Superoxide Excess in Hypertension and Aging
A Common Cause of Endothelial Dysfunction
There is evidence in humans that hypertension and aging similarly impair endothelial function, although the mechanism remains unclear. Superoxide anion (O2−) is a major determinant of nitric oxide (NO) bioavailability and thus endothelial function. We sought to determine the relationship between endothelial function, O2−, and age in normotensive Wistar-Kyoto (WKY) and stroke-prone spontaneously hypertensive rats (SHRSP). Aortic rings were removed from female WKY and SHRSP at 3 to 4 months (young) and 9 to 12 months (old). O2− generation by aortic rings was measured before and after removal of the endothelium or incubation with NG nitro-l-arginine methyl ester, diphenyleneiodonium, or apocynin. Levels of p22phox were studied with immunohistochemistry and used as a marker of NAD(P)H oxidase expression. NO bioavailability was significantly lower in old WKY compared with young WKY (P=0.0009) and in old SHRSP compared with young SHRSP (P=0.005). O2− generation was significantly greater in old WKY compared with young WKY (P=0.0001). Removal of the endothelium and NG nitro-l-arginine methyl ester treatment resulted in a significant reduction in O2− generation in old SHRSP (P=0.009 and 0.001, respectively). Diphenyleneiodonium significantly reduced O2− generation in 12-month WKY (P=0.008) and 12-month SHRSP (P=0.009). Apocynin attenuated O2− generation by older WKY (P=0.038) and SHRSP (P=0.028). p22phox was increased in older animals compared with young. We conclude that NO bioavailability decreases with age in female WKY and SHRSP. O2− generation increases with age in WKY and is higher in SHRSP and may contribute to the reduced NO by scavenging. NAD(P)H oxidase may contribute to the age-related increase in O2−.
There is evidence that in animal models and in humans, impaired endothelial function and a decrease in nitric oxide (NO) bioavailability may occur in hypercholesterolemia,1 2 , diabetes,3 and hypertension4 5 6 despite normal or increased NO production by the endothelium.6 A decrease in NO bioavailability may also occur with aging.7 8 9 10
In a number of animal models of disease, including hypertension11 12 and hypercholesterolemia,13 an increase in superoxide (O2−) occurs concurrent to the decrease in NO bioavailability. O2− rapidly reacts with NO, forming peroxynitrite and decreasing NO bioavailability.14 Thus, it has been proposed that elevations in O2− levels contribute to the impaired endothelial function associated with atherosclerotic disease.13 15
Taddei et al9 proposed that the endothelial dysfunction that occurs in hypertension represents an accelerated form of the dysfunction that occurs with aging. However, the effects of aging on O2− production are less well defined. Huraux and colleagues16 observed a negative correlation between O2− levels and age in human internal mammary arteries. In contrast, Berry et al17 found basal O2− production in human internal mammary arteries to be weakly but positively associated with age.
Potential vascular sources of O2− are endothelial NO synthase (eNOS),18 xanthine oxidase,19 and NAD(P)H oxidase.20 21 eNOS18 and NAD(P)H oxidase22 23 have been proposed to be involved in O2− production in different models of hypertension, whereas xanthine oxidase may be involved in O2− production in hypercholesterolemia.13 eNOS can be inhibited by arginine analogues such as NG nitro-l-arginine methyl ester (L-NAME). NAD(P)H oxidase is composed of at least 5 subunits, and apocynin can inhibit enzymatic activity by preventing association of the subunits. Diphenyleneiodonium (DPI) is a less specific inhibitor of flavin-containing oxidases, including NAD(P)H oxidase.
In this study, the hypothesis that both hypertension and aging result in increased levels of O2− and decreased NO bioavailability in blood vessels from Wistar-Kyoto rats (WKY) and stroke-prone spontaneously hypertensive rats (SHRSP) has been examined. The likely source(s) of O2− was also investigated.
Four groups of female rats were studied: 3- to 4-month-old WKY (n=28), 3- to 4-month-old SHRSP (n=28), 9- to 12-month-old WKY (n=48), and 9- to 12-month-old (n=46) SHRSP. Fewer young animals were used because studies comparing O2− production in young WKY and SHRSP animals had already been undertaken.12 The animals were obtained from the colonies established in Glasgow by brother-and-sister mating, as previously described.24 Blood pressure was measured by tail-cuff plethysmography 1 week before study, according to our published protocol.25 All experiments were approved by the Home Office according to regulations regarding experiments in animals in the United Kingdom.
The animals were given an overdose of barbiturate. The thoracic aorta and carotid arteries were removed, and periadventitial tissue was cleaned from the vessels. O2− was quantified in 4- to 5-mm segments by lucigenin chemiluminescence, as originally described by O’Hara et al13 and previously used by our group.12 17 In some experiments, the endothelium was removed by rubbing. In others, either L-NAME (0.1 mmol/L), DPI (0.1 mmol/L), or apocynin (3 mmol/L) was added 60 minutes before determining O2− generation. Control rings from the same animal were always assayed in parallel to each treatment. O2− generation was quantified against a standard curve of O2− generation by xanthine/xanthine oxidase. Tissue O2− was expressed as nanomoles per minute per milligram of wet weight.
Liochev and colleagues26 have reported that high concentrations of lucigenin may produce redox cycling leading to artificial increases in O2−. The concentration of lucigenin used for our initial studies (250 mmol/L) was relatively high; however, we wanted to be able to compare our results with these previously obtained in young animals.12 Studies in which a range of concentrations of lucigenin have been examined report no change17 or lower levels of O2− with lower concentrations of lucigenin but with any differences between experimental groups retained.27
NAD(P)H Oxidase Activity
Aortas and carotids were cleaned of any adhering connective tissue, rinsed, minced finely with scissors, and homogenized for 30 seconds with an Ultraturrax T8. The homogenate was centrifuged for 5 minutes at 1000g and the pellet discarded. Two milliliters of supernatant was taken for measurement of NAD(P)H oxidase activity by lucigenin chemiluminescence in the presence of 500 μmol/L NADH or NADPH and 25 μmol/L lucigenin. Protein concentrations were measured by the method of Bradford,28 and O2− generation was expressed as nanomoles per minute per milligram of protein.
Organ Bath Studies
Arteries were prepared as for measurement of O2−, except that they were cut into 2- to 3-mm rings. The rings were suspended under 1 g tension in individual 10-mL muscle baths containing physiological saline solution of the following composition (mmol/L): NaCl 130, KCl 4.7, NaHCO3 14.9, KH2PO4 1.18, MgSO4 0.7, H2O 1.17, CaCl2 0.2, H2O 1.6, glucose 5.5, and CaNa2 EDTA 0.03. The physiological saline solution was aerated with 95% O2/5% CO2, and indomethacin was added to a final bath concentration of 0.1 mmol/L to inhibit any prostanoid-mediated responses. Isometric tension was measured with Grass force transducers and displayed on a MacLab.
NO bioavailability was determined as previously described.6 Rings were constricted to their individual EC20 values to phenylephrine (PE). The NOS inhibitor L-NAME was added at a final concentration of 0.1 mmol/L. The increase in the contractile response was taken as a measure of NO bioavailability and expressed as a percentage of the PE EC20.
Small blocks of thoracic aortas from young and old rats were embedded in OCT and frozen at −70°C. Sections of 5 μm were cut, and immunohistochemistry was performed with standard techniques. Briefly, sections were blocked in 20% horse serum and then incubated overnight (in a humidified box) at 4°C with a monoclonal antibody against p22phox kindly supplied by Dr Mark Quinn. For negative control, the primary antibody was replaced with mouse IgG. Biotinylated anti-mouse (Vector Labs) at a dilution of 1:200 in 2% horse serum was incubated for 60 minutes followed by streptavidin conjugated to horseradish peroxidase. Color was developed by the addition of DAB (Sigma). The sections were lightly stained in hematoxylin and then dehydrated through alcohol and xylene. Sections were viewed and scored by an independent observer unaware of the age or genotype of the rats. Sections were scored as endothelial or medial, with 1 representing no staining, 2 representing faint brown, 3 moderate brown, and 4 intense brown.
Vessels from different animals were compared by 2-tailed unpaired t test, whereas vessels from the same animal with and without treatment were compared by a paired t test. Statistical significance was taken as P<0.05. Results are shown as mean±SEM, with 95% confidence intervals (CI) where significance was achieved. Bonferroni correction was applied for analysis of NADH/NADPH-driven O2− generation to allow for multiple comparisons.
The blood pressures (mm Hg±SEM) of the 4 groups of female rats studied were as follows: 3- to 4-month WKY, 117±1; 3- to 4-month SHRSP, 137±4; 9- to 12-month WKY, 115±2; and 9- to 12-month SHRSP, 141±3. Blood pressure was significantly higher in SHRSP than WKY at both ages (P<0.0001): 95% CI for 9- to 12-month WKY versus 9- to 12-month SHRSP, −33.6, −18.7, and for 3- to 4-month WKY versus 3- to 4-month SHRSP, −29.5, −11.8. No age-related effect was noted in either strain.
Basal NO Bioavailability
Addition of L-NAME caused an increase in the contractile response to PE in all groups studied. However, as shown in Figure 1a⇓, this increase (% of PE±SEM) was significantly lower in vessels for 9- to 12-month WKY (296±30, n=11) than for 3- to 4-month WKY (523±51, n=15, P=0.0009, 95% CI 144, 349) and in vessels from 9- to 12-month SHRSP (220±28, n=8) than for 3- to 4-month SHRSP (341±26, n=16, P=0.005; 95% CI, 41, 201).
O2− generation in aortas (nmol · min−1 · mg−1±SEM) was significantly higher in 9- to 12-month WKY (2.83±0.30, n=21) compared with 3- to 4-month WKY (1.06±0.2, n=7, P=0.001; 95% CI, 1.07, 2.54), but the difference between 9- to 12-month SHRSP (3.44±0.31 n=23) and 3- to 4-month SHRSP (2.98±0.49 n=9) did not reach statistical significance (Figure 1b⇑).
Similar increases in O2− levels with age and hypertension were observed in carotid arteries. O2− values of 0.88±0.18 (n=8) and 3.88±0.50 (n=12) were obtained in vessels from 3- to 4- and 9- to 12-month WKY, respectively (P=0.002; 95% CI, 1.39, 3.55), and 3.35±0.46 (n=12) and 4.89±0.88 (n=12) in 3- to 4- and 9- to 12-month SHRSP (P=0.19). These results are expressed per milligram of wet weight tissue. There was considerable hypertrophy of both carotid arteries and aortas from the older SHRSP, and it is possible that this resulted in an underestimation of O2− levels in these animals.
Sources of O2− in Aorta From 9- to 12-Month Animals
In older animals, incubation of the aortas with the NAD(P)H oxidase inhibitor DPI caused a significant decrease in O2− levels (nmol · min−1 · mg−1±SEM) from 2.13±0.30 to 0.89±0.18 (n=6, P=0.008) in WKY and from 3.04±0.43 to 1.19±0.14 (n=10, P=0.009; 95% CI, 0.39, 3.09) in SHRSP (Figure 2a⇓).
As shown in Figure 2b⇑, inhibition of NAD(P)H oxidase activity with apocynin also decreased O2− generation (nmol · min−1 · mg−1±SEM) in older animals, with levels being 1.86±0.25 and 1.06±0.36, respectively, in control and treated vessels from older WKY (n=7, P=0.038; 95% CI, 0.07, 1.78) and 2.29±0.53 and 1.44±0.43, respectively, in control and treated vessels from older SHRSP (n=7, P=0.028; 95% CI, 0.13, 1.57). In addition, apocynin had no significant effect in young WKY, with levels being 1.65±0.41 and 1.65±0.31 in control and treated vessels, respectively, but reduced O2− generation in aortas from young SHRSP from 2.36±0.47 to 1.48±0.27 (n=6, P=0.037; 95% CI, 0.08, 1.77).
The NOS inhibitor L-NAME had no significant effect on O2− generation in 9- to 12-month-old WKY, being 2.58±0.39 and 2.08±0.23 (n=9, P=0.08) in control and treated segments, respectively, but significantly reduced levels in 9- to 12-month-old SHRSP from 2.04±0.44 to 1.55±0.34 (n=6, P=0.02; 95% CI, 0.14, 0.79). Similarly, in WKY, the difference between control (3.42±0.34) and endothelium-denuded vessels (3.01±0.29, n=10) was not significant. In contrast, removal of the endothelium by rubbing decreased O2− levels in SHRSP from 3.63±0.38 to 2.79±0.18 (n=13, P=0.006; 95% CI, 0.30, 1.54).
NADH/NADPH-Driven O2− Production
In aortas and carotid arteries, NADH-driven O2− generation was greater than NADPH-driven O2− generation in all groups. Mean NADH- and NADPH-driven O2− generation was higher in older animals (Table⇓). This difference was significant for NADH-driven O2− generation in carotid arteries from 3- to 4-month versus 9- to 12-month WKY (P=0.038; 95% CI, 91, 5033) but failed to reach statistical significance in carotid arteries from SHRSP or in aorta from either WKY or SHRSP.
Representative sections from young and old WKY and SHRSP are shown in Figure 3⇓. Moderate brown staining was evident in the endothelium of the young vessels in both strains as 1±1 (Figure 3⇓, B and D), whereas the media was scored as 0±1 for both. In the older WKY rats (Figure 3C⇓), the endothelium scored 2±1, whereas that of the SHRSP (Figure 3E⇓) consistently scored 3. Moderate staining, 1±1, was present in the media of both old WKY and old SHRSP. Because much of the periadventitial tissue is routinely removed from these vessels, it was not possible to comment, reliably, on the staining patterns.
In these studies, we showed that both hypertension and aging result in a decrease in basal NO bioavailability and a corresponding increase in the generation of vascular O2− in female rats. We then went on to investigate the tissue and enzymatic sources of this excess O2−. In the older SHRSP but not WKY, both removal of the endothelium by rubbing and L-NAME treatment caused a significant reduction in O2− levels. Previously, we have made similar observations in young SHRSP,12 and the present results would substantiate our conclusion that eNOS is an important source of O2− in SHRSP.
However, eNOS is not the only the source of O2−. Both DPI and apocynin attenuated O2− production in vessels from SHRSP and older WKY. DPI is frequently used as an inhibitor of NAD(P)H pathways, although it has other actions, including inhibition of NOS.29 The vascular NAD(P)H oxidase consists of at least 5 subunits, with those that make up the membrane-bound cytochrome b558, p22phox, and gp91phox being important for the electron transport and the reduction of molecular oxygen to O2−. Apocynin acts by interfering with NAD(P)H subunit assembly in the membrane and is therefore a more specific inhibitor than DPI.20 Taken together, the inhibition of O2− production by these compounds would be consistent with a role for NAD(P)H oxidase as a source of O2−, particularly in older animals.
Further support for this hypothesis comes from the immunohistochemical studies that showed staining for p22phox in both WKY and SHRSP. Semiquantitatively, this staining was lowest in young WKY and highest in old SHRSP. However, both endothelial and vascular smooth muscle cell expression was upregulated in all the older rats.
As expected for vascular tissue, NADH generated tissue was greater than that generated by NADPH in both aortas and carotid arteries from all groups of animals studied. However, although NADH-generated O2− levels tended to be higher in the older animals, this only reached statistical significance for NADH-driven O2− generation in WKY carotid arteries. The immunohistochemical data suggested that p22phox levels were highest in the endothelium and lowest in vascular smooth muscle. The proportion of vascular smooth muscle was greater in blood vessels from older animals, which is likely to lead to an underestimation of the O2− generation per milligram of protein in the older animals. It is also possible that not all subunits of the NAD(P)H oxidase complex were upregulated to the same extent as p22phox in the older animals.
Although these studies indicate that NAD(P)H oxidase activity increases with age in female rats, these studies do not exclude an additional increase in O2− from other sources in the older animals. For example, although O2− generation from xanthine oxidase is negligible in young WKY and SHRSP, its contribution to O2− generation was not examined in older animals.12
In the studies reported here, a range of techniques was used to substantiate and extend our original findings. Taken together, these studies point to both eNOS and NAD(P)H oxidase as sources of O2− in SHRSP and suggest that the endothelium is an important source of O2− in both young and old SHRSP. In contrast, in young WKY, there is less endothelial involvement in O2− production. O2− generation by NAD(P)H oxidase appears to increase with age, and its primary source appears to be endothelium and adventitia.
All of the studies reported here were carried out in female rats. In contrast to female rats, we have previously observed no decrease in basal nitric oxide bioavailability with age in male WKY or SHRSP.10 Zalba et al22 found no difference in NAD(P)H-driven O2− production in aortas from 16- and 30-week-old male WKY, although an increase was observed in male SHR at 30 weeks. This could suggest that some of the age-related changes reported here are gender-specific. Decreased estrogen levels with age would provide a potential explanation because estrogen has been reported to act as an antioxidant decreasing LDL oxidation and uptake,30 to upregulate eNOS,31 and to decrease vascular O2− production.32 However, decreased estrogen levels are unlikely to be the cause of any of the age-related changes reported here. Most of the older animals used in our study were ex-breeders whose last litter had been weaned <1 month previously. Moreover, plasma estrogen levels do not differ significantly between 3- and 9-month-old animals (unpublished observations).
As with hypertension, the endothelial dysfunction with aging is due to reduced NO bioavailability as a result of scavenging by excess vascular O2− production. Endothelial NOS contributes significantly to O2− production in hypertensive animals, whereas NAD(P)H oxidase appears to be an important contributor to age-related increases in O2−.
This work was supported by the British Heart Foundation Program and Project Grants RG-97009 and PG-200023. The authors thank Emma Jardine for expert technical assistance.
- Received October 25, 2000.
- Revision received November 27, 2000.
- Accepted December 11, 2000.
Verhaar MC, Wever RM, Kastelein JJ, van Dam T, Koomans HA, Rabelink TJ. 5-methyltetrahydrofolate, the active form of folic acid, restores endothelial function in familial hypercholesterolemia. Circulation. 1998;97:237–241.
Rees DD, Ben-Ishay D, Moncada S. Nitric oxide and the regulation of blood pressure in hypertension-prone and hypertension-resistant Sabra rat. Hypertension. 1996;28:367–371.
McIntyre M, Hamilton CA, Rees DD, Reid JL, Dominiczak AF. Sex differences in the abundance of endothelial nitric oxide in a model of genetic hypertension. Hypertension. 1997;30:1517–1524.
Dohi Y, Thiel MA, Buhler FR, Luscher TF. Activation of endothelial L-arginine pathway in resistance arteries: effect of age and hypertension. Hypertension. 1990;16:170–179.
Taddei S, Virdis A, Mattei P, Ghiadoni L, Fasolo CB, Sudano I, Salvetti A. Hypertension causes premature aging of endothelial function in humans. Hypertension. 1997;29:736–743.
McIntyre M, Hamilton CA, Bohr DF, Reid JL, Dominiczak AF. Effects of age and gender on nitric oxide and superoxide in genetic hypertension. Hypertension. 1996;28:705. Abstract.
Bouloumie A, Bauersachs J, Linz W, Scholkens BA, Wiemer G, Fleming I, Busse R. Endothelial dysfunction coincides with an enhanced nitric oxide synthase expression and superoxide anion production. Hypertension. 1997;30:934–941.
Kerr S, Brosnan MJ, McIntyre M, Reid JL, Dominiczak AF, Hamilton CA. Superoxide anion production is increased in a model of genetic hypertension: the role of the endothelium. Hypertension. 1999;33:1353–1358.
O’Hara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest. 1993;91:2546–2551.
Rubanyi GM, Vanhoutte PM. Superoxide anions and hyperoxia inactivate endothelium-derived relaxing factor. Am J Physiol Heart Circ Physiol. 1986;250:H822–H827.
McIntyre M, Bohr DF, Dominiczak AF. Endothelial function in hypertension: the role of superoxide anion. Hypertension. 1999;34:539–545.
Huraux C, Makita T, Kurz S, Yamaguchi K, Szlam F, Tarpey MM, Wilcox JN, Harrison DG, Levy JH. Superoxide production, risk factors, and endothelium-dependent relaxations in human internal mammary arteries. Circulation. 1999;99:53–59.
Berry C, Hamilton CA, Brosnan MJ, Magill FG, Berg GA, McMurray JJV, Dominiczak AF. Investigation into the sources of superoxide in human blood vessels: angiotensin II increases superoxide production in human internal mammary arteries. Circulation. 2000;101:2206–2212.
Kerr S, McIntyre M, Brosnan MJ, Reid JL, Hamilton CA, Dominiczak AF. Endothelial generated superoxide contributes to endothelial dysfunction in SHRSP. Hypertension. 1998;32:796. Abstract.
Marczin N, Antonov A, Papapetropoulos A, Munn DH, Virmani R, Kolodgie FD, Gerrity R, Catravas JD. Monocyte-induced downregulation of nitric oxide synthase in cultured aortic endothelial cells. Arterioscler, Thromb, Vasc Biol. 1996;16:1095–1103.
Griendling KK, Sorescu D, Ushio-Fukai M. NAD(P)H oxidase: role in cardiovascular biology and disease. Circ Res. 2000;86:494–501.
Zalba G, Beaumont FJ, San Jose G, Fortuno A, Fortuno MA, Etayo JC, Diez J. Vascular NADH/NADPH oxidase is involved in enhanced superoxide production in spontaneously hypertensive rats. Hypertension. 2000;35:1055–1061.
Rajagopalan S, Kurz S, Munzel T, Tarpey M, Freeman BA, Griendling KK, Harrison DG. Angiotensin II-mediated hypertension in the rat increases vascular superoxide production via membrane NADH/NADPH oxidase activation: contribution to alterations of vasomotor tone. J Clin Invest. 1996;97:1916–1923.
Dominiczak AF, McLaren Y, Kusel J, Bell D, Goodfriend TL, Bohr DF, Reid JL. Lateral diffusion and fatty acid composition in vascular smooth muscle membrane from stroke-prone spontaneously hypertensive rats. Am J Hypertens. 1993;6:1003–1008.
Davidson AO, Schork N, Jaques BC, Kelman AW, Sutcliffe RG, Reid JL, Dominiczak AF. Blood pressure in genetically hypertensive rats. Influence of the Y chromosome. Hypertension. 1995;26:452–459.
Hough JL, Zilversmit DB. Effect of 17 beta estradiol on aortic cholesterol content and metabolism in cholesterol-fed rabbits. Arteriosclerosis. 1986;6:57–63.
Huang A, Sun D, Koller A, Kaley G. 17β-Estradiol restores endothelial nitric oxide release to shear stress in arterioles of male hypertensive rats. Circulation. 2000;101:94–100.
Barbacanne MA, Rami J, Michel JB, Souchard JP, Philippe M, Besombes JP, Bayard F, Arnal JF. Estradiol increases rat aorta endothelium-derived relaxing factor (EDRF) activity without changes in endothelial NO synthase gene expression: possible role of decreased endothelium-derived superoxide anion production. Cardiovasc Res. 1999;41:672–681.