Heterogeneous expression of the delayed-rectifier K⁺ currents i_K,r and i_K,s in rabbit sinoatrial node cells

Abstract

1. The electrical activity of sinoatrial node cells is heterogeneous. To understand the reasons for this, the density of the delayed-rectifier K⁺ current and its two components, i_K,r and i_K,s, as a function of the size (as measured by cell capacitance) of rabbit sinoatrial node cells was investigated using the whole-cell voltage-clamp technique at 35 °C.

2. i_K,r and i_K,s were isolated using E-4031 and 293B. Features of the E-4031-sensitive and 293B-insensitive currents corresponded well to those of i_K,r, while features of the E-4031-insensitive and 293B-sensitive currents corresponded well to those of i_K,s.

3. The densities of the outward current under control conditions and the drug-sensitive and -insensitive currents were significantly (P < 0.01) correlated with cell capacitance, with current densities being greater in larger cells.

4. The effects of partial blockade of i_K,r by 0.1 μm E-4031 on spontaneous action potentials were greater in smaller cells.

5. It is concluded that there are cell size-dependent differences in the density of the i_K,r and i_K,s components, and these may be involved in the heterogeneity of the electrical activity of single sinoatrial node cells as well as that of the intact sinoatrial node.

The sinoatrial (SA) node is known to be a heterogeneous tissue, and the electrophysiological properties of SA node cells vary in different regions of the SA node (Bleeker et al. 1980; Boyett et al. 2000). The action potential is usually first initiated in the centre of the SA node (distant from the surrounding atrial muscle). Normally, the function of the periphery of the SA node (at the border of the SA node with the atrial muscle) is to conduct the action potential from the centre to the surrounding atrial muscle, although it can take over the role as the leading pacemaker site (Boyett et al. 2000). Action potentials recorded from the centre of the rabbit SA node have a more positive take-off potential, slower upstroke, longer duration and less negative maximum diastolic potential than action potentials recorded from the periphery; the intrinsic pacemaker activity of the centre is also slower than that of the periphery (Bleeker et al. 1980; Kodama & Boyett, 1985; Kodama et al. 1997; Boyett et al. 1999). We have suggested that the regional differences in electrical activity are mainly the result of regional differences in the intrinsic properties of the cells (rather than electrotonic influences), because a similar heterogeneity in the action potential characteristics is observed in single SA node cells from the rabbit (Honjo et al. 1996). Furthermore, cell size increases from the centre to the periphery of the SA node (Bleeker et al. 1980; Masson-Pévet et al. 1984; Oosthoek et al. 1993; Boyett et al. 2000), and the smaller cells (presumably from the centre) show electrical activity characteristic of the centre, whereas the larger cells (presumably from the periphery) show electrical activity characteristic of the periphery (Honjo et al. 1996). We have suggested that cell size-dependent differences in the density of ionic currents underlie the regional differences in electrical activity within the SA node (Honjo et al. 1996, 1999; Lei et al. 2000). The lack of TTX-sensitive Na⁺ currents (i_Na) and a low density of the hyperpolarisation-activated current (i_f) in small cells (presumably from the centre) may be responsible, at least in part, for the slower upstroke and slower pacemaker activity of these cells (Honjo et al. 1996). Low densities of the transient and sustained components of the 4-aminopyridine (4-AP)-sensitive current in the small cells may be partially responsible for the longer duration of the action potential in the centre of the SA node compared to that observed in the periphery (Honjo et al. 1999; Lei et al. 2000).

The delayed-rectifier K⁺ current, i_K, is an important current in SA node cells. The activation of i_K plays an essential role in the repolarisation of the action potential, and its deactivation plays a key role in pacemaker depolarisation (Irisawa et al. 1993). Recently, investigations in rabbit SA node cells have shown that i_K is composed of two distinct components: a rapidly activating component, i_K,r, and a slowly activating component, i_K,s (Shibasaki, 1987; Ito & Ono, 1995; Verheijck et al. 1995; Lei & Brown, 1996; Sato et al. 1998). i_K,r is known to play a major role in the pacemaker activity of SA node cells, and partial or complete blockade of the current by E-4031 or dofetilide has been reported to slow or abolish pacemaker activity (Verheijck et al. 1995; Lei & Brown, 1996). We have hypothesised that i_K,r may be an important contributor to the regional heterogeneity in electrical activity in the SA node based on several lines of evidence (Kodama et al. 1999; Zhang et al. 2000): (1) small balls of tissue from the centre of the rabbit SA node are more sensitive to partial blockade of i_K,r than small balls of tissue from the periphery (Kodama et al. 1999); (2) in an immunocytochemical study of the ferret, ERG (the channel responsible for i_K,r) labelling was observed in the crista terminalis (where the periphery of the SA node is present in the rabbit) but was absent from the intercaval region (where the centre of the SA node is located in the rabbit; Brahmajothi et al. 1997); and (3) in mathematical models of action potentials in the centre and periphery of the SA node, we had to set the density of i_K,r to be 6.2 times lower in the centre than in the periphery to account for the difference in maximum diastolic potential and sensitivity to blockade of i_K,r (Zhang et al. 2000).

The present study was designed to test the hypothesis that there are cell size-dependent differences in the density of i_K,r and i_K,s in SA node cells.

METHODS

Cell isolation

Cells were isolated as described by Lei & Brown (1996) with a few minor modifications. Briefly, a 500-1000 g New Zealand White rabbit was killed by cervical dislocation and its heart removed quickly and placed in oxygenated Tyrode solution at 37 °C. Three to four strips of SA node tissue (∼3 mm × 0.5 mm) were cut perpendicular to the crista terminalis. The strips were placed for 5 min in Ca²⁺-free Tyrode solution at 37 °C, incubated for 30-40 min in 230 units ml⁻¹ collagenase (Type I, Sigma Chemical Co., Poole, UK) and 15 units ml⁻¹ elastase (Type IIA, Sigma) in Ca²⁺-free Tyrode solution at 37 °C, and then stored for at least 1 h in Kraft-Brühe (KB) medium at 4 °C. Single cells were released from the strips by glass pipette suction. In the present study, a total of 55 cells from 18 hearts were studied.

Whole-cell current and voltage clamp

During experiments, cells were superfused with Tyrode solution (plus appropriate drugs or blockers) at ∼1 ml min⁻¹ at 35 °C. The whole-cell patch-clamp technique was used for electrical recording from single SA node cells with amphotericin-permeabilised patches. Amphotericin (200 μg ml⁻¹) was added to the pipette solution just before use. Pipettes (tip diameter, ∼1-2 μm; resistance, 3-8 MΩ) were made from 1 mm diameter glass (Clark Electromedical Instruments, Reading, UK) using a Narishige pipette puller (PP-83, Narishige Scientific Instruments Laboratory, Tokyo, Japan). An Axopatch-1C patch-clamp amplifier (Axon Instruments Inc., Foster City, CA, USA) was used for current and voltage clamping. Cell capacitance (C_m) was obtained from the capacitance compensation control of the amplifier after the whole-cell capacity current (in response to 5 ms pulses to -70 mV at 100 Hz from a holding potential of -60 mV) was eliminated. In a previous study (Honjo et al. 1996), the accuracy of this method was checked by integrating the area of the uncompensated capacity current and fitting an exponential function to the decay of the uncompensated capacity current. The series resistance was electronically compensated (> 80 %) and the current signal was filtered by a low-pass Bessel filter with a cut-off frequency of 5 kHz (-3 dB). During experiments, electrical signals were displayed on an oscilloscope (5111A, Tektronix, The Netherlands) and a chart recorder (2007, Gould, France). Data were digitised using an AD/DA converter (Digidata 1200A, Axon Instruments Inc.) and stored on computer (sample rate, 1-2 kHz) for later analysis using pCLAMP version 6.2 software (Axon Instruments Inc.).

Solutions

Tyrode solution contained (mm): 140 NaCl, 5.4 KCl, 1.8 CaCl₂, 1 MgCl₂, 10 glucose and 5 Hepes, pH 7.4 (NaOH). Ca²⁺-free Tyrode solution was prepared as for normal Tyrode solution, but without CaCl₂. The KB solution contained (mm): 25 KCl, 80 l-glutamic acid, 20 taurine, 10 KH₂PO₄, 3 MgCl₂, 10 glucose, 0.5 EGTA and 10 Hepes, pH 7.4 (KOH). The pipette solution contained (mm): 140 KCl, 1.8 MgSO₄, 5 Hepes and 1 EGTA, pH 7.4 (KOH); amphotericin was added prior to use, as described above. Nisoldipine (Bayer Pharmaceutical Co., Newbury, UK) was added to Tyrode solution to block the L-type Ca²⁺ current, i_Ca,L, and E-4031 (1-[2-(6-methyl-2pyridyl)ethyl]-4-(4-methylsulphonylaminobenzoyl)piperidine; Eisai Pharmaceuticals, Tokyo, Japan) and chromanol 293B (293B; trans-6-cyano-4-4-(N-ethysulphonyl-N-methylamino)-3-hydroxy-2,2-dimethyl-chromane; gift from Hoechst AG, Frankfurt, Germany) were used to isolate i_K,r and i_K,s. Nisoldipine was dissolved in ethanol, E-4031 was dissolved in distilled water, and 293B was dissolved in distilled water and ethanol (50:50), to make 10 mm stock solutions. The nisoldipine and E-4031 stock solutions were stored at 4 °C, while the 293B stock solution was stored at room temperature.

Statistics

All results are presented as means ±s.e.m. (n = number of cells). Statistical significance was determined by Student's t test for paired or unpaired observations. Linear regression analysis was used for correlations. P < 0.05 was considered to indicate a significant difference. All statistical analysis was carried out using SigmaPlot version 4.0 software (Jandel Scientific, San Rafael, CA, USA).

RESULTS

Presence of E-4031-sensitive, E-4031-insensitive, 293B-sensitive and 293B-insensitive currents

E-4031 (3 μm), a selective i_K,r blocker, and 293B (10 μm), a selective i_K,s blocker, were used to dissect out i_K,r and i_K,s. Experiments were only carried out on cells showing spontaneous activity - they were spindle or spider shaped, with no obvious or even faint striations. Figure 1A-E shows an example of drug-sensitive and -insensitive currents. In the experiment shown in Fig. 1 (as well as those shown in Fig. 2 and Fig. 3), currents were evoked by 1 s voltage-clamp pulses to between -50 and +40 mV from a holding potential of -60 mV, in the presence of 300 nm nisoldipine to block i_Ca,L. Figure 1A shows the current recorded under control conditions, and Fig. 1B shows the current recorded after 5 min of exposure to 3 μm E-4031. E-4031 reduced both the outward current during the pulse and the outward tail current after the pulse. Figure 1C shows the E-4031-sensitive current (obtained by subtracting the current seen in Fig. 1B from that seen in Fig. 1A). The E-4031-sensitive current is similar to records of i_K,r published previously (Ito & Ono, 1995; Lei & Brown, 1996; Cheng et al. 1999). Figure 1D shows the current recorded after a 5 min exposure to 10 μm 293B (3 μm E-4031 still present). After the application of 293B the outward current during the pulse was further reduced (but not abolished), and the outward tail current after the pulse was abolished. Figure 1E shows the 293B-sensitive current (obtained by subtracting the current shown in Fig. 1D from that shown in Fig. 1B). The 293B-sensitive current is similar to records of i_K,s published previously (Lei & Brown, 1996; Cheng et al. 1999). This result confirms that both i_K,s and i_K,r are present in rabbit SA node cells.

Figure 1. E-4031-sensitive, E-4031-insensitive, 293B-sensitive and 293B-insensitive currents.

Currents were evoked by 1 s voltage-clamp pulses at 2.5 Hz to between -50 and +40 mV from a holding potential of -60 mV in the presence of 300 nm nisoldipine to block Ca²⁺ current. A-E, currents recorded from a SA node cell (membrane capacitance, C_m, is given): A, current under control conditions; B, current in the presence of 3 μm E-4031; C, E-4031-sensitive current (obtained by subtracting the current obtained in the presence of E-4031 from that obtained under control conditions); D, current obtained in the presence of both 3 μm E-4031 and 10 μm 293B; E, 293B-sensitive current (obtained by subtracting the current obtained in the presence of E-4031 and 293B from that obtained in the presence of E-4031 only. The dashed lines in this and other figures show zero current. F, activation curves. The amplitude of the tail current after a pulse (normalised to the maximum amplitude of the tail current) is plotted against the preceding pulse potential; tail current amplitude under control conditions (i_K,total; C_m, 34 ± 2 pF, 16-62 pF, n = 21) and the amplitude of the E-4031-sensitive (C_m, 32 ± 2 pF, range 16-58 pF, n = 10), E-4031-insensitive (C_m, 33 ± 3 pF, 16-58 pF, n = 13), 293B-sensitive (C_m, 37 ± 5 pF, 14-62 pF, n = 10) and 293B-insensitive (C_m, 37 ± 5 pF, 14-62 pF, n = 10) tail currents are plotted. Means ±s.e.m. shown. E-4031 and 293B data were obtained from different cells. The smooth curves shown were obtained by fitting the Boltzmann equation to the experimental data.

Figure 2. Relationship between the density of i_K,total and C_m.

A, examples of membrane current recorded from a small (25 pF) and a large (40 pF) cell under control conditions. The voltage-clamp protocol used was the same as for Fig. 1. B, relationship between the density of current at the end of the pulse and the pulse potential under control conditions for groups of small (C_m, 24 ± 1 pF, n = 9, ○) and large (C_m, 44 ± 2 pF, n = 12, •) cells (the range of C_m values as well as the number of cells used is given in the figure). C, relationship between the density of the tail current (i_K,total) after a pulse and the preceding pulse potential under control conditions for the groups of small (○) and large cells (•). Means ±s.e.m. are shown.

Figure 3. Relationship between the density of drug-sensitive and -insensitive currents and C_m.

A and B, examples of E-4031-sensitive (A) and -insensitive (B) currents recorded from small and large cells (C_m of cells given). The voltage-clamp protocol used was the same as for Fig. 1. The E-4031-sensitive current was obtained by subtracting the current obtained during 5 min of exposure to 3 μm E-4031 (the E-4031-insensitive current) from that obtained under control conditions. C and D, relationship between the density of E-4031-sensitive (C) and -insensitive (D) currents at the end of the pulse and the pulse potential for groups of small (C_m, 23.5 ± 1.5 pF, n = 6, ○) and large (C_m, 41 ± 3 pF, n = 7, •) cells (the range of C_m values as well as the number of cells used is given in the figure). E and F, relationship between the density of the E-4031-sensitive (E) and -insensitive (F) tail currents after a pulse and the preceding pulse potential for the groups of small and large cells. G and H, examples of 293B-sensitive (G) and -insensitive (H) currents recorded from small and large cells (the C_m of the cells is shown). The voltage-clamp protocol used was the same as for Fig. 1. The 293B-sensitive current was obtained by subtracting the current obtained during 5 min of exposure to 10 μm 293B (the 293B-insensitive current) from that obtained under control conditions. I and J, relationship between the density of 293B-sensitive (I) and -insensitive (J) currents at the end of the pulse and the pulse potential for groups of small (C_m, 22 ± 2 pF, n = 5, ○) and large (C_m, 52 ± 3 pF, n = 5, •) cells (the range of C_m values as well as the number of cells used is given in the figure). K and L, relationship between the density of the 293B-sensitive (K) and -insensitive (L) tail currents after a pulse, and the preceding pulse potential for the groups of small and large cells. In all graphs, means ±s.e.m. are shown.

Figure 1F shows activation curves that were constructed by plotting the amplitude of the outward tail currents (normalised to the maximum outward tail current amplitude) obtained at a holding potential of -60 mV against the potential obtained during the preceding 1 s pulse. The filled squares show data under control conditions - this is presumably the sum of both i_K,r and i_K,s and will henceforth be referred to as i_K,total. The upright filled triangles and filled circles show E-4031-sensitive and 293B-insensitive tail currents, respectively. Data on E-4031-sensitive and 293B-insensitive currents were obtained from different cells 5 min after the application of 3 μm E-4031 or 10 μm 293B; the E-4031-sensitive current was calculated by subtracting the current obtained in the presence of E-4031 from that obtained under control conditions. Since the outward tail current that occurred after a pulse was abolished in the presence of both E-4031 and 293B (Fig. 1D), it is possible that i_K,r and i_K,s are the only significant currents contributing to the outward tail current. If 3 μm E-4031 and 10 μm 293B selectively abolish i_K,r and i_K,s, respectively (see Discussion), it follows that E-4031-sensitive and 293B-insensitive tail currents should be measures of i_K,r. The putative activation curves for i_K,r obtained from the E-4031-sensitive and 293B-insensitive tail currents (Fig. 1F) are similar, as expected, and are similar to the activation curves for i_K,r published by others (see Discussion).

In Fig. 1F, the inverted filled triangles and open circles show E-4031-insensitive and 293B-sensitive tail currents (obtained in an analogous manner to the E-4031-sensitive and 293B-insensitive currents). It follows from the arguments above that E-4031-insensitive and 293B-sensitive tail currents should both be measures of i_K,s. The putative activation curves for i_K,s obtained from the E-4031-insensitive and 293B-sensitive tail currents (Fig. 1F) are similar, as expected, and are similar to the activation curves for i_K,s published by others (see Discussion).

In Fig. 1F, the continuous curves were obtained by fitting the experimental data by the Boltzmann equation: a = 1/(1 + exp[(V_m - V_1/2)/k]), where a is the activation variable, V_m is the membrane potential, V_1/2 is the half-activation voltage and k is the slope factor. Table 1 shows V_1/2 and k for the activation curves obtained from i_K,total, E-4031-sensitive, E-4031-insensitive, 293B-sensitive and 293B-insensitive tail currents.

Table 1.

Activation parameters of i_K,total, putative i_K,r and putative i_K,s

Current	Tissue	V1/2 (mV)	k (mV)	Vh (mV)	Temperature(°C)	Reference
iK,total	Rabbit SA node	−20.2 ± 0.8	9.3 ± 0.7	−60	35	This study (n = 21)
iK,total	Rabbit SA node	−25.1	7.4	−43	36–37	Shibasaki (1987)
iK,total	Rabbit SA node	−14.5	11.5	−50	36–37	Lei & Brown (1996)
Putative iK,r
E-4031 sensitive	Rabbit SA node	−27.7 ± 1.1	5.9 ± 1.1	−60	35	This study (n = 10)
293B insensitive	Rabbit SA node	−34.4 ± 1.5	6.0 ± 1.3	−60	35	This study (n = 13)
E-4031 sensitive	Rabbit SA node	−23.2	10.6	−60	35	Ono & Ito (1995)
Dofetilide sensitive	Rabbit SA node	−17.4	9.7	−50	36–37	Lei & Brown (1996)
E-4031 sensitive	Guinea-pig atrium	−19.3	5.2	−40	35	Sanguinetti & Jurkiewicz (1991)
E-4031 sensitive	Guinea-pig ventricle	−21.5	7.5	−40	35	Sanguinetti & Jurkiewicz (1990)
Putative iK,s
E-4031 insensitive	Rabbit SA node	+3.0 ± 1.2	15.9 ± 1.3	−60	35	This study (n = 13)
293B sensitive	Rabbit SA node	−3.7 ± 1.2	13.6 ± 1.3	−60	35	This study (n = 10)
Dofetilide insensitive	Rabbit SA node	+15.6	14.7	−50	36–37	Lei & Brown (1996)
E-4031 insensitve	Guinea-pig atrium	+24.0	15.7	−40	35	Sanguinetti & Jurkiewicz (1991)
E-4031 insensitive	Guinea-pig ventricle	+15.7	12.7	−40	35	Sanguinetti & Jurkiewicz (1990)

V_h, holding potential; k, slope factor; SA, sinoatrial.

correlation between the density of i_K,total and C_m

Figure 2 shows the relationship between the density of i_K,total under control conditions and cell size (as measured by C_m). Figure 2A shows two sets of current traces obtained from small and large SA node cells with C_m values of 25 and 40 pF, respectively. To test for a cell size-dependent difference in the density of i_K,total, 21 cells were divided into two groups: small cells with C_m < 30 pF and larger cells with C_m > 30 pF. Figure 2B shows current-voltage relationships for the two groups of cells: the density of the current at the end of the pulse is plotted against the pulse potential. The current at the end of the pulse will include i_K,total as well as other currents flowing during the pulse (i.e. the E-4031- and 293B-insensitive current shown in Fig. 1D). Figure 2C shows the density of the outward tail current after the pulse plotted against the pulse potential for the two groups of cells. As discussed above, the outward tail current is attributed to i_K,total alone. The density of both the pulse and tail currents was lower at each potential in the smaller cells. For example, at the end of a pulse to +40 mV, the density of the outward current was 13.2 ± 1.8 and 30.3 ± 4.4 pA pF⁻¹ (P < 0.01) in the groups of small and large cells, respectively. After a pulse to +40 mV, the density of the i_K,total tail current was 2.1 ± 0.2 and 6.6 ± 0.7 pA pF⁻¹ (P < 0.01) in the groups of small and large cells, respectively.

correlation between c_m and the density of E-4031-sensitive and -insensitive currents

Figure 3 shows the relationship between the density of E-4031-sensitive and -insensitive currents and cell size. The E-4031-sensitive current was obtained by subtracting the current obtained in the presence of 3 μm E-4031 for 5 min (the E-4031-insensitive current) from that obtained under control conditions. Figure 3A and B shows examples of E-4031-sensitive and -insensitive currents obtained from small and large cells. In Fig. 3C-F, the densities of E-4031-sensitive and -insensitive currents are plotted against the pulse potential for groups of small (< 30 pF) and large (> 30 pF) cells. The density of the current at the end of the pulse is plotted in Fig. 3C and D, and the density of the tail current is plotted in Fig. 3E and F. The E-4031-sensitive current observed both during and after the pulse is expected to be i_K,r. The degree of inward rectification displayed by the E-4031-sensitive current at the end of the pulse (Fig. 3C) is considered in the Discussion. The E-4031-insensitive current observed during the pulse is expected to include i_K,s and the E-4031-insensitive tail current is expected to be i_K,s. The densities of the E-4031-sensitive and -insensitive currents at the end of the pulse and the E-4031-sensitive and -insensitive tail currents were lower in the smaller cells at each potential. For example, in the groups of small and large cells, at the end of a pulse to +40 mV, the density of the E-4031-sensitive current was 3.6 ± 0.7 and 9.0 ± 0.9 pA pF⁻¹ (P < 0.01), respectively, and after the pulse to +40 mV, the density of the E-4031-sensitive tail current was 1.3 ± 0.4 and 4.2 ± 1.0 pA pF⁻¹ (P < 0.05), respectively. In the groups of small and large cells, at the end of a pulse to +40 mV, the density of the E-4031-insensitive current was 6.0 ± 1.4 and 14.4 ± 1.5 pA pF⁻¹ (P < 0.05), respectively, and after the pulse to +40 mV, the density of the E-4031-insensitive tail current was 1.4 ± 0.2 and 2.4 ± 0.5 pA pF⁻¹ (P < 0.05), respectively.

correlation between c_m and the density of 293B-sensitive and -insensitive currents

Figure 3 also shows the relationship between the density of 293B-sensitive and -insensitive currents and cell size. The 293B-sensitive current was obtained by subtracting the current obtained during 5 min of exposure to 10 μm 293B (the 293B-insensitive current) from that obtained under control conditions. Figure 3G and H shows examples of 293B-sensitive and -insensitive currents obtained from small and large cells. In Fig. 3I-L, the densities of 293B-sensitive and -insensitive currents are plotted against the pulse potential for groups of small (< 30 pF) and large (> 30 pF) cells. The density of the current at the end of the pulse is plotted in Fig. 3I and J, and the density of the tail current is plotted in Fig. 3K and L. The 293B-sensitive current observed both during and after the pulse is expected to be i_K,s. The 293B-insensitive current observed during the pulse is expected to include i_K,r and the 293B-insensitive tail current is expected to be i_K,r. The densities of the 293B-sensitive and -insensitive currents at the end of the pulse, and the 293B-sensitive and -insensitive tail currents were lower in the smaller cells at each potential. For example, in the groups of small and large cells, at the end of a pulse to +40 mV, the density of the 293B-sensitive current was 4.3 ± 1.1 and 9.2 ± 2.4 pA pF⁻¹ (P < 0.05), respectively, and after the pulse to +40 mV, the density of the 293B-sensitive tail current was 2.3 ± 0.5 and 6.6 ± 1.4 pA pF⁻¹ (P < 0.05), respectively. In the groups of small and large cells, at the end of a pulse to +40 mV, the density of the 293B-insensitive current was 2.6 ± 0.4 and 13.6 ± 2.3 pA pF⁻¹ (P < 0.01), respectively, and after the pulse to +40 mV, the density of the 293B-insensitive tail current was 2.3 ± 0.5 and 4.4 ± 1.4 pA pF⁻¹ (P < 0.05), respectively.

Summary

In Fig. 4, the density of the outward tail current at the holding potential of -60 mV after a 1 s pulse to +40 mV is plotted against C_m. Figure 4A shows the relationship between the density of i_K,total tail current under control conditions and C_m in 28 cells. The density of i_K,total tail current was significantly correlated with C_m (r² = 0.44, P < 0.01, n = 28); the density of i_K,total tail current was greater in larger cells. Figure 4B shows the relationship between the density of the E-4031-sensitive and 293B-insensitive tail current and C_m in 19 and 13 cells, respectively. Both the E-4031-sensitive and 293B-insensitive tail currents are likely to be i_K,r. The density of the E-4031-sensitive/293B-insensitive tail current was significantly correlated with C_m (r² = 0.31, P < 0.01, n = 32); the density of the E-4031-sensitive/293B-insensitive tail current was greater in larger cells. Figure 4C shows the relationship between the density of the E-4031-insensitive and 293B-sensitive tail current and C_m in 18 and 14 cells, respectively. Both the E-4031-insensitive and 293B-sensitive tail currents are likely to be i_K,s. The density of the E-4031-insensitive/293B-sensitive tail current was significantly correlated with C_m (r² = 0.33, P < 0.01, n = 32); the density of the E-4031-insensitive/293B-sensitive tail current was greater in larger cells.

Figure 4. Correlation between the densities of i_K,total, putative i_K,r, putative i_K,s and C_m.

Relationships are shown between the density of the tail current after a 1 s voltage-clamp pulse to +40 mV from a holding potential of -60 mV and C_m. The data were fitted with straight lines by linear regression and the r² and P values are given. A, tail current (i_K,total) under control conditions in 28 cells. B, E-4031-sensitive (•) and 293B-insensitive (□) tail current (putative i_K,r tail current) in 19 and 13 cells, respectively. The E-4031-sensitive current was obtained by subtracting the current obtained during 5 min of exposure to 3 μm E-4031 from that obtained under control conditions. The 293B-insensitive current was recorded during 5 min of exposure to 10 μm 293B. C, E-4031-insensitive (•) and 293B-sensitive (□) tail current (putative i_K,s tail current) in 18 and 14 cells, respectively. The E-4031-insensitive current was recorded during 5 min of exposure to 3 μm E-4031. The 293B-sensitive current was obtained by subtracting the current obtained during 5 min of exposure to 10 μm 293B from that obtained under control conditions.

Effect of E-4031 on spontaneous activity

Previously, the effects of a partial block of i_K,r by 0.1 μm E-4031 on rabbit SA node cells have been reported to be variable - spontaneous activity was blocked in some cells, but not others (Verheijck et al. 1995). In small balls of rabbit SA node tissue, we have shown that a partial block of i_K,r by 0.1 μm E-4031 abolishes spontaneous activity in tissue from the centre, but not from the periphery (Kodama et al. 1999). In the present study, we investigated the effect of partial blockade of i_K,r by 0.1 μm E-4031 (applied for 30 s) in small and large SA node cells: five small cells (C_m, 24 ± 1 pF, range 20-28 pF) and six large cells (C_m, 46 ± 3 pF, 39-54 pF). Figure 5A shows the effect of 0.1 μm E-4031 on the spontaneous activity of a small cell (C_m, 28 pF) - the membrane potential under control conditions and after 30 s exposure to 0.1 μm E-4031 is shown. In the presence of 0.1 μm E-4031, spontaneous activity ceased. After wash-off of E-4031 for 5 min, the spontaneous activity resumed (data not shown). The same effect was observed in the other small cells. Figure 5B shows the effect of 0.1 μm E-4031 on the spontaneous activity of a large cell (C_m, 48 pF) - the membrane potential under control conditions and after 30 s exposure to 0.1 μm E-4031 is again shown. In the presence of E-4031, spontaneous activity did not cease in either this cell or the five other large cells studied. The results are summarised in Table 2, which shows that in the six large cells, E-4031 significantly reduced the action potential amplitude, significantly prolonged the action potential, significantly decreased the maximum diastolic potential and significantly prolonged the cycle length (P < 0.05 in each case). The effects were reversible after wash-off of 0.1 μm E-4031 for 5 min (data not shown). For all cells studied, there was a significant correlation between the percentage slowing in firing rate and C_m (data not shown; r² = 0.79, P < 0.01, n = 11).

Figure 5. Effect of a 30 s exposure to 0.1 μm E-4031 on the spontaneous activity of small (A) and large (B) sinoatrial node cells.

The C_m of the cells is given in the figure. The membrane potential is shown under control conditions (○) and in the presence of E-4031 (•).

Table 2.

Effect of 0.1 μm E-4031 on action potential parameters of large SA node cells

	APA (mV)	APD100 (ms)	MDP (mV)	DDR (mV s−1)	Cycle length (ms)
Control	84 ± 6	140 ± 11	−66 ± 4	179 ± 22	250 ± 31
E-4031	68 ± 6*	187 ± 7*	−56 ± 6*	120 ± 15*	309 ± 24*

C_m, 46 ± 3 pF, range 39–54 pF, n = 6. APA, action potential amplitude; APD₁₀₀, action potential duration at 100% repolarisation; MDP, maximum diastolic potential; DDR, diastolic depolarisation rate

^*P < 0.05 vs. Control.

DISCUSSION

In the present study, the delayed-rectifier K⁺ current, i_K, was separated pharmacologically into two distinct components, a rapidly activating component (i_K,r) and a slowly activating component (i_K,s), through their sensitivities to E-4031 and 293B. The features of the E-4031-sensitive and 293B-insensitive currents corresponded well to those of i_K,r, while the features of the E-4031-insensitive and 293B-sensitive currents corresponded well to those of i_K,s. The densities of i_K,total and E-4031-sensitive, E-4031-insensitive, 293B-sensitive and 293B-insensitive currents were correlated with the size of cells as measured by their C_m. The current densities were greater in larger cells. The effects of partial blockade of i_K,r by 0.1 μm E-4031 on spontaneous activity were greater in smaller cells. These findings suggest that there are cell size-dependent differences in the densities of i_K,total, i_K,r and i_K,s; these differences could play a role in the regional differences in electrical activity within the SA node.

Selectivity of E-4031 and 293B

In the present study, 0.1 and 3 μm E-4031 were used. Verheijck et al. (1995) reported that in rabbit SA node cells, 0.1 μm E-4031 caused a partial blockade of i_K,r (∼78 % block) and 1 μm E-4031 blocked ∼96 % of i_K,r. In the study of Ito & Ono (1995) on rabbit SA node cells, 0.27 μm E-4031 caused half-maximal inhibition of i_K,r. In the study of Verheijck et al. (1995), 1 μm E-4031 had no effect on i_Ca and 10 μm E-4031 had no effect on i_f. In the study of Ito & Ono (1995), it was concluded that 3 μm E-4031 only blocked i_K,r and had no effect on i_Ca (Ca²⁺ current) and i_f. In guinea-pig ventricular cells, i_Ca has also been reported to be insensitive to 5 μm E-4031 (Sanguinetti & Jurkiewicz, 1990). In conclusion, all of the available evidence suggests that at the concentrations used, E-4031 is a selective blocker of i_K,r, although it remains a possibility that it has other, as yet unknown actions. E-4031 at a concentration of 0.1 μm is likely to block ∼27-78 % of i_K,r, whereas 3 μm E-4031 is likely to block ∼92-96 %.

In the present study, 10 μm 293B was used. The chromanol derivative 293B was first shown to inhibit a cAMP-regulated K⁺ channel in rat colon crypts (Lohrmann et al. 1995). A subsequent study showed that the I_sK channel may underlie the cAMP-regulated K⁺ channel and that 293B blocks the rat I_sK channel expressed in Xenopus oocytes (Suessbrich et al. 1996). I_sK is also known to be a i_K,s channel subunit. In the same study, 293B did not block the delayed- and inward-rectifying K⁺ channels, Kv1.1 and Kir2.1, respectively (Suessbrich et al. 1996). Busch et al. (1996) further investigated the specificity of 293B and found that it inhibited i_K,s in guinea-pig ventricular cells and the guinea-pig I_sK channel expressed in Xenopus oocytes with a similar IC₅₀ (2-6 μm). In contrast, 30 μm 293B had only a negligible effect on i_K,r in guinea-pig ventricular cells (Busch et al. 1996). Similarly, 30 μm 293B had no significant effect on the HERG channel (i_K,r channel) expressed in Xenopus oocytes (Busch et al. 1996). 293B at a concentration of 30 μm had no effect on i_Ca in guinea-pig ventricular cells or the cardiac i_Ca channel expressed in Xenopus oocytes (Busch et al. 1996). In conclusion, the available evidence suggests that at the concentration used, 293B is a selective blocker of i_K,s, although it remains a possibility that it has other, as yet unknown actions. At a concentration of 10 μm, 293B is likely to block ∼83 % of i_K,s (Busch et al. 1996).

isolation of i_K,r and i_K,s by E-4031 and 293B

Shibasaki (1987) was the first to record an i_K,r-like current in rabbit SA node (characterised by Ito & Ono, 1995, and Verheijck et al. 1995), and subsequently Lei & Brown (1996) and Sato et al. (1998) concluded that i_K,s is also present. Table 1 shows the V_1/2 and k for activation of i_K,r and i_K,s from some of these studies. In the present study, in terms of V_1/2 and k values, the absence or presence of inward rectification, and time dependence (most noticeable at positive potentials), the E-4031-sensitive/293B-insensitive currents and E-4031-insensitive/293B-sensitive currents are consistent with those of i_K,r and i_K,s, respectively (Table 1). In addition, the inward rectification displayed by i_K,r (E-4031-sensitive or 293B-insensitive current) was variable and, in general, weak. Cheng et al. (1999) also observed, in rabbit ventricular cells, a weak and variable inward rectification of i_K,r. Although the degree of inward rectification of i_K,r observed in the present study is similar to that observed by Shibasaki (1987) in rabbit SA node cells, it is less than that observed by Ito & Ono (1995), also in rabbit SA node cells. Two reasons for the difference between the present study and that of Ito & Ono (1995) are that in at least one experiment, Ito & Ono (1995) recorded i_K,r in a Ca²⁺-free, Mg²⁺-rich solution (inward rectification is dependent upon extracellular Ca²⁺ and Mg²⁺; Ho et al. 1998), and the extent of inward rectification is dependent upon phosphorylation (Heath & Terrar, 2000). In addition, the extent of inward rectification of i_K,r (in guinea-pig ventricular cells at least) is less when using the perforated patch-clamp technique (as used in the present study) than when using the conventional whole-cell patch-clamp technique (as used by Ito & Ono, 1995; B. M. Heath, personal communication). Cheng et al. (1999) reported that the kinetics of activation of i_K,s in rabbit ventricular cells were about 3 times faster than the activation of i_K,s in guinea-pig ventricular cells. Although the kinetics of i_K,s were not measured in the present study, they were comparable to those of i_K,s in rabbit ventricular cells.

heterogeneity of i_K,total, i_K,r and i_K,s

The present study has shown that the densities of: (1) the outward current during a pulse under control conditions, (2) i_K,total (tail current), (3) the E-4031-sensitive current (pulse and tail currents), (4) the E-4031-insensitive current (pulse and tail currents), (5) the 293B-sensitive current (pulse and tail currents), and (6) the 293B-insensitive current (pulse and tail currents) are significantly correlated with C_m such that they are greater in cells with a greater C_m. We conclude that the densities of i_K,r and i_K,s (and thus i_K,total) are correlated with C_m and are greater in larger cells. From this we also conclude (but see below) that the densities of i_K,r and i_K,s (and thus i_K,total) may be lower in the centre of the SA node than in the periphery (based on the assumption that cells are smaller in the centre than in the periphery). Several studies have shown important regional differences in i_K,total, i_K,r and i_K,s in the heart. In the canine ventricle, the density of i_K,total in sub-epicardial cells has been reported to be greater than that in mid-myocardial cells and similar to that in sub-endocardial cells; this was the result of a lower density of i_K,s in mid-myocardial cells - the density of i_K,r was similar in the three regions (Gintant, 1995; Liu & Antzelevitch, 1995). In the cat ventricle, the density of i_K,total was also greater in sub-epicardial cells than in sub-endocardial cells (Furukawa et al. 1992). In the guinea-pig ventricle, Bryant et al. (1998) showed that the density of i_K,total was greater in mid-myocardial and sub-epicardial cells than in sub-endocardial cells. Both i_K,r and i_K,s were significantly greater in mid-myocardial or sub-epicardial cells than in sub-endocardial cells (Bryant et al. 1998). In the rabbit ventricle, the density of i_K,s was greater in basal cells than apical cells, whereas the density of i_K,r was greater in apical cells than in basal cells (Cheng et al. 1999). In the rabbit, Sato et al. (1998) reported that i_K,r and i_K,s were present in SA node cells, whereas i_K,r currents were only present in atrioventricular node cells. It has been suggested that the regional differences in i_K and its components may underlie regional differences in action potential duration and regional differences in the effects of class III antiarrhythmic agents (Gintant, 1995; Bryant et al. 1998; Cheng et al. 1999).

Although we favour the possibility that the cell size-dependent differences may reflect regional differences, we cannot rule out the possibility that the cell size-dependent differences are the result of another factor. For example, it is possible that cells of different sizes are routinely adversely affected by the cell isolation procedure to different extents.

physiological significance of regional differences in the densities of i_K,r and i_K,s in the SA node

Verheijck et al. (1995) reported that in rabbit SA node cells, the partial blockade of i_K,r by 0.1 μm E-4031 arrested half of the cells, and in the other half it decreased the upstroke velocity of the action potential, increased the action potential duration, decreased the maximum diastolic potential, decreased the diastolic depolarisation rate and prolonged the cycle length (cf. Table 2). At higher concentrations of E-4031 (1 and 10 μm), spontaneous activity ceased (Verheijck et al. 1995). Comparable results were obtained by Ono & Ito (1995) and Lei & Brown (1996) from rabbit SA node cells. These investigations clearly show that i_K,r plays an important role in the pacemaker activity of the SA node. However, the possibility that there is a regional difference in the contribution of i_K,r to pacemaker activity in the SA node was not considered in these studies. In our present study, as well as in our previous study (Kodama et al. 1999), we have observed similar effects of E-4031, but we have also obtained evidence that the effects of E-4031 vary regionally. In a study of small balls of tissue from the centre and periphery of the SA node, partial block of i_K,r by 0.1 μm E-4031 arrested tissue from the centre of the SA node (Kodama et al. 1999). However, spontaneous activity in tissue from the periphery persisted, although the electrical activity was affected as above (Kodama et al. 1999). A higher concentration of E-4031 (1 μm) arrested the spontaneous activity in all of the tissue (Kodama et al. 1999). In the present study, partial blockade of i_K,r by 0.1 μm E-4031 arrested small cells, whereas spontaneous activity in large cells persisted, although electrical activity was affected as described above (Table 2). It is possible that the regional differences in the effects of E-4031 are the result of the presumed regional differences in the density of i_K,r: we have recently constructed mathematical models of the action potential in the centre and periphery of the rabbit SA node (Zhang et al. 2000). With the density of i_K,r lower in the central cell model than in the peripheral cell model, we were able to simulate the regional differences in the effects of E-4031 (Zhang et al. 2000). In the models, as in the experiments, complete blockade of i_K,r arrested spontaneous activity, whereas partial (50 %) blockade of i_K,r still arrested spontaneous activity in the central cell model, but not in the peripheral cell model (although electrical activity was modified as in the experiments; Zhang et al. 2000). The fact that complete blockade of i_K,r arrested spontaneous activity in both the experiments and the mathematical modelling demonstrates that i_K,r is essential, directly and indirectly, for spontaneous activity. In the peripheral cell model, a 50 % block of i_K,r did not arrest spontaneous activity because the remaining 50 % of i_K,r was still sufficient to sustain spontaneous activity, whereas in the central cell model, because the density of i_K,r was 6.2 times lower, after blockade the remaining 50 % of i_K,r was not sufficient to maintain the spontaneous activity (Zhang et al. 2000).

It is possible that a reduction in the density of i_K,r in the centre of the SA node explains why the maximum diastolic potential is more positive and, in part at least, why the action potential is longer in the centre of the SA node than in the periphery. Consistent with this, in our study of small balls of tissue from the centre and periphery of the SA node, after a partial block of i_K,r the electrical activity of the peripheral tissue was more like that of the central tissue under normal conditions (Kodama et al. 1999). We have suggested that the increase in action potential duration from the periphery to the centre of the SA node is a protective mechanism to help prevent re-entrant arrhythmias (Boyett et al. 1999).

It is interesting that in the rabbit there is a significant increase in the density of a number of currents from small to large SA node cells: the i_Na (Honjo et al. 1996), 4-AP-sensitive transient outward current (i_to; Lei et al. 2000), 4-AP-sensitive sustained outward current (Honjo et al. 1996; Lei et al. 2000), i_K,r (this study), i_K,s (this study) and i_f (Honjo et al. 1996). It is possible that the density of most currents may change with cell size in a similar manner. The cause of the putative regional differences in current densities is not known, but it is possible that they result from regional differences in the transcription of ion channel genes. Regional differences in current densities may not be restricted to the rabbit: Furukawa et al. (1999) observed a shift in the pacemaker in the dog SA node in response to various current blockers including E-4031 (such shifts are also seen in the rabbit - Boyett et al. 2000) and concluded that the roles of various currents, including i_K,r, vary regionally. The functioning of the intact SA node, as well as being influenced by regional differences in the density of the channels responsible for the currents noted above, is also expected to be influenced by known regional differences in gap junction channels (e.g. Kwong et al. 1998; Coppen et al. 1999).