Dietary Correction of Hypercholesterolemia in the Rabbit Normalizes Endothelial Superoxide Anion Production
Background We have shown that hypercholesterolemia increases vascular superoxide anion (O2−) production, which could be responsible for augmented inactivation of endothelium-derived vascular relaxing factor. We sought to determine whether this increased vascular O2− production is due to infiltration of macrophages into the intima and whether dietary treatment of hypercholesterolemia normalizes O2− production.
Methods and Results A specific and sensitive assay for O2− based on chemiluminescence of lucigenin was used; the amount of O2− produced by vascular ring segments was quantified based on known quantities of O2− produced by xanthine–xanthine oxidase standards. O2− production of aortic segments from normal rabbits (n=9), cholesterol-fed rabbits (1% cholesterol diet for 1 month, n=7), and rabbits fed a 1% cholesterol diet for 1 month followed by a normal diet for 1 month (regression rabbits, n=5) was measured. At the end of these diets, serum cholesterol levels were 1.5±0.2, 26.0±3.9, and 1.8±0.5 mmol/L (58±6, 1000±150, and 71±19 mg/dL) in the normal, cholesterol-fed, and regression animals, respectively. Vessels from normal rabbits with endothelium produced 0.32±0.06 nmol O2−/mg dry wt per minute, whereas those without endothelium produced approximately twice as much O2− (0.66±0.12 nmol O2− mg dry wt per minute. Vessels with endothelium from cholesterol-fed rabbits produced 4.5-fold more O2− than vessels from normal animals. This increased production of O2− was normalized by endothelial removal. This increased production of O2− was not due to infiltration of macrophages in the intima, because there was no correlation between vascular O2− production and macrophage infiltration assessed by immunohistochemistry with use of a specific antibody against rabbit macrophage. O2− production by vessels from regression rabbits was similar to that observed in normal animals, and as in the normal rabbits, endothelial removal increased O2− production. Aortic rings from these animals also were studied in organ chambers. Dietary lowering of cholesterol dramatically improved vasodilator responses to acetylcholine and A23187 (P<.05 versus cholesterol-fed rabbits).
Conclusions Dietary lowering of cholesterol not only improves endothelium-dependent vascular relaxation but also normalizes endothelial O2− production. Decreases of O2− production by dietary lowering of cholesterol not only may improve vasomotor control but also may improve other aspects of vascular integrity in atherosclerosis.
The endothelium modulates vasomotor tone by release of an endothelium-derived relaxing factor now known to be nitric oxide1 or a closely related compound.2 This important function of the endothelium is impaired in the presence of several common diseases, including hypercholesterolemia and atherosclerosis.3 4 5 6 The mechanisms underlying this abnormality of vascular function have been the subject of substantial research and probably are multifactorial. One contributing factor may be excessive production of superoxide anion (O2−),7 8 which can inactivate nitric oxide.9 Treatment with polyethylene-glycol superoxide dismutase improves endothelium-dependent vascular relaxation in cholesterol-fed rabbits, providing indirect evidence for increased vascular O2− production in hypercholesterolemia.8 Recently, using lucigenin chemiluminescence, we have found that aortas from hypercholesterolemic rabbits produce substantially more O2− than control rabbit aortas.7 In these studies, endothelial removal normalized O2− production, indicating that the source of the O2− was not the smooth muscle layer but the endothelial cell itself or monocyte-macrophages closely associated with the endothelium. Oxypurinol also normalized O2− production in these vessels, suggesting a role of xanthine oxidase as a source of O2−.
There has been substantial interest in therapeutic means to correct alterations of endothelial function in hypercholesterolemia. In hypercholesterolemic animals, cholesterol lowering via either dietary or pharmacological means can improve endothelium-dependent vascular relaxations.10 11 Recent studies in human subjects have shown that large coronary artery responses to acetylcholine can be converted to a modest degree of vasodilatation by 6 months of aggressive lipid-lowering therapy.12
If alteration of endothelium-dependent vascular relaxation is in part due to excess production of O2−, then it would follow that improvement of hypercholesterolemia would not only improve relaxations to endothelium-dependent vasodilators but also lower O2− production. The present experiments were performed to test this hypothesis. A second aim of this study was to determine whether the increase in O2− production in cholesterol-fed rabbit aortas correlated with the degree of macrophage infiltration.
Animal Model and Vessel Preparation
Three groups of New Zealand White rabbits were studied. The control group of animals (n=9) was fed standard rabbit chow. Plasma cholesterol level of control animals was 1.5±0.2 mmol/L (58±6 mg/dL). Age-matched rabbits (n=7) were fed a high cholesterol (1%) diet for 1 month. During this time, the plasma cholesterol increased to 26.0±3.9 mmol/L (1000±150 mg/dL) (P<.001 versus control). Other rabbits (regression, n=5) were fed a 1% cholesterol diet for 1 month and were then placed on a normal diet for the ensuing 1 month. Plasma cholesterol at the end of this time had decreased to 1.8±0.5 mmol/L (71±19 mg/dL) (P=NS versus control). On the day of study, the animals were killed by an overdose of intravenous sodium pentobarbital. The chest was then rapidly opened and the descending thoracic aorta removed. The vessel was placed in chilled Krebs-HEPES buffer, cleaned of excessive adventitial tissue, and cut into 5-mm ring segments. In some vessels, the endothelium was removed by inserting the closed tips of metal hemostat forceps into the ring segment and rolling it gently on moistened filter paper.
Measurements of Vascular Superoxide Anion Production
O2− production was measured using lucigenin chemiluminescence. The details of this assay have been published previously.7 13 14 Briefly, after preparation, the vessels were placed in a modified Krebs-HEPES buffer (mmol/L content: NaCl, 99.01; KCl, 4.69; CaCl2, 1.87; MgSO4, 1.20; K2HPO4, 1.03; NaHCO3, 25.0; Na-HEPES, 20.0; and glucose, 11.1; initially gassed with 95% O2 and 5% CO2, pH 7.4) and allowed to equilibrate for 30 minutes. Scintillation vials containing 2 mL of Krebs-HEPES buffer with 0.25 mmol/L lucigenin were placed into a scintillation counter switched to the out-of-coincidence mode. After 15 minutes, background counts were recorded, and a vascular segment then was added to the vial. Scintillation counts then were recorded 15 minutes later and the respective background counts subtracted. The vessels then were dried by placing them in a 90°C oven for 3 hours. Lucigenin scintillation counts were converted to absolute amounts of O2− with use of known quantities of xanthine and xanthine oxidase as described previously.7
Isometric Tension Studies
Ring segments from the same animals also were studied in organ chambers filled with Krebs-Henseleit buffer (mmol/L content: NaCl, 118.3; KCl, 4.69; CaCl2, 1.87; MgSO4, 1.20; K2HPO4, 1.03; NaHCO3, 25.0; and glucose, 11.1; pH 7.40). The buffer was aerated continuously with 95% O2 and 5% CO2 and maintained at 37°C. Each segment was suspended by means of metal stirrups, one of which was connected to a transducer to allow measurement of isometric tension. During the following hour, the vessels were gradually stretched to a resting tension of 5 g. In preliminary experiments, we found that this was an optimum resting tension for development of active contractile tone. After this, the vessels were constricted with 1 μmol/L phenylephrine. After development of a stable degree of active tension, the vessels were exposed to cumulative concentrations of either acetylcholine (1 nmol/L to 3 μmol/L) or the calcium ionophore A23187 (1 nmol/L to 3 μmol/L).
Additional studies were performed to attempt to determine the role of macrophages in the increased O2− production in cholesterol-fed rabbit aortas. Sixteen segments from 8 hypercholesterolemic rabbits (plasma cholesterol=29.6±5.5 mmol/L [1140±210 mg/dL]) and 4 segments from 2 regression rabbits (mean plasma cholesterol=1.8 mmol/L [68 mg/dL]) were studied. After measurement of O2− production, the segments were weighed (wet weight). Preliminary experiments from 12 aortic segments revealed that the dry weight of the aorta was 22% of the wet weight. We therefore used this value to estimate the dry weight of these vessels. After weighing, the aortas were fixed in 4% paraformaldehyde buffered with 0.1 mol/L NaPO4 (pH 7.4) for 3 hours at 4°C, cryoprotected in 15% sucrose PBS overnight, embedded in optimal cutting temperature compound (OCT, Miles Laboratories), frozen in liquid nitrogen, and stored at −70°C. Cryosections (7 μm) were thaw-mounted onto Superfrost/Plus slides (Fisher Scientific), refrozen, and stored at −70°C with desiccant until use. Immunohistochemistry was performed using the Vectastain ABC-AP system (Vector Laboratories) essentially as described by the manufacturer. Cryosections were pretreated with acetone and 1% gelatin-PBS and incubated at room temperature for 60 minutes with a mouse anti-rabbit macrophage monoclonal antibody (RAM 11 [Reference 15], Dako Platts, 1:50) followed by incubation with a biotinylated horse anti-mouse secondary antibody (Vector Laboratories, 1:400) for 30 minutes. The tissue sections were subsequently stained with use of Vector substrate kit I so that the final reaction product appeared red. The slides were counterstained with hematoxylin. The immunohistochemistry experiments were controlled by incubating some sections with the secondary antibody only.
To relate O2− production to the extent of macrophage infiltration, we examined both O2− production and macrophage infiltration in individual vascular segments. An additional 10 rabbits (8 hypercholesterolemic rabbits and 2 regression rabbits) were used in these experiments. We defined the extent of macrophage infiltration into four grades (Fig 1⇓). Macrophage grading was performed by one of the authors in a blinded fashion so that he was unaware of the lucigenin signal obtained from the vascular segments studied. Eight sections of the vascular segment in which O2− production had been determined were examined. A score was assigned to each vascular segment based on analysis of all sections from that segment, and the score assigned represented the highest number of macrophages observed: grade 0, no macrophages observed; grade 1, 1 to 3 macrophages in the histological section; grade 2, the infiltration of macrophage occupied less than one fourth of the circumferential area of lumen; grade 3, the infiltration of macrophage occupied one fourth to one half of the circumferential area of lumen; and grade 4, the infiltration of macrophage occupied more than one half of the circumferential area of lumen.
All reagents were purchased from Sigma Chemical Co except when specified.
Data are presented as mean±SEM. Differences in O2− production and vascular relaxation were compared using ANOVA, and when differences between groups were indicated, a Scheffé’s post hoc analysis was used. Probability values <.05 were considered significant.
Superoxide Anion Production
In aortic segments from normal rabbits (n=9), O2− production estimated by measuring chemiluminescence 15 minutes after exposure to lucigenin was 0.32 nmol/mg tissue per minute and was increased about twofold by removal of the endothelium (P<.05) (Fig 2⇓).
In aortic segments from hypercholesterolemic rabbits (n=7) with endothelium, O2− production was 4.5 times greater than that observed in normal aortas (P<.01) (Fig 2⇑). Furthermore, in contrast to the findings in normal aortas, removal of the endothelium from these hypercholesterolemic aortas did not increase but dramatically decreased O2− production (P<.05) (Fig 2⇑). O2− production of segments of hypercholesterolemic vessels without endothelium was similar to normal segments without endothelium.
In the regression group (n=5), O2− production was identical to values observed in normal rabbits (Fig 2⇑). Furthermore, as observed in normal vessels, O2− production was increased significantly by removal of the endothelium (Fig 2⇑). Thus, the endothelium in rabbits fed initially a high cholesterol and then a normal diet also appears to exert a protective role in limiting the total production of O2− from the vessel wall.
Endothelium-Dependent Vascular Relaxations
The degree of preconstriction by phenylephrine averaged 8.0±0.7 g, 8.6±0.6 g, and 8.4±0.4 g (P=NS) in normal, cholesterol-fed, and regression rabbits, respectively. Relaxations of vessels from the cholesterol-fed group to acetylcholine and the calcium ionophore A23187 were markedly reduced compared with control animals. After dietary lowering of cholesterol, responses to these agonists were markedly improved but were not as great as those observed in normal animals (Fig 3⇓).
Relation of Superoxide Anion Production to Macrophage Infiltration in the Intima
As can be seen in Fig 4⇓, there was no correlation between vascular O2− production and macrophage infiltration into the intima. Furthermore, in several vascular segments with very high O2− production, no macrophages were observed.
In the present experiments, diet-induced hypercholesterolemia was associated with a 4.5-fold increase in the vascular production of O2−. As in a previous study,7 endothelial removal normalized O2− production, suggesting that the source of the O2− was the endothelium or cells closely associated with the endothelium. The new findings are that this increased O2− production probably is not due to infiltration of macrophages into the intima and that with dietary correction and lowering cholesterol, the vascular production of O2− was normalized. In addition to correction of O2− production, endothelium-dependent vascular relaxations to acetylcholine and the calcium ionophore A23187 were greatly improved.
Lucigenin chemiluminescence, used in the present experiments, has proven to be useful for studies of vascular O2− production.7 13 14 As previously noted, this methodology is quite specific for O2− production. Chemiluminescence is not produced by H2O2 in concentrations lower than 1 mmol/L (approximately 1 million times higher than O2−). The technique does not detect hydroxyl radical, singlet oxygen, or nitroxyl anion in concentrations less than 1 μmol/L.7 14 Luminol has been used previously to measure O2−; however, it also may detect other radicals such as hydroxyl and nitric oxide.16 The reduction of ferricytochrome c has been used to detect O2−,17 but we have found it less sensitive than lucigenin. Unlike ferricytochrome c, lucigenin also detects both intracellular and extracellular O2−13 18 and thereby provides a more accurate estimate of total vascular production and release of the radical.
The finding that cholesterol lowering was associated with an improvement in endothelium-dependent vascular relaxation is in keeping with prior reports in monkeys,10 rabbits,11 and more recently in humans.12 As the O2− production was also decreased, the findings are also compatible with the hypothesis that excess O2− may contribute to alterations of endothelium-dependent vascular relaxation. O2− inactivates nitric oxide, leading to the formation of nitrate and nitrite, which are vasoinactive except in very high concentrations. Because the O2− in hypercholesterolemia appears to be released by the endothelium, it is reasonable to assume that the inactivation of nitric oxide also may occur within the endothelium or shortly after the nitric oxide has been released from the endothelium. The chemical reaction betwen O2− and nitric oxide also may lead to the production of the peroxynitrite radical,19 20 which is vasoactive21 but has an extremely short biological half-life, which reduces its capacity to diffuse to the adjacent vascular smooth muscle.
In the present studies, as in previous work,7 we found that removal of the endothelium from cholesterol-fed rabbits decreased O2− production. This finding suggests that the source of the O2− is either the endothelium or a cell type closely associated with the endothelium that is also removed in the denudation process. This consideration led us to examine the degree of macrophage infiltration in a separate set of vessels from cholesterol-fed rabbits and rabbits after dietary lowering of cholesterol. We also were able to obtain estimates of O2− production from the same vascular segments so that a correlation between O2− production and macrophage infiltration could be obtained. As is evident from Fig 4⇑, there was no obvious relation between vascular O2− production and the number of macrophages present. In a large number of these segments, no macrophages were observed. It is quite likely that this absence of macrophages was due to the short-term cholesterol feeding period and that had a longer feeding period been used, macrophages would have been more consistently observed. Nevertheless, these findings are compatible with the concept that the endothelium can serve as a source of O2− production.
In the present studies, cholesterol lowering markedly improved but did not completely correct endothelium-dependent vascular relaxation. In two previous studies in which we sought to lower O2− production, an almost identical effect was observed. Treatment of cholesterol-fed rabbits with polyethylene-glycolated superoxide dismutase8 and prevention of O2− production by oxypurinol7 improved endothelium-dependent vascular relaxation to about 75% or 80% of that observed in control animals. In the present experiments and in the studies with oxypurinol, direct evidence was obtained that the interventions completely normalized O2− production. It is therefore interesting to speculate that a portion (approximately 75% to 80%) of the defect in endothelium-dependent vascular relaxation is due to O2− inactivation of nitric oxide and that other mechanisms may account for the remainder of the defect.
The present studies have focused on control of vasomotor tone; however, reducing vascular O2− production by cholesterol lowering may have other therapeutic implications. The O2− participates in oxidation and modification of low-density lipoprotein.22 Other radicals derived from O2− and peroxynitrite such as the hydroxyl radical and hydrogen peroxide may modify membrane lipids via formation of peroxy radicals23 and are chemoattractant for neutrophils.24 Recently, it has become clear that cellular redox state plays an important role in modulation of gene transcription25 26 27 28 and that at least one endothelial cell adhesion molecule (VCAM-1) involved in the atherogenic process may be transcriptionally regulated by oxidant stress.29 Oxygen radicals also may induce vascular smooth muscle growth and proliferation,30 an important component of atherogenesis. Thus, reduction of O2− production by cholesterol lowering not only may improve endothelial regulation of vasomotion but probably has several other beneficial effects with regard to vascular homeostasis.
This work was supported by National Institutes of Health grants HL-32717, HL-39006, and HL-48667 and a Merit Grant from the Veterans Administration. We acknowledge the assistance of Ann B. Peterson and Cynthia M. Curry in the preparation of the manuscript.
- Received December 13, 1994.
- Revision received February 1, 1995.
- Accepted February 10, 1995.
- Copyright © 1995 by American Heart Association
Freiman PC, Mitchell GC, Heistad DD, Armstrong ML, Harrison DG. Atherosclerosis impairs endothelium-dependent vascular relaxation to acetylcholine and thrombin in primates. Circ Res. 1986;58:783-789.
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 200110. Circ Res. 1986;58:305-309.
Verbeuren TJ, Jordaens FH, Zonnekeyn LL, Van Hove CE, Coene MC, Herman AG. Effect of hypercholesterolemia on vascular reactivity in the rabbit. Circ Res. 1986;58:552-564.
Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest. 1993;91:2546-2551.
Mügge A, Elwell JH, Peterson TE, Hofmeyer TG, Heistad DD, Harrison DG. Chronic treatment with polyethylene glycolated superoxide dismutase partially restores endothelium-dependent vascular relaxations in cholesterol-fed rabbits. Circ Res. 1991;69:1293-1300.
Rubanyi GM, Vanhoutte PM. Oxygen-derived free radicals, endothelium and responsiveness of vascular smooth muscle. Am J Physiol. 1986;250:H815-H821.
Harrison DG, Armstrong ML, Freiman PC, Heistad DD. Restoration of endothelium-dependent relaxation by dietary treatment of atherosclerosis. J Clin Invest. 1987;80:1808-1811.
Osborne JA, Lento PH, Siegfried MR, Stahl GL, Fusman B, Lefer AM. Cardiovascular effects of acute hypercholesterolemia in rabbits. J Clin Invest. 1989;83:465-473.
Cherry PD, Omar HA, Farrell KA, Stuart JS, Wolin MS. Superoxide anion inhibits cGMP-associated bovine pulmonary arterial relaxation. Am J Physiol. 1990;259(Heart Circ Physiol 28):H1056-H1062.
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.
Tsukada T, Rosenfeld M, Ross R, Gown AM. Immunocytochemical analysis of cellular components in atherosclerotic lesions: use of monoclonal antibodies with the Watanabe and fat-fed rabbit. Arteriosclerosis. 1986;6:601-613.
Fridovich I. Quantitative aspects of the production of superoxide anion radical by milk xanthine oxidase. J Biol Chem. 1970;245:4053-4057.
Melinn M, McLaughlin H. Nitroblue tetrazolium reduction in lymphocytes. J Leukoc Biol. 1987;41:325-329.
Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Apparent hydroxyl radical production by peroxynitrite: implications for endothelial injury from nitric oxide and superoxide. Proc Natl Acad Sci U S A. 1990;87:1620-1624.
Liu S, Beckman JS, Ku DD. Peroxynitrite, a product of superoxide and nitric oxide, produces coronary vasorelaxation in dogs. J Pharmacol Exp Ther. 1994;268:1114-1121.
Steinbrecher UP. Role of superoxide anion in endothelial-cell modification of low-density lipoprotein. Biochim Biophys Acta. 1988;929:20-30.
Fong KL, McCay PB, Poyer JL. Evidence that peroxidation of lysosomal membranes is initiated by hydroxyl free radicals produced during flavin enzyme activity. J Biol Chem. 1973;248:7792-7797.
Zimmerman BJ, Grisham MB, Granger DN. Role of oxidants in ischemia/reperfusion induced granulocyte infiltration. Am J Physiol (Gastrointestinal). 1990;258:G185-G190.
Abate C, Patel L, Rauscher FI, Curran T. Redox regulation of fos and jun DNA-binding activity in vitro. Science. 1990;249:1157-1161.
Crawford D, Zbinden I, Amsted P, Cerutti P. Oxidant stress in duces the proto-oncogenes c-fos and c-myc in mouse epidermal cells. Oncogene. 1988;3:27-32.
Storz G, Tartaglia LA, Ames BN. Transcriptional regulator of oxidative stress-inducible genes: direct activation by oxidation. Science. 1990;248:189-194.
Marui N, Offermann MK, Swerlick R, Kunsch C, Rosen CA, Ahmad M, Alexander RW, Medford RM. Vascular cell adhesion molecule-1 (VCAM-1) gene transcription and expression are regulated through an antioxidant-sensitive mechanism in human vascular endothelial cells. J Clin Invest. 1993;92:1866-1874.
Rao GN, Berk BC. Active oxygen species stimulate vascular smooth muscle cell growth and proto-oncogene expression. Circ Res. 1992;70:593-599.