INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

DESPITE YEARS OF RESEARCH, the exact mechanism underlying exercise hyperemia remains unknown (25). The onset of dynamic exercise elicits a significant increase in blood flow to active skeletal muscles. The increase in skeletal muscle metabolism that occurs during exercise can lead to the release of vasoactive substances that act on the vasculature to increase blood flow to match metabolic demand. The predominant substances released include sodium, potassium, and hydrogen ions; inorganic phosphate; lactate; adenosine; and glucose. Furthermore, the cumulative release of these substances may well elicit an increase in osmolarity, as evidenced by increases in the osmolarity of venous (16, 18, 19, 23) and arterial (17) blood during exercise or muscle contraction. Several of these substances may play a role in the multiple mechanisms underlying exercise hyperemia (23, 26). For a substance to be considered the primary dilator agent of exercise hyperemia, it should, as proposed by Shepherd (25), have access to skeletal muscle resistance vessels in a concentration large enough to elicit dilation. For example, arterial infusions of hyperosmotic solutions, used to mimic the increase in osmolarity observed during exercise, elicit decreases in vascular resistance proportional to the increase in osmolarity (16, 18, 19, 27). These observations suggest that hyperosmolality may contribute to exercise hyperemia. However, methodological limitations prevent direct measurement of osmolarity in the vicinity of the vascular tissue during muscle contraction or arterial infusion so that the concentration of the dilator substance reaching the resistance vessels is unclear (19). Therefore, it is unclear whether changes in osmolarity have a direct effect on skeletal muscle arterioles. Recently, Ishizaka and Kuo (7) reported that porcine coronary microvessels, which are also thought to be sensitive to changes in metabolism, dilate to extraluminal increases in osmolarity via ATP-sensitive potassium (K_ATP) channels on endothelial cells. This finding is in accord with the current view that activation of potassium channels may contribute to exercise hyperemia in skeletal muscle (21); however, the mechanism underlying the response to increasing osmolality in skeletal muscle arterioles is not known. Therefore, one aim of this study was to characterize the response of isolated skeletal muscle arterioles to both extra- and intraluminal hyperosmolality within the range observed during exercise and to determine the role of K_ATP channels in this response.

Exercise training is associated with several adaptations, including an increase in skeletal muscle blood flow at rest and during exercise (14). Several investigators have also demonstrated that exercise training influences vascular reactivity to vasodilator and vasoconstrictor agents in coronary and skeletal muscle blood vessels ranging in size from arterioles to large arteries (11, 13, 14, 28). The vascular adaptations to exercise have been attributed, in large part, to improvements in endothelial function, i.e., increased basal and stimulated release of nitric oxide and vasodilator prostaglandins. These adaptations have been observed in skeletal muscle arterioles after only 4 wk of mild daily exercise (11, 28). Longer training regimens have also been shown to augment ion channel function in the coronary circulation, including a greater role for potassium channels in regulating vascular tone (6). On the basis of these observations and the mechanism for hyperosmolarity-induced dilation in coronary vessels, the second aim of this study was to determine whether daily exercise augments the response to hyperosmolality in rat skeletal muscle arterioles through an endothelium-dependent mechanism.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Blood vessel preparation. Skeletal muscle arterioles were prepared as previously described in detail (29). Briefly, male Wistar rats (387 ± 3 g) were anesthetized with pentobarbital sodium (50 mg/kg ip), and the gracilis or soleus muscle was removed and placed in a refrigerated dissecting dish containing ice-cold (4°C) MOPS-buffered physiological saline solution (PSS). Arterioles were dissected from adhering tissue by microscissors and transferred to a vessel chamber (Living Systems Instruments, Burlington, VT) containing PSS. Vessels, 1-2 mm in length, were mounted onto two glass micropipettes, cleared of clotted blood, and slowly pressurized to 80 mmHg using a pressure-servo-controlled syringe system (Living Systems Instruments). Total volume of the suffusion system, reservoir, and vessel chamber was 100 ml. PSS flow through the system was set at 40 ml/min. Steady-state arteriole diameter and peak changes in diameter were measured with a videomicrometer (Microcirculation Research Institute, Texas A&M University Health Science Center, College Station, TX) and continuously recorded using a computerized data-acquisition system (model MP1000, Biopac, Goleta, CA). All vessels were allowed to stabilize for 60 min in oxygenated (21% O₂-5% CO₂-74% N₂) PSS warmed to 37°C. Only those vessels that developed spontaneous tone during the stabilization period were utilized. The passive diameter (PD) of each vessel was determined at the end the experiment by superfusing the vessel with calcium-free PSS. Arterioles from soleus muscle were only used in experiments comparing responses between exercise and sedentary control rats.

Experimental protocols. Dilation of skeletal muscle arterioles to stepwise changes in osmolality was assessed by adding increasing concentrations of D-glucose to the superfusion solution before (i.e., control) and after addition of a given inhibitor or blocker to the superfusion or perfusion solution. Under zero-flow conditions, vessels were superfused with each glucose concentration for 15 min. After the initial concentration-response curve was completed, arterioles were washed and allowed to equilibrate for 30 min. The responses to sucrose or mannitol were also assessed in a limited number of arterioles to verify that the responses to glucose were related to changes in solution osmolality and not specific for glucose. In a separate group of vessels, intraluminal PSS was replaced with PSS containing 5-80 mM D-glucose. Vessels were perfused with each concentration of glucose at a rate of 4 µl/min. Vessel diameter did not change significantly at this flow rate. When a steady-state diameter was obtained, flow was stopped to ensure that the dilation was due to the change in osmolality and independent of flow. Changes in diameter due to increases in intraluminal osmolality were calculated on the basis of vessel diameter after flow had been established.

7

4

5

7

5

6

6

5

Exercise regimen. Previous work from this laboratory demonstrated that mild exercise training alters vascular function and augments endothelial function in isolated microvessels from rat skeletal muscle (11, 28). Therefore, additional experiments were conducted to determine the effect of 4 wk of daily exercise on the dilator responses to hyperosmolality. Rats were randomly assigned to exercise (Ex) or sedentary control (Sed) groups. Ex rats ran on a motor-driven treadmill (model 4215, Quinton, Seattle, WA) 5 days/wk for 4 wk. Exercise intensity was progressively increased to 40 min/day at a treadmill speed of 28 m/min at a 2° grade by the beginning of the fourth week. Rats exercised at this intensity for 1 wk. Sed rats were handled daily but not made to run on the treadmill. Gracilis or soleus muscles were excised 24 h or more after the last exercise bout. Arterioles from the soleus muscle were included in these experiments because exercise training has been shown to elicit hemodynamic and metabolic changes in this muscle (1).

Drugs and solutions. The MOPS-buffered PSS contained (in mM) 145 NaCl, 5.0 KCl, 1.0 MgSO₄, 1.0 NaH₂PO₄, 2.0 CaCl₂, 5.0 glucose, 2.0 pyruvic acid, 3.0 MOPS, and 0.02 EDTA (pH 7.4). The normal PSS contained (in mM) 118.3 NaCl, 24 NaHCO₃, 4.7 KCl, 1.2 MgSO₄, 1.2 KH₂PO₄, 1.9 CaCl₂, 11.1 glucose, and 0.02 EDTA (pH 7.4). EGTA (1 mM) was substituted for CaCl₂ in calcium-free PSS; all other components remained the same. DMSO vehicle had no effect on vascular responses in this preparation. All other drugs were dissolved in distilled water. All chemicals were purchased from Sigma Chemical (St. Louis, MO).

Data analysis. Osmolality (in mosmol/kgH₂O) of the PSS was measured using a vapor pressure osmometer (Wescor, Logan, UT). Data are expressed as means ± SE. Only one arteriole from each animal was studied. Arteriolar responses to hyperosmolality before and after treatment were compared using repeated-measures analysis of variance followed by a modified Student's t-test with a Bonferroni correction for multiple comparisons. Comparisons among the effect of osmotic agents and between exercise and sedentary groups were made using a one-way ANOVA. Differences in baseline diameters were compared using Student's paired or unpaired t-tests when appropriate. Statistical significance was set at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hyperosmolality and gracilis muscle arteriole dilation. All isolated arterioles from rat gracilis muscles developed spontaneous tone in response to a perfusion pressure of 80 mmHg (n = 82). The active diameter was 62 ± 2 µm, which is ~53% of the PD (121 ± 2 µm). Arterioles dilated in response to increasing concentrations of glucose added to the superfusion solution (Fig. 1). The maximal concentration of glucose increased diameter to ~76% of PD. Similar changes in diameter were elicited in response to superfusion with sucrose or mannitol (Fig. 1), verifying that this response is due to a change in solution osmolality and is not specific for glucose.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Effect of hyperosmolality on rat gracilis muscle arteriolar diameter. Changes in diameter are in response to extraluminal glucose (n = 9), sucrose (n = 4), or mannitol (n = 7). All agents produced comparable changes in vessel diameter. Values are means ± SE. Baseline diameters: glucose, 53 ± 3 µm; sucrose, 65 ± 7 µm; mannitol, 69 ± 4 µm.

The role of the endothelium and endothelial factors. The effect of endothelium removal on arteriole dilation to glucose-induced dilation is shown in Fig. 2A. In intact arterioles, glucose elicited a concentration-dependent increase in vessel diameter, which was completely abolished after endothelium removal. The maximal concentration of glucose only elicited a 3 ± 2 µm change in diameter after endothelium removal compared with a 28 ± 3 µm change in diameter before endothelium removal. Endothelium removal did not significantly alter baseline vessel diameter (endothelium intact, 52 ± 4 µm; endothelium denuded, 46 ± 4 µm; P = 0.06).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Role of the endothelium and endothelial factors on changes in rat gracilis muscle arteriolar diameter in response to extraluminal glucose. A: effect of endothelium removal (n = 9). B: effect of N-nitro-L-arginine (L-NNA). C: effect of indomethacin (INDO). The endothelium was removed by injection of 1 ml of air into the lumen of each vessel. +E, endothelium intact; E, endothelium denuded. L-NNA (104 M; n = 6) or INDO (105 M; n = 6) was added to the superfusion solution 30 min before and throughout the experiment. Values are means ± SE. Baseline diameters: +E, 52 ± 4 µm; E, 46 ± 4 µm; L-NNA, 43 ± 3 µm before vs. 42 ± 6 µm after; INDO, 56 ± 4 µm before vs. 62 ± 4 µm after. * P < 0.05 compared with +E.

Effect of glibenclamide on hyperosmolality-induced dilation. To assess the role of K_ATP channels in the dilation to hyperosmolality, isolated arterioles were incubated with 10⁶ M glibenclamide. Extraluminal glibenclamide did not alter baseline vessel diameter and had no effect on the responses to extraluminal glucose (Fig. 3A). In contrast, intraluminal administration of glibenclamide significantly attenuated the dilator response to extraluminal glucose (Fig. 4A). For example, the changes in diameter to the two highest concentrations of glucose after glibenclamide perfusion were ~25-30% of the control response. The dilator responses to increasing concentrations of mannitol were also determined before and after glibenclamide perfusion (Fig. 4B). Mannitol superfusion elicited concentration-dependent vasodilation that was reduced to ~33% of the control response after intraluminal administration of glibenclamide (maximum change in diameter = 33 ± 3 vs. 12 ± 1 µm). Intraluminal glibenclamide did not change baseline vascular diameter in any of the vessels tested.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Effect of the ATP-sensitive potassium channel inhibitor glibenclamide (106 M) on changes in rat gracilis muscle arteriolar diameter in response to hyperosmolality (A; n = 7) and pinacidil (B; n = 5). Responses to extraluminal glucose or pinacidil were obtained before and after treatment with extraluminal glibenclamide (+Glibsup). Vessels were allowed to equilibrate for 30 min after glibenclamide was added to the superfusate solution. Values are means ± SE. Baseline diameters: A, 45 ± 3 µm before vs. 54 ± 7 µm after; B, 103 ± 3 µm before vs. 101 ± 7 µm after. * P < 0.05 compared with control.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Effect of the ATP-sensitive potassium channel inhibitor glibenclamide (106 M) on changes in rat gracilis muscle arteriolar diameter in response to hyperosmolality and pinacidil. A: extraluminal glucose and intraluminal glibenclamide (+Glibin; n = 9). B: extraluminal mannitol and +Glibin (n = 6). C: extraluminal pinacidil and +Glibin (n = 4). Vessels were allowed to equilibrate for 30 min after glibenclamide was added to the perfusate. Values are means ± SE. Baseline diameters: A, 53 ± 4 µm before vs. 48 ± 4 µm after; B, 69 ± 4 µm before vs. 65 ± 2 µm after; C, 79 ± 6 µm before vs. 88 ± 4 µm after. * P < 0.05 compared with glucose or mannitol.

Effect of glibenclamide on the responses to K_ATP channel openers. The responses to the K_ATP channel opener pinacidil were compared before and after intra- or extraluminal administration of glibenclamide to verify adequate blockade of K_ATP channels and to differentiate between activation of endothelial and smooth muscle K_ATP channels. The effects of 10⁶ M glibenclamide on the dilator responses to pinacidil are presented in Figs. 3B and 4C. Addition of glibenclamide to the superfusate significantly attenuated the response to pinacidil (Fig. 3B), whereas intraluminal glibenclamide had no effect on the dilator responses to pinacidil (Fig. 4C). Similar results were obtained with the K_ATP channel opener diazoxide (3 × 10⁵ M). Glibenclamide superfusion significantly inhibited the response to diazoxide (change in diameter = 32 ± 6 µm before vs. 6 ± 5 µm after; P < 0.05; n = 5). In contrast, intraluminal glibenclamide did not alter the response to diazoxide (change in diameter = 20 ± 6 µm before vs. 24 ± 3 µm after; P > 0.05; n = 4).

Effect of intra-arteriolar glucose on gracilis muscle arteriolar diameter. A comparison of the dilator responses to intra- and extraluminal glucose is presented in Fig. 5A. Dilator responses to intraluminal glucose were observed at lower concentrations (5 and 10 mM) and were significantly greater than the responses to extraluminal glucose (maximum change in diameter = 56 ± 3 vs. 34 ± 1 µm). Whereas extraluminal administration of 20 mM glucose (the lowest concentration used) had little effect on vascular diameter, perfusion with 20 mM glucose resulted in a near-maximal dilation (86 ± 3% of PD).

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Changes in rat gracilis muscle arteriolar diameter in response to intraluminal glucose. A: comparison between responses to intraluminal (Glucosein; n = 7) or extraluminal (Glucosesup; n = 9) glucose. B: effect of +Glibin (106 M) on the responses to Glucosein (n = 6). In A, intraluminal physiological saline solution (PSS) was replaced with PSS containing 5-80 mM D-glucose. Data for Glucosein are from Fig. 1. For B, vessels were perfused with glibenclamide and allowed to equilibrate for 30 min after administration of the inhibitor. Intraluminal PSS was replaced with PSS containing glibenclamide and 10, 20, or 60 mM D-glucose. In A and B, vessels were perfused with each concentration of glucose at a rate of 4 µl/min. Values are means ± SE. Baseline diameters: A, Glucosesup, 53 ± 3 µm; Glucosein, 62 ± 4 µm; B, 66 ± 4 µm before vs. 65 ± 8 µm after. * P < 0.05 compared with Glucosesup.  P < 0.05 compared with Glucosein.

Effect of daily exercise on hyperosmolality-induced dilation. The efficacy of the exercise program is demonstrated by differences in body weight and treadmill performance test times between Ex and Sed rats. Body weights for Ex rats were significantly less than those of age-matched Sed rats (397 ± 6 vs. 418 ± 5 g; P < 0.05), whereas wet heart weight-to-body weight ratios were comparable (3.39 ± 0.11 vs. 3.31 ± 0.16 g/kg). Total run times during the treadmill performance test were also significantly higher in Ex rats compared with Sed (59 ± 3 vs. 40 ± 1 min; P < 0.05). Body weights (292 ± 5 vs. 290 ± 5 g) and total run times (46 ± 1 vs. 44 ± 2 min) were comparable at the start of the exercise program.

7

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effect of 4 wk of daily exercise on changes in rat gracilis (A; n = 7 exercise, n = 6 sedentary) and soleus (B; n = 4 exercise, n = 5 sedentary) muscle arteriolar diameter in response to hyperosmolality. Values are means ± SE. Baseline diameters: A, exercise, 43 ± 2 µm; sedentary, 41 ± 3 µm; B, exercise, 50 ± 2 µm; sedentary, 48 ± 4 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There are several novel findings from this study. First, rat skeletal muscle arterioles dilate to increases in either intra- or extraluminal osmolality. However, arterioles were significantly more sensitive to increases in intraluminal osmolality. Second, the dilator responses to hyperosmolality are endothelium dependent but are not due to the release of endothelial nitric oxide or vasodilator prostaglandins. Third, intraluminal administration of glibenclamide inhibited the dilator responses to hyperosmolality, suggesting that these responses are mediated primarily by activation of K_ATP channels in endothelial cells. Collectively, these data suggest that hyperosmolality-induced dilation is due to endothelial cell hyperpolarization after activation of endothelial cell K_ATP channels. Furthermore, these data provide evidence that physiological increases in intraluminal osmolality that often accompany exercise can modulate vascular tone and imply that plasma hyperosmolality may contribute to functional hyperemia. Fourth, we also found that 4 wk of daily exercise do not affect the response to hyperosmolality.

Skeletal muscle blood flow is determined by a variety of factors but is invariably linked to the metabolic state of the muscle (4). Changes in skeletal muscle metabolism, like the increase that occurs during exercise, can be accompanied by increases in tissue and plasma osmolarity (16, 18, 19, 23). For example, Lundvall (16) concluded that increases in venous osmolarity correlated with increases in skeletal muscle blood flow during exercise and that the increases in venous osmolarity reflected changes in interstitial osmolarity. Scott et al. (23) also reported that vascular resistance decreased by 50% and venous osmolarity increased by 26 mosM in the exercising gracilis muscle. Furthermore, relaxation to hyperosmotic solutions containing glucose or sucrose has been observed in basilar (22) and coronary arteries (10) and coronary microvessels (7), suggesting that conducting and resistance vessels respond to increases in osmolarity. In the present study, isolated skeletal muscle arterioles were exposed to increasing concentrations of glucose, sucrose, or mannitol, raising osmolality to ~330 mosmol/kgH₂O. Each of these agents elicited significant and comparable increases in diameter (Fig. 1), suggesting that this dilation is likely due to an increase in osmolality and is not specific for glucose.

Although in this study the responses to hyperosmolality were comparable among the osmotic agents used, the site of administration (intraluminal vs. extraluminal) significantly affected the magnitude of the responses. Arterioles were much more sensitive to intraluminal glucose, dilating at concentrations well below the extraluminal concentration needed to elicit a similar change in diameter (Fig. 5). The response to a specific increase in osmolality was also greater when administered intraluminally. The difference between responses to intraluminal and extraluminal glucose is probably determined, in part, by the location of the "osmoreceptor," the cellular unit that detects the change in osmolality, and the concentration of glucose reaching that site. Because sucrose, mannitol, and, presumably, glucose do not freely enter or diffuse across cells (2, 8, 15), the vascular smooth muscle may act as a diffusion barrier. Therefore, the concentration of osmotic agents reaching the lumen of the vessel may be significantly less than that added to the superfusion solution. This corresponds to the data of Lundvall (16), who demonstrated that the increase in interstitial osmolarity was significantly greater than that in the venous effluent, yet changes in skeletal muscle blood flow correlated well with the increase in venous osmolarity. Nevertheless, the responses to intraluminal and extraluminal glucose indicate that skeletal muscle arterioles respond to physiological increases in osmolality that can modulate vascular tone. It is interesting to note that the range of intraluminal glucose concentrations (5-10 mM) that elicited dilation in skeletal muscle arterioles in this study is comparable to the changes in arterial blood osmolarity measured in the cat hindlimb during heavy exercise and well below the 22 mosmol/kgH₂O increase measured in humans during intense exercise. This range of intraluminal glucose is also similar to that observed in diabetes mellitus (20), the early stages of which are associated with small-vessel dilation (5, 30) due, in part, to the hyperosmotic effects of elevated blood glucose levels (15).

Several mechanisms have been proposed for the dilator responses to hyperosmolarity in vascular smooth muscle, including hyperpolarization due to changes in ion permeability (2, 8, 18), changes in calcium mobilization (10, 22), and impaired excitation-contraction coupling (9). Of these mechanisms, hyperpolarization due to a change in the potassium ion gradient has received the most experimental support. Recently, Ishizaka and Kuo (7) reported that endothelial cell K_ATP channels mediated the response to hyperosmolarity in coronary microvessels. This potential mechanism was addressed in the present study by examining the effect of the K_ATP channel inhibitor glibenclamide on the dilator responses to hyperosmolality in isolated skeletal muscle arterioles. Perfusion with glibenclamide significantly attenuated the responses to both intra- and extraluminal glucose and extraluminal mannitol (Figs. 4 and 5), confirming and extending the findings of Ishizaka and Kuo. These data suggest that K_ATP channels on endothelial cells mediate the dilation to hyperosmotic solutions in skeletal muscle arterioles.

The role of the endothelium and K_ATP channels in the response to hyperosmolality is supported by several additional findings from the present study. First, endothelium removal completely inhibited the response to glucose (Fig. 2), indicating the endothelial dependence of this response. In contrast, incubation with L-NNA or indomethacin had no effect on the responses to hyperosmolality of skeletal muscle (Fig. 2) and coronary arterioles (7). Therefore, the release of known endothelial factors, nitric oxide and vasodilator prostaglandins, does not appear to be involved in the arteriolar response to hyperosmolality. Although these observations does not eliminate the possibility that an unidentified endothelium-derived hyperpolarizing factor(s) mediates hyperosmolality-induced dilation, they also support the postulate that an "alternative" endothelium-dependent pathway mediates hyperosmolality-induced dilation. Support for the conclusion that this pathway involves endothelial cell K_ATP channels is based on the observation that intraluminal glibenclamide significantly attenuated the response to increases in intra- and extraluminal osmolality (Figs. 4 and 5), whereas superfusion of the vessels with the K_ATP channel inhibitor glibenclamide had no effect on hyperosmolality-induced dilation (Fig. 3). Furthermore, the lack of an effect of glibenclamide superfusion on the dilator responses to hyperosmolality also indicates that intraluminal glibenclamide was selectively inhibiting K_ATP channels on endothelial cells. In contrast, intraluminal glibenclamide had no effect on the response to extraluminal pinacidil, a response that was significantly inhibited by extraluminal glibenclamide, suggesting that hyperosmolality and the K_ATP channel opener pinacidil elicit dilation by activating different populations of K_ATP channels (endothelial cell vs. vascular smooth muscle). Ishizaka and Kuo (7) reported similar findings for glibenclamide on hyperosmolarity-induced dilation in isolated porcine coronary arterioles. In addition, they demonstrated that iberiotoxin and low concentrations of barium chloride did not alter the responses to glucose, indicating that calcium-sensitive and inward rectifying potassium channels were not involved in this response. Thus hyperosmolality-induced dilation appears to be mediated by activation of endothelial cell K_ATP channels and membrane hyperpolarization.

In a separate group of experiments, we examined the effect of daily exercise on the responses to hyperosmolality. Lundvall and colleagues (16, 17) demonstrated that osmolarity in both venous and arterial blood increases during muscle contraction in an intensity-dependent manner, with the peak increase occurring within ~5 min after the onset of exercise. This increase in venous osmolarity is well correlated with increases in skeletal muscle blood flow, suggesting that hyperosmolarity may contribute to exercise hyperemia (16-19, 23). One of the adaptations associated with exercise training is an increase in skeletal muscle blood flow at rest and during exercise (14), which is generally attributed to improved endothelial function (11, 13, 14, 28). Previously, functional adaptations of skeletal muscle arterioles to exercise have been demonstrated after 4-8 wk of daily exercise, including augmented responses to flow and/or shear stress and endothelium-dependent vasodilator agents (11, 13, 28). Sessa et al. (24) also demonstrated that endothelial nitric oxide synthase mRNA is elevated in canine aorta after 1 wk of daily exercise, suggesting that short-term mild exercise can alter endothelial function. Despite evidence for improved endothelial function after 4 wk of daily exercise (the present study and Refs. 11, 28), the responses to hyperosmolality were comparable in arterioles from Ex and Sed rats. These results imply that short-term daily exercise does not upregulate all endothelium-dependent responses in skeletal muscle arterioles. These adaptations may be dependent on the stimulus (shear stress vs. hyperosmolality) and the underlying mechanism (nitric oxide vs. K_ATP channels) involved. Unfortunately, as with previous studies (18, 19), local changes in osmolality during exercise were not measured directly due to methodological limitations. Therefore, the magnitude of the changes in osmolality during acute exercise and throughout the 4-wk exercise period are not known. Thus one potential explanation for the lack of an exercise effect on the responses to hyperosmolality is that the stimulus, namely the local increase in osmolality, diminished over the 4 wk of exercise due to other adaptations associated with daily exercise, such as hypervolemia.

In summary, the results of this study demonstrate that rat skeletal muscle arterioles dilate to physiological increases in intraluminal osmolality, whereas relatively large increases in extraluminal osmolality also dilated skeletal muscle arterioles. Responses were not augmented in arterioles from rats after 4 wk of mild daily exercise. Therefore, responses to hyperosmolality may not change with daily exercise as readily as those mediated by other vasodilator stimuli. The difference in sensitivity to intra- and extraluminal changes in osmolality emphasizes the physiological contribution of changes in intraluminal osmolality on the regulation of vascular tone. The results also indicate that the responses to hyperosmolality are endothelium-dependent and primarily mediated by K_ATP channels in endothelial cells. On the basis of the responses to intraluminal changes in osmolality, these data suggest that hyperosmolality-induced vasodilation may contribute to the hyperemia that often accompanies changes in skeletal muscle metabolism.