text
string
predicted_class
string
confidence
float16
Reference decadic attenuance values determined for each lot of NISTmAb are given in Table S2 in the ESM. The combined standard uncertainty associated with the Dcorr measurement includes multiple contributions as described in the ESM [14, 15]. Metrological traceability is to the decadic logarithm of the derived unit of regular spectral transmittance through the NIST Transfer Spectrophotometer (TS), which is qualified against the HAS II National Spectrophotometer via control standard SRM 2031 at 280 nm. The Dcorr reference values reflect measurements conducted using the specific quartz cuvette (b = 0.05092 cm). In an effort to report a more universal reference value amenable to direct comparison with other spectrophotometric determinations (e.g. alternative path length and/or dilutions), Dcorr was further used along with Eq. 1 to calculate reference mass concentration values reported in Table S3 (see ESM).
study
100.0
The reference values listed in Table S3 (see ESM) and Fig. 2 are based specifically on measured decadic total attenuance at 280 nm assuming a theoretical extinction coefficient (ε) of 1.42 (mL mg−1 cm−1) . The extinction coefficient was calculated according to the method reported by Pace et al. , and further corrected for glycan mass fraction via a correction factor of 0.977 . Uncertainty associated with the theoretical extinction coefficient has not been fully evaluated; therefore the reported value is not traceable to the SI unit of mass. A comprehensive uncertainty budget is reported in the ESM such that a future experimental determination of ε may be incorporated if deemed necessary via stakeholder feedback. The combined standard uncertainty provided in Table S3 (see ESM) is intended to represent, at the level of one standard deviation, the effect of the combined components of uncertainty including Type A measurement uncertainty and Type B components related to the analysis, consistent with the International Standards Organization/Joint Committee for Guides in Metrology (ISO/JCGM) Guide . Each of the three 14HB-D lots were measured to have a concentration to within ±2uc of one another (Fig. 2) which demonstrates that good inter-lot reproducibility was achieved. These data indicate the lifecycle management plan using bulk homogenization produced highly reproducible lots, and therefore future lots of RM 8671 are expected to be consistent with respect to concentration.Fig. 2Mean concentration determined for each lot of NISTmAb using UV-Vis spectrophotometry. Error bars represent ±2uc based on statistical treatment of data as described in ESM (n = 10 for RM 8671 lots, n = 2 for PS 8670)
study
100.0
NISTmAb charge heterogeneity was evaluated by the qualified CZE assay , wherein mAb charge variants are separated according to differential electrophoretic mobility in free solution within a uniform electric field applied across a buffer-filled fused silica capillary. The results of the CZE analysis of RM 8671 were consistent in all salient features to those observed previously with PS 8670, having three charge groups: the main group, which comprises the majority of the sample; the basic variants, which migrate toward the cathode more rapidly than the main group; and the acidic variants, which migrate toward the cathode less rapidly than the main charge group . The charge purity of the NISTmAb is given as the relative abundance of the main charge group with respect to all detected charge species. The results for RM 8671 compared to PS 8670 method qualification results are presented in Table S6 (see ESM) and Fig. 3.Fig. 3Homogeneity of RM 8671 lots by charge purity measured by CZE. Error bars represent ±3uc based on ANOVA analysis for each lot (n = 3 for RM 8671 lots) or from intermediate precision data for PS 8670 Qualification
study
100.0
The inter-vial homogeneity of the three lots of RM 8671 was assessed from the CZE data. The CV for main peak purity is less than 0.2% in each of the lots as calculated from the results listed in ESM Table S6. Standard deviation calculated for intra-vial variation compared to inter-vial variation on the raw data for main peak purity was also nearly identical (e.g. intra-vial SD = 0.109 vs. inter-vial SD = 0.131 for lot 14HB-D-001 main peak purity). Collectively this indicates inter-vial homogeneity with respect to charge purity was achieved.
study
100.0
The three lots of RM 8671 were found to conform to previously observed results for PS 8670 for all charge parameters whose values (ESM Table S6) agreed to within ±3uc, as shown in Fig. 3 for the main peak purity. The one-way two-tailed ANOVA, however, did reveal a minor difference from PS 8670. RM 8671 lots were found to contain slightly increased basic variants (≈1%) relative to PS 8670. This difference may be attributable to expected minor variations in the cellular expression/production of batches of material used for PS 8670 versus RM 8671. The three lots originating from the homogenized bulk (RM 8671), however, were shown to be consistent with one another in terms of main peak purity, basic variant relative abundance, and acidic peak relative abundance. Figure 3 demonstrates that good inter-lot reproducibility was achieved for each of the three 14HB-D lots, and that each material was observed to have main peak purity to within ±3uc of one another. Despite a small decrease in apparent charge purity versus PS 8670, it is the consistency of the commercial (RM 8671) lots that are important in ensuring a consistent product to stakeholders, which the current data indicate will be true considering the statistical equivalence of all three commercial lots. The current data suggest that the bulk homogenization method used to produce the RM 8671 lots was successful in homogenizing any minor process related variability (e.g. C-terminal lysine occupancy) and produced highly consistent, reproducible lots, thereby ensuring long term consistency in product quality attributes.
study
100.0
The size heterogeneity and monomeric purity of NISTmAb RM 8671 were analyzed under non-denaturing conditions by SEC with UV detection according to the qualified protocol . The resultant chromatograms for all three RM 8671 lots were consistent in all salient features to that observed previously with PS 8670. The parameters considered were monomeric purity (main peak relative area), high molecular weight (HMW) relative area (RA), and low molecular weight (LMW) RA. Figure 4 and Table S7 (see ESM) present results for PS 8670 during method qualification compared to the results during value assignment for RM 8671. The CV for monomeric purity is approximately 0.1% in each of the individual lots as calculated from the results listed in ESM Table S7. Standard deviation calculated for intra-vial replicates compared to inter-vial variation on the raw data for main peak purity was also nearly identical (e.g. intra-vial SD = 0.049 vs. inter-vial SD = 0.049 for lot 14HB-D-001 monomeric purity). Collectively this indicates inter-vial homogeneity with respect to monomeric purity was achieved for each individual lot.Fig. 4Homogeneity of RM 8671 lots by monomeric purity (top panel) and high molecular weight relative area (bottom panel), measured by SEC. Error bars represent ±3uc based on ANOVA analysis for each lot (n = 3 for RM 8671 lots) or from intermediate precision data for PS 8670 Qualification
study
100.0
Homogeneity of RM 8671 lots by monomeric purity (top panel) and high molecular weight relative area (bottom panel), measured by SEC. Error bars represent ±3uc based on ANOVA analysis for each lot (n = 3 for RM 8671 lots) or from intermediate precision data for PS 8670 Qualification
study
99.9
It is clear from Fig. 4 that RM 8671 does not conform to the values determined for PS 8670. The relative area of the monomer has decreased from ≈98.7% in PS 8670 to ≈96.7% in RM 8671 and the relative area of the high molecular weight species has increased from ≈1% to ≈3% (ESM Table S7). Despite a small increase in the % HMW versus PS 8670, the three lots of RM 8671 were shown to be consistent with one another in terms of monomeric purity, high molecular weight relative area and low molecular weight relative area. Figure 4 demonstrates that good inter-lot reproducibility was achieved for each of the three 14HB-D lots, and that each material was observed to contain % monomer, % HMW, and % LMW to within ±3uc of one another. Alterations in the homogenization and vial filling process and/or an increased presence of HMW species in one of the constituent batches, versus the PS 8670 material, are thought to have resulted in the increased HMW species. Despite a small increase in the % HMW versus PS 8670, it is the inter-lot homogeneity of the commercial (RM 8671) lots that is important in ensuring a consistent product to stakeholders, which the current data indicate will be true considering the statistical equivalence of all three commercial lots with respect to aggregates based on SEC.
study
100.0
CE-SDS is the micro electrophoretic analogue of traditional slab-gel size-based separations (e.g. SDS-PAGE). Analytes are complexed with SDS and injected into a narrow-bore glass capillary filled with an uncrosslinked polymer sieving matrix. A high voltage applied across the capillary drives the anionic SDS complexes toward the detection window; their migration is retarded in a size-dependent manner by differential interaction with the sieving matrix. NISTmAb RM 8671 monomeric purity was measured by CE-SDS under non-reducing conditions (nrCE-SDS) according to the qualified method protocol . Glycan occupancies of the heavy chain and relative abundance of non-reducible species were measured by CE-SDS under reducing conditions (rCE-SDS) according to the qualified method protocol . The electropherograms of the CE-SDS analyses of RM 8671 were consistent in all salient features to that observed previously with PS 8670 . The results for RM 8671 compared to PS 8670 method qualification results are presented in Table S8 (see ESM) and Fig. 5. In addition to the calculated physicochemical reference values listed in ESM Table S8, it was determined that method performance (migration times and relative abundance) of individual, resolved components (monomer, light chain, heavy chain, etc.) may be useful to the stakeholder in comparing performance of orthogonal assays likely to be developed with the NISTmAb. Therefore additional Tables S9, S10, S11, and S12 are included in the ESM for PS 8670 (from the qualification exercise) and RM 8671 lots 14HB-D-001, 14HB-D-002, and 14HB-D-003 respectively.Fig. 5Homogeneity of RM 8671 lots by monomeric purity, measured by nrCE-SDS. The measured range for PS 8670 during qualification is included for comparison. Error bars represent ±3uc based on ANOVA analysis for each lot (n = 3 for RM 8671 lots) or from intermediate precision data for PS 8670 Qualification
study
100.0
Homogeneity of RM 8671 lots by monomeric purity, measured by nrCE-SDS. The measured range for PS 8670 during qualification is included for comparison. Error bars represent ±3uc based on ANOVA analysis for each lot (n = 3 for RM 8671 lots) or from intermediate precision data for PS 8670 Qualification
study
99.9
Inter-vial homogeneity of the three lots of RM 8671 was also evaluated for each lot using CE-SDS. The CV for monomeric purity ranges from 0.3% to 0.8% calculated from the results listed in ESM Table S8. Standard deviation calculated for intra-vial variation compared to inter-vial variation on the raw data for main peak purity was also similar (e.g. intra-vial SD = 0.053 vs. inter-vial SD = 0.137 for lot 14HB-D-001 monomeric purity). An F-test for equivalence of variance confirmed the standard uncertainties were statistically similar. Collectively these indicate inter-vial homogeneity with respect to monomeric purity was achieved.
study
100.0
The three lots of RM 8671 were found to conform to previously observed results for PS 8670 for all size parameters measured to within the ±3uc control range listed in ESM Table S8. The one-way two-tailed ANOVA, however, did reveal a minor (nearly negligible, <0.03%) difference from PS 8670 in the glycan occupancy. This difference may be attributable to expected minor variations in the cellular expression/production of batches of material used for PS 8670 versus RM 8671.
study
100.0
All three lots of RM 8671 were shown to be consistent with one another in terms of monomeric purity, thioether content, and glycan occupancy demonstrating good inter-lot homogeneity for these properties (ESM Table S8). Figure 5, for example, demonstrates that each material was observed to contain monomeric purity within ±3uc of one another. The current data suggest that the bulk homogenization method used to produce the RM 8671 lots was successful in producing highly consistent, reproducible lots, thereby ensuring long term consistency in product quality attributes.
study
100.0
Flow imaging can be used for the analysis of subvisible particles (2 μm to 100 μm in size) in suspension. As a sample stream passes through a flow cell positioned in the field of view of a microscopic system, bright-field images are captured in successive frames. The digital images of the particles present in the sample are stored in a database that can be retrieved and analyzed for count, size, transparency, and various other morphological parameters. Protein particle concentration is reported for one or more size bins as an equivalent circular diameter (ECD), defined as the diameter of a polystyrene microsphere with the same image area as the observed particle. The subvisible protein particle content of RM 8671 lots was evaluated according to the optimized protocol described previously . Figure 6 shows representative images obtained for various particles in the three lots.Fig. 6Representative images of proteinaceous particles and their sizes (in equivalent circular diameter) obtained from 14HB-D-001, 14HB-D-002, and 14HB-D-003 analyses
study
100.0
Table S13 (see ESM) shows the mean subvisible particle concentrations (ECD ≥ 2 μm) in the PS 8670 and the three lots of RM 8671. Each of the vials of PS 8670 was obtained from the same rack of material, such that vial to vial variability due to the fill process can be assumed to be negligible. The RM 8671 vials, on the other hand, were obtained from multiple racks across the fill sequence. The CV (as calculated from ESM Table S13) for the inter-vial variability of each of the RM 8671 lots is on the same order as observed for PS 8670, indicating little to no vial-to-vial heterogeneity with regard to protein particle content was introduced by the filling process. From Table S13 (see ESM), the variability appears to be high even between vials in a particular lot; however, no clear trend among the racks from which the samples were drawn was observed. Due to the fragile nature of protein particles and the sensitivity of particle production to small changes in vial storage or handling, higher variability in particle concentration is expected. It should be noted that even if the concentration of particles were dissimilar from one vial to another, the amount of protein present in the particles is very minute, on the order of tens to several hundreds of ng/mL, corresponding to <0.005% protein concentration in solution .
study
100.0
All three lots of RM 8671 were shown to be consistent with one another, as well as to the PS 8670 material with respect to total particle content (within ±3SD). This indicates that the fill finish sequence did not significantly introduce particulates in the size range ≥ 2 μm and that consistent production of future lots can be expected. Table S13 does, however, show a potential trend of increasing particle content with increasing lot number. Trending of additional lots of material may be useful in the future to discern if this is indeed a trend or within typical lot-to-lot variability.
study
99.94
To obtain a size distribution of particles ranging from 1 nm to 1 μm in size, dynamic light scattering (DLS) was used. Particles in a solution move randomly by Brownian motion and scatter light. By analyzing the fluctuations in the intensity of the scattered light as a function of time, the diffusion coefficient of the particles and consequently their size can be calculated. The hydrodynamic diameter of RM 8671 lots was evaluated according to the optimized protocol described previously . Table S14 (see ESM) shows the mean hydrodynamic diameters of the samples from the three lots, which ranged from 9.83 nm to 9.96 nm.
study
100.0
All three lots of RM 8671 were shown to be consistent with one another, as well as to the PS 8670 material with respect to the mean hydrodynamic radius (within ±3SD). Each of the three lots also showed similar particle profiles in the nanometer range supporting the assertion that the vial filling process for the RM 8671 material appears to be homogeneous across all racks.
study
99.94
The peptide mapping method examines RM 8671 primary structure by monitoring its trypsin digested peptides, which are resolved using reverse-phase ultrahigh pressure liquid chromatography instrumentation coupled to an ultraviolet wavelength detector and a high-resolution mass spectrometer with electrospray ionization source (together “LC-UV-MS/MS”). A chromatographic trace, or peptide map, results from the signal generated as peptides eluting from the LC column pass through the UV and MS detectors producing “peaks”. Differing amino acid sequences give each peptide unique chromatographic properties and the presence of online MS/MS detection provides confident assignment of primary structure based on mass and fragmentation consistent with the predicted amino acid sequence.
study
99.94
LC-UV-MS/MS peptide mapping was performed on all three lots of RM 8671 in parallel with PS 8670 according to the optimized protocol described previously . Alignment of the TIC and UV traces of the RM 8671 digest with the PS 8670 reference peptide map showed a high degree of similarity upon visual inspection (Fig. 7). No trace had a unique or missing peak as compared to the reference map.Fig. 7Alignment of PS 8670 chromatogram with RM 8671 lots 14-HB-D-001, 14-HB-D-002, and 14HB-D-003 chromatograms. Tryptic digests of PS 8670 and RM 8671 lots 14-HB-D-001, 14-HB-D-002, and14HB-D-003 (traces ordered top to bottom) were analyzed by LC-UV-MS and the similarity of the resulting (A) TIC and (B) UV chromatograms compared against the reference peptide map generated from the PS 8670 digest (top, black trace in panels a and b). The initial five minutes of the UV traces are not shown due to the large difference in scale between the relative levels of absorbance of peaks detected during the 0 min to 5 min period and the 5 min to 90 min period
study
100.0
Alignment of PS 8670 chromatogram with RM 8671 lots 14-HB-D-001, 14-HB-D-002, and 14HB-D-003 chromatograms. Tryptic digests of PS 8670 and RM 8671 lots 14-HB-D-001, 14-HB-D-002, and14HB-D-003 (traces ordered top to bottom) were analyzed by LC-UV-MS and the similarity of the resulting (A) TIC and (B) UV chromatograms compared against the reference peptide map generated from the PS 8670 digest (top, black trace in panels a and b). The initial five minutes of the UV traces are not shown due to the large difference in scale between the relative levels of absorbance of peaks detected during the 0 min to 5 min period and the 5 min to 90 min period
study
100.0
Mean TIC retention times were calculated across quadruplicate injections of the PS 8670 digest and data from three of the injections were used to calculate mean UV retention times. The difference between means of the PS 8670 reference map peak retention times and the corresponding peaks for the three lots of RM 8671 was <2% for all peaks in the TIC and UV chromatograms, indicating a high degree of similarity between PS 8670 and the RM 8671 lots.
study
100.0
To further confirm the identity of NISTmAb RM 8671, data from its tryptic digest were submitted for peptide identification. Calculation of the sequence coverage for each tryptic digest produced the same results as those described previously for PS 8670 and lot 14HB-D-001 . For all materials analyzed, sequence coverage of 96.89% was achieved for the heavy chain and 100% for the light chain. This included full coverage of the complementarity-determining regions (CDRs). All post translational modifications identified for the three RM 8671 lots were consistent with those previously reported for PS 8670 . The mass spectrometry results indicated nearly complete pyroglutamination of the N-terminus, moderate levels of the loss of C-terminal lysine, and low levels of glycation, oxidation, and deamidation as previously reported . The similarity of the TIC and UV peptide maps of PS 8670 and RM 8671 as well as the matching peptide identifications confirmed the identity of RM 8671 as conforming to PS 8670.
study
100.0
Each vial of RM 8671 contains 800 μL of a nominal 10 mg/mL IgG1κ monoclonal antibody (NISTmAb) in 12.5 mmol/L L-histidine, 12.5 mmol/L L-histidine -HCl- (pH 6.0) (Formulation Buffer). RM 8671 is packaged and supplied to the user in internal threaded polypropylene cryovials which have been double packaged in a cardboard box with an insert designed to securely fit the cryovials, and sealed in a laminate foil pouch to prevent CO2 ingress. The double packaged RM is shipped on dry ice and should remain frozen during shipment. The material should be stored in a frozen state at −80 °C immediately upon receipt and in all cases storage of the material at −80 °C is preferred. In reality, no fill volume/format will be amenable to all assays and end user purposes. Therefore, a series of stability samples (preparation described in ESM) were evaluated to determine the most extreme storage conditions (referred to here as alternative storage conditions) under which the sample yielded a measurement result within ±3uc for a given method.
study
99.94
Storage stability evaluation was performed on RM 8671 lot 14HB-D-001 and extrapolated for all other RM 8671 lots. Samples reserved for thaw/freeze stability and accelerated stability analyses were reserved as indicated in the ESM (see Table S15 and Table S16). T/F samples were tested using the qualified physicochemical assays or optimized informational value assays (at a minimum) described above after undergoing an additional one to five T/F cycles with freeze temperature at −80 °C or −20 °C. Accelerated stability samples were analyzed using the qualified physicochemical assays or optimized informational value assays described above at the (0, 7, and 28) day time points described in the ESM (see Table S16). The one day time points were also analyzed with SEC to provide more fine-detail in excursions from the control limits at elevated temperatures. Control charts for alternative storage stability evaluations displaying measurement results for the various stability conditions, with control range taken as the mean value of the unstressed material (±2uc for UV and ±3uc for physicochemical assays), are depicted in Fig. 8. A larger control range of ±3uc was utilized for physicochemical assays compared to the concentration measurement because of the fewer degrees of freedom and increased measurement variability observed during value assignment. In addition, the UV concentration measurement is expected to be less sensitive to degradation, yet deviations in concentration may have significant impact on other physicochemical measurements.Fig. 8Physicochemical method results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze conditions using (a) UV-Vis, (b) CZE, (c) SEC, and (d) nrCE-SDS. The black lines indicate a control range taken as the mean value of unstressed 14HB-D-001 ± 2uc for UV-Vis and ±3uc for physicochemical assays
study
100.0
Physicochemical method results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze conditions using (a) UV-Vis, (b) CZE, (c) SEC, and (d) nrCE-SDS. The black lines indicate a control range taken as the mean value of unstressed 14HB-D-001 ± 2uc for UV-Vis and ±3uc for physicochemical assays
study
99.94
Control charts for alternative storage stability evaluations with respect to informational value properties are displayed in Fig. 9 and Fig. 10. Informational value measurement results utilized a stability control range taken as the mean value of the unstressed material ±3SD.Fig. 9Mean particle concentration (ECD ≥ 2 μm) results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze (right) conditions using flow imaging. The black line indicates the upper control limit taken as the mean value of unstressed 14HB-D-001 + 3SD and the error bars represent ±1SD for the specific sample. Note that the 40 °C 28 day sample (54,627 ml−1) is indicated with a straight line extending off of the graphFig. 10Mean hydrodynamic diameter results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze (right) conditions using DLS. The black lines indicate the control range taken as the mean value of unstressed 14HB-D-001 ± 3SD and the error bars represent ±1SD for the specific sample. Note that the 40 °C 28 day sample (143 nm) is indicated with a straight line extending off of the graph
study
100.0
Mean particle concentration (ECD ≥ 2 μm) results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze (right) conditions using flow imaging. The black line indicates the upper control limit taken as the mean value of unstressed 14HB-D-001 + 3SD and the error bars represent ±1SD for the specific sample. Note that the 40 °C 28 day sample (54,627 ml−1) is indicated with a straight line extending off of the graph
study
99.9
Mean hydrodynamic diameter results for 14HB-D-001 under accelerated stablity (left) and thaw/freeze (right) conditions using DLS. The black lines indicate the control range taken as the mean value of unstressed 14HB-D-001 ± 3SD and the error bars represent ±1SD for the specific sample. Note that the 40 °C 28 day sample (143 nm) is indicated with a straight line extending off of the graph
study
99.94
Recommended maximum storage T/F cycles and maximum storage time at a given temperature were set (for each method individually) as the most extreme data point that remained within the control range. NISTmAb RM 8671 was shown to be quite stable with respect to T/F according to UV, SEC, CE-SDS, and CZE. In all cases the material produced a measurement result within the control range for up to 5 T/F cycles. According to Fig. 9, however, the T/F conditions appeared to show particle concentrations near or outside of the defined control regions, therefore it is recommended that T/F not be performed for samples intended for flow imaging analysis.
study
100.0
Elevated storage temperature and time did affect the measurement results in some cases, as expected. Incubation at 40 °C for extended periods produced the most noticeable change, with an increase in protein particulate content, apparent hydrodynamic radius, and apparent concentration. SEC results at 40 °C indicated an initial decrease in HMW species, followed by an increase in LMW species beyond 7 days, resulting in the parabolic behavior of Fig. 8C under this condition. The kinetics were slowed under refrigerated conditions, however, SEC results did fall outside of the control range when stored at 4 °C for 28 days. Measurement results for all other assays were shown to be within the control limits during refrigerated storage for up to 28 days. In all cases, storage of the material at −80 °C is preferred. If aliquot preparation and/or storage at other than recommended conditions are necessary, the maximum alternative storage condition evaluated that yielded measurement results within the control range for each method is listed in ESM Table S17 for reference values and ESM Table S18 for informational values and peptide mapping.
study
100.0
A comprehensive lifecycle management program including a two-tiered in-house reference standard approach was modeled after industry best practices . In order to serve as a common framework for pre-competitive innovation, a Reference Material must be shown to be homogeneous, stable, and fit for its intended use. The current manuscript describes in detail the stratified sampling plan that was used in association with rigorously qualified analytical methods to assign reference values and ensure the quality of the NISTmAb RM 8671 [9, 11, 12]. Reference value assignment was performed in concert with the in-house primary sample PS 8670 as a system suitability control, the material for which the most historical analytical and biophysical attribute information was available [6–8]. The initial methods selected were intended to be those that were most pertinent to the stakeholder in ensuring a quality material, and thus included chemical and colloidal stability/purity indicating assays along with a comprehensive primary amino acid sequence confirmation.
study
99.94
The homogeneity assessment of each individual lot was made at the time value assignment analyses were performed using CZE, SEC, CE-SDS, FI, and DLS. There was no apparent trend in the data with respect to vial/rack position within a lot or when plotted against the sequence in which the samples were prepared for any of the methods. Furthermore, comparison of each of the initial three lots of RM 8671 demonstrated statistical equivalence in all of the attributes measured. Although each lot of RM 8671 is assigned its own individual reference values, the intra- and inter-lot homogeneity indicates a successful homogenization and vial filling process resulting in consistent content and quality. It is therefore expected that consistent measurement results will be attainable for RM 8671 for the foreseeable future provided the samples are handled upon receipt as directed.
study
100.0
Stability with respect to physicochemical attributes was evaluated based on extended thaw/freeze testing and storage at various temperatures. In all cases, except for particle generation with multiple T/F cycles, the ability of the material to withstand several T/F cycles and extended storage at 4 °C (as indicated in Table S17 and Table S18 in the ESM) provides users a variety of options to suit their individual testing needs. Optimal storage for extended periods is in a frozen state at −80 °C. Prior to analysis, the vial should be removed from the −80 °C freezer and thawed at room temperature for approximately 30 min or until no residual ice crystals remain. Once thawed, the vial should be gently inverted five times to alleviate any concentration gradients that may have formed during the freezing process. The vial should then be briefly spun in a mini-centrifuge to settle any solution that may otherwise remain adhered to the lid or internal threads of the vial. When handled according to the optimum storage conditions, RM 8671 is expected to result in physicochemical performance within the assigned control range (ESM Table S17) for each qualified physicochemical reference value method described previously in this series [11, 12].
study
100.0
RM 8671 has demonstrated stability and homogeneity with respect to the specified physicochemical properties using the qualified methods described herein, and is consequently suitable as a Reference Material. Significant effort was made to ensure lot-to-lot homogeneity and stringent method optimization to produce measurements with minimal sample preparation artifacts. Ultimately the values, from a metrological perspective, are method-specific and intended as a baseline for comparison of results from orthogonal, yet related assays that may be performed in a given user’s laboratory. This measurement approach, however, is indeed fit-for-purpose as the measurement technologies selected are representative of the current state-of-the-art in biopharmaceutical characterization. The availability of a Reference Material along with the well-defined analysis parameters reported throughout this publication series provides for the first time a foundation upon which longitudinal advances in biopharmaceutical characterization can be benchmarked. Future measurement technologies, regardless of their mechanistic principles, can be traced back to a consistent quality material of the same constituent product quality attributes.
review
64.6
Advances in technology are inevitable, and will ultimately be the future of higher-resolution product understanding. The pre-competitive nature of NISTmAb RM 8671 makes it a useful, open-innovation platform to demonstrate novel analytical and biophysical technologies and/or competencies within an organization. It affords a common platform for discussion, harmonization and advancement of analytical and biophysical technologies as may be achieved through inter-laboratory comparisons. The presence of a widely available industry quality material, analyzed by countless permutations of every analytical tool available, will produce the most expansive and dynamic single protein characterization dataset to date. Continued operation in the pre-competitive space offers an unprecedented opportunity to harness these big data for developing foundational tools such as complex inter-method data integration, modeling, and data visualization to guide next-generation biopharmaceutical development.
other
99.9
Enactment of the NISTmAb RM 8671 quality plan is described herein, demonstrating identity, quality, and stability of this material with respect to its material properties. The NISTmAb represents a first-of-its-kind Reference Material intended for use in evaluating measurement technologies to inform on specific attributes during therapeutic protein characterization. With this unique role in mind, a novel lifecycle management plan was developed including a two-tier in-house reference standard system, industry-relevant concentration measurement, qualified physicochemical assays, and relevant informational measurements. RM 8671 was verified to be homogeneous both within and between vialing lots, demonstrating the robustness of the lifecycle management plan. It was analyzed in concert with the in-house primary sample PS 8670 to provide a historical link to this seminal material. RM 8671 was verified to be fit for its intended purpose as a technology innovation tool, external system suitability control, and cross-industry harmonization platform. The pre-competitive nature and long term availability of RM 8671 should provide a platform for monitoring the evolution of therapeutic protein characterization.
study
99.94
Dengue is considered one of the most important diseases transmitted by mosquitos, its incidence has increased at an alarming rate and it has become a public health problem over the last fifty years (Bhatt et al. 2013). Among the causes of this increase are social and demographic changes such as population growth and urbanisation together with the lack of programs for surveillance, prevention, and vector control. Dengue is transmitted by the bite of a mosquito (Aedes aegypti or Aedes albopictus) infected with one of the four dengue virus (DENV) serotypes, which are genetically and antigenically related viruses. DENV is a member of the family Flaviviridae, genus Flavivirus. The genome consists of a positive single-stranded RNA (ssRNA) of about 11 kb with a single open reading frame encoding one polyprotein that is processed by cellular and viral proteinases to form three structural and seven non-structural (NS) proteins (Chambers et al. 1990). DENV RNA is flanked by 5′ and 3′ untranslated regions (UTRs) that present high degrees of consensus sequences among the four serotypes. The 3′ UTR contains two secondary hairpin structures known as 3′-stem loops (3′-SLs) and a small hairpin that are important elements for viral replication/translation, and are the most conserved structures in Flavivirus RNAs. Cellular proteins such as the translation elongation factor-1 alpha (EF-1α) and the polypyrimidine tract-binding (PTB) protein are reported to interact with the 3′-SL, suggesting that both proteins may function as chaperones to maintain the RNA structure in a conformation that favours DENV replication (de Nova-Ocampo et al. 2002). Furthermore, Alvarez et al. (2008) described a sequence located upstream of the translation initiation codon in the 5′ UTR, complementary to a region present in the 3′-SL, designated as the 5′ and 3′ upstream AUG region (5’-3’UAR), which is also essential for RNA circularisation and thus for viral RNA synthesis by the viral RNA-dependent RNA polymerase (RdRp).
review
57.28
Since the discovery of miRNAs, a class of endogenous non-coding cellular RNAs of 18-25 nucleotides that binds to a complementary sequence in the 3′ UTR of target mRNAs to regulate target gene expression mostly at the post-transcriptional level, their potential application for the treatment of infectious diseases is being investigated (Ghosh et al. 2009). In this context, some miRNAs are involved in the host antiviral response, acting as factors with antiviral activity. For example, the miRNA let-7b inhibits Hepatitis C Virus (HCV) RNA expression and replication by targeting the host factor insulin-like growth factor 2 mRNA-binding protein 1 (IGF2BP1) (Cheng et al. 2013). The 3′ end of the Human immunodeficiency virus type 1 (HIV-1) mRNA is also reported to be a target for a group of miRNAs expressed in resting CD4+ T cells (miR-28, -125b, -150, -223, and -382) that may inhibit translation and cause a latent infection in these cells (Huang et al. 2007). Recently, six single miRNAs targeting the highly conserved regions of the DENV-2 genome were shown to efficiently inhibit virus replication (Xie et al. 2013). However, DENV infection significantly induced the expression of miR-146a, thereby facilitating viral replication by targeting tumour necrosis factor receptor-associated factor 6 (TRAF6) and diminishing interferon-β (IFN-β) production (Wu et al. 2013). Additionally, incorporation of miR-122-MRE (miRNA recognition element) confers an inhibitory susceptibility to miR-122 targeting the DENV replicon, suggesting that DENV can be engineered to exert the desired replication restriction effect for avoiding infection of vital tissues/organs (Lee et al. 2010). Recently it was reported that DENV infection up-regulates the expression of miR-30e* but overexpression of this miRNA simultaneously suppresses DENV replication by promoting IFN-β production (Zhu et al. 2014). Overexpression of miR-548g-3p suppresses the replication of all four DENV serotypes (Wen et al. 2015) and we have demonstrated that overexpression of synthetic miR-133a suppresses DENV-2 replication, possibly through regulating the expression of PTB, a miRNA-133a target, since its protein was found to be upregulated during DENV infection (Castillo et al. 2016). Furthermore, we found that the 3′ UTR of the RNA of all four DENV serotypes is targeted by miR-133a. However, members of the Flavivirus genus were recently reported to produce a subgenomic flavivirus RNA (sfRNA), a positive-sense non-coding RNA that accumulates to high levels in infected mammalian and insect cells (Lin et al. 2004). Interestingly, this sfRNA acts as an RNA interference (RNAi) suppressor in both insect and mammalian cells, inhibiting the RNase Dicer (Schnettler et al. 2012). This is why the discovery of cellular miRNAs recognising target sequences in the DENV genome would be vital, possibly leading to a treatment for dengue, for which there is no vaccine or therapeutic drug yet.
study
99.94
Although the biological functions of miR-744 and miR-484 are poorly understood, they appear to be highly conserved in vertebrates. The gene encoding miR-744 locates at chromosome 17p12 and is implicated in the cellular process leading to human disease development. The function of miR-744 depends on the cell type and it acts as either a tumour suppressor or a tumour promoter. The gene encoding miR-484 locates at chromosome 16p13 and is essential for cerebral cortex development. Both, miR-744 and miR-484 are mainly expressed in the brain, but they can also be found in the liver and blood. miR-484 was significantly differentially expressed in the serum of patients with early breast cancer versus healthy controls; however, no correlation could be established between miR-484 levels and the histopathological parameters of breast cancer (Zearo et al. 2014). miR-744 is suggested to play an inhibitory role in many cancers, including colon, breast, and gastric cancer. It was recently reported that miR-744 inhibits the growth, migration, invasion, proliferation, and metastasis of gastric cancer by targeting Bcl-2 protein expression (Chen & Liu 2016). Furthermore, a lower miR-744 expression level was reported in patients with hepatocellular carcinoma (HCC) and this was associated with HCC recurrence and prognosis (Tan et al. 2015). However, studies regarding the role, effect, and significance of miR-484 and miR-744 on viral infections are still insufficient. Therefore, the main objective of this study was to determine the role of miR-484 and miR-744 in DENV infection and to examine whether DENV infection alters the expression of both these miRNAs. While our results were obtained using Vero, a cell line highly susceptible to DENV infection, we have also observed that macrophages and dendritic cells, which are both targets of DENV infection, also express miR-484 and miR-744 (unpublished observations). Therefore, we explored the relationship between DENV and cellular miRNAs using bioinformatics tools. We then overexpressed miR-484 or miR-744 in Vero cells to test their role in DENV replication, and finally determined the effect of DENV infection and the effect of the 3′ UTR of DENV RNA on miR-484 and miR-744 expression in Vero cells. We observed that overexpression of either miR-484 or miR-744 diminishes the replication of all four DENV serotypes. Finally, infection and transfection of Vero cells with a plasmid construct encoding the 3′ UTR of DENV RNA resulted in downregulation of endogenous miR-484 and miR-744. This is one of the first studies to demonstrate the effect of two cellular miRNAs on regulating the replication of the 4 DENV serotypes. Our study thus helps us to better understand the relationship between host cells and DENV infection. Although additional studies are required to unambiguously confirm the function of miR-484 and miR-744, our preliminary data suggest that these miRNAs might play a role in DENV replication.
study
99.94
Bioinformatics predictions - For the computational analysis, we followed a strategy established in our laboratory (Castillo et al. 2016). Briefly, the sequences of DENV-1 to -4 were downloaded from GenBank, and the free algorithm MicroInspector (www.ncbi.nlm.nih.gov/pmc/articles/PMC1160125/) was used to scan for possible targets of human miRNAs collected in miRBase. Only those target sites common to the four reference sequences were selected. The RNAhybrid program (http://bibiserv.techfak.uni-bielefeld.de/rnahybrid/) was used to verify the findings of MicroInspector.
study
100.0
Cell lines - Mosquito C6/36 HT cells obtained from the ATCC were cultured as previously described (Castillo et al. 2016). Vero cells (CCL-81) were grown in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% heat-inactivated foetal bovine serum (FBS), 1% L-glutamine, 1% vitamins, 1% non-essential amino acids, and 1% Penicillin/Streptomycin (Sigma-Aldrich Chemical Co, St. Louis, MO, USA), at 37ºC with 5% CO2.
study
99.94
Viral stocks and titration - The reference strains of DENV-1, DENV-2 New Guinea C (NGC), DENV-3, and DENV-4 were provided by the Centers for Disease Control (CDC, CO, USA). Viral stocks were obtained by inoculating a monolayer of C6/36 HT cells in a 75-cm2 tissue culture flask with the virus at a multiplicity of infection (MOI) of 0.05 diluted in 1 mL of L-15 medium supplemented with 2% FBS. After 3 h of adsorption, 10 mL of L15 medium supplemented with 2% FBS was added and the cells were cultured for five days at 34ºC without CO2. The supernatant was then removed from the cells and centrifuged for 5 min at 400 x g to pellet the cell debris. The supernatant was aliquoted and stored at -70ºC for future use. Clinical isolates of DENV-1 (strain Bga-07), DENV-2 (strain 109-05), and DENV-4 (strain Bga-06) were obtained from patients with dengue haemorrhagic fever in Antioquia, Colombia (kindly provided by Dr FJ Díaz, Grupo Inmunovirología, Facultad de Medicina, Universidad de Antioquia) and used to clone the DENV 3′ UTR (Castillo et al. 2016). Virus titration was performed by flow cytometry, as previously described (Castillo et al. 2016). Briefly, C6/36 HT cells were seeded in 12-well plates and cultured overnight at 34ºC without CO2. They were then infected with 10-fold serial dilutions of the virus, and at 24 h post-infection (hpi) they were harvested and resuspended in PBS. For flow cytometry analyses, the cells were fixed using a Fixation/Permeabilisation buffer (eBioscience, San Diego, CA, USA), centrifuged and washed twice with PBS, and stained with the monoclonal antibody 4G2 (kindly provided by Dr P Desprès, Institut Pasteur, Paris) in a final volume of 100 mL. As the secondary antibody, fluorescein isothiocyanate (FITC)-labelled goat anti-mouse IgG antibody (Invitrogen, Life Technologies, CA, USA) was used. The cells were analysed with a FACScan flow cytometer using the FACSdiva software. The percentage of infected cells in each sample and the total number of cells seeded per well were used to calculate the final virus titre.
study
100.0
Generation of pGUD plasmid constructs - pGUD plasmids containing the 3′ UTRs of DENV-1, -2, or -4 downstream of green fluorescent protein (GFP) in the pEGFP-C1 (Clontech, CA, USA) construct were previously described (Castillo et al. 2016). Briefly, the 3′ UTRs of the three DENV serotypes were amplified by polymerase chain reaction (PCR) from viral RNA obtained from infected cell culture supernatants using specific primers (forward: 5′ GAATTCGTAGGTGCGGCTCATTGATTGGGCTAAC 3′ and reverse: 5′ GTCGACGAACCTGTTGATTCAACAGCACC 3′). A stop codon (indicated in bold in the 5′ end of the forward primer), and a restriction site for EcoRI and SalI (underlined) at the 5′ end of the forward and reverse primers respectively, were incorporated during amplification. The PCR products were purified and cloned into the pEGFP-C1 construct, using the EcoRI and SalI enzymes (Thermo Scientific, NH, USA). The constructs generated were designated pGUD1, pGUD2, and pGUD4 for the DENV-1, DENV-2, and DENV-4 3′ UTR respectively. We were unable to amplify the 3′ UTR of DENV-3 RNA from the available DENV-3 isolates.
study
100.0
Expression of miR-484 and miR-744 - Vero cells were seeded at a final concentration of 1.5 × 105 cells/well in 24-well plates and transiently transfected with pGUD1, pGUD2, or pGUD4, at a final concentration of 0.5 µg/well, using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. Alternatively, Vero cells were infected with each DENV serotype at an MOI of 3. In both cases, the cells were harvested at 8, 16, 24, 32, 48, and 72 hpi and total RNA extraction was performed using the Trizol reagent (Invitrogen) following the manufacturer’s instructions. The RNA concentration was measured using a NanoDrop spectrophotometer (Nano Drop Technologies, CA, USA). Reverse transcription of endogenous miR-484 and miR-744 in transfected and DENV-infected Vero cells was carried out with 10 ng RNA to produce cDNA using the TaqMan® MicroRNA Reverse Transcription Kit (Applied Biosystems, CA, USA). For miRNA reverse transcriptase quantitative-PCR (RT-qPCR), experiments were carried out using the Taqman microRNA Assay (Applied Biosystems) and TaqMan® Universal Master Mix II (Applied Biosystems), according to manufacturer’s instructions, and performed in triplicate on a Bio-Rad CFX96 real-time Detection System (Bio-Rad, CA, USA). The expression of both miRNAs was normalised to 18S rRNA, and the expression levels were determined via the comparative threshold cycle (Ct) method using 2-ΔΔCt. Three independent replicates were performed for each experiment.
study
100.0
Analysis of the effect of miR-484 and miR-744 expression on DENV replication - To evaluate the effect of miR-484 and miR-744 on DENV replication, Vero cells were seeded at a density of 2 × 105 cells/well in 12-well plates. On the following day, the cells were transiently transfected with pEZX-MR03-miR-484 (Homo sapiens miR-484 stem-loop expression clone, cat. HmiR0264-MR03) or pEZX-MR03-miR-744 (H. sapiens miR-744 stem-loop expression clone, cat. HmiR0509-MR03) both obtained from GeneCopoeia, Inc. (Rockville, Maryland, USA), at a final concentration of 0.5 µg/mL using Lipofectamine 2000 according to the manufacturer’s instructions. A scrambled miRNA or empty vector (Cat. CmiR0001-MR03, Rockville, Maryland, USA) was used as the negative control. These constructs contain the GFP reporter gene facilitating the determination of transfection efficiency and the miRNA overexpression level by fluorescence microscopy. At 24 hpt, the cells were challenged with each DENV serotype separately at an MOI of 3. Alternatively, the cells were first challenged with each DENV serotype separately and after 24 h, they were transfected with pEZX-MR03-miR-484 or pEZX-MR03-miR-744 as described above. After 3 h of virus adsorption and shaking every 30 min, the medium was removed, the cells were washed twice with PBS, and the medium was replaced with DMEM supplemented with 2% FBS. The effect of either miRNA on DENV replication was assessed at 72 hpi by flow cytometry and RT-qPCR, and by western blot in the case of DENV-2. All experiments were performed in triplicate.
study
100.0
Quantification of DENV infection by flow cytometry - At the indicated time points post-infection, Vero cells were harvested and analysed by flow cytometry as described above. The infected cells were expressed as the percentage of infected cells over the total number of cells analysed. All the experiments were performed in triplicate.
study
99.94
Quantification of DENV RNA copy number by RT-qPCR - Viral RNA was extracted from culture supernatants using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions. The viral copy number was determined by RT-qPCR using the following DENV-specific primers against the conserved sequences in the core gene of DENV RNA: forward: 5′ CAA TAT GCT GAA ACG CGA GAG AAA 3′, and reverse: 5′ CCC CAT CTA TTC AGA ATC CCT GCT 3′ (Castillo et al. 2016). Calculation of the genomic RNA copy number was performed based on a standard curve, as previously described (Sachs et al. 2011). All experiments were performed in triplicate.
study
100.0
Viral protein detection by western blotting - miRNA-transfected and DENV-2-infected cells were detached with trypsin and lysed using Lysis solution (Applied Biosystems). Total proteins were quantified using the BCA Protein Assay (Pierce, Thermo Scientific, NH, USA). Total protein (50 µg) was loaded onto an SDS-PAGE gel and then transferred to a nitrocellulose membrane after electrophoresis. A primary mouse monoclonal antibody against the DENV NS1 protein (Thermo Scientific) and a secondary anti-mouse IgG antibody conjugated with horseradish peroxidase (HRP; Santa Cruz Biotechnology, USA) were used for detection. Finally, the signals were detected using the chemiluminescence ECLTM detection system (Pierce). Band intensities were quantified by densitometry using the image processing software ImageJ 1.49, freely available online.
study
99.94
DENV 3′ UTR contains potential miR-484 and miR-744 binding sites - We were interested in determining whether any interaction occurs between DENV RNA and human host miRNAs, since it was previously reported that viral genomes can be targeted by human miRNAs. We therefore investigated whether the DENV 3′ UTR contains potential miRNA binding sites. The sequences of the 3′ UTRs plus 374 nucleotides of the coding region of NS5 of all four DENV serotypes were aligned. In total, 108 miRNAs for DENV-1, 80 for DENV-2, 94 for DENV-3, and 89 for DENV-4 were predicted to target the 3′ UTR using the MicroInspector software (data not shown). Since it is reported that the 3′ UTR sequence across all DENV serotypes is moderately conserved, and considering the hypothesis that a functional miRNA target site would be conserved across all DENV serotypes, only those target sites common to the four reference sequences were selected. Among the miRNAs that fulfilled these criteria, miR-484 and miR-744 were predicted in our analysis (Fig. 1A) using the RNAhybrid program. The StarMir program predicted the potential binding sites for miR-484 and miR-744 in the DENV 3′ UTR as the target sequences (Fig. 1B). Each of the candidate sites was assigned a logistic probability as a measure of confidence in the predicted site. Interestingly, the location of the target sequence was in a loop of the 3′ UTR known as the 3′-SL that contains the elements known as the 3′ cyclisation sequence (3’CS) and the region upstream of the AUG (3′UAR) (Fig. 1C). Secondary structure was predicted using the Mfold program (http://unafold.rna.albany.edu/?q=mfold). Using the RNAfold webserver we observed that this secondary structure is the natural-mode structure that is most stable for that region in the 3′ UTR of DENV RNA (the free energy of the thermodynamic ensemble was -140.64 kcal/mol) (http://unafold.rna.albany.edu/?q=mfold/rna-folding-form). We then wondered whether the miR-744 and miR-484 binding sites on the 3′ UTR of DENV RNA (Fig. 1C) are accessible in the presence of the SL and hairpin. Using the RNAup webserver (http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAup.cgi) we found that the total free energy of the miR-484-3′ UTR was -8.92 kcal/mol with an opening energy (accessibility) of 5.16 kcal/mol, whereas for the miR-744-3′ UTR the total energy was -7.91 kcal/mol with an opening energy of 8.13 kcal/mol, suggesting that miRNA-3′ UTR binding would proceed spontaneously.
study
100.0
Fig. 1: the 3′ UTR of Dengue virus (DENV) RNA contains target sequences for cellular miR-484 and miR-744. (A) Alignment of a fragment of the 3′ UTR of the four DENV serotypes. The miR-484 and miR-744 target sites common to the four serotypes are indicated; (B) the target sites for miR-484 and miR-744 fulfil the seed sequence (shaded nucleotides, first nucleotides from the 5′ end of the miRNA). Structures predicted using RNAhybrid; (C) location of the miR-484 and miR-744 target sequences in the 3′-SL, CS, and 3′UAR regions of the 3′ UTR. The secondary structure of DENV RNA was predicted using the Mfold program (Zuker 2003). The sequences and position numbers in (B) and (C) correspond to DENV-1 RNA.
study
100.0
Endogenous miRNAs alter the expression of GFP fused to the 3′ UTR of DENV RNA - Since our bioinformatics analyses suggested that the DENV 3′ UTR might be targeted by cellular miRNAs, it was important to determine whether the DENV 3′ UTR actually functions as a miRNA target. Based on our findings, the first approach that we used to validate our results was to follow a previously reported strategy (Castillo et al. 2016). The 3′ UTRs of DENV-1, -2, and -4 were individually cloned into the plasmid pEGFP-C1 and the respective constructs, pGUD1, pGUD2, and pGUD4 were obtained. These constructs were subsequently transfected into Vero cells; the empty pEGFP-C1 vector served as the control. Transfection efficiency was determined by fluorescence microscopy and all three plasmids had comparable transfection efficiencies (higher than 50%). We assumed that if endogenous cellular miRNAs that recognise the 3′ UTR of the viral genome exist, a decrease in the GFP expression level compared to pEGFP-C1 should be observed. The GFP expression levels were then assessed by western blotting. Reduced expression of GFP-fused to the 3′ UTR of DENV-1 (pGUD1), DENV-2 (pGUD2), and DENV-4 (pGUD4) versus the control was also observed using the monoclonal anti-GFP antibody (Fig. 2A). Fig. 2B shows the fold change measured by densitometry. A statistically significant inhibitory effect was observed in the expression of GFP fused to the 3′ UTR of DENV-1, DENV-2, and DENV-4 compared with that in the control (pEGFP-C1). To confirm whether among these endogenous miRNAs, miR-484 and miR-744 might have a target site in the DENV 3′ UTR we performed a co-transfection assay using each generated pGUD construct and the plasmids expressing miR-484 and miR-744. Co-transfection of pEGFP-C1 and the miRNA plasmids served as controls. As shown in Fig. 2C-D, significantly reduced GFP expression was observed using miR-484 but unexpectedly with miR-744, GFP reduction was only observed with pGUD1. This might be explained by the difference of the total free energy and opening energy found for miR-744 using the RNAup webserver, which is higher than that for miR-484 and at a lower total energy, and the possibility of an RNA-RNA interaction is greater, as was observed for miR-484. However, the fact that overexpression of miR-744 does not reduce the levels of GFP (Fig. 2D) demands further investigation. In any case, since we observed that Vero cells possess endogenous miRNAs capable of reducing GFP expression (Fig. 2A-B), we wondered whether miR-744 and miR-484 are conserved in humans and monkeys. As a first approach to answer this question, both miR-484 and miR-744 were amplified from HeLa and Vero cells using the same probe. We observed that miR-484 and miR-744 were amplified (data not shown) and although this does not explain the results observed in Fig. 2B, it suggests that the nucleotide sequence of these two miRNAs is conserved in monkeys and humans.
study
100.0
Fig. 2: cellular miRNAs alter the expression of GFP-fused to the 3′ UTR of Dengue virus (DENV) RNA. (A) Vero cells were transfected with pGUD1, pGUD2, pGUD4, or pEGFP-C1 and the GFP expression was determined by western blotting after 24 h, using an anti-GFP antibody; (b) band intensities quantified by densitometry; (C) Vero cells were co-transfected with pGUD1, pGUD2, pGUD4, or pEGFP-C1 and pEZX-mR03-miR-484 or pEZX-mR03-miR-744, and GFP expression was then determined by western blot after 24 h using an anti-GFP antibody; (D) band intensities quantified by densitometry. Data are shown as Median and Range (two way ANOVA). Three independent replicates were performed for each experiment. (*) Statistically significant difference compared to the control (p < 0.05).
study
100.0
Overexpression of miR-484 or of miR-744 suppresses DENV infection - Since the DENV 3’UTR contains target sites for miR-484 and miR-744 and since the expression level of GFP fused to the DENV 3’UTR decreased in Vero cells, we hypothesized that both miRNAs might affect DENV infection. Therefore, the effect of miR-484 and miR-744 overexpression on the DENV life cycle was investigated. The lentiviral plasmids pEZX-MR03-miR-484, pEZX-MR03-miR-744, and the pEZX-MR01-control (scrambled) were transfected into Vero cells, followed by challenge with each DENV serotype at an MOI of 3. Alternatively, the cells were first infected with each DENV serotype and then transfected with each plasmid carrying the gene encoding the corresponding miRNA. The percentage of Vero-infected cells, the DENV RNA copy number, and NS1 expression were evaluated 72 hpi under both conditions. As shown in Fig. 3A, when challenge with DENV occurred 24 h after miR-484 or miR-744 overexpression, the percentage of infected cells was significantly reduced for each of the four DENV serotypes compared with that in the scrambled control. Interestingly, the strongest effect was observed with miR-484 overexpression that reduced the infection of DENV-3 and DENV-4 by up to 60%. For DENV-2, a similar effect was observed with either miRNAs. However, when DENV RNA was quantified by RT-qPCR in the Vero culture supernatants, no difference was observed (Fig. 3B). Similar results were observed when the viral RNA within the infected cells was quantified (data not shown). To confirm our finding that miR-484 and miR-744 play a role in the modulation of DENV infection, the expression of DENV-2 NS1 in infected cells was evaluated by western blotting using anti-NS1 antibodies. Overexpression of miR-484 or miR-744 strongly reduced the amount of NS1 in DENV-2-infected cells compared with that in control infected cells (Fig. 3C). Fig. 3D shows the fold change measured by densitometry, indicating a statistically significant decrease in the expression of DENV-2 NS1 upon overexpression of either miRNA.
study
100.0
Fig. 3: overexpression of miR-484 and miR-744 modulates Dengue virus (DENV) replication. (A) Vero cells were transfected with pEZX-mR03-miR-484, pEZX-mR03-miR-744, or the empty vector pEZX-mR03 (scramble) and after 24 h, challenged independently with each of the four DENV serotypes at an MOI of 3; (B) quantification of DENV RNA copy number in the supernatants by reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). (C, D) DENV-2 NS1 expression by western blotting for each condition evaluated and the band intensities quantified by densitometry, respectively. Ponceau red was used as the loading control for western blotting, as previously reported (Romero-Calvo et al. 2010, Gilda & Gomes 2013, Rivero-Gutiérrez et al. 2014); (E) Vero cells were first challenged independently with each DENV serotype and then transfected with pEZX-mR03-miR-484, pEZX-mR03-miR-744, or the empty vector pEZX-mR03. The percentage of infected cells was evaluated at 72 hpi by flow cytometry. The data are expressed as the percentage of infected Vero cells compared with those in the scrambled infected Vero cells, defined as 100% infection. Results are shown as Median and Range (two way ANOVA, p < 0.005); (F) Quantification of DENV RNA copy number in the supernatants by RT-qPCR. (G, H) DENV-2 NS1 expression by western blotting for each condition evaluated and band intensities quantified by densitometry, respectively. Ponceau red was used as the loading control for western blotting, as previously reported (Romero-Calvo et al. 2010, Gilda & Gomes 2013, Rivero-Gutiérrez et al. 2014). The data are shown as Median and Range (two way ANOVA). Three independent replicates were performed for each experiment. (*) Statistically significant difference compared to the control (p < 0.05).
study
100.0
In contrast, when Vero cells were first challenged with DENV and then transfected independently with pEZX-MR03-miR-484, pEZX-MR03-miR-744, or the pEZX-MR01-control, there was a significant decrease in the percentage of infected cells (less than 50%) only for DENV-4 in the presence of miR-484 (Fig. 3E). Likewise, under these conditions, no change was observed in the number of DENV RNA copies, either in Vero culture supernatants (Fig. 3F) or in the infected cells (data not shown). To verify that both miR-484 and miR-744 indeed affect DENV replication, the level of NS1 expression was determined. DENV-2 NS1 expression was markedly suppressed by miR-484 overexpression, and though the effect of miR-744 overexpression was more modest, NS1 expression was still decreased to about 50% (Fig. 3G). However, as shown in Fig. 3H, when DENV-2 NS1 expression was measured by densitometry, a statistically significant decrease was noted with the overexpression of either miRNA. Taken together, these findings confirm the function of host miR-484 and miR-744 as inhibitors of DENV infection and protein synthesis. Nevertheless, additional studies are required to confirm the antiviral function of these two miRNAs.
study
100.0
DENV-1 to -4 infection downregulates the expression of endogenous miR-484 and miR-744 at the early stages of infection - To better evaluate changes in miR-484 and miR-744 expression, Vero cells were challenged with DENV-1 to -4 at an MOI of 3 or were mock-infected and the expression level of miR-484 and miR-744 was then evaluated at 8, 16, 24, 32, 48, and 72 hpi by RT-qPCR. These stages were considered as early (8-16 h), medium (16-48 h), and late (48-72 h) time-points post-infection. The Ct values were normalised to an uninfected control and to the 18S rRNA (∆∆Ct) to obtain the fold change in expression, according to the manufacturer’s instructions; as previously reported an RQ (relative quantification = 2-∆∆Ct) was considered significant at a minimum of two-fold change. The RT-qPCR results show that miR-484 or miR-744 expression was modulated after DENV infection; indeed, as shown in Fig. 4A-D, miR-484 and miR-744 were downregulated at early time points with all DENV serotypes, and depending on the DENV serotype this level was either maintained or increased progressively at the medium time-points, and finally increased progressively at the late time-points post-infection, except for the level of miR-744 in response to DENV-2 and DENV-4 infection. Taken together, these results suggest that all four DENV serotypes can modulate the endogenous expression of miR-484 and miR-744 at different time-points of infection.
study
100.0
Fig. 4: miR-484 and miR-744 expression levels are downregulated in Dengue virus (DENV)-infected Vero cells. The expression pattern of both miRNAs at different times post-infection in Vero cells infected with DENV-1 to 4 was determined by reverse transcriptase quantitative polymerase chain reaction (RT-qPCR). The results are expressed as the fold change for each miRNA expression level in each sample relative to the mock-infected samples and normalised to 18S rRNA using the 2-ΔΔCt method; (A-D). Data are shown as median and error from three repeated experiments. Three independent replicates were performed for each experiment. (*) Statistically significant difference compared to the control (p < 0.05).
study
100.0
Endogenous miR-484 and miR-744 expression is downregulated by the DENV 3′ UTR - It is reported that sfRNA derived from the DENV 3′ UTR is abundantly expressed during infection by all flaviviruses in cultured cells (Lin et al. 2004) and functions as an RNAi suppressor during flavivirus replication; sfRNA has also been shown to interfere in vitro with the processing of the double-stranded RNA (dsRNA) template by Dicer (Schnettler et al. 2012). Because DENV infection downregulates miR-484 and miR-744 expression in Vero cells, we examined whether the DENV 3′ UTR is involved in this process. Thus, Vero cells were transfected with the pGUD1, pGUD2, or pGUD4 constructs, with the pEGFP-C1 empty vector serving as control. The effect of the 3′ UTR on endogenous miR-484 and miR-744 expression was examined at 12, 24, 48, and 72 h post-transfection (hpt) by RT-qPCR, and for the analysis, each time-point was compared with the control (pEGFP-C1) and the levels of significance were determined (*p < 0.05). As shown in Fig. 5A, the 3′ UTR of DENV-1 RNA (pGUD1), DENV-2 RNA (pGUD2), and DENV-4 RNA (pGUD4) induced a greater than 2-fold change, which is statistically significant, in the expression of miR-484, at all-time-points evaluated, except for DENV-2 at 48 hpt (Fig. 5A). Similar results were obtained for miR-744 (Fig. 5B), but since the observed fold change was less pronounced than that with miR-484, possibly because we used the human miR-744 with the model Vero cells of monkey origin. Thus, we wondered if miR-744 and miR-484 are conserved in humans and monkeys. Firstly, to answer this question, we amplified both miR-484 and miR-744 in HeLa cells and Vero cells, by RT-qPCR using the same probe and observed that both miR-484 and miR-744 were amplified (data not shown). Although this finding does not explain the results in Figs 2B and 5B, it allows us to suggest that the nucleotide sequences of these two miRNAs are possibly conserved between monkeys and humans. Interestingly, the 3′ UTR of DENV-1 RNA induced the strongest decrease in miR-484 and miR-744 expression at 24 hpi (Fig. 5A-B).
study
100.0
Fig. 5: miR-484 and miR-744 expression is downregulated in Vero cells expressing the Dengue virus (DENV) 3′ UTR. Vero cells were transfected with the pGUD1, pGUD2, or pGUD4 constructs or with the empty vector pEGFP-C1 and their effect on miRNA expression was determined. The expression of miR-484 (A) and miR-744 (B) was evaluated at 12, 24, 48, and 72 hpt by reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) and normalised to the untransfected control and to 18S RNA (2-ΔΔCt). Data from RT-qPCR are shown as median and bars are presented from three independent experiments. (*) Statistically significant difference compared to the control (p < 0.05).
study
100.0
Several high-throughput studies have provided strong evidence that miRNAs might negatively or positively regulate the viral life cycle or play a critical role in host-virus interactions (Huang et al. 2007). These important features make miRNAs potential therapeutic targets in the treatment of several infectious diseases such as dengue, for which neither therapeutic treatments nor vaccines are currently available. Studies exploring the interaction between DENV and cellular miRNAs will be important for providing insights into the cellular defences against DENV. However, very few reports show the effect of cellular miRNAs on DENV infection or on the changes in miRNA expression during DENV infection. In this study, we focused on the effect of DENV infection on the expression of miR-484 and miR-744 and on the involvement of these two miRNAs on DENV replication when overexpressed. We report that the 3′ UTR of all four DENV serotypes present target sequences for cellular miRNAs, including miR-484 and miR-744. Interestingly, these miRNAs target a sequence localised in the 3′-SL of the 3′ UTR of DENV RNA. Since this 3′-SL contains the 3′CS and 3′UAR elements that are critical for efficient host and viral protein recruitment involved in viral replication, we suggest that the interaction of the 3′ UTR with miR-484 or miR-744 might alter the viral life cycle. This notion is strengthened by the results obtained in Vero cells expressing GFP-fused to the 3′ UTR of DENV plasmids that showed decreased GFP expression (Fig. 2A-B). However, further studies are needed to confirm this hypothesis, since overexpression of GFP-fused to the 3′ UTR of the DENV constructs in the presence of a plasmid containing the miR-744 gene did not yield the same results, except with pGUD1, i.e. we did not observe decreased GFP expression (Fig. 2C-D). However, it should be considered that although our in silico prediction of miR-484 and miR-744 interaction with DENV RNA relies on sequence complementarity and site homology, the seed sequence complementarity (shaded nucleotides) is lower for miR-744 than that for miR-484 (black circles) as shown in Fig. 1B, indicating that these two miRNAs have different sequences, including the seed sequence present in the first nucleotides from the 5′ end.
study
99.94
It has been shown that miR-484 and miR-744 may be expressed in different human cells including macrophages and dendritic cells (our unpublished observations), which are both DENV target cells, and that their expression can be regulated by external factors that alter cell homeostasis, such as diseases (Tan et al. 2015). However, despite the different studies on the expression of these miRNAs, thus far, no specific function has been attributed to either miRNA, but miR-744 was shown to enhance the IFN-I signalling pathway by targeting protein tyrosine phosphatase 1B, a ubiquitously expressed phosphatase, in primary human renal meningeal cells (Zhang et al. 2015).
study
99.94
Since miR-484 and miR-744 overexpression suppresses DENV-2 protein production such as NS1, we suggest that both miRNAs might affect DENV replication, but simultaneously, DENV infection decreases the expression of miR-744 and miR-484. Although the mechanisms through which miR-484 and miR-744 inhibit DENV replication remain uncertain, the DENV 3′ UTR is known to contain conserved sequences involved in viral RNA circularisation and host-protein interaction. Consequently, such an effect might be associated with interference in these events. For example, PTB is reported to interact with the 3′ UTR of DENV RNA and is required for efficient DENV propagation in Vero cells (de Nova-Ocampo et al. 2002). Furthermore, we recently reported that miR-133a inhibits DENV replication by decreasing PTB expression (Castillo et al. 2016), a direct and functional target of miR-133a. Interestingly, the target sites of miR-133a, miR-484, and miR-744 are located in a single region, highlighting a ‘hotspot’ of potential MRE. The interaction of these miRNAs with the 3′ UTR of DENV RNA might lead to failure in the recruitment of PTB, thus affecting the circularisation of the viral genome, with consequences on virus replication or RNA translation. Our data are consistent with those of a previous report showing that cellular miRNAs suppress/inhibit DENV replication (Wen et al. 2015). Furthermore, antisense oligomers or small interfering RNAs directed against target sequences in the 3′ UTR of DENV RNA are also known to inhibit virus replication and translation, presumably by blocking either the RNA-RNA interactions by steric interference, or the RNA-protein interaction complexes involved in the synthesis or translation of viral RNA (Holden et al. 2006). Similar results were observed using a related strategy for the avian leucosis virus subgroup J (ALV-J) (Wang et al. 2013), where the authors showed that the activity of the luciferase-reporter gene carrying the 5′ and 3′ UTR of ALV-J decreased with the host-encoded gga-miR-1650 that interacts with the 5′ UTR. The demonstration that host miR-484 and miR-744 share a target site in the 3′ UTR of the RNA in all four DENV serotypes suggests that these miRNAs might be part of the host antiviral response against DENV. However, it is necessary to emphasize that the anti-DENV activity of miR-744 was much stronger compared to the activity observed with miR-744. Notably, this is an unexpected result since both miRNAs share some degree of sequence homology, although the miR-484 seed region shares complete homology with the 3′ UTR of the DENV recognition site. In line with our results, Wu et al. (2014) also recently reported that DENV-2 infection decreases miR-223 expression in Vero cells, but that overexpression of miR-223 suppresses DENV-2 replication, suggesting that miR-223 may be present as an antiviral factor against DENV-2. Furthermore, our results gain importance because miR-744 was reported to direct the post-transcriptional regulation of TGF-β1, which is crucial in inflammation (Martin et al. 2011), and this cytokine is reported to be involved in severe dengue disease (Chen et al. 2009). In addition, a large body of evidence has indicated that TGF-β1 gene polymorphisms are associated with increased susceptibility to dengue haemorrhagic fever and higher virus load, and high TGF-β1 production has been linked to protection or a mild clinical outcome of dengue infection (Chen et al. 2009).
study
99.94
Although our results suggest that miR-484 and miR-744 possess antiviral activity, we also observed that both DENV infection and the expression of the 3′ UTR of DENV RNA can downregulate miR-484 and miR-744. Although it is still not clear how DENV promotes miR-484 and miR-744 downregulation, previous findings have identified the pathway by which DENV downregulates the expression of cellular miRNAs. Indeed, NS4B from all 4 DENV serotypes appears to inhibit the processing of Dicer, a protein involved in miRNA biogenesis of dsRNA (Kakumani et al. 2013). Similar behaviour has been reported for other viral proteins such as HIV-1 Nef, a regulatory and accessory protein that interacts with the RNAi pathway protein, Ago2, thereby suppressing miRNA silencing concentrated in the multivesicular bodies where HIV-1 actively replicates (Aqil et al. 2013). Interestingly, several recent reports have provided evidence that certain non-coding RNAs may function as competing endogenous RNAs (ceRNAs) in modulating the concentration and biological functions of miRNAs. These ceRNAs generally share miR-response elements with the transcripts of several important genes and prevent these mRNAs from being degraded. In accordance with these results, it was recently proposed that HBV mRNA acts as a ceRNA for miRNA-15a to regulate TGF-signalling, which contributes to the development of HBV-related hepatocellular carcinomas (HCCs) (Liu et al. 2015). Based on these results and because it was demonstrated that the West Nile virus (WNV) sfRNA suppresses the siRNA- and miR-induced RNAi pathway (Schnettler et al. 2012), we suggest a role for DENV RNA, through its 3′ UTR or sfRNA, that would act as a sponge for the hybridisation of endogenous miRNAs, and as a potential mechanism to reduce their expression; such a mechanism was demonstrated for HBV mRNAs possessing a miR-15a/16-complementary site that acts as a sponge to bind and sequester endogenous miR-15a/16 (Liu et al. 2013).
study
100.0
We propose that the interaction of viral RNA with the deregulated RNA-induced silencing complex (RISC) containing the mature miRNAs could prevent genome circularisation, thus affecting viral RNA translation. This could explain the reduction in the percentage of infected cells observed at 72 hpi, compared to that in the control (scrambled) cells, but only when either miR-744 or miR-484 were first overexpressed, whereas no significant effect was observed when the cells were first infected, and at 24 hpi either miRNA was overexpressed with no reduction in DENV RNA in either strategy. This is interesting since the expression of miR-744 and miR-484 was decreased during the first 48 hpi but later, their expression was increased mainly at 72 hpi in Vero cells infected with DENV-1 or DENV-3. Why the expression of these two miRNAs increases with a delay only in response to DENV infection but not when the 3′ UTR fused downstream of the GFP ORF was overexpressed, is unknown. We suggest that this effect is due to an immune interaction caused by DENV infection; e.g., in natural killer cells, TLR7 stimulation was reported to induce the expression of miR-744 (Voynova et al. 2015). Further related studies are being performed in our laboratory to examine and confirm these possible mechanisms of miR-484 and miR-744 action in DENV infection, using primary human cells, such as macrophages/dendritic cells, or in patients with dengue. However, further studies will be required to lend weight to this possibility and to determine whether miR-484 and miR-744 are possible therapeutic targets for all DENV serotypes, even more so if we consider that the present study was carried out in monkey cells. In summary, we propose that DENV can escape the antiviral activity of miR-484 and miR-744, downregulating their expression, and simultaneously, that miRNAs target an unknown cellular factor that promotes DENV replication. Therefore, our results imply that DENV infection modulates host miR expression for their own benefit.
study
99.94
miR-484 and miR-744 can be considered as two possible restriction host factors against DENV infection, but the virus might have evolved to resist inhibition by endogenous human miRNAs during productive replication. This finding suggests that DENV RNA is a target of cellular miRNAs. Furthermore, this study contributes to a better understanding of the relationship between host miRNAs and DENV, and recommends further studies to decipher the biological functions of this interaction to develop a therapy based on these two miRNAs for the control of infection and treatment of dengue illnesses.
study
99.94
Clinically isolated syndrome (CIS) is the first manifestation of multiple sclerosis (MS), a chronic inflammatory autoimmune disease of the central nervous system (CNS) affecting over 2.5 million people worldwide. After CIS, the delay in manifestation of a new relapse, which corresponds to the clinically definite MS (CDMS) according to McDonald’s criteria 2005 , varies from several months to more than 10 years and should be associated with the brain and spinal cord pathology at the time of a relapse . In about a third of the cases, CIS patients will not experience a new relapse activity over long-term follow-up .
review
99.8
Different clinical and brain imaging predictors have been evaluated for their sensitivity and specificity to predict occurrence of a relapse during follow-up. In this context, a number of studies have shown that abnormal MRI findings are the most informative predictors of future disease activity [5, 6].
review
99.75
On the other hand, the role of cerebrospinal fluid (CSF) biomarkers for the prediction of a relapse remains to be explored. It is also unclear if MRI predictors together with CSF biomarkers may improve prediction of CDMS when applied in a single model. Apart from oligoclonal bands (OCBs), several markers in the CSF appear to be more specific for neuro-inflammatory and neuro-degenerative pathophysiological processes in MS, such as inflammation and immune dysfunction, or cell type, such as B cells . Some of these markers have been shown to predict conversion to MS in patients with CIS, e.g. polyspecific intrathecal B-cell response of IgG antibodies against neurotropic viruses such as measles, rubella, and varicella zoster . CSF IgG heavy-chain bias was detected in patients with CIS who converted to MS within 6 months of the CIS presentation , and increased CSF concentrations of B cell recruiting chemokine CXCL13 were shown to be a good predictor of conversion to MS in patients with CIS over 2 years of follow-up . It has been suggested that Tau and neurofilaments, CSF markers of axonal damage, might be even more specific than MRI for predicting conversion of CIS to MS. Shorter time to CDMS conversion was also associated with high concentrations of CSF chitinase 3-like 1, which is up-regulated during inflammation .
review
99.75
Nevertheless, up to date none of these CSF markers can be recommended for routine implementation in clinical practice, mainly because of methodological limitations, invasiveness of the spinal tap, absence of conclusive data from small studies or other technical problems. Therefore large multicentre studies are needed to confirm the importance of CSF biomarkers as a tool in the clinical practice . CSF is the compartment in the closest proximity to CNS parenchyma and might reflect immunopathology in CNS. However, studies usually fail to provide longitudinal CSF data because repeated lumbar punctures are difficult to justify. Therefore, single CSF samples are only snapshots of immune response at the time of collection.
review
99.8
In this context, peripheral blood as a mirror of immune reaction in CNS is much easier to measure longitudinally. CD4+ T cells and later CD4 + Th1 cells are the most studied cell populations in MS because of their potential role in the pathogenesis of the disease, as well as they are used to assess the effect of different therapies used in MS. More recently, regulatory subpopulations such as Th2 cells, regulatory CD4+ T cells and NK cells were studied for their relationship in disease prognosis and radiologically confirmed MS activity. Changes in their effector populations were described . The association of CD4 + CD45RO+ IL-17A+ cells to clinical and radiological disease activity was reported. [17, 18] Differences in naive CD4 T-cell biology, notably in TCR and TLR signalling pathways, identified patients with MS with more rapid conversion to secondary progression .
review
99.8
In our study, we followed-up patients with CIS who were treated with interferon beta, irrespective of the further disease course (patients with both subsequent relapse and no relapse were included) for at least 48 months. Peripheral blood immunophenotyping was used to find early changes in lymphocyte subsets, which could predict the development of relapse.
study
100.0
The study enrolled patients after first clinical event suggestive of MS within 4 months from the first clinical event. The inclusion criteria were:18–55 years of age, enrolled within 4 months from the clinical event, Expanded Disability Status Scale (EDSS) ≤3.5, presence of ≥2 T2-hyperintense lesions on diagnostic MRI, and presence of ≥2 oligoclonal bands in CSF obtained at the screening visit prior to steroid treatment. The exclusion criteria were: occurrence of a second relapse before the baseline visit, pregnancy and symptoms that could possibly be attributed to neurological diseases different from MS (e.g. neuromyelitis optica).
study
99.94
All patients included with CIS were treated with intravenous steroids (3–5 g of methylprednisolone) following screening and preceding baseline, and subsequently at baseline started the treatment with 30 mg of intramuscular IFNb-1a once a week. Two hundred and twenty CIS patients were enrolled in the study; analyses in this laboratory part of study were limited to 191 subjects with immunophenotyping flow cytometry data available. Clinical assessments (EDSS) and peripheral blood assessments were obtained at baseline, 6, 12, 24, 36 and 48 months. In the case of relapses, patients were evaluated within 4 days from the onset of new symptoms.
study
100.0
We used two clinical outcomes in this analysis – 1) development of a new clinical relapse, and 2) confirmed disability progression. A new clinical relapse was defined as patient-reported symptoms or objectively observed signs typical for an acute inflammatory demyelinating event in the central nervous system with duration of at least 24 h, in the absence of fever or infection. Confirmed disability progression (CDP) was defined as an increase in EDSS by 1.0 point (if baseline EDSS >0) or 1.5 points (if baseline EDSS = 0) confirmed after 6 months .
study
99.94
Whole peripheral blood (PB) samples (2 ml/subject) were collected in ethylenediamine tetra-acetic acid (EDTA) test tubes. The following monoclonal antibodies from BD Biosciences (San Jose, CA, USA) were used: CD3 FITC, CD4 PE-Cy7, CD5 FITC, CD8 APC Cy7, CD19 PE, CD45RA FITC, HLA-DR PE Cy7, CD16 + 56 PE and CD45 PerCP. Whole PB samples were labelled with appropriate volumes of conjugated MoAb for 20 min at room temperature, and then lysed with 2 ml of lysis solution (FACS Lysis Solution, BD). Cells were washed twice and analyzed on a standard FACSCanto instrument (BD) with DIVA software (BD). All results were expressed as percentages of each subset out of total lymphocytes.
study
99.94
Percentages of lymphocyte populations do not have normal distribution and therefore, the parametric tests cannot be applied. Moreover, the results at the end of the study were influenced by different treatment modalities in study patients. For statistical assessment, only individual differences between baseline values and values at the time of relapse were used and compared with various indicators.
study
100.0
Predictive values of measured variables were investigated through receiver operating characteristic (ROC) curves. Based on these curves, thresholds which maximised the sum of sensitivity and specificity for each measured parameter were defined. A Kaplan-Meier estimator of survival probability and an asymptotic log-rank test were used to test differences between subsets of patients. For each lymphocyte population and the clinical outcome (relapse or CDP) three different indicators were considered: a) the population relative value, b) the population relative value change compared to baseline and c) the population relative value change compared to a measurement one year before.
study
100.0
For all indicators the following analyses were performed:An ROC curve to predict the clinical outcome at one year was constructed and an optimal threshold was chosen.At every time-point a cohort of patients was divided into two groups according to the chosen threshold and survival curves (Kaplan-Meier estimators) were constructed for each of these groups. Further, an asymptotic log-rank test was performed to test difference between these survivals probability curves (see Figures and Tables). P-values less than 0.05 were considered as statistically significant.
study
100.0
At every time-point a cohort of patients was divided into two groups according to the chosen threshold and survival curves (Kaplan-Meier estimators) were constructed for each of these groups. Further, an asymptotic log-rank test was performed to test difference between these survivals probability curves (see Figures and Tables). P-values less than 0.05 were considered as statistically significant.
study
99.94
The ROC curves were constructed from all the time points at once. As a predicted outcome we considered a relapse or disease progression, respectively that appeared in (and only in) a year after the corresponding measurement. The events were analysed from the year end. The same threshold applied to measurements before baseline, at baseline, at 6 M, etc. Every time it divided patients into two groups. For those groups the survival probability curves were constructed and compared.
study
99.94
Of the 220 patients enrolled in the study, 191 had available clinical and laboratory follow-up data (181 till the end of study, not only till relapse or disability progression) and were included in the analysis. During the first year of follow-up, second disease relapse was observed in 55 patients, during the second year additional 25 patients relapsed, and 21 and 13 patients, respectively, relapsed during the third and the fourth year of the follow-up. At 48 months, 114 (59.6%) patients experienced the second clinical attack, 69 (36.1%) experienced the third clinical attack and 37 (19.3%) had confirmed disability progression (CDP).
study
99.94
Representation of different lymphocyte subpopulations in patients who did not experience a relapse was compared to patients who relapsed within the respective time period (6, 12, 24, 36 and 48 months). The same comparison was done for patients who showed CDP. Patients who underwent both events (relapse and CDP) were included in both subgroups. Table 1a, b shows the number of patients in each subgroup and median ± standard deviation of measured lymphocyte subpopulations. Absolute counts of lymphocyte subpopulations are shown in Table 2a, b.Table 1a, b Median ± standard deviation of lymphocyte subpopulations (in % of all lymphocytes) in different subgroupsanT lyCD4+CD8+B lyNK cellsCD5+all patients at first relapse19171.8 ± 744.6 ± 723 ± 611.1 ± 414 ± 772 ± 7all patients at baseline (BL)a 19171.5 ± 743 ± 824 ± 79.3 ± 416.1 ± 771 ± 9all patients at 48 months (48)b 18172.5 ± 745 ± 721.7 ± 615.3 ± 69.6 ± 573.9 ± 6patients without relapse at BL7771.2 ± 742.3 ± 823.4 ± 79.3 ± 416.5 ± 871.2 ± 8patients without relapse at 48b 7773 ± 646.2 ± 622.1 ± 614.8 ± 59.45 ± 474.5 ± 5patients with relapse at BL11471.7 ± 843.5 ± 724.5 ± 69.3 ± 415.4 ± 770.7 ± 10patients with relapse at 48b 10472.1 ± 845.9 ± 821.4 ± 615.8 ± 79.8 ± 573.6 ± 7.patients without CDP at BL15471.3 ± 842.2 ± 824.5 ± 79.1 ± 417 ± 770.8 ± 9patients without CDP at 48b 14872.5 ± 745.7 ± 822.2 ± 615.3 ± 69.9 ± 573.6 ± 6patients with CDP at BL3773.4 ± 647 ± 723.1 ± 610 ± 413 ± 672.8 ± 6patients with CDP at 48b 3374.6 ± 747.8 ± 720.5 ± 516.3 ± 68.4 ± 375.3 ± 6bnDR+ in T lyCD45RA+CD45RA+ in CD4+CD5+ in B lyCD5+ B lyall patients at first relapse19110.8 ± 965.4 ± 1158.8 ± 1920 ± 162.1 ± 2all patients at baseline (BL)a 19112.1 ± 1165.2 ± 1056.4 ± 2018.1 ± 191.7 ± 2all patients at 48 months (48)b 18112.2 ± 665.1 ± 844.8 ± 1211.1 ± 71.7 ± 1patients without relapse at BL7712.9 ± 1164.9 ± 957.5 ± 2023.3 ± 181.9 ± 3patients without relapse at 48b 7712.1 ± 664.3 ± 845.2 ± 1211.4 ± 81.9 ± 2patients with relapse at BL11411.8 ± 1063.4 ± 1155.6 ± 2118 ± 141.7 ± 2patients with relapse at 48b 10412.8 ± 566.2 ± 844.8 ± 1310 ± 61.7 ± 2patients without CDP at BL15412.5 ± 1064.8 ± 1054.9 ± 2018.1 ± 201.7 ± 2patients without CDP at 48b 14812.4 ± 666.1 ± 845.8 ± 1311.6 ± 81.9 ± 1patients with CDP at BL379.8 ± 962.4 ± 1260.5 ± 2316.7 ± 141.9 ± 2patients with CDP at 48b 3311.1 ± 562.7 ± 940.7 ± 119.7 ± 51.4 ± 1Double positive populations (HLA-DR + CD3+, CD45RA + CD4+, CD5 + CD19+) are expressed as the percentage of the first mentioned subpopulation in the basic population. CD5+ B lymphocytes are expressed as the originally measured value (percentage of total lymphocytes) to enable comparison with previously published results. aThe laboratory test were done before intravenous steroids treatment, interval between this examination and baseline was 2–3 months bThe results at 48 months of following should be influenced by treatment (natalizumab n = 18, fingolimod n = 3, copaxone n = 6)T ly – T lymphocytes (CD3+), B ly – B lymphocytes (CD19+), NK cells (CD3-CD16 + 56+) Table 2a, b Median ± standard deviation of lymphocyte subpopulations (absolute count in 103/ml) in different subgroupsanT lyCD4+CD8+B lyNK cellsCD5+all patients at first relapse1911.3 ± 0.40.8 ± 0.30.4 ± 0.20.2 ± 0.10.2 ± 0.11.3 ± 0.4all patients at baseline (BL)a 1911.3 ± 0.40.8 ± 0.30.5 ± 0.20.2 ± 0.10.3 ± 0.21.3 ± 0.4all patients at 48 months (48)b 1811.3 ± 0.50.8 ± 0.30.4 ± 0.20.3 ± 0.20.2 ± 0.11.3 ± 0.5patients without relapse at BL771.3 ± 0.30.8 ± 0.20.4 ± 0.20.2 ± 0.10.3 ± 0.21.3 ± 0.3patients without relapse at 48b 771.3 ± 0.40.8 ± 0.20.4 ± 0.20.3 ± 0.10.2 ± 0.11.4 ± 0.4patients with relapse at BL1141.3 ± 0.50.8 ± 0.30.5 ± 0.20.2 ± 0.10.3 ± 0.11.3 ± 0.5patients with relapse at 48b 1041.3 ± 0.50.8 ± 0.30.4 ± 0.20.2 ± 0.20.2 ± 0.11.3 ± 0.6patients without CPD at BL1541.3 ± 0.40.8 ± 0.30.4 ± 0.20.2 ± 0.10.2 ± 0.11.3 ± 0.4patients without CPD at 48b 1481.3 ± 0.50.8 ± 0.30.4 ± 0.20.3 ± 0.20.2 ± 0.11.4 ± 0.5patients with CPD at BL371.2 ± 0.40.8 ± 0.20.3 ± 0.20.2 ± 0.10.2 ± 0.11.2 ± 0.4patients with CPD at 48b 331.1 ± 0.50.8 ± 0.30.3 ± 0.20.3 ± 0.20.1 ± 0.11.2 ± 0.6bnDR+ CD3+CD45RA+CD45RA+ CD4+CD5+ B lyall patients at first relapse1910.1 ± 0.11.1 ± 0.40.5 ± 0.20.04 ± 0.05all patients at baseline (BL)a 1910.2 ± 0.11.2 ± 0.40.5 ± 0.30.03 ± 0.05all patients at 48 months (48)b 1810.2 ± 0.11.1 ± 0.50.4 ± 0.20.03 ± 0.03patients without relapse at BL770.2 ± 0.11.2 ± 0.30.5 ± 0.20.03 ± 0.05patients without relapse at 48b 770.2 ± 0.11.1 ± 0.40.4 ± 0.20.03 ± 0.04patients with relapse at BL1140.2 ± 0.11.2 ± 0.50.5 ± 0.30.03 ± 0.04patients with relapse at 48b 1040.1 ± 0.11.1 ± 0.60.4 ± 0.20.03 ± 0.03patients without CDP at BL1540.2 ± 0.11.2 ± 0.40.5 ± 0.30.04 ± 0.05patients without CPD at 48b 1480.1 ± 0.11.1 ± 0.50.4 ± 0.20.03 ± 0.03patients with CPD at BL370.1 ± 0.21 ± 0.40.4 ± 0.20.04 ± 0.04patients with CPD at 48b 330.1 ± 0.11 ± 0.60.3 ± 0.20.02 ± 0.03 aThe laboratory test were done before intravenous steroids treatment, interval between this examination and baseline was 2–3 months bThe results at 48 months of following should be influenced by treatment (natalizumab n = 18, fingolimod n = 3, copaxone n = 6)T ly – T lymphocytes (CD3+), B ly – B lymphocytes (CD19+), NK cells (CD3-CD16 + 56+)
study
100.0
Double positive populations (HLA-DR + CD3+, CD45RA + CD4+, CD5 + CD19+) are expressed as the percentage of the first mentioned subpopulation in the basic population. CD5+ B lymphocytes are expressed as the originally measured value (percentage of total lymphocytes) to enable comparison with previously published results.
study
98.75
The results are shown for a better orientation and to complement the statistical analysis, but a few trends were observed in the results, the most noticeable ones were the increase in B lymphocytes and the decrease in NK cells. These trends were present in all subgroups, regardless of the clinical status (conversion to CDMS or no conversion). However, a profound decrease in B lymphocytes and an increase in NK cells was observed between the study entry and the baseline results and of at least 30 days prior to initiation of interferon therapy (of at least 30 days after steroid administration).
study
100.0
We aimed to test the possibility of finding a threshold value that would distinguish patients with a higher or lower probability of relapse or CDP (the “value” in Figs. 1-4). More importantly, we aimed to measure changes in lymphocyte subpopulations longitudinally in each patient separately so that the disease course could be predicted. Therefore, we tested the ratio of current values versus values obtained a year before compared to baseline (Ratio vs. LY and Ratio vs. BL Figs. 1-4).Fig. 11) ROC curves of CD19+ B lymphocytes to predict relapse at one year. AUC for the population relative value = 0.505, AUC for the population relative value change compared to baseline = 0.513, AUC for the population relative value change compared to a measurement one year before = 0.495. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
1) ROC curves of CD19+ B lymphocytes to predict relapse at one year. AUC for the population relative value = 0.505, AUC for the population relative value change compared to baseline = 0.513, AUC for the population relative value change compared to a measurement one year before = 0.495. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
99.94
For the assessment of lymphocyte subpopulations, relative percentage is usually used, but absolute counts can be also calculated. As interferon beta decreases the overall lymphocyte count, including absolute values of respective subpopulations, we selected relative values for the statistical analysis. Since only results obtained before a relapse or CDP were statistically assessed, no other treatment than interferon beta could have influenced the parameters, and treatment was the same in all patients. The results obtained from patients after therapy escalation (18 natalizumab, 3 fingolimod) and also after changing for copaxone (6 patients) were excluded from this part of our study. Also methylprednisolone pulses were the same in all patients and the treatment was given at least 30 days before baseline.
study
100.0
In B lymphocytes, higher levels (the population relative value in our study was 9.5% of the total count of lymphocytes) were a positive predictive factor. (Figure1) The decrease of B lymphocytes below this level increased the probability of a relapse from month 6 until the end of the follow-up.
study
100.0
Analysis of NK cells (Fig. 2) stressed the importance of long-term follow-up of representation of this population. The decrease of NK cells below 10.8% of the total count of lymphocytes increased the probability of CDP. Overall, a decreasing trend in the percentage of this population in peripheral blood was present, but transient more significant increase (of more than 28.7%) increased the probability of EDSS worsening in the first and third year of follow-up.Fig. 21) ROC curves of CD3-CD16 + 56+ NK cells to predict CDP at one year. AUC for the population relative value = 0.672, AUC for the population relative value change compared to baseline = 0.507, AUC for the population relative value change compared to a measurement one year before = 0.541. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
1) ROC curves of CD3-CD16 + 56+ NK cells to predict CDP at one year. AUC for the population relative value = 0.672, AUC for the population relative value change compared to baseline = 0.507, AUC for the population relative value change compared to a measurement one year before = 0.541. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
Percentages of naïve (CD45RA+) lymphocytes did not vary much (Table 1), but the statistical analysis revealed a significant relationship between this particular population and EDSS worsening. (Fig. 3) The probability of CDP was lower if the percentage of naïve lymphocytes was higher than 64.65% of total count of lymphocytes, and this was valid from the first year of follow-up. A more prominent decrease (to less than 95.9% of the baseline value) was a negative factor. However, the results were not significant when assessed with regard to the second relapse.Fig. 31) ROC curves of naive lymphocytes CD45RA+ to predict CDP at one year. AUC for the population relative value = 0.677, AUC for the population relative value change compared to baseline = 0.591, AUC for the population relative value change compared to a measurement one year before = 0.499. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
1) ROC curves of naive lymphocytes CD45RA+ to predict CDP at one year. AUC for the population relative value = 0.677, AUC for the population relative value change compared to baseline = 0.591, AUC for the population relative value change compared to a measurement one year before = 0.499. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
99.94
Due to the fact that different subpopulations were included within CD45RA+ lymphocytes that may vary and influence the total results, we focused on one of the subpopulations – CD4+ naïve cells. (Fig. 4) In this subpopulation, the results were also statistically more significant when related to EDSS worsening. The percentage above 52.3% of total count of lymphocytes seemed to be important for protection against the clinical deterioration and the decrease of more than 30.5% as compared to the baseline value was a negative prognostic factor. No significant results were found in the context of the relapse.Fig. 41) ROC curves of naive helper cells CD45RA+ in CD4+ to predict CDP at one year. AUC for the population relative value = 0.662, AUC for the population relative value change compared to baseline = 0.628, AUC for the population relative value change compared to a measurement one year before = 0.523. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
1) ROC curves of naive helper cells CD45RA+ in CD4+ to predict CDP at one year. AUC for the population relative value = 0.662, AUC for the population relative value change compared to baseline = 0.628, AUC for the population relative value change compared to a measurement one year before = 0.523. 2) Survival curves and asymptotic log-rank tests for groups divided by the threshold
study
100.0
The explanation for the fact that we found more significant changes in case of the CDP assessment than in the relapse one could be our incapability to guarantee examination at the same interval from relapse, as the examination was performed in predefined (yearly) intervals, irrespective of the presence and/or timing of relapse. We discovered CDP parameter to be more consistent for such an evaluation.
study
99.9
Various immunological markers with a potential of predicting conversion to CDMS or a relapse were not the only parameters analysed during a long-term follow-up of the CIS patients, which complemented results of radiological investigation of the same cohort. .
study
99.94
The results of our study corresponded with our long-time observation of immunophenotyping patterns in MS patients, i.e. higher level of CD4+ lymphocytes with a decreased expression of CD45RA molecules and lower proportion of B lymphocytes and NK cells. Almost all these results fit within normal values because of wide variability of these parameters. The same problem makes parametric statistical tests impossible for use in our and all such studies and this is the reason why we cannot use the value of these subsets as a MS diagnostic marker. We therefore decided to look at our data at an individual patient level and compared individual differences in observed subsets during 48 months of follow up in patients with newly diagnosed CIS treated with methylprednisolone at the beginning of the study and treated with interferon β either for the whole study (48 months) or until the time of a relapse or disability progression (before treatment escalation).
study
100.0
The key role in MS pathogenesis had been assigned to CD4+ cells for a long time and so this population was the most studied one, but low numbers of patients and different methods caused rather conflicting results . In general, the levels of CD4+ cell population in MS patients (unless influenced by treatment (e.g. fingolimod)), are at the upper limit of normal range. Values of CD4+ T lymphocytes in our CIS patients at baseline were lower than we commonly found in MS patients in our laboratory and so we assumed significant augmentation in this subset during the disease progression to CDMS, but only a tendency to increase during the whole time of follow-up was found in all groups. This supports our anticipation, but we cannot designate whole CD4 + CD3+ lymphocyte subpopulation as a predictive factor for conversion to MS.
study
100.0
In 1986, Mossman and Coffman presented the concept of distinct T helper cell subsets and MS was later considered a CD4+ Th1-mediated autoimmune disease but not surprisingly the T-cell biology in vivo is more complex than a simple dichotomy. In addition to Th1/Th2, the Th17 subset and different classes of regulatory T cells (Treg) are involved in MS. Recent studies have shown that the T cells mediating MS can be heterogeneous, with Th17 cells predominating in some individuals and Th1 cells in others The plasticity of different T cell subsets and emerging evidence that subset-signature cytokine expression is not as stable as initially believed strongly support our attempt to find a more reliable marker as a predictor of conversion to CDMS.
review
99.5