Mechanisms of cardiac arrhythmias and sudden death in transgenic rabbits with long QT syndrome

Long QT syndrome (LQTS) is a heritable disease associated with ECG QT interval prolongation, ventricular tachycardia, and sudden cardiac death in young patients. Among genotyped individuals, mutations in genes encoding repolarizing K⁺ channels (LQT1:KCNQ1; LQT2:KCNH2) are present in approximately 90% of affected individuals. Expression of pore mutants of the human genes KCNQ1 (KvLQT1-Y315S) and KCNH2 (HERG-G628S) in the rabbit heart produced transgenic rabbits with a long QT phenotype. Prolongations of QT intervals and action potential durations were due to the elimination of I_Ks and I_Kr currents in cardiomyocytes. LQT2 rabbits showed a high incidence of spontaneous sudden cardiac death (>50% at 1 year) due to polymorphic ventricular tachycardia. Optical mapping revealed increased spatial dispersion of repolarization underlying the arrhythmias. Both transgenes caused downregulation of the remaining complementary I_Kr and I_Ks without affecting the steady state levels of the native polypeptides. Thus, the elimination of 1 repolarizing current was associated with downregulation of the reciprocal repolarizing current rather than with the compensatory upregulation observed previously in LQTS mouse models. This suggests that mutant KvLQT1 and HERG interacted with the reciprocal wild-type α subunits of rabbit ERG and KvLQT1, respectively. These results have implications for understanding the nature and heterogeneity of cardiac arrhythmias and sudden cardiac death.

The long QT syndrome (LQTS) is a familial, usually autosomal dominant disease, characterized by prolongation of the QT interval in the surface ECG, spontaneous polymorphic ventricular tachycardia (VT, often termed torsade de pointes [TdP]), and sudden cardiac death in young patients (1–3). At least 9 different forms of the disease have been described so far, but in the vast majority of successfully genotyped patients, the underlying causes are mutations in the voltage-gated K⁺ channels, either in the KQT-like subfamily, member 1 (KCNQ1), causing LQT1, or in the voltage-gated K⁺ channel, subfamily H, member 2 (KCNH2), causing LQT2, and a dominant negative mechanism has been suggested in many of these mutations (4–7). In these patients, ventricular arrhythmias and sudden death are often triggered by emotional stress and/or physical exercise (e.g., swimming). In female LQTS patients, an increase in the incidence of arrhythmias has been reported with menses and during the first months postpartum (2, 8). Besides symptomatic treatment with beta blockers, which reduces the frequency of arrhythmias, and the implantation of automated defibrillators to terminate the fatal arrhythmias, no causal therapy is currently available (2, 9).

In experimental studies, transgenic mouse models of the LQTS (10–14) have shown at the cellular level that dominant negative mutations in genes coding for repolarizing K⁺ channels lead to a prolongation of the duration of the action potential, with a variable incidence of early afterdepolarizations, regional electrical heterogeneities, and reentrant arrhythmias (10, 15–18). Moreover, the phenotype of mice expressing pore mutant dominant negative transgenes in the heart was similar to that of the knockout of the targeted channel (10, 15–18). Of note, the downregulation of the K⁺ currents was associated with electrical remodeling characterized by upregulation of other K⁺ currents (14, 16, 17). In the mouse, however, the main repolarizing currents are carried by different channels compared with humans, which limits the applicability of studies in the mouse to human patients (19). In contrast, the rabbit heart bears marked similarity to the human heart (20), and the same K⁺ currents — namely the rapidly activating component (I_Kr) and the slowly activating component (I_Ks) — determine the repolarization of the action potential. Thus, the rabbit represents a more physiologically relevant model system to study diseases affecting these channels.

Here we show that expression of pore mutants of the human genes KCNQ1 (KvLQT1-Y315S) and KCNH2 (HERG-G628S) in the rabbit heart under the control of the rabbit β-myosin heavy chain (β-MyHC) promoter produced LQT1 and LQT2 phenotypes in the rabbits, respectively. Characterization of the rabbits revealed that prolongations of the QT interval and action potential duration at 90% repolarization (APD₉₀) were due to the elimination of I_Ks and I_Kr currents, respectively. Only the LQT2 phenotype was associated with sudden cardiac death due to polymorphic VTs. Optical mapping studies revealed that spatial dispersion of repolarization underlies the reentrant arrhythmias in LQT2 rabbits. Of note, we also observed downregulation of the remaining complementary I_Kr and I_Ks currents, respectively, in cardiomyocytes derived from LQT1 and LQT2 rabbits as compared with those from littermate control (LMC) rabbits. Thus, both rabbit models were associated with downregulation rather than with compensatory upregulation of the reciprocal repolarizing currents. These observations suggest that mutated KvLQT1 and HERG polypeptides might interact in vivo with the corresponding WT α subunits of RERG and rabbit KvLQT1, respectively.

Creation of transgenic LQT1 and LQT2 rabbits. To generate transgenic rabbits, the FLAG-tagged KvLQT1-Y315S (21) and HERG-G628S (5) cDNAs (kind gifts from J. Barhanin, Institut de Pharmacologie Moléculaire et Cellulaire, Valbonne, France, and M.C. Sanguinetti, University of Utah, Salt Lake City, Utah, USA) were first cloned under the control of the CMV promoter. The mutant cDNAs were then cloned under the control of rabbit β-MyHC promoters (22) (a kind gift of J. Robbins, Children’s Hospital Research Foundation, Cincinnati, Ohio, USA) linked to the human growth hormone polyadenylation signal at the carboxyl terminus (Figure 1A). The β-MyHC-FLAG-KvLQT1-Y315S (LQT1) or β-MyHC-FLAG-HERG-G628S (LQT2) insert was isolated and used to inject rabbit embryos. The offspring were screened by PCR (Figure 1B). Three male founders were identified: 2 were positive for the HERG-G628S (LQT2) transgene; 1 died at the age of 1 week, and 1 was positive for the KvLQT1-Y315S transgene (LQT1). Using Southern blots, the transgene copy number per diploid genome was estimated by comparing the intensity of the hybridization band with that of the control. LQT1 (KvLQT1-Y315S) rabbits ranged from 1–2 copies and LQT2 (HERG-G628S) ranged from 2 to 3 copies (Figure 1C). Expression studies in CHO cells demonstrated that both mutations coded for polypeptides that migrated similarly to the WT polypeptides (Figure 1D). Immunoblot analyses of crude membrane extracts derived from the hearts of these animals revealed that the corresponding transgene polypeptides were detectable with anti-FLAG antibody (Figure 1D). Immunoprecipitations of these polypeptides with anti-FLAG antibody followed by immunoblots with the specific antibodies (Figure 1D) revealed similar-sized reactive polypeptides. Of note, most of the KvLQT1-Y315S polypeptides aggregated and formed a large complex that migrated very slowly (Figure 1D). Additional control experiments revealed that IgG did not IP these peptides (Figure 1E). We have investigated the level of expression of the endogenous channel polypeptides in crude membrane heart extracts. The results showed that the antibodies cross-reacted with the native peptides and that the steady-state level of KvLQT1 polypeptides expressed in the free wall and septum of LQT2 rabbits and RERG polypeptides in LQT1 rabbits was similar to those of LMC rabbits. Thus, the overexpression of the transgene polypeptides did not suppress the expression level of the native polypeptides (Figure 1E).

Figure 1

Transgenic constructs and expression studies. (A) Schematic drawings of the mutations (top) in KvLQT1 (left) and HERG (right) polypeptides and transgenic constructs (bottom). (B) PCR of genomic DNA of the founders: + and – denote positive control (constructs) and negative control. Numbers correspond to animals. Rabbits 2 (for KvLQT1) and 33 (HERG) tested positive. (C) Southern blot analyses of genomic DNA. Numbers in brackets indicate rabbit numbers. The sizes of the inserts incorporated into the rabbit genome (right panel) were identical to those of the plasmids (left panel). (D) Left panel shows Western blots of membrane lysates of CHO cells transfected with empty vector (control), KvLQT1, KvLQT1-Y315S, HERG, and HERG-G628S, and heart lysates from LMC, LQT1, and LQT2. Right panels show IP with anti-FLAG antibody of LMC, LQT1 (KvLQT1-Y315S), and LQT2 (HERG-G628S) crude heart lysates probed with anti-HERG and anti-KvLQT1 antibodies, respectively. The apparent molecular weight is 75 kDa for KvLQT1 and 135 kDa and 155 kDa for HERG. (E) Crude heart membranes were prepared from sections (LV and septum [S]) of LMC, LQT1, and LQT2. 200 μg samples were immunoblotted with polyclonal HERG antibody (top) and monoclonal KvLQT1 antibody (bottom). Crude membranes from CHO cells transfected with HERG and KvLQT1 cDNA served as positive controls. Anti-HERG antibody reacted with a 155-kDa polypeptide representing the endogenous RERG expression in LMC and LQT1 hearts. The anti-KVLQT1 mAb detected a 75-kDa band representing the endogenous rabbit KvLQT1 channel polypeptides in LMC and LQT2 hearts. Upper right panel shows crude membranes from LQT2 or LQT1 rabbit hearts, which were reacted with either anti-FLAG mAb (right lane) or mouse IgG (left lane). Anti-HERG antibody detected a 155-kDa polypeptide, while IgG failed to precipitate this peptide. Lower right panel shows anti-KvLQT1 mAb, which reacted with a 75-kDa band while IgG failed to precipitate it.

Phenotypic characterization. At the age of 4–5 months, the positive founders and age- and sex-matched littermates (2 male positive founders and 6 male controls) were characterized phenotypically with a 12-lead surface ECG. The transgenes caused a marked QT prolongation with an abnormal T wave morphology (Figure 2A) and an altered QT/RR relationship (Figure 2B). These founders were then crossbred with WT New Zealand white (NZW) (WT) rabbits, and the vertical transmission of the transgenes was confirmed by PCR of their genomic DNA extracted from ear biopsies of their offspring.

Figure 2

Phenotypic characterization in sedated animals. (A) Sample ECGs (lead II) of the founders and an LMC rabbit with midazolam sedation (2 mg/kg, i.m.). Note the markedly prolonged QT interval and the lack of an isoelectric T-P line in both transgenic rabbits. (B) QT/RR relationship (mean of all 12 leads per animal) under midazolam sedation in the 2 founder rabbits and 6 LMCs. Dotted lines denote the 95% confidence intervals of the linear regression derived from the LMCs. (C) RR and QT intervals (mean of 12-lead surface ECG per animal) in ketamine/xylazine-sedated male rabbits (n = 13 LMC, 9 LQT1, and 10 LQT2 animals). (D) Effect of isoproterenol on QT index in isoflurane-sedated LMC, LQT1, and LQT2 rabbits. Solid bars denote baseline (base); hatched bars denote isoproterenol (Iso). **P < 0.01.

Using the offspring of the founders (F2–F3 generation), we recorded 12-lead surface ECGs in 13 LMC, 9 LQT1, and 10 LQT2 animals sedated with ketamine/xylazine. Baseline data of the major ECG parameters are presented in Table 1 and Figure 2C. In transgenic rabbits, the QT_peak and total QT interval were significantly prolonged, with no significant difference in either T_peak–T_end duration or the RR intervals or QRS and PR duration.

Table 1

Baseline values of ECG parameters measured in sedated LMC, LQT1, and LQT2 rabbits

Awake, free-moving animals (11 LMC, 8 LQT1, and 6 LQT2 males) were monitored telemetrically after at least 3 weeks of recovery from the surgical implantation of the transmitter (Figure 3A). Figure 3B displays the QT/RR relationship of all measured QT/RR pairs in these animals. The mean RR interval was 299.3 ± 26.4 ms in LMC, 302.3 ± 32.4 ms in LQT1, and 280.7 ± 13.2 ms in LQT2 animals (P = NS). The mean QT intervals were significantly different in LMC and transgenic animals: LMC, 152.7 ± 11.2 ms; LQT1, 176.6 ± 17.3 ms; and LQT2, 175.9 ± 15.78 ms (ANOVA: P < 0.005; LMC vs. LQT1, P < 0.01; LMC vs. LQT2, P < 0.05) (Figure 3C).

Figure 3

Phenotypic characterization in awake, free-moving animals. (A) Sample telemetric ECGs (lead II) of awake, free-moving male rabbits. (B) QT/RR relationship in awake, free-moving rabbits recorded approximately every 20 minutes for 24 hours in 11 LMC, 8 LQT1, and 6 LQT2 male rabbits. Lines indicate linear regression derived from the mean of all individual regression lines per genotype. (C) RR and QT intervals (mean of 72 telemetric measurements/animal during 24-hour monitoring) of awake, free-moving male rabbits. *P < 0.05.

Linear regression of the QT/RR relation was performed for each animal, and the resulting regressions were averaged per genotype, resulting in the equation for QT intervals of LMCs: QT_LMC = 86 + 0.22*RR. The slope of similar regressions for LQT animals was steeper in LQT1 (QT_LQT1 = 80 + 0.32*RR; P < 0.05 vs. LMCs) and more marked in LQT2 animals (QT_LQT2 = 35 + 0.50*RR; P < 0.001 vs. LMCs), reflecting a more pronounced QT prolongation of the LQT2 rabbits at slow heart rates (Figure 3B).

To determine the genotype-specific drug response and the arrhythmogenic effect of sympathomimetic stimulation, we treated LQT rabbits and LMCs with isoproterenol. Adult male LQT1 (n = 11), LQT2 (n = 9), and LMC rabbits (n = 6) were exposed to the β-adrenoceptor agonist isoproterenol (0.1–0.2 μg/min continuously i.v. under general anesthesia with isoflurane. Consistent with observations during clinical electrophysiological studies in human LQTS patients that did not show spontaneous polymorphic VT or inducibility for sustained ventricular arrhythmias with isoproterenol (23), none of the LQT1, LQT2, or LMC rabbits developed spontaneous polymorphic VT under β-adrenergic stimulation with isoproterenol. Interestingly, as seen in human LQT1 patients (24–26), the absolute QT interval duration did not shorten under isoproterenol stimulation in LQT1 (baseline: 222 ± 39 ms, vs. isoproterenol: 221 ± 34 ms) rabbits, although the heart rate increased (baseline: 338 ± 41 ms vs. isoproterenol: 276 ± 33 ms; P < 0.0001). This is in contrast with a catecholamine-induced shortening of the absolute QT-interval duration in LQT2 (baseline: 307 ± 61 ms, vs. isoproterenol: 242 ± 33 ms; P < 0.05; Table 2). To assess heart-rate–corrected isoproterenol-induced changes in the QT interval, we established genotype-specific heart-rate–correction formulas for the QT (QTexp) under isoflurane (I_Ks blocker): LMC: QTexp = 92 + 0.40 * RR; LQT1: QTexp = 74 + 0.42 * RR; LQT2: QTexp = 76 + 0.65 * RR. The QT index was calculated as follows: percentage of QT observed/QT expected. We observed a significant isoproterenol-induced prolongation of the heart-rate–corrected QT index in LQT1 rabbits. In contrast, in LQT2 rabbits, we observed a significant shortening of QT index (Figure 2D), demonstrating a differential pharmacogenomic effect of the drug in these LQT rabbits. Due to the length of the procedure, all rabbits were anesthetized with the I_Ks blocker isoflurane (27). Consequently, we observed a blunted response to isoproterenol in LMC rabbits without the expected QT shortening.

Table 2

Effect of isoproterenol on sedated animals

Echocardiography was performed in 8–10 rabbits from each group. LQT2 animals were younger than LQT1 animals (20.6 ± 4.9 months vs. 27.2 ± 4.2 months; P < 0.05), but neither transgenic group was significantly different from LMC (23.5 ± 5.0 months). No difference in LV ejection fraction, LV wall thickness, LV diameter, and intraventricular septum thickness was found in transgenic animals compared with LMC rabbits. However, the posterior wall in LQT2 animals was slightly thinner compared with LQT1 animals but not LMC animals, likely due to their younger age (Table 3). Minor valve insufficiencies of the tricuspid, pulmonary, and mitral valves were seen in 3–5 animals per group; no other abnormalities were detected.

Table 3

Echocardiographic data

Staining of LMC and transgenic hearts with H&E revealed normal myocardium in all 3 groups without myofibrillary disarray, fibrosis, or other morphologic changes in the transgenic rabbit hearts (Figure 4).

Figure 4

Histology of transgenic and LMC animals. Sample H&E-stained histological sections of the middle of the LV free wall. Left panels, LMC; middle panels, LQT1 rabbit; right panels, LQT2 rabbit. Original magnification, ×10 (top panels); ×20 (middle panels); ×40 (bottom panels). Scale bars: 10 μm. Boxes in the top row indicate areas of further magnification shown in the middle and bottom rows.

Incidence of arrhythmias and sudden death in transgenic rabbits. We used the Kaplan-Meier method to estimate survival free of sudden death (defined as any death without an obvious noncardiac cause) of LMC and transgenic rabbits. Rabbits in this cohort did not undergo instrumentation or surgery, were handled only for routine animal care, and were mated to generate offspring. The survival of 26 LQT1 and 34 LQT2 rabbits was compared with that of 51 LMC rabbits, with an average follow-up of 288 ± 126.8 days for LMC animals, 340 ± 104.6 days for LQT1 animals, and 266 ± 118 days for LQT2 animals. One sudden death occurred in LMC controls vs. none in the LQT1 rabbits (P = NS) vs. 14 deaths in the LQT2 group (P < 0.0001) (Figure 5A). No sex differences were seen in the mortality rates of the groups. Sudden death was precipitated in some animals by stress (mating and handling) or by lactation in females.

Figure 5

Sudden death and spontaneous arrhythmias. (A) Kaplan-Meier survival curve of 51 LMC, 26 LQT1, and 34 LQT2 rabbits. (B) Sample ventricular extrasystoles (bigeminy) with R-on-T phenomenon in the LQT2 founder rabbit during ECG (leads I, II, III) recording with midazolam sedation. Asterisks denote ventricular extrasystoles. This rabbit died suddenly approximately 2 weeks after this recording. (C) Telemetric recording of a polymorphic VT leading to the death of a LQT2 rabbit. The torsade is initiated by repetitive short-long-short sequences due to ventricular extrasystoles. (D) Spontaneous polymorphic tachycardia in a ketamine/xylazine-anesthetized rabbit. Top panel shows ECG lead II. Bottom panel shows simultaneous arterial pressure recorded from the ear artery.

The reasons for sudden death were further defined by implanting ECG transmitters in an additional 10 LMC, 9 LQT1, and 7 LQT2 male rabbits and continuously monitoring the animals for arrhythmias: 2 of 7 LQT2 animals died of spontaneous polymorphic VT (Figure 5C) versus none of the LMC or LQT1 animals during 4–6 weeks of monitoring. Additionally, spontaneous polymorphic VT was recorded in one LQT2 rabbit during anesthesia, leading to a dramatic drop in blood pressure (Figure 5D) with prompt recovery after a spontaneous return to a normal sinus rhythm.

When analyzing 16 episodes of nonsustained and sustained polymorphic VT (in 3 male LQT2 rabbits and 4 female LQT2 rabbits), we found no significant differences in the heart rate directly prior to the onset of polymorphic VT episodes compared with baseline (before polymorphic VT: 197.5 ± 58/min, at baseline 218 ± 39/min), although in 3 of 16 polymorphic VT episodes, the preceding heart rate was significantly slower (84/min, 138/min, 126/min). In 1 episode, the polymorphic VT was preceded by an episode of AV 2:1 block. All 16 episodes started with an R-on-T phenomenon; in 11 of 16, we observed short-long-short episodes initiating the polymorphic VT as has been described in human patients with acquired and congenital LQTS (28). All episodes of polymorphic VT were preceded by longer periods with multiple single premature ventricular contractions or short nonsustained VTs, and 5 episodes of polymorphic VT succeeded longer periods of ventricular bigeminy. These findings suggest that early afterdepolarizations (EADs) play a major role in the induction of arrhythmias.

Cellular electrophysiology of ventricular myocytes. Action potentials and various K⁺ currents in myocytes isolated from the LV apex were investigated. The action potential was significantly longer in both LQT1 and LQT2 rabbit myocytes compared with the controls (Figure 6A). The averaged APD₉₀ was 354.1 ± 30.1 ms (n = 22) in LMC, 499.9 ± 45.7 ms in LQT1 (n = 14), and 533.1 ± 54.2 ms in LQT2 (n = 14) animals (both P < 0.05 vs. LMC). To provide functional correlates with molecular targets in the 2 animal models, we first examined I_Kr and I_Ks currents in these myocytes. I_Ks currents were selectively abolished in cardiomyocytes derived from LQT1 hearts, whereas I_Kr currents were selectively abolished in LQT2 hearts (Figure 6B). Further characterization of I_Kr and I_Ks (Figure 4B) revealed that the current densities of I_Kr in LQT1 myocytes and of I_Ks in LQT2 myocytes were also decreased as compared with the LMC currents. At +10 mV, for example, tail I_Ks in LQT2 was 0.31 ± 0.03 pA/pF as compared with 0.41 ± 0.04 pA/pF in LMCs (P < 0.05); tail I_Kr in LQT1 animals was 0.28 ± 0.01 pA/pF as compared with 0.42 ± 0.02 pA/pF in LMCs (P < 0.05). In contrast, the current densities of the other 2 prominent K⁺ currents, the transient outward (I_to) and the inward rectifier (I_K1), did not differ among the LMC, LQT1, and LQT2 groups (Figure 6C).

Figure 6

Cellular electrophysiology. (A) APD of rabbit ventricular myocytes. Left panel shows typical action potential recordings (0.1 Hz) from LMC, LQT1, and LQT2 rabbits. Right shows averaged APD (APD₉₀, mean ± SEM) of LMC (354.05 ± 30.07 ms, n = 22), LQT1 (499.88 ± 45.71 ms, n = 14), and LQT2 rabbits (533.14 ± 54.22 ms, n = 14); *P < 0.05. (B) Isolation and quantification of I_Kr and I_Ks. Left panel shows original recordings of control, LQT1, and LQT2 rabbits as indicated. After a recording without drugs (a), the cells were perfused with 5 μM E-4031 (b) and I_Kr was defined as the E4031-sensitive current (d). Secondary to E4031 application, the cells were further perfused with 30 μM chromanol 293B (c) and I_Ks was defined as chromanol-sensitive current (e). Right panels shows quantification of I_Kr and I_Ks . Current amplitudes measured at the end of repolarization (I_Ks or I_Kr) and the peak of the tail (I_Ks tail or I_Kr tail) were plotted against membrane voltages. All currents were normalized to cell capacitance. Open circles depict control myocytes (n = 20 from 6 rabbits), filled circles depict LQT1 myocytes (n = 17 from 5 rabbits), and filled triangles depict LQT2 myocytes (n = 12 from 3 rabbits). The downregulation of I_Kr (in LQT1) or I_Ks (in LQT2) was significant compared with controls (P < 0.05 by 2-way ANOVA). (C) I_to current and I_K1 currents. Standard current-voltage relationship (IV curve) of I_to (left panel) or quasi-IV curve of I_K1 (right panel) revealed no significant differences in peak I_to currents (n = 6–8) or I_K1 currents (right, n = 10–12) among LMC, LQT1, and LQT2 rabbits.

Optical mapping of intact hearts. We next investigated the mechanism of arrhythmias in LQT rabbits using optical mapping of Langendorff-perfused hearts maintained at 37°C. We found that the ΔAPD (defined as the absolute difference between maximum and minimum APD across the anterior epicardial surface) in LMCs was 23 ± 9 and in LQT1 rabbits was 18.0 ± 6 ms. In contrast, ΔAPD in LQT2 rabbits was 41 ± 17 ms (P < 0.05). The increase of APD dispersion in LQT2 rabbits suggests that nonuniformity of repolarization correlates with arrhythmogenesis. The APDs at different stimulation cycle lengths and the dispersion of repolarization (ΔAPD) are summarized in Figure 7, D and E.

Figure 7

APD and APD dispersion in transgenic rabbits. (A–C) Typical raw data of action potential traces are shown in the left columns. APDs were measured as described in Methods and mapped in each group. Isochronal lines were drawn every 5 ms with lighter color representing shorter APD. (D) Mean APD in each group at basic cycle length of 350 and 300 ms. LMC (n = 6), LQT1 (n = 3), and LQT2 rabbits (n = 4). LQT2 and LQT1 rabbits show statistically significant differences in APD compared with LMC; *P < 0.05. (E) APD dispersion. APD dispersion was calculated by differences between APD_max and APD_min. LQT2 rabbits show greater dispersion compared with LMC. ^†P < 0.05.

To test for inducibility of arrhythmias, hearts were stimulated from LV epicardium at a cycle length (CL) equal to 350 ms initially, and cycle lengths were reduced in 10-ms steps until either 2:1 block or VT or ventricular fibrillation (VF) occurred. Most LMCs (n = 6 out of 7) and all LQT1 animals (n = 4 out of 4) were not inducible by this method. However, VF was triggered by programmed stimulation in most LQT2 rabbits (n = 3 out of 4; P < 0.05). One LQT2 rabbit showed very long APDs and could not be stimulated at a basic cycle length shorter than 350 ms. Pacing close to refractoriness induced VT/VF in LQT2 rabbits; however, induction of arrhythmia was not consistently seen at any specific paced heart rate. Examination of VF initiation showed that a premature beat during burst stimulation encountered refractoriness across an area of enhanced dispersion of APD, causing unidirectional block and consequently formation of reentry (see Figure 8 for details). In contrast, LMC and LQT1 animals showed smooth conduction without rotation or conduction blocks (Figure 8, D and E). Thus, APD dispersion is a major mechanism underlying LQT2-related arrhythmias (Figures 7 and 8).

Figure 8

Activation maps of VF initiation in an LQT2 heart. (A) APD map. The red dotted line marks the interventricular septum. The APD map in LQT2 shows increased dispersion, mostly in the mid LV region. (B) Trace of action potentials during initial period of VF. (C) Series of activation maps marked in panel B. Isochronal lines are drawn every 2-ms interval, and lighter color represents earlier activation. Panel no. 1 shows activation pattern of the paced beat. The following beats (nos. 2 and 3) encounter conduction block where the APD is longer (see panel A). Therefore, the LV activated via the apical free wall where APD is shorter and tissue is recovered from the previous beat. The next activation (no. 4) appears from the RV and propagates toward the mid LV. A similar wave front (no. 5) encounters conduction block (red straight line) in the region where APD is longer and forms a rotation (no. 6), initiating VF (no. 7 and no. 8). This initiation highlights APD dispersion as an important mechanism in arrhythmia formation in this LQT2 model. (D and E) Traces and activation maps from LMC and LQT1 hearts. Traces and maps were taken at the shortest cycle length that did not cause 2:1 block. Both LMC and LQT1 showed uninterrupted conduction from the stimulation sites to the rest of the heart without rotation or conduction blocks.

Transgenic and knockout mouse models have significantly improved our understanding of cardiac excitation (19), but because of significant differences in the length of the APD and the repolarization phase of the action potential of cardiomyocytes derived from mice and humans, none of the previously available mouse models fully emulates human LQTS (19). Unlike in the mouse heart, the rabbit action potential is remarkably similar to the human heart in that the same K⁺ outward currents determine the dynamics of the repolarization phase of the action potential. Moreover, the fibrillatory rotors resemble those of the human heart (29). Here, we present the first transgenic rabbit models for the 2 most common human LQTS, types 1 and 2. These rabbits were created by expressing dominant negative K⁺ channel α subunits cloned from affected humans under the control of a rabbit β-MyHC (22, 30). We show that these mutated polypeptides consistently eliminated the currents encoded by their respective WT proteins, producing a dominant negative effect. Consequently, these constructs caused a significant prolongation of the APD by selectively eliminating I_Ks (in LQT1 animals) and I_Kr (in LQT2 animals). Furthermore, as opposed to the data from studies of various strains of genetically modified mice (14, 16), our data show no compensatory upregulation of other repolarizing currents in these rabbits.

Clinically, the elimination of I_Ks and I_Kr has differential effects: LQT1 rabbits show QT prolongation but neither spontaneous arrhythmias nor an increase in sudden death. LQT2 animals have a more pronounced QT prolongation at slow heart rates and show both spontaneous arrhythmias and a significant increase in sudden death after the age of approximately 200 days, consistent with the end of puberty in these animals. Additionally LQT2 females exhibit a high incidence of sudden death during lactation, an observation similar to the increased incidence of arrhythmias in women with the LQT2 syndrome postpartum (8).

The QT/RR relationship was highly individual among our transgenic and LMC rabbits during telemetric monitoring — similar to that in healthy human subjects (31). Therefore, we calculated individual regression lines and averaged these per group. We used linear regressions because nonlinear regressions did not yield markedly better correlations, an observation that has similarities to studies in humans (31). We are aware, however, that neither linear nor nonlinear equations, such as Bazett formula, correct the QT interval properly to all ranges of heart rates, an issue that has been widely debated in the literature. Analysis of the QT/RR slope in the 2 transgenic lines and the controls revealed a pattern similar to that of the QT/RR slope as observed in affected humans with a steeper QT/RR relationship in LQT2 compared with LQT1 and LMC. This bears striking similarity to the results of studies in humans using Holter monitoring and exercise tests (32–34) in different populations.

Adult LQT2 patients have a higher risk of sudden death compared with LQT1 patients, and the risk is increased in those with a longer QT and in women (9, 35). LQT2 patients with a pore mutation have the highest risk of arrhythmic events (36). Here, we show that LQT2 animals with a mutation in the pore region (HERG-G628S) have a high incidence of spontaneous arrhythmias leading to sudden cardiac death (SCD). In congenital or acquired human LQTS, increased spatial dispersion of repolarization plays a major role in propagating and maintaining malignant arrhythmias (37, 38), and EADs due to prolonged APD have been assumed to trigger these arrhythmias. Experimental models in transgenic mice have shown increased apex-to-base dispersion of repolarization underlying mouse arrhythmias (17, 39), and pharmacological LQT models in rabbits and dogs have reproduced this finding as transmural dispersion of repolarization in wedge preparations (40–43). Studies in experimental models (43, 44), in mathematical simulations (45, 46), and in affected humans with LQTS (47) have demonstrated the occurrence of phase 2 and 3 EADs in vitro, in silico, and in vivo, hinting toward M cells and/or Purkinje fibers as the origin of EADs (44, 46).

Here, we demonstrate an increased spatial dispersion of repolarization across the anterior epicardial surface as the electrical basis for spontaneous malignant arrhythmias in LQT2 animals. In contrast, LQT1 animals did not have increased dispersion, despite a similar degree of APD prolongation. Thus, lack of spatial dispersion of repolarization correlates with absence of spontaneous arrhythmias or sudden cardiac death. Moreover, further studies into the mechanism of VF induction revealed that programmed stimulation induced VF in LQT2 rabbits but not in LQT1 rabbits. Importantly, imaging of the initiation of the arrhythmias (Figure 8) demonstrated that when a premature beat encountered an area of prolonged APD, conduction block occurred, resulting in wave breakup and formation of reentry and VF. These events support the notion that spatial dispersion of repolarization plays a key role in the pathogenesis of VF in LQT2 rabbits.

Additionally, all documented episodes of SCD in these animals were induced by R-on-T extrasystoles, and most were preceded by short-long-short sequences, a common mechanism of initiating polymorphic VT in LQT patients (48). It has been suggested that the initiating extrasystoles in short-long-short sequences represent phase 3 EADs (48). Our optical mapping data do not address the mechanisms underlying the trigger of SCDs. EADs were rarely observed under current experimental conditions of a retrogradely perfused isolated heart preparation. Several groups used different conditions in drug-induced LQT rabbit models to trigger EADs, including bradycardia by AV node ablation, low K⁺, low Mg²⁺, and adrenergic stimulation by α and/or β agonists (49–51). Further studies are required to verify that these conditions can induce EADs in the transgenic model of LQT and to investigate mechanisms of EADs and their preferential locations, such as Purkinje or M cells.

In pharmacological rabbit models of LQTS2, abnormal calcium release from the sarcoplasmic reticulum was shown to cause EADs and trigger polymorphic VT in accordance with the hypothesis that Ca overload may underlie these arrhythmias (44). The involvement of abnormal Ca²⁺ handling in transgenic LQT2 rabbits is the subject of ongoing investigations.

Rabbit I_Kr and I_Ks are regionally differentially expressed, causing regional APD differences after pharmacological block (52). Recently, rabbits with AV-block and tachycardic or bradycardic pacing were found to have QT prolongation due to differential downregulation of repolarizing K⁺ currents (53), which leads to differential susceptibility for ventricular arrhythmias: whereas the combined reduction of I_Ks and I_Kr was proarrhythmic, the downregulation of I_Ks alone was not. Additional pharmacological block of I_Kr using dofetilide caused polymorphic VT and death in these animals, suggesting a crucial role of I_Kr in the genesis of arrhythmias in these rabbits. This supports our finding that the transgenic elimination of I_Ks alone in LQT1 does not result in spontaneous arrhythmias, whereas the elimination of I_Kr leads to spontaneous arrhythmias and sudden death in LQT2 rabbits.

Cellular electrophysiology studies also showed that the pore mutants specifically eliminated the targeted currents (I_Ks in LQT1 cells and I_Kr in LQT2 rabbits). In addition, we observed a remodeling effect associated with downregulation of the second major repolarizing currents. Thus, I_Kr is downregulated in LQT1 rabbits, whereas I_Ks is reduced in LQT2 rabbits. These effects were channel specific, as the expression of either I_to or I_K1 was unchanged. Pull-down experiments in our laboratory confirmed weak interactions between KCNQ1 and KCNH2 polypeptides and also confirmed that a small fraction of these peptides could be coprecipitated in transient cotransfection experiments in CHO cells, similar to previously published studies (54). The lack of change in the steady state level of these polypeptides suggests that abnormal processing or trafficking to the membrane underlies the reduction in the density of these currents. Taken together, these observations support the possibility that KCNH2 and KvLQT1 polypeptides interact directly (or indirectly) in vivo and that a mutation of either of these channels could have a dominant negative effect on the other’s currents. These phenomena can explain the severe clinical phenotype of the pore mutants in some LQTS patients (36, 55). We propose that genetic models such as LQT2 rabbits may help identify mechanisms of sudden cardiac death, while LQT1 rabbits may serve as a sensitive assay system for detecting effects of drugs on HERG in vivo.

Limitations. As with every transgenic model, results obtained in 1 species do not necessarily transfer into other species, even though these models resemble the human phenotype of LQTS1 and LQTS2 more than any of the previously available models. Additionally, we cannot exclude a nonspecific effect of the presence of the transgene per se, as due to financial constraints, we did not create transgenic control rabbits with an inert gene.

The authors thank J. Barhanin and M.C. Sanguinetti for the KvLQT1-Y315S and HERG-G628S clones, respectively, and J. Robbins for the rabbit β-MyHC promoter cDNA. The authors wish to thank Max Lang for his support of the project. G. Koren is a recipient of NIH grant RO1 HL046005-14; M. Brunner and K. Odening were supported in part by grants from the German Cardiac Society (Max Schaldach and St. Jude Medical Stipendium). The authors thank Jun Zhou for his valuable comments in reviewing the manuscript.

Address correspondence to: Gideon Koren, Cardiovascular Research Center, Cardiology Division, Rhode Island Hospital, Coro West 51031 Hoppin Street, Providence, Rhode Island 02903, USA. Phone: (401) 444-0392; Fax: (401) 444-4061; E-mail: Gideon_Koren@Brown.edu.

Nonstandard abbreviations used: APD, action potential duration; EAD, early afterdepolarization; I_K1, inward rectifier current; I_Kr, rapidly activating delayed rectifier K⁺ current; I_Ks, slowly activating delayed rectifier K⁺ current; I_to, transient outward current; LMC, littermate control; LQT1, LQTS, type 1; LQTS, long QT syndrome; β-MyHC, β-myosin heavy chain; NZW, New Zealand white; QTexp, genotype-specific expected QT interval at a given heart rate; SCD, sudden cardiac death; VF, ventricular fibrillation; VT, ventricular tachycardia.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J. Clin. Invest.118:2246–2259 (2008). doi:10.1172/JCI33578

Michael Brunner, Xuwen Peng, and Gong Xin Liu contributed equally to this work.

1. Moss, A.J. 2003. Long QT syndrome. JAMA. 289:2041-2044.
   View this article via: CrossRef PubMed Google Scholar
2. Priori, S.G., Bloise, R., Crotti, L. 2001. The long QT syndrome. Europace. 3:16-27.
   View this article via: CrossRef PubMed Google Scholar
3. Roden, D.M. 2008. Long-QT syndrome. N. Engl. J. Med. 358:169-176.
   View this article via: CrossRef PubMed Google Scholar
4. Curran, M.E., et al. 1995. A molecular basis for cardiac arrhythmia: HERG mutations cause long QT syndrome. Cell. 80:795-803.
   View this article via: CrossRef PubMed Google Scholar
5. Sanguinetti, M.C., Curran, M.E., Spector, P.S., Keating, M.T. 1996. Spectrum of HERG K+-channel dysfunction in an inherited cardiac arrhythmia. Proc. Natl. Acad. Sci. U. S. A. 93:2208-2212.
   View this article via: CrossRef PubMed Google Scholar
6. Sanguinetti, M.C., et al. 1996. Coassembly of K (V)LQT1 and minK (IsK) proteins to form cardiac I (Ks) potassium channel. Nature. 384:80-83.
   View this article via: CrossRef PubMed Google Scholar
7. Sanguinetti, M.C., Jiang, C., Curran, M.E., Keating, M.T. 1995. A mechanistic link between an inherited and an acquired cardiac arrhythmia: HERG encodes the IKr potassium channel. Cell. 81:299-307.
   View this article via: CrossRef PubMed Google Scholar
8. Seth, R., et al. 2007. Long QT syndrome and pregnancy. J. Am. Coll. Cardiol. 49:1092-1098.
   View this article via: CrossRef PubMed Google Scholar
9. Sauer, A.J., et al. 2007. Long QT syndrome in adults. J. Am. Coll. Cardiol. 49:329-337.
   View this article via: CrossRef PubMed Google Scholar
10. Jeron, A., et al. 2000. Inducible polymorphic ventricular tachyarrhythmias in a transgenic mouse model with a long Q-T phenotype. Am. J. Physiol. Heart Circ. Physiol. 278:H1891-H1898.
    View this article via: PubMed Google Scholar
11. Guo, W., Li, H., London, B., Nerbonne, J.M. 2000. Functional consequences of elimination of i (to,f) and i (to,s): early afterdepolarizations, atrioventricular block, and ventricular arrhythmias in mice lacking Kv1.4 and expressing a dominant-negative Kv4 alpha subunit. Circ. Res. 87:73-79.
    View this article via: PubMed Google Scholar
12. Demolombe, S., et al. 2001. Transgenic mice overexpressing human KvLQT1 dominant-negative isoform. Part I: Phenotypic characterisation. Cardiovasc. Res. 50:314-327.
    View this article via: CrossRef PubMed Google Scholar
13. Brunner, M., et al. 2001. Characterization of mice with a combined suppression of I (to) and I (K,slow). Am. J. Physiol. Heart Circ. Physiol. 281:H1201-H1209.
    View this article via: PubMed Google Scholar
14. Guo, W., et al. 2005. Targeted deletion of Kv4.2 eliminates I (to,f) and results in electrical and molecular remodeling, with no evidence of ventricular hypertrophy or myocardial dysfunction. Circ. Res. 97:1342-1350.
    View this article via: CrossRef PubMed Google Scholar
15. London, B., et al. 1998. Long QT and ventricular arrhythmias in transgenic mice expressing the N terminus and first transmembrane segment of a voltage-gated potassium channel. Proc. Natl. Acad. Sci. U. S. A. 95:2926-2931.
    View this article via: CrossRef PubMed Google Scholar
16. Zhou, J., et al. 2003. Regional upregulation of Kv2.1-encoded current, IK,slow2, in Kv1DN mice is abolished by crossbreeding with Kv2DN mice. Am. J. Physiol. Heart Circ. Physiol. 284:H491-H500.
    View this article via: PubMed Google Scholar
17. Baker, L.C., London, B., Choi, B.R., Koren, G., Salama, G. 2000. Enhanced dispersion of repolarization and refractoriness in transgenic mouse hearts promotes reentrant ventricular tachycardia. Circ. Res. 86:396-407.
    View this article via: PubMed Google Scholar
18. Nerbonne, J.M., Nichols, C.G., Schwarz, T.L., Escande, D. 2001. Genetic manipulation of cardiac K (+) channel function in mice: what have we learned, and where do we go from here? Circ. Res. 89:944-956.
    View this article via: CrossRef PubMed Google Scholar
19. Nerbonne, J.M. 2004. Studying cardiac arrhythmias in the mouse — a reasonable model for probing mechanisms? Trends Cardiovasc. Med. 14:83-93.
    View this article via: CrossRef PubMed Google Scholar
20. Vermeulen, J.T., et al. 1994. Triggered activity and automaticity in ventricular trabeculae of failing human and rabbit hearts. Cardiovasc. Res. 28:1547-1554.
    View this article via: CrossRef PubMed Google Scholar
21. Chouabe, C., et al. 1997. Properties of KvLQT1 K+ channel mutations in Romano-Ward and Jervell and Lange-Nielsen inherited cardiac arrhythmias. EMBO J. 16:5472-5479.
    View this article via: CrossRef PubMed Google Scholar
22. James, J., et al. 2000. Transgenic modeling of a cardiac troponin I mutation linked to familial hypertrophic cardiomyopathy. Circ. Res. 87:805-811.
    View this article via: PubMed Google Scholar
23. Bhandari, A.K., et al. 1985. Electrophysiologic testing in patients with the long QT syndrome. Circulation. 71:63-71.
    View this article via: PubMed Google Scholar
24. Ackerman, M.J., et al. 2002. Epinephrine-induced QT interval prolongation: a gene-specific paradoxical response in congenital long QT syndrome. Mayo Clin. Proc. 77:413-421.
    View this article via: PubMed Google Scholar
25. Noda, T., et al. 2002. Gene-specific response of dynamic ventricular repolarization to sympathetic stimulation in LQT1, LQT2 and LQT3 forms of congenital long QT syndrome. Eur. Heart J. 23:975-983.
    View this article via: CrossRef PubMed Google Scholar
26. Shimizu, W., Kamakura, S. 2001. Catecholamines in children with congenital long QT syndrome and Brugada syndrome. J. Electrocardiol. 34(Suppl.):173-175.
27. Suzuki, A., Bosnjak, Z.J., Kwok, W.M. 2003. The effects of isoflurane on the cardiac slowly activating delayed-rectifier potassium channel in Guinea pig ventricular myocytes. Anesth. Analg. 96:1308-1315.
    View this article via: PubMed Google Scholar
28. Viskin, S., et al. 1996. Mode of onset of torsade de pointes in congenital long QT syndrome. J. Am. Coll. Cardiol. 28:1262-1268.
    View this article via: CrossRef PubMed Google Scholar
29. Panfilov, A.V. 2006. Is heart size a factor in ventricular fibrillation? Or how close are rabbit and human hearts? Heart Rhythm. 3:862-864.
    View this article via: CrossRef PubMed Google Scholar
30. James, J., et al. 2000. Genetic manipulation of the rabbit heart via transgenesis. Circulation. 101:1715-1721.
    View this article via: PubMed Google Scholar
31. Malik, M., et al. 2002. Relation between QT and RR intervals is highly individual among healthy subjects: implications for heart rate correction of the QT interval. Heart. 87:220-228.
    View this article via: CrossRef PubMed Google Scholar
32. Nemec, J., et al. 2004. Heart rate dependence of the QT interval duration: differences among congenital long QT syndrome subtypes. J. Cardiovasc. Electrophysiol. 15:550-556.
    View this article via: PubMed Google Scholar
33. Swan, H., et al. 1999. Sinus node function and ventricular repolarization during exercise stress test in long QT syndrome patients with KvLQT1 and HERG potassium channel defects. J. Am. Coll. Cardiol. 34:823-829.
    View this article via: CrossRef PubMed Google Scholar
34. Viitasalo, M., et al. 2002. Differentiation between LQT1 and LQT2 patients and unaffected subjects using 24-hour electrocardiographic recordings. Am. J. Cardiol. 89:679-685.
    View this article via: CrossRef PubMed Google Scholar
35. Hobbs, J.B., et al. 2006. Risk of aborted cardiac arrest or sudden cardiac death during adolescence in the long-QT syndrome. JAMA. 296:1249-1254.
    View this article via: CrossRef PubMed Google Scholar
36. Moss, A.J., et al. 2002. Increased risk of arrhythmic events in long-QT syndrome with mutations in the pore region of the human ether-a-go-go-related gene potassium channel. Circulation. 105:794-799.
    View this article via: CrossRef PubMed Google Scholar
37. Antzelevitch, C. 2007. Ionic, molecular, and cellular bases of QT-interval prolongation and torsade de pointes. Europace. 9(Suppl. 4):iv4-iv15.
38. Antzelevitch, C. 2007. Role of spatial dispersion of repolarization in inherited and acquired sudden cardiac death syndromes. Am. J. Physiol. Heart Circ. Physiol. 293:H2024-H2038.
    View this article via: CrossRef PubMed Google Scholar
39. London, B., et al. 2007. Dispersion of repolarization and refractoriness are determinants of arrhythmia phenotype in transgenic mice with long QT. J. Physiol. 578:115-129.
    View this article via: CrossRef PubMed Google Scholar
40. Antzelevitch, C., Shimizu, W. 2002. Cellular mechanisms underlying the long QT syndrome. Curr. Opin. Cardiol. 17:43-51.
    View this article via: CrossRef PubMed Google Scholar
41. Shimizu, W., Antzelevitch, C. 1999. Differential response of transmural dispersion of repolarization and torsade de pointes to beta-adrenergic agonists and antagonists in three models of the long QT syndrome. J. Electrocardiol. 32(Suppl.):150.
42. Shimizu, W., Antzelevitch, C. 2000. Differential effects of beta-adrenergic agonists and antagonists in LQT1, LQT2 and LQT3 models of the long QT syndrome. J. Am. Coll. Cardiol 35:778-786.
    View this article via: CrossRef PubMed Google Scholar
43. Yan, G.X., et al. 2001. Phase 2 early afterdepolarization as a trigger of polymorphic ventricular tachycardia in acquired long-QT syndrome: direct evidence from intracellular recordings in the intact left ventricular wall. Circulation. 103:2851-2856.
    View this article via: PubMed Google Scholar
44. Choi, B.R., Burton, F., Salama, G. 2002. Cytosolic Ca2+ triggers early afterdepolarizations and Torsade de Pointes in rabbit hearts with type 2 long QT syndrome. J. Physiol. 543:615-631.
    View this article via: CrossRef PubMed Google Scholar
45. Yinbin, J., Lin, Y., Hong, Z., Yecho, H., Dazong, J. 2005. Mechanism of reentry trigged by EADs in long-QT syndrome: a simulation study. Conf. Proc. IEEE Eng. Med. Biol. Soc. 7:7068-7071.
    View this article via: PubMed Google Scholar
46. Viswanathan, P.C., Rudy, Y. 2000. Cellular arrhythmogenic effects of congenital and acquired long-QT syndrome in the heterogeneous myocardium. Circulation. 101:1192-1198.
    View this article via: PubMed Google Scholar
47. Shimizu, W., et al. 1991. Early afterdepolarizations induced by isoproterenol in patients with congenital long QT syndrome. Circulation. 84:1915-1923.
    View this article via: PubMed Google Scholar
48. Noda, T., et al. 2004. Classification and mechanism of Torsade de Pointes initiation in patients with congenital long QT syndrome. Eur. Heart J. 25:2149-2154.
    View this article via: CrossRef PubMed Google Scholar
49. Carlsson, L., et al. 1993. Proarrhythmic effects of the class III agent almokalant: importance of infusion rate, QT dispersion, and early afterdepolarisations. Cardiovasc. Res. 27:2186-2193.
    View this article via: CrossRef PubMed Google Scholar
50. Carlsson, L., Abrahamsson, C., Drews, L., Duker, G. 1992. Antiarrhythmic effects of potassium channel openers in rhythm abnormalities related to delayed repolarization. Circulation. 85:1491-1500.
    View this article via: PubMed Google Scholar
51. Zabel, M., et al. 1997. Electrophysiologic features of torsades de pointes: insights from a new isolated rabbit heart model. J. Cardiovasc. Electrophysiol. 8:1148-1158.
    View this article via: CrossRef PubMed Google Scholar
52. Cheng, J., et al. 1999. Heterogeneous distribution of the two components of delayed rectifier K+ current: a potential mechanism of the proarrhythmic effects of methanesulfonanilideclass III agents. Cardiovasc. Res. 43:135-147.
    View this article via: CrossRef PubMed Google Scholar
53. Tsuji, Y., Zicha, S., Qi, X.Y., Kodama, I., Nattel, S. 2006. Potassium channel subunit remodeling in rabbits exposed to long-term bradycardia or tachycardia: discrete arrhythmogenic consequences related to differential delayed-rectifier changes. Circulation. 113:345-355.
    View this article via: CrossRef PubMed Google Scholar
54. Ehrlich, J.R., et al. 2004. KvLQT1 modulates the distribution and biophysical properties of HERG. A novel alpha-subunit interaction between delayed rectifier currents. J. Biol. Chem. 279:1233-1241.
    View this article via: CrossRef PubMed Google Scholar
55. Moss, A.J., et al. 2007. Clinical aspects of type-1 long-QT syndrome by location, coding type, and biophysical function of mutations involving the KCNQ1 gene. Circulation. 115:2481-2489.
    View this article via: CrossRef PubMed Google Scholar
56. Fedorov, V.V., et al. 2007. Application of blebbistatin as an excitation-contraction uncoupler for electrophysiologic study of rat and rabbit hearts. Heart Rhythm. 4:619-626.
    View this article via: CrossRef PubMed Google Scholar
57. Choi, B.R., Jang, W., Salama, G. 2007. Spatially discordant voltage alternans cause wavebreaks in ventricular fibrillation. Heart Rhythm. 4:1057-1068.
    View this article via: CrossRef PubMed Google Scholar
58. Efimov, I.R., Huang, D.T., Rendt, J.M., Salama, G. 1994. Optical mapping of repolarization and refractoriness from intact hearts. Circulation. 90:1469-1480.
    View this article via: PubMed Google Scholar