Estrogen Receptor Activation Via Activation Function 2 Predicts Agonism of Xenoestrogens in Normal and Neoplastic Cells of the Uterine Myometrium1

[CANCER RESEARCH 59, 3090–3099, July 1, 1999] Estrogen Receptor Activation via Activation Function 2 Predicts Agonism of Xenoestrogens in Normal and Neoplastic Cells of the Uterine Myometrium1

Deborah S. Hunter, Leslie C. Hodges, Peter M. Vonier, Robin Fuchs-Young, Marco M. Gottardis, and Cheryl L. Walker2 The University of Texas M. D. Anderson Cancer Center, Science Park Research Division, Smithville, Texas 78957 [D. S. H., L. C. H., R. F-Y., C. L. W.]; The Tulane-Xavier Center for Bioenvironmental Research, Tulane University, New Orleans, Louisiana 70112 [P. M. V.]; and Bristol Myers Squibb, Endocrine Oncology Pharmaceutical Research Institute, Princeton, New Jersey 08543 [M. M. G.]

ABSTRACT compounds from a biochemical standpoint, and yet their physiological effects are not predictable solely on the basis of structure or receptor The possible contribution of endocrine disrupters to human disease, affinities. It is known that a given ligand can mimic estrogen function particularly those compounds that modulate the estrogen receptor (ER), (agonist), oppose it (antagonist), or have a mixed function; the agonist/ has recently drawn considerable attention. The tissue specificity of effects mediated by the ER is well recognized, although the mechanism of this antagonist dichotomy is determined, to a large extent, by the target tissue, specificity is not understood sufficiently to predict the effects of a partic- but the molecular determinants have yet to be fully defined. ular ligand in different target tissues. Although the divergence of ER- Endogenous estrogens have critical functions not only in the female mediated effects in the breast, bone, and uterine endometrium has been reproductive tract and mammary gland but also in bone, the cardio- described, a frequently overlooked site of estrogen action is the smooth vascular system, and the central nervous system during development muscle of the uterus. The uterine myometrium is the tissue of origin of an and throughout life. There is controversy about whether estrogen extremely common hormone-responsive tumor, uterine leiomyoma, a tu- replacement therapy after the onset of menopause puts women at mor with a significant impact on women’s health and a possible environ- increased risk for hormonally responsive cancer, even as it protects mental influence. This report describes an in vitro/in vivo system for identifying the effects of ER ligands in the myometrium and elucidating against bone loss and cardiovascular disease (3). For breast cancer their mechanism of action. Several natural and synthetic xenoestrogens patients and for women at high risk for breast cancer development, were evaluated at the cellular and molecular level for their ability to antiestrogen therapy has proven beneficial, although opposing estro- mimic estrogen action in uterine myometrial tissues. Diethylstilbestrol, gen action is detrimental in some tissues. As a result, considerable coumestrol, genistein, naringenin, and endosulfan were able to activate the energy has been focused on developing agents with the capacity to AF2 function of the ER in vitro and demonstrated agonist activity in modulate hormone activity in a tissue-specific manner (4, 5). A estrogen-responsive myometrial cells, as determined by induction of pro- well-known example of differential effects of a synthetic ER ligand is liferation and increased message levels of progesterone receptor. Com- the compound tamoxifen, which is used clinically to inhibit estrogen- pounds that could not activate AF2 function (4-hydroxy-tamoxifen, LY117018, and LY317783) did not act as estrogen agonists. For agonists, responsive breast cancer and is currently undergoing clinical trials for rank order of potency was predicted by receptor affinity; however, en- use as a prophylactic agent. Although tamoxifen is an effective dosulfan displayed a surprising degree of activity, despite negligible re- estrogen antagonist in mammary tissue, it can act as an estrogen ceptor binding. Additionally, diethylstilbestrol and tamoxifen demon- agonist in the uterine endometrium, an effect that is observable both strated prototypical agonist and antagonist effects, respectively, in the at the biochemical level and biologically in an increased risk of intact myometrium of sexually mature rats. The results presented here endometrial carcinoma in women undergoing tamoxifen therapy (6). suggest that some exogenous ER ligands may mimic the effects of endog- Thus, a beneficial modulation of ER-mediated effects in one tissue enous estrogens on uterine leiomyoma and may contribute to a complex may have profound and sometimes adverse effects in another tissue. hormonal milieu that impacts both normal and neoplastic myometrium. Investigations of the mechanism of the tissue-specific effects of tamoxifen have had a tremendous impact on our understanding of INTRODUCTION ligand-dependent activation of the ER as well as the effects of estrogen agonism and antagonism in mammary tissue. This knowledge, Compounds with the biochemical capacity to disrupt normal endo- however, has not translated into an equivalent understanding of ER crine function are numerous and abundant in the environment and in ligand effects in other tissues. Relatively little is known about the dietary sources and are used as pharmacological agents. The adverse effect of ER modulation in the myometrial compartment of the uterus sequelae of exposure to compounds that mimic or oppose steroid (7, 8). The uterine myometrium is the site of an extremely common hormone function include developmental abnormalities, reproductive neoplasm, uterine leiomyoma, also known as fibroids. These tumors dysfunction, and proliferative disorders, including malignancies (1, occur in as many as 77% of adult women (9), an incidence that is 2). The mechanism by which endocrine disrupters produce these arguably higher than any other type of neoplasm. Despite the benign disorders is poorly understood, despite considerable study. Exogenous nature of these tumors, leiomyoma may result in abnormal bleeding, 3 ER ligands are among the most thoroughly investigated of xenobiotic abdominal pain, and infertility, making them a frequent cause of gynecological problems (10). Leiomyomas are responsive to ovarian Received 12/29/98; accepted 5/3/99. hormones and can be reduced in volume by the maintenance of a The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with hypo-ovarian hormonal milieu using gonadotropin-releasing hormone 18 U.S.C. Section 1734 solely to indicate this fact. agonists; however, this therapy rapidly results in adverse effects in 1 This work was supported in part by NIH Grants CA72253 and ES06658 (to C. L. W.) and ES07784, CA094480, and CA16672. other tissues, particularly a dramatic loss in bone density, that is not 2 To whom requests for reprints should be addressed, at The University of Texas M. D. completely reversible (11). Surgical intervention, often necessitating Anderson Cancer Center, Science Park Research Division, Box 389, Smithville, TX hysterectomy, is generally the only treatment option. As a result, 78957. Phone: (512) 237-9525; Fax: (512) 237-2475. 3 The abbreviations used are: ER, estrogen receptor; PR, progesterone receptor; DES, leiomyomas are the most common cause of hysterectomy in the ␤ diethylstilbestrol; E2,17-estradiol; OH-tam, 4-hydroxy-tamoxifen; ELT, Eker rat United States, accounting for Ͼ250,000 surgeries annually. leiomyoma tumor-derived; hER, human ER; GAPDH, glyceraldehyde-3-phosphate dehy- drogenase; ␤-gal, ␤-galactosidase; ERE, estrogen response element; AF, activation func- Recent increases in the incidence of hormonally related cancers tion; BrdUrd, bromodeoxyuridine. have stimulated research into a potential role for endocrine disrupters in 3090

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1999 American Association for Cancer Research. ER AGONISM OF XENOESTROGENS IN UTERINE MYOMETRIUM cancer development (12). The striking frequency with which leiomyomas were grown in Grace’s Insect Medium (Mediatech, Inc., Herndon, VA), are observed calls into question the influence of environmental factors, supplemented with 10% fetal bovine serum (Life Technologies, Inc., Grand including exogenous ER ligands. Unanswered questions have emerged Island, NY). Confluent Sf9 cells in 150-cm2 flasks were infected with the hER regarding whether estrogen agonists are potential promoters of leiomy- baculovirus and incubated for 5 days at 25°C. Cells were then suspended in a oma growth and whether exposure to an exogenous ligand can either buffer consisting of 20 mM Tris (pH 7.4), 1 mM EDTA, 1 mM EGTA, 10 mM NaPO ,50mM NaF, and 1 mM NaVO , and cells were lysed with three rounds predispose to or protect against the development of leiomyoma. To date, 4 4 of freeze-thaw. The extracts were brought to 400 mM NaCl and 10% glycerol the effects of xenoestrogens on the myometrium have not been evaluated. and incubated for 20 min on ice. The whole cell extract was prepared by Here, we have used the Eker rat model of spontaneous uterine leiomyoma centrifugation for 10 min at 15,000 ϫ g. For preparation of ammonium sulfate (13) to investigate the effects of a panel of suspected ER modulators on precipitates, the whole-cell extract was brought to 40% ammonium sulfate and the uterine myometrium. The nine compounds selected represent a vari- incubated on ice for 30–60 min. The precipitates were collected by centrifu- ety of synthetic and naturally occurring chemicals, of varied structure, gation at 15,000 ϫ g for 5 min and flash-frozen for later use. that have not been characterized for their effects on the myometrium. Competition Binding Assays. Recombinant hER (in ammonium sulfate Some have been previously categorized as estrogen agonists or antago- precipitates) was dissolved in binding buffer consisting of 10 mM Tris (pH nists in other tissues (Table 1). Each compound was tested for its affinity 7.4), 1 mM EDTA, 1 mM EGTA, 10 mM NaPO4,50mM NaF, 1 mM NaVO4, ␥ for and ability to activate ER␣ in vitro, and its ability to induce prolif- 10% glycerol, 10 mg/ml -globulin, 0.5 mM phenylmethylsulfonyl fluoride, 30 mg/ml aprotinin, and 0.2 mM leupeptin. Reactions were incubated at 25°C for eration in an estrogen-sensitive myometrial tumor-derived cell line. Ad- 1 h in the presence of 32 nM 17␤-[3,4,6,7-3H]estradiol (84 Ci/mmol; DuPont- ditionally, compounds demonstrating ER-mediated transactivation were NEN Products, Boston, MA) and radioinert chemicals or vehicle (DMSO). tested for their ability to induce expression of an endogenous estrogen- Reactions contained a final volume of 102 ml. Reactions were terminated by responsive gene, the PR, in myometrial cells. Finally, a prototypical incubation with a 50% slurry of hydroxyapatite (Bio-Rad, Hercules, CA) for agonist (DES) and a partial agonist/antagonist (tamoxifen) were evalu- 10 min at 4°C and centrifugation for 1 min at 15,000 ϫ g. Samples were ated for their effects on the normal myometrium in an in vivo assay. Six washed three times in binding buffer plus 0.05% Triton X-100. After three of nine xenoestrogens in this panel exhibited agonism in one or more of washes, the samples were resuspended in 100% ethanol, and the bound [ 3H]E2 these assays. The mechanism of in vitro transactivation by these com- was measured using liquid scintillation counting. The data are representative of pounds was correlated to agonist activity to define the molecular deter- two independent experiments with three replicates. minants of estrogen agonism in myometrial tissue. Proliferation Assays

MATERIALS AND METHODS Cells were plated into 24-well plastic cell culture dishes (Corning, Corning, NY) at 2000–3000 cells per well in DF8 medium and allowed to proliferate for Chemical Compounds 48–72 h. The media was then aspirated, and the wells were washed twice with 1ϫ PBS, followed by the addition of DF8-Basal medium containing the test com- Test compounds were solubilized in either ethanol or DMSO. Sources of pounds. Either ethanol or DMSO was used as a vehicle control, depending on the compounds were as follows: E , OH-tam, naringenin, and DES were from 2 solubility of the test compound. The relative volume of the vehicle plus test ␣ ␤ Sigma Chemical Co. (St. Louis, MO); endosulfan isomers and were from compound did not exceed 0.1% of the solution. At each time point, the medium Chem Service (West Chester, PA); genistein was from Aldrich Chemical was aspirated and the cells were rinsed twice with 1ϫ PBS and detached with Company (Milwaukee, WI); coumestrol was from Acros Organics (Fairlawn, trypsin-EDTA (Life Technologies, Inc.). Cells were resuspended in DF8 medium NJ); and LY117018 and LY317783 were both gifts from Eli Lilly and Co. and counted with a Coulter Counter (Coulter Electronics, Hialeah, FL). (Indianapolis, IN). Northern Analysis Cell Lines Cells were plated in DF8 medium and switched to DF8-Basal medium 8–12 ELT3 and ELT6 rat uterine leiomyoma cell lines (previously characterized; h before the addition of chemicals. The following concentrations of chemicals Ref. 24) were maintained in DF8 medium supplemented with 10% FCS were administered in basal medium with 0.1% ethanol or DMSO: DES, 10 nM; (Hyclone Laboratories Inc., Logan, UT), as described previously (24), at 37°C 4-OH-tam, 1 nM; naringenin, 10 ␮M; LY317783, 1 nM; LY117018, 1 nM; and 5% CO . Experiments in which cells were grown in the presence of test 2 genistein, 10 ␮M; coumestrol, 1 ␮M;E,10nM; endosulfan-␣, 100 nM; and compounds were conducted using serum-free, phenol red-free DF8-Basal 2 endosulfan-␤, 100 nM. These doses had previously produced stimulation in the medium (containing 1% BSA; Sigma). proliferation assays; if no proliferative dose had been observed, the highest Ligand Binding Studies dose that did not significantly inhibit proliferation was chosen for this experiment. Cells were harvested after 24 h from logarithmic-phase cultures; RNA Preparation of Receptor. Recombinant hER was produced in Sf9 insect was extracted as described previously (26). Five ␮g of polyadenylated RNA cells and prepared as ammonium sulfate precipitates (25). Briefly, Sf9 cells were isolated by oligodeoxythymidylate cellulose chromatography and sepa-

Table 1 Test panel of xenoestrogens Compound Chemical characteristics Estrogenic effects Ref. DES Synthetic hydroxystilbene drug previously used to prevent miscarriages Increased risk of adenocarcinoma of the vagina in women treated in utero 14 Promotes structural and functional defects in the reproductive tracts of 15 mice treated in utero OH-tam Synthetic triphenylethylene; active form of therapeutic compound Exhibits tissue and species specific agonistic behavior 16 tamoxifen LY117018 Synthetic benzothiophene; raloxifene analogue Antiestrogenic in breast and endometrium 17 Estrogenic in bone and cardiovascular system LY317783 Synthetic napthalene; raloxifene analogue Same as LY117018 17 Genistein Natural isoflavone present in soybeans Concentration-dependent effect on growth of breast cancer cells 18 Naringenin Natural hydrogenated flavone present in citrus fruits Inhibits proliferation of breast cancer cells 19 Inhibits E2-induced hypertrophy of uterine wet weight of immature rats 19 Coumestrol Natural coumestan present in some grains Stimulates growth of breast cancer cells in vitro 20 Stimulates transcriptional activity of hER 21 Endosulfan Chlorinated hydrocarbon present in some pesticides; ␣ and ␤ isomers Promotes enlarged uteri in pregnant rats 22 Estrogenic on MCF-7 cells 23 3091

rated by gel electrophoresis on 1% agarose with formaldehyde. The RNA was of sacrifice was determined by vaginal and ovarian histology (30) or by ␮ transferred to a 0.45- m nitrocellulose membrane (Schleicher & Schuell, evaluation of serum levels of E2 by RIA (Diagnostic Systems Laboratories, Dassel, Germany) by standard methods (27). The membrane was then baked, Webster, TX) according to the manufacturer’s instructions. BrdUrd immuno- prehybridized, and hybridized at 42°C for 48–72 h in 40% formamide-3ϫ staining was performed as described previously (24). Care and handling of rats SSC. The cDNA probes for rat PR (28) and rat GAPDH were labeled by were according to NIH guidelines in facilities accredited by the Association for random priming (Prime-It; Stratagene, La Jolla, CA) with [␣-32P]dCTP. The the Accreditation of Laboratory Animal Care. membrane was washed with four changes of 0.5ϫ SSC-0.1% SDS at 45°C for 2 h. The membrane was dried and exposed to X-ray film with an intensifying RESULTS screen at Ϫ70°C. The relative increase or decrease of the intensity of signal was quantitated by absorbance using Bioimager (Millipore, Ann Arbor, MI) in ER Binding Affinities. To determine the ability of these xe- each lane and compared to the intensity of signal from cells treated with noestrogens to bind ER, we performed competitive binding analysis vehicle. using recombinant hER␣ produced in an Sf9 baculoviral system.

Reporter Gene Assays Dissociation of radiolabeled E2 from the hER was used to accurately rank the relative binding affinities of compounds in the study (Fig. 1). ELT3 cells were transiently cotransfected with the vitERE-tk-LUC plasmid IC50s (the concentration of compound necessary to displace 50% of or C3-tk-LUC (29), along with hER, and a constitutive (cytomegalovirus bound E ) were determined from the displacement curves. OH-tam, promoter-driven) ␤-gal reporter (gift from Dr. Andrew Butler, University of 2 DES, and LY317783 demonstrated affinities for the receptor that were Texas, M.D. Anderson Cancer Center). Cells were plated in 12-well plates at 25,000 cells per well one day before calcium phosphate transfections. Calcium 5–6-fold that of E2, whereas LY117018 had an affinity that was 2-fold ϳ phosphate precipitates are prepared using standard procedures (27) using a that of E2. Genistein and coumestrol had 100-fold less affinity, and ϳ solution of 2 ␮g per well of plasmid DNA (at a reporter:receptor:␤-gal ratio of naringenin had 3000-fold less affinity for the receptor than E2. The 9:1:1). After transfection, cells were washed with PBS, and phenol red-free, IC50 could not be determined for the endosulfan isomers because serum-free basal medium was added, along with test compound or 0.1% displacement of E2 from the receptor was minimal at the limits of vehicle control. Transfected cells were incubated with test compound for 40 h, compound solubility. With this information, it was possible to rank then washed, lysed, and harvested for assay of reporter gene activity using order the panel of potential ER ligands (Table 2) based on receptor Tropix (Bedford, MA) chemiluminescence reporter assays. Luminescence was affinity as compared to E2, given an arbitrary value of 100. detected using a Dynex-MLX luminometer (Chantilly, VA), and luciferase In Vitro Proliferation. ELT3 cells have been shown previously to values for each of triplicate assay wells were normalized to ␤-gal values for the proliferate in response to E in culture (31). The ability of xenoestro- same well as a control for transfection efficiency. 2 gens to induce proliferation in this ER-positive cell line was assessed In Vivo Administration of OH-tam and DES over logarithmic and semilogarithmic dose ranges. Results of proliferative assays are shown in Fig. 2. The effective concentrations for Female, non-mutation-bearing Eker (Long Evans) rats were maintained on compounds inducing proliferation varied from picomolar to micro- a 12-h light/dark cycle with food and water provided ad libitum. Sexually molar ranges, over which a dose response was generally observed mature, 2.5–5-month-old rats were treated with tamoxifen via a 100-mg time-release s.c. implant, 25-mg DES implant, or vehicle control implant. After until an inhibitory threshold was reached. The relative potency of 60 days of treatment, animals were injected with BrdUrd (Sigma) at 100 mg/kg proliferative compounds was assessed by determining the lowest dose with the ability to induce proliferation at a level that was Ն25% that and euthanized 2 h later by CO2 asphyxiation. Cardiac blood and reproductive tract tissues were harvested from these animals; tissues were fixed in neutral of 10 nM E2 during logarithmic growth. By these criteria, DES, buffered formalin for at least 48 h. Reproductive stage of animals at the time coumestrol, genistein, naringenin, and endosulfan-␣ exhibited agonist

3 Fig. 1. Inhibition of tritiated-E2 binding to Sf9 cell extract by radioinert chemicals. Recombinant hER was incubated with 32 nM [ H]E2 in the presence of increasing concentrations of radioinert chemicals. Data points, two independent experiments of three replicates each. Inset,IC50s represent the concentration of competitor that displaced 50% of receptor-bound Ϯ radiolabeled E2 the range of the means of two independent experiments with three replicates each. 3092

Table 2 Comparison of proliferative stimulation, transactivation, and ER binding of affinitya

Compound Proliferative EC25 Compound Transactivation EC50 Compound Relative binding affinity

E2 10 pM E2 100 pM LY317783 610 DES 100 pM DES 1 nM DES 550 Coumestrol 10 nM Coumestrol 100 nM OH-tam 490 Genistein 100 nM Genistein 1 ␮M LY117018 200 ␮ ␮ Naringenin 10 M Naringenin 10 M E2 100 Endosulfan-␣ 10 ␮M Endosulfan-␣ 10 ␮M Coumestrol 2.4 Endosulfan-␤ None Endosulfan-␤ 10 ␮M Genistein 0.8 OH-tam None OH-tam None Naringenin 0.036 LY317783 None LY317783 None Endosulfan-␣ ––– LY117018 None LY117018 None Endosulfan-␤ ––– a Compounds are shown in rank order of potency in each assay. Proliferative EC25 was the lowest dose achieving 25% of 10 nM E2 activity in proliferation assay, calculated from triplicate wells. Transactivation EC50 was the lowest dose achieving 50% of 10 nM E2 activity. (The mean 10 nM E2 was calculated from two to five experiments in which the test 3 compound was analyzed.) Relative binding affinity was calculated from the displacement of [ H]E2 relative to unlabeled E2, which was arbitrarily designated as 100.

activity by inducing proliferation in ELT3 cells, and endosulfan-␤, high concentrations required to induce transactivation. Rank order of LY117018, LY317783, and OH-tam did not. potency in transactivation (Table 2), which was determined by the For genistein, which stimulated proliferation at concentrations of average compound dose that produced a fold-induction at least 50% Ͻ ␮ 10 M, a dramatic antiproliferative effect was seen in the 40–100 that of E2, coincided with the order of compounds as stimulators of ␮M concentration range (Fig. 2 and data not shown). A similar proliferation. inhibition has been previously reported in MCF-7 cells (18). Treat- Previous studies have shown that the ability of “partial” or “mixed” ment of ER-negative ELT6 cells with genistein (Fig. 3) revealed a agonists, such as tamoxifen, to mediate ER transactivation is specific similar growth suppression in this range of concentrations, but pro- to both promoter context and cell type and is mediated by the AF1 liferative stimulation was not observed at any dose. These data sup- domain of the ER (35, 36). A prototypical AF1-responsive promoter port previous indications that agonist activity of genistein is ER is the C3 (complement protein-3) promoter, which can be transcrip- mediated, whereas antiproliferative effects at the higher dose are tionally activated in vitro and in vivo in some cell types by OH-tam independent of the ER and may be mediated, instead, by inhibition of (37). To determine whether xenoestrogens could function as agonists tyrosine kinase or topoisomerase activity (18, 32–34). in the context of C3 transactivation in a myometrial cell background, Reporter Gene Transactivation. To determine whether xe- we cotransfected ELT3 cells with hER␣ and the C3-tk-LUC reporter

noestrogens are able to directly activate gene transcription of a clas- plasmid and then treated them with varying concentrations of E2 and sical estrogen-responsive promoter element (the vitellogenin ERE) in other ligands. E2 was minimally but reproducibly able to induce a myometrial cell background, ELT3 cells were transiently transfected expression of this reporter in repeated assays (Fig. 5). In contrast, in

with a firefly luciferase reporter gene driven by a single vitellogenin control experiments, C3-tk-LUC could be induced 16-fold by E2 in ERE linked to a minimal viral promoter. This ERE has been previ- Ishikawa cells derived from the human uterine endometrium (data not ously characterized in other cell types as requiring the function of both shown), confirming that the responsiveness of this promoter is gen- activation domains of ER␣ (AF1 and AF2; Ref. 29). Transcriptional erally low in the myometrial cell background. The ability of xe- activation of most estrogen-responsive genes has been shown to noestrogens to transactivate C3-tk-LUC is shown in Table 3, in order require both of these transactivation functions. Along with this re- of concentration yielding the maximum response in a representative ␣ ␤ ␤ porter (vitERE-tk-LUC), plasmids containing hER and -gal (CMV- experiment, with E2 being most potent and endosulfan- being the ␤–gal) under the control of constitutive promoters were cotransfected least potent. These data are compared to the fold-induction observed with the respective functions of increasing the measurable transcrip- using the vitERE-tk-LUC reporter at the same concentration of test tional response and normalizing for transfection efficiency. Normal- compound. Although the level of induction varied in magnitude, the

ized luciferase activity in treated cells was reported as fold induction concentrations of E2, coumestrol, naringenin, genistein, and the en- over baseline activity of the reporter in untreated cells that were dosulfans that were required for maximal stimulation remained or- identically transfected. Activity was tested over increasing logarith- dered roughly in the same range as activation of vitERE-tk-LUC. In mic and semilogarithmic dose ranges, increasing until an inhibitory repeated experiments, OH-tam, LY317783, and LY117018 were con- dose was reached or until the limits of compound solubility resulted in sistently able to induce C3-tk-LUC activity over basal levels and did exceeding the upper bound of 0.1% vehicle. The results of multiple not exhibit the suppression of activity observed with the AF2-requir- ␮ experiments are summarized in Fig. 4. DES and E2 were roughly ing vitERE-tk-LUC, even at doses as high as 100 M. The consistent equipotent at inducing transactivation of the reporter, at a range of ability of these xenoestrogens to induce C3-tk-LUC activity over doses 3 logs below that of coumestrol or genistein. Naringenin and the basal levels contrasts with that of the pure antiestrogen ICI182780 that endosulfan isomers were able to activate the receptor at dose ranges did not produce any increase in C3-tk-LUC reporter gene activity in that were ϳ100 times higher than coumestrol or genistein. Notably, ELT3 cells (data not shown). These results suggest that ER-mediated although the effective doses were nearly identical for naringenin and transcriptional activity of the C3 promoter in cells of myometrial the endosulfans, the maximum values for naringenin were consistently origin is reduced relative to the endometrium and that agonists and higher. Examination of the ␤-gal values (data not shown) for these nonagonists in this cell type are equally capable of stimulating tran- treated cells revealed an inhibitory effect by endosulfans at doses of scription via this promoter. In this tissue, estrogen agonism could not Ͼ5 ϫ 10Ϫ5, indicating a general suppression of gene expression at be discriminated at the level of AF1 activity and was better predicted this dose level. LY317783, LY117018, and OH-tam could not induce by transactivation via the AF2-requiring vitellogenin ERE. transactivation at any dose tested; a 30–90% inhibition of baseline Induction of PR. Compounds that induced transactivation of vit- luciferase activity was observed for each of these compounds, irre- ERE-tk-LUC were tested for their ability to up-regulate the endoge- spective of dose. All xenoestrogens with the ability to activate vit- nous PR to verify that transactivation could occur in the context of an ERE-tk-LUC displayed agonist activity in the proliferation assay, with endogenous promoter. PR induction reflects ER-mediated gene regu- the exception of endosulfan-␤, which inhibited proliferation at the lation, which is often but not always indicative of estrogen agonism 3093

Fig. 2. Effects of test compounds on proliferation in the estrogen-responsive ELT3 cell line. Growth kinetics were determined by counting cell number in triplicate wells of vehicle, 10 nM E2, or varying concentrations of test compound in basal medium. Cell counts were taken on successive days prior to confluence. Each graph is a representative experiment that was independently .significant difference (at P Յ 0.05) from control values by Student’s t test of the logarithm of the cell number ,ء .repeated up to three times. Data points, means of triplicate samples; bars, SE

(38, 39). ELT3 cells were incubated with the lowest dose of com- In Vivo Proliferation of Myometrium. To determine whether the pound that consistently induced proliferation, and PR message levels proliferative or antiproliferative activity of xenoestrogens observed in in these cells were compared with that of cells grown in basal medium vitro could also occur in intact myometrial tissue, a prototypical alone. The Northern blot shown in Fig. 6 is representative of repeated agonist (DES) and partial agonist/antagonist (tamoxifen) were evalu- experiments showing increased levels of PR message by E2, DES, ated in an in vivo assay. Young, sexually mature female Eker rats were coumestrol, genistein, naringenin, and endosulfan after normalization exposed to either tamoxifen (which is converted in vivo to OH-tam; to GAPDH. In repeated experiments (Fig. 6 and data not shown), Ref. 40), DES, or vehicle control via 60-day time release s.c. implant. densitometric quantitation and normalization of the 7.2- and 3.5-kb At the conclusion of treatment, reproductive tract tissues were exam- bands suggested increased message levels for each of these com- ined for changes in histology and for the presence of proliferating pounds, at levels varying between 2- and 20-fold, with the absolute 4 cells in the myometrial layers. In previous studies (41), we and others level depending on which PR band was measured. These results confirm the up-regulation of an endogenous estrogen-responsive gene 4 K. D. Burroughs, R. Fuchs-Young, B. Davis, and C. L. Walker. Altered hormonal in myometrial cells by compounds capable of inducing vitERE-tk- control of proliferation and apoptosis during myometrial maturation and the development LUC reporter gene activity. of uterine leiomyomas, submitted for publication. 3094

larged, with profound disruption of the normal tissue architecture. The appearance of the uterus in DES-treated animals was variable. In some regions, the myometrial and stromal layers of the uterine wall were reduced to a thickness of a few cell layers, concomitant with highly dilated lumen. In other areas, both circular and longitudinal layers of the myometrium were enlarged to several times their normal size, with only the stromal layer being diminished in thickness. In DES-treated animals, myometrium was highly proliferative, as evidenced by in- corporation of BrdUrd. Although DES and OH-tam might be expected to interfere with normal cycling, all treated animals showed evidence

of recent ovulation. Endogenous E2 levels in DES-treated animals were in the range of 8–9 ng/ml, well below the levels of 20–40 ng/ml, which occur during proestrus and are associated with luminal enlargement and myometrial proliferation in untreated animals (41),4 indicating that the observed proliferation in the myometrium was due to DES treatment rather than the influence of endogenous estrogens.

DISCUSSION

Fig. 3. Treatment of ER(Ϫ) ELT6 cells with genistein. ELT6 cells were treated for 7 Estrogen agonist activity of DES, coumestrol, genistein, narin- ␣ ء ␮ days with 0.1% vehicle, 10 nM E2, or 10–100 M genistein in basal medium. , significant genin, and endosulfan- was demonstrated by the dose-dependent difference from control by Student’s t test at P Ͻ 0.05. Data shown are representative of proliferation of ELT3 tumor-derived uterine myometrial cells. Addi- two separate experiments. tionally, each of these compounds was able to transactivate an AF2- requiring reporter gene and increase expression of endogenous PR in have demonstrated that myometrial cells proliferate in untreated fe- ELT3 cells, consistent with ER-mediated agonist activity in this cell male rats only during proestrus, the period of highest levels of type. Proliferation was not observed in response to the synthetic Ͼ endogenous ovarian steroids ( 20 pg/ml E2; Ref. 42). Animals sac- ligands OH-tam, LY117018, or LY317783. None of these three com- rificed during proestrus exhibit the uterine morphology that is typical pounds activated the AF2-requiring, vitERE-tk-LUC reporter gene, during this stage: an enlarged lumen, columnar luminal epithelium, but all could stimulate an AF1-requiring C3-tk-LUC reporter to some and a limited degree of myometrial proliferation. Controls sacrificed extent, suggesting that AF2-mediated transactivation is required for during other stages of the reproductive cycle are virtually devoid of agonism in this cell type and is a better predictor of agonist activity in proliferating cells in the myometrium. Uteri from DES- and tamox- this tissue than transcriptional activation mediated by AF1. ifen-treated animals are shown in Fig. 7. As compared to controls, Comparison of ER-binding affinity, stimulation of proliferation, uteri from each of three tamoxifen-treated animals were unusually and transactivation in estrogen-responsive leiomyoma cells showed small, with an immature, quiescent appearance, and they exhibited a that the rank order for potency of compounds as transactivators and complete absence of proliferation in the myometrium. In contrast, mitogens was identical. Interestingly, ER transactivation was uteri from four of five DES-treated animals were dramatically en- achieved by both endosulfan isomers, despite negligible receptor

Fig. 4. Reporter gene activity using vitERE-tk-LUC. Shown is a summary of in vitro transactivation experiments using the vitellogenin ERE reporter in ELT3 Eker rat leiomyoma cells. Luminescence data were calculated as: fold induction ϭ mean normalized luciferase activity in triplicate treated wells/mean normalized luciferase activity in triplicate wells with basal medium only. Data points, mean fold inductions from three to six separate experiments. 3095

may not require interaction with the ER (44). In this investigation, in vitro transactivation by endosulfans suggests mediation by ER␣ and classical response elements, although an indirect mechanism could be involved. For example, it is possible that these compounds can increase the activity of ligand-free receptor; however, this type of mechanism is difficult to substantiate. Although reporter and proliferation assays measure the effects of xenoestrogens cultured cells, in vivo administration of two compounds, one that exhibited agonist activity in ELT3 cells (DES) and one that did not (tamoxifen), showed that the intact myometrium of a normal female rat could also be influenced by xenoestrogen exposure. These results imply that agonistic or antagonistic effects of exogenous ER ligands observed in assays using tumor-derived myometrial cells may also be relevant to the normal myometrium of an intact animal. Together, these results suggest that exogenous ER ligands can act as estrogen agonists in myometrial cells. Although environmental estrogens have been implicated epidemiologically and experimentally as risk factors for both neoplasia and developmental disorders (45), no environmental compound has been firmly implicated in the etiology of uterine leiomyoma. Given the high incidence of these tumors, an Fig. 5. Reporter gene activity using C3-LUC. In vitro transactivation of the C3-LUC increase in tumor frequency in an exposed population would be reporter by E2. Each treatment was tested in triplicate in each of three experiments. difficult to detect. Previous studies have shown, however, that women significance at ,ء .Columns, mean fold induction of the three experiments; bars, SE P Յ 0.05 by Student’s t test. harboring uterine leiomyomas have higher serum burdens of some organochlorine pesticides than unaffected women and that tissue organochlorine levels are higher in leiomyomas than in normal myo- Table 3 Reporter gene activity of test compounds using C3-tk-LUC compared to vitERE-tk-LUC a metrium (46). A common feature of organochlorines is their persist- ence and tendency to accumulate in body lipid, a characteristic that Maximum fold Fold induction at induction, equivalent dose, makes them arguably suspect as etiological agents in cancer of repro- Compound Concentration C3-tk-LUC vitERE-tk-LUC ductive tissues. In contrast, dietary phytoestrogens are generally con-

E2 1nM 1.8 10.3 sidered antitumorigenic, despite their well-demonstrated estrogenicity LY317783 10 nM 1.7 0.15 in tissues such as the breast (21, 47). Historically, phytoestrogen LY117018 10 nM 1.5 0.15 OH-tam 10 nM 1.4 0.50 effects were first noted as a toxicity, resulting in infertility and DES 100 nM 2.4 13.9 developmental disorders in grazing livestock (reviewed in Ref. 48), ␮ Coumestrol 1 M 2.2 10.1 and thus, the potential for adverse effects to mammals exposed to Naringenin 10 ␮M 1.7 4.6 Genistein 10 ␮M 3.5 18.9 these compounds is well established. However, phytoestrogen-rich Endosulfan-␣ 10 ␮M 1.9 1.5 b foods, such as soy, are statistically associated with a reduced risk of Endosulfan-␤ 20 ␮M 1.8 1.6 cancer (12). Although a mechanism for this protective effect is un- a Maximum fold induction of luminescence was calculated for each compound using the C3-tk-LUC reporter in a representative experiment. Compounds are listed according known, recent studies suggest that a suppressive role for phytoestro- to potency as determined by the concentration at which this maximum induction was gens may rely on the developmental window of exposure (49, 50), an observed. Induction of C3-tk-LUC is compared with increase or decrease in vitERE-tk- observation with vital implications for the role of endogenous hor- LUC activity over control at the same concentration of compound. b Concentration, 10 ␮M. mones in tumor development.

binding. The magnitude of difference between effective concentrations was also preserved between the two functional assays, with the exception of endosulfan-␤, which was transactivating but did not induce proliferation. Endosulfan ␤ may have failed to elicit a proliferative response, due to a partial toxicity at the concentrations required to activate the ER. In most cases, proliferation was more sensitive to toxicity than transactivation in terms of dose, perhaps due to the longer incubation period with the test compounds in the proliferative assay. For compounds that could activate transcription via ER, the relative binding affinities exhibited the same rank order, with one notable exception. DES, while having 5.5 times greater affinity

for the receptor, was slightly less potent than E2 in both stimulation of proliferation and transactivation. Additionally, for coumestrol, genistein, naringenin, and the endosulfan isomers, although the rank order of affinity predicts potency, the magnitude of difference between binding affinities is inconsistent with that observed in the functional Fig. 6. Northern analysis of PR message in ELT3 cells. Northern analysis was performed on 5 ␮g of polyadenylated RNA isolated from ELT3 cells treated with assays. Especially interesting are the endosulfans, each of which were concentrations of test compounds that stimulated maximum proliferation in vitro. A, ␮ positive in one or both of the functional assays, despite having a very treatments using ethanol as a solvent: 0.1% ethanol, 10 nM E2,10nM DES, 10 M naringenin, 100 nM endosulfan-␣, and 100 nM endosulfan-␤. B, treatments using DMSO limited affinity for the receptor. Other studies have showed similar as a solvent: 10 ␮M genistein, 1 ␮M coumestrol, and 0.1% DMSO. Densitometric analysis results (43), and some investigators have proposed that these effects was performed after normalizing to signal for GAPDH. 3096

Fig. 7. Myometrial proliferation in rats treated in vivo with vehicle, tamoxifen, or DES. A, example of a BrdUrd-stained uterine horn of vehicle control-treated rat during proestrus. (l) and (c), longitudinal and circular layers of the myometrium, respectively; endometrial stroma and luminal epithelium lie to the left of the myometrial layers. Arrows, BrdUrd-positive cells in the myometrium, some of which are associated with glandular structures or vasculature. Magnification, ϫ100. B, BrdUrd-stained uterine horn of vehicle control-treated rat during estrus. Myometrial tissue (at the periphery) is nearly devoid of BrdUrd-stained cells, although a few stained cells are visible in the stroma and immediately beneath the luminal epithelium. Magnification, ϫ40. C, BrdUrd-stained uterine horn of tamoxifen-treated rat. Note the small size of the horn in cross-section as compared to uterus from vehicle-treated rat and the absence of BrdUrd-stained cells. Magnification, ϫ40. D, H&E-stained uterine horn of DES-treated rat. (l) and (c), longitudinal and circular layers of the myometrium, respectively, with condensed stromal tissue and thickened luminal epithelium with numerous invaginations visible above the myometrial layers. Magnification, ϫ100. E, BrdUrd-stained uterine horn of DES-treated rat. Note BrdUrd-positive cells in the longitudinal layer and dramatic enlargement of the tissue as compared to untreated uteri at the same magnification. Arrows, BrdUrd-positive cells in the longitudinal myometrium. Magnification, ϫ100. F, BrdUrd-stained uterine horn of DES-treated rat. The two myometrial layers are greatly enlarged, with an abundance of BrdUrd-stained cells. The proximal boundary of the myometrium is adjoined by a very thin stromal layer and an effaced luminal surface. Magnification, ϫ100.

In terms of agonist activity, the profile of ER ligands in the uterine observed using the C3 promoter, an observation that is consistent with myometrium appears to more closely resemble that of mammary the ability of tamoxifen and raloxifene to act as agonists in some cell tissue than uterine endometrium. For example, the therapeutic com- types (51–53). Additionally, the proliferative potential of test compounds OH-tam and the two raloxifene analogues (LY117018 and pounds for estrogen-sensitive ELT3 cells was most closely predicted LY317783) did not exhibit estrogen agonism in myometrial cells and, by the ability of the compounds to activate transcription via the furthermore, were able to inhibit the baseline activity of the ER on the AF2-mediated vitellogenin ERE rather than by the AF1-mediated C3 vitellogenin ERE in estrogen-free conditions. This inhibition was not promoter. The C3 promoter was only weakly active in ELT3 cells, 3097

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1999 American Association for Cancer Research. ER AGONISM OF XENOESTROGENS IN UTERINE MYOMETRIUM suggesting that AF1-mediated transactivation may be less important and normal myometrial tissues. In addition, the results of this study in the repertoire of the uterine myometrium than in the endometrium. are consistent with the necessity for activation of the AF2 function of Because the agonistic activity of tamoxifen is mediated via AF1 (35), the ER for agonist activity in this tissue. These results suggest that the this observation predicts that tamoxifen would not mimic estrogen response of the myometrium to a mixed function ligand differs from function in this tissue, consistent with in vitro and in vivo evidence that of other uterine compartments; this distinction is important and from this study. Thus, the myometrium responds to tamoxifen in a implies that each tissue must be independently evaluated for the similar manner to mammary tissue, where tamoxifen acts as an effects of a potential estrogen or antiestrogen. antagonist, in contrast with the response of endometrial cells, in which both E2 and tamoxifen can stimulate the transcription of C3 and promote proliferation. ACKNOWLEDGMENTS Controlled clinical studies have not clearly defined the effect of We thank Kalyn Sowell and Dr. Lezlee Coghlan for their contributions to pharmacological ER modulators on the uterine myometrium. Here, the in vivo studies and Michelle Gardiner for secretarial assistance. We are also OH-tam and the raloxifene analogues showed no evidence of estro- grateful to Dr. Kevin Burroughs for helpful insight and to Dr. Tim Zacharew- genicity in myometrial cells, a quality that could be clinically relevant, ski for critical review of the manuscript. given increasing numbers of perimenopausal and postmenopausal women on treatment regimens for breast cancer chemotherapy or prophylaxis. In contrast, DES and each of the environmental com- REFERENCES pounds exhibited estrogenicity in multiple assays, with their order of 1. Stancel, G. M., Boettger-Tong, H. L., Chiappetta, C., Hyder, S. M., Kirkland, J. L., potency predicted by receptor binding affinity relative to E .Itis Murthy, L., and Loose-Mitchell, D. S. Toxicity of endogenous and environmental 2 estrogens: what is the role of elemental interactions? Environ. Health Perspect., 103 notable that the estrogenic effects of compounds used here (excluding (Suppl. 7): 29–33, 1995. DES) were observed at concentrations 100 to 100,000 times greater 2. Crisp, T. M., Clegg, E. D., Cooper, R. L., Wood, W. P., Anderson, D. G., Baetcke, K. P., Hoffmann, J. L., Morrow, M. S., Rodier, D. J., Schaeffer, J. E., Touart, L. W., than the equipotent concentration of E2. The probability of achieving Zeeman, M. G., and Patel, Y. M. Environmental endocrine disruption: an effects concentrations of this magnitude from exposures other than pharma- assessment and analysis. Environ. Health Perspect., 106 (Suppl. 1): 11-56, 1998. cological ones is an important factor for assessing biological effects, 3. Gambrell, R. D., Jr. The menopause: benefits and risks of hormone replacement therapy. Comprehensive Ther., 20: 580–585, 1994. but this probability is not easily determined. Efforts to measure or 4. Tonetti, D. A., and Jordan, V. C. Targeted anti-estrogens to treat and prevent diseases estimate the circulating levels or tissue levels of exogenous ER in women. Mol. Med. Today, 2: 218–223, 1996. ligands (54) have not satisfactorily resolved this issue. It could be 5. McDonnell, D. P., Clemm, D. L., Hermann, T., Goldman, M. E., and Pike, J. W. Analysis of estrogen receptor function in vitro reveals three distinct classes of argued that the impact of an exogenous estrogens with a receptor antiestrogens. Mol. Endocrinol., 9: 659–669, 1995. affinity in the micromolar range on the myometrium of a mature 6. Friedl, A., and Jordan, V. C. What do we know and what we don’t know about tamoxifen in the human uterus. Breast Cancer Res. Treat., 31: 27–39, 1994. female with high levels of endogenous estrogens may be inconse- 7. Quarmby, V. E., and Korach, K. S. The influence of 17␤-estradiol on patterns of cell quential due to competition for receptor binding. However, exposure division in the uterus. Endocrinology, 114: 694–702, 1984. to exogenous ER ligands may have a much more significant impact in 8. Rein, M. S., and Nowak, R. A. Biology of uterine myomas and myometrium in vitro. Semin. Reprod. Endocrinol., 10: 310–319, 1992. individuals in which endogenous estrogen levels are very low; for 9. Cramer, S. F., and Patel, A. The frequency of uterine leiomyomas. Am. J. Clin. example, children and postmenopausal women. Pathol., 94: 435–438, 1990. Our ability to predict whether the effects of ER ligand exposure are 10. Buttram, V. C., Jr., and Reiter, R. C. Uterine leiomyomata: etiology, symptomatol- ogy, and management. Fertil. Steril., 36: 433–445, 1981. beneficial or hazardous, in the absence of empirical data, is still very 11. Johansen, J. S., Riis, B. J., Hassager, C., Moen, M., Jacobson, J., and Christiansen, C. poor but is improving as we discover the molecular mechanisms of The effect of a gonadotropin-releasing hormone agonist analog (nafarelin) on bone metabolism. J. Clin. Endocrinol. Metab., 67: 701–706, 1988. tissue specificity. Some tissue-specific effects may reflect the partic- 12. Rao, G. N. Influence of diet on tumors of hormonal tissues. In: J. Huff, J. Boyd, and ipation of the recently discovered ER␤ (55), although a role for this J. C. Barrett (eds.), Cellular and Molecular Mechanisms of Hormonal Carcinogenesis, receptor is not expected in the myometrium, where ER␤ expression is pp. 41–56. New York, NY: Wiley-Liss, Inc., 1996. 13. Everitt, J. I., Wolf, D. C., Howe, S. R., Goldsworthy, T. L., and Walker, C. Rodent low to undetectable (56, 57). Additionally, whereas most mechanistic model of reproductive tract leiomyomata. Clinical and pathological features. Am. J. studies of ligand effects have been conducted with EREs, other studies Pathol., 146: 1556–1567, 1995. have noted ER-mediated transactivation via other genetic elements, 14. Herbst, A. L., Ulfelder, H., and Poskanzer, D. C. Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young women. such as Sp-1 (58) and AP-1 (59), each of which carries the potential N. Engl. J. Med., 284: 878–881, 1971. for its own pattern of ligand-specific effects. It has recently become 15. Newbold, R. R., and McLachlan, J. A. Diethylstilbestrol-associated defects in murine genital tract development. In: J. A. McLachlan (ed.), Estrogens in the Environment II: apparent that the tissue-specific effects of ER ligands on receptor- Influences on Development, pp. 288–318. New York: Elsevier, 1985. mediated responses are governed by the participation of steroid co- 16. Furr, B. J., and Jordan, V. C. The pharmacology and clinical uses of tamoxifen. activator and corepressor proteins (60–64). The interaction and ex- Pharmacol. Ther., 25: 127–205, 1984. 17. Black, L. J., Sato, M., Rowley, E. R., Magee, D. E., Bekele, A., Williams, D. C., pression pattern of these comodulators is not fully understood but is Cullinan, G. J., Bendele, R., Kauffman, R. F., Bensch, W. R., Frolik, C. A., Termine, likely to explain the ability of the ER to effect a wide variety of J. D., and Bryant, H. U. Raloxifene (LY139481 HCI) prevents bone loss and reduces cellular responses and vary those responses to given ligands in a serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J. Clin. Invest., 93: 63–69, 1994. tissue-specific manner. These data and those of others suggest that the 18. Wang, T. T., Sathyamoorthy, N., and Phang, J. M. Molecular effects of genistein on ability of a ligand to activate the ER in a particular cellular context can estrogen receptor mediated pathways. Carcinogenesis (Lond.). 17: 271–275, 1996. 19. So, F. V., Guthrie, N., Chambers, A. F., Moussa, M., and Carroll, K. K. Inhibition of predict agonist activity in that tissue. Knowing more about the mo- human breast cancer cell proliferation and delay of mammary tumorigenesis by lecular determinants of this agonism will help improve our ability to flavonoids and citrus juices. Nutr. Cancer, 26: 167–181, 1996. modulate ER-mediated responses without unwanted side effects. 20. Martin, P. M., Horwitz, K. B., Ryan, D. S., and McGuire, W. L. Phytoestrogen interaction with estrogen receptors in human breast cancer cells, Endocrinology, 103: In summary, these experiments describe an in vitro/in vivo model 1860–1867, 1978. system for determining the effects of ER ligands in uterine myome- 21. Miksicek, R. J. Interaction of naturally occurring nonsteroidal estrogens with ex- trium and identifying their mechanism of action. Six of the nine pressed recombinant human estrogen receptor. J. Steroid Biochem. Mol. Biol., 49: 153–160, 1994. xenoestrogens in this panel (DES, genistein, coumestrol, naringenin, 22. Gupta, P. K., Chandra, S. V., and Saxena, D. K. Teratogenic and embryotoxic effects endosulfan-␣, and endosulfan-␤) exhibited evidence of agonist activ- of endosulfan in rats. Acta Pharmacol. Toxicol. (Copenh.), 42: 150–152, 1978. 23. Soto, A. M., Chung, K. L., and Sonnenschein, C. The pesticides endosulfan, toxa- ity in myometrial cells of the Eker rat in multiple assays, demonstrat- phene, and dieldrin have estrogenic effects on human estrogen-sensitive cells. Envi- ing the potential for an impact of xenoestrogens on both transformed ron. Health Perspect., 102: 380–383, 1994. 3098

24. Howe, S. R., Gottardis, M. M., Everitt, J. I., Goldsworthy, T. L., Wolf, D. C., and 44. Das, S. K., Tan, J., Johnson, D. C., and Dey, S. K. Differential spatiotemporal Walker, C. Rodent model of reproductive tract leiomyomata. Establishment and regulation of lactoferrin and progesterone receptor genes in the mouse uterus by characterization of tumor-derived cell lines. Am. J. Pathol., 146: 1568–1579, 1995. primary estrogen, catechol estrogen, and xenoestrogen. Endocrinology, 139: 2905– 25. Obourn, J. D., Koszewski, N. J., and Notides, A. C. Hormone- and DNA-binding 2915, 1998. mechanisms of the recombinant human estrogen receptor. Biochemistry, 32: 6229– 45. Daston, G. P., Gooch, J. W., Breslin, W. J., Shuey, D. L., Nikiforov, A. I., Fico, T. A., 6236, 1993. and Gorsuch, J. W. Environmental estrogens and reproductive health: a discussion of 26. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J., and Rutter, W. J. Isolation of the human and environmental data. Reprod. Toxicol., 11: 465–481, 1997. biologically active ribonucleic acid from sources enriched in ribonuclease, Biochem- 46. Saxena, S. P., Khare, C., Farooq, A., Murugesan, K., Buckshee, K., and Chandra, J. istry, 18: 5294–5299, 1979. DDT and its metabolites in leiomyomatous and normal human uterine tissue. Arch. 27. Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., Toxicol., 59: 453–455, 1987. and Struhl, K. Current Protocols in Molecular Biology, Vol. 1. New York, NY: John 47. Whitten, P. L., Lewis, C., Russell, E., and Naftolin, F. Potential adverse effects of Wiley & Sons, Inc., 1987. phytoestrogens. J. Nutr., 125 (Suppl.): 771S–776S, 1995. 28. Park, O. K., and Mayo, K. E. Transient expression of progesterone receptor messen- 48. Adams, N. R. Detection of the effects of phytoestrogens on sheep and cattle. J. Anim. ger RNA in ovarian granulosa cells after the preovulatory luteinizing hormone surge. Sci., 73: 1509–1515, 1995. Mol. Endocrinol., 5: 967–978, 1991. 49. Murrill, W. B., Brown, N. M., Zhang, J. X., Manzolillo, P. A., Barnes, S., and 29. Tzukerman, M. T., Esty, A., Santiso-Mere, D., Danielian, P., Parker, M. G., Stein, Lamartiniere, C. A. Prepubertal genistein exposure suppresses mammary cancer and R. B., Pike, J. W., and McDonnell, D. P. Human estrogen receptor transactivational enhances gland differentiation in rats. Carcinogenesis (Lond.), 17: 1451–1457, 1996. capacity is determined by both cellular and promoter context and mediated by two 50. Hsieh, C-Y., Santell, R. C., Haslam, S. Z., and Helferich, W. G. Estrogenic effects of functionally distinct intramolecular regions. Mol. Endocrinol., 8: 21–30, 1994. genistein on the growth of estrogen receptor-positive human breast cancer (MCF-7) 30. Yuan, Y. D., and Carlson, R. G. Structure, cyclic change, and function, vagina, and cells in vitro and in vivo. Cancer Res., 58: 3833–3838, 1998. vulva, rat. In: T. Jones, U. Mohr, and R. Hunt (eds.), Genital System, pp. 161–168. 51. Gallo, M. A., and Kaufman, D. Antagonistic and agonistic effects of tamoxifen: New York, NY: Springer-Verlag, 1987. significance in human cancer. Semin. Oncol., 24 (Suppl.): S1-71–S1-80, 1997. 31. Howe, S. R., Gottardis, M. M., Everitt, J. I., and Walker, C. Estrogen stimulation and 52. Dere, W. H. Clinical efficacy of SERMs. Aging (Milan), 10: 160–161, 1998. tamoxifen inhibition of leiomyoma cell growth in vitro and in vivo. Endocrinology, 53. Wakeling, A. E., O’Connor, K. M., and Newboult, E. Comparison of the biological 136: 4996–5003, 1995. effects of tamoxifen and a new antioestrogen (LY 117018) on the immature rat uterus. 32. Peterson, G., and Barnes, S. Genistein inhibits both estrogen and growth factor- J. Endocrinol., 99: 447–453, 1983. stimulated proliferation of human breast cancer cells. Cell Growth Differ., 7: 1345– 54. Adlercreutz, C. H., Goldin, B. R., Gorbach, S. L., Hockerstedt, K. A., Watanabe, S., 1351, 1996. Hamalainen, E. K., Markkanen, M. H., Makela, T. H., Wahala, K. T., and Adler- 33. Elsea, S., Westergaard, M., Burden, D., Lomenick, J., and Osheroff, N. Quinolones creutz, T. Soybean phytoestrogen intake and cancer risk [published erratum appears share a common interaction domain on topoisomerase II with other DNA cleavage- in J. Nutr., 125: 1960, 1995]. J. Nutr., 125 (Suppl.): 757S–770S, 1995. enhancing antineoplastic drugs. Biochemistry, 36: 2919–2924, 1997. 55. Pettersson, K., Grandien, K., Kuiper, G. G., and Gustafsson, J. A. Mouse estrogen 34. Pagliacci, M. C., Smacchia, M., Migliorati, G., Grignani, F., Riccardi, C., and receptor ␤ forms estrogen response element-binding heterodimers with estrogen Nicoletti, I. Growth-inhibitory effects of the natural phyto-oestrogen genistein in receptor ␣. Mol. Endocrinol., 11: 1486–1496, 1997. MCF-7 human breast cancer cells. Eur. J. Cancer, 11: 1675–1682, 1994. 56. Kuiper, G. G., Carlsson, B., Grandien, K., Enmark, E., Haggblad, J., Nilsson, S., and 35. Metzger, D., Losson, R., Bornert, J. M., Lemoine, Y., and Chambon, P. Promoter Gustafsson, J. A. Comparison of the ligand binding specificity and transcript tissue specificity of the two transcriptional activation functions of the human oestrogen ␣ ␤ receptor in yeast. Nucleic Acids Res., 20: 2813–2817, 1992. distribution of estrogen receptors and . Endocrinology, 138: 863–870, 1997. 36. Berry, M., Metzger, D., and Chambon, P. Role of the two activating domains of the 57. Petersen, D. N., Tkalcevic, G. T., Koza-Taylor, P. H., Turi, T. G., and Brown, T. A. ␤ oestrogen receptor in the cell-type and promoter-context dependent agonistic activity Identification of estrogen receptor 2, a functional variant of estrogen receptor beta of the anti-oestrogen 4-hydroxytamoxifen. EMBO J., 9: 2811–2818, 1990. expressed in normal rat tissues. Endocrinology, 139: 1082–1092, 1998. 37. Sundstrom, S. A., Komm, B. S., Xu, Q., Boundy, V., and Lyttle, C. R. The stimulation 58. Porter, W., Saville, B., Hoivik, D., and Safe, S. Functional synergy between the of uterine complement component C3 gene expression by antiestrogens, Endocrinol- transcription factor Sp1 and the estrogen receptor. Mol. Endocrinol., 11: 1569–1580, ogy, 126: 1449–1456, 1990. 1997. 38. Kraus, W. L., and Katzenellenbogen, B. S. Regulation of progesterone receptor gene 59. Webb, P., Lopez, G. N., Uht, R. M., and Kushner, P. J. Tamoxifen activation of the expression and growth in the rat uterus: modulation of estrogen actions by proges- estrogen receptor/AP-1 pathway: potential origin for the cell-specific estrogen-like terone and sex steroid hormone antagonists. Endocrinology, 132: 2371–2379, 1993. effects of antiestrogens. Mol. Endocrinol., 9: 443–456, 1995. 39. Horwitz, K. B., and McGuire, W. L. Estrogen control of progesterone receptor in 60. Horwitz, K. B., Jackson, T. A., Bain, D. L., Richer, J. K., Takimoto, G. S., and Tung, human breast cancer. Correlation with nuclear processing of estrogen receptor. L. Nuclear receptor coactivators and corepressors. Mol. Endocrinol., 10: 1167–1177, J. Biol. Chem., 253: 2223–2228, 1978. 1996. 40. Robinson, S. P., Langan-Fahey, S. M., Johnson, D. A., and Jordan, V. C. Metabolites, 61. Zhang, X., Jeyakumar, M., Petukhov, S., and Bagchi, M. K. A nuclear receptor pharmacodynamics, and pharmacokinetics of tamoxifen in rats and mice compared to corepressor modulates transcriptional activity of antagonist-occupied steroid hormone the breast cancer patient. Drug Metab. Dispos., 19: 36–43, 1990. receptor. Mol. Endocrinol., 12: 513–524, 1998. 41. Marcus, G. J. Mitosis in the rat uterus during the estrous cycle, early pregnancy, and 62. McDonnell, D. P., Dana, S. L., Hoener, P. A., Lieberman, B. A., Imhof, M. O., and early pseudopregnancy. Biol. Reprod. 10: 447–452, 1974. Stein, R. B. Cellular mechanisms which distinguish between hormone- and antihor- 42. Smith, M. S., Freeman, M. E., and Neill, J. D. The control of progesterone secretion mone-activated estrogen receptor. Ann. N. Y. Acad. Sci., 761: 121–137, 1995. during the estrous cycle and early pseudopregnancy in the rat: prolactin, gonadotropin 63. Smith, C. L., Nawaz, Z., and O’Malley, B. W. Coactivator and corepressor regulation and steroid levels associated with rescue of the corpus luteum of pseudopregnancy. of the agonist/antagonist activity of the mixed antiestrogen, 4-hydroxytamoxifen. Endocrinology, 96: 219–226, 1975. Mol. Endocrinol., 11: 657–666, 1997. 43. Vonier, P. M., Crain, D. A., McLachlan, J. A., Guillette, L. J., Jr., and Arnold, S. F. 64. Katzenellenbogen, J. A., O’Malley, B. W., and Katzenellenbogen, B. S. Tripartite Interaction of environmental chemicals with the estrogen and progesterone receptors steroid hormone receptor pharmacology: interaction with multiple effector sites as a from the oviduct of the American alligator. Environ. Health Perspect., 104: 1318– basis for the cell- and promoter-specific action of these hormones. Mol. Endocrinol., 1322, 1996. 10: 119–131, 1996.

3099

Downloaded from cancerres.aacrjournals.org on September 29, 2021. © 1999 American Association for Cancer Research. Estrogen Receptor Activation via Activation Function 2 Predicts Agonism of Xenoestrogens in Normal and Neoplastic Cells of the Uterine Myometrium

Deborah S. Hunter, Leslie C. Hodges, Peter M. Vonier, et al.

Cancer Res 1999;59:3090-3099.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/59/13/3090

Cited articles This article cites 53 articles, 5 of which you can access for free at: http://cancerres.aacrjournals.org/content/59/13/3090.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/59/13/3090.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/59/13/3090. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.