Physiological Genomics


Lau, Kim S., Robert W. Grange, Eiji Isotani, Ingrid H. Sarelius, Kristine E. Kamm, Paul L. Huang, and James T. Stull. nNOS and eNOS modulate cGMP formation and vascular response in contracting fast-twitch skeletal muscle. Physiol. Genomics 2: 21–27, 2000.—Nitric oxide (NO) from Ca2+-dependent neuronal nitric oxide synthase (nNOS) in skeletal muscle fibers may modulate vascular tone by a cGMP-dependent pathway similar to NO derived from NOS in endothelial cells (eNOS). In isolated fast-twitch extensor digitorum longus (EDL) muscles from control mice, cGMP formation increased ∼166% with electrical stimulation (30 Hz, 15 s). cGMP levels were not altered in slow-twitch soleus muscles. The NOS inhibitor Nω-nitro-l-arginine abolished the contraction-induced increase in cGMP content in EDL muscles, and the NO donor sodium nitroprusside (SNP) increased cGMP content ∼167% in noncontracting EDL muscles. SNP treatment but not electrical stimulation increased cGMP formation in muscles from nNOS−/− mice. cGMP formation in control and stimulated EDL muscles from eNOS−/− mice was less than that obtained with similarly treated muscles from control mice. Arteriolar relaxation in contracting fast-twitch mouse cremaster muscle was attenuated in muscles from mice lacking either nNOS or eNOS. These findings suggest that increases in cGMP and NO-dependent vascular relaxation in contracting fast-twitch skeletal muscle may require both nNOS and eNOS.

  • endothelial nitric oxide synthase
  • neuronal nitric oxide synthase
  • arteriolar relaxation

nitric oxide synthases (NOS) convert l-arginine to l-citrulline and NO in the presence of molecular oxygen and other cofactors (19). Several isoforms of NOS have been identified, including neuronal NOS (nNOS), endothelial NOS (eNOS), and inducible NOS (iNOS). NO generated in endothelial cells initiates a vasodilatory cascade in vascular smooth muscle tissue by activating soluble guanylyl cyclase (sGC) to promote cGMP formation (18). Elevated cGMP activates cGMP-dependent protein kinase (PKG), leading to decreased intracellular calcium concentration ([Ca2+]i) and subsequent relaxation. The resulting vasodilation increases blood flow in the affected vascular bed (18).

For exercise to be sustained, the necessary increase in blood flow to contracting skeletal muscle can be achieved in part by a decrease in skeletal muscle vascular resistance (1). The exercise-induced hyperemic response of skeletal muscle vasculature or active/functional hyperemia is a well-established phenomenon; however, the precise mechanisms that regulate skeletal muscle perfusion during exercise remain incompletely understood. Factors implicated as vasoactive agents to mediate functional hyperemia include low oxygen tension (27), adenosine (28), potassium, pH, magnesium, and osmolarity. Although these factors all have vasodilatory effects in contracting skeletal muscle, their precise contributions to mediating the exercise-induced hyperemia are not well defined.

NO produced by eNOS and nNOS may also contribute to functional hyperemia (12, 16), because NO is formed in contracting skeletal muscle (2). After stimulation of cell surface receptors or mechanical shear stress in contracting muscle, the vascular endothelium produces NO as increased [Ca2+]i leads to activation of Ca2+/calmodulin-dependent eNOS (24). Additionally, fast-twitch skeletal muscle fibers contain nNOS colocalized to the sarcolemma with dystrophin (4, 5), but the contribution of nNOS-derived NO to functional hyperemia is less clear. We considered the possibility that NO produced from Ca2+/calmodulin-dependent nNOS in contracting fast-twitch skeletal muscle fibers could diffuse to adjacent vascular smooth muscle cells to elicit a localized relaxation response via the NO/cGMP-mediated pathway. To this end, we have previously shown that cGMP formation is increased in contracting extensor digitorum longus muscle (EDL) in vitro from C57Bl6 but not mdx mice (16). In Duchenne's muscular dystrophy (DMD) and in dystrophin-deficient mdx mice, both nNOS expression and activity are greatly reduced (5). These reductions may explain both the decrease in cGMP and the attenuated arteriolar dilation observed in contracting muscles from mdx mice compared with C57Bl6 mice (16). In addition, NO produced by nNOS in skeletal muscle antagonizes sympathetic α-adrenergic vasoconstriction during muscle contraction in C57Bl6 but less so in mdx mice (32). This sympatholytic event appears to be fiber-type specific, because it occurs in contracting fast-twitch, glycolytic but not slow-twitch, oxidative muscles (31). Collectively, these findings (16, 31, 32) suggest that nNOS may play a specific functional role in the exercise-induced hyperemic response of contracting fast-twitch skeletal muscle fibers. However, in addition to the decrease in nNOS expression in mdx mice, the possibility that dystrophin-deficient muscles have coexisting derangements that interfere with the NO/cGMP signaling pathway could not be excluded.

In this report, we test the hypothesis that NO produced specifically by nNOS in contracting skeletal muscle stimulates the formation of cGMP and that this increase in cGMP is associated with a hyperemic response. We used fast-twitch (EDL and cremaster) muscles isolated from C57Bl6 mice as well as nNOS-deficient (nNOS−/−) and eNOS-deficient (eNOS−/−) mice to measure both cGMP formation and arteriolar relaxation in response to electrical stimulation. The use of nNOS−/− and eNOS−/− mice has the advantage of allowing assessment of individual contributions of nNOS or eNOS to cGMP formation and arteriolar responsiveness to skeletal muscle contraction. We also evaluated the potential signaling effects of NO on cGMP formation in slow-twitch soleus muscles, which have more eNOS and less nNOS compared with EDL muscles. We demonstrate that cGMP formation in contracting EDL muscles is nNOS dependent, but lack of eNOS also attenuates the response. We also show that the skeletal muscle vasodilatory response to contraction requires both nNOS and eNOS.


C57Bl6 mice (12–15 wk old) were purchased from Jackson Laboratories. The eNOS−/− (10) and nNOS−/− (11) mice were obtained from our colony, originally established with breeder pairs. Chemicals and biochemicals were purchased from Sigma. Antibody to cGMP was obtained from Calbiochem. 125I-cGMP was a kind gift from Dr. David Garbers at UT Southwestern.

Isolation of muscles.

Mouse EDL and soleus muscles were isolated and mounted on Grass FTO3.C force transducers and bathed in physiological salt solution (PSS) containing (in mM) 120.5 NaCl, 4.8 KCl, 1.2 MgSO4, 1.5 CaCl2, 1.2 Na2PO4, 20.4 NaHCO3, 10.0 dextrose, and 1.0 pyruvate at 30°C. The buffer was continuously gassed with 95% O2-5% CO2, resulting in pH 7.6 at room temperature (16). Muscles were either electrically stimulated at 30 Hz for 15 s, incubated with 1 mM Nω-nitro-l-arginine (NLA) for 30 min or with 10 μM sodium nitroprusside (SNP) for 30 s, or subjected to specific combinations. After specific treatments, muscles were quick-frozen by tongs prechilled in liquid nitrogen and stored at −80°C.

Radioimmunoassay of cGMP in muscle extracts.

Briefly, trichloroacetic acid-extracted muscle samples were applied to individual alumina columns, which were then sequentially washed with 10 ml of water, 5 ml of 0.6 N HCl in 95% ethanol, 5 ml of 50% ethanol, and 2 ml of water. Alumina columns were then piggy-backed over Dowex 50W-X8 columns. Three milliliters of 200 mM ammonium formate was added to the alumina columns to elute nucleotides into the Dowex resin. The piggy-backed columns were placed over vials, and 1 ml of 200 mM ammonium formate was added. After the liquid was passed through the alumina resin, cGMP was further eluted from the Dowex 50W-X8 resin with 3 ml of water, and samples were lyophilized overnight. cGMP standards contained 5, 10, 20, 40, 50, 100, and 200 fmol of cGMP. Lyophilized samples and standards were acetylated by redissolving with 25 μl of 4% triethylamine and 25 μl of 2% acetic anhydride, mixed, and placed in an ice bath. 125I-cGMP in 90 mM sodium acetate buffer, pH 4.75, with 20,000 cpm in 50 μl, and 20 μl of the diluted anti-cGMP antibody in 0.1% γ-globulin were added to each tube and mixed. To estimate nonspecific binding, 50 μl of 0.1% γ-globulin was added without anti-cGMP antibody. Samples were incubated overnight at 4°C. Heat-inactivated fetal bovine serum (50 μl) was added to each tube and immediately followed with 1 ml of 12% polyethylene glycol (PEG-8000) in 50 mM sodium acetate buffer, pH 6.2. Tubes were incubated for 1 h on ice and centrifuged (2,500 g, 30 min at 4°C). Immunoprecipitated complexes were washed with 1 ml of 12% PEG, then further incubated on ice for 20 min. After centrifugation, 125I radioactivity in immunoprecipitated complexes was measured in a gamma spectrometer.

Western blot analysis of nNOS and eNOS in muscle extracts.

Mouse skeletal muscles (EDL, soleus, or cremaster) were isolated from C57Bl6, nNOS−/−, and eNOS−/− mice, weighed, and homogenized in 20 vol of 50 mM Tris · HCl, pH 7.5, 1 mM EDTA, 0.1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml of pepstatin A, 2 μg/ml of leupeptin, 1 μg/ml of aprotinin, 1 mM N-α-p-tosyl-l-lysine chloromethyl ketone (TLCK), 2 μM tetrahydrobiopterin, and 20 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS). Protein determination of extracts was performed using a Bio-Rad DC Protein Assay kit. nNOS protein in muscle extracts was analyzed by electrophoresis of 20 μg of total protein from each muscle sample on a 6% SDS-polyacrylamide gel. eNOS protein was analyzed by electrophoresis of 60 μg of soluble protein (17,000 g supernatant fraction) from muscle extracts on a 6% SDS-polyacrylamide gel. Resolved proteins were transferred to nitrocellulose paper and subsequently probed for nNOS and eNOS. A positive nNOS marker prepared from mouse brain and an eNOS marker from mouse heart were also included. A polyclonal antibody raised against the NH2 terminus of rat brain nNOS was used (1:4,000) to detect nNOS, which includes μNOS, the alternatively spliced isoform specific to skeletal muscle (29), and a monoclonal eNOS antibody (Transduction Laboratories) was used (1:1,500) to detect eNOS. Horseradish peroxidase conjugated to goat anti-rabbit (nNOS) or anti-mouse (eNOS) IgG was used as secondary antibody. Immunoreactivity was detected by ECL chemiluminescence (Amersham). Relative nNOS and eNOS amounts were quantified by densitometric scanning of immunoblots. For relative comparisons, muscle samples were analyzed on the same immunoblot.

Vascular response measurements in mouse fast-twitch cremaster muscle.

Adult male mice were anesthetized with pentobarbital sodium (70 mg/kg ip), tracheotomized, maintained with supplemental anesthetic as needed via a jugular venous catheter, and kept at body temperature (37°C) by radiant heating. The right cremaster muscle was prepared and the microvasculature viewed as described elsewhere (16); like EDL, cremaster muscle consists predominantly of fast-twitch fibers (unpublished data). The prepared cremaster muscles were superfused with PSS containing (in mM) 131.9 NaCl, 4.7 KCl, 2.0 CaCl2, 1.2 MgSO4, 20 NaHCO3, 1.0 pyruvate, and 10.0 glucose. The buffer was equilibrated with 21% O2-5% CO2, balanced with N2, to give pH 7.4 at 37°C. Electrical field stimulation via silver foil electrodes was used to contract the muscle for 15 s at 30 Hz (0.2 ms duration, 1–3 V). Because flow-dependent mechanisms have been identified as possible contributors to local arteriolar responses, the muscles were clamped across their origins using a miniature arteriolar clamp to produce “no-flow” conditions comparable to those used in isolated EDL or soleus muscles. Following an equilibration period of 20–30 min, baseline arteriolar diameter was measured, and the muscle was then contracted at 30 Hz for 15 s. Arteriolar diameter was measured within 3 s of the completion of the stimulation. The protocol was completed in C57Bl6, nNOS−/−, and eNOS−/− mice, either under control conditions or with 1 mM NLA added to the superfusate. Separate groups of animals were used for each experimental condition. One vessel was selected for observation in each preparation using predetermined criteria to select either a second or third branch distal to the main inflow vessel and to ensure that all were from the same anatomic location in the microvasculature. Arteriolar diameters were measured offline from videotapes using an electronic caliper calibrated against a videotaped stage micrometer. Maximum diameter of each vessel was taken as that produced by the addition of 10−4 M adenosine to the suffusate at the end of the experiment.


cGMP data were analyzed by a two-way ANOVA; simple effects were examined with a one-way ANOVA. Duncan's post hoc test was used to determine differences between means. Changes in arteriolar diameter were compared using a one-way ANOVA and Newman-Keuls post hoc test. For all analyses, differences were considered significant for P < 0.05. Values are means ± SE.


Expression of nNOS and eNOS in mouse muscles.

Figure 1A shows the relative amounts of nNOS in EDL, soleus, and cremaster muscles from control and eNOS−/− mice. No nNOS was detected in muscles from nNOS−/− mice (data not shown). Amounts of nNOS present in muscles from control and eNOS−/− mice, including the higher molecular weight alternatively spliced μ-isoform (29), were not different. Compared with the fast-twitch EDL and cremaster muscles, which had similar amounts of nNOS, the nNOS content of soleus muscle was ∼30% (P < 0.05).

Fig. 1.

Relative amounts of nitric oxide synthase (NOS) in mouse skeletal muscles. Extensor digitorum longus (EDL), soleus, and cremaster muscles were isolated from control, nNOS−/−, and eNOS−/− mice as described in methods. Western blot analyses for nNOS (A) and eNOS (B) proteins were performed using 20 μg and 60 μg total protein, respectively, for each sample. nNOS protein was detected with an nNOS-specific polyclonal antibody, whereas eNOS protein was detected with a monoclonal antibody against eNOS. Volume measurements represent relative density units obtained from densitometric analysis of the immunoblots (insets).

Similar amounts of eNOS were present in muscles isolated from either control or nNOS−/− mice (Fig. 1B). No eNOS was evident in muscles from eNOS−/− mice (data not shown). In contrast to the distribution of nNOS, the highest expression of eNOS was found in slow-twitch soleus muscle. eNOS expression in EDL and cremaster muscles was similar and ∼40% of that found in soleus (P < 0.05). These results show that amounts of nNOS and eNOS are different in mouse fast-twitch and slow-twitch skeletal muscles. They also show that gene ablation of one NOS isoform does not alter expression of the other NOS isoform.

cGMP formation in mouse fast-twitch EDL skeletal muscle.

Under resting conditions, EDL muscles from control mice contained 4.50 ± 0.93 fmol cGMP/mg wet wt tissue (n = 6). Compared with resting values, cGMP levels increased 166% with electrical stimulation (P < 0.05; Fig. 2). The NOS inhibitor NLA did not affect the resting level of cGMP but completely inhibited the increase due to electrical stimulation. Treatment of the muscles with 10 μM SNP significantly increased cGMP content to 167% of resting values (P < 0.05; Fig. 2).

Fig. 2.

cGMP formation increased in contracting EDL muscle from control and eNOS−/− mice but not nNOS−/− mice. Isolated EDL muscles from 12- to 15-wk-old control, nNOS−/−, and eNOS−/− mice were subjected to the following conditions: PSS only (control), electrical stimulation of 30 Hz for 15 s (Stim), 1 mM Nω-nitro-l-arginine for 30 min (NLA), 1 mM NLA for 30 min followed by electrical stimulation of 30 Hz for 15 s (Stim/NLA), or 10 μM sodium nitroprusside (SNP) for 30 s. cGMP content was measured in extracts of quick-frozen muscles by radioimmunoassay as described under methods. For each mouse type, mean values ± SE are expressed relative to the respective control values (no treatment). cGMP contents were (in fmol cGMP/mg wet wt tissue) 4.50 ± 0.93 for C57Bl6 (n = 6), 4.50 ± 0.59 for nNOS−/− (n = 13), and 2.64 ± 0.31 for eNOS−/− (n = 19). *P < 0.05 compared with control value in each group. #P < 0.05 compared with stimulated value in same group.

Under resting conditions, cGMP content in EDL muscles from nNOS−/− mice was similar to that of control EDL (4.50 ± 0.59 fmol cGMP/mg wet wt tissue, n = 13). Direct electrical stimulation of nNOS−/− muscles with or without NLA did not significantly increase cGMP production (Fig. 2). However, the muscles were responsive to the NO donor SNP with cGMP levels increasing to 236% of resting values (P < 0.05; Fig. 2). Thus the absence of cGMP formation with stimulation in nNOS−/− EDL did not result from impairment of sGC activity but rather the absence of NO production.

In EDL muscles isolated from eNOS−/− mice, resting levels of cGMP (2.64 ± 0.31 fmol cGMP/mg wet wt tissue, n = 19) were significantly lower than those from control or nNOS−/− mice (P < 0.05). Electrical stimulation did not result in increased cGMP formation in EDL from eNOS−/− mice. The mean absolute cGMP value from stimulated eNOS−/− EDL (3.92 ± 0.38; n = 8) was similar to resting values for C57Bl6 or nNOS−/− muscles. NLA did not significantly alter cGMP content in resting muscles but diminished cGMP content with electrical stimulation (P < 0.05; Fig. 2). The value of cGMP measured in EDL muscles from eNOS−/− mice treated with 10 μM SNP (4.35 ± 0.41; n = 4) was significantly less than the corresponding value from nNOS−/− mice (10.94 ± 1.39; n = 13, P < 0.05). This result raises the possibility that some aspect of the cGMP signaling pathway in EDL muscles from eNOS−/− mice may be compromised.

cGMP formation in mouse slow-twitch soleus muscle.

In resting slow-twitch soleus muscles from control mice, cGMP content was similar to that of EDL muscles (4.46 ± 0.50 fmol cGMP/mg wet wt tissue, n = 10). However, unlike the response of fast-twitch EDL muscles, cGMP formation was not increased in electrically stimulated soleus (Fig. 3). The absence of increased cGMP formation in response to stimulation correlates to the reduced nNOS protein content in these muscles (Fig. 1A). Thus, the contraction-dependent increase in cGMP formation appears to be specific for fast-twitch fibers.

Fig. 3.

Stimulation of slow-twitch soleus muscle does not increase cGMP formation. Soleus muscles isolated from 12- to 15-wk-old C57Bl6 were electrically stimulated at 30 Hz for 15 s. cGMP content was measured in muscle extracts by radioimmunoassay. Mean cGMP content is expressed relative to the control (no stimulation) value ± SE. cGMP content in control muscles was 4.46 ± 0.50 fmol cGMP/mg wet wt tissue (n = 10).

Vascular response to skeletal muscle contraction.

In the three groups of mice, the vessels studied had similar maximal diameters of 35.4 ± 2.2, 33.8 ± 2.0, and 37.7 ± 2.6 μm in control, nNOS−/−, and eNOS−/− mice, respectively. All groups displayed high resting tone. The mean resting diameters without electrical stimulation for muscles from control animals were 8.0 ± 1.5 and 5.5 ± 1.6 μm in the absence and presence of NLA, respectively. The values for nNOS−/− muscles were 4.8 ± 0.8 and 3.5 ± 0.2 μm in the absence and presence of NLA, respectively, and for eNOS−/− muscles, resting diameters were 3.9 ± 0.2 and 4.1 ± 0.2 μm in the absence and presence of NLA, respectively. In control mice (n = 12), muscle contraction induced a significant increase in diameter (12.30 ± 2.05 μm), and the contraction-induced dilation was significantly reduced (2.68 ± 0.90 μm) in the presence of NLA (n = 12). The contraction-induced dilation produced in nNOS−/− muscles (6.29 ± 1.92 μm, n = 12) was significantly less than results obtained with control muscles but greater than control muscles in the presence of NLA. In contrast, results obtained with eNOS−/− muscles in the presence (0.93 ± 0.20 μm, n = 11) or absence (1.27 ± 0.48 μm, n = 11) of NLA were not different from the dilations observed in C57Bl6 mice in the presence of NLA (Fig. 4). In the presence of NLA, stimulation-induced arteriolar dilations were further decreased to 0.27 ± 0.1 μm (n = 12) in the nNOS−/− cremaster muscles.

Fig. 4.

Arteriolar dilation is reduced by NLA and in nNOS−/− and eNOS−/− mice. Arteriolar diameter in cremaster muscle was measured 3 s after cessation of electrical stimulation at 30 Hz for 15 s. Muscles were isolated from C57Bl6, nNOS−/−, and eNOS−/− mice and stimulated in the absence (Control, solid) or presence (NLA, open) of 1 mM NLA. Values are expressed as diameter change from precontraction baseline; error bars indicate SE. (For numbers in each group, see text.) *P < 0.05 compared with C57Bl6 control value. #P < 0.05 compared with nNOS−/− control value.


Functional hyperemia describes the matched increase in blood flow that supports the enhanced metabolism of contracting skeletal muscle. Among the many metabolic factors thought to mediate this response (27, 28), NO may also play a significant role (12, 26). NO is a potent vasodilator, which activates a cGMP-dependent relaxation cascade in vascular smooth muscle (18). Although eNOS is an important source of NO, we considered the possibility that nNOS localized at the sarcolemma of skeletal muscle fibers may also contribute NO to functional hyperemia during contraction. In the present study, we have evaluated the potential contribution of nNOS-derived NO to cGMP formation in contracting muscles in vitro and the potential link to arteriolar dilation in situ.

The significant increase in cGMP formation in electrically stimulated fast-twitch EDL muscles from control mice was totally abolished by the NOS inhibitor NLA, indicating a contraction-dependent activation of NOS (Ref. 16, and present study). In EDL muscles from nNOS−/− mice, cGMP content was not increased by electrical stimulation but was increased significantly by the NO donor SNP; therefore, the absence of cGMP formation with stimulation was due to the lack of nNOS and not an unresponsive sGC. These data support the hypothesis that nNOS activation during contraction is required for an increase in cGMP content.

To further discriminate the contribution of nNOS from that of eNOS in promoting cGMP production in contracting muscles, we employed eNOS−/− mice. Despite the nNOS contents comparable to control, cGMP formation was not significantly increased after stimulation. However, stimulation values were diminished by NLA treatment, indicating partial nNOS activity. Treatment with SNP produced a significant increase in cGMP content compared with the control value; however, the mean absolute cGMP responses of eNOS−/− EDL muscles following either stimulation or SNP treatment were lower compared with similar conditions for C57Bl6 EDL muscles. Moreover, the lower cGMP formation in SNP-treated eNOS−/− compared with nNOS−/− EDL muscles may indicate that the responsiveness of sGC to NO is compromised in fast-twitch muscles of eNOS−/− mice. Collectively, the cGMP data suggest that nNOS is activated during muscle contraction in vitro, but the responsiveness of sGC to NO may depend on the presence of eNOS.

cGMP formation in response to electrical stimulation is specific to fast-twitch and not slow-twitch muscles, because cGMP formation was not increased in stimulated soleus. Slow-twitch soleus muscles contain significantly less nNOS and more eNOS than either fast-twitch EDL or cremaster muscles. The amount of nNOS protein present in skeletal muscle fibers may be a critical determinant of its role in functional hyperemia. This apparent inverse relationship is demonstrated both in the differences between cGMP formation in stimulated EDL compared with soleus in vitro, as well as in the ability of electrically stimulated control vs. mdx hind limb to attenuate α-adrenergic vasoconstriction (sympatholysis) during contraction in situ (32). For example, there is significant sympatholysis in mouse hind limb from control mice, which have fast-twitch muscles that demonstrate abundant nNOS protein, but attenuated sympatholysis from mdx and nNOS−/− mice (32), which have fast-twitch muscles that demonstrate a marked decrease in nNOS content and activity (5).

In agreement with the role of NO in sympatholysis (32), NO contributes a significant component to the dilation in stimulated isolated fast-twitch cremaster muscles. However, the dilation was not completely inhibited by NLA. This result suggests, not surprisingly, that there are both NO-dependent and NO-independent dilatory mechanisms, including metabolic mediators of functional hyperemia (27, 28). The arterioles studied in all three groups of mice were of similar size (maximal diameter), and all displayed considerable resting tone due to the presence of 21% oxygen in the superfusate. Estimated dilatory capacities in these vessels were consequently quite large, being ∼27, 29, and 34 μm in control, nNOS−/−, and eNOS−/− mice, respectively. In stimulated cremaster muscles of control mice, the observed 2.5 times dilation from resting diameter was thus a dilatory response of ∼45% of the maximal dilatory capacity. Although nNOS−/− mice also increased their diameter by a similar 2.3 times, this response constituted approximately only 22% of their dilatory capacity, while in the eNOS−/− group, the response was about 10% of their dilatory capacity. Because we used systematic criteria independent of size to select vessels for observation, the finding that maximal diameters were not different among the groups suggests that there was no significant compensatory change in microvascular anatomy at this level of the microvasculature in the nNOS−/− or eNOS−/− animals. Thus, we cannot attribute differences in arteriolar responses to effects of compensatory differences in the microvascular network.

In stimulated nNOS−/− cremaster muscles, arteriolar diameter increased to a value greater than the NLA-blocked preparation from control mice. This contrasts with the finding in EDL muscle that cGMP did not increase with stimulation in nNOS−/− mice. Assuming that the responsiveness of the cremaster arterioles is similar to the EDL, this prompts several questions: was there an increase in cGMP that we could not detect in our assay due to sensitivity limitations; and, did NO act via a cGMP-independent as well as dependent pathway? In stimulated EDL muscles from nNOS−/− mice, cGMP may have increased to a level which we could not detect, but because of the amplification inherent to signal cascades, this was enough to produce a modest dilation.

It is possible that eNOS in endothelial cells might be activated by the mechanical stress of contraction (e.g., vessel compression) indirectly to produce NO. Other potential activators of eNOS include acetylcholine and adenosine. For example, during stimulation acetylcholine might be released from nerve terminals (14) or from endothelial cells (25), and adenosine may be formed in the interstitium (9). However, cGMP formation was not increased in stimulated nNOS−/− EDL where only eNOS was present, and stimulation was not sufficient to increase cGMP in soleus muscles, despite the presence of significant eNOS protein. In regard to the soleus, we cannot, however, discount that there are distinct differences in the regulation of functional hyperemia relative to fast-twitch muscle.

In C57Bl6 and nNOS−/− mice, a contribution from NO is suggested by the observation that the stimulation-dependent arteriolar dilation was attenuated by NLA. This is in accord with studies on human forearm blood flow during exercise, which showed a significant attenuation of increased forearm blood flow with infusion of a NOS inhibitor during exercise (6). The moderate arteriolar dilation that occurred in response to stimulation in nNOS−/− mouse cremaster may be attributed to both NO-independent and cGMP-independent pathways. Metabolic effectors associated with muscle contraction could account for the NO-independent relaxation (e.g., potassium, pH). However, because dilation was partially attenuated in the stimulated nNOS−/− cremaster by NLA, it is possible that an NO-dependent but cGMP-independent pathway was also activated. A number of studies have reported cGMP-independent vasodilatory responses to NO in several vascular tissues (8, 20). Recently, it was shown that NO blocks the production of 20-hydroxyeicosatetraenoic acid (20-HETE) by inhibiting the enzyme P4504A2. 20-HETE is a potent inhibitor of KCa channels (3), which leads to relaxation by hyperpolarization of smooth muscle cells and decreased [Ca2+]i; thus, inhibition of P4504A2 by NO can result in vasodilation independent of cGMP.

In stimulated eNOS−/− EDL, a modest increase in cGMP was blocked by NLA, yet there was a complete absence of effective dilation in stimulated eNOS−/− cremaster. This result suggests that there was a potential increase in NO with stimulation, but both NO-dependent and -independent relaxation pathways were compromised. nNOS and eNOS appear necessary for the complete contraction-induced NO-dependent arteriolar relaxation seen in control cremaster. This conclusion contrasts with reports indicating compensation by one isoform of NOS in the absence of the other isoform but supports the idea that compensation may depend on the location of the vessel in the circulatory system. For example, an nNOS-dependent mechanism appears to compensate for the absence of eNOS in mounting a vasodilatory response to acetylcholine in pial arterioles in eNOS−/− mice, because the vasodilation is blocked by 7-nitroindazole, an inhibitor of nNOS (21, 22). However, in a separate study with eNOS−/− mice, no nNOS-dependent compensatory relaxation mechanism was observed in carotid arteries treated with acetylcholine (7). In both cases, the sensitivity of carotid arteries and pial arterioles from eNOS−/− mice to SNP was increased (7, 21, 22), but this increased sensitivity was not observed in the eNOS−/− cremaster. Thus, compensatory mechanisms of nNOS or eNOS and related changes in sensitivity of sGC to NO may depend on the type and location of the vessel in the circulation. The attenuated cGMP formation with SNP in the eNOS−/− EDL identifies sGC as a possible candidate protein for decreased amount of activity in eNOS−/− fast-twitch skeletal muscles. Other possibilities in this cGMP signaling cascade should also be considered.

In summary, cGMP formation with muscle activation in vitro is due primarily to NO produced by nNOS and occurs in fast-twitch but not slow-twitch muscles. The cGMP response in fast-twitch muscle is attenuated in the absence of eNOS. In contracting isolated fast-twitch muscles, both nNOS and eNOS appear necessary for arteriolar dilation.


We thank Kimberley Burzynski and Patricia A. Titus for expert technical assistance.

This work was supported by National Heart, Lung, and Blood Institute Grants HL-06296 (to J. T. Stull) and HL-56574 (to I. H. Sarelius) and by a grant from the Muscular Dystrophy Association (to J. T. Stull).

Address for reprint requests and other correspondence: K. S. Lau, Dept. of Physiology, Univ. of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Boulevard, Dallas, TX 75235-9040 (E-mail: klau{at}



View Abstract