Inhibition of the Androgen Receptor by Mineralocorticoids at Levels Physiologically Achieved in Serum in Patients Treated with Abiraterone Acetate

Home , CYP17A1 inhibitor

ORIGINAL ARTICLE Inhibition of the androgen receptor by mineralocorticoids at levels physiologically achieved in serum in patients treated with abiraterone acetate

W Kim1, JO Jones2, M Diamond3, C Haqq4, A Molina5, EJ Small1 and CJ Ryan1

BACKGROUND: Abiraterone acetate (AA), a highly potent CYP17A1 inhibitor, has demonstrated marked clinical benefit in patients with metastatic castration-resistant prostate cancer (CRPC). Phase I trials of AA without prednisone showed significant elevation of serum mineralocorticoid concentrations. The aim of this study was to elucidate the biological significance of elevated mineralocorticoid levels on androgen receptor (AR) activity in prostate cancer (PC) cells. METHODS: Fluorescence resonance energy transfer (FRET) assay was used to assess the effect of mineralocorticoids on androgen- induced conformational change of the AR. LAPC4, LNCaP and LN-AR cells that were cultured and treated with androgens were exposed to mineralocorticoids at varying concentrations, including levels measured in the serum of AA-treated patients in a phase I trial. AR-dependent transcriptional activity and cell growth were measured in these cell lines to determine the biological impact of mineralocorticoids on PC cells. RESULTS: Corticosterone (CS) and deoxycorticosterone (DOC) inhibited androgen-induced conformational change of the AR in the FRET assay. CS inhibited AR-dependent transcriptional activity and cell growth at concentrations comparable to those measured in the serum of AA-treated patients. DOC inhibited AR transcriptional activity and cell growth at 10-fold greater concentrations than measured in the serum of AA-treated patients. CONCLUSIONS: Mineralocorticoids directly inhibit androgen-induced conformational change of the AR. CS inhibits AR transcriptional activity and PC cell growth at concentrations found in the serum of patients treated with AA. Further investigation of the potential therapeutic implications of mineralocorticoids in AA-treated CRPC patients is warranted.

Prostate Cancer and Prostatic Disease (2014) 17, 292–299; doi:10.1038/pcan.2014.27; published online 22 July 2014

INTRODUCTION studies of men with CRPC.10,11 Some of the notable adverse Retained androgen receptor (AR) activity despite castrate levels events associated with AA results directly from CYP17A1 inhibition of circulating testosterone (T) is a critical biological hallmark leading to decreased adrenal androgen synthesis with a com- associated with the progression of prostate cancer (PC) to its lethal pensatory rise in adrenocorticotropic hormone production.12,13 form.1 Ongoing androgen biosynthesis via cytochrome P450 17A1 This then leads to markedly increased levels of adrenal steroid (CYP17A1) is one process that contributes to the maintenance of hormones upstream of CYP17A1, most notably precursors of AR activity in castration-resistant prostate cancer (CRPC). CYP17A1, aldosterone with potent mineralocorticoid activity (Figure 1). In a complex of enzymes present in testes, adrenal glands and phase I studies of AA alone, on-treatment serum mineralocorticoid prostatic tissue, is the rate-limiting step in the conversion of levels (speciﬁcally corticosterone (CS) and deoxycorticosterone steroid hormones into androgens within these organs, and is most (DOC)) increased 11- to 85-fold over baseline levels, leading to responsible for the remaining circulating androgens in men after classical symptoms associated with mineralocorticoid excess, medical or surgical castration.2–4 In men treated with luteinizing including hypokalemia, hypertension and ﬂuid retention.12,13 hormone-releasing hormone agonists, B10% of baseline T As a result, the subsequent clinical development of AA in all remains due to peripheral conversion of adrenal steroids.5 In trials required concomitant administration of corticosteroids, addition, intracrine CYP17A1 activity is a source of androgen usually prednisone, to mitigate these toxicities by suppressing production leading to higher intratumoral androgen levels the adrenocorticotropic hormone feedback loop. However, compared with serum, and a potential contributor to disease despite the concomitant use of prednisone, many AA-treated progression.6–9 Therefore, CYP17A1 inhibition emerged as a patients in the phase III trials experienced adverse events related therapeutic target in CRPC. to mineralocorticoid excess.10,11 Now that AA has received broad Abiraterone acetate (AA; Zytiga, Janssen Biotech, Horsham, PA, regulatory approval and is widely utilized in the management USA) is a highly selective and potent inhibitor of CYP17A13 that of CRPC, it is relevant to annotate the potential physiologic effects has demonstrated profound clinical impact in pivotal phase III of mineralocorticoids on the AR. The AR, a member of the nuclear

1Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, San Francisco, CA, USA; 2Department of Molecular Pharmacology, City of Hope, Duarte, CA, USA; 3Department of Neurology, Washington University in St. Louis School of Medicine, St. Louis, MO, USA; 4Atara Biotherapeutics, Westlake Village, CA, USA and 5Janssen Research and Development, Los Angeles, CA, USA. Correspondence: Dr W Kim, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, 1600 Divisadero Street, San Francisco, CA 94115, USA. E-mail: [email protected] Received 23 March 2014; revised 27 May 2014; accepted 28 May 2014; published online 22 July 2014 AR inhibition by mineralocorticoids W Kim et al 293

Figure 1. The androgen biosynthesis pathway and the effect of CYP17 inhibition on mineralocorticoid and androgen levels. Abiraterone acetate (AA) inhibits the enzyme responsible for the conversion of pregnenolone to dehydroepiandrosterone, which results in increased production of hormones with mineralocorticoid activity and decreased production of androgens.

Figure 2. Fluorescence resonance energy transfer (FRET) with androgens: corticosterone (CS) and deoxycorticosterone (DOC) inhibit androgen receptor (AR) conformational change in the presence of androgens. HEK293 cells stably expressing the CFP-AR-YFP FRET reporter were treated with the indicated compounds. AR folding was quantiﬁed as the FRET:donor ratio in cells treated in quadruplicate. Cells were treated with 0.9 nM dihydrotestosterone (DHT) or 2.6 nM testosterone (T) alone, and with the combination of DHT and T for 24 h. The DHT/T combination caused AR folding with greater potency. Cells treated with the combination of DHT and T for 24 h were then treated with CS and DOC at various concentrations for 24 h. CS signiﬁcantly inhibited AR folding in DHT/T-treated cells at concentrations present in the serum of AA-treated patients. DOC demonstrated inhibition of AR folding, but at concentrations higher than present in the serum of AA-treated patients. AA, abiraterone acetate.

& 2014 Macmillan Publishers Limited Prostate Cancer and Prostatic Disease (2014), 292 – 299 AR inhibition by mineralocorticoids W Kim et al 294

Figure 3. Fluorescence resonance energy transfer (FRET) without androgens: corticosterone (CS) and deoxycorticosterone (DOC) demonstrate no signiﬁcant androgen receptor (AR) agonism in the absence of androgens. Neither CS nor DOC induced AR folding in the absence of androgens, indicating no signiﬁcant AR agonist activity.

hormone receptor superfamily, consists of a large group of ligand- in tissues derived from CRPC patients.18–20 The cells were also treated with related transcription factors that include the progesterone, a range of concentrations of CS and DOC, including those comparable to glucocorticoid and mineralocorticoid receptors.14 Promiscuous serum levels in AA-treated patients. After 24 h of exposure, cells were fixed binding to the AR by the ligands of these related receptors in 4% paraformaldehyde and read in PBS on a monochrometer-based has been demonstrated, albeit with lower affinity than that fluorescence plate reader (Safire, Tecan, San Jose, CA, USA). Each plate 15 contained untransfected, positive and negative controls. FRET:donor ratios of androgens. Whether mineralocorticoids such as CS and (representing the degree of AR folding) were calculated following DOC interact significantly with the AR, and whether potential background subtraction and correction for acceptor (YFP) contribution to downstream biological effects have any bearing on the clinical the FRET signal. efficacy, safety or resistance of AA, is unknown. The aim of this in vitro study is to evaluate the biological effects of mineral- Cell cultures ocorticoids on the AR, at concentrations present in the serum of AA-treated patients. Cell lines selected for in vitro studies were as follows: LAPC4, which expresses wild-type AR;21 LNCaP, which expresses an AR mutant with reduced specificity for DHT;22 and LN-AR, an LNCaP-derived cell line that overexpresses AR and recapitulates some features of CRPC.23 LAPC4 cells MATERIALS AND METHODS were maintained in phenol-red free Roswell Park Memorial Institute 1640 Patients and treatment media supplemented with antibiotics and 10% fetal bovine serum. LNCaP Serum mineralocorticoid levels were obtained from patients treated with and LN-AR cells were maintained in Roswell Park Memorial Institute 1640 AA for a minimum of 28 days on a phase I trial conducted at the University media supplemented with sodium bicarbonate, glutamine, HEPES, anti- of California, San Francisco, and the Dana-Farber Cancer Institute. Study biotics and 10% fetal bovine serum. enrollment occurred between July 2006 and December 2007. The primary objective of this dose-escalation trial was to determine the maximum Transcription assays tolerated dose of AA in CRPC patients. AA was administered in escalating dose cohorts of 250, 500, 750 and 1000 mg. Prednisone was not The ability of CS and DOC to inhibit endogenous AR activity was assessed administered concurrently. Secondary objectives included endocrine and in all cell lines. Cells were transfected with an androgen-responsive firefly pharmacokinetic evaluations. The findings from this study were reported luciferase reporter plasmid to measure AR-dependent transcription. Cells previously.12 Patient serum was not used in the cell culture experiments. were transfected using Lipofectamine Plus (Invitrogen, Life Technologies, Grand Island, NY, USA) with pRL-SV40 (Promega, Madison, WI, USA) and PSA-luciferase as previously described.24 Cells were replated 24 h later in Fluorescence resonance energy transfer (FRET) charcoal-stripped media. Cells were treated with a combination of 0.9 nM 16 FRET assays were performed as previously published. FRET accurately DHT and 2.6 nM T, and with a range of concentrations of CS and DOC, reflects AR activation status in the presence of different ligands, and including those comparable to serum levels in AA-treated patients. Cells has been used to identify novel antiandrogens.16,17 Briefly, HEK293 cells were also treated with bicalutamide at various concentrations for com- stably expressing the CFP-AR-YFP reporter were transferred to charcoal- parison of anti-AR activity. After 24 h, luciferase production was measured stripped, black, clear-bottomed 96-well plates. The cells were treated with (Dual Luciferase Assay Kit, Promega). Mean-effect plots (log[compound] 0.9 nM dihydrotestosterone (DHT) and 2.6 nM T, reflecting androgen levels versus log[fractional effect]) were generated to determine the IC50 values

Prostate Cancer and Prostatic Disease (2014), 292 – 299 & 2014 Macmillan Publishers Limited AR inhibition by mineralocorticoids W Kim et al 295

Figure 4. Androgen receptor (AR) transcriptional activity with androgens: inhibition of AR transcriptional activity by bicalutamide (BiC), corticosterone (CS) and deoxycorticosterone (DOC) in the presence of androgens. Cell lines were transfected with an androgen-responsive luciferase reporter. Luciferase production, depicted relative to the amount produced when stimulated with the combination of dihydrotestosterone (DHT)/T for each cell line, represents AR-dependent transcriptional activity. There was negligible luciferase production in the absence of androgen, and the combination of 0.9 nM DHT and 2.6 nM T stimulated luciferase production in all three cell lines after 24 h of exposure. Cells that were stimulated with the combination of androgens were then treated with physiologically relevant concentrations of CS and DOC for 24 h; BiC was used for comparison. CS inhibited luciferase production in all three cell lines at concentrations present in the serum of AA-treated patients. DOC inhibited luciferase production at concentrations 10-fold (in LAPC4 cells) and 100-fold (in LNCaP and LN-AR cells) higher than present in the serum of AA-treated patients. The IC50 of CS was comparable to that of BiC in all three cell lines. of each compound or combinations of compounds at constant ratios. upstream steroids.12 CS was increased 32- to 85-fold and Microsoft Excel was used to calculate the statistics for a line using the ‘least DOC was increased 11- to 25-fold over baseline serum levels squares’ method. The F statistic was used to determine whether the after 28 days of AA (maximum concentrations: CS 3.35 mM; DOC observed relationship between the dependent and independent variables 0.03 mM). occurred by chance. Only data with an r2 value greater than 0.95 and an F value that was greater than that indicated by the F table for a ¼ 0.05 were used for analysis. Mineralocorticoids inhibit androgen-induced AR conformational change Cell proliferation assays As heterologous hormones may inhibit AR transcription by For growth curves, cells were transferred to charcoal-stripped media 3 days alternative mechanisms (for example direct competition, activa- before they were split and plated at a density of B20 000 cells per well in tion of other NRs), FRET was utilized to assess the molecular 48-well plates, in quadruplicate. Twenty-four hours later, media with or conformational change of the AR induced by ligand binding. without androgens or other compounds were added. Media were changed The folding of AR is a discrete and critical step in AR signaling, daily, and cells were grown for 7 days in the presence or absence of dependent on ligand binding upstream of transcriptional the 0.9 nM DHT/2.6 nM T combination, and in the presence of CS, DOC or bicalutamide. Proliferation was determined by measuring the DNA content activation. HEK293 cells stably expressing the AR-FRET reporter of the cells in each well. Every 48 h, cells were fixed in 100% cold methanol, were in order to control for alternative mechanisms that may followed by staining for 5 min at RT with 0.2 ng ml À 1 406-diamidino-2- contribute to AR activity. As expected, DHT and T alone each phenylindole in PBS. Cells were washed with PBS and read on a caused AR folding, with greater potency observed with the DHT/T fluorescence plate reader using 365/439 excitation/emission wavelengths. combination (Figure 2). CS inhibited androgen-induced AR folding at concentrations present in the serum of AA-treated patients. DOC only significantly inhibited AR folding at concentrations RESULTS greater than those measured in the serum of AA-treated patients. Serum mineralocorticoid levels Neither CS nor DOC induced AR folding in the absence of As previously reported, AA inhibited androgen synthesis androgens, demonstrating that they lack AR agonist properties downstream of CYP17A1 and increased circulating levels of (Figure 3).

& 2014 Macmillan Publishers Limited Prostate Cancer and Prostatic Disease (2014), 292 – 299 AR inhibition by mineralocorticoids W Kim et al 296

Figure 5. Androgen receptor (AR) transcriptional activity without androgens: the effect of corticosterone (CS) and deoxycorticosterone (DOC) on AR transcriptional activity in the absence of androgens. In LAPC4 cells, CS and DOC did not induce signiﬁcant luciferase production, indicating no signiﬁcant agonist activity. In LNCaP cells, CS and DOC induces modest AR transcriptional activity, indicating AR agonism in the absence of androgens. BiC, bicalutamide.

Mineralocorticoids inhibit AR-dependent transcriptional activity in LAPC4 and LNCaP cells demonstrated limited growth in the PC cells absence of androgens but grew well with the DHT/T combination, In the absence of androgens, there was no significantly measurable as expected (Figure 6). The growth of LN-AR cells was not affected AR-dependent transcriptional activity in the three PC cell lines. AR by androgen exposure. At a concentration of 5 mM, CS inhibited the activity was stimulated as expected by the DHT/T combination. androgen-mediated growth of both LAPC4 and LNCaP cells by When cells were co-incubated with mineralocorticoids and the B50%, which was approximately equivalent to the effect observed DHT/T combination, both CS and DOC inhibited androgen- with the same concentration of bicalutamide. At the same stimulated AR transcriptional activity in all cell lines (Figure 4). CS concentration, CS also inhibited the growth of LN-AR cells, while had an IC50 of B3 mM in LAPC4 and LNCaP cells, and 4.3 mM in LN-AR bicalutamide demonstrated no growth inhibitory effect, as pre- cells, demonstrating that CS can inhibit androgen-stimulated AR viously reported.23 DOC did not cause statistically significant growth transcriptional activity at physiologically relevant concentrations, inhibition in any of the cell lines at concentrations up to 0.03 mM. even in the setting of mutant or amplified AR. DOC had an IC50 of In the absence of androgens, CS and DOC did not induce B0.2 mM in LAPC4 cells, and 8 mM in the LNCaP and LN-AR cells. cellular proliferation in LAPC4 cells (Figure 7), consistent with the Thus, the IC50 of DOC represents concentrations B10-fold (LAPC4) FRET and transcription results that these mineralocorticoids lack and 100-fold (LNCaP and LN-AR) higher than found in the serum of agonist activity in the setting of wild-type AR. In LNCaP cells, CS AA-treated patients. In comparison, the IC50 of the competitive AR and DOC induced modest cellular proliferation in the absence of antagonist bicalutamide was B1 mM in LAPC4 and LNCaP cells, and androgens, but only at supraphysiologic concentrations. 2.5 mM in LN-AR cells treated with DHT/T. In LAPC4 cells, CS and DOC did not elicit significant agonist activity. In LNCaP cells, both CS and DOC demonstrated modest DISCUSSION agonist activity in the absence of androgens (Figure 5). This These data show that mineralocorticoids inhibit androgen- finding is consistent with prior findings that CS and DOC can 25 induced AR conformational change, AR transcriptional activity induce mutant AR activity. and proliferation of PC cells. Specifically CS, a mineralocorticoid The effect of CS and DOC on endogenous AR targets was increased by 32- to 85-fold in the serum of AA-treated patients, evaluated by reverse transcription-quantitative polymerase chain demonstrated such activity at physiologically relevant concentra- reaction (Supplementary Figure 1); this was consistent with the tions. In comparison, DOC demonstrated such activity at luciferase data above (methods provided in Supplement). concentrations B10- to 100-fold greater than what is measurable in AA-treated men. Mineralocorticoids limit androgen-dependent and -independent Despite the profound clinical impact of AA in the treatment of growth of PC cells CRPC, disease progression is inevitable. To better understand both The observation that CS and DOC inhibit AR transcriptional activity drug sensitivity and resistance, it is important that the interaction in PC cells raised the question of whether mineralocorticoids can between mineralocorticoids and the AR be understood at the inhibit their growth. mechanistic level. Recently published investigations that implicate

Prostate Cancer and Prostatic Disease (2014), 292 – 299 & 2014 Macmillan Publishers Limited AR inhibition by mineralocorticoids W Kim et al 297

Figure 6. Cell growth with androgens: inhibition of cell growth by bicalutamide (BiC), corticosterone (CS) and deoxycorticosterone (DOC). Prostate cancer cells were treated as indicated for 7 days. Cellular proliferation was assessed using diamidino-2-phenylindole (DAPI) to measure the DNA content (in quadruplicate wells) after 7 days of treatment. Proliferation is shown relative to cells grown in the absence of androgens for 7 days (cellular proliferation ¼ 1). LAPC4 and LNCaP cells grew well in the presence of androgens (5- and 3-fold greater growth with androgens than without androgens, respectively). The growth of LN-AR cells was not affected by the presence of androgens. In LAPC4 and LNCaP cells, the growth inhibitory effect of 5 mM concentrations of CS and BiC was comparable. In LN-AR cells, CS significantly inhibited the growth of LN-AR cells, whereas BiC showed no effect. DOC did not significantly inhibit cell growth at concentrations up to 0.03 mM. DHT, dihydrotestosterone. promiscuous interactions between the ligands and receptors of prednisone daily must be considered. Lower doses of prednisone the nuclear hormone family as a potential cause of therapeutic may be a reasonable approach: in neoadjuvant studies where resistance, including glucocorticoid-mediated activation of mutant patients with high-risk PC were given AA for 3–6 months prior to AR 26 and activation of the glucocorticoid receptor in the setting radical prostatectomy, prednisone was administered at a dose of of AR blockade 27, support this approach. The findings here 5 mg daily.29,30 Only 4 of 83 patients (4.8%) experienced grade 3 that CS and DOC inhibit AR transcriptional activity in the presence adverse events related to mineralocorticoid excess, suggesting of androgens by blocking AR conformational change are pro- that the lower dose of prednisone is safe. The findings presented vocative, as mineralocorticoids have not previously been shown to here suggest that a lower dose of prednisone (and a resultant act as AR antagonists (in contrast, spironolactone and epleronone increase in mineralocorticoid levels) may impact AR signaling and are mineralocorticoid receptor antagonists that have been shown could potentially impact the clinical efficacy of AA. to have the capability of acting as AR agonists).26,28 That CS On the other hand, in the phase III AA studies in which demonstrated this antagonist effect at physiologic concentrations prednisone was administered at a dose of 5 mg twice daily, a measured in AA-treated patients suggests that this finding is significant proportion of patients still experienced symptoms clinically important. The results presented here suggest that associated with mineralocorticoid excess: 28–31% experienced mineralocorticoid excess may result in an off-target effect of AA, fluid retention, 17% experienced hypokalemia and 10–22% possibly contributing modestly to its clinical efficacy. The use of experienced hypertension.10,11 Thus, another question raised by cell lines representing wild-type, mutant and amplified AR this study that deserves further investigation is whether patients enhances these observations. who experience these symptoms have a different, and possibly However, to put these results in the proper clinical context, the improved, therapeutic response to AA. This would support the AR antagonist effects observed are modest compared with the development of novel approaches to inhibit the physiologic potentially adverse physiologic consequences of prolonged response to mineralocorticoid excess that spare the AR antagonist mineralocorticoid excess (that is hypertension and hypokalemia), effects of mineralocorticoids. and, on balance, these data do not support the use of AA Mineralocorticoids have previously been shown to exhibit AR without prednisone. Therefore, direct translational relevance agonism in the setting of mutant AR 25,31, and the data presented of these results to clinical practice is limited. However, in real here are consistent with those findings. Although CS and DOC world application, compliance with, and the side effects of, 10 mg demonstrated no significant agonist activity in LAPC4 cells, they

& 2014 Macmillan Publishers Limited Prostate Cancer and Prostatic Disease (2014), 292 – 299 AR inhibition by mineralocorticoids W Kim et al 298

Figure 7. Cell growth without androgens: effect of mineralocorticoids on cellular proliferation in the absence of androgens. Corticosterone (CS) and deoxycorticosterone (DOC) did not induce signiﬁcant cellular proliferation of LAPC4 cells. In LNCaP cells, CS and DOC demonstrated modestly increased cellular proliferation. BiC, bicalutamide.

induced modest transcriptional activity as well as cellular pro- with the AR; this observation may have important therapeutic liferation in the absence of androgens in LNCaP cells. Whereas implications. Enzalutamide and ARN-509 are potent next-generation alterations in the AR pathway are among the most common AR antagonists that are in clinical trials; enzalutamide has demon- events in CRPC, AR mutations remain a rare event, occurring in strated improved survival in CRPC in phase III trials.37,38 Although 8–20% of even heavily pretreated cases.32–34 Therefore, for the it is unlikely that mineralocorticoids themselves can be used with majority of CRPC patients, including those in whom the current therapeutic intent because of their adverse physiologic effects, it is utilization of abiraterone is most common (first-line, chemo- possible that drugs derived from mineralocorticoids may exert anti- therapy-naive), the physiology of mineralocorticoids on wild-type AR and anti-tumor properties and be developed as targeted AR is likely most relevant. However, with the emergence of therapies. In this study, the anti-AR effect of CS in the presence of AR mutations as a potential mechanism of treatment-related androgens is comparable to that of bicalutamide, a commercially resistance to novel AR-targeted therapies in CRPC,35 the available anti-androgen commonly used in the treatment of androgenic activity of mineralocorticoids in the mutant AR hormone-sensitive PC. Furthermore, CS demonstrated significant setting merits further investigation. Furthermore, if the potent growth inhibitory effect in all PC cell lines including LN-AR, whereas CYP17A1 inhibitory activity of abiraterone can significantly reduce bicalutamide showed no growth inhibitory effect in the setting of intratumoral androgen levels below what has been described amplified AR. In this era, when novel and highly effective drugs that previously,18–20 the agonist activity of CS and DOC in the absence target the AR and androgen production are being used to treat of androgens may prove relevant in treatment-related resistance CRPC, the issue of ligand-receptor promiscuity should be further to abiraterone. explored as a potential therapeutic opportunity. A limitation of this study is that, although mineralocorticoid concentrations were measured from clinical samples of AA-treated patients, the levels of these mineralocorticoids within tumor tissue CONFLICT OF INTEREST are unknown. Although AA monotherapy raises serum mineralo- AM is an employee of Johnson & Johnson and holds stock ownership. CH is a former corticoid concentrations,12,13 whether the same effect occurs within employee of Johnson & Johnson and holds stock ownership. CJR has received CRPC tissues is unclear. In addition, various studies have demon- honoraria from Janssen. JM is a member of the NCCN’s Prostate Cancer Early strated that androgen levels within PC tissue remain higher than Detection Panel and, on behalf of the NCCN, has received honoraria only for lectures serum concentrations following castration.18,36 Whether higher (usually annual meeting for guidelines update and occasional CME activity) or other activities (panel participation for guidelines adaptation for other countries); has levels of androgens affect mineralocorticoid concentrations within received an honorarium from the Genomic Health Scientific Advisory Board (related CRPC tissues, and with or without AA, is unknown. activity: presentation (no honorarium), collaboration with the North Carolina- The potential for endogenous mineralocorticoids to affect the Louisiana Prostate Cancer Project (no funded effort or honorarium)); is a cofounder growth of PC cells may be directly related to physical interaction and chief medical officer of AndroBioSys, with ownership interest (value: none); is

Prostate Cancer and Prostatic Disease (2014), 292 – 299 & 2014 Macmillan Publishers Limited AR inhibition by mineralocorticoids W Kim et al 299 associated with Medivation MTA for MDV3100 (no research support or honoraria); is a 18 Nishiyama T, Hashimoto Y, Takahashi K. The influence of androgen deprivation consultant to and collaborator on research at New York University School of therapy on dihydrotestosterone levels in the prostatic tissue of patients with Medicine; and was an initial investor (value: none) in and an advisory board member prostate cancer. Clin Cancer Res 2004; 10: 7121–7126. for Simulated Surgical Systems. The remaining authors declare no conflict of interest. 19 Mohler JL, Gregory CW, Ford OH 3rd, Kim D, Weaver CM, Petrusz P et al. The androgen axis in recurrent prostate cancer. Clin Cancer Res 2004; 10: 440–448. 20 Montgomery RB, Mostaghel EA, Vessella R, Hess DL, Kalhorn TF, Higano CS et al. Maintenance of intratumoral androgens in metastatic prostate cancer: a ACKNOWLEDGEMENTS mechanism for castration-resistant tumor growth. Cancer Res 2008; 68: 4447–4454. The authors thank Dr Charles Sawyers for his generous gift of the LAPC4 and LN-AR 21 Klein KA, Reiter RE, Redula J, Moradi H, Zhu XL, Brothman AR et al. Progression of cells, and Dr Ingo Mellinhoff for his generous gift of bicalutamide. JOJ was supported metastatic human prostate cancer to androgen independence in immunodefi- by NIH F32 CA123750. MD was supported by NCI CA131226. CJR was supported by cient SCID mice. Nat Med 1997; 3: 402–408. K23 CA115775. 22 Horoszewicz JS, Leong SS, Chu TM, Wajsman ZL, Friedman M, Papsidero L et al. The LNCaP cell line--a new model for studies on human prostatic carcinoma. Prog Clin Biol Res 1980; 37: 115–132. 23 Chen CD, Welsbie DS, Tran C, Baek SH, Chen R, Vessella R et al. Molecular REFERENCES determinants of resistance to antiandrogen therapy. Nat Med 2004; 10: 33–39. 1 Ryan CJ, Tindall DJ. Androgen receptor rediscovered: the new biology and tar- 24 Jones JO, Bolton EC, Huang Y, Feau C, Guy RK, Yamamoto KR et al. Non-com- geting the androgen receptor therapeutically. J Clin Oncol 2011; 29: 3651–3658. petitive androgen receptor inhibition in vitro and in vivo. Proc Natl Acad Sci USA 2 De Coster R, Wouters W, Bruynseels J. P450-dependent enzymes as targets for pro- 2009; 106: 7233–7238. state cancer therapy. J Steroid Biochem Mol Biol 1996; 56(1–6 Spec No): 133–143. 25 Chang CY, Walther PJ, McDonnell DP. Glucocorticoids manifest androgenic 3 Haidar S, Ehmer PB, Barassin S, Batzl-Hartmann C, Hartmann RW. Effects of novel activity in a cell line derived from a metastatic prostate cancer. Cancer Res 2001; 17alpha-hydroxylase/C17, 20-lyase (P450 17, CYP 17) inhibitors on androgen 61: 8712–8717. biosynthesis in vitro and in vivo. JSteroid Biochem Mol Biol 2003; 84: 555–562. 26 Richards J, Lim AC, Hay CW, Taylor AE, Wingate A, Nowakowska K et al. Interac- 4 Potter GA, Barrie SE, Jarman M, Rowlands MG. Novel steroidal inhibitors of human tions of abiraterone, eplerenone, and prednisolone with wild-type and mutant cytochrome P45017 alpha (17 alpha-hydroxylase-C17,20-lyase): potential agents androgen receptor: a rationale for increasing abiraterone exposure or combining for the treatment of prostatic cancer. J Med Chem 1995; 38: 2463–2471. with MDV3100. Cancer Res 2012; 72: 2176–2182. 5 Nishiyama T. Serum testosterone levels after medical or surgical androgen 27 Arora VK, Schenkein E, Murali R, Subudhi SK, Wongvipat J, Balbas MD et al. deprivation: a comprehensive review of the literature. Urol Oncol 2014; 32:38 Glucocorticoid receptor confers resistance to antiandrogens by bypassing e17–28. androgen receptor blockade. Cell 2013; 155: 1309–1322. 6 Stanbrough M, Bubley GJ, Ross K, Golub TR, Rubin MA, Penning TM et al. 28 Luthy IA, Begin DJ, Labrie F. Androgenic activity of synthetic progestins and Increased expression of genes converting adrenal androgens to testosterone in spironolactone in androgen-sensitive mouse mammary carcinoma (Shionogi) androgen-independent prostate cancer. Cancer Res 2006; 66: 2815–2825. cells in culture. J Steroid Biochem 1988; 31: 845–852. 7 Cai C, Balk SP. Intratumoral androgen biosynthesis in prostate cancer patho- 29 Taplin ME, Montgomery RB, Logothetis C, Bubley GJ, Richie JP, Dalkin BL et al. genesis and response to therapy. Endocr Relat Cancer 2011; 18: R175–R182. Effect of neoadjuvant abiraterone acetate (AA) plus leuprolide acetate (LHRHa) on 8 Dillard PR, Lin MF, Khan SA. Androgen-independent prostate cancer cells acquire PSA, pathological complete response (pCR), and near pCR in localized high-risk the complete steroidogenic potential of synthesizing testosterone from choles- prostate cancer (LHRPC): results of a randomized phase II study. J Clin Oncol 2012; terol. Mol Cell Endocrinol 2008; 295: 115–120. 30: 4521. 9 Locke JA, Guns ES, Lubik AA, Adomat HH, Hendy SC, Wood CA et al. Androgen 30 Efstathiou E, Davis JW, Troncoso P, Titus MA, Hoang A, Wen S et al. Cytoreduction levels increase by intratumoral de novo steroidogenesis during progression of and androgen signaling modulation by abiraterone acetate (AA) plus leuprolide castration-resistant prostate cancer. Cancer Res 2008; 68: 6407–6415. acetate (LHRHa) versus LHRHa in localized high-risk prostate cancer (PCa): 10 de Bono JS, Logothetis CJ, Molina A, Fizazi K, North S, Chu L et al. Abiraterone preliminary results of a randomized preoperative study. J Clin Oncol 2012; 30: and increased survival in metastatic prostate cancer. N Engl J Med 2011; 364: 4556. 1995–2005. 31 Uemura M, Honma S, Chung S, Takata R, Furihata M, Nishimura K et al. 5alphaDH- 11 Ryan CJ, Smith MR, de Bono JS, Molina A, Logothetis CJ, de Souza P et al. DOC (5alpha-dihydro-deoxycorticosterone) activates androgen receptor in Abiraterone in metastatic prostate cancer without previous chemotherapy. N Engl castration-resistant prostate cancer. Cancer Sci 2010; 101: 1897–1904. JMed2013; 368: 138–148. 32 Grasso CS, Wu YM, Robinson DR, Cao X, Dhanasekaran SM, Khan AP et al. 12 Ryan CJ, Smith MR, Fong L, Rosenberg JE, Kantoff P, Raynaud F et al. Phase I The mutational landscape of lethal castration-resistant prostate cancer. Nature clinical trial of the CYP17 inhibitor abiraterone acetate demonstrating clinical 2012; 487: 239–243. activity in patients with castration-resistant prostate cancer who received prior 33 Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS et al. Integrative ketoconazole therapy. J Clin Oncol 2010; 28: 1481–1488. genomic profiling of human prostate cancer. Cancer Cell 2010; 18: 11–22. 13 Attard G, Reid AH, Yap TA, Raynaud F, Dowsett M, Settatree S et al. Phase I clinical 34 Beltran H, Yelensky R, Frampton GM, Park K, Downing SR, MacDonald TY et al. trial of a selective inhibitor of CYP17, abiraterone acetate, confirms that castra- Targeted next-generation sequencing of advanced prostate cancer identi- tion-resistant prostate cancer commonly remains hormone driven. J Clin Oncol fies potential therapeutic targets and disease heterogeneity. Eur Urol 2013; 63: 2008; 26: 4563–4571. 920–926. 14 Gelmann EP. Molecular biology of the androgen receptor. J Clin Oncol 2002; 20: 35 Korpal M, Korn JM, Gao X, Rakiec DP, Ruddy DA, Doshi S et al. An F876L mutation 3001–3015. in androgen receptor confers genetic and phenotypic resistance to MDV3100 15 Marhefka CA, Moore BM 2nd, Bishop TC, Kirkovsky L, Mukherjee A, Dalton JT et al. (enzalutamide). Cancer Discov 2013; 3: 1030–1043. Homology modeling using multiple molecular dynamics simulations and docking 36 Page ST, Lin DW, Mostaghel EA, Hess DL, True LD, Amory JK et al. Persistent studies of the human androgen receptor ligand binding domain bound to intraprostatic androgen concentrations after medical castration in healthy men. testosterone and nonsteroidal ligands. J Med Chem 2001; 44: 1729–1740. J Clin Endocrinol Metab 2006; 91: 3850–3856. 16 Schaufele F, Carbonell X, Guerbadot M, Borngraeber S, Chapman MS, Ma AA et al. 37 Scher HI, Fizazi K, Saad F, Taplin ME, Sternberg CN, Miller K et al. Increased survival The structural basis of androgen receptor activation: intramolecular and inter- with enzalutamide in prostate cancer after chemotherapy. N Engl J Med 2012; molecular amino-carboxy interactions. Proc Natl Acad Sci USA 2005; 102: 9802–9807. 367: 1187–1197. 17 Jones JO, An WF, Diamond MI. AR inhibitors identified by high-throughput 38 Beer TM, Armstrong AJ, Sternberg CN, Higano C, Iverson P, Loriot Y et al. microscopy detection of conformational change and subcellular localization. ACS Enzalutamide in men with chemotherapy-naive metastatic prostate cancer (mCRPC): Chem Biol 2009; 4: 199–208. results of phase III PREVAIL study. JClinOncol2014; 32:LBA1.

Supplementary Information accompanies the paper on the Prostate Cancer and Prostatic Diseases website (http://www.nature.com/pcan)

& 2014 Macmillan Publishers Limited Prostate Cancer and Prostatic Disease (2014), 292 – 299