Aryl Hydrocarbon Receptor Promotes RORγt⁺ ILCs and Controls Intestinal Immunity and Inflammation

Abstract

Unlike adaptive immune cells that require antigen recognition and functional maturation during infection, innate lymphoid cells (ILCs) usually respond to pathogens promptly and serve as the first line of defense in infectious diseases. RAR-related orphan receptors (RORγt)⁺ ILCs are one of the innate cell populations that have recently been intensively studied. During the fetal stage of development, RORγt⁺ ILCs (e.g., lymphoid tissue inducer-LTi cells) are required for lymphoid organogenesis. In adult mice, RORγt⁺ ILCs are abundantly present in the gut to exert immune defensive functions. Under certain circumstances, however, RORγt⁺ ILCs can be pathogenic and contribute to intestinal inflammation. Aryl hydrocarbon receptor (Ahr), a ligand-dependent transcriptional factor, is widely expressed by various immune and non-immune cells. In the gut, the ligand for Ahr can be derived/generated from diet, microflora, and/or host cells. Ahr has been shown to regulate different cell populations in the immune system including RORγt⁺ ILCs, T helper (Th)17/22 cells, γδT cells, regulatory T cells (Tregs), Tr1 cells, and antigen presenting cells (APCs). In this review, we will focus on the development and function of RORγt⁺ ILCs, and discuss the role of Ahr in intestinal immunity and inflammation in mice and in humans. Better understanding the function of Ahr in the gut is important for developing new therapeutic means to target Ahr in future treatment of infectious and autoimmune diseases.

Introduction

The mammalian intestinal tract harbors trillions of commensal bacteria. The crosstalk between the microflora and the immune cells is important for the development of intestinal immune system and the maintenance of a healthy environment in the gut. The innate and adaptive immune systems exhibit surveillance functions to fight against invasive pathogens that can cause infectious diseases. However, excessive inflammatory responses can also result in damage in the intestinal tract and cause colitis. In the gut, Th17 cells, Th22 cells, γδT cells, and RORγt⁺ ILCs are main producers of effector cytokines interleukin (IL)-22 and/or IL-17A (hereafter referred to as IL-17). Th17 cells expressing the nuclear receptor RORγt (encoded by the Rorc gene) are one of the T helper cell subsets that mediates extracellular pathogen clearance but also causes autoimmunity when dysregulated (1–5). Th17 cells with both anti-microbial and pro-inflammatory properties are enriched in the intestinal lamina proprial layer and produce signature cytokines IL-17 and IL-22. Th22 cells were originally identified in humans (6, 7). It has recently been shown in mice that Th22 cells can be skewed by IL-6 in vitro and produce mainly IL-22 but little IL-17 (8). Although it remains to be determined whether Th22 and Th17 cells belong to the same subset of T helper cells with different effector cytokine properties, Th22 cells appear to be more effective than Th17 cells during the clearance of Citrobacter rodentium, a murine pathogen that models human enterohemorrhagic E. coli and enteropathogenic E. coli infections responsible for the deaths of several hundred thousand children in developing countries each year. Consistent with the protective role of Th22/Th17 cells, IL-22 has been shown to be an effector cytokine essential for Citrobacter rodentium clearance (3, 9). γδT cells are relatively rare in the lamina propria but are more enriched in the small and large intestinal intraepithelial lymphocytes, which mainly consist of TCRγδ and TCRαβCD8αα cells (10, 11).

γδT cells use different Vγ gene segments to encode γδTCRs at different peripheral sites (12). The intestinal γδT cells that participate in early host defense against pathogens predominantly express TCRVγ5 and can pair with multiple TCRVδ chains (13). A subset of innate lymphoid cells (ILCs) expressing RORγt is essential for gut immunity. RORγt⁺ ILCs and Th17 cells share a number of common features (e.g., transcription factor requirement, cytokine profile, and anatomic location). Given their production of IL-17 and/or IL-22, RORγt⁺ ILCs are also known as ILC17 or ILC22. In this review, we use the term RORγt⁺ ILCs to denote this population of cells. Aryl hydrocarbon receptor (Ahr) is a ligand-dependent transcriptional factor, which functions as an environmental sensor to recognize xenobiotic and/or endogenous compounds. Ahr has been implicated in the development and/or function of all the aforementioned cell populations. In this review, we discuss the development and function of RORγt⁺ ILCs as well as the crosstalk between RORγt⁺ ILCs and other cell populations in the gut. We focus on how Ahr regulates intestinal RORγt⁺ ILC development/maintenance and function, and discuss the potential role of Ahr in human intestinal diseases.

Various cell populations that express RORγt in the gut

Three major cell populations in the gut that express the transcription factor RORγt are Th17/Th22 cells, γδT cells and RORγt⁺ ILCs. These populations of cells share similar cytokine profiles characterized by the production of IL-17 and IL-22. Th17/Th22 cells are abundantly present in the gut under the steady state especially in the small intestinal lamina propria (1, 8, 14). Segmented filamentous bacteria (SFB), a type of commensal bacteria, have been reported to be a potent inducer for Th17 cell differentiation in the small intestines (1). Th17 cells are critical for controlling bacterial intrusion and fungi infection. However, Th17 cells have also been considered to be pathogenic in autoimmune diseases, such as human inflammatory bowel disease (IBD), by secreting pro-inflammatory cytokines such as IL-17, IL-17F, and IL-22 that contribute to tissue inflammation and damage (15–17). Interestingly, IL-17 has also been reported to be protective in CD45RB^hi T cell transfer colitis by inhibiting Th1 cytokines (e.g., IFN-γ), suggesting the intricate regulatory network among the cytokine-induced signaling pathways in the gut (18). IL-22 has dual functions in regard to either protective immunity or pathogenic inflammation in different disease settings. For example, IL-22 acts on gut epithelial cells to promote anti-microbial peptide secretion to clear certain bacterial infections. However, IL-22 can also promote severe inflammation in CD45RB^low T cell-induced colitis by causing mucosal thickening and epithelial hyperplasia (19). IL-17 has been reported to promote the intestinal tumorigenesis in mice bearing a heterozygous mutation in the adenomatous polyposis coli (Apc) gene (Apc(Min/+) mice) (20). Further study showed that IL-23 produced by antigen presenting cells upon microbial product stimulation induces IL-17 production from Th17 cells, which causes barrier deterioration and triggers inflammation and tumor growth (20, 21). γδT cells have been shown to play a dual role in intestinal inflammation (22). It has been reported that mice deficient in TCRδ have worsened DSS-induced colitis and TNBS-induced colitis, suggesting a protective role of γδT cells in gut inflammation (23, 24). However, it has also been shown that γδT cells can produce large amounts of IFN-γ in colitic TCRα^−/− mice (25). Although γδT cells are dispensable for initiating colitis in TCRα^−/− mice, γδT cells in the TCRα^−/− mice are essential for the colitis progression (26). Various immune cells in the gut form an intricate network to combat pathogen infections, while keeping autoimmunity at bay (Figure 1). Here, we will review the recent progress in understanding the development and function of RORγt⁺ ILCs in the context of intestinal immunity and inflammation.

Figure 1. Ahr-mediated transcriptional control of RORγt⁺ ILCs and the immune function of RORγt⁺ ILCs in the gut.

Th17/Th22 cells, γδT cells, and RORγt⁺ ILCs are major cell populations that co-express transcription factors Ahr and RORγt in the gut. Ahr ligands can be generated from diet (e.g., food containing Tryptophan), microflora (e.g., commensal or pathogenic bacteria), and/or host cells. In the gut, Ahr is important for the development of intraepithelial γδT cells, which have immune defensive function. Ahr is essential for the development and function of intestinal RORγt⁺ ILCs. Ahr also modulates IL-7/IL-7R expression, which confers the survival and maintenance of RORγt⁺ ILCs. Ahr regulates IL-22 expression by RORγt⁺ ILCs through cooperative interaction with RORγt, which directly binds at the Il22 promoter to induce transcription. Ahr controls the transcription of Kit that is crucial for RORγt⁺ ILC development by direct binding at the promoter region of the kit. Notch signaling pathway is also involved in the regulation of RORγt⁺ ILC development by Ahr. All the RORγt⁺ cell populations express IL-23R. An active interplay between RORγt⁺ ILCs and antigen presenting cells (i.e., dendritic cells (DCs) and macrophages (Mφ)) through various cytokines (e.g., IL-23, IL-1β, and TNFα) is crucial for the development and function of RORγt⁺ ILCs. RORγt⁺ cells produce large amounts of IL-22 and/or IL-17. Engagement of IL-22R that is expressed by intestinal epithelial cells and gut stem cells promotes anti-microbial peptide secretion and epithelial regeneration. A potential crosstalk between RORγt⁺ ILCs and Th17/Th22 cells mediated by Ahr may be important for the host to fight against bacterial infections, while keeping autoimmunity at bay.

RORγt⁺ ILCs: a subset of innate lymphoid cells

The innate lymphoid cells (ILCs) are referred to as cells that lack rearranged antigen receptors, and respond to pathogens in the first place without cell contact-based signals from antigen presenting cells. There are at least three types of ILCs: ILC1, ILC2, and RORγt⁺ ILCs (27–29). ILC1 cells contain NK cells, which secrete IFN-γ and function similarly to Th1 cells. NK cells are crucial for the clearance of intracellular pathogens (e.g., virus), requiring IL-15 signal for their development. T-bet is the transcription factor that controls the function of ILC1. ILC2 cells include natural helper (NH) cells, nuocytes, and innate helper type 2 (Ih2) cells. ILC2s are lineage marker negative but express CD117, Sca-1 and CD127. Human ILC2 cells express CRTH2, while mouse mature ILC2s express KLRG1 (30, 31). ILC2s are IL-17RB⁺IL-33R⁺, thus highly responsive to IL-25 and IL-33 stimulation, respectively. GATA3 is required for the differentiation and maintenance of ILC2 cells that secrete high levels of Th2-associated cytokines IL-5 and IL-13, suggesting a striking similarity between ILC2 and Th2 cells (32). ILC2 cells are enriched in mucosal tissues (e.g., the intestine, mesenteric lymph nodes, and the lung), playing a protective role in small intestinal helminth infection and contributing to allergic airway inflammation such as asthma (27).

RORγt⁺ ILCs were discovered about two decades ago with the characterization of prototypical lymphoid tissue-inducer (LTi) cells that are involved in secondary lymphoid organogenesis during the fetal stage of development (33, 34). Fetal LTi cells are developed from the common lymphoid progenitors (CLPs) residing in the fetal liver and are characterized as CD4⁻CD3⁻Sca-1⁺c-kit^intIL-7Rα⁺ cells (35). Fetal LTi cells localize in the spleen, blood and lymph node anlagen as early as E12.5, essential for the development of lymph nodes, Peyer’s patches, and nasal-associated lymphoid tissues (36). Bone marrow contains the progenitors for postnatal RORγt⁺ ILCs (37). Although they can be found in the mouse thymus and spleen (38, 39), the adult RORγt⁺ ILCs are mainly present in the intestinal lamina propria, where they are important for the development of cryptopathes and isolated lymphoid follicles (ILFs). Cryptopatches are clusters of cells that are mostly composed of RORγt⁺ ILCs. Microbiota-derived signals can convert cryptopatches to ILFs by accumulating T cells, B cells, and dendritic cells in the cluster. The formation of ILFs helps the generation of IgA-producing plasma cells for the clearance of microflora. Fetal RORγt⁺ ILCs can differentiate in situ in the fetal liver while adult CLPs can only differentiate into mature RORγt⁺ ILCs in the lamina proprial environment (37).

There are at least two types of RORγt⁺ ILCs: LTi cells and NKR-LTi cells. Similar to fetal LTi, adult LTi cells are lineage marker negative with (LTi4) or without the expression of CD4 (LTi0) (40). One sub-population of RORγt⁺ ILCs described recently expresses NK cell receptors such as NKp46 in mice and NKp44 in humans (41–44). This population of cells has been given different names such as NKR-LTi, NKp46⁺ ILC, NCR-22, and NK-22 (28, 40). The precursors of RORγt⁺ ILCs (i.e., LTi cells and NKR-LTi cells) seem to have differential expression of a gut-homing integrin called α4β7 (40). RORγt⁺α₄β₇⁺ fetal liver cells tend to develop into LTi cells, whereas RORγt⁺α₄β₇⁻ progenitors mainly generate NKR-LTi cells (40). LTi cells and NKR-LTi cells are also distributed differently in the gut-associated tissues. Unlike LTi cells that reside in the cryptopatches and ILFs of the gut, NKR-LTi cells are found in the intestinal lamina propria, gut intraepithelial compartment, mesenteric lymph nodes, and Peyer’s patches (41, 45). Although both LTi cells and NKR-LTi cells can produce similar effector cytokine (e.g., IL-22), it remains to be determined whether these two types of RORγt⁺ ILCs exert differential functions in gut immunity and inflammation.

Despite a limited number of studies defining the lineage relationship between NKR-LTi cells and LTi cells, increasing evidence supports that NKR-LTi cells and NK cells represent distinct developmental lineages. Two recent studies using fate-mapping approach showed that NK cells never express RORγt during their development (46, 47). Furthermore, NKR-LTi cells can be developed from NKR⁻RORγt⁺ innate lymphocytes (46). NKR-LTi cells can lose RORγt expression during development, in which IL-7 or IL-12/IL-15 positively or negatively regulate the maintenance of RORγt expression, respectively (46). Genome-wide expression profiling analysis further indicates that NKp46⁺RORγt⁺ cells are more similar to NKp46⁻RORγt⁺ cells (i.e., LTi cells) rather than NKp46⁺RORγt⁻ cells (i.e., classical NK cells) (45).

RORγt⁺ ILCs secrete effector cytokines IL-22 and/or IL-17, and are responsive to IL-1β and IL-23 stimulations. LTi cells produce both IL-22 and IL-17, but NKR-LTi cells only express IL-22 (27). Despite being a population that constitutes only less than 5% of lamina proprial lymphocytes in the gut, RORγt⁺ ILCs have been considered to be one of the major sources of IL-22, which plays a crucial role in intestinal immune defense against pathogens.

Cross-regulation between RORγt⁺ ILCs and T cells

RORγt⁺ ILCs strikingly resemble Th17 cells in their cytokine profile (e.g., production of IL-22 and IL-17). Co-evolution of two systems may be a fail-safe mechanism to implement redundancy in host immunity to certain infections especially at the mucosal surfaces. It has been shown that intestinal Th17 responses are enhanced by Citrobacter rodentium. Most recently, our group and others have shown that ILC-produced IL-22 is essential for the clearance of Citrobacter rodentium in the intestines (48–50). Interestingly, even in the lymphocyte-replete hosts, mice lacking RORγt⁺ ILCs died from the infection, highlighting an essential role for ILCs in gut immunity (9).

The crosstalk between ILCs and adaptive immune system (e.g., T cells) has recently been investigated. Despite rapid early ILC2 expansion after infection, analysis of ILC2 numbers in N. brasiliensis-infected Rag2^−/− mice showed that ILC2 numbers were not maintained in the absence of T cells, suggesting that T cells mediate prolonged ILC2 expansion, migration or survival through an as yet unknown mechanism (51). Compared to those in the immune competent mice, RORγt⁺ ILCs in Rag-deficient mice that lack both T and B cells produce higher level of IL-22, suggesting that the adaptive immunity may suppress the function of RORγt⁺ ILCs (52). It will be interesting to study the cross-regulation between RORγt⁺ ILCs and Th17/Th22 cells under the steady-state conditions and/or during infection/inflammation. Better characterization of different populations of RORγt⁺ ILCs will help elucidation of the interactions between RORγt⁺ ILCs and Th17/Th22 cells.

RORγt⁺ ILCs, antigen presenting cells, and gut microflora

In sharp contrast to an absolute requirement of gut flora in the small intestinal Th17 cell differentiation, the development of RORγt⁺ ILCs seems to be independent of gut microflora. Early reports suggest that microflora are required for the maintenance of NKR-LTi cells by stabilizing RORγt expression in NKR-LTi cells (41, 46). In addition, microflora can induce IL-7 expression, which seems to be specifically required for the development of NKR-LTi cells (46). Consistent with these data, NKR-LTi cells are absent in the fetal or newborn intestines but only appear after weaning when mice start to be populated by gut flora (40). However, recent data indicate that RORγt⁺ ILCs exist in mice treated with broad-spectrum antibiotics as well as in germ-free mice, suggesting that commensal bacteria are at least not necessary for the development of RORγt⁺ ILCs (40, 45). Indeed, IL-22-producing LTi cells are abundantly present in the fetal intestines that are sterile and free of indigenous bacteria. Unexpectedly, by inducing epithelial expression of IL-25, microflora have been shown to inhibit the function of RORγt⁺ ILCs (i.e., the production of IL-22) (52). This effect of inhibition is indirect because RORγt⁺ ILCs do not express IL-25R (i.e., IL-17BR) and requires the presence of CD11C⁺ dendritic cells (DCs) in a cell contact-dependent manner (52).

Microflora can also positively affect the function of RORγt⁺ ILCs through regulation of antigen presenting cells (APCs), especially DCs. Of note, neither the development of RORγt⁺ ILCs nor the production of IL-22 by ILCs under the steady state is dependent on IL-23 ((52); Qiu, Hao, and Zhou, unpublished). However, during infection (e.g., Citrobacter roduentium), enhanced production of IL-22 by RORγt⁺ ILCs requires IL-23 (9). Consistently, highly elevated expression of IL-22 by intestinal RORγt⁺ ILCs was observed upon in vitro IL-23 stimulation (41, 48, 53, 54). Furthermore, bacterial component flagellin has been found to induce IL-22 expression in RORγt⁺ ILCs through action of IL-23 produced by intestinal CD103⁺CD11b⁺ DCs, a population of DCs previously known to induce intestinal tolerance (55, 56). Under the steady state, commensal bacteria can induce the production of IL-1β by macrophages, which enhances IL-22 expression by RORγt⁺ ILCs (57, 58). IL-1β has also been shown to be important for the accumulation of pathogenic ILCs during innate colitis (59). IL-1β and IL-23 can synergize with each other to induce IL-17 production by RORγt⁺ ILCs in Helicobacter hepaticus-triggered intestinal inflammation (59). TNF-α produced by CD103⁻CD11b⁺ DCs can potentiate IL-23-induced innate IL-17 production by ILCs (60). Most recently, CX₃CR₁⁺ phagocytes (DCs and macrophages) have been shown to promote IL-22 production by RORγt⁺ ILCs and Cx₃cr₁-deficient mice died of Citrobacter rodentium infection due to reduced number of RORγt⁺ ILCs and decreased expression of IL-22 (61). Together, these data highlight the importance of cellular crosstalk between APCs and ILCs in gut immunity and inflammation.

Toll-like receptors (TLRs) are expressed by many innate cell populations (e.g., APCs) and can directly recognize structures of microbes. It has been reported that TLR2 protein is expressed by human CD127⁺RORγt⁺ and CD56⁺CD127⁺ LTi-like ILCs (62). Human LTi-like cells proliferated and enhanced the production of IL-22 in response to TLR2 ligand Pam3Cys stimulation (62). Besides TLR2, TLR1, 5, 6, 7, and 9 mRNA transcripts have also been detected in human LTi-like cells (62). These data raise an intriguing possibility that RORγt⁺ ILCs may also respond to pathogens directly, consistent with the prompt acting properties of ILCs during infection.

The transcriptional control of RORγt⁺ ILC development

The ROR family members include RORα, RORβ, and RORγ. While RORβ is not expressed in immune cells, RORα is highly expressed by Th17 cells and contributes to Th17 cell differentiation and function (63). Recent findings also showed that RORα is important for the ILC2 development (64, 65). The Rorc gene encodes both RORγ and RORγt with distinct promoters (66, 67). While RORγ is widely expressed by brain, liver, muscle, and adipose tissues, RORγt expression is restricted to lymphoid lineages of the immune system (68). Th17 cells and RORγt⁺ ILCs have many similarities in their cytokine profile (e.g., production of IL-22 and/or IL-17) and signaling pathways (e.g., functional enhancement by IL-23 stimulation), and share transcription factor RORγt as the “master-regulator”. RORγt is exclusively expressed by fetal LTi cells and is essential for their development. Accordingly, RORγt-deficient mice lack lymph nodes and Peyer’s patches. In adult mice, IL-22-producing intestinal ILCs are absent without RORγt. Recent studies demonstrate that IL-22 is mainly produced by RORγt⁺ cells but not RORγt⁻ cells in the innate cell populations in the gut (48). Thus, RORγt may represent a unique marker for the IL-22/IL-17-producing ILCs.

The inhibitor of DNA binding 2 (Id2) has been shown to be important for the development of fetal LTi cells (36, 69, 70). E proteins encoded by E2A gene, E12 and E47, are required for the development of committed B lymphocyte progenitors (71). Id2, which contains the helix-loop-helix (bHLH) domain but lacks basic DNA binding domain can form dimers with E proteins and inhibit the function of E protein, allowing the cells to adopt NK and LTi progenitor fates (72). Id2-deficient mice have no lymph nodes and Peyer’s patches (69). Id2-deficient mice lack RORγt⁺ LTi cells at E16.5, indicating a pivotal role for Id2 in controlling early lineage commitment of LTi cells. Although there might be other targets of Id2 (73, 74), lowering E protein activity in Id2^−/− mice by deletion of E2A restores LTi cells in the embryos (72), suggesting that E protein is a major target for Id2 to promote the development of LTi cells.

Runt-related transcription factor 1/Core binding factor 2 (Runx1/Cbfβ2) has also been reported to be critical for the development of fetal LTi (75). Runx complexes are composed of a DNA-binding Runx protein and its binding partner Cbfβ. Runx complexes have been shown to play a role in the development of many hematopoietic lineages including B cells, NKT cells, and T cells (76–79). In the fetal liver, the α₄β₇⁺ LTi progenitors can be separated into two populations based on the level of IL-7R expression (75). The α₄β₇⁺IL-7R^high cells are the population that expresses RORγt, and their development is regulated by RORγt and Id2 (75). The α₄β₇⁺IL-7R^high fetal LTi progenitors can be converted from α₄β₇⁺IL-7R^mid cells, a process requiring Runx1/Cbfβ2. Notably, Cbfβ2-deficient mice lack LTi cells but have NK cells, in agreement with the data showing that Cbfβ2 itself regulates the expression of RORγt in fetal LTi cells. As a result, both the Runx1-mutant mice and Cbfβ2-mutant mice lack lymph nodes and Peyer’s patches (75).

Thymocyte selection–associated high-mobility group box protein (Tox), has recently been shown to control the development of the NK and LTi lineages (80, 81). Tox is expressed by NK cells at specific stages and LTi cells (81). Tox^−/− mice have fewer small intestinal LTi cells at E18, resulting in defective lymphoid organogenesis at later stage of the development manifested by the absence of lymph nodes and decreased/smaller Peyer’s patches (81). Tox does not seem to function by regulating RORγt expression directly since the expression of RORγt mRNA is normal in the thymus of Tox^−/− mice. Notably, Id2 was decreased in mature NK cells in the bone marrow of Tox^−/− mice, suggesting that Id2 may be one of the targets of Tox in regulating LTi cell development (81). However, overexpression of Id2 in Tox^−/− bone marrow cells was not sufficient to restore the LTi population (81), arguing that Tox may also act downstream of Id2 for the development of NK and LTi cells, and/or other targets of Tox besides Id2 can regulate the development of NK and LTi cells.

The Notch receptor regulates different stages of T cell development in the thymus (82). Recently, a differential requirement of Notch in the development of bone marrow or fetal liver ILC progenitors has been reported (37, 70). Notch-2 but not Notch-1 is required for the adult but not fetal RORγt⁺ ILC development (37). Inhibition of Notch signaling by a γ-secretase inhibitor DAPT ablates the differentiation of bone marrow CLPs to mature RORγt⁺ ILCs in vitro, while fetal CLPs can be differentiated into RORγt⁺ ILCs normally in the presence of DAPT (37). RBP-Jκ is a molecule that associates with the soluble intracellular domain of Notch (NICD) and mediates the transcriptional output of Notch signaling (83). Selectively decreased NKR-LTi cells but less affected NKp46⁻RORγt⁺ ILCs was reported in RBP-Jκ-deficient mice (49), suggesting a differential requirement of Notch in programming certain RORγt⁺ ILC populations.

Ikaros is a member of the Krulppel family of zinc finger DNA-binding proteins, which regulates the development of various immune cell populations. Mice that express a dominant negative form of Ikaros lack mature T, B, and NK cells (84). Ikaros-null mutant mice have selective defects in the development of fetal and adult lymphoid systems. T and B lymphocytes are absent at the fetal stage, while postnatal thymocytes differentiate into aberrant CD4⁺ T cell with clonal expansion in the absence of functional Ikaros (85). Ikaros-null mice also lack lymph nodes, Peyer’s patches and lymphoid follicles in the gastrointestinal tract, suggesting a critical role of Ikaros in regulating the development of RORγt⁺ ILCs (e.g., fetal LTi cells). Considering the role of Ikaros in Notch signaling (86), the mechanism of the action of Ikaros in regulation of ILC development needs to be further investigated.

Other signaling pathways that regulate the development of RORγt⁺ ILCs

IL-7R is expressed by RORγt⁺ ILCs and IL-7/IL-7R signaling pathway is required for the formation of secondary lymphoid structures (87, 88). Enhanced IL-7 expression in the IL-7 transgenic mice results in accumulation of LTi cells and a five-fold increase of Peyer’s patch numbers (89). IL-7 also promotes de novo generation of adult RORγt⁺ ILCs from bone marrow cells and provides survival and proliferation signals to RORγt⁺ ILCs (46, 89). It has recently been shown that microflora are important for IL-7 production in the gut, which can stabilize the expression of RORγt in NKR-LTi cells (46). These finding suggest IL-7/IL-7R signaling is crucial for the development/maintenance of RORγt⁺ ILCs.

At the fetal stage, RORγt⁺ ILCs are important for the formation of secondary lymphoid organs including lymph nodes and Peyer’s patches. The TNF superfamily members lymphotoxin-α (LTα) and lymphotoxin-β (LTβ) expressed by RORγt⁺ ILCs are involved in this process (90). The heterotrimeric complex LTβ2LTα1 is the main ligand for LTβR, which is expressed on stromal organizer cells. The interaction between LTβ2LTα1 and LTβR can cause the increase of cell adhesion and chemokine molecules VCAM1, ICAM1, MADCAM1, CCL19, CCL21, and CXCL13, required for the recruitment of hematopoietic cells to the developing lymph nodes. Recently, it has been shown that LTβR signaling is required for the production of IL-22 by RORγt⁺ ILCs during Citrobacter Rodentium infection and mice genetically deficient in molecules of the lymphotoxin pathway died at the early stage of infection (91–93). The mechanism underlying the protective effect of lymphotoxin signaling involves an LT-driven positive feedback loop controlling IL-22 production by RORγt⁺ ILCs via LTβR signaling in DCs. Specifically, the expression of LTβR in dendritic cells is essential for the production of IL-23, a key cytokine to facilitate ILC function to enhance IL-22 production (93).

CXC chemokine receptor-5 (CXCR5), is needed for B cell homing to follicles in lymph nodes as well as in spleen. Cxcr5^−/− mice lack almost all Peyer’s patches and isolated lymphoid follicles (94). Adoptive transfer of fetal LTi cells restored the secondary lymphoid structures in Cxcr5^−/− mice (95). It has been shown that adoptive transfer of CD4⁺CD3⁻ cells purified from wildtype splenocytes that contain adult LTi cells has similar lymphoid structure-rescuing effect in Cxcr5^−/− mice, suggesting that compared to the fetal LTi cells, adult RORγt⁺ ILCs can exert similar functions to regulate lymphoid organogenesis (88). Mice infected with lymphocytic choriomeningitis virus (LCMV) have disrupted lymphoid organ structures due to damages caused by cytotoxic T cell responses (96). Restoration of the lymphoid structures requires the proliferation and accumulation of adult LTi cells to the damage sites during LCMV infection (96). This process is also dependent on LTβR signaling pathway, which mediates the interaction between LTi cells and stroma cells.

CCR6, a chemokine receptor, is expressed by LTi cells clustered in cryptopathes (97, 98) but not by NKR-LTi cells that are present outside cryptopatches (54, 98). Accordingly, Ccr6-defienct mice develop abnormal crytopatches and ILFs (97). Although the number/development of RORγt⁺ ILCs is unaltered in the absence of CCR6, IL-22 expression is significantly increased in Ccr6-deficient RORγt⁺ ILCs, indicating that CCL20-CCR6 signaling axis may also regulate ILC function (see below).

Effector functions of RORγt⁺ ILCs in the gut

One of the most important effector molecules produced by RORγt⁺ ILCs is IL-22. IL-22 receptor (IL-22R) is expressed exclusively by non-hematopoetic cells such as intestinal epithelial cells. Engagement of IL-22R signaling leads to secretion of anti-microbial peptides such as β-defensin, RegIIIβ, RegIIIγ, S100A8, and S100A9 that can control pathogenic bacterial infections (2, 3, 99). Il22-deficient mice succumbed to Citrobacter rodentium infection due to impaired production of RegIIIβ and RegIIIγ by gut epithelial cells. Accordingly, RegIIIγ-deficient mice have increased bacterial colonization at the intestinal epithelial surface, resulting in enhanced production of IFN-γ by CD4⁺ T cells and increased IgA level (100). IL-22 has recently been shown to have anti-parasite effect in the gut as well (101). IFN-γ is required for the control of intracellular apicomplexan parasite Eimeria falciformis (102). In Ifngr-deficient mice, neutralization of IL-22 alone can increase the parasite burden, suggesting an important anti-parasitic role of IL-22. Control of Candida albicans infection in the gastrointestinal tract has also been shown to require early activation of the IL-23-IL-22 axis in ILCs (103). Thus, IL-22 is an essential effector cytokine that promotes gut immunity.

In addition to its anti-microbial function, IL-22 is also critical for intestinal epithelial cell integrity mediated by mucin (e.g., Muc2). Muc2-deficient mice developed enhanced inflammation and cancer due to bacterial accumulation in epithelial cell layer and far down in the crypts, consistent with a notion that mucin can build a mucus barrier to prevent bacterial intrusion to the gut epithelia (104). The intestinal stem cells have been found to be a target of IL-22 (105). In a graft versus host disease (GVHD) model, recipient-derived IL-22 secreted by intestinal RORγt⁺ ILCs is crucial to protect the recipients from death, however no difference in inflammatory responses was observed between wildtype and Il22^−/− recipients (105). Immunohistochemistry staining suggested that IL-22R are expressed by intestinal stem cells at the crypt base between Paneth cells, suggesting a direct effect of IL-22 on the regeneration of epithelial cells (105). IL-22 can protect mice from DSS-induced colitis (106), in which IL-22 is upregulated. Consistently, IL-22 but not IL-6 acts through STAT3 to regulate the apoptosis and wound healing pathways in intestinal epithelial cells (107). Rorc-deficient mice, which lack RORγt⁺ ILCs, are susceptible to chronic DSS-induced colitis (108). This susceptibility is mainly due to increased number of tertiary lymphoid tissues (tLTs) that leads to inflammatory pathology upon epithelial cell damage. It is thought that in the absence of RORγt⁺ ILCs, compensatory elevation of tLTs is needed in Rorc-deficient mice to control the bacterial expansion after epithelial insult, while causing deregulated B cell responses and aggravated inflammation (108).

Besides their protective role in gut immunity against pathogen infections, certain RORγt⁺ ILCs can also exert pathogenic functions and are involved in inflammatory responses. Among them, Thy1^highSca1⁺RORγt⁺IL-23R⁺ innate lymphoid cells are enriched in the inflamed gut (109). Rag2^−/− mice infected with Helicobacter hepaticus or administered with anti-CD40 mAb develop colitis mediated by innate cells (109). Thy1^highSca1⁺RORγt⁺IL-23R⁺ cells secrete large amounts of IL-17 and IFN-γ in response to IL-23, which is a key cytokine produced by dendritic cells to drive the intestinal inflammation in this model (109). Depletion of Thy1⁺ cells in Rag^−/− mice completely abolished the innate colitis, indicating the pivotal role of ILCs in disease pathogenesis. Such IL-23-responsive ILCs are also present in the human intestines and play a pathogenic role in human Crohn’s disease (CD) (110). The Tbx21^−/−Rag2^−/− (TRUC) mice develop spontaneous colitis that is dependent on gut flora (111). The IL-23 responsive RORγt⁺ ILCs have recently been shown to be responsible for the disease pathogenesis in TRUC mice (60). T-bet is required for the colitogenic IFN-γ production by ILCs, however, without the expression of T-bet in TRUC mice, RORγt⁺ ILCs preferentially express pathogenic IL-17, thus leading to gut inflammation (60). IL-7R signaling is also crucial for the colitis pathogenesis in TRUC mice. T-bet suppresses the expression of IL-7R that is required for gut ILC homeostasis probably through directly binding to Il7ra gene locus. Consistent with the notion that IL-7R provides survival signal to RORγt⁺ ILCs (46, 89), IL-7R blockade dramatically reduced the severity of colitis in TRUC mice (60). Together, these data suggest that T-bet can play a protective role in controlling the pathogenicity of RORγt⁺ ILCs during certain innate immune cell-mediated colitis.

Ahr, an environmental sensor, functions in the immune system

Ahr is a ligand-dependent transcription factor that belongs to the basic helix-loop-helix transcription factor family members. The ligand for Ahr can come from environmental contaminants, therapeutics, naturally occurring chemicals, or small molecules from mammalian cells/tissues (112). Ahr ligands can be categorized into two major groups: xenobiotic and endogenous ligands. The most potent xenobiotic ligand for Ahr is 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), which has been used in various in vitro and in vivo studies (113, 114). Other xenobiotics known to be Ahr ligands include polychlorinated biphenyls (PCBs) and polycyclic aromatic hydrocarbons (PAH). Besides xenobiotics, endogenous Ahr ligands such as tryptophan derivative 6-formylindolo[3,2-b]carbazole (FICZ) and 2-(1′H-Indole-3′-carbonyl)-thiazole-4-carboxylic acid methyl ester (ITE) are also present (115, 116). Ahr protein is composed of a bHLH domain for DNA binding, the Per-Arnt-Sim (PAS) domain for ligand binding, and a Q-rich domain for co-activator recruitment and transactivation (117, 118). Without ligands, Ahr is retained in the cytoplasm and binds to the chaperone Hsp90, as well as co-chaperones ARA9 and p23. Binding to ligands leads to the conformational change of Ahr and its dissociation from chaperones. This event exposes the nuclear localization signal of Ahr. As a result, Ahr translocates to the nucleus and heterodimerizes with the aryl hydrocarbon receptor nuclear translocator (ARNT). The heterodimerization enhances the binding capacity of Ahr to the DNA elements called “dioxin responsive elements” (DRE) at the promoter and other regulatory loci of target genes. The most well-characterized target genes of Ahr include the cytochrome P450 family members Cyp1a1 and Cyp1a2 (119).

Although initial assessment of Ahr-deficient mice revealed few defects in the immune system (120), recent data suggest that Ahr plays an important role in both innate and adaptive immune responses. Ahr has been shown to regulate Th1/Th2 differentiation both in vitro and in vivo (121). Ahr has also been reported to function in Th17 cells, γδT cells, Tregs, Tr1 cells, and antigen presenting cells (e.g., dendritic cells and macrophages) (13, 122–127) (See accompanying reviews in the same issue). In the following sections, we will discuss the specific role of Ahr in controlling intestinal immune responses, focusing on RORγt⁺ ILCs.

Modulation of gut immune responses by Ahr ligands

Ahr ligands (both agonist and antagonist) have been used to examine the in vivo role of Ahr in the gut. Oral administration of Ahr agonist β-naphthoflavone (βNF) ameliorates DSS-induced colitis, the mechanism of which may be through inhibition of pro-inflammatory cytokine production such as TNF-α, IL-6, and IL-1β (128). It has been shown that FICZ, a UV photo-product of tryptophan, has a protective effect in various animal models of colitis (e.g., DSS-induced colitis, TNBS-induced colitis, and T cell transfer colitis) (115). The therapeutic effect of FICZ in TNBS-induced colitis is thought to be at least partially dependent on IL-22. Colonic cells produce relatively fewer amounts of IFN-γ, IL-17, and TNF-α in FICZ-treated mice, while blockade of IL-22 can ablate the downregulation of pro-inflammatory cytokines by FICZ (115). Amelioration of DSS-induced colitis may be mediated through epigenetic mechanisms (129). Demethylation of Foxp3 promoter and methylation of IL-17 promoter were observed upon Ahr ligand treatment (129). Of note, although ligand treatment offers exciting promise for future therapies, administration of ligands will inevitably activate Ahr in many tissues/cell types due to the broad expression pattern of Ahr. Furthermore, certain ligands may also exert indirect effects on the immune system through their metabolites (112). Thus, mechanistic insights await selective activation/deletion of Ahr in vivo in a cell/tissue-specific manner.

Ahr is well known to mediate the carcinogenicity of a family of environmental contaminants (i.e., xenobiotic ligands) (130). From an evolutionary perspective, Ahr exists as a sensor for endogenous ligands, rather than evolving directly to detect pollutants (e.g., dioxin). Consistent with this notion, we have recently shown in vivo efficacy of Ahr ligand (e.g., FICZ) administration to enhance the accumulation of intestinal RORγt⁺ ILCs (48). Recent data suggest that mice fed a synthetic diet that is free of phytochemicals (e.g., indole-3-carbinol (I3C) found in cruciferous vegetables) have a low level of in vivo Ahr activity as revealed by the decreased number of RORγt⁺ ILCs and γδT cells in the gut (13, 50), consistent with the data showing that tryptophan-derived phytochemicals under the influence of stomach acids can be converted to endogenous Ahr ligands (e.g., indolo[3,2-b]carbazole (ICZ) and 3,3-diindolylmethane (DIM)) (131). Intriguingly, another report suggests that ligands derived from the host cells may be important for the development of RORγt⁺ ILCs since deprivation of ligands from mouse diet had a minimal impact on the accumulation of RORγt⁺ ILCs in the gut (49). It remains to be determined whether these apparent discrepancies were due to experimental variations (e.g., mouse chow preparation and animal facility difference). Nevertheless, these data indicate the complexity of Ahr endogenous ligands in regulating RORγt⁺ ILCs and suggest that the precautions are needed in interpreting data involving Ahr ligand treatment.

Ahr controls the development and maintenance of RORγt⁺ ILC

We and others have recently reported that Ahr controls the development and function of RORγt⁺ ILCs (48–50). Unlike other transcription factors (e.g., Id2), Ahr seems to only affect the development of adult RORγt⁺ ILCs but not fetal RORγt⁺ ILCs (48–50). Consistently, Ahr-deficient mice show no apparent defects in secondary lymphoid organogenesis, such as the formation of lymph nodes and Peyer’s patches. In newborn mice, no difference in the accumulation of intestinal RORγt⁺ ILCs was found between Ahr^−/− mice and their littermate wildtype controls. The deficiency in RORγt⁺ ILCs only appears at the weaning age of Ahr^−/− mice, revealed by decreased proportion as well as total number of RORγt⁺ ILCs in the gut (48, 50). The defect in RORγt⁺ ILCs leads to marked reduction of cryptopatches and ILFs in both the small and large intestines of Ahr^−/− mice (50). Furthermore, mice with Ahr deletion specifically in RORγt⁺ ILCs showed similar phenotypes to Ahr-null mice, suggesting a cell-intrinsic role of Ahr in regulating RORγt⁺ ILCs ((50), and Qiu and Zhou, unpublished).

Although the precise mechanisms by which Ahr programs RORγt⁺ ILCs remain elusive, it has been shown that Ahr can modulate the development/maintenance of RORγt⁺ ILCs in multiple regulatory modes. We have shown that Ahr is important for the survival of RORγt⁺ ILCs (48). In agreement with the role of IL-7/IL-7R signaling pathway in the survival of RORγt⁺ ILCs (46, 89), IL-7R expression in RORγt⁺ ILCs is decreased, and IL-7 expression in small intestinal epithelial cells is also reduced in Ahr-deficient mice. Consequently, decreased expression of anti-apoptotic genes and enhanced apoptosis were observed in Ahr-deficient RORγt⁺ ILCs. Interestingly, Kiss et al showed reduced percentage of Ki67⁺ cells in RORγt⁺ ILCs from two-week old mice, suggesting that decreased ability of proliferation may also contribute to the defective expansion of NKR-LTi cells in the absence of Ahr at least at the early stage of mouse development. Ahr also regulates the expression of Kit, which can control RORγt⁺ ILC development (50). Consistent with this notion, Kit^Wv/Wv mice, in which Kit is functionally impaired, have reduced number of RORγt⁺ ILCs. Kit expression in mice decreases after birth but is further upregulated after weaning (50). Consistent with a requirement of Ahr for the development/maintenance of RORγt⁺ ILCs after the weaning age, Ahr is essential for the upregulation of Kit (50). ChIP experiment further showed that Ahr binds to the DREs located at the promoter of Kit, suggesting direct transcriptional regulation of Kit expression by Ahr. Ahr may also control the development of RORγt⁺ ILCs by regulating Notch expression (49). However, this effect seems to be restricted to NKR-LTi cells rather than NKp46⁻RORγt⁺ ILCs through an as yet unidentified mechanism.

Cooperative action of Ahr and RORγt in regulation of RORγt⁺ ILC function

Besides regulating the development/maintenance of RORγt⁺ ILCs, Ahr also controls the function of RORγt⁺ ILCs (i.e., production of IL-22). Ahr-deficient RORγt⁺ ILCs express less IL-22 on a per cell basis. Using ChIP assay, we recently demonstrated that binding of Ahr to the Cyp1a1 (a known Ahr direct target gene) locus is independent of RORγt. In contrast, without RORγt, Ahr is unable to bind to the Il22 locus, consistent with multiple DREs that are clustered with ROR responsive elements (ROREs) at the Il22 locus. With the co-expression of RORγt, Ahr binds to the Il22 locus efficiently. Using co-immunoprecipitation (co-IP) assay, we further detected the interaction between Ahr and RORγt (48), suggesting that Ahr may cooperate with RORγt to promote the Il22 transcription. Indeed, co-expression of Ahr and RORγt synergistically upregulates Il22. It remains to be determined whether Ahr indirectly binds to ROREs through interaction with RORγt to enhance the Il22 transcription. Alternatively, facilitated by RORγt, Ahr may acquire enhanced DNA binding activity, whereby directly binding to the DREs at the Il22 locus to induce transcription. It is worthy to note that although binding of Ahr at the Il22 locus is facilitated by RORγt, we have consistently observed reduction of Ahr binding at the Cyp1a1 locus when RORγt is co-expressed ((48) and Qiu and Zhou, unpublished). It is tempting to speculate that at least two types of Ahr target genes are present in ILCs, such as RORγt-dependent genes (e.g., IL-22) and RORγt-independent genes (e.g., Cyp1a1). RORγt may function as a pioneering factor to set a permissive chromatin structure at certain genes (e.g., Il22), and subsequent recruitment of Ahr by RORγt via protein-protein interaction induces gene transcription.

In addition to the direct mechanism of the action of Ahr in regulating Il22 gene transcription, the decreased IL-22 production by RORγt⁺ ILCs in Ahr-deficient mice could also be attributed to reduced responsiveness to IL-23 (48). Lower Il23r expression was consistently detected in Ahr-deficient RORγt⁺ ILCs (48). Since Il23r is a known direct target gene of RORγt (132), whether Il23r expression in RORγt⁺ ILCs is also regulated by cooperative action of Ahr and RORγt remains to be further determined.

Ahr and RORγt⁺ ILCs in human colitis

Crohn’s disease (CD) and ulcerative colitis (UC) are the two major types of human inflammatory bowel disease (IBD). These two types of IBD have distinct disease pathogenesis in that CD is thought to be mediated by Th1 cells, while UC is Th2 cell-dominant (133). Importantly, Th17 cells are considered to be involved in the pathogenesis of both diseases. Expression of Ahr has been examined in colitis patients (115). Compared to healthy controls, Ahr mRNA is decreased in Crohn’s disease patients but not in patients with UC (115). Ahr expression was found to be especially downregulated in the inflamed mucosa. Flow cytometry analysis demonstrated at the protein level that Ahr expression was decreased in CD3⁺, CD4⁺ CD56⁺, and CD25⁺ lamina proprial mononuclear cells (115). These findings indicate that Ahr is likely to play an important immune regulatory role in IBD. In contrast, IL-22 has been shown to be upregulated in both CD and UC patients (134, 135). Moreover, the level of IL-22 correlates with the IBD severity (136). Specifically, increased IL-22, but not IL-17 and IFN-γ was observed in CD patients (137). Despite the protective function of IL-22 in mouse models of colitis, IL-22 can stimulate the secretion of pro-inflammatory cytokines by human colonic sub-epithelial myofibroblasts in human IBD patients (135). Thus, the role of IL-22 in human IBD remains to be further explored.

RORγt⁺ ILCs are generally considered to be protective in gut immunity, presumably through production of IL-22 to promote tissue repair and/or to stimulate anti-microbial function of epithelial cells (2, 3, 99). Recent data have suggested that depletion of IL-22-producing RORγt⁺ ILCs results in the selective dissemination and survival of bacteria Alcaligenes spp. in peripheral tissues of mice (138). Considering Alcaligenes-specific systemic immune responses associated with Crohn’s disease and progressive hepatitis C virus infection in patients, identification of abundant presence of IL-22-producing RORγt⁺ ILCs in the intestine and gut-associated lymphoid tissues (GALTs) of healthy humans suggests an intriguing link between RORγt⁺ ILCs and pathogen control in humans (138). In humans, it has been shown that CD56⁻ ILCs express IL-17 and IFN-γ, whereas IL-22 is mainly produced by the CD56⁺ ILC compartment. Interestingly, selective increase in CD127⁺CD56⁻ ILCs in the inflamed intestine was observed in human colitis patients (110). Given the strong clinical relevance of RORγt⁺ ILCs, it is important to identify and characterize factor(s) that differentially regulate tissue-protective IL-22–producing ILC populations and pathogenic IFN-γ/IL-17–producing ILC populations in intestinal inflammation.

Future directions

The innate and adaptive immune systems in the gut play a fundamental role in eliminating pathogens while maintaining an immune-balanced environment during infections and autoimmune diseases. ILCs, T cells and antigen presenting cells may maintain an active crosstalk with each other and also interact with the gut flora to become educated and functional. Ahr is widely expressed by different types of immune cells in the gut and the intestinal environment is naturally enriched for Ahr ligands originated from food or generated by indigenous microflora, highlighting the gut as an important anatomical site for the physiological action of Ahr.

How Ahr expression and/or function are regulated in RORγt⁺ ILCs remains unclear. It is worth mentioning that although Ahr-deficient mice have marked reduction of RORγt⁺ ILCs, a significant number of RORγt⁺ ILCs is still present in the absence of Ahr (48). In the future, it is important to identify and characterize factors/signaling pathways that may bypass the requirement of Ahr in RORγt⁺ ILC development and/or maintenance. The cell-intrinsic action of Ahr requires further investigation in a tissue/cell type-specific manner. A dialogue among ILCs, T cells, and APCs needs to be further decoded. The signaling pathways that are regulated by Ahr in RORγt⁺ ILCs remain to be determined in the future.

Although microflora are apparently not required for the development of RORγt⁺ ILCs, it is unknown whether specific microflora population(s) can contribute to RORγt⁺ ILC programming in the gut. The impact of RORγt⁺ ILCs on shaping commensal gut flora is also an important research area for future study. The potential endogenous ligands for Ahr remain to be investigated. It is of special interest to understand the role of microflora in generating/modifying Ahr ligands. The intricate effect of various Ahr ligands either from exogenous origin or from diet may affect the strength of Ahr activation, causing differential effects in future clinical applications.