An integrated ceRNA network identifies miR-375 as an upregulated miRNA playing a tumor suppressive role in aggressive prostate cancer.

Prostate cancer (PCa) remains a significant cause of morbidity and mortality among men worldwide. A number of genes have been implicated in prostate tumorigenesis, but the mechanisms underlying their dysregulation are still incompletely understood. Evidence has established the competing endogenous RNA (ceRNA) theory as a novel regulatory mechanism for post-transcriptional alterations. Yet, a comprehensive characterization of ceRNA network in PCa lacks. Here we utilize stringent in-silico methods to construct a large ceRNA network across different PCa stages, and provide experimental demonstration for the competing regulation among protumorigenic SEC23A, PHTF2, and their corresponding ceRNA pairs. Using machine learning, we establish a ceRNA-based signature (ceRNA_sig) predictive of androgen receptor (AR) activity, tumor aggressiveness, and patient outcomes. Importantly, we identify miR-375 as a key node in PCa ceRNA network, which is upregulated in PCa relative to normal tissues. Forced expression of miR-375 significantly inhibits, while its inhibition promotes, aggressive behaviors of both AR+ and AR- PCa cells in vitro and in vivo. Mechanistically, we show that miR-375 predominantly targets genes possessing oncogenic roles (e.g., proliferation, DNA repair, and metastasis), and thus release targets with tumor suppressive functions. This action model well clarifies why an upregulated miRNA plays a tumor suppressive role in PCa. Together, our study provides new insights into understanding of transcriptomic aberrations during PCa evolution, and nominates miR-375 as a potential therapeutic target for combating aggressive PCa.