Endocrinology and Metabolism

Increased muscle proteolysis after local trauma mainly reflects macrophage-associated lysosomal proteolysis

Marie-Chantal Farges, Denis Balcerzak, Brian D. Fisher, Didier Attaix, Daniel Béchet, Marc Ferrara, Vickie E. Baracos


Rat gastrocnemius showed increased protein degradation (+75–115%) at 48 h after traumatic injury. Injured muscle showed increased cathepsin B activity (+327%) and mRNA encoding cathepsin B (+670%), cathepsin L (+298%), cathepsin H (+159%), and cathepsin C (+268%). In in situ hybridization, cathepsin B mRNA localized to the mononuclear cell infiltrate in injured muscle, and only background levels of hybridization were observed either over muscle cells in injured tissue or in uninjured muscle. Immunogold/electron microscopy showed specific staining for cathepsin B only in lysosome-like structures in cells of the mononuclear cell infiltrate in injured muscle. Muscle cells were uniformly negative in the immunocytochemistry. Matrix metalloproteinase-9 (granulocyte-macrophage gelatinase) mRNA and activity were not present in uninjured muscle but were expressed after trauma. There was no activation of the ATP-ubiquitin-proteasome-dependent proteolytic pathway in injured muscle, by contrast to diverse forms of muscle wasting where the activity of this system and the expression of genes encoding ubiquitin and proteasome elements rise. These results suggest that proteolytic systems of the muscle cells remain unstimulated after local injury and that lysosomal enzymes of the inflammatory infiltrated cells are likely to be the major participant in protein catabolism associated with local trauma.

  • injury
  • protein degradation

in a model of blunt trauma to muscle, we demonstrated a period of degeneration lasting ∼3 days, characterized by gross disruption of muscle cells, hemorrhage, inflammation, invasion of the injured site by mononuclear cells, and a 26% loss of previously existing muscle protein (12). A large increase in the process of protein catabolism occurs in injured muscle; however, it is not known which of the several distinct intracellular proteolytic systems of muscle (1, 3) might participate in this response. Lysosomal proteinases are thought to be responsible for degradation of membrane protein and certain soluble proteins in normal muscle (1). A lack of insulin and amino acids leads to activation of the lysosomal process (10). Cathepsin levels are elevated in denervation atrophy, muscular dystrophy, and inflammatory myopathies (23,25). Lysosomal enzymes of mononuclear phagocytes that enter the tissue during inflammation may be present (22). The elevation of lysosomal enzymes in muscle may not be confined to conditions of muscle wasting but also to differentiation and development processes occurring during regeneration (26,37). Muscle also contains Ca2+-activated proteinases, which promote overall degradation under conditions that raise cytosolic Ca2+ levels (4, 8). Cellular disruption would allow focal entry of extracellular Ca2+, possibly provoking Ca2+-dependent proteolysis after injury (4, 8). The bulk of proteolysis in normal muscle and degradation of the contractile proteins (1) involves an ATP-dependent system involving the cofactor ubiquitin and the proteasome complex (17). This may be the primary mechanism by which breakdown of muscle proteins increases in association with diverse forms of atrophy (1, 3, 17, 35, 38, 39, 42).

Matrix metalloproteinases (MMP) responsible for degradation of connective tissue are found in muscle. MMP are regulated at the levels of transcription and zymogen activation by plasmin or membrane type MMP and by tissue inhibitors of metalloproteinases (TIMP) (31). Skeletal muscle shows multiple MMP activities on gelatin zymography and also expresses mRNA encoding MMP-1, -2, -9, -14, and-16 and TIMP-1, -2, and -3, as well as plasminogen activator and its receptor (2). Physiological regulation of this system in muscle has not been extensively characterized; however, the activity, expression, and localization of MMP-9 have recently been reported after experimental injury induced in normal muscle by cardiotoxin injection and denervation (20, 21). This proteinase is mainly produced by inflammatory cells, including polymorphonuclear leukocytes, macrophages, eosinophils (36), and lymphocytes (29) and is involved in the migratory process of these cells in acute inflammation with remodeling and neovascularization (41).

The contribution of proteolytic systems to tissue catabolism after injury to nonmuscle tissue (i.e., systemic response to injury) has been studied. For example, muscle wasting after burn injury in rats has been attributed to the ATP-ubiquitin-proteasome-dependent system (11). Increased gene expression of elements of this system was also observed in peripheral muscle of head trauma victims (27). The relative roles of proteolytic systems after direct trauma to muscle are unknown, and we sought to clarify their nature through study of proteolytic activity, quantitation of mRNA encoding proteinases and their cofactors, and muscle incubation with specific proteinase inhibitors. Because initial studies (12,37) suggested the participation of proteinases potentially derived from muscle cells and/or mononuclear cells of the inflammatory infiltrate, we also determined the localization of the most increased proteolytic activity, cathepsin B, by immunocytochemistry and of its mRNA by in situ hybridization.


Experimental animals.

Studies were carried out in compliance with the guidelines of the Canadian Council on Animal Care. Male Sprague-Dawley rats (200–300 g) from a colony maintained at the University of Alberta were used. Rats were housed in individual wire mesh cages in a temperature (24°C) and humidity (80%)-controlled room on a 12:12-h light-dark cycle. Rats were fed ground laboratory chow (Continental Grain, Chicago, IL) containing 24% crude protein. Rats were killed by CO2 asphyxiation. Animals were allocated by initial body weight to the two treatment groups (control and injured) such that the mean body weights and SE of the groups were similar. Injured rats were administered a single impact trauma to the medial aspect of the right hindlimb. The procedure produced a moderate contusion of the medial gastrocnemius and was conducted while the rats were under general anesthesia (12). Control uninjured rats were also anesthetized. In some experiments, the tissue receiving the direct impact of the device (right medial gastrocnemius) as well as uninjured muscle on the contralateral (left) limb of the same animal were studied. Experiments were carried out with 6–10 rats per treatment. All of the described experiments were repeated at least twice. The results of each treatment are presented as mean values ± SE. Statistical comparisons were made by ANOVA followed by Duncan's test.

A time course study (6, 24, 48, and 72 h posttrauma) was done initially to determine the temporal sequence of induction of proteolysis. Because in vitro protein turnover measurement entails between-day variation, animals were injured at different times and then killed on the same day so that all incubations could be conducted at the same time. In all subsequent studies, control uninjured and injured rats were studied at 48 h after injury, when net protein mobilization and the process of protein catabolism occurred at the most rapid rate.

Lysosomal enzyme activities.

The presence of cytosolic inhibitors of cysteine proteinases (31) precludes direct assay of lysosomal cathepsins in unfractionated muscle extracts. Preparation and purification of lysosomal extracts were done as described previously (6,30). Muscles were homogenized with a polytron in 10 mM potassium phosphate buffer, pH 7.4, containing 0.25 M sucrose, 50 mM KCl, and 1 mM EDTA. An aliquot of homogenate was brought to 0.25% Triton X-100 in acetate buffer, pH 5.0, and stored at −20°C until further analysis of N-acetyl-β-d-glucosaminidase activity and protein content. The homogenate was centrifuged 10 min at 1,000g and then for 10 min at 2,500 g. The supernatant was centrifuged at 20,000 g for 20 min, and the pellet was resuspended in 30 mM sodium phosphate buffer, pH 5.8, and frozen overnight. After thawing, an aliquot was also made up to 0.2% in Triton X-100 and stored for determination ofN-acetyl-β-d-glucosaminidase activity. The supernatant recovered after 20-min centrifugation at 60,000g was designated the lysosomal extract and was used for determination of cathepsin B and B + L activity.N-acetyl-β-d-glucosaminidase activities were determined in lysosomal fractions to estimate the yield of lysosomes (7). Protein concentration was determined according to Bradford (9). Assays for Z-Arg-Arg-aminomethylcoumarin (NMec; cathepsin B) and Z-Phe-Arg-NMec (cathepsins B and L) hydrolysis were carried out according to Barrett (5).

Gelatin zymography.

To detect MMP-2 and MMP-9 activities present in control and injured muscles, samples were prepared and gelatin zymography conducted as described by Balcerzak et al. (2). Briefly, after extraction, soluble proteins (15 μg) were separated on a 15% SDS-PAGE gel containing gelatin (1 mg/ml). After migration, gels were washed in a Triton X-100 solution (2.5% in distilled water), incubated 20 h at 36°C in enzyme buffer (50 mM Tris · HCl, pH 7.5, 10 mM CaCl2, 0.05% Brij-35), and stained with naphthol blue-black solution.

Tissue RNA isolation, Northern hybridization analysis, and RT-PCR.

Total RNA was extracted from frozen samples by the guanidinium isothiocyanate-CsCl method (34). Purity and quantitation of RNA were determined by measures of absorbance at 260 and 280 nm. A cDNA insert encoding rat cathepsin B was subcloned into theEcoRI site of pGEM-blue (Promega, Madison, WI). Antisense riboprobes suited to Northern hybridization and in situ hybridization were generated from this plasmid pGEM rat cathepsin afterHindIII digestion and synthesis with T7 RNA polymerase. The riboprobe is complementary to 289 bases of rat cathepsin B mRNA. Northern hybridization with cathepsin B [32P]cRNA was performed as previously described (6, 42).

Dot blot hybridization was performed as described previously (38) using probes encoding rat cathepsins C and H, mouse cathepsins B, L, and D, rat calpain I, chicken ubiquitin, rat 20S proteasome subunits C2, C3, C5, C8, and C9, and human subunit S5a of the 19S complex. A probe encoding rat mitochondrial rRNA 12S (F17) was used as a control. Radioactivity in dot blots was quantitated using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).

RT-PCR was performed using the Superscript One-Step RT-PCR system (Life Technologies, Burlington, ON, Canada). The MMP-9 product (497 bp) was amplified using 2 μg of total RNA and 10 mM of the following primers: sense GGCAAGGATGGTCTACTGGC; antisense GACGCACATCTCTCCTGCCG. Custom primer synthesis was done for the specified sequences by the DNA Synthesis Laboratory in the Department of Biochemistry at the University of Alberta. The identity of the amplified product was determined by sequencing and comparison with the rat MMP-9 sequence (GenBank accession no. NM031055).

Light microscopy and in situ hybridization analysis.

For light and electron microscopy, control (uninjured) and injured animals were killed and immediately perfused in a single pass with phosphate-buffered physiological saline (PBS) and then with 0.5% glutaraldehyde-4% p-formaldehyde in 0.1 M cacodylate HCl buffer, pH 7.2. Samples of medial gastrocnemius ∼2 mm3(n = 5/muscle) were dissected from the injured area and from the same anatomical location on the contralateral (uninjured) limb. Muscle pieces were postfixed for 2 h at 4°C in a solution containing 4% p-formaldehyde and 4% sucrose in 0.1 M cacodylate HCl buffer, pH 7.2. Finally, tissue pieces were washed in the same buffer containing 7.5% sucrose. This fixation protocol was developed and employed for immunocytochemistry with the use of affinity-purified anti-cathepsin B (14, 24, 40).

Samples for light microscopy collected after fixation by perfusion were frozen in isopentane, cooled in liquid nitrogen, and mounted in OTC compound (Ames, Elkhart, IN) before sectioning (5 μm) in a cryostat (Reichert-Jung, Nussloch, Germany). To determine the general tissue architecture, these sections were stained with hematoxylin and eosin and mounted for light microscopy by standard procedures. In situ hybridization was carried out as described previously (15). Briefly, tissue sections were prepared by treatment with proteinase K (20 μg/ml; 8 min), blocking [10 mM dithiothreitol (DTT), 1.85 mg/ml iodoacetamide, 1.25 mg/ml N-ethylmaleimide for 25 min at 45°C] and acetylation (0.25% acetic anhydride in triethanolamine HCl, pH 8.0). Sections were incubated in a prehybridization medium containing 50% formamide, 5× PIPES, pH 6.8, 5× Denhardt's solution, 2% SDS, 0.25 mg/ml salmon sperm DNA, 0.25 mg/ml tRNA, and 0.1 mM DTT at 43°C for 2 h. The hybridization solution was of the same composition as that used for prehybridization, except that it contained no DTT and included 0.1 g/ml dextran sulfate and an antisense [35S]CTP-cathepsin B riboprobe (120,000 dpm/μl). Hybridization was at 43°C for 4–16 h. After hybridization and rinsing with 4× standard saline citrate (SSC; 0.15 M NaCl, 0.015 M Na citrate)-0.1% mercaptoethanol, sections were digested with ribonuclease A (40 μg/ml) and T1 (800 U/ml) for 30 min at 37°C. Finally, sections were counterstained with hematoxylin and eosin, and liquid emulsion microautoradiography was carried (15). Photomicrography was performed using light field optics. To check the specificity of in situ hybridization reactions, some sections were hybridized with an unrelated antisense riboprobe, [35S]CTP-glucagon, or with a labeled sense transcript of the cathepsin B cDNA.


For electron microscopy, tissue blocks were washed three times with PBS, pH 7.4, and then transferred to PBS containing 0.5% OsO4 for 10 min at 25°C. Tissues were dehydrated in a graded series of ethanol and then transferred to 100% propylene oxide. Samples were embedded in araldite CY212. Thin sections were cut on a Reichert Om U2 ultramicrotome and mounted on nickel grids. All of the following steps were conducted at 25°C. Grids bearing sections were floated face down on 1% NaIO4 for 1 h and then rinsed with distilled water three times for 10 min each. To block nonspecific binding, sections were incubated for 10 min in a solution of PBS that contained 1% bovine serum albumin (BSA), 1% gelatin, and 0.05% Tween-20 and transferred to a solution of 0.02 M glycine-PBS for 5 min. Affinity-purified IgG anti-cathepsin B was a generous gift of Dr. E. Kominami (Juntendo University, Tokyo, Japan). Antisera against cathepsin B were raised in rabbits and purified by affinity chromatography (22). These antibodies have been characterized by immunoblotting and immunohistochemistry (19,40). Affinity-purified IgG anti-cathepsin B was diluted to concentrations of 2.5, 5, or 10 mg/ml in PBS containing 1% BSA, 1% gelatin, and 0.05% Tween-20. The muscle sections were incubated with the primary antibody for 1 h at 25°C on a slow shaker and then washed with PBS containing 1% BSA and 1% gelatin. Protein A-gold (G-3766; Sigma, Oakville, ON, Canada) was used for immunocytochemical staining. Protein A was coupled to colloidal gold particles of the nominal size 10–15 nm. The final solution of protein A-gold was prepared in 0.5% BSA in 0.01 M PBS, pH 7.4, containing 0.05% Tween-20. The grids were placed on drops of protein A-gold at a 20-fold dilution of stock solution and incubated at room temperature for 1 h. The grids were rinsed with PBS and finally rinsed with distilled water. The sections were subsequently stained with 2% uranyl acetate for 30 min and with lead citrate for 5 min. The specimens were examined in a Hitachi H-7000 electron microscope at 75 kV. The specificity of the affinity-purified antibody was determined by immunoblotting on whole skeletal muscle obtained from injured and uninjured control animals. For a negative control for immunocytochemistry, the primary antibody was omitted.

Muscle incubations.

Incubations are routinely done with muscles that are small and thin enough to permit the diffusion of oxygen and substrates into the tissue. The use of surgically prepared longitudinal strips of larger muscles has been validated for studies of muscle protein turnover (43) as well as glucose transport (13). Because it would not have been possible to traumatize some of the more commonly incubated muscles, we dissected four thin longitudinal strips from control and injured medial gastrocnemius. Muscle strips weighed ∼20 mg and were ∼12 mm in length (i.e., dissected from origin to insertion of the gastrocnemius muscle) and not thicker than 1.2 mm. The strips were not mounted on a physical support because of the lack of long tendons on this muscle. Muscle strips were incubated in vitro as described extensively in other work from our laboratories (3, 4,42). Muscles were preincubated in individual flasks containing 3 ml of a modified Krebs-Ringer bicarbonate medium (KRB) composed as previously described (3). In all studies, tissues were incubated at 35°C in medium with 95% O2-5% CO2. KRB buffer contained 1.0 mM CaCl2, 8 nM bovine insulin, 0.5 mM cycloheximide, and 5 mM glucose. A control strip from each injured and uninjured muscle was incubated in this medium, and in a second strip dissected from the same muscles, lysosomal proteinases were inhibited by adding methylamine HCl (4). A separate experiment was conducted to study the possible involvement of proteasome-dependent proteolysis. Media were formulated as indicated above; however, Ca2+-dependent proteases were inhibited by deleting Ca2+ from the medium and adding 10 μg/ml Na dantrolene to prevent release of intracellular Ca2+ (3); methylamine HCL was added as indicated to inhibit lysosomal protease activity. In this approach, the contribution of the non-proteasome-dependent systems is first eliminated. We used a proteasome inhibitor MG 132 (carbobenzoxy-l-leucyl-l-leucyl-l-leucinal; Calbiochem, La Jolla, CA) (10 μM, final concentration) dissolved in dimethyl sulfoxide (35, 39). Control muscles were incubated in the same medium containing an equivalent amount of dimethyl sulfoxide but no inhibitor.

All muscles were preincubated for 1 h and then incubated for 2 h in fresh medium of identical composition. At the end of incubation, muscles were blotted and frozen in liquid nitrogen and stored at −20°C until analysis. Muscle protein mass was determined using the bicinchonic acid procedure (BCA Protein Assay; Pierce Chemical, Rockford, IL) after tissue solubilization in 1.0 N NaOH at 25°C.

Protein degradation was determined as the amount of tyrosine released by the tissue into the medium during incubation in the presence of cycloheximide to prevent amino acid reincorporation into proteins (3). Preliminary studies established that changes in intracellular pools of tyrosine during incubation were small and could be ignored for such measurements. Tyrosine release was linear for up to 120 min of incubation.


Gastrocnemius muscle preparation and effects of trauma on proteolysis.

Rates of protein catabolism in control uninjured gastrocnemius strips fell within a range from 2.09 to 2.55 nmol tyrosine · mg protein−1 · 2 h−1, and these values were similar to those reported for incubated epitrochlearis (3). In preliminary experiments, we determined that the four strips from each muscle had highly similar rates of catabolism (coefficient of variation, 3%). Tissue levels of free tyrosine were 0.625 ± 0.031 nmol/mg protein in control muscles and 0.685 ± 0.043 nmol/mg protein in injured muscles, and these values were not significantly different from each other before or after tissue incubation. Injury resulted in increased release of tyrosine into the incubation medium (Fig. 1). Because tissue levels of tyrosine did not change over the course of incubation and protein synthesis was inhibited, tyrosine appearing in the medium originated from proteolysis. At 6 h after injury, protein degradation tended to rise (+25%, P = 0.149 vs. control) and was significantly elevated at 24, 48, and 72 h.Day 2 posttrauma was the most catabolic day (+ 115% vs. control, P < 0.0001). In different experiments, the magnitude of the proteolytic response on day 2 showed some variation, from +75 to +115%. This time point, which reflected the peak rate of net protein mobilization (12) and the peak rate of protein catabolism, was selected for all further studies.

Fig. 1.

Time course of protein breakdown after injury. Proteolysis was estimated by the release of tyrosine into the incubation medium in the presence of cycloheximide. Medial gastrocnemius from control and injured rats (4 thin strips/muscle) were preincubated for 1 h in Krebs-Ringer bicarbonate buffer containing cycloheximide (0.5 mM), glucose (5 mM), insulin (8 nM), and CaCl2 (1.0 mM). Muscles were transferred to fresh media of the same composition and incubated for a further 3 h. The average value from each of the 4 strips was calculated for each animal. The 4 strips from each muscle had highly similar rates of catabolism (coefficient of variation, 3%). Values not sharing the same letters are significantly different (P < 0.01,n = 4).

Proteinase activity.

We tested for modifications of lysosomal proteinase activity. Similar lysosomal yields were obtained from control and injured muscles as demonstrated by glucosaminidase activity (Table1). Both cathepsin B and B + L activities in the medial gastrocnemius increased after trauma. The large increase in proteolysis observed on day 2 posttrauma was associated with a maximal activity of both cathepsin B and B + L in muscle homogenates (+327% and +123% vs. control, respectively). On day 7 posttrauma, cathepsin activities in lysosomal extracts returned to normal levels compared with control rats; however, these remained elevated in injured muscle homogenates (+96 and +53% for cathepsin B and B + L, respectively, vs. control).

View this table:
Table 1.

Lysosomal enzyme activities increase in injured gastrocnemius

Proteinase gene expression.

Injury was associated with increased expression of cathepsin B mRNA compared with a control gene, glyceraldehyde phosphate dehydrogenase (Fig. 2). Day 2, the most catabolic day, corresponded to the maximal expression of cathepsin B. Results of quantitative dot blot for proteinase gene expression are shown in Table 2. Lysosomal proteinase mRNAs increased (cathepsins B, H, L) or tended to increase (cathepsins C and D) in injured muscle. Of these, cathepsin B showed the largest increase, 6.72-fold, compared with control muscles. No significant changes were seen in expression of calpain, ubiquitin, or proteasome subunit mRNAs.

Fig. 2.

Cathepsin B and glyceraldehyde phosphate dehydrogenase (GAPDH) mRNA in injured (Inj) and uninjured (C, control) gastrocnemius. RNA was extracted from medial gastrocnemius of control rats and both injured and uninjured contralateral limbs of injured rats at 1, 2, or 3 days after injury. Each lane represents pooled RNA from 2 rats/treatment. Northern hybridization analysis was used to detect mRNA encoding cathepsin B and a control gene, GAPDH.

View this table:
Table 2.

Proteinase gene expression in injured and control muscles at 48 h posttrauma

Light microscopy and in situ hybridization.

In uninjured muscle, only scattered silver grains in a random pattern were observed (Fig. 3 B), and this was not different from muscle sections that were treated with an unrelated antisense riboprobe, [35S]CTP-glucagon, or with a sense transcript of the cathepsin B cDNA. In injured gastrocnemius, cathepsin B mRNA localized to the area of tissue damage (Fig.3 A). Dense clusters of silver grains were seen over and around mononuclear cell infiltrates localized in the widened interstitial spaces and around the damaged myofibers.

Fig. 3.

In situ hybridization analysis of cathepsin B mRNA in injured and control gastrocnemius. A: injured muscle at a site of focal damage, showing widened interstitial spaces and mononuclear cell infiltrate. Dense clusters of silver grains can be seen in cells comprising the mononuclear cell infiltrate.B: uninjured muscle. Bar, 50 μm.

Immunoelectron microscopy.

Because of the energy level of the isotope used in in situ hybridization, the silver grains are scattered about the source of radioactivity, and we chose the immunogold technique to more precisely localize cathepsin B at a higher level of resolution. At 2 days posttrauma, muscle damage was characterized by the presence of widened interstitial spaces (endomysial and/or perimysial areas) between the muscles fibers, mononuclear cell infiltration into the interstitial spaces, and disorganization of the muscle architecture. Specific myofiber damage was noted in cross section (Fig.4). In both control and injured muscles, muscle cells were scarcely stained with anti-cathepsin B, and this was not different from sections in which the primary antibody had been deleted. In the interstitial space, cells of the mononuclear cell infiltrate were stained, and gold particles were specifically localized over lysosome-like structures (Fig.5). Combined with the in situ hybridization, these results suggest that the observed increase of cathepsin B activity and mRNA in injured muscle reflects the invasion into the damaged muscle of phagocytes rich in this proteinase.

Fig. 4.

Immunocytochemical localization of cathepsin B in injured gastrocnemius muscle cells. Muscles were treated with affinity-purified anti-cathepsin B. A and B: injured muscle at site of focal damage, showing disorganized myofibrillar structures. No specific staining was observed. Bar, 5 μm.

Fig. 5.

Immunocytochemical localization of cathepsin B in mononuclear cell infiltrate of injured gastrocnemius. Muscles were treated with affinity-purified anti-cathepsin B. A: low-magnification view of mononuclear cell. B and C: gold particles are localized in lysosome-like structures of uniform electron density. Solid bar, 5 μm.

Gelatinase activities and MMP-9 gene expression.

We obtained the pattern of gelatinase activities in control and traumatized muscles using gelatin zymography 48 h after injury (Fig. 6). The conditions of SDS electrophoresis in zymography allow activity of both the uncleaved proenzymes and their cleaved active forms on the gelatin substrate. Uninjured muscles showed only MMP-2, with the two latent forms (66 and 60 kDa) and the active form (55 kDa). The pattern and intensity of MMP-2 activities in injured muscles were identical to those observed in control muscles. Traumatized muscles differed by the presence a strong band corresponding to pro-MMP-9 (100 kDa; Fig. 6 A). The RT-PCR technique confirmed the absence of MMP-9 expression in the control muscle and showed a localization of MMP-9 expression locally in the traumatized gastrocnemius (Fig. 6 B).

Fig. 6.

Gelatinase activities and ,matrix metalloproteinase (MMP)-9 gene expression in injured and uninjured gastrocnemius.A: MMP-2 and MMP-9 activities detected using the gelatin zymography technique in uninjured and injured muscles. Proteins (15 μg) were separated on a 15% SDS-PAGE gel containing gelatin (1 mg/ml). After migration, the gel was washed in a Triton X-100 solution (2.5% in distilled water), incubated 20 h at 36°C in enzyme buffer, and stained with naphthol blue-black solution. B: determination of MMP-9 expression by RT-PCR. Total RNA (2 μg) isolated from a positive control tissue rat lung (L), uninjured muscle (U), and injured muscle (I) was used. The 100-bp DNA ladder is presented in the lanes denoted M.

Inhibition of the proteolytic systems.

We compared untreated strips of each gastrocnemius with other strips from the same muscle incubated with inhibitor. Methylamine (20 mM), an inhibitor of lysosomal acidification, had no significant effect on protein degradation in uninjured muscles but suppressed the increase due to trauma by 67% (Table 3). An inhibitor of proteasome activity substantially inhibited total proteolysis in uninjured muscles but had no effect on the activation of proteolysis induced by injury (Table 4).

View this table:
Table 3.

Lysosomal protein catabolism is activated at 48 h posttrauma

View this table:
Table 4.

Proteasome inhibitor MG132 does not inhibit the activation of protein catabolism at 48 h posttrauma


Contribution of proteolytic systems to trauma-induced proteolysis.

After the experimental injury used here, a degenerative phase of ∼3 days duration is characterized by acute inflammation and muscle protein loss (12). The present study clearly demonstrates that a sustained increase in protein breakdown is one determinant of protein loss previously reported in acute muscle trauma (12). The specific aim of this study was to determine which proteolytic system(s) contribute to muscle degeneration after trauma. Using different approaches, we have shown that it is mainly the lysosomal proteolytic system that is activated during the degenerative phase. Lysosomal proteolysis accounted substantially (∼67%) for the overall increase in protein breakdown based on an in vitro approach with the use of an inhibitor. Concordant results were found with several different approaches. Increased activity of cathepsin B and B + L was found in the atrophying muscles. On day 2 after trauma, these enzyme activities increased both in muscle homogenates and in lysosome fractions in parallel with the rise in the lysosomal process in incubated muscles. Moreover, elevated mRNA levels coding for lysosomal cathepsins were markedly expressed in injured muscle, particularly onday 2 posttrauma.

Cellular source of lysosomal enzymes.

Various cellular components of the muscle tissue were altered after injury (12). At 1–2 days after injury, inflammation appeared to be fully established in the muscle, and this period was characterized by large numbers of mononuclear cells that had not previously been present. Mononuclear cells were seen both in the endomysial connective tissue and within some damaged muscle fibers (12). Mononuclear cells were seen beneath the basement membrane of the muscle cells in focal aggregates. These mononuclear cells had the distinctive morphology of phagocytes and may include tissue macrophages that were previously present, monocytes that were attracted to the site of injury and crossed the vascular wall to become macrophages, B lymphocytes, and cytotoxic T lymphocytes. Mononuclear phagocytes, when stimulated, synthesize and secrete >80 defined molecules, which serve to mediate the inflammatory, antibacterial, and antitumor activities of these cells. Hydrolytic enzymes, and particularly proteinases, figure prominently in the enzymes synthesized by phagocytes. Thus two potential sources for increased lysosomal enzyme levels in injured muscle may be considered: 1) infiltration of mononuclear cells into the muscle tissue and2) activation of the lysosomal system endogenous to the muscle cells. The cellular components released from mechanically disrupted muscle cells and vascular walls provide strong stimuli for the influx of inflammatory cells into the injured site, and these cells have a large capacity for phagocytosis and lysosomal degradation of proteins. We localized cathepsin B and cathepsin B mRNA to determine their potential cellular source(s). Cathepsin B mRNA, studied at the resolution of light microscopy, localized to cells of the mononuclear cell infiltrate and not in muscle cells. At the higher resolution of transmission electron microscopy, immunoreactive cathepsin B was detected only in lysosome-like structures within mononuclear cells.

The presence of inflammatory cell infiltrate and its possible contribution to proteolytic events after injury is additionally suggested by the appearance of MMP-9 activity and mRNA in the injured muscles. MMP-9 activity and mRNA were present only in the traumatized area dissected from the injured medial gastrocnemius. The infiltration of inflammatory cells is linked to the presence of the MMP-9 activity observed, as macrophages and lymphocytes are an important source of this particular gelatinase (29, 36). MMP-9 may be involved in the migration of the inflammatory cells into the traumatized area. By contrast, MMP-2, an MMP constitutively expressed in skeletal muscle and connective tissue cells (2), showed the same level of activity before and after injury.

Without the use of specific stains or markers, the origin of all of the new cells appearing within the injured site cannot be positively identified. Stauber et al. (37) studied cell infiltrates after forced lengthening of muscle. The authors offered evidence that some of the mononuclear cells observed in the injured muscle were phagocytes and some were of myofiber origin (satellite cells), and the latter were possibly responsible for the reestablishment of the myofiber after injury. The role of proteinases in the activities of satellite cells and muscle regeneration following injury is not clearly defined.

Lysosomal proteinases are known to eliminate specific proteins in normal muscle cells such as membrane proteins and soluble enzymes (10). The endogenous lysosomal enzymes of the myofiber are not thought to contribute to the degradation of the myofibrillar proteins actin and myosin under normal physiological and catabolic conditions (1, 3). However, this is likely due to low overall activity and/or accessibility to the substrate, since lysosomal enzymes such as cathepsin B are clearly capable of attacking myosin heavy chain, actin, and troponin T (28). We observed an inhibition of the trauma-associated rise in protein degradation of ∼70% in the presence of methylamine. This inhibitor neutralizes the pH of lysosomal vesicles, indicating that ∼70% of proteolysis occurs inside lysosome/endosomes and is not due to secreted (pro) cathepsins. That observation suggests a mononuclear cell penetration inside the muscle fiber and the participation of a phagocytic process in catabolism rather than liberation of free lysosomal enzymes and is concordant with our ultrastructural observations.

Methylamine did not inhibit ∼30% of the trauma-associated rise in protein degradation, and this could be due to several alternative types of proteolysis, which might be difficult to differentiate experimentally. The fraction not inhibited by methylamine could be from the activation of lysosomal proteinases released outside the mononuclear cells. It may also be due to muscle cellular disruption, increased Ca2+ entry into cells, and consequent activation of calpains (4, 8). The ATP-ubiquitin-proteasome-dependent system degrades myofibrillar proteins and accounts for the bulk of myofibrillar protein catabolism in normal muscle; this system is activated in a large variety of muscle-wasting conditions (reviewed in Ref. 1). By contrast, in our model of local trauma, a proteasome inhibitor, MG132, had no effect on the rise in protein degradation associated with injury, and there was no significant increase of mRNA encoding elements of the ATP-ubiquitin-proteasome system. This is surprising for several reasons. In skeletal muscles, this pathway was recently described to be the critical system responsible for the degradation of myofibrillar proteins in denervation atrophy and fasting, glucocorticoid treatment, disuse atrophy, metabolic acidosis, cancer, and sepsis. All of these conditions are associated with a marked rise in mRNA encoding ubiquitin, enzymes of ubiquitin conjugation, and/or proteasome subunits (1, 3, 17, 35, 38, 39, 42). This has stimulated interest in the ATP-ubiquitin-proteasome system as a possible common pathway for muscle catabolism in diverse forms of atrophy and as a possible site for therapeutic intervention. Highly localized muscle injury would appear to be one of the few instances recorded so far where induction of this system is not a major contributor to catabolic events. However, the lack of detectable activation of the ubiquitin-proteasome pathway reported here further supports a limited role for intramuscular proteolytic systems in the remodeling of muscle after direct injury.

Only a few results are available regarding the participation of proteolytic systems after trauma. In one of the few studies done to date in humans (27), mRNAs encoding multiple elements of the ATP-ubiquitin-proteasome system were increased in peripheral skeletal muscles of head trauma victims. Superficially, these results would seem to contradict those presented here, which suggest that proteolysis in traumatized muscle is largely lysosomal and is associated with mononuclear phagocytes. However, it seems that there is more likely to be both a systemic and a local proteolytic response to injury. A severe injury to the head would induce a systemic response by alterations in hormones such as glucocorticoids, which are known to activate ATP-ubiquitin-proteasome-dependent catabolism in peripheral muscle (1). Because the extent of the injury in our model was limited to a small part of the medial gastrocnemius, it is possible that a systemic response was small or not present and that, within the local environs of injury to muscle, proteolytic activity associated mainly with inflammatory cells prevailed. Although we did not obtain data on glucocorticoid levels in the injured animals, the absence of any reduction in food intake after injury (12) suggests that the overall level of stress was minimal.

A full appreciation of the role of locally and systemically produced inflammatory mediators in muscle injury and regeneration will permit therapeutic strategies to limit excess muscle catabolism and enhance regeneration. The results presented here clarify the role of proteinases in injury-induced local muscle catabolism. Lysosomal enzymes accounted in large part for the increased protein catabolism associated with muscle trauma, and these enzymes and their mRNA localized to cells of the inflammatory infiltrate, not to muscle cells, in the injured tissue. The lack of increase in mRNA of elements of the ATP-ubiquitin-proteasome-dependent proteolytic pathway, considered to be a key participant in diverse forms of muscle wasting, further supports a limited role of intramuscular proteolytic pathways in the remodeling of muscle after local injury.


This research was supported by the Natural Sciences and Engineering Research Council of Canada.


  • Address for reprint requests and other correspondence: V. E. Baracos, Dept. of Agricultural, Food and Nutritional Science, Univ. of Alberta, Edmonton, AB, T6G 2P5, Canada (E-mail:vickie.baracos{at}ualberta.ca).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • 10.1152/ajpendo.00345.2001


View Abstract