How Viruses Break B-Cell Tolerance and Trigger Autoimmune Disease

Viral infections are increasingly recognized as triggers of autoimmunity, yet the mechanisms connecting infection to the collapse of B-cell tolerance have remained fragmented across pathogen-specific studies. A review published in EMBO Reports by Michael Fraczek, Gielenny M. Salem, Ivonne Melano, Xuan Ning Sharon Tan, Charlotte S. M. Lim and Weiqiang Chen assembles mechanistic and clinical evidence from systemic, localized and neurotropic viral infections into a unified framework. Its central argument is that autoimmune outcomes after infection are determined less by which virus is involved than by the immunological and anatomical context in which tolerance fails.

B-cell tolerance is maintained through sequential checkpoints that normally eliminate or silence self-reactive clones. The problem begins with the inherent stochasticity of antibody diversification: 55 to 75 percent of newly generated immature B cells in the bone marrow express self-reactive B-cell receptors, according to landmark repertoire studies. Central tolerance in the bone marrow resolves this autoreactivity through three mechanisms: receptor editing, in which RAG1/2 reactivation drives secondary immunoglobulin light-chain rearrangements that replace autoreactive specificities; clonal deletion by apoptosis for cells with high-affinity binding to membrane-bound self-antigens; and functional silencing through anergy for cells encountering soluble self-antigens at intermediate affinity. Current evidence indicates receptor editing is the principal mechanism, with deletion occurring largely after editing has been exhausted. As a result of this stringent selection, fewer than 10 to 20 percent of developing bone marrow B cells successfully mature and exit into the peripheral circulation.

Central tolerance is further tuned by intracellular signaling. Elevated phosphoinositide 3-kinase (PI3K) signaling promotes immature B-cell survival and positive selection, whereas transient PI3K inhibition permits Foxo1 activation and induction of RAG1/2 for receptor editing. MicroRNAs fine-tune this threshold: miR-19 within the miR-17–92 cluster suppresses PTEN and thereby limits receptor editing, while miR-148a enhances survival of autoreactive B cells by suppressing pro-apoptotic regulators including Gadd45α, PTEN and Bim, a mechanism implicated in systemic lupus erythematosus and lupus-prone mouse models. Collectively, central tolerance reduces autoreactivity from roughly 75 percent in immature bone marrow B cells to around 40 percent upon peripheral egress.

Peripheral checkpoints then continue the pruning. Transitional B cells depend on the cytokine BAFF for survival, and strong autoreactive signaling at the T1 stage triggers clonal deletion, whereas chronic low-affinity engagement enforces anergy. The T3 compartment is enriched for functionally anergic, self-reactive clones that fail to enter the conventional mature pool. Among mature cells, follicular B cells face stringent regulation through BCR signaling thresholds and competition for survival factors, while marginal zone B cells operate with lower signaling thresholds and rapid response capacity. Despite this layered efficiency, roughly 40 percent of newly emigrated transitional B cells retain some autoreactivity, which is pruned to about 20 percent in the mature peripheral pool. Autoreactive cells that encounter self-antigen in secondary lymphoid organs are censored through follicular exclusion and rapid apoptosis outside follicles, and recent evidence confirms the periphery, rather than the bone marrow, is a major site of immature and transitional B-cell apoptotic clearance.

The germinal center adds further checkpoints. Within its dark and light zones, iterative affinity-based selection on follicular dendritic cells and competition for limiting T follicular helper cell signals preferentially expand high-affinity clones, while non-selected clones or those acquiring deleterious mutations after activation-induced cytidine deaminase activity undergo apoptosis. Because germinal center selection tracks antigen affinity rather than self-reactivity itself, some high-affinity autoreactive variants can evade localized deletion. A post-germinal-center apoptotic checkpoint therefore selectively eliminates mutated autoreactive memory B cells and plasma cells; disruption of this control, for example through enforced Bcl-2 expression, permits survival of mutated autoreactive clones, driving anti-nuclear antibody production, systemic autoimmunity and increased susceptibility to lymphomagenesis. Follicular regulatory T cells and the PD-1/PD-L1 axis add further restraint.

Extrafollicular responses, which bypass germinal centers to provide rapid early antibody defense during acute infection, are a critical vulnerability. These responses typically proceed with minimal somatic hypermutation, and their autoreactivity in healthy individuals is transient, reflecting activation of pre-existing self-reactive clones rather than de novo generation of autoreactive antibodies. In contrast, chronic autoimmune diseases such as lupus feature dysregulated extrafollicular responses with substantial AID activity, extensive mutation and sustained clonal expansion, driven by T peripheral helper cells providing IL-21 and CXCL13 within inflamed tissues. The review emphasizes that the strong association between extrafollicular responses and pathogenic autoantibodies likely reflects the relative absence of the stringent affinity- and survival-based checkpoints of germinal centers rather than active positive selection of autoreactive clones, though direct experimental comparisons remain limited.

Against this backdrop, the authors identify four convergent mechanisms by which viruses erode tolerance. First, inflammatory checkpoint erosion: during viremia, elevated type I interferon and pro-inflammatory cytokines reprogram B-cell activation thresholds and impair regulatory T cell and follicular regulatory T cell suppression, which normally depends on CTLA-4. In severe SARS-CoV-2 infection, high levels of CRP, IFN-I, IL-6, IL-18 and IL-1β correlate with poor clinical outcomes, and this cytokine storm, particularly IFN-I and IL-18, can promote extrafollicular responses that bypass germinal center selection. Excessive TLR7 signaling, a well-established B cell-intrinsic driver of lupus, is highlighted by de novo and somatic TLR7 gain-of-function variants that cause severe early-onset disease, and hyperactivating variants of the endosomal chaperone UNC93B1 produce similar phenotypes. When severe viral inflammation disrupts regulatory networks, unchecked IL-2 availability drives extrafollicular plasma cell differentiation through the mTOR–AKT axis, accelerating production of low-mutation pathogenic antibodies.

Second, viral infection frequently shunts B cells toward extrafollicular activation. In severe SARS-CoV-2 infection, germinal center formation is impaired and humoral immunity is dominated by extrafollicular responses, with critically ill patients showing expansion of T-bet+CD11c+ double-negative B cells inferred to arise through this pathway. During Epstein–Barr virus infection, autoreactive bystander double-negative B cells can be expanded by EBV-infected antigen-presenting B cells from lupus patients, while EBV-driven T-bet+, CXCR3+ age-associated B cells are enriched in inflamed meninges from multiple sclerosis patients and promote neuroinflammation. Third, molecular mimicry and polyreactivity allow antiviral responses to cross-react with host antigens: about 17 percent of IgG1 antibodies in SARS-CoV-2 infection cross-react with both viral and diverse self-antigens including nuclear, glomerular and neuronal targets, and the viral ORF8 protein mimics host IL-17A. Arthritogenic alphaviruses share homology with joint and connective tissue antigens, and EBNA1 mimics CNS self-antigens linked to multiple sclerosis, with EBV-mediated suppression of NK-cell cytotoxicity further preventing clearance of autoreactive B cells. Fourth, persistent antigen exposure and self-antigen leakage rescue autoreactive cells from deletion or anergy, whether through latent EBV reservoirs in memory B cells, persistent chikungunya viral RNA in synovial tissue, or blood-brain barrier disruption during Japanese encephalitis virus and HSV-1 infection that releases sequestered neural antigens, including via exosome-mediated release of NMDA receptor and GABA-B receptor antigens.

The clinical consequences differ with anatomical context. Among individuals without prior autoimmunity, SARS-CoV-2 infection increases the risk of new-onset autoimmune disease by about 40 percent, spanning multiple organ systems from multiple sclerosis and Graves’ disease to rheumatoid arthritis, type 1 diabetes and lupus. Conversely, rare loss-of-function TLR7 variants on the X chromosome are associated with a 5.3-fold increased risk of severe COVID-19, underscoring the same pathway’s protective role in antiviral immunity. Localized persistence produces organ-restricted disease: chikungunya causes chronic rheumatoid arthritis-like disease in roughly 36 percent of infected individuals, and over 50 percent of patients exposed to Ross River virus report chronic joint pain three years after infection, with disease severity correlating poorly with circulating autoantibody titers, pointing to tissue-resident B-cell responses within the joint. In immune-privileged sites, EBV infection confers a 32-fold increased risk of multiple sclerosis, up to 16 percent of Japanese encephalitis patients develop delayed neuropsychiatric symptoms with neuronal autoantibodies despite viral clearance, and 27 percent of patients with acute HSV-1 infection developed autoimmune encephalitis within 32 days, marked by robust autoantibodies against NMDA receptors.

The review closes with open questions that frame future work: where tolerance first fails during infection, when extrafollicular responses turn pathogenic, whether tolerance failure is transient or self-sustaining, and how tissue context shapes clinical outcomes. Resolving these will require integrated spatial transcriptomics, lineage tracing and longitudinal humoral profiling. The authors argue that such a unified framework could enable prediction of post-viral autoimmune risk, identification of vulnerable tissues, and therapeutic strategies that restore immune tolerance while preserving protective antiviral memory.

Subject of Research: Mechanisms by which viral infections disrupt B-cell tolerance checkpoints to promote autoimmune disease

Article Title: Viral subversion of B-cell tolerance: mechanisms linking infection to autoimmunity

Article References: Fraczek, M., Salem, G. M., Melano, I., Tan, X. N. S., Lim, C. S. M., & Chen, W. (2026). Viral subversion of B-cell tolerance: mechanisms linking infection to autoimmunity. EMBO Reports. https://doi.org/10.1038/s44319-026-00959-2

Image Credits: AI Generated

DOI: 10.1038/s44319-026-00959-2

Keywords: B-cell tolerance, autoimmunity, viral infection, germinal center, extrafollicular response, molecular mimicry, TLR7, SARS-CoV-2, Epstein-Barr virus, chikungunya, autoimmune encephalitis, checkpoint erosion