Mechanisms of Translation in Cellular Dynamics and Molecular Regulation of Advanced Prostate Cancer

Abstract

Prostate cancer is one of the most common cancers in men, and advanced prostate cancer is largely incurable and lethal. Mainstream treatment options for advanced prostate cancer all target the androgen hormone signaling pathway. The androgen receptor (AR) is a ligand-activated nuclear receptor that mediates widespread transcriptional changes in response to hormone signaling. AR-low prostate cancer is a rising concern as this subtype is incurable and lethal. Understanding how cellular populations respond to hormone independence can help us define essential characteristics of AR-low proliferation/survival. By scRNAseq of mouse prostates, we find that an intermediate epithelial cell type expands in cancer and AR loss. This cell state is castration resistant and has multiple potential cellular origins which create a highly heterogeneous epithelial compartment. Cellular subsets in human metastatic cancers exhibit striking similarities to these intermediate cells, suggesting a conserved role in mediating cancer aggressiveness and treatment resistance. We also find a highly pro-tumorigenic immune environment, likely recruited via a signaling cascade involving epithelial cells and tumor-associated macrophages. In addition, we characterize the dependence of AR-low castration-resistant prostate cancer (CRPC) on high translation initiation and find that inhibiting eIF4F function results in cell cycle arrest and apoptosis. To investigate molecular mechanisms of translation regulation in this system, we examined transcripts upregulated by eIF4F activity in AR-low mouse prostates. We find a guanine-rich translational element (GRTE) enriched in the 5’ UTRs of these mRNAs. We test the role of the GRTE in regulating translation of the candidate gene KFL5 and find it is required to maintain high eIF4F-dependent translation. We also find a similar G-rich functional element in 5’ UTRs of mRNAs translationally regulated by an HRAS oncogenic variant and eIF2B5 in murine skin cancer. We then confirm that the GRTE is functional in human cells and conduct a genome-wide analysis to characterize its conservation. Finally, we generate stable reporter cell lines to facilitate future high-throughput experiments.