NLRP3 Inflammasome and MS/EAE

Inflammasomes are cytosolic sensors that detect pathogens and danger signals in the innate immune system. The NLRP3 inflammasome is currently the most fully characterized inflammasome and is known to detect a wide array of microbes and endogenous damage-associated molecules. Possible involvement of the NLRP3 inflammasome (or inflammasomes) in the development of multiple sclerosis (MS) was suggested in a number of studies. Recent studies showed that the NLRP3 inflammasome exacerbates experimental autoimmune encephalomyelitis (EAE), an animal model of MS, although EAE can also develop without the NLRP3 inflammasome. In this paper, we discuss the NLRP3 inflammasome in MS and EAE development.


In�a��a�o�e�
In�ammasomes are cytosolic sensors that detect pathogens and stresses in order to mature and secrete proin�ammatory cytokines, such as interleukin-1 (IL-1 ) and IL-18. In�ammasomes are expressed in phagocytes, such as macrophages and dendritic cells (DCs), and form a multiprotein complex that activates caspase-1. Assembly of in�ammasomes that have clear physiological functions in vivo has been reported with relatively few NOD-like receptor (NLR) family members, such as NLRP1, NLRP3 (also called cryopyrin, CIAS1, NALP3), NLRC4 (IRAF), and AIM2 [1]. e NLRP3 in�ammasome is currently the most fully characteri�ed in�ammasome and is comprised of three different proteins: NLRP3, adapter protein apoptosis-associated speck-like protein (ASC), and procaspase-1. NLRP3 protein is autorepressed by an internal interaction between the NACHT domain and leucine-rich repeats (LRRs) (Figure 1(a)) [2,3]. Derepression of NLRP3 is essential for the interaction between NLRP3 and ASC through their Pyrin domains (PYD), followed by further interaction between ASC and procaspase-1 through CARD domains (caspase activation and recruitment domains) (Figure 1(b)). Oligomeri�ation of the NLRP3 in�ammasome heterotrimer unit leads to procaspase-1 self-cleavage to generate activated caspase-1, which processes maturation of IL-1 and IL-18 and elicits rapid release of those in�ammatory cytokines by cell death termed "pyroptosis" (Figure 1(c)). Molecular mechanism by which caspase-1 mediates pyroptosis is still elusive, but is distinguished from apoptosis and necrosis [4,5]. A molecule termed CARDINAL is known to be involved in the human NLRP3 in�ammasome [6] (Figure 1(b)); but its function is unknown and there is no mouse homolog of human CARDINAL. Critical role of CARDINAL in eliciting functions of in�ammasomes is �uestioned, because mouse in�ammasomes share basic biological functions with human in�ammasomes, despite the absence of CARDINAL. It is, at least, known that CARDINAL is dispensable for IL-1 production in human cells [7].
Multiple sclerosis (MS) is an autoimmune in�ammatory demyelinating disease of the central nervous systems (CNSs) mediated by myelin-speci�c autoreactive T cells. ere are a number of reports that strongly suggest the involvement of the NLRP3 in�ammasome (or in�ammasomes) in the development of MS. MS-like lesions were observed in a MWS patient who had a disease-susceptible mutation in the Nlrp3 gene [27,28]. Expression of caspase-1 is elevated in MS plaques and peripheral blood mononuclear cells of MS patients [29,30]. Abundance of caspase-1, together with that of IL-18, is also identi�ed in peripheral mononuclear cells from MS patients compared to those cells from healthy controls [31]. It is of note that various human SNPs of caspase-1 have already been identi�ed [32].
Quite a few studies showed correlation between severity of MS and IL-1 (and its receptor), which is a major cytokine processed by in�ammasomes [33][34][35][36][37][38][39][40]. Levels of IL-1 in the cerebrospinal �uid (CSF) in MS patients are upregulated and correlated with susceptibility and progression of relapseonset MS [33,38,39]. Treatment of MS patients with glatiramer acetate or IFN is known to elevate the levels of endogenous IL-1 receptor antagonist [41,42]. On the other hand, an NLRP3 in�ammasome activator, ATP, is detected by a purinergic receptor, P2X7R. Elevated P2X7R expression in spinal cords of MS patients was observed [43,44], and glatiramer acetate reduced the P2X7R expression levels [45]. Gain-of-function single nucleotide polymorphisms in the P2X7 receptor gene are considered to be associated to MS [46]. ose studies suggest the contribution of extracellular ATP to the development of MS. In addition to the levels of ATP, those of uric acid, which also activate the NLRP3 in�ammasome, are also upregulated in the CSF of MS patients [47], and serum uric acid level in patients is potentially associated with susceptibility of MS [48]. It should be noted that increased cathepsin B activity in brains of MS patients was reported as well [49]. Because the word �in�ammasomes� had not been used in earlier publications, the connection between MS and NLRP3 in�ammasome was less visible. As we recently demonstrated the presence of an NLRP3 in�ammasome-independent subset in experimental autoimmune encephalomyelitis (EAE), an animal model of MS [50], the NLRP3 in�ammasome may not be involved in the development of all kinds of MS, which is a multifactorial and heterogeneous disease. However, these studies strongly suggest the general involvement of the NLRP3 in�ammasome in MS.
Ting and colleagues �rst reported the critical role of the NLRP3 in�ammasome in EAE development using mice lacking gene encoding NLRP3 ( −/− ) [51]. Antigen . We also showed that −/− mice, which lack the gene encoding ASC, showed similar phenotypes to those in −/− mice [50,52]. Because ASC is shared with other in�ammasomes than NLRP3, the results suggest the ma�or contribution of the NLRP3 in�ammasomes in EAE pathogenicity among in�ammasomes. Caspase-1-de�cient mice are also resistant to EAE, supporting the involvement of in�ammasomes (most probably NLRP3 in�ammasome) in EAE pathogenicity [54,55].
A tie between the NLRP3 in�ammasome and EAE development is also supported by a number of other studies demonstrating enhanced levels of caspase-1, IL-1 , and IL-18 during EAE development [50,51,54]. NLRP3 in�ammasome requires activation to exert its function; therefore, we evaluated and con�rmed the activation of the NLRP3 in�ammasome during EAE progression by detecting active caspase-1 (p20) in splenocytes and in CNS and high levels of IL-1 and IL-18 in serum and in spleen [50]. IL-1R-de�cient mice and mice treated with a recombinant IL-1R antagonist (IL-1Ra) displayed mild EAE with reduced numbers of 17 cells [56,57]. In addition, IL-18 expression is increased in serum from mice that develop EAE [51], and IL-18 expression in EAE mice depends on the presence of ASC and NLRP3 [50]. IL-18-de�cient ( −/− ) mice display signi�cantly mild EAE [51,58]. Another report to suggest the involvement of the NLRP3 in�ammasome in EAE is amelioration of the disease by P2X7 receptor blockade, which prevents cells from detecting ATP [59]. Taken together, these studies demonstrated that the NLRP3 in�ammasome plays an important role in EAE.

Impact of IL-1 and IL-18 on EAE
IL-1 and IL-18 are cytokines matured by the NLRP3 in�ammasome. Involvement of IL-1 and IL-18 in EAE progression had long been speculated. IL-1 plays a role in reversing demyelination and breakdown of blood-brain barrier, which prevents in�ltration of T cells and other cells into the CNS in healthy individuals [60]. IL-1 also induces activation of microglia [61], which stimulate CNS-in�ltrated T cells by presenting self-antigens during EAE development. IL-1 , together with IL-23, promotes IL-17 expression both by T cells and CD4 + T cells [56,62]. (In particular, the critical role of 17 cells in EAE has been well documented.) IL-18 has been studied in the context of EAE. High Il18 mRNA levels were found in the brain and the spinal cord at the onset and throughout the course of EAE [63,64]. ere are con�icting results on EAE in IL-18-de�cient mice. Two groups showed signi�cantly mild EAE in −/− mice by defects in mounting autoreactive 1/17 and autoantibody responses [51,58], but another showed only slightly milder EAE in −/− mice compared to WT mice [65]. (e latter study also showed that IL-18 receptor-de�cient mice were completely resistant to EAE.) In both studies, −/− mice were reasonably backcrossed enough, and the backcross generations do not look to be a major issue. Rather, we suspect that the discrepancy could come from the distinct EAE inducition regimens among those studies. In a report that showed resistance to EAE in −/− mice, transfer of splenocytes from immunized −/− mice to recipients did not induce EAE [58]. e result is congruent with our �nding that transfer of peripheral T cells from immunized −/− and −/− mice to recipients failed to induce EAE [52]. ese results suggest a pathogenic role of IL-18, which is matured by in�ammasomes in APCs during the activation of peripheral T cells a�er immunization. NLRP3 in�ammasome-mediated IL-18 production is also known to exacerbate demyelination [53] and to promote IL-17 production by T cells as well as CD4 + T cells [66] during EAE progression. Based on these �ndings and the proin�ammatory character of IL-1 and IL-18, both cytokines are involved in EAE progression.

EAE Development by Immune Cell Chemotaxis Induced by the NLRP3 In�ammasome
More recently, the NLRP3 in�ammasome was shown to mediate EAE progression by inducing chemotactic ability of immune cells, rather than augmenting the 17 cell population [52]. Reduced 17 cell population in −/− and −/− miceappears to be one of mechanisms by which the absence of the NLRP3 in�ammasomes ameliorates EAE, but it is not a causal factor of the resistance to EAE in −/− and −/− mice. Instead, −/− and −/− mice developed mild EAE because 17 cells generated in those mice were not equipped to migrate into the CNS [52]; therefore, the quality of 17 cells matters in the resistance of −/− and −/− mice to EAE, rather than the quantity of 17 cells. 17 cells (and other T helper cells) from −/− and −/− mice express low levels of CCR2, CXCR6, and osteopontin (OPN) [52] (Figure 2). Interestingly, chemotaxis of APCs in immunized −/− and −/− mice was also impaired by diminished gene expression of CCL7/MCP3 (CCR2 ligand), CCL8/MCP2 (CCR2 ligand), CXCL16 (CXCR6 ligand), and 4 1 integrin (OPN receptor) [52] (Figure 2). Intriguingly, those molecules that are upregulated during EAE development by the presence of the NLRP3 in�ammasome are matching chemokine/receptor pairs between T cells and APCs. In vitro analyses to evaluate cellular chemotaxis indeed demonstrated that T cells and DCs from immunized −/− and −/− mice were severely compromised in migration ability [52].
Roles of these chemokines and their receptors in EAE development were reported. Previous studies showed that −/− mice recruit reduced numbers of mononuclear cells in the CNS and show mild EAE [67,68]. Actually, CCR2 expression in circulating CD4 + T cells is signi�cantly elevated during MS relapse [69,70]. On the other hand, roles of CCR2 ligands, CCL7/MCP3 and CCL8/MCP2, in EAE are less characterized, compared to other proteins detected in our study [52], but some MS studies suggest the involvement of MCP2 and MCP3 in the disease development. For example, both MCP2 and MCP3 are considered to be involved in the development of MS lesions in the CNS [71]. A genetic polymorphism was also reported in the promoter-enhancer region of the MCP3 gene of MS patients [72]. CXCR6 is required for neuroin�ammation by immune cell in�ltration in cortical injury sites [73].

Two Subtypes of EAE: NLRP3 Dependent or Independent
Two initial reports on the NLRP3 in�ammasome in EAE showed contrasting results. One showed susceptibility to EAE of −/− mice [85] and the other showed resistance of −/− mice [51]. e results appeared to be con�icting, but they do not: EAE can be induced with or without the NLRP3 in�ammasome depending on the intensity of immunization [50]. Such NLRP3 in�ammasome-independent EAE can be induced by aggressive immunization. For example, high doses of heat-killed Mycobacteria in complete Freund's adjuvant (CFA) are sufficient to induce EAE in −/− and −/− mice [50]. Similar observation was reported in a study of caspase-1-de�cient mice, in which disease susceptibility is associated with the number of immunization, the dose, and the MHC-binding affinity of antigen peptide [54]. erefore, intensive antigen presentation by APCs to T cells appears to break tolerance and this might induce NLRP3 in�ammasome-independent EAE. On the other hand, passive EAE, induced by an adoptive transfer of activated T cells to recipient mice, is NLRP3 in�ammasomedependent [50]. As passive EAE induction does not include CFA, the result suggests that adjuvant is not essential for activating the NLRP3 in�ammasome. Importantly, we further demonstrated that the treatment of IFN , which is a �rstline treatment for MS, is not effective when EAE progression is independent of the NLRP3 in�ammasome [50]. In human MS disease, such an arti�cial disease induction is not involved; therefore, it is currently not clear whether and how MS progresses in an NLRP3 in�ammasome-independent fashion.

Conclusion and Perspectives
Although EAE is an excellent model to understand MS, EAE and MS are not the same disease and the �ndings in EAE may not be able to fully apply to MS. However, without using animal models, it is almost impossible to understand the role of in�ammasomes in the development of those neuroin-�ammatory diseases. Congruent with a number of studies on MS patients, animal EAE models strongly suggested the involvement of in�ammasomes, NLRP3 in particular, in autoimmune neuroin�ammation. Because of the relatively recent identi�cation of the structure of in�ammasomes and its characterization, many studies of MS and EAE did not mention "in�ammasomes, " despite descriptions of caspase-1, IL-1 , and other in�ammasome-related molecules. As discussed in this paper, now we started to understand the role of the NLRP3 in�ammasome in EAE development. To develop better MS treatments, it may be necessary to understand innate in�ammation. Mechanistic understanding of innate in�ammation caused by the NLRP3 in�ammasome using EAE models and testing hypotheses obtained from EAE studies in MS are expected to bring a considerable progress in the efforts to treat MS.