Short Report

Immunology and Inflammation

c-Maf restrains T-bet-driven programming of CCR6-negative group 3 innate lymphoid cells

Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Germany
Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Germany
Chronic Immune Reactions, Deutsches Rheuma-Forschungszentrum, Germany
Therapeutic Gene Regulation, Deutsches Rheuma-Forschungszentrum, Germany
Chronic Inflammation, Deutsches Rheuma-Forschungszentrum, Germany
Belozersky Institute of Physico-Chemical Biology and Biological Faculty, M.V. Lomonosov Moscow State University, Russian Federation

Feb 10, 2020

Open access
Copyright information

Abstract
Introduction
Results and discussion
Materials and methods
Data availability
References
Article and author information
Metrics

Abstract

RORγt⁺ group 3 innate lymphoid cells (ILC3s) maintain intestinal homeostasis through secretion of type 3 cytokines such as interleukin (IL)−17 and IL-22. However, CCR6^- ILC3s additionally co-express T-bet allowing for the acquisition of type 1 effector functions. While T-bet controls the type 1 programming of ILC3s, the molecular mechanisms governing T-bet are undefined. Here, we identify c-Maf as a crucial negative regulator of murine T-bet⁺ CCR6^- ILC3s. Phenotypic and transcriptomic profiling of c-Maf-deficient CCR6^- ILC3s revealed a hyper type 1 differentiation status, characterized by overexpression of ILC1/NK cell-related genes and downregulation of type 3 signature genes. On the molecular level, c-Maf directly restrained T-bet expression. Conversely, c-Maf expression was dependent on T-bet and regulated by IL-1β, IL-18 and Notch signals. Thus, we define c-Maf as a crucial cell-intrinsic brake in the type 1 effector acquisition which forms a negative feedback loop with T-bet to preserve the identity of CCR6^- ILC3s.

Introduction

Tissue-resident, RORγt-dependent group 3 innate lymphoid cells (ILC3s) protect mucosal surfaces against infections and maintain the integrity of the epithelial barrier by secretion of cytokines such as IL-17 and IL-22 (Sonnenberg et al., 2012; Gronke et al., 2019; Hernández et al., 2015; Zheng et al., 2008; Gladiator et al., 2013; Satoh-Takayama et al., 2008; Zenewicz et al., 2008). In mice, ILC3s consist of two major subsets, CCR6⁺ ILC3s that include lymphoid tissue inducer (LTi) cells and CCR6^- ILC3s that can differentiate to cells expressing type 1 effector molecules, such as IFN-γ and the natural cytotoxic receptor NKp46 (Sawa et al., 2010; Klose et al., 2013). Functionally, fetal LTi and adult LTi-like CCR6⁺ ILC3s are essential for lymphoid organ development (Sun et al., 2000; Eberl et al., 2004), while NKp46⁺ CCR6^- ILC3s are implicated in type 1 inflammatory immune responses and thus may also have pathogenic functions during intestinal inflammation (Klose et al., 2013; Powell et al., 2012; Buonocore et al., 2010; Bernink et al., 2013; Vonarbourg et al., 2010; Rankin et al., 2016; Song et al., 2015).

The transcription factor (TF) RORγt is strictly required for the development of all ILC3s, as mice deficient for RORγt lack all ILC3 subsets (Eberl et al., 2004; Sanos et al., 2009). RORγt also controls the functionality of ILC3s by regulating the production of effector cytokines, such as IL-17 and IL-22 (Ivanov et al., 2006; Rutz et al., 2013). Interestingly, CCR6^- ILC3 co-express RORγt and the master regulator of type 1 immunity, T-bet. T-bet is key to the differentiation of NKp46⁺ CCR6^- ILC3s, as those cells fail to develop in mice lacking the gene encoding T-bet, Tbx21 (Klose et al., 2013; Rankin et al., 2013). Importantly, T-bet not only contributes to NKp46⁺ CCR6^- ILC3 development, but an increasing T-bet gradient enables functional plasticity of NKp46⁺ CCR6^- ILC3s by instructing a type 1 effector program in ILC3s (Klose et al., 2013; Sciumé et al., 2012; Klose et al., 2014; Cella et al., 2019). Tunable T-bet expression in NKp46⁺ CCR6^- ILC3s serves as a dynamic molecular switch from a type 3 to a type 1 phenotype (Klose et al., 2013). Once T-bet expression reaches a sufficient level, it can also act as a repressor of RORγt, resulting eventually in a full conversion of ILC3s to ILC1-like cells (referred to as ILC3-to-1 plasticity) (Vonarbourg et al., 2010; Cella et al., 2019; Bernink et al., 2015). Thus, the balance between RORγt versus T-bet expression dictates the fate and function of CCR6^- ILC3s (Fang and Zhu, 2017).

Importantly, the molecular mechanisms controlling the dynamic and quantitative co-expression of RORγt and T-bet in CCR6^- ILC3s are largely undefined. Several extrinsic signals were shown to promote or restrain T-bet-dependent plasticity, most prominently cues from the microbiota, IL-23, IL-7 and Notch signaling (Klose et al., 2013; Sanos et al., 2009; Rankin et al., 2013; Viant et al., 2016; Chea et al., 2016). Moreover, exposure to pro-inflammatory cytokines, such as IL-12, IL-15 and IL-18, was reported to further support transdifferentiation to an ILC1-like fate (Vonarbourg et al., 2010; Bernink et al., 2015; Satoh-Takayama et al., 2010). However, despite this, the intrinsic molecular mediators governing ILC3 plasticity have not been discovered yet.

In the past, our group and others could identify the AP-1 TF c-Maf as a central regulator of RORγt⁺ CD4⁺ T cells, including RORγt⁺ Foxp3⁺ Treg cells (Neumann et al., 2019; Xu et al., 2018; Wheaton et al., 2017), RORγt⁺ Th17 cells (Ciofani et al., 2012; Aschenbrenner et al., 2018; Tanaka et al., 2014) and RORγt⁺ γδ T cells (Zuberbuehler et al., 2019), both in mouse and human. Specifically, c-Maf was shown to directly bind and regulate key genes of RORγt⁺ T cells, including IL-22 and RORγt itself (Tanaka et al., 2014; Zuberbuehler et al., 2019; Rutz et al., 2011). Recently, a broad transcriptional network analysis also identified c-Maf as an important regulator of the ILC3-ILC1 balance, although the precise underlying molecular mechanisms have remained unclear (Pokrovskii et al., 2019).

Here, we demonstrate that c-Maf was essential for CCR6^- ILC3s to establish a physiological equilibrium between type 1 and type 3 effector states. c-Maf directly restrained T-bet expression, thereby preventing CCR6^- ILC3s from acquiring excessive type 1 effector functions. c-Maf expression itself was dependent on T-bet and tightly correlated with its expression level. Upstream, we identified IL-1ß- and IL-18-mediated NF-κB, as well as Notch signals, as potent extrinsic enhancers of c-Maf expression in CCR6^- ILC3s. Thus, our data define c-Maf as an integral regulator within the type 3-to-1 conversion program that acts as a cell-intrinsic gatekeeper of T-bet expression to maintain the function and lineage-stability of CCR6^- ILC3s.

Results and discussion

c-Maf specifically preserves the type 3 identity of CCR6^- ILC3s

Given the pivotal role of c-Maf in CD4⁺ T cells, we aimed to define its function in ILCs, which share a similar transcriptional program with T cells (Vivier et al., 2018). We first investigated the expression pattern of c-Maf in different ILC subsets of the small intestinal lamina propria (siLP) by staining for c-Maf. This analysis showed that ILC3s expressed higher levels of c-Maf when compared to ILC1s or ILC2s (Figure 1A, gating strategy see Figure 1—figure supplement 1). Among the ILC3 subsets, c-Maf was particularly highly expressed by NKp46⁺ CCR6^- ILC3s at levels comparable to RORγt⁺ CD4⁺ T cells (Figure 1B). Collectively, these data suggested a potential function of c-Maf in these cells.

Figure 1 with 1 supplement see all

Download asset Open asset

c-Maf is required to maintain the type 3 phenotype of CCR6^- ILC3s.

(A) Protein expression of c-Maf by ILC1s, ILC2s and ILC3s isolated from the siLP of naïve C57BL/6N wild-type mice as measured by flow cytometry (pregated on viable Lin^- CD127⁺ cells). Representative histograms show c-Maf expression. (B) Representative histograms showing expression of c-Maf by different siLP ILC3 subsets and RORγt⁺ T cells as comparison. (C) Frequencies and total numbers of ILC1s, ILC2s and ILC3s from siLP of *Rorc*^Cre *Maf*^fl/fl and control mice as measured by flow cytometry (pregated on viable Lin^- CD127⁺ cells). Representative flow cytometric profiles of RORγt vs. T-bet and GATA3 expression (left); right, quantification (n = 5–6, mean ± SEM, *p<0.05, **p<0.01). (D) Total numbers of different siLP ILC3 subsets of *Rorc*^Cre *Maf*^fl/fl and control mice as measured by flow cytometry. Representative flow cytometric profiles of NKp46 vs. CCR6 expression (left); right, quantification (n = 5–6, mean ± SEM, *p<0.05). (E) Expression of RORγt by different siLP ILC3 subsets from *Rorc*^Cre *Maf*^fl/fl and control mice. Representative histograms (upper panel) show RORγt expression. Graphs below show quantification of RORγt gMFI (n = 5–6, mean ± SEM, **p<0.01, ***p<0.001). (**F, G**) Frequencies of IL-17A and IL-22 positive cells among different siLP ILC3 subsets of *Rorc*^CreMaf^fl/fl and control mice after ex vivo restimulation with PMA/ionomycin and IL-23. Representative flow cytometric profiles of IL-17A or IL-22 vs. CCR6 are shown on the left; quantification on the right (n = 4, mean ± SEM, *p<0.05, **p<0.01). Data are representative of three independent experiments. Statistical differences were tested using an unpaired Students’ t-test (two-tailed).

In order to directly study the role of c-Maf in ILC3s, we crossed mice carrying floxed Maf alleles (Maf^fl/fl) to mice expressing Cre recombinase driven by the regulatory elements of the Rorc(γt) gene locus (Rorc-Cre^Tg), thereby generating mice with a specific deletion of c-Maf in RORγt⁺ ILC3s and T cells (Rorc^Cre Maf^fl/fl). Phenotypic analysis of siLP ILCs revealed decreased frequencies of ILC3s in the absence of c-Maf, whereas frequencies and total numbers of ILC1s and ILC2s were increased in Rorc^Cre Maf^fl/fl mice when compared to littermate controls (Figure 1C). Among ILC3s, we detected a selective loss of NKp46^- CCR6^- ILC3s in Rorc^Cre Maf^fl/fl mice, while total numbers of NKp46⁺ CCR6^- ILC3s and CCR6⁺ ILC3s were not significantly changed (Figure 1D).

Thus, c-Maf was selectively required for the maintenance of intestinal NKp46^- CCR6^- ILC3s. Its absence in ILC3s resulted in significant changes in the proportions of individual ILC subsets in the gut. Importantly, flow cytometric intracellular staining also revealed a significant downregulation of RORγt protein levels in c-Maf-deficient NKp46⁺ and NKp46^- CCR6^- ILC3s, which was not observed in CCR6⁺ ILC3s (Figure 1E).

Next, we tested the functionality of c-Maf-deficient ILC3s by assessing their capacity to produce the type 3 signature cytokines IL-17A and IL-22. In line with the decrease in RORγt expression, NKp46⁺ and NKp46^- CCR6^- ILC3s from Rorc^Cre Maf^fl/fl mice exhibited significantly reduced frequencies of IL-17A and IL-22 producers after ex vivo restimulation as compared to control cells (Figure 1F and G). Again, no differences were detected in CCR6⁺ ILC3s, corroborating the selective role of c-Maf for the homeostasis and function of CCR6^- ILC3s. Notably, c-Maf did not act as a repressor of IL-22 production in ILC3s, as it was shown for Th17 cells (Rutz et al., 2011), indicating essential differences in c-Maf function between ILCs and T cells.

In summary, these findings demonstrated a crucial requirement of c-Maf in maintaining the type 3 identity of CCR6^- ILC3s, including their expression of RORγt, IL-17A and IL-22.

c-Maf suppresses the acquisition of type 1 properties by CCR6^- ILC3s

RORγt⁺ CCR6^- ILC3s have the capacity to acquire type 1 effector characteristics, such as IFN-γ and NKp46 expression (Vonarbourg et al., 2010). This differentiation is facilitated by graded co-expression of T-bet, together with RORγt, enabling functional flexibility during inflammatory immune responses (Klose et al., 2013; Powell et al., 2012; Bernink et al., 2013; Rankin et al., 2013; Sciumé et al., 2012). However, the molecular mechanisms governing the T-bet-dependent type 1 programming of CCR6^- ILC3s are undefined.

Interestingly, c-Maf expression strongly correlated with T-bet expression in both NKp46⁺ and NKp46^- CCR6^- ILC3s (Figure 2A). More importantly, c-Maf deficiency resulted in strong upregulation of T-bet and NKp46 expression, both on the population (frequencies) and at single cell level (gMFI), in CCR6^- ILC3s (Figure 2B). These data, together with the selective loss of NKp46^- CCR6^- ILC3s, which are considered to contain precursors of NKp46⁺ CCR6^- ILC3s, in Rorc^CreMaf^fl/fl mice (Figure 1D), suggested an amplified type 1 conversion of CCR6^- ILC3s in the absence of c-Maf.

Figure 2 with 2 supplements see all

Download asset Open asset

CCR6^- ILC3s acquire a hyper type 1 phenotype in the absence of c-Maf.

(A) Representative flow cytometric profiles of T-bet vs. c-Maf expression by siLP NKp46⁺ and NKp46^- CCR6^- ILC3s. (B) T-bet and NKp46 expression (frequency and gMFI) by siLP CCR6^- ILC3s of *Rorc*^CreMaf^fl/fl and control mice. Representative flow cytometric profiles are shown on the left; quantification on the right (n = 4, mean ± SEM, **p<0.01, ***p<0.001). (C) Frequency of IFN-γ positive cells among siLP CCR6^- ILC3s of *Rorc*^CreMaf^fl/fl and control mice after ex vivo restimulation with PMA/ionomycin and IL-23. Representative flow cytometric profiles are shown on the left; quantification on the right (n = 4, mean ± SEM, ***p<0.001). (D) Analysis of mixed bone marrow chimeras of CD45.1⁺ wild-type and CD45.2⁺ *Rorc*^CreMaf^fl/f (Maf KO) bone marrow cells. Pregated on viable Lin^- CD127⁺ CD90.2⁺ RORγt⁺ cells. Recipient mice were CD90.1 positive. Representative flow cytometric profiles of T-bet vs. CCR6 expression are shown left; quantification on the right (n = 7, mean ± SEM, **p<0.01). All statistical differences were tested using an unpaired Students’ t-test (two-tailed). (E) NKp46⁺ CCR6^- ILC3s were sorted from siLP of *Rorc*^CreMaf^fl/fl and control mice and subjected to RNA sequencing. Vulcano plot showing comparison of gene expression between c-Maf-deficient and control NKp46⁺ CCR6^- ILC3s. Data represent the combined analysis of three biologically independent samples. Genes considered significant (FC > 1.5, FDR < 0.05) fall into the grey background, while selected genes are highlighted in red. (F) Gene set enrichment plots showing enrichment of ILC1 and NK cell signature genes in c-Maf-deficient vs. control NKp46⁺ CCR6^- ILC3s (FDR < 0.01). Normalized enrichment score (NES).

In accordance with the increase in T-bet expression, we also detected increased frequencies of IFN-γ producing cells within c-Maf-deficient CCR6^- ILC3s as compared to c-Maf-competent control cells (Figure 2C). Of note, Ki67 staining of c-Maf-deficient CCR6^- ILC3s was not altered, ruling out that proliferative differences facilitated the skewing towards a type 1 phenotype in the absence of c-Maf (Figure 2—figure supplement 1A).

In addition to the lack of c-Maf expression in ILC3s, Rorc^Cre Maf^fl/fl mice also harbour a c-Maf-deficient T cell compartment, due to the expression of RORγt during T cell development (Sun et al., 2000). Moreover, conditional deletion of c-Maf in T cells (Cd4^Cre Maf^fl/fl or Foxp3^Cre Maf^fl/fl) was reported to cause disturbances in intestinal homeostasis (Neumann et al., 2019; Imbratta et al., 2019), raising the possibility that changes in the gut microenvironment contributed to the ‘hyper type 1’ phenotype of CCR6^- ILC3s in Rorc^Cre Maf^fl/fl mice. To interrogate this scenario, we analysed intestinal CCR6^- ILC3s from Cd4^Cre Maf^fl/fl mice. The expression of T-bet by CCR6^- ILC3s was not altered when c-Maf was selectively deleted in T cells, excluding that T cell-dependent alterations affected the type 1 conversion of ILC3s in Rorc^Cre Maf^fl/fl mice (Figure 2—figure supplement 1B).

Interestingly, a small subset of CCR6^- ILC3s also co-expresses CD4 (ca. 2% of CCR6^- ILC3s, Figure 2—figure supplement 1C). Indeed, c-Maf staining of CD4⁺ CCR6^- ILC3s from Cd4^Cre Maf^fl/fl mice confirmed c-Maf deletion in a substantial fraction of cells (ca. 50%), allowing us to compare c-Maf-competent and c-Maf-deficient CCR6^- ILC3s within the same mouse (Figure 2—figure supplement 1C). We detected higher expression of T-bet and NKp46 in cells lacking c-Maf as compared to their c-Maf-expressing counterparts, indicating that the type 1 deviation of c-Maf-deficient CCR6^- ILC3s was a cell-intrinsic phenomenon (Figure 2—figure supplement 1C). Importantly, mixed bone marrow chimeras of wild-type and c-Maf-deficient cells confirmed the cell-intrinsic function of c-Maf in restraining T-bet expression (Figure 2D).

Together, these data identified c-Maf as a potent repressor of the T-bet-dependent type 1 conversion of CCR6^- ILC3s. Notably, c-Maf has also been shown to attenuate the differentiation of Th1 cells by a yet-to-be-defined mechanism (Ho et al., 1998), suggesting a broader role of c-Maf in restraining type 1 immune effector cells.

c-Maf globally restrains type 1 and NK cell specific gene programs in CCR6^- ILC3s

To better define c-Maf-dependent gene regulation, we performed RNA sequencing (RNA-seq) of siLP ILC3s of Rorc^CreMaf^fl/fl and control mice. We isolated NKp46⁺ CCR6^- ILC3s and as a comparison CCR6⁺ ILC3s for our analysis (sorting strategy see Figure 2—figure supplement 2A). Principal component analysis showed that c-Maf-deficient ILC3s clustered separately from their wild-type counterparts, demonstrating a unique role of c-Maf in shaping the transcriptome of ILC3s (Figure 2—figure supplement 2B).

We identified 941 genes as differentially expressed (FC > 1.5, p-value < 0.05) between c-Maf-deficient and control NKp46⁺ CCR6^- ILC3s (Figure 2E, Supplementary file 1). Interestingly, only ca. 30% of these genes (294 genes) were also found to be differentially expressed in CCR6⁺ ILC3s (Figure 2—figure supplement 2C, Supplementary file 2), indicating a distinct function of c-Maf in NKp46⁺ CCR6^- ILC3s. Consistent with our flow cytometry data, we detected a striking enrichment of type 1 and NK cell-related genes (e.g. Tbx21, Ifng, Nos2, Klri2, Klrc1, Klrc2, Gzmb, Gzmd, Fasl) in c-Maf-deficient NKp46⁺ CCR6^- ILC3s (Figure 2E). In line with this, Bcl6 and Irf8, both recently described as crucial transcription factors promoting ILC1/NK cell gene programs (Pokrovskii et al., 2019; Adams et al., 2018), were also significantly upregulated in the absence of c-Maf in NKp46⁺ CCR6^- ILC3s (Figure 2E, Figure 2—figure supplement 2D and E). Conversely, several type 3 effector genes (e.g. Il17a, Il17f, Il17re, Ccl20) were downregulated in NKp46⁺ CCR6^- ILC3s upon c-Maf-deficiency (Figure 2E).

To better understand the global c-Maf-dependent changes in gene expression programs, we performed gene set enrichment analysis (GSEA). Given the strong upregulation of type 1 and NK cell features in c-Maf-deficient CCR6^- ILC3s, we made use of published RNA-seq data comparing ILC1s and NK cells with CCR6^- ILC3s (Pokrovskii et al., 2019). In detail, we created sets of genes that were most highly overexpressed in ILC1s or NK cells, thereby defining ILC1 and NK cell gene signatures that distinctly separated those lineages from CCR6^- ILC3s. Importantly, when applied to GSEA, both gene signatures were significantly enriched in c-Maf-deficient NKp46⁺ CCR6^- ILC3s (Figure 2F), indicating a global shift in gene expression towards an ILC1/NK cell phenotype.

In summary, these data demonstrated that c-Maf was essential to globally balance type 1 and type 3 gene programs within CCR6^- ILC3s. This function of c-Maf was specific to NKp46⁺ CCR6^- ILC3s as opposed to CCR6⁺ ILC3s, most likely due to the particularly high c-Maf expression and the selective accessible of type 1 gene loci in these cells (Pokrovskii et al., 2019; Shih et al., 2016). In the absence of c-Maf, NKp46⁺ CCR6^- ILC3s downregulated type 3 effector genes, while overexpressing numerous genes encoding for type 1 and cytotoxic effector molecules. The latter finding is particularly interesting, since NKp46⁺ CCR6^- ILC3s are largely considered to be non-toxic cells (Melo-Gonzalez and Hepworth, 2017). Nevertheless, NKp46⁺ CCR6^- ILC3s share a considerable transcriptional overlap with ILC1s, which also exhibit a degree of cytotoxic capacity (Robinette et al., 2015; Cortez and Colonna, 2016). Thus, c-Maf-deficiency may result in marked cytotoxicity of NKp46⁺ CCR6^- ILC3s. More work is needed to precisely define the role of c-Maf for ILC3 functionality during homeostasis and in the context of intestinal inflammation.

c-Maf restrains T-bet expression by directly repressing the Tbx21 promoter

The strong upregulation of T-bet expression in c-Maf-deficient CCR6^- ILC3s raised the possibility that c-Maf acted as a direct repressor of T-bet, thereby restraining the type 1 differentiation program. Indeed, in silico analysis identified Maf response elements (MARE) within the Tbx21 promoter, as well as in a conserved distant Tbx21 enhancer (Kataoka et al., 1994; Yang et al., 2007), both regions accessible in NKp46⁺ CCR6^- ILC3s as evidenced by ATAC-sequencing (Shih et al., 2016; Figure 3A and B).

Figure 3 with 3 supplements see all

Download asset Open asset

c-Maf directly represses T-bet as part of a negative feedback loop.

(A) Representative ATAC sequencing tracks across the *Tbx21* locus of NKp46⁺ ILC3s (Shih et al., 2016). The *Tbx21* promoter and enhancer regions are highlighted in red (Yang et al., 2007). ATAC sequencing tracks were visualized using the WashU browser from the Cistrome project (Mei et al., 2017). (B) Positioning of putative *Maf* response elements (MARE) within the *Tbx21* promoter and enhancer (highlighted in blue). Depicted sequences represent selected regions of the *Tbx21* promoter and enhancer. Mutated sequences are shown in lower lines in red. (C) Schematic representation of plasmids containing the *Tbx21* promoter/enhancer linked to the firefly luciferase reporter gene (Luc). Blue dots indicate MARE sites within the *Tbx21* promoter and enhancer region. (D) Relative luciferase activity (RLU) of different reporter constructs driven by the *Tbx21* promoter alone or in combination with an enhancer sequence compared to a promoterless (w/o) control vector (pGL3 basic) (n = 3, mean ± SEM, *p<0.05, ***p<0.001). (E) Analysis of c-Maf-dependent suppression of luciferase activity. The MARE sites within the *Tbx21* promoter or the *Tbx21* promoter and enhancer were mutated in the *Tbx21* enhancer/promoter construct. Comparison of RLU between unmutated and mutated *Tbx21* enhancer/promoter constructs upon c-Maf overexpression (n = 3, mean ± SEM, ***p<0.001). All reporter assay data are pooled from three independent experiments. (F) Expression of c-Maf by siLP CCR6^- ILC3s from T-bet-deficient (T-bet^-/-) and -sufficient (T-bet^+/+) mice. Representative flow cytometric profiles are shown on the left; graph on the right shows quantification of c-Maf gMFI (n = 4, mean ± SEM, ***p<0.001). (**G–H**) Sort-purified siLP NKp46⁺ CCR6^- ILC3s from *Rorc*^Cre *R26*^EYFP mice were cultured in vitro for 36 hr in the presence of IL-7/SCF (w/o) or IL-7/SCF plus indicated cytokines. Subsequently, *Maf* expression was measured by qPCR (n = 4, mean ± SEM, *p<0.05, **p<0.01, ***p<0.001). In one condition the NF-κb inhibitor BMS-345541 (BMS) was added at 1 μM to the culture. Data are pooled from two independent experiments each with two replicate wells. (I) Sort-purified siLP NKp46⁺ CCR6^- ILC3s from *Rorc*^Cre *R26*^EYFP mice were cultured in the presence of IL-7/SCF on OP9 or OP9-DLL1 stromal cells as indicated. After 12 days, cells were analysed by flow cytometry for the cell surface expression of NKp46. Representative contour plots are shown on the left (pregated on CCR6^- ILC3s); graph on the right shows quantification of the frequency of NKp46⁺ cell among CCR6^- ILC3s (n = 10, mean ± SEM, ***p<0.001). Data are pooled from two independent experiments with 4 to 6 replicate wells. (J) Expression of c-Maf by NKp46⁺ CCR6^- ILC3s cultured on OP9 or OP9-DLL1 cells. Representative histogram is shown left; graph on the right shows quantification of c-Maf gMFI (n = 10, mean ± SEM, ***p<0.001). All statistical differences were tested using an unpaired Students’ t-test (two-tailed).

In order to study the transcriptional activity of c-Maf at these sites, we cloned the Tbx21 promoter alone or in conjunction with the Tbx21 enhancer upstream of a luciferase reporter (Figure 3C). Indeed, the Tbx21 promoter showed strong transcriptional activity when compared to a promoterless control vector (Figure 3D). The Tbx21 enhancer further increased this activity when cloned upstream of the Tbx21 promoter (Figure 3D). Importantly, upon exogenous overexpression of c-Maf, we detected a strong de-repression of the reporter signal when we mutated the MARE sites within the Tbx21 promoter (Figure 3B and E). Notably, mutating the Tbx21 enhancer did not result in further increase of reporter activity, indicating that c-Maf facilitated its suppressive function mainly by acting on the Tbx21 promoter (Figure 3B and E).

Thus, these data identified the Tbx21 promoter as a c-Maf-responsive region through which c-Maf directly controls T-bet expression.

c-Maf expression in CCR6^- ILC3s is dependent on T-bet

Despite the fact that c-Maf acted as a direct repressor of T-bet, we observed strong correlation of c-Maf expression with T-bet expression in CCR6^- ILC3s (Figure 2A). This finding let us hypothesize that c-Maf expression was co-regulated with T-bet expression as part of the type one conversion program. To explore this hypothesis, we studied c-Maf expression in CCR6^- ILC3s from T-bet-deficient mice. Indeed, in the absence of T-bet, c-Maf expression was strongly reduced when compared to T-bet-sufficient cells (Figure 3F), indicating that T-bet positively regulated its own repressor to establish an equilibrated state within the ILC3-to-ILC1 continuum.

In silico analysis of the Maf locus using ATAC-Seq (Shih et al., 2016) and T-bet ChIP-Seq (Gökmen et al., 2013) data further identified a striking overlap of T-bet binding peaks with open chromatin regions in NKp46⁺ CCR6^- ILC3s within two conserved non-coding sequences (CNS-0.5 and CNS-1) upstream of Maf, suggesting that T-bet directly controlled c-Maf expression in CCR6^- ILC3s (Figure 3—figure supplement 1).

IL-1β- and IL-18-mediated NF-κb and Notch signalling promote c-Maf expression in CCR6^- ILC3s

Next, we aimed to identify signals that control c-Maf expression in CCR6^- ILC3s. Several cytokines were shown to regulate the differentiation and function of T-bet⁺ CCR6^- ILC3s, including IL-1ß, IL-12, IL-15, IL-18 and IL-23 (Klose et al., 2013; Vonarbourg et al., 2010; Bernink et al., 2015; Satoh-Takayama et al., 2010). Thus, we short-term stimulated sort-purified NKp46⁺ CCR6^- ILC3s with these cytokines in vitro and subsequently assessed their Maf expression by qPCR. Among all tested conditions, the pro-inflammatory IL-1 family cytokines IL-1ß and IL-18 stood out as potent inducers of Maf expression (Figure 3G). IL-1ß/IL-18 stimulation also induced Ifng and Il22 expression (Figure 3—figure supplement 2A). In fact, c-Maf-deficient NKp46⁺ CCR6^- ILC3s expressed more Ifng in response to IL-1ß/IL-18 as compared to controls (Figure 3—figure supplement 2A), functionally connecting c-Maf expression downstream of these stimuli with the control of the type 1 effector response.

Since both IL-1ß and IL-18 signal via NF-κb and a conserved NF-κb binding site is present within CNS-0.5 upstream of Maf (Figure 3—figure supplement 1C and D), we hypothesized that NF-κb is involved in regulation of c-Maf expression downstream of IL-1ß/IL-18. Indeed, pharmacological inhibition of NF-κb signalling completely abrogated the cytokine-mediated induction of c-Maf expression (Figure 3H). In addition to IL-1ß/IL-18, we found that IL-12 suppressed Maf expression (Figure 3G). However, IL-12 did not interfere with the IL-1ß/IL-18-mediated c-Maf induction nor was c-Maf expression altered in NKp46⁺ CCR6^- ILC3s from Il12a^-/- mice as compared to controls (Figure 3—figure supplement 2B and C), questioning a dominant role of IL-12 in c-Maf regulation.

Of note, the cytokine-mediated induction of Maf expression was accompanied by a concomitant decrease in Tbx21 expression (Figure 3—figure supplement 2D). Yet, this suppression was independent of NF-κb and c-Maf, since Tbx21 expression was similarly suppressed by IL-1ß/IL-18 upon NF-κb inhibiton and in c-Maf-deficient NKp46⁺ CCR6^- ILC3s (Figure 3—figure supplement 2E and F). Taken together, these data demonstrate that, besides their reciprocal regulation, c-Maf and T-bet expression level are further critically controlled by cytokine signals.

In addition to cytokines, we also tested Notch signals as a potential extrinsic cue controlling both type 1 conversion of and c-Maf expression by ILC3s. Indeed, Notch was shown to be necessary for the induction and maintenance of T-bet and NKp46 expression in CCR6^- ILC3s (Rankin et al., 2013; Viant et al., 2016). Similarly, we could recently identify Notch signalling as a potent inducer of c-Maf expression in T cells (Neumann et al., 2019; Neumann et al., 2014). To test the role of Notch we cultured purified NKp46⁺ CCR6^- ILC3s on OP9 or OP9-DLL1 stromal cells, the latter ectopically express the Notch ligand Delta-like 1 (DLL1). As reported earlier, Notch signals were essential to drive type 1 conversion of NKp46⁺ CCR6^- ILC3s, as evidenced by reduction and loss of NKp46 expression in the absence of Notch (Figure 3I). Importantly, c-Maf showed a similar expression pattern, being significantly reduced in NKp46⁺ CCR6^- ILC3s cultured on OP9 cells as compared to cells cultured on OP9-DLL1 cells (Figure 3J). This downregulation of NKp46 and c-Maf expression in the absence of continuous Notch signalling was independent of potential survival promoting effects of Notch (Figure 3—figure supplement 3), establishing a molecular link between Notch signalling and type 1 conversion of NKp46⁺ CCR6^- ILC3s.

Concluding remarks

Collectively, our study adds c-Maf as a novel key factor to the complex transcriptional network that governs the differentiation and function of ILC3s. In line with the emerging concept that co-expression and cross-regulation of multiple master regulators determines the fate and function of ILCs (Fang and Zhu, 2017), we have uncovered an essential negative feedback loop between c-Maf and T-bet, which restrains the type 1 conversion of ILC3s. Given the antagonism between T-bet and RORγt, the c-Maf-dependent suppression of T-bet also indirectly stabilizes RORγt expression, thus preserving the type 3 identity of ILC3s. In addition, c-Maf may also directly contribute to RORγt expression in ILC3s, as it has been reported for T cells (Tanaka et al., 2014; Zuberbuehler et al., 2019).

Materials and methods

Animals

To generate conditional c-Maf-deficient mice, Rorc^Cre mice or Cd4^Cre mice were crossed to Maf^flox mice (provided by C. Birchmaier, MDC, Berlin, Germany) (Wende et al., 2012). Cd4^Cre (Lee et al., 2001), Rorc^Cre (Eberl and Littman, 2004), R26^EYFP (Srinivas et al., 2001) and T-bet-deficient mice (Szabo et al., 2002) were described before. Il12a^-/- mice were kindly provided by U. Schleicher, Erlangen. All mice were on a C57BL/6 background and bred and maintained under specific pathogen-free conditions at our animal facilities (FEM Charité Berlin, Germany). All animal experiments were in accordance with the ethical standards of the institution or practice at which the studies were conducted and were reviewed and approved by the responsible ethics committees (LAGeSo Berlin, I C 113 – G0172/14).

Antibodies

A list of antibodies used in this study is provided in Supplementary file 3.

Cell isolation from small intestine and flow cytometry

Request a detailed protocol

Small intestinal tissue was treated with HBSS buffer (without calcium and magnesium) containing 5 mM EDTA and 10 mM HEPES (pH 7.5) at 37°C for 30 min to remove epithelial cells, minced and digested in HBSS buffer (with calcium and magnesium) containing 10 mM HEPES, 4% FCS, 0.5 mg/ml Collagenase D, 0.5 mg/ml DNaseI (Sigma), 0.5 U/ml Dispase (BD) with constantly stirring at 37°C for 30 min. The supernatant was filtered and the remaining tissue was mashed through a 70 μm mesh. siLP cells were separated using a 40%/80% step-gradient (Percoll solution, GE Healthcare). Flow cytometry was performed according to previously defined guidelines (Melo-Gonzalez and Hepworth, 2017). In detail, single-cell suspensions were stained with different antibodies (Supplementary file 3). For cytokine analysis, cells were restimulated with PMA (Sigma, 10 ng/ml), ionomycin (Sigma, 1 μg/ml) and IL-23 (50 ng/ml) for 5 hr in TexMACS medium (Miltenyi Biotec) containing 10% FCS. After 1 hr of stimulation, Brefeldin A (Sigma, 5 μg/ml) was added to block cytokine secretion. For intracellular staining of cytokines and transcription factors, cells were first stained for surface markers and dead cells were labeled with Fixable Viability Dye eFluor780 (eBioscience). After that, cells were fixed in Fix/Perm buffer (eBioscience) at 4°C for 1 hr, followed by permeabilization (eBioscience) at 4°C for 2 hr in the presence of antibodies. Cells were acquired with a BD LSRFortessa X-20 and analysis was performed with FlowJo (Tree Star) software.

RNA-seq analysis

Request a detailed protocol

NKp46⁺ CCR6^- ILC3s and CCR6⁺ ILC3s were sorted from the siLP of 8–12 weeks old Rorc^Cre Maf^fl/fl or littermate Maf^fl/fl control mice using a BD FACSAria sorter (sorting strategy Figure 2—figure supplement 2A). RNA was isolated with the RNeasy Micro kit from Qiagen according to the manufacturer`s protocol. RNA libraries were prepared using the Smart-Seq V4 Ultra low Input RNA kit (Takara Clontech). Sequencing was performed on an Illumina Nextseq 500 generating 75 bp paired-end reads. Three biological replicates of each subset were sequenced. Raw sequence reads were mapped to the mouse GRCm38/mm10 genome with TopHat2 (Kim et al., 2013) in very-sensitive settings for Bowtie2 (Langmead and Salzberg, 2012). Gene expression was quantified either by HTSeq (Anders et al., 2015) for total RNA or featureCounts (Liao et al., 2014) for mRNA and analyzed using DESeq2 (Love et al., 2014). A cut-off (FC > 1.5 and p-value < 0.05) was applied for calling differentially expressed genes. Furthermore, differentially expressed genes were filtered for ‘gene_type = protein_coding’ before further analysis.

Gene set enrichment analysis (GSEA)

Request a detailed protocol

GSEA was performed using the GSEA tool from the Broad Institute. Gene sets used in this study were generated by taking the top upregulated genes (log₂FC > 2) from published differential gene expression analysis of RNA-seq data comparing ILC1s or NK cells with CCR6^- ILC3s (Pokrovskii et al., 2019).

Luciferase reporter assay

Request a detailed protocol

HEK293T cells were transfected with the pGL3 basic luciferase plasmid (Promega) containing the T-bet promoter alone or the T-bet promoter in combination with an upstream enhancer region (Yang et al., 2007), or the empty pGL3 basic in combination with an internal control pRL-TK Renilla plasmid (Promega). The T-bet enhancer/promoter plasmid was described before (Hosokawa et al., 2013) and kindly provided by H. Hosokawa (Tokai University, Japan). In order to assess gene regulation by c-Maf, putative Maf responsive elements in the promoter and enhancer were mutated using the Q5 Site-Directed Mutagenesis Kit (New England Biolabs). In addition to mutated reporter plasmids, cells were co-transfected with c-Maf coding sequence in pMSCV. Luciferase activity was measured on a SpectraMax i33 microplate reader (Molecular Devices) after 24 hr using dual luciferase assay system (Promega). Luciferase activity was determined relative to Renilla.

In vitro stimulation of NKp46⁺ CCR6^- ILC3s with cytokines

Request a detailed protocol

CD45⁺ Lineage^- (Lineage: anti-CD19, anti-Gr-1, anti-CD3, anti-CD5) RORγt^fm+ CD127⁺ NKp46⁺ CCR6^- cells were sort-purified from the siLP of 11–14 week old Rorc^Cre R26^EYFP mice. Sorted cells were transferred in complete RPMI medium to a 96 U bottom well plate at a density of 15.000 cells/well. Subsequently cells were cultured in the presence of IL-7 (20 ng/ml) and SCF (20 ng/ml) plus different cytokines (IL-1ß, IL-12, IL-15, IL-18, IL-23; each at 20 ng/ml) for 36 hr before mRNA analysis. For blocking NF-κb activation, BMS-345541, a selective inhibitor of IκB kinase (Burke et al., 2003), was added at a concentration of 1 μM to the culture.

In vitro culture of NKp46⁺ CCR6^- ILC3s on OP9-DLL1 cells

Request a detailed protocol

CD45⁺ Lineage^- (Lineage: anti-CD19, anti-Gr-1, anti-CD3, anti-CD5) RORγt^fm+ CD127⁺ NKp46⁺ CCR6^- cells were sort-purified from the siLP of 11–14 week old Rorc^Cre R26^EYFP mice. Sorted cells were transferred in complete RPMI medium to OP9 or OP9-DLL1 cells at a density of 10.000 cells/well and cultured in the presence of IL-7 (20 ng/ml) and SCF (20 ng/ml) for 12 days before flow cytometric analysis. OP9 cells are murine stromal cells derived from OP/OP mice used as feeder cells in lymphocyte differentiation assays. OP9-DLL1 cells are transfected with Notch ligand delta-like-1 (Schmitt and Zúñiga-Pflücker, 2002). Prior to adding isolated lymphocytes, confluent feeder cells were treated with 5 μg/ml Mitomycin C (Sigma) for 3 hr at 37°C and subsequently seeded on a 96 flat bottom well plate at a density of 50.000 cells/well.

Bone marrow chimeras

Request a detailed protocol

Bone marrow cells from wild-type CD45.1⁺ CD90.2⁺ C57BL/6 and CD45.2⁺ CD90.2⁺ Rorc^Cre Maf^fl/fl mice were mixed in a 1:1 ratio and intravenously injected into sub-lethally irradiated CD90.1⁺ wild-type recipient mice. Small and colonic lamina propria of reconstituted mice were analysed 6 weeks after cell transfer.

qPCR

Request a detailed protocol

mRNA for real-time qPCR was isolated with the RNeasy Plus Micro Kit according to the manual of the manufacturer (QIAGEN). Reverse transcription was done with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) as it is described in the manufacturer’s protocol. qPCR was performed using a Quant Studio five system (Applied Biosystems) and the SYBR Green PCR Master Mix Kit (Applied Biosystems). The mRNA expression is presented relative to the expression of the housekeeping gene hypoxanthine-guanine phosphoribosyl-transferase (HPRT). Real-time qPCR primer can be found in Supplementary file 4.

Statistical analysis

Request a detailed protocol

Data are the mean with SEM and summarize or are representative of independent experiments as specified in the text. Statistical analyses were performed using Prism software (GraphPad) with two-tailed unpaired Student's t test (except RNA-seq data).

Data availability

Sequencing data supporting the findings of this study have been deposited in the Gene Expression Omnibus (GEO) database under the GEO accession number: RNA-Seq: GSE143867.

The following data sets were generated

(2020) NCBI Gene Expression Omnibus
ID GSE143867. c-Maf restrains T-bet-driven programming of CCR6-negative group 3 innate lymphoid cells.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE143867

The following previously published data sets were used

1. Pokrovskii M
2. Hall JA
3. Ochayon DE
4. Yi R
5. Chaimowitz NS
6. Seelamneni H
7. Carriero N
8. Watters A
9. Waggoner SN
10. Littman DR
11. Bonneau R
12. Miraldi ER
(2019) NCBI Gene Expression Omnibus
ID GSE116092. Gene expression (RNA-seq) of innate lymphoid cells of the small intestine (SI) and large intestine (LI) lamina propria.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE116092
1. Gökmen MR
2. Dong R
3. Kanhere A
4. Powell N
5. Perucha E
6. Jackson I
7. Howard JK
8. Hernandez-Fuentes M
9. Jenner RG
10. Lord GM
(2013) NCBI Gene Expression Omnibus
ID GSE40623. ChIP-seq analysis of T-bet in WT mice (Th1 cells).

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE77695
1. Shih HY
2. Sciumè G
3. Mikami Y
4. Guo L
5. Sun HW
6. Brooks SR
7. Urban JF Jr
8. Davis FP
9. Kanno Y
10. O'Shea JJ
(2016) NCBI Gene Expression Omnibus
ID GSE77695. Integrated analysis of epigenome and transcriptome data from innate lymphoid cells and their progenitors.

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE77695

References

1. Adams NM
2. Lau CM
3. Fan X
4. Rapp M
5. Geary CD
6. Weizman OE
7. Diaz-Salazar C
8. Sun JC
(2018) Transcription factor IRF8 orchestrates the adaptive natural killer cell response
Immunity 48:1172–1182.

https://doi.org/10.1016/j.immuni.2018.04.018
- PubMed
- Google Scholar
1. Anders S
2. Pyl PT
3. Huber W
(2015) HTSeq--a Python framework to work with high-throughput sequencing data
Bioinformatics 31:166–169.

https://doi.org/10.1093/bioinformatics/btu638
- PubMed
- Google Scholar
(2018) An immunoregulatory and tissue-residency program modulated by c-MAF in human T_H17 cells
Nature Immunology 19:1126–1136.

https://doi.org/10.1038/s41590-018-0200-5
- PubMed
- Google Scholar
1. Bernink JH
2. Peters CP
3. Munneke M
4. te Velde AA
5. Meijer SL
6. Weijer K
7. Hreggvidsdottir HS
8. Heinsbroek SE
9. Legrand N
10. Buskens CJ
11. Bemelman WA
12. Mjösberg JM
13. Spits H
(2013) Human type 1 innate lymphoid cells accumulate in inflamed mucosal tissues
Nature Immunology 14:221–229.

https://doi.org/10.1038/ni.2534
- PubMed
- Google Scholar
1. Bernink JH
2. Krabbendam L
3. Germar K
4. de Jong E
5. Gronke K
6. Kofoed-Nielsen M
7. Munneke JM
8. Hazenberg MD
9. Villaudy J
10. Buskens CJ
11. Bemelman WA
12. Diefenbach A
13. Blom B
14. Spits H
(2015) Interleukin-12 and -23 control plasticity of CD127+ group 1 and group 3 innate lymphoid cells in the intestinal Lamina propria
Immunity 43:146–160.

https://doi.org/10.1016/j.immuni.2015.06.019
- Google Scholar
1. Buonocore S
2. Ahern PP
3. Uhlig HH
4. Ivanov II
5. Littman DR
6. Maloy KJ
7. Powrie F
(2010) Innate lymphoid cells drive interleukin-23-dependent innate intestinal pathology
Nature 464:1371–1375.

https://doi.org/10.1038/nature08949
- PubMed
- Google Scholar
1. Burke JR
2. Pattoli MA
3. Gregor KR
4. Brassil PJ
5. MacMaster JF
6. McIntyre KW
7. Yang X
8. Iotzova VS
9. Clarke W
10. Strnad J
11. Qiu Y
12. Zusi FC
(2003) BMS-345541 is a highly selective inhibitor of I kappa B kinase that binds at an allosteric site of the enzyme and blocks NF-kappa B-dependent transcription in mice
Journal of Biological Chemistry 278:1450–1456.

https://doi.org/10.1074/jbc.M209677200
- PubMed
- Google Scholar
1. Cella M
2. Gamini R
3. Sécca C
4. Collins PL
5. Zhao S
6. Peng V
7. Robinette ML
8. Schettini J
9. Zaitsev K
10. Gordon W
11. Bando JK
12. Yomogida K
13. Cortez V
14. Fronick C
15. Fulton R
16. Lin LL
17. Gilfillan S
18. Flavell RA
19. Shan L
20. Artyomov MN
21. Bowman M
22. Oltz EM
23. Jelinsky SA
24. Colonna M
(2019) Subsets of ILC3-ILC1-like cells generate a diversity spectrum of innate lymphoid cells in human mucosal tissues
Nature Immunology 20:980–991.

https://doi.org/10.1038/s41590-019-0425-y
- PubMed
- Google Scholar
1. Chea S
2. Perchet T
3. Petit M
4. Verrier T
5. Guy-Grand D
6. Banchi EG
7. Vosshenrich CA
8. Di Santo JP
9. Cumano A
10. Golub R
(2016) Notch signaling in group 3 innate lymphoid cells modulates their plasticity
Science Signaling 9:ra45.

https://doi.org/10.1126/scisignal.aaf2223
- PubMed
- Google Scholar
1. Ciofani M
2. Madar A
3. Galan C
4. Sellars M
5. Mace K
6. Pauli F
7. Agarwal A
8. Huang W
9. Parkhurst CN
10. Muratet M
11. Newberry KM
12. Meadows S
13. Greenfield A
14. Yang Y
15. Jain P
16. Kirigin FK
17. Birchmeier C
18. Wagner EF
19. Murphy KM
20. Myers RM
21. Bonneau R
22. Littman DR
(2012) A validated regulatory network for Th17 cell specification
Cell 151:289–303.

https://doi.org/10.1016/j.cell.2012.09.016
- PubMed
- Google Scholar
1. Cortez VS
2. Colonna M
(2016) Diversity and function of group 1 innate lymphoid cells
Immunology Letters 179:19–24.

https://doi.org/10.1016/j.imlet.2016.07.005
- PubMed
- Google Scholar
1. Eberl G
2. Marmon S
3. Sunshine MJ
4. Rennert PD
5. Choi Y
6. Littman DR
(2004) An essential function for the nuclear receptor RORgamma(t) in the generation of fetal lymphoid tissue inducer cells
Nature Immunology 5:64–73.

https://doi.org/10.1038/ni1022
- PubMed
- Google Scholar
1. Eberl G
2. Littman DR
(2004) Thymic origin of intestinal alphabeta T cells revealed by fate mapping of RORgammat+ cells
Science 305:248–251.

https://doi.org/10.1126/science.1096472
- PubMed
- Google Scholar
1. Fang D
2. Zhu J
(2017) Dynamic balance between master transcription factors determines the fates and functions of CD4 T cell and innate lymphoid cell subsets
The Journal of Experimental Medicine 214:1861–1876.

https://doi.org/10.1084/jem.20170494
- PubMed
- Google Scholar
(2013) Cutting edge: il-17-secreting innate lymphoid cells are essential for host defense against fungal infection
The Journal of Immunology 190:521–525.

https://doi.org/10.4049/jimmunol.1202924
- PubMed
- Google Scholar
1. Gökmen MR
2. Dong R
3. Kanhere A
4. Powell N
5. Perucha E
6. Jackson I
7. Howard JK
8. Hernandez-Fuentes M
9. Jenner RG
10. Lord GM
(2013) Genome-wide regulatory analysis reveals that T-bet controls Th17 lineage differentiation through direct suppression of IRF4
The Journal of Immunology 191:5925–5932.

https://doi.org/10.4049/jimmunol.1202254
- PubMed
- Google Scholar
(2019) Interleukin-22 protects intestinal stem cells against genotoxic stress
Nature 566:249–253.

https://doi.org/10.1038/s41586-019-0899-7
- PubMed
- Google Scholar
1. Heinz S
2. Benner C
3. Spann N
4. Bertolino E
5. Lin YC
6. Laslo P
7. Cheng JX
8. Murre C
9. Singh H
10. Glass CK
(2010) Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities
Molecular Cell 38:576–589.

https://doi.org/10.1016/j.molcel.2010.05.004
- PubMed
- Google Scholar
1. Hernández PP
2. Mahlakoiv T
3. Yang I
4. Schwierzeck V
5. Nguyen N
6. Guendel F
7. Gronke K
8. Ryffel B
9. Hoelscher C
10. Dumoutier L
11. Renauld JC
12. Suerbaum S
13. Staeheli P
14. Diefenbach A
(2015) Interferon-λ and interleukin 22 act synergistically for the induction of interferon-stimulated genes and control of Rotavirus infection
Nature Immunology 16:698–707.

https://doi.org/10.1038/ni.3180
- PubMed
- Google Scholar
1. Ho IC
2. Lo D
3. Glimcher LH
(1998) c-maf promotes T helper cell type 2 (Th2) and attenuates Th1 differentiation by both interleukin 4–dependent and –independent Mechanisms
The Journal of Experimental Medicine 188:1859–1866.

https://doi.org/10.1084/jem.188.10.1859
- PubMed
- Google Scholar
1. Hosokawa H
2. Tanaka T
3. Suzuki Y
4. Iwamura C
5. Ohkubo S
6. Endoh K
7. Kato M
8. Endo Y
9. Onodera A
10. Tumes DJ
11. Kanai A
12. Sugano S
13. Nakayama T
(2013) Functionally distinct Gata3/Chd4 complexes coordinately establish T helper 2 (Th2) cell identity
PNAS 110:4691–4696.

https://doi.org/10.1073/pnas.1220865110
- PubMed
- Google Scholar
(2019) Maf deficiency in T cells dysregulates T_reg - T_H17 balance leading to spontaneous colitis
Scientific Reports 9:6135.

https://doi.org/10.1038/s41598-019-42486-2
- PubMed
- Google Scholar
1. Ivanov II
2. McKenzie BS
3. Zhou L
4. Tadokoro CE
5. Lepelley A
6. Lafaille JJ
7. Cua DJ
8. Littman DR
(2006) The orphan nuclear receptor RORgammat directs the differentiation program of proinflammatory IL-17+ T helper cells
Cell 126:1121–1133.

https://doi.org/10.1016/j.cell.2006.07.035
- PubMed
- Google Scholar
(1994) Maf nuclear oncoprotein recognizes sequences related to an AP-1 site and forms heterodimers with both fos and jun
Molecular and Cellular Biology 14:700–712.

https://doi.org/10.1128/MCB.14.1.700
- Google Scholar
1. Kim D
2. Pertea G
3. Trapnell C
4. Pimentel H
5. Kelley R
6. Salzberg SL
(2013) TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions
Genome Biology 14:R36.

https://doi.org/10.1186/gb-2013-14-4-r36
- PubMed
- Google Scholar
1. Klose CS
2. Kiss EA
3. Schwierzeck V
4. Ebert K
5. Hoyler T
6. d'Hargues Y
7. Göppert N
8. Croxford AL
9. Waisman A
10. Tanriver Y
11. Diefenbach A
(2013) A T-bet gradient controls the fate and function of CCR6-RORγt+ innate lymphoid cells
Nature 494:261–265.

https://doi.org/10.1038/nature11813
- PubMed
- Google Scholar
1. Klose CSN
2. Flach M
3. Möhle L
4. Rogell L
5. Hoyler T
6. Ebert K
7. Fabiunke C
8. Pfeifer D
9. Sexl V
10. Fonseca-Pereira D
11. Domingues RG
12. Veiga-Fernandes H
13. Arnold SJ
14. Busslinger M
15. Dunay IR
16. Tanriver Y
17. Diefenbach A
(2014) Differentiation of type 1 ILCs from a common progenitor to all helper-like innate lymphoid cell lineages
Cell 157:340–356.

https://doi.org/10.1016/j.cell.2014.03.030
- PubMed
- Google Scholar
1. Langmead B
2. Salzberg SL
(2012) Fast gapped-read alignment with bowtie 2
Nature Methods 9:357–359.

https://doi.org/10.1038/nmeth.1923
- PubMed
- Google Scholar
1. Lee PP
2. Fitzpatrick DR
3. Beard C
4. Jessup HK
5. Lehar S
6. Makar KW
7. Pérez-Melgosa M
8. Sweetser MT
9. Schlissel MS
10. Nguyen S
11. Cherry SR
12. Tsai JH
13. Tucker SM
14. Weaver WM
15. Kelso A
16. Jaenisch R
17. Wilson CB
(2001) A critical role for Dnmt1 and DNA methylation in T cell development, function, and survival
Immunity 15:763–774.

https://doi.org/10.1016/S1074-7613(01)00227-8
- PubMed
- Google Scholar
1. Liao Y
2. Smyth GK
3. Shi W
(2014) featureCounts: an efficient general purpose program for assigning sequence reads to genomic features
Bioinformatics 30:923–930.

https://doi.org/10.1093/bioinformatics/btt656
- Google Scholar
1. Love MI
2. Huber W
3. Anders S
(2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2
Genome Biology 15:550.

https://doi.org/10.1186/s13059-014-0550-8
- Google Scholar
1. Mei S
2. Qin Q
3. Wu Q
4. Sun H
5. Zheng R
6. Zang C
7. Zhu M
8. Wu J
9. Shi X
10. Taing L
11. Liu T
12. Brown M
13. Meyer CA
14. Liu XS
(2017) Cistrome data browser: a data portal for ChIP-Seq and chromatin accessibility data in human and mouse
Nucleic Acids Research 45:D658–D662.

https://doi.org/10.1093/nar/gkw983
- PubMed
- Google Scholar
1. Melo-Gonzalez F
2. Hepworth MR
(2017) Functional and phenotypic heterogeneity of group 3 innate lymphoid cells
Immunology 150:265–275.

https://doi.org/10.1111/imm.12697
- PubMed
- Google Scholar
1. Neumann C
2. Heinrich F
3. Neumann K
4. Junghans V
5. Mashreghi MF
6. Ahlers J
7. Janke M
8. Rudolph C
9. Mockel-Tenbrinck N
10. Kühl AA
11. Heimesaat MM
12. Esser C
13. Im SH
14. Radbruch A
15. Rutz S
16. Scheffold A
(2014) Role of Blimp-1 in programing th effector cells into IL-10 producers
The Journal of Experimental Medicine 211:1807–1819.

https://doi.org/10.1084/jem.20131548
- PubMed
- Google Scholar
1. Neumann C
2. Blume J
3. Roy U
4. Teh PP
5. Vasanthakumar A
6. Beller A
7. Liao Y
8. Heinrich F
9. Arenzana TL
10. Hackney JA
11. Eidenschenk C
12. Gálvez EJC
13. Stehle C
14. Heinz GA
15. Maschmeyer P
16. Sidwell T
17. Hu Y
18. Amsen D
19. Romagnani C
20. Chang HD
21. Kruglov A
22. Mashreghi MF
23. Shi W
24. Strowig T
25. Rutz S
26. Kallies A
27. Scheffold A
(2019) c-Maf-dependent T_reg cell control of intestinal T_H17 cells and IgA establishes host-microbiota homeostasis
Nature Immunology 20:471–481.

https://doi.org/10.1038/s41590-019-0316-2
- PubMed
- Google Scholar
1. Pokrovskii M
2. Hall JA
3. Ochayon DE
4. Yi R
5. Chaimowitz NS
6. Seelamneni H
7. Carriero N
8. Watters A
9. Waggoner SN
10. Littman DR
11. Bonneau R
12. Miraldi ER
(2019) Characterization of transcriptional regulatory networks that promote and restrict identities and functions of intestinal innate lymphoid cells
Immunity 51:185–197.

https://doi.org/10.1016/j.immuni.2019.06.001
- PubMed
- Google Scholar
1. Powell N
2. Walker AW
3. Stolarczyk E
4. Canavan JB
5. Gökmen MR
6. Marks E
7. Jackson I
8. Hashim A
9. Curtis MA
10. Jenner RG
11. Howard JK
12. Parkhill J
13. MacDonald TT
14. Lord GM
(2012) The transcription factor T-bet regulates intestinal inflammation mediated by interleukin-7 receptor+ innate lymphoid cells
Immunity 37:674–684.

https://doi.org/10.1016/j.immuni.2012.09.008
- PubMed
- Google Scholar
1. Rankin LC
2. Groom JR
3. Chopin M
4. Herold MJ
5. Walker JA
6. Mielke LA
7. McKenzie AN
8. Carotta S
9. Nutt SL
10. Belz GT
(2013) The transcription factor T-bet is essential for the development of NKp46+ innate lymphocytes via the notch pathway
Nature Immunology 14:389–395.

https://doi.org/10.1038/ni.2545
- PubMed
- Google Scholar
1. Rankin LC
2. Girard-Madoux MJ
3. Seillet C
4. Mielke LA
5. Kerdiles Y
6. Fenis A
7. Wieduwild E
8. Putoczki T
9. Mondot S
10. Lantz O
11. Demon D
12. Papenfuss AT
13. Smyth GK
14. Lamkanfi M
15. Carotta S
16. Renauld JC
17. Shi W
18. Carpentier S
19. Soos T
20. Arendt C
21. Ugolini S
22. Huntington ND
23. Belz GT
24. Vivier E
(2016) Complementarity and redundancy of IL-22-producing innate lymphoid cells
Nature Immunology 17:179–186.

https://doi.org/10.1038/ni.3332
- PubMed
- Google Scholar
(2015) Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets
Nature Immunology 16:306–317.

https://doi.org/10.1038/ni.3094
- PubMed
- Google Scholar
1. Rutz S
2. Noubade R
3. Eidenschenk C
4. Ota N
5. Zeng W
6. Zheng Y
7. Hackney J
8. Ding J
9. Singh H
10. Ouyang W
(2011) Transcription factor c-Maf mediates the TGF-β-dependent suppression of IL-22 production in T(H)17 cells
Nature Immunology 12:1238–1245.

https://doi.org/10.1038/ni.2134
- PubMed
- Google Scholar
(2013) IL-22, not simply a Th17 cytokine
Immunological Reviews 252:116–132.

https://doi.org/10.1111/imr.12027
- PubMed
- Google Scholar
1. Sanos SL
2. Bui VL
3. Mortha A
4. Oberle K
5. Heners C
6. Johner C
7. Diefenbach A
(2009) RORγt and commensal microflora are required for the differentiation of mucosal interleukin 22-producing NKp46+ cells
Nature Immunology 10:83–91.

https://doi.org/10.1038/ni.1684
- PubMed
- Google Scholar
(2008) Microbial flora drives interleukin 22 production in intestinal NKp46+ cells that provide innate mucosal immune defense
Immunity 29:958–970.

https://doi.org/10.1016/j.immuni.2008.11.001
- PubMed
- Google Scholar
(2010) IL-7 and IL-15 independently program the differentiation of intestinal CD3-NKp46+ cell subsets from Id2-dependent precursors
The Journal of Experimental Medicine 207:273–280.

https://doi.org/10.1084/jem.20092029
- PubMed
- Google Scholar
1. Sawa S
2. Cherrier M
3. Lochner M
4. Satoh-Takayama N
5. Fehling HJ
6. Langa F
7. Di Santo JP
8. Eberl G
(2010) Lineage relationship analysis of RORgammat+ innate lymphoid cells
Science 330:665–669.

https://doi.org/10.1126/science.1194597
- PubMed
- Google Scholar
1. Schmitt TM
2. Zúñiga-Pflücker JC
(2002) Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro
Immunity 17:749–756.

https://doi.org/10.1016/S1074-7613(02)00474-0
- PubMed
- Google Scholar
1. Sciumé G
2. Hirahara K
3. Takahashi H
4. Laurence A
5. Villarino AV
6. Singleton KL
7. Spencer SP
8. Wilhelm C
9. Poholek AC
10. Vahedi G
11. Kanno Y
12. Belkaid Y
13. O'Shea JJ
(2012) Distinct requirements for T-bet in gut innate lymphoid cells
The Journal of Experimental Medicine 209:2331–2338.

https://doi.org/10.1084/jem.20122097
- PubMed
- Google Scholar
1. Shih HY
2. Sciumè G
3. Mikami Y
4. Guo L
5. Sun HW
6. Brooks SR
7. Urban JF
8. Davis FP
9. Kanno Y
10. O'Shea JJ
(2016) Developmental acquisition of regulomes underlies innate lymphoid cell functionality
Cell 165:1120–1133.

https://doi.org/10.1016/j.cell.2016.04.029
- PubMed
- Google Scholar
1. Song C
2. Lee JS
3. Gilfillan S
4. Robinette ML
5. Newberry RD
6. Stappenbeck TS
7. Mack M
8. Cella M
9. Colonna M
(2015) Unique and redundant functions of NKp46+ ILC3s in models of intestinal inflammation
The Journal of Experimental Medicine 212:1869–1882.

https://doi.org/10.1084/jem.20151403
- PubMed
- Google Scholar
1. Sonnenberg GF
2. Monticelli LA
3. Alenghat T
4. Fung TC
5. Hutnick NA
6. Kunisawa J
7. Shibata N
8. Grunberg S
9. Sinha R
10. Zahm AM
11. Tardif MR
12. Sathaliyawala T
13. Kubota M
14. Farber DL
15. Collman RG
16. Shaked A
17. Fouser LA
18. Weiner DB
19. Tessier PA
20. Friedman JR
21. Kiyono H
22. Bushman FD
23. Chang KM
24. Artis D
(2012) Innate lymphoid cells promote anatomical containment of lymphoid-resident commensal Bacteria
Science 336:1321–1325.

https://doi.org/10.1126/science.1222551
- PubMed
- Google Scholar
1. Srinivas S
2. Watanabe T
3. Lin CS
4. William CM
5. Tanabe Y
6. Jessell TM
7. Costantini F
(2001) Cre reporter strains produced by targeted insertion of EYFP and ECFP into the ROSA26 locus
BMC Developmental Biology 1:4.

https://doi.org/10.1186/1471-213x-1-4
- PubMed
- Google Scholar
1. Sun Z
2. Unutmaz D
3. Zou YR
4. Sunshine MJ
5. Pierani A
6. Brenner-Morton S
7. Mebius RE
8. Littman DR
(2000) Requirement for RORgamma in thymocyte survival and lymphoid organ development
Science 288:2369–2373.

https://doi.org/10.1126/science.288.5475.2369
- PubMed
- Google Scholar
(2002) Distinct effects of T-bet in TH1 lineage commitment and IFN-gamma production in CD4 and CD8 T cells
Science 295:338–342.

https://doi.org/10.1126/science.1065543
- PubMed
- Google Scholar
1. Tanaka S
2. Suto A
3. Iwamoto T
4. Kashiwakuma D
5. Kagami S
6. Suzuki K
7. Takatori H
8. Tamachi T
9. Hirose K
10. Onodera A
11. Suzuki J
12. Ohara O
13. Yamashita M
14. Nakayama T
15. Nakajima H
(2014) Sox5 and c-Maf cooperatively induce Th17 cell differentiation via rorγt induction as downstream targets of Stat3
The Journal of Experimental Medicine 211:1857–1874.

https://doi.org/10.1084/jem.20130791
- PubMed
- Google Scholar
1. Viant C
2. Rankin LC
3. Girard-Madoux MJ
4. Seillet C
5. Shi W
6. Smyth MJ
7. Bartholin L
8. Walzer T
9. Huntington ND
10. Vivier E
11. Belz GT
(2016) Transforming growth factor-β and notch ligands act as opposing environmental cues in regulating the plasticity of type 3 innate lymphoid cells
Science Signaling 9:ra46.

https://doi.org/10.1126/scisignal.aaf2176
- PubMed
- Google Scholar
1. Vivier E
2. Artis D
3. Colonna M
4. Diefenbach A
5. Di Santo JP
6. Eberl G
7. Koyasu S
8. Locksley RM
9. McKenzie ANJ
10. Mebius RE
11. Powrie F
12. Spits H
(2018) Innate lymphoid cells: 10 years on
Cell 174:1054–1066.

https://doi.org/10.1016/j.cell.2018.07.017
- PubMed
- Google Scholar
1. Vonarbourg C
2. Mortha A
3. Bui VL
4. Hernandez PP
5. Kiss EA
6. Hoyler T
7. Flach M
8. Bengsch B
9. Thimme R
10. Hölscher C
11. Hönig M
12. Pannicke U
13. Schwarz K
14. Ware CF
15. Finke D
16. Diefenbach A
(2010) Regulated expression of nuclear receptor rorγt confers distinct functional fates to NK cell receptor-expressing rorγt(+) innate lymphocytes
Immunity 33:736–751.

https://doi.org/10.1016/j.immuni.2010.10.017
- PubMed
- Google Scholar
1. Wende H
2. Lechner SG
3. Cheret C
4. Bourane S
5. Kolanczyk ME
6. Pattyn A
7. Reuter K
8. Munier FL
9. Carroll P
10. Lewin GR
11. Birchmeier C
(2012) The transcription factor c-Maf controls touch receptor development and function
Science 335:1373–1376.

https://doi.org/10.1126/science.1214314
- PubMed
- Google Scholar
(2017) Cutting edge: c-maf is required for regulatory T cells to adopt rorγt⁺ and Follicular Phenotypes
The Journal of Immunology 199:3931–3936.

https://doi.org/10.4049/jimmunol.1701134
- PubMed
- Google Scholar
1. Xu M
2. Pokrovskii M
3. Ding Y
4. Yi R
5. Au C
6. Harrison OJ
7. Galan C
8. Belkaid Y
9. Bonneau R
10. Littman DR
(2018) c-MAF-dependent regulatory T cells mediate immunological tolerance to a gut pathobiont
Nature 554:373–377.

https://doi.org/10.1038/nature25500
- PubMed
- Google Scholar
1. Yang Y
2. Ochando JC
3. Bromberg JS
4. Ding Y
(2007) Identification of a distant T-bet enhancer responsive to IL-12/Stat4 and IFNgamma/Stat1 signals
Blood 110:2494–2500.

https://doi.org/10.1182/blood-2006-11-058271
- PubMed
- Google Scholar
(2008) Innate and adaptive interleukin-22 protects mice from inflammatory bowel disease
Immunity 29:947–957.

https://doi.org/10.1016/j.immuni.2008.11.003
- PubMed
- Google Scholar
1. Zheng Y
2. Valdez PA
3. Danilenko DM
4. Hu Y
5. Sa SM
6. Gong Q
7. Abbas AR
8. Modrusan Z
9. Ghilardi N
10. de Sauvage FJ
11. Ouyang W
(2008) Interleukin-22 mediates early host defense against attaching and effacing bacterial pathogens
Nature Medicine 14:282–289.

https://doi.org/10.1038/nm1720
- PubMed
- Google Scholar
(2019) The transcription factor c-Maf is essential for the commitment of IL-17-producing γδ T cells
Nature Immunology 20:73–85.

https://doi.org/10.1038/s41590-018-0274-0
- PubMed
- Google Scholar

Article and author information

Author details

Caroline Tizian
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Data curation, Formal analysis, Validation, Investigation, Methodology, Designed and performed experiments, Analysed data

Contributed equally with
Annette Lahmann

Competing interests
No competing interests declared
Annette Lahmann

Chronic Immune Reactions, Deutsches Rheuma-Forschungszentrum, Berlin, Germany

Contribution
Data curation, Formal analysis, Validation, Investigation, Methodology, Designed and performed experiments, Analysed data

Contributed equally with
Caroline Tizian

Competing interests
No competing interests declared
Oliver Hölsken
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Formal analysis, Validation, Investigation, Methodology, Designed and performed experiments, Analysed data

Competing interests
No competing interests declared

"This ORCID iD identifies the author of this article:" 0000-0001-6086-9275
Catalina Cosovanu
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Formal analysis, Methodology, Performed experiments, Analysed data

Competing interests
No competing interests declared
Michael Kofoed-Branzk
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Methodology, Provided discussion, Proofread the manuscript

Competing interests
No competing interests declared
Frederik Heinrich

Therapeutic Gene Regulation, Deutsches Rheuma-Forschungszentrum, Berlin, Germany

Contribution
Formal analysis, Methodology, Performed computational analyses of 16S rRNA-seq data

Competing interests
No competing interests declared
Mir-Farzin Mashreghi

Therapeutic Gene Regulation, Deutsches Rheuma-Forschungszentrum, Berlin, Germany

Contribution
Resources, Supervision, Supervised sequencing experiments, Provided reagents and equipment for their execution

Competing interests
No competing interests declared

"This ORCID iD identifies the author of this article:" 0000-0002-8015-6907
Andrey Kruglov
1. Chronic Inflammation, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
2. Belozersky Institute of Physico-Chemical Biology and Biological Faculty, M.V. Lomonosov Moscow State University, Moscow, Russian Federation
Contribution
Resources, Methodology, Assisted in chimera experiments

Competing interests
No competing interests declared
Andreas Diefenbach
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Conceptualization, Resources, Supervision, Funding acquisition, Provided crucial reagents, Contributed to the design of the study, Proofread the manuscript

Competing interests
No competing interests declared
Christian Neumann
1. Laboratory of Innate Immunity, Department of Microbiology, Infectious Diseases and Immunology, Charité-Universitätsmedizin Berlin, Berlin, Germany
2. Mucosal and Developmental Immunology, Deutsches Rheuma-Forschungszentrum, Berlin, Germany
Contribution
Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Project administration, Conceived the project, Designed and performed most experiments, Analysed the data, Generated the figures and wrote the manuscript with the input of all co-authors

For correspondence
c.neumann@charite.de

Competing interests
No competing interests declared

"This ORCID iD identifies the author of this article:" 0000-0003-2202-1876

Funding

Deutsche Forschungsgemeinschaft (Priority Program 1937 "Innate Lymphoid Cells")

Andreas Diefenbach
Christian Neumann

European Regional Development Fund (ERDF 2014-2020)

Frederik Heinrich
Mir-Farzin Mashreghi

European Regional Development Fund (EFRE 1.8/11)

Mir-Farzin Mashreghi

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

This research was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) Priority Program 1937 ‘Innate Lymphoid Cells’ (to CN project number: 428192857 and to AD project number: DI764/9-2). FH and MFM were supported by the state of Berlin and the ‘European Regional Development Fund’ (ERDF 2014–2020, EFRE 1.8/11, Deutsches Rheuma‐Forschungszentrum). We acknowledge support from the DFG and the Open Access Publication Funds of Charité – Universitätsmedizin Berlin. We thank Irene Mattiola, Christoph S N Klose, Fabian Guendel-Rojas, Mario Witkowski, Pawel Durek and Katrin Lehmann for discussion, technical and experimental help and proofreading of the manuscript.

Ethics

Animal experimentation: All animal experiments were in accordance with the ethical standards of the institution or practice at which the studies were conducted and were reviewed and approved by the responsible ethics committees (LAGeSo Berlin, I C 113 - G0172/14).

Copyright

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.