Identification Of Clear Cell Renal Carcinoma Progression-related Key Genes Based On Expression Dynamics Analysis

BackgroundClear cell renal cell carcinoma(ccRCC)is the major subtype of renal malignancies.The diagnosis of ccRCC relies on computer tomography and magnetic resonance imaging.The five-year survival rate of patients with early diagnosis is about 90%,but the five-year survival rate of patients with advanced disease is only about 12%.At present,the specific prognostic biomarkers and classification markers of advanced ccRCC are still lacking,which has led to poor prognosis to a large extent.Therefore,screening and identification of driving molecular targets in the progression of ccRCC are critical to patient treatment and prognosis.In this study,we developed an expression dynamics analysis approach,combining multi-omics data of DNA methylation and copy number alteration analysis in order to identify key genes in progression of ccRCC.MethodsIntegrating transcriptome(n=1669),DNA methylation(n=577)and copy number data(n=832),we developed an approach to identify driver biomarkers by analyzing the omics-level dynamics of EpithelialMesenchymal Transition(EMT)-related genes in ccRCC.Independent training and validation cohorts were used to validate the clinical prognosis potential of identified biomarkers.ResultsWe first identified 504 expression dynamic changed genes,which are involved in key pathways associated with ccRCC,such as EMT,cell cycle,EGFR,and PI3 K / AKT signaling.Further analysis identified 229(90 gene promoters)abnormal expression quantitative trait methylation(eQTM)and 256 expression quantitative trait copy number(eQTCN)altered genes.Among them,FoxM1 is affected by both eQTM and eQTCN.FoxM1 copy number amplification(115/500,23% of patients)occurred in the amplification peak of chromosome 12p13.33,was enriched in patients with advanced ccRCC,and was associated with poor survival.Patients with distant metastasis of pT3 overexpressed by FoxM1 showed a reduction in overall survival of approximately 25% in both the training(log-rank P = 0.006)and validation(log-rank P = 0.018)cohorts.Better than individual gene expression signature or DNA methylation signature,eQTM-gene hybrid signature(cg00044170 and FoxM1)have great potential for early diagnosis of ccRCC in diagnostic training(area under the curve = 0.958)and validation data sets.ConclusionsFoxM1 may be a novel prognostic biomarker and shed light for early diagnosis of ccRCC in molecular level.