Single-cell profiling of myocardial epicardial adipose tissue - insights into coronary artery disease pathophysiology

Introduction Epicardial adipose tissue (EAT) has emerged as a new player in cardiovascular disease (CVD). While pericoronary EAT has been analyzed at the single cell level, the cellular composition and specific cellular functions of myocardial EAT (mEAT), which directly interfaces with the myocardium, remain poorly understood. Purpose This study aims to characterize the single-cell landscape of mEAT in coronary artery disease (CAD) patients to elucidate the roles of distinct cell populations in disease progression. Methods Single-nucleus RNA sequencing was performed on 21 mEAT samples from CAD patients undergoing coronary artery bypass grafting (n=10) and non-CAD patients undergoing surgery for aortic disease (n=9), along with forensic controls without CVD (n=2). Bioinformatic analysis was conducted using the R-based package "Harmony," yielding over 121,000 analyzed nuclei. Differentially expressed genes per cell cluster were analyzed, followed by gene set enrichment analysis (GSEA). Cell-cell interactions were assessed using the "CellphoneDB" package. Histological analyses of cryosections from CAD (n = 7) and non-CAD (n = 8) patients included adipocyte size measurements and Movat’s pentachrome staining to assess fibrosis. Results mEAT was composed of adipocytes (ADPC), stromal cells (SC), immune cells, vascular and mural cells, glial cells, and mesothelial cells (MC). Notably, two SC-associated cell clusters, initially classified as unidentified and exclusive to CAD patients, exhibited inflammatory signatures. GSEA identified upregulated pathways such as "TNFA signaling via NFKB" or "inflammatory response". One unidentified cluster displayed increased stromal (LAMA2, PDGFRA) and mesothelial (C3) markers and was classified as inflammatory stromal cells (ISCs). The second unidentified cell cluster expressed adipocyte-associated markers (PLIN4, PLIN5), suggesting a transitional pre-adipogenic state (inflammatory pre-adipogenic cells, IPACs). Additionally, an ADPC subcluster and a macrophage subcluster, both primarily detected in CAD patients, also exhibited inflammatory signatures. Histological analysis of mEAT confirmed adipocyte hypertrophy (p<0.001) and increased fibrosis (p<0.01) in CAD-mEAT, aligning with inflammation-related patterns observed in snRNA-seq data. Cell-cell communication analysis revealed enhanced interactions in CAD patients, particularly via THBS- and SEMA4-signaling, which are implicated in inflammation and fibrotic remodeling. Conclusions This study provides a comprehensive single-cell characterization of mEAT, revealing inflammation-driven cellular changes and altered cell-cell communication in CAD patients. Notably, we identified inflammation-associated cell populations and sub-populations that were exclusively present in CAD patients. These findings highlight potential mechanisms by which mEAT might contribute to CAD pathophysiology and suggest novel cellular targets for therapeutic intervention.