Role of Nox isoforms in angiotensin II-induced oxidative stress and endothelial dysfunction in brain

Abstract

Angiotensin II (Ang II) promotes vascular disease through several mechanisms including by producing oxidative stress and endothelial dysfunction. Although multiple potential sources of reactive oxygen species exist, the relative importance of each is unclear, particularly in individual vascular beds. In these experiments, we examined the role of NADPH oxidase (Nox1 and Nox2) in Ang II-induced endothelial dysfunction in the cerebral circulation. Treatment with Ang II (1.4 mg·kg⁻¹·day⁻¹ for 7 days), but not vehicle, increased blood pressure in all groups. In wild-type (WT; C57Bl/6) mice, Ang II reduced dilation of the basilar artery to the endothelium-dependent agonist acetylcholine compared with vehicle but had no effect on responses in Nox2-deficient (Nox2^−/y) mice. Ang II impaired responses to acetylcholine in Nox1 WT (Nox1^+/y) and caused a small reduction in responses to acetylcholine in Nox1-deficient (Nox1^−/y) mice. Ang II did not impair responses to the endothelium-independent agonists nitroprusside or papaverine in either group. In WT mice, Ang II increased basal and phorbol-dibutyrate-stimulated superoxide production in the cerebrovasculature, and these increases were abolished in Nox2^−/y mice. Overall, these data suggest that Nox2 plays a relatively prominent role in mediating Ang II-induced oxidative stress and cerebral endothelial dysfunction, with a minor role for Nox1.

the negative impact of chronic hypertension on the cerebral circulation and brain are great. Hypertension is a major risk factor for stroke and dementia, including Alzheimer's disease (17, 25). The renin-angiotensin system, and particularly angiotensin II (Ang II), underlie many of the changes in vascular structure and function that occur in animal models of hypertension as well as in patients with hypertension (1, 4, 9, 11, 17, 25, 36). Ang II also promotes atherosclerosis and vascular disease in the presence of other cardiovascular risk factors including diabetes and aging (3, 32, 35).

Production of reactive oxygen species (ROS) by vascular cells is increased in response to Ang II via several mechanisms, including activation of NADPH oxidases (9, 11, 21, 30, 36). For example, intravenous infusion of Ang II acutely increases blood pressure and levels of ROS (5–7, 24), and impairs vascular function including endothelium-dependent responses in brain (5, 7, 24). Endothelial dysfunction caused by Ang II can be reversed by scavengers of ROS and augmented by genetic deficiency in superoxide dismutases (5, 10, 13, 21, 24), providing strong evidence that oxidative stress is a key mediator of the vascular dysfunction.

A role for both Nox1- and Nox2-containing NADPH oxidases has been implicated in Ang II-induced oxidative stress and endothelial dysfunction in the peripheral circulation (14, 15, 27). To date, studies of NADPH oxidase in the cerebral circulation have focused on the role of Nox2, even though relatively high levels of Nox1 are expressed in brain vascular endothelium (2). For example, one report provided evidence that Nox2 mediates oxidative stress and endothelial dysfunction following acute treatment with Ang II in the cerebral microcirculation (24). In addition, there is evidence for Nox2 involvement in mediating oxidative stress and vascular dysfunction in models of diabetes and aging (9, 11, 28, 34). However, to our knowledge, the role of Nox2 in mediating vascular dysfunction following more chronic Ang II treatment in the cerebral circulation has not been studied. Furthermore, no studies have examined the functional importance of Nox1 in the cerebral vasculature. Because cerebral blood vessels appear to express relatively high levels of Nox1 (2), this isoform of Nox may exert important effects within the vessel wall. Thus the goal of this study was to examine the involvement of two sources of superoxide, Nox1- and Nox2-containing NADPH oxidases, in Ang II-induced endothelial dysfunction in cerebral arteries. In addition to NADPH oxidase, endothelial NO synthase (eNOS) is a potential source of superoxide if the enzyme becomes uncoupled (16, 37). Thus, as part of the effort to define sources of superoxide that mediate cerebrovascular dysfunction, we also examined whether eNOS is involved in Ang II-induced endothelial dysfunction.

METHODS

Experimental animals.

Mice deficient in Nox1 were derived from breeding male Nox^−/y (23) and female C57Bl/6J mice to generate male Nox1^+/y and female Nox1^+/− mice. These mice were then crossed, thus providing both genotypes and littermate controls. Only male mice were studied. Male Nox2-deficient (Nox2^−/y) and eNOS-deficient (eNOS^−/−) mice were purchased from Jackson Laboratories, using age-matched C57Bl/6J male mice as wild-type (WT) controls. Mice had access to regular chow and water ad libitum. All protocols and procedures conformed to the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of Iowa and the Animal Ethics Committee at Monash University.

Administration of Ang II and measurement of blood pressure.

Following anesthesia with ketamine/xylazine, an osmotic minipump (Alzet, model 1002) was placed subcutaneously in the midscapular region to administer vehicle (isotonic saline) or Ang II (1.4 mg·kg⁻¹·day⁻¹ for 7 days). With this dose of Ang II, a 7-day treatment produces a blood pressure increase comparable to that observed in other studies investigating vascular effects of Ang II treatment in models of Nox1 and Nox2 deficiency (23, 27, 38). Systolic blood pressure was measured using an automated tail-cuff device. Before surgery, mice were trained for 2–5 days, and baseline blood pressure was recorded, followed by implantation of pumps and subsequent measurements of arterial pressure.

Studies of basilar arteries.

Isolation and preparation of basilar arteries was described previously (10, 21, 22, 32). We initially recorded changes in diameter of the basilar artery in response to KCl (50 mM). To examine responses to acetylcholine (endothelium-dependent vasodilator) and nitroprusside (endothelium-independent vasodilator), arteries were constricted submaximally (∼40% of the response to KCl) using the thromboxane A₂ analog, U46619. After development of a stable baseline diameter, concentration-response curves were obtained. Papaverine (100 μM), another endothelium-independent vasodilator, was added at the end of each experiment, and responses were recorded. In some experiments, we also tested the effect of A23187, an endothelium-dependent vasodilator that acts in a receptor-independent manner. Although arteries were pressurized in these experiments, some preconstriction was necessary to induce sufficient tone to allow subsequent study of vasodilator responses. This approach is very common in the literature. Like the vast majority of studies of this type, the arteries were studied in the absence of flow.

Measurement of ROS.

L-012 (100 μM) chemiluminesence was used to measure superoxide levels in cerebral arteries (pooled samples of basilar artery, middle cerebral arteries, and the circle of Willis) in WT and Nox2^−/y mice. Measurements were made under basal conditions and in the presence of phorbol-12,13-dibutyrate (PdB;10 μM). PdB acutely increases Nox2-dependent superoxide production and was used to test whether superoxide levels could be further increased as a consequence of chronic Ang II treatment. All readings were corrected for background.

Drugs.

Acetylcholine, Ang II, nitroprusside, A23187, and papaverine were obtained from Sigma and were dissolved in saline. U46619 was obtained from Cayman Chemical (Ann Arbor, MI) and dissolved in 100% ethanol, with subsequent dilutions being made with Krebs buffer. Phorbol-12,13-dibutyrate was purchased from Calbiochem and prepared at 10 mM in dimethyl suplhoxide and diluted in Krebs-HEPES solution such that the final concentration of dimethyl sulfoxide was ≤0.1%. L-012 was purchased from Wako Pure Chemicals and was prepared at 100 mM in dimethyl sulfoxide and diluted in Krebs-HEPES solution such that the final concentration of dimethyl sulfoxide was ≤0.1%.

Statistical analysis.

All data are expressed as means ± SE. Vasodilator responses are expressed as % dilation (% of induced tone), with 100% representing the difference between the resting value under basal conditions and the constricted value with U46619. Vasoconstriction to KCl is expressed as % change in diameter over baseline. For experiments using L012-enhanced chemiluminesence, data were normalized to dry tissue weight and expressed as counts·s⁻¹·mg⁻¹. Changes in blood pressure were calculated by subtracting baseline blood pressure from the average blood pressure over days 5–7 of treatment. Comparisons of vasodilation and vasoconstriction, superoxide levels, and blood pressure were made using two-way ANOVA or Students t-test as appropriate. Statistical significance was accepted at P < 0.05.

RESULTS

Nox2 mice.

Blood pressure was similar under baseline conditions in Nox2^+/y and Nox2^−/y mice (Table 1). In WT mice, treatment with vehicle had no effect, whereas treatment with Ang II increased arterial pressure (Table 1). Similarly, in Nox2^−/y mice, vehicle treatment had no effect, whereas Ang II treatment increased blood pressure (Table 1), indicating that deficiency in Nox2 did not attenuate increases in arterial pressure in response to Ang II. These results are consistent with previous findings (38).

Table 1.

Various parameters in all treatment groups

	Baseline SBP, mmHg	Treatment Subgroup	ΔSBP, mmHg	Baseline Diameter, μm
Nox2+/y	109 ± 3 (n = 12)	Vehicle	−3 ± 6 (n = 6)	148 ± 7 (n = 6)
		Ang II	24 ± 5 (n = 6) *	155 ± 6 (n = 6)
Nox2−/y	103 ± 4 (n = 13)	Vehicle	−1 ± 5 (n = 6)	144 ± 4 (n = 6)
		Ang II	32 ± 9 (n = 7) *	141 ± 7 (n = 6)
Nox1+/y	122 ± 6 (n = 16)	Vehicle	−10 ± 7 (n = 10)	151 ± 9 (n = 11)
		Ang II	26 ± 8 (n = 6) *	162 ± 3 (n = 6)
Nox1−/y	109 ± 4 (n = 24) ***	Vehicle	−5 ± 7 (n = 12)	172 ± 9 (n = 12)
		Ang II	17 ± 6 (n = 12) *	153 ± 4 (n = 12)
eNOS+/+	116 ± 7 (n = 9)	Vehicle	7 ± 20 (n = 4)	138 ± 10 (n = 4)
		Ang II	35 ± 8 (n = 5)	143 ± 5 (n = 5)
eNOS−/−	136 ± 6 (n = 13) **	Vehicle	−1 ± 8 (n = 8)	107 ± 8 (n = 8) ****
		Ang II	29 ± 7 (n = 5) *	109 ± 9 (n = 8)

Baseline blood pressures and basilar artery diameters, as well as changes in systolic blood pressure (SBP) in response to vehicle and angiotensin (Ang) II treatment in all groups of mice. Nox, NADPH oxidase; eNOS, endothelial NOS.

^*Significant difference vs. corresponding vehicle (P < 0.05).

^**Difference vs. eNOS^+/+ (P = 0.052).

^***Difference vs. Nox1^+/y (P = 0.095).

^****Significant difference vs. vehicle-treated eNOS^+/+ (P < 0.05).

Baseline diameter of the basilar artery was similar in vehicle-treated Nox2^+/y and Nox2^−/y mice, and was similar in Nox2^+/y mice treated with either vehicle or Ang II (Table 1). Ang II treatment impaired responses to acetylcholine compared with vehicle treatment by up to 45% in Nox2^+/y mice (Fig. 1A). Vasomotor responses to papaverine (Fig. 1B) and KCl (Fig. 1D) were not affected by Ang II. There was a small but significant increase in response to nitroprusside following treatment with Ang II (Fig. 1C). Baseline diameter of the basilar artery was similar in Nox2^−/y mice treated with either vehicle or Ang II (Table 1). In contrast to Nox2^+/y mice, Ang II treatment had no effect on vasodilator responses to acetylcholine compared with vehicle treatment (Fig. 2A). Vasomotor responses to papaverine (Fig. 2B), nitroprusside (Fig. 2C), and KCl (Fig. 2D) were not significantly affected by Ang II.

Fig. 1.

Vascular responses to acetylcholine (A; vehicle, n = 6; Ang II, n = 6), papaverine (B; vehicle, n = 6; Ang II, n = 6), nitroprusside (C; vehicle, n = 6; Ang II, n = 6), and KCl (D; vehicle, n = 6; Ang II, n = 6) in Nox2^+/y mice. All data are means ± SE. Significant difference vs. vehicle: **P < 0.01; ***P < 0.001.

Fig. 2.

Vascular responses to acetylcholine (A; vehicle, n = 6; Ang II, n = 7), papaverine (B; vehicle, n = 6; Ang II, n = 7), nitroprusside (C; vehicle, n = 6; Ang II, n = 6), and KCl (D; vehicle, n = 6; Ang II, n = 6) in Nox2^−/y mice. All data are means ± SE.

Nox1 mice.

Baseline blood pressures tended to be lower in Nox1^−/y compared with Nox1^+/y mice, although this difference was not statistically significant (Table 1). In Nox1^+/y mice, vehicle treatment had no effect, whereas Ang II treatment increased blood pressure (Table 1), consistent with results in Nox2^+/y mice. In Nox1^−/y mice, treatment with vehicle had no significant effect on arterial pressure. Although Ang II treatment increased blood pressure in Nox1^−/y mice, the effect was not as great as in Nox1^+/y mice (Table 1). Lower baseline blood pressures and an attenuated pressor response to Ang II in Nox1^−/y mice are consistent with previous work (23).

Baseline diameter of the basilar artery was similar in vehicle-treated Nox1^+/y and Nox1^−/y mice, and similar in Nox1^+/y mice treated with either vehicle or Ang II (Table 1). Baseline vessel diameter was also similar in Nox1^−/y mice treated with either vehicle or Ang II (Table 1). Acetylcholine caused similar dilation in arteries from vehicle-treated Nox1^+/y and Nox1^−/y mice (Figs. 3A and 4A).

Fig. 3.

Vascular responses to acetylcholine (A; vehicle, n = 11; Ang II, n = 6), papaverine (B; vehicle, n = 11; Ang II, n = 7), nitroprusside (C; vehicle, n = 5; Ang II, n = 5), and KCl (D; vehicle, n = 9; Ang II, n = 8) in Nox1^+/y mice. All data are means ± SE. **Significant difference vs. vehicle (P < 0.01).

Fig. 4.

Vascular responses to acetylcholine (A; vehicle, n = 12; Ang II, n = 12), papaverine (B; vehicle, n = 12; Ang II, n = 12), nitroprusside (C; vehicle, n = 8; Ang II, n = 9), and KCl (D; vehicle, n = 12; Ang II, n = 12) in Nox1^−/y mice. All data are means ± SE. Significant difference vs. vehicle: *P < 0.05; **P < 0.01; ***P < 0.001.

Vasodilation to acetylcholine was impaired in Nox1^+/y mice treated with Ang II compared with vehicle by up to 65% (Fig. 3A), whereas Ang II had no effect on vascular responses to papaverine (Fig. 3B), nitroprusside (Fig. 3C), or KCl (Fig. 3D). In Nox1^−/y mice, Ang II treatment caused a small (∼25%) but significant inhibition of vasodilator responses to acetylcholine (Fig. 4A), whereas the response to papaverine (Fig. 4B) was unaffected. Vasodilation to nitroprusside was increased (Fig. 4C) and KCl-induced vasoconstriction was impaired in Ang II vs. vehicle-treated Nox1^−/y mice (Fig. 4D).

eNOS mice.

Baseline blood pressure was higher in eNOS^−/− mice compared with eNOS^+/+ mice (Table 1). In eNOS^+/+ mice, vehicle treatment had no effect, and although the effect was not statistically significant, the data indicate that Ang II treatment increased blood pressure (Table 1). Results in eNOS^+/+ mice are thus consistent with results in Nox2^+/y and Nox1^+/y mice. In eNOS^−/− mice, vehicle treatment had no effect, whereas Ang II treatment increased arterial pressure (Table 1).

Baseline diameter of the basilar artery was higher in vehicle-treated eNOS^+/+ mice compared with eNOS^−/− mice (Table 1). In eNOS^+/+ mice, baseline diameter was similar in mice receiving vehicle or Ang II (Table 1). Similar to Nox2^+/y and Nox1^+/y mice, Ang II impaired responses (by up to ∼60%) to acetylcholine in eNOS^+/+ mice compared with vehicle (Fig. 5A). There was no effect of Ang II treatment on vascular responses to papaverine (Fig. 5B). Responses to submaximal concentrations of nitroprusside were also modestly impaired by Ang II treatment (Fig. 5C).

Fig. 5.

Vascular responses to acetylcholine (A; vehicle, n = 4; Ang II, n = 5), papaverine (B; vehicle, n = 4; Ang II, n = 5), and nitroprusside (C; vehicle, n = 4; Ang II, n = 4) in eNOS^+/+ mice. **Significant difference vs. vehicle (P < 0.01).

Baseline diameter of the basilar artery in eNOS^−/− mice was similar in mice treated with vehicle or Ang II (Table 1). In contrast to Nox1^−/y and Nox2^−/y mice, Ang II markedly impaired responses (by ∼55%) to acetylcholine in eNOS^−/− mice (Fig. 6A). Responses to a second endothelium-dependent agonist (A23187) were also markedly impaired (by ∼65%) in eNOS^−/− mice (Fig. 6C). There was no effect of Ang II treatment on responses to papaverine (Fig. 6B), whereas responses to nitroprusside were impaired by Ang II treatment (Fig. 6D).

Fig. 6.

Vascular responses to acetylcholine (A; vehicle, n = 8; Ang II, n = 8), papaverine (B; vehicle, n = 7; Ang II, n = 8), A23187 (C; vehicle, n = 7; Ang II, n = 8), and nitroprusside (D; vehicle, n = 8; Ang II, n = 8) in eNOS^−/− mice. All data are means ± SE. Significant difference vs. vehicle: **P < 0.01; ***P < 0.0001.

Superoxide levels in cerebral vessels.

In this separate group of mice where effects of Ang II on superoxide levels were determined, blood pressure was also increased. In Nox2^+/y mice, resting blood pressure was 128 ± 5 mmHg (n = 16). Vehicle treatment had no effect, whereas Ang II treatment increased arterial pressure (Δ blood pressure in vehicle-treated mice: 3 ± 5 mmHg, n = 8; Δ blood pressure in Ang II-treated mice: 46 ± 2 mmHg, n = 8; P < 0.0001 vs. vehicle-treated). In Nox2^−/y mice, resting arterial pressure was 112 ± 5 mmHg (n = 10; P < 0.05 vs. WT mice). Vehicle treatment had no effect, whereas Ang II treatment increased arterial pressure (Δ blood pressure in vehicle-treated mice: 2 ± 6 mmHg, n = 5; Δ blood pressure in Ang II-treated mice: 45 ± 10 mmHg, n = 5; P < 0.01). These results are consistent with the blood pressure results we found in mice used for analysis of vascular function.

Levels of superoxide in cerebral arteries under basal conditions were increased in Ang II vs. vehicle-treated Nox2^+/y mice by approximately twofold, and this difference was abolished in Nox2^−/y mice (Fig. 7). In the presence of 10 μM PdB, superoxide levels were increased by ∼6.5-fold, confirming that acute Nox2 oxidase activation markedly increases superoxide in these vessels. Superoxide levels in the presence of PdB in cerebral arteries from Ang II compared with vehicle-treated WT mice were approximately twofold greater (Fig. 7), and this difference was abolished in Nox2^−/y mice, suggesting that Ang II increases cerebral vascular superoxide levels in a Nox2-dependent manner. Because the functional effect of Nox1 deficiency was modest, we did not repeat studies in those mice measuring superoxide.

Fig. 7.

Superoxide levels in cerebral arteries from vehicle and Ang II-treated mice under basal conditions and in the presence of 10 μM PdB in Nox2^+/y (vehicle, n = 9; Ang II, n = 9) and Nox2^−/y mice (vehicle, n = 7; Ang II, n = 7). All data are means ± SE. *Significant difference vs. vehicle-treated basal Nox2^+/y (P < 0.05). **Significant difference vs. vehicle-treated PdBu Nox2^+/y (P < 0.05). ***Significant difference vs. Ang II-treated basal Nox2^+/y (P < 0.05). ****Significant difference vs. vehicle and Ang II-treated PdBu Nox2^+/y (P < 0.05).

DISCUSSION

This study had several findings. First, we found that Ang II-induced endothelial dysfunction was partly dependent on expression of Nox1-containing NADPH oxidase. To our knowledge, this is the first demonstration (using any model) of a functional role for Nox1 in the cerebral circulation. Second, we found a major role for Nox2-containing NADPH oxidase in Ang II-induced oxidative stress and endothelial dysfunction, consistent with a previous study in the cerebral microcirculation where acute treatment with Ang II was used (24). Interestingly, endothelial dysfunction in response to Ang II was absent in Nox2-deficient mice despite a lack of effect of Nox2 deficiency on pressor effects of Ang II. These findings suggest that activity of Nox2 oxidase in the vessel wall, and not an elevation in blood pressure per se, is the mediator of oxidative stress and endothelial dysfunction, a concept consistent with previous studies (24). Third, although there is evidence to suggest that eNOS uncoupling may be an important mechanism of endothelial dysfunction in brain in other models (16, 37), we found no evidence for this mechanism in the current experiments. Studies of cerebral arteries (including the basilar artery) are important because these arteries are major resistance vessels in the brain (17, 18). In addition, these arteries are segments of the vasculature where some of the most important clinical complications of vascular disease occur.

Ang II-induced oxidative stress and endothelial dysfunction.

Acute or chronic treatment with Ang II causes oxidative stress and endothelial dysfunction in the carotid artery and cerebral circulation that is reversed by scavengers of ROS and is augmented by genetic deficiency in key antioxidants (5–7, 9, 10, 13, 21, 24). Consistent with these findings, we found that Ang II causes both increased superoxide levels and endothelial dysfunction in the cerebral circulation. Importantly, our finding of Ang II-induced endothelial dysfunction in WT mice using the model of Ang II infusion in the present study is consistent with previous findings in cerebral arteries using a genetic model of truly chronic Ang II-dependent hypertension (21). One mechanism by which superoxide impairs endothelium-dependent relaxation is by decreasing bioavailability of NO (9, 11, 17). The fact that endothelial dysfunction was inhibited by scavengers of ROS in several previous studies suggests impaired endothelium-dependent vasodilation in response to Ang II is mediated by superoxide.

Role of Nox isoforms.

Enzymatic subunits of NADPH oxidase, including Nox1 and Nox2, are expressed in cerebral blood vessels (reviewed in Ref. 9). For example, Nox2 is expressed in cerebral arteries (12, 31). We found that Nox2 deficiency under basal conditions has no effect on endothelial function, consistent with previous reports (24). Previous studies concluded that Ang II increases NADPH oxidase activity in the vasculature, resulting in increased production of superoxide (9, 24). In the present study, we found that Ang II produced oxidative stress and endothelial dysfunction in WT mice that was absent in Nox2^−/y mice, providing further support for the concept that superoxide is a key mediator of vascular dysfunction (11, 17). Our observations and those of others suggest that Ang II-induced endothelial dysfunction may be largely independent of increases in blood pressure (7, 10, 17). Direct effects of Ang II that impair vascular responses to endothelium-dependent stimuli have been seen in several models previously (9, 13, 24). In this context, and in relation to cerebrovascular disease, reductions in cerebral blood flow with age are prevented most effectively (and independent of effects on blood pressure) in hypertensive patients using antagonists of Ang II type 1 receptors compared with other anti-hypertensives (33).

Basal levels of superoxide were increased in response to Ang II in cerebral arteries, and this effect was abolished in Nox2^−/y mice. In addition, acute treatment of vessels with PdB increased superoxide levels, and this response was absent in Nox2^−/y mice, indicating that the increase in superoxide observed was Nox2 dependent. Furthermore, there was a marked increase in superoxide in cerebral arteries from Ang II vs. vehicle-treated WT mice following PdB treatment, perhaps reflecting increased Nox2 expression and/or activity following Ang II treatment and consistent with an important role for Nox2 activity during Ang II treatment in the cerebral circulation.

Nox1 is expressed in cerebral blood vessels (2, 9), and it has been suggested that levels of Nox1 in cerebral endothelium are relatively high compared with levels of Nox2 (2). Under basal conditions, genetic deletion of Nox1 had no effect on endothelial function. Thus Nox1 may not exert functional effects under normal conditions in the cerebral circulation. Although we observed a statistically significant effect of Ang II to attenuate endothelial function in Nox1^−/y mice, the effect was not as large as that observed in Nox1^+/y mice, suggesting some role for Nox1 in mediating endothelial dysfunction caused by Ang II. However, the role for Nox1 may be small in the context of the current model. Our functional and biochemical measurements using chemiluminescence suggest a much greater role for Nox2, further emphasizing the impact of this Nox isoform in the cerebral circulation. Taken together, these data suggest that Nox2-containing NADPH oxidase may be the predominant mediator of oxidative stress and endothelial dysfunction in brain in this model of hypertension. Very little is known regarding the functional importance of Nox1 in resistance vessels (including in the cerebral circulation). The present findings do not exclude the possibility that Nox1 plays a functionally more important role in other disease states.

Role of eNOS.

Under normal conditions, eNOS is a major mediator of endothelial function in cerebral arteries (19). Consistent with this concept, acute inhibition of eNOS inhibits the vast majority of the normal response of the basilar artery (including in the mouse) to acetylcholine (19, 22). In eNOS^−/− mice, responses to acetylcholine in the cerebral circulation are largely preserved due to compensatory mechanisms involving neuronal NOS (nNOS) (20, 29). Consistent with this concept, genetic deletion of eNOS had little effect on responses to acetylcholine in the current experiments. With this phenotype present, we hypothesized that, if Ang II-induced endothelial dysfunction was mediated by eNOS uncoupling, responses to acetylcholine would not be impaired following treatment of eNOS^−/− mice with Ang II. However, we found that, in the presence of eNOS deficiency, Ang II still produced endothelial dysfunction, suggesting no major role for eNOS uncoupling in this model. Thus these data are consistent with our other findings implicating Nox2 as a key source of superoxide in cerebrovascular dysfunction. Although our data do not support the possibility of eNOS being a key producer of superoxide during Ang II infusion, our data do not exclude a potential role for nNOS as a mediator of vascular dysfunction.

In conclusion, the present study supports the concept that NADPH oxidases play a major role in cerebrovascular disease and provides new insight into the relative importance of Nox1 and Nox2 in a model of Ang II-induced vascular disease. Of the two isoforms, Nox2 appears to be the more prominent mediator of the harmful effects of Ang II in the cerebral circulation during hypertension. Because Ang II plays an important role in promoting vascular disease in the face of other cardiovascular risk factors, the findings may have implications for other pathophysiological conditions as well (3, 32, 33, 35).

GRANTS

This work was supported by National Institutes of Health grants HL-38901, HL-62984, and NS-24621 and by a Bugher Foundation Award in Stroke from the American Heart Association (0575092N). S. Chrissobolis was supported by a CJ Martin Fellowship from the National Health and Medical Research Council of Australia (359282) and a postdoctoral fellowship from the American Heart Association (0725643Z).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: S.C., B.B., C.G.S., and F.M.F. conception and design of research; S.C. performed experiments; S.C. and F.M.F. analyzed data; S.C. and F.M.F. interpreted results of experiments; S.C. prepared figures; S.C. drafted manuscript; S.C., B.B., C.G.S., and F.M.F. approved final version of manuscript; C.G.S. and F.M.F. edited and revised manuscript.