Author:
Jin Yi,Huang Siwei,Wang Zhanwang
Abstract
BackgroundMuscle-invasive bladder cancer (MIBC) develops lymph node (LN) metastasis or distant metastasis, leading to recurrence and poor prognosis. The five-year survival rate of MIBC with LN or distant metastasis is only 8.1%; therefore, there is an urgent need to identify reliable biomarkers for prognosis and treatment regimen for patients with bladder cancer (BLCA).MethodsSEER database was used to select important clinical characteristics for MIBC. Then, weighted gene co-expression network analysis (WGCNA) was employed to identify differentially expressed genes (DEGs) to recognize significant co-expression modules by calculating the correlation between the modules and clinical data. Furthermore, Cox regression and lasso analysis were applied to screen prognostic hub genes and establish the risk predictive model. Bladder cancer cell lines (UMUC3 and 5637) were used for experimental validation in vitro.ResultsCox analysis of 122,600 MIBC patients showed that the N stage was the most important clinical factor. A total of 4,597 DEGs were calculated between N0 and N+ patients, and WGCNA with these DEGs in 368 samples revealed that expression of turquoise was positively and strongly correlated with the N stage. Eight genes were identified as important prognostic candidates using lasso regression based on Cox analysis and STRING database. Combining GEO datasets, literature, and clinical factors, we identified LAMA2 and RUNX2 as novel prognostic biomarkers. CCK8 assay showed that depletion of LAMA2 or RUNX2 significantly inhibited the proliferation of BLCA cells, and flow cytometry indicated that knockdown of LAMA2 or RUNX2 induced the apoptosis of BLCA cells. Transwell assay also showed that silencing of LAMA2 or RUNX2 weakened the migration and invasiveness of BLCA cells.ConclusionsWe constructed a new eight-gene risk model to provide novel prognostic biomarkers and therapeutic targets for BLCA.