An integrative multi-omics study to identify candidate DNA methylation biomarkers associated with gastric cancer prognosis

Abstract

Aberrant DNA methylation (DNAm) is the most well-defined epigenetic hallmark in gastric cancer (GC), which may be associated with a variety of risk factors exposure. In this study, leveraging the multi-omics data of Genome-wide association studies (GWAS), methylation quantitative trait locus (mQTL) and expression quantitative trait locus (eQTL) collected from The Cancer Genome Atlas (TCGA), Genotype-Tissue Expression (GTEx) Project and Gene Expression Omnibus (GEO) database, a joint analysis of cox proportional hazard regression and Summary-data-based Mendelian randomization (SMR) analysis were adapted to investigate the causal associations of DNAm, gene expression and the GC prognosis. The results showed the causal association of hypermethylation of cg16007185, down-regulation of TMX1, and poorer prognosis of GC patients. Mendelian randomization (MR) analysis revealed that exposure to PCB-99, a type of polychlorinated biphenyl, might lead to the hypermethylation of cg16007185. Mediation analysis showed the borderline mediation role of TMX1 in the association between cg16007185 and GC survival, with an indirect effect (IE) of 5.24% (P = 0.102). Weighted correlation network analysis (WGCNA) and enrichment analysis predicted that TMX1 was involved in the cell proliferation pathway. In vitro experiments validated that promoter hypomethylation could promote the TMX1 expression, which inhibited the proliferation and metastatic ability of GC cells. Overall, our results suggest that the hypermethylation of cg16007185, a result of PCB-99 exposure, may promote the poor prognosis of GC patients by decreasing the TMX1 expression.