The Transcription Factor STAT6 Mediates Direct Repression of Inflammatory Enhancers and Limits Activation of Alternatively Polarized Macrophages

•IL-4-activated STAT6 acts as a transcriptional repressor in macrophages•IL-4-STAT6-repressed enhancers associate with reduced LDTF and p300 binding•Inflammatory responsiveness of the IL-4-repressed enhancers is attenuated•IL-4 limits the LPS-induced inflammasome activation, IL-1β production, and pyroptosis The molecular basis of signal-dependent transcriptional activation has been extensively studied in macrophage polarization, but our understanding remains limited regarding the molecular determinants of repression. Here we show that IL-4-activated STAT6 transcription factor is required for the direct transcriptional repression of a large number of genes during in vitro and in vivo alternative macrophage polarization. Repression results in decreased lineage-determining transcription factor, p300, and RNA polymerase II binding followed by reduced enhancer RNA expression, H3K27 acetylation, and chromatin accessibility. The repressor function of STAT6 is HDAC3 dependent on a subset of IL-4-repressed genes. In addition, STAT6-repressed enhancers show extensive overlap with the NF-κB p65 cistrome and exhibit decreased responsiveness to lipopolysaccharide after IL-4 stimulus on a subset of genes. As a consequence, macrophages exhibit diminished inflammasome activation, decreased IL-1β production, and pyroptosis. Thus, the IL-4-STAT6 signaling pathway establishes an alternative polarization-specific epigenenomic signature resulting in dampened macrophage responsiveness to inflammatory stimuli. The molecular basis of signal-dependent transcriptional activation has been extensively studied in macrophage polarization, but our understanding remains limited regarding the molecular determinants of repression. Here we show that IL-4-activated STAT6 transcription factor is required for the direct transcriptional repression of a large number of genes during in vitro and in vivo alternative macrophage polarization. Repression results in decreased lineage-determining transcription factor, p300, and RNA polymerase II binding followed by reduced enhancer RNA expression, H3K27 acetylation, and chromatin accessibility. The repressor function of STAT6 is HDAC3 dependent on a subset of IL-4-repressed genes. In addition, STAT6-repressed enhancers show extensive overlap with the NF-κB p65 cistrome and exhibit decreased responsiveness to lipopolysaccharide after IL-4 stimulus on a subset of genes. As a consequence, macrophages exhibit diminished inflammasome activation, decreased IL-1β production, and pyroptosis. Thus, the IL-4-STAT6 signaling pathway establishes an alternative polarization-specific epigenenomic signature resulting in dampened macrophage responsiveness to inflammatory stimuli. Macrophage plasticity is ensured by dynamic and partially reversible responsiveness to pathogen-derived molecules as well as the cytokine and lipid microenvironment. The two well-characterized extreme functional outcomes of macrophage polarization are T helper 1 (Th1) cell-type cytokine interferon-gamma (IFN-γ)-induced classical or M(INF-γ)-type polarization with enhanced bactericidal capacity and Th2 cell-type cytokine interleukin-4 (IL-4)-induced alternative or M(IL-4)-type polarization with anti-inflammatory properties, but complex molecular cues can generate an entire spectrum of different activation states (Gordon and Martinez, 2010Gordon S. Martinez F.O. Alternative activation of macrophages: mechanism and functions.Immunity. 2010; 32: 593-604Abstract Full Text Full Text PDF PubMed Scopus (2818) Google Scholar, Murray et al., 2014Murray P.J. Allen J.E. Biswas S.K. Fisher E.A. Gilroy D.W. Goerdt S. Gordon S. Hamilton J.A. Ivashkiv L.B. Lawrence T. et al.Macrophage activation and polarization: nomenclature and experimental guidelines.Immunity. 2014; 41: 14-20Abstract Full Text Full Text PDF PubMed Scopus (3561) Google Scholar). The major determinant of macrophage plasticity is their specific transcriptional program dictated primarily by lineage-determining transcription factors (LDTFs) including ETS-domain transcription factor PU.1, CCAAT/enhancer binding proteins (C/EBPs), activator protein 1 (AP-1), or Runt-related transcription factor 1 (RUNX1) as well as extracellular signal-dependent transcription factors (SDTFs) including LPS-activated nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) or AP-1, IFN-γ-activated signal transducer and activator of transcription 1 (STAT1), or IL-4- and IL-13-activated STAT6; for a review see Glass and Natoli, 2016Glass C.K. Natoli G. Molecular control of activation and priming in macrophages.Nat. Immunol. 2016; 17: 26-33Crossref PubMed Scopus (315) Google Scholar. Despite the fact that polarization signals repress large sets of genes, the repressive activity of polarization-specific transcription factors has not been studied in detail (Bhatt et al., 2012Bhatt D.M. Pandya-Jones A. Tong A.J. Barozzi I. Lissner M.M. Natoli G. Black D.L. Smale S.T. Transcript dynamics of proinflammatory genes revealed by sequence analysis of subcellular RNA fractions.Cell. 2012; 150: 279-290Abstract Full Text Full Text PDF PubMed Scopus (331) Google Scholar, Martinez et al., 2013Martinez F.O. Helming L. Milde R. Varin A. Melgert B.N. Draijer C. Thomas B. Fabbri M. Crawshaw A. Ho L.P. et al.Genetic programs expressed in resting and IL-4 alternatively activated mouse and human macrophages: similarities and differences.Blood. 2013; 121: e57-e69Crossref PubMed Scopus (349) Google Scholar). Recently, a whole new spectrum of next-generation sequencing-based methods has evolved, enabling the characterization of the molecular features of transcriptional repression in macrophages at an unprecedented level (Hah et al., 2015Hah N. Benner C. Chong L.W. Yu R.T. Downes M. Evans R.M. Inflammation-sensitive super enhancers form domains of coordinately regulated enhancer RNAs.Proc. Natl. Acad. Sci. USA. 2015; 112: E297-E302Crossref PubMed Scopus (111) Google Scholar, Kang et al., 2017Kang K. Park S.H. Chen J. Qiao Y. Giannopoulou E. Berg K. Hanidu A. Li J. Nabozny G. Kang K. et al.Interferon-γ represses M2 gene expression in human macrophages by disassembling enhancers bound by the transcription factor MAF.Immunity. 2017; 47: 235-250.e4Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar). IL-4- or IL-13-induced alternative macrophage polarization occurs in a number of pathological processes including nematode infection, tumor development, lung inflammation, and fibrosis (Gordon and Martinez, 2010Gordon S. Martinez F.O. Alternative activation of macrophages: mechanism and functions.Immunity. 2010; 32: 593-604Abstract Full Text Full Text PDF PubMed Scopus (2818) Google Scholar). Given the complex immunological milieu that characterizes each of these conditions, alternatively polarized macrophages are likely to encounter inflammatory stimuli as well (Fort et al., 2001Fort M. Lesley R. Davidson N. Menon S. Brombacher F. Leach M. Rennick D. IL-4 exacerbates disease in a Th1 cell transfer model of colitis.J. Immunol. 2001; 166: 2793-2800Crossref PubMed Scopus (70) Google Scholar, Ruffell et al., 2012Ruffell B. Affara N.I. Coussens L.M. Differential macrophage programming in the tumor microenvironment.Trends Immunol. 2012; 33: 119-126Abstract Full Text Full Text PDF PubMed Scopus (627) Google Scholar). It has been shown that in vitro modeling of complex immunological microenvironment by IL-4 and IFN-γ co-stimulation leads to the attenuation of IFN-γ-induced transcriptional activation due to the effects of IL-4 on restrictive set of auxiliary transcription factors in mouse macrophages (Piccolo et al., 2017Piccolo V. Curina A. Genua M. Ghisletti S. Simonatto M. Sabò A. Amati B. Ostuni R. Natoli G. Opposing macrophage polarization programs show extensive epigenomic and transcriptional cross-talk.Nat. Immunol. 2017; 18: 530-540Crossref PubMed Scopus (122) Google Scholar). These results suggest that alternatively polarized macrophages exhibit an altered responsiveness to inflammatory signals. The underlying crosstalk at the epigenomic and transcriptional levels remained largely unexplored. One of the effector functions of macrophages is the integration of different danger signals with NLRP3 inflammasome activation (Rathinam and Fitzgerald, 2016Rathinam V.A. Fitzgerald K.A. Inflammasome complexes: emerging mechanisms and effector functions.Cell. 2016; 165: 792-800Abstract Full Text Full Text PDF PubMed Scopus (555) Google Scholar). Inflammasomes play key roles in the generation of secreted forms of proinflammatory IL-1β and IL-18 from their precursors. In parallel, macrophages undergo active NLRP3 inflammasome-dependent cell death termed “pyroptosis” (Rathinam and Fitzgerald, 2016Rathinam V.A. Fitzgerald K.A. Inflammasome complexes: emerging mechanisms and effector functions.Cell. 2016; 165: 792-800Abstract Full Text Full Text PDF PubMed Scopus (555) Google Scholar). The integration of this process to inflammatory epigenomic signaling is also not known. We sought to address these questions regarding the integration and regulation of the alternatively polarized macrophage phenotype by carrying out systemic genome-wide studies. We determined the STAT6-dependent IL-4-regulated genes in a time course in wild-type (WT) and Stat6−/− bone marrow-derived macrophages (BMDMs) using RNA-seq (Figure S1A). First, we examined the gene expression pattern of the 1,614 IL-4-regulated genes (Fc ≥ 2, p value < 0.05) and identified four IL-4-induced gene expression clusters based on expression dynamics and fold induction (Figures 1A and S1B; Table S1). We also found that a high proportion of IL-4-responsive genes (39%) were repressed. Repression by IL-4 was observed after 3 hr and remained attenuated at later time points (6, 24 hr) (Figure 1A, cluster E; Table S1). IL-4-mediated repression is dependent on STAT6 (Figure 1B). For validation, we measured the mRNA level of six IL-4-repressed (Abca1, Clec4d, Fos, Tlr2, Cd14, and Nlrp3) and three activated (Klf4, Hbegf, and Edn1) genes with RT-qPCR, and we confirmed the IL-4-mediated and STAT6-dependent regulation (Figures 1C and S1C). Filarial nematode infection is associated with the accumulation of alternatively polarized macrophages, exhibiting elevated expression of Ym1 and Fizz1/RELM-α (Anthony et al., 2006Anthony R.M. Urban Jr., J.F. Alem F. Hamed H.A. Rozo C.T. Boucher J.L. Van Rooijen N. Gause W.C. Memory T(H)2 cells induce alternatively activated macrophages to mediate protection against nematode parasites.Nat. Med. 2006; 12: 955-960Crossref PubMed Scopus (406) Google Scholar). In order to determine whether transcriptional repression in response to alternative polarization signals occurs in vivo, we compared the gene expression profile of peritoneal macrophages from Brugia malayi nematode-implanted mice (Ne-Mac) and thioglycollate-elicited peritoneal macrophages (Thio-Mac) utilizing publicly available RNA-seq data (Thomas et al., 2012Thomas G.D. Rückerl D. Maskrey B.H. Whitfield P.D. Blaxter M.L. Allen J.E. The biology of nematode- and IL4Rα-dependent murine macrophage polarization in vivo as defined by RNA-Seq and targeted lipidomics.Blood. 2012; 120: e93-e104Crossref PubMed Scopus (44) Google Scholar). Gene set enrichment analysis (GSEA) showed that the in vitro IL-4-repressed gene set was significantly enriched (FDR q-value < 0.1, NER: −2.38) among the genes that were downregulated in response to nematode infection in peritoneal macrophages (Figure 1D). In addition, all selected IL-4-STAT6-repressed genes were significantly downregulated during Brugia malayi-induced in vivo alternative macrophage polarization compared to thioglycollate-elicited peritoneal macrophages (Figure 1E). Next, we determined whether IL-4-STAT6 signaling represses gene expression at the transcriptional or post-transcriptional level. We assessed the immediate early effect of IL-4 on two serine phosphorylated forms of RNA polymerase II (RNAPII), the active histone mark H3K27Ac using chromatin immunoprecipitation sequencing (ChIP-seq), and nascent RNA expression by Global Run-On sequencing (GRO-seq) after 1 hr of exposure. Elongation-specific RNAPII-pS2 ChIP-seq revealed 5,931 gene bodies, exhibiting significantly changing read enrichments (3,008 downregulated and 2,923 upregulated, p ≤ 0.1) (Figure S2A and Table S2). RNAPII-pS2 binding showed positive correlation with transcription initiation-specific RNAPII-pS5 binding, H3K27Ac enrichment, and nascent RNA expression at the gene bodies (Figure S2B). Importantly, the gene bodies of IL-4-repressed genes (cluster E) showed attenuated RNAPII-pS2, RNAPII-pS5, and H3K27Ac enrichment and nascent RNA expression (Figures 2A, 2B, and S2C). In contrast, IL-4-dependent induction of RNAPII-pS2, RNAPII-pS5, and H3K27Ac enrichment as well as nascent RNA expression was detected at gene bodies of IL-4-induced genes (clusters A–C) (Figures 2A, S2C, and S2D). These results indicate that IL-4-STAT6 signaling directly represses gene expression, primarily at the transcriptional level during alternative macrophage polarization in vitro and in vivo. We also determined the STAT6 cistrome using a time course of 1, 6, and 24 hr of IL-4 stimulation (Figure S1A). STAT6 binding was negligible in unstimulated BMDMs (Figure 3A), but as little as 1 hr of stimulation dramatically induced the binding of STAT6, which was followed by a decline after 24 hr (Figure 3A). Comparing the STAT6 cistrome (20,119 genomic regions in IL-4-stimulated cells) to the RNAPII-pS5-positive genomic regions revealed that 60.5% of STAT6 peaks overlapped with the union of those genomic regions bound by RNAPII-pS5 in resting or IL-4-treated BMDMs (Figure 3B), suggesting that transcription could be directly regulated by STAT6 at these sites. Therefore, we next classified the RNAPII-pS5-positive STAT6 peaks based on IL-4-dependent regulation of RNAPII-pS5 binding, and we divided the STAT6-bound genomic regions into three different clusters: “repressor,” “neutral,” and “activator” STAT6 peak clusters (Figure 3C and Table S3). We noted that repressor and neutral STAT6 peaks showed typically lower occupancies if compared to the IL-4-induced RNAPII-pS5-associated activator STAT6 peaks (Figure S3A). Interestingly, IL-4-dependent regulation of RNAPII-pS2 binding as well as H3K27Ac enrichments showed similar patterns to RNAPII-pS5 in all three STAT6 clusters (Figures 3C and 3D). These findings support the conclusion that IL-4-activated STAT6 can be associated with either transcriptional activation or repression at different genomic loci. Next we assigned STAT6-bound genomic regions to genes in order to assess the correlation between IL-4-repressed enhancer activity (RNAPII-pS5 by ChIP-seq) and gene expression (mRNA by RNA-seq). For this analysis, we predicted the sub-topologically associated domains (subTADs) in which gene regulation by STAT6 might take place, using CTCF and RAD21 ChIP-seq datasets from BMDM, utilizing a previously described algorithm (Daniel et al., 2014Daniel B. Nagy G. Hah N. Horvath A. Czimmerer Z. Poliska S. Gyuris T. Keirsse J. Gysemans C. Van Ginderachter J.A. et al.The active enhancer network operated by liganded RXR supports angiogenic activity in macrophages.Genes Dev. 2014; 28: 1562-1577Crossref PubMed Scopus (67) Google Scholar, Rao et al., 2014Rao S.S. Huntley M.H. Durand N.C. Stamenova E.K. Bochkov I.D. Robinson J.T. Sanborn A.L. Machol I. Omer A.D. Lander E.S. Aiden E.L. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping.Cell. 2014; 159: 1665-1680Abstract Full Text Full Text PDF PubMed Scopus (3630) Google Scholar). As shown in Figure 3E, we found that repressor STAT6 peaks were tightly associated with the IL-4-repressed gene cluster (cluster E). In contrast, activator STAT6 peaks were associated with the immediate early IL-4-induced genes represented by clusters A–C (Figure 3E). These results suggest a tight connection between STAT6-dependent regulation of enhancer activity and neighboring gene expression in the same genomic compartment or transcription unit. To understand the IL-4-STAT6 signaling-mediated transcriptional regulation in more detail, we carried out analyses on individual genes and enhancers. For the selected repressed and activated genes, we identified at least one STAT6-bound enhancer showing reduced and induced H3K27 acetylation and RNAPII binding, respectively (Figure 3F). Enhancer RNA (eRNA) expression is a reliable marker of enhancer activity (Natoli and Andrau, 2012Natoli G. Andrau J.C. Noncoding transcription at enhancers: general principles and functional models.Annu. Rev. Genet. 2012; 46: 1-19Crossref PubMed Scopus (289) Google Scholar). Therefore, we measured eRNA expression at the repressor and activator STAT6 peaks by RT-qPCR. The expression of eRNAs were regulated in a similar manner as the enrichment of RNAPII-pS5 and RNAPII-pS2 and changes of H3K27Ac levels at the repressor and activator STAT6 sites in WT BMDMs (Figures 3F, 3G, and S3B). Importantly, IL-4-mediated regulation of eRNA expression was abolished in the absence of STAT6 at the examined enhancers (Figures 3G and S3B). Taken together, these results show that IL-4-activated STAT6 is required for the transcriptional repression characterized by decreasing RNAPII binding, histone acetylation, and consequently enhancer activity. In order to investigate whether the functional characteristics of STAT6 peaks (activator versus repressor) are influenced by their genomic localization and/or the DNA sequences they are associated with, we analyzed the genomic distribution of STAT6 peak clusters. We found only minor differences between the distinct STAT6 peak clusters regarding genomic localization relative to genes (Table S4). The majority of STAT6 peaks were localized in intergenic and intronic regions in the genome in all three clusters, and only about 10% of STAT6 binding sites were detected in promoter-proximal regions (Table S4). We also examined the enrichment of active histone mark H3K4m1 at the STAT6-bound genomic regions using a publicly available ChIP-seq dataset (Ostuni et al., 2013Ostuni R. Piccolo V. Barozzi I. Polletti S. Termanini A. Bonifacio S. Curina A. Prosperini E. Ghisletti S. Natoli G. Latent enhancers activated by stimulation in differentiated cells.Cell. 2013; 152: 157-171Abstract Full Text Full Text PDF PubMed Scopus (538) Google Scholar). Although H3K4m1 enrichment was observed at more than 98% of STAT6-bound genomic regions, it was not influenced by IL-4 treatment (Table S4, Figure S3C). These findings suggest that STAT6 primarily binds enhancers and that the functional characteristics of different STAT6 peak clusters cannot be explained by their genomic localization relative to genes. Next, we carried out de novo motif enrichment analysis of the sequences under the STAT6 peaks. PU.1, TRE, RUNX, and C/EBP motifs were enriched under all three clusters (Figure S3D). However, the canonical STAT6 motif was significantly under-represented under repressor and neutral STAT6 peaks if compared to the activator STAT6 peaks (Figures 3H and S3D). Plotting the motif scores for PU.1, TRE, RUNX, and C/EBP revealed no significant differences between the different STAT6 peak clusters (Figure S3E). In contrast, motif score for STAT6 was lower in the repressor and neutral STAT6 peak clusters compared to the activator STAT6 peak cluster (Figure 3I). Considering that the presence of STAT6 is needed for repression (Figure 1B), these findings raise the possibilities that STAT6 is bound without direct DNA contact or that it recognizes non-canonical STAT6-binding motifs at repressed enhancers. We investigated whether chromatin accessibility is affected at the STAT6-bound genomic regions by performing Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) experiments. Our genome-wide analyses showed increased chromatin accessibility at the activator STAT6-bound sites (Figure 4A), while significant reduction was detected in chromatin accessibility at the repressor STAT6-bound genomic regions (Figure 4A). These results suggest that both STAT6-mediated enhancer activation and repression are associated with the modification of chromatin structure during alternative macrophage polarization. Chromatin openness determines enhancer activity in different cell types (Shlyueva et al., 2014Shlyueva D. Stampfel G. Stark A. Transcriptional enhancers: from properties to genome-wide predictions.Nat. Rev. Genet. 2014; 15: 272-286Crossref PubMed Scopus (808) Google Scholar). Moreover, binding of macrophage LDTFs, PU.1, JUNB, IRF8, and C/EBPα are associated with active enhancers in macrophages (Glass and Natoli, 2016Glass C.K. Natoli G. Molecular control of activation and priming in macrophages.Nat. Immunol. 2016; 17: 26-33Crossref PubMed Scopus (315) Google Scholar). In addition, their binding motifs were among the most enriched transcription factor motifs under STAT6 peaks (Figure S3D). Therefore, we decided to determine whether IL-4-STAT6 signaling-mediated repression is associated with modified binding of LDTFs and examined their binding at repressed enhancers in the presence or absence of IL-4 using ChIP-seq. A high portion of the STAT6 cistrome overlapped with the examined LDTF cistromes except for JUNB, which showed moderated overlap (Table S4). Intriguingly, PU.1, JUNB, and C/EBPα binding was significantly decreased, while IRF8 binding was not modulated at the repressed STAT6-bound genomic regions after 1 hr IL-4 treatment in BMDMs (Figures 4B and S4A). In contrast, all four LDTFs showed significantly elevated binding at the IL-4-activated enhancers following IL-4 stimulation (Figures 4B and S4A). These findings suggest that IL-4-STAT6 signaling pathway modulates the binding of LDTFs at STAT6-activated and -repressed enhancers to opposite directions. The acetylation status and thus the activity of enhancers are tightly controlled by histone acetyltransferase (HAT) and histone deacetylase (HDAC) enzymes (Calo and Wysocka, 2013Calo E. Wysocka J. Modification of enhancer chromatin: what, how, and why?.Mol. Cell. 2013; 49: 825-837Abstract Full Text Full Text PDF PubMed Scopus (887) Google Scholar). Therefore, we examined the binding of the histone acetyltransferase p300 as well as classical histone deacetylases, including HDAC1, 2, and 3 at the STAT6-bound genomic regions after 1 hr of IL-4 exposure by ChIP-seq. We found that the majority of STAT6-bound genomic regions were either pre-loaded by p300 and classical HDACs or recruited these factors upon IL-4 stimulation (Table S4). The binding of p300 was significantly increased at STAT6-activated enhancers, but significantly reduced at STAT6-repressed enhancers upon IL-4 treatment (Figure 4C). Interestingly, genome-wide analysis of IL-4-modulated HDAC binding showed significantly enhanced HDAC1, 2, and 3 occupancy at STAT6-activated enhancers, while STAT6-repressed enhancers showed no effect to IL-4, but exhibited HDAC binding at the basal state (Figure 4D). Collectively, these results show that STAT6-repressed enhancers are bound by both HATs and HDACs at the steady state and that p300 binding is selectively reduced by IL-4, resulting in a changed equilibrium favoring HDAC activity. Direct interactions between classical HDACs and STAT transcription factors have been observed previously in numerous cell types influencing STAT-mediated direct transcriptional regulation (Icardi et al., 2012Icardi L. Mori R. Gesellchen V. Eyckerman S. De Cauwer L. Verhelst J. Vercauteren K. Saelens X. Meuleman P. Leroux-Roels G. et al.The Sin3a repressor complex is a master regulator of STAT transcriptional activity.Proc. Natl. Acad. Sci. USA. 2012; 109: 12058-12063Crossref PubMed Scopus (63) Google Scholar, Nusinzon and Horvath, 2003Nusinzon I. Horvath C.M. Interferon-stimulated transcription and innate antiviral immunity require deacetylase activity and histone deacetylase 1.Proc. Natl. Acad. Sci. USA. 2003; 100: 14742-14747Crossref PubMed Scopus (232) Google Scholar). In addition, HDAC3 has been shown to participate in the regulation of alternative macrophage polarization in vitro and in vivo (Mullican et al., 2011Mullican S.E. Gaddis C.A. Alenghat T. Nair M.G. Giacomin P.R. Everett L.J. Feng D. Steger D.J. Schug J. Artis D. Lazar M.A. Histone deacetylase 3 is an epigenomic brake in macrophage alternative activation.Genes Dev. 2011; 25: 2480-2488Crossref PubMed Scopus (213) Google Scholar). Thus, we hypothesized that HDAC3, which is present at repressed enhancers (Figure 4D), might also contribute to IL-4-STAT6-induced repression. Therefore, we decided to examine the role of HDAC3 using a dataset from Mullican et al., 2011Mullican S.E. Gaddis C.A. Alenghat T. Nair M.G. Giacomin P.R. Everett L.J. Feng D. Steger D.J. Schug J. Artis D. Lazar M.A. Histone deacetylase 3 is an epigenomic brake in macrophage alternative activation.Genes Dev. 2011; 25: 2480-2488Crossref PubMed Scopus (213) Google Scholar. Applying K-mean clustering method, we found 1,628 IL-4-repressed genes (p ≤ 0.05) in WT BMDMs (Figure S4B) and identified an IL-4-repressed gene cluster (cluster III, 371 genes) that showed attenuated repression in Hdac3fl/fl Lyz2 Cre BMDMs following IL-4 treatment (Figures 4E and S4B). Although the basal expression of these genes did not show major differences between WT and Hdac3fl/fl Lyz2 Cre BMDMs, the IL-4-induced repression was partially or completely abolished in the absence of HDAC3 (Figures 4E and S4B). In addition, enrichments of RNAPII-pS5 and RNAPII-S2 were reduced at these gene bodies after 1 hr of IL-4 treatment in WT BMDMs (Figure 4F). Interestingly, 325 STAT6-repressed enhancers were found within the subTADs of IL-4-HDAC3-repressed genes (Figure 4G). These enhancers were bound by HDAC3, but HDAC3 occupancy was not altered by IL-4 stimulation (Figures 4G and 4H). Our results indicate that HDAC3 is required for the IL-4-induced repression of a specific subset of genes. Due to the fact that HDAC3 is one of the key components of NCoR and SMRT corepressor complexes (Karagianni and Wong, 2007Karagianni P. Wong J. HDAC3: taking the SMRT-N-CoRrect road to repression.Oncogene. 2007; 26: 5439-5449Crossref PubMed Scopus (170) Google Scholar), we decided to determine whether the NCoR-SMRT complex itself participates in IL-4-STAT6-HDAC3-mediated repression as well. First, we determined the occupancy of NCoR and SMRT at HDAC3-bound enhancers using ChIP-seq data generated by others (Barish et al., 2012Barish G.D. Yu R.T. Karunasiri M.S. Becerra D. Kim J. Tseng T.W. Tai L.J. Leblanc M. Diehl C. Cerchietti L. et al.The Bcl6-SMRT/NCoR cistrome represses inflammation to attenuate atherosclerosis.Cell Metab. 2012; 15: 554-562Abstract Full Text Full Text PDF PubMed Scopus (96) Google Scholar). We found that the IL-4-STAT6-HDAC3-repressed enhancer set was bound by both NCoR and SMRT in unstimulated BMDMs (Figure 4G). Next, we investigated the requirement of NCoR in the IL-4-STAT6-HDAC3-mediated repression using Ncor1fl/fl Lyz2 Cre immortalized bone marrow-derived macrophages (iBMDMs). We selected four genes for this analysis (Fos, Lyz1, Lyz2, and Smad3) based on their IL-4-STAT6-HDAC3-dependent repression (Figures 4I, S4C, and S4D) and due to the fact that their enhancers were bound by HDAC3, NCoR, and SMRT (Figure S4C). Gene expression analysis showed that Fos and Lyz1 were expressed at a significantly higher level in unstimulated iBMDMs in the absence of NCoR compared to WT iBMDMs, while the basal expression of Lyz2 and Smad3 were not affected by NCoR (Figure 4J). In addition, IL-4-mediated repression of these genes was diminished in Ncor1fl/fl Lyz2 Cre iBMDMs (Figures 4J and S4E). In contrast, the basal expression and IL-4-induced repression of HDAC3-independent genes were not affected by NCoR, except for Abca1 (Figure S4F). Taken together, our findings suggest that IL-4-activated STAT6 mediates transcriptional repression via the NCoR-HDAC3 complex at a subset of genes representing one of the molecular mechanisms for STAT6-dependent transcriptional repression. Next we were wondering whether the repressive action of IL-4-STAT6 leaves its footprint on the epigenome and affects the subsequent response of the cells to other stimuli. Using KEGG pathway analysis, we identified 12 signaling pathways whose overrepresentation was specific to IL-4-repressed genes (Figure S5A). NOD-like receptor signaling and Toll-like receptor signaling among the top hits, which are known to be two major regulators of the inflammatory response (Figure S5A; Chen et al., 2009Chen G. Shaw M.H. Kim Y.G. Nuñez G. NOD-like receptors: role in innate immunity and inflammatory disease.Annu. Rev. Pathol. 2009; 4: 365-398Crossref PubMed Scopus (553) Google Scholar, Takeda et al., 2003Takeda K. Kaisho T. Akira S. Toll-like receptors.Annu. Rev. Immunol. 2003; 21: 335-376Crossref PubMed Scopus (4744) Google Scholar). In addition, upstream transcriptional regulator analysis with Ingenuity Pathway Analysis (IPA) software revealed that the LPS-activated p65 (RelA) is one of the most significantly inhibited transcriptional regulators upon IL-4 stimulation (Figure S5B). Interestingly, the majority of IL-4-STAT6-repressed genes included several members of NOD-like and Toll-like receptor signaling pathways showing attenuated mRNA expression following 24 hr of IL-4 stimulation and reduced STAT6 binding at the repressed enhancers (Figures 1A, 1B, S5C, and S5D). These results raised the possibility that IL-4 is able to modulate the subsequent inflammatory response of the macrophage epigenome via directly repressed enhancers following the dissociation of STAT6. In order to determine whether prior activation of IL-4-STAT6 signaling is able to influence the inflammatory program of macrophages, we performed RNA-seq as well as RNAPII-pS5-, RNAPII-pS2-, and p65-specific ChIP-seq experiments on I