J Clin Lab Anal. 2026 Apr 21:e70200. doi: 10.1002/jcla.70200. Online ahead of print.
ABSTRACT
BACKGROUND: Thoracic aortic aneurysm (TAA) is a life-threatening disease with limited therapeutic options, largely due to an incomplete understanding of its pathogenic mechanisms. Inflammation and immune dysregulation are central to TAA progression, yet the specific pathogenic cell subsets and signaling pathways remain unclear.
METHODS: Single-cell RNA sequencing was performed on aortic tissues from four healthy individuals and four TAA patients. Key immune cell populations were identified through clustering, macrophage subsetting, and M1/M2 polarization scoring. Machine learning (Random Forest and Boruta) across multiple bulk RNA-seq cohorts was used to identify robust biomarkers. Cell-cell communication was analyzed using CellChat and NewCCI. Molecular docking of 2988 FDA-approved drugs was conducted to screen potential CLEC5A inhibitors, followed by cellular validation experiments.
RESULTS: A total of 105,541 cells were profiled, revealing significant alterations in macrophage, dendritic cell, and NK cell proportions in TAA. CLEC5A+ macrophages emerged as the predominant pathogenic subset, exhibiting the highest M1 polarization score. Machine learning consistently identified CLEC5A and SPP1 as stable diagnostic biomarkers across four cohorts (AUC > 0.7). Cell-cell communication revealed strong CCL5-CLEC5A signaling from T/NK cells to macrophages. Molecular docking identified Dioscin as a potential CLEC5A inhibitor. Cellular experiments showed that CCL5-induced M1 polarization was CLEC5A-dependent and could be effectively blocked by Dioscin.
CONCLUSIONS: CLEC5A+M1-polarized macrophages are key pathogenic cells in TAA. CLEC5A serves as a robust biomarker, and targeting the CCL5-CLEC5A axis with Dioscin represents a promising therapeutic strategy.
PMID:42011731 | DOI:10.1002/jcla.70200