In Vitro  Electrophysiologic Effects of Morphine in Rabbit Ventricular Myocytes

The New Zealand white rabbits used in the current study were cared for in accordance with the Guide for the Care and Use of Laboratory Animals ,16and the experimental protocol was approved by the Ethics Committee on the Use of Animals in Teaching and Research, The University of Hong Kong (Pokfulam, Hong Kong SAR, China).

Ventricular myocytes were isolated from the hearts of rabbits with a procedure previously described.17Briefly, New Zealand white rabbits of either sex (1.5–2.5 kg) were anesthetized with pentobarbital (30 mg/kg, intravenous injection), and their hearts were quickly removed and placed in oxygenated Tyrode solution. Hearts were mounted on a Langendorff system and perfused for approximately 5 min with normal Tyrode solution and for an additional 10–12 min with Ca²⁺-free Tyrode solution. Then, perfusion solution was switched to the Ca²⁺-free Tyrode solution containing 0.5 mg/ml collagenase (CLS II; Worthington Biochemical, Freehold, NJ) and 1 mg/ml bovine serum albumin. The myocytes isolated from the softened heart were stored in high-K⁺storage solution. Myocytes were placed in the recording chamber (0.3 ml) mounted on the stage of an inverted microscope and superfused at approximately 2 ml/min with external solution. Only quiescent, rod-shaped cells showing clear striations were selected for experiments. The experiments were performed at 36°C for recording action potentials and delayed rectifier K⁺current (I^K, i.e. , I^Krand I^Ks) or room temperature (21°–22°C) for recording I^Ca.Land I^K1.

The high-K⁺storage solution contained 10 mm KCl, 10 mm KH²PO⁴, 20 mm glucose, 120 mm K-glutamate, 10 mm taurine, 0.5 mm EGTA, 10 mm HEPES, and 1.8 mm MgSO⁴(pH adjusted to 7.2 with KOH). The Tyrode solution contained 140 mm NaCl, 5.4 mm KCl, 1.0 mm MgCl², 1.8 mm CaCl², 0.33 mm NaH²PO⁴, 10 mm glucose, and 10 mm HEPES (pH adjusted to 7.4 with NaOH). When Tyrode solution was used to record K⁺currents, 200 μm Cd²⁺and 3 mm 4-aminopyridine were added to respectively block I^Ca.Land transient outward K⁺current (I^to). The pipette solution for K⁺current recordings contained 20 mm KCl, 110 mm potassium aspartate, 1.0 mm MgCl², 10 mm HEPES, 5 mm EGTA, 0.1 mm GTP, 5 mm Mg²ATP, and 5 mm sodium phosphocreatine (pH adjusted to 7.2 with KOH). An N -methyl-d-glucamine Tyrode solution, containing 140 mm N -methyl-d -glucamine, 5.4 mm CsCl, 1.0 mm MgCl², 1.8 mm CaCl², 0.33 mm KH²PO⁴, 10 mm glucose, and mm 10 HEPES, was used when I^Ca.Lwas measured (pH adjusted to 7.4 with CsOH). The pipette solution for I^Ca.Lrecording contained 20 mm CsCl, 110 mm cesium aspartate, 1.0 mm MgCl², 10 mm HEPES, 5 mm EGTA, 0.1 mm GTP, 5 mm Mg²ATP, and 5 mm sodium phosphocreatine (pH adjusted to 7.2 with CsOH). Morphine (sulfate salt) was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO) and dissolved in distilled water. All other chemicals were obtained from Sigma-Aldrich.

Based on the pharmacokinetics of morphine in humans, i.e. , a volume of distribution of 3.2 l/kg and 35% protein binding, the equilibrate plasma concentration is approximately 0.06 μm after intravenous injection of 0.2 mg/kg morphine.18We therefore used clinically relevant concentrations (0.01–1 μm) of morphine. To study the interaction of morphine with opioid receptors, effects of morphine on I^Ca.Land I^K1were determined in the presence of 5 μm naltrindole (NTD; a selective δ-opioid receptor antagonist), 5 μm nor-binaltorphimine (nor-BNI; a selective κ-opioid receptor antagonist), or 5 μm CTOP (a selective μ-opioid receptor antagonist).

The whole cell patch clamp technique was used17,19to record membrane current. Borosilicate glass electrodes (1.2 mm OD) were pulled with a Brown-Flaming puller (model P-97; Sutter Instrument Co., Novato, CA) and had a tip resistance of 2–3 MΩ when filled with pipette solution. The tip potentials were compensated before the pipette touched the cell. After a gigaseal was obtained, the cell membrane was ruptured by gentle suction to establish the whole cell configuration. Data were acquired with the use of EPC-9 (Heka Electronik, Lambrecht, Germany). Command pulses were generated by a 12-bit digital-to-analog converter controlled by Pulse software (Heka Electronik). The recording of ionic currents were low-pass filtered at 5 kHz and stored on the hard disk of a computer. The series resistance (Rs) was electrically compensated to minimize the capacitive surge on the current recording and voltage drop across the clamped membrane and was kept at a constant value during the current recording. Membrane capacitance was 116.5 ± 12.8 pF (n = 45). To account for differences in cell size, all mean data are expressed as current density (i.e. , normalized to cell membrane capacitance). For recording action potential, a perforated patch was applied using a procedure of back-filling amphotericin-B (160 μg/ml; Sigma-Aldrich) in the K⁺pipette solution.20Action potentials were recorded in current clamp mode. Recorded membrane potentials were corrected by 10 mV for the liquid junction potential between the pipette and external solutions.

Nonlinear curve fitting was performed using Sigmaplot (SPSS, Chicago, IL). Concentration–response effects of morphine were fit to the Hill equation: E = E^max/[1 + (EC⁵⁰/C)^b], where E is the effect at concentration C, E^maxis the maximal effect, EC⁵⁰is the concentration for half-maximal effect, and b  is the Hill coefficient. Curves for steady state activation (d ) and inactivation (f ) of I^Ca.Lwere fit to Boltzmann relations for activation and inactivation as follows: d  (or f ) = 1/{1 + exp[(V^0.5− V)/K]}, where V is the membrane potential, V^0.5is a midpoint of potential for activation or inactivation, and K is a slope factor.

Comparison of two means was performed using the paired and unpaired Student t  tests, and comparison of several means was performed using analysis of variance, then with a Dunnett post hoc  test for the comparison with control. Statistical differences were considered significant if the P  value was less than 0.05. Results are presented as mean ± sd.

Figure 1shows representative action potentials recorded at 1 Hz in current clamp mode from a rabbit ventricular myocyte in the presence of 0.01, 0.1, and 1 μm morphine and after washout of the drug. The APD was substantially increased, and the resting membrane potential was slightly hyperpolarized by application of morphine. The effects were partially reversed by washout of the drug for 10 min. The effects on action potential variables in each experiment exposed to three concentrations of morphine are summarized in table 1.

Because I^Ca.Lis mainly responsible for the plateau of action potentials of the cardiac muscle, we determined the effect of morphine on the current. Figure 2Adisplays I^Ca.Ltracings recorded from a representative cell with 300-ms voltage steps to between −40 and +60 mV from −50 mV as shown in the inset. Morphine at 0.1 μm substantially increased I^Ca.Lafter administration for 6 min. The effect was partially reversed upon washout for 10 min. Figure 2Bshows current–voltage relations of mean values of I^Ca.Ldensity (n = 12) in the absence and presence of 0.1 μm morphine. The density of I^Ca.Lat +10 mV corresponding to peak current was significantly increased by 23 ± 8% (from 5.9 ± 1.9 to 7.3 ± 1.7 pA/pF; P < 0.05).

Figure 2Cillustrates the concentration–response relation for the augmentation of I^Ca.Lby morphine. The mean data were fit to the Hill equation, and the EC⁵⁰(at +10 mV) was 0.47 μm with a Hill coefficient of 0.31, and the E^maxwas 47%.

Effects of morphine on steady state voltage dependence of I^Ca.Lactivation and inactivation were evaluated in rabbit ventricular cells. Voltage-dependent activation of I^Ca.Lwas determined, as described previously,21from the current–voltage relation of I^Ca.Lin figure 2Bon the basis of the formulation: d^t= I^t/[g^x(V^t− V^r)], where d^tis the activation variable and I^tis I^Ca.Lat a test potential V^t, g^xis the maximum conductance, and V^ris the reversal potential. Calculated values were normalized to the maximum value in each cell to obtain the activation variable d . Figure 3Aillustrates the voltage protocol and representative recordings used to assess I^Ca.Linactivation. The inactivation variable f ) was determined as I^Ca.Lat a given prepulse potential divided by the maximum I^Ca.Lin the absence of a prepulse. Figure 3Bshows the results obtained from analysis of voltage-dependent activation and inactivation before and after 0.1 μm morphine treatment. Mean data are shown by the symbols, and the curves shown are best-fit Boltzmann distributions. The midpoint of voltage (V^0.5) and slope factor for activation averaged −5.7 ± 1.2 mV and 6.4 ± 1.1 for control, and those for morphine (0.1 μm) were −6.7 ± 1.3 mV and 6.1 ± 1.0 (n = 9), respectively. The V^0.5and slope factor for inactivation averaged −20.1 ± 1.1 mV and −4.6 ± 0.9 for control and −20.4 ± 0.8 mV and −4.5 ± 0.7 for morphine (0.1 μm) treatment (n = 10), respectively. The steady state activation and inactivation of I^Ca.Lwere not affected by morphine. In addition, recovery of I^Ca.Lfrom inactivation, determined with the procedure as previously described,21was not influenced by 0.1 μm morphine, either (n = 7).

Figure 4shows the effects of 0.1 μm morphine on I^Ca.Lat +10 mV in the absence and presence of opioid receptor antagonists. These three types of opioid receptor antagonists themselves had no effect on I^Ca.L. NTD or nor-BNI, but not CTOP, abolished the action of morphine on the current, suggesting that the enhancement of I^Ca.Lby morphine is mediated by both δ- and κ-opioid receptors.

In addition, the prolongation of cardiac APD by morphine was prevented by NTD or nor-BNI, but not by CTOP. In cells with pretreatment of 5 μm NTD and 5 μm nor-BNI, APD⁵⁰(at 1 Hz) was 124.2 ± 31.6 and 121.7 ± 36.5 ms, respectively, and 122.9 ± 39.1 and 125.7 ± 35.3 ms after application of 0.1 nm morphine (n = 6; not significant). Nevertheless, in cells pretreated with 5 μm CTOP, APD⁵⁰was 126.8 ± 34.9 and 142.3 ± 41.3 ms in the absence and presence of 0.1 μm morphine, respectively (n = 8; P < 0.05). These results indicate that the prolongation of cardiac APD by morphine is related to the increase of I^Ca.Land mediated by δ- and κ-opioid receptors.

Figure 5displays the effect of morphine on I^K(I^Krand I^Ks) in a typical experiment. Morphine at 0.1 μm had no effect on I^K. In a total of six ventricular myocytes, the tail current (measured at −30 mV from the peak tail to the completed deactivation level) of I^Kwas 2.2 ± 0.8 pA/pF in control and 2.2 ± 0.9 after application of 0.1 μm morphine (n = 6; not significant). No change in I^Ktail current was observed even when the concentration of morphine was increased to 1 μm (2.3 ± 0.7 pA/pF, n = 6; not significant).

Figure 6Aillustrates representative I^K1tracings elicited by 300-ms voltage steps to between −120 and 0 mV from −40 mV during control, after the application of 0.1 μm morphine, and after washout of the drug. I^K1was increased after the application of 0.1 μm morphine for approximately 5 min. The effect was reversed only partially after washout of the drug for 10 min. Figure 6Bis the current–voltage relation of I^K1density (n = 12) in the absence and presence of 0.1 μm morphine. I^K1was significantly increased by morphine in both inward (−120 and −110 mV) and outward (−70 to −40 mV; P < 0.05) components. At −60 mV, I^K1was 2.8 ± 1.0 pA/pF during control. After the application of 0.01, 0.1, and 1 μm morphine, I^K1was increased to 3.1 ± 1.1, 3.5 ± 0.9, and 3.7 ± 1.0 pA/pF, respectively (i.e. , increased by 11 ± 7, 25 ± 9, and 32 ± 11%; P < 0.05 for 0.1 and 1 μm morphine).

Figure 6Csummarizes the mean data of morphine effect on I^K1at −60 mV in the absence and presence of opioid receptor antagonists. Pretreatment with 5 μm NTD, 5 μm nor-BNI, or 5 μm CTOP did not affect the action of morphine on the current, indicating that the increase of I^K1by morphine is not mediated by opioid receptors.

In the current study, we demonstrated that morphine at 0.01–0.1 μm prolonged APD and increased I^Ca.Lin isolated rabbit ventricular myocytes. The effects were prevented by blockade of δ- or κ-opioid receptors. At 0.1–1 μm, morphine also hyperpolarizes the resting membrane potential and enhances I^K1, but this effect was not mediated by opioid receptors. Based on the pharmacokinetics of morphine in humans, i.e. , a volume of distribution of 3.2 l/kg and 35% protein binding, the equilibrate plasma concentration is approximately 0.06 μm after intravenous injection of 0.2 mg/kg morphine.18Therefore, the concentrations at 0.01–0.1 μm used in the current study are within therapeutic concentrations.

It is well known that prolongation of action potential is dependent on increase of inward current (e.g. , I^Ca.L) or decrease of outward K currents (I^Kor I^K1). A previous report demonstrated that morphine prolonged cardiac APD in cardiac tissue,13but the ionic mechanisms are not understood. In the current study, using a whole cell patch clamp technique, we provided evidence that morphine prolonged cardiac APD by increasing I^Ca.L, not by decreasing I^Kor I^K1. The increase of I^Ca.Lby morphine had an EC⁵⁰of 0.5 μm with a Hill coefficient of 0.31. The small Hill coefficient suggested a wide effective concentration range of morphine.

The current observation on the increase of cardiac I^Ca.Lby morphine was supported by previous studies using ⁴⁵Ca²⁺or spectrofluorometry, in which morphine was found to induce transmembrane Ca²⁺influx in cardiac myocytes.9,22In contrast to the current finding, stimulation of opioid receptor with 0.01 μm leucine-enkephalin, a δ-opioid receptor agonist, was shown to reduce I^Ca.Lin rat ventricular myocytes.23This discrepancy may be due to the use of different opioid receptor agonists in the two studies: morphine in the current study and leucine-enkephalin in the previous study.23Actually, the opposing effects of opioids on voltage-dependent Ca²⁺channels (I^Ca) were documented in nerve cells. Most of the studies indicated inhibitory effects of opioids on I^Ca.L,24–26but stimulation was also observed.27 

It was reported that morphine at 30 μm did not affect I^Ca.Lin rat ventricular myocytes.14The discrepancies between these findings might be related to drug concentration, species, experimental procedures, opioid receptor subtypes involved, or other unknown factors. In the heart, L-type calcium channels play an important role in generating electrical activity and excitation–contraction coupling. It is well established that increases in I^Ca.Lproduce positive inotropic effects and induce triggered arrhythmias, which is highly undesirable in patients with acute myocardial infarction. However, increases in I^Ca.Lby Bay Y5959, a Ca²⁺promoter, have recently been shown to prevent the initiation of reentrant ventricular tachycardia in the epicardial border of healing infarcted canine heart,28indicating antiarrhythmic actions of an increase in I^Ca.L.

A previous study reported that activation of opioid receptors would hyperpolarize cell membrane in neurons.29In the current study, we showed that morphine at the therapeutic concentration of 0.1 μm increased I^K1in cardiac myocytes via  an opioid receptor–independent pathway. There is evidence that a reduction in I^K1is associated with arrhythmias. Patients with Andersen syndrome, a rare inherited disease with a mutation in a major contributor of the I^K1channel, Kir2.1, which causes a reduction in I^K1, exhibit complex ventricular ectopy and polymorphic ventricular tachycardia.30,31Therefore, it is likely that an increase in I^K1hyperpolarizes the resting membrane potential, thus producing a membrane stabilizing effect.

It should be noted that the concentrations of morphine used in the previous study29were much higher than those used in clinical practice. Recently, it was found that morphine at low concentrations produced preconditioning in brain Purkinje cells.32Yvon et al.  15reported that morphine at clinically relevant concentrations did not modify action potential variables in guinea pig ventricle in normoxic conditions but significantly attenuated the ischemia-induced depolarization and increase in amplitude of action potential and V^max, and therefore reduced the incidence of conduction block during ischemia and reperfusion-induced arrhythmias. Nevertheless, we observed that morphine at 0.1 and 1 μm hyperpolarized the resting membrane potential significantly and prolonged APD in single ventricular myocytes from rabbit hearts. Analyses of ionic currents showed that morphine increases both I^Ca.Land I^K1. The increase in I^Ca.Land I^K1would prolong and shorten the cardiac APD, respectively, thus resulting in a moderate prolongation of the APD. The effects of morphine on these two types of currents would be responsible for decreased ischemia-reperfusion–induced arrhythmias as observed by Yvon et al.  15Together with its analgesic and cardioprotective action, morphine at the therapeutic concentrations may be very useful in the management of acute myocardial infarction.

The following limitations must be considered in the assessment of the relevance of our study. The species differences, in vitro  study, healthy myocardium versus  diseased myocardium, normal physiologic conditions versus  abnormal conditions (e.g. , hypoxia, ischemia),33and interaction with other anesthetic agents34should all be taken into consideration. In addition, no clear explanation was available for the observed slow reversibility of morphine effect (by washout). Possible explanations for this phenomenon would be the tight binding of morphine with cell membrane protein and the participation of intracellular events. It has been found that several signal pathways (e.g. , protein kinase A, protein kinase C, protein kinase G) are involved in physiologic actions mediated by opioid receptors.35–37The slow reversibility is likely related to the fact that the effects from these pathways may require a longer time to disappear after washout of the receptor ligand.

It is interesting to note that blockade of either δ- or κ-opioid receptors abolished the effect of morphine. A similar phenomenon has been reported previously by us. The infarct-sparing effect of preconditioning of remifentanil, an ultrarapid-acting opioid receptor agonist, is abolished by blockade of κ- or δ-opioid receptors.38Further study is required to determine whether this is related to interaction of κ-opioid receptors with δ-opioid receptors.

In conclusion, the current study has demonstrated that morphine at a clinically relevant concentration of 0.1 μm increases I^Ca.Land I^K1in isolated rabbit ventricular myocytes. Its action on I^Ca.Lis mediated by δ- and κ-opioid receptors, whereas its effect on I^K1is not mediated by opioid receptors.

The authors thank Haiying Sun, B.Sc. (Research Assistant, Department of Medicine, The University of Hong Kong, Hong Kong, China), and Chi-Pui Mok (Technician, Department of Physiology, The University of Hong Kong, Hong Kong) for excellent technical support.