Infection with Epstein-Barr virus (EBV) increased the risk of subsequently developing multiple sclerosis (MS) by 32-fold in young adults, but produced no increase in risk after infection with other similarly transmitted viruses, according to a longitudinal analysis of more than 10 million US military personnel published in Science in 2022. Together with a separate 2018 Nature Genetics study showing that a viral protein made by EBV binds to nearly half of the human DNA regions associated with lupus risk, the findings position EBV, a virus carried by about 95% of adults, as a leading candidate cause of at least two major autoimmune diseases.
Why the link has been hard to prove
What EBV does in the body
EBV is one of the most common viruses in humans. Roughly 95% of adults worldwide carry a lifelong latent infection, most acquired in childhood or adolescence, and most carriers experience only mild illness (or none) at the time of primary infection. That near-universal prevalence has been the central methodological challenge in linking EBV to any specific disease: if almost everyone is infected, showing that infection precedes and causes a specific illness requires either finding a rare uninfected cohort large enough to compare against infected people, or catching individuals during the narrow window when they seroconvert from negative to positive. Both approaches were unavailable until very large longitudinally sampled cohorts became accessible.
The 2022 Science study on MS
The MS evidence
Bjornevik and colleagues, led by Alberto Ascherio at the Harvard T.H. Chan School of Public Health and working with the US Department of Defense Serum Repository, analyzed serum samples from more than 10 million active-duty US military personnel enrolled between 1993 and 2013. Military personnel are screened for HIV every two years, and the residual serum is banked, providing a longitudinal biobank rarely available in civilian populations. Of the cohort, 955 personnel were diagnosed with MS during their service, and 801 had sufficient stored samples for EBV serostatus analysis.
Of the 801 MS cases with adequate samples, only 35 were EBV-negative at their first serum sample. Of those 35, 34 seroconverted to EBV-positive before developing MS, a 97% seroconversion rate. By contrast, only about 57% of a matched control group of personnel who did not develop MS seroconverted during the same period. On formal analysis, the risk of MS increased 32-fold after EBV infection, but was unchanged after infection with cytomegalovirus (CMV), a herpesvirus transmitted similarly to EBV. Serum levels of neurofilament light chain, a blood biomarker of neuroaxonal damage that rises as MS develops, increased only after EBV seroconversion, not before.
The 2018 Nature Genetics study on lupus and other autoimmunity
The lupus evidence
A separate 2018 study by John Harley and colleagues at the Center for Autoimmune Genomics and Etiology at Cincinnati Children’s Hospital Medical Center approached the question from the opposite direction. Rather than tracking who develops autoimmune disease after EBV infection, they asked whether an EBV protein could bind to and switch on the human genes already known to raise autoimmune disease risk. Using a computational method they developed called Regulatory Element Locus Intersection (RELI), they mapped where a viral transcription factor called EBNA2 sits on human DNA in EBV-infected B cells, then compared those binding sites to the human genome regions linked to autoimmune disease risk.
The finding was striking: nearly half of the human genome regions statistically associated with systemic lupus erythematosus (SLE) risk were bound by EBNA2. Analogous EBNA2-anchored patterns were found for six other autoimmune conditions: MS, rheumatoid arthritis, inflammatory bowel disease, type 1 diabetes, juvenile idiopathic arthritis, and celiac disease. In an EurekAlert! press release, Harley described the finding as “surprised to see that nearly half of the locations on the human genome known to contribute to lupus risk were also binding sites for EBNA2.” The finding provided a mechanistic candidate for how EBV might trigger autoimmunity: by activating a subset of the exact human genes that predispose to autoimmune disease.
How EBV might cause autoimmune disease
How EBV might trigger autoimmunity
Several complementary mechanisms have been proposed, and the current evidence suggests more than one may operate. The first is molecular mimicry: some EBV proteins share structural features with human proteins, so antibodies made against the virus may cross-react with normal tissues, such as the myelin sheath in MS. The second is the EBNA2 gene-regulatory mechanism identified by Harley and colleagues, in which a viral protein activates human autoimmune risk genes directly. The third is EBV’s persistent presence in B cells: EBV establishes lifelong latency in a subset of B lymphocytes, where it can reactivate periodically and drive B cell proliferation, some of which may cross-react with self-antigens. The three mechanisms are not mutually exclusive.
What the evidence points toward in treatment and prevention
If EBV is a required cause of MS in most cases, then preventing or clearing EBV should reduce MS incidence. Two lines of clinical research follow this logic. Vaccine development is at an early stage: several EBV vaccine candidates are in human trials, including Moderna’s messenger RNA candidates mRNA-1189 and mRNA-1195 and a nanoparticle-based vaccine from the US National Institutes of Health. A licensed EBV vaccine remains years away, and any measurable impact on MS incidence would take longer still to detect. On the treatment side, anti-CD20 therapies (ocrelizumab, rituximab, and related agents) that deplete circulating B cells are already among the most effective drugs for MS. Their efficacy is consistent with the EBV hypothesis, because depleting B cells also depletes the reservoir in which EBV establishes latency.
What the evidence cannot yet answer
What this means, and doesn’t
Neither study proves that EBV is sufficient to cause MS or lupus. The vast majority of people infected with EBV never develop either disease, so additional factors, likely genetic susceptibility and other environmental exposures, must determine which infected individuals go on to develop autoimmunity. The 2022 Science study established that EBV infection precedes MS diagnosis by an average of five years, but the biological events that link the two are only partially understood. The 2018 Nature Genetics study identified a molecular mechanism (EBNA2 binding at risk loci) but did not directly demonstrate that EBNA2 binding causes the transcriptional changes underlying disease. Randomized prevention trials of an EBV vaccine will be the strongest test of the causal hypothesis in humans.
What the findings mean
Anyone who has had infectious mononucleosis (mono) or a subclinical EBV infection does not need to worry about impending MS or lupus. Most people who catch EBV never develop autoimmune disease, and even in the highest-risk populations the absolute incidence of MS remains low. What the two studies have changed is the strategic direction of research on these diseases. For decades, MS and lupus were considered idiopathic autoimmune conditions of unclear etiology. Now researchers have a specific, ubiquitous, and potentially preventable trigger to target. That framing shift extends to other autoimmune conditions where EBV has been implicated, and it explains why the modern anti-CD20 drugs are among the most effective treatments for MS: the drugs happen to hit the reservoir where the trigger persists.
Ascherio and colleagues concluded in the 2022 Science paper that “these findings cannot be explained by any known risk factor for MS and suggest EBV as the leading cause of MS.” A picture is emerging in which a lifelong latent infection can, in a susceptible subset of people, gradually bias the immune system toward attacking the body. That reframing changes how autoimmune disease is understood, and it explains why prevention focused on a single virus could ripple across several conditions at once.
References
- Bjornevik K, Cortese M, Healy BC, et al. Longitudinal analysis reveals high prevalence of Epstein-Barr virus associated with multiple sclerosis. Science, 2022; 375(6578): 296-301. DOI: 10.1126/science.abj8222
- Harley JB, Chen X, Pujato M, et al. Transcription factors operate across disease loci, with EBNA2 implicated in autoimmunity. Nature Genetics, 2018; 50(5): 699-707. DOI: 10.1038/s41588-018-0102-3