Abstract
Esophageal squamous cell carcinoma (ESCC) accounts for over 90% of all esophageal tumors. However, the molecular mechanism underlying ESCC development and prognosis remains unclear, and there are still no effective molecular biomarkers for diagnosing or predicting the clinical outcome of patients with ESCC. Here, using bioinformatics analyses, we attempted to identify potential biomarkers and therapeutic targets for ESCC. Differentially expressed genes (DEGs) between ESCC and normal esophageal tissue samples were obtained through comprehensive analysis of three publicly available gene expression profile datasets from the Gene Expression Omnibus database. The biological roles of the DEGs were identified by Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. Moreover, the Cytoscape 3.7.1 platform and subsidiary tools such as Molecular Complex Detection (MCODE) and CytoHubba were used to visualize the protein-protein interaction (PPI) network of the DEGs and identify hub genes. A total of 345 DEGs were identified between normal esophageal and ESCC samples, which were enriched in the KEGG pathways of the cell cycle, endocytosis, pancreatic secretion, and fatty acid metabolism. Two of the highest scoring models were selected from the PPI network using Molecular Complex Detection. Moreover, CytoHubba revealed 21 hub genes with a valuable influence on the progression of ESCC in these patients. Among these, the high expression levels of five genes-SPP1, SPARC, BGN, POSTN, and COL1A2-were associated with poor disease-free survival of ESCC patients, as indicated by survival analysis. Taken together, we identified that elevated expression of five hub genes, including SPP1, is associated with poor prognosis in ESCC patients, which may serve as potential prognostic biomarkers or therapeutic target for ESCC.