Castration-resistant prostate cancer is resensitized to androgen deprivation by autophagy-dependent apoptosis induced by blocking SKP2

Abstract

Resistance to androgen receptor (AR)-targeted therapies for prostate cancer (PCa) is characteristic of an aggressive subtype called castration-resistant prostate cancer (CRPC) and is often associated with tumor relapse. Both relapse and poor prognosis in patients with CRPC are associated with increased abundance of the E3 ubiquitin ligase SKP2. Therefore, we investigated the therapeutic potential of combined inhibition of AR and SKP2 for CRPC. We found that combined targeting of AR and SKP2 with small-molecule inhibitors decreased proliferation in two CRPC cell lines in culture and in xenografts in humanized mice. Furthermore, combined therapy in mice markedly decreased the growth of Pten/Trp53 double-knockout tumors, a particularly invasive model of CRPC, whereas disruption of either AR or SKP2 alone only modestly suppressed their growth. Mechanistically, the inhibition of SKP2 in CRPC cells induced autophagy-dependent apoptosis and promoted luminal-associated phenotypes, which promoted responsiveness to AR-targeted therapy. These effects were further enhanced by coinhibition of AR and were not induced by the AR inhibitor alone. Our findings indicate that targeting both AR and SKP2 signaling pathways is necessary to treat CRPC.

Fig. 1.. Activation of autophagy-dependent apoptosis in human PCa upon SKP2 KD.

(A) Scatterplots based on normalized androgen receptor (AR) versus normalized SKP2 mRNA expression in advanced prostate cancer (PCa). Linear fits were calculated and corresponding Pearson correlation is shown in localized PCa (GSE32269; r= 0.4339; P=0.0494). PCa gene expression dataset was retrieved from the GEO database (ncbi.nlm.nih.gov/gds/). The corresponding AR and SKP2 values were normalized and plotted. (B) Castration resistant prostate cancer (CRPC) clinical dataset (n=98 tumors) categorized by AR and neuroendocrine (NE) marker expression (GSE126078). The histogram displays the corresponding AR and SKP2 values. (C) Immunofluorescence (IF) staining for AR and SKP2 protein levels in tissue microarray (TMA) of primary prostate tumors (n=34 tumors). Scale bars are 25 μm. (D and E) Immunoblot analysis displays elevation in both apoptotic markers including cleaved PARP and cleaved caspase-3 and autophagy markers including LC3-I/II, Beclin-1 and phosphorylated AMPKα at Thr¹⁷² [p-AMPKα (T172)] upon SKP2 knockdown (shSKP2) in C4–2B and 22Rv1. Quantification analysis for the relative protein levels is displayed in the side panel (n=3 independent experiments). (F) Quantification of RT-qPCR analysis for SKP2 mRNA levels and autophagy promoting genes including MAP1LC3B, Beclin-1 and ATG5 relative to housekeeping gene Beta-actin in C4–2B (n=3 independent experiments). (G) IF images demonstrating the effects of SKP2 KD on autophagy in C4–2B. Cells were fixed and immunostained with anti-LC3 (red) and counterstained with DAPI for nuclear visualization. Scale bars are 20 μm. The relative fluorescence intensity is displayed in the right panel (n=3 sections from 3 independent experiments). (H and I) Apoptosis profile using flow cytometry in SKP2 KD for propidium iodide (PI) and Annexin V for C4–2B. Representative plot for percent of dead cells (n=7 independent experiments) and representative plot for percent of total cell pollution undergoing necrosis, early and late apoptosis (n=11 independent experiments). Comparison between groups was performed using either Student’s t-test or one-way ANOVA when comparing three or more groups. Bars indicate SEM. * P< 0.05, ** P< 0.01, *** P< 0.001, **** P<0.0001.

Fig. 2.. Pharmacological inhibition of SKP2 resensitizes human PCa cells to AR signaling inhibitor.

(A) Comparison on the effects of SKP2 (SZL P1–41) and AR (MDV3100) treatment individually or in combination on C4–2B proliferation (n=3 independent experiments/treatment group). (B) RT-qPCR analysis shows the mRNA levels of TMPRSS2 and KLK3 relative to housekeeping gene Beta-actin upon SKP2 (SZL P1–41) and/or AR (MDV3100) treatment in C4–2B (n=3 independent experiments/treatment group). (C) RT-qPCR analysis for the relative (to Beta-actin) mRNA levels for autophagy promoting genes including MAP1LC3B and BCN1 (Beclin-1) upon SKP2 (SZL P1–41) and/or AR (MDV3100) inhibition in C4–2B (n=3 independent experiments/treatment group). (D) Left panel, immunoblot analysis for C4–2B at 48 hours proceeding individual and/or combined SKP2 (SZL P1–41; 10 μM) and AR (MDV3100; 10 μM) treatment. Right panel, quantification of immunoblot analysis for the corresponding protein levels relative to controls (n=3 independent experiments). (E) Immunofluorescence (IF) images demonstrating the effects of single and/or combined treatment using SKP2 (SZL P1–41; 10 μM) and AR (MDV3100; 10 μM) inhibitors on autophagy promoting LC3 (red) protein in C4–2B. Nucleus was counterstained with DAPI. Scale bars are 20 μm. Right panel displays the relative fluorescence intensity for LC3 levels (n=3 sections from 3 independent experiments/treatment group). (F) TUNEL staining (left) for SKP2 (SZL P1–41; 10 μM) and AR (MDV3100; 10 μM) single and/or combined inhibition for promoting apoptosis in C4–2B. DAPI was used for nuclear staining. Scale bars are 25 μm. Right panel displays the relative TUNEL (red) intensity for treated C4–2B (n= ~500 cells from 3 independent experiments/treatment group). Comparison between groups was performed using either Student’s t-test or one-way ANOVA when comparing three or more groups. Bars indicate SEM. * P< 0.05, ** P< 0.01, *** P< 0.001, **** P<0.0001.

Fig. 3.. Combined therapeutic targeting of AR and SKP2 inhibits tumorigenesis in humanized PCa mouse models.

(A) Schematic representation of C4–2B xenograft in NOD/SCID mice. (B) Representative C4–2B biopsies with the corresponding tumor weights (n=5 mice/treatment group for a total of 20 mice) at end point for vehicle (DMSO), AR (MDV3100; 30mg/kg; three times/week; i.p. injection) and/or SKP2 (SZL P1–41; 30mg/kg; three times/week; i.p. injection) treated mice for ~30 days. (C) Immunoblot analysis for C4–2B tumors treated with individual or combined AR and SKP2 therapy. Lower panel shows quantification of protein levels for immunoblot analysis (n=3 independent experiments). (D) Immunohistochemistry (IHC) staining for C4–2B xenograft tumor sections for the indicated antibodies. Scale bars are 200 μm. Lower panel demonstrates quantification of IHC staining (n= ~500 cells from 3 mice tumors/treatment group). (E) TUNEL staining for C4–2B tumor sections. Scale bars are 25 μm. The corresponding fluorescence quantification is displayed in the bottom panel (n= ~500 cells from 3 mice tumors/treatment group). Comparison between groups was performed using either Student’s t-test or one-way ANOVA when comparing three or more groups. Bars indicate SEM. * P< 0.05, ** P< 0.01, *** P< 0.001, **** P<0.0001.

Fig. 4.. SKP2 abrogation elevates autophagy-dependent apoptosis signaling in vivo in mouse models.

(A) Immunoblot analysis for Skp2 knockout (KO) anterior prostate (AP) with the indicated genotypes. Quantification of protein levels relative to β-actin for apoptotic markers (cleaved PARP and cleaved caspase-3) and autophagy markers (LC3-I/II, p-AMPKα (Thr¹⁷²) and Beclin-1/β-actin) are displayed to the right (n=3 independent experiments). (B and C) Immunohistochemistry (IHC) staining of LC3 and cleaved caspase-3 in prostate tissues of mice with the indicated genotypes. Scale bars are 100 μm. The right panel displays a quantification analysis for IHC staining (n= ~500 cells from 3 mice tumors/treatment group). (D) Schematic and representative AP biopsies (top plot) for Pten^pc−/−; Trp53^pc−/− mutant mice treated for ~30 days with a vehicle (DMSO), individual or combined SKP2 (SZL P1–41; 30mg/kg) and AR (MDV3100; 30mg/kg) inhibitors. Corresponding AP tumor weights (n=15 mice tumors) for vehicle controls, individually treated mice or mice treated in combination (lower plot). (E) Immunofluorescence (IF) images demonstrating colocalization between AR, SKP2, and PCNA (proliferating cell nuclear antigen) in prostate tissues of Wt and Pten^pc−/−; Trp53^pc−/− mutant mice. SKP2 conditional KO in Pten^pc−/−; Trp53^pc−/−; Skp2^−/− mutant mice demonstrate decreased AR, SKP2 and PCNA colocalization, similarly to Pten^pc−/−; Trp53^pc−/− mutant mice after combined treatment with SKP2 (SZL P1–41) and AR (MDV3100). Right panel represents quantification of AR, SKP2, and PCNA relative fluoresce intensity for the AP indicated genotypes and treatment group (n= ~500 cells from 3 mice tumors/treatment group). Scale bars are 50 μm. (F) IF images for cellular lineage markers including luminal (CD24⁺) and basal (CD49f⁺) in Pten^pc−/−; Trp53^pc−/− mutant mice treated individually or in combination with SKP2 (SZL P1–41) and/or AR (MDV3100) inhibitors. Scale bars are 25 μm. Corresponding quantification for IF staining is displayed in the right panel (n= ~500 cells from 3 mice tumors/treatment group). (G) A working model for combined SZL P1–41 and MDV3100 treatment in Pten^pc−/−; Trp53^pc−/− mutant mice. Comparison between groups was performed using either Student’s t-test or one-way ANOVA when comparing three or more groups. Bars indicate SEM. * P< 0.05, ** P< 0.01, *** P< 0.001, **** P<0.0001.