The experiments were performed in strict accordance with the recommendation in the Guide for the Care and Use of Laboratory Animals from the Institute for Laboratory Animal Research, National Research Council, Washington, D.C., National Academy Press, 2011(13). Moreover, all experimental protocols were approved by the Experimental Animal Care and Institutional Animal Ethical Committee of Guizhou Medical University (Guiyang, China).

In total, 88 male Sprague-Dawley rats weighting between 200-250 g, were purchased from Hunan Silaike Jingda Laboratory Animal Co., Ltd. The rats were acclimatized at 25˚C with a 12-h light/dark cycle and relative humidity of between 40 and 70%, and had free access to food and water.

The rats were randomly divided into the following four groups: i) A sham-operated group comprising of healthy rats without heart failure that received 4 weeks of intraperitoneal saline treatment; ii) a myocardial infarction group (MI group) comprising of rats with coronary ligation-induced heart failure that received 4 weeks of intraperitoneal saline treatment; iii) an MI plus Exendin-4 group (MI+Ex-4 group) comprising of rats with heart failure that received an intraperitoneal injection of 10 µg/kg/day (14) exendin-4 (the dose was selected based on the results of a previous study by the authors) (9) for 4 weeks and iv) an MI plus exendin9-39 and exendin-4 group (MI+Ex-4+Ex9-39 group) comprising of rats with heart failure that received an intraperitoneal injection of 150 µg/kg/day (15) exendin9-39 and 10 µg/kg/day exendin-4 for 4 weeks. Each group comprised 22 rats. The drugs and saline were administered beginning at 24 h after surgery, and 1/2 of the total dose was administered every 12 h. The experimental groups and treatment schedules are presented in Fig. 1.

The coronary ligation procedure performed herein was described in a previous study by the authors (16). The rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (40 mg/kg, Sigma-Aldrich; Merck KGaA). After full anesthesia was achieved when pain reflex disappeared, artificial respiration was initiated using a volume-constant rodent ventilator. The chest was opened through a left thoracotomy at the third intercostal space, after which left anterior descending (LAD) artery ligation was performed using a 7-0 silk suture. The rats in the sham-operated group underwent the same surgical procedure, but did not undergo coronary artery ligation.

Echocardiography (Vevo 770; Visual Sonics) was performed under anesthesia with an intraperitoneal injection of pentobarbital sodium (40 mg/kg, Sigma-Aldrich; Merck KGaA) to assess the left ventricular function 4 weeks after the administration of saline/Ex-4/Ex-4+Ex9-39. Atrial analyses included left atrial (LA) dimension at end diastole and systole and LA fractional shortening. A pulse-wave Doppler was used to assess mitral valve flow as an index of diastolic function.

Systemic hemodynamic analyses were measured with 2.0 F Millar catheters inserted into the right carotid artery and then advanced into the left ventricle in rats anesthetized with an intraperitoneal injection of pentobarbital sodium (40 mg/kg, Sigma-Aldrich; Merck KGaA). Following the hemodynamic measurements, the rats were euthanized by cervical dislocation under the aforementioned anesthesia. The lungs and tibiae of the rats were dissected and measured for the lung weight/tibia length (LW/TL, mg/mm) ratio calculation.

The rats (n=12 in each group) were anesthetized as aforementioned, heparinized (heparin sodium, 400 U, i.p., Sigma-Aldrich; Merck KGaA). Their hearts were rapidly harvested, rinsed and perfused in a Langendorff perfusion system (ADInstruments) with oxygenated HEPES-buffered Tyrode solution containing the following components (mmol/l): NaCl, 135; KCl, 5.4; CaCl₂, 1.8; MgCl₂, 1; Na₂HPO4, 0.3; HEPES, 10 and glucose, 10 (pH 7.3) at a constant pressure (70-90 cm H₂O) and temperature (37.0±0.5˚C). The hearts were perfused for 20 min prior to microelectrode array recording, monophasic action potential (MAP) recordings and atrial tachyarrhythmia susceptibility.

MEA recording was conducted on the Langendorff-perfused hearts for extracellular electrophysiological analysis and excitation propagation observations, which were performed 4 weeks after MI. An array of 32 monopolar electrodes (32 Map) capable of covering a 3x3-mm square (inter-electrode distance of 500 µm) was placed on the left atrial appendage (LAA). The 32 Map signals were obtained and processed using Cardio 2D_2.6.2 and Cardio 2D+_2.4.2 software (Multi Channel Systems MCS GmbH). The total activation time (TAT) was defined as the longest possible period in which the excitation was passed from the first excited electrode to the last excited electrode. The TAT difference of each heartbeat was measured for the TAT dispersion as an index of conduction homogeneity. TAT dispersion was estimated by the standard deviation of the total activation time in the mapped area in 60 sec. Conduction velocity was assessed by linear regression analysis of the isochrones distance vs. activation time in MEA recording (17).

A pair of platinum stimulating electrodes were placed on the epicardial surface of the right atrial appendage for programmed stimulation, and custom-made-Ag-AgCl electrodes consisting of two 0.25-mm Teflon-coated silver wires were positioned on the LAA for MAP recording. The 20, 50 and 90 action potential durations (APD₂₀, APD₅₀ and APD₉₀) were calculated at a pacing cycle length (PCL) of 250 msec by performing at least eight successive MAPs. Burst pacing (2-msec pulse width at 20 Hz, 2-sec burst duration) was used to induce atrial tachyarrhythmias, which were defined as continuous atrial contractions lasting >2 sec. Sustained atrial tachyarrhythmia was defined as contractions lasting >30 sec.

The rats were euthanized by cervical dislocation under anesthesia (intraperitoneal pentobarbital sodium, 40 mg/kg, Sigma-Aldrich; Merck KGaA) following the administration of the drugs for 4 weeks. The hearts were excised and arrested in diastole with 10% KCl, fixed in 10% buffered formalin at room temperature for 24 h and embedded in paraffin. The tissues were subsequently cut into 3-µm thick sections and stained with Masson's trichrome at room temperature for 2 h to assess interstitial collagen deposition. All heart sections were examined, and images were captured via light microscopy (magnification, x100 and x400; BX-50; Olympus Corporation). Data analysis was performed using Image-Pro Plus 6.0 software (Media Cybernetics, Inc.). The collagen volume fraction (CVF) was calculated as a percentage of the interstitial collagen area.

Plasma GLP-1 (n=6; cat. no. EZGLP1T-36K; Sigma-Aldrich; Merck KGaA) and ANP (n=6; cat. no. E-EL-R0017; Elabscience, Inc.) concentrations were assessed using post-caval blood samples using enzyme immunoassay kits.

Total protein was extracted from the rat LAA tissue using RIPA lysis buffer (Beyotime Institute of Biotechnology). Protein concentration was examined using a Bicinchoninic Acid Protein Assay kit (cat. no. AS1086; Wuhan Aspen Biotechnology Co., Ltd.). Protein samples (40 µg/lane) were loaded onto an 8-12% gel, separated using SDS-PAGE and transferred onto polyvinylidene difluoride membranes. The membranes were subsequently incubated with primary antibodies against the GLP-1 receptor (cat. no. ab111125; 1:1,000, Abcam), collagen I (Col I; cat. no. ab34710; 1:200; Abcam), collagen III (Col III; cat. no. ab184993; 1:2,000; Abcam), transforming growth factor (TGF)-β1 (cat. no. ab179695; 1:600; Abcam), total AKT (1:1,000, Abcam, ab8805), phosphorylated AKT (cat. no. ab38449; 1:2,000, Abcam), total PI3K (cat. no. ab154598; 1:1,000, Abcam) and phosphorylated PI3K (cat. no. ab191606; 1:200; Abcam) overnight at 4˚C. Following which, membranes were incubated with horseradish peroxidase (HRP)-conjugated rabbit anti-goat immunoglobulin G (IgG) (cat. no. AS1108; 1:10,000; Wuhan Aspen Biotechnology Co., Ltd.) or HRP-conjugated goat anti-rabbit IgG (cat. no. AS1107; 1:10,000; Wuhan Aspen Biotechnology Co., Ltd.) secondary antibodies for 2 h at room temperature. Chemiluminescence reagent (ECL kit; Beyotime Institute of Biotechnology) was used to visualize the proteins, and quantitative densitometric analysis was performed using Image-Pro Plus 6.0 software (Media Cybernetics, Inc.).

RT-qPCR was used to analyze GLP-1 receptor mRNA expression. Total RNA was isolated from the LAA tissues using a Total RNA Extraction kit (RN0401; Aidlab Biotechnologies Co., Ltd.). RNA samples (4 µg) were reverse transcribed into cDNA using the Transcriptor First Strand cDNA Synthesis kit (cat. no. R101-01; Vazyme Biotech Co., Ltd.). The reverse transcription temperature protocol was as follows: 25˚C for 5 min, 50˚C for 15 min, 85˚C for 5 min and 4˚C for 10 min. qPCR was performed using using SYBR Green master mix (Roche Diagnostics). The following thermocycling conditions were used for the qPCR: 50˚C for 2 min, 95˚C for 10 min; followed by 40 cycles of denaturation at 95˚C for 30 sec, annealing and extension at 60˚C for 30 sec. The data were analyzed using the 2^-ΔΔCq method (18). GLP-1 receptor mRNA expression levels were normalized to GAPDH expression levels. The following primers used were used in the present study: GAPDH, forward: 5'-ACAGCAACAGGGTGGTGGAC-3' and reverse: 5'-TTTGAGGGTGCAGCGAACTT-3; and GLP-1 receptor, forward: 5'-TCTTCTGCAACCGAACCTTT-3' and reverse: 5'-TGCCCTTGGAGCACACTACT-3.

Statistical analysis was performed using SPSS 22.0 software (IBM Corp). All experiments were independently performed three times. Numerical data are expressed as the means ± standard error of the mean and were analyzed using one-way analysis of variance (ANOVA), followed by the Bonferroni's multiple comparison test. Categorical data, which included ATA incidence and sustained ATA incidence, were analyzed by the χ² test and Fisher's exact test. P<0.05 was considered to indicate a statistically significant difference.

Post-MI left atrium dimensions were increased and atrial fractional shortening decreased compared with sham rats (Fig. 2A). Furthermore, the LW/TL ratios and plasma ANP concentration were increased following MI (Fig. 2B). Hemodynamic measurements revealed reduced contractility (dP/dtmax) and relaxation (dP/dtmin) in the rats with MI compared with sham rats (Fig. 2C). These abnormalities were attenuated by exendin-4 treatment. Moreover, the co-administration of exendin9-39 and exendin-4 diminished the protective effects of exendin-4 on cardiac function.

To determine whether exendin-4 treatment exerts effects on atrial electrophysiology, the APDs and atrial tachyarrhythmia inducibility were measured in Langendorff-perfused hearts. It was found that the APD₅₀ and APD₉₀ were prolonged following MI compared with sham rats; however, the APD₅₀ and APD₉₀ were shorter in the exendin-4 group compared with the MI group (Fig. 3A). Additionally, atrial tachyarrhythmias occurred in two out of 12 sham-operated hearts and 11/12 hearts with MI. However, atrial tachyarrhythmias occurred in five out of 12 hearts treated with exendin-4 (Fig. 3B). Furthermore, the incidence of sustained atrial tachyarrhythmias inducibility was decreased in the hearts treated with exendin-4 compared with the hearts with MI (Fig. 3B). The effects of exendin-4 on atrial electrophysiology were abolished by the co-administration of exendin9-39.

Representative excitation transmission maps of the surface of the LAA in each group are presented in Fig. 4A. In the rats in the sham-operated group, propagation was conducted rapidly from the first excited electrode to the last excited electrode. Following MI, the signal proceeded slowly, a change accompanied by an increased total activation time and a decreased conduction velocity compared with sham rats (Fig. 4C). Exendin-4 treatment increased the conduction velocity and decreased the total activation time compared with MI rats, which were indicative of improvement in conduction (Fig. 4C). Representative tracings for TAT dispersion in the four groups are presented in Fig. 4B. Compared with the sham-operated hearts, the hearts with MI exhibited prolonged TAT dispersion (Fig. 4B and C). Exendin-4 treatment restored TAT dispersion to a level similar to that of the control value (Fig. 4B and C), and there was no significant difference between sham rat and MI+Ex-4 rats in TAT dispersion. As shown in Fig. 4A, the activation pattern was uniform in the sham-operated hearts; however, the activation pattern was more heterogeneous in the hearts with MI. Exendin-4 treatment ameliorated the anisotropic conduction properties. However, the GLP-1R antagonist partly diminished the beneficial effects of exendin-4 on excitation transmission.

As atrial fibrosis is one of the main atrial structural alterations responsible for the maintenance and progression of AF (19), the expression levels of profibrotic markers in the atrium were analyzed. Specifically, Col I, Col III and TGF-β1 expression levels were examined using western blotting and the CVF was assessed by Masson's trichrome staining. It was demonstrated that the Col I, Col III and TGF-β1 protein expression levels, and the CVF, were higher in the rats with MI compared with those who underwent the sham operation (Fig. 5A and B). However, these increases were attenuated by exendin-4 treatment (Fig. 5A and B). The expression of Col I, Col III and TGF-β1 protein, and the CVF were increased in MI+Ex-4+Ex9-39 rats compared with MI+Ex-4 rats. Thus, the inhibition of GLP-1R by exendin9-39 partly diminished the suppressive effects of exendin-4 on atrial fibrosis (Fig. 5A and B).

To determine whether the GLP-1 receptor plays a role in atrial arrhythmogenesis in the MI-induced heart failure model, atrial GLP-1 receptor expression and plasma GLP-1 levels were assessed using western blotting analysis and ELISA. The plasma GLP-1 levels and atrial GLP-1 receptor mRNA and protein expression levels were significantly downregulated following MI (Fig. 6A and B). Exendin-4 treatment increased the plasma GLP-1 levels, and the atrial GLP-1 receptor mRNA and protein expression levels (Fig. 6A-C). However, these effects were partly neutralized by co-treatment with exendin9-39 (Fig. 6A-C).

To investigate the mechanisms underlying the effects of exendin-4 on atrial fibrosis, the PI3K/AKT signaling pathway was assessed. Compared with the atrial tissues from sham-operated rats, the tissues from rats with MI exhibited increased phosphorylated PI3K and AKT levels compared with sham rats, although the total PI3K and AKT levels remained unaltered between the two groups. The increased expression levels of phosphorylated PI3K and AKT were attenuated by exendin-4 treatment (Fig. 6B and C), and all the effects of exendin-4 were neutralized by the co-administration of exendin9-39 (Fig. 6B and C).