Identification and verification of diagnostic biomarkers related to endoplasmic reticulum stress for atherosclerosis

Abstract Background Endoplasmic reticulum stress (ERS) is linked to the progression of atherosclerosis (AS). The function of ERS-related genes (ERSRGs) in AS remains ambiguous, hence, this study sought to examined their association with AS. Methods Datasets relevant to AS were acquired from the Gene Expression Omnibus (GEO) database. Initially, differentially expressed genes (DEGs) were identified between AS and controls. Weighted gene co-expression network analysis (WGCNA) derived AS-related module genes. The intersection of DEGs, AS-related module genes, and ERSRGs was subsequently evaluated by three machine learning algorithms. After processing expression analysis and receiver operating characteristic (ROC) curve evaluation, biomarkers were ultimately found. A nomogram was developed and validated to predict the incidence of AS based on biomarkers. Finally, single cell RNA sequencing (scRNA-seq) analysis was utilized to identify distinct cell subpopulations, followed by cell communication and pseudotime analysis. Results ANKRD1, BDNF, HLA-B, NLRP3, NOD2, and XAF1 were recognized as biomarkers for AS. The nomogram developed using biomarkers demonstrated exceptional efficacy in predicting AS occurrence. The scRNA-seq analysis annotated 6 cell subpopulations, including NK cells, B cells, T cells, vascular smooth muscle cells (VSMCs), endothelial, and macrophages. VSMCs exhibited increased communication in AS samples compared to control samples. Moreover, we noted that HLA-B exhibited elevated expression during the prophase and anaphase of VSMCs differentiation, whereas XAF1 had increased expression in the later stages of VSMCs differentiation. Conclusion This study found 6 biomarkers (ANKRD1, BDNF, HLA-B, NLRP3, NOD2, and XAF1) associated with ERS in AS, providing novel diagnostic targets for the condition.