Regulatory B and T lymphocytes in multiple sclerosis: friends or foes?

Tregs can be subdivided into thymus-developed, “natural” Tregs that mediate tolerance to self-antigens and “induced” Tregs derived from conventional CD4(+) T cells following non-self antigenic exposure [48]. Natural Treg production requires stable expression of FoxP3 and high-affinity binding of HLA/self-peptide complex on thymic antigen-presenting cells (APCs) to T cell receptor (TCR). Natural Tregs can be also sub-classified into CD45RA(+)“naïve” Tregs and CD45RO(+)“memory” Tregs [49]. ‘Induced’ or ‘adaptive’ Tregs (iTregs) are generated from naïve T cells in the presence of transforming growth factor-β (TGF-β) or retinoic acid and produce the anti-inflammatory cytokine IL-10 [50‐52]. Despite the phenotypic and functional overlaps with natural Tregs, iTregs demonstrate apparent differences in stability and gene expression [53]. Type 1 regulatory T cells (Tr1), are a subpopulation of Tregs-expressing CD4(+)CD49(+)LAG-3(+)IL-10(+) that exert significant immunosuppressive effects [54‐56]. In addition, CD8(+) Tregs (Tr2) and IL-17-producing Tregs that share some common features with Tr1 also exist [57]. iTregs-expressing RORγt, which is the master regulator of antimicrobial type 3 immunity are termed type 3 Tregs (Tr3) [58, 59]. These cells constitute the major population of colonic Tregs, require bacterial antigens for differentiation and are distinct from thymus-derived Tregs.

Tregs have a pivotal function in regulating the immune system by controlling the number and function of effector cells. Thus, they play a major role in suppressing unwanted autoreactive immune responses, such as in the case of autoimmunity [60]. Interestingly, it has been indicated that Tregs can modulate both adaptive and innate immune systems, and once activated they specifically regulate immune responses at multiple levels and by various mechanisms. These suppressive mechanisms can be organized into major groups, including cell–cell contact-dependent suppression, inhibitory cytokine release (such as IL-10 and TGF-β), modulation of APC function, cytolysis, metabolic disruption and induction of suppressor cells or “infectious tolerance” [53].

In addition to IL-10, the inhibitory cytokine IL-35 also contributes to regulatory T cell function [61, 62]. IL-35 belongs to IL-12 family of cytokines that includes IL-12, IL-23, IL-27 and IL-35. Of these, IL-12 and IL-23 have pro-inflammatory roles, whereas IL-35 appears to exert a more regulatory function by inducing the expansion of Tregs and Bregs subsets and inhibiting Th17 cell differentiation [63]. IL-35-producing Tregs represent a distinct effector population from the IL-10-producing iTregs which also have different transcription factor dependency, as differentiation regulator Blimp1 is essential for IL-10 production, but not for IL-35, whereas Foxp3 is important for IL-35 but dispensable for IL-10 production [64]. Recently, it was demonstrated that the IL-12p35 alpha subunit of IL-35 efficiently suppressed encephalitogenic T cell responses and ameliorated experimental autoimmune encephalomyelitis (EAE), a well-characterized murine model of MS [65]. IL-12p35 inhibited the expansion of pathogenic Th17 and Th1 cells and mediated the expansion of Tregs and Bregs [65].

The role of Tregs in MS has been thoroughly investigated in EAE [66]. EAE shares many features with the human disease and has thus revealed much information that led to the development of many approved therapies for MS [67, 68]. A correlation was found between antigen-specific Tregs and disease resistance [69]; similarly, transfer of Tregs to EAE-induced mice reduced the severity of the disease [70]. In addition, depletion of CD25(+) T cells reduced the antigen burden required to induce EAE and prevented disease recovery [71]. Furthermore, Tregs were also involved in the regulation of cell transmigration across the BBB [72]. The EAE model system was also exploited for in vivo silencing of certain microRNAs such as miR26a, which increases the expression of Th17-related cytokines and establishes more severe EAE [73]. In contrast, overexpression of miR26a is associated with decreased expression of Th17-related cytokines, positive correlation with Treg FoxP3 and less severe disease [73].

Many treatment regimens increase Tregs and have also been proven quite successful at the experimental level (in EAE) [74]. For instance, IDO was shown to upregulate Tregs via tryptophan catabolite and to suppresses encephalitogenic T cell responses [75]. Further evidence suggested that vitamin A and its active metabolites (all-trans-retinoic acid and 9-cis-retinoic acid) could restore the imbalance of Th17 and Treg cells and can be considered as a promising target in the prevention of EAE [76]. Expansion of Tregs also appears promising. Lately, it was shown that engineered clonal MBP-specific Tregs ameliorated EAE in myelin oligodendrocyte glycoprotein (MOG)-immunized DR15 transgenic mice [77]. Administration of antigen encapsulated within tolerogenic nanoparticles (tNPs) comprising biodegradable polymer is also capable of inducing Ag-specific Tregs [78]. tNP-treated mice did not develop EAE following adoptive transfer of encephalitogenic T cells [78].

It has become apparent that Tregs are also implicated in the pathophysiology of MS in humans [53, 74]. Although increased frequencies of Tregs are found in the cerebrospinal fluid but not peripheral blood of MS patients [79], alterations in Treg homeostasis [80, 81] and their functional impairment are documented [82‐84]. Interestingly, their functional defects are more profound in RRMS than in SPMS [85, 86]. Analysis of the thymic export activity in MS patients revealed impaired release of newly formed T cells into the periphery resulting in an imbalance of circulating Tregs [87]. This thymic functional impairment is compensated by peripheral post-thymic expansion, creating a shift from naïve Tregs to memory Tregs in MS patients. Researchers argue that this shift may account for the impaired suppressive function of Tregs in MS.

Another possible mechanism for the functional failure of Tregs appears to be pro-inflammatory cytokines, such as IL-12, which are up-regulated in MS [88]. IL-12 has the ability to change the phenotype and function of Tregs by inducing IFN-γ production. IFN-γ-producing Tregs display a decline in their suppressive activity in vitro, as IFN-γ blockade significantly boosted their suppressive ability but did not affect control Tregs [89]. The increased percentages of Th1-like Tregs may partly account for the lack of suppressive function Tregs of MS patients. These data illustrate the phenomenon of enhanced Tregs plasticity toward a pro-inflammatory, cytokine-producing effector phenotype [90]. Skewed IFN-γ-producing Th1-like Tregs play significant role in MS and also other autoimmune diseases [53, 91] and malignancies [92]. Importantly, IL-12 dependent IFN-γ production of Tregs could also be mimicked in vitro in Tregs from healthy subjects creating Th1-like Tregs that resembled a classical Th1 phenotype [89]. Moreover, Th17-like Tregs which expanded in the presence of IL-6 and IL-1β have also been documented [93, 94].

Apart from FoxP3(+) Tregs, the role of Tr1 cells’ role in MS seems to be equally important. In MS, Tr1 cells were reduced in CD46-activated T cells [83], which are known to acquire a Tr1 phenotype [95]. Furthermore, IL-10 production from CD46- activated T cells was almost absent, while IFN-γ production was not affected in these cells. It can, therefore, be concluded from these findings that MS is associated with multiple defects in regulatory T cell populations [83].

B cells can play a regulatory role in EAE pathophysiology, as mice with genetically deficient B cells cannot recover from the disease, whereas transfer of IL-10-producing B cells suppresses EAE symptoms [124, 125]. For instance, Bregs, transduced into mice with EAE, accumulated in the spleen and mesenteric lymph nodes, leading to an expansion of Tregs and Tr1 cells in vivo [132]. Importantly, Tregs and Tr1 s were also enriched in the CNS of the same littermates. In the EAE model again, treatment with MOG protein fused to reovirus protein σ1 (MOG–pσ1), resulted in an expansion of IL-10-producing B220(+)CD5(+) Bregs, which restored Tregs and facilitated the rapid improvement of EAE [133]. Additionally, PD-L1^Hi Bregs transferred to afflicted animals suppressed the disease. In total, Bregs, in contrast to effector B cells, protect from the development of EAE, by suppressing pro-inflammatory cytokines and the transmigration of activated cells to the CNS [97, 134, 135].

There is no consensus on Breg numbers in autoimmune diseases. In most diseases or disease states, Bregs are reduced [136‐140] but increased numbers were also reported [105]. In MS in particular, Bregs are reported to be numerically decreased [141, 142], unaltered [143, 144] or increased [145]. A representative phenotypic flow cytometric analysis of Bregs in RRMS is shown in Fig. 1. Irrespective of their numbers, Bregs function is impaired in MS patients, as IL-10 production and suppressive function of B cells are reduced [21, 146‐148]. In addition, the proportion of naïve Bregs in disease relapses is reduced, leading to an increased memory/naïve ratio [141]. Whether this reduction is the cause or the consequence of disease relapse remains to be seen. Recent data also have indicated that reduced peripheral blood Breg levels were not associated with the Expanded Disability Status Scale score in MS [149].

A novel type of Bregs, CD19(+)CD25(+) cells, was described in both healthy subjects and MS patients [112, 145]. This new subtype seems to be numerically increased in MS compared to healthy controls, and also in relapse compared to disease remission [135]. It is apparent that much more research is needed to illuminate the role of different Breg subsets in MS [150].

B cell-depleting therapies in MS focus on two main targets, CD20 and CD19. Monoclonal anti-CD20 includes rituximab, ocrelizumab and ofatumumab, which differ in their CD20 epitope recognition and in the intensity of their action [151]. CD20 is expressed on most B cells, from PreB to IgG memory B cells, while leaving plasmablasts and ProB cells mostly lack expression. All anti-CD20 therapies cause an almost complete extinction of B cell subtypes in peripheral blood [152]. B cell repopulation begins several months post-treatment and appears to be inclined toward a more naïve and regulatory phenotype [153]. B cell depletion also suppresses T_H1 and T_H17 responses and increases circulating Tregs. In general, this therapy has shown to ameliorate disease symptoms and activity, and reduce relapse rate, including reduction in gadolinium-enhancing lesions on brain MRI [154, 155].

Anti-CD19 monoclonal therapy also looks promising. IgG1 anti-CD19 antibody (MEDI-551) [156] recently entered a phase II clinical trial in MS. CD19 is expressed on all B cells and is progressively lost in terminally differentiated plasma cells [157]. As a result, it comes as no surprise that MEDI-551 induced a longer-lasting B cell depletion than rituximab, while also reducing immunoglobulin serum levels, including autoantibodies [155]. Anti-CD19 therapy in EAE-induced mice suppressed disease severity and duration and increased circulating Tregs, whereas potentially protective CD1d(hi)CD5(+) Bregs displayed resistance to depletion [158]. Hence, MEDI-551 is expected to have similar effects to anti-CD20 therapy on MS patients and could be approved for the treatment of MS in the future especially targeting autoreactive CD19(+)CD20(−) plasma cells that would be resistant to CD20 mAb treatment [159].

Alemtuzumab is a humanized IgG1 monoclonal antibody that targets the CD52 (Campath-1 antigen), a 12 amino acid glycoprotein anchored to glycosylphosphatidylinositol, which is widely expressed on the cell surface of mature immune cells. Anti-CD52 induces a rapid and prolonged depletion of lymphocytes from the circulation, which results in a profound immunosuppression status followed by an immune reconstitution phase [160]. In EAE, anti-CD52 treatment abrogated B cell infiltration and disrupted existing B cell aggregates in the CNS [161]. Recently, it has been shown that it can also ameliorate colitis through suppressing Th1/17 mediated inflammation and promoting Tregs differentiation in IL-10 deficient mice [162]. Much of the research on alemtuzumab focuses on the lymphocyte repopulation progress and shows that CD4(+)CD25(+)CD127(low) Treg cells preferentially expand within the CD4(+) lymphocytes, reaching their peak within 1 month [163]. A recent study demonstrated that alemtuzumab increased anti-inflammatory cytokines (such as IL-10 and TGF-β) while diminishing pro-inflammatory cytokines (such as IFN-γ, IL-17, IL-6 and TNF-α) within 6 months of treatment and increased Tregs percentage and function after 24 months post-treatment [164]. In addition, alemtuzumab seems to affect B cells, as it increased the percentage of repopulated naïve/immature B cells [165, 166]. Alemtuzumab may thus be a promising therapy for MS [6]; however, it also causes loss of immune-tolerance leading to secondary autoimmunity such as Graves’ disease and marked anti-drug antibody responses [160, 167].

Natalizumab, approved for the treatment of MS, is a humanized IgG4 monoclonal antibody. It mainly binds to the α4-chain of α4β1 integrin heterodimer—also known as very late activating antigen-4 (VLA-4), on the surface of leucocytes and inhibits binding of VLA-4 to vascular cell adhesion molecule-1 (VCAM-1) and, consequently, the attachment of leucocytes to the inner lining of cerebral vascular walls and their crossing of the BBB. This crossing directly diminishes IgM and partially IgG in the CSF [168] as well as in the serum [169]. Thus, natalizumab modulates B cell functions, but appears to be unable to restore the suppressive function of Tregs while marginally decreasing their percentages in MS [170, 171].

Tocilizumab is a humanized IgG1 anti-IL-6 receptor monoclonal antibody approved by the FDA for the treatment of rheumatoid arthritis, active systemic juvenile idiopathic arthritis and polyarticular juvenile idiopathic arthritis. IL-6 is known to induce plasmablast production of anti-aquaporin 4 (Aqp-4) antibodies in vitro and may account for neuromyelitis optica (NMO) disease activity [172]. IL-6 concentration is increased in the CSF of NMO patients [173] and IL-6 induces pro-inflammatory Th17 cells in both NMO and MS patients [174]. Thus, tocilizumab modulates Th17 cells and plasmablasts. Although there are no data on its effect on Tregs and Bregs, tocilizumab appears to be an attractive candidate therapeutic agent for MS.

Fingolimod is an approved therapeutic agent for RRMS. It has a potent pharmacological action because it functions as an unselective agonist of sphingosine 1-phosphate receptors (S1PR) and as a selective antagonist of the S1P1 subtype by induction of receptor downregulation [175]. Since S1P1 is fundamental in the regulation of lymphocyte trafficking, its downregulation leads to redistribution of the immune cells to secondary lymphoid tissues, resulting in the depletion from the circulation and therefore immunosuppression [175]. It prevents lymphocyte egress from secondary lymphoid tissues, promoting loss of CCR7-expressing T cells and increase in Treg numbers and their suppressive function on T cell proliferation [176, 177]. It may affect B cells, as it leads to increased percentage of plasma cells and a shift toward a more naïve and transitional B cell phenotype. In addition, treatment with fingolimod increases both Breg numbers and function, indicated by a boost in IL-10 production [8]. According to recent data, fingolimod significantly enhances CD19(+)BTLA(+)IL−10(+) B cells in RRMS patients, which may relate to amelioration of symptoms [142].

Dimethyl fumarate (DMF) is an approved therapeutic agent for MS. Its in vivo metabolite monomethyl fumarate (MMF) can bind to brain endothelium cells leading to activation of nuclear factor (erythroid-derived 2)-related factor 2 (Nrf2) and downregulation of vascular cell adhesion molecule 1 (VCAM-1) [178].This can be mediated via the G-protein-coupled receptor (GPCR) hydroxycarboxylic acid receptor 2 (HCA2), a known molecular target of MMF. Studies have documented the binding of DMF to HCA2 on dendritic cells followed by the inhibition of pro-inflammatory cytokines production in vitro and in MS murine models [179]. Although its precise mechanism of action remains unclear, evidence indicates that activation of HCA2/GPR109A pathway can decrease immune responses and may enhance anti-inflammatory functions in the intestinal mucosa, possibly leading to reduction in CNS tissue damage in MS patients [180]. In addition, it causes depletion of circulating lymphocytes in peripheral blood [181, 182]. More precisely, DMF alters lymphocyte subsets homeostasis in MS patients, decreasing absolute lymphocyte counts, but does not affect all subsets uniformly [183]. CD8(+) T cells are mainly affected, with reductions in the CD4(+) cells, particularly within the pro-inflammatory T-helper Th1 and Th17 subsets also occurring, creating a bias toward more anti-inflammatory Th2 and regulatory subsets [183]. Both naïve and memory B cells were diminished in certain patients, while Bregs were increased after 4–6 months of therapy and remained in higher numbers at 12 months post-treatment. Also, IL-10 production was elevated in some patients [184]. Other studies showed a skewing from memory CD8(+) and CD4(+) T cells toward their naïve counterparts together with a curtail on T_H1 cells in dimethyl fumarate-treated RRMS patients [9] and an anti-Inflammatory shift in B Cell subsets [183, 185]. These limited data appear very promising.

Teriflunomide is an approved oral therapeutic agent for MS relapses. Its main mechanism of action involves the suppression of the de novo synthesis of pyrimidines in rapidly proliferating cells such as T and B lymphocytes [186]. Pyrimidine synthesis inhibition leads to halt of the cell cycle in G1 phase and it thus has anti-proliferative results, reducing autoantigen-specific immune responses. In a recent study of teriflunomide in murine EAE, a significant increase in CD39(+) Treg concentration was observed, along with decrease in APCs of Peyer patches [187] [reviewed by [10] ].

Glatiramer acetate (GA), a random polymer consisting of four amino acids of the myelin basic protein, is considered a first-line treatment for MS. It prevents disease relapses and patient disability. This agent shifts T cells from a T_H1 to a T_H2 response [188, 189]. GA induces a Treg phenotype and increases FoxP3 expression while restoring Treg function [190]. B cells from GA-treated EAE mice also increased production of IL-10 and reduced expression of co-stimulatory molecules [191]. Importantly, the therapeutic effect of GA in EAE was abrogated in B cell-deficient mice [191, 192]. Another study demonstrated that B cell IL-10 expression was restored, and IL-6 production was diminished after glatiramer acetate treatment. There was also altered proliferation in response to CD40L and an increased immunoglobulin production by B cells [193].

IFNβ-1b and IFNβ-1a are disease-modifying agents for RRMS, affecting multiple immunological processes. IFNβ suppresses the ability of APCs to present antigens and stimulate T cells [194] and prevents T cells from crossing the BBB, while channeling autoreactive T cells into lymphoid tissues [195]. In addition, IFNβ has the ability to induce Tregs, probably due to a shift to Treg-promoting cytokines, such as IL-4, IL-5 and IL-13 [195]. Transitional Bregs are thought to increase as well, as a result of IFN-β therapy [24, 196]. Moreover, the treatment causes Th17 death [197], reduces TNF and increases IL-27 production, known to slow down EAE progression. Thus, IFN-β therapy both impedes pro-inflammatory cells and cytokines and promotes anti-inflammatory ones in MS (Reviewed by [198]).

Apart from specialized therapies, there are other agents with immunomodulatory properties that could prove useful as supportive and/or complementary treatments for MS. One example is HMG-CoA reductase inhibitors (statins), which are a class of lipid-lowering medications known to have immunomodulatory properties. For instance, atorvastatin increases Tregs and reduces clinical disease activity in patients with rheumatoid arthritis. It also displayed anti-inflammatory effects on peripheral blood [199]. In a recent study, atorvastatin and lovastatin enhanced Tregs numbers, but also FoxP3 mRNA levels 30 days post-treatment. However, Treg numbers returned to standard levels after 45 days of treatment. Nevertheless, increased values of TGF-β, FoxP3, CTLA-4 and GITR-expressing Tregs were observed [200]. Simvastatin also regulated TGF-β signal transduction, leading to an increase of Tregs [201], and reduced pro-inflammatory cytokines in patients with rheumatoid arthritis. Statins have an effect on MS as well. Simvastatin suppressed mononuclear cell responses, reduced IFN-γ, TNF-α and IL-2 production and inhibited the antigen-presenting capacity of macrophages [202]. Furthermore, atorvastatin combined with glatiramer acetate showed synergistic immunomodulatory effects in MS [203].

Vitamin D or 25-hydroxy vitamin D (25(OH)D)—the main vitamin D metabolite measured in blood—is known to have immunomodulatory properties. Vitamin D affects both B and T lymphocytes. It inhibits T cell proliferation and reduces IFN-γ, IL-2 and IL-17 expression [204], while increasing IL-10 and Tregs [205]. It also inhibits plasma cell production and increases IL-10-Bregs. Recent evidence from the EAE mouse model indicated that vitamin D-induced dendritic cells could ameliorate symptoms by enhancing the proportions of regulatory lymphocytes and reducing T-helper type 1 and type 17 cells [206]. Vitamin D deficiency is associated with an increased incidence of MS [207]. During MS relapse, 25(OH)D levels are generally decreased [208]. In a small study, vitamin D supplementation led to a significant reduction of the number of newly active brain lesions [209] (reviewed by [210]).