INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

NITRIC OXIDE (NO) is an endogenous modulator of skeletal muscle function, regulating processes that range from force generation to glucose transport, from mitochondrial respiration to local blood flow (25). NO is produced by the enzymatic action of NO synthase (NOS), which catalyzes the conversion of L-arginine to NO and citrulline. Skeletal muscle constitutively coexpresses two NOS isoforms: the neuronal, or type I, isoform (8, 11, 12, 17) and the endothelial, or type III, isoform (12, 13, 18). Each protein is a distinct gene product with unique distribution in muscle tissue. Type I NOS is associated with the sarcolemma of fast-twitch muscle fibers (8, 11, 17) and is present in interfascicular nerves (26). Type III NOS is expressed by all skeletal muscle fibers, where it associates with mitochondria, and by the vascular endothelium (18). The specific effects of type I versus type III NOS have not been established in skeletal muscle. Pharmacological inhibitors are not isoform specific and cannot be used to resolve this issue (24). An alternative approach is to study genetically altered mice that have a targeted disruption in the isoform of interest. In the current study, we used this genetic approach to assess the functional importance of type III NOS. Limb and ventilatory muscles were isolated from type III NOS-deficient mice, wild-type controls, and heterozygous animals and were studied in vitro to test three hypotheses.

Hypothesis 1. NO release is diminished in type III NOS-deficient muscle. Skeletal muscles composed of predominantly fast-type muscle fibers have been shown to release NO at low rates under passive conditions (1, 17) and at higher rates during repetitive contraction (1). We hypothesized that muscles of type III NOS-deficient animals would release less NO than muscles of wild-type controls under these two conditions. We tested this hypothesis using two muscles, diaphragm and soleus. The diaphragm is a predominantly fast-twitch muscle that expresses both type I and type III NOS; type III NOS levels are similar in the slow-twitch soleus but type I levels are lower (12, 13). We anticipated that diaphragm would produce NO at higher rates than soleus but that type III NOS deficiency would have similar effects on NO release by the two muscles.

Hypothesis 2. Skeletal muscles from type III NOS-deficient mice exhibit a leftward shift of the force-frequency relationship. The force-frequency relationship of diaphragm falls to the right of the relationship measured in soleus. We postulated that this difference is caused, in part, by an exaggerated influence of type III NOS on diaphragm function (17). Pharmacological blockade of NOS activity increases the force of diaphragm contractions at submaximal stimulus frequencies and shifts the force-frequency curve leftward (8, 17, 20, 26), minimizing the difference between diaphragm and soleus (17). We hypothesized that type III NOS deficiency would have similar effects on the diaphragm, shifting the force-frequency relationship to the left of the relationship in wild-type diaphragm. Because type III NOS levels are similar in soleus, we anticipated that contractile function would be similarly affected by genetic deficiency.

Hypothesis 3. The fall in force that occurs during repetitive contractions is accelerated in type III NOS-deficient muscle. Murrant and coworkers (21-23) reported that NO donors slow the decline of force during repetitive contractions of isolated mouse muscles in vitro. We reasoned that the converse would happen if exercising muscles were deficient in type III NOS: force should fall more rapidly. As in other experiments, we expected type III NOS deficiency to have similar effects on diaphragm and soleus.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experimental animals. All procedures were conducted in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved in advance by the Institutional Review Board of Baylor College of Medicine. Mice deficient in type III NOS were generated using embryonic stem cell technology as detailed previously (9). In brief, the murine Nos3 locus was disrupted by targeted deletion of exon 1 and part of the promotor region. Clones developed from the AB2.1 ES cell line were injected into blastocysts to obtain chimeras. The chimeras were bred to C57Bl/6 females to obtain heterozygous (Nos3^m/+) mice that were crossed to obtain homozygous (Nos3^m/Nos3^m) animals. Alterations in gene structure were confirmed by Southern blot analysis; type III NOS deficiency was established by immunohistochemistry; loss of type III NOS function was demonstrated by vascular ring assay. In the current experiments, we studied 38 littermate mice from heterozygote-heterozygote matings: 14 wild type, 13 heterozygous, and 11 type III NOS-deficient animals. Animals were between 6 and 8 wk old, of either sex, and weighed 17-30 g. The mice were fed standard rat chow ad libitum and were maintained on a 12:12-h light-dark cycle. Investigators were blinded to the genotype of the animals until data collection was complete.

Muscle preparations. Each mouse was anesthetized by intraperitoneal injection (0.4 ml/kg) of a combination anesthetic containing ketamine 42.8 mg/ml, xylazine 8.6 mg/ml, and acepromazine 0.7 mg/ml. Under deep general anesthesia, a tracheotomy was performed and the animal was ventilated using a Harvard small animal ventilator and a fraction of inspired O₂ of 1.0. We excised the diaphragm and both soleus muscles from each animal. Two fiber bundles were cut from each diaphragm. Sutures were tied to tendons of the soleus and to the central tendon and rib of the diaphragm fiber bundles so as not to impede muscle contraction.

Protocols. For incubation under passive conditions, one of each muscle type was tied to a plastic rod at resting length. The muscle was placed in a glass tube with 5 ml Krebs-Ringer solution bubbled with 95% O₂-5% CO₂ and incubated at 37°C. Samples of 200 µl were collected in Eppendorf tubes after 30 and 80 min and were stored at 80°C for future measurement of NO derivatives.

Contractile properties of soleus and diaphragm were measured in vitro. Each muscle was secured between platinum electrodes (4 × 37 mm) in a muscle bath filled with Krebs-Ringer solution containing 0.025 mM D-tubocurarine chloride and bubbled with 95%-5% O₂-CO₂ at room temperature. The muscle tendon was tied to an isometric force transducer mounted on a micrometer by which muscle length could be adjusted. Muscle contraction was stimulated directly using supramaximal voltage (measured characteristics: 13 V, 550 mA). Transducer output was displayed on an oscilloscope from which contractile properties were recorded. After adjusting muscle length for maximum twitch force (optimal length), the Krebs-Ringer solution was adjusted to 37°C and maintained at this temperature for the remainder of the experiment. After 30 min thermoequilibration, twitch force, time-to-peak twitch force (TPT), and twitch half-relaxation time (1/2RT) were recorded using a stimulus duration of 0.2 ms. Tetanic forces then were measured every 2 min at stimulation frequencies of 300 (maximal tetanic force), 15, 300, 30, 300, 40, 300, 50, 300, 80, 300, 120, 300, 160, 300, 200, and 300 Hz using a train duration of 300 ms. After 10 min recovery, fatigue was induced by 10 min of intermittent tetanic contractions stimulated using frequencies of 30 (diaphragm) or 15 Hz (soleus) and 0.5 trains/s; force was recorded every 30 s. Aliquots of the bathing medium were collected at standardized time points during each contractile study for later analysis of NO derivatives.

After contractile measurements, optimal length of the mounted muscle was measured and recorded. The muscle was trimmed of bone and connective tissue, blotted dry, and weighed. In post hoc calculations, force measurements were normalized for muscle cross section according to Close (6). Force-frequency data were expressed as relative forces by dividing the submaximal tetanic force at each stimulation frequency by the average of maximal tetanic forces generated immediately before and after the submaximal contraction (16).

NO measurements. NO was determined by the chemiluminescence method (3) as total NO derivatives in the bathing medium, including NO, nitrite, nitrate, and nitroso compounds. We used a commercial NO analyzer (model 7020, ANTEK, Houston, TX) and a solution of 0.1 M vanadium chloride in 2 N HCl at 90-95°C as the reducing agent. A constant flow of helium gas carried the NO that effluxed from each sample through a cold trap to reduce humidity, through 0.1 N sodium hydroxide to remove acidic gases, and into the reaction chamber where NO reacted with ozone to generate light. Luminescence measured by a photomultiplier tube was directly proportional to NO content of the sample. Signals from the analyzer were recorded on a strip chart and integrated using a Waters 746 Data Module (Millipore, Milford, MA). A standard curve for sodium nitrite was generated daily and was linear over a range of 163 nmol to 163 µmol.

Statistics. Data are expressed as means ± SE. One-way ANOVA or Kruskal-Wallis ANOVA on ranks were performed on parametric and nonparametric data, respectively, to compare differences among groups for an individual variable (15). Two-way repeated-measures ANOVA with a Tukey test was used to test parameters that were measured repeatedly over time (15). Comparisons were considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal characteristics and muscle dimensions. Data in Table 1 demonstrate that animal weights and gender distributions were similar among groups. The lengths, weights, and cross-sectional areas of diaphragm fiber bundles and soleus muscles also were evaluated. None of these parameters differed significantly among groups.

NO release by muscles. The average rates of NO release under passive conditions are detailed in Table 2. There were no differences among genotypes or between diaphragm and soleus. During repetitive contractions, there were no differences in NO release between wild-type and type III NOS-deficient muscles (data not shown). Grouped data in Fig. 1 show that NO was released at higher rates during active contraction. NO release increased approximately eightfold for diaphragm and three- to fourfold for soleus, such that NO release by active diaphragm exceeded that of active soleus. Control experiments confirmed that NO signals were muscle derived. In the absence of muscle, NO levels did not increase over time in the bathing medium nor were NO levels altered by electrical stimulation (data not shown)

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Average rates of nitric oxide (NO) release during passive incubation (solid bars) and active contraction (open bars). Values are averages ± SE of rates measured in wild-type and type III NO synthase (NOS)-deficient muscles, which did not differ between groups for either muscle under either condition. Heterozygous muscles released NO at significantly lower rates during active contraction (data not shown) and were excluded from these calculations. NO release by diaphragm and soleus were not different under passive conditions but were during active contraction (P < 0.001); * P < 0.05 vs. passive incubation.

Contractile properties. Table 3 shows twitch characteristics and maximal tetanic forces measured in all groups. Diaphragm exhibited no differences among genotypes in any parameter. For soleus, force production during twitch and/or maximal tetanic contractions was unaffected by genotype. Twitch timing was modestly prolonged in heterozygous and type III NOS-deficient muscles; small increases were observed in TPT (heterozygous only) and 1/2RT (both groups). Data in Fig. 2 show force-frequency relationships for diaphragm and soleus; both were unaffected by genetic modifications. During the force-frequency protocol, maximal tetanic force was measured repeatedly as an index of muscle stability (Fig. 3). Maximal force did not decrease significantly in any group studied. Finally, the acute fatigue protocol had similar effects on diaphragm and soleus (Fig. 4). Force decreased rapidly over the initial 300 s followed by a slower decline over the last half of the protocol. Overall, force decreased by ~45% for both muscles. There were no differences among genotypes.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Force-frequency relationships measured in diaphragm (A) and soleus (B) muscles from wild-type (), heterozygous (), and type III NOS-deficient () animals. Values are average forces expressed as percentages of maximal tetanic force (%Po); SE bars obscured by data symbols.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Changes in Po developed by diaphragm (A) and soleus (B) during force-frequency protocols. Values are average forces expressed as percentages of maximal force for wild-type (), heterozygous (), and type III NOS-deficient () muscles; SE bars are obscured by data symbols.

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Fatigue of diaphragm (A) and soleus (B) during repetitive contractions. Values are average forces ± SE developed by muscles from wild-type (), heterozygous (), and type III NOS-deficient () animals.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current data confirm previous reports that fast-twitch muscles release NO under basal and active conditions (1, 17) and extend those observations to show that slow-twitch muscle (soleus) responds in a similar fashion. Neither NO release by isolated muscles nor isometric contractile function was altered in type III NOS-deficient muscles. Our findings indicate that the type III NOS isoform is not an essential modulator of these processes.

Animal model. Gregg and coworkers (9) developed the type III NOS-deficient mice used in this study. The mutant Nos3 gene was created by targeted disruption of the entire first exon and part of the promotor region. Successful introduction of this construct into the germ lines has been confirmed by Southern blot analysis of genomic DNA using either 3'or 5' flanking probes. Vascular endothelium does not stain for the type III NOS isoform using rabbit polyclonal anti-type III NOS antibodies. Vascular function also is altered. Homozygous mice are mildly hypertensive, with mean arterial pressure increased ~15 mmHg, and preconstricted aortic rings from type III NOS-deficient animals do not relax with the addition of acetylcholine 10⁵ M. In addition to these stereotypical effects, type III NOS deficiency also has been found to influence development. Homozygous offspring are consistently fewer, due to fetal loss, and have lower body weights at birth; ~10% of surviving offspring have limb malformations. However, by the age at which we conducted our study (6-8 wk), growth and behavior of type III NOS-deficient animals are similar to control.

NO release by muscle. Constitutive sources of NO in wild-type muscle include the type I and type III NOS isoforms. In theory, either could contribute to NO release in vitro. The fact that NO was released at normal rates by type III NOS-deficient muscles indicates the type III isoform is not essential for this process. It may be that NO derivatives produced by type III NOS do not diffuse out of wild-type muscle. Intracellular localization of the enzyme and its close association with muscle mitochondria make this explanation attractive. NO synthesized within solid tissue can react with a variety of biological targets, including thiols, heme centers, and mitochondria-derived superoxide anions (27), thereby limiting diffusion from the tissue. An alternative explanation is that type I NOS activity may have been increased in type III-deficient muscle such that one isoform exactly compensated for loss of the other. However, previous studies of NOS-deficient animals argue against this mechanism; animals deficient in type I NOS do not upregulate the type III isoform (7).

This is the first report of NO production by mouse muscle, the first demonstration that slow-twitch muscle releases measurable amounts of NO, and the first direct comparison of NO release by ventilatory and limb muscles. Our data substantiate earlier reports that NO is released by rat diaphragm (17) and rat extensor digitorum longus (EDL; Ref. 1) in vitro. Under passive conditions, we found that mouse diaphragm and soleus released NO at nearly identical rates. These rates increased three to eightfold during active contraction, which also is consistent with data from rat EDL (1). The mechanism whereby muscle contraction increases NO synthesis has not been established, but factors that favor this response include increases in intracellular calcium levels, shear stress, and intramuscular pressure.

Use of the chemiluminescence technique requires that extraneous sources of nitrite and nitrate be rigorously controlled. We measured and controlled for basal nitrite/nitrate levels in the buffer and confirmed that this background noise was not altered by prolonged incubation under experimental conditions (90 min trials at 37°C with O₂-CO₂ equilibration and plastic support rod in place). Furthermore, we determined that the nitrite/nitrate levels of muscle-free medium were not altered by the protocol of electrical field stimulation used to activate muscle contraction. These data were used to identify the nitrite/nitrate signal that derived from the muscle specimens. More-detailed evaluation of the muscle-derived signal was not possible due to limited availability of these transgenic animals. However, related studies in our laboratory have shown that the nitrite/nitrate signal obtained from excised mouse muscle is inhibited by pharmacological NOS blockade (L-arginine derivatives depress nitrite/nitrate release, D-arginine homologues do not; data not shown). Such data confirm that the muscle-derived signal obtained under these conditions primarily reflects endogenous NOS activity.

The absolute rates of NO release that we measured were greater than the 1-3 pmol · mg¹ · min¹ reported by Kobzik et al. (17) and Balon and Nadler (1) for passive muscle. This difference is likely to reflect the use of divergent methodologies. We used a chemiluminescence technique adapted to measure all NO derivatives, including nitrite, nitrate, and nitroso compounds of proteins and thiols. In contrast, Kobzik and coworkers (17) measured NO release as that component of the cytochrome c reduction assay that was inhibited by NOS blockade. Only a subset of NO derivatives reduces cytochrome c, limiting the sensitivity of this assay. Sensitivity is further limited by muscle-derived oxidants, e.g., hydrogen peroxide, that can reoxidize cytochrome c and reverse the NO signal. Balon and Nadler (1) detected NO release using a chemiluminescence technique that selectively measured nitrite accumulation. Nitrite represents ~20% of the oxidized NO products that accumulate from enzymatic synthesis (14); nitrite measurements therefore are expected to yield a lower absolute value than the present method.

Contractile properties. NO has marked effects on the contractile function of skeletal muscle in vitro. In unfatigued muscles, NOS inhibitors increase force production at submaximal stimulus frequencies and shift the force-frequency relationship leftward; NO donors reverse this effect (8, 17, 26). During repetitive tetanic contractions, Murrant and colleagues (21-23) showed that force production by mouse soleus or by mouse EDL falls progressively and that exogenous NO can slow this decline. We reasoned that the effects of type III NOS deficiency would resemble the effects of NOS inhibition and would be the opposite of NO donors. Instead, type III NOS deficiency had no discernable effects. This suggests that the type III isoform expressed by muscle fibers does not modulate contractile function. Nor does NO from endothelial cells appear to influence muscle fiber contraction under the present conditions.

Perspectives