Gateau P., Schalch W.. Plasma Lutein and Zeaxanthin and Other Carotenoids as Modifiable Risk Factors for Age-Related Maculopathy and Cataract: The POLA Study. Investig. Ophthalmol. Vis. Sci. (2006.0) 47 2329-2335. DOI: 10.1167/iovs.05-1235 52. Huang L.L., Coleman H.R., Kim J., de Monasterio F., Wong W.T., Schleicher R.L., Ferris F.L., Chew E.Y.. Oral Supplementation of Lutein/zeaxanthin and Omega-3 Long Chain Polyunsaturated Fatty Acids in Persons Aged 60 Years or Older, with or without AMD. Investig. Ophthalmol. Vis. Sci. (2008.0) 49 3864-3869. DOI: 10.1167/iovs.07-1420 53. Tan J.S.L., Wang J.J., Flood V., Mitchell P.. Dietary Fatty Acids and the 10-Year Incidence of Age-Related Macular Degeneration: The Blue Mountains Eye Study. Arch. Ophthalmol. (2009.0) 127 656-665. DOI: 10.1001/archophthalmol.2009.76 54. Carpentier S., Knaus M., Suh M.. Associations between Lutein, Zeaxanthin, and Age-Related Macular Degeneration: An Overview. Crit. Rev. Food Sci. Nutr. (2009.0) 49 313-326. DOI: 10.1080/10408390802066979 55. Bazan N.G., Molina M.F., Gordon W.C.. Docosahexaenoic Acid Signalolipidomics in Nutrition: Significance in Aging, Neuroinflammation, Macular Degeneration, Alzheimer’s, and Other Neurodegenerative Diseases. Annu. Rev. Nutr. (2011.0) 31 321-351. DOI: 10.1146/annurev.nutr.012809.104635 56. Liu A., Chang J., Lin Y., Shen Z., Bernstein P.S.. Long-Chain and Very Long-Chain Polyunsaturated Fatty Acids in Ocular Aging and Age-Related Macular Degeneration. J. Lipid Res. (2010.0) 51 3217-3229. DOI: 10.1194/jlr.M007518 57. Querques G., Forte R., Souied E.H.. Retina and Omega-3. J. Nutr. Metab. (2011.0) 748361 12. DOI: 10.1155/2011/748361 58. De Meester F., Watson R.R., Zibadi S.. Omega-6/3 Fatty Acids: Functions, Sustainability Strategies and Perspectives (2012.0) 59. Souied E.H., Delcourt C., Querques G., Bassols A., Merle B., Zourdani A., Smith T., Benlian P.. Oral Docosahexaenoic Acid in the Prevention of Exudative Age-Related Macular Degeneration: The Nutritional AMD Treatment 2 Study. Ophthalmology (2013.0) 120 1619-1631. DOI: 10.1016/j.ophtha.2013.01.005 60. Chong E.W.-T., Kreis A.J., Wong T.Y., Simpson J.A., Guymer R.H.. Dietary ω-3 Fatty Acid and Fish Intake in the Primary Prevention of Age-Related Macular Degeneration: A Systematic Review and Meta-Analysis. Arch. Ophthalmol. (2008.0) 126 826-833. DOI: 10.1001/archopht.126.6.826 61. Mir S.M., Kanjilal S., Ahmed S.U.. Omega-3 Fatty Acids in Inflammatory Diseases. Omega-3 Fat. Acids (2016.0) 141-155. DOI: 10.1007/978-3-319-40458-5_11 62. Calder P.C.. N-3 Polyunsaturated Fatty Acids, Inflammation, and Inflammatory Diseases. Am. J. Clin. Nutr. (2006.0) 83 1505S-1519S. DOI: 10.1093/ajcn/83.6.1505S 63. Christen W.G., Schaumberg D.A., Glynn R.J., Buring J.E.. Dietary ω-3 Fatty Acid and Fish Intake and Incident Age-Related Macular Degeneration in Women. Arch. Ophthalmol. (2011.0) 129 921-929. DOI: 10.1001/archophthalmol.2011.34 64. D’Alessandro A., Lampignano L., De Pergola G.. Mediterranean Diet Pyramid: A Proposal for Italian People. A Systematic Review of Prospective Studies to Derive Serving Sizes. Nutrients (2019.0) 11. DOI: 10.3390/nu11061296 65. Wrolstad R.E.. Food Carbohydrate Chemistry (2012.0) 66. Kaushik S., Wang J.J., Flood V., Tan J.S.L., Barclay A.W., Wong T.Y., Brand-Miller J., Mitchell P.. Dietary Glycemic Index and the Risk of Age-Related Macular Degeneration. Am. J. Clin. Nutr. (2008.0) 88 1104-1110. DOI: 10.1093/ajcn/88.4.1104 67. Chiu C.-J., Milton R.C., Klein R., Gensler G., Taylor A.. Dietary Carbohydrate and the Progression of Age-Related Macular Degeneration: A Prospective Study from the Age-Related Eye Disease Study. Am. J. Clin. Nutr. (2007.0) 86 1210-1218. DOI: 10.1093/ajcn/86.4.1210 68. Wongwarawipat T., Papageorgiou N., Bertsias D., Siasos G., Tousoulis D.. Olive Oil-Related Anti-Inflammatory Effects on Atherosclerosis: Potential Clinical Implications. Endocr. Metab. Immune Disord. Drug Targets (2018.0) 18 51-62. DOI: 10.2174/1871530317666171116103618 69. Schwingshackl L., Christoph M., Hoffmann G.. Effects of Olive Oil on Markers of Inflammation and Endothelial Function—A Systematic Review and Meta-Analysis. Nutrients (2015.0) 7 7651-7675. DOI: 10.3390/nu7095356 70. Buckland G., Travier N., Barricarte A., Ardanaz E., Moreno-Iribas C., Sánchez M.-J., Molina-Montes E., Chirlaque M.D., Huerta J.M., Navarro C.. Olive Oil Intake and CHD in the European Prospective Investigation into Cancer and Nutrition Spanish Cohort. Br. J. Nutr. (2012.0) 108 2075-2082. DOI: 10.1017/S000711451200298X 71. Nocella C., Cammisotto V., Fianchini L., D’Amico A., Novo M., Castellani V., Stefanini L., Violi F., Carnevale R.. Extra Virgin Olive Oil and Cardiovascular Diseases: Benefits for Human Health. Endocr. Metab. Immune Disord. Drug Targets (2018.0) 18 4-13. DOI: 10.2174/1871530317666171114121533 72. Casas R., Estruch R., Sacanella E.. The Protective Effects of Extra Virgin Olive Oil on Immune-Mediated Inflammatory Responses. Endocr. Metab. Immune Disord. Drug Targets (2018.0) 18 23-35. DOI: 10.2174/1871530317666171114115632 73. Bordoni A., Danesi F., Dardevet D., Dupont D., Fernandez A.S., Gille D., Nunes Dos Santos C., Pinto P., Re R., Rémond D.. Dairy Products and Inflammation: A Review of the Clinical Evidence. Crit. Rev. Food Sci. Nutr. (2017.0) 57 2497-2525. DOI: 10.1080/10408398.2014.967385 74. Gopinath B., Flood V.M., Louie J.C.Y., Wang J.J., Burlutsky G., Rochtchina E., Mitchell P.. Consumption of Dairy Products and the 15-Year Incidence of Age-Related Macular Degeneration. Br. J. Nutr. (2014.0) 111 1673-1679. DOI: 10.1017/S000711451300408X 75. Klurfeld D.M.. Research Gaps in Evaluating the Relationship of Meat and Health. Meat Sci. (2015.0) 109 86-95. DOI: 10.1016/j.meatsci.2015.05.022 76. Chong E.W.-T., Simpson J.A., Robman L.D., Hodge A.M., Aung K.Z., English D.R., Giles G.G., Guymer R.H.. Red Meat and Chicken Consumption and Its Association with Age-Related Macular Degeneration. Am. J. Epidemiol. (2009.0) 169 867-876. DOI: 10.1093/aje/kwn393 77. Ersoy L., Ristau T., Lechanteur Y.T., Hahn M., Hoyng C.B., Kirchhof B., den Hollander A.I., Fauser S.. Nutritional Risk Factors for Age-Related Macular Degeneration. BioMed Res. Int. (2014.0) 2014 413150. DOI: 10.1155/2014/413150 78. Seddon J.M., Rosner B., Sperduto R.D., Yannuzzi L., Haller J.A., Blair N.P., Willett W.. Dietary Fat and Risk for Advanced Age-Related Macular Degeneration. Arch. Ophthalmol. (2001.0) 119 1191-1199. DOI: 10.1001/archopht.119.8.1191 79. Cross A.J., Pollock J.R.A., Bingham S.A.. Haem, Not Protein or Inorganic Iron, Is Responsible for Endogenous Intestinal N-Nitrosation Arising from Red Meat. Cancer Res. (2003.0) 63 2358-2360. PMID: 12750250 80. Lu M., Kuroki M., Amano S., Tolentino M., Keough K., Kim I., Bucala R., Adamis A.P.. Advanced Glycation End Products Increase Retinal Vascular Endothelial Growth Factor Expression. J. Clin. Investig. (1998.0) 101 1219-1224. DOI: 10.1172/JCI1277
--- title: Convenient and Sensitive Measurement of Lactosylceramide Synthase Activity Using Deuterated Glucosylceramide and Mass Spectrometry authors: - Michele Dei Cas - Linda Montavoci - Sara Casati - Nadia Malagolini - Fabio Dall’Olio - Marco Trinchera journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049619 doi: 10.3390/ijms24065291 license: CC BY 4.0 --- # Convenient and Sensitive Measurement of Lactosylceramide Synthase Activity Using Deuterated Glucosylceramide and Mass Spectrometry ## Abstract Lactosylceramide is necessary for the biosynthesis of almost all classes of glycosphingolipids and plays a relevant role in pathways involved in neuroinflammation. It is synthesized by the action of galactosyltransferases B4GALT5 and B4GALT6, which transfer galactose from UDP-galactose to glucosylceramide. Lactosylceramide synthase activity was classically determined in vitro by a method based on the incorporation of radiolabeled galactose followed by the chromatographic separation and quantitation of the product by liquid scintillation counting. Here, we used deuterated glucosylceramide as the acceptor substrate and quantitated the deuterated lactosylceramide product by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). We compared this method with the classical radiochemical method and found that the reactions have similar requirements and provide comparable results in the presence of high synthase activity. Conversely, when the biological source lacked lactosylceramide synthase activity, as in the case of a crude homogenate of human dermal fibroblasts, the radiochemical method failed, while the other provided a reliable measurement. In addition to being very accurate and sensitive, the proposed use of deuterated glucosylceramide and LC-MS/MS for the detection of lactosylceramide synthase in vitro has the relevant advantage of avoiding the costs and discomforts of managing radiochemicals. ## 1. Introduction The various classes of mammalian glycosphingolipids present on cell membranes are all derived from lactosylceramide (LacCer) (Figure 1). The biosynthesis of such a key molecule depends on an enzymatic activity known as LacCer synthase (EC 2.4.1.274), which is able to transfer galactose in a β1,4 linkage from UDP-galactose to glucosylceramide (GlcCer) (Figure 2). At present, two members of the large family of B4GALTs (β1,4 galactosyltransferases) are able to act as LacCer synthases: B4GALT5 and B4GALT6 [1,2]. The two human enzymes share $70\%$ identity at the aminoacid sequence level. The role of these two enzymes has been elucidated in mice via the phenotypic study of knock-out (KO) animals for b4galt5, b4galt6, or both. B4galt6 KO mice were healthy [3], whereas b4galt5 KO mice died at an early embryonic age [4], and conditional KO mice that only lacked b4galt5 in the brain were also healthy [5]. Double KO mice, conditional b4galt5 and null b4galt6, were born normally but were severely compromised in the central nervous system and died by four weeks of age [5]. Altogether, these data suggested that B4GALT5 is the major LacCer synthase [3,6] that can be rescued by B4GALT6, at least in the mouse brain. In addition to being a key intermediary in complex glycosphingolipid biosynthesis, LacCer has been reported to play a specific role in several signalling pathways, such as those orchestrated by platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), tumor necrosis factor-α (TNF-α), and oxidized low-density lipoprotein (LDL), thus affecting inflammation and atherosclerosis (reviewed in [7]). Moreover, LacCer specifically synthesized by B4GALT6 was found to be responsible for mediating neuroinflammation via interplay with cytosolic phospholipase A2 in astrocytes [8,9]. B4GALT5 was found to be modulated in several cancers, including colorectal [10], breast [11], and hepatocellular carcinoma [12], as well as in acute myeloid leukemia [13]. LacCer was also reported to be involved in diabetes [14], inflammatory bowel disease [15], adipocyte differentiation [6,16], and immunological functions of neutrophils [17]. Despite LacCer synthase’s pleiotropic relevance in both physiological and pathological conditions, a convenient enzymatic assay in vitro is still lacking. The current procedure is based on the use of radioactive UDP-Gal as a donor substrate, followed by the isolation of the radioactive LacCer product by various chromatographic techniques and quantitation by liquid scintillation counting. The only alternative proposed so far relies on the fluorescent acceptor substrate NBD-glucosylceramide and the detection of the NBD-LacCer product by HPLC [18]. Such a procedure was found to be very sensitive, but the required instrumentation is not routinely available, and the fluorescent substrate is quite different from its natural counterpart. Both restrictions have limited the application of such a procedure. In the present article, we reported an approach based on the use of deuterated GlcCer as an acceptor and the detection of the deuterated LacCer product in the reaction by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS). Our aim was to set a very sensitive method, avoiding the costs and troubles of radioactive UDP-Gal manipulation and supply. ## 2.1. Detection of LacCer Synthase Activity by Radiolabeled UDP-Gal and Chromatography Previous literature data have shown that LacCer synthase activity is due to the expression of B4GALT5, as well as the expression of B4GALT6 but to a lesser extent since the former is considered more efficient and relevant than the latter [5,6]. We determined the expression levels of their transcripts in the model cell line HEK-293T and in human dermal fibroblasts. We found that both transcripts are expressed at low levels in these two cell types when compared with that of the UGCG transcript (Figure 3), the one encoding GlcCer synthase, the enzyme forming the immediate biosynthetic precursor of LacCer. We thus measured LacCer synthase activity in three alternative sources: HEK-293T cells, HEK-293T cells transiently transfected with pcDNA3-B4GALT5 (named HEK-B4GALT5), and human dermal fibroblasts. Notably, upon transfection, the B4GALT5 transcript increases 105-fold in HEK-293T cells (Figure 3). Using tritiated UDP-Gal as the donor substrate and unlabeled GlcCer as the acceptor substrate, strong synthase activity was detected only in HEK-B4GALT5 cells. In fact, reactions lacking the GlcCer substrate incorporated 10–20-fold less tritiated Gal than those with the complete reaction mixture (Figure 4). Conversely, in the case of both untransfected (or mock-transfected) HEK-293T and dermal fibroblasts, the incorporation of tritiated Gal was not significantly different with or without acceptor GlcCer at any tested amounts of homogenate proteins (Figure 4). As expected, LacCer synthase activity in HEK-B4GALT5 required Mn2+ and was slightly stimulated by CDP-Choline (not shown). Since LacCer synthase is a Golgi apparatus resident activity, we tried to make it measurable in fibroblasts by preparing an enriched Golgi fraction for use as the enzyme source. Starting from large amounts of cells, we prepared a subcellular fraction enriched over 20-fold in sialyltransferase and galactosyltransferase activities measured with asialofetuin or ovalbumin as acceptor substrates, respectively. Using such a fraction as the enzyme source, we measured a LacCer synthase activity of 7.5 pmol/mg/min (Table 1), confirming that the activity was actually present in fibroblasts. Due to low expression levels, the assay’s procedure was unable to detect LacCer synthase in the crude homogenate. The attempt to measure LacCer synthase activity in vitro using GlcCerd7 as the acceptor substrate and unlabeled UDP-Gal as the donor substrate was performed under the same reaction conditions as in the radioisotopic assay, with the HEK-B4GALT5 homogenate as the enzyme source. As a negative control, we prepared reaction mixtures lacking GlcCerd7 or containing homogenate from untransfected HEK-293T cells. By LC-MS/MS, a peak corresponding to LacCerd7 was detected in the reaction mixture containing HEK-B4GALT5 homogenate that appeared negligible with mock-transfected cells, being more than 200-fold lower (Figure 5, note the different scales used in each panel). The enzymatic conversion of GlcCerd7 to LacCerd7 was linear with the amount of HEK-B4GALT5 homogenate (Figure 6 and Figure 7 panel C). In the absence of GlcCerd7, no peak at all at the size of LacCerd7 was detectable at any homogenate protein concentration. ## 2.2. Characterization of LacCer Synthase Activity Detected by GlcCerd7 and LC-MS/MS LacCer synthase activity in HEK-B4GALT5 cells assayed in vitro using GlcCer d7 and LC-MS/MS was found to be completely dependent on Mn2+, with an optimal concentration of 10 mM, and slightly stimulated by CDP-Choline, with an optimal concentration of 2.5 mM (Figure 7 panel A). The reaction was linear up to 60 min of incubation (Figure 7 panel B) and had homogenate protein concentrations up to 0.2 mg/mL (Figure 7 panel C). The activity was saturated at a UDP-Gal concentration over 0.8 mM (Figure 7 panel D), showing an apparent Km value 205 μM for the donor, and by a GlcCerd7 concentration over 10 μM (Figure 7 panel E), with an apparent Km of 3.37 μM for the acceptor. The corresponding Km values calculated via the non-linear regression of the Michaelis–Menten equation were 214.4 ($95\%$ confidence interval, 149.1–308.0) for the donor and 2.47 ($95\%$ confidence interval, 1.88 to 3.19) for the acceptor. Under such optimized conditions, LacCer synthase activity was also determined using crude homogenates prepared from non-transfected HEK-293T or human fibroblasts as the enzyme source. LacCer synthase activity was found to be linear up to 1.5 mg/mL homogenate protein in both fibroblasts and HEK-293T cells (Figure 7 panel F). The calculated specific activity ranged from 0.5 to 1.0 pmol/mg/min, being highest in neonatal fibroblasts undergoing less than 5 passages, lowest in adult fibroblasts with more than 15 passages, and intermediate in HEK-293T cells (Figure 8). A direct comparison of this new method with the classical radiochemical method was performed using HEK-B4GALT5 and HEK-293T and with dermal fibroblasts as the enzyme source at the same fixed protein concentration (Figure 4). LacCer synthase activity was detectable by both methods and with comparable values only in HEK-B4GALT5 cells that express very high activity. Conversely, in HEK-293T cells and dermal fibroblasts, the activity was much lower and detected by the new method only. In fact, at such low activity levels, the background of the radiochemical method overcomes the true activity value. The calculated precision, limit of detection, and limit of quantification of both methods are presented in Table 2. ## 3. Discussion We have set a novel assay procedure for determining LacCer synthase activity in biological samples based on the conversion of deuterated GlcCer (GlcCerd7) in LacCerd7 that is then quantitated by LC-MS/MS. Compared with the classical radiochemical assay, the new procedure requires the same reaction conditions, and it provides comparable results but offers several advantages. First, it avoids the costs and all the problems associated with the manipulation of radioisotopes. In addition, GlcCerd7 is an isotope of GlcCer without the relevant modifications present in the fluorescent substrate previously proposed as alternative to the radiochemical method [18]. Moreover, the procedure appeared even more sensitive than the classical one, allowing us to measure LacCer synthase activity in crude homogenates of human fibroblasts or model cell lines in which the radioisotopic method failed. In fact, it was necessary to prepare an enriched Golgi fraction to detect the activity with such a method. Our data indicated that the radiochemical assay performed in biological sources that poorly express LacCer synthase activity results in high background incorporation. This is probably due to the presence of several endogenous unspecific substrates and other galactosyltransferases. Conversely, the same sources lack lipid compounds with a molecular mass identical to that of LacCerd7, making the background close to zero. Notably, in a 20-fold enriched Golgi fraction from fibroblasts, the specific activity of LacCer synthase, determined with the radiochemical method (7.5 pmol/mg/min), corresponds to a value about 20-fold higher than that measured in crude homogenates using the novel approach (0.5 pmol/mg/min). Coupling such high specificity with the high sensitivity of LC-MS/MS, the procedure proposed here provides an unprecedented tool for functional studies to address the role of B4GALT$\frac{5}{6}$ and the cognate LacCer product in several pathological and physiological conditions relevant to current research topics [7]. Moreover, the LC-MS/MS analysis of the reaction mixture is able to simultaneously characterize and quantify both the enzymatic product LacCerd7 and the substrate acceptor GlcCerd7, making each point value more accurate and replicates more comparable. The use of deuterated substrates is emerging as a promising alternative for measuring in vitro the activity of enzymes involved in glycosphingolipid biosynthesis, as we reported very recently for UGCG [19]. ## 4.1. DNA Constructs Human B4GALT5 and B4GALT6 cDNAs were obtained by PCR using total RNA extracted from Huh-7 or Hep-3B cells [20] in the presence of Phusion High-fidelity Taq polymerase (ThermoFischer Scientific, Rome, Italy, Italian distributor), according to the manufacturer’s protocol. The following primer pairs were used: B4GALT5 (forward) 5′-CGCGAAGCTTGCGATCGCCATGCGCGCCCGCCGGG and (reverse) 5′-CGCGTCTAGAGTTTAAACTCAGTACTCGTTCACCTGAGCC; and B4GALT6 (forward) 5′-CGCGAAGCTTGCGATCGCCATGTCTGTGCTCAGGCGGATG and (reverse) 5′-CGCGTCTAGAGTTTAAACTTAATAGTCTTCGATTGGAGCTAACTC, both containing HindIII and XbaI restriction sites (italicized), respectively. Annealing was at 64 °C, and 30 cycles of amplification were run. The obtained fragments were purified by spin columns, digested with HindIII, repurified, digested with XbaI, purified again, and ligated to the pcDNA3 vector that was previously digested/purified with the same enzymes. E. coli Max-efficiency DH5α (ThermoFischer Scientific) was transformed by an aliquot of ligation reactions, and the obtained colonies were inoculated in liquid cultures to prepare plasmid DNA. The obtained clones were first assessed by restriction digestion and then confirmed by direct DNA sequencing. ## 4.2. Cell Culture, Transfection, and Processing HEK-293T cells were grown in DMEM supplemented with $10\%$ foetal bovine serum and transiently transfected with plasmid DNA in the presence of Fugene-HD (Promega, Madison, WI, USA), as previously reported [21]. Upon transfection, cells were harvested by trypsinization, pelleted, washed twice with PBS, and aliquoted. A small aliquot (<0.5 × 106 cells) was used for total RNA extraction using a Relia-Prep RNA kit (Promega), and the remaining cells (2–10 × 106 cells) were resuspended in a 0.1 M Tris/HCl buffer, pH 7.5, containing $0.5\%$ TritonX-100, vortexed to homogeneity, and stored in aliquots at −80 °C. After thawing on ice, the crude homogenate was used as the enzyme source for the in vitro assay. Protein content was determined by the bicinconic acid method (BCA protein kit, ThermoFischer Scientific). Homogenates for enzyme assay were kept at a protein concentration of about 10–20 mg/mL. Human neonatal (P10857) and adult (P10858) fibroblasts were purchased from Innoprot (Bizkaia, Spain) and cultured in DMEM supplemented with $10\%$ foetal bovine serum and antibiotics. Cell harvesting and aliquot preparation were performed, as was the case for transfected HEK-293 T cells. ## 4.3. Reverse Transcription Quantitative Real-Time Polymerase Chain Reaction First strand cDNA was synthesized from 0.5 to 1 μg of total RNA by a commercial kit (GoScript oligodT mix, PromegaItalia srl, Milano, Italy). Control reactions were prepared by omitting the reverse transcriptase in the reaction. cDNAs were diluted 1:4, v/v, with mQ water, and 1–2.0 μL of first strand reactions was amplified in a volume of 20 μL using Sybr Premix Ex Taq (Tli RNase H Plus, Takara, Sandiano, Italy, Italian distributor), ROX as the reference dye, and a StepOnePlus instrument (Applied Biosystem Life Technologies, Waltham, MA, USA) as reported [22]. Primer sequences for amplification were as follows: B4GALT5 (forward) 5′-GAGAACAATCGGTGCTCAGG, (reverse) 5′-GGGCCCTTCATGGAAGGG; B4GALT6 (forward) 5′-CCGGAAAACTTCACATACTCAC, (reverse) 5′-GAACTGCCACCTTCCATCTG; UGCC (forward) 5′-GTGATAGTGGAATAAGAGTAATTCC, (reverse) 5′-TGAAGTTCCAAAATATACCTGCTC. Annealing temperature was 60 °C. The amounts of amplified target cDNAs were calculated as ΔCt with respect to GAPDH [22]. ## 4.4. LacCer Synthase Activity Assay GlcCer, purified as reported [23], or GlcCerd7 (C15 Glucosyl(β) Ceramide-d7 (d18:1-D$\frac{7}{15}$:0, 330729P-1MG], Merck, Rome, Italy, Italian distributor), with 20–800 pmol dissolved in chloroform/methanol, 2:1 (v/v), and 15 μg Triton-X100 in the same solvent were placed together at the bottom of a 0.6 mL microcentrifuge tubes and allowed to dry at RT under hood and then kept at −20 °C until used. A reaction solution was prepared and added to each tube in order to obtain the following final concentrations: 0.2 M Tris/HCl pH 7.0, 0.4 mM UDP-Gal, 10 mM MnCl2, and 2.5 mM CDP-choline. In the case of the radio isotopic method, UDP-Gal was brought to a specific radioactivity of 10 mCi/mmol by adding UDP-[3H]Gal [24]. In an ice bucket, various amounts of protein (see results) were added to each tube, which already contained water, to a final volume of 20 μL. The reactions were started by placing the tubes at 37 °C and incubating them for 10–60 min. The reactions were stopped by placing the tubes on ice and then storing them at −20 °C. In the case of the radio isotopic method, the mixture was assayed by descending chromatography, and radioactivity was measured by liquid scintillation as reported [21]. ## 4.5. Reaction Product Characterization and Quantification by LC–MS/MS Samples for the determination of LacCer synthase activity were precipitated by the addition of pure methanol (75 μL) and centrifuged at 13,400 rpm for 10 min. The precipitates were discarded, and pure extracts (5 μL) were directly injected in LC-MS/MS. The samples were analyzed by a high-sensitivity LC–MS/MS consisting of a QTrap 5500 triple quadrupole linear ion trap mass spectrometer (Sciex, Darmstadt, Germany) equipped with an electrospray ionization (ESI) source and coupled with an Agilent 1200 Infinity pump Ultra High-Pressure Liquid Chromatography (UHPLC) system (Agilent Technologies, Palo Alto, CA, USA). Chromatographic separation was carried out on a reverse-phase Acquity UPLC BEH C8 column at 1.7 µm particle size, and100 × 2.1 mm (Waters, Franklin, MA, USA) at 30 °C using linear gradient elution with two solvents: $0.2\%$ formic acid and 2 mM ammonium formate in water (solvent A) and $0.2\%$ formic acid and 1 mM ammonium formate in acetonitrile (solvent B). Solvent A and B were $20\%$ and $80\%$ at 0.00 min, respectively. Solvent B increased to $90\%$ from 0.00 to 3.00 min, held at $90\%$ from 3.00 to 6.00 min, increased to $99\%$ from 6.00 to 10.00 min, held at $99\%$ from 10.00 to 12.00 min, and then decreased back to $80\%$ from 12.00 to 12.10 and held at $80\%$ until 15.00 min for re-equilibration. The flow rate was kept constant at 0.40 mL/min during the analysis. The separated analytes were detected with a triple quadrupole MS operated in multiple reaction monitoring (MRM) mode via a positive ESI using the following precursor ion and product ions transition: GlucCerd7 C15:0 (m/z 693.6 > 271.3, CE 45 eV, DP 65 eV) and LacCer d7 C15:0 (m/z 855.6 > 271.3, CE 60 eV, DP 65 eV). Data acquisition and processing were performed using Analyst®1.7.1 and MultiQuant®2.1.1 software (Sciex, Darmstadt, Germany), respectively. The declustering potential (DP) and collision energy (CE) are compound-dependent parameters that should be investigated before every mass spectrometry experiment. In particular, DP is a voltage applied—in MS ion source—that helps prevent ions from clustering together, whereas CE is the voltage produced to induce the dissociation of the molecular ion from their fragments’ ions. One of the most popular targeted types of MS experiments is multiple reaction monitoring (MRM). With the use of a triple quadrupole mass spectrometer, this method first targets the interested protonated form of the molecular ion before fragmenting it to create different daughter ions depending on the applied CE. ## 4.6. Preparation of Golgi Apparatus Fraction from Fibroblasts About 2.5 × 107 fibroblasts were detached by trypsinization, washed twice with PBS, resuspended with 0.3 mL of 10 mM Tris/HCl pH 7.5 containing 0.1 mM EDTA and 0.25 M sucrose, and homogenized by passing them several times through a 27G needle of a 1.0 mL syringe. The obtained homogenate was spun at 2000 g for 10 min at 4 °C. The supernatant was removed, and the pellet was rehomogenized and spun as above. The two supernatants were combined, $40\%$ sucrose was made in the same buffer (final volume about 1.4 mL) and placed at the bottom of an ultracentrifuge tube and overlayered with 1 mL each of $35\%$ and $25\%$ sucrose solutions prepared in the same buffer, and they were covered with 0.25 M sucrose solution. The tube was spun at 54,000 rpm in a Beckmann Ti 55 rotor for 90 min at 4 °C. Materials migrating at the 35–$25\%$ sucrose interface were collected, diluted 1:8 (v/v) with PBS, and spun again at 54,000 rpm in a Beckmann Ti 55 rotor for 60 min at 4 °C to obtain the Golgi membrane fraction that was resuspended with 0.05 mL of 0.1 M Tris/HCl pH7.0 containing $0.2\%$ Triton-X100. Sialyltransferase activity with asialofetuin and galactosyltransferase activity with ovalbumin were determined, as reported with other glycoprotein substrates [25]. ## 4.7. Equations and Statistical Analysis LacCer synthase catalyzes a reaction that uses two substrates (GlcCer and UDP-Gal) and forms two products (LacCer and UDP). Consequently, Michaelis–Menten kinetics could be used only by keeping either substrate at saturating concentrations and varying the concentrations of the other. For a graphical description and apparent kinetic constant calculations, the Hanes–Woolf equation was used: [S]/$v = 1$/Vmaxapp * [S] + Kmapp/Vmaxapp. Linear regression was obtained via Microsoft Excel. R2 values were 0.990 (UDP-Gal saturation curve) and 0.982 (GlcCerd7 saturation curve). Non-linear regression of the Michaelis–Menten equation was obtained via GraphPad Prism (Dotmatics Scientific, Boston, MA, USA). The specificity of LacCer synthase assays was assessed by using the response of a blank sample at the limit of quantitation (LOQ). The LOQ was determined as the lowest concentration with a signal-to-noise ratio of the instrumental response ≥10; the limit of detection (LOD) was determined as the lowest concentration with a signal-to-noise of ≥3. Precision of the methods was determined by assaying six biological replicates and calculating the coefficient of variation (CV%), which should be appropriate if it is lower than $15\%$. For transcripts quantitation, qPCR was performed in duplicate twice, starting from cDNA prepared from two independent transfections of HEK-293T cells or fibroblast cultures. For LacCer synthase activity determination, assays were performed in duplicate twice, starting from two individual cell homogenates consisting of distinct cell plates of growing cells or independent transfections of HEK-293T cells. ## References 1. Nomura T., Takizawa M., Aoki J., Arai H., Inoue K., Wakisaka E., Yoshizuka N., Imokawa G., Dohmae N., Takio K.. Purification, CDNA Cloning, and Expression of UDP-Gal: Glucosylceramide Beta-1,4-Galactosyltransferase from Rat Brain. J. Biol. Chem. (1998) 273 13570-13577. DOI: 10.1074/jbc.273.22.13570 2. Kumagai T., Sato T., Natsuka S., Kobayashi Y., Zhou D., Shinkai T., Hayakawa S., Furukawa K.. Involvement of Murine β-1,4-Galactosyltransferase V in Lactosylceramide Biosynthesis. Glycoconj. J. (2010) 27 685-695. DOI: 10.1007/s10719-010-9313-2 3. Tokuda N., Numata S., Li X., Nomura T., Takizawa M., Kondo Y., Yamashita Y., Hashimoto N., Kiyono T., Urano T.. Β4GalT6 Is Involved in the Synthesis of Lactosylceramide with Less Intensity than Β4GalT5. Glycobiology (2013) 23 1175-1183. DOI: 10.1093/glycob/cwt054 4. Kumagai T., Tanaka M., Yokoyama M., Sato T., Shinkai T., Furukawa K.. Early Lethality of Beta-1,4-Galactosyltransferase V-Mutant Mice by Growth Retardation. Biochem. Biophys. Res. Commun. (2009) 379 456-459. DOI: 10.1016/j.bbrc.2008.12.078 5. Yoshihara T., Satake H., Nishie T., Okino N., Hatta T., Otani H., Naruse C., Suzuki H., Sugihara K., Kamimura E.. Lactosylceramide Synthases Encoded by B4galt5 and 6 Genes Are Pivotal for Neuronal Generation and Myelin Formation in Mice. PLoS Genet. (2018) 14. DOI: 10.1371/journal.pgen.1007545 6. Rakhshandehroo M., van Eijkeren R.J., Gabriel T.L., de Haar C., Gijzel S.M.W., Hamers N., Ferraz M.J., Aerts J.M.F.G., Schipper H.S., van Eijk M.. Adipocytes Harbor a Glucosylceramide Biosynthesis Pathway Involved in INKT Cell Activation. Biochim. Biophys. Acta Mol. Cell Biol. Lipids (2019) 1864 1157-1167. DOI: 10.1016/j.bbalip.2019.04.016 7. Chatterjee S., Balram A., Li W.. Convergence: Lactosylceramide-Centric Signaling Pathways Induce Inflammation, Oxidative Stress, and Other Phenotypic Outcomes. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22041816 8. Mayo L., Trauger S.A., Blain M., Nadeau M., Patel B., Alvarez J.I., Mascanfroni I.D., Yeste A., Kivisäkk P., Kallas K.. Regulation of Astrocyte Activation by Glycolipids Drives Chronic CNS Inflammation. Nat. Med. (2014) 20 1147-1156. DOI: 10.1038/nm.3681 9. Chao C.-C., Gutiérrez-Vázquez C., Rothhammer V., Mayo L., Wheeler M.A., Tjon E.C., Zandee S.E.J., Blain M., de Lima K.A., Takenaka M.C.. Metabolic Control of Astrocyte Pathogenic Activity via CPLA2-MAVS. Cell (2019) 179 1483-1498.e22. DOI: 10.1016/j.cell.2019.11.016 10. Chatterjee S.B., Hou J., Bandaru V.V.R., Pezhouh M.K., Syed Rifat Mannan A.A., Sharma R.. Lactosylceramide Synthase β-1,4-GalT-V: A Novel Target for the Diagnosis and Therapy of Human Colorectal Cancer. Biochem. Biophys. Res. Commun. (2019) 508 380-386. DOI: 10.1016/j.bbrc.2018.11.149 11. Tang W., Li M., Qi X., Li J.. Β1,4-Galactosyltransferase V Modulates Breast Cancer Stem Cells through Wnt/β-Catenin Signaling Pathway. Cancer Res. Treat. (2020) 52 1084-1102. DOI: 10.4143/crt.2020.093 12. Han Y., Li Z., Wu Q., Liu H., Sun Z., Wu Y., Luo J.. B4GALT5 High Expression Associated with Poor Prognosis of Hepatocellular Carcinoma. BMC Cancer (2022) 22. DOI: 10.1186/s12885-022-09442-2 13. Wu Y., Zhao B., Chen X., Geng X., Zhang Z.. Circ_0009910 Sponges MiR-491-5p to Promote Acute Myeloid Leukemia Progression through Modulating B4GALT5 Expression and PI3K/AKT Signaling Pathway. Int. J. Lab. Hematol. (2022) 44 320-332. DOI: 10.1111/ijlh.13742 14. Novgorodov S.A., Riley C.L., Yu J., Keffler J.A., Clarke C.J., Van Laer A.O., Baicu C.F., Zile M.R., Gudz T.I.. Lactosylceramide Contributes to Mitochondrial Dysfunction in Diabetes. J. Lipid Res. (2016) 57 546-562. DOI: 10.1194/jlr.M060061 15. Filimoniuk A., Blachnio-Zabielska A., Imierska M., Lebensztejn D.M., Daniluk U.. Sphingolipid Analysis Indicate Lactosylceramide as a Potential Biomarker of Inflammatory Bowel Disease in Children. Biomolecules (2020) 10. DOI: 10.3390/biom10071083 16. Hocine T., Blaise S., Hachet C., Guillot A., Sartelet H., Maurice P., Bennasroune A., Martiny L., Duca L., Romier-Crouzet B.. Lactosylceramide Induced by Elastin-Derived Peptides Decreases Adipocyte Differentiation. J. Physiol. Biochem. (2020) 76 457-467. DOI: 10.1007/s13105-020-00755-z 17. Iwabuchi K., Nakayama H., Hanafusa K.. Lactosylceramide-Enriched Microdomains Mediate Human Neutrophil Immunological Functions via Carbohydrate-Carbohydrate Interaction. Glycoconj. J. (2022) 39 239-246. DOI: 10.1007/s10719-022-10060-0 18. Hayashi Y., Horibata Y., Sakaguchi K., Okino N., Ito M.. A Sensitive and Reproducible Assay to Measure the Activity of Glucosylceramide Synthase and Lactosylceramide Synthase Using HPLC and Fluorescent Substrates. Anal. Biochem. (2005) 345 181-186. DOI: 10.1016/j.ab.2005.05.029 19. Dei Cas M., Casati S., Roda G., Sardi P., Paroni R., Fonzo A., Trinchera M.. A Sensitive Method for Determining UDP-Glucose: Ceramide Glucosyltransferase (UGCG) Activity in Biological Samples Using Deuterated Glucosylceramide as Acceptor Substrate. Glycobiology (2022) 33 88-94. DOI: 10.1093/glycob/cwac081 20. Zulueta A., Caretti A., Signorelli P., Dall’Olio F., Trinchera M.. Transcriptional Control of the B3GALT5 Gene by a Retroviral Promoter and Methylation of Distant Regulatory Elements. FASEB J. (2014) 28 946-955. DOI: 10.1096/fj.13-236273 21. Indellicato R., Parini R., Domenighini R., Malagolini N., Iascone M., Gasperini S., Masera N., Dall’Olio F., Trinchera M.. Total Loss of GM3 Synthase Activity by a Normally Processed Enzyme in a Novel Variant and in All ST3GAL5 Variants Reported to Cause a Distinct Congenital Disorder of Glycosylation. Glycobiology (2019) 29 229-241. DOI: 10.1093/glycob/cwy112 22. Aronica A., Avagliano L., Caretti A., Tosi D., Bulfamante G.P., Trinchera M.. Unexpected Distribution of CA19.9 and Other Type 1 Chain Lewis Antigens in Normal and Cancer Tissues of Colon and Pancreas: Importance of the Detection Method and Role of Glycosyltransferase Regulation. Biochim. Biophys. Acta Gen. Subj. (2017) 1861 3210-3220. DOI: 10.1016/j.bbagen.2016.08.005 23. Trinchera M., Fiorilli A., Ghidoni R.. Localization in the Golgi Apparatus of Rat Liver UDP-Gal:Glucosylceramide Beta 1----4galactosyltransferase. Biochemistry (1991) 30 2719-2724. DOI: 10.1021/bi00224a021 24. Salvini R., Bardoni A., Valli M., Trinchera M.. Beta 1,3-Galactosyltransferase Beta 3Gal-T5 Acts on the GlcNAcbeta 1-->3Galbeta 1-->4GlcNAcbeta 1-->R Sugar Chains of Carcinoembryonic Antigen and Other N-Linked Glycoproteins and Is down-Regulated in Colon Adenocarcinomas. J. Biol. Chem. (2001) 276 3564-3573. DOI: 10.1074/jbc.M006662200 25. Indellicato R., Domenighini R., Malagolini N., Cereda A., Mamoli D., Pezzani L., Iascone M., Dall’olio F., Trinchera M.. A Novel Nonsense and Inactivating Variant of ST3GAL3 in Two Infant Siblings Suffering Severe Epilepsy and Expressing Circulating CA19.9. Glycobiology (2020) 30 95-104. DOI: 10.1093/glycob/cwz079
--- title: 'Predictors of Inadequate Health Literacy among Patients with Type 2 Diabetes Mellitus: Assessment with Different Self-Reported Instruments' authors: - Marija Levic - Natasa Bogavac-Stanojevic - Dragana Lakic - Dusanka Krajnovic journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049631 doi: 10.3390/ijerph20065190 license: CC BY 4.0 --- # Predictors of Inadequate Health Literacy among Patients with Type 2 Diabetes Mellitus: Assessment with Different Self-Reported Instruments ## Abstract Introduction: Obtaining, understanding, interpreting, and acting on health information enables people with diabetes to engage and make health decisions in various contexts. Hence, inadequate health literacy (HL) could pose a problem in making self-care decisions and in self-management for diabetes. By applying multidimensional instruments to assess HL, it is possible to differentiate domains of functional, communicative, and critical HL. Objectives: Primarily, this study aimed to measure the prevalence of inadequate HL among type 2 diabetes mellitus patients and to analyze the predictors influencing health literacy levels. Secondly, we analyzed if different self-reported measures, unidimensional instruments (Brief Health Literacy instruments (BRIEF-4 and abbreviated version BRIEF-3), and multidimensional instruments (Functional, Communicative and Critical health literacy instrument (FCCHL)) have the same findings. Methods: The cross-sectional study was conducted within one primary care institution in Serbia between March and September 2021. Data were collected through Serbian versions of BRIEF-4, BRIEF-3, and FCCHL-SR12. A chi-square test, Fisher’s exact test, and simple logistic regression were used to measure the association between the associated factors and health literacy level. Multivariate analyses were performed with significant predictors from univariate analyses. Results: Overall, 350 patients participated in the study. They were primarily males ($55.4\%$) and had a mean age of 61.5 years (SD = 10.5), ranging from 31 to 82 years. The prevalence of inadequate HL was estimated to be $42.2\%$ (FCCHL-SR12), $36.9\%$ (BRIEF-3) and $33.8\%$ (BRIEF-4). There are variations in the assessment of marginal and adequate HL by different instruments. The highest association was shown between BRIEF-3 and total FCCHL-SR12 score (0.204, $p \leq 0.01$). The total FCCHL-SR12 score correlates better with the abbreviated BRIEF instrument (BRIEF-3) than with BRIEF-4 (0.190, $p \leq 0.01$). All instruments indicated the highest levels for the communicative HL domain and the lowest for the functional HL domain with significant difference in functional HL between the functional HL of FCCHL-SR12 and both BRIEF-3 and BRIEF-4 ($$p \leq 0.006$$ and 0.008, respectively). Depending on applied instruments, we identified several variables (sociodemographic, access to health-related information, empowerment-related indicators, type of therapy, and frequency of drug administration) that could significantly predict inadequate HL. Probability of inadequate HL increased with older age, fewer children, lower education level, and higher consumption of alcohol. Only high education was associated with a lower probability of inadequate HL for all three instruments. Conclusions: The results we obtained indicate that patients in our study may have been more functionally illiterate, but differences between functional level could be observed if assessed by unidimensional and multidimensional instruments. The proportion of patients with inadequate HL is approximately similar as assessed by all three instruments. According to the association between HL and educational level in DMT2 patients we should investigate methods of further improvement. ## 1. Introduction The World Health Organization (WHO) and the International Diabetes Federation estimated that in 2019, 463 million people worldwide suffered from diabetes mellitus (DM) and that the number of DM patients will increase to 700 million by 2045. Although the highest incidence rates are registered in developed countries, a significant increase in the number of patients is expected in developing countries, including Serbia [1]. According to an estimate by the Institute for Public Health of Serbia, approximately 770,000 people—$12.0\%$ of the adult population—suffer from DM in the Republic of Serbia [2]. According to the estimates of domestic experts and based on the results of international studies, $43\%$ [330,000] of patients with type 2 DM (DMT2) have not been diagnosed and are not aware of their disease [3,4,5]. The number of people with DM is many times higher ($95\%$) compared to people with type 1 DM (DMT1). In Serbia, as well as in the world’s developed countries, DM is the fifth leading cause of mortality [6] and the fifth-highest cause of disease burden [7]. The main goal of today’s health system is to promote and maintain good health while enabling people to take care of their health and to participate more in making decisions related to their health [8,9,10]. Participation in health decisions could be moderated by specific patients’ skills and by the situational demands and complexities experienced by patients in their attempts to obtain, understand, and use health information or health services. As a social determinant of health, HL is a crucial driver to approaching health information competently and effectively, making self-care decisions, and taking self-management actions for diabetes. Recognizing that individual responses to information will result in different learning outcomes and associated behavioral and health outcomes led us to research that focuses on the categorization of functional health literacy (FHL), communicative/interactive health literacy (IHL), and critical health literacy (CHL). FHL is often required to meet the immediate and necessary goals of clinical care as it is related to the specific skills required to achieve outcomes that are determined primarily by those providing healthcare [11]. In such circumstances, specific skills to manage prescribed activities in chronic therapies could be required of any patient with chronic diseases at a point of decision-making. In contrast, IHL and CHL require the development of transferable skills, including obtaining, understanding, evaluating (interpreting), and acting on (applying) health information, which enable patients to engage with and make health decisions in various contexts. The development of specific and transferable skills offers a greater opportunity to optimize HL’s contribution in mediating self-care, low adherence, and medication management for a chronic disease such as diabetes. Several studies found that a small number of DMT2 patients had adequate levels of HL [12,13,14,15,16,17]. Some studies showed a connection between low-level HL, limited knowledge about the disease [18,19], and poor glycemic control [19,20,21]. Diabetic individuals with lower HL use preventive health services less, are more at risk of misdiagnosis, experience difficulties managing chronic diseases, have poor drug and treatment compliance, and have poorer health outcomes [15,17]. As self-care for a chronic disease such as diabetes often relies on information given in verbal instructions, printed educational materials, and patient education courses [12], low-literacy patients may have problems finding and following these instructions when they are to be integrated into everyday life. Inadequate HL (InHL) has been associated with poorer health states, broader inequalities, and higher health system costs. Hence, measuring changes to the specific skills required by DMT2 patients for decision-making and the more generic transferable skills that enable well-informed and more autonomous health decision-making could be considered crucial for diabetes self-management [12,21]. Based on the previous considerations, different self-reported and objective measures are needed if we want to assess both FHL as well as IHL and CHL. It is already proven that instruments vary in how they operationalize the concept of HL into a measurable construct, and many address limited sets of conceptual dimensions of HL. Measures that have the broadest measurement scopes are considered the most suitable for application in diabetes reference. In this research, three instruments were used: the Functional, Communicative, Critical, Health Literacy instrument (FCCHL) and the Brief Health Literacy Screening Instruments with three questions (BRIEF with three (BRIEF-3) and four questions (BRIEF-4)). The use of the BRIEF instruments in addition to FCCHL would serve as an additional confirmation of the distribution of diabetics in the category of those with high (adequate/AHL) or low (inadequate/InHL) HL. Keeping in mind that WHO claims HL is one of the most critical determinants of health [21], this study aims to assess HL and its domains among DMT2 patients in Serbia and identify predictors of InHL. Furthermore, the authors analyzed if different self-reported measures, screening unidimensional (BRIEF-3 and BRIEF-4 instruments), and multidimensional (FCCHL-SR12) have the same findings. ## 2. Method A cross-sectional study was conducted from March to September 2021 using non- probability sampling within one healthcare center in the municipality of Belgrade. A convenient sample of primary care patients with DMT2 was used. Before the survey, authors recruited five research assistants. To ensure that they were familiar with the purpose, process, and procedure of applying the instrument, the authors systematically trained three pharmacy graduates and two doctors as research assistants. Throughout data collection, the interviewers (researchers and assistants) explained the purpose and significance of the study to the participants and obtained written informed consent. The survey was conducted in a doctor’s office after the completed examinations. Patients filled out the questionnaires themselves (self-administrated measurements), and interviewers were at their disposal during this time. They did not receive any payment for filling out the instrument. All data were anonymous and were entered into the database as such. The required number of participants was calculated based on the population of DM patients in the Belgrade [80,241] area. An estimated percentage of high HL from the literature (which was $36\%$ [22,23,24]) was used for the calculation, and $95\%$ confidence interval with an error of $5\%$ were used for calculation. Based on these parameters, the required sample was at least 353. The required sample size of 353 was increased by $10\%$ due to potential dropouts (accounting for the non-responder rate) during the study. ## 2.1. Sample and Data Collection The target population was patients diagnosed with DMT2 at least six months before the start of the study who knew the Serbian language, were over 18 years old, and voluntarily agreed to participate with signed informed consent. Exclusion criteria were medical background (such as doctors, nurses, pharmacists). Three instruments were used, which participants filled out anonymously and voluntarily after receiving comprehensive information from the interviewer). The interviewers first explained the aim and the course of the research; those patients who wanted to participate filled in the questionnaires, and one medical parameter, HbA1c value, was taken from the medical documentation on the participants’ agreement. ## 2.2. Instruments Participants were given three instruments at the same time to complete in the following order: general instrument, instruments for rapid screening of health literacy (BRIEF-3 and BRIEF-4), the 12-item Serbian version of the three-dimensional instrument for the assessment of health literacy (FCCHL-SR12). The general instrument is a specially created instrument that includes questions related to the social and demographic characteristics of the participants, health-related issues, health behaviors, as well as empowerment-related indicators. The instruments’ questions were generated using a combination of literature review methods and analysis of instruments used in similar research conducted in diabetic populations. In line with the Sørensen model [22], many outcomes associated with HL were collected: self-assessed health status (excellent, very good, good, so-so/fair, bad), duration of diabetes, type of therapy for diabetes, and frequency of therapy. To measure health status, participants were asked about other long-term illnesses (illnesses that have lasted or are expected to last for at least 6 months), HbA1c value, coded in ≤$7\%$ and >$7\%$, and do not know/refuse. In addition to including questions on gender, birth year, educational level, and marital status, the questionnaire also has questions on the number of family members living in the same household, employment status (currently having a paid job), financial situation, and the number of children. Health behaviors (measured using three items on physical activity, tobacco use, and alcohol consumption), access to health-related information (a primary source of information), and empowerment-related indicators (perceived interest in one’s health and perceived self-assessment of one’s health in general) were recorded as well [23,24]. FCCHL is a multidimensional instrument that can be considered the most valuable and comprehensive instrument in screening for InHL [25]. BRIEF instruments focus on function, but contain the interaction, comprehension, and self-efficacy dimensions of HL. BRIEF instruments are primarily used for quick assessment, and they have many advantages in comparison to other instruments, including that they are less likely to cause anxiety and shame [22,25,26,27,28,29]. All three instruments can be considered the best for measuring FHL in patients with DMT2 and other chronic diseases [23,24]. They are self-reported instruments and are quick, easy, and inexpensive to administer. The authors used culturally adapted versions of these measures developed in previous studies for the Serbian population that are valid and reliable. BRIEF-3 consists of three questions and BRIEF-4 of four, which are evaluated on a five-point scale of 1–5. Both instruments can be administered and scored in less than two minutes. Each item is worth 1 to 5 points depending on their response, and summated responses provide a total score from a minimum of 3 to a maximum of 15 in BRIEF-3 and from a minimum of 4 to a maximum of 20 in BRIEF-4. We used the same scoring system as other researchers [28,29,30], categorizing individuals’ functional health literacy as inadequate HL/InHL (points 3–9), marginal HL/MHL (points 10–12), and adequate HL/AHL (points 13–15) in BRIEF-3 and InHL (points 4–12), MHL (points 13–16) and AHL (17–20) in BRIEF-4. The 12-item Serbian version of the three-dimensional FCCHL (FCCHL-SR12) is a multidimensional scale that uses subscales to measure different but related aspects to capture the complexity of HL that is context- and content-specific for the assessment of three dimensions of HL (functional, communicative, and critical) as well as total HL. It consists of 12 items and uses a 4-point Likert scale (1–4). The instrument contains questions that ask how often (never to often) patients have trouble with reading or understanding leaflets from healthcare providers/hospitals or pharmacies (FHL; 4 items) and how difficult it is (easy to rather difficult) to perform specific actions concerning health information (IHL (4 items) and CHL (4 items)). Since there are no defined cutoffs for InHL, MHL, and AHL for FCCHL-SR12, authors used defined levels according to the same principle as for BRIEF instruments. The level of HL measured using BRIEF was categorized as InHL (up to $60\%$ of a maximum score), MHL (up to $80\%$ of a maximum score), or AHL (more than $80\%$ of maximum score). The levels of FHL, IHL, CHL, and total HL measured using the FCCHL-SR12 instrument were defined in the same way. The FCCHL-SR12 score range is 12–48, and the scoring is as follows: 12–28 (InHL), 29–38 (MHL), and 39–48 (AHL) Investigating the dimensionality and validity of the FCCHL-SR12 was reported elsewhere [31], showing acceptable psychometric properties. A 3-dimensional 12-item version of the FCCHL had acceptable psychometric properties. Confirmatory factor analysis provided a good statistical and conceptual fit for the data. The analysis of the internal consistency of the FCCHL-SR12 was satisfactory for the health literacy total scores (Cronbach’s alpha was 0.767) and also acceptable for the three correlated subscales (Cronbach’s alpha equaled 0.792, 0.748, and 0.796 for FHL, IHL, and CHL, respectively) [31]. ## 2.3. Statistical Analysis Data were managed and analyzed using IBM SPSS Statistics for Windows, version 27.0. Armonk, NY, USA: IBM Corp software package. The significance level was set at a $95\%$ confidence level, with a p value of less than 0.05. Descriptive statistics, such as frequencies (absolute and relative), were used to describe the sample. Mean and standard deviation (SD) were employed for continuous data (such as age). To determine the relationship between HL scales, Spearman’s correlation coefficient was calculated. The distributions of sociodemographic, health-related, and health behavior characteristics along with health-related information and empowerment-related indicators through different HL levels were compared via chi-square test. Univariate and multivariate logistic regression analysis were used to determine independent predictors of inadequate HL as evaluated by all three instruments. ## 2.4. Ethical Considerations The study adhered to the ethical standards in line with the International Ethical Guidelines for Health-related Research Involving Humans (the Council for International Organizations of Medical Sciences, CIOMS, 2016) and Declaration of Helsinki (World Medical Association, 2013). Approvals to conduct the research were obtained from the ethics committee of the primary healthcare institution in Belgrade. ## 3. Results A total of 385 people were approached, of which about $90\%$ agreed to participate. The final sample consisted of 350 DMT2 participants. Participants were predominantly male ($55.4\%$) and had a mean age of 61.5 (SD = 10.5) years, ranging from 31 to 82 years. Participants’ data are presented in Table 1. ## Distributional Properties The mean scores for each domain of the FCCHL-SR12, BRIEF-4, and BRIEF-3 instruments are presented in Table 2. Regarding the FCCHL-SR12 instrument, the highest levels were for the IHL domain, and the lowest was for the FHL domain. Higher levels were found for BRIEF-4 instruments in comparison to the BRIEF-3 instrument. Items in all instruments showed no skewness or kurtosis in the distribution of scores. In BRIEF-4 and BRIEF-3 instruments, kurtosis was negative and indicated the small outliers in a distribution. The distributions of scores for FHL, IHL, CHL, FCCHL-SR12, BRIEF-4, and BRIEF-3 are presented in Table 2. The relationship between the total FCCHL-SR12 score and those of BRIEF instruments was investigated. The weak correlation was shown for total FCCHL-SR12 with BRIEF-3 ($r = 0.204$, $p \leq 0.01$) and BRIEF-4 ($r = 0.190$, $p \leq 0.01$). Both BRIEF instruments measure FHL, so we evaluated the association between BRIEFs and FCCHL-SR12 FHL domain. Neither the BRIEF-4 nor the BRIEF-3 instrument demonstrated a good correlation with the FHL domain ($r = 0.034$, $$p \leq 0.526$$ and 0.037, $$p \leq 0.490$$, respectively). Figure 1 shows the distribution of InHL, MHL, and AHL as assessed by different instruments and the FHL domain of FCCHL-SR12. Concerning the level of knowledge measured using different instruments, there is a significant difference between FCCHL-SR12 and BRIEF-3 ($$p \leq 0.003$$), in contrast to FCCHL-SR12 and BRIEF-4, where no statistically significant difference was observed ($$p \leq 0.192$$). The proportion of patients with InHL is approximately similar, but there are variations in the assessment of MHL and AHL. The difference in AHL ranges from $3.6\%$ for FCCHL-SR12 to $14.8\%$ for BRIEF-3. Additionally, the difference in HL levels between FHL of FCCHL-SR12 and both BRIEF-3 and BRIEF-4 was significant ($$p \leq 0.006$$ and 0.008, respectively). Since all three instruments similarly assess InHL, agreement in HL levels determined by BRIEFs and FCCHL-SR12 questionnaires was observed. The results are shown as a heat map for a cross table (Figure 2). Even though a similar number of participants have InHL for both instruments, 116 ($33.3\%$) (FCCHL-SR12) vs. 118 ($33.8\%$) (BRIEF-4) (Figure 1); only 49 of them were classified to have InHL with both instruments (Figure 2a). Using BRIEF-3 and FCCHL-SR12 instruments, only 55 participants were concurrently ranked in the InHL group (Figure 2b). BRIEF-3 and BRIEF-4 overestimate InHL in 12 patients and underestimate AHL in 4 patients evaluated by FCCHL-SR12. Table 3 shows the distribution of HL levels concerning the participants’ sociodemographic characteristics. All instruments identified the dependence of InHL on education level, exercise, and alcohol consumption. InHL was more prevalent in less-educated patients, in those who exercised rarely, and those who often consumed alcohol. HL levels also depended on gender, number of children, employment status, and interest in health as determined by BRIEF instruments. InHL were the most prevalent in males, in participants with two or more children, unemployed and pensioners and in participants who were not interested in health. However, when FCCHL-SR12 was used, HL literacy levels depended on the type of therapy, frequency of drug administration, and smoking status. InHL was more prevalent in patients who used diet and drugs as a therapy, who less frequently administrated drugs, and who were smokers. Furthermore, considering that all three instruments identify persons with HL in a similar percentage but the number of participants who were classified in the same way was small, the participants were divided into those with InHL and those with MHL and AHL to examine the predictors for InHL. Sociodemographic characteristics of participants (gender, marital status, children, education, employment, income, therapy, frequency of administration, health behaviors (exercise, alcohol, smoking), access to health-related information, and empowerment-related indicators (interest in health and self-estimation of health status)) were used as predictors of InHL. Predictors of InHL assessed by BRIEF-4 or BRIEF-3 were age, male gender, education level, employment status, and number of children. Probability of InHL increased with older age, number of children, and if participants were unemployed, contrary to female gender and high education, which decreased it. If InHL was assessed by BRIEF-3, additional predictors were exercise level and smoking status, reducing the probability of IHL. In the case of FCCHL-SR12—predictors were education level, smoking status, alcohol consumption, type of therapy, and frequency of drug administration. Data are shown in Table 4. Additionally, all significant predictors were included in multivariate analysis to assess independent predictors of InHL. Education was a significant independent predictor of InHL level for all three instruments. High education was associated with a lower probability of InHL. If the independent predictors of BRIEF-3 and BRIEF-4 are compared, it can be seen that the common predictors (except education) were age. Additionally, the number of children is an independent predictor for BRIEF-3. A higher number of children and older age were associated with a higher probability of InHL. Alcohol was an independent predictor for FCCHL-SR12. Lower consumption of alcohol is associated with a lower probability of InHL levels. ## 4.1. Health Literacy Levels Results showed that using the FCCHL-SR12 instrument among the DMT2 patients in Serbia, HL scored the best on IHL, followed by CHL and FHL, providing support for Nutbeam’s model on the three types of HL and levels of complexity. Patients in Serbia may be more functionally illiterate, but given that they have to learn to manage their illness, they are better at communication literacy. These results were not in accordance with [17,32,33], who found CHL to be the most difficult. As an explanation as to why CHL was not the most difficult among the participants in Serbia is that CHL comprises more advanced cognitive skills, but chronic patients may have developed self-management skills to manage the therapy and the disease. Perhaps, some issues are already defined in DM because there are treatment protocols, etc., have an influence. Considering that all patients who met the criteria were offered an opportunity to participate, even those with low literacy (e.g., FHL, who would have refused if it was electronic or post-survey), they had the same chance to be included in this type of study design. This is the explanation as to why in the sample we have many patients with low FHL. However, it is obvious that these people are either well cared for by their caregivers or that the patients are well trained to manage the disease and have developed IHL and CHL. In addition, the FCCHL scale is validated for a specific group of patients in Serbia—for diabetics—so the question is whether the same results would be obtained when measuring HL in the general population. With the use all three instruments, the proportion of patients with InHL was similar (FCCHL-SR12 detects HL similarly to BRIEF-3 and in some percentages more than BRIEF-4), but there are variations in the assessment of AHL (it ranges from $3.6\%$ for FCCHL-SR12 to $14.8\%$ for BRIEF-3) and MHL. In addition, a very weak correlation was shown between FCCHL-SR12 and both BRIEFs. FCCHL-SR12 detects a very small percentage of adequate HL. A very weak correlation between BRIEF-3 and FCCHL-SR12 was confirmed by the poor agreement of the instruments. Unlike other studies investigating FCCHL [17, 32, 33], higher correlation was found between IHL and CHL than between FHL and those two subscales. In this research, most patients had MHL ($63.3\%$, $53.0\%$, and $48.3\%$ measured using FCCHL-SR12, BRIEF-4, and BRIEF-3, respectively) with a broader range of AHL (from $3.4\%$ for FCCHL-SR12 to $14.83\%$ for BRIEF-3). Some previous studies that evaluated HL levels measured using different instruments in the same population showed that a low proportion of patients had AHL levels, with a reported prevalence ranging from $15\%$ to $40\%$. Many of these studies were performed in developed western countries (the USA and the UK) [29,34,35]. However, there has been limited explanation of the observed differences in the prevalence, and there was no effort to look at this problem globally. The proportion of patients with InHL in our study was from $33\%$ to $37\%$, and it was similar to some studies conducted in the US ($32.8\%$, $26.3\%$, $37.2\%$), Brazil ($26.7\%$), and the Marshall Islands ($24\%$) [36]. The study with the highest prevalence of InHL ($82\%$) was conducted in 2012 in Taiwan [37], and the lowest prevalence of InHL ($7.3\%$) was reported in 2011 in Switzerland [38] among DMT2 patients. It is found that the there is a need for countries to measure the burden of InHL in DMT2 patients using one standardized tool. A standardized method of measuring HL would allow for a direct comparison of findings between countries [39]. ## 4.2. Patients Characteristics and Predictors for InHL After investigation of HL across levels of social and demographic characteristics of the participants, health-related issues, health behaviors, and empowerment-related indicators, it is shown that DMT2 patients with higher education, males, the employed, and patients interested in their health had significantly higher HL than their counterparts. This finding was valid when applying the FCCHL-SR12, BRIEF-3, and BRIEF-4 instruments. The absence of an association between HL and gender was found in the research from Finbråten et al. [ 40], while in research from Hussein et al. [ 41], InHL was more likely to be higher in females. A significant difference was found in HL with regards to age with the BRIEF-3 and BRIEF-4 instruments. These results are in accordance with Al Sayah et al. [ 42], van der Heide et al. [ 19], and Hussein et al. [ 41]. Hence, higher age indicates lower HL level, which is in contrast with Al Sayah et al. [ 25] and Vandenbosch et al. [ 43] and in line with the findings from Heijmans et al. [ 17]. Furthermore, with the BRIEF-3 and BRIEF-4 instruments, age is shown as an independent predictor. The reasons for not detecting those differences by FCCHL-SR 12 might be related to the multidimensionality of the instrument rather than the measurement scope. Regardless of the instruments used, significant differences in HL were found in relation to education. People with DMT2 who had completed a university-level education reported a significantly higher HL than those with secondary school and lower educational levels. This is in line with the findings from Heijmans et al. [ 17], van der Heide et al. [ 19], Hussein et al. [ 41], Vandenbosch et al. [ 4], Berkman et al. [ 44], and Jeppesen et al. [ 45], who used different instruments; and those that used FCCHL to measure HL in DMT2 population were Nutbeam [11], Al Sayah et al. [ 25], and Finbråten et al. [ 40]. However, the average age of the sample was relatively high, so differences related to age may not have been evident. In the research conducted by Abdullah et al. in Malaysia, there was no significant association between educational level and HL [46]. Higher ages and lower education are in direct correlation with lower capacity of people to make sound decisions in the context of their everyday life; their ability to protect, maintain, and increase control over their illness and health is diminished. Poor health and worse health outcomes are consistently found among patients with more complex care need; these findings highlight the potential role of HL in this relationship. Differing from other research from Jeppesen et al. [ 45] and Finbråten et al. [ 40], an association between HL and health behaviors (alcohol consumption and smoking habits) was found with FCCHL-SR12 and confirmed by the combination of BRIEF-3 and FCCHL-SR12. Recommendations about smoking and alcohol consumption have been promoted among people with diabetes, and therefore, information on smoking and alcohol risk might be easier to understand regardless of HL level compared to other health behaviors. Number of children is an independent predictor for the abbreviated BRIEF, and it is associated with a higher probability of inadequate FHL, as parents with two or more children reported lower levels of FHL. The study conducted among the parents of preschool children in Serbia reported higher total pharmacotherapy scores (PTHL) among higher families using the PTHL-SR instrument [47]. In the research conducted in 2011, the number of children was associated with a lower probability for InHL [48]. The reasons for these differences are not clear and point out the need for future research. It may be that different results depending on yet-to-be elucidated factors, such as other parental characteristics. In chronic diseases such as diabetes, in which the patients should provide their own diabetes management (compliance with medication regimen, diet, blood glucose level measurement, insulin administration, foot care, etc.), the individual must be informed about HbA1c and its value as one of the most important metabolic markers in diabetes. The value of HbA1c being above the targeted value shows insufficient compliance with the treatment and care. We found InHL among patients with HbA1c>7, assessed using FCCHL-SR12. InHL leads to poorer diabetes self-care management skills, which may affect the control of HbA1c levels. To ensure sufficient diabetes self-care management skills, it is necessary for patients with diabetes to possess a high level of HL. These skills are needed in day-to-day decision-making, such as when measuring blood sugar levels, and patients with diabetes need to respond with the appropriate action for the reading they receive. Healthcare providers can improve the self-care management skills of patients with diabetes by enhancing their health literacy level through educational means, both face-to-face and through written information. This finding addresses the importance of not only treating the individual’s disease but also assessing and strengthening their health literacy level. Al Sayah et al. [ 42], Finbråten et al. [ 40], and Lee et al. [ 48] did not find any significant difference in HL related to HbA1c using the same FCCHL instrument. However, Finbråten et al. found significant differences in HL between those patients reporting and those not reporting their latest HbA1c levels [40]. Some previous studies reported inconsistent results of possible association between HL and HbA1C, investigated by Bailey et al. [ 12], Haun et al. [ 29], Franzen et al. [ 38], Al Sayah et al. [ 42], Jeppesen et al. [ 45], and Bains et al. [ 49]. The reason for that inconsistency might be in a fact that different instruments were used in those studies—mostly those that are limited to FHL. With the FCCHL-SR12 instrument, DMT2 patients who use insulin had a higher level of HL in comparison with those on tablets, which is in line with an expectation that they have higher knowledge and are engaged more actively in therapy management. ## 4.3. Strengths and Weakness In Serbia, studies on HL are insufficient, especially in the primary healthcare sector. To our knowledge, this is the first study that evaluates HL with instruments which are validated and have satisfactory validity and reliability, so that they can be used in population research. The strength of the study is that it was the first study to assess patients’ health literacy in DMT2 in primary healthcare, which accounts for $80\%$ of health-related decisions. Therefore, the findings from this study did not only reveal levels of HL literacy among this patient population but also suggested recommendations for healthcare professionals. Although the sample size was interpreted as sufficient for discussions and drawing sufficiently trustworthy conclusions, some limitations must be acknowledged. Data were collected using self-administrated measurements, which could be quite challenging for those people with InHL, especially those with low FHL, as the items require reading comprehension abilities. It is necessary to consider the possibility that those who chose to participate in this study had a higher HL. However, the recruitment approach and the pandemic measures could have assured the inclusion of individuals with very low HL levels. Patients who would refuse to participate in postal survey or online survey or who are illiterate or have problems with understanding the content of the instrument may have participated. This bias could have led to recruiting more individuals with low levels of HL, so the reported prevalence could be an overestimation of the true prevalence in this population. Due to the COVID pandemic and special measures issued on the doctors’ visits, some participants preferred that researchers filled out the survey instruments. In non-pandemic circumstances, those people might refuse to self-administer the instrument. In addition, the instruments were quite lengthy and could have been quite stressful and/or time-consuming to fill out, which also might be a reason for non-responses or drop-outs, if the research would have been done in non-pandemic circumstances. Still, during the validation process of the FCCHL-SR12, all participants (regardless of educational level) reported that the items were clearly stated. The study was conducted at a single healthcare center at the capital of Belgrade, which could limit the generalization of the results. This is a typical primary healthcare institution for those patients, and it fully represented a demographically diverse population. The adequate age and gender distribution of the sample reflected the targeted population well and could be considered representative of elderly people of DMT2 in the country [50]. The mean age of people with DMT2 in this study was relatively high (61.5 years). In Serbia, the incidence of DMT2 is significantly higher in the elderly and in men. According to the Statistical Office of the Republic of Serbia, in 2021, observed by gender, $51.3\%$ were women and $47.7\%$ were men. At the same time, the process of demographic aging of the population is manifested as low participation of young people and a high and continuously growing share of the elderly in the total; according to the data, the share of people over 65 years of age is $21.3\%$ [49]. However, the educational level in the sample was higher than that of the general Serbian population; according to the 2011 census in Serbia, $16.2\%$ of inhabitants have higher education, $49\%$ have a secondary education, $20.7\%$ have an elementary education, and $13.7\%$ have not completed elementary education [51]. Another limitation of the study is related to the cross-sectional design, making it impossible to discuss the role of causality in the associations found between predictors of HL levels and patients’ characteristics. Prospective study designs could more accurately describe the relationship between the two. Because the study did not use a randomized, controlled design, it may fail to account for confounding variables that introduced measurement errors. Additionally, three perception-based measures were included, and a performance-based measurement of HL was not, making interpretation challenging. ## 5. Conclusions Beyond previous measures focusing only on FHL (BRIEFs instruments), FCCHL-SR12 includes three levels of HL, each of which might have different effects on patient outcomes. BRIEF instruments appeared to point in the same direction as FCCHL-SR12. Both BRIEF and FCCHL-SR12 can be easily administered in a healthcare setting. Among primary care DMT2 patients in Serbia, MHL was dominant, with a low proportion of patients with AHL. Our findings indicate that parental status with more children and greater frequency of alcohol use are predictors of InHL and confirm those of previous studies showing that InHL is also associated with lower educational level and higher ages. However, by applying all three instruments, we found only education as an independent predictor. Future investigation is necessary to support those findings on a larger population and to improve our understanding of factors that could help primary care practitioners in the future to more easily identify patients who need help with the management and application of information for better control of their disease. ## References 1. 1. International Diabetes Federation IDF Diabetes Atlas9th ed.International Diabetes FederationBrussels, Belgium2019Available online: https://www.diabetesatlas.org(accessed on 2 January 2023). IDF Diabetes Atlas (2019.0) 2. 2. Institute for Public Health of Serbia “Dr. Milan Jovanović Batut” Results of the 2019 Population Health Survey in SerbiaInstitute for Public Health of Serbia “Dr. Milan Jovanović Batut”Belgrade, Serbia2019. Results of the 2019 Population Health Survey in Serbia (2019.0) 3. Ylihärsilä H., Lindström J., Eriksson J.G., Jousilahti P., Valle T.T., Sundvall J., Tuomilehto J.. Prevalence of Diabetes and impaired glucose regulation in 45- to 64-year-old individuals in three areas of Finland. Diabet. Med. (2005.0) 22 88-91. DOI: 10.1111/j.1464-5491.2005.01420.x 4. Ford E.S.. Risks for all-cause mortality, cardiovascular disease, and diabetes associated with the metabolic syndrome: A summary of the evidence. Diabetes Care (2005.0) 28 1769-1778. DOI: 10.2337/diacare.28.7.1769 5. McEwan P., Williams J.E., Griffiths J.D., Bagust A., Peters J.R., Hopkinson P., Currie C.J.. Evaluating the performance of the Framingham risk equations in a population with diabetes. Diabet. Med (2004.0) 21 318-323. DOI: 10.1111/j.1464-5491.2004.01139.x 6. 6. Institute for Public Health of Serbia Health and Statistical Yearbook of the Republic of SerbiaInstitute for Public Health of SerbiaBelgrade, Serbia2020. Health and Statistical Yearbook of the Republic of Serbia (2020.0) 7. Atanasković-Marković Z., Bjegović V., Janković S., Kocev N., Laaser U., Marinković J., Marković-Denić L., Pejin-Stokić L., Penev G., Stanisavljević D.. The Burden of Disease and Injury in Serbia (2003.0) 8. National Curriculum for Nursing Education. (2008.0) 9. The Coordination Reform. Report to the Storting no. 47 (Norwegian Parliament). (2009.0) 10. National Health and Care Service Plan. Report to the Storting no. 10 (Norwegian Parliament). (2011.0) 11. Nutbeam D.. Discussion Paper on Promoting, Measuring and Implementing Health Literacy—Implications for Policy and Practice in Non-Communicable Disease Prevention and Control (WHO GCM/NCD Working Group 3.3 2017) 12. Bailey S.C., Brega A.G., Crutchfield T.M., Elasy T., Herr H., Kaphingst K., Karter A.J., Moreland-Russell S., Osborn C.Y., Pignone M.. Update on Health Literacy and Diabetes. Diabetes Educ. (2014.0) 40 581-604. DOI: 10.1177/0145721714540220 13. Baur C., Harris L., Squire E.. The US National action plan to improve health literacy: A model for positive organizational change. Stud. Health Technol. Inform. (2017.0) 240 186-202. PMID: 28972518 14. Dewalt D.A., Berkman N.D., Sheridan S., Lohr K.N., Pignone M.P.. Literacy and health outcomes: A systematic review of the literature. J. Gen. Intern. Med. (2004.0) 19 1228-1239. DOI: 10.1111/j.1525-1497.2004.40153.x 15. Osborn C.Y., Cavanaugh K., Wallston K.A., White R.O., Rothman R.L.. Diabetes numeracy: An overlooked factor in understanding racial disparities in glycemic control. Diabetes Care (2009.0) 32 1614-1619. DOI: 10.2337/dc09-0425 16. Gazmararian J.A., Williams M.V., Peel J., Baker D.W.. Health literacy and knowledge of chronic disease. Patient Educ. Couns. (2003.0) 51 267-275. DOI: 10.1016/S0738-3991(02)00239-2 17. Heijmans M., Waverijn G., Rademakers J., van der Vaart R., Rijken M.. Functional, communicative and critical health literacy of chronic disease patients and their importance for self-management. Patient Educ. Couns. (2015.0) 98 41-48. DOI: 10.1016/j.pec.2014.10.006 18. Kim S., Love F., Quistberg D.A., Shea J.A.. Association of health literacy with self-management behavior in patients with diabetes. Diabetes Care (2004.0) 27 2980-2982. DOI: 10.2337/diacare.27.12.2980 19. Van der Heide I., Wang J., Droomers M., Spreeuwenberg P., Rademakers J., Uiters E.. The relationship between health, education, and health literacy: Results from the Dutch Adult Literacy and Life Skills Survey. J. Health Commun. (2013.0) 18 172-184. DOI: 10.1080/10810730.2013.825668 20. Cavanaugh K., Huizinga M.M., Wallston K.A., Gebretsadik T., Shintani A., Davis D., Gregory R.P., Fuchs L., Malone R., Cherrington A.. Association of numeracy and diabetes control. Ann. Intern. Med. (2008.0) 148 737-746. DOI: 10.7326/0003-4819-148-10-200805200-00006 21. Tang Y.H., Pang S.M., Chan M.F., Yeung G.S., Yeung V.T.. Health literacy, complication awareness, and diabetic control in patients with type 2 diabetes mellitus. J. Adv. Nurs. (2008.0) 62 74-83. DOI: 10.1111/j.1365-2648.2007.04526.x 22. Zeković M., Krajnović D., Marinković V., Tasić L.. The Complexity of Adherence Issue: A Review of its scope and determinants. Acta Med. Median. (2016.0) 55 51-57. DOI: 10.5633/amm.2016.0109 23. Osterberg L., Blaschke T.. Adherence to medication. New Engl. J. Med. (2005.0) 353 487-497. DOI: 10.1056/NEJMra050100 24. Miller H.N., Hill M., Kottke T.. The multi-level compliance challenge, recommendations for a call to action. J. Am. Heart Assoc. (1997.0) 77 308-314 25. Al Sayah F., Williams B., Johnson J.A.. Measuring health literacy in individuals with diabetes, a systematic review and evaluation of available measures. Health Educ. Behav. (2013.0) 40 42-55. DOI: 10.1177/1090198111436341 26. Levic M., Krajnovic D., Bogavac-Stanojevic N.. The Instruments Used to Assess Health Literacy and Pharmacotherapy Literacy of Diabetes Mellitus Type 2 Patients, A Scoping Review. Front. Public Health (2021.0) 9 1424-1440. DOI: 10.3389/fpubh.2021.747807 27. Sørensen K., Van den Broucke S., Fullam J., Doyle G., Pelikan J., Slonska Z., Brand H.. Health literacy and public health: A systematic review and integration of definitions and models. BMC Public Health (2012.0) 12. DOI: 10.1186/1471-2458-12-80 28. Wallace L.S., Rogers E.S., Roskos S.E., Holiday D.B., Weiss B.D.. Brief report, Screening items to identify patients with limited health literacy skills. J. Gen. Intern. Med. (2006.0) 21 874-877. DOI: 10.1111/j.1525-1497.2006.00532.x 29. Haun J., Luther S., Dodd V., Donaldson P.. Measurement variation across health literacy assessments, implications for assessment selection in research and practice. J. Health Commun. (2012.0) 3 141-159. DOI: 10.1080/10810730.2012.712615 30. Haun J., Noland-Dodd V., Varnes J., Graham-Pole J., Rienzo B., Donaldson P.. Testing the BRIEF Health Literacy Screening Tool. Fed Pract. (2009.0) 26 24 31. Levic M., Bogavac-Stanojevic N., Krajnovic D.. Cross-Cultural Adaptation and Validation of the Functional, Communicative and Critical Health Literacy Instrument (FCCHL-SR) for Diabetic Patients in Serbia. Healthcare (2022.0) 10. DOI: 10.3390/healthcare10091667 32. Ishikawa H., Takeuchi T., Yano E.. Measuring functional, communicative, and critical health literacy among diabetic patients. Diabetes Care (2008.0) 31 874-879. DOI: 10.2337/dc07-1932 33. Van der Vaart R., Drossaert C.H., Taal E., ten Klooster P.M., Hilderink-Koertshuis R.T., Klaase J.M., van de Laar M.A.. Validation of the Dutch functional, communicative and critical health literacy scales. Patient Educ. Couns. (2012.0) 89 82-88. DOI: 10.1016/j.pec.2012.07.014 34. Rothman R., Malone R., Bryant B., Horlen C., DeWalt D., Pignone M.. The Relationship Between Literacy and Glycemic Control in a Diabetes Disease-Management Program. Diabetes Educ. (2004.0) 30 263-273. DOI: 10.1177/014572170403000219 35. Schillinger D., Grumbach K., Piette J., Wang F., Osmond D., Daher C., Palacios J., Sullivan G.D., Bindman A.B.. Association of health literacy with diabetes outcomes. JAMA (2002.0) 288 475-482. DOI: 10.1001/jama.288.4.475 36. Abdullah A., Liew S.M., Salim H., Ng C.J., Chinna K.. Prevalence of limited health literacy among patients with type 2 diabetes mellitus: A systematic review. PLoS ONE (2019.0) 7. DOI: 10.1371/journal.pone.0216402 37. Chen G.-D., Huang C.-N., Yang Y.-S., Lew-Ting C.-Y.. Patient perception of understanding health education and instructions has moderating effect on glycemic control. BMC Public Health (2014.0) 14. DOI: 10.1186/1471-2458-14-683 38. Franzen J., Mantwill S., Rapold R., Schulz P.J.. The relationship between functional health literacy and the use of the health system by diabetics in Switzerland. Eur. J. Public Health (2014.0) 24 997-1003. DOI: 10.1093/eurpub/ckt202 39. Sørensen K., van den Broucke S., Pelikan J.M., Fullam J., Doyle G., Slonska Z., Kondilis B., Stoffels V., Osborne R.H., Brand H.. Measuring health literacy in populations, illuminating the design and development process of the European Health Literacy Survey Questionnaire (HLS-EU-Q). BMC Public Health (2013.0) 13. DOI: 10.1186/1471-2458-13-948 40. Finbråten H.S., Guttersrud Ø., Nordström G., Pettersen K.S., Trollvik A., Wilde-Larsson B.. Explaining variance in health literacy among people with type 2 diabetes, the association between health literacy and health behaviour and empowerment. BMC Public Health (2020.0) 20. DOI: 10.1186/s12889-020-8274-z 41. Hussein S.H., Almajran A., Albatineh A.N.. Prevalence of health literacy and its correlates among patients with type II diabetes in Kuwait, a population-based study. Diabetes Res. Clin. Pract. (2018.0) 141 118-125. DOI: 10.1016/j.diabres.2018.04.033 42. Al Sayah F., Majumdar S.R., Johnson J.A.. Association of Inadequate Health Literacy with Health Outcomes in Patients with Type 2 Diabetes and Depression, Secondary Analysis of a Controlled Trial. Can. J. Diabetes (2015.0) 39 259-265. DOI: 10.1016/j.jcjd.2014.11.005 43. VandenBosch J., Broucke S.V.D., Schinckus L., Schwarz P., Doyle G., Pelikan J., Muller I., Levin-Zamir D., Schillinger D., Chang P.. The impact of health literacy on diabetes self-management education. Health Educ. J. (2018.0) 77 349-362. DOI: 10.1177/0017896917751554 44. Berkman N.D., Sheridan S.L., Donahue K.E., Halpern D.J., Crotty K.. Low health literacy and health outcomes, an updated systematic review. Ann. Intern. Med. (2011.0) 155 97-107. DOI: 10.7326/0003-4819-155-2-201107190-00005 45. Jeppesen K.M., Coyle J.D., Miser W.F.. Screening questions to predict limited health literacy, a cross-sectional study of patients with diabetes mellitus. Ann. Fam. Med. (2009.0) 7 24-31. DOI: 10.1370/afm.919 46. Abdullah A., Ng C.J., Liew S.M., Ambigapathy S., Paranthaman V., Chinna K.. Prevalence of limited health literacy and its associated factors in patients with type 2 diabetes mellitus in Perak, Malaysia—A cross-sectional study. BMJ Open (2020.0) 10 039864. DOI: 10.1136/bmjopen-2020-039864 47. Ubavić S., Bogavac-Stanojević N., Jović-Vraneš A., Krajnović D.. Understanding of Information about Medicines Use among Parents of Pre-School Children in Serbia, Parental Pharmacotherapy Literacy Questionnaire (PTHL-SR). Int. J. Environ. Res. Public Health (2018.0) 15. DOI: 10.3390/ijerph15050977 48. Lee Y.-J., Shin S.-J., Wang R.-H., Lin K.-D., Lee Y.-L., Wang Y.-H.. Pathways of empowerment perceptions, health literacy, self-efficacy, and self-care behaviors to glycemic control in patients with type 2 diabetes mellitus. Patient Educ. Couns. (2016.0) 99 287-294. DOI: 10.1016/j.pec.2015.08.021 49. Bains S.S., Egede L.E.. Associations between Health Literacy, Diabetes Knowledge, Self-Care Behaviors, and Glycemic Control in a Low-Income Population with Type 2 Diabetes. Diabetes Technol. Ther. (2011.0) 13 335-341. DOI: 10.1089/dia.2010.0160 50. 50. Institute for Public Health of Serbia “Dr. Milan Jovanović Batut” Incidence and Mortality of Diabetes in SerbiaInstitute for Public Health of Serbia “Dr. Milan Jovanović Batut”Belgrade, Serbia2021. Incidence and Mortality of Diabetes in Serbia (2021.0) 51. National Health Survey of the Republic of Serbia. (2021.0)
--- title: A Comprehensive Analysis Revealing FBXW9 as a Potential Prognostic and Immunological Biomarker in Breast Cancer authors: - Shiyi Yu - Zhengyan Liang - Zhehao Fan - Binjie Cao - Ning Wang - Rui Wu - Haibo Sun journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049633 doi: 10.3390/ijms24065262 license: CC BY 4.0 --- # A Comprehensive Analysis Revealing FBXW9 as a Potential Prognostic and Immunological Biomarker in Breast Cancer ## Abstract The WD40 repeat-containing F-box proteins (FBXWs) family belongs to three major classes of F-box proteins. Consistent with the function of other F-box proteins, FBXWs are E3 ubiquitin ligases to mediate protease-dependent protein degradation. However, the roles of several FBXWs remain elusive. In the present study, via integrative analysis of transcriptome profiles from The Cancer Genome Atlas (TCGA) datasets, we found that FBXW9 was upregulated in the majority of cancer types, including breast cancer. FBXW expression was correlated with the prognosis of patients with various types of cancers, especially for FBXW4, 5, 9, and 10. Moreover, FBXWs were associated with infiltration of immune cells, and expression of FBXW9 was associated with poor prognosis of patients receiving anti-PD1 therapy. We predicted several substrates of FBXW9, and TP53 was the hub gene in the list. Downregulation of FBXW9 increased the expression of p21, a target of TP53, in breast cancer cells. FBXW9 was also strongly correlated with cancer cell stemness, and genes correlated with FBXW9 were associated with several MYC activities according to gene enrichment analysis in breast cancer. Cell-based assays showed that silencing of FBXW9 inhibited cell proliferation and cell cycle progression in breast cancer cells. Our study highlights the potential role of FBXW9 as a biomarker and promising target for patients with breast cancer. ## 1. Introduction Although advances have been achieved in understanding their mechanisms, cancers remain a major threat to humans and a serious public health problem worldwide [1]. Owing to the public easily accessed database, pan-cancer analysis has been performed on many proteins and non-coding RNAs [2,3], which could provide valuable information for cancer studies given that hallmarks are shared in cancers of various origins [4]. Ubiquitination is a crucial biological process for protein turnover [5,6]. This process is achieved via the cooperation of the E1 enzyme, E2 enzyme, and E3 ligase [7,8]. F-box proteins are the most well-characterized E3 ligases among hundreds [9]. Generally, F-box proteins specifically bind to phosphorylated substrates, leading to proteolytic ubiquitination (Lysine 48, Lysine 11) or non-proteolytic ubiquitination (Lysine 63) [10]. F-box proteins can be divided into three classes, namely FBXL (22 members), FBXO (37 members), and FBXW (10 members) [11]. For FBXWs, these proteins contain a WD-40 domain, which is pivotal for its recognition of substrates [12]. Several members of FBXWs have been recently identified as oncogenes or tumor suppressors according to cell background [13,14,15]. For example, FBXW7 mediated the ubiquitination of BGN to regulate peritoneal metastasis of gastric cancer [13], while the elevation of FBXW4 was observed in acute myeloid leukemia and its expression was associated with poor prognosis of patients [14]. FBXW1 was involved in the maintenance of cancer stem cells in glioblastoma via targeting GLI2 [15]. However, the roles and functions of most members of FBXWs have not been extensively examined in cancers. In the current study, we performed a comprehensive bioinformatic analysis of FBXWs in several large cohorts covering most cancer types. Our analysis reveals the expression pattern and potential roles of FBXWs in the prognosis, stemness, and immune infiltration of cancers. The potential pathways and targets of FBXW9 were also studied. We also performed cell-based assays to validate the pro-cancer function of FBXW9 in breast cancer cells. ## 2.1. Analysis of mRNA Levels of FBXWs in Tumors of Various Cancer Types and Normal Tissues All members of FBXWs contain an F-box domain and multiple WD40 domains (Figure 1A). The mRNA levels of all 10 members of FBXWs were retrieved from TCGA datasets including tumors of 17 cancer types and normal tissues. The expression of most FBXWs was positively correlated with the others in all cancers (Figure S1). All FBXWs were dysregulated in at least one cancer type (Figure 1B). Of all FBXW members, FBXW9 was the most frequently upregulated across 17 cancer types ($\frac{14}{17}$), while FBXW11 was the most frequently downregulated gene in cancers ($\frac{10}{17}$) (Figure 1B). Noticeably, the most significant upregulation of FBXW9 was observed in invasive breast carcinoma (BRCA) ($$p \leq 9.10$$ × 10−19), followed by colon adenocarcinoma (COAD) ($$p \leq 2.30$$ × 10−18), bladder urothelial carcinoma (BLCA) ($$p \leq 7.60$$ × 10−10), and 11 other cancer types ($p \leq 0.001$ for all) (Figure 1C). In contrast, downregulation of FBXW11 was observed in BLCA, BRCA, kidney chromophobe (KICH), and 7 other cancer types (Figure 1D). To examine the protein expression of FBXW9 in tumors and normal tissues, we analyzed immunohistochemistry (IHC) data in the Human Protein Atlas database. FBXW9 protein was not detected in normal tissues (urinary bladder, breast, and cervix). In contrast, low or medium staining of FBXW9 protein was observed in tumors of corresponding types (BLCA, BRCA, and cervical squamous cell carcinoma and endocervical adenocarcinoma (CESC)) (Figure 1E). In contrast, IHC staining of FBXW11 showed that FBXW11 was expressed in the urinary bladder, breast, and lung, but was not expressed in BLCA, BRCA, and lung adenocarcinoma (LUAD) (Figure 1F). We further evaluated FBXW9 mRNA expression in our collected breast tumors, which also showed upregulation of FBXW9 mRNA in tumors compared with normal tissues (Figure 1G). ## 2.2. Analysis of the Association between mRNA Levels of FBXWs and the Prognosis of Patients with Cancer To study the impact of FBXWs on the prognosis of patients, we collected overall survival data of patients with various cancer types and evaluated the correlation between the expression of FBXWs and prognosis. Based on the TIMER 2.0 database, several cancer types were further divided into subtypes. For example, invasive breast cancer was divided into Basal, Luminal A, Luminal B, and Her2, which gave rise to 41 cancer types. All mRNAs of FBXWs were associated with altered prognosis of cancers of at least four types (Figure 2A,B). To our interest, the expression of FBXW9 was associated with prognosis alternation of many cancer types ($\frac{14}{41}$), followed by FBXW5 ($\frac{10}{41}$) and FBXW4 ($\frac{9}{41}$) (Figure 2A,B). In detail, FBXW9 was favorable ($\frac{7}{41}$) and hazardous ($\frac{7}{41}$) to the overall survivals of patients with different cancers, indicating FBXW9 is a double-edged sword in cancers (Figure 2A,B). FBXW4 and FBXW5 were associated with decreased risk of most cancer types examined ($\frac{7}{41}$). FBXW10, however, was consistently associated with poor prognosis in seven cancer types with significance (Figure 2A,B). Inherently, in BRCA-Basal and BLCA, high expression of FBXW9 predicted poor overall survival with hazard ratios (HR) of 1.88 and 1.53, respectively (Figure 2C). In contrast, high expression of FBXW9 was associated with good prognosis in rectum adenocarcinoma (READ) (HR = 0.36) and kidney renal papillary cell carcinoma (KIRP) (HR = 0.36) (Figure 2C). Overall, these results indicate the important and complicated roles of FBXWs in different cancers. ## 2.3. Correlation of FBXW Expression and Immune Infiltration across Multiple Cancer Types Recent studies have shown that several proteins can regulate the function of immune cells to support or suppress cancer progression [16]. We, therefore, analyzed the prognostic value of FBXWs in patients receiving anti-PD1 therapy. In the pan-cancer cohort, high expression of FBXW1, FBXW4, and FBXW7 was strongly associated with prolonged overall survival time in these patients, while FBXW8, FBXW9, and FBXW11 showed the opposite effect (Figure 3). Next, we evaluated the correlation between expression of FBXWs with immune cells in the tumor microenvironment. Overall, FBXWs were positively correlated with the infiltration of all types of immune cells (Figure 4A,B). Specifically, FBXW2, FBXW5, and FBXW12 were mainly negatively associated with the infiltration of immune cells, while FBXW1, FBXW4, FBXW7, and FBXW11 were dominantly positively correlated with the infiltration of immune cells (Figure 4A,B). The impact of FBXW8 and FBXW9 on immune cells was dependent on the cancer context and type of immune cells (Figure 4B,C). For example, in BRCA, FBXW9 was significantly negatively correlated with B cells, CD8+ T cells, macrophages, neutrophils, and dendritic cells while positively correlated with CD4+ T cells (Figure 4C). To investigate the immunosuppressors relevant to FBXW9, the expression of several known immunosuppressors (NECTIN2, CD274, PDCD1LG2) was detected in breast cancer cells [17]. In SUM159 and MDA-MB-231 cells, we showed that the knockdown of FBXW9 decreased NECTIN2 mRNA levels but not CD274 or PDCD1LG2 (Figure 4D,E). We also searched for a correlation between mRNA levels of FBXWs and stroma scores in breast cancer using the CPTAC dataset. mRNA levels of FBXW9 and FBXW4 were negatively associated with stroma score ($p \leq 0.001$) in breast cancer (Figure 4F). The data implied that FBXWs might be involved in the response to immunotherapy via alteration of the tumor microenvironment. ## 2.4. Analysis of Correlation between FBXW Expression and Stemness Score Cancer stem cells are a small group of cells mediating initiation and drug resistance in cancer cells [18]. To define the association between FBXWs and the stemness of cancers, we collected the machine learning-based stemness score of each sample in the CPTAC dataset. We found that four FBXWs (FBXW-1, 8, 11, and 12) were negatively correlated with stemness score while two FBXWs (FBXW-4 and 9) were positively correlated with stemness score (Figure 5A,B). Since MYC activity was one of the most well-characterized contributors to the stem cell trait in cancer cells [19], we acquired the MYC activity score based on the RABIT transcription factor regulatory impact of each sample from the TCGA-BRCA dataset. Different from the stemness score, four FBXWs (FBXW1, 4, 8, and 9) were positively correlated with the MYC activity while two FBXWs (FBXW7 and 10) were negatively correlated with the MYC activity (Figure 5C). After overlap, no FBXW was both negatively correlated with the stemness score and the MYC activity. Inherently, FBXW9 was positively correlated with both the stemness score and the MYC activity (Figure 5D). In detail, the association between FBXW9 expression and stemness score, and FBXW9 expression and MYC activity were plotted (Figure 5E,F). Due to the critical role of cancer stem cells in mediating chemotherapy resistance, we next analyzed the association between FBXW9 expression and relapse-free survival (RFS) of patients with breast cancer. As expected, patients with high expression of FBXW9 showed significantly shorter RFS compared with those with low expression of FBXW9 (Figure 5G). These data further showed that FBXW9 was the potential contributor to stemness in cancer cells. ## 2.5. Analysis of Potential Substrates of FBXW9 As an E3 ligase, FBXW9 might exert its function by mediating the ubiquitination of substrate proteins. To explore the targets of FBXW9, we used Ubibrowser to perform structure-based predictions. A total of 36 proteins were predicted as potential substrates of FBXW9, and TP53 was the hub gene (Figure 6A). Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis indicated that these proteins were involved in cancer-related processes including cell cycle and transcriptional misregulation in cancer. To our interest, the involvement of basal cell carcinoma (Figure 6B) was consistent with its potential role in basal breast cancer as aforementioned. Gene ontology (GO) analysis further showed that these proteins were mainly localized in the nucleus and were involved in transcription and kinase activities (Figure 6C). As TP53 was the hub gene of putative substrates, we next detected the mRNA expression of p21, a well-characterized target of the TP53 pathway involved in cell cycle arrest [20], and two cell cycle drivers (CCNA2, CCNB1) in breast cancer cells. Indeed, the knockdown of FBXW9 increased p21 mRNA expression while mRNA levels of CCNA2 and CCNB1 were decreased in breast cancer cells (Figure 6D). Moreover, we confirmed the protein expression of p21 was increased in breast cancer cells upon FBXW9 silencing (Figure 6E). The data implied a TP53-regulated cell cycle progression was critical for the function of FBXW9, at least in breast cancer cells. ## 2.6. Establishment of Signaling Network with Co-Expressed Genes of FBXW9 To explore the signaling network regulated by FBXW9, we ranked cases in CPTAC according to the protein expression of FBXW9 and picked 10 cases with the highest expression of FBXW9 and 10 cases with the lowest expression of FBXW9. DESeq2 was applied to discover the differentially expressed genes between these two groups. A total of 834 upregulated genes and 2285 downregulated genes were distributed in the FBXW9 high-expression group compared with the counterpart (Figure 7A). Using transcription factor protein–protein interaction networks (PPIs), it was revealed that these genes were targets of well-known oncogenic transcription factors such as ESR1 and MYC (Figure 7B). As MYC is a well-known target of TP53 [21], the transcriptome data implied an FBXW9/TP53/MYC axis in breast cancer. KEGG analysis further indicated that these genes were involved in many oncogenic pathways such as the MAPK signaling pathway and insulin signaling pathway (Figure 8A), which are attributed to several cancer types such as breast cancer, non-small cell lung cancer, colorectal cancer, and hepatocellular carcinoma (Figure 8A). Again, GO analysis indicated that these genes were involved in protein ubiquitination and kinase activities (Figure 8B). Moreover, we used Metascape to establish the signaling network regulated by FBXW9. The plot showed that these differentially expressed genes were involved in several pivotal biological processes for cell migration and proliferation, such as the regulation of cytoskeleton organization, and regulation of mitotic nuclear division (Figure 8C). The signaling network analysis manifested that FBXW9 was potentially involved in cancer progression, particularly in breast cancer. ## 2.7. FBXW9 Silence Inhibited Cell Proliferation and Cell Cycle Progression in Breast Cancer Cells As evidenced by the bioinformatic data, FBXW9 might contribute to cancer progression including breast cancer. Cell proliferation and colony-forming assays showed that FBXW9 silence notably inhibited cell proliferation (Figure 9A) and decreased the number of cell colonies (Figure 9B) in SUM159 and MDA-MB-231 cells. Flow cytometry analysis further revealed that FBXW9 silence increased the proportion of cells accumulated in the G0/G1 phase, and decreased the proportion of cells in the S phase, indicating alteration of cell cycle distribution (Figure 9C). These data partly confirmed the oncogenic role of FBXW9 in breast cancer. ## 3. Discussion An imbalance of protein degradation and synthesis leads to the accumulation of oncoproteins or loss of tumor-suppressive proteins, resulting in the initiation and progression of cancers [22]. As a subfamily of E3 ligases, all members of FBXWs contain an F-box motif for forming the SCF complex and several WD40 motifs for interaction with substrates [23]. In addition to these two critical domains, FBXW1 and FBXW11 also contain a D domain for dimerization [24]. Compelling evidence suggested that FBXW1 and FBXW2 were involved in cancer development [25,26]. However, the potential roles of several FBXWs remain elusive. In the present study, we comprehensively analyzed FBXWs in a variety of cancer types. Consistent with the well-characterized anti-tumor function of FBXW4 [27], our analysis also showed that the expression of FBXW4 was decreased in several cancer types, and was associated with a good prognosis of patients. To provide novel insights, the current analysis also revealed prevalent upregulation of FBXW9 and downregulation of FBXW11 in multiple cancer types. Compared with other members, the expression of FBXW9 was associated with overall survival in more cancer types. These data suggested that our analysis could provide profiled information for the expression and prognostic roles of FBXWs in cancers, and implied certain members of FBXWs such as FBXW9 to be potential biomarkers in human cancers. Interestingly, the current pan-cancer analysis also provided novel findings uncovering the upregulation of FBXW9 in the majority of cancer types examined, as well as the downregulation of FBXW11 in multiple cancer types. FBXW9 is located in 19p13.13 with 10 exons. A previous study showed that FBXW9 was involved in synaptic transmission in C. elegans with an unknown mechanism [28]. Unlike other members of FBXWs, the function and substrates of FBXW9 have not been studied in human diseases yet. According to the protein sequence and structure, we predicted several targets of FBXW9. These putative targets were relevant to TP53 signaling and involved in cell cycle processes, such as TP53, LATS2, and CDKN1C. Via analysis of transcriptome data of breast cancer, FBXW9 was related to the activities of several cell cycle regulators including MYC and ESR1. Interestingly, MYC activity was also regulated by TP53 [21]. In breast cancer cells, we confirmed that the knockdown of FBXW9 inhibited cell proliferation and induced the arrest of cell cycle progression. Members of FBXWs showed distinct roles in the prediction of prognosis in patients receiving anti-PD1 therapy, suggesting FBXWs’ different functions in the tumor microenvironment. The percentage of immune cells and non-immune stroma cells infiltrated into tumors could reflect the sensitivity of tumor cells to immune therapy [29]. Generally, infiltration of T cells favors the survival of patients, while stroma cells, in contrast, are linked to poor prognosis of patients [17]. In the Clinical Proteomic Tumor Analysis Consortium (CPTAC) dataset, FBXW4, 7, and 9 were associated with stroma infiltration. Of them, FBXW4 and 9 were strongly negatively correlated with stroma infiltration. For immune cells, FBXWs could be divided into three groups according to their correlation with the infiltration of immune cells. Consistent with the pro-survival role in patients receiving anti-PD1 therapy, FBXW1, FBXW4, and FBXW7 were mainly positively correlated with the infiltration. FBXW9, in contrast, was associated with a poor prognosis in patients receiving anti-PD1 therapy. We experimentally found that FBXW9 repressed the expression of NECTIN2, a regulator of the tumor microenvironment [17], in breast cancer cells. Cancer stem cells are pivotal for cancer initiation and drug resistance [30]. According to data from breast cancer, it was found that most FBXWs were negatively correlated with stemness except FBXW5 and 9, which were positively correlated with cancer stemness. Consistent with our findings, FBXW1 was verified as an E3 ligase mediating TAZ/YAP degradation, and, thus, contributed to the initiation property and drug resistance of cancer cells [31,32]. Additionally, FBXW9 was positively correlated with the activity of MYC signaling according to RABIT transcription factor regulatory impact, which is a well-known contributor to cancer stemness [33]. This was further validated by analyzing differentially expressed genes in the FBXW9 high-expression group. The data support a potential role of FBXW9 in breast cancer stem cells that needs further experiments for validation. ## 4.1. Pan-Cancer FBXW Expression Analysis The TIMER database (https://cistrome.shinyapps.io/timer/, access date: 10 February 2023) was used to explore the transcript expression of FBXWs in cancers from TCGA datasets. The expression pattern of FBXWs was depicted on a heatmap composed of examined cancer types listed in Abbreviation. Pearson correlation analysis was used to analyze the relation between FBXWs across cancer types. The immunohistochemistry data of FBXW9 and FBXW11 in BLCA, BRCA, CESC, LUAD, and corresponding normal tissues were retrieved from the Human Protein Atlas database (www.proteinatlas.org, access date: 10 January 2023). ## 4.2. Pan-Cancer Analysis of the Association between FBXW Expression and Prognosis TIMER 2.0 (http://timer.cistrome.org/, access date: 10 February 2023) [34] was used to study the association between FBXW expression and the prognosis of 41 cancer types including subtypes of certain major ones. The exploration module was selected and FBXWs were input into the box. The cut-off was selected as $p \leq 0.05.$ ## 4.3. Pan-Cancer Analysis of FBXW Expression and Cancer Immunology The correlation between FBXWs’ mRNA expression and stroma score was studied using the cBioportal database (https://www.cbioportal.org/, access date: 10 November 2022) relying on the CPTAC dataset [35]. The correlation between FBXW9 mRNA levels and immune cell infiltration was explored using the TIMER database. The association between the expression of FBXWs and patients receiving anti-PD1 therapy was analyzed through a Kaplan–Meier plot (http://kmplot.com/analysis/index.php?p=service&cancer=immunotherapy, access date: 10 February 2023). ## 4.4. Analysis of FBXWs’ mRNA Expression and Stemness of Breast Cancer The cBioportal database was used to acquire the stemness score and mRNA expression of FBXWs from each case in CPTAC. The XENA database (https://xenabrowser.net/, access date: 10 November 2022) was used to study the association between MYC activity and mRNA expression of FBXWs in the TCGA-BRCA dataset. The MYC activity was calculated using the RABIT transcription factor regulatory impact [36]. ## 4.5. Prediction of Substrates of FBXW9 The UbiBrowser 2.0 software (http://ubibrowser.bio-it.cn/ubibrowser_v3/, access date: 5 November 2022) [37] was chosen to predict substrates of FBXW9. The protein interaction was analyzed using the STRING database (https://cn.string-db.org/, access date: 5 November 2022). The predicted substrates were further analyzed through KEGG and GO analyses using the DAVID database (https://david.ncifcrf.gov/summary.jsp, access date: 5 November 2022). ## 4.6. Pathway Enrichment Analysis of FBXW9-Correlated Genes in Breast Cancer We downloaded proteomic data on breast cancer from CPTAC. The cases were ranked according to FBXW9 protein expression. After that, we selected 10 cases with the highest FBXW9 expression (FBXW9 high-expression group) and 10 cases with the lowest FBXW9 expression (FBXW9 low-expression group). We next downloaded transcriptome data of 20 cases including the FBXW9 high- and low-expression groups. The DEseq2 package was used to analyze differentially expressed genes between the two groups. The cut-off was $p \leq 0.05$ and \|Fold change\| > 1. The differentially expressed genes were again sent to KEGG analysis and GO analysis using the DAVID database. They were also used to build a signaling network with Metascape (https://metascape.org/gp/index.html#/main/step1, access date: 5 January 2023). ## 4.7. Cell Culture Triple-negative breast cancer cell lines SUM159 and MDA-MB-231 were purchased from ATCC and maintained in DMEM with $10\%$ FBS (Hyclone, UT, USA). The cells were cultured in a humid incubator (37 °C, $5\%$ CO2). ## 4.8. siRNA-Mediated Gene Silencing Specific siRNAs targeting FBXW9 and control siRNA were purchased from GenePharma (Shanghai, China). The siRNAs were transfected into SUM159 and MDA-MB-231 cells using LipoFectamine RNAiMax (Invitrogen, CA, USA) reagent according to the manufacturers’ protocol. ## 4.9. Cell Proliferation and Colony Forming Assays The proliferation ability of cells was determined with a CCK-8 kit (Dojindo, Tokyo, Japan) following the manufacturer’s protocol. For the colony-forming assay, cells were transfected with siRNAs and then plated in 6-well plates and cultured for 7 days. The cells were stained with crystal violet. ## 4.10. Cell Cycle Analysis Cell cycle analysis was conducted as previously described [38]. In short, Cells were fixed in $70\%$ ethanol for 1 h at 4 °C, and then were treated with RNase and stained with PI for 30 min at room temperature. The cells were then subjected to flow cytometry analysis using ModFit software (Version 3.1). ## 4.11. Human Tissues Collection The breast tumors and matched normal tissues were collected from surgery from 25 patients in the Affiliated Hospital of Yangzhou University from 2021 to 2022. Written informed consent was acquired from all participants and the study was under the supervision of the Ethic Committee of the Affiliated Hospital of Yangzhou University (YXYLL-2021-07). No therapy was received before the surgery. The samples were immediately subjected to RNA extraction using TRIzol reagent. ## 4.12. RT-qPCR RNA was extracted from cells with Trizol reagent (Invitrogen). The RNA was reverse-transcribed with HiScript Reverse Transcriptase (Vazyme, Nanjing, China) and RT-qPCR was performed with ChamQ Universal SYRB qPCR Master Mix (Vazyme). The primer sequences were as follows: FBXW9-F:5′-TAGGGCGGTGCGATGATTC-3′; FBXW9-R:5′-CGGATTTTGGCGGACTGAGA-3′; p21-F: 5′-TGTCCGTCAGAACCCATGC-3′; p21-R:5′-AAAGTCGAAGTTCCATCGCTC-3′; CCNA2-F:5′-CGCTGGCGGTACTGAAGTC-3′; CCNA2-R: 5′-GAGGAACGGTGACATGCTCAT-3′; CCNB1-F:5′-AATAAGGCGAAGATCAACATGGC-3′; CCNB1-R:5′-TTTGTTACCAATGTCCCCAAGAG-3′; NECTIN2-F:5′-GGATGTGCGAGTTCAAGTGCT-3′; NECTIN2-R:5′-TGGGACCCATCTTAGGGTGG-3′; CD274-F:5′-TGGCATTTGCTGAACGCATTT-3′; CD274-R:5′-TGCAGCCAGGTCTAATTGTTTT-3′; PDCD1LG2-F:5′-ATTGCAGCTTCACCAGATAGC-3′; PDCD1LG2-R:5′-AAAGTTGCATTCCAGGGTCAC-3′; 18S-F:5′-GTAACCCGTTGAACCCCATT-3′;18S-R:5′-CCATCCAATCGGTAGTAGCG-3′. ## 4.13. Western Blotting Protein was extracted from cells using RIPA lysis buffer (Beyotime, Shanghai, China). Lysates were loaded on an SDS-PAGE gel. After electrophoresis, proteins were transferred to a PVDF membrane. The membrane was blocked with $5\%$ non-fat milk and then incubated with the primary and secondary antibodies sequentially. The blots were developed with Enhanced ECL buffer (Beyotime). The information of antibodies was as follows: FBXW9 (Novus Biologicals, CO, USA), p21 (Proteintech, IL, USA), GAPDH (Proteintech), HRP-conjugated anti-mouse (Proteintech), HRP-conjugated anti-rabbit (Proteintech). ## 4.14. Statistical Analysis The data were analyzed and graphed using GraphPad Prism 6 and presented as the mean ± SD of three repeats. Two-group comparison was performed with Student’s t-test. A $p \leq 0.05$ was considered statistically significant. ## 5. Conclusions In conclusion, the comprehensive analysis of FBXWs suggested that several novel E3 ligases might be involved in cancer progression. Importantly, the data strongly suggested a critical role of FBXW9 in several cancer types. FBXW9 might govern cell cycle, stemness, and immune infiltration in cancers. In basal breast cancer, FBXW9 promoted cell cycle progression to facilitate cell growth. However, further studies to explore the molecular mechanisms of FBXWs are needed in the future. ## References 1. Siegel R.L., Miller K.D., Fuchs H.E., Jemal A.. Cancer Statistics, 2021. CA Cancer J. Clin. (2021) 71 7-33. DOI: 10.3322/caac.21654 2. Luo Z., Liu W., Sun P., Wang F., Feng X.. Pan-cancer analyses reveal regulation and clinical outcome association of the shelterin complex in cancer. Brief. Bioinform. (2021) 22 bbaa441. DOI: 10.1093/bib/bbaa441 3. Li Y., Jiang T., Zhou W., Li J., Li X., Wang Q., Jin X., Yin J., Chen L., Zhang Y.. Pan-cancer characterization of immune-related lncRNAs identifies potential oncogenic biomarkers. Nat. Commun. (2020) 11 1000. DOI: 10.1038/s41467-020-14802-2 4. Hanahan D.. Hallmarks of Cancer: New Dimensions. Cancer Discov. (2022) 12 31-46. DOI: 10.1158/2159-8290.CD-21-1059 5. Hershko A., Ciechanover A.. The ubiquitin system. Annu Rev. Biochem. (1998) 67 425-479. DOI: 10.1146/annurev.biochem.67.1.425 6. Bian C., Luan Z., Zhang H., Zhang R., Gao J., Wang Y.. miR-154-5p Affects the TGFbeta1/Smad3 Pathway on the Fibrosis of Diabetic Kidney Disease via Binding E3 Ubiquitin Ligase Smurf1. Oxid. Med. Cell. Longev. (2022) 2022 7502632. DOI: 10.1155/2022/7502632 7. Senft D., Qi J., Ronai Z.. Ubiquitin ligases in oncogenic transformation and cancer therapy. Nat. Rev. Cancer (2017) 18 69-88. DOI: 10.1038/nrc.2017.105 8. Xiong B., Huang J., Liu Y., Zou M., Zhao Z., Gong J., Wu X., Qiu C.. Ubiquitin-specific protease 2a promotes hepatocellular carcinoma progression via deubiquitination and stabilization of RAB1A. Cell. Oncol. (2020) 44 329-343. DOI: 10.1007/s13402-020-00568-8 9. Wang Z., Liu P., Inuzuka H., Wei W.. Roles of F-box proteins in cancer. Nat. Rev. Cancer (2014) 14 233-247. DOI: 10.1038/nrc3700 10. Frescas D., Pagano M.. Deregulated proteolysis by the F-box proteins SKP2 and beta-TrCP: Tipping the scales of cancer. Nat. Rev. Cancer (2008) 8 438-449. DOI: 10.1038/nrc2396 11. Tekcham D.S., Chen D., Liu Y., Ling T., Zhang Y., Chen H., Wang W., Otkur W., Qi H., Xia T.. F-box proteins and cancer: An update from functional and regulatory mechanism to therapeutic clinical prospects. Theranostics (2020) 10 4150-4167. DOI: 10.7150/thno.42735 12. Zhang Q., Mady A.S.A., Ma Y., Ryan C., Lawrence T.S., Nikolovska-Coleska Z., Sun Y., Morgan M.A.. The WD40 domain of FBXW7 is a poly(ADP-ribose)-binding domain that mediates the early DNA damage response. Nucleic Acids Res. (2019) 47 4039-4053. DOI: 10.1093/nar/gkz058 13. Huang G., Xiang Z., Wu H., He Q., Dou R., Yang C.. The lncRNA SEMA3B-AS1/HMGB1/FBXW7 Axis Mediates the Peritoneal Metastasis of Gastric Cancer by Regulating BGN Protein Ubiquitination. Oxid. Med. Cell. Longev. (2022) 2022 5055684. DOI: 10.1155/2022/5055684 14. Han Q., Zhang Q., Song H., Bamme Y., Song C., Ge Z.. FBXW4 Is Highly Expressed and Associated with Poor Survival in Acute Myeloid Leukemia. Front. Oncol. (2020) 10 149. DOI: 10.3389/fonc.2020.00149 15. Wei P., Jiang J., Xiao M., Zeng M., Liu X., Zhao B., Chen F.. The transcript ENST00000444125 of lncRNA LINC01503 promotes cancer stem cell properties of glioblastoma cells via reducing FBXW1 mediated GLI2 degradation. Exp. Cell Res. (2022) 412 113009. DOI: 10.1016/j.yexcr.2022.113009 16. Lei M., Luo C., Zhang J., Cao W., Ge J., Zhao M.. A m6A methyltransferase-mediated immune signature determines prognosis, immune landscape and immunotherapy efficacy in patients with lung adenocarcinoma. Cell. Oncol. (2022) 45 931-949. DOI: 10.1007/s13402-022-00697-2 17. Zou Y., Ye F., Kong Y., Hu X., Deng X., Xie J., Song C., Ou X., Wu S., Wu L.. The Single-Cell Landscape of Intratumoral Heterogeneity and The Immunosuppressive Microenvironment in Liver and Brain Metastases of Breast Cancer. Adv. Sci. (2022) 10 2203699. DOI: 10.1002/advs.202203699 18. Yang Z., Zhao N., Cui J., Wu H., Xiong J., Peng T.. Exosomes derived from cancer stem cells of gemcitabine-resistant pancreatic cancer cells enhance drug resistance by delivering miR-210. Cell. Oncol. (2019) 43 123-136. DOI: 10.1007/s13402-019-00476-6 19. Lawson D.A., Bhakta N.R., Kessenbrock K., Prummel K.D., Yu Y., Takai K., Zhou A., Eyob H., Balakrishnan S., Wang C.-Y.. Single-cell analysis reveals a stem-cell program in human metastatic breast cancer cells. Nature (2015) 526 131-135. DOI: 10.1038/nature15260 20. Waldman T., Kinzler K.W., Vogelstein B.. p21 is necessary for the p53-mediated G1 arrest in human cancer cells. Cancer Res. (1995) 55 5187-5190. PMID: 7585571 21. Olivero C.E., Martínez-Terroba E., Zimmer J., Liao C., Tesfaye E., Hooshdaran N., Schofield J., Bendor J., Fang D., Simon M.D.. p53 Activates the Long Noncoding RNA Pvt1b to Inhibit Myc and Suppress Tumorigenesis. Mol. Cell (2020) 77 761-774.e8. DOI: 10.1016/j.molcel.2019.12.014 22. Popovic D., Vucic D., Dikic I.. Ubiquitination in disease pathogenesis and treatment. Nat. Med. (2014) 20 1242-1253. DOI: 10.1038/nm.3739 23. Fuchs S.Y., Spiegelman V.S., Kumar K.G.. The many faces of beta-TrCP E3 ubiquitin ligases: Reflections in the magic mirror of cancer. Oncogene (2004) 23 2028-2036. DOI: 10.1038/sj.onc.1207389 24. Bi Y., Cui D., Xiong X., Zhao Y.. The characteristics and roles of β-TrCP1/2 in carcinogenesis. FEBS J. (2021) 288 3351-3374. DOI: 10.1111/febs.15585 25. Ren C., Han X., Lu C., Yang T., Qiao P., Sun Y.. Ubiquitination of NF-kappaB p65 by FBXW2 suppresses breast cancer stemness, tumorigenesis, and paclitaxel resistance. Cell Death Differ. (2022) 29 381-392. DOI: 10.1038/s41418-021-00862-4 26. Lim Y.X., Lin H., Chu T., Lim Y.P.. WBP2 promotes BTRC mRNA stability to drive migration and invasion in triple-negative breast cancer via NF-kappaB activation. Mol. Oncol. (2022) 16 422-446. DOI: 10.1002/1878-0261.13048 27. Lockwood W.W., Chandel S.K., Stewart G.L., Erdjument-Bromage H., Beverly L.J.. The Novel Ubiquitin Ligase Complex, SCFFbxw4, Interacts with the COP9 Signalosome in an F-Box Dependent Manner, Is Mutated, Lost and Under-Expressed in Human Cancers. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0063610 28. Sun Y., Hu Z., Goeb Y., Dreier L.. The F-Box Protein MEC-15 (FBXW9) Promotes Synaptic Transmission in GABAergic Motor Neurons in C. elegans. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0059132 29. Becht E., Giraldo N.A., Lacroix L., Buttard B., Elarouci N., Petitprez F., Selves J., Laurent-Puig P., Sautes-Fridman C., Fridman W.H.. Estimating the population abundance of tissue-infiltrating immune and stromal cell populations using gene expression. Genome Biol. (2016) 17 218. DOI: 10.1186/s13059-016-1070-5 30. Cordenonsi M., Zanconato F., Azzolin L., Forcato M., Rosato A., Frasson C., Inui M., Montagner M., Parenti A.R., Poletti A.. The Hippo Transducer TAZ Confers Cancer Stem Cell-Related Traits on Breast Cancer Cells. Cell (2011) 147 759-772. DOI: 10.1016/j.cell.2011.09.048 31. Ji J., Xu R., Zhang X., Han M., Xu Y., Wei Y., Ding K., Wang S., Huang B., Chen A.. Actin like-6A promotes glioma progression through stabilization of transcriptional regulators YAP/TAZ. Cell Death Dis. (2018) 9 517. DOI: 10.1038/s41419-018-0548-3 32. Sui Y., Liu Z., Park S.H., Thatcher S.E., Zhu B., Fernandez J.P.. IKKbeta is a beta-catenin kinase that regulates mesenchymal stem cell differentiation. JCI Insight (2018) 3 e96660. DOI: 10.1172/jci.insight.96660 33. Aceto N., Sausgruber N., Brinkhaus H., Gaidatzis D., Martiny-Baron G., Mazzarol G., Confalonieri S., Quarto M., Hu G., Balwierz P.J.. Tyrosine phosphatase SHP2 promotes breast cancer progression and maintains tumor-initiating cells via activation of key transcription factors and a positive feedback signaling loop. Nat. Med. (2012) 18 529-537. DOI: 10.1038/nm.2645 34. Li T., Fu J., Zeng Z., Cohen D., Li J., Chen Q., Li B., Liu X.S.. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. (2020) 48 W509-W514. DOI: 10.1093/nar/gkaa407 35. Krug K., Jaehnig E.J., Satpathy S., Blumenberg L., Karpova A., Anurag M., Miles G., Mertins P., Geffen Y., Tang L.C.. Proteogenomic Landscape of Breast Cancer Tumorigenesis and Targeted Therapy. Cell (2020) 183 1436-1456.e31. DOI: 10.1016/j.cell.2020.10.036 36. Jiang P., Freedman M.L., Liu J.S., Liu X.S.. Inference of transcriptional regulation in cancers. Proc. Natl. Acad. Sci. USA (2015) 112 7731-7736. DOI: 10.1073/pnas.1424272112 37. Wang X., Li Y., He M., Kong X., Jiang P., Liu X., Diao L., Zhang X., Li H., Ling X.. UbiBrowser 2.0: A comprehensive resource for proteome-wide known and predicted ubiquitin ligase/deubiquitinase–substrate interactions in eukaryotic species. Nucleic Acids Res. (2021) 50 D719-D728. DOI: 10.1093/nar/gkab962 38. Sun H., Lin D.-C., Cao Q., Pang B., Gae D.D., Lee V.K.M., Lim H.J., Doan N., Said J.W., Gery S.. Identification of a Novel SYK/c-MYC/MALAT1 Signaling Pathway and Its Potential Therapeutic Value in Ewing Sarcoma. Clin. Cancer Res. (2017) 23 4376-4387. DOI: 10.1158/1078-0432.CCR-16-2185
--- title: Metformin Attenuates the Inflammatory Response via the Regulation of Synovial M1 Macrophage in Osteoarthritis authors: - Meng Zheng - Yuanli Zhu - Kang Wei - Hongxu Pu - Renpeng Peng - Jun Xiao - Changyu Liu - Xuying Sun journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049635 doi: 10.3390/ijms24065355 license: CC BY 4.0 --- # Metformin Attenuates the Inflammatory Response via the Regulation of Synovial M1 Macrophage in Osteoarthritis ## Abstract Osteoarthritis (OA), the most common chronic inflammatory joint disease, is characterized by progressive cartilage degeneration, subchondral bone sclerosis, synovitis, and osteophyte formation. Metformin, a hypoglycemic agent used in the treatment of type 2 diabetes, has been evidenced to have anti-inflammatory properties to treat OA. It hampers the M1 polarization of synovial sublining macrophages, which promotes synovitis and exacerbates OA, thus lessening cartilage loss. In this study, metformin prevented the pro-inflammatory cytokines secreted by M1 macrophages, suppressed the inflammatory response of chondrocytes cultured with conditional medium (CM) from M1 macrophages, and mitigated the migration of M1 macrophages induced by interleukin-1ß (IL-1ß)-treated chondrocytes in vitro. In the meantime, metformin reduced the invasion of M1 macrophages in synovial regions brought about by the destabilization of medial meniscus (DMM) surgery in mice, and alleviated cartilage degeneration. Mechanistically, metformin regulated PI3K/AKT and downstream pathways in M1 macrophages. Overall, we demonstrated the therapeutic potential of metformin targeting synovial M1 macrophages in OA. ## 1. Introduction Osteoarthritis (OA), the most common inflammatory joint disease, mainly affects the diarthrodial joints [1]. It caused a growing economic burden to society. The incidence of OA is rising due to an increasingly aging population. OA is characterized by cartilage degeneration, osteophyte formation, subchondral bone remodeling, and synovial hyperplasia [2]. Increasing evidence indicates that macrophages represent a significant culprit in OA. In response to various stimuli, synovial macrophages differentiate into two phenotypes: pro-inflammatory M1 type and anti-inflammatory M2 type [3]. M1 macrophages are typically linked to the immune response to bacteria and intracellular pathogens and develop in inflammatory situations driven by Toll-like receptor (TLR) and interferon signaling [4]. TLRs regulate the expression of proinflammatory cytokines, and their interaction with HMGB1 increases insulin resistance in type 2 diabetes [5]. In both human and mouse OA synovium, M1 but not M2-polarized macrophages accumulated [6]. The degenerative process of OA is worsened by M1 macrophage-driven synovitis [7]. In comparison to osteoarthritic chondrocytes in monoculture, osteoarthritic chondrocytes cocultured with activated macrophages expressed significantly higher levels of matrix metalloproteinases (MMPs), IL-1ß, tumor necrosis factor-α (TNF-α), and interferon-γ (IFN-γ), suggesting that proinflammatory macrophages may exacerbate the abnormal matrix degradation and cytokine secretion already associated with osteoarthritic chondrocytes [8]. Metformin is a first-line therapy for type 2 diabetes due to its glucose-lowering actions via suppression of hepatic gluconeogenesis, enhancement of glucose uptake in muscle and adipose tissue, and the reduction in intestinal glucose absorption [9]. Multiple research groups have reported findings indicating the putative preventive benefits of metformin in OA development [10]. A favorable effect of metformin on chondroprotection, immunomodulation, and the pain reduction in knee osteoarthritis was consistently supported by pre-clinical and human trials [11]. However, how metformin attenuates cartilage degeneration is not clear. One study suggested that metformin lowered monocyte differentiation and prevented inflammation, oxidative stress, polarization, foam cell formation, and apoptosis in macrophages [12]. Others found in palmitate-stimulated bone marrow-derived macrophages (BMDMs), metformin caused a shift in the ratio of M1 macrophages to M2 macrophages [13], indicating metformin’s role in macrophage reprogramming. We hypothesized that metformin might reduce inflammation by regulating synovial macrophage polarization, thus further relieving synovitis and protecting chondrocytes from degeneration. To prove our hypothesis, cell culture and DMM model were used to identify whether metformin could rescue OA pathogenesis by targeting synovial macrophages. Our study uncovered the protective effects of metformin in treating OA in the way of anti-synovitis in macrophages. ## 2.1. Increased M1-Polarized, but Not M2-Polarized Macrophages in Synovial Tissues from OA Patients and DMM Mice To investigate the involvement of synovial macrophages in the development of OA, we investigated the synovial inflammation and macrophage morphology changes in human normal and OA synovium. Severe synovial hyperplasia and inflammatory cell infiltration were observed in human OA synovium, with considerably higher synovitis scores than in normal samples (Figure 1A,E). The morphological characteristics of accumulated synovial macrophages were further identified via histomorphometry. Compared to controls, F$\frac{4}{80}$ (macrophage marker)-positive cells and INOS (M1-like macrophage marker)-positive cells were significantly elevated in OA synovial tissue (Figure 1B,C,F). Although the M2-like macrophage CD206 positive cell number increased, the ratio of CD206 positive macrophages did not differ in normal and OA synovium because the total cell number was also elevated in OA synovium (Figure 1D,F). These results indicated that M1-polarized macrophages, but not M2-polarized macrophages, increased in the synovium of joints in patients with OA compared to normal controls. DMM surgery was further performed to create an OA model in mice. Similarly, no apparent cell penetration was found in the synovium of sham knees. Still, the synovium of OA knees became hyperplastic and hypertrophic, with a higher synovitis score in DMM mice (Figure 1G,K). Notably, DMM surgery reliably increased the macrophage number in synovial lining layers. The ratio of macrophages with M1-like characteristics increased, but the ratio of M2-like macrophages remained unchanged (Figure 1H–J,L). These data indicated that macrophages accumulated around OA synovium, with increased M1-like polarization, suggesting their possible significance in the pathophysiology and progression of OA. ## 2.2. Metformin Inhibits Inflammatory Cytokine Levels Produced by M1 Macrophages To determine if metformin alters inflammatory activities in M1 polarized macrophages, BMDMs were stimulated with lipopolysaccharide (LPS) (100 ng/mL) and IFN-γ (20 ng/mL) for 12 h and exhibited the M1 polarization phenotype [14,15,16]. Then the cells were incubated with different concentrations of metformin (Met) from 0.25 mM to 4 mM. Macrophages without intervention served as M0 control. The CCK-8 assays demonstrated that low-dose metformin has no toxic effects on BMDMs (Figure 2A). CCK-8 results also showed the viability of M1 macrophages treated with high-dose metformin (10 mM to 100 mM) was impaired from 40 mM (Figure S1A). Western blot results showed that LPS and IFN-γ stimulation without metformin treatment markedly increased the expression of the inflammatory cytokines and markers such as INOS, CD86, IL-1ß, and IL-6, consistent with the M1 phenotype of polarized macrophages. However, metformin dose-dependently reduced the expression of these cytokines and markers induced by M1-polarized macrophages, with the lowest expression at 3 mM and no longer decreased at 4 mM. (Figure 2B,C). The results from the qPCR analysis of the mRNA levels of IL-1ß, IL-6, TNF-α, and INOS were significantly elevated, and they were dose-dependently downregulated by metformin, consistent with Western blot results (Figure 2D). On the contrary, high-dose metformin enhanced the M1 polarization of macrophages and exacerbated inflammation (Figure S1B). ## 2.3. Metformin Inhibits Macrophage-Derived CM-Induced Chondrocyte Degeneration and Cell Migration The Western blot results demonstrated that chondrocytes cultured with M1 macrophages CM (excluding LPS and IFN-γ) increased the expression of MMP3 and MMP13, and decreased the expression of collagen type II alpha 1 (Col2a1) and SRY-Box Transcription Factor 9 (Sox9), indicating that M1 macrophages might release multiple destructive factors that contribute to cartilage degeneration in OA. However, M1-CM exposed chondrocytes in the presence of metformin for 24 h showed increased Col2a1 and Sox9 expression and decreased MMP13 and MMP3 levels compared with M1-CM group (Figure 3A,B). Results from the qPCR analysis of mRNA levels of Col2a1, Sox9, MMP13, and MMP3 were consistent with Western blot results (Figure 3C). These findings suggested that metformin reversed the inflammatory effects caused by destructive factors produced by M1 macrophages, which protected the cartilage from degeneration. We co-cultured BMDMs and chondrocytes using Transwell inserts to determine whether more M1 macrophages migrate to chondrocytes in inflammatory environments in vitro. BMDMs were seeded on the upper chamber and were induced to M1 macrophages by LPS + IFN-γ before the co-culture system. More M1 macrophages migrated toward chondrocytes in the presence of IL-1ß. However, metformin inhibited M1 macrophage migration toward chondrocytes caused by IL-1ß (Figure 3D,E). This phenomenon suggested that chondrocytes in the inflammatory environment might secrete certain M1 macrophage chemokines. Additionally, metformin blocked such chemotaxis. The results confirmed that metformin significantly inhibited M1 macrophage migration at 24 h in the scratch wound assay (Figure 3F,G). ## 2.4. Potential Regulatory Role of Metformin in PI3K/AKT Signaling in M1 Macrophages To investigate the mechanism related to the anti-inflammatory and anti-migration role of metformin in M1 macrophages. Three biological replicates of M1 polarization of mouse BMDMs and three biological replicates of M1 macrophages treated with metformin (3 mM for 16 h were used to perform transcriptome analysis via high-throughput RNA sequencing (RNA-Seq). For identifying differentially expressed genes (DEGs), a p-value less than 0.05 was used. The final set of DEGs included genes with absolute values of logarithmic fold change (log2FC) > 1.0. Figure 4A shows that 307 DEGs were upregulated and 950 DEGs were downregulated in M1 macrophages treated with metformin compared to untreated M1 macrophages. According to kyoto encyclopedia of genes and genomes (KEGG) pathway analysis, the PI3K-AKT signaling pathway was the most significantly influenced in metformin-treated M1 macrophages (Figure 4B). PI3K/AKT pathway-related downstream genes were significantly altered as shown in heat map (Figure 4C). Additionally, the distribution of genes in several gene set pathways was displayed using gene set enrichment analysis (GSEA). Results showed that the control M1 macrophages had a high enrichment of genes relevant to the PI3K-AKT signaling pathway, pointing to a potential regulatory role for metformin in the PI3K-AKT signaling pathway (Figure 4D). The downstream mTOR pathways were also enriched, although they did not reach a significant difference (Figure 4E). The PI3K-AKT signaling and downstream mTOR, FoxO1, GSK-3ß, and IKK pathways were further tested, phosphorylation level of mTOR, FoxO1, AKT, GSK-3ß increased after LPS + IFN-γ exposure. NF-κb was also upregulated by LPS + IFN-γ, indicating a remarkable inflammation response in the presence of LPS + IFN-γ. However, mTOR, FoxO1, AKT, GSK-3ß, and NF-κb activation were notably reduced by metformin (Figure 4F,G). These results identified that metformin attenuates inflammatory response in macrophages by targeting PI3K-AKT and downstream mTOR, FoxO1, GSK-3ß, and NF-κb pathways. ## 2.5. Metformin Attenuates the Development of OA in DMM Mice We then ask whether metformin inhibits OA development in mice by targeting M1-polarized macrophages. DMM surgery was performed in mice to induce OA. DMM mice exhibited more severe cartilage degradation, as evidenced by Safranin O staining. MicroCT and HE staining showed more osteophyte formation and thicker synovial lining cells in DMM mice compared to sham-operated mice. While metformin dose-dependently reduced cartilage degradation (Figure 5A,B). Less osteophyte formation (Figure 5C,D) and milder synovitis were found after 200 mg/kg/d metformin treatment (Figure 5E,F). F$\frac{4}{80}$ and INOS-positive macrophages (inflammatory M1 type) increased, and CD206-positive macrophages (anti-inflammatory M2 type) remained unchanged after DMM surgery. Metformin dose-dependently decreased the amount of total and M1 macrophages in the synovium although the ratio of M2 type macrophages in the synovium was unchanged (Figure 5G–L). Furthermore, MMP13-positive chondrocytes were reduced in articular cartilage in both 100 mg/kg/d or 200mg/kg/d metformin-gavaged mice (Figure 5M,N). Those results confirmed that metformin attenuated cartilage degeneration by reducing synovitis through modulating macrophage reprogramming. ## 3. Discussion Metformin usage has been linked to lower rates of mortality and cardiovascular disease in type 2 diabetes (T2D) patients with chronic kidney disease (CKD) [17]. Mineral bone disease (MBD) is a frequent consequence of chronic kidney disease (CKD) [18]. Metformin has a potential role in preventing OA in patients with CKD [19]. Long-term exposure to a diabetic environment alters bone metabolism and impairs bone microarchitecture [20]. The antidiabetic drug GLP-1 receptor agonists (GLP-1Ras) could increase BMD and decrease the implications of bone fragility [21,22]. Another antidiabetic drug, SGLT-2 inhibitors, might have detrimental effects on skeletal integrity [23]. The therapeutic effects of more antidiabetic drugs on osteoarthritis are worth further exploration. In recent years, OA treatment has expanded to include more and more complementary therapies [24]. Routine management includes healthy exercise and weight management [25]. Non-steroidal anti-inflammatory drugs (NSAIDs), acetaminophen, duloxetine, and other medications have been licensed by the Osteoarthritis Research Society International (OARSI) to improve symptoms. Interventional treatment involves injecting corticosteroids, hyaluronic acid, and platelet-rich plasma into the affected joint. Surgical interventions, such as knee replacement and arthroscopy, are generally performed when a less invasive treatment is unsuccessful. Cell therapy uses stem cells to regenerate damaged articular cartilage. Gene therapy using genes expressing cartilage growth factor and anti-inflammatory cytokines is of interest in treating OA [26]. The understanding of the pathophysiology of OA is still changing; it has gone from being thought of as a disease that only affects cartilage, to one that affects the entire joint [1], including the synovium [27]. Macrophages play a significant role in the development of OA. High levels of macrophages are a conspicuous characteristic of inflammatory lesions, and an early sign of active OA, which exhibits an increased number of sublining macrophages in the synovium [13,28,29]. The depletion of these macrophages from inflamed tissue offers a significant therapeutic advantage. The degree of synovial macrophage infiltration correlates with the degree of joint erosion [30]. Although it was reported that metformin has therapeutic effects in treating OA, it is still unclear as to whether it affects the tissues around the joint, especially the inflammatory response in the synovium during OA development. In the present study, we confirmed that M1-like macrophages accumulated in the synovium, both in human samples and in DMM surgical mice, consistent with previous studies [31]. M1 macrophages increased the expression of inflammatory cytokines and markers such as INOS, CD86, IL-1ß, and IL-6. At the same time, metformin could reduce these cytokine levels induced by M1 macrophages. Stimulating chondrocytes with M1 macrophages, CM increased the expression of the breakdown enzymes such as MMP13 and MMP3, while decreasing Col2a1 and Sox9 levels. The degenerative phenotypes of chondrocytes cultured with M1 macrophage CM were identical to those of an IL-1ß-treated condition. However, metformin significantly reversed these degenerative cartilage markers caused by M1 macrophages CM. Macrophages were also responsive to IL-1ß. Chondrocytes with IL-1ß treatment induced more M1 macrophages to migrate to the bottom chamber. To our surprise, metformin could inhibit such a migration of the M1 macrophages, which suggested that metformin might delay OA phenotypes by hampering M1 macrophage migration. It deserves more research. The protective effect of metformin was further investigated in DMM mice. We found that metformin reduced the infiltration of M1 macrophages into the synovium in DMM mouse models, and alleviated synovitis and osteophyte formation, suggesting that metformin also has an indirect protective effect on chondrocytes by inhibiting the inflammatory secretion and migration of M1 macrophages. RNA-seq analysis was conducted in M1 macrophages treated with or without metformin for 16h. KEGG pathway analysis showed that the PI3K/AKT signaling pathway was markedly inhibited in metformin-treated M1 macrophages compared to the untreated group. PI3K/AKT is one of the most critical pathways that controls many cellular processes, including cell division, autophagy, survival, and differentiation [32]. AKT is activated by the phosphorylation of the transcription factor IκB kinase (IKK), a positive regulator of NF-kB, regulating the expression of genes with antiapoptotic activity [33]. Glycogen synthase kinase-3 (GSK3), the mammalian target of rapamycin (mTOR), are both AKT targets involved in protein synthesis, glycogen metabolism, and cell cycle regulation [34]. The phosphorylation of FoxO1 regulates NF-κB nuclear translocation by activating PI3K/AKT during aging [35]. Western blotting showed that changes in the downstream IKK, mTOR, NF-κB, and GSK-3ß pathways were also regulated. These results indicated that metformin could regulate many aspects of M1 macrophages through the PI3K/AKT signaling pathways. Indeed, a lot of studies have explored the effects of macrophages on OA by regulating synovitis. However, the specific mechanism needs to be further explored. The study reconfirmed that macrophages play an important role in the development of OA synovitis. The increased infiltration of M1 macrophages is accompanied by the aggravation of OA symptoms. Metformin can significantly inhibit the infiltration of M1 macrophages in OA. This finding has been rarely reported so far, which further provides the basis for metformin in the treatment of OA. There are many limitations in our study. First, although metformin attenuated M1 macrophage migration, we cannot exclude the direct role of metformin on chondrocytes in vivo. The protective effect of metformin may probably consist of the direct role in chondrocytes, and the indirect part in macrophage migration in the synovium, which significantly inhibited the inflammatory response in the tissue around the joint. Second, the precise mechanism underlying how metformin affects macrophage migration still needs further exploration. AMP-activated protein kinase (AMPK) has a significant role in its mechanism of action. Is the AMPK pathway also related to PI3K/AKT signaling? Despite the PI3K/AKT pathways, STAT6 was also vital in regulating macrophage polarization [36]. We will examine the exact mechanisms in our future studies. ## 4.1. Reagents and Antibodies Recombinant murine IL-1β and recombinant murine M-CSF protein were purchased from R&D Systems (Minneapolis, MN, USA). Recombinant murine IFN-γ was purchased from Peprotech (Rocky Hill, NJ, USA). Lipopolysaccharides (LPS) were obtained from Biosharp (Beijing, China). Metformin hydrochloride was purchased from CTAO (Eugene, OSU, USA). Anti-p85-PI3K, MTOR, p-MTOR, FoxO1, p-FoxO1, AKT, p-AKT, GSK3ß, p-GSK3ß, NF-κB, p-NF-κB, IκBα, p-IκBα, IκB kinase (IKK)-β, and p-IKKα/β antibodies were supplied by Cell Signaling Technology (Beverly, MA, USA). MMP3, MMP13, INOS, and CD206 antibodies were obtained from Abcam (Cambridge, UK). Col2a1 and Sox9 antibodies were purchased from Abclonal (Wuhan, China). IL-1ß, IL-6, CD86, and F$\frac{4}{80}$ antibodies were purchased from Proteintech (Wuhan, China). GAPDH, ß-Actin, and secondary antibodies were purchased from Boster (Wuhan, China). ## 4.2. Human Synovial Sample Collection Normal human synovium tissues were collected from six objects injured in car accidents, and who had no history of arthritic diseases (five men and one woman, aged 45 ± 2.8 years old). Human OA synovium tissues were collected from six objects who underwent knee replacement surgery (two men and four women, average age 64.67 ± 5.87 years old). People with cancer, diabetes or other severe diseases in the last 5 years were excluded. The collection of human samples was approved by the Ethics Committee of the Tongji Hospital (TJ-IRB20210127). Informed written agreement was obtained before the collection. ## 4.3. Isolation and Culture of Primary Chondrocytes After the carbon dioxide euthanasia of 5-day-old neonatal mice, knee articular cartilage was dissected. Then, the articular cartilage was cut into small pieces using scissors, followed by $0.25\%$ trypsin digestion for 30 min, and then further digested with $0.25\%$ collagenase II dissolved in DMEM/F12 (Hyclone, Logan, UT, USA) growth media for 6–8 h at 37 °C. After centrifuging the digested chondrocytes (1500 rpm for 5 min), the supernatant was discarded, and the resuspended cells were cultured in DMEM/F12 supplemented with $10\%$ fetal bovine serum (FBS) (Gibco, Carlsbad, CA, USA) at 37 °C in an incubator. ## 4.4. Culture of Bone Marrow-Derived Macrophages (BMDMs) and the Induction of Macrophage Polarization Bone marrow cells were flushed out from the long bones of the mice with α-modified Eagle’s medium (α-MEM; Hyclone, Logan, UT, USA). Then, cells were cultured in α-MEM/$10\%$ FBS supplemented with 30 ng/mL M-CSF for 24 h. The non-adherent macrophages were cultured for another 3 days in the same medium. Concentrations of 100 ng/mL LPS and 20 ng/mL IFN-γ were used to induce M1 macrophage polarization. Western blot and qPCR were used to confirm the polarization phenotypes of the macrophages. ## 4.5. Protein Extraction and Western Blot Macrophages administered with metformin (0.25–4 mM), and chondrocytes cultured with CM from macrophages were lysed for 30 min in ice-cold RIPA lysis buffer (Boster, Wuhan, China) with $1\%$ protease and phosphatase inhibitor. The cell lysates were then immediately configured for 30 min at 10,000× g at 4 °C. Proteins of equal concentrations were put onto SDS-PAGE (10–$15\%$) for electrophoresis before being transferred to polyvinylidene fluoride (PVDF) membranes. Each membrane was first blocked with $5\%$ non-fat milk in TBST (Tris-buffered saline with $0.1\%$ Tween 20 buffer) for 1 h, and then primary antibodies were incubated overnight at 4 °C on a shaker. Proteins were detected using an enhanced chemiluminescence kit (Thermo Fisher Scientific, Waltham, MA, USA) in the ChemiDoc XRS System after being incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h (Bio-Rad Laboratories, Hercules, CA, USA). ## 4.6. RNA Isolation and qPCR Total RNA was extracted from cells using Trizol (Takara, Otsu, Shiga, Japan) according to the manufacturer’s instructions, and cDNA was synthesized using the HiScript II Q RT Supermix for quantitative polymerase chain reaction (qPCR) (Vazyme, Nanjing, China). Next, qPCR was performed with ChamQ SYBR Color qPCR (Vazyme, Nanjing, China). All the data were normalized to GAPDH. The primer sequences used are listed below: GAPDH, 5′-ACGGGAAGCTCACTGGCATGGCCTT-3′ (sense), 5′-CATGAGGTCCACCACCCTGTTGCTG-3′ (antisense); IL-1ß, 5′-GAAATGCCACCTTTTGACAGTG-3′ (sense), 5′-TGGATGCTCTCATCAGGACAG-3′ (antisense); IL-6, 5′-TTCACAAGTCGGAGGCTT-3′ (sense), 5′-CAGTTTGGTAGCATCCAT-3′ (antisense); TNF-α, 5′- CAGGCGGTGCCTATGTCTC-3′ (sense), 5′- CGATCACCCCGAAGTTCAGTAG-3′ (antisense); INOS, 5′-CAGGAGGAGAGAGATCCGATTTA-3′ (sense), 5′-GCATTAGCATGGAAGCAAAGA-3′ (antisense); CD86, 5′-TCAATGGGACTGCATATCTGCC-3′ (sense), 5′-GCCAAAATACTACCAGCTCACT-3′ (antisense); Col2a1, 5′-CTTCACAGCGGTAGATCCCAG-3′ (sense), 5′-ACCAGGGGAACCACTCTCAC (antisense); MMP3, 5′-ACTCCCTGGGACTCTACCAC-3′ (sense), 5′-GGTACCACGAGGA CATCAGG-3′ (antisense); MMP13, 5′-TGATGGACCTTCT GGTCTTCTGG-3′ (sense), 5′-CATCCACATGGTTGGGAAGTTCT-3′ (antisense); Sox9, 5′-CAGCCCCTTCAACCTTCCTC-3′ (sense), 5′-TGATGGTCAGCGTAGTCGTATT-3′ (antisense). The 2−ΔΔCq method was used to determine the relative mRNA levels of the target genes. ## 4.7. Cell Viability Assay BMDMs were seeded onto 96-well plates for M1 polarization, and subsequently stimulated for 24 h with metformin. Cell proliferation was assessed using the Cell Counting Kit-8 Assay (CCK-8, TargetMol, Wellesley Hills, MA, USA). Cells were incubated for 1 h in CCK-8 media to determine the proliferation rate, and their absorbance at 450 nm was measured using a microplate reader (BioTek, Winooski, VT, USA). ## 4.8. Collection of Conditional Medium (CM) BMDMs were seeded on 6-well plates at a density of 2 × 105 cells per well. M1 polarization was induced as described above. Cultures were treated with 3 mM metformin. After 12 h of M1 induction, cells were washed entirely with phosphate-buffered saline (PBS) three times, and cultured in 2 mL of serum-free DMEM for an additional 12 h. This step removed LPS and IFN-γ from CM. The CM was collected and centrifuged for 5 min at a speed of 1000× g. After that, the supernatant was aliquoted and kept at minus 80 °C until it was used as CM mixed with DMEM/F12 (1:1) to treat chondrocytes. ## 4.9. Transwell Assay The cell migration test was carried out utilizing 24-well transwell inserts with Polycarbonate (PC) membrane with 8.0 μm holes (Corning, NY, USA). Briefly, 2 × 104 BMDMs suspended in 0.1 mL of serum-free media were seeded onto the top chambers with M1 polarization induction. A total of 1 × 105 primary chondrocytes were seeded on the lower chambers. After a 24 h incubation, cells that remained on the top chamber were removed using cotton swabs, while cells that had migrated to the other side of the transwell were fixed with $4\%$ paraformaldehyde (PFA) for 10 min and stained with $1\%$ crystal violet solution. Images were captured using an inverted microscope (Eclipse TS100, Nikon, Japan). The migration was evaluated by counting the number of cells that traveled through the membranes in 5 randomly chosen areas of each transwell. ## 4.10. RNA Sequencing and Gene Set Enrichment Analysis Total RNA was extracted from M1 macrophages treated with or without metformin (3mM). RNA purity was evaluated using the NanoDrop™ One/OneC (Thermo Scientific, Waltham, MA, USA). RNA quantification was detected using the Qubit™ RNA HS Assay Kit (Thermo Scientific, Waltham, MA, USA). RNA integrity was assessed using the Agilent 4200 TapeStation (Agilent Technologies, Palo Alto, CA, USA). Then, the libraries were constructed using the TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) according to the manufacturer’s instructions. AMPure XP beads (Beckman Coulter, Bria, CA, USA) were used to purify the fragments during library generation. These libraries were sequenced on an Illumina PE150, and 125 bp/150 bp paired-end reads were generated. Transcriptome sequencing and analysis were conducted by the Wuhan Biobank institution (Wuhan, China). Then, the clean reads were mapped to the reference genome using Hisat2. The FPKM of each gene was calculated using Cufflinks. HTSeqcount obtained the read counts of each gene. Expression analysis was performed using the DESeq R package. Hierarchical cluster analysis was performed to demonstrate the expression pattern of genes in different groups. KEGG pathway enrichment and GSEA analysis were performed using R, based on the hypergeometric distribution. ## 4.11. Dissociation of the Medial Meniscus (DMM) Surgery and Animal Gavage Eight-week-old C57/BL6 male mice ($$n = 24$$) purchased from Gempharmatech Company (Jiangsu, China) were used for our in vivo experiment. Mice were housed in the animal facility of Tongji Hospital, with less than five mice per cage, and free access to food and water. The surgical process was approved by The Ethics Committee of Tongji Hospital. DMM was performed to induce OA, as previously reported [37]. Before the surgery, the mice were anesthetized, and the right knee was cleaned and shaved. Then, the joint capsule was opened, and the fat pad was removed so that the medial meniscus-tibial ligament could be seen under a microscope. The ligament was cut before stitching up the joint capsule and skin wound. Mice were randomly divided into four groups: [1] Sham group: a sham operation involves cutting on the right side of the knee without damaging the joint capsule [38]. [ 2] DMM surgery group. [ 3] Metformin gavaged at a dosage of 100 mg/kg/d post-surgery, or [4] Metformin gavaged at a dosage of 200 mg/kg/d for 8 weeks post-surgery. ## 4.12. Micro-Computed Tomography (MicroCT) Imaging MicroCT analysis was performed on knees from sham-operated and DMM surgical mice using a VivaCT 40 scanner at 15 μm resolution, 70 kVP, and 112 μA X-ray energy (Scanco, Wangen-Brüttisellen, Switzerland). Following the manufacturer’s recommendations, the three-dimensional images were rebuilt. The periarticular osteophytes were chosen as the region of interest (ROI). On the medial and lateral sides of the tibia and femur, the ROI size was computed blindly on all four knee condyles (0–3), and the average was used for statistical analysis. ## 4.13. Safranin O–Fast Green, and H&E Mouse knee joint specimens were collected immediately after scanning. After decalcification, dehydration, and paraffin embedding of the samples, tissues were cut into 5-μm sections and stained using a Safranin O and Fast Green staining kit (Solarbio, Beijing, China), and hematoxylin and eosin solutions (H&E; Servicebio, Wuhan, China), as previously described [39]. ImageJ (version 1.53k) was used to determine and to measure the Safranin O-positive cartilage areas. H&E staining was used to evaluate synovial activation by scoring synovial lining cell thickness (0–3); the medial and lateral compartments of the joint were scored separately, and the sum of the two scores was presented (maximum site score 6) [40]. Cartilage degeneration was graded in Safranin-O/Fast Green-stained sections using a 0–6 semi-quantitative scoring system [41]. Two blinded, independent graders assessed each section, and the average score was used for statistical analysis. ## 4.14. Immunohistochemistry and Immunofluorescence After deparaffinization and hydration, sections were immersed in citrate buffer (10 mM citric acid, pH 6.0) overnight at 60 °C to unmask the antigen. A $3\%$ hydrogen peroxide was used to inhibit endogenous peroxidase activity before immunohistochemical staining. The sections were incubated with primary antibodies overnight at 4 °C after being blocked with $3\%$ goat serum at 37 °C for 1 h. For immunohistochemistry, the sections of horseradish peroxidase-conjugated secondary antibodies were stained for immunohistochemical staining. Next, the chromogen was observed using 3, 3-diaminobenzidine, and the sections were counterstained with hematoxylin. For immunofluorescence, sections were stained with Alexa 488- or Alexa 594-dyed secondary antibodies after primary antibodies incubation. DAPI (Boster, Wuhan, China) was utilized to label the nuclei. A fluorescence microscope (EVOS FL Auto, Thermo Fisher Scientific, Waltham, MA, USA) was used to acquire the images. Three areas of synovium were chosen for high-magnification imaging, and a mean value was determined for the positive staining of synovial intimal lining macrophages. ## 4.15. Scratch Assay A 10 mL pipette was used to scratch the M1 macrophage monolayer in a 6-well plate. Before taking the photos, the cells were given one wash to eliminate the suspension cells. Then, the culture medium was supplemented with or without 3 mM metformin. The scratches were imaged once more at the same location after 24 h to assess cell migration. Picture J was used to gauge the width of the marks. The scratch closure rates were determined using the formula (width 0 h − width 24 h)/width 0 h × $100\%$. ## 4.16. Statistical Analysis All experimental data were independently repeated at least three times. All data were presented as mean ± SEM. Student’s t-test was used to assess differences between two groups, while one-way analysis of variance (ANOVA) and Tukey’s multiple comparison tests were used to investigate differences between three groups using GraphPad Prism v. 8.0 (Graphpad Software Inc., San Diego, CA, USA) software. All statistical tests were two-sided. The significance level was set at $p \leq 0.05.$ ## 5. Conclusions Metformin attenuates the migration of M1 macrophages to the chondrocytes in vitro, and diminishes the infiltration of M1 macrophages in OA synovial tissues. Metformin plays a great therapeutic role in OA targeting M1 macrophages. ## References 1. Martel-Pelletier J., Barr A.J., Cicuttini F.M., Conaghan P.G., Cooper C., Goldring M.B., Goldring S.R., Jones G., Teichtahl A.J., Pelletier J.-P.. Osteoarthritis. Nat. Rev. Dis. Primers (2016) 2 16072. DOI: 10.1038/nrdp.2016.72 2. Glyn-Jones S., Palmer A.J.R., Agricola R., Price A.J., Vincent T.L., Weinans H., Carr A.J.. Osteoarthritis. Lancet (2015) 386 376-387. DOI: 10.1016/S0140-6736(14)60802-3 3. Sun Y., Zuo Z., Kuang Y.. An Emerging Target in the Battle against Osteoarthritis: Macrophage Polarization. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21228513 4. Sica A., Mantovani A.. Macrophage plasticity and polarization: In vivo veritas. J. Clin. Investig. (2012) 122 787-795. DOI: 10.1172/JCI59643 5. Behl T., Sharma E., Sehgal A., Kaur I., Kumar A., Arora R., Pal G., Kakkar M., Kumar R., Bungau S.. Expatiating the molecular approaches of HMGB1 in diabetes mellitus: Highlighting signalling pathways via RAGE and TLRs. Mol. Biol. Rep. (2021) 48 1869-1881. DOI: 10.1007/s11033-020-06130-x 6. Zhang H., Lin C., Zeng C., Wang Z., Wang H., Lu J., Liu X., Shao Y., Zhao C., Pan J.. Synovial macrophage M1 polarisation exacerbates experimental osteoarthritis partially through R-spondin-2. Ann. Rheum. Dis. (2018) 77 1524-1534. DOI: 10.1136/annrheumdis-2018-213450 7. Lv Z., Xu X., Sun Z., Yang Y.X., Guo H., Li J., Sun K., Wu R., Xu J., Jiang Q.. TRPV1 alleviates osteoarthritis by inhibiting M1 macrophage polarization via Ca/CaMKII/Nrf2 signaling pathway. Cell Death Dis. (2021) 12 504. DOI: 10.1038/s41419-021-03792-8 8. Samavedi S., Diaz-Rodriguez P., Erndt-Marino J.D., Hahn M.S.. A Three-Dimensional Chondrocyte-Macrophage Coculture System to Probe Inflammation in Experimental Osteoarthritis. Tissue Eng. Part A (2017) 23 101-114. DOI: 10.1089/ten.tea.2016.0007 9. Rena G., Hardie D.G., Pearson E.R.. The mechanisms of action of metformin. Diabetologia (2017) 60 1577-1585. DOI: 10.1007/s00125-017-4342-z 10. van den Bosch M.H.J.. Osteoarthritis year in review 2020: Biology. Osteoarthr. Cartil. (2021) 29 143-150. DOI: 10.1016/j.joca.2020.10.006 11. Lim Y.Z., Wang Y., Estee M., Abidi J., Udaya Kumar M., Hussain S.M., Wluka A.E., Little C.B., Cicuttini F.M.. Metformin as a potential disease-modifying drug in osteoarthritis: A systematic review of pre-clinical and human studies. Osteoarthr. Cartil. (2022) 30 1434-1442. DOI: 10.1016/j.joca.2022.05.005 12. Feng X., Chen W., Ni X., Little P.J., Xu S., Tang L., Weng J.. Metformin, Macrophage Dysfunction and Atherosclerosis. Front. Immunol. (2021) 12 682853. DOI: 10.3389/fimmu.2021.682853 13. Jing Y., Wu F., Li D., Yang L., Li Q., Li R.. Metformin improves obesity-associated inflammation by altering macrophages polarization. Mol. Cell. Endocrinol. (2018) 461 256-264. DOI: 10.1016/j.mce.2017.09.025 14. Lee M.K.S., Moore X.-L., Fu Y., Al-Sharea A., Dragoljevic D., Fernandez-Rojo M.A., Parton R., Sviridov D., Murphy A.J., Chin-Dusting J.P.F.. High-density lipoprotein inhibits human M1 macrophage polarization through redistribution of caveolin-1. Br. J. Pharmacol. (2016) 173 741-751. DOI: 10.1111/bph.13319 15. Cho D.-I., Kim M.R., Jeong H.-y., Jeong H.C., Jeong M.H., Yoon S.H., Kim Y.S., Ahn Y.. Mesenchymal stem cells reciprocally regulate the M1/M2 balance in mouse bone marrow-derived macrophages. Exp. Mol. Med. (2014) 46 e70. DOI: 10.1038/emm.2013.135 16. Gunassekaran G.R., Poongkavithai Vadevoo S.M., Baek M.-C., Lee B.. M1 macrophage exosomes engineered to foster M1 polarization and target the IL-4 receptor inhibit tumor growth by reprogramming tumor-associated macrophages into M1-like macrophages. Biomaterials (2021) 278 121137. DOI: 10.1016/j.biomaterials.2021.121137 17. Kawanami D., Takashi Y., Tanabe M.. Significance of Metformin Use in Diabetic Kidney Disease. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21124239 18. Reiss A.B., Miyawaki N., Moon J., Kasselman L.J., Voloshyna I., D’Avino R., De Leon J.. CKD, arterial calcification, atherosclerosis and bone health: Inter-relationships and controversies. Atherosclerosis (2018) 278 49-59. DOI: 10.1016/j.atherosclerosis.2018.08.046 19. Neven E., Vervaet B., Brand K., Gottwald-Hostalek U., Opdebeeck B., De Maré A., Verhulst A., Lalau J.-D., Kamel S., De Broe M.E.. Metformin prevents the development of severe chronic kidney disease and its associated mineral and bone disorder. Kidney Int. (2018) 94 102-113. DOI: 10.1016/j.kint.2018.01.027 20. Murray C.E., Coleman C.M.. Impact of Diabetes Mellitus on Bone Health. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20194873 21. Cai T.-T., Li H.-Q., Jiang L.-L., Wang H.-Y., Luo M.-H., Su X.-F., Ma J.-H.. Effects of GLP-1 Receptor Agonists on Bone Mineral Density in Patients with Type 2 Diabetes Mellitus: A 52-Week Clinical Study. Biomed. Res. Int. (2021) 2021 3361309. DOI: 10.1155/2021/3361309 22. Iorga R.A., Bacalbasa N., Carsote M., Bratu O.G., Stanescu A.M.A., Bungau S., Pantis C., Diaconu C.C.. Metabolic and cardiovascular benefits of GLP-1 agonists, besides the hypoglycemic effect (Review). Exp. Ther. Med. (2020) 20 2396-2400. DOI: 10.3892/etm.2020.8714 23. Egger A., Kraenzlin M.E., Meier C.. Effects of Incretin-Based Therapies and SGLT2 Inhibitors on Skeletal Health. Curr. Osteoporos. Rep. (2016) 14 345-350. DOI: 10.1007/s11914-016-0337-9 24. Radu A.-F., Bungau S.G., Negru P.A., Marcu M.F., Andronie-Cioara F.L.. In-depth bibliometric analysis and current scientific mapping research in the context of rheumatoid arthritis pharmacotherapy. Biomed. Pharmacother. (2022) 154 113614. DOI: 10.1016/j.biopha.2022.113614 25. Radu A.-F., Bungau S.G., Tit D.M., Behl T., Uivaraseanu B., Marcu M.F.. Highlighting the Benefits of Rehabilitation Treatments in Hip Osteoarthritis. Medicina (2022) 58. DOI: 10.3390/medicina58040494 26. Jang S., Lee K., Ju J.H.. Recent Updates of Diagnosis, Pathophysiology, and Treatment on Osteoarthritis of the Knee. Int. J. Mol. Sci. (2021) 22. DOI: 10.3390/ijms22052619 27. Scanzello C.R., Goldring S.R.. The role of synovitis in osteoarthritis pathogenesis. Bone (2012) 51 249-257. DOI: 10.1016/j.bone.2012.02.012 28. Wu C.L., Harasymowicz N.S., Klimak M.A., Collins K.H., Guilak F.. The role of macrophages in osteoarthritis and cartilage repair. Osteoarthr. Cartil. (2020) 28 544-554. DOI: 10.1016/j.joca.2019.12.007 29. Hügle T., Geurts J.. What drives osteoarthritis?-synovial versus subchondral bone pathology. Rheumatology (2017) 56 1461-1471. DOI: 10.1093/rheumatology/kew389 30. Udalova I.A., Mantovani A., Feldmann M.. Macrophage heterogeneity in the context of rheumatoid arthritis. Nat. Rev. Rheumatol. (2016) 12 472-485. DOI: 10.1038/nrrheum.2016.91 31. Li D., Ruan G., Zhang Y., Zhao Y., Zhu Z., Ou Q., Huang H., Chen J., Han W., Tang S.. Metformin attenuates osteoarthritis by targeting chondrocytes, synovial macrophages and adipocytes. Rheumatology (2022) keac467. DOI: 10.1093/rheumatology/keac467 32. Jafari M., Ghadami E., Dadkhah T., Akhavan-Niaki H.. PI3k/AKT signaling pathway: Erythropoiesis and beyond. J. Cell. Physiol. (2019) 234 2373-2385. DOI: 10.1002/jcp.27262 33. Fresno Vara J.A., Casado E., de Castro J., Cejas P., Belda-Iniesta C., González-Barón M.. PI3K/Akt signalling pathway and cancer. Cancer Treat. Rev. (2004) 30 193-204. DOI: 10.1016/j.ctrv.2003.07.007 34. Blume-Jensen P., Hunter T.. Oncogenic kinase signalling. Nature (2001) 411 355-365. DOI: 10.1038/35077225 35. Kim D.H., Kim J.Y., Yu B.P., Chung H.Y.. The activation of NF-kappaB through Akt-induced FOXO1 phosphorylation during aging and its modulation by calorie restriction. Biogerontology (2008) 9 33-47. DOI: 10.1007/s10522-007-9114-6 36. Yu T., Gan S., Zhu Q., Dai D., Li N., Wang H., Chen X., Hou D., Wang Y., Pan Q.. Modulation of M2 macrophage polarization by the crosstalk between Stat6 and Trim24. Nat. Commun. (2019) 10 4353. DOI: 10.1038/s41467-019-12384-2 37. Tao T., Luo D., Gao C., Liu H., Lei Z., Liu W., Zhou C., Qi D., Deng Z., Sun X.. Src Homology 2 Domain-Containing Protein Tyrosine Phosphatase Promotes Inflammation and Accelerates Osteoarthritis by Activating β-Catenin. Front. Cell. Dev. Biol. (2021) 9 646386. DOI: 10.3389/fcell.2021.646386 38. Abusarah J., Benabdoune H., Shi Q., Lussier B., Martel-Pelletier J., Malo M., Fernandes J.C., de Souza F.P., Fahmi H., Benderdour M.. Elucidating the Role of Protandim and 6-Gingerol in Protection Against Osteoarthritis. J. Cell. Biochem. (2017) 118 1003-1013. DOI: 10.1002/jcb.25659 39. Runge W.O., Ruppert D.S., Marcellin-Little D.J., Dahners L.E., Harrysson O.L.A., Weinhold P.S.. Bone changes after short-term whole body vibration are confined to cancellous bone. J. Musculoskelet. Neuronal Interact (2018) 18 485-492. PMID: 30511952 40. Krenn V., Morawietz L., Burmester G.R., Kinne R.W., Mueller-Ladner U., Muller B., Haupl T.. Synovitis score: Discrimination between chronic low-grade and high-grade synovitis. Histopathology (2006) 49 358-364. DOI: 10.1111/j.1365-2559.2006.02508.x 41. Glasson S.S., Chambers M.G., Van Den Berg W.B., Little C.B.. The OARSI histopathology initiative-recommendations for histological assessments of osteoarthritis in the mouse. Osteoarthr. Cartil. (2010) 18 S17-S23. DOI: 10.1016/j.joca.2010.05.025
--- title: Short-Term Decreasing and Increasing Dietary BCAA Have Similar, but Not Identical Effects on Lipid and Glucose Metabolism in Lean Mice authors: - Yuchen Sun - Bo Sun - Zhishen Wang - Yinfeng Lv - Qingquan Ma journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049642 doi: 10.3390/ijms24065401 license: CC BY 4.0 --- # Short-Term Decreasing and Increasing Dietary BCAA Have Similar, but Not Identical Effects on Lipid and Glucose Metabolism in Lean Mice ## Abstract Branched-chain amino acids (BCAA) showed multiple functions in glycolipid metabolism and protein synthesis. However, the impacts on the metabolic health of low or high dietary BCAA remain controversial due to the various experimental conditions. Gradient levels of BCAA were supplemented in lean mice for four weeks: 0BCAA (without BCAA), $\frac{1}{2}$BCAA (half BCAA), 1BCAA (regular BCAA), and 2BCAA (double BCAA). The results showed that the diet without BCAA caused energy metabolic disorders, immune defects, weight loss, hyperinsulinemia, and hyperleptinemia. $\frac{1}{2}$BCAA and 2BCAA diets reduced body fat percentage, but $\frac{1}{2}$ BCAA also decreased muscle mass. $\frac{1}{2}$BCAA and 2BCAA groups improved lipid and glucose metabolism by affecting metabolic genes. Meanwhile, significant differences between low and high dietary BCAA were observed. The results of this study provide evidence and reference for the controversy about dietary BCAA levels, which indicates that the main difference between low and high BCAA dietary levels may present in the longer term. ## 1. Introduction Obesity is a growing threat to human health worldwide, leading to severe metabolic complications, including diabetes mellitus, insulin and leptin resistance, and other metabolic disorders. A primary cause of obesity is improper food consumption, specifically of diets with high fat (HFD) [1]. Therefore, dietary intervention is an effective means to prohibit obesity. Amino acids (AAs), as critical nutritive substances in diets, can not only play a vital role in nutrition, but also exhibit various functions in immuno-stimulation, metabolism, or physiology. The proposal and application of the concept of functional amino acids have been proven effective in mitigating diseases impairing animal health [2]. For instance, dietary threonine supplementation notably reduces fat mass, improves insulin resistance, and facilitates lipid metabolic health in obese mice [3]. The increased dietary intake of aromatic amino acids (phenylalanine, tryptophan, and tyrosine) improved lipid metabolism and insulin tolerance by stimulating bile acid synthesis [4]. Moreover, leucine (Leu), isoleucine (Ile), and valine (Val), which are called branched-chain amino acids (BCAAs), are always research hot spots because of their association with type 2 diabetes, insulin resistance, and obesity [5,6]. Substantial evidence indicates that dietary BCAA supplementation can improve muscle development, insulin sensitivity, and lipid accumulation in high-fat, diet-induced, obese (DIO) mice [7,8,9]. However, the notion that BCAA supplementation could be positive for lipid metabolism or insulin function remains controversial [10]. Reduced consumption of BCAAs effectively antagonized obesity and insulin resistance [11]. A recent report showed that the adverse metabolic effects in lean mice were induced by Ile and Val, not Leu [12]. Yet, our previous study demonstrated that elevated dietary Leu, rather than Ile and Val, worsen the lipid metabolism in lean mice [13]. These results are affected by various experimental conditions, such as the mouse model, diet composition, or the duration of the test. Thus, more evidence is needed to verify the effects of dietary BCAA regulation under different conditions. In this study, we explore the effect of four different dietary BCAA levels (without, low, regular, and high BCAA levels included) by purified diet on lipid and glucose metabolism in normal lean mice under the same conditions. The total BCAA deprivation experiment was designed to indirectly investigate the role of BCAAs in body development, which was also a preliminary study for the following individual BCAA-removed trials. ## 2.1. Different Dietary BCAA Levels Influenced the Ratio of Muscle and WAT in Lean Mice Lean mice were provided diets with four different BCAA levels for four weeks (Figure 1A). There were no significant differences in body weight among the $\frac{1}{2}$BCAA, 1BCAA, and 2BCAA groups. Completely removing dietary BCAA significantly reduced body weight during the experimental period (Figure 1B). However, no marked difference in the average daily feed intake among the four groups was observed (Figure 1C). At the end of the fourth week, liver histopathological observation and relative liver weight showed no significant difference among experiment groups with different dietary BCAA levels (Figure 1D,E). The diet without BCAA significantly decreased the relative organ weight of the spleen, gastrocnemius muscle, and WAT, and increased the relative kidney weight. We also observed a lower relative weight of WAT in the $\frac{1}{2}$BCAA group compared with the 1BCAA and 2BCAA groups, a lower relative weight of gastrocnemius muscle and a higher relative weight of BAT in $\frac{1}{2}$BCAA and 2BCAA groups compared with 1BCAA group (Figure 1E). Half or doubled dietary BCAA levels showed the same trend in WAT and BAT. ## 2.2. Complete Deprivation of Dietary BCAA Impaired Glucose Homeostasis Different levels of BCAA had no significant effect on glucose homeostasis in lean mice (Figure 2A,B). However, non-BCAA and low-BCAA diets decreased the area under the curve (AUC) of ITT (Figure 2C,D). BCAA removal reduced the GLU level and increased the serum insulin level, thus inducing a poor performance in HOMA-IR (Figure 2E–G). The 2BCAA group had a lower GLU level and higher insulin level in the serum compared with the $\frac{1}{2}$BCAA and 1BCAA groups, which did not affect the HOMA-IR and ISI (Figure 2E–H). ## 2.3. Different Dietary BCAA Levels Altered the Serum Indexes in Lean Mice Different dietary BCAA levels had no significant influence on serum TC, TG, and HDL-C in lean mice (Figure 3). Mice fed with a diet without BCAA had a lower LDL-C and ADPN, and higher FFA and LEP in the serum compared with the control group. Half the dietary BCAA level elevated FFA, ADPN, and LEP serum levels. Compared with the 1BCAA group, a 2BCAA diet decreased BUN and ADPN levels and increased LEP level. Besides, the FFA and ADPN levels in the $\frac{1}{2}$BCAA group were superior to those in the 2BCAA group. ## 2.4. Different Dietary BCAA Levels Have Different Effects on Metabolic Genes To further determine the variations in transcriptome among groups with different dietary BCAA levels, transcriptome analyses were performed. The volcano plot showed differentially expressed genes in the liver of mice (Figure S1A). As shown in the violin plot, there was no distinct difference in the abundance of differentially expressed genes (DEGs) in different treatments (Figure S1B). The heatmap and hierarchical clusters analysis (Figure S1C) highlighted significant differences in the expression levels of the top 50 DEGs among the four groups. The counts of DEGs in 0BCAA vs. 1BCAA (455 up-regulated genes and 1127 down-regulated genes), $\frac{1}{2}$BCAA vs. 1BCAA (175 up-regulated genes and 188 down-regulated genes), 2BCAA vs. 1BCAA (327 up-regulated genes and 107 down-regulated genes), and $\frac{1}{2}$BCAA vs. 2BCAA (63 up-regulated genes and 203 down-regulated genes) were demonstrated (Figure S1D). GO functional enrichment analysis was performed to investigate the potential functions of the DEGs (Figure 4). Enriched GO terms targeted by DEGs between 0BCAA and 1BCAA were mainly related to an immune function and inflammation. Insulin secretion and response to insulin were also significantly enriched. Besides, terms related to lipid metabolism, carbohydrate metabolism, and muscle development were markedly enriched among $\frac{1}{2}$BCAA, 1BCAA, and 2BCAA groups: “fat cell differentiation”, “fatty acid metabolic process”, “lipid catabolic process”, “cholesterol transport”, “skeletal system development”, “muscle tissue development”, “glycogen (starch) synthase activity”, and so on. KEGG enrichment analysis revealed that dietary BCAA levels could influence the pathways involved in lipid, carbohydrate, protein, and insulin metabolism (Figure 5), including “fatty acid metabolism”, “fat digestion and absorption”, “cholesterol metabolism”, “bile secretion”, “starch and sucrose metabolism”, “carbohydrate digestion and absorption”, “protein digestion and absorption”, “PPAR signaling pathway”, “pancreatic secretion”, “insulin signaling pathway”, and “insulin resistance”. Seventeen genes related to lipid and glucose metabolism affected by different BCAA levels were selected for validation through real-time PCR (Table 1). The results of real-time PCR validation are shown in Figure 6A, consistent with the transcriptomic analyses, which proved that the results from transcriptomic analyses can be trusted. The secondary verification was performed by western blot, and the relative protein expression of HSL agreed with the data in the study (Figure 6B,C). ## 3. Discussion Our previous studies demonstrated that BCAAs had different impacts on lipid metabolism in mice under different conditions (lean or obese) [9,13], which was also proved by other researchers [12]. However, the results from these trials always had more or less differences. The never-ceased controversies highlight the importance of test conditions. In this study, the effects of $\frac{1}{2}$BCAA and 2BCAA levels on lipid and glucose metabolism in lean mice are considered favorable, but achieved through different mechanisms. Meanwhile, the deprivation of total BCAA caused severe immune and metabolic defects. Furthermore, almost no significant linear regression relationship between dietary levels and each physiological change was observed. Mice without BCAA intake kept losing weight during the whole group, showing the importance of BCAAs in body development. Meanwhile, the feed intake in the 0BCAA group was as much as the other groups, suggesting their collapsed energy metabolism. The nutritional and metabolic block forced the undernourished mice to break down fats and muscle to meet the energy requirement, which caused the relative decrease in muscle weight, WAT weight, and serum glucose levels, and increased serum FFA. The decreased relative spleen weight after BCAA deprivation may be the main sign of health problems. Transcriptome analysis also indicated that the diet without BCAA triggered marked suppression of the immune system. The present results are not surprising because the deficiency of essential amino acids results in impaired metabolism and immune function [14]. However, researchers paid little attention to the impact of BCAAs on immune resistance. When dealing with diabetes or obesity through dietary BCAA intervention, the immune affections of BCAAs should be considered together. Lipid accumulation is an obvious sign of obesity development, which is also the direct reflection of BCAAs affecting lipid metabolism. Previous studies demonstrated that BCAA supplementation maintained muscle mass and inhibited obesity development in mice [7,15,16,17]. However, it has been reported that reducing BCAA consumption decreased WAT and muscle mass [11,12]. No changes in body weight and body composition were observed when intervening with leucine [18]. We observed that $\frac{1}{2}$BCAA and 2BCAA dietary levels both reduced relative WAT weight and increased relative BAT weight in lean mice. In addition, the reduced muscle percentage in the $\frac{1}{2}$BCAA group confirmed that BCAA promotes muscle development and suppresses muscle loss [19]. Leptin has an anti-obesity effect that decreases adiposity but maintains muscle mass [20,21]. Serum leptin levels elevated by $\frac{1}{2}$BCAA and 2BCAA diets were perhaps a reason for the WAT loss. In addition, BAT is a thermogenic organ that enhances fat consumption and regulates BCAA catabolism [22]. Meanwhile, branched-chain aminotransferase (Bcat2), the rate-limiting BCAA catabolic enzyme, was up-regulated in both $\frac{1}{2}$BCAA and 2BCAA groups. The reason could be that both lowering and heightening dietary levels activated BCAA catabolism in healthy mice. Moreover, few genes related to the browning of WAT in lean mice were affected by different dietary BCAA levels. Insulin resistance is closely related to the incidence of type 2 diabetes. Supplementation of dietary BCAAs could moderately weaken the insulin sensitivity and metabolic efficiency of the mice but cannot induce insulin resistance [23,24]. We found that halving and doubling dietary BCAA levels did not influence glucose homeostasis in healthy lean mice, but high BCAA level showed the facilitation effect of BCAAs on insulin secretion [25]. Complete deprivation of BCAAs leads to hyperinsulinemia, hyperleptinemia, and higher HOMA-IR, which may not mean insulin resistance or leptin resistance, but because of the abnormal health and metabolism status in mice from the 0BCAA group. Still, hyperinsulinemia and hyperleptinemia were undoubtedly involved in the physical degeneration in the 0BCAA group. Different BCAA levels changed genes associated with lipid and glucose metabolism differently. Fibroblast growth factor 21 (Fgf21) promotes lipid metabolism, hepatic gluconeogenesis and insulin sensitivity without affecting hepatic glycogen breakdown [26,27,28]. Additionally, Fgf21 needs to take effect with enough adipose tissue as an indispensable mediator [29], which explains the low Fgf21 level in the 0BCAA group. However, Fgf21 in the $\frac{1}{2}$BCAA and 2BCAA groups was not significantly affected, assuming Fgf21 may not be the most critical signaling pathway in the liver when facing BCAA treatment. Fatty acid synthase (Fasn) is a complex multifunctional enzyme that plays an essential role in the synthesis of fatty acids, which was only suppressed in the 2BCAA group, but not the $\frac{1}{2}$BCAA group. The unique up-regulation of peroxisome proliferator-activated receptor alpha (Pparα), and uncoupling protein 2 (Ucp2) that regulates lipid metabolism, proved the positive effects of high BCAA levels. Nevertheless, the increased expression level of the Pparα–Fabp1 axis was always found in DIO mice, and the suppression of the Pparα–Fabp1 axis was considered as a treatment for nonalcoholic steatohepatitis [30], suggesting the potential threat of a high-BCAA diet. Carboxyl ester lipase (Cel) and colipase (Clps) control the hydrolysis and absorption of lipids, which were down-regulated in DIO mice [31] and up-regulated by increasing dietary BCAA levels in lean mice. Additionally, high dietary BCAA levels stimulated glycogen metabolism by enhancing glycogenesis (glycogen synthase 2, Gys2 [32]) and glycogenolysis (hexokinase 2, Hk2 [33]; muscle glycogen phosphorylase, Pygm [34]). Intriguingly, though the improvement of half-dietary BCAAs on genes involved in lipid and glucose metabolism is not as significant as that in the 2BCAA group, the $\frac{1}{2}$BCAA group significantly ameliorated insulin sensitivity. The loss of cytochrome P450 family 8 subfamily B member 1 (Cyp8b1) was proved to improve glucose homeostasis [35], and the up-regulation of Acat2 could reduce toxic polar lipids and improve insulin sensitivity [36], which may explain the better glucose homeostasis in many studies when reducing dietary BCAA levels. Taken together, the difference in gene expression between the $\frac{1}{2}$BCAA and 2BCAA groups may be the key to clarifying the mechanism when using dietary BCAA intervention. These differences may be enlarged after four weeks and perhaps reverse their advantages in metabolism, which warrants further studies. In summary, complete deprivation of BCAAs resulted in metabolic disorders and finally caused the development of weight loss, hyperinsulinemia, and hyperleptinemia in lean mice within four weeks. Increasing and decreasing dietary BCAA levels reduced the body fat rate, but only maintained muscle mass in mice through the high BCAA diet. Further, doubled and half dietary BCAA levels improved metabolic status via regulating lipid and glucose metabolism-related genes. The differences between the $\frac{1}{2}$BCAA and 2BCAA groups may provide references for the controversy about BCAA intervention. At least in the present study, low and high BCAA levels in the diet achieved favorable effects on metabolic health in lean mice before the fourth week. Further research is needed to determine the impact of single BCAA removal or long-term BCAA regulation. ## 4.1. Animals and Diets The study was approved by Northeast Agricultural University Animal Science Research Ethics Committee (NEAU-[2011]-9) (Approval Code: NEAUEC20200202; Approval Date: 3 April 2020). All mice were purchased from HFK Biotechnology Co., Ltd. (Beijing, China) and housed in a temperature-controlled (22 ± 2 °C) and humidity-controlled (55 ± $5\%$) environment, on a 12 h light/dark cycle with free access to food and water. Before the beginning of the experiment, a total of 60 eight-week-old male C57BL/6J mice were acclimated to a control diet (amino acid-customized diet) for 7 days and then randomly divided into following diets for 4 weeks: 0BCAA (without BCAA), $\frac{1}{2}$BCAA (half the normal BCAA level), 1BCAA (normal BCAA level, control group), and 2BCAA (twice the normal BCAA level). Mice selected for each group have a similar average initial body weight. Body weight and food intake were recorded weekly. To formulate isonitrogenous and isoenergetic diets, and minimize the interference of other amino acids, dietary nitrogen was balanced by proportionally supplementing amino acid mixtures, except for BCAAs (Table 2). ## 4.2. Glucose Tolerance Test (GTT) and Insulin Tolerance Test (ITT) In the third week, the glucose tolerance test (GTT) and insulin tolerance test (ITT) were performed following an intraperitoneal glucose (2 g/kg) and insulin injections (0.75 units/kg) after 12 h or 6 h starvation of the mice, respectively. Blood samples were collected from the tail vein at 0, 15, 30, 60 and 120 min after the injection, and glucose levels were measured with a glucose meter (Roche Diagnostics, Shanghai, China). The fasting serum glucose levels were determined by a colorimetric assay (Nanjing Jiancheng Bioengineering Institute, Nanjing, China), and insulin levels were assessed using a commercial ELISA kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China), according to the manufacturer’s instructions. The status of insulin resistance was applied by a homeostasis model assessment of insulin resistance (HOMA-IR) and the improved insulin sensitivity index (ISI). The HOMA-IR index was calculated using the following formula: HOMA-IR = [fasting glucose levels (mmol/L)] × [fasting serum insulin (mU/L)]/22.5. The ISI × 100 index was calculated using the following formula: ISI × 100 = 1/[fasting glucose levels (mmol/L)] × [fasting serum insulin (mU/L)] × 100. ## 4.3. Sample Collection At the end of the experiment, the mice were deprived of food overnight. Blood samples were obtained from the orbital vein in mice anesthetized with ether. Blood was centrifuged at 3000× g for 15 min at 4 °C, to collect serum that was stored at −80 °C until use. The mice were sacrificed by neck breaking. The liver, spleen, kidney, white adipose tissue (WAT), brown adipose tissue (BAT), and gastrocnemius muscle were quickly removed, weighed and immediately frozen in liquid nitrogen and stored at −80 °C until analysis. ## 4.4. Histological Analysis Histological analysis of the liver in mice was carried out with the H&E staining method [9]. The liver fragments fixed in $4\%$ paraformaldehyde buffer were dehydrated and embedded in paraffin to produce random 6 μm thick cuts and stained with hematoxylin and eosin for visualization through an Olympus microscope (Tokyo, Japan). ## 4.5. Serum Parameter Determination Serum levels of triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), glucose, free fatty acids (FFA), and blood urea nitrogen (BUN) were determined by enzymatic methods, using commercial diagnostics kits (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). The adiponectin (ADPN) and leptin (LEP) levels were assessed using a commercial ELISA kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China). ## 4.6. Transcriptomic Analysis and Data Processing The total RNA samples isolated from experimental groups were subjected to gene expression analysis. The Qubit®RNA Assay Kit in Qubit®2.0 Flurometer (Life Technologies, Carlsbad, CA, USA) was used to measure the RNA concentration, the RNA Nano 6000Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, Santa Clara, CA, USA) was used to assess RNA integrity. The sequencing libraries were generated using NEBNext®UltraTMRNA Library Prep Kit for Illumina® (NEB, Ipswich, MA, USA), following the manufacturer’s recommendations, and index codes were added to attribute the sequences to each sample. Then, mRNA was purified, and its purity was re-assessed using poly-oligo-attached magnetic beads. A stringent significance threshold (p-value < 0.05) was used to limit the false-positive findings, and the fold change ≥1 was chosen to screen a manageable number of genes. The Gene Ontology (GO) Database was used to detect the significant function of differentially expressed genes (DEG) from the level of biological process (BP), cellular component (CC) and molecular function (MF). The canonical pathways of DEGs were analyzed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. ## 4.7. Quantitative Real-Time PCR Analysis The total RNA of liver tissue samples was extracted using the Trizol reagent (TaKaRa, Dalian, China). Before cDNA synthesis, the purity and concentration of RNA were evaluated by absorbance at $\frac{260}{280}$ nm. The cDNA was obtained by reverse transcription using PrimeScriptTM RT kit (Takara, Dalian, China). Quantitative real-time PCR was performed using an SYBR Premix Ex TaqTM Kit (Takara, Dalian, China) on an ABI 7500 Fast Real-Time PCR System (Applied Biosystems, Bedford, MA, USA). Calculations were performed using a comparative method (2−ΔΔCt), and β-actin was used as the internal control. The sequences of primers for the PCR are shown in Table 3. ## 4.8. Western Blot Analysis The tissues were homogenized at 4 °C in RIPA lysis buffer with a $1\%$ protease inhibitor cocktail, as described previously [9]. The tissue homogenates were centrifuged (12,000× g, 15 min, 4 °C), and a Pierce BCA Protein Assay Kit (Thermo Scientific, Waltham, MA, USA) was used to determine the total protein concentration of lysate. Equal amounts of proteins were diluted with the loading buffer and heated in boiling water for 5 min. The protein sample was separated on a $12\%$ SDS−polyacrylamide gel electrophoresis and transferred to PVDF membranes overnight. After blocking with $3\%$ BSA in Tris−Tween buffered saline for 2 h at room temperature, the membranes were incubated with the following antibodies overnight at 4 °C: fatty acid synthase (FAS) (1:1000, Absin) and fibroblast growth factor 21 (FGF21) (1:1000, Absin). The membranes were washed with TBST three times and incubated for 3 h with the peroxidase-conjugated secondary antibody (1:2000, Cell Signaling). The band was visualized by ECL detection systems (Thermo Scientific, Waltham, MA, USA). The images were detected on a Fujifilm LAS-3000 (Tokyo, Japan). The Image J software 1.80 (National Institutes of Health, Bethesda, MD, USA) was used to quantify the protein densitometry. ## 4.9. Statistical Analysis Data were analyzed with the one-way analysis of variance (ANOVA), using the general linear model procedure in SPSS 26.0 (IBM SPSS Statistics, Chicago, IL, USA, 2019), Image J 1.80 (National Institutes of Health, Bethesda, MD, USA, 2021), and GraphPad Prism 8.3.0 (GraphPad Software, La Jolla, CA, USA, 2012). For multiple comparisons, a Tukey’s multiple comparison test, LSD multiple-comparison test, and Duncan’s multiple comparison test were used. The results are presented as the mean and standard error of the mean for the effect of protein restriction. Differences were considered significantly different at $p \leq 0.05$, with a trend toward significance at $p \leq 0.10.$ ## References 1. Miyamoto J., Igarashi M., Watanabe K., Karaki S.-I., Mukouyama H., Kishino S., Li X., Ichimura A., Irie J., Sugimoto Y.. Gut microbiota confers host resistance to obesity by metabolizing dietary polyunsaturated fatty acids. Nat. Commun. (2019) 10 4007. DOI: 10.1038/s41467-019-11978-0 2. Wu G.. Functional amino acids in nutrition and health. Amino Acids (2013) 45 407-411. DOI: 10.1007/s00726-013-1500-6 3. Ma Q., Zhou X., Sun Y., Hu L., Zhu J., Shao C., Meng Q., Shan A.. Threonine, but Not Lysine and Methionine, Reduces Fat Accumulation by Regulating Lipid Metabolism in Obese Mice. J. Agric. Food Chem. (2020) 68 4876-4883. DOI: 10.1021/acs.jafc.0c01023 4. Ma Q., Chen J., Zhou X., Hu L., Sun Y., Wang Z., Yue Z., Shan A.. Dietary supplementation with aromatic amino acids decreased triglycerides and alleviated hepatic steatosis by stimulating bile acid synthesis in mice. Food Funct. (2020) 12 267-277. DOI: 10.1039/D0FO02364G 5. McCormack S.E., Shaham O., McCarthy M.A., Deik A.A., Wang T.J., Gerszten R.E., Clish C.B., Mootha V.K., Grinspoon S.K., Fleischman A.. Circulating branched-chain amino acid concentrations are associated with obesity and future insulin resistance in children and adolescents. Pediatr. Obes. (2012) 8 52-61. DOI: 10.1111/j.2047-6310.2012.00087.x 6. Rousseau M., Guénard F., Garneau V., Allam-Ndoul B., Lemieux S., Pérusse L., Vohl M.-C.. Associations Between Dietary Protein Sources, Plasma BCAA and Short-Chain Acylcarnitine Levels in Adults. Nutrients (2019) 11. DOI: 10.3390/nu11010173 7. Zhang Y., Guo K., LeBlanc R.E., Loh D., Schwartz G.J., Yu Y.-H.. Increasing Dietary Leucine Intake Reduces Diet-Induced Obesity and Improves Glucose and Cholesterol Metabolism in Mice via Multimechanisms. Diabetes (2007) 56 1647-1654. DOI: 10.2337/db07-0123 8. Eller L.K., Saha D.C., Shearer J., Reimer R.A.. Dietary leucine improves whole-body insulin sensitivity independent of body fat in diet-induced obese Sprague–Dawley rats. J. Nutr. Biochem. (2013) 24 1285-1294. DOI: 10.1016/j.jnutbio.2012.10.004 9. Ma Q., Zhou X., Hu L., Chen J., Zhu J., Shan A.. Leucine and isoleucine have similar effects on reducing lipid accumulation, improving insulin sensitivity and increasing the browning of WAT in high-fat diet-induced obese mice. Food Funct. (2020) 11 2279-2290. DOI: 10.1039/C9FO03084K 10. Newgard C.B., An J., Bain J.R., Muehlbauer M.J., Stevens R.D., Lien L.F., Haqq A., Shah S., Arlotto M., Slentz C.. A Branchedchain Amino Acid-Related Metabolic Signature that Differentiates Obese and Lean Humans and Contributes to Insulin Resistance. Cell Metab. (2009) 9 565-566. DOI: 10.1016/j.cmet.2009.05.001 11. Fontana L., Cummings N.E., Apelo S.I.A., Neuman J.C., Kasza I., Schmidt B.A., Cava E., Spelta F., Tosti V., Syed F.A.. Decreased Consumption of Branched-Chain Amino Acids Improves Metabolic Health. Cell Rep. (2016) 16 520-530. DOI: 10.1016/j.celrep.2016.05.092 12. Yu D., Richardson N.E., Green C.L., Spicer A.B., Murphy M.E., Flores V., Jang C., Kasza I., Nikodemova M., Wakai M.H.. The Adverse Metabolic Effects of Branched-Chain Amino Acids Are Mediated by Isoleucine and Valine. Cell Metab. (2021) 33 905-922.e6. DOI: 10.1016/j.cmet.2021.03.025 13. Zhou X., Chen J., Sun B., Wang Z., Zhu J., Yue Z., Zhang Y., Shan A., Ma Q., Wang J.. Leucine, but not isoleucine or valine, affects serum lipid profiles and browning of WAT in mice. Food Funct. (2021) 12 6712-6724. DOI: 10.1039/D1FO00341K 14. Hou Q., Zhao F., Liu W., Lv R., Khine W.W.T., Han J., Sun Z., Lee Y.-K., Zhang H.. Probiotic-directed modulation of gut microbiota is basal microbiome dependent. Gut Microbes (2020) 12 1736974. DOI: 10.1080/19490976.2020.1736974 15. Nishimura J., Masaki T., Arakawa M., Seike M., Yoshimatsu H.. Isoleucine Prevents the Accumulation of Tissue Triglycerides and Upregulates the Expression of Pparalpha and Uncoupling Protein in Diet-Induced Obese Mice. J. Nutr. (2010) 140 496-500. DOI: 10.3945/jn.109.108977 16. Kamei Y., Hatazawa Y., Uchitomi R., Yoshimura R., Miura S.. Regulation of Skeletal Muscle Function by Amino Acids. Nutrients (2020) 12. DOI: 10.3390/nu12010261 17. O’Rielly R., Li H., Lim S.M., Yazbeck R., Kritas S., Ullrich S.S., Feinle-Bisset C., Heilbronn L., Page A.J.. The Effect of Isoleucine Supplementation on Body Weight Gain and Blood Glucose Response in Lean and Obese Mice. Nutrients (2020) 12. DOI: 10.3390/nu12082446 18. Binder E., Bermúdez-Silva F.J., André C., Elie M., Romero-Zerbo S.Y., Leste-Lasserre T., Belluomo L., Duchampt A., Clark S., Aubert A.. Leucine Supplementation Protects from Insulin Resistance by Regulating Adiposity Levels. PLoS ONE (2013) 8. DOI: 10.1371/journal.pone.0074705 19. Magne H., Savary-Auzeloux I., Rémond D., Dardevet D.. Nutritional strategies to counteract muscle atrophy caused by disuse and to improve recovery. Nutr. Res. Rev. (2013) 26 149-165. DOI: 10.1017/S0954422413000115 20. Buettner C., Muse E.D., Cheng A., Chen L., Scherer T., Pocai A., Su K., Cheng B., Li X., Harvey-White J.. Leptin Controls Adipose Tissue Lipogenesis Via Central, Stat3-Independent Mechanisms. Nat. Med. (2008) 14 667-675. DOI: 10.1038/nm1775 21. Yoshida R., Noguchi K., Shigemura N., Jyotaki M., Takahashi I., Margolskee R.F., Ninomiya Y.. Leptin Suppresses Mouse Taste Cell Responses to Sweet Compounds. Diabetes (2015) 64 3751-3762. DOI: 10.2337/db14-1462 22. Yoneshiro T., Wang Q., Tajima K., Matsushita M., Maki H., Igarashi K., Dai Z., White P.J., McGarrah R.W., Ilkayeva O.R.. BCAA catabolism in brown fat controls energy homeostasis through SLC25A44. Nature (2019) 572 614-619. DOI: 10.1038/s41586-019-1503-x 23. Maida A., Chan J.S., Sjøberg K.A., Zota A., Schmoll D., Kiens B., Herzig S., Rose A.J.. Repletion of branched chain amino acids reverses mTORC1 signaling but not improved metabolism during dietary protein dilution. Mol. Metab. (2017) 6 873-881. DOI: 10.1016/j.molmet.2017.06.009 24. Lee J., Vijayakumar A., White P.J., Xu Y., Ilkayeva O., Lynch C.J., Newgard C.B., Kahn B.B.. BCAA Supplementation in Mice with Diet-induced Obesity Alters the Metabolome without Impairing Glucose Homeostasis. Endocrinology (2021) 162 bqab062. DOI: 10.1210/endocr/bqab062 25. Wang T.J., Larson M.G., Vasan R.S., Cheng S., Rhee E.P., McCabe E., Lewis G.D., Fox C.S., Jacques P.F., Fernandez C.. Metabolite Profiles and the Risk of Developing Diabetes. Nat. Med. (2011) 17 448-453. PMID: 21423183 26. Kharitonenkov A., Shiyanova T.L., Koester A., Ford A.M., Micanovic R., Galbreath E., Sandusky G., Hammond L., Moyers J., Owens R.. Fgf-21 as a Novel Metabolic Regulator. J. Clin. Investig. (2005) 115 1627-1635. DOI: 10.1172/JCI23606 27. Li H., Wu G., Fang Q., Zhang M., Hui X., Sheng B., Wu L., Bao Y., Li P., Xu A.. Fibroblast growth factor 21 increases insulin sensitivity through specific expansion of subcutaneous fat. Nat. Commun. (2018) 9 272. DOI: 10.1038/s41467-017-02677-9 28. Barb D., Bril F., Kalavalapalli S., Cusi K.. Plasma Fibroblast Growth Factor 21 Is Associated with Severity of Nonalcoholic Steatohepatitis in Patients with Obesity and Type 2 Diabetes. J. Clin. Endocrinol. Metab. (2019) 104 3327-3336. DOI: 10.1210/jc.2018-02414 29. Veniant M.M., Hale C., Helmering J., Chen M.M., Stanislaus S., Busby J., Vonderfecht S., Xu J., Lloyd D.J.. FGF21 Promotes Metabolic Homeostasis via White Adipose and Leptin in Mice. PLoS ONE (2012) 7. DOI: 10.1371/journal.pone.0040164 30. Yan T., Luo Y., Yan N., Hamada K., Zhao N., Xia Y., Wang P., Zhao C., Qi D., Yang S.. Intestinal Peroxisome Proliferator-Activated Receptor A—Fatty Acid-Binding Protein 1 Axis Modulates Nonalcoholic Steatohepatitis. Hepatology (2022) 77 239-255. DOI: 10.1002/hep.32538 31. Ali F., Ismail A., Esa N.M., Pei C.P., Kersten S.. Hepatic genome-wide expression of lipid metabolism in diet-induced obesity rats treated with cocoa polyphenols. J. Funct. Foods (2015) 17 969-978. DOI: 10.1016/j.jff.2015.06.047 32. Flannery C.A., Choe G.H., Cooke K.M., Fleming A.G., Radford C.C., Kodaman P.H., Jurczak M.J., Kibbey R.G., Taylor H.S.. Insulin Regulates Glycogen Synthesis in Human Endometrial Glands through Increased GYS2. J. Clin. Endocrinol. Metab. (2018) 103 2843-2850. DOI: 10.1210/jc.2017-01759 33. Bramer S.A., Macedo A., Klein C.. Hexokinase 2 drives glycogen accumulation in equine endometrium at day 12 of diestrus and pregnancy. Reprod. Biol. Endocrinol. (2017) 15 4. DOI: 10.1186/s12958-016-0223-4 34. Migocka-Patrzałek M., Lewicka A., Elias M., Daczewska M.. The effect of muscle glycogen phosphorylase (Pygm) knockdown on zebrafish morphology. Int. J. Biochem. Cell Biol. (2019) 118 105658. DOI: 10.1016/j.biocel.2019.105658 35. Kaur A., Patankar J.V., de Haan W., Ruddle P., Wijesekara N., Groen A.K., Verchere C.B., Singaraja R.R., Hayden M.R.. Loss of. Diabetes (2014) 64 1168-1179. DOI: 10.2337/db14-0716 36. Wang Y.-J., Bian Y., Luo J., Lu M., Xiong Y., Guo S.-Y., Yin H.-Y., Lin X., Li Q., Chang C.C.Y.. Cholesterol and fatty acids regulate cysteine ubiquitylation of ACAT2 through competitive oxidation. Nature (2017) 19 808-819. DOI: 10.1038/ncb3551
--- title: PCSK9 Inhibitors Reduce PCSK9 and Early Atherogenic Biomarkers in Stimulated Human Coronary Artery Endothelial Cells authors: - Rahayu Zulkapli - Suhaila Abd Muid - Seok Mui Wang - Hapizah Nawawi journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049668 doi: 10.3390/ijms24065098 license: CC BY 4.0 --- # PCSK9 Inhibitors Reduce PCSK9 and Early Atherogenic Biomarkers in Stimulated Human Coronary Artery Endothelial Cells ## Abstract Despite reports on the efficacy of proprotein convertase subtilisin-Kexin type 9 (PCSK9) inhibitors as a potent lipid-lowering agent in various large-scale clinical trials, the anti-atherogenic properties of PCSK9 inhibitors in reducing PCSK9 and atherogenesis biomarkers via the NF-ĸB and eNOS pathway has yet to be established. This study aimed to investigate the effects of PCSK9 inhibitors on PCSK9, targeted early atherogenesis biomarkers, and monocyte binding in stimulated human coronary artery endothelial cells (HCAEC). HCAEC were stimulated with lipopolysaccharides (LPS) and incubated with evolocumab and alirocumab. The protein and gene expression of PCSK9, interleukin-6 (IL-6), E-selectin, intercellular adhesion molecule 1 (ICAM-1), nuclear factor kappa B (NF-ĸB) p65, and endothelial nitric oxide synthase (eNOS) were measured using ELISA and QuantiGene plex, respectively. The binding of U937 monocytes to endothelial cell capacity was measured by the Rose Bengal method. The anti-atherogenic effects of evolocumab and alirocumab were contributed to by the downregulation of PCSK9, early atherogenesis biomarkers, and the significant inhibition of monocyte adhesion to the endothelial cells via the NF-ĸB and eNOS pathways. These suggest the beyond cholesterol-lowering beneficial effects of PCSK9 inhibitors in impeding atherogenesis during the initial phase of atherosclerotic plaque development, hence their potential role in preventing atherosclerosis-related complications. ## 1. Introduction Various atherogenesis biomarkers of early atherosclerosis have been proposed to predict cardiovascular events and mechanisms. The early stage of atherosclerosis development involves inflammation, endothelial injury, and endothelial activation before the formation of atherosclerotic plaques [1]. The overexpression of proinflammatory cytokines (interleukin-6; IL-6) and adhesion molecules (intercellular adhesion molecule 1; ICAM-1 and E-selectin) by endothelial cells, mediated via nuclear factor-kappa beta (NF-ĸB) activation, will occur before the adherence of circulating monocytes on endothelial cells [2,3,4]. Moreover, inflammation, endothelial activation, and oxidative stress have been identified as critical events in the initiation and progression of atherosclerosis [5]. In July and August 2015, the FDA approved novel lipid-lowering therapy (LLT), namely alirocumab (human IgG1/κ mAb, genetically engineered in Chinese hamster ovary cells) and evolocumab (fully human IgG2 mAb), which aimed to reduce low-density lipoprotein cholesterol (LDL-C) [6,7,8]. These novel therapies are monoclonal antibodies that inactivate the proprotein convertase subtilisin-Kexin type 9 (PCSK9) enzyme that binds to LDLR to initiate the reuptake and catabolism of low-density lipoprotein cholesterol receptor (LDLR) by body cells. Diminished LDLR will reduce LDL-C uptake from the blood into the cells, causing increased LDL-C levels in the blood and leading to an increased risk of coronary artery disease (CAD). The PCSK9 inhibitor is a lipid-lowering agent prescribed for severe hypercholesterolemia patients. Alirocumab and evolocumab have shown striking LDL-C reductions of up to $60\%$ and are documented to be relatively safe [6,9,10,11]. PCSK9-bound LDLR will degrade within the cell cytoplasm, thus preventing it from being recycled to the cell surface for LDL-C reuptake [12,13]. In normal conditions, PCSK9 is co-expressed with LDLR by the same transcription factor (sterol-regulatory element-binding protein 2) [14] to modulate the cellular reuptake of circulatory LDL-C so that the LDL-C in plasma is not depleted by the cells. By inhibiting the PCSK9, the LDLR will reuptake the circulatory LDL-C unhindered, significantly reducing the LDL-C level [15,16]. There are various reports on the efficacy of PCSK9 inhibitors in LDL-C lowering in a variety of comprehensive large-scale clinical trials such as PROFICIO (Program to Reduce LDL-C and Cardiovascular Outcomes Following Inhibition of PCSK9 in Different Populations) [17,18,19,20,21,22], FOURIER (Further Cardiovascular Outcomes Research With PCSK9 Inhibition in Subjects with Elevated Risk) [23,24,25] and ODYSSEY (Cardiovascular Outcomes After an Acute Coronary Syndrome) [26,27,28]. PCSK9 inhibition attenuates atherosclerosis progression and lowers the risk for acute cardiovascular events [29,30,31]. In addition, preclinical studies also revealed that PCSK9 has pleiotropic effects other than plasma LDL-C regulation and could be a key molecule in the pathophysiology of atherosclerosis [32]. However, the molecular mechanisms through which PCSK9 inhibition might confer atheroprotection beyond low-density lipoprotein (LDL) lowering remain scarce. Furthermore, inflammation-stimulating agents, lipopolysaccharides (LPS), have been shown to increase PCSK9 expression significantly and act as an inflammatory trigger, which causes an increase in the uptake of oxidized LDL into the subendothelial space [33,34,35], mimicking the process of atheroma formation in the blood vessels. Therefore, this study aimed to investigate the effects of PCSK9 inhibitors on PCSK9 and early atherogenesis biomarker expression of inflammation, endothelial activation, oxidative stress, and monocyte-endothelial binding in LPS-stimulated HCAEC compared to controls. ## 2.1. Cytotoxic Effects of PCSK9 Inhibitors on HCAEC The incubation of LPS-stimulated HCAEC with PCSK9 inhibitors (evolocumab or alirocumab) at 1, 10, and 100 µg/mL for 24 h exhibited more than $90\%$ cell viability (Figure 1). Similarly, the incubation of LPS-stimulated HCAEC with media (untreated) and LPS alone (negative control) for 24 h also did not show any reduction in cell viability (Figure 1). Therefore, the selected concentrations of PCSK9 inhibitors were regarded as non-toxic to HCAEC, and they were subsequently used on HCAEC in this study. ## 2.2. Effects of PCSK9 Inhibitors on PCSK9 Protein Expression PCSK9 inhibitors exhibit significant reducing effects on PCSK9 protein suppression on LPS-stimulated HCAEC compared to LPS control (Figure 2a). Both exhibited a reduction in PCSK9 protein expression across all concentrations (1–100 µg/mL) ($p \leq 0.05$ and $p \leq 0.001$). Alirocumab and evolocumab work best at the lowest concentration tested (1 µg/mL) in downregulating PCSK9 protein expression ($p \leq 0.001$, $p \leq 0.05$). According to AUC analysis, alirocumab ($14.0\%$) is the most potent PCSK9 inhibitor for inhibiting PCSK9 protein expression compared to evolocumab ($5.7\%$) (Table 1, Figure 2a). ## 2.3. Effects of PCSK9 Inhibitors on Endothelial Activation Protein Expression (ICAM-1 and E-Selectin) There was a significant reduction in endothelial activation protein expression by the co-incubation with a PCSK9 inhibitor at different concentrations. Specifically, evolocumab (10 µg/mL) and alirocumab (100 µg/mL) suppressed ICAM-1 significantly compared to the LPS control ($p \leq 0.01$) (Figure 2b). Similarly, LPS stimulated-HCAEC treated with alirocumab or evolocumab showed E-selectin reduction at various concentrations, suggesting both inhibitors exhibited dose-independent effects. Evolocumab evidenced suppression only at the highest concentration tested, 100 µg/mL ($p \leq 0.01$). Alirocumab reduced E-selectin at all concentrations tested except for 10 µg/mL ($p \leq 0.01$). The most potent PCSK9 inhibitor for E-selectin protein expression inhibition was alirocumab ($11.9\%$), followed by evolocumab ($3.7\%$) (Table 1, Figure 2c). ## 2.4. Effects of PCSK9 Inhibitors on IL-6 Protein Expression Alirocumab and evolocumab, across all concentrations (1–100 µg/mL), showed no significant reductions in IL-6 protein expression compared to LPS-stimulated HCAEC (Figure 2d). Instead of downregulating IL-6, IL-6 protein expression was upregulated with a PCSK9 inhibitor ($p \leq 0.05$, $p \leq 0.001$) compared to the LPS controls. However, the untreated and unstimulated cells show a very minimal IL-6 protein expression compared to the LPS control, which ensures the study’s validity. ## 2.5. Effects of PCSK9 Inhibitor on PCSK9 Gene Expression The co-incubation of evolocumab at 1 and 10 µg/mL with LPS-stimulated HCAEC significantly downregulated the PCSK9 mRNA ($p \leq 0.05$ and $p \leq 0.01$, respectively). On the contrary, instead of downregulating the level of PCSK9 expression, Alirocumab caused an upregulation of PCSK9 mRNA across all concentrations compared to LPS controls (Figure 3a). ## 2.6. Effects of PCSK9 Inhibitor on Endothelial Activation Gene Expression (ICAM-1 and E-Selectin) The co-incubation of LPS and evolocumab down-regulated both endothelial activation gene expression, ICAM-1 ($p \leq 0.01$ and $p \leq 0.05$) and E-selectin ($p \leq 0.05$) mRNA at 10 and 100 µg/mL, respectively. In contrast, alirocumab enhanced the endothelial activation gene expression above LPS controls ($p \leq 0.05$) (Figure 3b,c). Thus, the AUC analysis only showed the percentage of inhibition of evolocumab for downregulating ICAM-1 and E-Selectin gene expression ($57.4\%$ and $126.9\%$, respectively) (Table 1). ## 2.7. Effects of PCSK9 Inhibitor on IL-6 Gene Expression All concentrations of both PCSK9 inhibitors did not suppress IL-6 mRNA expression when compared to LPS controls (Figure 3d), which was consistent with IL-6 protein expression (Figure 2d). ## 2.8. Effects of PCSK9 Inhibitor on NF-ĸB p65 Protein and Gene Expression Evolocumab and alirocumab showed a reduction trend of NF-κB p65 protein compared to LPS controls, although not a statistically significant one. It can be seen through AUC analysis that evolocumab ($21.6\%$) inhibits the NF-κB p65 protein expression at a higher percentage than alirocumab ($16.0\%$) (Table 1). The negative control showed an opposite trend of NF-κB p65 protein expression (up-regulating) (Figure 4a). The same applies to gene expression; the co-incubation of LPS with evolocumab at 10 μg/mL down-regulated the mRNA level of NF-κB p65 but was not significant. However, instead of downregulating, alirocumab enhanced NF-κB p65 gene expression compared to LPS alone at all different concentrations (1–100 µg/mL) (Figure 4b). ## 2.9. Effects of PCSK9 Inhibitor on eNOS Protein and Gene Expressions Co-incubation of LPS alone reduces eNOS in terms of protein and gene expressions. Meanwhile, the unstimulated/untreated cells showed elevated protein ($p \leq 0.001$) and gene expression ($p \leq 0.001$). PCSK9 inhibitors enhanced eNOS protein expressions at different concentrations, where evolocumab at 10 μg/mL ($p \leq 0.05$) and alirocumab at 1 μg/mL ($p \leq 0.05$) compared to the LPS control. The potency of both PCSK9 inhibitors in inhibiting eNOS protein is almost similar, where evolocumab was at $18.5\%$ and alirocumab was at $18.6\%$ (Figure 5a, Table 1). In contrast with gene expression, only evolocumab showed an upregulation trend of eNOS mRNA at 1–100 µg/mL, although it was not significant. Alirocumab, on the other hand, showed the opposite trend against evolocumab. The most potent PCSK9 inhibitor for upregulating the eNOS biomarker was alirocumab ($63.5\%$) compared to evolocumab ($18.6\%$) (Figure 5b, Table 1). ## 2.10. Effects of PCSK9 Inhibitors on Monocytes and LPS-Stimulated HCAEC Interaction A monocyte adhesion assay was performed to explore the effects of PCSK9 inhibitors on monocytes and endothelial cell interactions. After co-incubation with LPS alone for 24 h, the adhesion of U937 monocytes to HCAEC was markedly increased. The evolocumab treatment at 100 μg/mL ($p \leq 0.001$) reduced monocyte adhesion to LPS-stimulated HCAEC compared to LPS controls. A significant reduction was observed at 10 and 100 µg/mL ($p \leq 0.001$) compared to LPS controls for alirocumab. An AUC analysis showed that alirocumab exhibited $51.54\%$ inhibition of monocyte adhesion, whilst for evolocumab it was $29.15\%$. The control group, monocytic U937, evidenced minimal adherence to the unstimulated HCAEC (Figure 6, Table 1). ## 3. Discussion At present, there is an abundance of clinical or in vivo studies on PCSK9 inhibitors. Despite the importance of the in vitro studies, there is a scarcity of research related to PCSK9 inhibitors. To the best of our knowledge, this is the first report to describe the anti-atherogenic effects of PCSK9 inhibitors in HCAEC. HCAEC is the most appropriate cell line to be used in anti-atherogenic studies due to several factors: [1] PCSK9 is expressed by the endothelial cell [36]; [2] Atherosclerosis is a disease related to the arteries; and [3] HCAEC is the cell that lines the innermost layer of the coronary artery blood vessels. The greater use of multi-LLT regimens is associated with lower LDL-C levels and better outcomes [37]. Statins, which are functionally known as Hydroxymethylglutaryl-CoA (HMG-CoA) reductase inhibitors, have been used as first-line drugs to treat hypercholesterolemia that primarily decreases LDL-C and triglyceride (TG) levels [38,39]. However, statins also increase PCSK9 activity. While statins efficiently lower cholesterol levels, their efficacy diminishes with a rise in PCSK9 activity. Thus, it is paramount to study the mechanism of other LLTs that can reduce LDL-C and PCSK9, such as evolocumab and alirocumab. The MTS assay demonstrated that PCSK9 inhibitors (evolocumab and alirocumab) up to 100 µg/mL did not exhibit any toxic effects on the viability of HCAEC. PCSK9 inhibitors were considered safe, as the viability was more than $90\%$. Whilst studies to validate and compare the safety of PCSK9 are lacking, Safaeian et al. [ 2019, 2020] [40,41] did report on the non-cytotoxic effects of evolocumab on human umbilical vein endothelial cells (HUVEC) at 0.5–100 µg/mL using an MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay. Despite the consensus on the cytotoxic effects of PCSK inhibitors, the comparison of the cell viability may be incomparable due to the different cell lines used. In general, LPS-stimulated HCAEC, treated with evolocumab and alirocumab, exhibited a reduction of PCSK9, E-selectin, ICAM-1, and NF-ĸB biomarkers at different treatment concentrations. In contrast, for eNOS, the biomarker expression was marginally upregulated but not significant. For the monocyte binding to endothelial, alirocumab showed a higher percentage of inhibition than evolocumab. Therefore, this study suggested that the mechanism that leads to the improvement of endothelial function may be due to anti-PCSK9 and anti-endothelial activation that was mediated via the NF-κB p65 and eNOS pathways. This is in consensus with the study by Di Minno et al., which observed an improvement in endothelial function after evolocumab therapy in patients with familial hypercholesterolemia (FH) [42]. This is also supported by Maulucci et al., who had similar findings in subjects with increased cardiovascular risk [43]. Excess dietary fat consumption increases hepatic PCSK9 expression [44]. PCSK9 could accelerate atherosclerosis through mechanisms beyond the degradation of the hepatic LDLR. Several clinical studies suggested that PCSK9 is involved in atherosclerotic inflammation [45,46]. PCSK9 is ubiquitously expressed in many tissues and cell types [47]. Feingold et al., [ 2008] [33] reported that LPS stimulation (5 mg/kg body weight, intraperitoneally) in an animal model (Female C57BL/6 mice) increased PCSK9 mRNA levels by 2.5-fold (4 h) and 12.5-fold (38 h). In an in vitro model, Ding et al., [ 2015] [22] did not specifically report the increase in PCSK9 expression. Still, they reported the dose-dependent relationship between LPS (0.001–1 µg/mL) stimulation and PCSK9 expression on aortic endothelial cells. PCSK9 expression was more significant in smooth muscle cells (SMC) than in endothelial cells (EC) upon LPS stimulation [34,48]. PCSK9 expression in HCAEC is yet to be reported on. Our in vitro study showed the capability of LPS in upregulating the PCSK9 expression compared to the normal control (unstimulated) in HCAEC. The PCSK9 protein and gene expression increased 1.2-fold with LPS stimulation (24 h). The LPS-stimulated HCAEC were co-incubated with PCSK9 inhibitors to observe the efficacy of PCSK9 reduction using HCAEC in an in vitro model. Evolocumab showed a promising significant downregulation effect on the PCSK9 protein (decreased 1.2-fold) and gene expression (decreased 1.5-fold) compared to the LPS control. For alirocumab, it reduced the PCSK9 protein downregulation (decreased 1.2-fold), but the gene was upregulated. This might be due to the increase of hepatocyte nuclear factor 1α (HNF1α) gene transcription, which is critical in regulating PCSK9 gene transcription [49]. HNF1α promotes PCSK9 transcription by binding with the HNF1 motif, which is located upstream of sterol regulatory element 1 (SRE1) in the PCSK9 promoter [49]. Yang et al. [ 50] reported the effects of chitosan oligosaccharides, which downregulated the PCSK9 gene, but upregulated HNF1α and the sterol regulatory element-binding proteins (SREBP). However, none of the studies on alirocumab reported on HNF1α. Based on our findings, we postulated that the HNF1α and SREBP might not be the primary PCSK9 regulators for alirocumab. Endothelial activation has been established as one of the critical events in atherosclerosis initiation and progression [5]. In unstimulated endothelial cells, E-selectin is undetectable [51], unlike ICAM-1, which is present in healthy arteries [52]. Upon stimulation with LPS, ICAM-1 gene expression did not change at 0.5 h but increased two- to three-fold at 12 h [53]. The E-selectin mRNA and protein increased in the lymphatic endothelial cells with LPS at more than two-fold levels compared with human umbilical vein endothelial cells (HUVEC) [53]. This is in consensus with our findings that showed that evolocumab significantly downregulated the ICAM-1 protein at 10 µg/mL and E-selectin at 100 µg/mL ($p \leq 0.01$). The gene expression for ICAM-1 and E-selectin was significantly downregulated at the same concentration, 10–100 µg/mL. For alirocumab, the ICAM-1 and E-selectin protein was reduced. However, ICAM-1 and E-selectin gene expression did not reflect the protein expression. ICAM-1 and E-selectin protein expressions were upregulated compared to LPS controls. This discrepancy might be attributed to changes in the regulatory mechanism, which involves the rate of translation and protein breakdown. mRNAs initially translated may later be temporarily repressed [54]. The high expression of endothelial cellular adhesion molecules (CAM) such as ICAM-1 and E-selectin has been consistently observed in atherosclerotic plaques, as it plays a vital role in inducing monocyte recruitment into the intima. E-selectin is an adhesion receptor that slows leukocyte rolling, and its expression is restricted to endothelial cells [51]. This unstable binding is further facilitated by ICAM-1, which promotes transmigration by rearranging the endothelium cytoskeleton and weakening the strength of endothelial cell junctions [55]. ICAM-1 is a transmembrane immunoglobulin protein predominantly expressed on endothelial cells and is crucial in leukocyte recruitment and transmigration [56]. The early stages in the development of atherosclerosis involve inflammation and endothelial activation, thus making the role of inflammation in atherosclerotic plaque formation crucial. The endothelial cells’ over-expression of proinflammatory cytokines and adhesion molecules is mediated via the activation of NF-ĸB during these stages [57]. NF-ĸB is involved in a proinflammatory signalling pathway responsible for expressing proinflammatory genes, including cytokines, chemokines, and adhesion molecules. One of the most highly induced NF-ĸB-dependent cytokines is IL-6 [58]. Unlike other biomarkers, the protein and gene expression for inflammatory biomarkers (IL-6) of both PCSK inhibitors showed an increase in LPS controls. The AUC of alirocumab is −$119.2\%$ (IL-6 protein) and −$1373.1\%$ (IL-6 gene). For evolocumab, the AUC is −$132.7\%$ (IL-6 protein) and −$156.7\%$ (IL-6 gene). The NF-ĸB p65 showed a reduction trend but was not statistically significant for protein and gene expressions. Therefore, in these cases, it can be deduced that PCSK9 inhibitors may not reduce the IL-6 inflammatory biomarkers through the NF-ĸB p65 pathway. Leucker et al. [ 59] reported that unchanged inflammatory biomarkers, including high-sensitivity, C-reactive protein, IL-6, interferon-gamma, tumour necrosis factor-alpha, and soluble CD163 were unchanged in dyslipidemia patients treated with evolocumab. eNOS, when bound to its co-factor, tetrahydrobiopterin (BH4), will produce nitric oxide (NO). Before the development of atherosclerotic plaques, the NO, which serves as an endothelial vasodilator, is impaired. A defect in NO production or activity has been proposed as a significant mechanism of endothelial dysfunction and a contributor to atherosclerosis [60]. In addition to NO, eNOS produces a superoxide anion if its function is altered. This phenomenon is referred to as “eNOS uncoupling”, as eNOS is not coupled with its cofactor and has been found to play an essential role in the process of various cardiovascular diseases [61]. When endothelial cells are under oxidative stress, oxidation of BH4 to dihydrobiopterin (BH2) will occur. BH2 will bind to eNOS, eNOS becomes uncoupled, and, as a result, the production of reactive oxygen species (ROS) will increase [62]. LDL may be oxidized by ROS released by vascular cells within the arterial wall [63]. ROS generation followed the same pattern as PCSK9 expression, and it is dose-dependent in EC and SMC upon stimulation [5]. In this study, as expected, the eNOS protein and mRNA ($p \leq 0.001$ and $p \leq 0.01$) in the control groups (untreated/unstimulated) were found to be highest compared to LPS alone. The eNOS protein expression was upregulated with the co-incubation of alirocumab and evolocumab at specific concentrations. The eNOS mRNA was only upregulated for evolocumab. However, the upregulation is not significant. Further investigation on the coupling of eNOS is warranted. Monocyte adhesion to the endothelium, followed by monocyte extravasation, is an initial stage of atherogenesis development. It is an important event in vascular inflammation. Circulating monocytes are critically involved in the progression of atherosclerosis upon migration to the tissue and differentiate into macrophages. The adhesion cascade is a strongly regulated process, and monocyte adherence will increase in number during chronic inflammation. This process is predominantly mediated by cellular adhesion molecules, including ICAM-1 and E-selectin, which are expressed by activated endothelial cells in response to several inflammatory stimuli, including LPS. Evolocumab and alirocumab (1–100 µg/mL) can suppress monocyte adhesion to LPS-stimulated HCAEC in a dose-dependent manner compared to controls. Alirocumab showed the highest reduction in monocyte binding to HCAEC at 100 µg/mL compared to LPS alone. In this study, both PCSK9 inhibitors, alirocumab ($51.54\%$) and evolocumab ($29.15\%$), showed a prominent percentage of monocyte inhibition. Thus, this suggested their pivotal role in mediating the LDL-C lowering effects in PCSK9 inhibitors, parallel with the reduction of ICAM-1 and E-selectin biomarkers in these in vitro studies. Despite debates on the cost-effectiveness of PCSK9 inhibitors since 2015, a definite advantage of PCSK9 inhibitors is their better tolerance compared to the lipid-lowering drugs used to date. They cause significantly fewer muscle symptoms than statins [64,65]. As these drugs become more widely used in clinical practice, accumulating scientific and clinical data indicate that PCSK9 inhibitors have an excellent safety and tolerability profile, with a low occurrence of side effects. In the FOURIER [2017] study, the percentage of reported adverse events following evolocumab treatment was much lower than in earlier trials. More importantly, there was no significant difference in the number of adverse events between the trial and placebo groups [31]. ## 4.1. Cell Culture HCAEC were obtained from Cell Applications Inc. (San Diego, CA, USA) and were developed from a normal human coronary artery of a 20-year-old Caucasian. The cell lines were grown in the MesoEndo cell growth medium (Cell Application Inc., San Diego, CA, USA) and detached using Accutase (Nacalai Tesque, Kyoto, Japan). The cell lines were incubated in a $90\%$ humidified atmosphere containing $5\%$ CO2 (Galaxy 170 R, Eppendorf, Hamburg, Germany). Cells were grown in a 25 or 75-cm2 flask (BD Falcon, London, UK). The cells were passaged when they reached more than $80\%$ confluency. All of the experiments conducted in this research used HCAEC from passages five to nine. ## 4.2. Cell Viability HCAEC cell viability was determined following the MTS assay (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4sulfophenyl)-2Htetrazolium) that used the CellTiter 96® AQueous One Solution Cell Proliferation Assay kit (Promega, Madison, WI, USA). The HCAEC were seeded into a 96 wells culture plate (1 × 104 cells/well) and incubated overnight at 37 °C for 24 h in humidified $5\%$ CO2 before being treated with different concentrations of alirocumab and evolocumab (1, 10, and 100 μg/mL). Next, 10 µL of the MTS reagent (5 mg/mL MTS) was added to each well and incubated for another 4 h at 37 °C. The absorbance was measured at 490 nm using a microplate reader (Agilent, Santa Clara, CA, USA). The experiment was conducted in three biological replicates. The viability of the cells was measured by comparing the treated wells with the control wells and calculated using the following formula:Cell viability (%)=Sample absorbance−Blank absorbanceControl absorbance−Blank absorbance×$100\%$ ## 4.3. Treatment of HCAEC The PCSK9 inhibitors were diluted into various working concentrations (1, 10, and 100 μg/mL) with culture media. HCAEC were treated with 1, 10, and 100 μg/mL of PCSK9 inhibitors together with 1 μg/mL of LPS, followed by 24 h of incubation at 37 °C in $5\%$ CO2. LPS was used to imitate the release of the inflammatory mediator that contributes to the increased release of PCSK9 and the pathological process of atherogenesis. ## 4.4.1. Cell Culture Supernatant Collection LPS-stimulated HCAEC was co-incubated with alirocumab and evolocumab at concentrations of 1, 10, and 100 μg/mL for 24 h. The supernatant was collected and centrifuged for 20 min at 1000× g at 2–8 °C. The pellet formed at the bottom of the microcentrifuge tube was discarded. The supernatant was used to carry out the assay was kept at −80 °C before use. ## 4.4.2. Nuclear Lysates Collection (Nuclear Extraction) for NF-ĸB Assay The nuclear extraction was performed following the manufacturer’s instructions using the CHEMICON® Nuclear Extraction kit (CHEMICON® International, Temecula, CA, USA). The treated cells were washed with 1× PBS solution, followed by the addition of Accutase to detach the cell, and the volume of cells in the cell pellet was counted and estimated using the cell counter. Five cell pellet volumes of ice-cold 1× cytoplasmic lysis buffer containing 0.5 mM DTT and $\frac{1}{1000}$ dilution or inhibitor cocktail were added. The cell pellet was resuspended by gently inverting the tube. The cell suspension was incubated on ice for 15 min and centrifuged at 250× g for 5 min at 4 °C. The supernatant was discarded, and the cell pellet was resuspended in two volumes of ice-cold 1× cytoplasmic lysis buffer. A syringe with a small gauge needle (27 gauge) was used to draw the cell suspension prepared from the sample tube into the syringe, and the contents were then ejected back into the sample tube. The drawing and ejecting were repeated approximately five times. The disrupted cell suspension was centrifuged at 8000× g for 20 min at 4 °C. The supernatant was discarded. The remaining pellet contains the nuclear portion of the cell lysate. The nuclear pellet was resuspended in $\frac{2}{3}$ of the original cell pellet volume of ice-cold nuclear extraction buffer containing 0.5 mM DTT and $\frac{1}{1000}$ protease inhibitor cocktail. A fresh syringe with a 27-gauge needle was used. A rotator or orbital shaker (low speed) was used to gently agitate the nuclear suspension at 4 °C for 30–60 min. The nuclear suspension was centrifuged at 16,000× g for 5 min at 4 °C. The supernatant containing the nuclear lysates was transferred to a fresh tube to carry out the assay. ## 4.4.3. Cell Lysates Collection for eNOS Assay The cell lysates were collected following the protocol from Proteintech® on cell and tissue lysate preparation (Proteintech®, Rosemont, IL, USA). The treated cells were pelleted by centrifugation for 5 min at 1000× g (approximately 2000 rpm) at 4 °C. The cells were washed three times with ice-cold 1× PBS, then a chilled RIPA buffer with a protease inhibitor was added. In general, 100 μL RIPA buffer was added for approximately every 106 cells in the pellet (count cells before centrifugation). The pellet was vortexed occasionally until it homogenized with the buffer and was kept on ice for 30 min. The sample was sonicated to break the cells up and to shear the cell’s DNA. The sonication time was adjusted to 1 min at a power of about 180 watts (in rounds of 10 s sonication/10 s rest for each cycle). The sample was kept on ice during the sonication. ## 4.5.1. Quantitation of PCSK9, IL-6, ICAM-1, and E-Selectin in the Supernatant The supernatant’s concentrations of IL-6, soluble ICAM-1, soluble E-selectin, and PCSK9 were performed with a commercially available standard ELISA kit (Elabscience, Houston, TX, USA), according to the manufacturer’s instructions. The absorbance was measured at 450 nm using a microplate reader (Agilent, Santa Clara, CA, USA). ## 4.5.2. Measurement of NF-κB p65 Protein in Nuclear Lysates Protein The nuclear lysates collected were quantified for NF-ĸB using the Elabscience® Human NF-ĸB p65 ELISA kit (Elabscience, Affymetrix, Houston, TX, USA). All procedures were performed according to the manufacturer’s instructions. The absorbance of the samples was measured at 450 nm using a microplate reader (Agilent, Santa Clara, CA, USA). ## 4.5.3. Quantitation of eNOS Protein in Cell Lysates The eNOS protein concentration in cell lysates was quantitated using the Elabscience® Human eNOS ELISA kit (Elabscience, Houston, TX, USA). All procedures were conducted following the manufacturer’s instructions. The absorbance was read using a microplate reader at 450 nm with a reference wavelength set at 570 nm (Agilent, Santa Clara, CA, USA). ## 4.6.1. RNA Extraction The total RNA was extracted from the treated cell pellets using the RNA extraction kit from Macherey-Nagel (Duren, Germany). RNA purity and concentration were determined by nanodrop. Buffer RA1 (350 µL) was added to the cell pellet and vortexed vigorously. The mixture was placed in the NucleoSpin® filter (violet ring) on a 2 mL collection tube and centrifuged for 1 min at 11,000× g. The NucleoSpin® filter (violet ring) was discarded, and 350 µL ethanol ($70\%$) was added to the homogenized lysates and mixed by pipetting up and down (five times). The lysates were pipetted up two to three times and then transferred to a NucleoSpin® RNA Column (light blue ring) in a collection tube. The column was centrifuged for 30 s at 11,000× g, and the collection tube was discarded. The column was placed in a new collection tube. A three hundred fifty µL membrane desalting buffer (MDB) was added and centrifuged at 11,000 for 1 min to dry the membrane. The Dnase reaction mixture (95 µL) was applied directly onto the center of the silica membrane of the column and incubated at room temperature for 15 min. Buffer RAW2 (200 µL) was added to a NucleoSpin® RNA column and centrifuged at 11,000 for 30 s. The column was placed in a new collection tube. Buffer RA3 (600 µL) was added to a NucleoSpin® RNA column and centrifuged at 11,000 rpm for 30 s. The flowthrough was discarded, and the column was placed back into a 2 m collection tube. Buffer RA3 (250 µL) was added to a NucleoSpin® RNA column and centrifuged at 11,000 rpm for 2 min to dry the membrane completely. The column was placed into a nuclease-free collection tube. The RNA was eluted in 60 µL Rnase-free H2O and centrifuged at 11,000 rpm for 1 min. ## 4.6.2. Quantitation of PCSK9, Inflammation, Endothelial Activation, NF-κB, and eNOS Genes A QuantiGene Plex 96-well assay (Thermofisher Scientific, Waltham, MA, USA) measured gene expression according to the manufacturer’s protocol. In triplicates, RNA was transferred to the assay hybridization plate, and contained a working bead mix and probe sets. Hybridization was performed for 20 h at 54 °C ± 1 °C, shaking at 600 rpm. Next, the mixtures were transferred to a 96-well magnetic separation plate. The beads were hybridized with a preamplifier probe, an amplifier probe, a label probe, and Streptavidin conjugated R-Phycoerythrin (SAPE). SAPE fluorescence was measured with the Luminex FlexMap three-dimensional instrument (Luminex Corporation, Austin, TX, USA) to indicate the volume of mRNA transcripts captured by the beads. Fold-changes will be taken as the relative ratios between the normalized reference values of all treatment groups and the untreated group’s values. The target-specific RNA molecules of LPS-stimulated HCAEC were PCSK9: NM_174936; IL-6: NM_000600; ICAM-1:NM_000201; E-Selectin: NM_000450; NF-ĸB p65: NM_003998; eNOS: NM_000603. ## 4.6.3. Monocyte Binding Assay The binding of U937 monocytes to the endothelial cell capacity was measured by the Rose Bengal method [66]. HCAEC were seeded in 96-well microtiter plates at a seeding density of 1 × 105 cells/mL and incubated overnight in a humidified incubator set at 37 °C and $5\%$ CO2. Cells were treated with alirocumab and evolocumab at concentrations of 1, 10, and 100 μg/mL for 24 h with 5 × 105 cells/mL monocytes U937 added afterwards and incubated for 1 h at 37 °C. After three washes to remove non-adherent monocytes, $0.25\%$ of Rose Bengal stain in phosphate-buffered saline (PBS) was added to each well for 10 min at 25 °C. The excess stains were washed away three times with PBS [supplemented with $10\%$ fetal bovine serum (FBS)], and the stain was released from the cells with a solution of ethanol: PBS (1:1 v/v) for 1 h at 25 °C. Monocyte-endothelial cell adhesion was calculated from the difference in absorbance at 570 nm between wells that contain monocytes and HCAEC and wells that contain HCAEC only. ## 4.6.4. Statistical Analysis Results were reported as mean ± standard deviation (SD). All results were analyzed using IBM SPSS Statistic 26. A one-way analysis of variance (ANOVA) followed by a Bonferroni post hoc analysis was used. Two-sided $p \leq 0.05$ was considered significant. The percentage (%) of inhibition against LPS controls for each biomarker was obtained from the area under the curve (AUC) analysis using Graph Version 4.3. ## 5. Conclusions This in vitro study demonstrated that the anti-atherogenic properties of PCSK9 inhibition are mediated by endothelial activation and the capability of PCSK9 inhibitors to suppress the binding of monocytes to endothelial cells. These two factors are essential to atheroma formation. These findings will give researchers and pharmaceutical companies a broader view in developing prospective medications and providing better management plans for hypercholesterolemia patients. ## References 1. Jebari-Benslaiman S., Galicia-Garcia U., Larrea-Sebal A., Olaetxea J.R., Alloza I., Vandenbroeck K., Benito-Vicente A., Martin C.. Pathophysiology of Atherosclerosis. Int. J. Mol. Sci. (2022.0) 23. DOI: 10.3390/ijms23063346 2. Mestas J., Ley K.. Monocyte-endothelial cell interactions in the development of atherosclerosis. Trends Cardiovasc. Med. (2008.0) 18 228-232. DOI: 10.1016/j.tcm.2008.11.004 3. Tan J.N., Husain K., Jubri Z., Chan K.M., Jantan I., Mohd Fauzi N.. Gynura procumbens (Lour.) Merr. extract attenuates monocyte adherence to endothelial cells through suppression of the NF-κB signaling pathway. J. Ethnopharmacol. (2022.0) 10 294. DOI: 10.1016/j.jep.2022.115391 4. Mohd Ariff A., Abu Bakar N.A., Abd Muid S., Omar E., Ismail N.H., Ali A.M., Mohd Kasim N.A., Mohd Nawawi H.. Ficus deltoidea suppresses endothelial activation, inflammation, monocytes adhesion, and oxidative stress via NF-κB and eNOS pathways in stimulated human coronary artery endothelial cells. BMC Complement. Med. Ther. (2020.0) 17. DOI: 10.1186/s12906-020-2844-6 5. Marchio P., Guerra-Ojeda S., Vila J.M., Aldasoro M., Victor V.M., Mauricio M.D.. Targeting Early Atherosclerosis: A Focus on Oxidative Stress and Inflammation. Oxid. Med. Cell. Longev. (2019.0) 1 8563845. DOI: 10.1155/2019/8563845 6. Stein E.A., Gipe D., Bergeron J.. Effect of a monoclonal antibody to PCSK9, REGN727/SAR236553, to reduce low-density lipoprotein cholesterol in patients with heterozygous familial hypercholesterolaemia on stable statin dose with or without ezetimibe therapy: A phase 2 randomised controlled trial. Lancet (2012.0) 380 29-36. DOI: 10.1016/S0140-6736(12)60771-5 7. Ridker P.M., Revkin J., Amarenco P., Brunell R., Curto M., Civeira F., Flather M., Glynn R., Gregoire J., Jukema J.. Cardiovascular efficacy and safety of bococizumab in high-risk patients. N. Engl. J. Med. (2017.0) 376 1527-1539. DOI: 10.1056/NEJMoa1701488 8. PCSK9 Inhibitors 9. Roth E.M., McKenney J.M., Hanotin C., Asset G., Stein E.A.. Atorvastatin with or without an antibody to PCSK9 in primary hypercholesterolemia. N. Engl. J. Med. (2012.0) 367 1891-1900. DOI: 10.1056/NEJMoa1201832 10. Zhang J., Tecson K.M., Rocha N.A., McCullough P.A.. Usefulness of alirocumab and evolocumab for the treatment of patients with diabetic dyslipidemia. Baylor University Medical Center Proceedings (2018.0) Volume 11 180-184. DOI: 10.1080/08998280.2018.1441255 11. Dixon D.L., Pamulapati L.G., Bucheit J.D., Sisson E.M., Smith S.R., Kim C.J., Wohlford G.F., Pozen J.. Recent Updates on the Use of PCSK9 Inhibitors in Patients with Atherosclerotic Cardiovascular Disease. Curr. Atheroscler. Rep. (2019.0) 16 16. DOI: 10.1007/s11883-019-0778-6 12. Dong B., Wu M., Li H., Kraemer F.B., Adeli K., Seidah N.G., Park S.W., Liu J.. Strong induction of PCSK9 gene expression through HNF1α and SREBP2: Mechanism for the resistance to LDL-cholesterol lowering effect of statins in dyslipidemic hamsters. J. Lipid. Res. (2010.0) 51 1486-1495. DOI: 10.1194/jlr.M003566 13. Tang Y., Li S.L., Hu J.H., Sun K.J., Liu L.L., Xu D.Y.. Research progress on alternative non-classical mechanisms of PCSK9 in atherosclerosis in patients with and without diabetes. Cardiovasc. Diabetol. (2020.0) 19 33. DOI: 10.1186/s12933-020-01009-4 14. Zulkapli R., Yusof M.Y.P.M., Abd Muid S., Wang S.M., Firus Khan A.Y., Nawawi H.A.. Systematic Review on Attenuation of PCSK9 in Relation to Atherogenesis Biomarkers Associated with Natural Products or Plant Bioactive Compounds in In Vitro Studies: A Critique on the Quality and Imprecision of Studies. Int. J. Environ. Res. Public Health (2022.0) 19. DOI: 10.3390/ijerph191912878 15. Desai N.R., Kohli P., Giugliano R.P., O’Donoghue M.L., Somaratne R., Zhou J., Hoffman E.B., Huang F., Rogers W.J., Wasserman S.M.. AMG145, a monoclonal antibody against proprotein convertase subtilisin kexin type 9, significantly reduces lipoprotein(a) in hypercholesterolemic patients receiving statin therapy: An analysis from the LDL-C Assessment with Proprotein Convertase Subtilisin Kexin Type 9 Monoclonal Antibody Inhibition Combined with Statin Therapy (LAPLACE)-Thrombolysis in Myocardial Infarction (TIMI) 57 trial. Circulation (2013.0) 128 962-969. DOI: 10.1161/CIRCULATIONAHA.113.001969 16. Massimiliano R., Maria F.G., Nicola F., Alberto C.. Lipoprotein(a) and PCSK9 inhibition: Clinical evidence. Eur. Heart J. (2020.0) 22 53-56. DOI: 10.1093/eurheartj/suaa135 17. Giugliano R.P., Desai N.R., Kohli P., Rogers W.J., Somaratne R., Huang F., Liu T., Mohanavelu S., Hoffman E.B., McDonald S.T.. Efficacy, safety, and tolerability of a monoclonal antibody to proprotein convertase subtilisin/kexin type 9 in combination with a statin in patients with hypercholesterolaemia (LAPLACE-TIMI 57): A randomised, placebo-controlled, dose-ranging, phase 2 study. Lancet (2012.0) 380 2007-2017. DOI: 10.1016/S0140-6736(12)61770-X 18. Blom D.J., Hala T., Bolognese M., Lillestol M.J., Toth P.D., Burges L., Ceska R., Roth E., Koren M.J., Ballantyne C.M.. A 52-week placebo-controlled trial of evolocumab in hyperlipidemia. N. Engl. J. Med. (2014.0) 370 1809-1819. DOI: 10.1056/NEJMoa1316222 19. Koren M.J., Lundqvist P., Bolognese M., Neutel J.M., Monsalvo M.L., Yang J., Kim J.B., Scott R., Wasserman S.M., Bays H.. Anti-PCSK9 monotherapy for hypercholesterolemia: The MENDEL-2 randomized, controlled phase III clinical trial of evolocumab. J. Am. Coll. Cardiol. (2014.0) 63 2531-2540. DOI: 10.1016/j.jacc.2014.03.018 20. Koren M.J., Scott R., Kim J.B., Knusel B., Liu T., Lei L., Bolognese M., Wasserman S.M.. Efficacy, safety, and tolerability of a monoclonal antibody to proprotein convertase subtilisin/kexin type 9 as monotherapy in patients with hypercholesterolaemia (MENDEL): A randomised, double-blind, placebo-controlled, phase 2 study. Lancet. (2012.0) 380 1995-2006. DOI: 10.1016/S0140-6736(12)61771-1 21. Raal F., Scott R., Somaratne R., Bridges I., Li G., Wasserman S.M., Stein E.A.. Low-density lipoprotein cholesterol-lowering effects of AMG 145, a monoclonal antibody to proprotein convertase subtilisin/kexin type 9 serine protease in patients with heterozygous familial hypercholesterolemia: The reduction of LDL-C with PCSK9 inhibition in heterozygous familial hypercholesterolemia disorder (RUTHERFORD) randomized trial. Circulation (2012.0) 126 2408-2417. DOI: 10.1161/CIRCULATIONAHA.112.144055 22. Chaudhary R., Garg J., Shah N., Sumner A.. PCSK9 inhibitors: A new era of lipid lowering therapy. World J. Cardiol. (2017.0) 9 76-91. DOI: 10.4330/wjc.v9.i2.76 23. Sabatine M.S., Giugliano R.P., Keech A., Honarpour N., Wang H., Liu T., Wasserman S.M., Scott R., Sever P.S., Pedersen T.R.. Rationale and design of the Further cardiovascular Outcomes Research with PCSK9 Inhibition in subjects with Elevated Risk trial. Am. Heart J. (2016.0) 173 94-101. DOI: 10.1016/j.ahj.2015.11.015 24. Keech A.C., Oyama K., Sever P.S., Tang M., Murphy S.A., Hirayama A., Lu C., Tay L., Deedwania P.C., Siu C.W.. Efficacy and Safety of Long-Term Evolocumab Use Among Asian Subjects-A Subgroup Analysis of the Further Cardiovascular Outcomes Research with PCSK9 Inhibition in Subjects with Elevated Risk (FOURIER). Trial. Circ. J. (2021.0) 85 2063-2070. DOI: 10.1253/circj.CJ-20-1051 25. Murphy S.A., Pedersen T.R., Gaciong Z.A., Ceska R., Ezhov M.V., Connolly D.L., Jukema J.W., Toth K., Tikkanen M.J., Im K.. Effect of the PCSK9 Inhibitor Evolocumab on Total Cardiovascular Events in Patients with Cardiovascular Disease: A Prespecified Analysis from the FOURIER Trial. JAMA Cardiol. (2019.0) 4 613-619. DOI: 10.1001/jamacardio.2019.0886 26. Sinnaeve P.R., Schwartz G.G., Wojdyla D.M., Alings M., Bhatt D.L., Bittner V.A., Chiang C.E., Correa Flores R.M., Diaz R., Dorobantu M.. Effect of alirocumab on cardiovascular outcomes after acute coronary syndromes according to age: An ODYSSEY OUTCOMES trial analysis. Eur. Heart J. (2020.0) 21 2248-2258. DOI: 10.1093/eurheartj/ehz809 27. Hagstrom E., Steg P.G., Szarek M., Bhatt D.L., Bittner V.A., Danchin N., Diaz R., Goodman S.G., Harrington R.A., Jukema J.W.. Apolipoprotein B, Residual Cardiovascular Risk After Acute Coronary Syndrome, and Effects of Alirocumab. Circulation. (2022.0) 146 657-672. DOI: 10.1161/CIRCULATIONAHA.121.057807 28. Landmesser U., McGinniss J., Steg P.G., Bhatt D.L., Bittner V.A., Diaz R., Dilic M., Goodman S.G., Jukema J.W., Loy M.. Achievement of ESC/EAS LDL-C treatment goals after an acute coronary syndrome with statin and alirocumab. Eur. J. Prev. Cardiol. (2022.0) 29 1842-1851. DOI: 10.1093/eurjpc/zwac107 29. Nicholls S.J., Puri R., Anderson T., Ballantyne C.M., Cho L., Kastelein J.J.. Effect of evolocumab on progression of coronary disease in statin-treated patients: The GLAGOV randomised clinical trial. JAMA (2016.0) 316 2373-2384. DOI: 10.1001/jama.2016.16951 30. Sabatine M.S., Giugliano R.P., Keech A.C., Honarpour N., Wiviott S.D., Murphy S.A.. Evolocumab and clinical outcomes in patients with cardiovascular disease. N. Engl. J. Med. (2017.0) 376 1713-1722. DOI: 10.1056/NEJMoa1615664 31. Karagiannis A.D., Liu M., Toth P.P., Zhao S., Agrawal D.K., Libby P., Chatzizisis Y.S.. Pleiotropic Anti-atherosclerotic Effects of PCSK9 Inhibitors from Molecular Biology to Clinical Translation. Curr. Atheroscler. Rep. (2018.0) 10 20. DOI: 10.1007/s11883-018-0718-x 32. Shapiro M.D., Fazio S.. PCSK9 and atherosclerosis-lipids and beyond. J. Atheroscler. Thromb. (2017.0) 24 462-472. DOI: 10.5551/jat.RV17003 33. Feingold K.R., Moser A.H., Shigenaga J.K., Patzek S.M., Grunfeld C.. Inflammation stimulates the expression of PCSK9. Biochem. Biophys. Res. Commun. (2008.0) 374 341-344. DOI: 10.1016/j.bbrc.2008.07.023 34. Ding Z., Liu S., Wang X., Deng X., Fan Y., Shahanawaz J., Shmookler Reis R.J., Varughese K.I., Sawamura T., Mehta J.L.. Cross-talk between LOX-1 and PCSK9 in vascular tissues. Cardiovasc. Res. (2015.0) 107 556-567. DOI: 10.1093/cvr/cvv178 35. Luquero A., Badimon L., Borrell-Pages M.. PCSK9 Functions in Atherosclerosis Are Not Limited to Plasmatic LDL-Cholesterol Regulation. Front. Cardiovasc. Med. (2021.0) 23 639727. DOI: 10.3389/fcvm.2021.639727 36. Ragusa R., Basta G., Neglia D., De Caterina R., Del Turco S., Caselli C.. PCSK9 and atherosclerosis: Looking beyond LDL regulation. Eur. J. Clin. Investig. (2021.0) 51 e13459. DOI: 10.1111/eci.13459 37. Tromp T.R., Hartgers M.L., Hovingh G.K., Vallejo-Vaz A.J., Ray K.K., Soran H., Freiberger T., Bertolini S., Harada-Shiba M., Blom D.J.. Homozygous Familial Hypercholesterolaemia International Clinical Collaborators. Worldwide experience of homozygous familial hypercholesterolaemia: Retrospective cohort study. Lancet. (2022.0) 19 719-728. DOI: 10.1016/S0140-6736(21)02001-8 38. Sizar O., Khare S., Jamil R.T., Talati R.. Statin Medications (2022.0) 39. Al-Khateeb A., Zahri M.K., Mohamed M.S., Sasongko T.H., Ibrahim S., Yusof Z., Zilfalil B.A.. Analysis of sequence variations in low-density lipoprotein receptor gene among Malaysian patients with familial hypercholesterolemia. BMC Med. Genet. (2011.0) 12. DOI: 10.1186/1471-2350-12-40 40. Safaeian L., Vaseghi G., Jabari H., Dana N.. Evolocumab, a proprotein convertase subtilisin/kexin type 9 inhibitor, promotes angiogenesis in vitro. Can. J. Physiol. Pharmacol. (2019.0) 97 352-358. DOI: 10.1139/cjpp-2018-0542 41. Safaeian L., Mirian M., Bahrizadeh S.. Evolocumab, a PCSK9 inhibitor, protects human endothelial cells against H. Arch. Physiol. Biochem. (2020.0) 3 1-6. DOI: 10.1080/13813455.2020.1788605 42. 42. Lebeau P.F., Byun J.H., Platko K., Al-Hashimi A.A., Lhotak S., MacDonald M.E., Mejia-Benitez A., Prat A., Igdoura S.A., Trigatti B.. PCSK9 knockout exacerbates diet-induced non-alcoholic steatohepatitis, fibrosis and liver injury in mice. JHEP Rep. (2019.0) 1 418-429. DOI: 10.1016/j.jhepr.2019.10.009 43. Di Minno A., Gentile M., Iannuzzo G., Calcaterra I., Tripaldella M., Porro B., Cavalca V., Di Taranto M.D., Tremoli E., Fortunato G.. Endothelial function improvement in patients with familial hypercholesterolemia receiving PCSK-9 inhibitors on top of maximally tolerated lipid lowering therapy. Thromb. Res. (2020.0) 194 229-236. DOI: 10.1016/j.thromres.2020.07.049 44. Maulucci G., Cipriani F., Russo D., Casavecchia G., Di Staso C., Di Martino L.F.M., Ruggiero A., Di Biase M., Brunetti N.D.. Improved endothelial function after short-term therapy with evolocumab. J. Clin. Lipidol. (2018.0) 12 669-673. DOI: 10.1016/j.jacl.2018.02.004 45. Ferri N., Tibolla G., Pirillo A., Cipollone F., Mezzetti A., Pacia S., Corsini A., Catapano A.L.. Proprotein convertase subtilisin kexin type 9 (PCSK9) secreted by cultured smooth muscle cells reduces macrophages LDLR levels. Atherosclerosis (2012.0) 220 381-386. DOI: 10.1016/j.atherosclerosis.2011.11.026 46. Tang Z.-H., Peng J., Ren Z., Yang J., Li T.-T., Li T.-H., Wang Z., Wei D.-H., Liu L.-S., Zheng X.-L.. New role of PCSK9 in atherosclerotic inflammation promotion involving the TLR4/NF-κB pathway. Atherosclerosis (2017.0) 262 113-122. DOI: 10.1016/j.atherosclerosis.2017.04.023 47. Schlüter K.D., Wolf A., Schreckenberg R.. Coming Back to Physiology: Extra Hepatic Functions of Proprotein Convertase Subtilisin/Kexin Type 9. Front. Physiol. (2020.0) 11 598649. DOI: 10.3389/fphys.2020.598649 48. Ding Z., Liu S., Wang X., Deng X., Fan Y., Sun C., Wang Y., Mehta J.L.. Hemodynamic shear stress via ROS modulates PCSK9 expression in human vascular endothelial and smooth muscle cells and along the mouse aorta. Antioxid. Redox. Signal. (2015.0) 22 760-771. DOI: 10.1089/ars.2014.6054 49. Li H., Dong B., Park S.W., Lee H.S., Chen W., Liu J.. Hepatocyte nuclear factor 1alpha plays a critical role in PCSK9 gene transcription and regulation by the natural hypocholesterolemic compound berberine. J. Biol. Chem. (2009.0) 284 28885-28895. DOI: 10.1074/jbc.M109.052407 50. Yang X., Zhang J., Chen L., Wu Q., Yu C.. Chitosan oligosaccharides enhance lipid droplets via down-regulation of PCSK9 gene expression in HepG2 cells. Exp. Cell. Res. (2018.0) 366 152-160. DOI: 10.1016/j.yexcr.2018.03.013 51. Ilse T., Anna E., Daniel J.K., Jaap D., van B.. Chapter Five—Leukocytes Crossing the Endothelium: A Matter of Communication. Int. Rev. Cell Mol. Biol. (2016.0) 322 281-329. DOI: 10.1016/bs.ircmb.2015.10.005 52. Hadad N., Tuval L., Elgazar-Carmom V., Levy R., Levy R.. Endothelial ICAM-1 protein induction is regulated by cytosolic phospholipase A2α via both NFκB and CREB transcription factors. J. Immunol. (2011.0) 186 1816-1827. DOI: 10.4049/jimmunol.1000193 53. Sawa Y., Tsuruga E.. The expression of E-selectin and chemokines in the cultured human lymphatic endothelium with lipopolysaccharides. J. Anat. (2008.0) 212 654-663. DOI: 10.1111/j.1469-7580.2008.00892.x 54. Shyu A.B., Wilkinson M.F., van Hoof A.. Messenger RNA regulation: To translate or to degrade. EMBO J. (2008.0) 27 471-481. DOI: 10.1038/sj.emboj.7601977 55. Lawson C., Wolf S.. ICAM-1 signaling in endothelial cells. Pharmacol. Rep. (2009.0) 61 22-32. DOI: 10.1016/S1734-1140(09)70004-0 56. Auerbach S.D., Yang L., Luscinskas F.W., Ley K.. Endothelial ICAM-1 functions in adhesion and signaling during leukocyte recruitment. Adhesion Molecules: Function and Inhibition. Progress in Inflammation Research (2007.0) 99-116. DOI: 10.1007/978-3-7643-7975-9_4 57. Barnes P.J., Karin M.. Nuclear factor-κB—A pivotal transcription factor in chronic inflammatory diseases. N. Engl. J. Med. (1997.0) 336 1066-1071. DOI: 10.1056/NEJM199704103361506 58. Brasier A.R.. The nuclear factor-kappaB-interleukin-6 signalling pathway mediating vascular inflammation. Cardiovasc. Res. (2010.0) 86 211-218. DOI: 10.1093/cvr/cvq076 59. Leucker T.M., Gerstenblith G., Schar M., Brown T.T., Jones S.R., Afework Y., Weiss R.G., Hays A.G.. Evolocumab, a PCSK9-Monoclonal Antibody, Rapidly Reverses Coronary Artery Endothelial Dysfunction in People Living with HIV and People with Dyslipidemia. J. Am. Heart Assoc. (2020.0) 9 e016263. DOI: 10.1161/JAHA.120.016263 60. Davignon J., Ganz P.. Role of endothelial dysfunction in atherosclerosis. Circulation (2004.0) 109 27-32. DOI: 10.1161/01.CIR.0000131515.03336.f8 61. Sies H.. Oxidative stress: A redox biology and medicine concept. Redox Biol. (2015.0) 4 180-183. DOI: 10.1016/j.redox.2015.01.002 62. Landmesser U., Dikalov S., Price S.R., McCann L., Fukai T., Holland S.M., Mitch W.E., Harrison D.G.. Oxidation of tetrahydrobiopterin leads to uncoupling of endothelial cell nitric oxide synthase in hypertension. J. Clin. Investig. (2003.0) 111 1201-1209. DOI: 10.1172/JCI200314172 63. Sangle G.V., Shen G.X.. Signaling mechanisms for oxidized LDL-induced oxidative stress and the upregulation of plasminogen activator inhibitor-I in vascular cells. Clin. Lipidol. (2010.0) 5 221-232. DOI: 10.2217/clp.10.6 64. Stroes E., Colquhoun D., Sullivan D., Civeira F., Rosenson R.S., Watts G.F., Bruckert E., Cho L., Dent R., Knusel B.. Anti PCSK 9 antibody effectively lowers cholesterol in patients with statin intolerance: The GAUSS 2 randomized, placebo controlled phase 3 clinical trial of evolocumab. J. Am. Coll. Cardiol. (2014.0) 63 2541-2548. DOI: 10.1016/j.jacc.2014.03.019 65. Moriarty P.M., Jacobson T.A., Bruckert E., Thompson P.D., Guyton J.R., Baccara-Dinet M.T., Gipe D.I.. Efficacy and safety of alirocumab, a monoclonal antibody to PCSK-9, in statin-intolerant patients: Design and rationale of ODYSSEY ALTERNATIVE, a randomized phase 3 trial. J. Clin. Lipidol. (2014.0) 8 554-561. DOI: 10.1016/j.jacl.2014.09.007 66. Gamble J.R., Vadas M.A.. Endothelial adhesiveness for blood neutrophils is inhibited by transforming growth factor-beta. Science (1988.0) 242 97-99. DOI: 10.1126/science.3175638
--- title: Integrating Environmental Data with Medical Data in a Records-Linkage System to Explore Groundwater Nitrogen Levels and Child Health Outcomes authors: - Christine M. Prissel - Brandon R. Grossardt - Gregory S. Klinger - Jennifer L. St. Sauver - Walter A. Rocca journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049688 doi: 10.3390/ijerph20065116 license: CC BY 4.0 --- # Integrating Environmental Data with Medical Data in a Records-Linkage System to Explore Groundwater Nitrogen Levels and Child Health Outcomes ## Abstract Background: The Rochester Epidemiology Project (REP) medical records-linkage system offers a unique opportunity to integrate medical and residency data with existing environmental data, to estimate individual-level exposures. Our primary aim was to provide an archetype of this integration. Our secondary aim was to explore the association between groundwater inorganic nitrogen concentration and adverse child and adolescent health outcomes. Methods: We conducted a nested case-control study in children, aged seven to eighteen, from six counties of southeastern Minnesota. Groundwater inorganic nitrogen concentration data were interpolated, to estimate exposure across our study region. Residency data were then overlaid, to estimate individual-level exposure for our entire study population ($$n = 29$$,270). Clinical classification software sets of diagnostic codes were used to determine the presence of 21 clinical conditions. Regression models were adjusted for age, sex, race, and rurality. Results: The analyses support further investigation of associations between nitrogen concentration and chronic obstructive pulmonary disease and bronchiectasis (OR: 2.38, CI: 1.64–3.46) among boys and girls, thyroid disorders (OR: 1.44, CI: 1.05–1.99) and suicide and intentional self-inflicted injury (OR: 1.37, CI: >1.00–1.87) among girls, and attention deficit conduct and disruptive behavior disorders (OR: 1.34, CI: 1.24–1.46) among boys. Conclusions: Investigators with environmental health research questions should leverage the well-enumerated population and residency data in the REP. ## 1. Introduction Several previous environmental health studies have examined environmental exposures and health outcomes at an aggregated level [1,2,3]. Aggregation is often used because detailed residency data are unavailable through research databases in the United States [4]. Other studies have avoided the need to aggregate environmental data, by collecting individual exposure or residency data [5,6]. A common approach to collecting exposure and residency data, is through participant or household surveys [6,7]. With survey response rates often concerningly low, this method of data collection can be problematic [8]. Moreover, researchers employing survey methods must consider the generalizability of study results when notable proportions of the target population are not responding, and the potential recall bias from survey responders [9]. Other studies have employed samplers to ascertain location and individual exposures [10]. These studies are typically financially and time burdensome and therefore may not be practical for all environmental health research questions. When sufficient environmental data are available, spatial methods can be used to estimate individual exposure [11]. However, inconsistent or missing individual residency and health data can remain obstacles (e.g., unknown residency, inconsistent follow-up, inability to verify residency over time, unknown disease status, etc.). In part due to these obstacles, relatively few observational environmental health studies have used a well-enumerated population, with robust residency and health data, within the United States (US). The Rochester Epidemiology Project (REP) offers a unique infrastructure to overcome these obstacles and contribute to the field of environmental public health tracking (EPHT). The REP is a comprehensive medical records-linkage system, that covers nearly all Olmsted County residents and most residents in the 26 surrounding counties [12,13,14]. The REP’s well-enumerated population and the processes of verifying and collecting residency data of a dynamic population over time, provides a rare opportunity to integrate environmental and medical data [14,15]. Few REP studies have examined the environment in which the REP population resides and its relationship to various health outcomes [16,17,18,19]. MacLaughlin and colleagues recently examined human papillomavirus (HPV) vaccine uptake disparities, by estimated individual-level socioeconomic status [19]. Although this study did not examine environmental exposures, it did leverage the residential data available for the REP population [19]. Other studies have examined similar research questions at a group-level [16,17,18]. Rutten et al. estimated the odds of HPV vaccine initiation and series completion, based on geographical location, after adjusting for both US Census block and group level covariates [16]. Another study used participant residency to assign a corresponding US Census block group area deprivation index (ADI) [17]. REP researchers have also examined the association between US Census block group ADI and the likelihood of HPV vaccine initiation and completion [18]. Although these studies primarily focused on neighborhood demographic data, they have paved the way for the REP to emerge into the field of environmental epidemiology [16,17,18,19]. Currently, no environmental data are stored within the REP infrastructure [12,20]. This manuscript provides an example of how environmental data and spatial methods can be leveraged to estimate the effect of individual-level environmental exposures on clinically relevant outcomes, in a well-enumerated population, with robust residency and health data. No study to date has used the REP infrastructure in conjunction with environmental data to estimate individual-level environmental exposures and to examine a research question of interest. Nitrogen contamination of groundwater has become an increasing problem [21,22]. Inorganic nitrogen in the groundwater is both a direct health determinant and a proxy for other contaminants of concern [22]. These contaminants are often associated with human use of pesticides, a myriad of chemicals and chemical mixtures with the purpose of killing unwanted creatures, plants, and fungi [7,21,23]. High nitrogen contamination of groundwater is one mechanism by which environmental changes are reflective of human pesticide use [21]. Exposure to pesticides has been linked to numerous adverse health outcomes [7,24,25,26,27,28,29]. An increased risk of various cancers, diabetes, poor neurologic and cardiovascular health, endocrine disruption, and allergic effects have been linked to pesticide exposure [24,26,29,30,31]. Although children and adolescents are resilient, they are not impervious to the adverse health outcomes of pesticide exposure [7,32]. Previous studies have provided evidence of the harmful effects of pesticides on child and adolescent health [28,31,33,34,35,36,37,38,39]. However, studies of pesticide and fertilizer exposure among the general child and adolescent population have been sparse [25]. Furthermore, there have been limited studies examining the association of pesticide exposure and child and adolescent health outcomes within southeastern Minnesota—an area including both urban and rural regions [40]. This study explored the relationship in children and adolescents between groundwater inorganic nitrogen concentrations and 21 medical conditions, defined by the clinical classifications software (CCS) [41], to provide a proof-of-concept example of integrating address data and environmental data to estimate individual-level environmental exposures. This proof-of-concept, of how to efficiently estimate individual-level environmental exposure, will help advance EPHT and public health. The processes described in this study can be used to improve environmental hazard surveillance, help identify environmental health disparities, and inform targeted public health interventions for communities disproportionately affected by environmental hazards. The objectives of this study are twofold. Our primary objective was to provide a proof-of-concept example, integrating REP data and environmental data to estimate individual-level environmental exposure. Our secondary objective was to explore the association between groundwater inorganic nitrogen concentration (a marker of pesticide use) and child and adolescent health outcomes, using two groundwater data sources and a well-defined REP cohort. ## 2.1. Study Population We conducted a nested case-control study in southeastern Minnesota, using Minnesota Department of Agriculture (MDA), Olmsted County Public Health Services (OCPHS), and REP data, to explore the association between groundwater inorganic nitrogen concentration (mg/L), as a marker for fertilizer and pesticide exposure, and adverse health outcomes in children and adolescents, aged 7–18 years. A six-county area was identified in southeastern Minnesota (Olmsted, Goodhue, Filmore, Houston, Wabasha, and Winona counties), based on where MDA and OCPHS groundwater data overlapped with the REP region. MDA and OCPHS data were used to estimate inorganic nitrogen concentration (mg/L) exposure at the individual-level. We determined the complete resident population of all persons aged 7 to 18 years old on 1 January 2017, using the personal timelines available through the REP census [12,20]. Further exclusions to the population were made because parents or guardians (or adolescents themselves in the case of the 18-year-olds) had denied use of medical record data for research, under Minnesota research authorization [42]. Furthermore, persons were excluded when we were unable to link their mailing address with a geolocation (latitude and longitude), or when they had limited medical record information available in the 5-year period before 1 January 2017, thus limiting our ability to define case status for the case-control study design (see Figure 1) [13,15,20]. ## 2.2. Inorganic Nitrogen Concentration Inorganic nitrogen naturally occurs in groundwater [43]; however, nitrogen levels of ≥3 mg/L suggest human sources have contaminated the groundwater [23]. Groundwater inorganic nitrogen [nitrate (NO3) + nitrite (NO2)] concentrations were used as a marker of topsoil fertilizer and pesticide use. MDA groundwater samples from 2007–2018 (n ~ 516) and OCPHS groundwater samples from 2007–2017 (n ~ 5608) were used. Both MDA and OCPHS groundwater sample data contained inorganic nitrogen concentrations [nitrate (NO3) + nitrite (NO2)] and sample location. MDA and OCPHS inorganic nitrogen groundwater sample locations, with corresponding nitrogen concentrations, were plotted onto our six-county region using the ArcGIS (Geographic Information Systems) Pro 2.4 software. An ordinary kriging interpolation was used to estimate inorganic nitrogen concentrations at unsampled locations within our six-county region [44,45]. Previous studies have used ordinary kriging to estimate groundwater nitrate concentrations [44,45]. REP geolocation information (latitude and longitude) of residential addresses for our study population were obtained. Residential latitude and longitude were plotted onto our six-county region. We overlaid the residential location map layer onto our interpolated inorganic nitrogen concentration map layer (i.e., kriging), to estimate inorganic nitrogen concentration for each study participant. This process is illustrated in Figure 2. The individual inorganic nitrogen estimates were exported to a Statistical Analysis System (SAS) compatible file. Using the SAS 9.4 software, participants were categorized into high (≥3 mg/L) and low (<3 mg/L) inorganic nitrogen exposure, and association analyses were performed for each of the 21 conditions of interest [23]. High (≥3 mg/L) levels of inorganic nitrogen suggest human sources have contaminated the groundwater [23]. ## 2.3. Clinical Conditions in the Study Population Twenty-one CCS conditions were identified as outcomes of interest, using a multi-step process. The CCS diagnosis categorization system groups all possible International Classification of Diseases (ICD-9 and ICD-10) diagnosis codes into 285 clinically meaningful categories, across all body systems [41]. To select conditions most relevant to our pediatric population, we first conducted a thorough literature review of pesticide exposure and child–adolescent health outcomes. The evidence from the scientific literature was matched to specific CCS disease categories. Second, two authors reviewed the full list of CCS categories in detail and identified any additional conditions that could plausibly be related to pesticide exposure. Third, we excluded any CCS set of diagnostic codes that included the word “other” (e.g., other skin conditions), as these diagnosis categories are extremely broad and often contain a myriad of diagnoses [41]. Finally, only CCS conditions that could affect both boys and girls were examined. For each person, all electronically available diagnostic codes in the 5-year period before 1 January 2017 were included, to define the case status (i.e., diagnoses from 1 January 2012 to 31 December 2016). Persons were considered “cases” if they had two codes separated by more than 30 days, from among the codes in a defined CCS category. Controls were all persons in the full population of children and adolescents. ## 2.4. Statistical Analyses Controls were defined as REP persons who were at risk of one or more CCS conditions at baseline in our study population ($$n = 29$$,270). This definition allows controls to estimate the exposure prevalence in the entire study population at baseline [46,47,48]. A logistic regression model was used to estimate the effect of high vs. low groundwater inorganic nitrogen on various child and adolescent health outcomes (CCS conditions). Analyses were carried out overall, and separately for boys and girls. Overall analyses were adjusted for age (continuous, centered at the midpoint of 12.5 years), sex, non-white race, and rurality. Rural and urban categories were determined using the rural–urban commuting area (RUCA) codes, available from the Economic Research Service, part of the US Department of Agriculture [40]. These potential confounders were identified through a review of scientific literature and scientific reasoning. We reasoned that sex, as a surrogate for cultural and behavioral gender norms, may impact nitrogen exposure and the likelihood of seeking medical attention for health conditions [49,50,51]. If the child or adolescent was not seen at a hospital or clinic, then they would not have received a diagnostic code for a CCS condition, and therefore would not be counted as a case in this study. We reasoned that age may impact nitrogen exposure, through hygiene, play, and other behaviors that change throughout childhood and adolescent years, and that age may also impact our adverse health outcomes of interest, through the number of health care visits [49,52,53]. We used non-white race to adjust for potential racial disparities and race as a social construct, reasoning that historical and current racial disparities may impact adolescent nitrogen exposure and access to health care [53,54,55]. Lastly, we adjusted analyses for rurality (rural vs. urban) for two reasons. First, previous studies have shown a difference in clinical care utilization by rurality [56,57,58]. Second, other studies have provided evidence of potential different rural and urban behavioral norms related to pesticide and fertilizer use and exposure [39,58,59]. Plotted points were enlarged and skewed in Figure 2, to ensure confidentiality of study participants. Additionally, sex was examined as a possible effect modifier. Our hypothesis for this was two-fold. First, we hypothesized that sex may be an effect modifier, through biological and hormonal differences in the older age stratum (i.e., children who have experienced puberty). We did not anticipate this hypothesis to hold for the younger age stratum. Secondly, we hypothesized that sex may act as a surrogate variable for gender-based cultural and behavioral norms among children and adolescents. Analyses performed separately by sex strata were adjusted for age (continuous), non-white race, and rurality. ## 3.1. Study Demographics Our study population included 29,270 children, ranging in age from 7 to 18 years, who resided in a 6-county REP region of southeastern Minnesota on 1 January 2017 (Table 1). Children and adolescents were split approximately evenly between girls ($48.8\%$) and boys ($51.2\%$), and the sex distribution was similar across rural versus urban strata. Age was reported, for descriptive purposes, as two 6-year periods. However, in analyses, age was entered in the models as a continuous variable, centered at the mid-point of 12.5 years. Children within the age range of 7 to 12 years made up $52.9\%$ of our population, whereas adolescents aged 13 to 18 made up $47.1\%$. The proportion of our population who was of non-white race varied substantially across urban and rural strata. In the urban strata, $22.8\%$ identified as non-white race, whereas only $9.2\%$ identified as non-white race in the rural strata. This resulted in an overall study distribution of $78.1\%$ of participants identifying as white and $21.9\%$ identifying as non-white. The distribution of high and low nitrogen exposure was markedly different between urban and rural strata. We found that $76.2\%$ of children and adolescents living in rural areas had high exposure to nitrogen, whereas only $26.7\%$ of children living in urban areas had high exposure to nitrogen. ## 3.2. Results for Child and Adolescent Boys and Girls After adjusting for sex, age, rurality, and non-white race, the odds of chronic obstructive pulmonary disease and bronchiectasis among children with high nitrogen exposure was 2.38 (CI: 1.64–3.46) times the odds among children with low nitrogen exposure (Table 2, All children column). A positive association between high nitrogen exposure and chronic obstructive pulmonary disease and bronchiectasis was also observed in the sex specific strata; the odds ratios were 2.63 (CI: 1.69–4.10) for girls and 2.15 (CI: 1.38–3.35) for boys (Table 2, sex specific columns). ## 3.3. Sex Specific Results In addition, after adjusting for age, rurality, and non-white race, we observed that the following conditions had notably different odds among boys and girls. The odds of thyroid disorders among girls with high nitrogen exposure was 1.44 (CI: 1.05–1.99) times the odds with low nitrogen exposure (Table 2, girls only column). Moreover, among girls, after adjusting for age, rurality, and non-white race, we observed that the odds of suicide and intentional self-inflicted injury with high nitrogen exposure was 1.37 (CI: >1.00–1.87) times the odds with low nitrogen exposure (Table 2, girls only column). For child and adolescent boys, after adjusting for age, rurality, and non-white race, we observed the odds of attention deficit conduct and disruptive behavior disorders with high nitrogen exposure was 1.34 (CI: 1.24–1.46) times the odds with low nitrogen exposure (Table 2, boys only column). ## 4. Discussion The primary objective of this study was to provide an archetype of how REP medical and residency data, along with existing environmental data, can be leveraged to test clinically relevant hypotheses. We recognize that these geospatial methods are commonly used in environmental health studies. However, the novelty of this study is using these methods within a well-enumerated population, with robust residency and medical data [12]. Leveraging existing environmental data, and applying appropriate geospatial methods, will allow researchers to estimate individual-level environmental exposures at a significantly lower financial and time cost, compared to using study specific sampling. In part due to the limited environmental data maintained within the REP infrastructure, few studies to date have leveraged the REP robust medical and residency data. However, a few recent REP studies have investigated neighborhood demographics and associated health outcomes [16,17,18,19]. To advance public health, it is critical that clinically relevant studies begin to examine the impact of the places people work, live, and play [60]. In response, the REP has begun to develop the REP environmental exposure data dictionary. This searchable tool allows REP investigators to identify environmental data collected within the REP region and provides summary information about these environmental data (e.g., variables available, dates data were collected, data request form and contact, etc.). Our exploratory results support further investigation of the association between high inorganic nitrogen concentrations and chronic obstructive pulmonary disease and bronchiectasis among children and adolescents. Specifically, for girls, further investigation is needed to understand the association of high nitrogen concentration, thyroid disorders, and suicide and intentional self-inflicted injury. For boys, we found evidence that the association of nitrogen concentration and attention deficit conduct and disruptive behavior disorders should be explored further. A few CCS conditions showed reduced odds for those with high groundwater inorganic nitrogen concentrations (i.e., headache, including migraine, vision defects, dizziness or vertigo, nausea and vomiting, mood and developmental disorders). There is little to no biologic plausibility that high nitrogen concentrations are causing the reduced odds of these conditions [31,32,59]. Rather, we hypothesize that these odds reductions are due to decreased screening and diagnosing of these conditions in rural compared to urban regions. Healthcare behaviors and access differences among rural and urban residents impact screening, diagnosis, and treatment of these conditions. For example, if a child or adolescent was not diagnosed with nausea and vomiting at a healthcare facility, then they were not defined as a case, even if they had experienced nausea and vomiting. Moreover, a priori testing suggested that high nitrogen concentrations were highly correlated with rural residency. A limitation of this study is that groundwater samples were not collected for the purpose of this study. We worked extensively with partners from the OCPHS environmental health department and MDA to understand the data and their collection processes. Investigators of future studies, that use these or other environmental data, are strongly encouraged to partner with owners of environmental data, to ensure accurate understanding, use, and interpretation. Groundwater nitrogen values may vary, due to geology differences throughout our study region [61]. This limitation was a motivator for analyzing nitrogen exposure as a dichotomous variable, based on levels that suggest that human sources have contaminated the groundwater (i.e., high, ≥3 mg/L and low, <3 mg/L) [23], rather than examining nitrogen exposure as a continuous measurement. Therefore, we were unable to assess a dose–response relationship for nitrogen levels and outcomes. Moreover, it is important to note that the groundwater samples used in this study are not synonymous with drinking well water samples. Other studies have used well data to examine nitrogen exposure from drinking water. However, our study focused on the health risks of indirect exposure, and therefore was not restricted to well water samples. Our exploratory results do not provide causal evidence for the association of high inorganic nitrogen concentrations and adverse child and adolescent health conditions. Some CCS conditions require pesticide breakdown products to cross the blood brain barrier [62,63]. High inorganic nitrogen levels suggest that human-made chemicals were used [23], but we did not examine levels of specific breakdown products (e.g., atrazine). However, our exploratory results can be used to identify signals for further investigation. Additionally, we did not adjust for education or family income, because education attainment correlates with the age of our study population, and family income was hypothesized to affect child exposure through residency (Figure A1 in Appendix A). Age and residency were adjusted for in our models. Finally, we excluded $8\%$ of children because they did not have sufficient medical record information to assess medical conditions. A lack of medical record information may indicate recent movement into our population (immigration), extremely low to no utilization of, or access to, care, or children who are particularly healthy. ## 5. Conclusions The primary purpose of this study was to provide an archetype of how REP medical and residency data, along with existing environmental data, can be leveraged to improve our understanding of the complex relationships between our environment and health. This knowledge may inform evidence-based public health policies and interventions, aimed at reducing the impact of environmental hazards on public health. The REP provides a unique infrastructure, with robust data on a well-enumerated population over time [12]. This longstanding infrastructure, combined with environmental data sources, allows for investigators to estimate individual-level environmental exposures. This study provides a path forward for future investigators to delve into the field of EPHT and examine additional environmental exposures and health outcomes. The REP has supported the effort to consolidate environmental data sources that overlap with the REP 27-county region for future investigator use. We anticipate that the proof-of-concept demonstrated in this study will be utilized in future research, to monitor environmental hazards, assess their impact on public health, and investigate their potential burden on the healthcare system. ## References 1. Arcaya M.C., Tucker-Seeley R.D., Kim R., Schnake-Mahl A., So M., Subramanian S.. Research on neighborhood effects on health in the United States: A systematic review of study characteristics. Soc. Sci. Med. (2016.0) 168 16-29. DOI: 10.1016/j.socscimed.2016.08.047 2. Pekkanen J., Pearce N.. Environmental epidemiology: Challenges and opportunities. Environ. Health Perspect. (2001.0) 109 1-5. PMID: 11171517 3. Reynolds P., Von Behren J., Gunier R.B., Goldberg D.E., Hertz A., Harnly M.E.. Childhood cancer and agricultural pesticide use: An ecologic study in California. Environ. Health Perspect. (2002.0) 110 319-324. DOI: 10.1289/ehp.02110319 4. Nattinger A.B., Pezzin L.E., Sparapani R.A., Neuner J.M., King T.K., Laud P.W.. Heightened attention to medical privacy: Challenges for unbiased sample recruitment and a possible solution. Am. J. Epidemiol. (2010.0) 172 637-644. DOI: 10.1093/aje/kwq220 5. Gibb R., Browning E., Glover-Kapfer P., Jones K.E.. Emerging opportunities and challenges for passive acoustics in ecological assessment and monitoring. Methods Ecol. Evol. (2019.0) 10 169-185. DOI: 10.1111/2041-210X.13101 6. Hillhouse J., Turrisi R., Jaccard J., Robinson J.. Accuracy of self-reported sun exposure and sun protection behavior. Prev. Sci. (2012.0) 13 519-531. DOI: 10.1007/s11121-012-0278-1 7. Roberts J.R., Karr C.J., Health C.o.E., Paulson J.A., Brock-Utne A.C., Brumberg H.L., Campbell C.C., Lanphear B.P., Osterhoudt K.C., Sandel M.T.. Pesticide exposure in children. Pediatrics (2012.0) 130 e1765-e1788. DOI: 10.1542/peds.2012-2758 8. Blumenberg C., Barros A.J.. Response rate differences between web and alternative data collection methods for public health research: A systematic review of the literature. Int. J. Public Health (2018.0) 63 765-773. DOI: 10.1007/s00038-018-1108-4 9. Althubaiti A.. Information bias in health research: Definition, pitfalls, and adjustment methods. J. Multidiscip. Healthc. (2016.0) 9 211. DOI: 10.2147/JMDH.S104807 10. El-Nahhal I., El-Nahhal Y.. Pesticide residues in drinking water, their potential risk to human health and removal options. J. Environ. Manag. (2021.0) 299 113611. DOI: 10.1016/j.jenvman.2021.113611 11. Nuckols J.R., Ward M.H., Jarup L.. Using geographic information systems for exposure assessment in environmental epidemiology studies. Environ. Health Perspect. (2004.0) 112 1007-1015. DOI: 10.1289/ehp.6738 12. Rocca W.A., Grossardt B.R., Brue S.M., Bock-Goodner C.M., Chamberlain A.M., Wilson P.M., Finney Rutten L.J., St Sauver J.L.. Data Resource Profile: Expansion of the Rochester Epidemiology Project medical records-linkage system (E-REP). Int. J. Epidemiol. (2018.0) 47 368-368j. DOI: 10.1093/ije/dyx268 13. Rocca W.A., Yawn B.P., St Sauver J.L., Grossardt B.R., Melton L.J.. History of the Rochester Epidemiology Project: Half a century of medical records linkage in a US population. Mayo. Clin. Proc. (2012.0) 87 1202-1213. DOI: 10.1016/j.mayocp.2012.08.012 14. St Sauver J.L., Grossardt B.R., Leibson C.L., Yawn B.P., Melton L.J., Rocca W.A.. Generalizability of epidemiological findings and public health decisions: An illustration from the Rochester Epidemiology Project. Mayo. Clin. Proc. (2012.0) 87 151-160. DOI: 10.1016/j.mayocp.2011.11.009 15. St Sauver J.L., Grossardt B.R., Yawn B.P., Melton L.J., Rocca W.A.. Use of a medical records linkage system to enumerate a dynamic population over time: The Rochester epidemiology project. Am. J. Epidemiol. (2011.0) 173 1059-1068. DOI: 10.1093/aje/kwq482 16. Finney Rutten L.J., Wilson P.M., Jacobson D.J., Agunwamba A.A., Radecki Breitkopf C., Jacobson R.M., St Sauver J.L.. A Population-Based Study of Sociodemographic and Geographic Variation in HPV Vaccination. Cancer Epidemiol. Biomark. Prev. (2017.0) 26 533-540. DOI: 10.1158/1055-9965.EPI-16-0877 17. Chamberlain A.M., Finney Rutten L.J., Wilson P.M., Fan C., Boyd C.M., Jacobson D.J., Rocca W.A., St Sauver J.L.. Neighborhood socioeconomic disadvantage is associated with multimorbidity in a geographically-defined community. BMC Public Health (2020.0) 20. DOI: 10.1186/s12889-019-8123-0 18. Kurani S., MacLaughlin K.L., Jacobson R.M., St Sauver J.L., Jenkins G.D., Fan C., Jacobson D.J., Inselman J., Zhu X., Griffin J.M.. Socioeconomic disadvantage and human papillomavirus (HPV) vaccination uptake. Vaccine (2022.0) 40 471-476. DOI: 10.1016/j.vaccine.2021.12.003 19. MacLaughlin K.L., Jacobson R.M., Sauver J.L.S., Jacobson D.J., Fan C., Wi C.-I., Rutten L.J.F.. An innovative housing-related measure for individual socioeconomic status and human papillomavirus vaccination coverage: A population-based cross-sectional study. Vaccine (2020.0) 38 6112-6119. DOI: 10.1016/j.vaccine.2020.07.026 20. St Sauver J.L., Grossardt B.R., Yawn B.P., Melton L.J., Pankratz J.J., Brue S.M., Rocca W.A.. Data resource profile: The Rochester Epidemiology Project (REP) medical records-linkage system. Int. J. Epidemiol. (2012.0) 41 1614-1624. DOI: 10.1093/ije/dys195 21. Morrissy J.G., Currell M.J., Reichman S.M., Surapaneni A., Megharaj M., Crosbie N.D., Hirth D., Aquilina S., Rajendram W., Ball A.S.. Nitrogen contamination and bioremediation in groundwater and the environment: A review. Earth-Sci. Rev. (2021.0) 222 103816. DOI: 10.1016/j.earscirev.2021.103816 22. Burri N.M., Weatherl R., Moeck C., Schirmer M.. A review of threats to groundwater quality in the anthropocene. Sci. Total Environ. (2019.0) 684 136-154. DOI: 10.1016/j.scitotenv.2019.05.236 23. 23. Minnesota Department of Health Well Management Section Nitrate-Nitrogen in Well WaterMinnesota Department of HealthSt. Paul, MN, USA2019. Nitrate-Nitrogen in Well Water (2019.0) 24. Bassil K.L., Vakil C., Sanborn M., Cole D.C., Kaur J.S., Kerr K.J.. Cancer health effects of pesticides: Systematic review. Can. Fam. Physician (2007.0) 53 1704-1711. PMID: 17934034 25. Roingeard C., Monnereau A., Goujon S., Orazio S., Bouvier G., Vacquier B.. Passive environmental residential exposure to agricultural pesticides and hematological malignancies in the general population: A systematic review. Environ. Sci. Pollut. Res. (2021.0) 28 43190-43216. DOI: 10.1007/s11356-021-14789-3 26. Sanborn M., Kerr K.J., Sanin L.H., Cole D.C., Bassil K.L., Vakil C.. Non-cancer health effects of pesticides: Systematic review and implications for family doctors. Can. Fam. Physician (2007.0) 53 1712-1720. PMID: 17934035 27. Townsend M., Peck C., Meng W., Heaton M., Robison R., O’Neill K.. Evaluation of various glyphosate concentrations on DNA damage in human Raji cells and its impact on cytotoxicity. Regul. Toxicol. Pharmacol. (2017.0) 85 79-85. DOI: 10.1016/j.yrtph.2017.02.002 28. Pinos H., Carrillo B., Merchan A., Biosca-Brull J., Perez-Fernandez C., Colomina M.T., Sanchez-Santed F., Martin-Sanchez F., Collado P., Arias J.L.. Relationship between Prenatal or Postnatal Exposure to Pesticides and Obesity: A Systematic Review. Int. J. Environ. Res. Public Health (2021.0) 18. DOI: 10.3390/ijerph18137170 29. Evangelou E., Ntritsos G., Chondrogiorgi M., Kavvoura F.K., Hernández A.F., Ntzani E.E., Tzoulaki I.. Exposure to pesticides and diabetes: A systematic review and meta-analysis. Environ. Int. (2016.0) 91 60-68. DOI: 10.1016/j.envint.2016.02.013 30. Parrón T., Requena M., Hernández A.F., Alarcón R.. Association between environmental exposure to pesticides and neurodegenerative diseases. Toxicol. Appl. Pharmacol. (2011.0) 256 379-385. DOI: 10.1016/j.taap.2011.05.006 31. Buralli R.J., Dultra A.F., Ribeiro H.. Respiratory and Allergic Effects in Children Exposed to Pesticides-A Systematic Review. Int. J. Environ. Res. Public Health (2020.0) 17. DOI: 10.3390/ijerph17082740 32. Dereumeaux C., Fillol C., Quénel P., Denys S.. Pesticide exposures for residents living close to agricultural lands: A review. Environ. Int. (2020.0) 134 105210. DOI: 10.1016/j.envint.2019.105210 33. Muñoz-Quezada M.T., Lucero B.A., Barr D.B., Steenland K., Levy K., Ryan P.B., Iglesias V., Alvarado S., Concha C., Rojas E.. Neurodevelopmental effects in children associated with exposure to organophosphate pesticides: A systematic review. Neurotoxicology (2013.0) 39 158-168. DOI: 10.1016/j.neuro.2013.09.003 34. Van Maele-Fabry G., Gamet-Payrastre L., Lison D.. Household exposure to pesticides and risk of leukemia in children and adolescents: Updated systematic review and meta-analysis. Int. J. Hyg. Environ. Health (2019.0) 222 49-67. DOI: 10.1016/j.ijheh.2018.08.004 35. Van Maele-Fabry G., Hoet P., Lison D.. Parental occupational exposure to pesticides as risk factor for brain tumors in children and young adults: A systematic review and meta-analysis. Environ. Int. (2013.0) 56 19-31. DOI: 10.1016/j.envint.2013.02.011 36. Van Maele-Fabry G., Lantin A.-C., Hoet P., Lison D.. Residential exposure to pesticides and childhood leukaemia: A systematic review and meta-analysis. Environ. Int. (2011.0) 37 280-291. DOI: 10.1016/j.envint.2010.08.016 37. González-Alzaga B., Lacasaña M., Aguilar-Garduño C., Rodríguez-Barranco M., Ballester F., Rebagliato M., Hernández A.. A systematic review of neurodevelopmental effects of prenatal and postnatal organophosphate pesticide exposure. Toxicol. Lett. (2014.0) 230 104-121. DOI: 10.1016/j.toxlet.2013.11.019 38. Wigle D.T., Turner M.C., Krewski D.. A systematic review and meta-analysis of childhood leukemia and parental occupational pesticide exposure. Environ. Health Perspect. (2009.0) 117 1505-1513. DOI: 10.1289/ehp.0900582 39. Garry V.F.. Pesticides and children. Toxicol. Appl. Pharmacol. (2004.0) 198 152-163. DOI: 10.1016/j.taap.2003.11.027 40. Rural-Urban Commuting Area Codes (RUCA) 41. Clinical Classifications Software (CCS) for ICD-10-PCS 42. Yawn B.P., Yawn R.A., Geier G.R., Xia Z., Jacobsen S.J.. The impact of requiring patient authorization for use of data in medical records research. J. Fam. Pract. (1998.0) 47 361-365. PMID: 9834771 43. Meah M., Harrison N., Davies A.. Nitrate and nitrite in foods and the diet. Food Addit. Contam. (1994.0) 11 519-532. DOI: 10.1080/02652039409374250 44. Tipping R.G.. A Hydrochemical Survey of Groundwater Flow in the Rochester Metropolitan Area, Minnesota. OFR14-05 (2014.0) 45. Arslan H.. Spatial and temporal mapping of groundwater salinity using ordinary kriging and indicator kriging: The case of Bafra Plain, Turkey. Agric. Water Manag. (2012.0) 113 57-63. DOI: 10.1016/j.agwat.2012.06.015 46. Langholz B.. Case-control studies= odds ratios: Blame the retrospective model. Epidemiology (2010.0) 21 10-12. DOI: 10.1097/EDE.0b013e3181c308f5 47. Knol M.J., Vandenbroucke J.P., Scott P., Egger M.. What do case-control studies estimate? Survey of methods and assumptions in published case-control research. Am. J. Epidemiol. (2008.0) 168 1073-1081. DOI: 10.1093/aje/kwn217 48. Labrecque J.A., Hunink M.M., Ikram M.A., Ikram M.K.. Do case-control studies always estimate odds ratios?. Am. J. Epidemiol. (2021.0) 190 318-321. DOI: 10.1093/aje/kwaa167 49. Rand C.M., Shone L.P., Albertin C., Auinger P., Klein J.D., Szilagyi P.G.. National health care visit patterns of adolescents: Implications for delivery of new adolescent vaccines. Arch. Pediatr. Adolesc. Med. (2007.0) 161 252-259. DOI: 10.1001/archpedi.161.3.252 50. Vaidya V., Partha G., Karmakar M.. Gender differences in utilization of preventive care services in the United States. J. Women’s Health (2012.0) 21 140-145. DOI: 10.1089/jwh.2011.2876 51. Marcell A.V., Klein J.D., Fischer I., Allan M.J., Kokotailo P.K.. Male adolescent use of health care services: Where are the boys?. J. Adolesc. Health (2002.0) 30 35-43. DOI: 10.1016/S1054-139X(01)00319-6 52. Pascale A., Laborde A.. Impact of pesticide exposure in childhood. Rev. Environ. Health (2020.0) 35 221-227. DOI: 10.1515/reveh-2020-0011 53. Mulye T.P., Park M.J., Nelson C.D., Adams S.H., Irwin C.E., Brindis C.D.. Trends in adolescent and young adult health in the United States. J. Adolesc. Health (2009.0) 45 8-24. DOI: 10.1016/j.jadohealth.2009.03.013 54. Delgado R., Stefancic J.. Critical Race Theory: An Introduction (2017.0) Volume 20 55. Lasser K.E., Himmelstein D.U., Woolhandler S.. Access to care, health status, and health disparities in the United States and Canada: Results of a cross-national population-based survey. Am. J. Public Health (2006.0) 96 1300-1307. DOI: 10.2105/AJPH.2004.059402 56. Cyr M.E., Etchin A.G., Guthrie B.J., Benneyan J.C.. Access to specialty healthcare in urban versus rural US populations: A systematic literature review. BMC Health Serv. Res. (2019.0) 19. DOI: 10.1186/s12913-019-4815-5 57. Chen X., Orom H., Hay J.L., Waters E.A., Schofield E., Li Y., Kiviniemi M.T.. Differences in rural and urban health information access and use. J. Rural. Health (2019.0) 35 405-417. DOI: 10.1111/jrh.12335 58. Zhang P., Too G., Irwin K.L.. Utilization of preventive medical services in the United States: A comparison between rural and urban populations. J. Rural. Health (2000.0) 16 349-356. DOI: 10.1111/j.1748-0361.2000.tb00485.x 59. Arcury T.A., Chen H., Quandt S.A., Talton J.W., Anderson K.A., Scott R.P., Jensen A., Laurienti P.J.. Pesticide exposure among Latinx children: Comparison of children in rural, farmworker and urban, non-farmworker communities. Sci. Total Environ. (2021.0) 763 144233. DOI: 10.1016/j.scitotenv.2020.144233 60. Frumkin H.. Environmental Health: From Global to Local (2016.0) 61. Chae G.-T., Kim K., Yun S.-T., Kim K.-H., Kim S.-O., Choi B.-Y., Kim H.-S., Rhee C.W.. Hydrogeochemistry of alluvial groundwaters in an agricultural area: An implication for groundwater contamination susceptibility. Chemosphere (2004.0) 55 369-378. DOI: 10.1016/j.chemosphere.2003.11.001 62. Rosas L.G., Eskenazi B.. Pesticides and child neurodevelopment. Curr. Opin. Pediatr. (2008.0) 20 191-197. DOI: 10.1097/MOP.0b013e3282f60a7d 63. Burns C.J., McIntosh L.J., Mink P.J., Jurek A.M., Li A.A.. Pesticide exposure and neurodevelopmental outcomes: Review of the epidemiologic and animal studies. J. Toxicol. Environ. Health Part B (2013.0) 16 127-283. DOI: 10.1080/10937404.2013.783383
--- title: Activation of TLRs Triggers GLP-1 Secretion in Mice authors: - Lorène J. Lebrun - Alois Dusuel - Marion Xolin - Naig Le Guern - Jacques Grober journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049702 doi: 10.3390/ijms24065333 license: CC BY 4.0 --- # Activation of TLRs Triggers GLP-1 Secretion in Mice ## Abstract The gastrointestinal tract constitutes a large interface with the inner body and is a crucial barrier against gut microbiota and other pathogens. As soon as this barrier is damaged, pathogen-associated molecular patterns (PAMPs) are recognized by immune system receptors, including toll-like receptors (TLRs). Glucagon-like peptide 1 (GLP-1) is an incretin that was originally involved in glucose metabolism and recently shown to be rapidly and strongly induced by luminal lipopolysaccharides (LPS) through TLR4 activation. In order to investigate whether the activation of TLRs other than TLR4 also increases GLP-1 secretion, we used a polymicrobial infection model through cecal ligation puncture (CLP) in wild-type and TLR4-deficient mice. TLR pathways were assessed by intraperitoneal injection of specific TLR agonists in mice. Our results show that CLP induces GLP-1 secretion both in wild-type and TLR4-deficient mice. CLP and TLR agonists increase gut and systemic inflammation. Thus, the activation of different TLRs increases GLP-1 secretion. This study highlights for the first time that, in addition to an increased inflammatory status, CLP and TLR agonists also strongly induce total GLP-1 secretion. Microbial-induced GLP-1 secretion is therefore not only a TLR4/LPS-cascade. ## 1. Introduction There is an extensive interface between the gastrointestinal tract and its surrounding environment. Therefore, in addition to nutrient absorption, the gastrointestinal tract plays an important role as a barrier against the microbial components of the gut. Even if the gut microbiota is crucial for human physiology, with roles including non-digestible carbohydrate metabolism, antimicrobial protection and immunomodulation [1], it has to be kept separate from the rest of the inner body. Given these conditions, it is consistent that the gut hosts the most developed immune system in the organism: the gut-associated lymphoid tissues (GALTs) [2]. Alterations in the gut barrier and subsequent pathogen invasion activate inflammatory processes and can have disastrous consequences. When the gut barrier is damaged, the host’s innate immune system rapidly recognizes and responds to pathogen-associated molecular patterns (PAMPs) from microbes via specific receptors, including toll-like receptors (TLRs) [3], which induce immune cell recruitment and cytokine secretion. Glucagon-like peptide 1 (GLP-1) is an incretin produced by enteroendocrine cells (EECs). It was originally described for its ability to stimulate insulin secretion and for its implication in glucose metabolism [4], but in addition to these glucoregulatory functions, GLP-1 also exerts anti-inflammatory effects [5]. GLP-1 agonists have been shown to have protective effects on the cardiovascular system [6,7,8], and, in humans, GLP-1 plasma levels are increased in sepsis and critical illness [9]. In patients undergoing cardiac surgery with cardiopulmonary bypass, a situation known to alter the gut barrier [10], we have shown that GLP-1 was associated with poor clinical outcomes [11]. In addition, it has been suggested that GLP-1 has a role in maintaining gut homeostasis [12], but the literature on this is scarce. Liraglutide, a GLP-1 agonist, has been shown to attenuate intestinal ischemia/reperfusion injury in mice [13]. Moreover, it was suggested to have a protective role in gut permeability [14]. GLP-1 secretion mechanisms have been extensively described [15], and gut microbiota is involved in these secretory pathways through different mechanisms. Short-chain fatty acids produced by gut bacteria increase GLP-1 secretion through G-protein-coupled receptors GPR41/GPR43 [16]. Microbe-metabolized secondary bile acids increase GLP-1 secretion [17]. Bacterial components such as peptides from Akkermansia muciniphila [18] and from Staphylococcus epidermidis [19] have also been associated with enhanced GLP-1 secretion. Moreover, we and others have previously shown that inflammatory bacterial-derived lipopolysaccharides (LPS) increase GLP-1 plasma levels [20,21]. LPS molecules are able to directly activate the TLR4 in enteroendocrine L cells, leading to a significant and rapid increase in GLP-1 secretion upon gut-barrier alterations [22]. This LPS/GLP-1 cascade underlines the critical role of EECs as the key mucosal sensor of gut injury [23]. In different studies aiming to mimic sepsis using the cecal ligation puncture (CLP), which is one of the most frequently used models, TLR expression was modulated in the lungs, liver, kidneys and intestines [24,25,26]. EECs express multiple TLRs [27] and are targets of pro-inflammatory cytokines such as interleukin-6 (IL6) and tumor necrosis factor alpha (TNFA), which are known to stimulate GLP-1 secretion [15]. The aim of this study was therefore to investigate whether, besides TLR4, the only TLR involved to date in GLP-1 secretion, other TLRs could also be implicated in the control of GLP-1 secretion. For this purpose, the CLP model was performed in wild-type (WT) and TLR4-deficient (Tlr4-/-) mice. In addition, after mice were injected with TLR-specific agonists, we assessed cytokines, TLR expression and GLP-1 secretion. Our data suggest for the first time that the activation of TLRs other than TLR4 is involved in GLP-1 secretion. ## 2.1. CLP Induces Systemic Inflammation and Modulates Inflammatory Gene Expression in the Gut The quantification of plasma cytokines revealed a major rise in inflammatory markers after CLP (Figure 1A). Cytokine mRNA levels were significantly increased in the ileum of CLP-treated mice compared to sham mice (Figure 1B). An analysis of the cytokines in the ileum showed that IL6, interleukin-1 beta (IL1B) and C-C motif chemokine ligand 2 (CCL2) were significantly increased after CLP (Figure 1C). The mRNA levels in TLRs (1 to 9) were then assessed in the gut after CLP treatment (Figure 1D), showing a significant induction of Tlr2 and Tlr7 mRNA levels. ## 2.2. TLR Agonists Increase Cytokines and Expression of TLRs Intraperitoneal injections of different TLR agonists induced inflammatory status and increased TLR gene expression in the gut (Figure 2). Plasma levels of pro- or anti-inflammatory cytokines were assessed (Figure 2A–E). Most of the analyzed cytokines were increased upon stimulation with specific TLR agonists. One can note that only TLR4, through its activation by LPS, led to an increase in all the cytokines. The mRNA gene expression of these cytokines was then measured in the ileum (Figure 2F–J). Here again, the mRNA expression of many cytokines increased through activation by TLRs, with TLR4 being the receptor that induced increases in all mRNA levels. Lastly, mRNA gene expression of the different TLRs was quantified in the ileum (Figure 2K). Among the nine studied TLRs, four had either a reduced (Tlr5 and Tlr6) or an identical (Tlr1 and Tlr9) mRNA expression. ## 2.3. GLP-1 Secretion Is Mediated through Multiple TLRs Total plasma GLP-1 concentrations were assessed in CLP-induced and sham control mice. As shown in Figure 3A, CLP significantly increased GLP-1 secretion 3 h after the onset of CLP. The increased secretion was maintained at 6 and 24 h (data not shown). In Tlr4-/- mice, meaning in the absence of a functional TLR4 receptor, CLP still induced a significant increase in GLP-1 secretion compared to sham control mice (Figure 3B), suggesting the involvement of TLRs other than TLR4 in CLP-induced GLP-1 secretion. The total GLP-1 plasma levels were also investigated after the injection of different TLRs agonists compared to sodium chloride (NaCl)-injected mice (Figure 3C). Most of the TLRs agonists tested were able to induce the secretion of GLP-1. ## 3. Discussion Here, we show that a CLP model induces a rapid (3 h) and significant increase in GLP-1 plasma levels and that this effect is associated with changes in TLR gene expression in the ileum. Moreover, CLP-induced GLP-1 secretion also occurs in Tlr4-/- mice, suggesting that PAMPs other than LPS induce a significant increase in GLP-1 secretion through the activation of TLRs. GLP-1 is a gut-derived hormone generally described in the literature as playing an important role in glucose homeostasis by increasing insulin secretion, suppressing glucagon expression and also inhibiting gastric emptying and reducing food intake. GLP-1 is produced by EECs, which represent only $1\%$ of intestinal epithelial cells. Despite this small proportion, the hormonal secretion of EECs is essential for the coordination of food intake, digestion and absorption [28]. However, an intestinal immuno-endocrine axis has been described for EECs, which may play a key role in the gut immune response to both pathogens and commensal bacteria [23]. Moreover, inflammatory bowel diseases (IBDs) are associated with an increase in the number of these cells in the ileum [29]. Cells from the gut’s immune system express a wide range of peptide receptors produced by EECs [30]: ghrelin increases the activation and proliferation of T lymphocytes [31], and cholecystokinin stimulates the production of acetylcholine by B cells and thus has an effect on neutrophil recruitment [32]. In addition, both in vitro and in vivo studies demonstrate that EECs also express certain TLRs [27,33], and the activation of these TLRs induce hormone secretions: GLP-1 after TLR4 activation [22]; and cholecystokinin after TLR4, TLR5 and TLR9 activation. GLP-1 has also been described as an endogenous immune modulator with a wide range of physiological effects, such as stimulation of anti-inflammatory signaling [5]. In two previous studies, we showed that (i) pro-inflammatory molecules located on the surface of Gram-negative bacteria, LPS, are able to enhance GLP-1 plasma levels when injected into mice or humans; and (ii) when the gut barrier is damaged after a gut ischemia/reperfusion experiment, endogenous LPS naturally present in the gut are involved in this rapid GLP-1 secretion, at least partly through the activation of TLR4 receptors [20,22]. Why GLP-1 is released upon gut-barrier dysfunction is currently unknown. Contrary to GLP-2, whose role in gut homeostasis and, in particular, gut permeability is well documented [34], the literature regarding GLP-1 in intestinal homeostasis and, in particular, in regulating the barrier function is scarce [35]. Different pathological situations known to cause a gut-barrier disturbance are associated with enhanced GLP-1 secretion. For instance, in IBDs such as Crohn’s disease or ulcerative colitis, GLP-1 plasma levels are dysregulated [36]. In IBD patients, GLP-1 serum levels are increased [37]. Compared to healthy controls, Crohn’s disease patients have upregulated fasting GLP-1 levels, while stimulated GLP-1 secretion is reduced [38]. Sepsis is another situation in which GLP-1 secretion is enhanced. Previous studies have demonstrated that the endogenous GLP-1 system is activated during sepsis [9] and have associated GLP-1 levels with predicted mortality [39,40]. The CLP model used in this study is a widely used animal model for sepsis [41]. Gut-derived sepsis causes intestinal barrier dysfunction and induces bacterial translocation and the development of multiple-organ-dysfunction syndrome [42]. Whether in this study or in our previous one [22], the results obtained with Tlr4-/- mice suggest that other molecules derived from the gut, and certainly other TLRs, are involved in the GLP-1 secretion that occurs rapidly after gut-barrier alterations. Previous studies have suggested an important role for TLRs in the regulation of gut permeability during infection [43]. Nevertheless, not all TLRs have the same effect on gut permeability. Whereas TLR1 and TLR2 signaling sustain epithelial integrity [44,45], the activation of epithelial TLR4 increases barrier permeability and leaky gut in both human cell lines and mouse models [46,47,48]. However, the presence of functional TLR4 receptors has been shown to protect the intestinal mucosa from damage induced by ischemia/reperfusion treatment in mice [49,50]. Alterations in the intestinal epithelium of mice treated with dextran sodium sulphate (DSS) are associated with a strong increase in intestinal permeability [51], and the subsequent repair capacities of the intestinal mucosa are reduced by the antagonization of the TLR4 receptor. Tlr4-/- mice are also more sensitive to DSS and show greater bacterial translocation than WT mice [52]. Overall, these studies suggest that LPS/TLR4 engagement may provide beneficial effects in terms of gut-barrier integrity. Interestingly, in the experimental conditions described here, we observed the most robust GLP-1 secretion with LPS, the TLR4 agonist. As discussed above, GLP-1 exerts anti-inflammatory properties and could contribute to gut-barrier regulation. Mice with a GLP-1 receptor deficiency are more sensitive to DSS-induced intestinal injuries [12]. One may assume that LPS-induced GLP-1 secretion could contribute to the beneficial effects of TLR4 and, more largely, PAMP-induced GLP-1 secretion. Moreover, TLR2-, TLR5- and TLR9-deficient mice are also more susceptible to chemically induced colitis [45,53,54]. Considering the anti-inflammatory properties of GLP-1, treatments for several disorders are based on the administration of GLP-1 agonists. For instance, in type 2 diabetes patients, semaglutide and liraglutide are associated with a reduction in cardiovascular events [6,55]. In preclinical studies of IBD, GLP-1 agonists have been shown to improve colitis by significantly improving the colon’s weight-to-length ratio, lowering the histopathological score and reducing pro-inflammatory cytokines levels [56,57]. A recent study has shown that chronic pretreatment with liraglutide in mice challenged with intestinal ischemia increased their overall survival [13]. In a human case report, the treatment of IBD (ulcerative colitis) patient with liraglutide led to a full remission of colitis symptoms [58]. In type 2 diabetes patients treated with GLP-1-based therapies, the disease course of IBD is improved compared with patients taking other antidiabetics [59]. Liraglutide treatment of mice undergoing sepsis induction improved the cardiovascular consequences of the dysregulated inflammation observed in sepsis [60]. Sepsis is associated with a hyperglycemic state that needs to be tightly controlled [61]. Indeed, hyperglycemia increases pathogen invasion [62] and, thus, nosocomial mortality [63]. GLP-1 agonists are currently under investigation as a means of controlling hyperglycemia and inflammation during sepsis [64]. Besides the exogenous administration of GLP-1 agonists, the increase of endogenous GLP-1 secretion can also be considered from a therapeutic point of view. For instance, probiotics are known to promote GLP-1 secretion and alleviate both type 2 [65,66] and type 1 [67] diabetes symptoms. Probiotic-related increases in GLP-1 secretion and the associated beneficial effects can be mediated through short-chain fatty acid production [68]. The present study has some limitations. It focuses mainly on TLR activation, and we cannot exclude that other pathways/receptors are involved. Indeed, muramyl dipeptide, a bacterial component, has been shown to improve insulin sensitivity and glucose tolerance in diet-induced obese mice via an increase in GLP-1 secretion by acting through the nucleotide oligomerization domain 2 (NOD2) receptor [69]. Moreover, it is well established (and observed in the present study) that pro-inflammatory cytokines are secreted upon activation of TLRs. Since it has been demonstrated that IL6 and TNFA induce the secretion of GLP-1 [15], we can assume that the observed TLR-induced GLP-1 secretion could be indirectly mediated through cytokine expression. In this study, both the CLP model and the injection of TLR agonists induced cytokine release. Therefore, whether the above-described GLP-1 secretion is a direct result of TLR activation of EECs or an indirect pathway still needs to be determined. It is generally assumed in the literature that TLR ligands are able to induce GLP-1 secretion even though, to date, only the LPS/TLR4 pathway has been demonstrated to be involved in this secretion [20,22]. This study is the first to clearly highlight that the activation of other TLRs also leads to a GLP-1 secretion. ## 4.1. Experimental Animals and Samplings Male wild-type (WT) C57BL/6J mice and male mice deficient in TLR4 (Tlr4-/-) on C57BL/6J background were used. Mice were 8-to-10 weeks of age and were housed in a controlled environment (light from 7 am to 7 pm, constant humidity and temperature) and were given a standard chow diet (A03 diet, Safe SAS, Augy, France). Every sampling procedure was performed under inhaled anesthesia (isoflurane) titrated to maintain spontaneous breathing. Blood was collected by intracardiac puncture in EDTA-coated tubes. Plasmas were separated by centrifugation at 8000× g for 10 min at 4 °C and were stored at −20 °C until use. Ileum tissues were immediately snap-frozen (immersion in liquid nitrogen) after harvest and stored at −80 °C until utilization. ## 4.2. Cecal Ligation Puncture (CLP) Treatment CLP treatment was performed as previously described [70] on mice under inhaled anesthesia (isoflurane) titrated to maintain spontaneous breathing. Anesthetized mice were placed on a heating pad during the surgical procedure. Mice abdomens were shaved and cleaned with alcohol. A midline laparotomy was made, the cecum was isolated and ligated below the ileocecal valve with a 4–0 suture and without causing bowel obstruction. A perforation was made in the ligated cecum by a single puncture with a 21-gauge needle and gently squeezed to externalize a small amount of fecal matter. Cecum was then returned into the peritoneal cavity. The abdominal cavity of the mice was closed by suture and wound clips. Mice were then resuscitated with a subcutaneous injection of 0.4 mL of NaCl. ## 4.3. Drugs Administration in Mice Mice were fasted for 4 h and intraperitoneally injected with different TLRs agonists (10µL/g): Pam3CysSerLys4 (Pam3CSK4; TLR$\frac{1}{2}$ agonist; 1 mg/kg; tlrl-pms, Invivogen, San Diego, CA, USA), lipoteichoic acid (LTA; TLR2 agonist; 1 mg/kg; tlrl-pslta, Invivogen, San Diego, CA, USA), Pam2CysSerLys4 (Pam2CSK4; TLR$\frac{2}{6}$ agonist; 1 mg/kg; tlrl-pm2s, Invivogen, San Diego, CA, USA), polyinosinic-polycytidylic acid (Poly (I:C); TLR3 agonist; 1 mg/kg, tlrl-picw, Invivogen, San Diego, CA, USA), lipopolysaccharides (LPS; TLR4 agonist; 1 mg/kg; tlrl-pb5lps, Invivogen, San Diego, CA, USA), flagellin (TLR5 agonist, 1 mg/kg; tlrl-bsfla, Invivogen, San Diego, CA, USA), resiquimod (R848; TLR$\frac{7}{8}$ agonist; 1 mg/kg; tlrl-r848, Invivogen, San Diego, CA, USA) and CpG oligodeoxynucleotides (CpG ODN; TLR9 agonist; 0.6 mg/kg; tlrl-dls01, Invivogen, San Diego, CA, USA). ## 4.4. Real-Time Quantitative PCR Ileum was immediately snap-frozen (immersion in liquid nitrogen) after harvest and stored at −80 °C until RNA extraction. Total RNA was isolated using an RNA extraction kit (RNeasy Mini Kit, 74106, Qiagen, Hilden, Germany). RNA extraction included a DNAse treatment step. RNA was quantified using the NanoDrop 1000 spectrophotometer (ThermoFisher Scientific, Waltham, MA, USA), and 500 ng of RNA from each sample was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Multiscribe® reverse transcriptase, 4368813, Applied Biosystems, Waltham, MA, USA) according to the manufacturer’s instructions. Quantitative PCRs were performed using StepOnePlus (Real-Time PCR System, Applied Biosystems, Waltham, MA, USA) and TaqMan® technology (4324018, Applied Biosystems, Waltham, MA, USA). The mRNA level was normalized to levels of Rplp0 mRNA, and the results were expressed as relative expression levels, using the 2-ΔΔCt method. ## 4.5. Plasma and Tissues Biochemical Analyses Cytokines and total GLP-1 concentrations were determined by commercially available ELISA Kits (M-CYTOMAG-70K, Millipore, Burlington, MA, USA; EZGLP1T-36K, Millipore, Burlington, MA, USA). Ileum segments were homogenized in RIPA buffer (89901, ThermoFisher Scientific, Waltham, MA, USA) containing antiproteases and antiphosphatases (11873580001, Roche, Penzberg, Germany). Total protein concentration was measured with a commercial colorimetric assay kit (Thermo Scientific™ Pierce™ BCA assay kit, Waltham, MA, USA). ## 4.6. Statistics Data were collected using Microsoft Excel for Office 365. Data are presented as means ± SEM. Statistical analyses were performed using Prism 6.0 (GraphPad, San Diego, CA, USA). To decide whether to use parametric or non-parametric statistics, the normality of distributions were assessed with the Shapiro–Wilk test (under $$n = 7$$, distributions were considered to be non-normal). The statistical significance of differences between two groups was evaluated with the Mann–Whitney U test or Student’s t-test (a statistical correction was applied when variances were different between groups). For more than two groups, a Kruskal–Wallis test was performed. A value of $p \leq 0.05$ was considered statistically significant (NS, not significant; * $p \leq 0.05$, $p \leq 0.01$, * $p \leq 0.001$ and ** $p \leq 0.0001$). ## 5. Conclusions Overall, we report that activation of TLRs either through a polymicrobial infection model (CLP) or with specific agonists leads to increases in GLP-1 secretion. We confirm the secretory effects of LPS, and we extend this secretory capacity to others PAMPs through the activation of other TLRs than TLR4. The multiple-TLRs-induced GLP-1 secretion observed in this study supports a tiny link between GLP-1 and gut inflammatory processes. Even if the secretion of GLP-1 has increased the association to inflammatory processes such as IBD or sepsis, its actual systemic and gut functions in this context need more investigations. ## References 1. Jandhyala S.M., Talukdar R., Subramanyam C., Vuyyuru H., Sasikala M., Reddy D.N.. Role of the Normal Gut Microbiota*. World J. Gastroenterol.* (2015) 21 8787-8803. DOI: 10.3748/wjg.v21.i29.8787 2. Mörbe U.M., Jørgensen P.B., Fenton T.M., von Burg N., Riis L.B., Spencer J., Agace W.W.. Human Gut-Associated Lymphoid Tissues (GALT); Diversity, Structure, and Function. Mucosal Immunol. (2021) 14 793-802. DOI: 10.1038/s41385-021-00389-4 3. Abreu M.T., Fukata M., Arditi M.. TLR Signaling in the Gut in Health and Disease. J. Immunol. (2005) 174 4453-4460. DOI: 10.4049/jimmunol.174.8.4453 4. Drucker D.J., Habener J.F., Holst J.J.. Discovery, Characterization, and Clinical Development of the Glucagon-like Peptides. J. Clin. Investig. (2017) 127 4217-4227. DOI: 10.1172/JCI97233 5. Rowlands J., Heng J., Newsholme P., Carlessi R.. Pleiotropic Effects of GLP-1 and Analogs on Cell Signaling, Metabolism, and Function. Front. Endocrinol. (Lausanne) (2018) 9 672. DOI: 10.3389/fendo.2018.00672 6. Iorga R.A., Bacalbasa N., Carsote M., Bratu O.G., Stanescu A.M.A., Bungau S., Pantis C., Diaconu C.C.. Metabolic and Cardiovascular Benefits of GLP-1 Agonists, besides the Hypoglycemic Effect (Review). Exp. Ther. Med. (2020) 20 2396-2400. DOI: 10.3892/etm.2020.8714 7. Vandemark C., Nguyen J., Zhao Z.-Q.. Cardiovascular Protection with a Long-Acting GLP-1 Receptor Agonist Liraglutide: An Experimental Update. Molecules (2023) 28. DOI: 10.3390/molecules28031369 8. Merza N., Akram M., Mengal A., Mustafa Rashid A., Mahboob A., Faryad M., Fatima Z., Ahmed M., Ansari S.A.. The Safety and Efficacy of GLP-1 Receptor Agonists in Heart Failure Patients: A Systematic Review and Meta-Analysis. Curr. Probl. Cardiol. (2023) 48 101602. DOI: 10.1016/j.cpcardiol.2023.101602 9. Perl S.H., Bloch O., Zelnic-Yuval D., Love I., Mendel-Cohen L., Flor H., Rapoport M.J.. Sepsis-Induced Activation of Endogenous GLP-1 System Is Enhanced in Type 2 Diabetes. Diabetes/Metab. Res. Rev. (2018) 34 e2982. DOI: 10.1002/dmrr.2982 10. Braun J.-P., Buhner S., Kastrup M., Dietz E., Langer K., Dohmen P., Lochs H., Spies C.. Barrier Function of the Gut and Multiple Organ Dysfunction after Cardiac Surgery. J. Int. Med. Res. (2007) 35 72-83. DOI: 10.1177/147323000703500107 11. Nguyen M., Tavernier A., Gautier T., Aho S., Morgant M.C., Bouhemad B., Guinot P.-G., Grober J.. Glucagon-like Peptide-1 Is Associated with Poor Clinical Outcome, Lipopolysaccharide Translocation and Inflammation in Patients Undergoing Cardiac Surgery with Cardiopulmonary Bypass. Cytokine (2020) 133 155182. DOI: 10.1016/j.cyto.2020.155182 12. Yusta B., Baggio L.L., Koehler J., Holland D., Cao X., Pinnell L.J., Johnson-Henry K.C., Yeung W., Surette M.G., Bang K.W.A.. GLP-1R Agonists Modulate Enteric Immune Responses Through the Intestinal Intraepithelial Lymphocyte GLP-1R. Diabetes (2015) 64 2537-2549. DOI: 10.2337/db14-1577 13. Zou Z., Wang Z.. Liraglutide Attenuates Intestinal Ischemia/Reperfusion Injury via NF-ΚB and PI3K/Akt Pathways in Mice. Life Sci. (2022) 309 121045. DOI: 10.1016/j.lfs.2022.121045 14. Nozu T., Miyagishi S., Kumei S., Nozu R., Takakusaki K., Okumura T.. Glucagon-like Peptide-1 Analog, Liraglutide, Improves Visceral Sensation and Gut Permeability in Rats. J. Gastroenterol. Hepatol. (2018) 33 232-239. DOI: 10.1111/jgh.13808 15. Brubaker P.L.. The Molecular Determinants of Glucagon-like Peptide Secretion by the Intestinal L Cell. Endocrinology (2022) 163 bqac159. DOI: 10.1210/endocr/bqac159 16. Tolhurst G., Heffron H., Lam Y.S., Parker H.E., Habib A.M., Diakogiannaki E., Cameron J., Grosse J., Reimann F., Gribble F.M.. Short-Chain Fatty Acids Stimulate Glucagon-like Peptide-1 Secretion via the G-Protein-Coupled Receptor FFAR2. Diabetes (2012) 61 364-371. DOI: 10.2337/db11-1019 17. Parker H.E., Wallis K., le Roux C.W., Wong K.Y., Reimann F., Gribble F.M.. Molecular Mechanisms Underlying Bile Acid-Stimulated Glucagon-like Peptide-1 Secretion. Br. J. Pharm. (2012) 165 414-423. DOI: 10.1111/j.1476-5381.2011.01561.x 18. Cani P.D., Knauf C.. A Newly Identified Protein from Akkermansia Muciniphila Stimulates GLP-1 Secretion. Cell Metab. (2021) 33 1073-1075. DOI: 10.1016/j.cmet.2021.05.004 19. Tomaro-Duchesneau C., LeValley S.L., Roeth D., Sun L., Horrigan F.T., Kalkum M., Hyser J.M., Britton R.A.. Discovery of a Bacterial Peptide as a Modulator of GLP-1 and Metabolic Disease. Sci. Rep. (2020) 10 4922. DOI: 10.1038/s41598-020-61112-0 20. Nguyen A.T., Mandard S., Dray C., Deckert V., Valet P., Besnard P., Drucker D.J., Lagrost L., Grober J.. Lipopolysaccharides-Mediated Increase in Glucose-Stimulated Insulin Secretion: Involvement of the Glucagon-like Peptide 1 (GLP1) Pathway. Diabetes (2014) 63 471. DOI: 10.2337/db13-0903 21. Kahles F., Meyer C., Möllmann J., Diebold S., Findeisen H.M., Lebherz C., Trautwein C., Koch A., Tacke F., Marx N.. GLP-1 Secretion Is Increased by Inflammatory Stimuli in an IL-6–Dependent Manner, Leading to Hyperinsulinemia and Blood Glucose Lowering. Diabetes (2014) 63 3221-3229. DOI: 10.2337/db14-0100 22. Lebrun L.J., Lenaerts K., Kiers D., Pais de Barros J.-P., Le Guern N., Plesnik J., Thomas C., Bourgeois T., Dejong C.H.C., Kox M.. Enteroendocrine L Cells Sense LPS after Gut Barrier Injury to Enhance GLP-1 Secretion. Cell Rep. (2017) 21 1160-1168. DOI: 10.1016/j.celrep.2017.10.008 23. Worthington J.J., Reimann F., Gribble F.M.. Enteroendocrine Cells-Sensory Sentinels of the Intestinal Environment and Orchestrators of Mucosal Immunity. Mucosal Immunol. (2018) 11 3-20. DOI: 10.1038/mi.2017.73 24. Aravanis C.V., Kapelouzou A., Vagios S., Tsilimigras D.I., Katsimpoulas M., Moris D., Demesticha T.D., Schizas D., Kostakis A., Machairas A.. Toll-Like Receptors -2, -3, -4 and -7 Expression Patterns in the Liver of a CLP-Induced Sepsis Mouse Model. J. Investig. Surg (2020) 33 109-117. DOI: 10.1080/08941939.2018.1476630 25. Krivan S., Kapelouzou A., Vagios S., Tsilimigras D.I., Katsimpoulas M., Moris D., Aravanis C.V., Demesticha T.D., Schizas D., Mavroidis M.. Increased Expression of Toll-like Receptors 2, 3, 4 and 7 MRNA in the Kidney and Intestine of a Septic Mouse Model. Sci. Rep. (2019) 9 4010. DOI: 10.1038/s41598-019-40537-2 26. Bakopoulos A., Kapelouzou A., Tsilimigras D.I., Katsimpoulas M., Schizas D., Aravanis C., Balafas E., Mavroidis M., Pavlakis K., Machairas A.. Expression of Toll-like Receptors (TLRs) in the Lungs of an Experimental Sepsis Mouse Model. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0188050 27. Bogunovic M., Davé S.H., Tilstra J.S., Chang D.T.W., Harpaz N., Xiong H., Mayer L.F., Plevy S.E.. Enteroendocrine Cells Express Functional Toll-like Receptors. Am. J. Physiol. Gastrointest. Liver Physiol. (2007) 292 G1770-G1783. DOI: 10.1152/ajpgi.00249.2006 28. Gribble F.M., Reimann F.. Function and Mechanisms of Enteroendocrine Cells and Gut Hormones in Metabolism. Nat. Rev. Endocrinol. (2019) 15 226-237. DOI: 10.1038/s41574-019-0168-8 29. Harrison E., Lal S., McLaughlin J.T.. Enteroendocrine Cells in Gastrointestinal Pathophysiology. Curr. Opin. Pharmacol. (2013) 13 941-945. DOI: 10.1016/j.coph.2013.09.012 30. Genton L., Kudsk K.A.. Interactions between the Enteric Nervous System and the Immune System: Role of Neuropeptides and Nutrition. Am. J. Surg. (2003) 186 253-258. DOI: 10.1016/S0002-9610(03)00210-1 31. Lee J.H., Patel K., Tae H.J., Lustig A., Kim J.W., Mattson M.P., Taub D.D.. Ghrelin Augments Murine T-Cell Proliferation by Activation of the Phosphatidylinositol-3-Kinase, Extracellular Signal-Regulated Kinase and Protein Kinase C Signaling Pathways. FEBS Lett. (2014) 588 4708-4719. DOI: 10.1016/j.febslet.2014.10.044 32. Reardon C., Duncan G.S., Brüstle A., Brenner D., Tusche M.W., Olofsson P.S., Rosas-Ballina M., Tracey K.J., Mak T.W.. Lymphocyte-Derived ACh Regulates Local Innate but Not Adaptive Immunity. Proc. Natl. Acad. Sci. USA (2013) 110 1410-1415. DOI: 10.1073/pnas.1221655110 33. Palazzo M., Balsari A., Rossini A., Selleri S., Calcaterra C., Gariboldi S., Zanobbio L., Arnaboldi F., Shirai Y.F., Serrao G.. Activation of Enteroendocrine Cells via TLRs Induces Hormone, Chemokine, and Defensin Secretion. J. Immunol. (2007) 178 4296-4303. DOI: 10.4049/jimmunol.178.7.4296 34. Drucker D.J., Yusta B.. Physiology and Pharmacology of the Enteroendocrine Hormone Glucagon-like Peptide-2. Annu. Rev. Physiol (2014) 76 561-583. DOI: 10.1146/annurev-physiol-021113-170317 35. Drucker D.J.. Mechanisms of Action and Therapeutic Application of Glucagon-like Peptide-1. Cell Metab. (2018) 27 740-756. DOI: 10.1016/j.cmet.2018.03.001 36. Hunt J.E., Holst J.J., Jeppesen P.B., Kissow H.. GLP-1 and Intestinal Diseases. Biomedicines (2021) 9. DOI: 10.3390/biomedicines9040383 37. Keller J., Binnewies U., Rösch M., Juul Holst J., Beglinger C., Andresen V., Layer P.. Gastric Emptying and Disease Activity in Inflammatory Bowel Disease. Eur. J. Clin. Investig. (2015) 45 1234-1242. DOI: 10.1111/eci.12542 38. Lucotti P., Lovati E., Lenti M.V., Valvo B., Sprio E., Aronico N., Giuffrida P., Dell’Aera D., Pasini A., Ubezio C.. Abnormal Post-Prandial Glucagon-like Peptide Release in Patients with Crohn’s Disease. Clin. Res. Hepatol. Gastroenterol. (2021) 45 101533. DOI: 10.1016/j.clinre.2020.08.011 39. Brakenridge S.C., Moore F.A., Mercier N.R., Cox M., Wu Q., Moldawer L.L., Mohr A.M., Efron P.A., Smith R.S.. Persistently Elevated Glucagon-Like Peptide 1 Levels Among Critically-Ill Surgical Patients After Sepsis and Development of Chronic Critical Illness and Dismal Long-Term Outcomes. J. Am. Coll Surg. (2019) 229 58-67.e1. DOI: 10.1016/j.jamcollsurg.2019.04.014 40. Lebherz C., Schlieper G., Möllmann J., Kahles F., Schwarz M., Brünsing J., Dimkovic N., Koch A., Trautwein C., Flöge J.. GLP-1 Levels Predict Mortality in Patients with Critical Illness as Well as End-Stage Renal Disease. Am. J. Med. (2017) 130 833-841.e3. DOI: 10.1016/j.amjmed.2017.03.010 41. Zanotti-Cavazzoni S.L., Guglielmi M., Parrillo J.E., Walker T., Dellinger R.P., Hollenberg S.M.. Fluid Resuscitation Influences Cardiovascular Performance and Mortality in a Murine Model of Sepsis. Intensive Care Med. (2009) 35 748-754. DOI: 10.1007/s00134-008-1360-9 42. Assimakopoulos S.F., Triantos C., Thomopoulos K., Fligou F., Maroulis I., Marangos M., Gogos C.A.. Gut-Origin Sepsis in the Critically Ill Patient: Pathophysiology and Treatment. Infection (2018) 46 751-760. DOI: 10.1007/s15010-018-1178-5 43. Gibson D.L., Ma C., Rosenberger C.M., Bergstrom K.S.B., Valdez Y., Huang J.T., Khan M.A., Vallance B.A.. Toll-like Receptor 2 Plays a Critical Role in Maintaining Mucosal Integrity during Citrobacter Rodentium-Induced Colitis. Cell Microbiol. (2008) 10 388-403. DOI: 10.1111/j.1462-5822.2007.01052.x 44. Kamdar K., Johnson A.M.F., Chac D., Myers K., Kulur V., Truevillian K., DePaolo R.W.. Innate Recognition of the Microbiota by Toll-like Receptor-1 Promotes Epithelial Homeostasis and Prevents Chronic Inflammation. J. Immunol. (2018) 201 230-242. DOI: 10.4049/jimmunol.1701216 45. Cario E., Gerken G., Podolsky D.K.. Toll-like Receptor 2 Controls Mucosal Inflammation by Regulating Epithelial Barrier Function. Gastroenterology (2007) 132 1359-1374. DOI: 10.1053/j.gastro.2007.02.056 46. Dheer R., Santaolalla R., Davies J.M., Lang J.K., Phillips M.C., Pastorini C., Vazquez-Pertejo M.T., Abreu M.T.. Intestinal Epithelial Toll-Like Receptor 4 Signaling Affects Epithelial Function and Colonic Microbiota and Promotes a Risk for Transmissible Colitis. Infect. Immun. (2016) 84 798-810. DOI: 10.1128/IAI.01374-15 47. Guo S., Al-Sadi R., Said H.M., Ma T.Y.. Lipopolysaccharide Causes an Increase in Intestinal Tight Junction Permeability in Vitro and in Vivo by Inducing Enterocyte Membrane Expression and Localization of TLR-4 and CD14. Am. J. Pathol. (2013) 182 375-387. DOI: 10.1016/j.ajpath.2012.10.014 48. Nighot M., Al-Sadi R., Guo S., Rawat M., Nighot P., Watterson M.D., Ma T.Y.. Lipopolysaccharide-Induced Increase in Intestinal Epithelial Tight Permeability Is Mediated by Toll-Like Receptor 4/Myeloid Differentiation Primary Response 88 (MyD88) Activation of Myosin Light Chain Kinase Expression. Am. J. Pathol. (2017) 187 2698-2710. DOI: 10.1016/j.ajpath.2017.08.005 49. Tatum P.M., Harmon C.M., Lorenz R.G., Dimmitt R.A.. Toll-like Receptor 4 Is Protective against Neonatal Murine Ischemia-Reperfusion Intestinal Injury. J. Pediatr. Surg. (2010) 45 1246-1255. DOI: 10.1016/j.jpedsurg.2010.02.093 50. Chen L.-W., Chang W.-J., Chen P.-H., Liu W.-C., Hsu C.-M.. TLR Ligand Decreases Mesenteric Ischemia and Reperfusion Injury-Induced Gut Damage through TNF-α Signaling. Shock (2008) 30 563. DOI: 10.1097/SHK.0b013e31816a3458 51. Kitajima S., Takuma S., Morimoto M.. Changes in Colonic Mucosal Permeability in Mouse Colitis Induced with Dextran Sulfate Sodium. Exp. Anim. (1999) 48 137-143. DOI: 10.1538/expanim.48.137 52. Fukata M., Michelsen K.S., Eri R., Thomas L.S., Hu B., Lukasek K., Nast C.C., Lechago J., Xu R., Naiki Y.. Toll-like Receptor-4 Is Required for Intestinal Response to Epithelial Injury and Limiting Bacterial Translocation in a Murine Model of Acute Colitis. Am. J. Physiol.-Gastrointest. Liver Physiol. (2005) 288 G1055-G1065. DOI: 10.1152/ajpgi.00328.2004 53. Rose W.A., Sakamoto K., Leifer C.A.. TLR9 Is Important for Protection against Intestinal Damage and for Intestinal Repair. Sci. Rep. (2012) 2 574. DOI: 10.1038/srep00574 54. Chassaing B., Ley R.E., Gewirtz A.T.. Intestinal Epithelial Cell Toll-like Receptor 5 Regulates the Intestinal Microbiota to Prevent Low-Grade Inflammation and Metabolic Syndrome in Mice. Gastroenterology (2014) 147 1363-1377.e17. DOI: 10.1053/j.gastro.2014.08.033 55. Del Olmo-Garcia M.I., Merino-Torres J.F.. GLP-1 Receptor Agonists and Cardiovascular Disease in Patients with Type 2 Diabetes. J. Diabetes Res. (2018) 2018 4020492. DOI: 10.1155/2018/4020492 56. Bang-Berthelsen C.H., Holm T.L., Pyke C., Simonsen L., Søkilde R., Pociot F., Heller R.S., Folkersen L., Kvist P.H., Jackerott M.. GLP-1 Induces Barrier Protective Expression in Brunner’s Glands and Regulates Colonic Inflammation. Inflamm. Bowel Dis. (2016) 22 2078-2097. DOI: 10.1097/MIB.0000000000000847 57. Zatorski H., Sałaga M., Fichna J.. Role of Glucagon-like Peptides in Inflammatory Bowel Diseases—Current Knowledge and Future Perspectives. Naunyn-Schmiedeberg’s Arch. Pharm. (2019) 392 1321-1330. DOI: 10.1007/s00210-019-01698-z 58. Lourie J.. A Novel Use of Liraglutide: Induction of Partial Remission in Ulcerative Colitis and Ankylosing Spondylitis. Clin Med Rev Case Rep. (2019) 6 281. DOI: 10.23937/2378-3656/1410281 59. Villumsen M., Schelde A.B., Jimenez-Solem E., Jess T., Allin K.H.. GLP-1 Based Therapies and Disease Course of Inflammatory Bowel Disease. eClinicalMedicine (2021) 37 100979. DOI: 10.1016/j.eclinm.2021.100979 60. Helmstädter J., Keppeler K., Aust F., Küster L., Frenis K., Filippou K., Vujacic-Mirski K., Tsohataridis S., Kalinovic S., Kröller-Schön S.. GLP-1 Analog Liraglutide Improves Vascular Function in Polymicrobial Sepsis by Reduction of Oxidative Stress and Inflammation. Antioxidants (2021) 10. DOI: 10.3390/antiox10081175 61. Hirasawa H., Oda S., Nakamura M.. Blood Glucose Control in Patients with Severe Sepsis and Septic Shock. World J. Gastroenterol. (2009) 15 4132-4136. DOI: 10.3748/wjg.15.4132 62. Jafar N., Edriss H., Nugent K.. The Effect of Short-Term Hyperglycemia on the Innate Immune System. Am. J. Med. Sci. (2016) 351 201-211. DOI: 10.1016/j.amjms.2015.11.011 63. Leonidou L., Mouzaki A., Michalaki M., DeLastic A.L., Kyriazopoulou V., Bassaris H.P., Gogos C.A.. Cytokine Production and Hospital Mortality in Patients with Sepsis-Induced Stress Hyperglycemia. J. Infect. (2007) 55 340-346. DOI: 10.1016/j.jinf.2007.05.177 64. Yang F., Zeng F., Luo X., Lei Y., Li J., Lu S., Huang X., Lan Y., Liu R.. GLP-1 Receptor: A New Target for Sepsis. Front. Pharm. (2021) 12 706908. DOI: 10.3389/fphar.2021.706908 65. Xiao R., Wang R., Li S., Kang X., Ren Y., Sun E., Wang C., He J., Zhan J.. Preliminary Evaluation of Potential Properties of Three Probiotics and Their Combination with Prebiotics on GLP-1 Secretion and Type 2 Diabetes Alleviation. J. Food Qual. (2022) 2022 e8586843. DOI: 10.1155/2022/8586843 66. Pegah A., Abbasi-Oshaghi E., Khodadadi I., Mirzaei F., Tayebinai H.. Probiotic and Resveratrol Normalize GLP-1 Levels and Oxidative Stress in the Intestine of Diabetic Rats. Metab. Open (2021) 10 100093. DOI: 10.1016/j.metop.2021.100093 67. Wei S.-H., Chen Y.-P., Chen M.-J.. Selecting Probiotics with the Abilities of Enhancing GLP-1 to Mitigate the Progression of Type 1 Diabetes in Vitro and in Vivo. J. Funct. Foods (2015) 18 473-486. DOI: 10.1016/j.jff.2015.08.016 68. Yadav H., Lee J.-H., Lloyd J., Walter P., Rane S.G.. Beneficial Metabolic Effects of a Probiotic via Butyrate-Induced GLP-1 Hormone Secretion. J. Biol. Chem. (2013) 288 25088-25097. DOI: 10.1074/jbc.M113.452516 69. Williams L., Alshehri A., Robichaud B., Cudmore A., Gagnon J.. The Role of the Bacterial Muramyl Dipeptide in the Regulation of GLP-1 and Glycemia. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21155252 70. Klein A., Deckert V., Schneider M., Dutrillaux F., Hammann A., Athias A., Le Guern N., Pais de Barros J.-P., Desrumaux C., Masson D.. Alpha-Tocopherol Modulates Phosphatidylserine Externalization in Erythrocytes: Relevance in Phospholipid Transfer Protein-Deficient Mice. Arter. Thromb. Vasc. Biol. (2006) 26 2160-2167. DOI: 10.1161/01.ATV.0000235699.98024.11
--- title: FTIR Analysis of Renal Tissue for the Assessment of Hypertensive Organ Damage and proANP31–67 Treatment authors: - Leonardo Pioppi - Niki Tombolesi - Reza Parvan - Gustavo Jose Justo da Silva - Raffaele Altara - Marco Paolantoni - Assunta Morresi - Paola Sassi - Alessandro Cataliotti journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049716 doi: 10.3390/ijms24065196 license: CC BY 4.0 --- # FTIR Analysis of Renal Tissue for the Assessment of Hypertensive Organ Damage and proANP31–67 Treatment ## Abstract The kidneys are one of the main end organs targeted by hypertensive disease. Although the central role of the kidneys in the regulation of high blood pressure has been long recognized, the detailed mechanisms behind the pathophysiology of renal damage in hypertension remain a matter of investigation. Early renal biochemical alterations due to salt-induced hypertension in Dahl/salt-sensitive rats were monitored by Fourier-Transform Infrared (FTIR) micro-imaging. Furthermore, FTIR was used to investigate the effects of proANP31–67, a linear fragment of pro-atrial natriuretic peptide, on the renal tissue of hypertensive rats. Different hypertension-induced alterations were detected in the renal parenchyma and blood vessels by the combination of FTIR imaging and principal component analysis on specific spectral regions. Changes in amino acids and protein contents observed in renal blood vessels were independent of altered lipid, carbohydrate, and glycoprotein contents in the renal parenchyma. FTIR micro-imaging was found to be a reliable tool for monitoring the remarkable heterogeneity of kidney tissue and its hypertension-induced alterations. In addition, FTIR detected a significant reduction in these hypertension-induced alterations in the kidneys of proANP31–67-treated rats, further indicating the high sensitivity of this cutting-edge imaging modality and the beneficial effects of this novel medication on the kidneys. ## 1. Introduction Chronic kidney disease (CKD) continues to be a major health concern and a leading cause of death, with high blood pressure being a primary cause of renal failure [1,2]. CKD is characterized by tubular atrophy, vasculopathy, glomerulosclerosis, increased renal interstitial fibrosis, and reduced capacity for renal regeneration [3,4,5]. Renal damage and hypertension are known independent predictors of morbidity and mortality [6]. To date, there are no effective treatments to prevent the worsening of renal impairment, as few medications have been proven to have protective effects on the renal parenchyma. It has been suggested that the lack of effective therapies is, at least in part, due to the late diagnosis of CKD and that earlier detection of renal impairment could increase the efficacy of an intervention [7]. The histopathological analysis of kidney biopsies is the gold standard for an accurate diagnosis of renal damage [8,9]. Most of the late parenchymal lesions can be identified by microscopy [10]. However, the study of early renal biochemical alterations induced by hypertension is still an area of investigation. Fourier-transform infrared (FTIR) microspectroscopy allows for fast and label-free biochemical imaging of many biological tissues [11], and micro-FTIR analyses of the kidneys have accurately detected early biochemical alterations related to cancerous tissue [12,13] and other nephropathies [14,15]. Thus, this technique can be used to detect small changes in biochemical composition. However, the application of FTIR for the assessment of hypertension-related kidney lesions has not been reported. We recently demonstrated that vibrational spectroscopic techniques are efficient diagnostic tools that are capable of detecting early chemical changes that precede the structural and functional cardiac alterations induced by prolonged hypertension [16]. In the current study, we applied FTIR imaging with hierarchical cluster analysis (HCA) and principal component analysis (PCA) to assess, at the molecular level, the early structural changes in the kidney that are induced by chronic hypertension. To achieve this aim, we used Dahl/salt-sensitive (DSS) rats and monitored the kidney damage at six weeks of salt-induced hypertension. In the present work, FTIR analysis was also applied to investigate the effect of proANP31–67, a linear fragment of pro-atrial natriuretic peptide [17]. This drug was recently developed to enhance renal function [18]; however, its effects on renal tissue have not been fully characterized. Here, we investigated the actions of proANP31–67 on renal biochemistry and the sensitivity of FTIR for detecting such effects in DSS hypertensive rats. ## 2.1. Renal Functional and Structural Assessment Six weeks of a high-salt diet compared to a normal diet resulted in more pronounced diuresis, maintained glomerular filtration rate (GFR), higher sodium excretion, and tended to increase albuminuria (Figure 1a–d). The high-salt diet resulted in an increase in blood pressure, which was not reduced by proANP31–67. On histological analysis, the high-salt diet tended to increase fibrosis and perivascular collagen deposition (Figure 1e,f). ## 2.2. Spectroscopic Characterization of Renal Tissue Unstained histological slides were examined by conventional microscopy for the identification of kidney structure and compartments (Figure 2a). Hierarchical cluster analysis (HCA) of FTIR spectra revealed a clear segmentation of such compartments, depending on the spectral range used for the distance measurement: if a range between 900 cm−1 and 1400 cm−1 was used, clusters showed a different composition for blood, vessel walls (internal and external), and glomeruli/tubules (Figure 2b,c). The 900–1400 cm−1 range of IR absorption spectrum is characterized by the presence of signals characteristic of carbohydrates (1000–1150 cm−1), lipids (1150–1200 and 1330–1340 cm−1), proteins (1240–1300 cm−1), and free amino acids (1396 cm−1) [11,19]. To identify the mean composition of the different areas, we visually inspected the second derivative spectra that were obtained from classes segmented by HCA (Figure 1d). These spectral profiles showed that blood is characterized by a high content of free amino acids and lipids; internal vessel walls by a high content of lipids and proteins; external walls by the highest content of proteins; glomeruli and tubules by the highest content of glycogen. Based on optical images and supported by the results of HCA, we selected spectra from regions identified as glomeruli, tubules, and vessel walls in the kidneys of NT, HT, and proANP31–67-treated rats. For the analysis of vessels, only the results from NT and HT tissues were compared as the patency of the proANP31–67-treated vessels did not allow for unambiguous interpretation. For a range between 1000 cm−1 and 1800 cm−1, principal component analysis (PCA) showed that the first three principal components (i.e., PC1, PC2, and PC3) explain more than $80\%$ of the total variance (Figure 3a). For each of these, the most intense contributions were highlighted in the respective loading plots (Figure 3b). Regarding PC1, the most relevant regions are 1430–1480 cm−1 and 1620–1740 cm−1, which are assigned to both lipids (signals at 1430–1480 cm−1 and 1700–1740 cm−1) and proteins (Amide I band at 1620–1700 cm−1). The loadings of PC2 showed intense features in a region between 1450 cm−1 and 1700 cm−1 with a larger contribution from the protein signals (Amide II and Amide I bands at 1485–1700 cm−1); on the other hand, loadings of PC3 showed several distinct spectral features in a range between 1000 cm−1 and 1150 cm−1, which may be attributed to carbohydrates and glycoproteins. We found that a combination of PC1 and PC2 describes the major differences seen between the blood vessel spectra of NT and HT samples (Figure 3c). In particular, HT samples showed a shift in the positive direction of the PC2 score axis, suggesting that important changes in the protein content characterize vessel composition in HT compared to NT. Both PC2 and PC3 differentiate the positive scores of NT and proANP31–67-treated samples from the negative ones of HT in the spectral profiles of tubules and glomeruli (Figure 3d,e). The same marker bands were observed to differentiate the mean spectral profiles of HT from those of NH and pANP (Figure S1). The PCA of spectra from single anatomic groups also showed similar results (Figure S2), confirming the common features of glomeruli and tubules. Thus, for the latter, we selected a range between 1000 cm−1 and 1200 cm−1 to perform the analysis (Figure S3a) and observed that more than $60\%$ of the total variance is explained by a component (PC1) with an intense feature at 1030 cm−1 (Figure S3b) whose scores are negative for NT and proANP31–67-treated, and positive for HT (both tubules and glomeruli) data. Since the spectral markers of HT are located on specific bands of the FTIR spectrum (protein bands for blood vessels and glycoprotein/carbohydrate bands for glomeruli and tubules), we performed a univariate analysis of FTIR spectra by selecting specific marker bands to monitor the lipid (2840–2860 cm−1), protein (1485–1700 cm−1), glycoprotein/carbohydrate (1000–1128 cm−1), and free amino acid (1358–1416 cm−1) contents. In Figure 4, an example of FTIR images representing the ratio of integrated intensities of lipids to glycoprotein–carbohydrate (L/CG) bands is shown. The upper panels of Figure 4 shows the L/CG intensity ratio in tubules from NT, proANP31–67-treated, HT, and kidneys; the lower panels show the glomeruli of the same samples. All the panels were normalized to the same range of intensity ratio (scale on the right), which allows for a comparison of the composition of the different tissues with respect to the L/CG content. Both tubules and glomeruli of HT samples were associated with a reduced L/CG ratio compared to NT samples, whereas proANP31–67-treated samples showed values that were intermediate between NT and HT samples. The intensity ratios (L/CG, P/FAA, L/FAA) between the different marker bands for internal and external regions of blood vessels, glomeruli, and tubules are displayed in Figure 5, as these structures have a different composition according to HCA and/or PCA. In HT kidneys, the L/CG was reduced in both the external and internal regions of blood vessels (BV_Ext and BV_Int, respectively) when compared to NT kidneys ($p \leq 0.01$). The P/FAA was reduced in the BV_Ext ($p \leq 0.01$) and tended to increase in the BV_Int of HT kidneys compared to NT ($p \leq 0.05$). The L/FAA in HT kidneys was markedly increased, both in the BV_Ext and BV_Int, compared to NT kidneys ($p \leq 0.01$). In the glomeruli, L/CG, P/FAA, and L/FAA were lower in HT and were restored with proANP31–67 compared to NT ($p \leq 0.01$). In the tubules, both L/CG and P/FAA tended to be lower in HT compared to NT and restored with proANP31–67 ($p \leq 0.01$), while L/FAA increased in HT compared to NT and was reduced with proANP31–67 ($p \leq 0.05$). A negative correlation was observed between the glomerular P/FAA ratio and the extent of kidney fibrosis (Figure 5d and Table S1) and no correlation was observed between the other renal parameters (i.e., GFR, 24-h urine excretion, urinary-Na excretion, and urinary-albumin excretion) with any of the intensity ratios (Table S1). ## 3. Discussion In the current study, we investigated kidney structural changes in a rat model of hypertension using FTIR imaging. Using HCA, FTIR was able to distinguish the spectral features of primary renal structures, such as blood vessels, tubules, and glomeruli. The capability to distinguish these histological structures based on their spectral markers allows for structure-specific analysis. Although renal functional parameters (i.e., GFR, 24-h urine excretion, and urinary-Na excretion) remained within the normal range after six weeks of sustained hypertension, the renal parenchyma tended toward remodeling, and FTIR analysis revealed biochemical changes induced by hypertension on blood vessels, as well as changes in glomeruli and tubules. Specifically, changes in amino acids and protein content were observed in blood vessels, changes in lipid and carbohydrate/glycoprotein concentrations were detected in glomeruli, and changes in carbohydrate/glycoprotein content were observed in tubules. These changes, although associated with renal structural alterations, were not associated with the changes in renal function as indicated by GFR. Indeed, GFR was similar in the untreated HT and NT rats; thus, these alterations may constitute an early FTIR marker profile of hypertension-induced organ damage. Of note, treatment with proANP31–67 tended to improve renal structure compared to HT-untreated rats (Figure 1e,f). Importantly, FTIR detected these changes associated with proANP31–67-treatment and could differentiate between treated versus untreated HT kidneys (Figure 2, Figure 3, Figure 4 and Figure 5), even if the treatment did not reduce blood pressure. Fibrosis and vessel wall thickening are common histological alterations encountered as a response to hypertension [20]. In the PCA of our spectroscopic data, more protein and free amino acids were detected in HT vessel walls. These results may be explained by the progression of fibrosis. The FAA signal at 1396 cm−1 is characteristic of free amino acids as well as terminal COO- groups of protein molecules [21,22]. An increase in intensity for the FAA and P bands is likely attributed to increased collagen concentration; in particular, the FAA signal could relate to the short-chain collagen fraction. Taken together, the decreased P/FAA value in HT samples as compared to NT samples (Figure 5), and the negative correlation between P/FAA and the level of fibrosis (Figure 5d), suggest that the short-chain species increase is greater than the increase in long-chain collagen. In a recent study, a significant amount of perivascular collagen deposition (perivascular fibrosis) was observed in the kidney cortex of HT compared to NT rats [18]. Our data confirmed these findings and demonstrated that the molecular structures of the deposits are rich in amino acids/short-chain proteins. One possible explanation for this finding is the accumulation of non-fibrillar short-chain type VIII collagen in the outer vessel walls, as previously reported in diabetic nephropathy [23]. Second-harmonic-generation microscopy of human and mouse kidneys has revealed that nephropathy-related fibrosis and glomerulosclerosis correspond to an accumulation of different extracellular proteins [24]. Our results showed that, with hypertension, different biochemical changes occurred in vessels as compared to glomeruli and tubules. Recent findings in hypertensive DSS rats revealed that metabolic stress primarily affects glomeruli, whereas tubules showed changes related to altered amino acid handling [25]. The strong decrease in L/CG and L/FAA ratios that we observed was more prominent in glomeruli than in tubules (see Figure 5). This finding is consistent with prior reports of a reduction in glomerular lipid membranes in the setting of hypertension, perhaps due to oxidation processes [26]. The decrease in the L/CG ratio may also be a consequence of altered metabolism, as has been previously shown by Wang et al. [ 27] in a similar rat model. Following the CG marker band, univariate analysis and PCA indicated that the kidney parenchyma of HT rats has a higher carbohydrate and/or glycoprotein content compared to NT rats. This alteration affected both glomeruli and tubules and was substantially prevented by the administration of proANP31–67 (Figure 5a). Using isolated glomeruli from DSS hypertensive rats, Domondon et al. observed a decline in mitochondrial respiratory function, together with increased oxidative stress and reduced antioxidant capacity [26]. This impairment in mitochondrial function could explain our observation that carbohydrates accumulate in these structures. The increase in carbohydrate and glycoprotein content might also be explained by chronic glomerular ischemia and subsequent tubular atrophy induced by early hypertensive kidney damage [28]. Our results indicated that hypertension-related changes in carbohydrate and glycoprotein contents of the renal parenchyma are readily detectable by FTIR analysis. ProANP31–67 treatment resulted in normalization of the carbohydrate, protein, and amino acid content in all analyzed renal compartments. These effects were associated with a reduction in fibrosis, as assessed by conventional histological analysis. These favorable effects on renal chemical components and the reduced remodeling both resulted in enhanced GFR. Despite the clear renal protective effects of proANP31–67, it is possible that proANP31–67 had other direct effects on the renal vasculature, as we observed altered patency of the vessels in the kidneys of the treated animals. The pathophysiological consequences of such altered vascular patency were not investigated as they were beyond the scope of the current investigation. ## 4.1. Experimental Group and Study Protocol The experimental protocol was approved by the Committee for Animal Research of the Norwegian Food Safety Authority (Mattilsynet, protocol number 12582). DSS rats were purchased from Charles River Laboratories (Wilmington, MA, USA) and housed in a room with a 12:12 h light:dark cycle at a temperature of 21 °C and a humidity level of $55\%$. Consistent with the American Veterinary Medical Association (AVMA) Guidelines for the Euthanasia of Animals [2020], animals were sacrificed via deep anesthesia ($5\%$ isoflurane), exsanguination, and organ excision. Nine DSS male rats (~150 g initially) were used; three rats per group were recognized as the minimum number of animals needed to achieve reproducibility in terms of results, given the established consistency of the model and the high sensitivity of FTIR to monitor the chemical composition of cells and tissues [13,15,19]. The DSS rat develops high blood pressure within two days and cardiorenal symptoms within four weeks when chronically fed a high-salt diet [18]. All rats were given a normal ($0.3\%$ NaCl) salt diet (Special Diets Services, United Kingdom) until seven weeks of age. At that stage, three rats were kept normotensive (NT), while six were fed a high-salt ($4\%$ NaCl) diet (Envigo, TD.92034; Madison, WI, USA) for six weeks to induce hypertension. Starting four weeks before sacrifice (two weeks after initiation of the high-sodium diet), three hypertensive (HT) rats were chronically treated with subcutaneous (s.c.) infusion of proANP31–67 (50 ng/kg per day) via an Alzet osmotic mini-pump and three served as HT controls receiving vehicle. Parameters of renal function were assessed as we previously described [18]. ## 4.2. Histochemistry Kidneys were excised, rinsed in PBS and blotted on gauze, and fixed in $10\%$ formalin for 24 h. The right kidney was embedded in paraffin and cut into 4 µm sections, which were stained with Masson’s trichrome (Polysciences, Inc., Warrington, PA, USA) to assess collagen deposition. Stained sections were scanned (20× magnification) with AxioScan Z1 (Carl Zeiss, Jena, Germany) to obtain whole cross-sections for collagen quantification. Perivascular fibrosis was defined as the area of collagen surrounding the vessel wall indexed to the vessel lumen area (PVCA/LA), averaged over all quantifiable images of vessels taken from the renal cortex per section (mean, 4.73 ± 0.88 quantifiable images of arteries). Only rounded (not collapsed) vessels were used for quantification. Quantification was performed using ZEN2 blue edition (Carl Zeiss). All histological quantifications were performed independently by an experienced researcher who was blinded to group identity. ## 4.3. FTIR Measurements Deparaffined tissue sections were used. FTIR spectra were collected in transmission mode by using a Bruker Tensor 27 spectrometer and a Hyperion 3000 microscope equipped with a 15× Cassegrain objective and a 64 × 64 pixel focal-plane-array detector. The spectral range between 900 cm−1 and 3800 cm−1 was recorded at 6 cm−1 resolution. Spectral images of 180 µm × 180 µm with a pixel resolution of 2.8 µm were collected by averaging 256 measurements. At least three different areas from both medullary and cortical regions of each sample were analyzed, with particular reference to tubules and renal corpuscles. To reduce the influence of the baseline, which is remarkably variable due to the irregularity of the tissue, FTIR images were processed in second derivative spectroscopy; before HCA and PCA, a vector normalization in the 900–1800 cm−1 range was executed. IR spectrum of proANP31–67 was also measured and compared to tissue spectra; due to the low peptide concentration, we did not find any clear indication of proANP absorptions in the spectra of treated samples (Figure S4). ## 4.4. Statistical Analysis The spectral dataset was analyzed using a proprietary R-package specifically developed for the hyperspectral analysis of big datasets, which includes tools for HCA and PCA [29]. Both procedures were initially tested on the entire spectral range (900–3800 cm−1) and then restricted to subranges to enhance the sensitivity of each technique after exploring several options. HCA was used on FTIR images of control samples to identify different renal structures, selecting the 900–1400 cm−1 spectral range for the distance measurement. The dataset of second derivative spectra obtained at each pixel of FPA exposure was processed to the computation of distance matrix by the Euclidean method and clusterization using Ward’s method. PCA was applied to include the different types of samples (NT, HT, and proANP31–67) after the selection of the spectra from different renal structures, such as internal and external regions of vessels, glomeruli, and tubules. These selections were guided by optical images and submitted to quality tests. The spectra arising from the empty spaces in the samples were rejected by setting a minimum threshold equal to $\frac{1}{10}$ of the maximum integral intensity of the Amide I absorption band at 1620–1700 cm−1. Spectra with insufficient signal-to-noise ratio were also treated according to a quality test based on the intensity of the Amide I peak and the noise in a region between 1800 and 1900 cm⁻¹. We rejected spectra having a ratio lower than 80. The 1000–1800 cm−1 and 1000–1200 cm−1 spectral ranges were used for PCA of the entire dataset; the 1000–1800 cm−1 range was also used to analyze the spectra of specific tissue structures, i.e., vessels, glomeruli, and tubules. In both cases, a centered, unscaled principal component analysis was executed to obtain loadings and scores. With respect to the spectra treated by PCA, a univariate analysis was also performed by an estimate of integral intensities in the 2840–2860, 1485–1700, 1000–1128, and 1358–1416 cm−1 spectral regions. Integration of peak areas was performed by Opus 8.1 software from Bruker Optics; mean values and standard deviations of integrated area, correlation, and statistical significance by estimation of the p value were obtained by OriginPro 2021b software from OriginLab Corporation, Northampton, MA, USA. The sequence of data manipulation is described in Figure S5. ## 5. Conclusions In the present study, we report that FTIR spectroscopy provided molecular images of renal tissue that demonstrate alterations secondary to early-stage hypertension. Renal structures, such as blood vessels, glomeruli, and tubules, were chemically distinguished without any staining by assessing FTIR spectral features using HCA. We identified different spectral markers of hypertension in these structures. Univariate and multivariate analyses confirmed the same result, indicating that FTIR spectroscopy is an efficient and selective diagnostic tool for early renal tissue alterations. The major biochemical alterations after six weeks of sustained hypertension were detected in the renal parenchyma. Specifically, HT tubules and glomeruli showed an increase in carbohydrate and glycoprotein content as compared to NT. Of note, FTIR was able to detect the effect of chronic therapy with proANP31–67, which showed potential beneficial effects on the kidney function and structure. Remarkably, we observed a moderate biochemical change in carbohydrate and glycoprotein content in proANP31–67-treated rats, confirming its protective effect and opening new avenues of investigation into its mechanisms of action. Our study supports the potential utility of FTIR as an innovative method for the detection of early chemical changes in the renal parenchyma. These changes are otherwise underdiagnosed or missed as they precede changes in renal function such as a reduction in GFR. FTIR assessment of these chemical modifications of renal tissue could be useful to monitor the evolution of renal disease as well as the effects of therapeutic interventions. ## References 1. Lager D.. Heptinstall’s Pathology of the Kidney. Am. J. Surg. Pathol. (2008) 32 344. DOI: 10.1097/PAS.0b013e3181468b05 2. Folkow B., Göthberg G., Lundin S., Ricksten S.-E.. Structural “Resetting” of the Renal Vascular Bed in Spontaneously Hypertensive Rats (SHR). Acta Physiol. Scand. (1977) 100 270-272. DOI: 10.1111/j.1748-1716.1977.tb05949.x 3. Wadei H.M., Textor S.C.. The Role of the Kidney in Regulating Arterial Blood Pressure. Nat. Rev. Nephrol. (2012) 8 602-609. DOI: 10.1038/nrneph.2012.191 4. Mennuni S., Rubattu S., Pierelli G., Tocci G., Fofi C., Volpe M.. Hypertension and Kidneys: Unraveling Complex Molecular Mechanisms Underlying Hypertensive Renal Damage. J. Hum. Hypertens (2014) 28 74-79. DOI: 10.1038/jhh.2013.55 5. Shankland S.J.. The Podocyte’s Response to Injury: Role in Proteinuria and Glomerulosclerosis. Kidney Int. (2006) 69 2131-2147. DOI: 10.1038/sj.ki.5000410 6. Weldegiorgis M., Woodward M.. The Impact of Hypertension on Chronic Kidney Disease and End-Stage Renal Disease Is Greater in Men than Women: A Systematic Review and Meta-Analysis. BMC Nephrol. (2020) 21. DOI: 10.1186/S12882-020-02151-7 7. Shlipak M.G., Tummalapalli S.L., Boulware L.E., Grams M.E., Ix J.H., Jha V., Kengne A.P., Madero M., Mihaylova B., Tangri N.. The Case for Early Identification and Intervention of Chronic Kidney Disease: Conclusions from a Kidney Disease: Improving Global Outcomes (KDIGO) Controversies Conference. Kidney Int. (2021) 99 34-47. DOI: 10.1016/j.kint.2020.10.012 8. Walker P.D.. The Renal Biopsy. Arch. Pathol. Lab. Med. (2009) 133 181-188. DOI: 10.5858/133.2.181 9. Esposito V., Mazzon G., Baiardi P., Torreggiani M., Semeraro L., Catucci D., Colucci M., Mariotto A., Grosjean F., Bovio G.. Safety and Adequacy of Percutaneous Kidney Biopsy Performed by Nephrology Trainees. BMC Nephrol. (2018) 19. DOI: 10.1186/s12882-017-0796-y 10. D’Agati V.D., Mengel M.. The Rise of Renal Pathology in Nephrology: Structure Illuminates Function. Am. J. Kidney Dis. (2013) 61 1016-1025. DOI: 10.1053/j.ajkd.2012.12.019 11. Movasaghi Z., Rehman S., Rehman I.U.. Fourier Transform Infrared (FTIR) Spectroscopy of Biological Tissues. Appl. Spectrosc. Rev. (2008) 43 134-179. DOI: 10.1080/05704920701829043 12. Šablinskas V., Urbonienė V., Ceponkus J., Laurinavicius A., Dasevicius D., Jankevičius F., Hendrixson V., Koch E., Steiner G.. Infrared Spectroscopic Imaging of Renal Tumor Tissue. J. Biomed. Opt. (2011) 16 096006. DOI: 10.1117/1.3622292 13. Pucetaite M., Velicka M., Urboniene V., Ceponkus J., Bandzeviciute R., Jankevicius F., Zelvys A., Sablinskas V., Steiner G.. Rapid Intra-Operative Diagnosis of Kidney Cancer by Attenuated Total Reflection Infrared Spectroscopy of Tissue Smears. J. Biophotonics (2018) 11 e201700260. DOI: 10.1002/jbio.201700260 14. Palaniappan P.L.R.M., Vijayasundaram V.. Arsenic-Induced Biochemical Changes in Labeo Rohita Kidney: An FTIR Study. Spectrosc. Lett. (2009) 42 213-218. DOI: 10.1080/00387010902893033 15. Esteve E., Luque Y., Waeytens J., Bazin D., Mesnard L., Jouanneau C., Ronco P., Dazzi A., Daudon M., Deniset-Besseau A.. Nanometric Chemical Speciation of Abnormal Deposits in Kidney Biopsy: Infrared-Nanospectroscopy Reveals Heterogeneities within Vancomycin Casts. Anal. Chem. (2020) 92 7388-7392. DOI: 10.1021/acs.analchem.0c00290 16. Tombolesi N., Altara R., da Silva G.J.J., Tannous C., Zouein F.A., Stensløkken K.O., Morresi A., Paolantoni M., Booz G.W., Cataliotti A.. Early Cardiac-Chamber-Specific Fingerprints in Heart Failure with Preserved Ejection Fraction Detected by FTIR and Raman Spectroscopic Techniques. Sci. Rep. (2022) 12 1-12. DOI: 10.1038/s41598-022-07390-2 17. da Silva G.J.J., Altara R., Booz G.W., Cataliotti A.. Atrial Natriuretic Peptide31-67: A Novel Therapeutic Factor for Cardiovascular Diseases. Front. Physiol. (2021) 12 691407. DOI: 10.3389/fphys.2021.691407 18. Altara R., da Silva G.J.J., Frisk M., Spelta F., Zouein F.A., Louch W.E., Booz G.W., Cataliotti A.. Cardioprotective Effects of the Novel Compound Vastiras in a Preclinical Model of End-Organ Damage. Hypertension (2020) 75 1195-1204. DOI: 10.1161/HYPERTENSIONAHA.120.14704 19. Talari A.C.S., Martinez M.A.G., Movasaghi Z., Rehman S., Rehman I.U.. Advances in Fourier Transform Infrared (FTIR) Spectroscopy of Biological Tissues. Appl. Spectrosc. Rev. (2017) 52 456-506. DOI: 10.1080/05704928.2016.1230863 20. Jain M.. Hypertensive Renal Disease: Histological Aspects. Clin. Queries Nephrol. (2013) 2 23-28. DOI: 10.1016/j.cqn.2013.02.002 21. Barth A.. The Infrared Absorption of Amino Acid Side Chains. Prog. Biophys. Mol. Biol. (2000) 74 141-173. DOI: 10.1016/S0079-6107(00)00021-3 22. Barth A., Zscherp C.. What Vibrations Tell Us about Proteins. Q Rev. Biophys. (2002) 35 369-430. DOI: 10.1017/S0033583502003815 23. Rüger B.M., Hasan Q., Greenfaill N.S., Davis E.F., Dunbar P.R., Neaie T.J.. Mast Cells and Type VIII Collagen in Human Diabetic Nephropathy. Diabetologia (1996) 39 1215-1222. DOI: 10.1007/BF02658509 24. Strupler M., Hernest M., Fligny C., Martin J.-L., Tharaux P.-L., Schanne-Klein M.-C.. Second Harmonic Microscopy to Quantify Renal Interstitial Fibrosis and Arterial Remodeling. J. Biomed. Opt. (2008) 13 054041. DOI: 10.1117/1.2981830 25. Rinschen M.M., Palygin O., Guijas C., Palermo A., Palacio-Escat N., Domingo-Almenara X., Montenegro-Burke R., Saez-Rodriguez J., Staruschenko A., Siuzdak G.. Metabolic Rewiring of the Hypertensive Kidney. Sci. Signal. (2019) 12 9760. DOI: 10.1126/scisignal.aax9760 26. Domondon M., Polina I., Nikiforova A.B., Sultanova R.F., Kruger C., Vasileva V.Y., Fomin M.V., Beeson G.C., Nieminen A.L., Smythe N.. Renal Glomerular Mitochondria Function in Salt-Sensitive Hypertension. Front. Physiol. (2020) 10 1588. DOI: 10.3389/fphys.2019.01588 27. Liang M., Sun Q., Sun N., Tian Z.. Mitochondrial Dysfunction and Altered Renal Metabolism in Dahl Salt-Sensitive Rats. Kidney Blood. Press Res (2017) 42 587-597. DOI: 10.1159/000479846 28. Varma V.K., Kajdacsy-Balla A., Akkina S.K., Setty S., Walsh M.J.. A Label-Free Approach by Infrared Spectroscopic Imaging for Interrogating the Biochemistry of Diabetic Nephropathy Progression. Kidney Int. (2016) 89 1153-1159. DOI: 10.1016/j.kint.2015.11.027 29. Tombolesi N.. Combined FTIR and Raman Spectroscopic Study for the Characterization of Biological Tissues. Ph.D. Thesis (2022)
--- title: Secretory Factors from Calcium-Sensing Receptor-Activated SW872 Pre-Adipocytes Induce Cellular Senescence and A Mitochondrial Fragmentation-Mediated Inflammatory Response in HepG2 Cells authors: - Lautaro Briones-Suarez - Mariana Cifuentes - Roberto Bravo-Sagua journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049719 doi: 10.3390/ijms24065217 license: CC BY 4.0 --- # Secretory Factors from Calcium-Sensing Receptor-Activated SW872 Pre-Adipocytes Induce Cellular Senescence and A Mitochondrial Fragmentation-Mediated Inflammatory Response in HepG2 Cells ## Abstract Adipose tissue inflammation in obesity has a deleterious impact on organs such as the liver, ultimately leading to their dysfunction. We have previously shown that activation of the calcium-sensing receptor (CaSR) in pre-adipocytes induces TNF-α and IL-1β expression and secretion; however, it is unknown whether these factors promote hepatocyte alterations, particularly promoting cell senescence and/or mitochondrial dysfunction. We generated conditioned medium (CM) from the pre-adipocyte cell line SW872 treated with either vehicle (CMveh) or the CaSR activator cinacalcet 2 µM (CMcin), in the absence or presence of the CaSR inhibitor calhex 231 10 µM (CMcin+cal). HepG2 cells were cultured with these CM for 120 h and then assessed for cell senescence and mitochondrial dysfunction. CMcin-treated cells showed increased SA-β-GAL staining, which was absent in TNF-α- and IL-1β-depleted CM. Compared to CMveh, CMcin arrested cell cycle, increased IL-1β and CCL2 mRNA, and induced p16 and p53 senescence markers, which was prevented by CMcin+cal. Crucial proteins for mitochondrial function, PGC-1α and OPA1, were decreased with CMcin treatment, concomitant with fragmentation of the mitochondrial network and decreased mitochondrial transmembrane potential. We conclude that pro-inflammatory cytokines TNF-α and IL-1β secreted by SW872 cells after CaSR activation promote cell senescence and mitochondrial dysfunction, which is mediated by mitochondrial fragmentation in HepG2 cells and whose effects were reversed with Mdivi-1. This investigation provides new evidence about the deleterious CaSR-induced communication between pre-adipocytes and liver cells, incorporating the mechanisms involved in cellular senescence. ## 1. Introduction Cell senescence is described as a mechanism whereby a dividing cell enters a stable cell cycle arrest upon a stressing stimulus while remaining metabolically active. Senescent cells show the senescence-associated secretory phenotype (SASP) which involves the production of pro-inflammatory cytokines and other signaling factors and they become unresponsive to mitogenic and apoptotic signals [1]. Depending on the context, cell senescence can result in beneficial or detrimental effects to the organism. In young individuals or upon acute damage, senescence-related cell cycle arrest and SASP contribute to tumor suppression, wound healing, and tissue homeostasis [2]. Obesity is described as a chronic low-grade inflammatory process, where hypertrophic adipocytes accumulate in adipose tissue. This organ has an important endocrine role, secreting pro- or anti-inflammatory factors such as tumor necrosis factor α (TNF-α) and interleukin-1β (IL-1β) or interleukin-10 (IL-10) and adiponectin, respectively. The pro-inflammatory cytokines are associated with dysfunctional adipose tissue, which alters local and systemic metabolism, leading to obesity-related comorbidities [3,4]. Adipose tissue dysfunction has also been related to age-related diseases [3]. Adipose progenitor cells or “pre-adipocytes” are a cell population within adipose tissue with the ability to proliferate and differentiate into mature adipocytes, maintaining homeostasis through tissue plasticity and expandability [5,6,7]. An inflammatory environment can alter the homeostasis of pre-adipocytes, reducing their differentiation capacity and increasing their proliferation and pro-inflammatory cytokine expression [8,9]. Multiple inflammatory pathways have been described in adipose tissue, including the activation of the calcium-sensing receptor (CaSR), a G-protein coupled receptor that responds to several ligands and signals through multiple pathways. Pre-adipocytes express CaSR, and its activity is involved in the release of pro-inflammatory cytokines TNF-α and IL-1β [10,11,12], which may promote stress-mediated cellular senescence development in other tissues [13,14]. Aging and obesity modulate the cellular senescence program, leading to the accumulation of senescent cells that promote tissue dysfunction, chronic inflammation, and fibrosis in the liver [4,15]. Alterations in liver cell function are important because of its role in the control of metabolism, biomolecule secretion, and energy homeostasis [16,17]. The liver is rich in mitochondria and is responsible for $15\%$ of the total body oxygen consumption. Depending on the nutrient demand, the liver can oxidize various nutrients, such as carbohydrates and lipids, which are metabolized in mitochondria. Indeed, the liver can provide glucose to other tissues and, during fasting, its mitochondria can oxidize the fatty acids released from adipose tissue, producing ketones and ATP. Therefore, mitochondria are key players for hepatocyte energy homeostasis and their dysfunction is involved in the development of metabolic diseases [18,19]. It has been suggested that cell senescence and mitochondrial dysfunction are key hallmarks of aging [20]. During cell senescence, dysfunctional mitochondria accumulate due to a reduction in mitophagy [21]. Moreover, the senescent phenotype in hepatocytes is associated with decreased lipid oxidation, which favors hepatic inflammation and lipid accumulation (steatosis) [22]. The inflammatory environment generated by cytokines such as IL-1β and TNF-α alters the liver’s morphology and function, promotes fibrosis, mitochondrial dysfunction, and increases the number of senescent cells [22,23,24]. However, it is unknown whether the secretion products of pre-adipocytes after CaSR activation promote cell senescence and mitochondrial dysfunction in hepatocytes. Our aim was to investigate if CaSR activation induces pro-inflammatory secretion in the human pre-adipocyte cell line SW872, leading to pro-inflammatory cytokine expression, cellular senescence, and mitochondrial dysfunction in HepG2 cells. ## 2.1. Pro-Inflammatory Secretion Factors from CaSR-Activated SW872 Pre-Adipocytes Induce Cell Senescence in HepG2 Cells To evaluate the pro-senescence effect of pre-adipocyte-secreted factors, we obtained conditioned medium (CM) from SW872 cells treated with the CaSR activator cinacalcet 2 µM for 16 h (CMcin). As the control, we used CM from SW782 cells treated with vehicle alone (CMveh). The senescence-associated β-galactosidase (SA-β-GAL) assay (a marker of residual lysosomal activity associated with cell senescence) [13,25] showed that CMcin-treated HepG2 cells displayed higher SA-β-GAL expression, as well as cytomegaly compared to CMveh-treated cells ($p \leq 0.05$) (Figure 1a). We quantified SA-β-GAL expression both as the count of positive cells and colorimetry (Figure 1b,c). To test whether these changes depended on CaSR activation, we obtained CM from SW872 cells treated with cinacalcet in the presence of the CaSR inhibitor calhex 231 10 µM (CMcin+cal), which abolished the effect of CMcin (Figure 1b,c). To further characterize the senescent phenotype in our model, we evaluated the p53/p21/p16 signaling cascade. Cell cycle regulation by p53, p21, and p16 can be activated by stress signals, such as DNA damage or mitochondrial dysfunction [26,27,28]. CMcin-treated HepG2 cells showed increased protein levels of p16 and p21, but not p53, as assessed by Western blot analysis ($p \leq 0.05$) (Figure 2a–d). Again, treatment with CMcin+cal failed to induce any change, thereby supporting the role of CaSR activation in this process. Cell cycle arrest is a hallmark of cell senescence, which we evaluated by the location of the protein Ki67. This protein is diminished in the nucleus of senescent cells, and together with other markers, allows us to assess cell senescence [29]. As expected, HepG2 cells treated with CMcin showed a decreased Ki67 immunofluorescence signal compared with the control condition ($p \leq 0.05$). This change did not take place in the CMcin+cal condition (Figure 3a,b). Another key feature of senescent cells is the senescence-associated secretory phenotype (SASP), which consists of the induction of pro-inflammatory cytokines and chemoattractants, such as IL-1β and CCL2, respectively [30,31]. Our results show that HepG2 cells treated with CMcin increased the relative abundance of both IL-1β and CCL2 mRNA compared to CMveh ($p \leq 0.05$), while CMcin+cal prevented these changes (Figure 4a–c). To explore the mediators of this pro-senescence effect, we separately immunoprecipitated IL-6, IL-1β, or TNF-α from the SW872-derived CMcin, given the evidence of their release by preadipocytes [10,11,32] in addition to their association with stress-associated cell senescence [33,34,35]. Our approach specifically reduced the levels of each cytokine from the CM (Figure 5a). We verified the effectiveness of the procedure by observing that each cytokine was enriched in the pellet of their respective immune complex (Figure 5b). Immunoprecipitation of either TNF-α (CMTNF-α[−]) or IL-1β (CMIL-1β[−]) prevented the increase in SA-β-GAL in CMcin-treated HepG2 cells, while IL-6 immunoprecipitation (CMIL-6[−]) did not evoke a significant effect (Figure 5c–e). ## 2.2. Secretion Products from CaSR-Activated SW872 Pre-Adipocytes Alter the Mitochondrial Dynamics and Function of HepG2 Cells Given that cell senescence is characterized by a general decrease in cell function including mitochondrial dysfunction [27,36], we evaluated key factors for mitochondrial dynamics. Our results show that the protein levels of OPA1, which maintains mitochondrial cristae structure, and PGC-1α, which promotes mitochondrial biogenesis, significantly decrease in HepG2 cells exposed to CMcin ($p \leq 0.05$) compared to CMveh. In the case of MFN2 protein levels, which mediates mitochondrial fusion, there was a trend to decrease ($p \leq 0.08$). Conversely, the protein levels of DRP1, which participates in mitochondrial fission, increased upon treatment with CMcin (Figure 6a–f). As with previous results, these observations were reversed in the CMcin+cal condition. Next, we analyzed the mitochondrial transmembrane potential (Δψ) through immunofluorescence using the Δψ-sensitive probe MitoTracker Orange (MTO) normalized by mtHsp70 as a marker of mitochondrial mass. We observed that HepG2 cells exposed to CMcin had a decreased MTO/mtHsp70 fluorescence ratio ($p \leq 0.05$), indicative of decreased Δψ (Figure 7a,b). In addition, CMcin-treated cells presented a higher number of mitochondria per cell compared to the CMveh condition and a trend towards reduced average mitochondrial area ($p \leq 0.08$) (Figure 7c,d). Altogether, these results indicate that CMcin treatment in pre-adipocytes induces the secretion of factors that promote mitochondrial fragmentation and the decrease of bioenergetics in HepG2 cells, thus suggesting mitochondrial dysfunction. ## 2.3. Inhibition of Mitochondrial Fission in HepG2 Cells Prevents Cell Senescence Induced by Cinacalcet-Treated SW872 Pre-Adipocyte Secretion Factors We addressed whether the observed mitochondrial fragmentation contributes to the development of MCcin-induced cell senescence in HepG2 cells by inhibiting mitochondrial network fragmentation using the DRP1 inhibitor Mdivi-1 during the last 24 h of the treatment with SW872 CM. As expected, Mdivi-1 treatment prevented CMcin-induced mitochondrial fragmentation, maintaining the mitochondrial number similar to CMveh-treated cells (Figure 8a,c,d). Similarly, Mdivi-1 treatment prevented the decrease in Δψ observed upon CMcin treatment (Figure 8b). We next investigated how Mdivi1 affects the mitochondrial bioenergetics of HepG2 treated with CMcin. We observed that while CMcin did not significantly alter baseline respiration compared to CMveh, it decreased both non-ATP associated and maximal capacity respiration rates, confirming mitochondrial dysfunction. In both cases, treatment with Mdivi-1 reverted said changes (Figure 9a–c). Finally, we assessed whether mitochondrial fragmentation inhibition counteracts CMcin-induced cell senescence, evaluated as cell cycle arrest and SASP in HepG2 cells. Indeed, Mdivi-1 treatment prevented the decrease in Ki67 nuclear fluorescence (Figure 10a,b) induced by CMcin, as well as the increase in IL-1β and CCL2 mRNA levels (Figure 11a,b). Altogether, these results highlight the importance of mitochondrial fragmentation for the progression of cell senescence induced by CaSR-activated pre-adipocytes. ## 3. Discussion The present work evaluated whether CaSR activation in SW872 pre-adipocytes promotes pro-inflammatory signaling that induces senescence and mitochondrial dysfunction in HepG2 cells. Our observations indicate that CaSR activation in the SW872 cell line stimulated the production of factors that increased senescence, the production of pro-inflammatory cytokines, cell cycle arrest, and mitochondrial dysfunction markers in HepG2 cells. In pre-adipocytes, CaSR activation has been shown to promote the secretion of the pro-inflammatory cytokines TNF-α and IL-1β [10,11]. Elevated circulating TNF-α is associated with an increase in cellular senescence markers, such as SA-β-GAL, p21, and p53 in the liver and kidneys [37]. Our observations are consistent with other investigations where TNF-α was able to increase SA-β-GAL [37]. Moreover, we also observed that removing IL-1β prevented an increase in SA-β-GAL. Thus, our results highlight that both IL-1β and TNF-α in the conditioned medium from CaSR-activated pre-adipocytes contribute to cellular senescence development in the HepG2 cell line. Studies in macrophages have shown that exposure to conditioned medium from senescent pre-adipocytes increase p16 and p21 expression, which was attributed to the secretory products that make up the SASP, such as TNF-α and CCL2 [38,39]. Other reports on human vascular smooth muscle cell cultures have shown that IL-1β exposure increases p16 and p21 expression through a mechanism involving SIRT1 deregulation [40]. Decreased activation of the SIRT1 metabolic pathway has been observed in models of aging and inflammation. In rats, SIRT1 activation promoted mitochondrial biogenesis and the endogenous antioxidant system through PGC1a and NRF2 activity on neuronal tissue [41]. A study in cells from the nucleus pulposus showed that TNF-α promoted the cellular senescence markers SA-β-GAL and p21, a change that was related to greater activity of NF-κB [42]. Unlike p16 and p21, in our experiments we did not observe significant changes on p53, a protein that modulates p21 expression. In relation to this, it has been described that p21 can be regulated in a p53-independent manner through ERK$\frac{1}{2}$ and p38 MAPK [43]. It is well-accepted that the inflammatory activity of TNF-α and IL-1β initiates stress-driven cellular senescence. The process begins with an increase in cellular senescence markers such as SA-β-GAL and greater expression of pro-inflammatory cytokines and chemoattractants such as IL-1β and CCL2 [44]. In this context, the increased expression of IL-1β and CCL2 triggers the spread of inflammation and the attraction of immune cells, which eliminate senescent cells [14,44,45]. Studies in animal models have shown that the low expression of OPA1 is related to the increase in the pro-inflammatory cytokines TNF-α and IL-1β in the circulation [46]. In the liver of OPA1-deficient mice, the markers for cellular senescence, SA-β-GAL and p21, were increased [46]. We observed low expression of PGC-1α in CMcin-treated HepG2 cells, which is consistent with results in mouse liver tissue showing that stimuli such as a high-fat diet decreases PGC-1α expression and increases cellular senescence markers, such as SASP [47]. Moreover, the low expression of PGC-1α is related to reduced protection against oxidative stress and cellular senescence [48]. In the CMcin+cal treatment, MFN2 recovered its expression and DRP1 decreased compared to CMcin. Low expression of MFN2 has been associated with increased fragmentation of the mitochondrial network and lower membrane potential [49,50]. It has been reported that pro-inflammatory stimuli such as TNF-α increase mitochondrial dysfunction markers such as the fragmentation of the mitochondrial network [51]. In senescent cells, mitochondrial alterations such as fragmentation in the mitochondrial network or loss in the Δψ are common features [52,53]. Our results coincide with those reported in colon and melanocyte cell lines exposed to overexpression of TNF-α, which decreased the Δψ and increased the fragmentation of the mitochondrial network. After an early process of apoptosis, the remaining cells show mitochondrial dysfunction and increased cytokine expression, which is associated with cellular senescence [53]. The evidence on altered mitochondrial dynamics and cellular senescence is under discussion because the results may vary depending on the cell type under study [22,54]. Based on our findings, we propose that the secretion products in CMcin affect mitochondrial morphology through decreased expression of proteins that promote organelle biogenesis and fusion, as well as increased fragmentation of the mitochondrial network. In addition, treated cells presented an alteration in mitochondrial bioenergetics, showing a lower Δψ and lower respiratory rate. These data suggest that CMcin damages the mitochondrial network, and thus limits the biogenesis and bioenergetics of this organelle. Losing mitochondrial network integrity favors the development of mitochondrial dysfunction, and in our HepG2 cells treated with CMcin this indicator increased. In experiments in HepG2 cells where Ki67 was pharmacologically reduced, Ki67 was recovered after Mdivi-1 treatment [55]. Furthermore, in a rat model of ischemia, Mdivi-1 increased Δψ and the authors attribute this behavior to the regulatory action of the reagent on mitophagy mediated by PINK/Parking proteins [56]. It was also shown in LPS-exposed cardiomyocytes that Mdivi-1 contributes to the recovery of Δψ [57]. The effects of Mdivi-1 on the Δψ regulation would be related to mitochondrial membrane integrity maintenance. Our results show that Mdivi-1 inhibited the increase in IL-1β and CCL2 expression caused by CMcin. Published studies in animal models exposed to amyloid β reveal that treatment with Mdivi-1 protects hippocampal cells from increased expression of IL-1β, which changes in mitochondrial dynamics, decreasing the MFN2 and OPA1 proteins [58]. In a cell model, it has been observed that treatment with Mdivi-1 before LPS exposure impaired the increase in pro-inflammatory cytokines such as IL-1β and CCL2 [59]. The present work addresses a line of research that innovates on the relationship between CaSR activation in pre-adipocytes and the development of the senescent phenotype and mitochondrial dysfunction in hepatic cells. It highlights and supports previous research on the importance of pre-adipocytes as potential promoters of deleterious effects in other cell types, further evidencing the participation of mitochondrial dysfunction as a regulator of the senescent phenotype. Future studies should also include a wider use of CaSR inhibition strategies such as the CaSR-negative modulator calhex 231 and possibly others, and furthermore with CaSR gene downregulation via siRNA, in order to better understand the involvement of this receptor in the observed effects. Further characterization of the senescent phenotype would have been desirable, such as cell cycle analysis through flow cytometry or chromatin remodeling or DNA damage via fluorescence microscopy. In addition, we did not explore other functional outcomes in senescent HepG2 cells, such as lipid uptake or insulin signaling. Finally, this model of cellular communication can be used with other cell types. Thus, the effects of conditioned media from CaSR-activated pre-adipocytes can be assessed in smooth muscle or endothelial cells as a future research direction. Our results indicate that inhibiting mitochondrial network fragmentation protects HepG2 cells from changes related to CMcin exposure, seen in the recovery of Δψ and the decreased expression of pro-inflammatory cytokines. In summary, the conditioned medium from pre-adipocytes after CaSR activation promotes senescence and mitochondrial dysfunction in hepatocytes. Our investigation provides evidence about the communication between pre-adipocytes and liver cells mediated by CaSR activation, raising new questions such as further pre-adipocyte conditioned medium characterization or the consequences of CaSR activation at the organismal level on liver senescence phenotype and disease. ## 4.1. Cell Culture A human liposarcoma-derived SW872 pre-adipose cell line (HTB-92, ATCC, Manassas, VA, USA) was grown in DMEM/F12 medium. A human hepatocellular carcinoma-derived HepG2 cell line (HB-8065, ATCC, Manassas, VA, USA) was cultured in MEM medium. All media were supplemented with $10\%$ fetal bovine serum and antibiotics (penicillin–streptomycin) and the cells were maintained in a 37 °C $5\%$ CO2 atmosphere. The medium was changed for a fresh medium twice per week. ## 4.2. SW872 Cells’ Conditioned Media Collection SW872 cells were cultured at a 10,000 cells/cm2 density in 100 mm plastic culture dishes. To obtain conditioned media, the cells were treated with either vehicle (DMSO) or 2 µM cinacalcet with or without 30 min of pre-incubation with the negative allosteric CaSR modulator calhex-231 (10 µM). After 16 h, the media were replaced with fresh DMEM/F12 and conditioned for 24 h. The conditioned media were centrifuged at 800× g for 10 min at 4 °C and stored at −80 °C until use. ## 4.3. HepG2 Cells’ Exposure to Conditioned Media and Mdivi1 HepG2 cells were cultured at 9000 cells/cm2 density in culture dishes, according to each experimental protocol. The cells were washed twice with PBS and the media were replaced with fresh MEM media containing $50\%$ of the conditioned media from SW872 cells. The HepG2 cells were then grown for 5 days under standard culture conditions, replacing the media for fresh conditioned media after 3 days. For treatment with Mdivi1, the cells were conditioned as described above, and during the last 24 h of exposure they were treated with vehicle (DMSO) of 50 µM of Mdivi1. ## 4.4. Immunoprecipitation IL-6, IL-1β, or TNF-α were immunoprecipitated from the CMcin. For that, 10 μg of IL-6, IL-1β, or TNF-α antibodies (Table 1) were incubated with 200 μL of hydrated Protein A-Sepharose CL-4B (17-0963-03, Sigma, St. Louis, MO, USA) for 16 h to obtain a solution with the immobilized antibody. Then, the immobilized antibody solution was mixed with 1 mL of CMcin 12 h at 4 °C in a rotary mixer. The sample was then centrifuged at 3000× g for 5 min at 4 °C. Supernatants were used to treat HepG2 cells and evaluate SA-β-GAL. To analyze cytokine protein levels in the immunoprecipitate, the pellets were resuspended in 30 µL of Laemmli loading buffer and evaluated through Western blot analysis. ## 4.5. HepG2 SA-β-GAL Activity The cells were cultured in 96-well plates and observed under an inverted phase contrast microscope before and after treatment for 5 days. Images were recorded with a Motic AE2000 camera and Motic Images Plus 2.0 ML software (Motic, Vancouver, BC, Canada). The cultures were treated with the senescence-associated β-galactosidase staining kit (Cell Signaling Technology, Danvers, MA, USA) following the manufacturer’s instructions. After incubating overnight at 37 °C, the cells were examined by microscopy and the respective blue staining in positive cells was recorded. Next, the number of positive cells was normalized by the total number of cells in the photograph to obtain an indicator of the percentage of senescent cells in each experiment. Then, X-gal blue products were dissolved with DMSO and the absorbance was measured at 630 mn [60]. ## 4.6. RNA Isolation, Reverse Transcription, and mRNA Expression by RT-PCR The cells were cultured in 6-well plates and after experimentation, Trizol (Invitrogen, Life Technologies, Carlsbad, CA, USA) was used to lyse the HepG2 cells and isolate total RNA following the manufacturer’s instructions. RNA was reverse-transcribed into complementary DNA (cDNA) using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA). The cDNA levels were analyzed with a step-one real-time PCR system using the SYBR FAST qPCR kit (Applied Biosystems, Foster City, CA, USA) and specific primers for: il-1β (forward: 5′ GGACAAGCTGAGGAAGATGC 3′; reverse: 5′ TCGTTATCCCATGTGTCGAA 3′, NM_000576), ccl2 (forward: 5′ TGTCCCAAAGAAGCTGTGATCT 3′; Reverse: 5′ GGAATCCTGAACCCACTTCTG 3′; NM_002982), and gapdh (forward: 5′ GAAGGTGAAGGTCGGAGTCAAC 3′; reverse: 5′ CAGAGTTAAAAGCAGCCCTGGT 3′; NM_020). Thermal cycling consisted of an initial pre-incubation cycle of 20 s at 95 °C, followed by 40 cycles of 30 s at 95 °C. The results were normalized for the gapdh gene and relative mRNA levels were calculated using the Pfaffl method [61]. ## 4.7. Protein Levels by Western Blot Analysis The cells were cultured in 6-well plates and after experimentation, they were washed three times with cold PBS and lysed with NP40 buffer (Invitrogen™, Carlsbad, CA, USA), complete protease inhibitor (#11697498001, Sigma-Aldrich, St. Louis, MO, USA), and PhosSTOP phosphatase inhibitor (#4906845001, Sigma-Aldrich, St. Louis, MO, USA). The lysed cells were centrifuged at 12,000× g for 15 min and the supernatant was stored at −80 °C until use. At concentrations of 1 μg/μL, the proteins were subjected to electrophoresis in SDS-polyacrylamide gels under denaturing conditions and were electrotransferred to a 0.2 μm pore PVDF membrane. The membranes were then blocked with TBS $5\%$ skim milk $0.05\%$ Tween 20 (#7949, Sigma, St. Louis, MO, USA). The following proteins were detected after 16 h of incubation with the primary antibodies: p16, p21, p53, PGC1α, MFN2, DRP1, OPA1, and β-actin (Table 1). The detection of immune complexes was performed with the incubation of anti-rabbit IgG, anti-mouse IgG, or anti-goat IgG peroxidase-conjugated secondary antibodies, depending on the origin of each primary antibody (Table 2). The membranes were incubated for 1 min with chemiluminescence reagents (#20-500-500A and 20-500-500B, Biological Industries, Cromwell, CT, USA), the digitized luminescent signal was detected in a LI-COR C-Digit scanner 3600 (LI-COR Biosciences, Lincoln, NE, USA), and the intensity of the bands was quantified with the ImageJ program (National Institutes of Health, USA). The expression of each protein was normalized by β-actin and expressed as arbitrary units. ## 4.8. Immunofluorescence The HepG2 cells were seeded in 12-well plates with 0.17 mm and treated as indicated. For analysis of mitochondrial membrane potential, the cells were loaded with MitoTracker Orange 400 nM (M7510, Invitrogen™, Carlsbad, CA, USA) for 25 min at 37 °C. Then, the culture medium was removed and the cells were washed twice with PBS. The cells were fixed with $4\%$ paraformaldehyde at 4 °C for 15 min. The fixation medium was removed and washed twice with cold PBS. To permeabilize the cells, they were incubated with PBS $0.1\%$ Triton X-100 for 10 min. After permeabilization, the cells were washed twice with cold PBS and blocked with PBS $3\%$ BSA for 1 h at room temperature. The samples were treated with primary antibodies in PBS $3\%$ BSA overnight at 4 °C in a humid chamber protected from light. The primary antibodies and their dilutions were: anti-Ki-67 (#9449, Cell Signaling, proliferation marker) 1:400 or anti-mtHsp70 (MA3-028, Thermo Fisher Scientific, mitochondrial marker) 1:500, according to the assay. The next day, the coverslips were washed twice with cold PBS to remove excess primary antibodies, and the samples were incubated for 1 h with an Alexa Fluor 488 secondary antibody (A32723, Thermo Fisher Scientific, Waltham, MA, USA) 1:600 in a chamber protected from light at room temperature. After washing with PBS, the samples were incubated with the fluorescent probe Hoechst (B2261, Merck (Rahway, NJ, USA), nucleus marker) 1:1000 diluted in PBS BSA $3\%$ for 15 min at room temperature in a chamber protected from light. The coverslips were then washed twice with cold PBS to remove excess probes and mounted on a clean slide with Dako fluorescence mounting medium (S3023, Dako-Agilent, Santa Clara, CA, USA). The samples were stored at 4 °C protected from light until analysis. ## 4.9. Image Capture and Processing Images were taken with a Zeiss LSM-5, Pascal 5 Axiovert 200 confocal microscope, with a Plan-Apochromat 40×/1.5 Oil DIC objective, using 365 (for Hoechst), 488 (for Alexa Fluor 488), and 555 nm (for MitoTracker Orange) excitation lasers. For each independent experiment, it averaged the signal from 5 to 15 cells. The analysis was performed on 1 focal plane corresponding to the equator of the cells. The pixel size was 68 nm, following the Nyquist sampling criterion. The images thus obtained were deconvolved, the background noise was subtracted, a median filter was applied, and fluorescence intensity was quantified using the ImageJ software. ## 4.10. Oxygraphy HepG2 cells were seeded in 60 mm dishes. After the 5 days of treatment, the cells were trypsinized and resuspended in a Clark electrode chamber (Strathkelvin Instruments, North Lanarkshire, Scotland) to quantify the oxygen consumption of living cells. Basal respiration was measured for 5 min at 25 °C, then 10 μg/mL of oligomycin was added to assess non-ATP-associated respiration for 5 min, followed by 200 nM CCCP to quantify maximal respiration capacity for another 5 min. Then, the cells were recovered to quantify total proteins using a BCA kit to normalize oxygen consumption rates. ## 4.11. Statistical Analysis Data are shown as the mean ± the standard error of the mean (SEM). Differences between conditions were evaluated with the non-parametric Friedman test and multiple comparisons were made with the Dunn’s test. The Wilcoxon signed rank test was used to detect differences in the results obtained through real-time PCR. A p value less than 0.05 was considered as a significant change, while p values less than 0.1 were considered a trend. ## References 1. Hernandez-Segura A., Nehme J., Demaria M.. Hallmarks of Cellular Senescence. Trends Cell Biol. (2018) 28 436-453. DOI: 10.1016/j.tcb.2018.02.001 2. Burton D.G.A., Faragher R.G.A.. Cellular Senescence: From Growth Arrest to Immunogenic Conversion. Age (2015) 37 27. DOI: 10.1007/s11357-015-9764-2 3. Spritzer P.M., Lecke S.B., Satler F., Morsch D.M.. Adipose Tissue Dysfunction, Adipokines, and Low-Grade Chronic Inflammation in Polycystic Ovary Syndrome. Reproduction (2015) 149 R219-R227. DOI: 10.1530/REP-14-0435 4. Hotamisligil G.S.. Inflammation, Metaflammation and Immunometabolic Disorders. Nature (2017) 542 177-185. DOI: 10.1038/nature21363 5. Cawthorn W.P., Scheller E.L., MacDougald O.A.. Adipose Tissue Stem Cells Meet Preadipocyte Commitment: Going Back to the Future. J. Lipid Res. (2012) 53 227-246. DOI: 10.1194/jlr.R021089 6. Gesta S., Tseng Y.-H., Kahn C.R.. Developmental Origin of Fat: Tracking Obesity to Its Source. Cell (2007) 131 242-256. DOI: 10.1016/j.cell.2007.10.004 7. Ouchi N., Parker J.L., Lugus J.J., Walsh K.. Adipokines in Inflammation and Metabolic Disease. Nat. Rev. Immunol. (2011) 11 85. DOI: 10.1038/nri2921 8. Gerhardt C.C., Romero I.A., Cancello R., Camoin L., Strosberg A.D.. Chemokines Control Fat Accumulation and Leptin Secretion by Cultured Human Adipocytes. Mol. Cell Endocrinol. (2001) 175 81-92. DOI: 10.1016/S0303-7207(01)00394-X 9. Zamboni M., Nori N., Brunelli A., Zoico E.. How Does Adipose Tissue Contribute to Inflammageing?. Exp. Gerontol. (2021) 143 111162. DOI: 10.1016/j.exger.2020.111162 10. D’Espessailles A., Mora Y.A., Fuentes C., Cifuentes M.. Calcium-Sensing Receptor Activates the NLRP3 Inflammasome in LS14 Preadipocytes Mediated by ERK1/2 Signaling. J. Cell. Physiol. (2018) 233 6232-6240. DOI: 10.1002/jcp.26490 11. Mattar P., Bravo-Sagua R., Tobar N., Fuentes C., Troncoso R., Breitwieser G., Lavandero S., Cifuentes M.. Autophagy Mediates Calcium-Sensing Receptor-Induced TNFalpha Production in Human Preadipocytes. Biochim. Biophys. Acta Mol. Basis Dis. (2018) 1864 3585-3594. DOI: 10.1016/j.bbadis.2018.08.020 12. Cifuentes M., Fuentes C., Tobar N., Acevedo I., Villalobos E., Hugo E., Ben-Jonathan N., Reyes M.. Calcium Sensing Receptor Activation Elevates Proinflammatory Factor Expression in Human Adipose Cells and Adipose Tissue. Mol. Cell Endocrinol. (2012) 361 24-30. DOI: 10.1016/j.mce.2012.03.006 13. Aravinthan A., Shannon N., Heaney J., Hoare M., Marshall A., Alexander G.J.M.. The Senescent Hepatocyte Gene Signature in Chronic Liver Disease. Exp. Gerontol. (2014) 60 37-45. DOI: 10.1016/j.exger.2014.09.011 14. Chandler H., Peters G.. Stressing the Cell Cycle in Senescence and Aging. Curr. Opin. Cell Biol. (2013) 25 765-771. DOI: 10.1016/j.ceb.2013.07.005 15. Ferrucci L., Fabbri E.. Inflammageing: Chronic Inflammation in Ageing, Cardiovascular Disease, and Frailty. Nat. Rev. Cardiol. (2018) 15 505-522. DOI: 10.1038/s41569-018-0064-2 16. Hashimoto S.. Glucose Metabolism and Liver. The Liver in Systemic Diseases (2016) 77-103 17. Robinson M.W., Harmon C., O’Farrelly C.. Liver Immunology and Its Role in Inflammation and Homeostasis. Cell Mol. Immunol. (2016) 13 267-276. DOI: 10.1038/cmi.2016.3 18. Morio B., Panthu B., Bassot A., Rieusset J.. Role of Mitochondria in Liver Metabolic Health and Diseases. Cell Calcium. (2021) 94 102336. DOI: 10.1016/j.ceca.2020.102336 19. Shum M., Ngo J., Shirihai O.S., Liesa M.. Mitochondrial Oxidative Function in NAFLD: Friend or Foe?. Mol. Metab. (2021) 50 101134. DOI: 10.1016/j.molmet.2020.101134 20. López-Otín C., Blasco M.A., Partridge L., Serrano M., Kroemer G.. The Hallmarks of Aging. Cell (2013) 153 1194-1217. DOI: 10.1016/j.cell.2013.05.039 21. Korolchuk V.I., Miwa S., Carroll B., von Zglinicki T.. Mitochondria in Cell Senescence: Is Mitophagy the Weakest Link?. EBioMedicine (2017) 21 7-13. DOI: 10.1016/j.ebiom.2017.03.020 22. Ogrodnik M., Miwa S., Tchkonia T., Tiniakos D., Wilson C.L., Lahat A., Day C.P., Burt A., Palmer A., Anstee Q.M.. Cellular Senescence Drives Age-Dependent Hepatic Steatosis. Nat. Commun. (2017) 8 15691. DOI: 10.1038/ncomms15691 23. Luedde T., Schwabe R.F.. NF-ΚB in the Liver—Linking Injury, Fibrosis and Hepatocellular Carcinoma. Nat. Rev. Gastroenterol. Hepatol. (2011) 8 108-118. DOI: 10.1038/nrgastro.2010.213 24. Papatheodoridi A.-M., Chrysavgis L., Koutsilieris M., Chatzigeorgiou A.. The Role of Senescence in the Development of Nonalcoholic Fatty Liver Disease and Progression to Nonalcoholic Steatohepatitis. Hepatology (2020) 71 363-374. DOI: 10.1002/hep.30834 25. Nagano T., Nakano M., Nakashima A., Onishi K.. Identification of Cellular Senescence-Specific Genes by Comparative Transcriptomics. Sci Rep. (2016) 6 31758. DOI: 10.1038/srep31758 26. González-Gualda E., Baker A.G., Fruk L., Muñoz-Espín D.. A Guide to Assessing Cellular Senescence in Vitro and in Vivo. FEBS J. (2021) 288 56-80. DOI: 10.1111/febs.15570 27. Kumari R., Jat P.. Mechanisms of Cellular Senescence: Cell Cycle Arrest and Senescence Associated Secretory Phenotype. Front. Cell. Dev. Biol. (2021) 9 645593. DOI: 10.3389/fcell.2021.645593 28. Sun X., Kaufman P.D.. Ki-67: More than a Proliferation Marker. Chromosoma (2018) 127 175-186. DOI: 10.1007/s00412-018-0659-8 29. Biran A., Zada L., Abou Karam P., Vadai E., Roitman L., Ovadya Y., Porat Z., Krizhanovsky V.. Quantitative Identification of Senescent Cells in Aging and Disease. Aging Cell (2017) 16 661-671. DOI: 10.1111/acel.12592 30. Hansel C., Jendrossek V., Klein D.. Cellular Senescence in the Lung: The Central Role of Senescent Epithelial Cells. Int. J. Mol. Sci. (2020) 21. DOI: 10.3390/ijms21093279 31. Yoshimoto S., Loo T.M., Atarashi K., Kanda H., Sato S., Oyadomari S., Iwakura Y., Oshima K., Morita H., Hattori M.. Obesity-Induced Gut Microbial Metabolite Promotes Liver Cancer through Senescence Secretome. Nature (2013) 499 97-101. DOI: 10.1038/nature12347 32. Renovato-Martins M., Moreira-Nunes C., Atella G.C., Barja-Fidalgo C., de Moraes J.A.. Obese Adipose Tissue Secretion Induces Inflammation in Preadipocytes: Role of Toll-Like Receptor-4. Nutrients (2020) 12. DOI: 10.3390/nu12092828 33. Li Y., Lu L., Xie Y., Chen X., Tian L., Liang Y., Li H., Zhang J., Liu Y., Yu X.. Interleukin-6 Knockout Inhibits Senescence of Bone Mesenchymal Stem Cells in High-Fat Diet-Induced Bone Loss. Front. Endocrinol. (2021) 11 622950. DOI: 10.3389/fendo.2020.622950 34. Romero A., San Hipólito-Luengo Á., Villalobos L.A., Vallejo S., Valencia I., Michalska P., Pajuelo-Lozano N., Sánchez-Pérez I., León R., Bartha J.L.. The Angiotensin-(1-7)/Mas Receptor Axis Protects from Endothelial Cell Senescence via Klotho and Nrf2 Activation. Aging Cell (2019) 18 e12913. DOI: 10.1111/acel.12913 35. Wang Q., Chen Q., Sui J., Tu Y., Guo X., Li F.. Celecoxib Prevents Tumor Necrosis Factor-α (TNF-α)-Induced Cellular Senescence in Human Chondrocytes. Bioengineered (2021) 12 12812-12820. DOI: 10.1080/21655979.2021.2003661 36. Gorgoulis V., Adams P.D., Alimonti A., Bennett D.C., Bischof O., Bishop C., Campisi J., Collado M., Evangelou K., Ferbeyre G.. Cellular Senescence: Defining a Path Forward. Cell (2019) 179 813-827. DOI: 10.1016/j.cell.2019.10.005 37. Desdín-Micó G., Soto-Heredero G., Aranda J.F., Oller J., Carrasco E., Gabandé-Rodríguez E., Blanco E.M., Alfranca A., Cussó L., Desco M.. T Cells with Dysfunctional Mitochondria Induce Multimorbidity and Premature Senescence. Science (2020) 368 1371-1376. DOI: 10.1126/science.aax0860 38. Kumar R., Sharma A., Padwad Y., Sharma R.. Preadipocyte Secretory Factors Differentially Modulate Murine Macrophage Functions during Aging Which Are Reversed by the Application of Phytochemical EGCG. Biogerontology (2020) 21 325-343. DOI: 10.1007/s10522-020-09861-3 39. Parvizi M., Ryan Z.C., Ebtehaj S., Arendt B.K., Lanza I.R.. The Secretome of Senescent Preadipocytes Influences the Phenotype and Function of Cells of the Vascular Wall. Biochim. Biophys. Acta (BBA)-Mol. Basis Dis. (2021) 1867 165983. DOI: 10.1016/j.bbadis.2020.165983 40. Zhao J., He X., Zuo M., Li X., Sun Z.. Anagliptin Prevented Interleukin 1 β (IL-1 β) -Induced Cellular Senescence in Vascular Smooth Muscle Cells through Increasing the Expression of Sirtuin1. Bioengineered (2021) 12 3968-3977. DOI: 10.1080/21655979.2021.1948289 41. Chuang Y.C., der Chen S., Jou S.-B., Lin T.K., Chen S.F., Chen N.C., Hsu C.Y.. Sirtuin 1 Regulates Mitochondrial Biogenesis and Provides an Endogenous Neuroprotective Mechanism Against Seizure-Induced Neuronal Cell Death in the Hippocampus Following Status Epilepticus. Int. J. Mol. Sci. (2019) 20. DOI: 10.3390/ijms20143588 42. Chen Z., Yang X., Zhou Y., Liang Z., Chen C., Han C., Cao X., He W., Zhang K., Qin A.. Dehydrocostus Lactone Attenuates the Senescence of Nucleus Pulposus Cells and Ameliorates Intervertebral Disc Degeneration via Inhibition of STING-TBK1/NF-ΚB and MAPK Signaling. Front. Pharmacol. (2021) 12 641098. DOI: 10.3389/fphar.2021.641098 43. Hou X., Shi J., Sun L., Song L., Zhao W., Xiong X., Lu Y.. The Involvement of ERK1/2 and P38 MAPK in the Premature Senescence of Melanocytes Induced by H. J. Dermatol. Sci. (2022) 105 88-97. DOI: 10.1016/j.jdermsci.2022.01.002 44. Wang Y., Che M., Xin J., Zheng Z., Li J., Zhang S.. The Role of IL-1β and TNF-α in Intervertebral Disc Degeneration. Biomed. Pharmacother. (2020) 131 110660. DOI: 10.1016/j.biopha.2020.110660 45. Huang Z., Lan J., Gao X.. Feprazone Mitigates IL-1β-Induced Cellular Senescence in Chondrocytes. ACS Omega (2021) 6 9442-9448. DOI: 10.1021/acsomega.0c06066 46. Tezze C., Romanello V., Desbats M.A., Fadini G.P., Albiero M., Favaro G., Ciciliot S., Soriano M.E., Morbidoni V., Cerqua C.. Age-Associated Loss of OPA1 in Muscle Impacts Muscle Mass, Metabolic Homeostasis, Systemic Inflammation, and Epithelial Senescence. Cell Metab. (2017) 25 1374-1389.e6. DOI: 10.1016/j.cmet.2017.04.021 47. Barroso W.A., Victorino V.J., Jeremias I.C., Petroni R.C., Ariga S.K.K., Salles T.A., Barbeiro D.F., de Lima T.M., de Souza H.P.. High-Fat Diet Inhibits PGC-1α Suppressive Effect on NFκB Signaling in Hepatocytes. Eur. J. Nutr. (2018) 57 1891-1900. DOI: 10.1007/s00394-017-1472-5 48. Xiong S., Patrushev N., Forouzandeh F., Hilenski L., Alexander R.W.. PGC-1α Modulates Telomere Function and DNA Damage in Protecting against Aging-Related Chronic Diseases. Cell Rep. (2015) 12 1391-1399. DOI: 10.1016/j.celrep.2015.07.047 49. Filadi R., Pendin D., Pizzo P.. Mitofusin 2: From Functions to Disease. Cell Death Dis. (2018) 9 330. DOI: 10.1038/s41419-017-0023-6 50. Putti R., Sica R., Migliaccio V., Lionetti L.. Diet Impact on Mitochondrial Bioenergetics and Dynamics. Front. Physiol. (2015) 6 109. DOI: 10.3389/fphys.2015.00109 51. Zhang L., Gan X., He Y., Zhu Z., Zhu J., Yu H.. Drp1-Dependent Mitochondrial Fission Mediates Osteogenic Dysfunction in Inflammation through Elevated Production of Reactive Oxygen Species. PLoS ONE (2017) 12. DOI: 10.1371/journal.pone.0175262 52. Ji S., Zheng Z., Liu S., Ren G., Gao J., Zhang Y., Li G.. Resveratrol Promotes Oxidative Stress to Drive DLC1 Mediated Cellular Senescence in Cancer Cells. Exp. Cell. Res. (2018) 370 292-302. DOI: 10.1016/j.yexcr.2018.06.031 53. Tyciakova S., Valova V., Svitkova B., Matuskova M.. Overexpression of TNFα Induces Senescence, Autophagy and Mitochondrial Dysfunctions in Melanoma Cells. BMC Cancer (2021) 21 507. DOI: 10.1186/s12885-021-08237-1 54. Ogrodnik M., Salmonowicz H., Jurk D., Passos J.F.. Expansion and Cell-Cycle Arrest: Common Denominators of Cellular Senescence. Trends Biochem. Sci. (2019) 44 996-1008. DOI: 10.1016/j.tibs.2019.06.011 55. Niu Z., Zhang W., Gu X., Zhang X., Qi Y., Zhang Y.. Mitophagy Inhibits Proliferation by Decreasing Cyclooxygenase-2 (COX-2) in Arsenic Trioxide-Treated HepG2 Cells. Environ. Toxicol. Pharmacol. (2016) 45 212-221. DOI: 10.1016/j.etap.2016.06.006 56. Huang Y., Gao X., Zhou X., Zhang Y., Tan Z.T., Zhu S.B.. Remote Ischemic Postconditioning Inhibited Mitophagy to Achieve Neuroprotective Effects in the Rat Model of Cardiac Arrest. Neurochem. Res. (2021) 46 573-583. DOI: 10.1007/s11064-020-03193-x 57. Zhu Y., Kuang L., Wu Y., Deng H., She H., Zhou Y., Zhang J., Liu L., Li T.. Protective Effects of Inhibition of Mitochondrial Fission on Organ Function After Sepsis. Front. Pharmacol. (2021) 12 712489. DOI: 10.3389/fphar.2021.712489 58. Batista A.F., Rody T., Forny-Germano L., Cerdeiro S., Bellio M., Ferreira S.T., Munoz D.P., de Felice F.G.. Interleukin-1β Mediates Alterations in Mitochondrial Fusion/Fission Proteins and Memory Impairment Induced by Amyloid-β Oligomers. J. Neuroinflamm. (2021) 18 54. DOI: 10.1186/s12974-021-02099-x 59. Nair S., Sobotka K.S., Joshi P., Gressens P., Fleiss B., Thornton C., Mallard C., Hagberg H.. Lipopolysaccharide-Induced Alteration of Mitochondrial Morphology Induces a Metabolic Shift in Microglia Modulating the Inflammatory Response in Vitro and in Vivo. Glia (2019) 67 1047-1061. DOI: 10.1002/glia.23587 60. Lan F., Lin Y., Gao Z., Cacicedo J.M., Weikel K., Ido Y.. Activation of LKB1 Rescues 3T3-L1 Adipocytes from Senescence Induced by Sirt1 Knock-down: A Pivotal Role of LKB1 in Cellular Aging. Aging (2020) 12 18942-18956. DOI: 10.18632/aging.104052 61. Pfaffl M.W.. A New Mathematical Model for Relative Quantification in Real-Time RT-PCR. Nucleic. Acids Res. (2001) 29 e45. DOI: 10.1093/nar/29.9.e45
--- title: 'The Impact of Medical Nutrition Intervention on the Management of Hyperphosphatemia in Hemodialysis Patients with Stage 5 Chronic Kidney Disease: A Case Series' authors: - Elena Moroșan - Violeta Popovici - Viviana Elian - Adriana Maria Dărăban - Andreea Ioana Rusu - Monica Licu - Magdalena Mititelu - Oana Karampelas journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049720 doi: 10.3390/ijerph20065049 license: CC BY 4.0 --- # The Impact of Medical Nutrition Intervention on the Management of Hyperphosphatemia in Hemodialysis Patients with Stage 5 Chronic Kidney Disease: A Case Series ## Abstract The treatment and interdisciplinary management of patients with chronic kidney disease (CKD) continue to improve long-term outcomes. The medical nutrition intervention’s role is to establish a healthy diet plan for kidney protection, reach blood pressure and blood glucose goals, and prevent or delay health problems caused by kidney disease. Our study aims to report the effects of medical nutrition therapy—substituting foods rich in phosphorus-containing additives with ones low in phosphates content on phosphatemia and phosphate binders drug prescription in stage 5 CKD patients with hemodialysis. Thus, 18 adults with high phosphatemia levels (over 5.5 mg/dL) were monitored at a single center. Everyone received standard personalized diets to replace processed foods with phosphorus additives according to their comorbidities and treatment with prosphate binder drugs. Clinical laboratory data, including dialysis protocol, calcemia, and phosphatemia, were evaluated at the beginning of the study, after 30 and 60 days. A food survey was assessed at baseline and after 60 days. The results did not show significant differences between serum phosphate levels between the first and second measurements; thus, the phosphate binders’ initial doses did not change. After 2 months, phosphate levels decreased considerably (from 7.322 mg/dL to 5.368 mg/dL); therefore, phosphate binder doses were diminished. In conclusion, medical nutrition intervention in patients with hemodialysis significantly reduced serum phosphate concentrations after 60 days. Restricting the intake of processed foods containing phosphorus additives—in particularized diets adapted to each patient’s comorbidities—and receiving phosphate binders represented substantial steps to decrease phosphatemia levels. The best results were significantly associated with life expectancy; at the same time, they showed a negative correlation with the dialysis period and participants’ age. ## 1. Introduction According to the World Health Organization, chronic kidney disease (CKD) is a progressive loss of kidney function [1]. The International Organization Kidney Disease: Improving Global Outcomes (KDIGO) established the criteria for CKD as the glomerular filtration rate (GFR) < 60 mL/min per 1.73 m2 for >3 months [2]. The more recent data from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) [3] show that diabetes and high blood pressure are the two most common causes of CKD. The CKD incidence is slightly more common in women ($14\%$) than men ($12\%$). People aged 65 or older are most affected ($38\%$); the ones in the age range of 45 to 64 are the second regarding CKD relevance ($12\%$), followed by those ages between 18 and 44 years ($6\%$) [4]. It has a progressive evolution into kidney failure—GFR < 15 mL/min per 1.73 m2 or treatment by dialysis—affecting more than $10\%$ of people worldwide [5]. Kidney failure—known as the end stage of kidney disease (ESKD), diminishes the quality of life and causes premature death, meaningfully associated with cardiovascular diseases [6]. The treatment consists of kidney replacement therapy (KRT): dialysis (hemodialysis, peritoneal dialysis, hemofiltration, hemodiafiltration), and kidney transplant [7]. On the other hand, CKD treatment and interdisciplinary approaches continue improving long-term patient outcomes [8]. Thus, a CKD patient’s healthcare team must include a registered dietician as an essential medical professional. His or her role is to establish a healthy diet plan for kidney protection, reach blood pressure and blood glucose goals [8], and prevent or delay health problems caused by kidney disease [9]. This food and nutrition specialist performs a medical nutrition therapy (MNT) [10], which can help slow CKD progression, prevent or treat complications and improve quality of life [11]. Therefore, CKD patients can protect their bones and blood vessels by limiting fluids [12], eating a low-protein diet [13,14,15], and diminishing the intake of salt [16], potassium [17], phosphorus [18], and other electrolytes. When the blood phosphate level is high [19], CKD affects calcium-phosphate metabolism and bone homeostasis, inducing osteomalacia, renal osteodystrophy, and vascular and ectopic calcification, resulting in chronic kidney disease mineral bone disorder (CKD MBD) [18,20]. Moreover, various experimental studies highlighted the mechanisms by which phosphate in excess may adversely affect the cardiovascular system [21,22]. Thus, it may directly contribute to vascular damage by pro-inflammatory actions on the vascular smooth muscle cell leading to endothelial dysfunction and promoting vascular calcification [23,24]. Furthermore, a high dietary phosphate content may contribute to atherogenesis [25] and has also been linked to a more rapid progression of CKD to ESKD [19,26,27]. Using data from the US Renal Data System, Block et al. [ 4] found an increased risk of death (relative risk, 1.27) associated with serum phosphate levels > 6.5 mg/dL. In addition, excess phosphates could inhibit the renal transformation of 25(OH) vitamin D to 1.25(OH)2 vitamin D, leading to fibroblast growth factor 23 (FGF23) [28] and parathyroid hormone (PTH) [29] hypersecretion. In current clinical practice, the management of hyperphosphatemia is based on four main strategies: (i) restriction of dietary phosphate intake [30,31]; (ii) reduction of its intestinal absorption; (iii) phosphate removal by dialysis; and (iv) treatment and prevention of renal osteodystrophy. Previous studies have proven that managing eating behavior through MNT determines a favorable evolution of hyperphosphatemia in dialysis patients and a decrease in the treatment dose with phosphate binders. Our study aims to demonstrate the impact of a diet low in processed foods [32] with a low content of inorganic phosphates [33] and phosphate binder drug administration on phosphatemia and implicitly on the prognosis of stage 5 CKD patients with hemodialysis. To achieve this, we used qualitative methods to explore the CKD patients’ adherence to MNT and quantitative ones—corresponding laboratory parameter measurements and analyzing their evolution. ## 2.1. Study Design Our clinical research is conducted as a case series [34,35], a descriptive study of stage 5 CKD patients on hemodialysis [36], included in an MNT program with low inorganic phosphate content for 2 months. The aim is to observe the impact of a low-phosphorus diet associated with phosphate binders’ agents on phosphatemia. The Romanian Agency of Medicines and Medical Devices approved our study by Authorization for Conducting Clinical Studies with Therapeutic Benefit $\frac{171}{03.11.2019.}$ ## 2.2. Descriptive Analysis of the Patients’ Series Patients from a tertiary center in Bucharest with stage 5 CKD treated by hemodialysis and having high serum phosphates levels (>5.5 mg/dL) were included in this MNT program. ## 2.2.1. Inclusion criteria The selected series contains patients of both sexes, aged over >18 years, with stage 5 CKD on hemodialysis treatment for at least 12 months and constant serum phosphorus levels >5.5 mg/dL during the previous 3 months. In addition, they should have good cognitive function, and the ability to read and write is required. ## 2.2.2. Exclusion criteria The following patients have been excluded: those with early CKD and PTH > 1500 pg/mL, with infectious diseases, enteral or parenteral therapy, malabsorption, and cognitive or physical limitations. ## 2.2.3. Details Forty individuals were initially planned, but only eighteen ($45\%$) accepted the protocol and consented to rigorously respect all the study phases. Each participant had three study visits. The intact PTH levels were measured for the first time to assess the inclusion criteria. The monitored parameters before the MNT initiation were established during the first one: Kt/V, calcemia, and phosphatemia. According to their values, the treatment with phosphate binders was prescribed upon guideline recommendations [37]. Pending MNT, each participant completed a survey regarding food preferences and eating behavior. After 30 days, the second evaluation consisted of laboratory findings, measuring the levels of previously mentioned parameters. According to their values, the medication should be adjusted. After 60 days from the beginning, the third evaluation qualitatively investigated the diet and compliance with the nutrition changes through a survey. The levels of all parameters were also measured, and the drug prescription was modified. ## 2.3. Medical Nutrition Intervention All 18 patients included in this study performed the following procedures:Completing the questionnaire regarding food preferences and their consumption frequency at each of the three study visits. They received an instruction to edit a food diary structured for seven days regarding the type of food and liquids ingested, the frequency of consumption, and details about gastroenterology where applicable. The patient had to write down the meals immediately after consumption to reduce the risk of omissions. The food diary was analyzed for each food category, thus obtaining a profile of the patient’s eating habits before the start of the nutrition intervention. From this analysis, the parameters of the frequency of consumption of specific food categories resulted. After analyzing the data obtained in the previously mentioned stages, the nutrition intervention plan was presented to the patients. They received written instructions regarding the allowed and recommended foods and those that should be avoided daily. They were also informed about gastro-technics adapted to the food condition. These aspects were adapted monthly depending on the measurements of specific biochemical parameters. The medical nutrition intervention consisted of the following:The nutritionist recommended three meals daily and two snacks in 2 h, customized according to patients’ schedules and comorbidities. All patients were advised not to consume foods with a high phosphorus, potassium, or sodium content for 4 h before collecting samples for specific analyses. Foods containing additives and mainly inorganic phosphates and which are generally processed: processed and matured dairy products, acidic carbonated drinks, frozen bakery and pastry products, processed meat and sausages, canned fish or meat, frozen doughs, processed sweets, chocolate and products with a high cocoa content, oilseeds, and fruits, dried legumes, were eliminated from the diet. The instructions related to gastro-technics were the following: preparing food mainly by boiling, increasing the contact surface/permeability of the food in the immersion liquid, repeating the boiling process depending on the category of processed food, and avoiding extremely high temperatures in the cooking process. Recommendations consisted of increasing the proportion of foods of natural origin, unprocessed, containing phosphorus, potassium, and organic calcium: fresh meat prepared at home, freshwater fish, fresh products with low-fat content, eggwhite, sweets prepared at home, bread made from white flour prepared at home or with a reduced range of additives, cooked vegetables, and roots, olive oil, and vegetable fats, fruits with a low potassium content. Recommendations were made for consulting food products’ labels and avoiding consuming those with additives and/or preservatives. Personalized advice was focused on water intake according to BCM measurements to avoid hyperhydration. Personalized treatment with phosphate binder drugs was prescribed. The patients received a food diary in which the food consumed during a week was registered and presented at the next visit. To adjust further recommendations, the diet composition was analyzed and corroborated with the measured parameters (Kt/V, phosphatemia, and calcemia). ## 2.4. Clinical Laboratory Analyses Parathyroid hormone—commonly measured once every 3 or 6 months—was evaluated in this first step, aiming to prescribe suitable phosphate binders because it ensures the regulation of the calcium and phosphorus distribution in the body. Intact PTH level was measured using an immunoradiometric assay [38,39,40]. A colorimetric method was performed to assess serum calcium levels using 1,8-Dihydroxy-3,6-disulpho-2,7-naphthalene-bis (azo)-dibenzenearsonic acid (Arsenazo III) at neutral pH, as previously described [41]. Serum phosphorus levels were quantified through spectrophotometry using ammonium molybdate in an acid medium [42]. The Kt/V indicator is investigated monthly in the case of hemodialysis patients. Its variation from one evaluation to another represents the value of the urea clearance (Kt) normalized for the urea distribution volume (V) [43]; more precisely, it is a measure of what dose of dialysis is administered [44], measured through a standard method, previously displayed [45]. ## 2.5. Data Analysis Statistical Package for the Social Sciences (IBM SPSS 20) [46] was used for statistical analysis and graphical representations. Data analysis was carried out, considering all three measurements. Data was provided by the clinic respecting the terms of confidentiality: age, sex, height, weight, dialysis period, vascular access route, PTH, Kt/V level, phosphatemia, calcemia, phosphate binder treatment, and food preferences. Friedman and Wilcoxon tests and the descriptive analysis of the variables (mean, standard deviation, minimum and maximum) were used. We applied the Friedman test because the samples (or variables) are pairs, each subject being tested at least three times (three measurements for Kt/V level, serum phosphate, serum calcium, phosphate binder administration); the dependent variable (DV) is quantitative or ordinal. The Wilcoxon test first calculates the difference between the scores of the two variables for each subject (after, before), establishes the sign for each difference, and finally, ranks the differences in absolute value. The values around zero will be ignored because they do not provide information. Effect size (r) of the Wilcoxon test. The effect size was calculated using Equation [1]:[1]r=z2n where z is the Wilcoxon test result, and n is the number of individuals. The analysis of variables, respectively, the correlation between phosphatemia and all other ones, was performed through principal component analysis, using XLSTAT 2022.2.1. by Addinsoft (New York, NY, USA). ## 3. Results and Discussion Eighteen stage 5 CKD patients on hemodialysis ($50\%$ women and $50\%$ men) were enrolled in the study with an age range of 35–73 years and a mean age of 58.22 ± 10.719 years. Their median height is 170.94 ± 6.584 cm, and their mean dry weight is 75.25 ± 22.446 kg. Their dialysis period belongs to a large domain of 2–15 years, with a mean of 6.33 ± 4159 years (Table 1). The hemodialysis period is essential in our descriptive analysis, influencing patient compliance. Therefore, more receptivity and ability to understand the recommendations were observed in patients with shorter dialysis periods than those with more than 5 years—the resistance to guidance increased due to unwanted trial experiences and low results obtained. A recently published crossectional study [47] reports that of 800 eligible patients, only 497 consented and completed all assessments. The remaining 303 eligible patients refused the agreement and/or were reluctant to research procedures because they participated in many research studies [47] without substantial results. In another clinical trial, of 56 estimated participants, only 32 agreed to participate and completed their interview ($57\%$) [48]. Parathyroid hormone (PTH)—commonly measured once every 3 or 6 months—was evaluated in this first step to suitably prescribe the phosphate binders because it ensures the regulation of the calcium and phosphorus distribution in the body. The PTH level range was 30–1393 pg/mL, with a mean of 371.89 pg/mL and an SD of 366.45. Serum-intact PTH values between 100 and 300 pg/mL do not predict the degree of bone turnover in dialysis patients (8 of the 18 patients have PTH values > 300 pg/mL). The patients enrolled in the study also present the following comorbidities: diabetes type I [3], diabetes type II [2], arterial hypertension stage II [6], chronic HCV hepatitis with cirrhosis [2], heart failure [3], and neoplasms [2]. The dietary recommendations considered all these associated conditions. The access path for dialysis is an arteriovenous fistula (AFV, $67\%$ of patients) and a central venous catheter (CVC) for the remaining $33\%$. Most patients have AVF accesses over 12 weeks old with few reparative interventions, making them superior to CVC vascular prostheses. The initial food survey displays the patients’ food preferences, marked in the increasing order of consumption frequency. Thus, the following notations are available: 1 = very rarely—once per month, 2 = occasionally—once per week, 3 = rarely—once to every few days, 4 = often—once per day, and 5 = very often—several times per day (Table 2). The parameters with three measurements are displayed in Table 3. The Kt/V values do not report a significant variation from month to month, which shows us, depending on the mean value, that the dialysis dose had a slight decrease. The evolution of the calcemia registers an increase in the mean value in the second evaluation, reaching the value of 9.23 mg/dL. The third evaluation reaches approximately the mean value of the first one. The mean values recorded in all measurements fall within the recommended reference range of 8.5 mg/dL–10.2 mg/dL. The results were statistically analyzed using Friedman and Wilcoxon tests and presented in Tables S1 and S2 from the Supplementary Material. After 2 months, the ordered foods ingested by patients with increased frequency changed compared to the initial survey. Foods with significant phosphorus intake rank lower, indicating that the nutrition recommendations have been followed (Table 2). Moreover, some foods with high phosphate content (processed dried vegetables and spice blends, canned meat, pasta, oil seeds (nuts, almonds, peanuts), canned fish, and whole grains) were substituted by other ones more suitable according to MNT (white and red meat boiled, potatoes, eggwhite, rice, cream cheese). The phosphatemia did not have significantly different values in the first month of MNT. The favorable evolution of the serum phosphate level following the medical nutrition intervention was obtained after 2 months for 16 of the 18 patients. The nonparametric tests Friedman and Wilcoxon (Tables S3 and S4 from the Supplementary Material) reported significant differences between all three measurements (Friedman test results show χ2 = 16.778, $p \leq 0.001$ for 18 patients). Negative ranks are found for 16 of the 18 subjects; this means a decrease in the level of phosphate measured in the third month compared to the measurement in the first month on 16 patients. The two positive ranks show an increase in the serum phosphate level compared to the first month in two patients (Tables S2 and S3 from the Supplementary Material). The positive correlation between the phosphatemia level and restricted foods’ frequency of consumption, dialysis period, age, calcemia, and Kt/V is displayed in Figure 1. Thus, before MNT, at the first evaluation (Figure 1A), phosphatemia is substantially correlated with the dialysis period ($r = 0.999$, $p \leq 0.05$) and PTH value ($r = 0.998$, $p \leq 0.05$). It is strongly positively correlated with the age, Kt/V, and restricted foods frequency of consumption r = [0.878–0.994], $p \leq 0.05$ (Figure 1A and Figure S1 from Supplementary Materials). On the final evaluation, after 2 months of MNT, the high positive correlation between phosphatemia and the other parameters (dialysis period, consumption frequency of cheese/yogurt, snacks, fast food, chocolate, pastry, processed dried legumes) is more statistically significant: $r = 0.999$; $p \leq 0.05$ (Figure 1B). Moreover, all previous observations were supported by the places of different phosphate levels in the correlations biplots from Figure 1 (A and B). A general PCA correlation biplot between variable parameters in both phases is displayed in Figure 2. The two principal components explain $96.92\%$ of total data variance, with $88.21\%$ attributed to the first (PC1) and $8.71\%$ to the second (PC2). PC1 is associated with all variables. Phosphates level also reports a good correlation with the frequency of consumption of processed cheeses and snacks ($r = 0.799$–0.805, $p \leq 0.05$). Available data of the entire study show a statistically significant high positive correlation between phosphatemia and the greatest part of the parameter evaluated (intake frequency of various restricted foods, age, and dialysis period): r = [0.883–0.967], $p \leq 0.05$ (PCA Correlation Matrix from the Supplementary Material). De Fornasari et al. obtained similar results in 67 participants monitored for 90 days [49]. The phosphatemia level significantly decreased (from 7.2 to 5.0 mg/dL); the reported values are similar to those in Table 3. Calcemia and Kt/V values reported insignificant changes in the final evaluation. Two phosphate binder drugs and one food supplement were prescribed to decrease serum phosphate levels (Table 4). Renagel 800 mg (Sanofi, Berkshire, UK) contains Sevelamer hydrochloride [50]. OsvaRen $\frac{435}{235}$mg (Vifor Fresenius Medical Care Renal Pharma UK Ltd.) has calcium acetate 435 mg, equivalent to 110 mg of calcium, and magnesium carbonate, heavy 235 mg, equivalent to 60 mg of magnesium [51,52]. Prodial (Bioeel®, Targu Mures, Romania) is a food supplement with a calcium/magnesium composition similar to OsvaRen [53]. For 60 days, the phosphate binder treatment did not change. Thus, for two months, three patients followed treatment with OsvaRen, six tablets daily, eleven patients with Prodial, also six tablets daily, and four received Renagel, six tablets daily (Table 4). In the final evaluation, the MNT intervention showed its efficacy by adjusting the phosphate-chelator treatment with lower doses (from six tablets per day to five, four, and three tablets per day). There were significant differences between the recommended phosphate binder doses in the three months [χ2 = 26; $p \leq 0.001$], according to Friedman and Wilcoxon tests (Tables S5 and S6 from the Supplementary Material). For 13 patients, the phosphate-chelator dose was decreased in the third month (13 negative ranks). For five, the doses remained the same, so the result of the Wilcoxon test (Table S6 from the Supplementary Material) shows a significant difference between both recommended doses (the third evaluation versus the first one). One-third of monitored patients ($33\%$) had their daily dose decreased from six tablets of Prodial per day to five tablets per day. The most significant dose diminution was recorded in $11\%$ of patients, from six Prodial tablets per day to three tablets (Table 4). Data analysis allows for a global image regarding compliance with the MNT of participants. This aspect varies inversely proportionally to the age of individuals and the dialysis period; a reason supporting this observation could be a loss of health-being hope due to a long period of failed trials and associated comorbidities. Numerous clinical studies reported a low adherence of participants to phosphate binders treatment [54,55,56]. Khor et al. showed that phosphate binders—mainly calcium-based ($92.9\%$)–were prescribed to $98.2\%$ of patients, but only $59.5\%$ adhered to the phosphate binder drug prescription [47]. Another research team highlighted the effect of individual health education—conducted by a physician with short-term nutrition training—on hyperphosphatemia and adherence to MNT [57]. Moreover, a recent clinical trial reported that management of hyperphosphatemia can be achieved through a mobile application (similar to dietitian management), with the additional benefit of titrating phosphate binder doses according to individual meal choice [58]. Performing a multicenter cross-sectional study, Kurita et al. [ 28] highlighted the positive relationship of the CKD stage with health expectation, evidencing hope’s essential role in the psychological and physiological manifestations of adherence to the treatment. On the other hand, a recent qualitative study including patients with CKD in stages 2–4, conducted by Rivera et al. [ 48], reported multiple factors correlated with treatment adherence. First, patient factors, which include numerous comorbidities, chronic diseases, motivation, and outlook. Second, healthcare team factors: attentiveness, care, availability/accessibility, empathy, and communication style. Third, treatment planning factors: lack of a clear plan, proactive and dynamic research, provider-focused treatment objectives, and shared decision-making. As a result, the response of a CKD patient could be very different: positive feedback, lack of information, perceived capability deficit, and disagreement with treatment [48]. The present study was conceived as a case series (also known as clinical series) [35], one of the most common types of medical study designs that describe the experience of a small group of people [59]. Because it did not have a control group or any form of randomization employed, our clinical study could be considered relatively efficient and cost-saving [60]. This study’s results are closer to those obtained in routine clinical practice and may be more relevant than a randomized trial [61]. The external validity [36] is also appreciable in our case series because we included a diverse patient range (with different ages, comorbidities, and other particularities). Thus, our results could be applied to clinical practice in other centers. However, the lack of a control group is the main limitation of our study; thus, we cannot establish whether the outcomes are exclusively the effect of MNT or whether other variables cause them. Moreover, our clinical series was not a randomized or double-blind study; therefore, a rigorously established protocol is missing, and data collection is incomplete, leading to bias concerns. For example, the amount of phosphate intake could not be quantified because the patients prepared their meals at home, and the foods had different provenance; hence, it was impossible to do their biochemical analysis. Finally, this short-term clinical study lasted only 2 months. All these previously mentioned aspects limit our study’s generalizability. Further research will confirm these results with extensive groups of patients in more extended periods, following suitable protocols. ## 4. Conclusions Following a low phosphorus diet during medical nutrition intervention, our stage 5 CKD patients on hemodialysis revealed good outcomes: their serum phosphorus levels decreased, thus diminishing the received phosphate binders’ doses after 60 days. Our study shows that restricting the intake of processed foods containing phosphate additives—by individualized diets according to each patient’s comorbidities—associated with phosphate binder drug administration is essential to decrease phosphatemia levels. The best results were positively correlated with health-related expectancy and negatively correlated with the dialysis period and the age of participants. ## References 1. World Health Organisation, Chronic Kidney Disease 2. Levey A.S., Eckardt K.U., Dorman N.M., Christiansen S.L., Hoorn E.J., Ingelfinger J.R., Inker L.A., Levin A., Mehrotra R., Palevsky P.M.. Nomenclature for kidney function and disease: Report of a Kidney Disease: Improving Global Outcomes (KDIGO) Consensus Conference. Kidney Int. (2020.0) 97 1117-1129. DOI: 10.1016/j.kint.2020.02.010 3. National Institute of Diabetes and Digestive and Kidney Diseases 4. Barreto F.C., Barreto D.V., Massy Z.A., Drüeke T.B.. Strategies for Phosphate Control in Patients with CKD. Kidney Int. Rep. (2019.0) 4 1043-1056. DOI: 10.1016/j.ekir.2019.06.002 5. Kovesdy C.P.. Epidemiology of chronic kidney disease: An update 2022. Kidney Int. Suppl. (2022.0) 12 7-11. DOI: 10.1016/j.kisu.2021.11.003 6. Sarnak M.J., Amann K., Bangalore S., Cavalcante J.L., Charytan D.M., Craig J.C., Gill J.S., Hlatky M.A., Jardine A.G., Landmesser U.. Chronic Kidney Disease and Coronary Artery Disease: JACC State-of-the-Art Review. J. Am. Coll. Cardiol. (2019.0) 74 1823-1838. DOI: 10.1016/j.jacc.2019.08.1017 7. Lameire N.H., Levin A., Kellum J.A., Cheung M., Jadoul M., Winkelmayer W.C., Stevens P.E., Caskey F.J., Farmer C.K.T., Ferreiro Fuentes A.. Harmonizing acute and chronic kidney disease definition and classification: Report of a Kidney Disease: Improving Global Outcomes (KDIGO) Consensus Conference. Kidney Int. (2021.0) 100 516-526. DOI: 10.1016/j.kint.2021.06.028 8. Jassim R.A., Mihele D., Dogaru E.. Study regarding the influence of Vitis vinifera fruit (Muscat of Hamburg species) on some biochemical parameters. Farmacia (2010.0) 58 332-340 9. Kistler B.M., Moore L.W., Benner D., Biruete A., Boaz M., Brunori G., Chen J., Drechsler C., Guebre-Egziabher F., Hensley M.K.. The International Society of Renal Nutrition and Metabolism Commentary on the National Kidney Foundation and Academy of Nutrition and Dietetics KDOQI Clinical Practice Guideline for Nutrition in Chronic Kidney Disease. J. Ren. Nutr. (2021.0) 31 116-120.e1. DOI: 10.1053/j.jrn.2020.05.002 10. de Waal D., Heaslip E., Callas P.. Medical Nutrition Therapy for Chronic Kidney Disease Improves Biomarkers and Slows Time to Dialysis. J. Ren. Nutr. (2016.0) 26 1-9. DOI: 10.1053/j.jrn.2015.08.002 11. Ikizler T.A., Burrowes J.D., Byham-Gray L.D., Campbell K.L., Carrero J.J., Chan W., Fouque D., Friedman A.N., Ghaddar S., Goldstein-Fuchs D.J.. KDOQI Clinical Practice Guideline for Nutrition in CKD: 2020 Update. Am. J. Kidney Dis. (2020.0) 76 S1-S107. DOI: 10.1053/j.ajkd.2020.05.006 12. Palmer S.C., Hanson C.S., Craig J.C., Strippoli G.F.M., Ruospo M., Campbell K., Johnson D.W., Tong A.. Dietary and fluid restrictions in CKD: A thematic synthesis of patient views from qualitative studies. Am. J. Kidney Dis. (2015.0) 65 559-573. DOI: 10.1053/j.ajkd.2014.09.012 13. Kalantar-Zadeh K., Rhee C.M., Joshi S., Brown-Tortorici A., Kramer H.M.. Medical nutrition therapy using plant-focused low-protein meal plans for management of chronic kidney disease in diabetes. Curr. Opin. Nephrol. Hypertens. (2022.0) 31 26-35. DOI: 10.1097/MNH.0000000000000761 14. Foucan L., Merault H., Velayoudom-Cephise F.L., Larifla L., Alecu C., Ducros J.. Impact of protein energy wasting status on survival among Afro-Caribbean hemodialysis patients: A 3-year prospective study. Springerplus (2015.0) 4 1-10. DOI: 10.1186/s40064-015-1257-3 15. Kim H., Caulfield L.E., Garcia-Larsen V., Steffen L.M., Grams M.E., Coresh J., Rebholz C.M.. Plant-based diets and incident CKD and kidney function. Clin. J. Am. Soc. Nephrol. (2019.0) 14 682-691. DOI: 10.2215/CJN.12391018 16. Iseki K.. Nutrition and quality of life in chronic kidney disease patients: A practical approach for salt restriction. Kidney Res. Clin. Pract. (2022.0) 41 657-669. DOI: 10.23876/j.krcp.21.203 17. Picard K., Barreto Silva M.I., Mager D., Richard C.. Dietary potassium intake and risk of chronic kidney disease progression in predialysis patients with chronic kidney disease: A systematic review. Adv. Nutr. (2020.0) 11 1002-1015. DOI: 10.1093/advances/nmaa027 18. Cannata-Andía J.B., Martin K.J.. The challenge of controlling phosphorus in chronic kidney disease. Nephrol. Dial. Transplant. (2016.0) 31 541-547. DOI: 10.1093/ndt/gfv055 19. O’Seaghdha C.M., Hwang S.J., Muntner P., Melamed M.L., Fox C.S.. Serum phosphorus predicts incident chronic kidney disease and end-stage renal disease. Nephrol. Dial. Transplant. (2011.0) 26 2885-2890. DOI: 10.1093/ndt/gfq808 20. Shah A.. Phosphorus and other aspects of CKD-MBD in the conservative management of chronic kidney disease. Panminerva Med. (2017.0) 59 124-132. DOI: 10.23736/S0031-0808.17.03302-X 21. Gutiérrez O.M.. Recent Advances in the Role of Diet in Bone and Mineral Disorders in Chronic Kidney Disease. Curr. Osteoporos. Rep. (2021.0) 19 574-579. DOI: 10.1007/s11914-021-00710-x 22. Disthabanchong S.. Phosphate and Cardiovascular Disease beyond Chronic Kidney Disease and Vascular Calcification. Int. J. Nephrol. (2018.0) 2018 3162806. DOI: 10.1155/2018/3162806 23. Lestariningsih L., Hardisaputro S., Nurani A., Santosa D., Santoso G.. High sensitivity C-reactive protein as a predictive of intima-media thickness in patients with end-stage renal disease on regular hemodialysis. Kidney Int. Rep. (2020.0) 5 S107. DOI: 10.1016/j.ekir.2020.02.267 24. Miura S., Yoshihisa A., Takiguchi M., Shimizu T., Nakamura Y., Yamauchi H., Iwaya S., Owada T., Miyata M., Abe S.. Association of Hypocalcemia with Mortality in Hospitalized Patients with Heart Failure and Chronic Kidney Disease. J. Card. Fail. (2015.0) 21 621-627. DOI: 10.1016/j.cardfail.2015.04.015 25. Hase H.. Atherosclerotic heart disease in patients with advanced chronic kidney disease. J. Med. Soc. Toho Univ. (2019.0) 66 22-26 26. Doshi S.M., Wish J.B.. Past, Present, and Future of Phosphate Management. Kidney Int. Rep. (2022.0) 7 688-698. DOI: 10.1016/j.ekir.2022.01.1055 27. Chang W.X., Xu N., Kumagai T., Shiraishi T., Kikuyama T., Omizo H., Sakai K., Arai S., Tamura Y., Ota T.. The impact of normal range of serum phosphorus on the incidence of end-stage renal disease by a propensity score analysis. PLoS ONE (2016.0) 11. DOI: 10.1371/journal.pone.0154469 28. Wolf M.. Update on fibroblast growth factor 23 in chronic kidney disease. Kidney Int. (2012.0) 82 737-747. DOI: 10.1038/ki.2012.176 29. Ureña-Torres P.A., Vervloet M., Mazzaferro S., Oury F., Brandenburg V., Bover J., Cavalier E., Cohen-Solal M., Covic A., Drüeke T.B.. Novel insights into parathyroid hormone: Report of The Parathyroid Day in Chronic Kidney Disease. Clin. Kidney J. (2018.0) 12 269-280. DOI: 10.1093/ckj/sfy061 30. High_Phosphorus_Foods.pdf 31. Low_Phosphorus_Foods.pdf 32. Nishime K., Sugiyama N., Okada K.. A new disease concept in the age of processed foods—Phosphorus-burden disease; CKD–MBD concrete analysis and the way to solution. Nutrients (2021.0) 13. DOI: 10.3390/nu13082874 33. Tuominen M., Karp H.J., Itkonen S.T.. Phosphorus-Containing Food Additives in the Food Supply—An Audit of Products on Supermarket Shelves. J. Ren. Nutr. (2022.0) 32 30-38. DOI: 10.1053/j.jrn.2021.07.010 34. Abu-Zidan F.M., Abbas A.K., Hefny A.F.. Clinical “case series”: A concept analysis. Afr. Health Sci. (2012.0) 12 557-562. DOI: 10.4314/ahs.v12i4.25 35. Carey T.S., Boden S.D.. A critical guide to case series reports. Spine (2003.0) 28 1631-1634. DOI: 10.1097/01.BRS.0000083174.84050.E5 36. Kooistra B., Dijkman B., Einhorn T.A., Bhandari M.. How to design a good case series. J. Bone Jt. Surg. (2009.0) 91 21-26. DOI: 10.2106/JBJS.H.01573 37. Mason D.L., Godugu K., Nnani D., Mousa S.A.. Effects of sevelamer carbonate versus calcium acetate on vascular calcification, inflammation, and endothelial dysfunction in chronic kidney disease. Clin. Transl. Sci. (2022.0) 15 353-360. DOI: 10.1111/cts.13151 38. Kritmetapak K., Pongchaiyakul C.. Parathyroid Hormone Measurement in Chronic Kidney Disease: From Basics to Clinical Implications. Int. J. Nephrol. (2019.0) 2019 5496710. DOI: 10.1155/2019/5496710 39. Cavalier E., Vasikaran S., Bhattoa H.P., Heijboer A.C., Makris K., Ulmer C.Z.. The path to the standardization of PTH: Is this a realistic possibility? A position paper of the IFCC C-BM. Clin. Chim. Acta (2021.0) 515 44-51. DOI: 10.1016/j.cca.2020.12.022 40. Parathyroid Hormone (PTH), Intact—Test 41. Shin D.H., Yoon K.P., Kwon S.K., Kang C.G.. Determination of Total Calcium in Serum with Arsenazo III Method. Ann. Lab. Med. (1994.0) 14 12-19 42. Connell E.. Tietz Textbook of Clinical Chemistry and Molecular Diagnostics (5th edn), Saunders/Elsevier, Philadelphia, PA, USA. Ann. Clin. Biochem. Int. J. Lab. Med. (2012.0) 49 615. DOI: 10.1258/acb.2012.201217 43. Locatelli F., Buoncristiani U., Canaud B., Köhler H., Petitclerc T., Zucchelli P.. Dialysis dose and frequency. Nephrol. Dial. Transplant. (2005.0) 20 285-296. DOI: 10.1093/ndt/gfh550 44. Deira J., Murea M., Kalantar-Zadeh K., Casino F.G., Basile C.. Does delivering more dialysis improve clinical outcomes? What randomized controlled trials have shown. J. Nephrol. (2022.0) 35 1315-1327. DOI: 10.1007/s40620-022-01246-8 45. Ahrenholz P., Taborsky P., Bohling M., Rawer P., Ibrahim N., Gajdos M., MacHek P., Sagova M., Gruber H., Moucka P.. Determination of dialysis dose: A clinical comparison of methods. Blood Purif. (2011.0) 32 271-277. DOI: 10.1159/000330340 46. Riklikienė O., Spirgienė L., Kaselienė S., Luneckaitė Ž., Tomkevičiūtė J., Büssing A.. Translation, cultural, and clinical validation of the lithuanian version of the spiritual needs questionnaire among hospitalized cancer patients. Medicina (2019.0) 55. DOI: 10.3390/medicina55110738 47. Khor B.H., Sualeheen A., Sahathevan S., Chinna K., Gafor A.H.A., Bavanandan S., Goh B.L., Morad Z., Daud Z.A.M., Khosla P.. Association of dietary patterns with serum phosphorus in maintenance haemodialysis patients: A cross-sectional study. Sci. Rep. (2020.0) 10 12278. DOI: 10.1038/s41598-020-68893-4 48. Rivera E., Clark-Cutaia M.N., Schrauben S.J., Townsend R.R., Lash J.P., Hannan M., Jaar B.G., Rincon-Choles H., Kansal S., He J.. Treatment Adherence in CKD and Support From Health care Providers: A Qualitative Study. Kidney Med. (2022.0) 4 100545. DOI: 10.1016/j.xkme.2022.100545 49. de Fornasari M.L.L., dos Santos Sens Y.A.. Replacing Phosphorus-Containing Food Additives With Foods Without Additives Reduces Phosphatemia in End-Stage Renal Disease Patients: A Randomized Clinical Trial. J. Ren. Nutr. (2017.0) 27 97-105. DOI: 10.1053/j.jrn.2016.08.009 50. Suki W.N., Zabaneh R., Cangiano J.L., Reed J., Fischer D., Garrett L., Ling B.N., Chasan-Taber S., Dillon M.A., Blair A.T.. Effects of sevelamer and calcium-based phosphate binders on mortality in hemodialysis patients. Kidney Int. (2007.0) 72 1130-1137. DOI: 10.1038/sj.ki.5002466 51. Electronic Medicines Compendium, Osvaren 435 mg/235 mg Film-Coated Tablets 52. Plagemann T., Prenzler A., Mittendorf T.. Considerations about the effectiveness and cost effectiveness of therapies in the treatment of hyperphosphataemia. Health Econ. Rev. (2011.0) 1 1. DOI: 10.1186/2191-1991-1-1 53. Prodial in Catalogul de Produse BIOEEL 54. Guo H., Sriyuktasuth A., Pongthavornkamol K.. Factors Associated with Phosphate Binder Adherence among Patients Undergoing Hemodialysis in China. Nurs. Sci. J. Thail. (2022.0) 40 123-137 55. Hjemås B.J., Bøvre K., Mathiesen L., Lindstrøm J.C., Bjerknes K.. Interventional study to improve adherence to phosphate binder treatment in dialysis patients. BMC Nephrol. (2019.0) 20. DOI: 10.1186/s12882-019-1334-x 56. Karamanidou C., Clatworthy J., Weinman J., Horne R.. A systematic review of the prevalence and determinants of nonadherence to phosphate binding medication in patients with end-stage renal disease. BMC Nephrol. (2008.0). DOI: 10.1186/1471-2369-9-2 57. Chen Y., Li Z., Liang X., Zhang M., Zhang Y., Xu L., Zhong L., Shi W.. Effect of individual health education on hyperphosphatemia in the Hakkas residential area. Ren. Fail. (2015.0) 37 1303-1307. DOI: 10.3109/0886022X.2015.1073072 58. Teong L.F., Khor B.H., Ng H.M., Sahathevan S., Purba K.R., Narayanan S.S., Gafor A.H.A., Goh B.L., Bee B.C., Yahya R.. Effectiveness of a Nutritional Mobile Application for Management of Hyperphosphatemia in Patients on Hemodialysis: A Multicenter Open-Label Randomized Clinical Trial. J. Pers. Med. (2022.0) 12. DOI: 10.3390/jpm12060961 59. El-Gilany A.-H.. What is case series?. Asploro J. Biomed. Clin. Case Rep. (2018.0) 1 10-15 60. Hartz A., Marsh J.L.. Methodologic Issues in Observational Studies. Clin. Orthop. Relat. Res. (2003.0) 413 33-42. DOI: 10.1097/01.blo.0000079325.41006.95 61. Slim K., Nini E., Forestier D., Kwiatkowski F., Panis Y., Chipponi J.. Methodological index for non-randomized studies (Minors): Development and validation of a new instrument. ANZ J. Surg. (2003.0) 73 712-716. DOI: 10.1046/j.1445-2197.2003.02748.x
--- title: First-Trimester Screening for HELLP Syndrome—Prediction Model Based on MicroRNA Biomarkers and Maternal Clinical Characteristics authors: - Ilona Hromadnikova - Katerina Kotlabova - Ladislav Krofta journal: International Journal of Molecular Sciences year: 2023 pmcid: PMC10049724 doi: 10.3390/ijms24065177 license: CC BY 4.0 --- # First-Trimester Screening for HELLP Syndrome—Prediction Model Based on MicroRNA Biomarkers and Maternal Clinical Characteristics ## Abstract We evaluated the potential of cardiovascular-disease-associated microRNAs for early prediction of HELLP (hemolysis, elevated liver enzymes, and low platelets) syndrome. Gene expression profiling of 29 microRNAs was performed on whole peripheral venous blood samples collected between 10 and 13 weeks of gestation using real-time RT-PCR. The retrospective study involved singleton pregnancies of Caucasian descent only diagnosed with HELLP syndrome ($$n = 14$$) and 80 normal-term pregnancies. Upregulation of six microRNAs (miR-1-3p, miR-17-5p, miR-143-3p, miR-146a-5p, miR-181a-5p, and miR-499a-5p) was observed in pregnancies destined to develop HELLP syndrome. The combination of all six microRNAs showed a relatively high accuracy for the early identification of pregnancies destined to develop HELLP syndrome (AUC 0.903, $p \leq 0.001$, $78.57\%$ sensitivity, $93.75\%$ specificity, cut-off > 0.1622). It revealed $78.57\%$ of HELLP pregnancies at a $10.0\%$ false-positive rate (FPR). The predictive model for HELLP syndrome based on whole peripheral venous blood microRNA biomarkers was further extended to maternal clinical characteristics, most of which were identified as risk factors for the development of HELLP syndrome (maternal age and BMI values at early stages of gestation, the presence of any kind of autoimmune disease, the necessity to undergo an infertility treatment by assisted reproductive technology, a history of HELLP syndrome and/or pre-eclampsia in a previous gestation, and the presence of trombophilic gene mutations). Then, $85.71\%$ of cases were identified at a $10.0\%$ FPR. When another clinical variable (the positivity of the first-trimester screening for pre-eclampsia and/or fetal growth restriction by the Fetal Medicine Foundation algorithm) was implemented in the HELLP prediction model, the predictive power was increased further to $92.86\%$ at a $10.0\%$ FPR. The model based on the combination of selected cardiovascular-disease-associated microRNAs and maternal clinical characteristics has a very high predictive potential for HELLP syndrome and may be implemented in routine first-trimester screening programs. ## 1. Introduction HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets) is a very rare pregnancy-related disorder with a general prevalence ranging from $0.2\%$ to $0.9\%$ [1,2,3,4,5,6,7,8]. It occurs separately or in an association with severe pre-eclampsia (PE), where the incidence increases to 4–$24\%$ [2,6,9,10,11,12,13]. High maternal and fetal morbidity and mortality were reported due to the appearance of severe maternal and neonatal complications [1,9,10,14,15,16,17,18,19,20,21,22,23,24,25,26]. The diagnosis and the severity of HELLP syndrome are evaluated using two classification systems: the Tennessee classification and the Mississippi classification [20,27,28,29,30,31]. Both classification systems are based on the number of platelets in the blood (PLT), the serum levels of aspartate (AST) or alanine (ALT) aminotransferases and lactate dehydrogenase (LDH), and signs of hemolysis in a peripheral blood smear [20,27,28,29,30,31]. In the Tennessee classification, the platelet count is usually below 100 × 109/L, serum AST levels are usually above 70 IU/L, and serum LDH levels are usually above 600 IU/L. While complete (full) HELLP syndrome requires the fulfilment of all diagnostic criteria, in incomplete (partial) HELLP syndrome only one or two of the abovementioned criteria are diagnosed [27,28,29]. In the Mississippi classification, the most severe HELLP syndrome (also termed class 1) is represented by a platelet count below 50 × 109/L, moderate HELLP syndrome (class 2) by a blood platelet number in the range of 50 × 109/L to 100 × 109/L, and the mild form of HELLP syndrome (class 3) by a platelet count between 100 × 109/L and 150 × 109/L. Class 1 and class 2 HELLP syndrome are further defined by serum AST or ALT levels above 70 IU/L, while class 3 HELLP syndrome is defined by levels above 40 IU/L. In all cases, serum LDH levels are above 600 IU/L [20,30,31]. Several risk factors predisposing to the development of HELLP syndrome have already been identified. These factors are Caucasian ethnicity, nulliparity, history of gestational hypertensive disorders or HELLP syndrome in a previous gestation, multiple pregnancy at an advanced maternal age, and increased levels of mean arterial pressure (MAP) assessed in first-trimester screening [32,33,34,35,36,37]. The pathophysiology of HELLP syndrome has not yet been fully discovered. However, the activation of endothelial cells, increased production of antiangiogenic factors and microvascular injury resulting in microangiopathic anemia, periportal hepatic necrosis, and thrombocytopenia are probably induced by the ischemia occurring in placental tissue [8,37,38,39,40,41,42,43,44,45,46]. Up to now, only a few models for the prediction of HELLP syndrome have been developed. A logistic regression model (LRM) based on racial origin, nulliparity, history of HELLP syndrome, and PE showed an area under the curve (AUC) of 0.800, $75.0\%$ sensitivity, and $79.0\%$ specificity. The detection rate, with a $10.0\%$ false-positive rate (FPR), reached $55.0\%$ of cases [36]. Recently, a neuro-fuzzy model for the identification and prediction of HELLP syndrome has been developed. This novel model reached an AUC of 0.829 and a precision level of 0.685, but only seven pregnant women with HELLP syndrome were included in the study [47]. Unfortunately, the maternal clinical variables involved in this model were not stated. The aim of our study was to develop an efficient early predictive model for HELLP syndrome that could be implemented in the current algorithms of first-trimester screening. Since HELLP syndrome usually develops in the third trimester of gestation, the availability of an early predictive model for HELLP syndrome is desirable. Initially, we focused on the identification of risk factors associated with the later development of HELLP syndrome occurring separately or accompanying severe PE. Afterwards, we were interested to see whether there was any other potential way to improve the detection rate of our novel HELLP predictive model based on maternal clinical characteristics only. Recently, we reported an altered expression profile of microRNAs associated with the cardiovascular system in pregnant women affected with chronic hypertension [48] and in normotensive early pregnancies with subsequent onset of gestational hypertension (GH) [48], PE [48], fetal growth restriction (FGR) [49], small for gestational age (SGA) [49], preterm delivery [50], and/or gestational diabetes mellitus (GDM) [51]. Therefore, we were interested to determine whether an altered expression profile of microRNAs associated with the cardiovascular system might also be present in pregnancies developing HELLP syndrome. We performed at early gestational stages whole peripheral blood gene expression profiling of 29 selected microRNAs demonstrated previously to play a crucial role in the development and maintenance of homeostasis in the cardiovascular system and in the pathophysiology of cardiovascular and cerebrovascular diseases (Table 1) [51]. Finally, we examined the reliability of our novel early predictive model for HELLP syndrome based on the cardiovascular-disease-associated microRNA biomarkers combined with maternal clinical characteristics identified in our study as independent risk factors for the development of HELLP syndrome. ## 2.1. Identification of Risk Factors for the Development of HELLP Syndrome The clinical characteristics of cases (HELLP-syndrome pregnancies) and controls (normal-term pregnancies) are outlined in Table 2. The following independent risk factors for the development of HELLP syndrome with or without PE were identified at early gestational stages: the presence of any autoimmune disease, an infertility treatment by assisted reproductive technologies, the occurrence of HELLP syndrome and/or PE in a previous gestation, the presence of mutations in trombophilic genes, and the positivity of first-trimester PE/FGR screening by the Fetal Medicine Foundation (FMF) algorithm. ## 2.2. Altered Expression Profiles of MicroRNAs during the First Trimester of Gestation in Pregnancies Developing HELLP Syndrome Whole peripheral blood first-trimester expression profiles of microRNAs were compared between pregnancies that developed HELLP-syndrome and normal-term pregnancies. Increased levels of miR-1-3p ($p \leq 0.001$), miR-17-5p ($$p \leq 0.010$$), miR-143-3p ($$p \leq 0.011$$), miR-146a-5p ($p \leq 0.001$), miR-181a-5p ($$p \leq 0.001$$), and miR-499a-5p ($p \leq 0.001$) were detected during the first trimester of gestation in pregnancies developing HELLP syndrome (Figure 1). Individual microRNAs differentiated between normal-term pregnancies and pregnancies developing HELLP syndrome with various sensitivities at a $10.0\%$ FPR. MiR-1-3p ($64.29\%$) and miR-499a-5p ($50.0\%$) showed the best sensitivities, miR-146a-5p ($35.71\%$) and miR-181a-5p ($28.57\%$) showed moderate sensitivities, and miR-17-5p ($14.29\%$) and miR-143-3p ($14.29\%$) showed the lowest sensitivities (Figure 1). ## 2.3. The Prediction Model for HELLP Syndrome—The Combination of Six MicroRNAs Only The prediction model for HELLP syndrome based on the combination of six microRNAs only (miR-1-3p, miR-17a-5p, miR-143a-3p, miR-146a-5p, miR-181a-5p, and miR-499a-5p) identified pregnancies developing HELLP syndrome with relatively high accuracy (AUC 0.903, $p \leq 0.001$, $78.57\%$ sensitivity, $93.75\%$ specificity, cut-off > 0.1622). A total of $78.57\%$ of pregnancies destined to develop HELLP syndrome was revealed at early stages of gestation with a $10.0\%$ FPR (Figure 2). ## 2.4. The Prediction Model for HELLP Syndrome Based on Selected Maternal Clinical Characteristics Only The prediction model for HELLP syndrome based on the combination of six selected maternal clinical characteristics only (maternal age and BMI values at early gestational stages, the presence of any autoimmune disease, an infertility treatment by assisted reproductive technologies, the occurrence of HELLP syndrome and/or PE in a previous gestation, and the presence of mutations in trombophilic genes) identified pregnancies developing HELLP syndrome with a relatively high accuracy (AUC 0.862, $p \leq 0.001$, $71.43\%$ sensitivity, $97.50\%$ specificity, cut-off >0.1244). A total of $71.43\%$ of pregnancies destined to develop HELLP syndrome was revealed at early gestational stages with a $10.0\%$ FPR. The addition of another maternal clinical characteristic to the prediction model (the positivity of first-trimester PE/FGR screening by FMF) also revealed $71.43\%$ of HELLP pregnancies with a $10.0\%$ FPR (AUC 0.849, $p \leq 0.001$, $71.43\%$ sensitivity, $97.50\%$ specificity, cut-off > 0.1406) (Figure 3). ## 2.5. The Full Prediction Model for HELLP Syndrome Based on the Combination of Six MicroRNAs and Selected Maternal Clinical Characteristics The full prediction model for HELLP syndrome based on the combination of six microRNAs (miR-1-3p, miR-17a-5p, miR-143a-3p, miR-146a-5p, miR-181a-5p, and miR-499a-5p) and six selected maternal clinical characteristics (maternal age and BMI values at early gestational stages, the presence of any autoimmune disease, an infertility treatment by assisted reproductive technologies, the occurrence of HELLP syndrome and/or PE in a previous gestation, and the presence of mutations in trombophilic genes) identified pregnancies developing HELLP syndrome with relatively high accuracy (AUC 0.979, $p \leq 0.001$, $100.0\%$ sensitivity, $86.25\%$ specificity, cut-off > 0.0494). A total of $85.71\%$ of pregnancies destined to develop HELLP syndrome was revealed at early gestational stages with a $10.0\%$ FPR. The addition of another maternal clinical characteristic to the prediction model (the positivity of first-trimester PE/FGR screening by FMF) revealed $92.86\%$ of HELLP pregnancies at a $10.0\%$ FPR (AUC 0.975, $p \leq 0.001$, $92.86\%$ sensitivity, $92.50\%$ specificity, cut-off > 0.1110) (Figure 4). ## 3. Discussion Initially, we focused on the identification of risk factors associated with later development of HELLP syndrome occurring separately or accompanying severe PE. We identified an increased incidence of HELLP syndrome in patients with already diagnosed autoimmune diseases, such as SLE, APS, SS, RA, T1DM, and coeliac disease, which has not yet been reported. Furthermore, we observed a higher incidence of HELLP syndrome in patients undergoing an infertility treatment by assisted reproductive technology (ART), which was also reported as an independent risk factor for the onset of hypertensive disorders during pregnancy, such as GH or PE [52,53,54] and FGR [55]. Qin et al. [ 56] pointed to the fact that singleton pregnancies undergoing assisted reproductive technologies are at a higher risk of adverse outcomes and recommended that they be managed as high-risk pregnancies. The ART singleton pregnancies had a significant risk of pregnancy-induced hypertension, GDM, placenta previa, placental abruption, antepartum hemorrhage, postpartum hemorrhage, polyhydramnios, oligohydramnios, cesarean sections, preterm birth, small for gestational age, perinatal mortality, and congenital malformation [56]. Similarly to other researchers [32,33,34,35,36,37], we confirmed that a history of HELLP syndrome and/or PE in a previous gestation represents a risk factor predisposing to the development of HELLP syndrome. Furthermore, we demonstrated that the presence of trombophilic gene mutations is more frequent in pregnancies developing HELLP syndrome. This finding is congruent with the observations of Muetze et al. [ 39], who reported that mutation in factor V Leiden is associated with HELLP syndrome in women of Caucasian descent. Unsurprisingly, we also detected a higher incidence of HELLP syndrome in patients with positive first-trimester PE/FGR screening by FMF [57,58,59,60]. All of the maternal risk factors identified by our research group were placed in a model together with maternal age and BMI values at early gestational stages to assess their common predictive potential for later development of HELLP syndrome. This prediction model for HELLP syndrome identified pregnancies developing HELLP syndrome with relatively high accuracy, since it was able to reveal $71.43\%$ of cases with a $10.0\%$ FPR. The addition of another maternal clinical characteristic to the prediction model (the positivity of first-trimester PE/FGR screening by FMF) did not yield a better detection rate, since it was able to detect the same proportion of pregnancies with HELLP syndrome ($71.43\%$ of cases with a $10.0\%$ FPR). Our model based on maternal risk factors only showed better performance than the logistic regression model demonstrated previously by Oliveira et al. [ 36]. This model was based on racial origin, nulliparity, and the occurrence of HELLP syndrome and PE in a previous gestation, and reached a detection rate of $55.0\%$ of cases only, with a $10.0\%$ FPR. A recently developed neuro-fuzzy model for HELLP syndrome identification and prediction [47] also showed a lower discrimination power than our novel model based on six or seven basic maternal clinical characteristics, including maternal age and BMI values at early gestational stages, the presence of any autoimmune disease, an infertility treatment by assisted reproductive technologies, the occurrence of HELLP syndrome and/or PE in a previous gestation, and the presence of mutations in trombophilic genes, or the positivity of first-trimester PE/FGR screening by FMF [57,58,59,60]. Afterwards, we were interested to see whether there was any other potential way to improve the detection rate of our novel HELLP predictive model based on maternal clinical characteristics only. Therefore, we evaluated the predictive potential of microRNAs that play a crucial role in the development and maintenance of homeostasis in the cardiovascular system and in the pathophysiology of cardiovascular and cerebrovascular diseases (Table 1) [51]. Gene expression of preselected microRNAs associated with the cardiovascular system was retrospectively studied in peripheral blood during the first trimester in pregnancies subsequently developing HELLP syndrome and in normal-term pregnancies selected as a matched control group based on the equality of the period of biological sample storage and gestational age at sampling. Currently, the upregulation of six microRNAs associated with the cardiovascular system (miR-1-3p, miR-17-5p, miR-143-3p, miR-146a-5p, miR-181a-5p, and miR-499a-5p) was observed during the early gestational stages in pregnancies developing HELLP syndrome. The combination of these six microRNA biomarkers only identified pregnancies developing HELLP syndrome with relatively high accuracy. A total of $78.57\%$ of pregnancies destined to develop HELLP syndrome was revealed at early gestational stages with a $10.0\%$ FPR. This is an optimistic result if we take into consideration that any knowledge of maternal clinical characteristics is required to achieve such a high discrimination power. Nevertheless, the availability of early screening for HELLP syndrome in clinical practice depends on the successful validation of microRNA biomarkers in consecutive prospective cohort studies and the acquisition of CE and IVDR certifications. The advantage of such a screening based on microRNA biomarkers only is that additional information about maternal characteristics is not needed. Interestingly, apart from HELLP-syndrome pregnancies, miR-1-3p also showed an altered early expression profile in pregnancies destined to develop SGA [49] and GDM [51]. MiR-143-3p displayed an aberrant expression profile in pregnancies developing PE [48]. Besides HELLP syndrome, an altered early expression profile of miR-146a-5p appeared in pregnancies developing PE [48] and FGR or SGA [49]. An aberrant early expression profile of miR-181a-5p was present in most pregnancies regardless of the type of pregnancy-related complication: GH or PE [48], FGR or SGA [49], GDM [51], or HELLP syndrome. MiR-499a-5p expression profiles were also observed to be dysregulated in early stages of gestation in pregnancies destined to develop GDM [51]. When these six cardiovascular-disease-associated microRNAs were added to the predictive model based on six maternal clinical characteristics identified by our research group as risk factors for the onset of HELLP syndrome, the discrimination power increased to $85.71\%$ with a $10.0\%$ FPR. The addition of another maternal clinical characteristic to the prediction model (the positivity of first-trimester PE/FGR screening by FMF) did not increase the AUC but significantly increased the sensitivity with a $10.0\%$ FPR. Using this approach, $92.86\%$ of HELLP pregnancies were finally detected. To our knowledge, no studies on the early prediction of HELLP syndrome during the first trimester through screening of extracellular microRNAs in maternal body fluids (plasma/serum) or peripheral blood samples are currently available. Just one study describing the identification of differentially expressed microRNAs in serum samples of patients with clinical manifestations of HELLP syndrome is available. Upregulation of miR-122, miR-758, and miR-133a was detected in a group of patients with HELLP syndrome [61]. Concerning miR-122 and miR-758, we did not examine their expression levels in maternal peripheral venous blood leukocytes. In addition, we studied miR-133a-3p, whose expression levels did not differ in early stages of gestation between pregnancies developing HELLP syndrome and those with normal courses of gestation delivering at term. Recently, bioinformatics analysis of microarray data has identified hub genes (KIT, JAK2, LEP, EP300, HIST1H4L, HIST1H4F, HIST1H4H, MMP9, THBS2, and ADAMTS) as diagnostic biomarkers of HELLP syndrome. MiR-34a-5p was demonstrated to be most associated with hub genes [46]. Unfortunately, the expression profile of miR-34a-5p was not studied in our group of patients. ## 4.1. Patient Cohort The peripheral blood sampling was performed within the framework of the first-trimester prenatal screening between 10 and 13 gestational weeks within the period November 2012–May 2018. In total, 4187 samples were collected from Caucasian singleton pregnancies. In the end, 3028 pregnant women delivered on site. In all, 14 of 3028 pregnant women were diagnosed with HELLP syndrome. The diagnosis and the severity of HELLP syndrome were assessed using two classification systems: the Tennessee classification and the Mississippi classification [20,27,28,29,30,31]. Complete (full) HELLP syndrome developed in 3 cases and incomplete (partial) HELLP syndrome was diagnosed in 11 cases. The most severe form of HELLP syndrome (also termed class 1) was present in 3 cases, the moderate form (class 2) in 4 cases, and the mild form (class 3) in 7 cases. Seven pregnancies were diagnosed with HELLP syndrome only, and seven pregnancies had HELLP syndrome associated with severe PE. Clinical management guidelines issued by the American College of Obstetricians and Gynecologists (ACOG) including diagnostic criteria for pre-eclampsia were followed [62]. The selection of controls was performed with respect to the equality of gestational age at the time of sample collection and the period of storage of biological samples. The control group consisted of 80 normal-term pregnancies. The control group delivered healthy newborns after 37 gestational weeks with weights over 2500 g. ## 4.2. Processing of Samples Processing of samples, reverse transcription (RT), and real-time qPCR analyses were performed as described previously [48,49,50,51]. In detail, leukocyte lysates were prepared from 200 µL peripheral blood samples using the QIAamp RNA Blood Mini Kit (Qiagen, Hilden, Germany) and were stored in a mixture of RLT buffer and β-mercaptoethanol (β-ME) at −80 °C. The MirVana microRNA Isolation kit (Ambion, Austin, USA) was used to isolate RNA fractions highly enriched for small RNAs. Gene expression of microRNAs associated with the cardiovascular system (miR-1-3p, miR-16-5p, miR-17-5p, miR-20a-5p, miR-20b-5p, miR-21-5p, miR-23a-3p, miR-24-3p, miR-26a-5p, miR-29a-3p, miR-92a-3p, miR-100-5p, miR-103a-3p, miR-125b-5p, miR-126-3p, miR-130b-3p, miR-133a-3p, miR-143-3p, miR-145-5p, miR-146-5p, miR-155-5p, miR-181a-5p, miR-195-5p, miR-199a-5p, miR-210-3p, miR-221-3p, miR-342-3p, miR-499a-5p, and miR-574-3p) was studied. RT and real-time qPCR analyses were performed via TaqMan MicroRNA Assays (Applied Biosystems, Branchburg, NJ, USA) on a 7500 Real-Time PCR System under standard TaqMan PCR conditions. The relative expression of microRNA genes was assessed using the delta-delta Ct method [63]. The endogenous controls (RNU58A and RNU38B) were used to normalize microRNA gene expression data [64,65]. ## 4.3. Statistical Analysis MicroRNA gene expression was compared between cases and controls using the Mann–Whitney test. Box plots displaying the medians, 75th and 25th percentiles, outliers (circles), and extremes (asterisks) were produced using Statistica software (version 9.0; StatSoft, Inc., Tulsa, OK, USA). Receiver operating characteristic (ROC) curves displayed the areas under the curves (AUCs), cut-off-point-associated sensitivities, specificities, positive and negative likelihood ratios (LR+, LR−), and sensitivities at a $10.0\%$ false-positive rate (FPR) (MedCalc Software bvba, Ostend, Belgium). To select the best microRNA combinations, logistic regression with subsequent ROC curve analyses was applied (MedCalc Software bvba, Ostend, Belgium). This statistical approach was also applied to develop novel early predictive models for HELLP syndrome based on the combination of appropriate microRNAs only, maternal clinical characteristics only, and the combination of appropriate microRNAs and maternal clinical characteristics (MedCalc Software bvba, Ostend, Belgium). ## 5. Conclusions Consecutive, large-scale retrospective analyses have to be performed to verify the reliability of our novel early predictive model for HELLP syndrome based on the combination of microRNAs associated with the cardiovascular system (miR-1-3p, miR-17a-5p, miR-143a-3p, miR-146a-5p, miR-181a-5p, and miR-499a-5p) and maternal clinical characteristics (maternal age and BMI values at early gestational stages, the presence of any autoimmune disease, an infertility treatment by assisted reproductive technologies, the occurrence of HELLP syndrome and/or PE in a previous gestation, the presence of mutations in trombophilic genes, and, alternatively, the positivity of first-trimester PE/FGR screening by FMF). The model based on the combination of six cardiovascular-disease-associated microRNA biomarkers and six maternal clinical characteristics has a very high discrimination power. It is able to detect $85.71\%$ of cases with a $10.0\%$ FPR. The addition of another maternal clinical characteristic to this particular prediction model (the positivity of first-trimester PE/FGR screening by FMF) may significantly increase the detection rate to $92.86\%$ of cases. In addition, consecutive prospective cohort studies are needed to validate the suggested early predictive model for HELLP syndrome. The availability of early screening for HELLP syndrome in clinical practice depends on the successful validation of microRNA biomarkers in consecutive prospective cohort studies and the acquisition of CE and IVDR certifications. Only high-risk pregnancies identified firstly by the early predictive model based on maternal characteristics may be further screened using microRNA biomarkers to achieve a reasonable cost–benefit ratio for the early predictive model for HELLP syndrome. ## 6. Patents National patent application, Industrial Property Office, Czech Republic (patent No. PV 2022-505). ## References 1. Weinstein L.. Syndrome of hemolysis, elevated liver enzymes and low platelet count; a severe consequence of hypertension in pregnancy. Am. J. Obstet. Gynecol. (1982.0) 142 159-167. DOI: 10.1016/S0002-9378(16)32330-4 2. Kirkpatrick C.A.. The HELLP syndrome. Acta. Clin. Belg. (2010.0) 65 91-97. DOI: 10.1179/acb.2010.020 3. Haram K., Svendsen E., Abildgaard U.. The HELLP syndrome: Clinical issues and management. A Review. BMC Pregnancy Childbirth (2009.0) 26. DOI: 10.1186/1471-2393-9-8 4. Waterstone M., Bewley S., Wolfe C.. Incidence and predictors of severe obstetrics morbidity: Case-control study. BJM (2001.0) 322 1089-1093. DOI: 10.1136/bmj.322.7294.1089 5. Martin J.N., Rinehart B.K., May W.L., Magann E.F., Terrone D.A., Blake P.G.. The spectrum of severe preeclampsia: Comparative analysis by HELLP (hemolysis, elevated liver enzyme levels, and low platelet count) syndrome classification. Am. J. Obstet. Gynecol. (1999.0) 180 1373-1384. DOI: 10.1016/S0002-9378(99)70022-0 6. Gasem T., Al Jama F.E., Burshaid S., Rahman J., Al Suleiman S.A., Rahman M.S.. Maternal and fetal outcome of pregnancy complicated by HELLP syndrome. J. Matern. Fetal. Neonatal. Med. (2009.0) 22 1140-1143. DOI: 10.3109/14767050903019627 7. Isler C.M., Rinehart B.K., Terrone D.A., May W.L., Magann E.F., Martin J.N.. The importance of parity to major maternal morbidity in the eclamptic mother with HELLP syndrome. Hypertens. Preganncy (2003.0) 22 287-294. DOI: 10.1081/PRG-120024032 8. Barton J.R., Sibai B.M.. Care of pregnancy complicated by HELLP syndrome. Obstet. Gynecol. Clin. N. Am. (1991.0) 18 165-179. DOI: 10.1016/S0889-8545(21)00266-7 9. Sibai B.M.. Maternal morbidity and mortality in 442 pregnancy with hemolysis, elevated liver enzymes, and low platelets (HELLP syndrome). Am. J. Obstet. Gynecol. (1993.0) 169 1000-1006. DOI: 10.1016/0002-9378(93)90043-I 10. Khumsat R., Wongwananurak T., Boriboonhirunsarn D.. Incidence and risk factors of HELLP syndrome in Thai pregnant women with severe pre-eclampsia. Thai J. Obstet. Gynaecol. (2008.0) 16 192-198 11. Abraham K.A., Connolly G., Farrell J., Walshe J.J.. The HELLP syndrome, a prospective study. Ren. Fail (2001.0) 23 705-713. DOI: 10.1081/JDI-100107367 12. Wolf J.L.. Liver disease in pregnancy. Med. Clin. N. Am. (1996.0) 80 1167-1187. DOI: 10.1016/S0025-7125(05)70484-3 13. Abildgaard U., Heimdal K.. Pathogenesis of the syndrome of hemolysis, elevated liver enzymes, and low platelet count (HELLP): A review. Eur. J. Obstet. Gynecol. Reprod. Biol. (2013.0) 166 117-123. DOI: 10.1016/j.ejogrb.2012.09.026 14. Raval D.S., Co S., Reid M.A., Pildes R.. Maternal and neonatal outcome of pregnancies complicated with maternal HELLP syndrome. J. Perinatol. (1997.0) 17 266-269. PMID: 9280089 15. Visser W., Wallenburg H.C.S.. Temporising management of severe pre-eclampsia with and without the HELLP syndrome. Br. J. Obstet. Gynaecol (1995.0) 102 111-117. DOI: 10.1111/j.1471-0528.1995.tb09062.x 16. Harms K., Rath W., Herting E., Kuhn W.. Maternal hemolysis, elevated liver enzymes, low platelet count, and neonatal outcome. Am. J. Perinatol. (1995.0) 12 1-7. DOI: 10.1055/s-2007-994387 17. Audibert F., Friedman S.A., Frangieh A.Y., Sibai B.M.. Clinical utility of strict diagnostic criteria for the HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets) syndrome. Am. J. Obstet. Gynecol. (1996.0) 175 460-464. DOI: 10.1016/S0002-9378(96)70162-X 18. Van Pampus M.G., Wolf H., Westenberg S.M., van der Post J.A.M., Bonsel G.J., Treffers P.E.. Maternal and perinatal outcome after expectant management of the HELLP syndrome compared with preeclampsia without HELLP syndrome. Eur. J. Obstet. Gynecol. Reprod. Biol. (1998.0) 76 31-36. DOI: 10.1016/S0301-2115(97)00146-2 19. Aloizos S., Seretis C., Liakos N., Aravosita P., Mystakelli C., Kanna E., Gourgiotis S.. HELLP syndrome: Understanding and management of a pregnancy-specific disease. J. Obstet. Gynaecol. (2013.0) 33 331-337. DOI: 10.3109/01443615.2013.775231 20. Backes C.H., Markham K., Moorehead P., Cordero L., Nankervis C.A., Giannone P.J.. Maternal preeclampsia and neonatal outcomes. J. Pregnancy (2011.0) 2011 214365. DOI: 10.1155/2011/214365 21. Ganzevoort W., Rep A., de Vries J.I., Bonsel G.J., Wolf H.. Prediction of maternal complications and adverse infant outcome at admission for temporizing management of early-onset severe hypertensive disorders of pregnancy. Am. J. Obstet. Gynecol. (2006.0) 195 495-503. DOI: 10.1016/j.ajog.2006.02.012 22. Van Lieshout L.C.E.W., Koek G.H., Spaanderman M.A., van Runnard Heimel P.J.. Placenta derived factors involved in the pathogenesis of the liver in the syndrome of haemolysis, elevated liver enzymes and low platelets (HELLP): A review. Pregnancy Hypertens. (2019.0) 18 42-48. DOI: 10.1016/j.preghy.2019.08.004 23. Liu Q., Ling G.J., Zhang S.Q., Zhai W.Q., Chen Y.J.. Effect of HELLP syndrome on acute kidney injury in pregnancy and pregnancy outcomes: A systematic review and meta-analysis. BMC Pregnancy Childbirth (2020.0) 20. DOI: 10.1186/s12884-020-03346-4 24. Pavlis T., Aloizos S., Aravosita P., Mystakelli C., Petrochilou D., Dimopoulos N., Gourgiotis S.. Diagnosis and surgical management of spontaneous hepatic rupture associated with HELLP syndrome. J. Surg. Educ. (2009.0) 66 163-167. DOI: 10.1016/j.jsurg.2009.04.001 25. Sibai B.M.. The HELLP syndrome (hemolysis, elevated liver enzymes, and low platelets): Much ado about nothing?. Am. J. Obstet. Gynecol. (1990.0) 162 311-316. DOI: 10.1016/0002-9378(90)90376-I 26. Sibai B.M.. Diagnosis, controversies, and management of the syndrome of hemolysis, elevated liver enzymes, and low platelet count. Obstet. Gynecol. (2004.0) 103 981-991. DOI: 10.1097/01.AOG.0000126245.35811.2a 27. Sibai B.M.. Imitators of severe pre-eclampsia/eclampsia. Clin. Perinatol. (2004.0) 31 835-852. DOI: 10.1016/j.clp.2004.06.007 28. Martin J.N., Blake P.G., Perry K.G., McCaul J.F., Hess L.W., Martin R.W.. The natural history of HELLP syndrome: Patterns of disease progression and regression. Am. J. Obstet. Gynecol. (1991.0) 164 1500-1509. DOI: 10.1016/0002-9378(91)91429-Z 29. Martin J.N., Rose C.H., Briery C.M.. Understanding and managing HELLP syndrome: The integral role of aggressive glucocorticoids for mother and child. Am. J. Obstet. Gynecol. (2006.0) 195 914-934. DOI: 10.1016/j.ajog.2005.08.044 30. Malmström O., Morken N.H.. HELLP syndrome, risk factors in first and second pregnancy: A population-based cohort study. Acta. Obstet. Gynecol. Scand. (2018.0) 97 709-716. DOI: 10.1111/aogs.13322 31. Leeners B., Neumaier-Wagner P.M., Kuse S., Mütze S., Rudnik-Schöneborn S., Zerres K., Rath W.. Recurrence risks of hypertensive diseases in pregnancy after HELLP syndrome. J. Perinat. Med. (2011.0) 39 673-678. DOI: 10.1515/jpm.2011.081 32. Habli M., Eftekhari N., Wiebracht E., Bombrys A., Khabbaz M., How H., Sibai B.. Long-term maternal and subsequent pregnancy outcomes 5 years after hemolysis, elevated liver enzymes, and low platelets (HELLP) syndrome. Am. J. Obstet. Gynecol. (2009.0) 201 e1-e5. DOI: 10.1016/j.ajog.2009.06.033 33. Hupuczi P., Rigó B., Sziller I., Szabó G., Szigeti Z., Papp Z.. Follow-up analysis of pregnancies complicated by HELLP syndrome. Fetal. Diagn. Ther. (2006.0) 21 519-522. DOI: 10.1159/000095665 34. Fitzpatrick K.E., Hinshaw K., Kurinczuk J.J., Knight M.. Risk factors, management, and outcomes of hemolysis, elevated liver enzymes, and low platelets syndrome and elevated liver enzymes, low platelets syndrome. Obstet. Gynecol. (2014.0) 123 618-627. DOI: 10.1097/AOG.0000000000000140 35. Oliveira N., Poon L.C., Nicolaides K.H., Baschat A.A.. First trimester prediction of HELLP syndrome. Prenat. Diagn. (2016.0) 36 29-33. DOI: 10.1002/pd.4694 36. Pusl T., Beuers U.. Intrahepatic cholestasis of pregnancy. Orphanet. J. Rare Dis. (2007.0) 2 26. DOI: 10.1186/1750-1172-2-26 37. Haram K., Mortensen J.H., Nagy B.. Genetic aspects of preeclampsia and the HELLP syndrome. J. Pregnancy (2014.0) 2014 910751. DOI: 10.1155/2014/910751 38. Stojanovska V., Zenclussen A.C.. Innate and Adaptive Immune Responses in HELLP Syndrome. Front. Immunol. (2020.0) 11 667. DOI: 10.3389/fimmu.2020.00667 39. Muetze S., Leeners B., Ortlepp J.R., Kuse S., Tag C.G., Weiskirchen R., Gressner A.M., Rudnik-Schoeneborn S., Zerres K., Rath W.. Maternal factor V Leiden mutation is associated with HELLP syndrome in Caucasian women. Acta. Obstet. Gynecol. Scand (2008.0) 87 635-642. DOI: 10.1080/00016340802112740 40. Yu H., Yang Z., Ding X., Wang Y., Han Y.. Effects of serum from patients with early-onset pre-eclampsia, HELLP syndrome, and antiphospholipid syndrome on fatty acid oxidation in trophoblast cells. Arch. Gynecol. Obstet. (2015.0) 292 559-567. DOI: 10.1007/s00404-015-3669-2 41. Kongwattanakul K., Saksiriwuttho P., Chaiyarach S., Thepsuthammarat K.. Incidence, characteristics, maternal complications, and perinatal outcomes associated with preeclampsia with severe features and HELLP syndrome. Int. J. Womens. Health (2018.0) 10 371-377. DOI: 10.2147/IJWH.S168569 42. Adorno M., Maher-Griffiths C., Grush Abadie H.R.. HELLP Syndrome. Crit. Care. Nurs. Clin. N. Am. (2022.0) 34 277-288. DOI: 10.1016/j.cnc.2022.04.009 43. Wallace K., Harris S., Addison A., Bean C.. HELLP Syndrome: Pathophysiology and Current Therapies. Curr. Pharm. Biotechnol. (2018.0) 19 816-826. DOI: 10.2174/1389201019666180712115215 44. Asadikalameh Z., Maddah R., Maleknia M., Nassaj Z.S., Ali N.S., Azizi S., Dastyar F.. Bioinformatics analysis of microarray data to identify hub genes, as diagnostic biomarker of HELLP syndrome: System biology approach. J. Obstet. Gynaecol. Res. (2022.0) 48 2493-2504. DOI: 10.1111/jog.15363 45. Lam M.T.C., Dierking E.. Intensive Care Unit issues in eclampsia and HELLP syndrome. Int. J. Crit. Illn. Inj. Sci. (2017.0) 7 136-141. DOI: 10.4103/IJCIIS.IJCIIS_33_17 46. Gupta M., Feinberg B.B., Burwick R.M.. Thrombotic microangiopathies of pregnancy: Differential diagnosis. Pregnancy Hypertens (2018.0) 12 29-34. DOI: 10.1016/j.preghy.2018.02.007 47. Moreira M.W.L., Rodrigues J.J.P.C., Al-Muhtadi J., Korotaev V.V., de Albuquerque V.H.C.. Neuro-fuzzy model for HELLP syndrome prediction in mobile cloud computing environments. Concurr. Comput. Pract. Exp. (2021.0) 22 e4651. DOI: 10.1002/cpe.4651 48. Hromadnikova I., Kotlabova K., Krofta L.. Cardiovascular Disease-Associated MicroRNA Dysregulation during the First Trimester of Gestation in Women with Chronic Hypertension and Normotensive Women Subsequently Developing Gestational Hypertension or Preeclampsia with or without Fetal Growth Restriction. Biomedicines (2022.0) 10 256. PMID: 35203467 49. Hromadnikova I., Kotlabova K., Krofta L.. First-Trimester Screening for Fetal Growth Restriction and Small-for-Gestational-Age Pregnancies without Preeclampsia Using Cardiovascular Disease-Associated MicroRNA Biomarkers. Biomedicines (2022.0) 10. DOI: 10.3390/biomedicines10030718 50. Hromadnikova I., Kotlabova K., Krofta L.. First Trimester Prediction of Preterm Delivery in the Absence of Other Pregnancy-Related Complications Using Cardiovascular-Disease Associated MicroRNA Biomarkers. Int. J. Mol. Sci (2022.0) 23. DOI: 10.3390/ijms23073951 51. Hromadnikova I., Kotlabova K., Krofta L.. Cardiovascular Disease-Associated MicroRNAs as Novel Biomarkers of First-Trimester Screening for Gestational Diabetes Mellitus in the Absence of Other Pregnancy-Related Complications. Int. J. Mol. Sci. (2022.0) 23. DOI: 10.3390/ijms231810635 52. Hernández-Díaz S., Werler M.M., Mitchell A.A.. Gestational hypertension in pregnancies supported by infertility treatments: Role of infertility, treatments, and multiple gestations. Fertil. Steril. (2007.0) 88 438-445. DOI: 10.1016/j.fertnstert.2006.11.131 53. Almasi-Hashiani A., Omani-Samani R., Mohammadi M., Amini P., Navid B., Alizadeh A., Khedmati Morasae E., Maroufizadeh S.. Assisted reproductive technology and the risk of preeclampsia: An updated systematic review and meta-analysis. BMC Pregnancy Childbirth (2019.0) 19. DOI: 10.1186/s12884-019-2291-x 54. Monseur B.C., Morris J.R., Hipp H.S., Berghella V.. Hypertensive disorders of pregnancy and infertility treatment: A population-based survey among United States women. J. Assist. Reprod. Genet. (2019.0) 36 1449-1456. DOI: 10.1007/s10815-019-01490-1 55. Zhu J.L., Obel C., Hammer Bech B., Olsen J., Basso O.. Infertility, infertility treatment, and fetal growth restriction. Obstet. Gynecol. (2007.0) 110 1326-1334. DOI: 10.1097/01.AOG.0000290330.80256.97 56. Qin J., Liu X., Sheng X., Wang H., Gao S.. Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy outcomes in singleton pregnancies: A meta-analysis of cohort studies. Fertil. Steril. (2016.0) 105 73.e1–85.e6. DOI: 10.1016/j.fertnstert.2015.09.007 57. O’Gorman N., Wright D., Poon L.C., Rolnik D.L., Syngelaki A., de Alvarado M., Carbone I.F., Dutemeyer V., Fiolna M., Frick A.. Multicenter screening for pre-eclampsia by maternal factors and biomarkers at 11–13 weeks’ gestation: Comparison with NI-CE guidelines and ACOG recommendations. Ultrasound Obs. Gynecol. (2017.0) 49 756-760. DOI: 10.1002/uog.17455 58. O’Gorman N., Wright D., Syngelaki A., Akolekar R., Wright A., Poon L.C., Nicolaides K.H.. Competing risks model in screening for preeclampsia by maternal factors and biomarkers at 11–13 weeks gestation. Am. J. Obs. Gynecol. (2016.0) 214 103.e1–103.e12. DOI: 10.1016/j.ajog.2015.08.034 59. Stratification of Pregnancy Management 11–13 Weeks’ Gestation 60. Mazer Zumaeta A., Wright A., Syngelaki A., Maritsa V.A., Da Silva A.B., Nicolaides K.H.. Screening for pre-eclampsia at 11-13 weeks’ gestation: Use of pregnancy-associated plasma protein-A, placental growth factor or both. Ultrasound Obstet. Gynecol. (2020.0) 56 400-407. DOI: 10.1002/uog.22093 61. Stubert J., Koczan D., Richter D.U., Dieterich M., Ziems B., Thiesen H.J., Gerber B., Reimer T.. miRNA expression profiles determined in maternal sera of patients with HELLP syndrome. Hypertens. Pregnancy (2014.0) 33 215-235. DOI: 10.3109/10641955.2013.858743 62. Gestational Hypertension and Preeclampsia: ACOG Practice Bulletin, Number 222. Obstet. Gynecol. (2020.0) 135 e237-e260. DOI: 10.1097/AOG.0000000000003891 63. Livak K.J., Schmittgen T.D.. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods (2001.0) 25 402-408. DOI: 10.1006/meth.2001.1262 64. Vandesompele J., De Preter K., Pattyn F., Poppe B., Van Roy N., De Paepe A., Speleman F.. Accurate normalization of re-al-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol. (2002.0) 3 1-2. DOI: 10.1186/gb-2002-3-7-research0034 65. Hromadnikova I., Kotlabova K., Hympanova L., Krofta L.. Gestational hypertension, preeclampsia and intrauterine growth restriction induce dysregulation of cardiovascular and cerebrovascular disease associated microRNAs in maternal whole peripheral blood. Thromb. Res. (2016.0) 137 126-140. DOI: 10.1016/j.thromres.2015.11.032
--- title: 'Major Causes of Death among Older Adults after the Great East Japan Earthquake: A Retrospective Study' authors: - Takako Fujimaki - Yuko Ohno - Anna Tsutsui - Yuta Inoue - Ling Zha - Makoto Fujii - Tetsuya Tajima - Satoshi Hattori - Tomotaka Sobue journal: International Journal of Environmental Research and Public Health year: 2023 pmcid: PMC10049726 doi: 10.3390/ijerph20065058 license: CC BY 4.0 --- # Major Causes of Death among Older Adults after the Great East Japan Earthquake: A Retrospective Study ## Abstract This retrospective study investigated the 3-year impact of the Great East Japan Earthquake (GEJE) of 2011 on deaths due to neoplasm, heart disease, stroke, pneumonia, and senility among older adults in the primarily affected prefectures compared with other prefectures, previous investigations having been more limited as regards mortality causes and geographic areas. Using death certificates issued between 2006 and 2015 ($$n = 7$$,383,253), mortality rates (MRs) and risk ratios (RRs) were calculated using a linear mixed model with the log-transformed MR as the response variable. The model included interactions between the area category and each year of death from 2010 to 2013. The RRs in the interaction significantly increased to 1.13, 1.17, and 1.28 for deaths due to stroke, pneumonia, and senility, respectively, in Miyagi Prefecture in 2011, but did not significantly increase for any of the other areas affected by the GEJE. Moreover, increased RRs were not reported for any of the other years. The risk of death increased in 2011; however, this was only significant for single-year impact. In 2013, decreased RRs of pneumonia in the Miyagi and Iwate prefectures and of senility in Fukushima Prefecture were observed. Overall, we did not find evidence of strong associations between the GEJE and mortality. ## 1. Introduction The Great East Japan Earthquake (GEJE), which occurred on 11 March 2011, was followed by a large tsunami that struck the northeastern coastal areas of Japan, causing an accident within the Fukushima Daiichi nuclear power plant. Approximately 19,000 people died or are still missing due to the GEJE [1]. In total, over 400,000 survivors were evacuated or stranded a week after the GEJE due to damage to their homes and restricted access to areas surrounding the nuclear power plant [1]. The majority of evacuees were in the Fukushima, Miyagi, and Iwate prefectures, and people aged ≥65 years accounted for over $20\%$ of the evacuees [1]. Vulnerability to disasters is associated with impaired mobility, reduced sensory function, chronic health problems, social isolation, and economic constraints [2,3,4]. In addition, older adults are more likely to have chronic diseases, and the interruption of treatment, stress, lack of food and clean water, extreme heat and cold, and infections due to disaster exacerbate chronic diseases [5,6,7]. In Japan, with its rapidly aging population [8], it is crucial to elucidate the impact of earthquakes on older adults to prepare for the occurrence of disasters. Several studies have investigated mortality from cardiovascular disease and/or stroke following earthquakes, such as those pertaining to the Northridge Earthquake in 1994 [9], the Great Hanshin Awaji Earthquake (GHAE) in 1995 [10,11,12], the Niigata-Chuetsu Earthquake in 2004 [13], the GEJE in 2011 [14], and both the GHAE and the GEJE [15]. However, many of these studies were limited as regards both causes of death and geographic areas covered. Our study additionally focused on older adults who are especially vulnerable to disasters and at high risk of death. This study investigated the impact of the GEJE on mortality from five major causes among older adults in the severely affected prefectures compared with other prefectures in Japan. We hypothesized that mortality from the five causes would have increased following the earthquake among older adults in severely affected prefectures compared with those in other prefectures. ## 2.1. Study Design and Data Sources This retrospective study used population-based death certificate data, which included information on sex, age, and residential address (prefecture). To obtain the relevant statistical data, an application to the Ministry of Health, Labour and Welfare of Japan was required. There were two available population data sources based on the Japanese census data. One estimate by the National Cancer Center of Japan was compiled for the age categories 0–99 years and ≥100 years [16]. Another estimate by the Statistics Bureau of Japan was only compiled for the age categories 0–84 years and ≥85 years [17]. We required population data by the 5-year age categories (0–94 and ≥95 years) for this study. Therefore, we selected population data from the National Cancer Center. The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Osaka University (Approval No. 15272). ## 2.2. Participants Figure 1 shows the flow chart of included and excluded participants. We obtained 12,092,057 death certificates registered between 2006 and 2015. From these data, we excluded those with foreign nationality and/or a residential address outside Japan ($$n = 85$$,978). Japanese nationals with a domestic residential address accounted for 12,006,079 of these cases. We also excluded those with unknown or missing data for the age of death ($$n = 435$$) and those with a mismatch between death and registration years ($$n = 7885$$). A total of 11,997,759 deaths were analyzed. To investigate the mortality rate per 100,000 individuals (MR), we selected adults ≥ 65 years of age ($$n = 10$$,290,956). For this group, we selected five major causes of death: neoplasm; heart disease, except hypertensive disease; stroke; pneumonia; and senility [18] ($$n = 7$$,383,253). For deaths due to senility, the ages 65–74 years were excluded from the analysis due to there being a limited number of death registrations or no deaths ($$n = 3958$$). We analyzed 7,379,295 deaths to investigate the MR. Because the 10th version of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) was applied to classify the primary cause of death in the data, we identified the cause of death using the following ICD-10 codes: C00–97, D00–09, D18.0, D32–33, D35.2–D35.4, and D37–D48 for neoplasm; I01–I02.0, I05–I09, I20–I25, I27, and I30–I52 for heart disease; I60–I69 for stroke; J09–J18 for pneumonia; and R54 for senility. We defined the three prefectures (the Fukushima, Miyagi, and Iwate prefectures) with more than 100 deaths from the GEJE as severely affected prefectures and the remaining 44 prefectures as unaffected prefectures. Figure 2 highlights the location of these severely affected prefectures. ## 2.3. Statistical Analysis First, we summarized the population and death data from 2006 to 2015. Second, we obtained the MR for each sex, age category (65–69, 70–74, 75–79, 80–84, 85–89, 90–94, and ≥95 years), and prefecture to observe annual trends in mortality. The MR for each group was calculated by dividing the number of deaths by the population number and multiplying by 100,000. We excluded data with zero MR. The subgroup with zero MR was the senility group and was only $0.2\%$. Third, we used linear mixed models (LMMs) to estimate risk ratios (RRs) and $95\%$ confidence intervals (CIs) in the cause of death in affected prefectures for several single years after the GEJE. The response variable was the log-transformed MR. Adjustment for multiplicity was not performed. The model included the following fixed effects: area category—Fukushima, Miyagi, Iwate, and the remaining 44 prefectures; year of death—each year from 2006 to 2015; age category—65–69, 70–74, 75–79, 80–84, 85–89, 90–94, and ≥95 years; sex—male and female; and interaction terms between the area category and each year of death from 2010 to 2013. In addition, a random effect of intercept for all 47 prefectures in Japan was included. Of the observed years, 2006–2015, we focused on the results for the years 2010–2013. The year of death 2010 was included as a year with no impact from the GEJE. The years of death 2011–2013 were selected to observe RRs after the GEJE, as most of the previous studies investigating long-term deaths after earthquakes had not determined any impact beyond 3 years. A p-value < 0.05 was considered statistically significant. Statistical analysis was performed using SAS 9.4® (SAS Institute Inc., Cary, NC, USA), and the MIXED procedure of SAS was used to perform LMM analyses. ## 3. Results The characteristics of the population data are summarized in Table A1. The percentages of the mean annual population aged ≥65 years were $25.6\%$, $22.9\%$, $27.7\%$, and $23.6\%$ in the Fukushima, Miyagi, Iwate, and the other 44 prefectures, respectively. The percentage of males aged ≥65 years (40.7–$42.7\%$) was lower than that of females in all area categories. Table 1 summarizes the characteristics of the death data by age category, cause of death, and sex. The mean annual mortality rates (AMRs) for all causes of death in people aged 0–64 years were 202, 198, 237, and 183 deaths per 100,000 population in Fukushima, Miyagi, Iwate, and the other 44 prefectures, respectively. The mean AMRs for those aged ≥65 years were 3862, 3610, 3836, and 3575 deaths per 100,000 population in Fukushima, Miyagi, Iwate, and the other 44 prefectures, respectively. The older the age category, the higher its AMRs. The area category with the highest mean AMR was Fukushima Prefecture for deaths due to neoplasm, heart disease, and senility, Iwate Prefecture for stroke, and the other 44 prefectures for pneumonia. A comparison of males and females aged ≥65 years showed that AMRs were higher in males for all causes of death. Figure 3 shows annual trends in the MRs by cause of death and prefecture for males and females, respectively. These graphs also illustrate the major annual trends for the other unaffected prefectures ($$n = 44$$). The MRs by cause of death were highest for neoplasm, followed by heart disease, stroke or pneumonia, and senility. The annual trends increased for deaths due to senility and decreased for deaths due to other causes. The MRs were higher in males for all causes of death, and there were large sex differences for deaths due to neoplasm, stroke, and pneumonia. The MRs were generally higher among the older age categories than younger age categories across all causes of death. In a comparison between affected and unaffected prefectures, the MRs due to several causes of death were higher in the affected prefectures than in other prefectures in 2011; however, the results for the sub-age categories differed by age category. Table 2 summarizes the highlights of the RRs and $95\%$ CIs for fixed effects and interaction terms by cause of death. Table A2 lists the RRs for all the fixed effects and interaction terms. Considering the interaction terms between the area category and the year of death, the RRs for neoplasm showed minor variations from 2010 to 2013, ranging between 0.95 and 1.01 in the three affected prefectures. The RRs for heart disease in the three affected prefectures were slightly lower than those in the 44 prefectures in 2010 (0.99 [$95\%$ CI, 0.88–1.12], 0.97 [0.86–1.09], and 0.96 [0.85–1.08] for the Fukushima, Miyagi, and Iwate prefectures, respectively), but were higher than those in the 44 prefectures in 2011 (1.06 [$95\%$ CI, 0.95–1.20], 1.03 [0.91–1.15], and 1.06 [0.95–1.20], respectively). The RRs for stroke between 2010 and 2013 were the highest in 2011 in the Fukushima, Miyagi, and Iwate prefectures (1.08 [$95\%$ CI, 0.96–1.21], 1.13 [1.01–1.27], and 1.07 [0.96–1.20], respectively). Of these, the RR was significantly higher only in Miyagi Prefecture. The RRs for pneumonia between 2010 and 2013 were higher in the Fukushima, Miyagi, and Iwate prefectures than in the 44 prefectures in 2011 (1.10 [$95\%$ CI, 0.99–1.23], 1.17 [1.04–1.31], and 1.04 [0.93–1.17], respectively), although the RR was significant only in Miyagi Prefecture in that year. The RRs were significantly lower in the Miyagi and Iwate prefectures in 2013 (0.86 [$95\%$ CI, 0.77–0.97 and 0.88 [0.78–0.98], respectively). In 2011, the RR for senility was significantly higher only in Miyagi Prefecture (1.28 [$95\%$ CI, 1.07–1.54]), and slightly lower in the Fukushima and Iwate prefectures (0.98 [$95\%$ CI, 0.81–1.17 and 0.97 [0.81–1.16], respectively) compared to the 44 prefectures. For fixed effects, the RR for stroke was significantly high only in Iwate Prefecture (1.37 [$95\%$ CI, 1.07–1.74]). ## 4. Discussion We investigated the impact of the GEJE on mortality in older adults across three severely affected prefectures and compared it to the 44 unaffected prefectures in Japan using the enumerated data. Because no apparent changes were observed in the graphs of MRs post-GEJE (Figure 3), statistical analysis was performed using an LMM. As a result, we found that there were significant increases in the RRs for stroke, pneumonia, and senility in Miyagi Prefecture in 2011, while there were no significant increases in the other areas. Although we presented the results based on an LMM, we also applied Poisson and negative binomial mixed models. However, these models failed to converge on several causes of death and could not provide estimates. When obtaining estimates, these models produced estimates very similar to those provided by the LMM; however, the LMM yielded more conservative results. Therefore, we have presented our results using the LMM in a unified manner for all causes of death to describe the changes in area-specific MRs during the years shown in Figure 3. This study showed that the fixed effect for the RR for deaths due to stroke was significantly higher in Iwate Prefecture. However, the RR in the interaction terms between the area category and each year from 2010 to 2013 for deaths due to stroke significantly increased only in Miyagi Prefecture in 2011, and not in Iwate Prefecture. Several previous studies have shown post-earthquake increases in the incidence of death due to stroke [11,15] and the incidence of stroke [19,20,21]. Meanwhile, previous studies have noted no significant increases in the incidence of death caused by stroke [9] and the incidence of stroke [22]. Thus, the impacts of earthquakes on the incidence of stroke were inconsistent. This inconsistency may be related to a variety of factors. Reasons for differences in the impact on the RR among affected prefectures could not be clarified. Hypertension is a known risk factor for stroke [23,24]. Elevated blood pressure after disasters is referred to as disaster hypertension, which has been shown to affect individuals after earthquakes [25,26,27,28,29,30,31,32,33]. One previous study reported that the mean blood pressure increased in both evacuees and non-evacuees after the GEJE, and the change was more pronounced in evacuees compared to that in non-evacuees [33]. Several studies showed that an elevation in blood pressure peaked 1–2 weeks after the earthquake [25,27,31] and that it required 3–6 weeks to recover to the same level of blood pressure as before the earthquake [25,27,28]. These studies suggest that it is important for residents, especially evacuees, to undergo screening from the early phases after the earthquakes to detect elevated blood pressure and subsequently ensure appropriate blood-pressure control. In the RRs for mortality from heart disease, there was no significant change across all interaction terms. One previous study reported that there was no long-term impact on cardiovascular death, hospitalization for heart failure, and acute myocardial infarction for the three-year follow-up after the GEJE [14]. As regards short-term impact within 15 weeks after the earthquake, there was no significant increase in the incidence of acute myocardial infarction and takotsubo cardiomyopathy [34]. However, numerous previous studies have reported an increase in the incidence of death due to coronary heart disease [9,10,11,12,13,15] and the incidence of coronary heart disease [21,35,36,37,38,39], heart failure [21,34,40,41], out-of-hospital cardiac arrest [21,42,43], and takotsubo cardiomyopathy [44] after devastating earthquakes. Risk factors have been identified for heart disease although many of these factors were not considered in this study. Therefore, future studies should consider these risk factors. The RRs for death due to pneumonia significantly increased in Miyagi Prefecture in 2011; meanwhile, they decreased in Miyagi Prefecture and Iwate Prefecture in 2013. The reason behind this decrease could not be identified on the basis of previous studies. Several previous studies observed an increase in the incidence of pneumonia in areas severely affected by the GEJE [21,45,46]. Several studies reported that more patients with pneumonia were living in shelters following the GHAE [47] and the GEJE [48]. The number of evacuees after the GEJE was the highest in Miyagi Prefecture, followed by the Fukushima and Iwate prefectures [1]. The importance of proper living conditions, hygiene, and nutrition in shelters has been recommended in the Guidelines for the Management and Prevention of Cardiovascular Diseases in Disasters published after the GEJE [49]. The RRs for death due to senility significantly increased only in Miyagi Prefecture in 2011 and significantly decreased in Fukushima Prefecture in 2013. No previous studies mention a direct impact on deaths due to senility after earthquakes. One previous study reported that physical activity and muscle strength decreased in older adults living in temporary housing compared with those living at home following the GEJE [50]. Another study reported that damage to housing was significantly associated with cognitive decline among older survivors [51]. Meanwhile, factors contributing to the decrease in 2013 remain unclarified. There was a small variation and no significant increase in the RRs for mortality from neoplasm. Although many medical facilities were affected by the GEJE, it was unknown exactly how many patients with cancer consequently experienced a delayed diagnosis or interruption in treatment. A study reported that hospitalization for the progression of lung cancer increased slightly during the first 2 months following the GEJE, but not significantly [45]. This study had several limitations. First, the data did not include information on other important influencing factors on death, such as comorbidities, daily activity levels, and whether the participants were evacuated after the GEJE. Second, a comparison between the seriously damaged coastal and the less-damaged inland areas in each prefecture was not included in the data. Lastly, the interpretation of p-values should be exploratory only, since the estimated p-values were not adjusted for multiple tests. Therefore, future studies are required to assess the association between the cause of death following a disaster and individual background characteristics. ## 5. Conclusions We observed increases in the RRs of death due to stroke, pneumonia, and senility in the Miyagi Prefecture in 2011 among older adults, whereas decreases in the RRs of death due to pneumonia n the Miyagi and Iwate prefectures and senility in Fukushima Prefecture in 2013. The reason for this decrease could not be identified in previous studies. No strong association was observed between the GEJE and mortality; however, our results suggest this would have been significant only for the single-year impact. Future studies are warranted to assess the association between the cause of death following a disaster and individual background characteristics. ## References 1. The Great East Japan Earthquake: A Story of Devastating Natural Disaster, a Tale of Human Compassion 2. Fernandez L.S., Byard D., Lin C.C., Benson S., Barbera J.A.. Frail elderly as disaster victims: Emergency management strategies. Prehosp. Disast. Med. (2002.0) 17 67-74. DOI: 10.1017/S1049023X00000200 3. Pekovic V., Seff J., Rothman M.. Planning for and responding to special needs of elders in natural disasters. Generations (2007.0) 31 37-41 4. Cornell V.J., Cusack L., Arbon P.. Older people and disaster preparedness: A literature review. Aust. J. Emerg. Manag. (2012.0) 27 49-53 5. Mokdad A.H., Mensah G.A., Posner S.F., Reed E., Simoes E.J., Engelgau M.M.. Chronic Diseases and Vulnerable Populations in Natural Disasters Working Group. When chronic conditions become acute: Prevention and control of chronic diseases and adverse health outcomes during natural disasters. Prev. Chronic Dis. (2005.0) 2 A04. PMID: 16263037 6. Aldrich N., Benson W.F.. Disaster preparedness and the chronic disease needs of vulnerable older adults. Prev. Chronic Dis. (2008.0) 5 A27. PMID: 18082016 7. Tomio J., Sato H.. Emergency and disaster preparedness for chronically ill patients: A review of recommendations. Open Access Emerg. Med. (2014.0) 6 69-79. PMID: 27147882 8. Annual Report on the Aging Society (Summary) 9. Kloner R.A., Leor J., Poole W.K., Perritt R.. Population-based analysis of the effect of the Northridge Earthquake on cardiac death in Los Angeles County, California. J. Am. Coll. Cardiol. (1997.0) 30 1174-1180. DOI: 10.1016/S0735-1097(97)00281-7 10. Kario K., Ohashi T.. Increased coronary heart disease mortality after the Hanshin-Awaji earthquake among the older community on Awaji Island. Tsuna Medical Association. J. Am. Geriatr. Soc. (1997.0) 45 610-613. DOI: 10.1111/j.1532-5415.1997.tb03096.x 11. Kario K., Ohashi T.. After a major earthquake, stroke death occurs more frequently than coronary heart disease death in very old subjects. J. Am. Geriatr. Soc. (1998.0) 46 537-538. DOI: 10.1111/j.1532-5415.1998.tb02485.x 12. Ogawa K., Tsuji I., Shiono K., Hisamichi S.. Increased acute myocardial infarction mortality folldue the 1995 Great Hanshin-Awaji earthquake in Japan. Int. J. Epidemiol. (2000.0) 29 449-455. DOI: 10.1093/ije/29.3.449 13. Nakagawa I., Nakamura K., Oyama M., Yamazaki O., Ishigami K., Tsuchiya Y., Yamamoto M.. Long-term effects of the Niigata-Chuetsu earthquake in Japan on acute myocardial infarction mortality: An analysis of death certificate data. Heart (2009.0) 95 2009-2013. DOI: 10.1136/hrt.2009.174201 14. Miyata S., Sakata Y., Miura M., Yamauchi T., Onose T., Tsuji K., Abe R., Oikawa T., Kasahara S., Sato M.. CHART-2 Investigators. Long-term prognostic impact of the Great East Japan Earthquake in patients with cardiovascular disease— Report from the CHART-2 Study. J. Cardiol. (2017.0) 70 286-296. DOI: 10.1016/j.jjcc.2016.10.018 15. Takegami M., Miyamoto Y., Yasuda S., Nakai M., Nishimura K., Ogawa H., Hirata K.I., Toh R., Morino Y., Nakamura M.. Comparison of cardiovascular mortality in the Great East Japan and the Great Hanshin-Awaji Earthquakes—A large-scale data analysis of death certificates. Circ. J. (2015.0) 79 1000-1008. DOI: 10.1253/circj.CJ-15-0223 16. Population for the Cancer Registration 17. Statistics Bureau of Japan 18. Handbook of Health and Welfare Statistics 2019 19. Tsuchida M., Kawashiri M.A., Teramoto R., Takata M., Sakata K., Omi W., Okajima M., Takamura M., Ino H., Kita Y.. Impact of severe earthquake on the occurrence of acute coronary syndrome and stroke in a rural area of Japan. Circ. J. (2009.0) 73 1243-1247. DOI: 10.1253/circj.CJ-08-0812 20. Omama S., Yoshida Y., Ogasawara K., Ogawa A., Ishibashi Y., Nakamura M., Tanno K., Ohsawa M., Onoda T., Itai K.. Influence of the great East Japan earthquake and tsunami 2011 on occurrence of cerebrovascular diseases in Iwate, Japan. Stroke (2013.0) 44 1518-1524. DOI: 10.1161/STROKEAHA.111.000442 21. Aoki T., Fukumoto Y., Yasuda S., Sakata Y., Ito K., Takahashi J., Miyata S., Tsuji I., Shimokawa H.. The Great East Japan Earthquake Disaster and cardiovascular diseases. Eur. Heart J. (2012.0) 33 2796-2803. DOI: 10.1093/eurheartj/ehs288 22. Nozaki E., Nakamura A., Abe A., Kagaya Y., Kohzu K., Sato K., Nakajima S., Fukui S., Endo H., Takahashi T.. Occurrence of cardiovascular events after the 2011 Great East Japan Earthquake and tsunami disaster. Int. Heart J. (2013.0) 54 247-253. DOI: 10.1536/ihj.54.247 23. Gosmanova E.O., Mikkelsen M.K., Molnar M.Z., Lu J.L., Yessayan L.T., Kalantar-Zadeh K., Kovesdy C.P.. Associa tion of systolic blood pressure variability with mortality, coronary heart disease, stroke, and renal disease. J. Am. Coll. Cardiol. (2016.0) 68 1375-1386. DOI: 10.1016/j.jacc.2016.06.054 24. Kernan W.N., Ovbiagele B., Black H.R., Bravata D.M., Chimowitz M.I., Ezekowitz M.D., Fang M.C., Fisher M., Furie K.L., Heck D.V.. Guidelines for the prevention of stroke in patients with stroke and transient ischemic attack: A guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. (2014.0) 45 2160-2223. DOI: 10.1161/STR.0000000000000024 25. Minami J., Kawano Y., Ishimitsu T., Yoshimi H., Takishita S.. Effect of the Hanshin-Awaji earthquake on home blood pressure in patients with essential hypertension. Am. J. Hypertens. (1997.0) 10 222-225. DOI: 10.1016/S0895-7061(96)00328-7 26. Kario K., Matsuo T., Kobayashi H., Yamamoto K., Shimada K.. Earthquake-induced potentiation of acute risk factors in hypertensive elderly patients: Possible triggering of cardiovascular events after a major earthquake. J. Am. Coll. Cardiol. (1997.0) 29 926-933. DOI: 10.1016/S0735-1097(97)00002-8 27. Saito K., Kim J.I., Maekawa K., Ikeda Y., Yokoyama M.. The great Hanshin-Awaji earthquake aggravates blood pressure control in treated hypertensive patients. Am. J. Hypertens. (1997.0) 10 217-221. DOI: 10.1016/S0895-7061(96)00351-2 28. Kario K., Matsuo T., Shimada K., Pickering T.G.. Factors associated with the occurrence and magnitude of earthquake-induced increases in blood pressure. Am. J. Med. (2001.0) 111 379-384. DOI: 10.1016/S0002-9343(01)00832-4 29. Kamoi K., Tanaka M., Ikarashi T., Miyakoshi M.. Effect of the 2004 Mid-Niigata Prefecture earthquake on home blood pressure measurement in the morning in type 2 diabetic patients. Clin. Exp. Hypertens. (2006.0) 28 719-729. DOI: 10.1080/10641960601013575 30. Kario K.. Disaster hypertension— its characteristics, mechanism, and management. Circ. J. (2012.0) 76 553-562. DOI: 10.1253/circj.CJ-11-1510 31. Watanabe K., Tani Y., Tanaka K., Hayashi Y., Asahi K., Nakayama M., Watanabe T.. Acute changes in home blood pressure after the Great East Japan Earthquake among patients with chronic kidney disease in Fukushima City. Clin. Exp. Nephrol. (2013.0) 17 718-724. DOI: 10.1007/s10157-012-0750-z 32. Tani Y., Nakayama M., Tanaka K., Hayashi Y., Asahi K., Kamata T., Ogihara M., Sato K., Matsushima M., Watanabe T.. Blood pressure elevation in hemodialysis patients after the Great East Japan Earthquake. Hypertens. Res. (2014.0) 37 139-144. DOI: 10.1038/hr.2013.116 33. Ohira T., Hosoya M., Yasumura S., Satoh H., Suzuki H., Sakai A., Ohtsuru A., Kawasaki Y., Takahashi A., Ozasa K.. Evacuation and risk of hypertension after the Great East Japan Earthquake: The Fukushima health management survey. Hypertension (2016.0) 68 558-564. DOI: 10.1161/HYPERTENSIONAHA.116.07499 34. Aoki T., Takahashi J., Fukumoto Y., Yasuda S., Ito K., Miyata S., Shinozaki T., Inoue K., Yagi T., Komaru T.. Effect of the Great East Japan Earthquake on cardiovascular disease—-report from the 10 hospitals in the disaster area. Circ. J. (2013.0) 77 490-493. DOI: 10.1253/circj.CJ-12-1594 35. Leor J., Kloner R.A.. The Northridge earthquake as a trigger for acute myocardial infarction. Am. J. Cardiol. (1996.0) 77 1230-1232. DOI: 10.1016/S0002-9149(96)00169-5 36. Suzuki S., Sakamoto S., Miki T., Matsuo T.. Hanshin-Awaji earthquake and acute myocardial infarction. Lancet (1995.0) 345 981. DOI: 10.1016/S0140-6736(95)90727-0 37. Nakamura A., Nozaki E., Fukui S., Endo H., Takahashi T., Tamaki K.. Increased risk of acute myocardial infarction after the Great East Japan Earthquake. Heart Vessels. (2014.0) 29 206-212. DOI: 10.1007/s00380-013-0353-y 38. Yamaki T., Nakazato K., Kijima M., Maruyama Y., Takeishi Y.. Impact of the Great East Japan Earthquake on acute myocardial infarction in Fukushima prefecture. Disaster Med. Public Health Prep. (2014.0) 8 212-219. DOI: 10.1017/dmp.2014.37 39. Tsai C.H., Lung F.W., Wang S.Y.. The 1999 Ji-Ji (Taiwan) earthquake as a trigger for acute myocardial infarction. Psychosomatics (2004.0) 45 477-482. DOI: 10.1176/appi.psy.45.6.477 40. Nakamura M., Tanaka F., Nakajima S., Honma M., Sakai T., Kawakami M., Endo H., Onodera M., Niiyama M., Komatsu T.. Comparison of the incidence of acute decompensated heart failure before and after the major tsunami in Northeast Japan. Am. J. Cardiol. (2012.0) 110 1856-1860. DOI: 10.1016/j.amjcard.2012.08.020 41. Nakano M., Kondo M., Wakayama Y., Kawana A., Hasebe Y., Shafee M.A., Fukuda K., Shimokawa H.. Increased in cidence of tachyarrhythmias and heart failure hospitalization in patients with implanted cardiac devices after the great East Japan earthquake disaster. Circ. J. (2012.0) 76 1283-1285. DOI: 10.1253/circj.CJ-12-0261 42. Sado J., Kiyohara K., Iwami T., Kitamura Y., Ando E., Ohira T., Sobue T., Kitamura T.. Three-year follow-up after the Great East Japan Earthquake in the incidence of out-of-hospital cardiac arrest with cardiac origin. Circ. J. (2018.0) 82 919-922. DOI: 10.1253/circj.CJ-17-1003 43. Kitamura T., Kiyohara K., Iwami T.. The Great East Japan Earthquake and out-of-hospital cardiac arrest. N. Engl. J. Med. (2013.0) 369 2165-2167. DOI: 10.1056/NEJMc1306058 44. Watanabe H., Kodama M., Okura Y., Aizawa Y., Tanabe N., Chinushi M., Nakamura Y., Nagai T., Sato M., Okabe M.. Impact of earthquakes on Takotsubo cardiomyopathy. JAMA (2005.0) 294 305-307. DOI: 10.1001/jama.294.3.305 45. Yamanda S., Hanagama M., Kobayashi S., Satou H., Tokuda S., Niu K., Yanai M.. The impact of the 2011 Great East Japan Earthquake on hospitalisation for respiratory disease in a rapidly aging society: A retrospective descriptive and cross-sectional study at the disaster base hospital in Ishinomaki. BMJ Open (2013.0) 3 e000865. DOI: 10.1136/bmjopen-2012-000865 46. Aoyagi T., Yamada M., Kunishima H., Tokuda K., Yano H., Ishibashi N., Hatta M., Endo S., Arai K., Inomata S.. Characteristics of infectious diseases in hospitalized patients during the early phase after the 2011 great East Japan earthquake: Pneumonia as a significant reason for hospital care. Chest (2013.0) 143 349-356. DOI: 10.1378/chest.11-3298 47. Matsuoka T., Yoshioka T., Oda J., Tanaka H., Kuwagata Y., Sugimoto H., Sugimoto T.. The impact of a catastrophic earthquake on morbidity rates for various illnesses. Public Health (2000.0) 114 249-253. DOI: 10.1016/S0033-3506(00)00339-5 48. Ohkouchi S., Shibuya R., Yanai M., Kikuchi Y., Ichinose M., Nukiwa T.. Deterioration in regional health status after the acute phase of a great disaster: Respiratory physician’ experiences of the Great East Japan Earthquake. Respir. Investig. (2013.0) 51 50-55. DOI: 10.1016/j.resinv.2012.12.003 49. Guidelines for disaster medicine for patients with cardiovascular diseases (JCS 2014/JSH 2014/JCC 2014)—Digest Version. Circ. J. (2016.0) 80 261-284. DOI: 10.1253/circj.CJ-66-0121 50. Moriyama N., Urabe Y., Onoda S., Maeda N., Oikawa T.. Effect of residence in temporary housing after the Great East Ja pan Earthquake on the physical activity and quality of life of older survivors. Disaster Med. Public Health Prep. (2017.0) 11 701-710. DOI: 10.1017/dmp.2017.19 51. Hikichi H., Aida J., Kondo K., Tsuboya T., Matsuyama Y., Subramanian S.V., Kawachi I.. Increased risk of dementia in the aftermath of the 2011 Great East Japan Earthquake and Tsunami. Proc. Natl. Acad. Sci. USA (2016.0) 113 E6911-E6918. DOI: 10.1073/pnas.1607793113
--- title: 'Ethnic minority representation in UK COVID-19 trials: systematic review and meta-analysis' authors: - Mayur Murali - Leher Gumber - Hannah Jethwa - Divolka Ganesh - Jamie Hartmann-Boyce - Harpreet Sood - Francesco Zaccardi - Kamlesh Khunti journal: BMC Medicine year: 2023 pmcid: PMC10049782 doi: 10.1186/s12916-023-02809-7 license: CC BY 4.0 --- # Ethnic minority representation in UK COVID-19 trials: systematic review and meta-analysis ## Abstract ### Background The COVID-19 pandemic has highlighted health disparities affecting ethnic minority communities. There is growing concern about the lack of diversity in clinical trials. This study aimed to assess the representation of ethnic groups in UK-based COVID-19 randomised controlled trials (RCTs). ### Methods A systematic review and meta-analysis were undertaken. A search strategy was developed for MEDLINE (Ovid) and Google Scholar (1st January 2020–4th May 2022). Prospective COVID-19 RCTs for vaccines or therapeutics that reported UK data separately with a minimum of 50 participants were eligible. Search results were independently screened, and data extracted into proforma. Percentage of ethnic groups at all trial stages was mapped against Office of National Statistics (ONS) statistics. Post hoc DerSimonian-Laird random-effects meta-analysis of percentages and a meta-regression assessing recruitment over time were conducted. Due to the nature of the review question, risk of bias was not assessed. Data analysis was conducted in Stata v17.0. A protocol was registered (PROSPERO CRD42021244185). ### Results In total, 5319 articles were identified; 30 studies were included, with 118,912 participants. Enrolment to trials was the only stage consistently reported (17 trials). Meta-analysis showed significant heterogeneity across studies, in relation to census-expected proportions at study enrolment. All ethnic groups, apart from Other ($1.7\%$ [$95\%$ CI 1.1–$2.8\%$] vs ONS $1\%$) were represented to a lesser extent than ONS statistics, most marked in Black ($1\%$ [0.6–$1.5\%$] vs $3.3\%$) and Asian ($5.8\%$ [4.4–$7.6\%$] vs $7.5\%$) groups, but also apparent in White ($84.8\%$ [81.6–$87.5\%$] vs $86\%$) and Mixed $1.6\%$ [1.2–$2.1\%$] vs $2.2\%$) groups. Meta-regression showed recruitment of Black participants increased over time ($$p \leq 0.009$$). ### Conclusions Asian, Black and Mixed ethnic groups are under-represented or incorrectly classified in UK COVID-19 RCTs. Reporting by ethnicity lacks consistency and transparency. Under-representation in clinical trials occurs at multiple levels and requires complex solutions, which should be considered throughout trial conduct. These findings may not apply outside of the UK setting. ### Supplementary Information The online version contains supplementary material available at 10.1186/s12916-023-02809-7. ## Background Since the emergence of the Coronavirus disease (COVID-19) pandemic in January 2020 in the United Kingdom (UK), longstanding health disparities affecting ethnic minority communities have come to light [1]. Emerging evidence has shown that ethnic minorities have had the highest rate of diagnosis [2], severe disease requiring advanced respiratory support [3] and mortality [4–6]. Several reasons for the observed differences have been proposed including higher rates of social deprivation; higher rates of pre-existing health conditions (for example type 2 diabetes and cardiovascular disease); greater frequency of living in large or multi-generational households; and poorer access to health services [7–11]. To combat severe acute respiratory distress syndrome 2 (SARS-CoV-2) and limit its transmission and complications, many randomised controlled trials (RCTs) have been conducted globally to determine effective treatments and develop vaccines which have been subsequently rolled out at a population level. Landmark trials have provided compelling evidence for several vaccines such as BNT162b2 messenger ribonucleic acid (mRNA) [12], ChAdOx1 nCoV-19 (AZD1222) [13] and mRNA-1273 [14], and therapies including dexamethasone [15] and sotrovimab [16]. A well-designed and conducted RCT is considered the gold standard (Level I) to evaluate the causal effect of medical interventions. Like other study types, RCTs also depend upon participation of all groups to improve generalisability and validity of the findings. There is growing concern about the lack of diversity in trials across health and clinical research over the last few years [17–20]. This may stem from anxieties around the implications of participation within ethnic minority communities, added costs of participation (such as travel and parking), language barriers, knowledge gaps and lack of diversity within the research team [21–24]. Given the disproportionate impact of COVID-19 on minority individuals, inclusion of ethnic minority populations in COVID-19 trials is vital to understanding differences of interventions in disease severity and outcomes as well as addressing critical gaps in knowledge. Therefore, this systematic review aimed to assess the representation of different ethnic groups in UK-based COVID-19 vaccine and therapeutic trials and compared them to nationally available data on ethnic minority populations in the UK. ## Methods This systematic review and meta-analysis was conducted and reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines [25] (Supplementary Materials, Appendix 3 and 4). A protocol was registered in advance with the International Prospective Register of Systematic Reviews (PROSPERO, CRD42021244185). Ethical approval was not required. ## Search strategy and selection criteria A search strategy was developed to identify COVID-19 RCTs that were published in MEDLINE (Ovid) and Google Scholar (Supplementary Materials, Appendix 1). This deviated from the search strategy described in the protocol but was considered comprehensive following discussion with a subject librarian and the review team. We included articles available in English, published between 1st January 2020 and 4th May 2022 in peer-reviewed journals. RCTs were included if they [1] explored COVID-19 vaccine or therapies (medical devices and treatments), [2] were conducted in the UK or reported data separately for the UK population and [3] had a minimum sample size of 50 adults (Table 1). The minimum sample size of 50 patients was a pragmatic decision taken by the authors. We restricted to UK-based studies due to the diversity of the population and good quality ethnicity data at the population level. Table 1A summary of the inclusion criteria for the meta-analysis and meta-regressionPICOS criteriaInclusion criteriaPopulationAdults ≥ 18 yearsInterventionCOVID-19 vaccine or therapeutic treatmentComparatorAny or noneOutcomesAnyStudy designRandomised controlled trial (any phase) with a minimum sample size of 50 and conducted in UK or reporting UK dataTime periodPublished in peer-reviewed journal between 1st January 2020 and 4th May 2022 Our main outcome was percentage of each ethnic group at different trial stages, for the following: people approached for inclusion; people screened for inclusion; people determined eligible for inclusion; people determined ineligible for inclusion; people enrolled in the trial; people followed up at primary endpoint; people followed up at longest follow-up. However, on review of the data, it became apparent that we could only assess the percentage of ethnic groups enrolled in the included trials due to lack of data availability, though we could report on the number of trials that reported proportions of ethnic groups at each trial stage. We also investigated the number of trials reporting effect estimates by ethnicity at each trial stage as a secondary outcome. Search results were saved to Rayyan (Qatar Computing Research Institute), a systematic review web-based application. Abstracts were independently screened for inclusion by four review authors (M.M, L.G, H.J and D.G). An online discussion was held between the authors to compare results and adjudicate any discrepancies. Where discrepancies could not be resolved by discussion, they were referred to a second review author. Following exclusion of studies which did not meet the inclusion criteria and duplicates, full-text screening was carried out independently and in duplicate (by M.M, L.G, H.J and D.G) and data was extracted into piloted proforma. Due to the nature of the review questions, we did not assess risk of bias. ## Data analysis Data were collected on participant demographics (age and sex), type of intervention (vaccine or treatment), total number of participants and general study characteristics. To assess our main outcome, we extracted additional data on the reporting of ethnic diversity of participants at each stage of the trial. We also documented the approach to recruitment, and whether efforts were made to recruit from ethnic minority communities. We documented the enrolment period for each trial to investigate whether recruitment of ethnic minorities changed over time. All studies which reported any ethnicity data were included in the final analysis. The percentage of each ethnic group within each trial was calculated as a proportion and mapped against national population statistics using Office of National Statistics (ONS) 2011 data for each outcome. We initially anticipated that heterogeneity in reporting would preclude statistical synthesis and had planned to only calculate percentage of each ethnic group in each trial and map this against national population statistics using forest plots, as reported in our protocol. However, after data extraction, we found more similarities and detailed reporting than we had anticipated, and hence conducted the following post hoc statistical analyses. A DerSimonian-Laird random-effects meta-analysis and a meta-regression to assess changes in recruitment over time were conducted in Stata v 17.0, following a logit transformation of study-specific nonzero proportions and confidence intervals (obtained with Wilson’s method). The meta-regression was conducted to reflect the change in recruitment practices that may have occurred during the pandemic. A $p \leq 0.05$ was deemed indicative of statistical significance. ## Role of the funding source This study was funded by the South Asian Health Foundation. In addition, it was supported by members (K.K, F.Z) of the National Institute for Health Research (NIHR) Applied Research Collaboration East Midlands (ARC-EM). ## Results A total of 5319 studies were identified through the database search, and 5096 studies were excluded during the abstract screening phase. After removal of duplicates and those that did not meet the inclusion criteria, we reviewed 132 trials for full-text screening, after which a further 102 trials were excluded (Fig. 1) and a total of 30 were included (see Table 2 for a list of identified articles and their characteristics). The most common reasons for exclusion at full text stage were because data were not reported by country, meaning UK data could not be extracted, or the included paper did not assess a COVID-19 vaccine or therapeutic. Of the included studies, 21 were for therapeutic trials and 9 were COVID-19 vaccine trials. One study that met the inclusion criteria did not include any data on ethnicity. Fig. 1Study selectionTable 2A total of 30 studies were included in the meta-analysis and meta-regression following a MEDLINE (Ovid) and Google Scholar searchNoTitleAuthorsYearInterventionTotal participantsMean age% maleVaccine/ TherapeuticEnrolment period1Safety and efficacy of NVX-CoV2373 Covid-19 vaccineHeath PT et al2021NVX-CoV237314,$0395651.56\%$VaccineSep 2020—Nov 20202Inhaled budesonide for COVID-19 in people at high risk of complications in the community in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trialYu LM et al2021Budesonide$30066446.57\%$TherapeuticNov 2020–Mar 20213Doxycycline for community treatment of suspected COVID-19 in people at high risk of adverse outcomes in the UK: a randomised, controlled, open-label, adaptive platform trialButler CC et al2021Doxycycline$17926144.16\%$TherapeuticJul 2020–Dec 20204Azithromycin for community treatment of suspected COVID-19 in people at increased risk of an adverse clinical course in the UK (PRINCIPLE): a randomised, controlled, open-label, adaptive platform trialButler CC et al2021Azithromycin$138860.743.22\%$TherapeuticJul–Nov 20205Safety and efficacy of inhaled nebulised interferon beta-1a (SNG001) for treatment of SARS-CoV-2 infection: a randomised, double-blind, placebo-controlled, phase 2 trialMonk PD et al2020Interferon beta-$110157.159.18\%$TherapeuticMar–May 20206Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UKVoysey M et al2021ChAdOx1 nCoV-19 (AZD1222)$754838.69\%$VaccineApr–Nov 20207Efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine against SARS-CoV-2 variant of concern 202,$\frac{012}{01}$ (B.1.1.7): an exploratory analysis of a randomised controlled trialEmary KRW et al2021ChAdOx1 nCoV-19 (AZD1222)$853440.65\%$VaccineMay–Nov 20208Lopinavir-ritonavir in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trialHorby PW et al2020Lopinavir-ritonavir$504066.361.05\%$TherapeuticMar–Jun 20209Inhaled budesonide in the treatment of early COVID-19 (STOIC): a phase 2, open-label, randomised controlled trialRamakrishnan S et al2021Budesonide$1394542.45\%$TherapeuticJul–Dec 202010Convalescent plasma in patients admitted to hospital with COVID-19 (RECOVERY): a randomised controlled, open-label, platform trialRECOVERY collaborative2021Convalescent plasma11,$55863.564.28\%$TherapeuticMay 2020–Jan 202111Tocilizumab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trialRECOVERY collaborative2021Tocilizumab$411663.667.40\%$TherapeuticApr 2020–Jan 202112Human safety, tolerability, and pharmacokinetics of molnupiravir, a novel broad-spectrum oral antiviral agent with activity against SARS-CoV-2Painter WP et al2021Molnupiravir$13038.783.85\%$Therapeutic13Azithromycin in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trialRECOVERY collaborative2021Azithromycin$776365.262.08\%$TherapeuticApr–Nov 202014Single-dose administration and the influence of the timing of the booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine: a pooled analysis of four randomised trialsVoysey M et al2021ChAdOx1 nCoV-19 (AZD1222)$894840.87\%$VaccineApr–Dec 202015Safety and immunogenicity of ChAdOx1 nCoV-19 vaccine administered in a prime-boost regimen in young and old adults (COV002): a single-blind, randomised, controlled, phase $\frac{2}{3}$ trialRamasamy MN et al2020ChAdOx1 nCoV-19 (COV002)$55260.550.18\%$VaccineMay–Aug 202016Safety and immunogenicity of the ChadOx1 nCoV-19 vaccine against SARS-CoV-2: a preliminary report of a phase $\frac{1}{2}$, single-blind, randomised controlled trialFolegatti PM et al2020ChadOx1 nCoV-$1910773550.23\%$VaccineApr–May 202017Dexamethasone in Hospitalized Patients with Covid-19RECOVERY collaborative2021Dexamethasone$642566.163.37\%$TherapeuticMar–June 202018Effect of Hydroxychloroquine in Hospitalized Patients with COVID-19: Preliminary results from a multi-centre, randomized, controlled trialRECOVERY collaborative2020Hydroxychloro- quine$471665.462.21\%$TherapeuticMar–June 202019Effect of noninvasive respiratory strategies on intubation or mortality among patients with acute hypoxaemic respiratory failure and COVID-19: The RECOVERY-RS randomized clinical trialPerkins G et al2022Non-invasive ventilatory strategies$127356.766.30\%$TherapeuticApr 2020–May 202120An online breathing and wellbeing programme (ENO Breathe) for people with persistent symptoms following COVID-19: a parallel group, single blind, randomised controlled trialPhilip KEJ et al2022Online breathing and wellbeing programme$15049.517.69\%$TherapeuticApr–May 202121Colchicine for Covid-19 in the community (PRINCIPLE): a randomised, controlled, adaptive platform trialDorward J et al2022Colchicine138161TherapeuticMar–May 202122Inspiratory muscle training enhances recovery post COVID-19: a randomised clinical trialMcNarry MA et al2022Inspiratory muscles$1484611.49\%$Therapeutic23Casirivimab and imdevimab in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled open-label, platform trialRECOVERY collaborative2022Casirivimab and imdevimab$978562.63\%$TherapeuticSep 2020–May 202124Namilumab or inflixmab compared with standard of care in hospitalised patients with COVID-19 (CATALYST): a randomised, multicentre, multi-arm, multistage, open-label, adaptive, phase 2, proof of concept trialFisher BA et al2022Namilumab or inflixmab$14658.461.64\%$TherapeuticJun 2020–Feb 202125Immunogenicity, safety, and reactogenicity of heterologous COVID-19 primary vaccination incorporating mRNA, viral-vector and protein-adjuvant vaccines in the UK (Com-COV2): a single blind, randomised, phase 2, non-inferiority trialStuart ASV et al2022mRNA, viral-vector and protein-adjuvant vaccines$5326239.47\%$VaccineApr–May 202126Safety and immunogenicity of seven COVID-19 vaccines as a third dose (booster) following two doses of ChAdOx1 nCov-19 or BNT162b2 in the UK (COV-BOOST): a blinded, multicentre, randomised, controlled, phase 2 trialMunro APS et al2021ChAdOx1 nCov-19 or BNT162b$2288349.84\%$VaccineJun 202127Safety and immunogenicity of heterologous versus homologous prime-boost schedules with adenoviral vectored and mRNA COVID-19 vaccine (Com-COV): a single-blind, randomised, non-inferiority trialLiu X et al2021heterologous v homologous prime-boost with adenoviral vectored and mRNA COVID-19 vaccine$46357.854.21\%$VaccineFeb 202128Azithromycin versus standard care in patients with mild-to-moderate COVID-19 (ATOMIC2): an open-label, randomised trialHinks TSC et al2021Azithromycin$29545.951.53\%$TherapeuticJun 2020–Jan 202129Favipiravir, lopinavir-ritonavir or combination therapy (FLARE): a randomised, double blind, 2 × 2 factorial placebo-controlled trial of early antiviral therapy in COVID-19Lowe DM et al2022Favipiravir, lopinavir-ritonavir or combination therapy$2404051.25\%$TherapeuticOct 2020–Nov 202130Aspirin in patients admitted to hospital with COVID-19 (RECOVERY): a randomised, controlled, open-label, platform trialRECOVERY collaborative2022Aspirin14,$89259.261.78\%$TherapeuticNov 2020–Mar 2021 The total number of patients included in the meta-analysis was 118,912. The mean age of participants was 61.1 (range 35–66.3) years, and $55.03\%$ (range 11.49–$83.85\%$) were male. The trials enrolled patients between March 2020 and November 2021. Reporting of ethnicity through different stages of the trial was limited and inconsistent (Table 3). Of the 30 trials in the review, none reported data on those approached for inclusion by ethnicity. No trials reported data on those screened for inclusion by individual ethnicity, though seven trials, all from the RECOVERY (Randomised Evaluation of COVID-19 Therapy) group (https://www.recoverytrial.net/), reported these data as “White”, “BAME” (Black, Asian and minority ethnic) or “Unknown”, which was not further disaggregated. The percentage of “White” participants was under-represented in all 7 trials compared with ONS population statistics at this stage, with approximately a quarter of participants documented as “BAME” or “Unknown” in all trials (Table 4). Similarly, no trials reported data on those either deemed eligible for inclusion or deemed ineligible for inclusion by individual ethnicity, though the same seven trials from the RECOVERY group reported these data as “White”, “BAME” or “Unknown”. The proportion of those deemed eligible was similar to that at the screening stage, while the proportion of those deemed ineligible compared to the screening stage was more varied, with no clear pattern indicating whether “White” or “BAME” participants were more likely to be ineligible. Table 3A summary of the reporting by ethnicity through different stages of COVID-19 clinical trials, including those approached, screened and deemed eligible or ineligible for inclusion; those enrolled in the trial; those followed up at the primary end point and longest follow-up; and reporting of effect estimatesStage of trialNo. of trials reporting by ethnicity (no. participants)ReferencesNo. of trials reporting as “White”, “BAME”, “non-white”, “Other” (no. participants)ReferencesApproached for inclusion00Screened for inclusion07 [60,179][8, 10, 13, 17, 18, 23, 26]Eligible for inclusion07 [60,179][8, 10, 13, 17, 18, 23, 26]Ineligible for inclusion07 [60,179][8, 10, 13, 17, 18, 23, 26]Enrolled in trial17 [52,747][1–4, 6, 7, 12, 14–16, 19, 20, 24, 25, 27–29]11 [64,722][5, 8–11, 13, 17, 18, 23, 26, 30]Followed up at primary endpoint09 [64,434][8–11, 13, 17, 18, 23, 26]Followed up at longest follow-up00Effect estimates011 [79,740][1, 8–11, 13, 17–19, 23, 26]Table 4A summary of the RECOVERY group’s reporting of ethnicity at different trial stages, including those approached for inclusion; those deemed eligible or ineligible for inclusion; and those followed up at the primary endpoint. The RECOVERY trials reported ethnicity as “White”, “BAME” or “non-white”, and did not disaggregate this data furtherTrial noNo. of people approached for inclusion reported by ethnicityNo. of people deemed eligible for inclusion reported by ethnicityNo. of people deemed ineligible for inclusion reported by ethnicityNo. of people followed up at primary endpoint reported by ethnicityWhite (%)BAME (%)Unknown (%)White (%)BAME (%)Unknown (%)White (%)BAME (%)Unknown (%)White (%)BAME (%)Unknown (%)86060 [75]1351 [17]692 [9]3781 [75]865 [17]394 [8]2279 [74]486 [16]298 [10]37818651010,810 [75]2271 [16]1241 [9]8914 [77]1720 [15]924 [8]1896 [69]551 [20]317 [11]89141720113127 [76]732 [18]257 [6]139301 [75]1852 [15]1180 [10]5939 [77]1109 [14]715 [9]3362 [74]743 [16]465 [10]59391109176015 [74]1420 [17]697 [9]4689 [73]1147 [18]589 [9]1326 [78]273 [16]108 [6]4689 [73]1147 [18]589 [9]185860 [73]1365 [18]690 [9]3479 [74]857 [18]380 [8]2381 [74]508 [16]310 [10]34798572310,039 [77]1809 [14]1184 [9]7601 [78]1293 [13]891 [9]2438 [75]516 [16]293 [9]760112933016,189 [76]3260 [15]1900 [9]11,129 [75]2378 [16]1385 [9]5690 [80]882 [12]515 [7]11,1292378 Seventeen trials reported the number of people enrolled in the trial by individual ethnicity, though one of these only reported participants as “White”, “Black” or “Other”. A further eleven grouped enrolled participants into “White” vs “BAME”, “White” vs “non-white” or “White” vs “Other” (in this case meaning all non-White participants). The percentage of those enrolled in these studies is discussed in detail in the meta-analysis. In the seven RECOVERY trials for which there is complete data, there were no loss of participants from those deemed eligible for inclusion and subsequently enrolled in the study. None of the trials reported those followed up at the primary endpoint by individual ethnicity, though eight trials, all from the RECOVERY group, reported this data as “White”, “BAME” or “Unknown”, with none reporting loss to follow-up from those enrolled in the trial. One further trial reported the primary outcome for White participants only. No trials reported by ethnicity at the longest follow-up, though all the RECOVERY trials made reference to “further analyses specified at 6 months”. None of the trials reported effect estimates by ethnicity, though eleven trials report this data for “White”, “BAME” or “Other” groups, and do not disaggregate this further. Three studies documented specific strategies to improve recruitment of ethnic minority groups, and three studies mentioned recruiting from ethnic minority communities, though no details were provided (Table 5). None of these trials recruited a higher proportion of participants from ethnic minority communities compared to those not reporting recruitment strategies. Table 5A summary of the strategies used to improve recruitment of ethnic minority groups in COVID-19 trialsTrial noStrategies to improve recruitment2. Yu LM er al“To increase recruitment from ethnic minority and socially deprived communities, which have been disproportionately affected by COVID-19, we used several outreach strategies, including the appointment in September, 2020, of an expert working with ethnic minorities; active collaboration with community, religious and health organisations; and promotion in multiple languages through a range of media. ”4. Butler CC et al“Given the increased risk from COVID-19 among Black, Asian, and minority ethnic communities, we actively reached out to a range of religious and community organisations at national and regional levels to increase participation from diverse backgrounds. ”19. Perkins G et al.“Collection and reporting of race and ethnicity based on fixed categories and mandated by funder due to disproportionate effect of COVID-19 infection on non-white population.” – no specific recruitment strategy described21. Doward J et al“Several community outreach strategies were implemented aiming to increase recruitment of those from ethnically diverse communities and socioeconomically deprived backgrounds, who have been disproportionally affected by COVID-19.”26. Munro APS et al.“Recruitment of those identifying as black or minority ethnic was particularly encouraged. ”—no specific recruitment strategy described The meta-analysis, summarised in Fig. 2, shows that at enrolment to the trial, all ethnic groups, apart from Other ($1.7\%$ [$95\%$ confidence interval (CI) 1.1–$2.8\%$] vs ONS $1\%$, Fig. 3), were represented to a lesser extent than that suggested by 2011 ONS statistics. This was most marked in the Asian ($5.8\%$ [$95\%$ CI 4.4–$7.6\%$] vs ONS $7.5\%$, Fig. 4) and Black ($1\%$ [$95\%$ CI 0.6–$1.5\%$] vs ONS $3.3\%$, Fig. 5) groups, though also apparent in the Mixed ($1.6\%$ [$95\%$ CI1.2–$2.1\%$] vs ONS $2.2\%$, Fig. 6) and White ($84.8\%$ [$95\%$ CI 81.6–$87.5\%$] vs ONS $86\%$, Fig. 7) ethnic groups. The meta-analysis shows significant variation across studies, in relation to census-expected proportions of patients recruited from different ethnic groups. Fig. 2Summary of meta-analysis of enrolment to trialsFig. 3Seventeen trials documented enrolment of Other participants. Overall effect shows Other participants were over-represented when compared to ONS statistics ($1.7\%$ [$95\%$ CI 1.1–$2.8\%$] vs ONS $1\%$)Fig. 4Sixteen trials documented enrolment of Asian participants. Overall effect shows Asian participants were under-represented when compared to ONS statistics ($5.8\%$ [$95\%$ CI 4.4–$7.6\%$] vs ONS $7.5\%$)Fig. 5Seventeen trials documented enrolment of Black participants. Overall effect shows Black participants were under-represented when compared to ONS statistics ($1\%$ [$95\%$ CI 0.6–$1.5\%$] vs ONS $3.3\%$)Fig. 6Fifteen trials documented enrolment of Mixed participants. Overall effect shows Mixed participants were under-represented when compared to ONS statistics ($1.6\%$ [$95\%$ CI1.2–$2.1\%$] vs ONS $2.2\%$)Fig. 7Twenty-nine trials documented enrolment of White participants. Overall effect shows White participants were under-represented when compared to ONS statistics ($84.8\%$ [$95\%$ CI 81.6–$87.5\%$] vs ONS $86\%$) Figure 8 illustrates the results of the meta-regression, which shows that recruitment in the Black ethnic group improved from May 2020 to June 2021 (from an estimated 0.26 to $1.92\%$, $$p \leq 0.009$$). There were no statistically significant temporal trends in the other groups (Asian ($$p \leq 0.234$$), White ($$p \leq 0.914$$), Other ($$p \leq 0.528$$) and Mixed ($$p \leq 0.722$$).Fig. 8Results of the meta-regression. Recruitment in the Black ethnic group improved from May 2020 to June 2021 (from an estimated $0.26\%$ to $1.92\%$, $$p \leq 0.009$$). There were no statistically significant temporal trends in the other groups ## Discussion We conducted an extensive review of literature to determine any disparities in the representation of ethnic minority groups in UK COVID-19 clinical trials. Our meta-analysis findings demonstrate that in 30 trials of over 100,000 participants, the Asian, Black, Mixed and White ethnic groups were represented to a lesser extent than that suggested by 2011 ONS statistics. There is significant heterogeneity in the proportions of participants recruited from different ethnic groups across studies, though the Asian and Black groups demonstrate the greatest proportion of studies below the percentages demonstrated by the 2011 census. These results might indicate one of two things: first, that Asian and Black groups were enrolled at lower percentages than population averages in more studies than White or Other groups, though whether this interpretation withstands scrutiny is unclear, as the White ethnic group were also not over-represented in the majority of the data. This may lead us to a second conclusion: that Asian, Black and Mixed ethnicities were more likely to be classified as “Other”, grouped into problematic “non-White” or “BAME” categories, or not recorded at all (for example, as “Unknown”). If the first interpretation is correct, it suggests that ethnic minority groups are more likely to be under-represented in COVID-19 trials. This continues a trend of poor recruitment from European trials when compared to North American trials, though neither have shown a temporal improvement in representation of ethnic minority participants [27]. There is also a large body of evidence which has identified previous racial and ethnic enrolment disparities in other types of medical research including trials on cancer, diabetes [20, 28] and cardiovascular disease [29] over the last decade. This may be particularly concerning as the hospital population during the pandemic did not reflect ONS population statistics, with a higher proportion of inpatients from ethnic minority communities [30], in theory providing a larger pool for research teams to recruit from. This raises the important consideration of what population triallists should aspire to map to. Should trial recruitment aim to be representative of the general population, or of the population to whom the interventions are most relevant? If the latter, we might expect vaccine trials, which are designed for whole populations, to map to ONS data, but to use a different reference point for treatments for severe disease which affect a greater proportion of ethnic minority groups, such as tocilizumab for severe COVID-19 infection. If the second interpretation is true, it raises questions about data accuracy and reporting. There is a distinct lack of consistency in the reporting of results by ethnicity, with many studies continuing to use the term “BAME”, which is no longer favoured due to its emphasis on certain ethnic groups, to the exclusion of others [26]. Harmonisation of data is made difficult by differences in ethnicity coding internationally [31], and there have been calls for more detailed and consistent ethnicity coding [32]. Moreover, one could legitimately question the utility of presenting results for genetically, phenotypically, and culturally heterogenous groups under one umbrella (“BAME” or “non-white”), and indeed the authors suggest that in many cases this is to bolster numbers, increasing the likelihood that subgroup analyses are statistically significant. It is important to clarify why all groups, apart from “Other” appear to be under-represented in the data. The weighted averages do not include individuals grouped under terms such as “BAME”, “non-White” or “Other” (referencing non-White), as this disaggregated data was not available. Therefore, all groups appear to be under-represented as a proportion of the total. The meta-regression showed that over the course of the pandemic, recruitment in the Black group improved over the study period ($$p \leq 0.009$$), while no significant temporal trends were seen in the other ethnic groups. Improved recruitment amongst the Black ethnicity could be due to the recognition that COVID-19 disproportionately affected ethnic minorities, leading to calls to increase and encourage recruitment from these communities [33], as illustrated in Table 5. Recruitment to trials, however, is far from the only issue. Our findings show limited reporting by ethnicity at all stages. Enrolment to trials was best reported, with seventeen of the 30 studies breaking participants down by individual ethnic groups. The meta-analysis highlighted that while twenty-nine reported participant enrolment for the White ethnicity, only 17, 16 and 15 studies reported participant enrolment for the Black, Asian and Mixed ethnicity groups respectively. Some of the other studies grouped individuals from minority communities as “BAME” and in these studies “BAME” representation was higher than UK ONS data (if including Asian, Black, Mixed and Other groups from ONS statistics, though such definitions in these studies were not always clear). It is important to highlight that no information was available on the representation of individual ethnicities within these studies, or indeed the “non-White” grouping used by other studies. The over-representation of this grouping of ethnicities is in direct contrast to the studies that reported enrolment data by individual ethnic groups, where a lower proportion of Black, Asian and Mixed participants were enrolled when compared to ONS data in the majority of the studies ($\frac{14}{17}$, $\frac{12}{16}$ and $\frac{12}{15}$, respectively). Understanding enrolment disparities and data absenteeism in RCTs is vital as a lack of diversity can bias the results and limit generalisability to underrepresented populations. It is acknowledged that genetic polymorphisms can affect responses to vaccines [34] and medicinal therapeutics, such as antihypertensives, heart failure medications and warfarin [35]. Interventions which have been predominantly tested in White populations may not be as effective in other ethnicities [20, 36], and indeed a lack of representation in trials for vaccines and therapeutics fuels mistrust and vaccine hesitancy amongst minority communities [22]. A variety of reasons for the underrepresentation of ethnic minorities have been proposed. These can broadly be grouped into three categories: those occurring at system, individual and interpersonal levels. Barriers at the healthcare system and hospital level include restrictive study designs, financial costs associated with running trials and lack of community engagement. Commonly reported individual barriers revolve around lack of comfort, lack of knowledge on the research process and the study, logistics and time and resource constraints [37]. Doctor-patient relationships, including a lack of support if problems arose [22], language barriers [23], and mistrust (particularly suspicions of a hidden agenda [22]) play an important role at an interpersonal level [38, 39]. Overcoming these barriers is key to improving recruitment and participation in medical research, and tailored strategies will need to be implemented to improve participation in research of ethnic minority groups. A one-size fits all approach is inadequate as barriers identified vary between community groups [22]. These issues need to be approached at the conceptualisation of a trial, and inclusion of ethnic minorities should be considered at all stages of the research process [23]. Previous studies have shown that community-based [40, 41] and multimedia interventions [42] can be effective in increasing participation in research. Despite this, we found only three studies designed specific strategies to improve recruitment of ethnic minority groups (Table 5). None of these trials recruited a higher proportion of participants from ethnic minority communities, suggesting the issue requires a more complex solution. Two further studies made mention of recruiting from Black, Asian or minority ethnic communities, one mandated by the funder; however, no details were made available for how this was pursued. The Consolidated Standard of Reporting Trials (CONSORT) statement provides guidance to researchers on reporting findings from RCTs. It advocates providing baseline demographic and clinical features of participants [43]. However, it does not specify which sociodemographic characteristics should be captured or presented. In order to combat underrepresentation, the National Institute for Health and Care Research recently developed the Innovations in Clinical Trial Design and Delivery for the Under-served (INCLUDE) framework, a tool that provides specific guidance to researchers on improving recruitment of minority groups [44]. Such an approach appears to have borne fruit in North America, where the Food and Drug Administration (FDA)-published guidance emphasises collecting racial and ethnic data in clinical trials, and appears to have improved the collection of such data [27]. In addition to using such frameworks, we call for consistency in the reporting of ethnicity in randomised controlled trials; first, in the terminology used to describe ethnic groups, both in the UK and internationally, to enable comparison of data across trials [45]. Second, to avoid grouping multiple ethnic groups under one term, ensuring better reporting of effect estimates by different ethnic groups. And third, collecting and reporting ethnicity data at every stage of a trial, increasing transparency and improving our understanding of where failures to recruit or retain participants occur. Our study had several strengths. Our search strategy was robust, and our data collection methods were piloted and rigorous. Each full-text article was reviewed by two authors. We performed a meta-analysis and meta-regression, which provided a better estimate of the percentage and improved the generalisability of our findings. Limitations include that our systematic review assessed only UK-based studies, thus our findings are not generalisable to other countries. However, this was to allow comparison of our findings with high-quality data on proportion of patients from different ethnic groups in our national datasets. Second, some multinational studies that recruited from UK centres did not break down their results by individual country, meaning these were excluded, which may have biased the results. Third, the least biased pooled estimate for the analyses was for the White ethnicity, which was available for all studies; for other ethnicities, missing data may have impacted on the pooled estimate (for example, where ethnicities have been grouped into “BAME” rather than reported separately). Fourth, the population statistics we used were based on 2011 ONS data, which may not reflect today’s population statistics. Data from the 2021 census is awaited, but it is likely that changes to the demographic composition of the UK in this time will shine a harsher light on poor representation from ethnic minority communities. A further sub-analysis of interest is an examination of the characteristics of the organisation hosting the trial, and the region where the study took place, as there is significant heterogeneity in ethnic diversity in the UK. However, many of the studies we included were national or multi-site and did not disaggregate their results by region or site, making such an analysis impossible. This represents an area for future research. Finally, we have used one approach to the analysis of proportions, while others are available. ## Conclusions This systematic review of 30 trials with over 100,000 participants shows that Asian, Black and Mixed ethnic groups are either under-represented to a greater extent or incorrectly documented as “BAME”, “Other”, “non-white” or “Unknown” in UK COVID-19 RCTs. Underrepresentation in clinical trials occur at the system, individual and interpersonal level and require complex solutions, which should be approached at trial conception and considered throughout the research process. Reporting of trials by ethnicity lacks consistency, with obsolete terminology in common use, and grouping of multiple ethnicities commonplace despite genetic and phenotypic differences. Few trials report specific methods for recruiting participants from ethnic minorities. Those conducting trials need to make use of available frameworks for recruiting patients and report data that are consistent in terminology and have greater transparency. ## Supplementary Information Additional file 1: Appendix 1. Search strategy. Appendix 2. Numbers used for the analysis. Appendix 3. PRISMA abstract checklist. Appendix 4. PRISMA checklist. Appendix 5. PRISMA 2020 flow diagram for new systematic reviews which included searches of databases and registries only. ## References 1. 1.Khunti K, Singh AK, Pareek M, et al. Is ethnicity linked to incidence or outcomes of covid-19? BMJ. 2020;369:m1548. 2. Mulholland RH, Sinha IP. Ethnicity and COVID-19 infection: are the pieces of the puzzle falling into place?. BMC Med (2020.0) 18 206. DOI: 10.1186/s12916-020-01669-9 3. 3.Intensive Care National Audit & Research Centre (ICNARC). ICNARC report on COVID-19 in critical care: England, Wales & Northern Ireland. 23 October 2020. ICNARC Case Mix Programme Database, Intensive Care National Audit & Research Centre. 2020. 4. 4.Public Health England. COVID-19: review of disparities in risks and outcomes. 2020. https://www.gov.uk/government/publications/covid-19-review-of-disparities-in-risks-and-outcomes. Accessed 27 Jan 2022. 5. Williams DR, Cooper LA. COVID-19 and Health Equity—A New Kind of “Herd Immunity”. JAMA (2020.0) 323 2478. DOI: 10.1001/jama.2020.8051 6. Pan D, Sze S, Minhas JS. The impact of ethnicity on clinical outcomes in COVID-19: A systematic review. EClinicalMedicine (2020.0) 23 100404. DOI: 10.1016/j.eclinm.2020.100404 7. Chen JT, Krieger N. Revealing the Unequal Burden of COVID-19 by Income, Race/Ethnicity, and Household Crowding: US County Versus Zip Code Analyses. J Public Health Manag Pract (2021.0) 27 S43-S56. DOI: 10.1097/PHH.0000000000001263 8. 8.Razai MS, Kankam HKN, Majeed A, et al. Mitigating ethnic disparities in covid-19 and beyond. BMJ. 2021;372:m4921. 9. 9.New Policy Institute. Accounting for the Variation in the Confirmed Covid-19 Caseload across England: An analysis of the role of multi-generation households, London and time. New Policy Institute. 2020. https://www.npi.org.uk/publications/housing-and-homelessness/accounting-variation-confirmed-covid-19-caseload-across-england-analysis-role-multi-generation-households-london-and-time. Accessed 27 Jan 2022. 10. Katikireddi SV, Lal S, Carrol ED. Unequal impact of the COVID-19 crisis on minority ethnic groups: a framework for understanding and addressing inequalities. J Epidemiol Community Health (2021.0) 75 970. DOI: 10.1136/jech-2020-216061 11. Nafilyan V, Islam N, Ayoubkhani D. Ethnicity, household composition and COVID-19 mortality: a national linked data study. J R Soc Med (2021.0) 114 182-211. DOI: 10.1177/0141076821999973 12. Polack FP, Thomas SJ, Kitchin N. Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine. N Engl J Med (2020.0) 383 2603-2615. DOI: 10.1056/NEJMoa2034577 13. Voysey M, Clemens SAC, Madhi SA. Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK. The Lancet (2021.0) 397 99-111. DOI: 10.1016/S0140-6736(20)32661-1 14. Baden LR, El Sahly HM, Essink B. Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine. N Engl J Med (2021.0) 384 403-416. DOI: 10.1056/NEJMoa2035389 15. Dexamethasone in Hospitalized Patients with Covid-19. N Engl J Med (2021.0) 384 693-704. DOI: 10.1056/NEJMoa2021436 16. Gupta A, Gonzalez-Rojas Y, Juarez E. Early Treatment for Covid-19 with SARS-CoV-2 Neutralizing Antibody Sotrovimab. N Engl J Med (2021.0) 385 1941-1950. DOI: 10.1056/NEJMoa2107934 17. Shaw AR, Perales-Puchalt J, Johnson E. Representation of Racial and Ethnic Minority Populations in Dementia Prevention Trials: A Systematic Review. J Prev Alzheimers Dis (2021.0) 9 1-6 18. Hoel AW, Kayssi A, Brahmanandam S. Under-representation of women and ethnic minorities in vascular surgery randomized controlled trials. J Vasc Surg (2009.0) 50 349-354. DOI: 10.1016/j.jvs.2009.01.012 19. Berger JS, Melloni C, Wang TY. Reporting and representation of race/ethnicity in published randomized trials. Am Heart J (2009.0) 158 742-747. DOI: 10.1016/j.ahj.2009.08.018 20. Khunti K, Bellary S, Karamat MA. Representation of people of South Asian origin in cardiovascular outcome trials of glucose-lowering therapies in Type 2 diabetes. Diabet Med (2017.0) 34 64-68. DOI: 10.1111/dme.13103 21. Yancey AK, Ortega AN, Kumanyika SK. Effective recruitment and retention of minority research participants. Annu Rev Public Health (2006.0) 27 1-28. DOI: 10.1146/annurev.publhealth.27.021405.102113 22. Ekezie W, Czyznikowska BM, Rohit S. The views of ethnic minority and vulnerable communities towards participation in COVID-19 vaccine trials. J Public Health (2021.0) 43 e258-e260. DOI: 10.1093/pubmed/fdaa196 23. Bodicoat DH, Routen AC, Willis A. Promoting inclusion in clinical trials—a rapid review of the literature and recommendations for action. Trials (2021.0) 22 880. DOI: 10.1186/s13063-021-05849-7 24. Willis A, Isaacs T, Khunti K. Improving diversity in research and trial participation: the challenges of language. Lancet Public Health (2021.0) 6 e445-e446. DOI: 10.1016/S2468-2667(21)00100-6 25. Page MJ, McKenzie JE, Bossuyt PM, The PRISMA. statement: an updated guideline for reporting systematic reviews. BMJ (2020.0) 2021 n71 26. 26.Khunti K, Routen A, Pareek M, et al. The language of ethnicity. BMJ. 2020;m4493. 27. Zhang J, Van Spall HGC, Wang Y. Enrollment of black, indigenous, and other people of color in multicountry randomized controlled trials of diabetes conducted in North America and Europe. Diabetes Care (2022.0) 45 e116-e117. DOI: 10.2337/dc22-0261 28. Bowe T, Salabati M, Soares RR. Racial, Ethnic, and gender disparities in diabetic macular edema clinical trials ophthalmol. Retina (2022.0) 6 531-533. PMID: 35131526 29. Tahhan AS, Vaduganathan M, Greene SJ. Enrollment of older patients, women, and racial/ethnic minority groups in contemporary acute coronary syndrome clinical trials: a systematic review. JAMA Cardiol (2020.0) 5 714. DOI: 10.1001/jamacardio.2020.0359 30. 30.Gardiner T, Cooke G, Fidler S, et al. The under-representation of BAME patients in the COVID-19 Recovery trial at a major London NHS Trust. J Infect. 2020;82(4);84-123. 31. Routen A, Akbari A, Banerjee A. Strategies to record and use ethnicity information in routine health data. Nat Med (2022.0) 28 1338-1342. DOI: 10.1038/s41591-022-01842-y 32. Khunti K, Routen A, Banerjee A. The need for improved collection and coding of ethnicity in health research. J Public Health (2021.0) 43 e270-e272. DOI: 10.1093/pubmed/fdaa198 33. Treweek S, Forouhi NG, Narayan KMV. COVID-19 and ethnicity: who will research results apply to?. The Lancet (2020.0) 395 1955-1957. DOI: 10.1016/S0140-6736(20)31380-5 34. Jethwa H, Wong R, Abraham S. Covid-19 vaccine trials: Ethnic diversity and immunogenicity. Vaccine (2021.0) 39 3541-3543. DOI: 10.1016/j.vaccine.2021.05.017 35. Johnson JA. Ethnic Differences in Cardiovascular Drug Response: Potential Contribution of Pharmacogenetics. Circulation (2008.0) 118 1383-1393. DOI: 10.1161/CIRCULATIONAHA.107.704023 36. Zhang J, Van Spall HGC, Wang Y. Twenty-year trends in racial and ethnic enrollment in large diabetes randomized controlled trials. BMC Med (2022.0) 20 294. DOI: 10.1186/s12916-022-02501-2 37. Clark LT, Watkins L, Piña IL. Increasing Diversity in Clinical Trials: Overcoming Critical Barriers. Curr Probl Cardiol (2019.0) 44 148-172. DOI: 10.1016/j.cpcardiol.2018.11.002 38. Hamel LM, Penner LA, Albrecht TL. Barriers to Clinical Trial Enrollment in Racial and Ethnic Minority Patients With Cancer. Cancer Control J Moffitt Cancer Cent (2016.0) 23 327-337. DOI: 10.1177/107327481602300404 39. Clark LT, Watkins L, Piña IL. Increasing Diversity in Clinical Trials: Overcoming Critical Barriers. Curr Probl Cardiol (2019.0) 44 148-172. DOI: 10.1016/j.cpcardiol.2018.11.002 40. McCaskill-Stevens W, McKinney MM, Whitman CG. Increasing minority participation in cancer clinical trials: the minority-based community clinical oncology program experience. J Clin Oncol (2005.0) 23 5247-5254. DOI: 10.1200/JCO.2005.22.236 41. Duda C, Mahon I, Chen MH. Impact and costs of targeted recruitment of minorities to the National Lung Screening Trial. Clin Trials (2011.0) 8 214-223. DOI: 10.1177/1740774510396742 42. Jacobsen PB, Wells KJ, Meade CD. Effects of a brief multimedia psychoeducational intervention on the attitudes and interest of patients with cancer regarding clinical trial participation: a multicenter randomized controlled trial. J Clin Oncol (2012.0) 30 2516-2521. DOI: 10.1200/JCO.2011.39.5186 43. Moher D, Hopewell S, Schulz KF. CONSORT 2010 Explanation and elaboration: updated guidelines for reporting parallel group randomised trials. BMJ (2010.0) 340 c869-c869. DOI: 10.1136/bmj.c869 44. Treweek S, Banister K, Bower P. Developing the INCLUDE Ethnicity Framework—a tool to help trialists design trials that better reflect the communities they serve. Trials (2021.0) 22 337. DOI: 10.1186/s13063-021-05276-8 45. 45.Routen A, Akbari A, Banerjee A, et al. Strategies to record and use ethnicity information in routine health data. Nat Med. Epub ahead of print 31 May 2022. 10.1038/s41591-022-01842-y.
--- title: Diet Control and Swimming Exercise Ameliorate HFD-Induced Cognitive Impairment Related to the SIRT1-NF-κB/PGC-1α Pathways in ApoE-/- Mice authors: - Wei Wei - Zhicheng Lin - PeiTao Xu - Xinru Lv - Libin Lin - Yongxu Li - Yangjie Zhou - Taotao Lu - Xiehua Xue journal: Neural Plasticity year: 2023 pmcid: PMC10049848 doi: 10.1155/2023/9206875 license: CC BY 4.0 --- # Diet Control and Swimming Exercise Ameliorate HFD-Induced Cognitive Impairment Related to the SIRT1-NF-κB/PGC-1α Pathways in ApoE-/- Mice ## Abstract High-fat diet- (HFD-) induced neuroinflammation may ultimately lead to an increased risk of cognitive impairment. Here, we evaluate the effects of diet control and swimming or both on the prevention of cognitive impairment by enhancing SIRT1 activity. Twenty-week-old ApoE-/- mice were fed a HFD for 8 weeks and then were treated with diet control and/or swimming for 8 weeks. Cognitive function was assessed using the novel object recognition test (NORT) and Y-maze test. The expression of sirtuin-1 (SIRT1), peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1α), brain-derived neurotrophic factor (BDNF), nuclear factor kappa B p65 (NF-κB p65), interleukin-1β (IL-1β), and tumour necrosis factor-α (TNF-α) in the hippocampus was measured by western blotting. The levels of fractional anisotropy (FA), N-acetylaspartate (NAA)/creatine (Cr) ratio, choline (Cho)/Cr ratio, and myo-inositol (MI)/Cr ratio in the hippocampus were evaluated by diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MRS) using 7.0-T magnetic resonance imaging (MRI). Our results showed that cognitive dysfunction and hippocampal neuroinflammation appeared to be remarkably observed in apolipoprotein E (ApoE)-/- mice fed with HFD. Diet control plus swimming significantly reversed HFD-induced cognitive decline, reduced the time spent exploring the novel object, and ameliorated spontaneous alternation in the Y-maze test. Compared with the HFD group, ApoE-/- mice fed diet control and/or subjected to swimming had an increase in FA, NAA/Cr, and Cho/Cr; a drop in MI/Cr; elevated expression levels of SIRT1, PGC-1α, and BDNF; and inhibited production of proinflammatory cytokines, including NF-κB p65, IL-1β, and TNF-α. SIRT1, an NAD+-dependent class III histone enzyme, deacetylases and regulates the activity of PGC-1α and NF-κB. These data indicated that diet control and/or swimming ameliorate cognitive deficits through the inhibitory effect of neuroinflammation via SIRT1-mediated pathways, strongly suggesting that swimming and/or diet control could be potentially effective nonpharmacological treatments for cognitive impairment. ## 1. Introduction Cognitive impairment has become increasingly commonplace and has been regarded as one of the major health challenges in the past several decades [1], increasing personal, social, and economic burdens. Due to the lack of effective medical therapeutic approaches, patients in advanced stages of cognitive dysfunction are getting worse and not being controlled. Therefore, nonpharmacologic approaches to maintain and improve cognitive decline are drawing increasing attention, with interest in healthy diet and lifestyle behaviours that enhance memory formation [2]. Accumulating evidence has shown that a HFD and less physical activity result directly in cognitive dysfunction [3, 4]. Although cognitive impairment is not the inevitable consequence of an unhealthy lifestyle, early neuroprotection against cognitive dysfunction through therapeutic lifestyle modifications may slow the development of cognitive decline [5]. These findings suggest that a healthy lifestyle, such as physical exercise and diet control, may be the major nonpharmacological means for improving cognitive impairment. The relationship between an unhealthy dietary lifestyle and cognitive impairment is worthy of further discussion. It has been reported that a HFD harms the ultrastructure and function of the hippocampus and promotes the decline of cognitive function [6, 7], which is closely associated with neuroinflammation [8]. Aberrantly expressed neuroinflammation factors (e.g., TNF-α and IL-1β) are found in the hippocampus of HFD-fed mice and potentially lead to memory and learning performance loss [9, 10]. The same neuropathological change occurs in individuals with physical inactivity and poor physical performance [11]. Physical activity is an integral part of affecting human growth and overall health [12] and has a positive effect on the regulation of cognitive function [13]. It was demonstrated that aerobic exercise could upregulate the expression of synaptic plasticity-related proteins through inhibitory NF-κB and IL-1β release in the hippocampus, suggesting that aerobic exercise might alleviate neuroinflammation to improve cognitive function [14]. SIRT1 is an NAD+-dependent class III histone enzyme deacetylase involved in cellular senescence, metabolism homeostasis, neuroinflammation, and ageing by activating PGC-1α/BDNF [15] and interrupting the NF-κB pathway [16]. PGC-1α is a transcriptional activator and plays a crucial role in mitochondria–related energy metabolism [17]. Activation of PGC-1α depends deeply on SIRT1-mediated deacetylation [18]. Deacetylation of PGC-1α upregulates the synthesis and secretion of BDNF, a beneficial factor of brain function, to inhibit cognitive decline [19]. Exercise and dietary restriction elevated the expression of PGC-1α and BDNF, which contributed to enhancing brain function and improving cognitive decline [20]. In addition, SIRT1 deacetylates histones in the promoter region of NF-κB p65 and inhibits the release of the inflammation-related factors IL-1β and TNF-α, ultimately reducing inflammation [21, 22]. Our previous study confirmed that swimming exercise and diet control attenuated HFD-induced learning and memory deficits, which was closely associated with high expression of SIRT1 and low expression of NF-κB p65 [23]. Thus, we hypothesized that diet control and/or physical exercise had a positive impact on the improvement in cognitive deficits related to the SIRT1-mediated pathways. ApoE can effectively regulate the redistribution of lipoprotein and cholesterol. It is the main apolipoprotein of lipid metabolism in the central nervous system and plays an important neuroprotective role. ApoE knockout (ApoE-/-) mice fed a high-fat diet are more prone to metabolic disorders in vivo than wild-type mice. The association of diet control and swimming exercise with high-fat diet-fed ApoE-/- mice and the SIRT1-NF-κB/PGC-1α pathway has not been thoroughly investigated. Therefore, this study established the APOE-/- mouse high-fat model to explore the correlation between diet control and swimming exercise and between the high-fat diet ApoE-/- mice and the SIRT1-NF-κB/PGC-1α pathway. ## 2.1. Animals ApoE-/- male mice (aged 20 weeks and weighing 28-34 g) were purchased from Nanjing University-Nanjing Institute of Biomedicine. All mouse experimental protocols were approved by the Animal Management System and Use Committee of Fujian University of Traditional Chinese Medicine (permission number: FJTCMIACUC2020091). The experimental procedures were strictly in accordance with the international animal provisions of the protection and use guidelines. All animals were housed in group cages under a 12 h dark-light cycle at 23 ± 2°C with free access to food and water. ## 2.2. Reagents and Instruments The instruments and reagents used in these experiments were as follows: a 7.0T animal magnetic resonance instrument (Bruker, Germany); novel object recognition and test system (Boster Bioengineering, China); Y labyrinth video analysis system (Shanghai Xinruan Information Technology, China); high-fat feed ($21\%$ fat, $0.15\%$ cholesterol, Jiangsu Medisen Biomedicine, China); isoflurane (Shenzhen Reward, China); immunohistochemistry kit and DAB staining kit (Boster Bioengineering, China); and haematoxylin staining solution and eosin staining solution (Beijing Solebao Technology, China). The primary antibodies in this study were as follows: anti-SIRT1 (Proteintech, USA), anti-TNF-α (Proteintech, USA), anti-β-actin (Proteintech, USA), anti-IBA1 (Proteintech, USA), anti-GFAP (Proteintech, USA), anti-PGC-1α (Abcam, USA), anti-BDNF (Abcam, USA), anti-IL-1β (Abcam, USA), NF-κBp65 antibody (Cell Signaling Technology, USA), goat anti-mouse/anti-rabbit secondary antibody (Proteintech, USA), and antibody diluent (Beijing Biyuntian, China). ## 2.3. Grouping Experimental grouping was performed as described previously [23]. ApoE-/- mice were randomly divided into 5 groups (5 in each group) as follows: control group (CON), high-fat diet group (HFD), diet control group (DC), swimming exercise group (SE), and diet control + swimming exercise group (DS). After a week of adjustable feeding, the mice in the HFD, DC, SE, and DS groups were fed a high-fat diet ($21\%$ fat, $0.15\%$ cholesterol), and the control group was given a normal standard diet for 8 weeks. After that, the HFD group and SE group continued to be fed a high-fat diet, and the remaining groups were switched to a normal standard diet. ## 2.4. Swimming Exercise The swimming exercise protocol consisted of two phases: adaptation and training [24]. The time-adaptation phase lasted for the first week, and then, all mice were subjected to swimming training for 50 minutes once daily, 6 days per week for a total of 7 weeks. ApoE-/- mice exercised in the round pool (100 × 80 × 60 cm, 35-36°C) for 10 min on the first day of the adaptation phase. The swimming period was extended by 10 min every day until ApoE-/- mice were able to swim for 50 min per day. ## 2.5. NORT NORT is a simple way to evaluate recognition memory via the difference in the exploration time of novel objects, as described below. The size of the novel object device is an open box with dimensions of 72 × 72 × 25 cm. Three prepared objects, namely, A1, A2, and B, were used, where the size and shape of objects A1 and A2 are the same but different from object B. There are two phases in NORT. In the first phase (acceptance and adaptation phase), each mouse was free to explore for ten minutes in the open box with objects A1 and A2 placed into the symmetric corners on the same side. In the second phase (test phase), A2 was replaced with B, and the mice were allowed to explore freely for 5 min after 1 h and 24 h of habituation. The observation result is expressed as the discrimination index using the following formula: discrimination index = the number of times exploring object B/total number of times exploring objects A1 and B. ## 2.6. Y-Maze Test Y-maze spontaneous alternations were performed to evaluate short-term memory. The Y-labyrinth apparatus contained three equal-size black arms with a 120° angle between two adjacent arms. The dimensions of each arm were 30 cm in length × 8 cm in width × 15 cm in height. The mice were placed in the intersection of the three arms and allowed to move freely for 8 minutes. The number of times the mouse entered the different arms was continuously recorded. The result was expressed by the spontaneous alternate reaction rate using the following formula: the maximum number of alternations/(total number of arms − 2) × $100\%$. ## 2.7. MRI The mice were anaesthetized with 2-$3\%$ isoflurane inhalation and placed in a stable and comfortable prone position. The heart rate and body temperature were determined with a detector placed on the abdomen and a thermometer inserted into the rectum. Head coils were applied to fix the mice during the magnetic resonance scan. After adjusting to a suitable position, T2W1 and DTI were performed on the brains of the mice using the RARE sequence. The T2W1 scanning parameters were as follows: TE = 35 ms, TR = 4200 ms, averages = 4, slice thickness = 0.5 mm, and field of view = 20 mm × 20 mm. The EPI program for DTI scanning had the following parameters: TE = 25 ms, TR = 12000 ms, averages = 2, slice thickness = 0.5 mm, and field of view = 20 mm × 20 mm. Finally, MRS scanning was performed. The bilateral hippocampus was selected as the region of interest on the T2-weighted transverse, coronal, and sagittal planes, with a size of 1 mm × 1 mm × 1 mm. After shimming and suppressing water, the FWHM was less than 20 before scanning. The MRS-specific parameters were as follows: TE = 144 ms, TR = 1500 ms, and averages = 256. ## 2.8. Western Blot Hippocampal tissue was isolated from mouse brains and lysed with RIPA lysis buffer, and the supernatant was obtained and centrifuged at 12,000 × g for 10 min at 4°C. A BCA quantitative kit was used to determine the protein concentration of the obtained supernatant according to the manufacturer's directions. Fifty micrograms of protein from each group was loaded and separated by electrophoresis on $8\%$, $10\%$, and $12\%$ SDS–PAGE gels, followed by transfer onto PVDF membranes (0.22 μm). After blocking for 2 h at 25°C with $5\%$ skim milk, the membrane was incubated with antibodies (SIRT1 1: 1000, PGC-1α 1: 1000, BDNF 1: 1000, NF-κB 1: 1000, IL-1β 1: 1000, and TNF-α 1: 1000) overnight at 4°C. On the following day, the membrane was blotted with the corresponding diluent of the secondary antibody for 1 h at 4°C after washing with TBST solution for 3 times. The ultrasensitive ECL luminescent developer was added to the bands, and the data were analysed by Image Lab software. ## 2.9. Statistical Analysis The data in this study are presented as the mean ± SEM and were analysed using SPSS 22.0 statistical software. One-way ANOVA or T test was applied for comparisons among groups. $p \leq 0.05$ indicated that the differences were statistically significant. All experiments were repeated at least five times. ## 3.1. Diet Control and/or Swimming Exercise Reduced HFD-Induced Weight Gain ApoE-/- mice aged 20 weeks were fed a HFD and started diet control and/or received swimming exercise at 28 weeks (Figure 1(b)). The body weight of ApoE-/- mice was measured weekly. Significantly higher body weight gain of mice was found in the HFD group than in the CON group ($p \leq 0.05$) (Figure 1(a)). The increased body weight gain was gradually reduced at 28 weeks in the DC, SE, and DS groups in contrast to the HFD group ($p \leq 0.05$) (Figure 1(a)). The lowest body weight gain was found in the DS group (Figure 1(a)). ## 3.2. Diet Control and/or Swimming Exercise Ameliorated HFD-Induced Cognitive Deficits To verify whether diet control and/or swimming exercise contributed to improve HFD-induced cognitive impairment. NORT was used to evaluate recognition memory performance. The HFD group displayed a lower 1 h discrimination index (1 h DI) and 24 h discrimination index (24 h DI) than the CON group ($p \leq 0.05$) (Figure 2(a)). The 1 h DI and 24 h DI increased gradually in the DC group, SE group, and DS group ($p \leq 0.05$) (Figure 2(a)). The 1 h DI and 24 h DI of the DS group were higher than those of the SE and DC groups ($p \leq 0.05$) (Figure 2(a)). The Y-maze test was used to assess working memory (Figure 2(b)). The spontaneous alternate reaction rate of the HFD group was significantly reduced compared with that of the CON group ($p \leq 0.05$) (Figure 2(b)). The spontaneous alternate reaction rate increased gradually in the DC, SE, and DS groups compared to the HFD group ($p \leq 0.05$) (Figure 2(b)). The results demonstrated that consumption of a HFD for a long time leads to memory and learning performance loss, and diet control and/or swimming exercise significantly alleviated cognitive decline in ApoE-/- mice. ## 3.3. Diet Control and/or Swimming Exercise Ameliorated the HFD-Induced Decrease in FA in the Hippocampus It has been demonstrated that DTI is a novel MRI technique in the hippocampus to detect the impact of damage on white matter structural integrity, which is closely correlated with cognitive decline. High FA indicates that the white matter tracts are intact. In this study, a significant decrease in FA levels was found in the bilateral HFD-fed ApoE-/- mouse hippocampus compared with the CON group ($p \leq 0.05$) (Figures 3(a) and 3(b)). There was a remarkable upward trend of FA in the DC, SE, and DS groups ($p \leq 0.05$) (Figures 3(a) and 3(b)). The data suggested that diet control and/or swimming exercise significantly alleviated cognitive decline in ApoE-/- mice by maintaining the structural integrity of white matter in the hippocampus. ## 3.4. Diet Control and/or Swimming Exercise Ameliorated HFD-Induced Neurometabolic Abnormalities in the Hippocampus Structural-metabolic variations in the hippocampus are responsible for the initiation of cognitive dysfunction. To clearly explore the alterations in metabolites, magnetic resonance spectroscopy was applied to evaluate NAA/Cr, Cho/Cr, and MI/Cr (Figure 4(a)). There were significant decreases in NAA/Cr and Cho/Cr and an increase in MI/Cr in the bilateral HFD-fed ApoE-/- mouse hippocampus compared with the CON group ($p \leq 0.05$) (Figures 4(b)–4(g)). NAA/Cr and Cho/Cr were higher in the DC group, SE group, and DS group than in the HFD group ($p \leq 0.01$, $p \leq 0.05$) (Figures 4(b)–4(e)). Moreover, MI/Cr decreased gradually in the DC, SE, and DS groups in contrast to that in the HFD group ($p \leq 0.01$, $p \leq 0.05$) (Figures 4(f) and 4(g)). These results demonstrated that the neurometabolic abnormalities of the hippocampus in HFD-fed ApoE-/- mice were strongly associated with cognitive impairment, which could be improved by diet control and/or swimming exercise. ## 3.5. Diet Control and/or Swimming Exercise Inhibited HFD-Induced Neuroinflammation in the ApoE-/- Mouse Hippocampus through the SIRT1-Mediated Pathway To investigate the impact of diet control and/or swimming exercise on HFD-induced neuroinflammation, we first tested the expression of neuroinflammation cytokines (NF-κB p65, IL-1β, and TNF-α) in the ApoE-/- mouse hippocampus. Western blot analysis showed NF-κB p65, IL-1β, and TNF-α overexpression in the hippocampus of HFD-induced ApoE-/- mice compared to the CON group ($p \leq 0.05$), whereas the expression of NF-κB p65, IL-1β, and TNF-α was significantly inhibited in the DC, SE, and DS groups ($p \leq 0.05$) (Figures 5(c), 5(e), and 5(f)). To further confirm the protective role of the SIRT1-mediated pathway on neuroinflammation, we explored SIRT1, PGC-1α, and BDNF expression. Western blot analysis showed that SIRT1, PGC-1α, and BDNF expression was significantly downregulated in the HFD group compared to the CON group ($p \leq 0.05$), and the expression of SIRT1, PGC-1α, and BDNF was increased in the DC, SE, and DS groups ($p \leq 0.01$, $p \leq 0.05$) (Figures 5(a), 5(b), and 5(d)). Therefore, in view of the data above, we speculated that SIRT1 activated by diet control and/or swimming exercise might exert a crucial role in anti-neuroinflammation related to downregulation of NF-κB p65 and upregulation of PGC-1α/BDNF expression. ## 4. Discussion Owing to changes in dietetic habits, a high-fat diet is recognized as the biggest health threat worldwide. These unhealthy lifestyles are characterized by excess calorie intake and low calorie expenditure, which contribute to an increased risk for chronic diseases, such as obesity, type 2 diabetes, atherosclerosis, ischaemic stroke, and cardiovascular disease [25, 26]. In addition, both the HFD-induced neuroinflammatory response and physical inactivity are closely related to poor cognitive performance [27, 28]. Compelling evidence has demonstrated that HFD-induced obesity results in brain inflammation and potentially leads to memory loss [29], and similar findings were observed in the present study. From our point of view, diet control and aerobic exercise are considered the most promising therapeutic options to improve memory deficits. The hippocampus is located between the thalamus and medial temporal lobe in the brain and is primarily responsible for cognitive functions [30]. The hippocampus is vulnerable to the HFD-induced inflammatory response, and a HFD impairs the memory-consolidation process [31]. TNF-α and IL-1β, as well-known proinflammatory cytokines, were observed to be distinctly increased in the hippocampus of mice fed a HFD for 8 weeks [32]. Increasing evidence indicates that consumption of a HFD for more than one week is sufficient to generate learning and memory deficits and hippocampal plasticity impairment [33]. Furthermore, energy expenditure far below caloric intake under a sedentary lifestyle is the major contributor to obesity, which aggravates HFD-induced hippocampus-dependent memory loss [11]. TNF-α and IL-1β are confirmed to be the key regulators of synaptic plasticity and memory loss [34, 35]. As expected, higher levels of TNF-α and IL-1β were observed in the hippocampus of ApoE-/- mice in the HFD group, further suggesting that exposure to a HFD accelerated neuroinflammation in the hippocampus and resulted in cognitive dysfunction. Diet control and/or swimming exercise suppressed the neuroinflammatory response in the hippocampus and improved cognitive deficits. Neuroinflammation affects the structure and function of the hippocampus. Microstructural changes in the hippocampus are one of the characteristic biological markers of cognitive deficits and impaired performance in learning memory [36]. DTI is a novel MRI-based imaging technique to detect early subtle alterations in white matter structure that are closely associated with the pathophysiology of cognitive impairment [37]. DTI can measure the direction and extent of the diffusion of water molecules in a three-dimensional space, reflecting the diffusion anisotropic characteristics of living tissues in white matter. Currently, DTI has been widely used to evaluate the white matter microstructure in individuals [38]. FA, a measure of the degree of directionality of water diffusion, is the main indicator of DTI [39]. Lower levels of FA in white matter generally imply reduced microstructural white matter integrity [40]. Decreased FA in the hippocampus corresponded to worse memory performance in older healthy individuals [41]. Our data revealed a remarkable decrease in FA in the bilateral hippocampus of HFD-fed ApoE-/- mice, indicating that the HFD promoted microstructural white matter integrity alterations in the hippocampus. However, microstructural alterations in the hippocampus and cognitive impairment were improved in mice with diet control and/or swimming exercise. This finding was consistent with accumulating evidence that diet control and/or aerobic exercise ameliorated HFD-induced cognitive decline by improving microstructural changes in the hippocampus. MRS is a noninvasive imaging technique to detect and quantify metabolic changes in vivo. The main neurometabolites detected in MRS are NAA, Cho, MI, and Cr [42]. Aberrant metabolism was one of the features displayed before the occurrence of microstructural alterations in the brain [43]. It was reported that positive correlations of NAA/Cr and Cho/Cr with cognitive impairment and decreases in NAA/Cr and Cho/Cr were found in patients with MCI subsequently converting to AD [44]. It was demonstrated that metabolite abnormalities in the pathologic progression of AD were characterized by an increase in the MI/Cr ratio [45]. Therefore, progression in cognitive decline is accompanied by decreases in NAA/Cr and Cho/Cr and increases in MI/Cr. However, NAA is found almost exclusively in neurons and is recognized as a pivotal indicator of neuronal loss and dysfunction. A decreased concentration of NAA or a reduction in the NAA/Cr ratio reflects damage to neuronal microstructure and reduced neural metabolism, which are regarded as promoters of the occurrence of cognitive decline [46]. NAA is the marker of integrity maintenance in neurons, whereas MI is closely associated with CNS inflammation [47]. Our study revealed that lower levels of NAA/Cr and Cho/Cr and higher levels of MI/Cr were observed in the bilateral hippocampus of HFD-fed mice, and diet control and/or swimming exercise reversed this tendency, indicating that the HFD could impair hippocampal-dependent cognition. Diet control and/or swimming exercise improved cognition by affecting neurometabolites in the hippocampus. SIRT1 has been confirmed to be involved in the regulation of neuroinflammation [48]. As an NAD+-dependent deacetylase, SIRT1 can deacetylate lysine residues on various substrates to regulate anti-inflammatory and neuroprotective effects, such as NF-κB and PGC-1α [16]. NF-κB, as a sensitive transcription factor in the inflammatory pathway, plays a pivotal role in neuroinflammatory responses. Activation of NF-κB can regulate the release of TNF-α and IL-1β in the hippocampus, which is responsible for inflammatory diseases of the central nervous system [49]. NF-κB is a family of transcriptional complexes composed of five constituent proteins: RelA (p65), RelB, c-Rel, NF-κB1 (p50), and NF-κB2 (p52) [50]. In addition, the members of NF-κB have a C-terminal transactivation domain (TAD), including RelA (p65), RelB, and c-Rel. The predominant form of NF-κB is a heterodimer of RelA (p65) and NF-κB1 (p50) in the nucleus. SIRT1 can deacetylate lysine residue 310 on RelA (p65), which plays a dominant role in regulating cellular processes by binding to target genes [51]. SIRT1-mediated acetylation of RelA (p65) inhibits the transcriptional activity of NF-κB, as well as the expression of TNF-α and IL-1β [21]. It has been confirmed that PGC-1α is a downstream target of SIRT1 [52]. PGC-1α, first described in the oxidative stress system, binds to numerous transcription factors involved in the regulation of oxidative metabolism and mitochondrial biogenesis [53]. There is compelling evidence that PGC-1α plays an essential role in the neuroprotective effect [54] and that swimming exercise can upregulate the expression of PGC-1α [55]. Overexpression of PGC-1α induced hippocampal BDNF expression, leading to improved cognitive impairment [56]. Exercise-induced PGC-1α was accompanied by an increase in BDNF release [57], indicating that swimming exercise could ameliorate cognitive decline through SIRT1/PGC-1α/BDNF. In this study, HFD increased the expression of NF-κB p65, TNF-α, and IL-1β and inhibited SIRT1, PGC-1α, and BDNF expression in the hippocampus of ApoE-/- mice. However, diet control and/or swimming exercise increased the expression of SIRT1, PGC-1α, and BDNF and suppressed NF-κB p65, TNF-α, and IL-1β expression in the hippocampus. Our findings indicated that diet control and/or swimming exercise improved HFD-induced cognitive impairment through antineuroinflammation, which was associated with activating the SIRT1-NF-κB and SIRT1-PGC-1α-BDNF pathways. ## 5. Conclusions In summary, our results further confirmed that HFD and low physical activity are regarded as threats to health and contribute to cognitive decline through neuroinflammation and microstructural changes in the hippocampus. Dietary control and physical activity ameliorate HFD-induced cognitive impairment and abnormal neurometabolism, which is associated with SIRT1-mediated NF-κB pathways and PGC-1α-BDNF pathways. ## Data Availability The experiment data are available from the corresponding author upon reasonable request. ## Conflicts of Interest No potential conflict of interest is declared by the authors. ## Authors' Contributions Wei Wei and Zhicheng Lin contributed equally to this work. ## References 1. Athilingam P., Visovsky C., Elliott A. F., Rogal P. J.. Cognitive screening in persons with chronic diseases in primary care. (2015) 30 547-558. DOI: 10.1177/1533317515577127 2. Bhatti G. K., Reddy A. P., Reddy P. H., Bhatti J. S.. Lifestyle modifications and nutritional interventions in aging-associated cognitive decline and Alzheimer's disease. (2020) 11. DOI: 10.3389/fnagi.2019.00369 3. Boitard C., Maroun M., Tantot F.. Juvenile obesity enhances emotional memory and amygdala plasticity through glucocorticoids. (2015) 35 4092-4103. DOI: 10.1523/JNEUROSCI.3122-14.2015 4. Cai H., Li G., Hua S., Liu Y., Chen L.. Effect of exercise on cognitive function in chronic disease patients: a meta-analysis and systematic review of randomized controlled trials. (2017) Volume 12 773-783. DOI: 10.2147/CIA.S135700 5. Rojer A. G., Ramsey K. A., Gomes E. S.. Objectively assessed physical activity and sedentary behavior and global cognitive function in older adults: a systematic review. (2021) 198, article 111524. DOI: 10.1016/j.mad.2021.111524 6. Molteni R., Barnard R. J., Ying Z., Roberts C. K., Gómez-Pinilla F.. A high-fat, refined sugar diet reduces hippocampal brain-derived neurotrophic factor, neuronal plasticity, and learning. (2002) 112 803-814. DOI: 10.1016/S0306-4522(02)00123-9 7. Lindqvist A., Mohapel P., Bouter B.. High-fat diet impairs hippocampal neurogenesis in male rats. (2006) 13 1385-1388. DOI: 10.1111/j.1468-1331.2006.01500.x 8. Orchard T. S., Gaudier-Diaz M. M., Weinhold K. R., Courtney DeVries A.. Clearing the fog: a review of the effects of dietary omega-3 fatty acids and added sugars on chemotherapy-induced cognitive deficits. (2017) 161 391-398. DOI: 10.1007/s10549-016-4073-8 9. Alzoubi K. H., Mayyas F. A., Mahafzah R., Khabour O. F.. Melatonin prevents memory impairment induced by high-fat diet: role of oxidative stress. (2018) 336 93-98. DOI: 10.1016/j.bbr.2017.08.047 10. Cordner Z. A., Tamashiro K. L.. Effects of high-fat diet exposure on learning & memory. (2015) 152 363-371. DOI: 10.1016/j.physbeh.2015.06.008 11. Lee T. H., Yau S. Y.. From obesity to hippocampal neurodegeneration: pathogenesis and non-pharmacological interventions. (2020) 22. DOI: 10.3390/ijms22010201 12. Neufer P. D., Bamman M. M., Muoio D. M.. Understanding the cellular and molecular mechanisms of physical activity- induced health benefits. (2015) 22 4-11. DOI: 10.1016/j.cmet.2015.05.011 13. Kirk-Sanchez N. J., McGough E. L.. Physical exercise and cognitive performance in the elderly: current perspectives. (2014) 9 51-62. DOI: 10.2147/CIA.S39506 14. Li J., Liu Y., Liu B.. Mechanisms of aerobic exercise upregulating the expression of hippocampal synaptic plasticity-associated proteins in diabetic rats. (2019) 2019. DOI: 10.1155/2019/7920540 15. Xu J., Jackson C. W., Khoury N., Escobar I., Perez-Pinzon M. A.. Brain SIRT1 mediates metabolic homeostasis and neuroprotection. (2018) 9. DOI: 10.3389/fendo.2018.00702 16. Jiao F., Gong Z.. The beneficial roles of SIRT1 in neuroinflammation-related diseases. (2020) 2020. DOI: 10.1155/2020/6782872 17. Cheng C. F., Ku H. C., Lin H.. PGC-1. (2018) 19 p. 3447. DOI: 10.3390/ijms19113447 18. Khan S. A., Sathyanarayan A., Mashek M. T., Ong K. T., Wollaston-Hayden E. E., Mashek D. G.. ATGL-catalyzed lipolysis regulates SIRT1 to control PGC-1. (2015) 64 418-426. DOI: 10.2337/db14-0325 19. Xia D. Y., Huang X., Bi C. F., Mao L. L., Peng L. J., Qian H. R.. PGC-1. (2017) 9. DOI: 10.3389/fnagi.2017.00065 20. Torma F., Bori Z., Koltai E.. Eating habits modulate short term memory and epigenetical regulation of brain derived neurotrophic factor in hippocampus of low- and high running capacity rats. (2014) 107 54-60. DOI: 10.1016/j.brainresbull.2014.07.003 21. Yeung F., Hoberg J. E., Ramsey C. S.. Modulation of NF-. (2004) 23 2369-2380. DOI: 10.1038/sj.emboj.7600244 22. Yang H., Zhang W., Pan H.. SIRT1 activators suppress inflammatory responses through promotion of p65 deacetylation and inhibition of NF-. (2012) 7. DOI: 10.1371/journal.pone.0046364 23. Wang X., Yang J., Lu T.. The effect of swimming exercise and diet on the hypothalamic inflammation of ApoE-/- mice based on SIRT1-NF-. (2020) 12 11085-11099. DOI: 10.18632/aging.103323 24. Pellegrin M., Berthelot A., Houdayer C., Gaume V., Deckert V., Laurant P.. New insights into the vascular mechanisms underlying the beneficial effect of swimming training on the endothelial vasodilator function in apolipoprotein E-deficient mice. (2007) 190 35-42. DOI: 10.1016/j.atherosclerosis.2006.02.001 25. Bray G. A., Heisel W. E., Afshin A.. The science of obesity management: an endocrine society scientific statement. (2018) 39 79-132. DOI: 10.1210/er.2017-00253 26. Dempsey P. C., Matthews C. E., Dashti S. G.. Sedentary behavior and chronic disease: mechanisms and future directions. (2020) 17 52-61. DOI: 10.1123/jpah.2019-0377 27. Pistell P. J., Morrison C. D., Gupta S.. Cognitive impairment following high fat diet consumption is associated with brain inflammation. (2010) 219 25-32. DOI: 10.1016/j.jneuroim.2009.11.010 28. Miyawaki C. E., Bouldin E. D., Kumar G. S., McGuire L. C.. Associations between physical activity and cognitive functioning among middle-aged and older adults. (2017) 21 637-647. DOI: 10.1007/s12603-016-0835-6 29. Tan B. L., Norhaizan M. E.. Effect of high-fat diets on oxidative stress, cellular inflammatory response and cognitive function. (2019) 11 p. 2579. DOI: 10.3390/nu11112579 30. Preston A. R., Eichenbaum H.. Interplay of hippocampus and prefrontal cortex in memory. (2013) 23 R764-R773. DOI: 10.1016/j.cub.2013.05.041 31. Spencer S. J., D'Angelo H., Soch A., Watkins L. R., Maier S. F., Barrientos R. M.. High-fat diet and aging interact to produce neuroinflammation and impair hippocampal- and amygdalar-dependent memory. (2017) 58 88-101. DOI: 10.1016/j.neurobiolaging.2017.06.014 32. Masetto Antunes M., Godoy G., Masi L. N., Curi R., Barbosa Bazotte R.. Prefrontal cortex and hippocampus inflammation in mice fed high-carbohydrate or high-fat diets. (2022) 25 110-113. DOI: 10.1089/jmf.2021.0026 33. Wang Z., Ge Q., Wu Y., Zhang J., Gu Q., Han J.. Impairment of long-term memory by a short-term high-fat diet via hippocampal oxidative stress and alterations in synaptic plasticity. (2020) 424 24-33. DOI: 10.1016/j.neuroscience.2019.10.050 34. Sugimura T., Yoshimura Y., Komatsu Y.. TNF. (2015) 96 37-44. DOI: 10.1016/j.neures.2015.02.005 35. Prieto G. A., Snigdha S., Baglietto-Vargas D.. Synapse-specific IL-1 receptor subunit reconfiguration augments vulnerability to IL-1. (2015) 112 E5078-E5087. DOI: 10.1073/pnas.1514486112 36. Nazem A., Sankowski R., Bacher M., al-Abed Y.. Rodent models of neuroinflammation for Alzheimer's disease. (2015) 12 p. 74. DOI: 10.1186/s12974-015-0291-y 37. Mayo C. D., Garcia-Barrera M. A., Mazerolle E. L.. Relationship between DTI metrics and cognitive function in Alzheimer's disease. (2019) 10. DOI: 10.3389/fnagi.2018.00436 38. Clemm von Hohenberg C., Pasternak O., Kubicki M.. White matter microstructure in individuals at clinical high risk of psychosis: a whole-brain diffusion tensor imaging study. (2014) 40 895-903. DOI: 10.1093/schbul/sbt079 39. Amlien I. K., Fjell A. M.. Diffusion tensor imaging of white matter degeneration in Alzheimer's disease and mild cognitive impairment. (2014) 276 206-215. DOI: 10.1016/j.neuroscience.2014.02.017 40. Kochunov P., Thompson P. M., Lancaster J. L.. Relationship between white matter fractional anisotropy and other indices of cerebral health in normal aging: tract-based spatial statistics study of aging. (2007) 35 478-487. DOI: 10.1016/j.neuroimage.2006.12.021 41. Mielke M. M., Kozauer N. A., Chan K. C.. Regionally-specific diffusion tensor imaging in mild cognitive impairment and Alzheimer's disease. (2009) 46 47-55. DOI: 10.1016/j.neuroimage.2009.01.054 42. Zhang N., Song X., Bartha R.. Advances in high-field magnetic resonance spectroscopy in Alzheimer’s disease. (2014) 11 367-388. DOI: 10.2174/1567205011666140302200312 43. Chen S. Q., Wang P. J., Ten G. J., Zhan W., Li M. H., Zang F. C.. Role of myo-inositol by magnetic resonance spectroscopy in early diagnosis of Alzheimer's disease in APP/PS1 transgenic mice. (2009) 28 558-566. DOI: 10.1159/000261646 44. Wang H., Tan L., Wang H. F.. Magnetic resonance spectroscopy in alzheimer’s disease: systematic review and meta-analysis. (2015) 46 1049-1070. DOI: 10.3233/JAD-143225 45. Kantarci K. J., Jack C. R., Xu Y. C.. Regional metabolic patterns in mild cognitive impairment and Alzheimer’s disease: a 1H MRS study. (2000) 55 210-217. DOI: 10.1212/WNL.55.2.210 46. Xia Y., Fu Y., Xu H., Guan J., Yi H., Yin S.. Changes in cerebral metabolites in obstructive sleep apnea: a systemic review and meta-analysis. (2016) 6. DOI: 10.1038/srep28712 47. Bladowska J., Kulej D., Biel A.. The role of MR imaging in the assessment of clinical outcomes in children with X-linked adrenoleukodystrophy after allogeneic haematopoietic stem cell transplantation. (2015) 80 181-190. DOI: 10.12659/PJR.893285 48. Zhang Y., Anoopkumar-Dukie S., Arora D., Davey A. K.. Review of the anti-inflammatory effect of SIRT1 and SIRT2 modulators on neurodegenerative diseases. (2020) 867, article 172847. DOI: 10.1016/j.ejphar.2019.172847 49. Liu Y., Zhang Y., Zheng X.. Galantamine improves cognition, hippocampal inflammation, and synaptic plasticity impairments induced by lipopolysaccharide in mice. (2018) 15 p. 112. DOI: 10.1186/s12974-018-1141-5 50. Park M. H., Hong J. T.. Roles of NF-. (2016) 5 p. 2. DOI: 10.3390/cells5020015 51. Kim M. Y., Koh D. I., Choi W. I.. ZBTB2 increases PDK4 expression by transcriptional repression of RelA/p65. (2015) 43 1609-1625. DOI: 10.1093/nar/gkv026 52. Rodgers J. T., Lerin C., Haas W., Gygi S. P., Spiegelman B. M., Puigserver P.. Nutrient control of glucose homeostasis through a complex of PGC-1alpha and SIRT1. (2005) 434 113-118. DOI: 10.1038/nature03354 53. Vega R. B., Huss J. M., Kelly D. P.. The coactivator PGC-1 cooperates with peroxisome proliferator-activated receptor. (2000) 20 1868-1876. DOI: 10.1128/MCB.20.5.1868-1876.2000 54. Cheng A., Wan R., Yang J. L.. Involvement of PGC-1. (2012) 3 p. 1250. DOI: 10.1038/ncomms2238 55. Gioscia-Ryan R. A., Battson M. L., Cuevas L. M., Zigler M. C., Sindler A. L., Seals D. R.. Voluntary aerobic exercise increases arterial resilience and mitochondrial health with aging in mice. (2016) 8 2897-2914. DOI: 10.18632/aging.101099 56. Han B., Jiang W., Liu H.. Upregulation of neuronal PGC-1. (2020) 10 2832-2848. DOI: 10.7150/thno.37119 57. Ieraci A., Madaio A. I., Mallei A., Lee F. S., Popoli M.. Brain-derived neurotrophic factor Val66Met human polymorphism impairs the beneficial exercise-induced neurobiological changes in mice. (2016) 41 3070-3079. DOI: 10.1038/npp.2016.120
--- title: 'The roles of the dietitian in an 18-week telephone and mobile application nutrition intervention for upper gastrointestinal cancer: a qualitative analysis' authors: - Sharni Testa - Kate Furness - Tammie Choi - Terry Haines - Catherine E. Huggins journal: Supportive Care in Cancer year: 2023 pmcid: PMC10049904 doi: 10.1007/s00520-023-07684-9 license: CC BY 4.0 --- # The roles of the dietitian in an 18-week telephone and mobile application nutrition intervention for upper gastrointestinal cancer: a qualitative analysis ## Abstract ### Purpose This study aimed to explore the patient-dietitian experience during an 18-week nutrition counselling intervention delivered using the telephone and a mobile application to people newly diagnosed with upper gastrointestinal (UGI) cancer to [1] elucidate the roles of the dietitian during intervention delivery and [2] explore unmet needs impacting nutritional intake. ### Methods Qualitative case study methodology was followed, whereby the case was the 18-week nutrition counselling intervention. Dietary counselling conversations and post-intervention interviews were inductively coded from six case participants which included fifty-one telephone conversations (17 h), 244 written messages, and four interviews. Data were coded inductively, and themes constructed. The coding framework was subsequently applied to all post-study interviews ($$n = 20$$) to explore unmet needs. ### Results Themes describing the roles of the dietitian were as follows: regular collaborative problem-solving to encourage empowerment, a reassuring care navigator including anticipatory guidance, and rapport building via psychosocial support. Psychosocial support included provision of empathy, reliable care provision, and delivery of positive perspective. Despite intensive counselling from the dietitian, nutrition impact symptom management was a core unmet need as it required intervention beyond the scope of practice for the dietitian. ### Conclusion Delivery of nutrition care via the telephone or an asynchronous mobile application to people with newly diagnosed UGI cancer required the dietitian to adopt a range of roles to influence nutritional intake: they empower people, act as care navigators, and provide psychosocial support. Limitations in dietitians’ scope of practice identified unmet patient’s needs in nutrition impact symptom management, which requires medication management. ### Trial registration 27th January 2017 Australian and New Zealand Clinical Trial Registry (ACTRN12617000152325). ## Introduction Patients with upper gastrointestinal (UGI) cancer (oesophagus, gastric, and pancreas) are vulnerable to malnutrition, with 48–$90\%$ diagnosed with malnutrition [1–3]. Symptoms of the cancer and its treatment are barriers to usual eating patterns that contribute to unintentional weight loss [4]. These symptoms include nausea, fatigue, dysphagia, oesophageal obstructions, early satiety, anxiety, depression, and anorexia [4]. Malnutrition impacts adversely on patients, with increased hospitalisation duration, reduced treatment efficacy, complications, and poorer survival [2, 5, 6]. Despite the importance of nutrition, these patients do not usually see a dietitian unless they are referred via the result of malnutrition screening, or directly from a nurse or doctor [3]. Evidence suggests, however, that $45\%$ of dietetic referrals for patients with cancer should be provided earlier [7]. Limited dietetic resourcing and funding is a barrier to providing patients with cancer regular dietetic support, and therefore resource effective solutions are required [2, 8]. The Telephone or Electronic Nutrition care Delivery (TEND) study was a randomised controlled trial (RCT) that sought to overcome these barriers by testing non-traditional delivery modes of nutrition care including the telephone and a mobile application, to enable intensive, frequent dietetic intervention close to the time of diagnosis [9]. The dietetic intervention was provided external to the participant’s usual care team, and therefore the research dietitian was not part of a multidisciplinary team. Identifying what people may need from a dietitian is important for designing best practice care [10]. Qualitative data describing people’s experience of care can enhance understanding of how an intervention sits within its context [10]. People with cancer are situated in a life-changing experience where diverse support needs are indicated [11]. The TEND study commenced before the rapid adoption of telehealth during the COVID-19 pandemic [12] and is a useful setting in which to explore the roles of a dietitian in delivering care via these delivery modes. Traditionally, dietitian-patient communications have been face-to-face; however, telephone and mHealth delivery remove traditional cues such as body language which may affect rapport and engagement between the dietitian and patient. Evidence is lacking, however, to explore the implications of these novel modes of delivery on the roles of the dietitian. Addressing this gap in evidence is important for delivering effective nutrition care to those undergoing treatment for cancer. The aim of the present study is to explore the patient-dietitian experience of an 18-week nutrition intervention (the TEND study) delivered using the telephone and a mobile application to people newly diagnosed with UGI cancer to elucidate the roles of the dietitian in this context. ## Materials and methods This analysis is set within the TEND study, of which the RCT protocol has been outlined previously [13]. In brief, The TEND study was a three-arm RCT exploring the impact of delivering an 18-week intensive nutrition intervention to patients newly diagnosed with UGI cancer (oesophageal, gastric, and pancreatic cancer). The intervention was delivered by an experienced oncology dietitian, who provided personalised nutrition counselling and goal setting, with behaviour change strategies as informed by the Behaviour Change Technique Taxonomy (v1) [13–15]. A detailed representation of the dietetic intervention has been published [14]. Participants were allocated to receive the intervention using either the telephone or a mobile application, myPace. The initial nutrition assessment was conducted on the telephone. For proceeding weeks, the patient and dietitian communicated using the allocated mode of delivery. At least fortnightly communication was expected for intervention fidelity. A sub-sample of participants completed a post-study telephone interview (semi-structured with a pilot-tested question guide). The post-study interview questions have been described previously [16]. Briefly, the interviews explored patient acceptability of the novel modes of intervention delivery. ## Methodology To examine the roles of the dietitian within the boundaries of the intervention, this study adopted Yin’s qualitative case study approach, from a post-positivist standpoint [17]. Although an uncommon choice, qualitative research methods may be used to align with a post-positivist viewpoint through researcher reflexivity and consideration of various explanations of the data [18, 19]. Case study analysis is an in-depth exploration of a phenomenon within a real-world situation [17, 20]. The ‘case’ for this study is the intervention, an approach that has been used previously [21–23]. This methodology was selected to allow an in-depth analysis of the patient-dietitian experience of the intervention whereby its boundaries included the dietitian, the patient, the frequency and duration of communication, and the tool for communication (i.e. a telephone or a mobile application). In this study, it was theorised that the dietitian’s key roles would be to provide the patient with psychosocial support, nutrition impact symptom management, nutrition optimisation, and pharmacological support. The validity of this theory was explored through analysing in-depth the patient-dietitian experiences within the intervention, as well as the post-study perspectives of patients, to determine the roles of the dietitian required during intervention delivery. The intervention delivery is explored in a real-world, uncontrolled context. ## Reflexivity The research team included a student dietitian (S.T.), the study dietitian (K.F.), an experienced dietitian and qualitative researcher (T.C.), and a nutrition scientist with a scientific inquiry background (C.E.H.). These unique perspectives challenged presumptions of the roles of the dietitian in the described patient experience. As an aspiring dietitian, the student researcher interpreted the data more positively; however, this was managed through conversations with other researchers. ## Case selection In total, six cases (i.e. experiences) of the intervention were selected using maximum-variation sampling [24]. Four of the six cases analysed were selected by the study dietitian, as these were participants who engaged in regular reviews with the dietitian, which is potentially a source of selection bias; therefore, an additional two were independently selected by S.T. to maximise variation of the intervention experiences, including mode of telehealth delivery, cancer type, and gender. ## Data collection Within-intervention phone calls, messages, emails, and post-intervention interviews were audio-recorded/stored. The conversations within the 18 weeks of the six intervention experiences (both written and audio-recorded) were the primary sources of data for the case study. Post-intervention interviews (transcribed) where available were used as a data source in the case analysis. All post-intervention interviews ($$n = 20$$) were available to further examine the theme related to unmet needs. ## Data analysis Data files from the six cases were uploaded to NVivo release 1.4.1 (QSR International, Melbourne, Australia) where the researcher (S.T.) inductively coded within-intervention phone calls, myPace messages, emails, and post-intervention interviews for each intervention experience sequentially. Dialogues from both the dietitian and patient were coded. A subset of transcripts were presented to another author (T.C.) for duplicate coding. Yin’s general strategy ‘Relying on theoretical propositions’ was used [17]. These propositions were used to help guide the coding process and were based on the theory of patient empowerment, the Supportive Care Framework, and anticipatory guidance [11, 25–27]. Audio files were coded directly on NVivo. An annotation was made each time a section of data was coded to provide context to the section coded. After coding was completed for an intervention experience, Yin’s analytic technique ‘Explanation building’ was used [17]. This involved reflecting on the case of the intervention and looking back to the initial theoretical propositions to see if changes were indicated. After this, Yin’s general analytic strategy ‘Examining rival explanations’ provided a step to reflect on alternative views of the data [17]. Codes were examined to create groups and sub-groups leading to the identification of key themes. S.T. consulted with C.E.H. to discuss code interpretation and theme identification. During this analytic process, a coding framework was developed. The coding framework was subsequently used to guide analysis of the post-intervention interviews ($$n = 20$$) to gain a deeper understanding of the unmet needs of participants relating to symptom management. This information was extracted and included in the paper by the study dietitian (K.F.). ## Results Characteristics of the patients receiving each intervention experience are described in Table 1. In total, 51 telephone conversations (17 h), 244 written messages, and four interviews were analysed from six participants. Table 1Characteristics of six patients who received an 18-week nutrition intervention for upper gastrointestinal cancer. These patients’ experiences of the intervention were analysed as part of the present case studyIntervention delivery modeParticipant IDCancer typeGenderAge (years)Completed weeksaPost-study interview (yes/no)Mobile application1OesophagealMale5713Noc2bGastricFemale709Yes3GastricFemale6114YesTelephone4PancreaticMale8815Noc5OesophagealMale5618Yes6PancreaticFemale5515YesaA completed week was defined as any week in the intervention where both the patient and dietitian communicated with each other. Weeks where the dietitian and patient planned no intervention are not includedbIntervention completed using myPace (7 weeks) and email (4 weeks), whereby both email and myPace were used to communicate for 2 weekscPost-study interview not completed as interviews were delivered to a subset of the RCT participants, and all interviews had been completed before this participant completed their intervention ## Key roles of the dietitian The themes describing key roles of the dietitian were [1] regular collaborative problem-solving to encourage empowerment, [2] reassuring care navigation (including anticipatory guidance), [3] rapport building via reliable psychosocial support, and [4] role limitations lead to unmet nutrition needs. These themes are detailed below and can be seen in Fig. 1.Fig. 1The key roles of the dietitian during an 18-week nutrition intervention delivered using the telephone and a mobile application, close to time of diagnosis for cancer of the upper gastrointestinal tract. The blue boxes represent the key roles of the dietitian, i.e. three key themes of this study. The green boxes represent any sub-themes of the key roles portrayed by the dietitian in the intervention. The grey boxes show the limitation of the role of the dietitian in the TEND study, the fourth key theme, and highlight that patient advocacy is impacted when working external to a multidisciplinary team ## Regular collaborative problem-solving to encourage empowerment Patients experienced nutrition impact symptoms such as nausea, taste changes, obstructions, fatigue, and loss of appetite that impacted on eating. Symptoms experienced varied from patient-to-patient, and week-to-week, depending on factors such as cancer type, treatment modality, and cancer progression. These symptoms and the consequent loss of weight were a source of anxiety for patients. Patients were uncertain of how to manage eating and drinking with these symptoms and required the dietitian’s support.…I have also lost my appetite and find it difficult to find a meal that I can eat and enjoy without feeling nauseous… I don’t know what to do. I have lost about 3 kilograms in weight since I last wrote. I will see the [usual care] dietitian at the trial centre on Monday and hope she will have an answer for me…. ( Participant 2, myPace, week 10) Dietary advice provided by the dietitian (verbally and in written education material) to achieve goals of weight stability in cancer was sometimes in conflict with dietary advice received prior to cancer diagnosis for other health issues. Having regular contact with the dietitian helped patients prioritise dietary instructions and removed the angst that came with trying to decipher conflicting advice leading to greater confidence to self-manage nutrition needs. On many occasions, the patient was unnecessarily worried about their weight status, eating and drinking strategies, or food safety practices, and reassurance from the dietitian alleviated this concern. The prompt and regular delivery of reassurance by the dietitian was empowering for the patient. The dietitian also played an active role in encouraging self-confidence. This included illustrating to the patient when they were managing an issue appropriately and recognising patient achievements. Other forms of empowerment included encouraging social connection and patient participation in self-care. At study cessation, there was a sense that patients felt empowered in their own self-management capabilities. I’ve learnt so much from interacting with [the dietitian] … and I think that’s the best part of it, that a patient has a chance to interact with a professional like that, and be able to help themselves. ( Participant 2, myPace and email, post-study interview) ## A reassuring care navigator Cancer symptoms were worrying to the patient and the dietitian gave reassurance and prompted them to contact their medical practitioner to improve symptom management. This was a priority for the dietitian as poor symptom control impacts adversely on oral intake. Patients had concerns related to the use of some medications treating nutrition impact symptoms and the dietitian promoted understanding of these medications based on prescribing recommendations. When medication had been prescribed to a patient, the dietitian was able to support patients to apply the education information they received in hospital. For example, regular discussion with the dietitian was required to better understand how to dose pancreatic enzyme replacement therapy (PERT) with changes to oral intake.[The dietitian] kind of set me – Set me right on how I should be taking it [PERT]… Um, things that I was unaware of, that I thought I was taking them properly and, you know, I wasn’t. ( Participant 6, telephone, post-study interview) Patients also discussed feedback received from other health professionals with the dietitian, indicating trust in her opinion.… I’ve just had a call from [anonymous] in the Chemo Day Unit & apparently my pre-chemo blood test shows my current blood sugar levels are at 7.1. Normal should be between 4 & 6, so I have been referred to the Diabetes Unit & they will be in touch with me to arrange an appointment… Is this a normal process for people with a slightly high reading or…………….? ( Participant 1, mobile application, week 16) The dietitian was able to provide interpretation of this feedback and reassure them when they received positive news; i.e., the patient was unable to interpret the information in this way themselves. Additionally, the dietitian validated the patient of their right to play an active role in their treatment, i.e. to be informed, to make choices that aligned with their values and needs, and to speak up to healthcare providers where required, promoting patient-centred care. ## Anticipatory guidance to minimise the consequences of nutrition impact symptoms Because the intervention allowed for regular consultations, the dietitian was able to provide patients with information about what to expect along their cancer treatment. This included preparation for post-operative interventions, e.g. enteral feeding and fasting-related weight loss. Preoperative eating recommendations, for example carbohydrate loading, were also provided. On the other hand, when patients began stabilising their weight and hoping to gain, the dietitian realigned their expectations, with weight maintenance being the priority. From the patient perspective, this aspect of preparation was acknowledged. …I think I got looked after; I really do. Preparation-wise. Yep, umm, and my weight sort of went up before surgery, so that was good. As soon as the chemo stopped, I started to put some weight on. ( Participant 3, mobile application, post-study interview) Preparatory advice relating to nutrition impact symptoms and their management was also provided by the dietitian. This included preparing patients for symptoms they may experience with chemotherapy/radiotherapy or after surgery and explaining the likely manifestation of different symptoms with treatment progression. The dietitian also prepared patients for any notable side effects they may experience with a new medication. ## Rapport building via reliable psychosocial support Evidence of rapport between the dietitian and the patient was apparent in both modes of the intervention. This was shown through moments of shared humour, and when a patient disclosed something personal or offered well wishes to the dietitian. Rapport building was enabled by the positive perspective, empathy, and reliability of the dietitian. Patients felt comfortable disclosing their personal feelings and events to the dietitian, such as admitting when they were experiencing low mood or were finding something challenging or frustrating. Soon after their diagnosis, one patient described experiencing ‘the couldn’t be bothereds’, i.e. not feeling energised to do anything, and the dietitian positioned the patient to feel comfortable expressing personal feelings through empathising with their experiences and emotions. The dietitian promoted positive affect. In the context of negative experiences that patients faced with eating and drinking, a role of the dietitian was to encourage patients to find enjoyment in eating. When patients disclosed exciting upcoming personal events, achievements, or other good news, the dietitian expressed excitement. Social support that patients received from family or friends was also highlighted and celebrated by the dietitian. From the patient perspective, the dietitian was valued as someone to talk to who was a professional and not family or friends. Enjoyment was found in these conversations and gratitude was expressed to the dietitian for ‘Being there’.…And thank you very much for what you’re doing …it’s good encouragement and that… Yes, it helps me think I’m getting somewhere. ( Participant 4, telephone, week 8) For the mobile application participants, comfort was found knowing that messages would be responded to, suggesting that the dietitian was valued as a reliable source of support.…It was so good, because, you know, she didn’t really know me - I mean, you just develop a friendship, there’s no doubt about it, the kindness in her voice and that, but um, you know, once it’s finished, it’s finished. And um, yeah, so you can open up. ( Participant 3, mobile application, post-study interview) ## Role limitations lead to unmet nutrition needs The case analysis revealed a myriad of unmet needs described by participants throughout the 18-week intervention period. To explore this further, an analysis of the post-intervention interviews ($$n = 20$$) was undertaken. Symptom managements, particularly those affecting the patient’s ability to eat and drink adequately, were those of primary concern for the dietitian to address. Many of these debilitating nutrition impact symptoms occurred at different intervals and were of differing intensity throughout the intervention period, and many individuals had multiple different symptoms competing for priority management during this time. Not being able to eat. Physically, not mentally. I mean, I used to force food down my throat to the point where I was, you know, dry retching and I couldn’t eat because I was losing so much weight. But eating, that was the most challenging. ( Participant 7, post-study interview) The centralised nature of the dietetic intervention meant that many of the physical symptoms of anorexia, dry retching, nausea, and diarrhoea were unable to be managed effectively without the collaboration of a multidisciplinary team where the dietitian could advocate for medications to be reviewed. Often, where medication was prescribed to manage these symptoms, its dosing and mechanism for action were not explained or understood by the patients. “I’m on four Creon [PERT]. Is that right?” She [the dietitian] said, “No, no, back to two.” Instantly the bowels stopped feeling bloated and gassy and all that jazz, it was much better for me. ( Participant 8, post-study interview) The dietitian provided recommendations for participants to contact their general practitioner, oncologist, or surgeon to prescribe medications; however, this required patient advocacy which was often delayed. Delays were caused by appointment time delays, a competing schedule of chemotherapy and/or other appointments, and physical access such as transport. ## Discussion This study explored the patient-dietitian experience during an 18-week nutrition intervention delivered using the telephone or a mobile application to people newly diagnosed with UGI cancer to elucidate the roles of the dietitian during the intervention. This study identifies important components of delivering supportive care via different modes for those undergoing treatment for UGI cancer. The roles of the dietitian were characterised by regular collaborative problem-solving to encourage empowerment, a reassuring care navigator (including anticipatory guidance), and rapport building via reliable psychosocial support. This analysis also revealed that role limitations led to unmet needs as the dietitian was constrained by poorly managed cancer symptoms that negatively impacted oral intake and subsequently weight stability. This study demonstrated that rapport can be built within the patient-dietitian relationship without face-to-face communication. It is well established that psychosocial support, for example empathy, compassion, positivity, reassurance, and care, is valued by patients in the dietitian-patient relationship [28, 29], and this study shows it can be achieved over the telephone and via asynchronous modes of communication. In addition to psychosocial support, previous research indicates that patients value having a point of contact in their cancer journey [30]. In the present study, it appeared that the dietitian fulfilled this role of a central point of contact for the patients, making parallels with the role of a patient care navigator. Patient care navigators can take the form of a nurse, social worker, case manager, and laypersons [31]. Broadly, their role is to facilitate patients and their family through their cancer experience to increase access to required resources and reduce anxiety. Their role may include problem-solving, facilitating symptom management, liaising with the multidisciplinary team, and making referrals [31]. Likewise, the dietitian in this study provided anticipatory guidance, problem-solving for symptom management, and guidance when contact with another health professional was indicated. Both health professionals and patients recognise the value of the oncology dietitian’s ability to prepare patients for common challenges that occur throughout the cancer journey [16, 28, 32, 33]. Evidence suggests that patient navigation services benefit patients through addressing worries and reducing incidence of hospital admissions and emergency department attendance [34–36]. This study proposes that the care navigator role does not need to be limited to nursing or social work, and that, with training and education, could be shared by the multidisciplinary team. This would not only provide more human resources for this role, but it would also act to benefit the health professional through broadening their knowledge and connection to the patient experience through continuity of care, ultimately promoting patient-centred care. Furthermore, use of telephones and electronic devices in the delivery of care navigation could be an effective and time-efficient way to deliver care navigation, enabling flexibility and ease of access between the patient and their navigator. The dietitian’s role as a care navigator and multidisciplinary team member was limited by the scope of the RCT as the dietitian could only recommend to the patient when and where to seek help. It was not within the scope of the intervention for the dietitian to actively contact patients’ individual healthcare team to advocate, discuss, and collaborate to optimise symptom and medication management, distress, and treatment progression. This highlights the importance of multidisciplinary teams being critical in cancer management and the need for the dietitian to be part of the team. Improving the safety and quality of health service provision to ensure that it is provided by the right person, at the right time, in the right place, and for the right cost is essential [37, 38]. The dietitian can play an integral role in extended or advanced practice roles to address the high prevalence of malnutrition and high levels of unmet needs particularly relating to nutrition impact symptoms, and act as a care navigator to guide individuals and their families/carers on their cancer journey. The case study design is a strength of this work; many studies have short-term interventions and provide little insight into how the intervention works. The use of within-intervention conversations was also a strength of this study as this data is not limited by memory or response bias. This detailed analysis reveals the roles of the dietitian needed to maximise oral intake during cancer treatment; furthermore, it reveals that sub-optimal cancer symptom management remains a significant barrier to achieving optimal oral intake, even when under the intensive care of a dietitian. A limitation is that this was a post hoc analysis, precluding direct questions of the patient and dietitian perspective of the dietitian’s role. ## Conclusion This study explored the patient-dietitian experience of an 18-week nutrition intervention delivered non-face-to-face using the telephone or a mobile application to patients with newly diagnosed UGI cancer. Regular collaborative problem-solving to encourage empowerment, reassuring care navigation (including provision of anticipatory guidance), and rapport building via reliable psychosocial support were key roles of the dietitian. It is evident that patients with UGI cancer are faced with nutrition impact symptoms that limit the capacity of the dietitian to support patients in isolation of the multidisciplinary team. Further, research is recommended to examine an advanced care role for dietitians in the management of nutrition impact symptoms. ## References 1. 1.Loeliger J, Kiss N (2015) Phase II malnutrition in Victorian Cancer services: summary report. Dept of Health and Human Services, State Government of Victoria, Melbourne 2. 2.Marshall K, Loeliger J (2013) Investigating practices relating to malnutrition in Victorian cancer services: summary report 2012. Dept of Health, State Government of Victoria. https://www2.health.vic.gov.au/about/health-strategies/cancer-care/cancer-projects/investigating-practices-relating-to-malnutrition-in-victorian-cancer-services. Accessed 4 June 2021 3. Silvers MA, Savva J, Huggins CE, Truby H, Haines T. Potential benefits of early nutritional intervention in adults with upper gastrointestinal cancer: a pilot randomised trial. Support Care Cancer (2014.0) 22 3035-3044. DOI: 10.1007/s00520-014-2311-3 4. 4.Cancer Council Victoria and Department of Health Victoria (2021) Optimal care pathway for people with oesophagogastric cancer. https://www.cancer.org.au/assets/pdf/oesophagogastric-cancer-optimal-cancer-care-pathway. Accessed 4 June 2021 5. Andreyev HJ, Norman AR, Oates J, Cunningham D. Why do patients with weight loss have a worse outcome when undergoing chemotherapy for gastrointestinal malignancies?. Eur J Cancer (1998.0) 34 503-509. DOI: 10.1016/S0959-8049(97)10090-9 6. Van Cutsem E, Arends J. The causes and consequences of cancer-associated malnutrition. Eur J Oncol Nurs (2005.0) 9 S51-63. DOI: 10.1016/j.ejon.2005.09.007 7. Lorton CM, Griffin O, Higgins K, Roulston F, Stewart G, Gough N, Barnes E, Aktas A, Walsh TD. Late referral of cancer patients with malnutrition to dietitians: a prospective study of clinical practice. Support Care Cancer (2020.0) 28 2351-2360. DOI: 10.1007/s00520-019-05042-2 8. Findlay M, Rankin NM, Shaw T, White K, Boyer M, Milross C, De Abreu LR, Brown C, Collett G, Beale P, Bauer JD. Best evidence to best practice: implementing an innovative model of nutrition care for patients with head and neck cancer improves outcomes. Nutrients (2020.0) 12 1465. DOI: 10.3390/nu12051465 9. Huggins CE, Hanna L, Furness K, Silvers MA, Savva J, Frawley H, Croagh D, Cashin P, Low L, Bauer J, Truby H, Haines TP. Effect of early and intensive telephone or electronic nutrition counselling delivered to people with upper gastrointestinal cancer on quality of life: a three-arm randomised controlled trial. Nutrients (2022.0) 14 3234. DOI: 10.3390/nu14153234 10. Rand L, Dunn M, Slade I, Upadhyaya S, Sheehan M. Understanding and using patient experiences as evidence in healthcare priority setting. Cost Eff Resour Alloc (2019.0) 17 20. DOI: 10.1186/s12962-019-0188-1 11. Fitch MI. Supportive care framework. Can Oncol Nurs J (2008.0) 18 6-24. DOI: 10.5737/1181912x181614 12. Rozga M, Handu D, Kelley K, Jimenez EY, Martin H, Schofield M, Steiber A. Telehealth during the COVID-19 pandemic: a cross-sectional survey of registered dietitian nutritionists. J Acad Nutr Diet (2021.0) S2212–2672 00036-38. DOI: 10.1016/j.jand.2021.01.009 13. Hanna L, Huggins CE, Furness K, Silvers MA, Savva J, Frawley H, Croagh D, Cashin P, Low L, Bauer J, Truby H, Haines T. Effect of early and intensive nutrition care, delivered via telephone or mobile application, on quality of life in people with upper gastrointestinal cancer: study protocol of a randomised controlled trial. BMC Cancer (2018.0) 18 707. DOI: 10.1186/s12885-018-4595-z 14. Furness K, Huggins CE, Hanna L, Silvers MA, Cashin P, Low L, Croagh D, Haines TP. A process and mechanism of action evaluation of the effect of early and intensive nutrition care, delivered via telephone or mobile application, on quality of life in people with upper gastrointestinal cancer: a study protocol. BMC Cancer (2018.0) 18 1181. DOI: 10.1186/s12885-018-5089-8 15. Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, Eccles MP, Cane J, Wood CE. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med (2013.0) 46 81-95. DOI: 10.1007/s12160-013-9486-6 16. Furness K, Huggins CE, Truby H, Croagh D, Haines TP. Attitudes of Australian patients undergoing treatment for upper gastrointestinal cancers to different models of nutrition care delivery: qualitative investigation. JMIR Form Res (2021.0) 5 e23979. DOI: 10.2196/23979 17. Yin RK. Case study research: design and methods (2009.0) 18. Grant BM, Giddings LS. Making sense of methodologies: a paradigm framework for the novice researcher. Contemp Nurse (2002.0) 13 10-28. DOI: 10.5172/conu.13.1.10 19. Monrouxe LV, Rees CE. When I say … quantification in qualitative research. Med Educ (2020.0) 54 186-187. DOI: 10.1111/medu.14010 20. Crowe S, Cresswell K, Robertson A, Huby G, Avery A, Sheikh A. The case study approach. BMC Med Res Methodol (2011.0) 11 100. DOI: 10.1186/1471-2288-11-100 21. 21.Lesage-Moussavou-Nzamba M, Houle J, Trudeau F (2020) Participants’ perspectives of a primary exercise-based prevention program for cardiac patients: a prepost intervention qualitative case study. Rehabil Res Pract. 10.1155/2020/6215428 22. Rotteau L, Albert M, Bhattacharyya O, Berta W, Webster F. Understanding decisions to scale up: a qualitative case study of three health service intervention evaluations. J Health Serv Res Policy (2021.0) 26 37-45. DOI: 10.1177/1355819620921892 23. Moll SE. The web of silence: a qualitative case study of early intervention and support for healthcare workers with mental ill-health. BMC Public Health (2014.0) 14 138. DOI: 10.1186/1471-2458-14-138 24. Given LM. The SAGE Encyclopedia of Qualitative Research Methods (2008.0) 25. Alpay LL, Henkemans OB, Otten W, Rövekamp TAJM, Dumay ACM. E-health applications and services for patient empowerment: directions for best practices in The Netherlands. Telemed J E Health (2010.0) 16 787-791. DOI: 10.1089/tmj.2009.0156 26. Aujoulat I, d’Hoore W, Deccache A. Patient empowerment in theory and practice: polysemy or cacophony?. Patient Educ Couns (2007.0) 66 13-20. DOI: 10.1016/j.pec.2006.09.008 27. Johnson EL, Frias JP, Trujillo JM. Anticipatory guidance in type 2 diabetes to improve disease management next steps after basal insulin. Postgrad Med (2018.0) 130 365-374. DOI: 10.1080/00325481.2018.1452515 28. McCarter K, Baker AL, Britton B, Halpin SA, Beck A, Carter G, Wratten C, Bauer J, Wolfenden L, Burchell K, Forbes E. Head and neck cancer patient experience of a new dietitian-delivered health behaviour intervention: ‘you know you have to eat to survive’. Support Care Cancer (2018.0) 26 2167-2175. DOI: 10.1007/s00520-017-4029-5 29. Sladdin I, Chaboyer W, Ball L. Patients’ perceptions and experiences of patient-centred care in dietetic consultations. J Hum Nutr Diet (2018.0) 31 188-196. DOI: 10.1111/jhn.12507 30. 30.Easley J, Miedema B, Carroll JC, O’Brien MA, Manca DP, Grunfeld E (2016) Patients’ experiences with continuity of cancer care in Canada: results from the CanIMPACT study. Can Fam Physician 62(10):821–827 31. McMullen L. Oncology nurse navigators and the continuum of cancer care. Semin Oncol Nurs (2013.0) 29 105-117. DOI: 10.1016/j.soncn.2013.02.005 32. Furness K, Huggins C, Croagh D, Haines T. Exploring the attitudes of health professionals providing care to patients undergoing treatment for upper gastrointestinal cancers to different models of nutrition care delivery: a qualitative investigation. Nutrients (2021.0) 13 1020. DOI: 10.3390/nu13031020 33. Hazzard E, Walton K, McMahon A-T, Milosavljevic M, Tapsell LC. Healthcare professionals’ perspectives on the role of dietitians within multidisciplinary head and neck cancer teams: a qualitative multi-site study. Nutr Diet (2021.0) 78 506-515. DOI: 10.1111/1747-0080.12664 34. Kline RM, Rocque GB, Rohan EA, Blackley KA, Cantril CA, Pratt-Chapman ML, Burris HA, Shulman LN. Patient navigation in cancer: the business case to support clinical needs. J Oncol Pract (2019.0) 15 585-590. DOI: 10.1200/JOP.19.00230 35. Rocque GB, Partridge EE, Pisu M, Martin MY, Demark-Wahnefried W, Acemgil A, Kenzik K, Kvale EA, Meneses K, Li X, Li Y, Halilova KI, Jackson BE, Chambless C, Lisovicz N, Fouad M, Taylor RA. The Patient Care Connect Program: transforming health care through lay navigation. J Oncol Pract (2016.0) 12 e633-e642. DOI: 10.1200/JOP.2015.008896 36. Rocque GB, Pisu M, Jackson BE, Kvale EA, Demark-Wahnefried W, Martin MY, Meneses K, Li Y, Taylor RA, Acemgil A, Williams CP, Lisovicz N, Fouad M, Kenzik KM, Partridge EE. Resource use and Medicare costs during lay navigation for geriatric patients with cancer. JAMA Oncol. (2017.0) 3 817-25. DOI: 10.1001/jamaoncol.2016.6307 37. 37.Australian Commission on Safety and Quality in Healthcare, About Us. https://www.safetyandquality.gov.au/about-us. Accessed 15 Aug 2022 38. Beesley VL, Janda M, Goldstein D, Gooden H, Merrett ND, O’Connell DL, Rowlands IJ, Wyld D, Neale RE. A tsunami of unmet needs: pancreatic and ampullary cancer patients’ supportive care needs and use of community and allied health services. Psychooncology (2016.0) 25 150-157. DOI: 10.1002/pon.3887

Subsets and Splits

No community queries yet

The top public SQL queries from the community will appear here once available.