Аналитическая проверка множественного пептида для многоантигенного серологического профилирования вируса Эпштейн-Барр по различным клиническим образцам.

Introduction

Epstein–Barr virus (EBV) is a γ-herpesvirus that infects more than 95% of adults worldwide and persists for life as a latent infection of B cells. While most primary infections are clinically silent, EBV is the causative agent of infectious mononucleosis (IM), a recognized cofactor in lymphoid and epithelial malignancies, and has been established as a necessary but not sufficient cause for the development of multiple sclerosis (MS), with seroconversion typically preceding clinical disease by a decade or more. EBV is also associated with additional autoimmune conditions, including systemic lupus erythematosus ( 1 – 6 ).

Conventional EBV serology relies on enzyme-linked immunosorbent assays (ELISA) for a limited panel of whole-protein antigens, typically EBNA-1, viral capsid antigen (VCA), and early antigen-D (EA-D) ( 6 , 7 ). These assays effectively classify exposure and infection stage but do not resolve antibody responses across the broader EBV proteome. Consequently, questions regarding how EBV-directed humoral responses are distributed across antigens, concentrated within specific regions of individual proteins, or differ between clinical states are difficult to address using conventional formats. Because whole-protein assays aggregate antibody binding across an entire antigen into a single measurement, individuals with similar antigen-level serology may recognize different linear regions of the same protein, information that is not retained in conventional ELISA measurements ( 8 – 10 ).

Multiplexed peptide microarrays provide a complementary approach by enabling simultaneous, semi-quantitative measurement of antibody reactivity against many short linear epitopes ( 8 – 14 ). For such platforms to support clinical research, two requirements must be met: first, analytical performance such as reproducibility, sensitivity, dynamic range, and the sources of technical variance must be characterized; second, the platform must behave predictably across the conditions encountered in clinical workflows. Most blood is collected as serum, EDTA plasma, and increasingly common cell-free DNA (cfDNA) blood collection tubes that incorporate formaldehyde-releasing stabilizers for nucleic acid preservation ( 15 – 20 ). Whether such stabilizers materially affect multiplexed antibody measurement is incompletely characterized and has direct implications for studies that rely on remnant cfDNA-tube specimens.

In this work, we report the analytical validation of an EBV peptide microarray comprising 108 peptides spanning nine major EBV antigens. The novelty of this study lies in the comprehensive analytical characterization of a multi-antigen EBV peptide microarray under conditions relevant to clinical and translational research applications. By evaluating reproducibility, detection characteristics, technical sources of variation, ELISA concordance, and matrix-dependent effects across serum, K2 EDTA plasma, and Streck cfDNA plasma, this work establishes a framework for broader translational use of peptide-array-based EBV serology. Through systematic comparison of these specimen types, we demonstrate reproducible, high-resolution antibody profiling while preserving peptide-level resolution beyond that obtainable with traditional antigen-level assays. We further benchmark composite antigen scores against conventional EBV serology and demonstrate broad EBV humoral reactivity across clinically heterogeneous cohorts including infectious mononucleosis (IM), multiple sclerosis (MS), and non-MS donors. The disease-cohort analyses were included to evaluate whether the analytically characterized platform captures known features of EBV immunobiology and to generate hypotheses for future studies. More detailed disease-specific investigations, including MS-focused multivariable modeling and biomarker-development studies, will be reported separately.

Materials and methods

Study cohorts and specimen collection

A total of 329 de-identified human blood specimens were used, comprising people with MS (pwMS, n=74), non-MS EBV+ controls (n=194), infectious mononucleosis (IM, n=39), colorectal cancer (CRC, n=12), and non-CRC healthy controls (HC, n=10) ( Table 1 ). Specimens were procured from multiple commercial vendors and academic biorepositories. A matrix sub-study evaluated pre-analytical effects across three collection methods: Streck cfDNA blood collection tubes (n=22; formaldehyde-releasing stabilizer), K2 EDTA plasma tubes (n=18), and serum separator tubes (n=10). The Streck sub-cohort comprised CRC patients (n=12) and non-cancer controls (n=10) and was used to assess matrix-associated effects on IgG and IgM detection. For benchmarking against established serology, paired ELISA results for EBNA-1 IgG, VCA-p18 IgG/IgM, and EA-D IgG were obtained from a SeraCare reference cohort and a Discovery Life Sciences validation cohort. Laboratory personnel were not formally blinded to clinical group assignment, which is acknowledged as a limitation; however, array processing and feature extraction were performed using standardized workflows before group-level statistical analysis.

Peptide array design

The array comprised 108 synthetic peptides representing immunodominant regions from nine EBV proteins: EBNA-1, EBNA-2, EBNA-3, EBNA-4, EBNA-6, BMRF1 (EA-D), BLRF2 (VCA-p23), BFRF3 (VCA-p18), and BLLF1 (gp350/220), covering both latent and lytic phases ( Supplementary Table 1 ) ( 21 – 23 ). Peptides were 15-mers with a 10-residue tiling overlap and a C-terminal Gly-Lys-Cys linker for oriented surface attachment ( 8 ). Peptides were synthesized by MilliporeSigma (The Woodlands, TX) and spotted in triplicate on Schott Nexterion H functionalized slides (Schott Applied Microarrays, Tempe, AZ). Positive control features included immunodominant peptides from polio and other common vaccine-derived antigens to serve as internal benchmarks of assay performance.

Binding assay and image acquisition

Specimens were diluted 1:400 in incubation buffer and incubated with array slides overnight at 4 °C with continuous agitation. The 1:400 working dilution was selected from a five-point dilution series to maximize signal-to-threshold ratio while avoiding high-dose hook effects observed in EBV-reactivated donors at 1:200 ( Supplementary Figure 5 ). After PBST washes, antibody binding was detected with DyLight555-conjugated goat anti-human IgG (1:1, 000) and DyLight650-conjugated goat anti-human IgM (1:10, 000), incubated 1 h at room temperature in the dark. Slides were washed sequentially with PBST, PBS, and HPLC-grade water, dried under nitrogen, and scanned on an Agilent DNA Microarray Scanner at 550 nm (IgG) and 640 nm (IgM). Fluorescence intensities were extracted with Spotxel (SICASYS Software, Germersheim, Germany); per-feature mean, median, standard deviation, and CV were retained for downstream analysis.

Data processing and statistical analysis

Analyses were performed in JMP (18.0, SAS Institute, Cary, NC) and Microsoft Excel (Microsoft, Redmond, WA). Analytical performance was assessed using two QC datasets: 54 technical replicate slides of a single control specimen (RBD-22-3014) spanning eight experiments and two laboratory locations (downtown Phoenix and the SanTan facility in Chandler, AZ), and a five-point dilution series (1:200–1:3, 200) of six donors representing EBV-acute, EBV-latent, EBV-reactivated, and EBV-seronegative states. Within-block CV, between-slide CV, detection threshold (mean + 2 SD of slide-global background), limit of detection (probit regression on the dilution series at P[detect]=0.95), and minimum detectable fold-change (MDFC; estimated two ways, an empirical, distribution-free 95th percentile of |log 2 ratio| across replicate-slide pairs, and a parametric power analysis at triplicate N = 3 replicates) were computed. Variance decomposition by linear mixed model on log 2 (MFU) evaluated laboratory location, slide-manufacturing batch, and scanner-calibration contributions, with per-feature Type II F-tests and Benjamini–Hochberg FDR correction (q<0.05). Full analytical methodology, equations, and per-factor results are provided in Supplemental Data. Formal post hoc batch-correction algorithms were not applied to the primary analytical dataset. Instead, batch effects were quantified directly through variance decomposition, which identified slide-manufacturing batch as the dominant technical factor. Future studies will incorporate bridge specimens, lot-anchored normalization strategies, and prospective normalization procedures to further reduce batch-associated variability.

Composite antigen scores (CAS) were created by summing peptide-level fluorescence intensities per antigen, following established methods for aggregating peptide microarray signals against single-protein reference assays ( 8 , 11 , 12 , 21 , 23 , 24 ). Because the number of constituent peptides differs across antigens, antigen-normalized scores were also computed as the per-sample fraction of total EBV reactivity, with antigen-specific maximum composites defined as (number of peptides) × (3 replicates) × (antigen-specific 95th-percentile RFU). Group comparisons used non-parametric tests: Mann–Whitney U for two-group comparisons and Kruskal–Wallis with Dunn’s multiple comparisons for multi-cohort analyses, with Benjamini–Hochberg FDR correction. P-values are reported with 95% confidence intervals where applicable.

For peptide-level analyses, statistical comparisons were performed between study cohorts (IM, MS, and non-MS) with multiple-testing corrections applied. Heatmaps were generated using log 2 -transformed, background-adjusted peptide-isotype features and CAS values after feature-wise scaling. Rows and columns were ordered by marginal sums rather than hierarchical clustering. Under a multiplicative signal structure, marginal-sum ordering recovers the underlying order of row and column values from MFI measurements while preserving interpretable display axes. Samples were therefore displayed according to total EBV reactivity and features according to cumulative reactivity across the study population. This deterministic ordering requires no distance metric or linkage criterion and was selected to facilitate visualization of the range of antibody reactivity across subjects and features. Because marginal-sum ordering is a one-dimensional ranking approach, no disease linkage, cluster assignment, or unsupervised clustering inference is implied by the heatmap display.

For ELISA benchmarking analyses, positivity thresholds for peptide-array features and composite antigen scores (CAS) were established using the SeraCare reference cohort (n=21) and subsequently applied without modification to all available specimens with paired ELISA results. Agreement between ELISA and peptide-array classifications was summarized using true-positive (TP), false-positive (FP), false-negative (FN), and true-negative (TN) counts.

Results

Array design and analytical performance

To enable multi-antigen EBV serological profiling, we constructed a peptide microarray of 108 overlapping 15-mer peptides spanning nine EBV proteins across latent and lytic phases ( Figure 1A ; Supplementary Table 1 ). The platform supports simultaneous semi-quantitative measurement of IgG and IgM reactivity at peptide-level resolution within a single assay.

Analytical performance was characterized across 54 technical replicate slides of a single control specimen run at two laboratory locations (downtown Phoenix and the SanTan facility in Chandler, AZ) over eight experimental batches. Within-block reproducibility (median CV ≈14.1% across three spot replicates, and ≈11.1% restricted to above-threshold features) and between-slide reproducibility (median CV ≈11.6%) were consistent with manufacturer specifications for the Nexterion H hydrogel substrate ( 25 ). The minimum detectable fold-change was approximately 2-fold and robust across estimation methods: a distribution-free empirical estimate (95 th percentile of |log 2 ratio| across replicate-slide pairs) and a conventional parametric power analysis at N = 3 replicates converged on a similar ≈2-fold value, with the two estimates expected to overlap increasingly as replicate positive control data accumulate. A five-point dilution series defined a working dilution of 1:400, which preserved signal-to-threshold ratio across donor classes while avoiding high-dose hook effects observed in EBV-reactivated donors at 1:200 ( Supplementary Figure 5 ). Variance decomposition identified slide-manufacturing batch as the dominant identifiable nuisance factor; laboratory location and scanner calibration contributed minimally. Per-feature Type II F-tests confirmed no significant laboratory-location effect, while batch significantly affected a substantial fraction of detected features at q<0.05, identifying batch control as the primary focus for ongoing process refinement. Full analytical performance details, including LOD, MDFC, spatial-effect analyses, and variance decomposition, are provided in Supplementary Table 3 and Supplementary Figures 2 – S8 .

Concordance with ELISA-based serology

To benchmark the assay against established serology, peptide-array classifications were compared with ELISA calls using positivity thresholds established from the SeraCare reference cohort and subsequently applied to all available specimens with paired ELISA data ( Table 2 ). Agreement was evaluated for both individual immunodominant peptides and composite antigen scores (CAS). For EBNA-1 IgG, peptide EBNA1_426–440 demonstrated complete agreement with ELISA classifications, correctly identifying all positive and negative specimens. The corresponding EBNA-1 CAS achieved identical performance, indicating that EBNA-1 CAS classification was largely driven by this dominant peptide. CAS values for EBNA-1 IgG showed strong concordance with ELISA (r² ≈ 0.84, p < 0.0001), in close agreement with prior EBV peptide-to-protein correlations ( 18 ). Individual EBNA-1 peptides including EBNA1_426–440 correlated with ELISA at the peptide level ( Supplementary Table 2 ; Supplementary Figure 1 ). These results agree with other reports of single EBNA1 peptide correlation with ELISA ( 26 ).

Agreement with ELISA was more variable for EA-D and VCA antigens ( Figure 1B ; Table 2 ). For EA-D IgG, BMRF1_206–220 showed higher overall agreement with ELISA than the corresponding CAS, while both approaches displayed lower agreement than observed for EBNA-1. Similarly, for VCA-p18 IgG and IgM, the best-performing individual peptides achieved agreement equal to or greater than the corresponding CAS. Across all antigens evaluated, individual immunodominant peptides matched or exceeded the classification performance of the corresponding antigen-level composite scores.

The weaker correlations observed for VCA-p18 IgM and EA-D IgG likely reflect both biological and assay-format differences. IgM responses are transient and more susceptible to pre-analytical variability, whereas EA-D responses may be temporally heterogeneous and include conformational or discontinuous epitopes that are not fully represented by short linear peptides ( 21 ). Cross-vendor comparison of VCA-p18 ELISA (Diasorin and Trinity Biotech) against the array yielded comparable correlations, supporting platform robustness across ELISA reagent sources ( Supplementary Figure 1 ). The multiplexed format additionally resolved closely related VCA-p18 and VCA-p23 within a single assay without cross-interference ( Supplementary Figure 1 ). Together, these comparisons demonstrate analytical concordance with established serology while providing antigen- and peptide-level resolution beyond what conventional ELISA captures. Accordingly, peptide-level results should be interpreted as complementary to, rather than interchangeable with, whole-protein ELISA. Benchmarking was focused on EBNA-1, VCA, and EA-D because these antigens form the basis of routine clinical EBV serology. Several additional antigens represented on the peptide array do not have broadly adopted clinical ELISA comparators and were included to expand epitope-level characterization of EBV humoral responses beyond the conventional diagnostic panel.

Cross-matrix performance: serum, EDTA plasma, and streck cfDNA tubes

Clinical and translational workflows increasingly include cfDNA blood collection tubes that incorporate formaldehyde-releasing stabilizers. Because remnant samples from these tubes are an attractive source for opportunistic serological analyses, we evaluated whether array measurements remain stable across collection matrices. Across serum, K2 EDTA plasma, and Streck-stabilized plasma, total IgG reactivity to EBV antigens was comparable both at antigen-level and peptide-level ( Figures 2A, B ). IgG signals remained well above the lower limit of quantitation across all three matrices, and volcano plots identified no EBV peptides with significant Streck-associated IgG differences.

IgM signals followed a different pattern. Total IgM reactivity was significantly lower in Streck plasma than in either serum or EDTA plasma, and volcano plot analysis identified multiple EBV peptides with significantly reduced IgM signal in the Streck matrix ( Figure 2C ). IgM signals remained detectable above the IgM lower limit of quantitation (≈11.4 RFU) but were systematically attenuated relative to fresh collection. This is consistent with the known susceptibility of pentameric IgM to fixative-induced cross-linking and conformational change ( 16 , 18 ). The practical implication for users with remnant Streck-collected material is straightforward: IgG-based EBV profiling on Streck plasma is well supported, whereas IgM measurements from Streck should be interpreted with caution or omitted depending on the study question.

The array captures broad EBV humoral reactivity across clinically heterogeneous cohorts

To verify performance across clinical states, we compared antigen-level reactivity profiles in IM, MS, and non-MS donors. All three cohorts showed extensive IgG reactivity across latent and lytic antigens, with the expected hierarchical pattern of dominant EBNA, VCA-p18, and EA-D responses characteristic of EBV-experienced donors ( Figures 3A, B , 4A ; Supplementary Figure 9 ). IgM responses were lower in magnitude and more variable across individuals, consistent with the expected dominance of class-switched IgG in established EBV infection ( Figure 4B ). Across cohorts, the array consistently captured broad polyclonal EBV reactivity, confirming that the platform is not biased toward any single clinical state.

Although individual antigen-level differences between MS and non-MS donors were detectable, broad EBV-wide reactivity did not, by itself, provide a discriminator of MS from non-MS donors. Visualization of peptide- and antigen-level reactivity after marginal-sum ordering revealed substantial overlap between MS, IM, and non-MS samples, with no clear disease-associated separation evident across the dominant axis of EBV reactivity ( Figures 5A, B ). This observation is consistent with the near-universal prevalence of EBV exposure in adults: the breadth of array-wide EBV reactivity reflects exposure history rather than MS-specific signal, and disease-associated structure, if present, requires analytical approaches beyond aggregate antigen-level summaries.

As a positive control for the array’s ability to capture previously reported EBV serological patterns, we examined the EBNA-1 C-terminal region (residues 386–420), which contains sequences that were previously associated with MS ( 4 , 18 , 23 – 25 ). These sequences overlap regions implicated in molecular mimicry and cross-reactive antibody responses involving GlialCAM, CRYAB, and MBP. Tiled peptides spanning this region (EBNA1_396-410, EBNA1_401-415, and EBNA1_406-420) showed cohort-associated differences that were generally consistent with prior reports ( Figures 6A–C ), indicating that the array reproduces established EBNA-1 reactivity patterns. These observations should not be interpreted as evidence of a single, simple MS-discriminating serological signature; rather, they suggest that the platform can detect previously reported cohort-level differences within the context of a broader EBV-reactive background. A more detailed analysis of MS-associated serological patterns across additional MS cohorts, including multivariable modeling, will be reported separately.

Discussion

We report the analytical validation of a 108-peptide microarray spanning nine EBV antigens. The platform supports multi-antigen, epitope-level serological profiling with reproducibility, detection performance, and ELISA concordance consistent with established benchmarks. Within-block and between-slide CVs were within the expected range, and the approximately 2-fold minimum detectable fold-change was concordant across empirical and parametric (N = 3) estimates. Together with ELISA concordance across EBNA-1, VCA-p18, and EA-D, these results position the platform as suitable for downstream EBV research applications. Slide-manufacturing batch was the dominant identifiable technical factor and is the appropriate focus for ongoing process control; laboratory location and scanner calibration contributed minimally under the conditions tested.

The analytical performance observed in this study supports use of the platform for exploratory and translational serology applications, particularly those aimed at detecting moderate-to-large differences in antibody reactivity. Additional validation and standardization will be required before consideration for regulated or diagnostic use. An important implication of the validation is that the peptide array provides information that is complementary to, rather than a replacement for conventional ELISA. Whole-protein ELISAs are well established for assessing EBV exposure and infection stage, whereas the peptide array provides higher-resolution characterization of antibody reactivity across individual linear epitopes within and across antigens. This peptide-level profiling can identify differences in the specific linear regions recognized by antibodies that may not be apparent from whole-protein measurements alone. As a result, individuals with similar ELISA measurements may exhibit distinct peptide-level reactivity patterns despite comparable antigen-level serology. The multiplexed format preserves information regarding the specific linear regions recognized by an individual’s antibody response, information that is not captured by conventional whole-protein ELISAs. Individuals with similar antigen-level serology may exhibit distinct peptide-level reactivity patterns, enabling higher-resolution characterization of EBV humoral responses, longitudinal monitoring of antibody responses, comparative analysis of latent and lytic antigen targeting, and hypothesis generation in EBV-associated diseases.

The linear-peptide format also represents an important limitation of the platform. Because the assay measures antibody binding to short, surface-immobilized linear peptides, it does not fully represent epitopes that depend on native protein folding, glycosylation, quaternary structure, or discontinuous conformational features. This limitation may contribute to differences observed between peptide-array and whole-protein serological measurements for some antigens. Accordingly, peptide-array results should be interpreted as a higher-resolution assessment of the linear-epitope component of EBV humoral immunity rather than as a direct substitute for conventional protein-based serology.

Cross-matrix evaluation provides practical guidance for studies that incorporate mixed or remnant clinical specimens. IgG reactivity was indistinguishable across serum, EDTA plasma, and Streck cfDNA tubes, supporting use of the array on opportunistic samples drawn into liquid-biopsy collection systems. IgM was selectively attenuated in Streck plasma but remained detectable above the IgM lower limit of quantitation. The pattern is mechanistically consistent with the known susceptibility of pentameric IgM to fixative-induced cross-linking and is in line with broader pre-analytical considerations for IgM measurement ( 16 , 18 ). For investigators relying on remnant Streck-tube material, the practical recommendation is direct: IgG measurements are robust, while IgM should be interpreted with caution or omitted depending on the analytical question. This distinction is important because it allows existing biorepositories of Streck-tube samples, often used for cfDNA NGS testing, to be productively repurposed for EBV IgG serology while transparently flagging the IgM limitation.

Application of the platform to clinically heterogeneous cohorts confirmed that broad EBV humoral reactivity is detected across IM, MS, and non-MS donors. The array readily captures the dominant latent and lytic antibody responses that define EBV exposure, and it does so consistently across cohorts. The data show that array-wide EBV reactivity does not separate MS from non-MS donors by itself. Visualization of peptide- and antigen-level reactivity after marginal-sum ordering demonstrated substantial overlap in reactivity across MS, IM, and non-MS samples, with no clear disease-associated separation along the dominant axis of EBV reactivity. This observation is unsurprising given the near-universal prevalence of EBV exposure in adults, but it has direct implications for biomarker development: a high-resolution multiplexed measurement is necessary but not sufficient for extraction of MS-relevant signals, and platform-level “EBV positivity” should not be conflated with an MS-specific serological signature. The EBNA-1 C-terminal tiling data are consistent with this interpretation. Our observations recapitulate previously reported cohort-level patterns in the C-terminal region ( 23 , 27 – 29 ), but also show that these signals are embedded within a broadly EBV-reactive background and are not readily captured by aggregate antigen-level measurements alone.

The disease-cohort findings should be interpreted within the context of the study design. The IM, MS, and non-MS analyses were included as exploratory biological applications of the analytically validated platform rather than as a dedicated diagnostic investigation. Several factors may have influenced the ability to detect cohort-associated peptide patterns, including cohort size, heterogeneous specimen sources and incomplete clinical annotation. In addition, specimen matrix, age distribution, and demographic metadata were not fully balanced across cohorts. Within these limitations, only a small number of peptides showed differences between cohorts, suggesting that broad EBV seroreactivity and unsupervised clustering alone are insufficient to define a distinct MS-associated serological signature in this dataset.

The analytical validation presented here establishes reproducibility, detection characteristics, matrix effects, and the major sources of technical variation for this EBV peptide microarray. Additional studies will be needed to support broader translational applications, including implementation of standardized reference materials, lot-to-lot comparability procedures, and assessment of reproducibility across operators and laboratories. Because slide-manufacturing batch represents the largest identifiable technical factor, future studies will incorporate bridge specimens, batch-aware study designs, and normalization approaches anchored to common reference samples and control features. Such efforts will facilitate evaluation of the platform in larger multi-site studies and biomarker-development settings.

Taken together, the platform performs as a validated multi-antigen, multi-isotype EBV serology tool. Its analytical characteristics, ELISA concordance, and demonstrated compatibility with multiple specimen types including remnant Streck plasma with noted limitation, support its use for EBV research, including longitudinal studies in MS and other EBV-associated diseases.

Although the current data do not support use as a standalone MS diagnostic assay, they demonstrate the potential utility of the platform for characterizing linear-epitope reactivity, identifying candidate immunodominant regions, monitoring response breadth over time, and generating hypotheses for future biomarker investigations. Future studies will further evaluate how peptide-level EBV reactivity relates to clinical, demographic, and host-immunogenetic factors. More detailed disease-focused analyses built upon this analytical foundation are ongoing and will be reported separately.

Acknowledgments

The authors thank the specimen providers and collaborating organizations that supported access to de-identified clinical samples and acknowledge technical and operational support that contributed to peptide array fabrication.

Funding Statement

The author(s) declared that financial support was received for this work and/or its publication. The authors declare that this work received funding from Robust Diagnostics, LLC. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.

Footnotes

Edited by: Guilherme Curty Lechuga , Oswaldo Cruz Foundation (Fiocruz), Brazil

Reviewed by: Yogita Jethmalani , National Institute of Allergy and Infectious Diseases (NIH), United States

Khang Le Quy , University of Oslo, Norway

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors upon reasonable request to the corresponding author, subject to standard data use agreements. Supplemental tables provide peptide content and summary analytical metrics supporting the findings reported in this article.

Ethics statement

Ethical approval was not required for the studies involving humans because this study used de-identified remanent samples that were purchased or provided by commercial sources or collaborators. Therefore, this study was deemed exempt from additional IRB approval. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements because samples were de-identified remnant samples provided by vendors or collaborators.

Author contributions

MP: Visualization, Writing – original draft, Writing – review & editing, Formal analysis. JGLH: Formal analysis, Writing – original draft, Writing – review & editing. ZG: Formal analysis, Resources, Writing – review & editing. IR: Formal analysis, Resources, Writing – review & editing. PS: Conceptualization, Data curation, Formal analysis, Methodology, Supervision, Writing – original draft, Writing – review & editing, Visualization. CD: Conceptualization, Project administration, Supervision, Writing – original draft, Writing – review & editing, Formal analysis, Methodology.

Conflict of interest

Authors MP, JGLH, PS, and CWD were employed by Robust Diagnostics, LLC, which develops peptide microarray technologies relevant to this work. Authors ZG and IR were employed by Reference Medicine, Inc.

The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2026.1919983/full#supplementary-material

References

1. Damania B, Kenney SC, Raab-Traub N. Epstein-Barr virus: Biology and clinical disease. Cell. (2022) 185:3652–70. doi: 10.1016/j.cell.2022.08.026

2. Bjornevik K, Cortese M, Healy BC, Kuhle J, Mina MJ, Leng Y, et al. Longitudinal analysis reveals high prevalence of Epstein-Barr virus associated with multiple sclerosis. Science. (2022) 375:296–301. doi: 10.1126/science.abj8222

3. Bjornevik K, Münz C, Cohen JI, Ascherio A. Epstein-Barr virus as a leading cause of multiple sclerosis: mechanisms and implications. Nat Rev Neurol. (2023) 19:160–71. doi: 10.1038/s41582-023-00775-5

4. Jog NR, James JA. Epstein barr virus and autoimmune responses in systemic lupus erythematosus. Front Immunol. (2020) 11:623944. doi: 10.3389/fimmu.2020.623944

5. Wong Y, Meehan MT, Burrows SR, Doolan DL, Miles JJ. Estimating the global burden of Epstein-Barr virus-related cancers. J Cancer Res Clin Oncol. (2022) 148:31–46. doi: 10.1007/s00432-021-03824-y

6. Klutts JS, Ford BA, Perez NR, Gronowski AM. Evidence-based approach for interpretation of Epstein-Barr virus serological patterns. J Clin Microbiol. (2009) 47:3204–10. doi: 10.1128/jcm.00164-09

7. Prevention CfDCa . Laboratory testing for epstein-barr virus (EBV) (2024). Available online at: https://www.cdc.gov/epstein-barr/php/laboratories/index.html (Accessed May 11, 2026).

8. Stafford P, Sherma ND, Peterson M, Diehnelt CW. A peptide microarray platform approach for discovery of immunodominant antibody epitopes. Anal Chem. (2024) 96:14524–30. doi: 10.1021/acs.analchem.4c02806

9. Price JV, Tangsombatvisit S, Xu G, Yu J, Levy D, Baechler EC, et al. On silico peptide microarrays for high-resolution mapping of antibody epitopes and diverse protein-protein interactions. Nat Med. (2012) 18:1434–40. doi: 10.1038/nm.2913

10. Hecker M, Fitzner B, Wendt M, Lorenz P, Flechtner K, Steinbeck F, et al. High-density peptide microarray analysis of IgG autoantibody reactivities in serum and cerebrospinal fluid of multiple sclerosis patients. Mol Cell Proteomics. (2016) 15:1360–80. doi: 10.1074/mcp.m115.051664

11. Sykes KF, Legutki JB, Stafford P. Immunosignaturing: a critical review. Trends Biotechnol. (2013) 31:45–51. doi: 10.1016/j.tibtech.2012.10.012

12. Stafford P, Wrapp D, Johnston SA. General assessment of humoral activity in healthy humans. Mol Cell Proteomics. (2016) 15:1610–21. doi: 10.1074/mcp.m115.054601

13. Johnston SA, Domenyuk V, Gupta N, Batista MT, Lainson JC, Zhao ZG, et al. A simple platform for the rapid development of antimicrobials. Sci Rep. (2017) 7:17610. doi: 10.1038/s41598-017-17941-7

14. Rabinowitz JA, Lainson JC, Johnston SA, Diehnelt CW. Non-natural amino acid peptide microarrays to discover Ebola virus glycoprotein ligands. Chem Commun (Camb). (2018) 54:1417–20. doi: 10.1039/c7cc08242h

15. Rai AJ, Vitzthum F. Effects of preanalytical variables on peptide and protein measurements in human serum and plasma: implications for clinical proteomics. Expert Rev Proteomics. (2006) 3:409–26. doi: 10.1586/14789450.3.4.409

16. Bowen RA, Hortin GL, Csako G, Otañez OH, Remaley AT. Impact of blood collection devices on clinical chemistry assays. Clin Biochem. (2010) 43:4–25. doi: 10.1016/j.clinbiochem.2009.10.001

17. Meddeb R, Pisareva E, Thierry AR. Guidelines for the preanalytical conditions for analyzing circulating cell-free DNA. Clin Chem. (2019) 65:623–33. doi: 10.1373/clinchem.2018.298323

18. Medina Diaz I, Nocon A, Mehnert DH, Fredebohm J, Diehl F, Holtrup F. Performance of streck cfDNA blood collection tubes for liquid biopsy testing. PloS One. (2016) 11:e0166354. doi: 10.1371/journal.pone.0166354

19. Kang Q, Henry NL, Paoletti C, Jiang H, Vats P, Chinnaiyan AM, et al. Comparative analysis of circulating tumor DNA stability In K. Clin Biochem. (2016) 49:1354–60. doi: 10.1016/j.clinbiochem.2016.03.012

20. Wan JCM, Massie C, Garcia-Corbacho J, Mouliere F, Brenton JD, Caldas C, et al. Liquid biopsies come of age: towards implementation of circulating tumour DNA. Nat Rev Cancer. (2017) 17:223–38. doi: 10.1038/nrc.2017.7

21. Loebel M, Eckey M, Sotzny F, Hahn E, Bauer S, Grabowski P, et al. Serological profiling of the EBV immune response in Chronic Fatigue Syndrome using a peptide microarray. PloS One. (2017) 12:e0179124. doi: 10.1371/journal.pone.0179124

22. Cortese M, Leng Y, Bjornevik K, Mitchell M, Healy BC, Mina MJ, et al. Serologic response to the epstein-barr virus peptidome and the risk for multiple sclerosis. JAMA Neurol. (2024) 81:515–24. doi: 10.1001/jamaneurol.2024.0272

23. Vasilenko N, Tieck MP, Michel T, Schembecker S, Schwarz P, Guenther A, 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

24. Holenya P, Lange PJ, Reimer U, Woltersdorf W, Panterodt T, Glas M, et al. Peptide microarray-based analysis of antibody responses to SARS-CoV-2 identifies unique epitopes with potential for diagnostic test development. Eur J Immunol. (2021) 51:1839–49. doi: 10.1002/eji.202049101

25. SCHOTT . SCHOTT NEXTERION® Slide H Protocol Document. Tempe, AZ: SCHOTT; (2024).

26. Fachiroh J, Paramita DK, Hariwiyanto B, Harijadi A, Dahlia HL, Indrasari SR, et al. Single-assay combination of Epstein-Barr Virus (EBV) EBNA1- and viral capsid antigen-p18-derived synthetic peptides for measuring anti-EBV immunoglobulin G (IgG) and IgA antibody levels in sera from nasopharyngeal carcinoma patients: options for field screening. J Clin Microbiol. (2006) 44:1459–67. doi: 10.1128/jcm.44.4.1459-1467.2006

27. Lanz TV, Brewer RC, Ho PP, Moon JS, Jude KM, Fernandez D, et al. Clonally expanded B cells in multiple sclerosis bind EBV EBNA1 and GlialCAM. Nature. (2022) 603:321–27. doi: 10.1038/s41586-022-04432-7

28. Sattarnezhad N, Kockum I, Thomas OG, Liu Y, Ho PP, Barrett AK, et al. Antibody reactivity against EBNA1 and GlialCAM differentiates multiple sclerosis patients from healthy controls. Proc Natl Acad Sci USA. (2025) 122:e2424986122. doi: 10.1073/pnas.2424986122

29. Vietzen H, Kühner LM, Berger SM, Furlano PL, Bsteh G, Berger T, 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

Associated Data

Supplementary Materials

Data Availability Statement

The raw data supporting the conclusions of this article will be made available by the authors upon reasonable request to the corresponding author, subject to standard data use agreements. Supplemental tables provide peptide content and summary analytical metrics supporting the findings reported in this article.