Group 3 innate lymphoid cells (ILC3s) have dual roles in intestinal health, acting in both protective and pathogenic capacities, and importantly, modulations in this population of innate lymphoid cells have been implicated in inflammatory bowel disease. Further, subpopulations of ILC3s have been described as serving specific functions in maintaining homeostasis or responding to infection, and aberrant activation of one or more of these subpopulations could exacerbate inflammatory bowel disease. However, the signals that enforce the protective and pathogenic features of ILC3s are not fully elucidated. In this article, we show that IL-21, a cytokine primarily produced by CD4 T cells, acts on a subpopulation of intestinal ILC3s to promote a protective phenotype. IL-21 signaling does not affect the MHC class II–expressing ILC3 subset but promotes ILC3s that express Tbet and are poised to produce IL-22. Consistent with a protective phenotype, IL-21 deficiency dampens cytokine-induced IL-17A production. We show that exacerbated colitis develops in mice lacking the IL-21 receptor, in agreement with a protective role for IL-21 signaling on ILC3s. To our knowledge, these data reveal a novel role for IL-21 in shaping innate lymphoid cell responses in the intestine and provide one mechanism by which effector CD4 T cells can influence innate immunity.
Innate lymphoid cells (ILCs) are a diverse group of innate immune cells that lack an Ag receptor and are especially populous in the body’s barrier organs, such as the lungs, skin, and intestine (1–4). Because these cells lack an Ag receptor, ILCs are triggered to respond by stimulation with cytokines. ILCs can be divided into groups that parallel the effector CD4 T cell subsets based on the expression of transcription factors and production of cytokines (5, 6). Group 1 ILCs (ILC1s), which include NK cells as well as a non-NK ILC1 subset, express Tbet and/or Eomes and produce IFN-γ and/or cytotoxic granules. Group 2 ILCs (ILC2s) express GATA-3 and produce IL-13 and IL-5 in response to IL-25 and IL-33. Finally, group 3 ILCs (ILC3s) express RORγt and produce IL-22, IL-17A, GM-CSF, and/or IFN-γ in response to IL-23 and IL-1β. Importantly, functionally distinct subpopulations of ILC3s have been identified using single-cell sequencing, and it is yet undetermined how these discrete populations are differentially regulated to effect both protective and pathogenic immune responses (7).
The common γ-chain (γc-chain) cytokine family consists of six members, IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21, and is so named because all use the γc-chain receptor, CD132, in combination with specific receptor chains for signaling (8). This family of cytokines is known to impact multiple cell types within the immune system, including helper CD4 T cells and ILCs (9). For example, classic NK cells are reliant on IL-15 for their development, and specific ILC subsets require IL-7 for homeostasis (5). All ILC populations express the γc-chain, and, notably, these cells are absent in γc-deficient mice (5, 10). This indicates an important role for this family of cytokines in ILC biology; however, it is unclear if all γc-chain cytokines influence ILC development and function.
Given the similarity of ILC3s to Th17 CD4 T cells, we posited that these subsets may share additional cytokine networks besides IL-23, IL-17A, and IL-22. IL-21 acts to solidify the Th17 lineage and drives their function via upregulation of RORγt, Ahr, IL-23R, IL-17A, and IL-22 (11–14); however, the influence of IL-21 signaling on ILC3s has not yet been investigated. To investigate the effect of IL-21 on the regulation of ILC3 phenotype and function, we examined these cells in mice lacking either IL-21 or the IL-21 receptor. In the absence of IL-21 signaling, the bulk ILC3 population in the intestine is not compromised, but the proportion of plastic ILC3s is reduced, with fewer NKp46+ ILC3s and decreased Tbet expression. Additionally, IL-21 is necessary for optimal IL-22 production by ILC3s, and this is consistent with exacerbated colitis in mice lacking the IL-21 receptor on ILCs. Together, these data highlight an important role for IL-21 in calibrating ILC3 subsets in the intestine and emphasize the complex network on signals that regulate ILC3 homeostasis.
Materials and Methods
Wild-type (WT) C57BL/6 and B6.129S7-Rag1tm1Mom/J (Rag1−/−) mice were purchased from The Jackson Laboratory. IL-21R−/− and IL-21–deficient (IL-21−/−) mice were generated as described (15). Rag1−/−IL-21R−/− and Rag1−/−IL-21R+/− mice were generated at the University of Alabama at Birmingham (UAB). Mice were used at 6–12 wk of age, and all experiments were performed with sex- and age-matched groups. For colitis experiments, littermate controls were used as recipient mice. All animals were bred and maintained according to Institutional Animal Care and Use Committee regulations.
Isolation of CD4 T cells and B cells
Freshly explanted spleens were mechanically disrupted into single-cell suspensions using mesh screens. Erythrocytes were subsequently removed by lysis using 0.83% (w/v) NH4Cl. CD4 T cells were isolated using Dynabeads FlowComp Mouse CD4 kit, and B cells were isolated using Dynabeads Mouse CD43 (Untouched B Cells) kit (Invitrogen), according to the manufacturer’s instructions.
Isolation of lamina propria cells
Lamina propria lymphocytes were isolated using the Lamina Propria Dissociation Kit (Miltenyi Biotec), according to the manufacturer’s protocol. Cells were filtered through 100-μm cell strainers in PBS with 0.5% FBS. Cells were resuspended in 5 ml of 30% Percoll (GE Healthcare) and centrifuged at 25°C at 800 relative centrifugal force for 20 min. Cells were collected at the bottom of the Percoll gradient, washed, filtered through 70-μm filters, and resuspended in complete media (RPMI 1640 with 10% FCS, 2 mM l-glutamine, 100 IU/ml penicillin, 100 μg/ml streptomycin, 1× nonessential amino acids, 1 μM sodium pyruvate, and 2.5 μM 2-ME).
Except where otherwise indicated, ILC3s were identified with the following markers: Lineage− (CD4, TCRβ, TCRγδ, CD8α, CD8β, CD19, CD11b, CD11c, DX5, Gr-1, Ter119, NK1.1) CD45+ Thy1.2+ RORγt+ GATA-3lo.
Abs and flow cytometry
Isolated cells were stained in PBS for 20–30 min on ice. Fc-receptor blockade was performed with anti-CD16/32 mAbs prior to cell surface staining. For the detection of transcription factors and cytokines, cells were stimulated for 4 h with 10 ng/ml rIL-23 (eBioscience) and 10 ng/ml rIL-1β (Peprotech). Brefeldin A (10 μg/ml; BD Biosciences) was added for the last 3 h of stimulation. Cells were fixed and permeabilized for intracellular staining using the Foxp3 Staining Kit (eBioscience), according to the manufacturer’s protocol. Dead cells were excluded from analysis using LIVE/DEAD Fixable Aqua Dead cell stain (Invitrogen). Doublets were excluded prior to analysis. Samples were acquired using an LSRII flow cytometer (BD Biosciences), and data were analyzed using FlowJo software (Tree Star).
Biotinylated mAbs from BioLegend were used as lineage markers: CD4 (GK1.5), TCRβ (H57-597), TCRγδ (UC7-13D5; eBioscience), CD8α (53-6.7), CD8β (YTS156.7.7), CD19 (6D5), CD11b (M1/70), CD11c (N418), DX5, Gr-1 (RB6-8C5), Ter119, and NK1.1 (PK136). Streptavidin-APC-eFluor 780 (eBioscience) was used to stain lineage-positive cells.
Additional mAbs used are as follows: CD45 (30-F11; eBioscience), Thy1.2 (53-2.1; eBioscience), KLRG1 (2F1; eBioscience), IL-7Rα (A7R34; eBioscience), NKp46 (29A1.4; eBioscience), MHC class II (MHC II) (M5/114.15.2; eBioscience), Tbet (eBio4B10; eBioscience), RORγt (B2D; eBioscience), GATA-3 (BD L50-823), IL-17A (eBio17B7; eBioscience), and IL-22 (Poly5164; BioLegend).
For FACS sorting, pooled lamina propria cells were stained and sorted using a FACSAria II cell sorter (BD Biosciences) in the UAB Comprehensive Arthritis, Musculoskeletal, Bone, and Autoimmunity Center flow cytometry facility. ILC2s were identified as lineage−CD45+Thy1.2intKLRG1+; ILC3s were identified as lineage−CD45+Thy1.2hiKLRG1−IL-7Rα+.
Phosphorylated STAT3 staining
For the detection of p-STAT3, cells were stained for lineage markers in PBS for 30 min, then stimulated for 30 min with 10 ng/ml rIL-21 (eBioscience). Surface marker and p-STAT3 staining was performed following fixation with 4% paraformaldehyde and permeabilization with 100% methanol. p-STAT3 staining was achieved with Phospho-Stat3 (Y705) Ab (Cell Signaling).
ILC3s were FACS sorted from small intestine (SI) lamina propria, and 1 × 104 cells per well were plated in 100 μl of complete media supplemented with 10 ng/ml rIL-7 (BioLegend), 10 ng/ml stem cell factor (BioLegend), and ± 50 ng/ml rIL-21 (eBioscience). Cells were cultured overnight, and RNA was extracted for real-time PCR, as described below.
RNA isolation, cDNA synthesis, and real-time PCR
RNA was extracted from cells using the RNeasy Micro Plus Kit (Qiagen), and cDNA was synthesized using the iScript cDNA Synthesis Kit (Bio-Rad), according to the manufacturer’s instructions. Real-time PCR was performed using the iQ SYBR-Green Supermix (Bio-Rad) and the primer pairs listed below, with β2-microglobulin as the housekeeping gene. Reactions were run in triplicate on the iQ5 Multicolor Real-Time PCR Detection System (Bio-Rad). Relative gene expression was calculated according to the ΔΔ threshold cycle, and data were normalized to WT splenic CD4 T cells.
The following primer sequences were used for real-time PCR analyses: B2m forward (FWD): 5′-CCTGCAGAGTTAAGCATGCCAG-3′, B2m reverse (REV): 5′-TGCTTGATCACATGTCTCGATCC-3′; Rorγt FWD: 5′-CCGCTGAGAGGGCTTCAC-3′, Rorγt REV: 5′-TGCAGGAGTAGGCCACATTACA-3′; Gata3 FWD: 5′-CAGAACCGGCCCCTTATCA-3′, Gata3 REV: 5′-CATTAGCGTTCCTCCTCCAGA-3′; Il21r FWD: 5′-GTGACCCCGTCATCTTTCAGA-3′, Il21r REV: 5′-CCGCTGTGCTCCCTGTACA-3′; and Tbx21 FWD: 5′-ACCAGAGCGGCAAGTGGG-3′, Tbx21 REV: 5′-TGGACATATAAGCGGTTCCCTG-3′.
Lymphocytes were isolated from spleens and lymph nodes of WT mice and stained for CD4, CD25, and CD45RB. CD4+CD45RBhiCD25− cells were FACS sorted, and 5 × 105 cells per mouse were injected i.p. Mice were monitored for weight loss and sacrificed after 4 wk. Large intestines (LI) were excised, opened longitudinally, and fixed in formalin. Tissue was embedded in paraffin, sectioned, and stained with H&E by the UAB Comparative Histology Laboratory. Slides were scored in a blinded fashion by a veterinary pathologist for signs of disease.
Statistical significance was calculated as indicated in figure legends using Prism software (GraphPad). All t tests performed are two tailed, and p values <0.05 were considered significant, unless specifically indicated otherwise in the text.
IL-21 influences the composition of the intestinal ILC3 population by regulating Tbet levels
To determine how IL-21 signaling impacts ILC3s, we compared several characteristics of ILC3s from the lamina propria of WT and IL-21−/− mice. Throughout this study, ILC3s are identified as lineage− CD45+ Thy1.2+ RORγt+ GATA-3lo cells (Supplemental Fig. 1A; see 2Materials and Methods for the description of the lineage panel). The frequency of Tbet+ ILC3s is decreased in both the SI and LI of IL-21−/− mice, with the per-cell expression level of Tbet also reduced in the SI (Fig. 1A, 1B), suggesting that IL-21 promotes Tbet expression. To corroborate this finding, we sorted ILC3s from the SI lamina propria of WT mice, confirmed their identity as RORγt+ ILC3s with quantitative real-time PCR (Supplemental Fig. 1B, 1C), and cultured them overnight in the presence or absence of rIL-21. ILC3s cultured in the presence of IL-21 express higher levels of Tbx21 mRNA, which encodes Tbet, than ILC3s cultured in the absence of IL-21 (Fig. 1C). This confirms that IL-21 induces Tbet expression in ILC3s. Because Tbet has been shown to control the expression of NKp46 on ILC3s (16, 17), we next investigated whether this subset of ILC3s is impacted by IL-21 deficiency. Importantly, no differences in the total number of ILC3s were observed (Fig. 1E). In the absence of IL-21 signaling, the frequency of Tbet+NKp46− ILC3s is decreased, with no impact on the frequency of Tbet+NKp46+ ILC3s. Unsurprisingly, we observed a reciprocal increase in Tbet−NKp46− ILC3s in the absence of IL-21 (Fig. 1D, 1F), indicating that IL-21 regulates ILC3 plasticity via Tbet expression but not NKp46 expression.
A subset of ILC3s that are Tbet−NKp46− has been shown to express MHC II and is remarkable for its ability to dampen CD4 T cell–driven microbiota-dependent intestinal inflammation (18). Importantly, TCR–MHC II binding is one mechanism by which CD4 T cells engage in crosstalk with ILCs (18–20); therefore, we investigated whether CD4 T cell–derived IL-21 modulates MHC II expression. Despite our finding of an increased frequency of the Tbet−NKp46− subset of ILC3s in the absence of IL-21, we did not observe an appreciable difference in the frequency of MHC II+ ILC3s in IL-21−/− mice, implying that MHC II expression marks a unique subset of cells (Fig. 2). This is consistent with a recent publication showing that MHC II+ ILC3s are a distinct subset, based on transcriptional and functional profiling (7). We do observe a slight decrease in the level of MHC II in the SI and a trend in the LI indicating that ILC3s from IL-21−/− mice express lower levels of MHC II on a per-cell basis. However, it is unlikely that this slight change in MHC II expression would affect the ability of ILC3s to successfully present Ag to CD4 T cells. This demonstrates that IL-21 does not impact the development of MHC II+ ILC3s.
IL-21 promotes a protective ILC3 phenotype by preferentially augmenting an IL-22–producing ILC3 subset
A salient feature of ILC3s is the ability to produce IL-22, which can act on intestinal epithelial cells to induce production of antimicrobial peptides, such as RegIIIβ and RegIIIγ, to protect against bacterial infections and commensal-induced inflammation (4). In addition, IL-22 has been linked to the promotion of epithelial survival and repair in the intestine as well as amelioration of intestinal inflammation and induction of healing in certain models of inflammatory bowel disease (IBD) (1, 21–23). Further, IL-22–producing ILC3s are shown to be a distinct subset from those that express MHC II (7). Because we saw that the NKp46−Tbet− population was affected by IL-21 signaling, but the MHC II+ ILC3s showed no difference in population size or MHC II expression, we hypothesized that IL-21 is specifically influencing the IL-22–producing subset of ILC3s. IL-22–producing ILC3s are capable of producing other cytokines, such as GM-CSF (24, 25), IFN-γ (26), and IL-17A (27–30), but the regulation of this heterogeneity has yet to be established. To ascertain the role of IL-21 signaling on ILC3s, cells were isolated from the LI lamina propria of WT and IL-21−/− mice, cultured with media alone or IL-23 and IL-1β, and ILC3s were evaluated for cytokine production. Following stimulation with IL-23/IL-1β, the frequency of IL-22+IL-17A− ILC3s, which is the most abundant fraction during homeostasis, is reduced in IL-21−/− mice (Fig. 3A, 3B). In contrast, the small fraction of IL-22−IL-17A+ ILC3s is increased in the absence of IL-21, following cytokine stimulation (Fig. 3A, 3B). Finally, not only is the frequency of IL-22+IL-17A− ILC3s decreased in the absence of IL-21, but we also note a reduction in the amount of IL-22 produced on a per-cell basis following IL-23/IL-1β stimulation (Fig. 3C, 3D). This indicates that IL-21 regulates ILC3 production of IL-22 and IL-17A.
ILC3s express the receptor for and respond to IL-21
Our data show that distinct features of ILC3s are modulated by the presence of IL-21, leading us to hypothesize that IL-21 acts directly on a specific subset of ILC3s. IL-21 signals via its heterodimeric receptor, consisting of the IL-21R chain and γc-chain (8). Because it is known that all ILCs express the γc-chain, we examined the expression of IL-21R on subsets of ILCs. ILC2s and ILC3s were sorted from the SI and LI lamina propria of WT mice (Supplemental Fig. 1B). Subset identity was confirmed via real-time PCR, with ILC2s expressing low levels of RORγt and high levels of GATA-3 and ILC3s expressing high levels of RORγt and low levels of GATA-3 (Supplemental Fig. 1C). Strikingly, ILC3s, but not ILC2s, express Il21r mRNA at levels similar to those of B cells (Fig. 4A), which have been shown to harbor elevated levels of the IL-21R (31). These data indicate that ILC3s are capable of responding to IL-21.
It has been shown that STAT3 signaling is critical for the production of IL-22 by ILC3s and, thus, for innate defense against Citrobacter rodentium infection (32). Because IL-21 has been shown to be a potent inducer of STAT3 activation (14), we tested ILC phosphorylated STAT3 in response to IL-21. Cells were isolated from the mesenteric lymph nodes (mLNs) and LI lamina propria of WT mice, cultured in the presence or absence of IL-21 for 30 min, and then ILCs were evaluated for expression of p-STAT3. Stimulation with IL-21 induced a 2-fold and a 4.5-fold increase in the frequency of p-STAT3+ cells in mLNs and LI, respectively, compared with cells cultured with media alone (Fig. 4B, 4C). This is also reflected in the total number of p-STAT3+ ILC3s. This effect is specific to IL-21 signaling, as p-STAT3 is not induced by IL-21 in ILCs that lack IL-21R (data not shown). Collectively, these data indicate that IL-21 can induce STAT3 activation in ILCs via IL-21R and are consistent with our previous observations that discrete subsets of ILC3s are modulated by IL-21.
Loss of IL-21R by innate immune cells is associated with exacerbated CD4 T cell–mediated colitis
IBD is characterized by chronic intestinal inflammation, the causes of which are still largely unknown. There is much evidence that the IL-23/IL-17A axis plays a role in driving intestinal inflammation, as genome-wide association studies have identified polymorphisms in IL-23R and STAT3 that are associated with IBD (33–35). Recent studies have revealed that ILCs are innate sources of IL-17A that drive intestinal inflammation (27, 29). In contrast, several groups have shown that IL-22–producing ILC3s are protective against CD4 T cell–driven intestinal inflammation (23, 36, 37). Because we found that ILC3s have enhanced IL-22 yet reduced IL-17A production in the presence of CD4 T cell–derived IL-21, we posited that IL-21 signaling on ILCs would be protective in the CD45RBhi transfer model of colitis. To test this, we transferred naive CD4 T cells (CD4+ CD45RBhi CD25−) from WT mice into Rag1−/−IL-21R−/− recipients or Rag1−/−IL-21R+/− littermate controls and evaluated disease severity by weight loss and intestinal histology score. Recipient mice deficient in IL-21R expression lost weight more rapidly than their littermate controls (Fig. 5A). These mice also had more severe intestinal pathologic disease score in both the cecum and colon, as evidenced by epithelial hyperplasia, goblet cell loss, crypt exudate, and inflammatory cell accumulation (Fig. 5B, 5C). These data support a protective role for IL-21 signaling in nonadaptive cells, including ILC3s, in CD4 T cell–mediated colitis.
It is critical to understand the cytokine networks active in regulating ILC3 subsets to fully elucidate their function and role in the complex intestinal niche. To our knowledge, in this study, we have revealed IL-21 as a new player in ILC3 regulation; IL-21 signaling on ILC3s results in reduced IL-17A production and augmented IL-22 and Tbet expression, culminating in a protective variety of ILCs that can dampen pathogenesis. Further, our data suggest a novel mechanism by which IL-21 production by activated CD4 T cells directs the phenotype and function of ILCs, implying a unique function for this innate cell subset beyond the recruitment and activation of the adaptive immune response.
IL-21 is associated with chronic inflammation in the intestine, and single-nucleotide polymorphisms in the IL-2/IL-21 gene locus are linked with the development of IBD (38), yet how IL-21 contributes to disease is still unknown. IL-21 is known to impact numerous cell types, including many that reside in the intestine (14, 39), and it has been shown that during Th17 differentiation, IL-21 supports the upregulation of IL-23R and promotes IL-17A expression (13). In this study, we show that IL-21 actually decreases IL-17A and augments IL-22 production by ILC3s, illustrating an additional mechanism by which IL-21 may contribute to intestinal homeostasis and disease. Moreover, these data reveal both parallel and divergent regulatory networks that exist between ILC3s and Th17 cells. Interestingly, we find that stimulation of ILC3s with IL-21 does not directly enhance IL-22 production (data not shown); however, ILC3s from mice with intact IL-21 respond to stimulation with IL-23 and IL-1 with higher IL-22 production. Therefore, IL-21 possibly gives a priming signal that enhances the responsiveness of ILC3s, perhaps through upregulation of IL-23R or IL-1R1. Further investigation will be necessary to delineate the effector molecules influencing IL-21–mediated ILC3 activation.
The influence of tissue microenvironments, especially the commensal microbes of the intestine, on the phenotype and function of ILC populations is becoming increasingly evident (40). Recent transcriptome analysis reported little to no IL-21R expression by ILC3s isolated from the SI (41), which is in stark contrast to our data demonstrating striking IL-21R expression by ILC3s in the intestine. Because commensal bacteria and their metabolic products play a role in many aspects of ILC development, function, and phenotype (40, 42, 43), it is possible that this environmental factor influences the proportion of specific ILC3 subsets. Importantly, single-cell analysis of ILCs was recently used to track transcriptome responses to changes in microbial colonization and demonstrated that the abundance and composition of the microbiota has an impact on the diversity and phenotype of intestinal ILC3s (7). It will be important to evaluate the influence of the microbiota on the ability of the ILC population to detect proinflammatory signaling molecules such as IL-21, as this could influence treatment plans for ILC-mediated autoinflammatory conditions.
In this study, we show that IL-21 can regulate the phenotype and function of the ILC3 population. Interestingly, recent reports reveal that IL-21R is expressed by ILC1s in the intestine, salivary glands, and, to a lesser extent, in the spleen and liver (7, 41), yet the effect of IL-21 on this ILC subset is unknown. ILC1s have been found to be associated with the inflamed lesions in IBD patients, and these cells can produce the proinflammatory cytokines IFN-γ and TNF-α (44, 45). Future work investigating whether IL-21 is required for the development, maintenance, or function of ILC1s and how this affects the progression of inflammatory disease will enhance understanding of the overall impact of IL-21 on ILC responses.
Finally, our data highlight the heterogeneity and diversity of the ILC3 population. In line with recent evidence that there is functional compartmentalization under the ILC3 umbrella (7, 46), we show that cytokine signaling specifically directs the activity of IL-22–producing ILC3s without influencing MHC II+ ILC3s. Further, IL-21 signaling on this subset of ILC3s is necessary to protect against T cell–mediated intestinal inflammation. The demonstration that IL-21 acts on this small but crucial subset of ILCs may be of critical importance to understanding the mechanisms by which chronic intestinal inflammation is mediated and sustained.
We thank members of the Harrington and Zajac laboratories for comments and technical assistance; Y. Belkaid and C. Wilhem for helpful discussions; Enid Keyser of the UAB Comprehensive Arthritis, Musculoskeletal, Bone, and Autoimmunity Center Cytometry Facility for cell sorting (National Institutes of Health P30 AR048311 and P30 AI27667); and Trenton Schoeb for pathologic scoring of slides.
This work was supported by National Institutes of Health Grants R01 DK084082 and R01 AI113007 (to L.E.H.), R01 AI049360 and U19 AI109962 (to A.J.Z.), TL1 TR000167 (to C.H.P.), and T32 AI07051 (to S.J.D.) and was partially supported by T32GM008361 (to C.H.P.).
The online version of this article contains supplemental material.
Abbreviations used in this article:
inflammatory bowel disease
innate lymphoid cell
group 1 ILC
group 2 ILC
group 3 ILC
- MHC II
MHC class II
mesenteric lymph node
University of Alabama at Birmingham
The authors have no financial conflicts of interest.