In normotensive rats, an increase in dietary salt leads to decreased arteriolar responsiveness to acetylcholine (ACh) because of suppressed local nitric oxide (NO) activity. We evaluated the possibility that generation of reactive oxygen species in the arteriolar wall is responsible for this loss of NO activity. Arteriolar responses to iontophoretically applied ACh were examined in the superfused spinotrapezius muscle of Sprague-Dawley rats fed a low-salt (LS; 0.45%) or high-salt diet (HS; 7%) for 4–5 wk. Responses to ACh were significantly depressed in HS rats but returned to normal in the presence of the oxidant scavengers superoxide dismutase + catalase or 2,2,6,6-tetamethylpiperidine-N-oxyl (TEMPO) + catalase. Arteriolar responses to the NO donor sodium nitroprusside were similar in HS and LS rats. Arteriolar and venular wall oxidant activity, as determined by reduction of tetranitroblue tetrazolium, was significantly greater in HS rats than in LS rats. Exposure to TEMPO + catalase reduced microvascular oxidant levels to normal in HS rats. These data suggest that a high-salt diet leads to increased generation of reactive oxygen species in striated muscle microvessels, and this increased oxidative state may be responsible for decreased endothelium-dependent responses associated with high salt intake.
- nitric oxide
- free radicals
high dietary salt intake can lead to changes in microvascular structure and function that are unrelated to any changes in arterial pressure (2, 3, 9, 15,17, 23, 24). For example, M. A. Boegehold (2, 3) has previously reported that in skeletal muscle of normotensive rats fed a high-salt diet, endothelium-dependent arteriolar responses to acetylcholine (ACh) and increased shear stress are reduced because of a suppression of local nitric oxide (NO) activity. Reduced endothelium-dependent responses associated with high salt intake have more recently been documented in rat cerebral and cremaster muscle arterioles (9, 10, 24) and in small feed arteries of rat gracilis muscle (23). The reduced influence of NO on arteriolar tone in rats fed a high-salt diet is apparently not caused by a decrease in vascular smooth muscle responsiveness to NO, as judged by normal arteriolar responses to the NO donor sodium nitroprusside (SNP) in these animals (2,3, 9, 10, 23,24). It is more likely that high salt intake leads to either a decrease in endothelial release of NO or the premature inactivation of released NO.
Widespread inactivation of NO could occur with high salt intake if this diet leads to the generation of reactive oxygen species (ROS) in or near the microvascular endothelium. These highly reactive molecules can be formed in the microvasculature via the enzymatic activity of cyclooxygenase, xanthine oxidase, NAD(P)H oxidase, or NO synthase (6, 41). Superoxide anion generated by one or more of these pathways can rapidly oxidize endothelium-derived NO to interfere with its normal influence on vascular tone (14,33).
Increased ROS generation and the consequent inactivation of endothelium-derived NO has been implicated in the depressed endothelium-dependent vascular responses associated with hypertension (35, 38, 39), diabetes (5, 30), and ischemia-reperfusion (26, 27). The purpose of this study was to investigate the possibility that a high-salt diet leads to increased formation of ROS in the microvascular wall and consequently depresses the endothelium-dependent regulation of these vessels.
Surgical Preparation and Intravital Microscopy
All surgical and experimental procedures for this study were approved by the West Virginia University Animal Care and Use Committee. Weanling male Sprague-Dawley rats (Harlan Sprague Dawley, Indianapolis, IN) were placed on a whole-grain diet containing either 0.45% NaCl [low salt (LS)] or 7% NaCl [high salt (HS)] by weight (LS, TD8831; HS, TD92100; Teklad, Madison, WI). All rats were studied 4–5 wk after being placed on their respective diets.
Each rat was anesthetized with thiopental sodium (100 mg/kg ip) and placed on a heating mat to maintain a 37°C rectal temperature. The trachea was intubated to ensure a patent airway, and the right carotid artery was cannulated to measure arterial blood pressure. The spinotrapezius muscle was surgically exteriorized as previously described (3). With this approach the muscle is gently drawn away from the body wall without disturbing any feed vessels or surface fascia and is secured at its in situ length over a transparent observation pedestal. Throughout the surgical preparation and subsequent experimental period, the exteriorized muscle was continuously superfused with an electrolyte solution (119 mM NaCl, 25 mM NaHCO3, 6 mM KCl, and 3.6 mM CaCl2) warmed to 35°C and equilibrated with 95% N2–5% CO2. Superfusate flow rate was set at 4–6 ml/min to minimize equilibration with atmospheric O2(4).
The rat was then transferred to the stage of an Olympus BX50WI intravital microscope fitted with a charge-coupled device video camera (Dage-MTI, Michigan City, IN). Video images were displayed on a Panasonic high-resolution video monitor and stored on videotape for off-line analysis. Observations were made with an Olympus 20× water-immersion objective (final video magnification: 770×). Arteriolar inner diameters were measured during videotape replay with a video image-shearing monitor (IPM, San Diego, CA) or video caliper (Microcirculation Research Institute, Texas A&M University, College Station, TX).
The blood supply to the spinotrapezius muscle is carried by arteriolar branches of three small feed arteries that interconnect within the muscle to form a structure known as the “arcade bridge” (34). In the experiments described below, we studied the arcade arterioles that branch directly from the arcade bridge and interconnect to form an extensive anastomosing network.
Protocol 1: Arteriolar responses to ACh and SNP.
To determine the effect of high salt intake on endothelium-dependent dilation, we evaluated the responsiveness of arcade arterioles to iontophoretically applied ACh (Sigma Chemical, St. Louis, MO) in each dietary group. A micropipette, beveled to an outer tip diameter of 2–3 μm, was filled with 0.025 M ACh in distilled water and placed in light contact with the arteriolar wall. A hold current of 40 nA was used to prevent passive diffusion of ACh from the micropipette tip, and net ejection currents of 5, 20, and 40 nA (delivered in random order) were used to apply ACh. For each current dose of ACh, the vessel was continuously observed during a 2-min control period, a 2-min application period, and a 2-min recovery period.
In separate groups of LS and HS rats, arteriolar smooth muscle responsiveness to NO was evaluated by iontophoretically applying the NO donor SNP (Sigma) to the arteriolar wall. Beveled micropipettes were filled with 0.05 M SNP in distilled water, and currents of 5, 20, and 40 nA were used to apply SNP in random order as described above. After either ACh or SNP was applied, passive arteriolar diameters were measured in the presence of adenosine (Sigma), which was added to the superfusate to achieve a final concentration of 10−4 M.
Protocol 2: Effect of oxygen radical scavengers on arteriolar responses to ACh.
To evaluate the possible role of ROS in the decreased endothelium-dependent dilation that we observed in rats fed high salt (see results), we repeated the arteriolar ACh applications in the presence of superoxide and hydrogen peroxide scavengers. In initial experiments, superoxide dismutase (SOD; 50 U/ml; Sigma) and catalase (Cat; 50 U/ml; Sigma) were combined in the muscle superfusate. After a 30-min equilibration period, ACh was applied to individual arterioles as described above.
In a second series of experiments, arteriolar responses to ACh were evaluated before and during exposure to 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO; 1 × 10−4 M; Sigma) in combination with Cat (50 U/ml). As with SOD/Cat, these scavengers were added to the muscle superfusate, and a 30-min equilibration period was allowed before ACh applications were repeated. TEMPO belongs to a family of a nitroxide spin labels that act as membrane-permeable SOD mimics (11, 29,31). For these experiments, we first conducted control experiments in LS rats to verify that our chosen superfusate concentrations of TEMPO and Cat were sufficient to scavenge oxygen radicals in the immediate vicinity of the microvascular wall. Micropipettes beveled to an 8-μm tip diameter were filled with hypoxanthine (HX; 1 × 10−5 M; ICN Biochemicals, Cleveland, OH) and xanthine oxidase (XO, 0.005 U/ml, ICN Biomedicals, Aurora, OH), which spontaneously produce superoxide anion (7). The pipette tip was positioned close to an arteriole, and HX/XO was applied directly to the arteriolar wall using 1-min pressure pulses (10 or 20 psi) from a Picospritzer II ejection system (General Valve, Fairfield, NJ). Arteriolar diameter responses to HX/XO were evaluated before and after 30 min of continuous exposure to the TEMPO/Cat superfusate.
Protocol 3: Measurement of microvascular oxidant activity.
Microvascular wall ROS levels were quantified in LS and HS rats using a procedure described by Swei et al. (38). The exteriorized muscle was exposed for 1 h to a superfusate containing 2% tetranitroblue tetrazolium (TNBT), a dye that forms insoluble blue-black formazan deposits when reduced by ROS. The muscle was then rinsed with normal superfusate for 10 min, fixed with 10% Formalin solution (pH 7.4; Fisher Scientific, Fair Lawn, NJ), and excised. After being stored in Formalin overnight, the muscle was dehydrated in solutions of progressively increasing alcohol content and then cleared with methyl salicylate (Fisher Scientific) so that vessels could be more easily visualized. The muscle was then transferred to a microscope slide and viewed under an Olympus BHMJ microscope with a Nikon 40× water-immersion objective (final video magnification: 2,920×).
Images of arcade arterioles and their paired arcade venules were captured, digitized, and analyzed with a MetaMorph 3.5 imaging system (West Chester, PA). Using a 1 × 5-μm photometric window, we made a series of gray-value measurements along the vessel wall and in avascular regions immediately adjacent to the wall. To assess microvascular wall formazan content (an index of oxidant activity), we used the measured gray values to calculate microvascular wall light absorption (A) as follows: A = ln (It/Io), were It is the vessel wall gray value and Io is the gray value for the adjacent avascular region. To produce positive controls for ROS, we continuously exposed muscles from additional LS rats to TNBT in the superfusate while microvascular superoxide anion was generated by intra-arterial infusion of HX (4.4 mM) and XO (0.15 U/min) for 1 h. Muscles were then prepared and analyzed as described above.
In separate groups of LS and HS rats, experiments were conducted to verify that ROS are responsible for the differences in microvascular wall light absorption that we observed between dietary groups (seeresults). The exteriorized muscle was continuously exposed to TEMPO/Cat at the superfusate concentration given earlier beginning 30 min before TNBT application and continuing throughout the 1-h TNBT exposure period. Microvascular wall images were analyzed, and light absorption was calculated as described above.
Data and Statistical Analysis
Arteriolar diameter and arterial pressure measurements were digitized, stored, and analyzed using Polyview Lab Manager (Grass Instrument Division/Astro-Med, West Warwick, RI). Data were collected at a rate of 100 samples/s during control, application, and recovery periods. Resting vascular tone (T) was calculated as T = [(D pass −D c)/D pass] × 100, whereD pass is the passive vessel diameter under adenosine and D c is the average resting diameter measured during the 2-min control period. A tone value of 0% represents a vessel that is completely passive, whereas a value of 100% represents a vessel that is constricted to the point of closure. Arteriolar responses to vasoactive agents were determined by comparing the steady-state diameter reached during the application period with the immediately preceding control diameter. All responses are expressed in terms of percent change from control.
All data are expressed as means ± SE, and statistical analysis was carried out using commercially available software (SigmaStat, Jandel Scientific). Comparisons between groups were made using two-way analysis of variance (ANOVA) in combination with Student-Newman-Keuls post hoc analysis to isolate pairwise differences. One-way ANOVA was used to determine differences within a group subjected to repeated measures. P values ≤0.05 were considered significant for all tests.
General characteristics of all rats used in this study (protocols 1–3) are shown in Table1. At the time of use, the HS rats were slightly older than the LS rats, but they were also slightly smaller, suggesting that the high-salt diet may have been somewhat less palatable than the low-salt diet. Mean arterial pressure was not significantly different between the two groups. Table2 displays the characteristics of all arterioles studied in vivo (protocols 1 and 2). There were no significant differences in resting or passive arteriolar diameter or in calculated arteriolar tone between groups.
Under the normal superfusate (Fig.1 A), arterioles in LS rats exhibited dose-dependent dilations in response to 5-, 20-, and 40-nA current doses of ACh (steady-state diameters averaging 18 ± 7, 31 ± 8, and 37 ± 8% above control, respectively). In contrast, arterioles in HS rats did not respond to ACh at 5 or 20 nA (steady-state diameters averaging 0.4 ± 1 and 1.5 ± 2% below control) and dilated significantly less than those in LS rats when challenged with 40-nA ACh (12 ± 5% above control). In the presence of SOD and Cat (Fig. 1 B), arterioles in HS rats no longer showed a suppressed response to ACh at any current dose, with dilations in response to the 5-, 20-, and 40-nA applications averaging 27 ± 4, 42 ± 6, and 43 ± 11% above control compared with dilations of 32 ± 7, 48 ± 5, and 47 ± 6% above control for arterioles in LS rats.
As shown in Fig. 2, there were no differences between LS and HS rats in arteriolar responses to the NO donor SNP. SNP applied at the 5-, 20-, and 40-nA current doses produced dilations of 62 ± 17, 99 ± 27, and 143 ± 13% above control in LS rats and dilations of 66 ± 13, 108 ± 22, and 113 ± 19% above control in HS rats.
The superfusate concentration of TEMPO/Cat chosen for this study was sufficient to attenuate the effect of HX/XO on individual arterioles in LS rats (Fig. 3). Application of HX/XO caused a dose-dependent dilation under the normal superfusate but not in the presence of TEMPO/Cat. TEMPO/Cat significantly reduced the dilator response to HX/XO delivered at 10 psi (from 80 ± 8 to 50 ± 7% above control) and at 20 psi (from 139 ± 27 to 62 ± 12% above control).
Figure 4 summarizes arteriolar responses to ACh in each dietary group before and during exposure to TEMPO/Cat. Before TEMPO/Cat was added to the superfusate, ACh delivered at 5, 20, and 40 nA dilated arterioles by 20 ± 2, 30 ± 3, and 36 ± 3% in LS rats but by only 3 ± 1, 11 ± 2, and 14 ± 3% in HS rats. As in the first set of experiments (Fig. 1), these differences in responsiveness between groups were significant at each current dose. In the presence of TEMPO/Cat, there were no significant differences between LS and HS rats in arteriolar responses to ACh at 5, 20, or 40 nA (dilations of 19 ± 1, 33 ± 3, and 37 ± 4% in LS rats vs. dilations of 15 ± 1, 28 ± 2, and 39 ± 4% in HS rats).
After exposure to TNBT, the control light absorption value for microvessel walls (arterioles and venules combined) was −0.04 ± 0.01 (dimensionless units) in LS rats and −0.09 ± 0.01 in HS rats (Fig. 5). This difference was significant. Because the absorption value is inversely related to formazan content (and therefore ROS activity), we infer that under resting conditions ROS activity is significantly higher in HS microvessels than in LS microvessels. Infusion of HX/XO significantly reduced the absorption value in LS vessels (to −0.07 ± 0.01) but not in HS vessels (−0.08 ± 0.01).
Figure 6 illustrates the light absorption values for arterioles versus venules in LS and HS rats. In muscle that had not been exposed to TEMPO/Cat, arteriolar wall absorption values were significantly lower in HS rats (−0.07 ± 0.01) than in LS rats (−0.03 ± 0.004). In these specimens, venular wall absorption values were also lower in HS rats (−0.11 ± 0.01) than in LS rats (−0.05 ± 0.01). These findings suggest greater ROS activity in both the pre- and postcapillary microvessels of HS rats. In muscles that had been treated with TEMPO/Cat, arteriolar wall absorption values were significantly increased in HS rats (to −0.02 ± 0.01) but not in LS rats (−0.03 ± 0.004). In contrast, treatment with TEMPO/Cat increased venular wall absorption values in both LS and HS rats (to −0.02 ± 0.01 in LS and −0.04 ± 0.01 in HS rats). In the presence of TEMPO/Cat, there were no significant differences between HS and LS rats in either arteriolar or venular wall absorption values.
The importance of ROS in vascular physiology and pathophysiology is becoming increasingly evident. Superoxide anion and hydrogen peroxide can influence cellular homeostasis by altering gene expression, signaling pathways, or the redox state of key cellular components (33, 37, 40,41). In the vessel wall, the reduction of molecular O2 to superoxide is facilitated by enzymes such as XO, cyclooxygenase, NAD(P)H oxidase, and NO synthase (6,41). Cellular damage from ROS is normally prevented by the activity of antioxidant enzymes (such as SOD and Cat) that convert superoxide and hydrogen peroxide to O2 and H2O. A change in the activity of any one these enzymes has the potential to alter oxidative balance in the vascular wall.
The major findings of the current study are as follows.1) In normotensive rats fed a high-salt diet, the responsiveness of arcade arterioles to ACh was suppressed, whereas arteriolar responsiveness to the NO donor SNP was unchanged.2) Exposure of the vascular bed to ROS scavengers reversed the suppressed responsiveness to ACh in rats fed high salt but had no effect on arteriolar responsiveness to ACh in rats fed low salt.3) Quantification of oxidative activity in the microvascular wall revealed that arterioles and venules in rats fed high salt had significantly higher oxidant levels than those in rats fed low salt.4) In rats fed high salt, exposure to ROS scavengers significantly lowered arteriolar and venular oxidant activity to the same levels seen in rats fed low salt. These findings are consistent with the hypothesis that ingestion of a high-salt diet leads to increased ROS generation in the microvascular wall and that this increased oxidant activity is responsible for decreased arteriolar responsiveness to endothelium-dependent dilators.
The effect of a high-salt diet on normal physiological mechanisms is an important issue for our society because health organizations such as the American Heart Association, National Academy of Sciences, National Heart, Lung, and Blood Institute, and World Health Organization recommend lowering sodium intake for better health (20). Researchers investigating the consequences of a diet high in salt have observed structural and functional changes in the microvasculature that are independent of any change in blood pressure. A reduction in microvessel density and structural degeneration of microvascular walls has been observed in the cremaster muscle of rats fed high salt (12, 15). The decrease in circulating angiotensin II levels that accompanies high salt intake is apparently responsible for this network restructuring (17). M. A. Boegehold (2, 3) has previously shown that arterioles in the spinotrapezius muscle of rats fed a high-salt diet display reduced endothelium-dependent responses to increased shear stress and exogenous ACh. These reduced responses are due to a selective loss of NO activity. The correlation between a high-salt diet and decreased microvascular responsiveness to ACh has also been observed in the small feed arteries of the rat gracilis muscle (23) and arterioles in the rat cerebral cortex and cremaster muscle (9, 10, 24). Our current findings confirm this salt-dependent reduction in responsiveness to ACh (Figs. 1 and 4) and, because ROS can effectively scavenge endothelial NO (18, 33,39), provide a plausible explanation for the loss of NO activity that underlies this effect.
To investigate the influence of ROS on endothelium-dependent dilation in each dietary group, we examined the effect of ROS scavengers on arteriolar responses to ACh. We found that exposing the vessels to SOD/Cat for as little as 30 min could reverse the suppressed responses to ACh seen in HS rats (Fig. 1). These ROS scavengers presumably eliminated ROS that had been prematurely inactivating NO. Scavengers of ROS have also been found to reverse the deficit in NO activity associated with other abnormal conditions. For example, exposure to SOD increases NO activity in mesenteric resistance arteries and cultured aortic cells from the stroke-prone spontaneously hypertensive rat (13, 39). The loss of arteriolar NO activity following acute increases in luminal pressure can also be reversed with SOD (18). Because we were somewhat concerned about the possible inaccessibility of exogenously applied SOD to intracellular superoxide (22), we repeated these experiments using TEMPO, a more membrane-permeable SOD mimic (11,31). We first confirmed that the chosen concentration of TEMPO/Cat was sufficient to scavenge ROS in the vicinity of arterioles by demonstrating that it was capable of greatly attenuating the arteriolar dilation caused by direct application of HX/XO (Fig. 3). As seen with SOD/Cat, the combination of TEMPO/Cat reversed the suppressed arteriolar responses to ACh in HS rats, whereas arteriolar responses in LS rats were not altered (Fig. 4).
We used the compound TNBT to visualize the level of oxidative stress in vessels from both dietary groups. When reduced, TNBT forms insoluble blue-black formazan deposits that can be quantified by calculating vessel wall light absorption. The formation of these deposits can be prevented by SOD (1, 8), which has led to the conclusion that superoxide anion is specifically responsible for TNBT reduction (8, 28). However, because superoxide anion is also the precursor of hydrogen peroxide and ultimately hydroxyl radical (32, 40), its removal will also reduce the formation of other ROS that could contribute to TNBT reduction. We therefore considered the presence of formazan deposits to reflect general oxidant activity rather than superoxide activity per se. Our findings suggest that vessels in HS rats had a significantly higher level of oxidant activity than those in LS rats (Figs. 5 and 6). As shown in Fig. 6, this elevated activity was found in both the arterioles and venules of HS rats. In HS rats, venules were found to have a higher formazan content than arterioles. There is also evidence of increased postcapillary oxidant activity in salt-sensitive hypertension. Swei et al. (38) found evidence of arteriolar and venular formazan formation after TNBT exposure in Dahl-S rats fed high salt, with higher formazan levels in the venules than the arterioles. Arteriolar-venular differences in the activity of oxidant or antioxidant enzymes may account for this difference. As also shown in Fig. 6, exposure to TEMPO/Cat prevented arteriolar and venular formazan formation in HS rats (i.e., light absorption values not different from those in LS rats), verifying that the differences between HS and LS rats were due to differences in oxidant activity. Interestingly, TEMPO/Cat also significantly reduced the light absorption value for venules in LS rats, suggesting a modest basal production of ROS that could possibly play a role in normal signaling pathways (36).
The responsiveness of arterioles to the NO donor SNP was not different between the two dietary groups (Fig. 2), suggesting that the ability of vascular smooth muscle to respond to NO is not altered by the consumption of a high-salt diet. This inability of dietary salt to change vascular responsiveness to NO is in agreement with previous reports (2, 9, 23,24). Because SNP is converted to NO by enzymes found in the vascular smooth muscle membrane (21), our finding of normal responsiveness to SNP in HS rats suggests that NO inactivation by ROS does not occur to any appreciable extent within the vascular smooth muscle cell. If it did, then arteriolar responses to SNP also would have been reduced in HS rats. A more likely site of ROS generation and activity in HS rats is the microvascular endothelium, as suggested by Swei et al. (38).
As mentioned above, the combination of ROS scavengers that we used should lead to reduced levels of superoxide anion as well as its reactive metabolites. Therefore, the identity of the specific oxidant responsible for reduced endothelium-dependent dilation in the current study is unclear. Superoxide anion itself can rapidly inactivate NO (14, 19, 33), and this reaction occurs at a rate approximately three times faster than the interaction of superoxide with SOD, suggesting a preferential interaction between superoxide and NO in the vascular wall (16). Experimentally generated superoxide anion inactivates NO released from the arteriolar endothelium in rat cremaster muscle (19). In rat gracilis muscle arterioles, a transient increase in intravascular pressure eliminates the contribution of endothelium-derived NO to flow- and agonist-induced dilation, and this effect has also been attributed to the generation of superoxide anion (18). The selective loss of the NO-mediated components of these responses is similar to what this laboratory has reported in the spinotrapezius muscle of rats fed a high-salt diet (2,3), raising the possibility that superoxide could act in a similar fashion during high salt intake. Hydroxyl radical can also inactivate NO under some conditions (25); this radical is responsible for the loss of endothelium-dependent relaxation in the cat cerebral vasculature and general cerebrovascular injury in the rat following ischemia-reperfusion (25, 26). Finally, hydrogen peroxide formed by the dismutation of superoxide anion is also capable of oxidizing NO (25).
The sequence of events that link high salt intake to increased microvascular oxidant levels is unclear, but one interesting possibility is that dietary salt may change the expression or activity of enzymes important for the regulation of cellular oxidant activity. Future studies are planned to evaluate this possibility.
We gratefully acknowledge the expert technical assistance of Kimberley Wix in this study.
This work was supported by National Heart, Lung, and Blood Institute Grants HL-44012 and HL-52019.
Address for reprint requests and other correspondence: M. A. Boegehold, Dept. of Physiology, West Virginia Univ. School of Medicine, PO Box 9229, Robert C. Byrd Health Sciences Center, Morgantown, WV 26506-9229 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society