Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation

Article

Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation Graphical Abstract

Authors Shuo Wang, Pengyan Xia, Yi Chen, ..., Zhinan Yin, Zhiheng Xu, Zusen Fan

Correspondence [email protected] (S.W.), [email protected] (Z.F.)

In Brief A subpopulation of innate lymphoid cells, called ILCregs, are found to have a regulatory role in intestinal homeostasis and innate immune defenses akin to Treg cells.

Highlights d

ILCregs exist in mouse and human intestines

d

ILCregs contribute to the resolution of innate intestinal inflammation

d

ILCregs suppress the activation of ILC1s and ILC3s via secretion of IL-10

d

Autocrine TGF-b1 is required for the expansion of ILCregs during inflammation

Wang et al., 2017, Cell 171, 1–16 September 21, 2017 ª 2017 Elsevier Inc. http://dx.doi.org/10.1016/j.cell.2017.07.027

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Article Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation Shuo Wang,1,7,* Pengyan Xia,1,7 Yi Chen,3 Yuan Qu,1 Zhen Xiong,1 Buqing Ye,1 Ying Du,1 Yong Tian,2,4 Zhinan Yin,5 Zhiheng Xu,6 and Zusen Fan1,2,8,* 1CAS Key Laboratory of Infection and Immunity, CAS Center for Excellence in Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 100101 Beijing, China 2University of Chinese Academy of Sciences, 100049 Beijing, China 3Department of Gastrointestinal Surgery, State Key Laboratory of Biotherapy, West China Hospital, Sichuan University, 610041 Chengdu, China 4Key Laboratory of RNA Biology of CAS, Institute of Biophysics, Chinese Academy of Sciences, 100101 Beijing, China 5The First Affiliated Hospital, Biomedical Translation Research Institute and Guangdong Province Key Laboratory of Molecular Immunology and Antibody Engineering, Jinan University, 510632 Guangzhou, China 6State Key Laboratory of Molecular Developmental Biology, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, 100101 Beijing, China 7These authors contributed equally 8Lead Contact *Correspondence: [email protected] (S.W.), [email protected] (Z.F.) http://dx.doi.org/10.1016/j.cell.2017.07.027

SUMMARY

An emerging family of innate lymphoid cells (termed ILCs) has an essential role in the initiation and regulation of inflammation. However, it is still unclear how ILCs are regulated in the duration of intestinal inflammation. Here, we identify a regulatory subpopulation of ILCs (called ILCregs) that exists in the gut and harbors a unique gene identity that is distinct from that of ILCs or regulatory T cells (Tregs). During inflammatory stimulation, ILCregs can be induced in the intestine and suppress the activation of ILC1s and ILC3s via secretion of IL-10, leading to protection against innate intestinal inflammation. Moreover, TGF-b1 is induced by ILCregs during the innate intestinal inflammation, and autocrine TGF-b1 sustains the maintenance and expansion of ILCregs. Therefore, ILCregs play an inhibitory role in the innate immune response, favoring the resolution of intestinal inflammation. INTRODUCTION The intestine represents a major gateway for potential pathogens, which also contains dietary antigens and an extensive and diverse microbial flora that need to be tolerated. Due to these requirements, the gut constitutes the largest lymphoid organ in the body, encompassing an extensive network of secondary lymphoid organs and is home to an enormous number of lymphocytes (Izcue et al., 2009). Intestinal immune responses take place in the epithelium, lamina propria, and gut-associated lymphoid tissue (GALT). The gut harbors several intestine-specific subpopulations specialized in antigen presentation, antimicrobial immunity, and maintenance of tolerance. Breakdown in

the regulatory pathways leads to chronic intestinal inflammation (Kaser et al., 2010). For example, dysregulation of mucosal T cell responses may cause a loss of tolerance, leading to harmful intestinal inflammation reminiscent of human inflammatory bowel disease (IBD) (Strober et al., 2002). Moreover, the regulatory T cells (Tregs) play a critical role in the maintenance of intestinal homeostasis and self-tolerance (Asseman et al., 1999; Sakaguchi, 2005). However, how the intestinal innate immune system is regulated during the intestinal inflammation remains elusive. Innate lymphoid cells (ILCs) are located in mucosal surfaces to potentiate immune responses, sustain mucosal integrity and promote lymphoid organogenesis (Diefenbach et al., 2014; Eberl et al., 2015). Group 1 (ILC1) cells are characterized by their capacity to secrete interferon g (IFN-g) responding to interleukin-12 (IL-12), IL-15, and IL-18 (Gordon et al., 2012; Klose et al., 2014). Group 2 (ILC2) cells generate type 2 T helper (Th2) cell cytokines, such as IL-5, IL-9, and IL-13, in response to IL-25 and IL-33 stimulation (Brestoff et al., 2015; Moro et al., 2010). Group 3 (ILC3) cells produce IFN-g, IL-17, and IL-22 following stimulation with IL-1b and IL-23 (Buonocore et al., 2010; Klose et al., 2013). Given that ILCs produce substantial effector cytokines when stimulated, they play a critical role in the regulation of type 1, type 2, and type 3 (or Th17 cell) responses, controlling host protective immunity and intestinal homeostasis (Bedoui et al., 2016; Gasteiger and Rudensky, 2014; Shih et al., 2016). Accumulating evidence shows that inflammatory responses can be induced and regulated independently of adaptive immunity (Sonnenberg and Artis, 2015). Rag (V[D]J recombination activation gene)-deficient mice develop colitis after treatment with anti-CD40 antibody or dextran sodium sulfate (DSS) (Buonocore et al., 2010; Wirtz et al., 2007). Intraepithelial ILC1s undergo expansion and produce large amounts of IFN-g in the human Crohn’s disease. Blocking of ILC1 with anti-NK1.1 antibody in mice is able to attenuate colitis (Fuchs et al., 2013). In addition, Cell 171, 1–16, September 21, 2017 ª 2017 Elsevier Inc. 1

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Figure 1. ILCregs Exist in Mouse and Human Intestines (A) Flow cytometry analysis of ILCregs in different tissues of IL-10-GFP reporter mice. Lin
CD45+ lymphocytes were gated out for analysis of CD127 versus IL-10GFP. The total number of ILCregs was calculated and is shown as the mean ± SD. sLP, small intestine lamina propria; sIE, small intestine intra-epithelial

(legend continued on next page)

2 Cell 171, 1–16, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

ILC3s are implicated in the innate intestinal inflammation via secretion of IL-17 and IL-22. IL-17-producing ILC3s are responsible for the pathogenesis of Tbx21
/
Rag2
/
ulcerative colitis mice (Garrett et al., 2007; Powell et al., 2012). Depletion of ILC3s by anti-Thy1 antibody abrogates innate colitis (Buonocore et al., 2010), suggesting a critical role of ILC3s in the pathogenesis of intestinal inflammation. However, it is unknown whether a regulatory subset of ILCs exists and how they function in the regulation of intestinal inflammation. Here, we identify a novel regulatory subpopulation of ILCs (called ILCregs) that exists in the gut and expands in the intestine following pathogenic stimulation. ILCregs exert an inhibitory role in the innate immune response against intestinal inflammation. RESULTS ILCregs Exist in Mouse and Human Intestines Given that Treg cell-mediated suppression through secretion of their feature cytokines IL-10 and TGF-b (Izcue et al., 2009; Josefowicz et al., 2012), we then gated out Lin
CD45+CD127+ ILCs from multiple tissues in IL-10-GFP reporter mice and analyzed IL-10 expression (Kamanaka et al., 2006). We noticed that a unique subset of ILCs constitutively expressed IL-10 in the intestine (Figures 1A and S1A). These Lin
CD45+CD127+IL-10+ ILCs were mainly located in the lamina propria of the small intestine (sLP), and some of them in the small intestinal epithelium (sIE) and colon lamina propria (cLP) (Figure 1A). Lin
CD45+CD127+ IL-10+ ILCs were small in size and had a high nucleus/cytoplasm ratio and scanty cytoplasm, which are all characteristics of lymphoid morphology (Figures 1B and 1C). Of note, Lin
CD45+ CD127+IL-10+ ILCs were devoid of expression of CD4 and FoxP3 (Figure S1B), which are signature markers of Tregs (Sakaguchi et al., 2008), indicating that the Lin
CD45+CD127+IL-10+ ILC subset was different from the Tregs subset. In addition, Lin
CD45+CD127+IL-10+ ILCs also expressed ILC markers, such as CD25 (IL-2Ra) and CD90 (Thy1). This population also highly expressed IL-2Rg and Sca-1, but lacked ILC1 markers (NK1.1 and NKp46), ILC2 markers (ST2 and KLRG1), ILC3 markers (NKp46, CD4, and RORgt), or other leukocyte lineage markers (Gury-BenAri et al., 2016; Robinette et al., 2015) (Figures 1D and S1C). These Lin
CD45+CD127+IL-10+ ILCs represented

a new IL-10-producing ILC population that we have named regulatory ILCs (ILCregs). IL-10-producing ILCregs were further confirmed through imaging flow cytometry by using anti-IL-10 antibody (Figure 1E). Furthermore, ILCregs with co-expression of CD127 and IL-10 indeed existed in mouse small and large intestines by immunofluorescence staining (Figures 1F and S1D). These observations were further validated through immunohistochemical staining (Figure 1G). In addition, ILCregs also existed in the human intestine from biopsies (Figures 1H, 1I, and S1E). Collectively, the Lin
CD45+CD127+IL-10+ cells are a new ILC subset that exists in mouse and human intestines. ILCregs Are Distinct from ILCs and Tregs We isolated ILCregs from mouse sLP and performed transcriptome microarray assays. Comparing existing transcriptome datasets of ILCs with natural killer (NK) cells (GEO: GSE37448) and Tregs (GEO: GSE68009) highlights that ILCregs do not show any significant gene profile similarity to known ILC subsets and Tregs (Figure 2A). Of note, ILCregs do express high levels of Il10 and several ILC feature markers, including Il7r (encoding CD127), Il2ra (encoding CD25), Il2rg, and Ly6a (encoding Sca-1), which is consistent with our previous observations (Figure 1D). ILCregs uniquely express transcription factors, such as Id3 and Sox4 (Figure 2A), and they also highly expressed Id2, which is required for development of ILCs (Diefenbach et al., 2014; Spits and Cupedo, 2012). However, ILCregs lack typical transcription factors of other ILCs and Tregs, such as Rorc (encoding RORgt), Tbx21 (encoding T-bet), Gata3, and Foxp3, respectively (Serafini et al., 2015). Intriguingly, ILCregs constitutively express Tgfbr1, Tgfbr2, Il2rb, and Il2rg (Figure 2A), suggesting they responded to TGF-b and IL-2 signaling. Principal component analysis further verified that ILCregs harbor a unique gene profile (Figure 2B). Similarly, human ILCregs show distinct gene profiles of transcription factors and cytokines compared to other ILCs (Figure 2C). Additionally, some feature cytokines, such as IL-10 and transforming growth factor b1 (TGF-b1), are also highly expressed in human ILCregs as well. High expression levels of TGF-bRI, TGF-bRII, IL-2Rb, and Id2 (inhibitor of DNA binding 2) in ILCregs were validated by flow cytometry (Figures S1F and S1G). Moreover, Il2rg- or Id2-deficient

cells; cLP, colon lamina propria; mLN, mesenteric lymph nodes; PP, Peyer’s patches; BM, bone marrow. (Lin = CD3e, CD4, CD8, CD19, NK1.1, CD11b, CD11c, Gr1, Ter119). (B) Giemsa staining of sorted Lin
CD45+CD127+IL-10-GFP+ ILCreg cells from IL-10-GFP mice. Scale bar, 5 mm. (C) Ultrastructures of ILCregs were analyzed by electron microscopy. Lin
CD45+CD127+IL-10-GFP+ ILCregs were sorted out and visualized by electron microscopy. Scale bar, 2 mm. (D) Analysis of surface and transcription markers on ILCregs by flow cytometry. Gray histograms depict isotype control of each antibody. Red lines indicate ILCregs. Blue lines denote Lin
CD45+CD127+ cells as positive controls. (E) Imaging flow cytometry analysis of ILCregs. More than 90% of cells exhibited similar distribution of CD45, IL-10, and CD127 in Lin
CD45+IL-10+CD127+ gate, and three of them were shown. Scale bar, 5 mm. (F) Immunofluorescence staining of ILCregs in mouse small intestine. Intestinal sections were stained with primary antibodies against Lin (blue), CD127 (red), and IL-10 (green). DAPI (gray) denotes nucleus. The white arrowhead indicates ILCregs (Lin
CD127+IL-10+). Scale bar, 50 mm. (G) Immunohistochemical staining of ILCregs in mouse intestine by using anti-Lin (red) and anti-IL-10 (green) antibodies. The black arrowhead indicates ILCregs (Lin
IL-10+). Scale bar, 50 mm. (H) Analysis of ILCregs in human intestines by flow cytometry. Lin
CD45+ lymphocytes from human duodenum or colon were gated out for CD127 versus IL-10 analysis. (I) Immunofluorescence staining of ILCregs in human duodenum. The white arrowhead indicates ILCregs (Lin
CD127+IL-10+). Scale bar, 50 mm. All data are representative of at least three independent experiments. See also Figure S1.

Cell 171, 1–16, September 21, 2017 3

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

(legend on next page)

4 Cell 171, 1–16, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

mice abrogated ILCregs (Figure S1H), which are common characters similar to other ILC subsets (Serafini et al., 2015). In contrast, Rag1
/
mice show a normal number of ILCregs compared to wild-type (WT) mice (Figure S1H), and ILCregs of Rag1
/
mice do not express immature T cell markers (data not shown), suggesting they develop independent of Rag. By contrast, Id2 deficiency failed to impact the development of Tregs (Figure S1I). To further determine whether ILCregs are derived from myeloid or T cells, we generated Rosa26-STOPYFP;Lyz2-Cre;IL-10-GFP mice and Rosa26-STOP-YFP;CD4Cre;IL-10-GFP mice. We found that ILCregs were indeed negative for YFP signals in the above two mouse strains, whereas their respective lineage positive cells virtually displayed YFP signals (Figures S2A and S2B), indicating that ILCregs are not derived from myeloid or T cells. Recently, integrin a4b7+Id2high CHILP (common helper-like innate lymphoid precursor) was identified as a common precursor for all helper-like ILCs, including LTi and ILC1-3 subsets (Klose et al., 2014), whereas PLZF+ ILCP (common ILC precursor) is the precursor for ILC1-3s, only giving rise to ILC1s, ILC2s, and ILC3s (Ishizuka et al., 2016). Next, we crossed Rosa26-STOP-YFP mice with different Cre mice to determine the precursor of ILCregs. We observed that only Rosa26-STOP-YFP;Id2-CreERT2 mice displayed YFP-positive ILCregs in the presence or absence of DSS treatment (Figures S2C and S2D). However, ILCregs in Rosa26-STOP-YFP;PLZF-Cre mice or Rosa26-STOP-YFP; RORgt-Cre mice were YFP negative (Figures S2C and S2D). These data suggest that ILCregs are derived from Id2-driven CHILPs, but not ILCPs or ILC3s. To further validate IL-10-producing ILCregs represented a distinct lineage as opposed to a further differentiation state of ILC1, ILC2, or ILC3 cells, we established Id2- CreERT2;RosaDTR, PLZF-Cre;RosaDTR, and RORgt-Cre;RosaDTR mice, followed by adoptive transfer assays. With diphtheria toxin (DT) treatment, Id2-CreERT2;RosaDTR mice abrogated ILCreg cells (Figure S2E), suggesting ILCregs were generated from the progenitor CHILP (Klose et al., 2014). However, deletion of PLZF- or RORgt-expressing cells does not affect the development of ILCregs (Figure S2E), indicating that ILCregs do not differentiate from the progenitor of LTi, ILC1, ILC2, or ILC3 cells (Artis and Spits, 2015). To further verify the progenitor of ILCregs, we transferred CLP, aLP, CHILP, and ILCP progenitors into Rag1
/
Il2rg
/
mice for adoptive transfer assays. We noticed that CLP, aLP, and CHILP were able to produce ILCregs, whereas ILCPs could not generate ILCregs (Figure S2F). Depletion of CHILPs by DT treatment in adoptively reconstituted Rag1
/
Il2rg
/
mice abrogated generation of ILCregs (Figure S2G). Collectively, ILCregs are derived from CHILPs, but

not ILCPs, which are a separate subset distinct from Tregs and other ILC lineages. ILCregs Contribute to the Resolution of Innate Intestinal Inflammation Since ILCregs mainly reside in the intestine and develop independent of Rag, we next used innate intestinal inflammation models to monitor ILCreg functions. We treated Rag1
/
mice with several inflammatory stimuli, including DSS, anti-CD40 antibody, Salmonella typhimurium (S. typhimurium), and Citrobacter Rodentium (C. rodentium). We noticed that ILCregs could be induced in the intestines of Rag1
/
mice with these inflammatory stimuli (Figure 3A). Notably, ILCregs gradually expanded in the intestines during the process of inflammation (Figures 3B, 3C, and S3A), reaching a peak level at 8 days post-stimulation, which came up at peak inflammation (Figure S3B). By contrast, the number of ILC1s and ILC3s immediately increased at early stages post-stimulation and decreased soon afterward (Figure S3C), whereas the number of ILC2s did not change during the process of inflammation (Figure S3C). These data indicate that ILCregs can be induced by innate intestinal inflammation. It has been reported that IL-10-deficient mice display spontaneous colitis (Josefowicz et al., 2012; Ku¨hn et al., 1993). Since IL-10 was a signature cytokine of ILCregs, we then generated Rag1
/
Il10
/
mice to test the inhibitory role of IL-10 in the regulation of innate colitis. We observed that Rag1
/
Il10
/
mice caused more severe colitis with DSS stimulation than Rag1
/
mice alone (Figures 3D and 3E), indicating that IL-10 was involved in the suppression of innate colitis. We then transferred WT ILCregs into Rag1
/
Il10
/
mice, and ILCregs and IL-10 were successfully rescued in Rag1
/
Il10
/
mice (Figures S3D and S3E). Moreover, engrafted ILCregs could proliferate in vivo with DSS stimulation (Figure S3F). We observed that ILCreg transfer was able to prevent pathogenic innate colitis of their rescued Rag1
/
Il10
/
mice (Figures 3D–3F). Finally, ILCreg transfer exhibited a higher survival rate with DSS treatment (Figure 3G). Similar observations were achieved by infection with S. typhimurium (Figures S3G and S3H). Collectively, ILCregs are able to protect mice from innate immune colitis. ILCregs Produce IL-10 and TGF-b1 during Intestinal Inflammation Next, we analyzed cytokine expression profiles of ILCregs in Rag1
/
mice during inflammatory stimulation. We observed that ILCregs in Rag1
/
mice generated high levels of IL-10 and TGF-b1 after treatment with the previously used stimuli

Figure 2. ILCregs Harbor a Unique Gene Profile Distinct from ILCs and Tregs (A) Gene expression profiling of ILCregs. mRNA of ILCregs was extracted and analyzed by gene expression profiling by using Mouse Transcriptome Array (v.1.0) (Affymetrix). Genes related to mouse ILCs and mouse Tregs were selected out for heatmap analysis. Biological replicates from two independent experiments are shown. (B) Principal component analysis of the selected genes in (A) from the indicated cell populations. (C) Transcriptome microarray analysis of cytokines and transcription factors of human ILCregs. Human ILCregs were isolated, and mRNA of ILCregs was extracted followed by gene expression profiling by using Human Transcriptome Array (v.2.0) (Affymetrix). Genes related to human ILCs were selected for heatmap analysis. See also Figures S1 and S2.

Cell 171, 1–16, September 21, 2017 5

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Figure 3. ILCregs Favor the Resolution of Intestinal Inflammation (A) Rag1
/
mice were treated with 3% DSS in drinking water for the indicated days, 150 mg anti-CD40 antibody (intraperitoneally [i.p.]) on day 0, or infected by oral gavage of 5 3 104 colony-forming units (CFUs). S. typhimurium or 1 3 109 CFUs. C. rodentium on day 0. Cell numbers of ILCregs were calculated and are shown as means ± SD. n = 6 for each group. (B) Rag1
/
mice were treated with 3% DSS for the indicated days, and ILCregs were analyzed by flow cytometry. Lin
CD45+ lymphocytes were gated out for IL-10 versus CD127 staining for ILCregs. ILCregs were gated out for analysis of Ki67 by flow cytometry. (C) Immunofluorescence staining of ILCregs in mouse intestine after treatment with DSS by using antibodies against Lin (blue), CD127 (red), and IL-10 (green). The white arrowhead indicates ILCregs. Scale bar, 50 mm. (D) ILCregs from IL-10-GFP mice were isolated and adoptively transferred into Rag1
/
Il10
/
mice. After 2 weeks, mice were treated with 3% DSS, and body weight changes of mice were calculated and are shown as means ± SD. n = 10 for each group. **p < 0.01 by a non-parametric Mann-Whitney U-test. (E) Colitis score of the indicated mice by 8 days’ oral gavage of water (Ctrl) or DSS. **p < 0.01 by one-way ANOVA. n = 5 for each group. (F) H&E staining of colon of indicated mice treated like in (D). Scale bar, 50 mm. (G) Survival rates of adoptively transferred mice. n = 10 for each group. p values (Rag1
/
Il10
/
versus Rag1
/
Il10
/
+ILCreg) were analyzed by the KaplanMeier method. All data are representative of at least three independent experiments. See also Figure S3.

6 Cell 171, 1–16, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Figure 4. ILCregs Produce IL-10 and TGF-b1 during Intestinal Inflammation (A) Cytokine expression profiles of ILCregs after inflammatory stimulation. mRNA of ILCregs from Rag1
/
mice with indicated treatments for 8 days was extracted and subjected to RT-PCR. (B and C) Rag1
/
mice were stimulated with indicated reagents or pathogens for different days. ILCregs were isolated and subjected to mRNA extraction and RT-PCR of Il10 and Tgfb1 mRNA (B) or cultured in media for 24 hr followed by an ELISA of secreted IL-10 and TGF-b1 by ILCregs (C). Untreated Rag1
/
mice served as controls (Ctrl). (D and E) ILCregs inhibit cytokine production of ILC1s and ILC3s in vitro. Isolated ILC1s (Lin
CD127+NK1.1+NKp46+) or ILC3s (Lin
CD45+CD127+RORgt-GFP+ from RORgt-GFP mice) were incubated with ILCregs from DSS-treated mice in the presence or in the absence of the indicated reagents for 3 days. Intracellular cytokines were analyzed by flow cytometry (D), and secreted cytokines were analyzed by ELISA (E). Data are shown as means ± SD. ND, not detectable. **p < 0.01 by one-way repeated-measures (RM) ANOVA. (F and G) Flow cytometry analysis of activation markers on ILC1s (F) and ILC3s (G) after treatment with cytokines in the presence or absence of ILCregs from DSStreated mice. All data are representative of at least three independent experiments. See also Figure S4 and Table S1.

(Figures 4A–4C; Table S1). Of note, TGF-b1 was induced earlier than IL-10 during innate immune responses (Figures 4B and 4C). Moreover, ILCregs significantly expanded in the intestine with DSS stimulation (Figure S4A). Additionally, ILCregs secreted both IL-10 and TGF-b1 in the small and large intestines of Rag1
/
mice with DSS treatment (Figures S4A and S4B). Alto-

gether, ILCregs generate IL-10 and TGF-b during innate immune responses. Growing evidence indicates that ILCs mediate T-cell-independent innate colitis (Buonocore et al., 2010). Our above observations demonstrated that ILCregs protected against innate colitis. We thus hypothesized that ILCregs might inhibit the functions of

Cell 171, 1–16, September 21, 2017 7

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

ILCs to prevent innate immune responses. We isolated ILCregs from sLP of DSS-treated mice and incubated them with activated ILC1s, ILC2s, and ILC3s, respectively. We noticed that ILCregs were able to inhibit cytokine secretion by ILC1s (IFN-g) and ILC3s (IFN-g and IL-17A), but not by ILC2s (Figures 4D, 4E, S4C, and S4D). Of note, anti-IL-10 antibody treatment abrogated inhibitory roles of ILCregs (Figures 4D and 4E). However, ILCregs failed to suppress IL-22 secretion by ILC3s (Figures S4E and S4F). Furthermore, IL-10 alone could suppress the activation of ILC1s and ILC3s (Figures 4D and 4E). Of note, via titration assays, ILCregs effectively inhibited the activities of ILC1/3 at ratios around 1:2 to 2:1 (Figure S4G). Ki67 and CD69 are two feature molecules that denote the activation of ILCs (Munneke et al., 2014). We found that ILCregs also inhibited the expression of Ki67 and CD69 on activated ILC1s and ILC3s, respectively (Figures 4F and 4G). By contrast, addition of TGF-b1 did not impact the activation of ILC1s and ILC3s (Figures S4H and S4I). These data indicate that ILCregs suppress the activation of ILC1s and ILC3s via secretion of IL-10. ILCregs Suppress the Activation of ILC1s and ILC3s via Secretion of IL-10 To further validate the inhibitory role of ILCregs in vivo, we generated Rag1
/
Il2rg
/
mice that were devoid of T, B, NK, and ILC cells. We then transferred activated ILC1s (aILC1) or activated ILC3s (aILC3), together with ILCregs from DSS-treated mice into Rag1
/
Il2rg
/
mice (Figure 5A). We noticed that transfer of aILC1 or aILC3 alone was able to cause innate colitis (Figures 5A–5E). As expected, co-transfer of ILCregs could protect mice from innate colitis caused by activated ILCs (Figure 5A–5E). Finally, ILCreg transplantation suppressed the secretion of inflammatory cytokines by ILCs (Figures 5F and 5G). As expected, neutralization of IL-10 by anti-IL-10 antibody was able to block the inhibitory role of ILCregs (Figures 5A–5G). Next, we generated Il10raflox/flox;Id2-CreERT2 mice, followed by adoptive transfer assays. Transfer of IL-10Ra-deleted ILC1s and ILC3s with WT ILCregs into Rag1
/
Il2rg
/
mice caused much more severe colitis and more elevated activities of ILC1/3 (Figures S5A– S5C). In addition, we used CRISRP/Cas9 technology to delete IL-10 in ILCregs. We found that transplantation of IL-10-deleted ILCregs in adoptively reconstituted Rag1
/
Il2rg
/
mice also caused much more severe colitis and more elevated activities of ILC1/3 (Figures S5D–S5G). Collectively, ILCregs inhibit the activation of ILCs and protect against intestinal innate inflammation in an IL-10-dependent manner. We then isolated FoxP3-DTR-GFP Treg cells from FoxP3DTR-GFP mice and transferred them together with ILC1s and ILC3s into Rag1
/
Il2rg
/
mice (Figure S6A). With DT treatment, Tregs were almost depleted in recipient mice (Figure S6A). We noticed that mice with depletion of Tregs did not affect innate intestinal inflammation, which was comparable to that of mice with replete Tregs (Figures S6B and S6C). Consistently, depletion of Tregs did not impact the activities of ILC1s and ILC3s in engrafted mice (Figure S6D). Based on the microarray analysis, ILCregs constitutively expressed Lck (Figure 2A). We then generated Lck-Cre;RosaDTR;IL-10-GFP mice for isolation of ILCregs with DTR (named ILCregDTR cells), followed by adoptive transfer assays (Figure S6E). DT treatment completely depleted ILCreg

8 Cell 171, 1–16, September 21, 2017

cells in the intestine (Figure S6E). We observed that mice with depletion of ILCregs caused much more severe intestinal inflammation compared to those of vehicle-treated control mice (with ILCregs) (Figures S6F–S6J). Consequently, ILCreg depletion remarkably promoted the activation of ILC1s and ILC3s in engrafted mice (Figure S6H). Collectively, Tregs have no significant inhibitory effect on ILC1s and ILC3s-mediated innate inflammation, but ILCregs really attenuate ILC1s and ILC3s-mediated innate inflammation. In order to further define the physiological role of ILCregs in a lymphocyte replete setting, we isolated ILCregDTR cells from Lck-Cre;RosaDTR;IL-10-GFP mice and transferred them together with T, B, NK, and ILC cells of WT mice into Rag1
/
Il2rg
/
mice (Figure S6K). ILCregDTR, T, B, NK, and ILC cells were successfully reconstituted in Rag1
/
Il2rg
/
mice (Figure S6K). With DT treatment, ILCregs were almost lost in engrafted Rag1
/
Il2rg
/
mice (Figure S6L). Following DSS challenge, mice with depletion of ILCregs caused more severe intestinal inflammation compared to those mice with replete ILCregs (Figures S6M and S6N). Consequently, depletion of ILCregs markedly augmented the activation of ILC1s and ILC3s with DSS treatment (Figure S6O). We conclude that ILCregs play an inhibitory role in the regulation of innate immunity. Id3 Is Required for the Development of ILCregs Based on our above microarray analysis (Figure 2A), ILCregs uniquely, highly expressed the translational regulator Id3 (inhibitor of DNA binding 3). Id3 belongs to the Id family members that contain an HLH dimerization domain but lack the basic DNAbinding region. The Id proteins (Id1-4) modulate the DNA-binding activity of E proteins that act as transcriptional activators or repressors (Rivera et al., 2000). Accumulating evidence shows that Id2 and Id3 play critical roles in regulating the developmental progression of T cells (Li et al., 2004; Miyazaki et al., 2014). We further confirmed the expression of Id3 in ILCregs at mRNA and protein levels (Figures 6A and 6B). Next, we wanted to test the role of Id3 during ILCreg development. We generated Id3flox/flox;Id2-CreERT2 mice, followed by tamoxifen (TMX) treatment. We found that Id3 deletion abrogated ILCreg generation, but did not impact the development of other ILCs and T cells (Figure 6C), whose phenotype was different from that of Id2-deficient mice. In addition, we rescued WT ILCregs into Id3flox/flox;Id2-CreERT2 mice (Figure 6D). Of note, deficiency of Id3 abrogated ILCreg population and displayed severe colitis, which could be attenuated by rescue of WT ILCregs (Figures 6E and 6F). Activation of ILC1/3s was weakened after adoptive transfer of ILCregs (Figure 6G), indicating that deficiency of Id3 impairs ILCreg-related functions. We conclude that Id3 is a unique fate determination factor to drive the development and/or maintenance of ILCregs. Autocrine TGF-b1 Is Required for the Expansion of ILCregs during Inflammation Next, we wanted to examine which cytokine signaling induced the development and/or maintenance of ILCregs. As ILCregs expressed high levels of IL-2R and TGF-bR (Figure S1F), we then isolated ILCregs from IL-10-GFP reporter mice and incubated them with common g-chain-related cytokines and TGF-b1. We

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Figure 5. ILCregs Suppress the Activation of ILC1s and ILC3s via Secretion of IL-10 (A) IL-12/IL-18-activated ILC1s (aILC1) or IL-23-activated ILC3s (aILC3) isolated from CD45.1 mice were intravenously injected alone or together with ILCregs from DSS-treated CD45.2 mice (1:1 ratio) into Rag1
/
Il2rg
/
mice. Transferred mice were injected i.p. with PBS or 200 mg of anti-IL-10 antibody daily. Body weight changes were calculated and are shown as means ± SD. n = 5 for each group. **p < 0.01 by one-way ANOVA. (B and C) Colon images of indicated mice after adoptive transfer for 10 days. Colon length of mice was measured (C). **p < 0.01 by one-way ANOVA. n = 5 for each group. (D) Colitis score of mice treated as in (A) were analyzed. **p < 0.01 by one-way ANOVA. n = 5 for each group. (E) H&E staining of colon sections from the indicated mice after adoptive transfer for 10 days. Scale bar, 50 mm. The histological pictures showed inflammation (**). (F and G) Production of IFN-g by ILC1s or ILC3s and IL-17A by ILC3s from indicated mice were analyzed by flow cytometry (F) or ELISA (G). **p < 0.01 by a twotailed unpaired Student’s t test. All data represent at least three independent experiments. See also Figures S5 and S6.

Cell 171, 1–16, September 21, 2017 9

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Figure 6. Id3 Is a Fate Decision Factor for ILCreg Development (A and B) Expression levels of Id3 in different lymphocyte populations. Expression levels of Id3 in indicated populations were analyzed by RT-PCR for mRNA (A) or flow cytometry for protein levels (B). (C) Deletion of Id3 impairs generation of ILCregs. Id3flox/+;Id2-CreERT2 or Id3flox/flox;Id2-CreERT2 mice were treated with 50 mg/kg (i.p.) tamoxifin (TMX) for 5 consecutive days followed by flow cytometry analysis. Relative cell numbers of each population compared to WT mice were calculated and are shown as means ± SD. **p < 0.01 by a two-tailed unpaired Student’s t test. NS, not significant. (D) Indicated mice were treated with TMX followed by flow cytometry analysis. WT ILCregs were transferred into Id3flox/flox;Id2-CreERT2 mice for rescue of ILCregs. (E and F) H&E staining (E) and colitis scores (F) of colons of indicated mice treated for 7 days of 3% DSS in drinking water (+DSS). Scale bar, 50 mm. **p < 0.01 by one-way RM ANOVA. (G) Secreted cytokines of ILCs from the indicated mice like in (D) were analyzed by ELISA and are shown as means ± SD. **p < 0.01 by One Way RM ANOVA. All data are representative of at least three independent experiments.

noticed that IL-2 and TGF-b1 were able to promote expansion of ILCregs (Figure 7A), suggesting that IL-2 and TGF-b signaling are required for the proliferation and/or maintenance of ILCregs. Above, we demonstrated that TGF-b1 was induced in ILCregs during innate immune responses (Figures 4A–4C). Next, we sought to test whether autocrine TGF-b1 exerted its function on the development of ILCregs. We then generated Tgfb1flox/flox;Lck-Cre mice to delete TGF-b1 expression in

10 Cell 171, 1–16, September 21, 2017

ILCregs. We mixed bone marrow (BM) of WT (CD45.1) and Tgfb1flox/flox;Lck-Cre (CD45.2) mice to reconstitute sublethally irradiated Rag1
/
Il2rg
/
recipients. Interestingly, ILCregs from WT (CD45.1) and Tgfb1flox/flox;Lck-Cre (CD45.2) were successfully reconstituted in BM transplanted recipient mice (Figure S7A). Of note, numbers of CD45.1 and CD45.2 hematopoietic progenitors were comparable (Figure S7B). These data suggest that autocrine TGF-b1 is dispensable for the

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

(legend on next page)

Cell 171, 1–16, September 21, 2017 11

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

development of ILCregs. After treatment with DSS, the number of WT ILCregs (CD45.1) increased 2-fold in transplanted recipient mice compared to that with non-DSS treatment (Figure S7A). However, the number of TGF-b1-deleted ILCregs (CD45.2) dramatically reduced post-DSS treatment (Figure S7A). These data indicate that autocrine TGF-b1 is required for the expansion and survival of ILCregs during innate immune responses. Next, we generated Tgfb1flox/flox;CreERT2;IL-10-GFP mice to determine the role of TGF-b1 secreted by ILCregs in their expansion. We then isolated ILCregs from Tgfb1flox/flox;CreERT2;IL-10GFP mice and incubated them in media containing IL-2 following treatment with 4-hydroxytamoxifen (4-HT). With administration of 4-HT, ILCregs abrogated TGF-b1 production (Figure 7B). Consequently, the numbers of ILCregs were dramatically reduced (Figure 7C), suggesting autocrine TGF-b1 is required for the expansion of ILCregs. To further verify TGF-b1 acted as an autocrine manner in vivo, we transferred Tgfb1flox/flox;CreERT2;IL-10-GFP ILCregs, together with ILC1s and ILC3s, into Rag1
/
Il2rg
/
mice (Figure 7D). After administration of TMX, ILCregs were almost abolished in the intestine with DSS stimulation (Figures 7D, 7E, and S7C). These observations were further validated by in situ immunofluorescence staining (Figure 7F). Moreover, TMX-treated mice lost body weight quickly and displayed more severe colitis with DSS treatment (Figures 7G–7J). Finally, these mice produced substantial amounts of IFN-g and IFN-g/IL-17A by respective ILC1s and ILC3s (Figure 7K). Similar results were obtained by S. typhimurium infection (Figures S7D and S7E). Altogether, autocrine TGF-b1 by ILCregs sustains the maintenance/ survival and expansion of inflammation-activated ILCregs. The TGF-b receptor complex is a tetrameric structure, consisting of two type I TGF-b receptors (TGF-bRI) and two type II TGF-b receptors (TGF-bRII) (Travis and Sheppard, 2014). Deletion either of them is able to abrogate TGF-b signaling (Kang et al., 2009). We generated Tgfbr2flox/flox;CreERT2;IL-10-GFP

mice and performed adoptive transfer assays (Figure S7F). After TMX treatment, ILCregs almost lost in the intestine over DSS treatment (Figure S7G). Consequently, Tgfbr2 abrogation in ILCregs led to more severe colitis and more persistent intestinal damage (Figures S7H and S7I). In parallel, these Tgfbr2-deficient mice exhibited active ILC1s and ILC3s (Figure S7J). Therefore, TGF-b1 signaling is required for the expansion and survival of ILCregs. DISCUSSION It has been reported that innate lymphoid cells (ILCs) play important roles in innate immunity (Spits and Cupedo, 2012). ILCs are able to cause inflammation independent of T and B cells (Buonocore et al., 2010; Coccia et al., 2012; Powell et al., 2012). However, how the functions of ILCs in intestinal inflammation are regulated still remains elusive. Here, we show that a regulatory subset of ILCs (we called ILCregs) exist in mouse and human intestines. During inflammatory stimulation, ILCregs can be induced in the intestine and suppress the activation of ILC1s and ILC3s via secretion of IL-10, leading to protection against innate intestinal inflammation. Moreover, ILCregs also produce large amounts of TGF-b1, and this autocrine TGF-b1 is required for their expansion and survival during intestinal inflammation. Therefore, ILCregs play a regulatory role in favoring the resolution of innate intestinal inflammation. ILCs are rapidly activated at the early stage of innate immune response and implicated in eliminating pathogens-infected cells or provoking profound immune response (Eberl et al., 2015). Many studies reported that lack or dysfunction of ILCs by deletion of key transcription factors or signaling factors causes apparent pathologic disorders (Guo et al., 2015; Rankin et al., 2013; Spooner et al., 2013). Like all the immune responses, the activation of these cells also needs to be balanced by stringent regulations because the excessive activation may result in tissue

Figure 7. Autocrine TGF-b1 Is Required for the Expansion of ILCregs (A) ILCregs from IL-10-GFP mice were isolated, treated with indicated cytokines (2 ng/ml murine IL-2, 5 ng/ml murine IL-4, 5 ng/ml murine IL-7, 5 ng/ml murine IL-9, 5 ng/ml murine IL-15, 5 ng/ml murine IL-21, 5 ng/ml murine IL-10, and 5 ng/ml murine TGF-b1), and analyzed by flow cytometry. Fold changes of cell numbers were analyzed and are shown as means ± SD. (B) ILCregs from Tgfb1flox/flox;CreERT2;IL-10-GFP mice were isolated and cultured in media containing 2 ng/ml IL-2 with or without of 0.5 mm 4-hydroxytamoxinfen (4-HT) treatment for the indicated days. TGF-b1 amounts were analyzed by ELISA and are shown as means ± SD. **p < 0.01 by a two-tailed unpaired Student’s t test. (C) Cell numbers of ILCregs in (B) were calculated and are shown as means ± SD. **p < 0.01 by a two-tailed unpaired Student’s t test. (D) ILCregs were isolated from Tgfb1flox/flox;CreERT2;IL-10-GFP mice (CD45.2) and transferred with ILC1s (CD45.1) and ILC3s (CD45.1) into Rag1
/
Il2rg
/
mice. After 2 weeks, mice were treated with water (Ctrl) or 3% DSS (+DSS) or injection of tamoxinfen (TMX) with oral gavage of 3% DSS (+DSS/TMX) for 8 days. ILCregs were analyzed by flow cytometry. (E) Cell numbers of ILCregs in different tissues as treated in (D) are shown as means ± SD. **p < 0.01 by a two-tailed unpaired Student’s t test. (F) Immunofluorescence staining of ILCregs in the indicated mouse intestine sections. Scale bar, 50 mm. The white arrowhead indicates Lin
CD127+IL-10+ ILCregs. (G) Body weight changes of indicated mice were calculated and are shown as means ± SD. n = 6 for each group. **p < 0.01 by a non-parametric Mann-Whitney U-test. Mice transferred with ILCregs (from IL-10-GFP;CreERT2 mice), together with ILC1s and ILC3s, were treated with DSS and TMX and used as negative controls (Cre only+DSS/TMX). (H and I) Colitis score (H) and histology (I) of the indicated mouse colon were analyzed and are shown as means ± SD. **p < 0.01 by one-way ANOVA. n = 6 for each group. Scale bar, 50 mm. (J) Survival rates of the indicated adoptively transferred mice with administration of DSS and/or TMX. n = 10 for each group. p values (+DSS/TMX versus Cre only+DSS/TMX) were analyzed by Kaplan-Meier method. (K) ILC1s or ILC3s were isolated from indicated mice and secreted cytokines were analyzed by ELISA. Data are shown as means ± SD. **p < 0.01 by a nonparametric Mann-Whitney U-test. All data represent at least three independent experiments. See also Figure S7.

12 Cell 171, 1–16, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

damage. For example, ILCs are implicated in many inflammatory disorders, including inflammatory bowel diseases, allergic asthma, tissue fibrosis, and oncogenesis (Buonocore et al., 2010; Chang et al., 2011; Kirchberger et al., 2013; McHedlidze et al., 2013). Here, we have shown that ILCregs can inhibit the activation of ILC1s and ILC3s via secretion of IL-10 during the innate intestinal colitis. However, we found that ILCregs do not suppress the function of ILC2s in the process of intestinal inflammation. Of note, we noticed that Il10rb was highly expressed in ILC1s and ILC3s, but undetectable in ILC2s (data not shown). A recent study reported that ILC2s are involved in lung inflammation (Monticelli et al., 2016). Whether ILCregs exert tissue-specific regulation needs to be further investigated. Tolerance in the intestine relies on a wide array of independent immunosuppressive mechanisms. Accumulating evidence indicates that Tregs are able to utilize multiple mechanisms to inhibit immune responses in intestinal inflammation (Asseman et al., 1999; Izcue et al., 2009). Tregs can suppress the activation of other T cells either directly via a cell-contact-dependent fashion or indirectly through downregulating the activity of antigen-presenting cells (APCs) (Thornton and Shevach, 1998; Vignali et al., 2008). Tregs secreted effector cytokines, such as IL-10 and transforming growth factor (TGF)-b play a critical role in the regulation of intestinal inflammation (Maloy et al., 2003). Indeed, IL-10-deficient mice exhibit severe intestinal inflammation and elevated proinflammatory cytokines secretion (Ku¨hn et al., 1993). Moreover, polymorphisms at the Il10 locus confer risk for ulcerative colitis and Crohn’s disease (Franke et al., 2008, 2010), suggesting that IL-10 exerts a central role in the regulation of intestinal mucosal homeostasis. Here, we show that ILCregs uniquely secrete large amounts of IL-10 to downregulate the effector functions of ILC1s and ILC3s during the innate intestinal inflammation. However, Tregs have no apparent inhibitory roles in the suppression of ILC1s and ILC3s during the innate intestinal inflammation. More importantly, ILCregs were located very closely to ILCs in the small intestine (data not shown). The physical closer distance between ILCregs and ILCs might contribute to the regulatory activity of ILCregs to ILC1s and ILC3s. Of note, IL-22 exerts a protective role in mucosal immunity (Wang et al., 2014; Sugimoto et al., 2008). Here, we found that ILCregs inhibit the generation of IL-17A/IFN-g, but do not impact the production of IL-22 in ILC3s, which is beneficial to the protective function of ILCregs against innate inflammation. Thus, secretion of IL-22 and IL-17A/IFN-g by ILC3s might be separately regulated by ILCregs during the process of inflammation as Th17 cells do (Zheng et al., 2007). Accumulating evidence indicates that IL-27/IL-27R signaling triggers IL-10 production in a variety of immune cells (Saraiva and O’Garra, 2010). Here, we demonstrated that ILCregs express low levels of IL-27R, suggesting IL-27/IL-27R signaling may not participate in the regulation of IL-10 production in ILCregs. We observed that ILCregs are activated by IL-2 and TGF-b signaling. It has been reported that IL-2 and TGF-b signaling induces IL-10 generation in Tregs (Brandenburg et al., 2008; Maynard et al., 2007). Therefore, IL-2 and TGF-b signaling might be involved in the regulation of IL-10 production in ILCregs. How ILCregs induces IL-10 secretion needs to be further investigated.

All ILC lineages are originally derived from common lymphoid progenitors (CLPs) in the bone marrow, which also differentiate into T and B cells (Diefenbach et al., 2014). More recently, the progenitors that restrictedly differentiate into ILCs have been defined (Constantinides et al., 2014; Klose et al., 2014; Yang et al., 2015). ILCs are developed from the earliest a-lymphoid progenitor cells (aLPs, CXCR6+ integrin a4b7-expressing CLPs) (De Obaldia and Bhandoola, 2015), which differentiate into restricted CHILP cells (Klose et al., 2014). CHILPs generate all ILCs, including LTi cells, but they fail to give rise to conventional NK cells. Subsequently, their downstream precursor ILCPs (common precursor of ILCs), characterized by expression of the transcription factor PLZF, lose the ability to generate LTi cells and give rise to all ILC1, ILC2, and ILC3 subsets (Constantinides et al., 2014). RORgt (encoded by Rorc) drives differentiation of ILC3s from their precursor ILCPs (Montaldo et al., 2014). RORgt deletion causes a complete loss of ILC3s but not ILC1s or ILC2s. These observations indicate that the specifications of different ILC subsets are controlled by transcription factor networks. Here, we show that Id2 is required for the development of ILCregs as other ILCs, suggesting that ILCregs are derived from CHILPs. However, deletion of PLZF or RORgt did not affect the development of ILCregs, indicating that ILCregs were not derived from the all ILC precursor ILCPs. Our data suggest that CHILPs, but not from ILCPs, differentiate into ILCregs, validating that ILCregs are a distinct cell lineage. A recent study clustered all Lin
CD127+ cells into 23 transcriptionally homogeneous subgroups based on 6,637 differential genes (Gury-BenAri et al., 2016). Among these genes they used, however, several ILCreg feature genes were not included. We then chose 112 feature genes for ILCregs and ILCs and re-analyzed the single-cell RNA sequencing (RNA-seq) data (GEO: GSE85152). We noticed that ILCreg subpopulation was virtually clustered and distinct from the three existing ILC subpopulations (data not shown). Transforming growth factor-b (TGF-b) is a multifunctional cytokine that modulates cell survival, proliferation, and differentiation in a variety of settings ranging from embryogenesis to adult tissue homeostasis (Wu and Hill, 2009). TGF-b needs to bind its respective cell surface receptor to initiate TGF-b signaling for its effector functions. The TGF-b receptor complex is a tetrameric structure, consisting of two type I TGF-b receptors (TGF-bRI) and two type II TGF-b receptors (TGF-bRII) (Travis and Sheppard, 2014), deletion either of which can abrogate TGF-b signaling (Kang et al., 2009). For instance, TGF-b signaling plays critical roles in preventing uncontrolled immune responses via inducing development of Tregs (Curotto de Lafaille and Lafaille, 2009; Sakaguchi et al., 2008). A recent study showed that TGF-b signaling is required for the differentiation of salivary gland ILC1s through suppression of Eomes expression (Cortez et al. 2016). In this study, we show that ILCregs produce large amounts of TGF-b, and this autocrine TGF-b is required for the expansion and maintenance of inflammation-induced ILCregs. More importantly, autocrine TGF-b is dispensable for the development of ILCregs. Whether paracrine TGF-b affects the development and differentiation of ILCregs in the intestine needs to be further investigated.

Cell 171, 1–16, September 21, 2017 13

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

In sum, we have identified a new distinct regulatory subset of ILC cells (called ILCregs), and ILCregs exert inhibitory functions during the innate immune response, which contributes to resolution of innate intestinal inflammation. Hence, we believe that ILCregs may be used to develop new and potential therapies to restore immune tolerance in chronic inflammatory and autoimmune diseases for their effective treatment.

Received: December 7, 2016 Revised: June 16, 2017 Accepted: July 20, 2017 Published: August 24, 2017

STAR+METHODS

Asseman, C., Mauze, S., Leach, M.W., Coffman, R.L., and Powrie, F. (1999). An essential role for interleukin 10 in the function of regulatory T cells that inhibit intestinal inflammation. J. Exp. Med. 190, 995–1004.

Detailed methods are provided in the online version of this paper and include the following: d d d

d

d d

KEY RESOURCES TABLE CONTACT FOR REAGENT AND RESOURCE SHARING EXPERIMENTAL MODEL AND SUBJECT DETAILS B Mice and human tissues B Generation of inflammation mouse models METHOD DETAILS B Isolation of ILCreg cells B Flow cytometry assay B Imaging flow cytometry B Immunofluorescence assay B Adoptive transfer assays B ELISA assay B Histology and immunohistochemistry assay B Colitis scores and histologic analysis B Generation of IL-10 deficient ILCregs B Electron microscopy B Real-time PCR assay B Microarray analysis QUANTIFICATION AND STATISTICAL ANALYSIS B Statistical analysis DATA AND SOFTWARE AVAILABILITY

SUPPLEMENTAL INFORMATION Supplemental Information includes seven figures and one table and can be found with this article online at http://dx.doi.org/10.1016/j.cell.2017.07.027. AUTHOR CONTRIBUTIONS S.W. designed and performed the experiments, analyzed the data, and wrote the paper; P.X. performed experiments and analyzed data; Y.C., Y.Q., Z.X., B.Y., and Y.D. performed the experiments; Y.T, Z.Y., and Z.X. provided the genetic tool mice. Z.F. initiated the study and organized, designed, and wrote the paper. ACKNOWLEDGMENTS We thank Junying Jia, Lei Sun, Yan Teng, Shu Meng, Zhonglin Fu, Yapu Zhao, and Feng Zhang for technical support. We thank Drs. Flavell (Yale University), Yuan Zhuang (Duke University), and Chen Dong (Tsinghua University) for providing the genetic mouse stains. We thank Jing Li (Cnkingbio Company) for technical support. This work was supported by the National Natural Science Foundation of China (91640203, 31530093, 81330047, 81601361, 31471386, 31570872, and 31671531), the Strategic Priority Research Programs of the Chinese Academy of Sciences (XDB19030203 and XDA12020219), the Youth Innovation Promotion Association of CAS (2015073) (to S.W.), and the China Postdoctoral Science Foundation (2015M571141) (to P.X.).

14 Cell 171, 1–16, September 21, 2017

REFERENCES Artis, D., and Spits, H. (2015). The biology of innate lymphoid cells. Nature 517, 293–301.

Bedoui, S., Gebhardt, T., Gasteiger, G., and Kastenmu¨ller, W. (2016). Parallels and differences between innate and adaptive lymphocytes. Nat. Immunol. 17, 490–494. Brandenburg, S., Takahashi, T., de la Rosa, M., Janke, M., Karsten, G., Muzzulini, T., Orinska, Z., Bulfone-Paus, S., and Scheffold, A. (2008). IL-2 induces in vivo suppression by CD4(+)CD25(+)Foxp3(+) regulatory T cells. Eur. J. Immunol. 38, 1643–1653. Brestoff, J.R., Kim, B.S., Saenz, S.A., Stine, R.R., Monticelli, L.A., Sonnenberg, G.F., Thome, J.J., Farber, D.L., Lutfy, K., Seale, P., and Artis, D. (2015). Group 2 innate lymphoid cells promote beiging of white adipose tissue and limit obesity. Nature 519, 242–246. Buonocore, S., Ahern, P.P., Uhlig, H.H., Ivanov, I.I., Littman, D.R., Maloy, K.J., and Powrie, F. (2010). Innate lymphoid cells drive interleukin-23-dependent innate intestinal pathology. Nature 464, 1371–1375. Chang, Y.J., Kim, H.Y., Albacker, L.A., Baumgarth, N., McKenzie, A.N., Smith, D.E., Dekruyff, R.H., and Umetsu, D.T. (2011). Innate lymphoid cells mediate influenza-induced airway hyper-reactivity independently of adaptive immunity. Nat. Immunol. 12, 631–638. Coccia, M., Harrison, O.J., Schiering, C., Asquith, M.J., Becher, B., Powrie, F., and Maloy, K.J. (2012). IL-1b mediates chronic intestinal inflammation by promoting the accumulation of IL-17A secreting innate lymphoid cells and CD4(+) Th17 cells. J. Exp. Med. 209, 1595–1609. Constantinides, M.G., McDonald, B.D., Verhoef, P.A., and Bendelac, A. (2014). A committed precursor to innate lymphoid cells. Nature 508, 397–401. Cortez, V.S., Cervantes-Barragan, L., Robinette, M.L., Bando, J.K., Wang, Y., Geiger, T.L., Gilfillan, S., Fuchs, A., Vivier, E., Sun, J.C., et al. (2016). Transforming growth factor-beta signaling guides the differentiation of innate lymphoid cells in salivary glands. Immunity 44, 1127–1139. Curotto de Lafaille, M.A., and Lafaille, J.J. (2009). Natural and adaptive foxp3+ regulatory T cells: more of the same or a division of labor? Immunity 30, 626–635. De Obaldia, M.E., and Bhandoola, A. (2015). Transcriptional regulation of innate and adaptive lymphocyte lineages. Annu. Rev. Immunol. 33, 607–642. Diefenbach, A., Colonna, M., and Koyasu, S. (2014). Development, differentiation, and diversity of innate lymphoid cells. Immunity 41, 354–365. Eberl, G., Colonna, M., Di Santo, J.P., and McKenzie, A.N. (2015). Innate lymphoid cells. Innate lymphoid cells: a new paradigm in immunology. Science 348, aaa6566. Franke, A., Balschun, T., Karlsen, T.H., Hedderich, J., May, S., Lu, T., Schuldt, D., Nikolaus, S., Rosenstiel, P., Krawczak, M., and Schreiber, S. (2008). Replication of signals from recent studies of Crohn’s disease identifies previously unknown disease loci for ulcerative colitis. Nat. Genet. 40, 713–715. Franke, A., McGovern, D.P., Barrett, J.C., Wang, K., Radford-Smith, G.L., Ahmad, T., Lees, C.W., Balschun, T., Lee, J., Roberts, R., et al. (2010). Genomewide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci. Nat. Genet. 42, 1118–1125. Fuchs, A., Vermi, W., Lee, J.S., Lonardi, S., Gilfillan, S., Newberry, R.D., Cella, M., and Colonna, M. (2013). Intraepithelial type 1 innate lymphoid cells are a unique subset of IL-12- and IL-15-responsive IFN-g-producing cells. Immunity 38, 769–781.

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Gagliani, N., Amezcua Vesely, M.C., Iseppon, A., Brockmann, L., Xu, H., Palm, N.W., de Zoete, M.R., Licona-Limo´n, P., Paiva, R.S., Ching, T., et al. (2015). Th17 cells transdifferentiate into regulatory T cells during resolution of inflammation. Nature 523, 221–225.

Miyazaki, M., Miyazaki, K., Chen, S., Itoi, M., Miller, M., Lu, L.F., Varki, N., Chang, A.N., Broide, D.H., and Murre, C. (2014). Id2 and Id3 maintain the regulatory T cell pool to suppress inflammatory disease. Nat. Immunol. 15, 767–776.

Garrett, W.S., Lord, G.M., Punit, S., Lugo-Villarino, G., Mazmanian, S.K., Ito, S., Glickman, J.N., and Glimcher, L.H. (2007). Communicable ulcerative colitis induced by T-bet deficiency in the innate immune system. Cell 131, 33–45.

Montaldo, E., Teixeira-Alves, L.G., Glatzer, T., Durek, P., Stervbo, U., Hamann, W., Babic, M., Paclik, D., Sto¨lzel, K., Gro¨ne, J., et al. (2014). Human RORgt(+) CD34(+) cells are lineage-specified progenitors of group 3 RORgt(+) innate lymphoid cells. Immunity 41, 988–1000.

Gasteiger, G., and Rudensky, A.Y. (2014). Interactions between innate and adaptive lymphocytes. Nat. Rev. Immunol. 14, 631–639. Gordon, S.M., Chaix, J., Rupp, L.J., Wu, J., Madera, S., Sun, J.C., Lindsten, T., and Reiner, S.L. (2012). The transcription factors T-bet and Eomes control key checkpoints of natural killer cell maturation. Immunity 36, 55–67. Guo, X., Liang, Y., Zhang, Y., Lasorella, A., Kee, B.L., and Fu, Y.X. (2015). Innate lymphoid cells control early colonization resistance against intestinal pathogens through ID2-dependent regulation of the microbiota. Immunity 42, 731–743. Gury-BenAri, M., Thaiss, C.A., Serafini, N., Winter, D.R., Giladi, A., Lara-Astiaso, D., Levy, M., Salame, T.M., Weiner, A., David, E., et al. (2016). The spectrum and regulatory landscape of intestinal innate lymphoid cells are shaped by the microbiome. Cell 166, 1231–1246. Ishizuka, I.E., Constantinides, M.G., Gudjonson, H., and Bendelac, A. (2016). The innate lymphoid cell precursor. Annu. Rev. Immunol. 34, 299–316. Izcue, A., Coombes, J.L., and Powrie, F. (2009). Regulatory lymphocytes and intestinal inflammation. Annu. Rev. Immunol. 27, 313–338. Josefowicz, S.Z., Lu, L.F., and Rudensky, A.Y. (2012). Regulatory T cells: mechanisms of differentiation and function. Annu. Rev. Immunol. 30, 531–564. Kamanaka, M., Kim, S.T., Wan, Y.Y., Sutterwala, F.S., Lara-Tejero, M., Gala´n, J.E., Harhaj, E., and Flavell, R.A. (2006). Expression of interleukin-10 in intestinal lymphocytes detected by an interleukin-10 reporter knockin tiger mouse. Immunity 25, 941–952. Kang, J.S., Liu, C., and Derynck, R. (2009). New regulatory mechanisms of TGF-beta receptor function. Trends Cell Biol. 19, 385–394. Kaser, A., Zeissig, S., and Blumberg, R.S. (2010). Inflammatory bowel disease. Annu. Rev. Immunol. 28, 573–621. Kirchberger, S., Royston, D.J., Boulard, O., Thornton, E., Franchini, F., Szabady, R.L., Harrison, O., and Powrie, F. (2013). Innate lymphoid cells sustain colon cancer through production of interleukin-22 in a mouse model. J. Exp. Med. 210, 917–931. Klose, C.S., Kiss, E.A., Schwierzeck, V., Ebert, K., Hoyler, T., d’Hargues, Y., Go¨ppert, N., Croxford, A.L., Waisman, A., Tanriver, Y., and Diefenbach, A. (2013). A T-bet gradient controls the fate and function of CCR6-RORgt+ innate lymphoid cells. Nature 494, 261–265. Klose, C.S.N., Flach, M., Mo¨hle, L., Rogell, L., Hoyler, T., Ebert, K., Fabiunke, C., Pfeifer, D., Sexl, V., Fonseca-Pereira, D., et al. (2014). Differentiation of type 1 ILCs from a common progenitor to all helper-like innate lymphoid cell lineages. Cell 157, 340–356. Ku¨hn, R., Lo¨hler, J., Rennick, D., Rajewsky, K., and Mu¨ller, W. (1993). Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75, 263–274. Li, H., Dai, M., and Zhuang, Y. (2004). A T cell intrinsic role of Id3 in a mouse model for primary Sjogren’s syndrome. Immunity 21, 551–560. Maloy, K.J., Salaun, L., Cahill, R., Dougan, G., Saunders, N.J., and Powrie, F. (2003). CD4+CD25+ T(R) cells suppress innate immune pathology through cytokine-dependent mechanisms. J. Exp. Med. 197, 111–119. Maynard, C.L., Harrington, L.E., Janowski, K.M., Oliver, J.R., Zindl, C.L., Rudensky, A.Y., and Weaver, C.T. (2007). Regulatory T cells expressing interleukin 10 develop from Foxp3+ and Foxp3- precursor cells in the absence of interleukin 10. Nat. Immunol. 8, 931–941. McHedlidze, T., Waldner, M., Zopf, S., Walker, J., Rankin, A.L., Schuchmann, M., Voehringer, D., McKenzie, A.N., Neurath, M.F., Pflanz, S., and Wirtz, S. (2013). Interleukin-33-dependent innate lymphoid cells mediate hepatic fibrosis. Immunity 39, 357–371.

Monticelli, L.A., Buck, M.D., Flamar, A.L., Saenz, S.A., Tait Wojno, E.D., Yudanin, N.A., Osborne, L.C., Hepworth, M.R., Tran, S.V., Rodewald, H.R., et al. (2016). Arginase 1 is an innate lymphoid-cell-intrinsic metabolic checkpoint controlling type 2 inflammation. Nat. Immunol. 17, 656–665. Moro, K., Yamada, T., Tanabe, M., Takeuchi, T., Ikawa, T., Kawamoto, H., Furusawa, J., Ohtani, M., Fujii, H., and Koyasu, S. (2010). Innate production of T(H)2 cytokines by adipose tissue-associated c-Kit(+)Sca-1(+) lymphoid cells. Nature 463, 540–544. Munneke, J.M., Bjo¨rklund, A.T., Mjo¨sberg, J.M., Garming-Legert, K., Bernink, J.H., Blom, B., Huisman, C., van Oers, M.H., Spits, H., Malmberg, K.J., and Hazenberg, M.D. (2014). Activated innate lymphoid cells are associated with a reduced susceptibility to graft-versus-host disease. Blood 124, 812–821. Powell, N., Walker, A.W., Stolarczyk, E., Canavan, J.B., Go¨kmen, M.R., Marks, E., Jackson, I., Hashim, A., Curtis, M.A., Jenner, R.G., et al. (2012). The transcription factor T-bet regulates intestinal inflammation mediated by interleukin-7 receptor+ innate lymphoid cells. Immunity 37, 674–684. Rankin, L.C., Groom, J.R., Chopin, M., Herold, M.J., Walker, J.A., Mielke, L.A., McKenzie, A.N., Carotta, S., Nutt, S.L., and Belz, G.T. (2013). The transcription factor T-bet is essential for the development of NKp46+ innate lymphocytes via the Notch pathway. Nat. Immunol. 14, 389–395. Rivera, R.R., Johns, C.P., Quan, J., Johnson, R.S., and Murre, C. (2000). Thymocyte selection is regulated by the helix-loop-helix inhibitor protein, Id3. Immunity 12, 17–26. Robinette, M.L., Fuchs, A., Cortez, V.S., Lee, J.S., Wang, Y., Durum, S.K., Gilfillan, S., and Colonna, M.; Immunological Genome Consortium (2015). Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets. Nat. Immunol. 16, 306–317. Sakaguchi, S. (2005). Naturally arising Foxp3-expressing CD25+CD4+ regulatory T cells in immunological tolerance to self and non-self. Nat. Immunol. 6, 345–352. Sakaguchi, S., Yamaguchi, T., Nomura, T., and Ono, M. (2008). Regulatory T cells and immune tolerance. Cell 133, 775–787. Saraiva, M., and O’Garra, A. (2010). The regulation of IL-10 production by immune cells. Nat. Rev. Immunol. 10, 170–181. Serafini, N., Vosshenrich, C.A., and Di Santo, J.P. (2015). Transcriptional regulation of innate lymphoid cell fate. Nat. Rev. Immunol. 15, 415–428. Shih, H.Y., Sciume`, G., Mikami, Y., Guo, L., Sun, H.W., Brooks, S.R., Urban, J.F., Jr., Davis, F.P., Kanno, Y., and O’Shea, J.J. (2016). Developmental acquisition of regulomes underlies innate lymphoid cell functionality. Cell 165, 1120–1133. Sonnenberg, G.F., and Artis, D. (2015). Innate lymphoid cells in the initiation, regulation and resolution of inflammation. Nat. Med. 21, 698–708. Spits, H., and Cupedo, T. (2012). Innate lymphoid cells: emerging insights in development, lineage relationships, and function. Annu. Rev. Immunol. 30, 647–675. Spooner, C.J., Lesch, J., Yan, D., Khan, A.A., Abbas, A., Ramirez-Carrozzi, V., Zhou, M., Soriano, R., Eastham-Anderson, J., Diehl, L., et al. (2013). Specification of type 2 innate lymphocytes by the transcriptional determinant Gfi1. Nat. Immunol. 14, 1229–1236. Strober, W., Fuss, I.J., and Blumberg, R.S. (2002). The immunology of mucosal models of inflammation. Annu. Rev. Immunol. 20, 495–549. Sugimoto, K., Ogawa, A., Mizoguchi, E., Shimomura, Y., Andoh, A., Bhan, A.K., Blumberg, R.S., Xavier, R.J., and Mizoguchi, A. (2008). IL-22 ameliorates intestinal inflammation in a mouse model of ulcerative colitis. J. Clin. Invest. 118, 534–544.

Cell 171, 1–16, September 21, 2017 15

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Thornton, A.M., and Shevach, E.M. (1998). CD4+CD25+ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production. J. Exp. Med. 188, 287–296. Travis, M.A., and Sheppard, D. (2014). TGF-b activation and function in immunity. Annu. Rev. Immunol. 32, 51–82.

Wirtz, S., Neufert, C., Weigmann, B., and Neurath, M.F. (2007). Chemically induced mouse models of intestinal inflammation. Nat. Protoc. 2, 541–546. Wu, M.Y., and Hill, C.S. (2009). Tgf-beta superfamily signaling in embryonic development and homeostasis. Dev. Cell 16, 329–343.

Vignali, D.A., Collison, L.W., and Workman, C.J. (2008). How regulatory T cells work. Nat. Rev. Immunol. 8, 523–532.

Xia, P., Wang, S., Huang, G., Zhu, P., Li, M., Ye, B., Du, Y., and Fan, Z. (2014). WASH is required for the differentiation commitment of hematopoietic stem cells in a c-Myc-dependent manner. J. Exp. Med. 211, 2119–2134.

Wang, X., Ota, N., Manzanillo, P., Kates, L., Zavala-Solorio, J., Eidenschenk, C., Zhang, J., Lesch, J., Lee, W.P., Ross, J., et al. (2014). Interleukin-22 alleviates metabolic disorders and restores mucosal immunity in diabetes. Nature 514, 237–241.

Xia, P., Ye, B., Wang, S., Zhu, X., Du, Y., Xiong, Z., Tian, Y., and Fan, Z. (2016). Glutamylation of the DNA sensor cGAS regulates its binding and synthase activity in antiviral immunity. Nat. Immunol. 17, 369–378.

Wang, S., Xia, P., Chen, Y., Huang, G., Xiong, Z., Liu, J., Li, C., Ye, B., Du, Y., and Fan, Z. (2016a). Natural killer-like B cells prime innate lymphocytes against microbial infection. Immunity 45, 131–144.

Yang, Q., Li, F., Harly, C., Xing, S., Ye, L., Xia, X., Wang, H., Wang, X., Yu, S., Zhou, X., et al. (2015). TCF-1 upregulation identifies early innate lymphoid progenitors in the bone marrow. Nat. Immunol. 16, 1044–1050.

Wang, S., Xia, P., Huang, G., Zhu, P., Liu, J., Ye, B., Du, Y., and Fan, Z. (2016b). FoxO1-mediated autophagy is required for NK cell development and innate immunity. Nat. Commun. 7, 11023.

Zheng, Y., Danilenko, D.M., Valdez, P., Kasman, I., Eastham-Anderson, J., Wu, J., and Ouyang, W. (2007). Interleukin-22, a T(H)17 cytokine, mediates IL-23induced dermal inflammation and acanthosis. Nature 445, 648–651.

16 Cell 171, 1–16, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

STAR+METHODS KEY RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Antibodies Anti-mouse CD3 (17A2)

eBioscience

Cat# 17-0032-82 Cat# 16-0032-86

Anti- mouse CD4 (GK1.5)

eBioscience

Cat# 17-0041-81

Anti- mouse CD19 (1D3)

eBioscience

Cat# 17-0193-82

Anti- mouse NK1.1 (PK136)

eBioscience

Cat# 17-5941-82

Anti- mouse CD11b (M1/70)

eBioscience

Cat# 17-0112-81

Anti- mouse CD11c (N418)

eBioscience

Cat# 17-0114-81

Anti- mouse Gr1 (RB6-8C5)

eBioscience

Cat# 17-5931-81

Anti- mouse F4/80 (BM8)

eBioscience

Cat# 17-4801-80

Anti- mouse Ter119 (TER-119)

eBioscience

Cat# 17-5921-81

Anti- mouse CD45 (30-F11)

eBioscience

Cat# 11-0451-82 Cat# 48-0451-82

Anti- mouse CD127 (SB/199)

eBioscience

Cat# 46-1273-82

Anti- mouse IL-10 (JES5-16E3)

eBioscience

Cat# 12-7101-82

Anti- mouse CD25 (PC61.5)

eBioscience

Cat# 25-0251-81

Anti- mouse CD90 (HIS51)

eBioscience

Cat# 25-0900-82

Anti- mouse CD117 (2B8)

eBioscience

Cat# 25-1171-81

Anti- mouse NKp46 (29A1.4)

eBioscience

Cat# 12-3351-80

Anti- mouse ST2 (RMST2-33)

eBioscience

Cat# 12-9333-80

Anti- mouse KLRG1 (2F1)

eBioscience

Cat# 12-5893-82

Anti- mouse/human RORgt (AFKJS-9)

eBioscience

Cat# 12-6988-80

Anti- mouse CD122 (TM-b1)

eBioscience

Cat# 25-1222-80

Anti- mouse IL-5 (TRFK5)

eBioscience

Cat# 12-7052-81

Anti- mouse IFN-g (XMG1.2)

eBioscience

Cat# 12-7311-81

Anti- mouse IL-13 (eBio13A)

eBioscience

Cat# 25-7133-80

Anti- mouse IL-22 (IL22JOP)

eBioscience

Cat# 46-7222-80

Anti- mouse IL-17A (eBio17B7)

eBioscience

Cat# 25-7177-80

Anti- mouse TGF-b1 (TW7-16B4)

eBioscience

Cat# 25-9821-82

Anti- mouse CD69 (H1.2F3)

eBioscience

Cat# 47-0691-80

Anti- mouse Ki67 (SolA15)

eBioscience

Cat# 48-5698-80

Anti- mouse Gata3 (TWAJ)

eBioscience

Cat# 25-9966-41

Anti- mouse T-bet (eBio4B10)

eBioscience

Cat# 25-5825-80

Anti- mouse EOMES (Dan11mag)

eBioscience

Cat# 25-4875-80

Anti- mouse FoxP3 (FJK-16 s)

eBioscience

Cat# 25-5773-80

Anti- mouse IL-10RA (1B1.3a)

Biolegend

Cat# 112705

Anti- mouse Id3 (S30-778)

BD PharMingen

Cat# 564564

Anti- mouse TGF-bRI

Abcam

Cat# ab31013

Anti- mouse TGF-bRII

Abcam

Cat# ab186838

Anti- mouse Id2 (4E12G5)

Abcam

Cat#ab166708

Anti-human IL-10 (JES3-9D7)

eBioscience

Cat# 12-7108-81

Anti-human CD127 (eBioRDR5)

eBioscience

Cat# 25-1278-41 (Continued on next page)

Cell 171, 1–16.e1–e6, September 21, 2017 e1

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

Anti-human CD45 (2D1)

eBioscience

Cat# 11-9459-41

Human Hematopoietic Lineage Monoclonal Antibody

eBioscience

Cat# 22-7776-72

Salmonella (S.) Typhimurium

Institute of Microbiology, Chinese Academy of Sciences

Cat# 1.1194

Citrobacter (C.) Rodentium

Yan et al. Nat Immunol. 2012

strain DBS100

Lentivirus-lentiGuide-Puro-sgRNA-Il10

This paper

N/A

Normal human resected small intestine tissues

West China Hospital, Sichuan University, Chengdu, China

N/A

Normal human colon tissues

West China Hospital, Sichuan University, Chengdu, China

N/A

Recombinant mouse IL-2

PeproTech

121-12

Recombinant mouse IL-4

PeproTech

214-14

Recombinant mouse IL-7

PeproTech

217-17

Recombinant mouse IL-9

PeproTech

219-19

Recombinant mouse IL-10

PeproTech

210-10

Recombinant mouse IL-12

PeproTech

210-12

Recombinant mouse IL-15

PeproTech

210-15

Recombinant mouse IL-21

PeproTech

210-21

Recombinant mouse IL-33

PeproTech

210-33

Recombinant mouse TGF-b1

R&D Systems

P04202

Brefeldin A (BFA)

eBioscience

00-4506-51

7-AAD Viability Staining Solution

eBioscience

00-6993-50

Dextran sodium sulfate (DSS) (colitis grade)

MP Biomedicals

0216011080

Tamoxifen (TMX)

Sigma-Aldrich

T5648

4-Hydroxytamoxifen (4-HT)

Sigma-Aldrich

H6278

4’,6-diamidino-2-phenylindole (DAPI)

Sigma-Aldrich

D8417

Diphtheria toxin (DT)

Merck Calbiochem

322326

Bacterial and Virus Strains

Biological Samples

Chemicals, Peptides, and Recombinant Proteins

Critical Commercial Assays IFN gamma Mouse Uncoated ELISA Kit

Invitrogen

88-7314-88

IL-17A Mouse Uncoated ELISA Kit

Invitrogen

88-7371-88

IL-22 Mouse Uncoated ELISA Kit

Invitrogen

88-7422-88

IL-5 Mouse Uncoated ELISA Kit

Invitrogen

88-7054-88

IL-13 Mouse Uncoated ELISA Kit

Invitrogen

88-7137-77

TGF beta-1 Human/Mouse Uncoated ELISA Kit

Invitrogen

88-8350-88

Intracellular Fixation and Permeabilization Buffer Set

eBioscience

88-8824-00

Foxp3 / Transcription Factor Staining Buffer Set

eBioscience

00-5523-00

Microarray data of ILCregs

This paper

GEO:GSE101440

Microarray data of mouse ILCs

NCBI GEO DataSets

GEO:GSE37448

Microarray data of mouse Tregs

NCBI GEO DataSets

GEO:GSE68009

Microarray data of human ILCs

NCBI GEO DataSets

GEO: GSE63197; GEO: GSE78896; GEO: GSE45458

Mouse: C57BL/6

Beijing Vital River Laboratory Animal Technology, Beijing, China

Cat#213

Mouse: B6.129P2-Il10tm1Cgn/ NJU

Model Animal Research Center of Nanjing University, China

J002251

Deposited Data

Experimental Models: Organisms/Strains

(Continued on next page)

e2 Cell 171, 1–16.e1–e6, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

Model Animal Research Center of Nanjing University, China

J002216

Mouse: B6.Cg-Tg(Lck-cre)548Jxm/NJU

Model Animal Research Center of Nanjing University, China

J003802

Mouse: B6.129P2-Lyz2tm1(cre)Ifo/NJU

Model Animal Research Center of Nanjing University, China

J004781

Mouse: B6.129S7-Rag1

tmiMom

/ NJU

Mouse: B6.129S(Cg)-Id2tm1.1(cre/ERT2)Blh/ZhuJ

The Jackson Laboratory

016222

Mouse: B6(SJL)-Zbtb16tm1.1(EGFP/cre)Aben/J

The Jackson Laboratory

024529

Mouse: B6.129S(Cg)-Id2tm2.1Blh/ZhuJ

The Jackson Laboratory

016224

The Jackson Laboratory

003174

Mouse: B6.129S4-Il2rg

tm1Wjl

/J

Mouse: B6.FVB-Tg(Rorc-cre)1Litt/J

The Jackson Laboratory

022791

Mouse: B6.129(Cg)-Foxp3tm3(DTR/GFP)Ayr/J

The Jackson Laboratory

016958

Mouse: C57BL/6-Foxp3tm1Flv/J

The Jackson Laboratory

008374

Mouse: C57BL/6-Gt(ROSA)26Sortm1(HBEGF)Awai/J

The Jackson Laboratory

007900

Mouse: B6.Cg-Tg(UBC-cre/ERT2)1Ejb/2J

The Jackson Laboratory

008085

Mouse: B6(SJL)-Il10ratm1.1Tlg/J

The Jackson Laboratory

028146

Mouse: B6;129-Gt(ROSA)26Sortm1(CAG-cas9*,-EGFP)Fezh/J

The Jackson Laboratory

024857

Mouse: STOCK Tgfb1tm2.1Doe/J

The Jackson Laboratory

010721

Mouse: B6.129X1-Gt(ROSA)26Sortm1(EYFP)Cos/J

Shanghai Research Center For Model Organisms, China

SJ-006148

Mouse: STOCK-Tg(Cd4-cre)1Cwi/BfluJ

Shanghai Research Center For Model Organisms, China

SJ-022071

Mouse: B6.129S6-Il10tm1Flv/J

From Prof. Richard A. Flavell (Yale University)

N/A

Mouse: Id3flox/flox

From Profs. Yuan Zhuang (Duke University) and Chen Dong (Tsinghua University)

N/A

Mouse: Tgfbr2flox/flox

From Prof. Zhiheng Xu (Institute of Genetics and Developmental Biology, CAS)

N/A

Oligonucleotides Real-time PCR primers for Figures 4A and 4B

This paper

Table S1

Id3-F: 50 -CTCTATCTCTACTCTCCAACAT-30

This paper

N/A

Id3-R:50 -GTCGTCCAAGAGGCTAAG-30

This paper

N/A

sgRNA-Il10#1: 50 -TATTGTCTTCCCGGCTGTAC-30

This paper

N/A

sgRNA-Il10#2: 50 -GCATGTGGCTCTGGCCGACT-30

This paper

N/A

lentiGuide-Puro-sgRNA-Il10

This paper

N/A

pVSVg

AddGene

8454

psPAX2

AddGene

12260

Prim7

GraphPad

https://www.graphpad.com/ scientific-software/prism/

SigmaPlot 12.5

Systat Software

https://systatsoftware.com/ downloads/download-sigmaplot/

FlowJo 7.6.1

FlowJo

N/A

Recombinant DNA

Software and Algorithms

(Continued on next page)

Cell 171, 1–16.e1–e6, September 21, 2017 e3

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

LIMMA

Bioconductor

https://bioconductor.org/packages/ release/bioc/html/limma.html

IDEAS software

Merck

https://www.merckmillipore.com

Other BD FACS Aria III

BD Biosciences

N/A

Amnis ImageStream MakII

Merck

N/A

Tecnai spirit transmission electron microscope

FEI

N/A

Olympus FV1000 confocal microscope

Olympus

N/A

CONTACT FOR REAGENT AND RESOURCE SHARING Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Zusen Fan ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS Mice and human tissues Rag1
/
Il10
/
mice were obtained by crossing Rag1+/
mice with Il10+/
mice. Rag1
/
Il2rg
/
mice were generated by crossing Rag+/
mice with Il2rg+/
mice. All the mice are C57BL/6 background and maintained under specific pathogen-free conditions with approval by the Institutional Committee of Institute of Biophysics, Chinese Academy of Sciences. Both female and male mice were used in experiments. Age- and sex-matched littermates between 8 and 28 weeks of age were used. Exclusion criteria such as inadequate staining were pre-established. Mice were assigned randomly to experimental groups. To generate inducible TGF-b1 deletion in ILCregs, Tgfb1flox/+ mice were crossed with UBC-CreERT2 (CreERT2) mice or IL-10-GFP mice. Tgfb1flox/+;CreERT2 mice were crossed with IL-10-GFP;Tgfb1flox/+ mice to obtain IL-10-GFP;Tgfb1flox/flox;CreERT2 mice. IL-10-GFP;Tgfbr2flox/flox;CreERT2 mice were generated with a similar strategy. To generate ILCregs depleted mice, Lck-Cre mice were crossed with Rosa26-STOP-DTR (RosaDTR) mice to obtain Lck-Cre;RosaDTR mice. These Lck-Cre;RosaDTR mice were crossed with IL-10-GFP mice to generate Lck-Cre;RosaDTR;IL-10-GFP mice. ILCregs were isolated from Lck-Cre;RosaDTR;IL-10-GFP mice and transferred together with T, B, NK and ILC cells. After 7 days, engrafted cells were confirmed by flow cytometry. Transferred mice were administrated with 100 ng/mouse diphtheria toxin (DT) for every three days and treated with 3% DSS followed by further analysis. Human resected small intestine and colon tissues were obtained from West China Hospital, Sichuan University (Chengdu, China) with informed consent, according to the Institutional Review Board (IRB)–approved protocol. Generation of inflammation mouse models For DSS-induced colitis, Rag1
/
mice were treated with 3% DSS in drinking water for the indicated days. For anti-CD40-induced colitis, Rag1
/
mice were injected with 150 mg anti-CD40 antibody (clone:FGK45, i.p.) on day 0. For bacteria-induced-colitis, Rag1
/
mice were infected by oral gavage of 5x104 c.f.u. S. Typhimurium (from Institute of Microbiology, Chinese Academy of Sciences) or 1x109 c.f.u. C. Rodentium (from Prof. Baoxue Ge, Chinese Academy of Sciences) on day 0. Mice were monitored for indicated days followed by further analysis. METHOD DETAILS Isolation of ILCreg cells Intestines from IL-10-GFP mice were cut open longitudinally and Peyer’s patches were removed. Next, intestines were cleaned and cut into pieces. Epithelial layers were removed by incubation three times in 5 mM EDTA Ca2+ and Mg2+ free Hank’s medium for 20 min each at 37 C, and the epithelial cells were collected if needed. Then, intestines were cut into fine pieces and digested twice for 40 min each at 37 C with Collagenase II and III (1 mg/ml; Worthington), DNase I (200 mg/ml; Roche) and dispase (4U/ml; Sigma). Mononuclear cells were isolated with 40%–80% Percol gradient, and washed twice. Lin-CD45+ cells were sorted out by Magnetic Cell Sorting system (MACS, Miltenyi Biotec) Sorted cells were blocked with anti-CD16/32 antibody for 30 min on ice and then stained with antiCD45, anti-CD127, and lineage cocktail antibodies (lineage = CD3, CD4, CD8, CD19, NK1.1, CD11b, CD11c, Gr1, F4/80, and Ter119) on ice for 1 hr. After staining with 7AAD, cells were isolated by flow cytometer (FACS Aria III, BD), and ILCregs (Lin-CD45+CD127+IL10-GFP+7AAD-) were isolated and subjected to functional studies. Purity of ILCreg cells was over 98% for each assay that was determined by post sorting analysis of flow cytometry.

e4 Cell 171, 1–16.e1–e6, September 21, 2017

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Flow cytometry assay Intestinal cells were isolated and blocked with anti-CD16/32 antibody for 30 min on ice. Cells were then stained with surface markers for 1 hr on ice. For intracellular cytokine staining, cells were cultured in media containing indicated cytokines in the presence of Brefeldin A for 4 hr at 37 C. Cells were harvested for surface marker staining and fixed and permeablized by Intracellular Fixation & Permeablization buffer set (eBioscience) after surface marker staining, followed by staining with antibodies against intracellular antigens for flow cytometer (FACS Aria III, BD). We used anti-IL-10-PE antibody (JES5-16E3) to stain intracellular IL-10. Imaging flow cytometry Lymphocytes from mouse intestines were isolated and blocked with anti-CD16/32 antibody for 30 min. Cells were then stained with antibodies against Lin, CD127, CD45 followed by intracellular staining of IL-10. Nuclei were visualized by DAPI staining. Cells were analyzed by imaging flow cytometer (Amnis ImageStream MakII, Merck), and data were analyzed by IDEAS software (Merck) (Wang et al., 2016a). Immunofluorescence assay For in situ immunofluorescence of ILCreg cells, mouse intestines were treated as previously described (Xia et al., 2016). Briefly, intestines were cut open longitudinally and fixed in 4% paraformaldehyde (PFA) (Sigma-Aldrich) fixative for 8 hr, and rehydrated in 30% sucrose solution for 24 hr and frozen in OCT for sectioning. Intestinal sections were rehydrated in PBS and blocked in 10% donkey serum and anti-CD16/32 antibody. We used anti-IL-10 (ab9969, Abcam, Rabbit) and Biotin-conjugated anti-CD127 (A7R34, eBioscience, Rat) as primary antibodies. After staining with anti-IL-10 and Biotin-conjugated anti-CD127 antibodies for 2 hr at room temperature (RT) we used Alex488-conjugated donkey anti-Rabbit and Alex594-conjugated Streptavidin (Invitrogen) as secondary antibodies. Then, we used APC-conjugated lineage cocktails to stain the slides at RT for 1 hr. After washing for 3 times, nuclei were stained by DAPI if needed and the sections were subjected to dehydrating in EtOH gradient: 70%, 85%, 95%, and 100%. Mounted sections were analyzed by confocal microscopy (Olympus FV1000). Adoptive transfer assays ILCregs (Lin-CD45.2+CD127+IL-10-GFP+) were isolated from IL-10-GFP reporter mice (CD45.2). ILC1s (CD45.1+CD127+NK1.1+ NKp46+) were isolated from WT mice (CD45.1) and ILC3s (Lin-CD45.1+CD127+RORgt-GFP+) were isolated from RORgt-GFP reporter mice (CD45.1). ILCs were activated if needed as follows. ILC1s were cultured in media with IL-12 (10 ng/ml) and IL-18 (10 ng/ml) for 24 hr. ILC3s were cultured in media with IL-23 (10 ng/ml) for 24 hr. 5x104 ILCs alone or with 5x104 ILCregs were adoptively transferred into Rag1
/
Il2rg
/
mice. After 7-14 days, transferred mice were analyzed or treated with 3% DSS with or without administration of tamoxinfen (TMX) (50 mg/kg i.p. for five consecutive days) for indicated days for further assays. ELISA assay ILCregs or ILCs from mice with indicated treatment were isolated and cultured in complete media for 24 hr. Supernatants were collected and cytokines were analyzed by ELISA kit (eBioscience) by manufacturer’s instructions. Detection thresholds of different cytokines were 50 pg/ml for IFN-g, 50 pg/ml for IL-17A, 75 pg/ml for IL-22, 50 pg/ml for IL-4, 75 pg/ml for IL-5, 50 pg/ml for IL-13 and 50 pg/ml for TGF-b1. Values below detection thresholds were shown as ND (not detectable). The ELISA data were normalized to total cell numbers and shown as means ± SD per 5x103 cells. Histology and immunohistochemistry assay Mouse colons were cut open longitudinally and fixed in 4% paraformaldehyde (PFA) (Sigma-Aldrich) fixative for 8 hr for 12 hr. Fixed tissues were washed for twice using 75% ethanol and embedded in paraffin, followed by sectioning and staining with hematoxylin and eosin (H&E) according to standard laboratory procedures. Immunohistochemistry assay was performed as previously described (Wang et al., 2016a). Briefly, small intestines from mice were fixed in 4% PFA for 12 hr. Fixed tissues were rehydrated in 30% sucrose solution for 24 hr and frozen in OCT for sectioning. Intestine sections were blocked with H2O2 and anti-CD16/32 antibody and then stained with primary antibodies (anti-Lin, eBioscience, and anti-IL-10 antibodies, Abcam) and subjected to double immunohistochemistry staining with Polymer-HRP and AP Kit from GBI Labs according to the manufacturer’s instructions. Colitis scores and histologic analysis Colons of indicated mice were cut open longitudinally and fixed in 4% PFA followed by paraffin sectioning and H&E staining. A combined colitis score was calculated based on weight loss, appearance of the stool, intestinal bleeding, and histology (Gagliani et al., 2015). Weight loss was analyzed as follows: 0, 0%–4% weight loss or weight gain; 1, 4%–10% weight loss; 2, 10%–15% weight loss; 3, 15%–20% weight loss; 4, more than 20% weight loss. Appearance of the stool was scored as follows: 0, normal; 1, soft but still formed; 2, very soft; 3, diarrhea. Intestinal bleeding was assessed as follows: 0, negative hemoccult; 1, positive hemoccult; 2, blood traces in stool visible; 3, rectal bleeding. Scoring system for inflammation-associated histological changes are as follows: 0, no evidence of inflammation; 1, low level of inflammation with scattered infiltrating mononuclear cells (1–2 foci); 2, moderate inflammation with multiple foci; 3, high level of inflammation with increased vascular density and marked wall thickening; 4, maximal severity of inflammation with transmural leukocyte infiltration and loss of goblet cells.

Cell 171, 1–16.e1–e6, September 21, 2017 e5

Please cite this article in press as: Wang et al., Regulatory Innate Lymphoid Cells Control Innate Intestinal Inflammation, Cell (2017), http:// dx.doi.org/10.1016/j.cell.2017.07.027

Generation of IL-10 deficient ILCregs ILCregs from Rosa26-STOP-Cas9;Id2-CreERT2;IL-10-GFP mice were isolated and infected with lentivirus (lentiGuide-Puro-sgRNAIl10) carrying sgRNA against Il10 gene (#1: 50 -TATTGTCTTCCCGGCTGTAC-30 ; #2: 50 -GCATGTGGCTCTGGCCGACT-30 ). Infected cells were selected with 1 mg/ml puromycin (Invivogen) and transferred into Rag1
/
Il2rg
/
mice together with respective WT (CD45.1) lymphocytes. IL-10 expression in ILCregs (CD45.2) after treatment with 0.5 mm 4-hydroxytamoxinfen (4-HT) was detected prior to transplantation. Electron microscopy For transmission electron microscopy (TEM), ILCregs were isolated from IL-10-GFP mice and treated as previously described (Wang et al., 2016b). In brief, cells were fixed with 2.5% glutaraldehyde on ice for 2 hr followed by fixation in 2% osmium tetroxide. Briefly, cells were immersed in EPON812 resin after dehydrating with sequential washes in 50, 70, 90, 95 and 100% ethanol. Ultrathin sections were collected on copper grids and counterstained using uranyl acetate and lead citrate. Images were taken with a Tecnai spirit transmission electron microscope (FEI). Real-time PCR assay ILCreg cells were isolated from treated mice, mRNA was extracted by using Neasy Micro Kit (QIAGEN) according to the manufacturer’s instructions. mRNA quality was determined by A260/A280 (between 1.8 and 2.2) and A260/230 (> 1.7) ratios. RNA integrity was assessed by agarose gel electrophoresis. cDNAs were synthesized by using oligo dT followed by real-time PCR. Primers for real-time PCR in this study are as shown in Table S1. Microarray analysis Live ILCregs from IL-10-GFP mice were isolated by Magnetic Cell Sorting system (MACS, Miltenyi Biotec) and flow cytometer (FACS Aria III, BD). Purity of cells was over 95% that was determined by post sorting analysis with flow cytometry. Total RNA was extracted from ILCregs using standard RNA extraction protocol, followed by DNase incubation to remove nuclear DNA. RNA quality was monitored by NanoDrop ND-1000 and RNA integrity was tested by agarose gel electrophoresis. Total RNA was amplified with Whole Transcript (WT) Pico Reagent Kit (Affymetrix). Microarray assay was performed as described previously (Xia et al., 2014). Briefly, total RNA was subjected to labeling and array hybridization using GeneChip Mouse Transcriptome Array 1.0 according to manufacturer’s instructions (Affymetrix). Hybridized GeneChips were washed and stained in the Affymetrix Fluidics Station 450. GeneChips were scanned by using Affymetrix GeneChip Command Console (AGCC) that installed in GeneChip Scanner 3000 7G. Data were analyzed with Robust Multichip Analysis (RMA) algorithm using Affymetrix default analysis settings. Values presented are log2 RMA signal intensity. For human ILCregs, lymphocytes were isolated from human normal intestinal tissues from resected small intestines and Lin-CD45+ cells were then enriched by MACS (Lin = CD2, CD3, CD14, CD16, CD19, CD56, CD235a). Enriched cells were stained with Lineage cocktail, anti-CD45, and anti-CD127 antibodies followed by intracellular staining of IL-10. Lin-CD45+CD127+IL-10+ cells were isolated and subjected to RNA extraction with RNeasy FFPE Kit (QIAGEN). Total RNA was amplified with Whole Transcript (WT) Pico Reagent Kit (Affymetrix). Microarray assay was performed by using GeneChip Human Transcriptome Array 2.0 according to manufacturer’s instructions (Affymetrix). QUANTIFICATION AND STATISTICAL ANALYSIS Statistical analysis For statistical analysis, data were analyzed by using Sigma Plot or GraphPad Prism 7.0. Two-tailed unpaired Student’s t test, Non parametric Mann–Whitney U-test or One way ANOVA were used according to the type of experiments. Kaplan-Meier method was used for comparison of survival probabilities. P-values % 0.05 were considered significant (*, p < 0.05; **, p < 0.01; ***, p < 0.001); p > 0.05, non-significant (NS). All flow cytometry data were analyzed with FlowJo (Treestar). No statistical methods were used to predetermine sample size. DATA AND SOFTWARE AVAILABILITY The accession number for microarray data reported in this paper are GEO: mouse ILCs (GEO: GSE37448), mouse Tregs (GEO: GSE68009), human ILCs (GEO: GSE63197, GEO: GSE78896, GEO: GSE45458), and ILCregs (GEO: GSE101440).

e6 Cell 171, 1–16.e1–e6, September 21, 2017

Supplemental Figures

Figure S1. Identification of ILCregs in Mouse and Human Intestines, Related to Figures 1 and 2 (A) Gate strategy of ILCreg identification. Lamina propria cells from IL-10-GFP mice were isolated and blocked with anti-CD16/32 antibody followed by surface marker staining with antibodies against Lin, CD45, and CD127. Lymphocytes were gated out by FSC-A and SSC-A. Unconjugated cells were gated out by FSC-W and SSC-W. Lin CD45+ lymphocytes were gated out for CD127 versus IL-10-GFP analysis. (Lin = CD3e,CD4,CD8,CD19,NK1.1,CD11b,CD11c,Gr1,Ter119). (B) Analysis of Tregs and ILCregs in mouse intestines after treatment with anti-CD3 monoclonal antibody. Mouse lamina propria cells were isolated and stained with the indicated antibodies. CD127+IL-10+CD4 FoxP3 population was ILCregs, and CD127+IL-10+CD4+FoxP3+ population was Tregs. (C) Flow cytometry analysis of lineage markers on ILCregs. Gray histograms depict isotype control of each antibody. (D) Immunofluorescence staining of ILCregs in mouse colons. Scale bar, 50 mm. White arrowhead indicates ILCregs.

(legend continued on next page)

(E) Immunofluorescence staining of ILCregs in human colon. Scale bar, 50 mm. White arrowhead indicates ILCregs. (F) Flow cytometry analysis of markers on ILCregs. Gray histograms depict isotype control of each antibody. (G) Flow cytometry analysis of cytokines and transcription factors of ILCregs. Blue lines indicated positive controls. For TNFa and IFN-g, positive controls are ILC1s cultured in 10ng/ml IL-12/IL-18 for 24 hr. For IL-5 and IL-13, positive controls are ILC2s cultured in 10ng/ml IL-33 for 24 hr. For IL-17A and IL-22, positive controls are ILC3s cultured in 10ng/ml IL-23 for 24 hr. For Eomes, Gata3 and T-bet, positive controls are Lin CD45+CD127+ cells. (H) Analysis of ILCregs in different immunodeficient mouse strains. Cell numbers of ILCreg were calculated and shown as means ± SD (right panel). ***p < 0.01 by One Way ANOVA. (I) Analysis of Tregs in Id2 deficient mice. Tregs in lamina propria of small intestine from Id2+/+ or Id2 / mice were analyzed by flow cytometry. CD4+Foxp3+ gate indicates Tregs. Data are representative of at least three independent experiments.

Figure S2. ILCregs Are Derived from CHILPs, Related to Figure 2 (A) Lineage tracing of myeloid cells. ILCregs and myeloid cells (Lin+CD45+CD11b+Gr1+) from Rosa26-STOP-YFP;Lyz2-Cre;IL-10-GFP mice were analyzed by flow cytometry. (B) Lineage tracing of CD4+ T cells. ILCregs and Lin+CD45+ cells from Rosa26-STOP-YFP;CD4-Cre;IL-10-GFP mice were analyzed by flow cytometry. (C) Id2 is required for the development of ILCregs. The indicated mice for lineage tracing of ILCregs were established by crossing Rosa26-STOP-YFP mice with different Cre mice. YFP signals from indicated cell populations were analyzed. Tamoxifen (TMX) (50 mg/kg i.p. for five consecutive days) was injected into Rosa26-STOP-YFP;Id2-CreERT2 mice 2 weeks before flow cytometry analysis. (D) Analysis of YFP+ cells from Lin CD127+IL-10 cells and ILCregs from mice treated with or without 3% DSS for 7 days. Percentages of YFP+ cells were shown as means ± SD. **p < 0.01; NS, not significant by Two-tailed unpaired Student’s t test. (E) Depletion of CHILPs abrogates ILCreg generation. Mice with Cre expression in the indicated ILC progenitors (Id2-CreERT2, PLZF-Cre, and RORgt-Cre) were crossed with Rosa26-STOP-DTR (RosaDTR) mice. 5x104 CLPs (Lin CD127+c-KitlowSca-1low) isolated from the indicated mice (Id2-CreERT2;Rosa-STOP-DTR mice with administration of TMX for 5 days, PLZF-Cre;Rosa-STOP-DTR mice, and RORgt-Cre;Rosa-STOP-DTR mice) were transferred 1:1 with WT CLPs into

(legend continued on next page)

Rag1 / Il2rg / mice and treated with 100 ng/mouse diphtheria toxin (DT) for every 3 days. After 6 weeks, ILCregs and ILCs were analyzed by flow cytometry. **p < 0.01. NS, not significant by One Way RM ANOVA. (F) ILCregs are differentiated from CHILPs. 1x104 CLPs (Lin CD127+c-KitlowSca-1low), 1x104 aLPs (Lin CD127+a4b7+c-KitlowSca-1low), 1x104 CHILPs (Lin Id2+CD127+a4b7+CD25 from Id2-GFP mice) and 1x104 ILCP (Lin CD127+a4b7+PLZF+ from PLZF-GFP mice) were transferred into Rag1 / Il2rg / mice. After 2 weeks, ILCregs in small intestines were analyzed by flow cytometry. (G) 1x104 CHILPs isolated from Rosa-STOP-DTR;Id2-CreERT2 mice and transferred 1:1 with WT CHILPs into Rag1 / Il2rg / mice. Engrafted mice were treated with TMX for 5 days, followed by treatment with 100 ng/mouse DT for every 3 days. After 6 weeks, ILCregs and ILCs were analyzed by flow cytometry. **p < 0.01 by Two-tailed unpaired Student’s t test. Data are representative of at least three independent experiments.

Figure S3. ILCregs Attenuate Innate Colitis, Related to Figure 3 (A) Isotype control staining for anti-IL-10 (left panel) or anti-Ki67 antibody (right panel) by flow cytometry. (B) WT mice were treated with 3% DSS in drinking water for the indicated days or 150 mg anti-CD40 antibody (i.p.) on day 0, or infected by oral gavage of 5x104 c.f.u. S. Typhimurium or 1x109 c.f.u. C. Rodentium on day 0. Colitis scores of indicated mice with different treatments were analyzed and shown as means ± SD n = 5 for each group. (C) Cell numbers of ILC1s, ILC2s and ILC3s after different stimulation as (B) were analyzed by flow cytometry and shown as means ± SD. (D) Rescue of ILCregs in Rag1 / Il10 / mice. ILCregs were transferred into Rag1 / Il10 / mice. After 2 weeks, lymphocytes from lamina propria of small intestines were analyzed by flow cytometry. Lymphocytes were pre-gated by Lin CD45+. (E) Serum IL-10 levels in mice treated as in (D) with indicated days were analyzed by ELISA. Data were shown as means ± SD. (F) ILCregs were labeled with CellTrace (violet) and transferred into Rag1 / Il10 / mice for cell proliferation assays. After 7 days transferred mice were treated with 3% DSS for indicated days, CellTrace-labeled ILCregs were analyzed by flow cytometry. (G) ILCregs attenuate innate colitis induced by S.Typhimurium infection. ILCregs transferred mice were infected by oral gavage of 5x104 c.f.u. S. Typhimurium. After 8 days, colitis scores were analyzed. **p < 0.01 by One Way ANOVA. n = 5 for each group. (H) H&E staining of colon from indicated mice after S. Typhimurium infection. Scale bar, 50 mm. Data are representative of at least three independent experiments.

Figure S4. ILCregs Produce IL-10 and TGF-b1 and Suppress ILCs in an IL-10-Dependent Manner, Related to Figure 4 (A) Analysis ILCregs in mouse intestine from IL-10-GFP mice with oral gavage of water (Ctrl) or 3% DSS for 8 days by flow cytometry. Lin CD45+ intestinal lymphocytes were gated out for CD127 versus IL-10-GFP analysis. (B) Inducible expression of TGF-b1 of ILCregs after treatment with 3% DSS for 8 days. Intracellular TGF-b1 of ILCregs was analyzed by flow cytometry. (C and D) Isolated ILC2 cells (Lin Sca-1+cKit+KLRG1+ST2+) were incubated with indicated cytokines and ILCregs from DSS-treated mice for 3 days. Intracellular cytokines were analyzed by flow cytometry. (C) and secreted cytokines were analyzed by ELISA. (D) Data were show as means ± SD. ND, not detectable.

(legend continued on next page)

(E and F) ILCregs do not inhibit secretion of IL-22 by ILC3 cells. ILC3 cells from RORgt-GFP mice were incubated with ILCregs from DSS-treated mice in the presence or absence of indicated reagents for 3 days. IL-22 was analyzed by flow cytometry (E) or ELISA (F). Data were show as means ± SD. ND, not detectable. (G) ILC1s activated with IL-12/IL-18 or ILC3s activated with IL-23 were incubated with ILCregs from DSS-treated mice for 3 days at the indicated ratios. Secreted cytokines in supernatants were analyzed by ELISA and shown as means ± SD. (H) TGF-b1 does not affect cytokine secretion of ILC1s and ILC3s. ILC1s and ILC3s were incubated with 10 ng/ml TGF-b1 for 3 days. Intracellular cytokines of ILC1s and ILC3s were analyzed by flow cytometry. (I) Flow cytometry analysis of activation markers on ILC1s and ILC3s after treatment with cytokines in the presence or absence of 10 ng/ml TGF-b1 for 3 days. Data are representative of at least three independent experiments.

Figure S5. ILCregs Suppress ILC1 and ILC3 Cells by Secretion of IL-10, Related to Figure 5 (A) Deletion of IL-10Ra in ILC1/ILC3s. ILC1 and ILC3 cells from Il10raflox/flox;Id2-CreERT2 mice were isolated and transferred together with ILCregs (1:1) into Rag1 / Il2rg / mice. One week after transfer, TMX were injected into mice for five consecutive days followed by flow cytometry analysis. Mice treated with vehicle control (Ctrl) served as negative controls. (B) Colitis score of colons from indicated mice were calculated. **p < 0.01 by One Way RM ANOVA. (C) Secreted cytokines from ILC1 and ILC3 cells were analyzed by ELISA and shown as means ± SD. **p < 0.01 by One Way RM ANOVA. (D) Scheme of adoptive transfer assays with IL-10-deleted ILCregs (ILCregDIL-10). ILCregs from Rosa26-STOP-Cas9;Id2-CreERT2;IL-10-GFP mice were isolated and infected with lentivirus carrying sgRNA against Il10 gene. Infected cells were selected with 1 mg/ml puromycin and transferred into Rag1 / Il2rg / mice together with respective WT (CD45.1) lymphocytes. IL-10 expression in ILCregs (CD45.2) after treatment with 0.5 mm 4-hydroxytamoxinfen (4-HT) was detected prior to transplantation. (E) Mice engrafted with ILCregDIL-10 were treated with 3% DSS for 7 days in the presence (+TMX) or absence with TMX (Ctrl). Colitis scores of colons were calculated and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. (F) Colons from indicated mice were sectioned for H&E staining. Scale bar, 50 mm. (G) Secreted cytokines from ILC1 and ILC3 cells were analyzed by ELISA and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. All data are representative of at least three independent experiments.

Figure S6. Depletion of ILCregs, but Not Tregs, Causes Severe Innate Intestinal Inflammation, Related to Figure 5 (A) Generation of Tregs-depleted mice. FoxP3-DTR-GFP Tregs were isolated from FoxP3-DTR-GFP mice (CD45.2) and transferred together with ILC1s (from CD45.1 WT mice) and ILC3s (from CD45.1 RORgt-GFP mice) into Rag1 / Il2rg / mice (1:1 ratio). After 7 days, 100 ng DT per mouse was injected into transferred mice (i.p. every three days) and treated with 3% DSS for 6 days. CD4+CD25+FoxP3+ Tregs were analyzed by flow cytometry and cell numbers were shown as means ± SD. ***p < 0.001 by Two-tailed unpaired Student’s t test. (B) H&E staining of colon from mice after injection with vehicle control or DT after 6 days’ treatment of DSS. Scale bar, 50 mm. (C) Colitis scores of the indicated mice were shown as means ± SD. NS, not significant. (D) Secreted cytokines of ILCs from mice treated as above were analyzed by ELISA and shown as means ± SD. NS, not significant by Student’s t test. (E) Generation of ILCregs depleted mice. ILCregDTR cells were isolated from Lck-Cre;RosaDTR;IL-10-GFP mice (CD45.2) and transferred together with ILC1s and ILC3s (from CD45.1 mice) into Rag1 / Il2rg / mice. After 7 days, 100 ng DT per mouse was injected into transferred mice (i.p. every three days) and treated with 3% DSS for 6 days. ILCregs were analyzed by flow cytometry and cell numbers were shown as means ± SD. ***p < 0.001 by Two-tailed unpaired Student’s t test. (F) H&E staining of colon from transferred mice after injection with vehicle control or DT with 6 days’ treatment of DSS. Scale bar, 50 mm.

(legend continued on next page)

(G) Colitis scores of engrafted mice with the indicated treatment were shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. (H) Secreted cytokines of ILCs from mice treated as above were analyzed by ELISA and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. (I and J) WT mice were treated with vehicle control or 100 ng/mouse DT every three days with oral gavage of drinking water (Ctrl) or 3% DSS. After 6 days, colitis scores of indicated mice were analyzed and shown as means ± SD (I). NS, not significant by Two-tailed unpaired Student’s t test. Cytokines secreted by ILCs were analyzed by ELISA and shown as means ± SD (J). (K) Generation of ILCregDTR mice. ILCregs were isolated from Lck-Cre;RosaDTR;IL-10-GFP mice (CD45.2) and transferred together with T cells (CD3+TCRb+TCRgd+), B cells (CD19+B220+), NK cells (NK1.1+DX5+), ILC1s (CD127+NK1.1+NKp46+), ILC2s (Lin CD127+KLRG1+ST2+Sca1+) from LPL of WT mice (CD45.1) and ILC3s (Lin CD45+CD127+RORgt-GFP+) from RORgt-GFP (CD45.1) mice into Rag1 / Il2rg / mice. After 2 weeks, lymphocytes in engrafted mice were analyzed by flow cytometry. Cell numbers were normalized to those of WT mice and shown as means ± SD. (L) After administration of DT, ILCregs were analyzed by flow cytometry. ILCregs were almost lost. (M) H&E staining of colon from mice after injection with vehicle control or DT together with 6 days’ treatment of DSS. Scale bar, 50 mm. (N) Colitis scores of the indicated mice were shown as means ± SD. **p < 0.01 by One Way ANOVA. (O) ILCs were isolated from indicated mice. Secreted cytokines were analyzed by ELISA. The ELISA data were normalized to total cell numbers and shown as means ± SD. **p < 0.01 One Way ANOVA. Data are representative of at least three independent experiments.

Figure S7. TGF-b1 Is Required for the Maintenance of ILCregs during Innate Intestinal Inflammation, Related to Figure 7 (A) TGF-b1 is required for the maintenance of ILCregs during inflammatory responses. Bone marrow cells from WT (CD45.1) and Tgfb1flox/flox;Lck-Cre (CD45.2) mice were mixed and adoptively transferred into Rag1 / Il2rg / mice. After 8 weeks, chimeric mice were treated with 3% DSS for 8 days. ILCregs were analyzed by flow cytometry. Lin CD45.1+ or Lin CD45.2+ cells from mouse lamina propria cells were gated out for CD127 versus IL-10 analysis. Cell numbers of ILCregs were calculated and shown as means ± SD. ***p < 0.001 by Two-tailed unpaired Student’s t test. n = 6 for each group. (B) Abrogation of TGFb1 in ILCregs does not affect development of lymphocyte progenitors. Hematopoietic progenitors of WT (CD45.1) and Tgfb1flox/flox;Lck-Cre (CD45.2) chimeric mice were analyzed by flow cytometry. CD45.1 and CD45.2 cells from chimeric BM were gated out for analysis of hematopoietic progenitors. Absolute cell numbers of each cell subsets were calculated and shown as means ± SD. NS, not significant by One Way ANOVA. Long-term hematopoietic stem cell (LT-HSC, Lin CD127 Sca-1+c-Kit+CD34 Flk2 ), short-term hematopoietic stem cell (ST-HSC, Lin CD127 Sca-1+c-Kit+CD34+Flk2 ), multipotent progenitor cells (MPP, Lin CD127 Sca-1+c-Kit+CD34+Flk2+), common myeloid progenitor cells (CLP, Lin CD127+c-Kitlow Sca-1low). (C) ILCregs from Tgfb1flox/flox;CreERT2;IL-10-GFP mice were labeled with CellTrace (violet) and transferred together with ILC1s and ILC3s into Rag1 / Il2rg / mice. After 1 week, mice were treated with DSS or DSS/TMX for 5 days. CellTrace signals from Lin CD45+CD127+ cells were analyzed by flow cytometry. Deletion of TGF-b1 almost lost CellTrace-labeled ILCregs.

(legend continued on next page)

(D and E) TGFb1 is required for the maintenance of ILCregs during S. Typhimurium infection. ILCregs were isolated from Tgfb1flox/flox;CreERT2;IL-10-GFP mice and transferred with ILC1s and ILC3s into Rag1 / Il2rg / mice. After 2 weeks, mice were treated with PBS (Ctrl) or oral gavage of 5x104 c.f.u. S. Typhimurium (+S. Typhimurium) or injection of tamoxinfen (TMX) (50 mg/kg i.p. for five consecutive days) with oral gavage of 1x105 c.f.u. S. Typhimurium (+S. Typhimurium/ TMX). Colitis scores were analyzed 8 days after infection (D), and ILCs were isolated and cultured in complete media for 24 hr. Secreted cytokines were analyzed by ELISA and shown as means ± SD (E). **p < 0.01 by Two-tailed unpaired Student’s t test. n = 6 for each group. (F) Scheme of deletion TGF-bRII of ILCregs by administration of tamoxinfen (TMX). ILCregs were isolated from Tgfbr2flox/flox;CreERT2;IL-10-GFP mice and transferred with WT ILC1 and ILC3 cells into Rag1 / Il2rg / mice. After 2 weeks, mice were treated with water (Ctrl) or 3% DSS (+DSS) or injection of tamoxinfen (TMX) (50 mg/kg i.p. for five consecutive days) with oral gavage of 3% DSS (+DSS/TMX) for 8 days. (G) ILCregs from mice treated as in (F) were analyzed by flow cytometry. Cell numbers of ILCregs in different tissues were analyzed and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. (H) H&E staining of colon from indicated mice treated as in (F). Scale bar, 50 mm. (I) Colitis scores of indicated mouse colon were analyzed and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. n=6 for each group. (J) Cytokines secreted by ILC1s or ILC3s from indicated mice were analyzed by ELISA and shown as means ± SD. **p < 0.01 by Two-tailed unpaired Student’s t test. Data are representative of at least three independent experiments.