Multiple sclerosis (MS) is an autoimmune-mediated demyelinating disease of the central nervous system (CNS). Early evidence showed the presence of CD8⁺ T lymphocytes in the cerebrospinal fluid (CSF) and in perivascular leukocyte infiltration from white matter in chronic and active MS lesions (1–5). Thus, since there are few natural killer (NK) cells compared with T cells in the CSF of MS patients (6) and also effector T populations may be even more potent than NK cells in releasing cytotoxic granules (7); the expression of cytotoxic-associated molecules such as serine-protease granzyme-B (GzmB), during MS, seems to almost be exclusively originating from CD8⁺ T lymphocytes. Interestingly, Runx3, which is a crucial transcriptional factor related to the expression of cytotoxic molecules in effector CD8⁺ T subsets (8, 9), is reported as an MS-associated gene (10). In parallel, neurons express the mannose-6-phosphate receptor (M6PR), responsible for internalizing GzmB, which then makes them vulnerable to cell death triggered by this protease. In vitro evidence suggests that serine-protease inhibitors can dampen neuronal cell death associated with GzmB internalization (11). Supporting these findings, it was shown that MS patients exhibit higher GzmB levels in the CSF during relapses that tend to persist higher at 1–3 months into clinical remission (12). Also, increased circulating T lymphocytes with the ability to express GzmB were found in the peripheral blood from relapsing-remitting MS (RRMS) patients treated with fingolimod (FTY), and particularly during relapses, when compared to RRMS patients without FTY (13). Similarly, massive infiltration of cytotoxic CD8⁺GzmB⁺ T lymphocytes was found in the CNS parenchyma from two MS patients who suffered fulminant relapses after natalizumab (NTZ) discontinuation (14, 15). On the other hand, regarding progressive MS courses, not only the CSF from secondary progressive MS (SPMS) patients showed in vitro neurotoxicity due to the expression of GzmB (16) but also cytotoxic CD8⁺CD57⁺ T lymphocytes seem to be present in inflamed meninges in these patients with rapidly progressive disease (4). Altogether, these findings reinforce that cytotoxic mechanisms derived from CD8⁺ T lymphocytes are pivotal drivers of CNS damage during MS (12, 17, 18).

Despite this, successful outcomes in the last few years by the use of anti-CD20 monoclonal antibodies (mAbs) (rituximab, ocrelizumab, or ofatumumab) reassessed the importance of B cells during both relapsing-remitting (RRMS) and progressive MS courses (4). Indeed, oligoclonal band (OCB) synthesis, compartmentalized clonal expansion, and increased levels of chemoattractants for B cells and/or plasma cells in the CSF (19–22) were extensively described in MS patients. Nevertheless, since the CD20 molecule is not expressed on pro-B cells or differentiated plasma cells, the beneficial effect of anti-CD20 treatment appears to be extended beyond autoantibody production and release. For instance, in the last few years, increasing evidence supports that B subsets can express and release anti- and pro-inflammatory cytokines, evidencing their antibody-independent functions during MS pathophysiology (23, 24). Considering it, in the present study, we evaluated whether CD19⁺ B subsets may also exhibit the capacity to express and release GzmB similarly resembling the cytotoxic activity described for T lymphocytes in RRMS patients.

Herein, we demonstrated that CD19⁺ B cells from RRMS patients share the ability to express serine-protease GzmB, similarly resembling classical CD8⁺ T lymphocytes.

Regarding T lymphocytes, we show here that RRMS patients exhibit an increased percentage of circulating CD8⁺GzmB⁺ T lymphocytes when compared to healthy volunteers. Moreover, treated RRMS patients, particularly FTY and NTZ subgroups, showed higher CD8⁺GzmB⁺ T lymphocytes than healthy subjects.

Enhanced cytotoxic behavior derived from T lymphocytes has been suggested as a mechanism for controlling latent Epstein–Barr virus (EBV) infection and preventing viral replication during MS (27). However, sustained cytotoxic CD8⁺ T cell activity would also be implicated in CNS lesions during disease. Indeed, infiltration of CD8⁺GzmB⁺ T lymphocytes that respond against EBV-infected B cells/plasma cells was recently found in the CNS lesions from two MS patients who died after suffering fulminant relapses following NTZ withdrawal (14, 28).

In addition, Cencioni et al. (4) showed that cytotoxic CD57⁺ T subsets occur in inflamed meninges from progressive MS patients and are negatively correlated with disease progression/age of death. Interestingly, higher expression of programmed death-1 (PD-1) in circulating CD8⁺CD57⁺ T lymphocytes correlates with disease stability. In vitro blockade of PD-1 enhanced the release of IFN-γ, Perforin, and GzmB by these terminally differentiated cytotoxic T subsets from MS patients (4).

According to previous reports (4, 29), RRMS patients in our cohort showed increased expression of GzmB in CD8⁺ T lymphocytes markedly associated with senescent T phenotype exhibiting CD27^-, CD28^-, CD57⁺, and CD94⁺ markers.

Indeed, diverse evidence suggests that cytotoxic subsets including those exhibiting senescent CD28^- and CD57⁺ markers restrain the migration ability into inflamed tissues in response to chemokines and also to express and release GzmB and other pro-inflammatory cytokines supporting tissue damage in diverse conditions (30).

Considering this, in the last few years, these subsets seemed to have emerged as candidates for predicting disease worsening in several diseases. The prognostic value of CD4⁺CD28^- T subset during MS was recently suggested for progressive disease (31). However, there is still a lack of studies investigating the potential of cytotoxic behavior in subsets other than T lymphocytes and its possible implications regarding different MS clinical courses/treatments.

On the other hand, the role of B cells during MS has been more deeply investigated in the context of antibody-independent functions. For instance, IL-21, which is known to promote B cell differentiation to memory and plasma cells in the presence of both BCR or Toll-like receptor (TLR) signaling and CD40L co-stimulation, may also promote GzmB-secreting B cells in the absence of CD40L co-stimulation (26, 32, 33).

Thus, considering that B cells may differentiate into GzmB-producing cells upon insufficient T cell help, herein, we have provided evidence that this phenomenon may occur during MS. Similar to CD8⁺ T lymphocytes, we found no differences in total circulating CD19⁺ B cells. However, our results show an increased percentage of circulating CD19⁺GzmB⁺ B cells in RRMS patients vs. healthy. Treated RRMS subgroup patients showed higher amounts of CD19⁺GzmB⁺ B cells during FTY and NTZ when compared to patients during first-line immunomodulatory therapy (GA), untreated RRMS patients, and healthy donors. We were able to assess which B-cell markers in CD3^-CD19⁺ subsets were associated with the cytotoxic phenotype using flow cytometry high-dimensional analyses high-dimensional analyses. Accordingly, with previous literature, we observed that not CD20⁺ but CD38⁺ activated and CD138⁺ plasma cells seem to identify GzmB-expressing phenotype in B subsets (26, 34, 35). Moreover, Runx3, a master regulator associated with cytotoxic behavior in CD8⁺ T lymphocytes (8, 9) positively correlated with the GzmB-expressing phenotype.

As previously suggested by De Andrés et al. (36), these results reinforce a possible antibody-independent pathophysiological mechanism derived from B-cell subsets with the ability to express GzmB during MS. Beyond this, and considering the clinical efficacy of both FTY and NTZ, we may hypothesize that cytotoxicity may represent or even coexist with other tolerogenic functions in B subsets. Resembling our MS cohort, similar percentages of circulating CD19⁺GzmB⁺ B cells, in the absence of IL-10 coexpression, were described during HIV infection (37). Also, regulatory activity of GzmB-derived circulating CD19⁺ B cells was suggested due to degradation of TCR-ζ–chain that promotes a significant decrease in T-cell proliferation (32, 37, 38). As our results suggest, by now, it seems that GzmB expression is mainly derived from CD20^- B subsets with CD38⁺ and CD138⁺ markers. Indeed, beyond several changes regarding the total percentage of CD19⁺ B cells comprehending naive and memory phenotypes, as well as regulatory B subsets, increased circulating plasma cells were already described during highly effective MS treatments (39). Interestingly, reduced tumor necrosis factor (TNF)-α and enhanced interleukin (IL)-10 expression by B subsets were also reported during FTY. Yet, these regulatory IL-10-expressing B cells seem to be increased in the CSF from FTY-treated patients (24, 40).

It is noteworthy that despite CD20⁺ B cells being found in CNS lesions from different stages of the disease, many authors have proposed that B cells would take a later role in MS pathophysiology, since, in 2004, CD20⁺ B cells, CD138⁺ plasma cells, and follicular dendritic cells were described in tertiary lymphoid organs in inflamed meninges from progressive MS patients (5, 27).

Further investigation in progressive MS courses may identify whether or not GzmB-derived B cells occur during chronic disease pathogenesis. So far, cytotoxicity derived from B cells was shown to cause damage in oligodendrocytes and neurons (41, 42), eventually sustaining a silent and continuous CNS-restricted inflammatory process. Supporting this, anti-CD20 mAbs seem to be effective for managing progressive MS mainly during early disease course (24, 40, 43) and have also been suggested for mitigating the increased risk of relapses in RRMS patients after NTZ washout (44).

Thus, since anti-CD20 mAbs mainly deplete naive and memory B cells, preserving antibody-secreting (CD138+) plasma cells, cytotoxic behavior derived from CD20^- B subsets would be preserved during these treatments. Further investigations of cytotoxic behavior in CD19⁺ may address, for instance, eventual important mechanisms associated with the clinical efficacy of emerging anti-CD19 mAbs and oral drugs targeting Bruton’s tyrosine kinase (BTK) for MS patients (45).

The raw data supporting the conclusions of this article will be made available by the authors without undue reservation.

The studies involving human participants were reviewed and approved by the University of Campinas Committee for Ethical Research (CAAE: 53022516.3.0000.5404). The patients/participants provided their written informed consent to participate in this study.

VB, RQ, and ASM performed most of the experiments. CS, AD, RC, and CB diagnosed, treated, and selected MS patients as well as recruited all healthy donors. VB, AMM, ASM, MA, and BF designed and performed flow cytometry. VB, AMM, and AF analyzed flow cytometry data. AL and IS performed CBA experiments. AF and LS designed the experimental work. AF coordinated the study. VB, AMM, AF, and LS wrote the article with inputs from co-authors. All authors contributed to the article and approved the submitted version.

This work was supported by grants from Sao Paulo Research Foundation (FAPESP) (#2014/26431-0, 2015/22052-8, #2017/21363-5, #2019/06372-3, #2019/16116-4). This study was also partly financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES)-Finance Code 001 (2015/22052-8).

LS received a research grant from Biogen and a consultation honorarium from Biogen and Roche.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.