Interactions of abiraterone, eplerenone, and prednisolone with wild-type and mutant androgen receptor: a rationale for increasing abiraterone exposure or combining with MDV3100

Abstract

Prostate cancer progression can be associated with androgen receptor (AR) mutations acquired following treatment with castration and/or an antiandrogen. Abiraterone, a rationally designed inhibitor of CYP17A1 recently approved for the treatment of docetaxel-treated castration-resistant prostate cancer (CRPC), is often effective, but requires coadministration with glucocorticoids to curtail side effects. Here, we hypothesized that progressive disease on abiraterone may occur secondary to glucocorticoid-induced activation of mutated AR. We found that prednisolone plasma levels in patients with CRPC were sufficiently high to activate mutant AR. Mineralocorticoid receptor antagonists, such as spironolactone and eplerenone that are used to treat side effects related to mineralocorticoid excess, can also bind to and activate signaling through wild-type or mutant AR. Abiraterone inhibited in vitro proliferation and AR-regulated gene expression of AR-positive prostate cancer cells, which could be explained by AR antagonism in addition to inhibition of steroidogenesis. In fact, activation of mutant AR by eplerenone was inhibited by MDV3100, bicalutamide, or greater concentrations of abiraterone. Therefore, an increase in abiraterone exposure could reverse resistance secondary to activation of AR by residual ligands or coadministered drugs. Together, our findings provide a strong rationale for clinical evaluation of combined CYP17A1 inhibition and AR antagonism.

Figure 1. Eplerenone activates T877A-AR and spironolactone activates both T877A-AR and wild-type (WT)-AR

Sigmoidal dose response curves show activation of (A) WT-AR by R1881 and spironolactone and (B) T877A-AR by R1881, spironolactone, eplerenone, prednisolone, dexamethasone. Fold change from the DMSO control was plotted and EC₅₀s calculated using non-linear regression (GraphPad). EC₅₀ and 95% confidence intervals are given. C, LNCaP and VCaP prostate cancer cells in CSS were treated with eplerenone or spironolactone alone or in combination with 0.1, 1 or 5μM abiraterone, 10μM bicalutamide or 10μM MDV3100 or for 7 days and then analyzed for cell viability. Fold change from the DMSO control was then calculated and plotted. Significance is shown for stimulation by eplerenone or spironolactone compared to DMSO control and for inhibition by bicalutamide, MDV3100 or abiraterone when compared to stimulated levels. D, LNCaP and VCaP cells were treated with 0.1nM R1881 or 0.1 - 10μM eplerenone for 5 hours. RNA was extracted and cDNA synthesized for analysis by quantitative PCR to determine relative levels of PSA and TMPRSS2 mRNA expression. Significance compared to DMSO controls is shown. Data shown for all experiments are the mean (error bars, standard error of the mean, SEM) of 3 independent experiments of 16 replicates (A, B) or in duplicate (C, D). *, P<0.05; **, P<0.01; ***, P< 0.001, one-way ANOVA with Bonferroni correction.

Figure 2. Plasma concentrations (nmol/L) of prednisolone in 15 CRPC patients treated with abiraterone acetate measured using liquid chromatography/tandem mass spectrometry (LC/MS/MS)

The median concentration of 152nM (SD 100nM) is marked by the solid line. The 10nM limit above which activation of T877A-L701H-AR has been previously reported to occur is shown by the dashed line.

Figure 3. Inhibition of wild-type and mutant stimulated AR activity by abiraterone, bicalutamide and MDV3100

A, PC-3 cells were co-transfected with ARE3-luciferase and WT or mutant AR (T877A, D879G, W741C, M749L, R629Q). Cells were treated with 0.1 - 25μM abiraterone, 10μM bicalutamide or 10μM MDV3100 in CSS medium containing 0.1nM R1881 for 16 hours and then analyzed for luciferase activity. Fold change from the DMSO control was calculated and then percentage change relative to the R1881-stimulated DMSO control was determined. Data shown are representative of 3 independent experiments and represent mean and SEM of 8 replicates. B, COS-7 cells were co-transfected with GRE₂-TATA-Luc and the WT or mutant human expression plasmid pSVARo (T877A, G142V, P533S, T575A, H874Y, R629Q). Cells were treated with 0.1 - 5μM abiraterone or MDV3100 in CSS medium containing 10nM DHT for 24 hours. The luciferase activities were assayed in duplicate and normalized for the amounts of expressed AR determined immunologically by dot blot analysis and normalized for protein concentration. The change in normalized luciferase activity relative to cells incubated without any compound for each AR variant was determined. Data shown represent two or three independent experiments performed in quadruplicate. C, Dose-proportional inhibition of proliferation of LNCaP and VCaP cells by abiraterone, MDV3100 and bicalutamide. LNCaP and VCaP prostate cancer cells in CSS with 0.1nM R1881 were treated with 0.1, 1 or 5μM abiraterone, 10μM bicalutamide or 10μM MDV3100 for 7 days and then analyzed for cell viability. Fold change from the DMSO control was then calculated and plotted. Data shown are the mean (error bars, standard error of the mean, SEM) of 3 independent experiments in quadruplicate. D, LNCaP cells were treated with 0.1nM R1881 or 10μM eplerenone in combination with DMSO, 10μM bicalutamide, 10μM MDV3100 or 5μM abiraterone for 5 hours. RNA was extracted and cDNA synthesized for analysis by quantitative PCR to determine relative levels of PSA and TMPRSS2 mRNA expression. Data shown are the mean and SEM of 3 independent experiments in duplicate. Significance is shown for *, P<0.05; **, P<0.01; ***, P< 0.001; ****, P<0.0001 relative to DMSO control (one-way ANOVA with Bonferroni correction).

Figure 4. Displacement of [³H] R1881 by eplerenone and abiraterone in PC-3 cells transfected with WT or T877A mutant AR

PC-3 cells were transfected with WT AR (A) or T877A mutant AR (B) and then treated with CSS media containing 5nM of [³H] R1881 in combination with cold R1881, DHT, eplerenone, abiraterone or bicalutamide at the concentrations shown for 2 hours. Abiraterone was insoluble in cell media at concentrations greater than 25μM. Cell-associated radioactivity was measured and the data analyzed by nonlinear regression to determine the EC₅₀ for each test compound (GraphPad Prism). Data shown are the mean and SEM of three independent experiments in triplicate for percentage (%) [3H]-R1881 bound versus log10 of concentration (μM) of cold competitor. EC₅₀ and 95% confidence intervals are given. C, Inhibition of eplerenone-stimulated AR activation by bicalutamide, MDV3100 and abiraterone. PC-3 cells were co-transfected with ARE3-luciferase and T877A mutant AR. Cells were treated with DMSO (control) or eplerenone in combination with DMSO, 10μM bicalutamide, 10μM MDV3100 or 5μM abiraterone for 16 hours and then analyzed for luciferase activity. Fold change from the DMSO control was calculated. Data shown are from three independent experiments and represent mean and SEM of 13 replicates. D, Increased hormone levels reduce AR inhibition by MDV3100. PC-3 cells were co-transfected with ARE3-luciferase and WT-AR. Cells were treated with R1881 in combination with DMSO or MDV3100 at the concentrations indicated for 16 hours and then analyzed for luciferase activity. Fold change from the DMSO control was calculated. Data shown are from three independent experiments and represent mean and SEM of 24 replicates. ***, P< 0.01 relative to R1881 or DHT control with DMSO (one-way ANOVA with Bonferroni correction).