Fish Oil Treatment Decreases Superoxide Anions in the Myocardium and Coronary Arteries of Atherosclerotic Monkeys
Background This experiment sought to determine the effects of fish oil on superoxide anion production in the myocardium and coronary arteries of atherosclerotic monkeys. Recent evidence indicates that dietary supplementation with fish oil preserves normal vasomotion of atherosclerotic coronary arteries and reduces damage to the myocardium after ischemia and reperfusion, although the mechanisms remain unclear.
Methods and Results Adult male cynomolgus monkeys were fed an atherogenic diet with (n=15) or without (n=15) half the fat calories from fish oil. After 12 months, chemiluminescence of lucigenin was used to measure superoxide anion production in coronary arteries and myocardium after 1 hour of ischemia and 2 hours of reperfusion. The signals were calibrated with known quantities of xanthine and xanthine oxidase. Superoxide anion production in ischemic myocardium was (mean±SEM, nmol/mg dry wt per minute) 1±1 and 4±1 in monkeys fed fish oil and not fed fish oil, respectively (P<.05). Superoxide anion production in coronary arteries not exposed to ischemia and reperfusion was (nmol/mg dry wt per minute) 4±1 and 8±2 in monkeys fed fish oil and not fed fish oil, respectively (P<.05). Superoxide anion production in coronary arteries was (nmol/mg dry wt per minute) 5±2 and 16±3 in monkeys fed fish oil and not fed fish oil after ischemia and reperfusion, respectively (P<.05).
Conclusions Dietary supplementation with fish oil reduced vascular superoxide anion production and prevented the increase in vascular and myocardial superoxide anion production that accompanied ischemia and reperfusion. These phenomena may underlie some of the beneficial cardiovascular effects of fish oil.
Consumption of diets rich in fish oil has been associated with a reduced risk of coronary artery disease.1 Effects of fish oil on risk factors for coronary heart disease are variable.2 3 4 5 However, recent evidence indicates that dietary supplementation with fish oil reduces infarct size after ischemia and reperfusion6 7 and promotes coronary artery dilator responses,8 although the mechanisms remain unclear.
It has been proposed that invasion of inflammatory cells into the ischemic myocardium and the formation of oxygen free radicals during reperfusion may promote myocardial damage.9 ω-3 Fatty acids from fish oil can reduce superoxide anion production by inflammatory cells.10 However, there is no direct evidence that fish oil decreases superoxide production in the myocardium.
Recently it was reported that atherosclerosis increases the production of vascular superoxide anions and may, through this mechanism, inactivate nitric oxide and impair endothelium-mediated dilation of arteries.11 Shimokawa et al8 concluded that fish oil improves endothelium-dependent coronary dilator responses, presumably through nitric oxide–mediated mechanisms. However, there is no direct evidence that fish oil reduces superoxide anion production in arteries. Therefore, the goals of this study were to determine whether dietary fish oil reduces basal levels of superoxide anion in the myocardium and coronary arteries and whether fish oil diminishes the increase in superoxide anion production associated with ischemia and reperfusion. Such effects on superoxide anion production may explain in part the beneficial effects of fish oil on infarct size and coronary vasomotion.
Thirty adult male cynomolgus monkeys (Macaca fascicularis) were studied. The monkeys initially were housed at the Indonesian Primate Center in Bogor, Indonesia. The monkeys were fed an atherogenic diet containing 0.4 mg/cal cholesterol from egg yolk, 22% of calories from protein, 41% from fat (primarily beef tallow), and 37% from carbohydrate. The diet contained 43% saturated, 43% monounsaturated, and 14% polyunsaturated fat.
The monkeys consumed the atherogenic diet for 2 months and were divided into two groups (n=15 per group) with equivalent total plasma cholesterol (TPC) and HDL cholesterol (HDLC) concentrations. The diet of one group of monkeys was then changed so that one half the calories from fat was supplied by fish oil (purified lemuru fish oil from Indonesia; 18% eicosapentaenoic acid). The amount of fish oil fed to the monkeys supplied 0.5 g/kg per day of eicosapentaenoic acid. Since fish oil contains the antioxidant vitamin E (α-tocopherol), α-tocopherol was added to the diet without fish oil to equilibrate vitamin E concentrations between diets. The diet containing fish oil diet comprised 35% saturated, 36% monounsaturated, and 29% polyunsaturated fat.
The monkeys consumed their test diets for 6 months and were then shipped from Indonesia to the Comparative Medicine Clinical Research Center at Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC. They were quarantined for 3 months according to regulations of the Centers for Disease Control. The monkeys continued to eat their test diets during this time and for an additional 3 months after the quarantine period (total test diet time, 12 months) and were then studied. All procedures involving animals were conducted in compliance with state and federal laws, standards of the US Department of Health and Human Services, and guidelines established by the Institutional Animal Care and Use Committee.
Measurement of Blood Pressure and Plasma Lipids
Every 3 months during the experiment, blood pressures were measured with a model 8100 Dinamap Research Monitor12 and blood samples were collected from tranquilized, fasted monkeys for measurement of TPC,13 HDLC,14 and plasma triglyceride concentrations.15 Plasma concentrations of LDL cholesterol (LDLC) were determined indirectly by subtracting HDLC from the TPC concentrations. Samples taken in Indonesia were measured at the National Cardiac Center. Paired samples were sent to Bowman Gray for analysis. The correlation between results obtained at the National Cardiac Center and Bowman Gray was 0.90. Care was taken to ensure that individual plasma lipid concentrations returned to the levels measured in Indonesia at least 2 months before the monkeys were studied at Bowman Gray. Fatty acid distribution in the plasma of the monkeys was measured once, at the end of the dietary protocol just before necropsy.16 Fatty acid composition of the diet was measured by similar methods in Indonesia before the fish oil was sent to Bowman Gray.
On the day of the study, monkeys were anesthetized with ketamine hydrochloride (10 mg/kg IM) and sodium pentobarbital (30 mg/kg IV to effect). An endotracheal tube was inserted, and the lungs were ventilated with room air supplemented with 95% oxygen. Blood gases were monitored and maintained within the normal range. Catheters were inserted into the femoral artery for measurement of blood pressure and heart rate and into the femoral vein for administration of drugs. External patch leads were attached to the monkeys for monitoring the lead II configuration of the ECG. A thoracotomy was done through the fourth intercostal space, and the left side of the heart was exposed by packing back the left lung lobes and incising the pericardial sac. The left anterior descending coronary artery (LAD) was ligated distal to the first diagonal branch with a 0-Dexon suture that occluded both the artery and vein. Ligation was confirmed by lack of a pulse in the distal LAD and the appearance of a bluish color in the affected myocardium. The LAD was occluded for 1 hour, and the ligature was then removed. Reperfusion was confirmed by detection of a pulse in the distal LAD after removal of the ligature and by a return of more normal color to the myocardium.
Two hours after the onset of reperfusion, the monkeys were euthanatized by sodium pentobarbital 80 mg/kg IV. The heart was quickly removed (within 30 seconds after the pentobarbital was administered) and placed in oxygenated, chilled (20°C) Krebs-HEPES buffer (pH 7.4).11 The entire length of the epicardial LAD was quickly and carefully dissected from the myocardium and placed in fresh room-temperature Krebs-HEPES buffer. The LAD was then divided perpendicular to the long axis into distal (exposed to ischemia and reperfusion) and proximal (used to measure levels of superoxide anion production in unmanipulated artery) segments. Each segment (proximal and distal) was divided perpendicular to the long axis into two ring segments, and one segment was denuded of luminal endothelium by gentle rubbing with a closed forceps. Removal of endothelium was confirmed by light microscopic evaluation of random samples taken throughout the experiment. Four artery samples were analyzed for each monkey: two proximal (denuded and nondenuded) and two distal (denuded and nondenuded).
Samples (1 g each) of myocardium were taken from the distal anterior wall (ischemic) and proximal posterior wall (nonischemic) of the left ventricle and placed in room-temperature Krebs-HEPES buffer. The 1-g samples of myocardium were each divided into 10 (0.1-g) samples, which were assayed for superoxide dismutase by chemiluminescence, and the results were reported as the mean counts of these 0.1-g samples. One untreated atherosclerotic monkey was used to determine the optimum tissue sample size. Five 1-g full-thickness myocardial samples were counted whole; divided into 10, 5, 4, 3, and 2 equal-weight samples; and then recounted. All counts are reported as counts per minute per gram of tissue. Counts among samples were similar (10 000 to 11 000 cpm/g). Counts in tissue samples divided into segments were greater than the whole (from 10% to 25% higher). There were no differences in total counts (no further increase) from myocardium divided into 0.2- or 0.1-g samples. Therefore, myocardial samples of 0.1 to 0.2 g were determined to be optimal because counts were maximized and internal quenching was minimized. Total time from euthanatization of the monkey to detection of chemiluminescence in the myocardial samples was very consistent between monkeys (approximately 20 minutes). There was minimal handling of the tissues during this time.
To confirm that myocardial samples had been ischemic, random samples of myocardium from adjacent sampling sites were incubated with a 30% solution of triphenyltetrazolium chloride. A brick-red color denoted nonischemic sections, and a tan color denoted ischemic sections of myocardium.
Detection of Superoxide Radicals
Chemiluminescence of lucigenin was used to detect treatment effects on superoxide production by tissues, as described previously.11 17 The chemical specificity of this light-yielding reaction for superoxide ion was reported previously.18 In this study, sensitivity and specificity of this assay were determined with xanthine (100 to 400 nmol/L) and xanthine oxidase (0.002 U) to generate oxygen ions with or without superoxide dismutase (SOD) (0.5 U/mL). These validation studies have been described in detail.11 17 19 Combinations of xanthine and xanthine oxidase produced transient chemiluminescence signals in a manner dependent on the concentration of xanthine; these signals were abolished by SOD.
Superoxide yields from xanthine–xanthine oxidase reactions were determined as described previously.20 Under the conditions of our assay, the yield of superoxide anion was found to be equal to 30% of the total xanthine present. This value was used to calibrate the lucigenin signal obtained during reactions with xanthine–xanthine oxidase. Total luminescence was quantified by integration of the areas under the curves generated in experiments. The correlation between chemiluminescence and superoxide anion generated with a simple fit was 0.99.
Chemiluminescence was detected by tissue samples after these samples were incubated in Krebs-HEPES buffer maintained at 37°C for 30 minutes. The tissue samples were then gently transferred to glass scintillation vials containing 0.25 mmol/L lucigenin that had been dark adapted.11 The samples were counted in a single-channel scintillation counter in the out-of-sequence mode. Counts were obtained at 2-minute intervals at room temperature. Vials containing lucigenin plus components (with the exception of tissue) were counted, and these blank values were subtracted from the chemiluminescence signals obtained from the tissue. Chemiluminescence was reported as nanomoles of oxygen anion per milligram dry wt per minute. Lucigenin-mediated chemiluminescence with segments of aorta also was measured in the presence of SOD (6 μmol/L) or after a 30-minute incubation with the hydroxyl radical scavenger mannitol (10 mmol/L) to evaluate the specificity of this reaction. Addition of mannitol abolished the chemiluminescence signal. The amount of time between preparation of the aortic tissue and completion of chemiluminescence assays was consistent (approximately 45 minutes). There was minimal handling of tissue during this time.
Determination of Extent of Atherosclerosis
Segments of proximal and distal LAD were placed in 10% neutral buffered formalin for determination of plaque size (measured as intimal area).21 The artery segments were dehydrated in increasing concentrations of ethanol, placed in paraffin blocks, and cut longitudinally to the long axis of the artery. Histological sections were stained with Verhoeff–van Gieson’s stain and projected, and the cross-sectional area of the plaque lesion was measured with a digitizer. The extent of atherosclerosis was expressed as the mean cross-sectional area of intima in millimeters squared.
Reported values are mean±SEM. To determine the effect of treatment on experimental end points (plasma lipids, blood pressure, superoxide anion production, and plaque size), mean results were compared by Student’s t test. Simple correlations were used to compare two lipid concentrations from the two laboratories or correlations between chemiluminescence and superoxide anion generation. The accepted level of significance was P<.05.
Blood Pressure, Plasma Lipids, and Extent of Atherosclerosis
Plasma concentrations of TPC, LDLC, and triglycerides were lower in monkeys fed fish oil than in those not fed fish oil (all P values <.05, Table 1⇓). Consumption of the fish oil–containing diet increased the plasma concentration of 20- and 22-carbon ω-3 fatty acids, primarily at the expense of 18-carbon fatty acids (P<.05, Table 2⇓). There were no differences between treatment groups in resulting mean arterial pressures (75±6 mm Hg versus 78±8 mm Hg, P>.05).
There were no differences in plaque extent between treatment groups in the proximal or distal LAD (P>.05, Table 1⇑). Atherosclerosis extent was smaller in the distal than proximal segments of LAD in both treatment groups (P<.05, Table 1⇑).
Superoxide anion concentrations in the nonischemic myocardium were similar between treatment groups of monkeys (P>.05, Fig 1⇓). Dietary supplementation with fish oil prevented the increase in superoxide anion production in the myocardium after ischemia and reperfusion (P<.05, Fig 1⇓). Fish oil reduced superoxide anion production in nonischemic coronary arteries (P<.05, Fig 2⇓). Fish oil abolished the increase in superoxide anion production in coronary arteries after ischemia and reperfusion (P<.05, Fig 2⇓). Denudation of coronary arteries diminished the effects of fish oil on superoxide anion production (P>.05, Fig 3⇓). SOD applied extracellularly inhibited the lucigenin signal by approximately 70%.
The three major findings of this study in nonhuman primates are that (1) supplementing the diet with fish oil resulted in increased plasma concentrations of ω-3 fatty acids associated with reduced superoxide anion production in the myocardium after ischemia and reperfusion; (2) despite a lack of effect on plaque size, fish oil reduced levels of superoxide anions in atherosclerotic coronary arteries; and (3) fish oil prevented the increase in superoxide anion production in atherosclerotic coronary arteries after ischemia and reperfusion.
Effects of Fish Oil on Superoxide Anions
It is conceivable that fish oil reduced superoxide anion production in the myocardium and coronary arteries by separate mechanisms. It has been proposed9 that invasion of inflammatory cells into the ischemic myocardium and formation of oxygen free radicals during reperfusion may promote cell death in the myocardium. Inhibition of neutrophil infiltration into the ischemic myocardium has been shown to reduce infarct size.22 Fish oil has been reported to reduce prostaglandin and leukotriene formation23 24 and decrease free oxygen radical formation by neutrophils.10 In our study, dietary supplementation with fish oil increased the plasma concentrations of ω-3 fatty acids. Therefore, ω-3 fatty acids may favorably modify the inflammatory reaction in the myocardium and, through this mechanism, reduce superoxide anion production in the myocardium after ischemia and reperfusion.
This mechanism for inhibition of superoxide anion production by the myocardium may not apply to the coronary vasculature. Effects of fish oil on arterial superoxide anion production cannot be explained by effects on plaque size. However, there are numerous potential sources of superoxide anions in the wall of atherosclerotic coronary arteries that could be affected by fish oil. These include arachidonate metabolism, the cytochrome P450 system, electron transport, and xanthine dehydrogenase.11 Pagano et al25 reported recently that nicotinamide adenine dinucleotide phosphate oxidase is an important source of superoxide anion in the intima and adventitia of rabbit aorta. Additionally, superoxide anions may be produced by macrophages and other inflammatory cells in the intima and media of atherosclerotic arteries. Fish oil may have affected superoxide anion production by several of these mechanisms.
Denudation of coronary arteries eliminated the superoxide anion–inhibiting effects of fish oil, which suggests that fish oil affected superoxide anion production primarily in the endothelium or cells attached to the endothelium. This hypothesis is supported by recent evidence that the fatty acids of fish oil incorporate into the cell membrane phospholipids and cause competitive inhibition of arachidonate metabolism.23 24 However, denudation may have removed underlying or adherent inflammatory cells along with endothelial cells. Fisher et al10 noted that fish oil reduces oxygen radical formation by polymorphonuclear leukocytes. Huth et al26 recently reported that ω-3 fatty acids reduce adhesion of monocytes and platelets that may be sources of superoxide anion. Fish oil reduced plasma TPC and LDLC concentrations in monkeys in the present experiment. It is possible that lowering plasma concentrations of lipoproteins reduced the lipid composition, particularly concentrations of oxidized LDL in the vessel wall, and thereby improved endothelium-mediated vasodilation.
Ischemic damage to the myocardium is a major concern during cardiopulmonary bypass, coronary reperfusion in acute myocardial infarction, thrombolytic therapy, and coronary angioplasty. One of the major reasons that this study was done in atherosclerotic monkeys was to model more closely the disease state of the arteries in patients undergoing bypass and angioplasty procedures. Culp et al7 showed that dietary fish oil added to the diet for 4 to 6 weeks resulted in smaller infarctions after coronary artery occlusion. Similarly, Oskarsson et al6 reported that fish oil could reduce myocardial infarction size after 90 minutes of coronary artery occlusion followed by 6 hours of reperfusion. In both studies, questions were raised about the potential mechanisms by which fish oil limits damage to the myocardium after ischemia and reperfusion. Although infarct size was not measured in the present experiment, results indicate that suppression of superoxide anion production may explain in part the protective effects of fish oil on myocardial damage after reperfusion.
Results of recent studies in humans indicate that dietary supplementation with fish oil reduces the rate of restenosis after angioplasty.27 28 Active oxygen species have been shown to stimulate growth of vascular smooth muscle cells.29 Therefore, the inhibitory effects of fish oil on superoxide anion production may explain in part its beneficial effects on restenosis.
Reduction of superoxide anion production in atherosclerotic coronary arteries may have important implications in the pathogenesis of vasospasm and thrombosis. Nitric oxide is released by endothelial cells in response to various neurohumoral stimuli.30 Nitric oxide is rapidly inactivated by superoxide radical–generating systems31 and is protected by SOD.32 Nitric oxide has important protective actions in the vascular wall, including inhibition of smooth muscle contraction, smooth muscle proliferation, and platelet aggregation.30 Therefore, treatments that reduce superoxide anion production may reduce the risk of heart disease by promoting nitric oxide activity. Shimokawa et al8 showed that fish oil supplementation augments endothelium-mediated dilation of porcine coronary arteries, possibly through nitric oxide–mediated mechanisms. The results of the present experiment indicate that inhibition of vascular superoxide anion production may in part explain these effects.
The monkeys were transported from Indonesia to North Carolina in the middle of the experiment. Transport is a stressful situation for monkeys. Stress has been shown to be able to affect the extent of atherosclerosis in coronary arteries.33 However, it is unlikely that the stress of travel affected interpretation of the present results because both groups of monkeys were stressed. Additionally, we have shown that the effects of stress on endothelial function of atherosclerotic coronary arteries are transient.34
We used lucigenin to detect superoxide anion. The specificity of this technique was established previously.11 17 Lucigenin chemiluminescence is insensitive to hydroxyl radical, hydrogen peroxide, and nitric oxide. Additionally, the chemiluminescence of lucigenin has an advantage over other methods because it lends itself to measurements of superoxide production in intact vessels. By this method, it is not possible to determine whether superoxide dismutase production or scavenging is affected. A potential limitation of the present work is that the lucigenin signal produced by intact vessels is only 70% inhibited by SOD applied extracellularly. The explanation for this remains unclear but may relate to the fact that lucigenin produces chemiluminescence on the basis of both intracellular and extracellular superoxide anion. Therefore, SOD applied extracellularly may not completely inhibit the intracellular superoxide anion.
The percent of eicosapentaenoic acid in the fish oil diet was higher than that reported in other experiments.5 6 7 This is due, most likely, to the purified nature of the lemuru fish oil we used. Therefore, lemuru fish oil may be particularly useful in the prevention of myocardial damage after reperfusion.
The results of the present study indicate that consumption of a diet rich in fish oil inhibits production of superoxide anions in the coronary arteries and myocardium of atherosclerotic monkeys. Reduction in superoxide anion production may explain, in part, the reported beneficial effects of fish oil on myocardial infarction size and coronary artery vasomotion.
This work was supported in part by grants PO1-RR-08562 (National Center for Research Resources) and PO1-HL-45666, R01-HL-48667, and RO1-HL-39006 (National Heart, Lung, and Blood Institute) from the National Institutes of Health, Bethesda, Md. The authors thank Jamie Fox and Yuichi Ohara for their expert assistance with the lucigenin assay and other experimental procedures and Karen Potvin Klein for editing this manuscript.
Presented in part at the 68th Scientific Sessions of the American Heart Association, November 8-11, 1993, Atlanta, Ga, and published in abstract form (Circulation. 1993;88[suppl I]:I-466).
- Received July 7, 1994.
- Revision received August 31, 1994.
- Accepted September 23, 1994.
- Copyright © 1995 by American Heart Association
Connor WE, Harris WS, Goodnight SH. The hypolipaemic and antithrombotic effects of salmon oil. In: Perkins EG, Visek WJ, eds. Dietary Fats and Health. Champaign, Ill: American Oil Chemists Society; 1983:518.
Harris WS. Fish oils and plasma lipid and lipoprotein metabolism in humans: a critical review. J Lipid Res. 1989;30:785-807.
Fincham JE, Gouws E, Woodroof CW, van Wyk MJ, Kruger M, Smuts CM, van Jaarsveld PJ, Taljaard JJF, Schall R, Strauss JA deW, Benadé AJS. Atherosclerosis: chronic effects of fish oil and a therapeutic diet in nonhuman primates. Arterioscler Thromb. 1991;11:719-732.
Shimokawa H, Lam JY, Chesbro JH, Bowie EJW, Vanhoutte PM. Effects of dietary supplementation with cod-liver oil on endothelium-dependent responses in porcine coronary arteries. Circulation. 1987;76:898-905.
Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest. 1993;91:2546-2551.
Corbett WT, Schey HM, Lehner NDM, Greene AW. Standardization method for recording blood pressure in anesthetized Macaca fascicularis. Lab Anim. 1981;15:37-40.
Allain CC, Poon LS, Chan CSG, Richmond W, Fu PC. Enzymatic determination of total serum cholesterol. Clin Chem. 1974;20:470-475.
Manual of Laboratory Operations. Lipid Research Clinics Program: Vol. 1, Lipid and Lipoprotein Analysis. Bethesda, Md: National Heart, Lung, and Blood Institute, National Institutes of Health; 1974. US Dept of Health, Education and Welfare publication 75-628.
Fossati P, Principe L. Serum triglycerides determined colorimetrically with an enzyme that produces hydrogen peroxide. Clin Chem. 1982;28:2077-2080.
Omar HA, Cherry PD, Mortelliti MP, Burke-Wolin T, Wolin MS. Inhibition of coronary artery superoxide dismutase attenuates endothelium-dependent and -independent nitrovasodilator relaxation. Circ Res. 1991;69:601-608.
Fridovich I. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. J Biol Chem. 1970;245:4053-4057.
Romson JL, Hook BG, Kunkel SL, Abrams GD, Schork MA, Lucchesi BR. Reduction of the extent of ischemic myocardial ischemia and reflow in dogs. Am J Physiol. 1986;251:H314-H322.
Lee TH, Mencia-Huerta JM, Shih C, Corey EJ, Lewis RA, Austen KF. Effects of exogenous arachidonic, eicosapentaenoic, and docosahexaenoic acids on the generation of 5-lipoxygenase pathway products by ionophore-activated human neutrophils. J Clin Invest. 1984;74:1922-1933.
Pagano PJ, Ito Y, Tomheim K, Gallop P, Cohen RA. The superoxide generating system in the rabbit aorta resembles NAD(P)H oxidase. Circulation. 1993;88(suppl I):I-467. Abstract.
Huth PJ, Foxall TL, Nicolosi RJ, Weiner EJ. Effects of levels and ratios of dietary linoleic acid and marine or vegetable oil n-3 fatty acids on monocyte and platelet adhesion, hemostatic indices and plasma lipoproteins in cynomolgus monkeys. Circulation. 1993;88(suppl I):I-456. Abstract.
Rao GN, Burke BC. Active oxygen species stimulate vascular smooth muscle growth and proto-oncogene expression. Circ Res. 1992;70:593-599.
Flavahan NA. Atherosclerosis or lipoprotein-induced endothelial dysfunction: potential mechanisms underlying reduction in EDRF/nitric oxide activity. Circulation. 1992;85:1927-1938.
Rubanyi GM, Vanhoutte PM. Oxygen-derived free radicals, endothelium, and responsiveness of vascular smooth muscle. Am J Physiol. 1986;250:H815-H821.
Ignarro LJ, Byrns RE, Buga GM, Wood KS, Chaudhuri G. Pharmacologic evidence that endothelium-derived relaxing factor is nitric oxide: use of pyrogallol and superoxide dismutase to study endothelium-dependent and nitric oxide-elicited vascular smooth muscle relaxation. J Pharmacol Exp Ther. 1988;244:181-189.
Williams JK, Kaplan JR, Manuck SB. Effects of psychosocial stress on endothelium-mediated dilation of atherosclerotic arteries of cynomolgus monkeys. J Clin Invest. 1993;92:1819-1823.
Manuck SB, Kaplan JR, Matthews KA. Behavioral antecedents of coronary heart disease and atherosclerosis. Arteriosclerosis. 1986;6:2-14.