A Prostate Cancer Androgen Receptor Variant Gene Network With Prognostic And Therapeutic Value Identified Using Integrated Systems Analyses

Constitutively active androgen receptor (AR) splice variants such as AR-V7 represent a well-established mechanism of disease progression to castration-resistant prostate cancer (CRPC), the incurable form of PC. These AR variants lack the carboxyl-terminal ligand binding domain, the target of all approved drugs against AR. Consequently, a significant challenge is to identify novel therapeutic strategies that are effective against AR variants-expressing prostate tumors without requiring direct targeting of these variants, which is challenging due in part to the unstructured nature of the AR amino terminus. We used a multi-pronged, unbiased systems biology approach to identify downstream AR-V7-regulated hub genes that both drive cancer progression and feedback to enhance AR-V7 activity. Such genes likely encode prognostic markers as well as potential therapeutic targets acting within the AR-V7 network. Our systems biology approach utilized: 1) clinically relevant gene sets upregulated during human PC progression obtained by Weighted Gene-Co-expression Network Analysis (WGCNA); 2) an AR-V7 functional interactome from a high-throughput synthetic genetic array (SGA) screening in the yeast, Schizosaccharomyces pombe; and 3) PC cell AR-V7 transcriptome. We identified seven genes that were upregulated in human PC, functionally interacted with AR-V7 and were AR-V7 targets in PC. This gene set, composed of cell cycle and mitosis-regulating genes, not only encoded select proteins previously linked to PC, such as Cyclin B1 and Topoisomerase 2A (CCNB1 and TOP2A), but also proteins with little or no prior association with AR-V7 or PC. The seven genes comprised a signature that strongly correlated with patient Gleason score and predicted disease free-survival in large independent PC patient cohorts. In contrast, this gene set was not associated with survival metrics in other human cancers. Individual depletion of the expression of these genes decreased ligand-independent AR transcriptional activity and reduced CRPC cell proliferation. Furthermore, individual knock down of the members of the gene set affected the expression of the other members of the set indicative of the interrelationship of these genes. Consistent with CRPC dependence on these interrelated mitotic genes, combinatorial targeting of two of its members: TOP2A and CCNB1, led to a synergistic and selective inhibition of AR-driven CRPC cell proliferation but had no effect on non-tumorigenic prostate epithelial cells. This unbiased and novel gene discovery strategy, which is broadly applicable to other cancers, identified clinically-relevant interacting gene hubs forming the basis for both prognostic use and rational, combinatorial therapy for CRPC. Citation Format: Maria J. Martinez, Fiorella Magani, Eric R. Bray, Valeria Copello, Ning Zhao, Stephanie Peacock, David J. Wiley, Gennaro D9Urso, Kerry L. Burnstein. A prostate cancer androgen receptor variant gene network with prognostic and therapeutic value identified using integrated systems analyses [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1314.