The scientific backdrop to this work is a long-standing debate about what stimulates the malignant B cells in PCNSL. Even before the characteristic mutations in MYD88 and CD79B were identified, investigators suspected that stimulation of the B-cell receptor (BCR) by specific antigens might contribute to the disease. Previous research had already uncovered autoantigens as targets of PCNSL BCRs, most notably SAMD14 and neurabin-I, two proteins that share a highly homologous sterile alpha-motif domain. SAMD14 is involved in hematopoietic progenitor signaling, while neurabin-I is a scaffold protein enriched in neural tissue that binds F-actin and protein phosphatase-1 and participates in cytoskeletal organization and neuronal signaling. Remarkably, reactivity against these proteins was found in 8 of 12 PCNSL-derived BCRs, and antigen engagement induced BCR pathway activation and proliferation in experimental lymphoma models. A peculiar feature of these autoantigens is their immunogenic hyper-N-glycosylation at specific asparagine residues, and a similarly atypically modified protein, SEL1L3, has been identified as a BCR target in primary intraocular lymphoma.

Infectious triggers have been convincingly identified for several other lymphoma entities. Antigens of hepatitis C virus are implicated in splenic marginal zone lymphoma, Helicobacter pylori drives gastric extranodal marginal zone lymphoma, and bacterial antigens from Moraxella catarrhalis and Rothia mucilaginosa have been linked to nodular lymphocyte predominant Hodgkin lymphoma. In some of these diseases, eradication of the infectious trigger is part of standard therapy. For PCNSL, however, the evidence for infectious agents is far thinner. The best-established viral association is Epstein-Barr virus in immunodeficiency-associated PCNSL, whereas EBV is absent or detected only rarely in PCNSL arising in immunocompetent patients. Beyond EBV, no reproducible infectious trigger has been established, which provided the rationale for an unbiased screen of PCNSL BCRs against viral antigens.

The methodological approach was ambitious. The researchers generated recombinant antigen-binding fragments, or Fabs, of the B-cell receptors from 15 patients with PCNSL and screened them at concentrations of 1, 10, and 20 micrograms per milliliter on a custom-made VIROME microarray. This array represented 2,206 viral genes drawn from 158 viruses and was produced using nucleic acid programmable protein array technology at the Biodesign Institute of Arizona State University, which also performed the screening and analysis. Among the candidate antigens that produced reproducible signals, one pool showed reactivity against a Coxsackievirus A24-derived clone. Deconvolution by individual testing of the constituent Fabs identified a single reactive Fab among the 15, and subsequent dot-blot analysis and ELISA verified the Coxsackievirus A24 antigen clone CaCD00959607 as a specific target.

The specificity of this reactivity was reinforced by an extensive control panel. Recombinant Fabs from 153 patients with other lymphoma entities, including lymphoplasmacytic lymphoma, mantle cell lymphoma, angioimmunoblastic T-cell lymphoma, primary vitreoretinal lymphoma, Burkitt lymphoma, nodular lymphocyte predominant Hodgkin lymphoma, classic Hodgkin lymphoma cases with BCR expression, follicular lymphoma, and diffuse large B-cell lymphoma, were also tested against the Coxsackie A24 antigen. Reactivity was observed only in the PCNSL samples and was completely absent across all non-PCNSL lymphomas. None of the other viral antigens on the array demonstrated any relevant reactivity with PCNSL-derived receptors, and the reactive PCNSL Fab showed no cross-reactivity with the known autoantigens neurabin-I or SAMD14.

Validation in an independent cohort added further, though limited, support. In paired serum and cerebrospinal fluid samples from 22 patients with PCNSL, reactivity against the Coxsackievirus A24-derived antigen was detected in exactly one patient, present in both CSF and plasma, corresponding to a frequency of 4.55 percent. The antibodies were of IgG class, with an endpoint titer of 1:400 in plasma and 1:50 in CSF. Because most seroprevalence studies of Coxsackievirus A24 were conducted in post-endemic settings with different methodologies, the team examined 188 geographically matched healthy donors using the same recombinant antigen-based ELISA. Only one donor was weakly reactive, a prevalence of 0.53 percent. The numbers suggest enrichment in the PCNSL cohort, but with a single reactive individual in each group the authors caution that a chance association cannot be excluded.

Two observations complicate a straightforward infectious interpretation. First, the antigen-specific endpoint titer was eightfold lower in CSF than in plasma. Given the substantially lower total IgG concentration normally found in CSF, this relatively preserved reactivity might suggest a disproportionate representation of Coxsackievirus A24-reactive IgG in the CSF compartment, hinting at intrathecal antibody synthesis. However, endpoint titers are semiquantitative, and an antigen-specific antibody index incorporating paired total IgG concentrations and blood-CSF barrier function was not available, so intrathecal synthesis or CNS infection cannot be established. Second, cryopreserved lymphoma tissue from the reactive patient and two controls was tested for viral RNA, and no traces of Coxsackievirus A24 RNA were found. This argues against a persistent or latent infection maintaining a lymphomagenic stimulus, although the authors note that the tissue had been stored for a considerably long period, which may have influenced the result.

The most compelling part of the study is functional. Because no suitable EBV-negative PCNSL cell line exists, the researchers used a previously established doxycycline-inducible BCR expression system in the activated B-cell-like diffuse large B-cell lymphoma lines TMD8 and U2932, engineering the cells to express either the Coxsackievirus A24-reactive PCNSL BCR or a SAMD14/neurabin-I-reactive PCNSL BCR as an antigen-specific control. Following induction, cells were stimulated with the recombinant Coxsackievirus A24-derived antigen or SAMD14 at 5 micrograms per milliliter, with anti-IgM as a positive control and non-induced cells to confirm dependence on recombinant BCR expression. Stimulation with the Coxsackievirus A24-derived antigen induced phosphorylation of the signaling kinases SYK, BTK, PLCγ2, and BLNK, together with increased c-MYC expression, but only in cells expressing the Coxsackievirus A24-reactive receptor. To the authors’ knowledge, this represents the first demonstration that a viral antigen can activate a PCNSL-derived BCR, although the heterologous cell-line model demonstrates acute signaling without establishing persistent stimulation, enhanced tumor-cell survival, or causality.

Coxsackieviruses A and B belong to the Enterovirus genus and are common human pathogens with a wide spectrum of manifestations affecting skin, heart, and the central nervous system. Coxsackievirus B is among the most common causes of aseptic meningitis in children under two years of age, while coxsackievirus A is most often associated with hand, foot, and mouth disease and herpangina. Coxsackievirus A24 specifically has been linked chiefly to outbreaks of acute hemorrhagic conjunctivitis and, rarely, to acute flaccid paralysis, but not to CNS involvement, so these observations provide only limited evidence of neurotropism. Because enterovirus exposure is common, the isolated PCNSL reactivity could equally reflect previous infection, cross-reactivity, or chance. Published serological studies are difficult to compare because they were largely performed in outbreak-related settings using whole-virus neutralization assays rather than the recombinant antigen approach used here.

The authors and the field alike stress the limits of this exploratory work. The screening covered a large but incomplete set of viral constructs, the cohorts were small, the design was retrospective, and the rarity of PCNSL itself constrains statistical power. The functional experiments, while antigen-specific, cannot reproduce the complete biological context of PCNSL. What the study does establish is a genuine antigen recognition event: a recombinant Coxsackievirus A24-derived protein can engage a PCNSL-derived B-cell receptor and trigger its downstream signaling cascade, a biologically plausible interaction confirmed in an independent patient and largely absent from healthy and diseased controls. Whether viral antigens play any meaningful pathogenic role in this malignancy will require larger independent studies, molecular characterization of the recognized viral epitope, and functional assessment of cognate BCR signaling in settings that more faithfully model the disease. Until then, Coxsackievirus A24 stands as a tantalizing, unproven suspect in one of neuro-oncology’s most stubborn mysteries.

Subject of Research: B-cell receptor reactivity against Coxsackievirus A24 in primary central nervous system lymphoma

Article Title: Rare But Specific B‐Cell Receptor Reactivity Against Coxsackievirus A24 in Primary Central Nervous System Lymphomas

Article References: Rare But Specific B‐Cell Receptor Reactivity Against Coxsackievirus A24 in Primary Central Nervous System Lymphomas. (n.d.). https://doi.org/10.1002/jha2.70422

Image Credits: AI Generated

DOI: 10.1002/jha2.70422

Keywords: primary central nervous system lymphoma, Coxsackievirus A24, B-cell receptor, diffuse large B-cell lymphoma, viral antigens, VIROME microarray, BCR signaling, enterovirus, autoantigens, SAMD14, neurabin-I, cerebrospinal fluid