Exploring biomarkers for autophagy-mediated macrophage pyroptosis in atherosclerosis.

Abstract

This study tried to investigate the macrophage autophagy-related pyroptosis in atherosclerosis. The gene expression omnibus (GEO) dataset of GSE100927 was used for differentially expressed genes (DEG) screening, gene ontology (GO) and kyoto encyclopedia of genes and genomes (KEGG), CIBERSORT, weighted correlation network analysis (WGCNA), receiver operating characteristic (ROC), gene set enrichment analysis (GSEA), and correlation analysis, and GSE159677 was used for single-cell analysis, all conducted in R software. Protein-protein interaction (PPI) was constructed in STRING and analyzed in Cytoscape. Transcription factors, drugs, and tissue co-expression network were explored in NetworkAnalyst. A total of 110 autophagy-related DEG (DEATG) were identified, and GO/KEGG revealed the top items enriched in autophagy, phagosome and lysosome. CIBERSORT showed 11 cell types were markedly differentially expressed (p < .05). WGCNA found the turquoise and yellow module were positively correlated with macrophage M0 (corr = 0.5, P = 6e-6) and M2 (corr = 0.54, P = 1e-6), respectively. Then 35 immune-related DEATG were identified, and functional analysis showed immune effector process, interleukin-6 and myeloid cell activation were enriched besides autophagy. PPI and MCC algorithm identified 6 hub genes in regulating macrophage-related autophagy, and ROC indicated high prediction value (area under curve = 0.961). GSEA enriched 6 common pathways associated with autophagy and atherosclerosis pathogenesis, and immune correlation suggested these hub genes were correlated with macrophages M0/M1, monocytes and T cells. Then venn plot found 3 central genes in mediating macrophage autophagy-associated pyroptosis in atherosclerosis, and single-cell analysis demonstrated cell distribution, then validated in THPA human samples. Our data discovered hub genes responsible for macrophage autophagy-mediated pyroptosis in atherosclerosis, and functional analysis with immune cell distribution evidenced their high phenotype-trait prediction value.