INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Skeletal muscles generate substantial amounts of reactive oxygen species (ROS) both at rest (1, 2) and during contraction (3, 4). These products of oxidation are thought to be formed within mitochondria during electron transport and cross cell membranes via anion channels (5). ROS may be essential for normal physiologic control processes in muscle under quiescent conditions (6) and, at low levels, appear to induce endogenous antioxidant production, an effect that could explain the benefit of training (7). However, during periods of intense contractile activity, ROS peroxidize polyunsaturated fatty acids, which causes loss of cell membrane integrity and increases microvasculature permeability (8). Strenuous exercise may also produce levels of ROS high enough to overcome normal antioxidant defenses and denature proteins important in the contractile and excitation-contraction coupling (ECC) processes (9). Recovery requires a prolonged period for repair and resynthesis of these proteins. Such degrading effects suggest that ROS contribute to muscle fatigue (8).

The diaphragm fatigues when subjected to high work loads. There is both direct and indirect evidence that oxygen-derived free radicals play a contributory role to this complex response (10). Using the intraphrenic infusion of free-radical-generating solutions as a tool, Nashawati and colleagues (11) found a marked reduction in diaphragmatic contractility. Diaz and colleagues (3) used a salicylate trapping method to demonstrate that in vitro diaphragm muscles produce hydroxyl radical during fatigue. Reid and colleagues (4) used the reduced cytochrome c method to measure O₂ production in both resting and actively contracting rat diaphragm bundles, and they reported a significant increase in O₂ production under both conditions. Anzueto and colleagues (12) measured reduced glutathione (GSH), glutathione disulfide (GSSG), and the ratio of GSSG to total GSH and found that these indirect measurements of lipid peroxidation were elevated in the diaphragms exposed to resistive loading. More recently, Borzone and colleagues (13) used electron spin resonance to provide direct evidence that resistive breathing increases free radical formation. Other studies have shown that pretreatment of the diaphragm with free radical scavengers or antioxidants attenuates fatigue (14).

The purpose of this study was to evaluate and quantify the amount of O₂ produced in the isolated, retrograde-perfused rat diaphragm and to relate these measurements to the contractile performance of the diaphragm at rest, during fatiguing stimulation, and during recovery from fatigue. We demonstrate that the amount of O₂ produced was temporally correlated to the decrement in force production seen with fatiguing stimulation. We measured the reduction of cytochrome c both in the presence and the absence of SOD, an established method to quantify O₂ production (15). This method has previously been used to measure the cumulative amount of O₂ produced at rest and during fatigue in small diaphragm muscle bundles. However, we determined the actual amount of O₂ produced on a minute-to-minute basis by measuring the reduction of cytochrome c in the perfusate of the isolated diaphragm, a preparation where there is sufficient muscle mass to make this possible. We also examined the contractile properties of the diaphragm at rest, during fatiguing stimulation, and during recovery from fatigue.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surgical Procedures

Intact diaphragms were isolated from Sprague-Dawley rats weighing approximately 225 g (Harlan Laboratories, Indianapolis, IN). The animals were killed by cervical dislocation without anesthesia. All procedures were performed in accordance with the "Guiding principles in the Care and use of Animals" approved by the American Physiological Society.

After appropriate surgical preparation (10), the diaphragm was separated from its intercostal insertions and lifted free from the carcass. The isolated muscle was suspended from the perfusing cannula and attached to an isometric force transducer (UC-3; Statham Instruments, Hato Rey, PR). Force production was monitored on a strip chart recorder, and the output of the force transducer was digitized and stored in a computer for later analysis with DaDiSP Software (DSP Development, Cambridge, MA). The resting muscle length (L_o) and stimulatory voltage (V_o) were set to produce a maximal isometric tension during the 20-min equilibration period. Diaphragm weight and maximal isometric tension averaged 1.2 ± 0.2 and 20.8 ± 0.5 g, respectively. Once optimized, L_o and V_o were maintained throughout the equilibration period.

The perfused diaphragm model was first described by Bülbring (16) in 1946 and has found acceptance in research on the physiology of the diaphragm (17). The model is appropriate for studies of the physiologic response of the diaphragm to fatiguing stimulation because it avoids the diffusional problems inherent in incubated preparations. Wermers and colleagues (18) established that in this model there was an adequate delivery of oxygen and substrate to central fibers, even under conditions of rapid repetitive contractions.

Perfusion Procedures

Each diaphragm was perfused at a constant flow rate of 2.47 ± 0.03 ml/min by means of a Sarns constant flow pump. A constant flow rate was required to present a constant concentration of cytochrome c to the muscle, making the precise calculations of O₂ production possible. This flow provides adequate oxygenation to maintain constant levels of contractility. Inactive control muscle preparations lost only 10% of their initial force production over a 90 min period. Because flow rate did not change during the fatiguing stimulation, we assume that muscle fatigue was not attributable to a limitation in the provision of an exogenous substrate such as glucose. The perfusion fluid, a modified Krebs-Henseleit solution composed of (in mM): NaCl, 118.0; NaHCO₃, 25.0; KCl, 4.7; CaCl₂, 2.5; MgSO₄, 1.2; KH₂PO₄, 1.1; dextrose, 11.0, was continuously aerated with a 95% O₂-5% CO₂ gas mixture, which maintained the pH at 7.35. The temperature of the fluid was held constant at 37° C. Although insulin is not a component of the Krebs-Henseleit perfusing solution the transmembrane transport of dextrose is maintained in this preparation (19).

Drug Administration and Superoxide Anion Analysis

The detection of O₂ is based on its established ability to reduce cytochrome c (15). Accordingly, cytochrome c and SOD (Sigma Chemical, St. Louis, MO) were prepared as 0.5 mM and 200 µg/ml stock solutions, respectively, and were introduced into the perfusate solution by means of a syringe pump set at a precise rate so that the final concentrations delivered to the contracting muscle were 0.05 mM and 20 µg/ml, respectively. Cytochrome c was introduced into the perfusate 15 min prior to the beginning of fatiguing stimulation. When used, SOD was introduced into the perfusate 5 min prior to fatiguing stimulation and remained in the perfusate throughout the fatiguing procedure and for 5 min into the recovery period.

After 5 min of infusion, to allow equilibration of cytochrome c, the perfusate was collected at 1-min intervals. The spectrum of the cytochrome c present in the collected samples was measured using a Beckman DU 640 spectrophotometer (Beckman Instruments, Irvine, CA) (Figure 1). When reduced, the absorption spectra of cytochrome c has a peak at 550 nm. The magnitude of this peak is directly related to the amount of reduced cytochrome c contained in the sample. The absolute magnitude of the 550 peak (P_OD550) was calculated as

View larger version (16K): [in this window] [in a new window]   Figure 1.   Schematic absorbance spectra of cytochrome c. The spectral peak c located at 550 nm indicates the presence of reduced cytochrome c. The dotted line connecting A and B denotes the absorbance spectra in the absence of reduced cytochrome c. The height of the reduced cytochrome c spectral peak (arrow) (POD550) is used to determine the amount of O2 produced by the isolated, perfused rat diaphragm. The height of the peak is calculated as described in the text.

where (A + B)/2 is the baseline absorbance at 550 nm (i.e., the calculated optical density when the amount of reduced cytochrome c is zero) and A,B, and C are the measured absorbances at 540, 550, and 560 nm.

The actual amount of O₂ released into the vascular bed of the diaphragm was determined by measuring the difference in the optical density (P_OD550) in the presence and absence of SOD and by assuming a molar extinction coefficient of 18.5 × 10³M¹cm¹ (20). In the presence of SOD there is a background reduction of cytochrome c that is not due to O₂. This "noninhibitable" P_OD550, averaged 0.0092 ± 0.0016 (n = 6).

The perfusate was collected in 1-min intervals and analyzed immediately in order to minimize time-related changes in the absorption spectrum. The entire spectrum, as well as P_OD550, exhibited a slow and steady increase with time. These changes did not appear to be related to the influence of light; absorbance readings of perfusate samples were not altered by brief exposures to bright fluorescent light. Rather, we suspect that the absorbance changes were the result of the development of cloudiness in the sample as CO₂ diffused from it and pH changed.

Stimulation Procedures

The suspended diaphragm was stimulated through two platinum electrode rings, which encircled the muscle; one was placed at the insertion of the vena cava and the other at the distal severed portion of the muscle. The stimulator (Model S88; Grass Instruments, Quincy, MA) was adjusted to produce electrical pulses of a constant voltage between the electrodes with a pulse duration of 1.2 ms. A voltage of approximately 100 volts (V_o) elicited a maximal contractile response. Each diaphragm was allowed to equilibrate for 20 min while being subjected to 500 ms trains of 80 Hz stimuli administered once every minute. Muscles that did not maintain constant tetanic tension during the last 10 min of the equilibration period were discarded. After equilibration, the diaphragms were subjected to a 5-min fatiguing protocol consisting of 500 ms 80 Hz trains delivered at a periodicity of one per second. The experiment was concluded by returning the periodicity of train stimulation to one per minute and allowing the muscles to recover for 30 min while monitoring the response to 500 ms trains of 80 Hz stimulation.

In order to determine whether or not electrical stimulation could reduce cytochrome c, the platinum electrodes were immersed directly into perfusate samples containing the cytochrome c, and the fluid was stimulated with the same stimulatory parameters administered to the diaphragm. In no case were we able to detect any stimulation-induced reduction in cytochrome c.

Statistical Analysis

All data were analyzed for statistical significance using SigmaStat software (Jandel Scientific Corp., Corte Madera, CA). The relationships between groups of data were determined by Student's unpaired t test; p  0.05 was accepted as indicating a statistically significant difference between groups. All data in this study are expressed as mean values ± SEM.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Superoxide in Nonfatigued Muscle

Superoxide is not produced in muscles that are stimulated at a very low periodicity and that, consequently, are not fatigued. The magnitude of the 550 nm absorption peak of cytochrome c (P_OD550 as shown in Figure 1) was determined in perfusate samples collected from such nonfatigued diaphragms. During equilibration, these diaphragms were stimulated to contract once every minute (500 ms 80 Hz trains) in order to confirm the stability of the muscles. Under these conditions P_OD550 did not change when SOD was added to the perfusate (Figure 2A). Force also was unchanged in the presence of SOD (Figure 2B). We conclude, therefore, that no detectable O₂ was produced by the diaphragms under these nonfatiguing conditions, i.e., any O₂ produced by the muscle under these conditions was scavenged by intracellular buffers and did not reach the vasculature and was not detected.

View larger version (18K): [in this window] [in a new window]   Figure 2.   The POD550 values in Panel A represent the magnitude of the 550 nm peak of the cytochrome c spectrum measured in samples of perfusate collected from isolated, perfused, rat diaphragms activated by 500 ms 80 Hz train stimulation at a periodicity of one per minute to continuously test the viability of the preparation. Muscles do not fatigue in response to stimulation at this low train frequency. The addition of 20 µg/ml SOD (arrow) did not cause significant alterations in the POD550 values. Panel B represents the grams of force produced by the resting diaphragms. The addition of 20 µg/ml SOD (arrow) did not cause significant alterations in the force produced by the muscles. The data are displayed as average values ± SEM (n = 6).

Superoxide Produced During Fatiguing Stimulation

A plot of P_OD550 of the muscle perfusate (expressed as a percentage of the mean equilibration value for each muscle) during the equilibration, fatiguing, and recovery periods is shown in Figure 3A. In the absence of SOD, P_OD550 in the perfusate significantly increased during the first minute of fatiguing stimulation and remained elevated until the recovery period, when levels returned toward control. In the presence of SOD, P_OD550 did not change significantly throughout the experiment. The calculated values of O₂, expressed as nmol/min, are displayed in Figure 3B. The O₂ appearing in the vascular bed (perfusate) of the active diaphragm during the first minute of the fatiguing protocol was 0.70 ± 0.17 nmol/min. The fall in the concentrations of O₂ during recovery became statistically different (p < 0.009) during the second minute (and every minute thereafter) of recovery when compared with those measured during the last minute of fatigue.

View larger version (19K): [in this window] [in a new window]   Figure 3.   Panel A presents the POD550 values measured in samples of perfusate collected from isolated, perfused, rat diaphragm in the absence or presence of 20 µg/ml SOD. The contractile conditions to which the diaphragms were subjected are the same as those described in the legend of Figure 2. For the purpose of normalizing the data from different experiments, the POD550 values are expressed as percentages of the mean equilibration POD550 value for each muscle. The data in the lower panel show the calculated levels of O2 produced by the isolated, perfused, rat diaphragms as described in METHODS. The amount of O2 generated by the muscle during equilibration is essentially equal to zero. The data are expressed as nmol/min and are displayed as average values ± SEM (n = 6). The asterisks indicate statistically significant differences (p  0.05): In Panel A the statistical comparison is made between values measured in the presence or absence of 20 µg/ml SOD. In Panel B the statistical comparison is made between values and the average O2 measured during equilibration. The fall in the concentrations of O2 during recovery became statistically different (p < 0.009) during the second minute (and every minute thereafter) of recovery when compared with those measured during the last minute of fatigue.

Contractile Response to Fatiguing Stimulation

The isometric tetanic force produced by the diaphragms fell precipitously in response to increasing the frequency of the train stimulation to once per second (Figure 4). At the end of the 5-min fatigue period, the average force at 80 Hz was 12 ± 2% of prefatigue levels. SOD, in the absence of cytochrome c, had no discernible effect on the contractile performance of diaphragm muscles in response to fatiguing stimulation with the tension falling to 13 ± 3% of prefatigue levels in the presence of the antioxidant. After 30 min of recovery, diaphragms with or without exposure to SOD regained only a fraction of their ability to develop tension; the average tension developed by control muscles in response to a single 80 Hz tetany returned to 51 ± 8% of prefatigue levels, whereas that of muscles exposed to SOD recovered to 48 ± 4% of prefatigue levels.

View larger version (20K): [in this window] [in a new window]   Figure 4.   The isometric force production in the isolated, perfused rat diaphragms, in the presence or absence of 20 µg/ml SOD. During equilibration and recovery, the diaphragms were subjected to 500 ms 80 Hz train stimulation at a periodicity of one per minute. At 10 min on the time line, the period of stimulation was increased to one per second to fatigue the muscle. The data are displayed as percentages of the force produced at the onset of the fatiguing protocol (the 10-min mark on the time line). The normalized values are displayed as average values ± SEM (n = 6).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Reactive oxygen species cause damage to a variety of cellular constituents, including DNA and cellular proteins. These molecules produce severe membrane damage by inducing lipid peroxidation chain reactions in cellular organelles (21). Indeed, a fatigue-related reduction in the ability of the diaphragm to develop force may be related to ROS-induced damage of cellular membranes involved in energy metabolism (e.g., mitochondria), in excitation-concentration coupling (e.g., sarcoplasmic reticulum), and/or in action potential propagation (i.e., the outer sarcolemmal membrane).

Recent in vivo studies have provided substantial indirect evidence of free radical production during exercise (22, 23). In humans, for example, experiments have shown that strenuous contractions increase the production of free radicals in the rectus abdominus (1) and that fatiguing exercise alters biochemical indices of oxidative stress, including glutathione status and markers of lipid peroxidation (24). In the diaphragm, in vitro studies have demonstrated that there is an increase in the production of ROS in response to electrical stimulation (3, 8) and that fatigue associated with an increased load is attenuated by the use of free radical scavengers (11, 25). In vivo studies of the diaphragm have also provided evidence of activation of the glutathione redox cycle after resistive breathing (26).

In this study, we monitored the levels of reduced cytochrome c in the perfusate from the isolated, perfused diaphragm muscle in order to evaluate the temporal relationship between diaphragmatic contractility and the amount of superoxide anion produced under nonfatiguing and fatiguing conditions. We found that diaphragms existing essentially at rest, i.e., stimulated to contract only once per minute, did not produce detectable levels of O₂ (Figure 2). This finding differs from the findings of Reid and colleagues (4) who found that passive (resting) diaphragm muscle bundles released significant amounts of O₂ into the bathing medium. However, it is difficult to determine how much of this increase in absorbance was due to O₂ in that the investigators did not repeat the experiment in the presence of SOD. In fact, the absorbance levels of their passive muscle and active muscle + SOD are very similar, suggesting that the increase in cytochrome c absorbance that they observed in resting muscle was not SOD-inhibitable.

Within the first minute of fatiguing the diaphragm with 500 ms 80 Hz tetanic trains administered once per second, there was a significant increase in the levels of SOD-inhibitable reduced cytochrome c in the medium perfusing the diaphragm. Within 2 min of cessation of this fatiguing stimulation, the levels of reduced cytochrome c began to fall significantly and the tetanic tension in response to stimulus trains administered only once per minute began to rise. These two changes were temporally related.

Westerblad and colleagues (27) have demonstrated that a single high frequency tetanic stimulus train maximally releases calcium from the sarcoplasmic reticulum and saturates the contractile apparatus with calcium. The result is a maximal contractile response that reflects the maximum force that can be produced by the contractile proteins under a given set of conditions. On the other hand, low frequency tetanic stimulus trains administered once per minute provide a less than saturating concentration of calcium to the contractile apparatus and produce a contractile response that reflects the efficiency of the excitation-contraction coupling process (ECC) under a given set of conditions. We found that the contractile responses to tetanic trains of 80 Hz stimuli were depressed by fatiguing stimulation, suggesting that both the function of the proteins important in the ECC process and the contractile proteins were compromised. The fact that tetanic force did not completely recover 30 min after fatiguing stimulation was terminated suggests that this fatigue protocol caused long-term damage to the muscle.

Our study found an inverse temporal relationship between diaphragmatic contractility during fatigue and the associated increase in the production of O₂; a finding that, in this case, is similar to that reported by Reid and colleagues (4). Reid's group reported significant amounts of O₂ produced by actively contracting diaphragm muscle bundles. Because of the small size of their muscle preparation, however, it was necessary for them to incubate the diaphragm bundles for 1 h before making absorbance measurements of the reduced cytochrome c in the incubating medium. In view of the fact that our study utilized a muscle of considerably greater mass, we were able to measure O₂ production on a minute-by-minute basis. The advantage of this greater resolution is that we were able to report, for the first time, significant changes in P_OD550 in the vascular compartment within the first minute of the fatiguing stimulation. We noted that the elevated production of O₂ remained relatively constant throughout the fatiguing period. To our knowledge, the temporal relationship between changes in O₂ production and corresponding changes in diaphragm contractility has not previously been described.

The suggestion of protein damage by O₂ production during fatiguing stimulation is based on the reports of Edwards and coworkers (28), Westerblad and colleagues (27) and Allen and colleagues (29). Edwards and coworkers (28) reported in 1977 that fatiguing stimulations had a long-term effect on muscle function, a finding confirmed by the experiments of Westerblad and colleagues (27). The recent findings of Allen and colleagues (29) imply that the long-term fatigue is due to damage to one or more of the proteins involved in the ECC process.

Reid and coworkers (30) have also reported the results of experiments which demonstrated that antioxidant pretreatment inhibited fatigue in human skeletal muscle. They found that the intravenous infusion of N-acetylcysteine (150 mg/kg) inhibited the development of fatigue in tibialis anterior during repetitive, low frequency electrical stimulation. These findings indicated that oxidative stress contributed to the loss of function in human muscle fatigue and that peripheral fatigue can be selectively inhibited by pretreating patients with an exogenous antioxidant. Shindoh and colleagues (31) reported similar findings with N-acetylcysteine in the development of electrically induced fatigue using in situ rabbit diaphragm preparations. During a 20-min protocol of low frequency stimulation via the phrenic nerve, isometric force declined more slowly in animals that had been pretreated by intravenous infusion of N-acetylcysteine 150 mg/kg than in untreated animals. The results of this study (31) led to speculation that oxygen free radicals produced during repetitive muscle stimulation may injure subcellular membranes and organelles. Other investigators have suggested that antioxidant treatment may protect muscle from such injury (32, 33).

It has also been reported that SOD diminishes the emission intensity of 2',7'-dichlorofluorescin when compared with controls and causes treated rat diaphragm muscle bundles to fatigue more slowly than control bundles (8). It was suggested that because SOD was restricted to the extracellular space it reduced the intracellular effects of superoxide by maintaining a gradient for diffusion from the cytosol and by preventing back-diffusion into the cell.

Antioxidant administration is not uniformly beneficial in ameliorating the effects of elevated levels of reactive oxygen species. Studies investigating the effects of ROS scavengers have utilized SOD and SOD conjugated with polyethylene glycol (PEG-SOD) (34). However, SOD and PEG-SOD, because of their molecular size, are too large to readily gain access to the interstitial space and/or intracellular compartment (35), thereby limiting their ability to scavenge intracellular radicals at the site where they damage proteins and lipids. A recent study by Kilgore and colleagues (34) demonstrated that SC-52608, a SOD mimic of much smaller size, was able to provide protection of rabbit hearts from reperfusion injury. SC-52608 proved to have a distinct advantage over SOD because it was able to scavenge superoxide directly from intracellular sources. In our study, SOD had no effect on the contractility of either the fatigued or the unfatigued perfused diaphragm muscle. This could be attributed to the fact that SOD is restricted to the vascular space and is unable to scavenge superoxide radicals at their site of production (i.e., within the cell). Beneficial effects observed by others may be due to the ability of SOD to scavenge ROSs from the vascular space and limit ROS-induced damage at this site.