| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Division of Pulmonary and Critical Care Medicine, The Johns Hopkins Medical Institutions, Baltimore, Maryland 21224
Submitted 25 March 2003 ; accepted in final form 15 August 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
early afterdepolarization; delayed afterdepolarization; flash photolysis; Ca2+ overload; triggered arrhythmia; ischemia-reperfusion arrhythmia; sarcoplasmic reticulum
As one of the complex influences that shape the cardiac action potential, Na+/Ca2+ exchange can have an adverse effect on contractility and arrhythmogenesis under pathophysiological conditions. For example, myocardial Ca2+ overload is characterized by waves of spontaneous SR Ca2+ release that propagate through the cytoplasm generating inward INaCa and delayed afterdepolarizations (DADs) (2, 3, 28, 38, 51, 58, 67). This mechanism is thought to underlie arrhythmias associated with myocardial ischemia-reperfusion (7, 49). A similar mechanism may also facilitate or trigger early afterdepolarizations (EADs) and long Q-T arrhythmias, although this possibility remains controversial (37, 48, 57, 73, 75). Earlier studies (26, 39, 40) suggested that EADs result solely from the reactivation or sustained opening of sarcolemmal L-type Ca2+ channels (window ICa) during prolonged action potentials. However, Ca2+ overload and spontaneous SR Ca2+ release have been implicated in the long Q-T syndrome and some reports (11, 25, 52, 72, 73) have suggested that spontaneous [Ca2+]i transients can induce EADs in cells exposed to high doses of catecholamines. Recent evidence from perfused hearts exposed to the class III antiarrhythmic agent E4031 also showed a close association between [Ca2+]i transients and EADs (9), implying that INaCa may be one of the main triggers for EADs rather than merely contributing to APD prolongation.
In the present study, we sought to clarify the role of INaCa during SR Ca2+ release on action potentials and afterdepolarizations in isolated guinea pig ventricular myocytes, which exhibit an action potential configuration and Na+/Ca2+ exchange activity similar to human myocytes (4, 5, 62). To evaluate the influence of inward INaCa on APD, INaCa was reversibly inhibited and SR Ca2+ loading was manipulated by substituting extracellular Na+ with Li+. This ion acts readily as a Na+ analog in sarcolemmal ion channels, but is not transported by the Na+/Ca2+ exchanger, and thereby blocks the efflux of cytsolic Ca2+ (15, 18, 35, 61). To examine directly the role of Ca2+ release on afterpolarization generation, the timing of SR Ca2+ release was precisely controlled by laser-flash photolysis of caged Ca2+, DM nitrophen [1-(2-nitro-4,5-dimethoxyphenyl)-1,2-diaminoethane-N,N,N,',N'-tetraacetate, 4 Na; (DMNP)] (12, 44). Our results using these techniques suggest that SR Ca2+ release and INaCa can cause large changes in the duration and kinetics of action potentials, eliciting both DADs and EADs. These findings support a role for INaCa as an arrhythmogenic mechanism with the capacity to alter the action potential configuration dramatically in cardiac ventricular cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
350 g) were euthanized by an intraperitoneal injection of pentobarbitone sodium (100 mg/kg) in accordance with guidelines of the Johns Hopkins University Animal Care and Use Committee. Myocytes were isolated by a modified method of Mitra and Morad (43). Briefly, the heart was removed and perfused retrogradely at 37°C for 5 min with nominally Ca2+-free Tyrode solution containing (in mM) 137 NaCl, 5 KCl, 1 MgCl2, 10 D-glucose, and 10 NaHEPES (HEPES neutralized to pH 7.4 with NaOH). This was followed by recirculation of the same solution containing (U/ml) 300 collagenase (type I) and 1 protease (type XIV) for 10 min and then perfusion with enzyme-free Tyrode solution containing 0.2 mM CaCl2 for a further 5 min to stop enzymatic digestion. The ventricles were cut radially, and myocytes were dispersed at room temperature for experiments within 8 h of isolation.
Evoked action potentials. Whole cell current-clamped myocytes, held at zero current, were superfused with Tyrode solution containing 2 mM CaCl2. The ruptured patch-clamp method was employed to ensure that fluorescent indicators and caged compounds equilibrated fully with the cytosol and to prevent their intracellular compartmentalization. Action potentials were stimulated through the patch pipettes (resistance 2.5–5 M
) by 5-ms depolarizing pulses of 1.5 nA or less, applied at 0.2–0.33 Hz. The electrode-filling solution contained (in mM) 75 K glutamate, 20 KCl, 0.5 MgCl2, 10 MgATP, 5 Na2-phosphocreatine, 0.1 NaGTP, 5 pyruvic acid, and 30 K HEPES (HEPES neutralized to pH 7.2 with KOH). Electrophysiological signals were stimulated, recorded, and analyzed with the use of pCLAMP software (version 8.1) coupled to an Axopatch amplifier (Axon Instruments; Foster City, CA). APD at 90% repolarization (APD90) was determined for comparison. By following the reasoning of LeGuennec and Noble (33), we conducted the experiments at room temperature (25°C) to prolong APD90 so that perturbations of the action potential plateau, in particular due to SR Ca2+ release activated by laser-flash photolysis of caged Ca2+, were readily resolvable. A previous report (53) indicates that the mammalian cardiac action potential is not substantially affected by temperature, other than in duration. Furthermore, EADs and DADs of similar characteristics have been reported at both body and room temperatures, suggesting that their occurrences are also insensitive to cooling (2, 11, 25, 40, 66).
Intracellular Ca2+ measurements. The Ca2+-sensitive fluorescent indicators indo-1 (60 µM) and fluo-3 (50–100 µM) were included in the patch pipette solutions to measure, respectively, whole cell or localized [Ca2+]i. Indo-1 fluorescence was excited at 360 ± 8 nm with the use of a Nikon Diaphot microscope-based microphotometry system (PTI; S. Brunswick, NJ) and fluorescence signals were recorded at 405 ± 15 and 485 ± 12.5 nm to obtain the indo-1 fluorescence ratio (defined as F405/F485, where F405 and F485 are fluorescence intensities measured at the two emitted light wavelengths). All measurements were corrected for background fluorescence recorded before establishing whole cell current clamp. Action potentials and [Ca2+]i transients were digitized at 1 kHz and low-pass filtered at 500 Hz. Confocal images were acquired using a Zeiss LSM 510 microscope with a Zeiss Plan-Neofluor x40 oil immersion objective (numerical aperture = 1.3), and the confocal pinhole was set to render spatial resolutions of 0.4 µm in the x-y axes and 0.9 µm in the z-axis. Fluo-3 was excited at 488 nm, and fluorescence was recorded at wavelengths >505 nm. The images were acquired in linescan mode, with the scan line oriented parallel to the long axis of the cell, containing 512 pixels/line (0.075 µm/pixel) that were scanned repeatedly at 2- or 3-ms intervals for 6 s. Photobleaching and laser damage to the cells were minimized by attenuation of the laser to 1% of its maximum power (25 mW) with the use of an acousto-optical tunable filter. Image acquisition was synchronized with current-clamp protocols, such that line scans were initiated 100 ms before the stimulation of an action potential. The precise timing of the stimulus was marked by a blanking light-emitting diode output into the transmitted light channel of the microscope. Confocal images were processed with the use of IDL software (Research Systems; Boulder, CO). Fluorescence signals were background subtracted, and expressed relative to resting fluorescence (F0) as F/F0.
Laser-flash photolysis of caged Ca2+. In experiments using caged Ca2+ to trigger intracellular Ca2+ release, the composition of the internal solution was modified by omitting Na2-phosphocreatine and replacing it with 2 mM Na4-DMNP (Calbiochem; San Diego, CA) plus 0.25–0.5 mM CaCl2. In addition, 2 mM NaCl was included to give 10 mM total intracellular Na+, and reduced glutathione (3 mM) was added to protect against cytosolic damage due to photolysis byproducts (44). Ca2+ was released from DMNP by flashing the cells with a single pulse of UV light at 337 nm generated from a N2 laser (model GL-3300, PTI). The laser energy (maximum = 2 mJ) was attenuated to about 300 µJ and delivered using a 400-µm-diameter UV-enhanced fiber-optic light guide (WPI; Sarasota, FL) positioned 100 µm from the myocyte of interest. Myocytes were loaded with fluo-3 (50 or 100 µM) to record changes in [Ca2+]i using confocal microscopy as described above. In these experiments, a second light-emitting diode flash was used to indicate the timing of the laser pulse.
Chemicals and miscellaneous. Unless indicated otherwise, the chemicals were purchased from Sigma-Aldrich (St. Louis, MO). Pentapotassium salts of fluo-3 and indo-1 were purchased from Molecular Probes (Eugene, OR). KB-R7943 (KBR, Tocris; Ellisville, MO) and ryanodine (Calbiochem) were dissolved in DMSO as 10 mM stock solutions and diluted with Tyrode solution to give final concentrations of 5 and 20 µM, respectively. For Li+ Tyrode solution, NaCl was substituted by equimolar LiCl, and the pH of HEPES was adjusted to 7.4 using LiOH instead of NaOH. Throughout this study, external solutions were changed rapidly (<1 s) using a concentration-clamp system. Data are expressed as means ± SE and were compared statistically using the appropriate paired or unpaired t-test. A P value of <0.05 is taken to indicate a statistically significant difference, and n indicates the number of cells studied unless indicated.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
10 stimulus intervals. The periods of APD abbreviation and overshoot coincided with gradual increases and decreases, respectively, in the peak amplitude of [Ca2+]i transients indicated by indo-1 fluorescence (Fig. 1A, bottom). Similar changes occurred to a lesser extent in resting [Ca2+]i. The progressive changes in [Ca2+]i contrast strongly with the essentially instantaneous alterations in APD90 (indicated by arrows in Fig. 1A), reflecting the dependence of [Ca2+]i transient amplitude on progressive changes in SR Ca2+ loading. Figure 1B shows individual action potentials and fluorescence transients at higher resolution. Under control conditions, the action potential triggered a rather small [Ca2+]i transient (Fig. 1, i). In contrast, the abbreviated action potential in Li+ Tyrode solution (Fig. 1, ii) evoked a [Ca2+]i transient of roughly fivefold larger amplitude than control. This transient was followed by a spontaneous SR Ca2+ release, which is a characteristic sign of SR Ca2+ overload. We attribute this SR Ca2+ overloading to Ca2+ influx (ICa and reverse Na+/Ca2+ exchange) during action potentials, not compensated by Ca2+ efflux via forward Na+/Ca2+ exchange (65). After solutions were switched back to Na+ Tyrode solution, the potentiated [Ca2+]i transients decayed in parallel with the overshoot of APD90 (Fig. 1, iii). At the start of the decay, spontaneous [Ca2+]i transients evoked "slow" action potentials that were initiated by gradual depolarizations and had smaller, less steep upstrokes than those observed during stimulated action potentials. Collectively, in six cells, mean APD90 of 427 ± 56 ms decreased significantly to 207 ± 15 ms (a change of 52%) after solutions were switched to Li+ Tyrode solution, accompanied by a small but significant depolarization of the resting potential of 2.9 ± 0.2 mV (27). When Na+ was returned to the Ca2+-overloaded cells, the mean overshoot in APD was 59 ± 12% based on the APD90 of the first and tenth action potentials after solutions were switched. Thus the results shown in Fig. 1 suggest that SR Ca2+ overload caused both spontaneous depolarizations and potentiated [Ca2+]i transients, which, acting via inward INaCa, considerably prolonged the APD due to enhanced Ca2+ efflux.
|
To elucidate further the involvement of INaCa in Em changes, Na+/Ca2+ exchange was inhibited either by eliminating SR Ca2+ release using 20 µM ryanodine or by using the antagonist KBR. Figure 2A, top, shows that exposing a cell to ryanodine caused rapid APD abbreviation and the substitution of Li+ for Na+ in the presence of ryanodine typically caused further APD shortening. Individual action potentials (Fig. 2A, bottom) indicate that ryanodine abbreviated the action potential plateau and slightly reduced the action potential amplitude. Mean APD90 significantly decreased from 622 ± 65 to 366 ± 30 ms (41%) in the presence of ryanodine (n = 5), with a further decrease to 218 ± 10 ms (24%) after solutions were switched to Li+ Tyrode solution (plus ryanodine). Parallel results were obtained when INaCa was directly inhibited using 5 µM KBR (Fig. 2B) except that the time course of APD changes was slower than for ryanodine. Inhibition of Na+/Ca2+ exchange resulted in progressive abbreviation of APD90, followed by further abbreviation after substitution of Li+ for Na+, suggesting that 5 µM KBR only partly inhibited Na+/Ca2+ exchange. Mean APD90 in seven cells was 492 ± 62 mV and shortened to 332 ± 30 ms (32%) in the presence of KBR, with further abbreviation to 252 ± 18 mV (16%) after Li+ was exchanged for Na+. The total reductions in APD90 due to ryanodine plus Li+ (65%) and KBR plus Li+ (48%) were quite similar to the inhibition produced by Li+ alone (i.e., 52%). It therefore appears that all three agents acted on APD primarily by inhibiting INaCa. Because Li+-containing solution may potentiate outward Na+/Ca2+ exchange in the presence of 10 mM intracellular Na+, the small further reduction in APD caused by Li+ in the presence of ryanodine or KBR might reflect the maximum influence of outward Na+/Ca2+ exchange on APD.
|
Effects on APD of [Ca2+]i oscillations in Li+ Tyrode solution. Inhibition of Na+/Ca2+ exchange by Li+ caused SR Ca2+ overload made evident spontaneous [Ca2+]i oscillations. After their onset, Ca2+ oscillations increased in frequency with the number of applied stimuli (Fig. 3A). Figure 3B shows that the mean oscillation frequency increased from 0.5 to 1.3 Hz over 10 stimulation intervals in Li+ Tyrode solution (n = 7). During this increase, the Ca2+ oscillations had no obvious effects on resting membrane potential, despite a maximum increase in the resting fluorescence ratio by a factor of two relative to control. This suggests that in guinea pig ventricular cells, other Ca2+-dependent currents were absent or had minimal effects on resting Em. Figure 3C shows that spontaneous [Ca2+]i transients nonetheless affected APD in Li+ Tyrode solution. This influence was quantified by calculating the difference in APD90 between the longest and shortest action potentials in the presence of Li+ for each cell (n = 6). By this measure, the mean difference in APD90 was 81 ± 11 ms (32 ± 3% of the longer duration action potentials). Careful examination of action potentials, such as those shown in Fig. 3C, reveals that [Ca2+]i oscillations that coincided with the stimulation pulse caused the most APD abbreviation (dotted lines). Because ICa is the principal inward current able to affect APD during superfusion with Li+-Tyrode solution, this phenomenon appears to reflect cytosolic Ca2+-dependent inactivation of ICa (1, 21, 60). The dependence of APD on [Ca2+]i changes during the normal action potential is therefore likely to be mediated by the Ca2+ sensitivity of both INaCa and ICa.
|
[Ca2+]i oscillations and afterdepolarizations. Figure 4 shows that a range of Em depolarizations could be attributable to spontaneous oscillatory SR Ca2+ release immediately after reintroduction of Na+-containing Tyrode solution. Figure 4A shows a [Ca2+]i oscillation that followed a stimulated action potential, eliciting a DAD that failed to reach the threshold for triggering a slow action potential. Fourteen similar responses were observed in six cells. More commonly, however, DADs exceeded the threshold for evoking slow action potentials (23 events). A typical example of this type is shown in Fig. 4B, in which a repetitive series of regularly timed Ca2+ oscillations and slow action potentials occurred immediately after the stimulated action potential. Uniquely, the stimulated action potential in Fig. 4B had an extra "dome" at the early plateau phase, suggesting that INaCa, due to a secondary SR Ca2+ release, was capable of further depolarizing the cell. This observation is supported by the occurrence of a secondary depolarization during phase 3 repolarization of the stimulated action potential shown in Fig. 4C, top. In this particular action potential, [Ca2+]i was already elevated at the point of stimulation, as a result of a preceding spontaneous Ca2+ release, but the membrane potential was not strongly depolarized due to ongoing superfusion with Li+ Tyrode solution. During repolarization of the same action potential, the switch from Li+- to Na+-Tyrode solution was coincident with a secondary Ca2+ transient (Fig. 4C, bottom) and a large secondary depolarization. The secondary depolarization strongly resembles a phase 3 EAD as modeled by Luo and Rudy (37). Three similar events were recorded from other cells. A closer inspection of Fig. 4C reveals that this typical phase 3 EAD arose at an Em of –28 mV and peaked at –10 mV. The upstrokes and peaks of the EAD and its associated [Ca2+]i transient coincided, with no detectable delays. When considered with the fact that no further membrane depolarization (above –10 mV) occurred after [Ca2+]i had begun to decay, these observations suggest that the EAD was caused by the underlying Ca2+ transient (not vice versa). For comparison, the two slow action potentials that followed the EAD in Fig. 4C were triggered by DADs and both had relatively fast upstrokes peaking at around +30 mV. It is also noticeable that [Ca2+]i was increasing at the start of the phase 3 EAD in Fig. 4C (at –28 mV), whereas at the equivalent point during repolarization of the subsequent spontaneous action potentials, [Ca2+]i was rapidly declining. Because the two spontaneous action potentials contained no phase 3 EADs, it follows that spontaneous [Ca2+]i transients during repolarization may be the crucial trigger of phase 3 EADs. On the basis of these characteristics, it is deduced that voltage-gated sarcolemmal ion channels contribute relatively little to phase 3 EADs, which would otherwise have depolarized to a more positive potential, likely driven by Ca2+ equilibrium potential (+130 mV).
|
To investigate the mechanisms underlying slow action potentials, the peak Em during subthreshold DADs was compared with the potential at which the voltage slope increased during suprathreshold DADs. The latter is termed the "take-off" potential (TOP) for slow action potentials and is most evident in the second slow action potential shown in Fig. 4C. The mean TOP was –46 ± 1mV(n = 23 events), which is significantly more positive than the peak of subthreshold DADs of –62 ± 2mV(n = 15 events). This observation suggests that the membrane conductance underlying the faster depolarization phase of triggered slow action potentials was available at Em above –46 mV, i.e., close to the activation threshold for voltage-gated Ca2+ channels (41).
Spatiotemporal relationships between [Ca2+]i and Em were examined more closely in myocytes imaged by confocal microscopy. Figure 5A shows a control action potential stimulated in Na+ Tyrode solution that elicited a synchronous fluorescence increase across the cell (Fig. 5A, middle). This Ca2+ release (Fig. 5A, bottom) occurred after a delay of 10 ms and peaked 80 ms after the beginning of the action potential upstroke (74). Compared with stimulated action potentials, spontaneous [Ca2+]i oscillations in Li+ Tyrode solution produced much less synchronous linescan fluorescence changes, consistent with the propagation of cytosolic Ca2+ waves (at a frequency of 1 Hz). The inset in Fig. 5B, top, illustrates that these Ca2+ oscillations caused small "ripples" in the resting potential of the cell, visible only at high amplification. The mean amplitude of the voltage oscillations was found to be 1.0 ± 0.2 mV in 15 exposures to Li+ (n = 6), confirming the limited extent to which [Ca2+]i oscillations affected resting Em in the absence of Na+/Ca2+ exchange. After Na+ was returned to the superfusion solution, [Ca2+]i oscillations at the resting potential triggered slow action potentials that had two distinct phases (Fig. 5C). First, an initial slow depolarization at the foot of the slow action potential appeared to be closely associated with the spontaneous wave of Ca2+ release. Second, a more rapid upstroke was associated with the more synchronous [Ca2+]i change (see dotted line). Such results support the view that slow action potentials were initiated by Ca2+ waves, leading to subsequent activation of L-type Ca2+ channels to generate the fast upstroke and trigger synchronized SR Ca2+ release.
|
Induction of afterdepolarizations by photorelease of Ca2+. To demonstrate the involvement of [Ca2+]i changes in EADs and DADs more conclusively, SR Ca2+ release was activated by photolysis of caged Ca2+ (DMNP) in cells imaged confocally. The spatially averaged [Ca2+]i transients triggered by action potentials in cells coloaded with DMNP and fluo-3 were similar to those observed in cells not loaded with the caged compound (Fig. 6A). Figure 6B shows, in a myocyte that had submaximal evoked SR Ca2+ release, that flash photolysis of DMNP during action potential repolarization (at approximately –40 mV) caused the immediate appearance of a secondary Ca2+ transient superimposed upon the transient activated by stimulation. The secondary SR Ca2+ release due to caged Ca2+ photolysis unequivocally elicited a phase 3 EAD. Similar responses were observed in three other cells, clearly showing that photorelease of Ca2+ during the action potential plateau can interrupt repolarization, causing phase 3 EADs similar to those due to spontaneous SR Ca2+ releases (Fig. 4C). When a laser flash was applied earlier in the time course of the action potential (Fig. 6C), the smaller secondary Ca2+ transient triggered by DMNP photolysis was associated with an immediate acceleration of repolarization. Repolarization to negative potentials triggered a prolonged Ca2+ release during which a pseudoplateau developed at Em –15 mV, followed by terminal repolarization as [Ca2+]i declined. Finally, laser flashes applied at the resting potential evoked DADs (n = 20). In four cells, these DADs exceeded the threshold for stimulating action potentials, typified by Fig. 6D. The striking fact that these Em changes closely mirrored [Ca2+]i changes further suggests that INaCa is a major determinant of action potential morphology.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
50% of normal APD in isolated guinea pig ventricular myocytes. In Ca2+ overload, spontaneous Ca2+ oscillations could evoke both DADs and EADs that disappeared as the level of SR Ca2+ receded, INaCa was inhibited, or SR Ca2+ release was abolished pharmacologically. The photolysis of caged Ca2+ during or after a stimulated action potential could also elicit, respectively, an EAD or DAD. Taken together, these findings strongly suggest that INaCa has major effects both on APD in guinea pig ventricular myocytes and on afterdepolarizations. Furthermore, our findings suggest that APD is reciprocally influenced by the effects of [Ca2+]i on the inactivation of ICa, but is less affected by other Ca2+-dependent conductances. Effects of INaCa on APD. During normal endothelial cell coupling, the amounts of Ca2+ influx and SR Ca2+ release depend on APD, which is itself modulated by [Ca2+]i acting via a complex interplay of Ca2+-sensitive ionic mechanisms. Forward Na+/Ca2+ exchange is the predominant Ca2+ efflux mechanism in cardiac myocytes and is the most important regulator of [Ca2+]i. It is thus quite expected that Ca2+ extrusion via this electrogenic mechanism may play a role in modulating APD. In the present experiments, we observed that inhibition of inward INaCa by substitution of Li+ for extracellular Na+ shortened the APD and, conversely, increasing inward INaCa by potentiating SR Ca2+ release during Ca2+ overload substantially lengthened the APD (Fig. 1). Moreover, APD abbreviation or prolongation developed within 1–2 stimulation intervals (3 s) of switch to or from Li+ Tyrode solution, respectively (see Fig. 1A). The absence of an appreciable lag indicates that the rapid alteration in the exchanger activity, instead of the gradual change in SR Ca2+ loading, was responsible for the APD changes (16, 65).
The notion that INaCa contributes to normal APD in guinea pig ventricular myocytes is further supported by the pharmacological elimination of Na+/Ca2+ exchange using KBR and by the inhibition of SR Ca2+ release using ryanodine (Fig. 2). Both of these agents mimicked the effect of Li+, abbreviating APD90 by 30–40%. KBR may have differential effects on forward and reverse Na+/Ca2+ exchange in some species, but the available evidence suggests that this is not the case in guinea pig cells (29, 36). Despite other side effects of KBR in addition to inhibiting the exchanger (56, 70), and the fact that ryanodine may eliminate Ca2+ release-induced ICa inactivation (60, 64), the APD abbreviation caused by all three independent methods unequivocally suggests that Ca2+ release-mediated INaCa is a major determinant of APD in guinea pig ventricular myocytes. These findings are consistent with previous reports showing the effects on APD of eliminating Na+/Ca2+ exchange by Li+ substitution or by eliminating SR Ca2+ release with the use of EGTA, BAPTA-AM, ryanodine, or caffeine in cardiac cells of several species (2, 13, 27, 30, 33, 34, 47, 59).
Influences of other Ca2+-dependent mechanisms on APD. In contrast to our experimental results, a study in rat cardiac myocytes found a slight increase of 10–20% in APD when SR Ca2+ release was abolished (20). A simulation study also suggested that reducing Na+/Ca2+ exchange might actually cause APD prolongation in guinea pig cardiac cells (21). These findings indicate that in addition to Ca2+ fluxes via Na+/Ca2+ exchange, counteracting mechanisms such as ICa inactivation may significantly affect APD (1, 21, 60). Figure 3 shows that APD variations occurred in phase with spontaneous [Ca2+]i oscillations in the presence of Li+. APD was shortest during action potentials evoked at the peak of the oscillatory Ca2+ transient, suggesting that an underlying variation in Ca2+ release-dependent inactivation of ICa was responsible. Consistent with this idea, flash photolysis of DMNP at an early phase of the action potential accelerated repolarization until Em fell <0 mV (Fig. 6C). Moreover, APD may also be influenced by Ca2+-activated Cl– or nonselective cation currents in some species (15, 31, 73). In the present study, Ca2+ oscillations evoked resting membrane depolarizations of about 1 mV in Li+ Tyrode solution (Fig. 5B) suggesting that the combined depolarizing effects of such Ca2+-dependent currents was negligible in guinea pig ventricular myocytes. It has to be emphasized, however, that the relative contribution of inward INaCa and ICa inactivation to APD during a Ca2+ release could be species and inotropic state dependent. For example, in species like rats, where APD is short, ICa inactivation is robust (60), and Na+/Ca2+ exchange activity is relatively low (62), ICa inactivation during SR Ca2+ release may have a greater influence on the AP configuration than in species with longer action potentials (20). Moreover, during steady stimulation when sarcolemmal Ca2+ fluxes are in balance, an increased ICa would lead to greater SR Ca2+ release and INaCa.
Afterdepolarizations evoked by SR Ca2+ release and INaCa. The chief focus of the present studies was to evaluate the possibility that inward INaCa can cause early and delayed afterdepolarizations. The rationale of superfusing myocytes with Li+ Tyrode solution was to increase SR Ca2+ loading sufficient to elicit spontaneous SR Ca2+ releases and to determine the influence of these [Ca2+]i transients on APD. In Na+ Tyrode solution, SR Ca2+ releases occurred simultaneously with DADs, slow action potentials, and EADs (Fig. 4). These results are consistent with several previous studies showing the coexistence of afterdepolarizations when cardiac cells were exposed to certain pharmacological agents (11, 38, 52). We also observed marked kinetic differences between EADs and DADs. The typical EAD shown in Fig. 4C was in phase with a [Ca2+]i transient, but the change in Em during the EAD was considerably slower than in an action potential triggered by a DAD, and the peak potential was less positive. Such findings are consistent with a mechanism in which EADs are attributable mainly to spontaneous SR Ca2+ release and INaCa. At the resting membrane potential, slow depolarizations due to Ca2+ oscillations either progressed to the upstroke of a slow action potential when an apparent TOP was reached or subsided if the threshold was unattained. Confocal linescan images (Fig. 5) suggested that these initial slow depolarizations were related to Ca2+ waves traveling across the cell (8, 11, 71) because DADs were not observed when INaCa was inhibited by Li+ substitution or when Ca2+ waves were abolished by ryanodine. The step from DADs to slow action potentials likely involved additional activation of ICa because the TOP of slow action potentials was similar to the voltage threshold for Ca2+ channel activation, and the peak potential was highly positive, consistent with the driving force for Ca2+. This notion is supported by the abrupt synchronization of SR Ca2+ release with the slow action potential upstroke, which resembles Ca2+ release triggered by the L-type Ca2+ channel during an evoked AP. As far as we are aware, similar images of [Ca2+]i changes during the transition from DADs to slow action potentials have not been reported previously.
Although INaCa driven by [Ca2+]i is the well-accepted mechanism for DADs (2, 37, 51, 67, 73), our results also represent the first direct demonstration that SR Ca2+ release is capable of evoking EADs. This was evident clearly when the primary events in the activation of EADs were the precisely controlled photolysis of caged Ca2+ or unevoked SR Ca2+ release during the repolarization phase of the action potential. Nonetheless, earlier studies had dissociated the mechanisms of EADs and DADs, suggesting that EADs do not require Ca2+ elevation but rather are related to Ca2+ entry via the L-type Ca2+ channels (11, 37, 40). This hypothesis, which has led to the widely held view that window ICa is the major cause of EADs, is strongly supported by the work of January and colleagues (23, 26, 39), showing that the voltage dependence and pharmacological sensitivity of EADs is correlated with that of L-type Ca2+ current. However, a great diversity of agents evoke EADs in cardiac preparations (39, 66, 73, 75) and because EADs are categorized phenomenologically based on their timing of occurrence, regardless of the methods of initiation, it is possible that the various EADs may not all be explained by a single unifying mechanism such as ICa reactivation. For example, it is almost certain that window ICa cannot explain the initiation of EADs below –45 mV, the threshold voltage for Ca2+ channel activation (10, 48). Moreover, because EADs arise after widely variable delays and vary in occurrence from one action potential to the next, it seems plausible that an underlying random process, such as spontaneous SR Ca2+ release, is involved in their mechanism (10, 46, 48, 52, 66, 68, 72, 76). Consistent with this idea, the association of EADs with APD prolongation permits sufficient time for the recovery of SR Ca2+ release from refractoriness induced by the initial [Ca2+]i transient triggered by the upstroke (63).
Despite our strong evidence that SR Ca2+ release can trigger EADs during Ca2+ overload, we cannot absolutely discount the involvement of L-type Ca2+ channels in the generation of EADs, especially under different physiological conditions and pharmacological manipulations. In fact, the kinetics of Ca2+ channels are strongly affected by SR Ca2+ release (1, 60), so it is possible that complex interactions among SR Ca2+ release, INaCa, and ICa are involved in shaping the characteristics of various forms of EADs. Hence, future experiments are required to further evaluate the involvement of INaCa in different EAD models.
Physiological and clinical implications. Afterdepolarizations are of great significance to arrhythmogenesis in cardiac cells, particularly in reperfusion-induced arrhythmias (7, 49), torsade de pointes, and human long Q-T syndrome (39, 40, 73, 75). In the present paper, we have shown that SR Ca2+ release acting via INaCa can be the primary underlying mechanism of both types of afterdepolarizations during Ca2+ overload. This mechanism is likely to be involved in catecholamine-induced arrhythmia (11, 25, 52, 72) and may contribute to arrhythmia induced by mechanical loading, which is known to increase Ca2+ transients and Ca2+ influx via mechanosensitive nonselective cation channels (20, 22, 55). The novel findings reported here could open new avenues for further research because the expression and density of INaCa have been convincingly shown to be upregulated in heart failure (9, 24, 50, 51). Studies in isolated spontaneously beating sinoatrial node cells have also shown that the chronotropic effect of 
-adrenergic stimulation is associated with changes in SR Ca2+ release (32) and INaCa was found to link [Ca2+]i transients to membrane depolarizations, implying that a DAD-like mechanism plays a modulatory role in the pacemaker potential. Interestingly, ablation of inward rectifier K+ channels by gene therapy was recently reported to transform ventricular myocytes into pseudopacemaker cells (42). Our observation of short trains of regular slow action potentials during Ca2+ waves (e.g., Fig. 5B) raises the possibility that repetitive pacemaker-like activity could arise by DAD-like spontaneous SR Ca2+ release and INaCa, if the depolarized resting potential of pseudopacemaker cells permits Ca2+ entry and SR Ca2+ overload. Further experiments are necessary to test this possibility. Finally, the EAD-like events we induced by SR Ca2+ release were superimposed on repolarization rather than during a stable prolonged action potential plateau (75). Therefore, further experiments are in hand to examine the morphology and mechanism of Ca2+-dependent EADs during more prolonged action potentials.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Pott, X. Ren, D. X. Tran, M.-J. Yang, S. Henderson, M. C. Jordan, K. P. Roos, A. Garfinkel, K. D. Philipson, and J. I. Goldhaber Mechanism of shortened action potential duration in Na+-Ca2+ exchanger knockout mice Am J Physiol Cell Physiol, February 1, 2007; 292(2): C968 - C973. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Ono, H. Hayashi, A. Kawase, S.-F. Lin, H. Li, J. N. Weiss, P.-S. Chen, and H. S. Karagueuzian Spontaneous atrial fibrillation initiated by triggered activity near the pulmonary veins in aged rats subjected to glycolytic inhibition Am J Physiol Heart Circ Physiol, January 1, 2007; 292(1): H639 - H648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Bers Altered Cardiac Myocyte Ca Regulation In Heart Failure. Physiology, December 1, 2006; 21(6): 380 - 387. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Pott, K. D. Philipson, and J. I. Goldhaber Excitation-Contraction Coupling in Na+-Ca2+ Exchanger Knockout Mice: Reduced Transsarcolemmal Ca2+ Flux Circ. Res., December 9, 2005; 97(12): 1288 - 1295. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Tuteja, D. Xu, V. Timofeyev, L. Lu, D. Sharma, Z. Zhang, Y. Xu, L. Nie, A. E Vazquez, J. N. Young, et al. Differential expression of small-conductance Ca2+-activated K+ channels SK1, SK2, and SK3 in mouse atrial and ventricular myocytes Am J Physiol Heart Circ Physiol, December 1, 2005; 289(6): H2714 - H2723. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Fauconnier, A. Lacampagne, J.-M. Rauzier, G. Vassort, and S. Richard Ca2+-dependent reduction of IK1 in rat ventricular cells: A novel paradigm for arrhythmia in heart failure? Cardiovasc Res, November 1, 2005; 68(2): 204 - 212. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
, Spontaneous sarcoplasmic reticulum (SR) Ca2+ releases and/or depolarizations. Bottom, time bar refers to all three traces.
