INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

THE IMPRESSIVE INCREASE in blood flow to the working skeletal muscle that occurs during acute exercise serves to augment the delivery of oxygen and nutrients and the removal of metabolic waste (11). This is accomplished by an increase in cardiac output along with redistribution of blood flow, resulting in dramatic increases in cardiac and skeletal muscle perfusion during exercise (11). Recent studies have focused on the possible role of endothelium-derived nitric oxide (NO) in exercise-induced dilatation of the systemic and muscle vasculature. NO is synthesized from L-arginine by an oxidoreductase, namely NO synthase (NOS). To date, three isoforms of NOS have been identified. Neuronal NOS (nNOS, NOS I) and endothelial NOS (eNOS, NOS III) are calcium-dependent isoforms that are expressed under physiological conditions (25). In contrast, the activity of NOS II (iNOS), the expression of which is induced by immunological stimuli, is Ca²⁺ independent (25). Recent reports suggest the presence of a mitochondrial isoform of NOS with characteristics intermediate between constitutive and inducible NOS (8, 39).

Experimental evidence, using the radiolabeled microspheres technique, has demonstrated an enhancement of blood flow in the hindquarter skeletal muscle of exercise-trained animals (13, 22, 26). Hirai and co-workers (12) reported that NO contributes to the exercise-induced hyperemic blood flow response in rat hindquarter muscle. In isolated hindquarter skeletal muscle arterioles of rat, short-term daily exercise was reported to increase ACh-induced synthesis of NO by the endothelium (37). However, in vivo measurements of swim training-induced changes in hindquarter vascular reactivity have not been performed.

Previous studies reported enhanced expression of eNOS mRNA in aortic endothelial cells of dogs (35) and of eNOS protein expression in the abdominal aortas of rats (6) exercised by running. A recent study by Roberts and co-workers (32) showed that acute exercise increases total NOS activity in the skeletal muscle. However, combined measurements of the alterations in NOS activity and expression in the aorta, as well as in the hindquarter skeletal muscle, heart, lung, and viscera of rats trained by swimming, have not yet been reported.

There are some data suggesting that blood flow to the splanchnic and renal vascular beds is reduced during exercise (12, 13, 22, 26). Moreover, blood flow to specific abdominal organs (kidney, liver, pancreas, stomach, and intestine) is regulated by NO, at least partially, during exercise (12). NO also plays a role in regulating blood flow through the mesenteric arterial bed (27). To date, the changes in vasoreactivity of the superior mesenteric arterial bed induced by swim training are unknown, and the accompanying changes in NOS activity and expression of the abdominal organs have not been studied.

We investigated the changes in rat hindquarter vascular reactivity induced by swim exercise training in vivo by measuring hindquarter blood flow (HQBF) and vascular conductance (HQC). Responses to the vasoconstrictor norepinephrine (NE) and the endothelium-dependent vasodilators ACh and bradykinin (BK) were studied. Responses to the NOS inhibitor N^G-nitro-L-arginine methyl ester (L-NAME) or the NO donor S-nitroso-N-acetyl-penicillamine (SNAP) were investigated as well. The responses to these vasoactive drugs were also studied in a visceral vascular bed, the isolated perfused superior mesenteric arterial bed. We also determined whether an organ-specific alteration in NOS activity and expression would occur in the hindquarter skeletal muscle, heart, aorta, lung, kidney, and liver by swim exercise training. These organs were selected because of expected variations in exercise-induced changes in blood flow, and consequently shear stress, within their vascular beds. We tested the hypothesis that swim exercise training enhances NO-mediated vasoreactivity of the hindquarter circulation and upregulates NOS in hindquarter skeletal muscles and cardiopulmonary organs but does not alter mesenteric vasoreactivity or visceral NOS.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animal care and experimental design. All experiments and protocols were performed in accordance with the regulations of the Canadian Council on Animal Care and approved by the Animal Care Committee at the University of Saskatchewan. All in vivo hemodynamic data (systemic arterial blood pressure and HQBF) were collected from three separate sets of sedentary or swim-trained rats. From an additional two sets of rats, blood samples were withdrawn for measurement of plasma nitrate and nitrite (NO_x) level, then the superior mesenteric arterial bed was isolated and tissue samples were collected for determination of NOS activities and protein expression.

Animals and swim exercise training. Studies were carried out in outbred male Sprague-Dawley rats (Charles River, St. Constant, QC, Canada) of 8-10 wk of age. Animals were kept under standard conditions in the animal house for at least 2 wk before the training program was started. Rats were randomized into two groups: one group underwent a swimming program for 3-4 wk in a 1-m²-surface area tank containing tap water ~50 cm in depth and maintained at 20-22°C (swim trained); the second group was placed in a 10-cm water tank that did not require the rats to swim under the same schedule (sedentary). On the first day of the training program, the swimming time was 10 min in duration, and this was then increased by 10 min each subsequent day until the maximum swim duration of 60 min was reached. On the last week, rats swam for 60 min twice a day, 5 days/wk. Swimming sessions were supervised and the number of rats per session (10 rats) was kept low to avoid "gang swimming." Body weight and water and food intake were measured before the start of the swimming program and then weekly during the entire training period.

In vivo measurements of arterial blood pressure, HQBF, and HQC. Rats were anesthetized with pentobarbital sodium (60 mg/kg ip). The trachea was intubated to facilitate spontaneous respiration. Body temperature was maintained at 36-37°C using a heating lamp table controlled by a rectal thermistor probe. A polyethylene cannula (PE-50, Clay Adams, Parsippany, NJ) filled with heparinized saline (10 IU/ml) was inserted into the left carotid artery for the measurement of arterial blood pressure by a pressure transducer connected to a Grass polygraph (model RPS, 7C8). A second PE-50 cannula was also inserted into the left femoral vein for the intravenous administration of drugs and supplemental anesthetic.

Determination of NOS activities. After removal of the mesenteric arterial bed, the right kidney, liver, atria, ventricles, thoracic aorta, a lobe from the right lung, and a sample of hindquarter skeletal muscle (left quadriceps femoris) were rapidly dissected. Excess fat and connective tissues were trimmed from the skeletal muscle, kidney, and aorta. All tissues were then quickly frozen in liquid nitrogen and stored at 80°C until analysis. After measurement of atrial and ventricular wet weights, the frozen tissues were used for the assay of NOS activities as previously described (18, 19). In brief, tissues were pulverized with mortar and pestle under liquid nitrogen, homogenized, and subsequently centrifuged (1,000 g, 5 min, 4°C). The resultant supernatant was used for measurement of NOS activities by the conversion of L-[¹⁴C]arginine to [¹⁴C]citrulline. NOS activity was determined as the rate of citrulline production that was normalized to the protein concentration of the homogenates, as determined by the bicinchoninic acid assay using bovine serum albumin as a standard, and was expressed as picomoles per minute per milligram protein. The limit of detection was 0.1 pmol · min¹ · mg protein¹.

Western blotting and densitometric analysis. Western blot analysis of tissue homogenates prepared as above was performed as previously described (18, 19) with a few modifications. Samples (60, 15, 12, or 7.5 µg protein per lane for skeletal muscle, lung, aorta, and atrial tissues, respectively) were size fractionated by polyacrylamide (9%) gel electrophoresis and transferred to nitrocellulose membranes by wet electroblotting. The membranes were blocked at room temperature in 10% skimmed dried milk in phosphate-buffered saline for 3-5 h and then incubated for 1 h with murine monoclonal antibodies (Transduction Laboratories, Lexington, KY) against either human eNOS (1:4,000) or human nNOS (1:2,000) and thereafter with horseradish peroxidase-conjugated polyclonal goat anti-mouse IgG (1:4,000, Transduction Laboratories) for 1 h. The immunoreactive proteins were detected by using an enhanced horseradish peroxidase-luminol chemiluminescence reaction kit (Western Blot Chemiluminescence Reagent Plus, NEN Life Science Products, Boston, MA). Autoradiographs were obtained by exposure to X-ray film (Kodak, X-OMAT) for 30-35 s, and the density of the bands was quantified by scanning densitometry using a ScanJet 6100C scanner (Hewlett Packard, Boise, ID) and SigmaGel measurement software (Jandel, San Rafael, CA) and expressed as a percentage relative to the positive external control. As external controls for eNOS and nNOS in each blot, we used a single ventricular homogenate sample from a sedentary 8-wk-old rat or rat pituitary lysate (Transduction Laboratories) as supplied by the manufacturer, respectively.

Plasma NO_x levels. Blood (0.5 ml) was collected from the femoral artery under light ether anesthesia 1 wk before the swimming program started for determination of pretraining plasma NO_x level. A posttraining blood sample (0.5 ml) was collected by puncture of the lower abdominal aorta 1 day after the last exercise episode under light ether anesthesia before removal of the mesenteric arterial bed. Blood was immediately placed in a 1.5-ml microcentrifuge tube containing 20 µl heparin sodium (10 IU/ml). Plasma was separated from blood by centrifugation at 2,000 g for 4 min at 4°C and stored at 20°C until analysis. After thawing, plasma was diluted 1:1 with deionized water. Then, 400 ml of the diluted plasma were deproteinized by centrifugal ultrafiltration (Millipore Ultrafree-MC microcentrifuge tubes UFC3, Bedford, MA). Ultrafiltrates were analyzed for total NO_x content by the method of Green et al. (9). The limit of detection was 0.1 µM for both nitrite and nitrate.

In vitro vasoreactivity of the superior mesenteric arterial bed. In two sets of animals anesthetized with ether, the superior mesenteric artery was cannulated and the gut was removed as described previously (23, 38). The mesenteric arterial bed was perfused through the cannula at a constant flow rate of 5 ml/min and superfused at 0.2 ml/min using two separate pumps with modified Krebs-Henseleit solution (in mM: 118 NaCl, 4.7 KCl, 1.2 MgCl₂ · 6H₂O, 1.0 NaH₂PO₄, 2.6 CaCl₂ · 2H₂O, 25 NaHCO₃, 11.1 glucose; 37°C; pH 7.35-7.45), oxygenated with a 95% oxygen-5% carbon dioxide gas mixture. The Krebs-Henseleit solution contained 1 µM indomethacin to block cyclooxygenase activity. A bubble trap removed any air bubbles from the perfusate (38). Drugs were injected into the perfusate, and arteriolar constriction or dilatation was determined by the changes in perfusion pressure recorded from a pressure transducer (Beckman, Palo Alto, CA) that was connected to a catheter that had been placed in the perfusion circuit just before the mesenteric arterial bed. Mean perfusion pressure was recorded after electronic integration of the pulsatile pressure signal. All reported perfusion pressure values were corrected by subtracting the pressure generated by the tubing circuit of the perfusion system.

7

9

7

12

10

9

6

Materials. All compounds were purchased from Sigma except L-[U-¹⁴C]arginine monohydrochloride (Amersham), L-NMMA acetate (Alexis, San Diego, CA), and BK (American Peptide, Sunnyvale, CA). Norepinephrine bitartrate was dissolved in 0.1 mM ascorbic acid and BK in physiological saline containing 0.1% bovine serum albumin.

Statistical analyses. Data are expressed as means ± SE. Values of n refer to the number of rats in each group. Experiments were performed according to a randomized block design. Significant differences between the groups were determined using either Student's unpaired t-test (NOS activity and expression) or two-way repeated-measures ANOVA (MABP, HQBF, and HQC and superior mesenteric bed vasoreactivity). Data analysis was done using a statistical software package (SPSS, version 9.0, SPSS, Chicago, IL). Differences were considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Indexes of training and baseline cardiovascular parameters. Food and water intake and body weight were not statistically different between the two groups of rats before swim training commenced (data not shown). The characteristics of sedentary and trained rats are summarized in Table 1. The body weight was significantly lower in trained rats compared with sedentary animals. This occurred despite the fact that both food and water intake were significantly higher in trained rats. Atrial and ventricular weights significantly increased with swim training. Accordingly, the ratios of atrial or ventricular weights to total body weight were markedly increased in trained rats. Under pentobarbital sodium anesthesia, the resting heart rate was significantly lower in trained rats than in sedentary controls. Basal MABP was not significantly different between the two groups. HQBF was significantly higher in trained than in sedentary rats, but HQC was not different among the two groups.

MABP, HQBF, and HQC. NE dose dependently increased MABP and decreased both HQBF and HQC in sedentary and swim-trained rats, but no statistically significant differences were noted between the groups (Fig. 1). L-NAME also caused a dose-dependent increase in MABP together with a decrease in HQBF and HQC in both groups of rats. However, the L-NAME-induced changes in all these parameters were significantly greater in swim-trained rats compared with sedentary controls (Fig. 2). ACh produced a short-lasting fall in MABP that was not statistically significant between groups. It increased HQBF and HQC in both groups in a dose-dependent manner, and these responses were significantly greater in swim-trained rats compared with the sedentary controls (Fig. 3). When tested at two doses, both BK and SNAP decreased MABP and increased the HQBF and HQC in both sedentary and trained rats to a similar extent (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Changes in () mean arterial blood pressure (MABP, A) and hindquarter blood flow (HQBF, B) and conductance (HQC, C) in response to intravenous administration of norepinephrine (NE) in anesthetized sedentary and swim-trained rats. Values are means ± SE; n = 6-8. Differences between the groups were not significant (NS).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Changes in MABP (A), HQBF (B), and HQC (C) in response to intravenous administration of NG-nitro-L-arginine methyl ester (L-NAME) in anesthetized sedentary and swim-trained rats. Values are means ± SE; n = 8 or 9. Differences between the groups were statistically significant (P < 0.05).

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Changes in MABP (A), HQBF (B), and HQC (C) in response to intravenous administration of ACh in anesthetized sedentary and swim-trained rats. Values are means ± SE; n = 6-8. Differences between the groups were statistically significant (P < 0.05) for changes in HQBF and HQC only.

NOS activities and protein expression in the hindquarter skeletal muscle. The Ca²⁺-dependent NOS activity in homogenates of hindquarter skeletal muscle from sedentary and trained rats is depicted in Fig. 4A. The enzyme activity increased significantly with swim training and was ~40% higher than that in sedentary rats. No significant differences in renal Ca²⁺-dependent NOS activity were detected between sedentary and swim-trained rats (0.16 ± 0.05 and 0.17 ± 0.07 pmol · min¹ · mg protein¹, respectively), and the activity in the liver of both groups was below the detection limit. Figure 4B shows a representative immunoblot, using the anti-human nNOS monoclonal antibody, of immunoreactive proteins in the hindquarter skeletal muscle of sedentary and trained rats as resolved by 9% SDS-PAGE. Two bands were observed, the upper one corresponding to the 160-kDa nNOS (20) and a lower band of a molecular mass estimated at 120-125 kDa. These blots demonstrate a marked enhancement of the expression of the 160-kDa nNOS, as well as the 120- to 125-kDa band, in the swim-trained rats compared with sedentary animals. In fact, the density of nNOS protein (160-kDa band) was 70% higher in these rats compared with sedentary controls when determined by densitometric analysis (Fig. 4C). No bands were seen when the resolved skeletal muscle proteins were incubated with an anti-eNOS monoclonal antibody (data not shown). Ca²⁺-independent NOS activity in the skeletal muscle and kidney was not detectable in either group of rats. In the liver, this activity was detectable but not significantly different between sedentary and trained rats (1.86 ± 0.79 vs. 1.52 ± 0.81 pmol · min¹ · mg protein¹, respectively).

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Changes in Ca2+-dependent nitric oxide synthase (NOS) activity and neuronal NOS (nNOS) protein expression in homogenates of hindquarter skeletal muscle (left quadriceps femoris) of sedentary and swim-trained rats. A: Ca2+-dependent NOS activity. B: Western blot of nNOS protein expression; 60 µg protein was loaded per lane. C: densitometric analysis of changes in nNOS protein expression. Data are presented as a percent relative to the external control for nNOS. Values are means ± SE; n = 5 or 6. * P < 0.05 vs. sedentary controls.

NOS activities and protein expression in the lung, aorta, atria, and ventricles. The Ca²⁺-dependent NOS activity in homogenates of lung, aorta, atria, and ventricular tissues is depicted in Fig. 5A. In sedentary rats, the enzyme activity was significantly different among the tissues in the following rank order: aorta > atria > ventricles > lung. Exercise training induced a marked increase in Ca²⁺-dependent NOS activity in the lung (45%), aorta (45%), and atria (150%). Surprisingly, ventricular Ca²⁺-dependent NOS activity did not change significantly with training. Figure 5B shows a representative blot of immunoreactive eNOS protein in lung, aorta, and atrial homogenates as resolved by 9% SDS-PAGE. By using an eNOS-specific monoclonal antibody, a single 135-kDa band was observed in all tissues. Enhanced expression of eNOS protein was seen in lung (65%), aorta (190%), and atria (55%) of trained rats compared with the same tissues in sedentary animals as determined by densitometric analysis of the blots (Fig. 5B). No bands were seen when lung, aorta, and atrial samples were incubated with anti-nNOS antibody under the same conditions (data not shown). Ca²⁺-independent NOS activity was not detectable in any of these tissues except for a low level in the lung. However, no significant difference was detected between sedentary and trained rats (0.20 ± 0.11 vs. 0.15 ± 0.04 pmol · min¹ · mg protein¹, respectively).

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Changes in Ca2+-dependent NOS activity and endothelial NOS (eNOS) protein expression in pulmonary, aortic, atrial, and ventricular homogenates from sedentary and swim-trained rats. A: Ca2+-dependent NOS activity. B: Western blot of eNOS protein expression in the lung, aorta, and atria; 15, 12 and 7.5 µg of proteins, respectively, were loaded per lane. S, sedentary; T, swim trained. C: densitometric analysis of changes in eNOS protein expression in sedentary and swim-trained rats. Data are presented as a percent relative to the same external control run in each blot. Values are means ± SE; n = 6 or 7. * P < 0.05 vs. sedentary controls.

Plasma NO_x levels. Before training, no significant differences in the level of plasma NO_x were detected between sedentary and swim-trained rats (21.0 ± 1.6 vs. 17.1 ± 2.2 µM, n = 9 for each group, P > 0.05). As expected, no significant changes in NO_x levels (16.1 ± 1.6 µM) were observed in rats after 3-4 wk of sedentary living. Swim exercise training did not elicit any change in plasma NO_x levels (18.4 ± 2.1 µM).

Superior mesenteric arterial bed vasoreactivity. Constant flow perfusion of the isolated mesenteric arterial bed resulted in a steady basal perfusion pressure after a 60-min stabilization period, which was similar in both sedentary and swim-trained rats (21.0 ± 1.1 vs. 21.7 ± 1.3 mmHg, respectively, n = 8-10). In a concentration-dependent manner, NE increased perfusion pressure in both groups with no statistically significant difference between them (Fig. 6A). Thirty-minute incubation with L-NAME had no effect on basal perfusion pressure (23.7 ± 0.9 and 22.4 ± 1.2 mmHg in swim-trained and sedentary rats, respectively, n = 8-10) but significantly potentiated the vasoconstrictor responses to NE, causing a leftward shift in the concentration-response curve to NE (Fig. 6A). However, no differences between sedentary and swim-trained rats were observed. Methoxamine raised perfusion pressure to a similar extent in both sedentary and swim-trained groups (100.0 ± 14.7 and 108.1 ± 7.0 mmHg, respectively, n = 6-10). ACh induced a concentration-dependent fall in perfusion pressure of the methoxamine-preconstricted mesenteric arterial bed; the response was similar in both groups (Fig. 6B). SNAP induced a concentration-dependent fall in perfusion pressure of the methoxamine-preconstricted mesenteric arterial bed; however, no significant differences between the two groups of rats were observed (Fig. 6C).

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Changes in perfusion pressure of the isolated superior mesenteric arterial bed in response to the administration of vasoactive drugs in sedentary and swim-trained rats. A: vasoconstrictor effect of NE administered before and after 30-min incubation with L-NAME. B: vasodilator effect of ACh in the methoxamine-preconstricted bed. C: vasodilator effect of S-nitroso-N-acetyl-penicillamine (SNAP) in the methoxamine-preconstricted bed. Values are means ± SE; n = 6-10. Differences between the groups were not significant. If not shown, error bars are within the height of the symbol.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study is the first to demonstrate an upregulation of Ca²⁺-dependent NOS activity in the hindquarter skeletal muscle, lung, atria, and aorta of rats trained by swimming. The increased enzyme activity was associated with enhanced expression of neuronal NOS in the hindquarter skeletal muscle and endothelial NOS in the lung, atria, and aorta. Upregulation of skeletal muscle nNOS occurred in conjunction with augmented vascular responses in the lower limbs to L-NAME and ACh, suggesting enhanced NO production within the hindquarter skeletal muscle and vasculature as a result of swim training.

The increases in atrial and ventricular weights and their ratios to total body weight, as well as the reduction in resting heart rate, confirm the efficacy of the swim training protocol implemented in this study. In addition, the fact that the body weight of trained rats was significantly lower than that of sedentary controls, despite a higher food intake, suggests an increased metabolic rate and energy expenditure in swim-trained animals. These changes are in accordance with previously published data in rats exercised by running (37) and are representative of the physiological adaptations of the body to regular physical exercise. Baseline HQBF was significantly higher in trained rats compared with sedentary animals, probably reflecting a change in vascular tone or an altered metabolic state in the resting lower limb muscles. However, HQC in swim-trained rats was not significantly different from control rats because MABP was somewhat elevated in this group, although this did not reach statistical significance.

Swim training was associated with an increase in basal and agonist-stimulated NO production in the hindquarter circulation. Inhibition of NOS activity with L-NAME reduced HQBF and HQC and increased MABP in swim-trained rats to a higher degree than that in sedentary controls. Moreover, the "classical" endothelium-dependent vasodilator, ACh, induced significantly higher changes in HQBF and HQC in swim-trained rats. In contrast, the vasoconstrictor responses evoked by NE and the vasodilatory responses induced by SNAP were similar in sedentary and swim-trained rats, suggesting that exercise training did not alter hindquarter vascular smooth muscle reactivity in a nonspecific fashion. These data support the findings of Sun et al. (37) whereby the in vitro responses of rat gracilis muscle arterioles to NE, SNAP, and sodium nitrite were unchanged by short-term daily exercise but the vasodilator response to ACh was enhanced. Therefore, swim training appears to evoke a specific increase in basal and agonist-stimulated NO production in the vascular endothelium, presumably due to increased blood flow to the lower limbs with increased vascular wall shear stress and endothelial NO production (4, 24). Alternatively, the biodegradation of endogenous NO may be lessened as a result of swim training, a notion in accordance with enhanced antioxidant defenses seen with exercise.

Unlike ACh, responses to another endothelium-dependent vasodilator, BK, were similar in both sedentary and trained rats. Both ACh and BK stimulate the release from endothelial cells of a number of vasodilator factors, including prostaglandins, NO, and endothelium-derived hyperpolarizing factor. Several explanations as to why the responses to BK were similar in the two groups are possible: 1) ACh and BK receptor-coupled mechanisms may be differentially affected by training, 2) alterations in vasoconstrictor responses to BK via BK₁ receptors on smooth muscle cells may have masked a possible enhanced NO release from endothelial cells (mediated by BK₂ receptors), and 3) the two doses of BK used were higher than those that may have revealed differences in blood flow and conductance between the groups.

The specific isoform of NOS that accounted for the increase in total NOS activity observed in the hindquarter circulation of swim-trained rats was an important aspect of the investigation. The increase in NOS activity in the hindquarter skeletal muscle of rats is in agreement with and extends the recent report of increased skeletal muscle total NOS activity by acute exercise (32). We report that this NOS activity is Ca²⁺ dependent. Because this activity measures in principle that from both eNOS and nNOS, we performed Western blot analysis using monoclonal antibodies directed against either of these to identify which NOS was responsible for the altered activity. Unlike previous reports that noted both eNOS and nNOS protein expression in skeletal muscles (2, 17, 20, 40), no bands were observed when skeletal muscle homogenates from sedentary rats were incubated with anti eNOS antibody, but two bands were seen when an anti-nNOS antibody was used. Our finding of swim training-induced enhancement of nNOS (160 kDa) protein expression in the rat's quadriceps femoris muscle confirms previous data reported by Balon and Nadler (2) in which chronic treadmill running of rats was found to increase nNOS protein expression in the soleus muscle. The identity of the lower molecular mass band (120-125 kDa) seen in both sedentary and swim-trained rats is unknown at the present time. It may represent one of the nNOS splice variants that have been identified in skeletal muscle (36) or a mitochondrial NOS that may have cross-reacted with this antibody (7). Notably, the expression of this band was also increased in swim-trained rats. The fact that eNOS protein was undetectable in this muscle in our hands may be due to the low level of eNOS protein in the homogenate as a result of low vascularity of the muscle sample, a limited distribution of eNOS in skeletal muscle sarcolemma (40), or, although unlikely, the Western blot analysis not being sensitive enough for eNOS detection in skeletal muscle because of the concentration or type of primary antibodies used.

In contrast to the findings with the skeletal muscle, swim training increased Ca²⁺-dependent NOS activity and eNOS expression in the lung, aorta, and atria. These data are in accordance with observations of increased eNOS mRNA (35) and protein expression (6) in aortic extracts of trained dogs and rats, respectively. In the lung and aorta, upregulation of eNOS may be a consequence of exercise-induced increase in the cardiac output and blood flow that would cause an increase in wall shear stress in the aorta and pulmonary vessels. The endothelial cells lining these vessels sense the increase in shear stress and respond to such stimulus when of sufficient intensity and duration by increasing eNOS protein to release more NO (4, 24, 28, 29). However, the fact that atrial but not ventricular eNOS was enhanced with training suggests that the increased level of stretching and/or dilatation in the atria compared with ventricles (1), rather than increased shear stress, may be the underlying mechanism for the upregulation of NOS in this heart chamber. These findings indicate that swim exercise training upregulates the capacity of tissues to enhance the biosynthesis of NO by increasing NOS activity and/or the expression of eNOS protein. Because the heart and lung play a key role in the physiological adaptations to endurance training, our data highlight the contribution of NO to this.

Whereas endothelium-derived NO regulates vascular tone and blood flow (25), little is known about the role of nNOS-derived NO in skeletal muscle. Several explanations as to why nNOS is upregulated in skeletal muscle as a result of swim training are possible. First, muscle metabolism appears to be sensitive to NO at several sites, including glucose uptake, glycolysis, and creatine kinase activity (30). In fact, studies have shown that NO modulates skeletal muscle glucose transport (2, 3, 34). Therefore, upregulation of nNOS can be part of an exercise-induced improvement in muscle glucose metabolism. Of relevance, specific proteins that play a key regulatory role in control of glucose flux into skeletal muscle (GLUT-4 and hexokinase II) were also upregulated by increased contractile activity (33, 41). Second, exogenous and endogenous NO reversibly inhibit oxygen consumption and cellular respiration as shown in isolated bovine cardiac muscle (42). Increased NO production, via nNOS, in the setting of exercise can be a physiological mechanism to limit oxygen consumption in the actively contracting muscle fibers when the demand for oxygen exceeds its supply. Third, nNOS has been localized in the adult rat skeletal muscle endplate (21), and NO was reported to modulate cholinergic nerve transmission in Torpedo synaptosomes (31). nNOS may, therefore, regulate neuromuscular conduction in the actively contracting skeletal muscle. This is supported by the findings of Tews et al. (40), who showed decreased expression of nNOS in denervated muscles and the restoration of physiological nNOS expression by reinnervation. Finally, enhanced production of NO by nNOS in skeletal myocytes may also contribute to increased skeletal muscle vasodilator tone in swim-trained animals, because NO can diffuse readily over several cell diameters.

Our data indicate that the postexercise plasma NO_x level in trained rats was not significantly different from the preexercise level. Jonsdottir and co-workers (14) reported a significant decline in plasma nitrate level of spontaneously hypertensive rats after 7 days of exercise training. The level returned to an approximate pretraining value at 14 and 21 days of exercise and then was increased above the pretraining level after 35 days of exercise. In humans, a higher resting plasma nitrate level has been observed in athletic subjects than in nonathletic controls (15). These controversies may be related to either a possible strain or species variation, the type or intensity of exercise training, the fact that plasma nitrate level is influenced by exogenous sources such as nitrate- or nitrite-containing food or water (16) or differences in the rate of NO_x excretion (16).

Mesenteric arterial bed responsiveness was similar in both sedentary and swim-trained rats. A number of investigators (12, 13, 22, 26) have demonstrated diminished blood flow to the skin, kidneys, and organs served by the splanchnic circulation during exercise together with an increased flow to the hindquarter skeletal muscles. These reductions in blood flow are proportional to the relative exercise intensity and are mediated by vasoconstriction of the renal and splanchnic circulations (22). This may explain the lack of changes in mesenteric vasoreactivity and renal NOS because of reduced blood flow and shear stress. In addition, 30-min incubation with the NOS inhibitor L-NAME did not alter basal perfusion pressure in either sedentary or trained rats, suggesting a minimal role for NO in maintaining mesenteric vascular tone under basal conditions.

Some limitations of this study warrant further discussion. First, we cannot exclude the possibility that the observations made in this study were due to general stress, at least in part, rather than swim training itself. Second, drug-induced changes in systemic blood pressure may have been modified by the baroreceptor reflex (10). Third, the cellular sites of enhanced NOS expression in skeletal muscle, lung, atria, and aorta remain unknown and would require immunohistochemical localization. Fourth, the relationship between upregulated skeletal muscle nNOS and enhanced ACh-mediated responses in the nearby hindquarter vessels was not investigated. Furthermore, swim training-induced alterations in other vasoactive factors in the mesenteric circulation, such as prostaglandins and endothelium-derived hyperpolarizing factor, were not determined.

In conclusion, this study reveals that long-term swim training in rats enhances NO generation in the lower limbs and upregulates both nNOS activity and protein expression in hindquarter skeletal muscle and eNOS activity and protein expression in aorta, lung, and atria. The mechanism by which nNOS is upregulated by swim training and the physiological significance of enhanced NO production in the skeletal muscles of trained animals require further study.