EXPERIMENTAL
AND
Extracellular WILLIAM
MOLECULAR
PATHOLOGY
52,69-86 (I!?%)
Matrix Changes following Skeletal Muscles
T. STAUBER, '*2V~~~~~~
K. FRITZ,'
Blunt Trauma to Rat
ANDBURKHARDTDAHLMANN~
Departments of ‘Physiology and ‘Neurology, West Virginia University, Health Science Center, Morgantown, West Virginia 26506; and ‘Biochemistry Department, Diabetes Forschungsinstitut, Aufm Hennekamp 65, D-#ooO Dusseldorf-l, Federal Republic of Germany Received June 12, 1989, and in revised form September 5, 1989 Myofiber injury-repair was studied in the rat following blunt trauma to the lower leg in order to understand how the inflammatory and regenerative responses of muscles are altered when myofiber rupture is accompanied by bleeding and clotting reactions. A contusion injury to the muscles of the lower hindlimb of the rat was induced by applying an impact force of 4.7 N-m/cm* to one leg. The gastrocnemius and soleus muscles were removed bilaterally and evaluated by histochemical and immunohistochemical techniques to document myoliber, vascular, and connective tissue alterations for several days following insult (6120 hr). A significant increase in wet weight of the gastrocnemius muscle was noted 24 hr postinjury as fluid accumulation and bruising were evident in the muscles resulting from bleeding and intlammation. Vascular disruption was confirmed by the localization of some plasma constituents (fibrinogen, albumin, and complement C,) throughout the interstitial space and even inside some of the damaged myotibers. Inflammation was present and persisted for 5 days as evidenced by continued mast cell degranulation and increased vascular permeability. Using antibodies to identify specific proteoglycans which appear or disappear at various times during muscle regeneration, muscle repair could be followed. The repair process required approximately 10 days for restoration of morphologically intact myofibers. Thus, myoliber repair processes appear to be maintained even after disruption of the vascular system and ischemia following blunt trauma. 8 1990Academic RUSS, ~ttc.
INTRODUCTION A muscle contusion is a common household, occupational, and athletic injury. Apparently, the recovery of muscle from a contusion takes longer than from damage induced by mechanical overload. This difference in recovery time of injured muscles may relate solely to the degree of ischemia imparted by the disruption of the vascular system rather than to any fundamental difference in the repair process of the individual tissues, although information is at present insufficient. If ischemia or lack of adequate blood flow determined the outcome, then muscles with contusion injuries would be more similar to free muscle grafts than exercise-induced myofiber rupture in their time course of recovery and their predisposition to fibrosis. Inflammation in and regeneration of injured tissues represent the physiological responses to any tissue subjected to trauma, and muscle is no exception. The inflammatory response of muscle, however, does not always follow the classic definition of inflammation. Inflammation should involve increased vascular permeability, kallikrein release, mast cell degranulation, and macrophage invasion. In muscle injury where the vascular system remains intact and no bleeding occurs, all of the inflammatory responses exist except for the increase in infiltrating macrophages (Stauber et al., 1988). The purpose of this study was to describe the damage-repair process in muscles which have been subjected to direct trauma and to document if the inflammatory 69 0014-48OOBO $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
70
STAUBER,
FRITZ,
AND
DAHLMANN
response of muscle is different when disruption of the vascular system occurs. Generally, histochemical techniques were employed to observe myofiber, vascular, and connective tissue elements immediately following the insult and for several days afterward. The repair process following blunt trauma was notably longer than that seen in exercise-induced myofiber rupture but contained similar components occurring over a different time course with a few distinct differences. The differences may serve to provide a basis for alternative, rehabilitation protocols or pharmacological interventions for individuals injured from impact trauma. MATERIALS AND METHODS Twenty-one male, SpragueDawley rats (Hilltop Lab Animals, Inc., Scottdale, PA), weighing approximately 300 g each, were used. The use and care of rats in this study was approved by the West Virginia University Animal Care and Use Committee. The animals were provided clean wire cages, food and water ad lib., and maintained on a 12-hr light/lZhr dark cycle at 25°C. All animals were anesthetized to unconsciousness with ethyl ether prior to the experiment. Both hindlimbs between the foot and knee were shaved of all fur for easier positioning and observation. The hindlimb was positioned and held manually on a padded surface with the tibia and fibula outside the area of impact to prevent bone damage. While unconscious, the rats received a single impact to the lateral surface of the right lower leg midway between the heel and knee. The injury was accomplished using an instrument developed and previously described by Stratton ef al. (1984). The impact was delivered by dropping a weighted solid aluminum cylinder (2.7 kg) with a shaped melamine tip (surface area 1.77 cm’) through a tubular guide stabilized by welded braces attached to a 40 x 40-cm* platform through a distance of 31.5 cm. The impact force delivered to the leg using these parameters equaled 4.7 N-m/cm*. The contralateral limb of each animal was not injured but remained untreated to serve as control. The animals recovered consciousness within 10 min and were returned to their cages. The animals were divided into groups of three for tissue removal. At selected times postinjury (6, 12, 24, 48, 72, 96, and 120 hr postinjury), the rats were killed by exsanguination under anesthesia. The wound was examined after removal of the skin and the diameter of the contusion was measured. Then the soleus and gastrocnemius muscles were removed from the animal, weighed, and cut to obtain the injured area for sectioning. With the injured area properly oriented, the muscle samples were mounted on chucks with optimal cutting temperature compound and frozen by submersion in isopentane cooled by liquid nitrogen. The muscles were sectioned at -20°C in an International CT1 cryostat at 4-pm thickness, collected on precleaned glass slides, and air-dried. The slides were stored in airtight containers at - 20°C until use. The Student t test was used to evaluate the changes in wet weights of the muscles. Macrophages were localized in traumatized and control muscles using an a-naphthyl acetate esterase kit (Sigma, St. Louis, MO) according to manufacturers instructions. The slides were observed by light microscopy and photographed. Localization of proteinases, proteins, and proteoglycans were performed by indirect immunohistochemical techniques using fluorescein-labeled second antibody. Antiserum to rat chymase was prepared by conventional methods of immunization in rabbits by Dahlmann et al. (1985). The antiserum was provided as affinity-purified immunoglobulin (IgG) fractions from Affi-gel 15 (Bio-Rad, Rich-
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
71
mond, CA). Antisera to rat fibrinogen, albumin, and complement C3 were obtained commercially (Organon Teknika-Cappel, Malvern, PA). Briefly, the immunohistochemical staining procedure for fibrinogen, albumin complement Cs, and chymase included washing the slides three times for 5 min with filter sterile 0.01 M sodium phosphate in 0.15 M NaCl, pH 7.4 (PBS). Fifty microliters of the appropriate dilution of antiserum was applied to each slide. The slides were incubated at room temperature for 30 min in a moist chamber to prevent evaporation. The antisera were diluted to approximately 1 mg/ml protein as determined by protein assay (Lowry et al., 1951) using albumin as a standard. Following incubation, all slides were washed for 5 min in PBS. Finally, 50 ~1 of fluoresceinlabeled goat F(ab’), anti-rabbit IgG at a 1:20 dilution was added to all experimental and control slides. After another 30 min of incubation and a final wash for 5 min in PBS, glass coverslips were applied over the tissue sections with 50% glycerine 50% PBS containing 1 mg/ml p-phenylenediamine (Sigma, St. Louis, MO) and viewed on a Leitz Orthoplan fluorescence microscope (Stauber et al., 1986). The localization of the plasma proteins was used to establish vascular permeability and cellular membrane integrity. The presence or absence of specific proteoglycans has been useful in identification of muscle injury (where chondroitin 6-sulfate-proteoglycan disappears) and regeneration (where embryonic forms of proteoglycans reappear in the extracellular matrix). Monoclonal antibodies directed against proteoglycans digested with chondroitinase ABC were provided by Dr. Bruce Caterson (University of North Carolina at Chapel Hill, Chapel Hill, NC). Antigen isolation, monoclonal antibody production, and characterization have been completely described previously (Caterson et uI., 1981, 1985). Digestion of the tissue with chondroitinase ABC results in a proteoglycan fragment of small oligosaccharide stubs with at least one characteristic unsaturated disaccharide (0, 4, or 6 sulfated) of the chondroitin sulfate isomer and the linkage region attached to the proteoglycan core protein (Caterson et al., 1987). Localization of specific proteoglycans was performed using a modification of the previously described technique (Bertolotto et al., 1986). The slides were washed for 5 min with 0.05 M Tris/HCl, 0.05 M NaCl, pH 8.0, and predigested with 0.5 units/ml chondroitinase ABC (Sigma, St. Louis, MO) diluted in the same buffer for 30 min at 37°C in a moist chamber. After a 5-min wash in PBS, 50 p.1of diluted specific monoclonal antiserum was added to each slide and incubated as before. Following incubation, all slides were washed again in PBS followed by the application of fluorescein-labeled goat F(ab’), anti-mouse IgG or IgM at a 1:20 dilution. After another 30 min of incubation the slides were washed with PBS and glass coverslips were applied as described above. The slides were viewed on the fluorescence microscope. Method specificity and antibody specificity for the immunohistochemical reactions were determined according to the criteria of reliability established by Petrusz et al. (1980). The controls for the fluorescent immunohistochemical experiments included normal and injured tissue sections either incubated with normal rabbit serum instead of specific antibody or culture media instead of monoclonal antibody followed by fluorescein-labeled antiserum. Controls for the monoclonal antisera to proteoglycans included fresh frozen sections of rat ear as positive controls treated in the same manner as the normal and injured muscle tissues. Each monoclonal antibody to proteoglycans was localized in the cartilag-
72
STAUBER,
FRITZ,
AND
DAHLMANN
inous portion of the rat ear. The specificity of the monoclonal antibodies was tested by excluding the chondroitinase ABC digestion prior to monoclonal antibody application followed by fluorescein-labeled second antibody. Another group received only chondroitinase ABC, diluted mouse serum, and fluoresceinconjugated second antibody. Normal and injured muscle sections incubated with monoclonal antibodies to proteoglycans without prior chondroitinase ABC digestion or vice versa were devoid of any specific fluorescent localization. Fluorescein-conjugated concanavalin A (Con A; Vector Laboratories, Burlingame, CA) was used to localize nonspecific glycoproteins rich in mannose residues such as those found in the pericellular basement membrane of the basal lamina (Gulati et al., 1983). Fluorescein-labeled Con A at a I:200 dilution in PBS was applied directly to slides of normal and injured muscles after an initial wash for 5 min in PBS. Following a second wash in PBS, coverslips were applied and the samples were photographed after being viewed under the microscope. RESULTS Following blunt trauma to the lower leg muscles, the animals did not exhibit any abnormal behavior. Food and water intake and ambulation appeared normal. At the specific time for tissue removal, the hindlimbs of each animal were examined for evidence of damage. The skin of the injured right calf was not broken but had a large ecchymotic area on the lateral surface of the calf. The calf area appeared enlarged with a palpable hardened area at the site of injury. Removal of the skin revealed fluid accumulation between the skin and the fascia, hematoma, and bruising of the gastrocnemius muscle extending to the Achilles tendon. The soleus muscle was also bruised but only at the immediate site of impact. The muscles were still attached to their origins and insertions and there was no evidence of bone damage. The diameter of the ecchymotic area measured on the lateral surface of the gastrocnemius muscle decreased with time. At 24 hr following trauma, a lesion of 15 mm in diameter was recorded. This decreased to 10 mm at 48 hr, and by 120 hr, it measured only 5 mm in diameter. A significant increase in the wet weight of the gastrocnemius muscles was recorded at 24 hr postinjury (P < 0.05; Table I). While the gastrocnemius muscle weight remained elevated above controls beyond 48 hr, these differences were not significant. By 120 hr postinjury, the weight of the gastrocnemius muscle had actually decreased. Microscopic evidence of muscle damage to the gastrocnemius included disorganization of the muscle architecture, widened interstitial spaces, fascicle disruption, and accumulation of mononuclear cells in the interstitial spaces. At 24 hr postinjury specific myoliber damage was noted with flocculent degeneration of the sarcoplasm and/or macrophage infiltration (esterase-positive cells) (Fig. lA, arrow). The interstitial space at this time probably contained fluid as few identifiable mononuclear cells appeared between myofibers. The increased fluid content would account for the increased turgor and wet weight observed in the muscle. Mononuclear cell infiltration increased in the interstitial spaces through 72 hr with further evidence of macrophage involvement (Fig. lB, arrows). Myofiber degeneration was observed in samples removed at 96 hr with fiber remnants and cell fragments throughout each tissue section (Fig. 1C). Fewer esterase-positive cells and necrotic fibers were observed past 96 hr; however, the muscle architecture remained disrupted through 120 hr (Fig. 1C). Evidence of damage to the
ECM
CHANGES
FOLLOWING
BLUNT
73
TRAUMA
TABLE I Wet Weights (in Grams) of Injured and Control Muscles
Time period 24 hr Gastrocnemius Soleus 48 hr Gastrocnemius Soleus 12 hr Gastrocnemius Soleus 96 hr Gastrocnemius Soleus 120 hr Gastrocnemius Soleus
Injured mean + SEM (n =3)
Control mean f SEM (n = 3)
1.43 rt 0.089* 0.22 2 0.046
1.17 2 0.026 0.21 + 0.018
1.29 f 0.176 0.20 f 0.019
1.15 2 0.015 0.19 * 0.011
1.23 f 0.168 0.20 + 0.02
1.13 2 0.066 0.19 ” 0.009
1.24 _’ 0.097 0.18 k 0.02
1.13 + 0.057 0.19 f 0.016
0.89 2 0.083 0.18 2 0.024
1.06 2 0.047 0.18 k 0.02
* Significant at P 0.05.
soleus muscle was limited to a few damaged myotibers per fascicle with cellular infiltration only in these damaged myofibers. By 120 hr there was little evidence of damage observed on sections of soleus muscle. Evidence of vascular disruption was observed using specific antisera for plasma constituents: fibrinogen, albumin, complement C,, and high molecular weight cysteine proteinase. Immediately following injury, all of these plasma proteins were localized throughout the widened interstitial spaces around damaged myofibers. In addition, immunoreactivity was present inside some myofibers of the gastrocnemius muscle, indicative of sarcolemmal rupture (Fig. 2A). Albumin and complement C, had similar distributions in both the injured gastrocnemius and soleus muscles (Fig. 2B). Chymase localization revealed degranulation of mast cells from 12 through 120 hr in gastrocnemius muscle samples. In muscle samples from uninjured muscles, mast cells (chymase positive) were evidence in both the endomysium and the perimysium around blood vessels (Fig. 2C) but without degranulation. Thus, only a localized mast cell response to damage was noted following trauma which was especially evident around vascular areas (Fig. 2D) with chymase remaining associated with the extruded granule matrix. Extracellular matrix disruption was observed in the gastrocnemius muscle immediately after injury followed by an increased widening of the interstitial space caused by the deposition of extracellular matrix components. In the soleus muscle, proliferation of extracellular matrix also was evident but on a smaller scale. Details of the reorganization of the extracellular matrix are provided below. Both the soleus and gastrocnemius muscles showed alterations in Con A staining which helped distinguish undamaged, absent, and regenerating fibers. In control muscles, Con A outlined the perimysium, endomysium and epimysium around myofibers, blood vessels, nerves, and capillaries (Fig. 3A). Following injury, the endomysium and perimysium appeared disrupted around myofibers. In
74
STAUBER,
FRITZ,
AND
DAHLMANN
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
75
FIG. 3. Histochemical localization of Con A in control and injured rat muscles. Cross sections of gastrocnemius muscle. (A) Control. (B) 24 hr postinjury; note the absence of stained material around the periphery of the myofibers (arrow). (C) 48 hr postinjury. (D) 72 hr postinjury, Con A-positive material fills areas around infiltrating cells (arrow). (E) % hr postinjury, small fibers appear among Con A-positive material (arrow). Cross sections of soleus muscle. (F) 24 hr postinjury. (G) 72 hr postinjury. (H) % hr postinjury. X240.
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
77
FIG. 4. Immunohistochemical localization of chondroitin f5-sulfate proteoglycan in control and injured rat muscles. Cross sections of gastrocnemius muscle. (A) Control. (B) 24 hr postinjury; note the proteoglycan positive cells (white arrow) and proteoglycan stained remnants from former myofibers (black arrow). (C) 72 hr postinjury. (D) % hr postinjury; note the smaller outlined myofibers in close proximity to mature fibers (arrow). (E) 120 hr postinjury. Cross sections of soleus muscle. (F) % hr postinjury. (G) 120 hr postinjury. (H) Gastrocnemius muscle, chondroitin &sulfate proteoglycan control without chondroitinase ABC predigestion. x 240.
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
79
80
STAUBER,
FRITZ,
AND
DAHLMANN
some cases, myofibers appeared separated from the matrix as Con A-positive material was absent around the periphery of these fibers (Fig. 3B, arrow). The muscle architecture showed continued disorganization through 48 hr (Fig. 3C), and by 72 hr, Con A-stained material tilled the areas of presumably absent myofibers (Fig. 3D, arrow). The apparent density of the matrix material continued to increase through 96 hr, filling in all of the wide extracellular space between existing myotibers. Small regenerating myofibers appeared in the midst of Con A-positive material at this time (Fig. 3E, arrow). In the injured soleus muscles, Con A-stained material appeared in the widened interstitial space around damaged myofibers while a more regular staining pattern was observed around the undamaged fibers of the section (Fig. 3F). Some damaged myotibers in the injured soleus muscles appeared with Con A-positive staining present inside the fibers. By 72 hr, these apparently damaged myofibers were more heavily stained with Con A-positive material (Fig. 3G). Through 96 hr, groups of small myofibers were seen surrounded by a thick Con A-stained matrix (Fig. 3H). Thus, some type of limited myofiber repair was consistently observed in all of the injured soleus muscles examined. Chondroitin 6-sulfate proteoglycan was localized in the endomysial area around myofibers, blood vessels, and nerves in the normal gastrocnemius muscle (Fig. 4A). This sulfated proteoglycan filled the endomysial space between normal myofibers. Following injury, the localization of chondroitin 6-sulfate proteoglycan revealed (1) a disrupted arrangement of proteoglycan material within the widened interstitial space, (2) the absence of a continuous proteoglycan matrix around disrupted myofibers, and (3) the presence of proteoglycan-positive cells in areas of absent myotibers (Fig. 4B, white arrow). Additionally, some proteoglycan matrix material remained, outlining the position of former myofibers (Fig. 4B, black arrow). At 72 hr, many proteoglycan-positive cells were evident near existing myofibers (Fig. 4C). However, there was no proteoglycan staining associated with the cell membrane of the existing myofibers. Groups of small myofibers outlined with chondroitin 6-sulfate proteoglycan matrix were seen in the damaged areas around existing myofibers at 96 hr (Fig. 4D). Interestingly, these small fibers (Fig. 4D, arrow) were always in close proximity to larger, mature fibers. By 120 hr after muscle trauma, a more organized myofiber architecture was observed. However, thickened areas of proteoglycan material were still evident between fibers (Fig. 4E). Chondroitin 6-sulfate proteoglycan localization was absent around the few damaged myofibers in the soleus muscle while it remained apparent in the endomysial region of undamaged myofibers (Fig. 4F). Like the gastrocnemius muscle, thickened regions of proteoglycan material were evident at 120 hr with proteoglycanpositive cells present in the interstitial spaces (Fig. 4G). No extracellular matrix staining was observed in muscle sections incubated with this monoclonal antibody whenever predigestion with chondroitinase ABC was excluded (Fig. 4H). Localization of unsulfated chondroitin and chondroitin 4-sulfated proteoglycan revealed less change. In control muscle, unsulfated chondroitin proteoglycan was localized around myofibers, capillaries, and blood vessels in the endomysial and perimysial regions (Fig. 5A). Following injury, no unsulfated chondroitin proteoglycan was observed until 72 hr when patchy fluorescent localization could be seen around myolibers in injured areas (Fig. 5B). No fluorescent localization was observed in the interstitial spaces among mononuclear cells, small regenerating
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
81
a2
STAUBER,
FRITZ,
AND
DAHLMANN
fibers, or the matrix material that filled this area. In contrast, chondroitin 4-sulfate could not be localized around or between normal or injured gastrocnemius myofibers at 24 hr (Fig. 5C). However, small fiber-like structures located in the interstitial space were stained around their periphery with positive material at 96 hr (Fig. 5D). No other chondroitin 4-sulfate proteoglycan-positive staining was observed in the injured areas of the tissue after this time. Fibronectin localization changed as early as 24 hr postinjury. In normal gastrocnemius muscle, fibronectin predominated in capillaries with some patchy localization in a thin band around myofibers (Fig. 6A). During the first 48 hr postinjury, tibronectin could be observed filling the widened endomysium around damaged fibers in addition to its presence inside damaged myofibers (Fig. 6B). Fibronectin localization was most obvious around small fibers through 96 hr in addition to its presence among the cellular infiltrate (Fig. 6C). Fibronectin filled the interstitial space by 120 hr in a pattern similar to that observed with Con A (Fig. 6D). Dense staining was again noticed around small fiber-like structures in close proximity to surviving, mature myofibers. DISCUSSION Skeletal muscle injury is generally followed by a series of processes that results in some form of repair to the damaged tissue. These processes include inllammation, satellite cell activation, myogenesis, ftbroblast proliferation, and connective tissue reorganization. The magnitude of the injury and the degree of disruption of the vascular supply may determine the time course, overall involvement of the various cell populations, and the final outcome of the repair of a functional muscle. The repair can vary from complete restoration of an essentially normal muscle (Fritz and Stauber, 1988; Stauber ef al., 1988) to a muscle with extensive fibrosis (Lehto et al., 1986) and restricted motion (e.g., contracture). To study the histopathology of injured skeletal muscles, a contusion injury was produced by impact trauma to the lower leg of a rat. Impact trauma to the leg resulted in extensive damage to the gastrocnemius and to a lesser degree the soleus muscles. The damage was complex in nature including crushed myofibers, disrupted fascia, and ruptured blood vessels with evidence of bleeding and hematoma formation. It was also possible, although not readily evident, that some local nerve damage resulted. Using histochemical techniques, indicators of inflammation, cell infiltration, and connective tissue reorganization were monitored for 5 days following trauma in order to document the early histopathology of muscle following a contusion injury. The extended durations of both the inflammation and the repair process were uniquely different than muscle injury from a mechanical overload (Stauber et al., 1988) which also has a complex, multifaceted repair mechanism involving muscle (Stauber et al., 1988), connective tissue (Fritz and Stauber, 1988), and even cells of the reticuloendothelial system (Stauber et al., 1988). An acute inflammatory response with swelling resulted from damage to the vascular, myofiber, and connective tissues of the gastrocnemius muscle. The increased wet weight recorded for the gastrocnemius muscle and the large ecchymotic area observed externally on the lateral surface of the muscle indicated that hemorrhage and fluid accumulation were present in the damaged muscle. Within 12 hr, the plasma components, fibrinogen, albumin, and complement C,, were localized throughout the interstitial space and even inside some of the myofibers,
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
83
84
STAUBER,
FRITZ,
AND
DAHLMANN
indicative of extensive vascular leakage and sarcolemmal damage. The presence of these and other humoral factors in the interstitial space would act as chemotactic agents attracting mononuclear cells responsible for removing damaged tissue-another feature of inflammation following injury. The extensive mononuclear cell infiltration observed included numerous macrophages in and around damaged myofibers. Together with the observed degranulation of mast cells, the release of a variety of known mediators of inflammation such as histamine, lysosomal enzymes, and interleukin-1 (Larsen and Henson, 1983) would have occurred. The duration of the inllammatory response was longer than that seen following myotiber rupture without bleeding (Stauber et al., 1988) as the serum components remained elevated in the gastrocnemius muscle over the entire 5-day period. The extended duration of the inllammatory response in concert with damage to the basal lamina probably resulted from vascular disruption and ischemia which together would impair the repair process. Evidence of a delay in the repair of the damaged muscle was also found in the sequence of extracellular matrix changes observed. In mature muscle, proteoglycans reside as a thin band around myofibers; however, following injury, proliferation of matrix material occurs similar to that seen during embryogenesis. The early disorganized appearance of the matrix with a loss of chondroitin 6-sulfate proteoglycan (an indication of degeneration) was probably the result of crushing and fluid accumulation leading eventually to digestion by hydrolytic enzymes. Chondroitin 6-sulfate proteoglycan was not observed again until 4 days later when it was seen surrounding small fibers indicating that the repair process was present. Prior to the reappearance of chondroitin 6-sulfate proteoglycan, unsulfated chondroitin proteoglycan was observed in the granulated area. Since unsulfated chondroitin proteoglycan is synthesized by perfusion-myoblasts and chondroitin 6-sulfate by postfusion cells (Hutchison and Yasin, 1986), the proteoglycan staining pattern was similar to that seen during muscle development and repair following exercise-induced overload (Fritz and Stauber, 1988). Thus, a similar repair process followed muscle contusion but required a longer time period. Fibronectin requires special consideration because two types of fibronectin are present and cannot be differentiated by present immunohistochemical techniques (Gulati et al., 1982; Kleinman et al., 1981). Fibronectin is a normal constituent of plasma and follows the expected route of albumin and other blood constituents when vascular damage is sustained. A separate tissue fibronectin is also present and resides as a thin band along with the proteoglycans around each myofiberpart of the extracellular matrix or attached to it. Both fibronectin and chondroitin 6-sulfate proteoglycan appeared more prominently around small presumptive regenerating fibers at approximately 4 days postinjury. Chondroitin 4-sulfate proteoglycan could only be localized at 96 hr. Taken together, this evidence would indicate that myoblast fusion had occurred around the third or fourth day since the presence of these matrix components correlates with observations of myotube fusion by others (Gulati et al., 1982; Kleinman et al., 1981). Muscle tissue repair seems to follow a prescribed pattern of processes that involve muscle and nonmuscle components. Overall the repair process is independent of the type of injury with regard to the nature of myogenesis and con-
ECM CHANGES FOLLOWING
BLUNT
TRAUMA
85
nective tissue proliferation but differs in time course if vascular flow is arrested. Thus, free muscle grafts, muscle damage following ischemia such as during tourniquet application, and contusion injury can, in many ways, be considered equivalent. Once adequate blood flow has been provided, the regeneration of muscle occurs. If the restoration of blood flow is slow, the connective tissue proliferation seemsto predominate and fibrosis can result. This conclusion supports a rationale that rehabilitation or pharmacologic intervention following injuries that result in ischemic damage to skeletal muscle should concentrate on techniques to restore blood flow and capillary regrowth while inhibiting connective tissue proliferation. ACKNOWLEDGMENTS This work was supported in part by funds from Comptex, Inc., to W.T.S. The authors thank Bruce Caterson, Ph.D., University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, for a liberal supply of monoclonal antibodies to unsulfated chondroitin, chondroitin Csulfated, and chondroitin 6-sulfated proteoglycans. The authors also thank Steve Stratton, Ph.D., San Antonio, Texas, for the loan and use of the traumatizing unit.
REFERENCES A., PALMUCCI, L., GAGLIANO, A., MONGINI, T., and TARONE, G. (1986). Immunohistochemical localization of chondroitin sulfate in normal and pathological human muscle. J. Neural. Sci. 73,233-244. CATERSON, B., BAKER, J. R., CHIUSTNER, J. E., KEARNEY, J. F., and STROHRER, R. L. (1981). The characterization of clonal antibodies directed against bovine nasal cartilage proteoglycan and link protein. In “Monoclonal Antibodies and T-cell Hybridomas” (G. J. Hammerling, B. Hammerling, and J. F. Keamey, Eds.), pp. 259-268. Elsevier/North-Holland, New York. CATERSON, B., CALABRO, T., and HAMPTON, A. (1987). Monoclonal antibodies as probes for elucidating proteoglycan structure and function. In “Biology of Proteoglycans” (T. N. Wight and R. P. Mecham, Eds.), pp. l-26. Academic, New York. CATERSON, B., CHRISTNER, J. E., BAKER, J. R., and COUCHMAN, J. R. (1985). Production and characterization of monoclonal antibodies directed against connective tissue proteoglycans. Fed. Proc. 44, 386-393. DAHLMANN, B., KUEHN, L., RUTSCHMANN, M., and REINAUER, H. (1985). Purification and characterization of a multicatalytic high-molecular-mass proteinase from rat skeletal muscle. Biochem. J. 228, 161-170. FRITZ, V. K., and STAUBER W. T. (1988). Characterization of muscles injured by forced lengthening. II. Proteoglycans. Med. Sci. Sports Exercise 20, 354-361. GULATI, A. K., REDDI, A. H., and ZALEWSKI, A. A. (1982). Distribution of tibronectin in normal and regenerating skeletal muscle. Anat. Rec. 204, 175-183. GULATI, A. K., REDDI, A. H., and ZALEWSKI, A. A. (1983). Changes in the basement membrane zone components during skeletal muscle fiber degeneration and regeneration. J. Cell Biol. 97, 957-%2. HUTCHISON, C. J., and YASIN, R. (1986). Developmental changes in sulphation of chondroitin sulphate proteoglycan during myogenesis of human muscle cultures. Dev. Biol. 115, 7883. KLEINMAN, H. K., KLEBE, R. J., and MARTIN, G. R. (1981). Role of collagenous matrices in the adhesion and growth of cells. J. Cell Biol. 88, 473-485. LARSEN, G. L., and HENSON, P. M. (1983). Mediators of inflammation. Annu. Rev. Immunol. 1, 335-359. LEHTO, M., JARVINEN, M., and NELIMARKKA, 0. (1986). Scar formation after skeletal muscle injury: A histological and autoradiographical study in rats. Arch. Orthop. Trauma Surg. 104, 366-370. LOWRY, 0. H., ROSENBROUGH, J. N., FARR, A. L., and RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. PETRUSZ, P., ORDRONNEAU, P., and FINELY, J. C. W. (1980). Criteria of reliability for light microscopic immunocytochemical staining. Histochem. J. 12, 333-348. STAUBER, W., FRITZ, V., DAHLMANN, B., GAUTHIER, F. KIRSCHKE, H., and ULRICH, R. (1985). BERTOLO~O,
86
STAUBER,
FRITZ,
AND
DAHLMANN
Fluorescence methods for localizing proteinases and proteinase inhibitors in skeletal muscle. Histochem. STAUBER,
J. 17, 787-796.
W. T., FRITZ, V. K., DAHLMANN, B., and REINAUER, H. (1986). Immunofluorescent localization of an alkaline proteinase in skeletal muscles from diabetic rats. Basic Appl. Histochem. 30, 147-152. STAUBER, W. T., FRITZ, V. K., VOGELBACH, D. W., and DAHLMANN, B. (1988). Characterization of muscles injured by forced lengthening. I. Cellular Infiltrates. Med. Sci. Sports Exercise 20, 345-353. STRATTON, S. A., HECKMANN, R., and FRANCIS, R. (1984). Therapeutic ultrasound: Its effects on the integrity of a nonpenetrating wound. J. Orthop. Sports Phys. Ther. 5, 271281.
AND
Extracellular WILLIAM
MOLECULAR
PATHOLOGY
52,69-86 (I!?%)
Matrix Changes following Skeletal Muscles
T. STAUBER, '*2V~~~~~~
K. FRITZ,'
Blunt Trauma to Rat
ANDBURKHARDTDAHLMANN~
Departments of ‘Physiology and ‘Neurology, West Virginia University, Health Science Center, Morgantown, West Virginia 26506; and ‘Biochemistry Department, Diabetes Forschungsinstitut, Aufm Hennekamp 65, D-#ooO Dusseldorf-l, Federal Republic of Germany Received June 12, 1989, and in revised form September 5, 1989 Myofiber injury-repair was studied in the rat following blunt trauma to the lower leg in order to understand how the inflammatory and regenerative responses of muscles are altered when myofiber rupture is accompanied by bleeding and clotting reactions. A contusion injury to the muscles of the lower hindlimb of the rat was induced by applying an impact force of 4.7 N-m/cm* to one leg. The gastrocnemius and soleus muscles were removed bilaterally and evaluated by histochemical and immunohistochemical techniques to document myoliber, vascular, and connective tissue alterations for several days following insult (6120 hr). A significant increase in wet weight of the gastrocnemius muscle was noted 24 hr postinjury as fluid accumulation and bruising were evident in the muscles resulting from bleeding and intlammation. Vascular disruption was confirmed by the localization of some plasma constituents (fibrinogen, albumin, and complement C,) throughout the interstitial space and even inside some of the damaged myotibers. Inflammation was present and persisted for 5 days as evidenced by continued mast cell degranulation and increased vascular permeability. Using antibodies to identify specific proteoglycans which appear or disappear at various times during muscle regeneration, muscle repair could be followed. The repair process required approximately 10 days for restoration of morphologically intact myofibers. Thus, myoliber repair processes appear to be maintained even after disruption of the vascular system and ischemia following blunt trauma. 8 1990Academic RUSS, ~ttc.
INTRODUCTION A muscle contusion is a common household, occupational, and athletic injury. Apparently, the recovery of muscle from a contusion takes longer than from damage induced by mechanical overload. This difference in recovery time of injured muscles may relate solely to the degree of ischemia imparted by the disruption of the vascular system rather than to any fundamental difference in the repair process of the individual tissues, although information is at present insufficient. If ischemia or lack of adequate blood flow determined the outcome, then muscles with contusion injuries would be more similar to free muscle grafts than exercise-induced myofiber rupture in their time course of recovery and their predisposition to fibrosis. Inflammation in and regeneration of injured tissues represent the physiological responses to any tissue subjected to trauma, and muscle is no exception. The inflammatory response of muscle, however, does not always follow the classic definition of inflammation. Inflammation should involve increased vascular permeability, kallikrein release, mast cell degranulation, and macrophage invasion. In muscle injury where the vascular system remains intact and no bleeding occurs, all of the inflammatory responses exist except for the increase in infiltrating macrophages (Stauber et al., 1988). The purpose of this study was to describe the damage-repair process in muscles which have been subjected to direct trauma and to document if the inflammatory 69 0014-48OOBO $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
70
STAUBER,
FRITZ,
AND
DAHLMANN
response of muscle is different when disruption of the vascular system occurs. Generally, histochemical techniques were employed to observe myofiber, vascular, and connective tissue elements immediately following the insult and for several days afterward. The repair process following blunt trauma was notably longer than that seen in exercise-induced myofiber rupture but contained similar components occurring over a different time course with a few distinct differences. The differences may serve to provide a basis for alternative, rehabilitation protocols or pharmacological interventions for individuals injured from impact trauma. MATERIALS AND METHODS Twenty-one male, SpragueDawley rats (Hilltop Lab Animals, Inc., Scottdale, PA), weighing approximately 300 g each, were used. The use and care of rats in this study was approved by the West Virginia University Animal Care and Use Committee. The animals were provided clean wire cages, food and water ad lib., and maintained on a 12-hr light/lZhr dark cycle at 25°C. All animals were anesthetized to unconsciousness with ethyl ether prior to the experiment. Both hindlimbs between the foot and knee were shaved of all fur for easier positioning and observation. The hindlimb was positioned and held manually on a padded surface with the tibia and fibula outside the area of impact to prevent bone damage. While unconscious, the rats received a single impact to the lateral surface of the right lower leg midway between the heel and knee. The injury was accomplished using an instrument developed and previously described by Stratton ef al. (1984). The impact was delivered by dropping a weighted solid aluminum cylinder (2.7 kg) with a shaped melamine tip (surface area 1.77 cm’) through a tubular guide stabilized by welded braces attached to a 40 x 40-cm* platform through a distance of 31.5 cm. The impact force delivered to the leg using these parameters equaled 4.7 N-m/cm*. The contralateral limb of each animal was not injured but remained untreated to serve as control. The animals recovered consciousness within 10 min and were returned to their cages. The animals were divided into groups of three for tissue removal. At selected times postinjury (6, 12, 24, 48, 72, 96, and 120 hr postinjury), the rats were killed by exsanguination under anesthesia. The wound was examined after removal of the skin and the diameter of the contusion was measured. Then the soleus and gastrocnemius muscles were removed from the animal, weighed, and cut to obtain the injured area for sectioning. With the injured area properly oriented, the muscle samples were mounted on chucks with optimal cutting temperature compound and frozen by submersion in isopentane cooled by liquid nitrogen. The muscles were sectioned at -20°C in an International CT1 cryostat at 4-pm thickness, collected on precleaned glass slides, and air-dried. The slides were stored in airtight containers at - 20°C until use. The Student t test was used to evaluate the changes in wet weights of the muscles. Macrophages were localized in traumatized and control muscles using an a-naphthyl acetate esterase kit (Sigma, St. Louis, MO) according to manufacturers instructions. The slides were observed by light microscopy and photographed. Localization of proteinases, proteins, and proteoglycans were performed by indirect immunohistochemical techniques using fluorescein-labeled second antibody. Antiserum to rat chymase was prepared by conventional methods of immunization in rabbits by Dahlmann et al. (1985). The antiserum was provided as affinity-purified immunoglobulin (IgG) fractions from Affi-gel 15 (Bio-Rad, Rich-
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
71
mond, CA). Antisera to rat fibrinogen, albumin, and complement C3 were obtained commercially (Organon Teknika-Cappel, Malvern, PA). Briefly, the immunohistochemical staining procedure for fibrinogen, albumin complement Cs, and chymase included washing the slides three times for 5 min with filter sterile 0.01 M sodium phosphate in 0.15 M NaCl, pH 7.4 (PBS). Fifty microliters of the appropriate dilution of antiserum was applied to each slide. The slides were incubated at room temperature for 30 min in a moist chamber to prevent evaporation. The antisera were diluted to approximately 1 mg/ml protein as determined by protein assay (Lowry et al., 1951) using albumin as a standard. Following incubation, all slides were washed for 5 min in PBS. Finally, 50 ~1 of fluoresceinlabeled goat F(ab’), anti-rabbit IgG at a 1:20 dilution was added to all experimental and control slides. After another 30 min of incubation and a final wash for 5 min in PBS, glass coverslips were applied over the tissue sections with 50% glycerine 50% PBS containing 1 mg/ml p-phenylenediamine (Sigma, St. Louis, MO) and viewed on a Leitz Orthoplan fluorescence microscope (Stauber et al., 1986). The localization of the plasma proteins was used to establish vascular permeability and cellular membrane integrity. The presence or absence of specific proteoglycans has been useful in identification of muscle injury (where chondroitin 6-sulfate-proteoglycan disappears) and regeneration (where embryonic forms of proteoglycans reappear in the extracellular matrix). Monoclonal antibodies directed against proteoglycans digested with chondroitinase ABC were provided by Dr. Bruce Caterson (University of North Carolina at Chapel Hill, Chapel Hill, NC). Antigen isolation, monoclonal antibody production, and characterization have been completely described previously (Caterson et uI., 1981, 1985). Digestion of the tissue with chondroitinase ABC results in a proteoglycan fragment of small oligosaccharide stubs with at least one characteristic unsaturated disaccharide (0, 4, or 6 sulfated) of the chondroitin sulfate isomer and the linkage region attached to the proteoglycan core protein (Caterson et al., 1987). Localization of specific proteoglycans was performed using a modification of the previously described technique (Bertolotto et al., 1986). The slides were washed for 5 min with 0.05 M Tris/HCl, 0.05 M NaCl, pH 8.0, and predigested with 0.5 units/ml chondroitinase ABC (Sigma, St. Louis, MO) diluted in the same buffer for 30 min at 37°C in a moist chamber. After a 5-min wash in PBS, 50 p.1of diluted specific monoclonal antiserum was added to each slide and incubated as before. Following incubation, all slides were washed again in PBS followed by the application of fluorescein-labeled goat F(ab’), anti-mouse IgG or IgM at a 1:20 dilution. After another 30 min of incubation the slides were washed with PBS and glass coverslips were applied as described above. The slides were viewed on the fluorescence microscope. Method specificity and antibody specificity for the immunohistochemical reactions were determined according to the criteria of reliability established by Petrusz et al. (1980). The controls for the fluorescent immunohistochemical experiments included normal and injured tissue sections either incubated with normal rabbit serum instead of specific antibody or culture media instead of monoclonal antibody followed by fluorescein-labeled antiserum. Controls for the monoclonal antisera to proteoglycans included fresh frozen sections of rat ear as positive controls treated in the same manner as the normal and injured muscle tissues. Each monoclonal antibody to proteoglycans was localized in the cartilag-
72
STAUBER,
FRITZ,
AND
DAHLMANN
inous portion of the rat ear. The specificity of the monoclonal antibodies was tested by excluding the chondroitinase ABC digestion prior to monoclonal antibody application followed by fluorescein-labeled second antibody. Another group received only chondroitinase ABC, diluted mouse serum, and fluoresceinconjugated second antibody. Normal and injured muscle sections incubated with monoclonal antibodies to proteoglycans without prior chondroitinase ABC digestion or vice versa were devoid of any specific fluorescent localization. Fluorescein-conjugated concanavalin A (Con A; Vector Laboratories, Burlingame, CA) was used to localize nonspecific glycoproteins rich in mannose residues such as those found in the pericellular basement membrane of the basal lamina (Gulati et al., 1983). Fluorescein-labeled Con A at a I:200 dilution in PBS was applied directly to slides of normal and injured muscles after an initial wash for 5 min in PBS. Following a second wash in PBS, coverslips were applied and the samples were photographed after being viewed under the microscope. RESULTS Following blunt trauma to the lower leg muscles, the animals did not exhibit any abnormal behavior. Food and water intake and ambulation appeared normal. At the specific time for tissue removal, the hindlimbs of each animal were examined for evidence of damage. The skin of the injured right calf was not broken but had a large ecchymotic area on the lateral surface of the calf. The calf area appeared enlarged with a palpable hardened area at the site of injury. Removal of the skin revealed fluid accumulation between the skin and the fascia, hematoma, and bruising of the gastrocnemius muscle extending to the Achilles tendon. The soleus muscle was also bruised but only at the immediate site of impact. The muscles were still attached to their origins and insertions and there was no evidence of bone damage. The diameter of the ecchymotic area measured on the lateral surface of the gastrocnemius muscle decreased with time. At 24 hr following trauma, a lesion of 15 mm in diameter was recorded. This decreased to 10 mm at 48 hr, and by 120 hr, it measured only 5 mm in diameter. A significant increase in the wet weight of the gastrocnemius muscles was recorded at 24 hr postinjury (P < 0.05; Table I). While the gastrocnemius muscle weight remained elevated above controls beyond 48 hr, these differences were not significant. By 120 hr postinjury, the weight of the gastrocnemius muscle had actually decreased. Microscopic evidence of muscle damage to the gastrocnemius included disorganization of the muscle architecture, widened interstitial spaces, fascicle disruption, and accumulation of mononuclear cells in the interstitial spaces. At 24 hr postinjury specific myoliber damage was noted with flocculent degeneration of the sarcoplasm and/or macrophage infiltration (esterase-positive cells) (Fig. lA, arrow). The interstitial space at this time probably contained fluid as few identifiable mononuclear cells appeared between myofibers. The increased fluid content would account for the increased turgor and wet weight observed in the muscle. Mononuclear cell infiltration increased in the interstitial spaces through 72 hr with further evidence of macrophage involvement (Fig. lB, arrows). Myofiber degeneration was observed in samples removed at 96 hr with fiber remnants and cell fragments throughout each tissue section (Fig. 1C). Fewer esterase-positive cells and necrotic fibers were observed past 96 hr; however, the muscle architecture remained disrupted through 120 hr (Fig. 1C). Evidence of damage to the
ECM
CHANGES
FOLLOWING
BLUNT
73
TRAUMA
TABLE I Wet Weights (in Grams) of Injured and Control Muscles
Time period 24 hr Gastrocnemius Soleus 48 hr Gastrocnemius Soleus 12 hr Gastrocnemius Soleus 96 hr Gastrocnemius Soleus 120 hr Gastrocnemius Soleus
Injured mean + SEM (n =3)
Control mean f SEM (n = 3)
1.43 rt 0.089* 0.22 2 0.046
1.17 2 0.026 0.21 + 0.018
1.29 f 0.176 0.20 f 0.019
1.15 2 0.015 0.19 * 0.011
1.23 f 0.168 0.20 + 0.02
1.13 2 0.066 0.19 ” 0.009
1.24 _’ 0.097 0.18 k 0.02
1.13 + 0.057 0.19 f 0.016
0.89 2 0.083 0.18 2 0.024
1.06 2 0.047 0.18 k 0.02
* Significant at P 0.05.
soleus muscle was limited to a few damaged myotibers per fascicle with cellular infiltration only in these damaged myofibers. By 120 hr there was little evidence of damage observed on sections of soleus muscle. Evidence of vascular disruption was observed using specific antisera for plasma constituents: fibrinogen, albumin, complement C,, and high molecular weight cysteine proteinase. Immediately following injury, all of these plasma proteins were localized throughout the widened interstitial spaces around damaged myofibers. In addition, immunoreactivity was present inside some myofibers of the gastrocnemius muscle, indicative of sarcolemmal rupture (Fig. 2A). Albumin and complement C, had similar distributions in both the injured gastrocnemius and soleus muscles (Fig. 2B). Chymase localization revealed degranulation of mast cells from 12 through 120 hr in gastrocnemius muscle samples. In muscle samples from uninjured muscles, mast cells (chymase positive) were evidence in both the endomysium and the perimysium around blood vessels (Fig. 2C) but without degranulation. Thus, only a localized mast cell response to damage was noted following trauma which was especially evident around vascular areas (Fig. 2D) with chymase remaining associated with the extruded granule matrix. Extracellular matrix disruption was observed in the gastrocnemius muscle immediately after injury followed by an increased widening of the interstitial space caused by the deposition of extracellular matrix components. In the soleus muscle, proliferation of extracellular matrix also was evident but on a smaller scale. Details of the reorganization of the extracellular matrix are provided below. Both the soleus and gastrocnemius muscles showed alterations in Con A staining which helped distinguish undamaged, absent, and regenerating fibers. In control muscles, Con A outlined the perimysium, endomysium and epimysium around myofibers, blood vessels, nerves, and capillaries (Fig. 3A). Following injury, the endomysium and perimysium appeared disrupted around myofibers. In
74
STAUBER,
FRITZ,
AND
DAHLMANN
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
75
FIG. 3. Histochemical localization of Con A in control and injured rat muscles. Cross sections of gastrocnemius muscle. (A) Control. (B) 24 hr postinjury; note the absence of stained material around the periphery of the myofibers (arrow). (C) 48 hr postinjury. (D) 72 hr postinjury, Con A-positive material fills areas around infiltrating cells (arrow). (E) % hr postinjury, small fibers appear among Con A-positive material (arrow). Cross sections of soleus muscle. (F) 24 hr postinjury. (G) 72 hr postinjury. (H) % hr postinjury. X240.
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
77
FIG. 4. Immunohistochemical localization of chondroitin f5-sulfate proteoglycan in control and injured rat muscles. Cross sections of gastrocnemius muscle. (A) Control. (B) 24 hr postinjury; note the proteoglycan positive cells (white arrow) and proteoglycan stained remnants from former myofibers (black arrow). (C) 72 hr postinjury. (D) % hr postinjury; note the smaller outlined myofibers in close proximity to mature fibers (arrow). (E) 120 hr postinjury. Cross sections of soleus muscle. (F) % hr postinjury. (G) 120 hr postinjury. (H) Gastrocnemius muscle, chondroitin &sulfate proteoglycan control without chondroitinase ABC predigestion. x 240.
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
79
80
STAUBER,
FRITZ,
AND
DAHLMANN
some cases, myofibers appeared separated from the matrix as Con A-positive material was absent around the periphery of these fibers (Fig. 3B, arrow). The muscle architecture showed continued disorganization through 48 hr (Fig. 3C), and by 72 hr, Con A-stained material tilled the areas of presumably absent myofibers (Fig. 3D, arrow). The apparent density of the matrix material continued to increase through 96 hr, filling in all of the wide extracellular space between existing myotibers. Small regenerating myofibers appeared in the midst of Con A-positive material at this time (Fig. 3E, arrow). In the injured soleus muscles, Con A-stained material appeared in the widened interstitial space around damaged myofibers while a more regular staining pattern was observed around the undamaged fibers of the section (Fig. 3F). Some damaged myotibers in the injured soleus muscles appeared with Con A-positive staining present inside the fibers. By 72 hr, these apparently damaged myofibers were more heavily stained with Con A-positive material (Fig. 3G). Through 96 hr, groups of small myofibers were seen surrounded by a thick Con A-stained matrix (Fig. 3H). Thus, some type of limited myofiber repair was consistently observed in all of the injured soleus muscles examined. Chondroitin 6-sulfate proteoglycan was localized in the endomysial area around myofibers, blood vessels, and nerves in the normal gastrocnemius muscle (Fig. 4A). This sulfated proteoglycan filled the endomysial space between normal myofibers. Following injury, the localization of chondroitin 6-sulfate proteoglycan revealed (1) a disrupted arrangement of proteoglycan material within the widened interstitial space, (2) the absence of a continuous proteoglycan matrix around disrupted myofibers, and (3) the presence of proteoglycan-positive cells in areas of absent myotibers (Fig. 4B, white arrow). Additionally, some proteoglycan matrix material remained, outlining the position of former myofibers (Fig. 4B, black arrow). At 72 hr, many proteoglycan-positive cells were evident near existing myofibers (Fig. 4C). However, there was no proteoglycan staining associated with the cell membrane of the existing myofibers. Groups of small myofibers outlined with chondroitin 6-sulfate proteoglycan matrix were seen in the damaged areas around existing myofibers at 96 hr (Fig. 4D). Interestingly, these small fibers (Fig. 4D, arrow) were always in close proximity to larger, mature fibers. By 120 hr after muscle trauma, a more organized myofiber architecture was observed. However, thickened areas of proteoglycan material were still evident between fibers (Fig. 4E). Chondroitin 6-sulfate proteoglycan localization was absent around the few damaged myofibers in the soleus muscle while it remained apparent in the endomysial region of undamaged myofibers (Fig. 4F). Like the gastrocnemius muscle, thickened regions of proteoglycan material were evident at 120 hr with proteoglycanpositive cells present in the interstitial spaces (Fig. 4G). No extracellular matrix staining was observed in muscle sections incubated with this monoclonal antibody whenever predigestion with chondroitinase ABC was excluded (Fig. 4H). Localization of unsulfated chondroitin and chondroitin 4-sulfated proteoglycan revealed less change. In control muscle, unsulfated chondroitin proteoglycan was localized around myofibers, capillaries, and blood vessels in the endomysial and perimysial regions (Fig. 5A). Following injury, no unsulfated chondroitin proteoglycan was observed until 72 hr when patchy fluorescent localization could be seen around myolibers in injured areas (Fig. 5B). No fluorescent localization was observed in the interstitial spaces among mononuclear cells, small regenerating
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
81
a2
STAUBER,
FRITZ,
AND
DAHLMANN
fibers, or the matrix material that filled this area. In contrast, chondroitin 4-sulfate could not be localized around or between normal or injured gastrocnemius myofibers at 24 hr (Fig. 5C). However, small fiber-like structures located in the interstitial space were stained around their periphery with positive material at 96 hr (Fig. 5D). No other chondroitin 4-sulfate proteoglycan-positive staining was observed in the injured areas of the tissue after this time. Fibronectin localization changed as early as 24 hr postinjury. In normal gastrocnemius muscle, fibronectin predominated in capillaries with some patchy localization in a thin band around myofibers (Fig. 6A). During the first 48 hr postinjury, tibronectin could be observed filling the widened endomysium around damaged fibers in addition to its presence inside damaged myofibers (Fig. 6B). Fibronectin localization was most obvious around small fibers through 96 hr in addition to its presence among the cellular infiltrate (Fig. 6C). Fibronectin filled the interstitial space by 120 hr in a pattern similar to that observed with Con A (Fig. 6D). Dense staining was again noticed around small fiber-like structures in close proximity to surviving, mature myofibers. DISCUSSION Skeletal muscle injury is generally followed by a series of processes that results in some form of repair to the damaged tissue. These processes include inllammation, satellite cell activation, myogenesis, ftbroblast proliferation, and connective tissue reorganization. The magnitude of the injury and the degree of disruption of the vascular supply may determine the time course, overall involvement of the various cell populations, and the final outcome of the repair of a functional muscle. The repair can vary from complete restoration of an essentially normal muscle (Fritz and Stauber, 1988; Stauber ef al., 1988) to a muscle with extensive fibrosis (Lehto et al., 1986) and restricted motion (e.g., contracture). To study the histopathology of injured skeletal muscles, a contusion injury was produced by impact trauma to the lower leg of a rat. Impact trauma to the leg resulted in extensive damage to the gastrocnemius and to a lesser degree the soleus muscles. The damage was complex in nature including crushed myofibers, disrupted fascia, and ruptured blood vessels with evidence of bleeding and hematoma formation. It was also possible, although not readily evident, that some local nerve damage resulted. Using histochemical techniques, indicators of inflammation, cell infiltration, and connective tissue reorganization were monitored for 5 days following trauma in order to document the early histopathology of muscle following a contusion injury. The extended durations of both the inflammation and the repair process were uniquely different than muscle injury from a mechanical overload (Stauber et al., 1988) which also has a complex, multifaceted repair mechanism involving muscle (Stauber et al., 1988), connective tissue (Fritz and Stauber, 1988), and even cells of the reticuloendothelial system (Stauber et al., 1988). An acute inflammatory response with swelling resulted from damage to the vascular, myofiber, and connective tissues of the gastrocnemius muscle. The increased wet weight recorded for the gastrocnemius muscle and the large ecchymotic area observed externally on the lateral surface of the muscle indicated that hemorrhage and fluid accumulation were present in the damaged muscle. Within 12 hr, the plasma components, fibrinogen, albumin, and complement C,, were localized throughout the interstitial space and even inside some of the myofibers,
ECM
CHANGES
FOLLOWING
BLUNT
TRAUMA
83
84
STAUBER,
FRITZ,
AND
DAHLMANN
indicative of extensive vascular leakage and sarcolemmal damage. The presence of these and other humoral factors in the interstitial space would act as chemotactic agents attracting mononuclear cells responsible for removing damaged tissue-another feature of inflammation following injury. The extensive mononuclear cell infiltration observed included numerous macrophages in and around damaged myofibers. Together with the observed degranulation of mast cells, the release of a variety of known mediators of inflammation such as histamine, lysosomal enzymes, and interleukin-1 (Larsen and Henson, 1983) would have occurred. The duration of the inllammatory response was longer than that seen following myotiber rupture without bleeding (Stauber et al., 1988) as the serum components remained elevated in the gastrocnemius muscle over the entire 5-day period. The extended duration of the inllammatory response in concert with damage to the basal lamina probably resulted from vascular disruption and ischemia which together would impair the repair process. Evidence of a delay in the repair of the damaged muscle was also found in the sequence of extracellular matrix changes observed. In mature muscle, proteoglycans reside as a thin band around myofibers; however, following injury, proliferation of matrix material occurs similar to that seen during embryogenesis. The early disorganized appearance of the matrix with a loss of chondroitin 6-sulfate proteoglycan (an indication of degeneration) was probably the result of crushing and fluid accumulation leading eventually to digestion by hydrolytic enzymes. Chondroitin 6-sulfate proteoglycan was not observed again until 4 days later when it was seen surrounding small fibers indicating that the repair process was present. Prior to the reappearance of chondroitin 6-sulfate proteoglycan, unsulfated chondroitin proteoglycan was observed in the granulated area. Since unsulfated chondroitin proteoglycan is synthesized by perfusion-myoblasts and chondroitin 6-sulfate by postfusion cells (Hutchison and Yasin, 1986), the proteoglycan staining pattern was similar to that seen during muscle development and repair following exercise-induced overload (Fritz and Stauber, 1988). Thus, a similar repair process followed muscle contusion but required a longer time period. Fibronectin requires special consideration because two types of fibronectin are present and cannot be differentiated by present immunohistochemical techniques (Gulati et al., 1982; Kleinman et al., 1981). Fibronectin is a normal constituent of plasma and follows the expected route of albumin and other blood constituents when vascular damage is sustained. A separate tissue fibronectin is also present and resides as a thin band along with the proteoglycans around each myofiberpart of the extracellular matrix or attached to it. Both fibronectin and chondroitin 6-sulfate proteoglycan appeared more prominently around small presumptive regenerating fibers at approximately 4 days postinjury. Chondroitin 4-sulfate proteoglycan could only be localized at 96 hr. Taken together, this evidence would indicate that myoblast fusion had occurred around the third or fourth day since the presence of these matrix components correlates with observations of myotube fusion by others (Gulati et al., 1982; Kleinman et al., 1981). Muscle tissue repair seems to follow a prescribed pattern of processes that involve muscle and nonmuscle components. Overall the repair process is independent of the type of injury with regard to the nature of myogenesis and con-
ECM CHANGES FOLLOWING
BLUNT
TRAUMA
85
nective tissue proliferation but differs in time course if vascular flow is arrested. Thus, free muscle grafts, muscle damage following ischemia such as during tourniquet application, and contusion injury can, in many ways, be considered equivalent. Once adequate blood flow has been provided, the regeneration of muscle occurs. If the restoration of blood flow is slow, the connective tissue proliferation seemsto predominate and fibrosis can result. This conclusion supports a rationale that rehabilitation or pharmacologic intervention following injuries that result in ischemic damage to skeletal muscle should concentrate on techniques to restore blood flow and capillary regrowth while inhibiting connective tissue proliferation. ACKNOWLEDGMENTS This work was supported in part by funds from Comptex, Inc., to W.T.S. The authors thank Bruce Caterson, Ph.D., University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, for a liberal supply of monoclonal antibodies to unsulfated chondroitin, chondroitin Csulfated, and chondroitin 6-sulfated proteoglycans. The authors also thank Steve Stratton, Ph.D., San Antonio, Texas, for the loan and use of the traumatizing unit.
REFERENCES A., PALMUCCI, L., GAGLIANO, A., MONGINI, T., and TARONE, G. (1986). Immunohistochemical localization of chondroitin sulfate in normal and pathological human muscle. J. Neural. Sci. 73,233-244. CATERSON, B., BAKER, J. R., CHIUSTNER, J. E., KEARNEY, J. F., and STROHRER, R. L. (1981). The characterization of clonal antibodies directed against bovine nasal cartilage proteoglycan and link protein. In “Monoclonal Antibodies and T-cell Hybridomas” (G. J. Hammerling, B. Hammerling, and J. F. Keamey, Eds.), pp. 259-268. Elsevier/North-Holland, New York. CATERSON, B., CALABRO, T., and HAMPTON, A. (1987). Monoclonal antibodies as probes for elucidating proteoglycan structure and function. In “Biology of Proteoglycans” (T. N. Wight and R. P. Mecham, Eds.), pp. l-26. Academic, New York. CATERSON, B., CHRISTNER, J. E., BAKER, J. R., and COUCHMAN, J. R. (1985). Production and characterization of monoclonal antibodies directed against connective tissue proteoglycans. Fed. Proc. 44, 386-393. DAHLMANN, B., KUEHN, L., RUTSCHMANN, M., and REINAUER, H. (1985). Purification and characterization of a multicatalytic high-molecular-mass proteinase from rat skeletal muscle. Biochem. J. 228, 161-170. FRITZ, V. K., and STAUBER W. T. (1988). Characterization of muscles injured by forced lengthening. II. Proteoglycans. Med. Sci. Sports Exercise 20, 354-361. GULATI, A. K., REDDI, A. H., and ZALEWSKI, A. A. (1982). Distribution of tibronectin in normal and regenerating skeletal muscle. Anat. Rec. 204, 175-183. GULATI, A. K., REDDI, A. H., and ZALEWSKI, A. A. (1983). Changes in the basement membrane zone components during skeletal muscle fiber degeneration and regeneration. J. Cell Biol. 97, 957-%2. HUTCHISON, C. J., and YASIN, R. (1986). Developmental changes in sulphation of chondroitin sulphate proteoglycan during myogenesis of human muscle cultures. Dev. Biol. 115, 7883. KLEINMAN, H. K., KLEBE, R. J., and MARTIN, G. R. (1981). Role of collagenous matrices in the adhesion and growth of cells. J. Cell Biol. 88, 473-485. LARSEN, G. L., and HENSON, P. M. (1983). Mediators of inflammation. Annu. Rev. Immunol. 1, 335-359. LEHTO, M., JARVINEN, M., and NELIMARKKA, 0. (1986). Scar formation after skeletal muscle injury: A histological and autoradiographical study in rats. Arch. Orthop. Trauma Surg. 104, 366-370. LOWRY, 0. H., ROSENBROUGH, J. N., FARR, A. L., and RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275. PETRUSZ, P., ORDRONNEAU, P., and FINELY, J. C. W. (1980). Criteria of reliability for light microscopic immunocytochemical staining. Histochem. J. 12, 333-348. STAUBER, W., FRITZ, V., DAHLMANN, B., GAUTHIER, F. KIRSCHKE, H., and ULRICH, R. (1985). BERTOLO~O,
86
STAUBER,
FRITZ,
AND
DAHLMANN
Fluorescence methods for localizing proteinases and proteinase inhibitors in skeletal muscle. Histochem. STAUBER,
J. 17, 787-796.
W. T., FRITZ, V. K., DAHLMANN, B., and REINAUER, H. (1986). Immunofluorescent localization of an alkaline proteinase in skeletal muscles from diabetic rats. Basic Appl. Histochem. 30, 147-152. STAUBER, W. T., FRITZ, V. K., VOGELBACH, D. W., and DAHLMANN, B. (1988). Characterization of muscles injured by forced lengthening. I. Cellular Infiltrates. Med. Sci. Sports Exercise 20, 345-353. STRATTON, S. A., HECKMANN, R., and FRANCIS, R. (1984). Therapeutic ultrasound: Its effects on the integrity of a nonpenetrating wound. J. Orthop. Sports Phys. Ther. 5, 271281.
