We created a mouse model with a prolonged Q-T interval and spontaneous arrhythmias by overexpressing the NH2terminus and first transmembrane segment (Kv1.1N206Tag) of a delayed rectifier potassium channel (LQT+/ − mouse). Analyses were performed using whole cell recordings of cardiac myocytes, surface electrocardiography, and programmed electrical stimulation. Action potential duration (APD) was prolonged to the same extent and was more highly variable in myocytes derived from LQT+/ − and LQT+/+ mice than in myocytes derived from wild-type (WT) FVB mice. Under ketamine anesthesia, the Q-T interval of both LQT+/+ and LQT+/ − mice was comparably prolonged versus that of WT mice. Stimulation of the right ventricle using an intracardiac catheter induced polymorphic ventricular tachyarrhythmias in 50% of the LQT+/ − mice and 36% of the LQT+/+ mice, whereas polymorphic ventricular tachyarrhythmias were not inducible in WT mice. The analyses of LQT+/ − and LQT+/+ mice indicate that prolongation of the Q-T interval in LQT mice is associated with prolonged APD, increased dispersion of APD among cardiocytes, and inducibility of polymorphic ventricular tachycardia, providing the substrate for spontaneous arrhythmias in these animals.
- transgenic mice
- long Q-T syndrome
- Q-T interval
- programmed stimulation
- ventricular arrhythmias
genetically engineered mouse models may increase our understanding of the pathophysiology of inherited human diseases caused by mutations in genes that control cardiac development and function. Prolongation of the Q-T interval, polymorphic ventricular tachycardia, and an increased incidence of sudden death characterize the congenital long-QT syndrome (LQTS) (15). The prolonged ventricular repolarization and dispersion of refractoriness in LQTS is related to loss-of-function mutations that occur in genes encoding α- or β-subunits of voltage-gated potassium channels in the heart such as KvLQT1,MinK or HERG (6, 16-19). However, the precise mechanism by which the mutations lead to cardiac arrhythmias is uncertain. Many pharmacological models of LQTS have been proposed; however, no genetic animal models are currently available.
Koren and colleagues (2) showed previously that an NH2-terminal fragment that contains the first transmembrane segment of the rat delayed rectifier K+ channel Kv1.1 (Kv1.1N206Tag) coassembles with Shaker-like K+ channels in vitro and inhibits the currents coded by Kv1.5 cDNA in a dominant negative manner when coexpressed inXenopus oocytes. Recently, Koren and co-workers (8) demonstrated that the mechanism of the dominant negative effect ofKv1.1N206Tag involves the trapping of Kv1.5 and Kv1.4 polypeptides in the endoplasmic reticulum. Cardiac myocytes derived from these mice have prolonged action potential duration (APD) because of a significant reduction in the current density of a rapidly activating, slowly inactivating 4-aminopyridine-sensitive outward K+ current (I slow) (12, 21). Overexpression of Kv1.1N206Tag in the heart of heterozygous transgenic mice (LQT+/ −) resulted in prolongation of the Q-T interval and spontaneous ventricular arrhythmias (21). Recently, several other models, using a dominant negative or gene knockout approach against Kv4, MinK, orMERG, with or without comparable Q-T prolongation, resulted in inhibition of the currents encoded by these genes (1, 4, 9). Moreover, the attenuation of Kv4-encoded currents also resulted in a significant prolongation of the Q-T interval (4).
In the present study, we characterized mice homozygous for the transgene and compared the effect of additional copies of the transgene on APD and Q-T intervals. Moreover, we used this model system to determine whether sustained ventricular arrhythmias can be reproducibly induced by programmed ventricular stimulation in the LQT mouse and to evaluate the surface electrocardiogram (ECG) morphology of any induced ventricular arrhythmias.
Derivation of the heterozygous (LQT+/ −) mice used in this study has been described previously (12). To characterize the effects of additional copies of the transgene, mice homozygous for the Kv1.1N206Tag transgene (LQT+/+) were obtained by breeding heterozygous male and female mice. Homozygosity of the F1 offspring was then evaluated by back-breeding with wild-type (WT) FVB mice. Studies were attempted in 19 WT, 20 LQT+/ −, and 16 LQT+/+animals. Baseline ECGs were obtained on all animals. However, only animals that completed the full stimulation protocol were included in the electrophysiological analyses (15 WT, 14 LQT+/ −, and 14 LQT+/+animals). The animals ranged in age from 12 to 30 wk. There were no significant differences in body weight [31.8 ± 2.4 vs. 32.9 ± 2.4 vs. 32.4 ± 3.7 g, P = not significant (NS)] or age (18 ± 5 vs. 20 ± 6 vs. 18 ± 5 wk, P = NS) among WT, LQT+/ −, and LQT+/+ animals, respectively. The average duration of the protocol was 59 ± 16 min.
All studies were performed under anesthesia with ketamine (50 mg/kg) and xylazine (5 mg/kg) during spontaneous respiration. The average ketamine dose used in control mice was 82.5 μg compared with 80 and 78.5 μg in LQT+/ − and LQT+/+ mice (P = NS). Studies were performed in accordance with the guidelines of the Harvard Medical Area Standing Committee on Animals, following approval by the Institutional Animal Care and Use Committee.
After the induction of anesthesia, a 12-lead electrocardiogram was obtained by placing needle electrodes subcutaneously in each limb (27-gauge needles) and by moving a single chest electrode to sequentially record the 6 V leads. The signal was amplified and filtered (from 1 Hz to 100 Hz; PPG ARC 6 recorder), digitized at a 500-Hz sampling rate using a 16-bit analog-to-digital converter (AT-M10–16XE-10; National Instruments), and stored on the hard drive of a personal computer. All ECG intervals (P, P-R, QRS, Q-T, Q-Tc, J-T, J-Tc, and R-R) and axes (P, QRS, T) were measured for each mouse by a blinded observer using the 12-lead printout. To correct the Q-T and J-T intervals (Q-Tc and J-Tc), we used the recently established correction formula for mice (12, 14).
Electrophysiological study protocol.
With the use of an operating microscope, the right internal jugular vein was dissected and encircled with two 5-0 silk sutures without disrupting the blood flow. An incision was made between the sutures with microscissors. A quadrupolar 2-F stimulation catheter (NuMed, Nicholville, NY) was advanced to the right ventricle. The heart rate and cardiac rhythm were monitored continuously throughout the procedure. A heating lamp was used to maintain body temperature.
The stimulation protocol we used was similar to electrophysiological protocols used in humans (13) and recently adapted for mice (5). Briefly, bipolar voltage pacing thresholds were measured at both locations using a 1-ms pulse width. Stimulation was subsequently performed using twice the diastolic voltage threshold. Ventricular refractory period was measured using pacing trains of 20 beats at a basic cycle length of 100 ms with a single extrastimulus. The S1 interval was decremented by 10 ms between trains. However, in a preliminary analysis, this 10-ms interval proved too coarse to detect a difference among the groups. Therefore, the protocol was modified to use a 10-ms decremental interval to establish a preliminary right ventricular effective refractory period (RVERP). The S1 interval was then increased by 8 ms from this preliminary value, followed by 2-ms decrements to determine the final RVERP. This modified protocol was performed in seven WT, seven LQT+/ −, and six LQT+/+animals.
To induce ventricular arrhythmias, we introduced extrastimuli following a 10- or 20-beat train with a basic cycle length of 100 ms. After the train, one extrastimulus was added. If arrhythmia was not inducible in the mouse, second and third stimuli were added in a stepwise fashion. In those mice in which arrhythmia was not inducible following the addition of the third extrastimulus, the protocol was repeated following a train with a basic cycle length of 70 ms. To assess the adrenergic effects on the susceptibility to ventricular arrhythmias that were not inducible in mice at baseline, 30 μg of isoproterenol (200 μg /ml) were given intraperitoneally to obtain a stable heart rate increase of at least 30% over the basal heart rate under anesthesia. After several minutes, the ventricular pacing protocol was repeated. Evoked groups of one to four extrasystoles were defined as ventricular premature depolarizations, and five or more consecutive complexes were defined as ventricular tachycardia. Ventricular tachycardia was defined as discrete complexes with a clearly definable rate, whereas ventricular fibrillation was defined as a morphology that changes from beat to beat with no clearly definable rate (20).
Southern and Northern blotting.
Tail DNA was digested with BamH I and transferred to a nylon membrane. The probe used contained a 505-bp PCR fragment derived from the αMyHC promoter (position 585 to 1090 bp) labeled with the DIG system (Boehringer). The film was exposed for 30 min. Total RNA was isolated using standard techniques. The probe used was a 600-bpEcoR I-Cla I fragment derived from Kv1.1. Film was exposed overnight.
Myocytes preparation and cellular electrophysiological study.
Single myocytes were obtained as described previously (21) with a few modifications. Briefly, Langendorff-perfused hearts were prepared from the control and age-matched LQT adult FVB mice and subjected to 3–5 min of perfusion with nominally Ca2+- free Tyrode solution [containing (in mM) 137 NaCl, 5.4 KCl, 1.2 MgCl2, 10 glucose, 10 HEPES, and 0.25 Na2-EDTA; pH 7.35], followed by a perfusion with enzyme-containing Tyrode solution [0.5 mg/ml collagenase (type I; Worthington, Lakewood, NJ), 0.25 mg/ml hyaluronidase (Sigma, St. Louis, MO), and 1 mg/ml fatty acid-free BSA (Sigma)] for 8–12 min. The ventricles were then chopped into small pieces, incubated at 37°C in the same enzyme solution for 5–10 min, and mechanically dispersed. After repeated washings by a series of centrifugations at 500 rpm for 2–3 min, followed by resuspensions in 0.25 and then 0.5 mM Ca2+-containing Tyrode solution supplemented with 1–2% FCS, cells were stored in normal Tyrode solution (1.0 mM Ca2+) containing 2% FCS. Calcium-tolerant, rod-shaped ventricular myocytes with clear striations were selected randomly for the electrophysiological studies. APD was evaluated at room temperature as described previously (21).
Continuous variables were compared using an ANOVA with Sheffé's post hoc analysis. A Levene test was performed to test for homogeneity of variance. Variables with unequal variances were first log transformed and then retested. If the Levene statistic remained significant, the untransformed variable was compared using the Kruskal-Wallis H test. A chi-square test was used for categorical variables. All data are presented as means ± SD.P values <0.05 were considered significant.
Southern blotting of tail DNA obtained from mice homozygous for the transgene (LQT+/+) revealed an increased intensity of the transgene relative to the native αMyHC gene in homozygotes compared with the heterozygotes (LQT+/ −) (Fig.1 A). Furthermore, breeding the LQT+/+ mice with control FVB mice resulted in F1 litters that were all heterozygous for the transgene (data not shown), confirming the presence of duplicate copies of the transgene in LQT+/+ mice. Northern blot analysis of total cardiac RNA revealed that Kv1.1N206Tag transcript is not detectable in WT mice. Moreover, in LQT+/+ mice, an increase in the steady-state levels of Kv1.1N206Tag transcript was observed (Fig. 1 B). To determine whether increased expression of the transgene transcript modulates the phenotype, we characterized the APD. Table 1 shows the APD parameters obtained from cardiac myocytes derived from WT, LQT+/ −, and LQT+/+ hearts. As reported previously (21), APD varied considerably with a significant difference in variance among the groups (Levene statistics: APD30, P < 0.02; APD50, P < 0.001; APD70, P = 0.001; APD90,P = 0.064; where APD30–APD90 are APDs at 30–90% of full repolarization). With the use of nonparametric testing, the APD30, APD50, APD70 and APD90 of cardiocytes derived from LQT+/ − and LQT+/+ mice were significantly prolonged compared with those derived from WT mice. However, there were no differences between the APDs of LQT+/ − and LQT+/+ mice (Fig.2 A). The frequency distribution of APD50 and APD70 (Fig. 2, B andC) reveals that both LQT+/ − and LQT+/+ myocytes tended to show a positive skew with a significantly larger variance, indicating that these mice have cardiocytes with substantially longer APDs as well as cardiocytes with APDs that overlap those of WT cardiocytes.
Representative lead I surface ECG recordings of five anesthetized WT (Fig. 3 A) and five LQT+/ − and LQT+/+ mice are shown (Fig. 3, B and C). A prolonged Q-T interval was found in LQT mice compared with that in WT mice (Table2). Comparable differences were detected in the J-T and J-Tc intervals (Table 2 and Fig. 3). Thus both the LQT+/ − and LQT+/+ mice were characterized by a marked and comparable prolongation of the Q-T interval. The average heart rate under ketamine anesthesia did not differ among the three groups (Table 2). The P-R and QRS intervals did not differ among the three groups. There was a tendency for a more leftward P-wave axis in the LQT mice compared with that in WT mice. This difference was statistically significant only for the comparison between LQT+/ − and WT mice. There were no significant differences in the axis of the QRS or T wave (Table 2).
Ventricular pacing thresholds were comparable in the WT, LQT+/ −, and LQT+/+ animals (0.6 ± 0.2 vs. 0.7 ± 0.3 vs. 0.6 ± 0.3 V, respectively, allP = NS). When ventricular refractory period was assessed with a 2-ms resolution, significant prolongation was demonstrated only in the LQT+/+ animals (WT: 53 ± 6; LQT+/ −: 56 ± 6; LQT+/+: 65 ± 8 ms; P < 0.02 for the main effect and P < 0.05 for WT vs. LQT+/+). Baseline Q-Tc intervals for these animals were 56 ± 4, 76 ± 4, and 77 ± 8 ms, respectively (P < 0.001). Both LQT+/ − and LQT+/+ animals had a significantly prolonged Q-Tc interval compared with WT animals (P < 0.05 for both post hoc comparisons).
Programmed ventricular stimulation induced ventricular premature depolarizations in 1 of 15 WT, 13 of 14 LQT+/ −, and 10 of 14 LQT+/+animals (P < 0.001, Fig. 4). Ventricular tachycardia was not inducible in any WT animals with up to three extrastimuli and isoproterenol but was inducible in 7 of 14 LQT+/ − and 5 of 14 LQT+/+animals (P < 0.01, Figs. 4,5 A, and 5 B, i,ii, and iii). The likelihood of induction in LQT+/ − did not differ significantly from that in LQT+/+ mice. Of the mice in which ventricular tachycardia was inducible, isoproterenol was required for induction in 5 of 7 LQT+/ − and 2 of 5 LQT+/+ mice. Ventricular arrhythmias had a polymorphic configuration with a duration of 300–6,488 ms (Fig. 5). All arrhythmias terminated spontaneously. Short-duration ventricular tachycardia was generally followed by sinus rhythm, whereas longer-duration ventricular tachycardia was followed by escape complexes with sinus bradycardia and eventual recovery of sinus rhythm, suggesting hemodynamic compromise during prolonged episodes of ventricular tachycardia (Fig. 5 A). However, all animals recovered from all episodes of ventricular tachycardia. No episodes of ventricular fibrillation were observed.
We used 12-lead surface ECG and programmed electrical stimulation of mice to further assess the role of the dominant negative mutantKv1.1N206Tag transgene in cardiac excitation. We report that under ketamine and xylazine anesthesia, homozygous and heterozygous LQT mice had prolongation of the Q-T interval. Analysis of APD revealed similar prolongation in both LQT+/ − and LQT+/+ mice. Analysis of the RVERP revealed that it is significantly prolonged in LQT+/+ mice, with an intermediate value in LQT+/ − mice that was not significantly different from that in either WT or LQT+/+ mice. Programmed electrical stimulation of the right ventricle using a transvenous catheter induced polymorphic ventricular tachycardia in a significant proportion of LQT+/ − and LQT+/+ mice, whereas ventricular tachycardia was not inducible in WT FVB mice. The results of this study demonstrate that the suppression of a specific potassium current in mouse ventricular myocytes by genetic manipulation resulted in prolongation of APD and Q-T interval and in inducible polymorphic ventricular arrhythmias that are similar to those observed in patients with LQTS.
Transgenic mouse models offer the opportunity to evaluate the physiological correlates of altered expression of voltage-gated potassium channels (11). In this study, we described the presence of Q-T prolongation in LQT mice. We evaluated the relationship between the transgene copy number, APD, and Q-T interval in mice. Doubling of the number of copies of the transgene resulted in an increase in the steady-state levels of the transcript. However, there was not further prolongation of the Q-T interval or APD with increasing copy number. These results confirm previous observations (12, 21) that the level of expression of the transgene transcript and polypeptide in heterozygous mice was sufficient to suppress the steady-state level of Kv1.5 polypeptide and inhibit most or all I slow (12, 21). The multiple copies of Kv1.1N206 transgene that are integrated into the genome of heterozygous LQT mice result in the expression of a sufficient amount of the dominant negative polypeptide. Thus, even in the presence of two wild-type alleles, the overexpressed truncated Kv1.1N206 polypeptide could suppress most of the expression ofI slow, and further duplication did not have an additive dominant negative effect.
Here we demonstrated that cells derived from LQT mice have longer APD with substantially larger variance. We speculate that the cells with shorter APD may originate from a different region of the heart. Indeed, Baker et al. (3) showed that APDs of epicardial cardiocytes in the WT mouse heart were heterogenous, with the mean APD75 at the apex shorter by 10 ms than the APD at the base. The heterogeneity of APDs was more pronounced in transgenic hearts, with the mean APD75 at the apex shorter by 29 ms than the APD at the base. Furthermore, the effective refractory period at the apex of the left ventricle is substantially shorter than at the base in Langendorff-perfused LQT+/ − hearts (3). We therefore studied the outward potassium currents expressed in cardiac myocytes derived from either the apex or the base of the left ventricle. Our studies show that LQT+/ −cardiocytes derived from these regions express similar densities of transient outward K+ currents (I to) (J Zhou, M Murata, P Buckett, P Hwang, and G Koren, unpublished observations). Interestingly, in LQT myocytes, the elimination of I slow was associated with a 70% induction of a TEA-sensitive current with slower inactivation kinetics (I Kslow2) (22). Ongoing studies in our laboratory suggest that a higher induction of I Kslow2 in apical cardiocytes may account for restricted prolongation of the APDs at the apex and an increase in the APD heterogeneity observed in LQT hearts.
Here we report that, as expected, RVERP is significantly prolonged in LQT+/+ mice compared with that in controls, whereas in LQT+/ − mice the difference did not achieve statistical significance. The lack of significance may arise from variable positioning of the catheter within different locations in the right ventricle, because there is a significant gradient of the effective refractory period between the apex and base (shorter at the apex) (3). The slightly shorter RVERP in LQT+/ − mice compared with LQT+/+ may have resulted from fortuitous placement of the catheter in a more apical location in LQT+/ − mice.
Cardiac action potential represents a balance between multiple depolarizing and repolarizing currents. Reduced outward currents through voltage-gated potassium channels or persistent depolarizing currents through voltage-gated sodium channels can prolong the APD and create the long Q-T phenotype (6, 10, 16-19). Slowing of the heart rate in the presence of action potential prolongation may lead to increased incidence of early afterdepolarization, which may trigger polymorphic ventricular tachyarrhythmias such as torsade de pointes. These arrhythmias can later deteriorate to ventricular fibrillation and sudden death. Several mechanisms have been proposed to explain the development of polymorphic ventricular tachycardia and ventricular fibrillation. Recent work suggests that fast polymorphic ventricular tachycardia (with or without torsade de pointes configuration) is caused by reentrant excitation, which may be initiated by subendocardial (triggered) activity (7).
Recently, Barry et al. (4) reported the generation and characterization of transgenic mice with the long Q-T phenotype following overexpression of a dominant negative Kv4 polypeptide in the heart. The expression of Kv4.2 pore mutant (Kv4.2W362F) in ventricular myocytes resulted in attenuation of I to. The APD and Q-T prolongations were both greater than those observed in our LQT mice. However, in contrast to our LQT mice, induction of arrhythmias could not be elicited (JM Nerbonne, personal communication). These observations clearly suggest that prolongation of the APD and Q-T interval are not sufficient for induction of arrhythmias. As previously described, our LQT+/ − mice are characterized by an increase in the frequency of ventricular premature beats and spontaneous episodes of ventricular tachycardia during conscious telemetric monitoring of mice heterozygous for the transgene (12). Baker et al. (3) also showed that programmed stimulation at the apex of hearts derived from our LQT+/ −mice perfused on a Langendorff system can elicit long-lasting reentrant ventricular tachycardia. Optical mapping experiments revealed that the ventricular tachycardia is initiated when the premature beat encounters a functional line of block that causes unidirectional propagation. Subsequent beats of ventricular tachycardia could be maintained for >30 min. That work established that spatial dispersion of repolarization with substantially longer effective refractory period at the base created functional block and reentry, which underlies the mechanism for ventricular tachycardia induced by programmed electrical stimulation. Here we demonstrated that programmed electrical stimulation of the right ventricle, in vivo, induced torsade de pointes-like polymorphic ventricular tachycardia. We speculate that programmed electrical stimulation of the right ventricle may induce more than one reentrant pathway, which may underlie the polymorphic and transient nature of the arrhythmia we observed in LQT mice. Thus these LQT mouse ventricles can sustain what is likely to be a reentrant arrhythmia with morphological features that are similar to clinically observed torsade de pointes.
Limitations of our study.
Polymorphic ventricular tachycardia cannot be induced in control mice. In that respect mouse hearts differ from those of larger mammals, including humans, in which polymorphic ventricular tachycardia may be induced even in normal hearts with a maximal stimulation protocol (S3) that includes isoproterenol. Initiation of ventricular tachycardia using programmed electrical stimulation is somewhat artificial; however, in every other regard, including the changing morphology and relatively high rates, this induced arrhythmia in LQT mice resembles clinical torsade de pointes. We recognize the limitations of a small animal model for arrhythmias. However, we would like to stress that these small animal models provide a unique opportunity to evaluate the mechanism of arrhythmias both in vivo and in vitro in intact hearts, where all cells are affected by the repolarization abnormality. This model will also provide an invaluable opportunity to assess the feasibility of in vivo gene transfer techniques as an approach to treatment. Such techniques, which can be somewhat nonuniform and therefore potentially of no value in the intact heart, can be evaluated in our model, which has a stable, readily assessable phenotype in terms of both a long Q-T interval on ECG and, more importantly, inducibility of ventricular tachycardia.
We have previously shown that ketamine anesthesia prolongs the Q-T interval in WT mice (14) and that this effect is enhanced in LQT+/ − mice (12). However, sustained arrhythmia can be induced in the Langendorff-perfused LQT+/ − mouse heart in the absence of ketamine. Furthermore, we have observed spontaneous ventricular arrhythmias in conscious, monitored LQT+/ − mice, suggesting that ketamine is not necessary for arrhythmia induction in this model.
In summary, the LQT mouse is the only genetic model with ventricular arrhythmias that resemble those observed in patients with LQTS. This model will be used to test for novel genetic and pharmacological therapeutic approaches to preventing channel-related arrhythmias in this and other models of LQTS.
During this study, A. Jeron was a recipient of a fellowship from the Deutsche Forschungsgemeinschaft (Je235/1-1). B. London was a recipient of National Heart, Lung, and Blood Institute (NHLBI) Grant HL-02843 and an American Heart Association (AHA) Grant in Aid. G. Koren was a recipient of NHLBI Grant HL-46005 and an AHA Established Investigator Award. M. Murata is an awardee of a Research Fellowship from the Japan Heart Institute.
Address for reprint requests and other correspondence: G. Koren, Cardiovascular Division, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115 (E-mail:).
Part of this study was presented as an abstract at the AHA Annual Meeting, Dallas, TX, November 1998.
Present address of A. Jeron: Medizinische Klinik II, Universität Regensburg, Franz-Josef Strauss-Allee 11, 93042 Regensburg, Germany.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society