Crosstalk between Atrial Cardiomyocytes and Epicardial Adipose Tissue in Atrial Fibrillation: Insights from Machine Learning Methods and Rat Atrial Fibrillation Model

doi:10.21203/rs.3.rs-3934007/v1

Download PDF

Article

Crosstalk between Atrial Cardiomyocytes and Epicardial Adipose Tissue in Atrial Fibrillation: Insights from Machine Learning Methods and Rat Atrial Fibrillation Model

https://doi.org/10.21203/rs.3.rs-3934007/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective: Investigating the Role of Epicardial Adipose Tissue as a Catalyst for Reversal and Reconstruction of Atrial Myocardial Cells in the Context of Dialogue with Atrial Myocardial Cells. Implications for Breakthroughs in Preventing Paroxysmal Atrial Fibrillation Progression.

Methods: We obtained three datasets (GSE41177, GSE31821, and GSE135455) associated with atrial fibrillation (AF) from the Gene Expression Omnibus (GEO) database, which were subsequently merged for comprehensive analysis. Differentially expressed genes (DEGs) were identified using the "limma" package in the R software. Candidate AF genes were selected through machine learning techniques, including the LASSO regression algorithm and SVM-RFE algorithm. The diagnostic efficacy of these genes was evaluated using Receiver Operating Characteristic (ROC) curves. Additionally, CIBERSORT was employed to investigate the proportions of infiltrating immune cells in each sample, while the Pearson method was applied to examine the correlation between genes and immune cells. Further validation of the DEGs were performed by PCR in atrial fibrillation rats.

Results: A total of 310 Differentially Expressed Genes (DEGs) were identified in atrial cardiomyocytes with epicardial adipose tissue. Using the LASSO regression and SVM-RFE algorithms, ID1, SCN4A, COL4A5, COLEC11, and SNAI2 were pinpointed as key genes associated with Atrial Fibrillation (AF). In both the training and validation datasets, these genes exhibited robust effectiveness. The immune infiltration analysis revealed that, in comparison to sinus rhythm (SR), atrial samples from patients with AF exhibited higher levels of neutrophils, while T cells follicular helper were relatively lower. Correlation analysis highlighted significant associations between ID1, SCN4A, COL4A5, COLEC11, SNAI2, and infiltrating immune cells. The outcomes of the RT- qPCR analysis in our investigation were consistent with the findings of bioinformatics analysis.

Conclusions: In summary, this study posits that ID1, SCN4A, COL4A5, COLEC11, and SNAI2 emerge as pivotal genes in Atrial Fibrillation (AF), exhibiting correlation with infiltrating immune cells. Furthermore, it underscores the indispensable roles played by infiltrating immune cells in the context of AF.

Health sciences/Cardiology

Health sciences/Diseases

Atrial fibrillation

Crosstalk

Atrial cardiomyocytes

Epicardial adipose tissue

Immune infiltration

Atrial fibrillation (AF) stands as the most prevalent persistent arrhythmia globally, ranking among the primary causes of stroke, heart failure, sudden death, and cardiovascular disease¹. Globally, an estimated 33.5 million individuals suffer from AF, a number steadily increasing with the aging of the population ². Despite its prevalence, the pathophysiology of many AF cases remains elusive, resulting in a scarcity of effective treatments ³. Only a minority of AF patients achieve normalized heart rhythm through catheter ablation or cardiac surgery ⁴. The elevated prevalence of AF and the limited treatment options contribute significantly to both public health and economic burdens ⁵. Consequently, there is an imperative need to enhance our understanding of AF's pathogenesis and develop improved methods in preventing and treating the progression of atrial fibrillation.

Epicardial adipose tissue (EAT) serves as the visceral fat reservoir of the heart, exhibiting high plasticity and direct contact with the heart muscle and coronary arteries. Due to its unique proximity, this tissue's secreted adipokines and pro-inflammatory molecules may directly impact heart and coronary artery metabolism⁶. Positioned adjacent to the myocardium, EAT infiltrates it, harbors a rich ganglion plexus, and actively releases cytokines and adipokines, influencing inflammation or remodeling ⁷. EAT serves as a local source of inflammatory mediators, potentially resulting in atrial collagen deposition and fibrosis ⁸. Simultaneously, EAT can infiltrate and induce atrial myocardial fat deposition, altering atrial electrophysiological characteristics. Additionally, EAT engages in structural remodeling through its endocrine and paracrine activities, contributing to atrial fibrillation^{9, 10}.

In recent years, the role of Epicardial Adipose Tissue (EAT) in atrial fibrillation (AF) has been widely studied¹¹. Literature reveals the following findings: ①Under normal circumstances, EAT is a metabolically active visceral fat deposited around the heart, showing high expression of genes such as UCP-1, PRDM16, PGC-1α, Parkin, and the brown adipocyte-specific marker CD137. It exhibits molecular characteristics of brown adipocytes, which are protective for atrial myocardium¹²; ②The expression of UCP-1 in EAT of AF patients significantly decreases compared to normal individuals, trending towards whitening¹³; ③Whitened EAT can directly infiltrate atrial and ventricular myocardium through adipocytes or fat, leading to increased dispersion of normal myocardial electrical signals, or fat cells serving as a direct source of abnormal signals^{14, 15}; ④Whitened EAT can also secrete a large number of cytokines, inflammatory factors, and chemokines through paracrine and endocrine pathways, inducing chronic inflammation¹⁶, oxidative stress¹⁷, mitochondrial dysfunction¹⁸, local autonomic nervous dysfunction¹⁹, myocardial cell death²⁰, and pathological changes in cell membrane ion channel function²¹. These changes ultimately promote the occurrence and progression of atrial fibrillation.

In this study, our objective is to identify potential key genes in the Atrial Cardiomyocytes and EAT of AF patients through the Gene Expression Omnibus (GEO) comprehensive database, utilizing a bioinformatic approach and RT-qPCR. Subsequently, we aim to analyze their expression, function, and interactions.

2.1 Atrial fibrillation datasets

Raw files from three registered microarray datasets, namely GSE31821, GSE41177, and GSE135445 (refer to Table 1), were retrieved from the NCBI GEO database (https://www.ncbi.nlm.nih.gov/geo/). All datasets originated from the Affymetrix Human Genome U133 Plus 2.0 Array [HGU133_Plus_2] microarray platform. Within each dataset, samples of human left atrial appendage (LAA) and epicardial adipose tissue (EAT) were selected from individuals with Atrial Fibrillation (AF) and those in Sinus Rhythm (SR). Subsequently, 36 AF samples and 8 SR samples were included for further analyses (Table 1).

2.2 Data preprocessing

The GSE135445 dataset underwent a log base 2 transformation. Probe names were converted to gene symbols using platform annotation information, followed by the normalization of expression values through the "normalizeBetweenArrays" function within the "limma" package in R software. This normalization process aimed to ensure a consistent distribution of expression values across the entire array set. Subsequently, all microarray data were merged, and batch effects were mitigated using the "combat" function from the "sva" package in R software.

2.3 Identification of differentially expressed genes (DEGs)

To assess differential expression, the "limma" package in R software was employed. A linear model was fitted, and standard errors were moderated using a simple empirical Bayes model. Subsequently, for each gene and contrast, a moderated t-statistic and log-odds of differential expression were computed. A gene was deemed differentially expressed between the Atrial Fibrillation (AF) and Sinus Rhythm (SR) samples when the P value was < 0.05, and the gene expression fold change (FC) value was ≥ 1.41 or ≤ 0.707 (|log2 FC| ≥0.5). Visualization was carried out through Volcano plots and heat map plots. Genes with log2FC > 0.5 were classified as up-regulated, while those with log2FC < − 0.5 were considered down-regulated.

2.4 GO and KEGG Enrichment Analysis

To elucidate the role of Differentially Expressed Genes (DEGs) in Atrial Fibrillation (AF) patients, the "clusterProfiler" package in R software was utilized for Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. Concurrently, Gene Set Enrichment Analysis (GSEA) was employed to identify pathways that are enriched in both AF patients and individuals in Sinus Rhythm (SR)

2.5 Identification of Key DEGs as Hub Genes in AF using Machine Learning Methods

The Least Absolute Shrinkage and Selection Operator (LASSO) is a regression analysis technique employing regularization to enhance predictive accuracy. The LASSO regression analysis was conducted using the "glmnet" package in R to identify genes significantly associated with the discriminative power of Atrial Fibrillation (AF) and Sinus Rhythm (SR). Support Vector Machine (SVM) is a supervised machine learning technique widely utilized for classification and regression analysis. To prevent overfitting, Recursive Feature Elimination (RFE) was employed to select the optimal genes from the metadata cohort. Consequently, for identifying the gene set with the greatest discrimination ability, SVM Recursive Feature Elimination (SVM-RFE) was utilized to screen suitable characteristics.

2.6 CIBERSORT Analysis

The computational approach employed in CIBERSORT (http://cibersort.stanford.edu/) utilizes deconvolutional algorithms based on genetic expressions. It serves to evaluate the variations of a specific gene group concerning the remaining genes within a specimen. Through the CIBERSORT algorithm, our team identified the immunoresponses of 22 distinct immune cells and examined their correlation with the expression of crucial genes in Sinus Rhythm (SR) and Atrial Fibrillation (AF) samples. The primary objective of this research was to ascertain the associations among these immune cells.

2.7 Rat Atrial Fibrillation Model Workstation and Reverse transcription‑quantitative polymerase chain reaction (RT‑qPCR) analysis

The study utilized 8-week-old Sprague-Dawley rats as experimental subjects. All aspects of mouse feeding, animal care, and experimental procedures adhered to the "Measures for the Quality Management of Experimental Animals" (Guo Ke CAI Zi [1997], No. 593), which were approved by the Experimental Animal Ethics Committee of Ningbo University and conducted in accordance with Chinese government guidelines. In order to investigate the biological functions of the DEGs more comprehensively, we created a rat atrial fibrillation model utilizing the YC-3 Bipolar Program-controlled Stimulator (Chengdu Instrument Factory, China), and the EP-Workmate Recording system v4.3.2(Abbott, USA). The specific parameters are as follows: ①square wave stimulus waveform; ②individual adjustment of the stimulus threshold to 3mA to ensure atrial capture; ③stimulus pulse width of 6 ms; ④ stimulus frequency of 20Hz; ⑤baseline stimulation time of 3 hours. All animal studies were approved by the Animal Care and Ethics Committee of Ningbo University (Zhejiang, China). The rat atrial fibrillation model workstation are presented in supplementary Fig. 1.

Samples of rat’s atrial tissue were selected from individuals with Rat Atrial Fibrillation (AF) and those in Sinus Rhythm (SR). The DEGs were assessed using the 2 − ΔΔCt formula, with GAPDH serving as the reference gene. Total RNA was extracted from the samples using TRIzol reagent (Thermo Fisher Scientific). RNA concentration was determined by measuring the absorbance ratio at 260/280 nm with a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific). Reverse transcription of RNA was carried out using the PrimeScript™ RT Reagent Kit with gDNA Eraser (RR047A; Takara, Tokyo, Japan), and cDNA was analyzed by qRT-PCR using SYBR® Premix Ex Taq™ (RR420A; Takara, Tokyo, Japan). The data were normalized to GAPDH levels and further analyzed using the 2 − ΔΔCT method. The primer sequences for all qPCR assays are presented in Supplementary Table 2.

2.8 Statistical Analysis.

Student's t-tests were employed to compare the genetic expressions between Atrial Fibrillation (AF) and Sinus Rhythm (SR). To assess the classification effects of crucial genes on AF and SR, Receiver Operating Characteristic (ROC) curves and Area Under the Curve (AUC) were calculated using the "pROC" package in R. Statistical analysis was conducted using R version 4.3.1.

3.1. Determination of DEGs in AF.

A total of 42 AF and 14 Sinus Rhythm (SR) samples from three Gene Expression Omnibus (GEO) datasets (GSE41177, GSE31821, and GSE135445) were retrospectively analyzed in this study. DEGs within the metadata were investigated using the limma package after mitigating batch effects. A total of 310 DEGs were identified: 216 genes exhibited significant upregulation, while 94 genes showed distinct downregulation (Fig. 2A, B).

3.2 GO and KEGG Enrichment Analysis.

To investigate the biological functions of the 310 DEGs in AF, we conducted GO and KEGG analyses using the ClusterProfile R package. The results indicated that the 310 DEGs were primarily associated with processes such as leukocyte migration, positive regulation of cytokine production, myeloid leukocyte migration, cell chemotaxis, leukocyte chemotaxis, reactive oxygen species metabolic process, neutrophil migration, granulocyte migration, neutrophil chemotaxis, and granulocyte chemotaxis (Fig. 3(A)). Additionally, KEGG pathway analyses revealed significant enrichment in pathways including Cytokine − cytokine receptor interaction, PI3K − Akt signaling pathway, Neutrophil extracellular trap formation, Human T − cell leukemia virus 1 infection, Human cytomegalovirus infection, Chemokine signaling pathway, Staphylococcus aureus infection, Phagosome, cGMP − PKG signaling pathway, Tuberculosis, Fluid shear stress and atherosclerosis, Kaposi sarcoma − associated herpesvirus infection, among others (Fig. 3(B)).

3.3. Determination and Verification of hub genes

Two distinct algorithms were employed to identify potential biological markers. The DEGs were determined using the LASSO regression method, resulting in the identification of 15 variables as hub genes for Atrial Fibrillation (AF) (Fig. 4(A)). Simultaneously, a subset of 6 features among the DEGs was recognized through SVM-RFE arithmetic (Fig. 4(B)). Five overlapping features (ID1, SCN4A, COL4A5, COLEC11, and SNAI2) between these two algorithms were ultimately selected (Fig. 4(C)). These five genes may play a crucial role in the progression of AF.

3.4 Expression and Diagnostic Significance of ID1, SCN4A, COL4A5, COLEC11, and SNAI2 in AF

Our investigation revealed distinct upregulation of ID1 (Fig. 5(A)), COLEC11 (Fig. 5(C)), and SNAI2 (Fig. 5(E)) in AF samples compared to Sinus Rhythm (SR), while COL4A5 (Fig. 5(G)) and SCN4A (Fig. 5(I)) exhibited significant downregulation in AF. To further assess the diagnostic value of ID1, SCN4A, COL4A5, COLEC11, and SNAI2, ROC analyses were performed. Four genes demonstrated robust capabilities in distinguishing AF from normal samples, including ID1 (Fig. 5(B), AUC = 0.937), COLEC11 (Fig. 5(D), AUC = 0.838), SNAI2 (Fig. 5(F), AUC = 0.808), COL4A5 (Fig. 5(H), AUC = 0.898), and SCN4A (Fig. 5(J), AUC = 0.915).

3.5 Association of ID1, SCN4A, COL4A5, COLEC11, and SNAI2 with Immunocyte Infiltration Levels

Our research delved into the coefficients of ID1, SCN4A, COL4A5, COLEC11, and SNAI2, along with the immunocyte infiltration status in Atrial Fibrillation (AF) and Sinus Rhythm (SR). The objective was to unveil the correlation between immunocyte infiltration status and the expression levels of ID1, SCN4A, COL4A5, COLEC11, and SNAI2. Immunocyte features were scrutinized using the CIBERSORT approach, and their composition in AF and SR, as well as their interrelationships, are graphically presented in Figs. 6(A) and 6(B). Additionally, compared to SR, left atrial appendage (LAA) and epicardial adipose tissue (EAT) samples from AF patients exhibited elevated neutrophil levels (p < 0.001) and diminished T cells follicular helper levels (p = 0.0149) (Figs. 6(C)).

Furthermore, we explored the correlation between the expressions of ID1, SCN4A, COL4A5, COLEC11, and SNAI2, and immune infiltrating levels. As illustrated in Fig. 7(A-E), ID1 showed correlation with T cells regulatory (Tregs) (r = 0.0292), T cells follicular helper (r = 0.0068), NK cells activated (p = 0.0068), Macrophages M2 (r = 0.0068), Macrophages M0 (r = 0.0368). SCN4A exhibited correlation with T cells gamma delta (r = 0.0460), T cells follicular helper (r = 0.0137), NK cells activated (r = 0.0182). COL4A5 demonstrated correlation with neutrophils (r = 0.0268), Mast cells activated (r = 0.0438). COLEC11 correlated with Macrophages M2 (r = 0.0139), eosinophils (r = 0.0107). Moreover, SNAI2 was correlated with various immune cells, indicating its potential role in AF progression.

3.6 Rat Atrial Fibrillation Model Workstation and the DEGs of RT-qPCR

The RT-qPCR was applied to confirm the expression levels of ID1, SCN4A, COL4A5, COLEC11, SNAI2 after we produced rat atrial fibrillation model. As demonstrated in Fig. 8A-C, the expression levels of ID1, COLEC11, SNAI2 were significantly higher in AF samples, as compared to SR (all p-value < 0.05). The expression of SCN4A (Fig. 8D) was significantly downregulated in AF patients when compared to SR (p-value < 0.05). The findings of the RT- qPCR analysis in our research were consistent with the results of bioinformatics analysis. However, there was no difference in the expression of COL4A5 between AF patients and SR.

In this study, we integrated gene expression profiles from 42 AF and 14 SR samples across three GEO datasets, identifying 216 DEGs with |log2 FC| ≥ 0.5 and P value < 0.05. Five potential crucial genes (ID1, SCN4A, COL4A5, COLEC11, and SNAI2) were identified, suggesting their pivotal roles in AF mechanisms.

EAT is also enriched in immune activation-related genes and T lymphocytes compared with subcutaneous adipose tissue in patients with chronic heart failure. In addition, CCL5 and GZMK were shown to be associated with T lymphocytes in EAT from chronic heart failure patients²². At the same time, in patients with atrial fibrillation, dysregulation of the autonomic nervous system, increased humoral function of EAT, local atrial lymphomonocytic infiltration, and increased inflammatory vesicle activity can be observed²³. In animal experiments, electrical remodeling, increased susceptibility to atrial fibrillation, immune cell activation, inflammatory infiltration, and fibrosis can be observed in mice induced by lipopolysaccharide. And it was confirmed that acute inflammation promotes α-microtubulin degradation and induces electrical remodeling, which induces AF²⁴. However, little research has been done on whether EAT in turn affects AF through immune infiltration. This may become one of the directions for future research.

ID1 and SNAI2 may contribute to the pathophysiological process of AF through the Hippo signaling pathway. This study identified significantly elevated ID1 and SNAI2 expression in EAT and left atrial appendage tissue of atrial fibrillation patients. KEGG analysis implicated ID1 and SNAI2 in the Hippo signaling pathway. ID1, SNAI2 and the Hippo pathway are linked to various diseases, with the latter particularly associated with fibrosis²⁵. YWAE enrichment in the Hippo pathway, as observed in the transcription-target gene regulatory network of atrial fibrillation, indicates the pivotal role of the Hippo pathway in atrial fibrillation pathogenesis²⁶. Cao et al. propose the significant involvement of the Hippo signaling pathway in the progression of atrial fibrillation ²⁷. Current studies indicate EAT's close association with atrial fibrillation, encompassing fibrosis and metabolic remodeling. In adipose tissue, the Hippo pathway directs cell transformation from adipocytes to myofibroblasts during inactivation. Dysregulation, leading to maladaptation of adipocytes to chronic energy excess, ultimately results in adipofibrosis. In obesity, upregulated kinase expression (STK3 and STK4) in the Hippo pathway enhances mitochondrial autophagy in adipocytes, disrupting systemic metabolism and elevating cardiovascular disease risk ²⁸. Yao et al. demonstrated that hUC-MSCs regulate fibrosis via the Hippo/YAP/ ID1 pathway and macrophage-dependent mechanisms, alleviating liver cirrhosis in mice ²⁹.Therefore, we propose ID1 and SNAI2's involvement in AF occurrence and development through the Hippo signaling pathway.

This study found a significant up-regulation of COL4A5 expression in EAT and left atrial ear tissues of patients with atrial fibrillation. KEGG analysis indicated the involvement of COL4A5 in the PI3K-Akt signaling pathway. The PI3K-Akt signaling pathway is closely linked to atrial fibrosis. In atrial fibrillation patients, miR-210-3p expression is significantly up-regulated. Glycerol-3-phosphate dehydrogenase 1 (GPD1L), a potential target gene of miR-210-3p, regulates atrial fibrosis through the PI3K-AKT signaling pathway³⁰. Animal experiments confirmed that H2S significantly down-regulates the PI3K-AKt-endothelial nitric oxide synthase (eNOS) pathway, reducing DM-induced atrial fibrosis, cardiac fibroblast proliferation, and migration, thereby inhibiting atrial fibrillation³¹. The PI3K-Akt signaling pathway is closely related to inflammation. Animal experiments demonstrated a significant up-regulation of the PI3K-AKT signaling pathway in brown adipose tissue of mice with CIA-induced arthritis compared to normal white adipose tissue. Pro-inflammatory cytokines tend to decrease after normal BAT transplantation into mice with CIA-induced arthritis³⁰. COL4A5 is closely associated with inflammation. Overexpression of miR-221-5p inhibits COL4A5 mRNA expression, promoting inflammation³².Therefore, we hypothesize that COL4A5 can mediate the inflammatory response in EAT and regulate atrial fibrosis through the PI3K-Akt signaling pathway, contributing to the pathogenesis and progression of atrial fibrillation.

COLEC11 is closely linked to the complement system ^33–35. Previously, it was believed that COLEC11 primarily participated in the pathophysiological mechanisms of kidney diseases. In individuals with IgA nephropathy, the presence of IgA1 immune complexes increased COLEC11 secretion, leading to the deposition of COLEC11-IgA1 antibody complexes on cell surfaces. Recent findings indicate a significant elevation in COLEC11 plasma levels in patients with DIC, suggesting a potential role in immune processes ³⁶. This study observed a significant up-regulation of COLEC11 expression in EAT and left atrial ear tissues of patients with atrial fibrillation. KEGG analysis implicated COLEC11 in the Phagosome signaling pathway, which is closely associated with atrial fibrillation. Lysosome and phagosome pathways play a crucial role in Ang II-induced AF development, involving antigen processing and presentation, chemokine signaling, and extracellular stroma-receptor interactions³⁷. Additionally, in adipose tissue, TMAO has been identified as a regulator of the metabolic state of vWAT through phagosomal and lysosomal pathways, reducing fat production and improving specific fatty acid composition³⁸. Therefore, we postulate that COLEC11 may contribute to the pathophysiological mechanisms of atrial fibrillation through the Phagosome pathway.

SCN4A encodes the α-subunit of the skeletal voltage-gated sodium channel, Nav1.4. This channel, expressed in skeletal muscle, influences muscle excitability ³⁹. Mutations in SCN4A lead to non-dystrophic myotonia and/or periodic paralysis. Additionally, SCN4A mutations contribute to hyperkalemic periodic paralysis and sudden infant death syndrome⁴⁰. A study highlighted the significant role of SCN4A variants in the pathophysiology of spontaneous or drug-induced type 1 ECG patterns and malignant arrhythmias in certain Brugada syndrome patients ^{41, 42}. Hence, we posit that SCN4A might be implicated in atrial fibrillation (AF) development, potentially through epicardial adipose tissue (EAT).

We found AF is related to neutrophil and T cells follicular helper. A mendelian randomization study show higher peripheral immune cell counts are related with AF closely, especially CD4 ⁺ T cell⁴³. Further, according to immune analysis, we found the expression of SCN4A and SNAI2 genes showed a positive correlation with T cells. SCN4A encodes one member of the sodium channel alpha subunit gene family. Voltage-gated sodium channels can reduce the procession of atherosclerotic lesions, which may associate with the change of monocyte/macrophage subsets⁴⁴. We suppose that SNAI2 can mediated AF via the change of monocyte/macrophage subsets. On the contrary, ID1 gene expression showed a negative correlation with T cells. In the immune system, ID1-4 plays a regulatory role in the activation of T and B cells⁴⁵. Studies have shown a correlation between high expression of ID1 and poor prognosis in patients with colorectal cancer. Additionally, high expression of ID1 maintains cancer stemness and prevents infiltration of CD8⁺ T cell⁴⁶. We assume that ID1 can mediate the process of AF via CD8⁺ T cell. Additionally, COL4A5 gene expression was also negatively correlated with neutrophils. However, SNAI2 gene expression was positive correlated with neutrophils. In colorectal cancer, we found neutrophil extracellular traps (NETs) and the increase expression of SNAI2. The activation of NETs may take part in the metastatic progression of colorectal cancer. NETs is associated with the increase expression of SNAI2⁴⁷. At the same time, we observe NETs is related with AF. NETs induced the autophagic apoptosis of cardiomyocytes, and NETs also led to mitochondrial injury by promoting mitochondrial depolarization and ROS production⁴⁸. We supposed that SNAI2 can exacerbate the progression of AF via NETs.

In conclusion, five pivotal candidate genes (ID1, SCN4A, COL4A5, COLEC11, and SNAI2) likely play significant roles in the onset and progression of atrial fibrillation. This study will provide new strategies and potential targets for the prevention and treatment of atrial fibrillation.

Acknowledgments

Not applicable.

Author contributions

HZ Miao and BJ Song performed the bioinformation analysis, designed and performed the experiments, and wrote the manuscript. L Tao analyzed data and performed some in vitro experiments. J Wang participated in the animal experiments. QJ Zhang and YC Bao (corresponding author) designed, supervised the study, and performed manuscriptediting.

FundingThis work was supported by grants from the Natural Science Foundation of Zhejiang Province (NO.LY20H020001)，the Natural Science Foundation of Ningbo City(NO. 2022J038).

Competing interests

The authors declare that they have no competing interests. The study was carried out in compliance with the ARRIVE guidelines (https://arriveguidelines.org).

Ethics approval

All animal studies were approved by the Animal Care and Ethics Committee of Ningbo University (Zhejiang, China, NO.11944).

Consent for publication

Not applicable.

Andrade JG, Aguilar M, Atzema C, et al. The 2020 Canadian Cardiovascular Society/Canadian Heart Rhythm Society Comprehensive Guidelines for the Management of Atrial Fibrillation. Can J Cardiol. Dec 2020;36(12):1847-1948.
Sagris M, Vardas EP, Theofilis P, Antonopoulos AS, Oikonomou E, Tousoulis D. Atrial Fibrillation: Pathogenesis, Predisposing Factors, and Genetics. Int J Mol Sci. Dec 21 2021;23(1).
Verheule S, Eckstein J, Linz D, et al. Role of endo-epicardial dissociation of electrical activity and transmural conduction in the development of persistent atrial fibrillation. Prog Biophys Mol Biol. Aug 2014;115(2-3):173-185.
Leventopoulos G, Koros R, Travlos C, et al. Mechanisms of Atrial Fibrillation: How Our Knowledge Affects Clinical Practice. Life (Basel). May 25 2023;13(6).
Sheikh A, Patel NJ, Nalluri N, et al. Trends in hospitalization for atrial fibrillation: epidemiology, cost, and implications for the future. Prog Cardiovasc Dis. Sep-Oct 2015;58(2):105-116.
Doukbi E, Soghomonian A, Sengenes C, et al. Browning Epicardial Adipose Tissue: Friend or Foe? Cells. Mar 14 2022;11(6).
Zhou M, Wang H, Chen J, Zhao L. Epicardial adipose tissue and atrial fibrillation: Possible mechanisms, potential therapies, and future directions. Pacing Clin Electrophysiol. Jan 2020;43(1):133-145.
Conte M, Petraglia L, Cabaro S, et al. Epicardial Adipose Tissue and Cardiac Arrhythmias: Focus on Atrial Fibrillation. Front Cardiovasc Med. 2022;9:932262.
Couselo-Seijas M, Rodriguez-Manero M, Gonzalez-Juanatey JR, Eiras S. Updates on epicardial adipose tissue mechanisms on atrial fibrillation. Obes Rev. Sep 2021;22(9):e13277.
Nalliah CJ, Bell JR, Raaijmakers AJA, et al. Epicardial Adipose Tissue Accumulation Confers Atrial Conduction Abnormality. J Am Coll Cardiol. Sep 8 2020;76(10):1197-1211.
Ernault AC, Meijborg VMF, Coronel R. Modulation of Cardiac Arrhythmogenesis by Epicardial Adipose Tissue: JACC State-of-the-Art Review. J Am Coll Cardiol. Oct 26 2021;78(17):1730-1745.
Sacks HS, Fain JN, Bahouth SW, et al. Adult epicardial fat exhibits beige features. J Clin Endocrinol Metab. Sep 2013;98(9):E1448-1455.
He Y, Ma N, Tang M, Jiang ZL, Liu H, Mei J. The differentiation of beige adipocyte in pericardial and epicardial adipose tissues induces atrial fibrillation development. Eur Rev Med Pharmacol Sci. Oct 2017;21(19):4398-4405.
Otsuka N, Okumura Y, Arai M, et al. Effect of obesity and epicardial fat/fatty infiltration on electrical and structural remodeling associated with atrial fibrillation in a novel canine model of obesity and atrial fibrillation: A comparative study. J Cardiovasc Electrophysiol. Apr 2021;32(4):889-899.
Anumonwo JMB, Herron T. Fatty Infiltration of the Myocardium and Arrhythmogenesis: Potential Cellular and Molecular Mechanisms. Front Physiol. 2018;9:2.
Packer M. Epicardial Adipose Tissue May Mediate Deleterious Effects of Obesity and Inflammation on the Myocardium. J Am Coll Cardiol. May 22 2018;71(20):2360-2372.
Antonopoulos AS, Margaritis M, Verheule S, et al. Mutual Regulation of Epicardial Adipose Tissue and Myocardial Redox State by PPAR-γ/Adiponectin Signalling. Circ Res. Mar 4 2016;118(5):842-855.
Burgeiro A, Fonseca AC, Espinoza D, et al. Proteostasis in epicardial versus subcutaneous adipose tissue in heart failure subjects with and without diabetes. Biochim Biophys Acta Mol Basis Dis. Jun 2018;1864(6 Pt A):2183-2198.
Romanov A, Pokushalov E, Ponomarev D, et al. Long-term suppression of atrial fibrillation by botulinum toxin injection into epicardial fat pads in patients undergoing cardiac surgery: Three-year follow-up of a randomized study. Heart Rhythm. Feb 2019;16(2):172-177.
Meng T, Cheng G, Wei Y, et al. Exposure to a chronic high-fat diet promotes atrial structure and gap junction remodeling in rats. Int J Mol Med. Jul 2017;40(1):217-225.
Lee KT, Tang PW, Tsai WC, et al. Differential effects of central and peripheral fat tissues on the delayed rectifier K(+) outward currents in cardiac myocytes. Cardiology. 2013;125(2):118-124.
Zhang XZ, Chen XL, Tang TT, et al. T lymphocyte characteristics and immune repertoires in the epicardial adipose tissue of heart failure patients. Front Immunol. 2023;14:1126997.
Papathanasiou KA, Giotaki SG, Vrachatis DA, et al. Molecular Insights in Atrial Fibrillation Pathogenesis and Therapeutics: A Narrative Review. Diagnostics (Basel). Aug 31 2021;11(9).
Ying H, Guo W, Tang X, et al. Colchicine attenuates the electrical remodeling of post-operative atrial fibrillation through inhibited expression of immune-related hub genes and stabilization of microtubules. Int J Biol Sci. 2023;19(9):2934-2956.
Shen H, Huang X, Zhao Y, et al. The Hippo pathway links adipocyte plasticity to adipose tissue fibrosis. Nat Commun. Oct 13 2022;13(1):6030.
Zhang H, Yang G, Zhong N, et al. Possible key microRNAs and corresponding molecular mechanisms for atrial fibrillation. Anatol J Cardiol. Jun 2020;23(6):324-333.
Cao Y, Cui L. Identifying the key microRNAs implicated in atrial fibrillation. Anatol J Cardiol. Jun 2021;25(6):429-436.
Cho YK, Son Y, Saha A, et al. STK3/STK4 signalling in adipocytes regulates mitophagy and energy expenditure. Nat Metab. Mar 2021;3(3):428-441.
Yao L, Hu X, Yuan M, et al. Human umbilical cord-derived mesenchymal stromal cells alleviate liver cirrhosis through the Hippo/YAP/Id1 pathway and macrophage-dependent mechanism. Int Immunopharmacol. Oct 2023;123:110456.
Hao H, Yan S, Zhao X, et al. Atrial myocyte-derived exosomal microRNA contributes to atrial fibrosis in atrial fibrillation. J Transl Med. Sep 5 2022;20(1):407.
Xue X, Ling X, Xi W, et al. Exogenous hydrogen sulfide reduces atrial remodeling and atrial fibrillation induced by diabetes mellitus via activation of the PI3K/Akt/eNOS pathway. Mol Med Rep. Sep 2020;22(3):1759-1766.
Zhao N, Yu H, Xi Y, et al. MicroRNA-221-5p Promotes Ricin Toxin-induced Inflammation via PI3K/Akt signaling pathway by targeting COL4a5. Toxicon. Jun 2022;212:11-18.
Hansen SWK, Aagaard JB, Bjerrum KB, et al. CL-L1 and CL-K1 Exhibit Widespread Tissue Distribution With High and Co-Localized Expression in Secretory Epithelia and Mucosa. Front Immunol. 2018;9:1757.
Wei M, Guo WY, Xu BY, et al. Collectin11 and Complement Activation in IgA Nephropathy. Clin J Am Soc Nephrol. Dec 2021;16(12):1840-1850.
Nauser CL, Howard MC, Fanelli G, Farrar CA, Sacks S. Collectin-11 (CL-11) Is a Major Sentinel at Epithelial Surfaces and Key Pattern Recognition Molecule in Complement-Mediated Ischaemic Injury. Front Immunol. 2018;9:2023.
Takahashi K, Ohtani K, Larvie M, et al. Elevated plasma CL-K1 level is associated with a risk of developing disseminated intravascular coagulation (DIC). J Thromb Thrombolysis. Oct 2014;38(3):331-338.
Wu YX, Han X, Chen C, et al. Time Series Gene Expression Profiling and Temporal Regulatory Pathway Analysis of Angiotensin II Induced Atrial Fibrillation in Mice. Front Physiol. 2019;10:597.
He GD, Liu XC, Hou XH, Feng YQ. The effect of trimethylamine N-oxide on the metabolism of visceral white adipose tissue in spontaneously hypertensive rat. Adipocyte. Dec 2022;11(1):420-433.
Huang CW, Lai HJ, Lin PC, Lee MJ. Changes of Resurgent Na(+) Currents in the Na(v)1.4 Channel Resulting from an SCN4A Mutation Contributing to Sodium Channel Myotonia. Int J Mol Sci. Apr 8 2020;21(7).
Mannikko R, Wong L, Tester DJ, et al. Dysfunction of NaV1.4, a skeletal muscle voltage-gated sodium channel, in sudden infant death syndrome: a case-control study. Lancet. Apr 14 2018;391(10129):1483-1492.
Bissay V, Van Malderen SC, Keymolen K, et al. SCN4A variants and Brugada syndrome: phenotypic and genotypic overlap between cardiac and skeletal muscle sodium channelopathies. Eur J Hum Genet. Mar 2016;24(3):400-407.
Dehghani-Samani A, Madreseh-Ghahfarokhi S, Dehghani-Samani A. Mutations of Voltage-Gated Ionic Channels and Risk of Severe Cardiac Arrhythmias. Acta Cardiol Sin. Mar 2019;35(2):99-110.
Feng Y, Liu X, Tan H. Causal association of peripheral immune cell counts and atrial fibrillation: A Mendelian randomization study. Front Cardiovasc Med. 2022;9:1042938.
Sun H, Jiang J, Gong L, et al. Voltage-gated sodium channel inhibitor reduces atherosclerosis by modulating monocyte/macrophage subsets and suppressing macrophage proliferation. Biomed Pharmacother. Dec 2019;120:109352.
Hidaka R, Miyazaki K, Miyazaki M. The E-Id Axis Instructs Adaptive Versus Innate Lineage Cell Fate Choice and Instructs Regulatory T Cell Differentiation. Front Immunol. 2022;13:890056.
Shang S, Yang C, Chen F, et al. ID1 expressing macrophages support cancer cell stemness and limit CD8(+) T cell infiltration in colorectal cancer. Nat Commun. Nov 23 2023;14(1):7661.
Stehr AM, Wang G, Demmler R, et al. Neutrophil extracellular traps drive epithelial-mesenchymal transition of human colon cancer. J Pathol. Apr 2022;256(4):455-467.
He L, Liu R, Yue H, et al. Interaction between neutrophil extracellular traps and cardiomyocytes contributes to atrial fibrillation progression. Signal Transduct Target Ther. Jul 26 2023;8(1):279.

Tables 1 to 2 are available in the Supplementary Files section

No competing interests reported.

Table1.xlsx
Table 1、Characteristics of datasets in this study Raw files from three registered microarray datasets, namely GSE31821, GSE41177, and GSE135445, were retrieved from the NCBI GEO database (https://www.ncbi.nlm.nih.gov/geo/). All datasets originated from the Affymetrix Human Genome U133 Plus 2.0 Array [HGU133_Plus_2] microarray platform. Within each dataset, samples of human left atrial appendage (LAA) and epicardial adipose tissue (EAT) were selected from individuals with Atrial Fibrillation (AF) and those in Sinus Rhythm (SR). Subsequently, 36 AF samples and 8 SR samples were included for further analyses.
Table2.xlsx
Table 2、Primers used in the present study for quantitative polymerase chain reaction analysis Primers used in the present study for quantitative polymerase chain reaction analysis.

Download PDF

Version 1

posted

You are reading this latest preprint version

Crosstalk between Atrial Cardiomyocytes and Epicardial Adipose Tissue in Atrial Fibrillation: Insights from Machine Learning Methods and Rat Atrial Fibrillation Model

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and Methods

2.1 Atrial fibrillation datasets

2.2 Data preprocessing

2.3 Identification of differentially expressed genes (DEGs)

2.4 GO and KEGG Enrichment Analysis

2.5 Identification of Key DEGs as Hub Genes in AF using Machine Learning Methods

2.6 CIBERSORT Analysis

2.7 Rat Atrial Fibrillation Model Workstation and Reverse transcription‑quantitative polymerase chain reaction (RT‑qPCR) analysis

2.8 Statistical Analysis.

3. Result

3.1. Determination of DEGs in AF.

3.2 GO and KEGG Enrichment Analysis.

3.3. Determination and Verification of hub genes

3.4 Expression and Diagnostic Significance of ID1, SCN4A, COL4A5, COLEC11, and SNAI2 in AF

3.5 Association of ID1, SCN4A, COL4A5, COLEC11, and SNAI2 with Immunocyte Infiltration Levels

3.6 Rat Atrial Fibrillation Model Workstation and the DEGs of RT-qPCR

4. Disscussion

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1