HDAC6 regulates androgen receptor hypersensitivity and nuclear localization via modulating Hsp90 acetylation in castration-resistant prostate cancer

Abstract

The development of castration-resistant prostate cancer (PCa) requires that under castration conditions, the androgen receptor (AR) remains active and thus nuclear. Heat shock protein 90 (Hsp90) plays a key role in androgen-induced and -independent nuclear localization and activation of AR. Histone deacetylase 6 (HDAC6) is implicated, but has not been proven, in regulating AR activity via modulating Hsp90 acetylation. Here, we report that knockdown of HDAC6 in C4-2 cells using short hairpin RNA impaired ligand-independent nuclear localization of endogenous AR and inhibited PSA expression and cell growth in the absence or presence of dihydrotestosterone (DHT). The dose-response curve of DHT-stimulated C4-2 colony formation was shifted by shHDAC6 such that approximately 10-fold higher concentration of DHT is required, indicating a requirement for HDAC6 in AR hypersensitivity. HDAC6 knockdown also inhibited C4-2 xenograft tumor establishment in castrated, but not in testes-intact, nude mice. Studies using HDAC6-deficient mouse embryonic fibroblasts cells showed that inhibition of AR nuclear localization by HDAC6 knockdown can be largely alleviated by expressing a deacetylation mimic Hsp90 mutant. Taken together, our studies suggest that HDAC6 regulates AR hypersensitivity and nuclear localization, mainly via modulating HSP90 acetylation. Targeting HDAC6 alone or in combination with other therapeutic approaches is a promising new strategy for prevention and/or treatment of castration-resistant PCa.

Figure 1

Effect of the HDAC6 inhibitor TSA on GFP-AR localization and expression in castration-resistant PCa cells. Castration-resistant C4-2 or androgen-sensitive LNCaP cells were transfected with GFP-AR and 4 h later were treated with ethanol (vehicle), 1 μm TSA, or 1 mm NaBut in ligand-free conditions. A, Localization of GFP-AR in C4-2 or LNCaP cells was assessed by fluorescence microscopy 16 h after treatment. B, Results are the average of five experiments, in which more than 100 cells were analyzed for each experiment. Error bars represent ± sd. *, A P value < 0.05 was generated using an unpaired t test in GraphPad Prism. C, Effect of TSA on GFP-AR protein levels in C4-2 cells. Whole-cell lysates of C4-2 cells with identical treatments as in panel A were subjected to SDS-PAGE, and GFP-AR levels were determined by Western blot analysis.

Figure 2

Effect of HDAC6 inhibitor TSA on endogenous AR expression and transcriptional activity in PCa cells. A, C4-2 and LNCaP cells were cultured in ligand-free conditions for 24 h before treatment with ethanol (vehicle), 1 μm TSA, or 1 mm NaBut in the absence (−) or presence (+) of 1 nm DHT. The cell lysates were collected 16 h after treatment and were immunoblotted with anti-AR, HDAC6, Hsp90, or PSA antibody. GAPDH was used as loading control. Total RNA from cells with identical treatment as panel A were used for real-time RT-PCR analysis to determine AR (B) and PSA (C) mRNA levels in C4-2 and LNCaP cells. Error bars represent ± sd. GAPDH was used to normalize the quantitative real-time PCR results.

Figure 3

Effect of HDAC6 knockdown on endogenous AR expression and its transcriptional activity in PCa cells. A, C4-2 cells were uninfected or infected with the lentivirus encoding for shRNA to HDAC6 (shHDAC6) or luciferase (shControl), and stable cells were cultured in ligand-free conditions for 24 h before treatment with (+) or without (−) 1 nm DHT for an additional 16 h. Then the cell lyastes were immunoblotted with anti-HDAC6, HDAC10, AR, Hsp90, PSA, or GAPDH antibody. B, Total RNA isolated from C4-2 cells identically treated as in panel A were used for real-time RT-PCR to determine PSA mRNA level. C, C4-2 cells with HDAC6 knockdown or shRNA control were cultured in conditions with or without DHT for 16 h, and then Hsp90 was immunoprecipitated (IP) from the cell lysates and immunoblotted (IB) with either anti-Hsp90 or antiacetylated lysine antibody. GAPDH served as loading control or internal control. D, C4-2 cells were cotransfected with shHDAC6 or shControl vector and wild-type (WT) or codon-switched (CS) HDAC6 expression vector, and Flag-tagged HDAC6 expression was detected by anti-Flag antibody. E, GFP-positive transfected C4-2 cells were sorted, and endogenous PSA mRNA was determined using CellsDirect One-Step qRT-PCR (Invitrogen, Carlsbad, CA). Error bars represent ± sd. A P value < 0.05 or < 0.01 was generated using an unpaired t test in GraphPad Prism. Acet, Acetylated lysine.

Figure 4

Effects of HDAC6 knockdown or knockout on AR intracellular localization. A, shControl and HDAC6-knockdown C4-2 cells were cultured in ligand-free conditions for 24 h after which AR localization was determined by immunostaining with an anti-AR antibody. Nuclei were stained with 1 μg/ml Hoechst 33342. B, Wild-type and HDAC6-knockout MEFs were transfected with GFP-AR, or HDAC6-knockout MEFs were cotransfected with GFP-AR and wild-type or Hsp90 mutants or HDAC6 expression vector. After 4 h, the cells were cultured in ligand-free or complete medium. Localization of GFP-AR was assessed by fluorescence microscopy after an additional 16 h in culture. Results are the average of five experiments, in which more than 100 cells were analyzed for each experiment. Error bars represent ± sd. A P value < 0.05 or < 0.01 was generated using an unpaired t test in GraphPad Prism. KO, Knockout; WT, wild type; Q, K294Q; R, K294R.

Figure 5

Effect of HDAC6 knockdown on PCa cell growth. Control and HDAC6-knockdown C4-2 cells were cultured in ligand-free (A) or complete (B) medium with an initial 50,000 cells per well on six-well plates in triplicate. The cell number was counted at the indicated time points. The results presented here are representative of three independent experiments. Error bars represent ± sd. *, A P value < 0.01 was generated using an unpaired t test in GraphPad Prism. C, Control and HDAC6-knockdown C4-2 cells were seeded at 10,000 cells per 10-cm dish in triplicate. After 2 wk in T-medium without or with the indicated concentrations of DHT, cell colonies were stained with crystal violet and colony numbers counted. The colony images are representative of three independent experiments. D, Quantitative analysis of colony formation assay. The colony numbers represent means.

Figure 6

HDAC6 knockdown inhibits C4-2 xenograft tumor growth. A, Xenograft tumor formation in intact mice, indicated by an arrow. B, Tumor growth is shown as a function of tumor volume at indicated days after s.c. injection. Error bars represent ± sem. *, A P value < 0.05 was generated using an unpaired t test in GraphPad Prism. C, Tumor take rate was calculated 12 wk after s.c. injection. D, Tissue lysates of xenografts collected from intact mice were immunoblotted with anti-HDAC6, anti-Hsp90, anti-AR, and anti-PSA antibodies. GAPDH served as loading control.