TH17 cytokines in autoimmune neuro-inflammation (original) (raw)

. Author manuscript; available in PMC: 2013 Jan 3.

Published in final edited form as: Curr Opin Immunol. 2011 Sep 9;23(6):707–712. doi: 10.1016/j.coi.2011.08.005

Abstract

It has been firmly established that IL-23 polarized TH17 cells are potent effectors in the pathogenesis of experimental autoimmune encephalitomyelitis (EAE). However, the relative importance of these cells in comparison to other encephalitogenic TH subsets, and the mechanisms that they employ to effect inflammatory demyelination, are topics of continuing investigation. Interestingly, deletion of individual ‘TH17 cytokines’, such as IL-17A, IL-17F, IL-22 and IL-21, does not phenocopy the complete EAE-resistance of IL-23-deficient mice. The instability of TH17 cells in vivo introduces an additional layer of complexity to their role in the context of relapsing or chronic disease. Recent data indicate that IL-23 drives the production of myeloid activating factors, such as GM-CSF, by myelin-reactive T cells and facilitates their accumulation in the CNS. This review discusses the above issues in relation to the use of TH17 cells and related factors as potential therapeutic targets and biomarkers in CNS autoimmune diseases such as multiple sclerosis (MS).

Introduction

Ever since the discovery that CD4+ T helper (TH) cells could be classified into subsets based on the cytokines they produce, it has become evident that in vivo pathogenic as well as protective immune responses tend to be skewed towards specific TH lineages [1]. The nature of the TH cell polarization pattern underlying MS has been the focus of intense research for 20 years, driven in large part by the quest to develop better drugs to treat that disorder. Hence, therapeutic agents that target a dominant encephalitogenic TH subset or its products could retain the efficacy of more globally immunosuppressive agents while minimizing the risk of opportunistic infections and secondary neoplasia. Although TH1 cells were initially thought to mediate inflammatory demyelination in MS, the discovery of TH17 cells has forced a reexamination of that dogma [2, 3]. Recent data generated by studies in MS and in its animal model EAE, suggest that the mechanism of action of TH17 cells in CNS autoimmunity is more complex than previously appreciated. Here we will discuss apparent paradoxes regarding the respective roles of TH17 cells and their signature cytokines, IL-17A, IL-17 IL-22, IL-21, in the pathogenesis of autoimmune demyelinating disease in mice and men.

The role of TH17 cells in EAE

For many years, TH1 cells were deemed responsible for the initiation of autoimmune demyelination. Conversely, TH2 cells were believed to have regulatory properties, leading some to propose ‘immune deviation’ from a TH1 to a TH2 response as a promising therapeutic strategy in MS [4]. This paradigm has since been challenged by the results of experiments with genetically engineered mice. Against expectations, mice deficient in IFNγ, IFNγ receptor or IL-18 succumb to EAE, while IL-4 deficient mice exhibit a clinical course comparable to that of wildtype controls [5–8]. The TH1 polarizing factor, IL-12 is composed of covalently bound p40 and p35 chains. Paradoxically, while mice deficient in the p40 chain are resistant to EAE [9], those deficient in the p35 chain are not [10••, 11]. These apparent discrepancies were resolved by the discovery of IL-23, a heterodimer composed of the IL-12 p40 chain complexed with a unique p19 chain [12]. IL-23 p19 deficient mice phenocopy IL-12p40 deficient mice in their complete resistance to EAE [13••], underscoring the importance of IL-23, as opposed to IL-12 or IFNγ, in EAE pathogenesis [14]. IL-17A (IL-17) was soon identified as one of the cytokines produced by T cells in response to IL-23 [15]. The increased frequency of IL-17 secreting cells in myelin immunized mice led to the conjecture that they constitute the critical effector population in autoimmune demyelinating disease. This was supported by the demonstration that IL-23 polarized, IL-17 secreting myelin-reactive CD4+ T cells are highly encephalitogenic following transfer into naïve syngeneic recipients [15, 16••, 17].

More recent experiments have indicated that the encephalitogenicity of IL-23 polarized cells is not solely attributable to the production of IL-17 itself or of other cytokines traditionally ascribed to the TH17 panel. The course of EAE is unperturbed, or only ameliorated, in IL-17A knock-out mice and in wild-type mice treated with neutralizing antibodies specific for IL-17A [18•, 19]. Conversely, ectopic expression of IL-17A did not worsen EAE [20]. Genetic targeting of other TH17 associated cytokines, such as IL-17F [20] or IL-22 [21], also failed to confer resistance to the disease. IL-21 was initially claimed to be a vital encephalitogenicity gene [22], but this finding was not reproduced by others [23, 24]. The dispensability of IL-17A is not explained by redundant activities of IL-17F, since mice treated with an IL-17 receptor-Fc fusion protein, that blocks both cytokines, only had a modest impact [19]. Furthermore, IL-17F knock-out mice succumb to EAE, despite treatment with an IL-17A neutralizing antibody, – albeit disease severity is reduced [20].

Based on the above discussion, it could be argued that it is more appropriate to view IL-17 as a surrogate marker of IL-23 modulation, rather than as a master effector cytokine, in the context of CNS autoimmunity [25]. This does not mean that IL-17 is completely irrelevant to the pathogenesis of inflammatory demyelinating disorders. Several lines of evidence indicate that, in certain instances, IL-17 facilitates leukocyte trafficking across the blood–brain barrier (BBB) and infiltration within the CNS. In some EAE models, CNS infiltrates induced by myelin-reactive TH1 cells are confined to the subpial white matter, whereas those induced by TH17 cells reactive to the same epitope extend into the parenchyma in finger-like projections [16••]. Kebir and colleagues reported that human TH17 cells migrate across BBB endothelial cell monolayers more efficiently than TH1 cells and that IL-17 and IL-22 directly stimulate CNS endothelial cells to express chemokines and downregulate tight junction proteins [26•]. IL-17 also stimulates astrocytes to produce ELR+ CXC chemokines that could, in turn, attract neutrophils to the BBB and activate them to release vasoactive substances [27]. Leukocyte infiltration of the brainstem, as opposed to the spinal cord, appears to be particularly dependent on IL-17 and impeded by IFNγ [6, 17, 28•, 29]. The biological basis for this regional difference in IL-17 dependence is unknown.

IL-17 independent mechanisms of action of IL-23 in CNS autoimmunity

The literature summarized in the former section suggests that IL-23 promotes EAE by IL-17 independent, as well as dependent, pathways in a context-dependent manner. TH17 cells have been described to develop in vitro after exposure to IL-6 and TGFβ [[reviewed in 30]], but this cytokine cocktail does not generate an encephalitogenic population [31••]; pathogenicity is dependent on IL-23 and not on the secretion of IL-17 by TGFβ/IL-6 manipulated T cells [31••]. Two independent reports showed that, while IL-23 receptor signaling is not required for the activation and expansion of myelin-reactive T cells in vivo, it is required for their accumulation in the CNS [31••, 32]. It remains to be demonstrated whether IL-23 directly facilitates T cell trafficking to the CNS via upregulation of chemokine receptors and adhesion molecules and/or promotes T cell survival/retention once they have crossed the BBB. IL-23 polarized, myelin-reactive T cells induce EAE after a shorter latency than IL-12 polarized effectors [16••, 33]. Some have speculated that TH17 cells initially access the CNS and create a microenvironment conducive to the subsequent entry of TH1 cells [33]. This could explain why TH1 cells do not cause EAE in actively immunized IL-23p19 deficient mice [13••]. However others obtained data proposing the opposite, namely that TH1 cells enter before TH17 cells [34].

Although IL-17A is largely dispensable for EAE development, there is growing evidence that other soluble factors produced by IL-23 polarized T cells play non-redundant roles in the development of demyelinating lesions. Granulocyte monocyte colony stimulating factor (GM-CSF) has been identified as an IL-23 driven cytokine [35••, 36••]. Furthermore, GM-CSF might enhance IL-23 production by antigen presenting cells, thereby forming a positive feedback loop [36••]. Autoreactive helper T cells specifically lacking GM-CSF fail to initiate neuro-inflammation despite expression of IL-17A or IFNγ, whereas GM-CSF secretion by IFNγ−/− IL-17A−/− helper T cells is sufficient to induce EAE [36••]. A critical role for GM-CSF in EAE was originally demonstrated by Bernard and colleagues [37]. They found that the resistance of GM-CSF deficient mice to EAE could be overcome by repeated systemic injections of recombinant GM-CSF following active immunization with myelin antigens. It was subsequently shown that the administration of recombinant GM-CSF triggers the mobilization of CD11b+Ly-6Chi monocytes into the circulation of those mice [38••]. These immature monocytes migrate to the CNS during EAE and acquire characteristics of dendritic cells by upregulating CD11c and MHC Class II in situ [38••, 39]. CNS CD11c+CD11b+ cells have potent antigen presenting capacities [40] and have been implicated in epitope spreading [41]. Consistent with these findings, bone marrow chimera experiments have pinpointed peripheral myeloid cells, as opposed to resident microglia, as the key GM-CSF responder population in EAE [36••]. A parallel mechanism by which GM-CSF potentiates EAE was revealed by the discovery that GM-CSF dependent CD103+ dermal dendritic cells are superior to other dendritic cell subsets in activating encephalitogenic T cells in the periphery [42]. GM-CSF is the only known T cell-derived cytokine which has a non-redundant function in EAE, and thus represents the first identified IL-23-driven encephalitogenic T cell cytokine.

The plasticity of TH17 cells and implications regarding their mechanism of action in EAE

The quest to define TH17 cells more accurately and to pinpoint those characteristics responsible for their pathogenicity in autoimmune disease is complicated by the plasticity of the TH17 ‘lineage’. Human, as well as murine, TH17 polarized cells can upregulate T-bet and IFNγ and downregulate IL-17 after activation with Th1 polarizing factors in vitro [43••, 44]. A similar phenomenon has been observed in vivo [45•, 46]. Using IL-17A reporter mice, Hirota and colleagues recently showed that up to two thirds of CNS-infiltrating TH17 cells (defined as CD4+ T cells that express IL-17A and/or have expressed IL-17A at any point in the past) express the TH1 signature cytokine, IFNγ in C57BL/6 mice with MOG35–55 induced EAE [45•, 47•]. These so called ‘ex-TH17’ cells could easily have been mistaken for ‘authentic’ TH1 cells in previous studies, particularly since they turn off expression of IL-17A and RORγt and upregulate the TH1 associated IL-12 receptor γ2 chain. The ability of encephalitogenic Th17 cells to lose IL-17 expression and upregulate IFNγ in vivo was originally shown in an independent study using IL-17F reporter mice [47•]. Interestingly, Hirota et al. found that the conversion of TH17 cells into IFNγ+ IL-17− cells was IL-23 dependent [45•].This raises the question of whether the acquisition of TH1-like characteristics and encephalitogenicity by ex-TH17 cells are causally linked. Kinetic studies indicate that TH17 cells pass through an IL-17A/IFN γ double positive stage during their transition into ex-TH17 cells[45•]. The presence of IL-17A/IFNγ double positive CD4+ T cells in perivascular infiltrates and the adjacent white matter of CNS tissues from patients with MS suggest that ex-TH17 cells might play a prominent role in the human disease as well [48]. Furthermore, a significant percentage of CD4+CD45RO+ lymphocytes isolated from the peripheral blood of acutely relapsing MS patients co-express IFNγ and IL-17 following activation in the presence of IL-23 [48]. Hence, IL-23 might drive the development of ex-TH17 cells during MS as well as EAE.

Whereas TH17 cells are highly unstable, the stability of GM-CSF expression by pathogenic T cells is unknown. In mice, TH1 and TH17 polarized cells acquire GM-CSF expression upon CNS invasion [36••]. Apparently in vivo, the TH17 transcription factor RORγt is involved in GMCSF expression by TH cells [35••, 36••], whereas in vitro, GM-CSF secretion can be governed by different cytokine receptors and transcription factors [35••]. The long-term commitment of the GM-CSF-secreting pathogenic phenotype and the molecular factors that drive their emergence remain to be investigated in detail.

TH17 cells in MS

The potential importance of TH17 cells in multiple sclerosis was heralded by early reports of IL-17 gene expression in peripheral blood (PB) and cerebrospinal fluid (CSF) mononuclear cells obtained from MS patients, as well as in MS plaques themselves [49, 50]. The latter observation was made as the result of a microarray analysis comparing fresh-frozen brain tissues obtained at autopsy from MS patients versus controls without CNS pathology [50]. Genes involved in IL-1 and IL-6 signaling, which drive human TH17 differentiation, were also differentially expressed in the MS lesions [50]. Subsequent studies have used flow cytometry and ELIspot to confirm the presence of T cells producing IL-17A protein at elevated frequencies in the blood and cerebrospinal fluid of untreated individuals with relapsing remitting MS [51–53]. In several of those studies, IL-17A expression was highest during a clinical relapse. Furthermore, two independent groups have documented the accumulation of CD4+ and CD8+ T cells expressing IL-17A protein in active MS lesions by immunohistochemical analysis [48, 54]. One study suggests that the increased frequency of TH17 cells in MS is secondary to elevated IL-23 production by monocyte derived dendritic cells[55].

Although it remains to be determined whether IL-17 and/ or IL-17 producing T cells directly mediate CNS tissue damage in MS, the administration of disease modifying agents has been associated with a reduction in peripheral TH17 responses that parallels clinical improvement. This was most recently shown to be the case for fingolimod (FTY-720), the first oral drug to be approved by the Food and Drug Administration to suppress relapses and delay disability progression in patients with relapsing forms of MS [56]. Compared to untreated MS patients or healthy volunteers, MS patients treated with fingolimod had a lower frequency of circulating CD3+ TH17 cells, as defined by production of IL-17A in response to stimulation with plate-bound anti-CD3 and anti-CD28 ex vivo [57]. Similarly, CD4+ T cells staining positive for intracellular IL-17A were reduced in the peripheral blood of MS patients following 6–12 months of treatment with recombinant IFNγ, commonly prescribed as a first line disease modifying agent in relapsing disease [52]. Several mechanisms of action of IFNβ in curtailing TH17 cells have been proposed. Durelli and colleagues reported that IFNβ enhances activation-induced apoptosis of TH17, but not of TH1, cells [52]. TH17 cells expressed higher levels of the Type I interferon receptor 1 chain than TH1 cells, possibly accounting for their increased sensitivity to the cytokine. In an independent study, IFNβ prevented the upregulation of RORc and IL-23 receptor by human CD4+CD45RA+ lymphocytes cultured under TH17 polarization conditions. By contrast, IFNβ had no effect on expression of the TH1 and TH2 associated transcription factors, T-bet and GATA-3 [58]. IFNβ could also block TH17 differentiation by suppressing IL-23 and IL-1β and inducing IL-27 secretion by dendritic cells [58, 59]. Collectively, these data are at odds with the findings by Axtell et al. that IFNβ therapy suppresses TH1-mediated, but exacerbates TH17 mediated, EAE and that an elevated level of IL-17F in the serum of MS patients is predictive of non-responsiveness to IFNβ [60]. Hence the ultimate usefulness of TH17-related factors as predictive or surrogate biomarkers in MS remains to be established.

It is also unclear why neutralizing antibodies specific for the IL-12/IL-23 p40 chain were ineffective in preventing inflammatory lesion formation or clinical exacerbations in relapsing remitting MS [61, 62]. One possible explanation is that, because of the blood-brain-barrier, anti-IL-12 p40 antibodies did not gain access to the CNS at high enough levels to achieve a notable therapeutic effect. Alternatively, IL-23 might be more important for the initial generation or expansion of autoreactive T cell clones at disease initiation, than for their long-term survival and maintenance in established MS [61]. These issues will only be addressed by future clinical trials of agents that anatagonize TH17-related factors and have high CNS penetrance.

Concluding remarks

Investigations into the mechanism of action and long-term stability of TH17 cells in EAE, and of their biological significance in MS, have raised more questions than they have answered. The collective data has shown that exposure of autoreactive T cells to IL-23 is a more accurate determinant of pathogenic potential than the secretion of individual TH17 associated effector cytokines. The IL-23 induced events that imprint an encephalitogenic program are only now being uncovered. In MS, the accumulation of IL-17 producing T cells in the circulation and CNS compartments correlates with disease activity. However, the failure of anti-IL-12/23 neutralizing antibodies to reduce MS relapse rates and to prevent radiological lesion formation belies the causality of that association. Nonetheless, the identification of myelin-reactive IL-23-driven T cells as encephalitogenic mediators represents an important advance in our understanding of the pathogenesis of autoimmune demyelination. Future research on the subject promises to yield novel candidate biomarkers and therapeutic targets in MS.

Papers of particular interest, published within the period of review, have been highlighted as:

• of special interest

•• of outstanding interest