Nuclear receptor-mediated transcriptional regulation in prostate cancer cells

Prostate cancer is the most commonly diagnosed cancer and the second leading cause of cancer-related death for men in the United States. In prostate carcinogenesis, the critical role of androgen receptor (AR) signaling has been well established and the transcriptional networks governed by AR have been under rigorous investigations. AR belongs to nuclear receptor (NR) superfamily which, by directly regulating gene expression in response to hormones, plays important roles in normal development and pathological processes. In addition to AR, other NRs, such as vitamin D receptor (VDR) and retinoic acid receptor alpha (RARA), have also been shown to affect prostate cancer development. However, the understanding of transcriptional programs governed by these NRs is fragmentary and how they interact with AR is poorly understood. In order to address these questions, we have used an integrated genomic approach to uncover the transcriptional networks governed by AR, VDR and RARA in the prostate cancer cell line LNCaP.;We first studied the androgen signaling and the vitamin D signaling in prostate cancer. We identified transcriptional targets of AR and VDR in LNCaP cells by chromatin immunoprecipitation followed by high-throughput DNA sequencing (ChIP-Seq) and gene expression microarray experiments. We found that AR and VDR antagonistically regulate a subset of cell cycle-related genes which are over-expressed in prostate cancer tumors. The expression balance of these genes is partially regulated through the competitive dynamics between AR and VDR binding to shared cis-regulatory elements. On such shared elements, we found that FOXA1 mediates this competition by serving as a pioneering factor for both AR and VDR binding. We also found significant enrichment of AR-, VDR-, and AR/VDR overlapping binding sites in prostate cancer risk-associated single-nucleotide polymorphism (SNP) intervals identified from genome-wide association studies (GWAS), providing genetic evidence to link AR, VDR and their crosstalk to prostate cancer susceptibilities. In particular, we found that the T/C alleles of rs339331 might affect prostate cancer risk by altering the signaling interaction between AR and VDR on an intronic regulatory region of RFX6, providing a novel example to connect non-coding genomic variations and the integration of sex and nutrition hormones to disease etiology.;We have defined the transcriptional networks governed by RARA in LNCaP cells with ChIP-Seq and gene expression profiling. We found that, in the androgen-depleted condition, activation of RARA induces the expression of cell cycle-related genes which are over-expressed in prostate cancer tumors. The up-regulation of these genes is partially mediated through E2F1, which is up-regulated by RARA activation and bound by RARA within 50 kb of its TSS. We found significant enrichment of LNCaP RARA binding sites in prostate cancer GWAS SNP intervals, indicating that some of these risk-associating genomic variations might change the activity of regulatory elements bound by RARA. This enrichment also underscores the importance of RARA-mediated transcriptional networks in prostate cancer etiology. Interestingly, we found that the regulatory region containing rs339331 is also bound by RARA, suggesting that this region might be a transcriptional regulatory hotspot that integrates signaling from multiple pathways. Furthermore, we propose that upon the activation of RARA, the up-regulation of RFX6 might act as a parallel oncogenic stimulation with E2F1 up-regulation to promote prostate cancer development. Taken together, we have proposed molecular mechanisms underlying the RAR-induced proliferation in LNCaP cells under the androgen-depleted condition, providing novel insights into targeting retinoic acid signaling in prostate cancer patients following androgen-depletion therapies.;To further understand the function of regulatory elements bound by NRs, we have developed a method we call "Enhancer-Seq", which can be used to measure enhancer activity in a high-throughput fashion. In the pilot study of Enhancer-Seq, we have made a reporter vector library containing about 100,000 self-transcribing potential enhancers that are binding sites of nuclear receptors and their cofactors in the breast cancer cell line MCF-7 and the LNCaP cell line. We have measured the activity of these enhancers in response to DHT treatment in LNCaP cells. Our analysis proved the functionality and reproducibility of Enhancer-Seq, despite the need for further technical improvements. We believe that this assay will provide a powerful tool to systematically study the transcriptional regulatory function of NR binding sites. In addition, this assay has application potential in a wide range of transcriptional regulation studies, including elucidation of enhancer evolution, identification of enhancer quantitative trait locus (QTL), and quantitative modeling of transcriptional regulation.