In Frontiers in physiology
Background: Calcific aortic valve disease (CAVD) is the most common valvular heart disease in the aging population, resulting in a significant health and economic burden worldwide, but its underlying diagnostic biomarkers and pathophysiological mechanisms are not fully understood. Methods: Three publicly available gene expression profiles (GSE12644, GSE51472, and GSE77287) from human Calcific aortic valve disease (CAVD) and normal aortic valve samples were downloaded from the Gene Expression Omnibus database for combined analysis. R software was used to identify differentially expressed genes (DEGs) and conduct functional investigations. Two machine learning algorithms, least absolute shrinkage and selection operator (LASSO) and support vector machine-recursive feature elimination (SVM-RFE), were applied to identify key feature genes as potential biomarkers for Calcific aortic valve disease (CAVD). Receiver operating characteristic (ROC) curves were used to evaluate the discriminatory ability of key genes. The CIBERSORT deconvolution algorithm was used to determine differential immune cell infiltration and the relationship between key genes and immune cell types. Finally, the Expression level and diagnostic ability of the identified biomarkers were further validated in an external dataset (GSE83453), a single-cell sequencing dataset (SRP222100), and immunohistochemical staining of human clinical tissue samples, respectively. Results: In total, 34 identified DEGs included 21 upregulated and 13 downregulated genes. DEGs were mainly involved in immune-related pathways such as leukocyte migration, granulocyte chemotaxis, cytokine activity, and IL-17 signaling. The machine learning algorithm identified SCG2 and CCL19 as key feature genes [area under the ROC curve (AUC) = 0.940 and 0.913, respectively; validation AUC = 0.917 and 0.903, respectively]. CIBERSORT analysis indicated that the proportion of immune cells in Calcific aortic valve disease (CAVD) was different from that in normal aortic valve tissues, specifically M2 and M0 macrophages. Key genes SCG2 and CCL19 were significantly positively correlated with M0 macrophages. Single-cell sequencing analysis and immunohistochemical staining of human aortic valve tissue samples showed that SCG2 and CCL19 were increased in Calcific aortic valve disease (CAVD) valves. Conclusion: SCG2 and CCL19 are potential novel biomarkers of Calcific aortic valve disease (CAVD) and may play important roles in the biological process of Calcific aortic valve disease (CAVD). Our findings advance understanding of the underlying mechanisms of Calcific aortic valve disease (CAVD) pathogenesis and provide valuable information for future research into novel diagnostic and immunotherapeutic targets for Calcific aortic valve disease (CAVD).
Lv Xiaoshuo, Wang Xiaohui, Liu Jingwen, Wang Feng, Sun Mingsheng, Fan Xueqiang, Ye Zhidong, Liu Peng, Wen Jianyan
2022
LASSO, SVM-RFE, bioinformatics, calcific aortic valve disease, immune cell infiltration, machine learning
