Identification of Mitochondrial Unfolded Protein Response-Related Genes and Diagnostic Biomarkers in Atherosclerosis by Integrative Multidimensional Analysis and Experimental Validation.

Background: Atherosclerosis (AS) is a common cardiovascular disease worldwide. The mitochondrial unfolded protein response (UPRmt) is a defense mechanism that enhances protein folding and degradation to maintain mitochondrial function and cellular homeostasis under stress. Research suggests a strong link between mitochondrial dysfunction and AS, particularly related to oxidative stress and inflammation. However, the exact relationship between UPRmt and AS is unclear. Identifying biomarkers associated with UPRmt is crucial for improving AS diagnosis and treatment. Methods: Microarray datasets related to AS were retrieved from the Gene Expression Omnibus (GEO) database. After integrating these datasets and eliminating batch effects, we obtained 101 AS and 67 control samples. Based on the expression levels of UPRmt-related genes (MRGs), the samples were classified into two subtypes and subjected to differential analysis, weighted correlation network analysis, and immune infiltration analysis. A predictive model was built using 12 machine learning algorithms to identify hub genes associated with UPRmt. Additionally, single-cell RNA-seq data and the CellChat algorithm were used to explore intercellular communication mechanisms mediated by these hub genes in AS. Mendelian randomization analysis was performed to identify biomarkers linked to AS. Molecular simulation techniques assessed the therapeutic potential of Iloprost. Finally, the expression and distribution of core genes were analyzed by RT-qPCR, Western blot, and immunofluorescence. Results: We identified seven hub genes at the intersection of UPRmt dysregulation and atherosclerosis. These genes showed consistent differential expression across cohorts and formed coherent mitochondria-stress modules. Their expression correlated with multiple immune-cell infiltration scores, including macrophage and T-cell signatures, and with inflammatory mediators. A classifier based on the seven-gene panel distinguished atherosclerotic from non-atherosclerotic samples across external datasets and remained robust after accounting for clinical covariates. Experimental assays confirmed altered expression of selected genes and their modulation under mitochondrial stress. Molecular simulation suggested that Iloprost can bind to the APOC1 protein's active pocket. Conclusion: ARHGAP25, CYTH4, ITGB7, APOC1, WDFY4, MARCO and PLCB2 are pivotal genes intimately linked to AS and the UPRmt. They potentially play crucial roles in mitochondrial dysfunction and immune regulation. As such, these genes may be promising biomarkers and therapeutic targets for AS.