In this study, we demonstrated local maturation to and persistence of CD38highCD27high ASCs as a specific hallmark of the MS brain, most abundantly in white matter lesions. B-cell maturation as analyzed in CNS compartments of people with MS was associated with in situ enhanced lesion activity and both intrathecal as well as local in situ IgG production. CXCR3 expression was revealed as an important determinant for intrathecal IgG production. B-cell maturation in MS lesions seemed dependent on local interaction with CD4+ memory T cells. Overall, we provide a complete overview of the ASC and B-cell phenotype and distribution in distinct compartments of the affected (MS) and unaffected (control) human brain. These findings provide a direct link between MS B cell phenotype, antibody-secreting function, and local interaction with T cells, which adds to existing indirect evidence in providing a rationale for therapies targeting these mechanisms.
CNS-infiltrating CXCR3+ B cells have been implicated in MS and further development of this B-cell population into ASCs is thought to be important for local pathology [19] [22]. Focusing on their maturation profile within the CNS, we identified a prominent population of CD38highCD27high ASCs within distinct postmortem CNS compartments of MS patients and found a relative enrichment in the CNS compared to peripheral blood, which was most profound in MS white matter lesions. Recently, high plasmablast frequencies were already detected in MS CSF versus blood [45], but we extend these data with our findings. The increased presence of ASCs in the MS CNS supports the hypothesis that CNS-infiltrated memory B cells in MS mature into antibody-producing plasmablasts and plasma cells upon reactivation [8] [19] [22]. Clonal overlap has been found between B cells in distinct CNS compartments and blood of MS patients, indicating that B cells likely migrate towards the CNS and traffic across the tissue barrier within the CNS [9] [21] [46] [47]. However, not much is known about the effector functions and maturation of these B-cell populations in the CNS and therefore we closely looked into this.
Consistent with the positive link between ASC presence and lesion activity, Ig gene expression, most importantly IgG, was elevated in active MS lesions. ASC presence within the CNS even correlated with OCB profiles and IgG index, indicating that ASCs are likely responsible for intrathecal IgG production. These observations are supported by previous research, showing that OCBs are resulting from CNS-infiltrated plasmablasts and plasma cells [45] [48] [49] [50]. Recent studies revealed that MS patients with limited presence of CD20+ B cells and CD138+ plasma cells in white matter tissue show reduced intrathecal IgG production, based on IgG CSF/plasma ratios and OCB profiles, and have less severe clinical outcomes [10] [51]. Therefore, intrathecally produced IgG antibodies are likely associated with an unfavorable pathological profile in MS patients. Combined with our data, ASC presence in the MS CNS may thus be linked to more severe disease activity.
We found that CD45 expression was downregulated on ASCs when moving closer to the brain parenchyma and was lowest within MS white matter lesions. This reduction is most likely related to the maturation process of ASCs, since long-lived plasma cells are found to lose CD45 surface expression during their lifespan, indicating increased persistence [52] [53] [54]. Pollok et al demonstrated that long-lived plasma cells are able to persist in the chronically inflamed CNS, which appears to be provided with survival niches for these cells [55]. Survival niches within the MS brain likely promote the maturation of local ASCs. In MS brain donors, ASCs within the CSF were dominated by CD138− plasmablasts, although CD138+ plasma cells were increased in CSF versus blood samples. Supporting these data, short-lived plasmablasts were previously found to be highly present in MS CSF and were identified as a major effector population that contributes to active MS inflammation [56]. Furthermore, ASCs within the CSF of MS patients highly expressed CXCR3 compared to peripheral blood, which is consistent with our previous findings showing preferential migration of CXCR3+ B cells into the CNS [22]. The ASCs are thus likely derived from CXCR3+ B cells, which is further supported by other studies [35] [40] [57]. CXCR3 expression on ASCs in MS CSF correlated with IgG index, suggesting that CXCR3+ ASCs are primarily responsible for intrathecal IgG production in MS patients. This mirrors the increased capacity of CXCR3-expressing IgG+ B cells that accumulate in the blood of natalizumab-treated MS patients to develop into ASCs in vitro [22] [35].
We investigated multiple factors that may determine B-cell maturation within the MS CNS, including age. Brioschi et al demonstrated that so-called peripheral age-associated B cells infiltrate the CNS of aged mice and likely mature into autoantibody-secreting plasma cells [58]. Fransen et al showed limited CD20+ B-cell presence in the human MS brain with increasing age, which could imply more development into ASCs [10]. We found relative frequencies of ex vivo ASCs to decrease with age in CSF, meninges and white matter of patients with end-stage MS. Interestingly, ex vivo ASCs in MS white matter lesions did not correlate with age. MS white matter lesions probably contain a different microenvironment compared to normal(-appearing) white matter, in which other aspects such as T-cell presence are critical for maturation into ASCs.
In the periphery, TFH cells are found to be important for triggering autoreactive B cells to develop into CNS-infiltrating CXCR3(T-bet)+ B cells under influence of IFN-γ [19] [22]. Within the CNS, ectopic lymphoid follicles probably represent a critical site for B-cell reactivation, especially for subsets that have been triggered by IFN-γ as shown in mice [59] [60]. Locally, IL-21 and CD40L signaling from TFH cells might already be sufficient to induce maturation of memory B cells into ASCs [22]. The positive correlation between total ASC / B cell ratios and CD4+ / CD8+ memory T cell ratios in MS white mater lesions, in contrast to white matter, suggested that the presence of CD4+ memory T cells is indeed important for B-cell maturation into ASCs within MS lesions. We actually showed that local B- and T-cell interaction in MS lesions likely involves follicular-like T cells based on CXCR5 expression. Here, CXCR3+ B cells could possibly act as potent antigen-presenting cells to receive signals from CD4+ T cells that drive their maturation. Strikingly, we demonstrated that local B-cell maturation may depend on the type of T cell that is present in the MS brain.
This work has some limitations. First, we studied tissue of autopsy samples from donors with advanced MS. The extrapolation to living MS donors earlier in the disease is uncertain. Nevertheless, since OCB-presence is frequent at diagnosis, and we earlier showed presence of B cells in MS diagnostic biopsy samples [10], our findings likely reflect pathological events occurring already at the onset of MS. Second, the use of liberase for proper digestion of the meninges and brain tissue in our isolation protocol affected the expression levels of CD27, CXCR3 and CD138 on ASCs. Therefore, information on the ASC phenotype in these compartments is limited. Third, Ig gene expression and total ASC / B cell ratios have not been measured within brain tissue from the same MS brain donors, so local B-cell maturation could not be linked directly to local Ig production in the tissue. Lastly, we characterized follicular-like T cells in MS lesions as being dominantly CD4+ in our flow cytometry analysis but not in our immunohistochemical stainings. Despite these limitations, our study provides an in-depth analysis of B-cell maturation within paired blood, CSF, meninges and brain tissue from a large and very unique cohort of MS brain donors.