Identification and validation of senescence-related genes in circulating endothelial cells of patients with acute myocardial infarction

Jie Xiang,Jun Shen,Ling Zhang,Baopeng Tang

DOI: https://doi.org/10.3389/fcvm.2022.1057985

2022-12-13

Abstract:Background: Acute myocardial infarction (AMI) is the main clinical cause of death and cardiovascular disease and thus has high rates of morbidity and mortality. The increase in cardiovascular disease with aging is partly the result of vascular endothelial cell senescence and associated vascular dysfunction. This study was performed to identify potential key cellular senescence-related genes (SRGs) as biomarkers for the diagnosis of AMI using bioinformatics. Methods: Using the CellAge database, we identified cellular SRGs. GSE66360 and GSE48060 for AMI patients and healthy controls and GSE19322 for mice were downloaded from the Gene Expression Omnibus (GEO) database. The GSE66360 dataset was divided into a training set and a validation set. The GSE48060 dataset was used as another validation set. The GSE19322 dataset was used to explore the evolution of the screened diagnostic markers in the dynamic process of AMI. Differentially expressed genes (DEGs) of AMI were identified from the GSE66360 training set. Differentially expressed senescence-related genes (DESRGs) selected from SRGs and DEGs were analyzed using Gene Ontology (GO) enrichment, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, and protein-protein interaction (PPI) networks. Hub genes in DESRGs were selected based on degree, and diagnostic genes were further screened by gene expression and receiver operating characteristic (ROC) curve. Finally, a miRNA-gene network of diagnostic genes was constructed and targeted drug prediction was performed. Results: A total of 520 DEGs were screened from the GSE66360 training set, and 279 SRGs were identified from the CellAge database. The overlapping DEGs and SRGs constituted 14 DESRGs, including 4 senescence suppressor genes and 10 senescence inducible genes. The top 10 hub genes, including FOS, MMP9, CEBPB, CDKN1A, CXCL1, ETS2, BCL6, SGK1, ZFP36, and IGFBP3, were screened. Furthermore, three diagnostic genes were identified: MMP9, ETS2, and BCL6. The ROC analysis showed that the respective area under the curves (AUCs) of MMP9, ETS2, and BCL6 were 0.786, 0.848, and 0.852 in the GSE66360 validation set and 0.708, 0.791, and 0.727 in the GSE48060 dataset. In the GSE19322 dataset, MMP9 (AUC, 0.888) and ETS2 (AUC, 0.929) had very high diagnostic values in the early stage of AMI. Finally, based on these three diagnostic genes, we found that drugs such as acetylcysteine and genistein may be targeted for the treatment of age-related AMI. Conclusion: The results of this study suggest that cellular SRGs might play an important role in AMI. MMP9, ETS2, and BCL6 have potential as specific biomarkers for the early diagnosis of AMI.