Identification of an ATP-sensitive K+ channel in rat cultured cortical neurons

To determine whether membranes of mammalian central neurons contain an ATP-sensitive K⁺ (K_ATP) channel similar to that present in pancreatic cells, the patch-clamp technique was applied to cultured neurons prepared from the neonatal rat cerebral cortex and hippocampus. In whole-cell experiments with hippocampal neurons, extracellular application of 0.5 mM diazoxide (a K_ATP channel activator) elicited a hyperpolarization concomitant with an increase in membrane conductance, whereas application of 0.5 mM tolbutamide (a K_ATP channel blocker) induced a depolarization with a decrease in conductance. Similar results were obtained with cortical neurons. In outside-out patch experiments with cortical neurons, a K⁺ channel sensitive to these drugs was found. The channel was completely blocked by 0.5 mM tolbutamide and activated by 0.5 mM diazoxide. The single-channel conductance was 65 pS under symmetrical 145 mM K⁺ conditions and 24 pS in a physiological K⁺ gradient. In inside-out patch experiments, this channel was demonstrated to be inhibited by an application of 0.2–1 mM ATP to the cytoplasmic surface of the patch membrane. These results indicate that the membranes of rat cortical neurons contain a K_ATP channel that is quite similar to that found in pancreatic cells. It is also suggested that the same or a similar K⁺ channel may exist in membranes of hippocampal neurons.     

Potassium channel openers act through an activation of ATP-sensitive K+ channels in guinea-pig cardiac myocytes

In a previous article (Escande et al. 1988a), we have shown that cromakalim (BRL 34915), a potassium channel opener (PCO), is a potent activator of ATP-sensitive K⁺ channels in cardiac cells. In the present article, the influence on K⁺ channels of two other potassium channel openers chemically unrelated to cromakalim, RP 49356 and pinacidil, has been investigated in patch-clamped isolated cardiac myocytes. In the whole-cell configuration, K⁺ currents were recorded in the presence of 50 M TTX and 3 M nitrendipine or 3 mM cobalt. Like cromakalim, RP 49356 or pinacidil activated a time-independent outward current at 33–35°C but not at 19–21°C, which showed little voltage-dependency in the potential range –60 to +60 mV. Its amplitude was a function of the agonist concentration, e.g. it was 2.1±0.4 nA at +60 mV with 30 M RP 49356 and 4.3±0.8 nA with 300 M. In control conditions, glibenclamide, a blocker of K⁺-ATP channels in pancreatic and heart cells, affected neither the inward rectifier,i _K1, nor the delayed K⁺ current,i _K. At 3 M, glibenclamide fully prevented the effects of 300 M RP 49356 or pinacidil. At lower concentrations, glibenclamide partially counteracted the activation by PCOs of a K⁺ current. In the cell-attached contiguration, externally applied RP 49356 or pinacidil caused opening of large channels which reversed around 0 mV in a high K⁺ external medium. In inside-out patches, both RP 49356 or pinacidil activated K⁺-ATP channels by increasing the time period for which the channels remained in the open state. It is concluded that, like cromakalim, RP 49356 and pinacidil are potent activators of K⁺-ATP channels in cardiac myocytes.     

Identification of an ATP-sensitive K+ channel in spiny neurons of rat caudate nucleus

On the somata of GABAergic spiny neurons in the caudate nucleus of the rat an ATP-sensitive K⁺ channel (K_ATP-channel) was identified. The K_ATP-currents in cell-attached patches were activated both by energy-depleting conditions (200 M cyanide) and by diazoxide (300 M) and were reversibly blocked by tolbutamide (EC₅₀=5 M). In inside-out patch membranes both ATP (1 mM) and its non-hydrolyzable analog AMP-PNP (adenylylimidodiphosphate; EC₅₀=27M) reversibly inhibited channel activity. These results demonstrate that the K_ATP-channel in spiny neurons displays properties characteristic of the K_ATP-channel in hippocampal, neocortical and nigral neurons and in pancreatic ß-cells.     

Disruption of actin cytoskeleton attenuates sulfonylurea inhibition of cardiac ATP-sensitive K+ channels

 Two actin filament-depolymerizing agents, DNase I and cytochalasin D, were used to examine the involvement of the cytoskeleton in the functional interaction between the sulfonylurea receptor (SUR) and the ATP-sensitive K⁺ (K_ATP) channels. Isolated rat ventricular cardiomyocytes were studied using open cell-attached patches for single-channel recording. Bath application of DNase I (100 μg/ml) or cytochalasin D (10 μM) stimulated the K_ATP channel activities (in presence of 30 μM ATP), and these channels became resistant to inhibition by tolbutamide (0.5 mM). After exposure to tolbutamide, the relative NP_o value was 0.09 ± 0.02 in control patches in absence of actin disrupters, and 0.67 ± 0.22* or 0.65 ± 0.10*, respectively, in cells treated with DNase I or cytochalasine D (*P < 0.05 vs. control). The inhibitory action of glibenclamide (10 μM) on the K_ATP channels was also attenuated by DNase I. Thus, the disruption of the actin cytoskeleton attenuates the ability of SUR to inhibit the opening of K_ATP channels. Received: 3 February 1997 / Accepted: 25 March 1997     

ATP4− and ATP·Mg inhibit the ATP-sensitive K+ channel of rat ventricular myocytes

K⁺ currents were recorded from ATP-sensitive channels in inside-out membrane patches excised from isolated rat ventricular myocytes. ATP-sensitive K⁺ channel inhibition could be evoked by ATP in the absence of magnesium where most ATP would be present as the free acid ATP^4–. Channel inhibition was enhanced when the same total concentration of ATP was applied in the presence of magnesium, where most ATP would be bound as ATP·Mg. Dose-response relationships for ATP-sensitive K⁺ channel inhibition evoked by ATP had a Hill coefficient of 2 andK _i of 17 and 30 M for ATP in the presence and absence of magnesium respectively. This was the obverse of the expected results if ATP^4– were to be the sole form of ATP to effect channel closure. ATP-sensitive K⁺ channel inhibition evoked by ATPS, AMP-PNP and AMP-PCP was also enhanced in the presence of magnesium. It is concluded that the ATP-sensitive K⁺ channel of rat ventricular myocytes binds and is closed by both the free-acid and divalent-cationbound forms of ATP.     

Positive cooperativity in activation of the cardiac muscarinic K+ channel by intracellular GTP

Concentration-dependent effects of intracellular GTP on activation of the muscarinic K⁺ channel were examined in inside-out patches of cardiac atrial myocytes. The pipette solution contained 0.1 M ACh. GTP (0.01–30 M) and 0.5 mM MgCl₂ were applied to the inside side of the patch membrane. K⁺ channels were activated with GTP concentration above 0.1 M. Channel activation reached a maximal value with 1–3 M GTP. It decreased at GTP concentrations larger than 3 M, probably due to desensitization. The dependence of the open probability of the channel on intracellular GTP showed a sigmoidal relationship with a Hill coefficient of around 3. A positive cooperative effect of intracellular GTP on the K⁺ channel may play an important role in amplifying the signal from the membrane receptor to the K⁺ channel.     

Aromatic aldehydes and aromatic ketones open ATP-sensitive K+ channels in guinea-pig ventricular myocytes

Patch-clamp techniques were used to study the effects of three carbonyl compounds, 3,4-dihydroxy-benzaldehyde, 2,3-dihydroxybenzaldehyde, and 2,4-dihydroxy-acetophenone, on the adenosine-5-triphosphate(ATP)-sensitive K⁺ channel current (I _K.ATP) in guinea-pig ventricular myocytes. 3,4-Dihydroxybenzaldehyde (0.5–1 mM) shortened the action potential duration, and this effect was inhibited by application of a specific blocker of I _K.ATP, glibenclamide. The shortening of the action potential duration was shown to be caused by a time-independent outward current. In the cell-attached patch configuration, all three compounds activated a kind of single-channel current, which showed an inward rectification at positive potentials and which had a linear current/voltage relation at negative potentials, having a conductance of 90 pS. The current reversed at about 0 mV in symmetrical K⁺ concentrations on both sides of the membrane. In excised patches this current was blocked by internal application of ATP. Thus we identified this channel as I _K.ATP. The activation effects of two aromatic aldehydes were stronger than that of the aromatic ketone. The effect of these compounds on I _K.ATP was not reduced by addition of cysteine (10 mM). In inside-out patches, 3,4-dihydroxybenzaldehyde increased the activity of I _K.ATP, which had been blocked by 0.5 mM MgATP in the presence of 0.5 mM ADP, but the activation effect was variable and much weaker than that in the cell-attached configuration, and was completely eliminated in the absence of ADP. These results suggest that these compounds: (a) modulate I _K.ATP perhaps through an intracellular mechanism, (b) bind covalently to proteins to form a Schiff base which may by responsible for the effects, and (c) may require an ADP-dependent process.     

Calcitonin-gene-related peptide activates the muscarinic-gated K+ current in atrial cells

A high density of nerve fibers containing calcitonin-gene-related peptide (CGRP) is present in the atria. Recently CGRP was reported to open ATP-sensitive K channels in arterial smooth muscle cells. This study examines whether CGRP activates a similar K⁺ channel in cardiac cells. In voltage-clamped whole cells loaded with GTP and ATP, CGRP reversibly evoked an inwardly rectifying K⁺ current. To identify the K⁺ channel that gives rise to this current, three types of K⁺ channel (resting, ATP-sensitive and acetylcholine-activated) were examined. CGRP failed to activate or inhibit the ATP-sensitive or the resting K⁺ channel. However, CGRP (0.1–1 M) caused activation of single channels with kinetics similar to that of the muscarinic K⁺ channel (35–40 pS conductance and approx. 1 ms mean open time in symmetrical 140 mM K⁺). In excised, inside-out (CGRP in pipette) or in outside-out (GTP in pipette) patches, the K⁺ current was activated by perfusion with GTP or CGRP, respectively, suggesting that CGRP activated the muscarinic K⁺ channel via GTP-binding protein. Treatment with pertussis toxin inhibited the activation of the K⁺ channel, suggesting that CGRP receptor may be coupled to a G_i or a G_o type of GTP-binding protein. Together with previous findings, these results suggest that CGRP modulates several types of ion channels to produce its cellular effects.This work was supported in part by HL40586 and American Heart Association grant-in-aid.     

Fast and slow blockades of the inward-rectifier K+ channel by external divalent cations in guinea-pig cardiac myocytes

The present patch-clamp study shows that external Mg²⁺, Ca²⁺ and Sr²⁺ decrease the unit amplitude of inward current through the inward-rectifier K⁺ channel in a concentration-dependent manner. Sr²⁺ produces a voltage-dependent flickering block as well, and the fractional electrical distance between the external orifice and the Sr²⁺ binding site () is 0.73. The decrease of unit amplitude is reversible and voltage independent while it does not increase the noise level on the open-channel current. Unit current decreased by Mg²⁺ or Ca²⁺ has a longer mean open time, which is inversely proportional to the unit amplitude. External Mg²⁺ does not decrease the amplitude of unit outward current. A surface potential shift, measured using voltage-dependent Cs⁺ block (=1.60), failed to explain the current decrease. Therefore, we conclude that (1) the external divalent cations cause an extremely fast channel block, which appears as a decreased amplitude of the unit current on the recording system; (2) the blocking site (fast site) is present near the external orifice of the channel, and it is separate from the blocking site (slow site) to which Cs⁺ and Sr²⁺ bind.     

Role of intracellular Mg2+ in the activation of muscarinic K+ channel in cardiac atrial cell membrane

Effects of intracellular Mg²⁺ in the activation of a muscarinic K⁺ channel were examined in single atrial cells, using patch-recording techniques. In cell-attached patch recordings, acetylcholine (ACh) or adenosine (Ado), present in the pipette, activated a specific population of K⁺ channels. In inside-out patches, openings of the K⁺ channel by ACh or Ado diminished and did not resume until Mg²⁺ was added to the perfusate which contained GTP or GTP-S, a non-hydrolyzable GTP analogue. Channel openings caused by GTP faded by removing Mg²⁺, while GTP-S-induced openings persisted steadily even when both Mg²⁺ and GTP-S were removed. In contrast to the case of GTP-induced channel openings, the GTP-S-induced openings were not inhibited by the A protomer of pertussi toxin with NAD. From these observations, we concluded: 1) Intracellular Mg²⁺ is essential for GTP to activate the GTP-binding protein. 2) Deactivation of the N protein may be caused by hydrolysis of GTP to GDP. This process may not require Mg²⁺. 3) During the activation by GTP analogues, the N protein may be dissociated into its subunits.     

Pinacidil activates the ATP-sensitive K+ channel in inside-out and cell-attached patch membranes of guinea-pig ventricular myocytes

Patch-clamp techniques were used to study the effects of pinacidil on the adenosine-5-triphosphate (ATP)-sensitive K⁺ channel current in guinea-pig ventricular myocytes. In inside-out patches, the ATP-sensitive K⁺ channel current could be recorded at an internal ATP concentration of 0.5 mM or less and almost complete inhibition was achieved by raising the concentration to 2 mM. Application of pinacidil (10–30 M) in the presence of 2 mM ATP restored the current, whereas 5 mM ATP antagonized the effect of pinacidil. The conductance of the channel at symmetrical K⁺ concentrations of 140 mM was 75 pS with a slight inward rectification at voltages positive to +40 mV. There was no significant change in the conductance after application of pinacidil. In 0.5 mM ATP, at –80 mV, both the distributions of the open time and the life-time of bursts could be fitted by a single exponential. An increase in ATP concentration decreased the mean life-time of bursts, whereas pinacidil increased it with little increase in the mean open time. Closed time distributions of the channel were fitted by at least two exponentials, with a fast and a slow time constant. An increase in ATP concentration markedly increased the slow time constant associated with a decrease in the number of bursts, whereas the effect of pinacidil was opposite to that of increased ATP. These results indicate that pinacidil increases the open-state probability of the ATP-sensitive K⁺ channel. In cell-attached patches, application of pinacidil (100–200 M) to the extracellular solution reversibly induced the channel activity, which showed similar properties to those of the ATP-sensitive K⁺ channel recorded in cell-free patches.     

ATP-sensitive K+ channels in rat ventricular myocytes are blocked and inactivated by internal divalent cations

K⁺ currents were recorded from ATP-sensitive channels in inside-out patches from isolated rat ventricular myocytes. In the absence of internal divalent cations the current voltage relationship could be described by constant-field assumptions with a permeability of 1.25×10^–13 cm²/s; outward currents saturated under a high driving force for K⁺ movement. Internal 0.1–5.0 mM Mg²⁺, 0.1 M Ca²⁺ and 10 mM Na⁺ each depressed the flux of K⁺ ions moving outwards through open channels. Internal 0.1–5.0 mM Mg²⁺, 0.1–1.0 M Ca²⁺ and 1–10 M Ba²⁺ and Sr²⁺ blocked K⁺ channel activity in a dose-and voltage-dependent manner. Run-down channels could be reactivated by Mg-ATP, but not by AMP-PNP, ATPS or Mg-free ATP which suggested that phosphorylation of the channels was involved in their activity. Ca²⁺ (>=1 M) and Sr²⁺ (1 mM) markedly inactivated K⁺ ATP channels, millimolar Ba²⁺ or Mg²⁺ were less effective. This suggested that the run down of the channels was a Ca²⁺-dependent dephosphorylation of the K⁺ channel protein.     

A stretch-activated K+ channel in the basolateral membrane ofXenopus kidney proximal tubule cells

The present study examined whether a basolateral potassium ion (K⁺) channel is activated by membrane-stretching in the cell-attached patch. A K⁺ channel of conductance of 27.5 pS was most commonly observed in the basolateral membrane ofXenopus kidney proximal tubule cells. Channel activity increased with hyperpolarizing membrane potentials [at more positive pipette potentials (V _p)]. Open probability (P _o) was 0.03, 0.13, and 0.21 atV _p values of 0, 40, and 80 mV, respectively. Barium (0.1 mM) in the pipette reducedP _o by 79% at aV _p of 40 mV. Application of negative hydraulic pressure (−16 to −32 cm H₂O) to the pipette markedly activated outward currents (fromP _o=0.01 to 0.75) at aV _p of −80 mV, but not inward currents at aV _p of 80 mV. The size of the activated outward currents (from cell to pipette) did not change by replacing chloride with gluconate in the pipette. These results indicate that a stretch-activated K⁺ channel exists in the basolateral membrane of proximal tubule cells. It may play an important role as a K⁺ exit pathway when the cell membrane is stretched (for example, by cell swelling).     

High-conductance K+ channel in apical membranes of principal cells cultured from rabbit renal cortical collecting duct anlagen

Using the patch clamp technique, one type of K⁺ channel was identified in the apical cell membrane of cultured principal cells of rabbit renal collecting ducts in the cell-attached or excised-patch configuration. The channel was highly selective for K⁺ over Na⁺ (typically 30-70-fold) and had a conductance of 180, SD±39 pS (n=6), referred to a situation of 140 mmolar K⁺-Ringer solution present on either surface of the patch membrane. Channel activity was completely blocked by Ba²⁺ (5 mmol/l) and partially inhibited by Na⁺. The latter was evidenced by a deviation from Goldman rectification at high cytoplasm-positive membrane potentials, which was observed when Na⁺ competed with K⁺ for channel entrace from the cytoplasmic surface. Channel open probability depended strongly on membrane voltage and cytoplasmic Ca²⁺ concentration. Open-close kinetics exhibited double exponential behaviour, with a strong voltage dependence of the slow open time constant. Infrequently also a substate conductance level was identified. The voltage and calcium dependence suggest that the channel plays a role in adjusting K⁺ secretion to Na⁺ absorption in the fine regulation of cation excretion in renal collecting ducts.     

Blockade of dopamine storage,but not of dopamine synthesis,prevents activation of a tolbutamide-sensitive K+ channel in the guinea-pig substantia nigra

The substantia nigra has one of the highest levels of ATP-sensitive K⁺ channel in the brain. Since this channel is controlled by cell metabolism, the aim of this study was to see how closely it is associated with nigral dopamine systems, which are decreased in Parkinson's disease. In a sub-population of neurons within the rostral substantia nigra pars compacta of the guinea-pig, a brief period of hypoxia resulted in a tolbutamide (100–500 M) sensitive hyperpolarisation [input resistance (IR) decrease from 144.88±14.04 M pre-hypoxia to 105.91±13.25 M during hypoxia]. Maximal blockade of this decrease was seen in presence of 500 m tolbutamide [IR decrease only from 161.35±32.82 M to 155.02±34.29 M]. Reserpine (which depletes dopamine stores) but not -methyl-para-tyrosine (which decreases de novo synthesis of dopamine) caused a marked attenuation of this hyperpolarisation [IR decrease only from 163.32±44.42 M pre-hypoxia to 154.42±50.97 M during hypoxia]. This observation suggests that blockade of dopamine storage, but not of de novo synthesis, leads to a loss of responsiveness of certain mid-brain neurons to hypoxia, rendering them potentially more susceptible to subsequent degeneration. The possible link between nigral dopamine systems and ATP-sensitive K⁺ channels is discussed.     

Probing a Ca2+-activated K+ channel with quaternary ammonium ions

A series of quaternary amonium (QA) ions were used to probe the gross architecture of the ion conduction pathway in a Ca²⁺-activated K⁺ channel from rat muscle membrane. The channels were inserted into planar phospholipid membranes and the single channel currents were measured in the presence of the different QA ions. Internally applied monovalent QA ions (e.g. tetramethylammonium and analogues) induced a voltage-dependent blockade with a unique effective valence of the block equal to 0.30, and blocking potency increases as the compound is made more hydrophobic. Blockade is relieved by increasing the K⁺ concentration of the internal or external side of the channel. The effective valence of block is independent of K⁺ concentration. These results suggest that, from the internal side, all monovalent QA ions interact with a site located in the channel conduction system. Divalent QA ions of the type n-alkylbis-,-trimethylammonium (bisQn) applied internally also block the channel in a voltage dependent fashion. For short chains (bisQ2-bisQ5), the effective valence decreases with chain length from 0.41 to 0.27, it remains constant for bisQ5 to bisQ6 and increases up to 0.54 for bisQ10. This dependence of block with chain length implies that 27% of the voltage drop within the channel occurs over a distance of 1 nm. Externally applied monovalent QA ions also block the channel. The site is specific for tetraethylammonium; increasing or decreasing the side chains in one methylene group decrease potency by about 400-fold. It is concluded that the Ca²⁺-activated K⁺ channel has wide mouths located at each end and that they are different in molecular nature.     

Polymyxin B has multiple blocking actions on the ATP-sensitive potassium channel in insulin-secreting cells

The action of polymyxin B (0.1 M) on ATP-sensitive K⁺ (K⁺ _ATP) channels in RINm5F insulin-secreting cells was investigated by patch-clamp techniques. Using inside-out patches, open-cells and outside-out patches, polymyxin B was found to block K⁺ _ATP channels by, on average, approximately 90–95% of the initial control level of channel activity. The effects were rapid in onset, sustained and readily reversible. Similar effects were found in patches excised from cells pretreated overnight with 1 M of the phorbol ester phorbol myristate acetate (PMA). External block of channels was associated with a marked decrease in single-channel current amplitude, whereas these effects were not seen when polymyxin B was added to the inside face of the membrane. In patches bathed with internally applied ATP (0.5 mM) and ADP (0.5 mM), polymyxin B inhibited channels but its actions were not reversible upon removal of the compound. However, when the same protocol was undertaken upon cells pre-treated with PMA, the effects of polymyxin B were readily reversed. Our data suggests that polymyxin B is a novel modulator of K⁺ _ATP channels, exhibiting multiple blocking actions that may possibly involve a direct effect upon the channel and indirect effects mediated through the inhibition of endogenous protein kinase(s).To be considered as equal first author.     

Inhibition of voltage-dependent Na+ and K+ currents by forskolin in nodes of Ranvier

The effect of forskolin on voltage-activated Na⁺ and K⁺ currents in nodes of Ranvier from the toad, Bufo marinus, has been examined using the vaseline-gap voltageclamp technique. Peak Na⁺ currents (I _Na) were reduced by 35% and the rate of decline of Na⁺ current during continuous depolarization was accelerated following treatment with 450 M forskolin. However, the voltage-dependence of steady-state inactivation as well as the rate of recovery from fast inactivation remained unchanged. Upon repetitive depolarization at 1–10 Hz, a further inhibition of I _Na (60%) was observed. This use-dependent or phasic inhibition recovers slowly at -80 mV ( 13 s) and had a voltage-dependence like that of activation of the Na conductance. Near maximal steady-state phasic inhibition occurred with depolarizing pulse durations of only 4 ms, consistent with a direct involvement of the open Na⁺ channel in the blocking process. Inhibition of the delayed K⁺ current (I _K) was characterized by a concentration-dependent reduction in steady-state current amplitude (IC₅₀ 80 M) and a concentration-independent acceleration of current inactivation. A similar inhibition of I _K was obtained with 1,9-dideoxyforskolin, a homolog which does not activate adenylate cyclase (AC). The results suggest that the inhibition of I _K and perhaps I _Na follows directly from drug binding and is not a consequence of AC activation.     

Solubilisation and reconstitution of the rabbit skeletal muscle sarcoplasmic reticulum K+ channel into liposomes suitable for patch clamp studies

Sarcoplasmic reticulum (SR) membrane vesicles have been prepared from rabbit skeletal muscle and solubilised using K⁺ cholate. Solubilised membrane proteins were reconstituted into small asolectin liposomes by dialysis against cholate-free solution. Large liposomes were produced by freezing and thawing at –80°C and room temperature, respectively. The liposomes were assayed for the SR K⁺ channel using the patch clamp technique. Channel density was modulated by varying protein: lipid ratios during reconstitution. Channels inserted into the membrane with a preferred orientation. The solubilised and reconstituted channel behaves ohmically over the holding potential range ±70 mV and has a conductance of 178.4±4.4 pS (mean ± SE,n=37) in 200 mM KCl. The channel has a selectivity sequence of K⁺>NH ₄ ⁺ >Rb⁺>Na⁺ and K⁺ conductance is blocked by hexamethonium and decamethonium. The opening probability of the reconstituted channel is voltage dependent. The conductance and gating characteristics displayed by the solubilised and reconstituted channel correlate well with those previously observed following the fusion of native SR membrane vesicles with planar phospholipid bilayers.     

G-protein-mediated regulation of a Ca2+-dependent K+ channel in cultured vascular endothelial cells

The purpose of the present study was to determine the mechanism by which bradykinin activates the small conductance, inwardly rectifying, Ca²⁺-activated K⁺ channel (K_Ca) found in cultured bovine aortic endothelial cells. Channel activity was studied using the patch-clamp technique in whole-cell, cell-attached, inside-out and outside-out configurations. Channel conductance at potentials positive to 0 mV was 10±2 pS and at potentials negative to 0 mV 30±3 pS (n=7) when examined in symmetrical K⁺ (150 mmol/l) solutions. The channel open probability (P _o) was only weakly voltage dependent changing approximately 0.2 units over 160 mV. In contrast, raising the intracellular Ca²⁺ concentration from 100 nmol/l to 10 mol/l at –60 mV produced a graded increase in channel P _o from 0.15 to 0.96; the concentration required for half-maximum response (apparent K_0.5) was 719 nmol/l. At a constant Ca²⁺ concentration, application of guanosine triphosphate (GTP) to the cytoplasmic surface of the patch increased channel P _o. This effect was dependent upon the simultaneous presence of both GTP and Mg²⁺, and was reversed by the subsequent application of the guanosine diphosphate (GDP) analogue, guanosine-5-O-(2-thiodiphosphate) (GDPS). The hydrolysis-resistant GTP analogue, guanosine-5-O-(3-thiotriphosphate) (GTPS), induced a long-lasting increase in channel P _o. In the presence of Mg²⁺-GTP, the apparent K_0.5 for Ca²⁺ decreased from a control value of 722 nmol/l to 231 nmol/l. Addition of bradykinin to outside-out patches previously exposed to intracellular Mg²⁺-GTP further enhanced K_Ca activity, shifting the apparent K_0.5 for Ca²⁺ from 228 nmol/l to 107 nmol/l. This activation by bradykinin was not observed in patches following prior exposure to GDPS. These results suggest that bradykinin can activate the K_Ca channel of vascular endothelial cells via a G-protein-mediated change in the sensitivity of the channel for Ca²⁺. We postulate that vasoactive agonists may use this mechanism to maintain an elevated K⁺ permeability as the intracellular Ca²⁺ concentration returns towards normal resting levels.