Identifying new driver genes for castration resistance for prostate cancer based on protein mass spectrometry and transcriptome big data

Abstract

Purpose Androgen is important in the formation, development and progress of prostate cancer. Deprivation of androgen, also named castration, is an effective way in prostate cancer therapies. For patients undergoing castration therapies, castration was gradually becoming ineffective in prostate cancer treatment and therefore resulting in castration resistance for prostate cancer, which is beyond efficient therapies. Moreover, the mechanism on how castration resistance was originated and what promoted castration resistance was rarely reported. We designed the project to further understand the mechanism for castration resistance. Methods AR-V7 and PRPF6 were selected to be calibration markers and then performed with protein mass spectrometry to detect AR-V7 and PRPF6 binding proteins. Afterwards, we used transcriptome big data from prostate cancer and castration-resistant prostate cancer and then applied combined analysis of association rules method, gene co-expression analysis, linear matrix migration method and functional deep filtering on identifying candidate driver genes which were causally correlated with castration resistance. The predicted driver genes were intersected with protein mass spectrometry filtering results for both AR-V7 and PRPF6. Afterwards, the predicted driver genes were theoretically and experimentally validated. Results The 5 candidate driver genes RPL28, RPS4X, RPL11, RPL23A and RPSA were deemed as driver genes for castration resistance and were validated on being highly correlated and playing important role in the progress of castration resistance. Conclusion The new driver genes might assist on further understanding the mechanism and bringing new insight on the prevention and therapy for castration resistance on prostate cancer.