Genome-Wide Analysis Reveals Unique Regulation of Transcription of Th2-Specific by GATA3

This information is current as Shu Horiuchi, Atsushi Onodera, Hiroyuki Hosokawa, of September 27, 2021. Yukiko Watanabe, Tomoaki Tanaka, Sumio Sugano, Yutaka Suzuki and Toshinori Nakayama J Immunol 2011; 186:6378-6389; Prepublished online 2 May 2011;

doi: 10.4049/jimmunol.1100179 Downloaded from http://www.jimmunol.org/content/186/11/6378

Supplementary http://www.jimmunol.org/content/suppl/2011/05/02/jimmunol.110017 Material 9.DC1 http://www.jimmunol.org/ References This article cites 54 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/186/11/6378.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 27, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Genome-Wide Analysis Reveals Unique Regulation of Transcription of Th2-Specific Genes by GATA3

Shu Horiuchi,*,1 Atsushi Onodera,*,1 Hiroyuki Hosokawa,* Yukiko Watanabe,* Tomoaki Tanaka,† Sumio Sugano,‡ Yutaka Suzuki,‡ and Toshinori Nakayama*

Differentiation of naive CD4 T cells into Th2 cells is accompanied by chromatin remodeling and increased expression of a set of Th2- specific genes, including those encoding Th2 cytokines. IL-4–mediated STAT6 activation induces high levels of transcription of GATA3, a master regulator of Th2 cell differentiation, and enforced expression of GATA3 induces Th2 cytokine expression. However, it remains unclear whether the expression of other Th2-specific genes is induced directly by GATA3. A genome-wide unbiased chromatin immunoprecipitation assay coupled with massive parallel sequencing analysis revealed that GATA3 bound to 1279 genes selectively in Th2 cells, and 101 genes in both Th1 and Th2 cells. Simultaneously, we identified 26 highly Th2-specific

STAT6-dependent inducible genes by DNA microarray analysis-based three-step selection processes, and among them 17 genes Downloaded from showed GATA3 binding. We assessed dependency on GATA3 for the transcription of these 26 Th2-specific genes, and 10 genes showed increased transcription in a GATA3-dependent manner, whereas 16 genes showed no significant responses. The transcrip- tion of the 16 GATA3-nonresponding genes was clearly increased by the introduction of an active form of STAT6, STAT6VT. Therefore, although GATA3 has been recognized as a master regulator of Th2 cell differentiation, many Th2-specific genes are not regulated by GATA3 itself, but in collaboration with STAT6. The Journal of Immunology, 2011, 186: 6378–6389. http://www.jimmunol.org/ ffector CD4 Th subsets play an important role in various transcription factor is upregulated, which in turn leads to the ex- types of infection, and the best characterized are Th1, Th2, pression of subset-specific cytokine genes. Furthermore, a more E and Th17 cells (1–3). IL-12–induced STAT4 activation or fundamental regulation of expression during Th cell differ- IL-4–induced STAT6 activation is required for the differentiation entiation, so-called chromatin remodeling, establishes the stable of Th1 or Th2 cells, respectively. Master transcription factors that expression of the subset-specific cytokine genes (9). regulate Th1/Th2/Th17 cell differentiation have been identified. GATA3 is predominantly expressed in T lymphocytes and the T cell-specific T-box transcription factor (T-bet) appears to be embryonic brain (10). GATA3 is required for T cell development a key factor for Th1 cell differentiation (4), GATA3 for Th2 cells in the thymus; its roles in the CD4 versus CD8 lineage choice and (5, 6), and retinoid-related orphan receptors gt and a for Th17 at the b-selection checkpoint are the best characterized (11). In by guest on September 27, 2021 cells (7, 8). During the differentiation of these subsets, the ex- peripheral CD4 T cells, the activation of STAT6 induces high- pression of a certain set of genes including a specific master level GATA3 mRNA expression (12), and we and others have recently reported that STAT6 binds to specific regions of the GATA3 gene to regulate its transcription (13, 14). In addition, the *Department of Immunology, Graduate School of Medicine, Chiba University, Chiba expression of GATA3 in Th2 cells is regulated by various † 260-8670, Japan; Department of Clinical Cell Biology, Graduate School of Medi- posttranscriptional mechanisms (15–17). Changes in histone modi- cine, Chiba University, Chiba 260-8670, Japan; and ‡Laboratory of Functional Genomics, Department of Medical Genome Sciences, Graduate School of Frontier fications, such as H3-K9/14 acetylation and H3-K4 methylation at Sciences, University of Tokyo, Chiba 277-8562, Japan the Th2 cytokine gene loci, occur during Th2 cell differentiation 1S.H. and A.O. contributed equally to this study. (18–20), and this is induced primarily by GATA3 in both CD4 and Received for publication January 24, 2011. Accepted for publication March 24, 2011. CD8 T cells (21). High-level expression of GATA3 is required for This work was supported by the Global COE Program (Global Center for Education producing large amounts of Th2 cytokines in established Th2 cells and Research in Immune System Regulation and Treatment); the Ministry of Health, (22–25). Labor, and Welfare (Japan); the Uehara Memorial Foundation; and the following STAT6 appears to exert highly specific functions because the grants from the Ministry of Education, Culture, Sports, Science, and Technology, Japan: Grants-in-Aid for Scientific Research on Priority Areas (17016010), Scientific phenotype of STAT6-deficient mice is largely confined to Th2 cell Research (B) Grant 21390147, Start-up Grant 19890041, and Start-up Grant defects, and the mice develop otherwise normally with normal 2056131. numbers of T cells (26, 27). Transcriptional profiling by DNA The sequences presented in this article have been submitted to the Gene Expres- microarray analysis also helped clarify the genes whose expres- sion Omnibus database (http://www.ncbi.nlm.nih.gov/geo) under accession num- ber GSE28292. sion is controlled by STAT6, and the results argue that STAT6 is Address correspondence and reprint requests to Dr. Toshinori Nakayama, Department clearly important for regulating Th2 status (28, 29). However, of Immunology, Graduate School of Medicine, Chiba University, 1-8-1 Inohana, these studies do not address whether these genes are directly Chuo-ku, Chiba 260-8670, Japan. E-mail address: [email protected] regulated by STAT6 or not. The recent development of chromatin The online version of this article contains supplemental material. immunoprecipitation (ChIP) assay coupled with massive parallel Abbreviations used in this article: ChIP, chromatin immunoprecipitation; ChIP-chip, sequencing (ChIP-Seq) has allowed the analysis of transcription ChIP with microarray technology; ChIP-Seq, ChIP assay coupled with massive par- allel sequencing; HPRT, hypoxanthine phosphoribosyltransferase; IP, immunoprecip- factor binding sites and the status of histone modifications on itation; NGFR, nerve growth factor receptor; qPCR, quantitative PCR; qRT-PCR, a genome-wide scale (30). Recently, STAT6-binding genes in Th2 quantitative RT-PCR; RS, recovery score; TES, transcriptional end site; TSS, tran- cells were identified in mouse and human systems (13, 31). scriptional start site; WT, wild-type. To study how Th2-specific genes are induced by GATA3 and/or Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 STAT6 in developing Th2 cells, we determined GATA3 recruitment www.jimmunol.org/cgi/doi/10.4049/jimmunol.1100179 The Journal of Immunology 6379 and histone modifications by ChIP-Seq analysis in Th1 and Th2 Ig ChIP]/input DNA). All the enrichments are expressed as a function of cells. Simultaneously, 26 highly Th2-specific inducible genes were the highest enrichment obtained on the locus (set to 10) (33). identified by a DNA microarray-based three-step selection process. ChIP-Seq and Illumina sequencing In addition, we examined enforced expression of GATA3 and constitutively active STAT6 (STAT6VT) in STAT6-deficient Th2 For ChIP-Seq analysis, IP and input samples were prepared using ChIP- Seq Sample Prep kit (Illumina). Adaptor-ligated DNA fragments were size cells to determine direct effects of GATA3 and STAT6 on Th2- fractionated by 12% acrylamide gel, and the 170- to 250-bp fraction was specific gene expression. Among the Th2-specific genes, tran- recovered. DNA thus obtained was amplified by 18 cycles of PCR. One scription of nine genes was directly regulated by GATA3 in the nanogram of DNA was used for the sequencing reaction of the Illumina absence of STAT6. However, other Th2-specific genes were reg- GAIIx, according to the manufacturer’s instructions. A total of 170,000– 250,000 clusters was generated per tile, and 36 cycles of the sequencing ulated not by GATA3 itself, but in collaboration with STAT6. reactions were performed. Short-read sequences were aligned to the mouse genome sequences (mm9 from University of California, Santa Cruz Ge- Materials and Methods nome Browser; http://genome.ucsc.edu/) using the Eland program. Se- Mice quences allowing no more than two mismatches per sequence were used for the analysis. To enumerate GATA3-bound genes, at least one peak C57BL/6 mice were purchased from CLEA (Tokyo, Japan). (with a 5-fold increase in signal intensity compared with input DNA) STAT6-deficient mice (27) were provided by S. Akira (Osaka Univer- detected on the gene locus was selected. MEME (34) was used for per- sity, Osaka, Japan). All mice used in this study were maintained under forming a de novo search of consensus-binding motifs for GATA3. specific pathogen-free conditions and ranged from 6 to 8 wk of age. All animal care was performed in accordance with the guidelines of Chiba Retroviral vectors and infection University. The pMX-IRES-human nerve growth factor receptor (NGFR) plasmid was Downloaded from Reagents generated, as previously described (16). The infected cells were enriched Recombinant mouse IL-12 was purchased from BD Pharmingen, and recombinant mouse IL-4 was from TOYOBO (Osaka, Japan).

Abs

The Abs used for the ChIP assay were anti-GATA3 mAb (Santa Cruz http://www.jimmunol.org/ Biotechnology), anti-acetylhistone H3-K9 (Upstate; 06-599), anti-trimethyl- histone H3-K4 (LP Bio; AR-0169), and anti-trimethylhistone H3-K27 (Upstate; 07-449). The Abs used for cytoplasmic staining were anti–IFN-g FITC (XMG1.2) and anti–IL-4 PE (11B11). The generation of effector Th1 and Th2 cells Effector Th1/Th2 cells were generated, as previously described (32). Splenic CD4 T cells were prepared using a magnetic cell sorter (Auto- MACS; Miltenyi Biotec) yielding a purity of .98%. Where indicated, cells from C57BL/6 mice were stimulated with immobilized anti-TCR by guest on September 27, 2021 mAb (H57-597; 3 mg/ml) and anti-CD28 mAb under Th1 or Th2 culture conditions for 5 d in vitro. Th1 conditions were as follows: 25 U/ml IL-2, 10 U/ml IL-12, and anti–IL-4 mAb. Th2 conditions were as follows: 25 U/ ml IL-2 and 100 U/ml IL-4. These cells were used as either effector Th1 or Th2 cells, respectively. Microarray data collection and analysis Total cellular RNA from cells cultured under Th1 and Th2 conditions was extracted with TRIzol reagent (Invitrogen), according to the manufacturer’s instructions. Approximately 5 mg RNA was labeled and hybridized to GeneChip Mouse Genome 430 2.0 arrays (Affymetrix), according to the manufacturer’s protocol. Expression values were determined with Gene- Chip Operating Software v1.2 software (Affymetrix). Quantitative RT-PCR Total RNA was isolated using the TRIzol reagent (Invitrogen). cDNA was synthesized using oligo(dT) primer and Superscript II RT (Invitrogen). Quantitative RT-PCR was performed, as described previously, using an ABI PRISM 7500 Sequence Detection System (32). The primers and TaqMan probes for the detection of the indicated genes and hypoxanthine phos- phoribosyltransferase (HPRT) were purchased from Applied Biosystems and Roche, respectively. The expression was normalized by the HPRT signal. For quantitative RT-PCR (qRT-PCR) of cytokine genes, mRNA from Th1 or Th2 cells that were restimulated by anti-TCR mAb (H57-597; 3mg/ml) for 4 h was used. ChIP assay

ChIP was performed using ChIP assay kits (Upstate Biotechnology), as FIGURE 1. Schematic view of ChIP-Seq experiments and screening previously described (32). Real-time quantitative PCR (qPCR) analysis procedure of Th2-specific STAT6-dependent inducible genes. Freshly was performed on an ABI Prism 7500 real-time PCR machine with Taq- isolated CD4 T cells from C57BL/6 WT mice were stimulated with im- Man probes and primers. To calculate the enrichment of each protein to a particular target DNA, values obtained (via the standard curve method) mobilized anti-TCR mAb and anti-CD28 mAb under Th1 or Th2 culture for each target were divided by the amount of the corresponding target in conditions for 5 d in vitro. These cells were used for genome-wide GATA3 the input fraction. Enrichments obtained from mock immunoprecipitations binding and histone modification (H3-K4Me3, H3-K9Ac, and H3- (IPs) performed in parallel with normal IgG were then subtracted from the K27Me3) analysis by ChIP-Seq and for gene expression analysis. To enrichment values obtained with specific Abs ([specific Ab ChIP 2 control identify Th2-specific genes, a three-step selection process was adopted. 6380 ROLE OF GATA3 IN Th2-SPECIFIC GENE INDUCTION by MACS with anti-human NGFR (clone C40-1457; BD Pharmingen), and Using these criteria, we identified 250 genes selectively bound by were subjected to a qRT-PCR assay. GATA3 or STAT6 dependency was GATA3 in Th1 cells and 1,279 genes in Th2 cells. GATA3 binding evaluated by the recovery score (RS), which is defined as the linear was detected in both Th1 and Th2 cells at 101 gene loci (Fig. 2A). equation below: We next compiled the localization of H3-K9Ac, H3-K4Me3, and RSð%Þ¼ 100 3 ðC 2 BÞ=ðA 2 BÞ; H3-K27Me3 marks for the GATA3-bound genes in relation to each where A indicates the signal intensity of Th2, B indicates the signal in- gene TSS 6 10 kb to visualize the tag density profile for each tensity of STAT6KO Th2, and C indicates the signal intensity of STAT6KO mark (Fig. 2B). Histone H3-K9Ac and H3-K4Me3 are permissive Th2 with overexpression of GATA3 or STAT6VT. marks for chromatin, and H3-K27me3 is a mark for the closed Accession number configuration of chromatin (30). In both Th1 and Th2 cells, H3- K9Ac (in blue) and H3-K4Me3 (in red) were highly enriched Normalized microarray raw data are available in the Gene Expression Om- nibus database (http://www.ncbi.nlm.nih.gov/geo) under accession number around the TSS, as previously reported (30, 36). We also found GSE28292. that H3-K27Me3 marks (in green) were relatively depleted around the TSS of the GATA3-bound genes. Of note, the levels of H3- Results K9Ac and H3-K4Me3 marks were lower at GATA3-bound genes Genome-wide analysis of GATA3-binding genes and histone in Th1 cells, compared with those in Th2 cells. We also compiled modifications in Th1 and Th2 cells the localization of GATA3 peaks for all target genes in the context 6 To globally identify the GATA3-binding genes and their histone of the TSS 10 kb to again visualize the tag density profile, and modification status in Th1 and Th2 cells, we performed ChIP-Seq a marked enrichment for binding was observed around the TSS analysis using Abs specific for GATA3, histone H3K9 acetylation (Fig. 2B, lower panels). To determine whether epigenetic marks Downloaded from (H3-K9Ac), histone H3K4 trimethylation (H3-K4Me3), and his- are globally aligned with GATA3 occupancy, we plotted cumu- tone H3K27 trimethylation (H3-K27Me3). The ChIP DNA frag- lative sequence frequencies associated with GATA3-bound sites ments were sequenced with an Illumina-Solexa 1G Genome for each of the histone marks (Fig. 2C). The analysis revealed the Analyzer, as described (30, 35). A total of 5,400,000∼9,000,000 enrichment of H3-K9Ac (in blue) and H3-K4Me3 (in red) around sequence tags was generated for Th1 and Th2 cells and mapped the GATA3 peak, whereas H3-K27Me3 marks (in green) were not to the mouse genome (Fig. 1). We identified 510 GATA3-bound enriched. http://www.jimmunol.org/ peaks in Th1 and 2,258 GATA3-bound peaks in Th2 cells. To Previous in vitro studies have indicated that GATA family enumerate GATA3-bound genes, the genes having at least one bind to a WGATAR motif (37). To determine whether significant peak (.5-fold increase in signal intensity compared such an element was identifiable in our genome-wide analysis with input DNA) detected in the gene locus were selected (here- of GATA3 binding, we carried out de novo searches of consensus after defined as GATA3-bound genes). Each gene locus was de- motifs of GATA3-bound regions in Th1 and Th2 cells using fined as a region from 5 kb upstream of the transcriptional start MEME (34). In the GATA3 peaks detected in Th2 cells, we found site (TSS) to 5 kb downstream of transcriptional end site (TES). 2257 motifs that include a GATA motif (Fig. 2D). For the GATA3 by guest on September 27, 2021

FIGURE 2. Genome-wide distribution of GATA3 binding sites and identification of consensus-binding motifs. A, Venn diagram showing the number of genes bound by GATA3 uniquely in Th1, Th2, or genes bound in both Th1 and Th2 cells. GATA3-bound genes were identified if at least one peak (5-fold increase in signal intensity compared with input DNA) of binding was present within the gene locus. B, Compiled tag density profiles across the TSS 6 10- kb flanking regions with 100-bp resolution for H3-K9Ac, H3-K4Me3, and H3-K27Me3 and GATA3 binding at GATA3-bound genes in Th1 and Th2 cells are shown. C, Compiled tag density profiles across the GATA3 peak 6 10-kb flanking regions at GATA3-bound genes with 100-bp resolution for H3-K9Ac, H3-K4Me3, and H3-K27Me3 in Th1 and Th2 cells are shown. D, MEME (34) was used to perform a de novo search of consensus-binding motifs for GATA3 in Th2 cells. The Journal of Immunology 6381 peaks in Th1 cells and those overlapped in Th1 and Th2 cells, we with changes in histone modifications between Th1 and Th2 cells did not identify a consensus motif for GATA3 as a WGATAR in a set of genes, but also shows that in the majority of cases, (Supplemental Fig. 1A). We also identified the ETS-binding motif GATA3 binding and changes in histone modifications do not co- (GGAA) on GATA3 peaks in Th2 cells in the MEME analysis incide. (Supplemental Fig. 1B). Selection of Th2-specific genes Histone modification patterns in the GATA3-bound genes The genome-wide surveys on GATA3 binding and histone mod- To assess the correlation between direct binding of GATA3 and ifications indicate a possibility that GATA3 binding is not always histone modifications, we evaluated global histone modification associated with changes in histone modifications, which may patterns at the 1380 GATA3-binding gene loci using the results of correlate with the status of transcription. An emerging concept ChIP-Seq analyses. We quantified the total tag counts of H3-K9Ac, coming from genome-scale studies is that only a fraction of factor- H3-K4Me3, and H3-K27Me3 at the promoter region, which is binding genes showed clear functional dependence, as evaluated defined as a region from 5 kb upstream to 3 kb downstream relative by gene expression changes (13, 38). Therefore, we next sought to to the TSS in this analysis. The tag counts were normalized by the determine the extent to which the status of transcription correlates total number of input DNA tags at the corresponding region. We with both the binding of GATA3 and histone modifications at clustered GATA3-bound genes on the basis of the ratio of tag counts specific gene loci. To investigate the role of GATA3 during Th2 in Th1 and Th2 cells (Th2/Th1 ratio) for each of the epigenetic cell differentiation, we focused on Th2-specific genes in which the marks (Fig. 3A). From this analysis, four epigenetic patterns for level of transcription was upregulated during Th2 cell differenti-

GATA3-bound genes emerged. We identified GATA3-bound gene ation and was preferentially higher in Th2 cells as compared with Downloaded from clusters with a Th2/Th1 ratio .2 (in red) for H3K9Ac (denoted Th1 cells. To identify Th2-specific genes, we adopted three-step H3K9Ac high, 7.6%) and H3K4Me3 (H3K4Me3 high, 4.2%), and selection processes (Fig. 1). On the first step, we assessed gene ,0.5-fold Th2/Th1 ratio (in green) for H3K27Me3 (H3K27Me3 expression profiles in in vitro generated Th1 and Th2 cells by low, 11.6%) (Fig. 3B). For the largest cluster (63% of the total a DNA microarray analysis, and compared the results with those GATA3-bound genes), GATA3 binding had minimal effects (be- reported previously (28). A list of the potential Th2-specific genes

tween 1- and 2-fold Th2/Th1 ratio, in pink; between 0.5- and 1- (185 genes) was obtained. On the second step, the 185 selected http://www.jimmunol.org/ fold Th2/Th1 ratio, in white) on any epigenetic modifications we genes were validated using DD cycle threshold qRT-PCR with investigated (denoted as indeterminate). We also obtained similar mRNA from wild-type (WT) and STAT6-deficient Th2 cells results from the GATA3-bound genes in Th1 cells and those (Supplemental Fig. 2). mRNA levels of 92 of the 185 candidate detected in both Th1 and Th2 cells (Supplemental Fig. 1C,1D). genes were increased by at least 2-fold in WT Th2 cells compared These results indicate that GATA3 binding correlates very well with Th2 cells lacking STAT6. Ninety-two STAT6-dependent by guest on September 27, 2021

FIGURE 3. GATA3-bound genes in Th2 cells form clusters that share common epigenetic signatures. A, The total tag count of each epigenetic mod- ification (H3K9Ac, H3K4Me3, and H3K27Me3) was computed across the promoter of each GATA3-bound gene in Th2 versus Th1 cells. Red indicates .2- fold Th2/Th1 ratio; pink indicates 1- to 2-fold Th2/Th1 ratio; white indicates 0.5- to 1-fold Th2/Th1 ratio; and green indicates ,0.5-fold Th2/Th1 ratio. B, The ratio of tag counts for three epigenetic modifications was used to cluster GATA3-bound genes in four patterns, as follows: H3K9Ac-high, H3K4Me3- high, H3K27Me3-low, and an indeterminate pattern. 6382 ROLE OF GATA3 IN Th2-SPECIFIC GENE INDUCTION genes were identified in Th2 cells. Then qRT-PCR analysis was Tube1 gene loci had higher levels of H3-K9Ac by at least 1.5-fold performed using Th1, Th2, and STAT6-deficient Th2 cells (the in Th2 cells (Fig. 5A, top). Cell surface molecules tended to have third step). mRNA levels of 38 genes of the 92 genes were in- the lowest levels of H3-K9Ac and H3-K4Me3 when the whole creased by at least 4-fold in WT Th2 cells compared with Th1 or gene locus was analyzed (Fig. 5A). Overall, the results of H3- STAT6-deficient Th2 cells, whereas mRNA levels for 10 of these K9Ac and H3-K4Me3 levels at the whole gene locus were simi- 38 genes were not increased by .2-fold in Th2 cells compared lar to those of the promoter region, although the levels were with fresh CD4 T cells, and so were also excluded (data not slightly higher in the promoter region (Fig. 5). The levels of H3- shown). We also excluded two genes that were not described as K27Me3 at the Epas1, GATA3, IL-4, IL-5, IL-13, IL-24, CCR8, the RefSeq genes (mm9 from University of California, Santa Cruz Ecm1, TNFRSF8, Cyp11a1, DUSP4, and TANC2 gene loci were Genome Browser; http://genome.ucsc.edu/). As a result, 26 genes higher by at least 1.5-fold in Th1 cells as compared with Th2 cells. were selected as highly Th2-specific STAT6-dependent inducible These results indicate that increased permissive histone mod- genes. The preferential expression of these genes in Th2 cells is ifications tend to associate with Th2-specific gene expression. shown in comparison with Th1 cells and STAT6-deficient Th2 However, changes in gene expression are not always accompanied cells (Fig. 4A) or freshly prepared CD4 T cells (Fig. 4B). We by epigenetic changes in histone modifications, and this is par- focused on these 26 genes for further study. ticularly evident in the cell surface molecule group. H3-K9Ac, H3-K4Me3, and H3-K27Me3 modifications at the 26 Analysis of GATA3 binding at the 26 Th2-specific STAT6- Th2-specific STAT6-dependent inducible gene loci dependent inducible gene loci

The chromatin status at the 26 previously selected Th2-specific To assess the possible correlation between the direct binding of Downloaded from gene loci was determined using the results of the ChIP-Seq GATA3 and transcription of the Th2-specific genes, we next analysis on histone modifications in Th1 and Th2 cells. We searched GATA3 binding at the 26 Th2-specific STAT6-dependent quantified the total tag counts of H3-K9Ac, H3-K4Me3, and H3- inducible genes using the results of GATA3 ChIP-Seq analysis. Of K27Me3 at each gene locus (5 kb upstream of the TSS to 5 kb the 26, GATA3 binding was detected at 17 gene loci (Fig. 6A), downstream of TES) (Fig. 5A). The raw tag count patterns are also leaving 9 gene loci with no discernible GATA3 binding (data not

shown (Supplemental Fig. 3). The histone modification tag counts shown). Thirteen genes showed a single peak, and 4 genes showed http://www.jimmunol.org/ were also quantified at each gene promoter, which is defined as more than one peak at their gene loci. GATA3 binding was a region from 5 kb upstream to 3 kb downstream relative to the strongly detected at the region named conserved GATA3 response TSS (Fig. 5B). The tag counts were normalized by the total element, which is located 1.6 kb upstream of the IL-13 gene TSS number of input tags at the corresponding region. Epas1, GATA3, (39). The GATA3 binding site in the GATA3 gene locus was very NFIL3, GZMA, IL-4, IL-5, IL-13, IL-24, CCR8, Ecm1, IL1r2, close to the previously identified STAT6 binding site (13, 14). Itgb3, RNF128, TNFRSF8, TMTC2, Cyp11a1, S100a1, and These GATA3 binding sites detected by ChIP-Seq analysis were by guest on September 27, 2021

FIGURE 4. Th2-specific genes evaluated by qRT-PCR. A and B, The gene expression levels in Th1, Th2, and STAT6-deficient Th2 cells (A), or in freshly prepared CD4 T cells and Th2 cells (B) were determined by qRT-PCR. The relative intensity (relative expression/HPRT; highest signal intensity = 10) (mean of two samples with SD) is shown. The Journal of Immunology 6383 Downloaded from http://www.jimmunol.org/

FIGURE 5. H3-K9Ac, H3-K4Me3, and H3-K27Me3 modifications at Th2-specific gene loci. The levels of H3-K9Ac, H3-K4Me3, and H3-K27Me3 by guest on September 27, 2021 modifications at the Th2-specific genes in Th1 (n) and Th2 (N) cells. The numbers of histone modification tags (normalized by input DNA tags) at the gene locus (A) or promoter region (B) were calculated based on the results shown in Supplemental Fig. 3. then confirmed by conventional ChIP assay with qPCR (ChIP- GATA3, G1), NGFRlow (expressing low levels of retrovirus- qPCR) in Th1, Th2, and STAT6-deficient Th2 cells, and Th2- induced GATA3, G2), and NGFRhigh (expressing high levels of specific increased binding of GATA3 was detected at all of the retrovirus-induced GATA3, G3) (Fig. 7A). The mRNA levels of testable regions (Fig. 6B). the 26 Th2-specific genes including 17 GATA3-bound genes (Fig. 7B) and 9 GATA3-nonbound genes (Fig. 7C) were then assessed Preferential GATA3 binding and H3K9Ac modification in the in these sorted populations. The mRNA expression level of 26 Th2-specific STAT6-dependent inducible genes GATA3 in the NGFRhigh population of the STAT6-defiecient Next, we studied GATA3 binding and histone H3K9Ac in the 92 GATA3 retrovirus-infected group was similar to that of WT Th2 Th2-specific and STAT6-dependent genes shown in Supplemental cells (data not shown) (23). In GATA3 high-expressing cells (G3), Fig. 2A. We divided the 92 genes into two groups, as follows: the the mRNA levels of IL-13, Cyp11a1, Asb2, JDP2, IL-5, TMTC2, 26 inducible genes and the 66 other genes, and compared the IL-4, Ccnjl, IL-24, and PTGIR were greatly increased (.40% frequency of the genes with GATA3 binding (Supplemental Fig. recovery based on the RS, see Materials and Methods) (Fig. 7D), 4A) and increased levels of H3K9Ac (Th2/Th1 ratio; Supple- and this increase occurred in a gene dosage-dependent manner. In mental Fig. 4B). The percentages of GATA3-binding target genes contrast, the other 16 Th2-specific genes showed little or no re- and also genes that displayed high H3K9Ac (Th2/Th1 ratio .1.5) covery even in the fraction of GATA3 high-expressing cells (G3) were higher in the 26 STAT6-dependent inducible genes as (see GATA3-low genes in Fig. 7B–D). Next, a constitutively active compared with the other 66 genes (GATA3 target genes, 65.4 form of STAT6, STAT6VT, was introduced into STAT6-deficient versus 31.8%, and high H3K9Ac genes, 73.1 versus 30.0%). CD4 T cells cultured under Th2 conditions (Fig. 8). Among the 16 Th2-specific genes that showed little or no recovery even in Enforced expression of GATA3 induces transcription of the GATA3 high-expressing cells (G3), STAT6VT introduction a subset of Th2-specific genes in STAT6-deficient Th2 cells resulted in .40% recovery of mRNA expression in all (GATA3, Next, we assessed whether the enforced expression of GATA3 or Itgb3, GZMA, Epas1, Crem, NFIL3, CCR8, TNFRSF8, S100a1, STAT6 induces the transcription of the 26 Th2-specific genes using Ecm1, Dusp4, Tube1, IL1r2, RNF128, F2R, and TANC2) (see Fig. a retrovirus gene introduction system. To examine GATA3 de- 8B, GATA3-low genes). Furthermore, among these 16 genes, we pendency semiquantitatively, we sorted WT and STAT6-deficient detected GATA3 binding on 8 genes, as follows: GATA3, Itgb3, Th2 cells overexpressing GATA3 into three populations by a GZMA, Epas1, Crem, NFIL3, CCR8, and TNFRSF8 (Fig. 6A). cell sorter, as follows: NGFR2 (expressing no retrovirus-induced Taken together, these results indicate that among Th2-specific 6384 ROLE OF GATA3 IN Th2-SPECIFIC GENE INDUCTION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 6. Analysis of GATA3 binding at Th2-specific genes. A, ChIP-Seq signal profiles at GATA3-bound Th2-specific genes are shown. Upper panel, Indicates the profiles in Th1; lower panel, indicates those in Th2 cells. The y-axis indicates the average base coverage. Asterisk indicates 5-fold greater peak value compared with input DNA (see Materials and Methods for details). Open squares show gene bodies between TSS and TES, Crem (chr18: 3266352– 3299476), Epas1 (chr17: 87153242–87230922), GATA3 (chr2: 9778705–9800227), JDP2 (chr12: 86940366–86980828), NFIL3 (chr13: 53062578– 53076408), GZMA (chr13: 113884035–113891189), IL-4 (chr11: 53425970–53432167), IL-5 (chr11: 53534296–53538605), IL-13 (chr11: 53444826– 53448204), IL-24 (chr1: 132778651–132783988), CCR8 (chr9: 120001251–120004024), Itgb3 (chr11: 104469314–104531785), TNFRSF8 (chr4: 144858879–144905050), TMTC2 (chr10: 104624720–105011535), Asb2 (chr12: 104559352–104594211), Ccnjl (chr11: 43342286–43400499), and Cyp11a1 (chr9: 57862824–57874830). B, A ChIP assay of the GATA3-bound Th2-specific genes was performed using anti-GATA3 Ab and control Ab in Th1, Th2, and STAT6-deficient Th2 cells. The levels of binding were assessed by a qPCR analysis. The relative intensity (specific Ab ChIP 2 control Ig ChIP)/input DNA (highest signal intensity = 10) (mean of two samples) is shown with standard deviations. The binding region detected in the GZMA gene was unable to be included because of difficulties encountered during the design of qPCR primers specific for this region.

STAT6-dependent genes, some genes are bound by GATA3, and results of ChIP-Seq analysis, quantitative ChIP, and qRT-PCR their transcription is directly regulated by GATA3, whereas others assays. Although GATA3 has been recognized as a master regu- are either not directly bound by GATA3 or not regulated by lator of Th2 cell differentiation, we found that the transcription of GATA3 itself, but in collaboration with other molecules, including many Th2-specific genes was not regulated by GATA3 itself, and STAT6 (Table I). therefore GATA3 is not the only factor responsible for the induction of the transcription of a set of Th2-specific genes. Discussion In this study, we investigated in detail the dependency of 26 Th2- We performed ChIP-Seq analyses to assess genome-wide GATA3 specific STAT6-dependent inducible genes on GATA3. To select the binding and histone modifications in Th1 and Th2 cells. Simul- genes that are preferentially induced in developing Th2 cells, com- taneously, 26 genes that are selectively induced in developing Th2 parisons between WT Th2 cells and Th1 cells and between WT Th2 cells were identified. The recruitment of GATA3 and histone cells and STAT6-deficient Th2 cells were adopted (Fig. 1). It is known modifications at these 26 gene loci were analyzed in detail using the that the IL-4/IL-4R/STAT6 pathway is not the only downstream The Journal of Immunology 6385 Downloaded from http://www.jimmunol.org/

FIGURE 7. GATA3 dose-dependent recovery of expression of Th2-specific genes in STAT6-deficient Th2 cells. A, Freshly prepared STAT6-deficient by guest on September 27, 2021 CD4 T cells were stimulated under Th2-skewed conditions and were infected on day 2 with retrovirus encoding GATA3 bicistronically with NGFR (pMxs- GATA3-IRES-NGFR). Profiles of electronically gated NGFR2 (gate G1), NGFRlow (gate G2), and NGFRhigh (gate G3) populations are shown. B and C, mRNA expression levels of Th2-specific genes were assessed by qRT-PCR analysis in the indicated populations. WT and STAT6-deficient Th2 cells infected with mock retrovirus vector (pMXs-IRES-hNGFR) were also included for comparison. mRNA profiles of G1,G2, and G3 populations were de- termined. mRNA levels were normalized by mRNA level of WT Th2 cells infected with mock retrovirus vector, defined as level 1 at each fraction. B, GATA3-bound target genes. C, GATA3 nontarget genes. D, The RS of the Th2-specific genes in GATA3-introduced STAT6-deficient Th2 cells (gate G3). pathway of IL-4R signaling (40, 41). Therefore, the current study mRNA expressions were strongly induced by STAT6VT in- may exclude IL-4/IL-4R–dependent and STAT6-independent genes. troduction, but not by GATA3 (GB-2 group in Table I). For these Although it has not been well studied whether these genes are im- eight genes, GATA3 may not be the only factor that regulates portant for Th2 cell function, the use of IL-4–deficient and/or IL-4R– transcription, but may cooperate with STAT6 or STAT6-dependent deficient T cells may allow us to address this issue. genes. Furthermore, we also identified GATA3-nonbinding Th2- We examined GATA3 binding and H3K9Ac levels in the 92 specific genes whose expression was not recovered very well by Th2-specific STAT6-dependent genes, and compared the frequency GATA3 introduction, but strongly induced by STAT6VT in- between the 26 Th2-specific inducible genes and the remaining 66 troduction (S100a1, Ecm1, DUSP4, Tube1, IL1r2, RNF128, F2r, genes (Supplemental Fig. 4). Preferential GATA3 binding and and TANC2) (GN-2 group in Table I). For these eight genes, increased H3K9Ac levels were detected in the 26 Th2-specific STAT6 could play a critical role in transcription regardless of the inducible genes. These results may indicate a critical role for expression levels of GATA3. Finally, an interestingly behaving GATA3 in the increased expression of Th2-specific genes during gene PTGIR was identified (GN-1 group in Table I). Although Th2 cell differentiation. PTGIR was categorized as a GATA3 nontarget gene, the tran- Several interesting features in the regulation of transcription of scription of PTGIR was highly dependent on GATA3 expression. Th2-specific genes have been revealed. The summary is depicted Although we cannot exclude low-level binding of GATA3 to the in Table I. We identified nine GATA3-bound target genes (IL-13, PTGIR gene, an interesting possibility is that the transcription of Cyp11a1, Asb2, JDP2, IL-5, TMTC2, IL-4, Ccnjl, IL-24) whose the PTGIR gene is dependent on other unknown genes that were mRNA expression was highly dependent on GATA3 in the ab- induced by GATA3. Alternatively, GATA3 may bind to a distant sence of STAT6 (GB-1 group in Table I). For these genes, the gene enhancer region that is beyond the gene boundaries used in induction of GATA3 appears to be necessary and sufficient for the our global analysis (from 5 kb upstream of TSS to 5 kb down- upregulation of expression during Th2 cell differentiation. We stream of TES). As for the PTGIR gene, the nearest GATA3 peak also identified eight GATA3-bound target genes (GATA3, Itgb3, was located 872 kb upstream of the PTGIR gene TSS (S. Horiuchi, GZMA,Epas1,Crem,NFIL3,CCR8, and TNFRSF8) whose A. Onodera, Y. Suzuki, and T. Nakayama, unpublished observations). 6386 ROLE OF GATA3 IN Th2-SPECIFIC GENE INDUCTION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 8. STAT6VT-induced recovery of expression of Th2-specific genes in STAT6-deficient Th2 cells. A, Freshly isolated STAT6-deficient CD4 T cells were infected with a retrovirus vector containing STAT6VT cDNA (pMXs-STAT6VT-IRES-hNGFR) and stimulated under Th2 conditions. The STAT6VT-introduced cells were harvested 5 d postinfection. For cytokine genes, freshly isolated STAT6-deficient CD4 T cells were stimulated under Th2 conditions for 2 d, and then the cells were infected with a retrovirus vector containing STAT6VT cDNA. The STAT6VT-introduced cells were harvested 3 d postinfection. hNGFR-positive infected cells were enriched by magnetic cell sorting, and mRNA expression was assessed by qRT-PCR analysis. The relative intensity compared with HPRT (mean of three samples with SD) is shown. B, The RS of the Th2-specific genes in STAT6VT-introduced STAT6- deficient Th2 cells.

GATA3, Itgb3, GZMA, Epas1, Crem, NFIL3, CCR8, We selected 26 Th2-specific genes that include previously TNFRSF8, Ecm1, Dusp4, Tube1, IL1r2, and TANC2 were clas- reported Th2-specific genes (IL-4, IL-13, IL-5, IL-24, GATA3, sified as GATA3-low/STAT6-high (GB-2 and GN-2 in Table I), CCR8, and TNFRSF8) (13, 42–44), and also new genes that were and each of these have been reported previously as STAT6-bound not reported previously (Fig. 4). The newly identified genes were genes (second column, Table I) (13). We further confirmed a sig- phenotype-defining transcription factors (Crem, Epas1, JDP2, and nificant STAT6-dependent recovery of these genes (Table I, right NFIL3), a cytokine (GZMA), cell surface molecules (Ecm1, F2r, column). Thus, STAT6 binding is most likely important for the IL1r2, Itgb3, PTGIR, RNF128, and TMTC2), signaling mole- transcription for these 13 genes. Among these 13 genes, GATA3, cules, and others with unknown functions (Asb2, Ccnjl, Cyp11a1, Itgb3, GZMA, Epas1, Crem, NFIL3, CCR8, and TNFRSF8 were Dusp4, S100a1, TANC2, and Tube1). The IL-4–inducible gene GATA3-bound genes (GB-2 in Table I), and therefore, GATA3 NFIL3 is a transcription factor that is required for NK cell de- may collaborate with STAT6 to regulate transcription. Indeed, velopment and maintenance (45, 46). The c-Jun dimerization each GATA3 peak was located within 500 bp of the STAT6 peak protein 2, JDP2, is a DNA-binding protein that is known to be in these GB-2 gene loci (13). In contrast, S100a1, RNF128, and involved in cell differentiation process in skeletal muscle cells F2R were neither GATA3-bound genes nor STAT6 bound. The and osteoclasts (47, 48). Th2-specific increase in transcription of these three genes could be Of the GATA3 peaks detected in Th2 cells, we found WGATAR regulated by other genes that were directly regulated by STAT6 in motifs, which include GATA (Fig. 2D). Furthermore, we identified a similar fashion to that proposed for GATA3 and PTGIR. the ETS-binding motif (GGAA) among GATA3 peaks in Th2 cells The Journal of Immunology 6387

Table I. Overview of data presented in this study, highlighting representative genes that belong to each cluster

ChIP-Seq

GATA3-dep. STAT6-dep. RefSeq Accession Gene Name Gene Product GATA3 STAT6 (13) Recovery Recovery GATA3 target genes (GB) GATA3 high genes (GB-1) NM_008355 IL-13 IL-13 s 3 201.35 218.76 NM_019779 Cyp11a1 Cytochrome P450, family 11, subfamily a, ss122.42 130.58 polypeptide 1 NM_023049 Asb2 Ankyrin repeat and SOCS box-containing 2 ss92.92 48.75 NM_030893 JDP2 Jun dimerization protein 2 ss78.41 68.47 NM_010558 IL-5 IL-5 s 3 75.90 11.89 NM_177368 TMTC2 Transmembrane and tetratricopeptide ss55.56 42.60 repeat containing 2 NM_021283 IL-4 IL-4 ss50.77 33.53 NM_001045530 Ccnjl Cyclin J-like ss42.66 36.09 NM_053095 IL-24 IL-24 ss41.26 111.00

GATA3 low genes (GB-2)

STAT6 high genes Downloaded from NM_008091 GATA3 GATA-binding protein 3 ss12.35 131.76 NM_016780 Itgb3 Integrin b3 ss26.34 142.46 NM_010360 GZMA Granzyme A ss25.74 113.82 NM_010137 Epas1 Endothelial PAS domain protein 1 ss20.06 88.21 NM_013498 Crem cAMP-responsive element modulator ss27.90 85.55 NM_017373 NFIL3 NF, IL-3, regulated ss22.32 71.37 NM_007720 CCR8 Chemokine (C-C motif) receptor 8 ss19.44 68.65 NM_009401 TNFRSF8 TNFR superfamily, member 8 ss216.18 66.62 http://www.jimmunol.org/

GATA3 nontarget genes (GN) GATA3 high genes (GN-1) NM_008967 PTGIR PG I receptor (IP) 33110.02 126.88

GATA3 low genes (GN-2) STAT6 high genes NM_011309 S100a1 S100 calcium-binding protein A1 3318.57 659.10 NM_007899 Ecm1 Extracellular matrix protein 1 3 s 2.16 289.46 NM_176933 DUSP4 Dual specificity phosphatase 4 3 s 28.20 209.64 by guest on September 27, 2021 NM_028006 Tube1 ε- 1 3 s 221.52 175.88 NM_010555 IL1r2 IL-1R, type II 3 s 221.22 146.32 NM_023270 RNF128 Ring finger protein 128 3320.85 118.21 NM_010169 F2r Coagulation factor II (thrombin) receptor 3329.09 109.11 NM_181071 TANC2 Tetratricopeptide repeat, ankyrin repeat, 3 s 26.89 56.80 and coiled-coil containing 2 In this study, integration of the three following biological readouts was used for the identification of gene clusters: 1) GATA3 binding; 2) changes in gene expression by enforced expression of GATA3; and 3) changes in gene expression by enforced expression of STAT6VT. These represent various actions of GATA3 and STAT6 involved in the transcription of Th2-specific genes. GATA3 or STAT6 dependency of the 26 genes was determined by the RS, as described in Materials and Methods. s, binding is detected; 3, binding is not detected; dep., dependent in the MEME analysis (Supplemental Fig. 1B), indicating a pos- (19, 20, 39). There is a previous report on genome-wide GATA3 sibility that ETS family members bound to the sequence close target analysis in primary human T cells using a ChIP with to the GATA3-binding motif. Indeed, ETS-1 is reported to play microarray technology (ChIP-chip) analysis (50). We compared an important role in regulating the expression of Th2 cytokines the results of this study with our own, and an interesting difference in Th2 cells (49). A consensus motif for GATA3 was not identi- was noted. IL-4 was identified as a GATA3 target in both analyses. fied in the GATA3 peaks detected in Th1 cells or those that However, many GATA3 target genes identified in the current overlapped in Th1 and Th2 cells (Supplemental Fig. 1A). How- study, including IL-5 and IL-13, were not listed in the previous ever, because the expression of GATA3 is lower in Th1 cells in genome-wide ChIP-chip analysis in human T cells (50). This comparison with Th2 cells, the binding peak analysis may in- discrepancy may be explained by the difference in experimental correctly identify nonspecific binding more frequently in Th1 cells. systems, as follows: 1) the difference between human and mouse; Our present genome-wide surveys on GATA3 binding and his- 2) the different analysis of ChIP-chip and ChIP-Seq; 3) the dif- tone modifications indicate that GATA3 binding does not always ference in the time course of gene expression in in vitro differ- correlate with changes in the active form of histone modifications in entiated human and mouse T cells; and 4) the use of a different Th2 cells. From this result, there is a possibility that GATA3 acts anti-GATA3 Ab. In fact, in the genome-wide analyses of STAT6 in not only as an activator, but also as a repressor in both Th1 and Th2 human and mice, STAT6 binding to certain genes, including cells. Thus, the set of GATA3-binding genes that has no active form GATA3, was different between the two species (13, 31). The of histone modification will be interesting to analyze in further difference in the results between ChIP-chip and ChIP-Seq analy- detail. ses was also reported (51, 52). Human Th2 cells were prepared Th2 cell differentiation is accompanied by histone modifications, 7 d after TCR stimulation (50). This procedure may not induce such as H3-K9/14 acetylation and H3-K4 methylation at the IL-4, high-level expression of many Th2-specific genes because our IL-13, and IL-5 loci, which are dependent on GATA3 expression previous studies pointed out that two cycles of TCR stimulation 6388 ROLE OF GATA3 IN Th2-SPECIFIC GENE INDUCTION were required for the high-level expression of IL-4 and IL-5 (53, 18. Lo¨hning, M., A. Richter, and A. Radbruch. 2002. Cytokine memory of T helper lymphocytes. Adv. Immunol. 80: 115–181. 54). We are not able to determine the reason for the discrepancy at 19. Ansel, K. M., I. Djuretic, B. Tanasa, and A. Rao. 2006. Regulation of Th2 present, and further studies are required. differentiation and Il4 locus accessibility. Annu. Rev. Immunol. 24: 607–656. In summary, our study confirms the notion that ChIP-Seq is an 20. Nakayama, T., and M. Yamashita. 2008. Initiation and maintenance of Th2 cell identity. Curr. Opin. Immunol. 20: 265–271. excellent technology to facilitate the analysis of genome-wide 21. Omori, M., M. Yamashita, M. Inami, M. Ukai-Tadenuma, M. Kimura, Y. Nigo, binding of transcription factors to target genes in a nonbiased H. Hosokawa, A. Hasegawa, M. Taniguchi, and T. Nakayama. 2003. CD8 T cell- fashion, and also indicates that additional accurate gene expression specific downregulation of histone hyperacetylation and gene activation of the IL-4 gene locus by ROG, repressor of GATA. Immunity 19: 281–294. analyses are required to address the true biological function of the 22. Yamashita, M., M. Ukai-Tadenuma, T. Miyamoto, K. Sugaya, H. Hosokawa, transcription factor at the target genes. The present study provides A. Hasegawa, M. Kimura, M. Taniguchi, J. DeGregori, and T. Nakayama. 2004. a platform for understanding how GATA3 regulates the tran- Essential role of GATA3 for the maintenance of type 2 helper T (Th2) cytokine production and chromatin remodeling at the Th2 cytokine gene loci. J. Biol. scription of Th2-specific genes during Th2 cell differentiation. Chem. 279: 26983–26990. 23. Inami, M., M. Yamashita, Y. Tenda, A. Hasegawa, M. Kimura, K. Hashimoto, N. Seki, M. Taniguchi, and T. Nakayama. 2004. CD28 costimulation controls Acknowledgments histone hyperacetylation of the interleukin 5 gene locus in developing th2 cells. We are grateful to Drs. Damon Tumes and Ellen Rothenberg for helpful J. Biol. Chem. 279: 23123–23133. comments and constructive criticisms in the preparation of the manuscript. 24. Pai, S. Y., M. L. Truitt, and I. C. Ho. 2004. GATA-3 deficiency abrogates the development and maintenance of T helper type 2 cells. Proc. Natl. Acad. Sci. We thank Kaoru Sugaya, Hikari K. Asou, Miki Katou, and Toshihiro Ito for USA 101: 1993–1998. excellent technical assistance. 25. Zhu, J., B. Min, J. Hu-Li, C. J. Watson, A. Grinberg, Q. Wang, N. Killeen, J. F. Urban, Jr., L. Guo, and W. E. Paul. 2004. Conditional deletion of Gata3 shows its essential function in T(H)1-T(H)2 responses. Nat. Immunol. 5: 1157–1165. Disclosures 26. Wurster, A. L., T. Tanaka, and M. J. Grusby. 2000. The biology of Stat4 and Downloaded from The authors have no financial conflicts of interest. Stat6. Oncogene 19: 2577–2584. 27. Takeda, K., T. Tanaka, W. Shi, M. Matsumoto, M. Minami, S. Kashiwamura, K. Nakanishi, N. Yoshida, T. Kishimoto, and S. Akira. 1996. Essential role of Stat6 in IL-4 signalling. Nature 380: 627–630. References 28. Chen, Z., R. Lund, T. Aittokallio, M. Kosonen, O. Nevalainen, and R. Lahesmaa. 1. Mosmann, T. R., and R. L. Coffman. 1989. TH1 and TH2 cells: different patterns 2003. Identification of novel IL-4/Stat6-regulated genes in T lymphocytes. J. of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 171: 3627–3635.

Immunol. 7: 145–173. 29. Lund, R. J., M. Lo¨yto¨ma¨ki, T. Naumanen, C. Dixon, Z. Chen, H. Ahlfors, http://www.jimmunol.org/ 2. Reiner, S. L., and R. M. Locksley. 1995. The regulation of immunity to Leish- S. Tuomela, J. Tahvanainen, J. Scheinin, T. Henttinen, et al. 2007. Genome-wide mania major. Annu. Rev. Immunol. 13: 151–177. identification of novel genes involved in early Th1 and Th2 cell differentiation. J. 3. Korn, T., E. Bettelli, M. Oukka, and V. K. Kuchroo. 2009. IL-17 and Th17 cells. Immunol. 178: 3648–3660. Annu. Rev. Immunol. 27: 485–517. 30. Barski, A., S. Cuddapah, K. Cui, T. Y. Roh, D. E. Schones, Z. Wang, G. Wei, 4. Szabo, S. J., B. M. Sullivan, C. Stemmann, A. R. Satoskar, B. P. Sleckman, and I. Chepelev, and K. Zhao. 2007. High-resolution profiling of histone methyl- L. H. Glimcher. 2002. Distinct effects of T-bet in TH1 lineage commitment and ations in the . Cell 129: 823–837. IFN-g production in CD4 and CD8 T cells. Science 295: 338–342. 31. Elo, L. L., H. Ja¨rvenpa¨a¨, S. Tuomela, S. Raghav, H. Ahlfors, K. Laurila, 5. Zheng, W., and R. A. Flavell. 1997. The transcription factor GATA-3 is neces- B. Gupta, R. J. Lund, J. Tahvanainen, R. D. Hawkins, et al. 2010. Genome-wide sary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 89: profiling of interleukin-4 and STAT6 transcription factor regulation of human 587–596. Th2 cell programming. Immunity 32: 852–862. 6. Lee, H. J., N. Takemoto, H. Kurata, Y. Kamogawa, S. Miyatake, A. O’Garra, and 32. Yamashita, M., K. Hirahara, R. Shinnakasu, H. Hosokawa, S. Norikane, N. Arai. 2000. GATA-3 induces T helper cell type 2 (Th2) cytokine expression M. Y. Kimura, A. Hasegawa, and T. Nakayama. 2006. Crucial role of MLL for the by guest on September 27, 2021 and chromatin remodeling in committed Th1 cells. J. Exp. Med. 192: 105–115. maintenance of memory T helper type 2 cell responses. Immunity 24: 611–622. 7. Ivanov, I. I., B. S. McKenzie, L. Zhou, C. E. Tadokoro, A. Lepelley, J. J. Lafaille, 33. Demers, C., C. P. Chaturvedi, J. A. Ranish, G. Juban, P. Lai, F. Morle, D. J. Cua, and D. R. Littman. 2006. The orphan nuclear receptor RORgammat R. Aebersold, F. J. Dilworth, M. Groudine, and M. Brand. 2007. Activator- directs the differentiation program of proinflammatory IL-17+ T helper cells. mediated recruitment of the MLL2 methyltransferase complex to the b-globin Cell 126: 1121–1133. locus. Mol. Cell 27: 573–584. 8. Yang, X. O., B. P. Pappu, R. Nurieva, A. Akimzhanov, H. S. Kang, Y. Chung, 34. Bailey, T. L., and C. Elkan. 1994. Fitting a mixture model by expectation L. Ma, B. Shah, A. D. Panopoulos, K. S. Schluns, et al. 2008. T helper 17 lineage maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. differentiation is programmed by orphan nuclear receptors ROR a and ROR g. Mol. Biol. 2: 28–36. Immunity 28: 29–39. 35. Wang, Z., C. Zang, J. A. Rosenfeld, D. E. Schones, A. Barski, S. Cuddapah, 9. Nakayama, T., and M. Yamashita. 2009. Critical role of the Polycomb and K. Cui, T. Y. Roh, W. Peng, M. Q. Zhang, and K. Zhao. 2008. Combinatorial Trithorax complexes in the maintenance of CD4 T cell memory. Semin. Immu- patterns of histone acetylations and methylations in the human genome. Nat. nol. 21: 78–83. Genet. 40: 897–903. 10. Yamamoto, M., L. J. Ko, M. W. Leonard, H. Beug, S. H. Orkin, and J. D. Engel. 36. Roh, T. Y., S. Cuddapah, K. Cui, and K. Zhao. 2006. The genomic landscape of 1990. Activity and tissue-specific expression of the transcription factor NF-E1 histone modifications in human T cells. Proc. Natl. Acad. Sci. USA 103: 15782– multigene family. Genes Dev. 4: 1650–1662. 15787. 11. Hosoya, T., I. Maillard, and J. D. Engel. 2010. From the cradle to the grave: 37. Ko, L. J., and J. D. Engel. 1993. DNA-binding specificities of the GATA tran- activities of GATA-3 throughout T-cell development and differentiation. Immu- scription factor family. Mol. Cell. Biol. 13: 4011–4022. nol. Rev. 238: 110–125. 38. Yu, M., L. Riva, H. Xie, Y. Schindler, T. B. Moran, Y. Cheng, D. Yu, 12. Ouyang, W., S. H. Ranganath, K. Weindel, D. Bhattacharya, T. L. Murphy, R. Hardison, M. J. Weiss, S. H. Orkin, et al. 2009. Insights into GATA-1- W. C. Sha, and K. M. Murphy. 1998. Inhibition of Th1 development mediated by mediated gene activation versus repression via genome-wide chromatin occu- GATA-3 through an IL-4-independent mechanism. Immunity 9: 745–755. pancy analysis. Mol. Cell 36: 682–695. 13. Wei, L., G. Vahedi, H. W. Sun, W. T. Watford, H. Takatori, H. L. Ramos, 39. Yamashita, M., M. Ukai-Tadenuma, M. Kimura, M. Omori, M. Inami, H. Takahashi, J. Liang, G. Gutierrez-Cruz, C. Zang, et al. 2010. Discrete roles of M. Taniguchi, and T. Nakayama. 2002. Identification of a conserved GATA3 STAT4 and STAT6 transcription factors in tuning epigenetic modifications and response element upstream proximal from the interleukin-13 gene locus. J. Biol. transcription during T helper cell differentiation. Immunity 32: 840–851. Chem. 277: 42399–42408. 14. Onodera, A., M. Yamashita, Y. Endo, M. Kuwahara, S. Tofukuji, H. Hosokawa, 40. Finkelman, F. D., S. C. Morris, T. Orekhova, M. Mori, D. Donaldson, A. Kanai, Y. Suzuki, and T. Nakayama. 2010. STAT6-mediated displacement of S. L. Reiner, N. L. Reilly, L. Schopf, and J. F. Urban, Jr. 2000. Stat6 regulation of polycomb by trithorax complex establishes long-term maintenance of GATA3 in vivo IL-4 responses. J. Immunol. 164: 2303–2310. expression in T helper type 2 cells. J. Exp. Med. 207: 2493–2506. 41. Nelms, K., A. D. Keegan, J. Zamorano, J. J. Ryan, and W. E. Paul. 1999. The IL- 15. Hosokawa, H., M. Y. Kimura, R. Shinnakasu, A. Suzuki, T. Miki, H. Koseki, 4 receptor: signaling mechanisms and biologic functions. Annu. Rev. Immunol. M. van Lohuizen, M. Yamashita, and T. Nakayama. 2006. Regulation of Th2 cell 17: 701–738. development by Polycomb group gene bmi-1 through the stabilization of 42. D’Ambrosio, D., A. Iellem, R. Bonecchi, D. Mazzeo, S. Sozzani, A. Mantovani, GATA3. J. Immunol. 177: 7656–7664. and F. Sinigaglia. 1998. Selective up-regulation of chemokine receptors CCR4 16. Shinnakasu, R., M. Yamashita, K. Shinoda, Y. Endo, H. Hosokawa, and CCR8 upon activation of polarized human type 2 Th cells. J. Immunol. 161: A. Hasegawa, S. Ikemizu, and T. Nakayama. 2006. Critical YxKxHxxxRP motif 5111–5115. in the C-terminal region of GATA3 for its DNA binding and function. J. 43. Mikhak, Z., M. Fukui, A. Farsidjani, B. D. Medoff, A. M. Tager, and A. D. Immunol. 177: 5801–5810. Luster. 2009. Contribution of CCR4 and CCR8 to antigen-specific T(H)2 cell 17. Yamashita, M., R. Shinnakasu, H. Asou, M. Kimura, A. Hasegawa, trafficking in allergic pulmonary inflammation. J. Allergy Clin. Immunol. 123: K. Hashimoto, N. Hatano, M. Ogata, and T. Nakayama. 2005. Ras-ERK MAPK 67-73.e3. cascade regulates GATA3 stability and Th2 differentiation through ubiquitin- 44. Del Prete, G., M. De Carli, F. Almerigogna, C. K. Daniel, M. M. D’Elios, proteasome pathway. J. Biol. Chem. 280: 29409–29419. G. Zancuoghi, F. Vinante, G. Pizzolo, and S. Romagnani. 1995. Preferential The Journal of Immunology 6389

expression of CD30 by human CD4+ T cells producing Th2-type cytokines. 50. Jenner, R. G., M. J. Townsend, I. Jackson, K. Sun, R. D. Bouwman, R. A. Young, FASEB J. 9: 81–86. L. H. Glimcher, and G. M. Lord. 2009. The transcription factors T-bet and 45. Kamizono, S., G. S. Duncan, M. G. Seidel, A. Morimoto, K. Hamada, GATA-3 control alternative pathways of T-cell differentiation through a shared G. Grosveld, K. Akashi, E. F. Lind, J. P. Haight, P. S. Ohashi, et al. 2009. Nfil3/ set of target genes. Proc. Natl. Acad. Sci. USA 106: 17876–17881. E4bp4 is required for the development and maturation of NK cells in vivo. J. 51. Gilchrist, D. A., D. C. Fargo, and K. Adelman. 2009. Using ChIP-chip and Exp. Med. 206: 2977–2986. ChIP-Seq to study the regulation of gene expression: genome-wide localiza- 46. Gascoyne, D. M., E. Long, H. Veiga-Fernandes, J. de Boer, O. Williams, tion studies reveal widespread regulation of transcription elongation. Methods B. Seddon, M. Coles, D. Kioussis, and H. J. Brady. 2009. The basic leucine 48: 398–408. zipper transcription factor E4BP4 is essential for natural killer cell development. 52. Zhang, Y., T. Liu, C. A. Meyer, J. Eeckhoute, D. S. Johnson, B. E. Bernstein, Nat. Immunol. 10: 1118–1124. C. Nusbaum, R. M. Myers, M. Brown, W. Li, and X. S. Liu. 2008. Model-based 47. Ostrovsky, O., E. Bengal, and A. Aronheim. 2002. Induction of terminal dif- Genome Biol. ferentiation by the c-Jun dimerization protein JDP2 in C2 myoblasts and rhab- analysis of ChIP-Seq (MACS). 9: R137. domyosarcoma cells. J. Biol. Chem. 277: 40043–40054. 53. Kaneko, T., H. Hosokawa, M. Yamashita, C. R. Wang, A. Hasegawa, 48. Kawaida, R., T. Ohtsuka, J. Okutsu, T. Takahashi, Y. Kadono, H. Oda, A. Hikita, M. Y. Kimura, M. Kitajiama, F. Kimura, M. Miyazaki, and T. Nakayama. 2007. K. Nakamura, S. Tanaka, and H. Furukawa. 2003. Jun dimerization protein 2 Chromatin remodeling at the Th2 cytokine gene loci in human type 2 helper (JDP2), a member of the AP-1 family of transcription factor, mediates osteoclast T cells. Mol. Immunol. 44: 2249–2256. differentiation induced by RANKL. J. Exp. Med. 197: 1029–1035. 54. Ito, T., A. Hasegawa, H. Hosokawa, M. Yamashita, S. Motohashi, T. Naka, 49. Strempel, J. M., R. Grenningloh, I. C. Ho, and D. Vercelli. 2010. Phylogenetic Y. Okamoto, Y. Fujita, Y. Ishii, M. Taniguchi, et al. 2008. Human Th1 differ- and functional analysis identifies Ets-1 as a novel regulator of the Th2 cytokine entiation induced by lipoarabinomannan/lipomannan from Mycobacterium bovis gene locus. J. Immunol. 184: 1309–1316. BCG Tokyo-172. Int. Immunol. 20: 849–860. Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021 1

Supplemental Figure legends

Figure S1. Genome-wide distribution of GATA3 consensus binding motifs and GATA3-bound genes form clusters that share common epigenetic signatures.

(A and B), MEME (34) was used to perform a de novo search of consensus binding motifs for GATA3 at the peaks in Th1 cells and peaks that were overlapped in both Th1 and Th2 cells (A). The binding motifs for ESTs in Th2 cells (B). (C and D), GATA3-bound genes in Th1 cells (C) and genes that were bound by GATA3 both in Th1 and Th2 cells (D) form clusters that share common epigenetic signatures. The total tag count of each epigenetic modification (H3K9Ac, H3K4Me3, and H3K27Me3) was computed across the promoter of each GATA3-bound gene in Th2 versus Th1 cells. The ratio of tag counts for three epigenetic modifications was used to cluster GATA3-bound genes in 4 patterns: H3K9Ac-high, H3K4Me3-high, H3K27Me3-low, and an indeterminate pattern.

Figure S2. Screening of Th2-specific genes by ΔΔC qRT-PCR.

Gene expression of the selected 185 genes in Th2, and STAT6-deficient Th2 cells was determined by a ΔΔC qRT-PCR. The relative signal intensity of Th2 compared to signal intensity of STAT6-deficient Th2 cells is shown in Log2 scale. (A), The results for genes that were highly expressed in Th2 cells (more than 2-fold). (B), The results for genes that were not highly expressed in Th2 cells (less than 2-fold).

Figure S3. ChIP-seq signal profiles of H3-K9Ac, H3-K4Me3, H3-K27Me3 modifications at Th2-specific genes.

ChIP-seq signal profiles for H3-K9Ac, H3-K4Me3, H3-K27Me3 modifications at GATA3-bound Th2-specific genes in Th1 and Th2 cells are shown. The y-axis indicates the average base coverage. Open squares show gene bodies between TSS and TES, Crem (chr18: 3266352 - 3299476), Epas1 (chr17: 87153242 - 87230922), GATA3 (chr2: 9778705 - 9800227), JDP2 (chr12: 86940366 - 86980828), NFIL3 (chr13: 53062578 - 53076408), GZMA (chr13: 113884035 - 113891189), IL-4 (chr11: 53425970 - 53432167), IL-5 (chr11: 53534296 - 53538605), IL-13 (chr11: 53444826 - 2

53448204), IL-24 (chr1: 132778651 - 132783988), CCR8 (chr9: 120001251 - 120004024), Ecm1 (chr3: 95538071 - 95543492), F2r (chr13: 96371744 - 96388388),IL1r2 (chr1: 40141613 - 40182070), Itgβ3 (chr11: 104469314 - 104531785), PTGIR (chr7: 17491839 - 17496254), RNF128 (chrX: 136145159 - 136207674), TNFRSF8 (chr4: 144858879 - 144905050), TMTC2 (chr10: 104624720 - 105011535), Asb2 (chr12: 104559352 - 104594211), Ccnjl (chr11: 43342286 - 43400499), Cyp11a1 (chr9: 57862824 - 57874830), TANC2 (chr11: 105451300 - 105790617), Tube1 (chr10: 38853829 – 38870864), DUSP4 (chr8: 35870664 - 35882948), and S100a1 (chr3: 90314956 - 90318252).

Figure S4. Correlation between Th2-specific gene expression and GATA3 binding or H3K9Ac level

(A), Frequency of GATA3 target genes is shown with gene expression levels of the 92 Th2-specific genes. These 92 genes are the same as those shown in Supplemental Fig. 1A. Th2-specific STAT6-dependent inducible genes (left, 26 genes) and other Th2-specific STAT6-dependent genes (right, 66 genes) are separately shown. (B), The levels of H3K9Ac histone modification of the 92 genes are shown as the Th2/Th1 ratio of H3K9Ac relative tag counts.