The role of dysregulation of mRNA alternative splicing (AS) in the development and progression of solid tumors remains to be defined. Here we describe the first comprehensive AS landscape in the spectrum of human prostate cancer (PCa) evolution. We find that the severity of splicing dysregulation correlates with disease progression and establish intron retention as a hallmark of PCa stemness and aggressiveness. Systematic interrogation of 274 splicing-regulatory genes (SRGs) uncovers prevalent genomic copy number variations (CNVs), leading to mis-expression of ~68% of SRGs during PCa development and progression. Consequently, many SRGs are prognostic. Surprisingly, androgen receptor controls a splicing program distinct from its transcriptional regulation. The spliceosome modulator, E7107, reverses cancer aggressiveness and inhibits castration-resistant PCa (CRPC) in xenograft and autochthonous PCa models. Altogether, our studies establish aberrant AS landscape caused by dysregulated SRGs as a hallmark of PCa aggressiveness and the spliceosome as a therapeutic vulnerability for CRPC.

Mitochondrial proteostasis, regulated by the mitochondrial unfolded protein response (UPR^mt), is crucial for maintenance of cellular functions and survival. Elevated oxidative and proteotoxic stress in mitochondria must be attenuated by the activation of a ubiquitous UPR^mt to promote prostate cancer (PCa) growth. Here we show that the 2 key components of the UPR^mt, heat shock protein 60 (HSP60, a mitochondrial chaperonin) and caseinolytic protease P (ClpP, a mitochondrial protease), were required for the development of advanced PCa. HSP60 regulated ClpP expression via c-Myc and physically interacted with ClpP to restore mitochondrial functions that promote cancer cell survival. HSP60 maintained the ATP-producing functions of mitochondria, which activated the β-catenin pathway and led to the upregulation of c-Myc. We identified a UPR^mt inhibitor that blocked HSP60’s interaction with ClpP and abrogated survival signaling without altering HSP60’s chaperonin function. Disruption of HSP60-ClpP interaction with the UPR^mt inhibitor triggered metabolic stress and impeded PCa-promoting signaling. Treatment with the UPR^mt inhibitor or genetic ablation of Hsp60 inhibited PCa growth and progression. Together, our findings demonstrate that the HSP60-ClpP–mediated UPR^mt is essential for prostate tumorigenesis and the HSP60-ClpP interaction represents a therapeutic vulnerability in PCa.

Accurate detection of somatic copy number variations (CNVs) is an essential part of cancer genome analysis, and plays an important role in oncotarget identifications. Next generation sequencing (NGS) holds the promise to revolutionize somatic CNV detection. In this review, we provide an overview of current analytic tools used for CNV detection in NGS-based cancer studies. We summarize the NGS data types used for CNV detection, decipher the principles for data preprocessing, segmentation, and interpretation, and discuss the challenges in somatic CNV detection. This review aims to provide a guide to the analytic tools used in NGS-based cancer CNV studies, and to discuss the important factors that researchers need to consider when analyzing NGS data for somatic CNV detections.

Significance: Reliable proteome-wide quantification in large biological cohorts is highly valuable for clinical and pharmaceutical research yet remains extremely challenging despite recent technical advancements. Specifically, elevated missing data levels and compromised quantitative quality are common issues for prevalent methods. Here, we describe an IonStar technique taking advantage of sensitive and selective MS1 ion current-base quantification via innovations in effective and reproducible quantitative feature generation. Compared with several label-free strategies, IonStar showed superior performance in large-cohort analysis, manifested by excellent accuracy/precision, extremely low missing data, and confident discovery of subtle protein changes. In a proof-of-concept study, we demonstrated that IonStar quantified >7,000 unique proteins in 100 brain samples with no missing data and excellent quantitative quality, which has not been achievable by existing methods.