Abstract
Pathological conditions, including ischemia and heart failure, are associated with altered sodium channel function and increased late sodium current (INa,L), leading to prolonged action potential duration, increased intracellular sodium and calcium concentrations, and arrhythmias. We used anemone toxin (ATX)-II to study the effects of increasing INa,L on intracellular calcium cycling in rat isolated hearts. Cardiac contraction was abolished using paralytic agents. Ranolazine (RAN) was used to inhibit late INa. Hearts were loaded with fluo-4-acetoxymethyl ester, and myocyte intracellular calcium transients (CaTs) were measured using laser scanning confocal microscopy. ATX (1 nM) prolonged CaT duration at 50% recovery in hearts paced at a basal rate of 2 Hz and increased the sensitivity of the heart to the development of calcium alternans caused by fast pacing. ATX increased the time required for recovery of CaT amplitude following a previous beat, and ATX induced spontaneous calcium release waves during rapid pacing of the heart. ATX prolonged the duration of repolarization from the initiation of the activation to terminal repolarization in the pseudo-electrocardiogram. All actions of ATX were both reversed and prevented by subsequent or prior exposure, respectively, of hearts to RAN (10 μM). Most importantly, the increased vulnerability of the heart to the development of calcium alternans during rapid pacing was reversed or prevented by 10 μM RAN. These results suggest that enhancement of INa,L alters calcium cycling. Reduction by RAN of INa,L-induced dysregulation of calcium cycling could contribute to the antiarrhythmic actions of this agent in both reentrant and triggered arrhythmias.
- SR, sarcoplasmic reticulum
- RAN, ranolazine
- INa,L, late component of the rapid Na+ current
- ATX-II, anemone toxin II
- ECG, electrocardiogram
- pECG, pseudo-electrocardiogram
- LV, left ventricle
- X-t, spatial-temporal
- F − t, fluorescence − time
- X-Y, two-dimensional spatial recording
- F, fluorescence intensity
- F0, fluorescence intensity at rest
- CaT, intracellular calcium transient
- +dF/dt, first derivative of the fluorescence increase with time
- −dF/dt, first derivative of the fluorescence decrease with time
- TD50, duration of the calcium transient at 50% peak magnitude
- TD90, duration of the calcium transient at 90% peak magnitude
- AR, alternans ratio
- CL, cycle length
- BCL, basic cycle length
- ECL50, effective cycle length at which alternans ratio is 0.5
- R50, calculated cycle length at 50% recovery of calcium release
- QT90, time from the onset of activation to 90% recovery of repolarization
- IKr, delayed rectifier K+ current
- Nai, intracellular Na+
- Cai, intracellular Ca2+
- APD, action potential duration; peak width at 50% maximum.
The pathology of calcium overload in the heart includes both mechanical and electrical dysfunction (Vassalle and Lin, 2004). At slow rates of pacing, the reopening of calcium channels late in the action potential plateau may facilitate the occurrence of early afterdepolarizations, whereas at fast rates, abnormalities of calcium handling and intracellular calcium transients become manifest, including spontaneous release of calcium from the sarcoplasmic reticulum (SR), transient inward current, delayed afterdepolarizations and aftercontractions, calcium alternans, and triggered arrhythmic activity. There is growing evidence that rate-dependent alternations in SR calcium release (i.e., calcium alternans) contribute to action potential duration alternans (Walker and Rosenbaum, 2003; Pruvot et al., 2004; Wasserstrom et al., 2008; Kapur et al., 2009a,b), which may be responsible for T-wave alternans, the development of repolarization gradients, and re-entrant arrhythmias.
As a consequence of Na,Ca-exchange, an increase of sodium influx can lead to increases of calcium influx and intracellular calcium concentration and calcium overload in the heart (Verdonck et al., 1991; Verdonck et al., 1993). Cardiac sodium channels that fail to inactivate quickly after a brief opening allow a “late” sodium ion current flow (INa,L) and are a significant source of sodium influx into myocytes during the plateau of the cardiac action potential (Makielski and Farley, 2006). A number of congenital and acquired pathophysiological conditions are known to induce an increase in cardiac INa,L, including ischemia and heart failure (Zaza et al., 2008). An increase of cardiac INa,L has been associated with electrophysiological and mechanical dysfunction, including action potential prolongation, afterdepolarizations (early and delayed), triggered arrhythmic activity, and ventricular tachycardia (e.g., torsades de pointes), a slowing of diastolic contractile relaxation, and a reduction of coronary flow (Song et al., 2004; Fraser et al., 2006; Wu et al., 2006; Sossalla et al., 2008). Isolated cardiac myocytes have been used to demonstrate that an increase of INa,L leads to an increase in the concentration of intracellular calcium (Sossalla et al., 2008). However, isolated myocyte preparations have limitations for study of calcium cycling and calcium alternans, including the potential contribution of cell damage to calcium overloading, the absence of cell-cell coupling and its potential stabilization of cellular membrane potential and intracellular ionic changes, and the lack of stability of function of isolated myocytes beyond a brief period of time. Therefore, to clarify the effects of enhanced INa,L on calcium cycling in the heart, we have used the Ca-sensing dye fluo-4, confocal laser scanning microscopy, and an intact perfused rat heart preparation to study the effects of INa,L on calcium cycling.
The sodium channel toxin anemone toxin (ATX)-II was used to increase INa,L in this study. ATX is known to be a selective activator of INa,L (el-Sherif et al., 1992; Wasserstrom et al., 1993), and its effects on cardiac preparations have been described previously (Wasserstrom et al., 1993; Boutjdir et al., 1994). Because technical reasons prevent measurement of the effect of ATX to increase INa,L in the intact heart, we have used ranolazine (RAN) to confirm the role of INa,L in this study. RAN is an antianginal drug (Chaitman, 2006) that has been shown to inhibit INa,L in cardiac myocytes (Antzelevitch et al., 2004; Song et al., 2004; Undrovinas et al., 2006) and to antagonize actions of ATX in cardiac preparations (Wu et al., 2004, 2006). The goal of this study was to investigate whether RAN could reverse or antagonize the changes in calcium cycling caused by increased INa,L in the intact heart.
Materials and Methods
Langendorff Perfusion of Whole Rat Heart.
Animal usage in this study conformed to National Institutes of Health Guidelines and was approved by the Northwestern University Animal Care and Use Committee. Hearts of adult Sprague-Dawley rats of either sex were excised, cannulated at the aorta, and perfused retrogradely with normal bicarbonate-buffered modified Tyrode's solution (pH 7.3) via a Langendorff apparatus. Temperature and coronary perfusion rate were maintained at 25 ± 1°C and 8 ml/min, respectively. Atria were either removed or crushed to prohibit the intrinsic sinus node or atrioventricular node-mediated rhythm and P-waves. The heart was then positioned in an experimental chamber on the stage of an inverted Zeiss LSM510 confocal microscope such that the left ventricle was facing down and loaded via recirculating perfusion with 10 to 15 μM fluo-4-AM (Invitrogen, Carlsbad, CA) for 20 min and then washed for 5 to 10 min with modified Tyrode's solution. This loading protocol was repeated twice to obtain the desired level of fluo-4 fluorescence intensity (F, a relative measure of Cai) as monitored by confocal scanning. After dye loading, platinum hook-electrodes were inserted into the LV apex for external pacing, and Ag/AgCl2 electrodes were placed on either side of the heart for pseudo-ECG (pECG) recordings. The paralytic agents cytochalasin-D (50 μM) and blebbistatin (15 μM) were then added to the perfusate, which was recirculated for the remainder of the experiment. Cytochalasin-D is known to cause modest slowing in the rise and decay times of Ca2+ transients in isolated rat ventricular myocytes at high concentrations (Undrovinas and Maltsev, 1998). We use this combination of agents because the kinetics of Ca2+ transients are very rapid in whole heart and are nearly identical to those reported in single cells, and our recordings cannot be made in their absence. Experiments were performed at 25 ± 1°C, because at this temperature, greater intracellular retention of fluo-4 allows longer recording periods.
Superfusion with ATX and RAN.
ATX (Alomone Labs, Jerusalem, Israel) was added in the specified sequence (from 1 μM stock in distilled water) to the recirculating perfusate to give a 1 nM final concentration. RAN (CV Therapeutics, Inc., Palo Alto, CA) was added in the specified sequence (from a 10 mM stock in ethanol) to the recirculation perfusate to give a 10 μM working concentration, unless otherwise stated.
Pacing Protocols and Data Acquisition.
Cai cycling data from whole rat heart were acquired using confocal fluorescence microscopy as described previously (Aistrup et al., 2006; Kapur et al., 2009a,b). In brief, hearts were paced at a basic cycle length (BCL) at least 100 ms longer than that at which instabilities (e.g., calcium alternans) developed in Cai cycling under control conditions (BCL = 500 ms). These baseline pacing rates were always faster than the intrinsic rate. Confocal spatial (X)-spatial (Y) framescan and spatial (X)-temporal (t) linescan F image recordings were acquired from the midsection of the LV. Data acquisition consisted of transverse linescans made across the short axis of 10 to 20 myocytes, one to three cell layers beneath the epicardial surface. A Plan-Neofluor 25× water-immersion objective with a numerical aperture of 0.8 (Carl Zeiss Inc., Thornwood, NY) was used, giving a maximal resolution of ∼0.3 μm at a wavelength of 517 nm (the emission maximum for fluo-4). X-t linescan acquisitions were typically recorded during the following CaT assay pacing protocol: 2.5 s of baseline pacing → 10 s of test pacing → 2 to 3 s pause.
A series of linescan recordings was obtained from a heart paced at increasingly shorter BCL by decreasing the BCL by 20 ms per successive run until either 2:1 conduction block or arrhythmia occurred. Recordings were made from a single site per heart in hearts sequentially exposed to no drug (control), ATX, and ATX + RAN, or to no drug (control), RAN, and RAN + ATX.
Data Analyses.
Multicellular transverse X-t linescan recordings acquired under each experimental condition were analyzed off-line using the LSM-Examiner (Carl Zeiss Inc.) or ImageJ (NIH) image processing software. First, individual myocytes in X-t linescans were identified by aligning the X-t linescans with their corresponding X-Y framescans to identify cell edges for transverse X-t linescans, and then X-t linescan images were extracted for each myocyte, and the respective cellular mean F versus time (F-t) profiles were generated. F-t profiles were then used to determine the relative minimum diastolic Cai as measured by the minimum diastolic F measured in the 2 to 3 s after pacing, giving a best approximation of nominal F (F0) in each cell. Data were calculated as F/F0 to normalize for differential dye loading and/or variability of intrinsic diastolic Cai levels among cells. For measurement of ATX- or RAN-induced changes in diastolic Cai levels, all raw F measurements were normalized to the average of whole site F0 in the last run of the control series. Baseline CaT characteristics were determined from the 2.5-s baseline preceding rapid pacing and included: 1) peak systolic Cai (Fsystolic, measured as the averaged maximal F/F0); 2) CaT magnitude (measured as F/F0); 3) CaT duration at 50 and 90% recovery (TD50 and TD90); 4) CaT peak width at 50% peak amplitude (PW); 5) CaT rise time from 10 to 90% peak; 6) CaT decay time from 90 to 10% peak; 7) maximal +dF/dt (rate of calcium release); 8) maximal −dF/dt (rate of calcium reuptake); and 8) the CaT integral—Cai × t (as measured by F × t per cycle). CaT alternans ratios (AR) (Wu and Clusin, 1997) were calculated as:for six successive CaTs at steady state (end of 10-s test train) and averaged.
The equation used to fit the relationship between AR and cycle length waswhere CL = cycle length and ECL50 = effective cycle length at which half-maximal alternans is achieved. Triggered calcium activity was assessed by measuring the latency from the last paced stimulus in a pulse train to the first spontaneous calcium wave that occurred during the pause following stimulation. Because there is no unequivocally identifiable T-wave in the rat pECG, we designated the first inflection point in the wave form after the last sharply defined R-wave deflection as the starting point for late-phase repolarization. Two successive pECG waveforms in a recording were aligned at these initiation points for late-phase repolarization, and the net voltage difference between them in late-phase repolarization was calculated as repolarization alternans.
Results
Effects of ATX and RAN on Basal CaT Characteristics.
ATX had little effect on CaTs during basal pacing (BCL = 500 ms). Figure 1 shows images obtained from an LV epicardial recording site under control conditions, during exposure to 1 nM ATX and 10 min after addition of RAN (10 μM). Note that recordings were made transversely across 14 myocytes, each separated by a dashed white line. Each myocyte showed a brief CaT in response to stimulation (Fig. 1A), under control conditions, that was not changed in any significant way during exposure to ATX or during addition of RAN. When the data from all experiments were summarized (Fig. 1B), there were no significant changes in magnitude, rise time, decay time, TD90, and ±dF/dt of the CaT. However, the value of TD50 (CaT duration) was prolonged by ATX. Interestingly, the prolongation by ATX of TD50 was reversed by subsequent addition of RAN.
When the order of drug exposure was reversed and the heart was exposed first to RAN and then to ATX, there were no changes in CaT characteristics, including TD50 (Fig. 2). Thus the original images showed no clear changes in basal CaTs between control, RAN, and ATX (Fig. 2A), and the summary data (Fig. 2B) also demonstrate that there were no significant changes in CaTs induced by ATX after RAN pretreatment.
Effects of ATX and RAN on Pacing-Induced Calcium Alternans.
One of the most important dynamic properties of calcium cycling in cardiac myocytes is the development of rate-dependent calcium alternans. Calcium alternans increases with increasing rate of pacing. Results of two representative experiments are shown in Fig. 3. When the heart was stimulated at a BCL of 300 ms, alternans was minimal in the absence of drug, but it was increased in hearts perfused with 1 nM ATX (Fig. 3A). The magnitude of alternans (indicated by AR) increased with decreasing BCL of stimulation (Fig. 3C). The estimated midpoint of the relationship between BCL and AR (ECL50) occurred at a BCL value of 256 ms in the absence of the drug. ATX shifted the relationship between BCL and AR to the right; thus, alternans developed at lower pacing rates in the presence than in the absence of ATX (Fig. 3C), and the ECL50 value increased to 288 ms. The results demonstrate that ATX increased the vulnerability of the heart to calcium alternans. When RAN was added to the perfusate in the presence of ATX, there was a shift in the development of alternans back to shorter BCL (Fig. 3C), and the value of ECL50 decreased to 262 ms. When a second heart was pretreated with RAN (Fig. 3B), causing no change in the ECL50 (Fig. 3D), subsequent addition of ATX resulted in almost no shift in the rate dependence of AR (Fig. 3D).
Summary data from all experiments show that ATX caused a significant shift in the ECL50 and that RAN significantly reversed this shift back toward control levels (Fig. 4A). Furthermore, ECL50 was not altered by RAN alone, and after exposure of the heart to RAN, there was no additional shift in ECL50 during subsequent exposure to ATX in the continued presence of RAN (Fig. 4B). The magnitudes of the shift in ECL50 caused by ATX alone and by ATX following pretreatment with (and in the continued presence of) RAN are shown in Fig. 4C. ATX alone produced a significant shift in ECL50, which was nearly eliminated by previous exposure to RAN (p < 0.01).
These data demonstrate that ATX increases vulnerability to rate-induced cellular calcium alternans in whole heart. This effect of ATX can be both reversed and prevented by RAN. Thus, RAN is highly effective in antagonizing the effect of ATX to increase cellular susceptibility to calcium alternans induced by rapid pacing.
Effects of RAN on ATX-Induced Changes in Restitution of SR Calcium Release.
We have recently demonstrated that one of the most important determinants of calcium alternans is the rate of recovery (restitution) of SR calcium release (Wasserstrom et al., 2008). In general, interventions that slow the recovery of SR calcium release, such as prolonged CaTs, seem to also cause an increased vulnerability to rate-induced calcium alternans. Therefore, we investigated the effects of ATX on the restitution of SR calcium release in the absence and presence of RAN.
The effects of ATX on the restitution of SR calcium release are shown in Fig. 5A. The confocal images show the last beat during basal pacing followed by a progressively earlier test beat (and shorter recovery interval). As the recovery interval was shortened from 350 to 300, 250, and 150 ms in control, there was a stepwise decrease in the magnitude of calcium release during the test beat. The data summarizing the relationship for fractional release as a function of length of the recovery interval in this experiment are shown in Fig. 5B. In the presence of ATX, the relationship between the recovery interval and fractional recovery of SR calcium release was shifted to the right to decreased fractional recovery at any given recovery interval (Fig. 5B). Recovery was incomplete, even at 350 ms, in the presence of ATX. However, the addition of RAN (10 μM) in the continued presence of ATX increased fractional recovery of calcium release at all of the tested recovery intervals. In a separate experiment, when a heart was exposed to RAN before it was exposed to ATX (Fig. 5C), there was no change in restitution caused by either agent, and fractional recovery at each tested recovery interval was virtually identical in all treatment groups (Fig. 5D). Summary data in Fig. 6A show that ATX induced a significant slowing in restitution, as indicated by the increase (from control) of the calculated midpoint of the sigmoidal fit of the relationship between duration of the recovery interval and fractional recovery of calcium release (R50). Addition of RAN in the continued presence of ATX led to a significant decrease of the R50 value toward the control rate of recovery. Pretreatment with RAN blocked the slowing in the recovery rate of SR calcium release induced by ATX (Fig. 6B). Consequently, the significant slowing in SR calcium release caused by ATX alone was almost completely abolished by pretreatment with RAN (Fig. 6C; p < 0.01).
A slowing in SR recovery may provide a possible explanation for why ATX causes a shift in the rate dependence of calcium alternans to slower heart rates. The fact that RAN reverses the slowing by ATX of SR recovery rates may explain why RAN can reverse the effect of ATX to alter the rate dependence of calcium alternans. Furthermore, the fact that RAN pretreatment prevents the slowing in SR recovery rate caused by ATX may also explain its efficacy in preventing the increased vulnerability to calcium alternans.
Suppression of Calcium Waves by RAN.
One of the most common features of SR calcium overload is the development of spontaneous calcium waves following stimulation. Furthermore, the coupling interval of these calcium waves following rapid pacing shortens progressively as the severity of SR calcium overload increases. We investigated how RAN affects the coupling interval of ATX-induced calcium waves as an indicator of its ability to reverse the effects of ATX to induce calcium overload.
Figure 7A shows a confocal image in which calcium waves develop in all myocytes during exposure to ATX during the pause following a test train at BCL = 200 ms. The last stimulated beats of the test train are indicated by asterisks. The latency to the onset of these waves varies among myocytes, with some showing a short and others a long latency to onset of calcium release after pacing is terminated. After the addition of RAN in the continued presence of ATX (Fig. 7B), calcium waves still occurred but did so with a longer latency following termination of stimulation. The summary data shown in Fig. 7C demonstrate that the addition of RAN in the presence of ATX was associated with a significant increase (compared to ATX alone) in the latency to the appearance of the first calcium wave. These data demonstrate that RAN can reduce the severity of SR calcium overload induced by ATX, an effect that would result in reduced likelihood of calcium overload-induced triggered activity.
Effects of RAN on ATX-Induced Changes in Repolarization.
In addition to changes in calcium cycling, we also investigated the effects of RAN on ATX-induced prolongation of the ECG wave form. Because we recorded only a pseudo-ECG, there was no obvious T wave, and the duration of repolarization was measured from the initiation of activation of the QRS complex to terminal repolarization. ATX is known to prolong action potential duration (Isenberg and Ravens, 1984; Hoey et al., 1994) and the duration of the QT interval by prolonging repolarization (Wu et al., 2004), an effect that is apparent in Fig. 8A. The duration of repolarization was measured as QT90, the time from the onset of activation to 90% recovery of repolarization. The prolongation of repolarization by ATX was partially reversed by the subsequent addition of RAN (Fig. 8A). The summary data in Fig. 8B show that ATX produced a significant slowing of repolarization that was partially and significantly reversed by RAN. In contrast, RAN alone had no effect on the ECG waveform, and the addition of ATX in the presence of RAN caused almost no change in the ECG (Fig. 8C). The summary data shown in Fig. 8D demonstrate that RAN alone had no effect on the duration of repolarization and that pretreatment with RAN prevented the effect of ATX to slow repolarization. These data demonstrate that RAN antagonizes the effects of ATX on global electrophysiological properties of the heart, in particular, the overall duration of the ECG and the time to terminal repolarization. Thus, in addition to antagonizing the effects of ATX to alter calcium cycling, our data show that a similar pattern of behavior, namely reversal of the effects of ATX on repolarization by RAN and prevention of ATX effects by pretreatment with RAN, also occur with repolarization.
Discussion
The role of INa,L and the therapeutic potential use of specific blocking agents are becoming increasingly important in the treatment of a number of forms of cardiac disease. Ranolazine blocks INa,L with an IC50 value of 6 μM. It blocks the delayed rectifier K+ current, IKr, with an IC50 of 12 μM. The effects of ranolazine have been confirmed in hearts/myocytes from several species, including mouse, rat, guinea pig, rabbit, dog and human, and there does not seem to be any significant difference in the potency of ranolazine among these different species. However, it is noteworthy that the drug block of IKr is not relevant in rat where this current is virtually absent. Ranolazine also inhibits peak INa and L-type Ca2+ channel current at much higher concentrations (>200 μM). It is the most selective inhibitor of INa,L available to date (Hale et al., 2008). The ranolazine concentration of 10 μM that was used in the experiments described in this article is a concentration that has been achieved in patients treated in clinical trials with 1500 mg of ranolazine b.i.d [Cmax of 5710 ng/ml, which is 13.3 μM (Jerling, 2006)]. It is at the upper end of the range of recommended therapeutic concentration (which is 2–8 μM) when patients are treated with ranolazine at a twice daily dose of 1000 mg.
There have been numerous preclinical studies of ranolazine, and the concentration of ranolazine most often used in these studies is 10 μM. Rarely is a response noted below a concentration of 1 μM. The results become less relevant clinically at concentrations above 10 μM and potential off-target effects (e.g., IKr block, peak INa block in depolarized cells) become of greater concern, although the effect of IKr block is countered effectively by INa,L block and thus is not readily apparent (i.e., repolarization reserve was increased, not reduced, by ranolazine, and the incidence of arrhythmias in patients treated with ranolazine was lower than in control patients in the MERLIN-TIMI 36 outcomes trial).
The important findings of this study were that ATX (1 nM), a selective enhancer of late INa (Wanke et al., 2009), caused marked changes in calcium cycling in the intact heart and that these changes were reversed or prevented by ranolazine, an inhibitor of late INa. ATX treatment led to a shift in the rate dependence of calcium alternans, a slowing in the rate of restitution of SR calcium release as a function of the recovery interval following a previous stimulation, spontaneous calcium waves following termination of a stimulus train, and a slowing of repolarization of the heart. Ranolazine (10 μM) both reversed and prevented these effects of ATX. These data indicate that late INa can be a cause of calcium alternans and calcium overload pathology in myocytes in the intact heart.
Effects of ATX on Calcium Cycling.
The effects of ATX observed in the present study are likely to have implications with regard to enhancement of INa,L because it may occur in any of a number of pathophysiological conditions, including the INa-dependent form of the long QT syndrome (LQT3) (Bennett et al., 1995). In fact, pharmacological activation of INa,L by ATX causes changes in INa kinetics and APD similar to those seen in LQT3 models (Isenberg and Ravens, 1984; el-Sherif et al., 1992). The finding that ATX prolonged QT90 duration in this study is consistent with the assumption that ATX increased late INa and APD. The effect of ATX to increase late INa would be expected to increase sodium influx and, via NCX, to increase calcium influx and/or decrease calcium efflux. Thus, it was rather surprising that ATX had such modest effects on CaTs during basal pacing (BCL of 500 ms), with the only significant change being an increase in TD50. The increase of TD50 could be secondary to an ATX-induced increase of Nai+ that leads to a greater calcium influx during the plateau phase of the AP, thus contributing to TD50 but not overall TD90. The finding that changes in calcium cycling were much greater in the presence of 1 nM ATX at a BCL of 300 ms than at a BCL of 500 ms suggests that sodium and calcium overloading occurred only when the diastolic interval was too short to permit a return to baseline of the intracellular concentrations of sodium and calcium following an action potential.
Despite the unimpressive effect of ATX on the characteristics of CaTs during basal pacing, ATX increased the susceptibility of the heart to the development of calcium alternans—i.e., ATX caused calcium alternans to develop at significantly slower pacing rates compared to control. This shift is probably the result of the action of ATX to induce a slowing of the recovery (restitution) of SR calcium release following a paced beat (Figs. 5 and 6). These data suggest that ATX slowed the rate of restitution. Although the mechanism of action of ATX to slow the rate of restitution of SR calcium release is unclear, both direct mechanisms involving prolonged repolarization as well as indirect mechanisms involving an Na-induced increase in Cai are likely contributors. Whatever the mechanism, these results suggest that an increase of INa,L leads to an increased susceptibility to develop calcium and APD alternans. If so, increased INa,L could contribute to the induction of re-entrant arrhythmias.
In addition, ATX caused the induction of calcium waves in many myocytes, which is consistent with an effect of ATX to cause intracellular calcium overload. The effect of ATX to cause prolongation of APD and early afterdepolarizations can be distinguished from the effect of ATX to cause calcium overload and delayed afterdepolarizations (Song et al., 2008). Calcium overload is the likely mechanism for triggered activity consequent to a sustained increase in Na secondary to an increase in INa,L.
Calcium Alternans.
An ATX-induced slowing in the restitution of calcium release was associated with the development of calcium alternans. Because a slowing in the recovery of SR release would also produce instability in calcium release at rapid rates, an alternating large-small-large-small pattern of SR calcium release develops when heart rate is increased. Prolongation by ATX of TD50 is a potential cause of the slowing of the rate of restitution of CaT amplitude (Aistrup et al., 2006; Wasserstrom et al., 2008).
Calcium alternans may be linked to AP alternans and the development of repolarization gradients, which are a substrate for reentrant arrhythmias (Walker and Rosenbaum, 2003; Weiss et al., 2006). Consequently, an intervention such as enhancement of INa,L that shifts the development of calcium alternans to slower heart rates is likely to increase re-entry, whereas an intervention or drug that reverses or blocks this effect (e.g., RAN) is likely to be antiarrhythmic.
Effects of RAN to Antagonize the Rate-Dependent Effects of ATX on Calcium Cycling.
The effects of ATX on both restitution and calcium alternans susceptibility were both reversed by RAN and prevented by pretreatment with this agent. We interpret these results to indicate that RAN inhibition of increased INa,L induced by ATX antagonizes the changes in rate-dependent calcium cycling induced by the toxin. Presumably, the increases in Na and Ca that developed during increased INa,L were responsible for the changes in restitution and calcium alternans development. The block of the INa,L then allowed Na, Cai, and CaT restitution to return to normal levels over time, and calcium alternans was reduced. Therefore, our results in the present study are consistent with the known effect of RAN to block INa,L and suggest that interventions that promote INa,L and subsequent rate-dependent changes in calcium cycling could be antagonized by RAN.
An interesting implication of these results is that RAN could be antiarrhythmic in the settings of ischemia and congestive heart failure where the vulnerability to calcium alternans is also increased (Kapur et al., 2009a; Wasserstrom et al., 2009; Wilson et al., 2009) and where enhanced late INa has been demonstrated (Maltsev et al., 2001, 2007; Undrovinas et al., 2002, 2006; Valdivia et al., 2005), although a link between these two effects remains to be investigated. Finally, it is important to note that RAN also blocks human ether a-go-go-related gene channels and thus might induce some of its electrophysiological actions independently from the blockade of INa,L. One of the advantages of use of the rat heart in the present study is that there is no appreciable human ether a-go-go-related gene current in the rat. Consequently, the effects of RAN to reverse or block the effects of ATX to disrupt calcium cycling in this study are almost exclusively the result of its action to inhibit the enhanced INa,L induced by ATX.
Footnotes
-
This work was supported by a grant from Gilead Sciences, Inc.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
doi:10.1124/jpet.109.156471
-
ABBREVIATIONS:
- Received May 20, 2009.
- Accepted August 11, 2009.
- © 2009 by The American Society for Pharmacology and Experimental Therapeutics
References
- Aistrup et al., 2006.↵
- Antzelevitch et al., 2004.↵
- Bennett et al., 1995.↵
- Boutjdir et al., 1994.↵
- Chaitman, 2006.↵
- el-Sherif et al., 1992.↵
- Fraser et al., 2006.↵
- Hale et al., 2008.↵
- Hoey et al., 1994.↵
- Isenberg and Ravens, 1984.↵
- Jerling, 2006.↵
- Kapur et al., 2009a.↵
- Kapur et al., 2009b.↵
- Makielski and Farley, 2006.↵
- Maltsev et al., 2001.↵
- Maltsev et al., 2007.↵
- Pruvot et al., 2004.↵
- Song et al., 2008.↵
- Song et al., 2004.↵
- Sossalla et al., 2008.↵
- Undrovinas et al., 2006.↵
- Undrovinas and Maltsev, 1998.↵
- Undrovinas et al., 2002.↵
- Valdivia et al., 2005.↵
- Vassalle and Lin, 2004.↵
- Verdonck et al., 1991.↵
- Verdonck et al., 1993.↵
- Walker and Rosenbaum, 2003.↵
- Wanke et al., 2009.↵
- Wasserstrom et al., 2008.↵
- Wasserstrom et al., 1993.↵
- Wasserstrom et al., 2009.↵
- Weiss et al., 2006.↵
- Wilson et al., 2009.↵
- Wu et al., 2006.↵
- Wu et al., 2004.↵
- Wu and Clusin, 1997.↵
- Zaza et al., 2008.↵