Progesterone Receptor Regulates Decidual Prolactin Expression in ...

0013-7227/99/$03.00/0 Endocrinology Copyright © 1999 by The Endocrine Society

Vol. 140, No. 10 Printed in U.S.A.

Progesterone Receptor Regulates Decidual Prolactin Expression in Differentiating Human Endometrial Stromal Cells* JAN J. BROSENS, NAOKI HAYASHI†,

AND

JOHN O. WHITE

Department of Reproductive Sciences and Medicine, Division of Pediatrics, Obstetrics, and Gynecology, Imperial College of Science, Technology, and Medicine, Hammersmith Hospital, London, United Kingdom W12 ONN ABSTRACT Human endometrial stromal (ES) cells in culture express PRL, a marker of decidualization, in response to sustained activation of protein kinase A (PKA). Cotreatment with the progestin medroxyprogesterone acetate (MPA) enhanced decidual PRL gene activation in the presence of elevated intracellular cAMP levels. This synergy became apparent, at protein and promoter level, after a lag period of 2 days and increased in a time-dependent manner thereafter. Pretreatment with cAMP advanced the time at which synergy between cAMP and MPA was apparent, suggesting that PKA activation sensitized ES cells to the effects of progestins. Analysis of the progesterone receptor (PR) indicated that PR-A was the predominant form in differentiating ES cells, but its abundance decreased markedly during the course of the decidualization response. The decline in PR levels was of functional relevance, as expression of PR-B or PR-A, by transient transfection, dramatically inhibited the activity of a decidual PRL promoter-reporter construct in response to cAMP. Furthermore, the expression of endogenous PRL protein in response to cAMP or cAMP

I

NITIATION of pregnancy represents the successful combination of two independent processes: embryo development and uterine differentiation. The latter is tightly regulated by ovarian sex steroids, estradiol and progesterone. After ovulation, extensive tissue remodeling occurs in the superficial, but not the basal, layer of the endometrium in response to elevated progesterone levels. This process involves secretory transformation of the glandular epithelium followed by decidualization of the stromal compartment in the late luteal phase of the cycle. At a molecular level these morphological events are controlled by highly coordinated activation of certain gene sets (1, 2). The sequential expression of these progesterone-dependent genes defines a limited period of uterine receptivity, controls trophoblast invasion, or, in the absence of pregnancy, maintains vascular integrity

Received February 26, 1999. Address all correspondence and requests for reprints to: Dr. Jan Brosens, Department of Reproductive Sciences and Medicine, Division of Pediatrics, Obstetrics, and Gynecology, Imperial College of Science, Technology, and Medicine, Hammersmith Hospital, Du Cane Road, London, United Kingdom W12 ONN. E-mail: [email protected]. * This work was supported by a Wellcome Trust Clinician Scientist Fellowship (54043; to J.J.B.), Wellcome Trust Project Grant 044874 (to J.O.W.), and The British Council, Kissei Pharmaceutical Co., and the Daiwa Foundation (to N.H.). † Present address: Department of Obstetrics and Gynecology, Saitama Medical Center, Saitama Medical School, 1981 Kamoda Kawagoe, Saitama 350, Japan.

plus MPA was substantially decreased by constitutive expression of green fluorescence protein-tagged PR, which was localized in the nucleus even in the absence of added ligand. Ligand-independent PR inhibition of the decidual PRL promoter was receptor specific, independent of known PR phosphorylation sites, and required minimally a functional DNA-binding domain. Transient expression of steroid receptor coactivator-1e (SRC-1e), but not SRC-1a, allowed synergy between cAMP and MPA without the requirement of sensitization by pretreatment with cAMP. This raised the possibility that SRC-1e was a component of cAMP-dependent sensitization of ES cells, but there was no evidence of altered messenger RNA expression of either SRC-1 isoform during decidualization. In conclusion, cellular PR levels determine the onset of the decidualization response. Initiation of this process requires elevated intracellular cAMP levels that sensitize ES cells to the actions of progestins through down-regulation of cellular PR levels and possibly via modulation of function of an intermediate factor(s) such as SRC-1e. (Endocrinology 140: 4809 – 4820, 1999)

before menstruation (2– 4). Disruption of the endometrial differentiation process is associated with a number of gynecological disorders, such as endometriosis, adenomyosis, and luteal phase defect, and is an important cause of infertility and menstrual disorders (5–7). The cellular responses to progesterone are predominantly mediated by the progesterone receptor (PR), a member of the superfamily of ligand-inducible transcription factors, which also include the receptors for other steroid hormones, retinoic acid, thyroid hormone, vitamins A and D, and a large subfamily of orphan nuclear receptors (8 –10). In the absence of ligand, PR is maintained in an inactive complex containing heat shock proteins (HSPs) (11). Ligand binding results in a distinct conformational change in the receptor, dissociation of the HSPs, dimerization, and binding to specific palindromic glucocorticoid/progesterone response elements (PREs) in the promoter region of target genes. All steroid receptors share the same modular structure consisting of a highly conserved DNA-binding domain (DBD), a C-terminal ligand-binding domain (LBD), and an N-terminal trans-activation domain (AF-1). The LBD of PR also contains a second transcriptional activation domain, AF-2, as well as signals for receptor dimerization, nuclear localization, and HSP binding (8 –10). In the human, as in most species, two isoforms of the PR exist, hPR-A and hPR-B, which arise from different promoter usage in a single gene (12). PR-B differs from PR-A only in that it contains an additional 164 amino acids at the

4809

4810

PROGESTERONE-DEPENDENT DECIDUALIZATION

N-terminus [B upstream sequence (BUS)]. The two PR isoforms display indistinguishable hormone- and DNA-binding properties and can heterodimerize with each other. However, several studies have shown that, depending on the cell and promoter context, PR-A and PR-B display remarkably different transcriptional activities, suggesting that they may have distinct physiological functions (13–15). In general, the hPR-A isoform is transcriptionally less active and can function as a dominant inhibitor of transcription by human (h) PR-B and various other steroid receptors. The functional differences between both isoforms are partially explained by the presence of a third trans-activation domain, AF-3, located in the unique N-terminal segment of the hPR-B (16). On the other hand, a repressor domain has been mapped to the first 140 amino acids of the PR-A that is thought to mediate hPR-A trans-repression of hPR-B-induced gene transcription. In the hPR-B, this trans-repression function is itself inhibited by BUS (17). A growing number of cofactors have been identified that mediate initiation (coactivators) or repression (corepressors) of target gene transcription by nuclear receptors (18 –20). Coactivators such as steroid receptor coactivator-1 (SRC-1), transcription intermediary factor-2, receptor-associated coactivator 3, cAMP response element-binding protein, and p300, not only function as bridging proteins between the agonist-bound nuclear receptor and components of the preinitiation complex, but also display histone acetyltransferase activity. Acetylation of histones is thought to remodel the chromatin, stabilize the preinitiation complex, and hence facilitate transcription. In contrast, antagonist-bound steroid receptors have been shown to recruit corepressors such as silencing mediator for retinoid and thyroid hormone and nuclear receptor corepressor. The receptor-corepressor complex is thought to recruit Sin3 and histone deacetylases, which results in compaction of the chromatin and inhibition of gene transcription (21, 22). Cell surface signal transduction pathways have been shown to modulate the activity of steroid hormone receptors on target genes. The spatial and temporal expression of progesterone-dependent genes in the endometrium suggests indeed that PR function is coordinated by locally produced factors. PRL, a widely used marker of decidual transformation, is first detectable in the superficial endometrial stromal compartment approximately 10 days after the postovulatory rise in circulating progesterone levels. The expression of PRL by differentiating endometrial stromal (ES) cells in vitro is a model of the interactions between PR and cytoplasmatic signaling pathways (23–29). Several studies have shown that the expression of PRL is dependent upon activation of the protein kinase A (PKA) pathway (23–26). Decidualization in vivo coincides indeed with the release of several factors by the endometrium (CRF, relaxin, and PGE2), ovary (relaxin), and pituitary (gonadotropins), which are all capable of inducing PRL in ES cells in vitro through sustained elevation of intracytoplasmic cAMP levels (25–29). In contrast, progestins are very weak inducers of decidual PRL expression in ES cells and require at least 8 –10 days of stimulation. However, progestins have been shown to play an important role in amplifying the decidual

Endo • 1999 Vol 140 • No 10

response, as demonstrated by their ability to synergistically enhance cAMP-induced PRL expression (23–29). Here we demonstrate that the maintenance of cellular PR levels is inhibitory to PRL gene expression and that synergy with progestin becomes apparent only in cAMPsensitized cells in which PR is down-regulated. Sensitization of ES cells by cAMP can be mimicked by transient expression of SRC-1, suggesting that modulation of the activity of this PR coactivator is integral to full activity of the decidual PRL promoter. Materials and Methods Materials DMEM-Ham’s F-12 mixture (DMEM/F12), collagenase (type 1), deoxyribonuclease (type 1), 8-bromo-cAMP, medroxyprogesterone acetate (MPA), 17b-estradiol, BSA, antibiotic-antimycotic solution, diaminobenzidine, Hoechst dye 33258 (bisbenzimidine), calf thymus DNA, bacitracin, phenylmethylsulfonylfluoride, pepstatin A, and leupeptin were all obtained from Sigma Chemical Co. (Poole, UK). FBS, G-418 sulfate penicillin-streptomycin (5000 and 5000 IU/ml) solution were bought from Life Technologies, Inc. (Uxbridge, UK). SV Total RNA Isolation System, Access RT-PCR System, ProFection Mammalian Transfection System, and Luciferase Assay System were purchased from Promega Corp. (Southampton, UK). Plasmid Maxi kits were obtained from QIAGEN (Crawley, UK); Galacto-Light Plus for measurement of bgalactosidase activity was purchased from Tropix (Bedford, MA); the Ready-To-Go T4 polynucleotide kinase and ECL Western blotting detection reagents were obtained from Amersham Pharmacia Biotech (St. Albans, UK).

Primary endometrial cell culture The ES cells from normal proliferative endometrial tissues were isolated from cycling women by endometrial biopsy at the time of diagnostic laparoscopy. The study was approved by Hammersmith and Queen Charlotte’s Hospital research and ethics committee, and patient consent was obtained before biopsy. Samples were collected in Earle’s Buffered Saline containing 100 U/ml penicillin and 100 mg/ml streptomycin. The tissues were washed twice in DMEM/F12, finely minced, and enzymatically digested with collagenase (134 U/ml) and deoxyribonuclease type 1 (156 U/ml) for 1 h at 37 C. After centrifugation at 400 3 g for 4 min, the pellet was resuspended in maintenance medium, a mixture of DMEM/F12, 10% FBS, 2% penicillin-streptomycin, and 1% l-glutamine. ES cells were separated from epithelial cells and passed into culture as described previously (23). Proliferating ES cells were cultured in maintenance medium until confluence. Confluent monolayers were treated in phenol red-free DMEM/F12 containing 2% dextrancoated charcoal-treated FBS (DCC-FBS) with 0.5 mm 8-bromo-cAMP, 1026 m MPA, or 1025 m of the type 1 antiprogestin ZK98.299 (onapristone). All experiments were carried out before the fourth cell passage.

PRL and DNA assays PRL levels in supernatant were measured by microparticle enzyme immunoassay (AxSYM system, Abbott Laboratories, North Chicago, IL). The coefficient of variation within assays was 2–3%, and that between assays was 6 – 8%. DMEM/F12 supplemented with DCC-FBS did not have measurable PRL concentrations. PRL levels were normalized to the DNA content of each culture flask at the end of the treatment period. DNA content was measured by quantitative fluorometric analysis at room temperature. Cells were solubilized with 0.02% SDS. Aliquots were then mixed with 1 mg/ml Hoechst 33258 in 1 3 SSC (standard saline citrate), and fluorescence was measured in a fluorometer at excitation 344 nm and emission 460 nm. Calf thymus DNA was used as standard.

SDS-PAGE, Western blot, and immunodetection Whole cell protein extraction was performed in high salt buffer [0.4 m KCl, 20 mm HEPES (pH 7.4), 1 mm dithiothreitol, and 20% glycerol]

PROGESTERONE-DEPENDENT DECIDUALIZATION containing a cocktail of protease inhibitors (0.5 mg/ml bacitracin, 40 mg/ml phenylmethylsulfonylfluoride, 5 mg/ml pepstatin A, and 5 mg/ml leupeptin). Protein concentrations were determined by Bradford assay (Bio-Rad Laboratories, Inc., Hemel Hempstead, UK). Equal amounts of total protein (200 mg) were separated on a 7.5% SDS-polyacrylamide gel before electrotransfer at 80 V onto nitro-cellulose membrane. Equivalent protein loading was confirmed by Ponceau S staining. Nonspecific binding sites were blocked in 5% nonfat dry milk in 0.1% PBS-Tween. Subsequently, the blot was incubated overnight at 4 C with primary monoclonal antibody to human PR that recognizes both PR isoforms (NCL-PGR, Novacastra Laboratories Ltd., Newcastle upon Tyne, UK). After 1-h incubation with a secondary peroxidase-labeled antibody, protein bands were visualized by enhanced chemiluminescence. Antibody-antigen complexes were removed by incubation at 55 C in stripping buffer (100 mm 2-mercaptoethanol, 2% SDS, and 62.5 mm Tris-HCl, pH 6.7), and the membranes were reprobed with antiserum to human tissue factor (TF; a gift from Dr. John McVeigh, London, UK) or actin (BioGenex Laboratories, Inc., San Ramon, CA).

RT-PCR The SV Total RNA Isolation System (Promega Corp.) was used to extract total RNA from ES cells. One microgram of total RNA was reverse transcribed and amplified in a single reaction with Access RTPCR System (Promega Corp.) according to the manufacturer’s instruction. Simultaneous amplification of SRC-1a and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) complementary DNAs (cDNAs) was performed by adding 10 pmol each of the following oligonucleotides to one reaction: SRC-1a-sense (corresponding to positions 4288 – 4308 of both SRC isoforms cDNAs), SRC-1a-antisense (antisense to positions 4614 – 4634 of the unique SRC-1a 39-UTR), GAPDH-59 (corresponding to positions 212–232), and GAPDH-39 (antisense to positions 790 – 810). The same GAPDH primer set was used for simultaneous amplification of SRC-1e and GAPDH. The SRC-1e oligonucleotides were SRC-1e-sense (corresponding to positions 3844 –3864 of SRC-1a and SRC-1e cDNA) and SRC-1e-antisense (antisense to positions 4474 – 4494 of the unique SRC-1e 39-UTR). The GAPDH cDNA, representing a nonregulated gene, served as an internal control. The negative control sample consisted of reaction mix and primers without RNA template. The reaction was allowed to continue for 29 cycles, which was within the exponential phase of the amplification reaction as determined by cycle profiling. Southern blots of the PCR products were successively hybridized with an internal 32P-labeled oligonucleotide common to both SRC isoforms (positions 4311– 4341) followed by a 32P-labeled GAPDH-59 oligonucleotide. pGEM markers were used for size determination.

Plasmid constructs Complementary DNAs hPR-A and hPR-B cloned into pSG5 expression vector were gifts from Dr. Pierre Chambon (Strasbourg, France). The hPR-B phosphorylation mutants (M1, M3, M9, MH, and Bck (1–5)), containing various sets of serine to alanine substitutions; the hPR-B DNA-binding mutant (hPR-B/DBM) in which cystine 587 was mutated to alanine; and the BUS-DBD-NLS, a mutant in which the B upstream segment is fused to the DBD and nuclear localization signal, were gifts from Dr. Kathryn Horwitz (Denver, CO). The hPR-B ligand-binding mutant (hPRB-B/LBM) was created by linearizing hPR-B at position 2427 with HindIII, and a frame shift was introduced by subcloning a 125-bp fragment, obtained by HindIII digestion of l DNA (NBL, Hitchin, UK), into this site. The plasmid pGFP-C1/hPR-B (GFP, green fluorescent protein) was created by subcloning, in-frame, a BamHI/XbaI fragment from the hPR-B expression vector into pGFP-C1 linearized with BamHI. The reporter vectors dPRL-3000/Luc, carrying 3000 bp of 59-flanking DNA to the decidua-specific promoter of the hPRL gene, and pMSG/ Luc, containing the mouse mammary tumor virus (MMTV) long terminal repeat, were provided by Dr. Birgit Gellersen (Hamburg, Germany). Plasmid pRSV-Ca, encoding the PKA a catalytic subunit, was a gift from Dr. Richard Maurer (Portland, OR). Expression vectors for human SRC-1a and SRC-1e were obtained from Dr. Malcolm Parker (London, UK). pCH110, a b-galactosidase expression vector, was purchased from Pharmacia Biotech (Piscataway, NJ).

4811

Transient transfections ES cells were plated at a density of 5 3 105 cells/well in 12-well plates. Transfections were performed by the calcium phosphate precipitation method in medium supplemented with 2% DCC-FBS. Details of the transfection protocol and treatments are indicated in each figure legend. Cell extracts were harvested, and luciferase and b-galactosidase activities were determined as previously described (23). The b-galactosidase measurements were used to ensure that the transfection efficiencies were comparable when different combinations of expression vectors were used in a given experiment.

Stable transfection and microscopic analysis Unpassaged ES cells, transfected in 6-well plates (Nunc, Inc., Copenhagen, Denmark) with 1 mg GFP or GFP-hPR, were selected for G-418 sulfate (1 mg/ml) resistance. Untransfected cells were used to assess the efficiency of antibiotic selection. Cells surviving G-418 sulfate treatment were pooled, plated in 24-well plates, and allowed to reach confluence. The cultures were assessed using phase contrast and fluorescence microscopy (Diaphot 300, Nikon, Kingston, UK) with a GFP filter (Nikon). Subcellular localization of GFP and GFP-hPR was assessed in untreated cells and after treatment with 8-bromo-cAMP, MPA, or their combination.

Results Modulation of PKA-mediated PRL protein expression by progestins and antiprogestins

PRL secretion by ES cells in response to treatment with the stable cAMP analog, 8-bromo-cAMP, increased progressively in the first 6 days, but thereafter a gradual decline was observed (Fig. 1A). The antiprogestin ZK98.299 (onapristone) partially inhibited PRL secretion in response to 8bromo-cAMP; the effects were apparent after a lag period of 4 days. The observation that an antiprogestin can modulate PRL expression in response to cAMP suggests that PR has an integral role in PKA-mediated ES cell differentiation even in the absence of added ligand. Treatment of ES cells with ZK98.299 alone failed to trigger PRL secretion (data not shown). As expected, MPA did induce PRL expression, but the levels were low and only detectable after 6 days of stimulation. However, MPA potentiated 8-bromo-cAMP induction of PRL after 2 days of treatment. Figure 1B illustrates that the level of synergy between MPA and 8-bromo-cAMP was modest in the first 6 days of treatment, but increased dramatically thereafter. Such temporal responses suggest that the nature of interaction(s) between the PR and PKA pathway may change as the differentiation process proceeds. MPA potentiates decidual PRL promoter activity in response to cAMP

Previously, we reported that the synergy between cAMP and MPA on PRL protein expression was mimicked at the messenger RNA (mRNA) level, but we failed to demonstrate such an effect on decidual PRL promoter activity (23). However, these initial transient transfection studies were carried out in medium containing 10% DCC-FBS, in contrast to the 2% DCC-FBS used for protein and mRNA analysis. Subsequently, we found that the presence of 10% serum greatly inhibited PRL protein expression in response to cAMP treatment and almost abolished the synergy between MPA and cAMP (data not shown). Promoter analysis studies were

4812


Endo • 1999 Vol 140 • No 10

FIG. 1. Time course of PRL induction in primary ES cells by cAMP, progestins, and antiprogestins. A, Confluent ES cells (passage 3) were maintained in phenol red-free DMEM-F12 supplemented with 2% DCC-FBS and received no further treatment (control) or were treated with 8-bromo-cAMP, MPA, ZK98.299, or a combination of these as described in Materials and Methods. The medium was changed every 48 h. The data represent the mean of PRL concentrations (6SD) in the supernatant of triplicate cultures after normalization for DNA content in each well at a given time point. B, Relative effect of MPA or ZK98.299 on 8-bromo-cAMP induced PRL expression. The data show the mean fold enhancement or inhibition by MPA or ZK98.299 of the cAMP response at a given time point. One representative experiment of four similar experiments is shown.

FIG. 2. Progestin enhances cAMP-induced decidual PRL promoter activity. ES cells were maintained in phenol redfree medium containing 2% DCC-FBS without any additional supplements or primed with 0.5 mM 8-bromo-cAMP, 1026 M MPA or a combination of these for 48 h. In the last 7 h of the pretreatment period, cells were transfected with dPRL-3000/ Luc (1 mg/well). A constitutively expressed b-galactosidase gene was included in the transfections (pCH110; 0.2 mg/well). Subsequently, the cells received no further treatment (control) or were treated with 8-bromo-cAMP, MPA, or a combination of these. Luciferase and galactosidase assays were performed as described in Materials and Methods. The results represent the mean (6SD) of triplicate measurements of one representative experiment.

therefore performed with ES cells that had been maintained in 2% DCC-FBS and primed with 8-bromo-cAMP, MPA or a combination of these, for 48 h. Figure 2 shows that pretreatment with cAMP markedly enhanced the promoter activity in response to subsequent cAMP treatment. Priming of ES cells with MPA plus cAMP further increased the reporter

gene response to subsequent cAMP treatment. Pretreatment with MPA alone had no discernible effect on subsequent promoter activity. Synergy between cAMP and MPA on decidual PRL promoter activity was apparent, but only if the differentiation process had been initiated by pretreating ES cells with cAMP


4813

for 48 h before transfection. Addition of MPA to cAMP pretreatment did not further increase the level of synergy between cAMP and MPA on subsequent promoter activity. Together these data suggest that during the initial period of the response, cAMP sensitizes ES cells to the actions of MPA. Hence, the synergy between MPA and cAMP on decidual PRL expression could be dependent upon the induction or modulation of an intermediate factor(s) by cAMP. An obvious candidate for such an intermediate factor is PR. PR expression in differentiating ES cells

There is considerable evidence that cAMP not only enhances PR trans-activation of model reporter constructs, but also induces PR expression in uterine and ovarian cells (30 – 32). Hence, we studied the expression of PR protein in differentiating ES cells during the period of sensitization of the decidual PRL promoter to MPA. When ES cells were maintained in phenol red-free medium containing 2% DCC-FBS for 48 h before treatment, only the PR-A isoform was detected by immunoblotting (Fig. 3A). Activation of PKA resulted initially in rapid up-regulation of the receptor with the appearance of multiple species, attributable to receptor phosphorylation (33, 34). Subsequently, PR-A levels declined in response to cAMP, and this process was consistently found to be accelerated by the addition of MPA. Undifferentiated ES cells maintained in 10% FBS expressed both PR isoforms, PR-A and PR-B (Fig. 3B). When treated in phenol red-free medium containing 2% DCC-FBS with cAMP or cAMP plus MPA the levels of both PR isoforms progressively declined during the decidualization response. The fall in cellular PR levels in response to 8-bromo-cAMP treatment was again accelerated by the addition of MPA. A distinct PR-A signal was observed in untreated ES cells maintained in low serum for 4 days (control, 96 h), indicating that the decline of the PR-A isoform during ES cell differentiation did not merely reflect altered serum levels or withdrawal of an estrogenic stimulus (phenol red). In contrast, down-regulation of PR-B is caused by the altered culture conditions, and this process is accelerated in the presence of MPA. The induction of TF, a decidual marker (3), correlated inversely with cellular PR levels. This reciprocal relationship between PR levels and expression of the decidual phenotype was further confirmed at the cellular level using confocal laser microscopy and double labeling for PR and TF (Li, X.-F., J. Brosens, and J. White, unpublished observations). Expression of PR in differentiated ES cells inhibits decidual PRL promoter activity

To establish whether the decline in cellular PR levels was an obligate component of the decidualization process, the effect of reexpression of PR was investigated. cAMP-primed ES cells were therefore transfected with hPR-B expression vector (400 ng plasmid/well) and the reporter construct dPRL-3000/Luc, and the response to treatment was compared with that of cells transfected with dPRL-3000/Luc and the empty control vector pSG5. In the absence of overexpressed hPR-B, addition of MPA to cAMP treatment markedly enhanced promoter activation (Fig. 4A). However, in the presence of hPR-B there was a dramatic repression of

FIG. 3. Induction of the decidualized phenotype correlates with the progressive decline in PR levels. A, Confluent cells were maintained in phenol red-free medium containing 2% DCC-FBS for 48 h before treatment with 8-bromo-cAMP or 8-bromo-cAMP plus MPA (C1M). PR-A, but not PR-B, was detected in whole cell extracts of ES cells, resolved by SDS-PAGE, by immunoblotting with a monoclonal PR antibody that recognizes both PR subtypes, as described in Materials and Methods. B, PR and TF expression in differentiating ES cells. Undifferentiated ES cells were maintained in phenol red-free DMEMF12 medium supplemented with 10% FBS until confluence. Some cultures were then harvested for total protein extraction (control, 0 h), and the medium on the remaining cultures was changed to phenol red-free DMEM-F12 containing 2% DCC-FBS and supplemented with 8-bromo-cAMP or 8-bromo-cAMP plus MPA (C1M) for 24, 48, or 96 h. Untreated cells, maintained in 2% DCC-FBS, were also harvested at 96 h (control, 96 h). Immunodetection of PR was first performed, and the membrane was subsequently stripped and incubated with antiserum to TF, and finally stripped again and reprobed with antiserum to actin to demonstrate equal loading.

cAMP-induced reporter gene transcription in the absence and presence of MPA. Expression of the catalytic a-subunit (Ca-subunit) of the PKA holoenzyme was sufficient to induce dPRL-3000/Luc activation (Fig. 4B), and coexpression of hPR-B (200 ng plasmid/well) inhibited promoter activity in response to Ca-subunit in the absence of ligand. However, at these lower PR expression vector input levels, addition of MPA or ZK98.299 further enhanced the inhibition of reporter gene transcription (Fig. 4B). Constitutive PR expression inhibits PRL protein secretion in differentiating ES cells

Next, we aimed to determine whether PR was also capable of inhibiting the expression of the endogenous PRL gene. Primary ES cells were transfected with plasmids encoding the GFP or a fusion protein in which GFP was coupled to the N-terminal domain of PR-B (GFP-hPR). Transient transfection studies showed that GFP-hPR was capable of inhibiting cAMP-dependent activation of the decidual PRL promoter as well as activating a classical PR-responsive promoter, MMTV-Luc, in the presence of MPA (data not shown). Epifluoresence studies demonstrated that GFP-hPR, but not

4814


Endo • 1999 Vol 140 • No 10

FIG. 4. Overexpression of PR inhibits decidual PRL promoter activity in differentiating ES cells in a ligand-independent manner. A, The decidualization process was initiated in ES cells by pretreatment with 8-bromo-cAMP for 48 h in medium containing 2% DCC-FBS. The cells were transiently transfected with dPRL-3000/Luc (1.2 mg/well) and hPR-B expression vector (hPR-B; 0.4 mg/well) or the empty expression vector pSG5 (0.4 mg/well). The b-galactosidase control vector (pCH110; 0.2 mg/well) was also transfected. The total amount of plasmid transfected in each well was 1.8 mg. Subsequently, the cells were incubated with no supplements (control) or with 8-bromo-cAMP, MPA, or a combination. Cells were harvested for reporter analysis 40 h later. The data represent the mean relative luciferase activity (6SD) of triplicate measurements, and one representative experiment of six similar experiments is shown. B, ES cells, pretreated with 8-bromo-cAMP for 48 h, were transfected with dPRL-3000/Luc (1.2 mg/well) and expression vectors encoding the a-subunit of PKA (pRSV-Ca; 0.2 mg/well) and hPR-B (0.2 mg/well). pSG5 was used as filler to ensure that the total amount of DNA transfected in each well was 1.6 mg/well. The cultures then received no additional supplements (control) or were treated with MPA or ZK98.299. Cells were harvested for reporter analysis after 40 h of treatment. The results represent the mean luciferase activity (6SD) of triplicate measurements.

GFP, was localized predominantly in the nucleus in the presence or absence of cAMP or MPA (Fig. 5A), indicating that cytoplasmic-nuclear shuttling was not a prerequisite for PR inhibition of the decidual PRL promoter. Subsequently, primary ES cells transfected with GFP or GFP-hPR were selected for resistance to G-418 and expanded to allow measurements of PRL protein production in response to decidualizing stimuli. Because of the limited life span of primary ES cells, subcloning of stably transfected cells was not feasible. However, microscopically 25–30% of the G418-resistant cells clearly expressed the transfected plasmids (data not shown). The level of PRL protein secretion in response to cAMP or cAMP plus MPA by ES cells, which constitutively expressed GFP-PR, was approximately 55% of the levels expressed by cells in which GFP had been stably transfected (Fig. 5B), demonstrating that elevated cellular PR levels inhibited activation of the endogenous decidual PRL gene. PR does not trans-activate the MMTV-promoter in ES cells in the absence of ligand

Unliganded PR is thought to be maintained in a transcriptionally inactive complex with HSPs. However, several studies have shown that elevated intracellular cAMP levels not only enhance the transcriptional activity of various steroid hormone receptors, but may also result in activation of target promoters in the absence of ligand (33–35). Hence, progesterone-independent PR repression of the decidual PRL promoter in ES cells may reflect ligand-independent PR activation in response to activation of the PKA pathway. To test this hypothesis ES cells were transfected with the MMTV-

reporter vector with or without an expression vector for the Ca-subunit of PKA (pRSV-Ca). In addition, hPR-A or hPR-B was coexpressed. As expected, the presence of the Ca-subunit of the PKA holoenzyme markedly enhanced the transcriptional activation of ligand-occupied PR receptor, and this effect was more pronounced for the hPR-B than the hPR-A isoform (Fig. 6). Overexpression of PKA-C a-subunit did not, however, result in PR trans-activation of the MMTV promoter in the absence of ligand (Fig. 6). This indicates that PR inhibition of the decidual PRL promoter occurs without acquisition of generalized ligand-independent transcriptional competence. PR repression of the decidual PRL promoter is independent of its phosphorylation status

PR, like all other steroid receptors, is a phosphoprotein and site-specific phosphorylation of the receptor may modulate its function. To test whether PR phosphorylation is required for the repression of the decidual PRL promoter in ES cells, we tested several sets of serine to alanine substitution mutants in hPR-B (34). Among the mutated residues were ones that have been identified as authentic phosphoserines in hPR-B, including (Ser81, Ser102, Ser162, Ser190, and Ser345); ones that have been identified as phosphoserines in the chicken PR (cPR) and bear homology to sites in hPR (Ser345 and Ser676); and ones that represent consensus phosphorylation sites for casein kinase II (CKII) and proline-directed kinases and have at least a Ser-Pro motif (Ser549, Ser552, Ser554, Ser558, and Ser561) (34, 36 –38). The level of inhibition of the decidual PRL promoter by the PR-B phosphomutants was comparable


4815

FIG. 5. Constitutive expression of PR in differentiating ES cells inhibits PRL protein expression. ES cells (passage 1) were transiently or stably transfected with the GFP expression vector, pGFP-C1, or the fusion expression vector, pGFP-C1/hPR, encoding an N-terminus-truncated hPR-B protein fused to the C-terminus of GFP (GFP-hPR) as described in Materials and Methods. A, Subcellular localization of GFP and GFP-PR. Epifluorescence microscopy of ES cells transiently transfected with GFP or GFP-PR showed cytoplasmatic localization of GFP. In contrast, GFP-PR was found to be nuclear and perinuclear in ES cells regardless of the presence or absence of treatment. B, ES cells (passage 1) were stably transfected with GFP or GFP-PR. Cells resistant to G-418 were treated in phenol red-free DMEM-F12 containing 2% DCC with 8-bromo-cAMP, MPA (1026 M), or a combination of these or remained untreated for 6 days. The medium was changed every 48 h. PRL concentrations in the supernatant were corrected for the DNA content of each culture flask at the end of the treatment period.

to that of the wild-type receptor (data not shown), indicating that the mechanism of PR repression of PRL expression is unlikely to involve receptor phosphorylation. PR domains required for repression of decidual PRL transcription

As none of the major phosphorylation sites in PR was implicated in its repression function, we further investigated the domains of the receptor necessary for inhibition of the decidual PRL promoter. The PR mutants tested included a LBD mutant (hPR-B/LBD-M), a deletion construct in the LBD region of the receptor at position 809, a DBD mutant (hPR-B/DBD-M), a point mutant containing a cystine to alanine substitution in the base of the first zinc finger (C587A), and the BUS-DBD-NLS construct in which the B upstream segment of hPR-B is fused to the DBD and the nuclear localization signal. Table 1 shows that deletion of the LBD or disruption of the DNA-binding activity of the receptor prevents its repression of cAMP-mediated decidual PRL promoter activation. Interestingly, the BUS-DBD-nuclear localization signal construct, which lacks the LBD, was as capable of suppressing decidual PRL gene transcription, suggesting that cooperation between these minimal regions of the receptor is sufficient to confer repression of the decidual PRL promoter. The trans-activation domains of PR (AF-1, located in the N-terminus of PR-A; AF-2, located in the C-terminus; and AF-3, located in the BUS region of PR-B) appear not to be essential for inhibition of decidual PRL transcription.

FIG. 6. PR is not transcriptionally competent in differentiating ES stromal cells in the absence of progestins. Undifferentiated ES cells were maintained in phenol red-free medium containing 2% DCC-FBS for 48 h before transfection with MMTV-Luc reporter construct (1 mg/well) and expression vectors for hPR-A or hPR-B (0.2 mg/well). In addition, pSG5 or pRSV-Ca, an expression plasmid encoding the Ca-subunit of PKA, was also cotransfected (0.2 mg/well). The cultures were not treated (control) or were supplemented with MPA or ZK98.299. Reporter gene activity was measured 40 h later, and the data represent the mean activity (6SD) of triplicate measurements.

4816

Endo • 1999 Vol 140 • No 10


TABLE 1. Domain specificity of hPR-mediated repression of dPRL-3000 promoter activity in ES cells Fold induction Expression vector

cAMP

cAMP/MPA

pSG5

93.0 6 16.5

177.4 6 32.3

hPR-B

3.7 6 0.5

5.9 6 1.2

hPR-B/DBD-M

148.5 6 23.2

258.1 6 32.6

BUS-DBD-NLS

6.1 6 0.7

3.7 6 0.1

hPR-B/LBD-M

125.9 6 24.1

234.3 6 59.2

4.9 6 1.1

7.3 6 1.2

hPR-A

Transfection analysis. cAMP-pretreated ES cells (48 h) maintained in low serum (2% DCC-FBS) were transfected for 7 h with dPRL-3000/Luc (1.2 mg/well), pCH110 (0.2 mg/well), and an expression vector for wild-type hPR-B, wild-type hPR-A, hPR-B DNA-binding domain mutant (hPR-B/DBD-M), hPR-B ligand-binding domain mutant (PR-B/LBM), B-upstream segment fused to DNA-binding domain, and nuclear localization signal (BUS-DBD-NLS) or the control expression vector pSG5 at a concentration of 0.4 mg/well. The left column shows the major structural domains of the various PR constructs and their amino acid borders. The cultures remained untreated (control) or were treated for 40 h with 0.5 mM 8-bromo-cAMP or 0.5 mM 8-bromo-cAMP plus 1 mM MPA. Triplicate cultures were harvested, and luciferase activity was normalized to b-galactosidase activity. The results show the mean fold induction (6SD) over the control value.

Ligand-independent inhibition of the decidual PRL promoter is specific to PR

The ability of other steroid hormone receptors to inhibit decidual PRL promoter activity in the absence of ligand is presented in Fig. 7. ES cells were primed with 8-bromocAMP for 48 h before transfection of the dPRL-3000/Luc construct and expression vectors for the human androgen receptor (hAR), human glucocorticoid receptor (hGR), human estrogen receptor a (hERa), hPR-A, hPR-B, or the control vector pSG5. Subsequently, the cells were maintained in the presence or absence of 8-bromo-cAMP for 40 h before analysis of promoter activity. In the absence of ligand only, PR-A and PR-B, but not ERa, GR, or AR, inhibited decidual PRL transcription in response to PKA activation, emphasizing the unique role of PR in regulating the onset of expression of this decidualization marker. SRC-1 potentiates PKA-dependent activation of dPRL promoter in the presence of progestin

We have shown that MPA potentiates PKA-dependent activation of the decidual PRL promoter, but only if the differentiation process in ES cells had been initiated by pretreatment with 8-bromo-cAMP for approximately 48 h (Fig. 2). In undifferentiated ES cells, progestins failed to enhance promoter activity in response to cAMP. Hence, we speculated that cAMP treatment may induce specific PR coactivator(s), which, in concert with declining PR levels, could account for the time-dependent stimulatory effect of progestins. SRC-1 is one of the best characterized members of the

growing family of coactivators that are capable of enhancing ligand-dependent receptor transcriptional activity (39 – 41). In the human two SRC-1 isoforms exist, SRC-1a and SRC-1e, which diverge at their C-termini. Recently, it has been shown that SRC-1e can enhance ER-dependent transcription to a far greater extent than SRC-1a, illustrating the important functional differences between both isoforms (40). When overexpressed in undifferentiated ES cells (no cAMP priming), SRC-1e, but not SRC-1a, significantly enhanced decidual promoter activity in response to 8-bromocAMP (Fig. 8A). In addition, in the presence of SRC-1e, decidual PRL gene transcription in response to 8-bromo-cAMP was now potently increased by MPA. Therefore, the expression of SRC-1e mimicked in unprimed ES cells the effects of cAMP pretreatment. However, ES cell differentiation in response to cAMP or cAMP plus MPA did not result in altered expression levels of the mRNA of either SRC isoform (Fig. 8B). Discussion PR control of decidual PRL expression during decidualization

Activation of PR by its cognate ligand does not exert direct transcriptional control of the decidual PRL gene, in keeping with the absence of consensus PREs in the promoter region (24). There is, however, compelling evidence to indicate that initiation of the decidualization response requires sustained PKA activation, although Telgmann et al. (25) demonstrated that PKA activation of the decidual PRL promoter is not


FIG. 7. Steroid receptor specificity of ligand-independent inhibition of cAMP-induced decidual PRL promoter activity. The dPRL-3000/ Luc reporter vector was transiently transfected into 8-bromo-cAMPprimed ES cells (48 h) together with expression vectors for hPR-B, hPR-A, hERa, hGR, hAR, or pSG5. Subsequently, the cultures received no further treatment (control) or 0.5 mM 8-bromo-cAMP for 40 h. Reporter gene activity was measured 40 h later, and the data represent the mean activity (6SD) of triplicate measurements.

mediated by the classical cAMP response element-binding protein pathway, but involves the induction of an as yet unidentified ES cell-specific factor(s). The role of progestins appears to lie in maintaining and enhancing the decidualization response. However, the ability of the antiprogestin ZK98.299 to partially inhibit cAMP-induced PRL secretion indicates a role for PR in the presence of activated PKA. Furthermore, the ability of the catalytic a-subunit of PKA holoenzyme to activate the decidual PRL promoter in a manner inhibitable by ZK98.299 provides additional evidence in support of a role for unliganded PR in regulating PRL gene expression. Antiprogestin inhibition of cAMP-induced PRL secretion was a delayed response, becoming apparent after a lag period of 2– 4 days. Similarly, MPA enhanced the decidual PRL promoter activity in response to PKA activation, but only if the decidualization process had been initiated through pretreatment of the cultures with cAMP for approximately 48 h. These observations imply that activation of the PKA pathway sensitizes ES cells to the actions of progestins and antiprogestins through induction or modulation of an essential intermediate factor(s) or through loss of a repressor function. The expression of decidual-specific genes, such as TF and PRL, correlated inversely with cellular PR levels and suggested, surprisingly, PR as a candidate repressor. Under culture conditions that allowed an optimal decidualization response (2% DCC-FBS), only the PR-A isoform was detected on Western blotting. Activation of the PKA pathway resulted initially in rapid up-regulation and apparent phosphorylation of PR-A followed by a progressive decline in receptor levels. The observations that PR-A is the predominant PR

4817

isoform during decidualization in vitro and that down-regulation of this receptor to new steady state levels is an integral component of this process are in agreement with in vivo studies. In human endometrium, both PR isoforms are expressed in the glandular and stromal compartments during the proliferative phase of the cycle (42– 44). After ovulation the expression of both isoforms decreases dramatically in the glandular compartment, but PR-A levels are relatively maintained throughout the cycle in stromal cells (43). However, the stromal compartment is characterized by considerable cellular heterogeneity. Colocalization studies, using double labeling for PR and PRL, have shown that there is a subpopulation of stromal cells in the decidua that retain high nuclear PR expression levels, but stain very weakly for PRL. In contrast, those cells that secrete PRL not only have the characteristic decidual morphology, but also display low immunoreactivity for PR (44). Our unpublished confocal microscopy studies show that these disparate cellular responses to a decidualizing stimulus are maintained in culture (Li, X.-F., J. Brosens, and J. White, unpublished observations). Down-regulation of PR may reflect a homeostatic mechanism to limit the cellular response to progestins or represents the progressive loss of a repressor function. To test the hypothesis that low cellular PR levels are an essential prerequisite for a full decidual response, we expressed PR in ES cells in which endogenous PR would have been down-regulated by priming with cAMP for 48 h. Transient expression of hPR-B or hPR-A markedly inhibited decidual PRL promoter-reporter activity in response to cAMP treatment independently of the presence or absence of ligand. Furthermore, PR repression of the endogenous genomic decidual PRL gene in response to cAMP was demonstrated by the markedly reduced PRL secretion of ES cells constitutively expressing GFP-hPR. Negative effects of PR on decidual gene expression are not without precedent, as Gao and Tseng (45) demonstrated that PR can also inhibit insulin-like growth factor-binding protein-1 (IGFBP-1) gene induction in uterine cells, suggesting that PR may have a general role in limiting expression of the decidual phenotype. Mechanism of PR inhibition of decidual PRL expression

It appears unlikely that PR acquires its ability to inhibit the decidual PRL gene in response to PKA-dependent phosphorylation. Phosphotryptic peptide mapping has identified at least nine phosphorylation sites in hPR-B, seven of which have been sequenced to date (36 –38). Although cAMP treatment of ES cells maintained in low serum conditions did result in apparent phosphorylation of the receptor, reflected by altered mobility upon gel electrophoresis, none of the PR phosphomutants tested in this study, which covered all of the sequenced authentic phosphorylation sites except for Ser294 and Ser400, were important for inhibition of the decidual PRL promoter. Takimoto and co-workers (34) demonstrated that these mutants have also only subtle effects on the transactivation potential of the receptor on simple or complex PR-responsive promoters. Hence, the role of PR phosphorylation sites in modulating function remains unresolved. PR inhibition of the decidual PRL promoter activity in response to elevated cAMP levels could be caused by several

4818


Endo • 1999 Vol 140 • No 10

FIG. 8. A, SRC-1e enhances decidual PRL promoter activity. cAMP-primed ES cells were transiently transfected with dPRL-3000/Luc and expression vectors for SRC-1a, SRC-1e, or the control vector pSG5. The cultures were not treated (control) or were supplemented with 8-bromo-cAMP, 8-bromo-cAMP plus MPA, or MPA alone. Reporter gene activity was measured 40 h later, and the data represent the mean activity (6SD) of triplicate measurements. B, RT-PCR analysis of SRC-1 transcripts in differentiating ES cells. Confluent ES cell cultures were maintained in medium supplemented with 2% DCC-FBS, and RNA was prepared from untreated cells and cells treated for 6 days as indicated in the figure. Simultaneous RTPCR amplifications were performed in a single reaction for SRC-1a and GAPDH or for SRC-1e and GAPDH cDNAs. The Southern blot of PCR products was successively hybridized with oligonucleotide probes common to both SRC-1 isoforms and GAPDH sequences as detailed in Materials and Methods. The sizes of the products are indicated on the right.

mechanisms. The first possibility is that PR is able to bind to a cognate DNA element in the promoter. However, the absence of consensus PREs in the promoter region tend to argue against the presence of high affinity binding sites for PR (24). A second possibility is that PR could exert its inhibition through the induction of a repressor molecule. Although PR inhibition did not require ligand, there was no evidence that the receptor could activate a model progesterone-responsive promoter in the absence of progestins. The lack of generalized transcriptional competence of the receptor in response to cAMP argues against, but does not exclude, the induction of a PR-dependent repressor during decidualization. Finally, PR could directly interfere with a cAMP-dependent transcription factor(s) that binds to the decidual PRL promoter. We demonstrated that PR repressor function was minimally dependent upon a functional DBD, but also that within the context of the wild-type receptor the LBD contributed to inhibition of decidual PRL gene. Multiple examples exist in the literature where nuclear receptors repress promoter ac-

tivity through direct physical interaction with other transcription factors, and often these protein-protein interactions are dependent upon functional DBD and LBD of the receptor. This is the case in AR repression of the gene that encodes for the a-subunit of the pituitary glycoprotein hormones LH and FSH (46), repression of activating protein-1 by GR (45), ER inhibition of interleukin-6 promoter by interacting with nuclear factor-kB and CCAAT/enhancer-binding protein-b (C/EBPb) (47), and GR repression of trans-activation by RelA(p65) subunit of nuclear factor-kB (48). PR also represses RelA activity, and the mechanism involved appears very similar to that of PR inhibition of decidual PRL promoter (49). In both cases inhibition of gene transcription was independent of the PR isoform and occurred in the absence of ligand, but could be enhanced by both progestins and antiprogestins and was dependent only on a functional DBD of PR. Finally, PR has also been shown to repress IGFBP-1 expression in uterine cells by a mechanism involving protein-protein interaction. Interestingly, analysis of PR repression of the


IGFBP-1 promoter activity in HEC-1B cells, an endometrial adenocarcinoma cell line, showed that it is ligand dependent and involves both transcription activation domains of PR (AF-1 and AF-2), but not the DBD (50). Whether PR inhibits IGFBP-1 expression in differentiating ES cells through a similar mechanism remains to be determined. The role of SRC-1 in enhancing decidual PRL promoter activity

It appears plausible that MPA modestly enhances cAMPinduced PRL secretion in the early stages of ES differentiation through accelerated down-regulation of cellular PR levels. However, the marked synergistic effect of MPA and cAMP on PRL protein expression seen in long term cultures (6 –10 days) is unlikely to be accounted for solely by declining levels of PR. We speculated that activation of the PKA pathway in ES cells may induce the expression of a specific PR coactivator(s), which, in concert with declining PR levels, may account for the switch from PR repression to activation of the decidual PRL gene. Evidence in favor of such a hypothesis is provided by the observation that disruption of the steroid hormone coactivator gene SRC-1 in mice markedly reduces the decidualization reaction in response to progesterone treatment and mechanical traumatization (51). There are two human isoforms of SRC-1, SRC-1a and SRC-1e, that differ in their C-termini. Although SRC-1a contains an additional nuclear receptor-binding motif (LXXLL motif), it is less effective than SRC-1e in enhancing ER-dependent gene transcription (40). Expression of either SRC-1 isoform in undifferentiated ES cells did not induce decidual PRL promoter activity in the presence or absence of progestins. However, in the presence of activated PKA, SRC-1e, but not SRC-1a, markedly enhanced transcriptional activity in response to MPA, mimicking the synergy seen in decidualized cells. There was no evidence of regulation of SRC1 mRNA by cAMP or MPA in ES cells, but we cannot exclude that decidualization is associated with postranscriptional regulation of SRC-1 function. The mechanism by which SRC-1e mediates progesteronedependent enhancement of the decidual PRL promoter activity is unknown. An attractive hypothesis is that during ES differentiation, PR and SRC-1e are integrated in a multiprotein complex that confers the progestin effect on decidual PRL expression through protein-protein interactions or through binding to the multiple degenerate PRE half-sites in the promoter region. Two alternative explanations for the progestin effect on PRL expression in long term cultures should be considered. First, it is possible that nongenomic progesterone effects are revealed during ES differentiation. However, the absence of a decidualization response in PR deficient mice tends to argue against this (52). Second, the cellular heterogeneity in PR expression in the decidua could indicate that paracrine interactions confer the progestin effect. In this model, the subpopulation of ES cells that fail to down-regulate PR, and hence are resistant to differentiation, would secrete a PR-dependent signal(s) to those ES cells that have acquired susceptibility to decidual transformation by lowering their cellular PR levels. ES cell differentiation in vivo

4819

is likely to be further regulated by interaction with epithelial and local immune cells (53). In conclusion, this study revealed that maintenance of elevated PR levels inhibits initiation of decidualization. This may be an important mechanism in preventing illicit activation of decidual genes and may contribute to the highly coordinated timing of the expression of the decidual phenotype in the late luteal phase of the menstrual cycle. Sustained PKA activation is required for induction of ES cell differentiation and not only results in down-regulation of cellular PR levels but also sensitizes ES cells to the actions of progesterone. If pregnancy occurs, elevated progesterone levels will maintain and enhance the decidual response. Although the mechanism by which this is achieved remains unresolved, our study suggests that steroid hormone receptor coactivators such as SRC-1 play an integral role in this process. Acknowledgments We thank Dr. Ian Mak for his assistance with the fluorescence microscopy studies. We are also indebted to Dr. Birgit Gellersen, Institute for Hormone and Fertility Research (Hamburg, Germany); Dr. Pierre Chambon, INSERM (Strasbourg, France); Dr. Kate Horwitz, University of Colorado Health Sciences Center (Denver, CO); Dr. Richard Maurer, University of Oregon (Portland, OR); Dr. Malcolm Parker, Imperial Cancer Research Fund (London, UK); and Dr. John McVeigh, Medical Research Council Clinical Sciences Center (London, UK), for their generous gifts of plasmids and other reagents. Drs. Nicholas Dibb and Stephen Franks are thanked for reviewing this manuscript.

References 1. Tabibzadeh S, Sun XZ, Kong QF, Kasnic G, Miller J, Satyaswaroop PG 1993 Induction of a polarized micro-environment by human T cells and interferon-g in three-dimensional spheroid cultures of human endometrial epithelial cells. Hum Reprod 8:182–192 2. Tabibzadeh S, Babakani A 1995 The signals and molecular pathways involved in implantation, a symbiotic interaction between blastocyst and endometrium involving adhesion and tissue invasion. Hum Reprod 10:1579 –1602 3. Lockwood CJ, Nemerson Y, Krikun G, Hausknecht V, Markiewicz L, Alvarez M, Guller S, Schatz F 1993 Steroid-modulated stromal cell tissue factor expression: a model for the regulation of endometrial hemostasis and menstruation. J Clin Endocrinol Metab 77:1014 –1019 4. Lockwood CJ, Krikun G, Hausknecht VA, Papp C, Schatz F 1998 Matrix metalloproteinase and matrix metalloproteinase inhibitor expression in endometrial stromal cells during progestin-initiated decidualization and menstruation-related progestin withdrawal. Endocrinology 139:4607– 4613 5. Brosens J, Brosens I 1998 Adenomyosis. In: Cameron I, Fraser I, Smith S (eds) Clinical Disorders of the Menstrual Cycle and Endometrium. Oxford University Press, New York, pp 297–312 6. Bulun SE, Noble LS, Takayama K, et al 1997 Endocrine disorders associated with inappropiately high aromatase expression. J Steroid Biochem Mol Biol 61:133–139 7. Daley DC, Maslar IA, Rosenberg SM, Tohan N, Riddick DH 1981 Prolactin production by luteal phase defect endometrium. Am J Obstet Gynecol 140:587–591 8. Parker MG 1993 Steroid and related receptors. Curr Opin Cell Biol 5:499 –504 9. Gronemeyer H 1993 Transcription activation by nuclear receptors. J Receptor Res 13:667– 691 10. Tsai M-J, O’Malley BW 1994 Molecular mechanisms of action of steroid/ thyroid receptor superfamily members. Annu Rev Biochem 63:451– 486 11. Pratt WB, Toft DO 1997 Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocr Rev 18:306 –360 12. Kastner P, Krust A, Turcotte B, Stropp U, Tora L, Gronemeyer H, Chambon P 1990 Two distinct estrogen-regulated promoters generate transcripts encoding the two functionally different human progesterone receptor form A and B. EMBO J 9:1603–1614 13. Tung L, Mohamed MK, Hoeffler JP, Takimoto GS, Horwitz KB 1993 Antagonist-occupied human progesterone B-receptors activate transcription without binding to progesterone response elements and are dominantly inhibited by A-receptors. Mol Endocrinol 7:1256 –1265 14. Vegeto E, Shahbaz MM, Wen DX, Goldman ME, O’Malley BW 1993 Human

4820

15. 16.

17. 18. 19. 20. 21. 22. 23.

24. 25.

26. 27. 28. 29.

30.

31.

32.

33. 34.


progesterone receptor A form is a cell- and promoter-specific repressor of human progesterone receptor B function. Mol Endocrinol 7:1244 –55 Wen DX, Xu Y-F, Mais DE, Goldman ME, McDonnel DP 1994 The A and B isoforms of the human progesterone receptor operate through distinct signalling pathways within target cells. Mol Cell Biol 14:8356 – 8364 Sartorius CA, Groshong SD, Miller LA, Powell RL, Tung L, Takimoto G, Horwitz KB 1994 A third transactivation function (AF3) of human progesterone receptors located in the unique N-terminal segment of B-isoform. Mol Endocrinol 8:1347–1360 Giangrande PH, Pollio G, McDonnell DP 1997 Mapping and characterization of the functional domains responsible for the differential activity of the A and B isoforms of the human progesterone receptor. J Biol Chem 272:32889 –32900 Horwitz KB, Jackson TA, Bain DL, Richer JK, Takimoto GS, Tung L 1996 Nuclear receptor coactivators and corepressors. Mol Endocrinol 10:1167–1177 Shibata H, Spencer TE, Onate SA, Jenster G, Tsai SY, Tsai MJ, O’Malley BW 1997 Role of co-activators and co-repressors in the mechanism of steroid/ thyroid receptor action. Recent Prog Horm Res 52:1–25 Jenster G 1998 Coactivators and corepressors as mediators of nuclear receptor function: an update. Mol Cell Endocrinol 143:1–7 Nagy L, Kao H-Y, Chakravarti D, Lin RJ, Hassig CA, Ayer DE, Schreiber SL, Evans RM 1997 Nuclear receptor repression mediated by a complex containing SMRT, mSin3A, and histone deacetalylase. Cell 89:373–380 Alland L, Muhle R, Hou H Jr, Potes J, Chin L, Schreiber-Agus N, DePhinho RA 1997 Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 387:49 –55 Brosens JJ, Takeda S, Acevedo CH, Lewis MP, Kirby PL, Symes EK, Krausz T, Purohit A, Gellersen B, White JO 1996 Human endometrial fibroblasts immortalised by simian virus 40 large T antigen differentiate in response to a decidualization stimulus. Endocrinology 137:2225–2231 Gellersen B, Kempf R, Telgmann R, DiMattia GE 1994 Nonpituitary human prolactin gene transcription is independent of Pit-1 and differentially controlled in lymphocytes and in endometrial stroma. Mol Endocrinol 8:356 –373 Telgmann R, Maronde E, Tasken K, Gellersen B 1997 Activated protein kinase A is required for differentiation-dependent transcription of the decidual prolactin gene in human endometrial stromal cells. Endocrinology 138:929 –937 Brar AK, Frank GR, Cedars MI, Handwerger S 1997 Progesterone-dependent decidualization of the human endometrium is mediated by cAMP. Endocrine 6:301–307 Tang B, Gurpide E 1993 Direct effect of gonadotropins on decidualization of human endometrial stromal cells. J Steroid Biochem Mol Biol 47:115–121 Ferrari A, Petraglia F, Gupride E 1995 Corticotropin releasing factor decidualizes human endometrial stromal cells isolated from human proliferative endometrium. Endocrinology 133:2197–2203 Moy E, Kimzey LM, Nelson LM, Blithe DL 1996 Glycoprotein hormone a-subunit functions synergistically with progesterone to stimulate differentiation of cultured human endometrial stromal cells to decidualized cells: a novel role for free a-subunit in reproduction. Endocrinology 137:1332–1339 Edwards DP, Weigel NL, Nordeen SK, Beck CA 1993 Modulators of cellular protein phosphorylation alter the trans-activation function of human progesterone receptor and the biological activity of progesterone antagonists. Breat Cancer Res Treat 27:41–56 Aronica SM, Katzenellenbogen BS 1991 Progesterone receptor regulation in uterine cells: stimulation by estrogen, cyclic adenosine 39, 59-monophosphate, and insulin-like growth factor 1 and suppression by antiestrogens and protein kinase inhibitors. Endocrinology 128:2045–2052 Clemens JW, Robker RL, Kraus L, Katzenellenbogen BS, Richards JS 1998 Hormone induction of progesterone receptor (PR) messenger ribonucleic acid and activation of PR promoter regions in ovarian granulosa cells: evidence for a role of cyclic adenosine 39,59-monophosphate but not estradiol. Mol Endocrinol 12:1201–1214 Beck CA, Weigel NL, Edwards DP 1992 Effect of hormone and cellular modulators of protein phosphorylation on transcriptional activity, DNA binding, and phosphorylation of human progesterone receptors. Mol Endocrinol 6:607– 620 Takimoto GS, Hovland AR, Tasset DM, Melville MY, Tung L, Horwitz KB 1996 Role of phosphorylation on DNA binding and transcriptional functions of human progesterone receptors. J Biol Chem 271:13308 –13316

Endo • 1999 Vol 140 • No 10

35. Kazmi SM, Visconti V, Plante RK, Ishaque A, Lau C 1993 Differential regulation of human progesterone receptor A and B form-mediated trans-activation by phosphorylation. Endocrinology 133:1230 –1238 36. Zhang Y, Beck CA, Poletti A, Edwards DP, Weigel NL 1994 Identification of phosphorylation sites unique to the B form of human progesterone receptor: in vitro phosphorylation by casein kinase II. J Biol Chem 269:31034 –31040 37. Zhang Y, Beck CA, Poletti A, Edwards DP, Weigel NL 1995 Identification of a group of Ser-Pro motif hormone-inducible phosphorylation sites in the human progesterone receptor. Mol Endocrinol 9:1029 –1040 38. Zhang Y, Beck CA, Poletti A, Clement IV JP, Prendergast P, Yip T, Hutchens TW, Edwards DP, Weigel NL 1997 Phosphorylation of human progesterone receptor by cyclin-dependent kinase 2 on three sites that are authentic basal phosphorylation sites in vivo. Mol Endocrinol 11:823– 832 39. On˜ate SA, Tsai SY, Tsai M-J, O’Malley BW 1995 Sequence and characterization of a coactivator for steroid hormone receptor superfamily. Science 270:1354 –1357 40. Kalkhoven E, Valentine JE, Heery DM, Parker MG 1998 Isoforms of steroid receptor co-activator 1 differ in their ability to potentiate transcription by the oestrogen receptor. EMBO J 17:232–243 41. Onate SA, Boonyaratanakornkit V, Spence TE, Tsai SY, Tsai M-J, Edwards DP, O’Malley BW 1998 The steroid receptor coactivator-1 contains multiple receptor interacting and activation domains that cooperatively enhance the activation function 1 (AF1) and AF2 domains of steroid receptors. J Biol Chem 273:12101–12108 42. Koshiyama M, Konishi I, Nanbu K, Nanbu Y, Mandai M, Komatsu T, Yamamoto S, Mori T, Fuji S 1995 Immunohistochemical localization of heat shock proteins hsp70 and hsp90 in the human endometrium: correlation with sex steroid receptors and ki-67 antigen expression. J Clin Endocrinol Metab 80:1106 –1112 43. Wang H, Critchley HO, Kelly RW, Shen D, Baird DT 1998 Progesterone receptor subtype B is differentially regulated in human endometrial stroma. Mol Hum Reprod 4:407– 412 44. Wang J-D, Zhu J-B, Shi W-L, Zhu P-D 1994 Immunocytochemical colocalization of progesterone receptor and prolactin in individual stromal cells of human decidua. J Clin Endocrinol Metab 79:293–297 45. Jonat C, Rahmsdorf HJ, Park KK, Cato AC, Gebel S, Ponta H, Herrlich P 1990 Antitumor promotion and antiinflammation: down-modulation of AP-1 (Fos/ Jun) activity by glucocorticoid hormone. Cell 62:1189 –1204 46. Heckert LL, Wilson EM, Nilson JH 1997 Transcriptional repression of the a-subunit gene by androgen receptor occurs independently of DNA binding but requires the DNA-binding and ligand-binding domains of the receptor. Mol Endocrinol 11:1497–1506 47. Stein B, Yang MX 1995 Repression of the interleukin-6 promoter by estrogen receptor is mediated by NK-kB and C/EBPb. Mol Cel Biol 15:4971– 4979 48. Wissink S, van Heerde EC, Schmitz ML, Kalkhoven E, van der Burg B, Baeuerle PA, van der Saag PT 1997 Distinct domains of the RelA NK-kB subunit are required for negative cross-talk and direct interaction with the glucocorticoid receptor. J Biol Chem 272:22278 –22284 49. Kalkhoven E, Wissink S, van der Saag PT, van der Burg B 1996 Negative interaction between RelA(p65) subunit of NK-kB and progesterone receptor. J Biol Chem 271:6217– 6224 50. Gao J, Tseng L 1997 Progesterone receptor (PR) inhibits expression of insulinlike growth factor-binding protein-1 (IGFBP-1) in human endometrial cell line HEC-1B: characterization of the inhibitory effect of PR on the distal promoter region of the IGFBP-1 gene. Mol Endocrinol 11:973–979 51. Xu J, Qui Y, DeMayo FJ, Tsai SY, Tsai M-J, O’Malley BW 1998 Partial hormone resistance in mice with disruption of the steroid receptor coactivator-1 (SRC-1) gene. Science 279:1922–1925 52. Ormandy CJ. Camus A, Barra J, Damotte D, Lucas BK, Buteau H, Edery M, Brousse N, Babinet C, Binart N, Kelly PA 1997 Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse. Genes Dev 11:167–178 53. Jikihara H, Handwerger, S 1994 Tumor necrosis factor-a inhibits the synthesis and release of human decidual prolactin. Endocrinology 134:353–357