Superoxide dismutase gene transfer reduces portal pressure in CCl4 cirrhotic rats with portal hypertension

BACKGROUND

Increased intrahepatic vascular tone in cirrhosis has been attributed to a decrease of hepatic nitric oxide (NO) secondary to disturbances in the post-translational regulation of the enzyme eNOS. NO scavenging by superoxide (O(2)(-)) further contributes to a reduction of NO bioavailability in cirrhotic livers.

AIM

To investigate whether removing increased O(2)(-) levels could be a new therapeutic strategy to increase intrahepatic NO, improve endothelial dysfunction and reduce portal pressure in cirrhotic rats with portal hypertension.

METHODS

Adenoviral vectors expressing extracellular superoxide dismutase (SOD) (AdECSOD) or beta-galactosidase (Adbetagal) were injected intravenously in control and CCl(4)-induced cirrhotic rats. After 3 days, liver O(2)(-) levels were determined by dihydroethidium staining, NO bioavailability by hepatic cGMP levels, nitrotyrosinated proteins by immunohistochemistry and western blot, and endothelial function by responses to acetylcholine in perfused rat livers. Mean arterial pressure (MAP) and portal pressure were evaluated in vivo.

RESULTS

Transfection of cirrhotic livers with AdECSOD produced a significant reduction in O(2)(-) levels, a significant increase in hepatic cGMP, and a decrease in liver nitrotyrosinated proteins which were associated with a significant improvement in the endothelium-dependent vasodilatation to acetylcholine. In addition, in cirrhotic livers AdECSOD transfection produced a significant reduction in portal pressure (17.3 (SD 2) mm Hg vs 15 (SD 1.6) mm Hg; p<0.05) without significant changes in MAP. In control rats, AdECSOD transfection prevents the increase in portal perfusion pressure promoted by an ROS-generating system.

CONCLUSIONS

In cirrhotic rats, reduction of O(2)(-) by AdECSOD increases NO bioavailability, improves intrahepatic endothelial function and reduces portal pressure. These findings suggest that scavenging of O(2)(-) might be a new therapeutic strategy in the management of portal hypertension.

Gene transfer after subarachnoid hemorrhage: a tool and potential therapy

This mini-review describes steps towards gene therapy to prevent vasospasm after subarachnoid hemorrhage, and summarizes some remaining obstacles. With recombinant adenoviruses, it is now possible to prevent vasospasm in experimental animals. If an adenoviral or other effective vector is demonstrated to be safe, it is likely that gene therapy will be used in patients to prevent vasospasm.

Mechanisms of Inducible Nitric Oxide Synthase–Mediated Vascular Dysfunction

Objective— Inducible nitric oxide synthase (iNOS) is expressed in arteries during inflammation and may contribute to vascular dysfunction. Effects of gene transfer of iNOS to carotid arteries were examined in vitro in the absence of systemic inflammation to allow examination of mechanisms by which iNOS impairs contraction and relaxation.

Methods and Results— After gene transfer of iNOS with an adenovirus (AdiNOS), constrictor responses to phenylephrine (PE) and U46619 were impaired. After AdiNOS, inhibition of soluble guanylate cyclase (sGC) with 1H-[1,2,4]oxadiazolo-[4,3,2]quinoxalin-1-one (ODQ) reduced the EC 50 for PE from 4.33±0.78 μmol/L to 1.15±0.43 μmol/L (mean±SEM). These results imply that iNOS impairs contraction by activation of the NO/cGMP pathway. Relaxation to acetylcholine (ACh) also was impaired after AdiNOS. Sepiapterin (300 μmol/L), the precursor for tetrahydrobiopterin (BH 4 ), improved relaxation to Ach. Because BH 4 is an essential cofactor for production of NO by both iNOS and endothelial nitric oxide synthase (eNOS), these results suggest that iNOS may reduce production of NO by eNOS by limiting availability of BH 4 . Next, we examined effects of expression of iNOS in endothelium and adventitia. Selective expression of iNOS in endothelium, but not adventitia, impaired contraction to phenylephrine and relaxation to acetylcholine.

Conclusions— We conclude that: (1) iNOS may impair contraction in part by activation of sGC; (2) iNOS impairs relaxation, at least in part, by limiting availability of BH 4 ; and (3) expression of iNOS in endothelium may be a more important mediator of vascular dysfunction than expression of iNOS in adventitia.

Gene transfer of inducible nitric oxide synthase impairs relaxation in human and rabbit cerebral arteries

BACKGROUND AND PURPOSE

These studies evaluated whether gene transfer of inducible nitric oxide synthase (iNOS) is a sufficient stimulus to produce vascular dysfunction in cerebral arteries.

METHODS

Intracranial (pial) arteries were dissected from human brain tissue obtained during elective surgery. Isolated human arteries were incubated in vitro with adenovirus containing iNOS (AdiNOS) or a nonexpressive transgene (control, AdBglII) (500 micro L, 3x10(9) plaque-forming units per milliliter), and vascular function was examined 24 hours later. In anesthetized rabbits, AdiNOS or AdBglII (300 microL 1x10(10)) was injected into the cisterna magna. Three days later, the basilar artery was removed, and reactivity was examined ex vivo.

RESULTS

In submaximally precontracted vessels, we observed impairment of NO-dependent relaxation in human cerebral arteries after gene transfer of iNOS. Maximum relaxation to bradykinin (1 micromol/L, an endothelium-dependent agonist) was 77+/-11% (mean+/-SE) after AdBglII and 31+/-22% (P<0.05) after AdiNOS. After AdiNOS, responses to nitroprusside (an endothelium-independent NO donor) also were impaired. Responses to both nitroprusside and bradykinin were improved by aminoguanidine (300 micromol/L), an inhibitor of iNOS. AdiNOS produced no change in vasoconstrictor responses to U46619. In basilar arteries from rabbits examined in vitro after gene transfer in vivo, responses to histamine, serotonin, and nitroprusside all were similar after AdiNOS or AdBglII. In contrast, relaxation to acetylcholine was significantly depressed after AdiNOS. Maximum relaxation to acetylcholine (10 micromol/L) was 90+/-3% after AdBglII and 68+/-5% (P<0.05) after AdiNOS. Relaxation of arteries after AdiNOS was improved by aminoguanidine.

CONCLUSIONS

These studies suggest that expression of iNOS may impair NO-dependent relaxation in both human and rabbit cerebral arteries.

Adenovirus-mediated overexpression of extracellular superoxide dismutase improves endothelial dysfunction in a rat model of hypertension

Gene transfer may be appropriate for therapeutic protocols targeted at the vascular endothelium. Endothelial dysfunction is the principal phenotype associated with atherosclerosis and hypertension. Oxidative stress has been implicated in the development of endothelial dysfunction. We have explored the ability of overexpressing anti-oxidant genes (superoxide dismutases; SODs) in vitro and in vivo to assess their potential for reversing endothelial dysfunction in a rat model, the stroke-prone spontaneously hypertensive rat (SHRSP). Western blotting and immunofluorescence assays in vitro showed efficient overexpression of MnSOD and ECSOD with respect to localisation to the mitochondria and extracellular surface, respectively. Transgene functional activity was quantified with SOD activity assays. MnSOD and ECSOD overexpression in intact SHRSP vessels in vivo led to endothelial and adventitial overexpression. Pharmacological assessment of transduced vessels following in vivo delivery by basal NO availability quantification demonstrated that the "null" adenovirus and MnSOD adenovirus did not significantly increase NO availability. However, AdECSOD-treated carotid arteries showed a significant increase in NO availability (1.91 +/- 0.04 versus 0.75 +/- 0.08 g/g, n = 6, P = 0.029). In summary, efficient overexpression of ECSOD, but not MnSOD in vivo, results in improved endothelial function in a rat model of hypertension and has important implications for the development of endothelial-based vascular gene therapy.

What's new in the cerebral microcirculation? Landis Award lecture

The first part of this paper focuses on unusual aspects of the cerebral circulation. Cerebral vessels have less smooth muscle and adventitia than other vessels, and the endothelial blood-brain barrier is unique. Because the wall of the arteries is thin, one might expect that the vessels are especially vulnerable to rupture. Pressure in intracranial arteries, however, is lower than in other arteries, because resistance of larger cerebral arteries is remarkably high. The low pressure in cerebral arteries presumably protects against rupture of the vessels. The second part of the paper summarizes some new insights into regulation of cerebral circulation. One concept is that "breakthrough" of autoregulation, with dilatation of cerebral vessels at high levels of pressure, is an active process, rather than a passive phenomenon. This conclusion is based on the finding that inhibitors of calcium-dependent potassium channels greatly attenuate the cerebral vasodilator response during acute hypertension. The third part of the paper focuses on effects of gene transfer to cerebral blood vessels. Gene transfer to intracranial and extracranial vessels is feasible and vasomotor function can be altered. Gene transfer has proven to be useful to study vascular biology, and we are optimistic that the approach will ultimately lead to gene therapy.

Cationic polymer and lipids augment adenovirus-mediated gene transfer to cerebral arteries in vivo

Adenovirus-mediated gene transfer to blood vessels is relatively inefficient because binding of adenovirus to vessels is limited. The authors have reported that incorporation of cationic polymer and lipids with adenovirus augments gene transfer to blood vessels ex vivo. In this study, the authors determined whether complexes of adenovirus and cations improve efficiency of gene transfer in vivo. Poly-L-lysine, lipofectamine, or lipofectin was complexed with adenovirus encoding beta-galactosidase. Optimum ratios of the cations per adenovirus were determined by gene transfer to fibroblasts. After injection of the adenovirus into the cisterna magna of anesthetized rabbits, transgene activity was greater in the adventitia of intracranial arteries and meninges after injection of the complexes than adenovirus alone. Thirty minutes after application of adenovirus with the cations, binding of adenovirus to fibroblast cells in vitro or the basilar artery in vivo (by Southern blot analysis) was augmented, which suggests that enhanced binding of virus contributes to augmentation of transgene expression. Thus, cationic polymer and lipids improve transgene expression in intracranial arteries, primarily in the adventitia, after adenovirus-mediated gene transfer in vivo. This strategy may be applicable to studies of gene transfer and eventually for gene therapy.

NO-dependent vasorelaxation is impaired after gene transfer of inducible NO-synthase

Proinflammatory stimuli produce expression of inducible NO-synthase (iNOS) within blood vessels and are associated with impaired endothelium-dependent relaxation. Gene transfer of iNOS was used to test the hypothesis that expression of iNOS in blood vessels produces impairment of NO-dependent relaxation as well as contraction. An adenoviral vector containing cDNA for murine iNOS, AdCMViNOS, and a control virus, AdCMVBglII, were used for gene transfer to rabbit carotid arteries in vitro and in vivo. After gene transfer of iNOS in vitro, contractile responses to KCl, phenylephrine, and U46619 were impaired. Relaxation in response to acetylcholine, ADP, A23187, and nitroprusside was also impaired. For example, maximum relaxation of vessels to acetylcholine (10 micromol/L) was 78+/-4% (mean+/-SE) after AdBglII (10(10.5) plaque-forming units) and 34+/-5% after AdiNOS (10(10.5) plaque-forming units, P<0.05). NO-independent relaxation in response to 8-bromo-cGMP and papaverine was not impaired after AdiNOS. Contraction and relaxation were improved in carotid arteries expressing iNOS by aminoguanidine and L-N-iminoethyl lysine, inhibitors of iNOS. After intraluminal gene transfer of iNOS in vivo, contraction of vessels in vitro was normal, but responses to acetylcholine were impaired. In summary, the major finding is that NO-dependent relaxation is impaired in arteries after gene transfer of iNOS in vitro and in vivo. Thus, expression of iNOS per se impairs NO-dependent relaxation.

Arachidonate dilates basilar artery by lipoxygenase-dependent mechanism and activation of K(+) channels

Dilatation of cerebral arterioles in response to arachidonic acid is dependent on activity of cyclooxygenase. In this study, we examined mechanisms that mediate dilatation of the basilar artery in response to arachidonate. Diameter of the basilar artery (baseline diameter = 216 +/- 7 micrometer) (means +/- SE) was measured using a cranial window in anesthetized rats. Arachidonic acid (10 and 100 microM) produced concentration-dependent vasodilatation that was not inhibited by indomethacin (10 mg/kg iv) or N(G)-nitro-L-arginine (100 microM) but was inhibited markedly by baicalein (10 micrometerM) or nordihydroguaiaretic acid (NDGA; 10 microM), inhibitors of the lipoxygenase pathway. Dilatation of the basilar artery was also inhibited markedly by tetraethylammonium ion (TEA; 1 mM) or iberiotoxin (50 nM), inhibitors of calcium-dependent potassium channels. For example, 10 microM arachidonate dilated the basilar artery by 19 +/- 7 and 1 +/- 1% in the absence and presence of iberiotoxin, respectively. Measurements of membrane potential indicated that arachidonate produced hyperpolarization of the basilar artery that was blocked completely by TEA. Incubation with [(3)H]arachidonic acid followed by reverse-phase and chiral HPLC indicated that the basilar artery produces relatively small quantities of prostanoids but large quantities of 12(S)-hydroxyeicosatetraenoic acid (12-S-HETE), a lipoxygenase product. Moreover, the production of 12-HETE was inhibited by baicalein or NDGA. These findings suggest that dilatation of the basilar artery in response to arachidonate is mediated by a product(s) of the lipoxygenase pathway, with activation of calcium-dependent potassium channels and hyperpolarization of vascular muscle.

Endothelial dysfunction and elevation of S-adenosylhomocysteine in cystathionine beta-synthase-deficient mice

Hyperhomocysteinemia is associated with increased risk for cardiovascular events, but it is not certain whether it is a mediator of vascular dysfunction or a marker for another risk factor. Homocysteine levels are regulated by folate bioavailability and also by the methyl donor S-adenosylmethionine (SAM) and its metabolite S-adenosylhomocysteine (SAH). We tested the hypotheses that endothelial dysfunction occurs in hyperhomocysteinemic mice in the absence of folate deficiency and that levels of SAM and SAH are altered in mice with dysfunction. Heterozygous cystathionine beta-synthase-deficient (CBS(+/-)) and wild-type (CBS(+/+)) mice were fed a folate-replete, methionine-enriched diet. Plasma levels of total homocysteine were elevated in CBS(+/-) mice compared with CBS(+/+) mice after 7 weeks (27.1+/-5.2 versus 8.8+/-1.1 micromol/L; P<0.001) and 15 weeks (23.9+/-3.0 versus 13.0+/-2.3 micromol/L; P<0.01). After 15 weeks, but not 7 weeks, relaxation of aortic rings to acetylcholine was selectively impaired by 35% (P<0.05) and thrombomodulin anticoagulant activity was decreased by 20% (P<0.05) in CBS(+/-) mice. Plasma levels of folate did not differ between groups. Levels of SAH were elevated approximately 2-fold in liver and brain of CBS(+/-) mice, and correlations were observed between plasma total homocysteine and SAH in liver (r=0.54; P<0.001) and brain (r=0.67; P<0.001). These results indicate that endothelial dysfunction occurs in hyperhomocysteinemic mice even in the absence of folate deficiency. Endothelial dysfunction in CBS(+/-) mice was associated with increased tissue levels of SAH, which suggests that altered SAM-dependent methylation may contribute to vascular dysfunction in hyperhomocysteinemia.

Gene transfer to blood vessels: a research tool and potential therapy

Gene transfer to blood vessels is remarkably effective in altering vasomotor function, and has proven to be a useful tool for studying vascular biology. Gene therapy for cardiovascular diseases is an attractive new approach, which will be used initially for diseases in which pharmacologic approaches are not effective. Preliminary data suggest that two possible targets for gene therapy are prevention of cerebral vasospasm after subarachnoid hemorrhage and treatment of pulmonary hypertension. Perhaps, as better vectors are developed, common clinical problems such as hypertension and hypercholesterolemia also may become targets for gene therapy.

Mechanisms that produce nitric oxide-mediated relaxation of cerebral arteries during atherosclerosis

BACKGROUND AND PURPOSE

The first goal of the present study was to examine the hypothesis that relaxation of cerebral arteries to nitric oxide in primates is dependent on activation of soluble guanylate cyclase (sGC). The second goal was to determine whether the role of sGC in mediating responses to nitric oxide is altered in atherosclerosis.

METHODS

Basilar arteries from normal and atherosclerotic monkeys were studied in vitro. After precontraction with prostaglandin F(2alpha) (0.1 to 1 micromol/L), concentration-response curves to authentic nitric oxide (1 nmol/L to 1 micromol/L), sodium nitroprusside (10 nmol/L to 10 micromol/L; a nitric oxide donor), and papaverine (10 nmol/L to 10 micromol/L; a non-nitric oxide, non-sGC-dependent stimulus) were generated in the presence and absence of 1H-[1,2,4]-oxadiazolo[4,3-a]quinoxalin-1-one (ODQ; 1 and 10 micromol/L; an inhibitor of sGC). The effect of ODQ on basal tone of basilar arteries from normal and atherosclerotic monkeys was also examined.

RESULTS

Nitric oxide, sodium nitroprusside, and papaverine produced relaxation that was similar (P:>0.05) in normal and atherosclerotic monkeys. ODQ produced marked inhibition (P:<0.05) of vasorelaxation in response to nitric oxide and nitroprusside but not papaverine. For example, relaxation of the basilar artery in response to nitric oxide (0.1 micromol/L) was inhibited by approximately 85% and 73% by ODQ (1 micromol/L) in normal and atherosclerotic monkeys, respectively. ODQ produced contraction of the basilar arteries, and the increase in tension to ODQ was greater in normal (2.7+/-0.3 g; mean+/-SE) than in atherosclerotic monkeys (1.4+/-0.4 g; P:<0.05). In contrast, contraction to prostaglandin F(2alpha) was similar in the basilar artery from normal and atherosclerotic monkeys.

CONCLUSIONS

These findings suggest that (1) relaxation of cerebral arteries in primates in response to nitric oxide is normally dependent, in large part, on activation of sGC and (2) the influence of sGC (via reduced production and/or activity of basal nitric oxide) on cerebral vascular tone is reduced in atherosclerosis.

Supplementation of atherogenic diet with B vitamins does not prevent atherosclerosis or vascular dysfunction in monkeys

BACKGROUND

Hyperhomocysteinemia is associated with increased risk of atherosclerotic and thrombotic vascular disease. In many patients, hyperhomocysteinemia can be treated or prevented by dietary supplementation with B vitamins, but the clinical benefit of B vitamins for the prevention of vascular disease has not been proven.

METHODS AND RESULTS

Using an atherogenic diet that produces both hyperhomocysteinemia and hypercholesterolemia, we tested the hypothesis that dietary supplementation with B vitamins (folic acid, vitamin B(12), and vitamin B(6)) would prevent hyperhomocysteinemia, vascular dysfunction, and atherosclerotic lesions in monkeys. After 17 months, plasma total homocysteine increased from 3.6+/-0.3 to 11.8+/-1.7 micromol/L in monkeys fed an unsupplemented atherogenic diet (P<0.01) but did not increase in monkeys fed an atherogenic diet supplemented with B vitamins (3.8+/-0.3 micromol/L). Serum cholesterol increased from 122+/-7 to 550+/-59 mg/dL in the unsupplemented group (P<0.001) and from 118+/-5 to 492+/-55 mg/dL in the supplemented group (P<0.001). Responses to endothelium-dependent vasodilators, both in resistance vessels in vivo and in the carotid artery ex vivo, were impaired to a similar extent in groups that did and did not receive vitamin supplements. Anticoagulant responses to the infusion of thrombin were also impaired to a similar extent in both groups. Vitamin supplementation failed to prevent intimal thickening in the carotid or iliac arteries.

CONCLUSIONS

These findings demonstrate that supplementation with B vitamins prevents hyperhomocysteinemia but is not sufficient to prevent the development of vascular dysfunction or atherosclerotic lesions in monkeys with marked hypercholesterolemia, even in the absence of preexisting atherosclerosis.

The future of gene therapy for stroke

New diagnostic and treatment strategies are being developed for stroke. Gene therapy has several potential advantages over classical pharmacologic therapy. Direct administration of DNA into the brain offers the advantage of producing high concentrations of therapeutic agents in a relatively localized environment. Gene transfer also provides longer duration of effect than traditional drug therapy. Recent studies indicate that gene transfer can produce functional proteins in brain parenchyma and cerebral blood vessels after stroke. In animal models, gene transfer may reduce effects of cerebral ischemia or subarachnoid hemorrhage. This review summarizes some current methods of gene transfer to the brain and recent progress that may lead to gene therapy for stroke.

Gene transfer of extracellular superoxide dismutase increases superoxide dismutase activity in cerebrospinal fluid

BACKGROUND AND PURPOSE

Copper-zinc superoxide dismutase (CuZnSOD) is expressed intracellularly, while extracellular SOD (EC-SOD) is released from cells. The purpose of this study was to determine whether gene transfer of CuZnSOD increases SOD activity predominantly in tissues, and gene transfer of EC-SOD increases SOD activity in cerebrospinal fluid (CSF). We also determined whether heparin or dextran sulfate releases EC-SOD into CSF.

METHODS

We injected recombinant adenoviruses expressing EC-SOD (AdEC-SOD), CuZnSOD (AdCuZnSOD), or beta-galactosidase (Adbeta-gal) into the cisterna magna of rabbits.

RESULTS

Total SOD activity in CSF was 39+/-11 U/mL (mean+/-SE) before virus injection. Three days later, total SOD activity in CSF increased to 148+/-22 U/mL after AdEC-SOD and 92+/-10 U/mL after AdCuZnSOD (P:<0.05 versus AdEC-SOD), with no change after Adbeta-gal (49+/-5 U/mL). EC-SOD protein was detected in CSF after AdEC-SOD but not AdCuZnSOD or Adbeta-gal. Injection of heparin or dextran sulfate into the cisterna magna increased total SOD activity 27-fold and 32-fold over basal values, respectively, in CSF of rabbits that received AdEC-SOD. In contrast to effects in CSF, total SOD activity in basilar artery and meninges was significantly higher after AdCuZnSOD and tended to be higher after AdEC-SOD than after Adbeta-gal.

CONCLUSIONS

-We have developed a method for intracranial gene transfer of CuZnSOD and EC-SOD. After gene transfer, CuZnSOD was expressed mainly in tissues, and EC-SOD was released into the CSF, especially after injection of heparin or dextran sulfate. Gene transfer of different isoforms of SOD may be useful in studies of cerebral vascular physiology and pathophysiology.

Tumor necrosis factor-alpha impairs contraction but not relaxation in carotid arteries from iNOS-deficient mice

We used mice deficient in expression of inducible NO synthase (iNOS -/-) to directly examine the role of iNOS in impaired vasoconstrictor responses following tumor necrosis factor-alpha (TNF-alpha). In iNOS +/+ mice, contraction of carotid arteries in response to prostaglandin F(2alpha) (PGF(2alpha)) was impaired following TNF-alpha (100 microg/kg ip)(n = 10, P < 0.01). In contrast to responses in wild-type mice, contraction to low concentrations of PGF(2alpha) were normal, but maximum contraction to PGF(2alpha) was impaired in arteries from iNOS -/- mice treated with TNF-alpha [0.35 +/-.0.02 g (n = 8) following vehicle and 0.25 +/- 0.02 g (n = 7) following TNF-alpha (P < 0.05)]. Aminoguanidine, a relatively selective inhibitor of iNOS, partially restored contraction to PGF(2alpha) in vessels from iNOS +/+ mice but had no effect in iNOS -/- mice injected with TNF-alpha, suggesting that a mechanism(s) other than iNOS contributes to impaired responses. In contrast to contractile responses, relaxation of the carotid artery in response to acetylcholine and nitroprusside was not altered following TNF-alpha in iNOS +/+ or iNOS -/-mice. Responses of carotid arteries from iNOS -/- mice and effects of aminoguanidine suggest that both iNOS-dependent and iNOS-independent mechanisms contribute to impaired contractile responses following TNF-alpha.

Gene transfer of calcitonin gene-related peptide prevents vasoconstriction after subarachnoid hemorrhage

We sought to determine whether adenovirus-mediated gene transfer in vivo of calcitonin gene-related peptide (CGRP), a potent vasodilator, ameliorates cerebral vasoconstriction after experimental subarachnoid hemorrhage (SAH). Arterial blood was injected into the cisterna magna of rabbits to mimic SAH 5 days after injection of AdRSVCGRP (8x10(8) pfu), AdRSVbetagal (control virus), or vehicle. After injection of AdRSVCGRP, there was a 400-fold increase in CGRP in cerebrospinal fluid. Contraction of the basilar artery to serotonin in vitro was greater in rabbits after SAH than after injection of artificial cerebrospinal fluid (P<0.001). Contraction to serotonin was less in rabbits with SAH after AdRSVCGRP than after AdRSVbetagal or vehicle (P:<0.02). Basal diameter of the basilar artery before SAH (measured with digital subtraction angiogram) was 13% greater in rabbits treated with AdRSVCGRP than in rabbits treated with vehicle or AdRSVbetagal (P:<0.005). In rabbits treated with vehicle or AdRSVbetagal, arterial diameter after SAH was 25+/-3% smaller than before SAH (P<0.0005). In rabbits treated with AdRSVCGRP, arterial diameter was similar before and after SAH and was reduced by 19+/-3% (P<0.01) after intracisternal injection of CGRP-(8-37) (0.5 nmol/kg), a CGRP(1) receptor antagonist. To determine whether gene transfer of CGRP after SAH may prevent cerebral vasoconstriction, we constructed a virus with a cytomegalovirus (CMV) promoter, which results in rapid expression of the transgene product. Treatment of rabbits with AdCMVCGRP after experimental SAH prevented constriction of the basilar artery 2 days after SAH. Thus, gene transfer of CGRP prevents cerebral vasoconstriction in vivo after experimental SAH.

IL-10 deficiency increases superoxide and endothelial dysfunction during inflammation

Little is known about the role of interleukin-10 (IL-10), an anti-inflammatory cytokine, in blood vessels. We used IL-10-deficient mice (IL-10 -/-) to examine the hypothesis that IL-10 protects endothelial function after lipopolysaccharide (LPS) treatment. The responses of carotid arteries were studied in vitro 6 h after injection of a relatively low dose of LPS (10 microgram ip). In IL-10 -/- mice, the maximum relaxation to ACh (3 microM) was 56 +/- 6% (means +/- SE) after LPS injection and 84 +/- 4% after vehicle injection (P < 0.05). Thus endothelium-dependent relaxation was impaired in carotid arteries from IL-10 -/- mice after LPS injection. In contrast, this dose of LPS did not alter relaxation to ACh in vessels from wild-type (IL-10 +/+) mice. Relaxation to nitroprusside and papaverine was similar in arteries from both IL-10 -/- and IL-10 +/+ mice after vehicle or LPS injection. Because inflammation is associated with increased levels of reactive oxygen species, we also tested the hypothesis that superoxide contributes to the impairment of endothelial function by LPS in the absence of IL-10. Results using confocal microscopy and hydroethidine indicated that levels of superoxide are elevated in carotid arteries from IL-10 -/- mice compared with IL-10 +/+ mice after LPS injection. The impaired relaxation of arteries from IL-10 -/- mice after LPS injection was restored to normal by polyethylene glycol-suspended superoxide dismutase (50 U/ml) or allopurinol (1 mM), an inhibitor of xanthine oxidase. These data provide direct evidence that IL-10 protects endothelial function after an acute inflammatory stimulus by limiting local increases in superoxide. The source of superoxide in this model may be xanthine oxidase.

Folate dependence of hyperhomocysteinemia and vascular dysfunction in cystathionine beta-synthase-deficient mice

Hyperhomocysteinemia is a risk factor for stroke, myocardial infarction, and venous thrombosis. Moderate hyperhomocysteinemia is associated with impaired endothelial function, but the mechanisms responsible for endothelial dysfunction in hyperhomocysteinemia are poorly understood. We have used genetic and dietary approaches to produce hyperhomocysteinemia in mice. Heterozygous cystathionine beta-synthase-deficient mice (CBS +/-), which have a selective defect in homocysteine transsulfuration, and wild-type (CBS +/+) littermates were fed either a control diet or a diet that is relatively deficient in folic acid for 6 wk. Plasma total homocysteine was 5.3 +/- 0.7 microM in CBS +/+ mice and 6.4 +/- 0.6 microM in CBS +/- mice (P = 0.3) given the control diet. Plasma total homocysteine was 11.6 +/- 4.5 microM in CBS +/+ mice and 25.1 +/- 3.2 microM in CBS +/- mice (P = 0.004) given a low-folate diet. In mice fed the control diet, relaxation of aortic rings in response to the endothelium-dependent vasodilator acetylcholine did not differ significantly between CBS +/+ mice and CBS +/- mice. In contrast, in mice fed a low-folate diet, maximal relaxation to acetylcholine was markedly impaired in CBS +/- mice (58 +/- 9%) compared with CBS +/+ mice (84 +/- 4%) (P = 0.01). No differences in relaxation to the endothelium-independent vasodilator sodium nitroprusside were observed among the four groups of mice. These data indicate that CBS-deficient mice are predisposed to hyperhomocysteinemia during dietary folate deficiency, and moderate hyperhomocysteinemia is associated with marked impairment of endothelial function in mice.

Gene transfer of endothelial nitric oxide synthase but not Cu/Zn superoxide dismutase restores nitric oxide availability in the SHRSP

OBJECTIVE

Previous studies from our group have shown a deficit in nitric oxide (NO) bioavailability and an excess production of the superoxide anion (O(2)(-)) in the stroke prone spontaneously hypertensive rat (SHRSP) compared to the normotensive Wistar Kyoto (WKY) strain. This present study has investigated whether adenoviral-mediated gene transfer of human eNOS or Cu/ZnSOD can alter the NO/O(2)(-) balance, thereby improving endothelial function.

METHODS

A recombinant adenovirus, Ad/Hu/eNOS, containing the human eNOS cDNA fragment was generated by homologous recombination in 293 cells. Ad/Hu/eNOS or Ad/Cu/ZnSOD was delivered into SHRSP carotid arteries in vivo, using a titre of 2x10(9)-2x10(10) plaque forming units (pfu)/ml, and the effect on gene expression was observed 24 h later.

RESULTS

Western blotting confirmed increased enzyme levels of eNOS and Cu/ZnSOD in the viral-infused vessels. Ex vivo, the pressor response to phenylephrine (PE) in the presence of L-NAME was increased in the eNOS-infused arteries relative to the contralateral controls, indicating restoration of basal NO availability to that observed in untreated control WKY rats. Infusion of the SOD virus produced a statistically insignificant increase in NO bioavailability.

CONCLUSIONS

Our results support our previous findings obtained using a bovine eNOS recombinant adenovirus, that recombinant adenoviral gene transfer of human eNOS has a significant effect on NO bioavailability. In contrast, AdCu/ZnSOD gene transfer does not elicit an effect in our model. These results indicate that short-term overexpression of a recombinant eNOS, but not Cu/ZnSOD gene, in carotid arteries of the SHRSP is an effective means of locally increasing NO bioavailability to improve endothelial function.

Calcium phosphate precipitates augment adenovirus-mediated gene transfer to blood vessels in vitro and in vivo

Adenovirus (Ad)-mediated gene transfer to blood vessels is relatively inefficient, probably because binding of adenovirus to the endothelium and adventitia seems to be limited. Association of calcium phosphate (CaPi) precipitates with adenovirus improves efficiency of gene transfer to some cells in culture and to mouse lung in vivo. In this study, we tested the hypothesis that CaPi is useful for adenovirus-mediated gene transfer to blood vessels. In fibroblast and endothelial cells in culture, Ad:CaPi coprecipitates greatly increased transgene expression. Ad:CaPi also enhanced transgene expression in both adventitia and endothelium of carotid arteries and aortae from rabbits studied ex vivo. After injection of Ad:CaPi into the cisterna magna of rabbits in vivo, the transgene product was markedly increased in leptomeninges of the ventral brain stem, including the adventitia of the basilar artery. We also examined mechanisms of enhanced gene transfer. Binding of adenovirus to fibroblast and endothelial cells in culture, and to the basilar artery in vivo, as determined using Southern blot analysis, was augmented by CaPi. Antibody to adenoviral fiber knob did not inhibit augmented transgene expression by Ad:CaPi. The finding suggests that improved adenoviral binding occurs primarily via a fiber-independent pathway. Thus, CaPi precipitates are useful for improvement of adenovirus-mediated gene transfer to blood vessels in vitro and in vivo.

Adenovirus-mediated gene transfer to cerebral circulation

Gene therapy may, be a promising approach for treatment of cerebrovascular disease. An adenoviral vector encoding beta-galactosidase was administered intracisternally or intraventricularly into the brain of rats. Efficient expression of the reporter gene was observed at the cerebral blood vessels and perivascular tissues. When the adenoviral vector was delivered into CSF of dogs suffering from subarachnoid hemorrhage, prominent expressions of transgene were observed. Introduction of the vector to the ischemic brain of rats provided efficient transgene expression in the peri-ischemic area. Therefore, gene transfer to the cerebral blood vessel and brain may be a promising approach for gene therapy of stroke. Atherosclerotic lesion plays an important role in stroke. We evaluated efficacy of adenovirus-mediated gene transfer to the atherosclerotic vessels from monkeys and rabbits using an ex vivo gene transfer system. Efficiency of transgene expression in the atherosclerotic endothelium was better than that of normal vessels in both animals. Thus, the endothelium of atherosclerotic vessels may be a good target for gene therapy. Next, we transfected atherosclerotic carotid arteries from rabbits with an adenoviral vector encoding endothelial nitric oxide synthase (eNOS). After overexpression of eNOS in the atherosclerotic arteries, the response to acetylcholine was augmented, showing similar relaxation with normal vessels. These results suggest that gene transfer to atherosclerotic vessels improves endothelial function, which may be a new therapeutic approach for cerebrovascular disease.

Potassium channels modulate cerebral autoregulation during acute hypertension

We tested the hypothesis that constriction of cerebral arterioles during acute increases in blood pressure is attenuated by activation of potassium (K(+)) channels. We tested the effects of inhibitors of calcium-dependent K(+) channels [iberiotoxin (50 nM) and tetraethylammonium (TEA, 1 mM)] on changes in arteriolar diameter during acute hypertension. Diameter of cerebral arterioles (baseline diameter = 46 +/- 2 microm, mean +/- SE) was measured using a cranial window in anesthetized rats. Arterial pressure was increased from a control value of 96 +/- 1 mmHg to 130, 150, 170, and 200 mmHg by intravenous infusion of phenylephrine. Increases in arterial pressure from baseline to 130 and 150 mmHg decreased the diameter of cerebral arterioles by 5-10%. Greater increases in arterial pressure produced large increases in arteriolar diameter (i.e., "breakthrough of autoregulation"). Iberiotoxin or TEA inhibited increases in arteriolar diameter when arterial pressure was increased to 170 and 200 mmHg. The change in arteriolar diameter at 200 mmHg was 20 +/- 3% and -1 +/- 4% in the absence and presence of iberiotoxin, respectively. These findings suggest that calcium-dependent K(+) channels attenuate cerebral microvascular constriction during acute increases in arterial pressure, and that increases in arteriolar diameter at high levels of arterial pressure are not simply a passive phenomenon.

Plasma concentration of asymmetric dimethylarginine, an endogenous inhibitor of nitric oxide synthase, is elevated in monkeys with hyperhomocyst(e)inemia or hypercholesterolemia

Hyperhomocyst(e)inemia is associated with endothelial dysfunction. Mechanisms responsible for endothelial dysfunction in hyperhomocyst(e)inemia may involve impaired bioavailability of endothelium-dependent nitric oxide. We tested the hypothesis that hyperhomocyst(e)inemia is associated with an elevated plasma concentration of asymmetric dimethylarginine (ADMA), an endogenous inhibitor of nitric oxide synthase. One group of adult cynomolgus monkeys was fed either a control or hyperhomocyst(e)inemic diet for 4 weeks in a randomized crossover design. The second group was fed an atherogenic diet that produces both hyperhomocyst(e)inemia and hypercholesterolemia for 17 months, followed by an atherogenic diet supplemented with B vitamins for 6 months to decrease plasma homocyst(e)ine concentration. Human endothelial cells were used to study the effects of methionine and homocysteine in the presence or absence of B vitamins or the methylation inhibitor S-adenosylhomocysteine on the formation of ADMA and its inactive stereoisomer, symmetric dimethylarginine. The hyperhomocyst(e)inemic diet produced 2- to 3-fold increases in plasma levels of homocyst(e)ine and ADMA (both P<0.05). The atherogenic diet also produced elevated plasma levels of homocyst(e)ine and ADMA (both P<0. 05). Supplementation of the atherogenic diet with B vitamins decreased the plasma levels of homocyst(e)ine but did not affect the plasma levels of ADMA or endothelial function. There was a strong correlation between plasma ADMA and homocyst(e)ine and a strong inverse correlation between ADMA and carotid artery relaxation to acetylcholine. ADMA release by cultured endothelial cells was significantly increased in the presence of methionine or homocysteine. This effect was blocked by S-adenosylhomocysteine but not by B vitamins. We conclude that plasma levels of ADMA are elevated in hyperhomocyst(e)inemia. Because ADMA acts as a competitive inhibitor of endothelial nitric oxide synthase, these findings suggest a novel mechanism for impaired endothelial function in hyperhomocyst(e)inemia.

Gene transfer of endothelial nitric oxide synthase improves relaxation of carotid arteries from diabetic rabbits

BACKGROUND

Diabetes mellitus is associated with impairment of NO-mediated vascular relaxation. The purpose of this study was to determine whether adenovirus-mediated gene transfer of endothelial NO synthase (eNOS) or Cu/Zn superoxide dismutase (SOD1) improves responsiveness to acetylcholine in alloxan-induced diabetic rabbits.

METHODS AND RESULTS

After 8 weeks, plasma glucose was greater in diabetic rabbits (418+/-35 mg/dL) (mean+/-SEM) than in normal rabbits (105+/-4 mg/dL). Carotid arteries were removed and cut into ring segments. Arteries were incubated for 2 hours with adenoviral vectors driven by a CMV promoter expressing beta-galactosidase (beta-gal), eNOS, SOD1, or vehicle. After incubation with virus, arteries were incubated for an additional 24 hours to allow transgene expression. Vascular reactivity was examined by recording isometric tension. After precontraction with phenylephrine, responses to the endothelium-independent vasodilator sodium nitroprusside were similar in diabetic and normal arteries. Endothelium-dependent relaxation to acetylcholine (3x10(-6) mol/L) was significantly less in arteries from diabetic animals (68+/-5%) than in normal vessels (90+/-3%). Adenoviral transfection of arteries with eNOS improved relaxation in response to acetylcholine in diabetic (EC(50) eNOS=0.64+/-0.12x10(-7) mol/L versus vehicle =1. 70+/-0.43x10(-7) mol/L) but not normal arteries. Vasorelaxation in response to acetylcholine was inhibited by N(omega)-nitro-L-arginine (100 micromol/L) in all groups. Responses to acetylcholine were unchanged after gene transfection of SOD1 or beta-gal in arteries from diabetic or normal rabbits.

CONCLUSIONS

Adenovirus-mediated gene transfer of eNOS, but not SOD, improves impaired NO-mediated relaxation in vessels from diabetic rabbits.

In vivo gene transfer of prepro-calcitonin gene-related peptide to the lung attenuates chronic hypoxia-induced pulmonary hypertension in the mouse

BACKGROUND

Calcitonin gene-related peptide (CGRP) is believed to play an important role in maintaining low pulmonary vascular resistance (PVR) and in modulating pulmonary vascular responses to chronic hypoxia; however, the effects of adenovirally mediated gene transfer of CGRP on the response to hypoxia are unknown.

METHODS AND RESULTS

In the present study, an adenoviral vector encoding prepro-CGRP (AdRSVCGRP) was used to examine the effects of in vivo gene transfer of CGRP on increases in PVR, right ventricular mass (RVM), and pulmonary vascular remodeling that occur in chronic hypoxia in the mouse. Intratracheal administration of AdRSVCGRP, followed by 16 days of chronic hypoxia (FIO(2) 0.10), increased lung CGRP and cAMP levels. The increase in pulmonary arterial pressure (PAP), PVR, RVM, and pulmonary vascular remodeling in response to chronic hypoxia was attenuated in animals overexpressing prepro-CGRP, whereas systemic pressure was not altered while in chronically hypoxic mice, angiotensin II and endothelin-1-induced increases in PAP were reduced, whereas decreases in PAP in response to CGRP and adrenomedullin were not changed and decreases in PAP in response to a cAMP phosphodiesterase inhibitor were enhanced by AdRSVCGRP.

CONCLUSIONS

In vivo CGRP lung gene transfer attenuates the increase in PVR and RVM, pulmonary vascular remodeling, and pressor responses in chronically hypoxic mice, suggesting that CGRP gene transfer alone and with a cAMP phosphodiesterase inhibitor may be useful for the treatment of pulmonary hypertensive disorders.

Gene transfer of calcitonin gene-related peptide to cerebral arteries

Overexpression of calcitonin gene-related peptide (CGRP), an extremely potent vasodilator, to blood vessels is a possible strategy for prevention of vasospasm. We constructed an adenoviral vector that encodes prepro-CGRP (Adprepro-CGRP) and examined the effects of gene transfer on cultured cells and cerebral arteries. Transfection of Adprepro-CGRP to Cos-7 and NIH-3T3 cells increased CGRP-like immunoreactivity in media and produced an increase in cAMP in recipient cells. Five days after injection of Adprepro-CGRP into the cisterna magna of rabbits, the concentration of CGRP-like immunoreactivity increased by 93-fold in cerebrospinal fluid. In basilar artery, cAMP increased by 2.3-fold after Adprepro-CGRP compared with a control adenovirus. After transfection of Adprepro-CGRP, contraction of basilar artery in vitro to histamine and serotonin was attenuated, and relaxation to an inhibitor of cyclic nucleotide phosphodiesterase 3-isobutyl-1-methylxanthine was augmented compared with nontransduced arteries or arteries transfected with a control gene. Altered vascular responses were restored to normal by pretreatment with a CGRP(1) receptor antagonist CGRP-(8-37). Thus gene transfer of prepro-CGRP in vivo overexpresses CGRP in cerebrospinal fluid and perivascular tissues and modulates vascular tone. We speculate that this approach may be useful in prevention of vasospasm after subarachnoid hemorrhage.

Gene transfer of endothelial nitric oxide synthase reduces angiotensin II-induced endothelial dysfunction

Angiotensin II stimulates vascular NADPH oxidase to produce superoxide, which can react with nitric oxide and impair vasomotor function. We tested the hypothesis that the overexpression of endothelial nitric oxide synthase (eNOS) or superoxide dismutase (SOD) would correct angiotensin II-induced endothelial dysfunction. We examined the effects of the gene transfer of eNOS or 2 isoforms of SOD to the aorta in angiotensin II-treated rabbits on vasomotor function. New Zealand White rabbits were treated for 1 week with angiotensin II (100 ng. kg(-1). min(-1)) or saline by osmotic minipumps. In angiotensin II-treated rabbits, mean blood pressure was 107+/-8 mm Hg; it was 67+/-5 mm Hg in saline-infused rabbits (P<0.05). In aortas from angiotensin II-treated rabbits, lucigenin-enhanced chemiluminescence demonstrated a 2.5-fold increase in superoxide levels, and the oxidative fluorescent probe hydroethidine indicated increased superoxide levels throughout the vascular wall, especially in the endothelium and adventitia. Maximal relaxation to acetylcholine was less in aortas from rabbits treated with angiotensin II (72+/-5% versus 87+/-4% in saline-treated rabbits; P<0.01), but responses to sodium nitroprusside were similar. Segments of the thoracic aorta were incubated in vitro with an adenoviral vector that expressed eNOS, copper zinc SOD (CuZnSOD), extracellular SOD (ECSOD), or beta-galactosidase. beta-Gal treatment with adenovirus containing the gene for eNOS (AdeNOS) but not adenovirus containing the gene for beta-gal (Adbeta-gal) (control virus) restored responses to acetylcholine (82+/-3% after AdeNOS and 67+/-4% after Adbeta-gal). Gene transfer of CuZnSOD or ECSOD did not improve the endothelium-dependent relaxation of the aorta in rabbits that received angiotensin II. Thus, gene transfer of eNOS, but not SOD, effectively restores vasomotor function in angiotensin II-infused rabbits.

Gene therapy of hypertensive vascular injury

Hypertension and vascular injury usually require prolonged treatment, and compliance is a key to efficacy for pharmacologically-based antihypertensive therapy. Gene therapy has the potential to be long lasting, with few side effects. Recent studies have provided promising results, in which hypertension can be treated by either augmentation of vasodilation or inhibition of vasoconstriction through gene transfer in experimental models. Gene transfer is also becoming useful for the study of mechanisms of physiologic and pathophysiologic conditions, including hypertension. In this mini-review, we summarize some recent studies in this area of research, and suggest some areas where progress is needed to advance the research toward gene therapy.

Approaches to enhance expression after adenovirus-mediated gene transfer to the carotid artery

The goal of this study was to enhance transgene expression after adenoviral-mediated gene transfer to the carotid artery. We used an adenoviral vector with a transgene that expresses beta-galactosidase, driven by the human cytomegalovirus (CMV) promoter/enhancer. The CMV promoter drives constitutive expression, and response elements within the enhancer allow inducible expression through binding of active transcription factors, such as cAMP response element binding protein (CREB) and nuclear factor kappa B (NFkappaB). Rings of rabbit carotid artery were incubated ex vivo with a replication-deficient adenovirus that expresses beta-galactosidase (AdCMV-betagal). Virus was removed from the medium, and forskolin or phorbol-12-myristate-13-acetate (PMA), which can induce activation of CREB or NFkappaB, respectively, were added to the medium. Pyrrolidine dithiocarbamate (PDTC) was used to inhibit activation of NFkappaB. Following incubation for 24 hours, beta-galactosidase activity was assessed by chemiluminescent reporter assay. Forskolin and PMA enhanced transgene expression in the carotid artery. Activity increased from 56+/-13 mU/mg protein (mean+/-SE) in rings of carotid treated with virus alone (10(9) pfu) to 159+/-23 mU/mg protein (P<0.05) in rings treated with forskolin, and to 189+/-40 mU/mg protein (P<0.05) in rings treated with PMA. Phorbol didecanoate, an inactive phorbol, did not affect expression of beta-galactosidase. After pre-incubation with PDTC prior to PMA, expression of beta-galactosidase was less than in rings incubated with PMA alone (29+/-11, P<0.05). Histochemical staining of carotid artery for beta-galactosidase demonstrated enhanced endothelial expression following administration of PMA. These findings suggest that expression after gene transfer to the carotid artery using an adenoviral vector with the CMV promoter/enhancer may be enhanced by PMA and forskolin, perhaps by activation of transcription factors.

Gene transfer of endothelial nitric oxide synthase (eNOS) in eNOS-deficient mice

Relaxation to acetylcholine (ACh) and calcium ionophore (A-23187) is absent in aortas from endothelial nitric oxide synthase (eNOS)-deficient (eNOS -/-) mice. We hypothesized that gene transfer of eNOS would restore relaxation to ACh and A-23187 in eNOS -/- mice. Aortic rings from eNOS -/- and eNOS +/+ mice were exposed in vitro to vehicle or adenoviral vectors encoding beta-galactosidase (lacZ) or eNOS. Histochemical staining for beta-galactosidase and eNOS demonstrated transduction of endothelial cells and adventitia. Vehicle-treated vessels from eNOS -/- mice did not relax to ACh or A-23187 compared with eNOS +/+ mice. In contrast, relaxation to nitroprusside (NP) was significantly greater in eNOS -/- mice than in eNOS +/+ mice. Gene transfer of eNOS, but not lacZ, to vascular rings of eNOS -/- mice restored relaxation to ACh and A-23187. In vessels from eNOS -/- mice that were transduced with eNOS, N(omega)-nitro-L-arginine (10(-4) M) inhibited relaxation to ACh and A-23187 but not NP. Thus vascular function can be significantly improved by gene transfer in vessels where a major relaxation mechanism is genetically absent.

Impaired anticoagulant response to infusion of thrombin in atherosclerotic monkeys associated with acquired defects in the protein C system

To examine the effects of atherosclerosis on the protein C anticoagulant pathway in vivo, we measured anticoagulant responses to intravenous administration of human alpha-thrombin or activated protein C (APC) in cynomolgus monkeys. Two groups of monkeys were fed either a control diet (n=18) or an atherogenic diet (n=12) that produces both hypercholesterolemia and moderate hyperhomocyst(e)inemia. A third group (n=8) was fed an atherogenic diet for 15 months, and then fed the atherogenic diet supplemented with B vitamins for 6 months to correct the hyperhomocyst(e)inemia. The plasma homocyst(e)ine level was higher in monkeys fed the atherogenic diet (9.6+/-1.0 micromol/L) than in monkeys fed the control diet (3.7+/-0.2 micromol/L) or the atherogenic diet with B vitamins (3.6+/-0.2 micromol/L) (P<0.001). Infusion of thrombin produced a much greater prolongation of the activated partial thromboplastin time in monkeys fed the control diet (52+/-10 seconds) than in monkeys fed the atherogenic diet either with (24+/-4 seconds) or without (27+/-5 seconds) supplemental B vitamins (P<0.02). Thrombin-dependent generation of circulating APC was higher in control (294+/-17 U/mL) than in atherosclerotic (240+/-14 U/mL) monkeys (P<0.05), although levels of fibrinogen, plasminogen, D-dimer, and thrombin-antithrombin complexes were similar in each group. Injection of human APC produced a similar prolongation of the activated partial thromboplastin time in control (31+/-3 seconds) and atherosclerotic (29+/-2 seconds) monkeys. These findings provide evidence for impaired anticoagulation, due partly to decreased formation of APC, in atherosclerosis. The blunted anticoagulant response to thrombin in hypercholesterolemic monkeys was not corrected by supplementation with B vitamins.

Gene transfer of endothelial nitric oxide synthase to the lung of the mouse in vivo. Effect on agonist-induced and flow-mediated vascular responses

The effects of transfer of the endothelial nitric oxide synthase (eNOS) gene to the lung were studied in mice. After intratracheal administration of AdCMVbetagal, expression of the beta-galactosidase reporter gene was detected in pulmonary airway cells, in alveolar cells, and in small pulmonary arteries. Gene expression with AdCMVbetagal peaked 1 day after administration and decayed over a 7- to 14-day period, whereas gene expression after AdRSVbetagal transfection peaked on day 5 and was sustained over a 21- to 28-day period. One day after administration of AdCMVeNOS, eNOS protein levels were increased, and there was a small reduction in mean pulmonary arterial pressure and pulmonary vascular resistance. The pressure-flow relationship in the pulmonary vascular bed was shifted to the right in animals transfected with eNOS, and pulmonary vasodepressor responses to bradykinin and the type V cGMP-selective phosphodiesterase inhibitor zaprinast were enhanced, whereas systemic responses were not altered. Pulmonary vasopressor responses to endothelin-1 (ET-1), angiotensin II, and ventilatory hypoxia were reduced significantly in animals transfected with the eNOS gene, whereas pressor responses to norepinephrine and U46619 were not changed. Systemic pressor responses to ET-1 and angiotensin II were similar in eNOS-transfected mice and in control mice. Intratracheal administration of AdRSVeNOS attenuated the increase in pulmonary arterial pressure in mice exposed to the fibrogenic anticancer agent bleomycin. These data suggest that transfer of the eNOS gene in vivo can selectively reduce pulmonary vascular resistance and pulmonary pressor responses to ET-1, angiotensin II, and hypoxia; enhance pulmonary depressor responses; and attenuate pulmonary hypertension induced by bleomycin. Moreover, these data suggest that in vivo gene transfer may be a useful therapeutic intervention for the treatment of pulmonary hypertensive disorders.

Adenovirus-mediated gene transfer is augmented in basilar and carotid arteries of heritable hyperlipidemic rabbits

BACKGROUND AND PURPOSE

There are major differences in susceptibility of intracranial and extracranial arteries to atherosclerosis. The goal of this study was to examine adenovirus-mediated gene transfer to basilar and carotid arteries of Watanabe heritable hyperlipidemic (WHHL) rabbits, which have spontaneous hypercholesterolemia and atherosclerosis, and normal New Zealand White (NZW) rabbits. We used 2 different adenoviral vectors, driven by either cytomegalovirus (CMV) or Rous sarcoma virus (RSV) promoters.

METHODS

Basilar and carotid arteries were removed from WHHL and NZW rabbits and cut into rings. The arteries were incubated with an adenoviral vector that expresses beta-galactosidase and is driven by either a cytomegalovirus (CMV) or Rous sarcoma virus (RSV) promoter (AdCMVbetagal or AdRSVbetagal). Arteries were incubated with virus for 2 hours, and then incubated in medium for 24 hours to allow expression of transgene. Transgene expression was assessed by enzyme activity (Galacto-Light assay) and by a histochemical method after X-Gal staining.

RESULTS

After gene transfer, beta-galactosidase was expressed in endothelium and adventitia but not media. There were moderately severe atherosclerotic lesions in carotid arteries and early lesions in basilar arteries. Enzyme activity after gene transfer with AdCMVbetagal (3x10(11) particles/mL) was greater in the basilar artery of WHHL than NZW (137+/-40 versus 25+/-10 mU/mg protein, P<0.05) (mean+/-SE) and in the carotid artery (133+/-27 versus 34+/-11 mU/mg protein, P<0.05). After gene transfer with AdRSVbetagal, transgene expression was similar in arteries from WHHL and normal NZW rabbits.

CONCLUSIONS

This is the first study to examine gene transfer to intracranial and extracranial arteries from atherosclerotic animals. The findings suggest that an adenoviral vector with a CMV, but not RSV, promoter provides greater transgene expression in the basilar and carotid arteries from spontaneously atherosclerotic rabbits than from normal rabbits.

Potassium channels mediate dilatation of cerebral arterioles in response to arachidonate

We tested the hypothesis that cerebral vasodilatation in response to arachidonate is dependent on activation of cyclooxygenase and cytochrome P-450 pathways and formation of endogenous reactive oxygen species and is mediated by activation of potassium channels. The diameter of cerebral arterioles was measured using cranial windows in anesthetized rats. Under control conditions [baseline diameter = 45 +/- 1 micrometer (mean +/- SE)], arachidonate (1-100 microM) and papaverine (10-50 microM) produced concentration-dependent vasodilatation. Cerebral vasodilator responses to arachidonate, but not papaverine, were abolished during topical application of indomethacin (10 microM, an inhibitor of cyclooxygenase) or catalase (100 U/ml, which inactivates hydrogen peroxide). In contrast, clotrimazole (10 microM) and 17-ODYA (20 microM), inhibitors of cytochrome P-450 activity, had no effect on dilator responses of cerebral arterioles to arachidonate. Superoxide dismutase (SOD, 100 U/ml) had no effect on vasodilator responses to papaverine or lower concentrations of arachidonate, whereas dilator responses to 100 microM arachidonate were inhibited modestly (by 22%) by SOD. Similarly, deferoxamine (1 mM) partly inhibited dilator responses to 10 and 100 microM arachidonate (by approximately 30% at each concentration). Tetraethylammonium ion (1 mM) or iberiotoxin (50 nM), inhibitors of calcium-activated potassium channels, markedly inhibited vasodilatation in response to arachidonate (by 70-90%) but not papaverine. These findings suggest that dilatation of cerebral arterioles in response to arachidonate is mediated largely by endogenously formed reactive oxygen species, which are generated from cyclooxygenase activity, and activation of calcium-activated potassium channels. Thus activation of potassium channels appears to be a major mechanism of cerebral vasodilatation in response to reactive oxygen species produced endogenously.

Improvement of relaxation in an atherosclerotic artery by gene transfer of endothelial nitric oxide synthase

Gene transfer with replication-deficient adenovirus is a useful tool to study vascular biology. We have reported that overexpression of endothelial nitric oxide (NO) in carotid arteries from normal rabbits augments vasorelaxation mediated by NO. In this study, we tested the hypothesis that adenovirus-mediated gene transfer of endothelial nitric oxide synthase (eNOS) improves impaired relaxation of atherosclerotic vessels. We used 2 replication-deficient adenoviruses: AdeNOS, which carries cDNA for eNOS, and Adbetagal, which expresses beta-galactosidase. Common carotid arteries from 10 New Zealand White (NZW; plasma cholesterol, 79+/-13 mg/dL) and 10 Watanabe heritable hyperlipidemic (WHHL; plasma cholesterol, 452+/-39 mg/dL) rabbits were incubated in organ culture with AdeNOS, Adbetagal, or vehicle alone. Carotid arteries from WHHL rabbits had mild to moderate atherosclerotic lesions. Histochemical staining for beta-galactosidase and immunohistochemistry for eNOS indicated transgene expression in the endothelium and adventitia in both NZW and WHHL rabbits. Expression of eNOS determined with Western blot analysis after incubation with AdeNOS tended to be higher in vessels from WHHL rabbits than NZW rabbits. Effects of transgene expression on vascular function were examined by recording isometric tension 1 day after transduction. After precontraction with phenylephrine, acetylcholine produced significantly less relaxation in vessels from WHHL rabbits than in vessels from NZW rabbits. Relaxation in response to acetylcholine was greater in carotid arteries from both NZW and WHHL rabbits that were transfected with AdeNOS than in vessels treated with vehicle or Adbetagal. Vasorelaxation in response to acetylcholine was inhibited by Nomega-nitro-L-arginine. Responses to sodium nitroprusside were similar after treatment with vehicle alone, Adbetagal, or AdeNOS in both groups of rabbits. Thus, overexpression of eNOS with an adenoviral vector improves impaired NO-mediated relaxation in atherosclerotic arteries.

Cationic polymer and lipids enhance adenovirus-mediated gene transfer to rabbit carotid artery

BACKGROUND AND PURPOSE

Improvement of efficiency of gene transfer to endothelium could be useful for several applications. We tested the hypothesis that cationic nonviral molecules augment adenovirus-mediated gene transfer to blood vessels, perhaps by alteration of the surface charge of adenovirus and facilitation of binding to endothelium.

METHODS

Carotid arteries from rabbits were incubated in vitro for 0.5 to 2 hours with an adenoviral vector alone or noncovalent complexes of adenovirus with poly-L-lysine (a cationic polymer) or lipofectin (a cationic lipid). Binding of adenovirus to the vessels was evaluated immediately after incubation with virus, and assay of transgene (ss-galactosidase) activity and histochemistry were performed 24 hours after gene transfer. To determine whether cationic molecules can be used to augment alteration of vascular function by adenovirus-mediated gene transfer, we also examined effects on gene transfer of endothelial nitric oxide synthase.

RESULTS

Assay of ss-galactosidase activity indicated that both cationic molecules increased transgene expression in vessels by approximately 5- to 6-fold. In contrast, when endothelium was removed from the vessels after gene transfer, poly-L-lysine and lipofectin did not significantly increase transgene activity. Histochemistry for ss-galactosidase also suggested that the adenovirus-cationic molecule complexes augmented transgene expression mainly in the endothelium. In addition, we found that complexing adenovirus with cationic molecules increased binding of adenovirus to the vessels. After gene transfer with recombinant adenovirus containing endothelial nitric oxide synthase, calcium ionophore (A23187) produced greater relaxation of vessels treated with adenovirus complexed with poly-L-lysine or lipofectin than those treated with adenovirus alone.

CONCLUSIONS

Cationic molecules improve the efficiency of adenovirus-mediated gene transfer to blood vessels.

Evidence that macrophages in atherosclerotic lesions contain angiotensin II

BACKGROUND

We have reported that human mononuclear leukocytes contain large amounts of angiotensin II (Ang II). The goal of the present study was to test the hypothesis that Ang II is present in monocyte/macrophages in atherosclerotic lesions.

METHODS AND RESULTS

Segments of thoracic aorta and left circumflex coronary artery were obtained from 3 groups of cynomolgus monkeys: normal, atherosclerotic, and regression. Samples of human coronary arterial atherosclerotic lesions were obtained from directional atherectomy. Sections were stained for Ang II with 3 different polyclonal rabbit anti-human Ang II antisera. In aorta and coronary arteries from normal monkeys, there was no or minimal anti-Ang II staining in endothelial cells. All sections from atherosclerotic monkeys displayed discrete, localized regions of staining for Ang II in intima-media. Macrophages were present throughout the atherosclerotic intima-media, and anti-Ang II staining appeared to colocalize with macrophages. All human coronary atherectomy samples stained positive for Ang II and macrophages. Staining for both Ang II and macrophages was observed in vascular lesions from all 5 monkeys after regression of atherosclerosis, but staining was less extensive than in atherosclerotic blood vessels from monkeys.

CONCLUSIONS

These findings suggest that Ang II is present in atherosclerotic lesions in monkeys and humans, colocalizes with macrophages in intima-media of atherosclerotic vessels from monkeys, and decreases in lesions in monkeys with regression of atherosclerosis.

Vascular effects of LPS in mice deficient in expression of the gene for inducible nitric oxide synthase

The inducible isoform of nitric oxide synthase (iNOS) is expressed after systemic administration of lipopolysaccharide (LPS). The importance of expression of iNOS in blood vessels is poorly defined. Because nitric oxide from iNOS may alter vasomotor function, we examined effects of LPS on vasomotor function in carotid arteries from iNOS-deficient mice. We studied contraction of the carotid artery from wild-type and iNOS-deficient mice in vitro 12 h after injection of LPS (20 mg/kg ip). Contractile responses to PGF2alpha (3-30 microM) and thromboxane A2 analog (U-46619; 3-100 nM) were evaluated using vascular rings from mice treated with vehicle or LPS. Maximum force of contraction generated by rings in response to PGF2alpha was 0.39 +/- 0.02 and 0.25 +/- 0.01 (SE) g (n = 14) in vehicle and LPS-treated wild-type mice, respectively (P < 0.001 vs. vehicle). Thus LPS reduced constrictor responses in wild-type mice. Thiocitrulline and aminoguanidine (inhibitors of iNOS) improved contractile responses from LPS-treated wild-type vessels. Indomethacin also improved constrictor responses in arteries from wild-type mice injected with LPS. In contrast, contraction of the carotid arteries in response to PGF2alpha and U-46619 was not impaired in LPS-treated iNOS-deficient mice, and contraction was not altered by inhibitors of iNOS. Expression of iNOS mRNA was confirmed using RT-PCR in carotid arteries from wild-type mice after injection of LPS but not vehicle. PCR products for iNOS were not observed in iNOS-deficient mice. These findings provide the first direct evidence that iNOS mediates impairment of vascular contraction after treatment with LPS.

Adenovirus-mediated gene transfer in vivo to cerebral blood vessels and perivascular tissue in mice

BACKGROUND AND PURPOSE

Gene transfer to cerebral blood vessels has been accomplished in rats and dogs by injection of replication-deficient adenovirus into cerebrospinal fluid. In this study we examined transgene expression after injection of adenovirus into the cerebrospinal fluid of mice. Responses were observed in ICR mice and C57BL/6 mice, which are outbred and inbred strains, respectively.

METHODS

We injected replication-deficient recombinant adenovirus expressing nuclear targeted beta-galactosidase, driven by either the Rous sarcoma virus promoter (AdRSV-betaGal) or the cytomegalovirus promoter (AdCMV-betaGal), into the cisterna magna of anesthetized ICR and C57BL/6 strains of mice. The brains were examined from 1 to 21 days after injection by chemiluminescent enzyme activity assay or histochemical staining.

RESULTS

After injection of AdRSV-betaGal, expression of beta-galactosidase in ICR mice peaked on day 7 and returned to basal by day 14. Expression of beta-galactosidase in C57BL/6 mice was maximal on days 7 to 14 and was minimal by day 21 after injection of AdRSV-betaGal. After injection of AdCMV-betaGal in C57BL/6 mice, peak expression of transgene occurred on day 1 and was greatly diminished by day 3. Transgene expression was observed primarily on the ventral surface of the brain, with preferential expression in leptomeninges and adventitia along the major cerebral arteries of that region.

CONCLUSIONS

Injection of recombinant adenovirus in the cisterna magna resulted in transgene expression in leptomeninges and perivascular tissue of cerebral blood vessels in two strains of mice. The CMV promoter elicited rapid but short-lived expression of the transgene, while the RSV promoter elicited slower, more sustained transgene expression. Expression of AdRSV transgene was prolonged in C57BL/6 mice compared with ICR mice. This approach for gene transfer may be useful to study cerebral vascular biology in genetically altered strains of mice.

Overexpression of human superoxide dismutase inhibits oxidation of low-density lipoprotein by endothelial cells

Oxidation of LDL in the subendothelial space has been proposed to play a key role in atherosclerosis. Endothelial cells produce superoxide anions (O2.-) and oxidize LDL in vitro; however, the role of O2.- in endothelial cell-induced LDL oxidation is unclear. Incubation of human LDL (200 microg/mL) with bovine aortic endothelial cells (BAECs) for 18 hours resulted in a 4-fold increase in LDL oxidation compared with cell-free incubation (22.5+/-1.1 versus 6.3+/-0.2 [mean+/-SEM] nmol malondialdehyde/mg LDL protein, respectively; P<0.05). Under similar conditions, incubation of LDL with porcine aortic endothelial cells resulted in a 5-fold increase in LDL oxidation. Inclusion of exogenous copper/zinc superoxide dismutase (Cu/ZnSOD, 100 microg/mL) in the medium reduced BAEC-induced LDL oxidation by 79%. To determine whether the intracellular SOD content can have a similar protective effect, BAECs were infected with adenoviral vectors containing cDNA for human Cu/ZnSOD (AdCu/ZnSOD) or manganese SOD (AdMnSOD). Adenoviral infection increased the content and activity of either Cu/ZnSOD or MnSOD in the cells and reduced cellular O2.- release by two thirds. When cells infected with AdCu/ZnSOD or AdMnSOD were incubated with LDL, formation of malondialdehyde was decreased by 77% and 32%, respectively. Two other indices of LDL oxidation, formation of conjugated dienes and increased LDL electrophoretic mobility, were similarly reduced by SOD transduction. These data suggest that production of O2.- contributes to endothelial cell-induced oxidation of LDL in vitro. Furthermore, adenovirus-mediated transfer of cDNA for human SOD, particularly Cu/ZnSOD, effectively reduces oxidation of LDL by endothelial cells.

Superoxide production in vascular smooth muscle contributes to oxidative stress and impaired relaxation in atherosclerosis

The endothelium is a source of reactive oxygen species in short-term models of hypercholesterolemia and atherosclerosis. We examined a chronic model of atherosclerosis for increased vascular production of superoxide (O2-.) and determined whether endothelial overexpression of superoxide dismutase (SOD) would improve endothelium-dependent relaxation. Superoxide generation was 3 times higher in isolated aortas from Watanabe heritable hyperlipidemic (WHHL) rabbits (2 to 4 years old) compared with aortas from New Zealand White (NZ) rabbits (43+/-10 versus 14+/-2 relative light units x min(-1) x mm(-2), n=9, P<0.05). After in vitro transduction with adenovirus containing the gene for CuZn-SOD (AdCMVCuZn-SOD) or extracellular SOD (AdCMVEC-SOD), endothelial O2-. levels in WHHL aortas were significantly reduced. Gene transfer of SOD to WHHL aortas, however, failed to improve the impaired relaxation to acetylcholine or calcium ionophore. By use of the oxidative fluorescent dye hydroethidine, an in situ assay indicated markedly increased generation of O2-. throughout the wall of WHHL aorta, especially within layers of smooth muscle. This finding was confirmed by demonstrating increased O2-. levels in smooth muscle cells cultured from WHHL aorta. We conclude that elevated O2-. levels in atherosclerotic vessels are not confined to the endothelium but occur throughout the vascular wall, including smooth muscle cells. Reduction in endothelial O2-. levels is not sufficient to improve endothelium-dependent relaxation. Generation of reactive oxygen species within the media may contribute to vasomotor dysfunction in atherosclerosis.

Increase in TUNEL positive cells in aorta from diabetic rats

TUNEL staining, which allows detection of fragmented DNA in situ, is commonly used as an indication of apoptosis. Recent studies suggest that apoptosis is increased in several pathophysiological conditions. In this study we examined the hypothesis that chronic diabetes is associated with an increase in TUNEL staining of the aorta. Diabetic rats were studied 4-5 months after injection of streptozotocin (50 mg/kg). Aorta of diabetic and control rats were examined for TUNEL staining, morphology by electron microscopy, and DNA contamination in RNA preparation by polymerase chain reaction (PCR). TUNEL staining of aortic sections showed a 6 fold increase of positive cells in the media of diabetic aorta (22 +/- 6%) (mean +/- SE) compared with aorta from age-matched controls (3.6 +/- 0.9%, p < 0.05). Electron microscopy demonstrated typical apoptotic cells and bodies in the media of aorta from diabetic but not control rats. DNA contamination was found in RNA prepared from diabetic aorta, which was detected using PCR, which is consistent with increased DNA fragmentation. Increased TUNEL staining was not observed in rats with hyperglycemia 3 days after injection of streptozotocin. In conclusion, severe chronic diabetes is associated with an increase in TUNEL staining, and perhaps apoptosis, in the aorta. We speculate that increased apoptosis may compensate for increased proliferative activity in diabetic blood vessels.

Responses of carotid artery in mice deficient in expression of the gene for endothelial NO synthase

We examined the hypotheses that responses to acetylcholine are impaired and responses to NO are enhanced in carotid artery from mice made deficient in endothelial nitric oxide synthase (eNOS) by gene targeting (eNOS-deficient mice). We also tested the hypothesis that deletion of one copy of the eNOS gene is sufficient to alter vascular responses. Vessels were studied in vitro from heterozygous (+/-) and homozygous (-/-) eNOS-deficient mice as well as wild-type [eNOS(+/+)] littermates. After precontraction with prostaglandin F2 alpha, acetylcholine produced marked relaxation of carotid arteries in eNOS(+/+) mice, with impaired vasorelaxation in eNOS(+/-) mice. For example, 1 microM acetylcholine relaxed carotid arteries by 55 +/- 5% (mean +/- SE) in eNOS(+/-) mice (n = 13) compared with 83 +/- 3% in eNOS(+/+) mice (n = 14, P < 0.001 vs. +/-). In contrast, acetylcholine caused no relaxation in carotid arteries from eNOS(-/-) mice (P < 0.001 vs. +/+ and +/-). Relaxation of the carotid artery in response to nitroprusside [a nitric oxide (NO) donor] was enhanced (P < 0.001) in eNOS-deficient mice. For example, in response to 10 nM nitroprusside, the carotid artery relaxed by 18 +/- 2% in eNOS(+/+) mice (n = 14), 33 +/- 2% in eNOS(+/-) mice (n = 13), and 47 +/- 4% in eNOS(-/-) mice (n = 5). Thus relaxation of the carotid artery is impaired with acetylcholine and enhanced with the NO donor nitroprusside in eNOS-deficient mice. Enhanced responses to NO may represent a compensatory response expressed in the absence of eNOS. The findings that vascular responses to acetylcholine and NO are altered in eNOS(+/-) mice compared with those observed in eNOS(+/+) mice suggest a "gene-dosing" effect.

Regulation of the cerebral circulation: role of endothelium and potassium channels

Several new concepts have emerged in relation to mechanisms that contribute to regulation of the cerebral circulation. This review focuses on some physiological mechanisms of cerebral vasodilatation and alteration of these mechanisms by disease states. One mechanism involves release of vasoactive factors by the endothelium that affect underlying vascular muscle. These factors include endothelium-derived relaxing factor (nitric oxide), prostacyclin, and endothelium-derived hyperpolarizing factor(s). The normal vasodilator influence of endothelium is impaired by some disease states. Under pathophysiological conditions, endothelium may produce potent contracting factors such as endothelin. Another major mechanism of regulation of cerebral vascular tone relates to potassium channels. Activation of potassium channels appears to mediate relaxation of cerebral vessels to diverse stimuli including receptor-mediated agonists, intracellular second messenger, and hypoxia. Endothelial- and potassium channel-based mechanisms are related because several endothelium-derived factors produce relaxation by activation of potassium channels. The influence of potassium channels may be altered by disease states including chronic hypertension, subarachnoid hemorrhage, and diabetes.

Mechanisms of bradykinin-induced cerebral vasodilatation in rats. Evidence that reactive oxygen species activate K+ channels

BACKGROUND AND PURPOSE

Relatively little is know regarding mechanisms by which reactive oxygen species produce dilatation of cerebral arterioles. The goal of this study was to test the hypothesis that vasodilator responses of cerebral arterioles to bradykinin, which produces endogenous generation of reactive oxygen species, involve activation of calcium-dependent potassium channels.

METHODS

We used a cranial window in anesthetized rats to examine effects of catalase (which degrades hydrogen peroxide) on responses to bradykinin. In addition, we examined effects of tetraethylammonium (TEA) and iberiotoxin, inhibitors of calcium-dependent potassium channels, on responses of cerebral arterioles to hydrogen peroxide, bradykinin, and papaverine.

RESULTS

In cerebral arterioles (baseline diameter = 40 +/- 1 microns) (mean +/- SE), hydrogen peroxide (10 and 100 mumol/L) produced concentration-dependent dilatation. TEA (1 mmol/L), an inhibitor of calcium-dependent potassium channels, produced marked inhibition of vasodilatation in response to hydrogen peroxide. For example, 100 mumol/L hydrogen peroxide dilated arterioles by 13 +/- 2% in the absence and 4 +/- 1% (P < .05 versus control) in the presence of TEA. Bradykinin (10 nmol/L to 1 mumol/L) also produced concentration-dependent dilatation of cerebral arterioles that was inhibited completely by catalase (100 U/mL). TEA or iberiotoxin markedly inhibited vasodilatation in response to bradykinin. For example, 100 nmol/L bradykinin dilated arterioles by 21 +/- 3% in the absence and 2 +/- 2% (P < .05 vs control) in the presence of iberiotoxin (50 nmol/L).

CONCLUSIONS

These findings suggest that dilatation of cerebral arterioles in the rat in response to hydrogen peroxide, or hydrogen peroxide produced endogenously in response to bradykinin, is mediated by activation of calcium-dependent potassium channels. Thus, activation of potassium channels may be a major mechanism of dilatation in response to reactive oxygen species in the cerebral microcirculation.