p68/DdX5 supports β-catenin & RNAP II during androgen receptor mediated transcription in prostate cancer

Abstract

The DEAD box RNA helicase p68 (Ddx5) is an important androgen receptor (AR) transcriptional co-activator in prostate cancer (PCa) and is over-expressed in late stage disease. β-Catenin is a multifunctional protein with important structural and signalling functions which is up-regulated in PCa and similar to p68, interacts with the AR to co-activate expression of AR target genes. Importantly, p68 forms complexes with nuclear β-Catenin and promotes gene transcription in colon cancer indicating a functional interplay between these two proteins in cancer progression. In this study, we explore the relationship of p68 and β-Catenin in PCa to assess their potential co-operation in AR-dependent gene expression, which may be of importance in the development of castrate resistant prostate cancer (CRPCa). We use immunoprecipitation to demonstrate a novel interaction between p68 and β-Catenin in the nucleus of PCa cells, which is androgen dependent in LNCaP cells but androgen independent in a hormone refractory derivative of the same cell line (representative of the CRPCa disease type). Enhanced AR activity is seen in androgen-dependent luciferase reporter assays upon transient co-transfection of p68 and β-Catenin as an additive effect, and p68-depleted Chromatin-Immunoprecipitation (ChIP) showed a decrease in the recruitment of the AR and β-Catenin to androgen responsive promoter regions. In addition, we found p68 immunoprecipitated with the processive and non-processive form of RNA polymerase II (RNAP II) and show p68 recruited to elongating regions of the AR mediated PSA gene, suggesting a role for p68 in facilitating RNAP II transcription of AR mediated genes. These results suggest p68 is important in facilitating β-Catenin and AR transcriptional activity in PCa cells.

Figure 1. Localisation and interaction of p68 and β-Catenin in PCa cells.

A. Cropped immuno-blot images show cytoplasmic and nuclear LNCaP and LNCaP-AI PCa cell lysates (+/− R1881 10 nM, 8 hours), probed sequentially with β-Catenin, AR, p68, α-tubulin and TATA binding (TBP) antibodies. B. Interaction of ectopic p68 and β-Catenin in COS-7 cells. Whole cell lysates of COS-7 cells transfected with pcDNA₃-p68-myc and pCS³⁺-Myc₆-β-Catenin constructs (+/− R1881 10 nM, 8 hours), were immunoprecipitated with β-Catenin and p68 antibody respectively. Cropped immuno-blots were probed sequentially with β-Catenin, p68 and myc antibody.C. Interaction of endogenous p68 and β-Catenin in the nucleus of LNCaP and LNCaP-AI PCa cells in the presence and absence of androgens. Cropped immuno-blot images of LNCaP and LNCaP-AI nuclear lysates, immunoprecipitated with p68 antibody (+/− R1881 10 nM, 8 hours), and probed sequentially with β-Catenin and p68. Extract samples contain either whole cell or nuclear lysate and protein G sepharose with no antibody present. Control (Con) samples contain antibody and protein G sepharose in extraction buffer only.

Figure 2. p68 interacts with RNAPII in PCa cells and occupies many regions of the PSA gene.

A. Cropped immuno-blot images of LNCaP nuclear lysates immunoprecipitated with RNAP II H5 (ser-2), RNAP II H14 (ser-5), RNAP II CTD mouse monoclonal and p68 antibody (+/− R1881 10 nM, 8 hours), probed sequentially with p68 and RNAP II antibody. Extract samples contain nuclear cell lysate and protein G sepharose with no antibody present. Control samples contain antibody and protein G sepharose in nuclear extraction buffer only. B. Recruitment of AR and C. p68 to regions of the PSA gene. LNCaP cells were treated with 10 nM R1881 and harvested at 0, 15, 30, 45, 90 & 120 minute time points. Samples were immunoprecipitated with AR, p68 or control IgG antibodies and recovered material processed by ChIP assay. QPCR data are representative of n = 3 independent ChIP assays normalised to input levels (+/− SE). (N.B. QPCR primers could not be optimised to all exonic and intronic regions of the PSA gene). The independent paired sample t test was used to compare enrichment in recruitment between different PSA regions and show significance. D. Diagram of PSA gene depicting exon/intron boundaries.

Figure 3. Knockdown of p68 does not alter β-Catenin mRNA or protein expression.

mRNA expression levels of p68 A. and β-Catenin B. in control (NS) and p68 siRNA-transfected LNCaP cells (+/− R1881 10 nM, 16 hours). QPCR data was normalised to GAPDH levels and fold change calculated relative to control (NS) (-R1881) mRNA levels (set as 1). The independent sample t test was used to compare differences in expression levels and show significance. C. Cropped immuno-blot images of lysates from LNCaP cells treated with 10 nM R1881 (16 hours) and transfected with control (NS) and p68 siRNA. Blots probed sequentially with β-Catenin, p68 and α-tubulin antibodies.

Figure 4. Depletion of p68 reduces AR and β-Catenin recruitment to promoter regions of androgen responsive genes.

LNCaP cells transfected with p68 or control (NS) siRNA, treated with 10 nM R1881 for 90 minutes and immunoprecipitated with either AR A. B. C. & D. or β-Catenin E. F. G. & H. antibodies (including a IgG control antibody). Recovered material was processed by ChIP assay and recruitment to PSA ARE I (A & E), ARE III (B & F), KLK2 (C & G) & TMPRSS2 (D & H) promoter regions assessed relative to 0 minute time point. p68 depletion in LNCaP cells showed reduced AR & β-Catenin recruitment after treatment with 10 nM R1881 for 90 minutes to all regions assessed compared to control (NS) siRNA cells. Results shown represent n = 3 independent experiments (+/− SD). The independent sample t test was used to compare differences in expression levels and show significance.

Figure 5. Over-expression of p68 additively enhances activity of β-Catenin mediated AR transcription.

COS-7 cells transiently transfected in triplicate with p(ARE)₃Luc reporter and PCMV-β-galactosidase plasmids together with mammalian expression vectors for AR, β-Catenin and p68 (+/−10 nM R1881). Luciferase activity was corrected for the corresponding β-galactosidase activity to give relative activity. The range of plasmid levels (+ and ++) corresponds to 50 and 100 ng respectively. Data shown relative to AR activity alone (−R1881) (set as 1), and representative of at least n = 3 luciferase assay experiments (+/− SE). The independent sample t test was used to compare differences in expression levels and show significance.