Identification and validation of five hypoxia- and immune-related hub targets in atrial fibrillation Authors: Jinri Weng, Dongdong Huang, Qile Wu, Zhilong Chen, Jinzao Chen, Zhihua Pang Published: 2026 ABSTRACT BackgroundAtrial fibrillation (AF) is the most common arrhythmia, with increasing incidence and prevalence. Hypoxia and immune response are closely related to the progression of AF. This study aims to explore hub hypoxia- and immune-related genes associated with AF.MethodsMultiple AF datasets from the Gene Expression Omnibus database were downloaded. Hypoxia- and immune-related genes associated with AF were identified by integrating differential expression analysis, weighted gene co-expression network analysis (WGCNA), and protein-protein interaction (PPI) network. Functional enrichment analysis was used to evaluate the functions of key genes. Immune cell infiltration analysis was conducted using CIBERSORT and xCell. To experimentally validate the identified hub genes, RT-qPCR was performed on whole blood samples collected from healthy controls and AF patients.ResultsA total of 13 crucial hypoxia- and immune-related genes associated with AF were identified. Among them, the expressions of IL6, JUN, FOS, FOSL2, and MAFF were significantly higher in the AF group than in the sinus rhythm (SR) group, which was further validated by RT-qPCR showing increased mRNA expression in clinical whole blood samples from AF patients. Notably, transcription factors JUN, FOS, FOSL2, and MAFF exhibited clear binding peaks at promoter region of IL6. Functional enrichment analysis by GSEA revealed that these five hub genes were simultaneously involved in the cAMP signaling pathway, and four of them (JUN, FOS, FOSL2, and MAFF) were also associated with the calcium signaling pathway, indicating a potential cAMP-dependent Ca²⁺ handling mechanism in AF pathogenesis. Immune cell infiltration analysis further showed that these genes were significantly correlated with differential infiltration of various immune cells, including gamma delta T cells, monocytes, macrophages, neutrophils, and NK cells.ConclusionThis study successfully identified JUN, FOS, IL6, FOSL2, and MAFF as crucial hypoxia- and immune-related genes associated with AF, providing potential new targets for the treatment of AF.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13019-026-04369-2. Introduction Atrial fibrillation (AF) is the most common arrhythmia. As an age-related condition, the incidence and prevalence of AF have been increasing annually due to the acceleration of population aging [1]. According to estimates by the Global Burden of Disease (GBD) 2019, the number of AF patients worldwide reached 59.7 million in 2019, which is double the figure reported in 1990 [2]. The pathological characteristic of AF is the loss of regular and orderly atrial electrical activity, replaced by rapid and disorderly fibrillation waves [3]. Currently, the treatment strategies for AF are mainly divided into two categories: rhythm control aimed at restoring and maintaining sinus rhythm (SR), and rate control aimed at regulating ventricular heart rate [4]. However, the available treatment options suffer from poor efficacy, potential cardiac and extracardiac side effects, and high recurrence rates following cardioversion or ablation [5–7]. Therefore, strengthening the understanding of AF mechanisms and searching for potential AF biomarkers is of great significance for the diagnosis and treatment of AF. More recently, the interaction between hypoxia and immunity in AF has attracted increasing attention. Extensive research has shown that myocardial hypoxia plays a crucial role in cardiac remodeling, which serves as a significant important substrate for the development of AF [8]. Compared to the sinus rhythm (SR) population, the expression levels of hypoxia markers, such as hypoxia-inducible factor (HIF)-1α and HIF-2α, are significantly upregulated in the atria of AF patients, accompanied by an increase in atrial fibrosis [9]. The role of the immune and inflammatory responses in AF has also received widespread attention [10]. Multiple pro-inflammatory cytokines, such as tumor necrosis factor alpha (TNF-α) and interleukin (IL)-6, are known to promote atrial fibrosis, thereby facilitating the occurrence and development of AF [11]. Meanwhile, hypoxia can influence the function of immunocytes. HIF has been shown to regulate inflammatory responses involving multiple immunocytes, including macrophages, neutrophils, regulatory T cells (Tregs), T helper 17 (Th17) cells [12], all of which are crucial in the progression of AF [13]. Despite these findings, the intricate interactions between hypoxia- and immune-related genes in the context of AF remain underexplored. Therefore, the aim of this study is to explore the hub genes involved in hypoxia and immune in AF, as well as to investigate their potential roles in the onset of AF, based on the data mining of publicly available AF datasets. Materials and methods Data sources The mRNA data in GSE41177 (including 32 AF samples and 6 SR samples), GSE79768 (including 14 AF samples and 12 SR samples), GSE115574 (including 28 AF samples and 31 SR samples) cohorts were downloaded from the Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) database. Hypoxia-related genes were acquired from the HALLMARK_HYPOXIA gene set in the MSigDB database (Table S1). Differential expression analysis Based on the limma package in R language (version 3.52.4), differentially expressed mRNAs were screened using |Log2FC| > 1 and FDR < 0.05 as criteria [14]. Weighted gene co-expression network analysis (WGCNA) Using the WGCNA function package in R language [15], the top 25% genes were screened based on the median absolute deviation (MAD) of gene expression for WGCNA. The Pearson correlation coefficients were calculated between each gene, and an appropriate soft threshold β was selected to construct a weighted network that conformed to the scale-free distribution characteristics. The adjacency matrix was transformed into a topological overlap matrix (TOM). A gene hierarchical clustering tree was generated to determine gene modules. The correlation between gene modules and sample phenotypes was evaluated, and the modules significantly related to phenotypes were identified, and the genes in the modules were obtained. Construction of protein-protein interaction (PPI) network STRING database (https://string-db.org/) (Version 11.0) [16] was applied to analyze the functional relation and interaction between proteins. Cytoscape (version 3.7.2) [17] was used to visualize the PPI network, and we employed the cytoHubba plugin to further screen key genes in the PPI network based on the maximal clique centrality (MCC). Functional enrichment analysis The Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), and Gene Set Enrichment Analysis (GSEA) analyses were performed using the clusterProfiler package in R language (version 4.7.1.2) [18]. P. adjust < 0.05 was used as the criteria to screen for significantly enriched pathways. Immune cell infiltration analysis The CIBERSORT (version 0.1.0) software [19] and xCell algorithm in TIMER2.0 database [20] were utilized to calculate the relative infiltration proportions of various immune cell types in the samples, respectively. Construction of miRNA-mRNA regulatory network The miRTarBase database(https://mirtarbase.cuhk.edu.cn/~MiRTarBase/miRTarBase_2025/index.php) was used to predict miRNA targeting the key genes, and the regulatory network was visualized using Cytoscape software. Clinical samples A total of 30 blood samples were collected from 10 healthy controls (SR) and 20 AF patients recruited at the Tianjin Union Medical Center, The First Affiliated Hospital of Nankai University. Written informed consent was obtained from all participants, and the study protocol was approved by the Ethics Committee of the Tianjin Union Medical Center, The First Affiliated Hospital of Nankai University (2026018). Detailed clinical information of the participants is presented in Table S2. Reverse transcriptase quantitative real-time PCR (RT-qPCR) Total RNA extraction from whole blood samples was carried out with the RNAprep Pure Blood Kit (DP443, Tiangen Biotech). Next, cDNA synthesis was performed utilizing the Evo M-MLV RT Premix Kit (AG11728, Accurate Biology). Quantitative real-time PCR was performed using SuperStar Universal SYBR Master Mix (CW3360M, CWBio) on the SLAN-96 S Real-Time PCR System (Shanghai Hongshi Medical Technology Co., Ltd.). β-actin was utilized as the internal control, and the relative expression levels of genes were determined using the 2−ΔCT method. Primer sequences are provided in Table S3. Other statistical analyses Wilcoxon rank-sum test was used to compare the differences in gene expression and immune cell infiltration between different groups. Spearman correlation analysis was performed using the cor function in R language. P < 0.05 was considered statistically significant. All above statistical analyses were conducted using R software (version 4.2.2). For the analysis of the experimental data, differences between groups were analyzed using Student’s t-test. Data are presented as mean ± SD, and statistical significance was set at p < 0.05. Results Identification of hypoxia-related genes associated with AF The differentially expressed genes (DEGs) were analyzed between AF and SR samples using the GSE41177 cohort. The analysis revealed a total of 10,183 DEGs, which included 10,181 upregulated genes and 2 downregulated genes (Fig. 1A). The identified DEGs were then intersected with 200 hypoxia-related genes, resulting in a total of 114 intersecting genes (Fig. 1B), which were considered as AF-hypoxia-related genes. The GO and KEGG enrichment analyses conducted on these 114 genes showed that there were 365 significantly enriched GO terms and 22 significantly enriched KEGG pathways (Fig. 1C and D, Table S4). Fig. 1Identification of hypoxia-related genes associated with AF. A. Volcano plot of DEGs between AF and SR samples in GSE41177 dataset. B. Heatmap of DEGs between AF and SR samples in GSE41177 dataset. C-D. GO and KEGG bar charts Identification of hypoxia-related genes associated with AF. A. Volcano plot of DEGs between AF and SR samples in GSE41177 dataset. B. Heatmap of DEGs between AF and SR samples in GSE41177 dataset. C-D. GO and KEGG bar charts Identification of immune-related gene modules in AF through WCGNA To explore the role of immune-related genes in AF, we utilized the GSE41177 cohort and employed the CIBERSORT and xCell algorithms to analyze the differences in immune cell infiltration between AF and SR samples (Fig. 2A and B). The CIBERSORT analysis results showed significant differences in infiltration of 8 immune cell types between the two groups (Fig. 2C). Additionally, the XCell analysis revealed significant infiltration differences in 10 immune cell types between groups (Fig. 2D). Fig. 2Significant difference in immune cell infiltration between AF and SR samples. A. Heatmap of immune cell infiltration levels calculated by CIBERSORT. B. Stacked column chart of immune cell infiltration proportion estimated by xCell. C-D. Box plots comparing immune cell infiltration differences between AF and SR groups based on (C) CIBERSORT and (D) xCell Significant difference in immune cell infiltration between AF and SR samples. A. Heatmap of immune cell infiltration levels calculated by CIBERSORT. B. Stacked column chart of immune cell infiltration proportion estimated by xCell. C-D. Box plots comparing immune cell infiltration differences between AF and SR groups based on (C) CIBERSORT and (D) xCell Subsequently, for the 32 AF samples in this cohort, WGCNA was performed based on the infiltration differences of 8 immune cell types displayed by CIBERSORT analysis as phenotypic feature. The analysis revealed no outliers among the samples (Fig. 3A). Furthermore, the optimal soft threshold β was determined to be 20 (Fig. 3B), resulting in the identification of a total of seven gene modules (Fig. 3C). The correlation analysis between gene modules and sample phenotypes showed that the 5 gene modules, including blue, brown, red, yellow, and turquoise, exhibited significant correlations with immune cell infiltration (Fig. 3D). Then, the 5,642 genes encompassed within above 5 gene modules were considered as AF-immune-related genes (Table S5). Fig. 3Immune-related gene modules were identified in AF through WCGNA. A. Sample clustering dendrogram. B. Determination of the soft threshold β. C. Gene module clustering dendrogram. D. Heatmap of the correlation between modules and immune cell types Immune-related gene modules were identified in AF through WCGNA. A. Sample clustering dendrogram. B. Determination of the soft threshold β. C. Gene module clustering dendrogram. D. Heatmap of the correlation between modules and immune cell types Five hypoxia- and immune-related genes were promising markers in AF The intersection of 5,642 AF-immune-related genes and 114 AF-hypoxia-related genes resulted in 35 overlapping genes (Fig. 4A), which were designated as hypoxia- and immune-related genes associated with AF. Then, a PPI network was constructed for these 35 genes, revealing a network comprised of 19 points and 128 edges, implying their more important roles (Fig. 4B). Fig. 4Screening of potential hub genes from overlapping hypoxia- and immune-related genes in AF. A. Venn diagram showing the intersection of AF-hypoxia-related genes and AF-immune-related genes. B. PPI network diagram of overlapping genes. C-F. PPI network diagrams of the top 15 genes under different algorithms, including (C) MCC, (D) DMNC, (E) MNC, and (F) EPC Screening of potential hub genes from overlapping hypoxia- and immune-related genes in AF. A. Venn diagram showing the intersection of AF-hypoxia-related genes and AF-immune-related genes. B. PPI network diagram of overlapping genes. C-F. PPI network diagrams of the top 15 genes under different algorithms, including (C) MCC, (D) DMNC, (E) MNC, and (F) EPC To further narrow the crucial targets, we included the 15 most important nodes from the network in our subsequent screening. Then, MCC, DMNC, MNC, and EPC algorithms were employed in PPI analysis, respectively (Fig. 4C to F). Based on the nodes obtained via these four different algorithms, totally 13 overlapped genes were identified, namely BHLHE40, HMOX1, JUN, ETS1, LOX, MAFF, ATF3, IL6, SERPINE1, TNFAIP3, FOSL2, FOS, and CXCR4. The expression of these 13 genes was validated based on the GSE41177 cohort. We noticed that the expressions of IL6, JUN, FOS, FOSL2, and MAFF in the AF group were significantly higher than those in the SR group (Fig. 5A). Meanwhile, the ROC curves of these 5 genes showed that their AUC values were all greater than 0.8, indicating a good performance in diagnosing AF samples (Fig. 5B). Furthermore, to further validate these findings, two independent public datasets, GSE79768 and GSE115574, were analyzed. The results showed that all five genes exhibited significantly higher expression levels in AF samples than in SR controls in both datasets (Fig. 5C to F). Meanwhile, ROC analyses yielded AUC values exceeding 0.68 for all five genes in both datasets, suggesting moderate to good diagnostic potential (Fig. 5D and F). Fig. 5The expressions of 5 hub hypoxia- and immune-related genes between AF and SR groups. A. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE41177 dataset. B. ROC curves of these five genes in the GSE41177 dataset. C. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE79768 dataset. D. ROC curves of these five genes in the GSE79768 dataset. E. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE115574 dataset. F. ROC curves of these five genes in the GSE115574 dataset. G. RT-qPCR analysis of IL6, JUN, FOS, FOSL2, and MAFF mRNA expression levels in whole blood samples from healthy controls (SR) and AF patients. *P < 0.05, ***P < 0.001 The expressions of 5 hub hypoxia- and immune-related genes between AF and SR groups. A. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE41177 dataset. B. ROC curves of these five genes in the GSE41177 dataset. C. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE79768 dataset. D. ROC curves of these five genes in the GSE79768 dataset. E. Box plots comparing the expression differences of IL6, JUN, FOS, FOSL2, and MAFF between AF and SR groups in the GSE115574 dataset. F. ROC curves of these five genes in the GSE115574 dataset. G. RT-qPCR analysis of IL6, JUN, FOS, FOSL2, and MAFF mRNA expression levels in whole blood samples from healthy controls (SR) and AF patients. *P < 0.05, ***P < 0.001 Additionally, RT-qPCR was performed on blood samples collected from healthy controls and AF patients. The results confirmed that the mRNA expression levels of IL6, JUN, FOS, FOSL2, and MAFF were all significantly upregulated in the AF group compared to the SR group (Fig. 5G). Upstream transcription factor and miRNA regulatory networks of the five hub hypoxia- and immune-related genes According to literature reports, among these 5 genes, JUN, FOS, FOSL2, and MAFF are all transcription factors [21, 22]. To further investigate the genes potentially regulated by these 4 transcription factors, Cistrome DB (http://cistrome.org/db/#/), a ChIP-seq public database, was used to predict the genes that bind to them. The JUN, FOS, FOSL2, MAFF, and their top 10 regulated genes were shown in Fig. 6A to D. In addition, it was found that there were obvious binding peaks between JUN and the IL6 promoter in the human breast dataset [23] (Fig. 6E). In the human breast dataset [24], there were clear binding peaks between FOS and the IL6 promoter (Fig. 6F). In the human lung dataset [25], distinct binding peaks were found between FOSL2 and the IL6 promoter (Fig. 6G). In the human liver dataset [26], there were clear binding peaks between MAFF and IL6 promoter upstream (Fig. 6H). The above results indicated that JUN, FOS, FOSL2, and MAFF, as transcription factors, may regulate the transcription of IL6 gene by binding to its promoter region. Fig. 6Predicted target genes regulated by the four hub hypoxia- and immune-related transcription factors. A-D. Network diagram of transcription factors (JUN, FOS, FOSL2, and MAFF) and their top 10 target genes. E-H. Binding peak diagrams of transcription factors (JUN, FOS, FOSL2, and MAFF) and their target gene IL6 Predicted target genes regulated by the four hub hypoxia- and immune-related transcription factors. A-D. Network diagram of transcription factors (JUN, FOS, FOSL2, and MAFF) and their top 10 target genes. E-H. Binding peak diagrams of transcription factors (JUN, FOS, FOSL2, and MAFF) and their target gene IL6 Subsequently, we investigated the upstream regulatory networks of the 5 hub genes to uncover additional details regarding their abnormal expressions in AF. The miRTarBase database was used to predict the miRNAs that may regulate the expression of these 5 genes. It was indicated that IL6, JUN, FOS, FOSL2, and MAFF could be regulated by 22, 38, 58, 94, and 17 miRNAs, respectively. The top 15 miRNAs ranked by the regulation relationship scoring (TDMDScore) were selected for presentation in Fig. 7A to E. Fig. 7Predicted miRNA regulatory networks of the five hub hypoxia- and immune-related genes. A-E. miRNA-mRNA regulatory networks of IL6 (A), JUN (B), FOS (C), FOSL2 (D), and MAFF (E) Predicted miRNA regulatory networks of the five hub hypoxia- and immune-related genes. A-E. miRNA-mRNA regulatory networks of IL6 (A), JUN (B), FOS (C), FOSL2 (D), and MAFF (E) Crucial hypoxia- and immune-related genes associated with AF might involve in cAMP-dependent Ca2+ handling To further decipher the roles of 5 crucial hypoxia- and immune-related genes in AF, we analyzed 32 AF samples from the GSE41177 cohort. Based on the expression levels of these 5 genes, the samples were stratified into high- and low-expression groups, followed by GSEA analysis between the groups (Table S6). The results showed that compared with the low-expression groups of IL6, JUN, FOS, FOSL2, and MAFF, the high-expression groups exhibited significant enrichment in 24, 45, 25, 37, and 42 pathways, respectively (Table S6). It was worth noting that the cAMP signaling pathway was significantly activated by all 5 genes simultaneously, while the calcium signaling pathway was significantly activated by 4 of the genes, with the exception of IL6 (Fig. 8A to E). These two signaling pathways have been widely reported in the field of AF mechanism research [27, 28]. Fig. 8GSEA analysis of the five hub genes. A-E. GSEA-enriched pathways for the five hub genes: IL6 (A), JUN (B), FOS (C), FOSL2 (D), and MAFF (E) GSEA analysis of the five hub genes. A-E. GSEA-enriched pathways for the five hub genes: IL6 (A), JUN (B), FOS (C), FOSL2 (D), and MAFF (E) Immune landscape differences between AF samples with differentially expressed hub hypoxia- and immune-related genes To further investigate the immune landscape of AF, the infiltration differences in 22 immune cell types between the high and low expression groups of these 5 genes were calculated basing on GSE41177 cohort (Fig. 9A to E). There were significant infiltration differences in M2 macrophages, gamma delta T cells, and monocytes between the high and low JUN expression groups (Fig. 9B). M1 macrophages and neutrophils were significantly differentially infiltrated between the high and low FOS expression groups (Fig. 9C). In the comparison of high and low FOSL2 samples, significant differences in the infiltration of CD8+ T cells and resting NK cells were identified (Fig. 9D). Furthermore, significant infiltration differences in resting mast cells and resting NK cells were observed between the high and low MAFF expression groups (Fig. 9E). Fig. 9Correlation between hub gene expression and immune cell infiltration. A-E. Box plots of immune cell infiltration differences between high and low expression groups of (A) IL6, (B) JUN, (C) FOS, (D) FOSL2, and (E) MAFF. (F). Scatter plots showing correlations between hub gene expression and immune cell infiltration proportions Correlation between hub gene expression and immune cell infiltration. A-E. Box plots of immune cell infiltration differences between high and low expression groups of (A) IL6, (B) JUN, (C) FOS, (D) FOSL2, and (E) MAFF. (F). Scatter plots showing correlations between hub gene expression and immune cell infiltration proportions The correlation between these 5 genes and the infiltration of 22 immune cell types was also analyzed. It was found that FOS was significantly correlated with infiltration of 3 immune cells, including gamma delta T cells, monocytes, and neutrophils. JUN and FOSL2 were significantly correlated with gamma delta T cell infiltration. MAFF and IL6 were significantly correlated with CD4 memory activated T cell infiltration (Fig. 9F). Discussion AF is the most prevalent cardiac arrhythmia. Its high incidence and potential for severe complications pose a significant threat to human health [29]. In recent years, although some progress has been made in the treatment of AF, such as electrical cardioversion and catheter ablation, the recurrence rate remains high [30, 31], underscoring the urgent need for novel therapeutic targets. With the in-depth understanding of the pathophysiological mechanisms of AF, the roles of hypoxia- and immune-related genes in the occurrence and development of AF have gradually attracted attention. In this study, we successfully identified JUN, FOS, IL6, FOSL2, and MAFF as key hypoxia- and immune-related genes associated with AF, providing potential new therapeutic targets for the management of this challenging disease. The 5 key genes identified, IL6, JUN, FOS, FOSL2, and MAFF, are involved in various cellular processes, including cell proliferation, differentiation, and apoptosis, which were critical in many human diseases including AF. IL-6, as a pro-inflammatory cytokine [32], has been extensively studied in the field of AF. Chen et al. found that IL-6 promoted the miR-210 expression, a hypoxia-induced miRNA, by regulating HIF-1α, thereby inhibiting the function of Tregs, and thus promoting atrial fibrosis in AF [33]. Additionally, it has shown that the inflammatory state of left atrial (LA) epicardial adipose tissue (EAT) in AF patients, characterized by increased IL-6 secretion, is positively correlated with atrial fibrosis, and HIF-1α might be involved in this process [34]. Collectively, these findings indicate that IL-6-mediated atrial fibrosis in AF may be closely related to hypoxia and immune signal transduction. JUN, FOS and FOSL2 belong to the transcription factor AP-1 family, which is a class of dimeric proteins based on basic leucine zipper (bZIP) structure, participating in processes such as hypoxia injury and immune response [21, 22]. In contrast, recent studies have revealed, MAFF, a member of the small Maf family, can be rapidly induced by oxidative stress and is capable of heterodimerizing with c-Fos, thereby modulating the transcription of target genes [35], suggesting its relationship with the AP-1 family. Many studies have shown that AP-1 participated in cardiac fibrosis by activating fibroblasts and increasing the expression of profibrotic cytokines [36]. In particular, IL-6 and AP-1 family members (such as c-Jun, FOS, and FOSL2) have been implicated in cardiac fibrosis by regulating the TGF-β/Smad signaling pathway [37–40]. Our study found that compared to the SR group, the expressions of IL6, JUN, FOS, FOSL2, and MAFF were significantly upregulated in the AF group. Moreover, as transcription factors, JUN, FOS, FOSL2, and MAFF were predicted to bind to the promoter upstream of IL6 gene. Therefore, we have reason to speculate that JUN, FOS, FOSL2, and MAFF may regulate IL6 transcription, activate the TGF-β/Smad signaling pathway, induce atrial fibrosis, and consequently trigger and promote AF, while the underlying mechanisms remain to be validated in future study. To further decipher the specific roles that these 5 key genes, IL6, JUN, FOS, FOSL2, and MAFF, in AF, we conducted GSEA analysis for each gene. We noticed that the cAMP signaling pathway was significantly activated by these 5 genes simultaneously, while the calcium signaling pathway was significantly activated by 4 genes, excluding IL6. The cAMP-dependent Ca2+ handling is recognized as a crucial factor in the induction and progression of AF [41, 42]. In patients with AF, the activity of the sympathetic nervous system (SNS) is enhanced, and the released norepinephrine stimulates β-adrenergic receptors (β-AR), leading to an increase in cAMP in atrial myocytes [43–45]. The increase in cAMP activates protein kinase A (PKA) and calcium/calmodulin-dependent kinase II (CaMKII) [46–48]. Both PKA and CaMKII can phosphorylate the L-type Ca2 + current (LTCC), ryanodine receptor 2 (RyR2), and phospholamban (PLB), leading to an increase in the expression of sarco/endoplasmic reticulum Ca2+-ATPase (SERCA2a). This cascade triggers an increase in the L-type Ca2+ current and Ca2+ overload in the sarcoplasmic reticulum and cytoplasm, all of which contribute to the occurrence and development of AF [49, 50]. Moreover, the activation of JNK/c-Jun signaling pathway has been shown to reduce the expression of connexin 43 (Cx43), while enhancing the activity of CaMKII and SERCA2a. This results in a decrease in atrial conduction velocity and abnormal Ca2+ handling in the sarcoplasmic reticulum, ultimately leading to the occurrence of AF [51–53]. In addition, numerous studies have shown that hypoxic and inflammatory environments can trigger calcium homeostasis abnormalities in cardiomyocytes [54, 55]. Herein, we propose that these 5 key genes, IL6, JUN, FOS, FOSL2, and MAFF, may induce and promote AF through the cAMP-dependent Ca2+ handling mechanism, while the precise mechanisms remain to be explored in future studies. A limitation of our study is that although we predicted the binding of JUN, FOS, FOSL2, and MAFF to the IL6 promoter using publicly available ChIP-seq data from Cistrome DB, these datasets were derived from breast, lung, and liver tissues, not from atrial or cardiac tissue. Transcription factor binding is highly tissue-specific and dependent on the chromatin context [56, 57]. Therefore, the binding peaks observed in non-cardiac tissues do not constitute definitive evidence that these transcription factors directly regulate IL6 expression in atrial fibrillation. Future studies employing cardiac-specific ChIP-seq or ChIP-qPCR, performed on human atrial tissue or mouse cardiac tissue, are warranted to definitively validate the direct binding of these transcription factors to the IL6 promoter. Conclusion In conclusion, through comprehensive bioinformatics analysis, we identified 5 key hypoxia- and immune-related genes (JUN, FOS, IL6, FOSL2, and MAFF) associated with AF. These genes were significantly upregulated in AF samples. Our findings may provide potential therapeutic targets for AF treatment. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3. Supplementary Material 4. Supplementary Material 4. Supplementary Material 5. Supplementary Material 5. Supplementary Material 6. Supplementary Material 6.