Bioinformatic identification of differentially expressed genes regulated by DNA-methylation in glioblastoma.

Abstract

DNA methylation-driven differentially expressed genes (DEGs) play potentially important roles in glioblastoma (GBM). In the present study, we applied bioinformatic analyses to identify key methylation-regulated DEGs (MeDEGs) in glioblastoma and elucidate their functions. Gene expression and methylation profile data from glioblastoma samples along with clinical information were obtained from GEO and TCGA databases. A total of 65 genes were identified as MeDEGs from the aforementioned data. Subsequently, gene ontology and kyoto encyclopaedia of genes and genomes enrichment analyses of these MeDEGs exhibited that MeDEGs were mostly enriched in several tumour-related terms such as 'activation of cysteine-type endopeptidase activity involved in apoptotic process' and 'phospholipid scrambling'. Kaplan-Meier survival analysis demonstrated significant correlation of CASP1, CFH and TTLL7 hyper-methylation with patient prognosis. Finally, CASP1 protein could indirectly interact with CFH protein, but interaction of TTLL7 protein with CASP1 or CFH protein was not evident. Based on gene set enrichment analysis, hyper-methylation of CASP1, CFH and TTLL7 were found enriched in tumour-related KEGG terms, such as 'RNA degradation', 'apyruvate metabolism' and 'nitrogen metabolism'. Methylation levels of CASP1, CFH and TTLL7 were addressed to negatively correlate with their mRNA levels in GBM cell lines. In sum, the present identification of MeDEGs associated with overall survival put forth potential molecular targets for translation towards improved diagnosis and treatment of GBM, and specifically, methylation levels of CASP1, CFH and TTLL7 genes could serve as key prognostic biomarkers in GBM.