Matrix Vol. 911989, pp. 200-205

Collagenases from Periarticular Ligaments and Tendon: Enzyme Levels during the Development of Joint Contracture

JUDITH HARPER1, DAVID AMIEL2, and ELVIN HARPER1

1 Departments of Chemistry (D-006) and 2 Surgery (M-030) University of California at San Diego, La Jolla, CA 92093, USA.

Abstract

Our laboratories have previously demonstrated that normal rabbit periarticular ligaments, anterior cruciate ligaments (ACL), medial collateral ligaments (MCL) and patellar tendon (PT) secrete collagenase. In this current study we examined these connective tissues following an immobilization period of 4 weeks. In the ligaments producing collagenase, activity was ex­pressed only in the control, not in the immobilized joint. Control and experimental patellar tendon samples produce collagenolytic activity, suggesting that the expression of enyzme is less affected in tendons as compared to ligaments. Characterization of these collagenases was carried out using an antiserum directed against rabbit synovial collagenase. We demonstrated that ligament (ACL) and tendon (PT) collagenases cross react with this antibody in a double immuno­diffusion assay. Protein blots of PT, ACL and MCL collagenases identified one major species (M, = 45,000) and a minor species (M, = 50,000) of immunoreactive proteins in all three connective tissues. Differences between control and experimental enzyme levels appear to be due to less col­lagenase protein being produced by immobilized ligaments.

Key words: collagenase, joint contracture, stress deprivation.

Introduction

There are many unresolved questions regarding the rate of recovery of mechanical and structural properties of the periarticular connective tissue, i. e. ligaments and tendons following joint immobilization. Stress deprivation and exercise play important roles in the healing process of these tissues (Vailas et al., 1981; Gelberman et al., 1982; Woo et al., 1987). During the time of contracture development, stress deprivation contributes to a weakening of the immobilized periarticular connective tissue (Amiel et al., 1982; Woo et al., 1987). Immobility causes significant increases in collagen metabolism, but the newly deposited collagen fibrils are laid down in a random orientation, resulting in less tensile strength (Akeson et al., 1980). The major protein in periarticular ligaments and tendon is colla­gen. Since collagenase is a metalloproteinase that plays a key role in collagen degradation, we have examined these

© 1989 by Gustav Fischer Verlag, Stuttgart

tissues for this enzyme. In our previous studies (Harper et al., 1988) we determined the baseline levels of collagenase for anterior cruciate ligaments - ACL, medial collateral ligaments - MCL, and patellar tendons - PT from normal rabbits. Understanding the biochemical events that occur in these periarticular connective tissues will allow clinicians and physical therapists to design a rational therapeutical program utilizing the modality of immobilization.

This present investigation describes the changes in colla­genase activity following 4 weeks of immobilization and its effects on the periarticular connective tissue. At this point in time, early alterations of different biochemical components of these connective tissues have been observed. These include a decrease in hydration, decrease in proteoglycan concentration, and increase in the reducible crosslinks (Akeson et al., 1974; Amiel et al., 1979), as well as an early development in joint stiffness.

Material and Methods

Immunochemicals

Agarose for immunodiffusion assays was purchased from FMC Corporation, Rockland, ME. Supplies for the protein blots were obtained from Bio-Rad Laboratories, Richmond, CA. These included the trans-blot apparatus, immuno-blot system kit, prestained SDS-PAGE standards and biotinylated standards kit. The antiserum to rabbit synovial collagenase was kindly provided by Dr. Constance Brinckerhoff, Dartmouth College, NH. All other biochemi­cals were reagent grade.

Animals

Four Adult New Zealand white male rabbits weighing 2.5 to 3.0 kg with closed epiphyses were used as an experi­mental animal model. The left knee was fixed in acute flexion by means of a 2.4 mm diameter threaded stainless steel pin for a period of 4 weeks (Amiel et aI., 1982; Woo et aI., 1987; Akeson et aI., 1980; Akeson et aI., 1974). The right knee was left free to serve as a control. After a 4-week period of immobilization the animals were sacrificed and the medial collateral ligaments (MCLs), anterior cruciate ligaments (ACLs) and patellar tendons (PTs) were isolated for biochemical studies. Nonligamentous tissue was removed by careful scraping and dissection.

Ligament and Tendon Cultures

Periarticular connective tissues (ACL, MCL and PT) were surgically removed from the knee joints of 4 New Zealand white rabbits. Each connective tissue was cultured separately in Dulbecco's modified Eagle's medium contain­ing 100 [lg/ml streptomycin, 100 units/ml penicillin and 5 [lglml fungizone. ACL and MCL were cut in half with a scalpel (2 x 1 cm); PT was sliced into fragments (4 x 1 cm). ACL and MCL (wet weights approximately 80 mg and 100 mg, respectively) were cultured in 35-mm tissue culture dishes; PT (wet weight approximately 400 mg) in 60-mm dishes. One ml of medium was added to the 35 -mm dishes, two ml to the 60-mm dishes, and they were incubated at 3rC in a humidified atmosphere of 95% air, 5% CO2 •

Cultures of ligaments and tendons were harvested at the same time every 2 days for a period of 14 days. Media were assayed for collagenolytic activity or collagenase inhibition using the [14C]-glycine peptide release assay.

Collagenase Assay

Collagenase activity was measured in the [14CJ-glycine peptide release assay (Nagai et aI., 1966) and modified as described by Harper et al. (1988). This assay employs neu­tral salt soluble guinea pig skin collagen purified after in vivo labelling with e4C-(U)] glycine (Miller and Rhodes,

Collagenase Levels during Joint Contracture 201

1982). Briefly, 50 [ll (1200-1600 cpm) of a solution (4 mgl ml) of collagen in 0.4 M NaCI were added to 400 f.ll capac­ity tubes and fibrils allowed to form by incubation for 18 to 24 h in a 37°C water bath. Medium to be assayed for enzyme activity was not concentrated nor activated prior to addition to the collagen substrate. An aliquot of medium (10-200 [ll) was incubated with substrate in the presence of 50 mM Tris-HCl, pH 7.4, 5 mM CaCI2, 500 mM NaCl, 0.01 % Brij-35. The assay was allowed to proceed in a 3rc water bath for 16 - 20 h. The reaction was terminated by centrifugation in a Beckman 152 microfuge for 5 min at room temperature. Aliquots of the supernatant were counted in Biofluor (New England Nuclear) on a Beckman LS-3133T scintillation counter. A trypsin control was included to correct for radioactivity released by any dena­tured or noncollagenous labelled protein cleavage, and col­lagenase activity expressed as cpm above the trypsin blank. One unit of collagenase degrades 1 [lg of collagen per min at 37°C.

Double Immunodiffusion

This procedure was based on the method of Ouchterlony (1958). Samples of rabbit skin, ligament and tendon colla­genases were incubated in the outer wells of plates prepared with 1.5% Sea Kern LE agarose in 0.15 M NaCl, 0.02% azide. Anti-collagenase serum or normal sheep serum were placed in the center well. The plates were allowed to develop for 18h at 40°C, washed in 0.15M NaCl, then stained in 0.05% Coomassie blue. Gels were de stained in 50% methyl alcohol, 10% acetic acid.

Protein Blotting

The method of Towbin et al. (1979) was applied to periarticular collagenases. Enzymes were electrophoresced on a 1.0-mm thick SDS-polyacrylamide gel slab (125 mm x 160 mm) according to the procedure of Laemmli (1967). The gel was run at 35 rnA for 2 h. It was then equilibrated in transfer buffer (8.4 mM Tris, 192 mM glycine, 20% methyl alcohol) for 15 - 30 min and blotted onto a nitrocellulose membrane (0.45 micron, Bio-Rad Laboratories) with the aid of a Trans-Blot apparatus. Room temperature transfer was carried out for 18 hat 70 rnA. Membranes or strips of membrane were incubated overnight with normal sheep serum (1 : 100) or sheep antiserum to rabbit synovial colla­genase (1: 100), diluted in 20 mM Tris-HCl, 500 mM NaCl, pH 7.5, 0.05% Tween 20, 0.02% azide, and 1 % gelatin. Alkaline phosphatase conjugates (1 : 2000 dilution) were then added, followed by substrates for the enzyme, nitro blue tetrazolium (NBT) and 5-bromo-4-chloro-3-indolyl phosphate (BCIP). These were diluted into 0.1 M sodium carbonate buffer, pH9.8 to give a final concentra­tion of 0.3 mglml NBT and 0.15 mg/ml Bcrp. Development of the membranes was stopped by placing them in distilled

202 J. Harper et al.

water. Pre-stained SDS-PAGE standards and biotinylated standards were included in every experiment to determine the efficiency of protein transfer and to estimate the molecular weights of the immunoreactive species.

Results

After 4 weeks of immobilization collagenase activity was qjJantitated for 14 days in culture. We chose 14 days since our previous studies (Harper et al., 1988) showed that col­lagenase activity appears after a lag period of 6 days and diminishes by 14 days. Trypsin activation was not attemp­ted. The units of collagenase were calculated for days 2, 4, 6,8, 10, 12 and 14, and presented in Tables I, II and III.

Collagenase activity in control and experimental ACL samples was measured and expressed in Table!. The con­trolligaments produced high levels of collagenase, an aver­age of 9.19 units per ligament, in contrast to experimental values of 0.25 unit. When compared to the previously determined baseline levels of collagenase in normal ACL (1.28 units), it appears that immobilization affects the col­lagen metabolism of both the free and the stress-deprived ligament. This observation can also be extended to the MCL (Table 11). Control samples secreted 7.75 units colla­genase per ligament, although only 0.21 unit of enzyme was present in experimental MCL's. Normal MCL's produce an average of 1.92 units collagenase.

Patellar tendons (PT) were also included in this study to determine if tendons and ligaments respond differently dur­ing stress deprivation. We have presented these results in Table III. Control cultures produced more collagenase (9.09 units) than the experimental samples (4.91 units). Some enzyme was present in all experimental limbs, indi­cating that the diminution in collagenase activity experi­enced by tendons is less than that observed for ligaments. This connective tissue maintained normal collagenase levels, since normal PT's produced an average of 8.07 units collagenase in our previous studies (Harper et al., 1988).

We have presented the data in Tables I-III as the average units of collagenase produced per ligament or tendon over a given 2 day period. We observed differences in total colla­genase activity from individual rabbits, however, ACL and MCL control samples consistently showed higher enzyme levels than their immobilized contralateral limbs. A similar trend was observed for PT samples, although not of the same magnitude.

We then proceeded to elucidate the possible reasons for the lower levels of collagenase present in immobilized periarticular connective tissues. Perturbations in the activa­tion of zymogen may occur, resulting in less total active enzyme. With the use of an antibody to rabbit synovial col­lagenase, we were able to determine whether any cross reactive protein was present in rabbit skin, ACL, and PT cultures from normal ligaments and tendons (Figure 1). All

Table I. Collagenase Activity in Anterior Cruciate Ligaments (ACL) Immobilized for 4 Weeks.

Days in culture 2 4 6 8 10 12 14 Total

Control 0 0.92 2.98 2.14 1.45 1.05 0.65 9.19 Experimental 0 0 0.04 0.10 0.05 0.06 0 0.25

Tissues from 4 rabbits were cultured for 14 days and media harvested every 2 days. Collagenase activity was measured with the [14C]-glycine peptide release assay. Trypsin activation was not attempted. Units of collagenase were calculated for days 2, 4, 6, 8, 10, 12 and 14, and an average value for each day given in all Tables. One unit of collagenase degrades 1 f,tg collagen per min at 37°C.

Table II. Collagenase Activity in Medial Collateral Ligaments (MCL) Immobilized for 4 Weeks.

Control Experimental

Days in culture 2 4 6 8 10 12 14 Total

o 0.06 0.75 1.06 1.89 1.50 2.49 7.75 o 0 0 0 0.05 0.08 0.08 0.21

Tissues from 4 rabbits were cultured for 14 days and media harvested every 2 days. Collagenase activity was measured with the [14C]·glycine peptide release assay. One unit of collagenase degrades 1 f,tg collagen per min at 37°C.

Table III. Collagenase Activity in Patellar Tendons (PT) Immobilized for 4 Weeks.

Control Experimental

2 4

o 0 o 0

Days in culture 6 8 10 12 14 Total

o o

0.82 2.89 1.71 3.67 9.09 o 0.68 2.34 1.89 4.91

Tissues from four rabbits were cultured for 14 days and media harvested every 2 days. Collagenase activity was measured with the [14C]-glycine peptide release assay. One unit of collagenase degrades 1 ~g collagen per min at 37°C.

three samples were precipitated by the antiserum, suggest­ing that common antigenic determinants exist. We examined the nature of the molecular species present in ligaments and tendon by protein blotting (Figure 2). The three connective tissues showed a major band with Mr =

45,000 and a minor species with Mr = 50,000. Since ACL cultures exhibited the greatest differences between control and experimental enzymes, we probed them with antiserum to determine whether zymogen was present. Utilizing this technique, we examined control and experimental culture media from ACL's of rabbits immobilized for 4 weeks. As Table! indicates, little active collagenase was observed in experimental ACL's, but significant activity was present in the controls. If zymogen was being synthesized by the immobilized rabbits but it was not activated, we should see

Fig. 1. Double Immunodiffusion of Col­lagenase Antiserum with Rabbit Skin, Lig-ament and Tendon Collagenases. Agarose (1.5%) plates were developed for 18 hat 4°C. Precipitin bands were stained with 0.05% Coomassie blue and destained in 50% methanol, 10% acetic acid. A = sheep antiserum to rabbit fibroblast collagenase; B = sheep antiserum to rabbit fibroblast collagenase (1 : 2 dilution ); C = normal sheep serum; D = normal sheep serum (1 : 2 dilution); 1 = rabbit patellar tendon (PT) collagenase; 2 = rabbit skin collagenase; 3 = rabbit anterior cruciate ligament (ACL) collagenase Note precipitin bands in A and B and lack of precipitin bands in C and D, indicating cross reaction of ligament, tendon and skin collagenase with the antibody.

Fig. 2. Protein Blot of Rabbit ACL, MCL and PT Collagenase. Collagenases from normal ACL, MCL and PT were elec­trophoresced on a 10% SDS polyacrylamide gel slab. The proteins were transferred overnight to a nitrocellulose membrane and reacted with sheep antiserum to rabbit synovial collagenase (1 : 100). The blot was developed using anti-sheep IgG alkaline phosphatase conjugate (1 : 2000) and substrates for alkaline phos­phatase. From left to right: 1) patellar tendon (21lg = 0.014 unit); 2) medial collateral ligament (2.1Ilg = 0.005 unit); 3) anterior cruciate ligament (21lg = 0.017 unit); 4) biotinylated molecular weight standards: phosphorylase b (97.4 kDa ), BSA (66.2 kDa), ovalbumin (43 kDa), carbonic anhydrase (31 kDa ), soybean tryp­sin inhibitor (21.5 kDa).

equal amounts of immunoreactive collagenase protein in control and experimental samples. We examined media from days 6 or 8, since the highest activity was present in control cultures at this point in time. Our results demon­strate that more collagenase antigen is produced by control ligaments, suggesting that less zymogen is synthesized by immobilized rabbits (Figure3). Even if experimental ACL colJagenases were bound to inhibitor, one should be able to

detect this enzyme, since enzyme-inhibitor complexes are

Collagenase Levels during Joint Contracture 203

1 2 3 4

dissociable on SDS-polyacrylamide gels systems (Cawston et al., 1983 and confirmed in our laboratory), leaving the collagenase free to react with the antiserum.

Discussion

During stress deprivation collagen turnover continues, but newly synthesized collagen is not laid down in the

204 ]. Harper et al.

Fig. 3. Protein Blot of Collagenases from Control and Immobilized Anterior Cruciate Ligaments. Equal volumes (15 ~l) of media from 4 control and 4 experimental ACL samples were examined for crossreactivity with antiserum to rabbit collagenase. The units of active collagenase, as measured by the [14C]-glycine peptide release assay, are indicated for each rabbit. Protein blotting was carried out as described in Figure 2. Lane 1 Control rabbit 1, 0.036 unit; Lane 2 Experimental rabbit 1, 0.000 unit; Lane 3 Control rabbit 2,0.024 unit; Lane 4 Experimental rabbit 2,0.002 unit; Lane 5 Control rabbit 3,0.066 unit; Lane 6 Experimental rabbit 3,0.000 unit; Lane 7 Control rabbit 4,0.039 unit; Lane 8 Experimental rabbit 4,0.002 unit.

correct array for proper joint integrity. Restoration of the structural and mechanical properties is impeded by exces­sive periods of immobilization. Since changes at the biochemical level often precede gross anatomical differences, we examined the amounts of the enzyme colla­genase present at an early stage of immobilization, 4 weeks. We found that periarticular ligaments and tendon ex­pressed lower levels of active collagenase in the immobilized state. These results indicate that there is less catabolic activity in these tissues, perhaps reflecting a sub­strate that is more susceptible to degradation. It has been demonstrated that at 4 weeks, the collagen synthesized by immobilized periarticular connective tissues has a higher number of reducible cross links (Amiel et al., 1979) thus indicating a less mature substrate. Collagen cross linked by glutaraldehyde (Harris and Farrell, 1972) or lysyl oxidase (Vater et al., 1979) is less susceptible to collagenase. Less

enzyme may be needed to degrade the immature collagen produced in experimental limbs. Our immunological studies demonstrate that less enzyme protein is detectable in experimental cultures. This suggests that synthesis of col­lagenase may be suppressed during immobilization, or that collagenase may be degraded more rapidly than the enzyme from control ligaments.

Little is known about the influence of stress deprivation on the normal levels of enzyme activities in periarticular connective tissues. An earlier study found that lysosomal enzymes from immobilized MCL's were elevated by 30% when compared to contralateral controls (Gamble et al., 1984). It was concluded that this increment in lysosomal hydrolases is responsible for excess proteoglycan degrada­tion. These enzymes may play an important role in stress deprivation. Collagen is initially cleaved at physiological pH by the enzyme collagenase. Stromelysin has the capacity

to degrade proteoglycans under similar conditions. The hydrolases may subsequently act on the breakdown pro­ducts generated by these neutral metalloproteinases.

Acknowledgements

We wish to thank the Easter Seal Foundation for their support of this project. We also acknowledge the expert typing of this manu­script by Mrs. Fae Hutzel. We express our appreciation to Dr. Constance Brinckerhoff for her gift of antiserum to rabbit col­lagenase.

References

Akeson, W.H., Amiel, D. and Woo, S.L.-Y.: Immobility effects on synovial joints: The pathomechanics of joint contracture. Bio­rheology 17: 95-110, 1980.

Akeson, W.H., Woo, S.L.-Y., Amiel, D. and Matthews, J.: Biomechanical and biochemical changes in the periarticular connective tissue during contracture development in the immobilized rabbit knee. Connect. Tiss. Res.2: 315 - 323, 1974.

Amiel, D., Woo, S.L.-Y. and Akeson, W.H.: Studies of the biomechanical joint stiffness and biochemical changes in con­nective tissue during the formation and development of a joint contracture. International Conference on Medical and Biologi­cal Engineering, Pt. 9,1979, pp. 76. 10.

Arniel, D., Woo, S. L.-Y., Harwood, F. L. and Akeson, W. H.: The effect of immobilization on collagen turnover in connective tissue: A biochemical-biomechanical correlation. Acta Orthop. Scand.53: 325-332, 1982.

Cawston, T.E., Murphy, G., Mercer, E., Galloway, W.A., Hazle­man, B. L. and Reynolds, J.J.: The interaction of purified rabbit bone collagenase with purified rabbit bone metalloproteinase inhibitor. Biochem. J. 211: 313-318,1983.

Gamble, J. G., Edwards, C. C. and Max, S. R.: Enzymatic adapta­tion in ligaments during immobilization. Amer. ]. Sports Med.12: 221-228, 1984.

Gelberman, R.H., Woo, S.L.-Y., Lothringer, K., Akeson, W.H. and Amiel, D.: Effects of early intermittent passive mobilization

Collagenase Levels during Joint Contracture 205

on healing canine flexor tendons. J. Hand. Surg. 7: 170-175, 1982.

Harper,]., Amiel, D. and Harper, E.: Collagenase production by rabbit ligaments and tendon. Conn. Tissue Res. 17: 253-259, 1988.

Harris, E.D., Jr. and Farrell, M.E.: Resistance to collagenase: A characteristic of collagen fibrils cross linked by formaldehyde. Biochem. Biophys. Acta 278: 133-141,1972.

Laemmli, U. K.: Cleavage of structural proteins during the assem­bly of the head of bacteriophage T4. Nature 227: 680-685, 1970.

Miller, E.J. and Rhodes, R. K.: Preparation and characterization of the different types of collagen. Methods in Enzymol.82: 33-64,1982.

Nagai, Y., Lapiere, C. and Gross, J.: Tadpole collagenase prepara­tion and purification. Biochemistry 5: 3123 - 3130, 1966.

Ouchterlony, 0.: Diffusion-in-gel methods for immunological analyis II. Progr. Allergy 6: 30-154, 1962.

Towbin, H., Staehelin, T. and Gordon, J.: Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. USA 76: 4350-4354, 1979.

Vailas, A.C., Tipton, C.M., Matthes, R.D. and Gart, M.: Physical activity and its influence on the repair process of medial colla­teralligaments. Connect. Tiss. Res. 9: 25 - 30, 1981.

Vater, C.A., Harris, E.D., Jr. and Siegel, R.C.: Native cross links in collagen fibrils induce resistance to human synovial colla­genase. Biochem.].181: 639-645, 1979.

Woo, S.L.-Y., Gomez, M.A., Site, T.]., Newton, P.O., Orlando, C.A. and Akeson, W.H.: The biomechanical and morphologi­cal changes in the medial collateral ligament of the rabbit after immobilization and remobilization. J. Bone ft. Surg., Vol. 69A: 1200-1211,1987.

Woo, S.L.-Y., Hardwood, F.L. and Akeson, W.H.: The effect of immobilization on collagen turnover in connective tissue: A biochemical-biomechanical correlation. Acta Orthop. Scand. 53: 325 - 332,1982.

Woo, S.L.-Y., Inoue, M., McGurk-Burleson, E. and Gomez, M.A.: Treatment of the medial collateral ligament injury. Amer.]. Sports Med.15: 22-29, 1987.

Dr. Judith Harper, Department of Chemistry, D-006 University of California at San Diego, La Jolla, CA 92 093, USA.