Differential conditions for early after-depolarizations and triggered activity in cardiomyocytes derived from transgenic LQT1 and LQT2 rabbits

Abstract

Non-technical summary

Long QT syndrome (LQTS) is a genetic disorder characterized by recurrent syncope and sudden cardiac death (SCD). Type 1 (LQT1) and Type 2 (LQT2) LQTS account for 90% of the genotyped mutations in patients with this disorder. These syndromes have been associated with different sympathetic modes for initiation of cardiac arrest. Using isolated cardiomyocytes and Langendorff-perfused hearts from transgenic rabbit models of LQT1 and LQT2, we have identified differential conditions and cellular mechanisms for the generation of early afterdepolarizations (EADs), abnormal depolarizations during the plateau and repolarization phase of action potentials and the hallmark of the arrhythmias in LQTS. These differences explain why different types of increased autonomic nervous system activity, i.e. sympathetic surge vs. high sympathetic tone, are associated with the initiation of polymorphic ventricular tachycardia in LQTS patients with different genetic background.

Abstract

Early after-depolarization (EAD), or abnormal depolarization during the plateau phase of action potentials, is a hallmark of long-QT syndrome (LQTS). More than 13 genes have been identified as responsible for LQTS, and elevated risks for EADs may depend on genotypes, such as exercise in LQT1 vs. sudden arousal in LQT2 patients. We investigated mechanisms underlying different high-risk conditions that trigger EADs using transgenic rabbit models of LQT1 and LQT2, which lack I_Ks and I_Kr (slow and fast components of delayed rectifying K⁺ current), respectively. Single-cell patch-clamp studies show that prolongation of action potential duration (APD) can be further enhanced by lowering extracellular potassium concentration ([K⁺]_o) from 5.4 to 3.6 mm. However, only LQT2 myocytes developed spontaneous EADs following perfusion with lower [K⁺]_o, while there was no EAD formation in littermate control (LMC) or LQT1 myocytes, although APDs were also prolonged in LMC myocytes and LQT1 myocytes. Isoprenaline (ISO) prolonged APDs and triggered EADs in LQT1 myocytes in the presence of lower [K⁺]_o. In contrast, continuous ISO perfusion diminished APD prolongation and reduced the incidence of EADs in LQT2 myocytes. These different effects of ISO on LQT1 and LQT2 were verified by optical mapping of the whole heart, suggesting that ISO-induced EADs are genotype specific. Further voltage-clamp studies revealed that ISO increases L-type calcium current (I_Ca) faster than I_Ks (time constant 9.2 s for I_Ca and 43.6 s for I_Ks), and computer simulation demonstrated a high-risk window of EADs in LQT2 during ISO perfusion owing to mismatch in the time courses of I_Ca and I_Ks, which may explain why a sympathetic surge rather than high sympathetic tone can be an effective trigger of EADs in LQT2 perfused hearts. In summary, EAD formation is genotype specific, such that EADs can be elicited in LQT2 myocytes simply by lowering [K⁺]_o, while LQT1 myocytes require sympathetic stimulation. Slower activation of I_Ks than of I_Ca by ISO may explain why different sympathetic modes, i.e. sympathetic surge vs. high sympathetic tone, are associated with polymorphic ventricular tachycardia in LQTS patients.

Introduction

Long-QT syndrome (LQTS) is a genetic disorder characterized by recurrent syncope and sudden cardiac death. Loss-of-function mutations in the two voltage-gated potassium channels, KvLQT1 and HERG, account for about 90% of the genotyped mutations in patients with this disorder. The LQT1 sydrome is caused by mutations in KvLQT1, the α-subunit that encodes the slow component of delayed rectifying K⁺ currents (I_Ks). The LQT2 syndrome is caused by mutations in the HERG channel, the α-subunit that encodes the fast component of delayed rectifying K⁺ currents (I_Kr; el-Sherif & Turitto, 1999; Chiang & Roden, 2000; Goldenberg et al. 2008).

Several studies have shown that cardiac events in LQT1 and LQT2 have distinctive patterns. Half of the lethal cardiac events associated with LQT2 occur during rest or sleep without arousal. Moreover, these patients are particularly sensitive to sympathetic surge, such as startling noises, including the telephone or alarm clock (Moss et al. 1999; Schwartz et al. 2001). By contrast, 68% of lethal cardiac events in LQT1 patients occur during periods of high sympathetic tone, such as exercise, while sustained exercise rarely induces these events in LQT2 patients. Such evidence suggests that different sympathetic activations (surge vs. tone) underlie sudden cardiac death in these syndromes.

In this study, we used our transgenic LQT1 and LQT2 rabbit models with overexpression of a dominant negative pore mutant of KCNQ1 (KvLQT1-Y315S) and HERG (HERG-G628S), which exhibit a phenotype of prolonged QT interval and sudden cardiac death (Brunner et al. 2008), to investigate the mechanisms underlying the induction of early after-depolarizations (EADs). Our results demonstrated that EADs were readily induced in LQT2 myocytes at lower physiological levels of extracellular potassium concentration ([K⁺]_o), while LQT1 myocytes showed no EAD formation in the same conditions. Formation of EADs in LQT1 myocytes required stimulation with isoprenaline (ISO). Optical mapping of Langendorff-perfused hearts showed differential conditions for generation of EADs: the LQT2 hearts were susceptible to sympathetic surge, while the LQT1 hearts were susceptible to high sympathetic tone. Computer simulation suggests that the susceptibility of LQT2 hearts to sympathetic surge is most likely due to a faster increase of L-type calcium current (I_Ca) by ISO compared with I_Ks, creating a narrow high-risk window of EAD formation when I_Ca is fully augmented, while I_Ks is still only partly augmented.

Methods

Preparation of rabbit ventricular myocytes

Ventricular myocytes were isolated from the septal portions of the hearts of transgenic New Zealand White rabbits (4–7 months old, weighing 2–4 kg) by standard enzymatic techniques (Brunner et al. 2008). The experimental procedure for isolation of myocytes conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH publication no. 85-23, revised 1996).

Patch-clamp recording and data analysis

Whole-cell recordings were obtained with an Axopatch-200B amplifier (Axon Instruments, Foster City, CA, USA) with standard patch-clamp techniques. As a standard bath solution Tyrode solution was used, containing (mm): 140 NaCl, 5.4 or 3.6 KCl, 0.33 NaH₂PO₄, 1 MgCl₂, 1 CaCl₂, 5 Hepes and 7.5 glucose (pH adjusted to 7.4 with NaOH). The pipette resistances for whole-cell recordings were 2–4 MΩ when filled with a solution comprising (mm): 120 KCl, 5 MgCl₂, 0.36 CaCl₂, 5 EGTA, 5 Hepes, 5 glucose, 5 K₂-ATP, 5 Na₂ creatine phosphate and 0.25 Na-GTP; pH adjusted to 7.2 with KOH. Action potential and currents were recorded at 34 ± 1°C. Capacitance and 60–80% of series resistance were routinely compensated. The sampling frequency was 2.5 kHz; the −3 dB cut-off frequency was 1 kHz. Action potentials were recorded under current clamp, holding at 0 pA; the stimulus pulse was 4 ms, and 1.2 × threshold strength at 0.25 Hz. Data were evaluated using pClamp 9.0 (Axon Instruments) and Origin 7.0 software (Origin Lab Corporation, Northampton, MA), and results are given as means ±sem; Student's unpaired t test or one-way ANOVA was used to compare the data as appropriate, and P < 0.05 was considered to indicate statistical significance.

Isolated perfused hearts and optical mapping

The whole-heart isolated method and optical mapping procedure have been described previously (Brunner et al. 2008). Briefly, littermate control (LMC) and transgenic rabbits (either sex, 3.5–5.5 kg) were injected with buprenorphene (0.03 mg kg⁻¹, i.m.), acepromazine (0.5 mg kg⁻¹, i.m.), xylazine (15 mg kg⁻¹, i.m.), ketamine (60 mg kg⁻¹, i.m.) and pentothal (35 mg kg⁻¹, i.v.) plus heparin (200 U kg⁻¹). The beating heart was harvested from the anaesthetized rabbit via sternotomy and retrogradely perfused through the aorta with (in mmol·L⁻¹): 130 NaCl, 24 NaHCO₃, 1.0 MgCl₂, 4.0 KCl, 1.2 NaH₂PO₄, 5 dextrose, 25 mannitol and 1.25 CaCl₂, at pH 7.4, gassed with 95% O₂ and 5% CO₂. Temperature was maintained at 37.0 ± 0.2°C, and perfusion pressure was adjusted to ∼60 mmHg with a peristaltic pump. Hearts were placed in a chamber to maintain temperature, and 5 μmol l⁻¹ blebbistatin was added to the perfusate to reduce movement artifacts (Fedorov et al. 2007). Hearts were stained with a voltage-sensitive dye, di-4 ANEPPS (Invitrogen, Carlsbad, CA, USA), using 25 μl of stock solution (1 mg·ml⁻¹ of DMSO) delivered through a bubble trap, above the aortic cannula. The ECG and perfusion pressure were continuously monitored. Hearts were monitored for adequate perfusion throughout the study by visual inspection for pink hue, homogeneous fluorescence and action potential shape (with prominent plateau phase). After 30 min of perfusion with blebbistatin, the right atrial appendage was incised and the triangle of Koch exposed. The apex of the triangle of Koch was cauterized until complete atrioventricular (AV) block was seen on ECG monitoring to slow down heart rate during ISO perfusion. Atrioventricular ablation ensures heart rate below 1 Hz (slower than 60 beats min⁻¹) during ISO perfusion. The hearts were then stimulated at 1 Hz and perfused with 100 nm ISO. The LQT1 and LQT2 hearts showed frequent EAD formation during ISO perfusion, and action potential duration (APD) measurements may include EADs in these models.

Computer simulation

Computer simulations were carried out using the Luo–Rudy phase 1 model (Luo & Rudy, 1991). The potassium current (I_K) from the Luo–Rudy phase 1 model was replaced by the slow and fast potassium currents, I_Ks and I_Kr, formulated by Zeng et al. (1995). The time constants of the L-type Ca²⁺ current and the activation time constant of the slow potassium current were adapted as in the study by Tran et al. (2009). To model the effects of isoprenaline, we increased the maximal conductance () of I_Ks from 0.433 to 0.866 mS μF⁻¹, following an exponential time course with time constant τ_Ks = 43.6 s, as measured from the voltage-clamp experiment (Fig. 5B). The maximal conductance () of the calcium channel varied from 1.0 to 2.0 mS μF⁻¹ following an exponential time course with time constant τ_Ca = 9.2 s, as measured from the voltage-clamp experiment (see Fig. 5B). All other model parameters were set to the standard values given in the original Luo–Rudy phase 1 model (Luo & Rudy, 1991).

Figure 5. Isoprenaline (100 nm) boosted both I_Ca and I_Ks in a time-dependent manner.

A, the voltage protocol used to record I_Ca and I_Ks at the same time. Holding potential is −50 mV, voltage first jump to −10 mV for 200 ms to record peak Ca²⁺ current, then further jump to +10 mV for 2 s to record slow activated I_Ks. Tail current of I_Ks could be best observed at −30 mV. The pulse was repeated every 6 s following the ISO perfusion. Recordings were done in the presence of 5 μm E-4031 to eliminate I_Kr. B, the current is shown as a function of time. The I_Ca was measured as peak current at −10 mV. Steady-state I_Ks was measured at the end of the 10 mV testing potential. They could best be fitted with one exponential function. The averaged time constant for I_Ca is 9.2 ± 1.15 s (n = 6) and for I_Kr 43.6 ± 9.82 s (n = 6).

Results

Extracellular K⁺-dependent EAD induction in LQT2 myocytes

Clinically, patients with low serum K⁺ concentration following diuretic use are at elevated risk for developing cardiac arrhythmia (Cohen et al. 1987). In either the congenital or drug-associated forms of QT prolongation, correction of extracellular potassium to the normal range can shorten the QT interval (Choy et al. 1977; Compton et al. 1996; Etheridge et al. 2003). We previously performed cellular electrophysiological studies using isolated rabbit cardiomyocytes with 5.4 mm K⁺ in the bath solution (Brunner et al. 2008). In those conditions, we did not detect triggered activity in single myocytes. Here we tested the effects of lowering extracellular K⁺ on the induction of EADs in single myocytes. Figure 1Aa shows an example of APD in three different concentrations of extracellular potassium. Clearly, lower [K⁺]_o would prolong action potential duration and lower the resting membrane potential. The average action potential duration at 90% repolarization (APD₉₀) was 450 ± 46.9 ms (n = 5) in 3.6 mm[K⁺]_o, but decreased to 366.4 ± 36.3 and 289.2 ± 34.2 ms in 5.4 and 7.2 mm[K⁺]_o, respectively, in line with clinical studies linking low K⁺ with further prolongation of QT interval. The resting membrane potentials at 3.6, 5.4 and 7.2 mm[K⁺]_o were −91.2, −81.9 and −74.3 mV, respectively, very close to K⁺ equilibrium potentials extrapolated from the Nernst equation. Figure 1Ab shows that EADs did not occur with 5.4 mm[K⁺]_o in myocytes from LQT2 rabbits, but did arise when 5.4 [K⁺]_o was replaced with 3.6 mm[K⁺]_o. Based on this initial test, we carried out the following experiments using 3.6 mm[K⁺]_o. Figure 1B shows action potential and EAD formation in the three groups. Early after-depolarizations were not detected among 25 LMC and 21 LQT1 myocytes. By contrast, we recorded EADs in 13 of 16 LQT2 myocytes, suggesting substantially different genotype-dependent responses to low [K⁺]_o. The average APD₉₀ at 0.25 Hz stimulation was 510.6 ± 25.5 ms (n = 25) in LMC myocytes, 979.5 ± 60.9 ms (n = 21) in LQT1 myocytes and 1292.1 ± 189.5 ms (n = 16) in LQT2 myocytes. Notably, the action potential was relatively stable in LMC myocytes, somewhat variable in LQT1 myocytes and unstable in LQT2 myocytes. In LQT2 myocytes, some of the action potentials were prolonged due to slow repolarization, while others had variable-length APD, often due to a typical phase 3 EAD formation. Of note, when measuring APD we only included action potentials in myocytes without EAD. The take-off potential of the EADs was −21.54 ± 0.99 mV (n = 12). Of 13 myocytes with EADs, one EAD triggered spontaneous membrane potential oscillation lasting several minutes (Figure 1Bc, bottom trace).

Figure 1. Effect of [K⁺]_o on action potential and EAD formation.

Aa, dependence of action potential and resting membrane potential on [K⁺]_o in a single ventricular myocyte from a littermate control animal (LMC). Ab, early after-depolarization (EAD) occurred when [K⁺]_o was switched from 5.4 to 3.6 mm in a single ventricular myocyte from an LQT2 rabbit. B, seven consecutive action potentials in LMC (Ba), LQT1 (Bb) and LQT2 myocytes (Bc, top trace) in 3.6 mm[K⁺]_o; Bc, bottom trace, membrane potential oscillation secondary to EAD in an LQT2 myocyte.

Relative contributions of I_Kr and I_Ks to action potential repolarization and its relationship to EADs

We further investigated potential mechanisms by which low [K⁺]_o could induce marked prolongation of APDs and EAD generation in the LQT2 genotype but moderate prolongation of APDs and no EADs in the LQT1 genotype. As phase 3 EAD has been linked to the reactivation of I_Ca (January & Riddle, 1989; January & Moscucci, 1992; Clusin, 2003), we initially hypothesized that the propensity of EAD formation in LQT2 myocytes was related to the possible remodelling of I_Ca in LQT2 myocytes, which is different from remodelling in LQT1 myocytes. However, after we quantified I_Ca in three groups of myocytes, we found that the peak I_Ca did not differ between LQT1 and LQT2, although both were smaller than in LMC myocytes, as shown in Figure 2Ab. The density of peak I_Ca in LMC, LQT1 and LQT2 was 18.19 ± 0.79 (n = 18), 16.13 ± 0.50 (n = 23, P < 0.05) and 14.80 ± 0.63 pA pF⁻¹ (n = 7, P < 0.05), respectively. Alternative explanations for differential prolongation of APDs could be related to the relative contributions of I_Kr and I_Ks to action potential repolarization. We applied an action potential clamp protocol to measure individual currents during repolarization as previously described (Jost et al. 2005). Figure 2B shows traces of I_Kr (red) and I_Ks (black) during action potential clamp at the pacing rate of 0.25 Hz. In these conditions, the current density of peak I_Kr was three times larger than peak I_Ks (1.05 ± 0.07 vs. 0.33 ± 0.09 pA pF⁻¹, n = 6). Both currents peaked at the late phase of the action potential, but more prominently for I_Kr. In addition, the deactivation current of I_Kr is larger, which could be seen clearly when the resting membrane potential returned; however, no clear deactivation current was seen in I_Ks.

Figure 2. Relative contributions of I_Kr and I_Ks to action potential repolarization.

A, L-type Ca²⁺ current (I_Ca) and its peak current I–V curve. Aa, original currents of L-type Ca²⁺ current in an LMC myocyte. Holding is at −50 mV, test potential is from −40 to +40 mV, pulses are 250 ms, and pulse intervals are 2 s. Ab, I–V curve of peak I_Ca in LMC, LQT1 and LQT2 myocytes. B, I_Kr and I_Ks isolated from LMC myocytes using an action potential stimulation protocol. I_Kr or I_Ks was defined as a current sensitive to 5 μm E-4031 or 30 μm chromanol 293B. The stimulation action potential pulse was acquired from an LMC myocyte in 3.6 mm[K⁺]_o; the action potential duration at 90% repolarization was around 500 ms. C, an example of the action potentials (top panels) and derivative of the membrane potential (bottom panels). The action potentials were from the sixth traces of each group in Figure 1B. The derivative of the membrane potential is the rate of membrane change (ΔV/Δt) and was also proportional to transmembrane current.

We further compared repolarizing rates during action potential in different genotypes. Figure 2C shows action potentials from LMC, LQT1 and LQT2 myocytes, and the bottom traces show first derivatives of action potentials, illustrating the speed with which action potentials repolarize during phase 3 of action potentials. The amplitude of LQT2 shows the slowest repolarization, indicating that I_Kr is a major repolarizing current and that removing I_Kr prolongs APD substantially more than removing I_Ks in these conditions.

Isoprenaline induced EAD and triggered activities

Clinically, syncope in LQT1 and LQT2 patients often occurs during conditions of physical or emotional stress (Schwartz et al. 2001), which indicates the involvement of sympathetic stimulation in the triggering of the symptom (Moss et al. 1991; Shimizu, 2002). Here, we used ISO as an adrenergic receptor stimulant to test its effect on APD and EAD formation in our genetic background rabbit models. With 50 nm ISO, APDs in LMC were prolonged by 29%, and APDs in LQT1 were further prolonged by 52% (Fig. 3A). No EADs were observed in 14 LMC myocytes (Fig. 3B); however, EADs were induced in seven of 19 LQT1 myocytes (Fig. 3C). In LQT2 myocytes, APD was shortened by the application of ISO. Figure 3D shows an LQT2 myocyte that had EADs before stimulation with ISO, but after perfusion the EADs were abolished with shortening APD. As in the LQT2 myocytes, EAD also triggered membrane potential oscillations in three LQT1 myocytes. In one LQT1 myocyte, the oscillations lasted several minutes, as shown in Figure 3E. Thus, continuous stimulation with ISO induced EAD and membrane potential oscillation in LQT1 myocytes, but may protect LQT2 myocytes from EAD formation and membrane potential oscillation.

Figure 3. Effect of isoprenaline (ISO; 50 nm) on action potential duration (APD) and EAD formation.

Experiments were done with 3.6 mm[K⁺]_o. A, ISO increased APD in LMC and LQT1 myocytes and shortened APD in LQT2 myocytes. B, ISO prolonged APD in an LMC myocyte. C, ISO prolonged APD and induced EAD in an LQT1 myocyte. D, ISO shortened APD in an LQT2 myocyte. E, spontaneous membrane potential oscillation (Eb) secondary to EAD (Ea) in an LQT1 myocyte.

To test the effect of ISO on APD and EAD formation at the organ level, we carried out optical mapping in Langendorff-perfused hearts. To slow the heart rate, the AV node was ablated in all hearts as described in the Methods section. Among the four hearts studied from each genotype, no EADs were seen in any group with baseline conditions at 1 Hz stimulation ([K⁺]_o 4 mm). After addition of isoprenaline, all four LQT1 hearts and three of four LQT2 hearts demonstrated EADs, but none of the four LMC hearts demonstrated EADs. A striking difference was seen in the behaviour of EADs in LQT1 and LQT2 hearts (Fig. 4A). In LQT1 hearts, EADs persisted throughout ISO exposure. In contrast, LQT2 hearts demonstrated EADs within the first 4–10 s of exposure to ISO, after which EAD incidence was diminished with significant APD shortening (Fig. 4A). Action potential duration differed among the three genotypes in baseline conditions, with a mean APD of 249 ± 48 ms in LMC hearts, 354 ± 69 ms in LQT1 hearts and 505 ± 91 ms in LQT2 hearts (P < 0.05 for all comparisons among the three groups). Exposure to ISO also had different effects on APD in the three groups (Fig. 4B). In LMC hearts, the APD was shortened by a mean of 52 ± 40 ms. In LQT2 hearts, the APD shortening was more pronounced, with a mean reduction in APD post-ISO of 124 ± 83ms (n.s. between LMC and LQT2). The opposite effect was seen in LQT1 hearts, where a prolongation in APD was seen post-ISO exposure. The mean APD prolongation in LQT1 hearts post-ISO was 157 ± 78 ms (P < 0.05 for LQT1 vs. LMC and LQT2).

Figure 4. Response to adrenergic stimulation in Langendorff-perfused hearts.

A, optical signal examples of action potentials from LMC, LQT1 and LQT2 hearts at baseline and over the initial 10 s exposure to isoprenaline (100 nm). The atrioventricular node was ablated in all hearts to maintain the slow heart rate during isoprenaline exposure. In LQT1, there was continuous prolongation of the action potential and persistent EADs, while LQT2 shows initial EADs, with subsequent shortening of the action potential at 5 and 10 s. B, change in action potential at 10 s of isoprenaline exposure from baseline. Action potential prolongation is seen in LQT1, while action potential shortening is seen in LQT2.

Time-dependent activation of I_Ca and I_Ks in the presence of isoprenaline

Our ISO experiments strongly indicated time-dependent changes in APD and EAD generation, at both the single-cell and organ levels. We reasoned that ISO can increase several ionic currents and calcium handling on different time scales, so that some ion channels can be augmented earlier than others, creating time-dependent APD and EAD generation. We investigated time-dependent increases in I_Ca and I_Ks, the two antagonizing currents during repolarization in cardiomyocytes, in the presence of ISO, a well-known stimulant of both I_Ca and I_Ks (Sanguinetti et al. 1991; Imredy et al. 2008).

We found differences in the time course of the increases in I_Ca and I_Ks, with faster augmentation of I_Ca. Figure 5A shows I_Ca and I_Ks recorded simultaneously. Figure 5B shows that both I_Ca and I_Ks increased in a time-dependent manner. Fitting the points of I_Ca or I_Ks with one exponential function resulted in a time constant of 9.2 s for I_Ca increase and 43.6 s for I_Ks increase. We next performed computer modelling studies of APD that were based on the results depicted in Figure 5B. At starting values of and , there were no EADs, and the APD was 468 ms (Fig. 6Ba). When and changed with the time constants mentioned above to mimic the effects of ISO, an EAD appeared at t = 7 s (Fig. 6Bb); more EADs appeared in the action potential (eight EADs with APD = 2376 ms at t = 35 s; Fig. 6Bc). With time, however, the EADs were suppressed, and after 180 s APD was 1642 ms (Fig. 6Bd). This scenario is qualitatively consistent with our experimental observations in intact hearts (Fig. 4A) in terms of the biphasic change of APD and reduction of EAD incidence with prolonged ISO treatment.

Figure 6. Computer simulation of transient EAD genesis due to isoprenaline in an LQT2 myocyte.

A, the maximal conductances of I_Ca and I_Ksversus time to mimic the effects of isoprenaline. A negative sign was added to the I_Ca conductance to agree with Fig. 5B. B, predicated action potentials at different time points as indicated by the letters in A.

Discussion

The hallmark of long-QT syndrome is EADs followed by polymorphic ventricular tachycardias. More than 13 genes are known to cause LQTS, but it is not clear whether the same or different conditions facilitate EADs and sudden cardiac deaths in various genotypes. Our results show that conditions that prolong APD, such as slow heart rate and slightly low [K⁺]_o, can cause dramatic prolongation of APD and EAD formation in LQT2 myocytes, while the same effects on LQT1 and LMC are relatively limited to slight prolongation of APD. Sympathetic stimulation also has disparate effects in LQT1 and LQT2, such as persistent EADs in LQT1 but brief EAD formations in LQT2 during initial perfusion of ISO, most probably due to rapid activation of I_Cavs. slow activation of I_Ks during sympathetic stimulation.

Prolongation of APD, regardless of causes, such as delayed inactivation of depolarizing currents (sodium current in LQT3) or lack of repolarizing currents, can produce the same outcome, promoting abnormal depolarizations during the plateau phase of action potentials (EADs). However, the exact conditions that facilitate EADs may be different in different genotypes, because repolarization of action potentials in many species, including humans, is brought about by two delayed rectifiers, I_Kr and I_Ks, based on relative activation and inactivation kinetics (Horie et al. 1990; Liu & Antzelevitch, 1995; Li et al. 1996). Their relative contributions to cardiac repolarization can be dependent on physiological needs, such as fast or slow heart rate. Owing to its rapid kinetics, I_Kr plays a dominant role in the initial repolarization process, especially at slow heart rates, while I_Ks works as a reserve to assist repolarization by balancing the increase of I_Ca during fast heart rate and/or sympathetic activation.

Given a slow heart rate, the relative contribution of I_Ks to repolarization is minimal, and I_Kr becomes the dominant repolarization current. Using an action potential voltage protocol, we demonstrated the function of both I_Kr and I_Ks during repolarization; however, in our experimental conditions the density of I_Kr was significantly higher than I_Ks, as previously reported by Jost et al. (2005). Here we demonstrated that the density of I_Kr was three times higher than that of I_Ks in rabbit cardiomyocytes, suggesting that I_Kr plays a dominant role in repolarization. Our results also showed that, given a slow heart rate, slightly lowering [K⁺]_o (to 3.6 mm) was enough to induce EADs in LQT2 rabbit myocytes but not in LMC or LQT1 myocytes. Low [K⁺]_o downregulates both I_Kr and inward rectifier K⁺ current I_K1 (Damiano & Rosen, 1984; Scamps & Carmeliet, 1989; Yang et al. 1997; Bailly et al. 1998; Guo et al. 2009) and has often been used in the study of EADs and triggered activities. Damiano & Rosen (1984) showed that EAD and spontaneously triggered activities were induced by 2–4 mm[K⁺]_o, and that elevated extracellular [K⁺]_o either inhibited EAD formation or changed the EAD from high membrane potential to low membrane potential.

Our results are consistent with reports from other laboratories that found no EAD formation in heart tissue treated with I_Ks blocker alone (chromanol 293B) to mimic the LQT1 model (Burashnikov & Antzelevitch, 2000). Blockade of I_Ks produces a homogeneous prolongation of repolarization and refractoriness across the ventricular wall and requires additional adrenergic stimulation to induce EADs (Shimizu & Antzelevitch, 1998). Our results again support the contention that prolongation of QT or APD due to loss of I_Ks is limited compared with the loss of I_Kr and must be accompanied by sympathetic stimulation to produce arrhythmias (Antzelevitch, 2005). In line with single LQT1 myocyte experiments, EADs were observed in perfused LQT1 hearts only in conditions of slow heart rate achieved by AV ablation and stimulation with an adrenergic agonist. Clinical studies (Tan et al. 2006) showed that Torsade de Pointes (TdPs) are pause dependent only in LQT2, not LQT1 types. Although slow heart rate is a necessary condition for EAD generation in both LQT1 and LQT2, our study shows that LQT1 requires adrenergic stimulation in addition to slow heart rate, suggesting that pause-dependent TdPs are more likely in LQT2. Interestingly, LQT2 hearts show diminishing EADs with adrenergic stimulation, suggesting that these two genotypes of LQTS may have different risk profiles for adrenergic-induced triggered activity.

Two major currents modulated by adrenergic stimulation are I_Ca and I_Ks. An increase of inward I_Ca prolongs the APD, while a boost of outward I_Ks would shorten the APD. The boosting effect on both I_Ca and I_Ks would explain how ISO affects APD in LQT1 and LQT2 myocytes. In LQT1 myocytes, the lack of I_Ks to antagonize I_Ca during ISO stimulation would be responsible for further APD prolongation and subsequent EAD formation. In addition, we found that the two currents have different response times to adrenergic stimulation, I_Ca being fast and I_Ks being slow. The average time constant for I_Ks increase is 43.6 s and for I_Ca it is 9.2 s, more than three times faster, creating a high-risk window when I_Ca increases but I_Ks is not yet fully activated by adrenergic stimulation; therefore, LQT2 myocytes can show a high incidence of EADs during the early phase of adrenergic stimulation or strong sympathetic surge. Our findings support this contention, because transient EAD occurred after administration of ISO to Langendorff-perfused LQT2 hearts. This may explain why LQT2 patients are at high risk of syncope during sudden sympathetic surges caused by emotional stress, such as telephone rings, alarm clock sounds or surprises. Continuous perfusion of single LQT2 myocytes with ISO caused APD shortening due to the augmentation of I_Ks. Consequently, the frequency of EAD was reduced in LQT2 myocytes.

In summary, EAD conditions are genotype specific, and EADs can be produced in LQT2 myocytes simply by lowering [K⁺]_o, while LQT1 myocytes require sympathetic stimulation. Slower ISO activation of I_Ks compared with I_Ca may explain the different sympathetic modes, i.e. sympathetic surge vs. high sympathetic tone, associated with polymorphic ventricular tachycardias in LQTS patients.

Limitations

The isolated cardiomyocytes and whole hearts allow direct studies of electrophysiological properties at the cellular and organ levels. Unfortunately, however, they do not fully recapitulate the in vivo properties of the heart, posing limitations. As a result, the action potential measurements were carried out at relatively slow rates (0.25 and 1 Hz). Optical mapping was also done after the AV node was ablated to maintain a slow heart rate. This may not reflect the situation in rabbits in vivo, which of course live with a fast heart rate (3 Hz); but ECG recordings immediately before polymorphic ventricular tachycardias in vivo often show very slow heart rate and AV block, which is in line with the slow stimulation in our experiment. Measurements of APD in the presence of ISO show large variation due to EAD formation and changes in previous diastolic intervals caused by previous EADs. Our computer modelling was able to recapitulate the experimental observations qualitatively at the whole-heart level. However, quantitative differences exist, probably due to the limitation by the electrophysiological data from isolated cardiomyocytes. Additionally, the data from our L-type Ca²⁺ current study are very preliminary. More detailed study, including Ca²⁺ current activation and inactivation kinetics, is warranted.

Glossary

Abbreviations

APD

action potential duration

APD₉₀

action potential duration at 90% repolarization

AV

atrioventricular

EAD

early after-depolarization

I_Ca

L-type calcium current

I_Kr

fast component of delayed rectifying K⁺ current

I_Ks

slow component of delayed rectifying K⁺ current

ISO

isoprenaline

[K⁺]_o

extracellular potassium concentration

LMC

littermate control

LQTS

long-QT syndrome.

Author contributions

G.X.L. did myocytes isolation, patch clamp experiments and data analysis (at Rhode Island Hospital); B.R.C. and O.Z. did optical mapping experiments and analysis (at Rhode Island Hospital); E.L. and Z.Q. did computer modelling studies (at UCLA); G.X.L., W.L. and G.K. analyzed data, wrote and revised the manuscript. All authors approved the final version for publication.