Identification of molecular marker associated with ovarian cancer prognosis using bioinformatics analysis and experiments

BACKGROUND: Ovarian cancer is one of the three major malignant tumors of the female reproductive system, and the mortality associated with ovarian cancer ranks first among gynecologic malignant tumors. The pathogenesis of ovarian cancer is not yet clearly defined but elucidating this process would be of great significance for clinical diagnosis, prevention, and treatment. For this study, we used bioinformatics to identify the key pathogenic genes and reveal the potential molecular mechanisms of ovarian cancer; we used immunohistochemistry to validate them. METHODS: We analyzed and integrated four gene expression profiles (GSE14407, GSE18520, GSE26712, and GSE54388), which were downloaded from the Gene Expression Omnibus (GEO) database, with the aim of obtaining a common differentially expressed gene (DEG). Then, we performed Gene Ontology (GO) analysis and Kyoto Encyclopedia of Gene and Genome (KEGG) pathway analysis using the Database for Annotation, Visualization, and Integrated Discovery (DAVID). We then established a protein-protein interaction (PPI) network of the DEGs through the Search Tool for the Retrieval of Interacting Genes (STRING) database and selected hub genes. Finally, survival analysis of the hub genes was performed using a Kmplotter online tool. RESULTS: A total of 226 DEGs were detected after the analysis of the four gene expression profiles; of these, 87 were upregulated genes and 139 were downregulated. GO analysis results showed that DEGs were significantly enriched in biological processes including the G2/M transition of the mitotic cell cycle, the apoptotic process, cell proliferation, blood coagulation, and positive regulation of the canonical Wnt signaling pathway. KEGG analysis results showed that DEGs were particularly enriched in the cell cycle, the p53 signaling pathway, the Wnt signaling pathway, the Ras signaling pathway, the Rap1 signaling pathway, and tyrosine metabolism. We selected 50 hub genes from the PPI network, which had 147 nodes and 655 edges, and 30 of them were associated with the prognosis of ovarian cancer. We performed immunohistochemistry on phosphoserine aminotransferase 1 (PSAT1). PSAT1 was highly expressed in cancer tissues, and its expression level was related to clinical stage and tissue differentiation in ovarian cancer. A Cox proportional risk model suggested that high expression of PSAT1 and late clinical stage were independent risk factors for survival and prognosis of ovarian cancer patients. CONCLUSION: The detection of DEGs using bioinformatics analysis might be crucial to understanding the pathogenesis of ovarian cancer, especially the molecular mechanisms of its development. The association between PSAT1 expression and the occurrence, development, and prognosis of ovarian cancer was further verified by immunohistochemistry. The PSAT1 expression can be used as a prognostic marker to provide a potential target for the diagnosis and treatment of ovarian cancer.