Analysis of 5-Methylcytosine Regulators and DNA Methylation-Driven Genes in Colon Cancer

Abstract

Background: Epigenetic-driven events are important molecular mechanisms of carcinogenesis. The 5-methylcytosine (5mC) regulators play important roles in the methylation-driven gene expression. However, the effect of the 5mC regulators on the oncogenic pathways in colon cancer (CC) remains unclear. Also, the clinical value of such epigenetic-driven events needs further research. Methods: The transcriptome and matching epigenetic data were obtained from The Cancer Genome Atlas dataset. The gene set variation analysis identified the oncogenic pathways adjusted by 5mC regulators. The “edgeR” and “methylmix” package identified the differential expression genes of DNA methylation-driven genes. The correlation between 5mC regulators or transcription factors and shortlisted genes was investigated by calculating the Spearman's rank correlation coefficient. Among them, the genes related to diagnosis were screened out based on differential gene expression in extracellular vesicles (EVs) by the “limma” package and histology by immunohistochemistry. Then, a risk signature was constructed by fitting the generalized linear model and validated by the receiver operating characteristic curve. Results: MYC targets pathway and phosphatidylinositol-3-kinase–AKT–mammalian target of rapamycin signaling pathway were identified as the hallmark-related pathways associated with 5mC regulators. Also, the P53 pathway was subject to the influence of regulators' expression. A five methylation-driven gene signature (FIRRE, MYBL2, TGFBI, AXIN2, and SLC35D3) was developed as the biomarker for CC diagnosis. Meanwhile, those genes positively related to 5mC regulators and interacted with their relevant or transcription factors. Conclusion: In general, 5mC regulators are positively related to each other and DNA methylation-driven genes, with the relationship of multiple active and inhibitory pathways related to cancer. Meanwhile, the signature (FIRRE, MYBL2, TGFBI, AXIN2, and SLC35D3) can prefigure prospective diagnosis in CC.