INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

EXERTION-INDUCED MUSCLE INJURY is generally described as the presence of structural abnormalities observable at the light- or electron-microscopic level, although molecular and cellular mechanisms are involved in the injury process. Injury, which is common after prolonged, high-intensity, and especially eccentric exercise (3), has been well characterized in limb muscles (3, 17). Contraction of respiratory muscles (RM) is both concentric (41) and relatively low intensity in nature, because only a small proportion of maximum voluntary contractile force is required to sustain resting ventilation (4). Therefore, it may be questioned whether injury of the RM can occur. Several studies have demonstrated that short- and long-term overloads can induce RM injury in animals (20, 21, 30, 31, 44). Interestingly, examination of intercostal biopsies taken before lung surgery showed that 75% of 22 moderate chronic obstructive pulmonary disease patients had RM injury defined as fiber targeting, splitting, and atrophy (7). In other words, RM injury has been described as a consequence of overloading, and, furthermore, the increased load persists as long as breathing continues. Other factors, such as poor nutrition, aging, disuse, hypoxemia, and corticosteroid treatment, which are common in chronic obstructive pulmonary disease patients, may potentiate RM injury (32).

The physiological significance of muscle injury is based on the fact that muscles are functionally impaired (i.e., they lose force, velocity, and endurance) when injured. Furthermore, it has been demonstrated that the relative amount of functional impairment observed in muscle strips examined in vitro is greater than the amount of fibers presenting injury after exercise, as seen in biopsies of diaphragm cross sections studied by light microscopy (17).

Exertion-induced muscle injury is followed by repair. Information on this phenomenon available from limb muscle studies is fragmented, obtained from different species with injuries caused by different agents, and investigated at imprecise times after injury (8). Therefore, it is not yet possible to establish a relationship between the extent or type of injury and recovery, nor is it possible to define the role of RM membrane injury and the time course of expression of genes participating in the muscle repair process. Hence, the objective of this work was to develop an in vivo animal model in which a unique, standardized diaphragm lesion was induced and the time sequence of cellular and molecular events was examined until the muscle cell membranes were repaired and muscle function was recovered. The rationale for the utilization of caffeine as the muscle membrane injury agent was based on both its easy implementation and reproducibility. Furthermore, the amount of injury could be regulated by changing the exposure time or caffeine concentration (15, 40). As well, caffeine induces muscle injury through its ability to activate membrane degrading pathways, such as phospholipolytic enzymes, a process similar to that observed in vivo after strong contractions (see DISCUSSION). This in vivo model allows the time sequence of recovery from diaphragm injury to be studied at the physiological and structural levels. It could be used as well to investigate the time course of molecular factors involved in muscle repair.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Solutions and Reagents

Determination of Optimal Caffeine Concentration and Exposure Time

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Dose-response data determining caffeine concentration and exposure time. A: 30 mM caffeine solution was applied for 1, 3, 10, and 20 min. B: caffeine solution was applied for 10 min at 1, 3, 10, 30, and 100 mM concentrations. Values are means ± SE.

Experimental Procedures

Animal preparation. The guidelines of the Canadian Council on Animal Care were respected for all aspects of the experimental protocols approved by the institutional animal care committee. Experiments were performed on adult (250-300 g) male Sprague-Dawley rats from Charles River (St. Constant, Quebec).

Diaphragm injury experimental model: initial surgical procedure. Animals were anesthetized with pentobarbital sodium (50 mg/kg) via intraperitoneal (IP) injection. During surgery, the animals were able to breathe spontaneously. A 3-cm incision along the linea alba was made caudally from the xyphoid appendix under aseptic conditions. The right hemidiaphragm was exposed by sectioning the diaphragm hepatic ligaments.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Suction cup system used for the application of caffeine solution to the rat diaphragm.

Time course of recovery after the diaphragm injury induction. After the muscle injury was created, nine groups of six animals were studied at 1, 4, 6, 12, 24, 48, 72, 96, and 240 h (10 days). The recovery period was defined as the time elapsed between the end of caffeine exposure and diaphragmatic dissection. Animals in the 1-, 4-, and 6-h groups were kept under anesthesia with supplementary pentobarbital sodium IP injections (5 mg/kg) administered when needed to maintain a depth of anesthesia that prevented responses to tactile stimuli. Animals from the other groups received analgesia (subcutaneous injections of 0.35 ml of 0.3 mg/ml buprenorphine hydrochloride) immediately after the abdominal suture and every 12 h for a maximum of 4 days. The last analgesic injection was never administered within 12 h before diaphragm dissection. Animals from the 1- to 12-h groups were kept in their cages in the laboratory until diaphragm removal. The 24-h to 10-day groups were returned to the animal care facility for recovery from injury.

General procedure. Before diaphragm dissection, the animals received 50 mg/kg pentobarbital sodium IP. The abdominal wall was always opened 20 min before diaphragm removal to cannulate the superior mesenteric vein with an 18-gauge catheter. One milliliter of 3% procion orange was then injected in 1 min. The catheter was sealed to avoid bleeding. Procion orange was used because it is a low-molecular-weight (mol wt = 631) fluorescent tracer dye to which normal intact muscle fibers are impermeable (27, 44). Cells showing intracytoplasmic fluorescence were counted to quantify muscle cell membrane injury. A 15-min period was allowed for vascular and tissue redistribution of procion orange. This was confirmed by yellow coloration of the limb extremities. The diaphragm was excised rapidly en bloc and immersed in equilibrated (95% O₂-5% CO₂, pH 7.38) regular Krebs solution chilled at 4°C for further dissection. The caffeine-exposed area was localized and dissected in two portions in the right hemidiaphragm. Similarly, the control area of the left hemidiaphragm was also dissected in two parts at the same level of the caffeine-exposed area. In some animals, EBD was used instead of procion orange. For its administration, the tail vein was catheterized, and 0.5 ml of a 3% (wt/vol) EBD was injected in 1 min. The remaining procedures were identical.

Processing of muscle samples. From both diaphragm samples, one rectangular muscle block of ~5 × 10 mm was dissected parallel to the long axis of the muscle fibers and frozen immediately at 80°C in an Eppendorf tube for subsequent molecular biology analysis. A second block was quick frozen in isopentane precooled with liquid nitrogen and preserved at 80°C for later sarcolemmal injury analysis.

Assessment of Diaphragm Strip Contractility

The mounted muscle strips were stimulated with a square-wave pulse stimulator (model S48, Grass Instruments, Quincy, MA) via two sets of platinum electrodes mounted in the Plexiglas muscle chamber on either side of the strips. The strips were superfused continuously (5-10 ml/min) with fresh, equilibrated regular Krebs solution throughout the experiment. Procion orange (0.15% wt/vol) was added to regular Krebs solution for the identification of fibers that underwent sarcolemmal injury. At the beginning of each protocol, the diaphragm strip was equilibrated at room temperature (22-25°C) to minimize temperature-dependent deterioration. Initially, submaximal unfused diaphragm strip contractions (stimulation current of 50-75 mA) were obtained with 100-ms trains of single pulses (0.3-ms duration) at a frequency of 10 Hz. Each train of single pulses was delivered at a rate of 0.01 per second for the entire duration of the in vitro contractility experiments. The stimulation current, the frequency of single pulses, and train duration were all progressively increased until final settings of 300-350 mA (supramaximal setting, which is 20% above maximal), 120 Hz, and 600 ms, respectively, were reached. After the supramaximal settings were established, the temperature of the muscle bath and the Krebs solution was slowly raised to 35°C. This period usually lasted a minimum of 10 min. During that period of time, resting tension was maintained at 1 g. After both the bath and the Krebs solution reached 35°C, resting muscle length was adjusted (lengthened) with the micrometer until maximum force was developed. This new resting length was defined as L₀. Once L₀ was identified, a force-frequency curve was obtained for each diaphragm strip in each group at the following frequencies: 120, 100, 50, 30, 20, 10, and 5 Hz. A 100-s resting period was allowed between contractions. The force signals were digitized (DT2821; Data Translation, Marlboro, MA) at a sampling rate of 1,000 Hz and recorded in a computer for later analysis. Force was expressed in newtons per square centimeter, and the muscle cross-sectional area was calculated as the ratio of trimmed muscle mass (g) to strip length (cm) × 1.056 g/cm³ (muscle density) (10). Immediately after the force-frequency experiment, the muscle strips were rinsed in regular Krebs solution for 5 min. Strips corresponding to the caffeine-exposed and control areas were both dissected in two parts. One portion was quick frozen in isopentane precooled with liquid nitrogen and preserved at 80°C for later sarcolemmal injury analysis. The second portion was fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate (pH 7.4) and preserved at 4°C for further embedding and ultrastructural analysis.

Determination of Muscle Injury

Quantification of membrane damage: procion orange technique. Serial sections (10 µm thick) were cut in the transverse plane with a cryostat (Leica Cryocut 1800, Heidelberg, Germany) maintained at 20°C. To assess the level of membrane damage, randomly selected slides were observed at ×100 and ×200 magnification and photographed under epifluorescence microscopy (Carl Zeiss D-7082, Obberkochen, Germany) to identify fibers with membrane permeability defects. Four microscopic fields per hemidiaphragm were analyzed. Muscle fibers demonstrating a clear increase in cytoplasmic fluorescence (i.e., fibers containing procion orange dye) were counted, and the percentage of dye-positive fibers on each diaphragm section was determined. Areas with sectioning artifacts (folds, tears, etc.) were avoided, and the edges of sections were also excluded to avoid fibers potentially damaged by muscle dissection. A minimum of 1,000 diaphragm fibers was counted in each sample.

EBD technique. EBD is mainly characterized by its ability to form a tight complex with serum albumin within seconds of its injection into the bloodstream (28). In view of this, areas of blue macroscopic staining were taken to correspond to albumin entry into muscle fibers (38). By fluorescent microscopy analysis, EBD staining displayed a bright red emission. Muscle fibers demonstrating a clear increase in cytoplasmic fluorescence (i.e., fibers containing EBD) were counted, and the percentage of dye-positive fibers on each diaphragm section was determined. Over 1,000 fibers were counted in each sample.

Ultrastructural Analysis

Hematoxylin and Eosin Staining

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Figure 3 presents typical pictures of transversally cut diaphragms stained with procion orange. Figure 3A is an example of a control diaphragm, which exhibits no intracellular fluorescence after procion orange staining. Figure 3B shows marked injury manifested by cells with cytoplasmic staining. It was taken at the 1-h post-caffeine treatment in the exposed area of the diaphragm. The damage observed in the upper and lower edges was caused by scissors used to cut the muscle strip. Figure 3C represents a diaphragm 96 h after caffeine exposure. In this case, only traces of procion can be seen in a few cells, suggesting that membrane repair was almost complete.

View larger version (102K): [in this window] [in a new window]   Fig. 3.   Transversal diaphragm sections stained with procion orange injected intravenously. A: control diaphragm containing no intracellular procion orange staining. Note the staining in interstitial tissues. B: 1-h post-caffeine-treated diaphragm. C: diaphragm 96 h after caffeine exposure. Only traces of procion can still be seen in a few cells. Scale bar = 200 µm.

Figure 4 depicts the percentage of fibers manifesting membrane injury as a function of recovery time. Compared with control diaphragms (1.3 ± 0.8%), the percentage of fibers with membrane damage was significantly (P < 0.05) higher in caffeine-treated diaphragms at 1, 4, 6, 12, and 24 h (32.9 ± 3.1, 30.3 ± 6.7, 20.7 ± 3.3, 13.4 ± 1.6, and 16.1 ± 4.0%, respectively). The difference was no longer significant 72 h after caffeine exposure (3.4 ± 0.8%). At 96 h and 10 days, the percentages of injured fibers were 2.4 ± 0.6 and 2.7 ± 0.8%, respectively, with no significant difference from the controls.

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Time course of the percentage of fibers showing membrane damage as a function of recovery time. Values are means ± SE.

Figure 5 presents the force-frequency curves obtained from muscle strips at 1, 24, 72, and 96 h and 10 days after caffeine exposure as well as from control strips. Diaphragm force recovered up to 90% of control values 10 days after injury induction. At this time, the tension developed by caffeine-treated strips was not significantly different from that found in control tissues (P < 0.05).

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Force-frequency (f) curves obtained from control (CTL) diaphragm strips and from strips at 1, 24, 72, and 96 h and 10 days after caffeine injury. Values are means ± SE.

Figure 6 shows the force generated by diaphragm strips at different recovery times after injury. Values are expressed as a percentage of the control group. When the muscle was stimulated at 120 Hz, the force at 1-h recovery was 26% of that of the controls. Furthermore, at 72 and 96 h, the force was still only 67% of that of the controls. At 10 days, muscle force returned to 90% of control values and was no longer significantly different (P > 0.05). For comparison, we added in the figure force and injury data published in a previous study (42). In the latter case, injury was induced by subjecting rat diaphragm strips to fatiguing tension-time indexes. Note the similarity of the results.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Relationship between force (expressed as %maximum) and the percentage of noninjured fibers. Each data point is the average of 6 animals. Post-caffeine injury () times were from 1 h to 10 days. Note that the loss of force is proportionally more important compared with the percentage of "membrane-injured" cells. The exercise-induced injury model () has been described previously (42).

Figure 7A is a typical example of a transversal diaphragm section in a healthy animal injected with EBD. It shows interstitial tissue staining without intracellular staining of muscle fibers. Figure 7B represents a 1-h post-caffeine-exposed diaphragm. The appearance of numerous fluorescent fibers suggests that membrane gaps were large enough to allow serum macromolecules such as albumin to enter the cytoplasm.

View larger version (128K): [in this window] [in a new window]   Fig. 7.   A: transversal diaphragm sections in a healthy animal injected with Evans blue dye. B: 1-h post-caffeine-exposed diaphragm stained in the same manner. Scale bar = 50 µm.

Figure 8 presents electron microscope illustrations of diaphragmatic samples at a magnification of ×10,000. Normal muscle ultrastructure was found in the control hemidiaphragm (Fig. 8A). Figure 8B exemplifies an abnormal ultrastructural diaphragmatic architecture observed 1 h after caffeine exposure. However, some areas were completely normal in the same specimen (Fig. 8C), indicating multifocal lesions.

View larger version (81K): [in this window] [in a new window]   Fig. 8.   Electron microscope illustrations of diaphragmatic samples at a magnification of ×10,000. A: normal muscle. B: abnormal ultrastructural muscle architecture found 1 h after caffeine exposure. C: some completely normal areas in the same caffeine-treated specimen. Scale bars = 1 µm.

Figure 9 gives examples of diaphragms stained with hematoxylin and eosin. Figure 9A shows a control diaphragm. In Fig. 9B, slight infiltration of inflammatory cells was already observable as early as 6 h after injury induction. Figure 9C reveals that, at 24 h, more interstitial infiltration is evident with myofiber infiltration. Figure 9D illustrates the diaphragm at 10 days postinjury where interstitial infiltration was decreased and centrally nucleated myofibers became noticeable (arrows).

View larger version (139K): [in this window] [in a new window]   Fig. 9.   Hematoxylin and eosin diaphragm staining. A: control diaphragm. B: diaphragm 6 h after injury induction. C: diaphragm 24 h after injury induction. D: diaphragm at 10 days where inflammatory infiltration was absent but where central nuclei were observed in some fibers (see arrows). Scale bars = 50 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have developed an in vivo model of membrane injury and repair in which a caffeine solution was applied acutely to the abdominal surface of the diaphragm of anesthetized rats. The extent of injury over the next 10 days was evaluated by determining the percentage of fibers into which procion orange leaked, with shifts occurring in the force-frequency curves of strips isolated from diaphragms. One hour after caffeine exposure, 33% of fibers had membrane injury, but diaphragmatic force fell by 70%. Although the extent of injury decreased to only 3% by 72 h, force generation returned to control values only after 10 days.

Rationale for the Use of Caffeine as Injury-inducing Agent

In the present study, we opted for caffeine exposure to induce diaphragmatic membrane injury because the injury pathway is similar to the one followed by neuromuscular-mediated contractions. Caffeine has been used to trigger rapid ultrastructural muscle damage (13, 16) through abrupt rises in intracellular Ca²⁺ (15). It can also induce sarcolemmal membrane injury, as shown in isolated rat hearts (39, 40) and in rat diaphragm strips (43). Caffeine applied to the outer muscle membrane mobilizes intracellular Ca²⁺ from the sarcoplasmic reticulum (SR) and promotes muscle contracture. However, factors contributing to caffeine-induced sarcolemmal permeabilization also seem to facilitate Ca²⁺ influx from the extracellular environment (43). Interestingly, all of these mechanisms of injury induction are similar to those observed in in vivo muscle contractions. Indeed, the process by which muscle exertion leads to injury is thought to operate through the initial creation of gaps in the muscle membrane (3), which allows the influx of extracellular Ca²⁺, causing further muscle contraction. The normal process of excitation-contraction coupling, such as that observed during regular exercise, causes intracellular Ca²⁺ to rise 100-fold from 10⁷ M (normal level in resting muscle) to 10⁵ M (3). This normal physiological increase in cytosolic free Ca²⁺ has been demonstrated by Duncan (13) by using a skinned muscle fiber preparation to cause rapid ultrastructural damage, indicating that it is not necessary to have unusually high concentrations of intracellular Ca²⁺ to induce skeletal muscle damage. The increase in cytosolic Ca²⁺ is also responsible for activation of phospholipase A₂ (PLA₂), which causes sarcolemmal damage (14). Thus both PLA₂ and ultrastructural damage may be activated almost simultaneously by an abrupt rise in cytosolic free Ca²⁺ (14). This abrupt rise may occur subsequent to a large Ca²⁺ influx, such as observed in cardiac muscle (the Ca²⁺ paradox), or may be due to Ca²⁺ liberation from the SR, as seen in malignant hyperthermia (14).

It is thought that abrupt increases in total muscle Ca²⁺ concentration occur either from the disruption of intracellular Ca²⁺ stores, such as in the SR, and/or from extracellular Ca²⁺ leaking into the cytoplasm through the t-tubular system and/or sarcolemmal disruption (5, 24). Of particular relevance to the present study is the fact that the rise in free intracellular Ca²⁺ increases PLA₂ activity and promotes sarcolemmal degradation (19). Howl and Publicover (19) have reported that the Ca²⁺ channel (L-type) agonist BAY K 8644 induces permeabilization of the sarcolemma in mouse (YP strain) diaphragm muscle fibers. In their experiments, the inhibition of either free radical generation (1 mM deferoximine), PLA₂ (5 µM mepacrine), or lipoxygenase (50 µM nordihydroguaiaretic acid) was relatively effective in reducing sarcolemmal damage (19). However, the inhibition of lipoxygenase was not as effective as the inhibition of PLA₂ and free radical production. In fact, Nethery et al. (26) recently demonstrated that the generation of reactive oxygen species in septic animals is dependent, in part, on PLA₂ activation.

The reduction of sarcolemmal damage by PLA₂ inhibition is also supported by other investigators, who reported that creatine kinase efflux was reduced in experimentally damaged skeletal muscle (14). Interestingly, however, inhibition of sarcolemmal damage does not necessarily prevent ultrastructural damage, both of which, as mentioned earlier, are dependent on a rise in intracellular Ca²⁺ (2). Thus we believe that PLA₂ activity triggered by the increase in cytoplasmic Ca²⁺, which was produced by local caffeine application, was responsible for the sarcolemmal damage observed in our experiments. High levels of cytosolic Ca²⁺ elicit sustained muscle contracture, which in itself leads to membrane injury. It may also stimulate the production of free radicals that peroxidate membrane lipids, causing membrane injury (2). Because the Ca²⁺ pathway leading to membrane injury is common to several modes of injury, we decided to develop a caffeine model to eventually study the mechanisms of muscle repair.

The rat diaphragm is particularly well suited for localized topical application of caffeine. It is thin (~20 cells thick), is held distended to resting length by elastic recoil of the lungs, and is easily accessible via abdominal laparotomy. An additional rationale for the use of caffeine is easy local application and control over the amount of injury by modifying its concentration (40) and exposure time (16). Moreover, we obtained a high and reproducible amount of injured fibers 1 h after caffeine exposure in the different experimental animals of our study (Figs. 1 and 4). In addition, our results were similar to those obtained with in vitro caffeine-exposed diaphragmatic strips (31.0 ± 4.4%) (43). Another advantage of our model is the possibility of exposing a limited diaphragmatic area. Therefore, a control area from the same animal can be studied.

Other Caffeine-induced Morphological Damages

Finally, we found ultrastructural damage in some diaphragmatic areas (Fig. 8), in line with previous studies that have used caffeine (13, 15, 16). Sarcomere injury has also been frequently reported in several animal models of contraction-induced injury (11, 31, 44), as well as in human models (34, 35). Such abnormalities are illustrated in Fig. 8 (electron microscope images). However, as recently argued by Devor and Faulkner (11), these images are not meant to be representative of the millions of sarcomeres in a single diaphragm muscle fiber. Light microscopy of transversal sections allows analysis of at least 1,000 fibers, but electron microscopy permits analysis of only a small number of fibers. For example, Roth et al. (34, 35), using electron microscopy, were able to analyze a limited number of fibers (10-40) in biopsies taken from the human vastus lateralis. Indeed, caffeine-induced damage was not present in all sections of the diaphragm. Therefore, the electron microscopic images are simply presented to demonstrate the type of ultrastructural damage that occurs 1 h after caffeine exposure and provide evidence of damaged sarcomeres in our model. Albumin-bound EBD staining was used here as an indicator of the wide range of membrane gaps formed in this injury model.

We found that diaphragm membrane injury gave rise to inflammatory cell infiltration within hours after caffeine exposure. This kind of inflammation has been reported in other diaphragm and limb muscle injury models (20, 31, 37), suggesting that caffeine reproduced phenomena that are usually observed in the course of muscle fiber injury and repair.

Morphological and Functional Recovery From Diaphragm Injury

We observed that the loss of force measured in vitro was markedly higher than the percentage of fibers with membrane damage. In addition, force recovery took more time than membrane repair. These observations are in line with several studies in limb (11, 17, 23) and respiratory (20, 21, 30, 31, 43) muscles. As previously argued (20, 29), many structural changes at the tissue, cellular, and molecular levels with different recovery patterns may inhibit force production. In this model, a 10-day period was necessary to recover 90% of the force developed by control, noninjured muscles. This is in agreement with previous observations in murine limb muscles (17). In the latter example, a peak force deficit of 50%, obtained after contraction-induced injury, took 14 days to recover to 80% of the control value (17). Depending on the extent of peak injury, full recovery of most aspects of normal structure and function may require between 7 and 30 days (17). Furthermore, it is noteworthy that the diaphragm force-injury curve obtained in our caffeine model is compatible with the one produced in vitro by subjecting diaphragm strips to fatiguing tension-time indexes (42).

In conclusion, we have developed an in vivo rat model that allows reproducible membrane diaphragm injury through the application of a local caffeine solution. This model allowed us to study the recovery of membrane injury, which was almost complete within 72 h. Diaphragm force recovery was slower, requiring 10 days to obtain values close to those of the controls. The model is potentially useful for investigating the time course expression of molecular factors involved in muscle repair.