A sialyltransferases-related gene signature serves as a potential predictor of prognosis and therapeutic response for bladder cancer

Abstract

Abstract Background Aberrant glycosylation, catalyzed by the specific glycosyltransferase, is one of the dominant features of cancers. Among the glycosyltransferase subfamilies, sialyltransferases (SiaTs) are an essential part which has close linkages with tumor-associated events, such as tumor growth, metastasis and angiogenesis. Considering the relationship between SiaTs and cancer, the current study attempted to establish an effective prognostic model with SiaTs-related genes (SRGs) to predict patients’ outcome and therapeutic responsiveness of bladder cancer. Methods RNA-seq data, clinical information and genomic mutation data were downloaded (TCGA-BLCA and GSE13507 datasets). The comprehensive landscape of the 20 SiaTs was analyzed, and the differentially expressed SiaTs-related genes were screened with “DESeq2” R package. ConsensusClusterPlus was applied for clustering, following with survival analysis with Kaplan–Meier curve. The overall survival related SRGs were determined with univariate Cox proportional hazards regression analysis, and the least absolute shrinkage and selection operator (LASSO) regression analysis was performed to generate a SRGs-related prognostic model. The predictive value was estimated with Kaplan–Meier plot and the receiver operating characteristic (ROC) curve, which was further validated with the constructed nomogram and decision curve. Results In bladder cancer tissues, 17 out of the 20 SiaTs were differentially expressed with CNV changes and somatic mutations. Two SiaTs_Clusters were determined based on the expression of the 20 SiaTs, and two gene_Clusters were identified based on the expression of differentially expressed genes between SiaTs_Clusters. The SRGs-related prognostic model was generated with 7 key genes (CD109, TEAD4, FN1, TM4SF1, CDCA7L, ATOH8 and GZMA), and the accuracy for outcome prediction was validated with ROC curve and a constructed nomogram. The SRGs-related prognostic signature could separate patients into high- and low-risk group, where the high-risk group showed poorer outcome, more abundant immune infiltration, and higher expression of immune checkpoint genes. In addition, the risk score derived from the SRGs-related prognostic model could be utilized as a predictor to evaluate the responsiveness of patients to the medical therapies. Conclusions The SRGs-related prognostic signature could potentially aid in the prediction of the survival outcome and therapy response for patients with bladder cancer, contributing to the development of personalized treatment and appropriate medical decisions.