Spontaneous Atrial Fibrillation in Transgenic Goats With TGF (Transforming Growth Factor)-β1 Induced Atrial Myopathy With Endurance Exercise
Abstract
Background:
There is increasing evidence that endurance exercise is associated with increased risk of atrial fibrillation (AF). However, it is unknown if the relationship between endurance exercise and AF is dependent on an atrial myopathy.
Methods:
Six cardiac-specific TGF (transforming growth factor)-β1 transgenic and 6 wild-type (WT) goats were utilized for these studies. Pacemakers were implanted in all animals for continuous arrhythmia monitoring and AF inducibility. AF inducibility was evaluated using 5 separate 10 s bursts of atrial pacing (160–200 ms). Three months of progressive endurance exercise (up to 90 minutes at 4.5 mph) was performed. Quantitative assessment of circulating microRNAs and inflammatory biomarkers was performed.
Results:
Sustained AF (≥30 s) was induced with 10 s of atrial pacing in 4 out of 6 transgenic goats compared with 0 out of 6 WT controls at baseline (P<0.05). No spontaneous AF was observed at baseline. Interestingly, between 2 and 3 months of exercise 3 out of 6 transgenic animals developed self-terminating spontaneous AF compared with 0 out of 6 WT animals (P<0.05). There was an increase in AF inducibility in both transgenic and WT animals during the first 2 months of exercise with partial normalization at 3 months (transgenic 67%; 100%; 83% versus WT 0%; 67%; 17%). These changes in AF susceptibility were associated with a decrease in circulating microRNA-21 and microRNA-29 during the first 2 months of exercise with partial normalization at 3 months in both transgenic and WT animals. Finally, MMP9 (matrix metallopeptidase 9) was increased during the second and third months of exercise training.
Conclusions:
This study demonstrates a novel transgenic goat model of cardiac fibrosis (TGF-β1 overexpression) to demonstrate that endurance exercise in the setting of an underlying atrial myopathy increases the incidence of spontaneous AF. Furthermore, endurance exercise seems to increase inducible AF secondary to altered expression of key profibrotic biomarkers that is independent of the presence of an atrial myopathy.
WHAT IS KNOWN?
Extreme endurance exercise can increase risk of atrial fibrillation.
TGF (transforming growth factor)-β1 plays a role in atrial fibrosis development.
WHAT THE STUDY ADDS?
Endurance exercise in the setting of an underlying atrial myopathy can increase risk of atrial fibrillation.
Endurance exercise alters expression of key profibrotic biomarkers (microRNA-21, microRNA-29, and MMP9 [matrix metallopeptidase 9]) independent of the presence of an atrial myopathy.
Introduction
Atrial fibrillation (AF) has high clinical relevance, being the most common sustained clinical cardiac arrhythmia. AF cases in the United States alone are projected to reach 15.9 million by 2050.1 Although AF is the most common arrhythmia in humans, its underlying mechanisms are not completely understood, particularly in the absence of traditional cardiovascular risk factors.2 AF is often the manifestation of an atrial myopathy with accompanying structural (ie, atrial fibrosis) and electrical (ie, altered expression of ion channels) remodeling.2–4 TGF (transforming growth factor)-β1 is an important regulator of cell growth, proliferation, and apoptosis.5 Altered expression of TGF-β1 is thought to play an important role in the development of atrial fibrosis and an arrhythmogenic substrate.6 Recently, we developed a novel TGF-β1 transgenic goat with an underlying atrial myopathy and increased susceptibility to AF when compared with wild-type (WT) control animals.7
Increasing evidence suggests a J-shaped relationship between exercise and AF.8 It is generally agreed upon that exercise at a moderate level provides a protective effect against many diseases, including AF. In contrast, high levels of endurance or high-intensity exercise training are emerging factors for AF episodes.9 Specifically, athletes participating in chronic endurance training have a higher incidence of AF when compared with nonendurance trained individuals despite a lack of cardiovascular risk factors.10 The relationship between exercise and AF are partly mediated by structural and electrical remodeling of the atria, yet the mechanisms underlying these changes remain poorly understood. MicroRNAs (miR) are short, noncoding RNA molecules that inhibit mRNA translation or facilitate mRNA degradation and have been linked to regulation of ion channels, myocardial fibrosis, and myocyte apoptosis.2,11–15 However, the effect of exercise on circulating miR involved in atrial remodeling has not been well studied. Endurance exercise can produce a proinflammatory response that could result in structural (ie, fibrosis) and electrical remodeling of the atrium. Recently, it was demonstrated that exercise-trained rats exhibited more fibrosis when compared with a control.16 Additionally, endurance exercise has been shown to increase TGF-β1 expression in the rat atria resulting in increased fibrosis.17 However, it remains unclear if the increased incidence of AF with endurance exercise is dependent on the presence of an atrial myopathy.
In the present investigation, we sought to compare the effect of endurance exercise in the presence and absence of an underlying atrial myopathy on the incidence of AF utilizing our previously developed TGF-β1 transgenic goat (transgenic) model compared with WT controls. Additionally, we investigated the role of circulating miR involved in atrial remodeling and circulating inflammatory markers as mediators of AF risk. We hypothesized that endurance exercise in the presence of an atrial myopathy will increase AF incidence when compared with exercise in the absence of an atrial myopathy. Additionally, we hypothesized that exercise will alter the expression of circulating miR and inflammatory markers involved in atrial remodeling.
Methods
The authors declare that all supporting data are available within the article (and its Data Supplement).
Production of Transgenic Goats
Domestic goats (Capra aegagrus hircus) were used for this study. All procedures were approved by and conducted following the guidelines put forth by the Utah State University Institutional Animal Care and Use Committee and conformed to the National Institutes of Health guidelines.18
The MHC (myosin heavy chain)-TGF-β1cys33ser fragment containing the goat αMHC promoter was subcloned into the pcDNA3.1D vector. The NEON transfection system was used to electroporate primary goat fetal fibroblasts from a WT Nubian/Boer background. Cells containing the transgene were used as donor cells for somatic cell nuclear transfer and were screened by polymerase chain reaction. Donor ovaries were collected from a local abattoir, oocytes were collected using a slicing method and allowed to mature for 22 hours at 38.5 C and 5% CO2.19 Somatic cell nuclear transfer, recipient synchronization, and embryo transfers were performed as described elsewhere.20 Goats between 12 and 27 months were used for baseline characteristic evaluation. The transgenic goats (transgenic) exhibit increased cardiac fibrosis as previously shown.7
Pacemaker Implantation
Dual-chamber pacemakers (courtesy of Boston Scientific, Inc) were implanted in 6 WT and 6 transgenic goats between the age of 12 and 17 months. Goats were anesthetized with a cocktail of ketamine (6 mg/kg) and diazepam (0.1–0.2 mg/kg) via intravenous administration. Immediately following anesthesia induction, the animals were intubated, and inhalant anesthesia was maintained with isoflurane dosed to effect (1.5%–4%) in inspired oxygen. Animals were given a prophylactic antibiotic Cefazolin (20 mg/kg) IV intraoperatively as well as an analgesic post-surgery (flunixin meglamine at 1.1 mg/kg). Under sterile conditions, the chest of the animal was shaved, and the surgical site was scrubbed. Central venous access was obtained, and a two 7F sheath inserted. Pacing leads were inserted through the sheaths and positioned in the right ventricle and right atrium, respectively. The pacing leads were connected to the pacemaker generator and placed subcutaneously in an infra-scapular position. The pacemakers were then programmed to monitor for spontaneous atrial and ventricular arrhythmias during the study periods. Additionally, the pacemakers were utilized to perform AF induction studies. Goats were allowed to recover after surgery for 2 weeks, then evaluated for cardiac arrhythmia. Five separate, 10 s bursts of atrial pacing (160–200 mS) were done to evaluate inducibility; any arrhythmias were recorded (Figure 1).
Figure 1. TGF (transforming growth factor)-β1 (transgenic [TG]) goat baseline characteristics.A, Example of a TG goat immediately after pacemaker implant surgery; goats were allowed a minimum of a 2-wk recovery period before being utilized for exercise. A typical representation of an intracardiac electrogram is shown in B. The top electrogram shows a pacing burst followed by no atrial fibrillation (AF); the bottom shows an AF episode after burst pacing. A indicates atrial and V, ventricular.
Echocardiography
Images were acquired from nonsedated goats using a Vivid-q ultrasound system (GE HealthCare). The examination was performed with the goat in right lateral recumbency with the right front leg extended as previously described.6 All variables were measured 3× on 3 different cycles to calculate an average for each variable. Systolic function was evaluated using percent ejection fraction using the Teicholz method.
Exercise Protocol
Goats were evaluated biweekly with the same induction protocol detailed above to evaluate the physiological response to the exercise stimulus. A large animal hot walker (Priefert, 8 horse panel walker) was utilized for this study (see Movie in the Data Supplement). Exercise started at 2.5 miles per hour (mph) for 20 minutes, this was increased to 4.0 mph for 60 minutes by the end of the first month, 4.5 mph for 70 minutes by the end of month 2, finally by the end of month 3, goats were performing at 4.5 mph for 90 minutes (Figure 2). Any sign of lameness or ill health was monitored, and animals were reevaluated before continuing.
Figure 2. Exercise protocol. After recovering from pacemaker implants, goats started exercising in a large animal hotwalker for 3 mo. The exercise began with goats walking at 2.5 miles per hour (mph) for 20 min; this was increased to 4.0 mph for 60 min by the end of the first month. We increased both speed and time by the end of month 2 to 4.5 mph for 70 min; by the end of month 3 all goats were performing at 4.5 mph for 90 min. A short movie clip of goat exercising is provided in the Data Supplement. Afib indicates atrial fibrillation; Fri, Friday; and Mon, Monday.
Serum and Tissue miR Isolation
Peripheral blood samples were collected from all animals on the study for miR analysis. Blood was collected in red-topped, nonadditive collection tubes, and allowed to coagulate (Becton, Dickinson, and Company). Once coagulated, the tubes were spun at 3000 rpm for 30 minutes. Serum was removed, placed in microcentrifuge tubes, and stored in a -80°C freezer. Right ventricular biopsies were collected at the time of pacemaker insertion. A sterilized bioptome was inserted into the right ventricle and positioned at the mid septum of the right ventricle under fluoroscopic and intracardiac ultrasound guidance. Biopsy samples were flash-frozen with liquid nitrogen and stored. Additional cardiac tissue samples were added to this analysis from transgenic and WT goats not part of the exercise study. miR analyzed included: miR-21, miR-26a, miR-30a, miR-133a, miR-328, miR-499. Serum was used for miR profiling with Quantitative miR assays (Applied Biosystems by Thermo Fisher Scientific). Total RNA, including miR, was purified from known volumes of serum or cardiac tissue of known weights by spin column chromatography using Norgen proprietary resin as the separation matrix (Norgen Biotek, Inc) with on the column DNAse (deoxyribonuclease) treatment. RNA elution was performed in 50 μL of elution solution. Purified RNA samples were stored at −80°C.
Quantification of specific miR was performed with TaqMan MicroRNA Assays that are designed to detect and accurately quantify mature miRs using Applied Biosystems Real-Time polymerase chain reaction instruments. In brief, 5 μL of total RNA is reversed transcribed in 15 μL RT (reverse transcriptase) reaction, and then 1.3 μL of cDNA was amplified using sequence-specific primers, according to the manufacturer’s recommendations.
Known amounts of synthetic HPLS purified miRNAs (Sigma-Aldrich) were assayed side by side with experimental samples to generate standard curves for quantitative assessment of the molecule of interest. Construction of standard curves was performed for each miR by plotting mean cycle threshold values (x axis) versus corresponding log concentrations of miRs (y axis) and then used for approximation of unknown miRs from corresponding cycle threshold values as was previously described.21,22
miR concentration was calculated per microliter of serum or per 1 g of weight, based on known initial volume or weight of the sample, elution volume of RNA, and volume of RNA eluate that was inputted in assay.
Data are presented as a molar concentration per microliter of serum or per gram of tissue. miR-16 was used as a positive control for assay performance. miR-16 is stably and abundantly expressed in plasma and been validated for use as a control for RNA yield and quality as well as an assay performance.23,24 All samples were tested in triplicates for quality assurance purposes.
Protein Isolation
Biomarkers were analyzed in serum and tissue samples with Enzyme immunoassays by MyBioSource, Inc. Tissue lysates were prepared in accordance with manufacturer’s instructions; total protein content was measured with Pierce BCA assay (Thermo Fisher Scientific) and normalized by protein content.
Serum Inflammatory Markers
Goat MMP9 (matrix metallopeptidase 9), TNF (tissue necrosis factor)-α, IL (interleukin) 1β, and 6 (IL-1β and IL-6, respectively) were performed with quantitative sandwich ELISA (MyBioSource, Inc). All the assays were run according to manufacturer instructions.
Immunohistochemistry
Cardiac tissue samples obtained from animals not on the exercise regime were obtained and analyzed for the presence of positive cells per field view and quantified as percentages (Figure 3). The presence of inflammatory cells, such as macrophages and dendritic cells (CD68+ and CD14+), in cardiac tissue samples was assessed by immunohistochemical staining using an in-house protocol.25 Briefly, tissue from each chamber of the heart was preserved in optimal cutting temperature (OCT compound, Thermo Fisher Scientific) blocks and sectioned on a cryostat. Slides were dipped in acetone for 5 minutes and allowed to air dry, then placed in a −80°C freezer for 24 hours. Slides were thawed, allowed to air dry for at least 15 minutes, rinsed in PBS solution twice for 5 minutes each time, and then emerged in 0.3% endogenous peroxidase solution for 10 minutes. All incubations were done at room temperature in a humidity chamber placed on a rotator (30 rpm) and between incubations slides were washed in PBS 3× for 5 minutes, except between the incubations with blocking solution and the primary antibody. Slides were blocked with 2% goat serum/1% BSA-PBS solution for 20 minutes at room temperature. Primary antibodies against CD68 (clone EMB11; Dako Antibody Solutions, Agilent) and MHC-II (clone TH14B; Washington State University, Monoclonal Antibody Facility) were diluted to a working concentration of 1:200 in 1% BSA-PBS and slides were incubated with 300 µL of the primary antibody solution for an hour. Sections were then treated with a biotinylated goat anti-mouse IgG (Vector Laboratories, 1:200 dilution) for 20 minutes, followed by a streptavidin peroxidase (Thermo Fisher Scientific) incubation of 20 minutes. Aminoethyl carbazole solution (Thermo Fisher Scientific) was then applied for 5 minutes per manufacturer’s protocol. A counterstain was done with Gill No. 2 hematoxylin for 1 minute and then rinsed in running water until clear. A coverslip was mounted using fluoromount G (Southern Biotech), images were obtained using a Zeiss Axio Observer (Zeiss, Gottingen, Germany), and images were analyzed using AxioVision software (Zeiss). Positive cells were quantified by percent of positive area occupied by cells of interest; 10 images per section were acquired, and the AutoMeasure function from the software was used to set positive area per field of view.
Figure 3. Immunohistochemical (IHC) staining. IHC staining of atria and ventricles show no statistical difference between macrophage and dendritic cell populations in age- and sex-matched transgenic (TG) and wild-type (WT) goats. Although no statistical difference was found between TG and WT animals, there does seem to be an increase in positive cells in both MHC (myosin heavy chain)-II and CD68 staining. TGF indicates transforming growth factor.
Statistical Analysis
Statistical analysis of data was performed using StatPlus (AnalystSoft, Inc). Normality was tested using the Shapiro-Wilks test and by visual inspection of the distribution of all data sets. Homogeneity of variance was assessed using the Levene test. When data were found to have failed the homogeneity of variance assumption, the Wilcoxon rank-sum test was performed in place of the Student t test or the Kruskal-Wallis test in place of the ANOVA. Statistical differences were assessed with 2-way ANOVA, Student t test, Wilcoxon Rank-sum test, Kruskal-Wallis test, and χ2 test as appropriate. All data are expressed as mean±SEM.
Results
Baseline Characteristics
As previously reported, the transgenic goats exhibited increased atrial fibrosis compared with WT goats.7 Resting heart rate was not different at baseline between transgenic and WT goats (147±12 versus 150±12 bpm; P=0.84). Baseline left and right atrial size and interventricular septal thickness were increased in transgenic compared with WT goats (Table 1). All other baseline echocardiographic parameters were not different between transgenic and WT goats. Baseline atrial and ventricular protein expression of Cx (connexin) 40 and 43, RyR2 (ryanodine receptor), NCX (sodium-calcium exchanger), LTCC (L-type calcium channel), and KvLQT1 (voltage-gated potassium channel––LQT1) (voltage-gated potassium channel––LQT1) channel were not different between transgenic and WT goats (Table 2). There were no baseline differences in all measured atrial or ventricular tissue miR levels (Table 3).
| Parameter | Control | TGF-β1 TG | P Value |
|---|---|---|---|
| Aorta diam, cm | 2.53±0.03 | 2.76±0.09 | 0.07 |
| LA diam, cm | 2.85±0.03 | 3.17±0.07 | 0.05 |
| RA max, cm | 2.40±0.20 | 3.24±0.10 | 0.02 |
| IVSd, cm | 0.83±0.05 | 0.96±0.03 | 0.05 |
| LVIDd, cm | 3.80±0.21 | 4.17±0.23 | 0.54 |
| LVPWd, cm | 1.00±0.07 | 1.00±0.03 | 0.84 |
| LVEF, % | 78±4 | 78±2 | 0.71 |
| RVIDd, cm | 1.13±0.16 | 1.13±0.10 | 0.83 |
Atria |
Ventricle | |||||
|---|---|---|---|---|---|---|
| Control | TGF-β1 TG | P Value | Control | TG | P Value | |
| LTCC, ng/mL | 1.30±0.45 | 1.28±0.10 | 0.89 | 1.19±0.04 | 1.28±0.06 | 0.71 |
| Connexin 40, ng/mL | 5.07±0.32 | 6.51±0.56 | 0.06 | 5.98E±0.17 | 5.25±0.38 | 0.12 |
| Connexin 43, ng/mL | 3.58±0.12 | 3.73±0.20 | 0.73 | 3.50±0.09 | 3.65±0.16 | 0.43 |
| KCNQ1, ng/mL | 0.75±0.07 | 0.87±0.08 | 0.34 | 0.89±0.04 | 0.94±0.07 | 0.71 |
| NCX, ng/mL | 0.8±0.07 | 0.80±0.06 | 0.53 | 0.74±0.05 | 0.72±0.03 | 0.81 |
| RyR2, pg/mL | 157.0±8.5 | 174.0±4.6 | 0.09 | 166.0±9.0 | 153.0±5.1 | 0.77 |
| Atria | Ventricle | |||||
|---|---|---|---|---|---|---|
| Control | TGF-β1 TG | P Value | Control | TGF-β1 TG | P Value | |
| miR-1, fmol/g | 1.48×109±2.70×108 | 1.22×109±2.61×108 | 0.57 | 1.78×109±2.32×108 | 1.67×109±7.63×107 | 0.74 |
| miR-21, fmol/g | 1.90×107±2.60×106 | 1.92×107±3.06×106 | 0.95 | 1.48×107±1.40×106 | 1.52×107±2.19×106 | 0.61 |
| miR-26a, fmol/g | 1.77×109±2.31×108 | 1.17×108±5.83×106 | 0.52 | 1.67×107±9.75×105 | 6.70×106±5.13×105 | 0.47 |
| miR-29a, fmol/g | 8.24×106±1.50×106 | 8.51×106±1.78×106 | 0.92 | 1.05×107±7.43×105 | 9.79×106±1.16×106 | 0.36 |
| miR-30a, fmol/g | 6.57×106±1.09×106 | 6.57×106±2.03×106 | 0.99 | 6.58×106±9.92×105 | 7.56×106±6.04×105 | 0.43 |
| miR-133a, fmol/g | 1.10×1015±1.62×1014 | 1.44×1015±2.95×1014 | 0.29 | 1.76×1015±3.79 x1014 | 2.41×1015±4.17×1014 | 0.16 |
| miR-328, fmol/g | 3.91×1015±4.77×1014 | 3.16×1015±2.91×1014 | 0.34 | 3.81×1015± 8.41×1014 | 4.44×1015±1.57×1015 | 0.79 |
| miR-499, fmol/g | 2.14×107±4.56×106 | 1.24×107±5.05×106 | 0.27 | 2.02×108±3.42×107 | 1.38×108±2.18×107 | 0.32 |
There was no significant difference in macrophages and dendritic cells (CD68+ and MHC-II) in cardiac tissue of transgenic animals when compared with WT animals. More specifically, CD68 and MHC-II expression were numerically greater in transgenic compared with WT animals, but this was not statistically different (Figure 3). Additionally, serum MMP9, TNF-α, IL-1β, and IL-6 levels were not different (TNF-α and IL-6) or trended lower (MMP9 and IL-1β) in transgenic compared with WT animals. These findings suggest inflammation is not a primary mechanism for baseline atrial structural remodeling in the transgenic animals.
Exercise and AF
No spontaneous AF was detected at baseline before the start of exercise training in both transgenic and WT goats. However, during the 3-month exercise period, there was a progressive increase in the incidence of pacemaker-detected, spontaneous AF in transgenic animals that peaked with a 50% incidence during the final month of exercise training (Figure 4). Interestingly, all spontaneous AF episodes occurred during nonexercising periods. In contrast, no spontaneous AF was detected in WT animals.
Figure 4. Spontaneous atrial fibrillation (AF). An intracardiac electrogram tracing from goat G1-4 (transgenic [TG]) shows a spontaneous AF episode that eventually self-terminated. As shown, there were no instances of spontaneous AF at baseline in any of the goats. Over the course of the exercise regime, there was a progressive increase in spontaneous episodes in the TG animals, culminating towards the end of the exercise period, with 50% of the TG animals exhibiting spontaneous episodes (P=0.04). No spontaneous AF was recorded post-exercise or in any wild-type goats. A indicates atrial; TGF, transforming growth factor and V, ventricular.
At baseline, 67% (4 of 6) transgenic goats had sustained (≥30 s) inducible AF episodes, compared with 0% (0 of 6) WT goats (Figure 5A). During the first 2 months of exercise training, there was an increase in the incidence of inducible sustained AF in both transgenic and WT goat groups. This increase in inducible, sustained AF declined at 3 months of exercise and returned to near baseline levels 1-month postexercise training. Additionally, total AF duration (inducible) was greater in transgenic animals at baseline and throughout exercise compared with WT animals (Figure 5B). There was also a nonstatistically different increase in total AF duration (via burst pacing) during the first 2 months of exercise that declined during the third month and returned to near baseline levels at 1-month post-exercise.
Figure 5. Inducible atrial fibrillation (AF).A, During the exercise period, transgenic (TG) and wild-type (WT) goats showed an increase of inducible AF episodes, which returned close to baseline post-exercise. At baseline, 67% (4 of 6) transgenic goats had sustained (≥30 s) inducible AF episodes, compared with 0% (0 of 6) WT goats. During the first 2 mo of exercise training, there was an increase in inducible sustained AF in both TG (100% at month 1 and 83% at month 2) and WT groups (33% at month 1 and 67% at month 2). The TG animals inducible, sustained episodes peaked after 1 mo of exercise training, tapering during months 2 and 3, and returning close to baseline levels postexercise training. The increase in inducible sustained AF episodes declined in WT animals at the end of the exercise period and returned to baseline levels 1-mo postexercise training. B, Similarly, total inducible AF duration (in seconds) was greater in TG animals at baseline and throughout exercise compared with WT animals. There was also a nonstatistically different increase in total AF duration (via burst pacing) during the first 2 mo of exercise that declined during the third month and returned to near baseline levels at 1-mo post-exercise (P=0.004 for group, P=0.43 for time; B). TGF indicates transforming growth factor.
Exercise and Serum miR Expression
Similar to tissue miR expression, there were no baseline differences in serum miR expression between transgenic and WT goats. However, there was a time-dependent decrease in miR-21 and miR-29 during the first 2 months of exercise, returning to baseline at 1-month post-exercise in both transgenic and WT animals (Figure 6A and 6B). As shown in Figure 6C, miR-26a did not change from baseline during exercise, but there was a significant increase in miR-26a expression at 1-month post-exercise in both transgenic and WT goats. Interestingly, miR-133a, which is implicated in decreasing TGF-β1 level, was not different between transgenic and WT animals (Figure 6D). There was variable expression of miR-133a during exercise but no difference in miR-133a at 1-month post-exercise compared with baseline. No significant differences were detected between transgenic and WT animals in response to exercise training in miR-30a and miR-328 (Figure 6E and 6F).
Figure 6. Serum microRNA (miR) expression. As exhibited in tissue miR expression, miR analysis from peripheral serum showed no baseline difference in expression between transgenic and wild-type goats. There was a time-dependent decrease in miR-21 (P=0.01 for time, P=0.78 for group; A), and miR-29 (P=0.02 for time, P=0.34 for group; B), during the first 2 mo of exercise, returning to baseline at 1-mo post-exercise in both groups. miR-26a did not change from baseline during exercise, but there was a significant increase in expression 1-mo post-exercise in both groups (P=0.001 for time, P=0.81 for group); (C) miR-133a was also found to be significantly different over time (P=0.01 for time, P=0.64 for group); (D) no significant differences were shown between animal groups or time points in miR-30a and miR-328 (E and F). TGF indicates transforming growth factor.
Exercise and Serum Inflammatory Markers
Inflammatory markers were analyzed from serum during the baseline, exercise training, and postexercise training periods. Among the 4 inflammatory markers selected, we found that MMP9 was the only marker significantly changed during exercise (Figure 7A). Specifically, there was a gradual increase in serum MMP9 during exercise training with a further increase at 1-month post-exercise. Exercise did not alter serum IL1-β, IL-6 or TNF-α (Figure 7B, 7C, and 7D). Finally, serum MMP9, IL1-β, and IL-6 concentrations were lower in transgenic compared with WT animals, suggestive of less overall inflammation in transgenic animals.
Figure 7. Serum inflammatory markers. Four inflammatory markers were analyzed from peripheral serum between transgenic (TG) and wild-type (WT) groups, as well as over the exercise time period. Significant differences were observed for MMP9 (matrix metallopeptidase 9; P=0.005 for time, P=0.001 for group, A) and IL (interleukin)-1β (P=0.0001 for group, B). The effect of exercise did not alter levels of IL1-β, IL-6, or TNF (tumor necrosis factor)-α (B, C, and D). It is important to note that the significant group difference in MMP9, IL1-β, and IL-6 show lower levels in TG animals compared with WT animals. TGF indicates transforming growth factor.
Discussion
This study utilized a novel transgenic goat model of cardiac fibrosis (TGF-β1 overexpression) to demonstrate that endurance exercise in the setting of an underlying atrial myopathy increases the incidence of spontaneous AF. This finding is consistent with our hypothesis that endurance exercise in the presence of an atrial myopathy will increase AF incidence when compared with exercise in the absence of an atrial myopathy. Additionally, endurance exercise seems to increase inducible AF secondary to altered expression of key profibrotic biomarkers (miR-21, miR-29, and MMP9) that is independent of the presence of an underlying atrial myopathy.
The presence of spontaneous AF in this study highlights the strength of this novel large animal transgenic model for cardiac fibrosis in investigating the underlying mechanism of AF. In particular, this model targets a specific genetic mechanism of myocardial fibrosis with cardiac-specific overexpression of TGF-β1, resulting in increased atrial fibrosis with relative sparing of the ventricle. This is in contrast to other large animal models of AF that require external provocateurs (ie, rapid pacing and pericarditis) that have concomitant remodeling of the ventricle.26,27 Furthermore, the presence of spontaneous AF in this model is novel and has important clinical relevance when attempting to parallel and understand human disease provoked by internal biochemical pathways and exposure to cardiovascular diseases. Our data support the premise that atrial fibrosis is a key mechanism of AF and is consistent with prior research supporting a mechanistic role of fibrosis in AF.28,29 Furthermore, baseline measures of inflammation and key miR involved in atrial remodeling were not different in transgenic and WT animals, suggesting that the key mechanism of atrial fibrosis in this model is overexpression of TGF-β1. It is possible that spontaneous AF increased during the exercise period because of vagal-mediated slowing of resting heart rate. Analysis of the HR preceding the onset of spontaneous AF in transgenic animals showed a reduction in (heart rate) HR before spontaneous AF when compared with baseline (121 versus 146 bpm). Of note, similar modulation of vagal activity with concomitant HR slowing would be expected in the WT animals exposed to the same exercise regimen. However, spontaneous AF was not observed in any of the WT animals, further supporting the conclusion that atrial fibrosis is a key mechanism in the development of spontaneous AF in this study.
In humans, endurance exercise and AF risk exhibit a J-shaped relationship.8 Specifically, moderate amounts of endurance exercise seem to not increase or perhaps even decrease AF susceptibility. In contrast, high amounts of endurance exercise are associated with increased incidence of AF. Our study provides important insights into this relationship. Specifically, endurance exercise at the intensity and duration utilized in this study only increased the incidence of spontaneous AF in the presence of an underlying atrial myopathy. This highlights the importance of atrial structural remodeling, creating an arrhythmogenic substrate as a key mechanism for the development of AF during endurance training. These data suggest that high levels of endurance exercise may increase risk of AF in individuals with an underlying atrial myopathy. Research is needed to determine whether the increase in AF with endurance exercise in the presence of an atrial myopathy is dependent on the exercise dose. More specifically, in individuals with an underlying atrial myopathy, does all exercise increase AF incident or is the previously described J-shaped relationship between exercise and AF shifted to the left?
Interestingly, regardless of the presence of an underlying atrial myopathy, endurance exercise seems to initially increase susceptibility to inducible AF during the first 2 months of exercise but returns to baseline levels with continued exercise up to 3 months of exercise training. Also, the time course of the expression of key profibrotic biomarkers (miR-21, miR-29) seems to provide some mechanistic support for the relationship between endurance exercise and susceptibility to inducible AF found in this study. For example, there was an initial decrease in miR-21 and miR-29 (promoting fibrosis) during the first 2 months of exercise that returned to baseline during the 3 months. This suggests that initially, exercise may trigger atrial remodeling that is proarrhythmic, but with continued exercise, there are adaptive changes that prevent additional remodeling or perhaps adaptive changes developed to minimize the biologic response to exercise-related stress. For example, exercise training has been associated with modulation of the autonomic nervous system, in particular, increasing vagal tone. Importantly, vagal tone has been shown to have an inverse relationship with inflammation.30 As such, increased vagal tone from exercise may prevent adverse atrial remodeling by blunting exercise-related inflammation. Our study exposed animals to 3 months of endurance exercise training and the effect of more prolonged (ie, >3 months) would have on atrial remodeling, and AF susceptibility is not known at this time. Further research is needed to clarify the exact time course of the exercise and AF relationship and the underlying mechanisms. Our data agree with a previous study in rats,16 where exercise increased AF episodes but returned to near baseline levels towards the end of the training.
As noted above, we identified a profibrotic decrease in miR-21 during the first 2 months of exercise that returned to near baseline level at 3 months of exercise. It is widely shown that reduced miR-21 is associated with a fibrotic state in many animal models (rat pericarditis) as well as humans.31 Additionally, miR-29 also had a time-dependent effect, decreasing during the first 2 months of exercise with a recovery back toward baseline at 3 months of exercise. Reduced miR-29 is thought to be profibrotic secondary to its regulation of cardiac fibroblast expression.32 Interestingly, miR-133a, which is implicated in decreasing TGF-β1 levels33 was not different between transgenic and WT animals. No significant differences were detected between transgenic and WT animals in miR-30a (downregulates fibrosis via Snail 1) and miR-328 (implicated in calcium channel handling).34
Chronic endurance exercise is associated with increased production and release of the proinflammatory cytokines IL1-β, IL-6, and TNF-α.35–37 We hypothesized that there would be a trend toward global inflammation in transgenic compared with WT animals. However, we did not observe this and postulate that this reduction in global inflammation in transgenic animals may be secondary to the anti-inflammatory effects TGF-β1 overexpression. In contrast, we found no significant effect of endurance exercise on serum concentrations of IL1-β, IL-6, or TNF-α. However, serum MMP9 was increased in response to endurance exercise in both transgenic and WT animals. These data suggest that inflammation was not likely a key mechanism in the development of spontaneous AF in the transgenic animals. Furthermore, with the exception of MMP9, it is unlikely that these inflammatory cytokines were mechanistically involved in altering the incidence of inducible AF during exercise in this investigation. The MMP family is implicated in the exchange of ECM to connective tissue.38 MMP9 is found to be aberrantly expressed in human patients39 and is considered to contribute to structural remodeling. We also analyzed circulating serum from these animals; it should be noted that more insight into this anti-inflammatory effect may be gained by evaluating atrial tissue directly after the exercise treatment.
The structural remodeling exhibited by this transgenic goat model shows that progressive fibrosis can facilitate the development of spontaneous arrhythmic episodes. Although atrial fibrosis has been associated with AF incidence and AF-related comorbidities in humans, the interplay with endurance athletics and AF risk remains unknown.40 Our data suggest that endurance athletics can serve as a second hit or trigger of arrhythmia in the setting of inherited or acquired predisposition.
In conclusion, this study highlights the relationship between endurance exercise and an atrial myopathy in the development of AF (Figure 8). Specifically, it suggests that individuals with an underlying atrial myopathy may have greater risk of developing AF than those with structurally normal atria when exposed to a physiological stressor. Further research is needed to determine the relevance of this observation in humans.
Figure 8. Transgenic (TG) and wild-type (WT) goat differences. TG goats at baseline have an increase in ECM and diffuse fibrosis in their atria when compared with WT control goats. There was also a progressive increase in spontaneous and inducible atrial fibrillation (AF) episodes when compared with their respective controls. MicroRNA (miR)-21 seems to track inversely to the incidence of AF in both groups. MMP9 indicates matrix metallopeptidase 9; and TGF, transforming growth factor.
Sources of Funding
This work was supported by Intermountain Medical and Research Foundation Grant (IMRF00523 to Dr Cutler), Utah Agricultural Experiment Station Grant (UTAO1328 to Dr Polejaeva), and Boston Scientific, Inc (Pacemakers).
Disclosures
J. Doleac is employed by Boston Scientific, Inc. Dr Cutler is a consultant and on the advisory board for Biosense Webster. The other authors report no conflicts.
Footnotes
References
- 1.
Andrade J, Khairy P, Dobrev D, Nattel S . The clinical profile and pathophysiology of atrial fibrillation: relationships among clinical features, epidemiology, and mechanisms.Circ Res. 2014; 114:1453–1468. doi: 10.1161/CIRCRESAHA.114.303211LinkGoogle Scholar - 2.
Jalife J, Kaur K . Atrial remodeling, fibrosis, and atrial fibrillation.Trends Cardiovasc Med. 2015; 25:475–484. doi: 10.1016/j.tcm.2014.12.015CrossrefMedlineGoogle Scholar - 3.
Kottkamp H . Human atrial fibrillation substrate: towards a specific fibrotic atrial cardiomyopathy.Eur Heart J. 2013; 34:2731–2738. doi: 10.1093/eurheartj/eht194CrossrefMedlineGoogle Scholar - 4.
Nassal MM, Wan X, Laurita KR, Cutler MJ . Atrial SERCA2a overexpression has no affect on cardiac alternans but promotes arrhythmogenic SR Ca2+ triggers.PLoS One. 2015; 10:e0137359. doi: 10.1371/journal.pone.0137359CrossrefMedlineGoogle Scholar - 5.
Verheule S, Sato T, Everett T, Engle SK, Otten D, Rubart-von der Lohe M, Nakajima HO, Nakajima H, Field LJ, Olgin JE . Increased vulnerability to atrial fibrillation in transgenic mice with selective atrial fibrosis caused by overexpression of TGF-beta1.Circ Res. 2004; 94:1458–1465. doi: 10.1161/01.RES.0000129579.59664.9dLinkGoogle Scholar - 6.
Everett THt, Olgin JE . Atrial fibrosis and the mechanisms of atrial fibrillation.Heart Rhythm. 2007; 4:S24–S27.CrossrefMedlineGoogle Scholar - 7.
Polejaeva IA, Ranjan R, Davies CJ, Regouski M, Hall J, Olsen AL, Meng Q, Rutigliano HM, Dosdall DJ, Angel NA, Sachse FB, Seidel T, Thomas AJ, Stott R, Panter KE, Lee PM, Van Wettere AJ, Stevens JR, Wang Z, MacLeod RS, Marrouche NF, White KL . Increased susceptibility to atrial fibrillation secondary to atrial fibrosis in transgenic goats expressing transforming growth factor-β1.J Cardiovasc Electrophysiol. 2016; 27:1220–1229. doi: 10.1111/jce.13049CrossrefMedlineGoogle Scholar - 8.
Elliott AD, Linz D, Verdicchio CV, Sanders P . Exercise and atrial fibrillation: prevention or causation?Heart Lung Circ. 2018; 27:1078–1085. doi: 10.1016/j.hlc.2018.04.296CrossrefMedlineGoogle Scholar - 9.
Müssigbrodt A, Weber A, Mandrola J, van Belle Y, Richter S, Döring M, Arya A, Sommer P, Bollmann A, Hindricks G . Excess of exercise increases the risk of atrial fibrillation.Scand J Med Sci Sports. 2017; 27:910–917. doi: 10.1111/sms.12830CrossrefMedlineGoogle Scholar - 10.
Kim JH, Baggish AL . Strenuous exercise and cardiovascular disease outcomes.Curr Atheroscler Rep. 2017; 19:1. doi: 10.1007/s11883-017-0636-3CrossrefMedlineGoogle Scholar - 11.
Huang ZP, Neppl RL, Wang DZ . MicroRNAs in cardiac remodeling and disease.J Cardiovasc Transl Res. 2010; 3:212–218. doi: 10.1007/s12265-010-9165-yCrossrefMedlineGoogle Scholar - 12.
Wang Z, Lu Y, Yang B . MicroRNAs and atrial fibrillation: new fundamentals.Cardiovasc Res. 2011; 89:710–721. doi: 10.1093/cvr/cvq350CrossrefMedlineGoogle Scholar - 13.
Cai B, Pan Z, Lu Y . The roles of microRNAs in heart diseases: a novel important regulator.Curr Med Chem. 2010; 17:407–411. doi: 10.2174/092986710790226129CrossrefMedlineGoogle Scholar - 14.
Harada M, Luo X, Murohara T, Yang B, Dobrev D, Nattel S . MicroRNA regulation and cardiac calcium signaling: role in cardiac disease and therapeutic potential.Circ Res. 2014; 114:689–705. doi: 10.1161/CIRCRESAHA.114.301798LinkGoogle Scholar - 15.
Bartel DP . MicroRNAs: genomics, biogenesis, mechanism, and function.Cell. 2004; 116:281–297. doi: 10.1016/s0092-8674(04)00045-5CrossrefMedlineGoogle Scholar - 16.
Guasch E, Benito B, Qi X, Cifelli C, Naud P, Shi Y, Mighiu A, Tardif JC, Tadevosyan A, Chen Y, Gillis MA, Iwasaki YK, Dobrev D, Mont L, Heximer S, Nattel S . Atrial fibrillation promotion by endurance exercise: demonstration and mechanistic exploration in an animal model.J Am Coll Cardiol. 2013; 62:68–77. doi: 10.1016/j.jacc.2013.01.091CrossrefMedlineGoogle Scholar - 17.
Benito B, Gay-Jordi G, Serrano-Mollar A, Guasch E, Shi Y, Tardif JC, Brugada J, Nattel S, Mont L . Cardiac arrhythmogenic remodeling in a rat model of long-term intensive exercise training.Circulation. 2011; 123:13–22. doi: 10.1161/CIRCULATIONAHA.110.938282LinkGoogle Scholar - 18. Guide for the Care and Use of Laboratory Animals. 8 ed. Washington, DC: National Academies Press (US); 2011.Google Scholar
- 19.
Yang M, Hall J, Fan Z, Regouski M, Meng Q, Rutigliano HM, Stott R, Rood KA, Panter KE, Polejaeva IA . Oocytes from small and large follicles exhibit similar development competence following goat cloning despite their differences in meiotic and cytoplasmic maturation.Theriogenology. 2016; 86:2302–2311. doi: 10.1016/j.theriogenology.2016.07.026CrossrefMedlineGoogle Scholar - 20.
Yang M, Perisse I, Fan Z, Regouski M, Meyer-Ficca M, Polejaeva IA . Increased pregnancy losses following serial somatic cell nuclear transfer in goats.Reprod Fertil Dev. 2018; 30:1443–1453. doi: 10.1071/RD17323CrossrefMedlineGoogle Scholar - 21.
Adachi T, Nakanishi M, Otsuka Y, Nishimura K, Hirokawa G, Goto Y, Nonogi H, Iwai N . Plasma microRNA 499 as a biomarker of acute myocardial infarction.Clin Chem. 2010; 56:1183–1185. doi: 10.1373/clinchem.2010.144121CrossrefMedlineGoogle Scholar - 22.
Mitchell PS, Parkin RK, Kroh EM, Fritz BR, Wyman SK, Pogosova-Agadjanyan EL, Peterson A, Noteboom J, O’Briant KC, Allen A, Lin DW, Urban N, Drescher CW, Knudsen BS, Stirewalt DL, Gentleman R, Vessella RL, Nelson PS, Martin DB, Tewari M . Circulating microRNAs as stable blood-based markers for cancer detection.Proc Natl Acad Sci U S A. 2008; 105:10513–10518. doi: 10.1073/pnas.0804549105CrossrefMedlineGoogle Scholar - 23.
Kroh EM, Parkin RK, Mitchell PS, Tewari M . Analysis of circulating microRNA biomarkers in plasma and serum using quantitative reverse transcription-PCR (qRT-PCR).Methods. 2010; 50:298–301. doi: 10.1016/j.ymeth.2010.01.032CrossrefMedlineGoogle Scholar - 24.
Xiao J, Jing ZC, Ellinor PT, Liang D, Zhang H, Liu Y, Chen X, Pan L, Lyon R, Liu Y, Peng LY, Liang X, Sun Y, Popescu LM, Condorelli G, Chen YH . MicroRNA-134 as a potential plasma biomarker for the diagnosis of acute pulmonary embolism.J Transl Med. 2011; 9:159. doi: 10.1186/1479-5876-9-159CrossrefMedlineGoogle Scholar - 25.
Rutigliano HM, Thomas AJ, Wilhelm A, Sessions BR, Hicks BA, Schlafer DH, White KL, Davies CJ . Trophoblast major histocompatibility complex class i expression is associated with immune-mediated rejection of bovine fetuses produced by cloning.Biol Reprod. 2016; 95:39. doi: 10.1095/biolreprod.115.136523CrossrefMedlineGoogle Scholar - 26.
Yoshizawa T, Niwano S, Niwano H, Tamaki H, Nakamura H, Igarashi T, Oikawa J, Satoh A, Kishihara J, Murakami M, Fukaya H, Ako J . Antiremodeling effect of xanthine oxidase inhibition in a canine model of atrial fibrillation.Int Heart J. 2018; 59:1077–1085. doi: 10.1536/ihj.17-391CrossrefMedlineGoogle Scholar - 27.
Zhang Y, Wang YT, Shan ZL, Guo HY, Guan Y, Yuan HT . Role of inflammation in the initiation and maintenance of atrial fibrillation and the protective effect of atorvastatin in a goat model of aseptic pericarditis.Mol Med Rep. 2015; 11:2615–2623. doi: 10.3892/mmr.2014.3116CrossrefMedlineGoogle Scholar - 28.
Nattel S . Molecular and cellular mechanisms of atrial fibrosis in atrial fibrillation.JACC Clin Electrophysiol. 2017; 3:425–435. doi: 10.1016/j.jacep.2017.03.002CrossrefMedlineGoogle Scholar - 29.
Rahmutula D, Marcus GM, Wilson EE, Ding CH, Xiao Y, Paquet AC, Barbeau R, Barczak AJ, Erle DJ, Olgin JE . Molecular basis of selective atrial fibrosis due to overexpression of transforming growth factor-β1.Cardiovasc Res. 2013; 99:769–779. doi: 10.1093/cvr/cvt074CrossrefMedlineGoogle Scholar - 30.
Sloan RP, McCreath H, Tracey KJ, Sidney S, Liu K, Seeman T . RR interval variability is inversely related to inflammatory markers: the CARDIA study.Mol Med. 2007; 13:178–184. doi: 10.2119/2006–00112.SloanCrossrefMedlineGoogle Scholar - 31.
Zhou Q, Maleck C, von Ungern-Sternberg SNI, Neupane B, Heinzmann D, Marquardt J, Duckheim M, Scheckenbach C, Stimpfle F, Gawaz M, Schreieck J, Seizer P, Gramlich M . Circulating microRNA-21 correlates with left atrial low-voltage areas and is associated with procedure outcome in patients undergoing atrial fibrillation ablation.Circ Arrhythm Electrophysiol. 2018; 11:e006242. doi: 10.1161/CIRCEP.118.006242LinkGoogle Scholar - 32.
Poudel P, Xu Y, Cui Z, Sharma D, Tian B, Paudel S . Atrial fibrillation: recent advances in understanding the role of microRNAs in atrial remodeling with an electrophysiological overview.Cardiology. 2015; 131:58–67. doi: 10.1159/000375403CrossrefMedlineGoogle Scholar - 33.
Wang J, Liew OW, Richards AM, Chen YT . Overview of microRNAs in cardiac hypertrophy, fibrosis, and apoptosis.Int J Mol Sci. 2016; 17:749–60.CrossrefMedlineGoogle Scholar - 34.
van den Berg NWE, Kawasaki M, Berger WR, Neefs J, Meulendijks E, Tijsen AJ, de Groot JR . MicroRNAs in atrial fibrillation: from expression signatures to functional implications.Cardiovasc Drugs Ther. 2017; 31:345–365. doi: 10.1007/s10557-017-6736-zCrossrefMedlineGoogle Scholar - 35.
Swanson DR . Atrial fibrillation in athletes: implicit literature-based connections suggest that overtraining and subsequent inflammation may be a contributory mechanism.Med Hypotheses. 2006; 66:1085–1092. doi: 10.1016/j.mehy.2006.01.006CrossrefMedlineGoogle Scholar - 36.
Lakin R, Polidovitch N, Yang S, Guzman C, Gao X, Wauchop M, Burns J, Izaddoustdar F, Backx PH . Inhibition of soluble TNFα prevents adverse atrial remodeling and atrial arrhythmia susceptibility induced in mice by endurance exercise.J Mol Cell Cardiol. 2019; 129:165–173. doi: 10.1016/j.yjmcc.2019.01.012CrossrefMedlineGoogle Scholar - 37.
Stumpf C, Simon M, Wilhelm M, Zimmermann S, Rost C, Achenbach S, Brem MH . Left atrial remodeling, early repolarization pattern, and inflammatory cytokines in professional soccer players.J Cardiol. 2016; 68:64–70. doi: 10.1016/j.jjcc.2015.08.013CrossrefMedlineGoogle Scholar - 38.
Lin CS, Pan CH . Regulatory mechanisms of atrial fibrotic remodeling in atrial fibrillation.Cell Mol Life Sci. 2008; 65:1489–1508. doi: 10.1007/s00018-008-7408-8CrossrefMedlineGoogle Scholar - 39.
Doñate Puertas R, Millat G, Ernens I, Gache V, Chauveau S, Morel E, Christin E, Couturier N, Devaux Y, Chevalier P . Atrial structural remodeling gene variants in patients with atrial fibrillation.Biomed Res Int. 2018; 2018:4862480. doi: 10.1155/2018/4862480CrossrefMedlineGoogle Scholar - 40.
Akoum N, Marrouche N . Assessment and impact of cardiac fibrosis on atrial fibrillation.Curr Cardiol Rep. 2014; 16:518. doi: 10.1007/s11886-014-0518-zCrossrefMedlineGoogle Scholar
eLetters(0)
eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.
Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.