EBNA1381-452 и межреактивные антителные ответные ответные действия, открывающие неизгладимые подписи в множестве склерозов.

Introduction

Multiple sclerosis (MS) is a chronic autoimmune demyelinating disease of the CNS. Both environmental and genetic factors contribute to MS pathogenesis, with infection by the Epstein-Barr virus (EBV) now considered the key environmental risk factor and a prerequisite for disease. 1,2

Substantial evidence points toward an aberrant immune control of the latent EBV infection in individuals with MS. 3,4 This is accompanied by both qualitative and quantitative changes in antibody titer against EBV nuclear antigen-1 (EBNA1) in plasma. 5 Qualitatively, the epitope repertoire differs. In healthy EBV carriers, the glycine-alanine repeat region (EBNA1 90-325 ) is the immunodominant epitope, 6 whereas patients with MS frequently mount an additional antibody response against the EBNA1 381-452 region. 7,8 This potentially reflects increased antigenic processing in poorly controlled latent EBV infection. 9 Therefore, we designate EBNA1 381-452 here as EBNA1 AE (alternative epitope). Recent findings suggest that a consequently elevated immune response to this alternative epitope may predict MS onset and differentiate MS from other conditions. 7,10 Of interest, antibodies against the EBNA1 AE region are temporarily present during infectious mononucleosis, an immunologic state defined by high EBV viral loads and the immune system actively controlling infection. 11

Antibodies against the EBNA1 AE region have the potential to cross-react with CNS autoantigens, such as glial cell adhesion molecule (GlialCAM), alpha-crystallin B (CRYAB), myelin basic protein (MBP), and anoctamin-2 (ANO2). 12-15 This molecular mimicry is an attractive mechanistic explanation for the involvement of EBV in MS. 1,2 Alternatively, an inadequately controlled population of pathogenic EBNA1-expressing EBV + B cells could initiate or drive MS, 16 in which case the EBNA1 AE antibody response could be a mere epiphenomenon. Both hypotheses are not mutually exclusive, and their relative contribution to pathogenesis may even differ in individual patients.

Recent studies indicate a high EBNA1 AE response in most of MS plasma, when measured repeatedly. However, it remains unclear whether this also reflects high cross-reactivity because studies simultaneously examining responses against EBNA1 AE and CNS autoantigens in plasma have yielded inconsistent results. 17-19

It remains unclear whether EBV or its associated immune response merely triggers disease (the “hit and run” hypothesis) or continues to play an active role throughout the disease course. 20 Although the ANO2-response in particular has been linked to disease severity, 21,22 broader evidence supporting EBNA1-cross-reactivity as a driver of disease remains limited. Specifically, information on the interrelation of EBNA1 AE and cross-reactive titers, the correlation of cross-reactive antibody titers with MS disease activity, and the presence of EBNA1-cross-reactive IgG responses in the CSF remain inconclusive or lacking.

To address these questions, we investigated IgG reactivity against four 20-mer peptides from the EBNA1 AE region and their corresponding autoantigens in plasma of patients with untreated relapse-onset MS and healthy spouse controls. We then extended this analysis to a paired serum-CSF MS cohort to determine the extent of the intrathecal IgG response. Finally, we correlated our findings with clinical characteristics.

Methods

Plasma Cohort

Patients with MS (n = 211) and healthy spouse controls (n = 63) were recruited in the University Hospitals Leuven. Peripheral blood was collected from which plasma was extracted by centrifugation (300 g, 20 minutes, 20°C) and stored at −80°C. Samples were collected between 2015 and 2025 (storage time 0–10 years, median 7 years). Inclusion criteria for patient samples were (1) confirmed diagnosis of relapsing-remitting MS according to the 2017 McDonald criteria 23 and (2) no use of disease-modifying treatment (DMT) at the time of sample collection.

Serum-CSF Cohort

Paired serum and CSF samples (n = 61) of patients with relapsing-remitting MS at the time of diagnostic lumbar puncture were retrieved from the University Hospitals Leuven biobank. The obtained samples were collected between 2013 and 2023 and stored at −20°C (storage time 1–11 years, median 4 years). Inclusion criteria were (1) confirmed relapsing-remitting MS diagnosis according to the 2017 McDonald criteria, 23 (2) no use of DMT at the time of sample collection, and (3) availability of paired serum and CSF of the same timepoint.

Clinical Data Collection

Expanded Disability Status Scale (EDSS) scores, clinical relapses, and DMT use were recorded during follow-up visits at the MS clinic by the same treating neurologist (B.D.). Paraclinical data on oligoclonal band (OCB) status, number of OCBs, CSF leukocyte count, IgG index, T2 lesions on diagnostic MRI (<10 or ≥10 T2 lesions), and no evidence of disease activity-3 status (NEDA-3, defined as no new or enlarging MRI lesions, no new relapse, and no EDSS score increase) at approximately 1 year after baseline were also collected ( Table 1 ). The MS severity score (MSSS) was calculated from EDSS score and disease duration. 24 The normalized age-related MS severity score (nARMSS) was calculated for patients with at least 2 EDSS measurements separated by a minimum of 2 years 25 The annualized relapse rate was calculated for patients with ≥2 years of follow-up.

EBV and CMV Serology

Serostatus for EBV (EBNA1 IgG and viral capsid antigen (VCA) IgG (Abnova KA1448/KA1444)) and cytomegalovirus (CMV) (Abnova KA1452) was determined using commercially available indirect ELISA kits, following the manufacturer's guidelines. Plasma and serum were diluted 1:100. Quantitative measurement in arbitrary units per milliliter (arbitrary units [AU]/mL) was performed using the standard curve provided by the manufacturer. For EBNA1-IgG, samples with optical density (OD) > 2.3 were further diluted (1:200, 1:400, or 1:800), following the manufacturer's guidelines, and AU were adjusted accordingly. Seropositivity was determined using the threshold specified by the manufacturer (13 AU/mL for EBNA1-IgG and 12 AU/mL for CMV-IgG).

Genotyping and Human Leukocyte Antigen Imputation

Individuals were genotyped using the Infinium high-throughput screening assay on Illumina Global Screening Array. PLINK v1.9 was used for standard quality control, with individual-level and variant-level quality control steps performed as detailed by Vandebergh et al. 26 Human leukocyte antigen (HLA) alleles were then imputed using the SNP2HLA package v1.0.3 and the T1DGC reference panel 27 (n = 119) or the 1000Genomes reference panel 28 (n = 27), including only individuals of non-Finnish European ancestry.

Indirect ELISA

96-well flat-bottom MaxiSorp plates (Thermo Fisher Scientific) were coated with 100 µL/well of 1 µg/mL recombinant peptide (Genscript) in bicarbonate buffer (amino acid sequences provided in eTable 1) and incubated overnight at 4°C. Plates were washed 5 times with × 1 phosphate-buffered saline -Tween 20 (PBS-T; Thermo Fisher Scientific). Blocking was performed for 4 hours at room temperature with 1% bovine serum albumin (BSA; Sigma Aldrich) in phosphate-buffered saline. Plasma and serum samples were diluted 1:100 in 1% BSA. Samples (100 µL/well, in duplicate) were incubated overnight at 4°C, followed by 5 washes with × 1 PBS-T. 100 µL of HRP-conjugated anti-human IgG secondary antibody (Thermo Fisher Scientific, 1:12.500) was added. Following incubation for 1 hours at room temperature, plates were washed 5 times and developed with 50 µL ultra-TMB substrate (tetramethylbenzidine; Thermo Fisher Scientific). The reaction was stopped after 10 minutes incubation in the dark with 50 µL of stop solution (0.16 M Sulfuric Acid; Thermo Fisher Scientific).

OD was measured at 450 nm (Biotek 800 TS microplate reader), and duplicate values were averaged. Measurements with coefficient of variation ≥ 25% between duplicates were excluded. Blanks (1% BSA only) were included on each plate. The background signal was negligible (OD ∼ 0.05) and uniform across all plates. CSF samples were measured likewise, but without dilution. Paired CSF and serum samples were tested on the same plate for each individual. CSF measurements for some antigens could only be performed if enough sample volume was available. All measured values are provided in eAppendix 1 and 2. We did not apply a fixed ELISA cutoff to define seropositivity. Instead, we report continuous IgG levels in OD and relative reactivity metrics consistent with approaches recently used. 12,17 This approach avoids arbitrary thresholds and emphasizes quantitative differences in antibody levels, which may better reflect biological variability.

Reiber Antibody Index Calculation

To quantitatively compare serum and CSF IgG reactivity, the Reiber antibody index (AI) was calculated. 29 Values ≥1.5 were considered indicative of intrathecal antibody synthesis. To ensure correct quantitative comparison of highly reactive samples, a serial dilution (60,000 pg/mL to 32.5 pg/mL) of rabbit-derived polyclonal antibodies against EBNA1 386-405 (Thermo Fisher Life Technologies, Ghent, Belgium) was added on the paired serum-CSF plates, developed with HRP-conjugated anti-rabbit IgG secondary antibody (Thermo Fisher Scientific, 1:5,000) and used to plot a four-parameter logistic standard curve. High-reactive samples exceeding the linear part of the standard curve were further diluted. Dilution-corrected AU/mL values, along with total IgG and albumin levels obtained from clinical files, were used to calculate the antibody index, using the formula AI = Q Spec /Q IgG = (IgG target-specific-CSF /IgG target-specific-serum )/( IgGtotal-CSF /IgG total-serum ) if Q IgG < Q Lim , or AI = Q Spec /Q Lim if Q IgG > Q Lim , with Q Lim being calculated as reported previously. 29,30 In patients with sufficient CSF volume available, total EBNA1-IgG (Abnova KA1448, 1:4 dilution) was measured in CSF, and the antibody index was calculated accordingly (eAppendix 3).

Statistical Analysis

All statistical analyses were performed using GraphPad Prism v10.4.0 (GraphPad Software) or R v4.2.1. Data normality was assessed using the Shapiro-Wilk test. As most data sets were not normally distributed, nonparametric tests were applied (Mann-Whitney U for 2 groups, Kruskal-Wallis with Dunn post hoc test for more than 2 groups). All p values are 2-sided and were considered significant if < 0.05. Logistic regression analysis was performed using disease status (MS vs control) as a dependent variable, using either continuous IgG levels or categorical IgG immune signatures as independent variables. The EBNA1 AE -low signature was used as a reference category. Multiple logistic regression models which included commercial EBNA1 IgG titer and sex as covariate were fitted as detailed in eTable 2. Group distributions were compared using either the χ 2 test or the Fisher exact test. Correlation analyses between IgG OD values and clinical parameters were performed using the Spearman method with correction for multiple testing according to the Benjamini-Hochberg method.

Output of all 9 IgG measurements in the plasma cohort was reduced to 2 dimensions by principal component analysis (PCA) for visualization, followed by unsupervised k -means clustering in R v4.2.1, using the prcomp() and kmeans() function, respectively (stats package v4.2.1). Before clustering, missing values were imputed with the median value for each parameter, and data were scaled and centered using the scale() function. Clustering was applied to OD values from all 9 ELISA's with k ranging from 2 to 6. The value for k was determined to provide a stable, biologically relevant clustering. For the plasma cohort, clustering at k = 4 resulted in 4 clusters of which 2 were merged because of their close resemblance. For the serum cohort, the same approach was used, except that OD values needed to be normalized to an internal control serum sample to minimize batch effects that were absent in the plasma cohort. For serum, k = 3 provided similar clusters as identified for plasma.

Standard Protocol Approvals, Registrations, and Participant Consents

Written informed consent was obtained from all participants, in accordance with the Declaration of Helsinki. The study was approved by the University Hospitals Leuven Ethics Committee (S50354, S60222 , and S59940).

Data Availability

All primary ELISA data used in this article are provided in eAppendix 1–3. Additional anonymized data not published within this article can be made available on reasonable request from a qualified investigator.

Results

EBNA1 and Cross-reactive IgG Reactivity Are Increased in MS Plasma

We collected plasma samples from a cohort of 211 patients with relapse-onset MS, who were not on DMT at the time of sampling (baseline). The cohort was age-matched with 63 healthy spouse controls, with comparable median ages at baseline (42 years). Women represented 71.6% of the MS group and 44.4% of controls. As carrying the MS risk allele HLA-DRB1*15:01 is associated with increased EBNA1 titers, 31 we evaluated carriership in individuals for whom genotype data were available (146/274, 53.3%). There was an expected higher prevalence of this haplotype in patients with MS (50/127, 39.4%, vs 5/19, 26.3% in controls- χ 2 p = 0.32). All patients had a relapsing-remitting disease course at onset, with a median disease duration of 6 years at baseline. Other characteristics are provided in Table 1 .

IgG reactivity to EBNA1 was increased in plasma of patients with MS in comparison with healthy controls ( Figure 1A ). One of the 211 patients was negative for EBNA1-IgG in plasma but was still considered EBV-seropositive based on a positive VCA-IgG test. Seven healthy controls were EBNA1-seronegative of which 3 (4.8%) tested negative for both EBNA1-IgG and VCA-IgG and were considered EBV-negative. We also tested IgG reactivity against cytomegalovirus (CMV), a herpesvirus with characteristics similar to EBV. In our cohort, no quantitative or qualitative difference was observed, with 90 of 211 (42.7%) patients and 24 of 63 healthy controls (38.0%) testing positive for CMV-IgG ( Figure 1B ).

Next, we tested IgG levels against 4 epitopes within the EBNA1 AE region, and their corresponding autoantigens: EBNA1 386-405 /GlialCAM 370-389 , EBNA1 393-412 /CRYAB 2-21 , EBNA1 409-428 /MBP 205-224 , and EBNA1 426-445 /ANO2 135-154 ( Figure 2A ). We tested GlialCAM with posttranslational phosphorylation at position 379 (GlialCAM 370-389pSer379 ) and unphosphorylated. In comparison with controls, increased reactivity against all 4 EBNA1 AE epitopes was observed in MS plasma, as well as against MBP and ANO2. For both GlialCAM peptides and CRYAB, the difference did not reach significance. However, especially for CRYAB, a strong trend ( p = 0.06) could be noted (eFigure 1). When using multiple logistic regression to correct for potential confounding factors (i.e., commercially measured EBNA1 IgG titer and sex), only the association between EBNA1 386-405 IgG and MS was not replicated ( Figure 2, B and C , eTable 2). CMV serostatus or HLA-DRB1*15:01 status did not influence peptide-specific titers in our cohort (eFigure 2 and 3). In conclusion, we found an overall increased antibody response to the EBNA1 AE region and corresponding autoantigens, supporting EBV molecular mimicry in MS peripheral blood. However, isolated antibody responses did not always reach significance in our data set.

Three Distinct EBNA1 AE IgG Signatures Are Differentially Distributed in MS

To assess the interrelation of EBNA1 AE response and cross-reactivity, we applied dimensional reduction and k -means clustering on the 9 epitope-specific plasma IgG levels to identify distinct immune signatures. Principal component analysis revealed that the first principal component was primarily influenced by cross-reactive IgG levels, while the second principal component was defined primarily by EBNA1 386-405 IgG and EBNA1 393-412 IgG levels (eFigure 4). Clustering allowed the identification of 3 immune signatures (Methods), with distinct EBNA1 AE and cross-reactive epitope-specificity profiles capturing the separation in the first 2 PCs: (1) EBNA1 AE -high cross-reactive (12.4%), (2) EBNA1 AE -high selective (17.2%), and (3) EBNA1 AE -low (70.4%) ( Figure 3A ). The EBNA1 AE -high cross-reactive signature exhibited higher IgG levels against all measured EBNA1 AE epitopes, and their respective autoantigens compared with the EBNA1 AE -low signature. By contrast, the EBNA1 AE -high selective signature was characterized by an increased anti-EBNA1 386-405 and EBNA1 393-412 IgG response that were also markedly higher than in the EBNA1 AE -high cross-reactive signature ( Figure 3B , eFigure 5). The EBNA1 AE -high cross-reactive immune signature was more prevalent in patients with MS (32/211–15%) than in healthy controls (2/63–3%) ( Figure 3C ). This association persisted in a multiple logistic regression corrected for confounding factors EBNA1 IgG titer and sex (odds ratio 5.250). This was not the case for the EBNA1 AE -high selective signature (eTable 2, Figure 3D ). In conclusion, we found that a high-antibody response against the EBNA1 AE region is not always accompanied by a cross-reactive antibody response. Only when this is the case, the response is significantly enriched in MS plasma.

EBNA1 AE Response Is Increased After a Recent Relapse

Next, we examined whether EBNA1 AE and cross-reactive responses were correlated with clinical parameters. When comparing IgG levels in patients who had recently experienced a clinical relapse (n = 48) with those who did not, we observed higher reactivity toward selected EBNA1 AE reactive responses, reaching significance for EBNA1 386-405 and EBNA1 426-445 ( Figure 4, A and B , eFigure 6A). No difference was noted for commercially measured EBNA1 IgG titer (eFigure 6B). For cross-reactive IgG levels, the trend was generally less clear (eFigure 6C). At the level of the previously described immune signatures, this translated into a nonsignificant trend toward a more frequent presence of EBNA1 AE -high selective and EBNA1 AE -high cross-reactive immune signatures in these patients ( Figure 4C ). Immune signatures were not related to subsequent disease activity within 1 year after baseline or to any other marker of disease severity, including T2 lesion load at diagnostic MRI, EDSS at last clinical follow-up, MSSS, nARMSS, and DMT response (eFigure 7), nor were IgG titers (eTable 3). In conclusion, our findings show increased reactivity against selected EBNA1 AE epitopes after recent clinical disease activity.

Plasma Immune Signatures Are Identified in Serum of a Paired Serum-CSF Cohort

To expand our focus toward the CNS compartment, a cohort of 61 paired serum and CSF samples from patients with MS at diagnostic lumbar puncture was included. This second cohort consisted of patients with untreated relapse-onset MS (65.6% female) with a median age of 30 years. In this diagnostic setting, most patients (80.3%) experienced a relapse in the 3 months before sample collection. All patients were EBNA1-seropositive, and 27 (44.2%) were also seropositive for CMV. A more detailed characterization can be found in Table 1 . The 2 cohorts were largely independent. 20 serum samples (32.8%) were obtained from the same patient, but at a different timepoint (median interval of 8.5 months, range from 2.4 to 62.6 months), while only 10 serum samples (16.4%) were also obtained at approximately the same timepoint as the plasma sample.

PCA and clustering of the 9 epitope-specific IgG levels in serum re-established 3 immune signatures that closely mirrored the epitope-specificity profiles observed in plasma: EBNA1 AE -high cross-reactive (34%), EBNA1 AE -high selective (33%), and EBNA1 AE -low (33%) (eFigure 8, A and B), confirming consistency across sample types and partially independent cohorts. The difference in composition observed relative to the plasma cohort might be explained by the more frequent sampling postrelapse in the serum cohort. In line with the plasma cohort, EBNA1 AE -high signatures in serum were numerically more frequent in patients sampled within 3 months after relapse, although not statistically significant (eFigure 8C). In conclusion, in serum of patients with MS, taken at diagnosis, and thus usually shortly after a relapse, we find similar immune signatures, with a higher proportion of EBNA1 AE -high-immune signatures.

IgGs Associated With the EBNA1 AE -High Cross-reactive Immune Signature Are Intrathecally Produced

To assess whether the peripheral immune signatures are also reflected within the CNS, we quantified IgG levels in CSF that was collected simultaneously with the serum sample. We observed that the IgG response in CSF did not always reflect the serum profile (eFigure 9).

This divergence between serum and CSF IgG values prompted us to investigate intrathecal production, using the Reiber antibody index. 29,32 This index provides an estimate of the relative abundancy of a specific IgG in the CSF, accounting for total IgG production and blood-CSF barrier integrity. Previous work has shown that for IgG against the total EBNA1 protein, only a fraction of patients with MS shows signs of intrathecal production (i.e., antibody index ≥1.5). 32,33 For EBNA1 386-405 and EBNA1 393-412 , a comparable pattern with general EBNA1 IgG was found, suggesting that for most patients, increased CSF IgG against these epitopes mainly derives from the periphery. By contrast, the antibody index for other antigens, especially autoantigens, was higher, indicating relatively more intrathecal production ( Figure 5 ). Correlating intrathecal measurements with clinical activity, CSF obtained after a recent relapse showed higher EBNA1 386-405 IgG reactivity ( p = 0.03). Yet, a similar (although nonsignificant) increase in serum IgG and similar antibody indices indicate these antibodies mainly derive from the periphery rather than intrathecal production ( Figure 6 ). No differences were observed for other antigens (eFigure 10).

A higher antibody index for some CNS epitopes and EBNA1 426-445 was associated with recurrent disease activity within 1 year (eFigure 11A). Moreover, 4 patients with mild disease activity not requiring DMT had lower antibody indices than patients who did need DMT (eFigure 11B). Despite the small sample size of this subgroup analysis, these results may suggest an association between intrathecal cross-reactive IgG production and disease activity. However, no association with T2 lesion load at diagnosis was observed (eFigure 11C). In conclusion, we found that CSF immune profiles do not always mirror the peripheral blood, with antibody indices indicative of intrathecal production for antibodies related to EBNA1 AE cross-reactivity.

Discussion

In this study, we aimed to provide more clarity on the potential role of EBNA1 AE molecular mimicry in the MS disease process.

First, we describe the presence and concordance between EBNA1 AE and cross-reactive responses in our MS and control cohorts. Our study confirms the elevated EBNA1 AE -specific IgG response in MS plasma. 7,10,17,19 Of interest, for the response to the most N-terminal peptide (EBNA1 386-405 ), we show that this effect might be driven by its correlation with the overall elevated EBNA1 antibody response. This indicates the added value of correcting for the total EBNA1 titer in future investigations. Our results further corroborate the increased IgG reactivity against ANO2 and MBP in the plasma of patients with MS. At the peptide level, we did not observe a significant association of GlialCAM or CRYAB reactivity alone with MS. This contrasts with previous studies, 12,19 but is consistent with another study with a moderate sized cohort. 18 This discrepancy may stem from differences in antibody detection method, sample size, or cohort characteristics (e.g., HLA-DRB1*15:01 carrier status). To assess interrelations between individual titers, we integrated antibody responses in immune signatures, after which we identified a distinct MS-specific association that comprised all cross-reactive epitopes, including GlialCAM and CRYAB. We identified 3 distinct reactivity patterns. As expected, there were individuals showing low IgG reactivity against the EBNA1 AE region and its corresponding autoantigens (EBNA1 AE -low). The EBNA1 AE -high cross-reactive signature exhibited high IgG responses against the entire EBNA1 AE region and its cross-reactive counterparts. Of interest, a third pattern (EBNA1 AE -high selective) displayed a high response against EBNA1 386-405 and EBNA1 393-412 (regions that are closest to the N-terminal GA-repeat region), but a low response against the other EBNA1 epitopes and autoantigens, including GlialCAM 370–389 and CRYAB 2-21 , the putative cross-reactive counterparts of the viral epitopes. We found the EBNA1 AE -high cross-reactive signature highly enriched in patients with MS, though not exclusively. This enrichment is consistent with the recently described association of both cross-reactive and EBNA1 AE responses with MS. 19 Our findings indicate that a high EBNA1 AE response does not necessarily mean that an equally high cross-reactive response is present, and that the response against EBNA1 AE might harbor mechanistically relevant heterogeneity. The expansion of the EBNA1 AE response to the C-terminal region and high cross-reactivity might be relevant to the MS disease process. Finally, we emphasize the use of relative measures in our study. The samples assigned in non–cross-reactive immune signatures might still harbor such responses, albeit at a lower extent.

The selective signature displays EBNA1 386-405 -reactive and EBNA1 393-412 -reactive IgG levels that were markedly higher compared with the cross-reactive phenotype. We hypothesize that an increased response toward the more N-terminal region of EBNA1 AE could prevent epitope spreading to the C-terminal region. This could be driven by more efficient antigen clearance or possibly by antibody-mediated masking of the adjacent epitopes. 34 However, higher antibody titers could also lead to increased uptake of antigen, resulting in increased processing and presentation of these antigens. Further functional immunologic studies are warranted especially because the C-terminal portion of our region of interest lies closest to the DNA-binding region, an attribute commonly associated with autoimmunity, e.g., in systemic lupus erythematosus. 35

Next, we observed an increase in a subset of EBNA1 AE -reactive IgGs in patients who recently experienced a clinical relapse. This finding might indicate a more active role for the EBNA1 AE response in driving MS disease activity. However, our cross-sectional study does not exclude the possibility that this finding is a mere epiphenomenon of general immune activation. This element, together with the recent observation that EBNA1 AE responses may fluctuate over time within 1 individual, 7,10 warrants longitudinal studies of the EBNA1 AE IgG response, in particular investigating its correlation with MS relapse activity.

Finally, we assessed intrathecal antibody production. Intrathecal immunoglobulin production is a well-known hallmark of MS, as illustrated by the presence of CSF oligoclonal bands, an increased IgG index and a polyspecific intrathecal IgG response (Measles-Rubella-Varicella Zoster [MRZ] reaction). 36 This intrathecal production of antiviral antibodies is believed to be a nonspecific representation of B-cell activation within the CNS. 37 It has been established that EBNA1-directed antibodies are paradoxically not part of this intrathecal response. 32,38 Our data confirm that for the routinely measured anti-EBNA1 response, no major intrathecal increase is observed and that IgGs against the N-terminal region of EBNA1 AE (EBNA1 386-412 ) behave similarly. However, IgGs against autoantigens and IgGs against the C-terminal region of EBNA1 AE (EBNA1 409-445 ) seem to be more often intrathecally produced. It should thus be noted that the antibody index for autoantigens is usually higher than the index for their corresponding EBNA1 antigens. This might indicate that at least in a subset of patients, an intrathecal shift occurs from a selective toward a broader, cross-reactive response. Potential chronic stimulation with EBNA1-antigen in the CNS compartment that might be a driver for this shift is a topic of active investigation. 1,39 Whether our CSF findings, including the potential association between cross-reactive intrathecal antibody production and subsequent disease activity, are more than a polyspecific bystander effect, similar to the MRZ reaction, is subject for further research.

Our study has several limitations. First, our cross-sectional study design precluded assessment of longitudinal changes in EBNA1 AE -directed IgG responses. Second, although well defined, our single-center cohort was relatively modest in size as opposed to large multicenter studies, especially in subgroup analyses (e.g., when comparing clinical characteristics). Our cohort was not perfectly balanced for sex differences, and HLA status was only known for part of the cohort. This might have introduced a confounding effect, which we attempted to mitigate by including confounders as covariates in logistic regression analyses. However, we stress the need for validation of the findings described here in larger, HLA-matched, independent populations. Third, we could not investigate the intrathecal immune response in a control population. Although our study establishes associations between EBNA1 AE immune signatures and MS clinical features, functional studies dissecting the precise mechanisms of cross-reactivity remain warranted to understand the pathogenicity of the humoral responses described here. Since EBNA1 and the examined autoantigens represent intracellular epitopes, elevated IgG titers more likely reflect underlying T-cell responses, as has been demonstrated for GlialCAM, ANO2, and CRYAB. 12,15,22 As the study design did not allow to investigate EBV viral load or the presence of EBV + B cells in our samples, the EBNA1 AE -reactive IgG responses described here could still be an epiphenomenon of a latent EBV-infected pathogenic B-cell population. 16,40 Nonetheless, our comprehensive evaluation of the EBNA1 AE response in both peripheral blood and CSF in combination with the clinical characterization of our cohort provides a valuable multidimensional view of EBNA1-related cross-reactivity in MS.

In conclusion, we identified distinct peripheral blood humoral immune signatures against the EBNA1 AE region. We observed that high EBNA1 AE responses are not necessarily associated with high-IgG responses against autoantigens and that only the cross-reactive immune signature was significantly enriched in patients with MS. We observed an increased IgG response against the EBNA1 AE region in the peripheral blood and CSF of patients who recently experienced a clinical relapse. IgG levels in CSF did not always mirror the peripheral blood, and cross-reactive IgGs were relatively overrepresented in CSF, raising the possibility of an intrathecal shift toward a more cross-reactive phenotype. Our findings contribute to the accumulating evidence of a role for EBNA1 AE molecular mimicry in MS pathology.

Acknowledgment

The authors thank Koen Poesen for his input on the ELISA setup, Patrick Dupont and Yang Feng for their assistance with analysis, Katleen Clysters and Cindy Thys for their help with sample collection, and all patients and their families for their willingness to participate in this study.

Author Contributions

D. De Wit: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. J. Beliën: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. M. David: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data. S. Swinnen: drafting/revision of the manuscript for content, including medical writing for content; study concept or design. N. Dedoncker: major role in the acquisition of data. L. Yshii: drafting/revision of the manuscript for content, including medical writing for content; study concept or design; analysis or interpretation of data. B. Dubois: drafting/revision of the manuscript for content, including medical writing for content; analysis or interpretation of data.

Study Funding

This work was funded by the Belgian Charcot Foundation (2023-0007)—Fondation Jean et Marie Henkens-Lecloux. L.Yshii is funded by KU Leuven Starting Grant (22/019), KU Leuven C1 (C14/24/148). B. Dubois is supported by KU Leuven Bijzonder Onderzoeks Fonds—Fundamenteel Klinisch Onderzoeker and D. De Wit (11P8K24N), J.B. (11A0523N, 11A0525N), and M.D. (11F9623N, 11F9625N) by PhD Fellowships of the Research Foundation Flanders (FWO).

Disclosure

The authors report no relevant disclosures. Go to Neurology.org/NN for full disclosures.

Glossary

antibody index

arbitrary unit

cytomegalovirus

alpha-crystallin B

disease -modifying treatment

Epstein-Barr virus Nuclear Antigen 1

Epstein–-Barr virus

Expanded Disability Status Scale

glial cell adhesion molecule

Human leukocyte antigen

high-throughput screening

myelin basic protein

multiple sclerosis

MS severity score

normalized age-related MS severity score

no evidence of disease activity

oligoclonal band

optical density

phosphate-buffered saline

principal component analysis

viral capsid antigen

References

1. Soldan SS , Lieberman PM . Epstein-barr virus and multiple sclerosis . Nat Rev Microbiol. 2023 ; 21 ( 1 ): 51 - 64 . doi: 10.1038/s41579-022-00770-5 35931816 PMC9362539

2. Thomas OG , Rickinson A , Palendira U . Epstein-barr virus and multiple sclerosis: moving from questions of association to questions of mechanism . Clin Transl Immunol. 2023 ; 12 ( 5 ): e1451 . doi: 10.1002/cti2.1451 PMC10191779 37206956

3. Laderach F , Munz C . Altered immune response to the epstein-barr virus as a prerequisite for multiple sclerosis . Cell. 2022 ; 11 ( 17 ): 2757 . doi: 10.3390/cells11172757 PMC9454695 36078165

4. Schneider-Hohendorf T , Gerdes LA , Pignolet B , et al . Broader epstein-barr virus-specific T cell receptor repertoire in patients with multiple sclerosis . J Exp Med. 2022 ; 219 ( 11 ): e20220650 . doi: 10.1084/jem.20220650 36048016 PMC9437111

5. Munger KL , Levin LI , O'Reilly EJ , Falk KI , Ascherio A . Anti-epstein-barr virus antibodies as serological markers of multiple sclerosis: a prospective study among United States military personnel . Mult Scler. 2011 ; 17 ( 10 ): 1185 - 1193 . doi: 10.1177/1352458511408991 21685232 PMC3179777

6. Rumpold H , Rhodes GH , Bloch PL , Carson DA , Vaughan JH . The glycine-alanine repeating region is the major epitope of the epstein-barr nuclear antigen-1 (EBNA-1) . J Immunol. 1987 ; 138 ( 2 ): 593 - 599 . 2432130

7. Vietzen H , Kuhner LM , Berger SM , et al . Early identification of individuals at risk for multiple sclerosis by quantification of EBNA-1(381-452)-specific antibody titers . Nat Commun. 2025 ; 16 ( 1 ): 6416 . doi: 10.1038/s41467-025-61751-9 40659624 PMC12259891

8. Mechelli R , Anderson J , Vittori D , et al . Epstein-barr virus nuclear antigen-1 B-cell epitopes in multiple sclerosis twins . Mult Scler. 2011 ; 17 ( 11 ): 1290 - 1294 . doi: 10.1177/1352458511410515 21757535

9. Munz C . Neuroinflammation as a result of non-neurotropic herpesvirus infection . Immunol Cell Biol. 2026 ; 104 ( 5 ): 443 - 451 . doi: 10.1111/imcb.70108 41913941 PMC13155061

10. Vietzen H , Kuhner LM , Berger SM , et al . Epstein-barr virus antibodies to differentiate multiple sclerosis from other neuroinflammatory diseases . JAMA Neurol. 2026 ; 83 ( 5 ): 461 - 470 . doi: 10.1001/jamaneurol.2026.0240 41801194 PMC12973216

11. McClain MT , Rapp EC , Harley JB , James JA . Infectious mononucleosis patients temporarily recognize a unique, cross-reactive epitope of epstein-barr virus nuclear antigen-1 . J Med Virol. 2003 ; 70 ( 2 ): 253 - 257 . doi: 10.1002/jmv.10385 12696112

12. Lanz TV , Brewer RC , Ho PP , et al . Clonally expanded B cells in multiple sclerosis bind EBV EBNA1 and GlialCAM . Nature. 2022 ; 603 ( 7900 ): 321 - 327 . doi: 10.1038/s41586-022-04432-7 35073561 PMC9382663

13. Jog NR , McClain MT , Heinlen LD , et al . Epstein barr virus nuclear antigen 1 (EBNA-1) peptides recognized by adult multiple sclerosis patient sera induce neurologic symptoms in a murine model . J Autoimmun. 2020 ; 106 : 102332 . doi: 10.1016/j.jaut.2019.102332 31515129 PMC6930324

14. Tengvall K , Huang J , Hellstrom C , et al . Molecular mimicry between anoctamin 2 and epstein-barr virus nuclear antigen 1 associates with multiple sclerosis risk . Proc Natl Acad Sci U S A. 2019 ; 116 ( 34 ): 16955 - 16960 . doi: 10.1073/pnas.1902623116 31375628 PMC6708327

15. Thomas OG , Bronge M , Tengvall K , et al . Cross-reactive EBNA1 immunity targets alpha-crystallin B and is associated with multiple sclerosis . Sci Adv. 2023 ; 9 ( 20 ): eadg3032 . doi: 10.1126/sciadv.adg3032 37196088 PMC10191428

16. Laderach F , Piteros I , Fennell E , et al . EBV induces CNS homing of B cells attracting inflammatory T cells . Nature. 2025 ; 646 ( 8083 ): 171 - 179 . doi: 10.1038/s41586-025-09378-0 40770101

17. Vietzen H , Kuhner LM , Berger SM , et al . Accumulation of epstein-barr virus-induced cross-reactive immune responses is associated with multiple sclerosis . J Clin Invest. 2024 ; 134 ( 21 ): e184481 . doi: 10.1172/JCI184481 39484720 PMC11527435

18. Vasilenko N , Tieck MP , Michel T , et al . In-depth analysis of serum antibodies against epstein-barr virus lifecycle proteins, and EBNA1, ANO2, GlialCAM and CRYAB peptides in patients with multiple sclerosis . Front Immunol. 2024 ; 15 : 1487523 . doi: 10.3389/fimmu.2024.1487523 39742283 PMC11685087

19. Sattarnezhad N , Kockum I , Thomas OG , et al . Antibody reactivity against EBNA1 and GlialCAM differentiates multiple sclerosis patients from healthy controls . Proc Natl Acad Sci U S A. 2025 ; 122 ( 11 ): e2424986122 . doi: 10.1073/pnas.2424986122 40063790 PMC11929495

20. Wahbeh F , Sabatino JJ . Epstein-barr virus in multiple sclerosis: past, present, and future . Neurol Neuroimmunol Neuroinflamm. 2025 ; 12 ( 6 ): e200460 . doi: 10.1212/NXI.0000000000200460 40953326 PMC12439688

21. Jons D , Grut V , Bergstrom T , et al . Seroreactivity against lytic, latent and possible cross-reactive EBV antigens appears on average 10 years before MS induced preclinical neuroaxonal damage . J Neurol Neurosurg Psychiatry. 2024 ; 95 ( 4 ): 325 - 332 . doi: 10.1136/jnnp-2023-331868 37802637 PMC10958269

22. Thomas OG , Rykaczewska U , Galesic M , et al . Anoctamin-2-specific T cells link epstein-barr virus to multiple sclerosis . Cell. 2026 ; 189 ( 2 ): 585 - 602 e38 . doi: 10.1016/j.cell.2025.12.032 41534529

23. Thompson AJ , Banwell BL , Barkhof F , et al . Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria . Lancet Neurol. 2018 ; 17 ( 2 ): 162 - 173 . doi: 10.1016/S1474-4422(17)30470-2 29275977

24. Roxburgh RH , Seaman SR , Masterman T , et al . Multiple sclerosis severity score: using disability and disease duration to rate disease severity . Neurology. 2005 ; 64 ( 7 ): 1144 - 1151 . doi: 10.1212/01.WNL.0000156155.19270.F8 15824338

25. Manouchehrinia A , Kingwell E , Zhu F , Tremlett H , Hillert J , Ramanujam R . A multiple sclerosis disease progression measure based on cumulative disability . Mult Scler. 2021 ; 27 ( 12 ): 1875 - 1883 . doi: 10.1177/1352458520988632 33487091 PMC8521354

26. Vandebergh M , Andlauer TFM , Zhou Y , et al . Genetic variation in WNT9B increases relapse hazard in multiple sclerosis . Ann Neurol. 2021 ; 89 ( 5 ): 884 - 894 . doi: 10.1002/ana.26061 33704824 PMC8252032

27. Rich S . Data From: T1DGC HLA Reference Panel for Imputation with SNP2HLA (T1DGC-Special) . NIDDK Central Repository ; 2015 . doi: 10.58020/63vn-7d44

28. Abi-Rached L , Gouret P , Yeh JH , et al . Immune diversity sheds light on missing variation in worldwide genetic diversity panels . PLoS One. 2018 ; 13 ( 10 ): e0206512 . doi: 10.1371/journal.pone.0206512 30365549 PMC6203392

29. Reiber H , Lange P . Quantification of virus-specific antibodies in cerebrospinal fluid and serum: sensitive and specific detection of antibody synthesis in brain . Clin Chem. 1991 ; 37 ( 7 ): 1153 - 1160 . 1855284

30. Brettschneider J , Czerwoniak A , Senel M , et al . The chemokine CXCL13 is a prognostic marker in clinically isolated syndrome (CIS) . PLoS One. 2010 ; 5 ( 8 ): e11986 . doi: 10.1371/journal.pone.0011986 20700489 PMC2916843

31. Huang J , Tengvall K , Lima IB , et al . Genetics of immune response to epstein-barr virus: prospects for multiple sclerosis pathogenesis . Brain. 2024 ; 147 ( 10 ): 3573 - 3582 . doi: 10.1093/brain/awae110 38630618 PMC11449136

32. Pache F , Otto C , Wilken D , et al . Broad analysis of serum and intrathecal antimicrobial antibodies in multiple sclerosis underscores unique role of epstein-barr virus . Neurol Neuroimmunol Neuroinflamm. 2025 ; 12 ( 1 ): e200332 . doi: 10.1212/NXI.0000000000200332 39602676 PMC11616972

33. Ali R , Trier NH , Houen G , Frederiksen JL . Epstein-barr nuclear antigen 1 antibody-based indices are increased in patients with multiple sclerosis . Mult Scler Relat Disord. 2024 ; 92 : 106173 . doi: 10.1016/j.msard.2024.106173 39577298

34. Zarnitsyna VI , Lavine J , Ellebedy A , Ahmed R , Antia R . Multi-epitope models explain how pre-existing antibodies affect the generation of broadly protective responses to influenza . Plos Pathog. 2016 ; 12 ( 6 ): e1005692 . doi: 10.1371/journal.ppat.1005692 27336297 PMC4918916

35. Han S , Zhuang H , Shumyak S , Yang L , Reeves WH . Mechanisms of autoantibody production in systemic lupus erythematosus . Front Immunol. 2015 ; 6 : 228 . doi: 10.3389/fimmu.2015.00228 26029213 PMC4429614

36. Stangel M , Fredrikson S , Meinl E , Petzold A , Stuve O , Tumani H . The utility of cerebrospinal fluid analysis in patients with multiple sclerosis . Nat Rev Neurol. 2013 ; 9 ( 5 ): 267 - 276 . doi: 10.1038/nrneurol.2013.41 23528543

37. Jarius S , Eichhorn P , Franciotta D , et al . The MRZ reaction as a highly specific marker of multiple sclerosis: re-evaluation and structured review of the literature . J Neurol. 2017 ; 264 ( 3 ): 453 - 466 . doi: 10.1007/s00415-016-8360-4 28005176

38. Ruprecht K , Wildemann B , Jarius S . Low intrathecal antibody production despite high seroprevalence of epstein-barr virus in multiple sclerosis: a review of the literature . J Neurol. 2018 ; 265 ( 2 ): 239 - 252 . doi: 10.1007/s00415-017-8656-z 29098417

39. Orr N , Steinman L . Epstein-barr virus and the immune microenvironment in multiple sclerosis: insights from high-dimensional brain tissue imaging . Proc Natl Acad Sci U S A. 2025 ; 122 ( 11 ): e2425670122 . doi: 10.1073/pnas.2425670122 40063794 PMC11929469

40. Munz C . Altered EBV specific immune control in multiple sclerosis . J Neuroimmunol 2024 ; 390 : 578343 . doi. 10.1016/j.jneuroim.2024.578343 38615370