REV-ERB regulates RORγt⁺ regulatory T cell specification and function through the Bhlhe40-c-Maf axis

Abstract

Foxp3⁺ regulatory T (Treg) cells co-expressing RORγt adopt specialized functions to restrain intestinal inflammation. However, despite extensive characterization, the factors governing RORγt⁺Foxp3⁺ Treg specialization remain unclear. Here, we report that transcriptional repressor REV-ERB is critical for the differentiation and function of colonic RORγt⁺Foxp3⁺ Treg cells. REV-ERB deficiency exacerbates both TNBS- and oxazolone-induced intestinal inflammation. Mechanistically, REV-ERB promotes RORγt expression through suppressing the expression of transcriptional repressor Bhlhe40, which in turn inhibits c-Maf, a key factor promoting colonic RORγt⁺Foxp3⁺ Treg differentiation and function. Moreover, this Bhlhe40-c-Maf axis downstream of REV-ERB also regulates the expression of core colonic Treg signature genes including IL-10 and CTLA-4, while REV-ERB additionally safeguards RORγt⁺Foxp3⁺ Treg functional stability by directly suppressing proinflammatory cytokine IL-17A production. Collectively, the present study identifies that REV-ERB along with the downstream Bhlhe40-c-Maf axis jointly controls the RORγt⁺Foxp3⁺ Treg differentiation and suppressive function, suggesting that modulating their activities may strengthen RORγt⁺Foxp3⁺ Treg function to ameliorate inflammatory bowel diseases.

Introduction

Intestinal homeostasis requires a balance between effective immunity against invading pathogens and tolerance to harmless antigens present in commensal microorganisms and food (Belkaid and Hand, 2014; Honda and Littman, 2016; Kamada et al., 2013). Foxp3⁺ regulatory T (Treg) cells are a pivotal CD4 T cell subset that maintains intestinal homeostasis and prevents inflammatory bowel diseases (IBDs) by suppressing exuberant innate and adaptive immune responses (Tanoue et al., 2016; Whibley et al., 2019; Panduro et al., 2016). Intestinal Treg cells are a heterogeneous population and can be classified into thymus-derived natural Treg and periphery Treg (pTreg) cells, which are induced from conventional naïve CD4 T cells in response to microbial or food antigen stimulation. Notably, pTreg cells are substantially enriched in the colon, mainly expressing the retinoic acid–related orphan receptor (ROR)-γt (Ohnmacht et al., 2015; Sefik et al., 2015), a gene that was initially identified as a critical factor for Th17 cell differentiation and function (Ivanov et al., 2006). In addition to RORγt, c-Maf and STAT3 play critical roles in RORγt⁺Foxp3⁺ Treg cell differentiation as depletion of either gene leads to a dramatic reduction of these cells (Neumann et al., 2019; Xu et al., 2018; Ohnmacht et al., 2015). RORγt⁺Foxp3⁺ Treg cells express high levels of IL-10, CTLA-4, and ICOS and are more effective than RORγt⁻Foxp3⁺ Treg cells in restraining Th2 cell–mediated or Th1/Th17 cell–mediated colon inflammation in different models of colitis (Yang et al., 2016; Sefik et al., 2015; Ohnmacht et al., 2015). However, RORγt⁺Foxp3⁺ Treg cells could also co-express proinflammatory cytokines including IL-17, IFNγ, and TNFα, which are typically associated with Th17 cells during IBD progression in humans and mice (Quandt et al., 2021; Esposito et al., 2010; Kryczek et al., 2011). Intriguingly, some of RORγt⁺Foxp⁺ Treg cell’s features are Foxp3-independent as a genetic tracing study showed that lineage-committed RORγt⁺Foxp3^null Treg-like cells are capable of suppressing colonic T cell expansion (van der Veeken et al., 2022). Therefore, the underlying regulatory network that specifically induces RORγt expression, promotes RORγt⁺Foxp3⁺ Treg cell differentiation, and thus confers their distinct characteristics remains elusive. Furthermore, the safeguard mechanism that ensures RORγt plays a collaborative role with Foxp3 rather than to induce proinflammatory cytokine expression in RORγt⁺Foxp3⁺ Treg cells is poorly understood.

Nuclear receptors are a superfamily of highly conserved transcription factors that regulate a wide variety of physiological processes such as metabolism, reproduction, circadian rhythm, and immune function (Kojetin and Burris, 2014; Scholtes and Giguere, 2022; Yang et al., 2006; Glass and Ogawa, 2006). The nuclear receptor REV-ERBs (consisting of REV-ERBα and REV-ERBβ) often coregulate gene expression with RORs as they are often co-expressed and usually recognize similar ROR-response element (RORE) DNA sequence (Kojetin and Burris, 2014; Sever and Glass, 2013). Due to the lack of the activation function 2 region (AF-2) in the carboxy-terminal tail of their ligand binding domain, REV-ERB generally functions as a transcriptional repressor to suppress the expression of ROR target genes (Sever and Glass, 2013). In the immune system, REV-ERB regulates Th17 cell differentiation and function as reported by our group and others (Chang et al., 2019; Amir et al., 2018; Yu et al., 2013). Notably, previous studies indicate REV-ERBα expression is strongly associated with RORγt and c-Maf expression in colonic Treg cells, suggesting a potential role of REV-ERB in the differentiation and function of these cells (Sefik et al., 2015; Xu et al., 2018). However, whether and how REV-ERB regulates the differentiation and function of RORγt⁺Foxp3⁺ Treg cells remains unexplored.

In this study, we reveal that REV-ERBα is highly expressed in colonic RORγt⁺Foxp3⁺ Treg cells and is essential for their differentiation and function. Deletion of REV-ERB in Treg cells led to diminished colonic RORγt⁺Foxp3⁺ Treg cell populations at steady state, and rendered mice more susceptible to 2,4,6-trinitrobenzenesulfonic acid (TNBS) and oxazolone-induced colitis. As a transcriptional repressor, REV-ERBα can directly repress the expression of proinflammatory cytokines IL-17A, and promote RORγt expression through the Bhlhe40-c-Maf axis in RORγt⁺Foxp3⁺ Treg cells. Taken together, our data indicate that REV-ERB plays a critical role in RORγt⁺Foxp3⁺ Treg cell differentiation and function, and promotes intestinal homeostasis.

Results

REV-ERB is highly expressed in colonic RORγt⁺Foxp3⁺ Treg cells

To delineate the role of REV-ERB in Treg cells, we first examined the expression pattern of REV-ERBα and REV-ERBβ (encoded by Nr1d1 and Nr1d2, respectively) along with RORγt in Treg cells and CD4⁺ conventional T (Tcon) cells isolated from secondary lymphoid organs (SLOs), Peyer’s patch, and small intestinal and colonic lamina propria (siLP and cLP). Notably, REV-ERBα mRNA was highly upregulated in Treg cells of cLP, but not of siLP or SLOs, while. REV-ERBβ mRNA level was slightly elevated in Treg cells of both cLP and siLP (Fig. 1 A). Colonic Treg cells are a heterogeneous population, including Helios⁺RORt⁻ Foxp3⁺ Treg cells mostly generated in the thymus and Helios⁻RORγt⁺ Foxp3⁺ Treg cells mostly derived from naïve T cells under the influence of gut commensal microbiota (Tanoue et al., 2016; Zhang et al., 2025). To further assess whether REV-ERB has preferential expression in RORγt⁺Foxp⁺ Treg cells, we compared the REV-ERB expression in colonic RORγt⁺Foxp3⁺ Treg, RORγt⁻Foxp3⁺ Treg, RORγt⁺Foxp3⁻ Tcon, and RORγt⁻Foxp3⁻ Tcon cells sorted from Rorc^GFP:Foxp3^Thy1.1 double reporter mice (Ivanov et al., 2006; Liston et al., 2008). As expected, we observed a higher expression level of Foxp3 in RORγt⁺Foxp3⁺ and RORγt⁻Foxp3⁺ Treg cells, and that of Rorc in RORγt⁺Foxp3⁺ Treg and RORγt⁺Foxp3⁻ Tcon cells (Fig. 1 B). Importantly, the. mRNA level of Nr1d1 was remarkably elevated in RORγt⁺Foxp3⁺ Treg cells compared with RORγt⁻Foxp3⁺ Treg cells (Fig. 1 B), in agreement with a previous study showing that RORγt⁺Foxp3⁺ Treg cells express a higher level of Nr1d1 mRNA than RORγt⁻Foxp3⁺ Treg cells (Sefik et al., 2015). Additionally, the mRNA of Nr1d1 was also increased in RORγt⁺Foxp3⁻ Tcon (Th17) cells (Fig. 1 B), consistent with that REV-ERBα is critical for Th17 cell development and function as reported by us and others (Chang et al., 2019; Amir et al., 2018).

Figure 1. REV-ERB deficiency leads to reduced size in colonic RORγt⁺Foxp3⁺ Treg cell population.

(A) RT-qPCR analysis of mRNA expression of Nr1d1, Nr1d2, and Rorc in Treg and Tcon cells isolated from indicated organs of Foxp3^Thy1.1 reporter mice by flow cytometry. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. (B) RT-qPCR analysis of mRNA expression of Nr1d1, Nr1d2, and Rorc in RORγt⁺Foxp3⁺, RORγt⁻Foxp3⁺ Treg cells and RORγt⁺Foxp3⁻, RORγt⁻Foxp3⁻ Tcon cells isolated from the colon of Rorc^GFP: Foxp3^Thy1.1 double reporter mice. **P < 0.01, NS (not significant). Data in A and B are representative of two independent experiments. Statistical significance was determined using two-way ANOVA. (C) scRNA-seq data from GSE240982 reanalyzed by UMAP recomputed to consider only Treg cells, color-coded by Louvain clusters (Seurat), and identified from signature expression of colonic Treg subsets: RORγt⁺ Treg, Helios⁺ Treg, DN Treg cells. (D) Violin plots of Nr1d1 and Nr1d2 expression in colonic Treg subsets. The violin represents the probability density at each value; each dot represents one cell. ****P < 0.0001. (E) Single-cell transcript levels of Nr1d1 and Nr1d2 illustrated in the UMAP plots. Transcript levels are color-coded as indicated. (F) Western blot analysis of REV-ERBα in colonic RORγt⁺Foxp3⁺, RORγt⁻Foxp3⁺ Treg cells and RORγt⁺Foxp3⁻, RORγt⁻Foxp3⁻ Tcon cells isolated from Rorc^GFP:Foxp3^Thy1.1 double reporter mice. Data are representative of two independent experiments. (G) Suppression of proliferation of CellTrace Violet-labeled wide-type (WT) naïve CD4⁺ responder T cells by WT and REV-ERB–deficient Treg cells. NS (not significant). Data are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA. (H) Flow cytometric analysis of Helios and RORγt expression in colonic Treg cells isolated from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/flFoxp3^cre mice (n = 7 mice per group). *P = 0.0466, **P = 0.0017. Data are representative of at least two independent experiments. Statistical significance was determined using Student’s t test analysis. Source data are available for this figure: SourceData F1.

We next evaluated the expression pattern of REV-ERB at single-cell resolution by exploring a recently published colonic Treg single-cell RNA sequencing (scRNA-seq) dataset (Zhu et al., 2024). Indeed, both Nr1d1- and Nr1d2-expressing cells were enriched in the RORγt⁺ Foxp3⁺ Treg cell population (Fig. 1, C–E). To further confirm whether the REV-ERB protein is also enriched in colonic RORγt⁺ Foxp3⁺ Treg cells, we performed western blot and found that the REV-ERBα protein was highly expressed in colonic RORγt⁺Foxp3⁺ Treg cells compared with RORγt⁻Foxp3⁺ Treg cells (Fig. 1 F). Altogether, the REV-ERB expression pattern in Treg cells suggested that REV-ERB may play a role in regulating colonic RORγt⁺Foxp3⁺ Treg cell differentiation and function.

REV-ERB deficiency leads to reduced colonic RORγt⁺Foxp3⁺ Treg cell population

To further investigate the role of REV-ERB in Treg cells in vivo, we generated mice with REV-ERB deficiency specifically in Treg cells by crossing Foxp3^YFP-cre mice with REV-ERBα^fl/fl/β^fl/fl mice (referred to as REV-ERB cKO mouse). As expected, the expression of the REV-ERBα protein in colonic Treg cells from REV-ERB cKO mice was abolished, confirming Treg-specific deletion of REV-ERBα (Fig. S1 A). The REV-ERB cKO mice appeared normal without any overt inflammatory disease phenotype. Compared with WT controls, the REV-ERB cKO mice had normal distributions of CD4⁺ T cell, CD8⁺ T cell, and Treg cell populations in the thymus and spleen, no abnormal T cell activation measured by the CD44⁺CD62L^lo activated/memory T cell populations, and no aberrant production of proinflammatory cytokines IFNγ and IL-17A by T cells (Fig. S1, B–G). In addition, the suppressive function of Treg cells was not affected by the deletion of REV-ERB as measured by the in vitro suppression assay (Fig. 1 G). These data collectively indicate that REV-ERB plays little role in the Treg cell development and REV-ERB–deficient Treg. function is largely normal in unchallenged mice.

We next examined whether REV-ERB deficiency led to changes in colonic Treg cell composition at the steady state as REV-ERB is highly expressed in colonic RORγt⁺ Treg cells. The frequency of total colonic Foxp3⁺ Treg cells, Foxp3 protein level per cell, and the frequency of IFNγ- and IL-17A–producing colonic Tcon cells were comparable between REV-ERB cKO mice and WT control mice (Fig. S1, H–J). In contrast, the frequency of RORγt⁺Foxp3⁺ colonic Treg cells was significantly reduced in REV-ERB cKO mice compared with their WT littermate controls, whereas. Helios⁺Foxp3⁺ Treg cells were concomitantly increased (Fig. 1 H). Altogether, the above data suggest that the elevated expression of REV-ERB is crucial for RORγt⁺Foxp3⁺ Treg cell homeostasis, albeit not to the extent that its deficiency in Treg cells led to spontaneous colon inflammation.

Mice with Treg-specific REV-ERB deficiency are more susceptible to colonic inflammation

RORγt⁺Foxp3⁺ Treg cells have been implicated in the better control of colon inflammation under different challenges (Ohnmacht et al., 2015; Sefik et al., 2015; Yang et al., 2016). Given that REV-ERB deficiency modestly reduced colonic RORγt⁺Foxp3⁺ Treg population even at steady state (Fig. 1 H), we next examined whether this deficiency leads to more significant defects in colonic Treg cells in a TNBS-induced colitis, a model of IBD (Wirtz et al., 2017). REV-ERB cKO and WT control mice were sensitized by skin TNBS exposure and challenged with intrarectal TNBS injection 7 days later to induce colon inflammation. Compared with WT controls, REV-ERB cKO mice showed severe body weight loss followed by a slower recovery (Fig. 2 A) and shorter colon length (Fig. 2 B). Collectively, these data showed that REV-ERB expression in Treg cells was crucial for the protection of mice from TNBS-induced colon inflammation.

Figure 2. Mice with Treg-specific REV-ERB deficiency are more susceptible to TNBS-induced colon inflammation.

(A) Body weight changes in REV-ERBα^fl/fl/β^fl/fl (n = 7) and REV-ERBα^fl/fl/β^fl/flFoxp3^cre mice (n = 5) after TNBS challenge. **P = 0.004820 (day 3 [D3]), *P = 0.018173 (day 4 [D4]). Data are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA. (B) Representative photos of colons (left) and length of colons (right) from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/flFoxp3^cre mice 4 days after TNBS challenge. Scale bars = 2 cm. *P = 0.0135. (C) Flow cytometric analysis and quantification of Foxp3⁺ Treg cells in the colonic CD4⁺ T cells. **P = 0.007563, *P = 0.0481. (D) Flow cytometric analysis of RORγt⁺Foxp3⁺ and Helios⁺Foxp3⁺ Treg cells in the colonic CD4⁺ T cells (n = 5–7 mice per group). ****P < 0.0001. (E) Flow cytometric analysis of IFNγ and IL-17 expression in the colon-infiltrating CD4⁺ T cells (n = 5 mice for ERBα^fl/fl/β^fl/fl, n = 7 mice for ERBα^fl/fl/β^fl/fl). ****P < 0.0001, *P = 0.0344. Data in B–E are representative of at least two independent experiments. Statistical significance was determined using Student’s t test analysis.

Furthermore, flow cytometric analyses of colonic Treg cells after TNBS challenge revealed that the frequency of total colonic Foxp3⁺ Treg cells from REV-ERB cKO mice was slightly higher, while the Foxp3 protein level was slightly lower compared with WT mice (Fig. 2 C). More strikingly, the frequency of RORγt⁺Helios⁻ Treg cells within the colonic Treg population was profoundly decreased in REV-ERB cKO mice, with a compensatory increase of RORγt⁻Helios⁺ Treg cells (Fig. 2 D), further indicating that REV-ERB is required for colonic RORγt⁺ Treg homeostasis in an inflammatory environment. Notably, we also observed a higher proportion of CD4⁺ Tcon cells producing IFNγ and IL-17 in REV-ERB cKO mice compared with WT mice (Fig. 2 E), suggesting that REV-ERB is important for colonic Treg cells to exert their immune suppressive function and to control gut inflammation.

We next utilized an oxazolone-induced colitis model to examine whether. REV-ERB is critical for colonic Treg functions. In contrast to a Th1/Th17-mediated response in TNBS-induced colitis, oxazolone-induced colitis is characterized by Th2 immune response (Wirtz et al., 2017). Compared with WT mice, REV-ERB cKO mice again displayed severe colon inflammation as indicated by more body weight loss, reduced colon length, and increased production of IL-13, as well as IFNγ, in colonic CD4⁺ Tcon cells. (Fig. S2, A–C). Furthermore, the frequency of colonic RORγt⁺Foxp3⁺ Treg cells was significantly reduced in REV-ERB cKO mice, while the frequency of Helios⁺Foxp3⁺ Treg cells was increased (Fig. S2, D and E), similar to mice challenged with TNBS-induced colitis. Consistently, the expression of c-Maf and CTLA-4 was reduced, while IL-17A expression was increased in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells (Fig. S2, F–H). These results further illustrate that REV-ERB expression is critical for RORγt⁺Foxp3⁺ Treg cells to suppress colonic inflammation.

Given that REV-ERB deficiency leads to reduction of RORγt⁺ Treg cells both at steady state and under challenges (Fig. 1 D; Fig. 2 D; and Fig. S2, D and E), we next examined whether REV-ERB deficiency affects colonic Treg cell survival and fitness. To do so, we mixed equal numbers of CD45.2⁺ WT or REV-ERB–deficient colonic Treg cells with CD45.1⁺ WT Treg cells, and co-transferred them with naïve CD45.1⁺CD4⁺ T cells into RAG1 KO recipient mice and analyzed the transferred cells 21 days later. Indeed, REV-ERB-deficient Treg cells are less abundant than WT Treg cells in the colon but not in the spleen, with a much more profound reduction of RORγt⁺ Treg cells (Fig. S3, A–C). Thus, REV-ERB deficiency leads to a defect in the survival of colonic RORγt⁺Helios⁻ Treg cells.

REV-ERB regulates the expression of core signature genes in colonic RORγt⁺Foxp3⁺ Treg cells

To investigate how REV-ERB regulates the transcriptional programs in Treg cells, we performed RNA-seq analysis of colonic Treg cells isolated from WT and REV-ERB cKO mice after TNBS challenge. We identified 1030 upregulated genes and 952 downregulated genes (false discovery rate [FDR] <0.05) in REV-ERB–deficient Treg cells compared with WT controls (Fig. 3 A). REV-ERB–deficient colonic Treg cells expressed a lower level of Treg core genes including. Foxp3, Ctla4, Icos, Gzmb, Il10, Lag3, Nt5e (encoding CD73), Entpd1 (encoding CD39), Il1rl1 (encoding IL33R), Lrrc32 (encoding GARP), and Itgae (encoding CD103). Furthermore, the expression of genes encoding RORγt and c-Maf, a transcription factor critical for normal RORγt⁺Foxp3⁺ Treg cell function and differentiation, was also reduced in REV-ERB–deficient Treg cells (Fig. 3 A). Previous studies showed that although they are generally stable and exert enhanced suppressive function, colonic RORγt⁺Foxp⁺ Treg cells could be proinflammatory in individuals with IBD (Quandt et al., 2021). In agreement with these observations, we found that the genes. encoding proinflammatory cytokines, including Il17a, Il17f, Il4, and Il5, were significantly upregulated in REV-ERB–deficient colonic Treg cells (Fig. 3 A). To assess whether REV-ERB has a global impact on colonic Treg signature gene expression, we performed gene set enrichment analysis (GSEA) using a published gene set of 364 colonic Treg signature transcripts identified previously (Sefik et al., 2015). Indeed, these colonic Treg signature transcripts were significantly enriched in WT colonic Treg cells than REV-ERB–deficient colonic Treg cells (Fig. 3 B). Next, we extended to examine whether REV-ERB regulates RORγt-dependent transcripts in colonic Treg cells. Again, GSEA revealed that RORγt-dependent signature genes were preferably enriched in WT colonic Treg cells (Fig. 3 B), suggesting that REV-ERB is required for colonic RORγt⁺Foxp3⁺ Treg signature gene expression.

Figure 3. REV-ERB regulates the expression of colonic Treg core signature genes.

(A) Volcano plot of differentially expressed genes comparing REV-ERB–deficient and WT colonic Treg cells. (B) GSEA plot for the “colonic Treg signature” (upper) and “RORγt signature in colonic iTreg” (lower) of related gene expression in colonic Treg cells from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/fFoxp3^cre mice. (C and D) Flow cytometric analysis of c-Maf (C) and CTLA-4 (D) expression in RORγt⁺Foxp3⁺ and RORγt⁻Foxp3⁺ Treg cells from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/flFoxp^cre mice (n = 5 mice per group); ****P < 0.0001 for c-Maf and CTLA-4 expression in RORγt⁺Foxp3⁺ cells, **P = 0.0021 for CTLA-4 expression in RORγt⁻Foxp3⁺ cells. Data are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA. (E and F) Flow cytometric analysis of IL-10 (E) and IL-17A (F) in the colonic Treg cells from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/flFoxp3^cre mice (n = 5 mice per group). **P = 0.0089, ***P = 0.0002. Data are representative of at least two independent experiments. Statistical significance was determined using Student’s t test analysis.

In line with the RNA-seq data, flow cytometric analyses verified that the protein level of c-Maf was significantly reduced in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells compared with their WT counterparts, but not in RORγt⁻Foxp3⁺ Treg cells (Fig. 3 C), consistent with its critical role in these cells. Furthermore, CTLA4 and IL-10, two genes that are critical for colonic Treg cells’ suppressive function, were significantly decreased in protein levels in colonic Treg cells from REV-ERB cKO. mice (Fig. 3, D and E). Conversely, the production of IL-17A in colonic Treg cells was significantly increased in REV-ERB–deficient colonic Treg cells compared with WT controls (Fig. 3 F). Overall, these data suggested that REV-ERB has a global impact on RORt⁺Foxp3⁺ Treg cell transcriptome by promoting the expression of Treg cell signature genes and transcription factors that are crucial for their differentiation and function, while inhibiting the production of proinflammatory cytokines.

REV-ERB coordinates with RORγt to regulate core gene expression in RORγt⁺Foxp3⁺ Treg cells

To decipher the molecular mechanism of how REV-ERB regulates gene expression in RORγt⁺Foxp3⁺ Treg cells, we developed an in vitro culture system to generate a large number of RORγt⁺Foxp3⁺ Treg cells from naïve T cells (Fig. S4, A–C). Although RORγt and Foxp3 can co-express stably in colonic RORγt⁺Foxp3⁺ Treg cells, it is known that RORγt and Foxp3 reciprocally antagonize each other during in vitro T cell differentiation in which Th17 and iTreg represent alternative cell fates (Korn et al., 2009). To overcome these obstacles and recapitulate the features of colonic RORγt⁺Foxp3⁺ Treg cells in vitro, we tweaked the standard TGF-β/IL-2 iTreg differentiation condition by adding vitamin C and IL-6 to induce the expression of Foxp3 and RORγt simultaneously (Fig. S4 A). Vitamin C maintains stable Foxp3 expression by keeping the CNS2 enhancer region of the Foxp3 locus in a hypo-demethylated state (Sasidharan Nair et al., 2016), while IL-6 is important for the induction of RORγt expression (Fig. S4 B). After optimization of the dose and duration of these cytokines, we were able to efficiently differentiate CD4⁺ naïve T cells into RORγt⁺Foxp3⁺ Treg cells in vitro (Fig. S4 C).

Given the high expression levels of REV-ERB in colonic RORγt⁺ Foxp3⁺ Treg cells (Fig. 1, A–F), we then investigated the transcription factor regulating REV-ERB expression using the newly established in vitro RORγt⁺ Treg differentiation system. In addition to RORγt, we also assessed STAT3’s role in REV-ERB expression since. STAT3. is known to drive RORγt⁺Foxp3⁺ Treg differentiation (Ohnmacht et al., 2015). We cultured naïve T cells in RORγt⁺ Treg differentiation conditions with concurrent CRISPR/Cas9-mediated knockout of STAT3 or RORγt. Differentiated Treg cells were sorted by FACS, and mRNA levels of Rorc, Stat3, Nr1d1, and Nr1d2 were determined by quantitative RT-PCR. As shown in Fig. S4 D, knockout of STAT3 did not affect Nr1d1 expression, whereas knockout of RORγt indeed reduced the mRNA level of Nrıd1 (Fig. S4 D). Consistently, RORγt binding peaks are located at the Nr1d1 locus in RORγt⁺ Treg cells differentiated in vitro (Fig. S4 E). Furthermore, we found RORγt binding to the Nr1d1 locus in Th17 cells by analyzing a published chromatin immunoprecipitation (ChIP)–seq dataset (GSE40918) (Ciofani et al., 2012) (Fig. S4 F). These in vitro data suggest that RORγt might directly regulate REV-ERB expression in Treg cells, therefore forming a positive feedback loop with REV-ERB in colonic Treg cells.

Next, we examined whether the in vitro–generated RORγt⁺Foxp3⁺ Treg cells could mirror the deficiency of RORγt⁺Foxp3⁺ Treg cells isolated from REV-ERB cKO mice. To this end, naïve CD4⁺ T cells from REV-ERB cKO mice and WT control mice were cultured in RORγt⁺Foxp3⁺ Treg differentiation condition for 4 days, and their induction efficiency was measured by FACS. REV-ERB deficiency led to a significant reduction in the induction of total Foxp3⁺ Treg cells. Interestingly, only the induction of RORγt⁺Foxp3⁺ Treg cells was reduced, while the induction of RORγt⁻Foxp3⁺ Treg cells was comparable between WT and REV-ERB–deficient Treg cells (Fig. 4 A). In a complementary experiment, we observed that activation of REV-ERB by its agonist SR9009 increased RORγt⁺Foxp3⁺ Treg cell differentiation in WT T cells but not in REV-ERB–deficient T cells, confirming that REV-ERB plays an important role in RORγt⁺Foxp3⁺ Treg cell differentiation (Fig. S5 A). Importantly, the decreased expression of c-Maf and IL-10 and increased production of IL-17A were observed in these in vitro–differentiated REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells (Fig. 4, B–D). These results suggested that the in vitro–generated RORγt⁺Foxp3⁺ Treg cells can recapitulate the key features of colonic RORγt⁺Foxp3⁺ Treg cells.

Figure 4. REV-ERB coordinates with RORγt to regulate core gene expression in RORγt⁺Foxp3⁺ Treg cells.

(A) Naïve CD4⁺ T cells from REV-ERBα^fl/fl/β^fl/fl and REV-ERBα^fl/fl/β^fl/flFoxp3^cre mice were cultured in RORγt⁺Foxp3⁺ iTreg cell differentiation conditions. 4 days later, flow cytometric analysis was performed to determine Foxp3 and RORγt expression in differentiated cells. **P = 0.0094 (%Foxp3⁺), ***P = 0.0001 (%RORγt⁺ in Foxp3⁺ iTregs). Data are representative of at least two independent experiments. Statistical significance was determined using Student’s t test analysis (Foxp3⁺ %) or two-way ANOVA (%RORγt⁺ in Foxp3⁺ iTregs). (B–D) Flow cytometric analysis of c-Maf, IL-10, and IL-17A expression in differentiated T cells (n = 3 biologically independent replicates per group). **P = 0.0024 (B), *P = 0.017116 (C), **P = 0.001697 (D). Data are representative of at least two independent experiments. Statistical significance was determined using Student’s t test analysis. (E) Venn diagrams of overlapping REV-ERBα and RORγt binding sites (numbers of peaks) in the genomic regions of WT and REV-ERB–deficient Treg cells. (F) CUT&RUN tracks showing H3K27Ac- and REV-ERBα–bound peaks at Il17a-Il17f loci, with underlines detected as peaks by HOMER. (G) Heatmap of RORγt CUT&RUN signals ± 3 kb centered on RORγt-bound sites in WT or REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells, ranked according to read density. (H) Histogram of RORγt CUT&RUN read density ± 2 kb around the center of the RORγt binding sites showing that ectopic expression of RORγt largely restored RORγt binding in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells.

To dissect the molecular mechanism underlying REV-ERB control of RORγt⁺Foxp3⁺ Treg gene expression, we performed REV-ERBα and RORγt CUT&RUN experiments to identify their genome-wide bound peaks in the in vitro–differentiated WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells. CUT&RUN (short for “Cleavage Under Targets & Release Using Nuclease”) is a method used to map specific interactions between proteins and DNA in a way that overcomes the limitations of conventional ChIP method (Skene and Henikoff, 2017). CUT&RUN data analyses revealed that there were 8858 REV-ERBα binding sites in WT RORγt⁺Foxp3⁺ Treg cells, whereas only 70 were found in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells (Fig. 4 E), validating the robustness of the CUT&RUN assay. Notably, we observed many REV-ERBα binding sites in the Ill7a-Il17f locus, with strong overlapping with H3K27Ac peaks (Fig. 4 F). Considering that REV-ERB functions mainly as a transcriptional repressor, and that the proinflammatory cytokine IL-17A was significantly elevated in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells, the binding of REV-ERB to the Il17a-Il17f loci, together with stronger H3K27Ac signals in REV-ERB–deficient Treg cells, suggests that REV-ERB directly suppresses IL-17A and IL-17F expression in RORγt⁺Foxp3⁺ Treg cells and prevents them from eliciting colon inflammation.

REV-ERB often competes with RORγt binding to the same RORE motif and coregulates their shared gene expression. However, RORγt CUT&RUN data analyses revealed that the total RORγt binding sites dramatically decreased from 6,139 in WT RORγt⁺Foxp3⁺ Treg cells to 1,865 in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells. (Fig. 4, E and G). Given that REV-ERB. deficiency leads to decreased RORγt expression, we next tested whether this reduction of RORγt binding sites was due to the reduced expression of RORγt. To this end, we ectopically expressed RORγt in WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells and performed RORγt CUT&RUN experiments. Analysis of the CUT&RUN data revealed that the overexpression of RORγt can essentially restore the RORγt binding in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells, to a similar level of that in WT RORγt⁺Foxp3⁺ Treg cells with vector control, yet still lower than RORγt binding in WT. RORγt⁺Foxp3⁺ Treg. cells. overexpressed. with RORγt (Fig. 4 H). Therefore, the decreased expression of RORγt largely accounted for the dramatic reduction of RORγt binding in REV-ERB–deficient Treg cells. Collectively, these data suggest that REV-ERB ensures the immune regulatory functions. of RORγt⁺Foxp⁺ Treg cells by both repressing the expression of RORγt-dependent genes encoding proinflammatory cytokines IL-17A and IL-17F, and coordinating with RORγt to regulate the expression of core genes in RORγt⁺Foxp3⁺ Treg cells.

Bhlhe40 is a critical repressor downstream of REV-ERB to induce RORγt⁺ Treg cell differentiation

We next sought to assess how REV-ERB promotes RORγt expression in RORγt⁺Foxp3⁺Treg cells. Since REV-ERB is a transcriptional repressor, it likely promotes RORγt expression by suppressing another repressor that regulates RORγt expression directly. REV-ERB was reported to regulate RORγt expression in Th17 cells via the repression of transcription factor NFIL3 by our group and others (Chang et al., 2019; Yu et al., 2013). However, NFIL3 was not differentially expressed in WT and REV-ERB–deficient colonic Treg cells, ruling out NFIL3’s role in REV-ERB-dependent RORγt expression in Treg cells (Fig. 5, A and B). To search for other repressors downstream of REV-ERB that can regulate RORγt expression, we cross-referenced genes associated with REV-ERBα binding peaks with genes upregulated in REV-ERB–deficient RORγt⁺ Treg cells, resulting in a list of 303 genes (Fig. 5 A). Only one transcriptional repressor, Bhlhe40, stood out in this list because its expression increased significantly in REV-ERB–deficient Treg cells, its gene locus was bound by REV-ERBα, and H3K27Ac signals were elevated in REV-ERB–deficient RORγt⁺ Treg cells (Fig. 5, B and C), suggesting that Bhlhe40 could be a key repressor downstream of REV-ERB to regulate RORγt expression.

Figure 5. Bhlhe40 is a critical repressor downstream of REV-ERB that regulates RORγt⁺ expression in Treg cells.

(A) Venn diagram of overlapping REV-ERBα–bound genes from CUT&RUN and genes with upregulated expression in REV-ERB–deficient Treg cells from RNA-seq. (B) RT-qPCR analysis of Bhlhe40 and Nfil3 expression in WT and REV-ERB–deficient Treg cells. ***P = 0.0002. Data are representative of two independent experiments. Statistical significance was determined using Student’s t test analysis. (C) CUT&RUN tracks showing H3K27Ac- and REV-ERBα–bound peaks at the Bhlhe40 locus, with underlines detected as peaks by HOMER. (D) Flow cytometric analysis of RORγt expression in Foxp3⁺ iTreg cells with ectopic expression of Bhlhe40 or control vector (n = 3 biologically independent replicates per group). ****P < 0.0001, *P = 0.0131. (E) Flow cytometric analysis of RORγt expression in Foxp3⁺ iTreg cells transduced with sgNT and sgBhlhe40. **P = 0.0094, ****P < 0.0001. Data in D and E are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA.

Bhlhe40 is increasingly recognized as a crucial regulator for an expanding set of immune cell populations(Cook et al., 2020; Lin et al., 2016; Li et al., 2019; Lin et al., 2014; Miyazaki et al., 2010). To test whether Bhlhe40 regulates RORγt⁺ Treg cell differentiation, we first assessed the impact of ectopic BHLHE40. expression on RORγt⁺ Treg cell differentiation. Flow cytometric analyses showed that an increase in Bhlhe40 expression significantly inhibited the induction of RORγt⁺ Treg cells from WT T cells, while it only mildly affected RORγt⁺Foxp3⁺ Treg differentiation from REV-ERB–deficient T cells (Fig. 5 D). In a complementary experiment, we employed CRISPR technology in which retrovirus expressing Cas9 and guide RNAs (gRNAs) targeting Bhlhe40 was transduced into WT and REV-ERB cKO naïve T cells during RORγt⁺Foxp3⁺ Treg differentiation. While naïve T cells from REV-ERB cKO mice transduced with control gRNA were deficient in the generation of RORγt⁺ Treg cells, transduction with a Bhlhe40 targeting gRNA fully rescued the RORγt expression defects in REV-ERB–deficient T cells to a comparable level in WT T cells (Fig. 5 E). To further determine whether the above Bhlhe40 targeting could restore Treg immunosuppressive function, we performed an in vitro suppression assay using RORγt⁺Foxp3⁺ Treg cells transduced with control gRNA, REV-ERB gRNAs, Bhlhe40 gRNA, or REV-ERB/Bhlhe40 gRNAs. Targeting REV-ERB modestly impaired Treg suppression, whereas Bhlhe40 knockdown significantly enhanced it. Strikingly, dual REV-ERB/Bhlhe40 targeting rescued the suppression defect in RORγt⁺Foxp3⁺ Treg cells transduced with REV-ERB gRNAs, restoring Treg function to control levels (Fig. S5, B and C). Taken together, these data suggested that Bhlhe40 is a key transcriptional repressor downstream of REV-ERB that regulates RORγt⁺Foxp3⁺ Treg cell differentiation.

Bhlhe40-c-Maf axis is essential for RORγt and core signature gene expression in RORγt⁺Foxp3⁺ Treg cells

To further assess whether Bhlhe40 directly regulates RORγt expression in RORγt⁺ Treg cells, we performed the Bhlhe40 CUT&RUN assay in WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells. In agreement with the increased expression of Bhlhe40 in REV-ERB–deficient RORγt⁺ Treg cells, CUT&RUN analysis revealed that there were 12274. Bhlhe40. binding sites in REV-ERB–deficient RORγt⁺ Treg cells, while there were only 4,269 in WT RORγt⁺Foxp3⁺ Treg cells, with 2,581 shared sites between them (Fig. 6 A). However, we did not observe Bhlhe40 directly binding to the Rorc locus in RORγt⁺ Treg cells (Fig. 6 B). Instead, strong Bhlhe40-bound peaks were detected in the promoter of the c-Maf gene (Fig. 6 C), in line with a prior study identifying a parallel regulation of c-Maf by Bhlhe40 in macrophages (Jarjour et al., 2019). Consistently, we observed that sgRNA knockdown of Bhlhe40 led to a significant increase in c-Maf expression in both REV-ERB–deficient and WT T cells (Fig. 6 D), suggesting that Bhlhe40 could suppress c-Maf expression directly.

Figure 6. Bhlhe40 regulates RORγt⁺Foxp3⁺ Treg cell differentiation through c-Maf.

(A) Venn diagram of overlapping BHLHE40-bound sites in WT and REV-ERB-deficient RORγt⁺Foxp3⁺ iTreg cells. (B and C) CUT&RUN tracks showing H3K27Ac- and BHLHE40-bound peaks at Rorc (C) and Maf (D) loci in RORγt⁺Foxp3⁺ Treg cells, with underlines detected as peaks by HOMER. (D) Flow cytometric analysis of c-Maf expression in Foxp3⁺ iTreg cells transduced with sgNT and sgBhlhe40 (n = 3 biologically independent replicates per condition). ***P = 0.001 (REV-ERBα^fl/fl/β^fl/fl), ****P < 0.0001 (REV-ERBα^fl/fl/β^fl/flFoxp3^cre). Data are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA. (E) Venn diagram of overlapping c-Maf–bound sites in WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells. (F) CUT&RUN tracks showing H3K27Ac- and c-Maf–bound peaks at the Rorc locus in RORγt⁺Foxp3⁺ Treg cells, with underlines detected as peaks by HOMER. (G) GSEA plots of c-Maf–upregulated and c-Maf–downregulated genes in WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells.

Given that c-Maf expression was reduced in REV-ERB–deficient Treg cells and Bhlhe40 inhibited c-Maf expression, we postulated that c-Maf promotes RORγt expression. To test this possibility, we performed c-Maf CUT&RUN experiment, which revealed 6401 c-Maf binding sites in WT Treg cells and 4055 binding sites in REV-ERB–deficient Treg cells, probably due to decreased c-Maf expression in REV-ERB–deficient Treg cells (Fig. 6 E). More importantly, c-Maf, but not REV-ERB or Bhlhe40, binds to the Rorc locus, suggesting that c-Maf directly regulates RORγt expression (Fig. 6 F). Consistently, GSEA revealed that c-Maf–upregulated genes were significantly enriched in the transcriptome of WT RORγt⁺Foxp3⁺ Treg cells, whereas c-Maf–downregulated genes were significantly enriched in the transcriptome of REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells (Fig. 6 G).

In addition to RORγt, the expression of other colonic Treg signature genes, such as IL-10, was also reduced in REV-ERB–deficient Treg cells both in vivo and in vitro. IL-10 is a critical anti-inflammatory cytokine that maintains gut homeostasis (Chaudhry et al., 2011). Considering the potent roles of Bhlhe40 (Huynh et al., 2018; Lin et al., 2014; Jarjour et al., 2019; Yu et al., 2018) and c-Maf (Gabryšová et al., 2018; Xu et al., 2009; Liu et al., 2018) to either inhibit or promote IL-10 expression in various. immune cell types, we next explored whether the Bhlhe40-c-Maf axis regulated IL-10 expression in RORγt⁺Foxp3⁺ Treg cells. Bhlhe40 CUT&RUN data revealed that multiple Bhlhe40 peaks were present around the Il10 locus, with some of them displaying a slightly higher binding intensity in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells (Fig. 7 A). Further examination of c-Maf binding revealed a significant reduction in REV-ERB–deficient RORγt⁺ Treg cells compared with WT Treg cells (Fig. 7 A). Consistently, WT RORγt⁺ Treg cells displayed modestly stronger H3K27Ac signals around the Il10 locus (Fig. 7 A). Together, these data suggest that the expression of IL-10 in RORγt⁺Foxp3⁺ Treg cells is coregulated by REV-ERB downstream factors, Bhlhe40 and c-Maf, with c-Maf playing a dominant role in directly upregulating IL-10 expression, while c-Maf’s expression is controlled by Bhlhe40. In support of this mechanism, CRISPR knockdown of Bhlhe40 significantly increased IL-10 expression, while overexpression of Bhlhe40. decreased IL-10 expression in both REV-ERB–deficient and WT T cells (Fig. 7, B and C). Taken together, the Bhlhe40-c-Maf axis downstream of REV-ERB plays critical roles in the differentiation of RORγt⁺Foxp3⁺ Treg cells and regulating their signature gene expression.

Figure 7. Bhlhe40-c-Maf axis is essential for IL-10 expression in RORγt⁺Foxp3⁺ Treg cells.

(A) CUT&RUN tracks showing H3K27Ac peaks, and BHLHE40-bound and c-Maf-bound peaks at the II10 locus between WT and REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells, with underlines detected as peaks by HOMER. (B) Flow cytometric analysis of IL-10 expression in Foxp3⁺ iTreg cells transduced with sgNT and sgBhlhe40. ***P = 0.0003, ****P < 0.0001. (C) Flow cytometric analysis of IL-10 expression in Foxp3⁺ iTreg cells with the ectopic expression of BHLHE40 or control vector (n = 3 biological replicates per group). ***P = 0.002. Data in B and C are representative of at least two independent experiments. Statistical significance was determined using two-way ANOVA.

Discussion

In the present study, we uncovered an essential role of REV-ERB. in the regulation of colonic RORγt⁺Foxp3⁺ Treg cell differentiation and function. In the steady state, Treg-specific deletion of REV-ERB specifically impaired colonic RORγt⁺Foxp3⁺ Treg cells, but not RORγt⁻Foxp3⁺ Treg cells. Furthermore, REV-ERB deficiency undermined RORγt⁺Foxp3⁺ Treg capability to suppress T cell–mediated colon inflammation in colitis models. In addition to suppressing the expression of proinflammatory cytokines IL-17A and IL-17F, REV-ERB promotes RORγt expression in Treg cells via suppressing the expression of the transcriptional repressor Bhlhe40, highlighting an unrecognized repressor-of-repressor mechanism (Choi et al., 2020) in regulating RORγt⁺Foxp3⁺ Treg cell differentiation and function. As Bhlhe40 is a critical regulator of circadian rhythm, our study further suggests that colonic RORγt⁺Foxp3⁺ Treg functionality is under diurnal regulation, arguing that cell-intrinsic circadian rhythm is a key feature of tissue Treg cells, as recently discovered in visceral adipose fat Treg cells (Xiao et al., 2022; Hemmers and Rudensky, 2015).

While it was reported that RORγt and Foxp3 have mutually antagonistic functions in an earlier study (Zhou et al., 2008), RORγt⁺Foxp3⁺ Treg cells co-expressing RORγt and Foxp3 are now generally believed to be a stable Treg subset with enhanced suppressive function (Yang et al., 2016). However, these cells can still express IL-17 in mice and humans, suggesting that negative regulators might play a role to counteract RORγt-mediated IL-17A/F cytokine expression in RORγt⁺Foxp3⁺ Treg cells. Our data clearly showed that REV-ERB is such a repressor that directly binds to Il17a-Il17f locus and likely represses their expression through these key binding/regulatory sites, thus preventing RORγt⁺Foxp3⁺ Treg cells from eliciting inflammation-associated pathology. Notably, we previously reported a similar role of REV-ERB in inhibiting IL-17A/F expression in Th17 cells, indicating that the antagonism relationship between REV-ERB and ROR on IL-17 expression is conserved in these two different T cell populations.

Our data also highlighted that REV-ERB positively regulates RORγt expression in RORγt⁺Foxp3⁺ Treg cells via repressing the transcriptional repressor. Bhlhe40, which in turn represses c-Maf, a positive regulator for RORγt⁺Foxp3⁺ Treg cell differentiation. Notably, REV-ERB regulates RORγt expression in Th17 cells via repressing a different transcription repressor NFIL3 (Yu et al., 2013), reflecting the intrinsic difference between these two cell populations. Furthermore, by taking advantage of our newly developed in vitro RORγt⁺Foxp3⁺ Treg cell differentiation system and the robustness of the CUT&RUN assay, we showed that REV-ERB can directly bind to the Bhlhe40 locus, which in turn binds to the c-maf locus. While previous studies identified a critical role of c-Maf for RORγt⁺Foxp3⁺ Treg differentiation, our CUT&RUN data further demonstrated that c-Maf indeed binds to the Rorc locus. These data suggest that REV-ERB regulates RORγt expression by inhibiting the expression of Bhlhe40, a negative regulator of c-Maf, and subsequently RORγt. In addition to RORγt expression, the CUT&RUN data also revealed RORγt-bound peaks were dramatically reduced in REV-ERB–deficient RORγt⁺Foxp3⁺ Treg cells, which might be explained by a moderate decrease in RORγt expression. Since our data also suggest that RORγt could directly regulate REV-ERB expression in Treg cells, it is plausible that the reciprocal regulation of RORγt and REV-ERB forms a positive feedback loop in colonic Treg cells.

Our data also revealed Bhlhe40 as a critical regulator downstream of REV-ERB in RORγt⁺Foxp3⁺ Treg cells. In a loss-of-function study, CRISPR-mediated deletion of Bhlhe40 rescued the defective RORγt⁺Foxp3⁺ Treg cell differentiation from REV-ERB–deficient T cells, while in a gain-of-function study, the ectopic expression of Bhlhe40 resulted in impaired differentiation of RORγt⁺Foxp3⁺ Treg cell from WT T cells. Furthermore, we showed that Bhlhe40 could directly bind to the c-maf and Il10 loci and inhibit their gene expression. Since both c-Maf and IL-10 are essential for normal colonic RORγt⁺ Treg cell function (Rubtsov et al., 2008), our study showed that Bhlhe40 is a key negative. regulator of RORγt⁺Foxp3⁺ Treg cells. A previous study using Bhlhe40 whole-body knockout mice indicates that Bhlhe40 contributes to maintaining Treg homeostasis through upregulation of CD25 (Miyazaki et al., 2010). Future studies need to assess the role of Bhlhe40 in RORγt⁺Foxp3⁺ Treg cells in vivo by utilizing Bhlhe40 conditional knockout mice.

In summary, we demonstrate here that nuclear receptor REV-ERB is a critical positive regulator of colonic RORγt⁺Foxp⁺ Treg cell differentiation and function. Although REV-ERB generally functions as a transcription repressor, REV-ERB promotes RORγt⁺Foxp3⁺ Treg cell differentiation by repressing the transcriptional repressor Bhlhe40, which in turn represses the expression of c-Maf, which positively regulates RORγt. Additionally, REV-ERB safeguards RORγt⁺Foxp3⁺ Treg’s immunoregulatory roles by inhibiting proinflammatory cytokine IL-17A/F production. Hence, our study provides mechanistic insight into how the REV-ERB-Bhlhe40-c-Maf axis modulates RORγt⁺Foxp3⁺ Treg cells, highlighting REV-ERB as a potential target for IBD. treatment.

Materials and methods

Mice

REV-ERBα^fl/fl/β^fl/fl mice (Cho et al., 2012), Foxp3^YFP-cre mice (Rubtsov et al., 2008), Foxp3^Thy1.1 (Liston et al., 2008), Rorc-GFP (Lochner et al., 2008), and Rosa-Cas9/Foxp3Thy1.1 mice (Liu et al., 2022) have been described previously. CD4cre transgenic, Ly5.1 congenic, and Rag1^−/− mice were purchased from the Jackson Laboratory. All mice used in the study are in the C57BL/6 genetic background. Mice were housed in specific pathogen-free facilities under a 12-h light/dark cycle, with ambient temperature of 20–26°C and humidity of 30–70% at the. Salk Institute for Biological Studies. Animal experiments were performed under the regulation of the Institutional Animal Care and Use Committee according to the institutional guidelines. To minimize circadian effects, we performed the experiments and harvested samples in the morning around 9–11:00 am unless otherwise stated.

Isolation of cLP lymphocytes

Mouse colon was opened longitudinally and rinsed in ice-cold wash buffer (1× Pen-Step, 2 mM EDTA, 20 mM HEPES in RPMI 1640). Afterward, the colon was cut into approximately 1-cm length sections. Dissociation of epithelial cells was performed by incubation of small colon sections with digestion buffer #1 (1× Pen-Step, 5% FBS, 5 mM EDTA, 1 mM DTT, 20 mM HEPES in RPMI 1640) with constant rotations at 250 RPM at 37°C for 30 min in a shaking incubator. Samples were then vortexed and washed with plain RPMI 1640 medium three times to remove the intraepithelial lymphocyte fraction. The remaining tissues were incubated in 10 ml digestion buffer #2 (1× Pen-Step, 20 mM HEPES, 0.1 mg/ml Liberase TL [Roche], 10 μl/ml DNase [Roche] in RPMI 1640) at 37°C with constant rotations at 250 RPM for 50 min in a shaking incubator. 10 ml of complete RPMI medium was added to stop the digestion. Afterward, the suspension was filtered through a sterile 100-μm cell strainer, and the undigested colon tissue was further ground with a syringe plunger and washed with another 20 ml of RPMI medium. The combined cell suspension was centrifuged at 350 g for 5 min at 4°C. To enrich lymphocytes, the cell pellets were resuspended in 8 ml of 44% Percoll solution, and loaded on top of 5 ml of 67% Percoll solution, centrifuged at 805 g for 20 min at 23°C without acceleration and brake. Cells in the middle layer were collected, pelleted, and resuspended in FACS buffer for staining and flow cytometric analysis.

Antibodies and flow cytometry

Single-cell suspension isolated from indicated organs was washed in cold FACS buffer and transferred to a round-bottom 96-well plate. Cell surface antibodies and Live/Dead Dye (Ghost Dye Red 780) were incubated in the cell suspension for 20 min at 4°C. For intracellular transcription factor staining, cells were then fixed and permeabilized using the Foxp3/Transcription Factor Staining Buffer Set (catalog# 00-5523-00; eBioscience). Antibodies against Foxp3, RORγt, Helios, c-Maf, and CTLA-4 were incubated for at least 1 h at 4°C. For intracellular cytokine staining, cells were cultured and restimulated with 50 ng/ml PMA and 1 μg/ml ionomycin for 1 h at 37°C followed by the addition of GolgiStop for another 4 h. Afterward, cell surface antigens were stained as indicated above, and then, cells were fixed and permeabilized using BD. Cytofix/Cytoperm Fixation/Permeabilization Solution Kit for 20 min at room temperature. Antibodies against IFNγ, IL-17, IL-4, IL-5, IL-13, and IL-10 were incubated for at least 1 h at 4°C. Flow cytometric analysis was performed on a BD FACS Aria II flow cytometer and analyzed using FlowJo software (Tree Star). Details of antibodies, viability dye, and dilutions are listed in Table S1.

TNBS- and oxazolone-induced colitis model

To induce colitis, mice were challenged either with TNBS or with oxazolone according to a previously published protocol (Wirtz et al., 2007). For TNBS-induced colitis, mice were shaved on the back and presensitized with 150ul 1% (wt/vol). TNBS solution (in 4:1 mixture of acetone and olive oil) to the skin. 7 days later, colitis was induced by intrarectal administration of 100 μl of 2.5% (wt/vol) TNBS solution (in 50% ethanol) by inserting a catheter into the colon 4 cm proximal to the anus.

For oxazolone-induced colitis, mice were presensitized by applying 150 μl of 3% (wt/vol) oxazolone (4-ethoxymethylene-2-phenyl-2-oxazoline-5-one, Aldrich) in acetone/olive oil on shaved skin. Mice were challenged intrarectally with 100 μl of 1% oxazolone in 50% ethanol under anesthesia 7 days after sensitization.

In vitro Treg suppression assay

Treg cells (CD4⁺CD25⁺Foxp3YFP⁺) were sorted from the spleen and lymph nodes of Foxp3^YFPCre and REV-ERBα^fl/fl/β^fl/flFoxp3^YFPCre (RVB cKO) mice by flow cytometry. Naïve CD4 T cells (CD4⁺CD25⁻CD44⁻CD62L^hi) were sorted from CD45.1⁺ congenic mice and then labeled with CellTrace Violet Cell Dye (catalog# C34557; Invitrogen). Treg and naïve CD4 T cells were mixed at different ratios in the presence of irradiated T cell–depleted splenocytes as antigen-presenting cells, and 1 μg/ml soluble anti-CD3 antibody to stimulate T cell proliferation. After 3 days of coculture, Treg suppression function was measured by the percentage of nondividing cells within the CD45.1⁺ effector T cell population. To assess REV-ERB and Bhlhe40 effects on RORγt⁺Foxp3⁺ iTreg immune suppressive function, RORγt⁺Foxp3⁺ iTreg cells transduced with viruses expressing gRNAs were sorted to perform an in vitro suppression assay as above.

Treg adoptive transfer experiment

To get enough colonic Treg cells, we used IL-2: IL-2 mAb (JES6–1A12) immune complex to boost Treg population. Then, WT and REV-ERB–deficient Treg cells (CD4⁺CD25⁺Foxp3YFP⁺) from the colon of Foxp3^YFPCre and REV-ERBα^fl/fl/β^fl/flFoxp3^YFPCre (RVB cKO) mice on a CD45.2 background were sorted respectively by flow cytometry, and mixed with Treg cells from CD45.1 (CD4⁺CD25⁺) background at a ratio of 1:1, and cotransferred with naïve CD4⁺CD45RB^hi T cells from CD45.1 mice into RAG KO recipients. Cells from the spleen and colon were harvested for analysis 21 days after transfer.

In vitro differentiation of RORγt⁺Foxp3⁺ iTreg cells

Naïve CD4⁺ T cells (CD4⁺CD25⁻CD44^loCD62L^hi) were isolated by depletion of CD25⁺CD44⁺ activated/memory cells followed by CD4-positive selection (Mouse CD4 Positive Selection Kit, catalog# 18952A; StemCell). To induce RORγt⁺Foxp3⁺ double-positive iTreg cells, the following condition was used unless stated otherwise: naïve CD4⁺ T cells were first polarized in the RPMI medium with 5 ng/ml TGFβ, 50 U/ml hIL-2, 2 μg/ml anti-CD3, 2 μg/ml anti-CD28, and 100 μg/ml vitamin C for 24 h; afterward, IL-6 (5 ng/ml) was added to the medium to induce RORγt expression until analysis. This condition was used for in vitro differentiation of RORγt⁺Foxp3⁺ iTreg cells. For some experiments, IL-6 was added after 48 h of differentiation to induce optimal c-Maf expression in RORγt⁺Foxp3⁺ iTreg cells as indicated.

CRISPR knockout experiments

Cloning of sgRNAs into the pSIRG-NGFR vector, and retrovirus production in HEK293T cells were performed as previously described (Liu et al., 2022). To examine the effects of REV-ERB, Bhlhe40 on RORgt+ Treg cell differentiation, immune suppressive function, or the effects of STAT3 and RORγt on REV-ERB expression, naïve CD4⁺ T cells from WT and REV-ERB cKO mice were simultaneously induced to differentiate into RORγt+ Treg cells, and transduced with retrovirus carrying Cas9 protein and the sgRNAs targeting the genes as indicated, or control non-targeting sgRNAs. Cells were subsequently analyzed by FACS or sorted for gene expression analysis. Details of sgRNA sequence are listed in Table S2.

RNA isolation, RT-qPCR, and RNA-seq

20,000–50,000 Treg or Tcon cells were FACS-sorted into TRIzol reagent (catalog# 15596018) from indicated organs, including the spleen, axillary and inguinal lymph nodes (pLN), mesentric lymph nodes (mLN), Peyer’s patches (PP), small intestine (SI), and colon by flow cytometry. RNA was then extracted and purified according to the manufacturer’s instructions. RNA concentration and integrity were determined by Bioanalyzer using RNA 6000 Pico Kit (Agilent). RNA-seq libraries were prepared from total RNA extracted above using Illumina TruSeq RNA Library Prep Kit v.2 following the manufacturer’s instructions, and paired-end sequencing (PE 150 bp) was performed on an Illumina NovaSeq sequencer.

For RT-qPCR, complementary DNA was synthesized with SuperScript IV. First-Strand Synthesis Kit (catalog# 18091050; Invitrogen), followed by qPCR on a ViiA 7 Real-Time PCR System using Power SYBR Green Master Mix (catalog# 4309155; Thermo Fisher Scientific). The relative quantification value was calculated as 2^−ΔCt relative to an internal control (Gapdh). Details of primer sequence are listed in Table S3.

Western blot

To assess REV-ERBα protein level in colonic RORγt⁺Foxp3⁺ Treg, RORγt⁻Foxp3⁺ Treg, RORγt⁺Foxp3⁻ Tcon, and RORγt⁻Foxp3⁻ Tcon cells, equal numbers of those cells were sorted from Rorc^GFP:Foxp3^Thy1.1 double reporter mice, and further lysed using radioimmunoprecipitation assay buffer (Pierce, Thermo Fisher Scientific) supplemented with a protease inhibitor cocktail (Sigma-Aldrich). All the extracts were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore) for immunoblot analysis with antibodies to REV-ERBα (generated in-house, Chang et al., 2019), Foxp3 (generated in-house, Liu et al., 2022), RORγt (catalog# 14–6988-82; Invitrogen), β-actin (catalog# SAB1305554–40TST; Sigma-Aldrich). Similarly, equal numbers of colonic Treg (CD4⁺Foxp3^YFP+) and Tcon (CD4⁺Foxp3^YFP−) cells that were sorted from WT and REV-ERB cKO mice were used to verify whether REV-REVα is specifically deleted in Treg cells.

CUT&RUN sample processing and sequencing

CUT&RUN samples were processed as described previously (Liu et al., 2023). Briefly, 300,000–500,000 in vitro–generated RORγt⁺Foxp3⁺ iTreg cells per biological replicate were collected into a V-bottom 96-well plate by centrifugation. Cells were washed twice in the antibody buffer (2 mM EDTA, 1× EDTA-free protease inhibitors, 0.5 mM spermidine, 1× permeabilization buffer from eBioscience Foxp3/Transcription Factor Staining Buffer Set), centrifuged at 1,900 rpm for 6 min at 4°C, and then incubated with H3K27Ac, Foxp3, RORγt, REV-ERBα, BHLHE40, or c-Maf antibodies on ice for 1 h. After two washes with buffer 1 (1× EDTA-free protease inhibitors, 0.5 mM spermidine, 1× permeabilization buffer), cell pellets were further incubated with pA/G-MNase enzyme (catalog# 15–1016; EpiCypher) at 1:20 dilution in 50 μl buffer 1 and incubated at 4°C for 1 h. Afterward, cells were washed twice in buffer 2 (0.05% [wt/vol] saponin, 1× EDTA-free protease inhibitors, 0.5 mM spermidine in PBS) and resuspended in 100 μl calcium buffer (2 mM CaCl₂ in buffer 2) on ice for 30 min to activate MNase. Afterward, 100 μl 2× stop buffer (20 mM EDTA, 4 mM EGTA in saponin buffer) was added and incubated in a 37°C incubator for 10–20 min to release cleaved chromatin fragments. Finally, the supernatant containing chromatin fragments was collected by centrifugation and DNA was extracted using a QIAGEN MinElute kit (catalog# 28004) according to the manufacturer’s instructions.

CUT&RUN libraries were prepared using the NEBNext Ultra II DNA Library Prep Kit for Illumina (E7645) according to the manufacturer’s instructions with minor modifications in the End Prep. step. (reduce temperature to 50°C and increase the reaction time to 1 h) to better save small DNA fragments for library construction. After quality control using a Bioanalyzer, libraries were pooled and sequenced on an Illumina NovaSeq sequencer (PE 150 bp).

Retroviral production and T cell transduction

500K HEK293T cells per well in a 6-well plate were transfected with 0.8 μg of packaging vector pCL-Eco (catalog# 12371; Addgene) and 1.2 μg of expression vector (MigR2 or MigR2-BHLHE40) or targeting vector (pSIRG-sgNT or pSIRG-sgBHLHE40) or MCC vector expressing Cas9. using Lipofectamine 3000 Transfection Kit (catalog# L3000001; Invitrogen) according to the manufacturer’s instructions. The viral supernatant was collected 48 and 72 h after transfection.

Naïve CD4⁺ T cells were cultured in the RORγt⁺Foxp3⁺ iTreg cell differentiation condition. Retrovirus transduction was performed 24 h later by incubating cells with viral supernatant as indicated in the presence of 4 μg/ml polybrene (catalog# TR-1003-G; Millipore) and centrifuged at 2,500 RPM for 90 min at 32°C. The effect of BHLHE40 overexpression or BHLHE40 knockout on the expression of c-Maf, RORγt, IL-10, and CTLA-4 in RORγt⁺Foxp3⁺ iTreg cells was analyzed by FACS.

GSEA

GSEA software (v.3.0) was used to perform the analyses using TPM values to compare the enrichment of the colonic Treg signature and RORγt signature (Sefik et al., 2015) in colonic Treg cells from WT and REV-ERB cKO mice.

RNA-seq data processing and analysis

Sequenced reads were quality-tested using FASTQC v.0.11.8 (Andrews, 2010) and were trimmed by Trim Galore v.0.4.4 (Felix, 2023). The trimmed reads were aligned to the mm10 mouse genome using STAR v.2.5.3a (Dobin et al., 2013) with default parameters. Raw or TPM (Transcript Per Million) gene expression levels were quantified using HOMER v.4.9.1 (Heinz et al., 2010) by calculating the uniquely aligned reads to the exons of RefSeq genes. Differential gene expression analysis was performed on the raw gene counts with the R package, DESeq2 (Love et al., 2014), using replicates to compute within-group dispersion. Differentially expressed genes were defined as having a FDR <0.05 and absolute fold change (FC) > 1.5 when comparing two experimental conditions. Principal component analysis (PCA) was carried out on normalized gene counts using the R prcomp function.

CUT&RUN data processing and analysis

Raw read quality was checked using FASTQC v.0.11.8 (Andrews, 2010), and the adaptors were trimmed with Trim Galore v.0.4.4 (Felix, 2023). Reads were mapped to the mm10 mouse genome using STAR v.2.5.3a (Dobin et al., 2013). Mapping was carried out using default parameters (up to 10 mismatches per read, and up to 9 multi-mapping locations per read). Peaks were identified using HOMER “findpeaks,” using input IgG condition as background reference and default parameters (fourfold enrichment over input control, fourfold enrichment over local tag count, FDR < 0.001, and style factor). HOMER “mergePeaks -d given” was applied to find all combinations of overlaps between peaks in each condition, and HOMER “annotatePeaks” was used to determine the distance to the nearest TSS. To calculate the distribution of normalized read coverage around the center of peaks or TSSs, HOMER annotatePeaks was used with a window size of ±3 kb and a bin size of 25 bp. The normalized read coverage was visualized in a histogram or heatmap. Normalized read densities were visualized using the UCSC genome browser (Kent et al., 2002). All the statistical analysis and graphics were carried out in R v.4.2.0 (https://www.R-project.org/).

Statistics

All statistical analyses were conducted with GraphPad Prism 10. In all studies, values are expressed as the mean ± standard error of the mean (SEM) or mean ± standard deviation (SD), and all n numbers represent biological repeats. Statistical analyses were performed with Student’s t test for Fig. 1 H; Fig. 2, B and C–E; Fig. 3, E and F; Fig. 4, B–D; Fig. 5 B; Fig. S1, C and E–J; Fig. S2, B, D, E, and H; and Fig. S5 B; or two-way ANOVA for Fig. 1, A, B, and G; Fig. 2 A; Fig. 3, C and D; Fig. 4 A; Fig. 5, D and E; Fig. 6 D; Fig. 7, B and C; Fig. S1 D; Fig. S2, A, C, F, and G; Fig. S3 C; Fig. S4, B and D; and Fig. S5, A and C. The results are indicated in the corresponding figures and legends (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; NS, not significant).

Supplementary Material

Online supplemental material

Fig. S1 shows additional analysis of T cell development and distribution in REV-ERB cKO and WT mice at steady state. Fig. S2 shows that REV-ERB cKO mice are more susceptible to oxazolone-induced colitis. Fig. S3 shows that REV-ERB deficiency leads to impaired fitness of colonic Treg cells. Fig. S4 shows the in vitro induction strategy for RORγt⁺Foxp⁺ T cells, a system we used to demonstrate that RORγt may directly regulate REV-ERBα expression. Fig. S5 shows that REV-ERB agonist SR9009 increases RORγt⁺Foxp3⁺ T cell differentiation, and targeting Bhlhe40 could restore REV-ERBa–deficient RORγt⁺ Treg cell function in vitro.

Data availability

The scRNA-seq dataset of colonic Treg cells (shown in Fig. 1) is from GSE240982. The ChIP-seq dataset showing RORγt binding peaks at the Nr1d1 locus in Th17 cells (shown in Fig. S4) is from GSE40918. The RNA-seq and CUT&RUN sequence data generated in this study (shown in Figs. 3, 4, 5, 6, and 7) have been deposited in the Gene Expression Omnibus under the accession no. GSE225203.