A second pathway for modulating glucocorticoid receptor transactivation properties

Molecular and Cellular Endocrinology 199 (2003) 129
/142 www.elsevier.com/locate/mce

A second pathway for modulating glucocorticoid receptor transactivation properties Shiyou Chen 1, S. Stoney Simons, Jr * Steroid Hormones Section, NIDDK/LMCB, National Institutes of Health (NIH), Building 8, Room B2A-07, Bethesda, MD 20892, USA Received 23 May 2002; accepted 14 August 2002

Abstract We recently reported that three factors (a cis -acting element and changing concentrations of receptor or coactivator TIF2) act at a common rate-limiting step to modulate the position of the dose
/response curve and the partial agonist activity of glucocorticoid receptors (GRs). The ability of saturating levels of GR, and added inhibitors, to prevent the actions of the three modulators (cis acting element, GR, and TIF2) but not the currently investigated C-terminal fragment of E1A-13S (E1A-133C) indicates that E1A133C alters GR properties via a second pathway that is downstream of the common step for the original three modulators. hSur2 binds to E1A-133C. We find that hSur2 modulates GR transactivation properties, thus suggesting that the effects of E1A-133C are due to the recruitment of hSur2. hSur2 also modifies GR activities in the presence of saturating GR concentrations, which is consistent with hSur2 acting downstream of the common step for the original three modulators. The H160Y mutation, which eliminates hSur2 binding to E1A, blocks most of the activity of E1A-133C. This suggests that the modulatory activity of E1A-133C is largely due to the binding of hSur2, which is a component of the Mediator complex. Collectively, these data support the existence of a new pathway for modulating GR transactivation processes, thereby increasing the number of cellular mechanisms that permit differential control of gene expression by endogenous levels of glucocorticoid hormones. Published by Elsevier Science Ireland Ltd. Keywords: Glucocorticoid receptor; Modulation of dose
/response curve; Partial agonist activity; E1A; hSur2

1. Introduction A prime feature of steroid hormone action is the ability of elevated levels of hormone to induce the transcription of target genes. The basic steps by which this is accomplished have been known for many years (Beato et al., 1996; Tsai and O’Malley, 1994). These steps include steroid entry into the cell by passive diffusion and binding to the intracellular receptor protein, association of the receptor
/steroid complex with biologically active DNA sequences (called hormone response elements or HREs), and recruitment of an array of cofactors that increase (for coactivators),

* Corresponding author. Tel.: /1-301-496-6796; fax: /1-301-4023572. E-mail address: [email protected] (S.S. Simons, Jr). 1 Present address: Department of Cell Biology, Georgetown University, Med-Dent Bldg., SW204, 3900 Reservoir Road, Washington, DC 20007, USA.

decrease (for corepressors), or facilitate (for coregulators or comodulators) gene expression (Glass and Rosenfeld, 2000; McKenna et al., 1999; Robyr et al., 2000). Many of these cofactors are associated with a variety of other molecules in large, multiprotein complexes (Fondell et al., 1996; Heinzel et al., 1997; Ogryzko et al., 1998; Rachez et al., 1999). Furthermore, multiple forms of these complexes exist, each with a slightly different composition (Ito et al., 1999; Jones et al., 2001; McKenna et al., 1998). Precisely how the plethora of transcription factors influence the total level of transactivation by steroid hormones is not yet clear. We have been equally interested in whether other properties of steroid receptors are affected. One important property of steroid receptor action is the sensitivity of induction, such as the concentration of ligand required for half of the maximal induction, which is usually called the EC50. Due to how the EC50 is calculated, the value of the EC50 is independent of the magnitude of the basal level of

0303-7207/02/$ - see front matter. Published by Elsevier Science Ireland Ltd. PII: S 0 3 0 3 - 7 2 0 7 ( 0 2 ) 0 0 3 3 3 - 7

130

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

gene expression, the total amount of induced gene expression, and the fold induction. The current model of steroid hormone action suggests that all target genes of a given receptor
/steroid complex have the same EC50 and are half maximally induced at the same concentration of steroid. However, numerous reports indicate that the position of the dose
/response curve, or the EC50, for all genes regulated by a given receptor
/steroid complex is not the same (Simons et al., 1992) and in some cases can be modulated (Chen et al., 2000; Giannoukos et al., 2001; Oshima and Simons, 1992a,b; Oshima et al., 1995; Song et al., 2001; Szapary et al., 1996, 1999). This modulation provides a simple mechanism for differentially controlling gene transcription: steroid-regulated genes that are induced to an equal level, but with different EC50s, will afford unequal amounts of response to the same concentration of endogenous steroid, even within the same cell. Another property of steroid receptors is their response to antagonists, which bind to the receptor and block the action of agonist steroids. Antagonists are clinically useful for inhibiting the actions of endogenous hormones in inflammation (Vayssiere et al., 1997), conception (Baulieu, 1989), and hormone dependent cancers (Crawford et al., 1989; Jordan and Murphy, 1990). In many cases, however, antagonists are multifunctional and possess partial agonist activity (Berry et al., 1990; Mercier et al., 1986; Oshima and Simons, 1992b; Tzukerman et al., 1994; Wasner et al., 1988). Restricting the activity of antagonists to those genes that one wants to suppress while maintaining partial agonist activity for other genes would greatly reduce the side effects of endocrine therapies that usually result from the indiscriminate repression of all responsive genes. It should be noted that the partial agonist activity, like the EC50, is independent of basal and induced levels, and of fold induction, because it is defined as the amount of activity above basal levels divided by the maximally induced activity of a full agonist. We have recently reported that overexpressed coactivators (TIF2/GRIP1 (Hong et al., 1996; Voegel et al., 1996), SRC-1 (Onate et al., 1995), and AIB1 (Anzick et al., 1997)), corepressors (NCoR (Horlein et al., 1995) and SMRT (Chen and Evans, 1995)), and homologous receptors modulate the position of the dose
/response curve of agonists, and the partial agonist activity of antagonists, for both glucocorticoid receptors (GRs; Chen et al., 2000; Song et al., 2001; Szapary et al., 1996, 1999) and progesterone receptors (PRs; Giannoukos et al., 2001; Song et al., 2001). Also, the presence of a cis acting DNA element, called a glucocorticoid modulatory element (GME) and originally isolated from the rat tyrosine aminotransferase gene (Oshima and Simons, 1992b), can alter the dose
/response curve, and partial agonist activity of antagonists, with endogenous GR (Chen et al., 2000; Collier et al., 1996; Jackson et al.,

1998; Oshima and Simons, 1992b, 1993; Plisov et al., 1998). Similarly, Reichardt et al. (2000) described transgenic mice with a 2-fold increase in GR gene dosage, which causes an increase of 60% or less in receptor protein and simultaneously produces a greater than 10-fold decrease in the EC50 for induction of thymocyte apoptosis. These two examples suggest that both a cis -acting element and changing levels of transcription factor can alter GR induction properties not only in transiently transfected cells but also under in vivo conditions. Additional studies indicated that all three of the above components act on GR via a common rate-limiting step or intermediate ‘X’. First, the ability of each of these species (GME and changing concentrations of GR or TIF2) to modulate the GR dose
/ response curve, and partial agonist activity of antagonists, is mutually competitive (Chen et al., 2000). These observations, plus the fact that the cis -acting GME element influences only GRE-bound receptors (Oshima and Simons, 1993), also argue that the rate-limiting step ‘X’ is after those steps that are directly dependent upon GR or TIF2 (e.g. amount and/or activity of GR or TIF2 protein, affinity of steroid binding to GR, and levels of DNA-bound GR). Second, the action of each component (GME, GR, and TIF2) is inhibited by overexpression of both a fragment of TIF2 and the 13S form of the adenovirus protein, E1A (E1A-13S; Chen et al., 2000). The inhibitory fragment of TIF2 contains one of the two transactivation domains (Voegel et al., 1996) and presumably acts by binding/inactivating a limiting factor, i.e. squelching. The causes for inhibition by E1A-13S are less obvious and may involve one of the many other proteins that interact with E1A-13S (Flint and Shenk, 1997). Subsequent preliminary studies with a C-terminal fragment of E1A (E1A-133C; Chakravarti et al., 1999) suggested that this fragment has properties diametrically opposed to those of the full length E1A and more like the coactivators. The purpose of this study, therefore, was to further examine the modulatory pathway that might be affected by E1A-133C. We now report that E1A-133C is indeed capable of modifying GR transcriptional properties, thus making E1A-133C a valuable synthetic probe of the mechanism of GR-mediated transactivation. Furthermore, we have identified an endogenous protein (hSur2) that not only binds to E1A-133C but also modulates GR activities in the absence of exogenous E1A-133C. The previously implicated, common intermediate ‘X’ in the actions of the GME, and increasing concentrations of GR or TIF2, are not rate-limiting for either hSur2 or E1A-133C, indicating that their action is mechanistically different from that of the GME, GR, and TIF2. A model for this second pathway of modulating both the dose
/response curve of GR-agonist complexes and the partial agonist

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

activity of GR-antagonist complexes is proposed which incorporates all of these results.

2. Materials and methods Unless otherwise indicated, all operations were performed at 0 8C. 2.1. Chemicals Dexamethasone (Dex) was purchased from Sigma. Dex-21-mesylate (Dex-Mes; Simons et al., 1980) and Dex-oxetanone (Dex-Ox; Pons and Simons, 1981) were synthesized as described. Yeast media (BIO-101 Inc; Vista, CA), 3-amino-l,2,4-triazole (Sigma, St Louis, MO), and 5-fluroorotic acid (Life Technologies, Rockville, MD) are commercially available. 2.2. Plasmids The Renilla null luciferase reporter was purchased from Promega (Madison, WI). (UAS)5tkLuc was from Clontech (Palo Alto, CA). GREtkLUC has been previously described (Sarlis et al., 1999). pSVLGR (Keith Yamamoto, UCSF, San Francisco), TIF2 and T1F2.7 (Hinrich Gronemeyer, IGBMC, Strasbourg), E1A-13S, E1A-13S/133C (originally called E1A-13S-N132) and E1A-13S/N108 (originally called E1A-13S-C109; Debatra Chakravarti, University of Pennsylvania, Philadelphia), and hSur2 (Arnold Berk, UCLA, Los Angeles) were received as gifts. E1A-133C/H160Y was generated by PCR using the overlap extension method. The template was obtained by digesting E1A-13S/133C plasmid with BamHI and EcoRI. The coding sequence of H160 (CAC) was converted to TAG using the 3? primer of 5?-ATCGAATTCTTATGGCCTGGGGCGTTTACAGCT-3? and the 5? primer of 5?-GAGCACCCCGGGCACGGTTGCAGGTCTTGTCATTATTACCGGAGGAATACGGGGG-3?. PCR amplified DNA fragments containing the mutated sites were recovered after agarose gel electrophoresis, purified with QiaGen Gel Remove kit, and then ligated to the other fragment from BamHI/ EcoRI-digested E1A-133C. The entire PCR-generated region was verified by sequence analysis (Bioserve, Laurel, MD).

131

nologies, Inc.) for 5 h, or FugeneTM 6 transfection reagent (Roche Diagnostics Corp., Indianapolis, IN) overnight, with GR-containing plasmid (pSVLGR), 1.5 mg of GREtkLUC, and 200 ng Renilla null luciferase (the total transfected DNA was adjusted to 3 mg/60 mm dish with pBSK/ DNA). For 24 well plates, all amounts of DNA were reduced to 10% of that used in the 60 mm dishes. In experiments with varying amounts of coregulator cDNA plasmids, equimolar amounts of the empty vector were cotransfected to control for artifacts of the vector DNA. With Lipofectamine, the medium was changed after 5 h and the cells were incubated overnight before being treated (24 h) with 1% ethanol9/steroids in media containing 10% FBS and then harvested in 1 /Passive Lysis Buffer (0.6 ml per dish, 0.15 ml per well, Promega). Aliquots (50 ml) of the cell lysates were used to assay for luciferase activity using the Dual-Luciferase Assay System from Promega according to the supplier’s recommendations. The relative Luciferase activity values were normalized for the cotransfected Renilla activity, to correct for differences in transfection efficiency, before being graphed. 2.4. Statistics Unless otherwise noted, all experiments were performed in triplicate several times. The error bars in graphs of individual experiments correspond to the S.D. of the triplicate values. KALEIDAGRAPH 3.5 (Synergy Software, Reading, PA) was used to determine a leastsquares best fit (R2 was ]/0.95) of the theoretical dose
/ response curve, given by the equation derived from Michaelis
/Menton kinetics of y /[free steroid]/([free steroid]/Kd) (where the concentration of total steroid is approximately equal to the concentration of free steroid because such a small portion is actually bound), to the experimental data to yield a single EC50 value. The EC50 values of n independent experiments are then analyzed for statistical significance by the two-tailed Student’s ttest using the program ‘INSTAT 2.03’ for Macintosh (GraphPad Software, San Diego, CA). In paired analyses, all comparisons in a given figure are performed at the same time with the same seeding of cells and then analyzed for n different experiments.

3. Results

2.3. Cell culture and transient transfections

3.1. Modulatory activity of E1A-133C

CV-1 cells were grown in Dulbecco’s modified Eagle medium (DMEM; Life Technologies, Inc.) supplemented with 10% fetal calf serum (FCS; Biofluids Inc., Rockville, MD) at 37 8C in a humidified incubator (5% CO2). Coregulator plasmids were transiently co-transfected into CV-1 cells using Lipofectamine (Life Tech-

Both the 13S form of E1A and the amino terminal fragment E1A-N108 (Fig. 1A) cause a right-shift in the dose
/response curve for Dex induction of a transfected GREtkLUC reporter, and a decrease in the partial agonist activity of the antiglucocorticoid Dex-Mes, by GR in CV-1 cells (Chen et al., 2000). It was, therefore,

132

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

Fig. 1. Ability of E1A fragments to modulate GR induction properties. (A) Structures are of the full length, and truncated, E1A-13S constructs used in this study, along with the various domains (N and CR1
/3) that have been identified (Chakravarti et al., 1999). Numbers above the top structure correspond to the amino acid of selected positions. (B) Modulation of GR dose
/response curve and partial agonist activity of antiglucocorticoids by E1A-133C. Triplicate plates of CV-1 cells were transiently transfected in 60 mm dishes with 1.5 mg of GREtkLUC and 40 ng GR with or without 200 ng of E1A-133C. In both series, the total molar amount of E1A vector (pCMX) was constant. Samples were then treated with the indicated concentrations of Dex or 1 mM Dex-Mes and the induced luciferase activities above basal levels (EtOH) determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The error bars indicate the S.D. of triplicate values in a representative experiment. (C) Effects of E1A-133C are saturable. Triplicate plates of CV-1 cells were transiently transfected as in (B) with 40 ng of GR, and 0, 200, or 1800 ng of E1A-133C, and treated with the indicated concentrations of Dex, or 1 mM Dex-Mes. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The error bars indicate the S.D. of triplicate values.

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

surprising to find that the C-terminal fragment of E1A (E1A-133C; Fig. 1A) affords the opposite behavior to give a 3.29/0.4-fold (S.E.M., n /9, P /0.0006) left-shift in the dose
/response curve and a 2.49/0.2-fold (S.E.M., n /9, P /0.0002) increase in the partial agonist activity of Dex-Mes (Fig. 1B). Similarly, the partial agonist activity of the antiglucocorticoid Dex-Ox (Lamontagne et al., 1984) was increased 2.49/0.3-fold (range, n/2) by E1A-133C (data not shown). These changes in the Dex dose
/response curve, and Dex-Mes partial agonist activity, depend upon the concentration of E1A-133C (data not shown). However, concentrations of E1A133C above 0.2 mg are without any additional effect (Fig. 1C), indicating that a plateau has been reached with 0.2 mg of E1A-133C plasmid. In fact, 0.2 mg gives the same values as 0.05 mg of E1A-133C for EC50 (1.049/0.12-fold increase for 0.2 vs. 0.05 mg E1A-133C) and Dex-Mes partial agonist activity (1.069/0.18-fold increase for 0.2 vs. 0.05 mg E1A-133C; both values9/ S.E.M., n /4). In addition to the above effects, E1A-133C also increases the total activity of 1 mM Dex by 5.89/0.7fold (S.E.M., n /12) while reducing the fold induction by a factor of 0.559/0.03 (S.E.M., n /12) due to an average 10.5-fold increase in basal level activity. These changes are important characteristics of E1A-133C (see also below) but are unrelated to the modulatory activity of E1A-133C in Fig. 1B and C. As discussed in Section 1 and elsewhere (Chen et al., 2000; Oshima and Simons, 1993; Song et al., 2001; Szapary et al., 1996, 1999; Zeng et al., 2000), it should be appreciated that the ability of a factor to modulate the position of the dose
/response curve and the amount of partial agonist activity is independent from the ability of that factor to change the basal level activity, the total levels of induced activity, and the fold induction. We also note that the left-shift seen in Fig. 1B is not due simply to the increased basal activity in the presence of E1A-133C, which can occur for genes with an EC50 below 1 nM Dex due to low levels of endogenous glucocorticoids in FCS (Mercier et al., 1983); just raising the arithmetic value of the basal level would cause a right-shift in the dose
/response curve. Similarly, our conclusions here and below are not compromised by the increases in basal level that occur, due to low levels of endogenous glucocorticoids in FCS, with genes that display very low EC50s (i.e. B/1 nM Dex (Mercier et al., 1983)) 3.2. Effects of E1A-133C are non-competitive with GR and TIF2 The ability of added GR or TIF2 to shift the Dex dose
/response curve to the left, or to increase the partial agonist activity of Dex-Mes, is not infinite but reaches a plateau value at high concentrations of GR or TIF2 (Chen et al., 2000). Furthermore, the effects of GR and

133

TIF2 are mutually competitive in that TIF2 is unable to increase the activities seen with saturating amounts of GR and vice versa. These results argue that added GR and TIF2 modulate the induction properties of GR via a common intermediate or step ‘X’ (Chen et al., 2000). We employed this kinetic approach to ask whether E1A133C modulates GR induction properties through the same rate-limiting step as GR and TIF2. First, we asked if E1A-133C can increase the activities of a subsaturating concentration of Dex, and saturating concentration of Dex-Mes, that are obtained with saturating levels of GR. The data of Fig. 2A indicate that E1A-133C causes a further shift in the Dex dose
/response curve, and an additional increase in the partial agonist activity of DexMes (1.449/0.10-fold, S.D., n/3, P /0.017), under conditions where added GR has no effect (Chen et al., 2000). Therefore, the presence of saturating amounts of GR does not prevent E1A-133C from further modifying the induction properties of GR. We next inquired if saturating concentrations of E1A133C can block the ability of TIF2 to modify GR induction properties and vice versa (Chen et al., 2000; Szapary et al., 1999). As shown in Fig. 2B for the average values of four to six experiments, 50 ng of E1A133C produces the maximal responses because increasing the amount of transfected E1A-133C plasmid to 200 ng does not afford any supplementary shift in the dose
/ response curve or increase in the partial agonist activity of Dex-Mes. In contrast, the addition of 200 ng of TIF2 plasmid, which is a saturating amount of TIF2 in CV-1 cells (Chen et al., 2000), to samples with 200 ng of E1A133C produced a 1.959/0.25 (9/S.E.M., n/5, P / 0.019) fold left-shift in the dose
/response curve and a 50% increase in the partial agonist activity of Dex-Mes (n /7, paired P /0.0009). These data indicate that saturating concentrations of E1A-133C do not prevent TIF2 from further modifying GR induction properties, and vice versa. Together, the above series of experiments argue that there is no competitive inhibition between E1A-133C and GR or TIF2. From these kinetic results, we conclude that the rate-limiting step, or intermediate, for E1A-133C modulation of the GR dose
/response curve and partial agonist activity of antagonists is different from that of ‘X’, which is utilized by GR and TIF2 (Chen et al., 2000). Furthermore, the absence of competitive inhibition between E1A-133C and GR or TIF2 implies that the rate-limiting step, or intermediate, for E1A-133C action is downstream of ‘X’. These results also exclude the possibility of E1A-133C indirectly affecting GR or coactivator. Any effects of E1A-133C on GR or coactivator levels (due to synthesis, assembly, proteolysis, etc.) either would have no effect (if GR or coactivator levels are increased) or would cause a rightshift in the dose
/response curve and a decrease in the partial agonist activity (if GR or coactivator levels are

134

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

Fig. 2. Ability of saturating concentrations of GR orTIF2 to compete with E1A-133C for the modulation of GR transactivation properties. (A) Activity of E1A-133C in the presence of saturating concentrations of GR. Triplicate samples of CV-1 cells were transiently transfected as in Fig. 1B with 900 or 1400 ng GR cDNA plus pCMX vector, or 1400 ng GR plus 200 ng of E1A-133C, and treated with the indicated concentrations of Dex, or 1 mM Dex-Mes. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The error bars indicate the S.D. of triplicate values. Similar results were obtained in two other experiments. (B) Activity of E1A-133C in the presence of saturating concentrations of TIF2. Triplicate plates of CV-1 cells were transiently transfected as in Fig. 1B with 40 ng of GR and 0, 50, or 200 ng E1A-133C, or 200 ng E1A-133C plus 200 ng of TIF2 cDNA. In each case, the molar amount of pCMX (for E1A-133C) and pSG5 (for TIF2) vector was adjusted to be constant. The samples were treated with the indicated concentrations of Dex, or 1 mM Dex-Mes. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The average values of four to seven experiments (9/S.E.M.) were combined and plotted. The symbol ** indicates P/ 0.0009 for 200 ng E1A-133C9/200 ng TIF2 plasmids. See text for relevance of vertical dashed line.

decreased). Instead, we see a further left-shift and increased partial agonist activity with E1A-133C. 3.3. Ability of inhibitors of modulation of GR induction properties to block the actions of E1A-133C Both TIF2.7 (Voegel et al., 1998), a fragment of TIF2 containing predominantly the first activation domain (AD1), and E1A-N108, the N-terminal 108 amino acids of E1A, were found to inhibit the modulation of the GR dose
/response curve, and partial agonist activity of

antagonists, by GR, GME, and TIF2. As both inhibitors block GR, GME, and TIF2 actions, they are thought to act at steps downstream from the common intermediate ‘X’ that is utilized by GR, GME, and TIF2 (Chen et al., 2000). We, therefore, inquired whether these inhibitors could similarly block the actions of E1A-133C, which also influences a step(s) downstream of ‘X’. In these experiments, we quantitated the effects of the inhibitors on the activity of a single subsaturating concentration of Dex, expressed as percent of maximal induction by saturating concentrations of Dex, instead

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

of performing complete dose
/response curves. This abbreviated assay is, however, just as revealing. As seen from Fig. 2B, a decrease in the activity of 1 nM Dex, when expressed as percent of maximal response (indicated by the dashed line), corresponds to a rightshift in the dose
/response curve. Using this approach, E1A-N108 is found to be an inhibitor of the actions of E1A-133C (Fig. 3A). However, E1A-N108 only partially blocks the effects of E1A-133C, as seen by the ability of E1A-133C to significantly increase the activities obtained in the presence of E1A-N108. The other inhibitor, TIF2.7, is not as effective as E1A-N108 (Chen et al., 2000). This lesser inhibitory activity of TIF2.7, even with higher amounts of TIF2.7 plasmid (600 ng), was confirmed in the present study (compare addition of TIF2.7 in Fig. 3B vs. N108 in Fig. 3A). High concentrations of TIF2.7 do blunt the actions of E1A-133C (Fig. 3B). As with E1A-N108, though, the inhibitory activity of TIF2.7 is partially overcome by added E1A133C. The fact that neither inhibitor is as effective in blocking the effects of E1A-133C as they were in inhibiting the modulatory effects of added GR or TIF2 (Chen et al., 2000), even with nearly saturating concentrations of TIF2.7, suggests that the inhibitors affect a step upstream of the site of action of E1A-133C and downstream of the common step X, which is shared by added GR and TIF2.

135

3.4. Modulation of GR induction properties by an E1A133C associated protein One of the proteins that binds to the CR3 region of E1A, and to E1A-133C (Fig. 1A), is hSur2 (Boyer et al., 1999). Interestingly, hSur2 is also part of a multiprotein complex that both appears to be the human homolog of the yeast complex Srb/Mediator and is capable of stimulating transcription in vitro with added E1A (Boyer et al., 1999; Wang et al., 2001). The mouse Mediator complex is closely related to the TRAP/DRIP/ SMCC complex (Ito et al., 1999). We, therefore, inquired whether hSur2 has intrinsic modulatory activity and could be responsible for the above actions of E1A-133C. Overexpression of hSur2 in CV-1 cells produces a 419/5% (S.E.M., n /10, P B/0.0001) decrease in basal level activity of the GREtkLUC reporter and a 349/11% (S.E.M., n /10, P /0.056) reduction in total transactivation by GR and 1 mM Dex. As a result, the inducibility by GR is slightly increased (1.309/0.11fold, S.E.M., n/10, P /0.025). When the ability of hSur2 to modulate GR induction properties was examined, we found that hSur2 causes a left-shift in the dose
/ response curve (Fig. 4A, average /2.519/0.53-fold, S.E.M., n/10, P /0.019). At the same time, the partial agonist activity of the antagonists Dex-Mes and Dex-Ox are increased by 869/31% (P /0.021) and, 799/24%

Fig. 3. Influence of known inhibitors of the modulation of GR induction properties on the actions of E1A-133C. Triplicate plates of CV-1 cells were transiently transfected as in Fig. 1B with either 100 ng of GR9/200 ng E1A-133C and/or E1A-N108 for (A) or 100 ng of GR9/200 ng E1A-133C and/ or the indicated amounts of TIF2.7 for (B). In each case, the molar amount of pCMX (for E1A constructs) and pSG5 (for TIF2.7) vector was adjusted to be constant. The samples were treated with 0.1 or 1000 nM Dex, or 1 mM Dex-Mes. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The average values of 2 (9/range) or 3 (9/S.E.M.) experiments were combined and plotted for (A) and (B), respectively.

136

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

Fig. 4. Ability of hSur2 to replace E1A-133C in modulating GR transactivation properties. (A) Modulation of GR dose
/response curve and partial agonist activity of antiglucocorticoids by hSur2. Triplicate plates of CV-1 cells were transiently transfected as in Fig. 1B with 100 ng of GR and either 1600 ng of hSur2 cDNA or a molar equivalent of the vector (pCS2/). The samples were treated with the indicated concentrations of Dex, or 1 mM antiglucocorticoid (Dex-Mes or Dex-Ox). The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. (B) Activity of hSur2 in the presence of saturating concentrations of GR. Triplicate samples of CV-1 cells were transiently transfected as in Fig. 1B with 100, 900, or 1400 ng GR cDNA plus pCS2/vector, or 900 ng GR plus 1600 ng of hSur2 cDNA, and treated with the indicated concentrations of Dex, or 1 mM Dex-Mes or Dex-Ox. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The error bars indicate the S.D. of triplicate values in a representative experiment.

(P /0.01), respectively (S.E.M., n /10). It thus appears that Sur2 reproduces the effects of E1A-133C. However, no synergism is seen after co-transfection of E1A-133C and hSur2. Instead, the addition of 200
/400 ng of hSur2 to cells with 10
/200 ng of E1A-133C produces a 479/9% (P /0.012) decrease in the activity of 1 nM Dex (indicative of a right shift in the dose
/response curve) and a 279/8% (P /0.049, n /4, S.E.M.) reduction in the partial agonist activity of 1 mM Dex-Mes (data not shown). A comparison of the GR transactivation responses to added E1A-133C and hSur2 is particularly revealing. First, E1A-133C and hSur2 are expressed from different

vectors (pCMX and pCS2/, respectively) and afford the same responses relative to internal vectors controls. Therefore, the modulation of GR activities by E1A133C and hSur2 is not an artifact of a specific vector. Second, Sur2 produces just the opposite effects of E1A133C on the basal and fully induced levels: both are decreased by Sur2 and increased by E1A-133C. Nevertheless, Sur2 and E1A-133C each cause the same leftshift of the dose
/response curve and increase in partial agonist activity. This definitively establishes a lack of correlation between the changes in basal activity (and total transactivation) by E1A-133C and hSur2 and their ability to modulate the position of the dose
/response

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

curve (and partial agonist activity) for GR transactivation. To further test the hypothesis that the activity of E1A-133C is mediated by hSur2, we asked if the ability of hSur2 to augment GR induction properties still occurs in the presence of saturating concentrations of GR, just as is seen for E1A-133C (see Fig. 2A). As shown in Fig. 4B, raising the total GR plasmid from 0.9 to 1.4 mg causes negligible further left shift in the dose
/ response curve (1.129/0.04-fold, S.E.M., n /6) or increase in the partial agonist activity of Dex-Mes (0.989/0.06-fold) or Dex-Ox (1.029/0.09-fold, S.E.M., n /6
/7). However, cotransfection of hSur2 causes an additional left-shift in the dose
/response curve (2.419/ 0.26-fold, P /0.0028), and a greater increase in the partial agonist activity of both Dex-Mes (1.389/0.08fold, P /0.0025) and Dex-Ox (1.329/0.10-fold, P / 0.025, S.E.M., n /6
/7). In all cases, the changes seen with 0.9 mg GR plus 1.6 mg hSur2 are also significantly different from those seen with 1.4 mg GR plasmid at P 5/0.046. We, therefore, conclude that GR does not competitively inhibit the actions of hSur2. In this respect, exogenous hSur2 reproduces the behavior of added E1A-133C in being able to modulate GR transactivation properties via a pathway that does not involve the intermediate ‘X’. These results also suggest that the actions of E1A-133C are mediated, at least in part, by hSur2 bound to E1A-133C. 3.5. Modulatory activity of E1A-133C with H160Y mutation The binding of hSur2 to the CR3 region of E1A is eliminated by the single point mutation H160Y (Boyer et al., 1999; Webster and Ricciardi, 1991). If all of the activity of E1A-133C is mediated by hSur2, we would predict that the mutant protein, E1A-133C/H160Y, would be inactive. By all criteria, E1A-133C/H160Y has markedly different properties than E1A-133C. The effects of E1A-133C/H160Y are much smaller than with E1A-133C for the basal activity (2.49/0.4 vs. 8.49/1.2fold increase (P /0.0006 by paired Mann
/Whitney test), n/4) and total transactivation (2.39/0.3 vs. 4.29/0.7-fold increase (P /0.0062 by paired t-test) n / 7, S.E.M. for all). The presence of the H160Y mutation causes no change in the fold induction by Dex (1.039/ 0.11-fold) as opposed to a 509/6% decrease with E1A133C (P /0.0009). Most importantly, the representative experiment of Fig. 5 shows that the ability of E1A-133C to shift the dose
/response curve to the left, and to increase the partial agonist activity, is significantly reduced by the H160Y mutation. In seven paired experiments, the magnitude of the left-shift was decreased from 2.909/0.31 to 1.489/0.17-fold (P /0.0015). At the same time, the H160Y mutation reduced the partial agonist activity of Dex-Mes from 529/3 to 419/

137

5% (paired P /0.03; activity in the absence of E1A133C /299/4%; all error limits are S.E.M.). The reduced activity of the H160Y mutant is probably not due to lower expression levels of the functional protein as higher amounts of transfected H160Y displayed no more activity (data not shown). These results suggest that the majority of the activity of E1A-133C is expressed through hSur2.

4. Discussion We recently determined that both the positioning of the dose
/response curve, or EC50, for gene induction by glucocorticoids and the partial agonist activity of antiglucocorticoids are modified via a common intermediate ‘X’ in response to three factors (Chen et al., 2000): GME (Chen et al., 2000; Collier et al., 1996; Jackson et al., 1998; Kaul et al., 2000; Oshima and Simons, 1992b), GR (Chen et al., 2000; Song et al., 2001; Szapary et al., 1996, 1999), and coactivators (Chen et al., 2000; Szapary et al., 1999). We now report that a Cterminal fragment of E1A (E1A-133C) and hSur2 also modulate these transactivation properties of GR. In contrast to the behavior of the originally described three modulators, saturating concentrations of GR or TIF2 are unable to prevent hSur2 or E1A-133C from producing additional changes in the GR dose
/response curve or partial agonist activity of antagonists and vice versa (Figs. 2 and 4B). Also, the inhibitors E1A-N108 and TIF2.7 are much less efficient blockers of the actions of E1A-133C than of the GME and added GR or TIF2 (Fig. 3 vs. Fig. 5 and Fig. 7 of Chen et al., 2000). Collectively, these results provide strong kinetic evidence that the actions of hSur2, and E1A-133C, do not proceed through the same rate-limiting step or intermediate ‘X’ that is employed by GR, GME, and TIF2 but utilize a second pathway or branch. These results also eliminate possible mechanisms where hSur2 and E1A-133C increase GR or coactivator levels because added GR or TIF2 are unable to cause any further effects. Decreases in GR or coactivator levels would cause a right shift in the dose
/response curve, and lower amounts of partial agonist activity (Szapary et al., 1999), which is exactly opposite from what is observed with added hSur2 and E1A-133C. The existence of a second pathway was first identified from studies with E1A-133C, which is a C-terminal fragment of E1A-13S (Chakravarti et al., 1999). hSur2 has recently been reported to bind to the CR3 domain of E1A (Boyer et al., 1999), which is retained in E1A-133C (Fig. 1A). Two additional lines of evidence argue that the actions of E1A-133C are largely due to its ability to bind hSur2. First, hSur2 is independently able to modify GR transactivation properties in a manner that is not competed by excess GR (Fig. 4B), just as is seen for

138

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

Fig. 5. Mutation of hSur2 binding site of E1A-133C reduces the modulatory activity of E1A-133C. Triplicate samples of CV-1 cells were transiently transfected as in Fig. 1B with 40 ng GR and 200 ng of E1A-133C9/H160Y plus the appropriate molar amounts of vector and treated with the indicated concentrations of Dex or 1 mM Dex-Mes. The induced luciferase activities above basal levels (EtOH) were determined and expressed as percent of the maximal activity with 1 mM Dex, as described in Section 2. The error bars indicate the S.D. of triplicate values in a representative experiment.

E1A-133C (Fig. 2A). Second, the H160Y mutation of E1A, which prevents hSur2 binding to E1A (Boyer et al., 1999; Webster and Ricciardi, 1991), simultaneously eliminates the majority of the activity of E1A-133C (Fig. 5). Thus, there appear to be two ways in which this second pathway for modulating GR activities can be accessed: changing the levels of (1) endogenous hSur2 or (2) exogenous E1A-133C. We also note that hSur2 and E1A-133C have identical effects on the dose
/response curve of agonists and the partial agonist activity of antagonist while being expressed from different vectors and having diametrically opposite effects on the basal and fully induced levels of GR-mediated gene transactivation. Therefore, we conclude that the effects of hSur2 and E1A-133C are specific and are independent both of the expression vector and of any additional changes in basal or fully induced gene expression. This same independence of changes in the GR dose
/response curve, and partial agonist activity, from variations in the total induced activity has been previously documented (Chen et al., 2000; Oshima and Simons, 1993; Song et al., 2001; Szapary et al., 1996, 1999; Zeng et al., 2000). This second pathway, which is utilized by hSur2 and E1A-133C, appears to join the pathway followed by GR, GME, and TIF2 at some point after the intermediate ‘X’ but before the observed changes in reporter activity (Fig. 6A). As ‘X’ is downstream of those steps that are affected by increased GR or TIF2 (e.g. changes in the levels of GR or TIF2 protein or activity and changes in amounts of DNA-bound GR), the effects of E1A-133C and hSur2 are also after these earlier steps. In fact, we predict that the site(s) of action of E1A-133C and hSur2 are so far downstream from the immediate

actions of GR and TIF2 that there is no physical contact between GR or TIF2 and E1A-133C or hSur2. A more precise mechanistic proposal is limited by the many unknown molecular details of both steroid receptor-mediated gene activation and hSur2 action. hSur2 is often a subunit of the Mediator complex (Boyer et al., 1999; Wang et al., 2001), which is a component of the RNA polymerase II holoenzyme (Myer and Young, 1998). Mediator is known to contact RNA polymerase II (Asturias et al., 1999). However, whether Mediator exists in cells only as part of a multi-component complex with RNA polymerase II (Maldonado et al., 1996; Neish et al., 1998; Ossipow et al., 1995; Pan et al., 1997) or is recruited to the polymerase (Chiba et al., 2000; Naar et al., 1999; Wang et al., 2001) is currently unresolved. E1A-13S is known to interact through its CR3 domain with other transcription factors such as CBP (Eckner et al., 1994) and several subunits of TFIID including TBP (Flint and Shenk, 1997). Thus the actions of E1A-133C in our system could result from the recruitment of Sur2 to modify the activities of a component of the transcriptional machinery, most likely a subunit of Mediator, that is different from the above intermediate ‘X’ (Fig. 6B). Alternatively, E1A-133C might increase the ability of hSur2 to titrate molecules that otherwise cause a right-shift in the dose
/response curve and a decrease in the partial agonist activity of antiglucocorticoids. We favor the former mechanism as Mediator complex containing hSur2, but not hSur2 alone, has appreciable stimulatory effects on in vitro transcription (Wang et al., 2001). Further studies of hSur2 and its interacting proteins of the Mediator complex are needed to determine additional mechanistic details.

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

139

Fig. 6. Model for modulation of GR transactivation properties by E1A-133C and associated molecules. (A) Rate-limiting step that is influenced by E1A-133C is different from that affected by other modulators of GR properties. The lack of mutual competitive inhibition of E1A-133C by excess GR orTIF2, which compete for each other and act through the common intermediate (X) (Chen et al., 2000), and the relative inactivity of the inhibitors TIF2.7 and E1A-N108 (Chen et al., 2000) each suggest that E1A-133C influences a step/intermediate that is downstream of ‘X’. The precise step that is inhibited by TIF2.7 and by E1A-N108 is shown to be the same only for ease of display and may be different. Similarly, the inability of excess GR to prevent hSur2 action suggests that hSur2 proceeds via the same pathway as E1A-133C. (B) Suggested molecular model of the species involved in E1A-133C modulation of GR induction properties. It is proposed that the intermediate ‘X’ that is the limiting factor with excess GREbound GR is an unidentified component of the transcriptional machinery. Different components comprise the Mediator complex (conglomerate of lightly shaded species). E1A-133C contacts yet other proteins in the transcriptional complex to recruit and/or activate Sur2, which interacts with Mediator. Sur2 and ‘X’ are proposed to alter the properties of other components and thereby eventually modulate the transactivation properties that are initiated by GR-bound agonists and partial antagonists. See text for further details.

The existence of a second pathway for modulating GR transactivation properties has significant mechanistic consequences. With widely separated dose
/response curves, or EC50s that differ by a factor of ]/ 20, a steroid concentration sufficient for more than 50% of maximal activity with the more sensitive gene will be inactive with the less sensitive gene (e.g. Fig. 4B). Changes in the partial agonist activity of antagonists are clinically important in the search for selective receptor modulators that will display fewer adverse side effects by retaining agonist activity for

some genes but not the one(s) to be repressed (Giannoukos et al., 2001; Song et al., 2001; Szapary et al., 1999). The ability to modulate these GR transactivation properties may convey many benefits to an organism. For example, a 2-fold increase in GR gene dosage in transgenic mice results in a 10-fold decrease in EC50 for the GR-mediated induction of thymocyte apoptosis (Reichardt et al., 2000). Therefore, it is perhaps not surprising that evolution has selected for multiple, possibly redundant, mechanistic pathways for such modulation in cells.

140

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

The ability of E1A-133C to cause a left-shift in the dose
/response curve, and an increase in the partial agonist activity of antiglucocorticoids, was unexpected in view of our earlier documentation of exactly opposite effects by the full length E1A (Chen et al., 2000). However, this phenomenon of reversed activities upon truncation of proteins is not unusual and has been seen in a variety of situations. SRC-1 is a potent coactivator of steroid receptor transactivation (Onate et al., 1995) but contains a strong repressive domain in the Cterminal 226 amino acids (Jenster et al., 1997). PSF is a corepressor that interacts with the DNA-binding domain of TR, and RXR, but displays intrinsic activation activity in the absence of its amino-terminal repressor domain (Mathur et al., 2001). Similarly, the nuclear receptor interacting protein NSD1 (Huang et al., 1998) along with the MYST family protein Tip60 (HIV TAT interacting protein of 60 kDa) and others (Sharma et al., 2000 and references therein) display activating or repressing activities, depending on what portions of the molecules are present. Finally, we have recently shown that not only GR, but also PR, induction properties can be modified by changing concentrations of receptor, coactivator, and corepressor (Giannoukos et al., 2001; Song et al., 2001). This indicates that the modulation of steroid receptor induction properties is not restricted to GR and may be general for the other steroid receptors. For this reason, it will be interesting to see if hSur2 and E1A-133C can also modulate PR induction processes via a second pathway.

Acknowledgements We thank Arnold Berk, Debatra Chakravati, G. Chinnadurai, Hinrich Gronemeyer, and Keith Yamamoto for generously providing plasmids, Roland Owens (NIDDK, NIH) for helpful discussions, and Yun-Bo Shi (NICHD, NIH) for critical review of the manuscript.

References Anzick, S.L., Kononen, J., Walker, R.L., Azorsa, D.O., Tanner, M.M., Guan, X.Y., Sauter, G., Kallioniemi, O.P., Trent, J.M., Meltzer, P.S., 1997. AIB1, a steroid receptor coactivator amplified in breast and ovarian cancer. Science 277, 965
/968. Asturias, F.J., Jiang, Y.W., Myers, L.C., Gustafsson, C.M., Kornberg, R.D., 1999. Conserved structures of mediator and RNA polymerase II holoenzyme. Science 283, 985
/987. Baulieu, E.E., 1989. Contragestion and other clinical applications of RU 486, an antiprogesterone at the receptor. Science 245, 1351
/ 1357. Beato, M., Chavez, S., Truss, M., 1996. Transcriptional regulation by steroid hormones. Steroids 61, 240
/251. Berry, M., Metzger, D., Chambon, P., 1990. Role of the two activating domains of the oestrogen receptor in the cell-type and promoter-

context dependent agonistic activity of the anti-oestrogen 4hydroxytamoxifen. EMBO J. 9, 2811
/2818. Boyer, T.G., Martin, M.E.D., Lees, E., Ricciardi, R.P., Berk, A.J., 1999. Mammalian Srb/mediator complex is targeted by adenovirus E1A protein. Nature 399, 276
/279. Chakravarti, D., Ogryzko, V., Kao, H.-Y., Nash, A., Chen, H., Nakatani, Y., Evans, R.M., 1999. A viral mechanism for inhibition of p300 and PCAF acetyltransferase activity. Cell 96, 393
/403. Chen, J.D., Evans, R.M., 1995. A transcriptional co-repressor that interacts with nuclear hormone receptors. Nature 377, 454
/457. Chen, S., Sarlis, N.J., Simons, S.S., Jr, 2000. Evidence for a common step in three different processes for modulating the kinetic properties of glucocorticoid receptor-induced gene transcription. J. Biol. Chem. 275, 30106
/30117. Chiba, N., Suldan, Z., Freedman, L.P., Parvin, J.D., 2000. Binding of liganded vitamin D receptor to the vitamin D receptor interacting protein coactivator complex induces interaction with RNA polymerase II holoenzyme. J. Biol. Chem. 275, 10719
/10722. Collier, C.D., Oshima, H., Simons, S.S., Jr, 1996. A negative tyrosine aminotransferase gene element that blocks glucocorticoid modulatory element-regulated modulation of glucocorticoid-induced gene expression. Mol. Endocrinol. 10, 463
/476. Crawford, E.D., Eisenberger, M.A., McLeod, D.G., Spaulding, J.T., Benson, R., Dorr, F.A., Blumenstein, B.A., Davis, M.A., Goodman, P.J., 1989. A controlled trial of leuprolide with and without flutamide in prostatic carcinoma. New Engl. J. Med. 321, 419
/424. Eckner, R., Ewen, M.E., Newsome, D., Gerdes, M., DeCaprio, J.A., Lawrence, J.B., Livingston, D.M., 1994. Molecular cloning and functional analysis of the adenovirus E1A-associated 300 kD protein (p300) reveals a protein with properties of a transcriptional adaptor. Genes Dev. 8, 869
/884. Flint, J., Shenk, T., 1997. Viral transactivating proteins. Annu. Rev. Genet. 31, 177
/212. Fondell, J.D., Ge, H., Roeder, R.G., 1996. Ligand induction of a transcriptionally active thyroid hormone receptor coactivator complex. Proc. Natl. Acad. Sci. USA 93, 8329
/8333. Giannoukos, G., Szapary, D., Smith, C.L., Meeker, J.E.W., Simons, S.S., Jr, 2001. New antiprogestins with partial agonist activity: potential selective progesterone receptor modulators (SPRMs) and probes for receptor- and coregulator-induced changes in progesterone receptor induction properties. Mol. Endocrinol. 15, 255
/270. Glass, C.K., Rosenfeld, M.G., 2000. The coregulator exchange in transcription functions of nuclear receptors. Genes Dev. 14, 121
/ 141. Heinzel, T., Lavinsky, R.M., Mullen, T.-M., So¨derstro¨m, M., Laherty, C.D., Torchia, J., Yang, W.-M., Brard, G., Ngo, S.D., Davie, J.R., Seto, E., Eisenman, R.N., Rose, D.W., Glass, C.K., Rosenfeld, M.G., 1997. A complex containing N
/CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 387, 43
/ 48. Hong, H., Kohli, K., Trivedi, A., Johnson, D.L., Stallcup, M.R., 1996. GRIP1, a novel mouse protein that serves as a transcriptional coactivator in yeast for the hormone binding domains of steroid receptors. Proc. Natl. Acad. Sci. USA 93, 4948
/4952. Horlein, A.J., Naar, A.M., Heinzel, T., Torchia, J., Gloss, B., Kurokawa, R., Ryan, A., Kamei, Y., So¨derstro¨m, M., Glass, C.K., Rosenfeld, M.G., 1995. Ligand-independent repression by the thyroid hormone receptor mediated by a nuclear receptor corepressor. Nature 377, 397
/404. Huang, N., vom Baur, E., Garnier, J.M., Lerouge, T., Vonesch, J.L., Lutz, Y., Chambon, P., Losson, R., 1998. Two distinct nuclear receptor interaction domains in NSD1, a novel SET protein that exhibits characteristics of both corepressors and coactivators. EMBO J. 17, 3398
/3412. Ito, M., Yuan, C.X., Malik, S., Gu, W., Fondell, J.D., Yamamura, S., Fu, Z.Y., Zhang, X., Qin, J., Roeder, R.G., 1999. Identity between TRAP and SMCC complexes indicates novel pathways for the

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142 function of nuclear receptors and diverse mammalian activators. Mol. Cell 3, 361
/370. Jackson, D.A., Collier, C.D., Oshima, H., Simons, S.S., Jr, 1998. Modulation of tat gene induction by glucocorticoids involves a neutralizing sequence. J. Steroid Biochem. Mol. Biol. 66, 79
/91. Jenster, G., Spencer, T.E., Burcin, M.M., Tsai, M.T., O’Malley, B.W., 1997. Steroid receptor induction of gene transcription: a two step model. Proc. Natl. Acad. Sci. USA 94, 7879
/7884. Jones, P.L., Sachs, L.M., Rouse, N., Wade, P.A., Shi, Y.B., 2001. Multiple N
/CoR complexes contain distinct histone deacetylases. J. Biol. Chem. 276, 8807
/8811. Jordan, V.C., Murphy, C.S., 1990. Endocrine pharmacology of antiestrogens as antitumor agents. Endocrinol. Rev. 11, 578
/610. Kaul, S., Blackford, J.A., Jr, Chen, J., Ogryzko, V.V., Simons, S.S., Jr, 2000. Properties of the glucocorticoid modulatory element binding proteins GMEB-1 and -2: potential new modifiers of glucocorticoid receptor transactivation and members of the family of KDWK proteins. Mol. Endocrinol. 14, 1010
/1027. Lamontagne, N., Mercier, L., Pons, M., Thompson, E.B., Simons, S.S., Jr, 1984. Glucocorticoid versus antiglucocorticoid activity: can a single functional group modification of glucocorticoid steroids always convey antiglucocorticoid activity. Endocrinology 114, 2252
/2263. Maldonado, E., Shiekhattar, R., Sheldon, M., Cho, H., Drapkin, R., Rickert, P., Lees, E., Anderson, C.W., Linn, S., Reinberg, D., 1996. A human RNA polymerase II complex associated with SRB and DNA-repair proteins. Nature 381, 86
/89. Mathur, M., Tucker, P.W., Samuels, H.H., 2001. PSF is a novel corepressor that mediates its effect through Sin3A and the DNA binding domain of nuclear hormone receptors. Mol. Cell. Biol. 21, 2298
/2311. McKenna, N.J., Nawaz, Z., Tsai, S.Y., Tsai, M.-J., O’Malley, B.W., 1998. Distinct steady-state nuclear receptor coregulator complexes exist in vivo. Proc. Natl. Acad. Sci. USA 95, 11697
/11702. McKenna, N.J., Lanz, R.B., O’Malley, B.W., 1999. Nuclear receptor coregulators: cellular and molecular biology. Endocrinol. Rev. 20, 321
/344. Mercier, L., Thompson, E.B., Simons, S.S., Jr, 1983. Dissociation of steroid binding to receptors and steroid induction of biological activity in a glucocorticoid responsive cell. Endocrinology 112, 601
/609. Mercier, L., Miller, P.A., Simons, S.S., Jr, 1986. Antiglucocorticoid steroids have increased agonist activity in those hepatoma cell lines that are more sensitive to glucocorticoids. J. Steroid Biochem. 25, 11
/20. Myer, V.E., Young, R.A., 1998. RNA polymerase II holoenzymes and subcomplexes. J. Biol. Chem. 273, 27757
/27760. Naar, A.M., Beaurang, P.A., Zhou, S., Abraham, S., Solomon, W., Tjian, R., 1999. Composite co-activator ARC mediates chromatindirected transcriptional activation. Nature 398, 828
/832. Neish, A.S., Anderson, S.F., Schlegel, B.P., Wei, W., Parvin, J.D., 1998. Factors associated with the mammalian RNA polymerase II holoenzyme. Nucleic Acids Res. 26, 847
/853. Ogryzko, V.V., Kotani, T., Zhang, X., Schlitz, R.L., Howard, T., Yang, X.J., Howard, B.H., Qin, J., Nakatani, Y., 1998. Histonelike TAFs within the PCAF histone acetylase complex. Cell 94, 35
/ 44. Onate, S.A., Tsai, S.Y., Tsai, M.-J., O’Malley, B.W., 1995. Sequence and characterization of a coactivator for the steroid hormone receptor superfamily. Science 270, 1354
/1357. Oshima, H., Simons, S.S., Jr, 1992a. Modulation of glucocorticoid induction of tyrosine aminotransferase gene expression by variations in cell density. Endocrinology 130, 2106
/2112. Oshima, H., Simons, S.S., Jr, 1992b. Modulation of transcription factor activity by a distant steroid modulatory element. Mol. Endocrinol. 6, 416
/428.

141

Oshima, H., Simons, S.S., Jr, 1993. Sequence-selective interactions of transcription factor elements with tandem glucocorticoid-responsive elements at physiological steroid concentrations. J. Biol. Chem. 268, 26858
/26865. Oshima, H., Szapary, D., Simons, S.S., Jr, 1995. The factor binding to the glucocorticoid modulatory element of the tyrosine aminotransferase gene is a novel and ubiquitous heteromeric complex. J. Biol. Chem. 270, 21893
/21910. Ossipow, V., Tassan, J.P., Nigg, E.A., Schibler, U., 1995. A mammalian RNA polymerase II holoenzyme containing all components required for promoter-specific transcription initiation. Cell 83, 137
/146. Pan, G., Aso, T., Greenblatt, J., 1997. Interaction of elongation factors TFIIS and elongin A with a human RNA polymerase II holoenzyme capable of promoter-specific initiation and responsive to transcriptional activators. J. Biol. Chem. 272, 24563
/24571. Plisov, S.Y., Poirot, M.E., Modarress, K.J., Cavanaugh, A.H., Edwards, D.P., Simons, S.S.J., 1998. Different populations of progesterone receptor
/steroid complexes in binding to specific DNA sequences: effects of salts on kinetics and specificity. J. Steroid Biochem. Mol. Biol. 67, 251
/266. Pons, M., Simons, S.S., Jr, 1981. Facile, high yield synthesis of spiro C17-steroidal oxetan-3?-ones. J. Org. Chem. 46, 3262
/3264. Rachez, C., Lemon, B.D., Suldan, Z., Bromleigh, V., Gamble, M., Naar, A.M., Erdjument-Bromage, H., Tempst, P., Freedman, L.P., 1999. Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex. Nature 398, 824
/828. Reichardt, H.M., Umland, T., Bauer, A., Kretz, O., Schutz, G., 2000. Mice with an increased glucocorticoid receptor gene dosage show enhanced resistance to stress and endotoxic shock. Mol. Cell. Biol. 20, 9009
/9017. Robyr, D., Wolffe, A.P., Wahli, W., 2000. Nuclear hormone receptor coregulators in action: diversity for shared tasks. Mol. Endocrinol. 14, 329
/347. Sarlis, N.J., Bayly, S.F., Szapary, D., Simons, S.S., Jr, 1999. Quantity of partial agonist activity for antiglucocorticoids complexed with mutant glucocorticoid receptors is constant in two different transactivation assays but not predictable from steroid structure. J. Steroid Biochem. Mol. Biol. 68, 89
/102. Sharma, M., Zarnegar, M., Li, X., Lim, B., Sun, Z., 2000. Androgen receptor interacts with a novel MYST protein, HBO1. J. Biol. Chem. 275, 35200
/35208. Simons, S.S., Jr, Pons, M., Johnson, D.F., 1980. a-Keto mesylate: a reactive thiol-specific functional group. J. Org. Chem. 45, 3084
/ 3088. Simons, S.S., Jr, Oshima, H., Szapary, D., 1992. Higher levels of control: modulation of steroid hormone-regulated gene transcription. Mol. Endocrinol. 6, 995
/1002. Song, L.-N., Huse, B., Rusconi, S., Simons, S.S., Jr, 2001. Transactivation specificity of glucocorticoid versus progesterone receptors: role of functionally different interactions of transcription factors with amino- and carboxyl-terminal receptor domains. J. Biol. Chem. 276, 24806
/24816. Szapary, D., Xu, M., Simons, S.S., Jr, 1996. Induction properties of a transiently transfected glucocorticoid-responsive gene vary with glucocorticoid receptor concentration. J. Biol. Chem. 271, 30576
/ 30582. Szapary, D., Huang, Y., Simons, S.S., Jr, 1999. Opposing effects of corepressor and coactivators in determining the dose
/response curve of agonists, and residual agonist activity of antagonists, for glucocorticoid receptor regulated gene expression. Mol. Endocrinol. 13, 2108
/2121. Tsai, M.-J., O’Malley, B.W., 1994. Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu. Rev. Biochem. 63, 451
/486. Tzukerman, M.T., Esty, A., Santiso-Mere, D., Danielian, P., Parker, M.G., Stein, R.B., Pike, J.W., McDonnell, D.P., 1994. Human

142

S. Chen, S.S. Simons, Jr / Molecular and Cellular Endocrinology 199 (2003) 129
/142

estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol. Endocrinol. 8, 21
/30. Vayssiere, B.M., Dupont, S., Choquart, A., Petit, F., Garcia, T., Marchandeau, C., Gronemeyer, H., Resche-Rigon, M., 1997. Synthetic glucocorticoids that dissociate transactivation and AP-1 transrepression exhibit antiinflammatory activity in vivo. Mol. Endocrinol. 11, 1245
/1255. Voegel, J.J., Heine, M.J.S., Zechel, C., Chambon, P., Gronemeyer, H., 1996. T1F2, a 160 kDa transcriptional mediatior for the liganddependent activation function AF-2 of nuclear receptors. EMBO J. 15, 3667
/3675. Voegel, J.J., Heine, M.J.S., Tini, M., Vivat, V., Chambon, P., Gronemeyer, H., 1998. The coactivator T1F2 contains three nuclear receptor-binding motifs and mediates transactivation through CBP binding-dependent and -independent pathways. EMBO J. 17, 507
/519.

Wang, G., Cantin, G.T., Stevens, J.L., Berk, A.J., 2001. Characterization of mediator complexes from HeLa cell nuclear extract. Mol. Cell. Biol. 21, 4604
/4613. Wasner, G., Oshima, H., Thompson, E.B., Simons, S.S., Jr, 1988. Unlinked regulation of the sensitivity of primary glucocorticoidinducible responses in mouse mammary tumor virus infected Fu5-5 rat hepatoma cells. Mol. Endocrinol. 2, 1009
/1017. Webster, L.C., Ricciardi, R.P., 1991. Trans-dominant mutants of E1A provide genetic evidence that the zinc finger of the trans-activating domain binds a transcription factor. Mol. Cell. Biol. 11, 4287
/ 4296. Zeng, H., Plisov, S.Y., Simons, S.S., Jr, 2000. Ability of the glucocorticoid modulatory element (GME) to modify glucocorticoid receptor transactivation indicates parallel pathways for the expression of GME and glucocorticoid response element activities. Mol. Cell. Endocrinol. 162, 221
/234.