Loss of Selective Endothelial Cell Vasoactive Functions Caused by Hypercholesterolemia in Pig Coronary Arteries

Transcription

1 903 Loss of Selective Endothelial Cell Vasoactive Functions Caused by Hypercholesterolemia in Pig Coronary Arteries Richard A. Cohen, Kevin M. Zitnay, C. Christian Haudenschild, and Leslie D. Cunningham The influence of hypercholesterolemia on the reactivity of coronary arteries was investigated after feeding a high-cholesterol diet to pigs for 9 weeks. After this duration of hypercholesterolemia, the fatty or intimal proliferative changes of atherosclerosis were not yet evident in the coronary arteries by light or electron microscopy. Changes hi isometric tension were compared in isolated ring segments of coronary arteries from normal and hypercholesterolemic animals. The endothelium failed to inhibit contractions caused by 5-hydroxytryptamine in coronary arteries from hypercholesterolemic animals, but it did so in normal vessels. In contracted arteries, endothelium-dependent relaxations caused by 5-hydroxytryptamine and substance P were reduced by hypercholesterolemia. In contrast, endothelium-dependent relaxations mediated by norepinephrine acting at a 2 -adrenoceptors and those caused by the calcium ionophore A23187 were unaffected. Endothelium-independent /J-adrenergic relaxations caused by norepinephrine, as well as those caused by nitroprusside, and papaverine also were unaffected by hypercholesterolemia. The loss of selective endothelial cell receptor-mediated relaxation suggests that it is not the ability of the coronary artery endothelium to elaborate vasodilators, but the initiation of the coronary artery endothelial cell response to 5-hydroxytryptamine and substance P that is affected by hypercholesterolemia. Thus, during hypercholesterolemia, selective endothelial cell dysfunction giving rise to abnormal coronary artery reactivity precedes the onset of coronary artery atherosclerosis. (Circulation Research 1988;63: ) By releasing endothelium-derived relaxing factor(s), the endothelium plays a critical role in the local control of blood vessels. 1 The tonic release of relaxing factors as well as that induced by several receptor-dependent agonists is important in modulating contraction of the smooth muscle.'- 3 Changes in the physiological responses of endothelial cells have been noted in atherosclerotic arteries, which suggests that vasospasm of atherosclerotic arteries could be due to lack of the normal vasodilator influence of the endothelium In atherosclerotic human coronary arteries, 43 there are attenuated endothelium-dependent relaxations induced by acetylcholine, but those caused by the calcium ionophore A23187 may 5 or may not 4 be From the Robert Dawson Evans Department of Clinical Research and the Mallory Institute of Pathology, Boston University Medical Center, Boston, Massachusetts. Supported by National Institutes of Health grants HL-31607, HL-30668, HL-38731, and HL R.A.C. is supported by an Established Investigator Award from the American Heart Association. Address for correspondence: Richard A. Cohen, MD, E411, University Hospital, 75 E. Newton Street, Boston, MA Received November 25, 1987; accepted May 27, inhibited. In rabbits made atherosclerotic by severe hypercholesterolemia, there is an impairment of relaxations induced by acetylcholine and adenosine triphosphate, 6-8 but those caused by A23187 are preserved. 4 In pig coronary arteries made atherosclerotic by a combination of cholesterol feeding and balloon catheter-induced intimal injury, endothelium-dependent relaxations caused by 5- hydroxytryptamine (5-HT) are markedly attenuated, whereas those to A23187 are minimally affected. 9 In monkey iliac arteries made atherosclerotic by prolonged hypercholesterolemia, endothelium-dependent relaxations caused by acetylcholine and thrombin are inhibited, whereas those to A23187 are only moderately decreased Thus, studies of atherosclerotic arteries have demonstrated that endothelium-dependent relaxations caused by some, but not all, agonists are inhibited, which suggests a selective alteration by atherosclerosis of endothelial cell function. While it is not unexpected that atherosclerosis and the accompanying intimal thickening will affect the influence of endothelial cells on vascular smooth muscle, it is unknown if hypercholesterolemia affects the vaso-

2 904 Circulation Research Vol 63, No 5, November 1988 TABLE 1. Ingredient High-Cholesterol Diet Constituents Cholesterol Sodium choleate Sodium chloride Lard Milk powder (28.5% butterfat) Peanut oil VitaPlus (vitamin) High-protein pig mash (Purina Hog-grower #2122) Total Amount per day 22 g Kg I2g 200 g 52 g 40g 10 g 800g 1,147 g active role of the endothelium before the onset of atherosclerosis. The purpose of this study therefore was to determine the effect of hypercholesterolemia induced for a relatively short duration and moderate degree on pig coronary artery reactivity with special emphasis on endotheliumndependent responses. Materials and Methods Seven domestic swine were fed a high-cholesterol diet (Table 1) modified from Nam et al 12 for a period of 8.8 ±0.8 weeks (range, 6-12). At the time the animals were killed, the cholesterol-fed pigs weighed 26 ±2.2 kg. Seven control pigs were fed regular pig mash for an average of 2-3 weeks, and their weights at the time of death did not differ significantly from the cholesterol-fed pigs. Pigs were killed by exsanguination following ketamine (20 mg/kg i.m.), acepromazine (0.8 mg/kg i.m.), chloralose (75 mg/kg i.v.), and sodium heparin (80 units/kg i.v.). At the time of death significant hypercholesterolemia was present in the experimental group (601 ±45 mg/dl) compared with the controls (94 ±7.9, p<0.05). No significant alteration in triglycerides occurred (34 ±4.5 vs. 34 ±6.3 mg/dl). Organ Chamber Experiments Rings of artery 5 mm long were suspended in organ chambers (25 ml) filled with physiological salt solution (PSS) of the following millimolar composition: NaCl 118.3, KC1 4.7, MgSO 4 0.6, KH 2 PO 4 1.2, CaCl 2 2.5, NaHCO 3 25, Ca-EDTA 0.026, and glucose The solution was maintained at 37 C and gassed with 20% O 2-5% CO 2-75% N 2. The endothelium was removed from some rings by rubbing of the luminal surface with a stainless steel wire; removal of endothelium was later confirmed by absence of relaxations caused by substance P. 13 The rings were mounted on stirrups connected to a strain gauge so as to measure isometric circumferential force. The rings were stretched gradually over 1 hour to 12 g resting tension that had been determined previously to be optimal tension for contraction of normal pig coronary arteries. 3 During the last 15 minutes of this period, sodium nitroprusside (10~ 7 M) was introduced to eliminate spontaneous active tone which is observed in rings without endothelium 3 and tension was again adjusted to 12 g. The sodium nitroprusside was then removed by rinsing for 30 minutes. Relaxations. To measure relaxations, the rings were contracted with prostaglandin F^ (PGFjJ. After the PGF^-induced contraction stabilized, increasing concentrations of 5-HT, substance P, norepinephrine, A23187, nitroprusside, or papaverine were added until a maximum response occurred. Following this, nitroprusside (10~ 6 M) was added, which caused a maximal relaxation in all rings; this was assured by adding higher concentrations in some studies. Responses to each agonist were obtained in pairs of rings, one with and one without endothelium. The order of agonists was randomized. The control pigs were studied concurrently. Relaxations are expressed as a percentage of the maximal relaxation caused by nitroprusside. In some studies, the concentration causing 50% maximal relaxation (IC50) was estimated by linear regression and is reported as the logarithm. Contractions. Contractions caused by 5-HT, PGF^, and potassium are reported in grams. The concentration of potassium causing fifty percent of the maximum contraction (EDJQ) was estimated by linear regression analysis. The negative logarithm of the EDso is reported as the pd Drugs The following agents were used: A23187, 5-HT creatinine sulfate, norepinephrine bitartrate, papaverine, propranolol hydrochloride, PGF^ (Tris salt), sodium nitroprusside, substance P (Sigma Chemical, St. Louis, Missouri); ketanserin bitartrate (Janssen Pharmaceutica, Beerse, Belgium); methiothepin maleate (Hoffman-LaRoche, Nutley, New Jersey); prazosin hydrochloride (Pfizer, Brooklyn, New York); and rauwolscine hydrochloride (Carl Roth, KG, Karlsruhe, FRG). Unless otherwise specified, drugs were dissolved in distilled water such that volumes of 100 ^1 were added to the organ chamber. Norepinephrine was dissolved in 0.1% ascorbic acid. Prazosin and methiothepin were solubilized in DMSO and diluted with water. All concentrations are the final molar concentrations reached in the organ chamber. Potassium concentration (10, 20, 40, and 120 mm) was increased by equimolar replacement of sodium. Data and Statistical Analysis The data was expressed as mean±sem. Unless otherwise indicated, n was seven. The geometric means of the concentrations were analyzed. Statistical evaluations were by way of Student's t test for paired comparisons of responses of rings of the same artery with or without endothelium, or for unpaired comparisons of responses of rings of arteries from normal and hypercholesterolemic animals. Values of p<0.05 were regarded as significant. Morphology After dissection from the heart, a ring of coronary artery from each heart was fixed by immersion in

3 Cohen et al Hypercholesterolemia and Coronary Artery Endothelium 905 TABLE 2. PGF^-Induced Contraction of Coronary Arteries From Normal and Hypercholesterolcmlc Pigs Normal Hypcrcholesterolemic Endothelium Concentration Contraction With Without With Without ± ±0.10* 5.7± ± ± ±0.4t * 3.9±0.5t Data are mean ± SEM (n = 7) of the concentrations (logm) of prostaglandin F^ (PGF^) used to contract pig coronary artery rings and the resulting contractions (g). *For arteries from both normal and hypercholesterolemic animals, a significantly lower concentration of PGFi, was required in rings denuded of endothelium to yield equal contractions to those in rings with endothelium. tpgf^-induced contractions of arterial rings with or without endothelium from hypercholesterolemic animals were significantly less than those of normal rings. 4% formaldehyde, 1% glutaraldehyde in isotonic phosphate buffer. Paraffin sections at 4 /tm were examined after staining with hematoxylin-eosin, modified trichrome, or van Gieson-elastin stains. For scanning electron microscopy, samples were dehydrated through graded alcohol, dried by the critical point method, sputter-coated with gold palladium alloy, and examined with a Jeol microscope. Results Spontaneous Tone and Contractions Caused by PGF 2a, 5-HT, and Potassium Following stretching to the optimal resting tension for contraction in the presence of nitroprusside (10~ 7 M), rinsing coronary artery rings denuded of endothelium from normal and hypercholesterolemic pigs resulted in spontaneous increases in tone. The tone developed in rings from normal pigs was significantly greater than in rings from hypercholesterolemic pigs (5.4±0.9 vs. 2.3±0.7 g). In neither group did rings with endothelium develop tone. PGF^ was used to bring the total tension of rings with and without endothelium to similar levels. Higher concentrations of PGF^ were required to do this in rings with endothelium than in rings denuded of endothelium from both normal and hypercholesterolemic pigs (Table 2). There was no significant difference in the concentrations used for rings of arteries from normal and hypercholesterolemic pigs. At these concentrations, PGFi, caused contractions of coronary artery rings with or without endothelium from hypercholesterolemic pigs that averaged significantly less than those of normal arteries (Table 2). In coronary arteries from both normal and hypercholesterolemic pigs contracted with PGF^, 5-HT (10~ 8 to 10~ 3 M) caused further contractions. In normal pigs, these contractions were of significantly less magnitude in rings with intact endothelium, while in coronary artery rings from hypercholesterolemic animals, the contractions caused by 5-HT were not different in rings with or without endothelium (Figure 1). In rings without endothelium, the 3 6 z g o CONTROL HYPERCHOLESTEROLEMIC S HYDROXYTRYPTAMINE.-IOQ M FIGURE 1. Response of coronary arteries from normal control and hypercholesterolemic pigs to 5-hydroxytryptamine (5-HT). Rings of artery were contracted with PGF 2a, and when the contractions stabilized, 5-HT was added in cumulative increasing concentrations. Data are mean±sem of the additional contraction in grams caused by each concentration of 5-HT in rings with (+) and without ( ) endothelium. In arteries from normal pigs (n=6), the rings without endothelium contracted significantly more than those with endothelium, while there was no significant influence of the endothelium on the contraction of rings of artery from hypercholesterolemic pigs (r\=5). contractions caused by 5-HT (3.2 x 10" 7 M and 10" 6 M) were significantly greater in normal coronary arteries than in those from hypercholesterolemic pigs (Figure 1). Contractions caused by potassium ( mm) were not significantly different in coronary artery rings denuded of endothelium from normal and hypercholesterolemic pigs (Table 3). Relaxations Caused by 5-HT, Substance P, Norepinephrine, A23187, Nitroprusside, and Papaverine In the presence of ketanserin (10~ 6 M), coronary artery rings with endothelium relaxed in response to 5-HT (10~ 8 to 10" 3 M); these relaxations were significantly less in arteries from hypercholesterolemic animals (Figure 2). In both normal and hypercholesterolemic pigs, methiothepin (10~ 6 M) blocked the relaxations caused by 5-HT. 5-HT caused minimal TABLE 3. Response of Normal and Hypercholesterolemic Pig Coronary Artery to Potassium* Potassium concentration (mm) pdj Normal (/i = 8) ± ± Cholesterol (n = 5) ± ±0.06 *Responses to 10, 20, and 40 mm potassium are expressed as percent of maximal contraction to potassium (120 mm) which were g and 18.0±2.4 g for normal and hypercholesterolemic pig arteries, respectively. The pdj is the log of the molar concentration causing 50% maximal contraction. There were no significant differences in responses to potassium of normal and hypercholesterolemic pig coronary arteries whether expressed as percent of maximum or in grams tension.

4 906 Circulation Research Vol 63, No 5, November 1988 CHOLESTEROL KETANSERIN (10 6 M> HYDROXYTRYPTAMINE,-log M FIGURE 2. Relaxations caused by 5-hydroxytryptamine (5-HT) of coronary arteries from normal and hypercholesterolemic pigs. Arteries pretreated for I hour with ketanserin (I0~ 6 M) were contracted with PGF 2a, and when the contraction stabilized, increasing concentrations of 5-HT were added. Data are mean±sem of relaxations expressed as percentage of the maximal relaxation caused by nitroprusside (IO~ 6 M) for rings with (+) and without ( ) endothelium. Control arteries with endothelium (n=6) relaxed significantly more than those from hypercholesterolemic animals (n=5), while in neither group did rings without endothelium relax significantly. or no relaxation in rings of normal or hypercholesterolemic coronary arteries denuded of endothelium. Coronary artery rings with endothelium from hypercholesterolemic pigs relaxed significantly less to substance P (3x 10" 12 to 3x 10" l0 M) than did those from normal pigs (Figure 3). In both normal and hypercholesterolemic pigs, coronary artery relaxations caused by substance P were abolished in rings denuded of endothelium. In rings with or without endothelium, the relaxations caused by norepinephrine (10"' to 3 x 10" 6 M) were not significantly different between coronary arteries from normal and hypercholesterolemic pigs (Figure 4, upper panel). In the presence of propranolol (10~ 6 M) and prazosin (5xlO~ 7 M), norepinephrine still relaxed rings with endothelium from normal and hypercholesterolemic pigs similarly (Figure 4, lower panel). The relaxations caused by norepinephrine (10~ 8 to 3 x 10" 6 M) in the presence of propranolol and prazosin were abolished after removing the endothelium from arteries of both normal and hypercholesterolemic pigs (Figure 4, lower panel). As in normal arteries, the further addition of rauwolscine (3 x 10~ 7 M) blocked the endothelium-dependent relaxations caused by norepinephrine (10~ 8 to 3 x 10~ 6 M) in rings from hypercholesterolemic pigs (Figure 4, lower panel). There were no significant differences in the endothelium-dependent relaxations induced by A23187 (10" 8 to 10" 6 M, Figure 5) of coronary arteries from normal and hypercholesterolemic pigs. Relaxations caused by nitroprusside or papaverine of coronary artery rings without endothelium -100 SUBSTANCE P,-log M FIGURE 3. Relaxations caused by substance P of coronary arteries from normal and hypercholesterolemic pigs. Arteries with intact endothelium were contracted with prostaglandin F 2a and substance P was added in cumulative increasing concentrations. Data are mean±sem of the relaxations expressed as percentage of the maximal relaxation caused by nitroprusside (IO~ 6 M). Asterisks indicate that rings from hypercholesterolemic animals (n=7) relaxed significantly less than those from normal animals (n=7). Rings without endothelium did not respond to substance P (not shown). from normal and hypercholesterolemic pigs were not significantly different. The log ICJO for nitroprusside was 8.1 ±0.2 (n = 7) in arteries from normal pigs and 7.8 ±0.1 (n = 5) in those from hypercholesterolemic pigs (p>0.l). The log IC50 for papaverine was -6.0 ±0.2 (n = 4) in arteries from normal pigs and -6.1 ±0.1 (n = 5) in those from hypercholesterolemic pigs (/?>0.2). Morphology By light microscopy high cholesterol diet of this duration and degree had no noticeable atherogenic effects on the coronary arteries. Specifically, foam cells were not observed on regular sections or fat stained frozen sections, and there was no intimal thickening (Figure 6). Scanning electron microscopy showed a continuous layer of endothelial cells with rare gaps and some microvillous surface changes (Figure 7). Comparable minimal alterations in the endothelial cell surface were observed in similarly fixed coronary artery rings of three control pigs. Discussion Effect of Hypercholesterolemia on Contractions of Pig Coronary Artery Pig coronary arteries without endothelium spontaneously develop myogenic tone 3 which may be due to the endogenous synthesis of contractile prostaglandins, since it is prevented by cyclooxygenase inhibitors. 15 The decreased generation of myogenic tone in arteries from hypercholesterolemic pigs is therefore consistent with the smaller contractions caused by exogenously added PGF^. There is a reduction in other receptor-mediated contractions of

5 Cohen et al Hypercholesterolemia and Coronary Artery Endothelium 907 NOREPMEPHRINE.-log M 6.5 7ft BJi 6,6 0H>- -oh O P < CONTROL eo CONTROL -100 (T5 6.0 C iod Or CHOtE8TEROL**5 NOREPMEPHRINE.-log M ,6 A23187,-lofl M FIGURE 5. Effect of hypercholesterolemia on the response to A Rings of coronary artery with (+) endothelium from normal (n=5) and hypercholesterolemic (n=5) pigs relaxed similarly. Rings without ( ) endothelium from neither group of animals relaxed in response to A FIGURE 4. Effect of hypercholesterolemia on the relaxations caused by norepinephrine. Data are mean±sem of the relaxations expressed as percentage of the maximal response to nitroprusside. Responses were obtained in the absence (upper panel) or the presence (lower panel) of propranolol (W 6 M) and prazosin (5xlO~ 7 M). Relaxations in rings with endothelium in the presence off}- and a r adrenoceptor blockade were abolished by removal of the endothelium or by the further addition of rauwolscine (shown for arteries from hypercholesterolemic animals). Regardless of the presence (+) or absence ( ) of endothelium or antagonists, there was no significant difference in the relaxations to norepinephrine of arteries from normal (n = 7) and hypercholesterolemic (n=7) arteries. coronary smooth muscle in cholesterol-fed pigs because contractions of rings without endothelium caused by 5-HT were also of less magnitude. The arteries from hypercholesterolemic animals could develop similar responses to a direct smooth muscle contractile stimulus since the sensitivity to and magnitude of contractions caused by potassiuminduced depolarization did not differ in the two groups of animals. Therefore, it is unlikely that arteries from hypercholesterolemic animals had decreased prostaglandin and 5-HT-induced contractions due to dissimilar passive mechanical characteristics. While the mechanism of the reduced contractions is unknown and merits further investigation, moderate decreases in the magnitude of agonistinduced contractions have been observed in the aortas of hypercholesterolemic rabbits. 7-8 Influence of Hypercholesterolemia on Endothelium-Dependent Coronary Artery Relaxations Relaxations caused by several agonists were used to assess functional abnormalities of the pig coronary artery endothelium in cholesterol-fed animals. In pig and dog 1718 coronary arteries, 5-HT causes relaxations that are dependent on the presence of an intact endothelium and are mediated by 5-HT-like receptors which may be blocked by methysergide or methiothepin In pig coronary arteries, the endothelium-mediated relaxations are observed only after preventing contractions mediated by 5-HT 2 receptors on the smooth muscle. 16 Reduction of the relaxations caused by 5-HT in coronary arteries from hypercholesterolemic pigs suggests an impairment of the production of or response of the smooth muscle to endothelium-derived relaxing factors). An impairment of function is also suggested by the lack of the ability of the endothelium of arteries from hypercholesterolemic pigs to reduce 5-HT-induced contractions of the smooth muscle observed in the absence of ketanserin. Substance P also causes endothelium-dependent relaxations of pig coronary arteries by releasing endothelium-derived relaxing factor(s). 313 Although a significant reduction in effectiveness of substance P was seen in arteries from cholesterol-fed animals, the reduction was less marked than that for 5-HT. Despite the fact that it was less potent than 5-HT or substance P in causing endothelium-dependent relaxations, norepinephrine caused normal relaxations of coronary arteries from hypercholesterolemic pigs that were blocked by the a 2 -adrenoceptor antagonist rauwolscine. 319 The normal a 2 -adrenoceptor-mediated relaxation indicates that the impairment in the endothelium-dependent response to 5- HT and substance P in cholesterol-fed pigs is specific for those agonists rather than affecting responses mediated by all endothelial cell receptors.

6 908 Circulation Research Vol 63, No 5, November 1988 FIGURE 6. Light micrograph of an immersion-fixed crosssectioned right coronary artery from a pig fed the high-cholesterol diet for 11 weeks with a terminal cholesterol level of 570 mg/dl. Note the endothelium and the absence of intimal thickening or foam cells. Modified trichrome strain; original magnification, X130. A23187 initiates endothelium-dependent relaxations by inducing calcium influx into endothelial cells, bypassing the initial steps in endothelial cell activation mediated by receptors. 20 Normal relaxations caused by the ionophore suggest that endothelial cells from hypercholesterolemic pigs can produce normal amounts of relaxing factors), which in turn initiates a normal smooth muscle relaxation. Tonic release of relaxing factors from endothelium inhibits myogenic tone and PGF^-induced contractions of pig coronary artery. 3 That these contractions were inhibited to a similar extent as those of normal arteries suggests that the endothelium of coronary arteries of cholesterol-fed pigs also tonically releases relaxing factor(s) in normal quantities. Normal relaxations of arteries from cholesterolfed arteries mediated by direct smooth muscle relaxants nitroprusside, papaverine, and norepinephrine (in the absence of propranolol), as well as by endothelial cell a 2 -adrenoceptors and A23187, indicate that differences in the magnitude of the contractions caused by PGF^ were not responsible for the differences observed in relaxations to 5-HT and substance P. The mechanism by which hypercholesterolemia selectively inhibited endothelial cell function in the FIGURE 7. Scanning electron micrograph of the endothelium of an immersion fixed right coronary artery of a pig fed highcholesterol diet for 6 weeks with a terminal cholesterol level of 546 mg/dl. The endothelium is continuous except for rare small gaps (arrows). Asterisk is on a cell with a villous surface. Original magnification, x2,000.

7 Cohen et al Hypercholesterolemia and Coronary Artery Endothelium 909 present study is unknown but could involve changes in the numbers of endothelial cell receptors for 5-HT and substance P or in the steps leading to activation of endothelial cells initiated by those particular receptors that are susceptible to hypercholesterolemia. The three receptor agonists employed in this study and A23187 are all likely to share a common mechanism of relaxation by activation of smooth muscle guanylate cyclase 2-21 as the relaxations are all inhibited by methylene blue (authors' unpublished observations). 5-HT and substance P do distinguish themselves in that their transmembrane signal transduction is known to be dependent on phosphatidylinositol turnover to increase cell calcium Cellular activation by way of a^-adrenoceptors and A are by contrast strictly dependent on the influx of extracellular calcium. The recent demonstration that exposure of the rabbit aorta in vitro to low density lipoproteins rapidly inhibits the endotheliumdependent relaxation caused by acetylcholine 27 suggests that the dysfunction observed in the present study could be due to a direct interaction of elevated levels of lipoproteins with endothelial cells. As low density lipoproteins have also been shown to activate cells by way of phosphatidylinositol metabolism, 28 interference with this pathway would provide a possible mechanism by which responses to 5-HT and substance P are selectively inhibited by hypercholesterolemia. In the present study, pig coronary arteries were studied at a time when hypercholesterolemia had produced no abnormal structural alterations; the minimal changes observed in the endothelial cells were also observed in similarly fixed specimens from normal pigs. Thus, the changes in coronary artery endothelium-dependent responses observed are due to hypercholesterolemia without atherosclerosis. Furthermore, the abnormalities in specific endothelium-dependent vasoactive functions observed in this study suggest that the similar abnormalities in previous studies of atherosclerotic human and animal arteries could represent changes that preceded the development of atherosclerosis. Therefore, rather than reflecting atherosclerosis per se, the changes in endothelial cell vasoactive function induced by hypercholesterolemia may implicate early changes in endothelial cell function important in the development of atherosclerosis. Acknowledgments The authors thank Arun Masih for technical assistance, Marie Ryan for preparation of the figures, and Katherine Boris for preparation of the manuscript. References 1. Furchgott RF, Zawadzki JV: The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature (Lond) 1980;288: Martin W, Villani GM, Jothianandan D, Furchgott RF: Selective blockade of endothelium-dependent and glyceryl trinitrate-induced relaxation by hemoglobin and by methylene blue in the rabbit aorta. J Pharmacol Exp Ther 1985; 232: Cohen RA, Zitnay KM, Weisbrod RM, Tesfamariam B: Influence of the endothelium on tone and the response of isolated pig coronary artery to norepinephrine. J Pharmacol Exp Ther 1988;244: Bossaller C, Habib GB, Yamamoto H, Williams C, Wells S, Henry PD: Impaired muscarinic endothelium-dependent relaxation and cyclic guanosine 5'-monophosphate formation in atherosclerotic human coronary artery and rabbit aorta. J Clin Invest 1987;79: Forstermann U, Mugge A, Alheid U, Haverich A, Frolich JC: Selective attenuation of endothelium-mediated vasodilation in atherosclerotic human coronary arteries. Circ Res 1988;62: Habib JB, Bossaller C, Wells S, Williams C, Morrisett JD, Henry PD: Preservation of endothelium-dependent vascular relaxation in cholesterol-fed rabbit by treatment with the calcium blocker PN Circ Res 1986;58:305-3O9 7. Verbeuren TJ, Jordaens FH, Zonnekeyn LL, Van Hove CE, Coene M-C, Herman AG: Effect of hypercholesterolemia on vascular reactivity in the rabbit. I. Endothelium-dependent and endothelium-independent contractions and relaxations in isolated arteries of control and hypercholesterolemic rabbits. Circ Res 1986^8: Jayakody L, Senaratne M, Thomson A, Kappagoda T: Endothelium-dependent relaxation in experimental atherosclerosis in the rabbit. Circ Res 1987 ;60: Yamamoto Y, Tomoike H, Egashira K, Nakamura M: Attenuation of endothelium-related relaxation and enhanced responsiveness of vascular smooth muscle histamine in spastic coronary arterial segments from miniture pigs. Circ Res 1987;61: Freiman PC, Mitchell GG, Heistad DD, Armstrong ML, Harrison DG: Atherosclerosis impairs endotheliumdependent vascular relaxation to acetylcholine and thrombin in primates. Circ Res!986;58: Harrison DG, Freiman PC, Armstrong ML, Marcus ML, Heistad DD: Alterations of vascular reactivity in atherosclerosis. Circ Res 1987;6l(suppl II): Nam SC, Lee WM, Jarmolych J, Lee KT, Thomas WA: Rapid production of advanced atherosclerosis in swine by a combination of endothelial injury and cholesterol feeding. Exp Mot Pathol 1973;l8: Cocks TM, Angus JA: Endothelium-dependent relaxation of coronary arteries by noradrenaline and serotonin. Nature 1983^05: van Rossum JM: Cumulative dose-response curves. II. Technique for the making of dose-response curves in isolated organs and the evaluation of drug parameters. Arch Int Pharmacodyn Ther 1963;143: Rubanyi G, Paul RJ: Two distinct effects of oxygen on vascular tone in isolated porcine coronary arteries. Circ Res 1985^6: Shimokawa H, Aarhus LL, Vanhoutte PM: Porcine coronary arteries with regenerated endothelium have a reduced endothelium-dependent responsiveness to aggregating platelets and serotonin. Circ Res 1987;61: Cohen RA, Shepherd JT, Vanhoutte PM: 5-Hydroxytryptamine can mediate endothelium-dependent relaxation of coronary arteries. Am J Physiol I983;245:HI077-H Cohen RA, Shepherd JT, Vanhoutte PM: Endothelium and asymetrical responses of the coronary artery wall. Am J Physiol I984;247:H403-H Angus JA, Cocks TM, Satoh K: Alpha-2 adrenoceptors and endothelium-dependent relaxation in canine large arteries. Br J Pharmacol!986;88: Singer HA, Peach MJ: Calcium and endothelial-mediated vascular smooth muscle relaxation in rabbit aorta. Hypertension 1982;4(suppl II):II-19-II Rapaport RM, Murad F: Agonist-induced endotheliumdependent relaxation in rat thoracic aorta may be mediated through cgmp. Circ Res 1983,52:

8 910 Circulation Research Vol 63, No 5, November Berridge MJ, Dawson RMC, Downcs CP, Heslop JP, Irvine RF: Changes in the levels of inositol phosphates after agonist-dependent hydrolysis of membrane phosphoinositides. Biochem J 1983;212: Conn PJ, Sanders-Bush E: Biochemical characterization of serotonin stimulated phosphoinositide turnover. Life Sci 1986;38: Shogakiuchi Y, Kanaide H, Kobayashi S, Nishimura J, Nakamura M: Intracellular free calcium transients induced by norepinephrine in rat aortic smooth muscle cells in primary culture. Biochem Biophys Res Commun 1986; 135: van Zwieten PA, Timmcrmans PBMWM: Alpha-adrenoceptor stimulation and calcium movements. Blood Vessels 1987; 24: Derian CK, Moskowitz MA: Polyphosphoinositide hydrolysis in endothelial cells and carotid artery segments. Bradykinin-2 receptor stimulation is calcium-independent. J BiolChem 1986^61: Andrews HE, Bruckdorfer KR, Dunn RC, Jacobs M: Lowdensity lipoproteins inhibit endothelium-dependent relaxation in rabbit aorta. Nature 1987;327: Block LH, Knorr M, Vogt E, Locher R, Vetter W, Groscurth P, Qiao BY, Pometta D, James R, Regenass M, Pletscher A: Low density lipoprotein causes general cellular activation with increased phosphatidylinositol turnover and lipoprotein catabolism. Proc Natl Acad Sci USA 1988; 85: KEY WORDS endothelial cell coronary artery substance P prostaglandin F^ hypercholesterolemia 5-hydroxytryptamine