When Louis Pasteur proposed that invisible germs could cause disease, researchers began to consider whether multiple sclerosis (MS) might also have an infectious origin [1, 2]. In 1894, Pierre Marie – formerly a student of Charcot – suggested that MS was caused by an infective agent and predicted that a “vaccine of Pasteur or lymph of Koch” would soon be developed to treat it [2].
Since then, many viruses have been considered for their potential role in MS pathogenesis. Early studies examined antibodies to different viruses in people with MS and healthy individuals [3]. Among the viruses studied, infection with Epstein-Barr virus (EBV) showed a clear association with MS [3].
“About one third of the population experience EBV infection later in life, and among these individuals, roughly one third develops infectious mononucleosis. Epidemiological data suggest that EBV is a condicio sine qua non for the development of MS, but it is not sufficient on its own. So, what are the other factors?” asks Professor Christian Münz from the Department of Viral Immunobiology, Institute of Experimental Immunology, University of Zurich, “There is definitely a synergy with specific genetic backgrounds, which creates a fertile ground in which MS can develop.”
MS arises from an interplay between environmental factors and genetic susceptibility, with HLA-DRB1*15:01 representing the strongest genetic risk factor and EBV playing a major contributing role [4, 5]. Almost everyone – around 90% of people worldwide – is affected by EBV, a common herpesvirus that primarily infects B lymphocytes and triggers a strong immune response [5]. Typically asymptomatic in children, EBV infection can cause infectious mononucleosis in adolescents and adults [5].
Symptomatic infectious mononucleosis is associated with an increased risk of MS in people who are genetically predisposed to the disease [6]. A landmark longitudinal study, led by Professor Alberto Ascherio of the Harvard Medical School, analysed data from more than 10 million U.S. military personnel followed over 20 years [7]. During their service, 955 individuals developed MS. The results showed that infection with EBV increased the risk of MS 32-fold [7].
In the latest issue of Cell, three independent studies offer new mechanistic insights into how EBV infection can initiate MS pathogenesis [8].
Is EBV a Cause of MS or a Consequence?
We spoke with Professor Tobias Derfuss from the Departments of Neurology and Biomedicine at the University Hospital of Basel, to put these new findings into perspective. He tells us, “Infectious mononucleosis is characterised by swollen lymph nodes and spleen, along with fatigue and fever that can last for several weeks. This clinically apparent EBV infection clearly increases the risk of developing MS later in life. What remains unclear is how the virus contributes to that risk. One possibility is that this association is an epiphenomenon. An immune system already predisposed to MS may be less able to control EBV, leading to infectious mononucleosis. In that case, EBV infection would not drive MS risk but rather reflect an underlying immune vulnerability – essentially a form of reverse causation. An alternative view is that there is a causal link, in which EBV infection actively triggers inflammatory processes in the brain that eventually lead to MS. Support for this idea comes from the temporal sequence observed in longitudinal studies: EBV infection occurs first, followed by an increase in serum concentrations of neurofilament light (sNfL), and only later by the clinical diagnosis of MS [7].”
An increase in sNfL levels can indicate ongoing neuroaxonal degeneration and represents one of the earliest detectable pathological changes associated with MS, having been detected up to 6 years before clinical onset [9]. In the study by Ascherio and colleagues involving U.S. military personnel, individuals who were initially EBV-negative but later seroconverted to EBV, and developed MS, initially showed sNfL levels comparable to those of healthy individuals. Their NfL levels increased only after EBV infection [7]. These findings strengthen the hypothesis that EBV infection plays a causal role in MS, rather than MS predisposing individuals to EBV infection.
Professor Münz adds, “EBV appears to preferentially affect a distinct subset of B cells known as atypical B cells (ABCs), which are more frequent in women and increase with age. These cells are expanded during chronic viral infections and may be more readily affected by EBV in individuals with a genetic background that limits efficient immune control of the virus, such as those at increased risk for MS. This may help explain why EBV infection acquired later in life – when ABCs are more abundant – is associated with a higher risk of MS. Notably, ABCs develop a pro-inflammatory phenotype that is associated with enhanced migratory capacity. In experimental mouse models, these cells can cross the blood-brain barrier (BBB), produce chemokines and recruit activated T cells, leading to the formation of a local inflammatory environment within the brain.”
Recently, professor Münz and his colleagues presented a novel humanised mouse model – engineered to carry human immune cells and genes – carrying HLA-DRB1*15:01, that allowed them to simulate an infectious mononucleosis-like EBV infection in vivo [10]. Results show that EBV expands oligoclonal T-bet+CXCR3+ B cells – a type of ABCs – that can colonise submeningeal brain regions and attract activated and inflammatory T cells to the central nervous system (CNS) and thereby promote MS pathogenesis [10].
EBV Molecular Mechanisms in MS
Molecular Mimicry. One proposed molecular mechanism linking EBV to MS is molecular mimicry, in which viral proteins resemble CNS components [6].
“One of the hypotheses today is that molecular mimicry plays a key role.” Professor Derfuss says, “Some EBV proteins share structural similarities with proteins expressed in the CNS. As a result, when the immune system mounts a response against EBV, it may inadvertently also target CNS proteins. This structural homology between viral and brain proteins could cause the immune system to mistakenly attack the brain instead of the virus.”
Lanz and colleagues previously showed that the EBV nuclear antigen 1 (EBNA1) has a molecular mimicry with glial cell adhesion molecule (GlialCAM), a molecule found in myelin [6]. More recently, researchers have shown that a brain protein called anoctamin-2 (ANO2) is a common autoimmune target in MS [11]. ANO2-specific T cells have been detected in over half of individuals with MS and present similarities with T-cell responses to EBNA1. Furthermore, activating these cells in mice can exacerbate their experimental immune encephalomyelitis, an MS-like disease [11].
When B Cells Are Hijacked. Professor Derfuss offers a different possible explanation for the association between EBV and MS: “When EBV infects B cells, it can immortalise them by altering their behaviour. In our paper, we describe a cascade of events that begins with a normal immune repertoire containing some autoimmune B cells – cells that recognise the body’s own tissues, including myelin. Unlike T cells, the B-cell repertoire is not tightly regulated, and autoimmune B cells can be normally present in healthy people, where they are usually harmless. Under inflammatory conditions, these cells can migrate to the brain and encounter myelin antigens, but they then normally die following activation-induced cell death (AICD). Problems arise when such cells become infected with EBV. The virus can transform them, placing them on “autopilot”. EBV-infected B cells express latent membrane protein 1 (LMP1), which allows them to escape normal immune control and survive. A second trigger is then required, which is the disruption of the BBB. Viral infections or head trauma can cause inflammation in the brain and temporarily increase BBB permeability, allowing immune cells to enter the CNS. When EBV-infected, myelin-reactive B cells gain access to the CNS, they fail to undergo normal cell death after antigen capture, persist locally, and induce inflammation and demyelination [12].”
EBV Alters Antigen Processing in B cells. Another hypothesis is that EBV can alter how B cells process and present antigens to the immune system, in particular to CD4+T cells [13]. In individuals carrying the HLA-DR15 haplotype, EBV-infected B cells can present altered fragments of myelin basic protein (MBP) to CD4+T cells [13]. This altered antigen presentation likely contributes to pathogenic immune mechanisms in MS [13].
From Pierre Marie’s Prediction to an EBV Vaccine
More than a century after Pierre Marie predicted that MS might one day be prevented by a vaccine, developing an immunisation strategy has remained an enduring challenge.
“In the end, it would be necessary to have a vaccination strategy against EBV.” Professor Derfuss concludes, “Such a strategy could unfold in two ways. One approach would be preventive: vaccinating children against EBV and then, over time, assessing whether MS incidence is reduced in the overall population over the following 20 years. The other approach would be therapeutic, targeting individuals who are already infected to lower the number of EBV-transformed B cells and eliminate the latent EBV-infected cells. Companies are currently working on both therapeutic and prophylactic vaccines. Demonstrating that such strategies can reduce MS risk would provide the strongest evidence for a causal link between EBV and MS.”
***
Written by Stefania de Vito
Special thanks to Professor Tobias Derfuss (Departments of Neurology and Biomedicine, University Hospital of Basel; Research Centre for Clinical Neuroimmunology and Neuroscience Basel, RC2NB, University of Basel) and Professor Christian Münz (Department of Viral Immunobiology, Institute of Experimental Immunology, University of Zurich) for their insights.
References
[1] Pasteur L Science 1881; 62: 420-422.
[2] Murray TJ Multiple Sclerosis: The History of a Disease. 2005: Demos New York.
[3] Bray FB et al. Arch. Neurol. 1983; 40(7): 406 – 408.
[4] International MS Genetics Consortium et al. Nature 2011; 476(7359): 214-9.
[5] Ascherio A et al. Jama 2001; 286(24): 3083-3088.
[6] Lanz TV et al. Nature 2022; 603(7900): 321-327.
[7] Bjornevik K et al. Science 2022; 375: 296-301.
[8] Steinman L & Zamvil SS Cell 2026; 189(2): 345-347.
[9] Bjornevik K et al., JAMA Neurol. 2020; 77: 58-64.
[10] Läderach F et al. Nature 2025; 646(8083): 171-179.
[11] Thomas OG et al. Cell 2026; 189: 585-602.
[12] Kim H et al. Cell 2026; 189: 603-619.
[13] Wang J et al. Cell 2026; 189: 569-584.