The experimental animals were male New Zealand white rabbits (Department of Zoology, Yanbian University; weight: 2.5–3 kg; n = 6 rabbits per group). The rabbits were anesthetized with 20% urethane (5 ml/kg) via the ear vein. The right external jugular vein was identified along the lower edge of the sternocleidomastoid muscle, and a “V” incision was made in the right external jugular vein. An electrode catheter was inserted into the right atrium along the right external jugular vein, and a J-shaped guidewire was inserted into the sleeve. The guidewire was slowly pushed out of the sleeve and fixed. The J-type guidewire is connected to the cathode of the electrode to stimulate the atrium. The electrode anode is implanted at the strongest beating of apex cordis as a reference electrode. Aatrial beating is stimulated by monopolar pacing. Animal experiments did comply with the ARRIVE guidelines, and all experimental protocols were approved by the animal ethics committee of Yanbian University. The secretion of ANP involves endocrine processes, and its secretion is related to the regulation of endocrine secretion. In contrast, most female mammals have a physiological cycle, in which there is a relationship between the physiological cycle and ANP secretion. It is difficult for us to control their physiological cycle, in order to avoid the disturbance of ANP secretion by the presence of the physiological cycle, we have studied the mechanism of ANP secretion using male animals. So the results from this study are focused on male rabbits, and female rabbits remain to be further investigated.

After preparation of the aforementioned experimental animals, an RM6240 physiological function experimental system (RM6240BD; Chengdu, Sichuan, China) was used to connect the limb leads and detect the normal ECG waveform. After stabilization, the positive electrode was connected to the J-shaped guidewire and subjected to electrical stimulation. The stimulation intensity was 3.7 V, the pulse width was 0.5 ms, the delay was 0 s, and the frequency was 10 Hz. The animals were randomly divided into two groups: the control group (P0, only intubated without electrical stimulation, n = 6) and the rapid pacing 8 h group (P8, n = 6). The atrial effective refractory period (AERP) was measured using an electrophysiological instrument with the S₁S₂ program incremental stimulation method. The S₁S₂ refractory period was scanned incrementally (8:1, step size: 2 ms), and the longest S₁S₂ interval that could not be transmitted down to the atrium was considered the AERP. AERP₁₅₀ and AERP₂₀₀ in the control and pacing groups were detected (the circumferences of S₁S₂ basic stimulation were 150 and 200 ms, respectively) (Figure 1).

For preparation of the isolated beating perfused rabbit atrial model, New Zealand white rabbits were used (n = 6 rabbits/group). Briefly, 20% urethane (5 ml/kg) was injected intravenously through the ear margin for anesthesia, and the isolated beating perfused rabbit atrial model was prepared as described previously (Cho et al., 1993). For this model, the abdominal cavity was opened, and the heart was removed and rinsed in normal saline. The left atrium was removed and quickly fixed using an atrial perfusion device developed by Cho et al. (1995). Next, the atrium inserted into the cannula was placed into a perfusion cylinder containing HEPES buffer (36.5°C). Oxygen was delivered to the irrigation fluid through a silicone tube on the device. HEPES buffer was perfused into the atrium at a flow rate of 1 ml/min using a peristaltic pump. The composition of the buffer was as follows: distilled water, 118 mM·NaCl, 4.7 mM KCl, 2.5 mM CaCl₂, 1.2 mM MgCl₂, 25 mM NaHCO₃, 10 mM glucose, and 0.10% bovine serum albumin. Electrode stimulation was applied through the atrial wall immediately after perfusion, with the following stimulation parameters: voltage, 35 V; wave width, 0.5 ms; low-frequency, 1.5 Hz; and high frequency, 4.0 Hz. The atrial pulse pressure was measured using a pressure sensor connected to the atrial catheter and recorded in an RM6240 biological signal acquisition system. The atrial stroke volume was calculated based on the change in the water column of the transparent cannula during atrial contraction.

Before the formal experiment, each atrium was perfused for 60 min to stabilize ANP secretion. The perfusate was collected at 2 min intervals at 4°C. The experiment cycle was 12 min.

Animals were divided into four groups (n = 6 rabbits/group; Figure 2), as follows: first, the low-frequency stimulation group was subjected to 1.5 Hz stimulation. The perfusate was collected as a control for the first 36 min. 8-Br-cGMP (1 mmol/L, ab141449, abcam) was then added, and the perfusate was collected for an additional 24 min. Then, the high-frequency stimulation group was subjected to 4.0 Hz stimulation. The perfusate was collected as a control for the first 36 min. 8-Br-cGMP was added, and the perfusate was collected for an additional 24 min. Next, for the high-frequency stimulation plus Akt inhibitor LY294002 group, the perfusate was collected as a control during the first 12 min. LY294002 (10 μmol/L, ab120243, Abcam) was added, and the perfusate was collected for 24 min. 8-Br-cGMP was added, and the perfusate was collected for 24 min. Finally, for the high-frequency stimulation plus GSK-3β inhibitor SB216763 group, the perfusate was collected as a control during the first 12 min. SB216763 (10 μmol/L, ab120202, Abcam) was added, and the perfusate was collected for 24 min. 8-Br-cGMP was added, and the perfusate was collected for 24 min.

After the RAP model was made, the atrial tissue from rabbits was removed immediately, fixed in 10% formalin fixation solution for 24 h, and then rinsed with tap water for 12 h. Alcohol gradient dehydration was performed, and the heart tissues were then embedded in paraffin, cut into 4-μm thick sections, and baked in an oven at 56°C overnight. The sections were then stained using an H&E staining kit (Beijing Solarbio Technology Co., Ltd., Beijing, China).

The rabbit’s atrial tissues were removed immediately after the RAP model was made, immersed in a precooled 2.5% glutaraldehyde solution, and cut into 1-mm³ tissue blocks, fixed in 2.5% glutaraldehyde for 2 h. Subsequently, samples were washed with precooled phosphate-buffered saline (PBS) three times, fixed with osmic acid for 2 h, and washed again with precooled PBS three times. The samples were then dehydrated with 50%, 70%, and 90% ethanol at 4°C for 30 min and penetrated with 90 and 100% acetone. Finally, the samples were embedded, sliced at a thickness of 70 nm, and stained with 3% uranium acetate lead citrate. Transmission electron microscopy was performed.

After 8 h of electrical stimulation, the atrial tissue from rabbits was removed immediately, and the atrial tissue was cut up. The atrial tissue was homogenized at 4°C by three 30-s bursts of maximal speed using a tissue tearor, and then centrifuged at 16,000 rpm at 4°C for 10 min. The supernatant was collected, and then ELISA was performed using an ELISA kit (Nanjing Sbjbio Biotechnology Co., Ltd., Nanjing, China), according to the manufacturer’s instructions. The absorbance was measured at 450 nm by an enzyme-labeling instrument. The cGMP concentration in atrial tissue was calculated according to the absorbance of cGMP based on a standard curve.

ANP standard, ¹²⁵I-ANP, and rabbit anti-ANP antibodies were dissolved and diluted with buffer from a ¹²⁵I-ANP radioimmunoassay kit (Beijing North Institute of Biology Co., Ltd., Beijing, China). ANP standard substance was dissolved with ANP buffer, according to ¹²⁵I-ANP radioimmunoassay kit instructions and a solution with a final concentration of 4.0 ng/ml was obtained, and dilution was made (4.0, 2.0, 1.0, 0.5, and 0.125 ng/ml). 1) Number the special test tubes, total T (total tracer test tubes), NBS (non-specific immune test tubes), S0 (control tubes), S1-6 (standard test tubes), and sample tubes to be tested. 2) Before adding sample, all samples to be tested and ¹²⁵I-ANP, rabbit anti-ANP antibody, buffer solution, and other reagents were mixed at 4°C, respectively. Added samples are as shown in the table (Table 1). 3) After the samples were added, they were mixed and incubated in a refrigerator at 4°C for 16–24 h. 4) All tubes were added with 500 μl of donkey antirabbit immune isolate except the total T. After incubating for 15 min at room temperature, centrifuge at 4°C for 18 min at 3,500 rpm and discard the supernatant. The radioactivity in each sediment was detected by using a counter.

The atrial tissues of rabbits in the control group and the rapid pacing stimulation group were collected right after the RAP model was made. The atrial tissue was placed in ice-cold phosphate buffer, and the atrial tissue was cut up. The atrial tissue was homogenized at 4°C by three 30 s bursts of maximal speed using a tissue tearor. The homogenate was centrifuged at 1,000 g for 10 min at 4°C, and supernatant collected. The supernatant was recentrifuged at 40,000 g for 60 min at 4°C, and precipitation was collected (membrane pellet). The membrane pellet was washed three times with Tris-HCl buffer containing EDTA. Protein contents were determined by a BCA kit (Thermo Fisher Scientific, MA, United States). After the assay was completed, it was divided into equal amounts of protein suspensions. Then aliquots of the protein suspension were incubated for 15 min at 37°C in Tris-HCl buffer, containing 3-isobutyl-1-methylxanthine (IBMX, Sigam I5879), GTP, ATP, creatine phosphate, creatine phosphokinase, MgCl₂, and ANP. Incubations were stopped by adding ice-cold sodium acetate and boiling for 5 min. The samples were then centrifuged at 10,000 g for 5 min at 4°C. ANP was added to the experimental group, and only Tris-HCl buffer was added to the blank group as the control group.

A total protein extraction kit (Millipore, Billerica, MA, United States) was used to obtain total protein from atrial tissues. BCA working solution (Thermo Fisher Scientific, MA, United States) was used to determine protein concentrations. Bromophenol blue was added, and the samples were boiled to denature the protein. Proteins were then separated by SDS-PAGE gel electrophoresis and transferred to polyvinylidene fluoride membranes. Membranes were blocked with 5% skim milk powder for 1.5 h and then incubated overnight at 4°C with the following primary antibodies: anti-NPR-A [1:1000], anti-VASP [1:1000], anti-p-VASP [(Ser 239) 1:1000], anti-Akt [1:1000], anti-p-Akt [(Thr 308) 1:1000], anti-GSK-3β [1:1000], anti-p-GSK-3β [ (Ser 9) 1:1000], and anti-β-actin [1:5000]; all from Santa Cruz Biotechnology, Santa Cruz (CA, United States). The membranes were then washed with TBST and incubated with HRP-secondary antibodies at room temperature for 1.5 h. Finally, images were acquired using enhanced chemiluminescence solution.

ImageJ software was applied for analysis of images. Data were statistically analyzed using GraphPad Prism software. Statistical significance was evaluated by t-tests, one-way ANOVA, and tow-way ANOVA. Data were presented as mean ± SE. Results with p < 0.05 were considered significant.

The forelimbs of rabbits trembled with electrode stimulation, and incontinence occurred during the late stages of the experiment. AERP₁₅₀ and AERP₂₀₀ were significantly shorter in the pacing group than that in the control group (Figures 3A,B), whereas AERPd increased (Figure 3C). Meanwhile, the changes in the heart structure in rabbits (Figures 3D–G) were confirmed. The cell structure was complete, and the myocardial fiber was evenly distributed and arranged vertically and horizontally (Figure 3D). The cell boundary distribution was unclear. The arrangements of myocardial fiber were disordered, and the normal network structure was lost and leukocyte infiltration was also observed (Figure 3E). The ultrastructure of cardiomyocytes in the P0 is intact (Figure 3F). The P8 showed incomplete nuclear membranes, disappearance of gap links, expansion of the sarcoplasmic reticulum, and discontinuity of the bilayer membrane (Figure 3G).

To determine the influences of RAP on ANP and cGMP in the rabbit atrium, we detected changes in ANP and cGMP by RIA and ELISA. Compared with the P0, secretion of atrial ANP increased significantly in the P8 (Figure 4A), whereas the concentration of cGMP decreased significantly (Figure 4B). Thus, rapid pacing stimulated ANP secretion and inhibited cGMP secretion in rabbits.

Studies have shown that ANP combined with NPR-A to activate PGC to increase the expression of cGMP (Moghtadaei et al., 2016). To test the NPR-A level in the rabbit atrium during RAP, we detected the NPR-A protein level using Western blotting. The results showed that there was no difference in the level of NPR-A between P0 and P8 (Figures 5A,B). We further investigated the cGMP production in the atrial membrane by RIA and found that the cGMP production among P8 groups decreased in the ANP group, though there was no significant difference between P0 and P8 in the vehicle group (Figure 5C). The production of cGMP relates to the activity of PGC. It showed that the PGC activity is decreased when the sensitivity of NPR-A to ANP is reduced.

To determine the influences of rapid pacing on the Akt/GSK-3β signaling pathway, we detected the Akt/p-Akt and GSK-3β/p-GSK-3β protein levels using Western blotting. We found that the phosphorylation levels of Akt and GSK-3β decreased significantly (Figure 6) among the atrium of rabbits in P8 compared with that in P0.

Frist, we measured VASP-phosphorylation as indicator of cGMP activity (Figure 7). In the isolated atrial perfusion model, the secretion of ANP in the perfusion fluid was detected by radioimmunoassay. Shown in Figures 8A,B are significant increase of ANP with high-frequency stimulation, while ANP secretion was inhibited after adding cGMP. Also shown in Figures 8C,D are the atrial stroke volume high-frequency stimulation reduced with high-frequency stimulation, and adding cGMP could further decrease the atrial stroke volume. In addition, the atrial pulse pressure was reduced by high-frequency stimulation and was further reduced by adding cGMP (Figures 8E,F).

To examine whether the Akt inhibitor LY294002 affects cGMP function, we made the isolated atrial perfusion model, LY294002 was added for two cycles after one cycle of 4.0 Hz stimulation and the perfusion fluid was collected for detection. We found that there were no significant changes in ANP secretion or atrial dynamics after adding LY294002 (Figure 9). Then the Akt inhibitor and cGMP were added for two cycles, and the perfusion fluid was collected for detection. The results showed that LY294002 reduced the inhibitory effects of cGMP on ANP secretion and enhanced the inhibitory effects of cGMP on atrial dynamics (Figure 9).

We also performed another test to determine the effects of the GSK-3β inhibitor SB216763 as GSK-3β is the downstream of Akt. The isolated atrial perfusion model was made, and the GSK-3β inhibitor SB216763 was added for two cycles after one cycle of 4.0 Hz stimulation, then the perfusion fluid was collected for detection. No significant changes in ANP secretion or atrial dynamics after the addition of SB are shown in Figure 10 . We next add SB216763 and cGMP for two cycles, then collected the perfusion fluid for detection. We found that SB216763 blocked the inhibitory effects of cGMP on ANP secretion and enhanced the inhibitory effects of cGMP on atrial dynamics (Figure 10).

To determine the changes in the cGMP/Akt/GSK-3β pathway after the isolated atrial perfusion experiment, we detected the Akt/p-Akt and GSK-3β/p-GSK-3β protein levels in atrial tissues by Western blotting. We found that Akt and GSK-3β phosphorylation levels were decreased significantly in atrial tissues from rabbits in the high-frequency stimulation group compared with that in the control group (Figure 11). After adding cGMP, Akt, and GSK-3β phosphorylation levels were higher than that in the control group (Figure 11). Furthermore, the Akt and GSK-3β phosphorylation levels were decreased after adding LY294002 and cGMP to the high-frequency stimulation group. When SB216763 and cGMP were added to the high-frequency stimulation group, the phosphor-Akt did not change; however, phosphor-GSK-3β was decreased (Figure 11). These results were consistent with the results from the RAP model, suggesting that the cGMP/Akt/GSK-3β signaling pathway was inhibited, and ANP secretion was regulated by the cGMP/Akt/GSK-3β signaling pathway during RAP.