Cardiovascular Research

Cardiovascular Research 2000 46(3):376-392; doi:10.1016/S0008-6363(00)00022-5
© 2000 by European Society of Cardiology

This Article Extract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Google Scholar Articles by Volders, P. G.A. Articles by Lazzara, R. Search for Related Content PubMed PubMed Citation Articles by Volders, P. G.A. Articles by Lazzara, R. Social Bookmarking          What's this?

Copyright © 2000, European Society of Cardiology

Progress in the understanding of cardiac early afterdepolarizations and torsades de pointes: time to revise current concepts

Paul G.A. Volders^a,*, Marc A. Vos^a, Bela Szabo^b, Karin R. Sipido^c, S.H.Marieke de Groot^a, Anton P.M. Gorgels^a, Hein J.J. Wellens^a and Ralph Lazzara^b

^aDepartment of Cardiology, Cardiovascular Research Institute Maastricht, Academic Hospital Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands
^bDepartment of Medicine, Cardiovascular Section, University of Oklahoma Health Sciences Center, and Department of Veterans Affairs Medical Center, Oklahoma City, USA
^cLaboratory of Experimental Cardiology, University of Leuven, Leuven, Belgium

^* Corresponding author. Tel.: +31-43-3875093; fax: +31-43-3875104 p.volders{at}cardio.azm.nl

Received 14 July 1999; accepted 25 January 2000

KEYWORDS Arrhythmia (mechanisms); Calcium (cellular); Long QT syndrome; Membrane potential; Impulse formation; SR (function)

	1 Introduction

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 
Afterdepolarizations are oscillations of the transmembrane potential that depend on the preceding action potential (AP) for their generation and can give rise to new APs when they reach a critical threshold for activation of a depolarizing current. This form of abnormal impulse generation is called ‘triggered activity’ [1].

Two types of afterdepolarizations have been distinguished: delayed (DADs) and early afterdepolarizations (EADs). DADs have been defined as "oscillations in membrane potential that occur after repolarization of an action potential" [2]. EADs are generated during the AP and have been defined as "oscillations at the plateau level of membrane potential or later during phase 3 of repolarization" [2]. Depending on the level of the membrane potential at which they are generated, EADs can trigger new APs that may appear as ectopic beats on the ECG. EADs can also augment electrical heterogeneity in regions of neighboring myocardium, which can lead to the formation of new APs via electrotonic interaction between areas that are still inexcitable and those that have already recovered from refractoriness [3]. Although the latter mechanism is reentrant rather than triggered activity, the occurrence of EADs is of pivotal importance for arrhythmogenesis under these circumstances. The clinical significance of EADs lies in their capacity to provide both the trigger (premature ectopic beats) and the substrate (electrical heterogeneity with nonuniform repolarization and refractoriness) for the initiation and perpetuation of torsades de pointes.

In this article, we discuss the evidence for a new concept of EAD formation, which includes an important role for cytoplasmic-[Ca²⁺]-dependent mechanisms, as schematically illustrated in Fig. 1. As a background, we will first review the recent literature on cellular Ca²⁺ homeostasis. Then, we introduce the classical view on EAD formation with a discussion of the contribution of window L-type Ca²⁺ current (I_CaL) or Na⁺ current (I_Na). Finally, the article will focus on [Ca²⁺]_cyt-dependent mechanisms for (at least some types of) EADs based on cellular, multicellular and in vivo experiments. These new concepts on EADs are then discussed in the clinical context of torsades de pointes.

View larger version (5K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 1 Schematic figure of the central theme of this article: [Ca2+]cyt-dependent mechanisms for EADs. A parallel is drawn between the generally accepted mechanism of [Ca2+]cyt-dependent ITI underlying DADs (left panel) and the primary role of [Ca2+]cyt-dependent inward current for the ‘conditioning phase’ underlying action-potential prolongation and EAD formation (right panel). For the upstroke phase of the EAD INa and/or ICaL are the major charge carriers.

 

	2 Cellular mechanisms of Ca2+ homeostasis and their relation to EADs

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 
2.1 Ca²⁺ homeostasis in normal cardiac cells
Ca²⁺ transients underlying excitation–contraction coupling in cardiac cells result mainly from Ca²⁺ release from the sarcoplasmic reticulum (SR) triggered by Ca²⁺ entry during the AP [4,5]. Under normal circumstances, Ca²⁺ entry into cardiac myocytes is carried primarily via I_CaL, whereas additional fractions can enter via T-type Ca²⁺ current (I_CaT) and on reverse mode Na⁺–Ca²⁺ exchange. All three pathways are capable of triggering SR Ca²⁺ release and contraction, but the relative contribution and efficiency is largest for I_CaL [6–12].

Modern concepts of excitation–contraction coupling rely on a ‘local-control theory’, which was originally developed from theoretical models to explain how global Ca²⁺ transients can be graded by the Ca²⁺ influx without uncontrolled regenerative release [13]. This theory was supported by morphological studies on the microarchitecture showing a close association between L-type Ca²⁺ channels and ryanodine receptors [14], and subsequently received experimental support from investigations with high-resolution confocal microscopy documenting local functional interaction between these channels [15,16]. It is now generally thought that cell-wide Ca²⁺ transients result from the spatial and temporal summation of local Ca²⁺ transients, termed ‘Ca²⁺ sparks’, i.e., Ca²⁺ release from a cluster of ryanodine receptors upon Ca²⁺ entry through single L-type channels. This fits well with the finding in previous work that the size of global Ca²⁺ transients was regulated by the magnitude of the whole-cell I_CaL [17,18]. In ventricular myocytes, a well-developed t-tubular system promotes homogeneous SR Ca²⁺ release throughout the cell [19]. The recent observation that Ca²⁺ sparks occur also in cardiac trabeculae under physiological conditions [20] indicates that the concepts obtained in studies on single myocytes are applicable to the function of ventricular tissue.

Ca²⁺ release from the SR is controlled by at least three regulators: (1) the magnitude of the triggering Ca²⁺ influx; (2) the state of the ryanodine receptors; and (3) the Ca²⁺ content within the SR. Recently, Eisner et al. [21] have convincingly argued, both experimentally and theoretically, that the state of the ryanodine receptors as a sole factor will not affect Ca²⁺ release from the SR. Indeed, changes in the gating of the ryanodine receptor and consequently in the size of the Ca²⁺ transient will feed back on transsarcolemmal Ca²⁺ fluxes and this feedback will adjust the SR Ca²⁺ content such that the size of the Ca²⁺ transient returns to its basal level. This process has been called ‘autoregulation’ and underscores the importance of the SR Ca²⁺ content. The SR Ca²⁺ content, and thus the amount of Ca²⁺ available for release, depends on: (1) Ca²⁺ uptake into the SR by the Ca²⁺ ATPase; (2) Ca²⁺ efflux (leak); (3) the Ca²⁺-buffering capacity of the SR; and (4) the [Ca²⁺]_cyt [22]. The Ca²⁺ content will affect release not only passively by its effect on the concentration gradient, but also by actively determining the fraction that will be released [23–25], suggesting that there may be a regulatory site for Ca²⁺ on the luminal (intra-SR) side of the ryanodine receptor.

Na⁺–Ca²⁺ exchange is the main Ca²⁺-extrusion mechanism for the beat-to-beat regulation of contraction and relaxation in cardiac myocytes [26,27]. Because of a stoichiometry of 3 Na⁺ ions:1 Ca²⁺ ion, inward current is generated when Ca²⁺ is extruded from the cell and outward current is generated when Ca²⁺ enters via this transporter. The direction and magnitude of Na⁺–Ca²⁺ exchange are dependent on the membrane potential and on the intra- and extracellular [Na⁺] and [Ca²⁺] in the direct vicinity of the exchanger protein [28–30]. The dynamics of the AP and the resultant local Na⁺ and Ca²⁺ signals change the reversal potential of the Na⁺–Ca²⁺ exchanger on a continuous basis. Under normal conditions, it functions predominantly in the inward mode during the decay phase of the Ca²⁺ transient generating inward current during most of the repolarization. This lengthens the AP duration (APD) [31–33].

2.2 Differences of [Ca²⁺] between the bulk cytoplasm and the subsarcolemmal space
Compelling evidence has accumulated that intracellular concentrations of Na⁺ and Ca²⁺ ions can differ considerably between the bulk cytoplasm and the subsarcolemmal space. The discussion on such concentration gradients was instigated in 1990 on the basis of experiments by Leblanc and Hume [6], who showed that reverse-mode Na⁺–Ca²⁺ exchange was capable of triggering Ca²⁺ release from the SR consequent to Na⁺ influx through tetrodotoxin-sensitive Na⁺ channels. Their data were interpreted as indicative of very high Na⁺ concentrations activating reverse Na⁺–Ca²⁺ exchange in a "functionally restricted intracellular space accessible to Na⁺ channels, the Na⁺–Ca²⁺ exchanger, and some of the SR", which was termed fuzzy space [34]. Carmeliet [35] reviewed the evidence for Na⁺ gradients in 1992 and new supportive data have since been published [36–38]. Not only Na⁺, but also Ca²⁺ gradients are present in cardiac cells. The fuzzy space is associated with the subsarcolemmal space in the diadic cleft, around the feet of the ryanodine receptors, and is some 10 nm wide. Within this restricted space, and with free diffusion of Ca²⁺ reduced by a factor of 5–7, Ca²⁺ fluxes through L-type channels and during Ca²⁺ release create large local gradients [39,40]. Experimental evidence for these gradients comes mostly from studies examining Ca²⁺-dependent membrane currents. Given that during voltage clamp I_Na–Ca is a linear function of the intracellular Ca²⁺ concentration [41], several groups have shown that during Ca²⁺-induced Ca²⁺ release from the SR the Ca²⁺ concentration sensed by the membrane rises and falls more quickly, and is higher than the bulk [Ca²⁺]_cyt [42,43]. This hysteresis is further accentuated during cellular Ca²⁺ loading and spontaneous Ca²⁺ release from the SR [43,44]. Inactivation of L-type Ca²⁺ current during Ca²⁺ release from the SR also exceeds the inactivation expected on the basis of cytoplasmic Ca²⁺ signals [45,46]. More direct measurements of local Ca²⁺ by membrane-bound Ca²⁺ indicators have been hampered by methodological problems [47], although in smooth muscle such gradients have been demonstrated [48].

Until now, the importance of functional intracellular Ca²⁺ gradients for the generation of EADs and DADs has received little attention in the literature (but see Trafford and co-workers [44,49]). However, based on the aforementioned data local Ca²⁺ dynamics and resultant activation of I_Na–Ca and other Ca²⁺-sensitive membrane currents may be of particular relevance.

2.3 Spontaneous Ca²⁺ release from the SR and mechanisms of DADs
Spontaneous Ca²⁺ release is usually observed when the Ca²⁺ content of the cells, and of the SR, is very high, a condition often referred to as ‘Ca²⁺ overload’. Experimental interventions as diverse as rapid pacing, prolonged depolarization during voltage clamp, high [Ca²⁺] or low [K⁺] in the superfusate, cardiac glycosides, β-adrenergic receptor stimulation (and many more) produce Ca²⁺ loading via direct or indirect modulation of I_CaL or Na⁺–Ca²⁺ exchange. All these interventions have in common that the peak [Ca²⁺]_cyt and the amplitude of the Ca²⁺ transient are increased during subsequent APs or voltage clamps. Depending on the nature and the intensity of stimulation, the baseline [Ca²⁺]_cyt during rest (‘diastolic [Ca²⁺]_cyt’) may vary from virtually unaltered to significantly upregulated.

Based on contraction measurements in rat ventricular myocytes, Capogrossi et al. [50] suggested that the first occurrence of spontaneous SR Ca²⁺ release denotes the state of maximal inotropy in the myocardium. However, in canine ventricular myocytes we found that the twitch potentiation during administration of isoproterenol was not maximal at the first manifestation of aftercontractions, indicating that the amount of SR Ca²⁺ accumulation necessary for spontaneous Ca²⁺ release is lower than that for maximal inotropy [51]. Species-related differences in SR Ca²⁺ handling may explain these paradoxical findings: in the rat, unlike the dog, the SR works at a near-maximal Ca²⁺ sequestration, which is evident from the frequent spontaneous Ca²⁺ releases at rest [52–55].

Evidence for the crucial role of SR Ca²⁺ content for spontaneous release also comes from the studies of Ca²⁺ sparks. Cheng and co-workers [52,56] demonstrated that spontaneous sparks occurred only at low rates under normal-loading conditions (0.0001/s). However, when Ca²⁺ overload was produced by high extracellular [Ca²⁺] (10 mmol/l), the rates increased around 4-fold, whereas the sparks’ amplitude and size increased 4.1- and 1.7-fold, respectively. Fusion of the local Ca²⁺ transients could set off global Ca²⁺ waves that propagated through the cell by recruiting more Ca²⁺ sparks along the wave front [52,56].

The exact mechanisms of spontaneous Ca²⁺ release are still under study. Several groups have now established that with an increase in the Ca²⁺ content of the SR, a larger fraction of the Ca²⁺ pool will be released [23–25], suggestive of a regulation of release by luminal Ca²⁺. Spontaneous release could result from an increased sensitivity of the release channel to a small increase in diastolic [Ca²⁺], either from a small Ca²⁺ influx across the sarcolemma, or due to a passive leak from the SR, perhaps related to saturation of the buffering capacity of the SR. Alternatively, spontaneous release could be unrelated to the normal Ca²⁺-induced Ca²⁺ release mechanism, as suggested by Fabiato [5], and perhaps be due to altered properties of the ryanodine receptor.

Independent of the underlying mechanism, Ca²⁺ imaging in cardiac myocytes has revealed two patterns of spontaneous release: (1) focal Ca²⁺ release with no or limited propagation along the SR; and (2) Ca²⁺ release propagating as a wave through the entire cell (e.g., see Refs. [52,56–61]). Numerous groups have investigated the interplay between these types of Ca²⁺ release, and arrhythmogenic transient inward current (I_TI) and DADs (e.g., see Refs. [49,62–70]). For focal release and Ca²⁺ waves, however, quantitative interpretation of such data remains difficult since membrane currents always represent a spatial average.

It is generally accepted that DADs are caused by I_TI evoked by spontaneous Ca²⁺ releases from the SR under conditions that favor accumulation of [Ca²⁺]_cyt and cellular Ca²⁺ overload [2,62]. These sudden increases in [Ca²⁺]_cyt modulate Ca²⁺-sensitive membrane currents via channel- or transporter-specific mechanisms. Three different ionic currents, alone or in combination, have been implicated in the generation of I_TI since its first description in 1976 [71]. These are I_Na–Ca [49,68,70,72–77], non-selective cation current (I_NS) [72,78–81], and Ca²⁺-activated Cl^– current (I_Cl(Ca)) [49,68,76,77].

Kass et al. [72] were among the first to speculate on the ionic basis of I_TI and proposed that this current, induced by strophanthidin in calf cardiac Purkinje fibers, might be carried by a ‘leak channel’ or that it might reflect "Ca²⁺ extrusion by an electrogenic Ca–Na exchange". The contribution of I_Na–Ca to I_TI has been estimated by comparing the current–voltage relationships of the two and by determining the characteristics of I_Na–Ca and I_TI during spontaneous Ca²⁺ release from the SR. The reversal potential of the Ni²⁺-sensitive I_Na–Ca in guinea-pig ventricular myocytes increased linearly with the log [Ca²⁺]_cyt [29] and high [Ca²⁺]_cyt during spontaneous Ca²⁺ release would thus shift the reversal potential of I_Na–Ca towards very positive membrane potentials. The current can then be inward at voltages relevant for the AP, as shown by Lipp and Pott [74], who concluded that I_Na–Ca was the major charge-carrying mechanism for I_TI. This conclusion was supported by Benndorf et al. [75].

2.4 Ionic mechanisms of EADs
2.4.1 Role of window I_Na and I_CaL
The appearance and mechanisms of EADs are diverse as they occur in various conditions. EADs present as transient retardations or reversals of the repolarization during phases 2 and 3 of the AP, often in the setting of AP prolongation [82]. Although the term ‘early afterdepolarization’ per se would not comprise a mere retardation of the repolarization, this is actually the priming event of the EAD formation, and must be incorporated into any theoretical mechanistic framework on EADs. At the ionic level, a decrease in outward current(s), an increase in inward current(s), or combinations of the two can unbalance the repolarization such that EADs arise. However, due to the very complex dynamics and overlapping of transsarcolemmal ionic fluxes during the AP, the impact of individual current alterations on repolarization characteristics cannot easily be estimated. In the congenital long-QT syndromes, one can pinpoint single current defects as the primary event [83]. However, in most conditions a more complex situation exists. Experimental electrophysiological studies on EADs have often relied on pharmacological agents and slow pacing modes to cause the initial action-potential prolongation via reductions of I_Kr [84–87], or slowed inactivation of I_Na [87–90], and I_CaL [91,92]. The major emphasis was on the currents generating the actual upstroke of the EAD and convincing arguments were found for the role of reactivating I_Na and I_CaL [93–95].

The significance of I_CaL for EADs is due to its voltage-dependent activation falling in the range of membrane potentials (–35 to 0 mV) where repolarization is delayed and EADs arise. Based on experiments with Bay K 8644, January and co-workers [91,92,96,97] concluded that the induction of EADs requires: "(i) a conditioning phase controlled by the sum of membrane currents present near the AP plateau and characterized by lengthening and flattening of the plateau within a voltage range where, (ii) recovery from inactivation and reactivation of L-type Ca²⁺ current to carry the depolarizing charge can occur" [92]. In the presence of pharmacological agents (e.g., Bay K 8644 or ATX-II) or channel modifications that slow inactivation of I_CaL or I_Na, these currents are obviously of prime importance. However, in the absence of such conditions, inward I_Na–Ca is likely to play a significant role in the initial delay in repolarization designated ‘conditioning phase’ [92] or ‘conditional phase’ [98]. The conditioning phase appears as the transient delay in repolarization that can, but does not have to be followed by an EAD upstroke or triggered AP. Based on our own work and accumulating evidence in the literature, a primary role for Na⁺–Ca²⁺ exchange is not only restricted to the conditioning phase of EADs related to Ca²⁺ overload and spontaneous Ca²⁺ release from the SR, but applies also to EADs that occur in the setting of less extreme Ca²⁺ loading.

2.4.2 Evidence for [Ca²⁺]_cyt-dependent mechanisms of EADs
Major observations support the contribution of [Ca²⁺]_cyt-dependent mechanisms to EADs. First, dependent on the mode of induction, EADs can appear jointly with DADs in consecutive APs or even in relation to the same AP (Fig. 2). A second argument derives from the observation that several types of EADs are induced or amplified by fast pacing rates or rate acceleration during bradycardia. Also, the influence of increased adrenergic stimulation can induce EADs via sudden changes of cellular Ca²⁺ homeostasis. Clinically, increased adrenergic tone and sudden rate accelerations are often involved in the initiation of torsades de pointes. A third argument is that experimental interventions known to attenuate Na⁺–Ca²⁺ exchange can also inhibit EADs under certain conditions.

View larger version (12K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 2 Spontaneous Ca2+ release from the SR underlying both EADs and DADs in canine ventricular myocytes exposed to 10–20 nmol/l isoproterenol at a pacing cycle length of 1000 ms. (A) [Ca2+]cyt activity measured ratiometrically (I340/380; lower trace) after loading with fura-2 AM (excitation at 340 and 380 nm). Three distinct peaks of [Ca2+]cyt accompany the second AP: (1) diastolic spontaneous Ca2+ release (i.e., delayed Ca2+ aftertransient) underlying DAD (*); (2) twitch-related Ca2+ release; and (3) systolic spontaneous Ca2+ release (i.e., early Ca2+ aftertransient) underlying EAD. Arrows indicate start of early Ca2+ aftertransient and same time point on AP, well before EAD upstroke. (B) Mechanical corollaries of (spontaneous) [Ca2+]cyt activity in a different cell: delayed aftercontraction, twitch contraction and early aftercontraction. The start of the early aftercontraction (arrow to upper trace) coincides with the initial delay in repolarization (reversal of dV/dt; arrows to middle and lower traces), well before EAD upstroke. *, DAD. dV/dt of the action-potential upstrokes are artificially truncated. Reproduced from Ref. [51].

 
In Table 1, experimental conditions associated with the generation of both EADs and DADs in transmembrane and/or monophasic-action-potential (MAP) recordings are categorized. These include: external solutions without K⁺ or with CsCl, administration of digoxin or Bay K 8644, reoxygenation after anoxia, and - and/or β-adrenergic receptor stimulation. Their impact on the Ca²⁺ homeostasis and AP configuration of myocardial cells is obviously diverse and would tentatively be linked to either EADs or DADs only, based on classical experiments. However, the copresence of both afterdepolarizations, as demonstrated in referred original articles, suggests common mechanisms for the two at least under those circumstances. Table 1 also indicates that concomitant factors such as pacing rate and APD have a typical involvement in the generation of EADs. Various types of EADs are evoked during fast pacing under conditions when the preceding APD is not or only slightly prolonged.

View this table: [in this window] [in a new window]   Table 1 Experimental conditions associated with the generation of delayed (DADs) and early afterdepolarizations (EADs) in monophasic (MAP) and/or transmembrane action-potential (TAP) recordings

 
We will focus on the EADs induced by isoproterenol. During intense β-adrenergic receptor stimulation, arrhythmogenic responses in ventricular myocytes are accompanied by spontaneous Ca²⁺ release from the SR during both diastole and systole (Fig. 2A), often in conjunction with the same AP [63,64,99]. The systolic spontaneous releases appear as early Ca²⁺ aftertransients following the (twitch-related) I_CaL-induced Ca²⁺ release and they have mechanical corollaries: early aftercontractions (Fig. 2B) [51,100]. Early Ca²⁺ aftertransients and their aftercontractions rise significantly earlier than the EAD upstrokes, and they coincide with an initial delay in repolarization (i.e., the conditioning phase) preceding the upstrokes (Fig. 2). In other words: early Ca²⁺ aftertransients and aftercontractions are not secondary to the currents that induce the EADs, rather they are the primary events. We have described that isoproterenol-induced EADs are completely suppressed by the cation Ni²⁺, whereas their accompanying early aftercontractions remain demonstrable [51]. These observations strongly suggest the contribution of spontaneous SR Ca²⁺ release and inward I_Na–Ca to the generation of this type of EADs. Priori and Corr [101] noted similar suppressive actions on EADs by perfusion with low-[Na⁺]_o solution or treatment with the (aspecific) Na⁺–Ca²⁺ exchange blocker benzamil.

The simulation study of Zeng and Rudy [98] on isoproterenol-induced EADs did not incorporate a secondary systolic rise of [Ca²⁺]_cyt and for that reason must have underestimated the contribution of inward I_Na–Ca. The Luo–Rudy model [98,102], while the most comprehensive to date, may have deficiencies with respect to all the factors operative in SR Ca²⁺ handling. Spontaneous Ca²⁺ release may not be accurately predicted.

Using fluorescent Ca²⁺ imaging, two groups [65,66,103] noted that the spatial features of Ca²⁺ transients associated with afterdepolarizations were different for DADs (heterogeneous pattern indicating focal, spontaneous SR Ca²⁺ release) versus EADs (homogeneous pattern suggesting I_CaL-induced Ca²⁺ release). This is in line with the concept of different ionic mechanisms underlying DADs and EADs. However, the same groups also emphasized that EADs can consist of different types with different etiology; late EADs induced by perfusion with [K⁺]_o-free solution [103] or isoproterenol [65] were associated with the heterogeneous [Ca²⁺]_cyt pattern akin to DADs.

Next, we analyze in detail the so-called late EADs initiating at membrane potentials more negative than the range at which I_CaL can be activated [104–106]. Although conceptually I_Na, I_CaT or I_NS could underlie these late EADs, there is increasing evidence that inward I_Na–Ca has the primary role. Interventions that cause cellular Ca²⁺ loading (such as the combined application of CsCl and epinephrine) augment these EADs [105–107]. In Purkinje fibers or ventricular myocytes, relatively low concentrations of these agents do not generate EADs when applied separately. The explanation for late-EAD generation with combined application is in increasing the APD by Cs⁺ and increasing SR Ca²⁺ loading and release by epinephrine. Inward I_Na–Ca is the main current responsible for the initiation of afterdepolarizations as demonstrated by the inhibitory effects of high extracellular [Ca²⁺] or ouabain (the latter being due to the effect of increased [Na⁺]_cyt on the exchanger).

Finally, we focus on the conditioning phase of EADs induced during congenital or acquired inhibition of I_Kr and/or I_Ks. We recall that the typical example of the congenital form is the long-QT syndrome, whereas acquired forms are seen during treatment with most, if not all, drugs that prolong repolarization. In the clinic, these syndromes are associated with an increased regional dispersion of ventricular repolarization, generation of EADs, premature ventricular beats and a serious risk of torsades de pointes (e.g., see Refs. [108–113] (for congenital long-QT syndromes) and [110,112,114–117] (for acquired forms)). Most often, torsades de pointes is precipitated by sudden accelerations or short–long–short cycle lengths during bradycardia [118–121], whether or not under the influence of an increased sympathetic tone [108]. The influences of increased sympathetic tone, rate accelerations and short–long–short sequences are consecutively addressed.

Patterson et al. [122] reported that the elicitation of slow-rate-dependent EADs during administration of the class III antiarrhythmics D,L-sotalol and clofilium in canine cardiac Purkinje fibers was only observed after the additional administration of epinephrine at 10–100 nmol/l (Fig. 3A). In the case of clofilium, fast-rate-dependent DADs were also seen. Epinephrine alone did not produce afterdepolarizations at these concentrations. The authors showed that inward I_Na–Ca contributed to the generation of EADs because a temporary increase of the extracellular [Ca²⁺], which attenuates inward I_Na–Ca [30], suppressed the EAD formation [122]. The inhibition and subsequent transient stimulation of inward I_Na–Ca was also obtained by substituting LiCl for NaCl, followed by the return to normal extracellular [Na⁺] [123]. Whereas EADs had been absent at baseline (class III agents only), this change of solutions potentiated EAD formation upon reintroduction of the normal [Na⁺] with EAD take-off potentials being more negative than –60 mV. These results indicated the important contribution of inward I_Na–Ca to class III EAD under circumstances relevant for the in vivo situation.

View larger version (12K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 3 [Ca2+]cyt-dependent mechanisms of EADs in vitro. (A) Late EADs arising from membrane potentials more negative than –35 mV during administration of epinephrine in a clofilium-pretreated canine Purkinje fiber. In the upper panel APs at baseline are depicted. In the middle panel, AP prolongation by clofilium is shown. No EADs or DADs are present. In the lower panel, upon the additional administration of epinephrine at low concentration, pronounced formation of EADs is present at cycle lengths greater than 1000 ms. In these recordings the cycle length is 2000 ms. Reproduced from Ref. [122]. (B) Acceleration-induced EADs. In the left panel, APs are recorded from canine LV M-tissue, treated with the IKr blocker E-4031, during a sudden increase of the rate from a cycle length of 2000 to 1000 ms. The transient appearance of EADs is accompanied by AP prolongation. In the right panel, the APs are superimposed, showing the last beat at the cycle length of 2000 ms, and some of the first nine APs at 1000 ms. Reproduced from Ref. [125].

 
Proarrhythmic effects of sudden rate acceleration have been related to the distinct electrical characteristics of midmyocardial (M) cells in the transmural ventricular myocardium. Sicouri et al. [124] found that rate acceleration in the presence of D-sotalol led to a paradoxical transient prolongation of the AP and the occurrence of EADs in left ventricular (LV) M-cells, but not in endo- or epicardial cells. This behavior caused a temporary increase in the transmural dispersion. Using the I_Kr blocker E-4031, Burashnikov and Antzelevitch [125] showed that sudden accelerations from initially slow pacing rates caused transient EAD activity if none existed before or resulted in augmentation of preexisting EADs in the majority of canine M-tissues (Fig. 3B) and to a lesser degree in Purkinje fibers. A large transmural dispersion of repolarization resulted from the absence of such effect in the endo- and epicardial layers. The investigators applied ryanodine, flunarizine and low extracellular [Na⁺] to examine the possible involvement of cellular Ca²⁺ loading and increased activity of the Na⁺–Ca²⁺ exchanger. In support of such a role, all three interventions abolished the acceleration-induced EADs and the APD prolongation [125]. With regard to the cellular correlates of the in vivo short–long–short sequence, they also reported on the ability of single premature ventricular beats to induce EADs and AP prolongation in M-cells [125]. In line with an earlier study [126], this potentiating effect could be explained by increased Ca²⁺ transients accompanying the post-extrasystolic beats, which fitted well with a central role for the Na⁺–Ca²⁺ exchanger. Similar interpretations were made by Viswanathan and Rudy [127] for pause-induced EADs in a simulation study on the initiation of torsades de pointes. Single pauses interrupting the steady-state pacing rates caused prolongation of the post-pause AP due to a smaller I_Ks (more deactivation) and an enhanced I_Na–Ca (larger post-pause Ca²⁺ transient). This delayed repolarization set the stage for reactivation of I_CaL to depolarize the membrane and generate the EAD upstroke [127].

	3 EADs and arrhythmogenesis in multicellular preparations and the intact animal heart

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 
In this part of the review, we will focus on EAD-related triggered activity in multicellular preparations and the intact animal heart. Because we will concentrate on the literature of the 1990s, the emphasis will be on direct evidence (microelectrodes or MAP catheters) concerning: (1) the occurrence and propagation of acquired EADs, and their consequences for arrhythmogenesis; (2) circumstances where EADs and DADs occur in the same experiment; (3) other circumstances where EADs appear to be linked to myocardial Ca²⁺ overload.

3.1 EADs in multicellular preparations: consequences for arrhythmogenesis
In Purkinje muscle preparations [85,94,128], a number of studies demonstrated more pronounced changes of the APD in response to either bradycardia and/or repolarization-delaying drugs in Purkinje fibers than in ventricular muscle. Under these specific circumstances, a marked dispersion of repolarization is often observed and is an important determinant of arrhythmogenesis. Furthermore, there is an increased incidence of EADs in Purkinje fibers, which can give rise to propagated ectopic beats on the basis of: (1) direct triggering of EADs in Purkinje fibers (Fig. 3A), most likely from phase-3 EADs; or (2) prolonged repolarization-dependent reexcitation. The capacity for reexcitation will depend on the difference in membrane potential between connected cells, but also on the excitability threshold of the already-repolarized cells.

Using canine LV-wedge preparations, Antzelevitch and co-workers [86,87,129,130] found similar results with the M-layer as the responsible cell type. The APD and EAD sensitivity of M-cells for bradycardia and/or class III agents was significantly higher than of the surrounding endo- and epicardium. It has to be stated that M-cells are not the electrophysiological or pharmacological equivalent of Purkinje fibers, but their responses to bradycardia and certain drugs are quite similar. Microelectrode recordings in LV-wedge preparations showed that transmural dispersion (developing through nonhomogeneous responses of the M-layers to chromanol 293B plus isoproterenol [130], D-sotalol [87], and ATX-II [87]) was the prerequisite for spontaneous or pacing-induced polymorphic arrhythmias, mimicking torsades de pointes in the intact heart. The initiating beats of the spontaneous arrhythmias most frequently seemed to originate from the deep subendocardium (either in Purkinje or in M-cells).

3.2 Simultaneous appearance of EADs and DADs in the intact heart
To the best of our knowledge, Patterson et al. [131] were the first to describe that EADs and DADs can appear together in the same intact heart, albeit at different time points after the administration of intravenous CsCl. At the first instance, Cs⁺ caused AP prolongation with the appearance of EADs. Thereafter, the MAP duration started to decrease and DADs appeared. Surprisingly, the coupling interval of the afterdepolarizations remained similar, indicating that depending on the APD an EAD (long APD) or a DAD (short APD) could occur. It was also this group that suggested that these EADs and DADs could be based on similar cellular mechanisms via alterations of myocardial Ca²⁺ homeostasis [131]. Since 1990, more publications have indicated the simultaneous appearance and common etiology of certain EADs and DADs in the intact heart [132,133]. Using floating microelectrodes in cats, Xie and Xie [134] again reported the occurrence of EADs and DADs at different time points after administration of CsCl. Most EADs arose during phase 3 of the transmembrane and monophasic AP. These investigations provided evidence that the afterdepolarizations recorded in transmembrane APs could be visualized simultaneously by MAP recordings. In a study by Xu et al. [133], it was shown in guinea pigs with the use of MAP recordings from the midmyocardium that digoxin induced DADs and phase-3 EADs, which were both suppressed by verapamil. Some types of EADs converted to DADs in the MAP after administration of the K⁺-channel opener pinacidil, which shortens the APD. These authors concluded that "late phase 3-EADs generated under conditions of Ca²⁺ overload and DADs share similar properties" [133].

We have described that chronic complete atrioventricular block (AVB) in the dog results in: (1) electrical remodeling (nonhomogeneous action-potential prolongation); (2) an enhanced contractile performance at the slow idioventricular heart rate; and (3) biventricular hypertrophy [135]. In some of these (anesthetized) dogs, short pacing trains can induce both EADs and DADs, and related ectopic beats under baseline conditions. A typical example is shown in Fig. 4A. Ryanodine abolishes both these EADs and DADs (not shown), suggesting a common cellular mechanism for the two. In experiments with class III antiarrhythmic drugs at clinically relevant doses, the dogs show the appearance of EADs in MAP recordings. As illustrated in Fig. 4B, these EADs can occur at the end of the plateau (phase 2) and during phase 3 of the repolarization, often in consecutive beats. More importantly, EADs and related ectopic beats can be induced by fast pacing modes leading to the induction of torsades de pointes. In correspondence with the in vitro experiments by Burashnikov and Antzelevitch [125], we further find that sudden increases in pacing rate (e.g., from a cycle length of 1500 to 1000 ms) cause the emergence or accentuation of EADs, which are transient and usually restricted to a period of seconds to 1 min of the new rate (Fig. 4C). These observations are not in contrast with the known APD adaptation to heart-rate increases during pacing or adrenergic activation, but rather uncover the delicate balance of ventricular repolarization in this animal model, which needs time to adapt to a new steady state.

View larger version (41K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 4 In vivo appearance of EADs and DADs in anesthetized dogs with chronic AVB. In all three panels, the ECG lead II and a LV endocardial MAP recording are shown. (A) DAD (closed asterisk), EADs (open asterisks), and triggered activity in consecutive MAPs after a pacing train of eight stimuli with an interstimulus interval of 300 ms under baseline conditions. Paced beats are indicated by S. Just before pacing the idioventricular cycle length is 1360 ms, increasing to 1990 ms thereafter. The first beat post-pacing shows EAD formation and a triggered AP on the MAP signal. (B) Induction of phase-2 and phase-3 (late) EADs in consecutive beats during administration of the class III antiarrhythmic agent ibutilide. The occurrence of EADs follows a MAP prolongation from 375 ms at baseline (PRE-DRUG) to 555 ms during ibutilide and is related to the triggering of multiple ectopic beats (as shown on the right). (C) Acceleration-induced accentuation of EADs during a sudden increase in the pacing rate from a cycle length of 1500 to 1000 ms. The emergence of EAD upstrokes is a transient phenomenon upon the rate change, which fades away within 1 min thereafter.

 
3.3 EADs occurring in conditions associated with myocardial Ca²⁺ overload
In the dog with chronic AVB and acquired QT prolongation, a relation between myocardial Ca²⁺ overload and the generation of class III-dependent EADs and torsades de pointes has been suggested by the attenuating (EADs) and preventive (pacing-induced torsades de pointes) effects of the agents ryanodine and flunarizine [136]. Similarly, in a rabbit model of torsades de pointes (co-administration of methoxamine and almokalant), Carlsson et al. [137] showed the prevention of spontaneous torsades de pointes when the rabbits were pretreated with nisoldipine and flunarizine.

	4 Clinical relevance of EADs in the congenital and acquired long-QT syndromes

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 
Early electrophysiological studies with recordings of endocardial electrograms and MAPs in patients with congenital and acquired long-QT syndromes implicated EADs as contributors to the T-U waves and generators of arrhythmias [138–141]. Later investigations with MAP recordings in humans indicated the appearance of EADs that were enhanced at longer cycle lengths and with adrenergic stimulation, and were diminished by β-adrenergic receptor blockers, Ca²⁺-channel blockers, and K⁺-channel openers [108,109,111,113,115–117,142–147]. Because EADs are related to the congenital and acquired long-QT syndromes, the recording of MAPs in patients is of help in the understanding of the pathogenesis, the influence of EADs on the surface ECG, and the effects of treatment.

4.1 Congenital long-QT syndromes
In congenital long-QT patients, the therapeutic benefit of β-adrenergic receptor blockade, left cardiac sympathectomy, and Ca²⁺-channel blockers implicates Ca²⁺ entry and Ca²⁺ loading as the triggering mechanism, but does not allow the identification of I_CaL or I_Na–Ca as the major current generator for EADs. Reduction of I_CaL by these therapeutic measures would also result in reduction of Ca²⁺ loading and Na⁺–Ca²⁺ exchange, and prevent spontaneous SR Ca²⁺ release. In all reports of EADs shown in MAPs in both congenital and acquired long-QT syndromes, these membrane responses occurred during phase 3 of the AP. In some studies EADs also appeared during phase 2, but when triggering was observed, it was observed to be timed with phase 3 [116,142,143]. While MAPs do not allow precise determination of the membrane potential at which the EADs originate, the relative location on phase 3 strongly suggests an initial potential that is often more negative than –35 mV, excluding a role for I_CaL in those cases. These observations implicate I_Na–Ca rather than I_CaL as a common current generator for EADs and triggering. An example of MAPs recorded in a patient showing late phase-3 EADs is shown in Fig. 5.

View larger version (16K): [in this window] [in a new window] [Download PowerPoint slide]   Fig. 5 Phase-3 EADs and triggered ventricular extrasystoles in a 65-year-old male patient with an idiopathic, non-familial, pause-dependent and adrenergic-dependent long-QT syndrome. Shown are the ECG lead V1, a MAP recording from the right ventricular outflow tract, and a bipolar electrogram from the right ventricular apex (RV). Both panels begin with the last three complexes of a train of eight beats during right-ventricular pacing at a cycle length of 600 ms. (A) Post-pacing pause of 1340 ms results in a prominent post-pause EAD in the MAP recording (left arrow) and a single triggered ventricular extrasystole. (B) Longer pause of 1430 ms results in a larger EAD (left arrow) and a run of four ventricular extrasystoles. The patient was not receiving any medication known to prolong repolarization and serum electrolytes were normal. Reproduced from Ref. [108].

 
Other arrhythmia-initiating mechanisms should be considered. Experimental evidence supporting the EAD hypothesis has been obtained for the LQT2 (reduced I_Kr) and LQT3 (augmentation of late I_Na) forms, whereas the arrhythmogenic contribution of EADs in LQT1 and LQT5 (both with reduced I_Ks) is still unclear. EADs have been recorded in MAPs of LQT1 patients, but only in the presence of adrenergic stimulation [147].

4.2 Acquired long-QT syndromes
The acquired long-QT syndromes are characteristically seen during bradycardia with the surface ECG showing accentuated T- and U-wave abnormalities. Irregularities in the rhythm, such as premature beats caused by EAD-related ventricular ectopy or atrial fibrillation, can lead to the induction of torsades de pointes. Most drugs that prolong and disperse ventricular repolarization may cause torsades de pointes, but the arrhythmia is usually seen in a typical clinical context, which includes female preponderance, bradycardia, irregular rhythm, and hypopotassemia and/or hypomagnesemia. In some patients with a (presumed) acquired long-QT syndrome, de novo gene mutations can be identified, which suggest the presence of a forme fruste of the congenital long-QT syndrome [148]. Clinical studies on the electrophysiological characteristics of the acquired long-QT syndrome have been conducted mostly in forms induced by drugs, such as quinidine [115], procainamide [116], disopyramide [117], and almokalant [149]. During these interventions the behavior of the repolarization phase was often dynamic, with T–U waves showing abnormal configurations such as bifidity, negativity, and 2:1 alternation [149]. The U-wave amplitude correlated strongly with the EAD amplitude, especially when the latter was recorded from a corresponding endocardial site (e.g., ECG lead V₂versus high right-ventricular MAP) [117].

β-Adrenergic receptor stimulation in the form of isoproterenol is used therapeutically to suppress acquired torsades de pointes, which is discordant with its proarrhythmic effects in the congenital long-QT syndromes. One explanation for this discordance is that the effect of β-adrenergic receptor stimulation on heart rate and enhancement of I_Ks, as well as other repolarizing currents, may override its effect to enhance Ca²⁺ entry and Ca²⁺ loading. In the drug-induced long-QT syndromes, I_Kr may be the responsible current more commonly, allowing for action-potential shortening with adrenergic enhancement of I_Ks [150].

4.3 Cardiac hypertrophy and failure
Ventricular arrhythmias and sudden death are major concerns in the clinical management of cardiac hypertrophy and failure [151,152]. These arrhythmias are often associated with action-potential prolongation and increased regional dispersion of repolarization [153]. The findings of QT-interval and repolarization labilities in patients with ischemic and nonischemic dilated cardiomyopathy [154,155], and other cardiac disorders with hypertrophy or failure, indicates that the substrate for arrhythmias is related, at least partly, to a lower repolarization reserve. Alterations in K⁺ currents [156,157] and cellular Ca²⁺ dyshandling [158,159] have indeed been found in patients with terminal heart failure, and similar ionic changes occur in animal models with compensated and decompensated myocardial overload [160–162]. Remarkably little (if any) information is available in the literature on the in vivo occurrence of EADs in patients with cardiac hypertrophy or failure, and on the putative involvement of EADs in proarrhythmia under these circumstances. In a number of large-animal models with ventricular hypertrophy, EADs were implicated in the triggering of ventricular tachyarrhythmias when the initiating beats could be captured by MAP recordings (e.g., [135,163]).

	5 Time to revise current concepts on EADs

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 
The complex ionic basis of EADs and the significance of EADs for in vivo arrhythmogenesis are currently being unraveled at rapid pace. Recent progress urges a change in concepts, which can facilitate the development of antiarrhythmic drugs and other interventions on a more rational basis.

With respect to the upstroke phase of EADs there is general consensus, based on classical experiments, that I_CaL and I_Na are the major charge carriers. A crucial addition to the understanding of EAD generation is the so-called ‘conditioning phase’. This is the transient perturbation of the normal course of repolarization that precedes the upstroke of most types of EADs. Slowed inactivation and/or reactivation of I_CaL and I_Na are contributors to this phase, but it is now becoming clear that [Ca²⁺]_cyt-dependent mechanisms also play a significant role in many cases. In the new concepts on EADs, increased cytoplasmic and/or subsarcolemmal Ca²⁺ cycling determine the ‘go or no go’ for EAD generation under various dynamic circumstances by setting the magnitude of inward I_Na–Ca. During systolic spontaneous Ca²⁺ release from the SR, as in some types of EADs, this magnitude is expected to be amplified. It has been demonstrated that in other types of EADs, such as those due to defective or inhibited K⁺ currents, the significance of I_Na–Ca is dependent on various momentary influences (e.g., heart-rate changes and neurohumoral input), but can become arrhythmogenic even at relatively low amplitudes and in the absence of spontaneous Ca²⁺ release.

It is important to recognize that the multifactorial substrate for EADs in situ is of a greater complexity than currently controllable by experimental design. In this regard, one should be very cautious to extrapolate the results on EADs induced under relatively steady-state conditions in the cellular electrophysiological laboratory to in vivo arrhythmias.

From experimental and clinical investigations, it becomes increasingly clear that the congenital and acquired long-QT syndromes (including cardiac hypertrophy and failure) each have specific features in terms of electrophysiological substrate, inciting events for torsades de pointes, and responses to antiarrhythmic therapies. The trigger roles of dynamic rate switches and adrenergic stimulation in many of these forms are compatible with the hypothesis that augmented Ca²⁺ entry and Ca²⁺ (over) loading of the SR are integral components of the mechanisms of increased dispersion of repolarization and EADs initiating ventricular tachycardia. The development of diagnostic electrophysiological protocols is obviously necessary to tailor therapeutic strategies for individual patients. Such strategies should incorporate the new concept of a [Ca²⁺]_cyt-dependent conditioning phase for action-potential prolongation and EAD formation, and could aim at controlling sudden disarrays of systolic Ca²⁺ handling or the magnitude of [Ca²⁺]_cyt-dependent inward current during the repolarization of the AP.

Time for primary review 19 days.

	Acknowledgements

 
Dr. Volders is supported by the Wynand M. Pon Foundation, Leusden, The Netherlands. Dr. Sipido is supported by the National Fund for Scientific Research, Belgium, and Dr. de Groot by the Netherlands Organization for Scientific Research (NWO 902-16-214). The authors thank R. Spätjens, BS, for helpful assistance time and again. A word of gratitude goes to Dr. S. Kuypers for collegiality.

	References

Top 1 Introduction 2 Cellular mechanisms of... 3 EADs and arrhythmogenesis... 4 Clinical relevance of... 5 Time to revise... References

 

1. Hoffman B.F., Rosen M.R. Cellular mechanisms for cardiac arrhythmias. Circ Res (1981) 49:1–15.[Free Full Text]
2. Wit A.L., Rosen M.R. The Heart and Cardiovascular System. Fozzard H.A., Haber E., Jennings R.B., Katz A.M., Morgan H.E., eds. (1992) New York: Raven Press. 2113–2163.
3. Antzelevitch C., Sicouri S. Clinical relevance of cardiac arrhythmias generated by afterdepolarizations: role of M cells in the generation of U waves, triggered activity and torsade de pointes. J Am Coll Cardiol (1994) 23:259–277.[Abstract]
4. Fabiato A. Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol (1983) 245:C1–C14.[ISI][Medline]
5. Fabiato A. Time and calcium dependence of activation and inactivation of calcium-induced release of calcium from the sarcoplasmic reticulum of a skinned canine cardiac Purkinje cell. J Gen Physiol (1985) 85:247–289.[Abstract/Free Full Text]
6. Leblanc N., Hume J.R. Sodium current-induced release of calcium from cardiac sarcoplasmic reticulum. Science (1990) 248:372–376.[Abstract/Free Full Text]
7. Sham J.S.K., Cleemann L., Morad M. Gating of the cardiac Ca²⁺ release channel: the role of Na⁺ current and Na⁺–Ca²⁺ exchange. Science (1992) 255:850–853.[Abstract/Free Full Text]
8. Kohmoto O., Levi A.J., Bridge J.H.B. Relation between reverse sodium–calcium exchange and sarcoplasmic reticulum calcium release in guinea pig ventricular cells. Circ Res (1994) 74:550–554.[Abstract/Free Full Text]
9. Grantham C.J., Cannell M.B. Ca²⁺ influx during the cardiac action potential in guinea pig ventricular myocytes. Circ Res (1996) 79:194–200.[Abstract/Free Full Text]
10. Sipido K.R., Maes M., Van de Werf F. Low efficiency of Ca²⁺ entry through the Na⁺–Ca²⁺ exchanger as trigger for Ca²⁺ release from the sarcoplasmic reticulum: a comparison between L-type Ca²⁺ current and reverse-mode Na⁺–Ca²⁺ exchange. Circ Res (1997) 81:1034–1044.[Abstract/Free Full Text]
11. Sipido K.R., Carmeliet E., Van de Werf F. T-type Ca²⁺ current as a trigger for Ca²⁺ release from the sarcoplasmic reticulum in guinea-pig ventricular myocytes. J Physiol (1998) 508:439–451.[Abstract/Free Full Text]
12. Zhou Z., January C.T. Both T- and L-type Ca²⁺ channels can contribute to excitation–contraction coupling in cardiac Purkinje cells. Biophys J (1998) 74:1830–1839.[Abstract/Free Full Text]
13. Stern M.D. Theory of excitation–contraction coupling in cardiac muscle. Biophys J (1992) 63:497–517.[Abstract/Free Full Text]
14. Sun X.H., Protasi F., Takahashi M., Takeshima H., Ferguson D.G., Franzini-Armstrong C. Molecular architecture of membranes involved in excitation–contraction coupling of cardiac muscle. J Cell Biol (1995) 129:659–671.[Abstract/Free Full Text]
15. López-López J.R., Shacklock P.S., Balke C.W., Wier W.G. Local calcium transients triggered by single L-type calcium channel currents in cardiac cells. Science (1995) 268:1042–1045.[Abstract/Free Full Text]
16. Santana L.F., Cheng H., Gómez A.M., Cannell M.B., Lederer W.J. Relation between the sarcolemmal Ca²⁺ current and Ca²⁺ sparks and local control theories for cardiac excitation–contraction coupling. Circ Res (1996) 78:166–171.[Abstract/Free Full Text]
17. Callewaert G., Cleemann L., Morad M. Epinephrine enhances Ca²⁺ current-regulated Ca²⁺ release and Ca²⁺ reuptake in rat ventricular myocytes. Proc Natl Acad Sci USA (1988) 85:2009–2013.[Abstract/Free Full Text]
18. Cleemann L., Morad M. Role of Ca²⁺ channel in cardiac excitation–contraction coupling in the rat: evidence from Ca²⁺ transients and contraction. J Physiol (1991) 432:283–312.[Abstract/Free Full Text]
19. Shacklock P.S., Wier W.G., Balke C.W. Local Ca²⁺ transients (Ca²⁺ sparks) originate at transverse tubules in rat heart cells. J Physiol (1995) 487:601–608.[Abstract/Free Full Text]
20. Wier W.G., ter Keurs H.E.D.J., Marban E., Gao W.D., Balke C.W. Ca²⁺ ‘sparks’ and waves in intact ventricular muscle resolved by confocal imaging. Circ Res (1997) 81:462–469.[Abstract/Free Full Text]
21. Eisner D.A., Trafford A.W., Dìaz M.E., Overend C.L., O’Neill S.C. The control of Ca release from the cardiac sarcoplasmic reticulum: regulation versus autoregulation. Cardiovasc Res (1998) 38:589–604.[Abstract/Free Full Text]
22. Shannon T.R., Bers D.M. Assessment of intra-SR free [Ca] and buffering in rat heart. Biophys J (1997) 73:1524–1531.[Abstract/Free Full Text]
23. Han S., Schiefer A., Isenberg G. Ca²⁺ load of guinea-pig ventricular myocytes determines efficacy of brief Ca²⁺ currents as trigger for Ca²⁺ release. J Physiol (1994) 480:411–421.[Abstract/Free Full Text]
24. Janczewski A.M., Spurgeon H.A., Stern M.D., Lakatta E.G. Effects of sarcoplasmic reticulum Ca²⁺ load on the gain function of Ca²⁺ release by Ca²⁺ current in cardiac cells. Am J Physiol (1995) 268:H916–H920.[ISI][Medline]
25. Bassani J.W.M., Yuan W., Bers D.M. Fractional SR Ca release is regulated by trigger Ca and SR Ca content in cardiac myocytes. Am J Physiol (1995) 268:C1313–C1319.[ISI][Medline]
26. Barry W.H., Rasmussen C.A.F. Jr., Ishida H., Bridge J.H.B. External Na-independent Ca extrusion in cultured ventricular cells: magnitude and functional significance. J Gen Physiol (1986) 88:393–411.[Abstract/Free Full Text]
27. Bassani R.A., Bassani J.W.M., Bers D.M. Mitochondrial and sarcolemmal Ca²⁺ transport reduce [Ca²⁺]_i during caffeine contractures in rabbit cardiac myocytes. J Physiol (1992) 453:591–608.[Abstract/Free Full Text]
28. Ehara T., Matsuoka S., Noma A. Measurement of reversal potential of Na⁺–Ca²⁺ exchange current in single guinea-pig ventricular cells. J Physiol (1989) 410:227–249.[Abstract/Free Full Text]
29. Beuckelmann D.J., Wier W.G. Sodium–calcium exchange in guinea-pig cardiac cells: exchange current and changes in intracellular Ca²⁺. J Physiol (1989) 414:499–520.[Abstract/Free Full Text]
30. Miura Y., Kimura J. Sodium–calcium exchange current: dependence on internal Ca and Na and competitive binding of external Na and Ca. J Gen Physiol (1989) 93:1129–1145.[Abstract/Free Full Text]
31. Schouten V.J.A., ter Keurs H.E.D.J., Quaegebeur J.M. Influence of electrogenic Na/Ca exchange on the action potential in human heart muscle. Cardiovasc Res (1990) 24:758–767.[Abstract/Free Full Text]
32. Janvier N.C., Boyett M.R. The role of Na–Ca exchange current in the cardiac action potential. Cardiovasc Res (1996) 32:69–84.[Abstract/Free Full Text]
33. Janvier N.C., Harrison S.M., Boyett M.R. The role of inward Na⁺–Ca²⁺ exchange current in the ferret ventricular action potential. J Physiol (1997) 498:611–625.[Abstract/Free Full Text]
34. Lederer W.J., Niggli E., Hadley R.W. Sodium–calcium exchange in excitable cells: fuzzy space. Science (1990) 248:283.[Free Full Text]
35. Carmeliet E. A fuzzy subsarcolemmal space for intracellular Na⁺ in cardiac cells? Cardiovasc Res (1992) 26:433–442.[ISI][Medline]
36. Main M.J., Grantham C.J., Cannell M.B. Changes in subsarcolemmal sodium concentration measured by Na–Ca exchanger activity during Na-pump inhibition and β-adrenergic stimulation in guinea-pig ventricular myocytes. Pflügers Arch (1997) 435:112–118.[CrossRef][ISI][Medline]
37. Fujioka Y., Matsuoka S., Ban T., Noma A. Interaction of the Na⁺–K⁺ pump and Na⁺–Ca²⁺ exchange via [Na⁺]_i in a restricted space of guinea-pig ventricular cells. J Physiol (1998) 509:457–470.[Abstract/Free Full Text]
38. Su Z., Zou A., Nonaka A., Zubair I., Sanguinetti M.C., Barry W.H. Influence of prior Na⁺ pump activity on pump and Na⁺/Ca²⁺ exchange currents in mouse ventricular myocytes. Am J Physiol (1998) 275:H1808–H1817.[ISI][Medline]
39. Langer G.A., Peskoff A. Calcium concentration and movement in the diadic cleft space of the cardiac ventricular cell. Biophys J (1996) 70:1169–1182.[Abstract/Free Full Text]
40. Soeller C., Cannell M.B. Numerical simulation of local calcium movements during L-type calcium channel gating in the cardiac diad. Biophys J (1997) 73:97–111.[Abstract/Free Full Text]
41. Barcenas-Ruiz L., Beuckelmann D.J., Wier W.G. Sodium–calcium exchange in heart: membrane currents and changes in [Ca²⁺]_i. Science (1987) 238:1720–1722.[Abstract/Free Full Text]
42. Lipp P., Pott L., Callewaert G., Carmeliet E. Simultaneous recording of indo-1 fluorescence and Na⁺/Ca²⁺ exchange current reveals two components of Ca²⁺-release from sarcoplasmic reticulum of cardiac atrial myocytes. FEBS Lett (1990) 275:181–184.[CrossRef][ISI][Medline]
43. Hüser J., Lipsius S.L., Blatter L.A. Calcium gradients during excitation–contraction coupling in cat atrial myocytes. J Physiol (1996) 494:641–651.[Abstract/Free Full Text]
44. Trafford A.W., Dìaz M.E., O’Neill S.C., Eisner D.A. Comparison of subsarcolemmal and bulk calcium concentration during spontaneous calcium release in rat ventricular myocytes. J Physiol (1995) 488:577–586.[Abstract/Free Full Text]
45. Sipido K.R., Callewaert G., Carmeliet E. Inhibition and rapid recovery of Ca²⁺ current during Ca²⁺ release from sarcoplasmic reticulum in guinea pig ventricular myocytes. Circ Res (1995) 76:102–109.[Abstract/Free Full Text]
46. Sham J.S.K., Cleemann L., Morad M. Functional coupling of Ca²⁺ channels and ryanodine receptors in cardiac myocytes. Proc Natl Acad Sci USA (1995) 92:121–125.[Abstract/Free Full Text]
47. Blatter L.A., Niggli E. Confocal near-membrane detection of calcium in cardiac myocytes. Cell Calcium (1998) 23:269–279.[CrossRef][ISI][Medline]
48. Etter E.F., Minta A., Poenie M., Fay F.S. Near-membrane [Ca²⁺] transients resolved using the Ca²⁺ indicator FFP18. Proc Natl Acad Sci USA (1996) 93:5368–5373.[Abstract/Free Full Text]
49. Trafford A.W., Dìaz M.E., Eisner D.A. Ca-activated chloride current and Na–Ca exchange have different timecourses during sarcoplasmic reticulum Ca release in ferret ventricular myocytes. Pflügers Arch (1998) 435:743–745.[CrossRef][ISI][Medline]
50. Capogrossi M.C., Stern M.D., Spurgeon H.A., Lakatta E.G. Spontaneous Ca²⁺ release from the sarcoplasmic reticulum limits Ca²⁺-dependent twitch potentiation in individual cardiac myocytes: a mechanism for maximum inotropy in the myocardium. J Gen Physiol (1988) 91:133–155.[Abstract/Free Full Text]
51. Volders P.G.A., Kulcsár A., Vos M.A., et al. Similarities between early and delayed afterdepolarizations induced by isoproterenol in canine ventricular myocytes. Cardiovasc Res (1997) 34:348–359.[Abstract/Free Full Text]
52. Cheng H., Lederer W.J., Cannell M.B. Calcium sparks: elementary events underlying excitation–contraction coupling in heart muscle. Science (1993) 262:740–744.[Abstract/Free Full Text]
53. Dìaz M.E., Cook S.J., Chamunorwa J.P., et al. Variability of spontaneous Ca²⁺ release between different rat ventricular myocytes is correlated with Na⁺–Ca²⁺ exchange and [Na⁺]_i. Circ Res (1996) 78:857–862.[Abstract/Free Full Text]
54. Satoh H., Blatter L.A., Bers D.M. Effects of [Ca²⁺]_i, SR Ca²⁺ load, and rest on Ca²⁺ spark frequency in ventricular myocytes. Am J Physiol (1997) 272:H657–H668.[ISI][Medline]
55. Parker I., Wier W.G. Variability in frequency and characteristics of Ca²⁺ sparks at different release sites in rat ventricular myocytes. J Physiol (1997) 505:337–344.[Abstract/Free Full Text]
56. Cheng H., Lederer M.R., Lederer W.J., Cannell M.B. Calcium sparks and [Ca²⁺]_i waves in cardiac myocytes. Am J Physiol (1996) 270:C148–C159.