Optical Mapping Approaches on Muscleblind‐Like Compound Knockout Mice for Understanding Mechanistic Insights Into Ventricular Arrhythmias in Myotonic Dystrophy
Background Cardiac arrhythmias are common causes of death in patients with myotonic dystrophy (dystrophia myotonica [DM]). Evidence shows that atrial tachyarrhythmia is an independent risk factor for sudden death; however, the relationship is unclear.
Methods and Results Control wild‐type (Mbnl1+/+; Mbnl2+/+) and DM mutant (Mbnl1−/−; Mbnl2+/−) mice were generated by crossing double heterozygous knockout (Mbnl1+/−; Mbnl2+/−) mice. In vivo electrophysiological study and optical mapping technique were performed to investigate mechanisms of ventricular tachyarrhythmias. Transmission electron microscopy scanning was performed for myocardium ultrastructural analysis. DM mutant mice were more vulnerable to anesthesia medications and program electrical pacing: 2 of 12 mice had sudden apnea and cardiac arrest during premedication of general anesthesia; 9 of the remaining 10 had atrial tachycardia and/or atrioventricular block, but none of the wild‐type mice had spontaneous arrhythmias; and 9 of 10 mice had pacing‐induced ventricular tachyarrhythmias, but only 1 of 14 of the wild‐type mice. Optical mapping studies revealed prolonged action potential duration, slower conduction velocity, and steeper conduction velocity restitution curves in the DM mutant mice than in the wild‐type group. Spatially discordant alternans was more easily inducible in DM mutant than wild‐type mice. Transmission electron microscopy showed disarranged myofibrils with enlarged vacuole‐occupying mitochondria in the DM mutant group.
Conclusions This DM mutant mouse model presented with clinical myofibril ultrastructural abnormality and cardiac arrhythmias, including atrial tachyarrhythmias, atrioventricular block, and ventricular tachyarrhythmias. Optical mapping studies revealed prolonged action potential duration and slow conduction velocity in the DM mice, leading to vulnerability of spatially discordant alternans and ventricular arrhythmia induction to pacing.
Myotonic dystrophy (dystrophia myotonica [DM]) is an autosomal dominant disorder characterized by progressive skeletal muscle weakness, myotonia, and multisystemic features. DM type I (DM1) is the more severe type of DM that is caused by unstable microsatellite (CTG)n expansion (n>50) in the 3′ untranslated region of the DMPK gene.1 DM type II is a milder form that is caused by (CCTG)n expansion (n>75) in the first intron of the CNBP gene.2 Arrhythmia is the second most common cause of death for patients with DM1, only after respiratory failure.3, 4 Similar to skeletal muscle, the length of (CTG)n expansion correlates with the severity of cardiac manifestations in most series.5, 6 While heart failure is found only in end‐stage DM patients, conduction block is more common and ranges from asymptomatic first‐degree atrioventricular block to severe Stokes‐Adams attack in early‐stage DM patients. The ECG and electrophysiological findings include prolongation of the PR interval, increased QRS complex duration, and prolongation of the His‐ventricular interval.7 Of note, signs of conduction block are associated with developing lethal ventricular tachycardia/fibrillation (VT/VF). DM patients who have a family history of sudden death or who are older than 40 years appear to have higher risk5, 8; however, adolescent patients with increases in heart rate during physical exercise may also experience sudden cardiac death.9 Although severe ECG abnormalities that are indicative of conduction block and sustained atrial tachyarrhythmias (eg, atrial fibrillation) could be independent risk factors for sudden death,7 the relationship between atrial arrhythmia‐conduction abnormality and ventricular arrhythmias remains unclear. This evidence implies that rate‐dependent factors and ventricular conduction abnormalities play a role in ventricular arrhythmogensis in DM patients. Although prophylactic permanent pacemaker implantation may reduce the incidence of sudden cardiac death in DM patients,10 sudden death caused by ventricular tachyarrhythmias still occurs in these patients, suggesting that ventricular tachyarrhythmias may be more frequent than expected7, 11 Therefore, an implantable cardioverter‐defibrillator that is capable of both pacing and defibrillation is recommended for patients at high risk for ventricular tachyarrhythmias. In recent series, relatively high incidences (14.3–33.3%) of ventricular tachycardia in patients with implantable cardioverter‐defibrillator implantation were observed and shocks were appropriately provided.12, 13
Accumulating evidence shows that pathogenic RNA could generate toxic effects throughout the sequestration and, subsequently, loss of function of muscleblind‐like (MBNL) proteins. The MBNL family consists of 3 members (MBNL1, MBNL2, and MBNL3) and they are RNA‐binding proteins critical for splicing regulations and polyadenylations during the developmental transition from fetal to adult.14, 15, 16, 17 Mbnl1 knockout (KO) mice recapitulate myotonia and skeletal muscle pathology18 and Mbnl2 KO mice reproduce abnormal sleep pattern and cognitive decline.19 While Mbnl1−/−; Mbnl2−/− double KOs are embryonic lethal, Mbnl1−/−; Mbnl2+/− (Mbnl1 KO; Mbnl2 heterozygous [HET]) mice show enhanced myotonia and cardiac abnormalities including conduction block, dilated cardiomyopathy, focal fibrosis, and sudden death.20 Although these mice showed reduced life span, there was no direct evidence for verifying the mechanisms of sudden death. In this study, we conducted simultaneously monitored intracellular calcium (Cai) and membrane voltage (Vm) mapping on Mbnl1 KO; Mbnl2 HET mice to investigate possible mechanisms of ventricular tachyarrhythmias.
Materials and Methods
Mbnl Compound KO Mouse Model
Original articles describing the generation of Mbnl1 and Mbnl2 single KOs,18, 19 as well as Mbnl1 KO; Mbnl2 HET compound KO mice, have been published.20 For current experiments, the Mbnl1 KO; Mbnl2 HET (DM mutant) mice were generated by crossing 2 single KO lines in the animal center of Chang Gung Memorial Hospital, Keelung, Taiwan, which is an AAALAC‐accredited institute conducting appropriate breeding and experiments for biomedical research that meet the standards of animal welfare. The siblings with Mbnl1+/+; Mbnl2+/+ were used as the wild‐type (WT) control group. All experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at Chang Gung Memorial Hospital, Keelung (IACUC No. 2014020501) and Linkou (IACUC No. 2015111001), Taiwan. All of the mice used in this study were 2 to 4 months of age, due to a reduced life expectancy (no more than 6 months) in this compound KO line. In total, 16 DM mutant and 18 WT control mice were used.
In Vivo Electrophysiological Study and Langendorff Heart Preparation
Mice were premedicated with xylazine (10 mg/kg IP) and zoletil (25 mg/kg IP), and anesthetized with isoflurane 1.0% to 2.0% in an anesthesia chamber. When the mice were fully anesthetized and unresponsive to physical stimuli, endotracheal intubation was performed for gas general anesthesia (isoflurane 0.5–1%) and the chests were opened via a median thoracotomy to expose the heart. To perform surface electrocardiography, electrodes were attached to limbs, and 3‐lead electrocardiography was performed with Axon Digidata (Molecular Devices, CA, USA) to record the PR interval and arrhythmias in vivo for 5 to 10 minutes. The hearts were then excised and rinsed in Tyrode's solution (in mmol/L: NaCl 125, KCl 4.5, NaH2PO4 1.8, NaHCO3 24, CaCl2 0.9, MgCl2 0.5, and glucose 5.5 in deionized water), cannulated via ascending aorta with a 20‐gauge stainless steel blunt needle and perfused at a flow rate of 1.5 to 3 mL/min to keep the average perfusion pressure around 80 cmH2O on a Langendorff apparatus with warm (37°C) oxygenated Tyrode's solution, which was bubbled with 95% O2/5% CO2 to maintain a pH of 7.40±0.05. The hearts were immersed in a temperature‐controlled tissue bath and maintained at 37°C. Pseudo‐ECG recordings were obtained using 3 spaced electrodes located in the bath surrounding the hearts.
Optical Mapping Studies
Dual optical mapping techniques were used to study the anterior aspect of the hearts. Calcium indicator Rhod‐2AM (5 μmol/L, dissolved in dimethyl sulfoxide via mixing with pluronic F‐127; Molecular Probes, OR, USA) and voltage indicator RH237 (1 μmol/L in 20 mL Tyrode's solution, dissolved in dimethyl sulfoxide; Molecular Probes) were used for Cai and Vm staining, respectively. Blebbistatin (15 μmol/L; Tocris Bioscience, Minneapolis, MN, USA) was added to the Tyrode's solution to reduce motion artifacts. The hearts were illuminated with a solid‐state, frequency‐doubled laser (Millennia, Spectra‐Physics Inc, Newport Corporation, Irvine, CA, USA) at a wavelength of 532 nm. Epifluorescence was collected simultaneously through 2 high‐speed cameras (MiCAM Ultima, BrainVision, Tokyo, Japan) at 1 ms per frame through a 580±20‐nm bandpass filter and a 715‐nm long‐pass filter for Cai and Vm images, respectively. Digital images (100×100 pixels) were gathered from the epicardium of the mapped ventricles (14×14 mm2 area), resulting in a spatial resolution of 140×140 μm2 per pixel. Baseline fluorescence level () of an individual pixel was first calculated for the duration of recording. The ratio of fluorescence () of an individual pixel was processed with both spatial (3×3 pixels Gaussian filter) and temporal (3 frames moving average) filtering to generate the maps. A bipolar lead was used to pace the lateral wall of the left ventricle at a double threshold. The effective refractory period was measured by giving a premature stimulus after 8 beats at a 200‐ms pacing cycle length (PCL). Arrhythmogenic alternans and conduction velocity (CV) were studied using a dynamic pacing protocol: decremental pacing at a PCL of 200 ms, then decreased in 10‐ms steps until loss of 1:1 capture or induction of ventricular tachyarrhythmias. Ventricular tachyarrhythmia inducibility was defined as the ability to provoke VT/VF with the dynamic pacing protocol and/or programmed extra stimuli (up to 4 premature stimuli). Sustained ventricular arrhythmia was defined as duration longer than 30 seconds.
Transmission Electron Microscopy of Cardiac Muscles
The mice were anesthetized and the hearts were harvested by thoracotomy. Primary fixation of cardiac muscle samples from the left ventricles were performed with 3% glutaraldehyde and 2% paraformaldehyde in 0.1 mol/L sodium cacodylate buffer for 1 hour at 4°C. Secondary fixation was performed with 1% osmium tetroxide in 0.1 mol/L cacodylate buffer at 4°C for 1 hour and followed by washing. For imparting additional contrast and further fixation, immersion using 2% uranyl acetate for 2 hours in a dark room was performed. The tissues were then dehydrated with a graded series of ethanol and immersed in a transient solvent of 100% propylene oxide. Specimens were embedded with a gradient infiltration of propylene oxide: Epon resin, followed by 100% resin under vacuum. Tissue blocks were sectioned into 75 nm thickness with ultramicrotome for examination under Hitachi H‐7500 transmission electron microscopy (Hitachi Ltd, Tokyo, Japan).
For protein extraction, the hearts were dissected and homogenized with CelLytic MT cell lysis reagent (Sigma‐Aldrich), followed by sonication and centrifugation. The protein concentrations were determined by protein assay (Bio‐Rad) and equal amounts (50 μg) of extracted proteins were loaded for analysis. The samples were resolved by 8% SDS‐PAGE and wet‐transferred to a polyvinylidene difluoride membrane overnight. The proteins were detected by immunoblotting using mouse monoclonal antibody anti‐Actin (Merck Millipore) and rabbit polyclonal antibody anti‐SCN5A (Aviva Systems Biology) as well as horseradish peroxidase–conjugated anti‐mouse or anti‐rabbit secondary antibody (Thermo Fisher Scientific) followed by chemiluminescent detection with Immobilon Western HRP Substrate (Merck Millipore).
Action potential duration (APD) was measured from 20% depolarization to 80% repolarization (APD80) at PCLs of 200, 150, 120, and 100 ms. Cai transient duration (CaiTD80) was measured by the same method.21 APD80 and CaiTD80 dispersion represent the SD of all available APD80 and CaiTD80 data at PCLs of 200, 150, 120, and 100 ms. We plotted APD restitution graphs of APD80 versus the preceding diastolic interval, defined as the interval between 80% repolarization and the onset of the next action potential, as previously described.22 The curves of APD restitution were constructed by polynomial fitting, and the steepest slopes of APD restitution curves were calculated at the diastolic interval. We used monoexponential fitting to compute time constant (τ) of the decay portion of the Ca2+ transient at a PCL of 200 ms. Vm and Cai alternans were defined as the differences in APD80 of 2 consecutive beats >4 ms and in Cai transient amplitude of 2 consecutive beats >10%. The phase was defined as positive for a short‐long APD80 and a small‐large Cai amplitude sequence (color coded by red) and as negative for a long‐short APD80 and a large‐small Cai amplitude sequence (color coded by green). Spatially discordant alternans (SDA) was evidenced by the presence of both red and green regions separated by a nodal line. To estimate CV, we measured the distance and conduction time between the earliest activation point and 2 epicardial points: one was from the pacing site to the left ventricular apex, and the other was along an axis parallel to the atrioventricular ring at PCLs of 200, 150, 120, 100, 90, 80, and 70 ms. When CV alternans occurred, we selected the slower ones for comparisons. We plotted CV restitution graphs of CV versus the preceding diastolic interval. The steepest curves of CV restitution graphs were constructed by polynomial fitting, and the slope of CV restitution curve was calculated at the shortest diastolic interval.
Continuous variables are shown in mean±SD values and categorical variables are presented as absolute numbers and percentages. Unpaired Student t test was performed for comparison between 2 groups. For the statistical comparisons of ventricular tachyarrhythmia inducibility, Fisher exact test was used. The Wilcoxon rank sum test was used to compare protein expression between the 2 groups. Statistical analyses were performed using SPSS statistical software for Windows version 22 (IBM Corp, Armonk, NY). A P<0.05 was considered statistically significant.
Twelve DM mutant mice (mean age 144±39 days) and 14 WT mice (mean age 158±52 days) (P=0.47) were included in the electrophysiological (including the following optical mapping) studies. Consistent with our previous report,20 the compound KO mice had significantly lower body weight compared with the WT mice (23.4±4.1 g versus 28.4±5.8 g, P=0.03). The mean dry heart weight was 0.22±0.04 g (DM mutant) versus 0.20±0.04 g (WT) (P=0.27). As a result, DM mutant mice had a significantly higher heart weight/body weight ratio than the WT mice (0.95% versus 0.71%, P<0.001).
Ultrastructural Analysis of DM Mutant Mice Revealed Morphological Defects Similar to Clinical DM Hearts
The DM mutant mice remained the same phenotypes of enhanced myotonia, muscular weakness, and sudden death after transferring from Gainesville, Florida (USA), to Keelung, Taiwan. To further investigate the ultrastructural abnormalities of KO hearts, we performed transmission electron microscopy studies on 3 pairs of WT and DM mutant mice, respectively. In contrast to parallel and tightly bounded myofibrils in WT mice (Figure 1A), mutant mice showed disarrangement and overcontraction of myofibrils with extended side‐to‐side junctions. In addition, numerous enlarged mitochondria accompanied with vacuolar spaces were also observed (Figure 1B). The enclosed areas in Figure 1A and 1B were magnified and shown in Figure 1C (WT) and 1D (DM mutant). Unlike the WT mice that showed organized segment and distinct Z lines, the myofibrils in DM mutant mice were destructed and replaced with vacuoles. These results further confirmed that the DM mutant mice used in this study possess typical DM cardiac manifestations, and these findings may underlie cardiac abnormalities including arrhythmia in DM mouse hearts.
DM Mutant Mice Showed Spontaneous Arrhythmogeneity During Baseline Recordings
Before general anesthesia, 2 of 12 DM mutant mice were apneic and underwent cardiac arrest soon after premedication of xylazine (10 mg/kg IP) and zoletil (25 mg/kg IP). The baseline ECG recordings revealed spontaneous episodes of incessant atrial tachycardia in nearly all (9 of 10) of the remaining mutant mice (Figure 2A‐a). In addition, nearly a third (3 of 10) of the DM mutant mice showed high‐grade atrioventricular block after intubation under general anesthesia (Figure 2A‐b). In contrast, none of the WT mice showed atrial tachycardia or atrioventricular block during in vivo baseline recording (P<0.001) (Figure 2B). Also, DM mutant mice had significantly longer PR intervals (44±6 ms, n=10) compared with the WT mice (38±4 ms, n=14) (P=0.014) (Figure 2C).
Optical Mapping Analysis Revealed No Change of Calcium Dynamics in DM Mutant Mice
The DM mutant mice had a significantly longer ventricular effective refractory period than WT mice (49±8 ms versus 39±9 ms, P=0.015) and longer APD80 at all PCLs than the WT mice (Figure 3A). In contrast, there was no significant difference of CaiTD80 at all PCLs between the 2 groups (Figure 4A). APD80 and CaiTD80 heterogeneity analyses showed no significant differences between the DM mutant and WT mice (Figures 3B and 4B). Representative recording of action potential and Cai tracings from the same heart are shown in Figures 3C and 4C. The Cai tracings and CaiTD80 maps show similar CaiTD80 between the 2 hearts. In addition to Cai transient duration, the τ values of Cai decay also showed no difference (WT 34±3 ms versus DM mutant 33±2 ms; P=0.26). The discrepancy of longer APD80 with similar CaiTD80 and Cai decay suggests that other ionic mechanisms, rather than Cai dynamic, may account for the APD80 prolongation in the DM mutant mice.
Slowing of CV in DM Mutant Mice
The DM mutant mice displayed significantly slower CV compared with the WT mice (Figure 5A). A representative example of isochronal maps obtained by dynamic pacing in both groups was shown in Figure 5B. Upon the stimulation with a series of PCLs at 200, 150, 120, 100, 90, 80, and 70 ms, the CVs measured both from the pacing site to left ventricular apex or the pacing site along atrioventricular ring were both significantly reduced. To describe the values in Figure 5B more specifically, the former CVs were 83, 76, 69, 69, 64, 59, and 52 cm/s in this particular WT mouse and 55, 51, 51, 48, 48, 43 and 42 cm/s in this particular DM mutant mouse and the latter were 88, 88, 77, 77, 77, 77, and 77 cm/s in this particular WT mouse and 59, 59, 53, 53, 48, 48, 41 cm/s in this DM mutant mouse, respectively. Note that functional conduction block was only induced in the DM mutant hearts (at a PCL of 70 ms, subpanel at right lower corner of Figure 5B), which could have perpetuated reentry formation and ventricular tachyarrhythmias.
DM Mutant Mice Showed Steeper CV Restitution Curve Slopes
APD restitution analyses showed no significant difference of APD restitution curve slope between the WT mice and DM mutant mice (1.125±0.172 versus 1.245±0.240; P=0.103 [Figure 6A]). Figure 6B shows representative APD restitution curves of a WT mouse and a DM mutant mouse. The DM mutant mice had significantly steeper slopes in CV restitution curves than the WT mice (Figure 6C). The slopes of CV restitution curves in WT versus DM mutant mice were 0.479±0.142 cm/s per ms versus 0.675±0.195 cm/s per ms and 0.478±0.145 cm/s per ms versus 0.654±0.137 cm/s per ms from the left ventricular lateral wall pacing site along the atrioventricular ring and to the left ventricular apex, respectively. Figure 6D shows representative CV restitution curves of a WT mouse and a DM mutant mouse.
The Induction of SDA was a Common Feature in Mbnl Compound KO Mice That Increased Susceptibility of Conduction Block
The spatially concordant Vm and Cai alternans could be induced simultaneously in all mice by decremental pacing in both groups, and there was no significant difference of the longest PCL to induce spatially concordant alternans (SCA) between the WT (94±10 ms) and DM mutant mice (98±14 ms, P=0.54). However, the DM mutant hearts showed a higher risk of SDA compared with the WT hearts: the SDA was more easily inducible in mutant (8 of 10, 80%) than WT mice (5 of 14, 36%; P=0.036). The longest PCL to induce SDA in DM mutant hearts was 79±6 ms, slightly longer than WT (70±10 ms, P=0.078). An example of SDA induction is shown in Figure 7. Progressive shortening of PCL to 80 ms induced SCA (Figure 7A, left panels). When PCL was further shortened to 70 ms, the SDA (Figure 7A, middle and right panels, and Figure 7C, left panel) was induced and the mapping field was separated by 2 nodal lines (arrows in Figure 7A, middle panel). These lines moved farther towards the pacing site at a shorter PCL of 60 ms. Note that Vm and Cai transients were out of phase in different regions of the heart during the SDA, which resulted in a dispersion of refractoriness and increased the susceptibility to functional conduction block.
The DM Mutant Mice are Predisposed to VT/VF Induction
The DM mutant mice were more susceptible to the induction of ventricular tachyarrhythmia. The inducibility of VT/VF was 9 of 10 (90%) in DM mutant mice and 5 of 14 (36%) in WT mice (P=0.011), and the sustainability of VT/VF were also significantly higher in mutant (7 of 10 [70%]), compared with WT (1 of 14 [7%]; P=0.002). A representative example of pseudo‐ECG findings of VT/VF induction was shown in Figure 2D. Both dynamic pacing and S1–S4 (300‐60‐60‐50 ms) pacing could induce sustained ventricular tachyarrhythmia in DM mutant hearts.
The possible causal relationship between SDA and ventricular tachyarrhythmia induction during dynamic pacing are illustrated in Figure 7. The pseudo‐ECG (Figure 7C, right upper panel) and Vm (Figure 7C, lower panel) recordings are shown simultaneously, and a fragmented Vm tracing was found at site “a” (located on the nodal line labeled in Figure 7B), at the initiation of ventricular tachyarrhythmia (red arrow in Figure 7C, lower panel a). The simultaneously recorded isochronal maps, corresponding to the short run ECG reading (labeled as the red bar) in Figure 7C (lower panel b), is shown in Figure 7D. The CV alternans occurred while the PCL was shortened to 50 ms (fast and slow CV at frames 75 and 125 ms, respectively). The functional conduction block (frame 175 ms) followed to generate reentrant wave fronts that eventually lead to sustained ventricular tachyarrhythmia (Video S1).
Immunoblotting of NaV1.5 Protein
To verify factors that may change the regulation of sodium current, we chose to analyze the expression level of the NaV1.5 (Scn5a) protein by immunoblotting. Previously, the mis‐splicing of SCN5A without affecting mRNA level has been reported in a DM patient study.23 Our normalized Western blot results showed no significant changes of NaV1.5 protein level between WT and DM mutant mice, which is compatible with the results seen in patients (Figure 8; n=4 in each group, P=0.564).
Cardiac manifestations are commonly seen in a variety of hereditary muscular dystrophies. Similar to other inherited muscular dystrophies, the cardiac manifestations in DM may come after skeletal muscular symptoms or they could be the presenting features. Compared with Duchenne muscular dystrophy, ventricular dysfunction and heart failure are much less common and only seen in end‐stage DM patients.24 The incidence of arrhythmia‐related sudden death is higher in DM1 patients. However, DM type II patients are also at risk for severe cardiac complications.13, 25 Pathological studies on DM patient heart samples revealed myocardial degeneration, fatty infiltration, or diffuse or focal fibrosis throughout myocardium or limited to the conduction system.26, 27 Electron microscopic studies showed disarray of myofibrils, defective mitochondrial morphology, indistinct Z lines, and numerous vacuoles.28, 29, 30 These features could also be seen in our Mbnl compound KO mice. The myofibril disarray may contribute to the cardiac presentations in DM patients, and fibrosis may serve as a substrate for reentrant arrhythmia.31, 32
Conduction block–related asystole or ventricular tachyarrhythmia may contribute to sudden death, and the latter may be more common in DM patients than expected.3, 7 To determine the mechanisms of ventricular tachyarrhythmias in DM, Merino et al33 performed invasive electrophysiological studies and concluded that bundle branch reentry accounts for the tachycardia. For these patients who had inducible ventricular tachyarrhythmias, they recommended radiofrequency catheter ablation as an effective intervention. In addition to conduction block, CV restitution and fixed electrophysiological heterogeneity of myocardium also contribute to the development of SDA.34, 35 Rapid heart rate and ectopic beats further enhance initiation of SDA, reentry, and development of arrhythmias.34 In our study, we performed simultaneous Vm and Cai optical mapping to investigate whether abnormal Cai dynamic plays a role in ventricular arrhythmogenesis. As shown in our mapping data, the differences of CaiTD80, Cai decay, and threshold of Cai alternans between mutant and WT hearts were insignificant. However, it is interesting that mutant hearts had significantly slower CV, steeper slope of CV restitution curve, and higher susceptibility to SDA induction, which also facilitated SDA development, reentry formation, and ventricular fibrillation induction.34 As shown in Figures 5 and 6, rate‐dependent conduction delay prolonged the time differences in depolarization between myocytes, thereby raising spatial gradients of activation and perpetuating the onset of arrhythmogenic SDA.36 When CV was reduced sufficiently for regaining excitability of the myocardial tissue ahead of the excitation wave front, reentrant circuits were formed and self‐generating. However, whether the reentry hypothesis could also be applied to polymorphic ventricular tachycardia, which is occasionally seen in DM patients, needs further investigation.13
From the molecular perspective, DM has a unique underlying mechanism supported by evidence from various mouse models.37, 38, 39, 40 Instead of perturbing the expression of the genes (eg, DM protein kinase), the pathogenic effects of extended microsatellite repeats are mediated through the hairpin structure formed by the transcribed (CC(U)G)n RNA. These double‐stranded hairpins interact with a variety of RNA‐binding proteins to form RNA foci and retain in the nuclei. Confocal images revealed these foci are only “colocalizing” with MBNL proteins, but not other RNA‐binding proteins.41 Although a variety of disease mechanisms have been suggested, potentially generated by the extended repeats (eg, repeat associated non‐ATG (RAN) translation),42 we hypothesized that the primary events in the RNA‐dominant model is the MBNL loss of function and proved with a series of Mbnl KO mouse models. Various transgenic mouse models expressing pathogenic repeats in specific tissues are available; however, mice simultaneously presenting myotonia and corresponding cardiac defects are extremely rare. Since myotonia is the hallmark symptom of DM, it is ideal for a DM mouse model to show this typical phenotype as a reliable control. To our knowledge, this is the first cardiac optical mapping study on DM mouse models and importantly, the Mbnl1−/−; Mbnl2+/− mice with robust myotonia.
Although various MBNL‐misregulated genes are related to calcium homeostasis or sensitive to calcium level (eg, Cacna1s, Tnnt2), several reports instead highlight the importance of sodium channel misregulation in DM arrhythmogenicity. In a study by Freyermuth et al,23 inclusion of the exon 6A of SCN5A (the gene that encodes NaV1.5 sodium channel) was increased in DM heart samples as well as in Mbnl1 KO; Mbnl2 HET mice. Through a series of electrophysiological analysis on different isoforms they concluded that the misregulation of SCN5A splicing accounts for DM arrhythmia.23 In concordance with this report, Wahbi et al43 also reported that Brugada syndrome may cause cardiac sudden death in DM patients and SCN5a mis‐splicing may play a role. Also, Pambrun et al44 reported that patients carrying the DM1 mutant allele from the paternal side could have exacerbated Brugada syndrome induced by sodium channel mutation coming from the maternal side. In addition to misregulation of SCN5A splicing, other DM animal models showed abnormal sodium channel gating.45, 46 Slow decay of sodium currents, late reopenings of sodium channels, and relatively depolarized resting membrane potential contribute to prolonged action potential in DM cardiomyocytes.45 It seems that exploring the dynamic nature of INa is promising for an integrated understanding of the mechanisms, which may directly link INa dysfunction to DM cardiac pathophysiology.
There are several limitations in this study. Although the phenotypic presentations are similar to clinical DM, the features of cardiac arrhythamias in this DM animal model might not be identical to arrhythmias in patients. Small and large animals have different expressions of calcium handling proteins and ion channels,47 leading to different action potential morphology and cardiac electrophysiology. This is still an observational study and characterization of an animal model rather than data from clinical patients, and therefore further investigation is required for clinically relevant mechanisms. Excitation‐contraction uncouplers are required to suppress motion artifacts in optical mapping studies, but these agents have significant electrophysiological effects, including APD prolongation, effective refractory period increase, and restitution curve alteration.48, 49 The electrophysiological presentations in optical mapping studies might be different from that in vivo. Ultrastructural analyses from transmission electron microscopy showed enlarged mitochondria with vacuole formation, and the changes may play a role in reduced cardiac contraction. It has been shown that DM protein kinase protects cells and mitochondria from oxidative stress and cell death in a DM1 mouse model.50 However, whether mitochondrion dysfunction directly influences cardiac conduction or calcium handling in our model needs further evaluation.
We confirmed that severe cardiac arrhythmias could develop in the Mbnl compound KO mice and that sudden death in this mouse line is reasonably cardiogenic. This lethal phenomenon was caused by a significantly slower CV that contributed to higher susceptibility of conduction disturbance and SDA. Since mutant mice have longer APD80 but no difference in CaiTD80 and Cai decay τ values compared with WT, ionic mechanisms other than Ca2+ may be involved in ventricular tachyarrhythmias in DM patients. These findings are highly compatible with current knowledge about DM arrhythmogenesis and may imply an INa dysfunction‐based mechanism of ventricular tachyarrhythmias. Clinically, since rapid ventricular rate may increase the risk of ventricular tachyarrhythmias in DM patients, cautious heart rate control with conduction system monitoring is indicated. Implantable cardioverter‐defibrillator and/or radiofrequency ablation rather than antiarrhythmic agents for tachyarrhythmias seems to be a suitable strategy based on the evidence of reduced CV and susceptibility to SDA induction observed in this DM mouse model.
Sources of Funding
This work was supported by Chang Gung Medical Research Grant, Keelung (CMRPG2D0181 and CMRPG2F0281 to Lee) and Linkou (CMRPG3F035 to Chou; CMRPG3D034 to Chang).
Video S1. Ventricular tachyarrhythmia was induced by dynamic pacing when pacing cycle length (PCL) was decreased from 60 to 50 ms. Note that the unidirectional conduction block and reentry formation preceded the onset of ventricular tachyarrhythmia. Multiple wavelets were noted after a few beats of meandering reentrant wave fronts.
The authors thank Dr Maurice Swanson, the professor in the Department of Molecular Genetics and Microbiology and the Center for NeuroGenetics, University of Florida, College of Medicine, for the generous agreement of transferring Mbnl1 and Mbnl2 KO mouse lines. The transmission electron microscopy studies were facilitated by the Microscope Core Laboratory, Chang Gung Memorial Hospital, Linkou, Taiwan. The authors also thank Carol Seah for her contributions on mouse husbandry and immunoblotting analyses.
This work was presented in part at the American Heart Association Scientific Sessions, November 7‐11, 2015, in Orlando, FL.
- ↵Fu YH, Pizzuti A, Fenwick RG Jr., King J, Rajnarayan S, Dunne PW, Dubel J, Nasser GA, Ashizawa T, de Jong P, Wieringa B, Korneluk R, Perryman MB, Epstein HF, Thomas Caskey C. An unstable triplet repeat in a gene related to myotonic muscular dystrophy. Science. 1992;255:1256–1258.
- ↵Liquori CL, Ricker K, Moseley ML, Jacobsen JF, Kress W, Naylor SL, Day JW, Ranum LP. Myotonic dystrophy type 2 caused by a CCTG expansion in intron 1 of ZNF9. Science. 2001;293:864–867.
- ↵Clarke NR, Kelion AD, Nixon J, Hilton‐Jones D, Forfar JC. Does cytosine‐thymine‐guanine (CTG) expansion size predict cardiac events and electrocardiographic progression in myotonic dystrophy? Heart. 2001;86:411–416.
- ↵Pelargonio G, Dello Russo A, Sanna T, De Martino G, Bellocci F. Myotonic dystrophy and the heart. Heart. 2002;88:665–670.
- ↵O'Rourke JR, Swanson MS. Mechanisms of RNA‐mediated disease. J Biol Chem. 2009;284:7419–7423.
- ↵Goodwin M, Mohan A, Batra R, Lee KY, Charizanis K, Gomez FJ, Eddarkaoui S, Sergeant N, Buee L, Kimura T, Clark HB, Dalton J, Takamura K, Weyn‐Vanhentenryck SM, Zhang C, Reid T, Ranum LP, Day JW, Swanson MS. MBNL sequestration by toxic RNAs and RNA misprocessing in the myotonic dystrophy brain. Cell Rep. 2015;12:1159–1168.
- ↵Kanadia RN, Johnstone KA, Mankodi A, Lungu C, Thornton CA, Esson D, Timmers AM, Hauswirth WW, Swanson MS. A muscleblind knockout model for myotonic dystrophy. Science. 2003;302:1978–1980.
- ↵Charizanis K, Lee KY, Batra R, Goodwin M, Zhang C, Yuan Y, Shiue L, Cline M, Scotti MM, Xia G, Kumar A, Ashizawa T, Clark HB, Kimura T, Takahashi MP, Fujimura H, Jinnai K, Yoshikawa H, Gomes‐Pereira M, Gourdon G, Sakai N, Nishino S, Foster TC, Ares M Jr., Darnell RB, Swanson MS. Muscleblind‐like 2‐mediated alternative splicing in the developing brain and dysregulation in myotonic dystrophy. Neuron. 2012;75:437–450.
- ↵Lee KY, Li M, Manchanda M, Batra R, Charizanis K, Mohan A, Warren SA, Chamberlain CM, Finn D, Hong H, Ashraf H, Kasahara H, Ranum LP, Swanson MS. Compound loss of muscleblind‐like function in myotonic dystrophy. EMBO Mol Med. 2013;5:1887–1900.
- ↵Chou CC, Zhou S, Miyauchi Y, Pak HN, Okuyama Y, Fishbein MC, Karagueuzian HS, Chen PS. Effects of procainamide on electrical activity in thoracic veins and atria in canine model of sustained atrial fibrillation. Am J Physiol Heart Circ Physiol. 2004;286:H1936–H1945.
- ↵Freyermuth F, Rau F, Kokunai Y, Linke T, Sellier C, Nakamori M, Kino Y, Arandel L, Jollet A, Thibault C, Philipps M, Vicaire S, Jost B, Udd B, Day JW, Duboc D, Wahbi K, Matsumura T, Fujimura H, Mochizuki H, Deryckere F, Kimura T, Nukina N, Ishiura S, Lacroix V, Campan‐Fournier A, Navratil V, Chautard E, Auboeuf D, Horie M, Imoto K, Lee KY, Swanson MS, Lopez de Munain A, Inada S, Itoh H, Nakazawa K, Ashihara T, Wang E, Zimmer T, Furling D, Takahashi MP, Charlet‐Berguerand N. Splicing misregulation of SCN5A contributes to cardiac‐conduction delay and heart arrhythmia in myotonic dystrophy. Nat Commun. 2016;7:11067.
- ↵Nguyen HH, Wolfe JT III., Holmes DR Jr., Edwards WD. Pathology of the cardiac conduction system in myotonic dystrophy: a study of 12 cases. J Am Coll Cardiol. 1988;11:662–671.
- ↵Forsberg H, Olofsson BO, Eriksson A, Andersson S. Cardiac involvement in congenital myotonic dystrophy. Br Heart J. 1990;63:119–121.
- ↵Harper PS. Myotonic Dystrophy. London, England: W.B. Saunders; 2001.
- ↵Turillazzi E, Neri M, Riezzo I, Di Paolo M, Evangelisti L, Fineschi V. Cardiac fibrosis, arrhythmia and sudden death in myotonic dystrophy type 1: could TGF‐β1 improve the predictive accuracy of patients at risk, opening new therapeutic challenges? Int J Cardiol. 2013;168:4976–4978.
- ↵Merino JL, Carmona JR, Fernandez‐Lozano I, Peinado R, Basterra N, Sobrino JA. Mechanisms of sustained ventricular tachycardia in myotonic dystrophy: implications for catheter ablation. Circulation. 1998;98:541–546.
- ↵Weiss JN, Karma A, Shiferaw Y, Chen PS, Garfinkel A, Qu Z. From pulsus to pulseless: the saga of cardiac alternans. Circ Res. 2006;98:1244–1253.
- ↵Jansen G, Groenen PJ, Bachner D, Jap PH, Coerwinkel M, Oerlemans F, van den Broek W, Gohlsch B, Pette D, Plomp JJ, Molenaar PC, Nederhoff MG, van Echteld CJ, Dekker M, Berns A, Hameister H, Wieringa B. Abnormal myotonic dystrophy protein kinase levels produce only mild myopathy in mice. Nat Genet. 1996;13:316–324.
- ↵Mankodi A, Logigian E, Callahan L, McClain C, White R, Henderson D, Krym M, Thornton CA. Myotonic dystrophy in transgenic mice expressing an expanded CUG repeat. Science. 2000;289:1769–1773.
- ↵Fardaei M, Larkin K, Brook JD, Hamshere MG. In vivo co‐localisation of MBNL protein with DMPK expanded‐repeat transcripts. Nucleic Acids Res. 2001;29:2766–2771.
- ↵Wahbi K, Algalarrondo V, Becane HM, Fressart V, Beldjord C, Azibi K, Lazarus A, Berber N, Radvanyi‐Hoffman H, Stojkovic T, Behin A, Laforet P, Eymard B, Hatem S, Duboc D. Brugada syndrome and abnormal splicing of SCN5A in myotonic dystrophy type 1. Arch Cardiovasc Dis. 2013;106:635–643.
- ↵Pambrun T, Mercier A, Chatelier A, Patri S, Schott JJ, Le Scouarnec S, Chahine M, Degand B, Bois P. Myotonic dystrophy type 1 mimics and exacerbates Brugada phenotype induced by Nav1.5 sodium channel loss‐of‐function mutation. Heart Rhythm. 2014;11:1393–1400.
- ↵Lee HC, Patel MK, Mistry DJ, Wang Q, Reddy S, Moorman JR, Mounsey JP. Abnormal Na channel gating in murine cardiac myocytes deficient in myotonic dystrophy protein kinase. Physiol Genomics. 2003;12:147–157.
- ↵Baker LC, Wolk R, Choi BR, Watkins S, Plan P, Shah A, Salama G. Effects of mechanical uncouplers, diacetyl monoxime, and cytochalasin‐D on the electrophysiology of perfused mouse hearts. Am J Physiol Heart Circ Physiol. 2004;287:H1771–H1779.