Vanhoutte, Paul M.

The endothelium can secrete both relaxing and contracting substances. One of the most powerful stimuli to the release of the former are thrombin and aggregating platelets. This contributes to the protective role of the endothelium against inappropriate intraluminal platelet aggregation and coagulation in blood vessels with an intact intima. Thrombin-induced, endothelium-dependent relaxations have been obtained in isolated arteries of different species, including humans. Endothelium-dependent relaxations can be evoked by autologous platelets in isolated blood vessels of the dog, pig and rat; they can be obtained in canine coronary arteries with human platelets. The major platelet-products involved in these endothelium-dependent relaxations are 5-hydroxytryptamine (serotonin) and the adenine nucleotides. Although platelet-activating factor (PAF) can evoke endothelium-dependent relaxation it only does so at concentrations much higher than those occurring under physiological conditions; since the relaxations are not prevented by PAF-antagonists, they are non-specific in nature.The receptor mediating the endothelium-dependent relaxations to serotonin released from the aggregating platelets can be subtyped as a S1~(5HT1) serotonergic receptor;those mediating the response to the adenine nucleotides as P2y-purinergic receptors. In the absence of the endothelium aggregating platelets cause contractions of vascular smooth muscle; these are mediated by a mixture of S1-like and S2~serotoner-gic receptors in coronary arteriesof the dog, and by S2-serotonergic receptors in those of the pig. Thus, in the porcine coronary artery, the S2-serotonergic antagonist ketanserin markedly enhances the platelet-induced endothelium-dependent relaxation. After previous (four weeks) injury, the regenerated endothelium of the porcine coronary artery loses the ability to respond to serotonin,and is unable to prevent the constrictionsevoked by aggregating platelets. The endothelium-dependent relaxations of porcine coronary arteries evoked by aggregating platelets are potentiated by chronic treatmentof the donor animals with cod liver oil. These studies emphasize the protective roleof the endothelial cells against the vasoconstriction (vasospasm) induced by aggregating platelets. This role is depressed after previous injury, and can be facilitatedby dietary adj ustments.

The effects of nitric oxide and acetylcholine (ACh) were investigated on the electrical and mechanical properties of vascular smooth muscle cells of the canine mesenteric artery. Isolated tissues with or without the endothelium were contracted with prostaglandin F2 alpha. Nitric oxide caused comparable concentration-dependent relaxations in rings with and without endothelium. ACh induced concentration-dependent relaxations only in arteries with endothelium. The relaxations to both nitric oxide and ACh were inhibited by methylene blue or oxyhemoglobin. Either in the presence or absence of prostaglandin F2 alpha, ACh caused transient hyperpolarization of the cell membrane of the vascular smooth muscle. The ACh-induced transient hyperpolarization was not observed after mechanical removal of the endothelial cells or after treatment with atropine. Nitric oxide (less than or equal to 8 X 10(-6) M) did not alter membrane potential, in either the presence or absence of the endothelium. The excitatory junction potentials generated by perivascular nerve stimulation were inhibited by ACh but not by nitric oxide. These results suggest that in the canine mesenteric artery 1) the endothelium-derived hyperpolarizing factor generated by ACh is not nitric oxide; 2) nitric oxide relaxes vascular smooth muscle by a direct effect; and 3) nitric oxide does not modify adrenergic neurotransmission.

The present study was designed to determine the role of the respiratory epithelium as a diffusion barrier and a modulator of the responsiveness of airway smooth muscle to bronchoactive agents. Segments of canine bronchi, with or without epithelium, were suspended in organ chambers and perfused intraluminally. The isometric tension was recorded. Acetylcholine, given intraluminally, induced significantly smaller contractions in bronchi with than in bronchi without epithelium. When this agonist was given extraluminally, no difference in contractions was noted between the tissues. In the presence of acetylcholine and phentolamine, norepinephrine, given either intra- or extraluminally, induced significantly larger relaxations in bronchi with than in bronchi without epithelium. High potassium given intraluminally induced contractions only in bronchi without epithelium; however, in the presence of ouabain, both tissues contracted similarly. When high potassium was given extraluminally, no difference in contraction between tissues with and without epithelium was noted. When [3H]acetylcholine and [3H]norepinephrine were perfused intraluminally, the accumulation of 3H radioactivity in the extraluminal solutions was significantly less in bronchi with than in bronchi without epithelium. These observations suggest that the epithelium acts as both a diffusion barrier and a modulator of the responses of canine airways to bronchoactive agents.

Experiments were designed to determine the effect of oxygen-derived free radicals in isolated canine basilar arteries. Rings with and without endothelium were suspended for isometric tension recording in modified Krebs-Ringer bicarbonate solution bubbled with 95% O2-5% CO2 (temperature = 37 degrees C; pH = 7.4). A radioimmunoassay technique was used to measure production of prostaglandins and thromboxane B2. Xanthine oxidase (1-9 mU/ml, in the presence of 10(-4) M xanthine) and hydrogen peroxide (10(-6) to 10(-4) M) caused concentration-dependent contractions. The removal of endothelium reversed these contractions into relaxations. Contractions to xanthine oxidase and hydrogen peroxide were inhibited in the presence of superoxide dismutase (150 U/ml), catalase (1,200 U/ml), indomethacin (10(-5) M), and SQ 29548 (10(-6) M) but not in the presence of deferoxamine (10(-4) to 10(-3) M) and dimethyl sulfoxide (10(-4) M). NG-monomethyl-L-arginine (3 x 10(-5) M) augmented the contractions to hydrogen peroxide. Xanthine oxidase stimulated production of 6-ketoprostaglandin F1 alpha, prostaglandin F2 alpha, prostaglandin E2, and thromboxane B2. The stimulatory effect was prevented by the removal of endothelial cells. These studies suggest that xanthine oxidase causes endothelium-dependent contractions mediated by: 1) hydrogen peroxide-induced stimulation of the endothelial metabolism of arachidonic acid via the cyclooxygenase pathway, leading to activation of prostaglandin H2-thromboxane A2 receptors, and 2) inactivation of basal production of nitric oxide by superoxide anions.

The effect of removing the airway epithelium on the responses of canine airways of decreasing diameter to contractile and relaxing agonists was explored. Three orders of canine bronchus were studied: second order (lobar bronchus), third order (segmental bronchus), and fourth order (subsegmental bronchus). Paired rings of tissue, with and without epithelium, were placed in organ chambers in physiological salt solution gassed with 95% O2–5% CO2 and maintained at 37 degrees C. For second- and third-order bronchi, epithelium removal caused significant left-ward shifts of the concentration-effect curves for 5-hydroxytryptamine, histamine, and acetylcholine. In fourth-order bronchi, there was no significant shift for any of the contractile agonists. Isoproterenol (during contractions evoked by acetylcholine) induced concentration-dependent relaxations that were significantly greater in bronchi with than in those without epithelium. This effect was most prominent in fourth-order bronchi. These results suggest that 1) the canine airway epithelium releases a relaxing factor, 2) in larger airways the major effect is reduction of contractile responses, and 3) in smaller airways the major effect is enhancement of relaxing responses.

Experiments were conducted in canine bronchi to determine whether the effect of epithelium removal on relaxations was affected by the contractile agent used to induce active force and the degree of contraction. Pairs of fourth-order bronchi with and without epithelium were suspended in organ chambers in physiological salt solution (95% O2-5% CO2, 37 degrees C). The bronchi were contracted to the concentration of acetylcholine or 5-hydroxytryptamine that resulted in a contraction that was 40 or 80% of the response to 10(-4) M of the agonist (ED40 or ED80). Epithelium removal reduced relaxations to isoproterenol and sodium nitroprusside during contraction to the ED80 but not the ED40 of acetylcholine. Responses to forskolin were not affected. Bronchi were significantly more sensitive to relaxing agonists in the presence of 5-hydroxytryptamine; there was no effect on epithelium removal or level of contraction. Thus 1) the influence of epithelium on bronchial relaxation is greatest during high degrees of cholinergic tone and 2) experimental conditions influence the effect of epithelium removal on relaxation.

Experiments were designed to compare the sensitivity of venous and arterial smooth muscle to endothelin and to determine whether contractions to the peptide could be inhibited by endothelium-derived relaxing factor and nitric oxide. Rings of canine left anterior descending coronary, femoral, and mesenteric arteries and femoral and saphenous veins with and without endothelium were suspended for measurement of isometric force. In the presence of indomethacin, phentolamine, and propranolol, endothelin initiated concentration-dependent increases in tension in all rings. The veins were more sensitive to the peptide than were the arteries. Endothelin depolarized the smooth muscle of the saphenous veins and mesenteric arteries; the threshold concentration for depolarization was approximately 100 times lower in the veins (10(-10) M) than in the arteries (10(-8) M). Removal of the endothelium enhanced the sensitivity only of venous smooth muscle to endothelin. However, stimulation of the endothelium in the arteries with either acetylcholine or the calcium ionophore A23187 rapidly inhibited the maximal tension developed to the peptide. Nitric oxide inhibited contractions to endothelin in arteries and veins without endothelium; the inhibition was greater in the arteries than in the veins. These results indicate that venous smooth muscle is more sensitive than arterial smooth muscle to endothelin. In both blood vessels, endothelium-derived relaxing factor(s) can inhibit contractions to the peptide.

Dietary supplementation with cod-liver oil significantly augments endothelium-dependent relaxations in porcine coronary arteries. The present study was designed to examine the effect of dietary administration of omega 3 polyunsaturated fatty acids (mainly eicosapentaenoic acid, the major component of fish oil) on endothelium-dependent relaxations in porcine coronary arteries. Male Yorkshire pigs were maintained 4 wk on a regular diet with or without supplementation with purified eicosapentaenoic acid (3.5 g/day) and docosahexaenoic acid (1.5 g/day). Endothelium-dependent relaxations were examined in vitro. In rings from the treated group, endothelium-dependent relaxations were augmented in response to bradykinin, serotonin, and ADP, but not to the calcium ionophore A23187. These augmentations were not altered by indomethacin but were significantly inhibited by methylene blue, an inhibitor of guanylate cyclase. In the treated group, endothelium-dependent relaxations to aggregating platelets also were significantly augmented; platelet-induced contractions of quiescent rings were inhibited more by the presence of the endothelium than in arteries from the control group. Bioassay experiments demonstrated that the release of endothelium-derived relaxing factor(s) by bradykinin and relaxations of the vascular smooth muscle to the factor(s) were greater in arteries from the treated group. These observations indicate that dietary omega 3 polyunsaturated fatty acids augment receptor-operated endothelium-dependent relaxations, partly due to the augmented release of endothelium-derived relaxing factor(s) and partly due to the augmented relaxation of the vascular smooth muscle to the factor(s).

Chronic increases in blood flow caused by an arteriovenous fistula augment endothelium-dependent relaxations to acetylcholine. To determine whether endothelial muscarinic receptors are altered, concentration-response curves to acetylcholine were obtained in the presence of pirenzepine in fistula- and sham-operated canine femoral arteries. Pirenzepine inhibited the response to acetylcholine in both arteries. The pA2 (log Kb) for the antagonist was the same. A bioassay system was used to assess release of endothelium-derived relaxing factor. Rings of femoral artery (without endothelium) from unoperated dogs relaxed more when superfused with perfusate derived from endothelium of fistula-operated arteries during acetylcholine stimulation. Rings without endothelium of sham- and fistula-operated arteries relaxed to the same extent when superfused with perfusate derived from the endothelium of unoperated femoral arteries. These results suggest that augmented relaxations to acetylcholine in canine arteries where blood flow is chronically elevated do not result from changes in the subtype of endothelial muscarinic receptors or in the sensitivity of the underlying smooth muscle to endothelium-derived relaxing factor(s). They are likely due to increased release of endothelium-derived relaxing factor(s) on muscarinic activation.

The aim of the present study was to investigate the influence of inhibitors of endothelium-dependent relaxing factor on the release of atrial natriuretic factor from isolated rat atria. Electrical stimulation of the tissue produced a frequency-dependent increase in secretion of atriopeptide from basal levels. Saponin (0.3 mg/ml for 45 min) augmented the basal release of the peptide. Methylene blue (10(-5)M), oxyhemoglobin (10(-6)M), or hydroquinone (10(-5)M), agents that inhibit the effects of endothelium-derived relaxing factor by different mechanisms, caused an increase in the basal secretion of atrial natriuretic factor. Blockade of cyclooxygenase with indomethacin (10(-5)M) did not alter the release of atriopeptide, suggesting that it is not controlled by endogenous prostaglandins. These results would be explained if the inhibitors tested were to interfere with a factor released continuously from either endocardial cells or endothelial cells of the atrial blood vessels and having the same nature as endothelium-derived relaxing factor. Such release may exert a tonic negative modulatory influence on the secretion of atrial natriuretic factor.