Abstract PR009: A dependency map of androgen receptor activity identifies drivers of prostate cancer growth

Abstract Background. The Androgen Receptor (AR) is an important therapeutic target in patients with advanced prostate cancer (PC). Androgen deprivation therapy and drugs designed to inhibit AR signaling (ARSI) cause tumor regression and extend survival. However, disease progression is nearly universal and is usually accompanied by reactivation of AR signaling. To date, the specific mechanisms by which the AR maintains cell survival or promotes cell proliferation have not been defined and thus downstream programs cannot be exploited for treatment benefit. In this study, we sought to define the transcriptional program regulated by the AR in PC and identify AR target genes/proteins capable of promoting cell survival and growth. Methods. The transcriptional programs of established and novel PC cell line models with AR activity (ARPC) were characterized upon androgen treatment. Comparative analyses defined a cohort of 1144 common and unique AR-upregulated genes (ARGs) across the models. A CRISPR-Cas9 knockout library targeting these ARGs and AR-V7 associated genes was generated and viability screens were performed across six ARPC lines. Gene dependencies were identified by MAGeCK analysis. Open reading frames (ORFs) of essential ARGs were used for establishing stable cell lines in LNCaPFGC cells and the influence of ARG overexpression on proliferation determined by Incucyte live-cell Imaging and Cell-titer-glo assays. Results. Phenotypic effects of escalating androgen doses and ARSI co-treatment suggested involvement of transcriptional function of AR in promoting PC cell growth. CRISPR-Cas9 knockout screens identified ARG dependencies encompassing a wide range of cellular networks. Collectively, 29 ARGs were commonly essential across all models tested (core essential genes; CEG). Individual lines displayed additional private dependencies. Several CEGs namely PAK1IP1, CENPN, NKX3-1, SPDEF and POTEM displayed high expression in normal human prostate tissue. Notably, deletion of the tumor suppressor gene NKX3-1 resulted in complete cell growth arrest across 4 different ARPC models. To determine phenotypic functionalities, we overexpressed CEG ORFs in LNCaPFGC cells. While majority of the CEGs had neutral or modestly enhanced proliferative fitness in the basal state, SPDEF, POTEM and GMPPB repressed LNCaPFGC growth. Conclusions. The AR transcriptionally regulates a cell autonomous genetic program that promotes PC survival and proliferation. A component of this program appears to function via a hierarchical network involving other transcription factors such as NKX3-1 that may exert survival and growth activity in addition to its known tumor suppressor functions. Current efforts are being directed toward developing a mechanistic understanding of cell-type context for the individual CEGs; identifying minimal and sufficient CEG combinations which can drive AR-mediated survival program; and identifying epistatic relationships between growth sustaining and repressing CEGs in tuning cell proliferation downstream of androgens. Citation Format: Arnab Bose, Ilsa Coleman, Dapei Li, Rabeya Bilkis, Brian Hanratty, Michael Nyquist, Jared Lucas, Sander Frank, Peter Nelson. A dependency map of androgen receptor activity identifies drivers of prostate cancer growth abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Innovations in Prostate Cancer Research and Treatment; 2026 Jan 20-22; Philadelphia PA. Philadelphia (PA): AACR; Cancer Res 2026;86 (2Suppl): Abstract nr PR009.