Identification of a Conserved GATA3 Response Element Upstream Proximal from the Interleukin-13 Gene Locus* 210

Differentiation of naive CD4 T cells into type 2 helper (Th2) cells is accompanied by chromatin remodeling of Th2 cytokine gene loci. Hyperacetylation of histone H3 on nucleosomes associated with the interleukin (IL)-4, IL-13 and IL-5 genes was observed in developing Th2 cells but not in Th1 cells. Histone hyperacetylation on IL-5 gene-associated nucleosomes was Th2-specific but occurred with delayed kinetics, and hyperacetylation on RAD50 gene-associated nucleosomes was T cell antigen receptor stimulation-dependent but not Th2-specific. The induction of the Th2-specific histone hyperacetylation was STAT6- and GATA3-dependent, and interestingly, it was accompanied by the expression of intergenic transcripts within the IL-13 and IL-4 gene loci. A conserved GATA3 response element (CGRE) containing four GATA consensus sequences was identified 1.6 kbp upstream from the IL-13 gene, corresponding with the 5′-border of the Th2-specific histone hyperacetylation region. The CGRE was shown to bind to GATA3, histone acetyltransferase complexes including CBP/p300, and RNA polymerase II. Also, the CGRE showed a significant enhancing effect on the Th2 cytokine gene promoters. Thus, the CGRE may play a crucial role for GATA3-mediated targeting and downstream spreading of core histone hyperacetylation within the IL-13 and IL-4 gene loci.


Differentiation of naive CD4 T cells into type 2 helper (Th2) cells is accompanied by chromatin remodeling of
Th2 cytokine gene loci. Hyperacetylation of histone H3 on nucleosomes associated with the interleukin (IL)-4, IL-13 and IL-5 genes was observed in developing Th2 cells but not in Th1 cells. Histone hyperacetylation on IL-5 gene-associated nucleosomes was Th2-specific but occurred with delayed kinetics, and hyperacetylation on RAD50 gene-associated nucleosomes was T cell antigen receptor stimulation-dependent but not Th2-specific. The induction of the Th2-specific histone hyperacetylation was STAT6-and GATA3-dependent, and interestingly, it was accompanied by the expression of intergenic transcripts within the IL-13 and IL-4 gene loci. A conserved GATA3 response element (CGRE) containing four GATA consensus sequences was identified 1.6 kbp upstream from the IL-13 gene, corresponding with the 5-border of the Th2-specific histone hyperacetylation region. The CGRE was shown to bind to GATA3, histone acetyltransferase complexes including CBP/p300, and RNA polymerase II. Also, the CGRE showed a significant enhancing effect on the Th2 cytokine gene promoters. Thus, the CGRE may play a crucial role for GATA3-mediated targeting and downstream spreading of core histone hyperacetylation within the IL-13 and IL-4 gene loci.
Upon stimulation with antigens, naive CD4 T cells differentiate into two distinct T helper cell subsets, Th1 and Th2 1 (1).
Th1 cells produce interferon (IFN)-␥ and tumor necrosis factor-␤ and direct cell-mediated immunity against intracellular pathogens. Th2 cells produce interleukin (IL)-4, IL-5, and IL-13 and are involved in humoral immunity and allergic reactions. The outcome of Th cell differentiation depends on the cytokine environment (2,3). Naive CD4 T cells stimulated with antigens in the presence of IL-12 differentiate into Th1 cells, whereas the presence of IL-4 drives differentiation into Th2 cells (4 -6). For Th1 cell differentiation, the IL-12-mediated activation of signal transducer and activator of transcription (STAT) 4 is required, whereas IL-4-mediated STAT6 activation is important for Th2 cell development (7)(8)(9). In addition to the cytokines mentioned above, T cell antigen receptor (TCR) stimulation by antigens is also indispensable for both Th1 and Th2 cell differentiation. We reported that efficient TCR-mediated activation of the p56 lck , calcineurin, and Ras-ERK mitogen-activated protein kinase signaling cascade was required for Th2 cell generation (10 -12).
Recent studies have identified several transcription factors that control Th1/Th2 cell differentiation. Among them, GATA3 appears to be a key factor for Th2 cell differentiation. GATA3 is expressed selectively in Th2 cells, and its ectopic expression induced Th2 cell differentiation even in the absence of STAT6 (13)(14)(15)(16). For Th1 cell differentiation, T-bet was recently identified as a key transcription factor (17). Analogous to GATA3 in Th2 cell differentiation, the ectopic expression of T-bet induced IFN-␥ production and suppression of Th2 cytokine expression.
Changes in the chromatin structure of the Th2 cytokine (IL-4/IL-5/IL-13) gene loci occur during Th2 cell differentiation (18 -21). Takemoto et al. (20) demonstrated the induction of DNase I-hypersensitive sites located between the IL-4 and IL-13 genes during Th2 cell differentiation. The highly conserved 400-bp noncoding sequence 1 (CNS1) was identified in this region (22), and an important role for CNS1 in the expression of Th2 cytokines was revealed in both a human YAC transgenic mouse (22) and a mutant mouse lacking the CNS1 region (23). In addition, Agarwal and Rao (18) identified five DNase I-hypersensitive sites that were induced in the vicinity of the IL-4 gene during Th2 cell differentiation. More recently, a 3Ј-distal IL-4 enhancer (V A ) containing an inducible DNase I-hypersensitive site was identified (24). Reiner and colleagues (19) reported that demethylation of the intron 2 region of the IL-4 gene was associated with cell cycle progression and Th2 cell differentiation (19). We recently reported that demethylation of this region is regulated by polycomb group genes (25) that are known to regulate transcriptional memory in Drosophila (26).
It is also well documented that the modification of histone N termini is crucial for the changes in chromatin structure seen in sites of gene expression (27). For example, hyperacetylation of histone H3 and H4 tails by histone acetyltransferases was suggested to mark the remodeling site (28) and to stabilize the binding of the SNF/SWI complex to nucleosomes (29). Furthermore, large domains of transcriptionally active chromatin were found to contain hyperacetylated histones and showed increased sensitivity to DNase I (30).
In the present study, we investigated histone hyperacetylation on nucleosomes associated with Th2 cytokine gene loci in developing Th2 cells. The results indicate that hyperacetylation of histones associated with the IL-4 and IL-13 gene loci but not with the RAD50 locus is Th2-specific and is dependent on GATA3. We identified a 71-bp conserved GATA3 response element (CGRE) at 1.6 kbp upstream from the IL-13 locus. The CGRE was shown to bind GATA3 in vitro and in vivo. Also, p300 and RNA polymerase II were precipitated with CGRE oligonucleotide in vitro. The roles for the CGRE in the Th2specific histone hyperacetylation of the IL-4-and IL-13-geneassociated nucleosomes are discussed.

EXPERIMENTAL PROCEDURES
Mice-C57BL/6 mice were purchased from SLC (Shizuoka, Japan). STAT6-deficient mice were kindly provided by Shizuo Akira (Osaka University, Osaka, Japan) (31). All mice used in this study were maintained under specific pathogen-free conditions and were about 4 weeks old. Animal care was in accordance with the guidelines of Chiba University.
Cell Cultures and in Vitro T Cell Differentiation-CD4 T cells were purified using magnetic beads and an Auto-MACS Sorter (Miltenyi Biotec), yielding purity of Ͼ98%. Enriched CD4 T cells (1.5 ϫ 10 6 ) were stimulated for 2 days with 3 g/ml immobilized anti-TCR mAb (H57-597) in the added presence of 25 units/ml IL-2, 100 units/ml IL-12, and anti-IL-4 mAb 11B11 (25% culture supernatant) for Th1-skewed conditions. For Th2-skewed conditions, cells were stimulated with immobilized anti-TCR mAb as above but in the presence of 25 units/ml IL-2, 100 units/ml IL-4, and anti-IFN-␥ mAb R4.6A2 (25% culture supernatant). The cells were then transferred to new dishes and cultured for another 3 days in the presence of only the cytokines present in the initial culture. The levels of Th1/Th2 cell generation were then assessed by intracellular cytokine staining with anti-IL-4 and anti-IFN-␥ as described by Yamashita et al. (10).
Chromatin Immunoprecipitation (ChIP) Assay-ChIP was performed using histone H3 and histone H4 ChIP assay kits (Upstate Biotechnology). Some assays employed anti-GATA3 mAb (Santa Cruz Biotechnology) as described under "Results." In brief, 1 ϫ 10 6 CD4 T cells were fixed with 1% paraformaldehyde at 37°C for 10 min. Cells were sedimented, washed, lysed with SDS lysis buffer (50 mM Tris-HCl, 1% SDS, 10 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, and 1 g/ml pepstatin A). The lysates were sonicated to reduce DNA lengths to between 200 and 1,000 bp. The soluble fraction was diluted, precleared with salmon sperm DNA/protein A-agarose, and then incubated with 6 l of antiserum specific for the acetylated forms of histones H3 or H4. Then immune complexes were precipitated with protein A-agarose. The precipitated DNA was eluted with an elution buffer (0.1 M NaHCO 3 containing 1% SDS). The eluted material was incubated at 65°C for 6 h to reverse the formaldehyde cross-links, and DNA was extracted with phenol and chloroform. Ethanol-precipitated DNA was solubilized in water (1 ϫ 10 6 cell equivalent/100 l). Semiquantitative PCR was performed with 3, 1, and 0.3 l of DNA samples (a 3-fold dilution) at 27 cycles. PCR products were resolved by agarose gel electrophoresis and visualized with ethidium bromide. Images were recorded and quantified using ATTO L&S analyzer (ATTO, Tokyo, Japan). Semiquantitative PCR was performed using the primers listed in the supplemental figure.
RT-PCR-Total RNA was isolated using Trisol reagent (Invitrogen). All samples were treated with RNase-free DNase I (Takara) at 37°C for 30 min under the following conditions: 20 g of RNA, 40 mM Tris-HCl, 8 mM MgCl 2 , 5 mM dithiothreitol, 0.4 unit/l RNase inhibitor (Promega), and 10 units of DNase I in a volume of 50 l. Then, a phenol/ chloroform extraction was performed and the RNA precipitated. Reverse transcription was done using Superscript II (Invitrogen). Control reactions were set up under identical conditions but without the inclusion of reverse transcriptase. The primers for detection of intergenic transcripts were the same as those used in ChIP assay.
Pull-down Assay-Purified CD4 T cells were stimulated with immobilized anti-TCR mAb for 3 days under Th2-skewed conditions. Then, the cells were pelleted, resuspended in buffer C (20 mM HEPES (pH 7.9), 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 1 mM dithiothreitol, 0.1% Nonidet P-40, 1 mM Na 2 V0 3 , 1 mM NaF, 1 mM ␤-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride), and lysed on ice for 15 min. Insoluble material was removed by centrifugation. The supernatant was diluted 1:3 with buffer D (buffer C without NaCl). The lysates were incubated with 10 g of poly(dI-dC)⅐poly(dI-dC) (Amersham Biosciences) and 70 l of streptavidin-agarose (Upstate Biotechnology) carrying biotinylated oligonucleotides. The reaction mixtures were incubated at constant rotation for 40 min at 4°C. After removal of the supernatant, the beads were washed twice with buffer C/buffer D (1:3). The bound protein was eluted by adding 50 ml of SDS sample buffer and separated on a SDS-polyacrylamide gel. Subsequently, protein was subjected to immunoblot analysis using mAbs specific for GATA3, p300, and RNA polymerase II (Santa Cruz Biotechnology). Retroviral Vectors and Infection-pMX-IRES-GFP plasmid was kindly provided by Toshio Kitamura (Tokyo University, Tokyo, Japan). The methods for the generation of virus supernatant and CD4 T cell infection were described previously (25). Infected cells were subjected to intracellular staining with anti-IL-4 and anti-IFN-␥ mAb, or to cell sorting. cDNA for human GATA3 was inserted into a multicloning site of pMX-IRES-GFP.
Luciferase Reporter Assay-A single copy of a 260-bp fragment spanning the CGRE site (CGRE260 WT) or CGRE260 with a single GATA binding site 3 mutation (CGRE260 M3) was cloned at the 5Ј of the IL-4 promoter (Ϫ766), IL-13 promoter (Ϫ254), and IL-5 promoter (Ϫ1200) in the luciferase reporter plasmid pGL2Basic (Promega). M12 cells (B cell line) were used for transfection by electoroporation (960 microfarads/ 230 V). For transfection, 10 g of each promoter reporter construct was used in combination with 10 g of either a GATA3 expression vector, pcDNA3.1-GATA3, or empty pcDNA3.1 vector (Invitrogen). In addition, 20 ng of a Renilla luciferase reporter vector, pRL-TK (Promega), was added into each transfection as an internal control for transfection efficiency. The transfected cells were stimulated with 10 ng/ml PMA and 1 M ionomycin (IL-4 and IL-13 reporter assay) or 10 ng/ml PMA and 100 M dibutyryl cAMP (IL-5 reporter assay) 2 h after transfection. 24 h later, cell extracts were prepared and subjected to luciferase assay using a Dual Luciferase Reporter System (Promega) according to the manufacturer's instructions.

Dynamics of Acetylation Status of Histone H3 in IL-4-and IL-13-associated Nucleosomes during Th2
Cell Differentiation-Because histone acetylation is one of the factors affecting chromatin structure, we first assessed the status of histone acetylation on nucleosomes associated with Th2 cytokine genes in developing Th2 cells. The ChIP assay using antibodies specific for acetylated histone H3 (at both lysine residues 9 and 14) was employed, and the relative levels of DNA associated with the acetylated histones were determined by semiquantitative PCR assay using the specific primers shown in Fig. 1. We included promoter regions of each gene and intergenic regions of the IL-4 and IL-13 loci which were described previously as regions containing regulatory elements or DNase I-hypersensitive sites (18 -20, 24).
The results of a ChIP assay of freshly prepared CD4 T cells and CD4 T cells stimulated under Th1-and Th2-skewed conditions for the times indicated are shown ( Fig. 2A). The intensities of bands representing the indicated DNA fragments were measured by densitometry, and the change of relative intensi-ties is shown in Fig. 2B. At all regions tested, histone H3 was hypoacetylated in freshly prepared CD4 T cells. After 1 day of stimulation, low but significant levels of acetylation occurred at all regions tested, and no difference between Th1 and Th2 culture conditions was noted. However, as shown in Fig. 2, A and B, the levels of histone acetylation associated with certain DNA fragments (two IL-4-related, CNS1, and two IL-13-related) were increased significantly in T cells cultured for 2-5 days under Th2-skewed conditions. The acetylation levels were clearly lower in the case of Th1-skewed cultures. Th2-specific histone H3 acetylation in the IL-5 promoter region was also detected but only after culture for 3 days. Two other primer pairs located in the promoter and intron 2 of the IL-5 gene were also examined with similar results (data not shown). Interestingly, histone H3 in RAD50-associated nucleosomes was acetylated equivalently under Th1-and Th2-skewed conditions (Fig.  2, A and B). As expected, increased hyperacetylation of histone H3 on the IFN-␥ promoter was observed in developing Th1 cells but not in Th2 cells. Histone hyperacetylation of the above regions was maintained for at least 9 days after the initial stimulation (data not shown). Stimulation with IL-4 alone resulted in no detectable increase in the levels of acetylation at all region tested (data not shown), suggesting a requirement of TCR-mediated signals. The acetylation status of histone H4 (at the N terminus of all four lysines) revealed the same patterns as those of histone H3 (data not shown).
To assess the levels of histone acetylation more quantitatively, further assays were performed where aliquots of the input DNA obtained were also subjected to PCR using specific primer pairs as before. The PCR product intensities obtained using input DNA were then used to normalize the signals obtained with immunoprecipitated (ChIP assay) DNA. Thus, 3-fold serial dilutions of DNA obtained from cultured T cells for 3 days under Th1-or Th2-skewed conditions were subjected to ChIP assay with input DNA normalization (Fig. 2C). The intensities of PCR bands generated by the indicated primer pairs were measured by densitometry, and the calculated ratios (immunoprecipitates with anti-acetylhistone H3/input DNA) are shown in Fig. 2D. The ratios in IL-13 and IL-4-related gene loci were 1.0 -1.5 and 3-5 times higher, respectively, in the Th2skewed cultures compared with the Th1-skewed cultures. The ratio in the case of the IL-5 promoter in Th2-skewed cultures was about 0.3, clearly lower than that of the IL-13 and IL-4 gene loci. As expected, essentially no difference was observed in RAD50 promoter groups, and an inverted pattern was observed in IFN-␥ promoter groups (Fig. 2D, bottom two groups). These results suggest that Th2-specific histone H3 hyperacetylation occurred on nucleosomes associated with the IL-4-and IL-13related genes and also on IL-5-associated nucleosomes but with delayed kinetics. In contrast, histone hyperacetylation on RAD50 gene-associated nucleosomes was TCR stimulation-dependent but not Th2-specific.
Histone Hyperacetylation on IL-4-and IL-13 Gene-associated Nucleosomes Is Dependent on STAT6 Signaling and GATA3 Expression-To investigate the molecular mechanism underly-ing the induction of Th2-specific histone hyperacetylation in the IL-4 and IL-13 gene-associated nucleosomes, we assessed the role of the STAT6 signaling pathway. CD4 T cells from STAT6-deficient mice failed to differentiate into Th2 cells, suggesting that IL-4-dependent STAT6 activation is required for Th2 cell differentiation (8). Thus, splenic CD4 T cells from STAT6-deficient mice were stimulated under Th2-skewed conditions for 5 days and ChIP assays performed. As shown in Fig.  3A, the acetylation of histone H3 in IL-4, IL-13, and IL-5 gene-associated nucleosomes was dramatically reduced in the STAT6-deficient CD4 T cells compared with wild type CD4 T cells. In contrast, the acetylation in RAD50 gene-associated nucleosomes was induced equivalently to that of wild type. The level of histone H3 acetylation in IFN-␥-associated nucleosomes was low in wild type CD4 T cells under Th2-skewed conditions and was slightly higher in STAT6-deficient T cells (Fig. 3A, bottom).
Next, the role of GATA3 expression on Th2-specific histone hyperacetylation was investigated using a retroviral vector (pMX-IRES-EGFP) encoding GATA3 bicistronically with EGFP (R-GATA3-GFP). GATA3 was introduced into normal CD4 T cells stimulated under Th1-skewed conditions or STAT6-deficient CD4 T cells under Th2-skewed conditions. As reported previously, the introduction of GATA3 into developing Th1 cells results in the generation of IL-4-producing cells and down-regulation of IFN-␥-secreting cell production (Fig. 3B, upper) (15). The infection of STAT6-deficient CD4 T cells with the GATA3 vector also induced significant levels of IL-4-producing cells (Fig. 3B, lower panels).
We sought to determine whether the ectopic expression of GATA3 would induce the Th2-specific histone hyperacetylation of IL-4-and IL-13-associated nucleosomes. Introduction of GATA3 was done as described above, and GFP-positive cells were separated by cell sorting. 500,000 infected cells from each culture were subjected to ChIP assay. Significant increases in histone acetylation in the IL-4-, IL-13-, and IL-5-related nucleosomes were detected in the GATA3-infected group (Fig.  3C). The acetylation status at the RAD50 locus was not affected by infection with the GATA3 vector. Moreover, the introduction of GATA3 into STAT6-deficient T cells resulted in increased acetylation at the IL-4 and IL-13 promoter-associated nucleosomes (Fig. 3D). The acetylation status of the RAD50 locus was not affected. Thus, the ectopic expression of GATA3 did, in fact, induce the Th2-specific histone hyperacetylation.
Long Range Th2-specific Hyperacetylation within the IL-4 and IL-13 Locus-Next, the locations of Th2-specific hyperacetylation within IL-4 and IL-13 loci were analyzed more precisely. A series of primers to amplify ϳ0.4-kbp DNA fragments (ϳ0.5-kbp interval) between the RAD50 and KIF3A genes was prepared. The real patterns of each ChIP assay (Fig.  4, bottom panels) and the ratio of band intensities in CD4 T cells from Th1-and Th2-skewed 3-day cultures (Fig. 4, middle panel) with the 52 selected primer pairs are depicted. The Th2-specific hyperacetylation was observed at least from 1.5 kbp upstream from IL-13 exon 1 (primer pair 10) to 16 kbp downstream from IL-4 exon 4 (primer pairs 45ϳ47). No significant Th2 specificity was observed in the groups with primers 1-9 and 48 -52. These results suggest that almost all histones in the IL-4-and IL-13-associated nucleosomes, including the intergenic nucleosomes, are hyperacetylated selectively under Th2-skewed culture conditions.
Intergenic Transcription Detected throughout the IL-4 and IL-13 Loci-Previous work on the ␤-globin gene cluster demonstrated that intergenic transcription is required for chromatin remodeling of chromosomal domains (32). We used the PCR primer pairs shown in Fig. 4 to probe for the intergenic tran- Splenic CD4 T cells were stimulated under Th1-or Th2-skewed conditions for 1, 2, 3, and 5 days, and the acetylation status of histone H3 in the nucleosomes associated with the indicated DNA regions was assessed by ChIP assay. An antiacetylated histone H3 antibody and specific primer pairs were used. Five independent experiments were performed, and similar results were obtained. B, quantitative representations of the results shown in A. Relative band intensities measured by densitometry are shown. C, acetylation status of histone H3 on Th2 cytokine locus-associated nucleosomes. Splenic CD4 T cells were stimulated under Th1-or Th2-skewed conditions for 3 days, and sonicated cell extracts were prepared as before for ChIP assay. Before immunoprecipitation for ChIP assay, aliquots (ϳ6 ϫ 10 2 cell equivalents) were removed for PCR to determine relative levels of input DNA. To quantify the relative levels of acetylation, 3-fold serial dilution series were made with both the input DNA and immunoprecipitated DNA samples before PCR, and the PCR product intensities were then assessed. Three independent experiments with different T cell preparations were performed with similar results. D, quantitative representations of the result shown in C. The intensities of the bands were measured at the highest concentration by densitometry, and the relative intensities (anti-acetylhistone-precipitated/ input DNA ratio) in each primer pair were calculated. Similar results were obtained by measurement of the intermediate sample concentration bands. scripts in T cells cultured for 2 days under Th1-or Th2-skewed conditions. To digest contaminating DNA, all RNA samples were treated with DNase I. In addition, control PCRs were done with RNA samples that had not been reverse transcribed (RT Ϫ , Fig. 5A). Interestingly, considerable amounts of transcription were detected when we used primers with IL-13, IL-4, and KIF3A intergenic regions (primer pairs 20, 21, 22, IL-4 enhancer, primer pair 32) but not in the intergenic region between the RAD50 and IL-13 gene loci (primer pairs 4 and 5) (Fig. 5A). Clearly, lower levels of transcription were observed in Th1-skewed cultures. To determine whether the induction of the intergenic transcription is STAT6-dependent, STAT6-deficient CD4 T cells were also examined. Significantly decreased levels in intergenic region transcripts were observed (Fig. 5B). We also detected large amounts of IL-4 and IL-13 transcripts in developing Th2 cells (data not shown).
Simultaneously, kinetics of the generation of Th2-specific intergenic transcription was assessed (Fig. 5, C and D). RT-PCR products from CNS1, IL-4 V A enhancer, and ␤-actin were run on the same gel to facilitate the quantification of transcription levels. The relative intensities of the intergenic transcripts (CNS1/␤-actin and IL-4V A enhancer/␤-actin) are shown in Fig.  5D. At the 16-h time point, no significant difference in the levels of intergenic transcripts was detected between Th1 and Th2 culture conditions. However, the levels of transcripts of both CNS1 and IL-4 V A enhancers were increased dramatically thereafter under Th2-skewed conditions (the 32-and 48-h time points in Fig. 5D). Interestingly, the kinetics of the generation of Th2-specific intergenic transcripts was similar to that of Th2-specific histone hyperacetylation of the IL-4 and IL-13 gene loci (Fig. 2B) and time course of the protein expression of GATA3 in developing Th2 cells (data not shown). Identification of a CGRE at 1.6 Kilobase Pairs Upstream from IL-13 Gene Exon 1-To investigate the mechanisms un-

FIG. 3. Ectopic expression of GATA3 induces histone H3 hyperacetylation on IL-4-, IL-13-, and IL-5-associated nucleosomes.
A, acetylation status of histone H3 in STAT6-deficient (STAT6 KO) CD4 T cells. Splenic CD4 T cells from STAT6-deficient mice were cultured under Th2-skewed conditions for 3 days, and ChIP assays were performed as described in Fig. 2. B, ectopic expression of GATA3 induces Th2 cell generation in a STAT6-independent manner. Freshly prepared splenic CD4 T cells were stimulated under the indicated conditions and infected on day 2 with retrovirus encoding GATA3 bicistronically with EGFP. Three days after the infection, the cells were restimulated, and intracellular IFN-␥/IL-4 profiles of electronically gated GFP ϩ populations were determined. The control represents infection with an EGFP-containing retrovirus vector. The percentages of cells present in the each quadrant are shown. C and D, histone H3 hyperacetylation on IL-4, IL-13, and IL-5 gene-associated nucleosomes is induced by ectopic expression of GATA3. Retrovirus-infected CD4 T cells were prepared as described in B. 500,000 GFP ϩ -infected cells were then collected by cell sorting, and the acetylation status of histone H3 at the indicated gene loci was determined by ChIP assay. Three (for C) and two (for D) independent experiments were done with similar results. derlying the GATA3-dependent Th2-specific hyperacetylation of the IL-4 and IL-13 gene loci, we searched genomic DNA sequences upstream from the mouse and human IL-13 gene locus and found an evolutionarily well conserved sequence (Fig.  6A). This conserved 71-bp sequence locates at 1.6 kbp upstream from the IL-13 gene exon 1 and contains four possible binding motifs for GATA and one for cAMP-responsible element-binding protein (CBP). We named this conserved GATA response element (CGRE).
Consequently, we studied the binding of GATA3 protein to CGRE DNA in vitro using biotinylated CGRE oligonucleotides. Nuclear extracts from CD4 T cells stimulated under the Th2skewed conditions for 3 days were incubated with CGRE-absorbed streptavidin beads. Then, the proteins bound to the beads were eluted, separated electrophoretically by SDS-PAGE, and the amount of GATA3 present was assessed by immunoblotting with anti-GATA3 mAb. Substantial amounts of GATA3 protein were precipitated from nuclear extracts of developing Th2 cells compared with a negative control (Fig. 6B, far left; No Oligo DNA). Only a marginal level of GATA3 was detected in extracts from STAT6-deficient CD4 T cells. These results indicate that GATA3 protein from developing Th2 cells binds to CGRE DNA in vitro. The transcriptional coactivator p300 and RNA polymerase II were also precipitated by CGRE DNA-absorbed beads (Fig. 6B, center, lower panels). Significantly low levels of binding of p300 and RNA polymerase II were observed in STAT6-deficient groups. The protein expression levels of GATA3, p300, and polymerase II in the wild type and STAT6-deficient CD4 T cells were confirmed by immunoblotting total nuclear extracts with specific antibodies (Fig. 6B,  right panels). As expected, the levels of GATA3 protein were significantly lower in STAT6-deficient CD4 T cells, and those of p300 or polymerase II were equivalent between wild type and STAT6-deficient T cells.
To assess the requirement of the four GATA consensus motifs in CGRE for GATA3 binding more precisely, CGRE oligonucleotides with a series of mutations were generated, and the efficiency of binding by GATA3 was assessed by pull-down assay (Fig. 6C). Among CGRE oligonucleotides with a single point mutation (M1-M4), significantly decreased binding was observed with M3. The efficiency of binding to other single mutant CGREs was not decreased dramatically, suggesting that GATA site 3 is important for the binding of GATA3. In addition, no effect was observed by mutation of the CBP binding region (M5). Consistent with the results of the above single mutations, decreased binding was observed in one CGRE oligonucleotide with two mutations (M7: mutations of GATA sites 3 and 4) but not in another double mutant (M6: mutations of GATA sites 1 and 2). In addition, the CGRE oligonucleotide with all four mutations failed to bind GATA3 (M8 and M9). Thus, GATA site 3 appears to be most critical for GATA3 binding to CGRE DNA. The levels of p300 and polymerase II protein were also diminished in the precipitates with M3 and M8 oligonucleotides (data not shown).
We performed ChIP assay with a specific mAb against GATA3, to determine whether GATA3 also bound to the CGRE in developing Th2 cells. To detect PCR bands in the appropriate agarose gel electrophoresis, primer pairs covering the 71-bp CGRE region (CGRE260) were used. The CGRE260 was found to be amplified in the anti-GATA3 immunoprecipitates from developing Th2 cells. A large amount of CGRE260 band was detected in the anti-GATA3 immunoprecipitates of wild type mice (Fig. 6D). The levels were reduced dramatically in the STAT6-deficient group (Fig. 6D, far right). Interestingly, essen-   5. Th2-specific, STAT6-dependent generation of intergenic transcripts throughout the IL-4 and IL-13 gene loci. A, Th2-specific intergenic transcripts between the IL-4 and IL-13 gene loci and between the IL-4 and KIF3A gene loci. Freshly prepared CD4 T cells were stimulated under the indicated conditions for 2 days, and total RNA was prepared. RNA samples were treated with RNase-free DNase I to eliminate any possible genomic DNA contamination, reverse transcribed (RT ϩ ), and then subjected to PCR with the indicated primer pairs. RT Ϫ represents PCR without reverse transcription. The numbers of the primer pairs are the same as those used in Fig. 4. B, STAT6-dependent intergenic transcripts. CD4 T cells from STAT6-deficient mice (STAT6KO) were cultured under Th2-skewed conditions for 2 days, and RT-PCR was performed as described in A. Freshly prepared and cultured CD4 T cells under Th2-skewed conditions from wild type B6 mice were also used. C, time course of the generation of Th2-specific intergenic transcripts. CD4 T cells were stimulated under Th1-or Th2-skewed conditions for 16, 32, and 48 h, and the amounts of intergenic transcripts were assessed by semiquantitative RT-PCR using the indicated primers. To normalize the band generated from each transcript, RT-PCR products of ␤-actin were run on the same gel. D, quantitative representations of the results shown in C. The intensities of the bands were measured at the highest concentration by densitometry, and relative intensities (CNS1/␤-actin and IL-4 V A enhancer/␤-actin ratios) are shown. Similar results were obtained from the measurement of bands at the middle concentration of C. tially no DNA fragment of either CNS1 or IL-4V A enhancer regions was amplified in the GATA3 immunoprecipitates of developing Th2 cells, despite the presence of several reasonably good GATA binding sites in these regions (22,24). Also, no binding was revealed in the RAD50 promoter region. Taken together, the CGRE appears to bind GATA3 in developing Th2 cells with a very high degree of efficiency.
The CGRE Exhibits GATA3-dependent Enhancer Activity-To determine whether the CGRE plays a functional role in GATA3-dependent transcription, single copies of the 260-bp fragment spanning CGRE site (CGRE260 WT) and CGRE260 with a single GATA site 3 binding site mutation (CGRE260 M3) were tested in transient transfection assays where they were placed at the 5Ј-end of the IL-4 promoter (Ϫ766), IL-13 promoter (Ϫ254), and IL-5 promoter (Ϫ1200) in a luciferase reporter plasmid. Introduction of the CGRE260 WT fragment significantly enhanced GATA3-dependent transcriptional activity of the IL-4 promoter (ϳ3.8-fold), IL-13 promoter (ϳ2.4fold), and IL-5 promoter (ϳ4.3-fold) (Fig. 6E, upper and middle). Consistent with the decreased GATA3 binding to CGRE (Fig. 6C), the enhancer activity of CGRE260 was decreased significantly by the M3 mutation (Fig. 6E, upper and middle). These results clearly suggest that CGRE site plays a significant functional role in Th2 cytokine gene transcription.

DISCUSSION
In this report, we have identified a CGRE at 1.6 kbp upstream from the IL-13 gene exon 1. The location of CGRE corresponds to the 5Ј-border of a large domain of Th2-specific hyperacetylation of nucleosomes within the IL-13 and IL-4 gene loci. The binding of GATA3 to the CGRE is demonstrated both in vitro and in vivo. In addition, a critical GATA binding motif for the binding of GATA3 was identified. More interestingly, Th2-specific histone hyperacetylation and transcription, including the generation of intergenic transcripts, appear to occur coordinately and proceed downstream from the CGRE site during Th2 cell differentiation.
The molecular basis for chromatin remodeling of the Th2 cytokine gene loci is not well understood. Recent results suggest a causal relationship between hyperacetylation of histones and transcriptionally competent chromatin (33)(34)(35). The large domains of transcriptionally active chromatin containing hyperacetylated histones showed an increased sensitivity to DNase I (30). Ectopic expression of GATA3 in developing Th1 cells or in STAT6-deficient CD4 T cells induced DNase I-hypersensitive sites that were observed only under Th2-skewed culture conditions (16,36). Thus, one might reason that GATA3 is a critical factor for chromatin remodeling of the Th2 cytokine gene loci. In this report, we show that GATA3 is essential for the induction of histone hyperacetylation on IL-13-, IL-4-, and IL-5-related nucleosomes (Fig. 3). GATA3-dependent histone hyperacetylation was observed on nucleosomes between 1.5 kbp upstream from IL-13 exon 1 (primer 10) and 16 kbp downstream from IL-4 exon 4 (Fig. 4). During the search of the DNA sequence around the 5Ј-border of the large domain of Th2specific hyperacetylation nucleosomes, we found a well conserved 71-bp CGRE containing four GATA binding motifs. Agarwal and Rao (18) reported a DNase-hypersensitive site (HS1) at ϳ2.0 kbp upstream from the IL-13 exon 1 (18). The location of CGRE corresponds roughly to the site of HS1. We detected a very high level of histone acetylation on the CGREassociated nucleosomes when GATA3 was overexpressed by retroviral introduction in developing Th2 cells. 2 Thus, the CGRE region appears to be important for GATA3-dependent chromatin remodeling processes, such as histone hyperacetylation.
Additional support for the importance of the CGRE region in GATA3-dependent chromatin remodeling is the finding of efficient binding of CGRE by GATA3 protein. We demonstrate that the 71-bp CGRE is bound very efficiently by GATA3 protein from developing Th2 cells in vitro (Fig. 6B). We also demonstrate one GATA site (GATA site 3 in Fig. 6A) that is critical for the binding by GATA3 (Fig. 6C). Furthermore, GATA3 bound to the CGRE260 region in developing Th2 cells as revealed by the anti-GATA3 ChIP assay (Fig. 6D). In sharp contrast, no detectable GATA3 binding of the CNS1 and IL-4 V A enhancer region was detected in the anti-GATA3 ChIP assay system (Fig. 6D), although CNS1 contains two GATA motifs (22), and the IL-4 V A enhancer region possesses four (24).
The detection of intergenic transcripts of the IL-13, IL-4, and KIF3A gene loci is of interest. These were seen only in the downstream region of CGRE where the Th2-specific GATA3dependent hyperacetylation was induced (Fig. 5A). No transcript was detected upstream from the CGRE site. The induction of the intergenic transcription was found to be STAT6dependent (Fig. 5B). Recent studies suggest that some histone acetyltransferase activity is associated with actively transcribing RNA polymerase II (33,37,38). In fact, we demonstrate the association of RNA polymerase II and p300 with CGRE (Fig.  6B). In addition, the CGRE260 fragment conferred GATA3dependent enhancer activities on a minimal SV40 promoter in transient transfection assays, 2 suggesting a capacity for binding various regulatory factors. The precise molecular mechanisms underlying the coordinate induction of histone hyperacetylation and the intergenic transcription within the IL-13 2 M. Yamashita and T. Nakayama, unpublished observation.
Biotinylated CGRE oligonucleotides were absorbed to streptavidin-agarose beads and then incubated with nuclear extracts. Precipitation with streptavidin-agarose beads without oligonucleotides (No Oligo DNA) was done as a control. Proteins interacting with the CGRE oligonucleotide were precipitated and analyzed by immunoblotting with the indicated mAbs. 3-Fold dilutions of nuclear extracts equivalent to 1, 3, and 10 ϫ 10 6 cells were analyzed. Immunoblotting analysis with total nuclear extracts of CD4 T cells from wild type and STAT6-deficient mice is shown in the right panel. Arbitrary densitometric units are shown under each band. Three independent experiments were performed with similar results. C, requirement of GATA binding sites in CGRE. CD4 T cells were stimulated under Th2-skewed condition for 3 days and nuclear extracts prepared. Biotinylated original CGRE and CGRE oligonucleotides with mutations (M1-M9) were absorbed by streptavidin-agarose beads and then incubated with nuclear extracts. M1 M2, M3, M4, M5, M6, M7, M8, and M9 were used. Bound nuclear extract proteins were then analyzed by immunoblotting with anti-GATA3 mAb. Arbitrary densitometric units are shown under each band. Four independent experiments with titerated doses of CD4 T cells were performed with similar results. D, GATA3 binds to CGRE containing nucleosomes in vivo. CD4 T cells from wild type and STAT6deficient (STAT6KO) mice were stimulated under Th2-skewed condition for 3 days, and the extent of GATA3 binding to nucleosomes associated with the indicated DNA regions was then assessed by ChIP assay. An anti-GATA3 mAb and specific primer pairs (shown in the supplemental figure) were used. To detect PCR bands in an appropriate agarose gel electrophoresis, primer pairs covering the 71-bp CGRE region (CGRE260) were used. Results of input DNA are also shown as a control. Three independent experiments were performed with similar results. E, CGRE acts as a GATA3-dependent transcriptional enhancer. M12 cells were transfected with the indicated promoter reporter constructs along with either pcDNA3.1-GATA3 or pcDNA3.1 negative control plasmid and then stimulated with PMA ϩ ionomycin (IL-4 and IL-13 reporter constructs) or PMA ϩ dibutyryl cAMP (IL-5 promoter) as described under "Experimental Procedures." The GATA3-dependent increase in luciferase activity detected in each promoter reporter construct is shown as the -fold increase (luciferase activity with pcDNA3.1-GATA3/luciferase activity without pcDNA3.1-GATA3). Three individual experiments were done, and the mean values Ϯ S.D. are shown. and IL-4 gene loci are unknown at this time. However, it is possible that GATA3 may associate with a complex containing CBP/p300 having intrinsic histone acetyltransferase activity and RNA polymerase II, and thus be recruited to the CGRE region, resulting in the induction of both transcription and histone hyperacetylation on 3Ј-downstream nucleosomes, including those associated with the IL-13 and IL-4 gene loci. There may be precedence in that GATA1 was reported to be associated with CBP and to induce histone hyperacetylation at the ␤-globin locus (39).
The kinetics of the histone hyperacetylation on IL-5-associated nucleosomes was different from that of IL-13 and IL-4, although the hyperacetylation was also STAT6-and GATA3dependent (Figs. 2 and 3). The level of both IL-5 transcription and protein secretion in developing Th2 cells correlated with the level of acetylation on IL-5-associated nucleosomes. 2 On the other hand, histone hyperacetylation on RAD50-associated nucleosomes was dependent on TCR stimulation, but it was not a Th2-specific consequence (Figs. 2-4). Thus, there appear to be at least three distinct control mechanisms for histone hyperacetylation within the Th2 cytokine gene cluster, one for IL-13 and IL-4, one for IL-5, and another for RAD50.
The Th2 cytokine genes IL-4, IL-5, and IL-13 are localized within a 125-kb region in mouse and human (23). These cytokines are coordinately expressed in Th2 cells, and the existence of a locus control region for Th2 cytokine genes has been proposed (40). However, there are report suggesting that IL-5 expression is controlled by a mechanism distinct from that for IL-4 and IL-13 (40). Furthermore, recent analysis by Flavell and colleagues (40) failed to identify locus control region activity on DNase I-hypersensitive sites using transgenic mice. Further studies on possible locus control region activity focusing on Th2-specific histone hyperacetylation may provide an additional view of the initial process for the coordinate Th2 cytokine expression established during Th2 cell differentiation.
In conclusion, we have identified the CGRE region as a possible functional target site of GATA3 for the induction of Th2-specific histone hyperacetylation of IL-4 and IL-13 gene loci. Th2-specific histone hyperacetylation occurs in a GATA3dependent manner and is accompanied by transcription, suggesting long range gene activation via GATA3-mediated targeting and extensive spreading of core histone acetylation.