Journal of Applied Physiology

Nonuniform effects of endurance exercise training on vasodilation in rat skeletal muscle

R. M. McAllister, J. L. Jasperse, M. H. Laughlin


Endurance exercise training (Ex) has been shown to increase maximal skeletal muscle blood flow. The purpose of this study was to test the hypothesis that increased endothelium-dependent vasodilation is associated with the Ex-induced increase in muscle blood flow. Furthermore, we hypothesized that enhanced endothelium-dependent dilation is confined to vessels in high-oxidative muscles that are recruited during Ex. To test these hypotheses, sedentary (Sed) and rats that underwent Ex (30 m/min × 10% grade, 60 min/day, 5 days/wk, 8–12 wk) were studied using three experimental approaches. Training effectiveness was evidenced by increased citrate synthase activity in soleus and vastus lateralis (red section) muscles (P < 0.05). Vasodilatory responses to the endothelium-dependent agent acetylcholine (ACh) in situ tended to be augmented by training in the red section of gastrocnemius muscle (RG; Sed: control, 0.69 ± 0.12; ACh, 1.25 ± 0.15; Ex: control, 0.86 ± 0.17; ACh, 1.76 ± 0.27 ml·min−1·100 g−1·mmHg−1; 0.05 < P < 0.10 for Ex vs. Sed during ACh). Responses to ACh in situ did not differ between Sed and Ex for either the soleus muscle or white section of gastrocnemius muscle (WG). Dilatory responses of second-order arterioles from the RG in vitro to flow (4–8 μl/min) and sodium nitroprusside (SNP; 10−7 through10−4 M), but not ACh, were augmented in Ex (vs. Sed; P < 0.05). Dilatory responses to ACh, flow, and SNP of arterioles from soleus and WG muscles did not differ between Sed and Ex. Content of the endothelial isoform of nitric oxide synthase (eNOS) was increased in second-order, fourth-order, and fifth-order arterioles from the RG of Ex; eNOS content was similar between Sed and Ex in vessels from the soleus and WG muscles. These findings indicate that Ex induces endothelial adaptations in fast-twitch, oxidative, glycolytic skeletal muscle. These adaptations may contribute to enhanced skeletal muscle blood flow in endurance-trained individuals.

  • endothelium
  • acetylcholine
  • flow
  • shear stress
  • nitric oxide synthase

it is well established that endurance-type exercise training induces an increase in maximal cardiac output; this additional cardiac output is primarily directed to skeletal muscle during exercise (26). Greater muscle blood flow during exercise is likely also a consequence of adaptations to training in vasculature supplying muscle. These adaptations include increased vascularity and changes in vascular control (cf. Ref. 18). Regarding the latter possibility, reduced vasoconstriction and/or enhanced vasodilation could permit greater blood flow to skeletal muscle during exercise.

In recent years, there has been considerable interest in determining whether one aspect of vascular control, endothelium-dependent dilation, is enhanced by exercise training. Available data for vasculature supplying skeletal muscle in healthy humans and animals are equivocal. Whereas some studies have demonstrated greater endothelium-dependent vasodilation after training (3, 4, 15, 17, 23, 3032), others have failed to do so (7, 8, 12, 13, 17, 22). Few of these studies (7, 8, 12, 31) examined vessels from muscles that would be recruited during training sessions. This is an important consideration, because it is well established that training-induced skeletal muscle fiber adaptations are confined to those muscles recruited during training (5). Furthermore, none of these studies compared vascular adaptations to training in skeletal muscles of differing fiber-type composition. Because muscles of different fiber-type composition differ in how their myocytes adapt to exercise training (5), it may be that their vasculature also adapts differently to training. Indeed, it has been demonstrated that changes in vascularity with training are muscle fiber type-specific (cf. Ref. 18); whether training-induced alterations in vascular control vary similarly with muscle fiber type is unknown.

The purpose of this study was to test the hypothesis that endurance exercise training enhances endothelium-dependent vasodilation in skeletal muscle. Furthermore, we hypothesized that increased endothelium-dependent vasodilation is confined to muscles active during endurance exercise training; that is, muscles composed of slow-twitch oxidative (SO) and/or fast-twitch oxidative glycolytic (FOG) fibers. To test these hypotheses, three complimentary approaches were employed to determine muscle-specific vasodilatory responses in sedentary and endurance exercise-trained rats. These experimental approaches included the following: the isolated, perfused rat hindlimb preparation, to determine the integrated dilatory response of conductance- and resistance-type vessels in situ; the isolated resistance vessel preparation, to determine dilatory responses of resistance vessels in vitro; and Western blotting, to determine expression of key mediators of endothelium-dependent dilatory responses in conductance and resistance vessels.


Animals and exercise training program.

Six series of male Sprague-Dawley rats (Charles River or Harlan) were housed three per cage in rooms with controlled temperature (20–21°C) and light (12:12-h light-dark cycle). Rats weighed 150–175 g initially. They were randomly assigned to either a sedentary (Sed) or exercise-trained (Ex) group in each series. Endurance exercise training was done on motorized rodent treadmills. Ex rats ran at 30 m/min up a 10% grade for 60 min/day, 5 days/wk, over 8–12 wk after a 2- to 4-wk period of progressing to a duration of 60 min/day. Sed remained cage confined throughout the training period. Sed also served as a control group for a concurrent study involving vascular control in altered thyroid states (21a). Treadmill running was approved by the Institutional Animal Care and Use Committees of Kansas State University and the University of Missouri.

Training efficacy.

Effectiveness of treadmill running in promoting an exercise-trained state was assessed by determining citrate synthase activity in selected hindlimb skeletal muscles. Activity of citrate synthase, a marker enzyme for oxidative capacity, was determined spectrophotometrically according to the method of Srere (28).

Isolated, perfused rat hindlimb preparation.

Determination of vasodilatory responses in skeletal muscle in situ using the perfused rat hindlimb preparation has been described in detail previously (21). In brief, the left hindlimb was perfused via the femoral artery with an oxygenated perfusate that contained bovine erythrocytes (hematocrit ∼40%), bovine serum albumin (4 g/100 ml), bovine insulin (100 μU/ml), glucose (5.0 mM), and pyruvate (0.15 mM). Total inflow to the hindlimb was ∼9 ml/min. Perfusate exited the hindlimb via the femoral vein and was recirculated. A constant-flow, variable-pressure approach permitted identification of peak vasodilatory responses (see below). Perfusion pressure was monitored continuously (Digi-Med blood pressure analyzer), and muscle- and other tissue-specific flows were determined (radiolabeled microspheres; see below) at three time points: control, at peak effect (i.e., peak decrease in perfusion pressure) of the endothelium-dependent vasodilator acetylcholine (ACh), and at peak effect of the endothelium-independent vasodilator sodium nitroprusside (SNP). Stock solutions of ACh (1.10 × 10−2 M) and SNP (2.68 × 10−2 M) were infused, at a rate of 0.15 ml/min, into the perfusate immediately proximal to the femoral arterial catheter. Infusion of these stock solutions at a rate of 0.15 ml/min, with a total inflow to the hindlimb of ∼9 ml/min, establishes femoral arterial concentrations of 1–2 × 10−4 M and 4–5 × 10−4 M for ACh and SNP, respectively (21).

Muscle- and other tissue-specific flows were determined using radiolabeled microspheres (46Sc, 85Sr, 113Sn, or 141Ce; New England Nuclear) at the time points listed above, as done previously (21). All muscles and other tissues of the left hindlimb were dissected, weighed, and counted in a gamma counter (Packard) after each experiment. Flows were calculated as follows: Math where counts/min (total) refers to the sum of counts per minute for all hindlimb muscles and other tissues, and flow (total) refers to total inflow to the hindlimb.

Isolated resistance vessel preparation.

Resistance vessels from the first-order SO-type soleus (1A arterioles) and second-order FOG-type gastrocnemius muscles (red section, RG; 2A arterioles), as well as the fast-twitch glycolytic (FG)-type white section of gastrocnemius muscle (WG; 2A arterioles), were dissected and studied as described in detail previously (12). The 1A arteriole supplying both the red and white sections of gastrocnemius muscle was also studied. As described previously (19), a 1A arteriole was defined as the first intramuscular resistance vessel; daughters of this 1A arteriole were defined as 2A arterioles. After dissection, a vessel was transferred to a chamber containing ice-cold physiological saline solution (PSS; see below). PSS was equilibrated with room air. Both ends of the vessel were cannulated with glass pipettes (30- to 40-μm diameter) and secured to the pipettes with suture. Cannulating pipettes were connected to independent reservoir systems. Intraluminal pressure was gradually increased to 80 cmH2O (1A arterioles) or 60 cmH2O (2A arterioles) over 30 min, and the vessel was gradually warmed to 37°C over 60 min. The vessel was visualized with an inverted light microscope (Nikon Diaphot 200), and inside diameter was monitored and recorded using a video caliper (Microcirculation Research Institute, Texas A&M University) and data-acquisition system (MacLab), respectively.

PSS contained (in mM) 145.0 NaCl, 4.7 KCl, 2.0 CaCl2, 1.2 MgSO4, 1.2 NaH2PO4, 5.0 glucose, 2.0 pyruvate, 0.02 EDTA, and 3.0 MOPS, pH 7.4. Dose-response relationships for ACh (10−9 to 10−4 M, in whole-log increments) and SNP (10−9 to 10−4 M, in whole-log increments) were determined by adding these agents to the vessel chamber in a cumulative fashion. The dose-response relationship for flow (change in pressure, 0–20 cmH2O, in 2- to 5-cmH2O increments) was also determined. Flow rates associated with the change in pressure values were determined for all cannulating pipettes used in experiments. Maximal diameter of a vessel was determined in Ca2+-free PSS at the end of an experiment.

Western blotting.

Protein contents of the endothelial isoform of nitric oxide (NO) synthase (eNOS) and the Cu/Zn-dependent isoform of superoxide dismutase (SOD-1), key mediators of endothelium-dependent dilation, were determined in multiple vessel types from Sed and Ex rats. We selected eNOS since previous studies by our laboratory (11, 12, 19, 21) and others (15, 17, 3032) indicated that NO accounts for a substantial portion of endothelium-dependent vasodilation in rodent skeletal muscle. Conductance-type vessels that perfuse hindlimb skeletal muscle were examined, including the abdominal aorta, as well as the iliac, femoral, and popliteal arteries. The mesenteric and renal arteries were also examined. Resistance-type vessels studied included soleus muscle feed artery and 1A through 4A arterioles, gastrocnemius muscle feed artery and 1A arteriole, and 2A through 5A arterioles from both the RG and WG. For conductance vessels, eNOS and SOD-1 protein contents in individual Sed and Ex vessels were determined; for resistance vessels, vessels from several Sed and several Ex rats were pooled before determination of eNOS and SOD-1 protein expression. Pooling of resistance vessels was necessary to obtain sufficient total protein for Western blotting.

Vessels were placed in Laemmli buffer solution, and they were touch mixed, centrifuged briefly (14,000 rpm × 1 min), and boiled for 2 min (3 cycles). After the second cycle, samples were sonicated (5 s). For each sample, either 5.0 μg (conductance vessels) or 2.0 μg (pooled resistance vessels) of total protein were loaded onto a 4–12% Tris gradient gel (Invitrogen). When sample volume permitted, samples were loaded in duplicate. Samples were subjected to electrophoresis under reducing conditions (Invitrogen) and then transferred to polyvinylidene difluoride membrane (Hybond-ECL, Amersham). Membranes were blocked for 1 h in 5.0% (wt/vol) nonfat milk solution (in Tris-buffered saline-Tween 20; TBST), and then incubated overnight with a primary antibody against either eNOS (Transduction Laboratories; 1:1,000 and 1:833 in 5.0% nonfat milk-TBST solution for conductance and resistance vessels, respectively) or SOD-1 (Stressgen; 1:1,000 in 5.0% nonfat milk-TBST solution for both conductance and resistance vessels). After washing, membranes were incubated for 1 h with a secondary antibody (horseradish peroxidase-conjugated anti-mouse, Sigma; 1:2,500 in 5.0% nonfat milk-TBST solution). After additional washing, enhanced chemiluminescence (Amersham) was used to visualize proteins on film, and protein contents were determined using densitometry software (National Institutes of Health).

Statistical analysis.

Vascular conductance was used as an index of vasomotor function in perfused rat hindlimb experiments because of its linear relationship with flow (20). Conductance was calculated as the quotient of flow (in ml·min−1·100 g−1) and perfusion pressure (in mmHg). Data for muscles and other tissues of the upper hindlimb were necessarily included in all flow calculations. They are not, however, presented in results because variable femoral arterial catheter placement did not permit reliable flow determinations in the upper portion of the hindlimb. Conductance data for each individual muscle/other tissue were analyzed by using two-way analysis of variance, with repeated measures across experimental time points (i.e., control, ACh, SNP; Ref. 29). The Tukey test was used for post hoc analysis (29). Other data related to perfused rat hindlimb experiments (i.e., mass and perfusion variable data) were analyzed using the unpaired t-test (29), as were citrate synthase activity data for ascertaining training efficacy.

Diameter data from isolated resistance vessel experiments were expressed, as done previously (11, 12, 19), as percent possible dilation calculated as follows: Math

Data are presented in this form to control for any baseline or maximal diameter differences between Sed and Ex. Data from isolated resistance vessel experiments were analyzed by using two-way analysis of variance, with repeated measures across doses and with contrasts used for post hoc analysis (29).

eNOS and SOD-1 protein content data for conductance vessels were analyzed by expressing each Ex rat's densitometric value on a given immunoblot relative to the mean densitometric value for all Sed rats on that immunoblot. Densitometric values for all Sed rats on a given immunoblot were also expressed relative to the mean value for Sed on that immunoblot, and relative values for Sed and Ex were compared by using the unpaired t-test (29). If samples were Western blotted in duplicate, mean values for a given rat were used in data analysis. For resistance vessels, as noted above, samples from Sed and from Ex rats were pooled before Western blotting. In one series, vessels from all Sed and all Ex rats (n = 5–9 each) were pooled; in a second series, vessels from three Sed and three Ex rats were pooled before Western blot analyses. Statistical analysis of resistance vessel data was otherwise conducted as per that of conductance vessel data.

All data are presented as means ± SE. P < 0.05 was considered significant for all analyses.


Training efficacy.

Certain hindlimb muscles of Ex exhibited greater citrate synthase activity than those of Sed, reflecting training effectiveness (Fig. 1). Muscles primarily composed of SO fibers (soleus muscle; Ref. 2) and FOG fibers (red section of vastus lateralis muscle; Ref. 2), but not one composed of FG fibers (white section of vastus lateralis muscle; Ref. 2), had higher citrate synthase activity in Ex than Sed in all series of experiments. In one series, because of experimental constraints, the vastus intermedius was sampled instead of the soleus muscle. This muscle also consists primarily of SO fibers (2). Similar to the soleus muscle in other series, citrate synthase activity was increased by training in this muscle (Sed, 41.0 ± 1.0 μmol·min−1·g−1, n = 10; Ex, 52.9 ± 2.0 μmol·min−1·g−1, n = 9; P < 0.05).

Fig. 1.

Citrate synthase activity in selected hindlimb muscles. Values are means ± SE; n for sedentary (Sed; open bars) and exercise trained (Ex; solid bars), respectively, are 26 and 25 for soleus muscle, 34 and 33 for red section of vastus lateralis muscle [VL (red)], and 35 and 32 for white section of vastus lateralis muscle [VL (white)]. *Different from Sed, P < 0.05.

Vasodilatory responses in situ.

Perfused hindlimb mass was less in Ex than in Sed (Table 1). Hematocrit, total inflow to the hindlimb, and perfusion pressure under control conditions, however, were similar between groups (Table 1). The decrease in pressure with ACh was greater in Ex than Sed, whereas that with SNP did not differ between groups (Table 1).

View this table:
Table 1.

Masses and perfusion conditions for perfused hindlimb experiments

Figure 2 presents conductance under control conditions, and with ACh and SNP administration, for selected hindlimb muscles and muscle sections. The RG (Fig. 2A), mixed gastrocnemius (MG; Fig. 2C), and soleus (Fig. 2D) muscles generally exhibited vasodilation, reflected by an increase in conductance, to both agents irrespective of training status. One exception to this pattern was that ACh elicited vasodilation in the RG and MG muscle sections of Ex but not Sed. For the MG, the difference between Sed and Ex for the response to ACh was significant; for the RG, the Sed-Ex difference was of borderline significance (0.05 < P < 0.10). The WG, on the other hand, did not exhibit vasodilation to either agent in Sed or in Ex (Fig. 2B). Values for conductance in other muscles and tissues of the lower hindlimb are presented in Table 2. Notably, significant vasodilation to ACh was demonstrated in Ex, but not Sed, plantaris and tibialis posterior muscles. The Sed-Ex difference in response to ACh was of borderline significance for the plantaris muscle (0.05 < P < 0.10).

Fig. 2.

Conductance (Cond) in red section of gastrocnemius muscle (RG; A), white section of gastrocnemius muscle (WG; B), mixed section of gastrocnemius muscle (MG; C), and soleus muscle (Sol; D). Values are means ± SE; n = 14 and 13 for Sed (control; open bars) and Ex (solid bars), respectively. ACh, acetylcholine; SNP, sodium nitroprusside. Note different scaling of y-axes for WG (B) and MG (C). *Different from control within same group, P < 0.05. †Different between Sed and Ex, P < 0.05.

View this table:
Table 2.

Muscle- and other tissue-specific conductance values for perfused hindlimb experiments

Vasodilatory responses in vitro.

Maximal diameters of gastrocnemius muscle 1A arterioles from Ex were greater than those from Sed (Sed, 224 ± 5 μm, n = 8; Ex, 257 ± 10 μm, n = 11; P < 0.05); other vessel types examined did not differ between Sed and Ex in maximal diameter (RG 2A arterioles: Sed, 134 ± 13 μm, n = 9; Ex, 146 ± 5 μm, n = 10; WG 2A arterioles: Sed, 134 ± 8 μm, n = 10; Ex, 146 ± 8 μm, n = 10; soleus 1A arterioles: Sed, 173 ± 8 μm, n = 18; Ex, 160 ± 6 μm, n = 21).

Figures 35 illustrate dose-dependent dilatory responses to ACh, flow, and SNP, respectively, of resistance vessels from the soleus (1A arterioles), RG (2A arterioles), and WG (2A arterioles) muscles and muscle sections. Responses of 1A arterioles supplying both the red and white sections of the gastrocnemius muscle are also shown. Although vasodilatory responses were generally similar between Sed and Ex, 2A arterioles from the RG of Ex exhibited greater dilation in response to both flow (Fig. 4; P < 0.05 for interaction effect of group and dose) and SNP (Fig. 5; P < 0.05 for interaction effect of group and dose). In addition, 1A arterioles from the gastrocnemius muscle of Ex dilated less in response to flow (Fig. 4; P < 0.05 for both group effect and interaction effect of group and dose).

Fig. 3.

Dose-dependent dilatory responses to ACh of second-order (2A) arterioles from RG muscle (RG 2A; A), 2A arterioles from WG muscle (WG 2A; B), gastrocnemius muscle first-order (1A) arterioles (G 1A; C), and soleus muscle 1A arterioles (Sol 1A; D). Values are means ± SE; n for Sed and Ex as per text of results. Values for Sed and Ex do not differ at any dose for any vessel type.

Fig. 4.

Dose-dependent dilatory responses to flow of RG 2A (A), WG 2A (B), G 1A (C), and Sol 1A (D). Values are means ± SE; n for Sed and Ex as per text of results. Scaling of x-axis is different for C because of larger cannulating pipettes required for G 1A arterioles. *Different between Sed and Ex, P < 0.05.

Fig. 5.

Dose-dependent dilatory responses to SNP of RG 2A (A), WG 2A (B), G 1A (C), and Sol 1A (D). Values are means ± SE; n for Sed and Ex are per text of results. *Different between Sed and Ex, P < 0.05.

eNOS and SOD-1 protein contents.

Figure 6A presents data for eNOS protein content in conductance-type vessels. eNOS protein expression was greater in Ex than Sed in the abdominal aorta, femoral artery, and renal artery (P < 0.05). eNOS expression did not differ between Sed and Ex in the iliac, popliteal, or mesenteric arteries. For resistance-type vessels from the soleus (Fig. 7A), eNOS protein content in Ex was similar to that in Sed in all vessel types examined. In the gastrocnemius muscle (Fig. 8A), on the other hand, 2A, 4A, and 5A arterioles in the RG exhibited significantly greater eNOS protein content in Ex than Sed, and the Sed-Ex difference was of borderline significance for 3A arterioles (0.05 < P < 0.10). The 1A arteriole supplying both the red and white sections of the gastrocnemius muscle also demonstrated borderline greater eNOS expression in Ex. No Sed-Ex differences were observed in WG vessels. SOD-1 protein content did not differ between Sed and Ex in any conductance vessel examined (Fig. 6B). Expression of SOD-1 protein was greater in Ex than Sed only in the feed artery supplying the gastrocnemius muscle (Fig. 8B). For all other resistance vessels, SOD-1 protein content either was not different between Sed and Ex or was reduced in Ex relative to Sed (WG4A; Figs. 7B and 8B).

Fig. 6.

A: endothelial nitric oxide synthase (eNOS) protein content in conductance vessels. Values are means ± SE; n = 16 and 16 for Sed (open bars) and Ex (solid bars), respectively, abdominal aortas (Aorta); n = 8 and 9 for Sed and Ex, respectively, iliac arteries (Iliac); n = 17 and 18 for Sed and Ex, respectively, femoral (Fem) and popliteal (Poplit) arteries; n = 10 and 9 for Sed and Ex, respectively, mesenteric (Mes) arteries; n = 8 and 9 for Sed and Ex, respectively, renal arteries (Renal). *Different from Sed, P < 0.05. B: Cu/Zn-dependent superoxide dismutase (SOD-1) protein content in conductance vessels. Values are means ± SE; n for Sed (open bars) and Ex (solid bars), respectively, as per A. Sed and Ex are not different for any vessel type.

Fig. 7.

A: eNOS protein content in soleus (S) resistance vessels. SFA, soleus feed artery. Values are means ± SE; for Sed and Ex, n = 4 groups of pooled vessels each. B: SOD-1 protein content in resistance vessels. Values are means ± SE; number of of pooled vessels as per A. There are no differences between Sed and Ex for any vessel type for either eNOS or SOD-1.

Fig. 8.

A: eNOS protein content in gastrocnemius (G) resistance vessels. GFA, gastrocnemius feed artery. Values are means ± SE; for Sed and Ex, n = 4 groups of pooled vessels each. 3A, third order; 4A, fourth order; 5A, fifth order. Differences between Sed and Ex are significant (P < 0.05) for RG 2A, 4A, and 5A arterioles; borderline significant differences (0.05 < P < 0.10) also indicated. B: SOD-1 protein content in resistance vessels. Values are means ± SE; numbers of groups of pooled vessels are as per A. Differences between Sed and Ex are significant (P < 0.05) for GFA and WG 4A; borderline significant differences (0.05 < P < 0.10) are also indicated.


This study found that endurance exercise training led to nonuniform augmentation of endothelium-dependent vasodilation in skeletal muscle. Increased vasodilation was confined to muscles and muscle sections that consist primarily of FOG fibers. Endothelial adaptations to training were apparent using three different, yet complimentary, experimental approaches. These findings contribute to our understanding of how vasculature supplying skeletal muscle adapts to exercise training.

Training efficacy.

The endurance exercise training program employed in this study was chosen to model programs used with humans in the prevention of and/or rehabilitation from chronic disease (1). Effectiveness of the training program was indicated by increased skeletal muscle citrate synthase activity in Ex, a hallmark of the exercise-trained state. Elevated citrate synthase activity was observed only in soleus and vastus lateralis (red section) muscles (i.e., SO- and FOG-type muscles, respectively; Fig. 1), as would be predicted from muscle recruitment patterns during endurance exercise (5).

Endothelium-dependent vasodilation: in situ responses.

As we have previously reported for normal (untrained) rats (19, 21), endothelium-dependent vasodilation was exhibited by high-oxidative, but not low-oxidative, skeletal muscle in Sed. In Ex, the decrease in perfusion pressure with administration of the endothelium-dependent vasodilating agent ACh was twice that in Sed, suggesting that endothelium-dependent vasodilation was augmented by training. Our experimental preparation permitted determination of muscle-specific responses. Conductance data indicated that significant vasodilation in response to ACh occurred in certain hindlimb muscles and muscle sections of Ex only, including the RG and MG (Fig. 2, A and C), plantaris (Table 2), and tibialis posterior (Table 2). In the MG, ACh-induced vasodilation in Ex was significantly greater than in Sed; in two other muscles (RG, plantaris), the difference between Sed and Ex was of borderline significance. Most of these muscles and muscle sections consist of a majority of FOG fibers (2). Although the MG is primarily composed of FG (i.e., low oxidative) fibers, it nonetheless contains significant populations of SO and FOG fibers (total, 35%; Ref. 2). It is therefore uncertain whether the training-induced increase in conductance in this muscle section during ACh administration should be assigned to its high- and/or low-oxidative portions.

Although previous studies have not examined effects of exercise training on vasodilatory responses in muscles of different fiber-type composition, findings from the present study are consistent with available data for specific muscles examined in earlier studies. A study involving arterioles isolated from the plantaris muscle demonstrated training-induced enhancement of endothelium-dependent, flow-induced dilation (31), consistent with our findings for the entire vascular bed of this muscle (Table 2). In a study involving isolated soleus muscle feed arteries, we found that endothelium-dependent dilatory responses to both pharmacological (ACh) and hemodynamic stimuli (increases in flow) were similar between sedentary and exercise-trained rats (12). Similar ACh-induced changes in conductance within the soleus muscle for Sed and Ex (Fig. 2D), which represent the integrated response of all vessels supplying this muscle, are in agreement with our previously reported data for the feed artery supplying this muscle (12). Furthermore, although the soleus muscle consists exclusively of high-oxidative fibers (2), our in situ, in vitro, and protein content data uniformly indicate that endurance-type training does not induce changes in endothelium-dependent vasodilatory capacity in this muscle. As we have speculated previously, the soleus muscle's predominantly slow-twitch contractile character (i.e., preponderance of SO fibers) and, therefore, high level of recruitment under resting conditions (11) may preclude endothelial adaptations to training. The attendant high blood flow at rest in this muscle limits the fold increase in flow during exercise to 2–3; in contrast, a FOG-type muscle such as the RG exhibits a 20-fold or more increase in blood flow during exercise (cf. Ref. 18). Magnitude of the exercise-induced increase in blood flow may be a key determinant of training-induced vascular adaptations.

Endothelium-dependent vasodilation: in vitro responses.

Nonuniform adaptations to exercise training were also found using the isolated resistance vessel preparation. For vessels from the soleus and WG muscles, the absence of training-induced adaptations in vitro is consistent with our in situ and protein content data. Flow-induced dilation of 2A arterioles from the FOG-type RG was enhanced with endurance-type training. Enhanced flow-induced dilation of 2A arterioles from the RG is consistent with data for mediators of endothelium-dependent dilation (see below). Because responses to SNP of 2A arterioles from the RG were also enhanced, we cannot exclude the possibility that increased sensitivity of vascular smooth muscle to NO was in part responsible for augmented flow-induced dilation of these arterioles. Augmented NO formation in the endothelium, coupled with increased sensitivity to NO, could result in enhanced flow-induced dilation. A training-induced increase in NO sensitivity has been reported previously for the coronary circulation (10).

ACh-induced dilation of 2A arterioles from the RG was similar in Sed and Ex. The borderline enhancement of ACh-induced vasodilation observed in the RG in situ cannot, therefore, be attributed to adaptations in 2A arterioles, but rather must be explained by adaptations in other branches of its arterial microcirculation. Because flow-induced dilation of 2A arterioles from the RG was enhanced by exercise training, it is of interest that our data for flow- and ACh-induced dilation of RG 2A arterioles are seemingly inconsistent with each other. Both flow and ACh induce vasodilation via the endothelium. It has, however, been demonstrated in both cultured endothelial cells (16) and isolated arterioles (25) that pharmacological agents and increased flow or shear stress induce endothelium-derived NO formation via different mechanisms. Agents such as ACh and bradykinin are Ca2+ dependent, whereas the endothelial response to a sustained increase in flow or shear stress is Ca2+ independent (16, 25). Adaptations of the endothelium to exercise training may involve increased flow or shear stress, and this may contribute to the findings that Sed-Ex differences were revealed by increased flow but not by ACh in vitro. It may be that training-induced endothelial adaptations are specific to the pathway(s) linking a given endothelial stimulus (i.e., pharmacological or physiological) to eNOS.

eNOS and SOD-1 protein contents.

Our data for eNOS protein content are generally consistent with our vasodilatory data obtained in situ and in vitro. For resistance-type vessels of Ex, only those from within the FOG-type RG exhibited increased eNOS protein expression. In addition, a resistance vessel supplying both the red and white sections of gastrocnemius muscle (i.e., 1A arteriole) demonstrated slightly greater eNOS protein expression in Ex than Sed. These findings for gastrocnemius muscle vessels are consistent with the well-established increase in blood flow to the red, but not white, section of this muscle that would occur during training sessions (cf. Ref. 18) and that may be the stimulus for vascular adaptions to training. Increases in eNOS protein content also occurred in some conductance-type vessels supplying hindlimb muscles, including the abdominal aorta and femoral artery. Conductance vessels would also be expected to experience increased blood flow during training. Our laboratory (3) and others (6, 14) have previously reported that endurance exercise training significantly increases expression of eNOS protein in rodent aorta. These changes in eNOS expression do not exclude the possibility that increased formation of vasodilatory prostaglandins and/or endothelium-derived hyperpolarizing factor(s) also contribute to augmented endothelium-dependent vasodilation. We elected to focus on eNOS because previous work indicted that increased vascular NO formation is associated with exercise training (15, 19, 30, 32). We also examined SOD-1 protein content, because increased expression of this protein could enhance NO bioavailability in the vascular wall (cf. Ref. 27). In contrast to a previous study examining porcine aorta (27), but in agreement with another involving rodent aorta (6), we did not observe increased SOD-1 protein content in most vessels, either conductance- or resistance-type, of Ex.

Our findings are also consistent with those from studies evaluating effects of exercise training on skeletal muscle vascularity. Gute and colleagues (9), for example, found that endurance-type training induced increases in capillarity in the RG and MG but not in the WG or soleus muscle. Although signals present during training leading to increased vascularity in skeletal muscle may be different from those associated with an altered vascular control mechanism such as endothelium-dependent dilation, the presence of such signals is likely associated with muscle recruitment during training (cf. Ref. 18). Thus the RG and MG, which are active during endurance exercise training and consist largely of FOG muscle fibers (2, 5), exhibit increases in both endothelium-dependent vasodilation (present study) and vascularity (9).

Study limitations.

Certain features of vascular adaptation to exercise training appear to differ between rodents and humans. First, a majority of studies suggest that endothelium-dependent vasodilation improves with training in humans only in the setting of disease (cf. Ref. 24), perhaps because impaired endothelial function is associated with a variety of chronic diseases (e.g., atherosclerosis). Thus training-induced improvement in endothelium-dependent vasodilation may actually represent normalization of impaired dilation. In rodents, as demonstrated by the present study and others, endothelium-dependent vasodilation is augmented by training even in the absence of disease. Second, vascular adaptations in humans may not be confined to vessels supplying muscle active during exercise training sessions. Reports of augmented endothelium-dependent dilation of upper extremity vessels with predominantly lower extremity exercise training exist (cf. Ref. 24). In the present study, we did not observe adaptations in vessels from muscles that were likely inactive during endurance exercise training (e.g., WG; Ref. 5). Furthermore, even in active muscles that did exhibit training-induced adaptations, a nonuniformity of these adaptations was observed. The RG was characterized by significantly greater eNOS content in 2A, 4A, and 5A arterioles, and borderline greater eNOS content in 3A arterioles (Fig. 8A). These observations indicate that caution is necessary when results of rodent studies of vascular adaptation to exercise training are applied to humans.


The present study demonstrates, via three different experimental approaches, that endothelial adaptations occur in vessels within skeletal muscles comprised primarily of FOG fibers as a result of endurance-type exercise training. If endothelium-dependent vasodilation contributes to exercise hyperemia, these adaptations may permit greater muscle blood flow during exercise. This possibility is consistent with the finding that maximal muscle blood flow is increased after a period of exercise training, and it could contribute to increases in aerobic capacity and performance that are characteristic of endurance-trained individuals (26).


This work was supported by National Heart, Lung, and Blood Institute Grants HL-57226 (R. M. McAllister) and HL-36088 (M. H. Laughlin) and by American Heart Association-Kansas Affiliate Grant AHA-KS-98-GB-25 (R. M. McAllister).


The technical assistance of Ivelisse Albarracin, Molly Edmonds, Ann Melloh, Pam Thorne, and Mike Zbreski is gratefully acknowledged. The efforts of numerous students in the exercise training of rats are also acknowledged. In addition, the provision of bovine blood by the staffs at Alta Vista (Kansas) Locker, Burlingame (KS) Locker and Meat Market, and Clay Center (Kansas) Locker is acknowledged.


  • 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. Section 1734 solely to indicate this fact.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 21a.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
  27. 26.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
View Abstract