Effects of acidosis and NO on nicorandil-activated K_ATP channels in guinea-pig ventricular myocytes

Abstract

1. Nicorandil is a hybrid compound of K⁺ channel opener and nitrate. We investigated a possible interaction of acidosis and nitric oxide (NO)-donors on the nicorandil-activated ATP-sensitive K⁺ channel (K_ATP) in guinea-pig ventricular myocytes using the patch-clamp technique.

2. In whole-cell recordings, external application of 300 μM nicorandil activated K_ATP in the presence of 2 mM intracellular ATP concentration ([ATP]_i) at external pH (pH_o) 7.4, but the activated current was decreased by reducing pH_o to 6.5–6.0.

3. Single-channel recordings of inside-out patches revealed decreased open-state probability (P_o) of K_ATP activated by nicorandil with reducing internal pH (pH_i) from 7.2 to 6.0, whilst the channel activity increased at low pH_i in the absence of nicorandil.

4. Application of NO donors, 1 mM-sodium nitroprusside (SNP) or -NOR-3 to the membrane cytoplasmic side at pH_i 7.2 increased the channel activity but decreased it at pH_i 6.5–6.0. Neither removal of the drugs nor application of NO-scavengers reversed depression of channel activity induced by NO-donors.

5. We conclude that an increase in pH_o and pH_i depresses rather than stimulates the nicorandil-activated K_ATP. Since NO-donors at low pH_i exhibited a similar trend, involvement of H⁺ and NO interaction can be considered as a mechanism of decreased K_ATP activated by nicorandil.

Introduction

Nicorandil is a potassium channel opener that targets the ATP-sensitive potassium channel (K_ATP) (Hiraoka & Fan, 1989; Nakayama et al., 1991; Takano & Noma, 1990). Based on its chemical structure nicorandil can be considered as a hybrid compound having nitrate and K⁺ channel opening properties (Taira, 1987). The drug is clinically available by virtue of its ability to dilate the coronary vessels and increase the coronary flow reserve in patients with chronic ischaemic heart disease (Frampton et al., 1992; Goldschimidt et al., 1996; Krumenacker & Roland, 1992). Recently much attention has been paid to its effects mimicking ischaemic preconditioning (Matsubara et al., 2000; Patel et al., 1999). Whilst the potency for nicorandil to activate K_ATP is not strong compared to other agents, the drug has a unique action requiring the presence of MgADP for the channel activation (Shen et al., 1991), whilst other studies demonstrated the channel activation by nicorandil in the absence of ADP at the single channel level (Takano & Noma, 1990). It has been further suggested that nicorandil could activate K_ATP in the absence of ADP in external acidotic condition suggesting more effective action at low pH_o than at normal pH_o (Jahangir et al., 1994). Other studies indicated that K_ATP activated by another KCO, pinacidil, was suppressed at low pH_o and increased at high pH_o, suggesting an external site of interaction for the channel and H⁺ (Kwok & Kass, 1994).

During myocardial ischaemia important changes in external and internal pH as well as changes in the ratio of intracellular ATP and ADP concentrations take place (Allen et al., 1985). The modulation of K_ATP has been studied to demonstrate that moderate increase in H⁺ concentrations, acidosis, itself has the ability to increase the activity of K_ATP (Cuevas et al., 1991; Davies, 1990; Fan & Makielski, 1993; Koyano et al., 1993). Therefore, nicorandil may exert its K⁺ channel opening potency more strongly during ischaemia or acidotic conditions than in normoxic state and normal pH. However, it is not known how the other aspect of nicorandil as nitrate interacts with the K_ATP activation. The present study was undertaken to examine a possible interaction between low pH and NO on the nicorandil-activated K_ATP activity.

Methods

The investigation was conducted in accordance to the guidelines for the care and use of laboratory animals at Tokyo Medical and Dental University.

Cell isolation

Single ventricular myocytes were isolated enzymatically from the ventricles of female guinea-pigs weighing 250–350 g. The technique for cell isolation used in our laboratory has been described previously (Hirano & Hiraoka, 1988). In brief, guinea-pigs were anaesthetized with Na-pentobarbital (30 mg Kg⁻¹, IP) and right after injected intravenously with heparin (300 units Kg⁻¹). The chest was opened under artificial respiration (Respirator Model 141, NEMI Corp, Medwey, MA, U.S.A.) and the aorta was canulated to perfuse retrogradely with the Tyrode solution before the heart was dissected out. After 6 min of perfusion on the Langendorff apparatus, the heart was perfused with a nominally Ca²⁺ free Tyrode solution for an additional 6 min. The perfusate was switched to Ca²⁺ free Tyrode solution containing collagenase (5 mg 50 ml⁻¹, Yakult, Tokyo, Japan). After 5–6 min perfusion, the heart was washed out with a high K⁺, low Cl⁻ solution. The temperature of the perfusates was maintained at 35–36°C. The isolated cells were obtained through a mesh (size 200 μM). Rod-shaped cells with clear margin and striation were used for the experiments.

Solutions

In the whole-cell configuration the bath solution was Tyrode solution, and its composition was (mM): NaCl 144, NaH₂PO₄ 0.33, KCl 4.0, CaCl₂ 1.8, MgCl₂ 0.53, glucose 5.5, and HEPES 5.0; the pH was adjusted to 7.3–7.4 by addition of NaOH. The Ca²⁺-free Tyrode solution was prepared by omitting CaCl₂ from the Tyrode solution. High K⁺ low Cl⁻ solution contained (mM): KOH 80, glutamic acid 70, taurin 15, KH₂PO₄ 10, HEPES 5, MgCl₂ 0.5, glucose 11, and K₂-ethyleneglycol-O-O′-bis(B-aminoethyl)-N,N,N′,N′-tetraacetic acid (EGTA) 0.5 (pH 7.4-KOH). The composition of the pipette solution was (mM): KCl 120, K₂ATP (Sigma Chemical Co., St. Louis, MO, U.S.A.) 2.0, HEPES 5.0 and K₄BAPTA (Dojin Co., Kumamoto, Japan) 5.0; the pH was adjusted to 7.2 with KOH. The final K⁺ concentration was kept constant at 150 mM.

In the case of single-channel recordings (inside-out patch configuration), the bath solution (intracellular medium) contained (mM): KCl 140, glucose 5.5, EGTA 2 and HEPES (or PIPES) 5; the pH was adjusted to 7.3 with KOH. When experiments were conducted in acidotic conditions, bath solutions having different pH values (7.2, 6.5 or 6.0) were adjusted with HEPES or PIPES buffer systems, accordingly. Final pH adjustment was done with 0.1 N HCl. When nucleotides were present in the bath, MgCl₂ 0.6–0.8 mM was added to the bath. The drugs were dissolved in the bath solution at the concentration indicated in the text. The pipette solution (extracellular medium) contained (mM): KCl 140, CaCl₂ 1.8, MgCl₂ 0.53, glucose 5.5 and HEPES 5; the pH was adjusted to 7.3 by adding KOH.

Electrophysiological measurements

Whole-cell current recordings

Membrane currents were recorded using the patch-clamp technique of whole cell configuration (Hamill et al., 1981), using a patch-clamp amplifier (Axopatch ID, Axon Instrument, Foster City, CA, U.S.A.). Glass patch electrodes were made from borosilicate capillary tubes with an outer diameter of 1.5 mm (Clark Electromedical Instruments, Pangbourne, England) using a microelectrode puller (Model PP-830, Narishige Co. Tokyo, Japan) and were heat-polished by a microforge (Model MF-830, Narishige Co., Tokyo, Japan). The electrode resistance was 2–4 MΩ when the pipettes were filled with an internal solution. The recording technique and the data acquisition systems have been described in previous reports (Hirano & Hiraoka, 1988). When the ramp voltage-clamp method was employed, an intelligent arbitrary function synthesizer (model 1731, NF Instruments, Yokohama, Japan) was used to supply the command pulse.

The temperature of the bath chamber was maintained at 34–35°C with a heating system (DTC-100TA, Dia Medical System, Tokyo, Japan). Before establishing contact between the electrode and the cell membrane, the junction potential was adjusted to zero at the level of the bath solution. At the end of each experiment the junction potential was verified again and, adjusted if a difference of more than ±2 mV existed between the first and second measurements. A time-interval of 2–3 min for sequential recording at different pH was established.

Single-channel recordings

Single-channel current recordings were carried-out at room temperature with conventional inside-out patch configuration (Hamill et al., 1981) using the same patch-clamp amplifier as in the whole-cell experiments. The current signals were recorded and stored simultaneously on a videocassette recorder (HR-S7700, Victor, Tokyo, Japan) and a thermal recorder (Omnicorder 8M14-3, NEC-Sanei Instruments, Ltd., Tokyo, Japan) through a PCM data recorder system (RP-882, NF Instruments, Yokohama, Japan) at a conversion rate of 40 kHz. Recorded signals were filtered off-line through a programmable eight-pole Bessel low-pass filter (48 dB/octave, 3625, NF Instruments, Yokohama, Japan) and digitized at a sampling frequency of 10 kHz and stored into a MO disk of a computer (Physio PC-01; Physio-tech, Tokyo, Japan) using an analogue-to-digital converter (Digidata 1200 Interface, Axon Instruments, Foster City, CA, U.S.A.) for later analysis. pCLAMP software (version 6.0.4, Axon Instruments) was used to generate voltage pulse protocols, data acquisition and analysis.

Single channel data analysis

Single channel records were analysed by pCLAMP 6.0 software program on a computer (Physio PC-01; Physio-tech, Tokyo, Japan). The unitary current amplitude of K_ATP was measured from the all point histograms through Gaussian fitting, or by selecting well-defined long opening and close transitions and measuring the magnitude of the corresponding current steps with horizontal cursors.

Mean patch current (I) was obtained over 30 s as a time averaged currents of K_ATP, measured as the difference between the baseline (a current level where all channels are in a close state) and the current level where the channels were in the open state with a half height criteria. Data points over the signals were delimited by cursors. The number of active channels in the patch (N), while initially perfusing nucleotide-free solution was regarded as the maximum open state of activity for later comparisons to assess the effect of nucleotides and/or drugs upon the active channels in the membrane-patch. Since the unitary current (i) changes at different pH levels, the I/i quotient was used as a proportional indicator of channel activity. Since K_ATP could be subject to run-down phenomenon, during each experiment, maximum channel activity was always verified regularly and measurements and calculations were corrected accordingly assuming a linear decrease in the number of active channels.

Drugs

All drugs were freshly prepared before every experiment and diluted into the test solution to obtain the final concentration as indicated in the text. Nicorandil (2-nicotinamidoethyl nitrate; a gift from Chugai Pharmaceutical Co., Tokyo, Japan) was diluted into the test solution. Glibenclamide (a gift from Hoescht Japan, Tokyo) was dissolved in 2% dimethyl sulphoxide (DMSO) and diluted into the test solution. The final concentration of DMSO contained in the test solution was less than 0.01%. We used two NO-donors: Sodium nitroprusside (SNP, Sigma Chemical Co. St. Louis MO, U.S.A.) and (±)-(E)-ethyl-2-[(E)-hydroxyimino]-5-nitro-3-hexeneamide (NOR-3, a gift from Fujisawa Pharmaceutical Co. Ltd. Osaka, Japan). They were directly dissolved into the perfusing solution at the concentration indicated in the text. Exposure to light and oxygen was minimized. NO-scavengers: 5–10 μM oxyhaemoglobin (HbO₂, Calzyme Laboratories, Inc. CA, U.S.A.) and 2–3 mM C₁₄H₁₆N₂NaO₄ (Carboxy-PTIO; Dojindo Kumamoto, Japan) were both directly dissolved into the perfusate to a final concentration as indicated in the text.

Statistical analysis

Data are expressed as mean±s.d. The significance for pH-effect as the only changing factor when comparing more than two groups was assessed by analysis of variances (ANOVA). Comparisons between two groups of data were evaluated by paired or unpaired student t-test, accordingly. A value of P<0.05 was considered significant.

Results

Reduction of external pH affects nicorandil-activated K_ATP current

Application of 300 μM nicorandil activated outward current at potentials positive to ∼60 mV elicited by a ramp voltage clamp at pH_o 7.4. The nicorandil-activated current was completely inhibited by 1 μM glibenclamide confirming the current as K_ATP (n=4; not shown). When pH_o was lowered from 7.4 to 6.8–6.0 in the presence of 2 mM [ATP]_i, the nicorandil-activated K_ATP current was somewhat decreased at pH_o 6.8 and further suppression was noted at pH_o 6.0 (Figure 1a). Figure 1b presents a summary of the results obtained from 10 different myocytes. The current at 0 mV in the control was 1.22±0.6 nA and it was increased to 3.96±1.97 nA in the presence of 300 μM nicorandil at pH_o 7.4 (P<0.01). The nicorandil-activated current was decreased to 1.96±0.95 nA at pH_o 6.8 in the presence of the drug (P<0.05 versus nicorandil at pH_o 7.4) and further decrease of the current was noted at pH_o 6.0 (1.06±0.68 nA; P<0.01 versus nicorandil at pH_o 7.4). As we could not detect any increase in the nicorandil-activated K_ATP current by the whole-cell configuration, we proceeded to examine the modulatory action of pH_i on the single K_ATP currents with inside-out patch configuration.

Figure 1.

Suppression of nicorandil-activated whole-cell current by lowering pH_o. (a) Superimposed traces of background I-V curves obtained from a typical case in a single myocyte, at 3 min intervals. ○: Control at pH_o 7.4. •: S6.8Addition of 300 μM nicorandil to the bath at pH_o 7.4 (Nicorandil, pH_o 7.4). ▴: Reducing pH_o to 6.8 in the presence of nicorandil (Nicorandil, pH_o 6.8). ▪: Further reduction of pH_o to 6.0 in the presence of nicorandil (Nicorandil, pH_o 6.0). (b) Summarized data from 10 myocytes revealed a significant decrease of the nicorandil-activated K⁺ current by reducing pH_o from 7.2 to 6.8 and 6.0. Each symbol corresponds to the value obtained from individual preparations.

Increased K_ATP activity by internal acidosis

First, we tested the low pH_i-induced modulation of the K_ATP activity in the absence of nicorandil under our experimental conditions, since reducing pH_i from ∼7.4 to 6.5–6.0 was shown to increase the channel activity (Cuevas et al., 1991; Davies, 1990; Fan & Makielski, 1993; Koyano et al., 1993; Vivaudou & Forestier, 1995). Figure 2a presents the effect of lowering pH_i on K_ATP current. When the internal face of the patch-membrane was exposed to the ATP-free solution at pH 7.2 many channels were kept in the open state, and the channels were quickly and largely inhibited by addition of 0.35 mM ATP_i. The open-state probability (P_o) of the channels was 89±10% in the ATP-free solution and it was 7±5% at 0.35 mM [ATP]_i (P<0.01). A significant increase of channel activity became apparent by decreasing pH_i from 7.2 to 6.0 (P_o; 7±5% versus 32±13%, respectively, n=9; P<0.01) in the presence of Mg²⁺ as previously reported. Run-down was excluded by observing near complete recovery of the initial Po upon removal of ATP from the internal solution (Figure 2b). As reported previously reduction of pH_i slightly decreased the unit amplitude of single-channel current (i) (data not shown).

Figure 2.

Increased K_ATP activity by lowering pH at the cytoplasmic side of the membrane (pH_i). (a) Single-channel record obtained with inside-out patch configuration. The membrane potential was held at −40 mV, while having a 140 mM symmetrical K⁺ concentrations at both sides of the membrane. Control K_ATP current was recorded in the absence of nucleotides at pH 7.2 in the bath (pH_i). The control current was of large amplitude and off-scaled. Addition of nucleotides, 0.35 mM [Mg.ATP]_i+0.12 mM [K.ADP]_i whilst keeping the same pH_i markedly suppressed the current. Mg²⁺ was also present at a 0.6–0.8 mM concentration. Whilst lowering pH_i to 6.5 at this [ATP]_i did not exhibit any change in channel activity, further reduction of pH_i to 6.0 increased significantly the P_o of the channels. Application of 4.0-μM glibenclamide suppressed the current. C represents the closed level of the channel. Downward deflection indicates inward currents in this and the following figures except Figure 3b. (b) The bar graph showing summarized data obtained from nine different patches under the same experimental condition as in (a).

Effect of acidosis on the nicorandil-activated K_ATP current

Application of 1 mM nicorandil to the internal solution in the presence of 0.25 mM ATP, 0.1 mM ADP and Mg²⁺ at pH_i 7.2, K_ATP was activated as similar as the removal of ATP. The nicorandil-activated K_ATP currents were decreased with reduction of pH_i to 6.5 and 6.0 (Figure 3a). Return of pH_i to 7.2 or removal of [ATP]_i restored the channel activity. Figure 3b represents the amplitude histograms of the records shown in Figure 3a. Reduction of pH_i decreased the channel activity. Figure 4 demonstrates the summary data from 10 similar experiments as shown in Figure 3. P_o of nicorandil-activated K_ATP at pH_i 7.2 was 84±16%. Decreasing pH_i from 7.2 to 6.5 and 6.0, resulted in a significant reduction of K_ATP activity (P_o; 57±20% at pH_i 6.5; P<0.01 and 38±12% at pH_i 6.0; P<0.01 when compared to pH_i 7.2, respectively). Upon return to pH_i 7.2, an almost complete recovery to the initial level of channel activity was restored (P_o=80±5%).

Figure 3.

Effects of low pH_i on nicorandil-activated K_ATP current. (a) Single channel current record with holding potential at −40 mV. The bath contained 0.25 mM Mg.ATP, 0.12 mM K.ADP, 0.8 mM Mg²⁺ and 1.0 mM nicorandil. Lowering pHi from 7.2 to 6.5 and 6.0 in the presence of nicorandil reduced the nicorandil-activated K_ATP current. These effects were reversible upon return of pH_i to 7.2. (b) Current histograms of 30 s-analysed segments taken from (a). The letters in brackets indicate the corresponding segment in (a).

Figure 4.

Summarized data for effects of low pH_i on nicorandil-activated K_ATP current. The bar graphs summarize data from 10 different patches. Reducing pH_i to 6.5 and 6.0 significantly decreased the channels-activity induced by 1.0 mM nicorandil compared at pH_i 7.2. Upon washout of nicorandil and returning to ATP- and ADP-free pH_i 7.2 solution, near complete recovery of the initial activity was achieved in all patches excluding the rundown of the channel. Experimental conditions same as those in Figure 3.

Effects of Nitric Oxide on K_ATP

Since nicorandil is a hybrid compound including NO in its chemical structure, we investigated a possibility that the reduced activity of the nicorandil-activated K_ATP at low pH_i might be caused by an interaction between H⁺ and NO. For this purpose, we examined effects of NO donors on K_ATP. We used two different NO-donors: SNP and NOR-3 (see Methods). Figure 5a shows the effect of 1 mM SNP in the presence of 0.25 mM [ATP]_i at pH_i 7.2. K_ATP activity increased significantly in the presence of the NO-donor. Figure 5b summarizes data from 11 patches. P_o was 17±11% in the absence versus 32±12% in the presence of SNP 5 minutes (P<0.05). Upon removal of the NO-donor, an increase in K_ATP current was kept ongoing to reach P_o values up to 73±38% following 10 min of washout (P<0.05). On the contrary, lowering pH_i from 7.2 to 6.0 at constant ATP and SNP concentration caused a significant reduction in K_ATP activity (P_o; 54±34% at pH_i 7.2 versus 27±19% at pH_i 6.0; P<0.01; n=11) (Figure 6a,b). Using NOR-3, a different NO donor, we observed similar tendency as with SNP. In the presence of 0.25 mM [ATP]_i at pH_i 6.0, the P_o was 28±15% in the absence, whilst in the presence of 1 mM NOR-3, P_o decreased to 8±6% (P<0.05; n=5). Washout of the drug did not reverse the current depression, similarly as with the application of SNP. On the other hand, we also observed that a switch in pH_i from 6.0–7.2 increased the activity of the channel (P_o; 20±16% at pH_i 6.0 versus 46±19% at pH_i 7.2; P<0.05; n=4) (Figure 6a). Contrary to what we observed at pH_i 7.2, removal of NO-donor at pH_i 6.0 did not get complete recovery of the channel activity to the previous level. This was not due to run-down of channel activity since the nucleotide-free solution at pH_i 7.2 restored complete recovery of the channel activity to the initial level.

Figure 5.

Increased K_ATP activity at normal pH_i induced by NO-donor. (a) Single channel current record obtained from a typical patch. Application of nucleotides and NO-donor, 1.0 mM SNP in the bath is indicated at the top, pH_i was kept constant at 7.2 throughout these experiments. In the absence of nucleotides a large current was observed and off-scaled in the record. Decreased channel activity upon addition of 0.25 mM Mg.ATP and 0.12 mM K.ADP was gradually increased by adding 1.0 mM SNP into the bath and continued after the removal of the NO-donor. Increasing [Mg.ATP]_i to 0.5 mM caused a similar P_o as the level observed at 0.25 mM [Mg.ATP]_i before the exposure to SNP, suggesting a reduction to the inhibitory effect of ATP. (b) The bar graphs of data from 11 patches. The channel activity expressed by P_o increased significantly by addition of SNP in the presence of nucleotides and washout of the drug increased P_o.

Figure 6.

Effects of SNP on K_ATP at low pH_i (a) Experimental conditions of pH_i, presence of nucleotides and SNP are indicated on the top. The channel activity at pH_i 6.0 in the presence of nucleotides was significantly reduced by the presence of 1.0 mM SNP and return of pH_i to 7.2 resulted in increase of channel P_o. (b) The bar graphs of summarized data from 11 patches. The pH_i at 6.0 and nucleotides concentrations were kept constant with 0.25 mM [Mg.ATP]_i, and 0.12 mM [K.ADP]_i throughout the experimental conditions. In all patches run-down was excluded by observing nearly complete recovery of activity when perfusing nucleotide-free, pH 7.2 solution after these treatments.

Effects of NO-donors and NO-scavengers on K_ATP

In the presence of NO-donors at low pH_i, we attempted to reverse their depressing effects on K_ATP by adding NO scavengers, either HbO₂ or carboxy-PTIO. However, 5–10 μM HbO₂ enhanced rather than reversed the depressing effects of NO donors (Figure 7). P_o was 60±5% in the presence of 1 mM NOR-3 alone versus 20±2% in the presence of HbO₂ (P<0.01; n=4). At low pH_i neither washout of HbO₂ nor removal of the NO donor altered the decreased P_o of the channels, but an increase of the pH_i to 7.2 could increase P_o. Run-down was excluded by observing complete recovery of channel activity upon removal of nucleotides at pH_i 7.2. Either at normal or low pH_i, the compound itself at concentrations ranging 5–10 μM showed a direct suppressing effect on the channel activity. This effect was more prominent at low pH_i revealing an additive action to that of the NO-donor. In two additional patches we tested the effect of HbO₂ on the nicorandil-activated K_ATP at normal pH_i and again we observed a clear reduction of activity, with its reversal upon washout of HbO₂. No significant changes in channel activity were observed by using carboxy-PTIO with a ratio 2 : 1 or 3 : 1 respect to the NO donor.

Figure 7.

Effects of NO scavenger, HbO₂ on K_ATP activity in the presence of NO-donor, NOR-3. Single channel records show that lowering pHi from 7.2 to 6.0 in the presence of nucleotides and Mg²⁺ increased K_ATP activity. Increased channel activity was somewhat depressed by addition of 1.0 mM NOR-3 and further decrease was noted in the presence of 5 μM of HbO₂. Neither removal of HbO₂ nor NOR-3 reversed the depressed activity of the channel, but a change in pH_i from 6.0–7.2 increased the P_o of K_ATP. Upon removal of the nucleotides complete recovery of the initial activity took place ruling out the presence of run-down.

Discussion

In the present study we examined two major issues: first, how an increase in external and internal H⁺ concentrations modulates the activity of nicorandil-activated K_ATP in guinea-pig ventricular myocytes. Second, the possibility that the modulation caused by acidosis in nicorandil-activated K_ATP may be related to H⁺ and NO interaction.

Effect of external pH on nicorandil-activated K_ATP current

Previous study demonstrated that lowering pH_o could enhance the ability for nicorandil to activate whole-cell K_ATP current in guinea-pig ventricular myocytes (Jahangir et al., 1994). The authors considered that the enhanced action of nicorandil during acidosis was not related to a different ionization state of the drug nor a change in the ATP/ADP ratio, but a direct effect of H⁺ modulating the interaction between the channel and the drug. Effectiveness of another KCO, pinacidil decreased at low pH_o and increased at high pH_o (Kwok & Kass, 1994). The findings were interpreted as H⁺ might modulate the binding site for pinacidil located at the extracellular side of the membrane. Thus, the interaction between H⁺ and the different types of KCO might be different depending on their chemical structures.

We tested a slightly different view from that of Jahangir et al. (1994) as to the nicorandil action on K_ATP. That is how the nicorandil-activated K_ATP could be modified by low pH_o, the condition prone to be associated with myocardial ischemia. It turned out that lowering pH_o decreased rather than increased the nicorandil-activated K_ATP in whole-cell currents. Since pH_o might lead to change in pH_i, and pH_i itself would have modifying action on K_ATP, we proceeded to explore the modulating mechanism by increased H⁺ concentrations with inside-out patches, where involvement of the second messenger system produced by the drug could be largely excluded.

Effect of internal pH on nicorandil-activated K_ATP

In the presence of MgATP and ADP in the bath (cytoplasmic face), lowering pH_i remarkably increased the P_o of the channels, the results consistent with previous reports (Cuevas et al., 1991; Davies, 1990; Fan & Makielski, 1993; Koyano et al., 1993; Vivaudou & Forestier, 1995). This action of low pH_i can be interpreted that an intrinsic proton binding site regulates ATP sensitivity for the channel inhibition in K_ATP (Fan & Makielski, 1993). Evidence of the importance of Mg²⁺ as a cofactor necessary to keep the channel in operative state in the presence of either ATP or nucleotide diphosphates has been provided as well as its contribution in the enhancement of KCO effect (Findley, 1988; Lederer & Nichols, 1989; Shen et al., 1991; Terzic et al., 1994; Tung & Kurachi, 1991; Vivaudou & Forestier, 1995). We used MgATP combined with small doses of ADP in all our subsequent experiments in order to keep the channel in operative state and minimize run-down. At physiologic pH_i 7.2 nicorandil markedly increased the P_o of K_ATP, but when lowering pH_i, the P_o of nicorandil-activated K_ATP was significantly decreased. Therefore, it appears that variations in pH_i directly affect the interaction between nicorandil and K_ATP rather than via producing the second messengers such as the production of cyclic GMP (Kojda & Kottenberg, 1999; Kubo et al., 1994; Murphy & Brayden, 1995). This suppressive action on K_ATP seems to overcome stimulating effect of low pH_i.

Modulation of K_ATP by nitric oxide

In the present study, we demonstrated at the single-channel level that NO donors of different chemical structures increased the P_o of K_ATP at pH_i 7.2. Furthermore, removal of NO donors did not abolish the increased activity but the activity was kept increasing, and higher doses of ATP were needed to inhibit the channels than those before the drug treatment, suggesting a reduction of sensitivity to the nucleotide-inhibitory effect. Although the activation of the channels by NO donors was not as large as with nicorandil, the increase in P_o was clear and significant. At low pH_i, however, the presence of NO donors decreased the P_o of the channels and removal of the NO donors did not reverse the phenomenon. Also, with raising pH_i from 6.0 to 7.2 in the presence of the NO donors, the increase of P_o was evident. Therefore, the effects of the NO donors at either pH_i 7.2 or pH_i 6.0 were very similar to those observed with nicorandil.

Recently, NO was reported to enhance the KCO-activated K_ATP current with undefined mechanism (Shinbo & Iijima, 1997). They demonstrated the enhancing action of NO with whole-cell and cell-attached patch recordings in guinea-pig ventricular myocytes at normal pH_o and pH_i. In their experiments the cyclic GMP pathway and/or metabolic inhibition did not appear to be responsible for NO effects. Their findings were similar to the present results, except they failed to demonstrate the enhancing action of NOR-3 on K_ATP in inside-out patches. Here, we were able to show a clear activation of the channels by application of two different NO donors to inside-out patches. At present, we could not find any proper reason to explain the discrepancy between the two studies, but a direct action of NO on the modulation of cardiac K_ATP might be supported. In cultured vascular smooth muscle cells, isosorbide dinitrate and atrial natriuretic factor were proven to modulate the gating of K_ATP and the cyclic GMP pathway was involved (Kubo et al., 1994). There are, however, several lines of evidence that direct effects of NO on channel modulation; for example, expressed cardiac L-type Ca²⁺ channels (Hu et al., 1997), calcium-regulated potassium channel in vascular smooth muscle cells (Bolotina et al., 1994) and potassium channels in colonic muscle cells (Koh et al., 1995). Therefore, it may be possible that NO could directly modulate K_ATP and our findings add new evidence for its role on the modulation of cardiac K_ATP.

Because of these similarity of actions between the NO donors and nicorandil, the suppression of the nicorandil-activated K_ATP at low pH_i can be attributed to a direct interaction of NO with H⁺ at ATP-binding inhibitory sites of the channels, where stimulating action of pH_i is supposed to interact on this site. This action may be different from the enhancing action of NO and nicorandil on K_ATP at normal pH_i, or some other mechanism to exhibit both stimulating and suppressive actions depending on pH may be involved. This issue is to be explored further. Dual modulations of cardiac K_ATP have been demonstrated in the actions of pinacidil (Fan et al., 1990), nucleotides diphosphates and Mg²⁺ (Findley, 1988; Lederer & Nichols, 1989; Tung & Kurachi, 1990; Terzic et al., 1994), and pH_i (Cuevas et at., 1991; Fan & Makielski, 1993; Koyano et al., 1993; Fan et al., 1993). The dualistic modulations by these factors develop in variable ways and different mechanisms seem to be involved. For examples, pinacidil exerts voltage-independent activation of the channels and voltage dependent inhibition from inside of the membrane. Nucleotide diphosphates cause channel inhibition without Mg²⁺ and stimulate the channels in the operative state in the presence of Mg²⁺. Mg²⁺ produces channel inactivation and run-down on the one hand, and on the other it promotes the channels into the operative states. Low pH_i promotes increased channel open probability but induces decreased single channel current amplitude.

The fact that NO-scavengers were not able to prevent or reverse the actions of NO-donors on the K_ATP is unclear and needs to be further investigated. In the case of HbO₂, the compound itself was shown to affect directly the channel at either normal or lower pH_i, and that explains the prominent enhancement of the depressive effect of NO-donors on the channel. Does the K_ATP have the ability to sense and respond to HbO₂ levels? As for carboxy-PTIO, the dose required to abolish the NO effect is much higher than that for HbO₂, (Donor: Scavenger ratio=3 : 1) and the release rate of NO is slower, perhaps that might explains why it did not affect NO-H⁺ interaction respect to K_ATP.

Possible pathophysiological implications

The openings of K_ATP play an pivotal role for action potential shortening and are partly responsible for extra-cellular K⁺ accumulation during early phase of myocardial ischemia and infarction, which may contribute to the development of arrhythmias (Venkatesh et al., 1992; Wilde & Janse, 1994). At the same time, pH_i may go down quickly. On the other hand, the openings of K_ATP may exert the protective action on the second severe ischemic insult after a brief period of ischemia (preconditioning) (Downey & Cohen, 1997; Gross, 1995; Grover, 1994). If the patients taking nicorandil as anti-anginal drug develop ischemic attack, a quick opening of K_ATP can be anticipated during early ischemic episode and yet, the opening effects are subsequently suppressed due to decreased pH_i, which may give a self-limiting factor for the prevention of further shortening of action potential duration and extracellular K⁺ accumulation. At the same time, K_ATP as the end effector of cardioprotection is also modulated by NO production (Bolli et al., 1997; Imagawa et al., 1999; Yao and Gross, 1993). Therefore, these interplay and final results in the presence of nicorandil must be evaluated with caution. Acidosis is known to be pro-arrhythmogenic by affecting multiple ionic currents including K_ATP (Orchard & Cingolani, 1994). Therefore, it might be necessary to further examine the pharmacological profile of these drugs in acidotic conditions.

Conclusions

The present study provides information concerning the effect of H⁺ at either side of the membrane in nicorandil-activated K_ATP. We present new evidence of a cyclic GMP-independent modulating effect of NO on K_ATP during acidosis. These findings may be relevant to the therapeutic action of these drugs, especially during ischemic or hypoxic conditions.

Abbreviations

[ATP]_i

intracellular ATP concentration

[H⁺]_o

external proton-concentration

[H⁺]_i

internal proton-concentration

K_ATP

ATP-sensitive K⁺ channel

KCO

K⁺ channel opener

pH_o

external pH

pH_i

internal pH

Po

open-state probability