Inhibitory Effects of Lidocaine and Mexiletine on Vasorelaxation Mediated by Adenosine Triphosphate–sensitive K⁺Channels and the Role of Kinases in the Porcine Coronary Artery

The institutional animal care and use committee (Wakayama University, Wakayama, Japan) approved this study. Adult pig hearts were obtained from a slaughterhouse immediately after death and put in ice-cold modified Krebs-Ringer’s bicarbonate solution (control solution, pH 7.4) of the following composition: 119 mm NaCl, 4.7 mm KCl, 2.5 mm CaCl², 1.17 mm MgSO⁴, 1.18 mm KH²PO⁴, 25 mm NaHCO³, and 11 mm glucose. The left descending coronary artery was dissected and cut into 2-mm-length rings. Endothelial cells were removed mechanically to avoid the modification mediated by endothelium-derived nitric oxide as well as endothelium-derived hyperpolarizing factor.

Each ring was connected to an isometric force transducer and suspended in an organ chamber filled with 10 ml control solution (37°C) bubbled with a 95% O²–5% CO²gas mixture. The artery was gradually stretched to the optimal point of its length–tension curve as determined by the contraction to a prostaglandin H²/thromboxane receptor agonist, U46619 (10⁻⁷m). Optimal tension was achieved at approximately 3.0 g. Several rings cut from same artery were studied in parallel. The endothelial removal was evaluated by the absence of relaxation induced by bradykinin (10⁻⁶m). During submaximal contraction in response to U46619 (10⁻⁷m), concentration–response curves to levcromakalim (10⁻⁸to 10⁻⁵m) were obtained in the absence or in the presence of glibenclamide, lidocaine, mexiletine, phorbol 12-myristate 13-acetate (PMA), calphostin C, Gö 6976, genistein and/or erbstatin A, which were added 15 min before the contraction to U46619. In some experiments, these compounds for pretreatment were used as combination. The vasorelaxation was expressed as a percentage of the maximal relaxation in response to papaverine (3 × 10⁻⁴m), which was added at the end of the experiments to produce the maximal relaxation (100%) of arteries.7 

Arterial rings were longitudinally cut and fixed on the bottom of an experimental chamber. The arteries were continuously perfused with control solution (37°C) bubbled with a 95% O²–5% CO²gas mixture. A glass microelectrode (tip resistance 40–80 MΩ) filled with 3 m KCl and held by a micromanipulator (Narishige, Tokyo, Japan), was inserted into a smooth muscle cell from the intimal side of the vessel.8,9The electrical signal was amplified using a recording amplifier (Electro 705; World Precision Instruments Inc., Sarasota, FL). The membrane potential was continuously monitored and recorded on a chart recorder (SS-250F-1; SENKONIC Inc., Tokyo, Japan). The validity of a successful impalement was assessed by a sudden negative shift followed by a stable negative voltage for more than 2 min.8,9As previous studies demonstrated, resting membrane potential of vascular smooth muscle cells obtained using this technique was approximately −40 mV.10Changes in membrane potentials produced by levcromakalim (10⁻⁵m) were continuously recorded. Glibenclamide, lidocaine, mexiletine, calphostin C, or erbstatin A was applied 15 min before membrane potential recordings.

Levcromakalim was a generous gift from GlaxoSmithKline plc (Greenford, United Kingdom), and calphostin C, dimethyl sulfoxide, erbstatin A, genistein, glibenclamide, Gö 6976, lidocaine, mexiletine, papaverine, PMA, and U46619 were purchased from Sigma (St. Louis, MO). Stock solutions of levcromakalim, glibenclamide, PMA, calphostin C, Gö 6976, genistein, and erbstatin A were prepared in dimethyl sulfoxide (3 × 10⁻⁴m), and other drugs were dissolved in distilled water.

Data are expressed as mean ± SD. Statistical analysis was performed using repeated-measures analysis of variance followed by Scheffé F test for multiple comparison. Differences were considered to be statistically significant when P  was less than 0.05.

During submaximal contraction to U46619 (10⁻⁷m), a selective K^ATPchannel opener, levcromakalim (10⁻⁸to 10⁻⁵m) induced concentration-dependent relaxation in the porcine coronary artery without endothelium (fig. 1). A selective K^ATPchannel antagonist, glibenclamide, completely abolished this vasorelaxation (fig. 1), whereas it did not affect the basal tone of the coronary artery. Maximal vasorelaxation induced by papaverine (3 × 10⁻⁴m) in each group in figure 1was 100% = 4.89 ± 1.28 or 4.24 ± 2.73 g for control rings and rings treated with glibenclamide, respectively (statistically insignificant).

Lidocaine and mexiletine (10⁻⁵to 10⁻⁴m) significantly reduced vasorelaxation in response to levcromakalim in a concentration-dependent fashion (fig. 2). Maximal vasorelaxation induced by papaverine (3 × 10⁻⁴m) in each group in figure 2(left) was 100% = 5.00 ± 1.05, 5.15 ± 2.17, 5.40 ± 2.17, or 4.85 ± 1.25 g for control rings and rings treated with 10⁻⁵m, 3 × 10⁻⁵m, or 10⁻⁴m lidocaine, respectively (statistically insignificant), and that in each group in figure 2(right) was 100% = 4.75 ± 1.50, 5.64 ± 1.77, 4.65 ± 1.16, or 4.58 ± 1.07 g for control rings and rings treated with 10⁻⁵m, 3 × 10⁻⁵m, or 10⁻⁴m mexiletine, respectively (statistically insignificant).

Protein kinase C inhibitors calphostin C (3 × 10⁻⁷m) and Gö 6976 (3 × 10⁻⁷m) partly restored vasorelaxation in response to levcromakalim in the coronary arteries treated with mexiletine (10⁻⁴m) (figs. 3 and 4). In contrast, in arteries treated with lidocaine (10⁻⁴m), these protein kinase C inhibitors did not produce any statistically significant effects on vasorelaxation, although we noticed a tendency for Gö 6976 to recover vasorelaxation to levcromakalim (figs. 3 and 4). These inhibitors themselves did not alter the vasorelaxation produced by levcromakalim (figs. 3 and 4). A phorbol ester, PMA (3 × 10⁻⁶m), impaired vasorelaxation in response to levcromakalim, which is completely recovered by calphostin C (3 × 10⁻⁷m) or Gö 6976 (3 × 10⁻⁷m) (fig. 5). Maximal vasorelaxation induced by papaverine (3 × 10⁻⁴m) in each group in figure 3(left) was 100% = 5.04 ± 0.97, 4.47 ± 1.16, 4.59 ± 0.99, or 4.40 ± 1.08 g for control rings and rings treated with Gö 6976, lidocaine, or Gö 6976 plus lidocaine, respectively (statistically insignificant), and that in each group in figure 3(right) was 100% = 4.63 ± 1.60, 4.6 ± 1.14, 4.25 ± 0.89, or 5.25 ± 1.39 g for control rings and rings treated with Gö 6976, mexiletine, or Gö 6976 plus mexiletine, respectively (statistically insignificant). Maximal vasorelaxation induced by papaverine in each group in figure 4(left) was 100% = 5.23 ± 1.67, 3.97 ± 1.46, 3.74 ± 1.21, or 4.51 ± 2.19 g for control rings and rings treated with calphostin C, lidocaine, or calphostin C plus lidocaine, respectively (statistically insignificant), and that in each group in figure 4(right) was 100% = 3.58 ± 1.52, 3.82 ± 0.92, 4.23 ± 0.53, or 3.26 ± 1.18 g for control rings and rings treated with calphostin C, mexiletine, or calphostin C plus mexiletine, respectively (statistically insignificant). Maximal vasorelaxation in each group in figure 5(left) was 100% = 5.67 ± 0.82, 4.67 ± 0.82, or 5.00 ± 0.71 g for control rings and rings treated with PMA or PMA plus calphostin C, respectively (statistically insignificant), and that in each group in figure 5(right) was 100% = 5.57 ± 0.79, 4.71 ± 0.76, or 4.43 ± 1.27 g for control rings and rings treated with PMA or PMA plus Gö 6976, respectively (statistically insignificant).

Genistein (10⁻⁶m) and erbstatin A (3 × 10⁻⁶m) partly restored vasorelaxation in response to levcromakalim in the coronary arteries treated with mexiletine (10⁻⁴m) but not in those treated with lidocaine (10⁻⁴m), whereas these inhibitors did not alter vasorelaxation (figs. 6 and 7). Maximal vasorelaxation induced by papaverine (3 × 10⁻⁴m) in each group in figure 6(left) was 100% = 4.47 ± 1.25, 3.91 ± 1.38, 3.71 ± 0.99, or 4.27 ± 1.31 g for control rings and rings treated with genistein, lidocaine, or genistein plus lidocaine, respectively (statistically insignificant), and that in each group of figure 6(right) was 100% = 4.00 ± 1.26, 3.59 ± 1.62, 4.21 ± 0.59, or 4.93 ± 1.08 g for control rings and rings treated with genistein, mexiletine, or genistein plus mexiletine, respectively (statistically insignificant). Maximal vasorelaxation in each group in figure 7(left) was 100% = 5.58 ± 1.57, 5.01 ± 1.17, 4.55 ± 1.10, or 4.78 ± 1.91 g for control rings and rings treated with erbstatin A, lidocaine, or erbstatin A plus lidocaine, respectively (statistically insignificant), and that in each group in figure 7(right) was 100% = 3.91 ± 0.84, 4.38 ± 1.31, 4.11 ± 0.58, or 3.70 ± 1.25 g for control rings and rings treated with erbstatin A, mexiletine, or erbstatin A plus mexiletine, respectively (statistically insignificant).

Levcromakalim (10⁻⁵m) produced hyperpolarization of smooth muscle cells of the coronary artery, which is abolished by glibenclamide (5 × 10⁻⁶m) and reduced by lidocaine (10⁻⁴m) as well as mexiletine (10⁻⁴m) (fig. 8). In the arteries treated with mexiletine (10⁻⁴m) but not lidocaine (10⁻⁴m), levcromakalim-induced hyperpolarization is significantly restored by calphostin C (3 × 10⁻⁷m) as well as by erbstatin A (3 × 10⁻⁶m) (fig. 9). Resting membrane potentials did not differ among the groups shown in figures 8 and 9(for fig. 8: control, −40.1 ± 6.4 mV; glibenclamide [5 × 10⁻⁶m], −45.7 ± 4.1 mV; lidocaine [10⁻⁴m], −40.7 ± 6.8 mV; mexiletine [10⁻⁴m], −43.1 ± 6.4 mV; for fig. 9: lidocaine [10⁻⁴m], −37.2 ± 6.1 mV; calphostin C [3 × 10⁻⁷m] plus lidocaine [10⁻⁴m], −35.2 ± 4.2 mV; erbstatin A [3 × 10⁻⁶m] plus lidocaine [10⁻⁴m], −32.8 ± 2.2 mV; mexiletine [10⁻⁴m], −40.6 ± 6.8 mV; calphostin C [3 × 10⁻⁷m] plus mexiletine [10⁻⁴m], −36.2 ± 1.1 mV; erbstatin A [3 × 10⁻⁶m] plus lidocaine [10⁻⁴m], −37.0 ± 1.9 mV).

In the coronary artery, glibenclamide (5 × 10⁻⁶m) abolished vasorelaxation as well as hyperpolarization in response to levcromakalim (10⁻⁵m), indicating that both compounds probably act on K^ATPchannels.11,12In the current study, we used this concentrations of glibenclamide because, in some specific blood vessel preparations, including rabbit portal veins and cultured aortic smooth muscle cells from neonatal rats, glibenclamide usually above 10⁻⁵m may produce somewhat nonselective effects on ion channels other than K^ATPchannels,13,14and previous studies have demonstrated that higher concentration of glibenclamide (> 10⁻⁵m) is needed to completely block K^ATPchannels in the vascular smooth muscle cell.15However, it is important to note that 10⁻⁵m glibenclamide or levcromakalim has been used as the selective antagonist and the opener of K^ATPchannels in electrophysiologic studies on vascular smooth muscle cells, including those from the porcine coronary artery.16,17In the current study, glibenclamide did not produce any effects on the basal tension and contraction in response to a prostaglandin H²/thromboxane receptor agonist, indicating that at least in the pig, K^ATPchannels may contribute to neither the resting tone nor vasocontraction to an agonist acting via  receptors in the coronary circulation. In contrast to this finding, a recent human study has documented that direct administration of glibenclamide to the large coronary artery during percutaneous coronary intervention provokes reduction of resting vessel diameter, suggesting that in humans, these channels may modulate resting tone of the large coronary artery.5We cannot rule out the possible involvement of species differences in the differential role of K^ATPchannels in the coronary circulation.

In the current study, a phorbol ester, PMA, impaired vasorelaxation in response to levcromakalim, which is completely recovered by selective inhibitors of protein kinase C, calphostin C, and Gö 6976. Calphostin C or Gö 6976 reportedly acts on regulatory and catalytic domains of protein kinase C, respectively, and the former has wide-range effects on protein kinase C isozymes, whereas the latter shows rather limited effects, especially on α and β¹isozymes.18,19Therefore, our results indicate that the activation of α and β¹isozymes of protein kinase C may have a role in the inhibitory effects of PMA in the vasorelaxation mediated by K^ATPchannels in the porcine coronary artery. Previous studies have documented the expression of these subtypes of protein kinase C in arterial smooth muscle cells of the porcine coronary.20In the current study, inhibitors of protein kinase C themselves did not affect vasorelaxation produced by levcromakalim, also supporting the conclusion that the nonselective inhibitory effects of kinase inhibitors in our experimental condition are negligible. Our results are in accord with those of previous studies on the cerebral arteries, portal veins, and mesenteric arteries showing that the activation of protein kinase C inhibits vasorelaxation as well as currents via  K^ATPchannels.4,21–26 

Class Ib antiarrhythmic drugs lidocaine and mexiletine attenuated vasorelaxation as well as hyperpolarization in response to levcromakalim in a concentration-dependent fashion, suggesting that these compounds may impair coronary vasodilation mediated by the activation of K^ATPchannels. Our previous studies on the rat aorta demonstrated the inhibitory effect of lidocaine as well as augmenting effects of mexiletine on the vasorelaxation induced by K^ATPchannel openers.7We do not have a clear explanation of these differential effects of lidocaine and mexiletine on the vasorelaxation mediated by K^ATPchannels between the coronary and the aorta. However, it is most likely that species as well as regional differences contribute to the differential affects of antiarrhythmic drugs.

In the porcine coronary artery, selective protein kinase C inhibitors calphostin C and Gö 6976 similarly restored vasorelaxation or hyperpolarization in response to levcromakalim in the coronary arteries treated with mexiletine but not in those treated with lidocaine. Mutual targets of the protein kinase C isozyme for calphostin C and Gö 6976 are reportedly α and β¹isozymes, and a protein kinase C activator similarly impaired vasorelaxation in response to levcromakalim, which is completely recovered by these inhibitors.18,19Therefore, it is natural to speculate that protein kinase C α and β¹isozymes may contribute to the inhibitory effect of mexiletine but not that of lidocaine on the activity of K^ATPchannels in the coronary artery.

The K^ATPchannel is a complex of two proteins: the sulfonylurea receptor (SUR) and the pore forming subunit, which belongs to the inward rectifier K⁺channel (Kir) family.27Because recent direct functional and biochemical studies have revealed that the SUR of K^ATPchannel is a primary target of the channel openers, the action of lidocaine and mexiletine on some components of SUR may have a role in these inhibitory effects.28In addition to this assumption, recent electrophysiologic studies have documented that activation of protein kinase C is capable of modulating the limited subtype of K^ATPchannel expressed in vascular smooth muscle cells (SUR 2B + Kir6.1).29,30More importantly, the activity of the channel subtypes produced by SUR 2B and Kir6.2 was not altered by the kinase, suggesting the crucial role of the Kir6.1 compartment of K^ATPchannels in the modulator effect of protein kinase C.25Therefore, mexiletine but not lidocaine may modulate Kir6.1 compartment of K^ATPchannels, leading to the inhibition of vasorelaxation mediated by these channels in the coronary artery. However, further electrophysiologic studies using mutation for each compartment in these channels are needed to clarify the exact mechanisms of class Ib antiarrhythmic drugs on K^ATPchannels.

Genistein and erbstatin A restored vasorelaxation or hyperpolarization in response to levcromakalim in the coronary arteries treated with mexiletine but not in those treated with lidocaine. Because previous studies demonstrated that these inhibitors, in the concentrations used in the current study, can be administered as inhibitors of tyrosine kinase, our results indicate that activation of tyrosine kinase may also contribute to the inhibitory effect of mexiletine but not that of lidocaine.31–33Previous studies on cerebral arteries and the portal vein demonstrated that protein tyrosine phosphorylation modulates the activity of K^ATPchannels.6,34,35The evidence that neither inhibitors of tyrosine kinase solely alter vasorelaxation produced by levcromakalim seems to neglect the nonselective inhibitory effects of kinase inhibitors in our experimental condition. These results support our conclusion that in the coronary artery, activity of tyrosine kinase may also have a role in the regulation of vasodilation mediated by K^ATPchannels and that lidocaine and mexiletine differentially modulate this vasorelaxation. In addition, it is important to note that in the coronary artery treated with mexiletine, erbstatin A as well as calphostin C restored changes in membrane potential by approximately 3.5 mV and that in the control artery, levcromakalim produces hyperpolarization by approximately −7.0 mV. Therefore, these results may also indicate that recovery of hyperpolarization obtained by two types of kinase inhibitors in the artery treated with mexiletine is additive.

The therapeutic ranges of plasma concentrations of lidocaine and mexiletine used as antiarrhythmic drugs were reported up to 5 × 10⁻⁵m and 10⁻⁵m for lidocaine or mexiletine, respectively.36,37When one considers the protein binding of these compounds, protein-unbound concentrations of these antiarrhythmic compounds should be lower than those examined in the current study. Therefore, our results may be clinically important from a toxicologic point of view. However, lidocaine and mexiletine may impair coronary vasodilation mediated by K^ATPchannels under some specific conditions in the clinical setting, because higher free plasma concentrations of drugs, which have the capability of protein binding, were reported in diseased infants with a low α-1 acid glycoprotein.38 

Class Ib antiarrhythmic drugs are frequently administered to treat ventricular arrhythmias, which can be seen in patients with ischemic heart disease or those receiving cardiopulmonary resuscitation.39,40It is well known that during hypoxia, acidosis, and ischemia, K^ATPchannels are activated, resulting in coronary arterial dilation, increased tolerance of cardiac myocytes toward ischemia, or both.2,3,41In addition, these antiarrhythmic drugs can be coadministered with clinically available K^ATPchannel openers to treat these patients.42,43Therefore, it may be speculated that lidocaine and mexiletine reduce these beneficial vasodilator effects via  K^ATPchannels, which have important roles in the regulation of coronary circulation during diverse pathophysiologic situations.

This is the first study examining the roles of kinases and the effects of class Ib antiarrhythmic drugs in vasodilation mediated by K^ATPchannels in the coronary circulation. Our results have clearly shown that lidocaine and mexiletine inhibit vasorelaxation as well as hyperpolarization via  K^ATPchannels in the coronary artery. In addition, the activation of protein kinase C and tyrosine kinase seems to contribute to the inhibitory effect of mexiletine but not in that of lidocaine. Class Ib antiarrhythmic drugs may reduce coronary vasodilation mediated by these channels via  the differential modulator effects on these kinases.

The authors thank Zvonimir S. Katusic, M.D., Ph.D. (Professor, Department of Anesthesiology and Molecular Pharmacology, Mayo Clinic, Rochester, Minnesota), for his review of this article.