EBV-Specific CD8+ T Cells in CSF: New Insights Into Early MS Immune Biology

Clonally expanded CD8+ T cells enriched in cerebrospinal fluid (CSF) from early, treatment‑naive MS/CIS included definitively EBV‑specific clonotypes in a Nature Immunology study.

Paired single‑cell RNA/TCR sequencing was coupled with orthogonal antigen‑discovery and validation workflows, offering a clinician‑relevant bridge from the long‑standing EBV–MS epidemiologic association to a plausible intrathecal immune mechanism—without implying a causal chain. The results strengthen biologic plausibility for EBV‑linked immune activity within CSF while leaving causality and disease directionality unresolved.

To capture early disease biology, investigators collected paired blood and CSF on the same day from 18 participants—relapsing‑remitting MS and clinically isolated syndrome (MS/CIS) alongside healthy and other neuroinflammatory comparators—and analyzed CD8+ TCR clonotypes as the practical unit of intrathecal expansion. Highly expanded CSF clonotypes were defined by frequency thresholds anchored to each individual’s repertoire, and the clinically emphasized subset required both marked CSF expansion (≥0.75% of an individual’s CSF T cell repertoire) and CSF enrichment (CSF frequency at least 2× the paired blood frequency). That filter yielded 23 highly expanded, CSF‑enriched CD8+ clonotypes for antigen mapping, intentionally focusing on a small, potentially disease‑relevant intrathecal effector population rather than diffuse repertoire noise.

Antigen specificity was assigned through three parallel pipelines designed to triangulate sequence, binding, and function rather than over‑weighting any single readout. An unbiased peptide–MHC (pMHC) yeast‑display screen nominated candidate peptide ligands for individual patient‑derived TCRs, generating mimotopes and prompting follow‑on checks against naturally occurring homologs. In parallel, TCR sequences were aligned to public specificity resources to flag plausible viral targets—useful for hypothesis generation but vulnerable to misassignment without experimental confirmation. Specificity was then tested directly with targeted pMHC tetramers and functional assays after CRISPR‑mediated TCR knock‑in into primary human CD8+ T cells; on this orthogonal validation step, three expanded CSF CD8+ clonotypes were confirmed as EBV‑reactive by tetramer binding and cytokine production, while most expanded clonotypes remained unassigned.

Where the evidence allowed epitope‑level resolution and HLA restriction, EBV targets were interpreted narrowly as proof of specificity—not proof of pathogenic relevance. One expanded CSF‑enriched clonotype exactly matched a known EBV specificity to EBNA3A193–201 (FLRGRAYGL) presented by HLA‑B*08:01 and showed robust tetramer staining plus cytokine responses with cognate peptide stimulation after TCR knock‑in. A second expanded CSF‑enriched clonotype was a near‑exact match to an EBV lytic antigen epitope, BZLF154–64 (EPLPQGQLTAY) presented by HLA‑B*35:01, again supported by tetramer binding and functional cytokine production in engineered CD8+ T cells. In addition, an expanded CSF‑enriched clonotype reproducibly activated against EBV‑transformed lymphoblastoid cell lines in an HLA‑A*29:02–restricted context without a pinned‑down peptide, underscoring that at least some intrathecal expanded CD8+ clones in early MS are EBV‑specific even as the broader antigen landscape remains largely unresolved.

Independent molecular assays added a second axis of plausibility by detecting EBV nucleic acids in CSF supernatant. Conventional PCR was positive in a subset of participants, whereas droplet digital PCR detected EBV in nearly all participants with variable abundance and the highest levels in MS/CIS. Transcript assays extended the signal beyond DNA alone, and BamHI‑W transcripts in particular were reported as significantly increased in MS/CIS compared with controls; this pattern was observed in patients who also carried expanded EBV‑reactive CSF clonotypes, linking the combined findings to EBV BamHI‑W transcripts in CSF as a marker compatible with heightened EBV activity in the intrathecal compartment. Even so, the molecular signal cannot distinguish local CNS/CSF reactivation from bystander DNA/RNA release or a downstream marker of B‑cell activation, and it does not support causal inference on its own.

Key Takeaways:

• EBV specificity can be demonstrated for a minority of highly expanded, CSF‑enriched CD8+ clonotypes in early, treatment‑naive MS/CIS using convergent sequence, binding, and functional assays.
• Validated EBV targets included EBNA3A193–201 (HLA‑B*08:01) and BZLF154–64 (HLA‑B*35:01), while other EBV‑reactive signals lacked peptide‑level resolution.
• EBV DNA and transcripts in CSF—especially increased BamHI‑W transcripts in MS/CIS—support biologic plausibility but do not establish local reactivation or causality.