Interplay between Estrogens, Progestins, Retinoic Acid and a Single

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Val. 269, No. 46, Issue of November 18, pp. 28955-28962, 1994 Printed in U.S.A.

Interplay between Estrogens, Progestins, Retinoic Acid and AP-1 on a Single Regulatory Site in the Progesterone Receptor Gene*

(Received for publication, December 30, 1993, and in revised form, September 6, 1994)

Jean-Frangois Savouret, Michel Rauch, Gerard RedeuilhS, Sokhavuth Sar, Anne Chauchereau, Kathryn Woodruff, Malcolm G. Parker$, and Edwin Milgromn From ZNSERM Unit 135 and VNSERM Unit 33, Hopital de BicBtre, 94275 Le Kremlin-BicBtre Cedex, France and the &Molecular Endocrinolom Laboratory. Imperial Cancer Research Fund, 44 Lincoln’s Inn Fields, ” _ ~

London WC2A 3PX, United Kingdom

Transcriptional regulation of the progesterone receptor gene involves induction by estrogens and down-regulation by progestins, retinoic acid, and AP-1 proteins. We have previously identified an intragenic (+698/+723) estrogen-responsive element present in the progesterone receptor gene, which binds the estradiol receptor and mediates estrogen and 4-OH tamoxifen induction. Pro- gesterone receptor gene expression was equally stimulated by estradiol and 4-OH tamoxifen in the presence of a NH, terminally deleted estrogen receptor mutant lack- ing activation function 1, suggesting that activation function 2 was the predominant activation domain. This was confirmed by the lack of activity of an estrogen receptor mutant deleted of activation function 2. Re- pression by progestins, retinoic acid, and AP-1 was mediated by the same estrogen responsive element al- though retinoic and progesterone receptors as well as AP-1 proteins did not bind to this element. Repression by these proteins appears to involve different transactivating regions of the estrogen receptor. Repression by retinoic receptors involved only activation function 2 whereas repression by progesterone receptor and AP-1 necessitated both functional domains. Since these proteins act without directly contacting the DNA, it seems likely that repression may be achieved by protein-protein interactions among different domains of the estrogen receptor and/or the transcriptional machinery.

The regulation of the intracellular progesterone receptor concentration plays an important role in the physiology of the estrous or menstrual cycles (Vu hai et al., 1977; Bayard et al., 1978) and in various diseases of the female genital tract (Brenner and West, 1975; Lessey et al., 1988; Garcia et al., 1988; Clarke, 1991). In most target tissues, estrogens and progesterone, respectively, induce and down-regulate progesterone receptor expression (Milgrom et al., 1973; Vu Hai et al., 1977; Mester and Baulieu, 1977; Honvitz and McGuire, 1978; Read et al., 1988; Alexander et al., 1989). The antiestrogen 4-OH tamoxifen (4-OHT)’ displays a partial agonistic effect on progesterone receptor expression in breast cancer cells (Honvitz et al., 1978a) and other cell types (Berry et al., 1990). We have identified previously the main regulatory site of the rabbit proges-

* 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.

Fax: 33-1-45212751. II To whom correspondence should be addressed. Tel.: 33-1-45213329;

The abbreviations used are: 4-OHT, 4-OH tamoxifen; CAT, chloram-

vitellogenin A 2 ; MOR, mouse estrogen receptor; AF, activation function; phenicol acetyltransferase; ERE, estrogen-responsive element; vitA2,

Ugl, uteroglobin; Prl, prolactin; BSA, bovine serum albumin.

terone receptor gene by using transient expression of chloram- phenicol acetyltransferase (CAT) constructs containing various putative regulatory regions, as well as DNase footprinting to localize the receptor binding sites (Savouret et al., 1991). This regulatory site consisted of an intragenic (+698/+723) estrogen- responsive element (ERE) which mediated both estrogen and 4-OHT induction. The sa,me ERE conferred progestin down- regulation, which was efficiently suppressed by the antiproges- tin RU 38486 (Savouret et al., 1991). Such regulation appeared to be specific for the progesterone receptor gene ERE since the Xenopus vitellogenin A2 (vitA2) gene ERE mediated neither 4-OHT induction nor progestin down-regulation. That this ERE bound the estrogen receptor but not the progesterone receptor suggested the latter exerted its activity through protein-protein interactions. Mutational analysis showed that the DNA and honnone-binding domains of the progesterone receptor were both required for down-regulation.

In addition to sex steroids, other regulators of progesterone receptor gene expression have been described. Retinoic acid has been shown to decrease the transcription of the progesterone receptor gene (Clarke et al., 1990). Growth factors and phorbol esters have been reported to decrease or increase progesterone receptor mRNA concentration in a cell-specific manner (Roos et al., 1986; Sumida and Pasqualini, 1988). We have investigated whether these regulators acted at different or identical site(s) in the progesterone receptor gene promoter and if they acted through DNA-protein or protein-protein interactions. Finally, we used deletion mutants of estrogen receptor to gain insight into this complex regulatory interplay.

MATERIALS AND METHODS Gene Constructions-The vitellogenin A2 gene ERE (-331/-297) TK-

CAT and the progesterone receptor gene ERE (+698/+729) TKCAT constructs have been previously described (Klein-Hitpass et al., 1986; Savouret et al., 1991). The rabbit progesterone receptor cDNA expression vector also has been described (Guiochon-Mantel et al., 1988). The mouse estrogen receptor (MOR) cDNA is 599 amino acids long. It has been cloned in the pJ3 expression vector to yield the wild-type (MOR 1-599) vector. In other constructions, the estrogen receptor cDNA expression vector was deleted of sequences corresponding to amino acids 1-121 (MOR 121-599) or 340-599 (MOR 1-339) (Lees et al., 1989). Point mutations in the mouse estrogen receptor that disrupt dimerization function have been obtained by Fawell et al. (1990b). Mutant L511R has leucine 511 mutated into arginine and mutant R507A has arginine 507 mutated into alanine. The following expression vectors also were used: RSV-c-Jun (Hirai et al., 1989), RSV-JunB (Ryder et al., 19881, RSV-JunD (Hirai et al., 1989), pSG-c-Fos (Wasylyk et al., 1988), retinoic receptors a and y (Zelent et al., 19891, retinoic receptor p (Brand et al., 19881, and retinoic receptor X a (Leid et al., 1992).

Double-stranded Synthetic Oligonucleotides Used in Gel Retardation Assays-The 37-base pair vitellogenin A2 ERE has been described (Kumar and Chambon, 1988). The rabbit progesterone receptor gene (+698/+723) ERE (Savouret et al., 1991) was AGCTTGAGTTCAGGTC- GACATGACTGAGCTG. The tyrosine aminotransferase gene GREPRE

28955

28956 Regulation of Progesterone Receptor Gene Expression (Jantzen et al., 1987) was AGCTTTGTACAGGATGTTCTAGCTG. The retinoic receptor gene p RARE (de Th6 et al., 1990) was AGCTTGT- AGGGTTCACCGAAAGTTCACTCG. The consensus Ap-1 site (Lee et al., 1987) was AGCTTGGTGACTCACCGGGTGAACG. The pS2 gene ERE (Berry et al., 1989) was AGCTTCCCTGCAAGGTCACGGTGG- CCACCCCGTG. The uteroglobin gene ERE (Slater et el., 1990) was AGCTTGGCCAGGTCACCATGCCCTCGGGG. The prolactin gene ERE (Maurer and Notides, 1987) was AGCTTGCATT'MTGTCACTATGTC- CTAGAGTGCG.

Proteins and Antibodies-Progesterone receptor was purified from rabbit uteri by immunopurification (Logeat et al., 1985), and estrogen receptor was purified from calf uteri by affinity chromatography (Redeuilh et al., 1987). Anti-mouse estrogen receptor polyclonal antiserum MP 16 was obtained from Fawell et al. (1990a). Other monoclonal antibodies were purchased from suppliers: anti-c-Jun and anti-c-Fos were from UBI, anti-retinoic receptor a was from Affinity BioReagents, and anti-human estrogen receptor H222 was from Abbott Laboratories (Erica test).

Cell Culture and Dansfections-HeLa cells were cultured as described (Savouret et al., 1991). Briefly, transfections were done with 0.1 pg of pCH 110 galactosidase vector as internal standard, 10 pg of CAT construct, 1 pg of estrogen receptor expression vector, and 2 pg of the other vectors unless specified otherwise. Total DNA was normalized to 20 pg with herring sperm DNA. CAT assays have been described, and accuracy of transcription initiation at the correct site in the TK promoter was previously determined by primer extension (Savouret et al., 1991).

Gel Retardation Experiments4OS-7 cells were transfected similarly with 10 pg of the various expression vectors (20 pg of total DNA). Transfected cells were harvested a t 4 "C in phosphate-buffered saline. The cell pellet was lysed by freeze-thawing in 20 mM Tris, pH 7.4,O.l mM EDTA, 400 mM KCl, 10% glycerol. Protein concentration was adjusted to 10 mg/ml, and the extracts were stored at -70 "C. Gel retardation assays were performed with 30 pg of whole cell lysate proteins or various amounts of purified proteins preincubated 15 min on ice in 30 pl in the presence of 5 mM MgCI,, 10 mM dithiothreitol, 3% glycerol, 1 pg of poly(d1-dC), and 70 mM KCl. Then, 1 ng of 32P-labeled oligonucleotide probe was added, and incubations were continued for 30 min on ice. When required, antibodies were added simultaneously to the labeled probe. For kinetic experiments, 180 fmol of estrogen receptor were challenged with 500 fmol of progesterone receptor, W a r , or c-dun. Kinetic assays were performed by incubation with the labeled probe for various times up to 30 min. Following the initial 30-min binding incubation, pseudo-first-order dissociation rate assays were performed by adding a 200-fold excess of unlabeled ERE and continuing the incubation for various times a t 20 "C up to 60 min. Electrophoresis was for 100 min a t 4 "C in a 5% polyacrylamide gel (ratio 29:l). The gel and running buffer were 0.5 mM dithiothreitol, 3% glycerol, 0.5 x TBE (45 mM Tris, 65 p~ boric acid, 1 p~ EDTA). The gels were fixed, dried, and autoradio- graphed with direct exposure Kodak films. Bands on the films corresponding to retarded oligonucleotides were quantitated with a Millipore BioImage station.

RESULTS

Multiple Negative Regulations Occur at the Level of a Single Intragenic ERE-We have identified previously an intragenic (+698/+723) estrogen-responsive element in the progesterone receptor gene. This ERE is responsible for both the estrogen induction of the progesterone receptor gene and the progestin- mediated down-regulation of its induced level of expression (Savouret et al., 1991). We used the (+698/+729) TKCAT construct to study the effects of retinoids and the AP-1 system. Phorbol ester (12-0-tetradecanoyl-phorbol-13-acetate) repressed estradiol induction occurring through the intragenic progesterone receptor ERE in HeLa cells (data not shown). By contrast, the estradiol induction of the vitA2 ERE was not repressed by 12-0-tetradecanoyl-phorbol-13-acetate. Similar results were obtained in MCF-7 cells (not shown). These data concur with the reported negative effects of phorbol esters on progesterone receptor expression in breast cancer cells (Roos et al., 1986). These latter experiments could not ascertain which members of the AP-1 family were responsible for the observed repression. We performed cotransfections of the (+698/+729) TKCAT construct in HeLa cells with the estrogen receptor ex-

A2 vit ERE -

.4 .8 1.6 3.2 6.4 pg oncogene expression vector

FIG. 1. Various members of the AP-1 proteins family inhibit the estradiol induction mediated by the ERE of the progesterone receptor (PR) gene, but not the induction mediated by the ERE of the vitA2 gene. (+698/+729) TKCAT (10 pg) or v i m TKCAT (2 pg) were transfected into HeLa cells with the estrogen receptor expression vector (1 pg) and increasing amounts of the various expression vectors for c-Jun, c-Fos, JunB, and Jun D. The repression elicited by the AP-I proteins is presented as residual percentages of the initial estradiol induction.

pression vector and increasing amounts of expression vectors encoding the various specific AP-1 proteins. Fig. 1 shows that all members of the AP-1 system (c-Jun, c-Fos, and variants Jun-B and Jun-D) are grossly equivalent in repression efi- ciency, Jun-D being slightly weaker. The estradiol induction of the vitA2 ERE CAT construct (vitA2 TKCAT) was not affected by the AP-1 family members in HeLa (Figs. 1 and 3) or MCF-7 cells (not shown). The effect of c-Jun on the progesterone receptor gene ERE was suppressed in a concentration-dependent manner by increasing the amount of transfected estrogen receptor (not shown). This inhibition thus depends on the ratio of estrogen receptor to AP-1 proteins.

To analyze the effects of retinoids, we transfected the (+698/ +729) TKCAT vector into MCF-7 cells. CAT activity was nor- mally induced when the endogenous estrogen receptor was stimulated by 10 nM estradiol. Addition of 100 nM all-trans retinoic acid had no effect on the basal level of CAT expression but suppressed the estradiol mediated induction (not shown). Again, these results could not discriminate between the various endogenous retinoic receptors (RARs or RXR). We therefore cotransfected (+698/+729) TKCAT and the estrogen receptor expression vector into HeLa cells. Under these conditions, endogenous retinoic receptors present in HeLa cells were unable to suppress the stimulatory effect of the high levels of transfected estrogen receptor in the presence of estradiol (not shown). Cotransfection of the individual retinoic receptors M a , 0, and y, in the presence of 1-100 nM of all-trans retinoic acid inhibited the induction of (+698/+729) TKCAT by estradiol. When an expression vector for RXRa was used, inhibition was only observed a t 1-10 PM retinoic acid (Fig. 2). Such high concentrations of retinoic acid have been described as necessary to activate RXRa (Mangelsdorf et al., 1990). When expression vectors for M a and RXRa were cotransfected simultaneously, concentrations as high as 10 PM retinoic acid were without effect. This suggested that ligand-dependent down-regulation was provoked only by homodimers of either retinoic receptor but not by RAR/RxR heterodimers. Again, as observed for progesterone receptor and the AP-1 family members, the various retinoic receptors were unable to repress the estradiol induction of the vitA2 TKCAT construct, showing the specificity of this regulation with regard to the progesterone receptor gene (Fig. 3).

Regulation of Progesterone Receptor Gene Expression 28957

Domains of the Estrogen Receptor Involved in the Regulation of Progesterone Receptor Gene Dunscription-The estrogen receptor contains two transactivation functions: one in the NH,- terminal domain called activation function-1 (AF-1, formerly TAF-l), and one in the COOH-terminal steroid-binding domain called activation function-2 (AF-2, formerly TAF-2) (Lees et al., 1989; Tora et al., 1989). 4-OHT is known to be a partial estrogen agonist in terms of progesterone receptor induction, and we have shown previously that it exerts its effect on the intragenic ERE of the progesterone receptor gene (Savouret et al., 1991). To determine the domains of the estrogen receptor involved in progesterone receptor gene regulation, the (+698/+729) TKCAT construct was cotransfected into HeLa cells with expression vectors for wild-type or mutant estrogen receptors (Lees et al., 19891, and the cells were treated with estradiol or 4-OHT. As shown in Fig. 4, AF-2-containing estrogen receptor mutant (MOR 121-599) was sufficient to elicit a maximal effect (4-fold over control) similar to that of the wild-type estrogen receptor whereas the estrogen receptor mutant containing only AF-1 activity (MOR 1-339) was totally inactive. These results indicate that AF-1 does not contribute to the activity of the estrogen receptor in this model. Surprisingly, the partial agonist activity of 4-OHT (2.5-fold over control) was observed even in the absence ofAF-1. Previously, it had been proposed that all agonist activities of 4-OHT were mediated via the AF-1 domain (Berry et al., 1990).

We then asked whether the AF-2 domain of the estrogen receptor also was sufficient to mediate the down-regulation of the progesterone receptor gene by progestins. As shown in Fig. 5A, this was not the case, since only the wild-type estrogen receptor supported down-regulation of progesterone receptor

t

0 0

rn .RXR 0

I I I I I I =+ 1 2 1 1 1 0 9 8 7 6 5 - log M IRA]

FIG. 2. Inhibition by retinoic receptors of the estradiol induction mediated by the progesterone receptor gene ERE. The effect of retinoic acid is dose-dependent. (+698/+729) TKCAT was transfected into HeLa cells with the estrogen receptor expression vector (1 pg) and 2 pg of the various expression vectors for the retinoic receptors. Cells were then treated with 10 nM estradiol alone or in combination with increasing amounts of all-trans retinoic acid. The inhibitory effects are shown as residual percentages of the initial estradiol induction.

PR FIG. 3. The estradiol induction E2 - + +

mediated by the vitA2 gene ERE is insensitive to down-regulation by - - + progestins, retinoic acid, or the AP-1 proteins. The vitA2 TKCAT construct (2 pg) was transfected into HeLa cells to- gether with the estrogen receptor expression vector (1 pg) and, in separate experiments, the expression vectors for progesterone receptor (2 pg), M a (2 pg), or cJun (1.6 pg) and c-Fos (1.6 pg) (alone

gene expression whereas the AF-2 containing mutant (MOR 121-599) supported estrogen induction but not progestin down- regulation. The deletion of the steroid-binding domain in the AF-1-containing mutant (MOR 1-339) abolished the estrogen induction which precluded any study in down-regulation.

Estrogen receptor deletion mutants, devoid of AF-1 or AF-2, also were used in the study of progesterone receptor gene repression in the presence of Ap-1 proteins or retinoic receptors. As shown in Fig. 5B, the behavior of AP-1 proteins was similar to that of the progesterone receptor, the AF-1 region being necessary to observe repression. In contrast, repression by the retinoic receptor occurred even in the absence ofAF-1 (Fig. 5C).

A possible mechanism for the interaction between estrogen receptor and progesterone receptor on the intragenic ERE of the progesterone receptor gene could involve the formation of heterodimers through the dimerization domains present in the steroid binding region of receptors (Fawell et al., 1990a). This hypothesis might also account for the failure of the constitu- tive, steroid-binding domain-deleted progesterone receptor mutant (A 662-930) to down-regulate progesterone receptor gene expression (Savouret et al., 1991). We used two estrogen receptor mutants (L511R and R507A) bearing single amino acid changes in the dimerization domain, which were shown by Fawell et al. (1990b) to abolish receptor dimerization. As seen in Fig. 5A, these estrogen receptor mutants did support progestin down-regulation when cotransfected in the presence of progesterone receptor, suggesting that heterodimerization could not account for the inhibitory effect.

Repressors of Progesterone Receptor Gene Dunscription Do Not Appear to Bind to the Intragenic ERE-We sought to un- derstand the mechanism of down-regulation and determine whether it takes place through competition for DNA-binding sites or protein-protein interactions, either directly on the estrogen receptor or indirectly. In gel retardation experiments, purified estrogen receptor bound to the progesterone receptor gene ERE (Fig. 6A), confirming previous observations using DNase I footprinting (Savouret et al., 1991). The binding appeared weaker than for the vitA2 ERE: scanning densitometry of the retarded bands showed the affinity of the estrogen receptor for the progesterone receptor gene ERE to be roughly 60-fold lower. The binding was nevertheless specific as it was displaced by an excess of unlabeled ERE oligonucleotide. More- over, an anti-estrogen receptor antiserum provoked the char- acteristic supershift of the retarded material.

Gel retardation experiments showed the absence of binding of purified progesterone receptor or d u n to the progesterone receptor gene ERE (Fig. 6 B and C). We have repeated these experiments with purified recombinant vaccinia virus-pro- duced retinoic receptor a (a gift from Dr. H. de "he, Hopital St Louis, Paris, France). Similarly, the progesterone receptor gene ERE displayed no retardation with the retinoic receptor (not shown).

RAR a c-JUN/c-FOS

E2 - + + E2 - + + + + RA - - + C-JUN - - - + +

C-FOS - - + - +

or c0mbined.a; shown). The ceils were then treated with hormones as indicated. ww-

28958 Regulation of Progesterone Receptor Gene Expression

MOR 1-599 MOR 121-599 MOR 1-33'>

E2 - + - - + - - + - OH-Tamox - - + - - + - - +

- o m I r

A MOR 1-599 MOR 121-599 MOR 1-339 L-511-R R-5074

E2 - + + - + + - + + - + + - + + P " + - - + - - + - - + " +

I

FIG. 4. The inductive effects of estradiol and 4-OHT on the ERE of the progesterone receptor gene are mediated by the AF-2 functional domain of the estrogen receptor. (+698/+729) TKCAT was transfected into HeLa cells with 1 pg of either wild-type estrogen receptor (MOR 1-599), the NH, terminally deleted AF-2 containing estrogen receptor (MOR 121-599) or the COOH terminally deleted AF-1 containing estrogen receptor (MOR 1-339) (1 pg each). Cells were treated with 10 nM estradiol or 1 p~ 4-OHT.

The possibility remained that low affinity binding of the repressors to the ERE could be responsible for down-regulation. Such a binding might be difficult to observe by gel retardation due to the disruption of complexes during electrophoresis. We thus used competition experiments to detect labile interactions. Gel retardation assays were performed with purified progesterone receptor and the TAT gene GREPRE. The binding was challenged with unlabeled DNA (the cognate GRE/ PRE, the progesterone receptor gene ERE, and an unrelated oligonucleotide bearing the AP-1-binding site). As seen in Fig. 7, very efficient competition was observed with the cognate GREPRE, but neither the progesterone receptor gene ERE nor the AP-1-binding site were able to compete for progesterone receptor binding. Similar experiments were performed with c-Jun and retinoic receptor a (either as purified recombinant proteins or overexpressed proteins in COS-7 cells). These proteins bound specifically to their cognate elements in a manner that could not be competed by the progesterone receptor gene ERE or by unrelated oligonucleotides (not shown). The exist- ence of a DNA binding interference mechanism by progestins, AP-1, and retinoic receptors in progesterone receptor gene repression is therefore unlikely.

We also analyzed the kinetics of purified estrogen receptor binding and/or dissociation from the progesterone receptor gene ERE by uon-rate" and "off-rate" gel retardation assays in the presence of purified progesterone receptor, c-Jun, or RARa. Neither the migration of the estrogen receptorERE band nor the rate constants of association and dissociation of estrogen receptor to the oligonucleotide bearing the progesterone receptor gene ERE were modified in the presence of repressors (not shown). These results demonstrate the absence of any observ- able direct association between the estrogen receptor and repressors of progesterone receptor gene expression.

The vitA2 and progesterone receptor gene EREs differ sig- nificantly in their ability to interact with the estrogen receptor as can be seen in Fig. 6A. This suggested to us that the lower affinity of the progesterone receptor gene ERE for the estrogen receptor might be linked to its down-regulation by progestins. We performed competition experiments of estrogen receptor binding to the vitA2 ERE with various EREs. The EREs in the vitA2 and uteroglobin (Ugl) genes strongly respond to estrogen stimulation and are not repressed by progestins, while the EREs in the progesterone receptor, pS2, and prolactin (Prl) genes elicit a moderate response to estrogens and are repressed

B MOR 1-599 E2 - + - + Jun - - + +

c MOR 1-599 E2 - + + RA- - +

MOR 121-599 MOR 1-339 - + - + - + - + " + + + - + +

MOR 121-599 MOR 1-339 - + + - + + - - + - - +

by the progesterone receptor gene ERE requires both AF-1 and AF-2 FIG. 5. A, progestin down-regulation of estradiol induction mediated

regions of estrogen receptor. (+698/+729) TKCAT was transfected into

The expression vectors for wild-type estrogen receptor (MOR 1-599), or HeLa cells with the expression vector for progesterone receptor (2 pg).

estrogen receptor mutants (MOR 121-599, MOR 1-339, L511R, and R507A) were cotransfected (1 pg) as indicated. Cells were treated with

progestin R 5020. B, AP-1 protein inhibition of estradiol induction me- 10 nM estradiol alone or in combination with 10 n~ of the synthetic

AF-2 regions of ER. (+698/+729) TKCAT was transfected into HeLa cells diated by the progesterone receptor gene ERE requires both AF-1 and

with the cJun expression vector (1.6 pg). Expression vectors for wild- type estrogen receptor or estrogen receptor deletion mutants were cotransfected as in A. Cells were treated with 10 nM estradiol as shown. C, retinoic acid inhibition of the estradiol induction mediated by the progesterone receptor gene ERE only requires the AF-2 region of estrogen receptor. (+698/+729) TKCAT was transfected into HeLa cells with the RARa expression vector (2 pg). Expression vectors for wild-type estrogen receptor or estrogen receptor deletion mutants were cotransfected as in A. Cells were treated with 10 nM estradiol alone or in combination with 0.1 PM all-trans retinoic acid as shown.

by progestins (Savouret et al., 199112 As shown in Fig. 8, the vitA2 and Ugl EREs efficiently competed for estrogen receptor binding to the vitA2 ERE while the progesterone receptor, pS2, and Prl EREs were less effective. Thus, the ability of a given ERE to undergo transcriptional repression appears to be cor- related with its lower affinity for estrogen receptor.

DISCUSSION

Multiple Regulations of the Progesterone Receptor Gene-The expression of regulatory proteins involved in the control of gene transcription is generally submitted to an array of positive and negative regulations, including autoregulation (Serfling, 1989). The progesterone receptor is induced by estrogens and down- ~

Woodruff, M. G. Parker, and E. Milgrom, unpublished results. J.-F. Savouret, M. Rauch, G. Redeuilh, S. Sar, A. Chauchereau, K.

Regulation of Progesterone Receptor Gene Expression 28959

. ? . . -4 "E)

M I 1

2 3 4 5 6 7 8 9 1 0 1 2 vit A2 ERE PR gene ERE PR gene ERE

3 4 5 1 2 3 4 5 6 7 CREPRE PR gene ERE AP-I RE

FIG. 6. Neither progesterone receptor nor c-Jun interact directly with the intragenic ERE of the progesterone receptor (PR) gene. A, binding of estrogen receptor ( E R ) to the progesterone receptor gene ERE compared to the binding of estrogen receptor to the vitA2 ERE. Lane 1, estrogen receptor (30 fmol) binds a 37-base pair oligonucleotide encompassing the vitA2 gene ERE (-331/-297). Lanes 2-9, estrogen receptor binds the progesterone receptor gene ERE. Lane 2, 30 fmol of estrogen receptor. Lane 3, 150 fmol of estrogen receptor. Lanes 4-9,300 fmol of estrogen receptor. Lane 5, the retarded estrogen receptorDNA complex is supershifted by addition of anti-estrogen receptor antiserum (1 pg of IgG). Lanes 6-9, competition for estrogen receptor binding by a lo-, loo-, 500-, and 1000-fold excess amount of homologous unlabeled oligonucleotide. Lane 10, 100 pg of bovine serum albumin (BSA) added as control. B, the progesterone receptor does not bind to the progesterone receptor gene ERE (lanes 1 and 2 ) but does bind to the tyrosine aminotransferase gene GREPRE (lanes 3 3 ) . Lane 1, 900 fmol of progesterone receptor. Lane 2,900 fmol of progesterone receptor plus anti-progesterone receptor monoclonal antibody (1 pg of I&). Lane 3, 100 pg of BSA added as control. Lane 4,900 fmol of progesterone receptor. Lane 5,900 fmol of progesterone receptor plus anti-progesterone receptor antiserum (1 pg of IgG). C, purified cJun protein does not bind to the progesterone receptor gene ERE (lanes 1 3 ) but does bind to the consensus AP-1 site (lanes P 7 ) . Lanes 1 and 4, c-Jun

Lane 7, 100 pg of BSA added as control. (1 pg) plus anti-c-Jun antiserum (5 pg of 1 8 ) . Lanes 2 and 5, cJun . Lanes 3 and 6, c J u n plus a 100-fold excess of competitor AP-1 oligonucleotide.

regulated by progestins in the female genital tract (Milgrom et al., 1973; Vu Hai et al., 1977) and breast cancer cells (Honvitz and McGuire, 1978) where 4-OHT is a partial agonist (Honvitz et al., 1978). Non-canonical regulations of progesterone receptor expression also have been described (Blankenstein et al., 1983; Korte and Isola, 1984; Ylikomi et al., 1984; and see Clarke, 1991 for review). Retinoids repress progesterone receptor expression in breast cancer cells (Clarke et al., 1990). Growth factors and phorbol esters act differentially according to the target tissue: they repress progesterone receptor expression in human breast cancer cells (Roos et al., 1986) and induce it in rodent uterus (Sumida and Pasqualini, 1988). This suggested to us that developmentally regulated AP-1 regulatory sites may be present in the progesterone receptor gene. Indeed, we have detected a consensus sequence for the AP-1-binding site in the upstream region of the progesterone receptor gene promoter. As a control for the observation of the AP-1 repression of progesterone receptor gene expression (see Fig. 11, we have transfected a TKCAT construct bearing this putative AP- 1-responsive element in HeLa cells. This construct was equally induced by phorbol esters and cotransfection with a c-Jun expression vector.' Thus, the regulation of progesterone receptor expression appears to be very complex as the functionality of discrete regulatory sites may vary under tissue-specific control. Our results also suggest that a given transactivator may act positively in the presence of its cognate binding site but become a repressor when the site is masked, conformationally altered, or absent. Supporting results already have been reported in the case of estradiol regulation of the prolactin gene by Adler et al. (1988).

Diversity in Repression-Negative control mechanisms of gene transcription have received much attention recently (re- viewed in Renkawitz, 1990). In many cases, the regulator binds DNA at its cognate regulatory element. The subsequent mechanism may be by competition with activating factors (Akerblom et al., 1988; Sakai et al., 1988) or abnormal fixation of the regulator (Drouin et al., 1989). In some cases, repression takes place without DNA binding. Such interferences between estrogen receptor and progesterone receptor (Meyer et al., 1989; Shemshedini et al., 19911, as well as the glucocorticoid receptor and cJun (Yang-Yen et al., 1990; Jonat et al., 1990; Schiile et al., 1990) have been reported. Coprecipitation experiments (Jonat et al., 1990) as well as cross-linking studies (Diamond et al., 1990; Yang-Yen et al., 1990) suggest a direct interaction between AP-1 proteins and glucocorticoid receptors. Two groups have reported recently that the shorter form A of the human progesterone receptor might be able to antagonize the transcriptional stimulation elicited by the full size form B, in a hormone-dependent manner and without binding to DNA (Tung et al., 1993; Vegeto et al., 1993).

Implications of Estrogen Receptor Domains in the Progester- one Receptor Down-regulation-We used estrogen receptor deletion mutants in transfection experiments to analyze the mechanisms of the progesterone receptor gene down-regulation. We observed that the three negative regulators do not act through the same mechanism: progesterone receptor and c-Jun require the NH,-terminal domain of estrogen receptor while RARs and RXR repress the NH, terminally deleted estrogen receptor as efficiently as the wild-type ER. This indicates that different repressors act through different mechanisms.


In the course of these experiments, we examined the ability of estradiol and 4-OHT to elicit transcriptional induction of the (+698/+729) TKCAT construct cotransfected with various estrogen receptor mutants. We observed that 4-OHT induction was mediated through AF-2, while the estrogen receptor mutant containing only AF-1 (MOR 1-339) was unable to induce the progesterone receptor gene. This was surprising as the previous report from Berry et al., (1990) linked the effects of 4-OHT to AF-1 function. Our findings suggest that this is not a uni- versal phenomenon and that particular ERE nucleotide sequences, conformation, or cell specificity dictate the outcome of

1 2 3 4 5 6

-Pq;--

fold excess: 10 100 500 10 100 500 10 100 500

competitor: GRE/PRE PR gene ERE non specific DNA (AP-1 RE)

FIG. 7. The intragenic ERE of the progesterone receptor (PR) gene shows no affinity for progesterone receptor. Immunopuri- fied progesterone receptor was used in gel retardation assays with the tyrosine aminotransferase GREPRE. Receptor-PRE complexes were challenged with increasing amounts of either unlabeled homologous oligonucleotide, oligonucleotide containing the progesterone receptor gene ERE sequence, or an unrelated oligonucleotide (AP-1-responsive element (RE)). Following electrophoresis, the gels were autoradio- graphed and the retarded bands quantitated (see “Materials and Meth- ods”). Lane I , 100 pg of BSA added as control. Lane 2: 500 fmol of progesterone receptor were supershifted by an anti-progesterone receptor antiserum (1 pg of I$). Lane 3, 500 fmol of progesterone receptor. Lanes 4-6, competition with the homologous unlabeled oligonucleotide (lo-, loo-, and 500-fold molar excess). Lanes 7-9, competition with progesterone receptor ERE. Lanes 10-12, competition with AP-1- responsive element.

such regulation. A Direct or Indirect Mechanism for Progesterone Receptor

Gene Repression?-Negative regulation of the progesterone receptor gene might result from direct interactions of the repressors with estrogen receptor, free or DNA-bound, or via indirect mechanisms involving other transcription factors or compo- nents of the preinitiation complex.

The differential requirement for the AF-1 and AF-2 domains of estrogen receptor might reflect direct binding of progesterone receptor and c-Jun to the AF-1 region but not to the AF-2 region, while RARs might equally interact with both transactivation functions. Functional differences between AF-1 and AF-2 have been established (Tora et al., 1989; Tasset et al., 1990). If true, this model would imply that interactions occur on the progesterone receptor gene ERE but not on the vitA2 gene ERE, possibly due to a different conformation of estrogen receptor bound to an imperfectly palindromic ERE. DNA- induced conformational modifications of DNA-bound proteins have been described for GCN4 (Weiss et al., 1990) and Fos-Jun (Pate1 et al., 1990). The possibility of conformational variance of the DNA-bound estrogen receptor also has been suggested (Kumar and Chambon, 1988; Martin et al., 1988; Sabbah et al., 1991).

We have compared the down-regulation of progesterone receptor gene induction by AP-1 proteins and retinoic receptors to its previously reported repression by progestins (Savouret et al., 1991). Gel retardation experiments showed that none of the repressors interacted directly with the progesterone receptor gene ERE. Competition experiments, more amenable to the study of weak interactions (i.e. those that cannot be sustained during electrophoresis), showed that the progesterone receptor gene ERE could not compete with the cognate binding elements of any repressor, thereby excluding DNA binding interference. Another possibility was that the repressors modified the kinetics of the interaction of the estrogen receptor with the progesterone receptor gene ERE. On- and off-rate analyses by gel retardation assays in the presence of the various repressors showed no indication of such a mechanism, which further suggested that competition for binding sites or protein-protein interactions in situ were unlikely.

On the other hand, indirect mechanisms, e.g. competition between receptors for receptor-associated (transactivating function intermediary factors) or transcriptional factors, already has been proposed by several authors (Meyer et al., 1989; Doucas et al., 1991; Shemshedini et al., 1991; McEwan et al., 1993; Pfitzner et al., 1993). These authors rely mainly on the squelching model wherein an endogenous factor is titrated through interaction with an overexpressed regulator. Indeed, Meyer et al. (1989) and Shemshedini et al. (1991) repressed the

1 2 3 4 5 6 7 8 9 FIG. 8. Comoetition of estroeen re-

ceptor bindiLg to the vitA2 eRE by vitA2, Ugl, progesterone receptor, pS2 and Prl gene EREs. Experimental conditions were as in Fig. 6A. estrogen receptor binding was challenged with increasing amounts of each unlabeled ERE. Lane 1, 100 pg of BSA, lane 2, estrogen receptor (30 fmol) binding supershifted by an anti-estrogen receptor antiserum. Lane 3, estrogen receptor (30 fmol) binding to the vitA2 ERE. Lanes 4-6, competition with unlabeled vitA2 ERE (lo-, loo-, and 500-fold molar excess). Lanes 7-9, competition with Ugl ERE. Lanes 10- 12, competition with progesterone receptor ERE. Lanes 13-15. comDetition with

10 1112 131415 16 1718

pS2 ERE. Lanes 16-18, competition with Prl ERE.

Regulation of Progesterone

estrogen receptor-mediated induction of the vitellogenin A2 ERE, respectively, by 10-20- and 50-fold molar excesses of cotransfected progesterone receptor. Similarly, Segars et al. (1993) showed "mediated down-regulation of estrogen receptor induction of the vitellogenin A2 ERE in MCF-7 cells by challenging the endogenous estrogen receptor with high amounts of transfected R X R . In initial experiments, we observed specific repression either with phorbol ester (acting on endogenous AP-1 proteins) or with physiological doses of progestins and retinoic acid acting on their endogenous receptors. In cotransfection experiments, the molar excess of the repressor expression vector over the estrogen receptor expression vector was maintained at a 2:l ratio, shown to be effective in the down-regulation experiments involving the progesterone receptor (Savouret et al., 1991). Under these conditions, estradiol induction of the control vitA2 TKCAT construct was never repressed by any of the negative regulators used in this study.

Finally, the target of down-regulation may be a component(s) of the basic transcriptional initiation complex (for a review see Guarente, 1989; Mitchell and Tjian, 1988). Recent data report direct interactions of transcriptional regulators with members of this complex. TFIIB has been shown to interact with COUP- TF, estrogen receptor, and progesterone receptor (Ing et al., 1992), as well as with VP 16 (Lin and Green, 1991; Roberts et al., 1993). Similarly, TFIID interacts directly with c-Re1 (a member of the NF-KB family) (Kerr et al., 1993). Brou et al. (1993) report indirect interactions between various regulators and a population of TFIID-associated factors.

Whatever the targets of down-regulation, it occurred to us that this regulatory mechanism might depend on differential kinetics of estrogen receptor binding to high affinity palindromic sites uersus imperfect sites. Indeed, the imperfect ERE in the progesterone receptor gene is a weaker estrogen receptor binder that the palindromic vitA2 gene ERE. Other imperfect EREs from both the pS2 (Berry et al., 1989) and prolactin genes (Maurer and Notides, 1987) bound purified estrogen receptor poorly in gel retardation experiments. When TKCAT constructs bearing these EREs were cotransfected with estrogen receptor and progesterone receptor expression vectors, they elicited a moderate induction of CAT activity upon estradiol treatment but supported repression of this induction by progestins.' Con- versely, the Ugl ERE (Slater et al., 1990), an efficient high affinity estrogen receptor binder, conferred a very strong estradiol response, comparable to that of the vitA2 ERE, and was insensitive to progestin repression (Savouret et al., 1991).' When these EREs were used as competitors for the binding of estrogen receptor to the vitA2 ERE, the strong inducers (vitA2 and Ugl) competed efficiently while the down-regulatable EREs (progesterone receptor, pS2, and Prl) were less effective, showing a lower affinity for estrogen receptor. Thus, the lower sta- bility of such estrogen receptor.ERE complexes correlates with their ability to be repressed by negatively acting transcriptional regulators, suggesting a link between the two phenom- ena. As these EREs display different nucleotide sequences, the influence of a particular nucleotide or group of nucleotides in a given position seems unlikely. Our data suggest that other genes bearing weak EREs might also be targets for down- regulation. Further studies will be necessary to pinpoint the exact localization of these repressions in the transcriptional complex.

Acknowledgments-We are grateful to Dr. Moshe Yaniv (Inst. Pasteur, Paris, France) for c-Jun and c-Fos vectors, to Dr. Laurent

tors, to Dr. Hughes de The (Hopital Saint Louis, Paris, France) for Bracco (Rhone-Poulenc Rorer, Vltry, France) for Jun-B and Jun-D vec-

M a , 6, RXRa vectors, to Dr. Hughes de The and Dr. Henk Stunnen- berg (EMBL, Heidelberg, Federal Republic of Germany) for purified vaccinia virus expressed recombinant RARa protein, and to Dr. Ronald

Receptor Gene Expression 28961

Evans (Salk Inst., La Jolla, CA) for R A R y vectors. We thank Dr. Robert Taylor (University of California, San Francisco, CA) for critical reading of the manuscript.

REFERENCES

Adler, S., Waterman, M. L., He, X., and Rosenfeld, M. G. (1988) Cell 52, 685495 Alexander, I. E., Clarke, C. L., Shine J., and Sutherland, R. L. (1989) Mol.

Beato, M., Chalepakis. G., Schauer, M., and Slater, E. P. (1989) J. Steroid Biochem.

Berry, M., Nunez, A. M., and Chambon, P. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,

Berry, M., Metzger, D., and Chambon, P. (1990) EMBO J. 9,2811-2818 Blankenstein, M. A., Blaauw, G, Lamberts, S. W. J. , and Mulder, E. (1983) Eur. J .

Brand, N., Petkovich, M., Krust, A,, Chambon, P., De The, H., Marchio,A., Tiollais,

Brenner, R. M., and West, N. B. (19751A. Reu. Physiol. 37,273-302 Brou, C., Chaudhary, S., Davidson, I., Lutz, Y., Wu, J., Egly, J. M., "ora L., and

Clarke, C. L. (1990) Mol. Cell. Endocrinol. 70, C29-C33 De Th6, H., del Mar Vivanco-Ruiz, M., Tiollais, P., Stunnenberg, H., and Dejean, A.

Desbois, C . , Aubert, D., Legrand, C., Pain, B., and Samarut, J. (1991) Cell 67,

Diamond, M. I., Miner, J. N., Yoshinaga, S. K., and Yamamoto, K. R. (1990) Science

Doucas, V., Spyrou, G., and Yaniv, M. (19911 EMBO J. 10,2237-2245 Drouin, J., Tritiri, M. A,, Plante, R. K., Nemer, M., Eriksson, P., and Wrange, 0.

Fawell, S. E., White, R., Hoare, S., Sydenham, M., Page, M., and Parker, M. G.

Garcia, E., Bouchard, P., De Bmx, J., Berdah, J., Frydman, R., Schaison, G., Fawell, S. E., Lees, J. A,, White, R., and Parker, M. G. (1990b) Cell 60,953-962

Milgrom, E., and Perrot-Applanat, M. (1987) J. Clin. Endocrinol. Metab. 67, 80-87

Endocrinol. 3, 1377-1386

32,737-748

1218-1222

Cancer Clin. Oncol. 19,365-370

P., and Dejean, A. (1988) Nature 332,850453

Chambon, P. (1993) EMBO J . 12,489499

(1990) Nature 343, 177-180

731-740

249, 1266-1272

(1989) Mol. Cell. Biol. 9, 5305-5314

(1990a) Proc. Natl. Acad. Sci. U. S. A. 87, 6883-6887

Gill, G., and Patashne, M. (1988) Nature 334, 721-724 Guiochon-Mantel, A., Loosfelt, H., Ragot, T., Bailly, A,, Atger, M., Misrahi, M.,

Hirai, S. I., Ryseck, R. P., Mechta, F., Bravo, R., and Yaniv, M. (1989) EMBO J. 8,

Horwitz, K. B., and McGuire, W. L. (1978) J. Biol. Chem. 253, 2223-2228 Honvitz, K. B., Zava, D. T., Thilagar, A. K., Jensen, E. M., and McGuire, W. L.

Horwitz, K. B., Koseki, Y., and McGuire, W. L. (1978b) Endocrinology 103,1742-

Ing, N. H., Beekman, J. M., Tsai, S. Y., Tsai, M. J., and OMalley, B. W. (1992)

Jantzen, H. M., Striihle, U., Gloss, B., Stewart, F., Schmid, W., Boshart, M.,

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

Kassis, J. A., Walent, J . H., and Gorski, J. (1986) Endocrinology 118, 603-608 Kerr, L. D., Ransone, L. J., Wamsley, P., Schmitt, M. J., Boyer, T. G., Zhou, Q., Berk,

Klein-Hitpass, L., Schorpp, M., Wagner, U., and Ryffel, G. U. (1986) Cell 46,

Perricaudet, M., and Milgrom, E. (1988) Nature 336,695-698

1433-1439

(1978a) Cancer Res. 38, 2434-2437

1751

J. Biol. Chem. 267, 17617-17623

Miksicek, R., and Schutz, G. (1987) Cell 49, 29-38

Herrlich, P. (1990) Cell 62, 1189-1204

A. J., and Verma, I. M. (1993) Nature 365,412419

1wa-1 n m Klein-Hitpass, L., Tsai, S. Y., Greene, G. L., Clark, J. H., Tsai, M. J., and O'Malley, -. . . - . . -

B. W. (19891 Mol. Cell. Bid . 9. 43-49 Korte, J. M., and Isola, J. J. (1988) Mol. Cell. Endocrinol. 58, 93-101 Kumar, V., and Chambon, P. (1988) Cell 55, 145-156

Lees, J. A., Fawell, S. E., and Parker, M. G. (1989) Nucleic Acids Res. 17, 5477- Lee, W.. Mitchell, P., and Tjian, R. (1987) Cell 49, 741-752

5488 Leid, M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M., Zacharewski, T.,

Chen, J. Y., Staub,A., Gamier, J. M., Mader, S., and Chambon, P. (1992) Cell 68,

Lessey, B. A., Killam, A. P., Metzger, D. A,, Haney,A. E , Greene, G. L., and McCarty, 377-395

Lin, Y. S., and Green, M. R. (1991) Cell 64, 971-981 K. S. J . R. (1988) J. Clin. Endocrinol. Metab. 67, 334-342

Logeat, F., Pamphile, R., Loosfelt, H., Jolivet, A., Fournier, A., and Milgrom E.

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

Mangelsdorf, D. J., Ong, E. S., Dyck, J. A., and Evans, R. M. (1990) Nature 345,

Martin, P. M., Berthois, Y., and Jensen, E. V. (1988) Proc. Natl. Acad. Sci. U. S . A.

Maurer, R. A., and Notides, A. C . (1987) Mol. Cell. Biol. 7, 4247-4254 Mester, J., and Baulieu, E. E. (1977) Eur. J. Biochem. 72,403414 Meyer, M. E., Gronemeyer, H., Tbrcotte, B., Bocquel, M. T., Tasset, D., and

Milgrom, E., Luu Thi, M. T., Atger, M., and Baulieu, E. E. (1973) J. Biol. Chem. 248,

Mitchell, P. J., and Tjian, R. (1989) Science 245, 371-378

Read, L. D., Snider, C. E., Miller, J. S., Greene, G. L., and Katzenellenbogen, B. S. Patel, L., Abate, C., and Curran, T. (1990) Nature 347, 572-575

Redeuilh, G., Moncharmont, B.. Secco, C., and Baulieu, E. E. ( 1987) J. Biol. Chem.

Renkawitz, R. (1990) Rends Genet. 6, 192-196 Roberts, S. G. E., Ha, I., Maldonado, E., Reinberg, D., and Green, M. R. (1993)

~I ~~ ~~

(1985) Biochemistry 24, 1029-1035

J. 9, 2827-2834

22P229

85,2533-2537

Chambon, P. (1989) Cell 57,433442

636M374

(19881 Mol. Endocrinol. 2, 263-271

262,6969-6975


Roos, W., Fabbro, D., Kiing, W., Costa, S. D., andEppenberger, U. (1986)Proc. Natl.

Ryder, R., Lau, L. F., and Nathans, D. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,

Sabbah, M., Gouilleux, F., Sola, B., Redeuilh, G., and Baulieu, E. E. (1991) R O C .

Sakai, D. D., Helms, S., Carlsted-Duke, J., Gustafsson, J. A,, Rottman, F. M., and

Savouret, J. E , Bailly, A,, Misrahi, M., Rauch, M., Redeuilh, G., Chauchereau, A,,

Schiile, R., Rangarajan, P., Kliewer, S., Ransone, L. J., Bolado, J., Yand, N., Verma,

Segars, J. H., Marks, M. S., Hirschfeld, S., Driggers, P. H., Martinez, E., Grippo, J.

Serfling, E. (1989) TIC 5,131-133 Shemshedini, L., Knauthe, R., Sassone-Corsi, P., Pomon, A,, and Gronemeyer, H.

(1991) EMBO J. 10,38393849 Slater, E. P., Redeuilh, G., Theis, K., Suske, G., and Beato, M. (1990)

Mol. Endocrinol. 4, 604-610

Nature 363, 741-744

Acad. Sci. U. S. A. 83, 991-995

1487-1491

Natl. Acad. Sci. U. S. A. 88, 390394

Yamamoto, K. R. (1988) Genes & Deu. 2, 1144-1154

and Milgrom, E. (1991) EMBO J. 10, 1875-1883

I. M., and Evans, R. M. (1990) Cell 62, 1217-1226

F., Wahli, W., and Ozato, K. (1993) Mol. Cell. Biol. 13, 2258-2268

Sumida, C., Lecerf, F., and Pasqualini, J. R. (1988) Endocrinology 122, 3-11 Tasset, D., Tora, L., Fromental, C., Scheer, E., and Chambon, P. (1990) Cell 62,

Tora, L., White, J., Brou, C., Tasset, D., Webster, N., Scheer, E., and Chambon, P.

Tung, L., Mohamed, M. K., Hoemer, J. P., Takimoto, G. S., and Horwitz, K. B.

1177-1187

(1989) Cell 59, 477487

Vegeto, E., Shahbaz, M. M., Wen, D. X., Goldman, M. E., OMalley, B. W., and (1993) Mol. Endocrinol. 7, 1256-1265

McDonnell. D. P. (1993) Mol. Endocrinol. 7. 1244-1255 ~ ~ ~~ . ~~~, ~~ ~ ~~~~ ~

Vu Hai, M. T., Logeat, F., Warembourg, M., and Milgrom, E. (1977)Ann. N. EAcad.

Weiss, M. A,, Ellenberger, T., Wobbe, C. R., Lee, J. P., Harrison, S. C., and Struhl, Wasylyk, C., Imler, J. L., and Wasylyk, B. (1988) EMBO J. 7, 2475-2483

Yang-yen, H. F., Chambard, J. L., Sun, Y. L., Smeal, T., Schmidt, T. J., Drouin, J.,

Ylikomi, T., Isola, J., Ratia. J.. Vaha-Tahlo. T.. and Tuohimaa. P. (1984) J. Steroid

Zelent, A,, Krust, A,, Petkovich, M., Kastner, P., and Chambon, P. (1989) Nature

~I ~~ ~~ ~~ ~ ~

Sci. 286, 199-209

K. (1990) Nature 347, 575-578

and Karin, M. (1990) Cell 62, 1205-1215

Biochem. 2 0 , 4 4 5 4 7

339, 714-717

Documents

Interplay between Estrogens, Progestins, Retinoic Acid and a Single