Androgen receptor modulation by non-androgenic factors and the basal transcription factor TAF1

There is considerable evidence to suggest that progression of prostate cancers to castration-resistant is due to inappropriate activation of the androgen receptor (AR) and hence, the AR is a good target for the treatment of this disease. In this study we used two approaches to investigate AR activation and inhibition. First, we developed high-throughput, non-invasive, cell-based screening assays to rapidly and biologically assess factors that modulate prostate cancer growth and affect AR activity. Using these assays, we found that differentiated osteoblast-like condition media enhanced prostate cancer cell growth, but not AR activity. In addition, we applied this system to screen compounds, selected through in silico approaches against a crucial pocket on the AR protein. The application of our in silico tools and our cell based screening assay resulted in identification of 17 compounds out of 4 millions that can inhibit AR activity. Importantly, some of these compounds are more potent than bicalutamide, which is one of the most potent antiandrogen drugs currently used to treat patients with metastatic prostate cancer. In the context of AR transcription target genes, the presence of AR coactivators is essential for AR activity. Using the repressed transactivator yeast two-hybrid system, we found that TATA binding protein-associated factor 1 (TAF1) interacted with the AR. In tissue microarrays, TAF1 was shown to steadily increase with duration of neoadjuvant androgen withdrawal and with progression to castration resistance. GST pull-down assays established TAF1/AR interaction and co-immunoprecipitation and ChIP assays revealed colocalization of TAF1 and AR on the prostate specific antigen (PSA) promoter in prostate cancer cells. With respect to modulation of AR activity, while full-length TAF1 showed enhancement of both AR and some generic gene transcriptional activity, selective AR coactivator activity was demonstrated in transactivation experiments using cloned TAF1 N-terminal kinase and ubiquitin-activating/conjugating (E1/E2) domains. In keeping with AR coactivation by the E1/E2 domain, TAF1 was found to greatly increase the cellular amount of poly-ubiquitinated AR. In conclusion, our results indicate that TAF1 is a coactivator of AR and its overexpression could be part of a compensatory mechanism adapted by cancer cells to overcome reduced levels of circulating androgens.Medicine, Faculty ofPathology and Laboratory Medicine, Department ofGraduat