S Jarius, K Ruprecht, I Metz, F B König, J Haas, W Brück, F Paul, B Wildemann
9 min
Recently, an autoantibody signature considered to be predictive of multiple sclerosis (MS) has been reported in an article by Zamecnik et al. published in Nature Medicine, which is characterized by immunoglobulin G (IgG) responses to peptides sharing the amino acid motif P-(SA)-x-(SGA)-R-(SN)-(LRKH). These results are highly important, all the more so as the same motif is present also in two proteins expressed by Epstein-Barr virus (EBV), a pathogen that likely plays a key role in MS pathogenesis. However, clinically relevant autoantibody responses often target conformational epitopes, and peptides often differ from their corresponding proteins in terms of conformation. We were therefore interested in whether these findings can be reproduced at protein level and may thus play a role also in vivo. Here, we report findings from complementary experiments employing a microarray covering nearly 10,000 human full-length proteins and using serum and cerebrospinal fluid samples from patients with a histopathologically confirmed diagnosis of MS. Our data show that prominent IgG responses to full-length proteins bearing the P-(SA)-x-(SGA)-R-(SN)-(LRKH) motif can indeed be found in a substantial proportion of MS patients, although considerable inter-patient variability exists in both the type and number of individual responses. Notably, these IgG responses were more pronounced in patients with histopathologically defined pattern II MS (which is characterized by intralesional IgG and complement deposition) and pattern III MS than in patients with pattern I MS in our study. New motif-bearing candidate antigens identified in this study include RBMY2FP, CHMP2B, SRSF8 (SFRS2B), NUS1 (NgBR, Nogo-B receptor), and RTN2. Further studies investigating the diagnostic, pathophysiological, therapeutic, and prognostic implications of this antibody signature, as well as the potential role of cross-reactivity with EBV-suggested by the presence of the motif of interest in both EBV BRRF2 and EBV envelope glycoprotein M-are warranted and may significantly advance our understanding of MS.
Recent research identified an autoantibody signature in multiple sclerosis (MS) patients characterized by reactivity to specific amino acid motifs present in both human proteins and Epstein-Barr virus (EBV). Because initial findings were based on peptide fragments—which may lack the complex conformational structures of full-length proteins—the authors sought to determine if these autoantibody responses could be validated using a microarray of full-length human proteins.
The researchers conducted an in silico re-analysis of a cohort of 19 MS patients with histopathologically confirmed disease patterns (I, II, and III). They utilized a microarray containing nearly 10,000 full-length human proteins, identifying 14 proteins that shared the motif P-(SA)-x-(SGA)-R-(SN)-(LRKH). They compared the reactivity of serum and cerebrospinal fluid samples against these proteins and investigated potential links to MS histopathology and EBV cross-reactivity.
The study successfully validated that MS patients show significant IgG responses to full-length proteins containing the motif, with approximately 58% of patients responding to at least one of the 14 identified proteins. These responses were notably more common in patients with pattern II and pattern III MS, which are often associated with antibody-mediated processes, compared to pattern I. The researchers identified several new motif-bearing candidate antigens, such as RBMY2FP and CHMP2B. Additionally, they observed that motif-bearing proteins lacking predicted B-cell epitopes were not recognized, and they highlighted instances of high-ranking responses that could suggest potential intrathecal antibody production.
This study strengthens the link between the recently proposed autoantibody signature and actual in vivo immune responses in MS. By moving beyond peptide fragments to full-length proteins, the authors demonstrate the clinical relevance of these autoantibodies and further support the hypothesis that molecular mimicry—potentially involving EBV—may play a role in the pathogenesis of MS. These results underscore the importance of conformational integrity when studying autoantigens in neuroinflammatory diseases.
Alex: How did they turn that broad screen into a test of this particular signature?
Sam: They searched the sequences of the array’s proteins for the proposed motif. Fourteen proteins matched. Some corresponded to peptides in the earlier study; others were additional candidate targets identified through this search.
Alex: A broad array can generate plenty of background binding. What distinguished a response they counted from one they set aside?
Sam: They removed targets considered nonspecific, then ranked each patient’s antibody responses using an assay-quality measure called the Z factor. It reflects signal separation relative to variability. Their strict analysis counted a response only above a cutoff of zero point eight.
Alex: Before we get to the positive responses, how large was the patient comparison?
Sam: They tested serum, the liquid portion of blood, from nineteen adults with biopsy-supported multiple sclerosis. The comparison group comprised eight adults with other inflammatory diseases of the central nervous system. They also tested a small number of cerebrospinal fluid samples, the fluid surrounding the brain and spinal cord.
Alex: Those disease controls make the comparison more informative than simply asking whether sick people differ from healthy people. What emerged under the strict threshold?
Sam: Fifty-eight percent of the multiple sclerosis patients recognized at least one motif-bearing protein. No qualifying responses appeared in the control group. But the patients differed considerably in which proteins they recognized, and how many.
Alex: That percentage counts patients, not all the protein tests. A patient could qualify through one target while failing to recognize the rest.
Sam: Yes, and no patient recognized the entire set. Some recognized several proteins; others recognized none. This is evidence for a heterogeneous antibody response, not a single uniform target shared across everyone with multiple sclerosis.
Alex: The biopsy confirmation also let them divide patients by tissue findings. Did that reveal anything beyond confirming the diagnosis?
Sam: They compared established tissue patterns, called patterns one, two, and three. Pattern two features antibody and complement deposition within lesions; complement is part of the immune system. Under the strict threshold, responses were more pronounced in patterns two and three than in pattern one.
Alex: That fits an antibody-related interpretation for pattern two, but it is not confined to that group. How stable was the apparent separation?
Sam: It depended on the threshold. The authors also used protein-specific cutoffs, with some lowered to the manufacturer-recommended zero point four. More responses qualified, including responses in most pattern-one patients, while the control samples remained negative.
Alex: So the negative controls persisted, but the impression that pattern one largely lacks this response weakened. That makes the threshold comparison essential to interpreting the subgroup story.
Sam: I would read it that way. The stricter analysis supports a narrower pattern of recognition. The more permissive analysis expands it substantially. Neither should be presented alone as the definitive frequency of this signature.
Alex: Give us one concrete target. What does a protein-level response look like in these results?
Sam: Consider the candidate protein called R-B-M-Y-two-F-P. It produced qualifying serum responses in several patients. Both multiple sclerosis cerebrospinal fluid samples also showed highly ranked antibody responses to that protein, and both were collected during acute relapse.
Alex: Finding an antibody in that fluid does not necessarily mean it was produced there. Did the blood comparison help distinguish those possibilities?
Sam: In one patient, the response ranked high in cerebrospinal fluid but very low in serum. The authors say this raises the possibility of antibody production within the central nervous system. It is a suggestive contrast, not proof, especially with so few fluid samples.
Alex: There is another distinction here. Recognizing a protein that contains the motif does not show that the antibody binds the motif itself.
Sam: That remains unresolved by this design. The study establishes recognition of motif-bearing whole proteins, but does not directly map the binding site. Likewise, the shared sequence with Epstein–Barr virus suggests possible cross-reactivity; these experiments do not demonstrate it.
Alex: Then what should we make of the viral connection? It motivates a mechanism, but it is not the mechanism tested here.
Sam: The letter highlights particularly close sequence similarity between one human target and a viral protein. That gives researchers a specific lead to investigate. But these data cannot establish that antiviral antibodies also bind the human target, or that such binding contributes to disease.
Alex: Beyond that mechanistic gap, what most limits confidence in the result?
Sam: The cohort is small, and the authors call for larger, independent confirmation. These are biopsy-supported cases, not a broad diagnostic-validation population. Also, the array lacked some proteins corresponding to the original peptide findings, so it could not test the earlier signature completely.
Alex: That leaves a useful validation step, but not a clinical performance claim. Who should read the full document, and where should they begin?
Sam: Researchers developing multiple sclerosis antibody biomarkers, or investigating antibody-related mechanisms, should read it. Start with the comparison between strict and protein-specific cutoffs, then the supplementary methods describing nonspecific-target removal and individual responses. The supplement matters because the main document is a short letter.
Alex: And what should everyone else carry away from it?
Sam: These results support the earlier peptide signature at the whole-protein level, in a small, tissue-confirmed cohort. They do not yet establish prediction, viral cross-reactivity, or disease causation. The line to remember is: whole-protein recognition strengthens the signature, but does not explain what it means.
Alex: Keep that distinction in mind when you read the next claim.