In the second series of experiments, adult (3-mo-old) male C57BL/6 mice were killed with an overdose of pentobarbitone sodium, and soleus and EDL muscles were rapidly removed from both limbs in random order. The muscles were mounted at constant length in a tissue bath containing oxygenated, mammalian Ringer solution (mM: 137 NaCl, 5 KCl, 2 CaCl₂, 1 MgSO₄, 1 NaH₂PO₄, 24 NaHCO₃, and 0.025 tubocurarine chloride). The solution was gassed continually with 95% O₂-5% CO₂ throughout the experiment. Temperature was maintained at 37°C. Muscles were randomly divided into three experimental groups: control with no additions (n = 10), incubated with oxypurinol (final concentration 0.67 mM, n = 10), and incubated with polyethylene glycol-tagged Cu,Zn-SOD (PEG-SOD, final activity 500 U/ml, n = 10). Oxypurinol and PEG-SOD were obtained from Sigma Chemical (Poole, Dorset, UK) and prepared daily in mammalian Ringer solution.

Statistical analyses were carried out using the Statistical Package (SPSS version 11.01). All data are presented as means ± SE. A one-way repeated-measures ANOVA was used to analyze the effects of the contraction protocol on the reduction of cytochrome c. A mixed-design factorial ANOVA was used to examine the effects of the ROS inhibitors. When Mauchley's test of sphericity indicated a minimum level of violation (Greenhouse-Geisser ε >0.75), data were corrected using the Huynh-Feldt ε; however, when sphericity was deemed to be violated (Greenhouse-Geisser ε <0.75), data were corrected using the Greenhouse-Geisser ε. Where a significant F value was observed, Tukey's honestly significantly difference post hoc analysis was performed to identify where the significant differences occurred. Statistical significance was set at 0.05 for all the tests.

The changes in cytochrome c reduction (presented as superoxide equivalents) in microdialysates from resting and contracting muscle from control mice, mice given an intravenous injection of vehicle, and mice treated with oxypurinol are shown in Fig. 1. The protocol of 180 isometric contractions induced a significant increase in cytochrome c reduction in microdialysates from the mice given saline or vehicle, but prior injection of mice with oxypurinol prevented this contraction-induced increase in the reduction of cytochrome c. XO activity was determined in the liver tissue of mice, since XO activity is known to be high in this tissue (21). Liver XO activity decreased significantly from 2,058.6 ± 388.3 and 1,926.3 ± 356.1 mU/g protein in the saline- and vehicle-treated mice, respectively, to 94.3 ± 36.8 mU/g protein in the oxypurinol-treated mice (n = 6, P < 0.01).

Table 1 shows the effect of incubation of EDL or soleus muscles with oxypurinol or PEG-SOD on the maximum tetanic force generated by EDL and soleus muscles. Treatment with oxypurinol caused a significant reduction in the maximum force generation by EDL and soleus muscles. Incubation with PEG-SOD caused a significant reduction in maximum force generation for EDL muscle but had no effect on soleus force generation. Oxypurinol and PEG-SOD treatments significantly reduced the peak twitch force generation by the EDL muscle (data not shown in detail).

Figure 2 shows the loss of force generation by EDL and soleus muscles over time during repetitive tetanic contractions at 40 Hz. EDL muscles lost a greater proportion of force over 5 min of repeated stimulations than soleus muscles, but the loss of force generation in EDL or soleus muscles was unaffected by exposure to oxypurinol or PEG-SOD. Similarly, neither agent significantly affected the recovery of force at 2 min following contractions (data not shown in detail).

The data presented in Fig. 1 and Table 1 show that exposure of muscle to oxypurinol induced a significant depression in maximum tetanic force generation by EDL and soleus muscles and a reduction in contraction-induced extracellular superoxide anion. It appeared feasible that the effect of oxypurinol to reduce muscle force generation may have been responsible for the reduced superoxide release, rather than any direct effect of oxypurinol on XO activity reducing superoxide release. To examine this possibility, we examined the force-frequency curve for untreated EDL muscles (Fig. 3A) and calculated the stimulation frequency required to produce a force equivalent to that seen in the oxypurinol-treated muscles. A reduction in stimulation frequency to 59 Hz was calculated to generate force equivalent to that achieved at 100 Hz in muscles from oxypurinol-treated mice. The data in Fig. 3B show that the increase in extracellular superoxide anion was unaffected by the frequency of stimulation and, hence, force generation.

Hellsten et al. (18) first showed that XO is an important source of ROS generation during exercise and demonstrated that a chronic exercise protocol increased XO activity in human muscle. The enzyme was found to be present mainly in microvascular endothelial cells and infiltrating leukocytes (19). It was subsequently demonstrated that plasma XO activities were higher in rats exercised to exhaustion than in control nonexercised rats (14), and a linear correlation between plasma XO activity and lactate concentration in exercised rats has been described (40). Allopurinol-induced inhibition of XO was also found to reduce markers of muscle damage associated with exhaustive exercise in rats (56) and humans (16), and XO-derived ROS were shown to be important in activating signaling pathways involved in muscle adaptations to exercise (14, 25).

In the present study, the role of XO in superoxide generation in the extracellular space of skeletal muscle in vivo has been examined. The effect of a nondamaging contraction protocol on the reduction of cytochrome c in microdialysates obtained from the gastrocnemius muscle extracellular space has been studied as a measure of net superoxide activity during a contraction protocol. The specificity of this assay for superoxide in muscle dialysates has been confirmed by Close et al. (9). Figure 1 shows that intravenous injection of oxypurinol prevented the contraction-induced increase in the reduction of cytochrome c in skeletal muscle microdialysates. Our results are consistent with data from Stofan et al. (52), who reported a partial reduction in superoxide anion release from the contracting diaphragm treated with oxypurinol. Matuszczak and colleagues (33) also showed a reduction in ROS formation in the cytosol of soleus muscles from allopurinol-treated mice. In addition to the ability to inhibit XO, allopurinol and oxypurinol have weak hydroxyl radical scavenging properties at higher concentrations than those used here (36). Hence, our data support the theory that XO is involved in superoxide radical release into the extracellular fluid during a nondamaging protocol of muscle contractions.

Release of superoxide from cultured muscle cells has been reported (34), but the data presented here suggest that the superoxide responsible for extracellular cytochrome c reduction may also be derived from XO activity in the endothelium of skeletal muscle. It has been suggested that muscle contraction alters the shear stresses applied to the vascular bed of the muscle and that this latter stimulus induces superoxide formation and release (52). Multiple free radical-generating pathways have been reported in skeletal muscle (22), and the present data argue that the XO pathway is important in superoxide formation in the extracellular fluid after a nondamaging protocol of muscle contractions in addition to the acknowledged activity of XO during and after tissue ischemia.

Contractile activity of skeletal muscle may lead to oxidation of many biomolecules, indicated by altered muscle and blood glutathione levels and an increase in protein, DNA, and lipid oxidation (41, 47). Several studies have examined whether increasing the intracellular levels of antioxidants within a muscle cell can provide protection against oxidation and reduce muscle fatigue (49), but Reid and co-workers (44) also demonstrated that, in nonfatigued skeletal muscle, ROS appear to have a positive effect on excitation-contraction coupling and are obligatory for optimal contractile function. Specifically, they demonstrated that addition of catalase and SOD resulted in a diminished in vitro muscle contractile performance in unfatigued muscle (44). Furthermore, addition of strong synthetic reducing agents or antioxidants such as DTT (2), N-acetylcysteine (29), and DMSO (45, 46) to skeletal muscle in an organ bath resulted in a reduction in skeletal muscle force production. These data have been supported by other studies where exposure to an ROS-generating system (XO and hypoxanthine) or hydrogen peroxide resulted in an increase in low-frequency-stimulated force generation by the nonfatigued diaphragm (44). The data in Table 1 show that prevention of superoxide generation by inhibition of XO activity depressed skeletal muscle force production in EDL and soleus muscles. Maximum tetanic force generation by EDL muscles was depressed by ∼57% and in soleus muscle by ∼38% after incubation with oxypurinol. Incubation of the muscles with PEG-SOD (Table 1) caused a 34% decrease in tetanic force generation by EDL muscles but had no effect on force generation by soleus muscles. Although these data may suggest that PEG-SOD is more effective in preventing superoxide-induced effects on muscle force production in type II fibers, they may also reflect an inability of the added PEG-SOD to compete for superoxide in the presence of other potential reactants such as NO, the concentrations of these reactants may vary between different muscle types.

Other workers have studied the effect of antioxidants on force generation by skeletal muscle and demonstrated a reduction in submaximal force generation of nonfatigued muscle at low frequencies of stimulation, but no similar effects on maximal tetanic force generation have been reported, except where very high (and potentially toxic) concentrations of antioxidants were used (8, 29, 42, 44). It is not immediately clear why the data reported here differ from those previously reported, but the published data are derived from multiple models and none are identical to the model studied here. In particular, there are differences in the muscle type (diaphragm, gastrocnemius, EDL, or soleus muscles), the species (rats, mice, or dogs), the nature and concentration of the antioxidant, and the temperature and incubation periods. Thus a variety of factors may have contributed to the different effects.

The temperature and the duration of the incubation period may be important aspects. Thus, in the present study, mouse muscle was studied at 37°C, whereas some previous data were obtained at lower temperatures. Muscle-derived ROS activities have been reported to be diminished at room temperature (3), and antioxidants that inhibit fatigue at 37°C have been reported to have no effect at 23°C (13). Our experimental model utilized an extended duration of exposure of the muscle to antioxidants at 37°C compared with most other studies.

Variations in the concentrations of antioxidant used might also explain some of the differences observed in our study. There is no consensus regarding the concentration of oxypurinol necessary to inhibit XO in skeletal muscle, but our protocol differs from that used by other authors. In the present experiments, the muscles were incubated with 0.67 mM oxypurinol, which was observed previously to inhibit XO activity in rats (39). We also determined the XO activity in the present study and found a significant decrease after this protocol of administration. In comparison, Stofan et al. (52) injected rats with oxypurinol (50 mg/kg ip) 12 h before the study, and an additional dose of oxypurinol (50 mg/l) was added to the diaphragmatic perfusate used during the in vitro experiments. These authors did not measure the XO activity. Supinski and colleagues (54) examined fatigue and free radical-mediated lipid peroxidation in the rat diaphragm during loaded breathing by administration of oxypurinol and showed no effect of oxypurinol on maximal diaphragm force generation in an in situ diaphragm preparation. Oxypurinol (50 mg/kg) was again administered on the day before the study, and an additional 50 mg/kg was given after 30 min of equilibration of the in situ model. The effects of allopurinol were also studied by Barclay and Hansel (4), who used in vitro and in situ models to examine mouse soleus and canine gastrocnemius-plantaris preparations. In the canine experiments, the authors determined the effect of 1 mM allopurinol on the fatigue rate of blood-perfused canine gastrocnemius in situ but saw no significant effect of allopurinol. In a later study, Matuszczak et al. (33) gave allopurinol to mice (50 mg·kg⁻¹·day⁻¹) and also observed no loss of maximum force generation by the soleus muscle. For studies of PEG-SOD, we examined the effect of 500 U/l on isolated muscles, whereas Callahan and colleagues (8) gave PEG-SOD by intraperitoneal injection at 2,000 U/kg to rats and also incubated skinned muscle fibers isolated from the diaphragm with 2,000 U/l PEG-SOD. Callahan et al. did not observe an effect on maximum tetanic force generation with either treatment protocol.

The loss of contractile function that was observed after exposure to oxypurinol in electrically stimulated EDL and soleus muscles or PEG-SOD in EDL muscles suggests that these agents alter processes within the intact muscle cell. In this study, any intracellular effect of the SOD is likely to have been indirect, because the molecular mass of PEG-SOD restricts it to extracellular distribution. XO is located in the skeletal muscle endothelium. Reid and colleagues (44) examined how addition of SOD and catalase to the extracellular medium might influence intracellular processes mediated by ROS. They proposed that the enzyme substrates hydrogen peroxide and superoxide anions can cross cell membranes; hence, their rapid removal by exogenous enzymes would establish an extracellular “sink” for these endogenous oxidants, maintaining a gradient for diffusion from the cytosol and preventing diffusion back into the cell. In support of this postulate, the present study has shown that nonfatigued muscles release superoxide to the extracellular fluid and that incubation with oxypurinol or PEG-SOD (9) lowered the extracellular superoxide. Other research groups have shown that incubation of nonfatigued muscles with SOD or catalase lowered the activities of ROS measured in the muscle cytosol by a nonspecific indicator (42).

The mechanisms involved in the modulation of force generation by ROS are not fully established, but published data indicate that the variation of force in response to shifts in the redox balance may be mediated by changes in myofibrillar calcium sensitivity (2) and/or a reduction in calcium release from the sarcoplasmic reticulum (1, 57). The proteins that determine the calcium sensitivity of the contractile process are troponin and the regulatory myosin light chain (7, 35). It is also possible that superoxide generated during contractile activity acts to decrease NO bioavailability in untreated muscles; hence, when superoxide generation is inhibited by oxypurinol or SOD, NO bioavailability is increased, with beneficial effects on contractile function. To test this possibility, we used the microdialysis technique to determine the NO levels in muscle extracellular space during the contractile protocol, but treatment with oxypurinol had no effect on microdialysate NO levels (data not shown in detail).

Muscle contractions increase ROS production, and it has been suggested that the increased generation of these species influences the intracellular redox state to induce a more oxidizing environment. Furthermore, these perturbations may result in oxidative modifications in contractile proteins that depress contractile function. In previous studies, SOD and catalase and other agents that reduce ROS (N-acetylcysteine, DMSO, DTT, allopurinol, and desferoxamine) have been reported to slow the loss of contractile force that occurs during repeated electrical stimulation of contractions in muscle preparations in vitro and in situ (4, 49, 53). These findings have not been universally observed: Shrier et al. (50) reported no effect of SOD, PEG-conjugated catalase, or desferoxamine on skeletal muscle fatigue. To explain these contradictory results, Reid et al. (42) proposed that the effects of ROS scavengers depend on the fatigue protocol used and that fatigue is delayed by these agents during contractile protocols using submaximal activation patterns but not when contraction is at maximal or near-maximal intensities (42). These authors suggested that low-frequency fatigue in vitro mimics the metabolic changes produced by peripheral fatigue in vivo: glycogen stores and phosphocreatine levels are depleted, intracellular pH falls, and P_i levels rise. Acidosis and increased P_i each inhibit actin-myosin interaction, precipitating contractile failure, and this type of fatigue is also associated with oxidative stress (51). In contrast, high-frequency fatigue was attributed to failure of impulse propagation across the sarcolemma; force declines rapidly under these conditions and recovers within seconds to minutes after the end of contractions. Previous work indicates that this latter form of fatigue is not mediated by ROS intermediates (42). Our data show that prevention of the contraction-induced formation of superoxide anions in the muscle extracellular space with the administration of oxypurinol or scavenging of the extracellular superoxide with PEG-SOD did not prevent the loss of muscle force generation that occurs with repeated stimulation at 40 Hz for 0.1 s every 5 s.

We also examined the possibility that the decreased generation of force by muscle after treatment with oxypurinol might account for the decrease in extracellular superoxide anion produced by the contracting muscle. Control muscles were electrically stimulated with a decreased stimulation frequency that produced a force generation equivalent to that in the oxypurinol-treated group, but no differences in the stimulation-induced increase in cytochrome c reduction were seen between the two stimulation protocols (Fig. 3). Some previous data also indicate that the release of superoxide from muscle cells is activated by contractions but not directly related to the frequency of stimulation (38), and the present data are in general agreement with this.

We conclude that XO is a source of the elevated superoxide anion detected in skeletal muscle extracellular space during nondamaging contractions, since intravenous administration of oxypurinol prevented the contraction-induced increase in cytochrome c reduction in microdialysates from the gastrocnemius muscle. Selective inhibition of XO-induced superoxide generation by treatment with oxypurinol also caused a significant decrease in the maximum tetanic force generated by EDL or soleus muscles. Similar data were obtained after treatment of EDL muscles with PEG-SOD. Despite the decreased force generation by nonfatigued, oxypurinol-treated muscles compared with untreated controls, neither pretreatment with PEG-SOD nor pretreatment with oxypurinol delayed the development of fatigue in soleus and EDL muscles subjected to repeated stimulation over 5 min. There has been considerable debate about whether XDH and/or oxidase enzymes are as abundant in human muscle as in rodent tissue, but the enzymes are acknowledged to be present in endothelial cells from humans and rodent models (18, 19). The efficacy of allopurinol against muscle damage in some models of exercise-induced muscle damage in humans (16) indicates that even if the enzyme location is limited to endothelial cells in humans, they are able to influence human muscle function. A number of studies have recently suggested that dietary supplementation with agents that prevent ROS formation or scavenge these species does not improve, or may reduce, exercise performance (11, 15, 30–32), and the data presented here indicating a positive effect of XO-derived superoxide anions on muscle force generation are in agreement with these studies.