Identification of prognostic and bone metastatic alternative splicing signatures in bladder cancer

Abstract

Background ： Bladder cancer, originating from the epithelium of the urinary bladder, was the second most common malignancy in the urinary system with a high metastasis rate and poor post-metastasis prognosis. Alternative splicing events (ASEs) were regarded as important markers of tumor progression and prognosis, however, their roles in bladder cancer bone metastasis haven’t been recognized. Methods ： In order to explore the mechanism of ASEs in bladder cancer bone metastasis, we downloaded the RNA sequencing data and ASEs data of 412 samples of primary BLCAs from The Cancer Genome Atlas (TCGA) and TCGASpliceSeq databases. The Cox regression analysis was used to identify overall survival-related ASEs (OS-SEs), then, based on the OS-SEs screened by Lasso regression, we constructed the predict model. Finally, univariate and multivariate independent prognostic analysis were performed to prove it as an independent prognostic factor. Results ： In this study, a predict model of OS in BLCA was constructed and the Area Under Curve of the model was 0.581. Its risk score was also proved to be an independent predictor with the good accuracy (P < 0.001). Among identified 390 SFs, Junction plakoglobin (JUP) was significantly correlated with overall survival and bone metastasis. In co-expression analysis, the co-expression pathway of ITGB4 was the glycosphingolipid biosynthesis ganglio series. Conclusions ： We speculated that JUP regulating the ITGB4 might play a key role in bone metastasis of bladder cancer through the glycosphingolipid biosynthesis ganglio series pathway (R = 0.220, P < 0.001), which was also related to prognosis.