Innate, innate-like and adaptive lymphocytes in the pathogenesis of MS and EAE

Abstract

Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system (CNS) in which the immune system damages the protective insulation surrounding the nerve fibers that project from neurons. A hallmark of MS and its animal model, experimental autoimmune encephalomyelitis (EAE), is autoimmunity against proteins of the myelin sheath. Most studies in this field have focused on the roles of CD4⁺ T lymphocytes, which form part of the adaptive immune system as both mediators and regulators in disease pathogenesis. Consequently, the treatments for MS often target the inflammatory CD4⁺ T-cell responses. However, many other lymphocyte subsets contribute to the pathophysiology of MS and EAE, and these subsets include CD8⁺ T cells and B cells of the adaptive immune system, lymphocytes of the innate immune system such as natural killer cells, and subsets of innate-like T and B lymphocytes such as γδ T cells, natural killer T cells, and mucosal-associated invariant T cells. Several of these lymphocyte subsets can act as mediators of CNS inflammation, whereas others exhibit immunoregulatory functions in disease. Importantly, the efficacy of some MS treatments might be mediated in part by effects on lymphocytes other than CD4⁺ T cells. Here we review the contributions of distinct subsets of lymphocytes on the pathogenesis of MS and EAE, with an emphasis on lymphocytes other than CD4⁺ T cells. A better understanding of the distinct lymphocyte subsets that contribute to the pathophysiology of MS and its experimental models will inform the development of novel therapeutic approaches.

Introduction

Multiple sclerosis (MS) is a chronic autoimmune disease caused by demyelination of the neurons in the central nervous system (CNS), which results in highly variable symptoms, such as muscle weakness, gait difficulties, fatigue, visual disturbances, and loss of coordination and speech, and might ultimately cause paralysis.^1–3 Approximately 2.5 million people worldwide are affected by this disease and the incidence of MS is two- to threefold higher in women than in men. MS can take multiple forms and the most common presentation of MS, which is referred to as relapsing–remitting MS, involves defined attacks of new or increasing neurological symptoms that are followed by periods of partial or complete recovery. Progressive MS, which can occur following the onset of the disease (primary progressive MS) or after an initial relapsing–remitting course (secondary progressive MS), is characterized by a progressive worsening of neurological function over time. The treatments of MS aim to reduce inflammation and return function after an acute attack to the baseline levels, to modify the disease course for the prevention of future attacks, and to manage the diverse symptoms of MS.⁴

Although the etiology of MS remains unknown, it is thought that this disease occurs in genetically predisposed individuals following exposure to an environmental trigger.⁵ Genome-wide association studies have revealed a critical role of immune factors in disease pathogenesis and many of the genes associated with MS susceptibility are also observed in other autoimmune diseases, such as type 1 diabetes, rheumatoid arthritis, and Crohn’s disease.^6,7 Environmental factors that can predispose individuals to MS include low levels of vitamin D, smoking, obesity, and infection with Epstein–Barr virus.^8,9

The pathogenesis of MS is mediated primarily by T cells.^1,10 T cells are primed to CNS autoantigens in the periphery and then cross the blood–brain barrier to activate microglia and macrophages. In concert, these cells induce the death of myelin-producing oligodendrocytes and directly damage the myelin sheath around nerve fibers to generate active lesions in the CNS. The key role of T cells, particularly CD4⁺ T cells, in MS has been confirmed in experimental autoimmune encephalomyelitis (EAE), which is the main animal model of MS and can be induced in a variety of mammalian species through immunization with myelin antigens accompanied by adjuvant or through the adoptive transfer of myelin-reactive T cells.¹¹ Studies on MS and EAE have largely focused on CD4⁺ T cells and many treatments for MS are based on targeting these cells. However, many other lymphocyte subsets have been implicated in disease pathogenesis^12,13 and these subsets include lymphocytes other than CD4⁺ T cells that belong to the adaptive immune system, lymphocytes that belong to the innate immune system, and innate-like B and T lymphocytes that exhibit properties of both innate and adaptive immunity (Fig. 1). Several of these cell types infiltrate the CNS of MS patients, have been implicated in the efficacy of disease-modifying treatments, and contribute to the pathogenesis of EAE. Here we review the role of distinct subsets of innate, innate-like, and adaptive lymphocytes in the pathogenesis of MS and EAE, with an emphasis on cells other than CD4⁺ T cells.

Fig. 1.

Subsets of innate, innate-like, and adaptive lymphocytes. The main subsets of innate, innate-like, and adaptive lymphocytes, and their key properties. Abbreviations: BCR B-cell receptor, ILC innate lymphoid cell, LTi lymphoid tissue inducer, MAIT mucosal-associated invariant T, MZB marginal zone B, NK natural killer, NKT natural killer T, TCR T-cell receptor, VDJ variable diversity joining

Adaptive lymphocytes

In addition to CD4⁺ T cells, which have been the main focus of studies on MS and EAE, the lymphocytes of the adaptive immune system include CD8⁺ T cells and B cells. Although each of these cell types can mediate disease, they also contain regulatory subsets that can suppress disease.¹⁴

CD4⁺ T cells

CD4⁺ T cells are found in the brain lesions of patients with MS.^15,16 The key role of these cells as effectors in disease was confirmed by the identification of human leukocyte antigen (HLA)-DR15 as the key genetic factor associated with disease susceptibility.¹⁷ In addition, CD4⁺ T cells have a critical role in the pathogenesis of EAE. The susceptibility of different mouse strains to EAE is correlated with their major histocompatibility complex (MHC) haplotype. The adoptive transfer of CD4⁺ T cells from mice immunized with myelin antigens can often induce disease in recipient animals. Both interferon (IFN)-γ-producing T helper (Th) 1 and interleukin (IL)-17-producing Th17 cells are pathogenic, and most studies have indicated that Th17 cells are the main culprits.^1,10 In EAE, these T cells are specific for peptides derived from immunizing myelin antigens, which include myelin basic protein (MBP), proteolipid protein, and myelin oligodendrocyte glycoprotein,¹⁸ and these target antigens might also have a role in MS.^1,10 Although the frequency of circulating myelin-reactive CD4⁺ T cells in patients with MS is similar to that in healthy individuals, the T cells from MS patients exhibit a pro-inflammatory phenotype characterized by Th17 cytokines¹⁹ and often exhibit spontaneous proliferation ex vivo.²⁰

Considering the critical role of CD4⁺ T cells in mediating CNS inflammation, it was surprising that a phase II trial with a monoclonal anti-CD4 antibody failed to reduce MS activity.²¹ This failure might have been due, in part, to the critical role of CD4⁺Foxp3⁺ regulatory T cells (Tregs) in suppressing autoreactive T-cell responses.^22,23 Tregs produce the anti-inflammatory cytokines IL-10 and transforming growth factor (TGF)-β, which suppress myelin-reactive CD4⁺ T cells. Consistent with their suppressive role in disease progression, CD4⁺ Tregs from MS patients exhibit impaired functions²⁴ and myelin-specific T cells from patients show a reduced capacity to produce IL-10 compared with those obtained from healthy controls.¹⁹

CD8⁺ T cells

One of the main findings identifying CD8⁺ T cells as effectors in MS is that these cells are found in MS lesions at greater numbers than CD4⁺ T cells.^15,16,25 Furthermore, although MHC class II alleles are the main genetic factors that predispose individuals to MS, MHC class I alleles have also been associated with MS.^26–29 HLA-A3 and -B7 are associated with a higher risk of MS, whereas HLA-A2 is associated with a reduced risk of MS. These findings suggest that CD8⁺ T cells might have pathogenic and protective functions in MS.

The CD8⁺ T cells in the CNS lesions and cerebrospinal fluid (CSF) of MS patients exhibit oligoclonal expansions,^30–32 suggesting antigen-driven selection. These cells express granzyme B, which suggests that they have cytotoxic activity, and produce the pro-inflammatory cytokines IFN-γ and IL-17.^33–35 One challenge in interpreting these findings is that some of the CD8⁺ T cells investigated in these studies likely represent MR1-restricted mucosal-associated invariant T (MAIT) cells expressing CD8α chains (see below).^36,37

The adoptive transfer of myelin antigen-specific CD8⁺ T cells in the EAE model induces disease in the recipient animals and causes CNS lesions and symptoms that are more similar to those found in human disease than those observed in CD4⁺ T-cell-mediated EAE.^38,39 Similar to the results found for CD4⁺ T cells, IL-17 was the relevant pathogenic cytokine produced by myelin-reactive CD8⁺ T cells. Furthermore, IL-17-producing CD8⁺ T cells are required for the development of Th17-cell-mediated EAE, which suggests the involvement of critical interactions between pathogenic CD8⁺ and CD4⁺ T cells in disease progression.⁴⁰

The possibility that CD8⁺ T cells, which have a pathogenic role, might also have suppressive functions in CNS autoimmunity is reinforced by the finding that the depletion of CD8⁺ T cells before the induction of EAE in mice results in disease exacerbation.⁴¹ This result was further supported by studies with MHC class I- and CD8α-deficient animals, which developed more severe disease than wild-type animals.^42–44 In some studies, neuroantigen-specific CD8⁺ T cells were shown to suppress EAE induction and even ameliorate established EAE upon adoptive transfer through a mechanism involving the direct lysis of pathogenic CD4⁺ T cells.^43,45 The majority of regulatory CD8⁺ Tregs responsible for these effects are restricted by classical MHC class I molecules. However, CD8⁺ Tregs restricted by the nonclassical MHC class I molecule Qa-1 that protect against EAE have also been described.⁴⁶ In some of these studies, Qa-1-restricted CD8⁺ Tregs mediate their effects by recognizing peptides derived from specific T-cell receptor (TCR) Vβ molecules on myelin antigen-reactive CD4⁺ T cells and thereby directly cause their lysis.^47,48 In addition, some Qa-1-restricted CD8⁺ T cells react with myelin-derived peptide antigens.⁴⁹ In mice, Qa-1-restricted CD8⁺ Tregs are induced during treatment with glatiramer acetate,⁵⁰ which is a peptide that mimics MBP and an Food and Drug Administration-approved MS treatment. The disease-ameliorating effects of glatiramer acetate are dependent on Qa-1-restricted CD8⁺ Tregs.⁵⁰

CD8⁺ T cells with regulatory functions in MS were identified many years ago.^51–53 Phenotypically distinct subsets of CD8⁺ Tregs have been identified in humans and these include IL-10-producing cells and cells with cytotoxic properties.^51,54 Some of these cells might be restricted by the nonclassical MHC class I protein HLA-E, which is the functional homolog of murine Qa-1. Glatiramer acetate induces HLA-E-restricted CD8⁺ Tregs in treated patients and these cells have strong suppressive potential against CD4⁺ T cells.⁵⁵

B cells

B cells, which are best known for their capacity to produce antibodies, infiltrate MS lesions in the CNS.^56,57 Many of these cells have a memory phenotype and are found in cellular immune aggregates that resemble tertiary lymphoid structures.⁵⁸ The abnormal presence of antibodies in the CNS represents the most consistent feature of patients with MS.^59,60 These antibodies show evidence of clonotypic expansion, which suggests an antigen-driven process. Some of these antibodies are directed against myelin antigens, which can cause demyelination and axonal damage in a complement-dependent manner.^61,62 In some patients, the removal of antibodies by plasmapheresis and immunoadsorption is beneficial, which suggests that antibodies can contribute to disease pathogenesis.^63,64 However, the finding that B-cell depletion with the anti-CD20 antibody rituximab limits relapses in MS but does not influence abnormalities in CSF antibodies⁶⁵ indicates that antibody-independent roles of B cells are involved in the beneficial effects of B-cell depletion.⁶⁶ Indeed, an important role of B cells is the presentation of antigens to T lymphocytes in the context of MHC class II.⁶⁷ Memory B cells in MS patients can efficiently present neuronal antigens to T cells.⁶⁸ A recent study identified RASGRP2, a guanine nucleotide exchange factor expressed by B cells and enriched in the brain, as an autoantigen presented by memory B cells in the periphery to self-reactive brain-homing T cells,⁶⁹ which suggests that B cell/T cell interactions outside of the brain might induce pathogenic T cells that can then enter the brain to attack CNS-resident antigens.⁷⁰ Another function of B cells is cytokine production and subsets of pro-inflammatory B cells that produce cytokines such as tumor necrosis factor (TNF)-α, lymphotoxin α, IL-6, and granulocyte-macrophage colony-stimulating factor (GM-CSF), as well as anti-inflammatory regulatory B cells (Bregs) that produce cytokines such as IL-10, IL-35, and/or TGF-β have been identified.^71,72 Multiple B-cell populations with regulatory properties have been described⁶² and these include cells with phenotypic markers of immature or mature B cells, plasma cells, or B1a cells, which constitute a subset of innate-like B cells (see below). In MS, B cells appear to be polarized toward a pro-inflammatory phenotype and defects in Bregs have been observed, at least in some studies.⁷³ Some MS therapies, including IFN-β, glatiramer acetate, and fingolimod (FTY720, which sequesters lymphocytes in lymph nodes), are associated with an anti-inflammatory shift in B-cell cytokine responses,⁷³ which suggests that the balance between pro- and anti-inflammatory B-cell activities is relevant to disease pathogenesis and treatment success.

In EAE, B cells have distinct pathogenic or regulatory functions, depending on the stage of disease.⁷⁴ B-cell-deficient mice exhibit an exacerbated disease course, which is due to the regulatory functions of B cells and their capacity to produce IL-10.^75,76 B-cell depletion using anti-CD20 antibodies before EAE induction causes exacerbated disease and this effect is associated with the loss of IL-10-producing Bregs.⁷⁷ However, B-cell depletion after EAE induction ameliorates disease severity due to the pathogenic role of memory B cells, which facilitate T-cell reactivation during the later disease stages.^77–81 In adoptive transfer studies, Bregs reduce EAE initiation but not ongoing EAE disease.⁸² Similarly, the adoptive transfer of Bregs expanded ex vivo from splenic B cells through the provision of CD40 and IL-21 receptor signals can inhibit disease, even in mice with established disease.⁸³ In similar experiments, Bregs induced in vitro or in vivo from bone marrow cells through the provision of Toll-like receptor (TLR) 9 signals can protect against disease when administered at the onset of clinical symptoms.^84,85 Bregs have also been implicated in the beneficial effects of glatiramer acetate against EAE.^86,87 In addition to IL-10, other mechanisms have been proposed for the suppressive functions of Bregs in EAE^88,89 and these include the induction of IL-10-producing Tregs⁹⁰ and the production of TGF-β, IL-35, programmed death 1 ligand 1 or Fas ligand.^88,89 Although most studies have implicated adaptive Bregs in these activities, they were unable to exclude a potential contribution of innate-like Bregs.

Innate lymphocytes

It is now appreciated that the innate immune system includes a variety of lymphoid subsets, called innate lymphoid cells (ILCs), which are strategically positioned in many tissues of the body to provide protection against infections and to control tissue inflammation.^91,92 ILCs have been divided into three distinct groups (Fig. 1) based on their expression of transcription factors, cytokine profiles, and functional properties. Group 1 ILCs include classical natural killer (NK) cells and ILC1 cells that produce IFN-γ, group 2 ILCs include ILC2 cells that produce IL-4, IL-5, and IL-13, and group 3 ILCs include lymphoid tissue inducer (LTi) cells implicated in the development of secondary lymphoid organs and ILC3 cells that produce IL-17 and IL-22. These cells generally have roles very early in the immune system, are highly responsive to activation by certain cytokines, and can interact with adaptive lymphocytes.⁹³ Several of these cell types, particularly NK cells, have been implicated in the progression of MS and EAE.

NK cells

NK cells are innate lymphocytes with cytolytic activity that is controlled via a variety of activating and inhibitory receptors.⁹⁴ These cells can produce the pro-inflammatory cytokines IFN-γ and TNF-α, the anti-inflammatory cytokine IL-10, and the growth factor GM-CSF in response to activation by cytokines such as IL-12, IL-15, and IL-18.⁹⁵ These cells can engage in extensive reciprocal interactions with autoreactive T cells.⁹⁶

In humans (but not in mice), NK cells can be distinguished by the expression of CD56, a neural cell adhesion molecule.⁹⁷ CD56^dim NK cells are more effective killers, whereas CD56^bright NK cells produce greater amounts and a wider variety of cytokines, including immunoregulatory cytokines.⁹⁸ In the CSF of MS patients, the CD56^bright population is expanded compared with the CD56^dim population⁹⁹ and this expansion likely serves as a countermeasure in response to the inflammatory process. In contrast, in the peripheral blood, the CD56^dim population is enhanced.¹⁰⁰ Several therapies, including IFN-β,¹⁰¹ daclizumab (an antibody directed against the IL-2 receptor α subunit, CD25),^102,103 and natalizumab (an antibody directed against the α4 integrin, which is part of the adhesion molecule very late antigen-4 expressed by effector lymphocytes),¹⁰⁴ induce an increase in the CD56^bright NK cell population in the circulation of patients. These findings are consistent with the suppressive effect of the CD56^bright NK cell population observed in MS.¹⁰⁵ These CD56^bright NK cells can lyse activated (but not resting) CD4⁺ T cells^106,107 and this finding suggests a mechanism for their suppressive activities in MS.

In EAE, the depletion of NK cells has resulted in contradictory results: some studies have provided evidence for disease exacerbation upon NK cell depletion^108–111 and other studies have demonstrated that NK cell depletion results in disease amelioration.¹¹² This contradiction is likely due to differences in the experimental systems employed, particularly the reagents used to deplete NK cells.¹¹³ CX3CL1 (fractalkine)-deficient mice, in which the homing of NK cells to the CNS is prevented, develop an exaggerated form of EAE,¹¹⁴ which suggests that NK cells exert protective effects in EAE. Several subsequent studies have provided evidence for a protective role of NK cells in EAE.¹⁰⁵ Mechanistically, it has been shown that a blocking antibody directed against the NK cell inhibitory receptor NKG2A, which interacts with Qa-1, can increase the killing of T cells, microglia, and macrophages.^115,116 As Qa-1 is downregulated during EAE, these findings suggest that the progression of EAE is associated with a loss of the tolerance of pathogenic T cells and myeloid cells to NK cell lysis. The expansion of NK cells by treatment with IL-2 in complex with anti-IL-2 antibodies ameliorates EAE,¹¹¹ which suggests a potential approach for the therapeutic targeting of NK cells in MS.

A recent study provided evidence for a role of NK cells in delaying the repair process in CNS inflammation.¹¹⁷ In particular, the study showed that NK cells are retained for an extended period of time within the CNS during the chronic phase of MS and EAE. Moreover, NK cells suppress the reparative properties of neural stem cells during EAE through a mechanism that involves Qa-1 downregulation in these cells, resulting in their lysis by NK cells.

Collectively, the findings discussed here are consistent with a suppressive role of NK cells during the initiation and progression of CNS inflammation, but these cells might hinder recovery from brain inflammation.

LTi cells

LTi cells are a subset of ILCs that are critically important for the development of secondary lymphoid organs.¹¹⁸ These cells also control the generation of organized follicle-like structures called tertiary lymphoid follicles that might develop during chronic inflammation in tissues, including the CNS.^58,119,120 Several studies have provided evidence showing that LTi cells are expanded in the CNS and circulation of patients with MS,^121–124 which suggests that these cells have a pathogenic role by promoting the generation of tertiary lymphoid follicles. Consistent with these findings, the treatment of patients with daclizumab decreases the number of LTi cells compared with untreated patients.¹²¹

LTi cells have also been identified in the meninges of the CNS of mice and these cells accumulate and become activated during EAE.¹²⁵ It has been proposed that LTi cells promote the progression of EAE,^125,126 even though conclusive studies remain to be performed.

Other ILC cells

ILC2 cells have been implicated in sex differences associated with EAE in the Swiss Jim Lambert (SJL) mouse strain^127,128 and the related studies proposed that a deficit in Th2-promoting ILC2 cells in female SJL mice, compared with male mice, is responsible for the increased susceptibility of female animals to EAE.

A recent study provided evidence for a pathogenic role of a subset of ILCs that is dependent on the transcription factor T-bet and expresses the natural cytotoxicity receptor NKp46 in EAE.^129,130 These cells, which localize to the meninges, induce the expression of pro-inflammatory mediators that facilitate the entry of pathogenic T cells into the CNS. The relevant ILCs responsible for these effects include both ILC1 and ILC3 cells (but not NK cells).

Innate-like T and B cells

In addition to subsets of lymphocytes that belong to either the innate or the adaptive immune system, several subsets of lymphocytes exhibit phenotypic and functional properties of both the innate and adaptive immune system, and are referred to as innate-like T and B cells or “inbetweeners” (Fig. 1).^131,132 These cells exhibit several common features, including the expression of antigen-specific TCRs or B-cell receptors with limited diversity, the expression of surface markers typically found on memory cells, a tendency for autoreactivity, a rapid and robust response to antigen and cytokines, and limited memory responses. Multiple T-cell subsets that belong to this family of lymphocytes, including γδ T cells, NKT cells, and MAIT cells, have been implicated in the pathogenesis of MS and EAE. Few studies have investigated the potential role of innate-like B cells in CNS autoimmunity.

γδ T cells

T cells expressing γδ TCRs represent ~2% of all T cells in peripheral blood and secondary lymphoid organs but represent a major lymphocyte population in the epithelia of the gut, skin, lung, and other organs.^133,134 Distinct subsets of γδ T cells characterized by the expression of specific Vγ TCR chains as well as unique functional properties are found in distinct mucosal locations. The antigen specificity of γδ T cells remains largely unknown and rarely involves antigen presentation via classical MHC molecules, although some γδ T cells recognize stress-induced nonclassical MHC class I molecules.¹³⁵ Unlike αβ T cells, which are dependent on cognate antigens for their activation, γδ T cells can be readily stimulated with TLR ligands and cytokines such as IL-1 and IL-23. In turn, these cells can produce a variety of pro- and anti-inflammatory cytokines.

γδ T cells have been found in MS lesions in the brain and CSF of patients with recent onset disease, where they are clonally expanded, which is suggestive of antigenic stimulation.^136,137 A potential role of γδ T cells in the pathogenesis of MS has been further suggested by the capacity of these cells to lyse oligodendrocytes.^138,139

In mice, γδ T cells can be found in the CNS during steady-state conditions^140,141 and their number increases during the development of EAE.¹⁴² This population is enriched in cells expressing the Vγ4, Vγ1, and Vγ6 chains. The Vγ4 subset primarily produces IL-17 and has pro-inflammatory functions, whereas the Vγ1 subset primarily produces IFN-γ and has immunoregulatory functions.¹⁴³ Although initial studies provided evidence for a pathogenic role of γδ T cells in EAE, more recent studies have suggested that these cells can have both pathogenic and beneficial effects on disease.^113,144,145 The depletion of γδ T cells using monoclonal antibodies suppresses EAE severity.^146–148 Consistent with these findings, mice that are genetically deficient in γδ T cells are protected against EAE.^149,150 A subsequent study investigated the role of γδ T cell subsets through the use of activating, Vγ chain-specific antibodies.¹⁵¹ Activated Vγ4 cells exacerbate EAE, whereas activated Vγ1 cells suppress EAE. Although the related mechanisms remain unclear, it has been proposed that Vγ4 cells promote the activity of encephalitogenic Th cells while suppressing Tregs,^149,152 whereas Vγ1 cells enhance Treg cell function during EAE.¹⁵¹

NKT cells

NKT cells are a subset of T lymphocytes that express TCRs together with surface markers that are characteristic of the NK cell lineage.^153–155 These cells recognize glycolipid antigens that are presented to them by antigen-presenting cells in the context of the MHC class I-related protein CD1d. Two subsets of NKT cells have been identified: type 1 or invariant NKT (iNKT) cells express semi-invariant TCRs (Vα14-Jα18/Vβ8, -7, or -2 in mice and Vα24-Jα18/Vβ11 in humans), whereas type 2 or diverse NKT (dNKT) cells express more diverse or oligoclonal TCRs. Although iNKT cells are predominant in mice, dNKT cells are predominant in humans. iNKT cells are specific for a variety of microbial and autologous glycolipid antigens and include subsets with polarized cytokine production profiles, which is a property acquired during their development in the thymus. Although the antigen specificity of dNKT cells remains less well defined, a substantial proportion of these cells react with sulfatide, a self-glycolipid found in myelin sheaths.¹⁵⁶ In addition to TCR-mediated activation, NKT cells are also highly responsive to stimulation by cytokines. These cells have been implicated in a variety of infectious, autoimmune, and inflammatory diseases.^157,158 Curiously, iNKT and dNKT cells often have opposing roles in an immune response.¹⁵⁹

The numbers and functional activities of iNKT cells are decreased in the circulation of MS patients,^160,161 and the numbers and functions of these cells are restored in patients in remission due to treatment with IFN-β.¹⁶² In patients treated with corticosteroids¹⁶³ or patients in remission in the absence of treatment,¹⁶⁴ iNKT cells adopt a Th2-biased cytokine production profile, which suggests potential suppressive activities against the disease process. Studies on the status of dNKT cells in MS have not yet been performed.

SJL mice, which develop a relapsing–remitting form of EAE, have defects in iNKT cell numbers and functions due to absence of the Vβ8 gene segment,^165,166 which is the main Vβ chain of murine iNKT cells. iNKT cells are found in the healthy CNS and remain unchanged following EAE induction.¹⁶⁷ In contrast, sulfatide-reactive dNKT cells become expanded by several fold in the CNS of mice with EAE.¹⁵⁶

EAE studies with NKT cell-deficient mice have provided divergent results. In CD1d-deficient mice, which lack both iNKT and dNKT cells, either no effect^167–169 or disease exacerbation has been observed.¹⁷⁰ In Jα18-deficient animals, which only lack iNKT cells, disease amelioration,¹⁷¹ no effect,¹⁶⁸ or disease exacerbation¹⁷² has been reported. Vα14-Jα18 TCR transgenic mice that overproduce iNKT cells are partially protected against EAE through a mechanism that involves iNKT cell infiltration into the CNS and inhibition of autoantigen-specific Th cell responses in the spleen.^173,174 Thus, the role of iNKT and dNKT cells in the natural progression of EAE remains unclear. Potential reasons for these divergent findings include differences in the mouse strains and experimental procedures employed, as well as altered effects on iNKT cells exerted by the gut microflora present in different housing facilities, which is a possibility that is supported by one EAE study.¹⁷⁵

Due to their expression of semi-invariant (for iNKT cells) or oligoclonal (for dNKT cells) TCRs, it has been possible to investigate the effects of specific glycolipid ligands for these cells on the progression of EAE.¹⁷⁶ For iNKT cells, many studies have employed the glycolipid α-galactosylceramide (α-GalCer), a synthetic version of a marine sponge-derived natural product that activates iNKT cells from a variety of mammalian species.¹⁷⁷ These studies were performed with a variety of EAE models, mouse strains, and treatment doses, routes, and schedules.^158,178 Using B6, PL/J, or B10.PL strains, treatment before or close to the time of EAE induction protected against disease in most studies, whereas treatment following EAE induction usually resulted in disease exacerbation.^{167–170,179} In SJL mice, which have numerical deficits in iNKT cells as well as a Th1 bias in iNKT cell cytokine production,^165,166 α-GalCer exacerbated disease.¹⁶⁹ Disease protection is typically associated with a shift in the autoantigen-specific T-cell response from Th1/Th17 toward Th2 and the opposite was observed in the cases of disease exacerbation.^165,166 A role for the Th2 cytokines IL-4 and IL-10 has also been reported.^167,169,180 The role of Th2 cytokines is further supported by the finding that α-GalCer analogs that induce a Th2-biased cytokine profile in iNKT cells exhibit improved suppressive activities in EAE.^181,182 Additional mechanisms that have been proposed to contribute to the therapeutic activities of α-GalCer against EAE include the induction of immune-suppressive myeloid lineage cells (dendritic cells, M2 macrophages, and myeloid-derived suppressor cells)^{172,183–185} and CD4⁺ Tregs.^176,186

The capacity of sulfatide, the ligand for a major subset of dNKT cells, to modulate EAE has also been investigated.¹⁷⁶ A single treatment with sulfatide before, during, or after EAE induction exerts protective effects in B6 and B10.PL mice.¹⁵⁶ This ligand has also been found to effectively reverse ongoing relapsing–remitting disease in SJL mice.¹⁸⁷ Disease protection is associated with reduced autoantigen-specific Th1 and Th17 responses. Surprisingly, the treatment efficacy also required the participation of iNKT cells.¹⁸⁷ A model in which sulfatide-activated dNKT cells induce an anergic response in iNKT cells and a tolerant response in peripheral dendritic cells and CNS-resident microglia has been proposed.

MAIT cells

MAIT cells are a subset of innate-like T cells that show substantial similarities to NKT cells.^36,37 These cells, which were originally identified in association with the gut mucosa, express semi-invariant TCRs (Vα19-Jα33 in mice and Vα7.2-Jα33 in humans) that are specific for bacterial-derived riboflavin (vitamin B2) antigens presented in the context of the MHC class I-related protein MR1. MR1-restricted T cells that express more diverse receptors have also been identified.¹⁸⁸ MAIT cells are rare in mice but are quite abundant in the peripheral blood of humans, where they constitute ~5% of the T-cell population. Similar to NKT cells, these cells are often activated by cytokines in an MR1-independent manner. These cells can produce a variety of predominantly pro-inflammatory cytokines and have been implicated in immune responses against microorganisms and in inflammatory and autoimmune diseases, including CNS autoimmunity.^189,190

MAIT cells have been identified in the CNS lesions of patients with MS.^191–195 Most studies have found a reduction in all or subsets of MAIT cells in the circulation of MS patients,^{193,195–197} although one study found an increase in these cells.³⁴ Another study also reported that MAIT cells in untreated MS patients show a reduced capacity to produce IFN-γ and TNF-α, and that this defect is overcome by treatment with fingolimod.¹⁹⁷ Finally, a recent study revealed that MAIT cells in the peripheral blood of MS patients produce increased levels of IL-17.¹⁹⁸ Collectively, these studies are most consistent with a pathogenic role for MAIT cells in MS.

In EAE, the available evidence points toward a protective role of MAIT cells in CNS inflammation.¹⁹⁹ MR1-deficient mice develop more severe EAE than wild-type animals and, conversely, Vα19-Jα33 TCR transgenic animals develop less severe EAE. Furthermore, the adoptive transfer of MAIT cells protects wild-type mice against EAE. The protective effects of murine MAIT cells in this model are associated with reduced Th1 and Th17 responses against CNS autoantigens and increased secretion of IL-10. As there is little evidence showing that human MAIT cells produce IL-10, it is unclear whether these mouse studies can be extrapolated to human MS patients.

Innate-like B cells

This group of lymphocytes includes B1 cells, which are enriched in the peritoneal and pleural cavities,²⁰⁰ and marginal zone B (MZB) cells, which segregate into the marginal zone of the spleen.²⁰¹ Both of these B-cell subsets can produce antibodies in a T-cell-independent manner. Some of these cells, particularly B1 cells that express the CD5 surface marker (B1a cells), are also capable of producing IL-10 and can thus function as Bregs.^72,202

B1 cells accumulate in the brain-draining cervical lymph nodes during EAE.²⁰³ The depletion of these cells during the induction phase of EAE exacerbates the pathology of the disease, whereas their depletion during the effector phase decreases the disease pathology.²⁰⁴ Whether these effects are directly mediated by IL-10-producing B1a cells remains unclear. The counts of MZB cells are substantially reduced in the brain-draining cervical lymph nodes of mice with EAE.²⁰³

Concluding remarks

The reviewed studies reveal that multiple different lymphocyte subsets contribute to the initiation, progression, and repair of CNS autoimmunity, and these subsets include cells that belong to either the innate or adaptive immune systems and cells that straddle these systems (Fig. 1). Several of these cell types contain subpopulations with pathogenic or suppressive functions, which need to be considered when investigating the contribution of individual subsets to disease pathogenesis. For example, the divergent roles of particular lymphocyte subsets often observed in the initiation, progression, or reparative stages of disease might be explained, at least in some cases, by the opposing contributions of pro- and anti-inflammatory subpopulations.

Most treatments for MS and EAE have been developed based on the critical role of pathogenic CD4⁺ T cells (Th1 and particularly Th17 cells) in disease, but accumulating evidence indicates that successful treatments are often associated with alterations in many other lymphocyte subsets. These findings not only suggest the involvement of a wide variety of lymphocyte subsets in disease pathogenesis but also identify these cells as potential targets for immunotherapy. Many of these cell types can influence the induction, CNS recruitment, and/or activity of pathogenic Th1/Th17 cells and protective Tregs. Several of these lymphocyte subsets are amenable to therapeutic targeting by biologics such as monoclonal antibodies or, in some cases (e.g., NKT cells), with cognate antigens. It is therefore critically important to investigate the precise functions of individual lymphocyte subsets in CNS autoimmunity, including EAE. Notably, when attempting to translate findings obtained in mice with EAE to human MS patients, it is important to keep in mind that several of the lymphocyte subsets that have emerged as novel immunotherapeutic targets exhibit divergent phenotypic and functional properties in mice and humans.

Competing interests

The authors declare no conflict of interest.