text
string | predicted_class
string | confidence
float16 |
|---|---|---|
We are grateful for Skoufias and his co-workers for their work on the TPPP/p25 project, in fact, the paper in the Scientific Reports1 was a challenge which inspired us to reconsider our data and carry out further experiments in order to suggest a molecular mechanism regarding the MT bundling by TPPP/p25.
|
other
| 99.9 |
Human recombinant TPPP/p25 forms and fragments possessing His-tags were expressed in E. coli BL21 (DE3) cells and isolated on HIS-Select™ Cartridge (Sigma-Aldrich) as described previously515. The purity of proteins was analysed by SDS-PAGE. Protein concentrations of the TPPP/p25 variants were determined on the basis of the absorbance at 280 nm using the extinction coefficients evaluated by ProtParam (http://web.expasy.org/protparam/), 10095 M−1 *cm−1 for FL TPPP/p25 and N-terminal truncated TPPP/p25; 5625 M−1 *cm−1 for C-terminal truncated TPPP/p25 and CORE, respectively.
|
study
| 100.0 |
The BiFC plasmids (pBiFC-VN1-173, pBiFC-VC155-238) were the gift of Prof. Péter Várnai (Semmelweis University, Budapest). To insert the FL TPPP/p25 cDNA in either the pBiFC-VN1-173 or pBiFC-VC155-238 the following primers for polymerase chain reaction (PCR) amplification were used: forward 5-GAAGGAGCTCGAGATATGGCTGACAAGGCTAAGC-3 and reverse 5-CCGTGGATCCCTACTTGCCCCCTTGCACCTTCTGGTCGTAGG-3 and pET21c-TPPP/p25 as a template. The PCR product, following purification and after digestion with XhoI-BamHI restriction enzymes, was inserted to XhoI and BamHI sites in the BiFC vectors. To insert the CORE segment of TPPP/p25 (TPPP/p25 delta 3–43/delta 175–219) cDNA in either the pBiFC-VN1-173 or pBiFC-VC155-238, the following primers for PCR amplification were used: forward 5- GGAATTC TATGGCTGCATCCCCTGAG-3 and reverse 5- ATGGATCCCTAGCCCGTGAACTTGGT-3′ and pET21c-TPPP/p25 as a template. The PCR product, following purification and after digestion with EcoRI-BamHI restriction enzymes, was inserted to EcoRI and BamHI sites in the BiFC vectors. The sequences of all construct were verified by restriction mapping and sequencing.
|
study
| 99.94 |
Sandwich ELISA assays were carried out similarly as described previously8. The plate was coated with 1 μg/mL (50 μL/well) mAb9 in 200 mM Na2CO3 buffer pH 9.6 overnight at 4 °C. The wells were blocked with 1 mg/mL bovine serum albumin in phosphate-buffered saline (PBS) for 1 h at room temperature. Next, the plate was incubated with serial dilutions of 5 μM TPPP/p25 for 1 h at room temperature in PBS. Where indicated, TPPP/p25 was pre-incubated with 100 μM DTE for 30 min at room temperature. In another experimental setup, different TPPP forms at 1 μM concentration were added to the plate coated with mAb in the absence and presence of DTE. Alternatively, 1 μM FL TPPP/p25 or CORE TPPP/p25 was preincubated with 100 μM DTE for 30 min at room temperature, then added to the plate coated with mAb; after removing the excess DTE, the different TPPP/p25 forms were added, and dimers were detected by biotinylated mAb. Then the plate was sequentially incubated with biotinylated mAb (1 μg/mL) and peroxidase conjugated avidin (Calbiochem) (2.5 μg/mL). Both antibodies were in PBS buffer containing 1 mg/mL bovine serum albumin, and incubated for 1 h at room temperature. Between each incubation step the wells were washed thrice with PBS containing 0.05% Tween 20 for 10 min. The presence of antibodies was detected using o-phenylenediamine as substrate. The reaction was stopped after 10 min with 1 M H2SO4, and the absorbance was read at 490 nm with an EnSpire Multimode Reader (Perkin Elmer). The binding constants (Kd) were evaluated from the saturation curves by non-linear curve fitting assuming single binding site hyperbola model using the Origin 8.0 software.
|
study
| 100.0 |
The assembly of tubulin (7 μM) was assessed in polymerization buffer (50 mM 2-(N-morpholino) ethanesulfonic acid buffer pH 6.6 containing 100 mM KCl, 1 mM DTE, 1 mM MgCl2 and 1 mM ethylene glycol tetraacetic acid) at 37 °C. The tubulin polymerization into MTs was induced by the addition of 3 μM of the different TPPP/p25 forms. In another set of experiments, the polymerization of 10 μM tubulin into MTs was induced by the addition of 20 μM paclitaxel then subsequently 3 μM TPPP/p25 proteins were added to the solution. The optical density was monitored at 350 nm by a Cary 100 spectrophotometer (Varian).
|
study
| 100.0 |
Tubulin (5 μM) was incubated with the different TPPP/p25 forms (10 μM) for 10 min at 37 °C in polymerization buffer. In another set of experiments, 10 μM paclitaxel-stabilized MTs was incubated with the different TPPP/p25 forms (10 μM) for 10 min at 37 °C in polymerization buffer. After centrifugation at 17000 g for 15 min at 37 °C, the pellet and the supernatant fractions were separated. The pellet fraction was washed and resuspended in buffer. Then the fractions containing 2-mercaptoethanol were analysed by SDS-PAGE, separated on a 13.5% gel and stained with Coomassie Brilliant Blue R-250.
|
study
| 100.0 |
HeLa cells (ATCC® CCL-2™, American Type Culture Collection) were grown in Dulbecco’s modified eagle medium supplemented with 10% fetal calf serum and 100 μg/ml kanamycin in a humidified incubator at 37 °C with 5% CO2. Cells were grown on 12-mm-diameter glass coverslips for microscopic analysis. For the evaluation of the BiFC signal, HeLa cells were transfected with different mVenus constructs of TPPP/p25 (0.3 μg of each plasmid) using Turbofect (Invitrogen) transfection reagent according to the manufacture’s protocol. Monoclonal tubulin antibody (Sigma T9026) and an Alexa 546 conjugated mouse secondary antibody (Thermo Fisher Scientific A11030) were used for the detection of the tubulin signal. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). Images of the mounted samples were acquired on a Leica DM IL 500 microscope equipped with Leica DFC 395 FX camera and HBO 100w lamp. The equipment software was Leica Application Suite 4.4.0. Chroma UV filter set (No. C40888), Chroma 41028 HQ NB GFP filter set (No. C21116) and Leica filter N2.1 (No. 513832) was used for DAPI, BiFC and Alexa 546 signal acquisition, respectively, using a HCX FL Fluotar 40x/0.75 (dry) objective.
|
study
| 99.94 |
Chromobox homolog 7 (CBX7) is a chromobox family protein which belongs to polycomb repressive complex 1 (PRC1). In cooperation with other polycomb group (PcG) proteins, CBX7 maintain target genes, which participate in stem cell self-renewal and developmental regulation, in a silenced state [1–4]. CBX7 recognizes H3K27me3, which is catalyzed by enhancer of zeste homolog 2 (EZH2). Then, CBX7 recruits other PRC1 proteins to the target chromatin marked by H3K27me3 and represses gene transcription consequently [5–7]. Besides the PRC-dependent mechanism, recent studies found that CBX7 epigenetically regulates gene expression via the interaction with histone deacetylase 2 (HADC2) [8, 9]. However, the precise role of CBX7 in tumorigenesis and tumor progression is still controversial.
|
study
| 97.6 |
Gil and colleagues first identified CBX7 and found that the lifespan of differentiated human cells was extended and the immortalization of mouse fibroblasts was established via increased CBX7 levels which was followed by downregulating expression of the Ink4a/Arf locus [1, 2]. CBX7 also cooperated with c-Myc and initiated lymphomagenesis . However, of note, CBX7 expression has a negative correlation with advanced tumor aggressiveness and poor prognosis in a wide range of tumor types, such as thyroid, colorectal, pancreas, breast, lung, and glioblastoma (GBM) [9–18]. The contradictory behavior of CBX7 in different kinds of tumors reflects the notion that cellular context plays an essential role in the effects of the CBX7.
|
review
| 99.5 |
GBM, a grade IV astrocytoma that originates in the brain, is the most aggressive and common primary neoplasm . Owing to short median survival time (around 12–15 months) and poor prognosis, elucidation of molecular mechanisms in GBM is urgently needed . Although it was reported that CBX7 was downregulated and inhibited cell migration in glioma cells, the effect of CBX7 on cell cycle disregulation, another typical characteristic of GBM, is still unclear. In this study, we tested the relationship among CBX7 mRNA level, tumor grade and poor prognosis in CGGA, TCGA, REMBRANDT and GSE16011 datasets. We found that CBX7 possessed prognostic value in glioma patients. Then bioinformatics analysis indicated that CBX7 played a tumor suppressor role in glioma progression, especially in cell cycle control. CBX7 overexpression impaired the proliferation, migration and invasion of GBM cells; downregulation of CBX7 had the opposite effects. Furthermore, CBX7 inhibited G1/S phase transition via binding and silencing CCNE1 promoter mediated by the recruitment of HDAC2. Finally, CBX7 attenuated tumor growth in a xenografted model. Taken together, CBX7 participates in G1/S checkpoint control and is a novel prognostic marker in glioma patients.
|
study
| 99.94 |
Considering the contradictory role of CBX7 in different tumors, the transcription level of CBX7 in four glioma databases (CGGA, TCGA, REMBRANDT and GSE16011) was analyzed. We found that CBX7 mRNA levels in high-grade gliomas (HGG) were significantly decreased compared to normal brain tissues (NBT) and low-grade gliomas (LGG). CBX7 expression was also decreased in LGG in comparison to NBT (Figure 1A). Further, the association between CBX7 mRNA levels and clinical progression of glioma patients was determined using overall survival rates in GBM patients (samples from TCGA) and HGG patients (samples from CGGA, REMBRANDT and GSE16011), and Kaplan-Maier analysis and log-rank comparison were performed. In TCGA database, there was no significant difference in clinical outcome between the high and low CBX7-expression groups (high versus low median survival, 418 versus 375 d, respectively; p = 0.4922). However, in the three other databases, decreased CBX7 expression was associated with a shorter survival period (Figure 1B). Twenty human glioma specimens were divided into LGG or HGG groups and four brain samples obtained from epilepsy surgery were used for control group. In these samples we found that protein and mRNA levels of CBX7 are also negatively associated with tumor grade (Figure 1C and 1D). We also found that CBX7 expression possesses subtype preferences within gliomas (Supplementary Figure 1). These multi-center data and tumor samples revealed that CBX7 is a potential prognostic factor in glioma patients.
|
study
| 99.94 |
(A) The transcriptional levels of CBX7 were tested in CGGA, TCGA, REMBRANDT, and GSE16011 glioma datasets. (B) Kaplan-Meier survival curve was used to estimate the survival outcomes of patients divided into two groups with CBX7 low/high level. (C) Western blot analysis was used to test the levels of CBX7 in different glioma samples and normal brain tissues. Tubulin was used as a loading control. (D) Q-PCR was performed to determined CBX7 levels among NBT, LGG and HGG samples.
|
study
| 100.0 |
We performed a Pearson correlation analysis to discover genes that were associated with CBX7 in CGGA, Rembrandt and GSE16011 datasets (R > 0.4). To clarify the associations between these genes and specific GO functional categories, DAVID Web tool (http://david.abcc.ncifcrf.gov/home.jsp) and KEGG pathway analysis were used. In total 594 upregulated genes and 579 downregulated genes were identified (Figure 2A). As is shown in Figure 2B, decreased gene expression profiles were more enriched in pathways related to cell cycle and cancer signaling pathways. Furthermore, GO enrichment analysis also showed that these downregulated genes were strongly enriched in the cell cycle regulation process (Figure 2C). The in silico prediction of CBX7 function was in line with tumor-associated phenotypes.
|
study
| 100.0 |
(A) CBX7 associated genes obtained from overlapping CGGA, GSE16011 and Rembrandt databases via correlation analysis were analyzed with KEEG pathway analysis and GO enrichment analysis. (B) Enrichment results of KEGG pathways analysis were shown in different colours (red stands for up-regulated genes and green represents down-regulated genes). (C) Biological processes enrichment results from GO database. The orders of different biological processes were based on their enriched number.
|
study
| 100.0 |
Bioinformatics database-based results demonstrated that CBX7 was closely related to the progression and prognosis of gliomas and participated in cell cycle regulation. However, this in silico evidence was insufficient to elucidate the role of CBX7 reliably. Therefore, we constructed siRNAs targeting CBX7 and a lentivirus vector expressing CBX7. We then assessed the CBX7 protein levels in four GBM cell lines (Figure 3A). Other typical PRC1 elements in U251, U87 and LN229 cells were also tested (Supplementary Figure 2A). U251 cells containing high concentrations of CBX7 were used in knockdown experiments of CBX7 expression via the transient transfection of siRNA targeting CBX7 (si-CBX7) as well as a control vector (si-ctrl). The efficiency of three different sequences si-CBX7 was tested in U251 cells and the optimal si-CBX7 sequences was chosen for the subsequent experiments (Supplementary Figure 2B). We further constructed two stable CBX7 expressing cell lines, U87 and LN229, using a lentiviral CBX7-expression vector (Lenti-CBX7) and control cell lines using a control lentiviral vector (Lenti-ctrl) in the parental cells. The efficiency of up/downregulation of CBX7 is shown in Figure 3A.
|
study
| 100.0 |
(A) The protein levels of CBX7 in four classic GBM cell lines were examined through western blot assay (up panel). After transfection with siRNAs or lentivirus vectors, CBX7 expression in three GBM cell lines was assessed by western blot (down panel). (B) Respective merged images of U251, U87 and LN229 cells transfected with siRNAs or lentivirus vectors. Data are means of three independent experiments + SEM. *P < 0.05. Scale bar = 200 mm. (C) Proliferative abilities of GBM cells infected with siRNAs or lentivirus vectors were tested via CCK8 assays. *P < 0.05, #P < 0.001. (D, E) Wound-healing assay and transwell assay were performed to evaluate the migrative and invasive ability of GBM cells transfected with siRNAs or lentivirus vectors. *P < 0.05, #P < 0.001.
|
study
| 100.0 |
To explore the biological function of CBX7 in glioma cells, EdU, CCK8, wound-healing and transwell assays were performed to study the influence of CBX7 on cell proliferation, migration and invasion. After incubation with si-CBX7 for 48 h, the growth rate of U251 cells was increased. Conversely, the two stable CBX7 expressing cell lines, U87 and LN229, showed impairment in proliferation capability (Figure 3B and 3C). A known malignant behavior of glioma cells is its strong tendency for local invasion. We found that si-CBX7 treatment potentiated tumor aggressiveness in U251 cells compared with si-ctrl treatment. Conversely, the migrative and invasive capability in CBX7-overexpressing cell lines was suppressed (Figure 3D and 3E). In order to eliminate the influence of differences among cell types, we performed part of experiments in U251 with the use of siRNA and lentiviral vector and the results were consistent with the previous studies (Supplementary Figure 2C–2F). These findings imply that CBX7 is a potential tumor suppressor gene in glioma cells.
|
study
| 100.0 |
Our bioinformatics analysis indicated that CBX7-associated genes were enriched in cell cycle-related regulators. Moreover, Fusco and colleagues reported that CBX7 bound to the CCNE1 promoter as a multiprotein complex that included HDAC2 and then suppressed CCNE1 expression in mice and HEK293 cells . Thus, we emphasized analyzing the influence of CBX7 on the cell cycle and the mechanism by which CBX7 regulates CCNE1 expression. Flow cytometry analysis showed that the downregulated CBX7 levels promoted G1/S transition, which was halted by the overexpression of CBX7 (Figure 4A). The levels of cyclin E1, CDK2, p53 and p21 were meansured using western blotting (Figure 4B). Depleted CBX7 levels triggered an improvement in cyclin E1 and CDK2 levels, while the restoration of CBX7 expression lead to decreased cyclin E1 and CDK2 expression (Figure 4C). We also found that the expression of CCNE1 was regulated by CBX7 at the transcriptional level (Figure 4D).
|
study
| 100.0 |
(A) Flow cytometic assay was used to assess the cell cycle in GBM cells transfected with siRNAs or lentivirus vectors and the right panel showed the percent of different cell cycle phases. (B) Cyclin E1, CDK2, p53 and p21 levels were tested after the regulation of CBX7 in different GBM cell lines. (C) Immunofluorescence staining of CBX7 (green) and Cyclin E1 (red) in in different GBM cells. Scale bar = 50 mm. (D) Relative expression of CCNE1 in CBX7 up/down regulated cells was studied by Q-PCR. *P < 0.05. (E, F) ChIP experiment was used to examine the interaction between CBX7 and CCNE1 promoter. The precipitate was collected and evaluated by both RT-PCR and q-PCR. *P < 0.05, #P < 0.001. (G) ChIP assay showed that the occupancy of CCNE1 promoter by CBX7 was increased in CBX7 overexpression U87 and LN229 cells. P < 0.05. (H) Co-IP analysis was performed to evaluate the interaction of CBX7 and HDAC2. Total cell lysates from GBM cells were co-immunoprecipitated by anti-CBX7 antibody and anti-HDAC2 antibody. Protein-antibody complexes were analyzed by western blot and the results showed that CBX7 binds to HDAC2 in three cell lines. (I) After treated with 1μM HDCA2 inhibitor FK228 for 24 h, the degree of histone acetylation (acetyl-histone H3 Lys56 and acetyl-histone H2B Lys20) and cyclin E1 levels in three GBM cell lines were tested via western blot.
|
study
| 100.0 |
Chromatin Immunoprecipitation (ChIP) assays were conducted to test whether the CBX7 could bind to the CCNE1 promoter in glioma cells. Protein-chromatin complexes were subjected to immunoprecipitation with anti-CBX7 or anti-IgG antibodies and then the precipitate was analyzed using reverse transcription-PCR (RT-PCR) and quantitative real-time-PCR (q-PCR) (Figure 4E and 4F). Then, U87 and LN229 cells with exogenous expression CBX7 protein were used to further assess the interaction between CBX7 and CCNE1 promotor. Along with the increased CBX7 levels, the enrichment of precipitated CCNE1 promotor fragment was ascending (Figure 4G). To verify the interaction between CBX7 and HDAC2, lysates of U251, U87 and LN229 cells were subjected to IP with anti-CBX7 antibodies or nonspecific IgG. The immunocomplexes were immunoblotted with anti-HDAC2 antibodies. Conversely, anti-HDAC2 antibodies were used to pull down the proteins combine to HDAC2. Then, the immunocomplexes were immunoblotted with anti-CBX7 antibodies to ascertain that HDAC2 combines to CBX7. Using co-immunoprecipitation (coIP) assays, we confirmed that CBX7 interacted with HDAC2 (Figure 4H). We tested the degree of histone acetylation and cyclin E1 levels in three GBM cell lines treated with 1μM HDCA2 inhibitor FK228 for 24 h and found that inhibition of HDAC2 rescued the levels of cyclin E1 (Figure 4I).
|
study
| 100.0 |
An orthotopic mouse model using U87 cells was performed to further determine the role of CBX7 in gliomagenesis. U87 cells were initially transfected with luciferase expressing lentiviruses and then these cells were transfected with Lenti-ctrl or Lenti-CBX7. After implantation of U87 cells in mice, intracranial tumor volume was measured weekly via in vivo optical imaging system. Results from bioluminescence imaging showed that growth of glioma was significantly suppressed by CBX7 overexpression (Figure 5A). Concomitant with inhibition of tumor formation, mice treated with Lenti-CBX7 had prolonged survival times (Figure 5B). In addition, immunohistochemistry (IHC) revealed that upregulated CBX7 was accompanied by decreased cyclin E1, which was consistent with the in vitro results (Figure 5C). In summary, these findings showed that CBX7 inhibits glioma cell proliferation in vivo.
|
study
| 100.0 |
(A) Tumor volumes (n = 6 mice/group) were assessed by noninvasive bioluminescence imaging system and representative images taken at different days were shown in left panel. Quantification of the bioluminescence signal was plotted in right panel. P < 0.05. (B) Overall survival was calculated through Kaplan-Meier survival curves and the log-rank test was applied to estimate p-value between two groups. (C) IHC sections came from tumor slices were stained with CBX7 and Cyclin E1 and the representative profiles were demonstrated.
|
study
| 100.0 |
Epigenetic alterations such as histone modifications and DNA methylation broadly participate in biological functions and gene regulation, and are essential mechanisms in cancer development apart from genetic abnormalities [21, 22]. PcG proteins act as epigenetic regulators of gene expression and play pivotal roles in both physiological and pathological cellular activities [23–32]. CBX7 participates in the interaction between PRC2 and PRC1; another two PcG elements, EZH2 and BMI-1 belonging to PRC2 and PRC1, respectively, are described as oncoproteins [33–35]. However, the role of CBX7 in malignancy is still inconclusive. Our bioinformatics analysis has revealed that CBX7 is strongly linked to control of cell division. Indeed, CBX7 has been shown to repress Ink4a/Arf locus and impair p14Arf/p53- and p16Ink4a/Rb-dependent pathways in prostate cancer, human follicular lymphomas and gastric cancer [1, 2, 4, 36]. However, Ink4a/Arf locuses in U251, U87, and LN229 cells undergo homozygous deletion and the expression of p16Ink4a in HGG is abrogated . This phenomenon partially impair the oncogenic effect of CBX7. Thus CBX7 may possesse dual functions in cell cycle regulation and maintain a balance between cell division and quiescence. We tested p16 levels in three types of patient-derived primary GBM cells (G1, G2 and G3) and one p16-positive cell type (G1) was found (Supplementary Figure 3A). Compared to NHA, the CBX7 level in G1 was similar but G1 had a higher level of cyclin E1 and a lower level of p16 (Supplementary Figure 3A). After the upregulation of CBX7, the p16 expression was decreased and the cyclin E1 level was impaired (Supplementary Figure 3B). Interestingly, cyclin E1 abundance showed no significant difference in siRNA or lentivirus vector transfected NHA cells. By contrast, p16 protein levels were obvious changed along with the alteration of CBX7 in NHA (Supplementary Figure 3B). We suspect that the expression of CCNE1 in normal cells such as NHAs has been fully inhibited so that the regulation CBX7 cannot further effect CCNE1 expression. In cancer cells, cyclin E1 levels are up-regulated and, conversely, p16 levels was limited via various ways so that CBX7 acts as a tumor repressor in part of GBM cells. We assumed that this balance favors cell quiescence, owing to the anomalous genetic characteristics in GBM. To address this issue, we tested the function of CBX7 in glioma in vitro and in vivo.
|
study
| 99.94 |
In our study, we showed that CBX7 acted as a glioma suppressor via G1/S arrest. First, we analyzed CBX7 mRNA levels and patient outcomes in four public databases including CGGA, TCGA, REMBRANDT and GSE16011 databases. In all cohorts, the expression of CBX7 was significantly lower in HGG compared with LGG or NBT. The result showed that there was a significant inverse correlation between CBX7 level and glioma grade. In addition, survival analysis portrayed CBX7 as a valuable biomarker in survival prediction because the reduced expression of CBX7 correlated with poor outcome in HGG patients.
|
study
| 100.0 |
Three datasets were overlapped to mine CBX7-associated biological pathways. Both KEGG and GO analyses indicated that genes that were negatively correlated to CBX7 were strongly associated with the cell cycle pathway. Next, three GBM cell lines with different CBX7 abundance were chosen to further study the biological functions of CBX7. We observed that increased CBX7 protein levels inhibited glioma cells proliferation, migration and invasion. Then, we verified that CBX7 overexpression arrested cells in the G0/G1 phase. Moreover, we demonstrated that the underlying mechanism in CBX7 repression of CCNE1 promotor activity requiring HDAC2 recruitment and resulting in impaired CCNE1 transcription. Finally, orthotopic glioma models in nude mice implanted with CBX7-overexpressing U87 cells were used. Non-invasive bioluminescence imaging and survival times of nude mice revealed that CBX7 behaved as a tumor suppressor gene in gliomas.
|
study
| 100.0 |
The interaction between cyclin E and CDK2 plays an essential role in the G1/S phase transition via phosphorylation of p27Kip. Misregulation of cyclin E/CDK2 complex occurs in various types of tumors and leads to uncontrolled cell growth [38–40]. Thus, the CBX7-induced suppression of CCNE1, which encodes cyclin E1, contributed to the attenuation of glioma progression. Considering the complexity of phosphorylated proteins downstream of the cyclin E/CDK2 complex, further investigation is warranted. Furthermore, the interaction between CBX7 and other PcG proteins such as EZH2, CBX6 and CBX8 will require additional clarification.
|
study
| 100.0 |
In conclusion, our in vitro and in vivo results validate the assumption that CBX7 is a tumor suppressor of gliomas. Moreover, CBX7 is a potential and novel prognostic biomarker in glioma patients. In addition to showing CBX7-induced phenotypic changes, we also found that CBX7 silences CCNE1 via the combination of CCNE1 promoter and the recruitment of HDAC2.
|
study
| 100.0 |
Whole genome mRNA expression microarray data and clinical information of 220 glioma samples and 5 normal brain tissues were obtained from Chinese Glioma Genome Atlas (CGGA) database (http://www.cgga.org.cn) and used as discovery set . Three validation sets are The Cancer Genome Atlas database (TCGA, http://cancergenome.nih.gov), Repository of Molecular Brain Neoplasia Data (REMBRANDT, https://caintegrator.nci.nih.gov/rembrandt/) and GSE16011 data (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE16011).
|
study
| 99.94 |
Human glioblastoma cells (LN229, U87, and U251) were purchased from the cell bank of the Chinese Academy of Sciences (Shanghai, China). Primary GBM cells (G1, G2 and G3) that derived from three primary GBM surgical specimens were maintained in DMEM supplemented with 10% FBS. Normal human astrocytes were purchased from ScienCell Research Laboratories (USA). Three different sequences of CBX7 siRNA were purchased from GenePharma (Shanghai, China). Lentivirus vectors and plasmids were purchased from GeneChem (Shanghai, China). FK228 was purchased from MedChemExpress (USA).
|
other
| 99.6 |
Four normal brain tissues, ten primary LGG and ten HGG samples were obtained from Department of Neurosurgery, the First Affiliated Hospital of Nanjing Medical University (Nanjing, China). The clinic pathological and general information of those patients are shown in Supplementary Table 1. Written informed consent was obtained from all patients. Our study was approved by the Institutional Review Board and the ethics committee of Nanjing Medical University and Harbin Medical University.
|
study
| 99.0 |
RIPA lysis buffer (KeyGEN, Jiangsu, China) was used to extract proteins from cells. Equal amounts of protein were separated by SDS-PAGE, followed by electrotransfer onto polyvinylidene difluoride membranes (Thermo Fisher Scientific, Massachusetts, USA). Immunoblot analysis used the following primary antibodies: CBX7 (ab21873), HDAC2 (ab12169), Cyclin E1 (ab3927), p16 (ab51243), p53 (ab28), Actin (ab8226), Tubulin (ab7291) and GAPDH (ab9485) were obtained from Abcam (Cambridge, UK). P21 (#2946), acetyl-histone H3 (Lys56) (#4243), acetyl-histone H2B (Lys20) (#2571) were purchased from Cell Signaling Technology (Massachusetts, USA).
|
study
| 99.44 |
SiRNA or lentivirus transfected cells were seeded in 6-well plates and cultured until reached 100% confluence. Then, sterile pipette tips were used to make scratches across the cell monolayer and the photographs of scratched areas were taken. 24 h later, different scratched areas were selected in each well and were photographed under inverted microscope (Nikon, Tokyo, Japan). The cells protruding from the border of the scratches were counted.
|
study
| 99.94 |
Plates with transwell inserts (Corning, New York, USA) which were pre-coated with 20 μg/μL Matrigel (BD Biosciences, New Jersey, USA) were used to test cell invasion. The upper chambers with 200 μL serum-free media were added with 50,000 siRNA or lentivirus transfected cells. In parallel, 900 μL DMEM media with 10% FBS was added to lower chamber of each well. After incubation for 24 h at 37°C, cells from the upper surface of the membrane were removed with a cotton swab and the penetrated cells were fixed with 4% methanol for 5 min and then stained with 0.1% crystal violet for 30 min. Six fields of cells were captured and counted randomly under inverted microscope (Nikon, Tokyo, Japan) at ×20 magnification in each well. Upper chambers without Matrigel were used to test cell migration.
|
study
| 99.94 |
SiRNA or lentivirus transfected cells were harvested and washed with PBS. Next, 75% ethanol was used to fix cells at −20°C overnight. Then, DNA in cells was stained by Hank's balanced salt solution including 50 mg/mL propidium iodide and 50 mg/mL RNaseA for 1 h at room temperature. Cell cycle of these cells was analyzed by Gallios flow cytometer (Beckman Countler).
|
study
| 99.9 |
SiRNA or lentivirus transfected cells were collected and lysed using lysis buffer supplemented with PMSF. Then, equal amounts of protein was subjected was subjected to anti-CBX7 antibody or anti-HDAC2 antibody following overnight incubation at 4°C. Then, protein-antibody immunoprecipitates were collected by protein A/G plus-agarose (Santa Cruz Biotechnology, Texas, USA).
|
study
| 99.94 |
ChIP assays were conducted using reagents commercially obtained from Upstate Biotechnology and performed according to the manufacturer's instructions. Briefly, glioma cells were then fixed with formaldehyde for 15 min. Then cells were lysed in SDS lysis buffer, and the chromatin DNA was extracted and sonicated into 200–1000 bp fragments. Purified DNA was used for PCR amplification. Enrichments at target sites were compared with IgG group.
|
study
| 99.94 |
CCNE1 promoter primer sequences: F: 5′-ACGTGACCGTTGTGAGTCAA-3′; R:5′-CAGAGAGAAGAAGAGAAAGCTGAT-3′. CCNE1 primer sequences: F: 5′-AAGGAGCGGGACACCATGA-3′; R: 5′-ACGGTCACGTTTGCCTTCC-3′. GAPDH primer sequences: F: 5′-TGTGGGCATCAATGGATTTGG3′; R: 5′-ACACCATGTATTCCGGGTCAAT-3′.
|
other
| 99.9 |
Cells were incubated with anti-CBX7 and anti-Cyclin E1 overnight and then with Alexa 488- or 568-labeled anti-mouse or anti-rabbit IgG antibody (Thermo Fisher Scientific, Massachusetts, USA) 2 h at room temperature. After treated with DAPI (Beyotime Biotechnology, Jiangsu, China) for 10 min, cells were examined with Zeiss axiophot photomicroscope (Carl Zeiss AG, Jena, Germany).
|
other
| 93.75 |
Animal experiments were approved by the Animal Management Rule of the Chinese Ministry of Health (documentation 55, 2001) and were in accordance with the approved guidelines and the experimental protocol of Nanjing Medical University. Luciferase stably expressed U87 cells were transfected with Lenti-CBX7 or Lenti-Ctrl. Twelve mice purchased from Cancer Institute of the Chinese Academy of Medical Science were randomly divided into CBX7 upregulated group or control group. To establish intracranial glioma model, each nude mouse was implanted with 1 × 106 U87Luciferase/CBX7 or U87Luciferase/Ctrl cells (day 0). The intensity of bioluminescence was positively correlated with the number of U87 cells. Tumor growth was assessed weekly by live animal bioluminescence imaging system. After four times testing (day 7, day 14, day 21 and day 28), the survival of mice was observed until day 60. Then mice were exposed to CO2 to achieve euthanasia. Brains tissue sections were incubated with antibodies including CBX7 and Cyclin E1.
|
study
| 100.0 |
T-test was employed to analyze differences in each two-group comparison and one-way ANOVA was performed to determine the difference among at least three groups. Kaplan–Meier analysis was used to assess the survival rate of mice. Heat maps were performed with Gene Cluster 3.0 and Gene Tree View software. Kaplan–Meier analysis was used to assess the survival rate of patients and mice. KEGG pathway and GO analysis were performed via DAVID (http://david.abcc.ncifcrf.gov/). P < 0.05 was considered statistically significant.
|
study
| 99.94 |
Endophthalmitis after vitreous surgery is an uncommon but a devastating complication. The reported incidence of endophthalmitis after vitreous surgery ranges from 0.05% to 0.02% with 20-G vitrectomy.[2–10] Initial reports showed higher rates of post-minimally invasive vitreoretinal surgery (MIVS) endophthalmitis, whereas recent reports have shown a declining trend.
|
review
| 99.9 |
A recent multicentric study from India reported incidence rates of endophthalmitis after vitrectomy surgery to be 0.052% with culture-positive endophthalmitis being 0.031%. Although worldwide large multicentric studies have been conducted, these are also not without limitations. These retrospective multicentric studies had among centers heterogeneous case selection criteria, had contrasting antibiotic prophylaxis practice patterns, had different infection control protocols, and had different laboratory and treatment protocols which could affect the anatomical and functional outcomes.
|
review
| 99.7 |
The purpose of this study was to describe the 20-year incidence (20 G and MIVS), causative organisms, and visual outcomes associated with endophthalmitis after pars plana vitrectomy from a single, tertiary eye care center in India. Furthermore, to identify the differences in clinical presentation between post-cataract surgery endophthalmitis and post-pars plana vitrectomy endophthalmitis, we compared our data with previous post-cataract surgery endophthalmitis studies after cataract surgery from India. To the best of our knowledge, this is the largest case series of incidence of post-vitrectomy endophthalmitis reported in literature.
|
clinical case
| 94.75 |
Approval of the Institutional Review Board of Sankara Netharalaya was obtained to analyze the hospital-based data, and the tenets of the Declaration of Helsinki were followed. This was a retrospective review of case records of patients with suspected postoperative infectious endophthalmitis after pars plana vitreous surgery performed at a tertiary eye care institute in India was done. Data were collected from the electronic/paper medical records between January 1995 and December 2015; the charts of all patients who developed endophthalmitis were reviewed. For cross-validation of our data, the endophthalmitis surveillance log of the Department of Microbiology was reviewed. Acute endophthalmitis after vitrectomy was defined as the presence of unusual inflammation in the anterior segment and vitreous cavity within 6 weeks of surgery along with microbiological evidence (staining/culture/PCR) of bacterial and fungal infection. Culture-positive endophthalmitis was considered to be present if microbiological organisms were isolated on culture. If the fungal, aerobic, and anaerobic culture results were found to be all negative, then this case was considered to be culture-negative endophthalmitis. Polymerase chain reaction (PCR) data on the presence or absence of eubacterial, Propionibacterium acnes, and fungal genomes were also noted.
|
study
| 99.94 |
Data such as presenting complaints, systemic illness, time of presentation, presenting and final visual acuity, clinical findings, intraoperative procedure, microbial profile of aqueous, vitreous, or any other tissue/material, antimicrobial susceptibility, management including both intravitreal antibiotics and vitrectomy, and follow-up were collected. Informed written consent was obtained to have their medical records used in research.
|
study
| 99.5 |
Before 2009, all cases underwent 20-G pars plana vitrectomy. After that the practice shifted to MIVS and the number of cases undergoing 20-G vitrectomy declined considerably. Of the 111,876 vitrectomies performed during the study period, 70,585 (63.1%) were 20-G surgeries and 41,291 (36.9%) were MIVS. Before the year 2000 the hospital infection control protocol included pre-operative antibiotics (sulfacetamide, 10.0%) eye drops four times per day for atleast 3 days prior to surgery. Intra-operatively, peri-ocular cleaning was done with 0.5% cetrimide solution (Cettan, Tansi Polish Unit, Chennai). After the year 2000, preoperative antibiotic protocol included instillation of ciprofloxacin (0.3%) or sulfacetamide (10.0%) eyedrops 2 hourly one day before surgery. Pre-operatively 5% povidone-iodine solution was instilled in conjunctival sac and was allowed a contact period of 3 minutes followed by peri-ocular cleaning with 1% povidone iodine solution before draping. The details of indication for surgery, additional procedures performed along with vitrectomy, gauge of vitrectomy instruments, lens status at the end of vitreous surgery, type of tamponade used during surgery, and surgical time were noted in all the cases who developed endophthalmitis.
|
study
| 99.94 |
On clinical suspicion of endophthalmitis, if vitreous surgery was not contemplated in the next 6 hours, an anterior chamber tap was performed and the samples were subjected to gram staining, KOH staining, bacterial and fungal cultures, antibiotic sensitivity, and PCR for P. acnes, eubacterial and panfungal genomes. If surgery was contemplated, vitreous specimen was collected during vitrectomy and similar microbiological tests were conducted. On the basis of microbiological results, initial antibiotics were given. However, in cases of initial negative results of staining, choice of antibiotics was based on clinical judgment. Scleral or corneal scraping were taken for microbiological evaluation, when indicated.
|
other
| 99.9 |
Statistical analysis was performed using SPSS Statistics for Windows version 21.0. Continuous variables were expressed as mean ± SD and were analyzed using Mann–Whitney U test. Categorical variables were analyzed using Chi-Square and Fisher exact test, as applicable. The P values of ≤0.05 were considered to be statistically significant.
|
other
| 99.5 |
Table 1 shows the incidence rates of endophthalmitis after pars plana vitrectomy in 20G and MIVS group. The overall incidence of clinically evident and culture-proven endophthalmitis after vitrectomy was 0.040% (4.0 cases per 10,000 surgeries) and 0.021% (2.1 cases per 10,000 surgeries), respectively. The incidence of clinically evident and culture-proven endophthalmitis after 20G vitrectomy was 0.057% (5.7 cases per 10,000 surgeries) and 0.031% (3.1 cases per 10,000 surgeries), and 0.012% (1.2 case per 10,000 surgeries, P < 0.05) and 0.005% (0.5 cases per 100,000 surgeries, P < 0.05) in the MIVS group, respectively. The rate of culture negative endophthalmitis was 0.019% (1.9/10,000 surgeries), 0.025% (2.5/10,000 surgeries) in 20G and 0.007% (0.7/10,000 surgeries) in the overall, 20G and MIVS group respectively. Fig 1 shows the trend in the rate of post-vitrectomy endophthalmitis for the study period of 1995–2015. Over the last decade the has been there is an indisputable decrease noticed in the rate of endophthalmitis post-vitrectomy; the rate decreasing from 0.082% for the study period of 2003–2006 to 0.01% for the study period of 2011–2015.
|
study
| 99.94 |
Table 2 shows the baseline characteristics and surgical details of the 45 postvitrectomy endophthalmitis patients in the 20-G versus MIVS group. Overall, the group had significant proportion of patients with altered immune status (diabetes, 46.7%; and those on peri-operative oral steroids, 11.1%). One (2.2%) patient had an infective focus (tooth caries). 7 (15.5%) of 45 eyes had intraocular surgery within 1 year prior to vitrectomy (MIVS, 3 eyes; and 20 G, 4 eyes). The most common indications for vitrectomy were vascular retinopathies (44.4%), followed by rhegmatogenous retinal detachment (17.8%). In 16 (35.5%) eyes, vitrectomy was combined with an anterior segment procedure. There were no differences between the two gauges in terms of duration of surgery, number of sclerotomies, use of tamponade, and the lens status at closure. Of the total 111,876 cases, 8 (of 19741 cases) and 37 (of 92135 cases) cases of endophthalmitis occurred before and after introduction of pre-operative povidone iodine drops. Thus, an incidence rate of endophthalmitis before and after the introduction of pre-operative povidone iodine drops were 0.041% and 0.040% respectively (p>0.05).
|
study
| 99.94 |
We compared our data with historical data on endophthalmitis after cataract surgery to identify the differences of clinical presentation for endophthalmitis after cataract surgery versus vitrectomy. Table 3 shows the difference in clinical presentation between postvitrectomy and postcataract surgery endophthalmitis groups: 73.4% versus 50% patients in the postvitrectomy and postcataract surgery groups, respectively, presented within the first week (P = 0.006). Presence of fibrin was a finding in 93.3% patients in the postvitrectomy group compared to 72.5% patients in the postcataract surgery group (P = 0.004). Corneal edema and absence of red fundal glow (Grade 5) was more common in the postcataract surgery group compared to the postvitrectomy group (P = 0.001). Only 44.4% patients in the postvitrectomy group had normal intraocular pressure (IOP). Low and high IOP was recorded in 8.9% and 46.6% patients in the postvitrectomy endophthalmitis surgery group. This is in contrast to that reported by Endophthalmitis Vitrectomy Study (EVS), where normal IOP was seen in 80.6% of patients. Postoperative hypotony (defined by IOP < 7 mmHg) was seen in one (20%) patient in the MIVS group. Presence of corneal ulcer and scleral necrosis involving the sclerotomies were findings specific to this study.
|
study
| 99.94 |
Fig 2 shows the various intraocular samples taken and microbiological results obtained. The mean number of samples obtained was 1.4 per patient. Of the total 63 intraocular samples obtained, aqueous, vitreous, and others constituted 33 (52.4%), 20 (44.4%), and 10 (22.2%) samples, respectively. Smear, culture, and PCR positivity in aqueous samples was 56.2% (18/32), 33.3% (11/33), and 100% (17/17), respectively, and in vitreous samples was 45% (9/20), 35% (7/20%), and 87.5% (7/8), respectively. Overall smear, culture, and PCR positivity in both the groups was 54.5% (30/55), 38.6% (22/57), and 96% (24/25), respectively, in our case series.
|
study
| 100.0 |
Fig 3 shows the microbiology findings of 45 postvitrectomy endophthalmitis patients. Culture results of 24 (53.3%) cases were found to be positive in our series. The culture proven rates were 55.0% for 20G and 40.0% for MIVS, respectively. The organisms isolated in the 24 culture-proven cases were, in decreasing frequency,: gram-negative organisms, 12/24 (50.0%); gram-positive organisms, 5/24 (20.8%); fungus, 4/24 (16.7%); P. acnes, 1/24 (4.2%); combined gram-negative and gram-positive organisms, 1/24 (4.2%); and combined bacterial and fungal, 1/24 (4.2%). The common organisms isolated in our case series were: Pseudomonas aeruginosa, 5/24 (20.8%); Staphylococcus epidermidis, 2/24 (8.3%); Staphylococcus aureus, 2/24 (8.3%); Aeromonas hydrophila, 2/24 (8.3%); Acinetobacter calcoaceticus, 3/24 (12.5%); and Aspergillus species, 5/24 (20.8%). In one case each, Enterococcus faecalis (4.2%), P. acnes (4.2%), Alcaligenes faecalis (4.2%), Corynebacterium (4.2%), Klebsiella ozaenae (4.2%), and Pseudomonas stutzeri (4.2%) was isolated.
|
study
| 99.94 |
S1 Table shows the antibiotic sensitivity of the cases with positive cultures. The antibiotic sensitivity of gram-negative organisms was as follows: amikacin, 9/11 (81.8%); ciprofloxacin, 9/11 (81.8%); ceftazidime, 7/9 (77.8%); gentamicin, 9/12 (75.0%); and cefatoxime, 7/11 (63.6%) In gram-positive organisms, the antibiotic sensitivity pattern was as follows: cefatoxime, 4/4 (100%); ciprofloxacin, 4/4 (100%); vancomycin, 3/3 (100%); clindamycin, 3/4 (75%); and ceftazidime, 2/3 (66.7%). Of the 21 cases whose culture results were found to be negative, PCR results were positive in 20 (95.2%) and smear results were positive in only 1 (4.8%) patient (gonococcal bacilli). The results of PCR were as follows: eubacterial only (45%); eubacterial + pan fungal (15%); eubacterial + P. acnes (30%); and eubacterial + pan fungal + P. acnes (10%).
|
study
| 100.0 |
Seven eyes were lost to follow up beyond one week and were excluded from further analysis. The follow-up period for the patients with endophthalmitis was 586.14 ± 825.15 days. Table 4 shows the overall visual outcomes in various subgroups. Overall, 13 (34.2%) patients had a favorable visual outcome (i.e., best-corrected visual acuity [BCVA] > 5/200), and 24 (63.2%) had an unfavorable visual outcome (BCVA < 5/200). Overall, 5 (13.1%) and 10 (26.3%) patients had a vision of better than or equal to 6/18 and 6/60, respectively. The globe could be salvaged in 27 (71.0%) cases. The most common cause of poor vision was macular pathology, including scar, epiretinal membrane, and thinning across the three groups. No statistically significant difference was observed in outcome in different subgroups: gram negative versus gram positive; those who had tamponade versus those who did not have tamponade. However, cases with positive culture (particularly gram-negative and fungus-infected) had more unfavorable outcomes (P < 0.05). Although no significant difference was observed in the clinical presentation between the groups (P > 0.05), patients in the culture-positive group presented relatively early compared to the culture-negative group (75.0% vs. 66.7% at week 1). A higher proportion of patients in the culture-negative group were treated with only intraocular antibiotics (IOAB) compared to those in the culture-positive group (52.4% vs. 25.0%; P = 0.120).
|
study
| 99.94 |
We report the clinical features, microbiological profile, and outcomes in 45 cases of endophthalmitis following 111,876 vitrectomies performed over a period of 20 years. The incidence of clinically evident and culture-proven endophthalmitis after 20G vitrectomy was 5.7 and 3.1 cases per 10,000 surgeries, respectively. Previous studies have shown varied incidences with 20-G vitrectomy: 7 cases per 10,000 surgeries in earlier studies and reducing to 2 cases per 10,000 surgeries reported in the later studies.[2–10]
|
study
| 99.94 |
The incidence of endophthalmitis after MIVS reported in literature in the last decade have shown a declining trend. In a large, multicenter retrospective Pan-American Collaborative Retina Study Group, Wu et al. found an incidence of 0.026% (2.6 cases per 10,000 surgeries) for endophthalmitis after MIVS during 2005–2009. Scott et al. in their multicenter retrospective study found an incidence of 0.048% (4.8 cases per 10,000 surgeries) for endophthalmitis after MIVS during 2007–2008. They also found that the rate of endophthalmitis after MIVS in their study period was marginally lower than that estimated for 2005–2006, suggesting the decreasing trend. Similar indisputable decreasing trend is evident in our study (Fig 1) with the rate decreasing from 0.082% to 0.010% in the last decade. Lower rates observed in our study compared with previous studies are possibly due to the various reasons. As the incidence of endophthalmitis in the MIVS group is calculated for 2009–2014, we ascribe these lower rates to the decreasing trend observed in the past. Also, as our data represent the incidence from a single center, data are possibly less heterogeneous in terms of inclusion criteria, antibiotic prophylaxis, standard infection prevention protocols, and surgical technique except for those that have evolved over time. Moreover, the inclusion criteria of our study is more robust. We included only those cases that had an unusual post-operative inflammation sufficient enough to warrant a microbiological evaluation of ocular fluids and had microbiological evidence (staining positive/culture positive/PCR positive) of the same.
|
study
| 99.94 |
Although we found that the risk of endophthalmitis is significantly lower after MIVS compared to conventional 20-G vitrectomy, caution is warranted in interpreting this finding. As the incidence rates of endophthalmitis after 20-G vitrectomy and MIVS represent different study periods, comparing the two may not reflect the true difference. Since the first introduction of 25-G vitrectomy by Fujii et al. in 2002, the incisional techniques and technology of MIVS have evolved considerably, exemplified by the reported of decreasing incidence of endophthalmitis in the last decade. At best, our results suggest no increased risk of endophthalmitis after MIVS.
|
study
| 99.9 |
The potential predisposing factors identified in our study were as follows: diabetes (46.7%), vitrectomy for vascular retinopathies (44.4%), vitrectomy combined with anterior segment procedures (35.5%), and absence of tamponade (59.5%). History of intraocular surgery was more often seen in patients in the MIVS group. Diabetes, and immune-compromised states, can theoretically increase the risk of endophthalmitis after vitrectomy. Complicated vitreoretinal surgeries requiring multiple exchange of instruments, longer operative time, increased incidence of cataract requiring combined anterior segment surgeries, and poor wound healing can increase the risk of endophthalmitis. Similar associations have been noted in other studies.[1–5,7,9]1–5,7,9
|
study
| 99.94 |
The differential surface tension of a tamponading agent compared to balanced salt solution helps to seal the sclerotomy wounds, minimizing hypotony and associated infections.[17–19] Although this study has not been designed to show the association of use of tamponade and endophthalmitis, we found absence of tamponading agent in all cases with endophthalmitis (n = 5) who underwent MIVS where sclerotomies were self sealing. Park et al. in their series found that vitrectomy for retinal detachment was associated with reduced risk of infection. They suggested that as in these cases the use of a tamponading agent probably results in better wound integrity.
|
study
| 99.94 |
Because most patients have varying degree of pain, redness, chemosis, lid edema, and anterior segment inflammation, recognizing endophthalmitis early is sometimes challenging. We found a different presentation of endophthalmitis after postvitrectomy compared to that after cataract surgery in terms of fibrin, corneal edema, and IOP, suggesting a differential response of eyes in these two conditions.[14–15] We found that a significant proportion of postvitrectomy endophthalmitis patients (73.4%) presented within 1 week of surgery, which is in accordance with the results of the previous studies. This implies that after vitrectomy, patients must be reviewed in the first week to allow early detection of this dreaded complication. With the presence of fibrin (present in 93.3% patients) and abnormal IOP (present in 55.6% patients, especially high IOP), clinician should suspect endophthalmitis as an important differential diagnosis. One additional finding in patients of postvitrectomy endophthalmitis was the presence of corneal ulcer and scleral necrosis in the area of sclerotomy. This indicates that meticulous examination of the sclerotomy site is important in the post-operative period.
|
study
| 99.9 |
We found 53.3% culture positivity in our case series, which is in accordance with previously reported rates between 44.4% and 66.7%. Park et al. suggested the importance of aqueous samples in isolating the microorganisms in endophthalmitis after vitrectomy. We found that aqueous was the most common intraocular sample obtained (52.4%) with a culture-positive yield similar to that of vitreous samples (33.3% vs. 35.0%), further validating its use.
|
study
| 100.0 |
An important distinctive feature of our study is the use of PCR for microbial analysis. Chiquet et al. found culture positivity in only 32% aqueous samples as compared to 61% on PCR in their postcataract surgery endophthalmitis case series study. Using culture and PCR combination, diagnosis can be made in 71% cases. Similarly, Therese et al. showed that by inclusion of PCR, detection of microorganisms increased from 46.5% to 75.8% cases. Likewise, Anand et al. showed a high sensitivity for detection of fungus using PCR in endophthalmitis.
|
study
| 99.94 |
Reports from Western countries have shown coagulase negative Staphylococcus as the most common organism in postoperative endophthalmitis.[1–3,5,6,9,13,22] Gram-negative bacilli were the most common group isolated in our study. Pseudomonas was the most common organism cultured. Our finding is in agreement with previous reports from Indian subcontinent. Sharma et al. have reported a high prevalence of gram-negative bacteria and fungus in postoperative endophthalmitis. Anand et al. and Jambulingam et al. have reported similar spectrum in postoperative endophthalmitis from our center. A similar high prevalence of gram-negative bacteria has been reported in other infections in subjects with diabetes. We believe both varied geographical distribution and predominance of subjects with diabetes in our study are responsible for this microbial spectrum.
|
study
| 99.94 |
In this study, a significant number of culture-negative cases could be effectively managed with intravitreal antibiotic injections with relatively better visual and anatomical outcomes. Because the genomic bacterial load predicts severity and outcomes, it is possible that a lower bacterial load favored better outcomes and corresponding inability to capture the microbe on culture. This seems most probable as 75% culture-positive patients presented within 1 week of surgery, in contrast to 66.7% culture-negative patients, owing to this difference in genomic bacterial load.
|
study
| 100.0 |
Overall, the functional outcomes in endophthalmitis after vitrectomy were poor, 24 (63.2%) had final vision <5/200, these results are in accordance with those of previous studies.[1–7,9,12,13] The outcomes are poorer than that of endophthalmitis following cataract surgery due to the possible underlying retinal pathology. The high prevalence of gram-negative and fungal microorganisms in our study partly explains the poor visual outcome in our series. Park et al. reported only 14.8% patients had vision better than 6/12. Scott et al. reported 7 of the 13 cases had a vision worse than 20/100.
|
study
| 99.94 |
The study has several strengths. First, it is the largest case series of endophthalmitis after vitrectomy reported in literature; Second, being a single-center study, it scores over multicenter studies that are confounded by their inherent heterogeneous study design. Third, it addresses the concerns regarding the fear of increased endophthalmitis rate with MIVS using the largest database of 41,921 MIVS surgeries till date. Fourth, the differences between culture-positive and culture-negative cases have been shown in endophthalmitis after vitrectomy. And finally, the use of PCR, a modern sensitive tool used in our study, to exclude probable noninfectious cases has been appropriately demonstrated.
|
study
| 99.94 |
In conclusion, using a large dataset, we showed that MIVS does not increase the risk of endophthalmitis, despite recent concerns. Meticulous surgical asepsis, use of endotamponade, suturing of sclerotomy in doubtful cases and judicious use of antibiotics are crucial to minimize the risk of infection. Surgeons should be aware of the differences in presentation, microbiological profile, and factors responsible for unfavorable outcome. The outcomes are often guarded despite appropriate treatment. The risk of poor outcomes is potentially higher in culture-positive cases, which need to be treated aggressively.
|
study
| 99.5 |
Diet is a primary factor used to characterize the ecological niches of species and populations, including classification along the generalist–specialist spectrum (Elton, 2001; Hutchinson, 1957). However, dietary generalism usually is coarsely characterized, with the role of nutrients in defining a species’ niche overlooked (Machovsky‐Capuska, Senior, Simpson, & Raubenheimer, 2016). This is problematic because nutrition plays a dominant role in determining which foods species consume, and thus which environments they inhabit (Raubenheimer, Simpson, & Tait, 2012). The field of nutritional ecology has demonstrated that, in particular, the macronutrients (protein, carbohydrate, and lipid) in the foods and diets of animals strongly influence their foraging behavior (Coogan et al., 2017; Rothman, Plumptre, Dierenfeld, & Pell, 2007) and ultimately fitness, including reproduction and longevity (Jensen et al., 2012; Solon‐Biet et al., 2014). Neglecting nutrition in niche theory might therefore limit our understanding of the ecological factors that determine the abundance and distribution of species.
|
review
| 99.8 |
Recently, a multidimensional framework was developed for integrating nutrition and ecological niche theory (Machovsky‐Capuska, Senior, et al., 2016). This approach characterizes the dietary niche of species across four functional levels: (1) the diversity of physical and ecological characteristics of foods a species can utilize (“food exploitation niche”); (2) the range of food macronutrient compositions a species can consume as part of its diet (“food composition niche”); (3) the range of dietary macronutrient compositions that a species is physiologically capable of persisting on (“fundamental macronutrient niche”); and (4) the range that it actually persists on, given ecological constraints such as food availability and competition (“realized macronutrient niche”).
|
review
| 82.9 |
The purpose of this study was to investigate patterns of dietary generalism in a large widely distributed mammalian omnivore, through the lens of macronutritional niche theory. To that end, we selected the brown bear (Ursus arctos) as an exemplar model species. The brown bear has an ecologically and geographically wide (i.e., circumpolar) distribution (Pasitschniak‐Arts, 1993). Across its range, the brown bear has a polyphagous diet consisting of a wide range of foods that vary in physical, ecological, and nutritional properties, depending upon both seasonal and local availability (Bojarska & Selva, 2012; Coogan, Raubenheimer, Stenhouse, & Nielsen, 2014). As apex predators, brown bears consume a range of animal prey, such as small and large mammals, insects, fish, and birds (e.g., Ciucci, Tosoni, Di Domenico, Quattrociocchi, & Boitani, 2014; Rigg & Gorman, 2005). Brown bears also consume a variety of graminoids and forbs, consume both soft mast (i.e., fruit) and hard mast (i.e., nuts), and possess the ability to dig for and consume belowground vegetation (e.g., roots and corms, Hamer & Herrero, 1987). Brown bears are also known to obtain a variety of both plant‐ and animal‐based foods from anthropogenic sources, including grain, livestock, and human food waste (Gunther et al., 2004; Murray, Fassina, Hopkins, Whittington, & St Clair, 2017). In the context of the multidimensional nutritional niche, the brown bear can thus be characterized as a generalist in both food exploitation (level 1 above) and food composition (level 2 above).
|
study
| 99.94 |
What are not known, however, are the fundamental and realized macronutrient niches (levels 3 and 4 above) of brown bear, which are important for understanding the relationships between their nutritional environments, adaptation, population persistence, and functional omnivory more generally. The fundamental macronutrient niche concept is also germane to understanding and implementing applied ecology, such as whether a population (or individual) is likely to persist in the face of substantial perturbations to its nutritional environment, due to factors such as climate change, translocation, and dispersal.
|
other
| 99.8 |
A macronutrient self‐selection study of captive brown bears found that they preferred a ratio of 17% protein to 83% nonprotein (carbohydrate + lipid) metabolizable energy (Erlenbach, Rode, Raubenheimer, & Robbins, 2014). The study provided strong evidence that this ratio is functionally significant, because compared with other dietary compositions it maximized mass gain (primarily fat mass) per unit energy intake, which is an important outcome for a hibernating mammal. Bears in that study preferred high‐lipid intake, but when confined to low‐fat diets would maintain the target ratio of protein to nonprotein energy by consuming carbohydrate.
|
study
| 99.94 |
Brown bears in the wild, however, may be precluded from foraging to meet such nutritional preferences, because foods high in lipid or carbohydrate necessary to maintain a balanced intake are generally most available during late summer and autumn (Coogan et al., 2014). That foods available to achieve the optimal macronutrient ratio for primarily fat mass gain co‐occur with the prehibernation hyperphagic period, is suggestive of the functional significance and selective pressures shaping their behavioral dietary preferences; the nutritional and energetic demands necessary for hibernation require the acquisition of sufficient food resources (Rigano et al., 2017), with higher demands for females birthing cubs (López‐Alfaro, Robbins, Zedrosser, & Nielsen, 2013). It is unclear, however, the extent to which bears regulate their diets in the wild.
|
study
| 99.94 |
Here, we infer the minimal fundamental macronutrient niche of brown bear using population diet estimates as indicators of realized macronutrient niches. To that end, we collected and synthesized data from the literature to estimate the proportions of macronutrients in the diets of brown bear populations, following the approach recently applied to a small carnivore (Martes martes; Remonti, Balestrieri, Raubenheimer, & Saino, 2016) and invasive omnivore (Sus scrofa; Senior, Grueber, Machovsky‐Capuska, Simpson, & Raubenheimer, 2016). We then applied a compositional statistical paradigm (Aitchison, 1982) to our data analysis of macronutrient proportions. We hypothesized that the brown bear's broad diet could be associated with the following functional adaptations of omnivory: The Nutrient Balancing Hypothesis predicts that a wide diet serves to increase the range of food options that can be combined to achieve a balanced macronutrient intake. This hypothesis predicts that the macronutrient composition of bear diets from different ecological and geographic populations is similar despite consuming different foods (i.e., a narrow fundamental macronutrient niche). This type of nutrient balancing has been observed in primates (Raubenheimer, Machovsky‐Capuska, Chapman, & Rothman, 2015; Rothman et al., 2007), badger (Meles meles; Remonti, Balestrieri, & Prigioni, 2011), and pine marten (Martes martes; Remonti et al., 2016).The Nutritional Generalism Hypothesis postulates that a broad diet combined with the ability to tolerate a wide range of dietary macronutrient intakes enables a species to occupy a diverse range of habitats. This hypothesis predicts variation in brown bear diet compositions among populations and a wide fundamental macronutrient niche. This type of nutrient balancing has been observed in gannets (Morus spp.; Tait, Raubenheimer, Stockin, Merriman, & Machovsky‐Capuska, 2014) and wild boar (Senior et al., 2016).The Seasonal Variation Hypothesis predicts that the proportion of macronutrients in the diets of brown bear will vary seasonally (Coogan et al., 2014). It is well known that the protein content of bear diets declines over the active season in several ecosystems (López‐Alfaro, Coogan, Robbins, Fortin, & Nielsen, 2015). However, the multivariate relationship between seasonal macronutrient proportions has been less well established (but see Coogan et al., 2014 and Costello et al., 2016). This hypothesis is nonmutually exclusive with either of the above hypotheses.The Prehibernation Optimal Diet Hypothesis predicts that the macronutrient composition of brown bear diets will be closer to the self‐selected optimal ratio for mass gain of captive bears during the prehibernation hyperphagic season, because selective pressure during this period will be the highest (i.e., behavioral and physiological adaptation).
|
study
| 99.94 |
The Nutrient Balancing Hypothesis predicts that a wide diet serves to increase the range of food options that can be combined to achieve a balanced macronutrient intake. This hypothesis predicts that the macronutrient composition of bear diets from different ecological and geographic populations is similar despite consuming different foods (i.e., a narrow fundamental macronutrient niche). This type of nutrient balancing has been observed in primates (Raubenheimer, Machovsky‐Capuska, Chapman, & Rothman, 2015; Rothman et al., 2007), badger (Meles meles; Remonti, Balestrieri, & Prigioni, 2011), and pine marten (Martes martes; Remonti et al., 2016).
|
study
| 99.94 |
The Nutritional Generalism Hypothesis postulates that a broad diet combined with the ability to tolerate a wide range of dietary macronutrient intakes enables a species to occupy a diverse range of habitats. This hypothesis predicts variation in brown bear diet compositions among populations and a wide fundamental macronutrient niche. This type of nutrient balancing has been observed in gannets (Morus spp.; Tait, Raubenheimer, Stockin, Merriman, & Machovsky‐Capuska, 2014) and wild boar (Senior et al., 2016).
|
study
| 99.94 |
The Seasonal Variation Hypothesis predicts that the proportion of macronutrients in the diets of brown bear will vary seasonally (Coogan et al., 2014). It is well known that the protein content of bear diets declines over the active season in several ecosystems (López‐Alfaro, Coogan, Robbins, Fortin, & Nielsen, 2015). However, the multivariate relationship between seasonal macronutrient proportions has been less well established (but see Coogan et al., 2014 and Costello et al., 2016). This hypothesis is nonmutually exclusive with either of the above hypotheses.
|
study
| 99.94 |
The Prehibernation Optimal Diet Hypothesis predicts that the macronutrient composition of brown bear diets will be closer to the self‐selected optimal ratio for mass gain of captive bears during the prehibernation hyperphagic season, because selective pressure during this period will be the highest (i.e., behavioral and physiological adaptation).
|
study
| 95.5 |
Furthermore, in addition to hypotheses specific to omnivory, we test the hypothesis that bear diets documenting anthropogenic food “subsidies” (e.g., livestock, agricultural crops, and supplemental feeds) are higher in nonprotein macronutrients than populations with natural diets (Coogan & Raubenheimer, 2016).
|
study
| 99.7 |
We started with studies collected from a global review of brown bear diets (Bojarska & Selva, 2012). We required estimates of the dietary proportion of mass of food consumed to calculate macronutrient compositions; thus, we included studies where foods were originally reported as the proportion of dry mass of diet (digestible dry matter; %DDM) or were estimable after applying fecal correction factors (CF) to percent fecal volume (%Vol) estimates. We updated our search to find studies published between 2012 and 2017 using Google Scholar and the search term brown bear diet, as well as searching within literature citing the aforementioned review. Other articles were obtained via ResearchGate (www.researchgate.net). We excluded studies where: (1) food categories were considered too broad to reasonably estimate the nutritional composition of the diet; (2) did not provide %Vol or %DDM estimates of diet; (3) did not cover the brown bear active season; and (4) there were imbalances in sampling that resulted in overestimating season‐specific food resources. All studies exceeded the lowest scat sample size (n = 95) cited in Bojarska and Selva (2012). This left us with a total of 18 papers providing data for 19 “populations” (Table 1). Populations were considered independent if samples were taken from different studies, countries, geographic regions, habitats, or years, following Senior et al. (2016).
|
review
| 99.9 |
The use of CFs is considered among the most suitable methods for brown bear diet assessment (Bojarska & Selva, 2012); however, they can result in variable outcomes depending upon their application, particularly for ungulates (López‐Alfaro et al., 2015; Persson, Wikan, Swenson, & Mysterud, 2001). Thus, where possible, we used %Vol estimates given in cited papers and applied our chosen CFs to re‐estimate the %DDM of foods in diets. We chose this approach to standardize the CFs used thereby minimizing diet variation across studies due to their differential application. For papers that gave seasonal estimates, we estimated %DDM for each season and from this we estimated the annual diet. We considered each seasonal diet to be representative for that time period, as opposed to weighting by sample size, to avoid biasing annual diets toward seasonal food items where sample size was not evenly distributed. For analysis of seasonal diets, we classified seasonal diet estimates as being one of four categories: spring; summer; autumn; or winter. We used the CFs presented in Hewitt and Robbins (1996), as applied by Fortin et al. (2013) and López‐Alfaro et al. (2015) to different food categories. We also applied CFs where available to specific food items within soft mast, hard mast, insects, and small mammal categories following Hewitt and Robbins (1996) and Bojarska and Selva (2012). In brief, CFs are applied by multiplying %Vol estimates of food items by their respective CF (i.e., %Vol * CF). These values are then summed across food items, and the %Vol * CF for each food item is expressed as a percentage of this sum to yield %DDM estimates (see Table 2 in Hewitt & Robbins, 1996). CFs are given in Table S1.
|
study
| 100.0 |
After estimating %DDM of food items in diets, we estimated the macronutrient composition of each food or food group using data collected from the literature and the USDA National Nutrient Database (US Department of Agriculture 2015; Table S1). Graminoids and forbs were condensed into one food category each. For other food categories, we obtained species‐specific food estimates or proxies, where possible. For animal prey, we used estimates of whole carcasses, because estimates of only muscle tissue likely overestimate protein and underestimate lipid content (Coogan et al., 2014), and brown bears tend to eat entire carcasses (Hilderbrand, Jenkins, Schwartz, Hanley, & Robbins, 1999). When possible we used total dietary fiber (TDF) estimates of indigestible carbohydrates to avoid differences in available carbohydrate estimates by subtraction and to more closely match the digestibility of bears (Pritchard & Robbins, 1990). Macronutrients were converted to percent metabolizable energy (Coogan et al., 2014, 2017) using conversion factors of 17 kJ/g for protein and carbohydrate and 37 kJ/g for lipid (Merrill & Watt, 1973). The proportions of macronutrient energy in foods were weighted by %DDM estimates to estimate their overall proportions in diets.
|
study
| 100.0 |
We were unable to obtain nutritional composition estimates of reported foods that were spatially and temporally contemporary with bear fecal samples in the published studies, which may induce error in macronutrient estimates of certain foods. This approach, however, is unlikely to significantly affect comparisons of macronutrient proportions between populations (Remonti et al., 2016; Senior et al., 2016).
|
study
| 100.0 |
We used graphical devices and associated theory from nutritional geometry to inform our analysis of macronutrient proportions in bear diets. Nutritional geometry is a multivariate graphical approach to examining nutrition based on state–space models and has been applied to a variety of species in both laboratory and free‐ranging settings (Raubenheimer et al., 2015; Simpson & Raubenheimer, 2012). Because our data set was compositional (i.e., a vector of non‐negative components which sum to a constant) and consisted of three components, we plotted bear diets within a simplex using mixture triangles (Raubenheimer, 2011). Specifically, we used conventional ternary diagrams, or equilateral mixture triangles (EMT), to visualize and interpret data. We provide right‐angled mixture triangle (RMT) plots for comparison (Figures S1 and S2). For information on the use of mixture triangles in the context of nutritional ecology, we refer the reader to Raubenheimer (2011).
|
study
| 99.94 |
We used compositional data analysis to analyze the proportions of macronutrients in bear diets. Compositional data analysis is a field of statistics that was developed to address concerns regarding using conventional statistics to analyze compositional data (Aitchison, 1982) and has been used across a variety of fields, including geosciences (Buccianti, Nisi, Martín‐Ferández, & Palarea‐Albaladejo, 2014), public health (Chastin, Palarea‐Albaladejo, Dontje, & Skelton, 2015), and meat science (Ros‐Freixedes & Estany, 2013). A full review of compositional data analysis is beyond the scope of this study; hence, we refer readers to the papers cited herein.
|
review
| 99.9 |
We used the R (v.3.4.1; R Core Team 2017) package {compositions} (v.1.40‐1; van den Boogaart, Tolosana, & Bren, 2014) for our compositional analysis in acomp geometry (van den Boogaart & Tolosana‐Delgado, 2008). We first examined annual diets, where we reported compositional descriptive statistics and variance for annual diets as the closed geometric mean and variance matrix of our centered log‐ratio (clr) transformed data set. The compositional geometric mean better represents the center of compositional data points than the arithmetic mean, and dispersion of compositional data is summarized using a variance matrix of pairwise log‐ratios (Aitchison, 2003). Conventional univariate measures of dispersion (e.g., SD of the arithmetic mean proportion of protein) are not considered informative for multivariate compositional data. For comparison, however, we report both conventional arithmetic and geometric means. We plotted predicted 2‐sigma and 3‐sigma region ellipsoids around the geometric mean.
|
study
| 100.0 |
We used a principal component analysis in acomp geometry (PCA.acomp) to examine variance in the proportions of macronutrients in annual population diets (Aitchison, 1983; Aitchison & Greenacre, 2002; Pawlowsky‐Glahn & Egozcue, 2001). PCA.acomp axes were plotted both within an EMT as curvilinear axes and using a biplot. In PCA.acomp biplots, the length of the link (i.e., distance between arrowheads) along a component relates to the SD of the log‐ratio of two components. Thus, the distance between links is used to evaluate relative variation between components.
|
study
| 100.0 |
To examine differences between seasons, we used linear models (LM) to examine changes in the proportion of macronutrients in bear diets using an isometric log‐ratio (ilr) data transformation following Tolosana‐Delgado and van den Boogaart (2011). The ilr transformation adjusts for changes in the proportion of one macronutrient consumed with the proportion of others consumed and allows for the use of conventional statistics on the transformed data, which is then back transformed into the original units for interpretation. Season was set as an ordered 3‐level categorical variable (spring, summer, and autumn), with spring set as the intercept category. Winter (n = 3) observations were dropped from the statistical analysis. Differences in seasonal diets were evaluated statistically in the global LM using an ANOVA. Model residuals were assessed for normality and homoscedasticity. Differences between individual seasons were assessed graphically by plotting geometric means and both 90% and 99% confidence regions within an EMT. For comparison with our compositional model, we created three separate univariate LMs of the effect of season (as an ordered factor) on the logit‐transformed (Warton & Hui, 2011) decimal proportion of each macronutrient. We followed the same ilr approach to examine differences in the annual and seasonal diets of populations receiving anthropogenic “subsidies” (e.g., agricultural crops, livestock, and supplemental feeding) versus natural diets (set as a binary explanatory variable).
|
study
| 100.0 |
To examine diets in relation to the behavioral preferences of captive bears, we plotted the mean proportion of protein (17%) as an isoportion line (“intake target” sensu Simpson & Raubenheimer, 2012) within EMTs. The preferred optimal ratio of macronutrients is likely to vary between bears (Erlenbach et al., 2014) and perhaps populations (Shafer et al., 2014); thus, we also plotted the associated ±4% SD isoportion lines around the mean protein intake to represent variance. We note as caveats that the preferred mean protein intake of captive bears was determined using conventional statistics which might differ from the compositional mean. Likewise, as mentioned previously, conventional SD estimates are not consistent with a compositional data analysis paradigm. Nonetheless, given the difficulty in determining macronutrient intake targets and related functional outcomes of free‐ranging animals (Machovsky‐Capuska, Coogan, Simpson, & Raubenheimer, 2016), adopting the optimal diet reference point of captive bears serves as a useful heuristic.
|
study
| 100.0 |
Across annual diets, the closed geometric mean proportion of macronutrient energy was 31.4% protein, 34.7% carbohydrate, and 33.9% lipid, which lies near to the mixture triangle's barycentre (Figure 1). The arithmetic mean (±SD) proportions of macronutrients were as follows: 31.0% (±10.7) protein; 36.1% (±14.9) carbohydrate; and 32.8% (±9.5) lipid. Thus, differences between the compositional and arithmetic means were relatively small, being 1.4% for carbohydrate, 1.1% for lipid, and 0.4% for protein.
|
study
| 100.0 |
Equilateral mixture triangle (EMT) depicting the proportions of macronutrients (protein = P, carbohydrate = C, and lipid = L) in annual bear diets (black dots). The geometric mean is shown by a red triangle. Ellipsoids predicting 2‐sigma and 3‐sigma regions are given in blue and red, respectively. Curvilinear principal component axes in acomp geometry (PCA.acomp) are shown with green lines. Component 1 (“horizontal” curve) explained 80.2% of the variance, and component 2 (“vertical” curve) explained 19.2% of variance. Corners represent 100% composition of the labeled macronutrient
|
study
| 99.94 |
Variability in the proportion of macronutrients in population diets is summarized in the variance matrix containing the pairwise log‐ratio variances (Table 2). Values close to zero indicate that the macronutrients in the ratio are highly proportional/codependent (i.e., relatively more constant). Protein and lipid in bear diets have log‐ratio variance closest to zero, implying that there is a higher proportional relationship between the consumption of the two macronutrients. Conversely, the highest log‐ratio variances occur with carbohydrate, which indicates that carbohydrate in bear diets is the least codependent on the other macronutrients. Following the 68‐95‐99.7 rule, ca. 95% and 99% of values are predicted to lie within 2 and 3 standard deviations of the mean; thus, the 2‐sigma and 3‐sigma regions in Figure 1 can serve as an estimate of the fundamental macronutrient niche of brown bears.
|
study
| 100.0 |
The first component of the PCA.acomp, associated with differences in ratios of carbohydrate with both lipid and protein, explained 80.8% of variance in the macronutrient proportions of bear diets (Figures 1 and 2). The remaining second component, associated with lipid and protein ratios, explained the remaining 19.2% of variance. In the PCA.acomp biplot (Figure 2), the link distance between protein and carbohydrate was greatest, indicating that most relative variation occurs between these two macronutrients. Carbohydrate and lipid also share a large amount of relative variation. The shorter link between protein and lipid indicates that their component ratio was relatively more constant.
|
study
| 100.0 |
Principal component analysis in Aitchison geometry (PCA.acomp) biplot of the proportion of macronutrients (protein = P, carbohydrate = C, and lipid = L) in annual brown bear diets. Numbers correspond to populations in Table 1. The relevant variables in the PCA.acomp biplot are the “links” (i.e., difference between two arrowheads) which represents the SD of log‐ratios between two components. Thus, the greatest relative variation occurred among protein and carbohydrate ratios, while the ratios of protein and lipid were relatively more proportional
|
study
| 100.0 |
Annual diets with anthropogenic subsidies were significantly different to those with natural foods based on an arbitrary α = 0.05 (ANOVA p < .001; Figure 3). The geometric mean percentage of macronutrients in diets with anthropogenic subsidies was 24.2% protein, 40.6% carbohydrate, and 35.2% lipid. For natural diets, mean proportions were 40.5% protein, 28.1% carbohydrate, and 31.4% lipid. The 99% confidence region of the mean anthropogenic diet included the isoportion line representing the preferences of captive bears, suggesting that such an annual diet is possible for bears consuming anthropogenic subsidies. The closer alignment of geometric means along the 1:1 isoproportion line for lipid and carbohydrate (radiating from the 100% protein corner) indicates a stronger decrease in protein relative to the other macronutrients. The separation of means and confidence regions between the 1:1 isoproportion line for protein and lipid (radiating from the 100% carbohydrate corner) shows that the ratio of lipid relative to protein is higher in anthropogenically subsidized populations. The shape of the confidence regions around means shows that there is more variation in carbohydrate in both groups.
|
study
| 100.0 |
EMT showing differences between the annual diets of bear populations receiving anthropogenic subsidies versus natural diets. Means (filled symbols) are shown with 90% and 99% confidence regions. The blue line represents the preferred optimal proportion of protein (17% ± 4) selected by captive bears. Isoproportion lines represent 1:1 proportions of protein and lipid (radiating from the 100% carbohydrate corner) and carbohydrate and lipid (radiating from the 100% protein corner)
|
study
| 99.9 |
The proportions of macronutrients in all bear diets varied significantly between seasons (ANOVA p < .001). Comparing spring to autumn, the geometric mean proportion of macronutrients in diets declined in protein (20.4%) and lipid (4.8%) and increased in carbohydrate (26.2%) (Table 3). Both spring and summer show overlapping confidence regions around mean compositions, suggesting they are not different (Figure 4). Autumn diets, however, lie distinctly in a higher carbohydrate, lower protein region of the simplex. Conventional univariate models were in agreement with compositional analysis: Protein showed a statistically significant linear decrease, and carbohydrate showed a significant increase, from spring to autumn (Table S4). Within‐season variability (i.e., interpopulation) in the proportion of macronutrients in population diets is summarized in the variance matrix in Table 3. For all seasons, protein and lipid had the highest codependence, with carbohydrate being the least codependent on the other macronutrients.
|
study
| 100.0 |
(A) Geometric mean decimal proportion of protein (P), carbohydrate (C), and lipid (L) in seasonal bear diets across all populations and partitioned into those receiving anthropogenic subsidies versus natural diets. (B) Matrix containing the geometric mean pairwise ratios of macronutrients in seasonal diets. (C) Variance matrix of log‐ratios among macronutrients in seasonal diets
|
study
| 99.94 |
EMT of the proportions of macronutrients (protein = P, carbohydrate = C, and lipid = L) in seasonal brown bear diets. The geometric mean for each season is shown by a filled symbol surrounded by 90% and 99% confidence regions. For reference, the blue line represents the preferred optimal proportion of protein (17% ± 4) selected by captive bears. A black isoproportion line represents 1:1 proportions of protein and lipid
|
study
| 99.94 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.