AF represents one of the most prevalent clinical cardiac arrhythmias, characterized by irregular, rapid, and disorganized electrical activity (11). Severe AF can result in compromised cardiac pumping function and hemodynamic disorders (12). AF is particularly prevalent in individuals aged 60 and above, with its incidence progressively rising with advancing age (13, 14). As the global population ages, the prevalence of AF continues to escalate. Additionally, AF is closely associated with a lot of risk factors, including hypertension, diabetes mellitus, obesity, alcohol consumption, and smoking (15, 16). However, the pathophysiological mechanisms underlying AF still needs further investigation. Existing research has indicated that atria in AF patients typically exhibits pronounced electrophysiological instability, characterized by marked alterations in atrial myocyte excitability and repolarization times (17, 18). This instability precipitates disordered conduction of electrical signals within the atria, thereby fostering the onset of AF (19). Furthermore, investigators have identified structural changes within the atria, such as atrial enlargement or fibrosis, in part of AF patients, which further exacerbate the severity of the arrhythmia (20, 21).
In recent years, a large of studies have demonstrated the pivotal role of genetic susceptibility in AF (22). Alterations in gene expression profiles can give rise to abnormal ion channel function, thereby perturbing normal cardiac electrophysiology (23). Furthermore, changes in gene expression profiles during structural remodeling also assume significance in AF (24). Profound comprehension of the genetic susceptibility to AF helps the tailoring of personalized therapeutic regimens, encompassing pharmacologic selection and surgical interventions for patients with AF (22). Moreover, better understanding of the alterations in gene expression profiles in AF patients facilitates the exploration of gene therapy approaches for AF (25). It is noteworthy that numerous investigators have adopted publicly available databases to elucidate novel biomarkers for AF, advancing the diagnosis and therapeutic strategies (26). Chu et al. conducted an analysis investigating the association between non-alcoholic fatty liver disease (NAFLD) and AF, employing bioinformatics methodologies to analyze differential gene expression profiles from datasets obtained from the GEO database (27). Their analysis identified 45 differentially expressed genes in both NAFLD and AF, and functional enrichment analysis found the pivotal roles of these shared differentially expressed genes in immune response and cytokine pathways (28).
Single-cell sequencing is a groundbreaking high-throughput sequencing technology that has significantly advanced our understanding of gene expression within distinct cellular subpopulations, unveiling heterogeneity and variability among different cells in a certain disease condition (29). This technology enables researchers to better understand the functionality individual cell subgroups. Single-cell sequencing has already been extensively used in various fields, including stem cell biology, cell differentiation, tumor heterogeneity, immunology, and developmental biology (30). Despite its relatively high cost, an increasing number of researchers are sharing single-cell sequencing data in public biological databases, facilitating its utility as a powerful research tool (31). In a study by Liu et al., the heterogeneity within peripheral blood mononuclear cells (PBMCs) from patients with rheumatoid arthritis (RA) was analyzed using single-cell sequencing. The investigators conducted sequencing on 10,483 cells, revealing distinct PBMC subpopulations. Subgroup analysis of T cells identified differentially expressed genes (DEGs), and validation in external datasets confirmed nine candidate genes highly associated with RA (CD8A, CCL5, GZMB, NKG7, PRF1, GZMH, CCR7, GZMK, GZMA), offering potential biomarkers for RA diagnosis and treatment (32). In another study by Xie et al., single-cell sequencing data from the GEO database was used to unveil 8 major cell types and 25 subgroups of colorectal cancer (CRC). Substantial differences were observed in metabolic patterns, immune phenotypes, and transcription factor regulation among subgroups within different major cell types. The investigators also identified biomarkers regulating lipid metabolism and immune-suppressive ligand-receptor pairs, leading to the construction of robust immunological risk models and clinical risk models for CRC prognosis. However, research employing single-cell sequencing for in-depth analysis of atrial fibrillation gene expression profiles remains limited (33).
In the present study, single-cell sequencing data from patients with AF were obtained from the GEO database. Subsequently, all cellular entities derived from AF samples were subjected to unsupervised clustering analysis, resulting in the categorization of these cells into 16 distinct clusters. These clusters were subsequently assigned to five distinct cell categories. Notably, our investigation involved the examination of cell-to-cell communication patterns. Our findings revealed that tissue stem cells exhibited the highest degree of cellular activity and engaged in more frequent interactions with other cell types. Tissue stem cells represent a class of stem cells characterized by self-renewal capacity and the ability to differentiate into various cell lineages to meet tissue demands (34, 35). Tissue stem cells are distributed throughout diverse tissues and organs, encompassing bone marrow, skin, muscle, nervous tissue, liver, lung, gastrointestinal tract, and more, each exhibiting distinct characteristics and functions (36, 37). Tissue stem cells ensure the stability of their population by giving rise to new stem cells through symmetric or asymmetric division (38). This self-renewal ability enables them to persist long-term and contribute to the sustained maintenance and repair of tissues. The self-renewal and differentiation processes of tissue stem cells are rigorously regulated, encompassing intracellular signaling pathways and extracellular signals from the microenvironment. Due to their self-renewal and differentiation potential, tissue stem cells hold considerable promise in the field of medicine. Indeed, numerous studies have explored the role of tissue stem cells in cardiovascular diseases. Gu et al. utilized single-cell RNA sequencing technology to investigate the diversity of mesenchymal stem cells derived from peripheral vascular adipose tissue, revealing two distinct subpopulations of tissue stem cells. Additionally, the researchers identified miR-378a-3p as a potential potent regulator of metabolic reprogramming, serving as a potential therapeutic target for vascular regeneration (39). Given the limited regenerative capacity of myocardial cells, human induced pluripotent stem cells have emerged as a crucial research tool in the investigation of AF and other cardiac arrhythmias. However, the precise role of tissue stem cells in the pathogenesis and progression of AF remains inadequately understood. Our study, for the first time, reports single-cell sequencing results from AF patients, suggesting a close interplay between tissue stem cells and other cell subtypes. This discovery may provide valuable insights for identifying potential therapeutic targets in the progression of AF.
WGCNA, a widely employed bioinformatics research methodology, is utilized for the exploration of gene patterns and co-expression modules in high-throughput gene expression data (40). By constructing a co-expression network among genes, WGCNA clusters genes with similar expression patterns into modules, thereby aiding in the elucidation of biological relationships and functions among genes. In this study, alongside the analysis of single-cell data from AF patients, we also acquired bulk RNA-seq data from the GEO database and conducted a detailed investigation into the roles of WGCNA gene modules in the pathogenesis of AF. Furthermore, we intersected genes in the gene modules most closely associated with AF occurrence with marker genes of tissue stem cells. Ultimately, we identified ABTB2, NAV2, and RBFOX1 as novel biomarkers for AF. Additionally, we evaluated the immune cell infiltration in AF patients using the CIBERSORT algorithm, which has found extensive application in numerous studies. In a recent study by Jiang et al., this algorithm was used to select the most closely related biomarkers to sepsis by machine learning methods from gene expression profile obtained from the GEO database. Furthermore, the study conducted a detailed analysis of the immune microenvironment changes in sepsis and investigated the correlation between sepsis biomarker expression levels and immune cell infiltration (41). Yu et al. also conducted an analysis of biomarkers and immune cell infiltration in systemic lupus erythematosus (SLE) using data from the GEO database (42). In our study, we similarly analyzed the relationship between these three novel AF biomarkers, ABTB2, NAV2, and RBFOX1, and immune cell infiltration. We found that these biomarkers may influence changes in the immune microenvironment of AF patients, thereby playing a role in the development and progression of AF. Simultaneously, we conducted an analysis of transcription factors and miRNAs that may regulate these AF biomarkers, which holds significant importance in identifying potential therapeutic targets for AF treatment.