Translational Mini-Review Series on B cell subsets in disease. B cells in multiple sclerosis: drivers of disease pathogenesis and Trojan horse for Epstein–Barr virus entry to the central nervous system?

Abstract

OTHER ARTICLES PUBLISHED IN THIS MINI-REVIEW SERIES ON B CELL SUBSETS IN DISEASE

Transitional B cells in systemic lupus erythematosus and Sjögren's syndrome: clinical implications and effects of B cell-targeted therapies. Clinical and Experimental Immunology 2012, 167: 7–14. Reconstitution after haematopoietic stem cell transplantation – revelation of B cell developmental pathways and lineage phenotypes. Clinical and Experimental Immunology 2012, 167: 15–25.

The recent success of therapies directed at B cells has highlighted their potential as central players in multiple sclerosis (MS) pathogenesis. Exciting new data showed that B cell depletion led to reduced clinical and magnetic resonance imaging (MRI) evidence of disease activity. However, the mechanisms of action remain unknown, but could involve autoantibody production, antigen presentation and/or cytokine production by B cells. Another exciting line of investigation in the field of MS comes from latent infection of memory B cells by Epstein–Barr virus (EBV). These cells are hijacked as ‘Trojan horses’ and ‘smuggle’ the virus into the central nervous system (CNS). Thus, these new anti B cell treatments will also be likely to have anti-viral effects. We briefly review recent findings in the field of MS pathogenesis, and highlight promising new targets for therapeutic intervention in MS.

Introduction

Multiple sclerosis (MS) is an inflammatory and neurodegenerative disorder of the central nervous system (CNS). While it consistently shows genetic associations with human leucocyte antigen D-related 2 (HLA-DR2), those with -A3 are more controversial. Its prevalence is higher towards the North and South Poles than the Equator, and migration studies have implicated a possible encounter with unknown environmental factors before the age of 15 years [1]. In most patients, MS follows a relapsing–remitting course (RRMS), often with substantial functional recovery between relapses. Progressive MS (PMS), whether primary (PPMS) or secondary (SPMS), is characterized by the gradual increase in disability from which there is little or no recovery. MS was considered a white matter disease, but more recent studies have shown that grey matter can also be seriously affected. MS is thought to be an autoimmune disorder, in which the immune cells enter the CNS and attack the myelin sheath covering the neurones, causing demyelination and, eventually, axonal damage. Demyelination leads to a variety of sensory and motor symptoms, such as optic neuritis, numbness, fatigue, spasticity, muscle weakness and cognitive impairment [2]. An autoimmune basis is supported by the mouse model experimental autoimmune encephalomyelitis (EAE), evoked by immunization with myelin antigens (e.g. spinal cord homogenate) in Freund's adjuvant. EAE is a T cell-driven disease. Work on the resulting MS-like disease in the mouse model has suggested novel potential pathogenetic pathways and therapeutic agents, but these could not always be translated to the human disease [3]. The pleiotropic function of B cells (Fig. 1) and their potential involvement in MS pathogenesis has been overshadowed by the emphasis on T cell research in the last decade. However, recent exciting results with B cell-depleting agents highlight the pathogenetic roles for key players other than T cells.

Fig. 1.

Pleiotropic function of B cells in the regulation of immune responses and multiple sclerosis (MS). B cells are involved in cellular and humoral immune responses. They also act as host for latent Epstein–Barr virus (EBV) infection.

Evidence for B cell involvement in MS

The presence of B cells in MS lesions

MS research is complicated by the inaccessibility of its target organ during life. Much of the work, therefore, has focused on post-mortem brains. It has been helped by the typical mixture of old and new white matter lesions in affected MS brains. Peripheral B and T cells are numerous in white matter lesions, being frequent in acute lesions and the active margins of chronic active lesions, rather than in inactive lesions [4–7]. The characteristic inflammatory infiltrates of B, T, dendritic and plasma cells are primarily perivascular [8–11]; however, CD8⁺ T cells, in particular, tend to invade into the surrounding parenchyma. T helper type 1 (Th1) and CD4⁺ and CD8⁺ T cells expressing interleukin (IL)-17 are found in perivascular areas [6,12]. CD4⁺ cells were found mainly in perivascular spaces and the meninges, where B cells were also detected [5,8,13–15].

The presence of oligoclonal bands in cerebrospinal fluid

Much information has come from analysing cerebrospinal fluid (CSF); it occupies the subarachnoid space just outside the pia mater that tightly ensheathes the brain and spinal cord and lines the ventricles. During life, tapping CSF is the most practical way of sampling the CNS milieu. In MS patients, there is evidence of persistent intrathecal B and plasma cell activation [16,17]. The characteristic oligoclonal immunoglobulin bands (OCBs) are defined as two or more independent immunoglobulin (Ig)G bands in the electrophoretic gamma region in CSF but not serum. They are found in most patients with MS and imply an immune-mediated pathology, possibly of infectious nature. However, OCBs are also present in other inflammatory diseases of the CNS, e.g. subacute sclerosing panencephalitis, where they are directed against measles virus [18]. The nature of OCBs in MS remains a mystery; we still do not know if they derive from a small but random sample of peripheral B cells or from some pathogenic response. However, B cell frequencies are very low in the CNS and only the arrival of new and sensitive techniques, such as polymerase chain reaction (PCR), enabled the analysis of their maturation and developmental status. Earlier studies analysed the diversity of the third complementarity determining region (CDR3 gene fragments) of these CSF B cells and found intrathecal expansion in MS patients. Furthermore, these B cells were T cell-dependent hypermutated post-germinal centre antibody-forming or memory cells that had been positively selected through their antigen receptor [19]. Interestingly, V(D)J genes utilized by peripheral and central B cells differed, which is indicative of compartmentalized clonal expansion [20]. Intensive analysis revealed that CSF antibodies did not bind to myelin-basic protein (MBP), proteolipid protein (PLP) [17] or common viruses [21]; instead, some of them bound to targets on oligodendrocytes and astrocytes [22]. Somatic hypermutation of Ig transcripts in the CNS imply a local antigen-driven T cell-dependent process [23]. More recent studies showed that B cells are antigen-experienced, and identified different clonotypes in different plaques from the same individual [22]. Mutated B cells from MS lesions might sequentially colonize germinal centres (GC) in secondary lymphoid organs, undergo reactivation and then invade other brain regions. GC are the classic sites where mature B cells respond to antigen-bearing follicular dendritic cells (plus helper T cells), hypermutate their antibodies through somatic hypermutation and then migrate from the dark to the light zone, where they also class-switch and generate memory and plasma cells.

In MS, clonally related B cells populate meninges, inflammatory lesions, normal appearing white matter and CSF and CNS-resident B cells shared between CSF and CNS produced antibodies, which can be detected in the CNS [24,25]. Indeed, there are follicle-like structures in the meninges in secondary progressive MS patients [13–15,26] that have attracted much recent attention. If their suspected GC functions are confirmed, they may provide novel clues to the pathogenesis of MS.

Latently Epstein–Barr virus (EBV)-infected B cells in the CNS and their potential role in neuroinflammation

Another interesting line of investigation is the role of B cells as hosts for EBV. First isolated from Burkitt's lymphomas in 1964 [27], its causal role in infectious mononucleosis (IM) was discovered by accident 4 years later. A laboratory technician working with lymphoma samples contracted EBV, seroconverted and developed IM. More than 90% of the population is infected with EBV by age of approximately 20 in Europe and much earlier in developing countries [28]. Whereas infection in childhood is mainly asymptomatic, the presentation is typical of IM in approximately half of first infections in young adults.

EBV is a B cell-trophic γ-herpesvirus which establishes latency in B cells [29], initially in submucosal B cells in tonsils [30], which are subsequently transformed into Ig-producing B cell blasts [31]. Cytotoxic T lymphocyte cells (CTLs) are pivotal in eliminating these highly expanded EBV-infected B blasts during acute disease [32,33]. EBV remains immunologically silent in small numbers of B cells (1/10⁵) in the blood [34]. The virus periodically reactivates, leading to virus shedding in the saliva and blood, but this is tightly controlled by CTLs to prevent lymphoproliferation. However, some infected cells may escape, leading to such neoplasms as Burkitt's lymphoma, Hodgkin's disease, nasopharyngeal carcinoma and post-transplant lymphoproliferative disorder (PTLD) [35,36]. Altogether, this virus is very successful at hijacking B cell biology.

Among autoimmune diseases, EBV infection has been implicated in systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA) [37], although the high levels of EBV seropositivity in adults make it hard to establish such associations unequivocally. In fact, several findings implicate EBV in MS. A history of symptomatic IM, as opposed to subclinical primary EBV infection, increases the risk of developing MS more than twofold [38,39]. In addition, increases in serum antibody titres to EBV nuclear antigen 1 (EBNA-1) precede the onset of MS symptoms by several years, and are associated with active magnetic resonance imaging (MRI) lesions in established disease [40,41]. Furthermore, a single MS patient-derived T cell receptor cross-recognizes peptides from myelin basic protein or EBV when presented by two different but related HLA-DR2 molecules [42]. Finally, a recent study reported an accumulation of EBV-infected B cells in post-mortem brain samples from patients with MS, but not in other inflammatory central nervous system diseases [26]. However, other studies could not confirm these findings [43–45], and found no evidence of active EBV infection, which may not be a characteristic feature of the MS brain. Furthermore, as the virus resides in memory B cells, which traffic into inflamed tissues, its presence could be a bystander phenomenon and easily misinterpreted. Hence, it remains controversial whether EBV is truly involved in the initiation or evolution of MS, e.g. as a result of changes in its behaviour or an underlying immunopathogenesis in MS, and what other environmental and genetic factors are also contributing.

If EBV is finally incriminated, how could latent infection play a role? We selected post-mortem white matter MS lesions of different activity, grouped according to B cell content and expression of the innate cytokine interferon (IFN)-α, which proved to be over-expressed in active lesions. We looked for the presence of latent EBV infection by in-situ hybridization, a highly sensitive and specific method that targets the small non-coding RNAs of EBV expressed during all latency programmes, and is used as the gold standard for EBV detection [46]. We identified EBV non-coding small RNA (EBER)⁺ cells in B cell-rich active MS lesions with IFN-α over-expression [47]. However, these findings were not exclusive to the MS brain, as EBER⁺ cells were also found in cases of stroke. We proposed a more indirect mechanism by which latent EBV infection could contribute to neuroinflammation: that these small RNAs bind to Toll-like receptor 3 and potentially other intracellular receptors such as retinoic acid-inducible gene 1 (RIG-I) and thus stimulate IFN-α production in active MS lesions (Fig. 2). A recent study showed that EBERs were indeed released from EBV-infected cells and acted as local immunomodulators [48]. Could innate activation triggered by latent EBV infection be part of the game? Perhaps we have to think differently – EBV might be more subtle than we anticipated. After all, it is a persistent virus selected to co-exist with the host rather than endanger it.

Fig. 2.

Recruitment of B cells into active lesion in multiple sclerosis (MS). The B cell infiltrate may contain the occasional Epstein–Barr virus (EBV)⁺ B cells. These cells contain EBV non-coding small RNAs (EBERs), which can be secreted in a complex with the cellular EBER binding lupus antigen (La). EBERs can bind to Toll-like receptor (TLR)-3 on neighbouring cells and elicit interferon (IFN)-α production, thereby contributing to an inflammatory milieu. These processes may help to sustain an inflammatory response in the central nervous system (CNS) and contribute to demyelination and axonal injury.

The success of B cell therapies in MS

In a small Phase II trial with rituximab (anti-CD20), there was a dramatic reduction of disease activity in RRMS patients within 48 weeks [49]. Rituximab is a genetically engineered chimeric ‘humanized’ molecule that targets CD20⁺ B cells and is used for treating B cell lymphoma. CD20 is present on B cells and pre-B cells but lost upon plasma cell differentiation [50, 51]. The primary end-point of this trial was mean gadolinium (Gd)-enhancing lesions (inflammatory activity) assessed by MRI from baseline to week 48. A decrease in disease activity was already noted at week 4 and most pronounced at week 12. Such very early treatment responses suggest that rituximab treatment may act directly via B cell lysis – or, indeed, on the inflammatory mechanisms – rather than by reducing pathogenic autoantibody levels. Indeed, rituximab does not affect serum and CSF antibody levels [52].

Interestingly, in a trial on PPMS, the primary end-point was not reached; however, there was a suggestion of an effect in subjects with evidence of active inflammation [53]. Treatment with rituximab led to predominance of circulating naive and immature B cells. In the CSF, T and B cell numbers were decreased, and resting B cells predominated.

Two additional humanized antibodies targeting different epitopes on CD20 are now being trialled in MS: ofatumumab and ocrelizumab [54]. Ocrelizumab appears to target mature B cells. It has reached Phase III for several autoimmune diseases, e.g. rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), and Phase II for MS. Those for RA and SLE were halted in May 2010 because of occasional serious/fatal opportunistic infections in high-dose arms, especially in subjects with Asian ancestry. The Phase II study in RRMS in October 2010 showed statistically significant reductions at week 24 in both lesion load (as measured by MRI activity) and relapse rate, compared to placebo, both doses (200 mg and 600 mg) being well tolerated.

The rituximab trial data were somewhat surprising, as a positive effect was already detected at week 12 post-treatment. This finding highlights the potential for an autoantibody-independent effect of B cell depletion on MS disease activity. B cells are important antigen-presenting cells. Physical interaction of B cells and T cells [major histocompatibility complex (MHC)/antigen/T cell receptor] occurs in the presence of co-stimulatory molecules such as CD40/CD40ligand, B7/CD28, OX40 ligand/OX40 on the surface of B cells and T cells, respectively [55]. B cell depletion in mice was found to impact on CD4⁺ T cell activation and expansion in vivo, which may explain its positive effect on multiple T cell-mediated autoimmune diseases, including MS [56] and type 1 diabetes [57]. It remains to be seen whether B cell-depleting strategies may alter the ratio of CTL : infected targets cells favourably, and thus enable better control of EBV infection. Furthermore, B cells have the ability to regulate T cell function and inflammation through cytokine production. A recent study found that B cells of MS patients had altered cytokine responses, e.g. increased ratio of lymphotoxin (LT) : IL-10 and increased secretion of tumour necrosis factor (TNF)-α and LT when exposed to the proinflammatory cytokine IFN-γ or bacterial cytosine–guanine dinucleotide (CpG)-DNA bound to Toll-like receptor 9 [58]. Interestingly, CD4 and CD8 T cells of MS patients produced significantly fewer proinflammatory Th1/Th17 cytokines after in vivo or ex vivo B cell depletion. B cell depletion may, therefore, be effective in reducing CNS inflammation. However, B cells also play an important role in immunoregulation. Animal studies highlight the importance of the IL-10-producing B cell subset (B10) in the suppression of autoimmunity and inflammation [59], which may explain why B cell depletion led to the worsening of inflammatory disease in some EAE models, with delayed production of IL-10 and emergence of regulatory T cells [60]. B cell depletion also exacerbated disease in myelin–oligodendrocyte glycoprotein peptide (MOG p35–55)-induced EAE in mice [61]. Hence, the relative contribution of B cells to EAE and MS may vary depending on the stage of disease progression, highlighting the existence of an intricate cross-talk between T and B cells.

Conclusion

Exciting MS treatments are currently in the pipeline, which reflect important roles for B cells as drivers of MS pathogenesis, an area overshadowed by the emphasis on T cell research in the last decade. Furthermore, the eradication of EBV⁺ B cells by B cell-depleting strategies is another interesting line of investigation. B cell depletion may also impact on the propensity of latent infections to contribute to neuroinflammation in the CNS, and we may want to test anti-viral strategies in MS directly using drugs which can cross the blood–brain barrier; however, treatment success may also depend on the stage of disease progression.

We are just starting to unravel the complex interplay between MS risk factors, which we hope will help us to understand the pathophysiology of MS.

Disclosure

J. Tzartos and G. Khan report no disclosure. U.-C. Meier has received research support from British Technology Group. G. Giovannoni has received consulting fees from Bayer-Schering Healthcare, Biogen-Idec, Fiveprime therapeutics, GlaxoSmithKline, Ironwood Pharmaceuticals, Merck-Serono, Novartis, Protein Discovery Laboratories, Teva-Aventis, UCB Pharma and Vertex; lecture fees from Bayer-Schering Healthcare, Biogen Idec, and Teva-Aventis; and grant support from Bayer-Schering Healthcare, Biogen-Idec, Merck-Serono, Merz, Novartis, Teva-Aventis, and UCB Pharma.