Abstract / Summary
Background: Pulmonary arterial hypertension (PAH) is a severe cardiovascular disease, with early diagnosis being difficult. Ion homeostasis imbalance contributes to PAH pathogenesis, but the systemic regulatory network involving ion homeostasis-related genes (IHRGs) and their diagnostic potential remain unclear.
Methods: This study utilized WGCNA and machine learning to identify potential PAH diagnostic biomarkers. WGCNA was applied to GSE117261 to identify IHRGs associated with PAH, followed by functional analysis via GO and KEGG pathways. Three machine learning algorithms-LASSO, SVM-RFE, and Boruta-were used to screen core genes, validated in GSE117261 and GSE113439 datasets. Additionally, bioinformatics analysis explored the genes' roles in PAH's immune microenvironment and regulatory mechanisms, complemented by in vitro hypoxia experiments validating the protective effects of S1PR1 and CA2 on endothelial cells.
Results: WGCNA identified 55 IHRGs enriched in ion homeostasis and immune pathways. Machine learning cross-validation narrowed down 9 core genes, with ABCG2, CA2, and S1PR1 showing strong diagnostic performance. In vitro experiments further confirmed that CA2 expression was significantly upregulated under hypoxia, whereas S1PR1 was downregulated. S1PR1 overexpression effectively suppressed endothelial cell apoptosis, mitigated inflammatory responses, and enhanced tight junction protein expression, while CA2 knockdown produced opposing effects.
Conclusion: This study identified ABCG2, CA2, and S1PR1 as potential diagnostic biomarkers for PAH, suggesting that IHRGs contribute to disease progression through modulating ion homeostasis and endothelial function. Among these, S1PR1 exhibited a definitive endothelial protective role, whereas CA2 may participate in hypoxia-induced endothelial adaptation, providing novel candidate targets for mechanistic research and early diagnosis of PAH.