Abstract

   Atrial fibrillation (AF) is the most prevalent arrhythmia in clinical
   practice, and obesity serves as a significant risk factor for its
   development. The underlying mechanisms of obesity-related AF remain
   intricate and have yet to be fully elucidated. We have identified FPR2
   as a potential hub gene involved in obesity-related AF through
   comprehensive analysis of four transcriptome datasets from AF patients
   and one transcriptome dataset from obese individuals, and its
   expression is up-regulated in both AF and obese individuals.
   Interestingly, ANXA1, the endogenous ligand of FPR2, was found to
   exhibit differential expression with AF and obesity. Specifically, it
   was observed to be down-regulated in AF patients but up-regulated in
   obese individuals. The susceptibility to AF in obese mice induced by
   high-fat diet (HFD) was increased following with the FPR2 blocker
   Boc-2.The administration of exogenous ANXA1 active peptide chain Ac2-26
   can mitigate the susceptibility to AF in obese mice by attenuating
   atrial fibrosis, lipid deposition, oxidative stress injury, and
   myocardial cell apoptosis. However, this protective effect against AF
   susceptibility is reversed by AAV9-shAMPK-mediated AMPK specific
   knockdown in the myocardium. The vitro experiments demonstrated that
   silencing ANXA1 exacerbated lipid deposition, oxidative stress injury,
   and apoptosis induced by palmitic acid (PA) in cardiomyocytes.
   Additionally, Ac2-26 effectively mitigated myocardial lipid deposition,
   oxidative stress injury, and apoptosis induced by PA. These effects
   were impeded by FPR2 inhibitors Boc-2 and WRW4. The main mechanism
   involves the activation of AMPK by ANXA1 through FPR2 in order to
   enhance fatty acid oxidation in cardiomyocytes, thereby ultimately
   leading to a reduction in lipid accumulation and associated
   lipotoxicity. Our findings demonstrate that the ANXA1-FPR2 axis plays a
   protective role in obesity-associated AF by alleviating metabolic
   stress in the atria of obese mice, thereby emphasizing its potential as
   a promising therapeutic target for AF.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12933-024-02545-z.

   Keywords: Atrial fibrillation, Obesity, ANXA1-FPR2, AMP-activated
   protein kinase, Lipotoxicity

Introduction

   Atrial fibrillation (AF) is the most common sustained arrhythmia that
   impairs quality of life and increases the risk for stroke, dementia,
   heart failure, and sudden cardiac death [[54]1]. Because of the
   multiformity of potential contributors and the limited understanding of
   the molecular mechanisms driving AF and its progression, the current
   rhythm control strategies is inadequate in some patients [[55]2]. The
   heart is the greatest energy-demanding organ, with 60 to 90% of its
   energy supplied by mitochondrial oxidation of long-chain fatty acids
   (FAs) and the remainder by glucose, lactic acid and ketone bodies
   [[56]3]. Metabolic activity is closely related to arrhythmias by
   affecting excitation-contraction coupling and ion channel/pump function
   [[57]4]. Impaired cellular energetic status not only predisposes to
   atrial arrhythmias, but atrial rhythm disturbances also influence
   metabolic activity [[58]5]. The rapid rates of electrical and
   contractile activity during AF will affect atrial energy demands,
   circulation and oxygen supply, and change the balance between metabolic
   demand and supply, which is metabolic stress to atria and a driving
   force for atrial electrical and structural remodeling [[59]3,
   [60]5–[61]7]. A growing body of evidence suggests that correcting
   cellular metabolism deficiencies may represent a novel and promising
   antiarrhythmic strategy [[62]5, [63]8].

   Obesity is a well-established risk factor for AF and the prevalence of
   which is increasing globally [[64]9–[65]11]. Although growing evidences
   have demonstrated that weight management is effective for primary and
   secondary AF prevention, the pathophysiological mechanisms linking
   obesity and AF remain unclear [[66]12, [67]13]. There are pathological
   metabolic abnormalities in cardiomyocytes in the condition of obesity,
   characterized by impaired glucose metabolism, increased beta-oxidation
   of FAs, lipid accumulation, defective mitochondrial oxidative capacity,
   and increased mitochondrial reactive oxygen species (ROS), which
   contribute to atrial remodeling and the triggering and maintenance of
   AF [[68]8, [69]11, [70]14, [71]15]. Our recent studies have shown that
   improving metabolic abnormalities of cardiomyocytes in obese mice
   induced by high-fat diet (HFD) can ameliorate the arrhythmogenic
   substrate and AF susceptibility [[72]6, [73]15, [74]16]. Therefore a
   comprehensive understanding of the interrelated mechanism of metabolic
   and arrhythmogenic substrate promises to identify new molecular targets
   for the treatment of obesity-induced AF.

   The formyl peptide receptor 2 (FPR2) belongs to the G protein-coupled
   receptor (GPCR) family; however, unlike typical GPCRs, it exhibits a
   remarkable ability to recognize diverse ligands with distinct
   structural characteristics, encompassing formyl peptides, non-formyl
   peptides, lipids, small molecules, and proteins [[75]17]. Human FPR2
   and its mouse homolog FPR2 are highly expressed in a variety of cells,
   including cardiomyocytes, and interact with peptide and lipid ligands
   to modulate inflammatory response, cytokine expression and lipid
   metabolism [[76]17–[77]20]. The latest research findings indicate that
   resolvin-D1, one of the ligands of FPR2, exhibits potential in
   preventing AF caused by left ventricular dysfunction. The underlying
   mechanism suggests that it may exert its effects through the inhibition
   of myocardial tissue inflammation via the FPR2 receptor [[78]21,
   [79]22]. These results highlight the therapeutic prospects of targeting
   the FPR2 receptor for AF treatment. Annexin A1 (ANXA1) is a member of
   the annexin superfamily and an endogenous ligand of FPR2. ANXA1-FPR2
   axis can alleviate the corresponding organ dysfunction by improving
   metabolic disturbance in different pathological conditions [[80]19,
   [81]20, [82]23]. However, the role of ANXA1-FPR2 in obesity-induced
   atrial metabolic disorders and its effect on obesity-related AF
   susceptibility remain unclear.

   In this study, we applied bioinformatics tools to execute conjoint
   analysis of AF and obesity transcriptome datasets with the goal of
   identifying key molecules involved in the pathological process of
   obesity-related AF. The results of bioinformatics analysis suggested
   that ANXA1-FPR2 axis may perform crucially in the pathology of
   obesity-induced AF. Therefore, using a mouse model of obesity induced
   by HFD, we further explored the role of ANXA1-FPR2 axis on
   obesity-induced AF susceptibility and elucidated its underlying
   mechanism. The experimental design flow is illustrated in (Supplemental
   Fig. [83]1A).

Materials and methods

Retrieve the transcriptome datasets of AF and obesity in gene expression
omnibus (GEO)

   The transcriptome datasets about AF and obesity were retrieved from the
   GEO Datasets. Four AF related atrial transcriptome datasets screened in
   our previous study were used for analysis, one of which was excluded
   because it was not on the same sequencing platform as the other
   datasets [[84]24]. We take “obesity” and “atrium” as keywords for the
   retrieval in GEO Datasets. The “Homo sapiens” and “Expression profiling
   by array or Expression profiling by high throughput sequencing” were as
   filter conditions for “Organism” and “Study type” respectively. The
   selection criteria for transcriptome datasets of obesity were as
   follows: (1) tissue samples of atria or appendage obtained from obese
   and matched controls without AF, (2) datasets were mRNA microarray or
   high-throughput sequencing results, and (3) the number of samples in
   each group was not less than three. According to the above screening
   criteria, an obesity-related atrial transcriptome dataset,
   [85]GSE159612, was included in the study [[86]25]. The detail
   information of these five GEO datasets were listed in (Supplemental
   Table [87]1).

Data processing and differentially expressed genes (DEGs) analysis

   The raw data (CEL file) of microarrays related to AF was processed as
   previously described [[88]24]. Prior to integrative meta-analysis,
   individual datasets were normalized by log2 transformation and
   R-mediated mean, followed by quantile normalization. The meta-analysis
   was performed using NetworkAnalyst [[89]26], a web-based tool for
   statistical integration and visualization. The processed and normalized
   datasets were uploaded to the “multiple gene expression data” input
   area in the web interface option of multiple gene expression data. They
   were then subjected to well-established ComBat procedures for removing
   study-specific batch effects. Principal component analysis (PCA) was
   employed to visualize and compare the clustering patterns of samples
   with and without batch effect adjustment, in order to evaluate the
   effectiveness of batch effect removal. The DEGs were identified through
   Fisher’s method (p <.05), fixed effects model (p <.001), and Vote
   Counting (Votes number < 2 and p <.05). Subsequently, the robust DEGs
   between the AF and control groups were identified by intersecting the
   DEGs obtained from each model algorithm. The “Define Custom Signature”
   tool in NetworkAnalyst was utilized to generate a heatmap visualization
   of the top 10 up- and downregulated genes. The merged data for
   meta-analysis using NetworkAnalyst is lised in (Supplemental Table
   [90]2).

   The gene expression matrix obtained from high-throughput sequencing of
   atrial tissue in individuals with obesity was downloaded. Inter-group
   differences were assessed using the limma algorithm, and original p
   values were adjusted using the Benjamini-Hochberg (BH) method. Gene
   expression fold changes (FC) were computed by subtracting the mean of
   normalized log2-based expression levels of the respective control
   groups from each subject’s normalized log2-based expression levels DEGs
   were filtered using an adjusted p <.05 and|log2 FC| > 0.5. Common DEGs
   between AF and obesity were identified. DEGs that were up-regulated or
   down-regulated in both AF and obesity groups were defined as
   obesity-related AF DEGs.

Functional enrichment analysis of DEGs

   Metascape ([91]https://metascape.org) was utilized to conduct Gene
   Ontology (GO) functional analysis and Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathway enrichment analysis [[92]27]. The cutoff
   criterion was set at p <.01.

Construction of protein–protein interaction (PPI) network and hub gene
screening

   STRING database ([93]http://string-db.org) was used to establish the
   PPI networks [[94]28]. The hub genes were identified through the use of
   the CytoHuba plugin within Cytoscape software (version 3.8.0), with the
   degree algorithm being employed to identify these hub genes in this
   study [[95]29, [96]30].

Gene set expression analyses

   Gene set enrichment analyses (GSEAs) were conducted using the Molecular
   Signature database (MSigDB) “c2.cp.kegg.v7.2.symbols.gmt” gene set
   [[97]31] to evaluate the enrichment of canonical pathways between FPR2
   high-expression and low-expression groups, utilizing Gene Set
   Enrichment Analyses software (Java version 4.0) [[98]32]. Signaling
   pathways with a normalized p <.05 and a false discovery rate (FDR)
   q < 0.25 were deemed significantly enriched.

Animals and experimental design

   Male C57BL/6J mice (6 weeks old) were purchased from Xi’an Jiaotong
   University (Xi’an, China). The animals were housed in a controlled
   environment with specific conditions (temperature: 22 ± 2 °C; relative
   humidity: 50% ± 10%; light cycle: 12-hour light/dark cycle with lights
   on at 7:00 a.m.) and provided ad libitum access to food and water. All
   experiments were performed in strict accordance with the
   recommendations of the Guide for the National Institutes of Health
   Guidelines for the Use of Laboratory Animals and approved by the Animal
   Care Committee of Xi’an Jiaotong University (The study received ethical
   approval from the Ethics Committee of Xi’an Jiaotong University, with
   an assigned approval number: 2022-922).

   After a one-week acclimatization period, a total of sixty mice were
   randomly assigned to four groups (n = 15 per group): Standard diet
   (STD; 10% fat, 70% carbohydrate, 20% protein; Cat. no. D12450B;
   Research Diets Inc.), STD and Ac2-26 (STD + Ac2-26), HFD (HFD; 60% fat,
   20% carbohydrate, 20% protein; Cat. no. D12492; Research Diets Inc.),
   and HFD + Ac2-26 (HFD + Ac2-26). A mouse model of obesity-induced AF
   was established by feeding an HFD for 10 weeks when obesity and AF
   susceptibility changed significantly in the HFD mice compared to the
   STD mice. The mice were intraperitoneally injected with Ac2-26
   (1 mg/kg; Topscience, Shanghai, China), an ANXA1 mimetic peptide, or
   vehicle for a duration of 10 weeks [[99]33]. At the end of the
   experiment, electrophysiological examination, echocardiography, and
   intraperitoneal glucose tolerance testing (IPGTT) were performed on
   each group before tissue sampling. After an overnight fast, atrial
   tissues and blood samples were collected from euthanized mice for
   further analysis.

Recombinant viral vector construction and injection

   The Adeno-associated virus (AAV) particles were procured from
   GenePharma Co., Ltd (Shanghai, China). Short hairpin sequences were
   synthesized as previously reported [[100]34], and cloned into the AAV9
   vector (AAV9-shAMPK 5′-ATGATGTCAGATGGTGAATTT
   AAGTTCTCTAAACCTACCATCTGACAT-3′ and AAV9-shNC: 5′-ACGACGTCAGCT GGTGCA GT
   AGTTCTCTACATGCACCAGCTGACGTCGT-3′). After 8 weeks of high-fat diet
   feeding, AVV9 (1.7 × 10^13 vg/mL) was administered via intravenous
   injection through the tail vein. After 4 weeks of intervention with
   AAV9 and Ac2-26, cardiac electrophysiological examination,
   echocardiography and IPGTT were performed.

Cell isolation, culture and treatment

   The neonatal rat cardiomyocytes (NRCMs) were isolated from the hearts
   of male neonatal Sprague-Dawley rats aged 1 to 3 days old. In brief,
   atrial tissue was surgically excised and subsequently digested with
   collagenase II (0.1%; Solarbio, Beijing, China) to obtain a cell
   suspension. NRCMs were then isolated through differential adherence and
   verified under a microscope. The isolated cells were then cultured in
   6-well plates at 37 °C with Dulbecco’s Modified Eagle Medium (DMEM)
   supplemented with 10% Fetal Bovine Serum (FBS) and 1%
   penicillin/streptomycin solution, while maintaining a CO[2]
   concentration of 5%. At 80% confluence, the cells were stimulated with
   palmitic acid (PA) at a final concentration of 200 µM for 24 h to
   simulate the effects of lipid overload [[101]15]. Cells were
   pre-treated with FPR2 antagonists WRW4 (10 µM, Topscience) or Boc-2 (10
   µM, Topscience) prior to being treated with 0.3 µM Ac2-26 (Topscience)
   [[102]35]. Compound C (CC, 0.5 µM; Sigma-Aldrich) was added to inhibit
   AMPK activation.

   The NRCMs were transfected with siRNAs targeting rat ANXA1 (GenePharm,
   Shanghai, China) using Entranster^TM-R4000 siRNA Transfection Reagent
   (Engreen Biosystem, Beijing, China) according to the manufacturer’s
   instructions. The sequence of ANXA1 siRNA is
   5’-CAAGAUCUAACCAGCAAAUTT-3’.

Electrophysiological and echocardiography examinations

   AF inducibility was evaluated through programmed transesophageal
   electrical stimulation, as previously described with minor
   modifications [[103]24]. Briefly, after inducing anesthesia with
   intraperitoneal injection of sodium pentobarbital (60 mg/kg), a
   continuous surface lead electrocardiogram (ECG) was recorded using a
   physiological signal-acquisition system (RM6240; Chengdu Instrument
   Factory). AF was then induced through high-frequency burst pacing
   (pulse width 1 ms; 2× threshold current; 30, 35, and 40 Hz,
   respectively) using a programmable electrical stimulator (VCS3001;
   MappingLab Ltd.). AF was defined as a period of rapid and irregular
   atrial rhythm lasting for at least 1 s on ECG.

   Noninvasive transthoracic echocardiography was performed using a Vevo
   2100 Imaging System (VisualSonics Inc.) to discern differences in
   cardiac structure and function between groups. All images were
   subsequently processed with Vevolab 3.1 software (VisualSonics).

IPGTT

   The glucose homeostasis was assessed by IPGTT. Overnight fasted mice
   were intraperitoneally injected with D-glucose (2 g/kg·BW), and blood
   glucose levels were measured at 0, 15, 30, 60, 90 and 120 min after
   glucose loading by the ACCU-CHEK Performa Glucometer (Roche
   Diagnostics, Indianapolis, United States) using tail-vein blood. The
   glucose tolerance was determined by calculating the area under the
   curve (AUC) using GraphPad Prism 8.0 software (GraphPad Software, San
   Diego, USA).

Detection of biomarkers in serum and tissue samples

   Biomarkers were quantified using commercially available kits in
   accordance with the manufacturer’s instructions, including fasted
   nonesterified fatty acids (NEFA) (Solarbio), insulin (Cusabio, Wuhan,
   China), malondialdehyde (MDA) (Beyotime Institute of Biotechnology,
   Shanghai, China), and superoxide dismutase (SOD) (Beyotime Institute of
   Biotechnology).

Analysis of histological staining and fluorescence

   Atrial tissue samples were fixed in 4% paraformaldehyde overnight,
   subsequently embedded in paraffin and sectioned into slices that were
   5 μm thick. Atrial sections were stained with Masson’s trichrome,
   selected antibodies (ANXA1; 1:200; Santa Cruz Biotechnology,
   California, United States and FPR2; 1:200; Santa Cruz Biotechnology),
   and TUNEL (in situ cell death detection kit, TMR red; Roche
   Diagnostics). Moreover, frozen Sect. (10 μm thickness) of atria samples
   were prepared and stained with oil red O. The sections were examined
   and captured under a light microscope (Nikon, Melville, New York,
   United States) or a fluorescence microscope (Leica, Bensheim, Germany),
   followed by further analysis using ImageJ software (National Institutes
   of Health).

Quantitative real-time PCR (RT-qPCR) analysis

   Total RNA was extracted from fresh atrial tissue using TRIzol (AG,
   Changsha, China) following the manufacturer’s instructions.
   Subsequently, single-stranded cDNA was synthesized with the assistance
   of Evo M-MLV RT Kit (AG). The SYBR green chimeric fluorescence method
   (Accurate Biology) was employed for the RT-qPCR reaction on a CFX96™
   Real-Time PCR System (Bio-Rad, Hercules, CA, United States). The
   2^−△△Ct comparative method was used to quantify results which were then
   normalized by β-actin. Primer sequences are listed in (Supplemental
   Table [104]3).

Western blot

   Atrial protein lysates were extracted and quantified using a
   bicinchoninic acid (BCA) protein assay The proteins underwent sodium
   dodecyl sulfate polyacrylamide gel electrophoresis and were
   subsequently transferred onto a polyvinylidene difluoride membrane,
   which was then incubated with primary antibodies targeting ANXA1
   (1:1000; ZEN-BIOSCIENCE, Chengdu, China), FPR2 (1:1000; Santa Cruz
   Biotechnology), Bax (1:1000; ZEN-BIOSCIENCE), Bcl2 (1:1000;
   ZEN-BIOSCIENCE), cytochrome complex (Cyt C; 1:1000; Santa Cruz
   Biotechnology), caspase3 (1:1000; Cell Signaling Technology/CST),
   phosphorylated c-Jun N-terminal kinase (p-JNK; 1:1000; CST), t-JNK
   (1:1000; Proteintech), Acetyl-CoA carboxylase (Acc; 1:1000;
   Proteintech), phosphorylated Acc (p-Acc; 1:1000; CST), AMP-activated
   protein kinase (AMPK; 1:1000; CST), phosphorylated AMPK (p‐AMPK;
   Thr172; 1:1000; CST), peroxisome proliferator–activated receptor alpha
   (PPARα; 1:1000; Santa Cruz Biotechnology), carnitine
   palmitoylransferase-1B (CPT1B; 1:1000; Proteintech), and
   Glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 1:1000; Proteintech).
   Subsequently, the membranes were incubated with secondary antibodies
   for 1–2 h. The resulting bands were detected using an ECL Kit
   (Millipore) and imaged with a Tanon imaging system (Tanon).
   Densitometry analysis was performed using Quantity One software
   (Bio-Rad) to quantify the results.

Statistical analysis

   The data were presented as mean ± SEM for animal or cell experiments,
   and all statistical analyses were conducted using GraphPad Prism
   software (version 8.0). The Kolmogorov-Smirnov normality test was
   performed to assess the distribution of the data. If the data followed
   a normal distribution, unpaired two-tailed Student t-tests were used to
   compare two groups, while One-way ANOVA or Two-way ANOVA with
   Bonferroni correction was employed for comparing multiple groups. In
   case of non-normal distribution, the Kruskal-Wallis test was utilized
   for between-group comparisons. A significance level of p <.05 was
   considered statistically significant.

Results

The hub gene FPR2 may be closely associated with obesity-related AF

   Between the AF and control groups, a total of 3323 DEGs were identified
   using the Fixed Effect Model algorithm in the Meta-analysis module
   (Supplemental Table [105]4), while 5003 DEGs were detected by Fisher’s
   method algorithm model (Supplemental Table [106]4). The Vote Counting
   algorithm model screening yielded a set of 3013 DEGs (Supplemental
   Table [107]4). We obtained a robust list of 1083 DEGs by taking the
   intersection of these sets (Fig. [108]1A, Supplemental Table [109]4),
   and (Fig. [110]1C) shows the top 10 up- and downregulated genes.
   Differential analysis identified 7394 DEGs between the obesity and
   control groups (Fig. [111]1B, Supplemental Table [112]5). Subsequently,
   we screened for 133 DEGs with synchronous expression trends in both AF
   and obesity patients by intersecting the DEGs from these two groups
   (Supplemental Table [113]6). Of these, 132 were significantly
   up-regulated while only one was down-regulated in both AF and obesity
   (Fig. [114]1D). Through the online analysis tool Metascape, functional
   enrichment analysis of DEGs with synchronous expression trends in AF
   and obesity revealed their significant involvement in regulating
   inflammatory response, apoptosis, and GPCR downstream signaling
   pathways (Fig. [115]1E).

Fig. 1.

   [116]Fig. 1
   [117]Open in a new tab

   Identification and validation of hub genes in obesity-related AF. A The
   DEGs were identified through Fisher’s method (p <.05), fixed effects
   model (p <.001), and Vote Counting (Votes number < 2 and p <.05).
   Subsequently, the robust DEGs between the AF and control groups were
   identified by intersecting the DEGs obtained from each model algorithm.
   B The volcano plot display DEGs in obesity patients and the control
   group. C The heat map displays the top 10 significantly up-regulated
   and down-regulated DEGs between patients with AF and controls. D 133
   DEGs with the same expression trend in AF and obesity patients. E
   Construction of PPI network and F hub gene screening. G, H
   Representative sections and quantification of FPR2 using
   immunofluorescence staining in the atrium of obese mice. I, J
   Representative sections and quantification of ANXA1 using
   immunofluorescence staining in the atrium of obese mice. K The
   concentration of ANXA1 in the serum of mice in each group. Data are
   shown as mean ± SEM. *p <.05; *** p <.001 between indicated groups.
   n = 5 for immunofluorescence staining analysis. The scale bar = 50 μm.
   AF, atrial fibrillation; DEGs, differentially expressed genes; PPI,
   protein–protein interaction

   The DEGs were utilized to construct a PPI network, and the hub gene of
   the network was identified as FPR2 due to its high degree of
   connectivity (Fig. [118]1F). The expression of FPR2 is significantly
   upregulated in the atria of both AF and obese patients. A receiver
   operating characteristic (ROC) curve was constructed using FPR2 gene
   expression from integrated AF and control transcript datasets,
   revealing that FPR2 expression has good diagnostic value for AF (AUC
   0.758, 95%CI 0.635–0.871) (Supplemental Fig. [119]1B). Additionally, AF
   patients were stratified into a low FPR2 expression group and a high
   FPR2 expression group based on the median value of FPR2 gene expression
   for further single-gene GSEA enrichment analysis. The results indicated
   that the high FPR2 expression group was primarily enriched in
   lymphocyte differentiation, extracellular matrix (ECM) receptor
   interaction, adipocytokine signaling pathway, lipid metabolism, and
   PPAR signaling pathway (Supplemental Fig. [120]1C-D). The findings
   imply that FPR2 may participate in the pathophysiological mechanisms
   underlying atrial inflammation and metabolism in patients with AF.

Validation of FPR2 and its endogenous ligand ANXA1

   The expression levels of FPR2 were investigated in a mouse model of
   HFD-induced obesity. Immunofluorescence and Western blot results
   revealed significant upregulation of FPR2 expression in the atria of
   HFD-induced obese mice compared to the control group (Fig. [121]1G-H,
   Supplemental Fig. [122]1E-F), which was consistent with the result of
   our bioinformatics analysis. We conducted further analysis on ANXA1,
   the endogenous ligand of FPR2, and observed no significant difference
   in its expression level within the integrated AF database (Supplemental
   Fig. [123]1G). However, in another high-throughput sequencing dataset
   for AF ([124]GSE69890), we discovered a notable down-regulation of
   ANXA1 expression in the atria of AF patients compared to that of the
   control group (Supplemental Fig. [125]1H, Supplemental Table [126]7).
   We further conducted a measurement of serum ANXA1 concentrations in
   patients with AF and matched controls, enrolling 30 AF patients and 26
   matched controls through screening (The study received ethical approval
   from the Ethics Committee of the Second Affiliated Hospital of Xi’an
   Jiaotong University, with an assigned approval number: 2021197). The
   basic information is presented in (Supplemental Table [127]8). Our
   findings indicate that serum ANXA1 levels were significantly lower in
   AF patients compared to the control group (Supplemental Fig. [128]1I).
   However, in the obesity-related dataset [129]GSE159612, ANXA1
   expression in the atria of obese patients was found to be significantly
   increased compared to that of the control group (Supplemental
   Fig. [130]1J). Moreover, both atrial ANXA1 expression and serum ANXA1
   concentration were significantly upregulated in HFD-induced obese mice
   when compared to the control group (Fig. [131]1I-K, Supplemental
   Fig. [132]1E, K).

Ac2-26 attenuates obesity-induced AF susceptibility and insulin resistance

   To investigate the potential involvement of the ANXA1-FPR2 axis in the
   pathogenesis of obesity-induced AF, we initially administered Boc-2, an
   inhibitor of FPR2, to obese mice for a 10 weeks (Supplemental
   Fig. [133]2A). The findings revealed that while there were no
   significant disparities in the susceptibility and duration of AF
   between the obese mice treated with Boc and those treated without it,
   the intervention with Boc-2 exhibited an ability to augment both the
   susceptibility and duration of AF (Supplemental Fig. [134]2B-D).
   Furthermore, Masson staining indicated that Boc-2 could exacerbate
   atrial fibrosis in obese mice (Supplemental Fig. [135]2E-F). The
   subsequent investigation focused on evaluating the impact of an
   additional 10-week supplementation of Ac2-26, an active peptide chain
   of ANXA1, on the susceptibility to AF in mice with diet-induced obesity
   (Fig. [136]2A). The baseline characteristics of the mice in each group
   are summarized in (Table [137]1). HFD mice exhibited progressive weight
   gain over time, whereas treatment with Ac2-26 resulted in a reduction
   of body weight specifically in obese mice but not lean counterparts
   (Fig. [138]2B-C). The AF susceptibility in obese mice was significantly
   higher than controls, and this phenomenon was significantly reduced
   with Ac2-26 (Fig. [139]2D-F).

Fig. 2.

   [140]Fig. 2
   [141]Open in a new tab

   The administration of Ac2-26 improves insulin resistance and reduces
   susceptibility to AF in obese mice induced by HFD. A Experimental
   design of this study. Mice were randomly divided into four groups: STD
   group, STD + Ac2-26 group, HFD group, and HFD + Ac2-26 group. B Weight
   gain curve and C the change of body weight. D Representative ECGs
   showing AF induction by transesophageal burst pacing. E AF frequency
   and F AF duration were assessed between the groups. G Fasting blood
   glucose, H random blood glucose, and I fasting serum insulin among
   groups. J HOMA-IR (insulin resistance index) were calculated. K GTT and
   L calculated AUC among groups. Data are shown as mean ± SEM. *p <.05;
   **p <.01; *** p <.001 between indicated groups. n = 6–14 for animal
   experiments. AF, atrial fibrillation; AUC, area under the curve; HFD,
   high‐fat diet; HOMA‐IR, homeostasis model assessment of insulin
   resistance; GTT, glucose tolerance test

Table 1.

   Physical characteristics, echocardiographic, electrophysiological
   parameters, and metabolism-related indicators in mice from STD,
   STD + Ac2-26, HFD, and HFD + Ac2-26 groups
   Parameter STD
   (n = 8–11) STD + Ac2-26
   (n = 8–12) HFD
   (n = 9–14) HFD + Ac2-26
   (n = 9–14)
   Physical characteristics
   BW (g) 28.16 ± 0.36* 26.40 ± 0.40 40.66 ± 1.15*^# 33.43 ± 1.27^#
   AW (mg) 14.22 ± 1.23 13.15 ± 1.16 15.75 ± 1.42 14.17 ± 1.13
   HW (mg) 143.01 ± 2.75 133.91 ± 4.85 147.14 ± 3.48 141.35 ± 2.59
   HW/BW (mg/g) 5.08 ± 0.10* 5.07 ± 0.17 3.67 ± 0.17*^# 4.31 ± 0.20^#
   BL (cm) 10.51 ± 0.06 10.45 ± 0.12 10.58 ± 0.18 10.85 ± 0.05
   HW/BL (mg/cm) 13.61 ± 0.26 12.78 ± 0.36 13.92 ± 0.35 13.02 ± 0.26
   Echocardiography
   IVSd (mm) 0.89 ± 0.05 0.86 ± 0.06 0.96 ± 0.06 0.83 ± 0.06
   IVSs (mm) 1.33 ± 0.05 1.27 ± 0.06 1.45 ± 0.07 1.20 ± 0.06
   LVIDd (mm) 3.44 ± 0.13 3.22 ± 0.16 3.22 ± 0.10 3.38 ± 0.11
   LVIDs (mm) 2.24 ± 0.12 2.12 ± 0.16 2.07 ± 0.13 2.19 ± 0.15
   LVPWd (mm) 0.93 ± 0.02 0.92 ± 0.03 1.01 ± 0.05^# 0.81 ± 0.04^#
   LVPWs (mm) 1.30 ± 0.06 1.29 ± 0.07 1.43 ± 0.06 1.27 ± 0.07
   LVEF (%) 65.95 ± 2.40 65.55 ± 3.294 66.20 ± 3.21 65.44 ± 3.89
   LVFS (%) 34.89 ± 1.78 34.62 ± 2.44 35.41 ± 2.37 35.75 ± 2.98
   Electrophysiology
   P (ms) 20.09 ± 0.57 20.00 ± 0.84 21.58 ± 0.74^# 19.63 ± 0.50^#
   PR (ms) 37.87 ± 1.06* 42.08 ± 1.21 43.35 ± 1.33* 41.67 ± 1.37
   QRS (ms) 23.12 ± 1.82 23.47 ± 2.13 24.86 ± 0.56 20.82 ± 1.08
   QT (ms) 65.36 ± 1.32 66.94 ± 1.97 68.05 ± 1.42 65.14 ± 1.01
   SACT (ms) 20.12 ± 1.13 20.10 ± 3.38 24.67 ± 6.29 21.50 ± 2.10
   AVERP140 (ms) 66.89 ± 2.60 66.11 ± 1.624 63.50 ± 4.22 64.50 ± 3.76
   Metabolism-related indicators
   TG (mmol/l) 0.96 ± 0.05* 0.95 ± 0.05 1.24 ± 0.07*^# 0.97 ± 0.05^#
   NEFA (mmol/l) 0.52 ± 0.05* 0.49 ± 0.08 0.77 ± 0.06*^# 0.55 ± 0.05^#
   MDA (µmol/mg prot) 8.60 ± 0.58* 8.22 ± 0.48 12.44 ± 0.98*^#
   9.77 ± 0.45^#
   SOD (U/mg prot) 520.70 ± 0.58* 564.27 ± 0.82 455.80 ± 5.96*^#
   501.67 ± 1.83^#
   [142]Open in a new tab

   AVERP, atrioventricular node effective refractory period; AW, atria
   weight; BL, body length; BW, body weight; HFD, high-fat diet; HW, heart
   weight; IVSd, interventricular septal thickness at end-diastole; IVSs,
   interventricular septal thickness at end-systole; LVIDd, left
   ventricular internal diameter at end-diastole; LVIDs, left ventricular
   internal diameter at end-systole; LVPWd, left ventricular posterior
   wall thickness at end-diastole; LVPWs, left ventricular posterior wall
   thickness at end-systole; LVEF, left ventricular ejection fraction;
   LVFS, left ventricular fractional shortening; NEFA, nonesterified fatty
   acids; SACT, sinoatrial conduction time; TG, triglycerides. All data
   are expressed as mean ± SEM and n represents number of animals in each
   group. *p <.05, STD versus HFD; ^#p <.05, HFD versus HFD + Ac2-26

   To investigate whether metabolic disorders, particularly insulin
   resistance, contribute to the Ac2-26-induced reduction in AF
   susceptibility in obese mice, we assessed the degree of insulin
   sensitivity among groups using measures such as fasting blood glucose
   (FBG), fasting serum insulins (FINs), GTT and homeostasis model
   assessment of insulin resistance (HOMA-IR). The HFD mice exhibited
   apparent insulin resistance, as evidenced by elevated FBG, FINs,
   HOMA-IR values, and impaired glucose tolerance compared to the controls
   (Fig. [143]2G-L). In mice treated with Ac2-26, the detrimental effects
   of HFD on insulin sensitivity were mitigated, as evidenced by a
   reduction in FBG and FINs, an improvement in HOMA-IR score and glucose
   tolerance (Fig. [144]2G-L). The findings suggest that Ac2-26 may
   mitigate HFD-induced AF susceptibility by ameliorating metabolic
   dysfunction in obese mice.

AMPK mediates Ac2-26 to reduce obesity-induced AF susceptibility

   Previous studies have demonstrated the significant involvement of AMPK
   in AF. Therefore, we conducted an evaluation to investigate whether
   Ac2-26 could mitigate AF susceptibility in obese mice by activating
   AMPK through FPR2. Specifically, AAV9-shAMPK was employed to
   specifically knock down AMPK expression in heart (Supplemental
   Fig. [145]2G-I), followed by a 4-week intervention with Ac2-26. The
   baseline characteristics of each group of mice are presented in
   (Supplemental Table [146]9). No notable differences were observed
   between the two groups regarding body weight, fasting glucose levels,
   and IPGTT results (Fig. [147]3A-E). However, after one month of Ac2-26
   treatment, a remarkable reduction in both AF susceptibility and
   duration was observed in the AAV-shNC group compared to AAV9-shAMPK
   group (Fig. [148]3F-H). Furthermore, the administration of Ac2-26 for a
   duration of one month resulted in a significant reduction in atrial
   diameter and atrial fibrosis within the shNC group, as compared to the
   shAMPK group (Fig. [149]3I-L). The findings suggest that the activation
   of AMPK by Ac2-26, independent of weight loss and improvement in
   systemic insulin resistance factors, plays a crucial role in reducing
   susceptibility to AF in obese mice.

Fig. 3.

   [150]Fig. 3
   [151]Open in a new tab

   AMPK is a key target for Ac2-26 to improve AF susceptibility in obese
   mice. A Weight gain curve and B the change of body weight. Fasting
   blood glucose (C), GTT (D) and (E) calculated AUC among groups. F
   Representative ECGs showing AF induction by transesophageal burst
   pacing. G AF frequency and H AF duration were assessed between the
   groups. I-L Echocardiography and Masson staining showed that Ac2-26 did
   not improve atrial enlargement and fibrosis in AAV9-shAMPK group
   compared with AAV9-shNC group. Data are shown as mean ± SEM. *p <.05
   between indicated groups. n = 3–10 for animal experiments. The scale
   bar = 50 μm. AF, atrial fibrillation; AAV, Adeno-associated virus; AUC,
   area under the curve; HFD, high-fat diet; GTT, glucose tolerance test;
   LAD, left atrial dimension

Ac2-26 attenuates obesity-induced atrial remodeling

   Since electrical and structural remodeling are crucial in the
   initiation and maintenance of AF, we further investigated whether HFD
   could impact atrial electrical and structural remodeling, as well as
   the role of Ac2-26. The baseline Surface ECG (Lead II) indicated that
   HFD prolonged the P-wave duration in obese mice, which was mitigated by
   Ac2-26 treatment (Table [152]1). However, there were no significant
   changes observed in atrioventricular node effective refractory period
   (AVERP) and sinoatrial conduction time (SACT) between Ac2-26-treated
   and untreated mice (Table [153]1). Next, we evaluated the
   transcriptional profiles of pivotal ion channels implicated in AF
   pathogenesis. HFD augmented Cav1.2, Cav1.3, Ryanodine receptor 2
   (RyR2), Kv3.1, Kr/hERG and KATP/Kir6.2 gene expression while Ac2-26
   intervention upregulated Nav1.5, Cav1.2 and Cav1.3 but downregulated
   Kr/hERG in obese mice (Fig. [154]4A-C). The typical atrial action
   potential morphology with the corresponding AF-related ion channel
   transcriptional profile is depicted in (Fig. [155]4D). Due to the
   insignificant impact of Ac2-26 on the up-regulated RyR2 in obese mice,
   we proceeded to investigate potential post-translational modifications
   of RyR2. The Western blot analysis revealed that treatment with Ac2-26
   significantly reduced the phosphorylation level of atrial RyR2 in
   HFD-induced obese mice (Fig. [156]4I-J).

Fig. 4.

   [157]Fig. 4
   [158]Open in a new tab

   Effect of Ac2-26 on atrial remodeling in obese mice. A-C
   Transcriptional profile of AF-related sodium channel (Nav1.5), calcium
   channels (Cav1.2, Cav1.3 and RyR2), and potassium channels (Kv4.3,
   Kv1.5, Kv3.1, Kr/hERG, Ks/Kv7.1, KACH/Kir3.4, Kir2.1, and KATP/Kir6.2)
   among groups (n = 4). D Typical shape of an atrial action potential
   showing principle currents and the corresponding subunits (ion
   channels). E-H The echocardiography and Masson staining results
   revealed left atrial enlargement and a significant increase in fibrosis
   in obese mice, whereas Ac2-26 exhibited a notable improvement in
   structural remodeling (n = 3–9). I, J The administration of Ac2-26
   resulted in a significant reduction in the phosphorylation level of
   RyR2 in obese mice (n = 4). Data are shown as mean ± SEM. *p <.05;
   **p <.01; *** p <.001 between indicated groups. The scale bar = 50 μm.
   HFD, high‐fat diet; RyR2, ryanodine receptor 2; LAD, left atrial
   dimension

   Furthermore, we investigated the atrial structural remodeling. The
   results of echocardiography and Masson staining demonstrated that HFD
   significantly increased atrial diameter and myocardial fibrosis
   compared to the control group, while Ac2-26 intervention notably
   reduced both parameters in obese mice induced by HFD (Fig. [159]4E-H).
   These findings suggest that Ac2-26 can ameliorate atrial electrical and
   structural remodeling in HFD-induced obesity.

Ac2-26 alleviated myocardial lipid accumulation and toxicity by promoting
lipid metabolism

   Further investigation in myocardial tissue revealed a significant
   increase in atrial myocyte apoptosis, as evidenced by TUNEL staining,
   in obese mice (Fig. [160]5A-B). Western blot analysis demonstrated that
   HFD promoted the expression of pro-apoptotic proteins Cyt C and cleaved
   caspase3/caspase3, up-regulated JNK phosphorylation levels, and reduced
   Bcl2/Bax levels, thereby exerting a pro-apoptotic effect. However,
   Ac2-26 significantly inhibited these effects and played an
   anti-apoptotic role in HFD-induced obese mice (Fig. [161]5C-G).
   Furthermore, HFD resulted in an elevation of MDA levels within the
   atria while simultaneously reducing SOD concentrations. Ac2-26 exhibits
   an antioxidative stress effect in the atrium of HFD-induced obese mice
   by reducing MDA levels and up-regulating SOD concentration
   (Table [162]1).

Fig. 5.

   [163]Fig. 5
   [164]Open in a new tab

   The promotion of fatty acid oxidation by Ac2-26 may potentially
   mitigate lipotoxicity in atrial myocytes of obese mice. A, B The tunel
   staining analysis revealed that Ac2-26 significantly attenuated
   apoptosis in atrial myocytes of obese mice (n = 3). C-G The Western
   blot analysis demonstrated that Ac2-26 significantly attenuated the
   expression of pro-apoptotic protein molecules in obese mice (n = 4). H,
   I Representative sections and quantification of lipid accumulation
   using Oil Red O staining (n = 3). J-L Transcriptional profile of,
   glucose metabolism-related genes (GLUT1, GLUT4, HK2, PFKM, PKM2, and
   PDK4), fatty acid metabolism‐related genes (CD36, FABPpm, FABP3, and
   CPT1B) and metabolism regulatory genes (HIF1α, PGC1α, FOXO1, PPARα,
   PPARγ and PPARδ) among groups (n = 4). M-Q Representative images and
   Western blot analysis of Acc, AMPK, CPT1B and PPARα among groups
   (n = 4). Data are shown as mean ± SEM. *p <.05; **p <.01; *** p <.001
   between indicated groups. The scale bar = 50 μm. AMPK, AMP‐activated
   protein kinase; Acc, Acetyl-CoA carboxylase; CD36, fatty acid
   translocase; CPT1B, carnitine palmitoylransferase-1B; FABP-pm, plasma
   membrane fatty acid-binding protein; FABP3, fatty acid binding protein
   3; FOXO1, forkhead box protein O1, GLUT, glucose transporter; HFD,
   high‐fat diet; HIF1α, hypoxia inducible factor-1α; HK2, hexokinase2;
   PDK4, pyruvate dehydrogenase kinase 4; PFKM, phosphofructokinase;
   PGC1α, peroxisome proliferator-activated receptor γ coactivator1α;
   PKM2, pyruvate kinase isozyme type M2; PPAR, peroxisome
   proliferators-activated receptor

   Our previous study has indicated that HFD-induced obesity in mice leads
   to abnormal lipid deposition in the myocardium, resulting in myocardial
   lipid toxicity and contributing to the pathogenesis of AF. Based on
   previous bioinformatics analysis, it has been suggested that FPR2 may
   play a role in lipid metabolism. Therefore, we conducted further
   investigations to determine whether Ac2-26 affects lipid metabolism in
   myocardial tissue. HFD mice showed an increased tendency in serum NEFA
   and TG levels, and Ac2-26 treatment decreased it in obese mice but not
   lean mice (Table [165]1). Furthermore, Oil Red O staining and
   transmission electron microscope revealed that HFD increased lipid
   accumulation in the atria, while Ac2-26 treatment alleviated lipid
   accumulation in HFD mice but not in lean mice (Fig. [166]5H-I and
   Supplemental Fig. [167]2J-K).

   Thereafter, we evaluated the transcription and translation profiles of
   the metabolism-related genes. We observed that treatment with Ac2-26
   resulted in increased transcription of genes related to glucose
   metabolism (pyruvate dehydrogenase kinase 4, PDK4), fatty acid
   metabolism (fatty acid translocase, CD36; plasma membrane fatty acid
   binding protein, FABPpm; CPT1B) and metabolic regulation (PPARα) in
   obese mice (Fig. [168]5J-L). Additionally, HFD was found to
   down-regulate the phosphorylation of Acc and AMPK, while Ac2-26
   up-regulated their phosphorylation (Fig. [169]5M-O). No significant
   effect of HFD on the protein levels of CPT1B and PPARα was observed,
   but Ac2-26 treatment significantly upregulated their expression in the
   atria of obese mice (Fig. [170]5M, P-Q).

ANXA1 knockdown aggravates PA-induced lipotoxicity in NRCMs

   To further investigate the impact of ANXA1 on lipid metabolism and
   lipid toxicity in NRCMs, siRNA was employed to knockdown ANXA1
   expression in NRCMs (Fig. [171]6A-C). Subsequently, the cells were
   treated with 200 µM PA for 24 h to simulate a high-fat environment
   (Fig. [172]6D). Compared to the bovine serum albumin (BSA) control
   group, PA intervention significantly increased cardiomyocyte apoptosis,
   promoted the production of ROS and MDA, and decreased mitochondrial
   membrane potential and SOD concentration in cardiomyocytes
   (Fig. [173]6E-L). Knockdown of ANXA1 further exacerbated ROS and MDA
   production, reduced SOD activity and mitochondrial membrane potential,
   and aggravated cardiomyocyte apoptosis (Fig. [174]6E-L).

Fig. 6.

   [175]Fig. 6
   [176]Open in a new tab

   ANXA1 knockdown aggravates PA-induced lipotoxicity in cardiomyocytes.
   A-C Determination of ANXA1 knockdown efficiency by RT-qPCR and Western
   blot (n = 4). D Experimental design of cell experiments. E, F
   Representative images and statistical analysis of flow
   cytometry-detected apoptosis level (n = 3). (G, H) Representative
   images and analysis of the subcellular localization of the oxidation
   products of DHE (n = 100). I, J Representative images and analysis of
   the mitochondrial membrane potential detected by JC-1 staining
   (red/green) (n = 100). (K-L) Cellular MDA and SOD concentrations among
   the groups (n = 5). M-T Representative images and Western blot analysis
   of t-RyR2, Acc, Ampk, CPT1B, PPARα and caspase3 among the groups
   (n = 4). Data are shown as mean ± SEM. *p <.05; **p <.01; *** p <.001
   between indicated groups. The scale bar = 50 μm. AMPK, AMP‐activated
   protein kinase; Acc, Acetyl-CoA carboxylase; BSA, bovine serum albumin;
   CPT1B, carnitine palmitoylransferase-1B; DHE, dihydroethidium; MDA,
   malondialdehyde; PA, palmitic acid; PPAR, peroxisome
   proliferators-activated receptors; RyR2, ryanodine receptor 2; RT‐qPCR,
   quantitative reverse transcription‐PCR; SOD, superoxide dismutase

   The Western blot results demonstrated that PA significantly attenuated
   AMPK phosphorylation and increased the ratio of cleaved
   caspase3/caspase3 in cardiomyocytes, while no significant effect was
   observed on the expression of t-RyR2, Acc, CPT1B and PPARα
   (Fig. [177]6M-T). Knockdown of ANXA1 further suppressed AMPK
   phosphorylation in PA-treated cardiomyocytes, downregulated PPARα and
   CPT1B protein expression levels in the PA group, and upregulated t-RyR2
   and Acc protein expression levels in both BSA and PA groups
   (Fig. [178]6M-T). In summary, ANXA1 knockdown in cardiomyocytes impeded
   fatty acid catabolism and exacerbated oxidative stress and lipotoxicity
   induced by PA.

ANXA1 attenuates PA-induced lipotoxicity via FPR2

   To further explore the function and mechanism of ANXA1 on
   cardiomyocytes, we conducted in vitro experiments utilizing Ac2-26 in
   conjunction with FPR2 inhibitors Boc-2 and WRW4 (Fig. [179]7A). Our
   findings revealed that high concentrations of Ac2-26 (> 30 µM)
   exhibited cytotoxicity towards NRCMs, whereas concentrations at 0.3 µM
   and 3 µM did not elicit this adverse effect. To minimize the potential
   side effects of Ac2-26, a concentration of 0.3µM was utilized, almost
   consistent with prior literature [[180]35, [181]36]. Ac2-26
   significantly attenuated PA-induced cardiomyocyte apoptosis by
   down-regulating ROS and MDA production, while up-regulating SOD
   concentration and mitochondrial membrane potential. However, these
   effects were attenuated when FPR2 was blocked with the antagonist Boc-2
   or WRW4 (Fig. [182]7B-I).

Fig. 7.

   [183]Fig. 7
   [184]Open in a new tab

   ANXA1 attenuates PA-induced lipotoxicity via FPR2. A Experimental
   design of cell experiments. B-C Representative images and statistical
   analysis of flow cytometry-detected apoptosis level (n = 4). D, E
   Representative images and analysis of the subcellular localization of
   the oxidation products of DHE (n = 100). F, G Representative images and
   analysis of the mitochondrial membrane potential detected by JC-1
   staining (red/green) (n = 100). H, I Cellular MDA and SOD
   concentrations among the groups (n = 3). J-Q Representative images and
   Western blot analysis of Acc, Ampk, CPT1B, PPARα, caspase3 and RyR2
   among the groups (n = 4). Data are shown as mean ± SEM. *p <.05;
   **p <.01; *** p <.001 between indicated groups. The scale bar = 50 μm.
   AMPK, AMP-activated protein kinase; Acc, Acetyl-CoA carboxylase; BSA,
   bovine serum albumin; CPT1B, carnitine palmitoylransferase-1B; DHE,
   dihydroethidium; MDA, malondialdehyde; PA, palmitic acid; PPAR,
   peroxisome proliferators-activated receptors; RyR2, ryanodine receptor
   2; SOD, superoxide dismutase

   Similarly, Ac2-26 was found to decrease the ratio of cleaved
   caspase3/caspase3 and RyR2 phosphorylation induced by PA. Moreover,
   Ac2-26 was observed to enhance the phosphorylation of AMPK and Acc, as
   well as increase the expression of PPARα and CPT1B. These effects were
   reversed by FPR2 blockers (Fig. [185]7J-Q). Therefore, ANXA1 may
   mitigate oxidative stress and lipotoxicity caused by PA in
   cardiomyocytes through promoting fatty acid metabolism via FPR2.

AMPK serves as a crucial effector molecule in the ANXA1-FPR2 pathway to
mitigate cellular lipotoxicity in NRCMs

   AMPK serves as a pivotal regulatory component in cellular energy
   metabolism. While PA can attenuate the phosphorylation level of AMPK in
   cardiomyocytes, ANXA1 is capable of activating it. Therefore, we
   further investigated whether AMPK was involved in ANXA1-FPR2-induced
   cardiomyocyte protective effect by inhibiting AMPK activity using AMPK
   inhibitor Compound C (Fig. [186]8A). We observed that treatment with
   Compound C significantly attenuated the effects of Ac2-26 on PA-induced
   oxidative stress, mitochondrial membrane potential, and apoptosis in
   NRCMs (Fig. [187]8B-I). The Western blot results demonstrated that
   Compound C counteracted the anti-apoptotic effect of Ac2-26 and
   elevated the ratio of Cleaved caspase3/caspase3, while also reinstating
   RyR2 phosphorylation levels (Fig. [188]8J-P). Additionally, Compound C
   significantly decreased the protein expression of both CPT1B and PPARα
   as well as the phosphorylation of Acc that was induced by Ac2-26
   (Fig. [189]8J-P).

Fig. 8.

   [190]Fig. 8
   [191]Open in a new tab

   AMPK functions as a pivotal effector molecule within the ANXA1-FPR2
   signaling pathway. A Experimental design of cell experiments. B-G
   Representative images and statistical analysis were conducted to assess
   the levels of apoptosis detected by flow cytometry, as well as the
   oxidation products of DHE and JC-1 staining (red/green) (n = 4-100). H,
   I Cellular MDA and SOD concentrations among the groups (n = 3). J-P
   Representative images and Western blot analysis of Acc, Ampk, CPT1B,
   PPARα, caspase3 and RyR2 among the groups (n = 4). Q Proposed model of
   ANXA1-mediated protection against lipid accumulation and lipotoxicity.
   Data are shown as mean ± SEM. *p <.05; **p <.01; *** p <.001 between
   indicated groups. The scale bar = 50 μm. AMPK, AMP-activated protein
   kinase; Acc, Acetyl-CoA carboxylase; BSA, bovine serum albumin; CD36,
   fatty acid translocase; CPT1B, carnitine palmitoylransferase-1B; DHE,
   dihydroethidium; FABP-pm, plasma membrane fatty acid-binding protein;
   HFD, high‐fat diet; MDA, malondialdehyde; PA, palmitic acid; PPAR,
   peroxisome proliferators-activated receptors; RyR2, ryanodine receptor
   2; SOD, superoxide dismutase

Discussion

   In this study, we have identified differential expression of ANXA1 and
   FPR2 in the atria of patients with AF and obese individuals for the
   first time. Subsequently, we investigated the role and potential
   mechanism of ANXA1-FPR2 in obesity-related AF. Specifically, FPR2 was
   found to be significantly upregulated in the atria of both AF and obese
   patients, while its endogenous ligand ANXA1 was significantly
   upregulated in obese patients but downregulated in AF. Supplementation
   with the active polypeptide Ac2-26 of ANXA1 significantly mitigated the
   susceptibility to AF in high-fat diet-induced obese mice; however, this
   protective effect was diminished in obese mice with specific myocardial
   AMPK knockdown by AAV9-shAMPK. The mechanism study revealed that
   ANXA1-FPR2 may participate in mitigating myocardial lipotoxicity by
   activating the downstream AMPK signaling pathway to facilitate fatty
   acid catabolism in cardiomyocytes (Fig. [192]8Q).

   ANXA1 and FPR2 are expressed in a variety of cell types, including
   cardiomyocytes, macrophages, and neutrophils. ANXA1 functions as an
   endogenous ligand of FPR2 and is released via autocrine and paracrine
   pathways [[193]37]. The ANXA1-FPR2 axis modulates diverse pathological
   processes, including metabolism, inflammatory response, cell
   proliferation, and cell death [[194]19, [195]37–[196]40]. The role of
   ANXA1-FPR2 in AF has not been previously investigated. After analyzing
   the transcriptome dataset, we observed a significant upregulation of
   FPR2 in the atria of AF patients with obesity. Furthermore, FPR2
   expression was also significantly increased in HFD-induced obese mice,
   which is consistent with previous studies [[197]18, [198]19, [199]41].
   These findings suggest that FPR2 plays a crucial role in the pathology
   of AF and obesity-related AF. Interestingly, the expression level of
   ANXA1, which serves as an endogenous ligand for FPR2, exhibited
   opposite trends in AF patients and obese patients. This phenomenon has
   prompted us to conduct further investigations into the role of
   ANXA1-FPR2 axis in AF, and our findings have indicated that ANXA1-FPR2
   exerts a protective effect against obesity-related AF. Although we
   initially employed the FPR2 blocker Boc-2 to investigate the impact of
   ANXA1-FPR2 blockade on AF susceptibility in obese mice, no
   statistically significant difference was observed. However, it did
   reveal that inhibiting the ANXA1-FPR2 signaling pathway augmented AF
   susceptibility and further exacerbated atrial fibrosis in obese mice.
   Therefore, we attribute the lack of statistical significance to a
   limited sample size. Based on our findings, we postulate that the
   down-regulation of ANXA1 expression in AF patients may create a
   conducive environment for the development of AF by exacerbating its
   pathological progression. This pathological change, in turn, could
   up-regulate FPR2 expression as a response to ANXA1 deficiency. Our
   hypothesis is that the reduction of ANXA1 expression in AF patients may
   be associated with decreased high density lipoprotein levels and
   impaired plasminogen/plasmin system, both of which can hinder ANXA1
   synthesis [[200]42, [201]43]. However, further investigation is
   required to elucidate the exact mechanism.

   There is mounting evidence of an interaction between metabolism and AF.
   During AF episodes, the frequency of atrial electrical and contractile
   activity increases by 4 to 6 times [[202]44], leading to a significant
   rise in atrial energy and oxygen consumption that results in metabolic
   stress on atrial myocytes. Conversely, metabolic dysfunction of these
   cells can further exacerbate atrial remodeling and provide a substrate
   for the initiation and perpetuation of AF [[203]4, [204]5, [205]44].
   Obesity is an independent risk factor for AF, partially attributable to
   metabolic disorders in the atria [[206]13]. Previous metabolomics
   studies have revealed that glucose metabolism is impaired, while fatty
   acid oxidation is predominant and abnormal lipid deposition occurs in
   atrial myocytes of obese mice. This phenomenon can be attributed to
   excessive intake of fatty acids and insufficient β-oxidation [[207]14].
   Similarly, our previous study has demonstrated that insufficient fatty
   acid β-oxidation in the atrium of mice with HFD-induced obesity leads
   to excessive intracellular lipid accumulation and myocardial
   lipotoxicity, including oxidative stress, DNA damage, inflammation, and
   insulin intolerance. The aforementioned factors play a pivotal role in
   the development of cardiac hypertrophy, fibrosis, gap junction
   remodeling, and myocardial injury, thereby contributing to the
   initiation and progression of AF [[208]15]. The present study also
   revealed notable disturbances in glucose and lipid metabolism in obese
   mice, including impaired glucose tolerance, atrial tissue lipid
   accumulation, and oxidative stress injury. However, Ac2-26 was found to
   enhance the expression of receptors and regulators associated with
   glucose and lipid metabolism, ameliorate insulin resistance, mitigate
   atrial lipid accumulation, and alleviate oxidative stress injury in
   obese mice. Similarly, comparable outcomes were observed in NRCMs, and
   the impact of Ac2-26 was hindered by FPR2 blockers. Consistent with our
   findings, previous studies have also demonstrated the potential of
   ANXA1-FPR2 to enhance mitochondrial function and promote fatty acid
   β-oxidation. Additionally, therapeutic effects on lipotoxicity in
   diabetic mice have been observed with the use of ANXA1 mimetic peptide
   Ac2-26 [[209]19]. Furthermore, upregulation of ANXA1 significantly
   suppressed cigarette smoke extract-induced inflammation, apoptosis, and
   oxidative stress in BEAS-2B cells [[210]40]. The presented evidence
   suggests that the ANXA1-FPR2 axis exerts a protective role in
   obesity-related AF by augmenting metabolic status, thereby reducing
   cardiomyocyte lipotoxicity and associated oxidative stress.

   AMPK is an energy sensor, and its role in AF pathogenesis is attracting
   more and more attention. The previous studies have demonstrated that
   the deletion of liver kinase B1 (LKB1-AMPK), an upstream regulator of
   the AMPK superfamily, leads to spontaneous AF in mice. The findings
   suggest a strong association between atrial remodeling in AF and
   metabolic abnormalities as well as oxidative stress, both of which are
   closely linked to the LKB1-AMPK pathway [[211]45]. It is currently
   believed that the induction of metabolic stress by AF itself leads to a
   reduction in ATP levels and an increase in oxidative stress, ultimately
   resulting in atrial remodeling characterized by electrical
   abnormalities, systolic dysfunction, and cardiac structural remodeling
   [[212]7].The activation of AMPK can enhance glycolysis and fatty acid
   oxidation, thereby augmenting energy production while suppressing
   energy expenditure to rectify metabolic abnormalities in AF [[213]46,
   [214]47]. Our discovery that AMPK serves as a crucial downstream
   effector of the ANXA1-FPR2 signaling axis in combating obesity-related
   AF aligns with previous findings observed in other disease models
   [[215]19, [216]40]. Moreover, despite the absence of significant
   differences in body weight and insulin resistance between AAV9-shNC and
   AAV9-shAMPK group, treatment with Ac2-26 significantly attenuated
   susceptibility to AF in AAV9-shNC group compared to the AAV9-shAMPK
   group. This suggests that Ac2-26’s effect on improving susceptibility
   to AF in obese mice is not solely reliant on systemic metabolic
   changes. The activation of AMPK not only promotes Acc phosphorylation,
   thereby reducing intracellular lipid synthesis, but also activates
   PPARα and CPT1B, which accelerates fatty acid β-oxidation. The ultimate
   result was a decrease in lipotoxicity and oxidative stress in
   cardiomyocytes, as evidenced by a reduction in ROS production,
   upregulation of mitochondrial membrane potential, and enhancement of
   apoptosis. The well-established role of oxidative stress in triggering
   activity and potentially facilitating the transition from paroxysmal AF
   to persistent AF is widely recognized [[217]45]. Triggered activity
   arises from an elevated occurrence of spontaneous Ca^2+ release events
   (Ca^2+ leak) through RyR2 located in the sarcoplasmic reticulum (SR)
   [[218]48]. The function of RyR2 is increased by oxidative stress,
   either through direct oxidation of RyR2 or via an oxidation-mediated
   activation of Ca2+/calmodulin-dependent protein kinase-II (CaMKII),
   leading to hyperphosphorylation of RyR2 [[219]49]. The phosphorylation
   level of RyR2 was found to be significantly elevated in our obese mice,
   and it was notably reduced following Ac2-26 treatment, thereby
   providing further evidence for the potential protective role of Ac2-26
   in AF by mitigating metabolic stress in obese mice.

   The obese mice exhibit atrial electrical and structural remodeling,
   characterized by the down-regulation of I[Na] and I[Ca, L] in
   cardiomyocytes, up-regulation of I[Kur], shortening of action potential
   duration (APD), and activation of the transforming growth factorβ
   (TGFβ)-mediated pro-fibrotic signaling pathway in the atria [[220]9,
   [221]50]. Our findings demonstrated a down-regulation of Nav1.5 ion
   channel transcriptional levels, responsible for I[Na], in the atria of
   obese mice; however, no significant difference was observed between the
   groups. Notably, Ac2-26 exhibited a remarkable ability to significantly
   up-regulate their expression in the atria of obese mice. The
   transcriptional levels of hERG and KATP channels were found to be
   upregulated in the atrium of obese mice. However, Ac2-26 only exhibited
   a reduction in the transcriptional level of hERG channel without
   exerting any significant effect on the transcriptional levels of other
   potassium channels. Additionally, a significant upregulation of Cav1.2
   and Cav1.3 transcription levels was observed in obese mice, while
   Ac2-26 further augmented the expression of both Cav1.2 and Cav1.3.
   According to the transcriptional levels of ion channels, Ac2-26 appears
   to have a beneficial effect on prolonging APD in cardiomyocytes.
   However, since we have not investigated the functionality of ion
   channels specifically in cardiomyocytes, it is currently impossible to
   ascertain the precise impact of Ac2-26 on the functional
   characteristics of APD and ERP in these cells. Therefore, further
   studies are warranted to elucidate the role of ANXA1 in modulating ion
   channel function within cardiomyocytes. The previous studies have
   demonstrated that recombinant ANXA1 has the potential to ameliorate
   cardiac diastolic dysfunction in immune dysregulation by modulating the
   phenotypes of fibroblasts and immune cells within the cardiac tissue in
   animal models of rheumatoid arthritis [[222]38]. We also observed that
   ANXA1 in obese mice improved atrial fibrosis. However, since our study
   primarily focuses on myocardial cell metabolism, we did not further
   investigate the mechanism of myocardial fibrosis. Based on our
   findings, we hypothesize that ANXA1 may exert its anti-fibrotic effect
   by augmenting the cardiomyocytes’ anti-apoptotic response. However,
   further investigation is required to elucidate the precise mechanism
   underlying ANXA1’s anti-fibrotic action.

Conclusions

   In conclusion, we have identified FPR2 as an obesity-associated AF hub
   gene through a comprehensive analysis of transcriptome data from AF and
   obesity patients. Notably, our study is the first to demonstrate that
   supplementation of FPR2 endogenous ligand Ac2-26 reduces AF
   susceptibility in obese mice. The detailed mechanism was investigated
   from the perspective of metabolic homeostasis, showing that Ac2-26
   promotes atrial fatty acid metabolism and mitigates oxidative stress
   and lipotoxicity in obese mice by activating AMPK. Nevertheless, more
   functional experiments are warranted to provide deeper insights into
   the underlying mechanism and to support activating ANXA1-FPR2 axis as a
   clinical treatment strategy for AF in the future.

Electronic Supplementary Material

   Below is the link to the electronic supplementary material.
   [223]Supplementary Material 1.^ (2MB, jpg)
   [224]Supplementary Material 2.^ (4.6MB, jpg)
   [225]Supplementary Material 3.^ (15.4KB, docx)
   [226]Supplementary Material 4.^ (13.4MB, xlsx)
   [227]Supplementary Material 5.^ (16.7KB, docx)
   [228]Supplementary Material 6.^ (415.7KB, xlsx)
   [229]Supplementary Material 7.^ (513.3KB, xlsx)
   [230]Supplementary Material 8.^ (10.6KB, xlsx)
   [231]Supplementary Material 9.^ (954KB, xlsx)
   [232]Supplementary Material 10.^ (15.1KB, docx)
   [233]Supplementary Material 11.^ (22.3KB, docx)

Author contributions

   Q. Z., H. W. and X.Q. took responsibility for the study design. P. L.,
   L. W., Y. W., J. F., Y.Z., H. L., B. F., L. J., H. G., Q. W., Y. F., B.
   F. and X. L. performed the experiments; P.L., L. W. and Y. W. analyzed
   the data. P. L. and L. W. prepared the manuscript. All the authors
   participated in drafting the manuscript and approved the final version
   for publication.

Funding

   This work was supported by the National Natural Science Foundation of
   China (Grant Number 81870257, 31871172 and 82400379) and Key Research
   and Development Program of Shaanxi (grant number 2021SF-132).

Availability of data and materials

   No datasets were generated or analysed during the current study.

Declarations

Competing interests

   The authors declare no competing interests.

Footnotes

   Publisher’s note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

   Peng Liu, Lu Wang and Yixin Wang contributed equallyto this work.

Contributor Information

   Hongtao Wang, Email: wht506@126.com.

   Xinghua Qin, Email: xinghuaqin@nwpu.edu.cn.

   Qiangsun Zheng, Email: zhengqiangsun@126.com.

References