The Prostate 67:1152^1162 (2007)

Androgen or Receptor-b Blockade Alters DHEA-,DHT-, and E2-Induced Proliferation and PSAProduction in Human Prostate Cancer Cells

Julia T. Arnold,1* Xunxian Liu,1 Jeffrey D. Allen,1 Hanh Le,1 Kimberly K. McFann,2 and Marc R. Blackman1 1Endocrine Section, Laboratory of Clinical Investigation, Division of Intramural Research, National Center for Complementary and Alternative Medicine, National Institutes of Health, Bethesda, Maryland 2University of Colorado Health Sciences Center, Denver,Colorado

BACKGROUND. (DHEA) is an endogenous that is metabolized to androgens and/or in the human prostate. DHEA levels decline with age, and use of DHEA supplements to retard the aging process is of unproved effectiveness and safety. LNCaP and LAPC-4 prostate cancer cells were used to determine whether DHEA- modulated proliferation and prostate specific antigen (PSA) production were mediated via the androgen receptor (AR) and/or ERb. 1 METHODS. Cells were treated with DHEA, DHT, or E2 and antagonists to AR (Casodex - bicalutamide) or ER (ICI 182,780) or siRNA to the respective receptors. Proliferation was assessed by MTT assay and PSA mRNA and protein secretion were measured by quantitative real-time PCR and ELISA. Associations of AR and ERb were analyzed by co-immunoprecipita- tion studies and fluorescent confocal microscopy. RESULTS. DHEA-, T-, and E2-induced proliferation of LNCaP cells was blunted by Casodex but not by ICI treatment. In LNCaP cells, Casodex and ICI suppressed hormone-induced PSA production. In LAPC-4 cells, DHT-stimulated PSA mRNA was inhibited by Casodex and ICI, and the minimal stimulation by DHEA was inhibited by ICI. Use of siRNAs confirmed involvement of AR and ERb in hormone-induced PSA production while AR-ERb co-association was suggested by immunoprecipitation and nuclear co-localization. CONCLUSIONS. These findings support involvement of both AR and ERb in mediating DHEA-, DHT-, and E2-induced PSA expression in prostate cancer cells. Prostate 67: 1152–1162, 2007. # 2007 Wiley-Liss, Inc.

KEY WORDS: DHEA; DHT; E2; AR; ERb; PSA; Casodex; ICI 182,780

INTRODUCTION [3]. Whether the effects of DHEA in LNCaP cells result from binding of DHEA or its androgenic metabolites Dehydroepiandrosterone (DHEA), the most abun- to the mutant AR are uncertain [2]. In comparison, dant endogenous adrenal steroid produced by men and women, declines markedly with aging. It is increas- ingly consumed as an over-the-counter dietary supple- Grant sponsor: Intramural Research Program of the National Center ment for its purported anti-aging effects, yet its for Complementary and Alternative Medicine, National Institutes of usefulness and long-term safety remain uncertain [1]. Health, Bethesda; Grant number: MD20892. DHEA stimulates cell growth and gene and protein *Correspondence to: Julia T. Arnold, PhD, Endocrine Section, LCI, expression in human LNCaP prostate cancer epithelial NCCAM, 9 Memorial Drive, Rm 1N105, Bethesda, MD 20892-0933. cells, and its effects are delayed and reduced compared E-mail: [email protected] Received 12 January 2007; Accepted 20 February 2007 with those of DHT and T, but similar to those of E2 [2]. DOI 10.1002/pros.20585 LNCaP cells harbor a mutated androgen receptor (AR) Published online 14 May 2007 in Wiley InterScience that allows for direct binding of DHEA and E2 to the AR (www.interscience.wiley.com).

2007Wiley-Liss,Inc. AR or ERb Blockade Modulates PSA Production 1153

LAPC-4 cells are human prostate cancer epithelial cells concentrations ranging from 0.1 to 10,000 nM with a that contain a wild-type AR, for which DHEA has final CDS concentration of 1%. Hormone and hormo- minimal affinity [4]. Both LNCaP and LAPC-4 cells also ne þ inhibitor induced cell growth were evaluated at express ERb [5,6] but no significant ERa. It has been 5 days and compared with values from untreated cells. suggested that some of DHEA’s effects may be mediated via ERb [7]. The present study evaluates the RNAExtraction and Real-Time contributions of AR and ERb-related pathways to QuantitativeRT-PCR DHEA-mediated prostate specific antigen (PSA) pro- duction in prostate cancer cells, by comparing the LNCaP or LAPC-4 cells were seeded in triplicate into effects of DHEA, DHT or R1881, and E2 on proliferation 6 or 24-well tissue culture plates pre-coated with and PSA gene and protein expression in LNCaP and Matrigel film (diluted 1:10; BD Biosciences, Beford, LAPC-4 cells, both in the absence and presence of AR MA) at a density of 2 106 cells/6 well using Treatment and ER inhibitors. Physical associations between the Media as described in the proliferation assays. After receptors are explored using co-immunoprecipitation 2–3 days, cells were treated with each hormone and immunofluorescence. Our data suggest that both (1–100 nM), plus Casodex (10 mM) or ICI 182,780 AR and ERb contribute to DHEA-, DHT-, and E2- (1 mM) for 2 days, and then harvested to extract RNA. induced PSA production in LNCaP cells and to DHT- Hormone concentrations were chosen based on pre- modulated PSA production in LAPC-4 cells. vious studies using these hormones in LNCaP cells [2]. Three replicates per point were assayed and data from MATERIALS AND METHODS three separate experiments are presented. Methods for RNA extraction, reverse transcription, real-time PCR, Cell Culture as well as primers for RPLPO and PSA have been LNCaP-FGC cells (androgen-dependent, prostate described previously [2]. adenocarcinoma cells) derived from lymph node metastasis (CRL1435; American Type Culture Collec- PSA ELISA tion, Manassas, VA) and LAPC-4 cells (generously provided by Dr. Charles Sawyers, UCLA) were grown LNCaP or LAPC-4 cells were seeded in triplicate into in DMEM:F12 (1:1) medium, (Invitrogen, Gaithers- 24-well tissue culture plates pre-coated with Matrigel 5 burg, MD) with penicillin (100 units/ml), streptomycin matrix film (1:10) at a density of 5 10 cells/well with (100 mg/ml), L-glutamine (292 mg/ml) (Invitrogen), media and hormone treatments similar to those used and 5% Fetal Bovine Serum (FBS, HyClone Labora- in the proliferation and gene expression assays. Cells tories, Inc., Logan, UT) at 378C in 5% CO2 and were grown in media containing 2% CDS for 2 or propagated at 1:5 dilutions. 3 days. The hormone receptor inhibitors ICI 182,780 (1 mM) and Casodex (10 mM) were added to cultures 30 min prior to addition of hormones (1–100 nM) and Cell Proliferation Assays were incubated for 3 days without refeeding. In the ICI Cell proliferation was estimated using the CellTiter 182,780 dose response experiments, cells were treated 1 96 Assay (MTT assay) (Promega, Madison WI), as with DHT at 10 nM and ICI from 0.1 to 10 mM. After previously described [2] with the following modifica- 3 days, media containing inhibitor and hormones was tions: LNCaP cells were seeded in replicates of 8 into 96- replaced and allowed to condition for 48 hr. Condi- well plates at a density of 20,000 cells/well in tioned media were collected and assayed directly or Treatment Media consisting of Medium 199 (phenol frozen at 808C. Total PSA ELISA kits (DSLabs, red-free): F12 phenol red-reduced media (Invitrogen) Webster, TX) were used to determine PSA concentra- (1:1) supplemented with penicillin (100 units/ml), tions as previously reported [2]. Each original triplicate streptomycin (100 mg/ml), glutamine (100 mg/ml) experimental sample was assayed in duplicate. PSA and 2% charcoal-dextran-treated FBS (CDS—Hyclone values were normalized to cell numbers as determined Labs) for 3 days to down-regulate basal hormonal by the modified MTT assay [2]. Three replicates per activity. Steroid hormones (DHEA, DHT, T, and E2; condition were assayed and data from three separate Sigma-Aldrich, St. Louis, MO), R1881 (PerkinElmer, experiments are presented. Wellesley, MA), and steroid receptor inhibitors, Casodex (kindly supplied by Astra Zeneca, Cheshire, SIRNA Methods England), and ICI 182–780, Diarylpropionitrile (DPN), Propyl pyrazole triol (PPT), and ZK 164,015, (Tocris, Based upon the sequence of the human AR (Gen- Ellisville, MO), were dissolved in ethanol (ethanol bank accession no. NM_000044) and ERb (NM_001437) concentrations not exceeding 0.02%) and added at custom steathTM RNAi oligos (Invitrogen, Carlsbad,

The Prostate DOI 10.1002/pros 1154 Arnold et al.

CA) 25 base pairs in length were designed. Specifically, Immunofluorescence and the sequences used were: AR: 50-ACCGAG GAG- Confocal Microscopy Analyses CUUUCCAGAAUCUGUU-30;ERb50-GUCAAGGC- LAPC-4 cells were cultured on glass cover slips CAUGAUCCUGCUCAAUU-30; Non-specific control: until they were 90% confluent. Media were changed 50-CCAUGGCGCCAAUUCCAAACAGUUU-30. to serum-reduced medium (F12 phenol red-free, M- LNCaP cells were seeded in 24-well plates at a 199 (1:1) and 0.5% CDS and cells were cultured for concentration of 1 105 cells/well in previously des- 2 more days. Cells were left untreated or treated with cribed Treatment Media. After 2 days, serum was DHEA (100 nM), DHT (10 nM), or E (100 nM) and ICI reduced to 1% and cells were treated with transfection 2 (1 mM) or Casodex (10 mM) for 24 hr. Samples were fixed media containing serum-free F12:199 media, oligonu- with 100% MeOH for 10 min at 208C. Cover slips were cleotides (10 pmol/ml AR, 200 pmol/ml ERb, 10 pmol/ washed with PBS (pH 7.4) once and blocked with PBS ml control) with OligofectamineTM (Invitrogen) and with 1% BSA and 0.25% gelatin for 1 hr. Then cells were 1.6 ml/pmol RNA for AR and control, 0.16 ml/pmol treated with both antibodies from Santa Cruz, CA RNA for ERb). Media was spiked with transfection against AR (Rabbit, N-20, 1:100) and ERb (Goat, N-19, media a total of three times every other day. Prior to the 1:50) for 30 min at room temperature. After six washes final treatment, media was replaced with fresh media with PBS with 0.25% gelatin, secondary antibodies of containing 10 nM DHT, DHEA, or E , or ethanol 2 anti-rabbit FITC (green) and anti-goat rhodamine, (red) control. Cells were incubated in hormone-containing (Santa Cruz (1:200 and 1:100, respectively)) for 30 min media for 48 hr. Media was analyzed for PSA at room temperature. Cover slips were rinsed with production by ELISA and cell lysates were used for PBS six times, and mounted onto slides with VECTA- Western blotting, PSA production was normalized to SHIELD HardSet Mounting Medium with DAPI (Vec- cell number using MTT analysis as described pre- tor Labs, Burlingame, CA). Confocal microscopic viously [2]. emitted signals were recorded with the LSM510 meta/axiovert 200 M microscope (Carl Zeiss Micro- Immunoprecipitation and Imaging, Inc.) at the Microscopy and Imaging Core Western Blot Analysis (National Institute of Child Health and Human Devel- opment) with the assistance of V. Schram and Dr. J. T. Cells were treated with steroid hormones and Russell. The plan-apochromat 63 oil (1.4 NA, DIC, hormone inhibitors or siRNA as above. Protein extrac- working distance ¼ 0.17 mm; Carl Zeiss MicroImaging, tion and Western analysis of steroid hormone receptor Inc.) objective lens (Carl Zeiss MicroImaging, Inc.) was levels for siRNA study were performed as described used with the lens immersion medium (Immersol previously [8] using antibodies to AR (Santa Cruz, CA), 518FF, n ¼ 1.518: Carl Zeiss MicroImaging, Inc.) for and ERb (Upstate, Lake Placid, NY). Immunoprecipita- image acquisition. Confocal images were built by tion (IP) protocols have been described previously collecting the intensities from the photo-multiplier [9] with the following modifications. Cells were tube using LSM 5 software version 3.2 (Carl Zeiss collected and lysed in Cell Lysis Buffer (Cell Signaling MicroImaging Inc.). Technologies, Inc., Danvers, MA). Lysates were cen- trifuged for 5 min at 2.5 K rpm at 48C. The lysates were Statistical Analysis normalized for protein concentration, 300 ul cell lysis buffer was added (described previously [9] and Data are expressed as mean values SEM derived samples rotated with either monoclonal AR antibody from 3 to 8 replicates within each of three separate (Santa Cruz: mouse, 441 (targets AA-299-315 region) or experiments. To delineate effects of hormones or polyclonal rabbit antibody, N-20 (targets N-terminus) inhibitors, one way analysis of variance (ANOVA) and Protein G or Protein A sepharose beads (Sigma- was performed using the Tukey-Kramer HSD adjust- Aldrich) overnight at 48C, and precipitated at 2.5 K ment for multiple comparisons. In order to compare rpm. Precipitates were washed three times in 0.1% values within the same treatment groups, Student t- NP40, 0.25% Gelatin, 1 TBS (pH 7.4), 1 mM EDTA, tests were performed. An adjusted P-value of 0.05 was 1mMNa3VO4, and 5 mM NaFl. Sample buffer was considered significant. added to the final precipitates, and the samples were heated for 5 min at 958C. The samples were loaded onto RESULTS NuPAGE (Invitrogen), and the running buffer and Inhibitor Effects on DHEA-Induced transfer buffer always included phosphatase inhibi- LNCaP Proliferation tors. Blots were probed with ERb (Santa Cruz: rabbit H150-polyclonal or goat N-19-polyclonal) or AR DHEA-induced proliferation of LNCaP cells was (rabbit, N-20). evaluated in the absence and presence of the AR

The Prostate DOI 10.1002/pros AR or ERb Blockade Modulates PSA Production 1155

studies with LAPC-4 cells repeatedly showed no significant induction of growth with any of the hormones tested; therefore these cells were not used for proliferation inhibition studies.

PSAGeneandProteinExpressionModulationby Both Casodex and ICI182,780 To determine AR versus ERb mediated pathways of DHEA’s actions, PSA gene and protein expression were measured in steroid hormone- and hormone þ inhibitor-treated LNCaP and LAPC-4 cells.

LNCaP Cells. In LNCaP cells, PSA mRNA stimulation by DHEA, DHT, and E2 was suppressed by both Fig. 1. LNCaP cell Proliferation in response to treatment with Casodex and ICI (Fig. 2A). DHEA-induced PSA DHEA, DHT T, and E2 and the AR Antagonist,Casodex, or the ER production was inhibited 70% (P < 0.01) by Casodex Antagonist,ICI182,780.LNCaPcellswereseededinto96-wellplates and 48% (P < 0.01) by ICI. In comparison, DHT- at 20,000 cells/wellinTreatmentMedia for 2^3 days.DHEA,DHT,T induced PSA production was reduced by 50% or E were added (100 nM) along with steroid receptor inhibitors, 2 (P < 0.01) in the presence of Casodex and by 74% with Casodex (CS-10uM) or ICI182^780 (ICI-100 nM or1 mM). Inhibitor < effects on hormone-induced cell proliferation were evaluated at ICI (P 0.01), whereas E2-induced PSA production 5daysusing theMTTassayandcomparedwithvalues fromuntreated decreased by 80% (P < 0.01) with Casodex and by 43% cells.Data are expressed as mean values SEM derived from eight (P < 0.05) with ICI. Basal concentrations of PSA mRNA replicates within each of four separate experiments. Statisticaldes- in untreated LNCaP cells were relatively high and ignations are consistent throughoutfigures anddistinguishbetween decreased by 69% (P < 0.01) in the presence of ICI. comparison of hormone treatment versus no hormone control (*) Steroid-induced LNCaP PSA protein production and comparison of inhibitor plus hormone versus hormone alone. was also inhibited by AR and ER blockade (Fig. 2B). þ ( ). P values are represented as follows: ***P < 0.0001,**P < 0.001, Basal PSA production in non-steroid-treated LNCaP þþþþ versus non-hormone-treated (0) controls; or P < 0.0001, < þþþ þþ þ cells was decreased 88% (P 0.01) by ICI but was not P < 0.001, P < 0.01, P < 0.05 versus non-inhibitor-treated significantly altered by Casodex. DHEA- and E - samples(control). 2 induced PSA production were inhibited by Casodex by 77% (P < 0.001) and 65% (P < 0.01) respectively, and by ICI by 92% and 90%, respectively (P < 0.001). Casodex treatment reduced DHT-induced PSA pro- duction by 37% which was of borderline significance antagonist, Casodex (10 mM), and the ER antagonist, ICI (P < 0.058), whereas ICI elicited a corresponding 59% 182,780 (100 nM or 1 mM) (Fig. 1) with DHT, T, and E2 (P < 0.001) reduction. There were no significant differ- used as steroid positive controls. These doses of the ences between the inhibition by ICI or Casodex of inhibitors were used throughout the study and were DHEA, DHT-or E2-induced PSA protein levels. Further chosen based on literature review as well as our evaluation found ICI inhibition to be dose dependant preliminary studies showing maximal inhibitory toward DHT (10 nM)-induced PSA protein expression effects with no toxicity (data not shown). LNCaP cell in LNCaP cells (Fig. 3). No decrease in DHT-induced proliferation increased in all hormone treatments (100 PSA protein expression was observed at 0.01–0.1 mM nM) compared to control as measured by MTT assay ICI, whereas significant inhibition was detected with (P < 0.0001). Casodex inhibited proliferation induced 1 mM ICI (35% P < 0.01) and 10 mM ICI (54% P < 0.001). by DHEA (52% inhibition P < 0.001), T (13% P < 0.01), and E2 (19% P < 0.05). ICI treatment (100 nM) was LAPC-4 Cells. Steroid receptor inhibitors were added slightly inhibitory to DHEA-induced growth (11% to DHEA, DHT, or E2-treated LAPC-4 cells to compare decrease P < 0.01). There was no effect of ICI on responses in prostate cancer cells containing a wild- DHT or E2-treated cultures. ICI showed a slight dose type AR with those in LNCaP cells containing a mutant related trend of increased proliferation but was only AR. PSA mRNA expression was increased by DHT significant in the control, and in T-treated cells (16%, 1500-fold (P < 0.0001) versus the low basal amounts of P < 0.05 and 21% P < 0.05, respectively). Parallel PSA found in non-treated LAPC-4 cells (compared to studies using 10 nM concentration of each hormone basal levels in LNCaP cells) (Fig. 4A). DHT-induced showed similar results (data not shown). Proliferation PSA mRNA was inhibited by Casodex (95%;

The Prostate DOI 10.1002/pros 115 6 Arnold et al.

Fig. 3. Effects of increasing concentrations of the ER Blocker,ICI 182,780, on DHT-induced PSA protein expression in LNCaP cells. LNCaP cells were seeded in triplicate into 24-well tissue culture plates pre-coated with Matrigel matrix film (1:10) at a density of 5 105 cells/well in media containing 2% CDS for 2 or 3 days. Cells were treatedwith DHTat10nM and ICI from 0.1to10 mMfor3days. Media containing inhibitor and DHT was replaced and allowed to condition for 48 hr.Conditioned media was collected and assayed by ELISA for PSA concentrations. Cell numbers were estimated using the MTT assay scaled to a 24 well format. þþþP < 0.001, þþP < 0.01,versusnon-inhibitor-treatedsamples(control). Fig. 2. PSA gene and protein expression in LNCaP cells treated with DHEA, DHT, and E2 and the AR Inhibitor,Casodex, or the ER Inhibitor,ICI182,780.Geneexpression(A)LNCaPcellswereplatedin (P < 0.05) but not by Casodex. In comparison, in control triplicate on tissue culture plates pre-coated with Matrigel film at a and E -treated cells, Casodex augmented PSA by 21- 6 2 density of 2 10 cells/6 well, inTreatment Media for 3 days prior to fold and 17-fold; P < 0.001 and P < 0.01, respectively), treatment with DHEA, DHT, or E2 (10 uM) and either ICI 182,780 whereas ICI exerted no significant effects. (1 mM), or Casodex (10 mM), or vehicle control.RNA samples were PSA protein expression in LAPC-4 cells was stimu- extracted, reverse transcribed, and assayed by real-time PCR and < expressed as fold change. Data from three experiments are lated 10-fold by DHT (P 0.0001), whereas neither expressed as % fold change versus control. Protein expression (B) DHEA nor E2 significantly induced PSA protein LNCaP cells were seeded in triplicate into 24-well tissue culture production (Fig. 4B). Casodex did not significantly plates pre-coated with Matrigel matrix film (1:10) at a density of affect DHT-induced PSA protein production, whereas 5 105 cells/well inTreatment Media for 2 or 3 days. DHEA, DHT, co-treatment with ICI decreased DHT-stimulated PSA or E2 (1 ^ 10 0 nM) and ICI (1 mM) or Casodex (10 mM) were added levels by 65% (P ¼ 0.0011). In E2-treated LAPC-4 cells, to cultures and incubated for 3 days.Media containing inhibitor and co-treatment with Casodex elicited a small but sig- hormones was replaced and allowed to condition for 48 hr.Condi- nificant increase in PSA production over treatment tionedmediawascollectedandassayedbyELISAforPSAconcentra- with E2 alone (P < 0.0001). tions.Cell numbers were estimatedusing the MTTassay scaled to a 24wellformat.Threereplicatesperconditionwereassayedanddata from three separate experiments arepresented.P values arerepre- Association Between AR and ERb sented as follows: ***P < 0.0001,**P < 0.001, versus non-hormone- treated (0) controls; þþþþP < 0.0001, þþþP < 0.001, þþP < 0.01, To determine the associations of AR and ERb,AR þP < 0.05 versus non-inhibitor-treated samples (control). from LAPC-4-treated cells was immunoprecipitated and probed for presence of ERb. LAPC-4 cells were treated with DHEA, and E2 at 100 nM and R1881 (at 10 nM) plus or minus inhibitors (concentrations as P < 0.0001) and by ICI (41%; P < 0.01). DHEA and E2- before) for 24 hr. Cells were lysed and incubated with induced PSA mRNA concentrations were very low AR 441 monoclonal antibody. Immunoprecipitated compared to DHT-stimulated levels, even so, DHEA complexes were analyzed by SDS page, and probed increased PSA mRNA 19-fold, an effect blocked by ICI with ERb antibody. Figure 5 is a representative blot

The Prostate DOI 10.1002/pros AR or ERb Blockade Modulates PSA Production 1157

Fig. 5. Co-immunoprecipitation of AR and ERb. LAPC-4 cells were cultured to confluence and treated with DHEA, R1881,and E2 plusICIorCasodexfor24hr.Cellswerecollected,lysed,andincu- bated with an AR antibody (mouse) and Protein G sepharose over- night.Immunoprecipitated(IP)complexeswere dissolvedinloading buffer.SampleswererunonSDS^PAGEandblottedwithanERb antibody(rabbit).BlotswerealsoprobedforAR(110kD)asacontrol. Datarepresentthree separate experiments.

with a polyclonal AR antibody targeting the N- terminus of the AR, this intensity change in ERb was not seen (data not shown). This suggests that AR and ERb are associated.

SiRNA Inhibition of AR and ERb

Fig. 4. PSA gene andprotein expression in LAPC-4 cells treated Association and co-participation of AR and ERb in with DHEA, DHT, and E2 and the AR Inhibitor,Casodex, or the ER steroid hormone-modulated PSA production was Inhibitor,ICI182,780.Geneexpression:(A)LAPC-4cellswereplated further evaluated using siRNAs to AR or ERb in in triplicate on tissue cultureplatespre-coatedwith Matrigelfilm at a LNCaP cells. Diminished AR or ERb protein levels density of 2 10 6 cells/6 well, inTreatment Media for 3 days prior to represented by Western Blot confirm effectiveness of treatment with DHEA, DHT, or E2 (10 mM) and either ICI 182,780 siAR and siERb, respectively (Fig. 6A). The levels of AR (1 mM), or Casodex (10 mM), or vehicle control. RNA samples were protein in control or siERb-treated LNCaP cells extracted, reverse transcribed and assayed by real-time PCR and increased with DHT treatment, and to a lesser extent, expressed as fold change. Data from three experiments are with E treatments, but were unaffected by DHEA. expressed as % change versus control. Protein expression: 2 (B)LAPC-4cellswereseededintriplicateinto24-welltissueculture Cells treated with siAR exhibited reduced AR levels plates pre-coated with Matrigel matrix film (1:10) at a density of after DHEA, DHT or E2 treatment, compared to 5 105 cells/well in Treatment Media for 2 or 3 days. DHEA, DHT, controls treated with a non-specific siRNA. AR protein or E2 (10 ^100 nM) and ICI (1 mM) and Casodex (10 mM) were added levels were slightly diminished with siERb treatment. to cultures and incubated for 3 days. Media containing inhibitor (Fig. 6A). ERb protein expression was decreased in cells and hormones was replaced and allowed to condition for 48 hr. treated with siERb compared to siAR or control siRNA Conditioned media was collected and assayed by ELISA for PSA and did not appear to differ in hormone treated versus concentrations.Cell numbers were estimated using the MTTassay control cells. scaled to a 24 well format. Three replicates per condition were Blocking steroid receptors by siRNA inhibited assayed and data from three separate experiments are presented. hormone-induced PSA protein secretion similar to that ***P < 0.0001, **P < 0.001, *P < 0.01 versu non-hormone-treated observed using pharmacological blockade by ICI (0) controls. þþþþP < 0.0001, þþþP < 0.001, þþP < 0.01, þP < 0.05 versus non-inhibitor-treated samples (control). 182,780. Treatment of LNCaP cells with DHEA, DHT, or E2 (10 nM each) increased production of PSA by 77% (P < 0.01), 80% (P < 0.001), and 63% (P < 0.01), respec- from three replicate experiments showing co-immuno- tively (Fig. 6B). In control samples treated with siRNA, precipitation of ERb with AR. ERb was found to be but not with hormones, neither siAR nor siERb exerted associated with the AR in all treatment groups. An any significant effect. In contrast, in samples treated increase in intensity was noted in cultures treated with with DHEA, siAR reduced PSA secretion by 71% R1881 þ ICI. When the same lysates were incubated (P < 0.01), and siERb reduced PSA production by 65%

The Prostate DOI 10.1002/pros 1158 Arnold et al.

Fig. 6. Effects of siAR or siERb administration on AR and ERb Fig. 7. Immunofluorescent determination of co-localization of proteinexpression and PSAproductionin LNCaPcells.LNCaPcells AR and ERb. LAPC- 4 cells were cultured on glass cover slips and were seededin 24-wellplates at a concentration of1 105 cells/well treatedwithDHEA(100nM),DHT(10nM),orE2 (10 0 nM) and ICI in previously described Treatment Media. Cells were transfected (1 mM) or Casodex (10 mM) for 24 hr.Slips wereprocessedfor immu- TM with siRNA oligonucleotides to AR, ERb with Oligofectamine a nofluorescent staining with both antibodies against AR (FITC- total of three times every other day. Prior to the final treatment, green) and ERb (rhodamine-red) with nuclei visualized using DAPI media was replaced with fresh media containing DHT, DHEA, E2 and were visualized by confocal microsopy.Data represented are (10 nM) or ethanol control.Cells were incubated in hormone con- from R1881-treated cells where both AR (E)andERb (D)show tainingmedia for 48 hr and analyzed for receptor levels by Western increased expression over control (A,B) and were found to be blotting (A) and for PSA production by ELISA (B). PSA production co-expressed in the nucleus as visualized in the merged image by was normalized to cell number using MTT analysis as above, thewhite staining (F).Noco-expressionofreceptorsisfound incon- **P < 0.001, *P < 0.01 versus non-hormone-treated (0) controls. trol (C)orR1881þICI (I)orR1881þCS (L) treated cultures. The þþþ þþ þ P < 0.001, P < 0.01, P < 0.05 versus non-inhibitor-treated fluorescence of both receptors was diminished in the R1881plus ICI samples(control). or Casodex-treatedcultures(G,H,J,K) andlessnuclearlocalization wasnoted.

(P < 0.05). Similarly, siAR and siERb reduced PSA confocal microscopy. AR and ERb levels were not production in DHT-treated samples by 52% (P < 0.05) increased significantly by treatment with DHEA or E2, and 62% (P < 0.001), respectively, and in E2-treated therefore, any alteration by the inhibitors was not samples, by 68% (P < 0.01) and 64% (P < 0.01), respec- evident (data not shown). In R1881-treated cells, both tively. There were no significant differences between AR (green fluorescence Fig. 7E) and ERb (red fluores- the siAR versus siERb inhibition on DHEA, DHT or E2- cence Fig. 7D) showed increased expression over modulated PSA production. Comparable results with control (Fig. 7 A,B) and were found to be co-expressed LAPC-4 cells were not obtained, as multiple attempts to in the nucleus as visualized in the merged image by the transfect the cells with the siRNA oligonucleotides white staining (Fig. 7F). The fluorescence of both were unsuccessful. receptors was diminished in the R1881 plus ICI or Casodex-treated cultures (Fig. 7G,H,K,L) and less Immunofluorescent Determination of nuclear localization was noted. Co-localization of AR and ERb DISCUSSION LAPC-4 cells were treated with steroid hormones plus ICI or Casodex, processed for immunofluorescent Interactions among androgens, estrogens, and their staining of AR and ERb and were visualized by receptors contribute to normal prostate development

The Prostate DOI 10.1002/pros AR or ERb Blockade Modulates PSA Production 1159 and function [10–13], and to dysregulation of prostate The AR antagonist, Casodex, inhibited DHEA-, T-, growth, potentially promoting hormone-independent and E2-induced proliferation of LNCaP cells. This was prostate cancer [14,15]. DHEA is an important source expected as these hormones are known to transactivate of both androgenic and estrogenic ligands in the the mutated AR in the LNCaP cells [26]. The ER human prostate. DHEA sulfate (DHEAS) is present antagonist, ICI at 100 nM was slightly inhibitory to in high levels in the prostate, as is the sulfatase DHEA-induced proliferation. The slight but repeatable that converts DHEAS to DHEA [16]. Receptors trend toward increased proliferation seen in the control forDHEAorDHEAShavenotbeendefinitively and T-treated cultures may not be of biological isolated [17]. Prostate cells contain 3b-and17b- significance (16–21% change), yet it does not confirm hydroxysteroid dehydrogenases and can metabolize previously reported growth inhibitory effects of ICI in DHEA to DHT [18], providing up to one sixth of total R1881-treated cultures [27]. Our data also contrast with prostatic DHT [19]. Thus, DHEA can behave as a weak reports of inhibition of proliferation in E2 and DHT- androgen, potentially promoting prostate cancer treated LNCaP cells using ICI 182,780 or an ERb growth, as shown by its stimulation of LNCaP cell antisense expression vector [28]. The differences may proliferation, and modulation of cellular PSA, AR, be due to different culture conditions where our ERb, and IGF axis gene and protein expression, in a medium has minimal factors added and includes only pattern similar to DHT and T, although on a lesser 1% CDS. This proliferation increase with ICI treatment scale and delayed in time [2]. Alternatively, DHEA was also found repeatedly in parallel studies with the can act on ERb in the prostate, potentially antagoniz- LAPC-4 cells (data not shown). ERb is considered to ing androgenic effects on prostate cancer growth. antagonize androgen-driven prostate proliferation, as DHEA effects on ERb may be indirect via its evidenced by increased prostatic hyperplasia in ERb metabolism to 7a-hydroxy-DHEA (7HD), a known knock-out mice [29]. The increase in growth could be ligand for ERb [7] or direct, as observed in competitive due to bound ICI inhibiting ERb antagonistic ability to receptor-binding assays, in which DHEA displayed a AR, thus the proliferation may continue unchecked. higher affinity for ERb thanforARorERa,withERb ERb’s function in cell proliferation either free or bound being the preferred target for the transcriptional by ICI is not clear. ICI bound to ERb in MCF-7 cells was effects of DHEA [20,21]. shown to stabilize ERb compared to the opposite effect DHEA’s potential importance in the prostate is on ERa where ICI binding induces degradation further elaborated in a hypothesis we are currently mediated via a proteasome-mediated pathway [30]. examining that alterations in preferential DHEA ER ligands change conformation of ERa and ERb [31]. metabolism to androgens or estrogens may perturb Binding of ICI may interfere with how ERb potentially prostate tissue homeostasis. Proinflamatory cytokines interacts with the AR, as our co-immunoprecipitation which are present in pre-cancerous lesions, such as data suggest. prostatic inflammatory atrophy (PIA) [22], or higher In LNCaP cells, DHEA-, DHT-, and E2-stimulated levels of TGFb that promote a reactive stroma pheno- PSA gene and protein expression were each inhibited type [23] may increase levels of metabolizing enzymes by Casodex and ICI. Blocking AR by Casodex or by (3b-HSD) which convert DHEA to androgens and si-AR inhibited effects of estrogenic hormones, and estrogens [24]. Thus the otherwise benign presence of blocking ERb by ICI or by si-ERb inhibited effects of DHEA in an altered tissue microenvironment could androgenic hormones. These data confirm and extend become one of the many factors contributing to prostate our previous report that DHEA and E2 can induce PSA cancer progression. production in these prostate cancer cells with a mutant To investigate the relative contribution of androge- AR [2], and further suggest that blocking ERb as well as nic versus estrogenic pathways involved in DHEA- the AR can modulate the effects of both . mediated effects in prostate cells, two human prostate In comparison, in LAPC-4 cells, DHEA induction of cancer cell models, LNCaP and LAPC-4 cells, were PSA mRNA was minimal, compared with the much evaluated for proliferation and PSA gene and protein greater effects of DHT. In this sense, LAPC-4 cells expression, both in the absence and presence of AR and would likely represent DHEA effects on the wild-type ER antagonists, Casodex and ICI 182,780. LNCaP cells, AR in normal prostate epithelium. DHEA is not a direct which contain a mutant AR, and LAPC-4 cells, which ligand for the wild-type AR. However, the presence of have a wild-type AR that does not bind DHEA or E2 steroid hormone metabolizing enzymes in LAPC-4 appreciably [25] both expressed ERb (by RT-PCR and cells could produce androgenic metabolites, such as Western Blot), whereas ERa was undetectable (unpub- or DHT, which, even at low concentra- lished observations), confirming prior reports of the tions, might be sufficient to stimulate PSA. The presence of ER isoforms in LNCAP [5] and LAPC-4 [6] inhibitory effect of ICI on DHEA is similar to that of cells. Zhu et al. [6] in LAPC-4 cells in the presence of AR, ERa,

The Prostate DOI 10.1002/pros 116 0 Arnold et al. and ERb (co-transfected 1:1:1). Estrogens were shown induced ERb-specific target in the prostate that would to inhibit DHT-induced PSA gene expression as well as assist in defining ERb activity. For the present study, cell growth, and ICI 182,780 inhibited DHT-induced pS2 mRNA levels were assayed by real-time PCR as PSA transcription activity. Our data complement an estrogen-responsive gene [5], but were found to the latter findings and further demonstrate that ICI have no significant changes in expression after E2- 182,780 diminishes DHEA and DHT-induced PSA treatment in either LNCaP or LAPC-4 cells (data not production in LAPC-4 and LNCaP cells, as well as E2- shown). induced PSA production in the LNCaP cells. DHEA- These data prompt the question of the role of ERb in induced PSA production in LAPC-4 cells was inhibited production of PSA, a prototypic androgen responsive by ICI and not by Casodex, implicating the ER in gene. Also, what are the potential mechanisms of DHEA-induced LAPC-4 PSA production, further interaction between AR and ERb? There has been no supporting a link between the functions of AR and report of an estrogen responsive element in the PSA ERb in hormonal responses. Casodex inhibited the promoter. Zhu et al. [6] commented on their unpub- DHT-induced increase in LAPC-4 PSA mRNA, lished gel shift data showing that neither ERa nor ERb whereas it elicited a small effect in control could bind directly to any of the three androgen and E2-treated cells. Casodex was not inhibitory to responsive elements in the human PSA promoter, DHT-induced PSA protein expression and was even decreasing the likelihood that the ligand-ER competes stimulatory in control or E2-treated cells. This switch of directly with ligand-AR to bind to the ARE. Our study Casodex from antagonist to agonist has been pre- attempted to determine if there was any physical viously reported [32] in a sub-line of LNCaP cells that association between AR and ERb by immunoprecipita- had become androgen independent. Other reports also tion of the AR from cell lysates and probing with an show mixed antagonist and agonist effects in Casodex- ERb antibody. The ERb protein was associated with the treated prostate cells [8,33,34]. The mechanistic basis of AR as seen by the co-immunoprecitpitation. The these altered functions of Casodex remains to be deter- increase in intensity in the R1881-ICI treatment may mined. Of note, in both cell types, blocking ERb, either be from the differential ERb association with the AR pharmacologically or by siRNA, was as effective as as a result of a conformational change, increasing blocking the AR in inhibiting steroid-modulated PSA exposure of the AR antibody epitope of the AR/ER production. This gives further support to a potential complexes and therefore pulling down more AR/ERb linkage between the two receptors. complex compared with other treatments. These The effects of ICI 182,780 on the AR have been well results were not seen when the AR/ERb complex characterized by Bhattacharyya et al. [27] who report included the polyclonal AR antibody N-20 which that ICI suppressed AR mRNA and protein levels and targeted the N terminus. Because the epitope of the transcriptional activity. Also ICI decreased AR gene AR monoclonal antibody is in the middle of the transcription and PSA mRNA and protein in both receptor (299–315 amino acids), the antibody may LNCaP and LAPC-4 cells similar to our results. While be more sensitive to change of association of the the authors showed that ICI neither binds directly to the complexes. Treatment with inhibitor or hormone may AR nor degrades the receptor as it does in the ERa, they change the associations, which may or may not be speculated that ICI may suppress AR transcription via detectable by this method. ERb and by potential estrogen response elements on the Further probes into the nature of the AR/ERb AR promoter. The current report of siERb treatment receptor interaction were made using immunofluor- decreasing PSA production provides further support escent confocal microscopic visualization. This analysis for the involvement of ERb in androgen-driven PSA showed that the AR and ERb levels were increased with production. R1881 treatment and co-localized to the nucleus. ICI In the current study, additional ERa/b and Casodex decreased both AR and ERb receptor and an ERa antagonist were assessed for their ability levels and nuclear translocation. Based on confocal and to reproduce the ICI 180,782-induced inhibition of co-immunoprecititation data it is possible that ICI PSA production in either LNCaP or LAPC-4 cells. DPN, treatment could affect AR-mediated transcription of an ERb selective agonist [35], PPT, an ERa selective PSA due to potential changes in the conformation of agonist [36], and ZK 164,015, an ERa antagonist [37] AR/ERb. These results do not distinguish between a were assayed for modulation of DHT-induced PSA direct versus indirect interaction of AR and ERb or production in both cell types in ELISA assays run in identify co-factors or co-regulators involved in the parallel to those reported herein. There were no protein complex that they share. Nonetheless, the data inhibitory or stimulatory effects on PSA production demonstrate an association between AR and ERb, the by either the ERa or b agonists or the ERa antagonist details and implications of which are subject of further (data not shown). We are unaware of any hormone- investigations.

The Prostate DOI 10.1002/pros AR or ERb Blockade Modulates PSA Production 1161

The intersection of the AR and ER pathways has White J, editors. Encyclopedia of dietary supplements, 1st ed. been reported to occur at the level of the membrane, New York, NY: Marcel Dekker, Inc.; 2005:167–176. where both receptors formed a complex with Src, and 2. Arnold JT, Le H, McFann KK, Blackman MR. Comparative stimulated LNCaP cell proliferation [38]. Our results effects of DHEA vs. testosterone, , and on proliferation and gene expression in human LNCaP suggest an additional interaction between the receptor prostate cancer cells. Am J Physiol Endocrinol Metab 2005; pathways at the genomic level, in PSA gene expression. 288(3):E573–E584. Interestingly a direct interaction between AR and ER 3. Veldscholte J, Berrevoets CA, Ris-Stalpers C, Kuiper GG, Jenster has been reported in a yeast and mammalian two- G, Trapman J, Brinkmann AO, Mulder E. The androgen receptor hybrid system [39], in which interactions were char- in LNCaP cells contains a mutation in the ligand binding domain acterized between the N-terminus of AR and the which affects steroid binding characteristics and response to a antiandrogens. J Steroid Biochem Mol Biol 1992;41(3–8):665– ligand-binding domain of ER ; there was, however, 669. no interaction between the ligand-binding domain of 4. Williams MR, Ling S, Dawood T, Hashimura K, Dai A, Li H, Liu ERb and the AR. JP, Funder JW, Sudhir K, Komesaroff PA. Dehydroepiandroster- The molecular mechanisms underlying AR-ERb one inhibits human vascular smooth muscle cell proliferation interactions in prostate cancer cells remain an active independent of ARs and ERs. J Clin Endocrinol Metab 2002; area of research. The current study was designed to 87(1):176–181. assess whether the DHEA effects in prostate cancer 5. Lau KM, LaSpina M, Long J, Ho SM. Expression of (ER)-alpha and ER-beta in normal and malignant cells occurred via androgenic or estrogenic pathways as prostatic epithelial cells: Regulation by methylation and invol- assessed by blocking respective receptors in response vement in growth regulation. Cancer Res 2000;60(12):3175– to DHEA, and comparing effects with those of DHT or 3182. E2. DHEA-induced PSA gene and protein expression 6. Zhu YS, Cai LQ, Huang Y, Fish J, Wang L, Zhang ZK, Imperato- were suppressed by ICI, but not by CS, in LAPC-4 cells, McGinley JL. Receptor isoform and ligand-specific modulation and were equally suppressed by both inhibitors in of dihydrotestosterone-induced prostate specific antigen gene expression and prostate tumor cell growth by estrogens. LNCaP cells. Both the AR and ERb pathways are J Androl 2005;26:4500–4508; discussion 509–510. involved in DHEA- and other steroid-mediated PSA 7. Martin C, Ross M, Chapman KE, Andrew R, Bollina P, Seckl JR, production in prostate cells. Habib FK. CYP7B generates a selective agonist in human prostate. J Clin Endocrinol Metab 2004;89(6): CONCLUSIONS 2928–2935. 8. Le H, Arnold JT, McFann KK, Blackman MR. Dihydrotestoster- In LNCaP and LAPC-4 cells, steroid hormone- one, Testosterone, but not DHEA or Estradiol, differentially induced PSA secretion was decreased by blocking AR modulate IGF-I, IGFBP-2 and IGFBP-3 Gene and protein or ERb, suggesting interaction or ‘‘cross-talk’’ between expression in primary cultures of human prostatic stromal cells. the mutant or wild-type AR and ERb in regulation of Am J Physiol Endocrinol Metab 2005;290(5):E952–E960. PSA mRNA expression in these cells. Whether the 9. Liu X, Rubin JS, Kimmel AR. Rapid, Wnt-induced changes in GSK3beta associations that regulate beta-catenin stabilization effects of DHEA on the AR or ERb occur directly or are mediated by Galpha proteins. Curr Biol 2005;15(22):1989– indirectly through the actions of DHEA’s metabolites 1997. are under current investigation. These data further 10. Ho SM. Estrogens and anti-estrogens: Key mediators of prostate support the concept that ERb is a potential target in the carcinogenesis and new therapeutic candidates. J Cell Biochem treatment of prostate cancer, and that further investiga- 2004;91(3):491–503. tions of the interactions between ERb and AR may 11. Prins GS, Marmer M, Woodham C, Chang W, Kuiper G, contribute to our understanding of the effects of DHEA, Gustafsson JA, Birch L. Estrogen receptor-beta messenger ribonucleic acid ontogeny in the prostate of normal and androgens and estrogens in prostate physiology and neonatally estrogenized rats. Endocrinology 1998;139(3):874–883. pathology. 12. Soronen P, Laiti M, Torn S, Harkonen P, Patrikainen L, Li Y, Pulkka A, Kurkela R, Herrala A, Kaija H, Isomaa V, Vihko P. Sex ACKNOWLEDGMENTS steroid hormone metabolism and prostate cancer. J Steroid Biochem Mol Biol 2004;92(4):281–286. This work was supported by the Intramural 13. Risbridger GP, Bianco JJ, Ellem SJ, McPherson SJ. Oestrogens Research Program of the National Center for Comple- and prostate cancer. Endocr Relat Cancer 2003;10(2):187–191. mentary and Alternative Medicine, National Institutes 14. Heinlein CA, Chang C. Androgen receptor in prostate cancer. of Health, Bethesda, MD20892. The authors thank Endocr Rev 2004;25(2):276–308. Dr. Shuk-Mei Ho and Dr. Irini Manoli for their con- 15. Arnold JT, Isaacs JT. Mechanisms involved in the progression of structive comments in reviewing this manuscript. androgen-independent prostate cancers: It is not only the cancer cell’s fault. Endocr Relat Cancer 2002;9(1):61–73. REFERENCES 16. Klein H, Molwitz T, Bartsch W. Steroid sulfate sulfatase in human benign prostatic hyperplasia: Characterization and 1. Alesci S, Manoli I, Blackman MR. Dehydrodepiandrosterone quantification of the enzyme in epithelium and stroma. J Steroid (DHEA). In: Coates P, Blackman MR, Cragg G, Levine M, Moss J, Biochem 1989;33(2):195–200.

The Prostate DOI 10.1002/pros 1162 Arnold et al.

17. Widstrom RL, Dillon JS. Is there a receptor for dehydroepian- 30. Peekhaus NT, Chang T, Hayes EC, Wilkinson HA, Mitra SW, drosterone or dehydroepiandrosterone sulfate? Semin Reprod Schaeffer JM, Rohrer SP. Distinct effects of the Med 2004;22(4):289–298. Faslodex on the stability of estrogen receptors-alpha and -beta in 18. Labrie F. Intracrinology. Mol Cell Endocrinol 1991;78(3):C113– the breast cancer cell line MCF-7. J Mol Endocrinol 2004;32(3): C118. 987–995. 19. Geller J, Albert JD. Adrenal androgen blockade in relapsed 31. Van Den Bemd GJ, Kuiper GG, Pols HA, Van Leeuwen JP. prostate cancer. Eur J Cancer Clin Oncol 1985;21(10):1127–1131. Distinct effects on the conformation of 20. Chen F, Knecht K, Birzin E, Fisher J, Wilkinson H, Mojena M, and beta by both the ICI 164,384 and ICI 182,780 leading to opposite effects on receptor stability. Biochem Moreno CT, Schmidt A, Harada S, Freedman LP, Reszka AA. Direct agonist/antagonist functions of dehydroepiandroster- Biophys Res Commun 1999;261(1):1–5. one. Endocrinology 2005;146(11):4568–4576. 32. Hobisch A, Hoffmann J, Lambrinidis L, Eder IE, Bartsch G, 21. Arnold JT, Blackman MR. Does DHEA exert direct effects on Klocker H, Culig Z. Antagonist/agonist balance of the nonster- oidal antiandrogen bicalutamide (Casodex) in a new prostate androgen and estrogen receptors, and does it promote or prevent prostate cancer? Endocrinology 2005;146(11):4565–4567. cancer model. Urol Int 2000;65(2):73–79. 22. De Marzo AM, Marchi VL, Epstein JI, Nelson WG. Proliferative 33. Hara T, Miyazaki J, Araki H, Yamaoka M, Kanzaki N, Kusaka M, inflammatory atrophy of the prostate: Implications for prostatic Miyamoto M. Novel mutations of androgen receptor: A possible carcinogenesis. Am J Pathol 1999;155(6):1985–1992. mechanism of bicalutamide withdrawal syndrome. Cancer Res 2003;63(1):149–153. 23. Singh H, Dang TD, Ayala GE, Rowley DR. Transforming growth factor-beta1 induced myofibroblasts regulate LNCaP cell death. 34. Yoshida T, Kinoshita H, Segawa T, Nakamura E, Inoue T, J Urol 2004;172 (6 Pt 1):2421–2425. Shimizu Y, Kamoto T, Ogawa O. Antiandrogen bicalutamide promotes tumor growth in a novel androgen-dependent 24. Simard J, Gingras S. Crucial role of cytokines in sex steroid prostate cancer xenograft model derived from a bicalutamide- formation in normal and tumoral tissues. Mol Cell Endocrinol treated patient. Cancer Res 2005;65(21):9611–9616. 2001;171(1–2):25–40. 35. Sun J, Baudry J, Katzenellenbogen JA, Katzenellenbogen BS. 25. Klein KA, Reiter RE, Redula J, Moradi H, Zhu XL, Brothman AR, Molecular basis for the subtype discrimination of the estrogen Lamb DJ, Marcelli M, Belldegrun A, Witte ON, Sawyers CL. receptor-beta-selective ligand, diarylpropionitrile. Mol Endocri- Progression of metastatic human prostate cancer to androgen nol 2003;17(2):247–258. independence in immunodeficient SCID mice. Nat Med 1997;3(4):402–408. 36. Frasor J, Barnett DH, Danes JM, Hess R, Parlow AF, Katzenellenbogen BS. Response-specific and ligand dose- 26. Zhao XY, Boyle B, Krishnan AV, Navone NM, Peehl DM, dependent modulation of estrogen receptor (ER) alpha activity Feldman D. Two mutations identified in the androgen receptor by ER beta in the uterus. Endocrinology 2003;144(7):3159– of the new human prostate cancer cell line MDA PCa 2a. J Urol 3166. 1999;162(6):2192–2199. 37. Biberger C, von Angerer E. 2-Phenylindoles with sulfur contain- 27. Bhattacharyya RS, Krishnan AV, Swami S, Feldman D. Fulves- ing side chains. Estrogen receptor affinity, antiestrogenic trant (ICI 182,780) down-regulates androgen receptor expression potency, and antitumor activity. J Steroid Biochem Mol Biol and diminishes androgenic responses in LNCaP human prostate 1996;58(1):31–43. cancer cells. Mol Cancer Ther 2006;5(6):1539–1549. 28. Maggiolini M, Recchia AG, Carpino A, Vivacqua A, Fasanella G, 38. Migliaccio A, Castoria G, Di Domenico M, de Falco A, Rago V, Pezzi V, Briand PA, Picard D, Ando S. Oestrogen Bilancio A, Lombardi M, Barone MV, Ametrano D, Zannini MS, Abbondanza C, Auricchio F. Steroid-induced androgen receptor beta is required for androgen-stimulated proliferation of LNCaP prostate cancer cells. J Mol Endocrinol 2004;32(3):777– receptor-oestradiol receptor beta-Src complex triggers prostate 791. cancer cell proliferation. EMBO J 2000;19(20):5406–5417. 29. Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler 39. Panet-Raymond V, Gottlieb B, Beitel LK, Pinsky L, Trifiro MA. JF, Sar M, Korach KS, Gustafsson JA, Smithies O. Generation Interactions between androgen and estrogen receptors and the and reproductive phenotypes of mice lacking estrogen receptor effects on their transactivational properties. Mol Cell Endocrinol beta. Proc Natl Acad Sci USA 1998;95(26):15677–15682. 2000;167(1–2):139–150.

The Prostate DOI 10.1002/pros