Role of Estradiol Receptor-α in Differential Expression of 2,3,7,8-Tetrachlorodibenzo-p-dioxin-Inducible Genes in the RL95-2 and KLE Human Endometrial Cancer Cell Lines

Archives of Biochemistry and Biophysics Vol. 368, No. 1, August 1, pp. 31–39, 1999 Article ID abbi.1999.1288, available online at http://www.idealibrary.com on

Role of Estradiol Receptor-a in Differential Expression of 2,3,7,8-Tetrachlorodibenzo-p-dioxin-Inducible Genes in the RL95-2 and KLE Human Endometrial Cancer Cell Lines N. R. Jana,* S. Sarkar,* M. Ishizuka,* ,† J. Yonemoto,* ,† C. Tohyama,† ,‡ and H. Sone* ,† ,1 *Chemical Exposure and Health Effects Research Team, Regional Environment Division, and ‡Environmental Health Sciences Division, National Institute for Environmental Studies 16-2 Onogawa, Tsukuba, Ibaraki 305 0053, Japan; and †CREST, JST, 418 Honcho, Kawaguchi, Saitama 332 0012, Japan

Received February 8, 1999, and in revised form May 3, 1999

The present study was conducted to investigate the mechanism of the response of human uterine endometrial carcinoma cells, RL95-2 and KLE, to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). RL95-2 cells were highly responsive to TCDD in terms of cytochrome P4501A1 (CYP1A1), cytochrome P4501B1 (CYP1B1), and plasminogen activator inhibitor-2 (PAI-2), whereas KLE cells showed little stimulatory effects only at high doses. Neither showed any growth inhibition upon exposure to TCDD. KLE cells expressed higher levels of aryl hydrocarbon receptor (AhR) than RL95-2 and gel mobility shift assay also identified more liganded AhR–ARNT complex bound to xenobiotic response elements (XRE). TCDD had no downregulatory effects on the expression of either AhR or the estradiol receptor (ER). Though both cell types expressed ER-a almost equally, immunofluorescence demonstrated a defect in its nuclear translocation in KLE cells where ER-a was mainly cytoplasmic and estradiol-17b (E 2) was unable to translocate it to the nucleus. However, both cells were nonresponsive to E 2 in terms of transcriptional activation and transient expression of normal ER-a restored the E 2 responsiveness. Transient expression of ER-a in KLE cells also restored its responsiveness to TCDD on transcriptional activation. Collectively, these results indicate that ER-a acts as a positive modulator in regulation of the TCDD-inducible genes. © 1999 Academic Press

1

To whom correspondence should be addressed at Chemical Exposure and Health Effects Research Team, Regional Environment Division, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305 0053, Japan. Fax: 181-298-50-2571. E-mail: [email protected]. 0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD 2) and its related polycyclic or halogenated aromatic hydrocarbons produce a number of biochemical and toxicological responses. Its effects include carcinogenesis, hepatotoxicity, immune suppression, tumor promotion, induction of hypo/hyperplasia, and reproductive and developmental toxicity. Typical biochemical responses are induction of CYP isozymes and several other drugmetabolizing enzymes, cytokines, and various other growth factors (1, 2). The aryl hydrocarbon receptor (AhR) has been identified as an initial cellular target for TCDD and related compounds, and most studies have indicated that their effects are mediated in this way (3). The AhR is a ligand-activated transcription factor and a member of the basic helix-loop-helix superfamily of DNA binding proteins. Unliganded AhR is part of a cytosolic protein complex containing heat shock protein 90 (HSP90) (4) and possibly another 46-kDa protein (5). Binding of the ligand to the AhR results in the release of HSP90 and translocation to the nucleus followed by dimerization to the aryl hydrocarbon receptor nuclear translocator (ARNT) (3, 4, 6). Once inside the nucleus, the AhR–ARNT heterodimer binds with specific cis-acting enhancers, called xenobi2

Abbreviations used: TCDD, 2,3,7,8-tetrachlorodibenzo-p-dioxin; AhR, aryl hydrocarbon receptor; ARNT, aryl hydrocarbon receptor nuclear translocator; CYP1A1, cytochrome P4501A1; CYP1B1, cytochrome P4501B1; EROD, ethoxyresorufin O-deethylase; XRE, xenobiotic response element; E 2, estradiol-17b; ER, estradiol receptor; ERE, estradiol response element; RT-PCR, reverse-transcription polymerase chain reaction; PAI-2, plasminogen activator inhibitor-2; LUC, luciferase; DMSO, dimethyl sulfoxide; FBS, fetal bovine serum; PBS, phosphate-buffered saline; DTT, dithiothreitol; DRE, dioxin response element; NRE, negative regulatory elements; EGFR, epidermal growth factor receptor. 31

32

JANA ET AL.

otic or dioxin response elements (XREs or DREs), that promote the activation of several genes (4, 7). TCDD has been reported to cause carcinogenic effects in liver and skin, but anticarcinogenic effects in hormone-dependent tissues such as mammary, uterus, and the pituitary of female rats fed the compound for a period of 2 years (8). Several in vivo studies have also suggested that TCDD exposure counteracts the effects of estrogen with regard to uterine hypertrophy, peroxidase activity, and binding activities of estradiol receptor (ER), progesterone receptor (PR), and epidermal growth factor receptor (8 –10). In ER-positive human breast cancer cell lines, TCDD inhibits 17b-estradiol (E 2)-dependent cell proliferation (11), secretion of tissue plasminogen activator (12), postconfluent focus formation (13, 14), and secretions of E 2-induced proteins like cathepsin-D or pS2 (11). None of these effects has been found in ER-negative breast cancer cells (11, 15). There is substantial evidence that TCDD does not interact directly with ER or PR; therefore, the antiestrogenic effects cannot be explained by direct interaction of TCDD with those receptors (16, 17). Rather, antiestrogenic activity might be explained by a decrease in amounts of ER (18, 19) or inhibition of estradiol-induced gene transcription by transcriptional interference with the liganded AhR–ARNT complex (20) or by interaction of the liganded AhR–ARNT complex with XRE elements found to be present in E 2-inducible genes (21). In addition to its several antiestrogenic actions, it is of growing concern that TCDD might be linked with endometriosis, the presence of endometrial-like tissue in locations outside the endometrial cavity. Rhesus monkeys or rodents exposed to TCDD in their diets manifested a dose-related increase in the severity and incidence of endometriosis (22, 23). In addition, several epidemiological studies have provided evidence in support of a possible association between TCDD and the prevalence of endometriosis in women (24, 25). Although the exact mechanism of endometriosis is not known, it has been suggested that impaired immune function or imbalance in the activity of sex hormones or growth factors might be linked to the etiology of this disease (26, 27). Among TCDD-inducible genes, cytochrome P4501A1 (CYP1A1) has merited considerable attention because of its association with both toxic outcomes and genesis of several cancers. Induction of CYP1A1 has been correlated with increased susceptibility to lung, breast, and uterine cancer development and can be used as a diagnostic marker (28 –30). It can also be used as a marker for exposure or responsiveness of particular tissues or cell types to various toxic compounds because of its rapid induction at very low doses (31, 32). In the present study, we have investigated the mechanism of different responses of two uterine endometrial

cell lines, RL95-2 and KLE, to TCDD with regard to their CYP1A1, CYP1B1, and PAI-2 induction and also the novel factors responsible for differences in their responses. MATERIALS AND METHODS Expression and reporter plasmids. The expression plasmid for the human ER-a HEO was a kind gift from Dr. P. Chambon (IGBMC, INSERM, France). The E 2-responsive reporter plasmid pGL3-3 (EREc38)-LUC (33), containing three head-to-tail tandem copies of the consensus estradiol response element (ERE), was also a gift, from Dr. C. M. Klinge (University of Louisville, U.S.A.). The TCDD-responsive reporter plasmid pGL3-1 (XRE)-LUC was prepared by cloning an oligonucleotide (same as used in the gel shift assay) containing single XRE elements into the BglII site of the pGL3 promoter vector (Promega, U.S.A.). Cell culture and treatments. RL95-2 (moderately differentiated human endometrial carcinoma), KLE (poorly differentiated human endometrial adenocarcinoma), and MCF-7 (human breast carcinoma) cell lines were obtained from the American Type Culture Collection (Rockville, MD) and routinely maintained in DMEM/ Ham’s F12 (1:1) medium supplemented with 10% fetal bovine serum (FBS), 100 units penicillin/ml, and 100 mg streptomycin/ml under standard conditions in a 37°C incubator with a humidified mixture of 5% CO 2 and 95% air. All tissue-culture reagents were obtained from Life Technologies, U.S.A. The media were refreshed twice a week and cells were passaged with trypsinization every week. For TCDD dose and time kinetic studies, cells were grown on 60-mm 2 tissue culture dishes in routine culture media, and at about 80 –90% confluency cells were treated with different concentrations of TCDD in 0.1% DMSO (v/v) for different time periods. The controls received 0.1% DMSO. Cell proliferation assays. Cells were plated at a density of approximately 2 3 10 4 cells/well in 24-well tissue culture plates in reduced growth medium of 5% FBS (phenol red-free medium and charcoal-stripped serum for experiments containing E 2). Cells were treated with either TCDD or E 2 alone or in different combinations for 8 days with the medium changed and cells redosed every 2 days. Cell proliferation was measured using the Cell Titer 96 AQ ueous onesolution cell proliferation assay system (Promega) following the instructions provided in the kit. Ethoxyresorufin O-deethylase (EROD) assays. Cells were harvested in ice-cold Tris/sucrose (10 /25 mM, pH 7.5) by scraping, collected by centrifugation, and homogenized in the same buffer. EROD activities of the crude homogenate were measured spectrofluorometrically as described earlier (34). Protein concentrations were measured using a Coomassie protein assay kit (Bio-Rad, CA). RT-PCR analysis. Total RNA was prepared from cells using Isogen (Nippon Gene, Japan) according to the manufacturer’s instructions and RT-PCR was carried out with a RT-PCR kit (TaKaRa Biomedicals, Japan). Briefly, 0.1 mg of total RNA was reverse transcribed in a final volume of 10 ml with the following profile: 30°C for 10 min, 55°C for 20 min, 95°C for 5 min, and 5°C for 5 min. PCR primers were designed using Oligo 5 software (National Bioscience, U.S.A.) and sequence information available through the National Centre for Biotechnology Information (National Library of Medicine, U.S.A.). Primer sequences are given in Table I. PCRs were carried out in thin-walled reaction tubes (Perkin–Elmer, U.S.A.) in a final volume of 50 ml using a Perkin–Elmer Gene Amp PCR System 2400 with the following profile: an initial denaturation step at 94°C for 4 min and then cycled through a 30-s denaturation step at 94°C, a 30-s annealing step at different temperatures (see Table I), a 45-s extension step at 72°C, and a final extension step at 72°C for 5 min. Products were separated on 1.5% agarose gel, stained with ethidium bromide, and photographed on a UV transilluminator. Photographic

ER-a IN REGULATION OF TCDD-INDUCIBLE GENE EXPRESSION

33

TABLE I

Oligonucleotide Primers Used for the RT-PCR Gene Expression Analysis

Primer set

Sequence (59-39)

CYP1A1F CYP1A1R CYP1B1F CYP1B1R PAI-2F PAI-2R ACTINF ACTINR AhRF AhRR ARNTF ARNTR ER-aF ER-aR ER-bF ER-bR

TAGACACTGATCTGGCTGCAG GGGAAGGCTCCATCAGCATC TGATGGACGCCTTTATCCTC ACCTGATCCAATTCTGCCTG TTCATCCTTCCGCTCTCTCAG CTTCAGTGCCCTCCTCATTCA GTGGGGCGCCCCAGGCACCA CTCCTTAATGTCACGCACGATTTC ATACCGAAGACCGAGCTGAA TTTTTGGTCCGGATTTCAAG TGGAATTCAAGGTGGAGGAG TGCCATGCGTAAGATGGTTA GAATCTGCCAAGGAGACTCG CAGCATCCAACAAGGCACT TGGAGTCTGGTCGTGTGAAG CACTTCTCTGTCTCCGCACA

images were converted into computer files using an Epson color imaging scanner GT9500 in combination with Adobe Photoshop 3.0 software. Preparation of nuclear extracts and gel mobility shift assay. Nuclear extracts were prepared following the procedure described by Struhl (35). Briefly, cells were washed and harvested in ice-cold PBS and pelleted by centrifugation. The cell pellets were washed once in hypotonic buffer (10 mM Hepes, pH 8.0, 1.5 mM MgCl 2, 10 mM KCl, 0.2 mM PMSF, 0.5 mM DTT) and the cells were resuspended in hypotonic buffer and allowed to swell on ice for 10 min and then homogenized using a Dounce homogenizer. The nuclei were then pelleted by centrifugation at 3300g for 15 min and resuspended in 200 ml of low-salt buffer (20 mM Hepes, pH 8.0, 25% glycerol, 1.5 mM MgCl 2, 10 mM KCl, 0.2 mM EDTA, 0.2 mM PMSF, 0.5 mM DTT). KCl solution (2.5 M) was added to the resuspended mixture dropwise to a final concentration of 0.4 M, and then the mixtures were incubated for 30 min with continuous gentle mixing at 4°C. The extracted nuclei were pelleted by centrifugation at 25,000g for 30 min at 4°C and the resulting supernatant (nuclear extract) was collected and immediately frozen at 270°C. Protein concentrations were determined with a Bio-Rad Coomassie protein assay kit. Gel shift assays were performed in 20 ml of reaction mixture containing 25 mM Hepes (pH 7.5), 4% Ficoll, 40 mM KCl, 0.5 mM DTT, 1 mM MgCl 2, 0.5 mM EDTA, 20% glycerol, 0.5 mg poly(dI– dC), 5 mg nuclear extracts, and 5 fmol 32P-labeled XRE DNA. The oligonucleotide primer sequences for XRE were as follows: 59-GATCCCCCTCGCGTGACTGCGAGCCCC-39 and 59-GATCGGGGCTCGCAGTCACGCGAGGGGG-39. Oligos were annealed and end labeled with T4 polynucleotide kinase using [g- 32P]ATP (;3000 Ci/mmol, Amersham, UK). The reaction mixtures were incubated at room temperature for 1 h and DNA– protein complexes were separated on 4% nondenaturing polyacrylamide gels, in TBE buffer at 4°C. Gels were dried and autoradiographed. Immunofluorescence analysis. Cells were grown in slide flasks, specially designed for immunostaining, for 24 h in phenol red-free cultured media containing charcoal-stripped serum before E 2 was added. After 1 h incubation with 10 nM E 2, cells were washed twice in PBS, fixed in methanol at 220°C for 10 min, and then again washed twice in PBS for 10 min before incubation with ER-a antibody (mouse monoclonal antibody obtained from Affinity Bioreagents, U.S.A.) diluted 1:100 in PBS containing 1% bovine serum albumin. After 4 h incubations, cells were washed three times with

Product size (bp)

Annealing temperature (°C)

146

56

GB K03191

214

60

GB U03688

794

60

GB M18082

540

56

EMB X00351

364

60

GB L19872

322

60

GB M69238

439

60

GB M12674

221

60

DBJ AB006590

Reference

PBS for 10 min each and exposed to TRITC-conjugated goat antimouse IgG (Sigma, U.S.A.) for 1 h. After further washing three times the cells were mounted in fluorescent mounting medium (Kirkegaard and Perry Laboratories, MD) and viewed using an Olympus 1X70 fluorescence microscope. Photographic prints were converted into computer files using an Epson color imaging scanner GT9500 in combination with Adobe Photoshop 3.0 software. Transfections and reporter assays. Cells were seeded onto 24-well tissue culture plates and grown in normal growth medium (for experiments dealing with E 2, cells were grown in phenol red-free media supplemented with 10% FBS that had been pretreated with dextrancoated charcoal). After 24 h, at 50 – 60% confluency, cells were transfected with either 200 ng of pGL3-3(EREc38)-LUC plasmid and 4 ng of pRL-SV40 plasmid (control reporter plasmid obtained from Promega) or 200 ng of pGL3-1(XRE)-LUC plasmid and 4 ng pRL-SV40 plasmid per well using LipofectAMINE plus reagents (Life Technologies, U.S.A.) according to manufacturer’s instructions. In some experiments, ER-a expression plasmids (100 –200 ng/well) were cotransfected with the above two reporter plasmids. Three hours after transfection, medium was exchanged with normal growth medium. Medium was changed again at 24 h after transfection and cells were treated with different chemicals for 24 h. Cells were then washed twice with PBS and incubated in 100 ml of passive lysis buffer for 15 min at room temperature with gentle shaking. Luciferase activities were analyzed for 15 s in 20-ml cell extracts with a dual luciferase reporter assay system (Promega) on a TD 20/20 luminometer (Turners Design) and data represented as relative luciferase activity (the ratio of Firefly to Renilla values). Statistics. Statistical analyses were performed using one-way analysis of variance followed by Student’s t test. A P value of ,0.05 was considered statistically significant.

RESULTS

The impact of TCDD on normal and E 2-dependent cell growth was investigated by cultivating the cells for 8 days with different doses either alone or with 10 nM E 2. TCDD at doses of 10 and 100 nM reduced MCF-7 breast cancer cell growth to 67 and 51%, respectively, of the control level. However, none of these doses had

34

JANA ET AL. TABLE II

Effect of TCDD on Normal and E 2-Induced Proliferation of MCF-7, RL95-2, and KLE Cells Cell proliferation a Treatments

MCF-7

RL95-2

KLE

0.1% DMSO 10 nM TCDD 100 nM TCDD 10 nM E 2 10 nM E 2 1 10 nM TCDD 10 nM E 2 1 100 nM TCDD

100 67* 6 9 51* 6 7 148 6 18 117** 6 12 85** 6 6

100 94 6 8 97 6 14 106 6 15 97 6 6 96 6 11

100 101 6 6 96 6 10 109 6 12 107 6 9 98 6 8

Note. Cells were treated for 8 days with the indicated doses of chemicals. a Values are mean percentages of DMSO control 6 SD; n 5 5. * P , 0.001 compared to DMSO control. ** P , 0.001 compared to E 2-treated group.

any inhibitory effect on RL95-2 or KLE endometrial cell growth (Table II). E 2 at a dose of 10 nM strongly stimulated MCF-7 cell growth to about 148% and this growth stimulation was inhibited to 117 and 85% by 10 and 100 nM TCDD, respectively. E 2 had no growth stimulatory effect on either of the endometrial cells (Table II). Figure 1 shows results for time-dependent effects of TCDD on CYP1A1 mRNA accumulation and EROD activity in RL95-2 and KLE cells. Control cells did not express any detectable levels of CYP1A1 mRNA and EROD activity. Exposure to 10 nM TCDD rapidly increased both CYP1A1 mRNA accumulation and EROD activity time-dependently in the RL95-2 cells, but only a very low level of induction was observed in KLE cells limited to longer exposure times. DMSO (0.1%) also slightly induced CYP1A1 mRNA for 12 h with maximal induction at 6 h, but no EROD activity could be detected. We have carefully performed the time course and found at 24 h there were practically no effects of DMSO on CYP1A1 mRNA levels; therefore, all further experiments were conducted with 24 h exposure to minimize any possible DMSO influence. DMSO also had a similar effect on CYP1B1 induction but had no effect on PAI-2 induction. It is still unclear how DMSO accomplishes this, but a similar observation was also reported in MCF-7 breast cancer cells (15). Figure 2 shows data for the dose-dependent effects of TCDD on CYP1A1 mRNA accumulation and EROD activity. RL95-2 cells exhibited strong dose-dependent induction of CYP1A1 mRNA and EROD activity even with the minimum dose of TCDD at 0.01 nM. However, only a very low level of induction was observed in KLE cells limited to high doses. To test whether other TCDD-responsive genes were also differentially expressed in RL95-2 and KLE cells, we investigated the TCDD-inducible expression of

CYP1B1 and PAI-2 mRNAs. As shown in Fig. 3, the CYP1B1 mRNA was not detected in the control RL95-2 cells but was induced by 1 and 10 nM TCDD. The PAI-2 mRNA was detectable in control cells and induced by TCDD dose-dependently. However, in KLE cells both were constitutively expressed and showed very little stimulation upon exposure of TCDD, only at high doses. Since the induction of CYP1A1 or CYP1B1 gene transcription requires functional liganded nuclear AhR–ARNT complex binding to XRE sequences, we analyzed the level of expression of AhR and ARNT and the binding properties of liganded AhR–ARNT complexes to the XRE sequences in both RL95-2 and KLE cell lines. Both cells expressed ARNT mRNA almost at similar levels; however, expression levels of AhR mRNA were higher in the less responsive KLE cells (Fig. 4A). The levels of expression of both AhR and ARNT were not altered by TCDD for 24 h. Figure 4B

FIG. 1. Time-dependent increase of CYP1A1 mRNA expression and EROD activity in TCDD-treated RL95-2 and KLE cells. Cells were treated with either 0.1% DMSO (D) or 10 nM TCDD (T) for the indicated times and then processed for RNA extraction and EROD assays as described under Materials and Methods. (A) RT-PCR products of CYP1A1 (146 bp) and b-actin (470 bp). (B) EROD activity. Results are means 6 SD from triplicate determinations.

ER-a IN REGULATION OF TCDD-INDUCIBLE GENE EXPRESSION

35

FIG. 3. Dose-dependent increase in expression of CYP1B1 (214 bp) and PAI-2 (794 bp) mRNAs in TCDD-treated RL95-2 and KLE cells. Treatments were the same as described in the legend to Fig. 2.

cells and similar E 2 treatment did not affect nuclear translocation. Transient transfection of E 2-responsive reporter plasmid pGL3-3(EREc38)-LUC and subsequent E 2 (10 nM) treatment for 24 h had no effect on E 2-responsive reporter gene activity on either cell line (Fig. 6). In MCF-7 cells, used as controls, E 2 (10 nM) caused about a 55-fold induction of reporter gene activity. Transient cotransfection of pGL3-3(EREc38)LUC with HEO (an ER-a expression plasmid) and subsequent E 2 treatment significantly increased reporter gene activity in both cell lines (Fig. 6). To test the possibility of whether the defective ER-a and therefore blocked nuclear translocation in the KLE FIG. 2. Dose-dependent increase of CYP1A1 mRNA expression and EROD activity in TCDD-treated RL95-2 and KLE cells. Cells were treated for 24 h with different doses of TCDD and then processed for RNA isolation and EROD assays. (A) RT-PCR products of CYP1A1 (146 bp) and b-actin (470 bp). (B) EROD activity. Values are means 6 SD of triplicate determinations.

illustrates a gel mobility shift assay of nuclear extracts from RL95-2 and KLE cells treated with 1 and 10 nM TCDD for 1 h. Retarded bands of liganded AhR– ARNT–XRE complexes were detectable in both cell lines with increased band intensities in TCDD-treated cells compared to DMSO controls. The specificity of the retarded bands was evaluated by competition experiments with 50-fold excess of unlabeled XRE, which significantly reduced their intensities. Stronger band intensities were observed in less responsive KLE cells than in RL95-2 cells. Figure 5A shows results of the RT-PCR analysis of ER-a and -b expression in the two cell lines. Both RL95-2 and KLE cells expressed ER-a, almost in equal amounts, and TCDD treatment at a maximum dose of 10 nM for 24 h had no effect on their expression levels. Neither of the cell lines expressed ER-b. Immunofluorescence analysis of ER-a demonstrated that in normal RL95-2 cells it was present in both cytoplasm and nucleus and E 2 (10 nM) treatment for 1 h translocated it to the nucleus (Fig. 5B). However, in the KLE, ER-a was predominantly found in the cytoplasm of control

FIG. 4. (A) RT-PCR analysis of AhR (364 bp) and ARNT (322 bp) mRNA transcripts in untreated and TCDD-treated RL95-2 and KLE cells. Total RNA (0.1 mg) was reverse transcribed and amplified using specific primers under the conditions described under Materials and Methods. (B) Gel mobility shift assay of liganded AhR–ARNT complex binding to 32P-labeled XRE. Cells were treated for 1 h with 0.1% DMSO or 1 or 10 nM TCDD. Nuclear extracts were prepared and binding assays were performed as described under Materials and Methods.

36

JANA ET AL.

FIG. 6. E 2 responsiveness of RL95-2 and KLE cells transiently transfected with an estrogen-responsive reporter plasmid, pGL3-3 (EREc38)-LUC. Cells were grown in steroid-free medium and transfected with pGL3-3 (EREc38)-LUC and an internal control plasmid, pRL-SV40, in the presence or absence of the ER-a expression plasmid HEO (100 ng/well). After transfection, cells were treated with either ethanol or 10 nM E 2 for 24 h and then processed for dual luciferase reporter assays as described under Materials and Methods. Results are means 6 SD of two independent experiments, each performed in triplicate. MCF-7 cells were used as positive control.

and subsequent TCDD (10 nM) treatment significantly increased reporter gene activity and this was positively correlated with the amount of ER-a used in the transfection (from 100 to 200 ng/well). TCDD-responsive reporter gene activity was not influenced significantly by cotransfection of ER-a.

FIG. 5. (A) RT-PCR analysis of ER-a (439 bp) and ER-b (221 bp) mRNA transcripts in untreated and TCDD-treated RL95-2 and KLE cells. Total RNA (0.1 mg) was reverse transcribed and amplified using specific primers under the conditions described under Materials and Methods. MCF-7 cells were used as positive control. (B) Immunofluorescence localization of ER-a in RL95-2 and KLE cells. Cells were grown in slide flasks specially designed for immunostaining, fixed, and incubated with ER-a antibody followed by TRITCconjugated goat anti-mouse IgG as described under Materials and Methods. (a) Control KLE cells stained with ER-a antibody. (b) KLE cells treated with 10 nM E 2 for 1 h and then stained with ER-a antibody. (c) Control RL95-2 cells stained with ER-a antibody. (d) RL95-2 cells treated with 10 nM E 2 for 1 h and then stained with ER-a antibody.

cells were the main cause for their minimal responsiveness to TCDD, transient transfection of normal ER-a expression plasmid along with a TCDD-responsive reporter plasmid, pGL3-1(XRE)-LUC, was performed. As shown in Fig. 7, transient transfection of pGL31(XRE)-LUC and subsequent treatment of 10 nM TCDD increased reporter gene activity several fold in MCF-7 and RL95-2 cells, but not in KLE cells. However, cotransfection of ER-a with pGL3-1(XRE)-LUC

FIG. 7. Effects of ER-a on TCDD responsiveness in KLE cells. Cells were grown in steroid-free medium and cotransfected with the TCDD-responsive reporter plasmid pGL3-1 (XRE)-LUC, an internal control plasmid, pRL-SV40, and increasing amounts of the ER-a expression plasmid HEO (100 –200 ng/well). After transfection, cells were treated with either 0.1% DMSO or 10 nM TCDD for 24 h and then processed for dual luciferase reporter assays as described under Materials and Methods. Results are means 6 SD of two independent experiments each performed in triplicate. *P , 0.001 compared to respective control. MCF-7 cells were used as positive control.

ER-a IN REGULATION OF TCDD-INDUCIBLE GENE EXPRESSION

DISCUSSION

In the present study, two uterine endometrial cell lines were applied as a model system to investigate the responsiveness of uterus to TCDD, using mainly CYP1A1 as a marker. The RL95-2 line of moderately differentiated endometrial cells proved highly responsive to TCDD, with dramatic induction of CYP1A1, and also CYP1B1 and PAI-2 upon exposure, wherein the KLE line of poorly differentiated endometrial cells was less responsive and only at high doses. In an attempt to identify the probable factor(s) responsible for this differential response, we first checked the steady-state level of expression of AhR and its dimerization partner. The results of RT-PCR revealed that ARNT mRNA was expressed in both untreated RL95-2 and KLE cells at essentially similar levels, while AhR mRNA expression was greater in the nonresponsive KLE case. These results were confirmed by a gel mobility shift assay finding that showed a stronger signal for nuclear AhR–ARNT–XRE complexes in TCDD-treated KLE cells. Thus, binding of TCDD to AhR, nuclear translocation, and heterodimerization with ARNT and the binding of the liganded AhR–ARNT complex to XRE appeared to be perfectly in order. Thus, the AhR-mediated pathway probably is not responsible for the different response and the involvement of some other factor(s) was indicated. Of course, an impaired transactivation function through the AhR or ARNT in KLE cells could still be one explanation for the low inducibility. Both AhR and ARNT have potent transactivation domains at their C-terminal portion and are responsible for regulation of CYP1A1 expression (36, 37). It has also been reported that TCDD nonresponsiveness in ER-negative MDAMB-231 breast cancer cells is partly associated with expression of a C-terminal variant ARNT protein (38). However, analysis of expression of full-length AhR and ARNT mRNA did not demonstrate any truncated variant form in KLE cells (data not shown). Whether a point mutation in the C-terminal transactivation domain of either AhR or ARNT could have altered the transactivation function remains to be clarified. Although TCDD-induced gene expression requires interaction of functional nuclear AhR–ARNT complexes with genomic recognition sequences called XREs or DREs, there are many other factors that can modulate the induction response. The induction of CYP1A1 in different cell lines is enhanced by cycloheximide, suggesting that a labile protein may play a role in negatively regulating the transactivation process (39). In some cell lines DNA methylation of the CpG dinucleotides within the XRE core sequence (59GCGTG-39) is involved in the silencing of CYP1A1 expression (40). There is also evidence for negative regulatory elements (NRE) identified in the 59-promoter

37

region of the CYP1A1 gene (41) as well as contributions of other trans-acting factors (42– 44). We have studied many of these possibilities including the DNA methylation status, expression, and interaction of upstream stimulatory factor-1 (USF-1) with XRE and the possible role of NRE sequence in the differential expression of genes in these two cell lines (data not shown), but none gave a satisfactory explanation. Thus, the most likely mechanism concerns ER-a function. While expression levels do not differ between cell lines, immunofluorescence analysis identified a defect in KLE cells where it was exclusively present in the cytoplasm. Transient transfection of TCDD-responsive reporter plasmid containing single XRE into KLE cells, in contrast to the RL95-2 case, did not increase TCDD-induced transcriptional activity. However, transient cotransfection of ER-a with TCDD-responsive reporter plasmid to the KLE cells restored TCDD-mediated transcriptional activity, suggesting that ER-a has a positive modulatory role on this process and has general importance in the activation of genes associated with the AhR gene battery. However, inducibility of endogenous CYP1A1 was not restored after transfection of ER-a and subsequent TCDD treatment. This is probably because of transient expression of ER-a. Since the regulation of CYP1A1 gene expression is a very complex process and requires the interplay of various transcription factors, transient expression might not be able to displace negative regulatory elements or facilitate the binding of critical transcription factors to the upstream promoter region. In human breast cancer cell lines, induction of CYP1A1 appears related to their ER-a contents (45) and Ah responsiveness is not only dependent on the expression of AhR but also on ER-a levels (45– 47). ER-negative breast cancer cell lines such as MDA-MB231 and Hs578T are normally Ah nonresponsive but transient transfection of ER-a into these cells restores their Ah responsiveness (46, 47). In-depth studies revealed that both N- and C-terminal transactivation domains of ER-a are responsible for Ah responsiveness (47). The mechanisms are unclear but several possibilities can be speculated: (i) ER-a might interact with liganded AhR–ARNT complexes directly or through some bridging factors; and (ii) it might displace negative regulatory factors or facilitate the binding of critical transcription factors to the upstream promoter region. It has been reported that ER-a does not interact directly with the liganded AhR–ARNT complex (48), but both of these can physically interact with Sp1 protein (49, 50). Whether this interaction really increases the transactivation potential of TCDD-inducible genes remains to be clarified. Though RL95-2 cells proved highly responsive to TCDD rather than KLE in terms of induction of several genes, neither line exhibited any cell growth inhibition.

38

JANA ET AL.

This is consistent with observations of others (51) that downregulation of epidermal growth factor receptor (EGFR) protein by TCDD does not occur in RL95-2 cells. TCDD has been shown to downregulate EGFR in reproductive tissues as well as in several other cell lines (52, 53) partly linked with cell growth inhibition. It is therefore possible that the nonresponsiveness of TCDD regarding cell growth inhibition is due to a failure to influence EGFR. The lack of any growth stimulatory effect of E 2 in either cell line is in agreement with our findings of no direct impact on transactivation of E 2-responsive genes. The E 2 insensitivity of RL95-2 cells has also been reported by others (54, 55). At present, there is no proper explanation for this E 2 insensitivity; it is probably a defect within ER-a (in DNA binding or transactivation domain) since transfection of normal ER-a restored its E 2 responsiveness. TCDD is known to downregulate the expression of AhR (56, 57) and ER (18, 19) in various in vivo and in vitro cell culture systems. However, in these two cell lines TCDD seems to have no major downregulatory effect. Since a noncompetitive RT-PCR has been used, there is a possibility that a small amount of variation cannot be detected. TCDD has been shown to upregulate the expression of CYP1B1 and PAI-2 in RL95-2 cells, which was consistent with the observation of others in the keratinocyte (58) and hepatocarcinoma cell lines (59). However, nonresponsive KLE cells expressed high constitutive levels of CYP1B1 and PAI-2. Constitutive expression of CYP1B1 in KLE cells could be expected to be one of the reasons for the failure of ER-a nuclear translocation since CYP1B1 metabolizes E 2. Probably that is not the case because MCF-7 breast cancer cells also expressed high constitutive levels of CYP1B1 (15), but still it is highly responsive to E 2. The upregulation of PAI-2 by TCDD strongly suggests that the association of TCDD with the incidence of endometriosis might be linked at least in part to its alteration of PAI-2 since recently it has been reported that the PAI-2 concentration in peritoneal fluid is higher in women with endometriosis than in controls (60). Taken together, we conclude that ER-a plays a very important role in the positive regulation of genes associated with the AhR gene battery and is mainly responsible for the different responses of the two uterine cell lines, RL95-2 and KLE, to TCDD. However, TCDD proved completely ineffective at inhibition of cell growth and prompted a major downregulation of AhR and ER in both cell lines. Furthermore, the potential role of ER-a in transactivation of AhR-regulated genes, especially CYP1A1, suggests that TCDD and its related compounds might create an additional burden of carcinogenicity in E 2 target tissues.

ACKNOWLEDGMENTS We thank Professor P. Chambon, IGBMC, INSERM, France, and Dr. C. M. Klinge, University of Louisville, U.S.A., for providing HEO and the pGL3-3(EREc38)-LUC plasmid, respectively. This work was supported in part by grants from the Science and Technology Agency to N.R.J. and S.S. and by grants from the CREST Tohyama Team, JST (M.I., J.Y., C.T., H.S.).

REFERENCES 1. Poland, A., and Knutson, J. C. (1982) Annu. Rev. Pharmacol. Toxicol. 22, 517–554. 2. Safe, S. H. (1995) Pharmacol. Ther. 67, 247–281. 3. Whitlock, J. P. (1993) Chem. Res. Toxicol. 6, 754 –763. 4. Hankinson. O. (1995) Annu. Rev. Pharmacol. Toxicol. 35, 307– 340. 5. Meyer, B. K., Pray-Grant, M. G., Vander-Heuvel, J. P., and Perdew, G. H. (1998) Mol. Cell. Biol. 18, 978 –988. 6. Hoffman, E. C, Reyes, H., Chu, F.-F., Sander, F., Conley, L. H., Brooks, B. A., and Hankinson, O. (1991) Science 256, 1193–1195. 7. Denison, M. S., Fisher, J. M., and Whitlock, J. P. (1989) J. Biol. Chem. 264, 16478 –16482. 8. Kociba, R. J., Keyes, D. G., Beger, J. E., Carreon, R. M., Wade, C. E., Dittenber, D. A., Kalnins, R. P., Frauson, L. E., Park, C. L., Barnard, S. D., Hummel, R. A., and Humiston, C. G. (1978) Toxicol. Appl. Pharmacol. 46, 279 –303. 9. Astroff, B., and Safe, S. (1990) Biochem. Pharmacol. 39, 485– 488. 10. Romkes, M., and Safe, S. (1988) Toxicol. Appl. Pharmacol. 92, 368 –380. 11. Biegel, L., and Safe, S. (1990) J. Steroid Biochem. Mol. Biol. 37, 725–732. 12. Gierthy, J. F., Lincoln, D. W., Gillespie, M. B., Seeger, J. I., Martinez, H. L., Dickerman, H. W., and Kumar, S. A. (1987) Cancer Res. 47, 6198 – 6203. 13. Gierthy, J. F., and Lincoln, D. W. (1988) Breast Cancer Res. Treat. 12, 227–233. 14. Spink, D. C., Lincoln, D. W., Dickerman, H. W., and Gierthy, J. F. (1990) Proc. Natl. Acad. Sci. USA 87, 6917– 6921. 15. Dohr, O., Vogel, C., and Abel, J. (1995) Arch. Biochem. Biophys. 321, 405– 412. 16. Romkes, M., Piskorska-Pliszczynska, J., and Safe, S. (1987) Toxicol. Appl. Pharmacol. 87, 306 –314. 17. Romkes, M., and Safe, S. (1988) Toxicol. Appl. Pharmacol. 92, 368 –380. 18. Harris, M., Zacharewski, T., and Safe, S. (1990) Cancer Res. 50, 3579 –3584. 19. Wang, X., Porter, W., Krishnan, V., Narasimhan, T. R., and Safe, S. (1993) Mol. Cell. Endocrinol. 96, 159 –166. 20. Kharat, I., and Saatcioglu, F. (1996) J. Biol. Chem. 271, 10533– 10537. 21. Gillesby, B. E., Stanostefano, M., Porter, W., Safe, S., Wu, Z. F., and Zacharewski, T. R. (1997) Biochemistry 36, 6080 – 6089. 22. Rier, S. E., Martin, D. C., Bowman, R. E., Dmowski, W. P., and Becker, J. L. (1993) Toxicol. Appl. Pharmacol. 21, 433– 441. 23. Cummings, A. M., Metcalf, J. L., and Birnbaum, L. (1996) Toxicol. Appl. Pharmacol. 138, 131–139. 24. Koninchx, P. R., Braet, P., Kennedy, S. H., and Barlow, D. H. (1994) Human Reprod. 9, 1001–1002. 25. Mayani, A., Barel, S., Soback, S., and Almagor, M. (1997) Hum. Reprod. 12, 373–375.

ER-a IN REGULATION OF TCDD-INDUCIBLE GENE EXPRESSION 26. Mori, H., Sawairi, M., Nakagawa, M., Itoh, N., Wada, K., and Tamaya, T. (1991) Am. J. Reprod. Immunol. 26, 62– 67. 27. Rana, N., Gebel, H., Braun, D. P., Rotman, C., House, R., and Dmowski, W. P. (1996) Fertil. Steril. 65, 925–930. 28. Anttila, S., Hietanen, E., Vainio, H., Camus, A., Gelboin, H. V., Park, S. S., Heikkila, L., Karjalainen, A., and Bartsch, H. (1991) Int. J. Cancer 47, 681– 685. 29. Pyykko, K., Tuimala, R., Aalto, L., and Perkio, T. (1991) Br. J. Cancer 63, 596 – 600. 30. Esteller, M. (1997) Carcinogenesis 18, 2307–2311. 31. Nebert, D. W., and Jones, J. E. (1989) Int. J. Biochem. 21, 243–252. 32. Safe, S. H. (1994) Crit. Rev. Toxicol. 24, 87–149. 33. Klinge, C. M., Silver, B. F., Driscoll, M. D., Sathya, G., Bambara, R. A., and Hilf, R. (1997) J. Biol. Chem. 272, 31465–31474. 34. Pohl, R. J., and Fouts, J. R. (1980) Anal. Biochem. 107, 150 –155. 35. Struhl, K. (1990) in Current Protocols in Molecular Biology (Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Siedman, J. G., Smith, J. A., and Struhl, K., Eds.), pp. 12.1.3–12.1.3, Wiley, NY. 36. Jain, S., Dolwick, K. M., Schmidt, J. V., and Bradfield, C. A. (1994) J. Biol. Chem. 269, 31518 –31524. 37. Whitelaw, M. L., Gustafsson, J. A., and Pollinge, L. (1994) Mol. Cell. Biol. 14, 8343– 8355. 38. Wilson, C. L., Thompsen, J., Hoivik, D. J., Wormke, M. T., Stanker, L., Holtzapple, C., and Safe, S. (1997) Arch. Biochem. Biophys. 346, 65–73. 39. Lusska, A., Wu, L., and Whitlock, J. P. (1992) J. Biol. Chem. 267, 15146 –15151. 40. Takahashi, Y., Suzuki, C., and Kamataki, T. (1998) Biochem. Biophys. Res. Commun. 247, 383–386. 41. Piechocki, M. P., and Hines, R. (1998) Carcinogenesis 19, 771– 780. 42. Watson, A. J., Weir-Brown, K. I., Bannister, R. M., Chu, F. F., Reisz-Porszasz, S., Fujii-Kuriyama, Y., Sogawa, K., and Hankinson, O. (1992) Mol. Cell. Biol. 12, 2115–2123.

39

43. Thomsen, J. S., Nissen, L., Stacey, S. N., Hines, R. N., and Autrup, H. (1991) Eur. J. Biochem. 197, 577–582. 44. Takahashi, Y., Nakayama, K., Itoh, S., Fujii-Kuriyama, Y., and Kamataki, T. (1997) J. Biol. Chem. 272, 30025–30031. 45. Vickers, P. J., Dufresne, M. J., and Cowan, K. H. (1989) Mol. Endocrinol. 3, 157–164. 46. Thomsen, J. S., Wang, X., Hines, R. N., and Safe, S. (1994) Carcinogenesis 15, 933–937. 47. Wang, W. L., Thomsen, J. S., Porter, W., Moore, M., and Safe, S. (1996) Br. J. Cancer 73, 316 –322. 48. Miller, C. A., Martinat, M. A., and Hyman, L. E. (1998) Nucleic Acids Res. 26, 3577–3583. 49. Kobayashi, A., Sogawa, K., and Fujii-Kuriyama, Y. (1996) J. Biol. Chem. 271, 12310 –12316. 50. Wang, F., Hoivik, D., Pollenz, R., and Safe, S. (1998) Nucleic Acid Res. 26, 3044 –3052. 51. Charles, G. D., and Shiverick, K. T. (1997) Biochem. Biophys. Res. Commun. 238, 338 –342. 52. Madhukar, B. V., Brewster, D. W., and Matsumura, F. (1984) Proc. Natl. Acad. Sci. USA 81, 7407–7411. 53. Hudson. L. G., Toscano, W. A., and Greenlee, W. F. (1985) Toxicol. Appl. Pharmacol. 77, 251–259. 54. Liu, Y. H., and Teng, C. T. (1991) J. Biol. Chem. 266, 21880 –21885. 55. Liu, Y. H., Lin, L., and Zarneger, R. (1994) Mol. Cell. Endocrinol. 104, 173–181. 56. FitGerald, C. T., Fernandez-Salguero, P., Gonzalez, F. J., Nebert, D., and Puga, A. (1996) Arch. Biochem. Biophys. 333, 170 –178. 57. Singh, S. S., and Perdew, G. H. (1993) Arch. Biochem. Biophys. 305, 170 –175. 58. Gaido, K. W., and Maness, S. C. (1995) Toxicol. Appl. Pharmacol. 133, 34 – 42. 59. Gohl, G. L. T., Munzel, P. A., Schrenk, D., Viebahn, R., and Bock, K. W. (1996) Carcinogenesis 17, 443– 449. 60. Bruse, C., Bergqvist, A., Carlstrom, K., Fianu-Jonasson, A., Lecander, I., and Astedt, B. (1998) Fertil. Steril. 70, 821– 826.