INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

TWO COMPONENTS of the transient outward current (I_to) related to the initial repolarization of cardiac action potential have been described in various cardiac tissues such as sheep (7) and calf Purkinje fibers (20), rabbit atrium (9) and ventricles (9, 15), and canine ventricles (44). One component is carried by K⁺ and is sensitive to 4-aminopyridine (4-AP). The other component, which is resistant to 4-AP, was shown to be abolished by Ca²⁺ channel blockers or agents interfering with Ca²⁺ release from the sarcoplasmic reticulum (SR), suggesting that it is primarily triggered by Ca²⁺-induced Ca²⁺ release (CICR) from the SR (15, 32, 44). More recently, Cl have been identified to be the charge carrier of the 4-AP-resistant component of I_to (39, 47-49) and thus referred to as the Ca²⁺-activated Cl current [I_Cl(Ca)] by Zygmunt and Gibbons (48). A 1.0- to 1.3-pS Cl-selective channel that is hypothesized to underlie I_Cl(Ca) has been identified at the single-channel level in canine ventricular myocytes (6). This channel is thought to be mainly gated by changes in the Ca²⁺ concentration on the cytoplasmic side of the membrane ([Ca²⁺]_i) (6, 47) and would thus belong to the intracellular ligand-gated class of ion channels. In rabbit ventricular myocytes, I_Cl(Ca) is a major part of I_to at physiological stimulation rates (15) and is hypothesized to play a significant role in the rate- and rhythm-dependent repolarization of the cardiac action potential (47). Recent evidence suggests that a CICR process triggered by reverse-mode Na/Ca exchange may also elicit I_Cl(Ca) in this preparation (22, 23).

CICR, a process characterized by small Ca²⁺ entry across the sarcolemma triggering the release of a much greater quantity of Ca²⁺ from the SR (8), is now widely accepted to be the main mechanism underlying the excitation-contraction coupling (EC coupling) in cardiac muscle (2, 34, 35). At least five distinct mechanisms are hypothesized to be able to elicit the release of Ca²⁺ from the SR in the heart: 1) Ca²⁺ entry through L-type calcium channels [I_Ca(L)] is recognized as the primary triggering source of Ca²⁺ eliciting CICR (2, 31, and for review see Ref. 1) and to replenish SR Ca²⁺ stores on a beat-to-beat basis (1); 2) Ca²⁺ entry through the Na⁺/Ca²⁺ exchanger operating in reverse mode in response to membrane depolarization (21, 28, 30) and/or promoted by transient accumulation of Na⁺ in a subsarcolemmal compartment caused by Na⁺ influx through Na⁺ channels (I_Na) during the upstroke of cardiac action potential (24, 25, 27, 29); 3) Ca²⁺ entry through T-type Ca²⁺ channels (41); 4) Ca²⁺ permeation through tetrodotoxin (TTX)-sensitive Na⁺ channels favored by activation of the adenylate cyclase pathway (36); and 5) a putative voltage-dependent mechanism that is distinct from CICR mediated by I_Ca(L), the Na/Ca exchanger, or I_Na (17).

Among these postulated mechanisms of cardiac EC coupling, the I_Na-mediated SR Ca²⁺ release initially reported in guinea pig ventricular myocytes remains controversial. Major arguments against this hypothesis come from experiments carried out in rat ventricular myocytes (4, 38), and these experiments suggest that a loss of voltage control during I_Na activation may bring membrane potential to voltages where I_Ca(L) is activated, which then triggers CICR. An alternative explanation proposed for these experimental discrepancies is that species differences exist in the density of the Na⁺/Ca²⁺ exchanger protein and its spatial distribution with respect to the ryanodine receptors (25, 37).

The purpose of this study was twofold: 1) to test the hypothesis that I_Na can trigger SR Ca²⁺ release in the rabbit ventricle, and 2) to evaluate whether I_Na-mediated SR Ca²⁺ release is capable of activating I_Cl(Ca). Our study shows that in the absence of -adrenergic stimulation, as well as L- and T-type Ca²⁺ currents, the TTX-sensitive Na⁺ current can elicit Ca²⁺ transients and 4-AP-insensitive I_to sharing many properties with those of the I_Cl(Ca) previously identified in this preparation. Our results support the notion that I_Na-mediated Ca²⁺ transients and I_Cl(Ca) play a significant role in EC coupling and modulation of cardiac action potential repolarization of the rabbit ventricle under physiological conditions. A preliminary account of these findings has been presented in abstract form (43).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell isolation. New Zealand White rabbits (2.8-3.5 kg) of either sex were stunned about 30 min after being given an intraperitoneal injection of heparin sodium (500 IU/kg). The heart was quickly removed and mounted on a Langendorff apparatus for retrograde perfusion with 1) Tyrode solution (see Solutions and drugs for the composition) containing 1.8 mM CaCl₂ for 5 min; 2) nominally Ca-free Tyrode solution for 5 min; and 3) nominally Ca-free Tyrode solution containing 223 U/ml collagenase (Type 2, Worthington Biochemical) for 40-55 min. All perfusates were saturated with 100% oxygen and warmed to 37°C. After perfusion with enzymes, the left ventricle was cut off, rinsed, and minced in a fresh enzyme-free Tyrode solution containing 0.1 mM CaCl₂ and 0.2% bovine serum albumin. Rod-shaped myocytes were mechanically dispersed by gentle agitation and then stored at room temperature in the same solution. Cells were used within 6-8 h after isolation.

Solutions and drugs. The Tyrode solution used for isolating cells contained (in mM) 136 NaCl, 5.0 KCl, 1.0 MgCl₂, 1.8 CaCl₂, 0.33 NaH₂PO₄, 10 glucose, and 10 HEPES; pH adjusted to 7.2 with NaOH. The nominally Ca²⁺-free Tyrode solution had the same composition except that CaCl₂ was omitted.

The normal external solution (NES) used to superfuse the myocytes contained (in mM) 136 NaCl, 5.4 KCl, 1.0 MgCl₂, 1.8 CaCl₂, 0.33 NaH₂PO₄, 5.5 glucose, and 10 HEPES; pH adjusted to 7.35 with NaOH. The Cs⁺-containing external solution had the following composition (in mM): 130 NaCl, 0.33 NaH₂PO₄, 10 CsCl, 1.0 MgCl₂, 2.5 CaCl₂, 10 HEPES, 5.5 glucose, 5 4-AP; pH adjusted to 7.35 with 75 µl of HCl (10 N) and NaOH. The Ca²⁺-free Cs⁺ external solution had the same composition except that CaCl₂ was substituted by MgCl₂ (2.5 mM) and 0.1 EGTA was added. The Na⁺-free external solution contained (in mM) 125 N-methyl-D-glucamine chloride (NMDG-Cl), 10 CsCl, 1.0 MgCl₂, 2.5 CaCl₂, 10 HEPES, 5.5 glucose, and 5 4-AP (pH 7.35 with CsOH).

The composition of the low-Cl pipette solution was as follows (in mM): 120 CsOH, 120 aspartic acid, 20 tetraethylammonium chloride (TEA-Cl), 5 ATP-Mg, 1.0 MgCl₂, 5 HEPES, 0.1 GTP-Na₂, and 5 phosphocreatine-Na₂; pH adjusted to 7.2 with CsOH. In some experiments, 100 mM potassium gluconate was used to replace cesium aspartate, because Ca²⁺ transients quickly ran down in a large proportion of cells dialyzed with high concentration of Cs⁺. The same observation has been reported by others (12). The 4-AP-resistant I_to recorded with both pipette solutions did not appear different. The high-Cl pipette solution had essentially the same composition as that of the low-Cl pipette solution except that cesium aspartate and potassium gluconate were replaced by 20 mM CsCl and 100 mM KCl. The pipette solution used for examining the effects of nicardipine on I_Ca(L) contained (in mM) 110 CsOH, 110 aspartic acid, 20 TEA-Cl, 1 MgCl₂, 5 ATP-Mg, 5 HEPES, and 10 EGTA; pH adjusted to 7.2 with CsOH.

Nicardipine, niflumic acid (Sigma Chemical), and KB-R7943 (46) were dissolved in DMSO (Sigma Chemical) to obtain stock solutions of 10, 200, and 1 mM, respectively. Ryanodine (10 mM, Calbiochem) and TTX (30 mM; Sigma Chemical) stock solutions were prepared in distilled water. Appropriate aliquots of the stock solutions were added to external solutions to obtain the desired concentrations. The final concentration of DMSO in the perfusate did not exceed 0.1%. Caffeine (Sigma Chemical), 4-AP (Sigma Chemical), and FS-2 (Alomone Labs) were directly dissolved in Cs⁺ external solutions.

Electrophysiological measurements. Macroscopic currents were recorded with the whole cell variant of the patch-clamp technique. Patch pipette resistances ranged from 2 to 3 M when filled with internal solutions listed above. The voltage command and current measurements were performed through a patch-clamp amplifier (Axopatch-1D, Axon Instruments) driven by a PC type of computer using pCLAMP software (version 5.5, Axon Instruments) and interfaced with a D/A and A/D converter (TL-1 DMA Interface, Axon Instruments). Membrane currents (along with the Indo-1 ratio signal, see Measurement of cytoplasmic Ca²⁺ concentration) were sampled at 2 kHz and low-pass filtered at 1-2 kHz. Series resistance was electronically compensated by 63% (6.08 ± 0.54 M and 2.13 ± 0.20 M before and after compensation, n = 16). Liquid junction potentials between bath and pipette solutions were measured to be 14.5 (n = 11, pipette negative) for low-Cl pipette and 0 mV (n = 5) for high-Cl pipette solutions. The voltage values given in the text for low-Cl pipette solution were corrected for this offset assuming that the junction potentials between pipette solutions and the cytoplasm were null.

Measurement of cytoplasmic Ca²⁺ concentration. To measure intracellular Ca²⁺ transients, Indo-1 (50 or 100 µM) was dialyzed into the cytoplasm through the patch pipette. The experimental apparatus constructed around a Nikon Diaphot inverted microscope has been previously described (42). Briefly, intracellular Indo-1 was excited at 340 nm, the fluorescence emitted at 400 nm and 500 nm was filtered separately at 60 Hz, and the ratio (R_400/500) was converted into digital format (2 kHz) and delivered to the computer simultaneously with membrane currents. In this study, R_400/500 was used as an index of cytoplasmic Ca²⁺ concentration ([Ca²⁺]_i). The background and cell autofluorescence were corrected by zeroing the photomultiplier tubes after seal formation and before gaining whole cell access. Cells were only exposed to ultraviolet light during voltage steps.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Efficacy of block of I_Ca(L) by nicardipine. To test the hypothesis that I_Na can induce a Ca²⁺ transient and I_Cl(Ca) in rabbit ventricular myocytes, it is a prerequisite to ensure that I_Ca(L) can be abolished under our experimental conditions. In this study, nicardipine was used to block I_Ca(L). The efficacy of I_Ca(L) inhibition by nicardipine was examined using a Na-free external solution to eliminate overlapping I_Na. Complete block of I_Ca(L) by 10 µM nicardipine could be achieved when the holding potential was equal to or more positive than 75 mV (corrected for the liquid junction potentials between bath and pipette solutions, n = 5). For the experiments carried out in the presence of 131 mM external Na⁺ presented in the following sections, myocytes were therefore held at 75 mV, and a prepulse (500 ms) to 105 mV was applied immediately before the test pulse in an attempt to increase the availability of Na⁺ channels while maintaining a satisfactory block of I_Ca(L) by nicardipine. The potency of nicardipine in blocking I_Ca(L) with the use of such a voltage-clamp protocol (Fig. 1A) was first examined in cells superfused with Na⁺-free solution. Figure 1A illustrates a representative time course of block by nicardipine of I_Ca(L) elicited by test pulses to +5 mV. The current traces depicted in the inset of Fig. 1A were recorded at the times indicated. Nicardipine (10 µM) abolished I_Ca(L) 3 min after being switched to a drug-containing medium. In a total of five cells, full block was achieved by 2.5-4 min (2.8 ± 0.3 min) after application of nicardipine. The mean current-voltage (I-V) relationship (n = 5) after exposure to nicardipine is plotted in Fig. 1B. The inset shows a family of membrane currents recorded at different depolarizing voltages in a cell exposed to nicardipine for 3 min. No inward current was apparent in the presence of nicardipine, indicating complete block of I_Ca(L) by the drug under our experimental conditions. Our data also suggest that T-type Ca²⁺ current is absent in the rabbit ventricle as previously reported by others (11) and did not contribute to SR Ca²⁺ release in our experiments.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Effects of nicardipine on L-type Ca2+ current [ICa(L)] in rabbit ventricular myocytes. A: representative time course of block of ICa(L) by 10 µM nicardipine (Nic) examined in a Na+-free (substituted by N-methyl-D-glucamine) external solution. Line represents a fit of data points to a spline function. Inset: examples of current traces (a-c) recorded at times indicated. Voltage-clamp protocol (inset) was similar to that used for membrane current (Im)recordings in the following studies except that the second brief prepulse for activation of Na2+ current (INa) was omitted. B: mean current-voltage relationship obtained from 5 cells exposed to 10 µM nicardipine using voltage-clamp protocol shown in inset. For each step, absolute current level was measured immediately following complete relaxation of capacitative discharge. A typical set of current traces recorded from one cell exposed to nicardipine is given in inset.

4-AP-resistant I_to recorded in absence of I_Ca(L). Figure 2A shows membrane currents and Ca²⁺ transients recorded from a cell in the absence and presence of 10 µM nicardipine. The cell was continuously superfused with 5 mM 4-AP. A prepulse (5 ms) to 45 mV from the first step command to 105 mV was used to activate I_Na. Under control conditions, an outward current was elicited during the test pulse (+40 mV) followed by a slower overlapping inward current component. A robust Ca²⁺ transient was also elicited during the step to +40 mV. Nicardipine abolished the inward current during the test pulse, leaving a transient outward current component. The Ca²⁺ transient was also significantly reduced but not abolished. In 24 cells, the amplitude of Ca²⁺ transients was reduced by 61 ± 3% in the presence of nicardipine (R_400/500: 0.70 ± 0.06 vs. 0.25 ± 0.03, P < 0.01); where R_400/500 = increase in R_400/500 measured with respect to its resting level), with resting Ca²⁺ unchanged (R_400/500: 0.66 ± 0.03 vs. 0.67 ± 0.04, P > 0.05). The inward component recorded in the absence of nicardipine during the test pulse is consistent with I_Ca(L). The outward component resistant to 4-AP and nicardipine was recorded in 64 of 81 cells (79%). The time to peak of the 4-AP- and nicardipine-insensitive I_to elicited at +35 and +40 mV ranged from 8 to 20 ms, with an average value of 12.8 ± 0.5 ms (n = 42).

View larger version (25K): [in this window] [in a new window]   Fig. 2.   4-Aminopyridine (4-AP)-resistant transient outward current (Ito) and Ca2+ transients recorded in absence of ICa(L). A: membrane currents and Ca2+ transients recorded from a cell before (C) and after 3 min exposure to 10 µM Nic, elicited by a depolarization to +40 mV after a brief prepulse to 45 mV to activate INa. B: membrane currents and Ca2+ transients elicited at different depolarizing voltages from another cell exposed to 10 µM Nic for 8 min. C: voltage dependence of 4-AP-resistant Ito () and Ca2+ signals () obtained from 7 cells with voltage-clamp protocol (insets) shown in B. Current amplitude was measured as difference between outward peak and sustained level at end of pulse. Line represents a linear fit to data points. R400/500 was not significantly voltage dependent. R400/500, ratio of fluorescence emitted at 400 and 500 nm used as an index of cytoplasmic Ca2+ concentration.

Charge carrier of 4-AP-resistant I_to recorded in absence of I_Ca(L). To test whether the 4-AP- and nicardipine-resistant I_to observed in this study reflects the same Cl conductance described by others in the same preparation with an intact I_Ca(L) (19, 48), we then examined the Cl selectivity of the current. To evaluate ionic selectivity, we measured the reversal potential (E_rev) of the current in myocytes dialyzed with pipette solutions containing two different Cl concentrations. Figure 3, A and B, shows two representative experiments performed with low (21 mM)- and high (141 mM)-Cl pipette solution, respectively. As in previous experiments, the cells were superfused with 5 mM 4-AP, and the recordings were made after 4 min of exposure to 10 µM nicardipine. After an initial 5-ms pulse to elicit I_Na and Ca²⁺ transient, a double test pulse protocol was applied (Fig. 3, A and B, inset), which was composed of a brief (8-12 ms) depolarization to +25 or +40 mV to fully activate the 4-AP- and nicardipine-resistant I_to, followed by a repolarizing step to different voltages to elicit deactivating tail currents. The E_rev was then determined by measuring tail current amplitude. Figure 3, A and B, shows that the current reversed near 35 mV and 0 mV with low-Cl and high-Cl pipette solution, respectively. The [Cl]_i-sensitive shift in E_rev of the 4-AP-resistant I_to supports the idea that the underlying channel is permeable to Cl. Similar results were obtained in six cells for each pipette Cl concentration. In Fig. 3, C and D, tail current amplitude was plotted against repolarizing voltage. Symbols indicate different experiments. The mean E_rev of the current was 42.0 ± 1.3 mV with low [Cl]_i and +0.3 ± 1.4 mV for the high [Cl]_i, similar to the predicted E_rev with low [Cl]_i (47.8 mV) and high [Cl]_i (1.2 mV), respectively. These results suggest that the 4-AP- and nicardipine-resistant I_to described here is mainly carried by chloride ions. A close correlation between the measured and predicted E_rev values also suggests, at most, minor contamination from Na⁺/Ca²⁺ inward tail currents to our measurements.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Chloride sensitivity of 4-AP- and Nic-resistant Ito. A and B: representative membrane currents and Ca2+ transients recorded using a double-test pulse protocol (inset) from two different cells exposed to 10 µM nicardipine, dialyzed with 21 mM (low [Cl]i) and 141 mM (high [Cl]i) Cl pipette solutions, respectively. C and D: plots of tail current as a function of voltage during second test pulse for each of 12 cells dialyzed with low (C, n = 6) and high (D, n = 6) Cl pipette solutions. Each symbol represents one cell. Amplitude of tail current was measured as difference between outward peak or current level at 5 ms after onset of repolarization and sustained current at end of pulse.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Effects of niflumic acid and caffeine on Nic-insensitive Ca2+ transients and 4-AP- and Nic-resistant Ito. A: membrane currents and Ca2+ transients elicited by voltage-clamp protocol (inset) from a cell in absence (left) and presence (right) of 100 µM niflumic acid. Cell was continuously superfused with 10 µM nicardipine for 7 min. B: membrane currents and Ca2+ transients recorded using voltage-clamp protocol (inset) from a different cell before (left) and 1 min after switching to 10 mM caffeine (right). Cell was continuously superfused with 10 µM Nic for 5 min.

Does 4-AP- and nicardipine-resistant Cl current depend on SR Ca²⁺ release? Previous studies have indicated that I_Cl(Ca) is dependent on functional SR Ca²⁺ release. Figure 4B shows the results of an experiment in which we tested the effects of 10 mM caffeine, which is known to quickly empty the SR Ca²⁺ stores (1) on 4-AP-resistant I_to and Ca²⁺ transient recorded in a myocyte bathed with 10 µM nicardipine. The I_to and Ca²⁺ transient were abolished 45 s after being switched to a caffeine-containing solution. The resting Ca²⁺ level was increased slightly as expected from the inhibitory effect of caffeine on SR Ca²⁺ uptake. Similar results were obtained in six myocytes. The resting ratio increased from 0.85 ± 0.01 to 1.12 ± 0.04 (P < 0.01) shortly after application of 10 mM caffeine. We also tested the effects of ryanodine, which inhibits SR Ca²⁺ release by locking the Ca²⁺ release channel in a subconductance state. In three cells incubated with 10 µM ryanodine for 45 min, neither I_to nor Ca²⁺ transient could be recorded in the presence of 5 mM 4-AP and 10 µM nicardipine. These results suggest that the small Ca²⁺ transient recorded after blocking I_Ca(L) is due to Ca²⁺ released from the SR, and that the 4-AP- and nicardipine-resistant I_to is dependent on the intracellular Ca²⁺ transient and will thus be referred to as I_Cl(Ca).

Role of I_Na in activation of nicardipine-resistant I_Cl(Ca). Figure 5A shows membrane currents and Ca²⁺ transients recorded from a cell exposed to nicardipine for 5 min. A transient outward I_Cl(Ca) was elicited by depolarization to +35 mV following activation of I_Na during a brief step to 45 mV, accompanied by a phasic Ca²⁺ transient. Cell exposure to 30 µM TTX almost abolished I_Na and strongly suppressed the Ca²⁺ transient and I_Cl(Ca). Similar results were obtained in nine cells. Figure 5B shows membrane currents and Ca²⁺ transients recorded from another cell to which two voltage-clamp protocols (test pulses shown in insets) were consecutively applied in the presence of nicardipine. Each test pulse was preceded by a train of five conditioning pulses. With a brief step to 45 mV, which elicited I_Na, a transient outward I_Cl(Ca) was evoked during the subsequent depolarization to +45 mV. When the membrane was directly depolarized from 105 to +45 mV (near the reversal potential for I_Na), neither I_Na nor I_Cl(Ca) was observed. Similar results were obtained in six cells. These findings suggest that fast Na⁺ entry through Na⁺ channels is required to activate the nicardipine-resistant I_Cl(Ca).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Role of INa in activation of Nic-insensitive ICl(Ca) and Ca2+ transients. A: membrane currents and Ca2+ transients recorded from a cell exposed to 10 µM nicardipine for 5 min before (left) and 1 min after switching to 30 µM tetrodotoxin (TTX, right). B: membrane currents and Ca2+ transients recorded from a cell during 2 consecutive voltage-clamp pulses (each preceded by 3 conditioning pulses) with (left) or without (right) second brief prepulse to 45 mV (see insets). Cell was exposed to 10 µM Nic for 6 min.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Voltage dependence of INa (A) and Ca2+ transients (B) in presence of 10 µM Nic. Pooled data obtained from 5 cells before () and after () application of 30 µM tetrodotoxin (TTX) are shown. Cells were held at 75 mV. A train of 3-5 conditioning pulses to +80 mV preceded each voltage-clamp protocol consisting of a prepulse of 500 ms to 105 mV and a test pulse to different voltages. Amplitude of Ca2+ transient was measured as difference of R400/500 between resting state and 50 ms after onset of test pulse.

Role of extracellular Ca²⁺ in activating nicardipine-resistant I_Cl(Ca). We then examined whether extracellular Ca²⁺ is required for the activation of nicardipine-resistant I_Cl(Ca). In these experiments, the solution surrounding myocytes was rapidly switched (<300 ms) from Ca²⁺-containing to Ca²⁺-free media immediately before the test pulses using a fast flow system to avoid SR Ca²⁺ depletion in the absence of external Ca²⁺. Both I_Cl(Ca) and Ca²⁺ transient induced by the test pulse were abolished by removal of extracellular Ca²⁺ (n = 6, data not shown). These results indicate that the I_Na-mediated Ca²⁺ transient and I_Cl(Ca) in rabbit myocytes have an absolute requirement for extracellular Ca²⁺, as similarly reported for I_Na-induced Ca²⁺ release from the SR in guinea pig ventricular myocytes (24, 27).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study reports a transient increase in [Ca²⁺]_i associated with the activation of I_Na in rabbit ventricular myocytes under conditions that prevent Ca²⁺ flow through L-type channels. These Ca²⁺ transients are sensitive to caffeine or ryanodine and depend on the presence of extracellular Ca²⁺ when I_Ca(L) is abolished by 10 µM nicardipine. Our study also demonstrates that I_Na-induced SR Ca²⁺ release is able to activate I_Cl(Ca) and thus potentially contribute to early repolarization in this species.

4-AP- and nicardipine-resistant I_to is carried by Ca²⁺-activated Cl channels. A 4-AP-resistant I_to was recorded in rabbit ventricular myocytes exposed to 10 µM nicardipine to block I_Ca(L). Its properties are consistent with those of the 4-AP-resistant I_to activated by Ca²⁺ current (6, 19, 39, 47-49) or reverse-mode Na⁺/Ca²⁺ exchange (22, 23): 1) it was abolished by caffeine or ryanodine, agents known to empty the SR Ca²⁺ stores (1), 2) the reversal potential of this current shifted in a manner consistent with that expected from manipulations of the transmembrane Cl gradient, and 3) it was inhibited by niflumic acid, a known blocker of I_Cl(Ca) in many cell types, including cardiac myocytes (6).

I_Na-mediated SR Ca²⁺ release. Since the first report on I_Na-induced Ca²⁺ transient in guinea pig ventricular myocytes (24), a potential role for this mechanism in cardiac EC coupling has been the subject of debate over the past few years. It has been proposed that a transient accumulation of Na⁺ in a subsarcolemmal compartment during the activation of I_Na, in combination with membrane depolarization, promotes Ca²⁺ influx through reverse-mode Na⁺/Ca²⁺ exchange, which would then trigger Ca²⁺ release from the SR (24, 26). This hypothesis was based on the following observations: 1) TTX suppressed I_Na and Ca²⁺ transients elicited by depolarizations from 70 to 40 mV and reduced Ca²⁺ transients elicited by action potentials in the absence (24) or presence of a Ca²⁺ channel blocker (27); and 2) the I_Na-mediated Ca²⁺ transient was dependent on a functional SR and extracellular Ca²⁺ (24). A role for the Na⁺/Ca²⁺ exchanger in coupling the activation of I_Na to SR Ca²⁺ release was further demonstrated in guinea pig ventricular myocytes by revealing a residual I_Na-induced Ca²⁺ transient in the absence of Ca²⁺ release from the SR (29). With the fast Ca²⁺ indicator fluo-3 and the line-scan mode of a confocal microscope, Lipp and Niggli (29) showed that this residual I_Na-induced Ca²⁺ transient was insensitive to verapamil but was abolished after substitution of extracellular Na⁺ by Li⁺. I_Na-induced contraction was also reported in cat ventricular myocytes following complete blockade of I_Ca(L) with 10 µM nifedipine (45). In contrast to the above studies, I_Na-induced SR Ca²⁺ release in rat ventricular myocytes was found to be sensitive to the Ca²⁺ channel blocker Cd²⁺ but insensitive to substitution of extracellular Na⁺ by Li⁺ (4, 38). It was concluded that apparent I_Na-induced Ca²⁺ release was an artifact resulting from the loss of voltage control occurring during the activation of I_Na, which would briefly bring membrane potential into the voltage range of activation of I_Ca(L), thereby triggering SR Ca²⁺ release. A recent study in guinea pig ventricular myocytes also argued against the presence of an I_Na-induced Ca²⁺ release mechanism by concluding that the inhibition of I_Na-related Ca²⁺ transient by TTX is due to a depletion of SR Ca²⁺ stores caused by a reduction of intracellular Na⁺ concentration following suppression of I_Na by TTX (40). Recently, Santana et al. (36) suggested an alternative mechanism whereby cardiac Na⁺ channels may become permeable to Ca²⁺ during stimulation of the adenylate cyclase pathway, leading to cAMP accumulation and subsequent activation of protein kinase A. The increase in Ca²⁺ permeability was shown to be sufficient to trigger SR Ca²⁺ release as manifested by the enhancement of the probability of detecting Ca²⁺ sparks during -adrenergic stimulation.