Research ArticleINNATE LYMPHOID CELLS

A circadian clock is essential for homeostasis of group 3 innate lymphoid cells in the gut

See allHide authors and affiliations

Science Immunology  04 Oct 2019:
Vol. 4, Issue 40, eaax1215
DOI: 10.1126/sciimmunol.aax1215

Keeping track of time

A number of bodily functions, from feeding to sleeping, are regulated by our circadian clocks. Teng et al. and Wang et al. report that expression of a number of genes in intestinal group 3 innate lymphoid cells (ILC3s) in mice is synchronized to the time of day. To probe the role of clock proteins in regulating ILC3 functions, the two groups studied mice lacking distinct clock proteins, BMAL1 and REV-ERBα. Both groups found that the loss of these clock proteins resulted in marked losses of intestinal ILC3s, but that these ILC3s produced higher levels of certain cytokines. How clock proteins fit within the larger network of transcription factors in regulating development and gene expression in ILC3s remains to be seen.

Abstract

Group 3 innate lymphoid cells (ILC3s) critically orchestrate host-microbe interactions in the healthy mammalian intestine and become substantially impaired in the context of inflammatory bowel disease (IBD). However, the molecular pathways controlling the homeostasis of ILC3s remain incompletely defined. Here, we identify that intestinal ILC3s are highly enriched in expression of genes involved in the circadian clock and exhibit diurnal oscillations of these pathways in response to light cues. Classical ILC3 effector functions also exhibited diurnal oscillations, and lineage-specific deletion of BMAL1, a master regulator of the circadian clock, resulted in markedly reduced ILC3s selectively in the intestine. BMAL1-deficient ILC3s exhibit impaired expression of Nr1d1 and Per3, hyperactivation of RORγt-dependent target genes, and elevated proapoptotic pathways. Depletion of the microbiota with antibiotics partially reduced the hyperactivation of BMAL1-deficient ILC3s and restored cellular homeostasis in the intestine. Last, ILC3s isolated from the inflamed intestine of patients with IBD exhibit substantial alterations in expression of several circadian-related genes. Our results collectively define that circadian regulation is essential for the homeostasis of ILC3s in the presence of a complex intestinal microbiota and that this pathway is disrupted in the context of IBD.

INTRODUCTION

The mammalian gastrointestinal tract is continuously exposed to both pathogenic and commensal microbes, and the immune system must carefully regulate immunity, inflammation, and tolerance at this site to maintain tissue homeostasis (13). Innate lymphoid cells (ILCs) have emerged as critical regulators of these processes due to their enrichment and tissue residency in the intestine, ability to rapidly respond to microbial stimuli, and multiple protective, antimicrobial, or immunomodulatory functions (48). In particular, group 3 ILCs (ILC3s) play numerous roles in regulating intestinal health through production of interleukin-22 (IL-22) as well as their ability to control adaptive immune responses to dietary antigens or the intestinal microbiota (916). Consistent with this, experimental manipulation of ILC3s in mice frequently results in spontaneous inflammation, microbial dysbiosis, or increased susceptibility to tissue damage and infection in the intestine (13, 14, 17, 18). Furthermore, humans with chronic intestinal inflammation or infections exhibit substantially impaired ILC3 responses (14, 15, 1922), implicating a critical role for these cells in regulating human health.

Although many advances have shed light on how ILC3s maintain intestinal health and become impaired in human diseases, there is a limited understanding of the cellular and molecular pathways that are essential for ILC3 homeostasis in the gut. One identified factor is the microbiota, which can support the development of T-bet+ ILC3 responses in the intestine (2325). Recent studies have demonstrated that the intestinal microbiota exhibits diurnal oscillation in luminal abundance, epithelial attachment, and metabolism (2629). This circadian cycle of the microbiota also influences the transcriptomic and epigenomic patterns of host cells both local and distal to the gut (27, 30). Mammalian hosts also exhibit diurnal oscillations or a circadian clock. These pathways are regulated by, or consist of, complex neuronal signaling from the synchronization of environmental light cues by the suprachiasmatic nuclei in the hypothalamus and release of hormones or metabolites to the periphery (27, 3133). This can, in turn, influence multiple immune responses and nonhematopoietic cells through entrainment and a cell-autonomous circadian clock (34, 35). In the intestine, it was recently identified that intestinal epithelial cells (IECs) exhibit diurnal oscillation in transcription patterns that are required to orchestrate normal lipid metabolism, and this was, in part, regulated by the microbiota, IL-22, and IEC-intrinsic expression of NFIL3 (36). However, it remains unclear whether ILC3s exhibit circadian regulation and whether this is important for their cellular homeostasis. Here, we report that ILC3s exhibit diurnal oscillations in expression of genes related to the circadian clock and classical effector functions and that ILC3-specific BMAL1 is required for cellular homeostasis in the presence of a diverse microbiota. We identify that ILC3s in the inflamed intestine of patients with inflammatory bowel disease (IBD) exhibit alterations in circadian gene expression. These findings reveal that circadian regulation critically orchestrates ILC3 homeostasis in the gut and suggest that disruption in this pathway drives the impaired ILC3 responses in the inflamed intestine of patients with IBD.

RESULTS

To begin to examine whether ILC3s exhibit circadian regulation, we sort-purified ILC3s from the small intestinal lamina propria (SI-LP) of healthy mice and profiled them with RNA sequencing. In comparison with the related GATA3+ group 2 ILCs (ILC2s) population, RORγt+ ILC3s exhibited enrichment in multiple transcripts associated with a cell-intrinsic circadian clock, including Per1, Per2, Per3, Cry1, Cry2, Nr1d1, and Nfil3 (Fig. 1A). To verify these findings, we sort-purified several immune cell subsets from the SI-LP of C57BL/6 mice at Zeitgeber time (ZT) 6 (fig. S1). Quantitative polymerase chain reaction (qPCR) analyses confirmed that among major immune cell subsets in the intestine, ILC3s exhibited significantly higher expression of key circadian genes, including Nr1d1, Per1, Cry2, Nfil3, and Arntl (Fig. 1B). We next examined whether ILC3s express these genes in a diurnal manner by sort-purifying SI-LP ILC3s at four time points during a 24-hour cycle. We observed by qPCR analyses that ILC3s express Nr1d1, Per1, Per2, Per3, and Cry2 in a diurnal oscillation pattern (Fig. 1C). Diurnal oscillations of peripheral cell types can be entrained by the suprachiasmatic nuclei in the hypothalamus in response to light cues (27, 3133). Consistent with this, a 9-hour phase shift in the light-dark cycle resulted in a comparable shift in the diurnal expression patterns of Nr1d1, Per1, Per2, and Cry2 (Fig. 1D). In contrast, depletion of a diverse microbiota with antibiotics (Abx) modestly affected the diurnal oscillations of these circadian genes in ILC3s (fig. S2A). Thus, our data indicate that intestinal ILC3s exhibit a cell-intrinsic circadian clock that is regulated, in part, by light cues.

Fig. 1 Intestinal ILC3s are enriched for, and exhibit diurnal oscillations in, circadian-related gene expression.

(A) Heatmap showing expression z scores of the indicated genes in ILC2s (CD45+LinCD90.2+CD127+CD27KLRG1+) and ILC3s (CD45+CD3εCD127+RORγtGFP+) from SI-LPs of naïve Rorc(γt)-GfpTG mice, as measured by RNA sequencing (Lineage markers: CD3ε, CD5, CD8α, NK1.1, CD11c, CD11b, B220). (B) qPCR analyses of indicated circadian genes from sort-purified T cells, B cells, dendritic cells (DC), macrophages (Mac), ILC2s, and ILC3s from SI-LP of C57BL/6 mice sacrificed at ZT6. The expression of each target gene was normalized to Actb2 (n = 7, pooled from two independent assays). (C) qPCR analyses of diurnal expression patterns in indicated circadian genes on sort-purified ILC3s from SI-LP of C57BL/6 mice sacrificed at ZT0, ZT6, ZT12, and ZT18 within a 24-hour cycle. The expression of each target gene at each time point was first normalized to Actb2 and then further normalized to the sample with lowest expression at ZT0 (n = 4, representative of two independent assays). (D) qPCR analyses of diurnal expression patterns in indicated circadian genes on sort-purified ILC3s from SI-LP of C57BL/6 mice housed in standard light:dark cycle (control) or mice housed in 9-hour advanced light:dark cycle (L:D phase advanced). Both groups of mice were sacrificed at ZT0, ZT6, ZT12, and ZT18 within a 24-hour cycle. The expression of each target gene at each time point was first normalized to Actb2 and then further normalized to the sample with lowest expression at ZT0 in the control group (n = 6 per group, pooled from two independent assays). Results are means ± SEM. Statistics are calculated by one-way ANOVA.

A role for the circadian clock has been implicated in the function of multiple other immune cell types (35). Therefore, we next examined whether the classical effector functions of ILC3s exhibit circadian regulation. ILC3s are characterized by expression of the master regulator transcription factor RORγt and high expression of the effector cytokines IL-17 and IL-22 (10, 37). We observed that Rorc, Il17a, and Il22 transcripts exhibited diurnal oscillations within a 24-hour cycle (Fig. 2A). We were also able to isolate differential numbers of ILC3s from the intestine in a diurnal manner, observed oscillations of Csf2 expression in ILC3s, and found a diurnal expression of Reg3b in IECs (fig. S2, B to D). We verified diurnal oscillation patterns in intestinal ILC3s at the protein level by the staining intensity of RORγt (Fig. 2B). To verify oscillations in IL-22, we developed a novel IL-22 reporter mouse to permit the analyses of IL-22 production without the need for ex vivo restimulation (fig. S3A) and validated that the enhanced green fluorescent protein (eGFP) signal could robustly represent IL-22 protein production by ILC3s (fig. S3, B and C). Consistent with the oscillations of Il22 transcript, we observed significant changes in IL-22–eGFP levels between ZT6 and ZT18 in intestinal ILC3s from the reporter mice (Fig. 2C). The diurnal oscillations of Rorc and Il22 transcripts were regulated by both light cues and microbiota-derived signals (Fig. 2, D and E, and fig. S2, E and F). Together, these data demonstrate that major ILC3 regulators and effector functions also exhibit diurnal oscillations across a 24-hour cycle.

Fig. 2 ILC3-specific effector and regulatory pathways exhibit diurnal oscillations.

(A) qPCR analyses of Rorc, Il17a, and Il22 expression on sort-purified ILC3s from SI-LP of C57BL/6 mice sacrificed at ZT0, ZT6, ZT12, and ZT18 within a 24-hour cycle. The expression of each target gene at each time point was first normalized to Actb2 and then further normalized to the sample with lowest expression at ZT0 (n = 4, representative of two independent assays). (B) Mean fluorescent intensity (MFI) of RORγt diurnal expression in ILC3s from large intestinal LP of mice from (A) (n = 4, representative of two independent assays). (C) Representative flow cytometry plots and quantification of frequency of IL-22–eGFP–expressing ILC3s in SI-LP of IL-22–eGFP mice. Mice were sacrificed at ZT6 and ZT18, and cells were gated on live CD45dimLinCD90.2+CD127+CD27KLRG1 (n = 10 to 11, pooled from three independent assays). (D) qPCR analysis of Il22 expression on sort-purified ILC3s from SI-LP of C57BL/6 mice housed in standard light:dark cycle or mice housed in 9-hour advanced light:dark cycle. Both groups of mice were sacrificed at ZT0, ZT6, ZT12, and ZT18 within a 24-hour cycle. The relative gene expression was normalized as Fig. 1D (n = 6 per group, pooled from two independent assays). (E) qPCR analysis of Il22 expression on sort-purified ILC3s from SI-LP of conventionally housed C57BL/6 mice or mice treated with Abx for 2 weeks. Both groups of mice were sacrificed at ZT0, ZT6, ZT12, and ZT18 within a 24-hour cycle. The gene expression at each time point was first normalized to Actb2 and then further normalized to the sample with lowest expression at ZT0 in the control group (n = 6 per group, pooled from two independent assays). Results are means ± SEM. Statistics are calculated by unpaired two-tailed Student’s t test.

We next examined the functional significance of these diurnal oscillation patterns by generating mice with an ILC3-specific deletion in BMAL1, a master regulator of the circadian cycle encoded by Arntl (38). This was accomplished by crossing Rorccre mice (37) with BMAL1fl/fl mice (39) to generate BMAL1ΔRorc mice that exhibit an efficient loss of BMAL1 protein in intestinal ILC3s (Fig. 3A). Although RORγt is also expressed in double-positive thymocytes and maintained in T helper 17 (TH17) cells, we did not observe a loss of BMAL1 protein in intestinal TH17 cells in these mice (Fig. 3A), suggesting a selective targeting of ILC3s. We found that ILC3s from BMAL1ΔRorc mice exhibited significantly altered expression of RORγt and NFIL3 protein (Fig. 3B), indicating a functional impairment in this population. The lack of ILC3-intrinsic BMAL1 also led to significant reductions in both the frequency and cell number of ILC3s in the small intestine (Fig. 3C). This is in contrast to a recent study demonstrating that intestinal TH17 cells, an adaptive counterpart of ILC3s, are modestly affected by BMAL1 deletion (40). We found a preferential requirement of BMAL1 in the T-bet+ subset of ILC3s (Fig. 3D), a population that requires the microbiota for development (2325). This was selective to small intestinal ILC3s, as the cellularity of intestinal ILC2s and TH17 cells, along with ILC3s in lymphoid tissues, remained comparable to littermate controls (fig. S4, A to C). There were also comparable Peyer’s patches and myeloid cell–derived cytokines in both groups of mice, and there was no impact on the homeostasis of ILC3s after selective deletion of BMAL1 in T cells (fig. S4, D to F). The frequency and total numbers of ILC3s in the large intestine were comparable in the context of cell-intrinsic BMAL1 deletion, but significant reductions were observed in the T-bet+ ILC3 subset (fig. S5, A and B). Littermate control and BMAL1ΔRorc mice exhibited comparable susceptibility with dextran sulfate sodium (DSS)–induced intestinal damage and inflammation (fig. S5, C and D), which is consistent with the potential redundancy of T-bet+ ILC3s in this model (41). Collectively, these data critically demonstrate that the master regulator of circadian cycles, BMAL1, is essential for the homeostasis of ILC3s selectively within the intestine.

Fig. 3 Cell-intrinsic BMAL1 is required for ILC3 homeostasis in the intestine.

(A) Representative histogram overlay of BMAL1 expression in SI-LP ILC3s and TH17 cells from BMAL1fl/fl and BMAL1ΔRorc mice. ILC3s were gated as live CD45+LinCD90.2+CD127+GATA-3RORγt+; TH17 cells were gated as live CD45+CD3ε+CD4+RORγt+IL-17A+. (B) Quantifications of MFI of RORγt and NFIL3 in SI-LP ILC3s from BMAL1fl/fl and BMAL1ΔRorc mice (representative of three independent assays). (C) Representative flow cytometry plots and quantifications of frequency and cell numbers of SI-LP ILC3s from BMAL1fl/fl and BMAL1ΔRorc mice. Cells were gated on live CD45+LinCD90.2+CD127+ (representative of four independent assays). (D) Representative flow cytometry plots and quantifications of frequency and cell numbers of T-bet+ ILC3s from SI-LP of BMAL1fl/fl and BMAL1ΔRorc mice. Cells were gated on live CD45+LinCD90.2+CD127+GATA-3RORγt+ (representative of four independent assays). Results are means ± SEM. Statistics are calculated by unpaired two-tailed Student’s t test.

To define how BMAL1 regulates ILC3 cellular homeostasis, we performed RNA sequencing on sort-purified populations from the small intestine of littermate control and BMAL1ΔRorc mice. Unbiased analyses revealed significant differences in transcriptional signatures, which were driven in part by altered expression of genes associated with circadian regulation, RORγt-mediated transcription, and apoptosis pathways (Fig. 4, A and B). Specifically, BMAL1 deficiency in ILC3s resulted in significantly reduced expression on key circadian-associated genes Nr1d1, Nr1d2, and Per3, whereas Cry1 was significantly increased and other circadian-related genes remained unchanged (Fig. 4C). NR1D1 (also known as Rev-erbα) is a transcriptional repressor that competes for binding to RORγt DNA consensus sequences (42, 43). Consistent with this, BMAL1-deificient ILC3s exhibited hyperactivation of RORγt-dependent target genes Il17a, Il17f, and Il22, as well as the proapoptotic pathways Bcl2l11 (encoding Bim) and Bax (Fig. 4C). We verified that BMAL1-deficient ILC3s exhibited substantially impaired expression of Nr1d1 and Per3 at both ZT6 and ZT18, whereas Per1 and Cry2 were comparable at these time points, and Il22 was notably increased at ZT6, relative to littermate controls (Fig. 4, D and E). Expression of Reg3b in the intestinal epithelium was comparable between both groups of mice, which is likely the result of significantly elevated frequencies of IL-22– and IL-17A–producing ILC3s but reduced total numbers (Fig. 4, F and G, and fig. S6A). Mice lacking ILC3-specific BMAL1 also exhibited comparable frequencies of proliferating Ki-67+ ILC3s but significantly increased levels of Bim+ ILC3s relative to littermate controls (Fig. 4, H and I). The reduced frequencies of ILC3s and increased levels of Bim+ ILC3s were observed as early as 4 weeks of age (fig. S6, B to D). These results collectively demonstrate that ILC3-specific BMAL1 critically promotes expression of several circadian-associated genes (Nr1d1 and Per3), limits hyperactivation of RORγt-dependent target genes (Il17a, Il17f, and Il22), and prevents elevated expression of proapoptotic pathways (Bim and Bax).

Fig. 4 BMAL1 deficiency results in circadian dysregulation, cytokine hyperproduction, and increased cell death in intestinal ILC3s.

(A) Principal components analysis (PCA) of genome-wide transcriptional profiles of sort-purified ILC3s from SI-LP of BMAL1fl/fl and BMAL1ΔRorc mice, as measured by RNA sequencing. n = 3 per group. (B) Volcano plot of differential expression between BMAL1fl/fl [positive log2(FC)] and BMAL1ΔRorc [negative log2(FC)] groups. Differentially expressed genes (defined as FDR < 0.1) are shown in red. FC, fold change. (C) Heatmap showing expression z scores of the indicated genes in ILC3s from SI-LP of BMAL1fl/fl and BMAL1ΔRorc mice. (D and E) qPCR analyses of indicated genes on sort-purified ILC3s from SI-LP of BMAL1fl/fl and BMAL1ΔRorc mice. Both groups of mice were sacrificed at ZT6 and ZT18. The expression of each target gene at each time point was first normalized to Actb2 and then further normalized to the sample with lowest expression at ZT6 in the BMAL1fl/fl group (n = 6 to 7 per group, pooled from two independent assays). N.D., nondetectable. (F and G) Representative flow cytometry plots and quantifications of frequency and cell numbers of IL-22+ ILC3s (F) and IL-17A+ ILC3s (G) from SI-LP of BMAL1fl/fl and BMAL1ΔRorc mice. Cells were gated on live CD45+LinCD90.2+CD127+GATA-3RORγt+ (representative of four independent assays). (H and I) Quantifications of frequency of Ki-67+ ILC3s (H) and MFI of Bim expression in ILC3s (I) from SI-LP of 8-week-old BMAL1fl/fl and BMAL1ΔRorc mice (representative of four independent assays). Results are means ± SEM. Statistics are calculated by unpaired two-tailed Student’s t test.

There is a dynamic and reciprocal regulation of ILC3s with intestinal microbiota, which promotes expression of RORγt and its associated target genes (18, 4446). Given our findings and previous reports that the microbiota exhibit diurnal oscillations (2629), we next examined whether the microbiota is driving the observed alterations of ILC3 homeostasis after loss of BMAL1. This was accomplished by administering Abx to BMAL1ΔRorc mice. After 2 weeks of Abx administration, we observed comparable protein levels of RORγt, but significantly reduced levels of IL-17A and IL-22 in ILC3s from Abx-treated BMAL1ΔRorc mice relative to nontreated BMAL1ΔRorc mice (Fig. 5A). Abx treatment also resulted in a partial restoration of ILC3 frequencies and a complete restoration of ILC3 numbers in BMAL1ΔRorc mice (Fig. 5B). These results demonstrate that circadian regulation is essential for orchestrating homeostasis between ILC3s and the microbiota and imply that hyperactivation of RORγt is associated with the reduction of intestinal ILC3s in the context of BMAL1 deficiency. In support of this, we identified that small-molecule inhibition of RORγt in purified ILC3 cultures was sufficient to reduce expression of BMAL1 as well as significantly reduce Bim+ ILC3s in the context of simulation with proinflammatory cytokines that are typically elicited by the microbiota (Fig. 5, C and D). Thus, hyperactivation and loss of intestinal ILC3s in the context of cell-intrinsic BMAL1 deletion are driven, in part, by the microbiota.

Fig. 5 Microbiota drives the dysfunction of ILC3s in the absence of BMAL1.

(A) Quantifications of MFI of RORγt in ILC3s and frequency of IL-17A+ and IL-22+ ILC3s from SI-LP of Abx-treated or control BMAL1fl/fl and BMAL1ΔRorc mice (representative of three independent assays). (B) Representative flow cytometry plots and quantifications of frequency and cell numbers of SI-LP ILC3s from Abx-treated or control BMAL1fl/fl and BMAL1ΔRorc mice. Cells were gated as Fig. 3C (representative of three independent assays). (C) qPCR analysis of Arntl expression on sort-purified SI-LP ILC3s from Rorc(γt)-GfpTG mice treated with 2 μM of the RORγt inhibitor GSK805 or vehicle control. The gene expression was normalized to Actb2 (n = 9, pooled from four independent assays). (D) Quantification of MFI of Bim in SI-LP ILC3s from C57BL/6 mice after incubation with 2 μM GSK805 or vehicle control in the presence of recombinant mouse IL-1β, IL-6, IL-12, and IL-23 (20 ng/ml each) (representative of two independent assays). Results are means ± SEM. Statistics are calculated by one-way ANOVA or paired two-tailed Student’s t test.

IBD is an inflammatory disease of humans that is accompanied with microbial dysbiosis and disruption of intestinal ILC3 responses (14, 15, 22). To investigate for a contribution of circadian regulation, we sort-purified ILC3s from the inflamed or noninflamed regions of intestinal resections from patients with IBD. We observed significantly reduced frequencies of ILC3s from the inflamed versus noninflamed intestine of patients with IBD, and this correlated with significantly altered expression of the circadian-related genes NR1D1, PER3, and NFIL3 (Fig. 6, A and B), whereas expression of ARNTL and CRY1 was not substantially altered (fig. S7). These results suggest that, in the context of chronic intestinal inflammation, human ILC3s exhibit alterations in circadian gene expression, which may contribute to the altered homeostasis of these protective cell types.

Fig. 6 Circadian regulation of ILC3s is disrupted in patients with IBD.

(A) Representative flow cytometry plots and quantification of frequency of ILC3s from distal noninflamed versus matched inflamed surgical resection tissues of patients with IBD. Cells were gated on live CD45+LinCD127+ (Lineage markers: CD3, CD5, CD11b, CD11c, CD19, FcεR1). (B) qPCR analyses of indicated circadian genes on sort-purified ILC3s from distal non-inflamed versus matched inflamed surgical resection tissues of patients with IBD. The expression of each target gene was normalized to GAPDH. Statistics are calculated by paired two-tailed Student’s t test.

DISCUSSION

Our results define that cell-intrinsic circadian regulation is critically required for the homeostasis of ILC3s in the intestine. The mammalian circadian clock plays important roles in regulating multiple immune cell functions (34, 35, 47). Recent studies have shed light on how circadian clock influences both innate and adaptive counterparts of immunity, as well as nonimmune cells to properly respond to environmental stimuli and maintain homeostasis (36, 48, 49). However, very little is known about the circadian clock and ILC3s. An important earlier study demonstrated that ILC2s exhibit circadian regulation and that this was important to support eosinophil homeostasis (50). Here, we provide evidence that intestinal ILC3s require a circadian clock to better sense and interact with the microbiota, adding another layer of complexity in the dialogue between ILCs and microbes (fig. S8). This may be the result of coordinating diurnal cycles to facilitate normal ILC3 responses and fine-tune the function of RORγt. In the context of patients with IBD, we identified that ILC3s in the inflamed intestine exhibit altered circadian gene expression. This is consistent with previous reports that characterized altered circadian gene expression in whole intestinal biopsies from patients with IBD (51). Given that our findings come from the same patient, it suggests that the altered expression is not being driven by body-wide disruptions in circadian cycles or entrainment, but rather the result of chronic inflammation or microbial dysbiosis. Inflammatory disruption of circadian cycles has been reported in multiple other cell types (5254). Moreover, it has become increasingly appreciated that there are certain signatures of immune responses and therapies that exert day and night differences (termed as chrono-immunotherapy), and our understanding of this is already being applied at different levels in diseases such as chronic obstructive pulmonary disease and rheumatoid arthritis (55, 56). Therefore, our critical finding that ILC3s in the inflamed intestine of patients with IBD exhibit altered circadian gene expression suggests that this is an important pathway in human health and disease and may hold a key for developing novel strategies to boost ILC3 responses in the context of impaired intestinal homeostasis or microbial dysbiosis.

MATERIALS AND METHODS

Study design

The objective of this study was to interrogate the role of circadian regulation on ILC3 homeostasis in the gut. To do this, we used a combination of ex vivo and in vivo assays with both mouse and human samples. We designed and performed the experiments mainly in the fields of cellular immunology and molecular biology. The number of replicates for each experiment is indicated in the figure legends.

Mice

Wild-type, BMAL1fl/fl, and Cd4cre mice on a C57BL/6 background were purchased from the Jackson Laboratory. C57BL/6 Rorccre and Rorc(γt)-GfpTG mice were provided by G. Eberl. BMAL1ΔRorc mice were generated by crossing Rorccre mice with BMAL1fl/fl mice, and BMAL1ΔCd4 mice were generated by crossing Cd4cre mice with BMAL1fl/fl mice. IL-22–eGFP mice were generated in collaboration with Cyagen US Inc. by modifying a bacterial artificial chromosome (BAC) as noted in fig. S3A and injection into fertilized C57BL/6 embryos. All mice were bred and maintained in specific pathogen–free facilities under a standard 24-hour light:dark cycle (light on for 12 hours from 6:00 a.m. to 6:00 p.m. as ZT0 to ZT12 at Weill Cornell Medicine). In one noted experimental set, mice were housed with a 9-hour advanced light:dark cycle, and the light was set on from 9:00 p.m. to 9:00 a.m. Mice were fed ad libitum. Sex- and age-matched littermates were used as controls in all experiments, and mice between 7 and 11 weeks of age were used for all experiments unless otherwise indicated. Mice were sacrificed at ZT5 to ZT6 unless otherwise indicated. All animal experiments were approved and are in accordance with the Institutional Animal Care and Use Committee guidelines at Weill Cornell Medicine.

Isolation of cells from the intestinal epithelium and lamina propria of mice and humans

Mouse intestines were removed, cleaned from remaining fat tissue, and washed in ice-cold phosphate-buffered saline (PBS) (Corning). Peyer’s patches on the small intestine were identified and completely eliminated. Intestines were opened longitudinally and washed in ice-cold PBS. Afterward, mucus was gently removed by forceps and intestines were cut into about 0.5-cm sections. Dissociation of epithelial cells was performed by incubation on a shaker in Hanks’ balanced salt solution (Sigma-Aldrich) containing 5 mM EDTA (Thermo Fisher Scientific), 1 mM dithiothreitol (DTT) (Sigma-Aldrich), and 2% heat-inactivated fetal bovine serum (FBS) two times for 20 min each at 37°C. After each incubation, samples were vortexed, and after the second incubation, the epithelial fraction was washed by ice-cold PBS and proceeded with qPCR analysis. The remaining samples were then washed by cold PBS, and enzymatic digestion was performed using dispase (0.4 U/ml; Thermo Fisher Scientific), collagenase III (1 mg/ml; Worthington), and deoxyribonuclease (DNase) I (20 μg/ml; Sigma-Aldrich) in 10% FBS/Dulbecco’s modified Eagle’s medium (DMEM) (Corning) on a shaker for 45 min at 37°C. Leukocytes were further enriched by a 40/80% Percoll gradient centrifugation (GE Healthcare).

Surgical resection samples from patients with IBD (see table S1) were obtained through Institutional Review Board–approved protocols from the Center for Advanced Digestive Care at Weill Cornell Medicine following informed consent. Single-cell suspensions from intestinal tissues were obtained by incubating tissues for 30 min at 37°C with shaking in stripping buffer (1 mM EDTA, 1 mM DTT, and 5% FBS) to remove the epithelial layer. Supernatants were then discarded. Tissues were then mechanically dissociated with a sterile scalpel. The lamina propria fraction was obtained by incubating the dissociated tissues for 1 hour at 37°C with shaking in collagenase D (2 mg/ml) (Roche), DNase I (0.1 mg/ml) (Sigma-Aldrich), and trypsin inhibitor (1 mg/ml) (Gibco) digestion solution. The remaining tissues were then filtered through a 70-μm cell strainer. All cells were then viably cryopreserved in 90% FBS and 10% dimethyl sulfoxide for side-by-side analysis at a later time point. After thawing and filtering through a 70-μm cell strainer, cells were stained with antibodies for flow cytometry acquisition.

Flow cytometry and cell sorting

Single-cell suspensions were incubated on ice with conjugated antibodies in PBS containing 2% FBS and 1 mM EDTA. Dead cells were excluded with Fixable Aqua Dead Cell Stain (Thermo Fisher Scientific). The staining antibodies for flow cytometry were purchased from Thermo Fisher Scientific, BioLegend, BD Biosciences, or Cell Signaling Technology. The following antibodies were used for mouse cell surface staining: CD45 (30-F11), CD3ε (145-2C11), CD5 (53-7.3), CD8α (53-6.7), NK1.1 (PK136), CD11c (N418), CD11b (M1/70), B220 (RA3-6B2), CD19 (eBio1D3), TCRβ (H57-597), CD64 (X54-5/7.1), MHC-II (M5/114.15.2), CD90.2 (30-H12), CD127 (A7R34), CD27 (LG.7F9), CD4 (GK1.5 or RM4-5), and KLRG1 (2F1). The following antibodies were used for mouse intracellular staining: T-bet (4B10), GATA3 (L50-823), RORγt (B2D or Q31-378), NFIL3 (S2M-E19), Ki-67 (SolA15), Bim (C34C5), IL-17A (eBio 17B7), and IL-22 (IL22JOP). Intracellular staining of mouse BMAL1 was conducted by first staining polyclonal anti-BMAL1 antibody (Abcam) and then further staining with fluorescent-conjugated donkey anti-rabbit immunoglobulin G antibody (BioLegend). Human samples were stained for CD3 (UCHT1), CD5 (UCHT2), CD11b (CBRM1/5), CD11c (3.9), CD19 (HIB19), CD45 (HI30), CD117 (104D2), CD127 (A019D5), FcεR1 (AER-37), and CRTH2 (BM16).

For intracellular transcription factor or cytokine staining, cells were stained for surface markers, followed by fixation and permeabilization before nuclear factor staining according to the manufacturer’s protocol (Foxp3 staining buffer set from Thermo Fisher Scientific). For intracellular cytokine staining, cells were incubated for 4 hours at 37°C in complete medium [DMEM with 10% FBS, 10 mM Hepes, 1 mM sodium pyruvate, nonessential amino acids, 80 μM 2-mercaptoethanol, penicillin (100 U/ml), and streptomycin (100 μg/ml), all from Gibco] and supplied with phorbol 12-myristate 13-acetate (50 ng/ml), ionomycin (750 ng/ml), and brefeldin A (10 μg/ml) (all from Sigma-Aldrich). All flow cytometry experiments were performed using a Fortessa flow cytometer and FACSDiva software (BD Biosciences) and analyzed with FlowJo V10 software (TreeStar) or sort-purified by using a FACSAria II cell sorter (BD Biosciences).

RNA sequencing analysis

ILC2s and ILC3s were sort-purified from small intestine of Rorc(γt)-GfpTG mice or BMAL1fl/fl and BMAL1ΔRorc littermates. Sorted cells were used to prepare RNA sequencing libraries by the Epigenomics Core at Weill Cornell Medicine using the Clontech SMARTer Ultra Low Input RNA Kit V4 (Clontech Laboratories). Sequencing was performed on an Illumina HiSeq 2500, yielding 50–base pair single-end reads.

Raw sequencing reads were demultiplexed with Illumina CASAVA (v1.8.2). Adapters were trimmed from reads using Flexbar (v2.4), and reads were aligned to the National Center for Biotechnology Information (NCBI) GRCm38/mm10 mouse genome using the STAR aligner (v2.5.2b) with default settings. Reads per gene were counted using Rsubread. Genes with at least 50 or more counts in at least two samples were tested for differential expression. Differential expression was assessed using DESeq2 version 1.22.2 with default parameters and with a false discovery rate (FDR) of 0.1. PCA was performed after applying the DESeq2 varianceStabilizingTransformation function, using the 500 genes with highest variance.

Quantitative PCR

IECs or sort-purified ILC3s were lysed in Buffer RLT (Qiagen) or single-cell lysis buffer (Clontech-Takara). RNA was extracted with RNeasy Mini kits (Qiagen) as per the manufacturer’s instructions. Reverse transcription of RNA was performed using Maxima Reverse Transcriptase or SuperScript VILO according to the protocols provided by the manufacturer (Thermo Fisher Scientific). Real-time PCR was performed on complementary DNA (cDNA) using SYBR Green Chemistry (Applied Biosystems). Reactions were run on a real-time PCR system (ABI7500; Applied Biosystems). Samples were normalized to Actb2, Hprt1, or GAPDH and displayed as a fold change or relative values compared with controls (see table S2 for detailed primer information).

In vivo administration of Abx

Ampicillin (0.5 mg/ml) (Santa Cruz Biotechnology) and gentamicin (0.5 mg/ml) (Gemini BioProducts) were continuously administered via drinking water for 2 weeks. Mice were 7 to 8 weeks of age when Abx treatment started.

In vitro stimulation with GSK805

GSK805 (small-molecule RORγt inhibitor) was purchased from EMD Millipore at a purity of ≥95%. ILC3s were sort-purified from SI-LP of Rorc(γt)-GfpTG mice and incubated for 6 hours at 37°C with GSK805 or control vehicle in complete medium, before proceeding with qPCR analysis. Bulk-isolated SI-LP cells from C57BL/6 mice were incubated for 4 hours at 37°C with GSK805 or control vehicle in complete medium; supplied with recombinant mouse IL-1β, IL-6, IL-12, and IL-23 (20 ng/ml each, all from Thermo Fisher Scientific); and analyzed by flow cytometry.

DSS-induced colitis

DSS (3.5 g) (MP Biomedicals) was dissolved in 100-ml drinking water. Six- to 7-week-old BMAL1fl/fl and BMAL1ΔRorc littermates were given ad libitum access to DSS-containing water for seven consecutive days; the water was changed to normal drinking water from day 8. The mice were sacrificed on day 13 for measuring colon length and histology analysis.

For histology, distal colonic tissues were fixed with 4% paraformaldehyde (Bioworld) and embedded in paraffin, and 5-mm sections were stained with hematoxylin and eosin by IDEXX BioResearch. Images were acquired using a Nikon Eclipse Ti microscope.

Statistical analysis

P values of mouse data sets were determined by one-way analysis of variance (ANOVA), unpaired or paired two-tailed Student’s t test with 95% confidence interval. For human data, significance was determined by a paired two-tailed Students’ t test with 95% confidence interval. All statistical tests were performed with GraphPad Prism V8 software. Data are shown as means ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

SUPPLEMENTARY MATERIALS

immunology.sciencemag.org/cgi/content/full/4/40/eaax1215/DC1

Fig. S1. Gating strategy for sorting immune cells from the small intestine of C57BL/6 mice.

Fig. S2. Circadian regulation of ILC3s by light cues and microbiota-derived signals.

Fig. S3. Construct design and experimental validation of BAC transgenic IL-22–eGFP reporter mice.

Fig. S4. BMAL1ΔRorc mice have intact ILC2 and TH17 cell responses in the intestine and similar cellularity of ILC3s in lymphoid tissues.

Fig. S5. BMAL1 deficiency affects T-bet+ ILC3s in the large intestine.

Fig. S6. BMAL1 deficiency impairs intestinal ILC3s at weaning.

Fig. S7. Analysis of circadian genes in intestinal ILC3s from patients with IBD.

Fig. S8. A circadian clock is essential for homeostasis of ILC3s in the gut.

Table S1. Clinical metadata associated with the human samples.

Table S2. Primers used in the study.

REFERENCES AND NOTES

Acknowledgments: We thank members of the Sonnenberg Laboratory for discussions and critical reading of the manuscript. We would also like to thank the Epigenomics Core of Weill Cornell Medicine, G. G. Putzel, and the Center for Advanced Digestive Care (CADC). Funding: Research in the Sonnenberg Laboratory is supported by the NIH (R01AI143842, R01AI123368, R01AI145989, and U01AI095608), the NIAID Mucosal Immunology Studies Team (MIST), the Crohn’s and Colitis Foundation of America, the Searle Scholars Program, the American Asthma Foundation Scholar Award, Pilot Project Funding from the Center for Advanced Digestive Care (CADC), an Investigators in the Pathogenesis of Infectious Disease Award from the Burroughs Wellcome Fund, a Wade F.B. Thompson/Cancer Research Institute CLIP Investigator grant, the Meyer Cancer Center Collaborative Research Initiative, and the Jill Roberts Institute (JRI) for Research in IBD. G.F.S. is a CRI Lloyd J. Old STAR. L.Z. and J.G. are supported by fellowships from the Crohn’s and Colitis Foundation (608975 and 519428). Author contributions: F.T. and G.F.S. conceived the project. F.T. performed most experiments and analyzed the data. J.G., L.Z., and C.C. helped with experiments. M.A.S. provided human samples, scientific advice, and valuable expertise. G.E. provided essential mouse models, scientific advice, and expertise. F.T. and G.F.S. wrote the manuscript, with input from all the authors. Competing interests: The authors declare that they have no competing interests. Data and materials availability: RNA sequencing data have been deposited in the Gene Expression Omnibus database under the accession number GSE135359.
View Abstract

Stay Connected to Science Immunology

Navigate This Article