Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Distinct Structural Requirements of GATA-3 for the Regulation of Thymocyte and Th2 Cell Differentiation This information is current as of June 16, 2017. Sung-Yun Pai, Bok Yun Kang, Amelia M. Sabadini, Emilio Parisini, Morgan L. Truitt and I-Cheng Ho J Immunol 2008; 180:1050-1059; ; doi: 10.4049/jimmunol.180.2.1050 http://www.jimmunol.org/content/180/2/1050 Subscription Permissions Email Alerts This article cites 29 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/180/2/1050.full#ref-list-1 Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html 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 © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 References The Journal of Immunology Distinct Structural Requirements of GATA-3 for the Regulation of Thymocyte and Th2 Cell Differentiation1 Sung-Yun Pai,*†§ Bok Yun Kang,2¶ Amelia M. Sabadini,* Emilio Parisini,‡§ Morgan L. Truitt,3¶ and I-Cheng Ho4§¶ D ynamic and controlled adaptive immune responses to novel infection or Ag exposure rely on the differentiation of CD4 T cells into distinct Th1 and Th2 effector subsets (1, 2). Each subset elaborates a characteristic set of cytokines while suppressing the production of cytokines typical of the opposite subset. In the case of Th1 cells, IFN-␥ is the canonical Th1 cytokine, whereas Th2 cells suppress the production of IFN-␥ and produce Th2 cytokines such as IL-4, IL-5, and IL-13. Appropriate Th1 and Th2 cell development is required for the clearance of viral, other intracellular, and parasitic infections, whereas imbalance of Th1 and Th2 cytokine production can result in autoimmunity, asthma, and allergy. The polarized cytokine profiles of Th1 and Th2 cells are primarily dictated by the mutually exclusive expression of the “master” Th1 and Th2 transcription factors, T-bet and GATA-3 (3–5). In addition to being required for the Th2 cell development (6, 7), GATA-3 is critical for the development of CD4 single positive (SP)5 thymocytes (8, 9). The Th2 cytokines each have distinct roles in Th function: induction of class switch of B cells in the case *Division of Pediatric Hematology-Oncology, Children’s Hospital, †Department of Pediatric Oncology, ‡Department of Medical Oncology, Dana-Farber Cancer Institute, § Harvard Medical School, and ¶Division of Rheumatology, Allergy, and Immunology, Department of Medicine, Brigham and Women’s Hospital, Boston, MA 02115 Received for publication November 5, 2007. Accepted for publication November 5, 2007. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by Grant R01 AI054451 (to I.-C.H.) and a K08 AI050601 Award (to S.-Y.P.) from the National Institutes of Health and a Charles H. Hood Foundation Child Health Research Grant (to S.-Y.P.). 2 Current address: Chonnam National University, Gwangju, Republic of Korea. 3 Current addresses: Biomedical Sciences Program, University of California, San Francisco, CA 94143. 4 Address correspondence and reprint requests to Dr. I-Cheng Ho, Smith Building, One Jimmy Fund Way, Boston, MA 02115. E-mail address: [email protected] 5 Abbreviations used in this paper: SP, single positive; FF, floxed. Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 www.jimmunol.org of IL-4, stimulation of eosinophils in the case of IL-5, and the induction of airway hyperreactivity in the case of IL-13. Yet, all three cytokines are induced when GATA-3 is overexpressed in developing Th1 cells (6, 14, 20, 23). Indeed on a per cell basis, not every cell makes each cytokine simultaneously (10), and one imagines that it would be advantageous to the organism to modulate the relative levels and local expression of these three cytokines. GATA-3 is a transcription factor, with two N-terminal transactivation domains and two C2-C2 C-terminal zinc fingers (11). The family of GATA transcription factors includes six mammalian members, GATA-1 through GATA-6, which bind a canonical GATA motif (12, 13). Although the zinc finger regions of all six members are highly conserved, the N-terminal regions and distribution of expression are divergent. GATA-1, GATA-2, and GATA-3 are expressed in hemopoietic cells, whereas GATA-4, GATA-5, and GATA-6 are expressed in nonhemopoietic organs such as lung, heart, and intestine. Deletional analysis has shown that the transactivation regions are required for IL-4 production (14, 15), IL-5 production (14), and for suppression of IFN-␥ production (14, 15) when GATA-3 mutants are expressed in developing Th1 cells. The effect of deletion of the N-terminal or Cterminal zinc fingers was somewhat less consistent, with one group showing the N-terminal finger to be dispensable for IL-4 production and another showing both fingers to be required for effects on both IL-4 and IFN-␥ (14, 15). Mutations in the 12 amino acids immediately following the C-terminal zinc finger (residues 343– 354) abrogate all effects of GATA-3 on IL-4, IL-5, IL-13, and IFN-␥, likely by interfering with DNA binding (16). Thus, studies of truncated non-full-length GATA proteins and even of point mutants generally abrogate all GATA-3 functions simultaneously and do not shed light on the biochemical mechanism by which a cell expressing GATA-3 could express one Th2 cytokine but not another. We hypothesized that distinct regions of GATA-3 are responsible for control of individual Th1 and Th2 cytokine expression, and that this level of regulation would occur despite intact binding Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 GATA-3, the only T cell-specific member of the GATA family of transcription factors, is essential for the intrathymic development of CD4ⴙ T cells and for the differentiation of Th2 cells. However, whether distinct biochemical features, unique to GATA-3 compared with other GATA family members, are required to drive T cell transcriptional programs or whether the T cell-specific functions of GATA-3 can simply be ascribed to its expression pattern is unclear. Nor do we understand the protein structural requirements for each individual function of GATA-3. In this study, we report that a heterologous GATA factor, GATA-4, was competent in supporting the development of CD4ⴙ T cells but could not fully compensate for GATA-3 in regulating the expression of Th cytokines. Specifically, GATA-3 was more potent than GATA-4 in driving the production of IL-13 due to a mechanism independent of DNA binding or chromatin remodeling of the IL-13 locus. The difference was mapped to a partially conserved region C-terminal to the second zinc finger. Converting a single proline residue located in this region of GATA-4 to its counterpart, a methionine of GATA-3, was sufficient to enhance the IL-13-promoting function of GATA-4 but had no effect on other cytokines. Taken together, our data demonstrate that the unique function of GATA-3 is conferred by both its cell type-specific expression and distinct protein structure. The Journal of Immunology, 2008, 180: 1050 –1059. The Journal of Immunology Materials and Methods 1200 rad. To obtain wild-type and GATA-3 knockout thymocytes, E17.5 embryos were generated by timed matings between homozygous floxed animals crossed with homozygous floxed/CD4-cre bearing animals on the C57BL/6 background. Fetal thymi were harvested individually and disaggregated in 2 mg/ml collagenase type IV (Worthington) in RPMI 1640 supplemented with HEPES at 37°C. Thymocytes were washed then resuspended in retroviral supernatant containing 40 ng/ml recombinant human IL-7, 100 ng/ml murine stem cell factor, 4 g/ml polybrene. Cells were centrifuged at room temperature for 1.5 h at 2500 rpm, incubated at 37°C for 2– 4 h, washed and cocultured with depleted wild-type lobes in hanging drop format using Terasaki plates for 1 day. Reaggregated lobes were then cultured in transwell format for an additional 3 days. All reaggregates were cultured in heat-sealed bags filled with an 80% oxygen, 5% carbon dioxide, 15% nitrogen gas mixture to optimize survival. Reaggregated lobes were dissociated with collagenase, and cells were analyzed by flow cytometry. The following Abs from BD Biosciences were used: CD4 PE-Cy7, CD8␣ PerCP, TCR- allophycocyanin. Cell purification, in vitro differentiation, and retroviral transduction of Th cells CD4 T cell purification by magnetic separation (Miltenyi Biotec) was performed using anti-murine CD4 beads. In vitro differentiation of T cells with anti-CD3 Ab and anti-CD28 Ab under Th1 skewing (IL-12 and-IL-4 Ab) or Th2 skewing (IL-4 and anti-INF-␥) conditions was performed as described (6). Cells were expanded with recombinant human IL-2 (National Cancer Institute Preclinical Repository) after 48 h. Cells were restimulated on day 6 or 7 with plate-bound anti-CD3 0.1 g/ml and anti-CD28 and expanded with IL-2 for an additional 6 days before intracellular cytokine staining. Retroviral supernatants were generated by cotransfection of retroviral constructs with CMV-env and RSV-gag-pol plasmids, a gift of G. Nolan (Stanford University, Stanford, CA) into 293T cells and collection of supernatants between 24 and 72 h of culture. T cell transduction at 24 –36 h by centrifugation with an equal volume of viral supernatant and polybrene was performed essentially as described (20). Intracellular cytokine staining and ELISA Mice conditionally deficient in GATA-3 and crossed to CD4-cre (9), or generated then crossed to OX40-cre (7), have been described. C57BL/6 mice were obtained from Taconic Farms. Mice were maintained in specific pathogen-free conditions. All experiments were performed in accordance with Institutional Animal Care and Use Committee guidelines and approved protocols at Dana-Farber Cancer Institute (Boston, MA). M12 B cells, EL-4 T lymphoma cells, and primary T cells were maintained in RPMI 1640 medium supplemented with 10% FCS, penicillin-streptomycin, HEPES, sodium pyruvate, nonessential amino acids, L-glutamine, and 50 M 2-ME. Fetal thymocytes and 293T cells were maintained in DMEM supplemented as described. Intracellular cytokine staining was performed after stimulation with PMA, ionomycin, and monensin as described (6, 10) using the following Abs: anti-IL-4 PE (11B11), anti-IL-5 PE (TRFK5), anti-IFN-␥ PE (XMG1.2), biotinylated anti-IL-13, and streptavidin-PerCP. All Abs were purchased from BD Biosciences except for biotinylated anti-IL-13 (R&D Systems). Cells were analyzed by flow cytometry using a FACSCanto (BD Biosciences) and either CellQuest or Flowjo software (Tree Star). Statistical analysis was performed using the two-tailed Student t test. For measurement of secreted cytokines, GFP⫹ cells were sorted on a MoFlo (Beckman Coulter), washed and restimulated with plate-bound antiCD3 and sandwich ELISA was performed for IL-4, IL-5, IFN-␥, and IL-13 as described (6, 21). All Abs were obtained from BD Biosciences except IL-13 (obtained from R&D Systems). Constructs and vectors Luciferase assays Retroviral GATA constructs bearing an N-terminal FLAG epitope were generated by standard cloning techniques using the BamHI site of pcDNA3 FLAG provided by J. M. Leiden (Abbott Laboratories) and bluntend ligated into the GFP-RV retroviral vector obtained from K. Murphy (Washington University, St. Louis, MO). The GATA-3 construct contains aa 2– 443 of murine GATA-3. The GATA-4 construct contains aa 8 – 441 of murine GATA-4. The G3G4 construct fuses aa 2–263 of GATA-3 to aa 217– 441 of GATA-4. The G4G3 construct fuses aa 8 –216 of GATA-4 to 264 – 443 of GATA-3. The G3G4G3 construct fuses aa 2–261 of GATA-3 to aa 212–303 of GATA-4 to aa 351– 443 of GATA-3. The G4G3G4 construct fuses aa 8 –215 of GATA-4 to aa 263–345 of GATA-3 to aa 299 to 442 of GATA-4. The G3⌬365–380⌬G4 construct fuses aa 1–364 of GATA-3 to aa 318 –333 of GATA-4 to aa 381 to 443 of GATA-3. The G3_1–364, G3_1–370, G3_1–380, G3_1–380_3S-A, G3⌬365–380⌬G4, G4_Pro⌬Met constructs were generated by site-specific mutagenesis using the QuikChange XL kit (Stratagene), according to the manufacturer’s protocol. Further details are available upon request. Transfection of cell lines by electroporation and measurement of luciferase activity was performed as described using a 2-kb IL-13 promoter luciferase construct (10). Animals and cells Reaggregate fetal thymic organ culture To obtain thymocyte depleted wild-type stroma, embryonic day 17.5 (E17.5) embryos were generated by timed matings between wild-type males and females on the C57BL/6 background. Fetal thymic lobes were incubated in transwells with 0.4-M pore size over supplemented DMEM with 1.35 mM deoxyguanosine (Sigma-Aldrich) for 5 days. Before reaggregation, lobes were washed then irradiated using a cesium source at Chromatin immunoprecipitation Chromatin immunoprecipitation was performed similar to published protocols (22). Briefly, transduced Th1 cells were sorted for GFP expression and expanded by two rounds of stimulation. Cells were fixed with 1% formaldehyde in medium then quenched with glycine. Fixed cells were washed, swelled in hypotonic buffer then nuclei were lysed in buffer containing 1% SDS. Nuclear lysates were sonicated using a Branson sonifier to generate fragments of ⬃500 bp. Sonicates were precleared with 50% protein G-Sepharose slurry (Amersham Biosciences) previously blocked with BSA and salmon sperm DNA, then incubated with Abs overnight at 4°C. Ab to acetylated histone H3 (Lys9/Lys14; Upstate Biotechnology) and control rabbit IgG (Santa Cruz Biotechnology) was used at 1 g/sample. Immunoprecipitates using protein G-Sepharose preblocked beads were washed and eluted in 1% SDS and 50 mM NaHCO3. After reversal of cross-links, DNA was precipitated, treated with proteinase K, and purified with phenol, chloroform, and isoamyl extraction. The following primer pairs were used: IL-13 promoter, 5⬘-ACCCAGAACCTGGAAACCCT-3⬘ and 5⬘-GTGGCCGCTAAAGGAAAGAGT-3⬘; IL-5 promoter, 5⬘-TTTC CTCAGAGAGAGAATAAATTGCTT-3⬘ and 5⬘-GCTGGCCTTCAGC AAAGG-3⬘; neurofilament promoter, 5⬘-CCACGGCGCTGAAGGA-3⬘ and 5⬘-CTGGTGCATGTTCTGGTCTGA-3⬘; HS V/CNS2, 5⬘-ATCA Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 to DNA because loss of binding would be expected to affect all Th1 and Th2 cytokine expression. To investigate this possibility, we took advantage of the fact that GATA-3 is the only GATA factor expressed in T cells, and surmised that the failure of other GATA members to substitute for GATA-3 function would indicate that structural elements outside those involved in DNA binding are required. This strategy has been used to study the role of GATA-1 in erythroid development. For instance expression of hemopoietic factors GATA-2 or GATA-3 relieves (17) or partially relieves (18) the early erythroid developmental block in GATA-1-deficient embryos, and in one case rescued embryonic lethality. In contrast, the expression of the nonhemopoietic factor GATA-4 in a similar system extended survival but failed to rescue embryonic lethality (19). We found that GATA-4 could substitute for GATA-3 during thymocyte development, and suppressed IFN-␥ production in developing Th1 cells but was much less efficient at promoting Th2 cytokine production. Fusion GATA factors in which domains of GATA-3 and GATA-4 were exchanged revealed that the post-zinc finger region of GATA-3 was specifically responsible for control of IL-13 production, despite equivalent ability to function in luciferase assays and induce the acetylation of histone H3 of the IL-13 locus. A single amino acid in GATA-4, when changed from proline to methionine, was sufficient to render GATA-4 capable of driving IL-13 production but did not affect the production of IL-4, IL-5, or IFN-␥. Thus we found that a previously unrecognized region of GATA-3 is important specifically for IL-13 production. Our findings support the hypothesis that distinct biochemical features of GATA-3 are responsible for modulating the dyscoordinate expression of Th1 and Th2 cytokines. 1051 1052 STRUCTURAL FEATURES OF GATA-3 CONTROLLING IL-13 EXPRESSION CGTCGTCTTACCCAAACA-3⬘ and 5⬘-TGTGGGAGAGCGTCTGA TCTG-3⬘; and Th2 intergenic, 5⬘-GGCAAGCCCCACACTGTT-3⬘ and 5⬘ATGAGACGCACCAGCACAGA-3⬘. All primers were designed using Primer Express software. All PCR were done on a MX3000P machine using Brilliant SYBR Green qPCR Master Mix (Stratagene) and primers yielded a single peak on dissociation curve. Relative DNA amounts were calculated by the expression 2⫺(Ctsample ⫺ CtTh2 intergenic), where Ct is the threshold cycle. Statistical analysis was performed using the two-tailed Student t test. Results A heterologous GATA factor can reconstitute CD4 thymocyte development in GATA-3-deficient thymus GATA-4 does not induce Th2 differentiation as well as GATA-3 The development of Th2 cells from naive Th precursors is driven by the up-regulation of GATA-3, and indeed, developing Th1 cells forced to express GATA-3 are driven into the Th2 lineage (14, 20, 23). To test whether another GATA factor can substitute for the function of GATA-3, we overexpressed GATA-3 or GATA-4 in developing Th1 cells, expanded the cells with a second round of stimulation with anti-CD3/anti-CD28, then tested cytokine FIGURE 1. The nonhemopoietic GATA factor GATA-4 can substitute for GATA-3 during fetal thymic development. A, Expression of GATA-3 in GATA-3-sufficient and -deficient fetal thymus augments CD4 SP thymocyte development. Embryonic day 17.5 thymocytes from wild-type animals carrying two floxed GATA-3 alleles (FF) or GATA-3 deficient animals (FF-CD4cre) were transduced to express GATA-3 using a bicistronic retroviral vector coexpressing GFP (GFP-G3). After 4 days in reaggregate fetal thymic organ culture, the CD4 and CD8 profile of thymocytes with very high expression of TCR were analyzed gating on GFP⫺ and GFP⫹ fractions. The percentage of cells in each quadrant is shown. B, The development of CD4 SP thymocytes depends on the dose of GATA-3. GATA-3-deficient thymocytes transduced with GFP-G3 vector were cultured as in A, and the percentage of CD4 SP and CD8 SP thymocytes in the GFP⫺, GFPlow (GFPlo), and GFPhigh (GFPhi) expressing fractions is graphed with each symbol representing an individual reaggregated thymic lobe. Dot plot on the left depicts the gating strategy. C, Expression of GATA-4- in GATA-3-sufficient and -deficient fetal thymus has effects similar to expression of GATA-3. FF and FF-CD4-cre thymocytes were transduced to express GATA-4 using a bicistronic retroviral vector coexpressing GFP (GFP-G4) and analyzed as described in A. D, Expression of GATA proteins in GATA-3-deficient fetal thymocytes results in augmentation of CD4 SP percentage, diminution of CD8 SP percentage, and increase in the CD4 to CD8 ratio. CD4 SP and CD8 SP percentages (left) in the GFP⫺ gate and GFP⫹ gate of GATA-3-deficient thymocytes transduced with GFP-G4 vector are shown. The CD4 to CD8 ratio (right) in the GFP⫺ and GFP⫹ fractions of GATA-3-deficient thymocytes transduced with empty vector (RV), GFP-G3 (G3), or GFP-G4 (G4). Each symbol represents an individual reaggregated thymic lobe. Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 Mice conditionally deficient in GATA-3 in thymocytes using the Cre-lox system are profoundly deficient in CD4 SP thymocytes, demonstrating the critical role of GATA-3 in the transcriptional control of CD4 SP development (9). To test whether another GATA factor can substitute for GATA-3 during thymocyte development, we overexpressed full-length murine GATA-3 and GATA-4 in developing GATA-3-sufficient and -deficient thymocytes cultured in reaggregate fetal thymic organ culture. Because the hemopoietic factors GATA-1 and GATA-2 are more similar to GATA-3, we chose the nonhemopoietic factor GATA-4 to take optimal advantage of the differences between the nonshared domains. Bicistronic retroviral vectors coexpressing GFP with either Nterminal FLAG-tagged murine GATA-3 or GATA-4 were created. Thymocytes from E17.5 fetuses bearing two “floxed” (FF) GATA-3 alleles or floxed animals that express Cre under the control of the CD4 promoter at the double positive stage of thymocyte development (FF-CD4cre) were transduced with these retroviruses and forced to express GATA-3 or GATA-4. In Fig. 1A, we show that untransduced (GFP⫺) FF fetal thymocytes have normal CD4 and CD8 SP development after 4 days in culture, whereas untransduced (GFP⫺) FF-CD4cre knockout fetal thymocytes have poor CD4 SP development and reversal of the CD4 to CD8 ratio, similar to adult animals (9). As expected, GFP⫹ thymocytes expressing GATA-3 in FF-CD4cre animals show restoration of CD4 SP development (Fig. 1A, right) and forced expression of GATA-3 in FF thymocytes showed a further increase in the CD4 SP percentage compared with GFP⫺ thymocytes in the same culture (54.1 vs 43.7%) consistent with published results (8). Indeed the development of CD4 SP cells appeared to depend on the level of GATA-3 because there was a progressive increase in CD4 SP percentage and decrease in CD8 SP percentage when GFP⫺, GFPlow, and GFPhigh cells were analyzed and compared (Fig. 1B). Forced expression of GATA-4 in GATA-3-sufficient FF and GATA-3-deficient FF-CD4cre fetal thymocytes showed restoration of CD4 SP development (Fig. 1C), similar to the effect of expressing GATA-3. Thus the low CD4 SP percentage seen in knockout FF-CD4cre GFP⫺ thymocytes is dramatically increased and the high CD8 SP percentage decreased in GFP⫹ GATA-4expressing thymocytes (Fig. 1D). This response is reflected in the CD4 to CD8 ratio increasing from 0.1– 0.2 in knockout GFP⫺ thymocytes to 3 in GFP⫹ thymocytes when either GATA-3 or GATA-4 is expressed, whereas expression of empty vector has no effect (Fig. 1D). Thus, GATA-4 can substitute for the function of GATA-3 during CD4 SP development. The Journal of Immunology 1053 production by intracellular cytokine staining after treatment with PMA and ionomycin (Fig. 2A). As shown in Fig. 2, B and C, Th1 cells infected with empty vector express little IL-4, IL-5, and IL-13 and express high levels of IFN-␥. Forced expression of GATA-3 results in high levels of IL-4, IL-5, and IL-13 and repression of IFN-␥ production. In contrast, forced expression of GATA-4 resulted in some increase in IL-4 and equivalent suppression of IFN-␥, but very little increase in IL-5 or in IL-13 production. The failure to drive IL-5 and IL-13 production was not due to poor protein expression; indeed Western blot using anti-FLAG Ab consistently showed much higher expression of GATA-4 than GATA-3 when equal amounts of cells were loaded (Fig. 2D). Thus GATA-4 does not function equivalently to GATA-3 in Th2 cytokine production. IL-13 production requires the C-terminal half of GATA-3 The inability of GATA-4 to drive IL-13 production is not due to failure to activate the IL-13 promoter The cell type-specific expression of IL-13 has been localized to the proximal promoter, which contains a critical GATA site (10). If GATA-4 was unable to bind to or activate the IL-13 promoter, this could explain our results in Fig. 2. A construct in which a 2-kb FIGURE 2. GATA-4 does not fully recapitulate the effects of GATA-3 on Th1 and Th2 cytokine production. A, Schema depicts the experimental culture conditions. Magnetically purified bulk murine CD4 Th precursor (Thp) cells were stimulated with Abs against CD3 and CD28, under Th1-polarizing conditions. Cells were transduced with bicistronic retroviral vectors coexpressing GATA proteins with GFP. Cells were cultured for 6 days, restimulated with Abs to CD3 and CD28, and cultured an additional 6 days. Cells were then stimulated with phorbol ester (PMA) and ionomycin, and intracellular cytokine staining was performed. For some experiments, GFP⫹ cells were sorted at day 5 or 6 of culture. B, GATA-4 induces little IL-5 and IL-13 expression compared with GATA-3, and the ability to induce IL-13 depends on the presence of the C-terminal half of GATA-3. CD4 T cells were transduced with empty vector (RV), or vectors expressing GATA-3 (G3), GATA-4 (G4), or fusion constructs of GATA-3 and GATA-4 (G3G4 and G4G3) and cultured as described in A. Portions derived from GATA-3 (black line) and from GATA-4 (open line) are depicted on the left, and the curved regions represent the N-terminal and C-terminal zinc fingers. The percentage of cells producing IL-4, IL-5, IL-13, and IFN-␥ in the GFP⫹ gate is shown. C, Summary of cytokine production by G3 and GATA constructs. For each experiment, the amount of cytokine production was calculated as a percentage of the production of control cells from that experiment, cells expressing GATA-3 in the case of IL-4, IL-5, and IL-13, and or cells transduced with empty vector (RV) in the case of IFN-␥. Data shown are average percentage and SD of three to nine experiments. D, Expression of GATA constructs. Developing Th1 cells transduced with empty vector or FLAG-tagged G3, G4, G3G4, and G4G3 vectors were sorted, and an equal number of GFP⫹ cells were subjected to Western blot analysis using anti-FLAG Ab. Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 To determine whether a particular structural domain is responsible for the difference between GATA-3 and GATA-4 in their ability to drive IL-13 production, constructs fusing the N-terminal half of GATA-3 with the C-terminal zinc finger containing regions of GATA-4 (G3G4) and vice versa (G4G3) were created. By Western blot using anti-FLAG Ab, we consistently saw that the G4G3 construct was more poorly expressed, and in some blots was barely visible (Fig. 2D and data not shown). Indeed, the G4G3 construct was the least efficient at driving IL-4 production, did not significantly drive IL-5 production, and failed to suppress IFN-␥ production (Fig. 2, B and C). However, despite the lower protein level, a high percentage of cells expressing the G4G3 construct did produce IL-13 (Fig. 2C). Thus GATA-4, unlike GATA-3, is unable to induce IL-13 production in developing Th1 cells, and IL-13 production is restored with a fusion construct containing the C-terminal half of GATA-3. 1054 STRUCTURAL FEATURES OF GATA-3 CONTROLLING IL-13 EXPRESSION fragment of the IL-13 promoter drives luciferase activity (10) was coexpressed with GATA-3 or GATA-4 in the EL-4 T lymphoma line and also in the M12 B cell line. In both lines, expression of GATA-4 and other GATA fusion constructs equally induced luciferase activity by the IL-13 promoter (Fig. 3A and data not shown). As the binding of transcription factors to promoters driving luciferase is artificial, we also examined the activity of the IL-13 promoter in the context of native chromatin. Transcriptionally active gene loci generally are characterized by the acetylation of histone H3 at the Lys9 and Lys14 positions, and acetylation of histone H3 can be detected at Th2 cytokine regulatory regions in both Th2 cells and in Th1 cells forced to express GATA-3 (24 – 26). Thus we examined the acetylation status of Lys9 and Lys14 of FIGURE 4. Zinc finger regions of GATA-3 are not sufficient to drive IL-13 production. A, IL-13 production by intracellular cytokine staining in Th1 cells cultured for 6 days, restimulated with Abs to CD3 and CD28, and cultured an additional 6 days, as in Fig. 2A with the constructs empty vector or constructs expressing full-length GATA-3 (G3), full-length GATA-4 (G4), or fusion constructs (G3G4, G4G3, G3G4G3, G4G3G4) depicted. The percentage of cells producing IL-13 in the GFP⫹ gate is indicated. The G3G4G3 construct was created such that the two zinc fingers and intervening amino acids of GATA-3 have been replaced with the corresponding section of GATA-4 (open white line). The G4G3G4 construct is full-length GATA-4 with the two zinc fingers and intervening amino acids replaced with the corresponding section of GATA-3 (black line). B, GFP⫹ cells from A were sorted and restimulated, and IL-13 was measured in the supernatants by ELISA. histone H3 by chromatin immunoprecipitation in sorted Th1 cells forced to express empty vector, GATA-3, or GATA-4. As expected, we consistently detected increased acetylation of histone H3 in critical regulatory regions of the Th2 cytokine genes, including the CNS2 (or HS V) region 3⬘ of the IL-4 gene, the IL-5 promoter and the IL-13 promoter in Th1 cells expressing GATA-3 compared with empty vector. We also saw equivalent acetylation Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 FIGURE 3. GATA-3 and GATA-4 function similarly in luciferase assays and induce Th2-specific histone H3 acetylation patterns. A, EL-4 cells were transfected with empty vector (RV) or constructs expressing fulllength GATA-3 (G3), full-length GATA-4 (G4), or fusion constructs (G3G4 and G4G3), as shown in Fig. 2B, as well as a construct in which the 2-kb IL-13 promoter was cloned upstream of the firefly luciferase gene. The luciferase activity of cells stimulated with PMA and ionomycin is represented as fold change over empty vector. B, Developing Th1 cells were transduced with empty vector or G3 or G4 vectors as described in Fig. 2B. GFP⫹ cells were sorted and expanded cells were subjected to chromatin immunoprecipitation using Ab to Lys9/Lys14 acetylated histone H3 (anti-AcH3) or rabbit IgG control (IgG). Precipitated DNA was amplified using primers to the CNS 2/HS V region of the IL-4 gene, the IL-13 promoter, or the IL-5 promoter. Amplification of a gene silent in T cells, the neurofilament (Nfm) promoter, was used as a control. All signals were normalized to the signal detected using primers to a Th2 intergenic region to control for input, and shown with arbitrary units. The average and SD of three experiments is shown. In all cases the differences between the anti-acetylated histone H3 signal in cells expressing G3 or G4 was not statistically significant (p ⬎ 0.05, determined by Student’s two-tailed t test). The Journal of Immunology 1055 Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 FIGURE 5. The terminal 73 amino acids of GATA-3 are dispensable, and residues 365–370 (NRKMSS) are required for all effect of GATA-3 on cytokine production. A, The post-zinc finger tail of GATA-3 (G3) and of deletion mutants are shown. The six amino acids NRKMSS from positions 365–370 are underlined. B, IL-13 production by Th1 cells expressing GATA-3 deletion mutants. The percentage of Th1 cells expressing empty vector (RV), full-length GATA-3 (G3), or deletion mutants, gating on GFP⫹ cells is shown. Cells were cultured for 6 days, restimulated with Abs to CD3 and CD28, and cultured an additional 6 days, as in Fig. 2A. C, GFP⫹ cells from B were sorted, restimulated, and supernatants analyzed for IL-13 production. D, Summary of cytokine production by GATA-3 and deletion mutants. Experiments were normalized as in Fig. 2C. The average and SDs of two or three experiments is shown. of histone H3 in Th1 cells expressing GATA-4 with no significant differences when compared with cells expressing GATA-3 ( p ⬎ 0.05). We saw no binding above IgG control using primers to a neural promoter expected to be silent in T cells (neurofilament promoter) (Fig. 3B). Thus, GATA-4 functions in luciferase assays and induces accessibility of the IL-13 promoter in T cells, yet is unable to induce IL-13 expression, in contrast to GATA-3. The ability to induce IL-13 expression localizes to the post-zinc finger tail of GATA-3, and six amino acids are critical for overall GATA-3 function The zinc fingers of GATA-3 and GATA-4, though very similar, are not identical, whereas the post-zinc finger tails of GATA-3 and GATA-4 are quite divergent. Either region could be responsible 1056 STRUCTURAL FEATURES OF GATA-3 CONTROLLING IL-13 EXPRESSION Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 FIGURE 6. Substitution of serine residues or of the 365–380 residue region with the corresponding region of GATA-4 does not affect cytokine production. A, The post-zinc finger regions of full length GATA-3 (G3), truncated GATA-3 (G3_1–380), truncated GATA-3 with serine mutations (G3_1–380_3S-A), and GATA-3 loss of function mutant (G3⌬365–380⌬G4) are shown. B, IL-13 production by Th1 cells expressing GATA-3 mutants. The percentage of Th1 cells expressing empty vector (RV), GATA-3 (G3), or mutants is shown as gating on GFP⫹ cells cultured for 6 days, restimulated with Abs to CD3 and CD28, and cultured an additional 6 days, as in Fig. 2A. The three asterisks represent serines 369, 370, and 372 that were mutated to alanine. C, IL-13 production in Th1 cells expressing G3⌬365–380⌬G4. Loops represent the two zinc finger regions. D, Summary of cytokine production by G3⌬365–380⌬G4. Experiments were normalized as in Fig. 2C, with the amount of IL-13 production calculated as a percentage of the production of control cells from that experiment, cells expressing GATA-3 in the case of IL-4, IL-5, and IL-13, and or cells transduced with empty vector (RV) in the case of IFN-␥. The average and SD shown is from seven experiments. E, IL-13 production is shown in GATA-3-deficient Th1 cells incapable of autoactivation expressing G3⌬365–380⌬G4. GATA-3-deficient Th cells were manipulated as in Fig. 2A. The percentage of IL-13-producing cells in the GFP⫹ gate is indicated. Portions of constructs derived from GATA-3 (black line) and GATA-4 (open line) are depicted. Loops represent the two zinc finger regions. for the observed difference in the ability to induce IL-13 expression. We therefore created constructs in which the zinc fingers of GATA-3 are substituted with those of GATA-4 (G3G4G3) and vice versa (G4G3G4). IL-13 production by both intracellular cytokine staining (Fig. 4A) and by ELISA of sorted GFP⫹ cells (Fig. 4B) showed high expression of IL-13 by the G3G4G3 construct The Journal of Immunology 1057 and poor expression by the G4G3G4 construct, arguing that the post-zinc finger tail of GATA-3 is the region responsible for driving IL-13 production. The 102 amino acid post-zinc finger tail of GATA-3 (aa 342– 443) has little structural similarity to GATA-4, particularly past the first 30 – 40 amino acids, precluding further obvious domain swaps. We thus constructed deletion mutants to determine the region of this tail critical for IL-13 production (Fig. 5A). Deletion of the terminal 54 (G3_1–380) or 64 amino acids (G3_1–370) had little effect on Th2 cytokine production or repression of IFN-␥ as shown by cytokine staining (Fig. 5, B and D) or by ELISA (Fig. 5C), a result consistent with published reports (16). However, loss of an additional six amino acids, NRKMSS (G3_1–364), resulted in loss of all function with poor Th2 cytokine production and failure to repress IFN-␥ (Fig. 5, B and D). Three of the amino acids in this six amino acid region are conserved in murine GATA-1 and murine GATA-2 (R, K, and S, underlined in Fig. 5A) and have been shown to contact DNA directly according to the NMR structure of chicken GATA-1 (27). Thus we show that GATA-3 cannot function in Th2 differentiation without these six amino acids, whereas the remainder of the post-zinc finger tail is dispensable. We then focused on differences in the post-zinc finger regions of GATA-3 and GATA-4 to understand how this region of GATA-3 specifically controls IL-13 expression. Alignment of the first 38 Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 FIGURE 7. Substitution of proline residue of GATA-4 with methionine results in gain of function exclusively for production of IL-13. A, The post-zinc finger regions of GATA-3 (G3), GATA-4 (G4), and proline to methionine mutant (G4_Pro⌬Met) are shown. The proline to methionine change is enlarged and in bold. Residues 365–380 in GATA-3 (NRKMSS) and the corresponding region in GATA-4 residues 318 –323 (KRKPKN) are underlined. B, Mutation of proline to methionine in GATA-4 is sufficient to induce IL-13 production. The percentage of Th1 cells expressing empty vector (RV), GATA-3, GATA-4, or proline to methionine mutant is shown, gating on GFP⫹ cells cultured as in Fig. 2A. C, Summary of cytokine production by mutation of proline to methionine in GATA-4 (G4_Pro⌬Met). Experiments were normalized as described in Fig. 2C. The average and SD of three to nine experiments is shown. D, Expression of proline to methionine mutant is similar to GATA-4. Western blot shows expression of the indicated proteins in equal numbers of sorted GFP⫹ Th1 cells using anti-FLAG Ab. 1058 STRUCTURAL FEATURES OF GATA-3 CONTROLLING IL-13 EXPRESSION A single mutation of proline to methionine in GATA-4 renders GATA-4 capable of inducing IL-13 expression The NMR structure of a fragment of chicken GATA-1 shows that the region between the end of the zinc fingers and NRKMSS is closely apposed to DNA (27); the structure of the remainder of the tail is unknown. By comparing all six GATA factors, we speculated that the presence of proline, a rigid amino acid, in GATA-4, GATA-5, and GATA-6 instead of a nonrigid amino acid residue, such as methionine (of GATA-2 and GATA-3) or alanine (of GATA-1), at the corresponding position, could significantly change the tertiary structure of the protein in that region and render GATA-4 incapable of functioning like GATA-3 (Fig. 7A). We hypothesized that substitution of the proline in full-length GATA-4 with methionine could be sufficient to drive IL-13 expression. Indeed this mutant (G4_Pro⌬Met) was capable of inducing IL-13 production ( p ⫽ 0.02 compared with GATA-4) (Fig. 7, B and C). Interestingly, this mutation did not improve the ability of GATA-4 to drive IL-4 or IL-5 production ( p ⫽ 0.72 or 0.16, respectively), and in fact was less efficient at suppressing IFN-␥ production ( p ⫽ 0.03), despite equivalent protein expression by Western blot (Fig. 7, C and D). Thus not only did this single point mutation render GATA-4 capable of inducing IL-13 expression, the effect was specific for IL-13 and did not extend to effects on IL-4, IL-5, or IFN-␥. Discussion GATA-3 is critical for both intrathymic CD4 T cell development and for Th2 differentiation. To gain insight into the molecular mechanism of how GATA-3 controls these two processes, we took advantage of the structural differences between GATA-3 and another GATA family member, GATA-4, which is not expressed in hemopoietic tissues, to define which regions of GATA-3 are responsible for these two effects. In T cells, aside from deletional analysis, little is known about the structural requirements for Th2 cytokine production (14, 15). One study concluded that GATA-4 functioned similarly to GATA-3 in its effects on IL-4, IL-5 and IFN-␥ production, but IL-13 was not examined (15). Heterologous factors have not been used to reconstitute intrathymic CD4 thymocyte development to our knowledge. We found that expressing GATA-4 in GATA-3 deficient double positive fetal thymocytes successfully reconstituted intrathymic CD4 T cell development, with normalization of the CD4 to CD8 ratio (Fig. 1). Although it was technically not feasible to check the level of expression of GATA-4 in transduced fetal thymocytes, we suspect that the level of expression of both retroviral GATA-3 and GATA-4 is above the endogenous level of GATA-3 because FF cells transduced with either also had an increase in CD4 to CD8 ratio as published previously (8). Various mice engineered to express 5, 20, or 50% of GATA-1 or GATA-3 in GATA-1-deficient animals have suggested that a threshold level of GATA-1 is necessary to drive the early embryonic erythroid transcriptional program (18, 28, 29). Our data showing that there is a progressive rise in CD4 SP percentage from GFP⫺ to GFPlow to GFPhigh populations (Fig. 1B) parallel these findings in early red cell development and underscore the importance of tight control over GATA-3 protein levels during thymocyte development. In contrast to intrathymic CD4 T cell development, the ability of GATA-3 when overexpressed to drive the production of Th2 cytokines such as IL-4, IL-5, and IL-13 and suppress the production of the Th1 cytokine IFN-␥ in mature CD4 T cells is not fully substituted by GATA-4 (Fig. 2). In particular we found that despite an abundance of GATA-4 by Western blot, IL-4, IL-5, and IL-13 production were all much lower in cells expressing GATA-4 than in cells expressing GATA-3, whereas the suppression of IFN-␥ was equivalent. Moreover, we found that a construct with the Cterminal half of GATA-3 fused to the N-terminal half of GATA-4, despite low expression by Western blot, specifically reconstituted the ability to drive IL-13 production (Fig. 2). These data strongly suggest that during mature T cell differentiation, distinct structural or biochemical features of GATA-3 not present in GATA-4 are required for each cytokine gene, and cannot be compensated by supraphysiologic GATA protein levels. Mechanistically, either factor can activate the IL-13 promoter when upstream of luciferase, and can promote accessibility of the promoter and other Th2 responsive sites in primary T cells as measured by binding of acetylated histone H3 (Fig. 3), and thus our findings are not likely explained by differences in DNA binding or inability to induce an open chromatin conformation. We hypothesized that amino acid differences between GATA-3 and GATA-4 are responsible for the action of GATA-3 in IL-13 production. Indeed, further domain swaps and deletional analysis localized the region of interest to the area C-terminal to the second zinc finger, a region previously not known to have functional significance. Because we found that 20 –30 amino acids immediately following the zinc fingers are particularly critical for all cytokinerelated effects of GATA-3, whereas the remaining 73 amino acids are dispensable when deleted (Fig. 5), we focused on the differences between the hemopoietic and nonhemopoietic GATA factors in this region. A single amino acid change in GATA-4, substituting a proline common to GATA-4, GATA-5, and GATA-6, to methionine (residue 368 of GATA-3), was sufficient to restore IL-13 production specifically, without affecting IL-4 or IL-5. Single amino acid changes upstream, in regions required for binding to DNA, have been shown to abrogate all Th2 cytokine production (16); to our knowledge this study is the only to report a single Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 amino acids of the post-zinc finger tail revealed that Ser369 is conserved in the hemopoietic GATA factors, GATA-1, GATA-2, and GATA-3, whereas Ser370 and Ser372 are conserved in GATA-2 and GATA-3. None of the serines are present in GATA-4, GATA-5, or GATA-6. We hypothesized that substitution of the serines with alanine would result in the inability to drive IL-13 production. However, such a mutant functioned equivalently to native GATA-3 (Fig. 6, A and B). The six amino acid NRKMSS region of GATA-3 (residues 365– 370) corresponds to amino acids KRKPKN in GATA-4 (residues 318 –323). Given the critical importance of this region for overall GATA-3 function, we tested whether substituting a 16-amino acid region containing NRKMSS with the corresponding region of GATA-4 would cause GATA-3 to lose function. As shown in Fig. 6, C and D, this mutant (G3⌬365–380⌬G4) functioned equivalently to GATA-3 in all respects. Forced expression of GATA-3 or GATA-4 by retrovirus has been shown to induce the expression of endogenous GATA-3 (15, 23). We speculated that if the G3⌬365– 380⌬G4 mutant was able to induce sufficient levels of endogenous GATA-3 by this autoactivation mechanism, IL-13 production could occur despite the mutant itself being nonfunctional. To test this hypothesis, we took advantage of mice conditionally deficient in GATA-3 in which floxed animals are crossed with OX-40-cre (7). Because endogenous GATA-3 is not expressed (7 and data not shown), autoactivation should not occur in this setting and the function of the mutant can be tested in isolation. CD4 T cells from these mice were differentiated under Th1-polarizing conditions and forced to express empty vector, GATA-3 or G3⌬365–380⌬G4 mutant. As shown in Fig. 6E, the mutant induced IL-13 production equivalently to overexpressed native GATA-3. The Journal of Immunology Acknowledgments We thank Dr. William Paul for providing the GATA-3 FF-OX40-cre mice and Dr. Roland Grenningloh for critical review of the manuscript. We also thank Dr. Michael Pazin and Dr. Andrea Wurster for sharing primers and giving technical advice for chromatin immunoprecipitation. Disclosures The authors have no financial conflict of interest. References 1. Mosmann, T. R., and R. L. Coffman. 1989. TH1 and TH2 cells: different patterns of lymphokine secretion lead to different functional properties. Annu. Rev. Immunol. 7: 145–173. 2. Paul, W. E., and R. A. Seder. 1994. Lymphocyte responses and cytokines. Cell 76: 241–251. 3. Szabo, S. J., S. T. Kim, G. L. Costa, X. Zhang, C. G. Fathman, and L. H. Glimcher. 2000. A novel transcription factor, T-bet, directs Th1 lineage commitment. Cell 100: 655– 669. 4. Zhang, D. H., L. Cohn, P. Ray, K. Bottomly, and A. Ray. 1997. Transcription factor Gata-3 is differentially expressed murine Th1 and Th2 cells and controls Th2-specific expression of the interleukin-5 gene. J. Biol. Chem. 272: 21597–21603. 5. Zheng, W. P., and R. A. Flavell. 1997. The transcription factor Gata-3 is necessary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 89: 587–596. 6. 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. USA 101: 1993–1998. 7. Zhu, J., B. Min, J. Hu-Li, C. Watson, A. Grinberg, Q. Wang, N. Killeen, J. Urban, L. Guo, and W. Paul. 2004. Conditional deletion of Gata3 shows its essential function in TH1-TH2 responses. Nat. Immunol. 5: 1157–1165. 8. Hernández-Hoyos, G., M. K. Anderson, C. Wang, E. V. Rothenberg, and J. Alberola-Ila. 2003. GATA-3 expression is controlled by TCR signals and regulates CD4/CD8 differentiation. Immunity 19: 83–94. 9. Pai, S. Y., M. L. Truitt, C. N. Ting, J. M. Leiden, L. H. Glimcher, and I. C. Ho. 2003. Critical roles for transcription factor GATA-3 in thymocyte development. Immunity 19: 863– 875. 10. Kishikawa, H., J. Sun, A. Choi, S. Miaw, and I. Ho. 2001. The cell type-specific expression of the murine IL-13 gene is regulated by GATA-3. J. Immunol. 167: 4414 – 4420. 11. Ho, I. C., P. Vorhees, N. Marin, B. K. Oakley, S. F. Tsai, S. H. Orkin, and J. M. Leiden. 1991. Human GATA-3: a lineage-restricted transcription factor that regulates the expression of the T cell receptor ␣ gene. EMBO J. 10: 1187–1192. 12. Ko, L., and J. Engel. 1993. DNA-binding specificities of the GATA transcription factor family. Mol. Cell Biol. 13: 4011– 4022. 13. Merika, M., and S. H. Orkin. 1993. DNA-binding specificity of GATA family transcription factors. Mol. Cell. Biol. 13: 3999 – 4010. 14. Lee, H. J., N. Takemoto, H. Kurata, Y. Kamogawa, S. Miyatake, A. O’Garra, and N. Arai. 2000. GATA-3 induces T helper cell type 2 (Th2) cytokine expression and chromatin remodeling in committed Th1 cells. J. Exp. Med. 192: 105–115. 15. Ranganath, S., and K. M. Murphy. 2001. Structure and specificity of GATA proteins in Th2 development. Mol. Cell. Biol. 21: 2716 –2725. 16. Shinnakasu, R., M. Yamashita, K. Shinoda, Y. Endo, H. Hosokawa, A. Hasegawa, S. Ikemizu, and T. Nakayama. 2006. Critical YxKxHxxxRP motif in the C-terminal region of GATA3 for its DNA binding and function. J. Immunol. 177: 5801–5810. 17. Takahashi, S., R. Shimizu, N. Suwabe, T. Kuroha, K. Yoh, J. Ohta, S. Nishimura, K. Lim, J. Engel, and M. Yamamoto. 2000. GATA factor transgenes under GATA-1 locus control rescue germline GATA-1 mutant deficiencies. Blood 96: 910 –916. 18. Tsai, F., C. Browne, and S. Orkin. 1998. Knock-in mutation of transcription factor GATA-3 into the GATA-1 locus: partial rescue of GATA-1 loss of function in erythroid cells. Dev. Biol. 196: 218 –227. 19. Hosoya-Ohmura, S., N. Mochizuki, M. Suzuki, O. Ohneda, K. Ohneda, and M. Yamamoto. 2006. GATA-4 incompletely substitutes for GATA-1 in promoting both primitive and definitive erythropoiesis in vivo. J. Biol. Chem. 281: 32820 –32830. 20. Ouyang, W., S. H. Ranganath, K. Weindel, D. Bhattacharya, T. L. Murphy, W. C. Sha, and K. M. Murphy. 1998. Inhibition of Th1 development mediated by GATA-3 through an IL-4-independent mechanism. Immunity 9: 745–755. 21. Grenningloh, R., B. Kang, and I. Ho. 2005. Ets-1, a functional cofactor of T-bet, is essential for Th1 inflammatory responses. J. Exp. Med. 201: 615– 626. 22. Boyd, K. E., J. Wells, J. Gutman, S. M. Bartley, and P. J. Farnham. 1998. c-Myc target gene specificity is determined by a post-DNA binding mechanism. Proc. Natl. Acad. Sci. USA 95: 13887–13892. 23. Ouyang, W., M. Lohning, Z. Gao, M. Assenmacher, S. Ranganth, A. Radbruch, and M. M. Murphy. 2000. Stat6-independent GATA-3 autoactivation directs IL4-independent Th2 development and commitment. Immunity 12: 27–37. 24. Avni, O., D. Lee, F. Macian, S. Szabo, L. Glimcher, and A. Rao. 2002. TH cell differentiation is accompanied by dynamic changes in histone acetylation of cytokine genes. Nat. Immunol. 3: 643– 651. 25. Fields, P., S. Kim, and R. Flavell. 2002. Cutting edge: changes in histone acetylation at the IL-4 and IFN-␥ loci accompany Th1/Th2 differentiation. J. Immunol. 169: 647– 650. 26. Yamashita, M., M. Ukai-Tadenuma, M. Kimura, M. Omori, M. Inami, M. Taniguchi, and T. Nakayama. 2002. Identification of a conserved GATA3 response element upstream proximal from the interleukin-13 gene locus. J. Biol. Chem. 277: 42399 – 42408. 27. Omichinski, J., G. Clore, O. Schaad, G. Felsenfeld, C. Trainor, E. Appella, S. Stahl, and A. Gronenborn. 1993. NMR structure of a specific DNA complex of Zn-containing DNA binding domain of GATA-1. Science 261: 438 – 446. 28. McDevitt, M., R. Shivdasani, Y. Fujiwara, H. Yang, and S. Orkin. 1997. A “knockdown” mutation created by cis-element gene targeting reveals the dependence of erythroid cell maturation on the level of transcription factor GATA-1. Proc. Natl. Acad. Sci. USA 94: 6781– 6785. 29. Takahashi, S., K. Onodera, H. Motohashi, N. Suwabe, N. Hayashi, N. Yanai, Y. Nabesima, and M. Yamamoto. 1997. Arrest in primitive erythroid cell development caused by promoter-specific disruption of the GATA-1 gene. J. Biol. Chem. 272: 12611–12615. Downloaded from http://www.jimmunol.org/ by guest on June 16, 2017 amino acid change having specific effects on one cytokine. Although this amino acid is sufficient, it does not appear to be necessary because full-length GATA-3, in which this region is replaced by the corresponding region of GATA-4, functions normally (Fig. 6). We conclude that the presence of proline in this position in GATA-4 renders the factor incapable of inducing IL-13 production, and that the remainder of the post-zinc finger tail of GATA-3, though dispensable when deleted, can compensate when methionine 368 in GATA-3 is replaced by proline, in the context of full-length GATA-3. Whether our data can be further explained by differential recruitment of an IL-13 specific binding partner or inability of the C-terminal region of GATA-4 to activate transcriptional machinery downstream of acetylated histone H3 remains to be investigated. That GATA-4 can substitute for GATA-3 during the early stages of T cell development but cannot during Th2 development mimics the pattern seen in studies of GATA-1 during erythroid development. When heterologous hemopoietic GATA factors such as GATA-2 or GATA-3 expressed under control of a genomic GATA-1 regulatory fragment were used to rescue GATA-1 knock down embryos, the block at the proerythroblast stage was relieved and the embryos survived lethality (17). Despite rescue from lethality, the adult animals are anemic and have abnormal red cell morphology, indicating that later development requires the distinct actions of GATA-1 (17). Based on these and our data, we speculate that the targets of GATA-3 during early T cell development, as yet unknown, are distinct and of lower stringency than the Th1 and Th2 cytokine genes. We also propose that the different structural requirements for the control of each Th1 and Th2 cytokine demonstrated in this study facilitates fine tuning of cytokine output on a single cell level, ensuring that not all Th2 cytokines are produced at the same levels simultaneously. Knowledge of the biochemical determinants of Th2 cytokine control could be exploited to design inhibitors of GATA-3 that would inhibit IL-13 production specifically, whereas leaving intrathymic CD4 T cell development, IL-4 production and therefore class switching of B cells intact. Specific control of individual Th2 cytokines could thus be useful in the treatment of human diseases, such as allergy and asthma. 1059