Steroid hormones have been implicated in the etiology and/or progression of epithelial ovarian cancer. As ovarian surface epithelial cells are growth inhibited by transforming growth factor β (TGF-β), we tested whether steroid hormones could regulate the expression of TGF-β1 or its receptors in ovarian cancer cells, as assessed by quantitative reverse transcription-PCR. Treatment of ovarian cancer HEY cells with 500 nm 5α-dihydrotestosterone (DHT), but not estradiol-17β or progesterone, for 60 h down-regulated the expression of mRNA for TGF-β receptors I and II (TβR-I and TβR-II), betaglycan, and endoglin but had no effect on TGF-β1 mRNA levels. Androgen receptor (AR) mRNA expression in HEY cells was compared to other ovarian cancer cell lines. OVCAR-3 cells expressed AR mRNA levels similar to that of androgen-responsive LNCaP prostate cancer cells, whereas SKOV-3 and HEY cells expressed only 3 and 0.01%, respectively. Western blot analysis and saturation binding assays confirmed the expression of AR protein in these three cell lines, but at the limit of detection in SKOV-3 and HEY cells. Treatment of SKOV-3 and HEY cells for 24 h with 1–50 nm DHT resulted in a dose-dependent down-regulation of TβR-II mRNA. The AR antagonist hydroxyflutamide did not reverse the effect of DHT on SKOV-3 cells but by itself down-regulated TβR-II mRNA. This apparent androgen-mimetic action of hydroxyflutamide and the ability of SKOV-3 and HEY cells to respond to DHT may be due to their expression of AR-associating protein 70, an AR co-activator reported to amplify AR transactivation and to result in agonist activity of AR antagonists. DHT was able to reverse TGF-β1 growth-inhibitory action in SKOV-3 cells and in a primary culture of ovarian cancer cells derived from ascites. Thus, androgens may promote ovarian cancer progression in part by decreasing TGF-β receptor levels, thereby allowing ovarian cancer cells to escape TGF-β1 growth inhibition.

The most common and lethal form of ovarian cancer is that derived from the ovarian surface epithelium. The etiology of epithelial ovarian cancer is poorly understood; however, steroid hormones have been implicated in the progression of this disease (1). Although 50–60% of ovarian cancer tumors express estrogen and/or progestin receptors, up to 95% of ovarian cancer tumors express ARs (2, 3, 4, 5, 6). In a nested case-control study, an increased risk of ovarian cancer was found in women with elevated circulating levels of androgens (7). Polycystic ovarian syndrome, a condition characterized by thecal cell hypertrophy and elevated circulating levels of androgens, has also been associated with an increased risk of ovarian cancer (8). Furthermore, in guinea pigs, long-term administration of testosterone, but not estrone, induced the formation of benign ovarian epithelial neoplasms (9).

A mechanism by which steroid hormones might stimulate the growth and/or progression of ovarian cancer is by blocking the action of TGF-β,4a potent inhibitor of ovarian epithelial cell growth in culture (10, 11). Development of resistance to TGF-β has been observed in many cancers (12) and can occur gradually during the multistep process of tumor progression (12, 13).

TGF-βs (TGF-β1, -β2, and -β3) mediate their effects on target cells by forming a heteromeric complex with transmembrane serine/threonine kinase receptors, TβR-II and TβR-I (14). Binding of TGF-β to constitutively active TβR-II leads to recruitment, interaction, and phosphorylation of TβR-I,which then propagates the signal to downstream Smad proteins (14). In addition, two other proteins can be found in TGF-β receptor complexes, betaglycan, and endoglin. Betaglycan binds TGF-β1, -β2, and -β3 directly and is referred to as type III receptor; it presents ligand to TβR-II but does not have a recognizable signaling domain (15). Endoglin is a dimeric transmembrane glycoprotein that binds TGF-β1 and -β3 via its association with TβR-II (16, 17). Overexpression of endoglin has been shown to modulate cellular responses to TGF-β1 (18). Thus, the down-regulation or loss of functional expression of receptors and/or signaling mediators for TGF-β would allow cancer cells to become resistant to this potent growth inhibitor.

In the present study, we tested the effects of steroid hormones on the expression of TGF-β and its receptors in ovarian cancer cells. As steroid hormones regulate ovarian function, we hypothesize that an imbalance in their action could promote ovarian cancer progression by decreasing the action of TGF-β.

Cell Cultures

HEY, SKOV-3, and HOC-7 cells were obtained from Dr. Alexander Marks (Banting and Best Institute, University of Toronto, Toronto,Ontario, Canada). OVCA-429 cells were obtained from Dr. Robert Kerbel(Sunnybrook and Women’s College Health Science Center, Toronto,Ontario, Canada). OVCAR-3, LNCaP, and DU-145 cells were obtained from American Type Culture Collection (Manassas, VA). These cells were maintained in RPMI 1640 (without phenol red) supplemented with 5%(v/v) FBS, 50 units/ml penicillin, 50 μg/ml streptomycin, and 0.625μg/ml amphotericin B (Fungizone).

All cell cultures were maintained at 37°C in a humidified 5%CO2 atmosphere. sFBS was substituted for normal FBS 48 h before beginning experiments. DHT, estradiol-17β, and progesterone (Sigma Chemical Co., St. Louis, MO) and OH-FLUT(Schering-Plow, Madison, NJ) were dissolved in ethanol before addition to culture medium. The final ethanol concentration was 0.001%. TGF-β1 (R&D Systems, Minneapolis, MN) was diluted in culture media and stored at 4°C before addition to DHT-treated cultures.

Primary ovarian cancer cells (OVCAS-16) were isolated from ascites of a patient diagnosed with papillary adenocarcinoma of the ovary, as described by Hirte et al.(19). Cells were propagated in DMEM/F12 media (without phenol red) supplemented with 3%sFBS, 5 μg/ml insulin, 5 × 10−5m ethanolamine, 5 ng/ml EGF, 5 × 10−5m phosphoethanolamine, 10 μg/ml transferrin,penicillin-streptomycin, and fungizone. The cells were shown by flow cytometry to express the epithelial ovarian cancer antigens 2G3 and M2A (19, 20). These monoclonal antibodies were provided by Dr. A. Marks. OVCAS-16 cells were maintained in the presence or absence of 10 nm DHT added daily from the time of isolation. Experiments were performed within 2 weeks and after one passage.

RNA Preparation, cDNA Synthesis, and Quantitative Analysis of mRNA by PCR

Cell cultures were washed in ice-cold PBS or serum-free medium, and total cellular RNA was isolated by the guanidinium thiocyanate extraction method (21). RNA concentration in all samples was measured in two independent determinations and at two dilutions, as its accuracy is essential for quantitative RT-PCR. cDNAs for TGF-βreceptors, AR, ARA70, and β-actin were generated using oligo(dT)primer, whereas an internal primer was used to generate TGF-β1 cDNA to increase the specificity of the reaction (22).

Quantitative analysis of PCR products was performed according to Murphy et al.(23). Selected primers for amplification of TGF-β1, TGF-β receptors, and β-actin have been described previously (22). Three sets of AR-specific primers spanning exons 1–4, 4–7, and 1–7 selected from the cDNA sequence (24) were originally tested in the AR-positive prostate cancer cell line LNCaP. As identical results were obtained with all three primer sets, forward primer 5′-GTCAAAAGCGAAATGGGCCCC-3′ and reverse primer 5′-CTTGGTGAGCTGGTAGAAGCG-3′, which yielded a 1031-bp PCR product, were chosen for the current study. Primers for ARA70, selected from the cDNA sequence (25), were forward primer 5′-GCTTTGCAGAGTGTGTGTGTG-3′ and reverse primer 5′-GGGGAAGTTATGTTCCTCCTG-3′, which yielded a product of 509 bp. PCR amplification conditions were as described previously (22)with an annealing temperature of 58°C for all except AR (64°C), and 28 cycles for all except betaglycan and AR (30 cycles). Aliquots (10 or 20 μl) of PCR products were fractionated on 4–20% gradient polyacrylamide gels in 0.5% Tris-borate/EDTA buffer (Helixx Technologies, Scarborough, Ontario, Canada) or on 1% agarose gels. Gels were transferred to nylon membranes and hybridized with specific 32P-dATP-labeled or 3′ end-labeled digoxigenin-11-ddUTP internal probes. The probes were as described previously (18, 22) with the addition of the internal probe, 5′-GCATCCTGGAGTTGACATTGG-3′, for AR and 5′-TGATAAGCCACTCCGACAAGG-3′ for ARA70. The blots were analyzed by computer-assisted densitometry as reported previously (18, 22). Data were plotted as log of densitometric units versus log of equivalent RNA concentration to identify a linear range of exponential amplification. Linear regressions were obtained using SigmaPlot scientific graphing software (version 2.00,Jandel Corp., San Rafael, CA). For TGF-β receptor mRNA expression,the amount of total RNA estimated from the linear range of amplification after steroid treatment was expressed as a percentage of that measured in cells treated with ethanol vehicle (control). All measurements were corrected for β-actin expression.

Cytosol Preparation and Analysis of AR Binding and Protein Expression

Cytosol extracts were prepared as described previously (26). For Western blot analysis, aliquots of cytosol protein (1–50 μg) were fractionated by 8% SDS-PAGE and transferred to Immobilon polyvinylidene difluoride membrane (Millipore Corp.,Bedford, MA). The membranes were preincubated for 1 h in blocking solution (50 mm Tris-HCl, 0.15 m NaCl, pH 7.4,0.2% (v/v) Tween 20, and 5% powdered skim milk) followed by 1 h with 1 μg/ml affinity-purified polyclonal antibody to AR, PAR-1 (27). Immunoreactive proteins were visualized with horseradish peroxidase-labeled goat anti-rabbit antisera (Amersham Pharmacia Biotech, Oakville, Ontario, Canada) at 1:10,000 dilution in blocking solution for 30 min at room temperature and detected by enhanced chemiluminescence (Amersham Pharmacia Biotech) on Kodak X-OMAT X-ray film.

Androgen saturation binding assays were performed as reported previously (26). All samples contained 1 μmtriamcinolone acetonide to block possible binding of[3H]R1881 to progesterone or glucocorticoid receptor (28).

Cell Proliferation Assays

SKOV-3 proliferation was determined by the MTT (Sigma) reduction method as described previously (26). Cells were treated with 10 nm DHT or vehicle daily beginning 24 h prior to seeding into 96-well microtiter plates at 2500 cells/well. At time 0(24 h after seeding) and at 48 h, 5 pm TGF-β1(optimal dose for SKOV-3 inhibition) or vehicle was added to each well. MTT reduction was quantified using 570 nm absorbance with an ELISA microplate reader (μQuant Biomolecular Spectrophotometer, BIO-TEK Instruments, Winooski, VT). Results were expressed as the mean ± SE of 6–8 replicates.

Proliferation of the primary ovarian cancer cells (OVCAS-16) was assessed by counting the number of viable cells, as these cells did not reduce the MTT reagent. Cells maintained in the presence or absence of 10 nm DHT or vehicle were seeded at 15,000 cells/well in 24-well plates. TGF-β1 (25 pm, optimal dose for OVCAS-16 inhibition) and/or 10 nm DHT was added every 24 h during proliferation assays. Cell counts were obtained using a hemacytometer, with four replicates per treatment group.

DHT Down-Regulates mRNA for TGF-β Receptors, but not TGF-β1, in the HEY Cell Line

To determine whether steroid hormones might regulate the expression of TGF-β1 and its receptors, we examined the effect of estradiol-17β,progesterone, and DHT treatment in HEY cells. Cells were grown in serum-free DMEM for 2 days, then treated with 500 nmestradiol-17β, progesterone, DHT, or ethanol vehicle (control), which was replaced every 12 h. Cells were harvested at 60 h, and total RNA was extracted and subjected to quantitative RT-PCR. TGF-β1 mRNA levels were unaffected by treatment with 500 nm of estradiol-17β, progesterone or DHT (Fig. 1,A). However, the autoradiograms from a representative experiment revealed that DHT down-regulated mRNA for TβR-I, -RII,betaglycan, and endoglin, whereas estradiol-17β and progesterone did not (Fig. 1 B). The expression of β-actin mRNA was unaffected by treatment with steroids, suggesting that a concentration of 500 nm had no nonspecific inhibitory effects on mRNA synthesis.

A summary of the results from 3–5 independent experiments is shown in Fig. 1,C. TGF-β receptor mRNA levels within the exponential range were quantified for each treatment and expressed as percentage of that measured in control cells. Expression of steady-state mRNA levels of TβR-I, TβR-II, betaglycan, and endoglin were markedly reduced to 9–18% that of control cells. Treatment of HEY cells with estradiol-17β or progesterone did not affect mRNA levels for any of the TGF-β receptors (Fig. 1 C).

Variable Expression of AR in Ovarian Cancer Cell Lines

Because androgenic responses should be mediated by AR, we examined its expression in HEY cells, as well as in several other ovarian cancer cell lines. RNA was extracted from HEY, SKOV-3, OVCAR-3, HOC-7, and OVCA-429 ovarian cancer cell lines, and AR mRNA was detected by RT-PCR. The AR-positive LNCaP and AR-negative DU-145 prostate cancer cell lines were included as positive and negative controls, respectively. A 1031-bp PCR product, corresponding to exons 1–7 of AR, was amplified in all of the ovarian cancer cell lines tested (Fig. 2,A); however, the apparent levels of expression were variable. High levels of AR mRNA were observed in LNCaP and OVCAR-3 cells, and intermediate levels were found in SKOV-3 cells. HEY, HOC-7, and OVCA-429 cells gave markedly less intense signals and required 100 ng instead of 20 ng per lane for detection (Fig. 2,A). As expected, no AR mRNA expression was detected in DU-145 cells (Fig. 2 A).

OVCAR-3, SKOV-3, and HEY cells expressing high, moderate, and low AR mRNA, respectively, were further analyzed by quantitative RT-PCR(Fig. 2, B and C). A linear range of exponential amplification was obtained with less than 0.4 ng of equivalent RNA for both LNCaP and OVCAR-3 cells, whereas the linear range of amplification for SKOV-3 cells was 0.8–20 ng of equivalent RNA. In comparison, HEY cells had barely detectable AR expression (Fig. 2,B). By determining the relative amount of RNA needed to obtain similar densitometric units, SKOV-3 and HEY cells were found to express 3 and 0.01%, respectively, of the level of AR mRNA measured in OVCAR-3 cells(Fig. 2,C). OVCAR-3 cells expressed levels of AR mRNA similar to that expressed by LNCaP cells. DU-145 cells did not express detectable AR mRNA. β-Actin levels were comparable for all cell lines, indicating an equivalent amount of RNA was amplified (Fig. 2 C).

Cytosol extracts from ovarian cancer cell lines were analyzed for AR protein by Western blot analysis and by saturation androgen binding assay (Fig. 3 and Table 1). LNCaP and DU-145 cells were included as positive and negative controls, respectively. Western blot analysis revealed a 110-kDa immunoreactive protein in OVCAR-3 and LNCaP cells, whereas no AR immunoreactive protein was detected in DU-145, HOC-7, or OVCA-429 cells, even at higher protein concentration (Fig. 3). A 110-kDa AR-immunoreactive protein was detected in SKOV-3 and HEY cells with 50μg of cytosol protein (Fig. 3).

Saturation binding assays were performed with[3H]R1881, a synthetic radiolabeled androgen (28), to further characterize AR in ovarian cancer cell lines and to determine whether the receptor was capable of ligand binding. Scatchard analysis revealed a single class of high-affinity(Kd < 0.1 nm) and low capacity (Bmax < 30 fmol/mg cytosol protein) androgen binding sites in OVCAR-3, SKOV-3, and HEY cytosol extracts (Table 1). The mean androgen binding capacity in OVCAR-3 cells was approximately 7% of that measured in the AR-positive LNCaP cells (Table 1). SKOV-3 and HEY cells had much lower levels of binding activity, near the limit of sensitivity of the assay (Table 1).

Low-Dose DHT Down-Regulates TβR-II mRNA Expression in Several Ovarian Cancer Cell Lines

Because TβR-II is the ligand-binding receptor and its expression and function are more frequently altered in many human cancers (12), we examined the effect of DHT treatment on its expression in ovarian cancer cell lines. OVCAR-3, SKOV-3, and HEY cells were treated with ethanol vehicle or with 1, 10, or 50 nm DHT, and TβR-II mRNA levels were compared by quantitative RT-PCR. LNCaP and DU-145 prostate cancer cells were included as controls. Autoradiograms from a representative experiment,performed after 24 h of DHT treatment, are shown in Fig. 4. A linear range of exponential amplification for TβR-II mRNA expression was detected between 0.08 and 10 ng of equivalent RNA per lane (Fig. 4). Reduction in TβR-II mRNA levels was observed in SKOV-3 and HEY cells with 1 and 10 nm DHT (Fig. 4,B). With OVCAR-3 cells, the effect was less apparent, whereas with DU-145 cells, no effect was noted, as expected (Fig. 4,B). LNCaP cells did not express detectable TβR-II mRNA (Fig. 4,A), in agreement with previous reports (29). Androgen treatment did not alter β-actin mRNA levels in the different cell lines tested,confirming the specificity of the TβR-II down-regulation (Fig. 4, A and B).

The results of several experiments are combined in Fig. 5. DHT down-regulated the expression of TβR-II mRNA in a dose-dependent manner. Treatment with 1, 10, or 50 nm DHT down-regulated the expression of TβR-II mRNA levels by 30, 80, and 90% in HEY cells and by 70, 90, and 99%, respectively in SKOV-3 cells measured 24 h after addition of DHT to the culture medium (Fig. 5). With OVCAR-3 cells, a decrease in TβR-II mRNA expression was only noted after treatment with 50 nm DHT; however, this did not attain statistical significance (Fig. 5). By 48 h after addition of DHT to the culture medium, the suppression of TβR-II mRNA levels was no longer observed, as shown for SKOV-3 cells (Fig. 5).

OH-FLUT Down-Regulates TβR-II mRNA in SKOV-3 Cells

To further demonstrate that down-regulation of TβR-II mRNA by DHT is mediated by AR, SKOV-3 cells were treated with ethanol vehicle(control) or with 1 or 10 nm DHT in the presence or absence of 1 or 0.1 μm OH-FLUT, an AR antagonist. Additional OH-FLUT was added 12 h later, and cells were harvested 24 h after addition of DHT. OH-FLUT at either dose did not block the effect of DHT on TβR-II mRNA expression. Surprisingly a 90% inhibition of TβR-II mRNA expression was observed with 0.1 and 1.0 μmOH-FLUT in the absence of DHT (Fig. 6), suggesting that OH-FLUT may act as an AR agonist in SKOV-3 cells.

AR Coactivator ARA70 Is Expressed in Ovarian Cancer Cell Lines

Several coactivator proteins that bind steroid receptors and enhance interaction with basal transcription factors have been identified that amplify the transcriptional activation potential of the receptor. One such protein, ARA70, was identified and reported to enhance AR transcriptional activity by approximately 10-fold in the presence of androgen (25). AR antagonists have also been shown to acquire agonist activity in the presence of ARA70 expression (30).

Expression of ARA70 mRNA was detected by RT-PCR as a 509-bp product in all ovarian cancer cell lines and in the two prostate cancer cell lines tested (Fig. 7,A). A linear range of amplification was attained between 3.2 and 400 pg of equivalent RNA with all cell lines, which expressed similar levels of ARA70 (Fig. 7, B and C).β-Actin mRNA levels were comparable for all cell lines (Fig. 7 C).

DHT Blocks the Growth-inhibitory Action of TGF-β1 in Ovarian Cancer Cells

Because low-dose DHT can down-regulate the expression of TβR-II in ovarian cancer cells, we first tested the ability of this steroid to block the growth-inhibitory action of TGF-β1 in SKOV-3 cells. Although it has been reported previously that this cell line is resistant to TGF-β1 (31), we observed that treatment with 5 pm TGF-β1 for 4 days resulted in a 40% decrease in growth (P < 0.05; Fig. 8 A). Maintenance of SKOV-3 cells in 10 nm DHT alone had no effect on growth but prevented the inhibitory effect of TGF-β1.

We then tested whether DHT could reverse growth inhibition by TGF-β1 of ovarian cancer cells derived from ascites. As reported previously for the vast majority of such cultures (19, 32), TGF-β1 inhibited the growth of OVCAS-16 cells. The levels of inhibition observed with an optimal concentration of 25 pm TGF-β1 were 29, 43, and 22% at days 1, 3, and 5, respectively (Fig. 8,B). These values were significantly different from the vehicle control group (P < 0.05). Maintenance of these cells in 10 nm DHT did not have a significant effect on growth; however, it prevented the inhibition by TGF-β1 at all time points, resulting in cell numbers that were not significantly different from control (Fig. 8 B).

In this study, we demonstrate that treatment of ovarian cancer cell lines with the nonaromatizable androgen DHT decreased mRNA expression for TGF-β receptors. DHT treatment also reversed the growth-inhibitory effect of TGF-β1 on SKOV-3 cells, as well as in a primary culture of ovarian cancer cells isolated from patient ascites. Because TGF-β is a potent inhibitor of cell proliferation, including that of ovarian surface epithelial cells (10) and ascites-derived ovarian cancer cells (11, 32), the attenuation of TGF-β-induced growth inhibition by a reduction in receptor expression may represent an important mechanism by which androgens promote ovarian cancer progression.

The ability of DHT to down-regulate TβR-II mRNA levels in SKOV-3 and HEY cells was somewhat surprising given the near-nondetectable levels of AR expressed by these cells. Our findings that high levels of estradiol or progesterone did not inhibit expression of TGF-βreceptor mRNAs in HEY cells indicate that the effect of DHT is specific. Although we were unable to block DHT-induced down-regulation of TβR-II mRNA expression with OH-FLUT, this cannot be taken as evidence of an AR-independent mechanism, as treatment of cells with the AR antagonist alone also resulted in decreased TβR-II expression.

The ability of HEY and SKOV-3 cells to respond to DHT, and the apparent androgen-mimetic action of OH-FLUT on TβR-II mRNA expression, may involve the actions of ARA70. This putative coactivator has a predilection for AR (25) and is highly expressed in ovarian cancer cell lines tested here and in ovarian cancer tissue.5Expression of ARA70 in prostate cancer cells has been shown to increase AR transactivation potential by approximately 10-fold (25), although the magnitude of this effect is controversial (33, 34). Thus, in ovarian cancer cell lines expressing low AR levels, ARA70 could increase their sensitivity to physiologically relevant levels of androgens. Furthermore, the observation that OH-FLUT acts as an androgen agonist in ovarian cancer cells is consistent with the finding by Miyamoto et al.(30) that ARA70 promotes agonist activity by antiandrogens when overexpressed in transfected DU-145 prostate cancer cells. These data raise the possibility that modulation of TβR-II mRNA by androgens may include potentiation of transcription involving ARA70,leading to increased androgen sensitivity of ovarian cancer cells.

Our findings of low levels of AR expression in ovarian cancer cell lines are compatible with those of Lau et al.(35), who recently reported little or no AR mRNA expression in ovarian cancer cells by semiquantitative RT-PCR. We detected AR mRNA in all five ovarian cancer cell lines tested, likely because our quantitative RT-PCR method is more sensitive, due to detection of PCR products by hybridization with labeled AR probes. Two of our cell lines, HOC-7 and OVCA-429, expressed very low levels of AR mRNA and barely detectable levels of receptor protein, at the limit of sensitivity of both Western blot and binding assays.

Whereas normal ovarian surface epithelium cells respond to TGF-β by growth inhibition, many immortalized ovarian cancer cell lines are relatively resistant (10, 32). The growth inhibition of SKOV-3 cells in this study was not observed until 96 h after initiation of TGF-β1 treatment. In contrast, growth inhibition by TGF-β1 was observed after 24 h in ovarian cancer cells isolated from patient ascites. Hurteau et al.(11) have shown that 19 of 20 primary ovarian cancer cells isolated from ascites fluid of patients were growth inhibited by TGF-β1. These findings were confirmed by Havrilesky et al.(32), who reported that 10 of 10 primary cultures were growth inhibited by TGF-β1. TGF-β has been shown to be present in ascites derived from patients with epithelial ovarian carcinoma (36). Furthermore, androgens are produced and secreted by ovarian adenocarcinomas into the ascites fluid (37) and may decrease the expression of TGF-β receptors on tumor cells and alter their responses to TGF-β. The down-regulation of TβR-II mRNA by androgens requires sustained levels of DHT, as indicated by return to control levels 48 h after addition of DHT to SKOV-3 cell cultures. Thus, once removed from the ascites and cultured in the absence of androgens, ovarian cancer cells may regain their sensitivity to TGF-β.

In summary, our data provide a novel mechanism whereby androgens may act to promote ovarian cancer cell growth and progression. The androgen-induced down-regulation of TGF-β receptors occurred in ovarian cancer cells that expressed very low AR levels. Further studies are therefore warranted to define the role of AR co-activators, such as ARA70, in modulating the sensitivity of ovarian cancer cells to DHT.

Fig. 1.

Effect of estradiol-17β (E2),progesterone (P4), and DHT on mRNA expression for TGF-β1 and TGF-β receptors. HEY cells were treated continuously for 60 h with ethanol vehicle (control) or 500 nmE2, P4, or DHT, and total RNA was extracted and subjected to quantitative RT-PCR using specific primers for TGF-β1,TβR-I, TβR-II, betaglycan, and endoglin cDNA. PCR products were fractionated on 4–20% gradient polyacrylamide gels, transferred onto nylon membranes, and hybridized with specific[32P]dATP-labeled internal probes. A, TGF-β1 mRNA levels. Conditions were optimized so that an exponential range of amplification was obtained. The autoradiogram shown is representative of two or three experiments performed. B, TGF-β receptor mRNA levels. Conditions were optimized for each set of primers, and corresponding β-actin was analyzed to ensure that equivalent amounts of cDNA were amplified. C, histogram summarizing 2–4 experiments on TGF-βreceptor mRNA levels, measured after steroid treatment and expressed as a percentage of ethanol vehicle control (mean ± SE) and corrected for β-actin. *, significantly different from control(P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test.

Fig. 1.

Effect of estradiol-17β (E2),progesterone (P4), and DHT on mRNA expression for TGF-β1 and TGF-β receptors. HEY cells were treated continuously for 60 h with ethanol vehicle (control) or 500 nmE2, P4, or DHT, and total RNA was extracted and subjected to quantitative RT-PCR using specific primers for TGF-β1,TβR-I, TβR-II, betaglycan, and endoglin cDNA. PCR products were fractionated on 4–20% gradient polyacrylamide gels, transferred onto nylon membranes, and hybridized with specific[32P]dATP-labeled internal probes. A, TGF-β1 mRNA levels. Conditions were optimized so that an exponential range of amplification was obtained. The autoradiogram shown is representative of two or three experiments performed. B, TGF-β receptor mRNA levels. Conditions were optimized for each set of primers, and corresponding β-actin was analyzed to ensure that equivalent amounts of cDNA were amplified. C, histogram summarizing 2–4 experiments on TGF-βreceptor mRNA levels, measured after steroid treatment and expressed as a percentage of ethanol vehicle control (mean ± SE) and corrected for β-actin. *, significantly different from control(P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test.

Close modal
Fig. 2.

AR mRNA expression in ovarian cancer cell lines. A, detection of AR mRNA by RT-PCR. Total RNA was extracted from OVCAR-3, SKOV-3, HEY, HOC-7, and OVCA-429 ovarian cancer cell lines and subjected to RT-PCR using primers specific for AR. The PCR products were fractionated on 1% agarose gels, transferred onto nylon membrane, and detected by chemiluminescence with a specific digoxigenin-11-ddUTP-labeled internal probe. The blot was overexposed to show detection of low levels of AR. AR-positive LNCaP and AR-negative DU-145 prostate cancer cell lines were used as controls. B, quantitation of AR mRNA by RT-PCR. Conditions were optimized for each cell line so that an exponential range of amplification was obtained. C, graphical analysis of AR mRNA levels. The hybridized bands were scanned and analyzed densitometrically and relative absorbance (ROD) plotted versus equivalent RNA concentration. Lines are drawn through the linear range of exponential amplification. Corresponding β-actin was also analyzed to validate that equivalent amounts of cDNA were amplified.

Fig. 2.

AR mRNA expression in ovarian cancer cell lines. A, detection of AR mRNA by RT-PCR. Total RNA was extracted from OVCAR-3, SKOV-3, HEY, HOC-7, and OVCA-429 ovarian cancer cell lines and subjected to RT-PCR using primers specific for AR. The PCR products were fractionated on 1% agarose gels, transferred onto nylon membrane, and detected by chemiluminescence with a specific digoxigenin-11-ddUTP-labeled internal probe. The blot was overexposed to show detection of low levels of AR. AR-positive LNCaP and AR-negative DU-145 prostate cancer cell lines were used as controls. B, quantitation of AR mRNA by RT-PCR. Conditions were optimized for each cell line so that an exponential range of amplification was obtained. C, graphical analysis of AR mRNA levels. The hybridized bands were scanned and analyzed densitometrically and relative absorbance (ROD) plotted versus equivalent RNA concentration. Lines are drawn through the linear range of exponential amplification. Corresponding β-actin was also analyzed to validate that equivalent amounts of cDNA were amplified.

Close modal
Fig. 3.

Western blot analysis of AR protein expression in ovarian cancer cell lines. Cells were cultured in medium supplemented with 5%sFBS for 48 h before harvesting. Cytosol was extracted, and proteins were separated by SDS-PAGE, transferred onto nylon membrane,and probed with affinity-purified PAR-1 (1 μg/ml). The amount of cytosol protein loaded in each lane and the position of the 110-kDa immunoreactive AR protein are indicated.

Fig. 3.

Western blot analysis of AR protein expression in ovarian cancer cell lines. Cells were cultured in medium supplemented with 5%sFBS for 48 h before harvesting. Cytosol was extracted, and proteins were separated by SDS-PAGE, transferred onto nylon membrane,and probed with affinity-purified PAR-1 (1 μg/ml). The amount of cytosol protein loaded in each lane and the position of the 110-kDa immunoreactive AR protein are indicated.

Close modal
Fig. 4.

Androgen-induced down-regulation of TβR-II mRNA expression in ovarian cancer cell lines. Cells were grown in culture with 5% sFBS and were harvested 24 h after treatment with ethanol vehicle (control) or 1 or 10 nm DHT. Total RNA was extracted and subjected to quantitative RT-PCR using primers specific for TβR-II cDNA. A, a representative set of blots. PCR products were run on 1% agarose gels,transferred to nylon membranes, and hybridized with a specific Digoxigenin-11-ddUTP-labeled internal probe. B,graphical analysis of TβR-II and β-actin mRNA levels.

Fig. 4.

Androgen-induced down-regulation of TβR-II mRNA expression in ovarian cancer cell lines. Cells were grown in culture with 5% sFBS and were harvested 24 h after treatment with ethanol vehicle (control) or 1 or 10 nm DHT. Total RNA was extracted and subjected to quantitative RT-PCR using primers specific for TβR-II cDNA. A, a representative set of blots. PCR products were run on 1% agarose gels,transferred to nylon membranes, and hybridized with a specific Digoxigenin-11-ddUTP-labeled internal probe. B,graphical analysis of TβR-II and β-actin mRNA levels.

Close modal
Fig. 5.

Summary of DHT effects on TβR-II mRNA levels in ovarian cancer cell lines. Bars, mean of several experiments for TβR-II mRNA levels measured after DHT treatment, expressed as a percentage of ethanol vehicle control corrected for β-actin(mean ± SE). *, significantly different from control(P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test.

Fig. 5.

Summary of DHT effects on TβR-II mRNA levels in ovarian cancer cell lines. Bars, mean of several experiments for TβR-II mRNA levels measured after DHT treatment, expressed as a percentage of ethanol vehicle control corrected for β-actin(mean ± SE). *, significantly different from control(P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test.

Close modal
Fig. 6.

OH-FLUT down-regulation of TβR-II mRNA expression in SKOV-3 cells. Cells were cultured for 24 h in medium supplemented with sFBS and ethanol vehicle (control) or 1 or 10 nm DHT in the presence or absence of 0.1 or 1 μm OH-FLUT. Bars, TβR-II mRNA levels measured by quantitative RT-PCR after treatment with DHT and/or OH-FLUT, expressed as a percentage of ethanol vehicle control corrected for β-actin.

Fig. 6.

OH-FLUT down-regulation of TβR-II mRNA expression in SKOV-3 cells. Cells were cultured for 24 h in medium supplemented with sFBS and ethanol vehicle (control) or 1 or 10 nm DHT in the presence or absence of 0.1 or 1 μm OH-FLUT. Bars, TβR-II mRNA levels measured by quantitative RT-PCR after treatment with DHT and/or OH-FLUT, expressed as a percentage of ethanol vehicle control corrected for β-actin.

Close modal
Fig. 7.

ARA70 mRNA expression in ovarian and prostate cancer cell lines. A, detection of ARA70 mRNA by RT-PCR. Total RNA was subjected to RT-PCR using primers specific for ARA70 cDNA. The PCR products were fractionated on 1% agarose gels, stained with ethidium bromide, and photographed. B, quantitation of ARA70 mRNA. The PCR products were transferred and detected by chemiluminescence with a specific digoxigenin-11-ddUTP-labeled internal probe. C, graphical analysis of ARA70 and β-actin mRNA levels.

Fig. 7.

ARA70 mRNA expression in ovarian and prostate cancer cell lines. A, detection of ARA70 mRNA by RT-PCR. Total RNA was subjected to RT-PCR using primers specific for ARA70 cDNA. The PCR products were fractionated on 1% agarose gels, stained with ethidium bromide, and photographed. B, quantitation of ARA70 mRNA. The PCR products were transferred and detected by chemiluminescence with a specific digoxigenin-11-ddUTP-labeled internal probe. C, graphical analysis of ARA70 and β-actin mRNA levels.

Close modal
Fig. 8.

Reversal of the growth-inhibitory action of TGF-β1 in ovarian cancer cells by DHT. A, SKOV-3 cell cultures were treated daily with 10 nm DHT beginning 48 h before the treatment with 5 pm TGF-β1 at 0 and 48 h. Growth of SKOV-3 cells was assessed by the MTT reduction assay with 6–8 replicates per treatment group. Histogram at right shows the percentage of control (vehicle only) for each treatment group. B, growth of primary ovarian cancer cells, OVCAS-16,derived from patient ascites was measured by counting viable cells using four replicates per treatment group. TGF-β1 (25 pm)and 10 nm DHT were added every 24 h to the culture medium, which was replaced after 3 days. The histogram at the right shows the percentage of control for each group. *,significantly different from control (P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test. White columns, with DHT; gray columns, without TGF-β1; black columns, with both DHT and TGF-β1.

Fig. 8.

Reversal of the growth-inhibitory action of TGF-β1 in ovarian cancer cells by DHT. A, SKOV-3 cell cultures were treated daily with 10 nm DHT beginning 48 h before the treatment with 5 pm TGF-β1 at 0 and 48 h. Growth of SKOV-3 cells was assessed by the MTT reduction assay with 6–8 replicates per treatment group. Histogram at right shows the percentage of control (vehicle only) for each treatment group. B, growth of primary ovarian cancer cells, OVCAS-16,derived from patient ascites was measured by counting viable cells using four replicates per treatment group. TGF-β1 (25 pm)and 10 nm DHT were added every 24 h to the culture medium, which was replaced after 3 days. The histogram at the right shows the percentage of control for each group. *,significantly different from control (P < 0.05) as assessed by ANOVA followed by Tukey’s multiple range test. White columns, with DHT; gray columns, without TGF-β1; black columns, with both DHT and TGF-β1.

Close modal

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was supported by grants from the National Cancer Institute of Canada and by a grant from the Natural Sciences and Engineering Research Council of Canada. A. E. was supported by a GENESIS Foundation Studentship and by an Educational Grant from the Division of Gynecological Oncology, the Princess Hospital Margaret Hospital University Health Network. M. L. is a Terry Fox Research Scientist of the National Cancer Institute of Canada.

4

The abbreviations used are: TGF-β,transforming growth factor β; AR, androgen receptor; ARA70,AR-associating protein 70; DHT, 5α-dihydrotestosterone; FBS, fetal bovine serum; sFBS, charcoal-stripped FBS; MTT,3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide;OH-FLUT, hydroxyflutamide; RT-PCR, reverse transcription-PCR; TβR-I,TGF-β receptor I.

5

P. V. C. Rittenberg, P. A. Shaw, and T. S. Brown. Activation of androgen receptor associated protein (ARA 70)expression in ovarian cancer, manuscript in preparation.

Table 1

Saturation binding analysis of AR in cytosol extracts

Cells were grown in culture medium supplemented with 5% sFBS for 48 h before harvesting. Cytosol was extracted and incubated overnight at 4°C with a range of [3H]R1881 concentrations(0.1–12.0 nm) in the presence or absence of 1μ m unlabeled DHT. Values for Bmax(saturation of binding) and Kd (binding affinity)represent the mean ± SE from five independent experiments, except for OVCAR-3 cells (n = 2). For SKOV-3 and HEY cells, specific [3H]R1881 binding was not observed in three of the five experiments; therefore, a value of 0 was assigned for these experiments for statistical purposes. In the two experiments in which binding was detected, Bmaxvalues (fmol/mg cytosol) were 7.06 and 7.33 for SKOV-3, and 11.11 and 3.51 for HEY. ND, binding not detected.

Cell line[3H]R1881 Binding
Mean Bmax ± SE (fmol/mg cytosol protein)Mean Kd ± SE (nm)
LNCaP 314 ± 71.6 0.324 ± 0.082 
OVCAR-3 20.6 ± 2.57 0.03 ± 0.02 
SKOV-3 2.88 ± 1.76 0.11 ± 0.02 
HEY 2.92 ± 2.16 0.10 ± 0.03 
HOC-7 ND  
OVCA429 ND  
DU-145 ND  
Cell line[3H]R1881 Binding
Mean Bmax ± SE (fmol/mg cytosol protein)Mean Kd ± SE (nm)
LNCaP 314 ± 71.6 0.324 ± 0.082 
OVCAR-3 20.6 ± 2.57 0.03 ± 0.02 
SKOV-3 2.88 ± 1.76 0.11 ± 0.02 
HEY 2.92 ± 2.16 0.10 ± 0.03 
HOC-7 ND  
OVCA429 ND  
DU-145 ND  

We acknowledge Drs. Joan Murphy and Barry Rosen for their support and for providing us with ascites fluid material.

1
Risch H. A. Hormonal etiology of epithelial ovarian cancer, with a hypothesis concerning the role of androgens and progesterone.
J. Natl. Cancer Inst.
,
90
:
1774
-1786,  
1998
.
2
Hamilton T. C., Davies P., Griffiths K. Androgen and oestrogen binding in cytosols of human ovarian tumours.
J. Endocrinol.
,
90
:
421
-431,  
1981
.
3
Kuhnel R., de Graaff J., Rao B. R., Stolk J. G. Androgen receptor predominance in human ovarian carcinoma.
J. Steroid Biochem.
,
26
:
393
-397,  
1987
.
4
Slotman B. J., Baak J. P., Rao B. R. Correlation between nuclear DNA content and steroid receptor status in ovarian cancer.
Eur. J. Obstet. Gynecol. Reprod. Biol.
,
38
:
221
-227,  
1991
.
5
Chadha S., Rao B. R., Slotman B. J., van Vroonhoven C. J., van der Kwast H. An immunohistochemical evaluation of androgen and progesterone receptors in ovarian tumors.
Hum. Pathol.
,
24
:
90
-95,  
1993
.
6
Cardillo M. R., Petrangeli E., Aliotta N., Salvatori L., Ravenna L., Chang C., Castagna G. Androgen receptors in ovarian tumors: correlation with oestrogen and progesterone receptors in an immunohistochemical and semiquantitative image analysis study.
J. Exp. Clin. Cancer Res.
,
17
:
231
-237,  
1998
.
7
Helzlsouer K. J., Alberg A. J., Gordon G. B., Longcope C., Bush T. L., Hoffman S. C., Comstock G. W. Serum gonadotropins and steroid hormones and the development of ovarian cancer.
J. Am. Med. Assoc.
,
274
:
1926
-1930,  
1995
.
8
Schildkraut J. M., Schwingl P. J., Bastos E., Evanoff A., Hughes C. Epithelial ovarian cancer risk among women with polycystic ovary syndrome.
Obstet. Gynecol.
,
88
:
554
-559,  
1996
.
9
Silva E. G., Tornos C., Fritsche H. A., El-Naggar A., Gray K., Ordonez N. G., Luna M., Gershenson D. The induction of benign epithelial neoplasms of the ovaries of guinea pigs by testosterone stimulation: a potential animal model.
Mod. Pathol.
,
10
:
879
-883,  
1997
.
10
Berchuck A., Rodriguez G., Oli G., Whitaker R., Boente M. P., Arrick B. A., Clarke-Pearson D. L., Bastf R. C. Regulation of growth of normal ovarian epithelial cells and ovarian cell lines by transforming growth factor-β.
Am. J. Obstet. Gynecol.
,
166
:
676
-683,  
1992
.
11
Hurteau J., Rodriguez G. C., Whitaker R. S., Shah S., Mills G., Bast R. C., Berchuck A. Transforming growth factor-β inhibits proliferation of human ovarian cancer cells obtained from ascites.
Cancer (Phila.)
,
74
:
93
-99,  
1994
.
12
Markowitz S. D., Roberts A. B. Tumor suppressor activity of the TGF-β pathway in human cancers.
Cytokine Growth Factor Rev.
,
7
:
93
-102,  
1996
.
13
Filmus J., Kerbel R. S. Development of resistance mechanisms to the growth-inhibitory effects of transforming growth factor-β during tumor progression.
Cur. Opin. Oncol.
,
5
:
123
-129,  
1993
.
14
Wrana J. L. TGF-β receptors and signaling mechanisms.
Miner. Electrolyte Metab.
,
24
:
120
-130,  
1998
.
15
Lopez-Casillas F., Cheifetz S., Doody J., Andres J. L., Lane W. S., Massague J. Structure and expression of the membrane proteoglycan betaglycan, a component of the TGF-β receptor system.
Cell
,
67
:
785
-795,  
1991
.
16
Barbara N. P., Wrana J. L., Letarte M. Endoglin is an accessory protein that interacts with the signaling receptor complex of multiple members of the transforming growth factor-β superfamily.
J. Biol. Chem.
,
274
:
584
-594,  
1999
.
17
Cheifetz S., Bellon T., Cales C., Vera S., Bernabeu C., Massague J., Letarte M. Endoglin is a component of the transforming growth factor-β receptor system in human endothelial cells.
J. Biol. Chem.
,
267
:
19027
-19030,  
1992
.
18
Lastres P., Letamendia A., Zhang H., Rius C., Almendro N., Raab U., Lopez L. A., Langa C., Fabra A., Letarte M., Bernabeu C. Endoglin modulates cellular responses to TGF-β1.
J. Cell Biol.
,
133
:
1109
-1121,  
1996
.
19
Hirte H. W., Clark D. A., Mazurka J., O’Connell G., Rusthoven J. A rapid and simple method for the purification of tumor cells from ascitic fluid of ovarian carcinoma.
Gynecol. Oncol.
,
44
:
223
-226,  
1998
.
20
Marks A., Ballinger J. R., Reilly R. M., Law J., Baumal R. A novel anti-seminoma monoclonal antibody (M2A) labelled with technetium-99m: potential application for radioimmunoscintigraphy.
Br. J. Urol.
,
75
:
225
-229,  
1995
.
21
Chomczynski P., Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction.
Anal. Biochem.
,
162
:
156
-159,  
1986
.
22
Jindal S. K., Ishii E., Letarte M., Vera S., Teerds K. J., Dorrington J. H. Regulation of transforming growth factor-α gene expression in an ovarian surface epithelial cell line derived from a human carcinoma.
Biol. Reprod.
,
52
:
1027
-1037,  
1995
.
23
Murphy L. D., Herzog C. E., Rudick J. B., Fojo A. T., Bates S. E. Use of the polymerase chain reaction in the quantiation of mdr-1 gene expression.
Biochemistry
,
29
:
10351
-10356,  
1990
.
24
Chang C., Kokontis J., Liao S. Structural analysis of complementary DNA and amino acid sequences of human and rat androgen receptors.
Proc. Natl. Acad. Sci. USA
,
85
:
7211
-7215,  
1988
.
25
Yeh S., Chang C. Cloning and characterization of a specific coactivator, ARA70, for the androgen receptor in human prostate cells.
Proc. Natl. Acad. Sci. USA
,
93
:
5517
-5521,  
1996
.
26
Heisler L. E., Evangelou A., Lew A. M., Trachtenberg J., Elsholtz H. P., Brown T. J. Androgen-dependent cell cycle arrest and apoptotic death in PC-3 prostatic cell cultures expressing a full-length human androgen receptor.
Mol. Cell. Endocrinol.
,
126
:
59
-73,  
1997
.
27
Puy L., MacLusky N. J., Becker L., Karsan N., Trachtenberg J., Brown T. J. Immunocytochemical detection of androgen receptor in human temporal cortex: characterization and application of polyclonal androgen receptor antibodies in frozen and paraffin-embedded tissues.
J. Steroid Biochem. Mol. Biol.
,
55
:
197
-209,  
1995
.
28
Zava D. T., Landrum B., Horwitz K. B., McGuire W. L. Androgen receptor assay with [3H]methyltrienolone (R1881) in the presence of progesterone receptors.
Endocrinology
,
104
:
1007
-1013,  
1979
.
29
Guo Y., Kyprianou N. Overexpression of transforming growth factor (TGF)-β1 type II receptor restores TGF-β1 sensitivity and signaling in human prostate cancer cells.
Cell Growth Differ.
,
9
:
185
-193,  
1998
.
30
Miyamoto H., Yeh S., Wilding G., Chang C. Promotion of agonist activity of antiandrogens by the androgen receptor coactivator, ARA70, in human prostate cancer DU145 cells.
Proc. Natl. Acad. Sci. USA
,
95
:
7379
-7384,  
1998
.
31
Zeinoun Z., Teugels E., de Bleser P. J., Neyns B., Geerts A., de Greve J. Insufficient TGF-β1 production inactivates the autocrine growth suppressive circuit in human ovarian cancer cell lines.
Anticancer Res.
,
19
:
413
-420,  
1999
.
32
Havrilesky L. J., Hurteau J. A., Whitaker R. S., Elbendary A., Wu S., Rodriguez G. C., Bast R. C., Berchuk A. Regulation of apoptosis in normal and malignant ovarian epithelial cells by transforming growth factor-β.
Cancer Res.
,
55
:
944
-948,  
1995
.
33
Gao T., Brantle K., Bol E., McPhaul M. J. RGF, (ARA70, ELE1) interacts with the human androgen receptor in a ligand-dependent fashion, but functions only weakly as a coactivator in cotransfection assays.
Mol. Endocrinol.
,
13
:
1645
-1656,  
1999
.
34
Alen P., Claessens F., Schoenmakers E., Swinnen J. V., Verhoeven G., Rombauts W., Peeters B. Interaction of the putative androgen receptor-specific coactivator ARA70/ELE1α with multiple steroid receptors and identification of an internally deleted ELE1β isoform.
Mol. Endocrinol.
,
13
:
117
-128,  
1999
.
35
Lau K. M., Mok S. C., Ho S. M. Expression of human estrogen receptor-α and -β, progesterone receptor, and androgen receptor mRNA in normal and malignant ovarian epithelial cells.
Proc. Natl. Acad. Sci. USA
,
96
:
5722
-5727,  
1999
.
36
Wilson A. P., Fox H., Scott I. V., Lee H., Dent M., Golding P. R. A comparison of the growth promoting properties of ascitic fluids, cyst fluids and peritoneal fluids from patients with ovarian tumors.
Br. J. Cancer
,
63
:
102
-108,  
1991
.
37
Ridderheim M., Mahlck C. G., Selstam G., Stendahl U., Backstrom T. Peritoneal fluid and hCG influence on steroid production from ovarian tumor tissue in vitro.
Anticancer Res.
,
14
:
1753
-1758,  
1994
.