Journal of Orthopaedic Research 930!%316 Raven Press, Ltd., New York 0 1991 Orthopaedic Research Society
Cartilage Synthesizes the Serine Protease Inhibitor PAI- 1: Support for the Involvement of Serine Proteases in Cartilage Remodeling Benjamin V. Treadwell, Michele Pavia, Christine A. Towle, Vernon J. Cooley, and Henry J. Mankin Orthopaedic Research Laboratories, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, U.S.A.
~~
Summary: The work described here demonstrates the synthesis by human articular cartilage of plasminogen activator inhibitor-1 (PAI-l), a potent inhibitor of the serine protease tissue plasminogen activator (tPA). We also present data demonstrating an increase in PAI-1 messenger ribonucleic acid (mRNA) in chondrocytes" exposed to the cytokine interleukin-1 (IL-1). Interestingly, this elevation of steady-state mRNA levels does not appear to result in an increase in synthesis of PAI-1 protein. Northern blot analysis reveals that of the two mRNA species (3.4 kb, 2.4 kb) previously reported for PAI-1, only the larger species (3.4 kb) appears to be synthesized by chrondrocytes. Our data demonstrate the IL- l-stimulated production by cartilage of tissue plasminogen activator. We also show evidence for the presence of plasminogen in cartilage. A scheme is presented indicating the probable importance of the serine proteases (tPA and plasminogen) and PAI-1 in cartilage degradation. Key Words: Cartilage-Proteases-Inhibitors-Interleukin1-Plasminogen activator.
The major macromolecular components of cartilage matrix are type I1 collagen and a series of highmolecular-weight proteoglycans (28). In the past two decades, however, analytical studies have demonstrated the presence of a number of quantitatively minor components, which include at least five additional collagen types and a number of other proteins? such as fibronectin, laminin, and thrombospondin (12,28,29,31,32,35,46,47). Cartilage is capable of synthesizing and remodeling matrix with its own machinery. Thus the chondrocyte synthesizes not only all the matrix compo-
nents but also the proteases required for degradation and the regulators (activators and inhibitors) of these proteases. In prior studies we have shown that interleukin-1 (IL-1) is a key component of this remodeling network; it has been shown to stimulate the synthesis of both stromelysin and collagenase by chondrocytes (11,16,17,27,4345). IL-1 has recently been demonstrated to be synthesized by the chondrocytes themselves (34). The work presented in this article further extends our knowledge regarding the action of IL-1 on the degradative cascade in cartilage. We demonstrate that the synthesis by chondrocytes of plasminogen activator inhibitor-1 (PAI-I) is, in part, regulated by IL- 1, which maintains the steady-state level of PAII-specific messenger ribonucleic acid (mRNA). We also demonstrate that cartilage contains two pro-
Received April 30, 1990; accepted October 31, 1990. Address correspondence and reprint requests to Dr. Benjamin V. Treadwell, Orthopaedic Research Laboratories, Massachusetts General Hospital, Jackson 1 1 , Boston, MA 02114, U.S.A.
309
B . V . TREADWELL ET AL.
310
tease components of the fibrinolytic cascade, tissue plasminogen activator (tPA) and plasminogen, and that IL-1-stimulated cartilage has marked elevation of tPA activity. We propose a scheme in which we postulate the probable importance of the serine proteases in cartilage degradation and demonstrate the effects of IL-1 on the activity of these enzymes.
periments, were incubated at 37°C in an atmosphere of 95% air, 5% CO,. The medium was harvested after 24 h. This period of time was chosen for convenience, since we found in previous experiments that the effects of IL-1 on the parameters investigated did not significantly vary between 6-24 h.
MATERIALS AND METHODS
Harvested media samples were prepared for SDS polyacrylamide gel electrophoresis as follows: medium (200 p.1) from each well was treated with 2 volumes of acetone at -20°C for 1 h. The precipitated material was collected by centrifugation. The pellets were lyophilized and resolubilized in 20 p1 of sample buffer (2.5% SDS, 1% glycerol, 10 mM TrisHC1, pH 7.0, and 20 pg/ml phenol red). Samples to be analyzed for enzyme activity were incubated at 37°C for 20 min before gel application. All other samples were heated at 90°C for 1 min in the presence of 1% 2-mercaptoethanol.
Protein A Sepharose CL-4B and human plasminogen were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Human plasminogen antiserum was purchased from CalBiochem Corp. (La Jolla, CA, U.S.A.). Human recombinant tPA and monoclonal antibody were from Genentech, Inc. (South San Francisco, CA, U.S.A.). Bovine PAI-1 antibody (raised in rabbits) was a gift from David J. Loskutoff (Scripps Clinical Research Foundation, La Jolla, CA, U.S.A.). Human recombinant IL1 alpha was a gift from Peter Lomedico (HoffmannLaRoche, Nutley, NJ, U.S.A.). Plasmid PAI-B6 was a gift from David Ginsburg (Howard Hughes Medical Institute, University of Michigan, Ann Arbor, MI, U.S.A.). Dulbecco’s modified Eagle medium (DMEM) was purchased from Gibco (Long Island, NY, U.S.A.). All electrophoresis reagents were purchased from Gallard Schlesinger (Carle Place, NY, U.S.A.). Tran ,%-label was purchased from ICN BioMedicals, Inc. (Irvine, CA, U.S.A.). Casein was from Fisher Scientific Co. (Fairlawn, NJ, U.S.A.).
Preparation of Samples for Sodium Dodecyl Sulfate (SDS) Gel Electrophoresis
SDS-Polyacrylamide-CaseinGel Electrophoresis for Detection of Plasminogen Activator (PA) Activity
The media samples, prepared as described above, were applied to an 11% polyacrylamide gel prepared according to the procedure of Laemmli (25), except that the acrylamide solution contained casein (1 mg/ml) and, where indicated, human plasminogen (5 pg/ml). Preparation of the gel containing casein and analysis of enzyme activity is essentially according to the procedure of Heussen and Dowdle (20). Preparation of Cartilage-Conditioned Medium Electrophoresis was run at a constant 200 V at 4°C. Following electrophoresis, the gels were Cartilage cores from the articular surface of norplaced in a solution of 2.5% Triton X-100 and agimal human knee joints were obtained at the time of tated for 1 h at room temperature. The gels were amputation using a cork borer (3 mm in diameter). then incubated 3 hr at 37°C in 20 mM Tris-HC1, 150 Amputations were performed on patients with mM NaCl, 10 mM ethylenediaminetetracetate chondrosarcomas in the hip area. Discs (1.5 mm (EDTA), 0.02% NaN,, pH 7.5. Immediately followthick) were sliced off from the outer layer of cores ing incubation, gels were stained in Coomassie blue of cartilage taken from the superficial surface of the and destained in 7% acetic acid. Evidence for PA femoral condyle using a specially constructed temactivity was the appearance of a clear band against plate, as described by Ollivierre and colleagues a dark blue background on the plasminogen gels and (34). The discs were placed in wells of a 96-well of these bands in the gel without plasthe absence Falcon plate containing 200 pl of DMEM, and, minogen. where indicated, 20 pCi Tran 35S-label(70% L - [ ~ ~ S ] methionine, 15% ~-[~~S]cysteine; specific activity SDS-Polyacrylamide-casein Gel Electrophoresis for 1,194 Cdmmol), 15 units of human recombinant ILDetection of Plasminogen 1, and cycloheximide (CX) (10 pg/ml) were added to The preparation of the sample and the polyacrylwells. The cartilage plugs, in triplicate for each conamide gel containing casein (no plasminogen) as dition, and, with a minimum of three separate ex-
J Orthop Res, Vol. 9, No. 3, 1991
311
FACTORS INVOLVED IN CARTILAGE DEGRADATION
well as the condition of electrophoresis were as described above. After the gels had been washed in 2.5% Triton X-100 for 1 h, they were placed in plastic bags containing 2 ml Tris-buffered saline (TBS) and tPA (5 pg/2 ml), where indicated. The bags containing the gel and solution were hermetically sealed and incubated for 3 h at 37°C with agitation. The tPA-dependent protease activity was detected on the Coomassie blue-stained gels as described in the section above.
in sample buffer containing 1% 2-mercaptoethanol. For detection of enzyme activity, the immunoprecipitates were analyzed on casein or casein plus plasminogen, as described above. For detection of immunoprecipitated radioactive proteins, gels were dried under vacuum following destaining, and the dried gel was placed in contact with x-ray film.
Immunoprecipitation of Specific Proteins Present in Cartilage-Conditioned Medium
Human articular cartilage was resected from the articular surface of a knee joint. The cells were liberated from the matrix by treatment with clostridal collagenase (0.1%) in DMEM for 18 h, as previously described (30). The isolated cells were rested as a suspension culture (1 x lo6 cells/ml) for 6-8 h in DMEM containing 5% fetal calf serum. Interleukin1 was added to one-half of the suspended cells at a concentration of 50 units/ml, and incubations were continued at 37°C in an atmosphere of 5% CO, for 18 h.
Cartilage-conditioned medium was removed from the well after incubation, adjusted to 0.02% NaN,, and incubated with antibody [anti-(PAI-1, plasminogen, or tPA)] or preimmune serum [normal rabbit serum (NRS)] according to the following procedure. Two hundred microliters of cartilage-conditioned medium was incubated with serum (or ascites for tPA antibody) equivalent to 40 pg IgG, either NRS or anti-PAI-1, -tPA, or -plasminogen, in the presence of 1 mM phenyl methyl sulfonyl fluoride (PMSF) and 10 mM EDTA for 1 h at 25°C and overnight at 4°C. To each sample was added 20 p1 of a protein-A Sepharose bead suspension [protein-A Sepharose equilibrated and suspended in an equal volume of buffer I (1% Triton X-100, 0.05% Tween 20,0.1% SDS, 0.02% NaN,, 300 mM NaCl, 10 mM EDTA, 20 mM Tris HCl, pH 7.5, and 10 Ulml aprotinin)]. Samples were vortexed at room temperature every 10 min for 1 h. Goat anti-mouse (40 pg) IgG was used to precipitate the monoclonal antibody to tPA. IgG-bound Sepharose and goat anti-mouse IgGanti-tPA complex was collected by centrifugation (8,OOOg for 1 min) in a microfuge. Supernatant was aspirated, and pellets were washed twice with 200 p1 of buffer I. The goat anti-mouse IgG complex pellet was immediately treated with SDS sample buffer as described below. The IgG-containing Sepharose pellets were resuspended in 20 p1 buffer I. This suspension was removed from tubes and overlayed onto 700 p1 of buffer I containing 10% glycerol. Protein-bound Sepharose beads were finally collected by centrifugation. The bound material was released from the protein A Sepharose beads, and the goat anti-mouse IgG complex was disrupted by either incubating the pellet in SDS sample buffer at 37°C for 20 min (if enzyme activity was to be determined) or heating at 90°C for 1 min
Isolation of Chondrocytes and Incubation with Interleukin-1
Preparation of Chondrocyte RNA for Northern Analysis Cellular RNA was extracted using the RNAzol Method (CinndBiotecx Laboratories International, Inc., Friendswood, TX, U.S.A.), which is a singlestep method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction (6). For Northern analysis, 10 pg of RNA per lane was applied to a formaldehyde (2.2 h4) agarose, (1%) gel. The RNA in the samples was separated by electrophoresis at 60 V for 5 h. The RNA was transferred by capillary action to a nitrocellulose membrane (Schleicher & Schuell Inc., Keene, NH, U.S.A.). After transfer, the RNA was immobilized by baking at 80°C in a vacuum oven for 1 h. PAI-1 mRNA was detected using a 2 kb Eco R1 cDNA fragment of plasmid PAI-B6 (15). The PAI-1 cDNA probe was prepared using a multiprime DNA labeling system with 32P-dCTP (Amersham International, Amersham, U.K.), according to the procedure of Feinberg and Vogelstein (13,14). The membrane was hybridized overnight at 42°C in 5 X SSC 50% formamide, 0.1% SDS, salmon sperm DNA 100 pg/ml, 10% dextran, and 1X Denhardt’s solution, and subsequently washed under stringent conditions using standard techniques (1). The blot was placed in contact with XAR film (Eastman Kodak Co., Rochester, NY, U.S.A.) and left at -40°C for 12 h with
J Orthop Res, Vol. 9, No. 3, 1991
B . V . TREADWELL ET AL.
312
1 2 3
intensification screens and the radioactive bands visualized after development in an X-Omat developer.
50K-
RESULTS IL-1-stimulated Synthesis of Tissue Plasminogen Activator Increased levels of plasminogen activator activity are evident in the medium conditioned by cartilage for 24 h in the presence of IL-1 (Fig. 1). The figure demonstrates plasminogen-dependent protease activity in cartilage-conditioned medium. This activity is markedly stimulated by IL-1, as is demonstrated in lane 2 ( + IL-1) compared with lane 1 ( - IL-1). Lane 2 shows a clear area (suggestive of protease activity) at M, 75-80 x lo3 that is not evident in the absence of plasminogen (lane 3). This in situ conversion of the latent plasminogen to active plasmin within the gel is demonstrative of PA activity. The PA activity was immunoprecipitated by antibody to the tPA species (lane 5). Lane 7 shows that no pro1
2
3
4
5
6
?
68K-
FIG. 2. Detection of plasminogen activator inhibitor-1 synthesis in cartilage-conditioned medium. Cartilage-conditioned medium taken from samples incubated in the presence of [35S]methionine for 24 h (2IL-1). The medium was subsequently incubated with a serum fraction and the precipitates separated on SDS-polyacrylamide gels. Lane 1: Control medium precipitated with normal rabbit serum. Lane 2: Control medium precipitated with rabbit anti-PAl-1. Lane 3: As in lane 2, except IL-1 was present during the 24-h incubation with cartilage.
tease activity was precipitated if normal mouse ascites fluid was substituted for the tPA antibody. Evidence suggesting that the stimulation of tPA activity by IL-1 (lane 2 versus lane 1) is the product of increased synthesis of the enzyme and not simply increased secretion is demonstrated by the results shown in Fig. 1, lane 4. This figure shows that the translational inhibitor cycloheximide prevents the stimulation of tPA activity with IL-1. Synthesis of Plasminogen Activator Inhibitor I
FIG. 1. Effect of IL-1 on synthesis of tPA. Cartilage-conditioned medium was applied to SDS polyacrylamide gels containing either casein and plasminogen or casein alone (lane 3) and separated by electrophoresis. Lane 1: Acetoneprecipitated 24-h cartilage-conditioned medium. Lane 2: As in lane 1, except IL-1 (30 units) was present with cartilage during the 24-h incubation step. Lane 3: Sample as in lane 2 applied to casein gel lacking plasminogen. Lane 4: As in lane 2, except cycloheximide was present during the 24-h incubation of cartilage with IL-1. Lane 5: Samples as in lane 2 immunoprecipitated with anti-human tPA. Lane 6 : Commercially available human recombinant tPA (0.2 ng). Lane 7: As in lane 5, except the tPA antibody was replaced by normal mouse serum.
J Orthop Res. Vol. 9, No. 3, 1991
Plasminogen activator inhibitor I synthesis by cartilage explants is demonstrated in autoradiographs shown in Fig. 2. The results show the synthesis and secretion from cartilage of a radioactive peptide immunoreactive with antibody raised to purified bovine endothelial cell PAI-1. The synthesis of this inhibitor by chondrocytes appears to be unchanged by the presence of IL-1 in the culture medium. This is indicated by the similar intensities of the immunoprecipitated radioactive band of M, 50,000 shown in the autoradiographs (lanes 2, 3) of Fig. 2. 11-1-treated Chondrocytes Have Increased Levels of PAI-1 mRNA In contrast to the lack of effect IL-1 has on the synthesis of PAI-1 protein, chondrocytes do respond to this cytokine at the mRNA level. Figure 3
FACTORS INVOLVED IN CARTILAGE DEGRADATION
1
2
3
3.4 2.4
-
FIG. 3. Northern analysis of cartilage RNA. RNA (10 pg) isolated from either human endothelial or cartilage cells was separated on agarose gels and subsequently tested for its ability to hybridize with a 32P-PAI-lcDNA probe. Lane 1: Endothelial cell RNA. Lane 2: Cartilage cell RNA. Lane 3: RNA isolated from IL-1-activated cartilage cells.
demonstrates a substantial (four- to fivefold) increase in the levels of mRNA coding for this inhibitor. This level of PAI-1 message, however, is considerably below that obtained with comparable amounts of endothelial cell RNA (lane 1). Furthermore, the endothelial cell PAI-1 mRNA is of two sizes-3.4 and 2.4 kb-whereas the chondrocyte contains only the 3.4 kb species.
313
(2,11,16,26,45) and affects the level of mRNA coding for at least one protease inhibitor, PAI-1. As suggested by the scheme shown in Fig. 5, the pivotal enzyme in this pathway is tPA. This enzyme is virtually undetectable in medium conditioned by unstimulated cartilage (see Fig. 1). PAI-1, a potent inhibitor of tPA (18,19,36), is present in cartilage and actively synthesized by it, thus allowing for the possibility that PAI-1 inactivates any tPA present. Although not described in this article, we have included in Fig. 5 the recent report showing that IL-1 stimulates the synthesis of phospholipase A,. This is a key enzyme involved in the synthesis of the precursor to the eicosanoids, compounds clearly implicated in connective tissue catabolism (4). There are several immunologically distinct types of plasminogen activator inhibitors reported in the literature (22). The inhibitor reported here (PAI-1) to be synthesized by cartilage is of the endothelial cell type (33). A second PA inhibitor (PAI-2) has been shown to be synthesized by human placenta
Plasminogen in Cartilage
The presence of plasminogen activity in cartilageconditioned medium is shown in Fig. 4. Cartilageconditioned medium was applied to a casein-containing SDS polyacrylamide gel and the components separated by electrophoresis. After electrophoresis the gel was incubated in buffer either with or without purified recombinant tPA. The gel incubated in the presence of tPA (lane 2) shows a clear band at M, 85,000. Interleukin-1 does not appear to stimulate the synthesis of this enzyme, but rather a slight inhibition appears to be evident (lane 3). We believe that this unexpected result may be due to an increase in the activation and subsequent degradation of plasminogen in the IL-1-treated cartilage. DISCUSSION
The results reported in this study demonstrate that the serine proteases tPA and plasminogen are present in cartilage and that there is a significant role for these proteases and their inhibitors in the control of cartilage degradation. Interleukin- 1 appears to play a key role in this pathway since it stimulates the synthesis of several proteases
FIG. 4. Presence of plasminogen in cartilage-conditioned medium. Acetone-precipitated samples were applied to an SDS polyacrylamide gel containing casein, and after electrophoresis the gel was incubated in buffer containing tPA. Lane 1: Cartilage-conditioned medium after electrophoresis on casein gel. tPA was omitted from the gel buffer incubation step. Lane 2: As in lane 1, but tPA was present during the gel incubation step. Lane 3: As in lane 2, but IL-1 was present during the 24 h incubation with cartilage. Lane 4: Commercially available plasminogen (0.2 pg) electrophoretically separated. The gel was incubated in the presence of tPA. Lane 5: As in lane 4, but tPA was omitted from the gel incubation step.
J Orthop Res, Vol. 9, No. 3, 1991
B . V . TREADWELL ET AL.
314
2 HOURS
INFLA MMA T I 0N
24-48 HOURS
Enzyme synthesis Inhibitor synthesis
CAR TILA GE DESTRUCTION
Enzyme activation
JOINT CAPSULE COLLAGENASE (A)
SYNOVIOCYTES FIBROBLASTS COLLAGENASE
NEUTROPHILS-IL1 MACROPHAGES / CHONDROCYTES
/ PLASMINOG
FIG. 5. Cartilage remodeling. A scheme produced from accumulated data to help explain cartilage remodeling and the role played by those factors believed to be involved (34). The synthesis by chondrocytes of phospholipase A, has been reported to be stimulated by IL-1 (4). The 92,000 gelatinase has been reported to be synthesized in response to IL-1 in bovine growth plate chondrocytes (17). (L) and (A) refer to the latent and active forms of the enzymes, respectively.
(probably macrophage derived) (8). Both these inhibitors have more recently found to be synthesized by a variety of cell types (30,38). A third inhibitor capable of inhibiting plasminogen activator is protease nexin. It was first isolated from fibroblastconditioned medium and has a more broad spectrum of inhibitory activity against serine proteases. Protease nexin-1 cannot be considered a true PA inhibitor, since it is more effective an inhibitor against thrombin and plasmin (21,23,4042). Of interest is the result we report of the increased levels of PAI-1 mRNA in cells treated with IL-1. The elevated levels of PAI-1 mRNA, however, do not appear to affect the level of expressed product. At present we do not know how to interpret this result except to say that it may be related to the species of PAL1 message (3.4 kb) affected by IL-1. This message species has been reported to be less stable than the other PAI-1 message of 2.4 kb (15). Another, and perhaps more likely, explanation attributes the lack of increased expressed product to the relatively (as compared to endothelial cell PAI-1 message) minor increase in mRNA levels with the cytokine. Speculation on the probable involvement of plas-
J Orthop Res, Vol. 9, No. 3, 1991
min in the degradative pathway of connective tissue, including cartilage, dates back to the 1930s (7,9,23,24). Recent information reported here and elsewhere allows us to present a more convincing scheme supporting a role for this enzyme in matrix degradation (2,5,27). Plasminogen, although perhaps not actively synthesized by the chondrocyte, is present in the matrix in significant quantities. The matrix appears to function as a sink for this protein, and it is not easily removed from the matrix. A possible explanation for this is the recent report of the synthesis by cartilage of a new matrix protein, thrombospondin (32), which contains as part of its structure a binding site for plasminogen. Thrombospondin has also been shown to bind tPA, heparin, fibronectin, laminin, and collagen (37,39). These properties suggest that the protein thrombospondin acts as a nucleus for the initiation site of cartilage degradation by plasmin. Further support for this hypothesis is the localized nature of matrix proteolysis (i.e. , pericellular), as has been reported by Dingle and Dingle (10). The results reported here and elsewhere showing that cartilage synthesizes tPA rather than urinary plasminogen activator (uPA) are somewhat surpris-
FACTORS INVOLVED IN CARTILAGE DEGRADATION
ing (2). A generally held belief is that uPA is involved in tissue destruction and tPA in thrombolysis (for review see 3 3 ) . This conclusion comes largely from studies on invasive connective tissue tumors, which have been demonstrated to synthesize large amounts of uPA. It appears that a major trigger for cartilage degradation is under the control of the IL-1 receptor. The cytokine induces the synthesis of tPA and metalloproteases in chondrocytes. Tissue plasminogen activator catalyzes the formation of the nonspecific and very active protease plasmin from the ubiquitous protein (160 pg/ml in serum) plasminogen. Plasmin is the probable activator of two additional matrix-degrading proteases, collagenase and stromelysin (5). The synthesis of these two metalloproteases is also markedly increased in IL- 1-activated chondrocytes (11,44). These results and the recent reports demonstrating the synthesis of al,antitrypsin by cartilage strongly suggest an important role for serine proteases and their inhibitors in cartilage remodeling and degradation. The primary role for collagenase, once activated by stromelysin (5) and by plasmin via tPA, is to digest the framework, whereas the noncollagenous matrix component is probably largely digested by the very active enzyme plasmin and the metalloprotease stromelysin. Acknowledgment: W e thank D. J. Loskutoff of t h e Scripps Institute, La Jolla, California, for the antibody to PAI-1, Peter Lomedico of Hoffmann-LaRoche for recombinant human IL-la, and Thomas Quertermous of Massachusetts General Hospital for his help in Northern blot analysis of PAI-1 mRNA.
REFERENCES 1. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, eds: Current Protocols in Molecular Biology. New York, Wiley, 1987 2. Bunning RAD, Crawford A, Richardson HJ, Opdenakker G, Van Damme J, Russell RGG: Interleukin 1 preferentially stimulates the production of tissue-type plasminogen activator by human articular chondrocytes. Biochim Biophys Acta 924:473482, 1987 3. Canipari R, O’Connell ML, Meyer G, Strickland S: Mouse ovarian granulosa cells produce urokinase-type plasminogen activator, whereas the corresponding rat cells produce tissue type plasminogen activator. J Cell Biol 105:977-981, 1987 4. Chang J, Gilman SC, Lewis AJ: Interleukin-1activates phospholipase A2 in rabbit chondrocytes: A possible signal for IL-1 action. J Zmmunol 136: 1283-1287, 1986 5 . Chengshi H, Wilhelm SM, Pentland AP, Manner BL, Grant GA, Eisen AZ, Goldberg GI: Tissue cooperation in a pro-
315
teolytic cascade activating human interstitial collagenase. Proc Natl Acad Sci 86:2632-2636, 1989 6. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium-thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156159, 1987 7. Chrisman OD, Southwick WO, Fessell JM: Plasmin and articular cartilage. Yale J Biol Med 34524-526, 1962 8. Cohen RL, Niclas J, Lee WMF, Wun T-C, Crowley CW, Levinson AD, Sadler JE, Shuman MA: Effects of cellular transformation on expression of plasminogen activator inhibitors 1 and 2. J Biol Chem 264:837543383, 1989 9. Curtiss PH Jr, Klein L: Destruction of articular cartilage in septic arthritis. I. In vitro studies. J Bone Joint Surg 45A: 797-806, 1963 10. Dingle JT, Dingle TT: The site of cartilage matrix degradation. Biochem J 190:431438, 1988 11. DiPasquale G, Caccese R, Pasternak R, Conaty J, Hubbs S, Perry K: Proteoglycan and collagen-degrading enzymes from human interleukin 1-stimulatedchondrocytes from several species: Proteoglycanase and collagenase inhibitors as potentially new disease-modifying antiarthritic agents. Proc Soc Exp Biol Med 183:262-267, 1986 12. Eyre DR, Apon S, Wu J-J, Ericsson LH, Walsh KA: Collagen type IX: Evidence for covalent linkages to type I1 collagen in cartilage. FEBS Lett 220:337-341, 1987 13. Feinberg AP, Vogelstein B: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1325-13, 1983 14. Feinberg AP, Vogelstein B: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Addendum Anal Biochem 137:26&267, 1984 15. Ginsberg D, Zeheb R, Yang AY, Raf€erty VM, Andreasen PA, Nielson L, Dan0 K, Lebo RV, Gelehrter TD: cDNA cloning of human plasminogen activator-inhibitor from endothelial cells. J Clin Invest 78:1673, 1986 16. Goldring MB , Krane SM: Modulation by recombinant interleukin 1 synthesis of types I and I11 collagens and associated procollagen mRNA levels in cultured human cells. J Biol Chem 262: 16724-16729, 1987 17. Hedden DM, Ehrlich MG, Towle CA, Treadwell BV: The effects of IL-1 on the growth plate chondrocyte. Trans Orthop Res Soc 14591, 1989 18. Hekman CM, Loskutoff DJ: Bovine plasminogen activator inhibitor 1: Specificity determinations and comparison of the active, latent, and guanidine-activated forms. Biochem 27~2911-2918, 1988 19. Hekman CM, Loskutoff DJ: Kinetic analysis of the interactions between plasminogen activator inhibitor 1 and both urokinase and tissue plasminogen activator. Arch Biochem Biophys 262: 199-210, 1988 20. Heussen C, Dowdle: Electrophoretic analysis of plasminogen activators in polyacrylamide gels containing sodium dodecyl sulfate and copolymerized substrates. Anal Biochem 102: 196-202, 1980 21. Howard EW, Knauer DJ: Human protease nexin-1. Further characterization using a highly specific polyclonal antibody. J Biol Chem 261584, 1986 22. Katagiri K, Okada K, Hattori H, Yano M: Bovine endothelial cell plasminogen activator inhibitor. Purification and heat activation. Eur J Biochem 176:81437, 1988 23. Lack CH, Anderson AJ, Ali SY: Action of plasmin on cartilage matrix in vivo. Nature 191:1402-1403, 1961 24. Lack CH, Ali SY: The degradation of cartilage by enzymes. In: Cartilage Degradation and Repair, ed by CAL Bassett, Washington, D.C., National Research Council-National Academy of Sciences, 1967, pp 61-75 25, Laemmli UK: Cleavage of structural proteins during the as-
J Orthop Res, Vol. 9, No.3, 1991
316
26.
27. 28. 29.
30.
31. 32. 33.
34.
35.
B . V . TREADWELL ET AL.
sembly of the head of bacteriophage T4. Nature (London) 22:680-684, 1970 Lin CW, Phillips SL, Brinckerhoff E, Georgescu HI, Bandara G, Evans CH: Induction of collagenase mRNA in lapine articular chondrocytes by synovial factors and interleukin-1. Arch Biochem Biophys 264:351-354, 1988 Martel-PelletierJ, Cloutier JM, Pelletier J-P: Possible role of plasmin and cathepsin B activating metalloproteinase in human OA joints. Trans Orthop Res SOC14:306, 1989 Mayne, R: Cartilage collagens. What is their function and are they involved in articular disease? Arthritis Rheum 32~241-246, 1989 Mayne R, Irwin MH: Collagen types in cartilage. In: Articular Cartilage Biochemistry: Workshop Conference, Hoechst-Werk Albert, Wiesbaden, ed by KE Kuettner, R Schleyerback, VC Hascall, New York, Raven Press, 1986, pp 23-35 Michel JB, Quertermous T: Modulation of mRNA levels for urinary and tissue type plasminogen activator and plasminogen activated inhibitors 1 and 2 in human fibroblasts by interleukin-1. J Zmmunol 143:89&895, 1989 Miller DR, Mankin HJ, Shoji H, D’Ambrosia RD: Identification of fibronectin in preparations of osteoarthritic human cartilage. Conn Tis Res 12:267, 1984 Miller RR, McDevitt CA: Thrombospondin is present in articular cartilage and is synthesized by articular chondrocytes. Biochem Biophys Res Comm 153:708-714, 1988 Mimuro J, Loskutoff DJ: Purification of a protein from bovine plasma that binds to type 1 plasminogen activator inhibitor and prevents its interaction with extracellular matrix. J Biol Chem 264:93&939, 1989 Ollivierre F, Gubler U, Towle CA, Laurencin C, Treadwell BV: Expression of IL-1 genes in human and bovine chondrocytes: A mechanism for autocrine control of cartilage matrix degradation. Biochem Biophys Res Comm 13 1 : s 911, 1986 Poole CA, Ayad S, Schofield I: Chondrons from articular cartilage: Immunolocalization of type VI collagen in the pencellular capsule of isolated canine tibia1 chondrons. J Cell Sci 90:635-645, 1988
J Orthop Res, Vol. 9, No. 3, 1991
36. Rifken DC, Collen D: Purification and characterization of the plasminogen activator secreted by human melanoma cells in culture. J Biol Chem 256:7035-7041, 1981 37. Robey PG, Young MF, Fisher LW, McClain TD: Thrombospondin is an osteoblast-derived component of mineralized extracellular matrix. J Cell Biol 108:719-727, 1989 38. Scarpati EM, Sadler JE: Regulation of endothelial cell coagulant properties. J Biol Chem 264:20705-20713, 1989 39. Schwartz BS: Monocyte synthesis of thrombospondin. The role of platelets. J Biol Chem 264:7512-7517, 1989 40. Scott RW, Baker JB: Purification of human protease nexin. J Biol Chem 258:10439, 1983 41. Scott RW, Bergman BL, Bajpai A, Hersh RT, Rodriguez H, Jones BN, Barreda C, Watts S, Baker JB: Protease nexin. Properties and a modified purification procedure. J Biol Chem 260:7029, 1985 42. Sprengers ED, Kluft C: Plasminogen activator inhibitors. Blood 69~381-387, 1987 43. Treadwell BV, Neidel J, Pavia M, Towle CA, Trice ME, Mankin HJ: Purification and characterization of collagenase activator protein synthesized by articular cartilage. Arch Biochem Biophys 251:715-723, 1986 44. Treadwell BV, Towle CA, Ishizue K, Mankin KP, Pavia M, Ollivierre FM, Gray DH: Stimulation of the synthesis of collagenase activator protein in cartilage by a factor present in synovial-conditioned medium. Arch Biochem Biophys 251:724-731, 1986 45. Tyler JA: Insulin-like growth factor 1 can decrease degradation and promote synthesis of proteoglycan in cartilage exposed to cytokines. Biochem J 260543-548, 1989 46. Van der Rest M, Mayne R: Type IX collagen proteoglycan from cartilage is covalently cross-linked to type I1 collagen. J Biol Chem 263:1615-1618, 1988 47. Wurster NB, Lust G: Fibronectin in osteoarthritic cartilage-a possible indication of phenotype modulation of the chondrocyte? In: Degenerative Joints, vol2, ed by G Verbruggen, EM Veys, Proceedings of the Second Conference on Degenerative Joint Diseases, Ghent, 29 November-1 December 1984. International Congress Series 668. Amsterdam: Excerpta Medica, 1985, pp 141-147
Cartilage Synthesizes the Serine Protease Inhibitor PAI- 1: Support for the Involvement of Serine Proteases in Cartilage Remodeling Benjamin V. Treadwell, Michele Pavia, Christine A. Towle, Vernon J. Cooley, and Henry J. Mankin Orthopaedic Research Laboratories, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, U.S.A.
~~
Summary: The work described here demonstrates the synthesis by human articular cartilage of plasminogen activator inhibitor-1 (PAI-l), a potent inhibitor of the serine protease tissue plasminogen activator (tPA). We also present data demonstrating an increase in PAI-1 messenger ribonucleic acid (mRNA) in chondrocytes" exposed to the cytokine interleukin-1 (IL-1). Interestingly, this elevation of steady-state mRNA levels does not appear to result in an increase in synthesis of PAI-1 protein. Northern blot analysis reveals that of the two mRNA species (3.4 kb, 2.4 kb) previously reported for PAI-1, only the larger species (3.4 kb) appears to be synthesized by chrondrocytes. Our data demonstrate the IL- l-stimulated production by cartilage of tissue plasminogen activator. We also show evidence for the presence of plasminogen in cartilage. A scheme is presented indicating the probable importance of the serine proteases (tPA and plasminogen) and PAI-1 in cartilage degradation. Key Words: Cartilage-Proteases-Inhibitors-Interleukin1-Plasminogen activator.
The major macromolecular components of cartilage matrix are type I1 collagen and a series of highmolecular-weight proteoglycans (28). In the past two decades, however, analytical studies have demonstrated the presence of a number of quantitatively minor components, which include at least five additional collagen types and a number of other proteins? such as fibronectin, laminin, and thrombospondin (12,28,29,31,32,35,46,47). Cartilage is capable of synthesizing and remodeling matrix with its own machinery. Thus the chondrocyte synthesizes not only all the matrix compo-
nents but also the proteases required for degradation and the regulators (activators and inhibitors) of these proteases. In prior studies we have shown that interleukin-1 (IL-1) is a key component of this remodeling network; it has been shown to stimulate the synthesis of both stromelysin and collagenase by chondrocytes (11,16,17,27,4345). IL-1 has recently been demonstrated to be synthesized by the chondrocytes themselves (34). The work presented in this article further extends our knowledge regarding the action of IL-1 on the degradative cascade in cartilage. We demonstrate that the synthesis by chondrocytes of plasminogen activator inhibitor-1 (PAI-I) is, in part, regulated by IL- 1, which maintains the steady-state level of PAII-specific messenger ribonucleic acid (mRNA). We also demonstrate that cartilage contains two pro-
Received April 30, 1990; accepted October 31, 1990. Address correspondence and reprint requests to Dr. Benjamin V. Treadwell, Orthopaedic Research Laboratories, Massachusetts General Hospital, Jackson 1 1 , Boston, MA 02114, U.S.A.
309
B . V . TREADWELL ET AL.
310
tease components of the fibrinolytic cascade, tissue plasminogen activator (tPA) and plasminogen, and that IL-1-stimulated cartilage has marked elevation of tPA activity. We propose a scheme in which we postulate the probable importance of the serine proteases in cartilage degradation and demonstrate the effects of IL-1 on the activity of these enzymes.
periments, were incubated at 37°C in an atmosphere of 95% air, 5% CO,. The medium was harvested after 24 h. This period of time was chosen for convenience, since we found in previous experiments that the effects of IL-1 on the parameters investigated did not significantly vary between 6-24 h.
MATERIALS AND METHODS
Harvested media samples were prepared for SDS polyacrylamide gel electrophoresis as follows: medium (200 p.1) from each well was treated with 2 volumes of acetone at -20°C for 1 h. The precipitated material was collected by centrifugation. The pellets were lyophilized and resolubilized in 20 p1 of sample buffer (2.5% SDS, 1% glycerol, 10 mM TrisHC1, pH 7.0, and 20 pg/ml phenol red). Samples to be analyzed for enzyme activity were incubated at 37°C for 20 min before gel application. All other samples were heated at 90°C for 1 min in the presence of 1% 2-mercaptoethanol.
Protein A Sepharose CL-4B and human plasminogen were purchased from Sigma Chemical Co. (St. Louis, MO, U.S.A.). Human plasminogen antiserum was purchased from CalBiochem Corp. (La Jolla, CA, U.S.A.). Human recombinant tPA and monoclonal antibody were from Genentech, Inc. (South San Francisco, CA, U.S.A.). Bovine PAI-1 antibody (raised in rabbits) was a gift from David J. Loskutoff (Scripps Clinical Research Foundation, La Jolla, CA, U.S.A.). Human recombinant IL1 alpha was a gift from Peter Lomedico (HoffmannLaRoche, Nutley, NJ, U.S.A.). Plasmid PAI-B6 was a gift from David Ginsburg (Howard Hughes Medical Institute, University of Michigan, Ann Arbor, MI, U.S.A.). Dulbecco’s modified Eagle medium (DMEM) was purchased from Gibco (Long Island, NY, U.S.A.). All electrophoresis reagents were purchased from Gallard Schlesinger (Carle Place, NY, U.S.A.). Tran ,%-label was purchased from ICN BioMedicals, Inc. (Irvine, CA, U.S.A.). Casein was from Fisher Scientific Co. (Fairlawn, NJ, U.S.A.).
Preparation of Samples for Sodium Dodecyl Sulfate (SDS) Gel Electrophoresis
SDS-Polyacrylamide-CaseinGel Electrophoresis for Detection of Plasminogen Activator (PA) Activity
The media samples, prepared as described above, were applied to an 11% polyacrylamide gel prepared according to the procedure of Laemmli (25), except that the acrylamide solution contained casein (1 mg/ml) and, where indicated, human plasminogen (5 pg/ml). Preparation of the gel containing casein and analysis of enzyme activity is essentially according to the procedure of Heussen and Dowdle (20). Preparation of Cartilage-Conditioned Medium Electrophoresis was run at a constant 200 V at 4°C. Following electrophoresis, the gels were Cartilage cores from the articular surface of norplaced in a solution of 2.5% Triton X-100 and agimal human knee joints were obtained at the time of tated for 1 h at room temperature. The gels were amputation using a cork borer (3 mm in diameter). then incubated 3 hr at 37°C in 20 mM Tris-HC1, 150 Amputations were performed on patients with mM NaCl, 10 mM ethylenediaminetetracetate chondrosarcomas in the hip area. Discs (1.5 mm (EDTA), 0.02% NaN,, pH 7.5. Immediately followthick) were sliced off from the outer layer of cores ing incubation, gels were stained in Coomassie blue of cartilage taken from the superficial surface of the and destained in 7% acetic acid. Evidence for PA femoral condyle using a specially constructed temactivity was the appearance of a clear band against plate, as described by Ollivierre and colleagues a dark blue background on the plasminogen gels and (34). The discs were placed in wells of a 96-well of these bands in the gel without plasthe absence Falcon plate containing 200 pl of DMEM, and, minogen. where indicated, 20 pCi Tran 35S-label(70% L - [ ~ ~ S ] methionine, 15% ~-[~~S]cysteine; specific activity SDS-Polyacrylamide-casein Gel Electrophoresis for 1,194 Cdmmol), 15 units of human recombinant ILDetection of Plasminogen 1, and cycloheximide (CX) (10 pg/ml) were added to The preparation of the sample and the polyacrylwells. The cartilage plugs, in triplicate for each conamide gel containing casein (no plasminogen) as dition, and, with a minimum of three separate ex-
J Orthop Res, Vol. 9, No. 3, 1991
311
FACTORS INVOLVED IN CARTILAGE DEGRADATION
well as the condition of electrophoresis were as described above. After the gels had been washed in 2.5% Triton X-100 for 1 h, they were placed in plastic bags containing 2 ml Tris-buffered saline (TBS) and tPA (5 pg/2 ml), where indicated. The bags containing the gel and solution were hermetically sealed and incubated for 3 h at 37°C with agitation. The tPA-dependent protease activity was detected on the Coomassie blue-stained gels as described in the section above.
in sample buffer containing 1% 2-mercaptoethanol. For detection of enzyme activity, the immunoprecipitates were analyzed on casein or casein plus plasminogen, as described above. For detection of immunoprecipitated radioactive proteins, gels were dried under vacuum following destaining, and the dried gel was placed in contact with x-ray film.
Immunoprecipitation of Specific Proteins Present in Cartilage-Conditioned Medium
Human articular cartilage was resected from the articular surface of a knee joint. The cells were liberated from the matrix by treatment with clostridal collagenase (0.1%) in DMEM for 18 h, as previously described (30). The isolated cells were rested as a suspension culture (1 x lo6 cells/ml) for 6-8 h in DMEM containing 5% fetal calf serum. Interleukin1 was added to one-half of the suspended cells at a concentration of 50 units/ml, and incubations were continued at 37°C in an atmosphere of 5% CO, for 18 h.
Cartilage-conditioned medium was removed from the well after incubation, adjusted to 0.02% NaN,, and incubated with antibody [anti-(PAI-1, plasminogen, or tPA)] or preimmune serum [normal rabbit serum (NRS)] according to the following procedure. Two hundred microliters of cartilage-conditioned medium was incubated with serum (or ascites for tPA antibody) equivalent to 40 pg IgG, either NRS or anti-PAI-1, -tPA, or -plasminogen, in the presence of 1 mM phenyl methyl sulfonyl fluoride (PMSF) and 10 mM EDTA for 1 h at 25°C and overnight at 4°C. To each sample was added 20 p1 of a protein-A Sepharose bead suspension [protein-A Sepharose equilibrated and suspended in an equal volume of buffer I (1% Triton X-100, 0.05% Tween 20,0.1% SDS, 0.02% NaN,, 300 mM NaCl, 10 mM EDTA, 20 mM Tris HCl, pH 7.5, and 10 Ulml aprotinin)]. Samples were vortexed at room temperature every 10 min for 1 h. Goat anti-mouse (40 pg) IgG was used to precipitate the monoclonal antibody to tPA. IgG-bound Sepharose and goat anti-mouse IgGanti-tPA complex was collected by centrifugation (8,OOOg for 1 min) in a microfuge. Supernatant was aspirated, and pellets were washed twice with 200 p1 of buffer I. The goat anti-mouse IgG complex pellet was immediately treated with SDS sample buffer as described below. The IgG-containing Sepharose pellets were resuspended in 20 p1 buffer I. This suspension was removed from tubes and overlayed onto 700 p1 of buffer I containing 10% glycerol. Protein-bound Sepharose beads were finally collected by centrifugation. The bound material was released from the protein A Sepharose beads, and the goat anti-mouse IgG complex was disrupted by either incubating the pellet in SDS sample buffer at 37°C for 20 min (if enzyme activity was to be determined) or heating at 90°C for 1 min
Isolation of Chondrocytes and Incubation with Interleukin-1
Preparation of Chondrocyte RNA for Northern Analysis Cellular RNA was extracted using the RNAzol Method (CinndBiotecx Laboratories International, Inc., Friendswood, TX, U.S.A.), which is a singlestep method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction (6). For Northern analysis, 10 pg of RNA per lane was applied to a formaldehyde (2.2 h4) agarose, (1%) gel. The RNA in the samples was separated by electrophoresis at 60 V for 5 h. The RNA was transferred by capillary action to a nitrocellulose membrane (Schleicher & Schuell Inc., Keene, NH, U.S.A.). After transfer, the RNA was immobilized by baking at 80°C in a vacuum oven for 1 h. PAI-1 mRNA was detected using a 2 kb Eco R1 cDNA fragment of plasmid PAI-B6 (15). The PAI-1 cDNA probe was prepared using a multiprime DNA labeling system with 32P-dCTP (Amersham International, Amersham, U.K.), according to the procedure of Feinberg and Vogelstein (13,14). The membrane was hybridized overnight at 42°C in 5 X SSC 50% formamide, 0.1% SDS, salmon sperm DNA 100 pg/ml, 10% dextran, and 1X Denhardt’s solution, and subsequently washed under stringent conditions using standard techniques (1). The blot was placed in contact with XAR film (Eastman Kodak Co., Rochester, NY, U.S.A.) and left at -40°C for 12 h with
J Orthop Res, Vol. 9, No. 3, 1991
B . V . TREADWELL ET AL.
312
1 2 3
intensification screens and the radioactive bands visualized after development in an X-Omat developer.
50K-
RESULTS IL-1-stimulated Synthesis of Tissue Plasminogen Activator Increased levels of plasminogen activator activity are evident in the medium conditioned by cartilage for 24 h in the presence of IL-1 (Fig. 1). The figure demonstrates plasminogen-dependent protease activity in cartilage-conditioned medium. This activity is markedly stimulated by IL-1, as is demonstrated in lane 2 ( + IL-1) compared with lane 1 ( - IL-1). Lane 2 shows a clear area (suggestive of protease activity) at M, 75-80 x lo3 that is not evident in the absence of plasminogen (lane 3). This in situ conversion of the latent plasminogen to active plasmin within the gel is demonstrative of PA activity. The PA activity was immunoprecipitated by antibody to the tPA species (lane 5). Lane 7 shows that no pro1
2
3
4
5
6
?
68K-
FIG. 2. Detection of plasminogen activator inhibitor-1 synthesis in cartilage-conditioned medium. Cartilage-conditioned medium taken from samples incubated in the presence of [35S]methionine for 24 h (2IL-1). The medium was subsequently incubated with a serum fraction and the precipitates separated on SDS-polyacrylamide gels. Lane 1: Control medium precipitated with normal rabbit serum. Lane 2: Control medium precipitated with rabbit anti-PAl-1. Lane 3: As in lane 2, except IL-1 was present during the 24-h incubation with cartilage.
tease activity was precipitated if normal mouse ascites fluid was substituted for the tPA antibody. Evidence suggesting that the stimulation of tPA activity by IL-1 (lane 2 versus lane 1) is the product of increased synthesis of the enzyme and not simply increased secretion is demonstrated by the results shown in Fig. 1, lane 4. This figure shows that the translational inhibitor cycloheximide prevents the stimulation of tPA activity with IL-1. Synthesis of Plasminogen Activator Inhibitor I
FIG. 1. Effect of IL-1 on synthesis of tPA. Cartilage-conditioned medium was applied to SDS polyacrylamide gels containing either casein and plasminogen or casein alone (lane 3) and separated by electrophoresis. Lane 1: Acetoneprecipitated 24-h cartilage-conditioned medium. Lane 2: As in lane 1, except IL-1 (30 units) was present with cartilage during the 24-h incubation step. Lane 3: Sample as in lane 2 applied to casein gel lacking plasminogen. Lane 4: As in lane 2, except cycloheximide was present during the 24-h incubation of cartilage with IL-1. Lane 5: Samples as in lane 2 immunoprecipitated with anti-human tPA. Lane 6 : Commercially available human recombinant tPA (0.2 ng). Lane 7: As in lane 5, except the tPA antibody was replaced by normal mouse serum.
J Orthop Res. Vol. 9, No. 3, 1991
Plasminogen activator inhibitor I synthesis by cartilage explants is demonstrated in autoradiographs shown in Fig. 2. The results show the synthesis and secretion from cartilage of a radioactive peptide immunoreactive with antibody raised to purified bovine endothelial cell PAI-1. The synthesis of this inhibitor by chondrocytes appears to be unchanged by the presence of IL-1 in the culture medium. This is indicated by the similar intensities of the immunoprecipitated radioactive band of M, 50,000 shown in the autoradiographs (lanes 2, 3) of Fig. 2. 11-1-treated Chondrocytes Have Increased Levels of PAI-1 mRNA In contrast to the lack of effect IL-1 has on the synthesis of PAI-1 protein, chondrocytes do respond to this cytokine at the mRNA level. Figure 3
FACTORS INVOLVED IN CARTILAGE DEGRADATION
1
2
3
3.4 2.4
-
FIG. 3. Northern analysis of cartilage RNA. RNA (10 pg) isolated from either human endothelial or cartilage cells was separated on agarose gels and subsequently tested for its ability to hybridize with a 32P-PAI-lcDNA probe. Lane 1: Endothelial cell RNA. Lane 2: Cartilage cell RNA. Lane 3: RNA isolated from IL-1-activated cartilage cells.
demonstrates a substantial (four- to fivefold) increase in the levels of mRNA coding for this inhibitor. This level of PAI-1 message, however, is considerably below that obtained with comparable amounts of endothelial cell RNA (lane 1). Furthermore, the endothelial cell PAI-1 mRNA is of two sizes-3.4 and 2.4 kb-whereas the chondrocyte contains only the 3.4 kb species.
313
(2,11,16,26,45) and affects the level of mRNA coding for at least one protease inhibitor, PAI-1. As suggested by the scheme shown in Fig. 5, the pivotal enzyme in this pathway is tPA. This enzyme is virtually undetectable in medium conditioned by unstimulated cartilage (see Fig. 1). PAI-1, a potent inhibitor of tPA (18,19,36), is present in cartilage and actively synthesized by it, thus allowing for the possibility that PAI-1 inactivates any tPA present. Although not described in this article, we have included in Fig. 5 the recent report showing that IL-1 stimulates the synthesis of phospholipase A,. This is a key enzyme involved in the synthesis of the precursor to the eicosanoids, compounds clearly implicated in connective tissue catabolism (4). There are several immunologically distinct types of plasminogen activator inhibitors reported in the literature (22). The inhibitor reported here (PAI-1) to be synthesized by cartilage is of the endothelial cell type (33). A second PA inhibitor (PAI-2) has been shown to be synthesized by human placenta
Plasminogen in Cartilage
The presence of plasminogen activity in cartilageconditioned medium is shown in Fig. 4. Cartilageconditioned medium was applied to a casein-containing SDS polyacrylamide gel and the components separated by electrophoresis. After electrophoresis the gel was incubated in buffer either with or without purified recombinant tPA. The gel incubated in the presence of tPA (lane 2) shows a clear band at M, 85,000. Interleukin-1 does not appear to stimulate the synthesis of this enzyme, but rather a slight inhibition appears to be evident (lane 3). We believe that this unexpected result may be due to an increase in the activation and subsequent degradation of plasminogen in the IL-1-treated cartilage. DISCUSSION
The results reported in this study demonstrate that the serine proteases tPA and plasminogen are present in cartilage and that there is a significant role for these proteases and their inhibitors in the control of cartilage degradation. Interleukin- 1 appears to play a key role in this pathway since it stimulates the synthesis of several proteases
FIG. 4. Presence of plasminogen in cartilage-conditioned medium. Acetone-precipitated samples were applied to an SDS polyacrylamide gel containing casein, and after electrophoresis the gel was incubated in buffer containing tPA. Lane 1: Cartilage-conditioned medium after electrophoresis on casein gel. tPA was omitted from the gel buffer incubation step. Lane 2: As in lane 1, but tPA was present during the gel incubation step. Lane 3: As in lane 2, but IL-1 was present during the 24 h incubation with cartilage. Lane 4: Commercially available plasminogen (0.2 pg) electrophoretically separated. The gel was incubated in the presence of tPA. Lane 5: As in lane 4, but tPA was omitted from the gel incubation step.
J Orthop Res, Vol. 9, No. 3, 1991
B . V . TREADWELL ET AL.
314
2 HOURS
INFLA MMA T I 0N
24-48 HOURS
Enzyme synthesis Inhibitor synthesis
CAR TILA GE DESTRUCTION
Enzyme activation
JOINT CAPSULE COLLAGENASE (A)
SYNOVIOCYTES FIBROBLASTS COLLAGENASE
NEUTROPHILS-IL1 MACROPHAGES / CHONDROCYTES
/ PLASMINOG
FIG. 5. Cartilage remodeling. A scheme produced from accumulated data to help explain cartilage remodeling and the role played by those factors believed to be involved (34). The synthesis by chondrocytes of phospholipase A, has been reported to be stimulated by IL-1 (4). The 92,000 gelatinase has been reported to be synthesized in response to IL-1 in bovine growth plate chondrocytes (17). (L) and (A) refer to the latent and active forms of the enzymes, respectively.
(probably macrophage derived) (8). Both these inhibitors have more recently found to be synthesized by a variety of cell types (30,38). A third inhibitor capable of inhibiting plasminogen activator is protease nexin. It was first isolated from fibroblastconditioned medium and has a more broad spectrum of inhibitory activity against serine proteases. Protease nexin-1 cannot be considered a true PA inhibitor, since it is more effective an inhibitor against thrombin and plasmin (21,23,4042). Of interest is the result we report of the increased levels of PAI-1 mRNA in cells treated with IL-1. The elevated levels of PAI-1 mRNA, however, do not appear to affect the level of expressed product. At present we do not know how to interpret this result except to say that it may be related to the species of PAL1 message (3.4 kb) affected by IL-1. This message species has been reported to be less stable than the other PAI-1 message of 2.4 kb (15). Another, and perhaps more likely, explanation attributes the lack of increased expressed product to the relatively (as compared to endothelial cell PAI-1 message) minor increase in mRNA levels with the cytokine. Speculation on the probable involvement of plas-
J Orthop Res, Vol. 9, No. 3, 1991
min in the degradative pathway of connective tissue, including cartilage, dates back to the 1930s (7,9,23,24). Recent information reported here and elsewhere allows us to present a more convincing scheme supporting a role for this enzyme in matrix degradation (2,5,27). Plasminogen, although perhaps not actively synthesized by the chondrocyte, is present in the matrix in significant quantities. The matrix appears to function as a sink for this protein, and it is not easily removed from the matrix. A possible explanation for this is the recent report of the synthesis by cartilage of a new matrix protein, thrombospondin (32), which contains as part of its structure a binding site for plasminogen. Thrombospondin has also been shown to bind tPA, heparin, fibronectin, laminin, and collagen (37,39). These properties suggest that the protein thrombospondin acts as a nucleus for the initiation site of cartilage degradation by plasmin. Further support for this hypothesis is the localized nature of matrix proteolysis (i.e. , pericellular), as has been reported by Dingle and Dingle (10). The results reported here and elsewhere showing that cartilage synthesizes tPA rather than urinary plasminogen activator (uPA) are somewhat surpris-
FACTORS INVOLVED IN CARTILAGE DEGRADATION
ing (2). A generally held belief is that uPA is involved in tissue destruction and tPA in thrombolysis (for review see 3 3 ) . This conclusion comes largely from studies on invasive connective tissue tumors, which have been demonstrated to synthesize large amounts of uPA. It appears that a major trigger for cartilage degradation is under the control of the IL-1 receptor. The cytokine induces the synthesis of tPA and metalloproteases in chondrocytes. Tissue plasminogen activator catalyzes the formation of the nonspecific and very active protease plasmin from the ubiquitous protein (160 pg/ml in serum) plasminogen. Plasmin is the probable activator of two additional matrix-degrading proteases, collagenase and stromelysin (5). The synthesis of these two metalloproteases is also markedly increased in IL- 1-activated chondrocytes (11,44). These results and the recent reports demonstrating the synthesis of al,antitrypsin by cartilage strongly suggest an important role for serine proteases and their inhibitors in cartilage remodeling and degradation. The primary role for collagenase, once activated by stromelysin (5) and by plasmin via tPA, is to digest the framework, whereas the noncollagenous matrix component is probably largely digested by the very active enzyme plasmin and the metalloprotease stromelysin. Acknowledgment: W e thank D. J. Loskutoff of t h e Scripps Institute, La Jolla, California, for the antibody to PAI-1, Peter Lomedico of Hoffmann-LaRoche for recombinant human IL-la, and Thomas Quertermous of Massachusetts General Hospital for his help in Northern blot analysis of PAI-1 mRNA.
REFERENCES 1. Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, eds: Current Protocols in Molecular Biology. New York, Wiley, 1987 2. Bunning RAD, Crawford A, Richardson HJ, Opdenakker G, Van Damme J, Russell RGG: Interleukin 1 preferentially stimulates the production of tissue-type plasminogen activator by human articular chondrocytes. Biochim Biophys Acta 924:473482, 1987 3. Canipari R, O’Connell ML, Meyer G, Strickland S: Mouse ovarian granulosa cells produce urokinase-type plasminogen activator, whereas the corresponding rat cells produce tissue type plasminogen activator. J Cell Biol 105:977-981, 1987 4. Chang J, Gilman SC, Lewis AJ: Interleukin-1activates phospholipase A2 in rabbit chondrocytes: A possible signal for IL-1 action. J Zmmunol 136: 1283-1287, 1986 5 . Chengshi H, Wilhelm SM, Pentland AP, Manner BL, Grant GA, Eisen AZ, Goldberg GI: Tissue cooperation in a pro-
315
teolytic cascade activating human interstitial collagenase. Proc Natl Acad Sci 86:2632-2636, 1989 6. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium-thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156159, 1987 7. Chrisman OD, Southwick WO, Fessell JM: Plasmin and articular cartilage. Yale J Biol Med 34524-526, 1962 8. Cohen RL, Niclas J, Lee WMF, Wun T-C, Crowley CW, Levinson AD, Sadler JE, Shuman MA: Effects of cellular transformation on expression of plasminogen activator inhibitors 1 and 2. J Biol Chem 264:837543383, 1989 9. Curtiss PH Jr, Klein L: Destruction of articular cartilage in septic arthritis. I. In vitro studies. J Bone Joint Surg 45A: 797-806, 1963 10. Dingle JT, Dingle TT: The site of cartilage matrix degradation. Biochem J 190:431438, 1988 11. DiPasquale G, Caccese R, Pasternak R, Conaty J, Hubbs S, Perry K: Proteoglycan and collagen-degrading enzymes from human interleukin 1-stimulatedchondrocytes from several species: Proteoglycanase and collagenase inhibitors as potentially new disease-modifying antiarthritic agents. Proc Soc Exp Biol Med 183:262-267, 1986 12. Eyre DR, Apon S, Wu J-J, Ericsson LH, Walsh KA: Collagen type IX: Evidence for covalent linkages to type I1 collagen in cartilage. FEBS Lett 220:337-341, 1987 13. Feinberg AP, Vogelstein B: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem 1325-13, 1983 14. Feinberg AP, Vogelstein B: A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Addendum Anal Biochem 137:26&267, 1984 15. Ginsberg D, Zeheb R, Yang AY, Raf€erty VM, Andreasen PA, Nielson L, Dan0 K, Lebo RV, Gelehrter TD: cDNA cloning of human plasminogen activator-inhibitor from endothelial cells. J Clin Invest 78:1673, 1986 16. Goldring MB , Krane SM: Modulation by recombinant interleukin 1 synthesis of types I and I11 collagens and associated procollagen mRNA levels in cultured human cells. J Biol Chem 262: 16724-16729, 1987 17. Hedden DM, Ehrlich MG, Towle CA, Treadwell BV: The effects of IL-1 on the growth plate chondrocyte. Trans Orthop Res Soc 14591, 1989 18. Hekman CM, Loskutoff DJ: Bovine plasminogen activator inhibitor 1: Specificity determinations and comparison of the active, latent, and guanidine-activated forms. Biochem 27~2911-2918, 1988 19. Hekman CM, Loskutoff DJ: Kinetic analysis of the interactions between plasminogen activator inhibitor 1 and both urokinase and tissue plasminogen activator. Arch Biochem Biophys 262: 199-210, 1988 20. Heussen C, Dowdle: Electrophoretic analysis of plasminogen activators in polyacrylamide gels containing sodium dodecyl sulfate and copolymerized substrates. Anal Biochem 102: 196-202, 1980 21. Howard EW, Knauer DJ: Human protease nexin-1. Further characterization using a highly specific polyclonal antibody. J Biol Chem 261584, 1986 22. Katagiri K, Okada K, Hattori H, Yano M: Bovine endothelial cell plasminogen activator inhibitor. Purification and heat activation. Eur J Biochem 176:81437, 1988 23. Lack CH, Anderson AJ, Ali SY: Action of plasmin on cartilage matrix in vivo. Nature 191:1402-1403, 1961 24. Lack CH, Ali SY: The degradation of cartilage by enzymes. In: Cartilage Degradation and Repair, ed by CAL Bassett, Washington, D.C., National Research Council-National Academy of Sciences, 1967, pp 61-75 25, Laemmli UK: Cleavage of structural proteins during the as-
J Orthop Res, Vol. 9, No.3, 1991
316
26.
27. 28. 29.
30.
31. 32. 33.
34.
35.
B . V . TREADWELL ET AL.
sembly of the head of bacteriophage T4. Nature (London) 22:680-684, 1970 Lin CW, Phillips SL, Brinckerhoff E, Georgescu HI, Bandara G, Evans CH: Induction of collagenase mRNA in lapine articular chondrocytes by synovial factors and interleukin-1. Arch Biochem Biophys 264:351-354, 1988 Martel-PelletierJ, Cloutier JM, Pelletier J-P: Possible role of plasmin and cathepsin B activating metalloproteinase in human OA joints. Trans Orthop Res SOC14:306, 1989 Mayne, R: Cartilage collagens. What is their function and are they involved in articular disease? Arthritis Rheum 32~241-246, 1989 Mayne R, Irwin MH: Collagen types in cartilage. In: Articular Cartilage Biochemistry: Workshop Conference, Hoechst-Werk Albert, Wiesbaden, ed by KE Kuettner, R Schleyerback, VC Hascall, New York, Raven Press, 1986, pp 23-35 Michel JB, Quertermous T: Modulation of mRNA levels for urinary and tissue type plasminogen activator and plasminogen activated inhibitors 1 and 2 in human fibroblasts by interleukin-1. J Zmmunol 143:89&895, 1989 Miller DR, Mankin HJ, Shoji H, D’Ambrosia RD: Identification of fibronectin in preparations of osteoarthritic human cartilage. Conn Tis Res 12:267, 1984 Miller RR, McDevitt CA: Thrombospondin is present in articular cartilage and is synthesized by articular chondrocytes. Biochem Biophys Res Comm 153:708-714, 1988 Mimuro J, Loskutoff DJ: Purification of a protein from bovine plasma that binds to type 1 plasminogen activator inhibitor and prevents its interaction with extracellular matrix. J Biol Chem 264:93&939, 1989 Ollivierre F, Gubler U, Towle CA, Laurencin C, Treadwell BV: Expression of IL-1 genes in human and bovine chondrocytes: A mechanism for autocrine control of cartilage matrix degradation. Biochem Biophys Res Comm 13 1 : s 911, 1986 Poole CA, Ayad S, Schofield I: Chondrons from articular cartilage: Immunolocalization of type VI collagen in the pencellular capsule of isolated canine tibia1 chondrons. J Cell Sci 90:635-645, 1988
J Orthop Res, Vol. 9, No. 3, 1991
36. Rifken DC, Collen D: Purification and characterization of the plasminogen activator secreted by human melanoma cells in culture. J Biol Chem 256:7035-7041, 1981 37. Robey PG, Young MF, Fisher LW, McClain TD: Thrombospondin is an osteoblast-derived component of mineralized extracellular matrix. J Cell Biol 108:719-727, 1989 38. Scarpati EM, Sadler JE: Regulation of endothelial cell coagulant properties. J Biol Chem 264:20705-20713, 1989 39. Schwartz BS: Monocyte synthesis of thrombospondin. The role of platelets. J Biol Chem 264:7512-7517, 1989 40. Scott RW, Baker JB: Purification of human protease nexin. J Biol Chem 258:10439, 1983 41. Scott RW, Bergman BL, Bajpai A, Hersh RT, Rodriguez H, Jones BN, Barreda C, Watts S, Baker JB: Protease nexin. Properties and a modified purification procedure. J Biol Chem 260:7029, 1985 42. Sprengers ED, Kluft C: Plasminogen activator inhibitors. Blood 69~381-387, 1987 43. Treadwell BV, Neidel J, Pavia M, Towle CA, Trice ME, Mankin HJ: Purification and characterization of collagenase activator protein synthesized by articular cartilage. Arch Biochem Biophys 251:715-723, 1986 44. Treadwell BV, Towle CA, Ishizue K, Mankin KP, Pavia M, Ollivierre FM, Gray DH: Stimulation of the synthesis of collagenase activator protein in cartilage by a factor present in synovial-conditioned medium. Arch Biochem Biophys 251:724-731, 1986 45. Tyler JA: Insulin-like growth factor 1 can decrease degradation and promote synthesis of proteoglycan in cartilage exposed to cytokines. Biochem J 260543-548, 1989 46. Van der Rest M, Mayne R: Type IX collagen proteoglycan from cartilage is covalently cross-linked to type I1 collagen. J Biol Chem 263:1615-1618, 1988 47. Wurster NB, Lust G: Fibronectin in osteoarthritic cartilage-a possible indication of phenotype modulation of the chondrocyte? In: Degenerative Joints, vol2, ed by G Verbruggen, EM Veys, Proceedings of the Second Conference on Degenerative Joint Diseases, Ghent, 29 November-1 December 1984. International Congress Series 668. Amsterdam: Excerpta Medica, 1985, pp 141-147
