Comprehensive investigation of alternative splicing and development of a prognostic risk score for prostate cancer based on six-gene signatures

Zhe-Xu Cao; Guang-An Xiao; Wei Zhang; Jin Ji; Chen Ye; Dan Liu; Qin-Qin Tian; Ying-Hao Sun Prof

doi:10.7150/jca.31725

Comprehensive investigation of alternative splicing and development of a prognostic risk score for prostate cancer based on six-gene signatures

J Cancer. 2019 Sep 7;10(22):5585-5596. doi: 10.7150/jca.31725. eCollection 2019.

Authors

Zhe-Xu Cao^{1

2}, Guang-An Xiao^{1

2}, Wei Zhang^{1

2}, Jin Ji^{1

2}, Chen Ye^{1

2}, Dan Liu^{1

2}, Qin-Qin Tian^{1

2}, Ying-Hao Sun Prof^{1

2}

Affiliations

¹ Department of Urology, Shanghai Changhai Hospital, Naval Medical University (Second Military Medical University), Shanghai 200433, China.
² Shanghai Key Laboratory of Cell Engineering, Naval Medical University (Second Military Medical University), Shanghai 200433, China.

Abstract

Purpose: To systematically document alternative splicing profiles of prostate cancer in relatively large populations in order to construct a prognostic predictors model for prostate cancer. Methods: Splicing data and clinical information of 495 prostate cancer patients were obtained from The Cancer Genome Atlas (TCGA). The SpliceSeq database was used to extract information regarding splicing events. Multiple bioinformatic tools were used for functional and pathway enrichment analysis as well as for construction of gene interaction networks. Candidate gene expression profiles were verified with clinical samples using QRT-PCR. Results: We detected a total of 44070 alternative splicing events of 10381 genes in prostate cancer. 7 and 14 KEGG pathways were enriched and were associated with overall and recurrence-free survival, respectively. The expression of 396 genes among the 1526 overall survival genes associated alternative splicing events were associated with overall survival. The expression of 483 genes among the 1916 recurrence-free survival genes associated alternative splicing events were associated with recurrence-free survival. Lastly, we constructed the prognosis risk score system based on the expression profiles of six-gene signatures which in combination had an AUC of 0.941 for overall survival associated alternative splicing events, followed by overall survival associated gene expressions with an AUC of 0.794, a recurrence-free survival associated gene expression with an AUC of 0.752 and recurrence-free survival associated alternative splicing events with an AUC of 0.735, indicating its strong ability to predict patient outcome. The expression profile of the six genes was also confirmed in different prostate cell lines and clinic samples. Conclusion: Our comprehensive investigation of alternative splicing not only provided insight into the biological pathways of alternative splicing involved in the development of prostate cancer but also revealed new potential biomarkers for prognosticating as well as novel therapeutic targets for development of prostate cancer treatment.

Keywords: TCGA; alternative splicing; prognosis; prostate cancer.