Inhibition of a Fast Inwardly Rectifying Potassium Conductance by Barbiturates

Abstract

Whole cell voltage clamp recordings were used to study the effects of two barbiturates, methohexital and pentobarbital, on inwardly rectifying K⁺ currents in the plasma membrane of a rat basophilic granulocyte cell line (RBL-1). Inwardly rectifying K⁺ currents are responsible for maintaining the resting membrane potential in a variety of cell types including skeletal and cardiac muscle, neurons, glia, blood cells, and endothelial cells. RBL-1 cells are unusual because the inward rectifier is the only apparent voltage-dependent current in these cells. Steps to command potentials between +80 and -120 mV evoked only this strongly rectifying, rapidly developing current at membrane potentials more hyperpolarized than the reversal potential for K⁺ ions. Extracellular Cs⁺ (10 mM) and Ba²⁺ (100 micro Meter and 1 mM) blocked this current in a reversible and voltage-dependent manner. The voltage threshold for activation of the inwardly rectifying K⁺ current is dependent on the extracellular K⁺ concentration as predicted by the Nernst equation. Methohexital and pentobarbital reversibly inhibited the current in a concentration-dependent fashion with 50% inhibitory concentration (IC₅₀) values of 145 micro Meter and 218 micro Meter respectively. The Hill slopes for both of these effects were approximately 1. The inhibition was not voltage dependent. These results indicate that fast inwardly rectifying K⁺ channels are potential molecular targets for barbiturates and could explain some of the diverse clinical effects of these drugs.

(Anesth Analg 1996;82:1242-6)

Barbiturates continue to be used extensively as intravenous anesthetics and anticonvulsants. The pharmacological properties of these drugs vary according to the chemical structure of the compound, from a potent short-acting sedative/anesthetic such as methohexital (MTX), to the weakly sedative but effective anticonvulsant phenobarbital. In common with many general anesthetics, the ability of barbiturates to potentiate gamma-aminobutyric acid type A (GABA_A) receptor-mediated inhibitory synaptic transmission is closely correlated with, and may underlie, the anesthetic properties of the drugs ^[1,2]. In addition, barbiturates have a variety of effects on other neurotransmitter receptors and some ion channels including excitatory amino acid receptors ^[3], adenosine (A₁) receptors ^[4], and voltage-gated calcium, sodium, and potassium channels ^[5,6]. Some of the effects on ion channels may contribute to the observed peripheral side effects of barbiturates, including decreased systemic vascular resistance and reduced myocardial contractility ^[7].

Inwardly rectifying K⁺ channels are found in a variety of tissues including glia, neurons, both cardiac and skeletal muscle, macrophages, and invertebrate oocytes ^[8-10]. The defining property of these channels is that they allow influx but very little efflux of K⁺ ions. This property of strong rectification accounts for the ability of these channels to determine the resting membrane potential, permit prolonged membrane depolarization, and buffer extracellular K⁺ by transporting the ion into cells ^[11]. A cell which predominantly expresses the inward rectifier current will maintain a resting membrane potential close to the potassium equilibrium potential (E_K). Activation of the large inward current at potentials below E (_K) prevents excessive hyperpolarization of the cell. Similarly, the small outward current which is present when the resting potential is close to E_K also helps maintain a membrane potential close to E_K. Electrophysiological experiments have identified several types of inwardly rectifying currents. These currents show varying degrees of rectification, and some are modulated by G protein-linked neurotransmitter receptors, e.g., the cardiac "K_ACh" channel ^[8]. Recently the deoxyribonucleic acid sequences for several inward rectifiers have been determined (or cloned). The deduced amino acid sequences for these channels are similar to one another but are different from the known sequences for other types of ion channels; therefore, inward rectifiers are considered to constitute a distinct "subfamily" of potassium channels ^[12-15]. In this paper, we have studied a "classic" inwardly rectifying K⁺ channel normally found in a rat basophilic granulocyte cell line (RBL-1). These cells do not express any other voltage-gated ion channels ^[10,16].

Methods

RBL-1 cells were obtained from American Type Culture Collection (Rockville, MD) and maintained in Dulbecco's modified Eagle medium supplemented with 2 mM L-glutamine and 10% fetal bovine serum. Approximately every 4 days the cells were passaged by trypsin treatment, followed by centrifugation and trituration through a fire-polished pipette, before being replated on 35-mm culture dishes or in 75-cm² sterile culture flasks.

Currents were recorded using the whole cell patch clamp technique and low-pass filtered at 1 kHz (-3 dB; Bessel filter, Frequency Devices Inc., Haverhill, MA). Signals were sampled and digitized at 2 kHz using the TL-1-125 analog-to-digital interface (Axon Instruments, Foster City, CA). Voltage commands, acquisition, and analysis were executed using AXOBASIC (Axon Instruments). To study the nature of the effect of barbiturates, current/voltage relationships were determined in the presence and absence of drug. These were obtained from holding potentials between -60 mV and 0 mV (depending on the extracellular [K⁺]), with steps of 200 ms duration to test potentials of -200 mV to +80 mV every 20 s. In 50 mM extracellular K⁺, the slope conductance values (partial I/partial V) were obtained from linear regression analysis of the current-voltage relationship between -120 mV and -80 mV. Inhibition of the current by drugs was calculated from the change in slope of the linear portion of the current-voltage relationship. The 50% inhibitory concentration (IC₅₀) values were derived from a fit of the dose-response data to the equation, R = 1 centered dot {[D]^n/IC₅₀ + [D]ⁿ}, where R = % inhibition of the steady state conductance, [D] = the concentration of drug, and n = the Hill slope, using a version of the Marquardt-Levenberg algorithm (NFIT; Island Products, Galveston, TX).

Patch pipettes were filled with a solution containing (in mM): K gluconate, 140; K₂ adenosine triphosphate (ATP), 5; MgCl₂, 2; EGTA, 1.1; CaCl₂ 0.1; HEPES/KOH, 5 (pH 7.2). The osmolarity of the patch pipette solution was measured using a freezing-point osmometer, and adjusted to 310 mOsm. Seal resistances were >or=to5 G Omega and pipette resistances, <or=to5 M Omega. The extracellular saline contained (mM): NaCl, 145; KCl, 3; D-glucose, 6; CaCl₂, 1.5; MgCl₂, 1; HEPES/NaOH, 10 (pH 7.4). The osmolarity was adjusted to 325 mOsm. For solutions containing higher K⁺ concentrations, the KCl replaced equimolar quantities of NaCl. All drugs were dissolved in the extracellular medium and applied by continuous perfusion of the tissue culture dish containing the cells. Exchange of the extracellular solution was complete in less than 5 min at which point the properties of the potassium currents were measured. Drugs were washed out by perfusion with extracellular solution until the size of the currents returned to the baseline level or for at least 10 min if the drug had no apparent effect. MTX (Brevital sodium) was obtained from Eli Lilly (Indianapolis, IN). Pentobarbital (PTB; Nembutal sodium) was obtained from Abbott Laboratories (North Chicago, IL). ATP was purchased from Calbiochem (San Diego, CA). All other chemicals were supplied by Sigma (St. Louis, MO).

Results

RBL-1 cells cultured as described expressed inwardly rectifying K^+-selective currents. As expected for an inwardly rectifying K^+-selective conductance, currents recorded in 3 mM external K⁺ solution were small and activated only by hyperpolarizing the membrane potential beyond -90 mV. On application of 50 mM external K⁺ solution, the size of the inward current for a given voltage step increased, and the zero current potential shifted in the depolarizing direction, as previously reported for other cells with significant inward rectifier currents ^[11,12]. The change in zero current potential was +66 +/- 3 mV (n = 6), which compares favorably with a theoretical prediction of +70 mV, obtained from the Nernst equation. The current-voltage relation showed strong inward rectification, with a significantly increased slope conductance in 50 mM K⁺ (33.5 +/- 7.2 nanosiemens), relative to that measured in 3 mM K⁺ (8.4 +/- 1.3 nanosiemens; P < 0.02; n = 6, Student's unpaired t-test). In 3 mM K⁺, at membrane potentials more negative than -140 mV, the current showed rapid "inactivation," such that the steady state current-voltage relationship exhibited a negative slope in this voltage range, whereas the instantaneous current-voltage relationship remained almost linear. These properties are typical of inward rectifiers identified in other tissues ^[8,11].

All further experiments were done in an external solution containing 50 mM K⁺ in order to maximize the amplitude of the recorded current. Under these conditions, the inwardly rectifying potassium current was inhibited by both Ba²⁺ and Cs⁺ in a voltage-dependent manner. The data were consistent with Ba²⁺ acting as an open channel blocker of this current. Although the steady-state current was completely inhibited by 1 mM Ba²⁺, there was less inhibition of the instantaneous current immediately upon hyperpolarization, indicating a time-dependent block by Ba²⁺. This effect of Ba²⁺ was concentration dependent; 10 mM Cs⁺ completely ablated the current at very negative test potentials but was less effective on currents activated by smaller voltage steps, consistent with a voltage-dependent action.

Inward rectifier currents in RBL-1 cells were reversibly inhibited by high micromolar concentrations of the short-acting anesthetic MTX Figure 1A and B. This effect did not appear to be voltage dependent, because it was observed at all membrane potentials. In addition, MTX did not affect the kinetics of the current Figure 1A. PTB, another barbiturate anesthetic, was also tested and had similar effects, namely a reversible, voltage-independent inhibition of the current. This indicates that inhibition of inward rectifier potassium currents is common to this class of drugs and not unique to MTX. The concentration-inhibition curves in Figure 2 show that MTX and PTB had comparable pharmacological effects. The inward rectifier current could always be completely inhibited in the presence of sufficient concentrations of drug and the slopes of the curves were roughly parallel with the calculated Hill coefficient being approximately one for both compounds (MTX = 1.4, PTB = 1.2). This indicates that it is unlikely that there was more than one component to the conductance studied and that the inhibition by barbiturates did not involve complex interactions such as cooperativity between multiple binding sites. MTX was more potent than PTB, the IC₅₀ values for MTX and PTB being 145 and 218 micro Meter, respectively. We also tested the effects of these compounds at a single concentration on inward rectifier currents in external solutions containing 3 mM K⁺, and observed that although the currents were much smaller in amplitude, the inhibition by barbiturates was comparable to their effects in high K⁺Figure 3. The concentrations chosen were in the linear part of the dose-response curve for the two drugs, in order to establish that comparable effects were observed in the region of greatest variability in the dose-response relationships.

Figure 1: Methohexital blocked the whole cell inwardly rectifying potassium conductance in a dose-dependent manner in rat basophilic granulocytes. a) Methohexital inhibited the current activated by a voltage step to -100 mV from a holding potential of -20 mV. b) The effect of methohexital is not voltage-dependent and is fully reversible. open square = control; fill square = +50 micro Meter methohexital; fill triangle = +500 micro Meter methohexital; and open triangle = washout. The currents were measured at the end of the 200-ms voltage step. All currents were recorded in 50 mM [K⁺]_e. These data are representative of at least six other experiments.

Figure 2: Concentration-response relationship for the inhibition of the rat basophilic granulocyte inwardly rectifying potassium conductance by methohexital (fill square) and pentobarbital (fill circle). The lines shown are derived from a fit to the equation, R = 1 centered dot [D]^n/IC ₅₀ +[D]ⁿ where R = % inhibition of the steady state current, [D] = the concentration of drug and n = the Hill slope, using a version of the Marquardt-Levenberg algorithm (NFIT; Island Products, Galveston, TX). The 50% inhibition concentration (IC₅₀) values were 145 micro Meter (methohexital) and 218 micro Meter (pentobarbital) and the respective Hill slopes were 1.4 and 1.2. Each point represents the mean +/- SEM for four to six experiments.

Figure 3: Effects of single concentrations of methohexital and pentobarbital on the inwardly rectifying potassium current measured at the end of the voltage step in rat basophilic granulocytes (RBL-1) in 3 mM (fill square) and 50 mM (open square) [K⁺]_e. Data represent the mean +/- SEM for four to six observations.

Discussion

We show that barbiturates inhibit the fast inwardly rectifying K⁺ current in RBL-1 cells. Several inward rectifiers have been identified by cloning the deoxyribonucleic acid complementary to messenger ribonucleic acid (complementary deoxyribonucleic acid) from a variety of tissues and studying the electrophysiological properties of the ion channel proteins by expressing them in heterologous systems such as Xenopus oocytes and other cells which do not normally have inward rectifier currents ^[12,13,15]. These channels share many common properties, some of which we have observed for the RBL-1 inward rectifier. The current in RBL-1 cells rectifies strongly (like IRK1, but distinct from the mildly rectifying channel ROMK1), and the slope conductance was greater in higher external [K⁺]. In low external K (⁺), time-dependent "inactivation" of the current was observed at strongly hyperpolarized potentials, possibly due to block of the channel by external Na^+[16]. We have also observed that the current is blocked by Ba²⁺ and Cs⁺ in a time-and voltage-dependent manner, as shown by Lewis et al. ^[10]. We have not attempted to identify the detailed mechanism of action of the barbiturates on the RBL-1 inward rectifier, but it is clearly distinct from the effects of Ba²⁺ and Cs⁺. It also appears different from the inhibitory effects of barbiturates on voltage-gated Ca²⁺ channels which is a use-dependent slow block (5; S. Gibbons, unpublished observation). There is no indication that the barbiturates modulate the voltage-dependence of the inward rectifier conductance; therefore, it is unlikely that the drugs affect the interaction of cations with the ion channel. The effects of both drugs could be reversed, although MTX took slightly longer to wash out than PTB (data not shown). This is consistent with the partitioning of MTX into the cell membrane and the slow release of residual drug due to its more lipophilic nature.

It would be useful to know the molecular identity of the RBL-1 inward rectifier. The RBL-1 current has superficially similar electrophysiological properties to the IRK channels which have been cloned from mouse macrophages (IRK1) ^[13] except that the current in RBL-1 cells and chromaffin cells is inhibited via G proteins ^[17], whereas IRK1 may not be ^[13].

Further knowledge of the individual protein species responsible for generating the RBL-1 inward rectifier current would indicate whether inhibition of such channels by barbiturates is a widespread phenomenon. Previous reports on such interactions are contradictory. The inwardly rectifying ATP-sensitive K⁺ current (K_ATP) in insulin-secreting cells is inhibited by barbiturates ^[18], whereas Baum ^[19] reported that the inward rectifier in acutely isolated guinea pig ventricular myocytes was not sensitive to 100 micro Meter MTX. In studies of cloned channels, MTX has little effect on either a hippocampal inward rectifier channel ^[26] or on IRK1, expressed in Xenopus oocytes (C. G. Nichols, J. L. Kugler, unpublished observations). The RBL-1 cell inward rectifier is inhibited by intracellular GTP gamma through activation of a pertussis toxin-sensitive G protein ^[17]. This appears to distinguish this channel from the cardiac K_ACh channel and IRK1, which are activated by G proteins ^[13]. Recent evidence, however, suggests that some inward rectifier channels may consist of combinations of at least two distinct polypeptides, which together determine the properties of the expressed channel ^[14]. This complicates considerably the extrapolation of data obtained with homomeric channels, (i.e., tetramers of a single polypeptide) to the study of native inward rectifier currents. We speculate that barbiturate inhibition of the RBL-1 cell inward rectifier may be dependent upon the presence of an auxiliary subunit lacking in the IRK1 and HRK1 homomeric channel expression experiments.

The potency of the barbiturates on the RBL-1 cell inward rectifier is rather low when compared with their potency for potentiating effects on the GABA (_A) receptor. GABA_A receptors are more sensitive to barbiturates than any other identified receptor or channel. For example, MTX and PTB inhibit binding of a selective GABA_A receptor antagonist, t-butylbicyclophosphorothionate, to rat brain synaptic membranes with IC₅₀ values of 12 micro Meter and 63 micro Meter, respectively ^[4]. In the rat, the approximate plasma concentration of PTB during anesthesia has been estimated at 80 micro Meter ^[20] and in humans, effective anesthesia is achieved with a plasma concentration of approximately 10 micro Meter MTX ^[21]. In electrophysiological experiments, 5 micro Meter MTX and 50 micro Meter PTB cause strong potentiation of the action of GABA ^[2]. These concentrations had little effect on the inwardly rectifying channel in this study. If one also takes into account the observation that the inhaled anesthetic, halothane, has no effect on inwardly rectifying potassium channels in either atrial or ventricular myocytes ^[22], it is extremely unlikely that the inhibition of inward rectifiers is relevant to the anesthetic properties of these barbiturates.

This does not mean, however, that the observation described here is irrelevant to observed side effects of these drugs. Barbiturates are more potent as inhibitors of the inward rectifier current in RBL-1 cells and the K_ATP current in insulin-secreting cells than as inhibitors of voltage-gated calcium channels, sodium channels, or outward potassium channels (see, e.g., discussion in ^[6,18]), all of which are considered possible contributors to the overall clinical effect of these drugs. In fact, anesthetics are known to interact with K_ATP channel blockers in modulating peripheral airway reactivity ^[23]. During anesthesia, the plasma concentration of PTB can increase to 50-200 micro Meter ^[20], which is sufficient to cause a 25%-50% inhibition of the RBL-1 K⁺ conductance Figure 2.

In conclusion, we suggest that block of inward rectifiers could contribute to the side-effects of barbiturate anesthetics. Indeed, selective inhibitors of inwardly rectifying K⁺ conductances such as disopyramide and flecainide have been used clinically because of the antiarrhythmic consequences of blocking inward rectifier channels ^[24]. This effect is associated with prolonged repolarization of cardiac myocytes and lengthening of the terminal phases of the action potential ^[25]. Modulation of inward rectifiers by barbiturates should be considered as a possible contributory factor in the side effects of these drugs, because of the ubiquity of inward rectifiers and the important roles they play in the cardiovascular and respiratory systems and in buffering of extracellular K⁺ in the central nervous system.