The depressing effect of tetracaine and ryanodine on the slow outward current correlated with that of contraction in voltage-clamped frog muscle fibres

The effects of tetracaine (10–50 M) and ryanodine (0.1–10 M) were tested on the slow outward K⁺ current (I _so) and the mechanical tension of isolated frog muscle fibres in a voltage-clamp device (double mannitol-gap) connected to a mechanoelectric transducer. In the concentration range tested, both drugs induced a simultaneous inhibition of tension and current. In all cases the effect on tension was twice that on current. The tetracaine-induced current and tension blocks were fully reversible and dose-dependent. In contrast the ryanodine effects on current and tension were not reversible and did not exhibit a dose dependence except for the delay before the onset of the response, which was shortened when the concentration was raised. Linear regression analysis of the time-dependent and dose-dependent effects of both drugs indicated a strong correlation between the decreases in tension and current. It is concluded that the slow outward current is partly under the control of the Ca²⁺ release from sarcoplasmic reticulum during contraction.     

3-Isobutyl-1-methylxanthine (IBMX) affects potassium permeability in rat sensory neurones via pathways that are sensitive and insensitive to [Ca2+]in

The effects of externally applied 3-isobutyl-1-methylxanthine (IBMX), in millimolar concentrations, on the membrane currents in dorsal root ganglia (DRG) neurones isolated from newborn rats were investigated using the amphotericin-based perforated patch-clamp technique. In some experiments, simultaneous measurements of intracellular Ca²⁺ concentration ([Ca²⁺]_in) were performed using fura-2 microfluorimetry. Applications of IBMX induced elevation of [Ca²⁺]_in resulting from Ca²⁺ release from caffeine-ryanodine-sensitive internal stores. In addition to Ca²⁺ release, IBMX produced a biphasic membrane current response comprised of an inward current transiently interrupted by outward current. The onset of the inward current slightly preceded the onset of the [Ca²⁺]_in transient, while the interrupting outward current developed synchronously with the [Ca²⁺]_in rise. The development of IBMX-induced outward current ultimately needed the [Ca²⁺]_in elevation. After the depletion of Ca²⁺ stores by IBMX or caffeine exposure, the subsequent IBMX challenge failed to produce both the [Ca²⁺]_in transient and outward membrane current, although the inward current remained unchanged. Both components of the IBMX-induced membrane current response had a reversal potential close to the K⁺ equilibrium potential and the IBMX-induced membrane current response disappeared while dialysing the cell interior with K⁺-free, Cs⁺-containing solutions suggesting their association with K⁺ channel activity. External administration of 10 mM tetraethylammonium chloride (TEA-Cl) evoked an inward current similar to that observed in response to IBMX; in the presence of TEA-Cl, IBMX application was almost unable to induce additional inward current. IBMX (5 mM) effectively (50%) inhibited K⁺ currents evoked by step depolarizations of membrane potential. We suggest that IBMX affects membrane permeability via activation of Ca²⁺-regulated K⁺ channels and direct inhibition of TEA-sensitive K⁺ channels.     

Single channel recordings from basolateral and apical membranes of renal proximal tubules

A new method is described, which enables the recording of single ionic channels from the basolateral as well as the luminal membrane of renal proximal tubules with the patch-clamp technique. Segments of late proximal tubules of rabbit kidney are cannulated and perfused from one end. The other end is open and freely accessible to a patch pipette. The patch electrode can be moved against lateral cell membranes or can be inserted through the open end into the lumen and brought to contact with the brush-border membrane. Both, in the basolateral as well as in the luminal membrane, giga seals can be achieved. In both membranes, K⁺ selective channels could be identified.Part of this work was presented at the 59th Meeting of the Deutsche Physiologische Gesellschaft, March 1984 in Dortmund.     

Changes of membrane currents in cardiac cells induced by long whole-cell recordings and tolbutamide

Single isolated myocytes were obtained from the ventricles of adult guinea pig hearts. The whole-cell recording configuration of the patch-clamp technique was used to measure membrane currents. A decrease (run-down) of the Ca²⁺ inward current and an increase of a time-independent K⁺ outward current were observed during long lasting (1–3 h) recordings. The time at which the outward current developed depended on the intracellular ATP concentration in the pipette, suggesting that this current is identical to the ATP-dependent K⁺ current described by Noma and Shibasaki (1985). However, the maximum outward current reached in the experiments was independent of the ATP concentration indicating a limited diffusion of ATP in the cell interior. In single-channel experiments on isolated patches of cell membrane and in whole-cell recordings the ATP-dependent K⁺ current could be blocked by the hypoglycaemic sulphonylurea tolbutamide. The IC₅₀ of 0.38 mM was about 50 times higher than that reported for pancreatic -cells (Trube et al. 1986). The Ca²⁺ inward current and the inwardly retifying K⁺ current were not affected by tolbutamide (3 mM).     

L-alanine evokes opening of single Ca2+-activated K+ channels in rat liver cells

Cellular uptake of neutral amino acids via Na⁺ cotransporters is known to be associated with an increased membrane K⁺ conductance mediated by an unknown mechanism that is essential for avoiding excessive cell swelling. We now demonstrate by patch-clamp single-channel current recording that exposure of rat liver cells to L-alanine, but not the poorly transported D-stereoisomer, evokes opening of single K⁺ channels and that this effect is reversible upon removal of the amino acid. The nature of the conductance pathways opened in the intact cell by L-alanine has been investigated in cell-free excised membrane patches where it can be shown that the K⁺-selective channels are opened by Ca²⁺ acting from the inside of the membrane at a concentration as low as 0.1 M.     

Intracellular potassium activity in epithelial cells of frog fundic gastric mucosa

Microelectrodes were used to measure membrane potential and intracellular potassium activity in surface epithelial cells (SEC) of frog (Rana esculenta) fundic gastric mucosa in vitro. Separate measurements were carried out by applying fine-tipped, single barrelled, KCl filled non-selective electrodes and liquid K⁺-selective electrodes. Membrane potentials with respect to the mucosal and serosal surfaces, measured with non-selective electrodes, were –54.5±1.0 S.E. mV (n=59) and –73.0±1.1 S.E. mV (n=59) respectively. The electrical potential difference referred to the mucosal surface, when measured with K⁺-sensitive electrodes, was +21.2±0.8 S.E. mV (n=35), and intracellular K⁺ activity was 98.5 mmol/l. Assuming that intracellular and extracellular K⁺ activity coefficients are equal (_K=_K), the K⁺ concentration is 135.0 mmol/l. The K⁺ equilibrium potential,E _K, was calculated as –90.0 mV i.e. more negative than both membrane potentials. This result indicates active potassium accumulation in the SEC and provides direct evidence of the presence of an active K⁺ pump in either both or in only one of the cell membranes.     

Characterization of an outward K+ current in freshly dispersed cerebral arterial muscle cells

This study was carried out to define some of the cellular ionic mechanisms controlling cerebral arterial muscle. Muscle cells were enzymatically dispersed from cat cerebral arteries. Cells were dialyzed and voltage-clamped using patch pipettes and whole-cell currents measured. Using pipette solutions allowing us to record K⁺ currents we identified an outward current elicited by depolarizing voltage steps beyond –20 mV. This outward current exhibited properties of delayed outward rectification having a peak macroscopic current at +90 mV of 504±236 pA. The current was sensitive to 4-aminopyridine, but was sensitive to tetraethylammonium only at very high doses. When CsCl was in the recording pipette, macroscopic outward currents could not be recorded. Variations in the extracellular Ca²⁺ concentration from 0.5 to 5.0 mM had no effect on current amplitude or voltage dependence; similarly the Ca²⁺ channel blockers nifedipine and Mn²⁺ were without effect on this outward current. The current inactivated slowly with no decay seen even with 3-s command pulses. Repetitive voltage pulses from –60 to +90 mV at a frequency of 1 Hz resulted in cumulative reduction, depressing peak current by 60% after ten pulses. Upon reduction of pH from 7.43 to 7.20 we observed a 350% increase in peak outward current in 7 of 12 cells studied in this regard. Thus, the cellular mechanism responsible for cerebral vascular dilation to acidosis and/or hypercapnia may involve an increase in outward K⁺ current.     

Intracellular calcium modulates basolateral K+-permeability in frog skin epithelium

Cytosolic calcium ([Ca²⁺]_i) has been suggested as a key modulator in the regulation of active sodium transport across electrically tight (high resistance) epithelia. In this study we investigated the effects of calcium on cellular electrophysiological parameters in a classical model tissue, the frog skin. [Ca²⁺]_i was measured with fura-2 in an epifluorescence microscope setup. An inhibition of basolateral potassium permeability was observed when cytosolic calcium was increased. This inhibition was reversible upon removal of calcium from the serosal solution.     

Whole-cell and perforated-patch recordings from O2-sensitive rat carotid body cells grown in short- and long-term culture

We are investigating transduction mechanisms in a major peripheral chemosensory organ, the rat carotid body, using short- and long-term dissociated cell cultures and patch-clamp, whole-cell recording. In this study membrane properties of cultured glomus or type I cells were characterized with conventional whole-cell recording and the new perforated-patch technique during control (160 Torr) and low-PO₂ (20 Torr) conditions. These cells contained voltage-gated channels typical of electrically excitable cells and had large input resistances (approx. 2 G). Under whole-cell voltage clamp the cells produced brief inactivating inward currents, which were largely abolished by 0.2–2.0 M tetrodotoxin, followed by prolonged outward currents, which were reduced by 5 mM tetraethylammonium or abolished by the substitution of Cs⁺ ions for K⁺ ions in the pipette. On exposure to hypoxia the outward K⁺ current was reduced typically by 15%–20% with both conventional whole-cell and perforated-patch recording, which minimizes washout of the cell's cytoplasm. This effect persisted in long-term culture and was specific, since the inward current was unaffected and, moreover, it did not occur in cultured small intensely fluorescent cells, which are closely related to glomus cells. These properties of cultured rat glomus cells are contrasted with those recently reported for freshly isolated rabbit glomus cells.     

Cl−-channels in the apical cell membrane of the rectal gland “induced” by cAMP

Isolated rectal gland tubules (n1000) of dogfish (Squalus Acanthias) were perfused in vitro. Individual channels in the apical and basolateral cell membrane were recorded with the patch clamp method. K⁺-channels were present in excised membrane patches of the basolateral membrane in stimulated (dbcAMP + forskolin + adenosine) and in nonstimulated state. Cl^–-channels were found only in patches of the apical cell membrane when the tubule was stimulated. Cell attached recordings and simultaneous transepithelial PD measurements were obtained while the segment was stimulated. It is shown that concomitant with the increase in lumen negative PD silent membrane patches of the apical cell membrane suddenly develop Cl^–-channel activity. It is concluded that stimulation of rectal gland tubules activates Cl^–-channels in the apical cell membrane.Supported by Deutsche Forschungsgemeinschaft Gr 480/8-2 Correspondence to first author at the above address     

CGRP-induced activation of KATP channels in follicular Xenopus oocytes

The two-microelectrode voltage-clamp technique was used to monitor K⁺ channel activity in Xenopus oocyte follicular cells, which are electrically coupled to the oocyte itself by gap junctions. Endogenous vasodilators such as calcitonin gene-related peptide (CGRP), vasoactive intestinal peptide (VIP), prostaglandin E₂ (PGE₂) and adenosine activate glibenclamide-ATP-sensitive K⁺ (K_ATP) channels in Xenopus oocyte follicular cells. The mechanism of action of CGRP was studied in detail. CGRP effects undergo a rapid desensitization. CGRP acts via CGRP_I receptors. Its effects are antagonized by the amino-truncated CGRP analog hCGRP(8–37). The second messenger for CGRP activation of K_ATP channels is cAMP. Phosphodiesterase inhibition by 3-isobutyl-1-methylxanthine enhances the CGRP response while adenyl cyclase inhibition by either 2,5-dideoxyadenosine or progesterone nearly completely depresses the CGRP response. Vasoconstrictors such as ACh and angiotensin II also have receptors in follicular cells. ACh strongly inhibits the CGRP activation of K⁺ channels as it inhibits the activation of K_ATP channels by P1060, but angiotensin II does not. It is concluded that as in vascular smooth muscle cells, CGRP and probably other hyperpolarizing vasodilators open K_ATP channels in follicular cells by protein kinase A activation.Thanks are due to C. Roulinat and F. Aguila for expert technical assistance. This work was supported by the Centre National de la Recherche Scientifique (CNRS).     

Spontaneous membrane potential oscillations in Madin-Darby canine kidney cells transformed by alkaline stress

High pH is known to be associated with normal cell growth and neoplastic transformation. We observed that Madin-Darby canine kidney (MDCK) cells grown under sustained alkaline stress (pH 7.7) develop foci composed of spindle-shaped cells lacking contact inhibition and exhibiting only poor adhesion to the culture support. Foci-developing (F) cells were cloned and grown in control medium (pH 7.4), where they maintained their neoplastic features indicating a stable pH-induced genetic transformation. After F cells had been fused to giant cells with polyethylene glycol, the cell membrane potential (V _m) was measured by means of microelectrodes. In contrast to non-transformed MDCK cells, V _M of F cells showed spontaneous biorhythmicity caused by periodic opening of Ca²⁺-activated K⁺ channels. Spiking activity was blunted by the Ca²⁺ channel blocker nifedipine, by the K⁺ channel blocker Ba²⁺, by the Na⁺/H⁺ exchange blocker amiloride and its analogue ethylisopropylamiloride, and by an extracellular pH of 7.6 and 6.8. We conclude that MDCK cells transformed by sustained alkaline stress have lost their stable plasma membrane potential but, instead, exhibit endogenous Ca²⁺- and pH-sensitive oscillations.     

A potassium current activated by lemakalim and metabolic inhibition in rabbit mesenteric artery

K⁺ channels which are inhibited by intracellular ATP (ATP_i) (K_ATP channels) are thought to be the physiological target site of the K⁺ channel opening drugs (2) and to underlie a variety of physiological phenomena including hypoxia induced vasodilation (3). However, electrophysiological evidence for ATP_i-regulated K⁺ currents in smooth muscle is scarce. We, therefore, investigated the effects of one K⁺ channel opener, lemakalim, and metabolic inhibition on the membrane conductance of freshly dissociated rabbit mesenteric artery smooth muscle cells, using the perforated-patch whole cell recording technique (6). The cells were metabolically inhibited with 1 mM iodoacetic acid and 50 M dinitrophenol. Both lemakalim (0.1–3 M) and metabolic inhibition activated a time-independent and glyburide sensitive K⁺ current at physiological membrane potentials. The similarities between the lemakalim and metabolic inhibition activated currents suggest that a single class of channels underlies both currents. These results are the first whole-cell current recordings to demonstrate the activation of a smooth muscle membrane conductance by metabolic inhibition, lending support to the view that hypoxia induced vasodilation arises from the activation of K_ATP channels.     

Quinidine blockade of calcium-activated potassium channels in dissociated gastric smooth muscle cells

The effects of quinidine, an antiarrhythmic alkaloid, on potassium-selective channels in enzymatically dissociated gastric smooth muscle cells fromRana pipiens andBufo marinus were investigated using excised patches and the patch-clamp technique. The predominant potassium channel in these cells is the calcium- and voltage-activated maxi-K channel with a single-channel conductance > 100 pS. Applications of quinidine (100–600 M) resulted in resolvable rapid flickerings between the open and blocked states with a corresponding reduction in open channel amplitude and an increase in open channel noise. The currentvoltage curves in the presence of internal quinidine and symmetrical potassium gradients displayed inward rectification. The time-constant of open-time distributions was found to decrease with increasing quinidine concentrations and membrane depolarization. The power-density spectrum of the channel current noise induced by internal quinidine showed a second Lorentzian component with a corner frequency larger than 300 Hz, suggesting that the noise is caused by rapid fluctuations between the open and blocked states. Apparent dissociation constants of 253 M and 209 M for membrane potentials of +20 mV and –60 mV, respectively, were obtained for the quinidine-induced blockade of Ca²⁺-activated K⁺ channels in these smooth muscle cells. Another potassium-selective channel with a single-channel conductance of 40 pS was completely blocked in the presence of 100 M qunidine. However, a 15 pS potassium channel was not affected by quinidine but was reversibly blocked by tetraethylammonium. Quinidine (500 M) was also observed to decrease the opening probability of a 40 pS potassium channel fromBufo marinus without affecting its channel amplitude. Thus, quinidine appears to have diverse mechanisms of action on potassium-selective channels in smooth muscle cells, ranging from totally ineffective to highly selective, as a slow blocker for some channels and as intermediate and fast blockers for others.     

Effects of intracellular Ca2+ chelating compounds on inward currents caused by Ca2+ release from sarcoplasmic reticulum in guinea-pig atrial myocytes

Ca²⁺ release from the sarcoplasmic reticulum (SR) of mammalian cardiac myocytes occuring either due to activation by a depolarization or the resulting transmembrane Ca²⁺ current (I _Ca), or spontaneously due to Ca²⁺ overload has been shown to cause inward current(s) at negative membrane potentials. In this study, the effects of different intracellular Ca²⁺ chelating compounds on I _Ca-evoked or spontaneous Ca²⁺-release-dependent inward currents were examined in dialysed atrial myocytes from hearts of adult guinea-pigs by means of whole-cell voltage-clamp. As compared to dialysis with solutions containing only a low concentration of a high affinity ethylene glycol-bis(-aminoethylether) N,N,N,N-tetraacetic acid (EGTA) like chelator (50–200 M), inward membrane currents (at –50 mV) due to evoked Ca²⁺ release, spontaneous Ca²⁺ release or Ca²⁺ overload following long-lasting depolarizations to very positive membrane potentials are prolonged if the dialysing fluid contains a high concentration of a low affinity Ca²⁺ chelating compound such as citrate or free adenosine 5-triphosphate (ATP). Without such a non-saturable Ca²⁺ chelator in the dialysing fluid, Ca²⁺-release-dependent inward currents are often oscillatory and show an irregular amplitude. With a low affinity chelator in a non-saturable concentration, discrete inward currents with constant properties can be recorded. We conclude that the variability in Ca²⁺-release-dependent inward current seen in single cells arises from spatial inhomogeneities of intracellular Ca²⁺ concentration ([Ca²⁺]_i) due to localized saturation of endogenous and exogenous high affinity Ca²⁺ buffers (e.g. [2]). This can be avoided experimentally by addition of a non-saturable buffer to the intracellular solution. This condition might be useful, if properties of Ca²⁺ release from the SR and/ or the resulting membrane current, like for example arrhythmogenic transient inward current, are to be investigated on the single cell level.     

Potassium conductance of smooth muscle cells from rabbit aorta in primary culture

Vascular smooth muscle cells were obtained from rabbit aorta and were studied in primary culture on days 1–7 after seeding with electrophysiological techniques. In impalement experiments a mean membrane potential difference (PD) of –50±0.3 mV (n=387) was obtained with Ringer-type solution in the bath. PD was depolarized by 6±0.3 mV (n=45) and 16±2 mV (n= 5) when the bath K⁺ concentration was increased from the control value of 3.6 mmol/l to 13.6 and 23.6 mmol/l, respectively. Ba²⁺ (0.1–1 mmol/l) depolarized PD. Tetraethylammonium (TEA, 10 mmol/l) depolarized PD only slightly but significantly. Verapamil (0.1 mmol/l) and charybdotoxin (10 nmol/l) had no effect on PD. The conductance properties of these cells were further examined with the patch-clamp technique. K⁺ channels were spontaneously present in cell-attached patches. When the pipette was filled with 145 mmol/l KCl, a mean conductance (g _K) of 209.6±4.6 mV (n=17) was read from the current/voltage curves at a clamp voltage (V _c) of 0 mV. After excision K⁺ channels were found in 129 patches with inside-out and in 50 with outside-out configuration. With KCl on one and NaCl on the other side the mean g _K at a V _c of 0 mV was 134.6±3.9 pS (n=179). The mean permeability was 0.89±0.03×10^–12 cm³/s. With symmetrical KCl solution the mean g _K was 227±6 pS (n=17). The conductance sequence was g _K g _Rb= g _Cs=g _Na=0. TEA blocked dose-dependently only from the outside.(1–10 mmol/l). Lidocaine (5 mmol/l) quinidine (0.01–1 mmol/l) and quinine (0.01–1 mmol/l) blocked from both sides. Charybdotoxin (0.5–5 nmol/l) blocked only from the extracellular side. Ba²⁺ blocked from the cytosolic side and the inhibition was increased by depolarization and reduced by hyperpolarization. At a V _c of 0 mV a half-maximal inhibition (IC₅₀) of 2 mol/l was obtained. Verapamil and diltiazem blocked from both sides, verapamil with an IC₅₀ of 2 mol/l and diltiazem with an IC₅₀ of 10 mol/l. The open probability of this channel was increased by Ca²⁺ on the cytosolic side at activities > 0.1 mol/l. Half-maximal activation occurred at Ca²⁺ activities exceeding 1 mol/l. The present data indicate that the vascular smooth muscle cells of rabbit aorta in primary culture possess a K⁺ conductance. In excised patches only a maxi K⁺ channel was detected. This channel has properties different from the macroscopic K⁺ conductance. Hence, it is likely that the K⁺ conductance of the intact cell is dominated by yet another and thus far not detected K⁺ channel.Supported by DFG Gr 480/10     

Tedisamil inhibits the delayed rectifier K+ current in single smooth muscle cells of the guinea-pig portal vein

Tedisamil is a new bradycardic agent with an inhibitory action on K⁺ channels in cardiac muscle, and secondary beneficial effects in experimentally induced cardiac ischemia. In whole-cell clamp studies in enzymatically dispersed, single smooth muscle cells from the guinea-pig portal vein, tedisamil inhibited the delayed rectifier K⁺ current (determined as the charge transferred through the cell membrane), the mean concentration for half-maximal inhibition being 2.9 M. In contrast to controls in the absence of drugs or in the presence of the classical K⁺ channel blockers barium, tetraethylammonium or 4-aminopyridine, the time course of the delayed rectifier K⁺ current in the presence of tedisamil could no longer be fitted by a single exponential, and signs of an accelerated inactivation by tedisamil were obtained. The slow onset of the response to tedisamil applied to the outside of the vascular myocytes, and the finding that tedisamil applied directly to the cytosol via the pipette was highly effective, suggest an intracellular site of action.     

Use of selective toxins to separate surface and tubular sodium currents in frog skeletal muscle fibers

The interaction between toxin from the venom of the scorpionTityus serrulatus and sodium channels in skeletal muscle membranes from the frogCaudiverbera caudiverbera was studied. Sodium current from cut sartorius muscle fibers is a complex signal in which early and late components are difficult to separate. External application of Tityus toxin initially blocked the early component in a voltage-dependent manner. Longer exposure to the toxin induced a complete blockade of the two components of the inward current. Application of tetrodotoxin to fibers pretreated with Tityus toxin at submaximal concentrations allowed the observation of the two distinct components of the inward current. Binding of¹²⁵I-labelled toxin to highly purified membrane fractions from the same muscle was used to establish the presence of high affinit receptors both in the transverse-tubular and in the surface membrane.     

Microelectrode study of voltage-dependent Ba2+ and Cs+ block of apical K+ channels in the skin of Rana temporaria

The blockage of the apical K⁺ channels in frog species Rana temporaria by Ba²⁺ and Cs⁺ is strongly voltage-dependent. The interaction of both blockers with the K⁺ channels was studied by recording relations between the K⁺ currents (I _K) and the transepithelial and intracellular potential. Mucosal Ba²⁺ and Cs⁺ depress I _K, hyperpolarize the cell and induce pronounced nonlinearities in the current/voltage (I/V) relations. The nonlinearities are caused by the voltage-dependent interaction of Ba²⁺ and Cs⁺ with the binding site. Consequently, the apical membrane resistance not only depends on the blocker concentration but also on the apical membrane potential. Also the fractional resistance, fR _a, and the voltage divider ratio, fV _a, will change with blocker concentration and voltage. Owing to this non-ohmic behaviour, measurements of fV _a in the presence of Ba²⁺ deviate markedly from the expected fR _a values. The inhibitory effect of Ba²⁺ and Cs⁺ was analysed at different transepithelial and apical membrane voltages. The relation between the Michaelis-Menten constants and the voltage could be fitted with equations based on Eyring rate theory with the assumption of a single binding site. With this model we calculated the relative electrical position of the binding site for the blocker (), referred to the extracellular side of the channel. We obtained for Ba²⁺, =0.34±0.05 and for Cs⁺, =0.81±0.01. Comparison of the results from apical and transepithelial I/V relations demonstrates that the analysis of the transepithelial data provides overestimated values of the Hill coefficient and results in an underestimation of .     

Heparin uncouples the muscarinic receptors from GK protein in the atrial cell membrane of the guinea-pig heart

The effects of heparin on activation of the G protein-gated muscarinic K⁺ channel were examined in atrial cells of guinea-pig heart. The inside-out patch clamp technique was used. The pipette solution contained 1.1 M acetylcholine (ACh). In the inside-out patches, intracellular GTP activated the muscarinic K⁺ channel. When heparin (0.05–5 units/ml) was further added to the intracellular side of the patch membrane, the channel openings were depressed in a concentration-dependent fashion. The effects of heparin were reversible after wash-out. Heparin did not affect GTP-S-induced activation of the K⁺ channel. Therefore, it is suggested that heparin may uncouple the muscarinic receptors from G_K protein in the cardiac atrial cell membrane.