Identification of differentially expressed mRNAs as novel predictive biomarkers for gastric cancer diagnosis and prognosis
Jian-Wei Zhou,Yi-Bing Zhang,Zhi-Yang Huang,Yu-Ping Yuan,Jie Jin
DOI: https://doi.org/10.4251/wjgo.v16.i5.1947
2024-05-08
World Journal of Gastrointestinal Oncology
Abstract:BACKGROUND Gastric cancer (GC) has a high mortality rate worldwide. Despite significant progress in GC diagnosis and treatment, the prognosis for affected patients still remains unfavorable. AIM To identify important candidate genes related to the development of GC and identify potential pathogenic mechanisms through comprehensive bioinformatics analysis. METHODS The Gene Expression Omnibus database was used to obtain the GSE183136 dataset, which includes a total of 135 GC samples. The limma package in R software was employed to identify differentially expressed genes (DEGs). Thereafter, enrichment analyses of Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were performed for the gene modules using the clusterProfile package in R software. The protein-protein interaction (PPI) networks of target genes were constructed using STRING and visualized by Cytoscape software. The common hub genes that emerged in the cohort of DEGs that was retrieved from the GEPIA database were then screened using a Venn Diagram. The expression levels of these overlapping genes in stomach adenocarcinoma samples and non-tumor samples and their association with prognosis in GC patients were also obtained from the GEPIA database and Kaplan-Meier curves. Moreover, real-time quantitative polymerase chain reaction (RT-qPCR) and western blotting were performed to determine the mRNA and protein levels of glutamic-pyruvic transaminase (GPT ) in GC and normal immortalized cell lines. In addition, cell viability, cell cycle distribution, migration and invasion were evaluated by cell counting kit-8, flow cytometry and transwell assays. Furthermore, we also conducted a retrospective analysis on 70 GC patients diagnosed and surgically treated in Wenzhou Central Hospital, Dingli Clinical College of Wenzhou Medical University, The Second Affiliated Hospital of Shanghai University between January 2017 to December 2020. The tumor and adjacent normal samples were collected from the patients to determine the potential association between the expression level of GPT and the clinical as well as pathological features of GC patients. RESULTS We selected 19214 genes from the GSE183136 dataset, among which there were 250 downregulated genes and 401 upregulated genes in the tumor samples of stage III-IV in comparison to those in tumor samples of stage I-II with a P -value < 0.05. In addition, GO and KEGG results revealed that the various upregulated DEGs were mainly enriched in plasma membrane and neuroactive ligand-receptor interaction, whereas the downregulated DEGs were primarily enriched in cytosol and pancreatic secretion, vascular smooth muscle contraction and biosynthesis of the different cofactors. Furthermore, PPI networks were constructed based on the various upregulated and downregulated genes, and there were a total 15 upregulated and 10 downregulated hub genes. After a comprehensive analysis, several hub genes, including runt-related transcription factor 2 (RUNX2 ), salmonella pathogenicity island 1 (SPI1 ), lysyl oxidase (LOX ), fibrillin 1 (FBN1 ) and GPT , displayed prognostic values. Interestingly, it was observed that GPT was downregulated in GC cells and its upregulation could suppress the malignant phenotypes of GC cells. Furthermore, the expression level of GPT was found to be associated with age, lymph node metastasis, pathological staging and distant metastasis (P < 0.05). CONCLUSION RUNX2 , SPI1 , LOX , FBN1 and GPT were identified key hub genes in GC by bioinformatics analysis. GPT was significantly associated with the prognosis of GC, and its upregulation can effectively inhibit the proliferative, migrative and invasive capabilities of GC cells.
oncology,gastroenterology & hepatology