Altered Na⁺ Channels Promote Pause-Induced Spontaneous Diastolic Activity in Long QT Syndrome Type 3 Myocytes

Arrhythmias result from abnormalities in impulse initiation or impulse conduction or a combination of both.¹ One of the possible cellular mechanisms for impulse initiation is triggered activity caused by afterdepolarizations. There are 2 types of afterdepolarizations in heart. Early afterdepolarizations (EADs) are generated near action potential plateau voltages before repolarization is complete and are driven, at least in part, by L-type calcium channel activity.² A second type of arrhythmic disturbance, generally referred to as “delayed after depolarization” (DAD) because of the close coupling between preceding action potential activity and subsequent depolarization that follows with a delay,³ can occur over more negative voltages. Both EADs and DADs are calcium-dependent events that can be exacerbated by changes in action potential waveform.^4,5 It is now clear that action potential prolongation, whether disease or drug induced, can lead to life-threatening cardiac arrhythmias, most likely through 1 or both of these pathways.⁶

The acquired and congenital long QT syndromes (LQTSs), caused either by drug- or mutation-induced changes in cardiac ion channel function, share a common phenotype of action potential (and hence concomitant QT interval) prolongation.⁷ Thus LQTS-associated arrhythmias are likely to occur via 1 or both of these Ca²⁺-dependent pathways. In most cases, action potential prolongation occurs in LQTSs (acquired and congenital) because of mutation or drug-induced loss of channel function. However, this is not the case for LQTS type 3 (LQT3), which is caused by mutations in SCN5A, the gene coding for the Na_v1.5 sodium channel α subunit. Here, mutation-induced changes in Na⁺ channel gating provide additional inward Na⁺ channel current that delays cellular repolarization.⁸ Though rare, LQT3 mutations are thought to be the most lethal of the prevalent LQTS mutations.⁹

Previous in vivo^10,11 and in silico^12,13 studies have provided support for the concept that LQTS-induced action potential prolongation can induce EAD-driven electrical activity, which is related in a gene or drug-dependent manner to heart rate and/or adrenergic input. However, little evidence has been presented to date that diastolic dysfunction may also be causally related to LQTS perturbation in electrical activity. Here we focused on a well-described LQT3 mutation in which amino acid residues 1505 to 1507 (KPQ) are deleted in the Nav1.5 channel (ΔKPQ mutation)¹⁴ that promotes persistent or sustained Na⁺ channel current (I_NaL) during prolonged depolarization through mutation-altered inactivation.¹⁵ We took advantage of myocytes from a previously described mouse heterozygous for a knock-in KPQ deletion (ΔKPQ)¹¹ to test the hypothesis that, in addition to previously reported proarrhythmic behavior in the form of rate-dependent EAD driven dysfunction, myocytes isolated from these mice are also predisposed to spontaneous diastolic DAD arrhythmic activity caused by mutation-altered Na⁺ entry.

Our results support this hypothesis and provide evidence for the first time that although all LQTS mutations, independent of the mutated gene, have a common phenotype (delayed ventricular repolarization); our results demonstrate a marked additional contribution of mutation-altered Na⁺ entry to the incidence of pause-dependent spontaneous diastolic activity that is unique to LQT3 and suggest that altered Na⁺ entry may contribute to the elevated lethality of LQT3 versus other LQTS variants.

Materials and Methods

An expended Materials and Methods section is available in the online data supplement at http://circres.ahajournals.org

Transgenic Mice and Isolation of Cardiac Ventricular Myocytes

Mice heterozygous for a knock-in KPQ deletion in SCN5A¹¹ were genotyped by PCR analysis to confirm the expression of the SCN5A ΔKPQ Na⁺ channels. Ventricular myocytes from adult mice (1 to 8 months of age) were dissociated as previously described.^16,17 The institutional Animal Care and Use Committee at Columbia University approved the protocols for all animal studies.

Electrophysiology

Current- and voltage-clamp protocols and solutions are described in detail in the online data supplement. All experiments were carried out at room temperature (22°C).

Data Analysis

pClamp 8.0 (Axon Instruments Inc), Excel (Microsoft), and Origin (Microcal Software) were used for data acquisition and analysis. Data are presented as mean values±SEM. Two-tailed Student’s t test was used to compare 2 means; a value of P<0.05 was considered statistically significant.

Results

The ΔKPQ mutation disrupts Na⁺ channel inactivation such that during prolonged depolarization, channels harboring this mutation are characterized by a small component of noninactivating Na⁺ channel activity.¹⁵ Knock-in mice, heterozygous for Na⁺ channels with the ΔKPQ mutation, express Na⁺ channel current with this phenotype that leads to prolongation of the ventricular action potential.¹¹ We begin this study by illustrating the impact of the ΔKPQ mutation on cellular action potentials and currents measured in myocytes isolated from ΔKPQ heterozygous mice and their wild-type (WT) littermates under our recording conditions. Figure 1 illustrates the presence of I_NaL (Figure 1B) and subsequent action potential prolongation (Figure 1D) in myocytes isolated from ΔKPQ knock-in compared with WT mice (Figure 1A and 1C) under our recording conditions. We found no significant differences in peak current density (I_NaP) in WT and ΔKPQ (−/+) myocytes either in low-sodium solutions (illustrated in Figure 1E) or in full-sodium external solutions (data not shown). However, in the ΔKPQ myocytes, a large component of I_NaL (1.35±0.17 pA/pF; n=12) was recorded (−10 mV, 150 ms), which was significantly greater than I_NaL in WT myocytes (0.10±0.03 pA/pF; n=7). Furthermore, we compared total cell capacitance, which is a reflection of myocyte surface area, of WT (155.4±6.9 pF; n=7) versus ΔKPQ (165.2±7.4 pF; n=12) myocytes and found no significant difference in this parameter. We also tested for but found no significant difference between L-type Ca²⁺ channel density in ΔKPQ and WT myocytes (Figure IV in the online data supplement). Current-clamp action potential recordings for the ΔKPQ myocytes are summarized in the Table 1. In comparison with recordings in WT myocytes, action potentials recorded in knock-in heterozygote myocytes displayed no significant differences in action potential peak voltage or resting membrane potential. However, the ΔKPQ myocyte action potentials were substantially prolonged at 50% (APD₅₀) and 90% (APD₉₀) of depolarization. APD₅₀ and APD₉₀ values were measured respectively as 20.6±3.6 ms and 242.4±32.4 ms in ΔKPQ myocytes as compared with 3±0.4 ms and 66.6±4.1 ms for WT myocytes (all P<0.001, WT versus ΔKPQ myocytes). Our results under our experimental conditions, thus, were consistent with those reported by Nuyens et al,¹¹ with the exception that we found no significant difference in peak current expression in the −/+ myocytes.

• Download figure
• Download PowerPoint

Pause-Induced Spontaneous Depolarizations

We next performed a series of current-clamp studies to determine whether the ΔKPQ myocytes were more prone to pause induced depolarizations at diastole. In each experiment, a train of preconditioning pulses was delivered with 3-ms, near-threshold current pulses followed by a 500-ms pause, a single near-threshold current pulse, and prolonged diastole (see supplemental Figure I). Using this protocol and varying the frequency of the preconditioning train pulses, we observed a mutation- and frequency-dependent incidence of spontaneous diastolic depolarizations. Figure 2 shows that as ΔKPQ heterozygous myocytes are paced at increasing rates during conditioning trains, spontaneous events occur and become larger, more frequent, and less delayed after the final stimulus. Only 1 of 5 ΔKPQ myocytes had spontaneous diastolic activity at 2 Hz and 2 of 5 at 5 Hz, whereas none of the WT myocytes displayed afterdepolarizations using this stimulus protocol. At 10 Hz, there is a significant difference (statistical power >95%, based on the difference and sample sizes) in DAD appearance comparing WT (0 of 13) versus ΔKPQ (5 of 7) myocytes (see the Table 2).

• Download figure
• Download PowerPoint

Voltage-Clamp Studies of Transient Inward Current

In the preceding current-clamp recordings, interpretation of the mechanism responsible for the afterdepolarizations was clouded by the concomitant increase in action potential duration (APD) present in myocytes with the LQT3 lesion as well as rate-dependent adaptation and cell-to cell variability in APD. As a result, we designed voltage-clamp protocols and ionic conditions to mimic both the short and long action potentials recorded in WT and mutant myocytes. Similar to the previous experiments, we applied trains of 100 conditioning voltage waveforms (either short- or long-preconditioning pulses) followed by a 500-ms pause, 1 extra voltage pulse, and then diastole. The short pulse experiment used a voltage step from −75 to −10 mV for 2 ms followed by a 38-ms ramp back to −75 mV, whereas the longer pulse consisted of a voltage step from −75 to −10 mV for 20 ms followed by a 70 ms ramp to −75 mV (supplemental Figure II). The results shown in Figure 3 illustrate the effect of pulse duration, preconditioning stimulus rate, and the presence of the mutation on spontaneous activity, which, under voltage-clamp conditions, is observed as transient inward current (I_TI). Even in the case of WT myocytes, increasing the duration and pulse frequency of the conditioning train voltage pulses can induce I_TI (compare Figure 3 WT traces short- versus long-pulse data), but we consistently found ΔKPQ myocytes more susceptible to I_TI induction in response to the same conditioning protocols. To quantify the combined magnitude and prevalence of the I_TI, the time integral of the transient inward current (∫I_TI) is calculated for the 5-second interval following the last stimulus (see the expanded Materials and Methods section in the online data supplement). This represents the total charge transported during the spontaneous events, which correlates with generation of after depolarizations and extra systole under current-clamp conditions.^18,19 This charge measurement is shown in Figure 3C and 3D, where it is evident that prolongation of the conditioning pulse and the presence of the ΔKPQ deletion mutation both predispose the myocytes to the generation of I_TIs.

• Download figure
• Download PowerPoint

Confocal imaging of Ca_i²⁺ transients in fluo 3-loaded myocytes indicated that transient inward currents correlated with measured spontaneous Ca²⁺ events at diastole following conditioning trains, and the transient currents were larger and more frequent in ΔKPQ myocytes (≈2.5 F/F₀ in 3 of 4 cells) than in WT myocytes (≈1.5 F/F₀ in 2 of 4 myocytes) (supplemental Figure VI). These experiments indicate that the diastolic transient inward currents correlated with Ca_i transients recorded at diastolic potentials following conditioning pulse trains. These are well-described properties of I_TIs generated by spontaneous sarcoplasmic reticulum Ca²⁺ release²⁰ and carried largely by Na⁺/Ca²⁺ exchange current.¹⁸

If the I_TIs generated in the ΔKPQ myocytes were caused by diastolic Ca_i²⁺ transients resulting from sarcoplasmic reticulum Ca²⁺ release, we reasoned that these events should be inhibited by increasing Ca²⁺ buffer capacity of intracellular recording solutions. In the preceding recordings, free [Ca²⁺] in the pipette filling solution was buffered to 100 nmol/L using only 50 μmol/L EGTA (Materials and Methods) to minimize effects of EGTA on conditioning train-dependent changes in Ca_i²⁺. To test the role of Ca_i²⁺ buffering on I_TI recorded in these experiments, the short- and long-pulse protocols were repeated with another pipette solution in which free [Ca²⁺] was again buffered to 100 nmol/L but with a high EGTA concentration (11 mmol/L). Figure 4 shows that high Ca²⁺ buffer suppresses I_TI formation. We found similar results in 5 of 5 ΔKPQ myocytes tested, indeed suggesting a role for Ca²⁺ in the genesis of the spontaneous activity.

• Download figure
• Download PowerPoint

After observing that the presence of I_TIs is greatly enhanced in the ΔKPQ myocytes, we wanted to determine whether the mutation-induced changes in Na⁺ channel activity contributed to I_TI induction, providing a pathway for arrhythmia susceptibility that might be linked specifically to LQT3 lesions. Therefore, voltage-clamp protocols were repeated in heterozygote and WT littermates in the absence or presence of tetrodotoxin (TTX) (30 μmol/L). Figure 5 shows current traces for both the short- and long-pulse voltage-clamp protocols in mutant and WT cells before and after treatment with TTX. In the ΔKPQ myocytes, application of TTX greatly reduces the magnitude and number of the observed transient currents. Examining the time integral of the I_TI, as shown in Figure 5, it is evident that TTX (30 μmol/L) significantly reduces the charge transported across the membrane. However, in WT myocytes, where I_TIs are observed with long-pulse conditioning trains, TTX application does not significantly affect the amount of charge carried by I_TI. These results are consistent with a distinct contribution of altered Na⁺ channel activity to I_TI generation in ΔKPQ myocytes in response to conditioning train protocols designed to resemble electrical activity in murine ventricle. Interestingly, I_TI can be induced by conditioning trains of prolonged depolarizing pulses even in WT myocytes, but in this case I_TI induction is TTX insensitive (see Discussion).

• Download figure
• Download PowerPoint

Extension to Human Rates

As noted above, our voltage-clamp experiments were designed to imitate the behavior of the mouse myocytes observed in current-clamp conditions; however, of much greater interest is whether or not mutation-altered Na⁺ channel activity contributes distinctly to induction of I_TI under voltage conditions, both pulse width and frequency, more relevant to human physiology. We thus developed voltage-clamp protocols in which, starting from a −75 mV holding potentials, conditioning pulses were imposed to −10 mV for either 100 or 400 ms and then ramped back to −75 mV over 100 ms. Conditioning trains of 10 short or long pulses were applied at 0.5 Hz, after a 500-ms postconditioning pause, a final long- or short-pulse waveform was imposed, followed by diastole, during which I_TI was measured. The typical TTX-sensitive currents recorded during these conditioning depolarizing waveforms at 0.5 Hz for both 100- and 400-ms duration protocols revealed not only the noninactivating Na⁺ channel current but also TTX-sensitive Na⁺ channel current during the repolarizing voltage ramp (see supplemental Figure V and Clancy et al⁴). As LQT3 patients are more sensitive to sleep induced arrhythmias and are prone to prolongation of QT duration upwards of 500 ms,²¹ we were interested in seeing whether the ΔKPQ lesion produced pause-induced transient inward currents at slow rates in the 400-ms-long preconditioning pulse. Figure 6 illustrates currents recorded in response to the 400-ms protocol at a preconditioning rate of 0.5 Hz for WT and ΔKPQ myocytes. For the ΔKPQ, but not WT myocytes, this protocol induces I_TI (Figure 6B), which is suppressed by TTX (Figure 6D). Figure 6E shows summary data for the effects of TTX on ∫I_TI in WT and ΔKPQ myocytes. These protocols did not induce I_TIs in WT myocytes. In contrast, ΔKPQ myocytes were prone to I_TI induction for both long- and short-pulse conditioning trains, and in all cases ∫I_TI was significantly reduced by TTX application.

• Download figure
• Download PowerPoint

I_TI Suppression by Flecainide

The results with TTX implicate a role of altered Na⁺ channel activity in I_TI induction in ΔKPQ myocytes without implicating a role of noninactivating Na⁺ channel activity in this process. Consequently, we tested the effects of flecainide on ΔKPQ myocyte I_TIs generated by conditioning trains. We chose to study flecainide because it not only preferentially inhibits late, noninactivating Na_v1.5 channel activity, like other local anesthetic drugs,^22,23 but also because it has been shown to be an effective therapeutic agent in the treatment of LQT3 mutation carriers.^24,25 To examine this, the 100- and 400-ms depolarizing pulses were used again at 0.5 Hz and ∫I_TI was calculated for any events occurring before and after flecainide (3 μmol/L) application. We first confirmed that late current is approximately 5-fold more sensitive to flecainide block than peak current in ΔKPQ myocytes and that a 3 μmol/L flecainide concentration blocks almost 50% of late current, I_NaL, while inhibiting less than 10% peak current (supplemental Figure III), and thus we chose to test this flecainide concentration on I_TI. Figure 7, which summarizes these experiments, shows that 3 μmol/L flecainide effectively inhibits I_TIs induced by 0.5-Hz conditioning trains of 100-ms (left panel) and 400-ms (right panel) pulses, as illustrated in Figure 7C and 7D compared with the drug free correlates with Figure 7A and 7B, respectively. The bar graphs in Figure 7E, which summarize data from multiple experiments for each protocol, plot mean charge carried by I_TI as a function of preconditioning waveform. Flecainide significantly reduced the charge and subsequently the arrhythmogenic potential of these spontaneous events in response to both conditioning protocols.

• Download figure
• Download PowerPoint

Discussion

The results presented in this study suggest a role for mutation-enhanced I_NaL in the generation of spontaneous diastolic activity in LQT3. In myocytes isolated from mice heterozygous for a knock-in KPQ deletion (ΔKPQ), diastolic arrhythmogenic activity is significantly more prevalent than in myocytes isolated from WT littermates. Because pause-dependent arrhythmias were previously reported in ECG measurements in these mice,¹¹ and because clinical, in vivo animal and computational studies have demonstrated correlation between the spontaneous arrhythmias in LQTSs following pauses in stimulation,^26,27 we focused our studies on pause-dependent activity that might occur in myocytes. Our results show that even in WT myocytes, prolonged depolarization can lead to spontaneous activity (I_TIs). Because I_TIs are known to be caused by sarcoplasmic reticulum calcium overload, our results indicate that even in WT myocytes that are stimulated at sufficiently high frequencies and pulse durations, sarcoplasmic reticulum overload can occur. Our data indicate that these I_TIs are not TTX-sensitive and are consistent with effects of action potential duration on intracellular calcium dynamics via non-Na⁺ channel pathways such as L-type calcium channels.²⁸ When the same protocols are applied to ΔKPQ myocytes, spontaneous diastolic activity is more pronounced and is markedly suppressed by Na⁺ channel blockade. We suggest not only that the induction of diastolic activity by prolonged depolarization is a potential contributor to arrhythmogenesis for all LQTS mutations but also that its further increase and sensitivity to Na⁺ channel blockade in ΔKPQ myocytes reveals a key role of altered Na⁺ channels in this process, which is unique to LQT3 mutations. This finding thus will impact therapeutic management of LQT3 patients and will provide additional strategies to develop treatments that are specific for distinct LQTS variants.

Clinical Significance

The influence of genotype on the clinical course of the LQTS has been the subject of multiple important and informative investigations,^9,21,29–32 and, although clear patterns such as gene-related lethality and arrhythmia triggers are suggested, in part because of the relatively small numbers of cases available for study, relationships between genetic background and risk stratification remain complex. This is particularly relevant to studies of LQT3 patients, in which the number of patients investigated is the smallest of the major LQT variants. Here, studies suggest that high lethality is associated with arrhythmic events in LQT3^9,29 and that, compared with LQT1 and LQT2 patients, arrhythmia risk is higher during sleep for LQT3 patients, but elevated for LQT1 patients during exercise (elevated sympathetic nerve activity).²¹ However, a close look at the available data indicate that, even in the case of LQT3, other factors may also exacerbate arrhythmia risk because in the previous study, it was found that of the studied triggers in LQT3 patients: 39% were during sleep, 13% during exercise, 19% during emotional events, and 29% by unknown causes.²¹ Furthermore, during sleep, heart rate and sympathetic nerve traffic decrease progressively in non-rapid eye movement sleep, but both can increase significantly and transiently during rapid eye movement sleep.³³ Extraction of a single risk factor for LQT3 sleep during sleep clearly is complex. Thus, it is important to investigate potential mutation-specific mechanisms that might contribute to arrhythmogenic activity under a range of conditions that include, but are not limited to, slow heart rate (stimulation rate) and prolonged action potential (voltage-pulse) duration.

In the case of our experimental analysis of I_TI induction in ΔKPQ myocytes, conditioning pulse stimulation rate and duration increase the likelihood of I_TI induction. However, it must be remembered that in these experiments, in contrast to physiological conditions in LQT3 patients, conditioning pulse duration was constant, whereas pulse rate (equivalent to heart rate) was varied. When voltage is not feedback controlled in LQT3 mutation carriers, QT interval (APD) will shorten with increasing heart rate and thus reduce, according to our experiments, the impact on subsequent I_TI induction.^13,34,35 In contrast, at slower heart rates, both the amplitude of I_NaL and APD are expected to increase, and hence our experiments predict that in LQT3 patients, it is under these conditions that I_TI induction would be most likely to occur. Because LQT3 mutation carriers are most susceptible to cardiac events during bradycardia and low SNS,²¹ our results suggest that diastolic spontaneous activity can be a contributory factor to the initiation of arrhythmias in these patients.

I_NaL and DADs

That I_TI induction is more sensitive to TTX inhibition in ΔKPQ than in WT myocytes implicates a role of altered Na⁺ channel activity in this phenomenon, but the potent effects of flecainide on I_TI in ΔKPQ myocytes, supports the hypothesis that mutation-enhanced I_NaL is a major pathway contributing to these events. This result has several implications that add to our understanding of the therapeutic mechanisms of Na⁺ channel blockade in LQT3 and raise the possibility of these mechanisms and therapies in other cardiovascular disorders. First, for LQT3, where flecainide and other Na⁺ channel blockers have been shown to be effective clinically, our data suggest that this efficacy has at least 2 pathways. Because I_NaL is more sensitive to flecainide and other previously tested Na⁺ channel blockers, these drugs shorten LQT3 altered APD (and QT interval) as has been demonstrated in vivo, in vitro, and in silico.^{13,24,25,36,37} However, in addition to the direct effects of correction of APD, particularly at slow heart rates, these drugs have the added benefit of reducing the contributions to I_TI induction of mutation-altered Na⁺ entry during the prolonged QT intervals of mutation carriers. This distinction is made clearly in the experiments of Figure 7, in which flecainide was found to effectively inhibit I_TI despite the fact that conditioning pulse (similar to QT) duration was constant. Thus, our results predict that development of compounds that in which the selectivity for block of I_NaL versus I_NaP is maximized would lead to improved therapeutic potential of future compounds. Furthermore, because there is growing evidence that more common cardiovascular pathologies, such as heart failure³⁸ and ischemia,³⁹ may also promote increases in I_NaL, our results suggest that augmented I_NaL and its inhibition by selective blockers may contribute to the genesis and subsequent management of arrhythmias in these diseases as well.

Footnotes