Modulation of Cell Signaling Pathways Can Enhance or Impair Glucocorticoid-induced Gene

THE JOURNAL OF BIOLOGICAL CHEMISTRY Val. 268, No. 30, Issue of October 25, pp. 22933-22940,1933 0 1993 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U. S. A.

Modulation of Cell Signaling Pathways Can Enhance or Impair Glucocorticoid-induced Gene Expression without Altering the State of Receptor Phosphorylation*

(Received for publication, May 21, 1993)

Marissa L. MoyerSBq, Kristina C. Borrorllll , Betty J. BonaS, Donald B. DeFrancoll , and Steven K. Nordeen$§** From the $Department of Pathology and §Program in Molecular Biology, University of Colorado Health Sciences Center, Denver, Colorado 80262 and the 11 Department of Biological Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania 15260

We have stably introduced expression vectors for the glucocorticoid receptor and a sensitive, hormone-responsive reporter (mouse mammary tumor virus-luciferase) into a human breast carcinoma-derived cell line. Employing this cell line, we have conducted a detailed examination of the induction of glucocorticoid-regulated genes and the phosphorylation of glucocorticoid receptor following pharmacologic manipulation of cell signaling pathways. The hormone response can be enhanced from 2 to 10-fold by activators of protein kinase A, protein kinase C, and inhibitors of protein phosphatase. Forskolin and 8-bromoadenosine 3’:5’- cyclic monophosphate (BrcAMP), but not BrcGMP, enhance the hormone effect, yet surprisingly, phosphodiesterase inhibitors, isobutylmethylxanthine and Ro20-1724, strongly inhibit hormone-mediated induction of the reporter gene. These treatments do not alter cellular receptor content, dexamethasone binding, nor hormone-mediated receptor down-regulation. Tryptic peptide analysis of 32P-labeled receptor reveals that neither BrcAMP, isobutylmethylxanthine, nor the tumor promoter and protein kinase C activator, 12-O-tetradecanoyl-phorbol-13-acetate,detectably alter the state of glucocorticoid receptor phosphorylation. The only agent which alters receptor phosphorylation is the protein phosphatase inhibitor okadaic acid, but only at concentrations higher than required for maximum effects on glucocorticoid receptor transactivation. W e propose that these effectors do not modify receptor directly but alter its interaction with transcription complexes.

The coupling of cell signaling pathways can coordinate processes that occur within distinct cellular compartments, combinatorially increase regulatory capacity, or provide a mechanism for cell- and tissue-specific response without the need to generate entirely new signaling pathways. Recent evidence has begun to link steroid action with other signal transduction pathways. For example, glucocorticoid-regulated

* This work was supported by National Institutes of Health Re- search Grants DK37061 (to S. K. N.) and CA43037 (to D. B. D.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll These authors contributed to this work in distinct but equal measure.

** TO whom correspondence should be addressed Dept. of Pathol- ogy, Box B216, University of Colorado Health Sciences Center, 4200 E. gth Ave., Denver, CO 80262. Tel.: 303-270-5463; Fax: 303-270-6721,

transcription can be modulated by the activation or expression of certain oncogenes, including nos, fos, jun, and ras (1- 10). Since these glucocorticoid response-modulating oncopro- teins include a cytoplasmic protein kinase (mos), transcription factors ( jun and fos), and a GTP-binding protein (ras) , it seems likely that various signal transduction pathways could impact upon nuclear functions of the glucocorticoid receptor. Although steroid receptors are phosphoproteins, es- tablishing a linkage between the modulation of steroid receptor phosphorylation and transcriptional activity has proven elusive. Agents which modulate kinase/phosphatase pathways enhanced the transcriptional activity of the chicken progesterone receptor or the orphan receptor, COUP, in a ligand- independent manner (11-13), a finding potentially of central importance to understanding the biological roles of the orphan receptor subclass of the steroid receptor family. In contrast to the results with chicken progesterone receptor, human progesterone receptor-mediated transcriptional activation was not induced in the absence of hormone by either BrcAMP’ or the protein phosphatase inhibitor, okadaic acid. However, both compounds stimulated hormone-dependent induction (14). Okadaic acid also stimulated transcriptional enhancement by hormone-activated rat glucocorticoid receptor in COS-1 cells. Nonetheless, little or no effect was seen on receptor phosphorylation, suggesting that the inhibitor was acting on some component of the signal transduction pathway other than the receptor itself (15).

To investigate the coupling of phosphorylation-dependent cell signaling pathways to glucocorticoid hormone action, we have employed a cell line, T47D(A1-2), derived from T47D human mammary carcinoma cells and, as previously described (16), engineered to express relatively high levels of rat glucocorticoid receptor (1-2 pmol/mg protein, -50,000-100,000/ cell). These cells also contain a stably integrated luciferase reporter gene under the control of a truncated mouse mammary tumor virus (MMTV) promoter. This promoter contains all sequences required for full glucocorticoid induction but lacks long terminal repeat sequences more than 224 base pairs upstream of the transcription start (17). In this study we have treated T47D(A1-2) cells with a variety of pharmacological agents which modulate different cell signaling pathways and have assessed their effects on both glucocorticoid regulation

’ The abbreviations used are BrcAMP, 8-bromoadenosine 3’:5’ cyclic monophosphate; CAMP, adenosine 3’:5’ cyclic monophosphate; IBMX, isobutylmethylxanthine; TPA, 12-0-tetradecanoyl-phorbol- 13-acetate; MMTV, mouse mammary tumor virus; CAT, chloramphenicol acetyltransferase; dexamethasone, 9a-fluoro-16a-methyl- Il~,l7a,21-trihydroxy-1,4-pregnadiene-3,2O-dione.

22933

22934 GR Phosphorylation and Coupling to Cell Signaling Pathways

of luciferase expression and phosphorylation of the glucocorticoid receptor.

EXPERIMENTAL PROCEDURES

Materials-Hormone and modulators were obtained as follows: BrcAMP, Boehringer Mannheim; dexamethasone and IBMX, Sigma; TPA, 4a TPA, (+)-7-octylindolactam V, LC Services; okadaic acid, Moana Bioproducts; forskolin, Calbiochem. Ro20-1724 was a gift of Dr. D. Cooper. Hormone and modulator stock solutions were ali- quoted and stored at -20 "C as follows: dexamethasone, 20 mM in ethanol; BrcAMP (100 mM) and IBMX (20 mM) in water neutralized with NaOH; forskolin (10 mM) in dimethyl sulfoxide; TPA (1 mg/ ml), 4a-TPA (1 mg/ml), (+)-7-octylindolactam V (4 mM), and Ro20- 1724 (10 mM) in dimethyl sulfoxide. Okadaic acid is obtained as a 125 p M solution in dimethylformamide.

Other reagents were obtained as follows: luciferin (sodium salt), Analytical Luminescence Laboratories; acetyl-coA synthetase and coenzyme A, Sigma; chloramphenicol, Boehringer Mannheim. Radi- olabeled reagents were obtained as follows: [3H]dexamethasone, Amersham Corp., [3H]acetate and [3ZP]orthophosphate, ICN.

Cell Culture and Transfection-The construction of T47D(A1-2) cells was described previously (16). The reporter gene plasmid, pHHluc, bearing a promoter derived from a fragment of the long terminal repeat of MMTV has also been previously described (17). The reporter plasmid was introduced into T47D cells along with the glucocorticoid receptor expression vector due to contamination at the initial preparation of the two plasmids. The presence of the pHHluc plasmid in the T47D(A1-2) clone which expresses high levels of glucocorticoid receptor was recognized only after the description of the cell line. pHHluc contains MMTV sequences to a HpaII site 224 base pairs upstream of the transcription start. This includes all sequences required for glucocorticoid responsiveness.

T47D(A1-2) cells were maintained in modified Eagle's medium supplemented with 5% fetal bovine serum, nonessential amino acids, 2 mM glutamine, 10 mM HEPES, 50 units/ml penicillin, 50 mg/ml streptomycin, and 200 pg/ml G418. The stock culture was maintained under G418 selection but cells split out for experiments were cultured without it. Culture medium was prepared by the Cell Culture Core Facility of the University of Colorado Cancer Center. For experiments assessing the modulation of dexamethasone-mediated luciferase induction, 0.8 X lo6 T47D(A1-2) cells were plated in 60-mm dishes on day 1. Cells were harvested on day 4 for assessment of luciferase activity. Prior to harvest, cells were treated as indicated in the figure legends. For transient transfection experiments, T47D(A1-2) cells were plated at 1.4 X lo6 cells/60-mm dish. The following day the growth medium was replaced by 1 ml of growth medium containing 500 pg/ml DEAE-dextran and 2 pg/ml plasmid. Incubation in the transfection solution was continued for 4 h at 37 "C whereupon the transfection solution was removed and replaced with 1 ml of dimethyl sulfoxide shock buffer. Dimethyl sulfoxide shock buffer is 15% dimethyl sulfoxide in HBS (137 mM NaCl, 5 mM KC1,0.7 mM Na2HP04, 6 mM glucose, 21 mM HEPES, pH 7.1) (18). Shock buffer was removed after 6 min and replaced with 2 ml of growth medium containing 100 p~ chloroquine. After 2 h the chloroquine medium was removed and replaced with 3 ml of growth medium. Forty-eight h after the addition of the transfection solution, cells were treated as indicated for 6 h at which point cells were harvested for assay of luciferase and chloramphenicol acetyltransferase (CAT) activity. All transfection and treatment conditions were performed in duplicate except as noted in Table I1 where conditions were done in quadruplicate. All data reported have been observed in multiple experiments.

Quantitation of Luciferase and CAT Activity-To prepare extracts for quantitation of luciferase and CAT activity, cell monolayers were rinsed twice with wash buffer (40 mM Tris-C1, pH 7.4, 150 mM NaC1, 1 mM EDTA). Cells were lysed by the addition of 0.5 ml of lysis buffer (20 mM K2HP04, pH 7.8, 5 mM MgCl,, 0.5% Triton x-100). The lysate was transferred to a microfuge tube and centrifuged for 2 min to pellet particulates. Luciferase assays were performed the day of harvest. Luciferase activity is not stable in lysis buffer when frozen at -70 "C. CAT activity is completely stable to freezing and storage in lysis buffer. The lysis buffer does not contain reducing agents such as dithiothreitol. As recently reported (19), and as we also have observed, thiols can cause high backgrounds in CAT assays, probably due to chemical reaction between thiols and acetyl-coenzyme A. Luciferase does not require reducing agents, at least in the short term, to maintain its activity. I t also should be noted that the Triton

X-100 in the lysis buffer does not inhibit CAT activity under the conditions used here.

Luciferase assays were performed using a Monolight 2001 lumi- nometer (Analytical Luminescence Laboratories). Extract (usually 50 pl) was added to 0.35 ml of luciferase assay buffer (100 mM K,HPO,, pH 7.8, 15 mM MgS04, 5 mM ATP, 1 mM dithiothreitol). Luciferase- mediated light output was assessed for 10 s with a built-in 2-s delay following injection of 100 g1 of 1 mM luciferin into the reaction chamber. For CAT assays, extract, 50 pl, was added to 200 p1 of CAT assay buffer. Duplicate aliquots, 100 pl, were removed at 4 h for assessment of CAT activity. The CAT assay, which employs [3H] acetate as a precursor, has been described in detail (20). The protein concentration of each extract was determined by dye binding using a commercial kit (Bio-Rad).

Glucocorticoid Receptor Content and Hormone Binding-For determination of dexamethasone binding in whole cell extracts of T47D(A1-2) ceIls, pairs of confluent T-175 flasks were treated with 1 mM BrcAMP, 100 ng/ml TPA, or vehicle and incubated for 6 h. The monolayers were then rinsed with Earle's salts then detached from the dish by incubating in Earle's salts with 1 mM EDTA for 5 min at 37 "C. The cells were then transferred to a centrifuge tube, the flasks again washed with Earle's salts and this rinse added to the centrifuge tube. The cells were pelleted by centrifugation, resuspended in Earle's salts, and again pelleted. This pellet was resuspended in TEDG (10 mM Tris-C1, pH 7.4, 1 mM EDTA, 1 mM dithiothreitol, 10% glycerol) and again pelleted. The pellets were resuspended in 1.5 ml of TEDG with proteinase inhibitors (21) and homogenized by 40 strokes in a Dounce homogenizer with a Teflon pestle. The homogenate was centrifuged at 40,000 revolutions/min for 30 min and the pellet discarded. Hormone binding assays were done on triplicate aliquots of supernatant derived from each pair of flasks. The assays employed 40 nM radiolabeled dexamethasone. Nonspecific binding was determined on duplicate aliquots containing a 100-fold excess of unlabeled dexamethasone. Binding was carried out at 4 "C for 17 h. Bound and unbound dexamethasone were separated using dextran-coated charcoal.

For determination of dexamethasone binding in cells, the growth medium in T-175 flasks of confluent T47D(A1-2) cells was removed and replaced with 15 ml of serum-free modified Eagle's medium containing 20 nM labeled dexamethasone and modulators as indicated. To conserve on radiolabeled dexamethasone, the 20 nM concentration was achieved by mixing labeled and unlabeled dexamethasone 1:4. A control flask contained an excess of unlabeled dexamethasone (3 p ~ ) . Incubation was continued for 6 h except in one flask where the cells were exposed to hormone for only 30 min. Nuclear and cytosolic extracts were prepared as described above with the exception that the homogenization was performed in 0.5 ml and the pellet of the high speed centrifugation was not discarded but instead resuspended in 0.5 ml of TEDG with 0.5 M NaCl and proteinase inhibitors then vortexed intermittently for 1 h at 4 "C to extract glucocorticoid receptors. The salt-extracted pellet was again centrifuged and the supernatant saved as nuclear extract. Receptor-bound dexamethasone in the cytosolic and nuclear extracts was determined by removing the unbound hormone with dextran-treated charcoal.

Potential effects of modulators on cellular glucocorticoid receptor content were also examined by Western blot analyses. Two T-175 flasks were treated for 6 h as indicated. Cells from each group of two flasks were pooled, and 1 ml of nuclear and cytosolic extracts were prepared as described above. Nuclear extracts were dialyzed overnight at 4 "C against 1 liter of TEDG. Dialyzed extracts were clarified by centrifugation at 40,000 revolutions/min for 20 min. One ml of extract was incubated at 4 "C for 4 h with Protein A-Sepharose beads coated with the anti-glucocorticoid receptor monoclonal antibody GR-7 (22). The beads were washed once with TEG (TEDG without dithiothreitol), three times with TEG + 0.3 M NaCI, and once more with TEG. The beads were resuspended in TEG, transferred to a new tube, washed once again with TEG, and pelleted. Protein bound to the beads was extracted with 150 p1 of sample buffer (125 mM Tris-C1, pH 6.8, 2% SDS, 20% glycerol, 10% 6-mercaptoethanol, and 0.01% bromphenol blue) at room temperature for 30 min, followed by boiling for 5 min. After centrifugation to remove the Sepharose beads, the supernatant was electrophoresed on a discontinuous SDS-polyacrylamide gel (7.5% separating gel). Western blot analysis was performed as previously described for progesterone receptors (23), except the antibody used was GR-7 (1:lOOO dilution) and 35S-protein A was employed for detection. The blot was exposed to autoradiographic film for 24 h.

"P-Labeling of Cells and Phosphopeptide Tryptic Mapping of Glu-

GR Phosphorylation and Coupling to Cell Signaling Pathways 22935

cocorticoid Recept~r-~~P-Labeling and tryptic mapping of glucocorticoid receptor phosphopeptides were done essentially as described previously (24) with the following changes. T47D(A1-2) cells were incubated at 37 "C, 5% C02, for 2 h in phosphate-free medium before adding 3 mCi of [32Pi] (carrier-free; 500 mCi/ml). Dexamethasone (1 PM) or other agents (100 ng/ml TPA, 0.5 mM IBMX, 1 mM 8- BrcAMP, or okadaic acid as indicated) were added along with 32P and cells incubated for 4-6 h at 37 "C. Cells were lysed and GR immunoprecipitated from cell-free lysates using BuGR2 ascites (25) diluted to 1:200 and Protein A-Sepharose as described previously (26). Pro- tein A-Sepharose pellets containing 32P-labeled glucocorticoid receptor were solubilized with SDS sample buffer and electrophoresed on 7.5% SDS-polyacrylamide gels.

Dried gel slices containing glucocorticoid receptor were digested with trypsin and resultant peptides separated in two dimensions by electrophoresis and chromatography (26). Separated peptides were quantified by direct counting of thin layer chromatography plates using an AMBIS Radioanalytic Imaging System (San Diego, CA).

RESULTS

Modulators of Cell Signaling Pathways Alter Hormone In- duction of Gene Expression-BrcARP and okadaic acid have been reported to stimulate ligand-independent expression mediated by the chicken progesterone receptor (11); whereas with the human progesterone receptor, both compounds enhance expression only in the presence of hormone (14). We have examined the effects of these compounds, BrcAMP, an activator of protein kinase A, and okadaic acid, an inhibitor of protein phosphatases 1 and 2A, on expression of luciferase directed by the hormone-responsive MMTV promoter. Treat- ment of T47D(A1-2) cells with a maximally stimulatory dose of dexamethasone results in a detectable induction of the low steady state levels of luciferase within 3-4 h increasing up to lo4 fold over basal by 24 h (cf. Fig. 1D). Treatment with BrcAMP or okadaic acid alone k )as little effect on basal levels

120

100

80

60

40

20

n w C Control BrcAMP Forskolin Control Okadaic Acid

Hours after dexamethasone

FIG. 1. A, T47D(A1-2) cells were treated with BrcAMP (1 mM), forskolin (10 p ~ ) , or vehicle in the presence (shaded bars) or absence (open bars) of dexamethasone (100 nM) for 6 h before harvesting the cells for assay of luciferase. B, cells were treated with okadaic acid (50 nM) or vehicle in the presence (shaded bars) or absence (open bars) of dexamethasone as in A. C, cells were treated with BrcAMP, TPA (100 ng/ml), or vehicle in the presence (shaded bars) or absence (open bars) of dexamethasone as in A. D, cells were treated with 100 nM dexamethasone and BrcAMP (1 mM, open symbols) or dexamethasone alone (closed symbols). Cells were harvested at the indicated times for assay of luciferase activity. Luciferase activity in dexamethasone-treated cells measured at 4 h was significantly greater (about 6-fold) than that observed in uninduced cells. Activity in cells treated with BrcAMP in addition to dexamethasone is twice that of cells that received hormone alone at the 4-h time point.

of luciferase expression, but when either is added with dexamethasone, the hormonal induction is enhanced 2-3-fold (Fig. 1, A and 23).

In order to test whether another protein kinase-directed signaling pathway could affect glucocorticoid-mediated transactivation, T47D(A1-2) cells were treated with the protein kinase C activator and tumor promoter, TPA. TPA has little effect on basal expression of the MMTV-luciferase reporter, however, TPA markedly enhances the dexamethasone induction, up to 10-fold (Fig. IC). The effect of TPA or the other compounds, added at the same time as dexamethasone, is evident as soon as a hormonal induction can be detected (3- 4 h). The fold enhancement by the modulators is relatively constant from this time through the first 24 h of induction (Fig 1D).

Specificity of the Modulators-Induction of the transcription of genes, such as collagenase, by TPA is mediated by the transcription factor AP-1 (27, 28). In light of the previous reports of antagonism between fos or jun, the components of AP-1, and glucocorticoids (2-lo), we were somewhat surprised that TPA led to such a strong enhancement of the glucocorticoid-mediated induction. Although TPA is a known activator of protein kinase C, and BrcAMP of protein kinase A, we sought to test the specificity of their effects on glucocorticoid induction of the MMTV promoter. The inactive stereoisomer of TPA fails to enhance the glucocorticoid induction of luciferase expression in T47D(A1-2) cells. Although structurally unrelated to TPA, the tumor promoter (-)-7-octylindolactam V is thought to activate protein kinase C by binding to the same site as TPA (29). Treatment of T47D(A1-2) cells with (-)-7-octylindolactam V enhances the hormone induction of luciferase as effectively as TPA, whereas about a 30-fold higher dose of the less potent stereoisomer, (+)-7-octylindolactam V, is needed to achieve a partial enhancement of the dexamethasone effect. That this effect is related to protein kinase C and not some property related to tumor promotion is indicated by the fact that the tumor promoter, thapsigargin, which acts via mobilization of intracellular calcium stores, has no effect on basal or hormone-induced luciferase expression.'

To test whether activation of protein kinase A can account for the effects of BrcAMP, we treated T47D(A1-2) cells with forskolin to directly activate adenyl cyclase and raise intracellular CAMP. Forskolin treatment enhances the dexamethasone induction to the same degree as BrcAMP (Fig. L4). BrcGMP had no effect on basal or hormone-induced luciferase expression (not shown).

Phosphodiesterase Inhibitors Impair the Hormone Re- sponse-We also sought to raise intracellular CAMP levels by inhibiting its catabolism. To do this we treated T47D(A1-2) cells with either of two widely used, structurally unrelated phosphodiesterase inhibitors, IBMX or Ro20-1724. Neither affected basal luciferase expression but, surprisingly, both compounds strongly impaired the hormone induction, inhibiting it, in some experiments, up to 90% (Fig. 2 A ) . Added together with forskolin, the inhibitory effect of IBMX on dexamethasone induction of luciferase was dominant (Fig. 2B). Similarly, when IBMX was added along with BrcAMP, TPA, or okadaic acid the hormone response was inhibited although some effect of strong enhancers like TPA may still be seen when compared with dexamethasone and IBMX alone (data not shown).

In light of the paradoxical opposing effects of IBMX and BrcAMP on the hormone response, we tested the effects of

M. L. Moyer, B. J. Bona, and S. K. Nordeen, manuscript in preparation.

22936 GR Phosphorylation and Coupling to Cell Signaling Pathways 40

] A

Control IBMX Ro20-1724

F + I

FIG. 2. A , T47D(A1-2) cells were treated with IBMX (0.5 mM), Ro20-1724 (0.5 mM), or vehicle in the presence (shaded bars) or absence (open bars) of dexamethasone (100 nM) for 6 h before harvesting the cells for assay of luciferase. B, cells were treated as indicated in the presence (shaded bars) or absence (open bars) of dexamethasone for 6 h.

-

a 5 C o n t r o l I B M X BrcAMP lBMX+ - O C o n t r o l IBMX TPA TPA+

.E 5000 BrcAMP IBMX

I IC

i? 2000 3 p 1000

Control __

IBMX Dex Dex+ Dex+ IBMX BrcAMP TPA

Dext

FIG. 3. A , T47D(A1-2) cells were transfected with the plasmid, aTthCAT. Expression of the CAT reporter gene in this plasmid is under the control of the promoter of the glycoprotein hormone a subunit gene (30). Transfected cells received the indicated treatments for 6 h prior to harvest of the cells. B, from the same experiment as A, but the cells were transfected with the plasmid, -73/+63 colCAT, which bears a TPA-inducible promoter from the collagenase gene (32). C , from the same experiment as A but the cells were transfected with the plasmid, pHHCAT, which bears a glucocorticoid-inducible promoter from MMTV.

the two on a CAMP-responsive promoter derived from the gene encoding the common a subunit of the glycoprotein hormones (30). Both IBMX and BrcAMP strongly induced expression directed by the a subunit promoter suggesting that both raise cAMP and activate protein kinase A as expected (Fig. 3A). When the two compounds are supplied together, the response was additive. These results support the idea that, while IBMX treatment raises intracellular levels of CAMP, the inhibitory activity of IBMX on glucocorticoid action must be mediated via effects on other signaling pathways. Similarly, IBMX-induced differentiation of 3T3 cells to adipocytes is not reproduced by raising intracellular cAMP (31).

Although no effects of IBMX or Ro20-1724 were seen in the absence of hormone, an inhibition of basal expression might be difficult to visualize over a short term if the message

and protein were sufficiently stable. To analyze whether the effects of the phosphodiesterase inhibitors are specific for steroid receptor-mediated inductions and do not represent a generalized inhibition of gene expression or of other promoter induction processes, T47D(A1-2) cells were transiently transfected with a reporter plasmid bearing the TPA-inducible collagenase promoter (32). As shown in Fig. 3B, the collagenase promoter is induced by TPA in transfected T47D(A1-2) cells. IBMX did not inhibit this induction. In fact, IBMX stimulated both basal expression from the collagenase promoter and the TPA induction about 2-fold. IBMX manifested an inhibitory action only upon the induction of the glucocorticoid-inducible MMTV promoter without affecting basal expression (Fig. 3C) indicative of a selective coupling with the hormone-response pathway.

These results with the transiently transfected MMTV-CAT gene mirror precisely the data with the stably incorporated MMTV-luciferase gene from the same cells assayed on the same extracts. The results shown in Fig. 3C indicate that the action of the modulators documented in Figs. 1 and 2 is not peculiar to the chromosomal location(s) of the stably integrated luciferase reporter gene in this cell line nor to a direct inhibitory effect on the luciferase protein itself. Furthermore, the action of modulators is evidenced on glucocorticoid-responsive promoters in addition to MMTV. We have observed similar phenomena using promoters bearing two copies of a synthetic, palindromic response element adjacent to a viral thymidine kinase promoter or to a minimal promoter with only a TATA box (Fig. 4). With these promoters, interest- ingly, we can see an effect on basal expression with BrcAMP even though the promoters do not contain a known cAMP response element. Expression in cells given BrcAMP and dexamethasone is greater than the sum of expression in the presence of either alone. Thus with these promoters, the effect of BrcAMP would more properly be termed a synergism rather

10000 - 3 8000 g m 2 6000 0. e

4000

P $ 2000

C

2000

I .-

1000 E c a

J= i5= a .9

[

-

1 -

T

1-

A

Control IBMX BrcAMP TPA

B

FIG. 4. T47D(A1-2) cells were transfected with the plasmid, pGRE2ElbCAT, in which CAT expression is directed by a minimal TATA-containing promoter with two copies of a

the indicated treatments for 6 h prior to harvest of the cells. CAT synthetic hormone-response element. Transfected cells received

activity directed by the transiently transfected reporter and luciferase activity directed by stably introduced MMTV-luciferase reporter were assessed on the same extracts.

GR Phosphorylation and Coupling to Cell Signaling Pathways

than an enhancement as seen for the MMTV promoter. Fig. 4B presents the luciferase expression data for the endogenous MMTV-luciferase reporter assayed on the same extracts showing that IBMX inhibited hormone induction and both TPA and BrcAMP enhanced it, just as with the transfected reporters.

The Actions of Modulators on Glucocorticoid Action Are Not via Alteration of Receptor Content or Ability to Bind Hor- mone-A straightforward mechanism that could potentially account for the enhancing or inhibiting action of modulators would be for the modulators to alter the cellular content of receptors by changing receptor synthesis or by altering glucocorticoid-mediated down-regulation of its receptor. In a hepatoma cell system, CAMP effects could be accounted for by increases in cellular receptor content (33). We felt that effects on receptor numbers were unlikely to explain the phenomena we were seeing since the effects were evident as soon as induced luciferase expression could be reliably quan- titated (54 h), particularly in the case of IBMX and TPA, where the effects were quite large to be accounted for by changes in synthesis or turnover of receptor. The experiments presented in Table I and Fig. 5 indicate that neither the enhancing effects of TPA, BrcAMP, or okadaic acid nor the inhibitory effect of IBMX can be ascribed to changes in receptor content, cellular distribution, or hormone binding capacity. The results depicted in Table IA indicate that neither TPA nor BrcAMP affect dexamethasone binding measured in whole cell extracts. In order to test whether modulators may manifest an effect only in the presence of hormone or may alter the down-regulation of receptor, T47D(A1-2) cells were incubated with labeled hormone in the presence or absence of different modulators. Analysis of binding of the hormone indicates that binding is predominantly nuclear, as expected, and that modulators have no significant effect on the quantity or distribution of receptor. Down-regulation of receptor levels is in evidence by 6 h of hormone exposure; dexamethasone binding was about half that seen in cells exposed to hormone for only 30 min. However, this process was unaltered by the modulators. The binding data is echoed by Western blot analyses of receptor content in nuclear and cytosolic extracts from cells treated with TPA or BrcAMP in the presence of dexamethasone (Fig. 5).

Alterations in Hormone-induced Expression by Protein Ki-

TABLE I Dexatnethnsone binding in T47D (Al-2) cells treated with modulators

In experiment A, binding of radiolabeled dexamethasone was assessed in whole cell extracts from cells treated with TPA (100 ng/ml) or BrcAMP (1 mM) for 6 h. In experiment B, cells were treated with 20 nM radiolabeled dexamethasone (dex') alone or in combination with modulators. Treatments were for 6 h except as indicated. Non- specific binding determined in cells given 150-fold excess of unlabeled dexamethasone (9 fmol/mg protein for both cytosolic and nuclear extracts) has been subtracted from the values aiven.

ReceDtor levels Treatment

Whole cell Cytosolic Nuclear f m o l / r n g protein

Experiment A Control TPA BrcAMP

Experiment B dex' (30 min) dex' dex* + TPA dex* + BrcAMP dex* + IBMX dex* + okadaic acid

1620 2170 1390

175 1283 16 578 91 61 1 58 453 86 652

155 461

Dex Dex

Control Dex B d M P TPA C N C N C N C N

+ +

GR (

22937

4-

4-

4-

FIG. 5. T47D(A1-2) cells were treated for 6 h with vehicle alone, dexamethasone (100 IIM) alone, or dexamethasone plus TPA (100 ng/ml) or BrcAMP (1 mM). Nuclear and cytosolic extracts were prepared as described under "Experimental Proce- dures." Glucocorticoid receptors were immunoprecipitated and the precipitates subjected to SDS-polyacrylamide gel electrophoresis. After transfer to nitrocellulose, receptor was visualized by incuhation with antibody and ["S]protein A.

nase C and Protein Kinase A Activators Are Not Correlated with Changes in Glucocorticoid Receptor Phosphorylation- The glucocorticoid receptor is a phosphoprotein, and specific sites become hyperphosphorylated upon addition of agonist (15,34-36). Although the effect of phosphorylation on receptor function is largely unknown, we have shown previously that changes in glucocorticoid receptor phosphorylation can be correlated with changes in i t s nucleocytoplasmic shuttling (24). Since protein phosphorylation has been implicated in the control of many signal transduction pathways (for a review, see Ref. 37), we set out to determine whether modulators of protein kinases and phosphatases which affect glucocorticoid receptor transactivation also influence glucocorticoid receptor phosphorylation. T47D(A1-2) cells were treated with hormone alone or with modulators in the presence of ['zP]orthophosphate. "P-Labeled glucocorticoid receptor was immunoprecipitated from lysates, gel purified, and analyzed by two-dimensional tryptic mapping. Nine major and minor phosphopeptides can be visualized as shown in Fig. 6.

Fig. 6 shows autoradiograms of tryptic maps from cells treated with dexamethasone (A), dexamethasone plus BrcAMP ( B ) , dexamethasone plus TPA (C), dexamethasone plus IBMX (D), or dexamethasone plus okadaic acid ( E ) . Comparison of maps from dexamethasone-treated cells to maps from cells treated with dexamethasone and activators of protein kinase A or protein kinase C reveals no significant changes in phosphorylation of glucocorticoid receptor. Indeed, IBMX and BrcAMP, which lead to opposite effects on hormone-induced expression (IBMX represses and BrcAMP increases), generate virtually identical tryptic maps of glucocorticoid receptor phosphopeptides. Phosphopeptide analysis of receptor dually labeled to steady state with ['%]methionine also indicates that no change in overall "P incorporation is observed upon treatment with modulators (data not shown).

Since we had previously seen hyperphosphorylation of glucocorticoid receptor in rat fibroblast cells with 100 nM okadaic acid (24), we performed an okadaic acid dose-response for glucocorticoid receptor phosphorylation. As can be seen in Fig. 7, treatment of T47D(A1-2) cells with no okadaic acid (A), or 50 nM okadaic acid (B) generated no observable change in receptor phosphorylation. Treatment with 100 nM okadaic acid caused a dramatic increase in the phosphate content of peptide d. Importantly, this is one of the two major glucocor-

22938 GR Phosphorylation and Coupling to Cell Signaling Pathways

I * *

- 8 O e

e B

FIG. 6. Glucocorticoid receptor phosphorylat ion in cel ls treated wi th dexamethasone p lus s igna l t ransduct ion e f fec- tors. The first panel shows a schematic representation of tryptic phosphopeptides. Major peptides that are phosphorylated in the absence of hormone are marked with a shaded circle. Peptides that become hyperphosphorylated upon hormone treatment are marked with a hatched circle. The minor peptides that are phosphorylated in the absence of hormone are marked with a plain circle. Glucocorticoid receptor from "P-labeled T47D(A1-2) cells was immunoprecipitated and analyzed by two-dimensional tryptic mapping. Autoradiographs of separated phosphopeptides are shown from cells treated for 6 h with dexamethasone ( A ) , dexamethasone plus BrcAMP ( E ) , dexamethasone plus TPA ( C ) , dexamethasone plus IBMX (D), or dexamethasone plus SO nM okadaic acid ( E ) . Peptides derived from an equal number of cells were analyzed for each treatment. The higher number of counts apparent in all spots in panel A compared to all other treatments is not a consistent finding.

ticoid receptor peptides that we have previously shown to be hyperphosphorylated in rat fibroblasts treated for 16 h with 100 nM okadaic acid and dephosphorylated in vitro by protein phosphatases 1 and 2A (24). Treatment with 200 nM okadaic acid caused an even greater increase in phosphorylation of peptide d and also caused a hyperphosphorylation of other peptides, including f and b and some other, previously uni- dentified, peptides. There is an overall increase in phosphorylation of glucocorticoid receptor upon treatment with 200 nM okadaic acid, as determined by 32P/35S dual labeling experiments (data not shown). Since this hyperphosphorylation does not occur at the same concentration of okadaic acid that the synergism in gene expression is seen, it is not likely to be responsible for an increased induction of gene expression. This conclusion is supported further by experiments like that of Table 11. T47D(A1-2) cells were treated with hormone for 4 h, until the hormone induction was first apparent, at which time okadaic acid or, for comparison, TPA was added. The incubation in the presence of modulator was continued for either 1 or 2 h more. The enhancement of luciferase activity by both compounds is evident by 1 h of treatment and is fully developed by 2 h. In contrast, no hyperphosphorylation of receptor by 100 mM okadaic acid is observed after 2 h of treatment (data not shown). Similarly, in our studies with monkey kidney cells transfected with glucocorticoid receptor (15), okadaic acid enhanced hormone-mediated gene expression without imposing a change in phosphorylation of the receptor that could be detected following 2 h of treatment. Indeed, the okadaic acid-induced hyperphosphorylation in

FIG. 7. Hyperphosphorylation of glucocorticoid receptor peptide d in okadaic acid-treated cells . In L'U'O "l'-laheled glucocorticoid receptor from T17D(A1-2) cells was immunoprecipitated and analyzed by two-dimensional tryptic mapping. Autoradiographs are shown from cells treated for 4 h with vehicle alone ( A ), 50 nM okadaic acid ( R ) , 100 nM okadaic acid ( C ) , or 200 nM okadaic acid (D). Peptide d becomes hyperphosphorylated with treatment of 100 or 200 nM okadaic acid and is marked by an arrowhead. The spot patterns are less complex than of the previous figure since the dexamethasone-induced spots are reduced or absent.

TABLE I1 Kinetics of the enhancement of the hormone induction okadaic

acid and TPA T47D(A1-2) cells were treated with dexamethasone (dex. 1 0 0 nM)

or vehicle. After 4 h (when induction begins to be detectable) cells were treated with TPA (100 ng/ml), okadaic acid (SO nM), or vehicle. After another 1 or 2 h. as indicated, cells were harvested for assessment of luciferase activity. Another set of hormone-treated cells received okadaic acid or TPA at the same time as dexamethasone and were harvested a t 6 h. The treatment column gives the total time the cells were exposed to each compound. Activities represent the mean obtained from duplicate or quadruplicate dishes. The standard deviation is presented for quadruplicates.

Treatment Luciferane activity

dex, 5 h dex + okadaic acid, 1 h dex + TPA, 1 h

dex, 6 h dex + okadaic acid, 2 h dex + TPA. 2 h

dex + okadaic acid, 6 h dex + TPA, 6 h

-dex + vehicle, 2 h -dex + okadaic acid, 2 h -dex + TPA. 2 h

l$ht unitn/M protein

682s 101 f 1s 123 f 17

131 f 9 265 f 41 s23 f 27

242 SSO

6.6 9.2 9.3

these cells was not apparent until 16 h of okadaic acid treatment (data not shown).

The repression of hormone-induced transactivation by glucocorticoid receptor seen with IBMX appears to be dominant over the effects of the activating agents, including forskolin (Fig. 2B), BrcAMP, TPA, and okadaic acid (data not shown).

GR Phosphorylation and Coupling to Cell Signaling Pathways 22939

We analyzed the two-dimensional tryptic map from cells treated with dexamethasone, IBMX, and okadaic acid to determine if treatment with IBMX represses the okadaic acid induced hyperphosphorylation of peptide d. Fig. 8 shows that the okadaic acid-induced phosphorylation is still present even when cells are treated with IBMX (panel D). Therefore, this provides further evidence that okadaic acid-induced hyperphosphorylation of peptide d exerts little or no effect on the transactivation activities of the glucocorticoid receptor.

DISCUSSION

Steroid receptor pathways are coupled to a number of signal transduction pathways in breast cancer-derived cells. The outcome of this coupling can be to enhance or to inhibit steroid action. In the experiments presented herein, signal transduction pathways were manipulated by pharmacological means. Activators of both protein kinase A and protein kinase C pathways as well as an inhibitor of protein phosphatases 1 and 2A enhanced glucocorticoid stimulation of hormone-responsive promoters. In contrast to the enhancement observed with BrcAMP and forskolin, activators of the protein kinase A pathway, treatment with either of two phosphodiesterase inhibitors, IBMX and Ro20-1724, strongly inhibited the steroid-mediated activation of gene expression. The inhibitory effect of IBMX is dominant. Expression from a CAMP-responsive promoter is induced both by IBMX and BrcAMP suggesting that both are raising cellular levels of CAMP. We suspect that phosphodiesterase inhibitors may lead to the

r ; *s.

FIG. 8. Okadaic acid-induced hyperphosphorylation of peptide d is not abrogated by treatment with IBMX. Autoradi- ographs are shown of two-dimensional phosphopeptide tryptic maps of glucocorticoid receptor from cells treated for 6 h with vehicle ( A ), 100 nM okadaic acid alone ( E ) , IBMX alone ( C ) , or IRMX plus 100 nM okadaic acid (D). Peptide d is marked with an arrowhead and is hyperphosphorylated in okadaic acid-treated cells both in the presence and absence of IBMX. The spot patterns are less complex than those of Fig. 6 since the cells were not hormone-treated and because a shorter autoradiographic exposure was used so as not to overexpose the spot representing peptide d.

activation (or inhibition) of a distinct pathway that is dominant over the pathway activated by BrcAMP and forskolin.

Controls performed using inactive, structurally related compounds and alternative activators of a given pathway suggest that the effects observed represent specific actions of the modulators. The influence of the modulators exhibits some cell type specificity. In a fibroblast cell line, activators of the protein kinase A pathway confer a stronger enhancement of glucocorticoid-induced gene expression than activators of the protein kinase C pathway,’just the opposite of the data shown herein with the breast cancer derived T47D(A1-2) cells. Mod- ulation is observed not only with the truncated MMTV promoter but with other hormone-responsive promoters as well. These data suggest that the interactions of steroid signaling pathways with other signaling pathways are intricate, highly regulated, and represent phenomena central to understanding hormone responsiveness. We have shown further that BrcAMP (but not TPA) treatment of a fibroblast cell line or breast cancer cell line convert.. the glucocorticoid and progesterone antagonist, RU486, into an agonist, indicating that the interaction of cellular signal transduction pathways with steroid receptors also has significant implications toward understanding mechanisms of resistance to therapeutic steroid an- tagonists (38, 39).

Three lines of reasoning suggested that the actions of these modulators we have examined could be mediated by altering receptor phosphorylation. The glucocorticoid receptor is a phosphoprotein, phosphorylated a t multiple sites, whose phosphorylation state changes upon activation by ligand (34- 36). Additionally, many cellular signal transduction pathways alter protein phosphorylation, including those presumed to be activated in these studies. Finally, there is ample precedent for the function of transcription factors to be regulated by phosphorylation, cf Ref. 37, for review. Activation of protein kinase A has recently been reported to enhance the transcriptional activity of hormone-activated estrogen (40) and retinoic acid (41) receptors, as seen for glucocorticoid and human progesterone receptors. Although phosphorylation of receptors has been invoked as a mechanism for protein kinase A effects, direct evidence has been lacking. A direct effect of phosphorylation on glucocorticoid receptor was suggested by a report indicating that, when overexpressed in COS cells along with protein kinase A, glucocorticoid receptor displayed increased sequence-specific DNA binding in oitm (42). In contrast, no changes were observed in DNA binding properties of progesterone (14) or estrogen (40) receptors from protein kinase A-activated cells. The retinoic acid receptor was reported to serve as a protein kinase A substrate in oitro (41) whereas no changes were observed in the overall level of phosphorylation of progesterone receptor upon BrcAMP treatment of T47D cells (14). However, high resolution examination of the phosphorylation state of receptors following activation of phosphorylation-dependent signal transduction pathways has been lacking. We have carefully examined glucocorticoid receptor phosphorylation following treatment with modulators and/or dexamethasone by two-dimensional phosphotryptic peptide analysis. As expected, hormone stim- ulates hyperphosphorylation a t specific sites, however, neither BrcAMP, TPA, nor IBMX, alone nor when added with hormone, detectably alters receptor phosphorylation. In contrast, okadaic acid, which like BrcAMP, enhances the hormone effect about 2-fold, does result in the hyperphosphorylation of one phosphopeptide. However, this change is only seen with treatments of a t least 4 h with a concentration of okadaic acid of 100 nM or greater. As the effect of okadaic acid to enhance glucocorticoid-induced gene expression is manifested

22940 GR Phosphorylation and Coupling

within 1 h at 50 nM okadaic acid, it appears unlikely that this transcriptional effect could be ascribed to the alteration in the state of receptor phosphorylation. Receptor functions, such as recycling, which are only manifested at later times following hormone binding and nuclear translocation, may well be influenced by okadaic acid-induced phosphorylation.

In summary, hormone-induced gene expression can be enhanced or inhibited by manipulation of different signal transduction pathways independently of receptor phosphorylation. We propose that modulators act through phosphorylation of transcription factors that interact with the glucocorticoid receptor. Phosphorylation of these factors would alter their interaction with receptor to enhance or inhibit receptor action. The factors may be part of the basal transcription apparatus or, alternatively, adaptors or coactivators which are the intermediaries between the receptor and the basal transcription apparatus (43). This model could account for our observations that RU486, normally an excellent glucocorticoid and progestin antagonist, may act as an agonist in cells treated with BrcAMP. Alteration of the interfacing of the transcription apparatus with the steroid receptor may allow a productive interaction that results in an induction of a hormone-responsive gene despite the suboptimal conformation of the receptor when complexed to RU486 (38). Disruption of the ability of a modulator to enhance receptor interaction with the transcription apparatus would create a cell relatively resistant to steroid action. This could account for the elevated frequency of dexamethasone-resistance observed in WEHI-7 cells defective for CAMP-dependent protein kinase activity (44).

We previously have reported indirect evidence for a role of adaptors/coactivators in glucocorticoid receptor mediated control of gene expression (45) as have others (46). Ap- proaches are being investigated to explore a role of transcriptional coactivators in the cross-coupling of steroid response pathways with other cellular signal transduction pathways.

Acknowledgments-We thank Drs. J. Hoeffler, V. Allgood, and M. Karin for the gift of plasmids, %TthCAT, PGREZElbCAT, and -73/ +63colCAT, respectively, Dr. J-A. Gustafsson for the GR-7 antibody, Drs. D. Edwards and C. Beck for helpful discussions and technical advice, and Dr. J. Cidlowski for advice on glucocorticoid receptor- ligand binding assays. N. Hart and C. Mraz are acknowledged for their assistance with the manuscript and B. Lieberman for preparation of figures.

REFERENCES 1. Qi, M., Hamilton, B. J., and DeFranco, D. (1989) Mol. Endocrinol. 3,1279-

2. Vacca, A., Screpanti, I., Maroder, M., Petrangeli, E., Frati, L., and Gulino,

3. Lucibello, F. C., Slater, E. P.. Jooss. K. U., Beato, M., and Muller, R. (1990)

1288

A. (1989) Mol. Endocrinol. 3 , 1659-1665

EMBO J. 9,2827-2834

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15. 16.

17. 18.

19. 20. 21.

22.

23.

24.

25. 26. 27.

28.

30. 29.

31.

32.

33.

34. 35.

36.

37. 38.

39.

40. 41.

42.

44. 43.

45. 46.

to Cell Signaling Pathways Diamond. M. I.. Miner. J. N.. Yoshinaea. S. K.. and Yamamoto. K. R.

(1990) ;Science 2 4 9 , i266-1272 Y l

Jonat, C., Rahmsdorf, H. J., Park, K.-K., Cato, A. C. B., Gebel, S., Ponta,

Yang-Yen, H.-F., Chambard, J.-C., Sun, Y.-L., Smeal, T., Schmidt, T. J., H., and Herrlich, P. (1990) Cell 6 2 , 1189-1204

Drouin. J.. and Karin. M. (1990) Cell 62. 1205-1215 Schule, R., hngarajan,’P., Kliewer, S., Ransone, L. J., Bolado, J., Yang,

N., Verma, I. M., and Evans, R. M. (1990) Cell 6 2 , 1217-1226 Shemshedini, L., Krauthe, R., Sassone-Corsi, P., Pornon, A,, and Grone-

meyer, H. (1991) EMBOJ., 10,3839-3849 Touray, M., Ryan, F., Jaggi, R., and Martin, F. (1991) Oncogene 6, 1227-

1234 Jehn, B., Costello, E., Marti, A,, Keon, N., Deane, R., Li, F., Friis, R. R.,

Burri, P. H., Martin, F., and Jaggi, R. (1992) Mol. Cell. Biol. 12 , 3890- 3902

Denner, L. A., Weigel, N. L., Maxwell, B. L., Schrader, W. T., and O’Malley,

Power. R. F.. Lvdon. J. P.. Conneelv. 0. M.. and O’Mallev. B. W. (1991) B. W. (1990) Science 250,1740-1743

Scie’we 252,-1546-1548’

W. (1991) Science 2 5 4 , 1636-1639

_ I “ . . ,

Power, R. F., Mani, S. K., Codina, J., Conneely, 0. M., and O’Malley, B.

Beck, C. A,, Weigel, N. L., and Edwards, D. P. (1991) Mol. Endocrinol. 6 ,

Somers, J. P., and DeFranco, D. B. (1992) Mol. Endocrinol. 6 , 26-34 Nordeen, S. K., Kuhnel, B., Lawler-Heavner, J., Barber, D. A., and Ed-

Nordeen, S. K. (1988) Bmtechnqws 6,454-457 Louata. M. A,. Cleveland. D. W.. and Sollner-Webb. B. (1984) Nucleic

607-620

wards, D. P. (1989) Mol. Endocrinol. 3,1270-1278

Acids Res. 12,5707-5717 , . .

Shepard, A. R., and Eberhardt, N. L. (1992) Biotechniqws 13 , 702-704 Nordeen. S. K.. Green. P. P.. and Fowlkes. D. M. (1987) DNA 6,173-178 Logeat, F. , Pamphile, ‘R., Loosfelt, H., Joiivet, A., Fournier, A., and Mil-

grom, E. (1985) Biochemistry 2 4 , 1029-1035 Ok et, S , Wikstrom, A,-C., Wrange, O., Anderson, B., and Gustaffson, J.- A. (19M) Proc. Natl. Acad. Sci. U. S. A. 8 1 , 1609-1613 Estes, P. A,, Suba, E. J., Lawler-Heavner, J., Elashry-Stowers, D., Wei, L.

L., Toft, D. O., Sullivan, W. P., Horwitz, K. B., and Edwards, D. P.

DeFranco, D. B., Qi, M., Borror, K. C., Garabedian, M. J., and Brautigan, (1987) Biochemistry 26,6250-6262

D. L. 11991) Mol. Endocrinol. 5. 1215-1228 ~ ” . ~ , .” ~ ~ ~. Gamzchu, B., and Harrison, R. (1984) Endocrinology 114,274-279 Ward, G. E., and Kirschner, M. W. (1990) Cell 61,561-577 Angel, P., Imagawa, M., Chiu, R., Stein, B., Imbra, R. J., Rahmsdorf, H. J.,

Lee, W., Mitchell, P., and Tjian, R. (1987) Cell 4 9 , 741-752

Jameson, J. L., Jaffe, R. C., Deutsch, P. J., Albenese, C., and Habener, J. Sugimura, T. (1982) Gann 73,499-507

Wang, H.-U., Watkins, D. C., and Malbon, C. C. (1992) Nature 3 5 8 , 334-

I ~~~~ ~~ ~

Jonat, C., Herrlich, P., and Karin, M. (1987) Cell 4 9 , 729-739

F. (1988) J. Biol. Chem. 263,9879-9886

Angel, P., Baumann, I., Stein, B., Delius, H., Rahmsdorf, H. J., and

Done. Y.. Aronsson. M.. Gustafsson, J.%., and Okret, S. (1989) J. Biol.

337

Herrlich, P. (1987) Mol. Cell. Biol. 7, 2 56 2266

C&m. 2 6 4 , 13679-13683 Orti, E., Bodwell, J. E., and Munck, A. (1992) Endocrine Reu. 13 , 105-128 Orti, E., Mendel, D. B., Smith, L. I., and Munck, A. (1989) J. Biol. Chem.

Tienrungroj, W., Sanchez, E. R., Housley, P. R., Harrison, R. W., and

Hunter. T.. and Karin, M. (1992) Cell 70,375-387

264,9728-9731

Pratt, W. B. (1987) J. Biol. Chem. 262,17342-17349

ma, B. J., and Moyer, M. L. (1993) Mol. Endocrinol. 7 , Nordeen, S . K., BI

Beck, C. A,, Weigel, N. L., Moyer, M. L., Nordeen, S. K., and Edwards, D.

Cho, H., and Katzenellenbogen, B. S. (1993) Mol. Endocrinol. 7, 441-452 Huggenvik, J. I., Collard, M. W., Kim, Y. W., and Sharma, R. P. (1993)

Rangarajan, P. N., Umesono, K., and Evans, R. M. (1992) Mol. Endocrinol.

Ptashne, M., and Gann, A. F. (1990) Nature 346,329-331 Gmol, D. J., Campbell, N. F., and Bourgeois, S. (1986) J. Bid. Chem. 2 6 1 ,

Bastian, L. S., and Nordeen, S. K. (1991) Mol. Endocrinol. 5,619-627 Meyer, M.-E., Gronemeyer, H., Turcotte, B., Bocquel, M.-T., Tasset, D.,

731-742

P. (1993) Proc. Natl. Acad. Sci. U. S. A. 90,4441-4445

Mol. Endocrinol. 7 , 543-550

6,1451-1457

4909-4914

and Chambon, P. (1989) Cell 57,433-442

Documents

Modulation of Cell Signaling Pathways Can Enhance or Impair Glucocorticoid-induced Gene