The role of nuclear receptors in regulation of Th17/Treg biology and its implications for diseases

Abstract

Nuclear receptors play an essential role in cellular environmental sensing, differentiation, development, homeostasis, and metabolism and are thus highly conserved across multiple species. The anti-inflammatory role of nuclear receptors in immune cells has recently gained recognition. Nuclear receptors play critical roles in both myeloid and lymphoid cells, particularly in helper CD4+ T-cell type 17 (Th17) and regulatory T cells (Treg). Th17 and Treg have a major impact on cellular fate through their interactions with cytokine signaling pathways. Recent studies have emphasized the interactions between nuclear receptors and the known cytokine signals and how these interactions affect the expression and function of master transcription factors in Th17 and Treg subsets. This review will focus on the most recent discoveries concerning the roles of nuclear receptors in regulating the Th17/Treg cell-fate determination.

The list of helper CD4+ T-cell subsets has grown extensively since the initial discovery of helper CD4+ T-cell type 1 (Th1) and type 2 (Th2).¹ Additions to this extending list include regulatory T cells (Treg), type 17 (Th17), follicular helper T cells (Tfh), and type 9 (Th9).^2,3,4 As this list has grown, molecular aspects have been elucidated through the identification of master transcription factors unique to each subset.

The differentiation of Th17 and Treg is tightly linked, similar to that of the Th1 and Th2 subsets that lead to the dichotomous fates of CD4+ T cells.⁵ Different cytokines have the potential to tip the subtle balance of Treg/Th17 lineages and alter the cellular fate. This cell-fate decision is highly context-dependent, with TGF-β1 acting as a required cytokine for both Th17 and Treg differentiation.⁶ IL-6- or IL-21-dependent STAT3 activation, together with TGF-β1, can induce differentiation of the Th17 subset that secretes proinflammatory cytokines IL-17A/F and IL-22.⁷ The Th17 subset is critical for the immune response against bacterial and fungal infections.⁸ Conversely, excessive levels of the Th17 subset in peripheral blood and lesions are associated with pathogenesis in patients with multiple sclerosis (MS), rheumatoid arthritis (RA), psoriasis, Crohn's disease, and ulcerative colitis.⁹ Canonical transcription factors that confer a Th17 phenotype in CD4+ T cells are ROR-γt and ROR-α, members of a nuclear receptor family of proteins.^7,10 ROR-γt and IL-17A/F expression levels can be further regulated by other transcription factors, including HIF-1α, STAT3, STAT1, and STAT5.^11,12,13,14

IL-2 and TGF-β1 induce and maintain the expression of Foxp3, which is a master regulator of Tregs.¹⁵ Tregs are critical for maintaining self-tolerance and homeostasis in a host organism. At the same time, under certain circumstances, the overproduction of Tregs hinders an effective immune response to tumors, and the loss of Tregs is associated with major autoimmune diseases.¹⁶ The Treg subset is further divided into natural Tregs (nTregs) derived from the thymus and induced Tregs (iTregs) derived from the periphery.¹⁷ Modes of Foxp3+ Treg-dependent suppression of immune mediators include the (i) inhibition of proinflammatory cytokines such as IFN-γ and IL-2 and (ii) trans-suppression of T cells, dendritic cells, and macrophages through contact-dependent and contact-independent mechanisms.¹⁸ Sustained Foxp3 expression is critical for the capacity of Tregs to negatively regulate the immune response, and its expression is tightly regulated by many transcription factors, including STAT5 and RAR.^19,20 More recently, it was found that Foxp3 expression can be further modified through HIF-1α, Stub1, and USP7 via posttranslational mechanisms.^14,21,22 Foxp3 can also antagonize ROR-γt, thereby inhibiting Th17 differentiation in favor of Treg differentiation.^23,24

Many recent studies have identified nuclear receptors as additional modulators of Th17/Treg development.^25,26 These nuclear receptors can either positively or negatively regulate cell differentiation and function. In this review, we begin with a brief, general introduction to how nuclear receptor subfamilies differentially regulate Th17/Treg differentiation and function in relation to other transcription factors. In general, this review focuses on iTregs (denoted henceforth simply as “Treg”), and nTregs are specified as such when relevant.

Nuclear receptor signaling and mechanisms of gene regulation

Nuclear receptor-dependent signaling differs from cytokine-mediated signaling in that it is mediated by membrane-anchored receptors. A nuclear receptor signal is initiated by a lipid-derived molecule that penetrates the plasma membrane.²⁷ Thus, the signaling cascade begins in the cytosol, and ligand-bound nuclear receptors subsequently translocate the signal into the nucleus by directing gene expression. A well-known subfamily of the nuclear receptor superfamily includes steroid receptors, which were discovered initially in 1985 with the glucocorticoid and estrogen receptors.²⁸ Since then, our knowledge of nuclear receptors has significantly expanded.

Later, nonsteroidal receptors with different activation mechanisms were discovered. Steroid receptors form homodimers and interact with ligands,²⁹ whereas nonsteroidal receptors form heterodimers and have a common receptor, RXR (retinoid-X-receptor), which recognizes 9-cis-retinoic acid.³⁰ This subfamily of nuclear receptors consists of many members, including thyroid hormone receptors (T3R), all-trans retinoic acid receptors (RAR), peroxisome proliferator of activated receptors (PPAR), and liver-X-receptors (LXR). Each nonsteroidal receptor subfamily consists of multiple subtypes, which are each expressed by different alleles and bind to a unique set of ligands²⁷ (Table 1). However, the ligands for many nuclear receptors are still unknown, and receptors with unidentified ligands are often termed ‘orphan receptors'.³¹ Examples include retinoid acid-related orphan receptors (ROR), nerve growth factor-induced B (NGFI-B) receptors, steroidogenic factor-1 (SF-1) receptors, and estrogen receptor-related (ERR) receptors.

Table 1. Nuclear receptors/agonists/antagonists and their effects on Th17/Treg differentiation.

Nuclear receptors	subtypes	Cross-talk across nuclear receptors	Ligands	Antagonists	Ligand Effects on Th17	Ligand Effects on Treg	References
ROR	ROR-α, -β, -γ	Unknown	Unknown	Digoxin SR1001	Positive regulator of Th17 development; deletion of ROR-γt and ROR-α impairs Th17 differentiation. Ectopic expression of ROR-γt and ROR-α enhances Th17 development	No direct effects by ROR-γt and ROR-α on Treg development (Foxp3 expression)	(7), (10), (39), (14)
RAR/RXR	RAR-α, -β, -γ	RXR	Natural ligands: all-trans retinoic acid (RAR), 9-cis-retinoic acids (RXR) Synthetic ligands: AM580 (RAR), LG268 (RXR)	Ro 41-5253 LE135 LE540	Inhibit Th17 differentiation	Enhance Treg differentiation	(41), (19), (46), (44)
PPAR	PPAR-α	RXR LXR	Natural ligands : palmitic acid, oleic acid, and leukotriene B4 (LTB4) Synthetic ligands: WY14643, Gemfibrozil, Fenofibrate	GW6471	Deletion of PPAR-α in CD4+ T cells enhances Th1/Th17 differentiation	Deletion of PPAR-α in Tregs impairs suppressive capacity	(31), (32), (33), (34)
	PPAR-γ		Natural ligands: α-, γ-linoleic acid, arachidonic acid, 13-HODE (hydroxyoctadeca-9Z,11E-dienoic acid) and 15-deoxy-delta-prostaglandin J2 (15dPGJ2) Synthetic ligands: pioglitazone, thioglitazone, rosiglitazone	GW9662	Ligand treatment inhibits Th17 differentiation; deletion of PPAR-γ enhances Th17 differentiation	Impaired development of adipocyte tissue specific Treg development; intact development of peripheral Tregs	(26), (27), (28), (29), (30)
	PPAR-β/δ		Natural ligands: prostaglandin A1 (PGA1) and prostaglandin D2 (PGD2), Synthetic ligands: GW501516, L165041	GSK0660 GSK3787	Ligand treatment inhibits Th17 differentiation; Genetic knock-out of PPAR-β/δ in mice results in exacerbated EAE	Unknown	(35), (36)
AHR	AHR1 AHR2 (nonmammalian vertebrate species)	LXR	Synthetic ligands: tetrachlorodibenzo-p-dioxin (TCDD), 3-methylcholanthrene, 6-formylindolo(3,2-b) carbazol Natural ligands: heme metabolites, indigoids, kynurenine (Kyn), 2-(1'H-indole-3'-carbonyl)-thiazole-4-carboxylic acid methyl ester (ITE), other tryptophan metabolites	CH-233191	FICZ treatment enhances Th17 differentiation in vitro and exacerbates pathogenesis in EAE; the lack of AHR gene results in less severe EAE	TCDD treatment induces Treg differentiation and ameliorates EAE	(37), (38), (39), (40), (41), (42), (43), (44)
NGFIB	Nr4a1 (Nur77), Nr4a2 (Nurr1), Nr4a3 (NOR-1)	Unknown	Unknown; orphan receptor	N/A	Deletion of all three isoforms results in complete abrogation of natural Treg(nTreg) development and the development of Th17	Deletion of Nr4a1 results in enhanced thymic Treg population; deletion of Nr4a2 inhibits iTreg development, while ectopic expression of Nr4a2 imparts functional iTreg	(45), (46), (47), (48), (49), (50)
VDR	VDR	Unknown	Natural ligands: 1,25-dihydroxyvitamin D3	N/A	Deletion of VDR leads to enhanced Th17 development	Deletion of VDR leads to impaired Treg development and loss of suppressive capacity by Tregs; addition of vitamin D enhances Treg development	(51), (52), (53)
LXR	LXR-α (NR1H3), LXR-β (NR1H2)	RXR AHR	Natural ligands: oxysterolSynthetic ligands: GW3965, T0901317	GSK2033	Deletion of LXR leads to enhanced Th17 development and severe pathogenesis during EAE; addition of LXR ligands impairs Th17 development	Addition of LXR enhances Treg development	(54), (55), (56)

The ligand-mediated activation of nuclear receptors can cause the induction or repression of gene expression in many different ways.³² The regulatory mechanism is unique in that many nuclear receptors modulate gene expression by recruiting transcriptional co-activators or co-repressors in processes known as transactivation and transrepression, respectively. Transactivation involves the recruitment of histone acetylases such as P300/CBP and steroid receptor co-activators (SRCs), whereas transrepression involves the recruitment of histone deacetylase (HDAC) through the interaction between nuclear receptors and SMRT/NcoR (silencing mediator for retinoid and T3R/ nuclear co-repressor) complexes.

The underlying themes of ligand-mediated nuclear receptor activation can be summarized by the following principles. First, nuclear receptors require the presence of natural or synthetic ligands to direct Th17 or Treg differentiation. In this context, genetic deletion of the receptor of interest negates the effect of the ligand on the cells, indicating the ligand is specific for the receptor. Second, even in the absence of synthetic ligands, the deletion of a nuclear receptor can affect Th17 or Treg differentiation. This is true for many nuclear orphan receptors, suggesting the existence of endogenous ligands. In this review, we address the most recent available findings for each individual subfamily of nuclear receptors concerning their effects on Th17 and Treg differentiation. Specifically, we address the phenotypes of nuclear receptor knockout (KO) mice and the effects of treatments with agonists and/or antagonists in in vitro and/or in vivo models.

Retinoic acid-related orphan receptors (ROR)

ROR is a member of a nuclear receptor superfamily with unidentified ligands. This superfamily consists of the α, β, and γ subtypes, each of which have several isoforms generated by alternative splicing.³³ ROR-γt (one splicing isoform of RORγ, also known as RORγ2) was found to be a canonical transcription factor required for Th17 differentiation.⁷ ROR-γt is uniquely expressed in IL-17A- and IL-17F-expressing Th17 cells that were signaled to differentiate by TGF-β1 and IL-6 or IL-21. Genetic deletion of murine ROR-γt in CD4+ T cells significantly decreased Th17 differentiation, as shown in both in vitro and in vivo studies. Nevertheless, the lack of ROR-γt does not completely abrogate Th17 differentiation, suggesting that an additional transcription factor promotes Th17 development. Another study found that the subtype ROR-α regulates Th17 differentiation as well and that the combined genetic ablation of both ROR-γt and ROR-α leads to a complete loss of Th17 development.¹⁰ In contrast, the overexpression of both ROR-γt and ROR-α synergistically enhances Th17 differentiation, suggesting that both receptor subtypes are critical for Th17 development. Both ROR-γt and ROR-α induce Il17a/f transcriptional activity, and ROR-γt interacts with HIF-1α and STAT3 form a transcriptional complex.¹⁴ Similar to the situation in murine Th17 cells, TGF-β1 combined with IL-6, IL-1β, or IL-23 induces ROR-γt expression in human naïve CD4+ T cells, which is required for human Th17 differentiation.^34,35 However, the ROR-α transcription levels were relatively low in differentiated human Th17 cells, and whether ROR-α plays a role in human Th17-cell development remains uncertain.³⁴

Although endogenous ligands for ROR-γt and ROR-α have not yet been discovered, synthetic derivatives have been found to effectively antagonize both ROR subtypes. Digoxin and its derivatives inhibit ROR-γt activity; furthermore, the compounds inhibit both human and murine Th17 development in vitro and ameliorate the pathogenesis of EAE in mice.³⁶ SR1001, an inhibitor of both ROR-α and ROR-γt, similarly obstructs both human and murine Th17 development by promoting NcoR recruitment and blocking SRC2 recruitment to the murine Il17a promoter.³⁷

In addition to promoting Th17 differentiation, ROR-γt inhibits human Treg differentiation by binding and modulating Foxp3-promoter activity.³⁸ However, siRNA-mediated ROR-γt knockdown in human Tregs did not alter the capacity of ROR-γt to suppress Treg differentiation. In contrast, ROR-γt and ROR-α do not appear to interfere with murine Treg development because ROR-γt/ROR-α deficiency in murine CD4+ T cells minimally affects HIF-1α-mediated Foxp3 downregulation.^14,39 Human ROR-γt binds the human Foxp3 promoter, but ROR-γt in mice does not bind the murine Foxp3 promoter, suggesting that ROR family members may regulate Foxp3 in a species-specific manner.³⁸

All-trans retinoic acid receptor (RAR)

RAR consists of the α, β, and γ subtypes, and its preferential ligand is the naturally produced all-trans retinoic acid (atRA), an active metabolite of Vitamin A. In addition, a RA synthetic derivative AM580 acts as a RAR agonist.⁴⁰ Once bound to its ligand, RAR forms a heterodimer with RXR and binds to RA response elements (RAREs) to regulate gene expression.⁴¹ atRA is known to enhance Treg and inhibit Th17 differentiation. The addition of atRA to murine CD4+ T cells enhances Foxp3 expression in vitro under Treg-polarizing conditions and inhibits IL-17A expression under Th17-polarizing conditions.¹⁹ The administration of atRA in vivo significantly inhibits Th17 and Th1 responses in an EAE model, although the molecule interestingly did not alter Foxp3 expression in CD4+ T cells. This result suggests that additional signals may counteract RA activity in vivo. There appears to be little redundancy among RAR subtypes because RAR-α deficiency in murine CD4+ T cells is sufficient to abrogate atRA-mediated Treg enhancement. It is not clear whether atRA has similar effects in humans, although atRA may stabilize Foxp3 expression and suppress Th17 differentiation in both murine and human nTregs.^42,43

Consistent with RAR-RXR heterodimer formation, treatment with an RXR agonist augments RAR-dependent induction of Foxp3 expression.⁴⁴ Furthermore, this effect was abrogated in the presence of an RXR antagonist. Several mutual and nonexclusive mechanisms have been proposed for atRA-dependent Foxp3 enhancement in T cells. Some results have shown that atRA directly regulates Foxp3 expression by enhancing the phosphorylation of Smad3 and inhibiting the expression of the IL-6 receptor in murine naïve CD4+ T cells.⁴⁵ In contrast, other researchers have shown that Smad3 is dispensable for atRA-mediated regulation, indicating that atRA functions indirectly by inhibiting AP-1/NFATc1-mediated Foxp3 regulation.⁴⁶ Additionally, atRA is reported to indirectly enhance Foxp3 expression by inhibiting cytokines such as IFN-γ, IL-4, and IL-21 from CD44^hi memory CD4+ T cells.⁴⁷

Identified RAR-α antagonists include Ro 41-5253 and LE-135. Ro 41-5253 effectively inhibits Foxp3 induction in murine CD4+ T cells under Treg-promoting conditions in vitro.⁴⁸ LE-135 appears to effectively mitigate atRA-dependent Foxp3 overexpression in colonic biopsies of ulcerative colitis patients and in colonic tissue from 2,4,6-trinitrobenzenesulfonic acid (TNBS)-induced murine colitis.⁴⁹ MS patients who received Vitamin A supplements showed decreased T-cell proliferation in response to peptide stimulation in vitro compared to placebo-treated patients.^50,51 Collectively, these studies present strong evidence that vitamin A supplements may be effective as an immunomodulatory agent in human autoimmune diseases.⁵²

Peroxisome proliferator of activated receptor (PPAR)

PPARs consist of the subtypes α, β, and γ similar to the case for ROR and RAR receptors. PPAR was discovered for its ability to induce rapid proliferation of peroxisomes in hepatocytes treated with anti-diabetic drugs such as fibrate and glitazone, which are synthetic ligands of PPAR.⁵³ Subsequently, it was found that PPAR induces acyl-CoA oxidase (ACO) and mediates peroxisomal beta-oxidation.⁵⁴ PPAR is part of a group of nuclear receptors with extensively identified natural and synthetic ligands, as shown in Table 1. Additionally, each subtype of PPAR appears to play a nonredundant role in Treg/Th17 development.

PPAR-γ was originally discovered as a critical mediator of adipocyte differentiation.⁵⁵ Recently, it was found that PPAR-γ is highly expressed in CD4+ T cells and regulates cellular differentiation. Treatment with pioglitazone, a PPAR-γ ligand, significantly impaired Th17 development from murine naïve CD4+ T cells under Th17-polarizing conditions in vitro.⁵⁶ In addition, in vivo experiments showed that pioglitazone-fed mice developed less severe paralysis in an EAE model, whereas PPAR-γ cKO mice exhibited exacerbated paralysis. Human in vitro experiments showed that pioglitazone treatment specifically inhibited Th17 development (i.e., showed expression of IL-17A/F, IL-21, IL-22, and IL-23 receptors) in peripheral T cells from both healthy and MS patients.⁵⁶ PPAR-γ has been shown to inhibit Il17a expression by recruiting NcoR and SMRT to the rorc promoter and blocking transcription.⁵⁶ Alternatively, PPAR-γ can inhibit the DNA binding activity of STAT3, a critical downstream mediator of IL-6.^57,58

At the same time, PPAR-γ is essential for the development of Tregs in highly enriched adipose tissue in mice.⁵⁵ Mice that are deficient in PPAR-γ in Foxp3+ cells have significantly reduced Tregs in visceral adipose tissue in comparison to wild-type littermates. Conversely, the development of Tregs in the spleen and lymph nodes does not change in these mice, suggesting that PPAR-γ may not directly regulate nTreg development in vivo. Another study found that ciglitazone, which is another PPAR-γ agonist, can enhance murine Treg development and its Th17-suppressive capacity in a murine graft-versus-host disease model.⁵⁹ In contrast, a PPAR-γ KO in Tregs in a murine colitis model increased the susceptibility of the mice to pathogenesis.⁶⁰

Murine CD4+ T cells deficient in PPAR-α or PPAR-β/δ showed enhanced expression of IFN-γ (Th1) and IL-17A (Th17) in an EAE model.⁶¹ It was found that these PPAR-α effects are gender-dependent, with male PPAR-α KO mice showing more severe paralysis and higher levels of IFN-γ and IL-17A secretion but with female PPAR-α KO mice showing pathogenic responses that were comparable to their WT littermates.⁶² Interestingly, in humans, the gender dependence of Th1 and Th17 development in EAE applies to PPAR-α as well as PPAR-γ.⁶³ In contrast, mice treated with a synthetic PPAR-α or PPAR-β/δ agonist are less susceptible to developing paralysis and show enhancement of the Th2 subset.^64,65

It is not clear whether PPAR-α directly regulates Treg development in mice, although we found that WT and PPAR-α KO CD4+ T cells are equally capable of expressing Foxp3 under Treg-polarizing conditions in vitro (unpublished data). In human T cells, PPAR-α agonists bezafibrate and GW7647 can stabilize Foxp3 expression through epigenetic modification of the Foxp3 promoter.⁶⁶ However, PPAR-α is critical for maintaining the suppressive functions of both murine and human Tregs in vitro.^66,67 The molecular mechanism behind PPAR-α- and PPAR-β/δ-dependent Th17/Treg development remains to be elucidated, although one possible scenario involves the interaction between PPAR subtypes.⁶⁸ Similar to other nuclear receptors, PPARs represent important therapeutic targets in major autoimmune diseases. For example, treatment with the PPAR-γ agonist pioglitazone was effective in treating patients with MS.⁶⁹ Additionally, PPAR-β/δ is highly expressed in plaques of psoriasis patients, and PPAR-β/δ antagonists GSK3787 and GSK0660 effectively ameliorated disease severity in a murine psoriasis model.⁷⁰

Aryl hydrocarbon receptor (AHR)

The aryl hydrocarbon receptor (AHR) is activated by environmental toxins and is required to metabolize toxins in the liver.⁷¹ Some of these toxins include halogenated aromatic hydrocarbons such as tetrachlorodibenzo-p-dioxin (TCDD) and polycyclic aromatic hydrocarbons such as 3-methylcholanthrene.⁷² In addition to these environmental ligands, endogenous ligands include dietary supplements and tryptophan metabolites such as 6-formylindolo[3,2-b]carbazole (FICZ).^73,74 Due to its critical role in protecting hosts from environmental toxins, AHR is highly conserved in vertebrates and only one subtype exists in mammals, although some nonmammalian organisms express two isoforms.⁷⁵

Similar to ROR function, AHR is critical for Th17 development. AHR is highly expressed in the murine/human Th17 cell subset, and FICZ treatment enhances IL-17A/F and IL-22 expression in murine/human CD4+ T cells in vitro without altering ROR-γt expression.⁷⁶ Furthermore, the genetic KO of AHR in mice ameliorated paralysis in an EAE model, even though Th17 differentiation in the AHR-KO CD4+ T cells in vitro remains relatively intact. In a subsequent study by the same group, it was found that CH-223191, an AHR antagonist, can inhibit IL-17A expression in murine Th17-polarized CD4+ T cells in vitro.⁷⁷ At a molecular level, AHR appears to directly interact with the Th17-inhibitory STAT1 and does not interact with Th17-promotor STAT3.⁷⁸

The role of AHR in Treg development remains unclear, as FICZ treatment reportedly had no effect on murine Treg development (i.e., Foxp3 expression) either in vitro and in vivo, suggesting its effects are specific to Th17.⁷⁶ In contrast, another group found that murine naïve CD4+ T cells from AHR-KO mice are less capable of Treg development compared to its WT littermates.⁷⁸ Furthermore, the ARH ligand TCDD can reportedly substitute for TGF-β1 in murine Treg differentiation in vitro.⁷⁹ In addition, TCDD treatment can generate human Tregs that are capable of Th17 suppression in vitro through Smad1 and Aiolos.⁸⁰ Consistent with this in vitro data, administration of TCDD to wild-type mice can suppress paralysis in EAE.

It is interesting that TCDD and FICZ, which are both ligands of AHR, exert different effects on Th17/Treg development. Nevertheless, AHR expression is less involved in the regulation of the Treg subset than in that of the Th17 subset,⁷⁶ and TCDD-dependent Treg enhancement could have resulted from a survival advantage of Tregs when faced with TCDD toxicity.⁸¹ However, more recent studies with endogenous ligands such as 2-(1′H-indole-3′-carbonyl)-thiazole-4-carboxylic acid methyl ester (ITE) and kynurenine (Kyn) have shown the critical roles of AHR in murine Treg development in vivo and in vitro.^82,83 Although AHR ligand-mediated Th17/Treg regulation clearly requires further examination, AHR is a promising therapeutic target for the treatment of human autoimmune diseases.

Nerve Growth Factor-Induced B (NGIB)

NGIB is an early gene induced by a variety of external signal mediators, including T-cell receptors (TCRs) on T cells.⁸⁴ NGIB is involved in diverse cellular processes such as metabolism, apoptosis, and DNA repair. The three subtypes of NGIB include Nr4a1 (Nurr1), Nr4a2 (Nurr77), and Nr4a3 (Nor1), which play nonredundant roles. The ligands for Nr4a1, Nra2, and Nra3 are not yet known; thus, these receptors are classified as orphan receptors.⁸⁵ The roles of NGIB in T cells have been recently recognized in the context of Th17 and Treg subsets, similar to many other nuclear receptors. Nr4a1 is essential for the clonal deletion of thymocytes and the development of Tregs in mice.⁸⁶ Nr4a1 KO mice have increased percentages of CD4+ and CD8+ T cells in the thymus compared with WT littermates, a phenotype that has been attributed to diminished levels of the pro-apoptotic molecule Bim. Concurrently, these mice express a higher percentage of nTregs in the thymus, suggesting a critical role of Nr4a1 in early T-cell development.

Nr4a2, another subtype of the NGIB subfamily, is critical for murine Th17 development.⁸⁷ The genetic ablation of Nr4a2 in CD4+ T cells resulted in a failure to induce Th17 differentiation in vitro. Similarly, siRNA-mediated gene silencing of Nr4a2 in T cells offered a level of protection against murine EAE. The mechanism of Nr4a2 is interesting, as the impaired Th17 response is not accompanied by decreased ROR-γt expression. Instead, Nr4a2 seems to be a modulator of IL-23R and thus expression of IL-17A/F and IL-21. In addition, Nr4a2 is required for murine Treg development and function.⁸⁸ Deletion of Nr4a2 in CD4+ T cells impairs Treg development, whereas the exogenous expression of Nr4a2 can lead to functional Tregs in vitro. This effect seems to apply to Nr4a2 but not to Nr4a1 and Nr4a3, suggesting nonredundant roles of the three NGIB subtypes in Treg development.

In contrast, NGIB subtypes appear to play redundant roles in the development of nTregs in mice.⁸⁹ Single KOs of these genes are not sufficient to impair thymic Foxp3+ T cells, suggesting some degree of redundancy in function. Conversely, KOs of all three isoforms result in severe autoimmunity in mice due to a loss of both thymic and peripheral nTregs and enhanced Th1 and Th17 responses. It is not clear why NGIB subtypes appear to have redundant roles in nTregs and not iTregs. It can be speculated that the loss of nTreg development is more detrimental to hosts on an evolutionary scale, and thus organisms develop redundant mechanisms to compensate for the potential loss of one subtype. Alternatively, this variance in redundancy of NGIB subtypes may reflect a more complex difference in the molecular natures of nTregs and iTregs.

Other families of nuclear receptors

Other nuclear receptors involved in Treg/Th17 development and function include the vitamin D receptor (VDR) and LXR. The active metabolite of Vitamin D is 1,25-dihydroxyvitamin D3 (1,25(OH)₂VD₃), which binds to VDR. VDR also forms a heterodimer with RXR and binds to vitamin D response elements (VDREs). The human Foxp3 promoter contains VDREs, and the addition of vitamin D can enhance Foxp3 expression in human T cells under Treg-polarized conditions and enhance its suppressive capacity through cell–cell contacts.⁹⁰ Another study found that 1,25(OH)₂VD₃ inhibits human Treg proliferation but has no effects on its suppression.⁹¹ In a murine model, deletion of VDR in CD4+ T cells leads to enhanced Th17 development with increased IL-17A and IL-21 expression.⁹² At the same time, VDR-KO CD4+ T cells have significantly decreased the expression of Foxp3 in CD4+ T cells under Treg-polarized conditions. Thus, VDR determines Treg/Th17 cell fate through the reciprocal inhibition of alternative fates.

Vitamin D has been implicated in major autoimmune diseases. Serum levels of 1,25(OH)₂VD₃ were significantly correlated with the suppressive capacity of peripheral Tregs in patients with MS.⁹³ Additionally, low vitamin D levels in serum are associated with an increased severity of inflammatory diseases, such as MS, RA, and inflammatory bowel disease.⁹⁴ Because the oral administration of vitamin D3 to patients with Crohn's disease was found to ameliorate disease severity,⁹⁵ vitamin D may be an important regimen for autoimmune diseases.⁹⁶

Finally, the LXR is a highly important nuclear receptor and consists of the two subtypes LXR-α and LXR-β. LXR is a crucial sensor of cholesterol in cells and regulates its metabolism through transcriptional regulation.⁹⁷ For example, LXR activation by ligands can antagonize the sterol regulatory element binding protein (SREBP) pathway for cholesterol synthesis.⁹⁸ Such changes can negatively affect T-cell activation and proliferation. Treatment with the LXR agonist T0901317 inhibits differentiation of murine CD4+ T cells, while treatment with the LXR antagonist GSK2033 enhances cellular proliferation and Th1/Th2/Th17 differentiation. Consistent with the antagonist-treatment results, murine CD4+ T cells deficient in LXR show increased Th17 development under polarizing conditions compared with WT CD4+ T cells in vitro.⁹⁹ In contrast, the administration of LXR ligands in mice inhibits Th17 development in vitro and suppresses EAE in vivo. Interestingly, Srebp-1, which is induced by LXR activation, suppresses IL-17A expression through physical interaction with AHR. Although LXR agonists have been effective in decreasing inflammation in atherosclerosis and some neurodegenerative diseases,^100,101 these compounds need to be examined for therapeutic efficacy in autoimmune diseases.

Summary

Although our understanding of the critical roles of nuclear receptor families has been growing, much remains to be investigated regarding the cross-talk between individual subfamily members and the effects of the specific interactions on Th17/Treg development. For example, the discovery that PPAR-β/δ activation can antagonize PPAR-γ activity⁶⁸ calls for a more comprehensive analysis of activation responses to agonist treatments or genetic deletions of particular receptors in other nuclear receptors. This type of analysis may reveal physical interactions among nuclear receptors such as AHR and LXR. Finally, although many studies have focused on the downstream effects of nuclear receptor signaling, the upstream modulators of nuclear receptors require further investigation (Figure 1).

Figure 1.

Schematic illustration of nuclear receptor-mediated Treg/Th17 differentiation: Pointed arrows represent positive regulation and blunt arrows represent negative regulation of the target. If known, specific mechanisms are indicated next to the arrows. Nuclear receptors are indicated in blue circles or boxes. Other transcription factors (non-nuclear receptors) are indicated by green circles. Overlapping nuclear receptors indicate direct physical interactions. Abbreviations: sterol regulatory element binding protein-1 (SREBP-1), liver-X-receptors (LXR), vitamin D receptor (VDR), aryl hydrocarbon receptor (AHR), peroxisome proliferator of activated receptor (PPAR), all-trans retinoic acid receptor (RAR), retinoic acid receptor-related orphan receptor (ROR).

One of the many challenges involved in studying nuclear receptors is the unique transrepression and transactivation mechanisms of gene regulation. These modes of regulation require additional co-activators and co-repressors for nuclear receptors to mediate gene expression. Additionally, each subfamily of nuclear receptors consists of different subtypes that may play redundant or nonredundant roles. For many subfamilies, the KO of a single subtype in mice results in the manifestation of a clear Th17/Treg phenotype (e.g., ROR, PPAR), but many other subfamilies do not follow this pattern, suggesting that certain genes are redundantly regulated by different subtypes (e.g., NGFIB). Finally, the context-dependent activation of nuclear receptors makes it difficult to determine the specificity of molecular regulation. For example, although PPAR-γ expression is highly induced in both Th2 and Th17 subsets, ligand treatment only affects Th17 differentiation. This result suggests that individual cytokines affect the responses of nuclear receptors and cause differential activation by natural and synthetic ligands.

Another emerging point of interest is the connection between nuclear receptors and cellular metabolism.¹⁰² It has been found that fatty acid oxidation functions as the major source of Treg energy but that Th17 cells obtain energy from glycolysis.¹⁰³ Additionally, it appears that the mammalian target of rapamycin (mTOR) and ERR-α, both of which are sensors of environmental nutrients, can differentially affect T-cell lineage decisions.^104,105,106 Nevertheless, it is not clear how these metabolic regulators modulate IL-17A and Foxp3 expression at the molecular level, and the interactions between nuclear receptors and these metabolic regulators of Th17/Treg remain to be explored.

Finally, nuclear receptors serve as important therapeutic targets. Alteration of the Treg/Th17 ratio is correlated with many autoimmune diseases.^107,108 Nevertheless, many agonists and antagonists are nonspecific to a single subtype of nuclear receptor, and thus off-target effects remain a complicating factor (Table 1). Thus, the development of highly specific agonists or antagonists is critical for successful treatment of autoimmune diseases. At the same time, the use of agonists/antagonists must be deeply examined because Th17 and Tregs are important mediators of anti-tumor immunity. It is known that deletion of Tregs and loss of their suppressive functions can enhance anti-tumor immune responses.¹⁰⁹ One study found that PPAR-α-KO mice exhibit a better control of murine B16-melanoma because of decreased Treg function.¹¹⁰ In contrast, the use of agonists that favor Treg development may impose a risk of attenuated anti-tumor immune responses. These questions warrant future investigation into the effects of nuclear receptors in cancer.