Impaired Vasorelaxation in Inbred Mice Is Associated with Alterations in Both Nitric Oxide and Super Oxide PathwaysChen C. · Korshunov V.A. · Massett M.P. · Yan C. · Berk B.C.
Cardiovascular Research Institute, Department of Medicine, University of Rochester, Rochester, N.Y., USA Corresponding Author
Recently, we showed that genetic factors determine flow-dependent vascular remodeling. Among five inbred mouse strains, the SJL strain developed the largest intima in response to low flow. Because SJL mice have a spontaneous mutation in superoxide dismutase 2 (SOD-2) we tested the hypothesis that strain-specific variations in vascular function are due to alterations in redox and nitric oxide (NO) pathways. Vasorelaxation to acetylcholine was significantly impaired in aortic rings from SJL compared to C3H or FVB mice (up to 40%). Relaxation to the endothelium-independent vasodilator sodium nitroprusside (SNP) in SJL mice was also significantly impaired at low concentrations, with decreases in sensitivity and maximal relaxation to SNP compared to C3H and FVB mice. Western blot analyses showed significantly decreased expression (∼40%) of eNOS, PKG and SOD-2 proteins in SJL vasculature compared to C3H. Intact aortas from SJL showed significantly increased nitrotyrosine and decreased SOD-2 expression compared to C3H by immunohistochemistry. Basal levels of superoxide in aortas from SJL were not significantly different than C3H as measured by dihydroethidine. In summary, relatively small alterations in redox (SOD-2) and NO pathways (eNOS and PKG) may contribute to significantly impaired vasorelaxation in SJL mice.
Copyright © 2007 S. Karger AG, Basel
Carotid intima-media thickening (IMT) predicts heart disease and stroke  and has a strong genetic component in humans . Mechanisms responsible for IMT are largely unknown but recent observations in patients show a key role for local hemodynamics (e.g. shear stress) in IMT progression [3, 4]. We have developed a mouse model of carotid IMT induced by low flow , and found dramatic strain-dependent differences in vascular remodeling among five inbred strains of mice . SJL mice exhibited the greatest intima formation due to increased inflammatory cell retention/proliferation and upregulation of proinflammatory cytokines (e.g. interleukin 18 and macrophage migration inhibitory factor) compared to C3H mice .
The vascular endothelium senses hemodynamic forces (shear stress) and transduces these signals into biochemical events that regulate vascular tone and structure . In fact endothelium is required for vascular remodeling, since endothelial denudation or blockade of endothelial nitric oxide (NO) inhibits the ability of arteries to remodel [9, 10]. Targeted deletion of the endothelial NO synthase (eNOS) gene in mice results in hypertension and impaired vascular remodeling [11, 12]. In contrast, augmenting NO by local gene delivery of eNOS improves endothelial function, and limits intimal proliferation .
Conversely, oxidative stress contributes significantly to cardiovascular pathophysiology by reducing bioavailability of NO . The major regulators of NO bioavailability in the vessel wall are the regulators of ROS including both NADPH oxidase and superoxide dismutase (SOD) . The reaction of superoxide with NO reduces the bioavailability of NO as a vasodilator by generating peroxynitrite (a product of NO + O2–), which itself may contribute adversely to vascular function . Decreased NO bioavailability is also associated with decreased protein kinase G (PKG) activity that may have long-term effects on vascular gene expression and structure . Importantly, SJL mice are known to have a 50% reduction in SOD-2 activity due to a valine to methionine substitution (Val138Met) when compared to BALB/c mice . Recent data suggested that SOD-2 protects against oxidative stress and endothelial dysfunction in the carotid arteries of ApoE–/– mice .
In the present study we tested the hypothesis that alterations in vascular reactivity among different inbred mouse strains (C3H, FVB and SJL) are due to decreased NO bioavailability caused by changes in nitrosative and oxidative stress. Here we show specific alterations in the NO pathway and SOD-2.
Male and female C3HeB/FeJ (C3H), FVB/NJ (FVB) and SJL/J (SJL) mice (16–18 weeks, 20–28 g, Jackson Laboratories, Bar Harbor, Me., USA) were used in accordance with the guidelines of the National Institutes of Health and American Heart Association for the care and use of laboratory animals. All experimental protocols were approved by the University of Rochester Animal Care and Use Committee.
The mouse vascular ring preparation was performed as described . Briefly, mice were anesthetized with an intraperitoneal injection of a mixture of ketamine (130 mg/kg) and xylazine (8.8 mg/kg). Subsequently, the thoracic aortas were removed and placed in Krebs buffer pH 7.4 (in mmol/l: 118.3 NaCl, 4.7 KCl, 2.5 CaCl2, 1.2 MgSO4, 1.2 KH2PO4, 25 NaHCO3 and 5.5 glucose) saturated with 95% O2 and 5% CO2. Vascular segments were dissected free of loose connective and adipose tissue and cut into rings of equal length (2 mm). The segments were suspended in a 10-ml organ bath (Multi Myograph System, Arhus, Denmark) maintained at 37°C and connected to a force transducer for measurement of isometric tension. In preliminary length-tension experiments, maximal responses to KCl were obtained at a resting tension of 12 mN. Therefore, resting tension was adjusted stepwise to reach a final resting tension of 12 mN in each experiment. Vessels were allowed to equilibrate for 45 min. Dose-response curves for KCl (5–100 mmol/l) and phenylephrine (Phe: 10–9 to 10–5 mol/l) were generated to determine the contractile responses. The concentration-response curves (10–9 to 10–5 mol/l) to acetylcholine (Ach) and sodium nitroprusside (SNP) were generated after the aortic rings were preconstricted to 70% of maximum with Phe. Doses were added after the response curve reached a plateau from the previous dose. Several previous publications show that vasorelaxation to Ach in mouse aortas is due to NO released via eNOS [21, 22]. We did not find differences in vascular reactivity that were gender dependent (data not shown).
In separate experiments, thoracic aortas were removed from mice and Western blots were carried out with tissue lysis as described previously . Monoclonal antibodies were used against eNOS (1:1,000, BD Transduction Laboratories, Franklin Lakes, N.J., USA), PECAM-1 and α-actin (1:1,000, Santa Cruz Biotechnology, Santa Cruz, Calif., USA), PKG (1:1,000, StressGen Biotechnologies, Ann Arbor, Mich., USA) and SOD-2 (1:5,000, StressGen Biotechnologies). Membranes were scanned by Odyssey infrared imaging system (Li-Cor Biosciences, Lincoln, Nebr., USA). Data were normalized to α-actin or PECAM-1. All data were analyzed by normalizing the ratio to 1.0 for the C3H strain.
Samples were obtained from mice utilized in a recently published study . Briefly, mice were perfusion fixed, thoracic aortas were harvested and embedded in paraffin, and cross-sections (4 μm) were made. Vessels were assessed by immunostaining for the following markers: nitrotyrosine (1:100; Upstate Biotechnology) or SOD-2 (1:400; StressGen) with hematoxylin counterstain. A high temperature (120°C) under pressure was used for antigen retrieval with 10 mmol/l citrate buffer (pH 6.0) for 20 min. Primary antibodies were incubated at 4°C overnight, followed by incubation with the secondary antibody for 30 min and streptavidin peroxidase complex for 30 min. The peroxidase-binding sites were verified by 3,3′-diaminobenzidine in chromogen solution (DakoCytomation, Glostrup, Denmark).
Superoxide was measured in situ as recently described . Briefly, in a separate experiment frozen sections of the thoracic aorta (30 μm thick) were incubated at room temperature for 30 min with dihydroethidine (DHE; 2 μmol/l) and protected from light. Relative intensity of fluorescence was determined using ImageJ software. Data were normalized to the cross-sectional area for each aorta. In addition, DHE staining was assayed after incubation with Tiron (superoxide scavenger; 10 mmol/l, 30 min at 37°C). Tiron caused a similar reduction (20–25%) in DHE staining in the media of SJL and C3H mice (data not shown), suggesting that DHE staining is measuring ROS generation.
All data are reported as means ± SEM. The ED50 or IC50 values were calculated by using a computer algorithm that assumes a linear curve between two points around 50% of the maximum response. One-way ANOVA with repeated measurements was used to test for statistically significant differences in the concentration-response curves from different mice strains using StatView software. Data from Western blots were analyzed using non-parametric Kruskal-Wallis test (rank sums). Statistical significance was accepted at p < 0.05.
Dose-response curves to KCl and Phe were generated to determine the contractile responses (fig. 1a, b). There were no significant differences in the maximal response to KCl among mouse strains (table 1). However, aortic rings from FVB exhibited a greater sensitivity to KCl than C3H and SJL with a significantly lower ED50 (fig. 1a, b; table 1). All three strains responded similarly to Phe as measured by EC50, although the maximal responses to Phe were significantly increased in FVB (fig. 1b; table 1). In summary, aortic rings from FVB were more sensitive to KCl and exhibited increased maximal vasoconstriction to Phe compared to C3H and SJL.
|Table 1. ED50 and IC50 and maximal responses of inbred mice to vasoactive agents|
|Fig. 1. Differences in concentration-response curves to vasoactive agents among inbred mouse strains (n = 7 mice per group, respectively). a KCl. b Phenylephrine (Phe). c Ach. d SNP. For vasorelaxation experiments, aortic rings were preconstricted with Phe (70% maximal vasoconstriction). Cumulative concentrations of each drug, log [mol/l], are shown on the x-axis. Values are means ± SEM. * p < 0.05 vs. SJL (ANOVA).|
To evaluate vasorelaxation for all three strains, concentration-response curves to Ach and SNP were generated (fig. 1c, d). Vasodilator responses to Ach differed among strains in the following order: C3H > FVB> > SJL (fig. 1c). Relaxation to Ach in SJL mice was significantly impaired at higher concentrations (10–7 to 10–5 mol/l) compared with FVB and C3H (fig. 1c). The maximal response to Ach was diminished in SJL by 40% compared to C3H (table 1). However, sensitivity to Ach was similar among these strains (table 1). The concentration-response curve to SNP, an endothelium-independent vasodilator, was also impaired in SJL compared to C3H and FVB (fig. 1d). Relaxation to SNP in SJL mice was significantly impaired at low concentrations (10–9 to 3 × 10–7 mol/l) compared with FVB and C3H (fig. 1d). The maximal relaxation to SNP was significantly diminished in SJL (∼15%) compared to C3H (table 1). Also, sensitivity to SNP was significantly decreased in SJL compared to C3H and FVB (table 1). Thus, endothelium-dependent and endothelium-independent vasorelaxation were significantly impaired in aortic rings from SJL compared to C3H or FVB mice.
A possible mechanism for the genetic differences in endothelium-dependent vasorelaxation in mice could be due to altered expression or function of eNOS. We measured eNOS expression in aortic homogenates by Western blots (fig. 2a). Western blot analyses revealed an immunoreactive band of similar molecular weight in all strains, but the intensity of this band was lower in SJL compared to C3H and FVB (fig. 2a). Relative expression of the endothelial-specific marker PECAM-1 to actin was similar among inbred strains, so we used PECAM-1 to normalize protein abundance in endothelial cells (fig. 2a, b). The relative eNOS expression was significantly decreased only in SJL (∼40%) compared to C3H, while FVB showed a slightly lower basal level of eNOS (fig. 2b).
|Fig. 2. Analyses of the protein expression in thoracic aortas from inbred mouse strains. a Western blots of eNOS (135 kDa), PECAM-1 (130 kDa), PKG (80 kDa), SOD-2 (24 kDa) and α-actin (42 kDa) in mouse aortas. b–d Normalized eNOS (b), PKG (c) and SOD-2 expression (d). An average of 3–5 independent experiments is shown. Values are means ± SEM. * p < 0.05 vs. C3H (Kruskal-Wallis Test).|
A signaling pathway that involves NO, cGMP and PKG induces relaxation in vascular smooth muscle. Because there were significant impairments in endothelium-independent vasorelaxation (assayed by SNP) in SJL compared to other strains, we measured PKG protein expression in these strains (fig. 2a). PKG expression normalized to total actin had a pattern similar to that of eNOS among inbred mouse strains: C3H ≈ FVB >> SJL (compare fig. 2c vs. b). These data showed that decreased vasorelaxation in SJL is associated with relatively lower expression of PKG in the aorta.
Previously, a 50% reduction in SOD-2 expression and activity in the liver was found in SJL mice due to a spontaneous mutation . To gain further insight into mechanisms responsible for differences in vasorelaxation among inbred mouse strains, we evaluated SOD-2 expression in the aortas. Western blot analyses revealed an immunoreactive band of similar molecular weight (25 kDa) for SOD-2 in all strains (fig. 2a). The intensity of this band was lower in SJL compared to C3H and FVB (fig. 2a). When normalized to the total actin, SOD-2 expression was significantly decreased in SJL (∼40%) compared to C3H (fig. 2d). As predicted by Western blot analyses, SOD-2 immunostaining was significantly reduced in the vessel wall from SJL compared to C3H or FVB mice (dark staining, compare Fig. 3c vs. a/b). Thus, SOD-2 expression in the vasculature was considerably lower in SJL compared to C3H or FVB mice.
|Fig. 3. Immunohistochemical analyses of the thoracic aortas from C3H (a, d), FVB (b, e) and SJL mice (c, f). a–c SOD-2. d–f Nitrotyrosine. Bracket shows area between internal and external elastic lamina. ×60. Bar = 20 μm.|
To evaluate the levels of oxidative and nitrosative stress in the aortas from inbred mouse strains we assessed nitrotyrosine immunoreactivity (fig. 3d, f). As predicted by lower expression of SOD-2 proteins (fig. 2, 3) there was a large increase in intensity of nitrotyrosine staining in the thoracic aorta from SJL compared to C3H and FVB (dark staining, compare fig. 3f vs. d/e). These data suggest increased oxidative stress in vessels from SJL compared to other mouse strains.
To assay basal levels of superoxide in the aorta we measured DHE fluorescence (fig. 4). In contrast to the lower expression of SOD-2 (fig. 2, 3), DHE fluorescence was only slightly greater in the aortic wall of SJL compared to C3H (compare fig. 4b vs. a). When DHE fluorescence was quantified in the media, it was not significantly different in SJL compared to C3H (fig. 4c). More impressive was the increased DHE fluorescence in the adventitia of SJL compared to C3H (white circles, fig. 4b vs. a).
|Fig. 4. Comparison of DHE fluorescence in the thoracic aortas between C3H and SJL mice. a C3H. b SJL. a, b Positive staining of the media of the thoracic aorta is white. White arrows point to internal and external elastic lamina to define the tunica media. White circles outline DHE-positive cells in the adventitia of the aorta from SJL mouse. ×10. c Normalized DHE fluorescence. Values are means ± SEM for 3 independent experiments.|
The major findings of the present study are that the SJL mouse strain showed impaired endothelium-dependent and endothelium-independent vasorelaxation compared to C3H and FVB mice associated with decreased expression of SOD-2, eNOS and PKG. Aortic responses to the endothelium-dependent vasodilator Ach were significantly less in SJL compared to C3H and FVB mice (fig. 1). In addition, SJL mice exhibited decreased sensitivity and decreased maximal responses to the endothelium-independent vasodilator SNP (table 1). The impaired vasorelaxation in SJL mice is likely mediated, in part, by alterations in both NO (decreased eNOS and PKG) and redox pathways (decreased SOD-2). Specifically, Western blots showed significantly lower (∼40%) eNOS, PKG and SOD-2 protein expression in SJL compared to C3H mice (fig. 2). Furthermore, immunohistochemical evaluation of intact aortas confirmed decreased SOD-2 expression and greater immunoreactivity for nitrotyrosine (fig. 3).
The NO/cGMP/PKG signaling cascade plays an essential role in vascular smooth muscle relaxation, and clinical studies indicate that decreased NO bioavailability contributes to the pathogenesis of vascular disease in humans, including atherosclerosis [24,25,26]. Here we found that responses to Ach were dramatically decreased in SJL compared to C3H and FVB (fig. 1c). These data support a previous finding of significant variation in endothelium-dependent vasorelaxation in aortic rings among seven normotensive inbred mouse strains (A/J, BALB/c, C3H, C57Bl/6, SWR, 129P3 and 129X1) . In the current study as well as the previous report , endothelial dysfunction in mice did not correlate with blood pressure: systolic blood pressure was similar in SJL (119 mm Hg) and C3H (116 mm Hg) . One plausible explanation for this is that blood pressure is determined by multiple factors including locally released endothelial factors, circulating neurohormones, and the sympathetic nervous system. Changes in any of these factors may compensate for the loss of eNOS protein expression and increased oxidative stress [11, 12]. For example, eNOS+/– and SOD-2+/– mice have systolic blood pressures that are comparable to wild-type mice [21, 27], while eNOS–/– mice are hypertensive (SOD-2–/– is embryonic lethal ), suggesting that a significant impairment in eNOS or SOD-2 is required to significantly alter systolic blood pressure. In contrast, changes in eNOS and SOD-2 in SJL mice are much more apparent in isolated vessels where there are fewer compensatory mechanisms. Therefore we focused on changes in the redox state and NO bioavailability via alterations in the NO pathway and SOD-2.
While multiple mechanisms regulate the function of eNOS including L-arginine and tetrahydrobiopterin availability, protein-protein interactions, and phosphorylation-dependent signal transduction events, reduced expression of eNOS is a key factor . For example, targeted inactivation of eNOS causes hypertension , and eNOS deficiency accelerates atherosclerosis in the ApoE–/– mouse . Indeed, we found decreased eNOS protein expression relative to the endothelial cell marker PECAM-1 in SJL (fig. 2b). Furthermore, we demonstrated that basal PKG protein expression was significantly decreased by 40% in SJL compared to C3H (fig. 2c). PKG is a major mediator of cGMP signaling in the cardiovascular system by phosphorylating contractile proteins directly and by phosphorylating transcription factors that regulate smooth muscle phenotype . Homozygous deletion of the PKG gene in mice abolishes NO-mediated vasorelaxation . The lower PKG expression may explain the attenuated relaxation in SJL mice in response to both vasodilators (Ach and SNP), since PKG is a key effector for cGMP-mediated vasorelaxation. However, we cannot rule out the possibility that alterations in PKG-dependent gene transcription may be involved in the reduced response to NO. Taken together, our data suggest that impaired vascular relaxation in SJL mice depends on the NO/cGMP/PKG pathway in the vessel wall.
A second mechanism for reduced vasorelaxation in SJL is decreased NO bioavailability, mediated in part by inactivation of NO by superoxide. A recent genetic survey of spontaneous mutations of the major antioxidant enzymes among 10 inbred strains showed that SJL mice have unique substitutions in SOD-2 leading to a 50% reduction in SOD-2 activity . We found that vascular SOD-2 protein expression was significantly decreased in SJL compared to C3H (fig. 2). However, the relative values of the superoxide (detected by DHE) in aortas from SJL mice were similar to those reported for SOD-2+/– mice , confirming that the 50% reduction in SOD activity in the SJL aorta causes only a small increase in basal superoxide generation. It is possible that during stress response (e.g. low shear stress) there may be more superoxide generated in SJL vessels. SOD exists as three isoforms: SOD-1 is copper-zinc dependent and localized to the cytosol, SOD-2 is manganese dependent and found in the mitochondria, and SOD-3 is extracellular. Since SOD-2–/– mice die due to dilated cardiomyopathy , SOD-2+/– mice are used as a model for mitochondrial dysfunction and oxidative stress. In line with our observations in the vessels from SJL mice, a recent study confirmed that SOD-2 protected against oxidative stress and endothelial dysfunction in the arteries of ApoE–/– mice . Finally, SOD-2+/– mice exhibited increased infarction after permanent cerebral ischemia  or ischemia/reperfusion in the brain  and heart . Unlike blood pressure , genetic differences in response to vasoactive substances were highly correlated with susceptibility to global cerebral ischemic injury in mice . In particular, an increased inflammatory response in SJL was a dominant mechanism for focal cerebral infarcts compared with BALB/c or C57Bl/6 mice . In addition, ascorbic acid moderately but significantly reduced a spontaneous lymphoproliferative process in SJL mice . Based on the present study and recently published data, we can speculate that the greater inflammation in response to focal cerebral ischemia  and the enhanced carotid IMT in SJL in response to low flow  are due to an imbalance in antioxidant enzymes (e.g. SOD-2).
In summary, the present study shows that the SJL inbred mouse strain is characterized by impaired vascular reactivity due, in part, to dual alterations in redox (SOD-2) and NO (eNOS and PKG) pathways. Our data suggest that relatively modest dysregulation (compared with complete gene deletion) in redox and NO pathways in SJL inbred mice may represent a useful model of vascular dysfunction in humans.
This study was supported in part by a grant from the NIH (HL-62826) to B.C.B. and in part by funds from the AHA (SDG 01300500T) to M.P.M. The authors thank Dr. Keigi Fujiwara for his help with DHE experiments and Sarah Mack for her help with mouse tissue processing.
Dr. Bradford C. Berk
University of Rochester, Box MED
601 Elmwood Avenue
Rochester, NY 14642 (USA)
Tel. +1 585 273 1946, Fax +1 585 273 1497, E-Mail Bradford_Berk@URMC.rochester.edu
Received: March 12, 2007
Accepted after revision: May 25, 2007
Published online: July 30, 2007
Number of Print Pages : 9
Number of Figures : 4, Number of Tables : 1, Number of References : 38
Journal of Vascular Research (Incorporating 'International Journal of Microcirculation')
Vol. 44, No. 6, Year 2007 (Cover Date: October 2007)
Journal Editor: Pohl, U. (Munich)
ISSN: 1018–1172 (print), 1423–0135 (Online)
For additional information: http://www.karger.com/JVR