Ca²⁺-activated Cl⁻ current in rabbit sinoatrial node cells

Abstract

The Ca²⁺-activated Cl⁻ current (I_Cl(Ca)) has been identified in atrial, Purkinje and ventricular cells, where it plays a substantial role in phase-1 repolarization and delayed after-depolarizations. In sinoatrial (SA) node cells, however, the presence and functional role of I_Cl(Ca) is unknown. In the present study we address this issue using perforated patch-clamp methodology and computer simulations. Single SA node cells were enzymatically isolated from rabbit hearts. I_Cl(Ca) was measured, using the perforated patch-clamp technique, as the current sensitive to the anion blocker 4,4′-diisothiocyanostilbene-2,2′-disulphonic acid (DIDS). Voltage clamp experiments demonstrate the presence of I_Cl(Ca) in one third of the spontaneously active SA node cells. The current was transient outward with a bell-shaped current-voltage relationship. Adrenoceptor stimulation with 1 μm noradrenaline doubled the I_Cl(Ca) density. Action potential clamp measurements demonstrate that I_Cl(Ca) is activate late during the action potential upstroke. Current clamp experiments show, both in the absence and presence of 1 μm noradrenaline, that blockade of I_Cl(Ca) increases the action potential overshoot and duration, measured at 20 % repolarization. However, intrinsic interbeat interval, upstroke velocity, diastolic depolarization rate and the action potential duration measured at 50 and 90 % repolarization were not affected. Our experimental data are supported by computer simulations, which additionally demonstrate that I_Cl(Ca) has a limited role in pacemaker synchronization or action potential conduction. In conclusion, I_Cl(Ca) is present in one third of SA node cells and is activated during the pacemaker cycle. However, I_Cl(Ca) does not modulate intrinsic interbeat interval, pacemaker synchronization or action potential conduction.

Atrial, Purkinje and ventricular myocytes from different species have been demonstrated to express a Ca²⁺-activated Cl⁻ current (I_Cl(Ca)) (for reviews, see Sorota, 1999; Hume et al. 2000). I_Cl(Ca) is the 4-aminopyridine insensitive part of the transient outward current and is also termed I_to2. I_Cl(Ca) is primarily activated by Ca²⁺ ions released from the sarcoplasmic reticulum (Sipido et al. 1993; Kawano et al. 1995). The current does not follow the time course of the intracellular Ca²⁺ transients, but reaches a peak within 10–20 ms and then inactivates in the following 100 ms (Zygmunt & Gibbons, 1991, 1992; Sipido et al. 1993). It is postulated that this transient behaviour may be due to an intrinsic Ca²⁺-dependent inactivation of I_Cl(Ca) or may be due to the time course of local, subsarcolemmal Ca²⁺ gradients (Sipido et al. 1993; Zygmunt, 1994; Kawano et al. 1995).

The reversal potential for Cl⁻ ions (E_Cl) in cardiomyocytes is approximately −50 mV (Sorota, 1999). Therefore, activation of I_Cl(Ca) has the potential, in theory, to generate an inwardly directed, depolarizing current at resting membrane potentials, whereas it generates an outwardly directed, repolarizing current during phase-1 and phase-2 of the action potential. Indeed, it has been demonstrated experimentally that I_Cl(Ca) strongly influences phase-1 repolarization of atrial (Wang et al. 1995) and ventricular action potentials (Hiraoka & Kawano, 1989; Kawano & Hiraoka, 1991), especially under conditions of adrenoceptor stimulation (Verkerk et al. 2001). Moreover, it has been demonstrated that I_Cl(Ca) contributes to the potentially arrhythmogenic delayed after-depolarizations in Purkinje and ventricular cells (Zygmunt et al. 1998; Verkerk et al. 2000). These findings indicate that I_Cl(Ca) plays a substantial role in atrial, Purkinje and ventricular electrophysiology.

The presence of I_Cl(Ca) in sinoatrial (SA) node cells is unknown. In SA node tissue, the sarcoplasmic reticulum is relatively sparse compared with atrial cells (Masson-Pévet et al. 1978). Nevertheless, cultured and freshly isolated mammalian SA node cells show clear intracellular Ca²⁺ transients (Li et al. 1997; Huser et al. 2000; Rigg et al. 2000; Bogdanov et al. 2001) as do amphibian pacemaker cells (Ju & Allen, 1999, 2000), indicating that the substrate for activating I_Cl(Ca) in SA node cells is present. The present study was designed to assess the presence and functional role of I_Cl(Ca) in single SA node cells of rabbit. We report that I_Cl(Ca) is present in one third of the rabbit SA node cells. Action potential clamp measurements show that I_Cl(Ca) is activated during the normal pacemaker cycle late during the action potential upstroke. Blockade of I_Cl(Ca) results in an increase of the action potential overshoot. Noradrenaline (1 μm) doubled the size of I_Cl(Ca), but even under such conditions, I_Cl(Ca) did not change the firing rate. We incorporated I_Cl(Ca) into our previously published model of a rabbit SA node cell (Wilders et al. 1991), and used this model to assess the functional role of I_Cl(Ca) in pacemaker activity. The computer simulations show that I_Cl(Ca) has minimal effects on synchronization of SA nodal cells and action potential conduction between SA nodal and atrial cells.

METHODS

Cell preparation

All experiments were carried out in accordance with guidelines of the local Institutional Animal Care and Use Committee. Single SA node cells were isolated from rabbit hearts by an enzymatic dissociation procedure as described by Verheijck et al. (1995). The hearts were obtained from New Zealand White rabbits (body weight: 3.0–3.5 kg), which were anaesthetized with a 1 ml kg⁻¹ intramuscular injection of Hypnorm (10 mg ml⁻¹ fluanisone and 0.315 mg ml⁻¹ fentanyl citrate; Jansen Pharmaceutics, Tilburg, The Netherlands). Small aliquots of cell suspension were put in a recording chamber on the stage of an inverted microscope. Cells were allowed to adhere for 5 min after which perfusion with Tyrode solution (36 ± 0.5 °C) was started. The Tyrode solution contained (mm): NaCl 140, KCl 5.4, CaCl₂ 1.8, MgCl₂ 1.0, glucose 5.5 and Hepes 5.0 (pH was adjusted to 7.4 with NaOH). Cells were dissociated from the entire SA node region and represent a mixed population of cells showing heterogeneity in shape (Denyer & Brown, 1990; Verheijck et al. 1998a). For our experiments, we selected spindle and elongated spindle-like cells displaying regular contractions.

Recording procedures

Membrane potentials and membrane currents were recorded using the amphotericin perforated-patch technique (Horn & Marty, 1988) to prevent run down of membrane currents by dilution of intracellular components. Patch pipettes were pulled from borosilicate glass and heat polished. The pipette solution contained (mm): potassium gluconate 125, KCl 20, amphotericin 2.2 and Hepes 10 (pH was adjusted to 7.2 with KOH). The potential between pipette and bath solution was adjusted to zero before a high-resistance seal was formed. All potentials were off-line corrected for the estimated 13 mV liquid junction potential. Membrane currents and potentials were filtered on-line with a cutoff frequency of 1 kHz, and digitized by a 12-bit analog-to-digital converter (NB-MIO-16, National Instruments Co. Austin, TX, USA) with a sampling frequency of 2 kHz. Data were stored and analysed by custom software.

Current-clamp experiments

To characterize action potentials, several action potential parameters were determined: action potential duration at 20, 50 and 100 % repolarization (APD₂₀, APD₅₀, and APD₁₀₀, respectively), maximal diastolic potential (MDP), action potential overshoot, cycle length and maximum upstroke velocity (dV/dt_max). Diastolic depolarization rate (DDR) was measured over the 50 ms time interval starting at the MDP + 1 mV. MDP + 1 mV was used rather than MDP because the time at which the MDP + 1 mV was reached could be determined more reliably than the time at which the MDP was reached. Cell capacitance (C_m) was estimated from the change in slope of the membrane potential (Δ(dV_m/dt)) upon 100 ms hyperpolarizing current pulses of 25 to 50 pA (ΔI_m), which were applied shortly after the action potential had reached its MDP. Cell capacitance was calculated as C_m = ΔI_m / (Δ(dV_m/dt), and ranged between 18 and 110 pF with a mean value of 51 ± 5.8 pF (n = 40; mean ± s.e.m.).

Voltage clamp experiments

The L-type Ca²⁺ current (I_Ca,L) and the delayed rectifier current (I_K) were elicited by series of depolarizing voltage clamp steps of 500 ms duration to membrane potentials ranging from −40 to +70 mV, with 10 mV increments. Voltage steps were applied once every 2 s from a holding potential of −40 mV. We defined I_Ca,L as inward peak current at the beginning of a depolarizing voltage clamp step (I_peak). I_K was defined as the quasi steady-state current at the end of the depolarizing voltage clamp steps (I_qss). I_peak and I_qss are largely due to I_Ca,L and I_K, respectively, but other currents might also contribute to these currents. The hyperpolarizing-activated current (I_f) was elicited by series of hyperpolarizing voltage clamp steps of 500 ms duration to membrane potentials ranging from −40 to −100 mV, with 10 mV decrements. Voltage clamp steps were applied once every 2 s from a holding potential of −40 mV. We defined I_f as the difference between the current at the end and start of the hyperpolarizing voltage clamp step. In addition, tail currents of the delayed rectifier current and I_f were analysed. As tail currents predominantly express deactivation of a current, they are expressed relative to the current level at the holding potential, i.e. −40 mV. The Ca²⁺-activated Cl⁻ outward current (I_Cl(Ca)) was elicited by a series of depolarizing voltage clamp steps of 500 ms duration to membrane potentials ranging from −40 to +70 mV, with 10 mV increments. Voltage steps were applied once every 2 s from a holding potential of −40 mV. I_Cl(Ca) was defined as the transient outward current sensitive to 4,4′diisothiocyanatostilbene-2,2′-disulphonic acid (DIDS; Sigma Chemical Co). DIDS was freshly prepared as a 0.5 m stock solution in DMSO (Merck) and kept in the dark. It was diluted in Tyrode solution for use at a final concentration of 0.2 mm. The action potential clamp technique, as introduced in SA node cells by Doerr et al. (1989), was used to determine the time course of the I_Cl(Ca) during the pacemaking cycle. Therefore, transmembrane potentials were recorded during stable spontaneous pacemaker activity (current clamp conditions), digitized at 10 kHz, and stored. One representative action potential waveform was selected and used as the command signal in voltage clamp conditions to drive the potential in the same cell. Continuous pacemaker activity was obtained by repeatedly applying the same waveform to the cell. At least 20 consecutive waveforms were applied in order to reach stable electrical activity. The membrane currents were recorded under control conditions, in the presence of 0.2 mm DIDS and after drug washout. All currents were normalized for cell size by dividing current amplitude by C_m. Adrenoceptor stimulation was induced by application of 1 μm noradrenaline (Centrafarm. Etten-Leur, The Netherlands).

Computer simulations

We used computer simulations to investigate the role of I_Cl(Ca) in pacemaker activity. In these simulations, the membrane potential (V_m) of a rabbit SA nodal cell was calculated from our detailed mathematical SA node model cell (Wilders et al. 1991), into which we incorporated equations for I_Cl(Ca). The formulation of I_Cl(Ca) was based on a model study of I_Cl(Ca) in atrial cells (Gomis-Tena & Saiz, 1999a, 1999b) and is simulated as a [Ca²⁺]_i-dependent conductance with (V_m-dependent) outward rectification through:

(1)

where R, T and F are universal gas constant, absolute temperature and Faraday constant, respectively, and where:

(2)

and

(3)

represent the [Ca²⁺]_i dependence and V_m dependence, respectively. Since our model does not discriminate between free and bound intracellular calcium, the [Ca²⁺]_i in eqn (2) is total intracellular calcium concentration. According to Gomis-Tena & Saiz (1999a, b), the intracellular and extracellular chloride concentrations, [Cl⁻]_i and [Cl⁻]_o, are set to the Lindblad et al. (1996) values of 30 and 132 mm, respectively.

Following the approach of Gomis-Tena & Saiz (1999a, b), I_Cl(Ca) is thus the product of: (a) the driving force for chloride ions, described by the Goldmann-Hodgkin-Katz equation, (b) the [Ca²⁺]_i dependence, tentatively described by a Hill-type equation with parameters K_m and n_h, and (c) the V_m dependence as is demonstrated in a variety of excitable and non-excitable cells using physiological [Cl⁻]_i (for reviews, see Sorota, 1999; Hume et al. 2000). Quantitative experimental data on the [Ca²⁺]_i dependence of I_Cl(Ca), however, are absent. In our computer simulations the binding constant K_m and exponent n_h (in eqn (2)) were set to 5 mm and 3, respectively. Using these two parameter settings, both the experimentally observed current-voltage relationship and the moment of activation during an action potential could be well fitted (see Figs 2, 6 and 7). In addition, the permeability P_Cl(Ca) (in eqn (1)) was set on 0.33 μL s⁻¹ to match the density of I_Cl(Ca) with our own experimental data (see Fig. 2 and Fig. 7). Thus some parameter values were not based on experimental data but were chosen because they allow a proper fit of our own experimental data.

Figure 2. Presence of the Ca²⁺-activated Cl⁻ current (I_Cl(Ca)) in SA node cells.

A, superimposed current traces of an SA node cell (C_m, 110 pF) elicited by voltage steps from −40 to +10 mV in absence (•) and presence (○) of 0.2 mm DIDS. Arrow indicates an outwardly directed transient current, which is superimposed on the inward peak current. This transient current is blocked by DIDS. B, DIDS-sensitive I_Cl(Ca) current obtained by digital subtraction of the two current traces of panel A. C, I_Cl(Ca) measured as the DIDS-sensitive current at potentials between −40 and +70 mV in the same cell as shown in panels A and B. The current activates around −20 mV and the peak amplitude increases upon further depolarization to +40 mV. At more depolarized potentials, the peak amplitude of I_Cl(Ca) decreases. D, average I-V relationship of I_Cl(Ca) in 7 cells.

Figure 6. Action potential clamp recordings of I_Cl(Ca) in an SA node cell.

A, complete cycle of pacemaker activity of an SA node cell (C_m, 81 pF) recorded under current clamp conditions. B, currents recorded during reapplied pacemaker activity under voltage clamp conditions (‘action potential clamp’) in absence of DIDS (upper panel), in presence of 0.2 mm DIDS (middle panel), and upon wash out of the drug (lower panel). Inset shows current obtained by digitally subtracting the current trace recorded under wash out conditions from the one recorded under control conditions. C, DIDS-sensitive current I_Cl(Ca) obtained by digital subtraction of compensation currents in presence and after wash out of 0.2 mm DIDS. D, reconstructed current (I_m) flow during the cycle of pacemaker activity. Current was calculated by I_m = -C_m × dV/dt, where C_m is the membrane capacity of the cell. Note the different current scale.

Figure 7. Computer simulations of I_Cl(Ca) under voltage clamp and current clamp conditions.

A, I_Cl(Ca) in response to voltage clamp steps as indicated (left) and associated I-V relationship (right). B, free-running membrane potential (V_m) of the Wilders et al. (1991) rabbit SA nodal cell model with and without I_Cl(Ca) (continuous and dashed lines, respectively). C, time course of I_Cl(Ca) and net membrane current (I_m) during the spontaneous activity of panel B.

For numerical integration of differential equations we applied a simple and efficient Euler scheme with a fixed time step of 5 μs. All software was compiled as a 32-bit Windows application using Compaq Visual Fortran 6.5 and run on a 667-MHz Alpha processor workstation (Microway Screamer).

Statistics

Results are expressed as means ± s.e.m. Data on action potential characteristics were obtained from 10 consecutive action potentials and averaged. Using statistical analysis software of Microsoft® Excel 98, statistical significance was determined by one-way or two-way analysis of variance (ANOVA) combined with Student's t test for paired or unpaired observations. A probability value of P < 0.05 was considered significant.

RESULTS

Sensitivity of SA node cation membrane currents to DIDS

In previous studies in atrial, Purkinje and ventricular cells, I_Cl(Ca) was defined as the current sensitive to DIDS (Zygmunt & Gibbons, 1991, 1992; Sipido et al. 1993). In a first series of experiments, we tested the sensitivity to 0.2 mm DIDS of the major cationic membrane currents, i.e., I_f, I_Ca,L (defined as the inward peak current at the beginning of a voltage clamp step, I_peak), and I_K (defined as quasi steady-state current at the end of a voltage clamp step, I_qss) in SA node cells. For Fig. 1, to avoid interference with the DIDS-sensitive I_Cl(Ca), we only used cells lacking I_Cl(Ca) (see below). Figure 1A shows current traces in the absence and presence of DIDS recorded upon voltage clamp steps to −100, 0 and +50 mV, from a holding potential of −40 mV. Figure 1, B-D, shows the average I-V relationship of I_f, I_peak, and I_qss of 10 SA node cells in the absence (•) and presence (○) of 0.2 mm DIDS. Current traces as well as the I-V relationships before and during the administration of 0.2 mm DIDS almost completely overlap. Therefore we conclude that DIDS does not significantly affect I_f, I_peak (predominantly reflecting I_Ca,L), or I_qss (predominantly reflecting I_K).

Figure 1. Effects of 0.2 mm DIDS on cationic membrane currents in sinoatrial (SA) node cells.

A, superimposed current traces of an SA node cell (Cm, 57 pF) elicited by 500 ms voltage steps from −40 mV to −100, 0, and +50 mV in absence and presence of DIDS. B-D, average current-voltage (I-V) relationship of the hyperpolarization-activated current (If, B), the inward peak current at the beginning of a voltage clamp step (Ipeak, C), and the quasi steady-state current at the end of a voltage clamp step (Iqss, D) in 10 cells in the absence (•) and presence (○) of DIDS. Both quasi steady-state and tail currents of Iqss and If are shown.

Presence of I_Cl(Ca) in SA node cells

In a second series of experiments, we tested the presence of I_Cl(Ca) in 21 single SA node cells. We applied a series of depolarizing voltage clamp steps from a holding potential of −40 mV to membrane potentials ranging from −40 to +70 mV in the absence and presence of 0.2 mm DIDS and recorded the resulting family of currents. Figure 2A shows two current traces elicited upon stepping to +10 mV recorded before (•) and during (○) the administration of 0.2 mm DIDS. Before administration, I_peak was accompanied by a transient outwardly directed current component (see arrow), which was blocked during the administration of DIDS. By digitally subtracting the two current traces, the DIDS-sensitive I_Cl(Ca) current was obtained (Fig. 2B). In 7 out of 21 cells, we observed such a DIDS-sensitive current. Figure 2C shows the current traces of this DIDS-sensitive current recorded between −40 and +70 mV with 10 mV increments on an expanded time scale. The current is transient and outwardly directed, with characteristics similar to those of I_Cl(Ca) in atrial, Purkinje and ventricular cells (for reviews, see Sorota, 1999; Hume et al. 2000). Figure 2D shows the average I-V relationship of I_Cl(Ca) in these 7 SA node cells. The I-V relationship was bell shaped with an activation threshold around −20 mV and a maximum near +40 mV.

Our results demonstrate that I_Cl(Ca) is present in one third of the cells tested. It might be argued that cells which display I_Cl(Ca) are not nodal but atrial cells since we have previously described that atrial cells are present in the SA node region (Verheijck et al. 1998a). Figure 3A shows an example of the electrical activity of a cell with I_Cl(Ca) (○) and one without I_Cl(Ca) (•). The two cells clearly show similar electrical activity. Table 1 contrasts action potential parameters of the 7 SA node cells with I_Cl(Ca) and the 14 SA node cells without I_Cl(Ca). No difference in any action potential parameter could be detected between both groups of cells. However, a flattening of the top of the action potential was consistently present in cells including I_Cl(Ca) (see Fig. 3A). Figs 3B-D shows that the average I-V relationships of I_f, I_peak, and I_qss of SA node cells with I_Cl(Ca) (○) and without I_Cl(Ca) (•) also do not differ from each other.

Figure 3. Electrical characteristics of SA node cells with and without I_Cl(Ca).

A, typical pacemaker activity of an SA node cell with I_Cl(Ca) (○; C_m, 29 pF) and an SA node cell without I_Cl(Ca) (•; C_m, 62 pF). B-D, average I-V relationship of I_f (B), I_peak (C), and I_qss (D) of 7 cells with (•) and 14 cells without (○) I_Cl(Ca). Both quasi steady-state and tail currents of I_qss and I_f are shown.

Table 1.

Action potential characteristics of SA node cells with and without I_Cl(Ca)

	No ICl(Ca) (n = 14)	With ICl(Ca) (n = 7)
APD20 (ms)	63 ± 11	70 ± 7
APD50 (ms)	92 ± 12	102 ± 10
APD100 (ms)	167 ± 17	171 ± 18
MDP (mV)	−61.0 ± 1.6	−60.4 ± 3.2
DDR (mVs−1)	92 ± 12	91 ± 18
Overshoot (mV)	23.3 ± 1.6	20.9 ± 2.9
dV/dtmax (V s−1)	5.9 ± 1.3	5.1 ± 1.5
Cycle length (ms)	316 ± 31	314 ± 38
Capacitance (pF)	49 ± 6	47 ± 10

Data are means ± s.e.m. APD₂₀, APD₅₀ and APD₁₀₀,action potential duration at 20, 50 and 100% repolarization, respectively. MDP, maximal diastolic potential; DDR,diastolic depolarization rate; Overshoot,action potential overshoot and dV/dt_max, maximum upstroke velocity.

Effect of I_Cl(Ca) blockade on electrical activity of SA nodal myocytes

Our voltage clamp experiments show that I_Cl(Ca) is present in one third of the regularly beating single SA node cells of rabbit. Recently, it was demonstrated that the increase in subsarcolemmal intracellular Ca²⁺ concentration in spontaneously active cells occurred concomitantly with the late phase of diastolic depolarization (Huser et al. 2000; Rigg et al. 2000; Bogdanov et al. 2001). This implies that I_Cl(Ca) may already be activated during the late phase of the diastolic depolarization and thus can play a role in regulating pacemaker activity. The selective action of DIDS on I_Cl(Ca) enabled us to evaluate the contribution of I_Cl(Ca) to spontaneous electrical activity.

The effects of DIDS on SA node action potentials were investigated on the 7 cells including I_Cl(Ca). Figure 4 shows a typical example of an SA node action potential recorded in the absence (•) and presence (○) of 0.2 mm DIDS. DIDS predominantly increased the action potential overshoot and prolonged the APD₂₀. Table 2 summarizes the effect of 0.2 mm DIDS on action potential parameters. DIDS significantly increased the action potential overshoot and prolonged APD₂₀, without affecting dV/dt_max, DDR, MDP, APD₅₀, APD₁₀₀ or cycle length. Thus, I_Cl(Ca) modulates the action potential overshoot of SA node cells, but it does not play a significant role in beating rate regulation under normal conditions.

Figure 4. Effects of blockade of I_Cl(Ca) in SA node cells.

Typical pacemaker activity of a I_Cl(Ca) expressing SA node cell (C_m, 29 pF) in absence (•) and presence (○) of 0.2 mm DIDS.

Table 2.

Effect of 0.2 mm DIDS on action potential characteristics of SA node cells expressing I_Cl(Ca) (n = 7)

	Control	DIDS
APD20 (ms)	70 ± 7	73 ± 7*
APD50 (ms)	102 ± 10	104 ± 9
APD100 (ms)	171 ± 18	172 ± 19
MDP (mV)	−60.4 ± 3.2	−60.3 ± 3.5
DDR (mV s−1)	91 ± 18	105 ± 25
Overshoot (mV)	20.9 ± 2.9	22.8 ± 3.3*
dV/dtmax (V s−1)	5.1 ± 1.5	5.1 ± 1.4
Cycle length (ms)	314 ± 38	316 ± 43

Data are means ± s.e.m.. APD₂₀, APD₅₀, and APD₁₀₀, action potention duration at 20, 50, and 100% repolarization, respectively. MDP, maximal diastolic potential; DDR, diastolic depolarization rate; Overshoot, action potential overshoot and dV/dt_max, maximum upstroke velocity.

^*P < 0.05.

Role of I_Cl(Ca) on electrical activity during adrenergic stimulation

It is well known that the calcium content of the SR is increased during adrenoceptor stimulation, mainly as a consequence of an increase in I_Ca,L (Osterrieder et al. 1982) and an increase in the loading of the SR (Wolska et al. 1996). It has been demonstrated that sympathetic stimulation with isoprenaline increased the amplitude of Ca²⁺ transient by about 75 % in guinea-pig SA node cells (Rigg et al. 2000) and by about 80 % in amphibian pacemaker cells (Bramich & Cousins, 1999; Ju & Allen, 1999). It is conceivable that this enhanced filling and subsequent release of Ca²⁺ from the SR during adrenergic stimulation may enhance I_Cl(Ca) and thereby increase its role in SA node electrophysiology.

In an additional series of experiments, we determined the effect of adrenoceptor stimulation on the density of I_Cl(Ca). Therefore, we compared the I_Cl(Ca) already measured in the 7 cells under control conditions with I_Cl(Ca) in another set of cells in presence of 1 μm noradrenaline. In 6 out of 16 cells, I_Cl(Ca) was observed in the presence of noradrenaline. The percentage of cells showing I_Cl(Ca) thus was comparable in the absence (33.3 %, n = 21) and presence (37.5 %, n = 16) of noradrenaline. This suggests that adrenoceptor stimulation could not ‘wake up’ I_Cl(Ca) in cells that had no apparent current at baseline. In 5 out of these 6 cells, the voltage clamp experiments could be completed before the cells showed any signs of run down. Figure 5A shows the mean I-V relationship of I_Cl(Ca) under control conditions (•) and in presence of noradrenaline (○). Both I-V relationships are bell shaped with an activation threshold around −20 mV and a peak close to +40 mV. However, density of I_Cl(Ca) is about doubled in the presence of noradrenaline.

Figure 5. Adrenoceptor stimulation and I_Cl(Ca) in SA node cells.

A, average I-V relationship of I_Cl(Ca) in absence (n = 7; •) and presence of 1 μm noradrenaline (n = 5; ○). B, typical pacemaker activity of a I_Cl(Ca) expressing SA node cell (C_m, 36 pF) under control conditions (▴), in the presence of 1 μm noradrenaline (•), and in the presence of both 1 μm noradrenaline and 0.2 mm DIDS (○).

In the same group of cells, we measured whether I_Cl(Ca) may influence pacemaker activity upon adrenoceptor stimulation. The effects of DIDS on SA node action potentials parameters in the presence of noradrenaline were investigated in the 6 cells expressing I_Cl(Ca). Figure 5B shows a typical example of SA node action potentials recorded under control conditions (▴), in the presence of noradrenaline (•), and in the presence of noradrenaline and 0.2 mm DIDS (○). Table 3 summarizes the average action potential parameters under these three conditions. In the SA node cells expressing I_Cl(Ca), noradrenaline induces the well-known positive chronotropic effect (for review, see Irisawa et al. 1993; Boyett et al. 2000). In addition, the cycle length and action potential overshoot were significantly decreased, while dV/dt_max and DDR were significantly increased. In the presence of noradrenaline, blockade of I_Cl(Ca) with DIDS significantly increased action potential overshoot and APD₂₀, without changing the other action potential characteristics. Our experiments thus show that adrenoceptor stimulation increased I_Cl(Ca) density and thereby enhanced its modulating role on the action potential overshoot. However, even under conditions of potentiated I_Cl(Ca) density, this current does not play a significant role in controlling beating rate.

Table 3.

Effects of 0.2 mm DIDS on action potential characteristics of SA node cells expressing I_Cl(Ca) (n = 6) in the presence of adrenergic receptor stimulation by 1 μM noradrenaline

	Control	Norad.	Norad. + DIDS
APD20 (ms)	72 ± 7	61 ± 4*	70 ± 5†
APD50 (ms)	105 ± 11	94 ± 7	100 ± 6
APD100 (ms)	160 ± 11	153 ± 5	155 ± 5
MDP (mV)	−62.0 ± 1.2	−61.4 ± 2.0	−61.6 ± 1.5
DDR (mV s−1)	91 ± 17	117 ± 17*	118 ± 17
Overshoot (mV)	19.3 ± 1.2	16.4 ± 1.3*	21.9 ± 1.6†
dV/dtmax (V s−1)	5.4 ± 0.6	6.8 ± 0.4*	6.6 ± 0.3
Cycle length (ms)	339 ± 16	267 ± 24*	270 ± 19

Data are means ± s.e.m.. APD₂₀, APD₅₀, and APD₁₀₀, action potential duration at 20, 50, and 100 % repolarization, respectively. MDP, maximal diastolic potential; DDR, diastolic depolarization rate; Overshoot, action potential overshoot; and dV/dt_max, maximum upstroke velocity, Norad., noradrenaline.

^*P < 0.05 noradrenaline vs. control

^†P < 0.05 noradrenaline + DIDS vs. noradrenaline.

Action potential clamp and DIDS sensitive current during electrical activity

The above experiments provide evidence that I_Cl(Ca) is active during spontaneous activity, but plays no role in regulating the intrinsic cycle length. Also, the above experiments suggest that I_Cl(Ca) is activated late during the fast action potential depolarization. Pharmacological blockade of individual currents provides useful information, but the changes in membrane potential induced by blockade of a single current may, in turn, affect other currents (Zaza et al. 1997), resulting in an overlap of primary and secondary effects of blockade, thus confusing the interpretation of results. Therefore, we carried out a final series of experiments in which we used the action potential clamp technique as introduced in SA node cells by Doerr et al. (1989) to avoid these problems. In these experiments, the time course of the DIDS-sensitive I_Cl(Ca) during the pacemaking cycle is measured while clamping the membrane potential with action potential waveforms recorded from the same cell under control conditions.

Figure 6 shows a typical example of an action potential clamp experiment. Figure 6A shows the membrane potential recorded during stable spontaneous pacemaking activity (current clamp conditions) in a single cell under control conditions. This waveform was used as the command signal to drive the membrane potential of the same cell in the voltage clamp mode. Figure 6B shows the membrane current recorded under action potential clamp conditions during control conditions (upper panel), in presence of 0.2 mm DIDS (middle panel), and after wash out of the drug (lower panel). The total membrane current recorded under control conditions was close to zero (Fig. 6B, top), indicating that there is no difference in the intrinsic and the applied voltage waveform. The current recorded after wash out of DIDS was also close to zero, except for a short period during the action potential upstroke and shortly thereafter (Fig. 6B, bottom). Apparently, a difference in intrinsic and the applied voltage occurs after wash out, which is reflected by a small compensation current. The inset shows the current obtained by digitally subtracting the current trace recorded after wash out of DIDS from the one recorded during control conditions. The resulting current shows similarities in shape with the nifedipine-sensitive current, I_Ca,L, as recorded by Zaza and coworkers (1997) during action potential clamp conditions. The current amplitude in our experiments, however, is much smaller than reported by Zaza et al. (1997), suggesting that the small compensation current during the wash out conditions is due to slight run down of I_Ca,L. An additional compensation current was observed in the presence of DIDS (Fig. 6B, middle panel, arrow). This DIDS-induced compensation current provides a mirror image of the contribution of the blocked component, i.e. I_Cl(Ca), to the normal pacemaking cycle. Figure 6C shows I_Cl(Ca), which was obtained by digitally subtracting the current trace recorded in presence of DIDS from the one recorded after wash out of the drug. I_Cl(Ca) measured using the action potential clamp technique is a transient outward current that becomes apparent near +5 mV and reaches a peak value of about 80 pA in the final phase of the action potential upstroke. The contribution of I_Cl(Ca) to the current generating the action potential waveform is limited to the final phase of the upstroke and the early plateau phase. Figure 6D shows the reconstructed net membrane current (I_m = - C_m dV/dt) of the action potential shown in Panel A. Note that I_Cl(Ca) is active in the phase when I_m declined from its large inward peak of about −600 pA to values near about −100 pA.

Using the action potential clamp technique, we observed I_Cl(Ca) in 4 out of 13 (30.8 %) single SA node cells tested, which agrees with our conventional voltage clamp experiments that show that about one third of the SA node cells exhibit I_Cl(Ca). The current was active late during the upstroke and early plateau phase. The average density of peak I_Cl(Ca) measured during the action potential clamp measurements is 1.16 ± 0.19 pA pF⁻¹ (n = 4), which is comparable with the current density measured during the conventional voltage clamp steps (Fig. 2D).

Role of I_Cl(Ca) in pacemaker activity

Our experiments demonstrate that I_Cl(Ca) is active late during the upstroke and early during the plateau phase of the action potential, resulting in a flattening of the top of the action potential in SA node cells. In atrial and ventricular cells, it was demonstrated that such reduction of action potential overshoot has implications for action potential conduction, especially under conditions of decreased intercellular coupling (Joyner et al. 1996; Wang et al. 2000; Huelsing et al. 2001). As a consequence of the reduced height of the early plateau phase of the action potential, the driving force for intercellular coupling current through gap junctions was found to be reduced, thereby increasing the minimum value of intercellular coupling conductance (G_c) required for successful action potential transfer (Joyner et al. 1996; Wang et al. 2000; Huelsing et al. 2001). To assess the functional role of I_Cl(Ca) in pacemaker activity, we carried out computer simulations using our previously published model of a rabbit SA node cell (Wilders et al. 1991), into which we incorporated the Gomis-Tena & Saiz (1999a, b)I_Cl(Ca) equations which are as set out in Methods.

Validation of I_Cl(Ca)

In a first set of simulations, we validated our model description of I_Cl(Ca) by comparing the model-generated I_Cl(Ca) with experimental data, both under voltage clamp and current clamp conditions. Figure 7A (left) shows the model I_Cl(Ca) in response to voltage clamp steps from a holding potential of-40 mV to test potentials ranging between −40 and +70 mV. We obtained current traces quite similar to those obtained with the same protocol in our experiments (Fig. 2C) by setting the model parameters K_m and n_h (eqn (2)) to 5 mm and 3, respectively. The model I-V relationship (Fig. 7A, right) is also quite similar to the one obtained experimentally (Fig. 2D). The value of the one remaining model parameter, i.e. P_Cl(Ca) (eqn (1)), was chosen such that the maximum current density in the model I-V relationship (1.43 pA pF⁻¹) was similar to the average maximum density that we observed experimentally (1.44 ± 0.45 pA pF⁻¹, n = 7). The differences in amplitude between the current traces of Fig. 2C and Fig. 7A (left) are largely due to the difference in size between the real cell (110 pF) and the model cell (32 pF).

Next, we carried out current clamp simulations with our SA nodal cell model, with the same I_Cl(Ca) equations and parameters as used in the voltage clamp simulations, to test to which extent the results of our current clamp and action potential clamp experiments could be reproduced. Figure 7B shows the spontaneous activity of the model cell with and without I_Cl(Ca) (continuous and dashed lines, respectively). Like the experimental case (Fig. 4), the most prominent effect of ‘blocking’ I_Cl(Ca) is a 4 mV increase in action potential overshoot. If we increase P_Cl(Ca) (eqn (1)) by a factor of 2, thus doubling I_Cl(Ca), as occurs during adrenergic stimulation (Fig. 5A), and then study the effects of ‘blocking’ I_Cl(Ca), the most prominent effect is a somewhat larger increase in action potential overshoot (7 mV), in accordance with the experimental observations (Fig. 5B). Figure 7C shows I_Cl(Ca) and the net membrane current (I_m) during the spontaneous activity of Fig. 7B. The time course of those currents resembles that observed experimentally in the action potential clamp experiment of Fig. 6. Also, with a value of 1.20 pA pF⁻¹, the peak amplitude of I_Cl(Ca) in the simulations compares well with the experimentally observed value of 1.16 ± 0.19 pA pF⁻¹ (n = 4).

Effects of I_Cl(Ca) on pacemaker synchronization

Subsequently, we tested the importance of I_Cl(Ca) for pacemaker synchronization between SA node cells, which is a prerequisite for normal action potential initiation in the SA node. Previously, we demonstrated that pacemaker synchronization in pairs of SA node cells at low G_c results mainly from the phase resetting effects of the action potential of one cell on the depolarization phase of the other cell, while at higher coupling conductances, the tonic, diastolic interaction prevails (Wilders et al. 1996; Verheijck et al. 1998b). Consequently, we hypothesized that I_Cl(Ca) may play a role in pacemaker synchronization under conditions of decreased intercellular coupling. To address this issue, we electrically coupled two model rabbit SA nodal cells via a variable G_c and estimated the critical value of G_c, i.e. the minimal value of G_c required for action potential synchronization. The critical G_c was determined in absence of I_Cl(Ca) and after incorporation of I_Cl(Ca) into the SA nodal cell model at distinct (difference in) cycle lengths. The cycle length was varied by changing the maximum conductance of I_f. As we reported previously (Wilders et al. 1996), critical G_c increases almost linearly with increasing difference in cycle length. The critical G_c in presence of I_Cl(Ca), however, did not differ in any significant way from the critical G_c in absence of I_Cl(Ca) (data not shown). Also, at higher values of G_c, I_Cl(Ca) had minimal effects on synchronization of the model cells. These results suggest that I_Cl(Ca) has a limited role in pacemaker synchronization.

Effects of I_Cl(Ca) on action potential conduction

In a final set of simulations, we assessed the importance of I_Cl(Ca) for action potential conduction between SA nodal and atrial cells. Previously, we demonstrated that our model rabbit SA nodal cell drives a real rabbit atrial cell at G_c > 0.55 nS (Joyner et al. 1998). In our simulations, we coupled the model SA node cell to a model rabbit atrial cell (Lindblad et al. 1996) and studied the critical G_c at which the SA node cell could successfully pace and drive the atrial cell. This critical G_c was determined in absence of I_Cl(Ca) and after incorporation of I_Cl(Ca) into the SA nodal cell model. The critical G_c in presence of I_Cl(Ca) did not differ significantly from the critical G_c in absence of I_Cl(Ca) (data not shown). Also at higher values of G_c, I_Cl(Ca) had minimal effects on action potential transfer between the model cells. These results suggest that I_Cl(Ca) has a limited role in action potential conduction between SA nodal and atrial cells.

DISCUSSION

In this paper we employed patch-clamp methodology and computer simulations to investigate the presence and functional role of I_Cl(Ca) in single SA node cells of rabbit. Using the amphotericin-permeabilized patch clamp technique, we found a transient I_Cl(Ca) in one third of the cells. The I-V relationship of I_Cl(Ca) is bell shaped with an activation threshold around −20 mV, and a maximum close to +40 mV. Using the action potential clamp technique, we demonstrate that I_Cl(Ca) is active late during the upstroke and early during the plateau phase of the action potential. Adrenoceptor stimulation with noradrenaline doubles I_Cl(Ca), which resulted in a decrease of the action potential overshoot. Both under control conditions and upon adrenoceptor stimulation, blockade of I_Cl(Ca) caused an increase in action potential overshoot and APD₂₀, but did not otherwise affect pacemaker activity. Our computer simulations demonstrate that I_Cl(Ca) has a neither a role in pacemaker synchronization nor in action potential conduction.

Selective effect of DIDS on I_Cl(Ca)

In the present study, we used 0.2 mm DIDS as a tool to demonstrate the presence of I_Cl(Ca) and to study the functional role of this current in SA node cells. DIDS, however, is not a specific blocker for the Ca²⁺-activated Cl⁻ conductance, but blocks also the swelling-activated Cl⁻ current (I_Cl(swell)) and carriers involved in pH_i regulation, i.e. the Na⁺-HCO₃⁻ cotransporter and the Cl⁻-HCO₃⁻ exchanger (Leem & Vaughan-Jones, 1998; Hume et al. 2000).

We cannot exclude the possibility that I_Cl(swell) also contributes to our DIDS-sensitive current, but we presume that the role of I_Cl(swell) is limited for the following reasons. Firstly, we found a transient DIDS-sensitive current with a bell shaped I-V relation. These characteristics compare to those of I_Cl(Ca) rather than to the characteristics of a swelling-activated Cl⁻ conductance, which is time independent and exhibits outward rectification (for review, see Vandenberg et al. 1996; Sorota, 1999; Hume et al. 2000). Secondly, the ability to observe a transient DIDS-sensitive current immediately after formation of the whole-cell perforated-patch configuration suggests that the observed current is unrelated to I_Cl(swell), which develops over time (Vandenberg et al. 1996). Thirdly, I_Cl(swell) was reported to be insensitive to β-receptor stimulation in SA node cells (Hagiwara et al. 1992), while in our experiments the DIDS-sensitive current was doubled in the presence of noradrenaline.

In our experiments, we expect no influences of DIDS on pH_i. We used a Hepes-buffered solution where pH_i is regulated mainly by DIDS-insensitive Na⁺-H⁺ exchange (Buckler et al. 1990; Leem & Vaughan-Jones, 1998). This is supported by the absence of changes in cycle length, while both alakalinization and acidification alter significantly the frequency of spontaneous activity in rabbit sinoatrial node (Satoh & Seyama, 1986).

Recently, it was demonstrated that DIDS (10 μm) inhibits the I_Ca,L and I_K in colonic myocytes (Dick et al. 1999). We found no indications that the main cation currents in SA node cells were significantly changed by as much as 0.2 mm DIDS (Fig. 1). Our findings agree with observations made in atrial and ventricular myocytes of dog, rabbit and sheep where the drug failed to influence the I_Ca,L, the I_K, the 4-aminopyridine sensitive transient outward current, the inward rectifier current, or the action potential when I_Cl(Ca) was already blocked by 20 mm internal EGTA (Zygmunt & Gibbons, 1992; Zygmunt, 1994; Kawano et al. 1995; Zygmunt et al. 1997; Verkerk et al. 2001). In addition, no effects of DIDS on [Ca²⁺]_i-transients were found (Sipido et al. 1993; Verkerk et al. 2001), further indicating that Ca²⁺-modulated membrane currents other than I_Cl(Ca) do not play a role in the effects of the drug on action potential configuration. Thus in our experiments, the effects of DIDS on the action potentials and membrane currents seem entirely attributable to blockade of I_Cl(Ca).

Presence of I_Cl(Ca) in SA node cells

It has been demonstrated that SA node cells contain at least two distinct Cl⁻ currents. Stretch-activated anion currents were observed in single rabbit SA node cells (Hagiwara et al. 1992; Arai et al. 1996; Lei & Kohl, 1998; Kohl et al. 1999). Moreover, an angiotensin-II-activated chloride current was found in rabbit SA node cells suggesting that SA node cells contain protein kinase-C-sensitive Cl⁻ channels (Bescond et al. 1994). In our experiments, we additionally demonstrate that I_Cl(Ca) is present in one third of the rabbit SA node cells (Figs 2, 5 and 6). It is unknown why I_Cl(Ca) is non-uniformly distributed among the SA node cells. All tested cells displayed regular contractions indicating that Ca²⁺ release from the SR, and thus the substrate for activating I_Cl(Ca), was significantly present.

The non-uniform distribution of I_Cl(Ca) may be related to the inhomogeneity of function and structure of the SA node (for review, see Bouman & Jongsma, 1995; Boyett et al. 2000). Functionally, a distinction has been made in the SA node between the site of earliest activation, i.e. the primary pacemaker area, and a group of nodal fibres that are triggered to excitation, i.e. the subsidiary pacemaker area. Action potentials in the primary pacemaker area have a lower amplitude, dV/dt_max, MDP and a higher DDR and action potential duration compared with those in the subsidiary pacemaker area (Bouman & Jongsma, 1995; Boyett et al. 2000). In our study, action potential characteristics did not differ significantly between isolated SA node cells with and without I_Cl(Ca) (Table 1). We therefore suggest that the role of pacemaker cell type in the non-uniform distribution of I_Cl(Ca) is limited. Structurally, Boyett and coworkers found a cell size-dependent variation in densities of various cation currents in isolated SA node cells (for review, see Boyett et al. 2000), although this is not a consistent finding (Wilders et al. 1996; Verheijck et al. 1998a; Wu et al. 2001; Mangoni & Nargeot, 2001). In the present study, we did not observe a cell size-dependent variation in density of the anion current, I_Cl(Ca). The cell capacitance did not differ significantly between SA node cells with and without I_Cl(Ca) (Table 1). We therefore suggest that the role of cell size in the non-uniformly distributed I_Cl(Ca) is limited. Recently, Wu et al. (2001) found a morphology-dependent variation in densities of a cation current in isolated SA node cells. They observed a smaller I_f in spindle cells compared with spider-like cells. In the present study, we only used spindle-like cells. Thus, a possible role of morphological differences in the non-uniform distribution of I_Cl(Ca) was excluded.

In our experiments, the pipette solution did not contain sodium. Sipido et al. (1997) and Faber & Rudy (2000) have demonstrated that this may have implications for the amplitude of Ca²⁺ transients. Sipido et al. (1997) performed experiments with low (0 mm) or high [Na⁺]_i (20 mm) and found in guinea-pig ventricular cells that Ca²⁺ transients were smaller using low [Na⁺]_i. Comparable findings were made by Faber & Rudy (2000) using a detailed theoretical model of a guinea-pig ventricular cell. In our experiments, a lower Ca²⁺ transient may have led to a smaller current density of I_Cl(Ca). However, we think that the absence of [Na⁺]_i is not responsible for the heterogeneity of presence of I_Cl(Ca) in SA node cells. All cells showed clear contractions, and upregulation of [Ca²⁺]_i transient by 1 μm noradrenaline could not activate I_Cl(Ca) in cells that had no apparent current at baseline. Moreover, we have used a similar pipette solution in previous studies on sheep Purkinje and ventricular cells (Verkerk et al. 2000, 2001), and found that I_Cl(Ca) was present in all cells tested.

It has been suggested, based on cell size-dependent variation, that the expression of Ca²⁺-handling proteins of the ryanodine receptor declines from the periphery to the centre of rabbit SA node cells (Musa et al. 1999). This suggests that intracellular Ca²⁺ regulation may vary between different parts of the SA node. In a recent report by the same group, it was found that intracellular Ca²⁺ transients as well as diastolic Ca²⁺ concentration depend on SA nodal cell size (Lancaster et al. 2001). It is tempting to speculate that the non-uniform distribution of I_Cl(Ca) in SA node cells is due to heterogeneity of intracellular Ca²⁺ regulation within the SA node. Further studies are required to clarify this issue.

Functional role of I_Cl(Ca) in SA node

Recently, it has become clear that Ca²⁺ released from the SR plays an important role in regulating pacemaker activity in mammalian SA node, probably by modulating membrane currents and exchangers (Hata et al. 1996; Li et al. 1997; Satoh, 1997; Rigg et al. 2000; Bogdanov et al. 2001). In the present study, our action potential clamp measurements demonstrate that Ca²⁺-activated Cl⁻ current only is present late during the upstroke of the action potential and early during the plateau phase (Fig. 6). Both this transient nature of the current and its time of occurrence during the pacemaker cycle suggest that I_Cl(Ca) plays a role in modulation of the action potential overshoot. This hypothesis was supported by the finding that blockade of the current increased the action potential overshoot and APD₂₀ significantly (Fig. 4, Table 2), which agrees with findings in atrial and ventricular cells (Wang et al. 1995; Hiraoka & Kawano, 1989; Kawano & Hiraoka, 1991). We found that the modulating role of I_Cl(Ca) on the action potential overshoot was increased during adrenoceptor stimulation (Fig. 5B, Table 3), most likely due to the potentiated I_Cl(Ca) (Fig. 5A). In dog, rabbit and sheep ventricular cells, I_Cl(Ca) also increases during adrenoceptor stimulation (Tseng & Hoffmann 1989; Zygmunt & Gibbons, 1991; Verkerk et al. 2001) and it was demonstrated that such potentiation prevents drastic action potential prolongation (Verkerk et al. 2001). In our experiments, however, blockade of I_Cl(Ca) failed to significantly affect APD₅₀ or APD₁₀₀, both under control conditions and upon adrenoceptor stimulation (Fig. 4 and Fig. 5, Tables 2 and 3). Nor did we observe significant effects of I_Cl(Ca) blockade on dV/dt_max, cycle length, and DDR under control conditions and adrenoceptor stimulation (Fig. 4 and Fig. 5, Tables 2 and 3). Our experiments thus in addition, show that even under conditions of enhanced I_Cl(Ca) density, this current does not play a significant role in controlling beating rate in single SA node cells.

In our study, DIDS caused a flattening of the top of the SA node action potential, which is more pronounced during adrenoceptor stimulation (Fig. 4 and Fig. 5, Tables 2 and 3). In atrial and ventricular cells, it was demonstrated that such reduction of action potential overshoot has implications for action potential conduction, especially under conditions of decreased cellular coupling (Joyner et al. 1996; Wang et al. 2000; Huelsing et al. 2001). Our computer simulations, however, neither revealed an important role of I_Cl(Ca) in pacemaker synchronization of SA node cells nor in action potential transfer between SA node and atrial cells.

Conclusions

We found an I_Cl(Ca) in single SA node cells which modulates importantly the action potential overshoot. However, we did not find a clear physiological importance of I_Cl(Ca) for the generation or conduction of the SA nodal action potential.