Long-Term Treatment With Nω-Nitro-l-Arginine Methyl Ester Causes Arteriosclerotic Coronary Lesions in Endothelial Nitric Oxide Synthase-Deficient Mice
Background— Nω-nitro-l-arginine methyl ester (l-NAME) is widely used to inhibit endothelial synthesis of NO in vivo. However, it is controversial whether the long-term vascular effects of l-NAME are mediated primarily by inhibition of endothelial NO synthesis. We addressed this point in mice that are deficient in the endothelial NO synthase gene (eNOS-KO mice).
Methods and Results— Wild-type and eNOS-KO mice received l-NAME in drinking water for 8 weeks. In wild-type mice, long-term treatment with l-NAME caused significant medial thickening and perivascular fibrosis in coronary microvessels but not in large coronary arteries. Importantly, in eNOS-KO mice, treatment with l-NAME also caused an extent of medial thickening and perivascular fibrosis in coronary microvessels that was comparable to that in wild-type mice and that was not prevented by supplementation of L-arginine. Vascular NO and cGMP levels were not significantly reduced by l-NAME treatment, and no expression of inducible or neuronal NO synthase was noted in microvessels of eNOS-KO mice, suggesting an involvement of NO-independent mechanisms. Treatment with l-NAME caused an upregulation of vascular ACE and an increase in cardiac lucigenin chemiluminescence that were comparable in both strains and that were abolished by simultaneous treatment with temocapril (ACE inhibitor) or CS866 (angiotensin II type 1 receptor antagonist) along with the suppression of vascular lesion formation.
Conclusions— These results provide the first direct evidence that the long-term vascular effects of l-NAME are not mediated by simple inhibition of endothelial NO synthesis. Direct upregulation of local ACE and increased oxidative stress appear to be involved in the long-term vascular effects of l-NAME in vivo.
Received April 15, 2002; revision received June 26, 2002; accepted June 26, 2002.
Endothelium-derived NO, synthesized by endothelial NO synthase (eNOS), has several important antiatherogenic actions.1–3⇓⇓ As pharmacological tools used to inhibit endothelial NO synthesis, l-arginine analogues have been widely used in vitro and in vivo. Among them, Nω-nitro-l-arginine methyl ester (l-NAME) is the most frequently used agent.1–3⇓⇓ Long-term oral treatment with l-NAME is known to cause arteriosclerotic coronary lesions, especially at microvascular levels in experimental animals.4,5⇓ This model with l-NAME is regarded as a useful animal model for examining the protective roles of endothelium-derived NO in the pathogenesis of arteriosclerosis.4,5⇓
However, it is controversial whether these vascular effects of l-NAME are caused primarily by the inhibition of endothelial NO synthesis for the following reasons: First, the importance of endothelium-derived NO decreases as the vessel size becomes smaller,6 whereas l-NAME-induced vascular lesions are prominent at microvascular levels.4 Second, long-term treatment with l-NAME does not reduce eNOS activity.7 Third, multiple actions of l-NAME other than simple inhibition of NO synthesis have been reported.8,9⇓ The most appropriate way to address this important issue is to use mice that are deficient in the eNOS gene and to examine whether long-term treatment with l-NAME causes coronary vascular lesions in those mice. In the present study, we were able to demonstrate that treatment with l-NAME causes a comparable extent of coronary arteriosclerotic lesions in wild-type and eNOS-deficient (eNOS-KO) mice.
The present study was reviewed and approved by the Ethics Committee on Animal Experiments and Care of the University of Occupational and Environmental Health.
Disruption of the eNOS gene was confirmed by polymerase chain reaction (PCR) of genomic DNA.10 C57BL/6 mice were used as a wild genotype control.
Experiments were performed in 8- to 10-week-old male mice (Charles River Japan, Inc, Yokoha, Japan) weighing 20 to 25 g. The following 13 groups were studied: wild-type and eNOS-KO mice that received untreated drinking water, l-NAME (1 mg/mL, Sigma Chemical Co) in drinking water, l-NAME plus l-arginine (70 mg/mL, Sigma) in drinking water, Nω-nitro-l-arginine (l-NNA, 1 mg/mL, Sigma) in drinking water, l-NAME plus temocapril (0.1 mg/mL, Sankyo Pharmaceutical Co) in drinking water, and l-NAME plus CS866 (5 mg/kg, Sankyo Pharmaceutical Co) in chow and wild-type mice that received l-NAME plus hydralazine (0.05 mg/mL) in drinking water for 8 weeks. The actual daily intake of water containing drugs was 4 to 6 mL. Age- and sex-matched control mice were used in all experiments.
Histological and Immunohistochemical Analyses
Paraffin slices were stained with Masson’s trichrome solutions or with mouse ACE antibody (Chemicon). The sections were scanned by using a light microscope equipped with a 2D analysis system (IBAS, Carl Zeiss). The extent of medial thickening and the extent of perivascular fibrosis of the coronary arteries were evaluated by the ratio of medial thickness to internal diameter and the ratio of perivascular fibrosis area to total vascular area, respectively. In each heart, >10 large epicardial coronary arteries (internal diameter 212±4 μm) and coronary microvessels (internal diameter 28±2 μm) were examined, and average values were used.
Measurement of NOx, NO, and cGMP
Nitrite plus nitrate (NOx) concentrations in plasma and urine were analyzed by the Griess method after 12 hours of fasting.11 NO release in the coronary arteries was directly measured with an NO-sensitive electrode (model No. 501, Inter Medical Co).12 Mouse hearts were isolated, and an NO-sensitive electrode was stuck and placed perpendicularly into the left anterior descending coronary artery in oxygenated Krebs-Ringer bicarbonate solution at 37°C. Tissue cGMP concentrations in the aorta were assessed by radioimmunoassay.13
Expression of iNOS and nNOS
The expression of inducible NO synthase (iNOS) was evaluated by immunostaining,14 Western blotting,15 and reverse transcriptase (RT)-PCR. Expression of neuronal NO synthase (nNOS) was examined by immunostaining,15 Western blotting,15 and in situ hybridization.
Change in redox state consistent with an oxidative stress was assessed by the lucigenin-enhanced chemiluminescence method.16
Results are presented as mean±SEM. Statistical analysis was performed by ANOVA followed by the Scheffé post hoc test. A value of P<0.05 was considered to be statistically significant.
Arterial systolic blood pressure (BP) was slightly but significantly higher in eNOS-KO mice (127±2 mm Hg) than in wild-type mice (107±1 mm Hg; P<0.05, n=10 each).
Long-term treatment with l-NAME (1 mg/mL) significantly increased BP in wild-type mice (129±3 mm Hg, P<0.05) but significantly decreased it in eNOS-KO mice (91±1 mm Hg; P<0.05, n=10 each).
Formation of Coronary Vascular Lesions
In wild-type mice, long-term treatment with l-NAME caused significant medial thickening and perivascular fibrosis in coronary microvessels but not in large coronary arteries (Figure 1A and 1C). Importantly, treatment with l-NAME also caused a comparable extent of medial thickening and perivascular fibrosis in microvessels (but not in large coronary arteries) in eNOS-KO mice and wild-type mice (Figure 1B and 1D). These vascular effects of l-NAME were not reversed by simultaneous treatment with l-arginine (70 mg/mL) in both strains (Figure 2A through 2D). Another l-arginine analogue, l-NNA (1 mg/mL), similarly induced a comparable extent of medial thickening and perivascular fibrosis in coronary microvessels in wild-type and eNOS-KO mice (Figure 3A through 3D).
In wild-type mice, antihypertensive treatment with hydralazine (0.05 mg/mL) normalized systolic BP (129±3 to 103±3 mm Hg) but failed to inhibit the formation of medial thickening (0.3±0.1 to 0.8±0.1) or perivascular fibrosis (0.2±0.1 to 0.6±0.1) in microvessels caused by the l-NAME treatment (n=5).
Long-term treatment with l-NAME significantly decreased plasma NOx concentrations (37±7 to 22±5 μmol/L) and urine NOx excretion (3.0±0.5 to 1.9±0.3 μmol/d), markers of systemic NO production, in wild-type mice (n=10 each). However, long-term treatment with l-NAME did not significantly affect local NO production, including basal NO release (assessed by the reduction in NO levels in coronary arteries of isolated hearts induced by acute administration of 10−3 mol/L NG-monomethyl-l-arginine [0.15±0.07 μmol/L without and 0.17±0.06 μmol/L with long-term l-NAME treatment, respectively]) or acetylcholine (10−5 mol/L)-stimulated NO release (2.7±1.2 μmol/L without and 3.2±1.1 μmol/L with l-NAME treatment, respectively) in coronary arteries of wild-type mice (n=7 each). Furthermore, treatment with l-NAME did not affect tissue cGMP concentrations in the aortas of wild-type mice (3.2±0.5 without and 5.3±1.0 pmol/mg wet wt with l-NAME treatment, n=5 each), another marker of local NO production.
Expression of iNOS and nNOS
Before treatment with l-NAME, no compensatory expression of iNOS or nNOS was noted in coronary microvessels by immunostaining (Figures 4A and 5⇓A), RT-PCR (data not shown), in situ hybridization (data not shown), or Western blotting (Figures 4B and 5⇓B). With our methods, we were able to detect the expression of iNOS protein (Figure 4B) and mRNA (data not shown) in the hearts and aortas taken from eNOS-KO mice 12 hours after intraperitoneal injection of lipopolysaccharide (10 mg/kg) and the expression of nNOS protein (Figure 5B) and mRNA (data not shown) in the brains of eNOS-KO mice, confirming the accuracy of our methods. By contrast, 4 and 8 weeks after treatment with l-NAME, expression of iNOS (Figure 4A and 4B) and nNOS (Figure 5A and 5B) was noted in the medial smooth muscle cells of large coronary arteries and coronary microvessels and the myocardium (but to a similar extent in both strains) by immunostaining, RT-PCR, in situ hybridization, and Western blotting.
To clarify whether the expression of iNOS may reflect vascular inflammation, the number of vascular inflammatory cells was examined. Treatment with l-NAME for 3 days induced a comparable extent of accumulation of inflammatory mononuclear leukocytes (MOMA-2-positive macrophages) into coronary microvessels of wild-type (5.2±0.6 without and 16.6±1.1/mm2 with l-NAME treatment, respectively) and eNOS-KO mice (7.4±0.9 without and 20.1±1.4/mm2 with l-NAME treatment, respectively).
Immunostaining revealed no nNOS protein expression in large coronary arteries, coronary microvessels, or myocardium of age-matched 18-week-old wild-type and eNOS-KO mice untreated with l-NAME (data not shown). Western blot analysis also showed no nNOS protein expression in the hearts of 18-week-old control wild-type or eNOS-KO mice (data not shown).
To elucidate whether nNOS-containing nerves might be induced in coronary microvessels after l-NAME treatment, the immunoreactivity of nNOS was examined with the 2 nerve-specific antibodies, protein gene product 9.5 (PGP)17 and neurofilament 200 (NF).18 After long-term treatment with l-NAME, the immunoreactivity of nNOS was observed mainly in the media of coronary microvessels in both strains, whereas that of PGP and NF was noted to a comparable extent before and after l-NAME treatment, mainly in the adventitia (data not shown).
Tissue ACE Expression
ACE immunoreactivity was slightly noted at the adventitia of coronary microvessels in both strains (Figure 6). By contrast, after the treatment with l-NAME, ACE immunoreactivity increased in both strains to a comparable extent, mainly in the adventitia. No such increase in ACE immunoreactivity was noted in mice cotreated with temocapril (0.1 mg/mL), which is an ACE inhibitor, or CS866 (5 mg/kg), which is an angiotensin II type 1 (AT1) receptor antagonist (Figure 6).
Treatment with l-NAME for 3 days significantly increased cardiac lucigenin chemiluminescence in both strains to a comparable extent (Figure 7A and 7B). The l-NAME-induced increase in lucigenin chemiluminescence was again normalized by cotreatment with temocapril or CS866 in both strains (Figure 7A and 7B).
Prevention of Coronary Vascular Lesion Formation
Medial thickening and perivascular fibrosis induced by l-NAME were also abolished by the simultaneous treatment with temocapril or CS866 in both strains (Figure 8A and 8B).
The novel findings of the present study were as follows: (1) In wild-type and eNOS-KO mice, long-term oral treatment with l-NAME induced a comparable extent of coronary microvascular lesions that was not reversed by the addition of l-arginine. (2) The treatment also caused a comparable extent of ACE upregulation and an increase in lucigenin chemiluminescence in both strains. (3) Those changes were not a consequence of the developed hypertension. (4) Treatment with l-NAME did not significantly reduce vascular NO release or cGMP concentrations. (5) No compensatory expression of iNOS or nNOS was noted in coronary microvessels of eNOS-KO mice. These results provide the first direct evidence that mechanisms other than simple inhibition of endothelial NO synthesis are involved in the long-term vascular effects of l-NAME in vivo.
No Effect of BP
Long-term treatment with l-NAME increased arterial BP in wild-type mice but decreased it in eNOS-KO mice. Irrespective of such changes in arterial BP, l-NAME caused a comparable extent of coronary vascular lesions in both strains. Furthermore, antihypertensive treatment with hydralazine failed to inhibit the coronary vascular lesion formation by l-NAME, indicating that long-term vascular effects of l-NAME were not caused by changes in arterial BP.
No Effect of l-NAME on Vascular NO Production
Plasma NOx concentrations and urine NOx excretion, markers of systemic NO production, were significantly decreased by long-term treatment with l-NAME in wild-type mice. However, NO release in coronary arteries and tissue cGMP concentrations in the aorta, markers of local NO production, were unaltered by treatment with l-NAME in wild-type mice. These findings are consistent with a previous study reporting that eNOS activity and local NO production are not reduced in the renal microvessels of rats after long-term oral administration of l-NAME.7 Therefore, it is possible that long-term treatment with l-NAME inhibits systemic NO production but does not affect local vascular NO production.
Involvement of NO-Independent Mechanisms in Long-Term Vascular Effects of l-NAME
The following lines of evidence support the involvement of NO-independent mechanisms in the long-term vascular effects of l-NAME. First, the effects of l-NAME could not be reversed by l-arginine. Second, l-NAME treatment did not decrease vascular NO or cGMP production. Third, l-NAME treatment caused a comparable extent of vascular lesion formation irrespective of the presence or absence of eNOS. Fourth, l-NAME treatment also caused a comparable extent of direct local ACE upregulation and oxidative stress irrespective of the presence or absence of eNOS. Fifth, the vascular lesions were induced in coronary microvessels of eNOS-KO mice, on which expression of other NO synthases was initially absent. However, because iNOS and nNOS were expressed after the initiation of l-NAME treatment, it is possible that the vascular effects of l-NAME might be mediated in part by inhibition of newly expressed iNOS and/or nNOS. It is also conceivable that the vascular effects of l-NAME might be mediated by inhibition of NO delivery from circulating sources, such as S-nitrosohemoglobin or other nitrosothiols, which might be derived from nNOS and/or iNOS. It remains to be fully elucidated whether the vascular effects of l-NAME observed in eNOS-KO mice are due to inhibition of other NO synthases. For this purpose, mice deficient in all 3 NO synthase isoforms would need to be developed.
Expression of iNOS After l-NAME Treatment
Expression of iNOS may reflect a vascular inflammatory response. Indeed, we have confirmed that treatment with l-NAME for 3 days induces an infiltration of inflammatory mononuclear leukocytes (MOMA-2-positive macrophages) into coronary microvessels of both strains. Previous studies have consistently reported that l-NAME administration induces inflammatory changes in the arterial wall in rats.19
Expression of nNOS After l-NAME Treatment
It has been reported that nNOS is a constitutively expressed enzyme and that in the vascular system, nNOS is present in perivascular nitridergic nerves. However, in the present study, nNOS was induced in coronary microvessels after l-NAME treatment. To determine whether nNOS-containing nerves might be induced in coronary microvessels after l-NAME treatment, we compared the immunoreactivity of nNOS with that of 2 nerve-specific antibodies, PGP17 and NF.18 After long-term treatment with l-NAME, the immunoreactivity of nNOS was observed mainly in the media of coronary microvessels in both strains, whereas that of PGP and NF was seen in the adventitia. In addition, after l-NAME treatment, no increase in immunoreactivity of PGP or NF was noted, in contrast to the increase in the immunoreactivity of nNOS. These results indicate that coronary microvessels were not innervated by nitridergic nerves after the initiation of treatment with l-NAME.
We examined the effect of age on nNOS expression by immunostaining and Western blot analysis. Immunostaining revealed no nNOS protein expression in large coronary arteries, coronary microvessels, or myocardium of age-matched 18-week-old wild-type and eNOS-KO mice not treated with l-NAME. Western blot analysis also showed no nNOS protein expression in the hearts of 18-week-old control wild-type and eNOS-KO mice. Thus, it is evident that nNOS is not induced in the course of aging.
We have demonstrated that nNOS is subject to expression regulation in arteriosclerotic vascular lesions in vivo. In a mouse carotid artery ligation model20 and a rat balloon-injury model,21 nNOS was absent before injury and was upregulated in vascular lesions after the injury, predominantly in the neointima and medial smooth muscle cells. We were unable to detect any immunoreactivity of nerve-specific antibodies in such vascular lesions. Induction of nNOS in the carotid arteries of spontaneously hypertensive rats15 and in human atherosclerotic blood vessels14 has also been reported.
Involvement of Renin-Angiotensin System in Long-Term Vascular Effects of l-NAME
Treatment with l-NAME significantly increased tissue ACE expression in perivascular areas and cardiac lucigenin chemiluminescence, which were normalized by ACE inhibition22 or AT1 receptor blockade22 along with the suppression of vascular lesion formation. Long-term administration of angiotensin II causes perivascular fibrosis in rats,23 which resembles the histopathologic changes seen in the present l-NAME model. Furthermore, AT1 receptor stimulation elicits vascular production of reactive oxygen species, including superoxide, which, in turn, induces arteriosclerotic vascular lesions.2 Dual increases in the production of superoxide anions and NO derived from newly expressed iNOS and/or nNOS after l-NAME treatment might exacerbate the arteriosclerotic process via peroxynitrite formation.2
Thus, enhanced tissue ACE expression and oxidative stress through AT1 receptors may contribute, at least in part, to the long-term vascular effects of l-NAME irrespective of the presence or absence of eNOS.
Multiple Nonspecific Effects of l-Arginine Analogues
Multiple nonspecific effects of l-NAME other than simple inhibition of NO synthesis have been reported. First, l-NAME antagonizes muscarinic acetylcholine receptors.8 Second, l-NAME has inhibitory effects on cytochrome c reduction in vitro.9 Third, l-NAME may impair the urea cycle, in which l-arginine is a substrate for arginase, one of the critical enzymes of the urea cycle.24 Indeed, we confirmed that l-NAME significantly increased plasma urea concentrations in mice (authors’ unpublished data, 2002). Thus, multiple mechanisms other than simple inhibition of NO synthesis appear to be involved in the long-term vascular effects of l-NAME.
In the present study, l-NNA caused a comparable extent of coronary microvascular lesions in both strains, as did l-NAME. Thus, it is possible that nonspecific vascular effects may not be limited to l-NAME but may also be extended to other l-arginine analogues. Indeed, other l-arginine analogues, such as l-NNA and NG-monomethyl-l-arginine, have also been reported to have multiple actions, including inhibition of cytochrome c reduction,9 endothelial generation of superoxide anions,25 and inhibition of the endothelium-independent relaxation induced by amiloride (an inhibitor of Na+-H+ exchange) and dibutyryl cAMP (a membrane-permeable cAMP analogue).26
Difference Between Short-Term and Long-Term Effects of l-NAME
In contrast to its long-term effects, l-NAME acutely inhibits endothelial NO synthesis in vitro and in vivo.1–3⇓⇓ This notion is supported by the fact that short-term effects of l-NAME can be reversed by either cotreatment with l-arginine (but not with d-arginine) or by the administration of NO donors or cGMP analogues.1–3⇓⇓ Although the precise mechanisms for the difference between the short-term and long-term vascular effects of l-NAME remain to be elucidated, it is conceivable that metabolites of l-NAME may be accumulated at higher concentrations in blood vessels, exerting various unknown effects other than simple inhibition of endothelial NO synthesis, and that other compensatory mechanisms for NO bioavailability (eg, endothelial superoxide dismutase system and cofactors for eNOS) may be upregulated.
In summary, in the present study, we were able to demonstrate that the long-term vascular effects of l-NAME are not mediated by the inhibition of endothelial NO synthesis. The present results warrant reevaluation of the previous studies that used long-term treatment with l-NAME in vivo.
- ↵Luscher TF, Vanhoutte PM. The Endothelium: Modulator of Cardiovascular Function. Boca Raton, Fla: CRC Press; 1990.
- ↵Cayatte AJ, Palacino JJ, Horten K, et al. Chronic inhibition of nitric oxide production accelerates neointima formation and impairs endothelial function in hypercholesterolemic rabbits. Arterioscler Thromb. 1994; 14: 753–759.
- ↵Buxton ILO, Cheek DJ, Eckman D, et al. NG-Nitro-l-arginine methyl ester and other alkyl esters of arginine are muscarinic receptor antagonists. Circ Res. 1993; 72: 387–395.
- ↵Tsutsui M, Milstien S, Katusic ZS. Effect of tetrahydrobiopterin on endothelial function in canine middle cerebral arteries. Circ Res. 1996; 79: 336–342.
- ↵Wilcox JN, Subramanian RR, Sundell CL, et al. Expression of multiple isoforms of nitric oxide synthase in normal and atherosclerotic vessels. Arterioscler Thromb Vasc Biol. 1997; 17: 2479–2488.
- ↵Boulanger CM, Heymes C, Benessiano J, et al. Neuronal nitric oxide synthase is expressed in rat vascular smooth muscle cells: activation by angiotensin II in hypertension. Circ Res. 1998; 83: 1271–1278.
- ↵Li Y, Zhu H, Kuppusamy P, et al. Validation of lucigenin (bis-N-methylacridinium) as a chemilumigenic probe for detecting superoxide anion radical production by enzymatic and cellular systems. J Biol Chem. 1998; 273: 2015–2023.
- ↵Day IN, Hinks LJ, Thompson RJ. The structure of the human gene encoding protein gene product 9:5 (PGP9.5), a neuron-specific ubiquitin C-terminal hydrolase. Biochem J. 1990; 268: 521–524.
- ↵Luvara G, Pueyo ME, Philippe M, et al. Chronic blockade of NO synthase activity induces a proinflammatory phenotype in the arterial wall: prevention by angiotensin II antagonism. Arterioscler Thromb Vasc Biol. 1998; 18: 1408–1416.
- ↵Tsutsui M, Morishita T, Tanimoto A, et al. Vasculoprotective role of neuronal nitric oxide synthase in the vascular lesion formation in mice. Circulation. 2000; 102 (suppl II): II-592.Abstract.
- ↵Horiuchi M, Tsutsui M, Morishita T, et al. Functional significance of neuronal nitric oxide synthase in arteriosclerotic vascular lesions induced by balloon injury in rats. Circulation. 2001; 104 (suppl II): II-207.Abstract.
- ↵Sanada S, Kitakaze M, Node K, et al. Differential subcellular actions of ACE inhibitors and AT1 receptor antagonists on cardiac remodeling induced by chronic inhibition of NO synthesis in rats. Hypertension. 2001; 38: 404–411.
- ↵Brusilow SW, Horwich AL. Urea cycle enzymes. In: Scriver CR, ed. The Metabolic Basis of Inherited Disease. New York, NY: McGraw-Hill; 1989: 629–663.
- ↵Heim KF, Thomas G, Ramwell PW. Effect of substituted arginine compounds on superoxide production in the rabbit aorta. J Pharmacol Exp Ther. 1991; 257: 1130–1135.