The expression and function of Ca(2+)-sensing receptors in rat mesenteric artery; comparative studies using a model of type II diabetes.

BACKGROUND AND PURPOSE: The extracellular calcium-sensing receptor (CaR) in vascular endothelial cells activates endothelial intermediate-conductance, calcium-sensitive K(+) channels (IK(Ca)) indirectly leading to myocyte hyperpolarization. We determined whether CaR expression and function was modified in a rat model of type II diabetes. EXPERIMENTAL APPROACH: Pressure myography, western blotting, sharp microelectrode and K(+)-selective electrode recordings were used to investigate the functional expression of the CaR and IK(Ca) in rat mesenteric arteries. KEY RESULTS: Myocyte hyperpolarization to the CaR activator calindol was inhibited by Calhex 231. U46619-induced vessel contraction elevated the extracellular [K(+)] around the myocytes, and inhibition of this 'K(+) cloud' by iberiotoxin was needed to reveal calindol-induced vasodilatations. These were antagonized by Calhex 231 and significantly smaller in Zucker diabetic fatty rat (ZDF) vessels than in Zucker lean (ZL) controls. Myocyte hyperpolarizations to calindol were also smaller in ZDF than in ZL arteries. In ZDF vessels, endothelial cell CaR protein expression was reduced; IK(Ca) expression was also diminished, but IK(Ca)-generated hyperpolarizations mediated by 1-EBIO were unaffected. CONCLUSIONS AND IMPLICATIONS: The reduced CaR-mediated hyperpolarizing and vasodilator responses in ZDF arteries result from a decrease in CaR expression, rather than from a modification of IK(Ca) channels. Detection of CaR-mediated vasodilatation required the presence of iberiotoxin, suggesting a CaR contribution to vascular diameter, that is, inversely related to the degree of vasoconstriction. Compromise of the CaR pathway would favour the long-term development of a higher basal vascular tone and could contribute to the vascular complications associated with type II diabetes.

Effects of methyl beta-cyclodextrin on EDHF responses in pig and rat arteries; association between SK(Ca) channels and caveolin-rich domains.

BACKGROUND AND PURPOSE: The small and intermediate conductance, Ca2+-sensitive K+ channels (SK(Ca) and IK(Ca), respectively) which are pivotal in the EDHF pathway may be differentially activated. The importance of caveolae in the functioning of IK(Ca) and SK(Ca) channels was investigated. EXPERIMENTAL APPROACH: The effect of the caveolae-disrupting agent methyl-beta-cyclodextrin (MbetaCD) on IK(Ca) and SK(Ca) localization and function was determined. KEY RESULTS: EDHF-mediated, SK(Ca)-dependent myocyte hyperpolarizations evoked by acetylcholine in rat mesenteric arteries (following blockade of IK(Ca) with TRAM-34) were inhibited by MbetaCD. Hyperpolarizations evoked by direct SK(Ca) channel activation (using NS309 in the presence of TRAM-34) were also inhibited by MbetaCD, an effect reversed by cholesterol. In contrast, IK(Ca)-dependent hyperpolarizations (in the presence of apamin) were unaffected by MbetaCD. Similarly, in porcine coronary arteries, EDHF-mediated, SK(Ca)-dependent (but not IK(Ca)-dependent) endothelial cell hyperpolarizations evoked by substance P were inhibited by MbetaCD. In mesenteric artery homogenates subjected to sucrose-density centrifugation, caveolin-1 and SK3 (SK(Ca)) proteins but not IK1 (IK(Ca)) protein migrated to the buoyant, caveolin-rich fraction. MbetaCD pretreatment redistributed caveolin-1 and SK3 proteins into more dense fractions. In immunofluorescence images of porcine coronary artery endothelium, SK3 (but not IK1) and caveolin-1 were co-localized. Furthermore, caveolin-1 immunoprecipitates prepared from native porcine coronary artery endothelium contained SK3 but not IK1 protein. CONCLUSIONS AND IMPLICATIONS: These data provide strong evidence that endothelial cell SK(Ca) channels are located in caveolae while the IK(Ca) channels reside in a different membrane compartment. These studies reveal cellular organisation as a further complexity in the EDHF pathway signalling cascade.

Structure-activity relationships of K+ channel openers.

Seven groups of synthetic agent, distinguished by a combination of their chemical and pharmacological characteristics exert some or all of their effects by opening plasmalemmal K+ channels primarily in smooth muscle. Progress over the past two years now allows broad structure-activity relationships to be formulated within many of the individual groups of agent. Gillian Edwards and Arthur Weston review the historical basis of these discoveries and comment on the significance of new developments. They focus on the search for tissue and channel selectivity, two factors likely to be important for the successful clinical deployment of these substances as antihypertensive and bronchodilator agents.

Potassium channel openers and vascular smooth muscle relaxation.

Potassium channel openers comprise a diverse group of chemical agents which open plasma-lemmal K-channels. They show selectivity for smooth muscle, although K-channels in cardiac and skeletal muscle, neurones and the pancreatic beta-cell are also affected at relatively high concentrations. In addition, at least one endogenous K-channel opener of vascular origin--endothelium-derived hyperpolarizing factor--exists and in man plays a role in modulating blood vessel tone. The type of K-channel involved in the actions of both exogenous and endogenous K-channel openers is still uncertain, although a prime candidate in smooth muscle seems similar to the [ATPi]-modulated K-channel in the pancreatic beta-cell. This review focuses attention on the action of these agents in vascular smooth muscle and on the possible clinical exploitation of their powerful vasorelaxant properties.

Effects of isoprenaline and phenylephrine on energy-rich phosphate compounds and glucose-6-phosphate in smooth and cardiac muscle.

1. Tension changes produced by phenylephrine and isoprenaline have been examined in four tissues-the guinea-pig taenia coli, the longitudinal muscle strip of rabbit duodenum, the abdominal aorta of the rabbit and the rat heart.2. Changes in the amounts of adenosine triphosphate (ATP), creatine phosphate (CP) and glucose-6-phosphate (G-6-P) associated with the addition of phenylephrine and isoprenaline have been measured in the four tissues using enzymatic fluorimetric analysis.3. An alpha-adrenoceptor-mediated increase in tension (phenylephrine; rabbit aorta) was associated with no change in the amounts of ATP, CP or G-6-P.4. An alpha-adrenoceptor-mediated decrease in tension (phenylephrine; guineapig taenia coli and rabbit duodenum) was associated with no change in the amounts of ATP, CP or G-6-P.5. A beta-adrenoceptor-mediated increase in force of contraction (phenylephrine and isoprenaline; rat heart) was associated with a reduction in the amounts of ATP and CP and an increase in the amount of G-6-P.6. A beta-adrenoceptor-mediated decrease in tension (isoprenaline; guinea-pig taenia coli and rabbit duodenum) was associated with an increase in the amounts of ATP and CP and no change in the amount of G-6-P.7. beta-Adrenoceptor-mediated metabolic changes were antagonized by propranolol.

The effects of isoprenaline and phenylephrine on oxygen consumption in isolated smooth muscle.

1. Simultaneous measurements of oxygen consumption and mechanical activity have been made in isolated preparations of guinea-pig taenia coli, rabbit abdominal aorta and the longitudinal muscle of rabbit duodenum.2. Mean resting oxygen consumption was greater in mechanically-active preparations (taenia coli and duodenum) than in the aorta which showed no spontaneous activity.3. The relaxation produced by isoprenaline in guinea-pig taenia coli and rabbit duodenum was accompanied by a dose-dependent fall in oxygen consumption.4. The relaxation produced by phenylephrine in guinea-pig taenia coli was accompanied by a dose-dependent fall in oxygen consumption whereas, in the rabbit duodenum, the transient fall in tension was not accompanied by any significant change in oxygen consumption.5. The increase in tension produced by phenylephrine in rabbit abdominal aorta was accompanied by an increase in oxygen consumption.6. Changes in mechanical activity and oxygen consumption produced by isoprenaline and phenylephrine were antagonized by propranolol and phentolamine respectively.7. It is concluded that the variations in oxygen consumption associated with spontaneous or drug-induced mechanical changes are simple reflections of altered mechanical activity. The possibility of additional direct metabolic actions of isoprenaline and phenylephrine is discussed.

Nerve-mediated inhibition of mechanical activity in rabbit duodenum and the effects of desensitization to adenosine and several of its derivatives.

1. Inhibition of mechanical activity in longitudinal muscle strips of rabbit duodenum was induced by perivascular and intramural nerve stimulation.2. The effects of perivascular stimulation were abolished by phentolamine + propranolol, guanethidine, reserpine and by tetrodotoxin. The effects of intramural stimulation were abolished only by tetrodotoxin.3. Noradrenaline, phenylephrine, isoprenaline, adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine monophosphate (AMP) and adenosine each produced an inhibition of mechanical activity. The relative potencies of these agonists were noradrenaline>isoprenaline>phenylephrine>ATP>ADP>adenosine>AMP.4. Exposure of tissues to high concentrations of either ATP or adenosine desensitized the tissue to further exposure to ATP, ADP, AMP and adenosine, whilst the inhibitory effects of noradrenaline, phenylephrine and isoprenaline and of perivascular and intramural stimulation were unaffected. The effects of desensitization were reversible.5. It was concluded that the effects of intramural stimulation were mediated neither by noradrenaline nor by adenosine or its derivatives.

The effect of desensitization to adenosine triphosphate on the peristaltic reflex in guinea-pig ileum.

Contractions of the longitudinal muscle of guinea-pig ileum were inhibited by adenosine triphosphate. This inhibition was prevented by specific desensitization of the tissue to adenosine triphosphate whilst the peristaltic reflex was unaffected. The results suggest that the transmitter substance released from the intramural inhibitory neurones activated during peristalsis is not adenosine triphosphate.

Induction of a glibenclamide-sensitive K-current by modification of a delayed rectifier channel in rat portal vein in insulinoma cells.

In insulinoma cells (RINm5F), the glibenclamide-sensitive K-current (IK(ATP)) which developed spontaneously or after exposure to levcromakalim or to butanedione monoxime was always accompanied by a reduction in the delayed rectifier current (IK(V)). At potentials over which IK(V) was fully activated, the total outward current remained constant. In rat portal vein, the delayed rectifier channel inhibitor, margatoxin, reduced the combined induction of IK(ATP) and inhibition of IK(V) by levcromakalim. These data suggest that the ATP-sensitive K-channel, K(ATP), is a voltage-insensitive state of the delayed rectifier, KV.

Effects of rubidium on responses to potassium channel openers in rat isolated aorta.

1. In a physiological salt solution (PSS) in which potassium (K) was replaced by rubidium (Rb), segments of rat aorta precontracted with 20 mM RbCl were fully relaxed by K-channel openers with an order of potency levcromakalim > cromakalim > aprikalim > RP 49356. These relaxations were inhibited by glibenclamide. 2. Segments of rat aorta bathed in normal PSS and precontracted with 20 mM KCl were also relaxed by these K-channel openers with an order of potency levcromakalim > cromakalim > aprikalim > RP 49356. These relaxations were glibenclamide-sensitive. However, the absolute potencies of the K-channel openers were approximately four times greater in normal PSS than in RbPSS. 3. In RbPSS, minoxidil sulphate relaxed segments of aorta precontracted with 20 mM RbCl by approximately 20% whereas in normal PSS it fully relaxed those contracted with 20 mM KCl. 4. In RbPSS, levcromakalim-induced relaxation of aortic segments precontracted with 20 mM RbCl was initially well-maintained but then faded by approximately 60% of the initial relaxation to a new, stable level. Subsequent exposure to RP 49356 or to higher concentrations of levcromakalim was without further relaxant effect. Similar changes were observed when RP 49356 was the initial relaxant and tissues were exposed to either RP 49356 or levcromakalim. In normal PSS, levcromakalim- or RP 49356-induced relaxation of contractions produced by 20 mM KCl was well-maintained. 5. In RbPSS, minoxidil sulphate-induced relaxation of aortic segments precontracted with 20 mM RbCl was well-maintained. Subsequent exposure to either levcromakalim or to RP 49356 produced further tissue relaxation. 6. In RbPSS, levcromakalim produced no detectable increase in either 86Rb- or 42K-efflux from rat aortic strips. In normal PSS, a significant increase in the exchange of both isotopes was detected.7. Levcromakalim hyperpolarized segments of rat aorta bathed both in normal PSS and after depolarization by the addition of 20 mM KCI. Exposure to RbPSS depolarized the tissue and under these conditions, levcromakalim had no effect on membrane potential.8. In Rb- and normal PSS, levcromakalim produced a similar degree of inhibition of the refilling of then or adrenaline-sensitive Ca store.9. It is concluded that millimolar concentrations of Rb inhibit the plasmalemmal ATP-sensitive K-channels (KATP) which are the target of the K-channel openers. The relaxant actions of the K-channel openers in both normal and Rb-PSS and the inhibition of these effects by glibenclamide may reflect a functional interaction between these agents at ATP-binding sites associated with both KATP and with intracellular structures including Ca stores.

Characteristics of the contractile response of rabbit aorta produced by cromakalim in calcium-free solution.

1 The effect of potassium channel opening compounds has been investigated in the smooth muscle of rabbit aorta under Ca-free conditions. Examination of the characteristics of the response has been performed using cromakalim as the prototype compound. 2 In order of potency, Ro 31-6930, cromakalim, minoxidil sulphate and pinacidil each produced a contraction in rabbit aortic strips bathed in Ca-free MOPS-buffered physiological salt solution (PSS). In contrast, forskolin, glyceryl trinitrate and nifedipine each failed to increase tension under identical conditions. Cromakalim also evoked contraction of bovine trachealis muscle bathed in Ca-free PSS. 3. The contractile response to cromakalim in rabbit aortic strips was of delayed onset (15-20 min) and reached a plateau after approximately 120 min (1.8 g maximum with 1 microM cromakalim). No cromakalim-induced tension changes were observed in either 1 mM or 2.5 mM Ca-containing PSS. 4. Raising the [KCl] of the Ca-free PSS to 65.9 mM fully inhibited the cromakalim-induced contraction in rabbit aortic strips. In addition, pretreatment of aortic strips with the sulphonylurea glibenclamide antagonized the subsequent mechanical response to cromakalim. 5. In Ca-free PSS, cromakalim (1 microM) stimulated 42K-efflux with a time-course corresponding to the contractile event. Glibenclamide (1 microM) inhibited this cromakalim-induced 42K-efflux. 6. In sharp microelectrode studies in bovine trachealis, cromakalim (10 microM) produced a sustained membrane hyperpolarization in normal PSS. In contrast, the cromakalim-induced hyperpolarization in Ca-free PSS was not sustained. The fading of the hyperpolarization was temporally correlated with the increase in tension under these experimental conditions. 7. It is concluded that the K-channel opener-induced smooth muscle contractile response revealed in Ca-free PSS is the consequence of K-channel opening. The nature of the detailed mechanism which underlies this contractile phenomenon remains to be determined.

Investigation of the spasmogenic effect of manganese on the guineapig isolated ileum preparation.

1. The mechanism of the manganese-induced spasm of the guinea-pig ileum was investigated using agents known to modify nerve function. The spasm was reduced by cooling, tetrodotoxin, procaine, Botulinus toxin (Type A), hyoscine and pempidine. It was potentiated by mipafox.2. In the presence of manganese, the release of acetylcholine from the ileum was greatly increased.3. Tetrodotoxin prevented the manganese-induced increase in acetylcholine output from the ileum but had no significant effect on the spontaneous acetylcholine output.4. It is suggested that the manganese-induced spasm of the ileum results from an action on intramural cholinergic nerves.

Theophylline antagonizes some effects of purines in the intestine but not those of intramural inhibitory nerve stimulation.

1 Hyoscine- and guanethidine-treated preparations of longitudinal muscle of rabbit duodenum, guinea-pig taenia caeci and fundic strip relaxed when exposed to noradrenaline, adenosine, adenosine triphosphate (ATP) or to field stimulation of their intramural nerves. 2 In guinea-pig taenia caeci and fundus, theophylline 100 mumol/l had no effect on responses to noradrenaline, adenosine, ATP and intramural nerve stimulation. 3 In rabbit duodenum, theophylline 100 mumol/l antagonized some responses to adenosine but had no effect on responses to noradrenaline, ATP and intramural nerve stimulation. 4 Theophylline 1 mmol/l itself relaxed the intestinal tissues and in the fundic strip and taenia caeci, these relaxant effects were associated with abolition of spike activity and cellular hyperpolarization. In the taenia caeci, the amplitude of inhibitory post-junctional potentials was reduced. 5 Theophylline 1 mmol/l antagonized the twitch suppression produced by adenosine and ATP in the transmurally-stimulated guinea-pig ileum but not that evoked by noradrenaline. 6 It is concluded that theophylline can selectively antagonize some actions of purines in the intestine but that it does not specifically antagonize the effect of intramural inhibitory nerve stimulation.

The accumulation of adenosine in rabbit intestinal muscle.

1 Strips of longitudinal muscle from rabbit intestine accumulated radioactivity when exposed to [(3)H]-adenosine.2 Accumulation of radioactivity was not sodium-dependent or ouabain-sensitive, but was reduced by cooling, zero glucose plus bubbling with N(2), 2,4,dinitrophenol, dipyridamole, hexobendine and lidoflazine.3 After 7 min exposure to [(3)H]-adenosine, the tissue was found to contain radioactive adenosine triphosphate (ATP), adenosine diphosphate (ADP), adenosine monophosphate (AMP) and adenosine itself in the approximate ratio 13: 6: 4: 1.4 In the presence of dipyridamole, hexobendine or lidoflazine (each 1 muM), the amounts of radioactive ATP, ADP, AMP and adenosine were reduced with the concentration of adenosine not significantly different from controls.5 It is concluded that energy-dependent uptake of adenosine does not occur in the longitudinal muscle of rabbit intestine. Adenosine enters the tissue by a passive process and rapidly becomes phosphorylated giving rise to apparently high tissue: medium ratios.6 The drugs dipyridamole, hexobendine and lidoflazine appear to reduce the accumulation of radioactivity by preventing the formation of adenosine phosphate derivatives.

Some effects of dipyridamole, hexobendine and lidoflazine on inhibitory processes in rabbit duodenum.

Some of the effects of dipyridamole, hexobendine and lidoflazine on mechanical responses in rabbit duodenum have been investigated. In concentrations known to inhibit tissue accumulation of adenosine and its metabolites, none of these agents potentiated inhibitory responses to intramural nerve stimulation or to application of adenosine, adenosine triphosphate or phenylephrine. These results neither support nor dispute the suggestion that adenosine or a related nucleotide is the intramural inhibitory transmitter but do show that tissue accumulation in rabbit duodenum is not an important feature in the termination of the action of adenosine.

Some effects of sodium nitroprusside, methoxyverapamil (D600) and nifedipine on rat portal vein.

1 The effects of methoxyverapamil (D600), nifedipine and sodium nitroprusside on noradrenaline-induced electrical and mechanical activity in rat portal vein have been examined. 2 D600 and nifedipine produced a concentration-dependent reduction in mechanical responses to noradrenaline whilst sodium nitroprusside had no effect. The effects of D600 and nifedipine were reversed by increasing the extracellular calcium concentration. 3 The mechano-inhibitory effects of D600 were accompanied by a marked reduction in electrical activity with some evidence of electromechanical uncoupling. 4 The mechano-inhibitory effects of nifedipine were associated with considerable electromechanical uncoupling. 5 It is concluded that in the concentrations used, D600 acts primarily by inhibiting calcium influx with some effect on electromechanical coupling whilst nifedipine interferes mainly with the coupling process. The inactivity of sodium nitroprusside suggests that the excitation-contraction coupling process in rat portal vein is relatively simple and further studies with this tissue seem indicated.

Adrenoceptors in the guinea-pig colon.

Guinea-pig colon was shown to contain both alpha- and beta-adrenoceptors. Isoprenaline-induced relaxation was mediated by beta-adrenoceptors alone whilst that produced by phenylephrine was mediated by both alpha- and beta- adrenoceptors.

Effect of apamin on responses to BRL 34915, nicorandil and other relaxants in the guinea-pig taenia caeci.

BRL 34915 (4-64 X 10(-7) M), isoprenaline (0.5-32 X 10(-8) M) and nicorandil (4-64 X 10(-6) M) produced a slowly-developing relaxation of spontaneous tone of the guinea-pig taenia caeci; with no after-contraction on washout. These inhibitory responses were unaffected by apamin (10(-7) M). Adenosine triphosphate (0.06-2 X 10(-3) M) and noradrenaline (1-16 X 10(-7) M) produced a rapid inhibition of spontaneous tone with a prominent after-contraction, especially on washout. Both the inhibitory effect and the rebound contraction were abolished by apamin (10(-7) M). Exposure to both BRL 34915 (64 X 10(-7) M) and to nicorandil (64 X 10(-6) M) produced an increase in the 86Rb efflux rate coefficient which was unaffected by apamin (10(-7) M). Exposure to isoprenaline (32 X 10(-8) M) had no effect on the 86Rb efflux rate coefficient. Exposure to noradrenaline (16 X 10(-7) M) produced an increase in the 86Rb efflux rate coefficient which was abolished by apamin (10(-7) M). The results confirm that both BRL 34915 and nicorandil are capable of opening potassium channels in smooth muscle but show that the channel is not apamin-sensitive.

The effects of BRL 34915 and nicorandil on electrical and mechanical activity and on 86Rb efflux in rat blood vessels.

The effects of the antihypertensive agent BRL 34915 on a variety of responses of the aorta and portal vein of the rat have been compared with those of nicorandil. On portal vein, BRL 34915 (0.01-50 X 10(-6) M) and nicorandil (0.1-500 X 10(-6) M) abolished spontaneous mechanical activity and reduced mechanical responses to noradrenaline (0.1-100 X 10(-6) M) and K+ (5-20 X 10(-3) M) but had little inhibitory effect on responses to K+ (40-80 X 10(-3) M). The onset of the reduced responses to noradrenaline was delayed by both agents. On portal vein, BRL 34915 (0.1-50 X 10(-6) M) and nicorandil (0.5-500 X 10(-6) M) abolished spontaneous electrical and mechanical activity, hyperpolarized the smooth muscle cells to a value close to their calculated potassium equilibrium potential and increased the 86Rb efflux rate coefficient. On aorta, BRL 34915 (0.2-0.8 X 10(-6) M) and nicorandil (8-32 X 10(-6) M) reduced mechanical responses to noradrenaline (0.001-1 X 10(-6) M) and K+ (5-20 X 10(-3) M) but had little inhibitory effect on responses to K+ (40-80 X 10(-3) M). On aorta, BRL 34915 (0.2-0.8 X 10(-6) M) increased the 86Rb efflux rate coefficient whereas nicorandil (8-32 X 10(-6) M) was without effect. It is concluded that the inhibitory actions of BRL 34915 on both aorta and portal vein result from the opening of membrane potassium channels. The resulting membrane shunt inhibits the effects of excitatory agents. The inhibitory effects of nicorandil result from a combination of the opening of potassium channels together with an additional, undefined action.

The contribution of Rb-permeable potassium channels to the relaxant and membrane hyperpolarizing actions of cromakalim, RP49356 and diazoxide in bovine tracheal smooth muscle.

1. Cromakalim (1 and 10 microM), RP49356 (5 and 50 microM) and diazoxide (100 and 300 microM) produced full relaxation of smooth muscle strips pre-contracted with 25 mM KCl. These agents caused membrane hyperpolarization and increased 42K and 86Rb efflux. The time taken to achieve the maximum change in each of these parameters (tmax) was less for the higher concentration levels of cromakalim, RP49356 and diazoxide than for the lower concentration levels. 2. Calculation of permeability (P) changes showed that cormakalim (1 and 10 microM) produced a greater rise in PK than PRb, although the PRb:PK ratio was similar at both concentration levels. Similarly RP49356 produced a greater change in PK than PRb. However, in contrast to cromakalim, this difference was more marked at the higher concentration (50 microM) and was reflected by a differential effect of the two concentrations of RP49356 on the PRb:PK ratio. Diazoxide (100 and 300 microM) produced similar changes in PK and PRb. 3. For cromakalim (1 and 10 microM) the tmax for the electrical and mechanical effects and also the profile of change in these parameters corresponded to changes in both PK and PRb. For RP49356 (5 microM), changes in tension and membrane potential were related to both changes in PK and PRb, whereas at 50 microM these responses more closely corresponded to changes in PK. For diazoxide (100 and 300 microM) the electrical and mechanical effects corresponded to changes in both PK and PRb. 4. The results show that changes in 42K and 86Rb efflux induced by cromakalim, RP49356 and diazoxide are good indicators of changes in membrane PK and PRb evoked by these agents. Furthermore, it is concluded that the K channels involved in the mechanical and electrical effects of cromakalim are represented by the opening of a single population through which Rb can pass less easily than K, whilst the K channels associated with actions of diazoxide are equally permeable to both K and Rb. In contrast, the relaxant and membrane hyperpolarizing actions of RP49356 may involve the opening of more than one group of K channels which differ in their permeability to Rb.