Comprehensive genomics in androgen receptor-dependent castration-resistant prostate cancer identifies an adaptation pathway mediated by opioid receptor kappa 1

Abstract

Castration resistance is a lethal form of treatment failure of prostate cancer (PCa) and is associated with ligand-independent activation of the androgen receptor (AR). It is only partially understood how the AR mediates survival and castration-resistant growth of PCa upon androgen deprivation. We investigated integrative genomics using a patient-derived xenograft model recapitulating acquired, AR-dependent castration-resistant PCa (CRPC). Sequencing of chromatin immunoprecipitation using an anti-AR antibody (AR-ChIP seq) revealed distinct profiles of AR binding site (ARBS) in androgen-dependent and castration-resistant xenograft tumors compared with those previously reported based on human PCa cells or tumor tissues. An integrative genetic analysis identified several AR-target genes associated with CRPC progression including OPRK1, which harbors ARBS and was upregulated upon androgen deprivation. Loss of function of OPRK1 retarded the acquisition of castration resistance and inhibited castration-resistant growth of PCa both in vitro and in vivo. Immunohistochemical analysis showed that expression of OPRK1, a G protein-coupled receptor, was upregulated in human prostate cancer tissues after preoperative androgen derivation or CRPC progression. These data suggest that OPRK1 is involved in post-castration survival and cellular adaptation process toward castration-resistant progression of PCa, accelerating the clinical implementation of ORPK1-targeting therapy in the management of this lethal disease.

Fig. 1. KUCaP2 is a PDX line as a model for androgen receptor-dependent castration-resistant prostate cancer.

a, b Individual growth curves of KUCaP2 tumors in intact (a) and castrated (b) mice. Arrow indicates surgical castration. c Serum prostate-specific antigen (PSA) levels in androgen-dependent (AD) and castration-resistant (CR) tumor-bearing mice (n = 4 each). d Western blotting of indicated proteins in AD and CR KUCaP2 tumors. Two antibodies for androgen receptor (AR) recognizing distinct epitopes on N-terminus (N20) and C-terminus (C19) were used. LNCaP, PC3, and 22RV1 prostate cancer cell lines act as controls. ACTB; β-actin. e, f Serum testosterone (e) and dihydrotestosterone (DHT, f) levels in mice bearing AD and CR KUCaP2 tumors. g, h Growth curves of AD (g) and CR (h) KUCaP2 tumors treated with control (siNTC) or siRNAs for AR (siAR#1 and siAR#2). **P < 0.01 (ANOVA). i, j Western blotting of indicated proteins in AD (i) and CR (j) KUCaP2 tumors treated with control or siRNAs for AR. k, l Representative microphotographs of hematoxylin and eosin (H&E) stain and immunohistochemical stains for AR and PSA in AD (k) and CR (l) KUCaP2 tumors treated with control or siRNAs for AR. Bars indicate 50 μm.

Fig. 2. Chromatin-immunoprecipitation sequence (ChIP seq) using antibody against androgen receptor (AR).

a Top: Experimental schemes. KUCaP tumors under androgen-dependent (AD) and castration-resistant (CR) growth were obtained. Bottom: LNCaP cells cultured in FBS, charcoal-strip FBS (CSFBS) and CSFBS supplemented with 1 nM dihydrotestosterone (CSFBS + DHT) as well as AILNCaP cultured in CSFBS were subject to CHiP seq. b Quality controls for ChIP seq binding data. Bar files were generated after MAT analysis of AR whole-genome ChIP seq raw data from LNCaP cells cultured in FBS, charcoal-strip FBS (CSFBS), and CSFBS supplemented with 1 nM dihydrotestosterone (DHT). AR-binding peaks at the promoter regions (unless otherwise indicated) of indicated genes are shown. KLK3 p; KLK3 (PSA) promoter, KLK3 e; KLK3 (PSA) enhancer, FKBP5 3', 3' UTR region of FKBP5. c Venn diagrams showing AR-binding sites in KUCaP2 AD and CR tumors (top), and LNCaP and AILNCaP cells (bottom). We identified 3131 differential AR-binding sites for KUCaP AD tumors, 1850 for KUCaP2 CR tumors, 2,938 for LNCaP cells, and 717 for AILNCaP cells, which were defined as having fold change ≤ 0.5 or ≥2 compared with each counterpart. There were 6102 AR-binding sites commonly identified in KUCaP2 AD and CR tumors and 1751 in LNCaP and AILNCaP cells. d Reactome pathways for genes exclusively identified for KUCaP2 CR tumors annotated by AR-binding sites in AR-ChIP seq with regard to entities p-values (−log[p-value]). e, f Venn diagrams depicting differentially and commonly identified genes harboring AR-binding site in AD (e) and CR (f) models including PDX (the present study), human PCa tissue, and cell lines.

Fig. 3. RNA sequence KUCaP2 AD and CR tumors.

Among 30 most significantly upregulated genes in RNA seq of KUCaP2 CR compared with AD tumors, seven genes that were reported to be frequently amplified in PCa were picked up. Likewise, among 30 most significantly downregulated genes in KUCaP2 CR tumors, five genes that were reported to be frequently altered (deep deletion, and truncating and missense mutations) in CRPC were picked up. a Dot plot of log₁₀[fkpm] indicating the 12 genes. b Landscape of alterations (amplification, deep deletion, and truncating and missense mutations) of the 12 genes primary PCa. c Summarized results from integrated genomics including RNA seq of KUCaP2 tumors (AD vs CR), cDNA microarray of KUCaP2 tumors (AD vs CR), ChIP seq of KUCaP2 tumors (AD and CR), and ChIP seq of AILNCaPs (four independent sublines) and LNCaP cells cultured in FBS (LNCaP_FBS), charcoal-stripped FBS (LNCaP_CSFBS), and CSFBS supplemented with 1 nM dihydrotestosterone (LNCaP_CSFBS + DHT). d–j Quantitative RT-PCR in human PCa cell lines. Expressions of TRPA1 (d), CP (e), CLSTN2 (f), MGLL (g), OPRK1 (h), NMNAT2 (i), and ROBO1 (j) normalized by that of GAPDH in the indicated PCa cell lines.

Fig. 4. AR suppression upregulates OPRK1, which underpin androgen-independent growth of androgen-independent derivatives of LNCaP cells.

a, b Expressions of AR in LNCaP, AILNCaP2 (AI2), AILNCaP3 (AI3), and AILNCaP4 (AI4) treated with siRNA for GFP (siControl) and AR (siAR#1 and siAR#2) assessed by quantitative RT-PCR (a) and western blot (b). **P < 0.01, Normalized by GAPDH (a) and β-actin (ACTB, b). c Expression of OPRK1 in LNCaP, AILNCaP2 (AI2), AILNCaP3 (AI3), and AILNCaP4 (AI4) treated with siRNA for GFP (siControl) and AR (siAR#1 and siAR#2) assessed by quantitative RT-PCR. **P < 0.01, Normalized by GAPDH. d Expression of AR (left) and OPRK1 (right) in LNCaP cultured in media supplemented with 10%FBS, charcoal-stripped FBS (CSFBS), and CSFBS plus 1 nM dihydrotestosterone. **P < 0.01, Normalized by GAPDH. e Chromatin immunoprecipitation (ChIP) using anti-AR antibody and control IgG followed by quantitative RT-PCR using two primers for OPRK1 promoter (OPRK1-a, and -b) in LNCaP cells cultured as in d. **P < 0.01. f, g ChIP using anti-AR antibody and control IgG followed by quantitative RT-PCR using two primers for OPRK1 promoter (OPRK1-a, and -b) in KUCaP2 AD (f) and CR (g) tumors. **P < 0.01. h Expression of AR in LNCaP, AILNCaP2 (AI2), AILNCaP3 (AI3), and AILNCaP4 (AI4) treated with siRNA for GFP (siControl) and OPRK1 (siOPRK1#1 and siOPRK1#2) assessed by quantitative RT-PCR. Normalized by GAPDH. i Expression of ORPK1 normalized by GAPDH (left) and cell proliferation (right) in LNCaP, AILNCaP2 (AI2), AILNCaP3 (AI3), and AILNCaP4 (AI4) treated with siRNA for GFP (siControl) and OPRK1 (siOPRK1#1 and siOPRK1#2). *P < 0.05, **P < 0.01. j Summarized cell proliferation rates of indicated cells with regard to the concentration of supplemented nor-BNI, OPRK1 inhibitor. *P < 0.05.

Fig. 5. AR suppression upregulates OPRK1, which underpins androgen-independent growth of androgen-independent derivatives of VCaP cells.

a, b Expressions of AR in VCaP treated with siRNA for GFP (siControl) and AR (siAR#1 and siAR#2) assessed by quantitative RT-PCR (a) and western blot (b). **P < 0.01, Normalized by GAPDH (a) and β-actin (ACTB, b). c Expression of OPRK1 in VCaP treated with siRNA for GFP (siControl) and AR (siAR#1 and siAR#2) assessed by quantitative RT-PCR. **P < 0.01, Normalized by GAPDH. d Expression of AR (left) and OPRK1 (right) in VCaP cultured in media supplemented with 10% FBS, charcoal-stripped FBS (CSFBS), and CSFBS plus 1 nM dihydrotestosterone (DHT) assessed by quantitative RT-PCR. **P < 0.01, Normalized by GAPDH. e Summarized results of quantitative RT-PCR for full-length AR (AR-FL), AR variant 7 (AR-V7), KLK3, and OPRK1 normalized by GAPDH in indicated cells. **P < 0.01. f Western blot of indicated proteins in VCaP, PC3, AIVCaP1, AIVCaP2, and AIVaP3. β-actin (ACTB) acted as loading control. g Summarized results of proliferation rates of VCaP and AIVCaP2 (AIVCaP) treated with siRNA for GFP (siControl) and AR (siAR#1 and siAR#2). h Summarized results of proliferation rates of VCaP and AIVCaP2 treated with siRNA for GFP (siControl) and OPRK1 (siOPRK1#1 and siOPRK1#2). i Summarized cell proliferation rates of VCaP and AIVCaP2 (AIVCaP) with regard to the concentration of supplemented nor-BNI, OPRK1 inhibitor. *P < 0.05. j Cell proliferation in VCaP (left) and AIVCaP2 (AIVCaP) treated with siRNA for GFP (siControl), AR (siAR) OPRK1 (siOPRK1), or siAR plus siOPRK1. *P < 0.05, **P < 0.01. k Summarized results of migration (scratch) assay with greater distance indicating higher migration ability. VCaP, AIVCaP, and AILNCaP cells were treated with indicated siRNA and subject to the assay. **P < 0.01.

Fig. 6. Pharmacological inhibition of OPRK1 suppresses tumor growth in multiple in vivo castration-resistant prostate cancer models.

a Mice bearing VCaP cell-derived xenograft were castrated. Four weeks later, when the tumor started castration-resistant growth, mice were untreated or treated with OPRK1 inhibitor nor-BNI (n = 5 each). b Mice bearing VCaP cell-derived xenograft were castrated and untreated or treated with nor-BNI (n = 5 each) at the same time. c AILNCaP cells were inoculated to castrated mice. When the cell-derived xenograft tumors were engrafted and started growing, mice were untreated or treated with nor-BNI (n = 5 each). d Single-sample GSEA (ssGSEA) showing differentially enriched gene sets in VCaP cells treated with siRNA for AR (siAR) alone vs those treated with siAR and siRNA for OPRK1 (siORPK1). A gene set involved in SMAD6 pathway (red column) is enriched as well as some gene sets involved in neuronal pathways (light blue) and G-protein-related pathways (magenta). e Expressions of six genes involved in the SMAD6 pathway (JEON_SMAD6_TARGETS_UP) were evaluated using quantitative RT-PCR in VCaP or AIVCaP cells under AR signal suppression between siNTC and siOPRK1 treatments. f Representative photomicrograph images of hematoxylin and eosin (H&E) and immunohistochemical stains for OPRK1 in non-cancer prostate (benign), castration-sensitive (CSPC) prostate cancer, prostate cancer treated with neoadjuvant androgen deprivation therapy (NAADT), and castration-resistant prostate cancer (CRPC) tissues. g Immunostainability of OPRK1 classified into negative (none), weakly (weak), and strongly (strong) positive in benign, hormone-naïve (HNPC), PCa after neoadjuvant hormone therapy (NAHT), and CRPC. h Immunostainability of OPRK1 in HNPC tissues with regard to AR immunostainability. i Immunostainability of OPRK1 in CRPC tissues with regard to AR immunostainability.