All animals were maintained in specific pathogen-free conditions in the American Association for Laboratory Animal Science-accredited facility at the University of Texas at San Antonio. All experiments were approved by the Institutional Animal Care and Use Committee at the University of Texas at San Antonio and performed in accordance with the relevant guidelines and regulations. Mice were fed and watered ad libitum.

C57BL/6 female mice were purchased from the Jackson Laboratory (Stock number 000664; Bar Harbor, ME, USA). 6–8 weeks old female mice were used in all experiments.

Active EAE was induced in C57BL/6 female mice by subcutaneous (s.c.) injection of 200 µg myelin oligodendrocyte glycoprotein (MOG)_35–55 peptide (United Biochemical Research, Seattle, WA, USA) in 50 µL of complete Freund’s adjuvant (CFA) containing 5 mg/mL Mycobacterium tuberculosis H37Ra (Difco Laboratories, Detroit, MI, USA). Mice also received intraperitoneal injections of 400 ng Bordetella pertussis toxin (PTx) (List Biological, Campbell, CA, USA) on days 0 and 2. For induction of passive EAE by adoptive transfer, donor mice were immunized s.c. with 200 µg of MOG_35–55 in CFA. Splenocytes and draining lymph node cells were collected from donor mice on day 10 and restimulated for 3 days at 37°C with 10 µg/mL of MOG_35–55 peptide in complete DMEM containing 10% fetal bovine serum, 20 ng/mL of recombinant mouse IL-23 (R&D Systems, Minneapolis, MN, USA), and 10 µg/mL of anti-IFN-γ monoclonal antibody (mAb) (Bio X Cell, West Lebanon, NH; R4-6A2). Recipient mice were injected with 4,500 restimulated IL-17-producing donor cells.

Mice were assigned clinical scores for EAE according to the following scale (16): 0, no abnormality; 1, limp tail; 2, moderate hind limb weakness; 3, complete hind limb paralysis; 4, quadriplegia or premoribund state; 5, moribund, death.

Mice were anesthetized intranasally using 3% isoflurane in a rodent anesthesia system (Harvard Apparatus, Holliston, MA, USA) and immediately inoculated intravaginally with 5 × 10⁴ inclusion-forming units (IFU) of C. muridarum in 5 µL of sterile SPG buffer, as previously described (17). Vaginal vaults of challenged mice were swabbed at 3-day intervals, and swabs were collected into Eppendorf tubes containing 4-mm glass beads (Kimble, Vineland, NJ, USA) and 500 µL of sterile SPG buffer. The tubes were vortexed for 1 min, and swab material was plated and incubated for 28 h with HeLa cells grown on coverslips in 24-well plates to facilitate Chlamydia replication and growth. The infected HeLa cells were fixed with 2% paraformaldehyde and permeabilized with 2% saponin. Cells were washed using PBS and incubated with DMEM containing 10% fetal bovine serum for 1 h to block non-specific binding. Thereafter, cells were washed and incubated with polyclonal rabbit anti-Chlamydia antibody for 1 h and then incubated for an additional 2 h with goat anti-rabbit immunoglobulin conjugated to fluorescein isothiocyanate (Sigma) plus Hoechst nuclear staining. The treated coverslip cultures were then washed and mounted onto Superfrost microscope slides (Fisher) by using FluorSave reagent (Calbiochem). Slides were visualized using a Zeiss Axioskop 2 Plus research microscope (Zeiss, Thornwood, NY, USA). The bacterial shedding was calculated and expressed as the number of IFU per animal.

Cytokine ELISPOT assay was performed and spots analyzed as described previously (18, 19). In brief, ELISPOT plates (Millipore, Billerica, MA, USA) were precoated with anti-mouse-IFN-γ mAb (eBioscience, AN-18) and anti-mouse-IL-17 mAb (Bio X Cell; 17F3). Splenocytes (5 × 10⁵ cells/well) and brain-infiltrating mononuclear cells (2–5 × 10⁴ cells/well) were restimulated with MOG_35–55 peptide in HL-1 medium (Lonza) at 37°C for 24 h. Biotinylated anti-mouse-IFN-γ mAb (eBioscience; R4-6A2) and anti-mouse-IL-17 mAb (BioLegend, TC11-8H4) were then added overnight at 4°C, followed by incubation with streptavidin alkaline phosphatase (Invitrogen, Waltham, MA, USA) for 2 h at room temperature and developing with BCIP/NBT Phosphatase Substrate (KPL, Gaithersburg, MD, USA). After plate developing, image analysis of spots was performed on a Series 2 Immunospot analyzer (Cellular Technology Limited, Cleveland, OH, USA). Results for antigen-specific spot-forming cells were normalized with a negative control containing peptide-free media. All measurements were performed in duplicate or triplicate.

Single-cell suspensions were obtained from brain tissues by mechanical isolation. Extracellular myelin was removed from all brain and spinal cord suspensions using myelin removal beads (Miltenyi Biotec), according to the manufacturer’s instructions. For surface staining, cell suspensions were blocked with 1% anti-mouse CD16/CD32 (eBioscience; 93) for 20 min on ice and then stained with fluorescently labeled antibodies for murine CD4 (BD Biosciences; RM4-5), CD11c (eBioscience; N418), CD11b (eBioscience; M1/70), or GR1 (BD Biosciences; RB6-8C5), for 45 min. Following staining, cells were washed with PBS, lysed for red blood cell (Beckman Coulter), and fixed using a fixative agent (eBioscience). All samples were analyzed or sorted on a BD FACSAria II (BD Bioscience). Data were analyzed using the FlowJo 10 Data Analysis software (FlowJo LLC) and BD FACSDiva software (BD Biosciences).

Protein was extracted from whole mouse brain using the RIPA Lysis Buffer Kit (Santa Cruz Biotechnology) as per the manufacturer’s protocol. Briefly, an appropriate amount of RIPA complete lysis buffer was added to cell pellet. The mixture was incubated on ice for 5 min, followed by centrifugation at 14,000 × g for 15 min at 4°C. The supernatant was collected as brain tissue lysate and stored at −80°C until further use. Protein concentration was determined using Invitrogen EZQ Protein Quantitation Kit (Invitrogen) and Pierce BCA Protein Assays (ThermoFisher Scientific). Protein from all mice (n = 157), spanning all time points, was pooled as reference material.

Samples were obtained from a single cheek-puncture for each individual mouse. Whole blood samples were collected using BD Vacutainer SST Blood Collection Tubes (BD Worldwide; Product Number: 367381) for serum separation as described previously (20). Specimens were centrifuged at 1,100 × g for 20 min after clotting for 30 min.

Commercial ELISA Kits for synapsin-1 (ABIN426058), synapsin-2 (ABIN852365), synaptotagmin-1 (ABIN1117297), neuronal enolase (ABIN1116159), glutamine synthetase (ABIN1570168), protein deglycase DJ-1 (ABIN819944), and NEFL (ABIN427109) were used as per the manufacturer’s suggested protocol (Antibodies-online.com Inc.). For brain tissue specimens, 300–500 µg of protein per sample was added to the corresponding well in each plate. For serum specimens, 40 µL of serum and 60 µL of sample-diluent per sample were added to the corresponding well in each plate. Plates were read at OD 450 nm for absorbance on a Synergy HT microplate reader (BioTek). Data were analyzed using the BioTek Gen5 software.

Experimental autoimmune encephalomyelitis was induced, serum was collected from naive and EAE mice, and a randomly generated number assigned to the samples. The serum concentration of protein biomarker candidates was then determined by ELISA by a second investigator blinded to the origin of the samples. The serum concentration of protein biomarker candidates was then used to stratify/predict EAE versus naive animals by the second and/or third investigator.

Microwave and magnetic proteomics were performed as previously described (14, 15). For isobaric tandem mass tagging (TMT) labeling, protein was pooled from all specimens by protein amount as reference material prior to sample preparation from individual specimens. A total of 50 mg of C8 magnetic beads (BcMg; Bioclone Inc.) were suspended in 1 mL of 50% methanol. Immediately before use, 100 µL of the beads were washed three times with equilibration buffer [200 mM NaCl, 0.1% trifluoroacetic acid (TFA)]. Whole cell protein lysate (25–100 µg at 1 µg/µL) was mixed with pre-equilibrated beads and one-third sample binding buffer (800 mM NaCl, 0.4% TFA) by volume. The mixture was incubated at room temperature for 5 min followed by removing the supernatant. The beads were washed twice with 150 µL of 40 mM triethylammonium bicarbonate (TEAB), and then, 150 µL of 10 mM dithiothreitol (DTT) was added. The bead–lysate mixture underwent microwave heating for 10 s. DTT was removed and 150 µL of 50 mM iodoacetamide added, followed by a second microwave heating for 10 s. The beads were washed twice and re-suspended in 150 µL of 40 mM TEAB. In vitro proteolysis was performed with 4 µL of trypsin in a 1:25 trypsin-to-protein ratio (stock = 1 µg/µL in 50 mM acetic acid) and microwave heated for 20 s in triplicate. The supernatant was used immediately or stored at −80°C. Released tryptic peptides from digested protein lysates, including the reference materials described above, were modified at the N-terminus and at lysine residues with the TMT-6plex isobaric labeling reagents (ThermoFisher Scientific). Each individual specimen was encoded with one of the TMT-127-130 reagents, while reference material was encoded with the TMT-126 and -131 reagents: 41 µL of anhydrous acetonitrile was added to 0.8 mg of TMT labeling reagent for 25 µg of protein lysate and microwave heated for 10 s. To quench the reaction, 8 µL of 5% hydroxylamine was added to the sample at room temperature. To normalize across all specimens, TMT-encoded cell lysates from individual specimens, labeled with the TMT-127-130 reagents, were mixed with the reference material encoded with the TMT-126 and -131 reagents in 1₁₂₆:1₁₂₇:1₁₂₈:1₁₂₉:1₁₃₀:1₁₃₁ ratios. These sample mixtures, including all TMT-encoded specimens, were stored at −80°C until further use.

Capillary LC/FT/MS/MS was performed with a split less nanoLC-2D pump (Eksigent, Livermore, CA, USA), a 50 μm-i.d. column packed with 7 cm of 3 µm-o.d. C18 particles, and a hybrid linear ion trap-Fourier-transform tandem mass spectrometer (LTQ-ELITE; ThermoFisher, San Jose, CA, USA) operated with a lock mass for calibration. The reverse-phase gradient was 2–62% of 0.1% formic acid in acetonitrile over 60 min at 350 nL/min. For unbiased analysis, the top six most abundant eluting ions were fragmented by data-dependent HCD with a mass resolution of 120,000 for MS and 15,000 for MS/MS. For isobaric TMT labeling, probability-based protein database searching of MS/MS spectra against the TrEMBL protein database (release December 29, 2012; 59,862 sequences) was performed with a 10-node MASCOT cluster (v. 2.3.02; Matrix Science, London, UK) with the following search criteria: peak picking with Mascot Distiller; 10 ppm precursor ion mass tolerance, 0.8 Da product ion mass tolerance, three missed cleavages, trypsin, carbamidomethyl cysteines as a static modification, oxidized methionine and deamidated asparagine as variable modifications, an ion score threshold of 20, and TMT-6-plex for quantification.

Function, network, and pathway analysis were performed with QIAGEN ingenuity pathways analysis (IPA) according to the manufacturer’s suggestions. Briefly, quantile MASCOT results were imported to IPA and IPA’s core analysis was performed on each data set. Differentially expressed proteins from the preclinical onset time points were mapped onto diseases and functions, canonical signaling pathways, and molecular networks. Quantified proteins in each category were visualized to investigate function, pathway and molecular network enrichment during these time points, where P values for dysregulation and enrichment were assigned by IPA.

Statistical analysis for relative expression change of proteins revealed by M2 proteomics was performed as described previously (15). M2 proteomics results for each technical replicate estimate peptide expression for individual mice, encoded in sample mixtures, relative to pooled reference material from all mice, and spanning all time points. Relative peptide expression levels were transformed to log base 2 for quantile normalization. We tested for changes in relative peptide expression with post-immunization time using a linear mixed-effect model in which time was treated as a multilevel factor. We tested all the pairwise differences in relative peptide expression between all disease time points, using an unpaired, unequal variance t-test on the replicate averages. Statistical analysis was performed with R v3.0.2 (R-Project, Vienna, Austria).

Additional statistical analysis, including Student’s t-test, ANOVA, and receiver operating characteristic (ROC), were performed using GraphPad Prism 7 (GraphPad Software) and SigmaPlot 12 (Systat Software Inc.) software. Statistical significance was determined as indicated in the text and/or figure legends. Unless otherwise indicated, normal distribution was assumed for all samples. P values: *≤0.05; **≤0.01; ***≤0.001; ****≤0.0001.

Clinical signs of MOG_35–55 peptide-induced EAE in C57BL/6 mice, a standard EAE model, usually become apparent by 10–11 days post-immunization and disease peaks around days 19–22 (Figure 1A). Confirming previous studies, statistical analysis of the CNS proteome over the course of EAE by M2 proteomics revealed two subsets of protein expression waves with a statistically significant expression change specific for pre-onset (onset-leading wave; black curve) and acute disease phase (onset-lagging wave; gray curve) of EAE (Figure 1B) (14, 15). Statistical analysis revealed proteins that were significantly up- or downregulated within each time point at the onset-leading wave (Figure 1C).

To correlate the CNS protein expression waves with cellular and inflammatory events over the course of EAE, we determined T-cell cytokine production and inflammatory infiltrates at representative time points, including during the pre-onset phase of EAE (onset-leading phase) comprised induction of disease (day 0), pre-onset (days 5 and 7), and acute disease (onset-lagging phase) comprised onset (day 10), peak disease (day 20), and at remission (days 25 and 30).

In line with previously published results (15, 19), large frequencies of MOG-reactive IFN-γ and IL-17 producing Th1 and Th17 cells could be detected in secondary lymphatic tissues as early as day 5 after immunization and the numbers peaked after the onset of clinical disease (Figures 1D,E). Frequencies of pathogenic Th17 cells in the CNS mirrored the disease trajectory (Figure 1F). Analysis of CD4⁺ T cells, CD11c⁺ dendritic cells, and CD11b⁺ macrophages/microglia in the CNS by flow cytometry corresponded to inflammatory cytokine production and clinical EAE (Figures 1G–I), consistent with previous studies (21). Of note, a substantial number of CD11b⁺ (macrophages/microglia) cells and GR1⁺ cells (monocytes/neutrophils) were present at the pre-onset phase of EAE (Figures 1I,J).

Based on these results, we hypothesized that the EAE onset-leading CNS proteome expression wave may be a result of early (subclinical) CNS inflammation and could reveal candidate serum NDICP biomarkers to predict disease onset in EAE.

The CNS proteome expression profile during the pre-onset phase of EAE included nearly 800 proteins and over 2,000 unique peptides with altered expression, in agreement with our previous studies (14, 15). High-confidence serum NDICP expression waves and protein biomarker candidates for the prediction of disease onset were selected based on the following prioritization criteria: (i) Statistically significant difference (P ≤ 0.05) in expression compared to control during the preclinical (pre-onset) stage at either day 5 or day 7 post-immunization (P = 0.05 to 1.10E−43); proteins which did not achieve statistical significance during either one of these time points (day 5 or day 7) were excluded. (ii) Biological/functional relevance for neuroinflammatory and demyelinating diseases based on bioinformatic analysis of the pre-onset M2 proteomics datasets (Figure S1 in Supplementary Material), in accordance to previous studies (22). (iii) CNS tissue-specific or primary expression (i.e., NDICPs) based on bioinformatic analysis with tools provided by the EMBL-EBI expression atlas (23), MOPED (24), and ProteomicsDB (25). Ubiquitously expressed or non-primarily CNS-expressed proteins were excluded (e.g., actin, tubulin, and enzymes of glycolysis). (iv) A human homolog to increase the utility of the candidate biomarkers for studies in human MS patients.

Shown in Table 1 are seven pre-onset phase NDICP biomarker candidates for predicting clinical onset after applying the selection criteria. These NDICPs included synapsin-2 (SYN2), synaptotagmin-1 (SYT1), neurofilament light (NEFL), glutamine synthetase (GS), protein deglycase DJ-1 (PARK7), enolase-2 (ENO2), and synapsin-1 (SYN1), which was previously reported by us (15). Additionally, we employed orthogonal ELISA-based measurements to confirm the expression waves for several NDICPs from the EAE CNS proteome that were revealed with M2 proteomics. Shown in Figure 2, the M2 proteomics results (black line, bottom) had similar expression trajectory trends to quantitative protein measurements by ELISA (gray line, top) for SYN2, SYT1, and PARK7. Statistical analysis showed a strong correlation of ELISA and M2 proteomics measurements for SYT1 and PARK7 and, to a lesser extent, for SYN2 (Figure S2 in Supplementary Material). Other proteins, which also showed similar expression trends during the pre-onset phase of EAE, showed a partial correlation trend between relative abundance (M2 proteomics) and concentration (ELISA) in CNS during the acute phase of the disease (days 10–25) (Figure S3 in Supplementary Material).

In summary, these results, together with our previously published data (14, 15), suggest that NDICP expression analysis with M2 proteomics is an efficient means to discover tissue-specific expression changes during the pre-onset phase of the disease. Moreover, these results provide evidence for the utility of applying a statistical and bioinformatic prioritization strategy to M2 proteomics results to identify novel protein biomarker candidates in EAE.

To determine the utility of the NDICP biomarker candidates revealed by our analysis (Table 1) for predicting onset of clinical disease, we induced EAE and collected serum longitudinally from these animals and from matched naive control mice at multiple time points corresponding to different clinical stages of disease. We then analyzed these samples for our high-confidence NDICP serum expression waves and protein biomarker candidates with ELISA.

As shown in Figure 3A, clinical EAE became apparent by 11 days post-immunization. Importantly, we observed specific serum NDICP expression patterns in EAE mice for SYN2, GS, ENO2, and SYT1 (Figures 3B–E). Specifically, the serum levels for SYN2, GS, and ENO2 were increased at the pre-onset phase of the disease, whereas the levels for SYT1 were decreased (Figures 3B–E). Notably, while the expression of these NDICPs was most strikingly altered during the pre-onset phase of EAE, they showed a trend toward returning to baseline levels at onset and peak disease (Figures 3B–E). Serum levels of these NDICPs showed statistically significant changes from the levels of naive mice for the same time points, particularly at the pre-onset phase of EAE (Figures 3B–E). Of note, serum levels of SYN2 peaked at day 5 after induction of EAE, whereas serum levels of GS peaked at day 7 (Figures 3B,C). Serum levels of ENO2 and SYT1 at days 5 and 7 post-immunization were similar (Figures 3D,E). Shown in Figure 3F is the serum fold change for the respective proteins at each pre-onset time point from EAE mice versus baseline levels of naive mice. Figure 3F shows that the magnitude of change in serum levels was strongly affected by EAE, particularly for SYN2 and GS, which showed over 50-fold change increase. Notably, the changes in serum protein levels for SYN2, ENO2, and SYT1 during EAE correlated with the expressions changes in the CNS (Figures 2 and 3, and not shown for ENO2). In contrast, the serum levels of other NDICP biomarker candidates (Table 1), including SYN1, NEFL, and PARK7, were not significantly altered during disease (Figure S4A–C in Supplementary Material; respectively). Importantly, in naive mice, serum levels of these NDICPs remained stable over time and no significant expression changes were noted within the same individuals during the observation period (Figures 3B–E, and not shown for SYN1, NEFL, and PARK7). Taken together, our results suggest that the preclinical onset (or onset-leading) phase of EAE is characterized by specific serum NDICP expression waves as indicated by altered serum abundance of SYN2, GS, ENO2, and SYT1.

To confirm the potential of serum NDICP expression waves and protein biomarker candidates as predictive tools for EAE onset, serum was collected at various time points from EAE mice and analyzed for the abundance of GS, SYN2, ENO2, and SYT1 by an investigator blinded to the EAE results. Shown in Figure S5 in Supplementary Material are EAE scores for three independent experiments and naive control animals, where serum was collected during the pre-onset phase of disease from EAE and naive animals (indicated by arrows). Shown in Figures 4A–D are serum levels of GS, SYN2, ENO2, and SYT1, respectively, in individual mice immunized for EAE (right panels; open circles) as compared with naive animals (left panels; filled circles). Importantly, serum levels of GS, SYN2, ENO2, and SYT1 stratified with 100%, 78%, 85%, and 80% accuracy, respectively, for healthy versus EAE animals and predicted the development of clinical EAE. Additionally, serum levels of GS and SYN2 were tightly controlled in naive mice (Figures 4A,B), whereas the levels of ENO2 and SYT1 showed more biological spread across multiple experiments (Figures 4C,D).

Next, we determined the sensitivity and specificity of the serum NDICP biomarker candidates and their predictive value by calculating ROC values. This model has been widely used to assess the sensitivity/specificity and predictive values of novel biomarkers (26). Using ROC analysis, we determined the discriminating power of GS, SYN2, ENO2, and SYT1 during the pre-onset phase of EAE as predictors for disease development. The results suggested that GS had the highest probability to discriminate between animals with EAE and healthy controls when analyzing serum samples from animals with unknown disease induction status (100%; i.e., no risk), followed by SYN2 (93%), ENO2 (85%), and SYT1 (77%), respectively (Figures 4E–H; percentages reflect AUC).

Taken together, the results strongly support the notion that the disease onset-leading serum expression wave of GS, SYN2, ENO2, and SYT1 can predict EAE clinical onset and discriminate between EAE mice versus healthy control animals.

Induction of active disease in EAE models requires the injection of neuroantigens in CFA and Bordetella PTx (27). Previously, it was reported that injection of CFA and PTx in mice triggers a complex set of signals in innate immune cells (16, 28). Moreover, adjuvant-induced signals can result in an increased BBB permeability and potentially promote the leakage of NDICPs into the blood circulation (29–33). Therefore, we investigated the effect of adjuvants and infection on the abundance of EAE onset-leading NDICPs in serum. Serum was collected during the pre-onset phase of disease (days 5 and 7 post-immunization) from individual naive mice, or animals immunized for EAE with MOG_35–55:CFA, CFA alone, and PTx alone. Additionally, serum was collected from individual mice in a C. muridarum infection model to investigate whether an infectious disease model that does not directly affect the CNS could alter the serum levels of EAE onset-leading NDICPs.

As shown in Figure 5A, clinical disease became apparent by 11 days post-immunization in the EAE group (open circles), whereas CFA (closed circles) and PTx (closed triangles) immunized mice did not develop clinical EAE. Shown in Figure 5B is the bacterial burden at the indicated time points post-infection with C. muridarum (black bars) and mock (gray bars). Bacterial burdens after infection with C. muridarum were self-limiting and decreased over time as previously reported (17). Shown in Figures 5C–F are the serum levels of the EAE onset-leading NDICPs measured in these animals.

The results show that injection with CFA and PTx had negligible effects on serum levels of GS and SYN2 (Figures 5C,D). Moreover, C. muridarum infection had minor and statistically not significant effects on the serum abundance of GS and SYN2 (Figures 5C,D). Similarly, PTx injection had no significant effect on ENO2 expression. Furthermore, CFA injection or C. muridarum infection resulted in increased levels of serum ENO2, but the increase was of considerably lower magnitude as compared with that observed in EAE mice (Figure 5E). Importantly, these serum levels were significantly different in EAE mice as compared with control mice. Serum levels of SYT1 were not significantly affected by C. muridarum infection. In contrast, CFA immunization or PTx injection increased SYT1 serum expression, whereas SYT1 expression decreased during EAE (Figure 5F).

Taken together, the results show that the effect of adjuvants (i.e., CFA and PTx) on serum abundance of GS and SYN2 was negligible in our studies. Moreover, levels of GS and SYN2 were not affected by infection with C. muridarum. Similarly, the serum expression of SYT1 remained relatively unaffected by C. muridarum infection. In contrast, serum levels of ENO2 and SYT1 were affected by the CFA adjuvant in our model, and SYT1 serum levels were affected by PTx as well. Finally, the expression of ENO2 was affected by peripheral infection with C. muridarum. Nonetheless, the levels of serum NDICPs were statistically different in EAE mice compared with control animals. Thus, for translational purposes, it will be important to consider the effects of infection or other non-specific stimuli on the expression of potential biomarkers for MS.

Th17 cells are strongly associated with the pathogenesis of many autoimmune diseases including EAE and MS (34). Therefore, we determined the effect of EAE induced by adoptive transfer of Th17 cells on the serum levels of GS, SYN2, ENO2, and SYT1 in our model.

As shown in Figure 6A, MOG_35–55-reactive donor T cells predominantly produced IL-17 after in vitro culture under Th17-inducing conditions prior to adoptive transfer. Upon adoptive transfer of these Th17 cells to naive recipient mice, clinical signs of EAE became apparent by day 7 post-transfer in agreement with previous studies (Figure 6B) (35). Based on our results from the active EAE model, which showed altered serum expression levels for GS, SYN2, ENO2, and SYT1 at the pre-onset phase of disease (Figures 3 and 4), we tested the serum abundance of these proteins at the pre-onset phase of passive EAE. Serum levels of GS, SYN2, ENO2, and SYT1 were determined at regular intervals before and after adoptive transfer. Of note, increased serum abundance was observed for SYN2 and GS but not for ENO2 and SYT1 at day 4 post-adoptive transfer (Figure 6C). Specifically, the serum levels for GS and SYN2 were significantly increased longitudinally at the pre-onset phase of the disease (Figures 6D,E). Furthermore, the changes in EAE onset-leading serum NDICPs coincided with the emergence of pathogenic Th17 in the CNS at this stage (Figure 6F).

Taken together, the results show that under conditions where adjuvant-mediated signals were absent, neuroinflammation caused by myelin-specific pathogenic Th17 cells promoted expression of SYN2 and GS during the pre-onset phase of EAE. Moreover, although EAE can be induced by both Th1 and Th17 cells (36), the results suggested that Th17-mediated EAE induction resulted in distinct pathophysiological mechanisms which resulted in upregulation of SYN2 and GS, but not ENO2 and SYT1.