Integrated bioinformatics, Mendelian randomization, and experimental validation identify IL-10 and MMP-19 as candidate regulators associated with thoracic aortic aneurysm

Objective To identify molecular candidates associated with thoracic aortic aneurysm (TAA) and evaluate convergent evidence for interleukin-10 (IL-10) and matrix metalloproteinase-19 (MMP-19) across transcriptomic, genetic, clinical-sample, and cell-based analyses. Methods Transcriptomic candidate discovery was performed using thoracic-aortic GEO datasets. GSE26155, containing both dilated and non-dilated thoracic-aortic samples, served as a tissue-level reference, while peripheral-blood GSE9106 provided supportive validation. Candidate genes were evaluated by functional enrichment, protein-protein interaction analysis, LASSO regression, and receiver operating characteristic (ROC) analysis. Expression levels were assessed by qRT-PCR in peripheral blood from 9 patients with TAA and 12 controls. Two-sample Mendelian randomization (MR) was conducted using European-ancestry GWAS data. Human aortic smooth muscle cells (HASMCs) exposed to angiotensin II (Ang II) were transfected with siRNAs targeting IL-10 or MMP-19 to assess cell viability, apoptosis, and extracellular matrix (ECM)-related markers. Results Thirty-one candidate genes were identified, and IL-10 and MMP-19 showed AUC values of 0.983 and 0.840, respectively, in the analyzed datasets. Peripheral-blood qRT-PCR showed higher IL-10 and MMP-19 expression in the small TAA cohort. At the nominal threshold, IVW MR estimates were positive for IL-10 (OR = 1.107, 95% CI 1.036-1.183; P = 0.003) and MMP-19 (OR = 1.233, 95% CI 1.052-1.445; P = 0.010); however, neither association remained statistically significant after false-discovery-rate correction (FDR-adjusted P = 0.054 and 0.231, respectively). In Ang II-treated HASMCs, knockdown of either target improved viability, reduced apoptosis-related changes, and partially restored extracellular matrix-related expression. Conclusion IL-10 and MMP-19 are convergent candidates associated with TAA across the analyzed datasets and experimental model. Because the supplementary cross-platform tissue comparison remains susceptible to dataset-phenotype confounding, the MR associations did not survive FDR correction, the clinical cohort was small, and the functional experiments were limited to an in vitro model, these findings are hypothesis-generating and require independent mechanistic and clinical validation.