Opposing Action of Estrogen Receptors and on Cyclin D1 Gene ...

THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2002 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 277, No. 27, Issue of July 5, pp. 24353–24360, 2002 Printed in U.S.A.

Opposing Action of Estrogen Receptors ␣ and ␤ on Cyclin D1 Gene Expression* Received for publication, February 23, 2002, and in revised form, April 30, 2002 Published, JBC Papers in Press, May 1, 2002, DOI 10.1074/jbc.M201829200

Meng-Min Liu‡, Chris Albanese§, Carol M. Anderson¶, Kristin Hilty¶, Paul Webb¶, Rosalie M. Uht‡¶储, Richard H. Price, Jr.‡, Richard G. Pestell§, and Peter J. Kushner‡¶** From the ‡Department of Medicine, University of California, San Francisco, California 94112-1640, the §Albert Einstein Cancer Center, Department of Developmental and Molecular Biology, Albert Einstein College of Medicine, Bronx, New York 10461, and the ¶Metabolic Research Unit, University of California, San Francisco, California 94143

Estrogen stimulates proliferation of epithelial cells in the female reproductive tract and mammary gland and in the prostate. In the female tissues and most likely in the prostate as well, it also plays a role in the development of cancer (for recent reviews, see Refs. 1– 4). Two related proteins, estrogen receptor ␣ (ER␣)1 and ␤ (ER␤), which function as transcription factors * This work was supported by National Institutes of Health Grant R01 CA80210 (to P. J. K.) and R01 CA70896, R01 CA75503, R01 CA86072, the Breast Cancer Alliance Inc., The Susan Komen Breast Cancer Foundation (to R. G. P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 储 Present address: Depts. of Pathology and Biochemistry and Molecular Genetics, University of Virginia, Charlottesville, VA 22908-0214. ** Consultant, Director, and holder of significant financial interests in KaroBio AB, a Swedish company that develops pharmaceuticals that target nuclear receptors. To whom correspondence should be addressed: Dept. of Medicine, 2200 Post St., Rm. C442, University of California, San Francisco, CA 94112-1640. Tel.: 415-476-6790; Fax: 415-885-7724; E-mail: [email protected]. 1 The abbreviations used are: ER␣ and -␤, estrogen receptor ␣ and ␤, respectively; ␣ERKO and ␤ERKO, ER␣ and ER␤ knock-out, respecThis paper is available on line at http://www.jbc.org

to regulate expression of target genes, carry out and modulate the effects of estrogen (2, 5). Studies with mice carrying disrupted estrogen receptors indicate that ER␣ mediates the major proliferative effects of estrogen (6). Thus, female mice in which ER␣ has been knocked out (␣ERKOs) lack estrogenprovoked proliferation of the uterus, cervix, and vagina and have rudimentary mammary glands (6 – 8). Male ␣ERKO mice are completely resistant to estrogen-provoked prostate metaplasia and cancer (9, 10). In contrast, ER␤ knockout (␤ERKO) female mice have full estrogen responses of the reproductive tract, and males have a full response of the prostate. Indeed, there is suggestive evidence that ER␤ may modulate the proliferative effects of ER␣. Thus, the ␤ERKOs are reported to have exaggerated estrogen responses in the uterus and to have spontaneous hyperplasia of the prostate, although the latter observation is not free of controversy (11–13). Furthermore, there is a progressive loss of ER␤ expression in prostate cancer and in mammary multistage carcinogenesis, consistent with a loss of a potential inhibitor of proliferation (14 –16). Estrogen receptors (ERs) regulate gene expression in two ways. In the classical mode of action, ERs bind directly to classical estrogen response elements (EREs) within promoters of estrogen-regulated target genes. Tethered on the DNA, the ERs recruit p160-CBP and possibly DRIP/TRAP coactivator complexes that remodel chromatin and mediate function of the transcriptional machinery (17). In a second mode of action, ERs regulate transcription at promoter elements that directly bind heterologous transcription factors. These promoter elements include AP-1 sites that bind Jun/Fos (18), variant cyclic AMPresponse elements (CREs) that bind c-Jun/ATF-2 proteins (19, 20), and Sp1 sites (21, 22). These sites for AP-1, CRE, or Sp1 do not bind directly ERs, and regulation is presumed to occur through protein-protein interactions. The two ERs have a similar domain structure and similar actions at EREs (5). The central DNA binding domain is highly conserved between ER␣ and ER␤, and the C-terminal ligand binding domain, which is responsible for high affinity ligand binding, dimerization, and hormone-dependent activation (AF2), is moderately conserved. Only the N-terminal region is poorly conserved; ER␣, but not ER␤, has a hormone-independent activation function (AF-1) (23). Given the conservation, it is not surprising that ER␣ and ER␤ interact with the same spectrum of EREs and exhibit similar patterns of gene activation at classical ERE-containing target genes. Thus, both ER␣ and ER␤ activate gene expression from either consensus or tively; ER, estrogen receptor; ERE, estrogen response element; AP-1, activating protein-1; CRE, cyclic AMP-response element; AF, activation function; CD1, cyclin D1; LUC, luciferase; E2, 17-␤-estradiol; DES, diethylstilbesterol; Ral, raloxifene; Tam, tamoxifen.

24353

Downloaded from http://www.jbc.org/ by guest on December 25, 2015

Induction of cyclin D1 gene transcription by estrogen receptor ␣ (ER␣) plays an important role in estrogenmediated proliferation. There is no classical estrogen response element in the cyclin D1 promoter, and induction by ER␣ has been mapped to an alternative response element, a cyclic AMP-response element at ⴚ57, with possible participation of an activating protein-1 site at ⴚ954. The action of ER␤ at the cyclin D1 promoter is unknown, although evidence suggests that ER␤ may inhibit the proliferative action of ER␣. We examined the response of cyclin D1 promoter constructs by luciferase assay and the response of the endogenous protein by Western blot in HeLa cells transiently expressing ER␣, ER␣K206A (a derivative that is superactive at alternative response elements), or ER␤. In each case, ER activation at the cyclin D1 promoter is mediated by both the cyclic AMP-response element and the activating protein-1 site, which play partly redundant roles. The activation by ER␤ occurs only with antiestrogens. Estrogens, which activate cyclin D1 gene expression with ER␣, inhibit expression with ER␤. Strikingly, the presence of ER␤ completely inhibits cyclin D1 gene activation by estrogen and ER␣ or even by estrogen and the superactive ER␣K206A. The observation of the opposing action and dominance of ER␤ over ER␣ in activation of cyclin D1 gene expression has implications for the postulated role of ER␤ as a modulator of the proliferative effects of estrogen.

24354

Opposing Cyclin D1 Regulation by Estrogen Receptors 50% of human mammary carcinomas, including those that are estrogen-responsive (43– 46). Targeted overexpression of cyclin D1 protein in mammary epithelial cells leads to ductal hyperproliferation and eventual tumor formation (47). Mice nullizygous for cyclin D1 show profound defects in estrogen-driven mammary lobuloalveolar development during pregnancy, indicating that the induction of the cyclin D1 gene plays a critical role in the maturation of this tissue (48 –50). It should be noted, however, that cyclin D1 gene knockout mice still had estrogenstimulated ductal elongation and branching during puberty and early pregnancy. Nonetheless, a recent study has shown that cyclin D1-deficient mice are resistant to breast cancers induced by the neu and ras oncogenes (51). Whereas the above studies suggest a pathway for ER␣ action at the cyclin D1 gene and subsequent proliferation, they leave the role of ER␤ unclear. The suggestion from the ER knock-out studies is that ER␤ may modulate the proliferative effects of ER␣, but the matter is controversial. Furthermore, there is some uncertainty as to the exact elements in the cyclin D1 promoter that mediate the effects of ER␣. We therefore sought to confirm the cis-elements in the cyclin D1 promoter that are required for the transcriptional activation of the gene targeted by human ERs and determine how ER␤ acts at the cyclin D1 promoter. In the present study, the response elements of the cyclin D1 promoter targeted by ERs were mapped to the CRE as well as the AP-1 site, with the CRE predominant. Unlike ER␣, ER␤ complexed to 17-␤-estradiol repressed cyclin D1 gene transcription and blocked ER␣-mediated induction when both receptors were present. This suggests that ER␤ may indeed modulate the proliferative effects of ER␣⫺estrogen by blocking its action at the cyclin D1 gene or other key pro-proliferative target genes containing CRE or AP-1 sites. EXPERIMENTAL PROCEDURES

Materials—HeLa cells were grown in Dulbecco’s modified Eagle’s/ F-12 Coon’s modified medium (Sigma) with 15 mM Hepes, L-glutamine (0.438 g/liter), NaHCO3 (1.338 g/liter), 8% iron-supplemented calf serum (Sigma), and penicillin/streptomycin. E2, tamoxifen (Tam), and diethylstilbestrol (DES) were purchased from Sigma. Imperial Chemical Industries (ICI) 182780 was a gift from Dr. A. Wakeling (Astra/ Zeneca, Macclesfield, UK). Raloxifene (Ral) was a gift from Paul Webb and was extracted from Evista. Construction of Reporter Genes and Expression Vector—Human cyclin D1 promoter reporter constructions used for luciferase assays in pA3LUC termed ⫺963CD1LUC, ⫺963AP1mtCD1LUC, ⫺163CD1LUC, and ⫺163⌬Sp1CD1LUC have been previously described (52–56). ⫺963CREmtCD1LUC and ⫺1745AP1/CREmtCD1LUC were generated by primer-based PCR and confirmed by sequencing. ERE-II-LUC and the full-length human ER␣ cDNA (HEO) in pSG5 expression vector have been previously described (27). The full-length human ER␤ cDNA in pCMV5 expression vector was moved into pSG5 vector as an EcoRI and HindIII fragment, and the orientation of the insert was confirmed by sequencing. The ER␣K206A mutant was introduced into the full-length ER␣ cDNA in pSG5 vector by PCR-based site-directed mutagenesis (QuikChange kit; Stratagene) and confirmed by sequencing. The ER␣K206A creates a point mutation in the codon coding for lysine 206 into alanine in the first zinc finger of the human ER␣ DNA binding domain. The oligonucleotide containing the desired nucleotide changes for ER␣K206A was 5⬘-CTGTGAGGGCTGCGCTGCCTTCTTCAAGAG3⬘, and its complementary strand is 5⬘-CTCTTGAAGAAGGCAGCGCAGCCCTCACAG-3⬘. Transfection and Luciferase Assays—Cells were grown to a density of not more than 5 ⫻ 104 cells/cm2. HeLa cells were transfected by electroporation. 2–3 million cells were trypsinized and resuspended in 0.5 ml of phosphate-buffered saline supplemented with 10% glucose and 10 ␮g/ml BioBrene (Applied Biosystems, Foster City, CA) in a single 0.4-cm gap electroporation cuvette with 2.5 ␮g of the luciferase reporter plasmid and the expression vector plasmid and 1.0 ␮g of actin-␤-galactosidase plasmid internal control. Cells were electroporated at 0.24 kV, 960 microfarads in a Bio-Rad Gene Pulser II apparatus (Bio-Rad). Following electroporation, the cells were immediately resuspended in


divergent EREs. There are nonetheless subtle differences between the two ERs. In particular, ER␤ requires higher levels of 17-␤-estradiol (E2) for activation at an ERE than does ER␣ (24, 25). In cells that have both ER␣ and ER␤, heterodimers are the dominant species. In these heterodimers, ER␤ functions as a transdominant inhibitor of ER␣ with subsaturating hormone levels, although ER␤ does not interfere with ER␣-activated transcription at saturating levels of hormone (24, 25). Despite their similar action at EREs, ER␣ and ER␤ have completely different effects at AP-1 sites. ER␣ activates and ER␤ inhibits transcription from an AP-1 site when the receptors are complexed to 17-␤-estradiol (26, 27). At Sp1 sites, the differences between the ERs are less dramatic; ER␣ activates with multiple ligands, but in a cell-specific manner, and ER␤ is nearly inactive (21, 22). To explain ER action at Sp1 sites, it has been proposed that ER binding to Sp1 increases the binding of Sp1 to its cognate element, thereby enhancing transcription (21, 22). The model for ER action at AP-1 sites is more complicated. It has been proposed that ER␣ is present at AP-1 sites through contact with p160-CBP coactivators that have been recruited by Jun/Fos. ER␣-estrogen is believed to trigger the ability of coactivator to stimulate transcription (18). It has been proposed that ER␤-raloxifene, in contrast, is not present at the AP-1 site within the promoter and functions instead by serving as a decoy for inhibitors that would otherwise dampen transcription (18). One important target gene through which estrogen-complexed ER␣ mediates its proliferative action on mammary cancer cells in culture is cyclin D1, a major regulator of entry into the proliferative stage of the cell cycle (28, 29). Thus, there is a strong correlation between increased proliferative response and increased levels of cyclin D1 mRNA with increased levels of ER␣ overexpression in MCF-7 breast cancer cells (30). The effect of ER␣⫺estrogen on cyclin D1 protein expression appears to be predominantly transcriptional with increased expression of cyclin D1 mRNA preceding changes in cyclin D1 protein (31, 32). The abundance of cyclin D1 rises during estrogen-provoked proliferation and declines following exposure to antiestrogens in ER␣-positive MCF-7 cells (33–35). Strikingly, the estrogenprovoked entry into the cell cycle is blocked by antisense cyclin D1 or by microinjection of anti-cyclin D1 antibodies, whereas the blockade imposed by antiestrogens can be overcome by cyclin D1 gene expression in MCF-7 cells (32, 36, 37). Although cyclin D1 gene transcription is directly inducible by estrogen, there is no ERE-related sequence in the promoter region (38). Instead, the cyclin D1 promoter contains multiple regulatory elements, including binding sites for AP-1, STAT5, NF-␬B, E2F, Oct1, Sp1, Myc/Max, Egr, Ets, CRE, and TCF/ LEF (see Ref. 39 and references therein). Sabbah et al. (20) showed that E2 induced reporter gene activity in MCF-7 cells transfected with a construct containing the ⫺973 to ⫹139 region of the cyclin D1 promoter. Deletion analysis of this promoter in ER-negative HeLa cells identified a variant CRE at ⫺57 as the E2-responsive region (20). This variant CRE appeared to bind Jun/ATF-2. Altucci et al. mapped the estrogen-responsive region to a fragment between ⫺994 and ⫺136 of the cyclin D1 promoter (40). Several potential binding sites for known transcription factors can be found in this region of the promoter including the AP-1 site at ⫺954, indicating that this AP-1 site has a potential contribution. Recently, Safe and colleagues (41) reported that the Sp1 site in the cyclin D1 promoter also contributed to the estrogen response with ER␣. The cyclin D1 gene and its induction by ER␣ also appear to play a key role in estrogen-provoked proliferation in vivo (42). The cyclin D1 gene is amplified in up to 20% of human breast cancers, whereas cyclin D1 protein is overexpressed in over

Opposing Cyclin D1 Regulation by Estrogen Receptors

24355

growth medium, plated into 12-well dishes at 1 ml/well, and treated with ligands (ICI 182780 (1 ␮M), raloxifene (1 ␮M), tamoxifen (5 ␮M), E2 (0.1 ␮M), DES (0.1 ␮M), or ethyl alcohol vehicle control). After 40 – 48 h of incubation at 37 °C, the cells were lysed by first removing the medium from the wells, washing with phosphate-buffered saline, and then adding 0.1 ml of lysis buffer consisting of 100 mM Tris (pH 7.8) and 0.2% Triton X-100 for 10 min at 4 °C. Luciferase and ␤-galactosidase activities were then measured using standard luciferase (Promega, Madison, WI) and ␤-galactosidase detection kits (Applied Biosystems, Bedford, MA). Luciferase activities were normalized for ␤-galactosidase activities. Individual transfections (each containing data from triplicate wells) were repeated three times. Western Blotting—HeLa cells were transfected with human ER␣, ER␤, and ER␣K206A expression plasmids by electroporation and treated with ligands as above. After 46 – 48 h of treatment, cells were washed with cold phosphate-buffered saline buffer and harvested in radioimmune precipitation buffer (1⫻ phosphate-buffered saline, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, and protease inhibitors phenylmethylsulfonyl fluoride, sodium orthovanadate, aprotinin (Sigma)). The whole cell extracts were separated by SDS-PAGE in 10% gels and electroblotted to nitrocellulose membranes. The mem-


FIG. 1. ER␣ activates the cyclin D1 promoter mainly with estrogens and does so through the CRE with minor contributions from the AP-1 site. A, expression of a luciferase reporter driven by the human cyclin D1 promoter (⫺963CD1LUC) or mutant in the AP-1 site, CRE, or both and activated by human ER␣ or empty vector (pSG5). HeLa cells were transfected by electroporation with 2.5 ␮g of luciferase reporter plasmid and 2.5 ␮g of expression vector plasmid and 1.0 ␮g of actin-␤-galactosidase plasmid internal control. Transfected cells were treated with ICI 182780 (ICI; 1.0 ␮M), Ral (1.0 ␮M), Tam (5.0 ␮M), E2 (0.1 ␮M), DES (0.1 ␮M), and an ethyl alcohol vehicle (Control). B, Western blot of endogenous cyclin D1 protein from HeLa cells transfected with an expression vector for ER␣ and treated with ligands as indicated (left). The mean OD for quantitative measure of cyclin D1 level from three different Western blots (right).

branes were incubated with a monoclonal antibody against human cyclin D1 (HD11; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and an anti-mouse horseradish peroxidase second antibody. The immune complexes were detected with an enhanced chemiluminescence detection kit (Invitrogen). Quantitation of Autoradiograms—Autoradiograms from Western blots were quantitated using NIH Image 1.62 (National Institutes of Health, Bethesda, MD). A gel quantification macro was used to generate density peaks for each band based on an internal optical density standard. Peaks were user-defined, and the area under the curve for each density peak was measured. Measurements were made on each of three different Western blots. Graphs represent mean values, and error bars represent the S.D. RESULTS

ER␣ Up-regulation of the Cyclin D1 Promoter Is Mediated Mainly by the CRE but Also by the AP-1 Site—We examined the transactivation properties of human ER␣ expressed in HeLa cells on a series of cyclin D1 gene promoter constructs driving luciferase. In particular, we compared a reporter plasmid that

24356



FIG. 2. ER␣K206A superactivates cyclin D1 gene expression exclusively with estrogens. A, expression of the cyclin D1 luciferase reporter in HeLa cells expressing ER␣, ER␣K206A, or empty vector (pSG5) and treated with ligands as indicated. B, Western blot of endogenous cyclin D1 protein from HeLa cells transfected with an expression vector for ER␣K206A and treated with ligands as indicated.

contained the wild-type cyclin D1 promoter (⫺963CD1LUC) with a similar reporter carrying mutations in the AP-1 site (⫺963AP1mtCD1LUC), in CRE (⫺963CREmtCD1LUC), or in both the AP-1 site and the CRE (⫺1745AP1/CREmtCD1LUC). The estrogens (17-␤-estradiol and diethylstilbesterol) up-regulated the ER␣ action at the wild-type cyclin D1 gene promoter, and antiestrogens had a modest but valuable effect (Fig. 1A). Induction with E2 and DES was around 3-fold. The estrogen response of the AP-1 mutant was slightly, but consistently, weaker than that of the wild-type cyclin D1 promoter. The response of the CRE mutant, in contrast, was much weaker than wild-type but was not abolished. The response of the AP-1/CRE double mutant was abolished. Thus, ER␣ up-regulates the cyclin D1 promoter with estrogens through both the CRE and the AP-1 site, which have partly redundant functions. We also examined the expression of the endogenous cyclin D1 gene regulated by ER␣ in HeLa cells. We used Western blot analysis using a cyclin D1-specific antibody of extracts from ER␣ transiently transfected HeLa cells. A modest increase of cyclin D1 protein levels was observed in HeLa cells transiently transfected with an expression vector for ER␣ and treated with estrogens and a smaller increase with antiestrogens (Fig. 1B). Cyclin D1 protein levels did not increase in HeLa cells transiently transfected with control vector after treatment with antiestrogens and estrogens (data not shown). ER␣K206A, a Mutant Selectively Superactive at Alternative Response Elements, Superstimulates Cyclin D1 Gene Expression Only with Estrogens—ER␣ is unique among the nuclear receptors in its ability to enhance AP-1-mediated transcription with the estrogen. Other nuclear receptors, such as glucocorticoid receptor and thyroid receptor, repress AP-1-dependent transcription in response to their cognate hormone. Interestingly, a point mutation at the base of the first zinc finger of the DNA binding domain converts glucocorticoid receptor and thyroid receptor from inhibitors to AP-1 activators (57). When we introduced the homologous mutation (K206A) into the first zinc finger of the DNA binding domain of human ER␣, we found that the mutant receptor (ER␣K206A) was superactive at target genes regulated by AP-1 but underactive at target genes

FIG. 3. ER␤ activates the cyclin D1 promoter with antiestrogens and does so through either the CRE or AP-1 site. A, expression of wild-type and mutant cyclin D1 luciferase reporters in HeLa cells expressing human ER␤ in the presence of the indicated ligands. B, Western blot of endogenous cyclin D1 protein in HeLa cells transfected with an expression vector for human ER␤ and treated with ligands as indicated. C, estrogens block activation of the cyclin D1 gene by ER␤ and antiestrogens. Activity of the cyclin D1 luciferase reporter in HeLa cells with human ER␤ and treated with ligands as indicated. Note the abolition of the activating effects of antiestrogens (ICI 182780 (ICI), Ral, and Tam) by estrogens (E2 and DES).

with classical EREs. This pattern occurred in cell culture and in transgenic mice.2 In accord with these results, ER␣K206A enhanced transcriptional activity at the cyclin D1 promoter about 20-fold with estrogens and was consistently 7–10-fold more active than ER␣ (Fig. 2A). No activation occurred with 2

B. Anderegg and R. M. Uht, unpublished results.


24357


FIG. 4. ER␤ inhibits ER␣ activation of the cyclin D1 gene with estrogen, and ER␣ modulates ER␤ activation of the cyclin D1 gene with antiestrogens. A, ER␤ inhibition of ER␣ transcriptional activity at the cyclin D1 promoter with 17-␤-estradiol (0.1 ␮M) monitored with cyclin D1 luciferase reporter, human ER␣ expression vector (200 ng), and various amounts of expression vector for human ER␤. B, expression of the endogenous cyclin D1 protein in HeLa cells transfected with an expression vector for ER␣, ER␤, both, or empty vector and treated with 17-␤-estradiol (0.1 ␮M) (above). The mean OD for quantitative measure of cyclin D1 level from three different Western blots is shown (below). C, ER␣ modulates ER␤ activation of the cyclin D1 gene with antiestrogens. Activity of the cyclin D1 luciferase reporter in HeLa cells transfected with expression vectors for ER␣, ER␤, both ER subtypes, or empty vector and treated with ligands as indicated. D, ER␤ modulates ER␣ activation at the ERE with estrogens. Activity of the ERE-II-LUC reporter in HeLa cells transfected with expression vectors for ER␣, ER␤, both, or empty vector and treated with ligands as indicated.

antiestrogens in contrast to the weak activation observed with ER␣. Because the transcriptional effect of ER␣K206A on cyclin D1 promoter was so profound, we explored whether expression of the endogenous cyclin D1 gene in HeLa cells might also respond to transient transfection with this superactive receptor. Cyclin D1 protein levels consistently rose in HeLa cells transiently transfected with ER␣K206A after treatment with E2 or DES (Fig. 2B) but never after treatment with antiestrogens. The induction of the endogenous gene, although modest in comparison with the response of cotransfected reporter gene,

was consistently observed. We conclude that the endogenous cyclin D1 gene of HeLa cells is induced by estrogen and ER␣K206A. ER␤ Up-regulates Cyclin D1 Gene Expression with Antiestrogens through the CRE and the AP-1 Site and Down-regulates Cyclin D1 Gene Expression with Estrogens—In a previous study, we showed that ER␤ was a potent transcriptional activator at target genes with AP-1 sites when bound to the antiestrogens tamoxifen, raloxifene, and ICI 164384. ER␤ repressed AP-1-regulated gene activity with estrogens (26). Thus, we investigated whether antiestrogens and estrogens also have

24358


opposite transcriptional action at the cyclin D1 promoter with ER␤. In HeLa cells, ER␤ slightly, but consistently, repressed transcription of the intact cyclin D1 promoter below basal


FIG. 5. Mutual inhibition exists between ER␣K206A and ER␤ action at the cyclin D1 promoter. A, activity of the cyclin D1 luciferase reporter with expression vector for ER␤, ER␣K206A, both, or empty vector in HeLa cells and treated with ligands as indicated. Note that whereas ER␣K206A activates with estrogens (E2 and DES) and ER␤ activates with antiestrogens (ICI 182780 (ICI), Ral, and Tam), when both receptors are present there is little activation with any ligand. B, expression of the endogenous cyclin D1 protein in HeLa cells transfected with an expression vector for ER␣K206A, ER␤, both, or empty vector and treated with 17-␤-estradiol (0.1 ␮M). C, ER␤ does not inhibit ER␣K206A activation at the ERE with estrogens. Activity of the ERE-II-LUC reporter in HeLa cells transfected with expression vectors for ER␤, ER␣K206A, both, or empty vector and treated with ligands as indicated.

levels after treatment with estrogen E2 or DES (Fig. 3A). ER␤ activated cyclin D1 promoter activity after treatment with antiestrogens ICI 182780, raloxifene, and tamoxifen. The activation by ER␤ and antiestrogens was potent and comparable with that produced by ER␣K206A and estrogens. ER␤-antiestrogen also activates the cyclin D1 promoter through either the AP-1 site or CRE. Some activity persists if either of the elements is mutated, although the activity of transcription is substantially weakened by mutation of the CRE. However, the AP-1/CRE double mutant of the cyclin D1 promoter almost abolishes the ER␤-antiestrogen transcriptional activity (Fig. 3A). Thus, like ER␣, ER␤ activation of the cyclin D1 promoter maps to both the CRE and the AP-1 sites, and the former plays a more important role. In a parallel experiment, we investigated the endogenous cyclin D1 protein levels regulated by ER␤. An increase of cyclin D1 protein levels was observed in HeLa cells transiently transfected with an expression vector for ER␤ and treated with antiestrogens. No increase in cyclin D1 protein levels occurred in ER␤-transfected cells after treatment with the estrogens E2 and DES (Fig. 3B). In ligand competition experiments, the estrogens E2 and DES reduced the induction by ER␤ and the antiestrogens raloxifene, ICI 182780, and tamoxifen to the basal level of transcription (Fig. 3C). Thus, the activation of the cyclin D1 promoter appears to be mediated by an ER␤-antiestrogen complex that can be competed with estrogens. In summary, the pharmacology of estrogens and antiestrogens is unusual at the cyclin D1 gene promoter with ER␤; the antiestrogens act as transcription activators, up-regulating the cyclin D1 promoter, and the estrogens act as transcription inhibitors, down-regulating the cyclin D1 promoter. ER␤ Inhibits ER␣ Transcriptional Activity at the Cyclin D1 Gene Promoter with Estrogen, and ER␣ Modulates ER␤ Transcriptional Activity at the Cyclin D1 Gene Promoter with Antiestrogens—ER␣ and ER␤ are co-expressed in some tissues such as the mammary gland, ovary, and possibly the prostate. Thus, in these tissues, the effect of each receptor could be affected by the presence of the other. Indeed, previous studies have shown that ER␣ and ER␤ form heterodimers that bind to and activate at ERE-regulated target genes (58 – 60) and that the presence of excess ER␤ can increase the concentration of estrogen that is required (24, 25). Considering the above observation of the opposite effect of ER␣ and ER␤ on cyclin D1 gene transcription with estrogens, we investigated whether the transactivation properties of ER␣ action at the cyclin D1 promoter are affected by the presence of ER␤. In HeLa cells transfected with ER␣, activation of the cyclin D1 promoter with E2 was completely inhibited by cotransfection with ER␤. Moreover, a low concentration of ER␤ (20 ng) inhibited a higher concentration of ER␣ (200 ng) action at the cyclin D1 promoter (Fig. 4A). We conclude that ER␤ is a potent inhibitor of ER␣ activation of cyclin D1 promoter. In transiently transfected HeLa cells, ER␤ with E2 also inhibited endogenous cyclin D1 protein levels that were up-regulated by ER␣ (Fig. 4B). Thus, ER␤ inhibits ER␣ activation of cyclin D1 gene expression both from reporter gene constructs and endogenous cyclin D1 protein with estrogens. Conversely, ER␣ modulates ER␤ transcriptional activity at the cyclin D1 promoter with antiestrogens (Fig. 4C). We conclude that ER␣ functions as a transdominant inhibitor of ER␤ activity at the cyclin D1 promoter with antiestrogens. Unlike at the cyclin D1 promoter, both ERs were transcriptional activators at target genes containing ERE when bound to the estrogens. ER␣, as previously noted, was more transcriptionally active than ER␤, and ER␤ acted as a transdominant modulator of ER␣ activity (Fig. 4D).


DISCUSSION

Previous studies have mapped ER␣ induction of the cyclin D1 promoter activity to the CRE at ⫺57, although an earlier report proposed a role for the AP-1 site at ⫺954 (20, 40). We have examined the roles of these elements in cyclin D1 induction mediated by ER␣ and ER␤. We find, consistent with previous studies of Sabbah et al. (20), that the variant CRE in the cyclin D1 promoter is the primary mediator of estrogen response with ER␣. The AP-1 site contributes relatively little. However, cyclin D1 activation by ER␤ or ER␣K206A is much stronger than that with ER␣ and reveals that there is some overlap of function between the CRE and the AP-1 site. With these strong activators, a robust response remains as long as either the CRE or AP-1 site is maintained. While this work was in progress, Safe and colleagues (41) reported that the Sp1 site also contributed to the estrogen response with ER␣. We have found a similar small contribution of the Sp1 site (data not shown). However, the Sp1 site in the cyclin D1 promoter, unlike the CRE or AP-1 site, does not by itself mediate robust induction with ER␤ or ER␣K206A. A notable finding of our studies is the differential ligand preference for activation of the cyclin D1 promoter by ER␣ and ER␤. We previously reported that ER␣ activated at AP-1 sites exclusively with estrogen in mammary cells and with both estrogen and tamoxifen in HeLa cells and Ishikawa cells (26). Only estrogen induces the gene containing AP-1 sites in mammary cells with ER␣ (20). For the most part, a similar pattern of action is seen at cyclin D1. Estrogen and tamoxifen (albeit weakly) induce cyclin D1 in HeLa cells. The exception to the previous pattern is that ICI 182780 also gives weak, but variable, activation of cyclin D1 gene expression in HeLa cells with ER␣. We are uncertain as to the origin of this action, which may be due to differences between the collagenase and cyclin D1 promoters or to differences in cell culture conditions. We note that mutations that interfere with ER␣ AF-1 activity awaken a previously dormant ability to activate AP-1 targets with ICI 182780 (27). It is possible that cell culture conditions that weaken AF-1 might have a similar effect. In contrast to the pattern with ER␣, where estrogens are the preferred ligand, ER␤ activates exclusively with antiestrogens

and represses cyclin D1 gene expression with estrogens. The contrast in ligand preference is especially stark between the superactive ER␣K206A and ER␤. Whereas ER␣K206A activates the cyclin D1 reporter gene only with estrogens (17-␤estradiol and DES), ER␤ activates this promoter only with antiestrogens (tamoxifen, raloxifene, and ICI 182780). The differential ligand preference for ER␣ or ER␤ activation at the cyclin D1 promoter is similar to the previously noted differential ligand preference at the AP-1-regulated collagenase promoter (26). The differential ligand preference is not restricted to activation of artificially constructed reporter genes. The expression of endogenous cyclin D1 protein in HeLa cells mediated by transiently expressed ERs and assayed by Western blotting shows a similar pattern. Also, ER␣ or ER␣K206A activation of the endogenous gene requires estrogen, whereas ER␤ activation requires antiestrogens. Together with the previous studies of the AP-1-regulated collagenase promoter, these studies suggest that differential ligand action between ER␣ and ER␤ may be a general feature of ER action at promoters with AP-1 site or CRE. Since the two ERs have opposite action at the cyclin D1 promoter in the presence of a given ligand, we examined the transactivation properties in cells where both ERs are expressed. Strikingly, the presence of ER␤ completely inhibits cyclin D1 gene activation by estrogens and ER␣ or even by estrogens and the superactive ER␣K206A. ER␤ blocks ER␣ activation both of reporter gene constructs and of the endogenous gene. There is a reciprocal inhibition of ER␤-antiestrogen activation by ER␣, but this inhibition is relatively inefficient unless the superactive receptor (ER␣K206A) is transfected. Whereas it has been previously reported that ER␤ modulates ER␣ activation at an ERE, the modulation only occurred with subsaturating 17-␤-estradiol concentrations and was only partial (24, 25). In contrast, ER␤ inhibits ER␣ activation of cyclin D1 with saturating 17-␤-estradiol concentrations, and the inhibition is nearly complete. These observations suggest that ER␤ has the ability to block estrogen activation of the cyclin D1 gene mediated by ER␣ if the two receptors are present in the same cell. Only a few studies have directly examined whether co-expression of ␣ and ␤ occurs in individual cells. The clearest example of co-expression is in rat mammary gland epithelial cells during puberty and again during lactation (61). In addition, there is suggestive evidence from the broad distribution of both receptors in some tissues that they must be co-expressed in at least some cells. Thus, both receptors may be present in human mammary gland epithelial cells, in epithelial cells of the uterus (although ␣ is much more abundant than ␤), and in ovarian epithelial and interstitial glandular cells (62). Clearly, the presence of both receptors in individual cells needs to be examined in light of the present findings. The observations here suggest the intriguing possibility that ER␤ could inhibit estrogen-induced proliferation by inhibiting ER␣ activation of the cyclin D1 gene and other pro-proliferative targets that might be similarly regulated by AP-1 or CRE. As noted in the introduction, there is provocative evidence that loss of ER␤ accompanies proliferative disease in the mammary gland and prostate and that knock-out of ER␤ leads to hyperproliferation in the mouse uterus and possibly the prostate, although the latter is more controversial. These observations suggest a potential modulatory role of ER␤ in estrogen-induced proliferation. However, it is unknown whether such modulation is via direct inhibition by ER␤ of ER␣ activation of gene expression, and the matter remains to be investigated. Acknowledgment—We thank Cathleen Valentine for technical help.


Mutual Inhibition Exists between ER␣K206A and ER␤ Action at the Cyclin D1 Promoter with Estrogens and Antiestrogens—ER␣K206A activates the cyclin D1 promoter much more strongly than dose wild-type ER␣. We thus investigated whether ER␤ was able to inhibit ER␣K206A superactivation at the cyclin D1 promoter with estrogens and conversely whether ER␣K206A can inhibit ER␤ activity at the cyclin D1 promoter with antiestrogens. In HeLa cells cotransfected with both ER␣K206A and ER␤ expression vectors along with the reporter driven by the cyclin D1 promoter, surprisingly, ER␤ nearly completely inhibited ER␣K206A luciferase transcription to basal level with the estrogens E2 and DES. However, ER␣K206A inhibited ER␤ transcriptional activity at the cyclin D1 promoter with antiestrogens (Fig. 5A). Thus, a mutual inhibition of transcriptional action at the cyclin D1 promoter exists between ER␣K206A and ER␤. We next asked whether ER␤ inhibits the endogenous cyclin D1 protein levels regulated by ER␣K206A. Strikingly, in transiently transfected HeLa cells, ER␤ with E2 also inhibited endogenous cyclin D1 protein levels that were up-regulated by ER␣K206A (Fig. 5B). Thus, ER␤ inhibits ER␣K206A activation of cyclin D1 gene expression both from reporter gene constructs and endogenous cyclin D1 protein with estrogens. Unlike at the cyclin D1 promoter, at an ERE, both ER␤ and ER␣K206A were transcriptional activators, and ER␤ did not repress ER␣K206A activity with estrogens (Fig. 5C).

24359

24360

Opposing Cyclin D1 Regulation by Estrogen Receptors REFERENCES

R. L. (1997) J. Biol. Chem. 272, 10882–10894 32. Prall, O. W., Rogan, E. M., Musgrove, E. A., Watts, C. K., and Sutherland, R. L. (1998) Mol. Cell. Biol. 18, 4499 – 4508 33. Watts, C. K., Brady, A., Sarcevic, B., deFazio, A., Musgrove, E. A., and Sutherland, R. L. (1995) Mol. Endocrinol. 9, 1804 –1813 34. Musgrove, E. A., Hui, R., Sweeney, K. J., Watts, C. K., and Sutherland, R. L. (1996) J. Mammary Gland Biol. Neoplasia 1, 153–162 35. Sutherland, R. L., Hamilton, J. A., Sweeney, K. J., Watts, C. K., and Musgrove, E. A. (1995) Ciba Found. Symp. 191, 218 –234 36. Lukas, J., Bartkova, J., and Bartek, J. (1996) Mol. Cell. Biol. 16, 6917– 6925 37. Carroll, J. S., Prall, O. W., Musgrove, E. A., and Sutherland, R. L. (2000) J. Biol. Chem. 275, 38221–38229 38. Herber, B., Truss, M., Beato, M., and Mueller, R. (1994) Oncogene 9, 1295–1304 39. Allan, A. L., Albanese, C., Pestell, R. G., and LaMarre, J. (2001) J. Biol. Chem. 276, 27272–27280 40. Altucci, L., Addeo, R., Cicatiello, L., Dauvois, S., Parker, M. G., Truss, M., Beato, M., Sica, V., Bresciani, F., and Weisz, A. (1996) Oncogene 12, 2315–2324 41. Castro-Rivera, E., Samudio, I., and Safe, S. (2001) J. Biol. Chem. 276, 30853–30861 42. Steeg, P. S., and Zhou, Q. (1998) Breast Cancer Res. Treat. 52, 17–28 43. Dickson, C., Fantl, V., Gillett, C., Brookes, S., Bartek, J., Smith, R., Fisher, C., Barnes, D., and Peters, G. (1995) Cancer Lett. 90, 43–50 44. Schuuring, E. (1995) Gene (Amst.) 159, 83–96 45. Frierson, H. F., Jr., Gaffey, M. J., Zukerberg, L. R., Arnold, A., and Williams, M. E. (1996) Mod. Pathol. 9, 725–730 46. Hosokawa, Y., and Arnold, A. (1998) Genes Chromosomes Cancer 22, 66 –71 47. Wang, T. C., Cardiff, R. D., Zukerberg, L., Lees, E., Arnold, A., and Schmidt, E. V. (1994) Nature 369, 669 – 671 48. Sicinski, P., Donaher, J. L., Parker, S. B., Li, T., Fazeli, A., Gardner, H., Haslam, S. Z., Bronson, R. T., Elledge, S. J., and Weinberg, R. A. (1995) Cell 82, 621– 630 49. Fantl, V., Stamp, G., Andrews, A., Rosewell, I., and Dickson, C. (1995) Genes Dev. 9, 2364 –2372 50. Fantl, V., Edwards, P. A., Steel, J. H., Vonderhaar, B. K., and Dickson, C. (1999) Dev. Biol. 212, 1–11 51. Yu, Q., Geng, Y., and Sicinski, P. (2001) Nature 411, 1017–1021 52. Albanese, C., Johnson, J., Watanabe, G., Eklund, N., Vu, D., Arnold, A., and Pestell, R. G. (1995) J. Biol. Chem. 270, 23589 –23597 53. Albanese, C., D’Amico, M., Reutens, A. T., Fu, M., Watanabe, G., Lee, R. J., Kitsis, R. N., Henglein, B., Avantaggiati, M., Somasundaram, K., Thimmapaya, B., and Pestell, R. G. (1999) J. Biol. Chem. 274, 34186 –34195 54. Watanabe, G., Albanese, C., Lee, R. J., Reutens, A., Vairo, G., Henglein, B. D., and Pestell, R. G. (1998) Mol. Cell. Biol. 18, 3212–3222 55. Watanabe, G., Howe, A., Lee, R. J., Albanese, C., Shu, I. W., Karnezis, A. N., Zon, L., Kyriakis, J., Rundell, K., and Pestell, R. G. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 12861–12866 56. Lee, R. J., Albanese, C., Stenger, R. J., Watanabe, G., Inghirami, G., Haines, G. K., Webster, M., Muller, W. J., Brugge, J. S., Davis, R. J., and Pestell, R. G. (1999) J. Biol. Chem. 274, 7341–7350 57. Starr, D. B., Matsui, W., Thomas, J. R., and Yamamoto, K. R. (1996) Genes Dev. 10, 1271–1283 58. Cowley, S. M., Hoare, S., Mosselman, S., and Parker, M. G. (1997) J. Biol. Chem. 272, 19858 –19862 59. Pettersson, K., Grandien, K., Kuiper, G. G., and Gustafsson, J. A. (1997) Mol. Endocrinol. 11, 1486 –1496 60. Tremblay, G. B., Tremblay, A., Labrie, F., and Giguere, V. (1999) Mol. Cell. Biol. 19, 1919 –1927 61. Saji, S., Jensen, E. V., Nilsson, S., Rylander, T., Warner, M., and Gustafsson, J. A. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 337–342 62. Pelletier, G., and El-Alfy, M. (2000) J. Clin. Endocrinol. Metab. 85, 4835– 4840


1. Signoretti, S., and Loda, M. (2001) Am. J. Pathol. 159, 13–16 2. Katzenellenbogen, B. S., and Katzenellenbogen, J. A. (2000) Breast Cancer Res. 2, 335–344 3. Dickson, R. B., and Stancel, G. M. (2000) J. Natl. Cancer Inst. Monogr. 27, 135–145 4. Tonetti, D. A., and Jordan, V. C. (1999) J. Mammary Gland Biol. Neoplasia 4, 401– 413 5. Pettersson, K., and Gustafsson, J. A. (2001) Annu. Rev. Physiol. 63, 165–192 6. Couse, J. F., and Korach, K. S. (1999) Endocr. Rev. 20, 358 – 417 7. Lubahn, D. B., Moyer, J. S., Golding, T. S., Couse, J. F., Korach, K. S., and Smithies, O. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 11162–11166 8. Bocchinfuso, W. P., Lindzey, J. K., Hewitt, S. C., Clark, J. A., Myers, P. H., Cooper, R., and Korach, K. S. (2000) Endocrinology 141, 2982–2994 9. Risbridger, G., Wang, H., Young, P., Kurita, T., Wang, Y. Z., Lubahn, D., Gustafsson, J. A., Cunha, G., and Wong, Y. Z. (2001) Dev. Biol. 229, 432– 442 10. Prins, G. S., Birch, L., Couse, J. F., Choi, I., Katzenellenbogen, B., and Korach, K. S. (2001) Cancer Res. 61, 6089 – 6097 11. Weihua, Z., Saji, S., Makinen, S., Cheng, G., Jensen, E. V., Warner, M., and Gustafsson, J. A. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 5936 –5941 12. Weihua, Z., Makela, S., Andersson, L. C., Salmi, S., Saji, S., Webster, J. I., Jensen, E. V., Nilsson, S., Warner, M., and Gustafsson, J. A. (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 6330 – 6335 13. Dupont, S., Krust, A., Gansmuller, A., Dierich, A., Chambon, P., and Mark, M. (2000) Development 127, 4277– 4291 14. Horvath, L. G., Henshall, S. M., Lee, C. S., Head, D. R., Quinn, D. I., Makela, S., Delprado, W., Golovsky, D., Brenner, P. C., O’Neill, G., Kooner, R., Stricker, P. D., Grygiel, J. J., Gustafsson, J. A., and Sutherland, R. L. (2001) Cancer Res. 61, 5331–5335 15. Pasquali, D., Rossi, V., Esposito, D., Abbondanza, C., Puca, G. A., Bellastella, A., and Sinisi, A. A. (2001) J. Clin. Endocrinol. Metab. 86, 2051–2055 16. Roger, P., Sahla, M. E., Makela, S., Gustafsson, J. A., Baldet, P., and Rochefort, H. (2001) Cancer Res. 61, 2537–2541 17. McKenna, N. J., Xu, J., Nawaz, Z., Tsai, S. Y., Tsai, M. J., and O’Malley, B. W. (1999) J. Steroid Biochem. Mol. Biol. 69, 3–12 18. Kushner, P. J., Agard, D. A., Greene, G. L., Scanlan, T. S., Shiau, A. K., Uht, R. M., and Webb, P. (2000) J. Steroid Biochem. Mol. Biol. 74, 311–317 19. Webb, P., Keneally, M.-R., Shinsako, J., Uht, R., Anderson, C., Paech, K., Scanlan, T. S., and Kushner, P. J. (1998) in Molecular Basis of Sex Hormone Receptor Function (Gronemeyer, H., Fuhrmann, U., and Parczyk, K., eds) pp. 121–140, Springer, Berlin 20. Sabbah, M., Courilleau, D., Mester, J., and Redeuilh, G. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 11217–11222 21. Saville, B., Wormke, M., Wang, F., Nguyen, T., Enmark, E., Kuiper, G., Gustafsson, J. A., and Safe, S. (2000) J. Biol. Chem. 275, 5379 –5387 22. Safe, S. (2001) Vitam. Horm. 62, 231–252 23. Delaunay, F., Pettersson, K., Tujague, M., and Gustafsson, J. A. (2000) Mol. Pharmacol. 58, 584 –590 24. Hall, J. M., and McDonnell, D. P. (1999) Endocrinology 140, 5566 –5578 25. Pettersson, K., Delaunay, F., and Gustafsson, J. A. (2000) Oncogene 19, 4970 – 4978 26. Paech, K., Webb, P., Kuiper, G. G., Nilsson, S., Gustafsson, J., Kushner, P. J., and Scanlan, T. S. (1997) Science 277, 1508 –1510 27. Webb, P., Nguyen, P., Valentine, C., Lopez, G. N., Kwok, G. R., McInerney, E., Katzenellenbogen, B. S., Enmark, E., Gustafsson, J. A., Nilsson, S., and Kushner, P. J. (1999) Mol. Endocrinol. 13, 1672–1685 28. Planas-Silva, M. D., and Weinberg, R. A. (1997) Mol. Cell. Biol. 17, 4059 – 4069 29. Prall, O. W., Rogan, E. M., and Sutherland, R. L. (1998) J. Steroid Biochem. Mol. Biol. 65, 169 –174 30. Wilcken, N. R., Prall, O. W., Musgrove, E. A., and Sutherland, R. L. (1997) Clin. Cancer Res. 3, 849 – 854 31. Prall, O. W. J., Sarcevic, B., Musgrove, E. A., Watts, C. K. W., and Sutherland,

GENES: STRUCTURE AND REGULATION: Opposing Action of Estrogen Receptors α and β on Cyclin D1 Gene Expression Meng-Min Liu, Chris Albanese, Carol M. Anderson, Kristin Hilty, Paul Webb, Rosalie M. Uht, Richard H. Price, Jr., Richard G. Pestell and Peter J. Kushner J. Biol. Chem. 2002, 277:24353-24360. doi: 10.1074/jbc.M201829200 originally published online May 1, 2002

Find articles, minireviews, Reflections and Classics on similar topics on the JBC Affinity Sites. Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 61 references, 29 of which can be accessed free at http://www.jbc.org/content/277/27/24353.full.html#ref-list-1


Access the most updated version of this article at doi: 10.1074/jbc.M201829200