The Estrogen Receptor (ER) Variant 5 Exhibits Dominant Positive Activity on ER-Regulated Promoters in Endometrial Carcinoma Cells
http://www.100md.com
内分泌学杂志 2005年第2期
Division of Endocrinology and Metabolism, Department of Internal Medicine (W.B., M.A.S.), Department of Microbiology (A.E.S.), and Department of Obstetrics and Gynecology (L.W.R.), University of Virginia, Charlottesville, Virginia 22908
Address all correspondence and requests for reprints to: Margaret Shupnik, Ph.D., Department of Internal Medicine/Endocrinology, P.O. Box 800578, Health Sciences Center, University of Virginia, Charlottesville, Virginia 22908. E-mail: mas3x@virginia.edu.
Abstract
Estrogen receptor (ER) is a ligand-inducible transcription factor that mediates the physiological effects of 17?-estradiol (E2). In the uterus, E2 is involved in tissue growth, maintenance, and differentiation. 5ER (5) is an ER variant protein expressed in uterine tumors but not in normal tissue. We examined the transcriptional activity of 5 and its modulation of human ER basal and E2-stimulated activity in Ishikawa cells, an endometrial cancer cell line. In transient transfection assays, 5 increased basal activity of an estrogen response element-containing promoter in the absence or presence of ER but lessened stimulation by ER and E2. Effects of 5 were not limited to model reporters, given that cyclin D1 and complement 3 promoters were similarly affected. Increases in basal transcription required dimerization and DNA binding of 5, whereas decreased E2 stimulation with ER required only DNA binding. Decreased ligand stimulation was not unique to E2 but also applied to the selective ER modulators tamoxifen and genistein. However, promoter stimulation by epidermal growth factor is retained with 5. The ER coactivator small nuclear ring finger protein is expressed in Ishikawa cells and uterine tumors, and it enhances effects of 5 alone and with ER on basal activity of an estrogen response element reporter. Thus, in the presence of 5 plus ER, there is a lower transcriptional response to E2 and SERMS, but stimulation by epidermal growth factor is retained. The expression of 5 in uterine carcinoma may provide a mechanism by which tumors could maintain expression of E2-responsive genes in the absence of E2.
Introduction
THE GONADAL STEROID 17?-estradiol (E2) exerts effects on differentiation, growth, and gene regulation in a number of target tissues, including the breast and the uterus (Refs.1 and 2 and references within). The physiological effects of E2 are not always beneficial, given that E2 exposure contributes to the development and growth of breast and uterine cancer (1). The estrogen receptors (ERs) and ? are ligand-inducible transcription factors that act as the primary mediators of E2 action and E2-regulated gene expression. ERs are members of the nuclear receptor superfamily, with conserved protein functional domains (1). The major conserved domains of the ER include an amino terminus activation function (AF)-1 that mediates ligand-independent transcription, and a DNA binding domain (DBD) that binds to estrogen response elements (EREs) on E2-responsive genes and contains a dimerization interface (1, 2). A central hinge region is required for nuclear localization of the ER. The carboxyl terminus is required for the binding of ligand and ligand-stimulated gene transcription via its AF-2. This domain also promotes the recruitment of coactivators and dimerization of ligand-bound ERs (1, 2).
ER activity can be modified through such mechanisms as the expression of ER isoforms, cross-talk between intracellular signaling pathways, and interaction of the ER with coregulator proteins (2). Occupancy of the ER by ligand results in receptor dimerization. Subsequent binding to an ERE in a target gene induces recruitment of various coregulators, which act to enhance (coactivators) or suppress (corepressors) gene expression. The complex of ligand-bound dimers and coregulators contacts the general transcriptional machinery to form a preinitiation complex for transcription. Many coactivators, including SRC-1 (3, 4), cAMP response element binding protein-binding protein (CBP) (5, 6), and small nuclear ring finger protein (SNURF, also known as RF4) (7, 8), stimulate transcription mediated by ER as well as other steroid receptors. Several proteins suppress ER transcription, but to date, the only ER-specific corepressor is the repressor of estrogen activity (REA), which competes with coactivators for binding to ligand-bound ER (9) and consequently suppresses ER-mediated transcription.
In vitro and in vivo evidence support the idea that ER-mediated transcription is also modulated by numerous growth factor signaling pathways (10, 11, 12, 13) and that this requires the N-terminal AF-1 domain (10, 11). The contribution of AF-1 to ER activity is dependent on cell and promoter context and plays an important role in ER activity in uterine cells (12). Direct phosphorylation of serine 118, a highly conserved amino acid residue, via MAPK, has been shown to mediate AF-1 activity in some cell types (13).
Alternate splicing of ER mRNA can produce transcripts that lack one or more exons (14). In normal and neoplastic uterine tissue, expression of full-length ER mRNA is generally greater than the mRNA expression of a given splice variant (15, 16). Occasionally, ER mRNA splice variants are translated into proteins that may exhibit activity that is markedly different from that of wild-type ER (14, 17, 18, 19, 20, 21, 22). These variants have the potential to act alone or in a dominant-positive or -negative fashion and alter the activity of the wild-type ER.
5ER (5) is an ER mRNA variant that was originally identified in primary breast cancer tumors that were classified as ER–/PR+ (22). 5 is also the only ER mRNA splice variant that is translated into protein in primary uterine tumors (23). The deletion of exon 5 produces a frameshift in the coding region between exons 4 and 6, and results in a premature stop codon. The 42-kDa truncated protein is composed of the AF-1, DNA binding, and hinge domains, as well as a unique sequence of eight amino acids at the carboxyl terminus (Fig. 1A). 5 also retains one of two dimerization interfaces and the nuclear localization signal of the ER. The ligand-binding and AF-2 regions are completely absent. 5 activity and its ability to act in a dominant-negative or -positive way with ER is dependent on cell type. For example, in HMT-3522S1 cells, an ER-negative human breast epithelial cell line, transient transfection with ER and 5 decreased the basal and E2-stimulated activity of ER (24). On the other hand, in yeast transfection systems (22), 5 alone had constitutive basal activity on an ERE reporter. 5 acts cooperatively with ER in osteosarcoma cells to significantly increase both the basal and E2-stimulated transcription of an ERE reporter in a dose-dependent manner (18). In MCF-7 cells, an ER-positive breast cancer line, 5 had no effect on ERE reporter activity (19).
FIG. 1. Functional protein domains of ER and 5 and Western blot analysis of hER and 5 expression in Cos-1 and Ishikawa cells. A, Exon and functional domains of ER and 5. Black lines under each cartoon represent the dimerization interfaces. The truncated splice variant 5 lacks a majority of the ligand-binding and AF-2 regions. Deletion of exon 5 results in a frameshift mutation and the production of a premature stop codon (horizontal lines). Diagonal stripes indicate the unique carboxyl terminus in 5. Arrows indicate sites of substituted amino acids residues in the 5 protein. Cos-1 cells were transiently transfected with ER, 5, ER and 5, or 5 mutants. B, Protein samples (30μg) from Cos-1 (left panel) or Ishikawa cells (right panel) were resolved on a 12% acrylamide gel. Blots were probed with 1D5, an N-terminal ER antibody, to visualize ER and 5. Blots were also probed with antibody for glucose-6-dehydrogenase (G6PDH) to confirm equal loading of protein. Arrows show the migration of ER and 5 proteins. C, Proteins (30 μg) from cells transfected with either wild-type or mutant 5 constructs were analyzed for receptor expression as above. Shown are representative blots from four to seven experiments.
We previously reported that primary endometrial carcinoma tumors express appreciable levels of 5 protein, but normal endometrial tissue does not (23). Because uterine cells allow expression of ER AF-1 activity (12), and the AF-1 region is contained in 5, we characterized the activity of 5 alone and the modulatory effects of 5 on ER activity in the presence and absence of both E2 and epidermal growth factor (EGF). We show that the overall effect of 5 in endometrial tumor cells is to increase the basal activity of an ERE-containing reporter. The transcriptional response of ER to E2 (fold-response) is decreased in the presence of 5, although the overall promoter activity is increased. In contrast, the transcriptional response of ER to EGF is not suppressed in the presence of 5. In endometrial cancer cells expressing 5 protein, the presence of this variant could result in stimulation of E2-sensitive genes in the absence of ligand.
Materials and Methods
ER constructs
5 cDNA was cloned and constructs prepared as previously described (23). All human ER (hER) constructs were cloned into the EcoR1 site of the pcDNA 3.1 expression vector containing a cytomegalovirus (CMV) promoter (Invitrogen, Carlsbad, CA). The plasmid CMV-ER contains the entire ER cDNA coding sequence. All mutant constructs were created by site-directed mutagenesis (QuikChange by Stratagene, La Jolla, CA) of CMV hER and CMV 5, and include E203G/G204S, S236E, and S118A. The E203G/G204S mutants were created using the following oligonucleotides: 5'ggagtctggtcctgtgggagctgcaaggccttcttcaag-3' and 5' cttgaagaaggccttgcagctcccacaggaccagactcc-3'. These mutations alter the DNA binding site of ER to that of the recognition site for the glucocorticoid receptor (25). The serine 118 residue was mutated using the following oligonucleotides: 5'-gataaaaacaggaggaaggagtgccaggcctgccggctct-3' and 5'-gagccggcaggcctggcactccttcctcctgtttttatc-3'. Substitution of the serine residue at position 118 with an alanine residue prevents phosphorylation and interferes with MAPK and ras-stimulated transcription via ER in HeLa and MCF-7 cells (13, 26). Substitution of the serine 236 residue with a negatively charged glutamic acid at position 236 mimics a phosphorylated residue and prevents receptor dimerization (27) and was mutated using the following nucleotides: 5'-gataaaaaacaggaggaaggagtcccaggcctgccggctc-3' and 5'-gagccggcaggcctgggactccttcctcctgtttttatc-3'.
Cell culture and transfection
Ishikawa cells are an endometrial tumor cell line; we are unable to detect ER or 5 by Western blot analysis, and stimulation of transcription in response to E2 requires transfection of ER. Cells were maintained in modified Eagle’s medium supplemented with 10% fetal bovine serum. Ishikawa cells were plated in six-well plates with DMEM containing no phenol red and supplemented with 5% stripped newborn calf serum to a final concentration of 4 x 105 cells/well. Serum and media were from Mediatech, Inc. (Herndon, VA). Cells were transfected using the calcium phosphate method for 16 h as previously described (28). Amounts of transfected hER, 5, and 5 mutants, or coactivator SNURF are indicated in figure legends. After transfection, cells were washed with PBS (pH 7.4) and treated with vehicle, 10 nM E2, 1 μM 4-hydroxytamoxifen (4-OHT), or 1 μM genistein (G), all from Sigma (St. Louis, MO), for 24 h. In EGF (100 ng/ml) experiments, cells were incubated in DMEM without phenol red or serum for 6–7 h before the addition of the peptide. The model ERE reporter plasmid (0.5 or 1 μg/well in all experiments) used in transfection experiments was (ERE) 2-vit-TK-luc, containing 2 EREs from the vitellogenin 2A promoter fused to the –105-bp thymidine kinase promoter (Vit-6) (29). Plasmid constructs containing the 963 bp of the cyclin D1 gene promoter (30) and 1.8 kb of the human complement 3 (C3) promoter gene promoter (31) fused to luciferase were used to determine the effect of 5 on E2-sensitive, physiological promoters. Each well was transfected with 0.5 μg ERE reporter as indicated and specified concentrations of the expression vectors CMV hER and CMV 5 or CMV 5 mutants. In all experiments, total DNA added was normalized with the control expression vector pcDNA 3.1.
Western blot analysis
For immunoblotting of Ishikawa cell extracts, cells were plated as above, and each well was collected in 1x gel loading buffer (50 mM Tris-Cl, pH 6.8; 2% sodium dodecyl sulfate (SDS), 10% glycerol) and protein concentration determined by BCA assay (Pierce, Rockford, IL). A total of 30 μg of each protein sample, was electrophoresed on a 12% SDS-containing polyacrylamide gel and transferred to a nitrocellulose membrane. The membranes were blocked in 15% nonfat dried milk in 1x Tris-buffered saline, 0.1% Tween 20, for 1 h. Blots were incubated with 1D5, an N-terminal hER monoclonal antibody (Dako, Carpinteria, CA) at a 1:100 dilution overnight, at 4 C. This incubation was followed by goat antimouse secondary antibody (1:15,000, Amersham; Buckinghamshire, UK) overnight. To normalize ER levels for total protein expression in the samples, membranes were incubated with Restore blot stripper (Pierce) for 30 min and probed with anti-glucose-6-dehydrogenase (Sigma) at a 1:1,000 dilution overnight. Incubation was continued with horseradish peroxidase (HRP)-conjugated antirabbit secondary antibody (Amersham, 1:20,000). Immunocomplexes were visualized using an enhanced chemiluminescence detection system (West Pico, Pierce). Cos-1 cells, an ER negative green monkey kidney cell line, were maintained in DMEM supplemented with 10% newborn calf serum. To validate protein expression of CMV hER, CMV 5, and CMV 5, these constructs (1 μg each) were transiently transfected into Cos-1 cells using calcium phosphate as above. Cell extracts were prepared and protein analyzed (30 μg/sample) as above. On each gel, lanes containing in vitro translated ER and 5 were included as positive controls.
To measure SNURF protein levels, Ishikawa cell lysate or primary endometrial tumor lysate was electrophoresed on a 12% denaturing polyacrylamide gel, transferred to a nitrocellulose membrane, and blocked as described above. Gels also contained equal amounts of protein extracts from Cos-1 cells or Cos-1 cells transfected with CMV-SNURF, used as negative and positive controls, respectively (32). Blots were incubated with the SNURF antibody K7979 (kind gift from J. Palvimo, Institute of Biomedicine, University of Helsinki, Helsinki, Finland) at a 1:20,000 dilution overnight, at 4 C, followed by a 60-min incubation with HRP-conjugated antirabbit secondary antibody (Amersham, 1:10,000). To normalize ER levels for total protein expression in the samples, membranes were incubated with Restore blot stripper (Pierce) for 30 min and probed with anti-?-actin (Sigma) at a 1:100,000 dilution overnight. Incubation was continued with HRP-conjugated antimouse secondary antibody (Amersham, 1:50,000).
Glutathione-S-transferase (GST) constructs and pull-down studies
GST-SNURF was kindly provided by J. Palvimo (7). Briefly, RNA was isolated from MCF-7 cells using the following primers: 5'-atggcccagaacttgaagg-3' and 5'-cctcatcaagggraagaaatga-3'. The PCR product was sequenced and subcloned into the EcoR1 site of the pGEX-2T GST fusion vector (Amersham). Proteins were isolated from BC21 cells inoculated with the expression vector for GST, GST-SNURF, or GST-REA. The transformed cells were induced with 100 nM isopropyl ?-D-thiogalactopyranoside and harvested. Bacterial lysate was incubated with glutathione beads (Sigma) at 4 C overnight. Proteins were resolved on a 12% denaturing acrylamide gel, and protein yield was determined by Coomassie staining. Assays were performed as previously described (27). Cell-free translation (TNT Rabbit Reticulocyte Transcription/Translation Kit, Promega; Madison, WI) of ER and 5 was performed in the presence of [S35] methionine (0.044 mCi/50 μl reaction; PerkinElmer, Boston, MA). [S35] methionine-labeled (0.02 mCi/150 μl reaction) ER and 5 were incubated with 1.0 μg GST or GST fusion protein in the presence of 10 μg BSA. Total reaction volume was adjusted to 150 μl with GST wash buffer (10 mM MgCl2, 150 mM KCl, 20 mM HEPES, 10% glycerol, and 0.12% Nonidet P-40). Samples were incubated for 90 min at 4 C. Proteins were electrophoresed on a 12% denaturing acrylamide gel. Gels were exposed to film at –80 C for 48 h. All chemicals were from Sigma.
Statistical analysis
Transfections were performed in triplicate, and each experiment was performed a minimum of three times. Western blots and GST pull-downs were performed at least three times. Data are expressed as mean ± SE of the mean. Data were analyzed by one-way ANOVA to determine differences among the means. Where differences were detected, a Tukey’s post hoc test was used to determine where the means were significant. Values were considered significant at P values 0.05. Analyses were performed using GraphPad Prism software, version 4 (GraphPad, San Diego, CA).
Results
Structure and expression of ER, 5, and 5 mutants
The functional protein domains and exons of ER and 5 are shown in Fig. 1A. Subsequent expression of ER, 5, and mutant proteins was evaluated in Cos-1 cells and Ishikawa cells by immunoblotting. Neither ER nor 5 proteins were detected by this method in Cos-1 cells or in our Ishikawa cells (Fig. 1B). Transiently transfected receptor constructs are easily detected, and expression of 5 does not appear to influence expression of ER protein. We observed equal expression of wild-type and mutant forms of 5 protein from equal amounts of transfected plasmid. (Fig. 1C); thus, potential differences in transcriptional response with these proteins will not be due to different levels of expressed 5 proteins.
5 Stimulates basal expression of an ERE-containing promoter but decreases E2 responses in the presence of ER
Because significant levels of 5 protein are found in endometrial tumors along with full-length ER, we examined the effects of 5 expression alone and on full-length ER transcriptional activity in a tumor cell line of endometrial origin. Untransfected Ishikawa cells express low basal and negligible E2-stimulated activity of an ERE reporter (Vit-6, Fig. 2A). In Ishikawa cells transfected with ER or 5 alone, 5 exhibits basal transcriptional activity on an ERE promoter that is significantly increased relative to no receptor or to ER alone (Fig. 2A). However, only full-length ER is responsive to E2 stimulation. To test the influence of 5 on ER transcription, cells were transfected with 0.5 μg ER expression vector and cotransfected with a 5 expression vector in increasing ratios relative to ER. In the presence of ER plus 5, basal ERE reporter activity was increased relative to ER alone (Fig. 2B). Although the degree of increase varied, promoter activity with 5 plus ER was consistently higher than with ER in all experiments. In subsequent studies with the ERE reporter, ratios of ER to 5 of 1:5 and 1:7 were used. Cotransfection of Ishikawa cells with ER and 5, and subsequent treatment with E2, resulted in a 5-mediated decrease in E2-stimulated reporter activity (Fig. 2C). At a 1.0:0.1 ratio of ER to 5, there was no significant difference in E2-stimulation (8.2 vs. 8.5) from cells transfected with ER alone; but at ratios of 1.0:0.5, E2 stimulation was reduced by approximately 40–50%. At the highest amount of 5 transfected, stimulation of ERE reporter activity was reduced by approximately 75% compared with ER alone. Thus, 5 exerts stimulatory effects on basal ERE reporter activity in the absence of ligand; but in the presence of full-length ER and 5, the response to E2 was reduced.
FIG. 2. 5 Stimulates basal ERE activity alone or with ER, and reduces E2 stimulation in the presence of ER in Ishikawa cells. A, Ishikawa cells were transiently transfected with 0.5 μg hER or increasing amounts of 5, along with 1 μg Vit-6-luciferase reporter, and treated for 24 h with vehicle or 10 nM E2. After collection of lysate, luciferase and protein assays were performed. Data are expressed as normalized luciferase activity and the mean ± SEM for three experiments with three samples per group. B and C, Ishikawa cells were cotransfected with 0.5 μg ER and 5 in increasing ratios of ER to 5 (1:0.1, 1:0.5, 1:1, 1:3, 1:5, and 1:7). Total transfected DNA was normalized using the expression vector pcDNA 3.1. After transfection for 16 h, cells were treated with 10 nM E2 for 24 h. Cells were collected and luciferase activity normalized as in 2A. B, Basal (vehicle-treated) values for each group. Figures above the bars are the fold-stimulation of vehicle-treated activity compared with ER alone. C, Both basal and E2-treated normalized luciferase values. Numbers above the black bars are the fold E2-stimulated values for each transfection group. Values represent the mean ± SEM of three experiments with three samples per group. In each panel: *, P < 0.01 compared with that treatment group’s basal activity. For panel B: *, P < 0.01 compared with activity with ER alone. a, P < 0.01 compared with the basal (untreated) activity of ER alone.
5 Responds to EGF and decreases ER responses to SERMS but not to EGF
We next tested whether 5 could also affect the ER transcriptional response to SERMS and growth factors. Ishikawa cells were transfected with ER alone, 5 alone, or ER and 5 and treated with 10 nM E2, 1 μM 4-OHT (an E2 agonist in uterine tissue), or 1 μM G (a phytoestrogen), or 100 ng/ml EGF (Fig. 3). All data are plotted relative to their own receptor (ER, 5, or ER plus 5) controls. However, relative to reporter activity in the absence of receptor, vehicle-treated promoter activity in the presence of ER, 5, or ER plus 5 was 3.3-, 4.4-, and 6.8-fold over reporter alone, respectively.
FIG. 3. 5 Is stimulated by EGF and reduces the ER transcriptional response to SERMS but not to EGF. Ishikawa cells were transiently transfected with 0.5 μg ER alone, 0.5 μg 5 alone, or ER and 5 (ratio of 1:7) and 1 μg of the Vit6-luciferase reporter. Transcriptional activity was assessed after 24 h treatment of transfected cells with 10 nM E2, 1 μM 4-OHT, 1 μM G, or 100 ng/ml EGF. Fold stimulation in the absence and presence of SERMS is shown. Values represent mean ± SEM of four experiments in each panel (three experiments in cells transfected with 5 alone). *, P < 0.01, treatment compared with that treatment group’s basal activity.
E2, 4-OHT, G, and EGF increased transcriptional activity of ER alone in Ishikawa cells. In cells transfected with 5, only EGF stimulated 5-mediated activity of the ERE reporter, approximately 1.4-fold. Cotransfection of 5 increased basal transcription from the ERE-containing promoter 4.4-fold, but lessened stimulation of the ER by E2 from 2.9- to 1.4-fold, and suppressed stimulation by both SERMS to nonsignificant levels. In contrast, the ER response to EGF was not reduced in the presence of 5 (Fig. 3B). Therefore, 5 may act generally to decrease the transcriptional response to ligands interacting with the ER at the ligand-binding domain (LBD) but allows stimulation by growth factors.
DNA binding and dimerization is required for 5 to increase ERE basal activity, and DNA binding is required to decrease E2 responsiveness in the presence of ER
Because E2 regulates many physiological events in the uterus, we focused our studies on the mechanism of 5 actions on ER in the absence and presence of E2 (Fig. 4A). To determine which regions of the 5 protein mediate the effects on ER activity, mutations were introduced into critical regions of the protein required for phosphorylation in the N-terminal region, DNA binding, and dimerization.
FIG. 4. Mutants of 5 differentially affect protein function. A, Dimerization and DNA binding of 5 is required for full stimulation of basal activity with ER, and DNA binding is required for 5 to reduce E2 stimulation with ER. Ishikawa cells were transiently transfected with 0.5 μg hER, 5 or 5 mutants alone, or 0.5 μg ER cotransfected with 5 at a 1:5 ratio. Mutants included: 5 S118A (phosphorylation mutant); 5 S235E (dimerization mutant), 5 E203G/G204S (DNA binding mutant). Each 5 mutation at the amino acid group is defined in the 5 mutant-alone groups and is abbreviated by the amino acid numbers in the ER plus 5 groups. Total transfected DNA was normalized using the expression vector pcDNA 3.1. Data are expressed as fold-stimulation relative to cells transfected with ER alone, with this value set at 1.0. Values represent mean ± SEM of six experiments, with three samples per group. *, P < 0.01, E2 compared with ERE basal activity in each group. a, P < 0.01, basal ERE activity compared with that with ER alone. B, Ishikawa cells were transiently transfected with reporter alone, 0.5 μg hER alone, 5 or 5 S118A alone, or ER with cotransfected 5 or 5 S118A in a 1:5 ratio. Cells were treated with vehicle or 100 ng/ml EGF for 24 h. Some cells were transfected with reporter (ERE), in the absence of any cotransfected receptor, and treated with or without EGF. *, P < 0.01 E2 compared with ERE basal activity in each group. a, P < 0.05, basal ERE activity compared with that with ER alone. #, P < 0.05, basal ERE activity compared with that of 5 alone.
In Ishikawa cells transfected with 5 S118A (phosphorylation mutant of 5) alone, basal activity of the ERE reporter was significantly increased compared with cells transfected with ER alone, and similar to cells transfected with 5 alone; transcriptional activity did not increase in the presence of E2. In Ishikawa cells cotransfected with ER and 5 S118A, basal activity was significantly increased compared with cells transfected with either ER or 5 alone. In the presence of E2, no significant increase in transcriptional activity was observed. Wild-type 5 and 5 S118A behave similarly when transfected into Ishikawa cells, indicating that the effects of 5 on basal and E2-stimulated ER activity are not mediated at the serine 118 site.
In Ishikawa cells transfected with only the dimerization mutant of 5, 5 S236E, basal activity on an ERE was not increased and was similar to that of cells transfected with ER alone; stimulation by E2 was not observed. Cotransfection of cells with ER and 5 S236E resulted in basal ERE activity that was greater than with ER alone but was significantly diminished compared with cells transfected with ER and wild-type 5. Reporter activity in cells transfected with ER plus 5236E was only slightly increased in the presence of E2. These data indicate that 5 must dimerize for full stimulation of basal ERE activity in the presence of ER plus 5. However, 5 dimerization is not required for the suppression of E2-stimulated ERE reporter activity, because this still occurs with this mutant.
In Ishikawa cells transfected with only the DNA binding mutant of 5, 5 E203G/G204S, basal activity was similar to that of cells transfected with ER alone, and stimulation by E2 was not observed. Cotransfection of cells with both ER and 5 E203G/G204S exhibited similar basal activity compared with cells transfected with ER alone. Reporter activity was significantly increased in the presence of E2. Thus, the 5-induced increase in basal ERE reporter activity is mediated, at least in part, by DNA binding and dimerization. The suppression of the E2-ER transcriptional response by 5 also appears to be mediated by DNA binding.
Serine 118 is required for EGF stimulation
We next examined the contribution of serine 118 to the EGF response by testing the phosphorylation mutant S118A (Fig. 4B). Serine 118 has previously been shown to be required for growth factor and ras-stimulated transcription of ER in HeLa cells (13). In Ishikawa cells transfected with the ERE reporter alone, EGF treatment did not stimulate luciferase activity. 5 significantly stimulates basal transcription from an ERE and is stimulated by EGF, whereas 5 S118A alone increases basal transcription from an ERE but is not stimulated by EGF, indicating this residue is important for stimulation to occur. In cotransfection experiments, the presence of 5 did not prevent significant stimulation of ER by EGF, but cotransfection with 5 S118A did. Thus, S118A is not required for effects on basal transcription but is required for growth factor stimulation.
The coactivator SNURF enhances effects of 5 on basal ERE activity
The ability of 5 to increase basal ERE activity, alone and with ER, may occur through several mechanisms. One possibility may be recruitment of coactivators that interact with the ER N terminal, hinge, or DBD regions that are retained in 5. One such coactivator is SNURF, which augments stimulation of ER by E2 in ZR-75 human breast cancer cells via interaction at the hinge and DBD regions of the receptor (8). To test possible effects of SNURF, Ishikawa cells were cotransfected with ER and 5 in the presence or absence of SNURF (Fig. 5). SNURF increases the basal activity of both ER and 5 on an ERE. SNURF alone had no effect on the promoter construct (not shown). SNURF does not preferentially increase the ER response to E2; however, because the basal activity in the presence of SNURF was higher, the overall activity in the presence of E2 plus SNURF was significantly increased. As in previous experiments, 5 increased basal ERE activity in the presence of ER. SNURF further increased basal activity. Thus, this coactivator would favor increased basal expression from an ERE and decreased fold-stimulation by E2 in the presence of 5 plus ER.
FIG. 5. The coactivator SNURF enhances the effect of 5 and ER on ERE transcription. Ishikawa cells were cotransfected with 0.5 μg ER and 2.5 μg 5, 1 μg Vit6-luciferase, and either 0.5 or 1.0 μg SNURF. Total transfected DNA was normalized using the expression vector pcDNA 3.1. Luciferase activity was determined after 24 h treatment with vehicle (white bars) or 10 nM E2 treatment (black bars). Experiments were performed five times with three samples per group. Shown is a representative experiment in which transfections were performed in triplicate. *, P < 0.05, E2 compared with vehicle in each group. a, P < 0.05, basal ERE activity compared with that with ER alone. #, P < 0.05, comparing basal ERE activity in ER/5 cotransfected cells with and without SNURF. SNURF also significantly stimulated basal activity of the reporter in the presence of ER, 5, or ER plus 5 (P < 0.01).
SNURF is expressed in uterine cells and interacts directly with 5
We examined SNURF expression in our clone of Ishikawa cells and in whole-cell lysate obtained from primary endometrial tumors. Because Cos-1 cells do not express endogenous SNURF, they were transfected with SNURF, and lysates from control and transfected cells were used as positive and negative controls. Figure 6A shows that SNURF is expressed in Ishikawa cells and endometrial tumor tissue. The doublet of SNURF seen in these tissues is observed for endogenous SNURF expression in other tissues (32) and may represent splice variants or posttranslational modifications of SNURF.
FIG. 6. SNURF is expressed in uterine cells and binds both ER and 5. A, SNURF is expressed in uterine cells. One hundred micrograms of Ishikawa cell lysate or endometrial tumor lysate were electrophoresed on a denaturing gel, transferred to nitrocellulose membrane, and immunoblotted for SNURF. Cos-1 cells do not express endogenous SNURF (negative control); Cos-1 cells transiently transfected with SNURF were used as a positive control. A representative blot of three separate experiments is shown. B, ER and 5 bind the coactivator SNURF. Approximately 1 μg GST or GST-SNURF fusion protein was incubated with [S35] methionine labeled ER or 5 for 90 min. Beads were washed, and bound proteins were eluted and electrophoresed on a 12% SDS-containing polyacrylamide gel. Migration of proteins bound to GST or GST-SNURF are indicated by arrows and visualized by autoradiography. Input in vitro translated (IVT) 5 and ER is shown and represents 1/8 input of radiolabeled protein in each reaction. A representative autoradiogram from six experiments is shown.
Although SNURF effects on ER transcription require the hinge and DBD regions of ER, direct SNURF binding to the receptor had not been demonstrated (8). Because the coactivator SNURF modulates 5 activity in our cell culture system, we investigated direct protein interactions between SNURF and ER and SNURF and 5 in GST pull-down experiments. GST and GST-SNURF fusion proteins were incubated with radiolabeled wild-type ER and 5. Minimal interaction was observed between GST and ER and GST and 5. GST-SNURF bound both ER and 5 in pull-down experiments (Fig 6B). Thus, the regions contained in the 5 protein are sufficient for interactions with SNURF, and SNURF could act to increase the basal activity of ER and 5-containing dimers by binding directly to the proteins.
5 Expression modulates the activity of complex physiological E2-sensitive promoters
To test whether physiologically E2-sensitive promoters in the uterus could be regulated by 5, we tested several complex physiological promoters, including the promoter of the cell cycle regulator and protooncogene cyclin D1 and the complement 3 promoter. E2 sensitivity of the cyclin D1 promoter is not mediated by a canonical ERE, but by a proximal cAMP response element and an AP-1 site (33, 34). Cyclin D1 promoter activity was significantly stimulated by E2 in cells transfected with ER. Basal activity of the cyclin D1 promoter was only slightly stimulated by 5 alone. However, there was up to a 1.7-fold increase in basal promoter activity in cells cotransfected cells with ER plus 5 compared with basal activity in cells transfected with ER alone. Under these conditions, the promoter was no longer stimulated significantly by E2. The E2 stimulation of the C3 promoter is mediated via three responsive elements, including one noncanonical ERE and two additional elements with no significant homology to known EREs (31), and C3 is regulated in uterus by E2 and during the menstrual cycle (35). In cells transfected with the C3 reporter construct and ER, the activity of the promoter was significantly stimulated (3.6-fold) by E2. Transfection with 5 alone stimulated basal activity of the promoter 1.5-fold, and cotransfection of 5 plus ER increased basal activity of the C3 promoter compared with either ER (2.2-fold) or 5 alone (2.7-fold). Increasing 5 resulted in a decreased E2 stimulation of promoter activity, as for cyclin D1. In both cases, overall promoter activity did not increase with increasing 5; thus, the net effect was increased promoter activity that did not require E2. These data indicate that, in addition to model reporters, 5 can also regulate the activity of physiologically relevant promoters.
Discussion
Splice variants of ER mRNA are commonly transcribed in normal and cancerous tissues (36) but are rarely translated into protein. Three notable exceptions are: the presence of 4 in primary ovarian tumors (20), 7 in primary tumors of the breast and in breast cancer cell lines (21), and 5 in breast (22) and primary endometrial tumors (23). Our studies focus on the activity of 5 in Ishikawa cells, an endometrial cancer cell line, and some of the transcriptional consequences for E2-regulated genes. We demonstrated previously that 5 mRNA is translated to protein in primary endometrial tumors but is not expressed in normal uterine tissue (23). This is physiologically significant, because activity of a functional 5 protein may affect the expression or activity of the wild-type receptor. Indeed, it has been shown that 5 modulates the activity of ER in a cell-specific manner. For example, in osteosarcoma cells stably transfected with ER, 5 increases both basal and E2-stimulated activity as well as cellular proliferation (18). In MCF-7 breast cancer cells stably transfected with 5, the variant did not modulate E2 responsive genes, such as the progesterone receptor, but did stimulate growth of transfected cells (19). In HeLa cells and breast epithelial cells, 5 had a dominant negative effect on ER stimulation of a reporter gene (14, 24). The varied effects of 5 indicate that its effects are dependent on cell and promoter context.
We investigated the activity of 5 alone and the ability of 5 to modulate ER activity using transient transfection studies in Ishikawa cells. We found that 5 alone consistently increased basal activity of the ERE-containing model promoter, as well as the basal activity of the physiological promoter, C3. The cyclin D1 promoter was not significantly stimulated by 5 alone, perhaps because it responds to E2 via an AP1 site and tethered proteins rather than by direct receptor binding to DNA. As expected, 5-mediated activity was not modulated by E2; however, EGF stimulated the activity of 5 alone. Basal activity mediated by 5 required both DNA binding and dimerization, and EGF stimulation required an intact serine 118 residue. This suggests that phosphorylation of this residue may be important for the EGF effect on 5, as has been reported previously for full-length ER (13, 37).
In endometrial tumors, 5 expression occurs in the presence of full-length ER. Therefore, we examined the influence of 5 on ER activity. In the presence of ER, addition of 5 increases basal activity of an ERE-containing promoter, as well as two physiological promoters, in a dose-dependent manner (Figs. 2B and 7). Moreover, coexpression of 5 plus ER increased reporter activity more than either receptor alone. The magnitude of this increase varied, perhaps as a result of differing expression of coactivator, corepressor, or other integrator proteins that modulate ER function. Similar to effects of 5 alone, increased basal activity with ER plus 5 required 5 DNA binding and dimerization functions of 5 but not serine 118. Collectively, these data indicate that 5 is indeed able to dimerize with itself and/or ER and to occupy an ERE to alter the transcription of E2-responsive genes in the absence of ligand. Our results on the importance of DNA binding for 5 function in a physiological system are supported by published data demonstrating binding of 5, with and without ER, to DNA in gel shift assays (14, 18). Our data indicate that 5 behaves in uterine cells in a dominant positive manner. This is in agreement with results from osteosarcoma cells (18), but in contrast with results in breast cells and HeLa cells (14, 24). This is unlikely to depend exclusively on any one specific physical property of 5 but also on the presence or activity of other proteins that modulate receptor activity. However, mutational analysis of 5 was not performed in the context of these other cell types, and direct comparisons cannot be made.
FIG. 7. 5 Modulates activity of the cyclin D1 promoter and the complement 3 promoters. A. Cells were transfected with 0.5 μg cyclin D1 promoter-luciferase construct and 2.5 or 3.5 μg 5 in the absence or presence of 0.5 μg ER and treated with 10 nM E2 for 12 h. *, P < 0.05, E2 stimulated activity compared with each group’s vehicle-treated control. a, P < 0.05, basal activity compared with basal activity of the cyclin D1 promoter in the presence of ER. B, Cells were transfected with 0.5 μg complement 3 promoter-luciferase construct and 2.5–3.5 μg 5 in the absence or presence of 0.5 μg ER and treated with E2 for 24 h. *, P < 0.001, E2-stimulated activity compared with group’s vehicle-treated control. a, P < 0.001, basal activity of the C3 promoter in the presence of ER plus 5 compared with ER alone. #, P <.05, basal activity of the C3 promoter with no receptor vs. with 5 alone.
In Ishikawa cells, we also observed that 5 reduced the fold stimulation of the ERE reporter by ER and E2 (Fig. 2). Similar effects were observed with the complex physiological promoters for cyclin D1 and C3. The effects of 5 on ligand-stimulated ER transcriptional activity occur more generally, because 5 expression also reduced the transcriptional response to the SERMS tamoxifen and G (Fig. 3). DNA binding of 5 is required for this effect (Fig. 4A). Because 5 does not contain an LBD, it cannot bind or respond to ligand. Thus, increased occupancy of the ERE by 5 or 5-ER dimers would result in a decreased response to ligand.
In contrast, the presence of 5 does not prevent ER-mediated stimulation of an ERE-containing reporter by EGF (Fig. 4B). Serine 118 was required for this response, because the 5 phosphorylation mutant could not be stimulated by EGF, and the mutation prevented EGF stimulation of the ER/5 complex. The magnitude of the response to EGF by ER alone or ER plus 5 was constant. However, because basal expression from an ERE reporter was increased, the total EGF response was greater in cells transfected with ER plus 5. Several groups have shown that ER can be activated by various growth factors, including IGF-1, EGF, TGF and heregulin, as well as signaling intermediates of the MAP kinase pathway (10, 11, 37, 38). EGF stimulates ER activity in HeLa cervical cancer cells (13) in the mouse uterus (39), and in Ishikawa cells (40), and this occurs primarily through the AF-1 region (13, 40). Because the 5 protein contains the AF-1 and the DBD, dimers containing 5 can respond to EGF in this cell line.
The dominant positive effect of 5 in endometrial carcinoma cells could occur through several mechanisms, including preferential expression or altered activity of specific coactivators, corepressors, or integrator proteins that stabilize DNA binding of complexes containing 5 and/or favor expression of AF-1 activity. Ishikawa cells express high levels of some coactivators and other regulatory molecules relative to MCF7 cells (41). The ratio of coactivators and corepressors in uterine cells contributes to the transcriptional responses to tamoxifen, which suppresses ER AF-2 activity but allows responses mediated by AF-1 (42). A second contributing factor to 5 activity may be the absence of the AF-2 region, which might release constraints in the full-length receptor and allow maximum expression of AF-1 activity. Finally, many corepressors, such as nuclear receptor corepressor (NCoR), silencing mediator of retinoic acid and thyroid hormone receptors (SMRT), and REA, preferentially bind to the C-terminal LBD, and this is missing in 5 (9, 43). Because REA can bind to ER in the presence of E2 and compete with coactivators for binding to the LBD, it may exert major effects on gene activity. A recent report demonstrated that mice in which the REA gene is disrupted have significant changes in endogenous uterine genes, and C3 expression is stimulated (44). Alternatively, 5 may titrate away an unknown repressor acting via the N terminal and stimulate activity of both ER and 5. Because the effects of 5 plus ER on basal activity are greater than either receptor alone, a combination of these or other possibilities may occur.
We hypothesized that dimers containing 5 would bind a particular class of coactivators and stimulate activity of this complex. Several coactivators have been reported to activate ER through the N-terminal AF-1 region. P300 binds to AF-1 and mediates synergism between AF-1 and AF-2 (45). CBP can bind to the full-length ER in the absence of E2 and stimulates AF-1 activity on an E2-sensitive promoter specifically through serines 104/106/118 (46). However, these studies tested protein-protein interactions via yeast or mammalian two-hybrid approaches, and the entire ER molecule was required for most responses. Other coactivators such as SRA (47), ini (48), arginine methyltransferase (49), and SNURF (8) have also been reported to stimulate transcription of E2-responsive genes via the AF-1 or DBD region of ER. Modulation of 5 activity by these and other coactivators could play a role in this system.
We tested one coactivator, SNURF, and its ability to modulate 5 activity. SNURF acts via the hinge and DBD of the androgen receptor (7), and stimulates ER activity in breast cancer cells through multiple regions, including the DBD and hinge (8). SNURF is expressed in Ishikawa cells and in human endometrial tumors (Fig 6), and increased basal activity of an ERE-containing promoter, in the presence of ER, 5, and particularly of ER plus 5 (Fig. 5), up to 10-fold. In the presence of SNURF, ERE-promoter reporter activity in cells transfected with ER plus 5 was increased, but the response to E2 was reduced. This may be because SNURF favors binding of 5, in homo- or heterodimers to DNA, and this variant form cannot respond to E2. Alternatively, SNURF binding to ER-5 heterodimers may alter or inhibit subsequent recruitment of coactivators or regulatory proteins in the presence of E2. These results are somewhat divergent from those of Saville et al. (8), who showed that SNURF amplifies the E2-stimulated (up to 5-fold) but not basal ER transcription of the simple ERE-TATA box reporter in breast cancer cells. Some differences in results between the two cell systems may arise because of the use of the simpler promoter vs. a more complex thymidine kinase promoter we used in our studies. The results may also be due to the preferential expression of AF-1 activity in uterine cells and the ability of SNURF to enhance basal expression through this pathway. Overall expression of SNURF may also play a role in the response, but this was not tested in the breast cancer cell studies. Using GST pull-down analysis, we demonstrated, for the first time, direct interaction between SNURF and ER, and SNURF and 5 (Fig. 6). This interaction is likely through the DBD, because SNURF binds to the DBD of several nuclear receptors (7, 8), and the DBD is contained in both ER and 5. Thus, stimulated basal activity from an ERE-containing promoter in the presence of 5 may occur through preferential recruitment of specific coactivators.
Modulation of ER activity by the 5 ER variant is important because the preferential expression of 5 in endometrial cancer may have significant physiological implications (23). In endometrial tumors, expression of variant protein occurs in addition to overall decreases in full-length ER and progesterone receptor expression (50). In Ishikawa cells, the expression of 5, in addition to full-length ER, increases the expression of ER-sensitive genes in the absence of ligand and reduces the response to ligand and SERMS. Therefore, endometrial tumors expressing 5 might be less responsive to antiestrogens, despite a steroid receptor profile that classifies it as being ER positive. It could thus be helpful to consider the expression of ER variants in clinical methods used to determine steroid receptor profiles. Because expression of 5 allows growth factor stimulation of the receptor, the growth factor pathway may become more influential in the regulation and maintenance of uterine tumors. In the case of endometrial cancers that express 5, compounds that interfere with the ability of 5 either to dimerize and bind DNA or otherwise disrupt signaling may prove useful.
Acknowledgments
The authors thank Dr. Jorma Palvimo (Institute of Biomedicine) for the SNURF expression vectors.
References
Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20:358–417
Nilsson S, Makela S, Treuter E, Tujague M, Thonsen J, Andersson G, Enmark E, Pettersson K, Warner M, Gustafsson J-A 2001 Mechanisms of estrogen action. Physiol Rev 81:1535–1565
Onate SA, Tsai SY, Tsai M-J, O’Malley BW 1995 Sequence and characterization of a co-activator for the steroid hormone receptor superfamily. Science 270:1354–1357
McInerney EM, Tsai MJ, O’Malley BW, Katzenellenbogen BS 1996 Analysis of estrogen receptor transcriptional enhancement by a nuclear hormone receptor coactivator. Proc Natl Acad Sci USA 93:10069–10073
Hanstein B, Eckner R, DiRenzo J, Halachmi S, Liu H, Searcy B, Kurokawa R, Brown M 1996 P300 is a component of an estrogen receptor co-activator complex. Proc Natl Acad Sci USA 93:11540–11545
Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon H, Montiminy M, Evans RM 1996 Role of CBP/P300 in nuclear receptor signaling. Nature 383:99–103
Moilanen AM, Poukka H, Karvonen U, Hakil M, Janne OA, Palvimo JJ 1998 Identification of a novel RING finger protein as a coregulator in steroid receptor-mediated gene transcription. Mol Cell Biol 18:5128–5139
Saville B, Poukka H, Wormke M, J?nne OA, Palvimo JJ, Stoner M, Samuido I, Safe S 2002 Cooperative coactivation of estrogen receptor a in ZR-75 human breast cancer cells by SNURF and TATA binding protein. J Biol Chem 277:2485–2497
Montano MM, Ekena K, Delage-Mourroux R, Chang W, Martini P, Katzenellenbogen BS 1999 An estrogen receptor coregulator that potentiates the effectiveness of antiestrogens and represses the activity of estrogens. Proc Natl Acad Sci USA 96:6947–6952
Ignar-Trowbridge DM, Pimentel M, Parker MG, McLachlan JA, Korach KS 1996 Peptide growth factor cross-talk with the estrogen receptor requires the A/B domain and occurs independently of protein kinase C or estradiol. Endocrinology 137:1435–1744
Aronica SM, Katzenellenbogen BS 1993 Stimulation of estrogen receptor mediated transcription and alteration on the phosphorylation state of the rat uterine receptor by estrogen, cyclic adenosine monophosphate and insulin like growth factor-1. Mol Endocrinol 6:743–752
Tzukerman MT, Esty A, Santiso-Mere D, Danielian Parker MG, Stein RB, Pike JW, McDonnell DP 1994 Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol 6:1043–1050
Kato S, Endoh H, Masuhiro Y, Kitamoto T, Uchiyama S, Sasaki H, Masushige S, Gotoh Y, Nishida E, Kawashima H, Metzger D, Chambon P 1995 Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270:1491–1494
Bollig A, Miksicek RJ 2000 An estrogen receptor- splicing variant mediates both positive and negative effects on gene transcription. Mol Endocrinol 14:634–649
Hu C, Hyder SM, Needleman DS, Baker VV 1996 Expression of estrogen receptor variants in normal and neoplastic human uterus. Mol Cell Endocrinol 118:173–179
Rice LW, Jazaeri AA, Shupnik MA 1997 Estrogen receptor mRNA splice variants in pre- and postmenopausal human endometrium and endometrial carcinoma. Gynecol Oncol 65:149–157
Castles CG, Fuqua SAW, Klotz DM, Hill SM 1993 Expression of a constitutively active estrogen receptor variant in the estrogen receptor-negative BT-20 human breast cancer cell line. Cancer Res 53:5934–5939
Chaidarun SS, Alexander JM 1998 A tumor-specific truncated estrogen receptor splice variant enhances estrogen-stimulated gene expression. Mol Endocrinol 12:1355–1366
Rae D, Parker MG 1996 Effects of an exon 5 variant of the estrogen receptor in MCF-7 breast cancer cells. Cancer Res 56:1556–1563
Park W, Choi JJ, Hwang ES, Lee JH 1996 Identification of a variant estrogen receptor lacking exon 4 and its coexpression with wild-type estrogen receptor in ovarian carcinomas. Clin Cancer Res 12:2029–2035
Fasco MJ, Keyomarsi K, Arcaro KF, Gierthy JF 2000 Expression of an estrogen receptor variant protein in cell lines and tumors. Mol Cell Endocrinol 162:167–180
Fuqua SA, Fitzgerald SD, Chamness GC, Tandon AK, McDonnell DP, Nawaz Z, O’Malley BW, McGuire WL 1991 Variant human breast tumor estrogen receptor with constitutive transcriptional activity. Cancer Res 51:105–109
Jazaeri O, Shupnik MA, Jazaeri AA, Rice LW 1999 Expression of estrogen receptor mRNA and protein variants in human endometrial carcinoma. Gynecol Oncol 74:38–47
Ohllson H, Lykkesfeldt AE, Madsen MW, Briand P 1998 The estrogen receptor variant lacking exon 5 has dominant negative activity in the human breast epithelial line HMT-3522S1. Cancer Res 58:4264–4268
Mader S, Kumar V, de Verneuil H, Chambon P 1989 Three amino acids of the oestrogen receptor are essential to its ability to distinguish an oestrogen from a glucocorticoid-responsive element. Nature 338:271–274
Migliaccio A, Di Domenico M, Castoria G, de Falco A, Bontempo P, Nola E, Auricchio F 1996 Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol-receptor complex in MCF-7 cells. EMBO J 15:1292–1300
Chen D, Pace PE, Coombes RC, Ali S 1999 Phosphorylation of the human estrogen receptor by protein kinase A regulates dimerization. Mol Cell Biol 19:1002–1015
Lin VY, Resnick EM, Shupnik MA 2003 Truncated estrogen receptor product-1 stimulates estrogen receptor transcriptional activity by titration of repressor proteins. J Biol Chem 278:38125–38131
Schreihofer DA, Resnick EM, Soh AY, Shupnik MA 1999 Transcriptional regulation by a naturally occurring truncated rat estrogen receptor, truncated ER product-1 (TERP-1). Mol Endocrinol 13:320–329
Albanese C, Johnson J, Watanabe G, Eklund N, Vu D, Arnold A, Pestell RG 1995 Transforming p21 mutants and c-Ets-2 activate the cyclin D1 promoter through distinguishable regions. J Biol Chem 270:23589–23597
Norris JD, Fan D, Wagner BL, McDonnell DP 1996 Identification of the sequences within the human complement 3 promoter required for estrogen responsiveness provides insight into the mechanism of tamoxifen mixed agonist activity. Mol Endocrinol 10:1605–1616
Curtin D, Ferris HA, Hakli M, Gibson M, Janne OA, Palvimo JJ, Shupnik MA 2002 SNURF increases basal and GnRH-stimulation of the rat LH? gene and prevents androgen suppression. Mol Endocrinol 18:1263–1276
Castro-Rivera E, Samuidio I, Safe S 2001 Estrogen regulation of cyclin D1 gene expression in ZR-75 breast cancer cells involves multiple enhancer elements. J Biol Chem 276:30853–30861
Liu M-M, Albanese C, Anderson CM, Hilty K, Webb P, Uht RM, Price RH, Pestell RG, Kushner PJ 2002 Opposing action of estrogen receptors and ? on cyclin D1 gene expression. J Biol Chem 277:24353–24360
Sundstrom DA, Komm BS, Ponce-de-Leon H, Yi Z, Teuscher C, Lyttle CR 1989 Estrogen regulation of tissue specific expression of complement C3. J Biol Chem 264:16941–16947
Yang J, Liu A, Chiou SK, Guzman R, Nandi D 2000 Estrogen receptor variants are present in many normal human tissues. Int J Mol Med 3:223–227
Bunone G, Briand PA, Miksicek RJ, Picard D 1996 Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation. EMBO J 15:2174–2183
Pietras RJ, Arboleda J, Reese DM, Wongvipat N, Pegram MD, Ramos L, Gorman CM, Parker MG, Sliwkowski MX, Slamon DJ 1995 HER-2 tyrosine kinase pathway targets estrogen receptor and promotes hormone independent growth in human breast cancer cells. Oncogene 10:2345–2346
Ignar-Trowbridge DM, Pimentel M, Teng CT, Korach KS, McLachlan JA 1995 Cross-talk between peptide growth factor and estrogen receptor signaling system. Environ Health Perspect 103(Suppl 7):35–38
Ignar-Trowbridge DM, Teng CT, Ross KA, Parker MG, Korach KS, McLachlan JA 1993 Peptide growth factors elicit estrogen receptor-dependent transcriptional activation of an estrogen-responsive element. Mol Endocrinol 8:992–998
Thenot D, Charpin M, Bonnet S, Cavailles V 1999 Estrogen receptor cofactors expression in breast and endometrial human cancer cells. Mol Cell Endocrinol 156:85–93
Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of SERMs. Science 295:2465–2468
Klinge CM 2000 Estrogen receptor interaction with co-activators and co-repressors. Steroids 65:227–251
Park S-E, Xu J, Frolova A, O’Malley BW, Katzenellenbogen BS, Genetic ablation of the repressor of estrogen receptor activity (REA) modulates response to estrogen in target tissues. Program of the 84th Annual Meeting of The Endocrine Society, New Orleans, LA, 2004 (Abstract P1-149)
Kobayashi Y, Kitamoto T, Yoshika M, Michiko W, Kase T, Metzger D, Yanagisawa J, Kato S 2000 p300 mediates functional synergism between AF-1 and AF-2 of estrogen receptor and ? by interacting directly with the N-terminal A/B domains. J Biol Chem 275:15645–15651
Dutertre M, Smith CL 2003 Ligand-independent interactions of p160/steroid receptor coactivators and CREB-binding protein (CBP) with estrogen receptor-: regulation by phosphorylation sites in the A/B region depends on receptor domains. Mol Endocrinol 17:1296–1314
Deblois G, Giguere V 2003 Ligand independent coactivation of ER AF-1 by steroid receptor RNA activator (SRA) via MAPK activation. J Steroid Biochem Mol Biol 85:123–131
Oltra E, Pfeifer I, Werner R 2003 Ini, a small nuclear protein that enhances the response of the connexin 43 gene to estrogen. Endocrinology 144:3148–3158
Qi C, Chang J, Zhu Y, Yeldandi AV, Rao SM, Zhu YS 2002 Identification of protein arginine methyltransferase 2 as a coactivator for estrogen receptor . J Biol Chem 297:28624–28630
Jazaeri AA, Nunes KJ, Dalton MS, Xu MH, Shupnik MA, Rice LW 2001 Well-differentiated endometrial adenocarcinomas and poorly differentiated mixed müllerian tumors have altered ER and PR isoform expression. Oncogene 20:6965–6969(Winnifred Bryant, Allison)
Address all correspondence and requests for reprints to: Margaret Shupnik, Ph.D., Department of Internal Medicine/Endocrinology, P.O. Box 800578, Health Sciences Center, University of Virginia, Charlottesville, Virginia 22908. E-mail: mas3x@virginia.edu.
Abstract
Estrogen receptor (ER) is a ligand-inducible transcription factor that mediates the physiological effects of 17?-estradiol (E2). In the uterus, E2 is involved in tissue growth, maintenance, and differentiation. 5ER (5) is an ER variant protein expressed in uterine tumors but not in normal tissue. We examined the transcriptional activity of 5 and its modulation of human ER basal and E2-stimulated activity in Ishikawa cells, an endometrial cancer cell line. In transient transfection assays, 5 increased basal activity of an estrogen response element-containing promoter in the absence or presence of ER but lessened stimulation by ER and E2. Effects of 5 were not limited to model reporters, given that cyclin D1 and complement 3 promoters were similarly affected. Increases in basal transcription required dimerization and DNA binding of 5, whereas decreased E2 stimulation with ER required only DNA binding. Decreased ligand stimulation was not unique to E2 but also applied to the selective ER modulators tamoxifen and genistein. However, promoter stimulation by epidermal growth factor is retained with 5. The ER coactivator small nuclear ring finger protein is expressed in Ishikawa cells and uterine tumors, and it enhances effects of 5 alone and with ER on basal activity of an estrogen response element reporter. Thus, in the presence of 5 plus ER, there is a lower transcriptional response to E2 and SERMS, but stimulation by epidermal growth factor is retained. The expression of 5 in uterine carcinoma may provide a mechanism by which tumors could maintain expression of E2-responsive genes in the absence of E2.
Introduction
THE GONADAL STEROID 17?-estradiol (E2) exerts effects on differentiation, growth, and gene regulation in a number of target tissues, including the breast and the uterus (Refs.1 and 2 and references within). The physiological effects of E2 are not always beneficial, given that E2 exposure contributes to the development and growth of breast and uterine cancer (1). The estrogen receptors (ERs) and ? are ligand-inducible transcription factors that act as the primary mediators of E2 action and E2-regulated gene expression. ERs are members of the nuclear receptor superfamily, with conserved protein functional domains (1). The major conserved domains of the ER include an amino terminus activation function (AF)-1 that mediates ligand-independent transcription, and a DNA binding domain (DBD) that binds to estrogen response elements (EREs) on E2-responsive genes and contains a dimerization interface (1, 2). A central hinge region is required for nuclear localization of the ER. The carboxyl terminus is required for the binding of ligand and ligand-stimulated gene transcription via its AF-2. This domain also promotes the recruitment of coactivators and dimerization of ligand-bound ERs (1, 2).
ER activity can be modified through such mechanisms as the expression of ER isoforms, cross-talk between intracellular signaling pathways, and interaction of the ER with coregulator proteins (2). Occupancy of the ER by ligand results in receptor dimerization. Subsequent binding to an ERE in a target gene induces recruitment of various coregulators, which act to enhance (coactivators) or suppress (corepressors) gene expression. The complex of ligand-bound dimers and coregulators contacts the general transcriptional machinery to form a preinitiation complex for transcription. Many coactivators, including SRC-1 (3, 4), cAMP response element binding protein-binding protein (CBP) (5, 6), and small nuclear ring finger protein (SNURF, also known as RF4) (7, 8), stimulate transcription mediated by ER as well as other steroid receptors. Several proteins suppress ER transcription, but to date, the only ER-specific corepressor is the repressor of estrogen activity (REA), which competes with coactivators for binding to ligand-bound ER (9) and consequently suppresses ER-mediated transcription.
In vitro and in vivo evidence support the idea that ER-mediated transcription is also modulated by numerous growth factor signaling pathways (10, 11, 12, 13) and that this requires the N-terminal AF-1 domain (10, 11). The contribution of AF-1 to ER activity is dependent on cell and promoter context and plays an important role in ER activity in uterine cells (12). Direct phosphorylation of serine 118, a highly conserved amino acid residue, via MAPK, has been shown to mediate AF-1 activity in some cell types (13).
Alternate splicing of ER mRNA can produce transcripts that lack one or more exons (14). In normal and neoplastic uterine tissue, expression of full-length ER mRNA is generally greater than the mRNA expression of a given splice variant (15, 16). Occasionally, ER mRNA splice variants are translated into proteins that may exhibit activity that is markedly different from that of wild-type ER (14, 17, 18, 19, 20, 21, 22). These variants have the potential to act alone or in a dominant-positive or -negative fashion and alter the activity of the wild-type ER.
5ER (5) is an ER mRNA variant that was originally identified in primary breast cancer tumors that were classified as ER–/PR+ (22). 5 is also the only ER mRNA splice variant that is translated into protein in primary uterine tumors (23). The deletion of exon 5 produces a frameshift in the coding region between exons 4 and 6, and results in a premature stop codon. The 42-kDa truncated protein is composed of the AF-1, DNA binding, and hinge domains, as well as a unique sequence of eight amino acids at the carboxyl terminus (Fig. 1A). 5 also retains one of two dimerization interfaces and the nuclear localization signal of the ER. The ligand-binding and AF-2 regions are completely absent. 5 activity and its ability to act in a dominant-negative or -positive way with ER is dependent on cell type. For example, in HMT-3522S1 cells, an ER-negative human breast epithelial cell line, transient transfection with ER and 5 decreased the basal and E2-stimulated activity of ER (24). On the other hand, in yeast transfection systems (22), 5 alone had constitutive basal activity on an ERE reporter. 5 acts cooperatively with ER in osteosarcoma cells to significantly increase both the basal and E2-stimulated transcription of an ERE reporter in a dose-dependent manner (18). In MCF-7 cells, an ER-positive breast cancer line, 5 had no effect on ERE reporter activity (19).
FIG. 1. Functional protein domains of ER and 5 and Western blot analysis of hER and 5 expression in Cos-1 and Ishikawa cells. A, Exon and functional domains of ER and 5. Black lines under each cartoon represent the dimerization interfaces. The truncated splice variant 5 lacks a majority of the ligand-binding and AF-2 regions. Deletion of exon 5 results in a frameshift mutation and the production of a premature stop codon (horizontal lines). Diagonal stripes indicate the unique carboxyl terminus in 5. Arrows indicate sites of substituted amino acids residues in the 5 protein. Cos-1 cells were transiently transfected with ER, 5, ER and 5, or 5 mutants. B, Protein samples (30μg) from Cos-1 (left panel) or Ishikawa cells (right panel) were resolved on a 12% acrylamide gel. Blots were probed with 1D5, an N-terminal ER antibody, to visualize ER and 5. Blots were also probed with antibody for glucose-6-dehydrogenase (G6PDH) to confirm equal loading of protein. Arrows show the migration of ER and 5 proteins. C, Proteins (30 μg) from cells transfected with either wild-type or mutant 5 constructs were analyzed for receptor expression as above. Shown are representative blots from four to seven experiments.
We previously reported that primary endometrial carcinoma tumors express appreciable levels of 5 protein, but normal endometrial tissue does not (23). Because uterine cells allow expression of ER AF-1 activity (12), and the AF-1 region is contained in 5, we characterized the activity of 5 alone and the modulatory effects of 5 on ER activity in the presence and absence of both E2 and epidermal growth factor (EGF). We show that the overall effect of 5 in endometrial tumor cells is to increase the basal activity of an ERE-containing reporter. The transcriptional response of ER to E2 (fold-response) is decreased in the presence of 5, although the overall promoter activity is increased. In contrast, the transcriptional response of ER to EGF is not suppressed in the presence of 5. In endometrial cancer cells expressing 5 protein, the presence of this variant could result in stimulation of E2-sensitive genes in the absence of ligand.
Materials and Methods
ER constructs
5 cDNA was cloned and constructs prepared as previously described (23). All human ER (hER) constructs were cloned into the EcoR1 site of the pcDNA 3.1 expression vector containing a cytomegalovirus (CMV) promoter (Invitrogen, Carlsbad, CA). The plasmid CMV-ER contains the entire ER cDNA coding sequence. All mutant constructs were created by site-directed mutagenesis (QuikChange by Stratagene, La Jolla, CA) of CMV hER and CMV 5, and include E203G/G204S, S236E, and S118A. The E203G/G204S mutants were created using the following oligonucleotides: 5'ggagtctggtcctgtgggagctgcaaggccttcttcaag-3' and 5' cttgaagaaggccttgcagctcccacaggaccagactcc-3'. These mutations alter the DNA binding site of ER to that of the recognition site for the glucocorticoid receptor (25). The serine 118 residue was mutated using the following oligonucleotides: 5'-gataaaaacaggaggaaggagtgccaggcctgccggctct-3' and 5'-gagccggcaggcctggcactccttcctcctgtttttatc-3'. Substitution of the serine residue at position 118 with an alanine residue prevents phosphorylation and interferes with MAPK and ras-stimulated transcription via ER in HeLa and MCF-7 cells (13, 26). Substitution of the serine 236 residue with a negatively charged glutamic acid at position 236 mimics a phosphorylated residue and prevents receptor dimerization (27) and was mutated using the following nucleotides: 5'-gataaaaaacaggaggaaggagtcccaggcctgccggctc-3' and 5'-gagccggcaggcctgggactccttcctcctgtttttatc-3'.
Cell culture and transfection
Ishikawa cells are an endometrial tumor cell line; we are unable to detect ER or 5 by Western blot analysis, and stimulation of transcription in response to E2 requires transfection of ER. Cells were maintained in modified Eagle’s medium supplemented with 10% fetal bovine serum. Ishikawa cells were plated in six-well plates with DMEM containing no phenol red and supplemented with 5% stripped newborn calf serum to a final concentration of 4 x 105 cells/well. Serum and media were from Mediatech, Inc. (Herndon, VA). Cells were transfected using the calcium phosphate method for 16 h as previously described (28). Amounts of transfected hER, 5, and 5 mutants, or coactivator SNURF are indicated in figure legends. After transfection, cells were washed with PBS (pH 7.4) and treated with vehicle, 10 nM E2, 1 μM 4-hydroxytamoxifen (4-OHT), or 1 μM genistein (G), all from Sigma (St. Louis, MO), for 24 h. In EGF (100 ng/ml) experiments, cells were incubated in DMEM without phenol red or serum for 6–7 h before the addition of the peptide. The model ERE reporter plasmid (0.5 or 1 μg/well in all experiments) used in transfection experiments was (ERE) 2-vit-TK-luc, containing 2 EREs from the vitellogenin 2A promoter fused to the –105-bp thymidine kinase promoter (Vit-6) (29). Plasmid constructs containing the 963 bp of the cyclin D1 gene promoter (30) and 1.8 kb of the human complement 3 (C3) promoter gene promoter (31) fused to luciferase were used to determine the effect of 5 on E2-sensitive, physiological promoters. Each well was transfected with 0.5 μg ERE reporter as indicated and specified concentrations of the expression vectors CMV hER and CMV 5 or CMV 5 mutants. In all experiments, total DNA added was normalized with the control expression vector pcDNA 3.1.
Western blot analysis
For immunoblotting of Ishikawa cell extracts, cells were plated as above, and each well was collected in 1x gel loading buffer (50 mM Tris-Cl, pH 6.8; 2% sodium dodecyl sulfate (SDS), 10% glycerol) and protein concentration determined by BCA assay (Pierce, Rockford, IL). A total of 30 μg of each protein sample, was electrophoresed on a 12% SDS-containing polyacrylamide gel and transferred to a nitrocellulose membrane. The membranes were blocked in 15% nonfat dried milk in 1x Tris-buffered saline, 0.1% Tween 20, for 1 h. Blots were incubated with 1D5, an N-terminal hER monoclonal antibody (Dako, Carpinteria, CA) at a 1:100 dilution overnight, at 4 C. This incubation was followed by goat antimouse secondary antibody (1:15,000, Amersham; Buckinghamshire, UK) overnight. To normalize ER levels for total protein expression in the samples, membranes were incubated with Restore blot stripper (Pierce) for 30 min and probed with anti-glucose-6-dehydrogenase (Sigma) at a 1:1,000 dilution overnight. Incubation was continued with horseradish peroxidase (HRP)-conjugated antirabbit secondary antibody (Amersham, 1:20,000). Immunocomplexes were visualized using an enhanced chemiluminescence detection system (West Pico, Pierce). Cos-1 cells, an ER negative green monkey kidney cell line, were maintained in DMEM supplemented with 10% newborn calf serum. To validate protein expression of CMV hER, CMV 5, and CMV 5, these constructs (1 μg each) were transiently transfected into Cos-1 cells using calcium phosphate as above. Cell extracts were prepared and protein analyzed (30 μg/sample) as above. On each gel, lanes containing in vitro translated ER and 5 were included as positive controls.
To measure SNURF protein levels, Ishikawa cell lysate or primary endometrial tumor lysate was electrophoresed on a 12% denaturing polyacrylamide gel, transferred to a nitrocellulose membrane, and blocked as described above. Gels also contained equal amounts of protein extracts from Cos-1 cells or Cos-1 cells transfected with CMV-SNURF, used as negative and positive controls, respectively (32). Blots were incubated with the SNURF antibody K7979 (kind gift from J. Palvimo, Institute of Biomedicine, University of Helsinki, Helsinki, Finland) at a 1:20,000 dilution overnight, at 4 C, followed by a 60-min incubation with HRP-conjugated antirabbit secondary antibody (Amersham, 1:10,000). To normalize ER levels for total protein expression in the samples, membranes were incubated with Restore blot stripper (Pierce) for 30 min and probed with anti-?-actin (Sigma) at a 1:100,000 dilution overnight. Incubation was continued with HRP-conjugated antimouse secondary antibody (Amersham, 1:50,000).
Glutathione-S-transferase (GST) constructs and pull-down studies
GST-SNURF was kindly provided by J. Palvimo (7). Briefly, RNA was isolated from MCF-7 cells using the following primers: 5'-atggcccagaacttgaagg-3' and 5'-cctcatcaagggraagaaatga-3'. The PCR product was sequenced and subcloned into the EcoR1 site of the pGEX-2T GST fusion vector (Amersham). Proteins were isolated from BC21 cells inoculated with the expression vector for GST, GST-SNURF, or GST-REA. The transformed cells were induced with 100 nM isopropyl ?-D-thiogalactopyranoside and harvested. Bacterial lysate was incubated with glutathione beads (Sigma) at 4 C overnight. Proteins were resolved on a 12% denaturing acrylamide gel, and protein yield was determined by Coomassie staining. Assays were performed as previously described (27). Cell-free translation (TNT Rabbit Reticulocyte Transcription/Translation Kit, Promega; Madison, WI) of ER and 5 was performed in the presence of [S35] methionine (0.044 mCi/50 μl reaction; PerkinElmer, Boston, MA). [S35] methionine-labeled (0.02 mCi/150 μl reaction) ER and 5 were incubated with 1.0 μg GST or GST fusion protein in the presence of 10 μg BSA. Total reaction volume was adjusted to 150 μl with GST wash buffer (10 mM MgCl2, 150 mM KCl, 20 mM HEPES, 10% glycerol, and 0.12% Nonidet P-40). Samples were incubated for 90 min at 4 C. Proteins were electrophoresed on a 12% denaturing acrylamide gel. Gels were exposed to film at –80 C for 48 h. All chemicals were from Sigma.
Statistical analysis
Transfections were performed in triplicate, and each experiment was performed a minimum of three times. Western blots and GST pull-downs were performed at least three times. Data are expressed as mean ± SE of the mean. Data were analyzed by one-way ANOVA to determine differences among the means. Where differences were detected, a Tukey’s post hoc test was used to determine where the means were significant. Values were considered significant at P values 0.05. Analyses were performed using GraphPad Prism software, version 4 (GraphPad, San Diego, CA).
Results
Structure and expression of ER, 5, and 5 mutants
The functional protein domains and exons of ER and 5 are shown in Fig. 1A. Subsequent expression of ER, 5, and mutant proteins was evaluated in Cos-1 cells and Ishikawa cells by immunoblotting. Neither ER nor 5 proteins were detected by this method in Cos-1 cells or in our Ishikawa cells (Fig. 1B). Transiently transfected receptor constructs are easily detected, and expression of 5 does not appear to influence expression of ER protein. We observed equal expression of wild-type and mutant forms of 5 protein from equal amounts of transfected plasmid. (Fig. 1C); thus, potential differences in transcriptional response with these proteins will not be due to different levels of expressed 5 proteins.
5 Stimulates basal expression of an ERE-containing promoter but decreases E2 responses in the presence of ER
Because significant levels of 5 protein are found in endometrial tumors along with full-length ER, we examined the effects of 5 expression alone and on full-length ER transcriptional activity in a tumor cell line of endometrial origin. Untransfected Ishikawa cells express low basal and negligible E2-stimulated activity of an ERE reporter (Vit-6, Fig. 2A). In Ishikawa cells transfected with ER or 5 alone, 5 exhibits basal transcriptional activity on an ERE promoter that is significantly increased relative to no receptor or to ER alone (Fig. 2A). However, only full-length ER is responsive to E2 stimulation. To test the influence of 5 on ER transcription, cells were transfected with 0.5 μg ER expression vector and cotransfected with a 5 expression vector in increasing ratios relative to ER. In the presence of ER plus 5, basal ERE reporter activity was increased relative to ER alone (Fig. 2B). Although the degree of increase varied, promoter activity with 5 plus ER was consistently higher than with ER in all experiments. In subsequent studies with the ERE reporter, ratios of ER to 5 of 1:5 and 1:7 were used. Cotransfection of Ishikawa cells with ER and 5, and subsequent treatment with E2, resulted in a 5-mediated decrease in E2-stimulated reporter activity (Fig. 2C). At a 1.0:0.1 ratio of ER to 5, there was no significant difference in E2-stimulation (8.2 vs. 8.5) from cells transfected with ER alone; but at ratios of 1.0:0.5, E2 stimulation was reduced by approximately 40–50%. At the highest amount of 5 transfected, stimulation of ERE reporter activity was reduced by approximately 75% compared with ER alone. Thus, 5 exerts stimulatory effects on basal ERE reporter activity in the absence of ligand; but in the presence of full-length ER and 5, the response to E2 was reduced.
FIG. 2. 5 Stimulates basal ERE activity alone or with ER, and reduces E2 stimulation in the presence of ER in Ishikawa cells. A, Ishikawa cells were transiently transfected with 0.5 μg hER or increasing amounts of 5, along with 1 μg Vit-6-luciferase reporter, and treated for 24 h with vehicle or 10 nM E2. After collection of lysate, luciferase and protein assays were performed. Data are expressed as normalized luciferase activity and the mean ± SEM for three experiments with three samples per group. B and C, Ishikawa cells were cotransfected with 0.5 μg ER and 5 in increasing ratios of ER to 5 (1:0.1, 1:0.5, 1:1, 1:3, 1:5, and 1:7). Total transfected DNA was normalized using the expression vector pcDNA 3.1. After transfection for 16 h, cells were treated with 10 nM E2 for 24 h. Cells were collected and luciferase activity normalized as in 2A. B, Basal (vehicle-treated) values for each group. Figures above the bars are the fold-stimulation of vehicle-treated activity compared with ER alone. C, Both basal and E2-treated normalized luciferase values. Numbers above the black bars are the fold E2-stimulated values for each transfection group. Values represent the mean ± SEM of three experiments with three samples per group. In each panel: *, P < 0.01 compared with that treatment group’s basal activity. For panel B: *, P < 0.01 compared with activity with ER alone. a, P < 0.01 compared with the basal (untreated) activity of ER alone.
5 Responds to EGF and decreases ER responses to SERMS but not to EGF
We next tested whether 5 could also affect the ER transcriptional response to SERMS and growth factors. Ishikawa cells were transfected with ER alone, 5 alone, or ER and 5 and treated with 10 nM E2, 1 μM 4-OHT (an E2 agonist in uterine tissue), or 1 μM G (a phytoestrogen), or 100 ng/ml EGF (Fig. 3). All data are plotted relative to their own receptor (ER, 5, or ER plus 5) controls. However, relative to reporter activity in the absence of receptor, vehicle-treated promoter activity in the presence of ER, 5, or ER plus 5 was 3.3-, 4.4-, and 6.8-fold over reporter alone, respectively.
FIG. 3. 5 Is stimulated by EGF and reduces the ER transcriptional response to SERMS but not to EGF. Ishikawa cells were transiently transfected with 0.5 μg ER alone, 0.5 μg 5 alone, or ER and 5 (ratio of 1:7) and 1 μg of the Vit6-luciferase reporter. Transcriptional activity was assessed after 24 h treatment of transfected cells with 10 nM E2, 1 μM 4-OHT, 1 μM G, or 100 ng/ml EGF. Fold stimulation in the absence and presence of SERMS is shown. Values represent mean ± SEM of four experiments in each panel (three experiments in cells transfected with 5 alone). *, P < 0.01, treatment compared with that treatment group’s basal activity.
E2, 4-OHT, G, and EGF increased transcriptional activity of ER alone in Ishikawa cells. In cells transfected with 5, only EGF stimulated 5-mediated activity of the ERE reporter, approximately 1.4-fold. Cotransfection of 5 increased basal transcription from the ERE-containing promoter 4.4-fold, but lessened stimulation of the ER by E2 from 2.9- to 1.4-fold, and suppressed stimulation by both SERMS to nonsignificant levels. In contrast, the ER response to EGF was not reduced in the presence of 5 (Fig. 3B). Therefore, 5 may act generally to decrease the transcriptional response to ligands interacting with the ER at the ligand-binding domain (LBD) but allows stimulation by growth factors.
DNA binding and dimerization is required for 5 to increase ERE basal activity, and DNA binding is required to decrease E2 responsiveness in the presence of ER
Because E2 regulates many physiological events in the uterus, we focused our studies on the mechanism of 5 actions on ER in the absence and presence of E2 (Fig. 4A). To determine which regions of the 5 protein mediate the effects on ER activity, mutations were introduced into critical regions of the protein required for phosphorylation in the N-terminal region, DNA binding, and dimerization.
FIG. 4. Mutants of 5 differentially affect protein function. A, Dimerization and DNA binding of 5 is required for full stimulation of basal activity with ER, and DNA binding is required for 5 to reduce E2 stimulation with ER. Ishikawa cells were transiently transfected with 0.5 μg hER, 5 or 5 mutants alone, or 0.5 μg ER cotransfected with 5 at a 1:5 ratio. Mutants included: 5 S118A (phosphorylation mutant); 5 S235E (dimerization mutant), 5 E203G/G204S (DNA binding mutant). Each 5 mutation at the amino acid group is defined in the 5 mutant-alone groups and is abbreviated by the amino acid numbers in the ER plus 5 groups. Total transfected DNA was normalized using the expression vector pcDNA 3.1. Data are expressed as fold-stimulation relative to cells transfected with ER alone, with this value set at 1.0. Values represent mean ± SEM of six experiments, with three samples per group. *, P < 0.01, E2 compared with ERE basal activity in each group. a, P < 0.01, basal ERE activity compared with that with ER alone. B, Ishikawa cells were transiently transfected with reporter alone, 0.5 μg hER alone, 5 or 5 S118A alone, or ER with cotransfected 5 or 5 S118A in a 1:5 ratio. Cells were treated with vehicle or 100 ng/ml EGF for 24 h. Some cells were transfected with reporter (ERE), in the absence of any cotransfected receptor, and treated with or without EGF. *, P < 0.01 E2 compared with ERE basal activity in each group. a, P < 0.05, basal ERE activity compared with that with ER alone. #, P < 0.05, basal ERE activity compared with that of 5 alone.
In Ishikawa cells transfected with 5 S118A (phosphorylation mutant of 5) alone, basal activity of the ERE reporter was significantly increased compared with cells transfected with ER alone, and similar to cells transfected with 5 alone; transcriptional activity did not increase in the presence of E2. In Ishikawa cells cotransfected with ER and 5 S118A, basal activity was significantly increased compared with cells transfected with either ER or 5 alone. In the presence of E2, no significant increase in transcriptional activity was observed. Wild-type 5 and 5 S118A behave similarly when transfected into Ishikawa cells, indicating that the effects of 5 on basal and E2-stimulated ER activity are not mediated at the serine 118 site.
In Ishikawa cells transfected with only the dimerization mutant of 5, 5 S236E, basal activity on an ERE was not increased and was similar to that of cells transfected with ER alone; stimulation by E2 was not observed. Cotransfection of cells with ER and 5 S236E resulted in basal ERE activity that was greater than with ER alone but was significantly diminished compared with cells transfected with ER and wild-type 5. Reporter activity in cells transfected with ER plus 5236E was only slightly increased in the presence of E2. These data indicate that 5 must dimerize for full stimulation of basal ERE activity in the presence of ER plus 5. However, 5 dimerization is not required for the suppression of E2-stimulated ERE reporter activity, because this still occurs with this mutant.
In Ishikawa cells transfected with only the DNA binding mutant of 5, 5 E203G/G204S, basal activity was similar to that of cells transfected with ER alone, and stimulation by E2 was not observed. Cotransfection of cells with both ER and 5 E203G/G204S exhibited similar basal activity compared with cells transfected with ER alone. Reporter activity was significantly increased in the presence of E2. Thus, the 5-induced increase in basal ERE reporter activity is mediated, at least in part, by DNA binding and dimerization. The suppression of the E2-ER transcriptional response by 5 also appears to be mediated by DNA binding.
Serine 118 is required for EGF stimulation
We next examined the contribution of serine 118 to the EGF response by testing the phosphorylation mutant S118A (Fig. 4B). Serine 118 has previously been shown to be required for growth factor and ras-stimulated transcription of ER in HeLa cells (13). In Ishikawa cells transfected with the ERE reporter alone, EGF treatment did not stimulate luciferase activity. 5 significantly stimulates basal transcription from an ERE and is stimulated by EGF, whereas 5 S118A alone increases basal transcription from an ERE but is not stimulated by EGF, indicating this residue is important for stimulation to occur. In cotransfection experiments, the presence of 5 did not prevent significant stimulation of ER by EGF, but cotransfection with 5 S118A did. Thus, S118A is not required for effects on basal transcription but is required for growth factor stimulation.
The coactivator SNURF enhances effects of 5 on basal ERE activity
The ability of 5 to increase basal ERE activity, alone and with ER, may occur through several mechanisms. One possibility may be recruitment of coactivators that interact with the ER N terminal, hinge, or DBD regions that are retained in 5. One such coactivator is SNURF, which augments stimulation of ER by E2 in ZR-75 human breast cancer cells via interaction at the hinge and DBD regions of the receptor (8). To test possible effects of SNURF, Ishikawa cells were cotransfected with ER and 5 in the presence or absence of SNURF (Fig. 5). SNURF increases the basal activity of both ER and 5 on an ERE. SNURF alone had no effect on the promoter construct (not shown). SNURF does not preferentially increase the ER response to E2; however, because the basal activity in the presence of SNURF was higher, the overall activity in the presence of E2 plus SNURF was significantly increased. As in previous experiments, 5 increased basal ERE activity in the presence of ER. SNURF further increased basal activity. Thus, this coactivator would favor increased basal expression from an ERE and decreased fold-stimulation by E2 in the presence of 5 plus ER.
FIG. 5. The coactivator SNURF enhances the effect of 5 and ER on ERE transcription. Ishikawa cells were cotransfected with 0.5 μg ER and 2.5 μg 5, 1 μg Vit6-luciferase, and either 0.5 or 1.0 μg SNURF. Total transfected DNA was normalized using the expression vector pcDNA 3.1. Luciferase activity was determined after 24 h treatment with vehicle (white bars) or 10 nM E2 treatment (black bars). Experiments were performed five times with three samples per group. Shown is a representative experiment in which transfections were performed in triplicate. *, P < 0.05, E2 compared with vehicle in each group. a, P < 0.05, basal ERE activity compared with that with ER alone. #, P < 0.05, comparing basal ERE activity in ER/5 cotransfected cells with and without SNURF. SNURF also significantly stimulated basal activity of the reporter in the presence of ER, 5, or ER plus 5 (P < 0.01).
SNURF is expressed in uterine cells and interacts directly with 5
We examined SNURF expression in our clone of Ishikawa cells and in whole-cell lysate obtained from primary endometrial tumors. Because Cos-1 cells do not express endogenous SNURF, they were transfected with SNURF, and lysates from control and transfected cells were used as positive and negative controls. Figure 6A shows that SNURF is expressed in Ishikawa cells and endometrial tumor tissue. The doublet of SNURF seen in these tissues is observed for endogenous SNURF expression in other tissues (32) and may represent splice variants or posttranslational modifications of SNURF.
FIG. 6. SNURF is expressed in uterine cells and binds both ER and 5. A, SNURF is expressed in uterine cells. One hundred micrograms of Ishikawa cell lysate or endometrial tumor lysate were electrophoresed on a denaturing gel, transferred to nitrocellulose membrane, and immunoblotted for SNURF. Cos-1 cells do not express endogenous SNURF (negative control); Cos-1 cells transiently transfected with SNURF were used as a positive control. A representative blot of three separate experiments is shown. B, ER and 5 bind the coactivator SNURF. Approximately 1 μg GST or GST-SNURF fusion protein was incubated with [S35] methionine labeled ER or 5 for 90 min. Beads were washed, and bound proteins were eluted and electrophoresed on a 12% SDS-containing polyacrylamide gel. Migration of proteins bound to GST or GST-SNURF are indicated by arrows and visualized by autoradiography. Input in vitro translated (IVT) 5 and ER is shown and represents 1/8 input of radiolabeled protein in each reaction. A representative autoradiogram from six experiments is shown.
Although SNURF effects on ER transcription require the hinge and DBD regions of ER, direct SNURF binding to the receptor had not been demonstrated (8). Because the coactivator SNURF modulates 5 activity in our cell culture system, we investigated direct protein interactions between SNURF and ER and SNURF and 5 in GST pull-down experiments. GST and GST-SNURF fusion proteins were incubated with radiolabeled wild-type ER and 5. Minimal interaction was observed between GST and ER and GST and 5. GST-SNURF bound both ER and 5 in pull-down experiments (Fig 6B). Thus, the regions contained in the 5 protein are sufficient for interactions with SNURF, and SNURF could act to increase the basal activity of ER and 5-containing dimers by binding directly to the proteins.
5 Expression modulates the activity of complex physiological E2-sensitive promoters
To test whether physiologically E2-sensitive promoters in the uterus could be regulated by 5, we tested several complex physiological promoters, including the promoter of the cell cycle regulator and protooncogene cyclin D1 and the complement 3 promoter. E2 sensitivity of the cyclin D1 promoter is not mediated by a canonical ERE, but by a proximal cAMP response element and an AP-1 site (33, 34). Cyclin D1 promoter activity was significantly stimulated by E2 in cells transfected with ER. Basal activity of the cyclin D1 promoter was only slightly stimulated by 5 alone. However, there was up to a 1.7-fold increase in basal promoter activity in cells cotransfected cells with ER plus 5 compared with basal activity in cells transfected with ER alone. Under these conditions, the promoter was no longer stimulated significantly by E2. The E2 stimulation of the C3 promoter is mediated via three responsive elements, including one noncanonical ERE and two additional elements with no significant homology to known EREs (31), and C3 is regulated in uterus by E2 and during the menstrual cycle (35). In cells transfected with the C3 reporter construct and ER, the activity of the promoter was significantly stimulated (3.6-fold) by E2. Transfection with 5 alone stimulated basal activity of the promoter 1.5-fold, and cotransfection of 5 plus ER increased basal activity of the C3 promoter compared with either ER (2.2-fold) or 5 alone (2.7-fold). Increasing 5 resulted in a decreased E2 stimulation of promoter activity, as for cyclin D1. In both cases, overall promoter activity did not increase with increasing 5; thus, the net effect was increased promoter activity that did not require E2. These data indicate that, in addition to model reporters, 5 can also regulate the activity of physiologically relevant promoters.
Discussion
Splice variants of ER mRNA are commonly transcribed in normal and cancerous tissues (36) but are rarely translated into protein. Three notable exceptions are: the presence of 4 in primary ovarian tumors (20), 7 in primary tumors of the breast and in breast cancer cell lines (21), and 5 in breast (22) and primary endometrial tumors (23). Our studies focus on the activity of 5 in Ishikawa cells, an endometrial cancer cell line, and some of the transcriptional consequences for E2-regulated genes. We demonstrated previously that 5 mRNA is translated to protein in primary endometrial tumors but is not expressed in normal uterine tissue (23). This is physiologically significant, because activity of a functional 5 protein may affect the expression or activity of the wild-type receptor. Indeed, it has been shown that 5 modulates the activity of ER in a cell-specific manner. For example, in osteosarcoma cells stably transfected with ER, 5 increases both basal and E2-stimulated activity as well as cellular proliferation (18). In MCF-7 breast cancer cells stably transfected with 5, the variant did not modulate E2 responsive genes, such as the progesterone receptor, but did stimulate growth of transfected cells (19). In HeLa cells and breast epithelial cells, 5 had a dominant negative effect on ER stimulation of a reporter gene (14, 24). The varied effects of 5 indicate that its effects are dependent on cell and promoter context.
We investigated the activity of 5 alone and the ability of 5 to modulate ER activity using transient transfection studies in Ishikawa cells. We found that 5 alone consistently increased basal activity of the ERE-containing model promoter, as well as the basal activity of the physiological promoter, C3. The cyclin D1 promoter was not significantly stimulated by 5 alone, perhaps because it responds to E2 via an AP1 site and tethered proteins rather than by direct receptor binding to DNA. As expected, 5-mediated activity was not modulated by E2; however, EGF stimulated the activity of 5 alone. Basal activity mediated by 5 required both DNA binding and dimerization, and EGF stimulation required an intact serine 118 residue. This suggests that phosphorylation of this residue may be important for the EGF effect on 5, as has been reported previously for full-length ER (13, 37).
In endometrial tumors, 5 expression occurs in the presence of full-length ER. Therefore, we examined the influence of 5 on ER activity. In the presence of ER, addition of 5 increases basal activity of an ERE-containing promoter, as well as two physiological promoters, in a dose-dependent manner (Figs. 2B and 7). Moreover, coexpression of 5 plus ER increased reporter activity more than either receptor alone. The magnitude of this increase varied, perhaps as a result of differing expression of coactivator, corepressor, or other integrator proteins that modulate ER function. Similar to effects of 5 alone, increased basal activity with ER plus 5 required 5 DNA binding and dimerization functions of 5 but not serine 118. Collectively, these data indicate that 5 is indeed able to dimerize with itself and/or ER and to occupy an ERE to alter the transcription of E2-responsive genes in the absence of ligand. Our results on the importance of DNA binding for 5 function in a physiological system are supported by published data demonstrating binding of 5, with and without ER, to DNA in gel shift assays (14, 18). Our data indicate that 5 behaves in uterine cells in a dominant positive manner. This is in agreement with results from osteosarcoma cells (18), but in contrast with results in breast cells and HeLa cells (14, 24). This is unlikely to depend exclusively on any one specific physical property of 5 but also on the presence or activity of other proteins that modulate receptor activity. However, mutational analysis of 5 was not performed in the context of these other cell types, and direct comparisons cannot be made.
FIG. 7. 5 Modulates activity of the cyclin D1 promoter and the complement 3 promoters. A. Cells were transfected with 0.5 μg cyclin D1 promoter-luciferase construct and 2.5 or 3.5 μg 5 in the absence or presence of 0.5 μg ER and treated with 10 nM E2 for 12 h. *, P < 0.05, E2 stimulated activity compared with each group’s vehicle-treated control. a, P < 0.05, basal activity compared with basal activity of the cyclin D1 promoter in the presence of ER. B, Cells were transfected with 0.5 μg complement 3 promoter-luciferase construct and 2.5–3.5 μg 5 in the absence or presence of 0.5 μg ER and treated with E2 for 24 h. *, P < 0.001, E2-stimulated activity compared with group’s vehicle-treated control. a, P < 0.001, basal activity of the C3 promoter in the presence of ER plus 5 compared with ER alone. #, P <.05, basal activity of the C3 promoter with no receptor vs. with 5 alone.
In Ishikawa cells, we also observed that 5 reduced the fold stimulation of the ERE reporter by ER and E2 (Fig. 2). Similar effects were observed with the complex physiological promoters for cyclin D1 and C3. The effects of 5 on ligand-stimulated ER transcriptional activity occur more generally, because 5 expression also reduced the transcriptional response to the SERMS tamoxifen and G (Fig. 3). DNA binding of 5 is required for this effect (Fig. 4A). Because 5 does not contain an LBD, it cannot bind or respond to ligand. Thus, increased occupancy of the ERE by 5 or 5-ER dimers would result in a decreased response to ligand.
In contrast, the presence of 5 does not prevent ER-mediated stimulation of an ERE-containing reporter by EGF (Fig. 4B). Serine 118 was required for this response, because the 5 phosphorylation mutant could not be stimulated by EGF, and the mutation prevented EGF stimulation of the ER/5 complex. The magnitude of the response to EGF by ER alone or ER plus 5 was constant. However, because basal expression from an ERE reporter was increased, the total EGF response was greater in cells transfected with ER plus 5. Several groups have shown that ER can be activated by various growth factors, including IGF-1, EGF, TGF and heregulin, as well as signaling intermediates of the MAP kinase pathway (10, 11, 37, 38). EGF stimulates ER activity in HeLa cervical cancer cells (13) in the mouse uterus (39), and in Ishikawa cells (40), and this occurs primarily through the AF-1 region (13, 40). Because the 5 protein contains the AF-1 and the DBD, dimers containing 5 can respond to EGF in this cell line.
The dominant positive effect of 5 in endometrial carcinoma cells could occur through several mechanisms, including preferential expression or altered activity of specific coactivators, corepressors, or integrator proteins that stabilize DNA binding of complexes containing 5 and/or favor expression of AF-1 activity. Ishikawa cells express high levels of some coactivators and other regulatory molecules relative to MCF7 cells (41). The ratio of coactivators and corepressors in uterine cells contributes to the transcriptional responses to tamoxifen, which suppresses ER AF-2 activity but allows responses mediated by AF-1 (42). A second contributing factor to 5 activity may be the absence of the AF-2 region, which might release constraints in the full-length receptor and allow maximum expression of AF-1 activity. Finally, many corepressors, such as nuclear receptor corepressor (NCoR), silencing mediator of retinoic acid and thyroid hormone receptors (SMRT), and REA, preferentially bind to the C-terminal LBD, and this is missing in 5 (9, 43). Because REA can bind to ER in the presence of E2 and compete with coactivators for binding to the LBD, it may exert major effects on gene activity. A recent report demonstrated that mice in which the REA gene is disrupted have significant changes in endogenous uterine genes, and C3 expression is stimulated (44). Alternatively, 5 may titrate away an unknown repressor acting via the N terminal and stimulate activity of both ER and 5. Because the effects of 5 plus ER on basal activity are greater than either receptor alone, a combination of these or other possibilities may occur.
We hypothesized that dimers containing 5 would bind a particular class of coactivators and stimulate activity of this complex. Several coactivators have been reported to activate ER through the N-terminal AF-1 region. P300 binds to AF-1 and mediates synergism between AF-1 and AF-2 (45). CBP can bind to the full-length ER in the absence of E2 and stimulates AF-1 activity on an E2-sensitive promoter specifically through serines 104/106/118 (46). However, these studies tested protein-protein interactions via yeast or mammalian two-hybrid approaches, and the entire ER molecule was required for most responses. Other coactivators such as SRA (47), ini (48), arginine methyltransferase (49), and SNURF (8) have also been reported to stimulate transcription of E2-responsive genes via the AF-1 or DBD region of ER. Modulation of 5 activity by these and other coactivators could play a role in this system.
We tested one coactivator, SNURF, and its ability to modulate 5 activity. SNURF acts via the hinge and DBD of the androgen receptor (7), and stimulates ER activity in breast cancer cells through multiple regions, including the DBD and hinge (8). SNURF is expressed in Ishikawa cells and in human endometrial tumors (Fig 6), and increased basal activity of an ERE-containing promoter, in the presence of ER, 5, and particularly of ER plus 5 (Fig. 5), up to 10-fold. In the presence of SNURF, ERE-promoter reporter activity in cells transfected with ER plus 5 was increased, but the response to E2 was reduced. This may be because SNURF favors binding of 5, in homo- or heterodimers to DNA, and this variant form cannot respond to E2. Alternatively, SNURF binding to ER-5 heterodimers may alter or inhibit subsequent recruitment of coactivators or regulatory proteins in the presence of E2. These results are somewhat divergent from those of Saville et al. (8), who showed that SNURF amplifies the E2-stimulated (up to 5-fold) but not basal ER transcription of the simple ERE-TATA box reporter in breast cancer cells. Some differences in results between the two cell systems may arise because of the use of the simpler promoter vs. a more complex thymidine kinase promoter we used in our studies. The results may also be due to the preferential expression of AF-1 activity in uterine cells and the ability of SNURF to enhance basal expression through this pathway. Overall expression of SNURF may also play a role in the response, but this was not tested in the breast cancer cell studies. Using GST pull-down analysis, we demonstrated, for the first time, direct interaction between SNURF and ER, and SNURF and 5 (Fig. 6). This interaction is likely through the DBD, because SNURF binds to the DBD of several nuclear receptors (7, 8), and the DBD is contained in both ER and 5. Thus, stimulated basal activity from an ERE-containing promoter in the presence of 5 may occur through preferential recruitment of specific coactivators.
Modulation of ER activity by the 5 ER variant is important because the preferential expression of 5 in endometrial cancer may have significant physiological implications (23). In endometrial tumors, expression of variant protein occurs in addition to overall decreases in full-length ER and progesterone receptor expression (50). In Ishikawa cells, the expression of 5, in addition to full-length ER, increases the expression of ER-sensitive genes in the absence of ligand and reduces the response to ligand and SERMS. Therefore, endometrial tumors expressing 5 might be less responsive to antiestrogens, despite a steroid receptor profile that classifies it as being ER positive. It could thus be helpful to consider the expression of ER variants in clinical methods used to determine steroid receptor profiles. Because expression of 5 allows growth factor stimulation of the receptor, the growth factor pathway may become more influential in the regulation and maintenance of uterine tumors. In the case of endometrial cancers that express 5, compounds that interfere with the ability of 5 either to dimerize and bind DNA or otherwise disrupt signaling may prove useful.
Acknowledgments
The authors thank Dr. Jorma Palvimo (Institute of Biomedicine) for the SNURF expression vectors.
References
Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20:358–417
Nilsson S, Makela S, Treuter E, Tujague M, Thonsen J, Andersson G, Enmark E, Pettersson K, Warner M, Gustafsson J-A 2001 Mechanisms of estrogen action. Physiol Rev 81:1535–1565
Onate SA, Tsai SY, Tsai M-J, O’Malley BW 1995 Sequence and characterization of a co-activator for the steroid hormone receptor superfamily. Science 270:1354–1357
McInerney EM, Tsai MJ, O’Malley BW, Katzenellenbogen BS 1996 Analysis of estrogen receptor transcriptional enhancement by a nuclear hormone receptor coactivator. Proc Natl Acad Sci USA 93:10069–10073
Hanstein B, Eckner R, DiRenzo J, Halachmi S, Liu H, Searcy B, Kurokawa R, Brown M 1996 P300 is a component of an estrogen receptor co-activator complex. Proc Natl Acad Sci USA 93:11540–11545
Chakravarti D, LaMorte VJ, Nelson MC, Nakajima T, Schulman IG, Juguilon H, Montiminy M, Evans RM 1996 Role of CBP/P300 in nuclear receptor signaling. Nature 383:99–103
Moilanen AM, Poukka H, Karvonen U, Hakil M, Janne OA, Palvimo JJ 1998 Identification of a novel RING finger protein as a coregulator in steroid receptor-mediated gene transcription. Mol Cell Biol 18:5128–5139
Saville B, Poukka H, Wormke M, J?nne OA, Palvimo JJ, Stoner M, Samuido I, Safe S 2002 Cooperative coactivation of estrogen receptor a in ZR-75 human breast cancer cells by SNURF and TATA binding protein. J Biol Chem 277:2485–2497
Montano MM, Ekena K, Delage-Mourroux R, Chang W, Martini P, Katzenellenbogen BS 1999 An estrogen receptor coregulator that potentiates the effectiveness of antiestrogens and represses the activity of estrogens. Proc Natl Acad Sci USA 96:6947–6952
Ignar-Trowbridge DM, Pimentel M, Parker MG, McLachlan JA, Korach KS 1996 Peptide growth factor cross-talk with the estrogen receptor requires the A/B domain and occurs independently of protein kinase C or estradiol. Endocrinology 137:1435–1744
Aronica SM, Katzenellenbogen BS 1993 Stimulation of estrogen receptor mediated transcription and alteration on the phosphorylation state of the rat uterine receptor by estrogen, cyclic adenosine monophosphate and insulin like growth factor-1. Mol Endocrinol 6:743–752
Tzukerman MT, Esty A, Santiso-Mere D, Danielian Parker MG, Stein RB, Pike JW, McDonnell DP 1994 Human estrogen receptor transactivational capacity is determined by both cellular and promoter context and mediated by two functionally distinct intramolecular regions. Mol Endocrinol 6:1043–1050
Kato S, Endoh H, Masuhiro Y, Kitamoto T, Uchiyama S, Sasaki H, Masushige S, Gotoh Y, Nishida E, Kawashima H, Metzger D, Chambon P 1995 Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science 270:1491–1494
Bollig A, Miksicek RJ 2000 An estrogen receptor- splicing variant mediates both positive and negative effects on gene transcription. Mol Endocrinol 14:634–649
Hu C, Hyder SM, Needleman DS, Baker VV 1996 Expression of estrogen receptor variants in normal and neoplastic human uterus. Mol Cell Endocrinol 118:173–179
Rice LW, Jazaeri AA, Shupnik MA 1997 Estrogen receptor mRNA splice variants in pre- and postmenopausal human endometrium and endometrial carcinoma. Gynecol Oncol 65:149–157
Castles CG, Fuqua SAW, Klotz DM, Hill SM 1993 Expression of a constitutively active estrogen receptor variant in the estrogen receptor-negative BT-20 human breast cancer cell line. Cancer Res 53:5934–5939
Chaidarun SS, Alexander JM 1998 A tumor-specific truncated estrogen receptor splice variant enhances estrogen-stimulated gene expression. Mol Endocrinol 12:1355–1366
Rae D, Parker MG 1996 Effects of an exon 5 variant of the estrogen receptor in MCF-7 breast cancer cells. Cancer Res 56:1556–1563
Park W, Choi JJ, Hwang ES, Lee JH 1996 Identification of a variant estrogen receptor lacking exon 4 and its coexpression with wild-type estrogen receptor in ovarian carcinomas. Clin Cancer Res 12:2029–2035
Fasco MJ, Keyomarsi K, Arcaro KF, Gierthy JF 2000 Expression of an estrogen receptor variant protein in cell lines and tumors. Mol Cell Endocrinol 162:167–180
Fuqua SA, Fitzgerald SD, Chamness GC, Tandon AK, McDonnell DP, Nawaz Z, O’Malley BW, McGuire WL 1991 Variant human breast tumor estrogen receptor with constitutive transcriptional activity. Cancer Res 51:105–109
Jazaeri O, Shupnik MA, Jazaeri AA, Rice LW 1999 Expression of estrogen receptor mRNA and protein variants in human endometrial carcinoma. Gynecol Oncol 74:38–47
Ohllson H, Lykkesfeldt AE, Madsen MW, Briand P 1998 The estrogen receptor variant lacking exon 5 has dominant negative activity in the human breast epithelial line HMT-3522S1. Cancer Res 58:4264–4268
Mader S, Kumar V, de Verneuil H, Chambon P 1989 Three amino acids of the oestrogen receptor are essential to its ability to distinguish an oestrogen from a glucocorticoid-responsive element. Nature 338:271–274
Migliaccio A, Di Domenico M, Castoria G, de Falco A, Bontempo P, Nola E, Auricchio F 1996 Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol-receptor complex in MCF-7 cells. EMBO J 15:1292–1300
Chen D, Pace PE, Coombes RC, Ali S 1999 Phosphorylation of the human estrogen receptor by protein kinase A regulates dimerization. Mol Cell Biol 19:1002–1015
Lin VY, Resnick EM, Shupnik MA 2003 Truncated estrogen receptor product-1 stimulates estrogen receptor transcriptional activity by titration of repressor proteins. J Biol Chem 278:38125–38131
Schreihofer DA, Resnick EM, Soh AY, Shupnik MA 1999 Transcriptional regulation by a naturally occurring truncated rat estrogen receptor, truncated ER product-1 (TERP-1). Mol Endocrinol 13:320–329
Albanese C, Johnson J, Watanabe G, Eklund N, Vu D, Arnold A, Pestell RG 1995 Transforming p21 mutants and c-Ets-2 activate the cyclin D1 promoter through distinguishable regions. J Biol Chem 270:23589–23597
Norris JD, Fan D, Wagner BL, McDonnell DP 1996 Identification of the sequences within the human complement 3 promoter required for estrogen responsiveness provides insight into the mechanism of tamoxifen mixed agonist activity. Mol Endocrinol 10:1605–1616
Curtin D, Ferris HA, Hakli M, Gibson M, Janne OA, Palvimo JJ, Shupnik MA 2002 SNURF increases basal and GnRH-stimulation of the rat LH? gene and prevents androgen suppression. Mol Endocrinol 18:1263–1276
Castro-Rivera E, Samuidio I, Safe S 2001 Estrogen regulation of cyclin D1 gene expression in ZR-75 breast cancer cells involves multiple enhancer elements. J Biol Chem 276:30853–30861
Liu M-M, Albanese C, Anderson CM, Hilty K, Webb P, Uht RM, Price RH, Pestell RG, Kushner PJ 2002 Opposing action of estrogen receptors and ? on cyclin D1 gene expression. J Biol Chem 277:24353–24360
Sundstrom DA, Komm BS, Ponce-de-Leon H, Yi Z, Teuscher C, Lyttle CR 1989 Estrogen regulation of tissue specific expression of complement C3. J Biol Chem 264:16941–16947
Yang J, Liu A, Chiou SK, Guzman R, Nandi D 2000 Estrogen receptor variants are present in many normal human tissues. Int J Mol Med 3:223–227
Bunone G, Briand PA, Miksicek RJ, Picard D 1996 Activation of the unliganded estrogen receptor by EGF involves the MAP kinase pathway and direct phosphorylation. EMBO J 15:2174–2183
Pietras RJ, Arboleda J, Reese DM, Wongvipat N, Pegram MD, Ramos L, Gorman CM, Parker MG, Sliwkowski MX, Slamon DJ 1995 HER-2 tyrosine kinase pathway targets estrogen receptor and promotes hormone independent growth in human breast cancer cells. Oncogene 10:2345–2346
Ignar-Trowbridge DM, Pimentel M, Teng CT, Korach KS, McLachlan JA 1995 Cross-talk between peptide growth factor and estrogen receptor signaling system. Environ Health Perspect 103(Suppl 7):35–38
Ignar-Trowbridge DM, Teng CT, Ross KA, Parker MG, Korach KS, McLachlan JA 1993 Peptide growth factors elicit estrogen receptor-dependent transcriptional activation of an estrogen-responsive element. Mol Endocrinol 8:992–998
Thenot D, Charpin M, Bonnet S, Cavailles V 1999 Estrogen receptor cofactors expression in breast and endometrial human cancer cells. Mol Cell Endocrinol 156:85–93
Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of SERMs. Science 295:2465–2468
Klinge CM 2000 Estrogen receptor interaction with co-activators and co-repressors. Steroids 65:227–251
Park S-E, Xu J, Frolova A, O’Malley BW, Katzenellenbogen BS, Genetic ablation of the repressor of estrogen receptor activity (REA) modulates response to estrogen in target tissues. Program of the 84th Annual Meeting of The Endocrine Society, New Orleans, LA, 2004 (Abstract P1-149)
Kobayashi Y, Kitamoto T, Yoshika M, Michiko W, Kase T, Metzger D, Yanagisawa J, Kato S 2000 p300 mediates functional synergism between AF-1 and AF-2 of estrogen receptor and ? by interacting directly with the N-terminal A/B domains. J Biol Chem 275:15645–15651
Dutertre M, Smith CL 2003 Ligand-independent interactions of p160/steroid receptor coactivators and CREB-binding protein (CBP) with estrogen receptor-: regulation by phosphorylation sites in the A/B region depends on receptor domains. Mol Endocrinol 17:1296–1314
Deblois G, Giguere V 2003 Ligand independent coactivation of ER AF-1 by steroid receptor RNA activator (SRA) via MAPK activation. J Steroid Biochem Mol Biol 85:123–131
Oltra E, Pfeifer I, Werner R 2003 Ini, a small nuclear protein that enhances the response of the connexin 43 gene to estrogen. Endocrinology 144:3148–3158
Qi C, Chang J, Zhu Y, Yeldandi AV, Rao SM, Zhu YS 2002 Identification of protein arginine methyltransferase 2 as a coactivator for estrogen receptor . J Biol Chem 297:28624–28630
Jazaeri AA, Nunes KJ, Dalton MS, Xu MH, Shupnik MA, Rice LW 2001 Well-differentiated endometrial adenocarcinomas and poorly differentiated mixed müllerian tumors have altered ER and PR isoform expression. Oncogene 20:6965–6969(Winnifred Bryant, Allison)