Hypercholesterolemia Impairs Endothelium-Dependent Relaxations in Common Carotid Arteries of Apolipoprotein E-Deficient Mice
Background and Purpose— The effects of Western-type fat diet on endothelium-dependent relaxations and vascular structure in carotid arteries from a mouse model of human atherosclerosis are not known. Our objective was to characterize the mechanisms underlying endothelial dysfunction in apoE-deficient mice.
Methods— C57BL/6J and apoE-deficient mice were fed for 26 weeks with a lipid-rich Western-type diet. Changes in the intraluminal diameter of pressurized common carotid arteries (ID 450 μm) were measured in vitro with a video dimension analyzer. Endothelial NO synthase protein content was evaluated by Western blotting. Intracellular cGMP and cAMP levels were determined by radioimmunoassay.
Results— No morphological changes were observed in the carotid arteries of apoE-deficient mice. However, endothelium-dependent relaxations to acetylcholine (10−9 to 10−5 mol/L) were impaired (maximal relaxation 52±7% versus 83±5% for control mice, P<0.05). Treatment of arteries with NO synthase inhibitor Nω-nitro-l-arginine methyl ester inhibited relaxations to acetylcholine to the same extent in apoE-deficient mice as in control mice. Preincubation of carotid arteries with cell-permeable superoxide dismutase mimetic Mn(III) tetra(4-benzoic acid)porphyrin chloride almost normalized NO-mediated relaxations to acetylcholine (75±5%, P<0.05). Endothelium-dependent relaxations to calcium ionophore and endothelium-independent relaxations to NO donor diethylammonium(Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate were unchanged in apoE-deficient mice. In addition, no changes in endothelial NO synthase protein expression and cGMP/cAMP levels were found in carotid arteries of apoE-deficient mice.
Conclusions— In carotid arteries of apoE-deficient mice, hypercholesterolemia causes impairment of receptor-mediated activation of eNOS. Increased superoxide anion production in endothelial cells appears to be coupled to activation of cholinergic receptors and is responsible for hypercholesterolemia-induced endothelial dysfunction. The apoE-deficient mouse carotid artery is a valuable new experimental model of endothelial dysfunction.
NO is a potent vasodilator synthesized by enzymatic activity of endothelial NO synthase (eNOS).1–3 NO production is activated by the stimulation of cell membrane receptors or by mechanical forces such as shear stress.1,4 Accumulating evidence suggests that alteration in the NO pathway and increased production of superoxide anion (O2−) play a central role in endothelial dysfunction induced by hypercholesterolemia.5,6
Atherosclerosis is a multifactorial disease that is triggered by numerous cardiovascular risk factors and is associated with impaired endothelial function that precedes structural vascular change.7 Mice homozygous for inactivated apoE gene provide a new model of human atherosclerosis. These mice develop spontaneous hypercholesterolemia and aortic atherosclerosis, which can be accelerated by a lipid-rich, Western-type diet.8,9
An impaired endothelium-dependent relaxation in response to acetylcholine has been observed in the aorta6,10–13 and in coronary artery14 of genetically altered hyperlipidemic mice. To our knowledge, there is no study concerning the mechanisms of endothelial dysfunction in common carotid artery of hypercholesterolemic apoE-deficient mice. We hypothesized that hypercholesterolemia impairs endothelium-dependent relaxation and that this effect is due to increased formation of superoxide anions.
Materials and Methods
Male C57BL/6J (control) mice and homozygous apoE-deficient mice (C57BL/6J-ApoETm1Unc) were obtained at the age of 4 to 5 weeks from Jackson Laboratory. Housing facilities and all experimental protocols were approved by the Institutional Animal Care and Use Committee of Mayo Clinic. To accelerate the development of spontaneous atherosclerotic lesions in apoE-deficient mice, both C57BL/6J and apoE-deficient mice were fed a lipid-rich Western-type diet for 26 weeks (0.15% cholesterol and 42% milk fat by weight, TD88137; Harlan Teklad).8,9 The mice were anesthetized (pentobarbital 60 mg/kg body wt IP) and killed. The body weights were 31±2 and 42±1 g for age-matched apoE-deficient and control mice, respectively. Blood samples were obtained through puncture of the right ventricle. The blood was immediately transferred to a tube containing heparin and centrifuged at 4°C for 10 minutes. Plasma was separated immediately at 4°C and kept at −80°C until assayed. Cholesterol was determined using a colorimetric-based assay on a Cobas Mira system. Total plasma cholesterol levels were 22.2±1.6 mmol/L in apoE-deficient mice and 6.4±0.7 mmol/L in C57BL/6J mice (P<0.05, n=10).
Carotid arteries were carefully removed and placed immediately into cold (4°C) modified Krebs-Ringer bicarbonate solution containing (in mmol/L) NaCl 118.6, KCl 4.7, CaCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 25.1, EDTA 0.026, and glucose 10.1. Carotid arteries were carefully dissected free from connective tissue in cold Krebs’ solution under a microscope (Carl Zeiss). For cGMP and cAMP analysis (see later), the common carotid artery was immediately frozen in liquid nitrogen and stored at −80°C until assayed.
Segments (4 mm long) of carotid arteries were transferred to small vessel chambers (Living Systems Instrumentation) filled with Krebs-Ringer bicarbonate solution. The solutions circulating from a 250-mL reservoir at a flow rate of 50 mL/min were aerated continuously with 94% O2 and 6% CO2 gas and kept at 37°C. Proximal and distal ends of the small vessels were mounted and sutured onto 2 small glass microcannulas (inflow and outflow cannula, respectively) positioned in the vessel chamber. The axial length of the vessel was carefully adjusted longitudinally under a microscope by positioning the afferent cannula. The transmural pressure of 50 mm Hg was set at a level, which was found to be optimal for contractions to U46619 (3×10−8 mol/L).15 The chamber was positioned on the stage of an inverted microscope (Nikon Diaphot-TMD) with a video camera. The amplified image was transmitted to a monitor and a video dimension analyzer (Living Systems Instrumentation), allowing for measurements and recording of lumen diameter. Carotid arteries were equilibrated for 60 minutes.
Common carotid rings from C57BL/6J and apoE-deficient mice were studied in parallel. Between each protocol, the system was washed out with Krebs’ solution and then equilibrated for 30 minutes. Endothelium-dependent relaxations to acetylcholine (10−9 to 10−5 mol/L) were first obtained after stabilization of submaximal contraction to thromboxane analog 9,11-dideoxy-11α,9α-epoxymethano-prostaglandin F2α (U46619; 3×10−8 to 10−7 mol/L). After washout, endothelium-dependent relaxations to acetylcholine were assessed in the presence of Nω-nitro-l-arginine methyl ester (L-NAME; 3×10−4 mol/L; 20 minutes before submaximal contraction with 10−8 mol/L U46619).
Concentration-dependent responses to diethylammonium(Z)-1-(N,N-diethylamino)diazen-1-ium-1,2-diolate (DEA-NONOate; 10−9 to 10−5 mol/L) or Ca2+ ionophore (A23187; 10−9 to 10−6 mol/L) were obtained in different carotid arteries after submaximal contraction to U46619 (3×10−8 to 10−7 mol/L).
In a separate protocol, a novel cell-permeable superoxide dismutase (SOD) mimetic, Mn(III) tetra(4-benzoic acid)porphyrin chloride (MnTBAP; 10−5 mol/L for 15 minutes), was used, and endothelium-dependent relaxation was tested using acetylcholine (10−9 to 10−5 mol/L).
Western Blot Analysis
Briefly, after collection and removal of connective tissue, 4 pairs of carotid arteries (n=1 experiment) were homogenized on ice in lysis buffer (pH 7.5) containing 50 mmol/L Tris-HCl, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 0.1% SDS, 0.1% deoxycholate, 1% IGEPAL, and a 1000-fold dilution of a mammalian protease inhibitor cocktail (all from Sigma Chemical Co). Equal amounts of protein (50 μg/lane) from the C57BL/6J and apoE groups were separated by SDS-PAGE and transferred to nitrocellulose membrane (Amersham) using a semidry electrophoretic transfer cell for Western analysis. For eNOS protein analysis, monoclonal anti-eNOS (1:100; Transduction Laboratories) was used. Bands were visualized by enhanced chemiluminescence using a commercially available kit (Amersham). Densitometry was carried out using NIH Image (Scion-Image; Scion-Corp), and the results were expressed in optical intensity per microgram of protein.
Measurements of cGMP and cAMP
Frozen carotid arteries were homogenized in a cooled solution of 100% ethanol, and 1 N HCl. cGMP and cAMP radioimmunoassay kits (Amersham) were used to perform the measurements.16 Aortic rings from C57BL/6J and apoE-deficient mice were analyzed in parallel, and the results were expressed as pmol/mg protein.
Right common carotid arteries were fixed in buffered formalin (4%) and kept for 24 hours at room temperature. Tissues were embedded in paraffin and processed for light microscopy using standard H&E staining.
Acetylcholine hydrochloride, A23187, and L-NAME were from Sigma Chemical Co. DEA-NONOate and U46619 were from Cayman Chemical Co, and MnTBAP was from BIOMOL Laboratories. DEA-NONOate and A23187 were prepared as stock solutions in 1.5 mol/L Tris buffer, pH 8.8, and DMSO, respectively. U46619 was dissolved in 1 part of 100% ethanol and then diluted with 9 parts of water. The remaining drugs were dissolved in distilled water. All drugs were then diluted in Krebs’ solution, and concentrations are expressed as final molar concentration (mol/L) in the organ bath.
Calculations and Statistical Analysis
For statistical analysis, sensitivity of the vessels to different drugs was expressed as negative logarithm of the concentration causing half-maximal relaxation or contraction (pD2 value). In addition, maximal contraction or relaxation (expressed as percentage of the decrease in the basal intraluminal diameter or of the increase in intraluminal diameter from the diameter obtained after precontraction, respectively) were determined for each individual concentration-response curve by nonlinear regression analysis. The concentration-response curves of the different groups were compared by ANOVA for repeated measurements followed by Bonferroni’s correction. For simple comparison between 2 values, a paired or unpaired Student’s t test was used, where appropriate. A value of P<0.05 was considered significant. In all experiments, n equals the number of mice from which blood vessels were obtained.
Histology and Vascular Structure
No atherosclerotic changes were observed in the carotid arteries of C57BL/6J mice (Figure 1A). Histological examination revealed neither atherosclerotic lesions nor intimal thickness in the common carotid arteries of apoE-deficient mice (Figure 1B). In addition, lumen diameter did not differ between the groups of mice (447±7 and 454±13 μm for C57BL/6J and apoE-deficient mice, respectively; n=7 and 9).
During contractions to U46619, endothelium-dependent relaxations to acetylcholine were significantly reduced in common carotid artery of apoE-deficient mice (P<0.01 versus C57BL/6J; Figures 2 and 3⇓A). The sensitivity to acetylcholine was slightly shifted to the right in apoE mice compared with C57BL/6J mice (pD2 7.0±0.2 and 7.3±0.1, respectively; P=0.11). In the presence of L-NAME (3×10−4 mol/L), relaxations to acetylcholine were reduced to the same extent in apoE-deficient mice as in control mice (P<0.01; Figure 3B).
A23187 also caused endothelium-dependent relaxations in common carotid arteries; however, no differences were found between apoE-deficient and C57BL/6J mice (P=0.46; Figure 4A). Endothelium-independent relaxations to the NO donor DEA-NONOate were also unaltered (P=0.74; Figure 4B).
Effect of SOD Mimetic
The cell-permeable SOD mimetic MnTBAP (10−5 mol/L) significantly improved endothelium-dependent relaxation to acetylcholine in carotid arteries from apoE-deficient mice (P<0.05; Figure 5A). MnTBAP had no effect on sensitivity or maximal relaxations to acetylcholine in C57BL/6J mice (Figure 5B). Contraction to U46619 (3×10−8 mol/L) was not affected by MnTBAP in control and apoE-deficient mice (52±2% versus 51±5% and 57±3% versus 51±7% with and without MnTBAP, respectively).
eNOS Protein Expressions
Western blot analysis showed that eNOS protein was expressed in common carotid arteries of both C57BL/6J and apoE-deficient mice. However, expression was similar between groups (P=0.54; Figure 6).
cGMP and cAMP Levels
Basal cGMP and cAMP levels were not different in carotid arteries from apoE-deficient and C57BL/6J mice (P=0.57 and 0.89, respectively; Figure 7).
This is the first study to examine endothelial function in the common carotid artery of apoE-deficient mice. Our results demonstrate that despite severe hypercholesterolemia, we did not detect any morphological change in the common carotid arteries of apoE-deficient mice. In contrast, hypercholesterolemia selectively impaired NO-mediated endothelium-dependent relaxations to acetylcholine, whereas eNOS protein expression and cyclic nucleotide levels were unchanged. SOD-mimetic normalized endothelial function in apoE-deficient mice, suggesting that increased formation of superoxide anion is an important mechanism responsible for endothelial dysfunction.
A previous study reported that the relaxation of the normal mouse carotid artery to acetylcholine is mediated by NO and activation of soluble guanylate cyclase.17 In the present study, apoE-deficient mice showed severe hyperlipidemia associated with impaired endothelium-dependent relaxations to acetylcholine. These observations are consistent with the previous study in the aorta and coronary arteries of genetically altered mice.10–12,14 It is important to note that relaxation to acetylcholine was substantially attenuated by an inhibitor of NOS, suggesting that the response is mediated by endothelium-derived NO. In addition, the L-NAME-resistant portion of relaxation to acetylcholine was unaffected by hypercholesterolemia in apoE-deficient mice.
Interestingly, in common carotid arteries of apoE-deficient mice, neither atherosclerotic lesions nor intimal thickening was found despite pronounced atherosclerotic lesions in the aortic arch, innominate artery,18 and external carotid arteries.8,19 The reason for this discrepancy is not known. Certain parts of arteries may be more prone to structural changes due to characteristics of wall hemodynamic and shear stress forces. Indeed, turbulent and complex flow patterns that occur at branching points favor the development of atherosclerosis. Although both atherosclerotic lesions and endothelial dysfunction are prevalent in large conduit arteries of genetically altered hyperlipidemic mice,10,11 in smaller arteries, only impairment of endothelium-dependent relaxations is present. Indeed, previous studies showed that epicardial atherosclerosis in humans is associated with impaired endothelium-dependent relaxations in coronary arteries, indicating that the pathophysiological consequences of atherosclerosis may extend into small arteries.20–23 In addition, there is a selective endothelial dysfunction in angiographically defined normal coronary arteries in patients with hypercholesterolemia followed by progressive worsening to a complete loss of endothelium-mediated vasodilation in angiographically defined atherosclerotic coronary arteries.20
Previous studies identified several mechanisms responsible for the reduced bioavailability of NO in arteries exposed to hypercholesterolemia: (1) enhanced degradation of NO by O2−,24 (2) functional abnormalities of NOS due to deficiency of substrate or cofactor,25,26 (3) a defect in the signal transduction pathways,27 and/or (4) alteration in eNOS protein expression.28 In the present study, treatment with the SOD mimetic MnTBAP normalized endothelium-dependent relaxations in the carotid artery of apoE-deficient mice. The selectivity of MnTBAP was demonstrated by the fact that it did not affect endothelium-dependent relaxations of acetylcholine in control C57BL/6J mice. It is important to note that the SOD mimetic used in this study scavenges not only O2− but also hydrogen peroxide and peroxynitrite,29,30 suggesting that excess production of these oxidants may also contribute to endothelial dysfunction in hypercholesterolemia. Because of the small size of the carotid arteries, we are unable to quantify O2− levels directly using lucigenin-enhanced chemiluminescence.
Oxygen free radicals such as O2− can cause a wide spectrum of cell damage, including lipid peroxidation, inactivation of enzymes, alteration of intracellular redox state, and damage to DNA, and are thought to participate in the pathogenesis of atherosclerosis.5 O2− reacts with NO to form peroxynitrite in a diffusion-limited process, and because peroxynitrite is much less effective as an activator of guanylyl cyclase, this reaction results in a marked reduction in the bioactivity of NO.28,31 Several sources of O2− have been identified in the vascular wall. Cyclooxygenase, xanthine oxidase, NAD(P)H oxidase, and eNOS enzyme activity can generate O2−.32–34 The exact source of O2− in carotid artery of apoE-deficient mice is unknown and remains to be determined.
Unlike endothelium-dependent relaxations to acetylcholine, activation of endothelial cells with calcium ionophore A23187 caused relaxation that was not affected by hypercholesterolemia. This finding is consistent with results obtained in isolated arteries of hypercholesterolemic rabbits, pigs, and humans.27 It supports the idea that in the carotid artery, endothelial dysfunction is not caused by a nonspecific loss of endothelial cell capacity to synthesize and release NO. This interpretation is further supported by the fact that basal production of the NO second messenger cGMP was not affected by hypercholesterolemia. However, because the mechanism of relaxation of mouse common carotid artery to A23187 has not been characterized, we cannot rule out the possibility that the response to A23187 is in part mediated by the release of an endothelium-derived relaxing factor other than NO.
There is evidence that in control carotid arteries, endothelium-dependent relaxations to acetylcholine are mediated by eNOS because inactivation of the eNOS gene in mice abolished the relaxations.17 In our study, Western blot analysis showed no change in eNOS protein expression in common carotid arteries of apoE-deficient mice, suggesting that impaired endothelial function is not due to the altered eNOS expression. This finding was in contrast to results obtained in human carotid arteries, in which a reduction of immunoreactive eNOS was found.28 The discrepancy may be related to the differential duration of the high fat treatment or duration of the atherosclerotic process (ie, months in experimental animals versus decades in patients).
Endothelium-independent relaxations were similar in apoE-deficient and control mice, further supporting our conclusion that the observed blunted responses to acetylcholine are not a consequence of reduced responsiveness of vascular smooth muscle cells to NO and/or changes in activation of soluble guanylyl cyclase and the subsequent formation of cGMP. In addition, basal production of cAMP is not impaired in apoE-deficient carotid arteries, suggesting that mechanisms responsible for control of cyclic nucleotide metabolism are not affected by hypercholesterolemia. These findings are in agreement with the results of previous studies demonstrating impaired endothelium-dependent relaxations to acetylcholine and normal vasodilation of vascular smooth muscle in animals or patients with hypercholesterolemia.14,22,35,36 They also suggest that increased production of O2− in endothelial cells is the most likely explanation for impaired cholinergic vasodilatation.
The results of the present study demonstrate that in mouse common carotid artery, hypercholesterolemia causes selective loss of endothelium-dependent relaxation to acetylcholine. Increased formation of superoxide anion appears to be an important mechanism underlying impairment of endothelial function. These alterations were observed in morphologically intact arteries, suggesting that impaired NO-mediated endothelium-dependent relaxation is a precursor of intimal thickening and development of atherosclerotic lesions. Our study provides new model of endothelial dysfunction in mouse carotid artery. This model will be very useful in future studies designed to characterize genetic and molecular mechanisms underlying early endothelial dysfunction in morphologically normal arteries.
This work was supported by National Heart, Lung, and Blood Institute grant HL-53524, National Institute of Neurological Disorders and Stroke grant NS-37491, and the Mayo Foundation. Dr d’Uscio is the recipient of a Swiss National Sciences Foundation stipend and an American Heart Association Northland Affiliate Fellowship.
- Received March 29, 2001.
- Revision received June 1, 2001.
- Accepted July 20, 2001.
Ignarro LJ, Byrns RE, Buga GM, Wood KS. Endothelium-derived relaxing factor from pulmonary artery and vein possesses pharmacologic and chemical properties identical to those of nitric oxide radical. Circ Res. 1987; 61: 866–879.
Lüscher TF, Noll G. The endothelium in coronary vascular control.In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine, 3rd ed. Philadelphia, Pa: WB Saunders; 1995: 1–10.
Harrison DG, Freiman PC, Armstrong ML, Marcus ML, Heistad DD. Alterations of vascular reactivity in atherosclerosis. Circ Res. 1987; 61 (suppl II): II-74–II-80.
Nakashima Y, Plump AS, Raines EW, Breslow JL, Ross R. ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994; 14: 133–140.
Barton M, Haudenschild CC, d’Uscio LV, Shaw S, Münter K, Lüscher TF. Endothelin ETA receptor blockade restores NO-mediated endothelial dysfunction and inhibits atherosclerosis in apoE-deficient mice. Proc Natl Acad Sci U S A. 1998; 95: 14367–14372.
Bonthu S, Heistad DD, Chappell DA, Lamping KG, Faraci FM. Atherosclerosis, vascular remodeling, and impairment of endothelium-dependent relaxation in genetically altered hyperlipidemic mice. Arterioscler Thromb Vasc Biol. 1997; 17: 2333–2340.
Kauser K, da Cunha V, Fitch R, Mallari C, Rubanyi GM. Role of endogenous nitric oxide in progression of atherosclerosis in apolipoprotein E-deficient mice. Am J Physiol. 2000; 278: H1679–H1685.
Lamping KG, Nuno DW, Chappell DA, Faraci FM. Agonist-specific impairment of coronary vascular function in genetically altered, hyperlipidemic mice. Am J Physiol. 1999; 276: R1023–R1029.
Katusic ZS. Endothelial L-arginine pathway and regional cerebral arterial reactivity to vasopressin. Am J Physiol. 1992; 262: H1557–H1562.
Tsutsui M, Milstien S, Katusic ZS. Effect of tetrahydrobiopterin on endothelial function in canine middle cerebral arteries. Circ Res. 1996; 79: 336–342.
Faraci FM, Sigmund CD, Shesely EG, Maeda N, Heistad DD. Responses of carotid artery in mice deficient in expression of the gene for endothelial NO synthase. Am J Physiol. 1998; 274: H564–H570.
Rosenfeld ME, Polinsky P, Virmani R, Kauser K, Rubanyi G, Schwartz SM. Advanced atherosclerotic lesions in the innominate artery of the apoE knockout mouse. Arterioscler Thromb Vasc Biol. 2000; 20: 2587–2592.
Seo HS, Lombardi DM, Polinsky P, Powell-Braxton L, Bunting S, Schwartz SM, Rosenfeld ME. Peripheral vascular stenosis in apolipoprotein E-deficient mice: potential roles of lipid deposition, medial atrophy, and adventitial inflammation. Arterioscler Thromb Vasc Biol. 1997; 17: 3593–3601.
Zeiher AM, Drexler H, Wollschlager H, Just H. Modulation of coronary vasomotor tone in humans: progressive endothelial dysfunction with different early stages of coronary atherosclerosis. Circulation. 1991; 83: 391–401.
Kuo L, Davis MJ, Cannon MS, Chilian WM. Pathophysiological consequences of atherosclerosis extend into the coronary microcirculation: restoration of endothelium-dependent responses by l-arginine. Circ Res. 1992; 70: 465–476.
Sellke FW, Armstrong ML, Harrison DG. Endothelium-dependent vascular relaxation is abnormal in the coronary microcirculation of atherosclerotic primates. Circulation. 1990; 81: 1586–1593.
Quyyumi AA, Dakak N, Andrews NP, Husain S, Arora S, Gilligan DM, Panza JA, Cannon RO3rd. Nitric oxide activity in the human coronary circulation: impact of risk factors for coronary atherosclerosis. J Clin Invest. 1995; 95: 1747–1755.
Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest. 1993; 91: 2546–2551.
Creager MA, Gallagher SJ, Girerd XJ, Coleman SM, Dzau VJ, Cooke JP. L-Arginine improves endothelium-dependent vasodilatation in hypercholesterolemic humans. J Clin Invest. 1992; 90: 1248–1253.
Flavahan NA. Atherosclerosis or lipoprotein-induced endothelial dysfunction: potential mechanisms underlying reduction in EDRF/nitric oxide activity. Circulation. 1992; 85: 1927–1938.
Oemar BS, Tschudi MR, Godoy N, Brovkovich V, Malinski T, Lüscher TF. Reduced endothelial nitric oxide synthase expression and production in human atherosclerosis. Circulation. 1998; 97: 2494–2498.
Szabo C, Day BJ, Salzman AL. Evaluation of the relative contribution of nitric oxide and peroxynitrite to the suppression of mitochondrial respiration in immunostimulated macrophages using a manganese mesoporphyrin superoxide dismutase mimetic and peroxynitrite scavenger. FEBS Lett. 1996; 381: 82–86.
Kukreja RC, Kontos HA, Hess ML, Ellis EF. PGH synthase and lipoxygenase generate superoxide in the presence of NADH or NADPH. Circ Res. 1986; 59: 612–619.
Schultz D, Harrison DG. Quest for fire: seeking the source of pathogenic oxygen radicals in atherosclerosis. Arterioscler Thromb Vasc Biol. 2000; 20: 1412–1413.
Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res. 2000; 86: 494–501.
Yang R, Powell-Braxton L, Ogaoawara AK, Dybdal N, Bunting S, Ohneda O, Jin H. Hypertension and endothelial dysfunction in apolipoprotein E knockout mice. Arterioscler Thromb Vasc Biol. 1999; 19: 2762–2768.
Vita JA, Treasure CB, Nabel EG, McLenachan JM, Fish RD, Yeung AC, Vekshtein VI, Selwyn AP, Ganz P. Coronary vasomotor response to acetylcholine relates to risk factors for coronary artery disease. Circulation. 1990; 81: 491–497.