B cells undergo several stages of maturation during adaptive immune responses. Shortly summarized, B cells are generated in the bone marrow (pre/pro B cells) and then released into the peripheral blood (Cyster and Allen, 2019). Naïve B cells (antigen inexperienced) then migrate from the peripheral blood to secondary lymphoid tissues such as the spleen or lymph nodes, where they undergo further differentiation in germinal center reactions (Kurosaki et al., 2015; Cyster and Allen, 2019). Once bound to an antigen, B cells undergo a series of receptor and Ig subclass expression changes with co-stimulatory signaling by, e.g., T cell help. After this differentiation process, different subsets of antigen-experienced B cells emerge: memory B cells and plasmablasts. When memory B cells reencounter specific antigens, they undergo expansion and differentiate into plasmablasts mostly in the germinal centers (Kurosaki et al., 2015). Plasmablasts develop subsequently in either short-lived plasma cells or long-lived plasma cells which can maintain antibody production for decades without antigen re-stimulation (Lightman et al., 2019).

In contrast, another heterogeneous B cell group has the ability to suppress immune responses and are named regulatory B cells as a functionally defined population (Catalán et al., 2021). A definitive set of phenotypic markers are still lacking (Catalán et al., 2021). IL-10, IL-35, and TGF-beta secretion, the cell surface proteins CD1d and PD-L1 characterize the anti-inflammatory properties of this cell group (Catalán et al., 2021). Immature transitional B cells, divided into T1, T2, and T3 subpopulations are an intermediate stage between immature cells from the bone marrow and mature cells in the periphery (Catalán et al., 2021). T1 and T2 subtypes constitute a significant source of functional regulatory B cells (Zhou et al., 2020; Catalán et al., 2021). Autoimmune diseases are prone to have a lower frequency of regulatory B cells (Zhou et al., 2020).

Double negative (DN) B cells constitute another B cell population that lacks expression of immunoglobulin D and CD27 surface markers and has shown to be associated with autoimmune diseases (Sanz et al., 2019; Ruschil et al., 2021). The class-switched IgD- phenotype may indicate an antigen-specific maturation. Some author suggests that in the absence of CD27, a transition from a naive B cell seems unlikely (Li et al., 2021a). However, transcriptome analysis points toward a continuum of naive B-cells, memory B cells and plasmablasts (Ruschil et al., 2020), although the exact origin and maturation pathway of DN B cells is still unclear. Jenks et al. (2018) found two subgroups of DN B cells mainly based on the expression of the follicular marker CXCR5 (DN1: CXCL5 + and DN2: CXCL5- subtype), which is involved in the migration of B cells into B-cell follicles. CXCL5 + DN B cells are mostly expanded in elderly healthy individuals, while CXCL5- DN B cells were markedly found in active systemic lupus erythematosus (SLE), a defined autoantibody associated disease. In SLE, it has been further shown that the CXCL5- DN B cell subset develops from an activated naïve B cell pool. The lack of CXCR5 point toward an extrafollicular maturation pathway (Tipton et al., 2015; Jenks et al., 2018; Sanz et al., 2019). Compelling evidence suggests that CXCL5- DN B cell subset represents a primed precursor population for antibody-secreting cells (Jenks et al., 2018; Sanz et al., 2019). Although our studies didn’t differentiate these two DN B cell subpopulations, we have also shown that DN B cells are as well up-regulated in various auto-inflammatory neurological diseases including myasthenia gravis, Guillain-Barreì syndrome and NMOSD but not consistently in MS (Ruschil et al., 2020). We and others could show that this population most likely represents a transient precursor B cell population undergoing differentiation into antibody-secreting cells (Ruschil et al., 2020). Along these lines, we could show that the proportion of peripheral DN B cells is increased after vaccination and DN B cell-derived recombinant antibodies showed binding of specific vaccines, providing indirect evidence of their antibody secretion capacity (Ruschil et al., 2020).

The development of auto-reactive properties of B cells in autoimmune diseases remains a topic of great debate. While molecular mimicry is a recognized mechanism that can mislead B cells toward self-antigens, impaired self-tolerance during B cell development also contributes to auto-reactivity. The initial B cell repertoire generated by random V(D)J recombination undergoes a bimodal removal of autoreactive clones due to exposure to self-antigens (Meffre and O’Connor, 2019). This exposure occurs initially in the bone marrow, the site of B cell generation, and later in the periphery when B cells encounter a new set of self-antigens, resulting in the removal of autoreactive clones (Goodnow, 1996; Wardemann et al., 2003; Meffre and O’Connor, 2019). Distinct mouse models have shown that developing self-reacting B cells can be silenced through the following mechanisms: (1) clonal deletion; (2) clonal unresponsiveness to antigen or anergy; and (3) “receptor editing” or antigen receptor gene replacement by continued V(D)J recombination (Meffre, 2011; Stoehr et al., 2011; Meffre and O’Connor, 2019). Several lines of evidence suggest that central tolerance is likely dysregulated in NMOSD (Cotzomi et al., 2019; Meffre and O’Connor, 2019). The identification of pathogenic anti-AQP4 clones, which originate from unmutated autoreactive naive B cells in patients with NMOSD, is in potential agreement with this scenario (Meffre and O’Connor, 2019). In contrast, MS patients exhibit distinct B cell tolerance patterns compared to other autoimmune diseases. Here, an impaired peripheral B cell tolerance checkpoint is believed to be the main culprit, leading to the peripheral buildup of polyreactive mature naïve B cells, as shown by Kinnunen et al. (2013). Consistent with this assumption, regulatory T cells (Tregs) in MS patients seem to exhibit impaired suppressive activity and abnormally secrete interferon gamma (IFNγ) (Dominguez-Villar et al., 2011).

Molecular mimicry arises when peptides from pathogens display structural similarities with self-antigens. The presence of diverse pathogens, with each having its own potential unique molecular mimic to a CNS antigen, may elucidate why researchers have struggled to link a specific virus to, e.g., multiple sclerosis (Libbey and Fujinami, 2014). However, Epstein-Barr virus (EBV) has been identified as a potential viral agent that may trigger the production of autoreactive antibodies targeting GlialCam (Lanz et al., 2022). Nonetheless, more extensive evaluation of these findings is required. To the best of our knowledge, there is presently no conclusive evidence of established pathogens incorporating molecular mimicry mechanisms in NMOSD.

Regarding other mechanisms of B cell-mediated autoimmunity, B cells contribute to the development of diabetes through recognition of self-antigens with autoreactive antibodies and presentation of self-antigens via MHC class II molecules to T cells (Serreze et al., 1996). These findings indicate that self-antigen presentation by autoreactive B cells, which evade tolerance, could be the catalyst for the onset of autoimmune disorders. In multiple sclerosis, HLA class II alleles of the DR2 haplotype, DRB1*1501, DRB5*0101, and DQB1*0602, are well established genetic risk factors for MS and show a functional redundancy in Ag presentation (Sospedra and Martin, 2006). Thus, B cells serving as antigen presenting cells may shape an autoreactive T cell repertoire by presenting autoantigens by DR2 HLA-DR molecules (Wang et al., 2020).

Finally, altered cytokine levels that may result from a misdirected B cell activation can provide a pathogenic milieu for autoimmunity. Shortly summarized, serum IL-6 concentrations are significantly elevated in patients with NMOSD and are higher than in healthy individuals and patients with MS (Fujihara et al., 2020). Serum cytokine levels in MS do not show a clear proinflammatory profile, and several cytokines have even been shown to be downregulated (Lepennetier et al., 2019; Melamud et al., 2022). Within the CSF compartment, IL6 is also upregulated in NMOSD patients while CXCL13 seems to be a consistently up-regulated B cell-associated cytokine in MS (Sospedra and Martin, 2006). However, B cells are also able to secrete anti-inflammatory cytokines such as IL10 which is also upregulated in the CSF of NMOSD patients (Kaneko et al., 2018).

With the discovery of oligoclonal bands in the cerebrospinal fluid (CSF) of patients with MS, evidence pointed toward a pathophysiological role of B cells with potentially disease-driving antibodies in the CSF (DiSano et al., 2021). In MS, the majority of B cells in CSF are antigen-experienced B cells (Harp et al., 2007; Eggers et al., 2017), and the frequency of memory B cells is increased in CSF compared to peripheral blood (Eggers et al., 2017). In addition, B cell infiltration has been found within the brain parenchyma (Machado-Santos et al., 2018) and also in leptomeningeal aggregates, which are strongly associated with cortical lesions (Fraussen et al., 2014; Martin et al., 2016; Jain and Yong, 2022). Although Th1/Th17 T cells have at times attracted attention as potential therapeutic targets due to their important role in EAE models, specific CD4, Th1/Th17 immunotherapies have largely failed to show a clear impact on MS relapses (Baker et al., 2017c). In contrast, the high efficacy of B cell depletion in multiple sclerosis, first demonstrated for the CD20-specific B cell depleting agent rituximab (Hauser et al., 2008), was surprising and again highlighted the role of B cells. Subsequent clinical trials with ocrelizumab (Hauser et al., 2017; Vermersch et al., 2022), ofatumumab (Hauser et al., 2020), and ublituximab (Steinman et al., 2022) provide further evidence for the efficacy of (CD20) B cell depletion not only in relapsing but also in primary progressive multiple sclerosis (Kappos et al., 2011; Hauser et al., 2017).

The exact role of B cells in the pathophysiology of MS remains controversial. Epstein-Barr virus (EBV) infections have recently been strongly associated with multiple sclerosis, with 97% of patients in a large cohort showing positive EBV serum titers or seroconversion prior to the development of multiple sclerosis (Bjornevik et al., 2022). In addition, another study suggested molecular mimicry between EBNA1–a prominent EBV antigen–and GlialCAM (glial cell adhesion molecule), suggesting a direct role of pathogenic antibodies in MS (Lanz et al., 2022). However, only a limited number of antibodies reacted against both targets, so further confirmation seems necessary. Other potentially interesting targets for MS antibodies that have been proposed in recent years are chloride-channel protein Anoctamin 2 (ANO2) (Tengvall et al., 2019), which is a transmembrane protein for modulation for neural-excitability; alpha-crystallin B (CRYAB), which is expressed by oligodendrocytes and may have a protective effect by down-regulating the innate immune system (Thomas et al., 2023). Another group recently found antibodies against conformational membrane complexes containing the myelin proteolipid protein 1 (PLP1) (Owens et al., 2023). In addition to these recently described targets, a large number of autoantibodies have been described against various CNS cell types, including neurons, oligodendrocytes and astrocytes, and even immune cells (Fraussen et al., 2014). Although some of these possible antigen-antibody interactions seem to point toward an antibody-driven role of B cells in multiple sclerosis, multiple antigens could not be confirmed in further analyses, and one or a subset of clear antibody targets such as AQP4 in NMOSD have not yet been identified. An alternative function of B cells could be centered around antigen presentation and T cell stimulation. Our group (Kowarik et al., 2021) and other studies (Stern et al., 2014; Ruschil et al., 2021) have demonstrated that B cells not only traverse the blood-brain barrier but also recirculate in the peripheral blood through cervical lymph node drainage (Figure 1.). B cells, potentially primed against antigens in the CNS compartment during relapse, could thus possess the capacity to re-enter germinal centers in the periphery and perpetuate autoimmune circuits (Ruschil et al., 2021). Altogether, it remains unclear whether B cells predominantly produce autoantibodies against specific targets or act as antigen-presenting cells that circulate between the CNS and peripheral compartments; however, the diversity and inconsistency of suggested antigen targets might point toward a substantial antigen-presenting role.

Peripheral blood B cell subsets including total B cell numbers, naïve, memory B cells, double negative B cells and plasmablasts during stable disease do not show significant differences when compared to healthy controls (Kemmerer et al., 2020; Ruschil et al., 2020). The prevalence of transitional/regulatory B cells is often low (Zhou et al., 2020) while DN B cells did not show a consistent up-regulation in MS (Fraussen et al., 2019; Ruschil et al., 2020). Besides the broad depletion of circulating B cells by anti-CD20 antibodies such as rituximab, ocrelizumab and ublituximab, several MS treatments have shown to also exert profound effects on peripheral B cells and cerebrospinal fluid (Tables 1, 3). The absolute number of B cells in MS treatments has been shown to be slightly reduced during dimethyl fumarate, fingolimod, and siponimod treatment, unchanged during glatiramer acetate and interferon beta treatment and increased during natalizumab treatment (Kemmerer et al., 2020; Traub et al., 2020). Natalizumab is a monoclonal antibody against the cell adhesion molecule α4-integrin, which is highly expressed in B-cells (Saraste et al., 2016). The increase in peripheral B cell number during natalizumab most likely relies on the egress of memory B cells from the marginal sinus of the spleen through the blockade of integrins by which memory B cells attach to the sinus (Kowarik et al., 2021). However, these cells are also impaired in their ability to cross the blood-brain-barrier so that natalizumab treatment has to be considered separately (Kowarik et al., 2021). Further differential flow cytometric analyses showed, that in most treatments, the fraction of naïve B cells is increased, while the percentage of memory B cells is significantly decreased (Kemmerer et al., 2020; Traub et al., 2020). Regulatory B cells show consistently elevated percentages during most treatments while DN B cells show unchanged percentages or an elevated proportion during fingolimod therapy (Kemmerer et al., 2020). Plasmablast percentages show different patterns or are unchanged during treatment with cladribine, interferon beta, dimethyl fumarate or glatiramer acetate. Further analyses by B cell repertoire mass sequencing or whole transcriptome analysis underlined that memory B cells are significantly affected during cladribine treatment (Ceronie et al., 2018; Rolfes et al., 2022; Ruschil et al., 2023) and also alemtuzumab treatment (Ruck et al., 2022). Data regarding changes in CSF immune cell subsets are limited and differences are difficult to assess due to the overall low number of immune cells. However, treatment with dimethyl fumarate, natalizumab, rituximab, ocrelizumab, and alemtuzumab resulted in reduced CSF B cell counts, whereas fingolimod did not alter the proportion of CSF B cells (Table 3). Plasmablasts were reduced during treatment with dimethyl fumarate, natalizumab and fingolimod (Table 3).

In contrast to these approved treatments, other drugs that also affect B cells have been shown to be ineffective or even worsen MS. For example, atacicept was stopped in the ATAMS trial because of a pronounced conversion to MS in patients with optic neuritis. Although the exact mechanisms regarding B cells was not elucidated in the study, an increase in IL15 provided some evidence suggesting stimulation of memory B cells as a possible explanation for the clinical outcomes observed in the study (Kappos et al., 2014). In addition, the use of anti-TNF blockers such as infliximab, which can stimulate memory B cell activity, has been associated with an increased incidence of MS in patients with chronic disease (Avasarala et al., 2021). Regarding B cell depleting treatments, it is important to note that regulatory B cells may also be depleted, but this does not seem to drastically limit the therapeutic potential.

Interestingly, numerous MS treatments influence T cell function and T cell subset distribution (Martin et al., 2016), however, a direct effect on B cell populations seems to have an even more relevant effect on the disease course. The consistent effect of MS treatments on the memory B cell subset could further underpin this assumption due to their frequent occurrence in the CSF and their ability to recirculate into the periphery and to repeatedly participate in germinal center reactions. Of note, memory B cells are the primary site of persistent latent EBV infection which could partially explain the association between EBV infections and multiple sclerosis (Tracy et al., 2012).

Neuromyelitis optica spectrum disorder has been recognized as a separate disease entity with the discovery of autoantibodies against the water-channel aquaporin-4 (AQP4-AB) (Lennon et al., 2004). It could be clearly demonstrated that AQP4-AB bind to AQP4 channels on astrocytes triggering an activation of the complement cascade, with granulocyte, eosinophil, and lymphocyte infiltration, resulting in astrocyte damage. As a secondary event, oligodendrocyte injury leads to demyelination and neuronal loss (Carnero Contentti and Correale, 2021). Lineage analysis of AQP4- specific B cells from the peripheral blood and CSF B cells of NMOSD patients showed a clonal relationship with memory B cells, plasmablasts and DN B cells in the periphery during active disease. Immunoglobulin transcriptome analysis further indicated that expanded DN B cells undergo antigen-specific B cell maturation and are closely linked to AQP4-specific CSF B cells (Kowarik et al., 2017). Although it is believed that mis-priming and/or escape from tolerance mechanisms of peripheral B cells and the peripheral secretion of AQP4-AB might initiate NMOSD disease pathology, AQP-4 specific CSF plasmablasts have been shown to originate from peripheral B cells and intrathecally secrete AQP4-AB during active disease and thus might contribute to disease exacerbation (Kowarik et al., 2015).

Besides the peripheral up-regulation and association of DN B cells and AQP4-reactive CSF plasmablasts in active NMOSD, DN B cells have received increasing attention in recent years, especially in SLE, where they have been found to be a marker of disease severity (Jenks et al., 2018; Szelinski et al., 2022). DN B cells are also elevated in the elderly, in infections and in other autoimmune diseases such as rheumatoid arthritis, Guillain-Barre syndrome and myasthenia gravis (Fraussen et al., 2019). DN B cells (CXCR5-) are extensively expanded in antibody-mediated autoimmune diseases such as SLE, where a worse disease course is correlated with an inflated population of DN B cells (CXCR5-), which are thought to represent plasmablasts precursors (Jenks et al., 2018; Szelinski et al., 2022). When co-cultured with Th cells, DN B cells have the capacity to differentiate into antibody-secreting cells (Janssen et al., 2020; Hoshino et al., 2022). Conversely, most of DN B cells in MS are not CXCR5-, indicating a different mechanism from that observed in NMOSD and SLE (Li et al., 2021a).

In the peripheral blood, plasmablasts (CD19intCD27highCD38highCD180-) have been shown to be up-regulated in NMOSD and secrete AQP4-AB following IL6 stimulation (Chihara et al., 2013). This dysregulatory shift toward antibody-secreting cells has been reaffirmed by different studies (Hoshino et al., 2022). As mentioned above, peripheral DN B cells have also been shown to be upregulated in the peripheral blood of NMOSD patients (Ruschil et al., 2020). Regulatory B cells are significantly reduced in AQP-4 positive patients compared to MS patients, possibly due to the high IL6 secretion, which subsequently inhibits the generation of regulatory B cells (Quan et al., 2013).

Most approved therapies currently target effector B cell lineages and the direct interaction caused by antibodies. B cell depletion, including the use of rituximab as an anti-CD20 antibody, as well as inebilizumab targeting CD19, has demonstrated effectiveness in treating NMOSD (Barreras et al., 2022; Nie and Blair, 2022). As DN B cells and plasmablasts lose CD20 expression, targeting the consistently expressed CD19 marker on both cell types may result in a more profound depletion and improve treatment effects (Agius et al., 2019). After receiving treatment with rituximab, the presence of regulatory B cells increases (Quan et al., 2015). Satralizumab and tocilizumab both inhibit the IL6 receptor, disrupting lymphocyte activation (Chu and Huang, 2022). Tocilizumab reduces memory B cells in the peripheral B cell subset, while regulatory B cells and plasmablasts remain unaffected (Traub et al., 2020). Eculizumab and ravulizumab are inhibitors of complement factor 5 and disrupt the complement signaling cascade initiated by anti-AQP4 antibodies. Eculizumab reduced the percentage of memory B cells in patients with myasthenia gravis (Li et al., 2021b).

The wide development of MS medications has led to experimental usage of these therapies in NMOSD in the past when no approved medications for NMOSD were available. Several medications have failed to show positive treatment effects or even worsened NMOSD disease course in single patients or small case series. Natalizumab and fingolimod appeared to increase the proportion of DN B cells in the periphery (Kemmerer et al., 2020), potentially clarifying why these drugs have not been shown to be effective in treating NMOSD. Along these lines, paradoxical rebound under rituximab therapy in NMOSD patients may be explained by an increase of CD20-negative DN B cell/plasmablasts and an asynchronous B cell depletion (Kowarik et al., 2017). Other approved MS drugs that were not effective or even harmful when used in single NMOSD cases included alemtuzumab, dimethyl fumarate, glatiramer acetate, interferon-β, fingolimod and natalizumab. Analyses of peripheral B cell subsets reveals that the mentioned medications might increase the proportion of plasmablasts, as wells as B cells supporting the pathophysiology of NMOSD. Some medications also increase serum interleukin-6 levels and serum BAFF levels, which could contribute to the pro-inflammatory reaction, worsening the disease course (Traub et al., 2020).