Frances Plane

Endothelium-dependent hyperpolarization: a role in the control of vascular tone

Endothelial-dependent relaxation of vascular smooth muscle cells evoked by a number of agonists, including cholinomimetics and substance P, is often accompanied by an increase (repolarization and/or hyperpolarization) in the membrane potential. This change in membrane potential appears predominantly to reflect the action of an endothelial-derived hyperpolarizing factor (EDHF), which is distinct from NO (or endothelial-derived relaxing factor), and is discussed in this article by Chris Garland and colleagues. In large conducting arteries, EDHF may provide a secondary system to NO, which assumes primary importance in some disease states such as pulmonary hypertension and atherosclerosis. However, in small resistance arteries (100-300 microns), EDHF appears to be a major determinant of vascular calibre under normal conditions, and may therefore be of primary importance in the regulation of vascular resistance.

Charybdotoxin and apamin block EDHF in rat mesenteric artery if selectively applied to the endothelium

In rat mesenteric artery, endothelium-derived hyperpolarizing factor (EDHF) is blocked by a combination of apamin and charybdotoxin (ChTX). The site of action of these toxins has not been established. We compared the effects of ChTX and apamin applied selectively to the endothelium and to the smooth muscle. In isometrically mounted arteries, ACh (0.01-10 μm), in the presence of indomethacin (2.8 μM) and N ω-nitro-l-arginine methyl ester (l-NAME) (100 μM), concentration dependently relaxed phenylephrine (PE)-stimulated tone (EC50 50 nM; n = 10). Apamin (50 nM) and ChTX (50 nM) abolished this relaxation ( n = 5). In pressurized arteries, ACh (10 μM), applied intraluminally in the presence of indomethacin (2.8 μM) andl-NAME (100 μM), dilated both PE-stimulated (0.3-0.5 μM; n = 5) and myogenic tone ( n = 3). Apamin (50 nM ) and ChTX (50 nM) applied intraluminally abolished ACh-induced dilatations. Bath superperfusion of apamin and ChTX did not affect ACh-induced dilatations of either PE-stimulated ( n = 5) or myogenic tone ( n = 3). This is the first demonstration that ChTX and apamin act selectively on the endothelium to block EDHF-mediated relaxation.In rat mesenteric artery, endothelium-derived hyperpolarizing factor (EDHF) is blocked by a combination of apamin and charybdotoxin (ChTX). The site of action of these toxins has not been established. We compared the effects of ChTX and apamin applied selectively to the endothelium and to the smooth muscle. In isometrically mounted arteries, ACh (0.01-10 micrometers), in the presence of indomethacin (2.8 microM) and Nomega-nitro-L-arginine methyl ester (L-NAME) (100 microM), concentration dependently relaxed phenylephrine (PE)-stimulated tone (EC50 50 nM; n = 10). Apamin (50 nM) and ChTX (50 nM) abolished this relaxation (n = 5). In pressurized arteries, ACh (10 microM), applied intraluminally in the presence of indomethacin (2.8 microM) and L-NAME (100 microM), dilated both PE-stimulated (0.3-0.5 microM; n = 5) and myogenic tone (n = 3). Apamin (50 nM ) and ChTX (50 nM) applied intraluminally abolished ACh-induced dilatations. Bath superperfusion of apamin and ChTX did not affect ACh-induced dilatations of either PE-stimulated (n = 5) or myogenic tone (n = 3). This is the first demonstration that ChTX and apamin act selectively on the endothelium to block EDHF-mediated relaxation.

Evidence that anandamide and EDHF act via different mechanisms in rat isolated mesenteric arteries

The endogenous cannabinoid, anandamide, has been suggested as an endothelium‐derived hyperpolarizing factor (EDHF). We found that anandamide‐evoked relaxation in isolated segments of rat mesenteric artery was associated with smooth muscle hyperpolarization. However, although anandamide‐evoked relaxation was inhibited by either charybdotoxin (ChTX) or iberiotoxin, inhibition of the relaxation to EDHF required a combination of ChTX and apamin. The relaxations induced by either anandamide or EDHF were not inhibited by the cannabinoid receptor (CB1) antagonist SRI41716A, or mimicked by selective CB1 agonists. Thus, anandamide appears to cause smooth muscle relaxation via a CB1 receptor‐independent mechanism and cannabinoid receptor activation apparently does not contribute to EDHF‐mediated relaxation in this resistance artery.

Evidence that different mechanisms underlie smooth muscle relaxation to nitric oxide and nitric oxide donors in the rabbit isolated carotid artery

The endothelium‐dependent relaxants acetylcholine (ACh; 0.03–10 μm) and A23187 (0.03–10 μm), and nitric oxide (NO), applied either as authentic NO (0.01–10 μm) or as the NO donors 3‐morpholino‐sydnonimine (SIN‐1; 0.1–10 μm) and S‐nitroso‐N‐acetylpenicillamine (SNAP; 0.1–10 μm), each evoked concentration‐dependent relaxation in phenylephrine stimulated (1–3 μm; mean contraction and depolarization, 45.8±5.3 mV and 31.5±3.3 mN; n=10) segments of rabbit isolated carotid artery. In each case, relaxation closely correlated with repolarization of the smooth muscle membrane potential and stimulated a maximal reversal of around 95% and 98% of the phenylephrine‐induced depolarization and contraction, respectively. In tissues stimulated with 30 mm KCl rather than phenylephrine, smooth muscle hyperpolarization and relaxation to ACh, A23187, authentic NO and the NO donors were dissociated. Whereas the hyperpolarization was reduced by 75–80% to around a total of 10 mV, relaxation was only inhibited by 35% (n=4–7 in each case; P<0.01). The responses which persisted to ACh and A23187 in the presence of 30 mm KCl were abolished by either the NO synthase inhibitor l‐NG‐nitroarginine methyl ester (l‐NAME; 100 μm) or the inhibitor of soluble guanylyl cyclase 1H‐[1,2,4]oxadiazolo[4,3‐a]quinoxalin‐1‐one (ODQ; 10 μm; 10 min; n=4 in each case; P<0.01). Exposure to ODQ significantly attenuated both repolarization and relaxation to ACh, A23187 and authentic NO, reducing the maximum changes in both membrane potential and tension to each relaxant to around 60% of control values (n=4 in each case; P<0.01). In contrast, ODQ almost completely inhibited repolarization and relaxation to SIN‐1 and SNAP, reducing the maximum responses to around 8% in each case (n=3–5; P<0.01). The potassium channel blockers glibenclamide (10 μm), iberiotoxin (100 nm) and apamin (50 nm), alone or in combination, had no significant effect on relaxation to ACh, A23187, authentic NO, or the NO donors SIN‐1 and SNAP (n=4 in each case; P>0.05). Charybdotoxin (ChTX; 50 nm) almost abolished repolarization to ACh (n=4; P<0.01) and inhibited the maximum relaxation to ACh, A23187 and authentic NO each by 30% (n=4–8; P<0.01). Application of ODQ (10 μm; 10 min) abolished the ChTX‐insensitive responses to ACh, A23187 and authentic NO (n=4 in each case; P<0.01 When the concentration of phenylephrine was reduced (to 0.3–0.5 μm) to ensure the level of smooth muscle contraction was the same as in the absence of potassium channel blocker, ChTX had no effect on the subsequent relaxation to SIN‐1 (n=4; P>0.05). However, in the presence of tone induced by 1–3 μm phenylephrine (51.2±3.3 mN; n=4), ChTX significantly reduced relaxation to SIN‐1 by nearly 50% (maximum relaxation 53.2±6.3%, n=4; P<0.01). These data indicate that NO‐evoked relaxation of the rabbit isolated carotid artery can be mediated by three distinct mechanisms: (a) a cyclic GMP‐dependent, voltage‐independent pathway, (b) cyclic GMP‐mediated smooth muscle repolarization and (c) cyclic GMP‐independent, ChTX‐sensitive smooth muscle repolarization. Relaxation and repolarization to both authentic and endothelium‐derived NO in this large conduit artery appear to be mediated by parallel cyclic GMP‐dependent and ‐independent pathways. In contrast, relaxation to the NO‐donors SIN‐1 and SNAP appears to be mediated entirely via cyclic GMP‐dependent mechanisms.

Investigation of the inhibitory effects of homocysteine and copper on nitric oxide‐mediated relaxation of rat isolated aorta

Elevated plasma levels of homocysteine (HC) and copper have both been associated with the development of inflammatory vascular diseases, such as atherosclerosis. In this study, the effects of a combination of HC and copper on nitric oxide (NO)‐mediated relaxation of isolated rat aortic rings were investigated. Exposure to HC (10–100 μM; 30 min) had no effect on relaxation to acetylcholine (ACh; 0.01–10 μM, n=4). Pre‐incubation of aortic rings with a higher concentration of HC for an extended period (1 mM; 180 min) significantly inhibited endothelium‐dependent relaxation (n=4), but this inhibition was prevented by the presence of the copper chelator bathocuprione (10 μM, 180 min, n=6). Exposure to HC (100 μM) and copper (10–100 μM; 30 min) caused a copper concentration‐dependent inhibition of endothelium‐dependent relaxation (n=4). This inhibitory effect was reduced in the presence of either superoxide dismutase (SOD; 100 u ml−1; n=4) or catalase (100 u ml−1; n=4), and further reduced by the presence of both enzymes (n=5). HC and copper (100 μM; 30 min) significantly inhibited endothelium‐independent relaxation to glyceryl trinitrate (0.01–10 μM; n=8). In contrast, HC (1 mM), alone or in combination with copper (100 μM), did not inhibit relaxation to the endothelium‐independent relaxant sodium nitroprusside (0.01–10 μM; n=4). These data indicate that the presence of copper greatly enhances the inhibitory actions of HC on NO‐mediated relaxation of isolated aortic rings. The reduction of inhibition by catalase and SOD indicates a possible role for copper‐catalyzed generation of superoxide and hydrogen peroxide leading to an increased inactivation or decreased production of endothelium‐derived NO.

Differential effects of acetylcholine, nitric oxide and levcromakalim on smooth muscle membrane potential and tone in the rabbit basilar artery

1 Endothelium‐dependent hyperpolarization of smooth muscle cells in isolated, pre‐contracted segments of rabbit basilar artery in response to acetycholine (100 μm) was abolished in the presence of glibenclamide (10 μm). 2 Acetylcholine‐evoked relaxation was unaffected by either glibenclamide or 65 mm potassium chloride, indicating that the change in membrane potential did not form an essential component of relaxation and that high concentrations of potassium did not inhibit the release or action of endothelium‐derived relaxing factor in this vessel. 3 Saturated solutions of nitric oxide (NO) gas in solution (150 μm), which evoked maximal relaxation of arterial segments pre‐contracted and depolarized by noradrenaline (10–100 μm), did not alter the membrane potential of either unstimulated or depolarized smooth muscle cells. 4 The potassium channel opener levcromakalim, evoked concentration‐dependent relaxation and hyperpolarization in pre‐constricted smooth muscle cells. The threshold concentrations for hyperpolarization and relaxation, the EC50 values and the maximally effective concentration of levcromakalim (around 30 nm, 150 nm and 10 μm, respectively) were not significantly different, and both components of the response were inhibited by glibenclamide (10 μm), indicating a close coupling between the two responses. 5 In the presence of 65 mm potassium chloride, the hyperpolarization to levcromakalim was abolished, while a small relaxation (25 ± 4%) persisted, indicating an additional mechanism for relaxation to this agent. 6 These results show that different mechanisms underlie the relaxant action of potassium channel openers, NO and endothelium‐derived factors in cerebral arteries and provide further evidence that in the basilar artery, in contrast to some other vessels, endothelium‐dependent hyperpolarization to acetylcholine is not important for smooth muscle relaxation.

The relative importance of nitric oxide and nitric oxide-independent mechanisms in acetylcholine-evoked dilatation of the rat mesenteric bed

1 The relative contribution of nitric oxide (NO) to acetylcholine‐induced smooth muscle relaxation was investigated in the rat perfused mesenteric vasculature and in isolated segments of second, third and fourth order arterial branches. 2 The EC50 values and maximal relaxation to acetylcholine were not significantly different in the sequential arterial branches, being approximately 0.05 μm and 85%, respectively. 3 The NO synthase inhibitor L‐NG‐nitro‐L‐arginine methyl ester (L‐NAME; 100 μm) reduced acetylcholine‐evoked endothelium‐dependent dilatation and relaxation in the perfused mesenteric bed and in isolated arterial segments. The maximum response to acetylcholine in both preparations was reduced by between 35% to 40% while the EC50 values were increased by 5–6 fold. L‐NAME had no effect on basal smooth muscle tone in either case. 4 In contrast, endothelium‐dependent dilatation of the perfused mesenteric bed evoked by A23187 (0.002–20 nmol), was unaffected by exposure to L‐NAME. The EC50 values and maximal responses elicited by A23187 (20 nmol) before and after exposure to L‐NAME were 0.96 ± 0.5 nmol and 67.0 ± 7.0% (n = 4), and 0.7 ± 0.4 nmol and 70.0 ± 5.0% (n = 4; P > 0.01), respectively. 5 Perfusion of the isolated mesenteric bed with raised K+‐Krebs buffer (25 μm) had no effect on basal tone, but reduced the amplitude of both acetylcholine‐ and A23187‐evoked dilatation. The maximum responses to acetylcholine (2 μmol) and A23187 (20 nmol) were reduced from 67.5 ± 7.3% and 65.4 ± 8.2% to 18.9 ± 11.0% (n = 5; P < 0.01) and 13.5 ± 12.0% (n = 4; P < 0.01), respectively. 6 Exposure of the mesenteric bed to L‐NAME in the presence of raised K+‐Krebs further reduced the maximal response elicited by acetylcholine to only 8.9 ± 2.8% (n = 4; P < 0.01). 7 These results indicate that acetylcholine‐evoked vasodilatation of the rat mesenteric vasculature is mediated by both NO‐dependent and ‐independent mechanisms. The relative contribution made by these mechanisms does not appear to differ in sequential branches of the mesenteric artery. In contrast, A23187‐evoked vasodilatation appears to be mediated predominantly by a NO‐independent mechanism which is sensitive to increases in the extracellular potassium concentration and may reflect the action of endothelium‐derived hyperpolarizing factor (EDHF).

Influence of contractile agonists on the mechanism of endothelium‐dependent relaxation in rat isolated mesenteric artery

This study demonstrates directly that the relative contribution of nitric oxide (NO) and an NO synthase‐independent repolarization to acetylcholine‐evoked relaxation in rat isolated mesenteric resistance arteries is determined by the processes which mediate pre‐contraction. Noradrenaline‐induced contractions were reversed by acetylcholine via both NO and NO synthase‐independent smooth muscle repolarization. In contrast, reversal of contractions to the thromboxane‐mimetic, U46619, by acetylcholine was entirely mediated by the actions of NO, independently of a change in membrane potential.

Multiple pathways underlying endothelium‐dependent relaxation in the rabbit isolated femoral artery

1 In isolated segments of the rabbit femoral artery stimulated with noradrenaline, both acetylcholine (1nM‐10μM) and the calcium ionophore A23187 (1nM‐100 μM) evoked endothelium‐dependent smooth muscle relaxation and hyperpolarization while bradykinin (0.01–100 nM) had no effect. 2 The nitric oxide synthase inhibitors, NG‐nitro‐L‐arginine (L‐NOARG; 100 μM; 20 min) or NG‐nitro‐L‐arginine methyl ester (L‐NAME; 100 μM; 20 min) each abolished the hyperpolarization and the majority of the relaxation to acetylcholine (maximal response reduced from 96.8 ±2.3% to 2.0 ± 1.4%). 3 The potassium channel blocker, glibenclamide (10 μM; 10 min) also abolished the change in membrane potential to acetylcholine but did not modify the smooth muscle relaxation. 4 In contrast, neither L‐NAME nor glibenclamide modified the comparable responses of the femoral artery to A23187, which were also unaffected by the cyclo‐oxygenase inhibitor, indomethacin (10 μM). 5 In artery segments stimulated with potassium chloride (25 mM), the maximal change in tension and membrane potential evoked by A23187 (100 μM) was significantly reduced from 95.0 ±4.5% and 23.0±2.0mV to 69.0 ±10.1% and 12.0±1.5 mV, respectively. Under these conditions L‐NAME further reduced the relaxation but not the accompanying hyperpolarization to A23187. 6 Endothelium‐denuded arterial segments sandwiched with endothelium‐intact ‘donor’ segments gave qualitatively similar relaxant responses to those described above for acetylcholine and A23187. 7 Exogenous nitric oxide (0.5–10μM) stimulated a transient relaxation in precontracted artery segments, which at concentrations above 5 μM was accompanied by smooth muscle hyperpolarization (maximum 8.5 ± 3.2 mV; n = 4). The hyperpolarization but not the relaxation to nitric oxide was abolished by either glibenclamide or 25 mM potassium. 8 These data indicate that in the femoral artery, acetylcholine‐induced relaxation can be attributed solely to the release of nitric oxide from the endothelium, which then stimulates relaxation independently of a change in smooth muscle membrane potential. In contrast, both the relaxation and hyperpolarization evoked by A23187 appear to be mediated predominantly by nitric oxide‐independent pathways which appear to involve a diffusible factor released from the endothelium. The results suggest that this diffusible hyperpolarizing factor can be released from endothelial cells in the femoral artery by A23187 but not by acetylcholine.

Interactions between endothelium-derived relaxing factors in the rat hepatic artery: focus on regulation of EDHF.

In rat isolated hepatic arteries contracted with phenylephrine, acetylcholine and the calcium ionophore A23187 each elicit endothelium‐dependent relaxations, which involve both nitric oxide (NO) and endothelium‐derived hyperpolarizing factor (EDHF). However, the contribution of prostanoids to these responses, and the potential interaction between EDHF and other endothelium‐derived relaxing factors have not been examined. In the presence of the NO synthase inhibitor NG‐nitro‐L‐arginine (L‐NOARG, 0.3 mM) and a mixture of charybdotoxin (0.3 μM) and apamin (0.3 μM), inhibitors of the target potassium (K) channel(s) for EDHF, acetylcholine and A23187 each induced a concentration‐dependent and almost complete relaxation, which was abolished in the additional presence of indomethacin (10 μM). Thus, in addition to EDHF and NO, a relaxing factor(s) generated by cyclo‐oxygenase (COX) contributes to endothelium‐dependent relaxation in the rat hepatic artery. The resting membrane potentials of endothelium‐intact and endothelium‐denuded vascular segments were −57 mV and −52 mV, respectively (P>0.05). In intact arteries, the resting membrane potential was not affected by L‐NOARG plus indomethacin, but reduced to −47 mV in the presence of charybdotoxin plus apamin. Acetylcholine and A23187 (10 μM each) elicited a hyperpolarization of 13 mV and 15 mV, respectively. The hyperpolarization induced by these agents was not affected by L‐NOARG plus indomethacin (12 mV and 14 mV, respectively), but reduced in the presence of charybdotoxin plus apamin (7 mV and 10 mV, respectively), and abolished in the combined presence of charybdotoxin, apamin and indomethacin. The NO donor 3‐morpholino‐sydnonimine (SIN‐1) induced a concentration‐dependent relaxation, which was unaffected by charybdotoxin plus apamin, but abolished by the selective soluble guanylate cyclase inhibitor 1H‐[1,2,4]oxadiazolo[4,3‐a]quinoxaline‐1‐one (ODQ, 10 μM). SIN‐1 (10 μM) did not alter the resting membrane potential in endothelium‐denuded vascular segments. The COX‐dependent relaxation induced by acetylcholine was abolished following exposure to 30 mM KCl, but unaffected by glibenclamide (10 μM). The prostacyclin analogue iloprost induced a concentration‐dependent relaxation, which was also abolished in 30 mM KCl and unaffected by the combined treatment with glibenclamide, charybdotoxin and apamin. Iloprost (10 μM) induced a glibenclamide‐resistant hyperpolarization (8 mV with and 9 mV without glibenclamide) in endothelium‐denuded vascular segments. Exposure to SIN‐1 or iloprost did not affect the EDHF‐mediated relaxation induced by acetylcholine (i.e. in the presence of L‐NOARG and indomethacin). Replacement of L‐NOARG with the NO scavenger oxyhaemoglobin (10 μM) or the soluble guanylate cyclase inhibitor ODQ (10 μM) or methylene blue (10 μM), which all significantly inhibited responses to endothelium‐derived NO, did not affect the acetylcholine‐induced relaxation in the presence of indomethacin, indicating that endogenous NO also does not suppress EDHF‐mediated responses. These results show that, in addition to EDHF and NO, an endothelium‐derived hyperpolarizing factor(s) generated by COX contributes significantly to endothelium‐dependent relaxation in the rat heptic artery. Neither this factor nor NO seems to regulate EDHF‐mediated responses. Thus, EDHF does not serve simply as a ‘back‐up’ system for NO and prostacyclin in this artery. However, whether EDHF modulates the NO and COX pathways remains to be determined.