Diastolic Spontaneous Calcium Release From the Sarcoplasmic Reticulum Increases Beat-to-Beat Variability of Repolarization in Canine Ventricular Myocytes After β-Adrenergic StimulationNovelty and Significance
Rationale: Spontaneous Ca2+ release (SCR) from the sarcoplasmic reticulum can cause delayed afterdepolarizations and triggered activity, contributing to arrhythmogenesis during β-adrenergic stimulation. Excessive beat-to-beat variability of repolarization duration (BVR) is a proarrhythmic marker. Previous research has shown that BVR is increased during intense β-adrenergic stimulation, leading to SCR.
Objective: We aimed to determine ionic mechanisms controlling BVR under these conditions.
Methods and Results: Membrane potentials and cell shortening or Ca2+ transients were recorded from isolated canine left ventricular myocytes in the presence of isoproterenol. Action-potential (AP) durations after delayed afterdepolarizations were significantly prolonged. Addition of slowly activating delayed rectifier K+ current (IKs) blockade led to further AP prolongation after SCR, and this strongly correlated with exaggerated BVR. Suppressing SCR via inhibition of ryanodine receptors, Ca2+/calmodulin-dependent protein kinase II inhibition, or by using Mg2+ or flecainide eliminated delayed afterdepolarizations and decreased BVR independent of effects on AP duration. Computational analyses and voltage-clamp experiments measuring L-type Ca2+ current (ICaL) with and without previous SCR indicated that ICaL was increased during Ca2+-induced Ca2+ release after SCR, and this contributes to AP prolongation. Prolongation of QT, Tpeak-Tend intervals, and left ventricular monophasic AP duration of beats after aftercontractions occurred before torsades de pointes in an in vivo dog model of drug-induced long-QT1 syndrome.
Conclusions: SCR contributes to increased BVR by interspersed prolongation of AP duration, which is exacerbated during IKs blockade. Attenuation of Ca2+-induced Ca2+ release by SCR underlies AP prolongation via increased ICaL. These data provide novel insights into arrhythmogenic mechanisms during β-adrenergic stimulation besides triggered activity and illustrate the importance of IKs function in preventing excessive BVR.
Enhanced cellular Ca2+ load, for example, during β-adrenergic receptor (βAR) stimulation, results in augmentation of Ca2+ release from the sarcoplasmic reticulum (SR), larger Ca2+ transients (CaT), and enhanced contractile force.1 Under certain conditions, Ca2+ load is increased beyond a certain threshold, leading to spontaneous Ca2+ release (SCR) from the SR during diastole.2 In turn, this diastolic leak causes a transient inward current (Iti), causing delayed afterdepolarizations (DADs) mainly attributable to activation of the electrogenic Na+-Ca2+ exchanger, with the Ca2+-activated Cl− current (IClCa) contributing in some species.3,4 Previous work has shown that both DADs and early afterdepolarizations (EADs) may share a common mechanism, at least during βAR stimulation and Ca2+ overload, namely SCR-induced currents.5 Both types of afterdepolarizations have been incriminated in the formation of ventricular tachycardia via triggered activity (TA) and by increasing dispersion of repolarization.6
Beat-to-beat variability (BVR) of repolarization duration occurs as an apparently random alteration in repolarization duration and can be observed at all levels from the action potential (AP) of the single cardiac myocyte to the QT interval on the body surface.7–9 Exaggerated BVR has been reported to be a more reliable indicator of arrhythmogenic risk than repolarization prolongation, per se, at least in several experimental ventricular tachycardia models10–12 and in selected human subjects.8,13
Although BVR has been investigated in multiple studies, the mechanisms underlying this phenomenon at the single-cell level remain to be fully elucidated. Pharmacological interventions influencing ion channels that operate during the AP plateau can markedly alter BVR.7,14 Despite the fact that inhibition of the slowly activating delayed rectifier K+ current (IKs) alone has minimal effects on both cellular AP duration (APD) and BVR,14 we recently have shown that during increased Ca2+ loading in myocytes subjected to blockade of IKs in combination with βAR stimulation, BVR is significantly enhanced, even before the occurrence of EADs and TA.14
In the present study, we investigated the relationship between SCR and BVR using a combined experimental and computational approach in both canine ventricular myocytes and in situ hearts subjected to βAR stimulation. We show that SCRs not only lead to Iti and DAD formation but also lead to a prolonged duration of AP via increased L-type Ca2+ current (ICaL), which in turn leads to increased BVR when analyzing multiple consecutive APs. Pharmacological interventions that inhibit SCR (either with reduced or with preserved systolic contraction) prevent this SCR-associated AP prolongation and reduce BVR.
This investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the United States National Institutes of Health (National Institutes of Health Publication 85-23, revised 1996). Animal handling was in accordance with the European Directive for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes (86/609/EU). Full details of methods, solutions, and interventions used are given in the online-only Data Supplement accompanying this article. A brief summary of the main aspects is provided.
Myocyte Isolation and Electrophysiology
Canine left ventricular (LV) myocytes were isolated as previously described.15 Transmembrane APs were recorded at ≈37°C using high-resistance (30–60 Mω) glass microelectrodes filled with 3 mol/L KCl. Myocyte contractions were recorded with a video edge motion detector.
We used the perforated patch-clamp technique under current-clamp or voltage-clamp control as previously described.16 Myocytes were stimulated to elicit APs (current clamp) or ICaL (voltage clamp). Changes in intracellular Ca2+ concentration ([Ca2+]i) were measured using Fluo-3 or Fura-2 AM.16
A recent model of the canine ventricular myocyte electrophysiology including βAR stimulation17 was extended with a method to induce diastolic SR Ca2+ release in 1 of 2 identical Ca2+ domains in a controlled fashion (Online Figures I and II). Simulations were performed in single cells as well as in a 1-dimensional homogeneous strand or 2-dimensional tissue of electrically coupled cells to determine the effects of diastolic Ca2+ release in the presence of electrotonic coupling.
In Vivo Dog Model of Drug-Induced Long-QT1 Syndrome
Torsades de Pointes (TdP) arrhythmias were induced in an in vivo dog model of long-QT1 (LQT1) syndrome as previously described.11 ECG, LV pressure, and LV monophasic APs (LV MAPs) were recorded throughout the experiments and analyzed offline.
Data Analysis and Statistical Comparisons
APD was quantified at 90% repolarization. BVR was quantified as variability of APD using the formula for 30 consecutive APs in the absence of TA.10 DADs were detected with a semiautomated method using a custom software script (online-only Data Supplement). When applicable, data were reported as mean±SEM of number experiments and significance was tested with either t test or 1-way ANOVA. Differences were considered statistically significant if P<0.05.
DADs Prolong Subsequent APs and Increase BVR
Under baseline conditions, a small beat-to-beat APD variability was observed (Figure 1A, left). βAR stimulation with isoproterenol (ISO) shortened APD and enhanced contraction (Figure 1A, middle). In the majority of cells, 100 nmol/L ISO caused DADs during pacing at 1000-ms cycle length (CL). APs after a DAD were prolonged, leading to increased BVR attributable to interspersed occurrence of normal and prolonged APs (Figure 1A, middle). Pharmacological inhibition of IKs by HMR1556 (500 nmol/L) in the presence of ISO further prolonged APD and increased BVR (Figure 1A, right). In separate experiments, CaT were recorded together with APs after HMR1556 and ISO. HMR1556 alone did not increase the CaT amplitude or lead to any SCR events. The combination of HMR1556 and ISO, however, led to large increases in CaT amplitude and SCR, as expected.18 APs after SCR were prolonged and had a reduced systolic CaT amplitude compared with those in which no diastolic release preceded the AP, indicating reduced SR Ca2+ load at the start of the AP (Figure 1B), consistent with the reduced cell shortening after an aftercontraction (Figure 1A).
APD prolongation after a DAD was significantly more pronounced in the presence of ISO plus HMR1556 compared with ISO alone at all CLs (eg, 41±6 ms vs 15±4 ms at CL=1000 ms; P<0.05; Figure 2A), and this coincided with a significant increase in BVR (17±2 ms vs 7±1 ms; P<0.05; Figure 2B). As such, HMR1556 was added in all subsequent experiments in the presence of ISO to maximize effects on APD, essentially creating a cellular model of drug-induced LQT1 syndrome.14 DADs showed typical rate-dependent properties, including increased amplitude and decreased AP–DAD coupling interval at shorter pacing CL, consistent with previous results.19 IKs blockade did not alter rate-dependent DAD properties (Online Figure III).
To determine whether DAD-like membrane-potential changes alone (ie, without underlying alterations in intracellular Ca2+) influenced the subsequent AP, current was injected to induce small alterations in the resting membrane potential (25-mV amplitude, 150-ms duration; Figure 2C, top). There was no significant difference between APDs in the absence or presence of these diastolic pulses (Figure 2D), independent of the timing of the pulse (not shown). In contrast, when these cells were stimulated with HMR1556 plus ISO, DADs occurred and subsequent APD was significantly prolonged (Figure 2C, bottom), indicating that alterations in the subsequent APD were influenced by DAD-related events other than membrane-potential alterations per se.
Coupling Between the Occurrence of DADs and EADs During β-Adrenergic Stimulation and IKs Blockade
In addition to DADs, EADs were observed. EADs occurred predominantly at 500-ms CL and exclusively in the presence of HMR1556 plus ISO. We observed early aftercontractions preceding EAD upstrokes by 21±5 ms (Figure 3A and 3C),5 consistent with the observations by Zhao et al20 in rabbit ventricular myocytes challenged with ISO and the ICaL agonist Bay K 8644, suggesting a role of SCR for EAD formation under these specific conditions.5 We quantified the occurrence of DADs in the beats preceding an EAD and found that DADs were significantly more likely to occur in the 2 beats directly preceding an EAD (53% and 95% of APs) compared with average DAD occurrence for all beats (34%; Figure 3A, bottom). Figure 3B shows the responses during ISO alone.
Blockade of Spontaneous SR Ca2+ Release Reduces BVR
We sought to further investigate how eliminating SCR would affect APD and BVR. Blockade of the ryanodine receptor (RyR) with ryanodine eliminated DADs and led to a drastic reduction in BVR, despite leading to an increased APD (Figure 4A and 4B). APD prolongation under these conditions was not associated with any arrhythmogenic events. A drastic reduction in systolic contraction was also seen, as expected, indicating that blockade of RyRs was achieved effectively.
Tetracaine reduces the open probability (Po) of RyR, leading to a decrease in the frequency of SCR in unstimulated rat ventricular myocytes21 and the abolishment of ISO-induced SCR in voltage-clamped rat ventricular myocytes.22 Similar to ryanodine, tetracaine led to an elimination of DADs in our experiments and a reduction in BVR at all CLs. However, unlike ryanodine, cell shortening remained largely unaltered, indicating that systolic Ca2+ release was largely intact (Figure 4C and 4D).
To determine the effects of increasing the Po of RyRs, as opposed to decreasing it, we used low concentrations of caffeine (maximally 500 µmol/L).23 Under these conditions, the percentage of APs showing DADs was increased at all CLs. Interestingly, this led to a significant decrease in BVR compared with HMR1556 and ISO alone, because DADs occurred before every AP (Figure 5, rightmost grey bars).
Several studies have shown that Ca2+/CaMKII is an important signaling molecule affecting Ca2+-induced Ca2+ release and that its inhibition can be antiarrhythmic.24 Curran et al25 demonstrated that SR Ca2+ leak during βAR stimulation is mediated by Ca2+/calmodulin-dependent protein kinase II (CaMKII). In our experiments, inhibition of calmodulin using W7 reduced DAD incidence in all cells (from 44% to 8% at 1000 ms CL; P<0.05) and completely abolished them in 3 of 7 cells. The decrease in DAD occurrence was paralleled by a decrease in BVR. Average APD was unaltered by W7, and systolic contractions remained intact. Application of the CaMKII inhibitor KN93 led to similar results (Figure 5, black bars).
Mg2+ has been shown to suppress both EADs and DADs in different models26,27 and is used clinically as a first-line agent against TdP. Increasing extracellular [Mg2+] from 1 mmol/L to 5 mmol/L abolished DADs induced after HMR1556 and ISO in 4 of 7 cells. In the remaining 3 cells, DAD incidence was significantly reduced. A concomitant reduction in BVR was seen. APDs were not significantly altered, and although systolic contractions were slightly reduced compared with beats that did not show SCR events before Mg2+ they remained stable, suggesting that the SR content remained constant (not shown). Similar results (albeit with reduced APD) were obtained with the class IC antiarrhythmic agent flecainide (Figure 5, striped bars), which has been shown to be effective in preventing catecholaminergic polymorphic ventricular tachycardia because of its effects on RyR and Na+ channels.28
Increases in [Ca2+]o Can Reinstate SCR Only When Residual RyR Function Is Available
We hypothesized that increases in [Ca2+]o after interventions that reduced DAD occurrence would lead to recurrence of SCR and DADs and increase BVR. Therefore, we raised the [Ca2+]o from 1.8 mmol/L to 3.6 mmol/L after application of ryanodine, tetracaine, or W7. This led to a reinduction of DADs after tetracaine or W7, and to a significant decrease of APD (Online Figure IVA–IVC). However, increasing [Ca2+]o after ryanodine did not lead to DADs, nor did it alter BVR, indicating that RyR function is required for SCR, and without this BVR cannot be altered. Interestingly, BVR was not significantly reincreased after W7 and high [Ca2+]o, whereas after tetracaine and high [Ca2+]o it was, despite DADs being reinstated under both conditions (Online Figure IVB and IVC). In the presence of W7, APs after SCR were no longer prolonged (Online Figure IVC, inset). These data suggest that calmodulin inhibition has a direct effect on the coupling between SCR and APD, possibly via ICaL.
Reduced Ca2+-Dependent Inactivation of ICaL Underlies APD Prolongation After SCR
We used a computational model of the canine ventricular myocyte17 to further investigate the ionic basis of the coupling between SCR and APD prolongation. After pacing to steady-state, a partial release of SR Ca2+ was induced during diastole. After this release, APD was prolonged, consistent with experimental observations (Figure 6A; 1000-ms CL in the presence of ISO). The prolonged APD was associated with a decreased CaT, an increase in the sustained component of ICaL, and increased IKs. The increase in IKs was attributable to the increased plateau potential resulting from the larger ICaL and it counteracted the repolarization delay. As such, APD prolongation was significantly larger when IKs was inhibited (Figure 6B). An increase in plateau Vm also was observed in experimental recordings (maximum amplitude of plateau was 114±1 mV without previous DAD and 118±1 mV after DADs at 1000 ms CL in the presence of HMR1556 plus ISO; P<0.05).
Several currents underlying the ventricular AP are modified by intracellular Ca2+. To determine the relative contribution of each component, we used the model to selectively inhibit each current for 1 final beat at steady-state after either a normal diastole or SCR. Inhibition of the main currents activated by SCR in the dog (INaCa and IClCa)3,29 only had a minor impact on the APD differences in the presence and absence of SCR (Figure 6C). In contrast, inhibition of ICaL or RyR was able to significantly reduce DAD-provoked APD prolongation. More specifically, when only Ca2+-dependent inactivation (CDI) of ICaL was inhibited, APD after SCR was no longer prolonged compared with APD without previous SCR (Figure 6C). Further investigation showed any differences in APD in these conditions were because of priming of CDI attributable to a transition of L-type Ca2+ channels to the Ca2+-dependent tier of the Markov model induced by the Ca2+ that was released during the SCR, before activation of inhibition.
Combined, our data implicate reduced SR Ca2+-release-dependent inactivation of ICaL as the mechanism underlying APD prolongation after SCR. This prediction was confirmed under voltage-clamp conditions in native canine ventricular myocytes when the extracellular solution was modified to isolate ICaL (Figure 7). During a voltage step to +10 mV from a holding potential of −80 mV, a significant increase in the integral of ICaL was observed after SCR: ∫ICaL=27.9±3.5 nC vs 33.0±3.5 nC in the absence or presence of SCR, respectively (P<0.05).
Relation Between Aftercontractions, Repolarization Prolongation, and Dispersion and Arrhythmogenesis in the Intact Canine Heart
To evaluate whether the single-cell mechanisms described above could affect arrhythmogenesis in vivo, we used a canine model of drug-induced LQT1.11 Application of a bolus of ISO during continuous HMR1556 infusion resulted in aftercontractions exclusively in the LV pressure signal before the development of ventricular premature beats and TdP (Figure 8A). In parallel with the occurrence of aftercontractions, we observed paradoxical QT prolongation during β-adrenergic heart rate acceleration, resulting in a significant increase of the QT interval in the last sinus beats before the extrasystoles triggering TdP (Figure 8B). Similarly, LV MAP duration was significantly prolonged in these beats and a significant increase in Tpeak-Tend interval was noted (Figure 8B), which could reflect an increased dispersion of repolarization. We focused on repolarization duration by comparing the beats before and after the first notable aftercontraction. QT, LV MAP, and Tpeak-Tend durations were significantly increased in the beat after the first aftercontraction, whereas the RR intervals were unchanged (Figure 8C). Furthermore, the ICaL inhibitor verapamil (0.4 mg/kg) prevented TdP induction by ISO during continuous HMR1556 infusion in 5 of 5 animals (Figure 8A), although some ventricular extrasystolic activity still could be observed (Online Figure VIII). Similar chronotropic responses were seen after ISO in the presence of verapamil but, in the absence of aftercontractions, QT, LV MAP, and Tpeak-Tend prolongation were significantly smaller when compared with HMR1556 and ISO alone (Figure 8B). These data illustrate the role of Ca2+-dependent regional repolarization prolongation for TdP induction in this model.
To further test the hypothesis that SCR at the single-myocyte level can contribute to the QT and Tpeak-Tend interval prolongation seen in vivo, we performed computer simulations of both a homogeneous 1-dimensional strand and a 2-dimensional tissue of electrically coupled myocytes. Consistent with our hypothesis, diastolic SR Ca2+ release increased repolarization duration and spatial dispersion of repolarization in both simulations (Figure 8D, Online Figures V–VI), and this effect was completely abolished by inhibition of ICaL CDI (Figure 8D, Online Figure VII).
In this study, we elucidated the relationship between SCR, AP prolongation, and BVR in canine LV myocytes and provide arguments for its arrhythmogenic significance in the in vivo beating heart. Our data indicate that under conditions of ISO-induced Ca2+ loading, SCR can occur over a wide range of pacing CLs and that APD after SCR is significantly prolonged. Reduced Ca2+-induced Ca2+ release-dependent inactivation of ICaL after SCR is involved in this AP prolongation. The increase in BVR was strongly dependent on the degree of APD prolongation after SCR. In anesthetized dogs subjected to similar conditions, we observed the occurrence of mounting aftercontractions in parallel with QT, Tpeak-Tend, and LV MAP prolongation just before TdP, which could be prevented by inhibition of ICaL and Ca2+ load with verapamil. SCR-related repolarization prolongation was reproduced in multicellular computer simulations. Our data align with recent data showing that local βAR stimulation synchronizes SCR in the normal rabbit heart, leading to Ca2+-mediated focal arrhythmia.30 In the setting of exaggerated spatio-temporal dispersion of repolarization, such focal activity may trigger TdP, as actually observed in the canine model of drug-induced LQT1 syndrome and βAR stimulation.11
Previous cellular studies31,32 have shown that large CaTs result in abbreviation of APD, whereas a small CaT after SR Ca2+ depletion corresponds to prolonged APD. In agreement, we found substantial APD prolongation after application of ryanodine because of reduced CDI of ICaL. Furthermore, ICaL CDI is sensitive to (partial) SR Ca2+ unloading by SCR. These changes in CDI are sufficient to modulate APD on a beat-to-beat basis over a wide range of CLs. Interestingly, Spencer and Sham reported the opposite effect in guinea pig ventricular myocytes, perhaps because of species differences in the balance of Na+-Ca2+ exchanger and ICaL.33 In the present study, we used Ca2+-sensitive fluorescent probes and cell shortening as Ca2+ indicators. Both measures reflect cytosolic Ca2+ levels, which are substantially different from the subsarcolemmal [Ca2+] influencing Ca2+-activated membrane currents.
The Importance of IKs Blockade
Burashnikov and Antzelevitch34 demonstrated that IKs block alone was insufficient to induce DADs or to modulate repolarization heterogeneity in canine transmural ventricular tissues.34 However, it amplified the effects of adrenergic stimuli. At the myocyte level, we previously have shown that IKs blockade (via KCNQ1 inhibition)35 has minimal effects on APD and BVR under baseline conditions. During βAR stimulation and IKs blockade, BVR is significantly increased and this is, at least partly, dependent on [Ca2+]i.14 Here, we extend this by determining an important role for CDI of ICaL, at least in the presence of SCR. Under baseline conditions, rapidly activating delayed-rectifier K+ current (IKr) is the main repolarizing current in canine ventricular myocytes and inhibition of IKr can cause EADs attributable to APD prolongation and subsequent ICaL reactivation.14 During additional βAR stimulation, the balance of the repolarization reserve is altered and the role of IKs becomes more prominent; enhanced IKs prevents repolarization instability and EAD generation by other proarrhythmic mechanisms (eg, IKr inhibition or augmentation of late INa). In contrast, inhibition of IKs during βAR stimulation leads to APD prolongation and increased BVR and EADs, indicating that under our conditions with βAR stimulation, IKs is a major contributor to repolarization. Bárándi et al36 previously have shown that the degree of APD prolongation induced by pharmacological block of repolarizing currents or augmentation of depolarizing currents depends on baseline APD. Consistent with these results, we found a more pronounced APD prolongation after SCR at slow CL and in the presence of IKs blockade. In our experiments, APD prolongation after SCR is particularly pronounced because the increase in ICaL also elevates the plateau potential, which increases IKs activation and offsets the prolongation induced by ICaL (Figure 6A, bottom). This compensating mechanism is absent when IKs is inhibited. Thus, IKs inhibition exacerbates the effect of SCR on APD prolongation.
The Role of SCR in Arrhythmogenesis
It has long been established that SCR-induced TA is a major arrhythmogenic mechanism during Ca2+ overload and its occurrence is increased in various pathological conditions.6,37 Here, we observed that diastolic SCR prolongs the subsequent APD, suggesting that even SCR below the threshold for TA has important electrophysiological effects that may be proarrhythmic. Regional prolongation of APD after SCR in the intact heart may cause increased spatial dispersion of repolarization between regions with the highest Ca2+ load (generating SCR) and regions with lower Ca2+ loading. In agreement with this hypothesis, we observed increased QT duration and Tpeak-Tend intervals after the first aftercontraction on a challenge with a bolus of ISO during IKs blockade in vivo. Regional heterogeneities in ion channel expression or intercellular conduction may further amplify this dispersion. Combined, these mechanisms can promote functional reentry. Stabilization of Ca2+ handling therefore may not only reduce the incidence of arrhythmogenic triggers but also prevent their reentrant perpetuation. Although current pharmacological interventions do not allow specific targeting of SCR or measurement of cellular Ca2+ handling in the in vivo dog heart, we found that ICaL inhibition with verapamil could prevent aftercontractions and TdP in our model of LQT1, and this was associated with a reduction in the paradoxical increase in QT and Tpeak-Tend intervals after ISO.
We also have shown that the combination of ISO and fast pacing can give rise to diastolic SCR, which prolongs APD sufficiently such that the next SCR occurs before the end of repolarization, generating an EAD. The common dependence of DADs and EADs on SCR in the presence of ISO previously has been described5,19 and is in agreement with recent findings in rabbit ventricular myocytes.20
Our results provide novel mechanistic insights on the coupling between SCR, APD prolongation, and EAD occurrence, and illustrate that diastolic SCR is a central element in both TA and repolarization instability. However, the ionic mechanisms of EAD generation, particularly the relative roles of INCX and ICaL, are complex and cannot be fully determined based on our data. Moreover, the measurement of the lag between the start of the aftercontraction and the EAD upstroke does not take into account the delay between SCR and activation of contraction or the SCR-induced slowing of repolarization in the priming phase before the EAD upstroke. This lag is likely because of differential Ca2+ thresholds for the activation of contraction vs the activation of membrane currents.
APD prolongation after SCR results from increased ICaL, which enhances Ca2+ loading via increased sarcolemmal Ca2+ influx. Enhanced loading along with reduced Ca2+ efflux serves to restore and fine-tune SR Ca2+ content to maintain Ca2+-induced Ca2+ release efficacy, as previously has been described in rat ventricular myocytes.31,32 However, during increased Ca2+ load, these mechanisms will readjust, promoting SCR, facilitating the occurrence of afterdepolarizations. Consistent with this, we found that the probability of observing a DAD was not significantly altered by the presence of a DAD on the previous beat. This strongly implies that under our experimental conditions, the reduction in myocyte Ca2+ load during a DAD is overcome by the increased sarcolemmal Ca2+ influx during the following prolonged AP. Thus, APD prolongation after SCR contributes to the vicious cycle of Ca2+ loading and, ultimately, overload. These results are in agreement with the recent modeling study by Morotti et al38 that established that ICaL predominantly inactivates because of CDI and that (strongly) reduced CDI can cause Ca2+ overload and DADs.
Inhibition of SCR by ryanodine and tetracaine previously has been described,19,22 and the results presented here agree with those data. We extend these observations by showing a concomitant decrease in BVR. Both ryanodine and tetracaine are useful for mechanistic studies, but because of their deleterious effects in vivo they cannot be used as therapeutic agents. In contrast, both magnesium and flecainide are commonly used antiarrhythmic agents. Here, we show that both these agents can lead to a reduction in BVR and arrhythmogenic events, most likely because of stabilization of Ca2+ handling in the single myocyte. Similarly, we confirm the usefulness of modulating Calmodulin/CaMKII during increased Ca2+ loading as an antiarrhythmic strategy.24 CaMKII phosphorylation has been shown to induce a different gating mode (mode 2) of the ICaL channel, thereby reducing inactivation and favoring Ca2+ entry and the occurrence of EADs and DADs.39 Interestingly, when CAMKII inhibition is applied, a disconnect appears between SCR events and APD. Whether this is attributable to alterations in CDI or because of effects on other CAMKII substrates, such as INaL, is not clear and further investigation is warranted. In this regard, the development of selective CaMKII modulators suitable for antiarrhythmic interventions in humans is awaited.
We have shown that after SCR, inactivating ICaL is increased and APD and QT intervals are prolonged, most likely because of reduced Ca2+-induced Ca2+ release-dependent inactivation of this current. The degree of APD prolongation is exacerbated by inhibition of IKs. This contributes to increased BVR and spatial dispersion of repolarization during ISO-induced diastolic Ca2+ release and, aside from DAD-mediated TA, may be an additional mechanism contributing to arrhythmogenesis. Pharmacological interventions that regularize SCR or inhibit SCR with or without preserved systolic contractions reduce BVR. Our data provide novel insights into arrhythmogenic mechanisms during increased Ca2+ loading.
The authors thank Drs Chris Pollard and Jean-Pierre Valentin, Department of Safety Pharmacology, Safety Assessment UK, AstraZeneca R&D, Alderley Park, United Kingdom, for active collaboration and for providing cardiac myocytes. Ongoing collaborations with Dr Yoram Rudy, Washington University, St. Louis, Missouri, and use of his computational resources are gratefully acknowledged.
Sources of Funding
P.G.A.V. is supported by a Vidi grant from the Netherlands Organization for Scientific Research (ZonMw 91710365). D.A. Eisner and A.W. Trafford are supported by the British Heart Foundation. D.M. Johnson was financially supported by AstraZeneca Ltd, United Kingdom.
In October 2012, the average time from submission to first decision for all original research papers submitted to Circulation Research was 12.5 days.
The online-only Data Supplement is available with this article at http://circres.ahajournals.org/lookup/suppl/doi:10.1161/CIRCRESAHA.112.275735/-/DC1.
Non-standard Abbreviations and Acronyms
- action potential
- action-potential duration
- β-adrenergic receptor
- beat-to-beat variability of repolarization duration
- Ca2+/calmodulin-dependent protein kinase II
- Ca2+-dependent inactivation
- cycle length
- delayed afterdepolarization
- early afterdepolarization
- long-QT syndrome type 1
- left ventricular
- LV MAP
- left ventricular monophasic action potential
- ryanodine receptor
- sarcoplasmic reticulum
- spontaneous Ca2+ release
- triggered activity
- torsades de pointes
- ventricular tachycardia
- Received August 8, 2012.
- Revision received November 5, 2012.
- Accepted November 13, 2012.
- © 2013 American Heart Association, Inc.
- Volders PG,
- Kulcsar A,
- Vos MA,
- Sipido KR,
- Wellens HJ,
- Lazzara R,
- Szabo B
- Ter Keurs HE,
- Boyden PA
- Zaniboni M,
- Pollard AE,
- Yang L,
- Spitzer KW
- Hinterseer M,
- Beckmann BM,
- Thomsen MB,
- Pfeufer A,
- Ulbrich M,
- Sinner MF,
- Perz S,
- Wichmann HE,
- Lengyel C,
- Schimpf R,
- Maier SK,
- Varró A,
- Vos MA,
- Steinbeck G,
- Kääb S
- Thomsen MB,
- Verduyn SC,
- Stengl M,
- Beekman JD,
- de Pater G,
- van Opstal J,
- Volders PG,
- Vos MA
- Gallacher DJ,
- Van de Water A,
- van der Linde H,
- Hermans AN,
- Lu HR,
- Towart R,
- Volders PG
- Volders PG,
- Sipido KR,
- Vos MA,
- Spätjens RL,
- Leunissen JD,
- Carmeliet E,
- Wellens HJ
- Katra RP,
- Laurita KR
- Zhao Z,
- Wen H,
- Fefelova N,
- Allen C,
- Baba A,
- Matsuda T,
- Xie LH
- Venetucci LA,
- Trafford AW,
- Díaz ME,
- O’Neill SC,
- Eisner DA
- Anderson ME
- Curran J,
- Hinton MJ,
- Ríos E,
- Bers DM,
- Shannon TR
- Myles RC,
- Wang L,
- Kang C,
- Bers DM,
- Ripplinger CM
- Trafford AW,
- Díaz ME,
- Negretti N,
- Eisner DA
- Spencer CI,
- Sham JS
- Bárándi L,
- Virág L,
- Jost N,
- Horváth Z,
- Koncz I,
- Papp R,
- Harmati G,
- Horváth B,
- Szentandrássy N,
- Bányász T,
- Magyar J,
- Zaza A,
- Varró A,
- Nánási PP
- Sipido KR,
- Volders PG,
- de Groot SH,
- Verdonck F,
- Van de Werf F,
- Wellens HJ,
- Vos MA
Novelty and Significance
What Is Known?
Beat-to-beat variability of ventricular repolarization (BVR) duration has been proposed as a more predictive marker of Torsades de Pointes (TdP) arrhythmia than repolarization duration alone. The cellular mechanisms of BVR remain to be fully elucidated.
During intense β-adrenergic receptor stimulation, Ca2+ load of the myocyte increases. Additional blockade of the slowly activating delayed-rectifier K+ current (IKs) amplifies this effect and exaggerates BVR.
Spontaneous sarcoplasmic reticulum Ca2+ release (SCR) during Ca2+ overload can cause membrane depolarization via Ca2+-dependent ion currents, resulting in afterdepolarizations and, potentially, triggered activity.
What New Information Does This Article Contribute?
In canine left ventricular myocytes, late diastolic SCR during β-adrenergic stimulation causes prolongation of the following action potential (AP), especially during IKs blockade. The irregular occurrence of SCR exaggerates BVR by interspersed AP prolongation. Pharmacological interventions that prevent or regularize SCR reduce BVR.
By sarcoplasmic reticulum unloading, late diastolic SCR decreases systolic sarcoplasmic reticulum Ca2+ release during the following AP. This reduces Ca2+-dependent inactivation of the L-type Ca2+ current, causing the AP prolongation.
In a canine model of drug-induced long-QT1 syndrome, β-adrenergic challenges cause paradoxical repolarization prolongation, exaggerated BVR, and aftercontractions just before the initiation of TdP. QT, monophasic AP, and Tpeak-Tend interval of the beats after aftercontractions are significantly prolonged. The Ca2+ antagonist verapamil prevents aftercontractions and arrhythmia.
A significant increase in BVR has been observed in animal models and selected human subjects before TdP onset. The in vivo mechanisms of this remain incompletely understood but reside partly in the cardiac myocyte. Cellular Ca2+ overload can augment BVR and plays a major role in arrhythmogenesis. Here, we aimed at elucidating Ca2+-dependent mechanisms of augmented BVR during β-adrenergic stimulation and IKs blockade, in effect mimicking long-QT1 syndrome. We show that APs prolong significantly after SCR, leading to increased BVR if the occurrence of SCR is irregular. An intact IKs prevents much of the repolarization prolongation. AP prolongation is driven by reduced Ca2+-dependent inactivation of L-type Ca2+ current resulting from decreased systolic sarcoplasmic reticulum Ca2+ release after SCR. Pharmacological interventions that prevent or regularize SCR reduce BVR, which suggests novel antiarrhythmic approaches against TdP under these conditions. We also demonstrate that during β-adrenergic stimulation and IKs blockade in vivo, paradoxical prolongation of repolarization and aftercontractions precede TdP, suggesting that the identified cellular mechanisms have relevance for arrhythmogenesis in the intact heart. Thus, myocardial Ca2+ overload and delayed and early afterdepolarizations not only can cause arrhythmias via triggered activity but also by increasing temporal and spatial dispersion of repolarization, promoting reentrant excitation.