INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

OXYGEN-DERIVED FREE RADICALS or reactive oxygen species (ROS) have been implicated in the pathogenesis of a wide spectrum of diseases as well as in the aging process. In addition to playing a role in direct tissue damage, their generation may also amplify the body's general inflammatory response and promote further cell injury (8). In general, the mammalian cell has adequate antioxidant reserves to cope with ROS production under normal physiological conditions. The system consists of antioxidant vitamins, thiol-containing low-molecular-weight compounds such as reduced glutathione (GSH), and antioxidant enzymes, including superoxide dismutase (SOD), GSH peroxidase (GPX), and catalase (Cat).

A number of paradigms, including rhythmic exercise and repeated eccentric muscle activity, have related ROS production to muscle inflammation and injury. During intense physical activity, the flow of oxygen through muscle cells is greatly increased at high levels of oxygen uptake. At the same time, the rate of ATP utilization exceeds the rate of ATP generation that can lead to free radical generation and has been implicated in fatigue, muscle soreness, myofibril disruption, and impairment of immune function (4, 46). In addition to rhythmic exercise, authors have speculated that ROS may also play a role in muscle damage associated with repeated eccentric contraction protocols (4, 36, 37). Repeated eccentric muscle contraction leads to an initial focal muscle injury, followed over the next 48-72 h by rising levels of creatine kinase (31), loss of force production (29), and inflammatory changes within the muscle (8, 40). As few as 15 high-intensity lengthening contractions can result in an immediate, 60% reduction in maximum isometric tetanic force production (43). This initial force decline is felt by most to be attributable to mechanical factors, whereas some have speculated (4, 36, 41) that ROS may play a causative role in the changes seen 48-72 h after the initial mechanical injury. This hypothesis is supported by a study in which animals pretreated with polyethylene glycol-SOD showed an attenuation of force loss at 3 days postinjury, compared with the nontreated control group (47). The protective effect of polyethylene glycol-SOD pretreatment was particularly true for aged animals, which also showed an attenuation of force loss both immediately and 3 days after the eccentric contraction protocol. Interestingly, studies on antioxidant supplementation and ROS-induced muscle damage have produced conflicting results (10, 11, 28, 30, 39, 44).

Our laboratory studies focus on the pathogenesis of acute muscle stretch injury. We have recently developed a new approach to create a minimally invasive and reproducible partial injury to the rabbit tibialis anterior (TA) muscle-tendon unit (5). In our in vivo model, an acute injury is produced in the TA muscle by stimulating the peroneal nerve and plantar flexing the ankle one time within its physiological range of motion. Varying the initial amount of tendon shortening before muscle stimulation and ankle rotation can control the amount of injury (5). It should be kept in mind that our model of acute stretch injury is different from the injury observed with exhaustive exercise and repeated eccentric muscle contractions. With exhaustive exercise, there is a disturbance between intracellular prooxidant and antioxidant homeostasis, which has been shown to result in oxidant stress damage (20). On the other hand, the injury associated with repeated eccentric muscle activity often leads to delayed-onset muscular soreness and morphological changes, rather than the immediate muscle hemorrhage and edema that occur along with frank muscle fiber disruption in our model (5, 32).

Other laboratories studying acute muscle stretch injury have shown a decline in the muscle's contractile properties and tensile strength in the first 24 h after injury (32). These observations have not been explained to the best of our knowledge. In addition, there is an intense inflammatory response that is accompanied by both phagocytic and monocytic cell infiltration of the injured tissue (5, 35). Our hypothesis is that ROS production is increased within the first 24 h after muscle stretch injury. There are a number of possible sources of ROS generation in damaged skeletal muscle. These include phagocytic cells, membrane-bound oxidases, mitochondrial respiratory chain, cytochrome P-450, and various cytosolic catalytic enzymes (46). In turn, these highly reactive molecules may promote direct tissue injury, e.g., lipid peroxidation damage, or perhaps even amplify the initial host inflammatory response by upregulation of various cytokines and growth factors. The primary purpose of the present study was to determine whether ROS production is increased in the first 24 h after acute muscle stretch injury. Second, we wanted to examine for indirect measures of ROS production, e.g., lipid peroxidation damage, and to determine the effects of the acute injury on antioxidant enzyme activity and glutathione status.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isokinetic Test System

View larger version (70K): [in this window] [in a new window]   Fig. 1.   Isokinetic injury and isometric torque device. [Modified from Best et al. (5)].

A number of pilot studies were conducted to determine the kinematics of muscle and tendon motion during tendon shortening and ankle rotation. A video-based image-analysis system was used to determine muscle-tendon excursion during both active and passive ankle rotation. An anesthetized rabbit was placed supine in the testing system (Fig. 1), with the lower extremity exposed; i.e., the skin and fascia were removed to permit direct visualization with the video system and access to the muscle-tendon unit. Surface markers (4-mm dots of black paint) were placed on the tendon (at the site of clamping for tendon shortening) and proximally at the muscle origin off the tibial tubercle. The maximum length change of the muscle-tendon unit was defined as the difference between the ankle and knee at 90° and full extension. The test was performed in both the active and passive states. The change in length from 0° (L₀ = 9.0 cm) to 90° (L_f = 10.3 cm) in the passive state was, therefore, 1.3 cm. In the active state (muscle stimulated at 50 Hz, 0.5 V, 0.5-ms pulse width), the length change was 1.7 cm; (L₀ = 8.6 cm, L_f = 10.3 cm). Therefore, we believe that 1.7 cm define the maximum range of TA muscle-tendon excursion while the muscle is being stimulated and the ankle rotated within its physiological range of motion. To produce injury and maintain the kinematics of ankle joint rotation and motion we, therefore, temporarily shortened the tendon; i.e., we lengthened the muscle fiber and muscle-tendon unit. Therefore, the total amount of length change during stimulation and stretch would be the amount due to shortening and the amount due to rotation. In the present study, this would be 1.3 cm (tendon shortening) + 1.7 cm (ankle rotation during muscle stimulation) for a total of 3.0 cm.

Animal Care and Injury Protocol

Muscle injury and tissue harvest for biochemical analysis were carried out by using an intramuscular anesthesia combination of ketamine, xylazine, and acepromazine to provide complete muscle-tendon relaxation (5). A 1-cm incision was made over the dorsum of the right foot just distal to the ankle joint to isolate the tendinous portion of the TA muscle-tendon unit. The peroneal nerve was isolated via a 4-mm skin incision at the knee.

The animal was placed supine in the test system (Fig. 1). The tendon was accessed through the foot incision and shortened 1.3 cm by using the tendon-shortening device. With the muscle-tendon unit in the shortened position, the TA muscle was stimulated with a 50-Hz pulse rate, 0.5-ms pulse width, and 0.5-V output to produce tetany (Fig. 2). At muscle tetany, the ankle was plantar flexed through 90° at 450°/s. This combination of stimulation and stretch within the ankle's physiological range of motion creates a reproducible and quantifiable injury to the muscle-tendon unit (5). The torque-angular displacement time behavior during this stimulation and stretch protocol was recorded, and data were stored on a 486 personal computer. A sham operation was performed in the left foot and knee. Ankle isometric torque was recorded immediately postinjury for all 18 animals (control and injured). The 11 animals killed at 24 h postinjury were retested for isometric torque before tissue harvest. For the 11 animals examined at 24 h postinjury, the ankle and knee incisions were closed with 4-0 Ethilon sutures, and the animal was returned to its cage.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Typical nonlinear torque-time curve generated during tibialis anterior (TA) muscle stimulation and ankle joint plantar flexion.

Tissue Preparation

Biochemical Analyses

ROS production. Muscle ROS generation was determined in fresh tissue homogenates by using dichlorofluoroscein (DCF) as a probe, according to LeBel and Bondy (26), as modified by Kim et al. (23). The 2',7'dichlorofluoroscin acetate (DCFH-DA) stock solution was made fresh every week by dissolving it in 1.25 mM methanol and kept in a dark room at 0°C. Fifty microliters of homogenate were added to a quartz cuvette containing 2,938 µl of 0.1 M phosphate buffer (pH 7.4), and 12 µl of 1.25 mM DCFH-DA (total volume = 0.3 ml). DCFH-DA was dissolved in methanol to aid in the transport across membranes. The assay mixture was incubated for 15 min at 37°C to allow the DCFH-DA probe to cross any membranes and for nonspecific esterases to cleave to diacetate groups. DCF formation was determined fluorometrically with a Hitachi F-2000 fluorescence spectrophotometer at excitation wavelength of 488 nm and emission wavelength of 525 nm at 37°C. Measurements were made every 15 min for 60 min, and linear DCF production rate was determined relative to the amount of protein added to the cuvette. A blank consisting of the appropriate buffer and 5.0 µM DCFH-DA without homogenate was used to correct autooxidation rate of DCFH-DA. The DCF assay was carried out within 1 h of tissue harvest. Duplicate samples were run to confirm the assay's repeatability, and the results were averaged. The units were expressed as nanomoles DCF formed per minute per milligram protein.

Lipid peroxidation. Peroxidative damage to cellular lipid constituents was determined by measuring malondialdehyde (MDA) in butanol extracts according to Urchiyama and Mihara (42), with modifications as follows: 10 mM butylated hydroxytoluene and 200 mM ferrous sulfate were included in the assay mixture. Sealed tubes were incubated for 15 min at 99°C. MDA content was calculated based on a standard curve using 1,1,3,3-tetraethoxypropane as a standard.

Antioxidant enzyme activities. Activities of GPX (EC 1.11.1.9), glutathione reductase (GR) (oxidized-glutathione oxidoreductase; EC 1.6.4.2.), and total SOD (mitochondrial + cytosolic) (EC 1.15.1.1) were determined as previously described (22). Cat (EC 1.11.1.6) was assayed at 20°C by the method of Aebi (1) .

Glutathione status. GSH and glutathione disulfide (GSSG) were analyzed in the muscle homogenates by using HPLC methods, as previously described (20). Protein concentrations in the tissue homogenates were determined by the Bio-Rad assay using BSA as the protein standard.

Mechanical Data

Histology

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mechanical Data

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Typical nonlinear torque-theta curve generated during TA muscle stimulation and ankle joint plantar flexion. Note that area within the curve (hysteresis loop) is energy loss of the system during injury.

Free Radical Production

Lipid Peroxidation (MDA)

Antioxidant Enzyme Activity

Glutathione Status

Histology

View larger version (137K): [in this window] [in a new window]   Fig. 4.   Hematoxylin and eosin stain of myotendinous junction at 24 h postinjury, demonstrating inflammation and presence of polymorphonuclear leukocytes along with some edema and fiber necrosis. Magnification ×100.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There is increasing evidence to suggest that oxygen-derived free radicals or ROS participate in the pathogenesis of a wide variety of disease processes. In addition, ROS are thought to play a role in muscle inflammation and damage seen with exhaustive exercise and perhaps even repeated eccentric muscle activity as well as in various pathological disorders of muscle (18). However, the exact role and mechanism by which free radicals act to cause this damage are unknown. Furthermore, the majority of studies to date investigating exhaustive exercise and repeated eccentric muscle activity have perhaps been limited by methodological problems of measuring free radical generation and, therefore, it has been difficult to generate consistent data and conclusions. The most significant finding in the present study was the observation that ROS production is increased in the first 24 h after acute single-stretch muscle injury. The increase in intracellular free radical levels after injury was observed despite few changes in antioxidant enzyme activity and biomarkers of oxidant stress damage.

The high reactivity of free radicals makes their direct detection in biological tissues difficult. Consequently, the majority of studies have examined markers of lipid peroxidation, protein oxidation, cellular redox status, and changes in antioxidant enzyme activity as indirect measures of free radical involvement in tissue damage. In light of this limitation, we were interested in providing more direct evidence for ROS production. To our knowledge, there are only three studies in the literature showing direct free radical production by skeletal muscle (6, 9, 19). These studies used electron spin resonance spectroscopy to measure free radical production.

We chose to measure ROS levels in tissue homogenates using the DCFH-DA intracellular probe, first described as a fluorometric assay of H₂O₂ (22). DCFH-DA is a stable, nonfluorescent molecule that readily crosses cell membranes. The diacetate group is cleaved by nonspecific cellular esterases to nonfluorescent DCFH, which can then be oxidized by ROS to highly fluorescent DCF. The method is rapid and quantitative (26) and has been used to determine ROS production in the diaphragm after chronic low-frequency electrical stimulation (34). Others have shown that DCF fluorescence is a sensitive, quantitative, and direct measure of ROS formation in rat synaptosomes (26). A major limitation of this assay is that DCFH competes with antioxidants as well as other cell macromolecules and, therefore, only serves as an estimation of ROS generation. The probe can be oxidized by a number of free radicals and, therefore, is not specific for any one molecule. Whereas H₂O₂ and several lipid hydroperoxides, in the presence of hematin, are reported to oxidize DCFH (7), the ability of other ROS, such as superoxide anion (O₂·) and hydroxyl radical, to stimulate DCF formation are inconclusive (26). In addition, DCF signals reflect not only the magnitude of oxidant produced, but in biological systems it is readily oxidized by nitric oxide derivatives. Nevertheless, in this experiment, we show increased ROS production in mechanically stretched skeletal muscle within 24 h of injury. Duplicate samples of the assay showed excellent repeatability (<2% difference between the duplicate samples). It should be noted, however, that in addition to increased levels within the injured leg there were also increased levels within the control leg samples at 24 h. This observation suggests that there was a generalized whole body systemic response to the trauma induced by a combination of the surgery and the injury. The 0-h animals were included in the present study to estimate baseline ROS production immediately after injury. This experimental design allowed us to examine both time (0 vs. 24 h) and treatment (injury vs. control) effects on ROS generation and measures of oxidant stress. Previous studies have demonstrated that the choice of homogenate buffer is important when interpreting results with the DCF assay (26). Our buffer did not contain respiration substrates for mitochondrial oxidative phosphorylation. Therefore, we propose that the assay measured predominantly nonmitochondrial sources of ROS generation, e.g., neutrophils, peroxisomes, and P-450 in our study.

No immediate or delayed lipid peroxidation was observed in the injured muscle in the present study. This is also in contrast to prominent increases in peroxidative damage to lipid components after exhaustive exercise (11) and ischemia-reperfusion (13). One possible explanation is that with our acute stretch-injury model ROS were generated primarily in the soluble fractions (cytosol) of the cell, which are geographically distant from lipid-rich membranes. The absence of lipid peroxidation damage in our model may also indicate that the main ROS produced were peroxides rather that O₂· or hydroxyl radical that are required for the initiation of peroxidative damage to lipids. Another possible explanation of our findings is the methods used to detect lipid peroxidation. A common method for MDA measurement involves a reaction with thiobarbituric acid (TBA) to produce a compound detected spectrophotometrically (42). TBA reacts with MDA as well as with other nonlipid compounds that lie within the same absorption spectra (15). Measurement of MDA by HPLC may have produced different findings (45).

The increase in GR and GPX activity observed 24 h postinjury is consistent with previous findings in an exercise model (20). These two enzymes are the two most important in the GSH-GSSG cycle and may be activated by increased hydrogen and/or lipid peroxide production (12). Notably, we also observed an increase in GSH and total glutathione contents in the 24-h injured leg. Therefore, it is possible that the glutathione system plays an important role in defending against oxidative damage from muscle stretch injury.

It is worth noting that our animal model of muscle injury is distinct from previous studies of oxidative stress and muscle damage (2, 35, 44, 47). The signs and symptoms associated with exercise-induced muscle damage after repeated lengthening contractions are familiar to anyone who has performed an unaccustomed bout of physical activity. Typically, the symptoms begin 24-48 h after the activity, and the individual often experiences pain, stiffness, and, in severe cases, swelling of the affected muscles (14). In our study, injury is created by using a single stretch and stimulation of the muscle-tendon unit rather than repeated eccentric muscle contractions. Therefore, our model is intended to mimic the clinical injury pattern that occurs with single stretch of the muscle-tendon unit, for example, the so-called "hamstring strain." We attempted to standardize the injury protocol by controlling as many variables as possible to make our comparisons between the two groups (0 and 24 h) valid. Body weight was restricted to control for the animal's age and muscle mass. Tendon shortening was very reproducible when the new customized system was used. Muscles with their ability to absorb mechanical energy serve an important role as shock absorbers for rapid changes in joint position. This property of energy absorption as well as peak torque during eccentric joint rotation and joint torque at tetany are different descriptors of the functional performance of the muscles. Of note is that the injury was reproducible for all of these measures in the two groups (0 and 24 h) of animals, giving a valid starting point for the other assays and comparisons. Furthermore, the hysteresis energy loss of the system during the stimulation and stretch protocol to injure the muscle was the same for both the 0- and 24-h groups. Finally, in all cases, the tissue sample was harvested from the same site at the myotendinous junction where a hematoma was always produced.

An important question not addressed in this study is the source of free radical production after injury. Potential sites for ROS production within muscle include the mitochondrial electron-transport system, membrane-bound oxidases, infiltrating phagocytic cells, cytochrome P-450, and various cytosolic catalytic enzymes (46). Some investigations of muscle injury show that neutrophils may potentiate injury in an ischemia-reperfusion model, as the fiber damage that occurs during reperfusion can be reduced if leukocytes are first depleted (24). This observation led Korthuis and colleagues (24) to suggest that membrane-bound xanthine oxidase is the principal source of free radicals in ischemia-reperfusion injury. More recently, eccentric exercise that leads to skeletal muscle damage has been associated with increased levels of xanthine oxidase in human skeletal muscle (16). In our model, injury is created without a measurable increase in the animal's metabolic rate. Accordingly, our injury probably represents a different type of injury from studies of exhaustive exercise and muscle damage in which the primary source for ROS generation is believed to be the mitochondrial electron-transport chain. Our histological findings in the 24-h tissue samples suggest an abundance of polymorphonuclear leukocytes (neutrophils) at the injury site. Neutrophils contain a number of enzymes, such as myeloperoxidase and NADPH oxidase, that generate free radicals. Other studies have shown that the O₂· causes activation of an extracellular chemotactic factor that can activate neutrophils and, in turn, attract additional neutrophils through superoxide production (33).

A second question that remains unanswered is what exact role free radicals play in muscle damage. It has been hypothesized that ROS contribute to a loss of force production by causing direct damage to cellular tissue (3). However, recent data suggest that alterations in contractile proteins involved in skeletal muscle excitation-contraction coupling may result from changes in the redox state (34). It is also becoming apparent that, in addition to promoting cellular damage, ROS may also initiate and/or amplify inflammation via the upregulation of genes involved in the inflammatory response. This may occur, in certain tissues at least, by the activation of certain transcription factors, such as nuclear factor-B, which amplify the inflammatory response by upregulating production of various cytokines such as interleukin-2 (17) and tumor necrosis factor- (38). Therefore, it is possible that free radicals injure cells and tissues directly through oxidative degradation as well as indirectly, by upregulation of genes involved in the inflammatory response.

In summary, the most significant observation in this study was the presence of increased ROS levels 24 h after mechanical muscle stretch injury. No overt oxidative damage was found in the injured muscle. However, there appeared to be adaptive changes in the cellular glutathione system as well as GPX and GR. The significance and sources of increased ROS levels after in vivo muscle injury remain to be determined.