Estrogen action in target cells: selective requirements for activation of different hormone response elements

ELSEVIER

Molecular and Cellular Endocrinology 112 (1995) 35-43

Estrogen action in target cells: selective requirements for activation of different hormone response elements Salman M. Hyder*aTb,Gregory L. Shipley

a,George M. Stancela

aLkpaltment of Pharmacology,Unioersiry of TexnsMedicalSchooJ P.O. Box 20708, Houston, TX 77225, USA bDepartmentof Obstetrics,Gynecologyand ReproductiwSciences, Uniwrsityof TexasHealth Sciences Center,Houston, TX 77225, USA Received 30 March 1995; revision received 1995; accepted 12 May 1995

Abstract RENEl cells, an estrogen receptor positive rat uterine endometrial cell line immortalized with the ElA oncogene, were analyzed for the presence of estrogen-dependent signal transduction pathways using the induction of transfected as well as endogenous genes. RENEl cells express the estrogen receptor as analyzed by Northern blots and ligand binding assays (40 fmoles/mg protein). The receptor system appears functional, based on the induction of reporter constructs containing the consensus estrogen response element (ERE) in transient transfection assays and alterations in endogenous transcripts visualized by utilizing differential display methodology. However, neither transfected reporter constructs containing the c-for ERE, nor the endogenous c-fos, c-&n, or c-rnyc genes are induced by estrogens in these cells despite being induced by estrogens in the uterus in vivo. In addition, estradiol did not induce endogenous c-fos expression or the activity of CAT reporters containing the c-fos ERE in a stable transfectant of RRNEl cells with a 3-fold elevation in estrogen receptor content. Under identical conditions, TPA and serum rapidly induce c-fos transcription in RENEl cells, indicating that the lack of inducibility by estradiol is not due to a general inhibitor of transcription of these genes. These results suggest that RBNBl

cells lack factors present in normal uterine cells which are required for the estrogenic induction of a specific subset(s) of EREs. These observations support the generally evolving hypothesis that steroid hormones may act through composite response elements via interactions palindromic response elements.

with other transcription

factors, in addition to functioning

as homodimers

at classical

Keywords: Rat uterine cell; Estrogen receptor; Estrogen response element; Estrogen; c-fos; TF’A; Serum

1. Introduction

Estrogens regulate cell proliferation and a variety of differentiated functions in target tissues such as the uterus, vagina, pituitary, and mammary epithelium. Such effects are thought to be due in large part to hormone-induced alterations in gene transcription, and a number of genes rapidly induced by estrogens have been identified in normal reproductive tissues

*Corresponding 7925911.

author, Tel: + 1 713 7925967; Fax: + 1 713

and in tumors derived from such tissues (Landers and Spelsberg, 1992; Stance1 et al., 1993; Weisz and Bresciani, 1993). Recently, estrogen response elements (EREs) have been identified in hormone responsive genes such as c-fos (Weisz and Rosales, 1990, Hyder et al., 1992) and c-jun (Hyder et al., 19951, which bind the estrogen receptor (ER) and are well suited to amplify tissue responses emanating from the initial ER-ERE interactions. However, the exact relationship between the induction of such genes and overall tissue responses such as proliferation is not clear at present. This relationship between hormone induced changes in the

0303-7207/95/$09.50 Q 1995 Elsevier Science Ireland Ltd. All rights reserved. SSDI 0303-7207(95)03581-Q

36

S.M. Hyder et al. /Molecular and Cellular Endocrinology 112 (1995) 35-43

expression of individual genes such as AP-1 compo nents and overall tissue responses is further complicated because such responses may be highly cell specific (Nephew et al., 1994), and because estrogen responsive genes such as c-fos and c-&n are subject to potential regulation by numerous peptide growth factors and other regulatory agents in vivo (Angel and Karin, 1991). A major drawback to defining such relationships in reproductive tissues such as the uterus has been the lack of suitable non-transformed cell culture systems in which hormonal responses faithfully mimic those produced in vivo. The recent establishment of immortalized uterine cell lines such as RENEl (Wiehle et al., 1990), which express the ER, provide a potential system to investigate the relationship between the expression of specific estrogen inducible genes and responses such as cell proliferation in non-transformed cells. As an initial approach to this question, we have analyzed this cell system for the expression of estrogen, TPA and serum-stimulated signal transduction pathways. We selected the expression of endogenous c-fos, c_iun and c-myc in these cells as the endpoint for target gene expression, since: (1) these genes are rapidly induced by estradiol in vivo (Murphy et al., 1987; Loose-Mitchell et al., 1988; Weisz and Bresciani, 1988; Nephew et al., 1994); (2) all the available evidence indicates that these are primary transcriptional responses to the hormone (LooseMitchell et al., 1988; Weisz and Bresciani, 1988; Chiappetta et al., 1992; Cicatiello et al., 1992, 1993); and (3) functional EREs for these genes have recently been defined (Weisz and Rosales, 1990; Hyder et al., 1992, 1995). Estrogen induction of c-fos expression is one of the most rapid and dramatic effects of estrogens observed in the uterine epithelial cells of intact animals (Papa et al., 1991). Surprisingly, however, estradiol does not induce c-fos expression in RENEl uterine epithelial cells, even though expression of this protooncogene can be induced by serum and TPA. Furthermore, estrogen receptors can regulate the expression of reporters containing the consensus ERE as well as endogenous RENEl genes. This suggests that individual EREs require different factors for estrogenic regulation and RENEl cells contain some but not all of such factors present in uterine cells in vivo.

17p and hygromycin were obtained from Sigma Chemical Company (St. Louis, MO). All radioactive isotopes for the synthesis of labeled probes were obtained from Amersham (Arlington Heights, IN). 2.2. Cell culture and treatment with reagents for expression of immediate earZygrowth responsive genes RENEl cells were kindly provided by Dr. Ronald Wiehle, Baylor College of Medicine, grown in DME/F12 supplemented with 5% (v/v> FCS, and were routinely passaged once a week. All experiments with RENEl cells were done in passages 5-25, and passages 5-20 for the selected clone RENEl 14 following transfection with the ER. For experiments involving estradiol treatment, cells were kept in DME/F12 supplemented with BSA and transferrin (Hyder et al., 1991) for 48 h with a media change after 24 h. Cells were then incubated with 20 nM estradiol for various times. Experiments were also repeated with media containing 5% (v/v) dextran-coated charcoal treated serum instead of BSA and transferrin in order to evaluate the requirement of any serum component to mediate estrogen effects. For experiments with serum or TPA activation of gene expression, cells were maintained overnight in 0.5% (v/v) serum and then treated with either 200 nM TPA or 20% (v/v) serum for the indicated times. RNA was then prepared from the cells and processed as described elsewhere (Loose-Mitchell et al., 1988). 2.3. RNA-blot analysis Synthesis of the 32P-labeled antisense c-fos, c-jun and c-myc riboprobes have been described previously (Kirkland et al., 1992). All probes utilized 32P-UTP and either Sp6 or ‘I7 polymerase for their synthesis. The probe for the estrogen receptor has previously been described (Scrocchi and Jones, 1991). The procedure for RNA blot analysis of mRNA was performed as previously described (Loose-Mitchell et al., 1988). All RNA-blot analyses were performed in duplicate.

2. Materials and methods

2.4. Transient transfections Cultures of RENEl cells were transfected with 10 pg of recombinant plasmid DNA by the calcium phosphate procedure (Hyder et al., 1992). Cells (approximately 50% confluent) in lo-cm plates were transfected, and CAT assays were performed as described previously (Hyder et al., 1992). All transfections were performed at least twice.

2.1. Materials Phenol red free DME/F12 was obtained from Gibco laboratories (Grand Island, NY) and used in all the cell culture studies. Fetal calf serum (FCS) was obtained from Hazelton Biologicals (Lanexa, KS). Bovine serum albumin, human transferrin, estradiol-

2.5. Stable transfections RENEl cells were transfected with the estrogen receptor constructs together with the construct expressing the gene for hygromycin resistance. The ER expression vector used in the generation of stable clones has been described previously (Bradshaw et al.,

S.M. Hyder et al. /Molecular am’ Cellular Endocrinology 112 (1995) 35-43

1988). Following transfections, cells were kept overnight in media containing 5% (v/v) FCS for recovery. The FCS was then replaced with 5% (v/v> dextran-coated charcoal treated serum to remove endogenous steroids that may be toxic to cells overexpressing the ER (Kushner et al., 1990). Twenty-four hours later, selection was started by continuously exposing cells to 200 pg/ml of hygromycin for 2 weeks with a change of media twice a week. Plates were then incubated until individual clones were easily identified. These were individually trypsinized with the use of cloning rings and placed in six well plates for further growth. After the first passage from the six-well plates, the hygromycin concentration was reduced to 50 pg/ml for maintenance of the cells and routine passaging. 2.6. Differential disphy Differential display RT-PCR was done by published procedures (Liang and Pardee, 1992; Liang et al., 1993). Briefly, 200 ng of total RNA was reverse transcribed in a 20 ~1 reaction volume using reverse transcriptase and buffers from Promega (Madison, WI) and (T)i2 VA as primer (V = A, C or G) in a MJR PTC 100 thermocycler (MJR Research, Inc., Watertown, MA) as follows: 35°C for 60 s and 99°C for 5 min. Reactants were removed from the cDNA using Microcon 30 cartridges (Amicon, Beverley, MA) following the directions of the manufacturer and recovered in a final volume of 20 ~1 water. Polymerase chain reactions utilized components from a Perkin Elmer PCR kit (Perkin Elmer/ABI, Foster city, CA). For isotopic labeling, 0.4 ~1 (I!33P-dATP (NEN, 3000 Ci/mmole) was added along with 2 ~1 of cDNA as template, the above 3’-primer (2.5 PM) and 5’-primer #l-TTGGGCTGGA (0.5 PM) (Genosys, Houston, TX) in a 20 ~1 reaction volume using the following program cycle: 94°C for 30 s; 40°C for 60 s, 72°C for 30 s for 30 cycles followed by 72°C for 5 min. The final products were recovered by precipitation and electrophoresed on a 7% Long Ranger (AT Biochem, Malvern, PA) denaturing gel at 60 W for 4 h. The gel was dried and exposed to X-Omat film (Eastman Kodak, Rochester, NY) for 18 h at room temperature. 3. Results We initiated our studies by first investigating the effects of estrogen on the regulation of c-f& transcript levels in the RENEl cells using culture conditions under which the hormone induces expression of this oncogene in other hormone responsive cells (Hyder et al., 1991). As shown in Fig. lA, estradiol does not increase c-fos mRNA levels in RENEl cells over a period of 120 min. This experiment used RENEl cells at passage number 5, and other experiments at

37

later passages also showed that c-fos could not be induced by estradiol (data not shown). In contrast to these findings, both the phorbol ester TPA (Figs. 1A and 1B) and fetal calf serum (Fig. 10 stimulate c-fos expression under exactly the same conditions. As in most other experimental systems, both ‘IPA (Fig. 1B) and fetal calf serum (Fig. 10 induce a transient increase in c-fos mRNA levels in RENEl cells. These results demonstrate that the signal transduction pathway for estrogen-mediated induction of c-fos expression is impaired in the RENEl cells, while the pathways mediating the TPA and serum effects on expression of this protooncogene remain intact. The experiments in Fig. 1 were performed with serum-free media (see ‘Materials and methods’). We also tested the ability of estradiol to induce c-fos expression in RENEl cells using media supplemented with 5% (v/v> fetal calf serum that had been stripped of endogenous steroid hormones by charcoal treatment. This approach was attempted in the event that c-fos induction by estrogens might require the presence of additional serum borne factors, but this approach also failed to demonstrate hormonal induction of protooncogene expression (data not shown). Thus our findings are similar to the observations of Liu and Teng (1994) who could not demonstrate c-fo.s induction by estradiol in RG925 human endometrial tumor cells. We next tested if the failure of estradiol to induce c-fos in the RENEl cells was due to the absence of functional ERs or other factors required for ER activity. To assess this possibility, we tested the ability of the endogenous RENEl estrogen receptors to induce hormone-dependent expression from CAT reporter constructs containing either the consensus ERE (Burch et al., 1988) or the murine c-fos ERE we have recently identified (Hyder et al., 1992). The data in Fig. 2 demonstrates that RENEl cells contain functional ERs since estrogen-dependent CAT activity is induced in these cells from the reporter constructs containing the vitellogenin ERE. However, estradiol does not induce CAT expression from plasmids containing the c-fos ERE in these cells. One possible explanation of these results is that the RENEl cells do not contain sufficient ER levels to activate the c-fos ERE. For example, the magnitude of estrogenic induction from both the lactoferrin ERE (Liu and Teng, 1994) and Cathepsin D ERE (Cavailles et al., 1991) is dependent upon cellular levels of ER. We therefore tested if increasing the number of estrogen receptors in RENEl cells would yield a measurable induction of the c-for ERE in this system. In order to perform these studies, we isolated a RENEl clone expressing elevated levels of receptor. For this purpose, we stably transfected an ER expression vector into RENEl cells as described in ‘Materi-

38

S.M. Hyder et al. /Molecular and CellukzrEndocrinology 112 (1995) 35-43

-

-

2.2KB

2.2KB

TlME(min)

cz,,

_-

(jo

go

TlllE(min) Fig. 1. Expression of c-fos mRNA in RJSNEl rat uterine endometrial cell line following stimulation with estradiol, TPA and FCS. (A) Cells were grown as descriid in ‘Materials and methods’ and then treated with estradiol (20 nM) for the indicated period of time or with 200 nM TPA for 30 min. RNA preparation and RNA-blot analysis was performed as described in ‘Materials and methods’. c-fos mRNA is indicated at 2.2 KB. (I?>Cells were maintained in phenol red free DME/F12 as in Panel (A) and then treated with 200 nM TPA for the indicated period of times. Preparation of RNA and RNA-blot analysis was as described in ‘Methods’. (0 Cells were maintained overnight in phenol red free DME/FlZ supplemented with 0.5% (v/v) FCS and then treated with 20% (v/v) FCS for the indicated times prior to analysis as described in ‘Materials and methods’.

SM. Hyder et al. /Molecular

39

and Cellular Endocrinology 112 (1995) 35-43

als and methods’. The cells were co-transfected with a selection plasmid expressing the gene for hygromycin resistance. Several hygromycin-resistant clones were isolated. The clone expressing the highest estrogen receptor level, designated as RENE A 14, was selected for further experiments. The RENEl Al4 clone expressed very high levels of the transfected ER mRNA (Fig. 3) and showed an approximate 3-fold increase in functional receptor levels as judged by ligand binding (120 fmoles/mg in RENEl Al4 cells vs 40 fmoles/mg protein in the parent RENEl cells). As seen in Fig. 3, the expression of the endogenous ER gene is also markedly increased in the RENEl Al4 cells, and neither the expression of the endogenous or transfected receptor genes is estrogen dependent. Thus, while we cannot conclude if the increased level of ER determined by ligand binding results from the expression of endogenous or transfected genes, it is nevertheless clear that RENEl Al4 cells express 3-fold higher levels of ER than the parent cells. Treatment of the RENEl Al4 cells with estrogen did not lead to an increase in c-fos mRNA levels as judged by Northern analysis (data not shown). Similarly, c-myc and c-&n transcript levels, which are induced in the uterus by estradiol (Murphy et al., 1987; Kirkland et al., 1992; Nephew et al., 1994), were not increased by the hormone in these cells. However, other experiments similar to those in Fig. 1 showed that c-fos is dramatically induced by the tumor promoter TPA in the RENEl Al4 cells (data not shown). We also examined the effect of the elevated receptor content in RENEl Al4 cells on the induction of CAT reporters containing either the consensus ERE or the c-fos ERE (Fig. 4), and compared the inducibility to that seen in the parental cells. Estrogen induces CAT activity driven by the vitellogenin ERE to the same degree in both the RENEl and RENEl Al4 cells indicating that the level of receptor in the parental cells is sufficient to produce a maximum response from the consensus element. However, estradiol failed to induce expression driven by the c-fos ERE in the RENEl Al4 cells despite the increased level of estrogen receptors in this clone. The experiments to this point established that estradiol induces expression of a transfected reporter containing the vitellogenin ERE, but does not increase the levels of the endogenous c-fos, c-jun, or c-myc mRNAs, which are dramatically induced by estrogen in vivo (Stance1 et al., 1993; Weisz and Bresciani, 1993; Hyder et al., 1994). This raised the question of whether estradiol was able to regulate the level of any endogenous transcripts in RENE 1 cells. To address this question, we isolated RNA from cells treated with 20 nM estradiol for short times and

-

+

-

FCNco

+

vit-ERE

RENEl Fig. 2. Selective induction of the vitellogenin ERE in RENJ3 cells. RENEl cells were transfected with plasmids containing either the vit-ERJZ (GGTCAnnnTGACC) or the fos-EFtE (GGTCAnnCAGCC,FCNco). After transfection, cells were plated in phenol red free DME/F12 containing 5% (v/v) DCC-treated FCS for 18 h. Cells were then treated with 20 nM estradiol (+) or vehicle alone ( -) for 24 h and cellular extracts prepared for CAT assay as described in ‘Materials and Methods’.

utilized RT-PCR-based differential display (Liang et al., 1992) to search for transcripts that were rapidly altered in response to hormone treatment in the RENEl cells. The results of this study are shown in Fig. 5. and it is clearly seen that the hormone rapidly affects the level of two transcripts in the RENEl cells. A band corresponding to the single arrowhead in Fig. 5 is markedly decreased within 30 min after exposure to estradiol, and another band is increased within 120 min (double arrowhead in Fig. 5). Since our analysis

40

S.M. Hyder et al. /Molecular and Cellular Endocrinology I12 (1995) 35-43

MCF-7

5

E 8

30

60 120

2 E 8

30

60 120

Duration of Ep Treatment (min) Fig. 3. RNA-blot analysis of estrogen receptor mRNA in RENEl, RENEl Al4 and MCF-7 cells. RENEl Al4 cells were selected from RENEl cells transfected with an ER expression vector as described in ‘Materials and Methods’ and contain 3-fold higher ER levels as measured by ligand binding. Cells were treated with 20 nM estradiol for the indicated times prior to RNA preparation and blot analysis. The blank arrow indicates the expression of ER mRNA in cells. Signal for the expression of the transfected ER mRNA is present just below the 1% marker. RNA from the MCF-7 cells was used as a positive control for hybridization signal of ER.

was done with only one of the approximately 80 primer combinations required to analyze the entire mRNA population, this result strongly implies that estradiol also regulates the expression of additional endogenous genes not observed in this experiment. 4. Discussion RENEl cells were isolated as a clone of rat uterine endometrial cells immortalized with the ElA oncogene that retains the expression of estrogen receptors. Our initial results confirmed that these cells contain estrogen receptors as indicated by the presence of receptor transcripts in Northern blots and by radiolabelled ligand binding assays. Thus, our initial intent was to use these cells to analyze the induction of genes activated by estrogens in the uterus of intact animals. In contrast to our expectations, we found that estradiol does not induce the expression of the endogenous AP-1 components c-fos and c-&n, or the nuclear protooncogene c-myc in this uterine cell line. This was surprising since estradiol induces massive increases in the expression of these genes in the uterus of intact animals (Weisz and Bresciani, 1993; Hyder et al., 1994). For example, the level of c-fos mRNA increases 20-50 fold in the rodent uterus within several hours of hormone treatment (Loose-

Mitchell et al., 1988). This in vivo induction appears to be a direct transcriptional activation (LooseMitchell et al., 1988; Weisz and Bresciani, 1988), and functional EREs have been defined in both c-fos (Weisz and Rosales, 1990; Hyder et al., 1992) and c-&n (Hyder et al., 1995). This lack of protooncogene inducibility does not appear to be due to a general defect in the estrogen receptor system in RENEl cells. The parental RENEl cells contain appreciable amounts of estrogen receptor (40 fmoles/mg protein), as judged by ligand binding, and the endogenous receptor is effective at inducing expression of CAT reporters containing the vitellogenin ERE. Furthermore, increasing the level of ER 3-fold by transfecting ER expression plasmid (i.e. the RENEl Al4 cells), did not enable estradiol to activate the c-fos ERE. Other experiments using RT-PCR-mediated differential display analysis indicated that estradiol rapidly regulates the levels of at least two RENEl cell transcripts. Parenthetically, it should be noted that this result was obtained with only one of 80 potential primer sets. Thus, it is highly likely that estradiol also controls expression of additional RENEl transcripts besides those observed in Fig. 5. While we have not established unequivocally that the changes seen in this experiment are due to direct transcriptional regulation, the results imply either that this is the case or

S.M. Hyder et al. /Molecular

41

and Celhlar Endocrirwlogy 112 (1995) 35-43

-123

-76

-

+

FCNco

-

+

-67

vit-ERE

RENElA14 Fig. 4. Induction of CAT activity from reporter plasmids containing either the vit-ERE (GGTCAnnnTGACC) or the fos-ERE (GGTCAnnCAGCC). RENEl and RENRl Al4 cells were transfected with the indicated constructs and exposed to 20 nM estradiol ( + ) or vehicle alone ( -) for 24 h. CAT assay was then performed as described in ‘Materials and methods’.

that the hormone receptor complex directly activates expression of a rapidly translated factor which secondarily produces the changes seen in Fig. 5. In either case, the results strongly suggest that estradiol treatment regulates the expression of one or more endogenous RENEl genes. Thus, RENEl cells contain functional estrogen receptors as evidenced by: (1) ligand binding; (2) induction of transfected reporters containing a consensus ERE; and (3) induction of endogenous gene transcripts in RENEl cells. Another possibility we considered to explain the lack of inducibility of c-fos and other protooncogenes was that RENEl cells produce a factor that inhibits

C 30 60120 Duration of E2 Treatment

(min)

Fig. 5. Differential display with total RNA prepared from RENEl cells treated with estradiol. The procedure and primer combination used for differential display is described in ‘Materials and methods’. Each sample was analyzed in duplicate. The single and double arrowheads point to the amplified portion of cDNAs which are rapidly regulated by estradiol in RENEl cells.

their induction. To test this possibility, we stimulated RENEl cells with the phorbol ester TPA or serum. Both agents produced rapid induction of c-fos, indicating that a general inhibitor is not present, and that the basal promoters of the various protooncogenes are functional in these cells. The observed estrogenic induction of the vitellogenin-CAT reporter and endogenous genes by differential display also indicates that the cells do not contain a factor which is a generalized inhibitor of estrogen receptor function.

42

SM. Hyder et al. /Molecular and Cellular Endocrinology 112 (1995) 35-43

Our results thus appear to suggest one of two possibilities: (1) the ElA protein, or a RENEl protein stimulated by ElA immortalization, selectively inhibits hormonal induction from a subset of EREs (e.g. those for c-fos, c-ju~, and c-myc), but not from the EREs of other endogenous RENEl genes or from the consensus ERE; or (2) the estrogen receptor alone cannot induce expression from a subset of elements such as the c-fos and c-&n EREs without another regulatory factor(s) that is absent from RENEl cells. The first possibility seems unlikely since ElA is commonly used to immortalize cell lines and we are unaware of any reports that this process has ever led to the production of a protein(s) that selectively interferes with the activity of endogenous response elements for any steroid hormone. It seems more likely that RENEl cells lack a factor that cooperates with the estrogen receptor to induce genes such as c-fos and c-j~n. One possibility is that RENEl cells lack a factor that selectively ‘activates’ the ER to regulate transcription from a peculiar subset of EREs; this factor itself could be under estrogen control. A second possibility is that another regulatory factor forms a heterodimer with the ER to selectively activate genes such as c-fos and c-jun. This possibility is also consistent with previous structural studies on the c-fos (Hyder et al., 1992) and c-jun (Hyder et al., 1995) EREs. Each of these elements contains two half-sites, one of which is identical to the half site of the consensus ERE, but the second which differs in three of five positions. This raises the possibility that the receptor binds to these elements as a heterodimer, in contrast to a receptor homodimer which binds to the consensus palindromic element (Kumar and Chambon, 1988; Schwabe et al., 1993). In the case of the c-&n ERE, this hypothesis is further supported by the finding that base changes in the ‘non-consensus’ half-site of the element abolish transcriptional activation without decreasing receptor binding (Hyder et al., 1995). Other studies with different cell lines also support the notion of receptor heterodimerization as an explanation for the selective induction of a subset of estrogen responsive genes (Kaneko et al., 1993). Clearly, further studies are required to prove or disprove this suggestion, but this interpretation seems most consistent with the findings in this report and previous studies of estrogenic induction of these AP-1 components (Hyder et al., 1992, 1995). In a broader sense, our findings with the estrogen receptor also support the evolving picture that steroid receptors may not act exclusively as homodimers via interactions with palindromic response elements. For example: (1) studies with the glucocorticoid receptor originally led to the suggestion that composite response elements might exert either positive or nega-

tive effects on transcription (Miner and Yamamoto, 1991); (2) some members of the steroid/thyroid receptor superfamily (e.g. thyroid and retinoid receptors) clearly function as heterodimers via interactions with different types of response elements (Kliewer et al., 1992); and (3) steroid receptors in some cases are thought to regulate transcription via protein-protein interactions with other factors rather than by direct DNA binding (Savouret et al., 19941. Our findings may thus have general significance for understanding estrogen action in a variety of target cells besides the RENEl uterine epithelial cells we have used in this study. In summary, we have studied estrogen-induced transcriptional responses in RENEl uterine epithelial cells. These cells contain receptors and factors sufficient for hormonal inducibility of the consensus ERE and some endogenous genes, but not for the induction of other endogenous genes or transfected reporters. RENEl cells thus provide a useful model system for some uterine responses to estrogens, but they do not mimic the full pattern of uterine responses to hormone administration seen in vivo. These cells may also prove useful for the identification of estrogen receptor-associated factors required for selective transcriptional regulation of individual EREs. Acknowledgements

We would like to thank Ms. Lata Murthy and Constance Chiappetta for excellent technical help during some aspects of this study. We would also like to thank Dr. Ming Tsai (Baylor College of Medicine) for providing the estrogen receptor expression vector. This research was supported by grant HD-08615 from the NIH. References Angel, P. and Karin, M. (1991) Biochim. Biophys. Acta 1072, 1299157. Bradshaw, M.S., Tsai, M.J. and O’Malley, B.W. (1988) J. Biol. Chem. 263, 8485-8490. Burch, J.B.E., Evans, M.I., Friedman, T.M. and O’Malley, B.J. (1988) Mol. Biol. Rep. 8, 1123-1131. Cavailles, V., Augereau, P. and Rochefort, H. (1991) Biochem. Biophys. Res. Commun. 174, 816-824. Chiappetta, C., Kirkland, J.L., Loose-Mitchell, D.S., Murthy, L. and Stancel, GM. (1992) J. Steroid Biochem. Mol. Biol. 41,113-123. Cicatiello, L., Ambrosino, C., Coletta, B., Scalona, M., Sica, V., Bresciani, F. and Weisz, A. (1992) J. Steroid. Biochem. Mol. Biol. 41, 523-528. Cicatiello, L., Sica, V., Bresciani, F. and Weisz, A. (1993) Receptor 3, 17-30. Hyder, S.M., Stancel, G.M. and Loose-Mitchell, D.S. (1991) Steroids 56,498-X)4. Hyder, S.M., Stancel, G.M., Nawaz, Z., McDonnell, D.P. and Loose-Mitchell, D.S. (1992) J. Biol. Chem. 267, 18047-18054. Hyder, S.M., Stance], GM. and Loose-Mitchell, D.S. (1994) Crit. Rev. Euk. Gene Expr. 4, 55-116.

SM. I-Q&r et al. /iUolecular and

Cellular Endocrinology

Hyder, SM., Nawaz, Z., Chiappetta, C., Yokoyama, K. and Stancel, G. M. (1995) J. Biol. Chem. 270,8506-8513. Kaneko, K.J., Glinas, C. and Gorski, J. (1993) Biochemistry 32, 8348-8359. Kirkland, J.L., Mm-thy, L. and Stancel, GM. (1992) Endocrinology 130, 3223-3230. Kliewer, S.A., Umesono, K., Mangelsdorf, D.J. and Evans, R.M. (1992) Nature 355,446-449. Kumar, V. and Chambon, P. (1988) Cell 55,145-156. Kushner, P.J., Hort, E., Shine, J., Bazter, J.D. and Greene, G. (1990) Mol. Endocrinol. 4, 1465-1473. Landers, J.P. and Spelsberg, T.C. (1992) Crit. Rev. Euk. Gene. Expr. 2, 19-63. Liang, P. and Pardee, AB. (1992) Science 257,967-971. Liang, P., Averboukh, L. and Pardee, A.B. (1993) Nut. Acids Res. 21, 3269-3275. Liu, Y. and Teng, C.T. (1994) Mol. Cell. Endocrinol. 101, 167-171. Loose-Mitchell, D.S., Chiappetta, C. and Stancel, G.M. (1988) Mol. Endocrinol. 2, 946-951. Miner, J.N. and Yamamoto, K.R. (1991) Trends B&hem. Sci. 16, 423-427.

112

(1995)

35-43

43

Murphy, L.J., Murphy, L.C. and Friesen, H.G. (1987) Endocrinology 120,1882-1888. Nephew, K.P., Tang, M. and Khan, S.A. (1994) Endocrinology 134, 1827-1834. Papa, M., Mezzogiomo, V., Bresciani, F. and Weisz, A. (1991) Biochem. Biophys. Res. Commun. 175,480-485. Savouret, J-F., Rauch, M., Redeuilh, G., Sar, S., Chauchereau, A., Woodruff, K, Parker, M. and Milgrom, E. (1994) J. Biol. Chem. 269,28955-28962. Schwabe, J.W.R., Chapman, L., Finch, J.T. and Rhodes, D. (1993) Cell 75,567-578. Scrocchi, L.A. and Jones, L.A. (1991) Endocrinology 129, 2251-2253. Stancel, G.M., Baker, V.V., Hyder, SM., Kirkland, J.L. and LooseMitchell, D.S. (1993) Gzford Rev. Reprod. Biol. 15, l-42. Weisz, A. and Bresciani, F. (1988) Mol. Endocrinol. 2, 816-824. Weisz, A. and Bresciani, F. (1993) Crit. Rev. Oncogenesis 4, 361-388. Weisz, A. and Rosales, R. (1990) Nucl. Acids Res. 18, 5097-5106. Wiehle, R.D., Helftenbien, G., Land, H., Neumann, K and Beato, M. (1990) Oncogene 5,787-794.