Ca²⁺-activated non-selective cation current in rabbit ventricular myocytes

Abstract

1. Oscillatory currents (OCs) were studied in isolated rabbit ventricular myocytes with whole cell mode voltage clamp using Na⁺-free intracellular and extracellular solutions under conditions where K⁺ currents were anticipated to be eliminated or minimized.

2. All OCs were dependent on release of Ca²⁺ from the sarcoplasmic reticulum (SR) because they were associated with intracellular Ca²⁺ ([Ca²⁺]_i) transients, and were suppressed by high concentrations of BAPTA (20 mmol l⁻¹) or pretreatment with the SR antagonist agents ryanodine (10 μmol l⁻¹) or thapsigargin (1 μmol l⁻¹).

3. The reversal potential (V_rev) for OCs shifted with changes in the calculated V_rev for Cl⁻ (E_Cl) but was between E_Cl and the calculated V_rev for elemental monovalent cations (E_Cat), indicating that more than one Ca²⁺-activated current contributed to OCs.

4. Addition of the Ca²⁺-activated Cl⁻ current (I_Cl(Ca)) antagonist, niflumic acid, shifted the OC V_rev to E_Cat, suggesting that I_Cl(Ca) and a Ca²⁺-activated non-selective cation current (I_CAN) contributed to the observed OCs.

5. A reduced niflumic acid-insensitive Ca²⁺-activated OC persisted following marked symmetrical reduction of Cl⁻ in the intracellular and extracellular solutions. Subsequent removal of all extracellular monovalent cations, by N-methyl-D-glucamine (NMDG) substitution, eliminated OCs and the inward holding current suggesting that I_CAN and I_Cl(Ca) accounted for all or most of the Ca²⁺-activated OC in the absence of Na⁺.

6. The OC V_rev was equal to E_Cl in the absence of monovalent elemental cations. Under these conditions niflumic acid eliminated all OCs.

7. Macroscopic OC is partially due to I_CAN in rabbit ventricular myocytes.

Ca²⁺-activated oscillatory currents (OCs) in cardiac muscle are likely to be important in the genesis of cardiac arrhythmias. Using physiological solutions, most observed OC seems to be accounted for by the electrogenic Na⁺-Ca²⁺ exchanger and by the Ca²⁺-activated chloride current (I_Cl(Ca)) (Zygmunt et al. 1998; Wu et al. 1999). Other potential sources of Ca²⁺-activated OC include inactivation of the L-type Ca²⁺ current (Sipido et al. 1995), inhibition of the inward rectifier current (I_K1) (Zaza et al. 1998) and a Ca²⁺-activated non-selective cation current (I_CAN) (Colquhoun et al. 1981; Ehara et al. 1988; Magishi et al. 1996).

I_CAN was first described in cultured rat cardiomyocytes using single channel recordings where it was found to lack voltage-dependent gating and to be equally permeant to Na⁺ and K⁺ (Colquhoun et al. 1981). Later workers using whole cell mode voltage clamp techniques also found evidence for I_CAN in Purkinje fibres (Ehara et al. 1988) and ventricular myocytes (Giles & Shimoni, 1989). Another non-selective cation current has also been recorded in ventricular myocytes subjected to oxidative stress (Jabr & Cole, 1995) or following lysophosphatidyl choline-induced elevation of intracellular Ca²⁺ concentration ([Ca²⁺]_i) (Magishi et al. 1996). However, some whole cell mode voltage clamp studies in ventricular myocytes have not supported the existence of I_CAN (Sipido et al. 1995; Laflamme & Becker, 1996). Failure to implicate I_CAN in some previous studies (Sipido et al. 1995; Laflamme & Becker, 1996) may be partially due to the lack of a specific antagonist, or differences in species or other experimental conditions. The present voltage clamp studies were performed in the absence of Na⁺ and after blockade of K⁺ currents in rabbit ventricular myocytes in whole cell mode. The hypothesis that I_CAN is evident as a macroscopic current was tested by independently manipulating E_Cat and E_Cl.

METHODS

Cell isolation

Single ventricular myocytes were prepared as previously described (Fedida et al. 1993). In brief, New Zealand White rabbits (2–3 kg) were killed by pentobarbitol overdose (50 mg kg⁻¹, i.v.), in accordance with institutional guidelines. The hearts were rapidly excised and placed in an ice-cold nominally Ca²⁺-free Hepes-buffered saline solution. The aorta was cannulated and the heart was perfused in Langendorff fashion with a nominally Ca²⁺-free perfusate for 10 min at 37°C. This was followed by a 25 min perfusion with collagenase-containing solution, and finally by a 10 min perfusion with a low Ca²⁺ (0.2 mmol l⁻¹) solution. After the enzyme was washed out, the left ventricle was coarsely minced and placed in a beaker containing low Ca²⁺ solution with 0.5 % (w/v) bovine serum albumin at 37°C. The left ventricular myocytes were dispersed by gentle agitation and then maintained in the standard saline solution containing 1.8 mmol l⁻¹ CaCl₂. The cells were used within 8 h of isolation.

Solutions

The standard Ca²⁺-containing saline solution was composed of (mmol l⁻¹): NaCl 135.0, Hepes (free acid) 10.0, NaH₂PO₄ 0.33, glucose 10.0, KCl 5.4, MgCl₂ 1.0, CaCl₂ 1.8. The nominally Ca²⁺-free solution was identical except CaCl₂ was omitted. The low Ca²⁺ solution contained 0.2 mmol l⁻¹ CaCl₂. The collagenase solution was prepared in nominally Ca²⁺-free saline containing 60 U ml⁻¹ type I collagenase (Worthington Biochemicals) and 0.1 U ml⁻¹ type XIV protease (Sigma Chemical Company). Solutions were adjusted to pH 7.4 with 10 n NaOH.

Four bath and pipette solution sets were prepared for the voltage clamp studies. E_Cat was calculated for monovalent elemental cations.

Condition A (E_Cl =−37 mV; E_Cat =+4.3 mV) used (mmol l⁻¹): choline chloride 42, Hepes 10, glucose 10, CaCl₂ 1.8, MgCl₂ 0.5, CsCl 120 (bath) and caesium aspartate (CsAsp) 90, CsCl 30, TEA 10, MgATP 5, Hepes 10 (pipette). The measured junction potential was 13 mV (pipette negative).

Condition B (E_Cl =+27 mV; E_Cat =−6 mV) used (mmol l⁻¹): choline chloride 42, Hepes 10, glucose 10, CaCl₂ 1.8, MgCl₂ 0.5, CsAsp 120 (bath) and CsCl 120, TEA 10, MgATP 5, Hepes 10 (pipette). The measured junction potential was −13.4 mV (pipette positive).

Condition C (E_Cl = 0 mV; E_Cat = 0 mV) used (mmol l⁻¹): Hepes 10, glucose 5, Ca(OH)₂ 1.8, MgSO₄ 1.0, CsAsp 140, TEACl 5.0 (bath) and CsAsp 140, MgATP 5.0, Hepes 10, TEACl 5.0 (pipette). Solutions for condition C were slightly modified for experiments involving substitution of bath Cs⁺ with NMDG as follows (mmol l⁻¹): Hepes 10, glucose 5, Ca(OH)₂ 3.0, MgSO₄ 1.0, NMDG 135, TEACl 5.0, pH adjusted to 7.4 with 1 n H₂SO₄ (bath) and CsAsp 135, MgATP 5.0, Hepes 10, TEACl 5.0, pH adjusted to 7.2 with 1 n CsOH (pipette). The measured junction potential was 1.5 mV (pipette negative). The E_Cat for monovalent cations deviated slightly from 0 mV for conditions A and B because the measured osmolality (Advanced Instruments, Inc., Needham, MA, USA) of CsAsp stock solutions was only 80 % of the predicted osmolality value. Thus, the activity of Cs⁺ for calculating E_Cat was determined by multiplying the amount of Cs⁺ added as CsAsp by 0.8.

Condition D (E_Cl = 0 mV) used (mmol l⁻¹): Hepes 10, glucose 5.0, CaCl₂ 1.8, MgCl₂ 0.5, NMDG 110, choline chloride 42 (bath) and NMDG 50, MgATP 5.0, Hepes 10, TEACl 107 (pipette). The pH was adjusted to 7.3 with 1 n HCl for condition D bath and pipette solutions and the measured junction potential was 1.1 mV (pipette negative).

For solution sets A-C the pH was 7.2 for the pipette solution and 7.4 for the bath solution. Unless otherwise noted all chemicals were from Sigma. Dofetilide was a generous gift from Pfizer Inc. (Sandwich, UK). Thapsigargin (Calbiochem) was added from a stock solution in DMSO (final DMSO concentration 0.01 %).

Voltage clamp

Isolated quiescent ventricular myocytes were studied at 32°C with patch clamp methodology in whole cell mode configuration (Hamill et al. 1981) using an Axopatch 200B amplifier (Axon Instruments). Patch electrode resistance was 1.0-2.5 MΩ when filled with the intracellular solution. The liquid junction potentials between the bath and pipette solutions were compensated electronically. The series resistance and cell membrane capacitance were compensated before and during the voltage clamp experiments. Currents were low-pass filtered at 2 kHz and sampled at 3.3 kHz using a 12-bit analog to digital converter (Digidata D 1200 B, Axon Instruments). The digitized signals were stored in a personal computer for later analysis with pCLAMP 6.03 (Axon Instruments) and SigmaPlot (SPSS Inc., Chicago, IL, USA). The cell membrane capacitance was measured from the integral of the current transient using a 10 mV depolarization or hyperpolarization step from the holding potential of 0 or −80 mV, respectively. Cells were loaded with Ca²⁺ by a voltage command step to +10 mV from −80 mV for 600 ms applied at 0.33 Hz. The current-voltage (I–V) relationship for the oscillatory current was obtained by voltage clamping the cell membrane to different test potentials between −80 and +60 mV for 1950 ms after the Ca²⁺ loading step before returning to −80 mV (Fig. 1a). Peak OC was measured in experiments with solution conditions A and B. Because the OC was less distinct for experiments using solution set C, steady-state current was measured at the end of the test pulse.

Figure 1. Oscillatory currents in response to the voltage potential protocol used in these experiments.

A, schematic depiction of the conditioning and test voltage command steps used. Cells were held at −80 mV and stimulated at 0.3 Hz with a square wave conditioning step to +10 mV, followed by a test command potential from −80 to +60 mV. B, representative tracing of the oscillatory currents in response to the protocol shown in A. The calculated Cl⁻ reversal potential was −37 mV and the calculated monovalent cation reversal potential was +4.3 mV (condition A, see Methods). The labelled test command cell membrane potentials are corrected for the liquid junction potentials.

Ca²⁺ fluorescence measurement

Intracellular Ca²⁺ was monitored during some experiments to determine the presence of [Ca²⁺]_i transients under conditions favourable to the OC by including the pentapotassium salt of the fluorescent Ca²⁺ indicator fluo-3 (Molecular Probes) in the pipette solution (100–150 μmol l⁻¹) as previously described, with minor modifications (Anderson et al. 1994; Wu et al. 1999). Cells were dialysed for an average of 7 min in whole cell mode to achieve an equilibrium between the pipette and intracellular solutions and steady-state signals during pacing. Voltage signals were low-pass filtered at 50 Hz prior to analysis. Fluo-3 [Ca²⁺]_i transients were stored using pCLAMP 6.03.

Data analysis

Data were analysed with Student's paired or unpaired t tests, as appropriate, and the null hypothesis was rejected for P < 0.05. Results are given as means ±s.e.m.

RESULTS

Oscillatory current

OCs were recorded in every cell studied (n = 88) under control conditions using solution sets A-D following stimulation with the conditioning voltage command (Fig. 1a). The OCs were discrete and had a clear V_rev (Fig. 1b), which was not affected by the order of application of test voltage command steps (data not shown).

Measured reversal potential for the oscillatory currents

The measured OC V_rev using solution set A was −8.9 ± 0.8 mV (Fig. 2a) and was 0.1 ± 1.0 mV using solution set B (Fig. 2b). Measured V_rev for both solution sets A and B was between the calculated V_rev for I_Cl(Ca) and I_CAN. Addition of the I_Cl(Ca) antagonist niflumic acid shifted the measured V_rev to the E_Cat (Fig. 2). Niflumic acid shifted the measured V_rev for solution set A (+5.1 ± 2.8 mV, P < 0.001 compared to control) and solution set B (-10.5 ± 3.4 mV, P = 0.019 compared to control) to E_Cat. The shift in the measured V_rev by niflumic acid suggests that one or more niflumic acid-insensitive currents is also important for the OC.

Figure 2. Current-voltage relationships for measured peak oscillatory currents obtained with the experimental protocol shown in Fig. 1A.

A, the calculated Cl⁻ reversal potential (E_Cl) was −37 mV and the calculated monovalent cation reversal potential (E_Cat) was +4.3 mV (condition A, see Methods). •, control values (n = 11); ○, values obtained after application of the Cl⁻ current antagonist niflumic acid (30 μmol l⁻¹, n = 3). B, data are displayed as in A. Conditions were the same as in A except that solutions were changed so that E_Cl was +27 mV and E_Cat was −6 mV (condition B, see Methods). Data are from 5 control and 5 niflumic acid-treated cells.

Intracellular Ca²⁺ dependence of the oscillatory currents

Intracellular Ca²⁺ transients occurred under conditions associated with OCs (Fig. 3) in all (n = 12) of the cells studied. The apparent magnitude of the [Ca²⁺]_i transients was not reduced near the measured OC V_rev (Fig. 3, middle panel), suggesting that the [Ca²⁺]_i transients were not a result of the OCs, but instead that the OCs were activated by the [Ca²⁺]_i transients. The probable initiating cause of the [Ca²⁺]_i transients was spontaneous release from SR stores due to intracellular Ca²⁺ overload because the [Ca²⁺]_i transients were prevented by treatment with 10 μmol l⁻¹ ryanodine (data not shown).

Figure 3. Representative recordings of the oscillatory currents simultaneously obtained with fluorescence transients in response to the experimental protocol shown in Fig. 1A.

The cells were dialysed with the Ca²⁺ indicator fluo-3 (150 μmol l⁻¹). The calculated Cl⁻ reversal potential was −37 mV and the calculated monovalent cation reversal potential was +4.3 mV (condition A, see Methods). The cell membrane test potential values, corrected for the liquid junction potential, are shown.

Enhanced intracellular Ca²⁺ buffering with BAPTA (20 mmol l⁻¹) completely suppressed OCs (n = 6) (Fig. 4b). The OCs were also prevented following pretreatment with the SR Ca²⁺ release channel antagonist ryanodine (10 μmol l⁻¹, n = 4) or the SR Ca²⁺-ATPase antagonist thapsigargin (1 μmol l⁻¹, n = 4) (Fig. 4C and D), further implicating the SR as the source of activator Ca²⁺ for the OCs. Taken together, these findings suggest that the OCs are activated by repetitive release of Ca²⁺ from the SR.

Figure 4. Dependence of oscillatory currents on activator Ca²⁺ from the sarcoplasmic reticulum.

Representative tracings for oscillatory currents measured in response to a command step to −65 mV. A, oscillatory currents under control conditions (condition A, see Methods) are outward during the conditioning step (+10 mV) and inward in response to the test command. B, oscillatory currents are prevented in a cell studied under identical conditions to A, but with enhanced intracellular Ca²⁺ buffering by addition of BAPTA (20 mmol l⁻¹) to the pipette solution. Oscillatory currents are also prevented under identical conditions to A except for the addition of ryanodine (10 μmol l⁻¹; C) or thapsigargin (1 μmol l⁻¹; D).

Currents in low symmetrical Cl⁻ solutions

Studies were performed in low symmetrical Cl⁻-containing solutions in order to minimize the contribution of I_Cl(Ca) to the observed currents without creating undefined junction potentials, as would occur in Cl⁻-free solutions. These OCs were clearly present, but less distinct (Fig. 5) than in the presence of higher Cl⁻ concentrations (Fig. 1), perhaps because Cl⁻ is a necessary counter-ion during Ca²⁺ fluxes across the SR (Townsend & Rosenberg, 1995). Like the previously observed OC, this current was prevented by enhanced intracellular Ca²⁺ buffering with BAPTA (20 mmol l⁻¹, n = 6) or the SR antagonist ryanodine (10 μmol l⁻¹, n = 4) (data not shown). Unlike the OC recorded in higher Cl⁻ conditions, this current was not sensitive to niflumic acid (10–20 μmol l⁻¹, n = 8) (Fig. 5a). The observed current was not primarily due to permeation of K⁺ channels because it was not eliminated by the I_K1 antagonist Ba²⁺ (1 mmol l⁻¹, n = 6) (Fig. 5b) or by dofetilide (1 μmol l⁻¹, n = 4), an antagonist of the rapid component of the delayed rectifier current (Fig. 5C). A component of outward OC was likely to be due to Cs⁺ permeation through K⁺ channels under these conditions as judged by the reduction in peak tail currents by Ba²⁺ and dofetilide following positive test potentials (Fig. 5b and C).

Figure 5. Representative currents recorded in low symmetrical Cl⁻ solutions.

Experiments were performed with the protocol shown in Fig. 1A with Cl⁻ reduced to 5 mmol l⁻¹ in both bath and pipette solutions (condition C, see Methods). A, a raw data tracing illustrates that application of niflumic acid (20 μmol l⁻¹) did not reduce oscillatory currents in response to the test command step to +60 (grey line) and −60 mV (black line). This tracing is representative of 8 cells studied. Subsequent data tracings, as in A, illustrate that oscillatory currents were not prevented by Ba²⁺ (1 mmol l⁻¹, representative of n = 6 cells; B) or dofetilide (1 μmol l⁻¹, representative of n = 4 cells; C). D, removal of bath solution monovalent cations by NMDG substitution (see Methods) eliminated oscillatory currents during the test commands (+60 and −60 mV shown) without inhibiting L-type Ca²⁺ current during the conditioning step. E, summary I–V relationship for steady-state current at baseline (n = 17) and for residual currents following substitution of monovalent cations in the bath solution with NMDG (n = 4), as shown in D. *P < 0.05, **P < 0.001.

The OC and most residual current was eliminated after equimolar replacement of monovalent cations in the bath solution by NMDG (Fig. 5D and E), suggesting that OCs in low symmetrical Cl⁻ conditions were due to an NMDG-impermeant pathway. The measured V_rev for OC was 4.1 ± 1.8 mV (Fig. 5E), close to the E_Cat (i.e. 0 mV). The steady-state I–V relationship was quasi-linear between −60 and +60 mV (Fig. 5E), consistent with previous reports of I_CAN (Colquhoun et al. 1981; Matsuda, 1983).

Currents in the absence of monovalent elemental cations

The experiments performed thus far support the assumption that I_CAN is due to monovalent elemental cations, based on the agreement between measured V_rev following niflumic acid and calculated E_Cat (Fig. 2). Nevertheless, it is possible that other cationic species used in the solution sets may have contributed to the OC. Solutions for condition D were designed to eliminate monovalent elemental cations and employed asymmetric concentrations of choline and TEA to distinguish I_Cl from current carried by these cations. Clear OCs were present under these conditions (Fig. 6a) with a V_rev equal to E_Cl (Fig. 6b). In the absence of monovalent elemental cations niflumic acid eliminated all OC (Fig. 6b), in contrast to the persistent OC seen after niflumic acid in low symmetrical Cl⁻ solutions (Fig. 5). These findings (1) indicate that choline and TEA do not appreciably contribute to I_CAN, (2) validate the method of calculating E_Cat, and (3) show that niflumic acid is a useful agent for distinguishing I_Cl from I_CAN under these experimental conditions.

Figure 6. Representative oscillatory currents recorded in the absence of monovalent elemental cations.

Experiments were performed with the protocol shown in Fig. 1A with E_Cl = 0 mV and asymmetric TEA and choline in the bath and pipette solutions (condition D, see Methods). A, raw data tracings show oscillatory currents in response to cell membrane voltage commands indicated (top). B, summary current-voltage relationship for control cells (n = 10) and cells following exposure to niflumic acid (20 μmol l⁻¹, n = 3). *P < 0.05.

DISCUSSION

Oscillatory Ca²⁺-activated currents

The identity of OCs recorded in the absence of Na⁺-containing solutions has been controversial, but evidence exists to support a role for a Ca²⁺-activated non-selective cation current (Colquhoun et al. 1981; Cannell & Lederer, 1986; Ehara et al. 1988; Giles & Shimoni, 1989; Han & Ferrier, 1992; Magishi et al. 1996). Other currents have been proposed to underlie OC including L-type Ca²⁺ current (Sipido et al. 1995) and I_Cl(Ca) (Zygmunt & Gibbons, 1991). Recently it was reported that the inward rectifier current (I_K1) is reduced by increased [Ca²⁺]_i and so it too could contribute to Ca²⁺-activated OC (Zaza et al. 1998). The OC measured under the experimental conditions of this study were distinct and had a clear reversal potential suggesting that they were due to (a) specific conductance(s). The OCs were activated by release of Ca²⁺ from the SR because they were prevented by treatment with ryanodine or thapsigargin under experimental conditions favourable to SR Ca²⁺ loading. Thus, several of the above-mentioned candidate Ca²⁺-activated currents could potentially underlie the OC in these experiments.

I_CAN and I_Cl(Ca) in rabbit ventricular myocytes

These experiments found strong evidence indicating that both I_Cl(Ca) and I_CAN contribute importantly to Ca²⁺-activated currents in rabbit ventricular myocytes in the absence of Na⁺ and under conditions where K⁺ conductances are anticipated to be markedly reduced or eliminated. I_Cl(Ca) is an outwardly rectifying current (Zygmunt & Gibbons, 1991) that is inhibited by niflumic acid (Fig. 2) (Collier et al. 1996). Under conditions of markedly reduced Cl⁻ (Fig. 5E) or in the presence of niflumic acid (Fig. 2) the I–V relationship for the Ca²⁺-activated currents became more linear, indicating elimination of most of the outwardly rectifying I_Cl(Ca). Elimination of I_Cl(Ca) under conditions of reduced Cl⁻ was also suggested by the lack of further response to niflumic acid (Fig. 5a). Using higher Cl⁻ concentrations (conditions A and B), the measured V_rev for the OC shifted with the E_Cl, and was always between the E_Cl and E_Cat (Fig. 2), indicating that both of these currents were likely to contribute to the OC. Niflumic acid shifted the measured V_rev for the OC to the E_Cat (Fig. 2). This shift in V_rev is consistent with I_CAN importantly contributing to the observed OC. Finally, the OC V_rev was identical to E_Cl in the absence of monovalent elemental cations consistent with the hypothesis that I_CAN and I_Cl(Ca) account for the major portion of OC in the absence of Na⁺. Under these conditions all OCs were eliminated by niflumic acid (Fig. 6).

The OCs that persisted in low Cl⁻ conditions were reduced but remained Ca²⁺ sensitive, had a near-linear I–V relationship from −60 to +60 mV, and a measured V_rev near E_Cat (Fig. 5E). Removal of monovalent elemental cations from the bath eliminated the OCs and most residual current (Fig. 5D and E), indicating that an NMDG-impermeant conductance, such as that presumed to underlie I_CAN, was responsible for the OCs after elimination of I_Cl(Ca). I_CAN was likewise not affected by choline or TEA because the OC V_rev was equal to E_Cl when these constituents were distributed asymmetrically (Fig. 6). The finding that NMDG eliminated all OCs, not only inward OCs, may indicate that NMDG is an I_CAN antagonist.

Study limitations

At present I_CAN is only demonstrable under extraordinary experimental conditions, in part due to the lack of a specific antagonist. Thus it is uncertain how I_CAN may participate in the genesis of cardiac arrhythmias or in the physiology of the action potential. While our results suggest that niflumic acid is a useful tool to distinguish I_CAN from I_Cl(Ca) based on (1) the lack of suppression of OCs in low symmetrical Cl⁻ solutions and (2) the complete suppression of OCs in the absence of elemental monovalent cations, niflumic acid is known to directly inhibit L-type Ca²⁺ current (Terracciano & MacLeod, 1997) and so could reduce the intracellular Ca²⁺ concentration. Thus, the apparent relative magnitudes of I_CAN and I_Cl(Ca) (Fig. 2) may be misleading if (1) niflumic acid significantly reduced the [Ca²⁺]_i and if these currents are activated over different [Ca²⁺]_i ranges, or (2) if direct block of I_Cl(Ca) by niflumic acid is enhanced at positive cell membrane potentials. Similarly, the apparent outward rectification of the OC in the absence of elemental monovalent cations (Fig. 6) may represent an intrinsic property of I_Cl(Ca), but may also be due to changes in [Ca²⁺]_i during the stimulation protocol. Thus, while these experimental findings provide strong evidence for the existence of a macroscopic I_CAN in rabbit ventricular myocytes, they do not allow for definitive quantitative comparisons between I_CAN and I_Cl(Ca).

Ca²⁺-activated currents other than I_CAN and I_Cl(Ca) in rabbit ventricular myocytes

Other candidate currents were not likely to account for the inward OCs observed in this study. L-type Ca²⁺ current was not responsible for the OCs in these experiments because (1) the OC occurred negative to the anticipated cell membrane potential for L-type Ca²⁺ current activation (Fig. 1); (2) the V_rev for OC was sensitive to Cl⁻ (Figs 2 and 6); (3) the OCs were eliminated after substitution of monovalent cations when L-type Ca²⁺ current persisted in the conditioning step (Fig. 5D and E). Likewise, under low Cl⁻ conditions I_K1 was not significantly involved in the inward currents because it was not diminished by Ba²⁺ (Fig. 5b) (Zaza et al. 1998). Finally, the lack of effect of dofetilide on the inward current in low Cl⁻ suggests that this current was not due to Cs⁺ permeation of other K⁺ channels. Some current following test potentials was likely through K⁺ channels normally responsible for I_K1 and I_Kr because the tail currents after Ba²⁺ (Fig. 5a) and dofetilide (Fig. 5C) were reduced. In contrast, no reduction was seen in the holding current or in tail currents following command steps to −60 mV. Several putative Ca²⁺-activated non-selective ion channels have recently been cloned from a variety of tissues, including heart (Suzuki et al. 1998). Further characterization of these ion channels is likely to help clarify some of these apparent discrepancies. These findings support the hypothesis that I_CAN is present as a macroscopic current in rabbit ventricular myocytes.