Title: A Point Mutation of the AF2 Transactivation Domain of the Glucocorticoid Receptor Disrupts Its Interaction with Steroid Receptor Coactivator 1
Abstract: Glucocorticoids cause a 10-fold increase in hepatic phosphoenolpyruvate carboxykinase (PEPCK) gene transcription through two low affinity glucocorticoid receptor (GR) binding sites and a complex array of accessory factor DNA elements and associated proteins. To analyze how co-activators interact with the GR in this context, we took advantage of the C656G GR mutant that binds ligand with very high affinity. This GR activates PEPCK gene transcription at a 500-fold lower dexamethasone concentration than does wild type GR. Transfected C656G GR containing additional mutations or deletions was tested on PEPCK gene expression in H4IIE hepatoma cells. We found that the AF2 domain is the only one of the three defined transactivation domains in GR that is required for PEPCK gene expression and that mutation of this domain disrupts the direct interaction of GR with steroid receptor coactivator 1 (SRC-1). These data help define the functional interaction between GR and SRC-1 and further define the role of the GR in glucocorticoid-mediated expression of the PEPCK gene. Glucocorticoids cause a 10-fold increase in hepatic phosphoenolpyruvate carboxykinase (PEPCK) gene transcription through two low affinity glucocorticoid receptor (GR) binding sites and a complex array of accessory factor DNA elements and associated proteins. To analyze how co-activators interact with the GR in this context, we took advantage of the C656G GR mutant that binds ligand with very high affinity. This GR activates PEPCK gene transcription at a 500-fold lower dexamethasone concentration than does wild type GR. Transfected C656G GR containing additional mutations or deletions was tested on PEPCK gene expression in H4IIE hepatoma cells. We found that the AF2 domain is the only one of the three defined transactivation domains in GR that is required for PEPCK gene expression and that mutation of this domain disrupts the direct interaction of GR with steroid receptor coactivator 1 (SRC-1). These data help define the functional interaction between GR and SRC-1 and further define the role of the GR in glucocorticoid-mediated expression of the PEPCK gene. glucocorticoid accessory factor chloramphenicol acetyltransferase DNA-binding domain dexamethasone glucocorticoid accessory factor binding site glucocorticoid receptor human GR rat GR mouse GR glucocorticoid response element hepatocyte nuclear factor phosphoenolpyruvate carboxykinase steroid receptor coactivator transactivation domain Dulbecco's modified Eagle's medium 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol murine mammary tumor virus-long terminal repeat Glucocorticoids (GCs)1play a central role in carbohydrate metabolism by increasing glucose production, decreasing glucose tolerance, and causing insulin resistance (1Sapolsky R.M. Romero L.M. Munck A.U. Endocr. Rev. 2000; 21: 55-89Crossref PubMed Scopus (5378) Google Scholar). GCs increase glucose production primarily by inducing the transcription of genes that encode gluconeogenic enzymes, including PEPCK, a rate-determining enzyme of gluconeogenesis. GCs increase the rate of transcription of the PEPCK gene by ∼10-fold (2Granner D. Pilkis S. J. Biol. Chem. 1990; 265: 10173-10176Abstract Full Text PDF PubMed Google Scholar) through a multicomponent glucocorticoid response unit. The glucocorticoid response unit consists of a tandem array of four accessory factor elements (gAF1–4) 2We originally designated the accessory factor elements AF1, AF2, and AF3 in the order of their discovery. The subsequent definition of transactivation domains of the nuclear receptor subfamily as AF1 and AF2 (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar) has created a potentially confusing situation. To avoid having to discuss the interaction of the AF1 (or AF2) transcription domain of the glucocorticoid receptor with the AF1 element (or with the AF2 and AF3 elements) in the glucocorticoid response unit, we now designate these elements as gAF1 to denote glucocorticoid accessory factor 1 element, for example.2We originally designated the accessory factor elements AF1, AF2, and AF3 in the order of their discovery. The subsequent definition of transactivation domains of the nuclear receptor subfamily as AF1 and AF2 (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar) has created a potentially confusing situation. To avoid having to discuss the interaction of the AF1 (or AF2) transcription domain of the glucocorticoid receptor with the AF1 element (or with the AF2 and AF3 elements) in the glucocorticoid response unit, we now designate these elements as gAF1 to denote glucocorticoid accessory factor 1 element, for example. that bind HNF4/COUP-TF, HNF3, COUP-TF, and C/EBPβ, respectively, and two non-consensus glucocorticoid response elements (GR1, GR2) (see Fig. 6,inset) (3Wang J.C. Stromstedt P.E. Sugiyama T. Granner D.K. Mol. Endocrinol. 1999; 13: 604-618PubMed Google Scholar, 4Granner D. O'Brien R. Imai E. Forest C. Mitchell J. Lucas P. Recent Prog. Horm. Res. 1991; 47: 319-346PubMed Google Scholar). The GR binds to the GR1 and GR2 elements 10–20 times less avidly than to a consensus GRE (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar), and GR1 and GR2 are unable to confer glucocorticoid responsiveness by themselves (6Imai E. Stromstedt P.E. Quinn P.G. Carlstedt-Duke J. Gustafsson J.A. Granner D.K. Mol. Cell. Biol. 1990; 10: 4712-4719Crossref PubMed Scopus (241) Google Scholar). Mutations that disrupt the binding of GR to GR1 result in a more severe reduction of the GC response than does disruption of binding to GR2 (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar).GR belongs to the superfamily of steroid/thyroid/retinoic acid receptor proteins that function as ligand-dependent transcription factors (7Evans R.M. Science. 1988; 240: 889-895Crossref PubMed Scopus (6291) Google Scholar). Two transactivation domains (referred to as τ1 and τ2) were originally identified in the human GR (hGR) (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar, 9Giguere V. Hollenberg S.M. Rosenfeld M.G. Evans R.M. Cell. 1986; 46: 645-652Abstract Full Text PDF PubMed Scopus (674) Google Scholar, 10Hollenberg S.M. Weinberger C. Ong E.S. Cerelli G. Oro A. Lebo R. Thompson E.B. Rosenfeld M.G. Evans R.M. Nature. 1985; 318: 635-641Crossref PubMed Scopus (1437) Google Scholar, 11Hollenberg S.M. Giguere V. Segui P. Evans R.M. Cell. 1987; 49: 39-46Abstract Full Text PDF PubMed Scopus (332) Google Scholar, 12Hollenberg S.M. Evans R.M. Cell. 1988; 55: 899-906Abstract Full Text PDF PubMed Scopus (545) Google Scholar). An additional region involved in transactivation, AF2, has been mapped (see Fig.2 A, inset) (13Danielian P.S. White R. Lees J.A. Parker M.G. EMBO J. 1992; 11: 1025-1033Crossref PubMed Scopus (716) Google Scholar, 14Halachmi S. Marden E. Martin G. MacKay H. Abbondanza C. Brown M. Science. 1994; 264: 1455-1458Crossref PubMed Scopus (565) Google Scholar, 15Boruk M. Savory J.G. Hache R.J. Mol. Endocrinol. 1998; 12: 1749-1763Crossref PubMed Scopus (76) Google Scholar). The τ2 and AF2 domains are both located in the ligand-binding domain (LBD) in the C-terminal regions of GR. The τ1 domain (also referred to as enh2 or AF-1; residues 77–262 in hGR or 106–318 in rat GR (rGR)) is located in the N-terminal region of the GR molecule and is generally considered a major region responsible for transactivation. τ1 makes contact with proteins in the basal transcriptional apparatus, including the TATA box-binding protein (TBP), possibly through an intermediary adapter protein(s) (16McEwan I.J. Wright A.P. Dahlman-Wright K. Carlstedt-Duke J. Gustafsson J.A. Mol. Cell. Biol. 1993; 13: 399-407Crossref PubMed Google Scholar, 17Henriksson A. Almlof T. Ford J. McEwan I.J. Gustafsson J.A. Wright A.P. Mol. Cell. Biol. 1997; 17: 3065-3073Crossref PubMed Scopus (65) Google Scholar, 18Ford J. McEwan I.J. Wright A.P. Gustafsson J.A. Mol. Endocrinol. 1997; 11: 1467-1475Crossref PubMed Scopus (70) Google Scholar). The τ1 domain is highly acidic and phosphorylated (19Almlof T. Wright A.P. Gustafsson J.A. J. Biol. Chem. 1995; 270: 17535-17540Abstract Full Text Full Text PDF PubMed Scopus (89) Google Scholar, 20Bodwell J.E. Webster J.C. Jewell C.M. Cidlowski J.A., Hu, J.M. Munck A. J. Steroid Biochem. Mol. Biol. 1998; 65: 91-99Crossref PubMed Scopus (116) Google Scholar) and it may undergo interdomain interactions consequent to DNA binding (21Kumar R. Baskakov I.V. Srinivasan G. Bolen D.W. Lee J.C. Thompson E.B. J. Biol. Chem. 1999; 274: 24737-24741Abstract Full Text Full Text PDF PubMed Scopus (91) Google Scholar). The τ2 domain (residues 526–556 in hGR or 544–573 in rGR) possesses transactivation potential in the context of full-length GR (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar), nuclear matrix targeting activity (23Tang Y. Getzenberg R.H. Vietmeier B.N. Stallcup M.R. Eggert M. Renkawitz R. DeFranco D.B. Mol. Endocrinol. 1998; 12: 1420-1431PubMed Google Scholar), and a binding domain for hsp90 (24Giannoukos G. Silverstein A.M. Pratt W.B. Simons S.S., Jr. J. Biol. Chem. 1999; 274: 36527-36536Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar). The C-terminal AF2 domain (sometimes called τc; residues 727–763 in hGR (25McEwan I.J. Wright A.P. Gustafsson J.A. Bioessays. 1997; 19: 153-160Crossref PubMed Scopus (160) Google Scholar) or 745–781 in rGR, determined also as 752–758 in hGR (26Williams S.P. Sigler P.B. Nature. 1998; 393: 392-396Crossref PubMed Scopus (574) Google Scholar)) forms part of the amphipathic α-helix 12 of the ligand binding domain and is responsible for the hormone-dependent interaction with coactivators of the p160 family, e.g. SRC-1 (27Onate S.A. Tsai S.Y. Tsai M.J. O'Malley B.W. Science. 1995; 270: 1354-1357Crossref PubMed Scopus (2042) Google Scholar, 28Ding X.F. Anderson C.M., Ma, H. Hong H. Uht R.M. Kushner P.J. Stallcup M.R. Mol. Endocrinol. 1998; 12: 302-313Crossref PubMed Google Scholar). This interaction with coactivators may occur because of the conformational change AF2 undergoes upon ligand binding (29Renaud J.P. Rochel N. Ruff M. Vivat V. Chambon P. Gronemeyer H. Moras D. Nature. 1995; 378: 681-689Crossref PubMed Scopus (1021) Google Scholar, 30Wagner R.L. Apriletti J.W. McGrath M.E. West B.L. Baxter J.D. Fletterick R.J. Nature. 1995; 378: 690-697Crossref PubMed Scopus (805) Google Scholar), but the exact residues have not been defined.Figure 2Involvement of the TADs of GR in Dex-induced PEPCK gene transcription. A, H4IIE cells were transiently transfected with 5 μg of the PEPCK-LUC reporter construct (the PEPCK gene promoter encompassing nucleotides −467 to +65 relative to the transcription start site, ligated to the luciferase reporter gene) and 2.5 μg of either a GR expression plasmid or an empty RSV-neo vector. Cells were treated with 1 nmDex for 18 h in serum-free DMEM and harvested for the assay of luciferase activity. The GR expression vectors used are as follows: coGR, the C656G GR mutant;menh2, the triple mutation E219K,F220L,W234R;denh2, the deletion 106–318 of the enh2/τ1/AF1 domain;mtau2, the S573A mutation; mAF2, the E773A mutation. The results represent the mean ± S.E. of at least five experiments. The asterisks indicate a statistically significant difference (p < 0.05) from coGR.Panel B shows that the C656G GR AF2 mutants are expressed equally. The C656G GR and the corresponding AF2 GR mutants (40 μg) were transfected into COS cells. Whole cell extracts were prepared and immunoprecipitated with GR antibodies for GR immunoblot analysis.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Two conditions had to be met in order to study the exact contribution of the three transactivation domains of GR to PEPCK gene transcription: 1) the contribution of the endogenous GR in H4IIE hepatoma cells used for analyzing GC-stimulated PEPCK gene transcription had to be eliminated; and 2) GR mutations that disrupt the transactivation potential of each TAD, while maintaining the ligand and DNA binding properties of GR, had to be constructed. In this study we employed a system that eliminates the influence of the endogenous GR by using GR C656G (31Chakraborti P.K. Garabedian M.J. Yamamoto K.R. Simons S.S., Jr. J. Biol. Chem. 1991; 266: 22075-22078Abstract Full Text PDF PubMed Google Scholar). This mutant receptor is maximally activated by glucocorticoids at 500 times less ligand than is required for activation of the endogenous, wild type GR, thereby eliminating the influence of the latter. We used this system and a series of mutations of C656G GR to demonstrate the unique importance of the AF2 domain for induction of the PEPCK gene by glucocorticoids.DISCUSSIONThree cysteine residues (Cys640, Cys656, and Cys661) are involved in the GR-ligand interaction. The point mutations C640S and C661S each reduces the affinity of glucocorticoid binding to GR. Surprisingly, mutation C656G increases the affinity of the GR for its ligand and also activates transcription at a lower ligand concentration (31Chakraborti P.K. Garabedian M.J. Yamamoto K.R. Simons S.S., Jr. J. Biol. Chem. 1991; 266: 22075-22078Abstract Full Text PDF PubMed Google Scholar). As we show in this study, half-maximal activation of expression from a PEPCK gene promoter-driven reporter construct in H4IIE cells is achieved at a 500-fold lower ligand concentration when C656G GR is used. Because significant activation occurs at 1 nm Dex, a concentration at which wild type GR is inactive vis-′a-vis the PEPCK gene, one can bypass the endogenous, wild type GR and thereby analyze the function of different regions of a transfected GR (Fig. 1).Two hGR domains, referred to as τ1 and τ2, are necessary for activation of transcription but are not specifically involved in either steroid or DNA binding (9Giguere V. Hollenberg S.M. Rosenfeld M.G. Evans R.M. Cell. 1986; 46: 645-652Abstract Full Text PDF PubMed Scopus (674) Google Scholar). These hGR transactivation domains were defined according to two criteria: 1) deletion of the domain decreases activity and 2) duplication of the domain increases activity (12Hollenberg S.M. Evans R.M. Cell. 1988; 55: 899-906Abstract Full Text PDF PubMed Scopus (545) Google Scholar). We show that deletion of the enh2/τ1/AF1 domain in rGR (Fig. 2) does not decrease the activation of PEPCK gene transcription promoted by glucocorticoids. This observation does not speak against the transactivation function of this domain but rather speaks for the fact that different activation surfaces of GR are employed in different GRE and promoter contexts. For example, deletion of the enh2/τ1/AF1 domain in the rGR does not affect the activation of the tyrosine aminotransferase promoter, but this mutation completely abrogates the activation of the MMTV-LTR promoter by hormone (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). Distinct GR surfaces also mediate the cytotoxic and cytostatic effects of glucocorticoids (43Rogatsky I. Hittelman A.B. Pearce D. Garabedian M.J. Mol. Cell. Biol. 1999; 19: 5036-5049Crossref PubMed Scopus (90) Google Scholar).Three amino acid substitutions in the enh2/τ1/AF1 domain (E219K, F220L, W234R) selectively disrupt the activation function of this N-terminal TAD in GR (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). This triple mutation does not influence PEPCK gene transcription (Fig. 2), and hence GC regulation of the PEPCK gene must be conferred by one or both of the C-terminal TADs. The hinge region of steroid receptors, which includes the τ2 domain, interacts with the general transcription factor, TAFII30, in vitro(44Jacq X. Brou C. Lutz Y. Davidson I. Chambon P. Tora L. Cell. 1994; 79: 107-117Abstract Full Text PDF PubMed Scopus (344) Google Scholar) and a novel coactivator, L7/SPA, in vivo (45Jackson T.A. Richer J.K. Bain D.L. Takimoto G.S. Tung L. Horwitz K.B. Mol. Endocrinol. 1997; 11: 693-705Crossref PubMed Scopus (383) Google Scholar), but no evidence exists for the interaction of the τ2 domain with the p160 coactivator family (SRC-1, GRIP1/TIF2, and p/CIP) (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). S573A in rGR is homologous to S561A in mGR that causes a loss of transactivation activity in the minimum τ2 fragment and the full-length GR (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). The disruption of the τ2 transactivation domain by point mutation S573A also does not diminish the activation of PEPCK gene (Fig. 2). By contrast, a point mutation, E773A (mAF2), in the AF2 transactivation domain, which is known to interact with coactivators, does decrease the response of the PEPCK reporter gene to Dex by 70%. However, the mAF2 GR does not affect activation of transcription through a PEPCK promoter construct that contains a palindromic GRE in the place of GR1. It is noteworthy that this replacement relieves the requirement for gAF1 and gAF3 (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar). These accessory factor elements bind HNF4/COUP-TF and COUP-TF, respectively, an interaction that affects the affinity and stability of GR binding to GR1, and thus perhaps influence the interaction of SRC-1 with GR (46Stafford J.M. Wilkinson J.C. Beechem J.M. Granner D.K. J. Biol. Chem. 2001; 276: 39885-39891Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). The palindromic GRE may therefore present a surface of GR that is sufficient, in itself, to interact with SRC-1.Although it is known that the AF2 domain of GR is involved in the interaction with SRC-1 in vivo (28Ding X.F. Anderson C.M., Ma, H. Hong H. Uht R.M. Kushner P.J. Stallcup M.R. Mol. Endocrinol. 1998; 12: 302-313Crossref PubMed Google Scholar), the GR residue(s) necessary for this interaction are unknown. We show here that the GR·SRC-1 interaction is interrupted when the E773A mutation is introduced into the GR (Fig. 6). There is specificity to this reaction because another AF2 mutation, Y753F, which displays reduced activation of the MMTV promoter (42Ray D.W. Suen C.S. Brass A. Soden J. White A. Mol. Endocrinol. 1999; 13: 1855-1863Crossref PubMed Scopus (49) Google Scholar), has no influence on the GR·SRC-1 interaction in the yeast two-hybrid system (Fig. 5).In this study we further demonstrate the utility of using C656G GR to eliminate the function of endogenous GR. This Dex-sensitive GR allowed us to identify the transactivation domains of GR necessary for transcription of the PEPCK gene, which was heretofore not possible. Instead of possessing high affinity GREs, the PEPCK gene promoter has two low affinity GR binding sites, GR1 and GR2, and thus requires accessory factor activity to stabilize GR binding and to recruit coactivators to the promoter. We have identified a residue, Glu773, that directs the interaction of GR with SRC-1 in the context of the PEPCK gene promoter. Mutation of this residue does not disrupt activation of the PEPCK gene if the low affinity GR1 site is changed to a high affinity, palindromic GRE. These data suggest that a context-dependent conformation of GR is essential for the recruitment of SRC-1 to the PEPCK gene promoter and efficient glucocorticoid-mediated induction of this gene. Glucocorticoids (GCs)1play a central role in carbohydrate metabolism by increasing glucose production, decreasing glucose tolerance, and causing insulin resistance (1Sapolsky R.M. Romero L.M. Munck A.U. Endocr. Rev. 2000; 21: 55-89Crossref PubMed Scopus (5378) Google Scholar). GCs increase glucose production primarily by inducing the transcription of genes that encode gluconeogenic enzymes, including PEPCK, a rate-determining enzyme of gluconeogenesis. GCs increase the rate of transcription of the PEPCK gene by ∼10-fold (2Granner D. Pilkis S. J. Biol. Chem. 1990; 265: 10173-10176Abstract Full Text PDF PubMed Google Scholar) through a multicomponent glucocorticoid response unit. The glucocorticoid response unit consists of a tandem array of four accessory factor elements (gAF1–4) 2We originally designated the accessory factor elements AF1, AF2, and AF3 in the order of their discovery. The subsequent definition of transactivation domains of the nuclear receptor subfamily as AF1 and AF2 (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar) has created a potentially confusing situation. To avoid having to discuss the interaction of the AF1 (or AF2) transcription domain of the glucocorticoid receptor with the AF1 element (or with the AF2 and AF3 elements) in the glucocorticoid response unit, we now designate these elements as gAF1 to denote glucocorticoid accessory factor 1 element, for example.2We originally designated the accessory factor elements AF1, AF2, and AF3 in the order of their discovery. The subsequent definition of transactivation domains of the nuclear receptor subfamily as AF1 and AF2 (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar) has created a potentially confusing situation. To avoid having to discuss the interaction of the AF1 (or AF2) transcription domain of the glucocorticoid receptor with the AF1 element (or with the AF2 and AF3 elements) in the glucocorticoid response unit, we now designate these elements as gAF1 to denote glucocorticoid accessory factor 1 element, for example. that bind HNF4/COUP-TF, HNF3, COUP-TF, and C/EBPβ, respectively, and two non-consensus glucocorticoid response elements (GR1, GR2) (see Fig. 6,inset) (3Wang J.C. Stromstedt P.E. Sugiyama T. Granner D.K. Mol. Endocrinol. 1999; 13: 604-618PubMed Google Scholar, 4Granner D. O'Brien R. Imai E. Forest C. Mitchell J. Lucas P. Recent Prog. Horm. Res. 1991; 47: 319-346PubMed Google Scholar). The GR binds to the GR1 and GR2 elements 10–20 times less avidly than to a consensus GRE (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar), and GR1 and GR2 are unable to confer glucocorticoid responsiveness by themselves (6Imai E. Stromstedt P.E. Quinn P.G. Carlstedt-Duke J. Gustafsson J.A. Granner D.K. Mol. Cell. Biol. 1990; 10: 4712-4719Crossref PubMed Scopus (241) Google Scholar). Mutations that disrupt the binding of GR to GR1 result in a more severe reduction of the GC response than does disruption of binding to GR2 (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar). GR belongs to the superfamily of steroid/thyroid/retinoic acid receptor proteins that function as ligand-dependent transcription factors (7Evans R.M. Science. 1988; 240: 889-895Crossref PubMed Scopus (6291) Google Scholar). Two transactivation domains (referred to as τ1 and τ2) were originally identified in the human GR (hGR) (8Evans R.M. Hollenberg S.M. Cold Spring Harbor Symp. Quant. Biol. 1988; 53: 813-818Crossref PubMed Google Scholar, 9Giguere V. Hollenberg S.M. Rosenfeld M.G. Evans R.M. Cell. 1986; 46: 645-652Abstract Full Text PDF PubMed Scopus (674) Google Scholar, 10Hollenberg S.M. Weinberger C. Ong E.S. Cerelli G. Oro A. Lebo R. Thompson E.B. Rosenfeld M.G. Evans R.M. Nature. 1985; 318: 635-641Crossref PubMed Scopus (1437) Google Scholar, 11Hollenberg S.M. Giguere V. Segui P. Evans R.M. Cell. 1987; 49: 39-46Abstract Full Text PDF PubMed Scopus (332) Google Scholar, 12Hollenberg S.M. Evans R.M. Cell. 1988; 55: 899-906Abstract Full Text PDF PubMed Scopus (545) Google Scholar). An additional region involved in transactivation, AF2, has been mapped (see Fig.2 A, inset) (13Danielian P.S. White R. Lees J.A. Parker M.G. EMBO J. 1992; 11: 1025-1033Crossref PubMed Scopus (716) Google Scholar, 14Halachmi S. Marden E. Martin G. MacKay H. Abbondanza C. Brown M. Science. 1994; 264: 1455-1458Crossref PubMed Scopus (565) Google Scholar, 15Boruk M. Savory J.G. Hache R.J. Mol. Endocrinol. 1998; 12: 1749-1763Crossref PubMed Scopus (76) Google Scholar). The τ2 and AF2 domains are both located in the ligand-binding domain (LBD) in the C-terminal regions of GR. The τ1 domain (also referred to as enh2 or AF-1; residues 77–262 in hGR or 106–318 in rat GR (rGR)) is located in the N-terminal region of the GR molecule and is generally considered a major region responsible for transactivation. τ1 makes contact with proteins in the basal transcriptional apparatus, including the TATA box-binding protein (TBP), possibly through an intermediary adapter protein(s) (16McEwan I.J. Wright A.P. Dahlman-Wright K. Carlstedt-Duke J. Gustafsson J.A. Mol. Cell. Biol. 1993; 13: 399-407Crossref PubMed Google Scholar, 17Henriksson A. Almlof T. Ford J. McEwan I.J. Gustafsson J.A. Wright A.P. Mol. Cell. Biol. 1997; 17: 3065-3073Crossref PubMed Scopus (65) Google Scholar, 18Ford J. McEwan I.J. Wright A.P. Gustafsson J.A. Mol. Endocrinol. 1997; 11: 1467-1475Crossref PubMed Scopus (70) Google Scholar). The τ1 domain is highly acidic and phosphorylated (19Almlof T. Wright A.P. Gustafsson J.A. J. Biol. Chem. 1995; 270: 17535-17540Abstract Full Text Full Text PDF PubMed Scopus (89) Google Scholar, 20Bodwell J.E. Webster J.C. Jewell C.M. Cidlowski J.A., Hu, J.M. Munck A. J. Steroid Biochem. Mol. Biol. 1998; 65: 91-99Crossref PubMed Scopus (116) Google Scholar) and it may undergo interdomain interactions consequent to DNA binding (21Kumar R. Baskakov I.V. Srinivasan G. Bolen D.W. Lee J.C. Thompson E.B. J. Biol. Chem. 1999; 274: 24737-24741Abstract Full Text Full Text PDF PubMed Scopus (91) Google Scholar). The τ2 domain (residues 526–556 in hGR or 544–573 in rGR) possesses transactivation potential in the context of full-length GR (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar), nuclear matrix targeting activity (23Tang Y. Getzenberg R.H. Vietmeier B.N. Stallcup M.R. Eggert M. Renkawitz R. DeFranco D.B. Mol. Endocrinol. 1998; 12: 1420-1431PubMed Google Scholar), and a binding domain for hsp90 (24Giannoukos G. Silverstein A.M. Pratt W.B. Simons S.S., Jr. J. Biol. Chem. 1999; 274: 36527-36536Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar). The C-terminal AF2 domain (sometimes called τc; residues 727–763 in hGR (25McEwan I.J. Wright A.P. Gustafsson J.A. Bioessays. 1997; 19: 153-160Crossref PubMed Scopus (160) Google Scholar) or 745–781 in rGR, determined also as 752–758 in hGR (26Williams S.P. Sigler P.B. Nature. 1998; 393: 392-396Crossref PubMed Scopus (574) Google Scholar)) forms part of the amphipathic α-helix 12 of the ligand binding domain and is responsible for the hormone-dependent interaction with coactivators of the p160 family, e.g. SRC-1 (27Onate S.A. Tsai S.Y. Tsai M.J. O'Malley B.W. Science. 1995; 270: 1354-1357Crossref PubMed Scopus (2042) Google Scholar, 28Ding X.F. Anderson C.M., Ma, H. Hong H. Uht R.M. Kushner P.J. Stallcup M.R. Mol. Endocrinol. 1998; 12: 302-313Crossref PubMed Google Scholar). This interaction with coactivators may occur because of the conformational change AF2 undergoes upon ligand binding (29Renaud J.P. Rochel N. Ruff M. Vivat V. Chambon P. Gronemeyer H. Moras D. Nature. 1995; 378: 681-689Crossref PubMed Scopus (1021) Google Scholar, 30Wagner R.L. Apriletti J.W. McGrath M.E. West B.L. Baxter J.D. Fletterick R.J. Nature. 1995; 378: 690-697Crossref PubMed Scopus (805) Google Scholar), but the exact residues have not been defined. Two conditions had to be met in order to study the exact contribution of the three transactivation domains of GR to PEPCK gene transcription: 1) the contribution of the endogenous GR in H4IIE hepatoma cells used for analyzing GC-stimulated PEPCK gene transcription had to be eliminated; and 2) GR mutations that disrupt the transactivation potential of each TAD, while maintaining the ligand and DNA binding properties of GR, had to be constructed. In this study we employed a system that eliminates the influence of the endogenous GR by using GR C656G (31Chakraborti P.K. Garabedian M.J. Yamamoto K.R. Simons S.S., Jr. J. Biol. Chem. 1991; 266: 22075-22078Abstract Full Text PDF PubMed Google Scholar). This mutant receptor is maximally activated by glucocorticoids at 500 times less ligand than is required for activation of the endogenous, wild type GR, thereby eliminating the influence of the latter. We used this system and a series of mutations of C656G GR to demonstrate the unique importance of the AF2 domain for induction of the PEPCK gene by glucocorticoids. DISCUSSIONThree cysteine residues (Cys640, Cys656, and Cys661) are involved in the GR-ligand interaction. The point mutations C640S and C661S each reduces the affinity of glucocorticoid binding to GR. Surprisingly, mutation C656G increases the affinity of the GR for its ligand and also activates transcription at a lower ligand concentration (31Chakraborti P.K. Garabedian M.J. Yamamoto K.R. Simons S.S., Jr. J. Biol. Chem. 1991; 266: 22075-22078Abstract Full Text PDF PubMed Google Scholar). As we show in this study, half-maximal activation of expression from a PEPCK gene promoter-driven reporter construct in H4IIE cells is achieved at a 500-fold lower ligand concentration when C656G GR is used. Because significant activation occurs at 1 nm Dex, a concentration at which wild type GR is inactive vis-′a-vis the PEPCK gene, one can bypass the endogenous, wild type GR and thereby analyze the function of different regions of a transfected GR (Fig. 1).Two hGR domains, referred to as τ1 and τ2, are necessary for activation of transcription but are not specifically involved in either steroid or DNA binding (9Giguere V. Hollenberg S.M. Rosenfeld M.G. Evans R.M. Cell. 1986; 46: 645-652Abstract Full Text PDF PubMed Scopus (674) Google Scholar). These hGR transactivation domains were defined according to two criteria: 1) deletion of the domain decreases activity and 2) duplication of the domain increases activity (12Hollenberg S.M. Evans R.M. Cell. 1988; 55: 899-906Abstract Full Text PDF PubMed Scopus (545) Google Scholar). We show that deletion of the enh2/τ1/AF1 domain in rGR (Fig. 2) does not decrease the activation of PEPCK gene transcription promoted by glucocorticoids. This observation does not speak against the transactivation function of this domain but rather speaks for the fact that different activation surfaces of GR are employed in different GRE and promoter contexts. For example, deletion of the enh2/τ1/AF1 domain in the rGR does not affect the activation of the tyrosine aminotransferase promoter, but this mutation completely abrogates the activation of the MMTV-LTR promoter by hormone (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). Distinct GR surfaces also mediate the cytotoxic and cytostatic effects of glucocorticoids (43Rogatsky I. Hittelman A.B. Pearce D. Garabedian M.J. Mol. Cell. Biol. 1999; 19: 5036-5049Crossref PubMed Scopus (90) Google Scholar).Three amino acid substitutions in the enh2/τ1/AF1 domain (E219K, F220L, W234R) selectively disrupt the activation function of this N-terminal TAD in GR (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). This triple mutation does not influence PEPCK gene transcription (Fig. 2), and hence GC regulation of the PEPCK gene must be conferred by one or both of the C-terminal TADs. The hinge region of steroid receptors, which includes the τ2 domain, interacts with the general transcription factor, TAFII30, in vitro(44Jacq X. Brou C. Lutz Y. Davidson I. Chambon P. Tora L. Cell. 1994; 79: 107-117Abstract Full Text PDF PubMed Scopus (344) Google Scholar) and a novel coactivator, L7/SPA, in vivo (45Jackson T.A. Richer J.K. Bain D.L. Takimoto G.S. Tung L. Horwitz K.B. Mol. Endocrinol. 1997; 11: 693-705Crossref PubMed Scopus (383) Google Scholar), but no evidence exists for the interaction of the τ2 domain with the p160 coactivator family (SRC-1, GRIP1/TIF2, and p/CIP) (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). S573A in rGR is homologous to S561A in mGR that causes a loss of transactivation activity in the minimum τ2 fragment and the full-length GR (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). The disruption of the τ2 transactivation domain by point mutation S573A also does not diminish the activation of PEPCK gene (Fig. 2). By contrast, a point mutation, E773A (mAF2), in the AF2 transactivation domain, which is known to interact with coactivators, does decrease the response of the PEPCK reporter gene to Dex by 70%. However, the mAF2 GR does not affect activation of transcription through a PEPCK promoter construct that contains a palindromic GRE in the place of GR1. It is noteworthy that this replacement relieves the requirement for gAF1 and gAF3 (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar). These accessory factor elements bind HNF4/COUP-TF and COUP-TF, respectively, an interaction that affects the affinity and stability of GR binding to GR1, and thus perhaps influence the interaction of SRC-1 with GR (46Stafford J.M. Wilkinson J.C. Beechem J.M. Granner D.K. J. Biol. Chem. 2001; 276: 39885-39891Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). The palindromic GRE may therefore present a surface of GR that is sufficient, in itself, to interact with SRC-1.Although it is known that the AF2 domain of GR is involved in the interaction with SRC-1 in vivo (28Ding X.F. Anderson C.M., Ma, H. Hong H. Uht R.M. Kushner P.J. Stallcup M.R. Mol. Endocrinol. 1998; 12: 302-313Crossref PubMed Google Scholar), the GR residue(s) necessary for this interaction are unknown. We show here that the GR·SRC-1 interaction is interrupted when the E773A mutation is introduced into the GR (Fig. 6). There is specificity to this reaction because another AF2 mutation, Y753F, which displays reduced activation of the MMTV promoter (42Ray D.W. Suen C.S. Brass A. Soden J. White A. Mol. Endocrinol. 1999; 13: 1855-1863Crossref PubMed Scopus (49) Google Scholar), has no influence on the GR·SRC-1 interaction in the yeast two-hybrid system (Fig. 5).In this study we further demonstrate the utility of using C656G GR to eliminate the function of endogenous GR. This Dex-sensitive GR allowed us to identify the transactivation domains of GR necessary for transcription of the PEPCK gene, which was heretofore not possible. Instead of possessing high affinity GREs, the PEPCK gene promoter has two low affinity GR binding sites, GR1 and GR2, and thus requires accessory factor activity to stabilize GR binding and to recruit coactivators to the promoter. We have identified a residue, Glu773, that directs the interaction of GR with SRC-1 in the context of the PEPCK gene promoter. Mutation of this residue does not disrupt activation of the PEPCK gene if the low affinity GR1 site is changed to a high affinity, palindromic GRE. These data suggest that a context-dependent conformation of GR is essential for the recruitment of SRC-1 to the PEPCK gene promoter and efficient glucocorticoid-mediated induction of this gene. Three cysteine residues (Cys640, Cys656, and Cys661) are involved in the GR-ligand interaction. The point mutations C640S and C661S each reduces the affinity of glucocorticoid binding to GR. Surprisingly, mutation C656G increases the affinity of the GR for its ligand and also activates transcription at a lower ligand concentration (31Chakraborti P.K. Garabedian M.J. Yamamoto K.R. Simons S.S., Jr. J. Biol. Chem. 1991; 266: 22075-22078Abstract Full Text PDF PubMed Google Scholar). As we show in this study, half-maximal activation of expression from a PEPCK gene promoter-driven reporter construct in H4IIE cells is achieved at a 500-fold lower ligand concentration when C656G GR is used. Because significant activation occurs at 1 nm Dex, a concentration at which wild type GR is inactive vis-′a-vis the PEPCK gene, one can bypass the endogenous, wild type GR and thereby analyze the function of different regions of a transfected GR (Fig. 1). Two hGR domains, referred to as τ1 and τ2, are necessary for activation of transcription but are not specifically involved in either steroid or DNA binding (9Giguere V. Hollenberg S.M. Rosenfeld M.G. Evans R.M. Cell. 1986; 46: 645-652Abstract Full Text PDF PubMed Scopus (674) Google Scholar). These hGR transactivation domains were defined according to two criteria: 1) deletion of the domain decreases activity and 2) duplication of the domain increases activity (12Hollenberg S.M. Evans R.M. Cell. 1988; 55: 899-906Abstract Full Text PDF PubMed Scopus (545) Google Scholar). We show that deletion of the enh2/τ1/AF1 domain in rGR (Fig. 2) does not decrease the activation of PEPCK gene transcription promoted by glucocorticoids. This observation does not speak against the transactivation function of this domain but rather speaks for the fact that different activation surfaces of GR are employed in different GRE and promoter contexts. For example, deletion of the enh2/τ1/AF1 domain in the rGR does not affect the activation of the tyrosine aminotransferase promoter, but this mutation completely abrogates the activation of the MMTV-LTR promoter by hormone (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). Distinct GR surfaces also mediate the cytotoxic and cytostatic effects of glucocorticoids (43Rogatsky I. Hittelman A.B. Pearce D. Garabedian M.J. Mol. Cell. Biol. 1999; 19: 5036-5049Crossref PubMed Scopus (90) Google Scholar). Three amino acid substitutions in the enh2/τ1/AF1 domain (E219K, F220L, W234R) selectively disrupt the activation function of this N-terminal TAD in GR (39Iniguez-Lluhi J.A. Lou D.Y. Yamamoto K.R. J. Biol. Chem. 1997; 272: 4149-4156Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar). This triple mutation does not influence PEPCK gene transcription (Fig. 2), and hence GC regulation of the PEPCK gene must be conferred by one or both of the C-terminal TADs. The hinge region of steroid receptors, which includes the τ2 domain, interacts with the general transcription factor, TAFII30, in vitro(44Jacq X. Brou C. Lutz Y. Davidson I. Chambon P. Tora L. Cell. 1994; 79: 107-117Abstract Full Text PDF PubMed Scopus (344) Google Scholar) and a novel coactivator, L7/SPA, in vivo (45Jackson T.A. Richer J.K. Bain D.L. Takimoto G.S. Tung L. Horwitz K.B. Mol. Endocrinol. 1997; 11: 693-705Crossref PubMed Scopus (383) Google Scholar), but no evidence exists for the interaction of the τ2 domain with the p160 coactivator family (SRC-1, GRIP1/TIF2, and p/CIP) (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). S573A in rGR is homologous to S561A in mGR that causes a loss of transactivation activity in the minimum τ2 fragment and the full-length GR (22Milhon J. Lee S. Kohli K. Chen D. Hong H. Stallcup M.R. Mol. Endocrinol. 1997; 11: 1795-1805PubMed Google Scholar). The disruption of the τ2 transactivation domain by point mutation S573A also does not diminish the activation of PEPCK gene (Fig. 2). By contrast, a point mutation, E773A (mAF2), in the AF2 transactivation domain, which is known to interact with coactivators, does decrease the response of the PEPCK reporter gene to Dex by 70%. However, the mAF2 GR does not affect activation of transcription through a PEPCK promoter construct that contains a palindromic GRE in the place of GR1. It is noteworthy that this replacement relieves the requirement for gAF1 and gAF3 (5Scott D.K. Stromstedt P.E. Wang J.C. Granner D.K. Mol. Endocrinol. 1998; 12: 482-491Crossref PubMed Scopus (65) Google Scholar). These accessory factor elements bind HNF4/COUP-TF and COUP-TF, respectively, an interaction that affects the affinity and stability of GR binding to GR1, and thus perhaps influence the interaction of SRC-1 with GR (46Stafford J.M. Wilkinson J.C. Beechem J.M. Granner D.K. J. Biol. Chem. 2001; 276: 39885-39891Abstract Full Text Full Text PDF PubMed Scopus (77) Google Scholar). The palindromic GRE may therefore present a surface of GR that is sufficient, in itself, to interact with SRC-1. Although it is known that the AF2 domain of GR is involved in the interaction with SRC-1 in vivo (28Ding X.F. Anderson C.M., Ma, H. Hong H. Uht R.M. Kushner P.J. Stallcup M.R. Mol. Endocrinol. 1998; 12: 302-313Crossref PubMed Google Scholar), the GR residue(s) necessary for this interaction are unknown. We show here that the GR·SRC-1 interaction is interrupted when the E773A mutation is introduced into the GR (Fig. 6). There is specificity to this reaction because another AF2 mutation, Y753F, which displays reduced activation of the MMTV promoter (42Ray D.W. Suen C.S. Brass A. Soden J. White A. Mol. Endocrinol. 1999; 13: 1855-1863Crossref PubMed Scopus (49) Google Scholar), has no influence on the GR·SRC-1 interaction in the yeast two-hybrid system (Fig. 5). In this study we further demonstrate the utility of using C656G GR to eliminate the function of endogenous GR. This Dex-sensitive GR allowed us to identify the transactivation domains of GR necessary for transcription of the PEPCK gene, which was heretofore not possible. Instead of possessing high affinity GREs, the PEPCK gene promoter has two low affinity GR binding sites, GR1 and GR2, and thus requires accessory factor activity to stabilize GR binding and to recruit coactivators to the promoter. We have identified a residue, Glu773, that directs the interaction of GR with SRC-1 in the context of the PEPCK gene promoter. Mutation of this residue does not disrupt activation of the PEPCK gene if the low affinity GR1 site is changed to a high affinity, palindromic GRE. These data suggest that a context-dependent conformation of GR is essential for the recruitment of SRC-1 to the PEPCK gene promoter and efficient glucocorticoid-mediated induction of this gene. We thank Catherine Caldwell for excellent technical assistance and Deborah Brown for manuscript preparation.