Identification of salivary inflammatory non coding RNA biomarkers for coronary artery disease through epicardial...

Epicardial adipose tissue (EAT), due to its proximity to the myocardium, contributes inflammatory profile in CAD onset and progression. Emerging evidence highlights the role of EAT-derived non-coding RNAs in modulating disease pathways. However, identifying EAT-specific biomarkers remains challenging for routine diagnosis. Therefore, this study aimed to identify non-invasive, multi-source biomarkers from EAT, plasma, and saliva, and their regulatory miRNAs and lncRNAs, through network analysis, regulatory mapping, and experimental validation.

Differential expression analyses identified upregulated genes and miRNAs in EAT. Protein interaction network was constructed for overlapping upregulated genes and downregulated miRNA targets along with inflammatory-related genes. Clustering, ontology, and topological analyses were performed to pinpoint hub biomarkers and regulatory non-coding RNAs. Key candidate molecules were then experimentally validated using saliva and plasma samples from a cohort of CAD patients (n = 30) and healthy controls (n = 30).

A total of 840 upregulated and 1817 downregulated genes were identified in EAT with 36 upregulated and 82 downregulated miRNAs. Among 343 overexpressed inflammatory genes, CD8A, IL7R, and CCL5 emerged as hub biomarkers. Regulatory analysis uncovered 18 miRNAs and 548 lncRNAs influencing these hubs. Notably, hsa-miR-582-3p exhibited the highest number of lncRNA interactions and was significantly downregulated in saliva and plasma of CAD patients (p < 0.05).

Our study establishes a link between EAT inflammation and salivary expression profiles, highlighting hsa-miR-582-3p as a promising non-invasive salivary biomarker for CAD. This integrative approach provides a valuable foundation for developing multi-source diagnostics in cardiovascular clinical settings.