Coronary artery disease (CAD) is a major global cause of mortality, with atherosclerosis being its fundamental pathological change. The primary mechanisms contributing to CAD are inflammation and the development of atherosclerotic plaques [1]. Recent studies have highlighted the significant contribution of epicardial adipose tissue (EAT) sharing the same microcirculation as the coronary artery in the progression, onset, and prognosis of CAD. The EAT which is located near the myocardium exerts a distinct impact on CAD [2]. EAT is primarily nourished by the coronary artery and is considered a unique adipose tissue depot. In addition to storing energy, EAT functions as both an immune and endocrine organ. It is mainly composed of stromal-vascular cells, fibroblasts, adipocytes, nerves, and immune cells [3]. Under normal conditions, EAT offers mechanical protection to the coronary arteries, prevents cardiac lipotoxicity, and provides immune support [4]. However, growing evidence highlights the connection between EAT inflammation and CAD pathogenesis [5]. Also, it is now widely accepted that certain cytokines released by EAT impair coronary arteries, atrial fibrillation and cardiomyocyte function through vasocrine or paracrine pathways. Also, EAT-related proatherogenic or proinflammatory factors induce atherogenic changes and vascular damage in endothelial cells and monocytes [2,3]. Therefore, EAT serves as a potential biomarker and therapeutic target for CAD.

Alternatively, several non-coding RNAs (ncRNAs) released from EAT such as long non-coding RNAs (lncRNAs) and microRNAs (miRNAs), may influence atrial remodelling by passively diffusing into the adjacent myocardium [6,7]. The lncRNAs represent a novel class of ncRNAs with more than 200 nucleotides and are highly associated with cardiovascular disorders [8]. Likewise, the small non-coding miRNA levels have a significant impact on biological processes, and the development of disease conditions [9]. Previously, studies such as Deng et al., 2022 have identified differentially expressed biomarkers like TCF21, CDH19, XG, and NNAT in EAT from CAD patients [3]. Additionally, Tan et al., 2020 have listed 10 biomarker genes, including CCND1, HP, HOXB7, HOXC8, COL1A1, CCL2, HOXA5, HOXB5, HOXC6, and TWIST1, which distinguish EAT from subcutaneous adipose tissue (SAT) in CAD patients [5]. Despite advancements, there remains a pressing need for multi-source, non-invasive biomarkers to enhance the sensitivity and specificity of disease diagnosis. In 2021, Bahbah et al. [10] reported several salivary biomarkers such as creatine kinase-MB, C-reactive protein, troponin I, and myoglobin for cardiovascular conditions. However, their potential inflammatory roles and regulatory mechanisms via ncRNAs have not been explored. Thus, identifying inflammatory biomarkers and their regulatory ncRNAs, which are potentially detectable in both plasma and saliva, is critical for developing non-invasive, multi-source diagnostic strategies, particularly for EAT-specific CAD pathogenesis. In recent decades, bioinformatics has been used widely to identify dysregulated genes making it a valuable strategy for uncovering molecular mechanisms and candidate biomarkers related to CAD [11].

Herein, we aimed to identify over-expressed genes in EAT, explore their regulatory miRNAs and lncRNAs, and cross-validate their expression levels in plasma and saliva for identifying multi-source biomarkers (Fig. 1). The differential gene expression (DGE) and differential miRNA expression (DEM) analyses revealed both upregulated and downregulated genes and miRNAs in EAT linked to CAD progression. Additionally, the target genes of downregulated miRNAs and the upregulated genes were mapped against a list of inflammatory genes. The overlapping genes were used to frame a protein level interactive network to explore their functional roles. Clustering and ontological analysis revealed key interconnected clusters involved in CAD progression. Topological analysis sorted by degree pinpointed hub inflammatory biomarkers in EAT-related CAD, and their corresponding suppressive miRNAs and lncRNAs were identified to uncover their regulatory significance. Additionally, the regulatory miRNAs associated with hub biomarkers were experimentally validated via RT-qPCR (reverse transcription-quantitative polymerase chain reaction) in clinical saliva and plasma samples.