Androgen Receptor Variants Mediate DNA Repair after Prostate Cancer Irradiation

Abstract

In prostate cancer, androgen deprivation therapy (ADT) enhances the cytotoxic effects of radiotherapy. This effect is associated with weakening of the DNA damage response (DDR) normally supported by the androgen receptor. As a significant number of patients will fail combined ADT and radiotherapy, we hypothesized that DDR may be driven by androgen receptor splice variants (ARV) induced by ADT. Investigating this hypothesis, we found that ARVs increase the clonogenic survival of prostate cancer cells after irradiation in an ADT-independent manner. Notably, prostate cancer cell irradiation triggers binding of ARV to the catalytic subunit of the critical DNA repair kinase DNA-PK. Pharmacologic inhibition of DNA-PKc blocked this interaction, increased DNA damage, and elevated prostate cancer cell death after irradiation. Our findings provide a mechanistic rationale for therapeutic targeting of DNA-PK in the context of combined ADT and radiotherapy as a strategy to radiosensitize clinically localized prostate cancer. Cancer Res; 77(18); 4745-54. ©2017 AACR.

Fig. 1

ARV mediates DDR and clonogenic survival. (A) R1-D567 and (B) R1-AD1 cells in the presence of vehicle (DMSO) or Enza (5μM), were subjected to IR (2Gy), fixed and stained with γ-H2AX (green) and 53BP1 (red) antibodies. Foci imaged by confocal microscopy were plotted as average foci per cell. (C) R1-D567, R1-AD1, C4-2 and 22Rv1 cells were cultured with vehicle or Enza for 24 hours, then treated with escalating doses of IR. Following fixation and staining 14 days later, colonies greater than 50 cells were counted. **p<0.01.

Fig. 2

ARVs are necessary for DDR and are recruited to DNA damage sites: (A) Clonogenic survival of 22Rv1 following knockdown of ARfl alone, ARv7 alone or both ARfl and ARv7 demonstrates that either ARVs or ARfl alone can mediate DDR. Western blot shows ARfl and ARV protein expression in 22Rv1 cells following siRNA knockdown (right), with β-actin control. (B) PLA shows IR-induced interaction between ARV and γ-H2AX in R1-D567 cells. (controls: AR and mIgG, AR and ARID1A and γ-H2AX and rIgG). All irradiated cells (2 Gy) contained >5 dots each. (C) Venn diagram (left) and heatmap (middle) and gene body plots (right) shows the increase in RNA polymerase II-enriched ChIP-seq peaks in R1-D567 upon IR treatment. (D) Venn diagram (left) and heatmap (middle) and gene body plots (right) shows the increase in RNA polymerase II transcription start site (TSS) in R1-D567 upon IR treatment.

Fig. 3

IR increases the interaction between ARV and DNA-PKc. (A) IP-MS analysis of proteins interacting with ARv567es in R1-D567 cells shows that PRKDC is the top binder to ARVs: a list of the top AR binders is tabulated. Ability of IR (10 Gy) to enhance interaction between ARV or ARfl and DNA-PKc is demonstrated by Co-IP (B) and PLA (C) at 16 hours after IR. The interaction between ARV and DNA-PKc in R1-D567 cells can be blocked by pretreatment with NU7441 (D).

Fig. 4

DNA-PKc inhibitors block DDR following IR and enhance IR-mediated PCa cell death: Pretreatment with NU7441 enhances the IR-induced olive tail moment in comet assays (A) and alters kinetics of DNA repair, as evidenced by persistence of γ-H2Ax foci at 24h in R1-D567 cells (p<0.0001) and decreased IR induced phospho-DNA-PKc at 1 h after IR (B). (C) Clonogenic survival assays indicate that NU7441 pretreatment dramatically decreases the survival of R1-AD1, R1-D567 and 22Rv1 cells (p≤ 0.02). (D) Pretreatment of R1-D567 xenografts (200 mm³) with NU7441 prior to a single IR dose (2Gy) decreased tumor growth more than IR alone (each xenograft is represented by a separate line, with each measurement of tumor volume represented by a dot. Groups of mice receiving the same therapy are grouped by color. Tumor growth rates after NU7441, IR or both all significantly differed from controls when assessed by split-plot ANOVA (p = 0.009, 0.003 and <0.001, respectively). Pretreated R1-D567 xenografts shows persistence of γ-H2AX and 53BP1 foci at 24h after IR. (Scale bar- 40 μm; blue- DAPI; green- γ-H2AX; red- 53BP1)(p < 0.001; n = 3).