EBV Reprograms B Cells in an Autoimmune-Like Fashion in Patients with COVID-19, 2026, Chen et al.

[SNT Gatchaman]

EBV Reprograms B Cells in an Autoimmune-Like Fashion in Patients with COVID-19

Spoiler: Authors

Daniel Guorui Chen; Dan Yuan; Yapeng Su; Andrew Magis; Helen Chu; Jason D Goldman; James R Heath

Epstein-Barr virus (EBV) reprograms B cells in autoimmune disease. Reprogrammed EBV+ B cells activate nearby B and CD4+ T cells, via upregulated antigen presentation and costimulatory machinery, to drive autoimmune pathology. EBV reactivation is a known correlate of long COVID, which is a heterogeneous condition that can bear similarities to autoimmune disease. However, the mechanisms underpinning this association remain unresolved.

We report on EBV metabolically reprogrammed B cells in patients with COVID-19. We find EBV+ B cells provide stimulatory signals to bystander B and CD4+ T cells. SARS-CoV-2 infected participants exhibiting elevated fractions of EBV+ B cells present, at convalescence, with dysregulated lipid profiles, increased autoantibody titers, and post-acute symptomology likely reflective of this metabolic reprogramming and cell-cell interactions.

Enrichment of our EBV+ B cell signatures seen in patients with COVID-19 is similar in patients with lupus and multiple sclerosis suggesting a potentially shared pathway of EBV-driven dysfunction across diseases.

Web | DOI | PDF | Preprint: BioRxiv | Open Access

 

[SNT Gatchaman]

In the past decade, large-scale studies have demonstrated a consistent positive association between prior EBV infection and increased autoimmune disease risk, particularly for multiple sclerosis (MS). […] Recent works have compounded on these findings by demonstrating a fundamental reprogramming of EBV-infected B cells towards a metabolically-distinct cell state (e.g. increased glycolysis) that markedly upregulates antigen presentation and co-stimulation machinery. Reprogrammed EBV-infected B cells can then leverage this machinery to activate autoreactive B and CD4 + T cells and thus facilitate autoimmune disease pathology, such as autoantibody production.

studies have consistently revealed a marked positive association between long COVID risk and EBV reactivation (or detection via viremia); these findings have been increasingly validated in larger patient populations such as the IMPACC cohort

We leveraged a reported computational framework in autoimmune disease to identify EBV-infected B cells in a primary cohort of 209 patients with COVID-19 (INCOV) and a validation cohort of 100 patients with COVID-19 (HAARVI)

Inspired by recent reports of marked B cell reprogramming by EBV in autoimmune disease, we hypothesized that EBV+ B cells may undergo a similar reprogramming in patients with COVID-19. To investigate this, we first classified single B cells into canonical B cell states by their marker gene expression (e.g. FCER2 in naïve B cells). We then interrogated for differences in the distribution of B cell states in EBV+ versus EBV- B cell compartments.

data suggest that patients with COVID-19 experience bona fide EBV reprogramming of their B cells towards a distinct transcriptional cell state. We confirmed these findings of EBV reprogrammed B cells in our independent validation cohort

 

[SNT Gatchaman]

The enrichment of antigen presentation and B cell activation processes in EBV reprogrammed B cells in the context of COVID-19 is reminiscent of the EBV reprogramming observed in certain autoimmune diseases.

In confirmation of our previous findings, identified EBV+ B cells once again enriched for cell populations (e.g. IgA+ plasmablasts) and gene signatures reported to be associated with EBV+ B cells in autoimmune contexts.

Together, these data suggest EBV+ B cells in patients with COVID-19 to have undergone canonical EBV reprogramming in terms of their transcriptomes (individual genes and pathways) and epigenomes, similar to EBV+ B cells in certain autoimmune diseases.

Consistently, in this study we found EBV+ memory B cells to present with the greatest upregulation of defining EBV reprogramming genes […] These genes were enriched for glycolytic and immunomodulatory functions (e.g. ENO1 49 and OAS150) and may suggest the existence of a core EBV reprogramming signature.

we performed metabolic flux analysis and found EBV+ memory B cells to significantly upregulate flux through the glycolysis pathway and decrease fatty acid breakdown (i.e. beta-oxidation).

With sufficient dysregulation, cellular metabolic reprogramming can lead to shifts in plasma metabolite levels. We assessed for such dysregulation in patients with increased proportions of EBV + B cells and were indeed able to find sets of metabolites whose levels during convalescence (2-3 months after infection) significantly correlated with patient EBV+ B cell percentages. The enrichment of these metabolites for dicarboxylic acids and carnitines, especially those of longer lengths, is consistent with our observation of reduced beta-oxidation in EBV+ memory B cells); these metabolites accumulate when beta-oxidation is impaired.

Thus, we find EBV+ memory B cells to undergo metabolic reprogramming at the cellular level, upregulated glycolysis programs and impaired beta oxidation, that is reflected by global metabolite changes in plasma.

 

[SNT Gatchaman]

The upregulation of costimulatory molecules in EBV+ B cells is reminiscent of reported mechanisms in autoimmune disease […] raise the possibility of shared molecular mechanisms between these conditions. To investigate this possibility, we queried for cell-cell interactions between EBV+ B cells and bystander EBV- B and CD4+ T cells in patients with COVID-19. Strikingly, we observed EBV+ B cells to be significantly enriched for ligand-receptor interactions that facilitate B cell survival and activation (e.g. BAFF:BAFFR, BAFF:TACI) and CD4+ T cell co-stimulation (e.g. CD80:CD28, CD86:CD28, CD70:CD27). […] we found patients with increased EBV+ B cell percentages to present with increased serum levels of autoantibodies and increased symptomology 2-3 months after acute disease resolution.

we interrogated for similarities between EBV+ B cells in COVID and in SLE […] We found the EBV+ B cell signature in COVID to significantly positively correlate with the expression of genes implicated in SLE (RNA, transcriptomic) and for patient EBV+ B cell percentages to significantly positively correlate with their polygenic SLE risk (DNA, genomic). […] As EBV reprogramming is also implicated in MS, we repeated this analysis in a public cohort of patients with MS and similarly found enrichment of our EBV+ B cell signatures from the COVID cohort.

 

[SNT Gatchaman]

we have four fundamental observations.

First, the identification of a non-zero (non-amplified) transcript signal within a B cell that is assignable to the EBV EBNA2 gene is sufficient to identify that cell as EBV-reprogrammed.

Second, that reprogramming, investigated here within the context of COVID-19, strongly resembles EBV reprogrammed B cells in autoimmune disease and is indeed associated with increased autoantibody titers and long COVID symptomology (e.g. difficulty concentrating).

Third, the scRNAseqassessed level of B cell reprogramming is at least semiquantitative; the impact in patients with a higher % of EBV transformed B cells is measurably greater.

Finally, reprogrammed B cells can exert significant influence over the global plasma environment of an individual.

Our finding that EBV+ B cells signatures in COVID-19 are similar to those observed in autoimmune disease, at the transcriptomic and epigenomic level, suggests the existence of a core EBV reprogramming gene module shared across diseases.

Recent works on SLE and MS demonstrated EBV+ B cells as drivers of autoimmune pathology by facilitating the expansion of both themselves and autoreactive double negative 2 (DN2, IgD- CD27-) and memory B cells into autoantibody secreting plasmablasts, secondary to stimulatory ligand-receptor interactions. Here, we find these same interactions to be present between EBV+ B cells and EBV- B and CD4+ T cells in patients with COVID-19.

Should DN2 B cells be stimulated by EBV+ B cells in patients with COVID-19, as suggested by the EBV+ and EBV- B cell interactions reported here, then there would exist a shared mechanism of autoantibody production across SLE, MS, and long COVID.

Together, these data suggest the clearance of EBV+ B cells, or disruption of their signaling, as a potential therapy for long COVID and, given potentially shared molecular mechanisms and clinical features, autoimmune disease.

 

[V.R.T.]

From nosing around looking for prolactin and IGHV3-30 links, see thresd here

A bit over my head so I may be jumping a few steps as I tend to but the thought was maybe prolactin can bypass some checkpointing in B cell maturation and selection which can lead to changes in b cell repertoire. So we could potentially see a situation where transitory elevated or spiky prolactin levels could lead to elevated IGHV3-30. A downstream effect of a downstream effect potentially.

Of note, JE says this is too tenuous to be worth investing in.

However, with that important warning of rampant speculation, here’s my thoughts…

I was envisioning a loop whereby, whatever the underlying cause, transitory prolactin spikes could change the B cell repertoire which itself triggers prolactin spikes through an effect on serotonin or dopamine receptors via something like GPCR autoantibodies. And maybe these B cells would only survive or survive more often during periods of these prolactin spikes, like illness or any physical stressors, maybe type I Interferon could fit in here too? The population could change and drop a bit over time through normal cell recycling processes but never get back to normal baseline without something to snap the cycle completely. It could explain some of the patterns we see in ME/CFS.

That said, B cell depleting therapies have not worked. Which means there may be a correlation but this may not be the full mechanism. The comment in the paper above which said “the expansion and activation of anti-DNA B cells in prolactin-treated R4A-γ2b BALB/c mice was dependent on the presence of CD4+ T cells” may be relevant here somehow. If they are required in this prolactin driven checkpointing change then just depleting the naive B cell population but leaving any CD4+ T cells (which also express prolactin receptors) and any lurking mature long lived plasma cells around wouldn’t stop the loop.

Thought this might be of tangental relevance as it discusses similar cell populations.