A Nonsteroidal Glucocorticoid Receptor Antagonist

17 downloads 141 Views 253KB Size Report
ish peroxidase-donkey antirabbit Ig in 0.1% BSA/0.1%. Tween 20/PBS was ..... Marschke K, Kong J, Farmer L, Jones T 1998 Synthesis and biological activity of a ...
0888-8809/03/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 17(1):117–127 Copyright © 2003 by The Endocrine Society doi: 10.1210/me.2002-0010

A Nonsteroidal Glucocorticoid Receptor Antagonist JEFFREY N. MINER, CURTIS TYREE, JUNLIAN HU, ELAINE BERGER, KEITH MARSCHKE, MASAKI NAKANE, MICHAEL J. COGHLAN, DAVE CLEMM, BEN LANE, AND JON ROSEN Department of Molecular and Cell Biology (J.N.M., J.H., D.C., J.R.), New Leads Discovery (C.T., E.B., K.M.), Ligand Pharmaceuticals, Inc., San Diego, California 92121; and Abbott Laboratories (M.N., M.J.C., B.L.) D-4NB J35 Pharmaceutical Discovery, Abbott Park, Illinois 60064-3535 Selective intracellular receptor antagonists are used clinically to ameliorate hormone-dependent disease states. Patients with Cushing’s syndrome have high levels of the glucocorticoid, cortisol, and suffer significant consequences from this overexposure. High levels of this hormone are also implicated in exacerbating diabetes and the stress response. Selectively inhibiting this hormone may have clinical benefit in these disease states. To this end, we have identified the first selective, nonsteroidal glucocorticoid receptor (GR) antagonist. This compound is characterized by a tri-aryl methane core chemical structure. This GR-specific an-

tagonist binds with nanomolar affinity to the GR and has no detectable binding affinity for the highly related receptors for mineralocorticoids, androgens, estrogens, and progestins. We demonstrate that this antagonist inhibits glucocorticoid-mediated transcriptional regulation. This compound binds competitively with steroids, likely occupying a similar site within the ligand-binding domain. Once bound, however, the compound fails to induce critical conformational changes in the receptor necessary for agonist activity. (Molecular Endocrinology 17: 117–127, 2003)

T

with a beneficial therapeutic profile (22–26). Similarly, the progesterone receptor ligands have been developed with both antagonist and partial agonist activity (27, 28). However, functional nonsteroidal antagonists that interact in cells with the remaining steroid receptors, GR or MR, have not yet been identified. Several nonsteroidal compounds have been identified that appear to bind directly to the GR, including 5,5-diphenylhydantoin (29) and ␤-Lapachone (30), although it is not yet clear whether these compounds bind the receptor in cells. Selective antagonists of GRs could be useful in treating hypercortisolemia associated with Cushing’s syndrome and other conditions in which the endogenous GR is hyperactivated either through higher glucocorticoid levels or increased receptor sensitivity (5). Other possible uses include a reduction of the immunosuppression associated with ongoing HIV infection, depression (31), and other stress-associated phenomena (32–34). Finally, a selective GR antagonist may be useful in treating diabetic patients. Glucocorticoids appear to play a pathological role in high serum glucose levels in diabetic patients by up-regulating the rate-limiting enzyme in gluconeogenesis, phosphoenol pyruvate carboxykinase. Regardless of their potential for therapeutic use, selective antagonists will likely prove useful in furthering our understanding of the GR signaling pathway itself. During the course of screening compound libraries for GR modulators, we discovered an antagonist, designated “AL082D06” (D06) that bound specifically to GR with nanomolar affinity. This antagonist is unlike the other frequently used steroidal antagonists for GR, RU-38486 (RU-486) and ZK-98299 (ZK-299), in that it

HE INTRACELLULAR RECEPTOR family is a large group of transcription factors defined by homology within their DNA-binding domains. Modulation of intracellular receptor activity in vitro and in vivo using synthetic ligands has yielded remarkable insights into the molecular mechanisms of receptor function (1, 2) and enormous benefit in disease treatment (3–7). Specific chemical structural classes that bind with high affinity to intracellular receptors and either mimic (agonists) or suppress (antagonists) the activity of the endogenous ligand have been identified (8). Ligands used clinically target the receptors for estrogen (ER) (9), progesterone (PR) (10), glucocorticoid (GR) (11– 13), androgen (AR) (10), and mineralocorticoid (MR) (14). Typically, these ligands are steroidal, consisting of the classic tetracyclic structure with various substituents decorating this core. Nonsteroidal compounds that interact with members of the steroid receptor subclass have been identified. Nonsteroidal compounds have been derived that have primarily antagonist properties; these include the partial ER antagonists, tamoxifen and raloxifene-related compounds (15, 16), and the AR antagonists, casodex and flutamide (17–19), as well as several more recently identified AR antagonists (20, 21). Additionally, selective nonsteroidal modulators for the androgen receptor have been identified, which also act as partial agonists and may represent a new class of agonist Abbreviations: AR, Androgen receptor; Dex, dexamethasone; ER, estrogen receptor; GR, glucocorticoid receptor; GRE, glucocorticoid response element; MMTV, mouse mammary tumor virus; MMTV:Luc, MMTV promoter driving a luciferase reporter; MR, mineralocorticoid receptor; PR, progesterone receptor; TAT, tyrosine amino transferase. 117

118 Mol Endocrinol, January 2003, 17(1):117–127

has no measurable binding affinity for the progesterone receptor. As has been described previously, the three-dimensional structure of the ligand defines not only its affinity for the receptor, but also the conformation of that receptor once it has associated with ligand (35–37). This new compound appears to bind directly to receptor without inducing the same conformational changes associated with steroidal ligands. These ligands prevent the occurrence of some of the earliest steps in receptor activation. We report here the molecular and cellular characterization of this antagonist.

RESULTS We conducted a high-throughput screen of a defined compound library using a GR-based assay. This screen revealed a nonsteroidal compound that exhibited strong antagonist activity against GR. We undertook the molecular and cellular characterization of this ligand. Transient transfection of GR into CV1 cells in the presence of the synthetic glucocorticoid, dexamethasone (Dex), and a glucocorticoid-responsive mouse mammary tumor virus (MMTV) promoter driving a luciferase reporter (MMTV:Luc) plasmid results in an increase in luciferase activity (ⱕ2000-fold) (Fig. 1). This activity can be inhibited with known antagonists like RU-486 and ZK-299 (Fig. 1), which compete with Dex for the ligand-binding pocket of the receptor but fail, in most cases, to induce an active conformation (1, 35, 36, 38). Addition of D06 causes a dose-dependent decrease in transcriptional activation from the MMTV: Luc reporter stimulated with half-maximal DEX concentrations. D06 acts to antagonize reporter activity using several glucocorticoid-responsive promoterreporter systems including the 3-kb tyrosine amino transferase (TAT) promoter and less complex promoters comprised of isolated glucocorticoid response element (GRE) sequences (data not shown) (37).

Miner et al. • Nonsteroidal GR Antagonist

This compound, bis(4-N,N-dimethylaminophenyl)(2chloro-5-nitrophenyl) methane, is related to a well known series of dyes such as Malachite green, a fungicide used in aquariums (Fig. 2). This general class of molecule has been described previously as ligands for the estrogen receptor (39). We used a competitive binding assay to determine the affinity of D06 for GR as well as the other steroid receptors. D06 competes with 3H-Dex for baculovirusexpressed GR with nanomolar affinity. Other intracellular receptors (AR, ER, PR, and MR) have no affinity for D06 in a similarly structured binding assay with the appropriate receptor and tritiated ligand (⬎2500 nM). This selectivity for binding is in contrast to the significant PR cross-reactivity of other known GR antagonists, RU-486 and ZK-299 (Fig. 2). We tested the functional specificity of D06 by assessing its agonist and antagonist activity against a battery of related intracellular receptors (Fig. 3). These results confirm the in vitro binding data and clearly indicate that D06 has no activation efficacy on the progesterone, androgen, mineralocorticoid, retinoid, glucocorticoid, or estrogen receptors (Fig. 3A). Furthermore, these data indicate that D06 is very efficacious at antagonizing GR activity but exhibits much weaker efficacy when tested against the other steroid receptors in contrast to the reference antagonists used as controls (Fig. 3B). D06 was tested for antagonist effects in cell-based models of transcriptional activation. We demonstrate that D06 can antagonize steroid-mediated induction of glutamine synthetase RNA in MG63 cells (data not shown) and TAT enzyme in human skin fibroblasts (Fig. 4A). We also tested D06 for effects on genes normally repressed by glucocorticoids. We measured repression using an E-Selectin promoter:luciferase construct. This plasmid contains the E-Selectin gene promoter upstream from the luciferase reporter. TNF and IL-1 strongly induce expression from this plasmid, and glucocorticoids are effective repressors of this induction (Fig. 4B), In contrast, D06 is unable to re-

Fig. 1. D06 Antagonist Activity on GR This shows the results of a cotransfection experiment with a concentration response of D06 (0.01 nM to 10 ␮M in log steps). The MMTV:Luc reporter and hGR expression vector are cotransfected into CV-1 cells and compound is added. Curves for Dex and D06 are shown in agonist mode (F, Dex; f, D06). In addition, D06 and ZK-299 are tested in antagonist mode (Œ, ZK-299 and F, D06) in the presence of an EC50 of Dex (3 ⫻ 10⫺10 M). Symbols for the antagonists experiment are as follows (Œ, ZK299; and F, D06). The experimental data for the experiment shown are the average of three independent replicates. This is a representative experiment from more than three similar experiments with the same result.

Miner et al. • Nonsteroidal GR Antagonist

Mol Endocrinol, January 2003, 17(1):117–127

119

Fig. 2. Tri-aryl Methane, D06, Exhibits Selective GR Binding Activity The structure of D06 is shown [bis(4-N,N-dimethylaminophenyl)(2-chloro-5-nitrophenyl) methane] together with the binding data for D06 covering a number of related intracellular receptors (GR, PR, AR, MR, ER). Reference steroidal agonists (Dex, hydrocortisone, and prednisolone) and antagonists (RU-486 and ZK-299) are shown for comparison. The data are shown as Ki (nM). NT, Not tested. K, 1000.

press transcription when added alone at any concentration (black squares). However, D06 is able to fully reverse the repression mediated by Dex at the Eselectin promoter. D06 is also capable of partially inhibiting Dex-mediated repression of IL-6 and collagenase protein production from untransfected human skin fibroblast cells using endogenous receptors (data not shown). In summary, D06 can act to inhibit both transcriptional activation and repression by receptor in a variety of cell types on a variety of genes. Antagonists like RU-486 can cause receptor to bind to GRE sequences and exhibit agonist activity when tested in certain cells at certain promoters (37). In contrast, ZK-299-bound GR exhibits no detectable DNA binding activity in vitro and does not activate transcription under any circumstances although this is controversial (40). We tested D06 for its effect on DNA binding and determined that much like ZK-299, D06 did not induce DNA binding by GR in vitro (Fig. 5A) and will inhibit both Dex- and RU-486-induced GR DNA binding activity (Fig. 5A and data not shown). To confirm the in vitro DNA binding results in cells, we used an in vivo competition assay to monitor DNA binding activity of antagonist-bound GR (40). The assay utilizes coexpression of a constitutively active Cterminal deletion of GR with wild-type GR. The constitutively active receptor binds to and activates transcription from the MMTV reporter. The wild-type protein will compete with this activity if it is bound to an antagonist like RU-486 that allows DNA binding, but not transcriptional activation. The results of this type of assay with D06 reveal that unlike either ZK-299 or RU-486, D06-bound GR does not compete with the constitutively active GR (Fig. 5B). This result demon-

strates a clear difference between the known GR antagonists and this nonsteroidal compound. The structure and function of the receptor are in part determined by the structure of the bound ligand (41, 42). Agonists as well as antagonists such as RU-486 induce a particular structural conformation in the receptor that can be detected by limited protease digestion (35, 36, 43, 44). In our hands, compounds such as ZK-299 induce a different, more protease-sensitive structure that is similar to the unbound state (37, 45). As shown in Fig. 6A, D06 produces a pattern that, like ZK-299, is highly sensitive to protease digestion. We are convinced that D06 is bound to the receptor under these conditions because much like ZK299, D06 inhibits the formation of a Dex-protected band at the same concentrations (Fig. 6B). Early in the GR signal transduction pathway, the receptor translocates to the nucleus (46). We compared the nuclear translocation activity of RU-486 and ZK-299 to D06 using an immunofluorescence assay in transfected COS cells (40). In agreement with previously published data for progesterone receptor (40) and for GR (47), Fig. 7 shows that both RU-486 and ZK-299 induced significant nuclear translocation in this immunofluoresence assay. In contrast, D06 exhibited only weak nuclear translocation, even when added at 10 ␮M.

DISCUSSION A nonsteroidal GR ligand, D06, has been identified. We have tested this compound in a variety of assays

120

Mol Endocrinol, January 2003, 17(1):117–127

Miner et al. • Nonsteroidal GR Antagonist

Fig. 3. D06 Activity Is GR Selective A, Agonist assay: This graph represents the data from a 96-well cotransfection experiment into CV1 cells with the specified transfected receptors and the MMTV:Luc (or MMTV:3⫻ERE:Luc for ER) promoter in the presence of D06 or appropriate reference steroidal agonists [GR-Dex (D, 100 nM); PR-progesterone (P, 100 nM], MR-aldosterone [100 nM (A); ER-estrogen (E, 100 nM); AR-DHT (Dht,100 nM)]. The results are shown as a percent agonist activity. This experiment has been repeated more than three times with similar results. The data shown are an average of three replicates within a single experiment. B, Antagonist: Identical to panel A except that the assays are done in the presence of an EC50 of each reference steroidal agonist [GR-Dex (3 ⫻ 10⫺10 ⫺9 ⫺9 ⫺9 ⫺10 M); PR-progesterone (1 ⫻ 10 ); MR-aldosterone (1 ⫻ 10 ); ER-estrogen (1 ⫻ 10 ); AR-DHT (3 ⫻ 10 M)]. Reference antagonists are shown to validate the assays [GR-RU486 (1 ␮M), PR-RU-486 (1 ␮M), MR-RU28318 (RU318) (1 ␮M), ER-Tamoxifen (Tam.) (1 ␮M), AR-flutamide (Flut.) (1 ␮M)]. D06 and the reference antagonists are used at two different concentrations designated by 7 for D06 (1 ␮M and 10 ␮M). The results are expressed such that maximal antagonist activity by the reference antagonist is set to 100%. These experiments were repeated three times with similar results.

to assess its agonist and antagonist activities. This compound is characterized by a tri-aryl methane core chemical structure and has strong (⬍250 nM) affinity for the GR and binds competitively with other known GR ligands. The most likely explanation for our competitive binding profiles is that D06 occupies the same hydrophobic pocket in the ligand-binding domain as steroids, although it is conceivable that binding could occur elsewhere on the receptor and this in turn alters the steroidal ligands’ affinity. Once bound, the compound is a very efficacious antagonist of receptor activity. It is capable of antagonizing both glucocorticoid-mediated transcriptional activation and repression in multiple cellular contexts. D06 is an antagonist in these assays with a potency of 200 nM. This compound does not have any effect in these assays in

the absence of steroid, classifying it as a pure antagonist. One of the problems with current attempts at glucocorticoid antagonist therapy is cross-reactivity with other steroid receptors. The currently available glucocorticoid antagonists mifepristone (RU-486) and onapristone (ZK-299) are also potent progesterone antagonists, making their clinical use problematic. We have demonstrated that D06 is selective for GR, with no measurable affinity for, and little or no activity on, the progesterone, mineralocorticoid, androgen, estrogen, and retinoid receptors (Fig. 2 and data not shown). Thus, this antagonist appears to be useful for selectively inhibiting GR in different contexts. The mechanism of antagonist ligands has been extensively reviewed (11, 48–50). Current literature sug-

Miner et al. • Nonsteroidal GR Antagonist

Mol Endocrinol, January 2003, 17(1):117–127

121

Fig. 4. Antagonism of GR-Mediated Activation and Repression A, TAT antagonist activity of D06. TAT activity was measured as described (41) in human skin fibroblasts in the presence of vehicle (F), an EC50 of Dex (F) (10 nM), or hydrocortisone (F) (HC 100 nM). Antagonists ZK299, RU-486, or D06 were added in increasing concentration. These experiments were repeated three times each with similar results and a representative experiment is shown. B, D06 inhibits DEX-mediated repression of E-selectin promoter activity. Luciferase activity is shown upon induction with TNF and IL-1␤ (open vertical bar). D06 in dose response in the absence of DEX is shown in the black squares. This demonstrates no effect of D06 on TNF-IL-1-induced luciferase activity. In the presence of half-maximal concentrations of DEX the luciferase activity is suppressed. Addition of D06 relieves repression by DEX (open squares). This is a representative experiment showing the average of three replicates within the experiment. This experiment has been repeated in this format at least three times.

122 Mol Endocrinol, January 2003, 17(1):117–127

Fig. 5. D06 Is an Antagonist to GR DNA Binding Activity A, In vitro DNA binding activity was measured using an EMSA. PA is radiolabeled GRE probe alone. In vitro translated GR was incubated with either solvent (⫺), Dex (10 nM), or D06 (1 and 10 ␮M). This assay produces strong hormonedependent DNA binding by GR in response to glucocorticoids, but not by D06. The antagonist experiment is similar except that GR was incubated with 10 nM Dex and increasing concentrations of D06 (0.01, 0.1, 1, and 10 ␮M) or ZK-299 (0.01 and 1 ␮M) as a positive antagonist control. B, Cellular competition assay of D06. This assay uses cotransfection of the MMTV:Luc reporter, together with a constitutively active truncation of GR (I550) and the wild-type receptor (RSVhGR). The first lane shows the basal level of transcription from the MMTV promoter without transfected constitutive activator. The second lane shows the signal generated in response to the constitutive activator in the presence of the solvent ethanol. Known antagonists such as RU-486 (10 ␮M) and ZK-299 (10 ␮M) will inhibit the constitutive activator by binding to wild-type GR and inducing competitive DNA binding significantly (P ⬍ 0.05). In contrast, shown in the last lane, D06 (10 ␮M) exhibits only weak, inconsistent inhibition that is not significantly different from vehicle plus constitutive GR (P ⬎ 0.05) (Fisher’s least significant differences). This experiment has been conducted four times with the same conclusion, although some differences between absolute levels of luciferase produced by the constitutive GR were found.

Miner et al. • Nonsteroidal GR Antagonist

gests that they induce a conformation in the receptor that differs from that induced by agonists. These changes affect the interaction of the receptor with critical components of the cell, be it other proteins or DNA. Many antagonists do, in fact, exhibit some agonist activity when bound to the receptor. Tamoxifen, an antagonist in breast tissue, has partial agonist activity in the uterus that may increase the risk of significant proliferative side effects. We established the point in the glucocorticoid signal transduction pathway at which D06 inhibits GR function. The known GR/PR antagonist RU-486 induces DNA binding activity of the receptor. In contrast, in simple gel shift assays, ZK-299 and D06 antagonize DNA binding by GR (Fig. 5A). We also examined the DNA binding activity in a cell-based assay used to measure GR DNA binding activity; (37). We have previously shown that under certain conditions, even ZK299-receptor complexes can occupy GREs in vivo (37). Using similar in vivo competition assays, D06 fails to induce the formation of receptor-GRE complexes, whereas ZK-299 and RU-486 are fully capable of inhibiting activation from a constitutively activated reporter (Fig. 5B). We interpret the results from the cellular competition assay with the view that this assay measures competition at the level of DNA, although it is possible that the competition we are seeing is competition for some type of limiting coactivator required for the constitutive activator to function. At present, we cannot distinguish between these models. This question is especially relevant for ZK-299 because it fails to induce DNA binding in vitro, but competes nicely in vivo. The interpretation of the D06 results is more straightforward because it fails to induce DNA binding and is extremely weak in the in vivo competition assay. In contrast to other known antagonists, D06 has little or no agonist activity in any assay we have devised; therefore, we consider it a pure, type 1 GR antagonist as defined by McDonnell et al. (51, 52). To roughly assess the conformation of the receptor when bound to D06, we used a protease digestion assay (Fig. 6). This assay demonstrated that D06 and ZK-299 generate a conformation different from both partial antagonists (RU-486) and full agonists (Dex) (37, 45, 53). Furthermore, this conformation is more sensitive to proteases than agonist-like conformations. Thus the receptor is in an antagonist conformation when bound to D06. After binding ligand and changing conformation, the receptor undergoes nuclear translocation. We tested the impact of D06 on this process. Confirming the results of others (54, 55), we show that the known steroidal agonists and antagonists all induce significant nuclear translocation under our assay conditions. This result is controversial because the nuclear translocation activity of RU-486 may be cell type specific (56, 57). In our assay, in contrast to RU-486 and the other steroids, D06 exhibited reduced nuclear translocation (Fig. 7). These results indicate that at least part of the antagonist activity of this class occurs by reducing the amount of GR transported to the nucleus (Fig. 7). Furthermore,

Miner et al. • Nonsteroidal GR Antagonist

Mol Endocrinol, January 2003, 17(1):117–127

123

Fig. 6. Protease Digestion Assay of D06-Bound GR A, The conformation induced by D06 was assayed by incubation of radiolabeled GR together with increasing concentrations of trypsin in the presence or absence of solvent, the reference agonist (Dex 1 ␮M), reference antagonists ZK299 (10 ␮M), RU-486 (10 ␮M), and with D06 (10 ␮M). Protected bands correspond to segments of the ligand-binding domain that are more resistant to protease in the presence of hormone. B, Protease digestion assay in antagonist mode. This assay uses 10 nM Dex to produce a small amount of the protected species denoted by an arrow. Addition of either antagonist D06 or ZK299 effectively competes with Dex for the receptor, which creates a less protease-resistant structure resulting in the disappearance of the band.

the portion of GR bound to the D06 that does reach the nucleus is unable to bind to DNA (Fig. 5, A and B). In summary, we have characterized a compound, D06, that is a full antagonist for GR, but not for other steroid receptors. D06 partially blocks GR translocation to the nucleus, and completely blocks DNA binding by the receptor. This molecule may be a useful template for the development of clinically useful oral antagonists of the GR. The search for selective antagonists of GR activity has been driven by both clinical and research needs. The selective inhibition of GR action may allow the differentiation of its activities from those of MR and PR when studying the physiology and cellular biology of glucocorticoid action. In addition, compounds of this type may provide better treatment for patients with a variety of cortisol-related endocrine disorders.

MATERIALS AND METHODS Cellular Assays Transfection. CV-1 cells (African green monkey kidney fibroblasts, American Type Culture Collection, Manassas, VA)

were grown in DMEM (BioWhittaker, Inc., Walkersville, MD) containing 10% (vol/vol) fetal calf serum (HyClone Laboratories, Inc., Logan, UT), 2 mM L-glutamine, and 55 ␮g/ml gentamycin. Cells were transiently transfected using the calcium phosphate coprecipitation method (58). Unless otherwise noted, 5 ␮g/ml of a human GR-expression plasmid vector (RSV:hGR), 5 ␮g/ml MMTV:Luc reporter plasmid, 2.5 ␮g/ml of pRSV-␤-Gal (␤-galactosidase) as a control for transfection efficiency, and 7.5 ␮g of filler DNA (pGEM4) at a final concentration of 20 ␮g/ml were precipitated and then added to the cells. The medium was changed 16 h later to contain 5% charcoal-stripped fetal calf serum and steroid ligands with or without test compounds (10 ␮M) for 24 h. Cells were then lysed and assayed as described previously (43, 59). The E-selectin transfection assay is similar except that 5 ␮g/ml E-sel/luc reporter plasmid was added instead of MMTV:Luc. The medium was changed 16 h after transfection to contain 10% charcoal-stripped fetal calf serum, TNF␣ (10 ng/ml), IL-1␤ (1 ng/ml), and test compounds (10 nM to 10 ␮M) with or without 0.32 nM Dex for 24 h. Cells were then lysed and assayed as described above. TAT Assay. TAT activity in H4IIE cells was measured as described previously (60). Preconfluent H4IIE cells in 96-well plates were incubated for 24 h with compound, washed with PBS, and lysed. Extracts were subjected to enzymatic assay as described (60). IL-6 was measured in confluent human skin fibroblasts in induction media (1.75% BSA/antibiotics/DMEM) after incubation with induction media for 4–6 h. Media were changed

124 Mol Endocrinol, January 2003, 17(1):117–127

Miner et al. • Nonsteroidal GR Antagonist

Fig. 7. Nuclear Translocation Induced by Hormone Cos cells transfected with a Glu-tagged GR expression vector (RSVhGRnxg) were monitored for receptor localization by immunofluoresence using anti-Glu antibodies in response to various treatments. Cells were assessed for receptor localization by a blinded analysis; a minimum of 60 cells were analyzed per experiment. Comparisons were made between a nuclear DNA stain and the anti-GR immunofluoresence to ensure proper localization of the nuclei and cytoplasm. Cells were scored 0–3 (0 ⫽ cytoplasmic entirely; 1⫽ C ⬎ N; 2 ⫽ C ⬍ N; 3 ⫽ nuclear entirely). The scores of all the cells counted for a given treatment were averaged, and the SEM was calculated. The results were graphically expressed as a percentage of nuclear translocation where 100% is complete translocation (a perfect score of 3). RU-486 and ZK-299 are used as antagonist controls. All compounds were used at saturating concentrations (D06, 10 ␮M; Dex, 1 ␮M; RU-486, 1 ␮M; ZK299 10 ␮M).

and cells incubated a further 1 h in induction media; compound. IL-1␤ was then added to a final concentration of 1 ng/ml in induction media (Roche Molecular Biochemicals, Indianapolis, IN), and cells were cultured for 24 h. Media were removed and added to Maxisorp Plate (Nunc) with capture antibody (IL-6-monoclonal mouse antihuman IL-6)-coated wells (M-620, Endogen, Inc., Boston, MA) and incubated at room temperature (RT) overnight. Plate was washed twice in PBS, blocked with 4% BSA/PBS, and incubated 1 h at RT. Secondary antibody-biotinylated monoclonal antihuman IL-6 (M-621-B, Endogen, Inc.), 500 ␮g/ml in 4% BSA/PBS, was added and incubated for 2 h at RT, and washed three times in PBS. A 1:5000 diluted ExtrAvidin-horseradish peroxidase solution (E-2886, Sigma, St. Louis, MO) in 4% BSA/PBS was added and incubated for 30 min at RT. Plates were washed three times in PBS, and substrate solution (One hundred microliters of 3,3⬘,5,5⬘ tetramethyl benzidine-hydrogen peroxide (Sigma) was added and incubated 15 min at RT. Reaction was stopped with 50 ␮l per well of 2 N H2SO4 and OD was read at 450 nm/540 nm. Collagenase was measured in confluent human skin fibroblasts induced in 1.75% BSADMEM with compounds for 1 h. IL-1␤ was added (Roche Molecular Biochemicals) in induction medium (final 1 ng/ml) and the cells were cultured for 24 h. Collagenase Assay. Culture supernatants were added to 0.1% BSA/PBS and incubated for 2 h at RT; after washing, polyclonal rabbit antihuman MMP-1 in assay buffer was added and incubated for 2 h at RT. After washing, horseradish peroxidase-donkey antirabbit Ig in 0.1% BSA/0.1% Tween 20/PBS was added and incubated for 1 h at RT. One hundred microliters of 3,3⬘,5,5⬘ tetramethyl benzidine-hydrogen peroxide were added and incubated for approximately 5–30 min at RT, after which 100 ␮l per well of stop solution (1 N H2SO4) were added and the OD read at 450/540 nm. Plasmids The constitutive activator I550, pRSV:hGRnx, was obtained from Ron Evans (Salk Institute, La Jolla, CA) (61). T7hGRnxg was constructed by first inserting the Glu-Glu tag epitope sequence (62, 63) into pRSVhGRnx at the KpnI and SalI sites; hGRnxg was then inserted into the pT7-link expression vector at the NcoI and BamHI sites (61, 64). MMTV:Luc was obtained from Ron Evans (Salk Institute, La Jolla, CA). MMTV: ERE3x:Luc contains three tandem estrogen response ele-

ments cloned into a version of the MMTV promoter in which the GREs have been deleted (59). A reporter construct containing 600 bp of the E-selectin promoter region fused to the luciferase gene (E-sel/luc) was used in the E-selectin repression assays. Protease Digestion Assay The protease digestion was performed essentially as described by Allan and colleagues (35, 36, 43) with minor modifications. The plasmid pGR107 containing the GR-wt cDNA was used to produce 35S-radiolabeled GR using the TNT system (Promega Corp., Madison, WI). After the translation reaction, an aliquot (25 ␮l) of the lysate was incubated for 20 min at room temperature in the presence or absence of test compound at a final concentration of 1 ␮M. Aliquots (5 ␮l) of ligand-treated receptor mixture were subsequently incubated for 10 min with 0.6 ␮l of a trypsin solution (Worthington Biochemical Corp., Freehold, NJ) yielding a final enzyme concentration of 5, 10, 25, and 50 ␮g/ml. After termination and electrophoresis, the gels were fixed with a 30% (vol/vol) methanol, 10% (vol/vol) acetic acid solution for 30 min, and then immersed in Amplify (Amersham Pharmacia Biotech, Arlington Heights, IL) for 30 min. EMSAs Human GR was prepared by in vitro translation using the T7 expression vector pT7hGRnxg in a TNT T7-coupled reticulocyte lysate system (Promega Corp.). Test compounds were added at the beginning of the translation reaction at the indicated concentration. The specific probe is based on a palindromic GRE and was formed by annealing oligonucleotides with the sequences 5⬘-TCGACAGAACATCATGTTCTGAGCTAC-3⬘ and 5⬘-TCGAGTAGCTCAGAACATGATGTTCTG-3⬘. The annealed oligonucleotide was labeled by filling in the overhanging ends with the Klenow fragment of DNA polymerase in the presence of [␣32P]dATP and dGTP. Binding reactions were performed as described (65). Reactions were incubated on ice for 5–10 min and then resolved on 4% polyacrylamide gels containing 0.25⫻ TBE [1⫻ TBE is 89 mM Tris borate, 1 mM EDTA (pH 8.0)] at 4 C and 20 V/cm, which were then dried and autoradiographed.

Miner et al. • Nonsteroidal GR Antagonist

Competitive Binding Assay Growth and purification of recombinant hGR baculovirus followed the protocol outlined by Summers and Smith (66). The extract and binding assay buffer consisted of 25 mM sodium phosphate, 10 mM potassium fluoride, 10 mM sodium molybdate, 10% glycerol, 1.5 mM EDTA, 2 mM dithiothreitol, 2 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), and 1 mM phenylmethylsulfonyl fluoride (pH 7.4), at room temperature. Intracellular receptors produced in this fashion exhibit reproducible interaction with known ligands at the published affinity. These preparations were subjected to extensive quality control experiments before the assays, covering receptor response, specificity, size, and reference ligand affinity. Receptor assays were performed with a final volume of 250 ␮l containing from 50–75 ␮g of extract protein, plus 1–2 nM [3H]Dex at 84 Ci/mmol and varying concentrations of competing ligand (0 to 10⫺5 M). Assays were set up using a 96-well minitube system, and incubations were carried out at 4 C for 18 h. Equilibrium under these conditions of buffer and temperature was achieved by 6–8 h. Nonspecific binding was defined as that binding remaining in the presence of 1000 nM unlabeled Dex. At the end of the incubation period, 200 ␮l of 6.25% hydroxyapatite were added in wash buffer (binding buffer in the absence of dithiothreitol and phenylmethylsulfonyl fluoride). Specific ligand binding to receptor was determined by a hydroxyapatite-binding assay according to the protocol of Wecksler and Norman (67). Hydroxyapatite absorbs the receptorligand complex, allowing for the separation of bound from free radiolabeled ligand. The mixture was vortexed and incubated for 10 min at 4 C and centrifuged, and the supernatant was removed. The hydroxyapatite pellet was washed two times in wash buffer. The amount of receptor-ligand complex was determined by liquid scintillation counting of the hydroxyapatite pellet after the addition of 0.5 mM EcoScint A scintillation cocktail from National Diagnostics (Atlanta, GA). After correcting for nonspecific binding, IC50 values were determined. The IC50 value is defined as the concentration of competing ligand required to reduce specific binding by 50%; the IC50 values were determined graphically from a log-logit plot of the data. Kd values for the analogs were calculated by application of the Cheng-Prussof equation (68, 69). Steroid standards are included in each assay, and resulting Kd values are determined by use of a modified ChengPrussoff equation (49, 50). MR, AR, PR, and ER␣ expression in the baculovirus system and binding assays was conducted similarly except that labeled ligands were aldosterone [1–2 nM 3H-aldosterone from Amersham Pharmacia Biotech (TRK 434), specific activity 60 Ci/mmol], DHT (1–2 nM 3H-DHT at 130 Ci/mmol), progesterone [2–3 nM 3H -progesterone (Sigma, 93 Ci/mmol], and estradiol [2–3 nM 3H-estradiol (NEN Life Science Products), 114 Ci/mmol], respectively. Each binding assay point is done in duplicate, and each full experiment is repeated three or more times. Nuclear Translocation Cos cells were transfected as above, and the immunofluoresence assay was done as described (70). Glu tag antibody was used to detect GLU-tagged receptor (BABCO, Berkeley, CA). The results were quantified by a blinded analysis of views of at least 60 cells defined as follows: N, entirely nuclear (3 points); C, entirely cytoplasmic (0 points); N ⬎ C, 2 points; and C ⬎ N, 1 point. Nuclei were localized by Hoechst stain 33342, which binds to DNA.

Acknowledgments We thank Emily Guido for her expert technical assistance. We appreciate the plasmids and helpful discussions provided

Mol Endocrinol, January 2003, 17(1):117–127

125

by Marc Elgort, Ron Evans, Keith Yamamoto, and David Pearce.

Received January 8, 2002. Accepted September 20, 2002. Address all correspondence and requests for reprints to: Jeffrey N. Miner, Ligand Pharmaceuticals, Molecular and Cellular Biology, 10275 Science Center Drive, San Diego, California 92121. E-mail: [email protected].

REFERENCES 1. Gass EK, Leonhardt SA, Nordeen SK, Edwards DP 1998 The antagonists RU486 and ZK98299 stimulate progesterone receptor binding to deoxyribonucleic acid in vitro and in vivo, but have distinct effects on receptor conformation. Endocrinology 139:1905–1919 2. Pham TA, Elliston JF, Nawaz Z, McDonnell DP, Tsai MJ, O’Malley BW 1991 Antiestrogen can establish nonproductive receptor complexes and alter chromatin structure at target enhancers. Proc Natl Acad Sci USA 88: 3125–3129 3. Nawaz Z, Tsai MJ, McDonnell DP, O’Malley BW 1992 Identification of novel steroid-response elements. Gene Expr 2:39–47 4. Bertagna X, Bertagna C, Laudat MH, Husson JM, Girard F, Luton JP 1986 Pituitary-adrenal response to the antiglucocorticoid action of RU 486 in Cushing’s syndrome. J Clin Endocrinol Metab 63:639–643 5. Nieman LK, Chrousos GP, Kellner C, Spitz IM, Nisula BC, Cutler GB, Merriam GR, Bardin CW, Loriaux DL 1985 Successful treatment of Cushing’s syndrome with the glucocorticoid antagonist RU 486. J Clin Endocrinol Metab 61:536–540 6. Labrie F, Simard J, de Launoit Y, Poulin R, Theriault C, Dumont M, Dauvois S, Martel C, Li SM 1992 Androgens and breast cancer. Cancer Detect Prev 16:31–38 7. Koide SS 1998 Mifepristone. Auxiliary therapeutic use in cancer and related disorders. J Reprod Med 43:551–560 8. Anstead GM, Carlson KE, Katzenellenbogen JA 1997 The estradiol pharmacophore: ligand structure-estrogen receptor binding affinity relationships and a model for the receptor binding site. Steroids 62:268–303 9. Gale KE, Andersen JW, Tormey DC, Mansour EG, Davis TE, Horton J, Wolter JM, Smith TJ, Cummings FJ 1994 Hormonal treatment for metastatic breast cancer. An Eastern Cooperative Oncology Group Phase III trial comparing aminoglutethimide to tamoxifen. Cancer 73: 354–361 10. Adams JB 1998 Adrenal androgens and human breast cancer: a new appraisal. Breast Cancer Res Treat 51: 183–188 11. Baulieu EE 1997 RU 486 (mifepristone). A short overview of its mechanisms of action and clinical uses at the end of 1996. Ann NY Acad Sci 828:47–58 12. Clayton LH, Dilley KB 1998 Cushing’s syndrome. Am J Nurs 98:40–41 13. Diez JJ 1998 [Pharmacologic treatment of Cushing’s syndrome]. Med Clin (Barc) 110:781–796 14. Funder JW 1998 Aldosterone action: fact, failure and the future. Clin Exp Pharmacol Physiol Suppl 25:S47–S50 15. Friedman ZY 1998 Recent advances in understanding the molecular mechanisms of tamoxifen action. Cancer Invest 16:391–396 16. Osborne CK 1998 Tamoxifen in the treatment of breast cancer. N Engl J Med 339:1609–1618 17. Studer UE, Mills RD 1998 Arguments against the long-term use of combined androgen blockade. Eur Urol 34:29–32 18. Schroder FH 1998 Antiandrogens as monotherapy for prostate cancer. Eur Urol 34:12–17

126 Mol Endocrinol, January 2003, 17(1):117–127

19. Wajsman Z 1998 Arguments for the long-term use of combined androgen blockade. Eur Urol 34:25–28 20. Kong JW, Hamann LG, Ruppar DA, Edwards JP, Marschke KB, Jones TK 2000 Effects of isosteric pyridone replacements in androgen receptor antagonists based on 1,2-dihydro- and 1,2,3,4-tetrahydro-2,2-dimethyl-6trifluoromethyl-8-pyridono[5,6-g]quin olines. Bioorg Med Chem Lett 10:411–414 21. Zhi L, Tegley CM, Marschke KB, Jones TK 1999 Switching androgen receptor antagonists to agonists by modifying C-ring substituents on piperidino[3,2-g]quinolinone. Bioorg Med Chem Lett 9:1009–1012 22. Higuchi RI, Edwards JP, Caferro TR, Ringenberg JD, Kong JW, Hamann LG, Arienti KL, Marschke KB, Davis RL, Farmer LJ, Jones TK 1999 4-Alkyl- and 3,4-dialkyl1,2,3,4-tetrahydro-8-pyridono[5,6-g]quinolines: potent, nonsteroidal androgen receptor agonists. Bioorg Med Chem Lett 9:1335–1340 23. Edwards JP, Higuchi RI, Winn DT, Pooley CL, Caferro TR, Hamann LG, Zhi L, Marschke KB, Goldman ME, Jones TK 1999 Nonsteroidal androgen receptor agonists based on 4-(trifluoromethyl)-2H-pyrano[3,2-g]quinolin-2one. Bioorg Med Chem Lett 9:1003–1008 24. Hamann LG, Mani NS, Davis RL, Wang XN, Marschke KB, Jones TK 1999 Discovery of a potent, orally active, nonsteroidal androgen receptor agonist: 4-ethyl-1,2,3,4tetrahydro-6-(trifluoromethyl)-8-pyridono[5,6-g]-quinoline (LG121071). J Med Chem 42:210–212 25. Edwards JP, West SJ, Pooley CL, Marschke KB, Farmer LJ, Jones TK 1998 New nonsteroidal androgen receptor modulators based on 4-(trifluoromethyl)-2(1H)-pyrrolidino[3,2-g] quinolinone. Bioorg Med Chem Lett 8:745–750 26. Hamann LG, Higuchi RI, Zhi L, Edwards J, Xiao-ning W, Marschke K, Kong J, Farmer L, Jones T 1998 Synthesis and biological activity of a novel series of nonsteroidal, peripherally selective androgen receptor antagonists derived from 1,2-dihydropyridono[5,6-g]quinolines. J Med Chem 41:623–639 27. Zhi L, Tegley CM, Pio B, West SJ, Marschke KB, Mais DE, Jones TK 2000 Nonsteroidal progesterone receptor antagonists based on 6-thiophenehydroquinolines. Bioorg Med Chem Lett 10:415–418 28. Zhi L, Tegley CM, Marschke KB, Mais DE, Jones TK 1999 5-Aryl-1,2,3,4-tetrahydrochromeno[3,4-f]quinolin-3-ones as a novel class of nonsteroidal progesterone receptor agonists: effect of A-ring modification. J Med Chem 42: 1466–1472 29. Katsumata M, Gupta C, Baker MK, Sussdorf CE, Goldman AS 1982 Diphenylhydantoin: an alternative ligand of a glucocorticoid receptor affecting prostaglandin generation in A/J mice. Science 218:1313–1315 30. Schmidt TJ, Miller-Diener A, Litwack G 1984 ␤-Lapachone, a specific competitive inhibitor of ligand binding to the glucocorticoid receptor. J Biol Chem 259: 9536–9543 31. Schulte HM, Chrousos GP, Gold PW, Booth JD, Oldfield EH, Cutler Jr JB, Loriaux DL 1985 Continuous administration of synthetic ovine corticotropin-releasing factor in man. Physiological and pathophysiological implications. J Clin Invest 75:1781–1785 32. Chrousos GP 1998 Stress as a medical and scientific idea and its implications. Adv Pharmacol 42:552–556 33. Chrousos GP, Gold PW 1998 A healthy body in a healthy mind–and vice versa–the damaging power of “uncontrollable” stress [editorial; comment]. J Clin Endocrinol Metab 83:1842–1845 34. Burgess E, Dorn LD, Haaga DA, Chrousos G 1996 Sociotropy, autonomy, stress, and depression in Cushing syndrome. J Nerv Ment Dis 184:362–367 35. Allan GF, Tsai SY, Tsai MJ, O’Malley BW 1992 Liganddependent conformational changes in the progesterone

Miner et al. • Nonsteroidal GR Antagonist

36.

37.

38. 39.

40.

41.

42.

43.

44.

45.

46.

47.

48. 49. 50.

51.

52.

53.

receptor are necessary for events that follow DNA binding. Proc Natl Acad Sci USA 89:11750–11754 Allan GF, Leng X, Tsai SY, Weigel NL, Edwards DP, Tsai MJ, O’Malley BW 1992 Hormone and antihormone induce distinct conformational changes which are central to steroid receptor activation. J Biol Chem 267: 19513–19520 Guido EC, Delorme EO, Clemm DL, Stein RB, Rosen J, Miner JN 1996 Determinants of promoter-specific activity by glucocorticoid receptor. Mol Endocrinol 10: 1178–1190 Mahajan DK, London SN 1997 Mifepristone (RU486): a review. Fertil Steril 68:967–976 Bindal RD, Carlson KE, Katzenellenbogen BS, Katzenellenbogen JA 1988 Lipophilic impurities, not phenolsulfonphthalein, account for the estrogenic activity in commercial preparations of phenol red. J Steroid Biochem 31:287–293 Delabre K, Guiochon-Mantel A, Milgrom E 1993 In vivo evidence against the existence of antiprogestins disrupting receptor binding to DNA. Proc Natl Acad Sci USA 90:4421–4425 Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ, Agard DA, Greene GL 1998 The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95:927–937 Tanenbaum DM, Wang Y, Williams SP, Sigler PB 1998 Crystallographic comparison of the estrogen and progesterone receptor’s ligand binding domains. Proc Natl Acad Sci USA 95:5998–6003 Vegeto E, Allan GF, Schrader WT, Tsai MJ, McDonnell DP, O’Malley BW 1992 The mechanism of RU486 antagonism is dependent on the conformation of the carboxyterminal tail of the human progesterone receptor. Cell 69:703–713 Beekman JM, Allan GF, Tsai SY, Tsai MJ, O’Malley BW 1993 Transcriptional activation by the estrogen receptor requires a conformational change in the ligand binding domain. Mol Endocrinol 7:1266–1274 Modarress KJ, Opoku J, Xu M, Sarlis NJ, Simons Jr SS 1997 Steroid-induced conformational changes at ends of the hormone-binding domain in the rat glucocorticoid receptor are independent of agonist vs. antagonist activity. J Biol Chem 272:23986–23994 DeFranco D, Morris Jr SM, Leonard CM, GluecksohnWaelsch S 1988 Metallothionein mRNA expression in mice homozygous for chromosomal deletions around the albino locus. Proc Natl Acad Sci USA 85:1161–1164 Prima V, Depoix C, Masselot B, Formstecher P, Lefebvre P 2000 Alteration of the glucocorticoid receptor subcellular localization by nonsteroidal compounds. J Steroid Biochem Mol Biol 72:1–12 Lazar G, Lazar Jr G, Husztik E, Duda E, Agarwal MK 1995 The influence of antiglucocorticoids on stress and shock. Ann NY Acad Sci 761:276–295 Thurlimann B 1998 Hormonal treatment of breast cancer: new developments. Oncology 55:501–507 Nordeen SK, Bona BJ, Beck CA, Edwards DP, Borror KC, DeFranco DB 1995 The two faces of a steroid antagonist: when an antagonist isn’t. Steroids. 60: 97–104 McDonnell DP, Clevenger B, Dana S, Santiso-Mere D, Tzukerman MT, Gleeson MA 1993 The mechanism of action of steroid hormones: a new twist to an old tale. J Clin Pharmacol 33:1165–1172 McDonnell DP, Clemm DL, Hermann T, Goldman ME, Pike JW 1995 Analysis of estrogen receptor function in vitro reveals three distinct classes of antiestrogens. Mol Endocrinol 9:659–669 Reichman ME, Foster CM, Eisen LP, Eisen HJ, Torain BF, Simons Jr SS 1984 Limited proteolysis of covalently labeled glucocorticoid receptors as a probe of receptor structure. Biochemistry 23:5376–5384

Miner et al. • Nonsteroidal GR Antagonist

54. Madan AP, DeFranco DB 1993 Bidirectional transport of glucocorticoid receptors across the nuclear envelope. Proc Natl Acad Sci USA 90:3588–3592 55. Defranco DB, Madan AP, Tang Y, Chandran UR, Xiao N, Yang J 1995 Nucleocytoplasmic shuttling of steroid receptors. Vitam Horm 51:315–338 56. Lindemeyer RG, Robertson NM, Litwack G 1990 Glucocorticoid receptor monoclonal antibodies define the biological action of RU 38486 in intact B16 melanoma cells. Cancer Res 50:7985–7991 57. Segnitz B, Gehring U 1990 Mechanism of action of a steroidal antiglucocorticoid in lymphoid cells. J Biol Chem 265:2789–2796 58. Graham FL, Van der Eb AJ 1973 A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52:456–467 59. Pham TA, Hwung YP, Santiso-Mere D, McDonnell DP, O’Malley BW 1992 Ligand-dependent and -independent function of the transactivation regions of the human estrogen receptor in yeast. Mol Endocrinol 6:1043–1050 60. Granner DK, Thompson EB, Tomkins GM 1970 Dexamethasone phosphate-induced synthesis of tyrosine aminotransferase in hepatoma tissue culture cells. Studies of the early phases of induction and of the steroid requirement for maintenance of the induced rate of synthesis. J Biol Chem 245:1472–1478 61. Hollenberg SM, Weinberger C, Ong ES, Cerelli G, Oro A, Lebo R, Thompson EB, Rosenfeld MG, Evans RM 1985 Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318:635–641 62. Yan M, Dai T, Deak JC, Kyriakis JM, Zon LI, Woodgett JR, Templeton DJ 1994 Activation of stress-activated

Mol Endocrinol, January 2003, 17(1):117–127

63.

64.

65.

66. 67. 68.

69.

70.

127

protein kinase by MEKK1 phosphorylation of its activator SEK1. Nature 372:798–800 Grussenmeyer T, Scheidtmann KH, Hutchinson MA, Eckhart W, Walter G 1985 Complexes of polyoma virus medium T antigen and cellular proteins. Proc Natl Acad Sci USA 82:7952–7954 Hollenberg SM, Giguere V, Segui P, Evans RM 1987 Colocalization of DNA-binding and transcriptional activation functions in the human glucocorticoid receptor. Cell 49:39–46 Eul J, Meyer ME, Tora L, Bocquel MT, Quirin-Stricker C, Chambon P, Gronemeyer H 1989 Expression of active hormone and DNA-binding domains of the chicken progesterone receptor in E coli. EMBO J 8:83–90 Summers MD, Smith GA 1987 A manual of methods for baculovirus vectors and insect cell culture procedures. Tex Agric Exp Stn [Bull]. 155 Wecksler WR, Norman AW 1979 An hydroxylapatite batch assay for the quantitation of 1␣,25-dihydroxyvitamin D3-receptor complexes. Anal Biochem 92:314–323 Cheng Y-CP, Prusoff WF 1973 Relationship between the inhibition constant (Ki) and the concentration of inhibitor which causes 50% inhibition (IC50) of an enzymatic reaction. Biochem Pharmacol 22:3099–3108 DeBlasi A, O’Reilly K, Motulsky HJ 1989 Calculating receptor number from binding experiments using same compound as radioligand and competitor. Trends Pharmacol Sci 10:227–229 Chen S, Wang J, Yu G, Liu W, Pearce D 1997 Androgen and glucocorticoid receptor heterodimer formation. A possible mechanism for mutual inhibition of transcriptional activity. J Biol Chem 272:14087–14092