Screening key genes and biomarkers in gastrointestinal metaplasia that progress to gastric cancer by integrated bioinformatics analysis.

Abstract

This study aimed to screen the potential candidate genes and relevant biological markers associated with gastrointestinal metaplasia that progresses to gastric cancer (GIM-GC). Microarray datasets (GSE78523) were downloaded from the GEO database. Differentially expressed genes (DEGs) between GIM-GC samples and healthy controls were identified. GO and KEGG pathway enrichment analyses were performed. STRING and Cytoscape were used to identify significant module and hub genes. Survival analysis was applied to identify key genes. A Venn diagram was built to find hub DEGs that differed in all three relevant comparisons (GIM-GC vs. healthy controls vs. GIM-NoGC). The clinical characteristics of the hub DEGs were evaluated using the Cancer Genome Atlas dataset. The study found 257 DEGs (217 upregulated and 40 downregulated). The upregulated DEGs were enriched in regulation of microvillus length and phospholipid binding and were components of the apical plasma membrane. Downregulated DEGs were involved in digestion and hormone activity and were found in the extracellular space. Fat digestion and absorption as well as gastric acid secretion were the pathways enrichment. The most important gene modules related mainly to O-glycan processing, extracellular exosome, hormone activity, and vitamin and fat digestion and absorption. Eleven hub genes were identified, of which APOB, FABP1, CDX2, GCG, HNF4A, SLC26A3, CFTR, MUC5AC, OLFM4, and SI were related to the prognosis. Olfactomedin-4 (OLFM4) was the most relevant DEG to identify GIM-GC. In conclusion: DEGs and hub genes are helpful to understand the molecular mechanisms of GIM-GC. OLFM4 may be a biological marker for GIM-GC.