Metalloproteinase and Activation of - Europe PMC

2 downloads 0 Views 3MB Size Report
enous peroxidase and nonspecific binding, respec- tively. The sections were then treated with a mono- clonal antibody to MT1-MMP (20 jig/ml, clone 114-.
American Journal of Pathology, Vol. 151, No. 1, July 1997 Copynight X3 American Society for Investigative Pathology

Expression of Membrane-Type 1 Matrix Metalloproteinase and Activation of Progelatinase A in Human Osteoarthritic Cartilage

Kazushi Imai,* Satoru Ohta,*t Tadami Matsumoto,t Noboru Fujimoto,t Hiroshi Sato,§ Motoharu Seiki,§ and Yasunori Okada* From the Departments of Molecular Immunology and Pathologe and Molecular Virology and Oncology,5 Cancer Research Institute, and the Department of Orthopedic Surgery,t School of Medicine, Kanazawa University, Kanazawa, Ishikawa, and Fuji Chemical Industries Ltd.,t Takaoka, Toyama, Japan

Matrix metaloproteinases (MMPs) are expressed in osteoarthritic (OA) cartilage and are thought to be involved in the degradation of cartilage extracelular matrix (ECM). Among these proteinases, MMP-2 (gelatinase A) demonstrates a wide range of substrate specificity against the ECM present in cartilage. Although MMP-2 expression increases in OA cartilage, the activation mechanism of the corresponding zymogen (proMMP-2) in cartilage is unknown. In this study, we examined the expression pattern of membrane-type I MMP (MTI-MMP) in human OA articular cartilage and its correlation with the activation of pro-MMP-2. Immunohistochemical studies demonstrate that MTI-MMP localizes to the chondrocytes in the superficial and transitional zones in aU of the samples examined directly correlating with cartilage degradation. Reverse transcription polymerase chain reaction confirmed the predominant expression of MT1MMP mRNA in the OA cartilage. In situ hybridization revealed the site of expression of MT1MMP in OA cartilage to be the chondrocytes. Through gelatin zymography and a sandwich enzyme immunoassay it was demonstrated that OA cartilage explants secrete signiflcantly higher levels ofpro-MMP-2 than normal samples. Pro-MMP-2 activation was enhanced in the OA

cartilage samples and correlated with MT1-MMP expression in the cartilage. Plasma membranes preparedfrom cultured chondrocytes with MT1MMP expression and those directly isolatedfrom OA cartilage could activate pro-MMP-2. MT1MMP gene expression in cultured chondrocytes was induced by treatment with interleukin-1 a and/or tumor necrosis factor-ae. These data suggest that cytokine-induced MTI-MMP in the chondrocytes may play a key role in the activation of pro-MMP-2 in the OA articular cartilage, leading to cartilage destruction through ECM degradation. (Am JPathol 1997, 151:245-256)

Osteoarthritis (OA) is a common disease and increases progressively with age.1 It is characterized by loss of proteoglycans and collagens from the cartilage matrix, leading to fibrillation and eventually complete loss of cartilage exposing the underlying subchondral bone.2 Increased proteolytic activity in the joint is one of the possible mechanisms leading to cartilage destruction.3 As migration of inflammatory cells is minimal in the initial stage of the disease, proteinases derived from chondrocytes are most likely to be responsible for cartilage matrix degradation. Among the proteinases identified in chondrocytes, matrix metalloproteinases (MMPs) are thought to play a key role in the extracellular matrix (ECM) degradation of the cartilage.3 Supported by grant-in-aid 07457049 (to Y. Okada) and 5780 (to K. Imai) from the Ministry of Education, Science, and Culture of Japan. K. Imai was a recipient of Research Fellowship from the Japan Society for the Promotion of Science for Young Scientists. Accepted for publication March 26, 1997. Address reprint requests to Dr. Yasunori Okada, Department of Molecular Immunology and Pathology, Cancer Research Institute, Kanazawa University, 13-1 Takara-machi, Kanazawa, Ishikawa 920,

Japan.

245

246

Imai et al

AJPJuly 1997, Vol. 151, No. 1

MMPs are zinc-dependent neutral endopeptidases with a wide spectrum of substrate specificity and consist of at least 14 different members. They can be subdivided into five groups according to their structural and enzymic properties: interstitial collagenases, gelatinases/type IV collagenases, stromelysins, membrane-type MMPs (MT-MMPs) and others.4 Among the MMPs, MMP-2 (gelatinase A) is important to complete collagen degradation after the specific cleavage of the triple helical region of the fibrillar collagen molecules by collagenases. MMP-2 also digests other substrates including aggrecan, link protein, decorin, fibronectin, and type X and XI collagens,5-10 all of which are components of the articular cartilage matrix. It is therefore likely that MMP-2 contributes to the cartilage matrix breakdown when produced in the articular cartilage. Actually, enhanced expression of pro-MMP-2 is reported in human OA articular cartilage.11 However, as most MMPs are secreted from cells as inactive zymogens (pro-MMPs), activation of pro-MMPs is another key step for the MMPs to function in vivo. Serine proteinases may be important in the activation of many pro-MMPs such as pro-MMP-1 (tissue collagenase), pro-MMP-3 (stromelysin 1), and pro-MMP-9 (gelatinase B).3 However, pro-MMP-2 is unique in that its activation is not caused by serine proteinases but achieved by membrane-associated activators.8'12 We have previously cloned MT1-MMP and identified it as an activator of pro-MMP-2.13 However, very little is known about MT1-MMP expression and its relation to the pro-MMP-2 activation in the cartilage. In the present study, we demonstrate for the first time that OA articular cartilage highly expresses MT1-MMP correlating with pro-MMP-2 activation whereas normal cartilage exhibits negligible MT1MMP expression and pro-MMP-2 activation. Plasma membrane fractions isolated from chondrocytes with MT1-MMP expression can activate pro-MMP-2. Both MT1-MMP expression and pro-MMP-2 activation in cultured chondrocytes are stimulated by treatment with interleukin (IL)-la and less efficiently by tumor necrosis factor (TNF)-a. Our data suggest that MT1 MMP inducible by the cytokines may contribute to the activation of pro-MMP-2 facilitating the ECM degradation in human OA. -

Materials and Methods Source and Histology of Articular Cartilages Non-osteophytic cartilage samples were obtained from 39 hip and 25 knee joints of 64 OA patients (mean age, 64 + 12 years) undergoing operations

for total hip or knee joint replacement. Normal control samples without macroscopic and microscopic changes were taken from 11 hip and 4 knee joints of 15 patients (74 + 9 years) with femoral neck fracture or an amputation operation due to atherosclerotic gangrene or malignant mesenchymal tumors. All of the samples were cut into slices (-3 mm thick), fixed with periodate-lysine-paraformaldehyde (PLP) or 4% paraformaldehyde fixatives for -24 hours at 40C, and embedded in paraffin wax after decalcification with 0.5 mol/L EDTA, pH 7.4. PLP-fixed paraffin sections (4 l,m thick) were stained with hematoxylin and eosin or toluidine blue and subjected to histological/ histochemical grading according to Mankin et al.14 The Mankin grading system, which indicates the morphological damage and loss of proteoglycans histochemically in articular cartilage, is used as an index of cartilage degradation.

Immunohistochemistry and in Situ Hybridization Sections from the PLP-fixed samples (64 from OA and 15 from normal cartilage) were treated with 0.3% H202 and 10% normal horse serum to block endogenous peroxidase and nonspecific binding, respectively. The sections were then treated with a monoclonal antibody to MT1-MMP (20 jig/ml, clone 1146G6; H. Ueno, H. Nakamura, K. Imai, H. Sato, M. Seiki, and Y. Okada, Cancer Res., in press), the antibody absorbed with purified recombinant MT1MMP15 or nonimmune mouse IgG (20 ,tg/ml). After reactions with biotinylated horse IgG to mouse IgG (Vector Laboratories, Burlingame, CA) and an avidin-biotin-peroxidase complex (DAKO, Glostrup, Denmark), the color was developed with 3,3'-diaminobenzidine tetrahydrochloride as described previously. 16 Immunostaining for MMP-2 was performed in a similar fashion using a monoclonal antibody to MMP-2 (2 ,ug/ml, clone 75-7F7).17 For in situ hybridization, the cDNA (177 bp) specific to MT1-MMP was prepared by polymerase chain reaction (PCR) using an antisense primer (5'TCGGCCCAAAGCAGCAGCTTC-3') and a sense primer (5'-CTTCATGGTGTCTGCATCAGC-3') and subcloned into Bluescript 11 KS ± (Stratagene, La Jolla, CA). Paraffin sections of the tissues fixed with 4% paraformaldehyde (six OA and three normal samples) were treated with 20 ,ug/ml proteinase K (Promega Biotec, Oakland, CA) and then 0.0025% acetic anhydride (Eastman Kodak, Rochester, NY) in 0.1 mol/L triethanolamine, pH 8.0 (Eastman Kodak), as described previously. 18 After prehybridization,

MT1-MMP Expression in OA Cartilage

247

AJPJuly 1997, Vol. 151, No. 1

the sections were hybridized with 35S-labeled antisense RNA or sense RNA for -12 hours at 500C. The slides were then treated with 20 ,tg/ml RNAse A, washed under stringent conditions (2X SSC, 0.5X SSC, and 0.1X SSC, twice each for 30 minutes at 500C), and air dried. They were dipped in NR-M2 emulsion (Konica, Tokyo, Japan) and subjected to autoradiography by exposing for 7 to 10 days.

Cultures of Cartilage Slices and Detection of MMP-2 Full-thickness cartilage slices with (18 cases) or without (7 cases) OA changes were cultured in serumfree minimal essential medium (MEM) containing 0.2% lactalbumin hydrolysate for 3 days. The culture media (6 ,ul/mg tissue dry weight) were subjected to gelatin zymography using SDS-polyacrylamide gels containing 0.2% gelatin. The intensity of the gelatinolytic bands corresponding to pro-MMP-2 of Mr 68,000 and active MMP-2 of Mr 62,000 was measured by computer-assisted image analysis according to the method of Davies et al,19 and the activation ratio (the ratio of the active MMP-2 form to proMMP-2 and active forms) was estimated. Pro-MMP-2 levels in the media were measured by a sandwich enzyme immunoassay specific to pro-MMP-2.17 Gelatinolytic activities were assayed in a solution assay using 14C-gelatin in the presence and absence of 1 mmol/L p-aminophenylmercuric acetate (APMA).8 The assay was performed in the presence of 2 mmol/L phenylmethylsulfonyl fluoride and 5 mmol/L N-ethylmaleimide to inhibit serine and cysteine proteinases. One unit of the activity degraded 1 ,tg of

gelatin/minute at 370C.

Isolation of Chondrocytes and Cell Culture Chondrocytes were obtained from OA or normal cartilages by digestion of the minced cartilages with 0.4% (w/v) Pronase (Calbiochem, La Jolla, CA) for 1 hour at 370C followed by 0.025% (w/v) bacterial collagenase type (Worthington Biochemical Corp., Freehold, NJ) for 12 to 16 hours at 370C.20 For monolayer cell cultures, isolated cells were plated on culture flasks at 5 x 104 cells/cm2 in MEM supplemented with 10% fetal bovine serum until the cells substantially reached confluence. The culture media of chondrocytes isolated from two OA cases were replaced with serum-free MEM containing 0.2% lactalbumin hydrolysate and then treated with 100 U/mI IL-ia (Dainippon Pharmaceutical Co., Osaka, Japan) and/or 500 U/ml TNF-a (Dainippon) for 36 hours.

Reverse Transcription (RT)-PCR Total RNA was isolated from chondrocytes (seven OA and four normal cartilage samples) using the acid guanidium-phenol-chloroform method.21 RNA extracted was treated with RNAse-free DNAse (Boehringer Mannheim, Mannheim, Germany) to eliminate DNA contamination in the samples and converted to a single-stranded cDNA using a random oligonucleotide hexamer (Takara, Otsu, Japan). Randomly primed cDNAs were prepared from 5 ,ug of total RNA by Moloney murine leukemia virus reverse transcriptase (Gibco BRL, Gaithersburg, MD) followed by PCR amplification. The primers prepared for in situ hybridization were utilized for the PCR amplification of MT1-MMP. As a control, for the PCR reaction, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was amplified with a 5'-CCACCCATGGCAAATTCCATGGCA-3' (forward primer) and a

5'-TCTAGACGGCAGGTCAGGTCCACC-3' (reverse primer). PCR amplification was performed by running 30 cycles under the following conditions: denatured at 94°C for 30 seconds, annealed for 30 seconds at 560C for MT1-MMP or at 500C for GAPDH, and extended at 720C for 1 minute.

Northern Blotting Total RNA was extracted by the acid guanidiumphenol-chloroform method from primary cultures of OA chondrocytes (two cases) stimulated with 100 U/ml IL-la and/or 500 U/ml TNF-a in serum-free MEM for 36 hours. RNA (15 ,ug) was electrophoresed on 1.0% formaldehyde-agarose gels and transferred to nylon filters (Hybond-N, Amersham, Tokyo, Japan) followed by cross-linking with ultraviolet light. Filters were hybridized with 32P-labeled probes for the MT1-MMP gene as described previously.22

Preparation of Chondrocyte Membrane Fractions Cultured chondrocytes (three OA cases and one normal case) and chondrocytes directly isolated from four OA and two normal articular cartilages by the sequential enzymic treatment were homogenized in 10 mmol/L Tris/HCI, pH 7.5, containing 0.25 mol/L sucrose, 2 mmol/L phenylmethylsulfonyl fluoride, and 5 mmol/L N-ethylmaleimide. The whole-cell homogenates were centrifuged at 3000 x g for 10 minutes to remove nuclei, and the cell organelleenriched fractions in the supernatants were pelleted by centrifugation at 100,000 x g for 90 minutes. The pellets were resuspended in 50 mmol/L Tris/HCI, pH

248

lmai et al

AJPJuly 1997, Vol. 151, No. 1

7.5, 0.15 mol/L NaCI, 10 mmol/L CaCI2, 0.02% NaN3, layered onto a discontinuous sucrose density gradient (20, 30, 50, and 60% sucrose), and centrifuged at 100,000 x g for 3 hours. The plasma-membraneenriched fractions at the 30 to 50% sucrose interface were collected and pelleted again. The pellets were dissolved in the same buffer and protein concentrations were determined by reading absorbance at 595 nm using dye reagent (BioRad, Hercules, CA). All of the procedures after homogenization were carried out at 40C. As a control, plasma membranes were prepared in the same way from MDA-MB-231 cells from a human breast carcinoma cell line that is known to express MT1-MMP.23

Immunoblotting The plasma-membrane-enriched fractions prepared from the cultured chondrocytes (three OA cases and one normal case) were subjected to SDS-polyacrylamide gel electrophoresis (8.5% total acrylamide) under reducing conditions. Proteins separated in the gels were electrophoretically transferred onto nitrocellulose filters. The filters were then incubated for -12 hours at 230C with a monoclonal antibody specific to MT1-MMP (8 ,ug/ml; clone 114-1F2)13 and reacted with biotinylated horse IgG to mouse IgG, and the color was developed with 3,3'-diaminobenzidine tetrahydrochloride as described previously.16

Pro-MMP-2 Activation by Chondrocyte Plasma Membranes Pro-MMP-2 (2.6 ng) purified from human rheumatoid synovial cells8 was incubated with the plasma membranes (200 ng) from cultured chondrocyte (three OA cases and one normal case), those isolated directly from the articular cartilages (four OA and two normal cases), or MDA-MB-231 cell membranes in 20 ,ul of 50 mmol/L Tris/HCI, pH 7.5, 0.15 mol/L NaCI, 10 mmol/L CaCI2, and 0.02% NaN3 for 8 to 10 hours at 370C. The samples were subjected to gelatin zymography (8.5% total acrylamide), and the gels stained with Coomassie brilliant blue R-250.

Statistical Analyses Mann-Whitney U test was used to compare the data of OA and normal samples. Gelatinolytic activities in culture media with or without activation by APMA were compared using Wilcoxon single-rank test. Simple linear regression or Spearman rank correla-

tion was used for analyses of relationships between different parameters recorded in this study.

Results Immunolocalization and mRNA Expression of MT1-MMP in Osteoarthritic Cartilage OA cartilage samples varied in their histological appearance, ranging from surface irregularities to marked fibrillation and fissuring. The distribution of the histological/histochemical scores ranged from 2 to 13 with a mean score (±SD) of 8.2 ± 3.1. Normal articular cartilage had little or no microscopic changes with grades 0 to 1. By immunohistochemistry, the localization of MT1MMP was observed in all of the OA samples examined (64/64 cases). MT1 -MMP mainly localized to the chondrocytes in the superficial and transitional zones, where proteoglycans were depleted from the matrix (Figure 1 B). The chondrocytes located in the radial zone also stained for MT1-MMP when the cartilage had deep fissures reaching this zone (Figure 1C). Clustered chondrocytes close to the fissures were also labeled with the antibody (Figure 1 C). The immunostaining was considered to be specific for MT1-MMP as the staining was abolished with the monoclonal antibody absorbed with recombinant MT1 -MMP (Figure 1 D) and no staining was observed with nonimmune mouse IgG (data not shown). Approximately 50% of the total chondrocytes on average (48.1 ± 28.6%) immunostained positively in OA cartilage samples. In contrast, MT1-MMP staining was found in 27% of the normal cartilage samples (4/15 cases) but only in a few chondrocytes (5.2 ± 3.4%) in the superficial zone (Figure 1A). The percentage of MT1-MMP-positive chondrocytes was significantly higher in the OA samples than in the normal cartilage (P < 0.01). A linear correlation was found between the percentage of immunostained chondrocytes and Mankin score (p = 0.847; Figure 2). Immunolocalization of MMP-2 co-localized with the MT1-MMP-positive chondrocytes in almost all of the OA samples examined (K. lmai and Y. Okada, manuscript in preparation). In normal cartilage, chondrocytes in the superficial zone also reacted with the anti-MMP-2 monoclonal antibody. Expression of MT1-MMP mRNA in cartilage was examined by RT-PCR and in situ hybridization. RTPCR using a specific primer pair amplified a single 177-bp product in all of the OA cartilage samples examined (seven of seven cases), but the products were detected in only one of the four normal control

MT1-MMP Expression in OA Cartilage

249

AJPJuly 1997, Vol. 151, No. 1

S 8..

,

S

..;WMBI' 4' ,> .2.;..os s2

Figure 1. Immunolocalization ofMT1-MMP in the normal and OA cartilages. Immunostaining was performed as described in Materials and Methods. A: Normal cartilage, Mankin grade 0. No staining is seen. B: OA cartilage, Mankin grade 4. Chondrocytes in the superficial and transitional zones stain (arrows). C: OA cartilage, Mankin grade 8. Chondrocytes in the transitional and radial zones strongly stain. Arrows indicate immunoreactive chondrocytes in clusters. D: OA cartilage (Mankin grade 5) stained with the monoclonal antibody absorbed with recombinant MT1-MMP. Bar, 100

p.m.

Imai et al

250

AJPJuly 1997, Vol. 151, No.

I

1 nn I IOU'

-

I-,

I

o

.10480,

*5: 60

9

co 0

28 o

0

0

co

1

40

.E3 20 O0

2

4

6 10 8 Mankin score

12

14

I

Figure 2. Correlation ofMT1-MMP immunostaining with histological grade. A positive, direct correlation between MT1-MMP immunoreactivity and Mankin score is found by Spearman rank correlation (p 0.849; P = 0.0001; n = 79).

1

2 3 4 5 6 7 8 9 10 11

MTI-MMP

GAPDHD Figure 3. Determination of MT1-MMP mRNA expression in OA and normal cartilage by RT-PCR. Total RNA was extracted from OA (lanes 1 to 7) and normal (lanes 8 to 11 ) cartilage and reverse-transcribed into cDNA followed by a PCR reaction using specific primer pairs for MT1 MMP or GAPDH.

cartilage samples (Figure 3). The products were considered not to be derived from DNA contamination, as the template RNA was obtained from total RNA treated with RNAse-free DNAse and amplification did not occur when total RNA without the RT reaction was used for the PCR templates (data not shown). Primers specific to GAPDH generated a specific single amplicon of 598 bp without additional bands (Figure 3). In situ hybridization was carried out to identify the cells expressing MT1-MMP mRNA in the articular cartilage. Chondrocytes in the superficial and transitional zone of the OA cartilage were labeled with the antisense RNA (Figure 4A), whereas specimens from normal cartilage were not labeled (data not shown). The sense probe gave only a background signal in the chondrocytes of OA and normal cartilage (Figure 4B).

Production and Activation of Pro-MMP-2 in the Cartilage Culture media of the explants from 18 OA and 7 normal cartilage samples was subjected to gelatin zymography. Although the secretion of pro-MMP-2 of Mr 68,000 was observed in most of the samples examined, the intensity of the lytic band in the gels

Figure 4. In situ hybridization qfMTI-MMP in OA cartilage. Paraffin sections u'ere reacted uith -IS-labe(levd antisense (A) or .sense (B) riboprobes as described in Materials and Metbods. Note intense labeling in the chondrocytes in the s.uperficial zone of the articular cartilag,e uwith the antisenseprobe (arrows) and not with the sense probe. Bar, 25 ,um.

MT1-MMP Expression in OA Cartilage

251

AJPJuly 1997, Vol. 151, No. 1

B

A 1 2

.4 1220

I

*

I

-I

I

;a

kDa

a _S9

I

O

0 0 O

cl 'aI

947768-

.5

._

a _o

1o

c-

c9 ,1

.0.C3 v

NOR

43-

OA

Figure 6. Quantity ofpro-MMP-2 (A) and gelatinolytic activity (B) in the culture media secreted from cartilage explants. A: Culture media were obtainedfrom explants of normal (NOR; n = 7) and OA (n = 18) cartilages. Pro-MMP-2 was measured by a sandwich enzyme immunoassayforMMP-2. B: Gelatinolytic activity in the media from normal (n = 7) and OA (n = 18) cartilage was determined using "4C-gelatin solution assays in the absence (-) or presence (+) of APMA. Bars indicate mean + SD. 4P < 0.01; *¶ < 0.05 (Mann-Whitney U test).

29-

B 100

0 ._ P-

60

u

0

>? 40-

00

0

0

0

)20

o

0

0

0

o~~~~~~~~~ 60 80 20 40

100

NIT1 inmmul o reactivities (%) Figure 5. Gelatin zymography of the culture media from cartilage explants (A) and correlation between pro-MMP-2 activation and MT1MMP immunoreactivity (B). A: Culture media were subjected to gelatin zymography as described in Materials and Methods. Seven representative OA samples (lanes 1 to 7) andfour normal cartilage samples (lanes 8 to 11 ) are shown. The major gelatinolytic activity of 68 and 62 kd corresponds to pro-MMP-2 and active MMP-2, respectively. ProMMP-9 with 92 kd is also observed in two OA samples (lanes 2 and 6). Protein standards include phosphorylase b (94 kd), human transferrin (77 kd), BSA (68 kd), IgG heavy chain (55 kd), ovalbumin (43 kd), and carbonic anhydrase (29 kd). B: A linear increase in the activation ratio of pro-MMP-2 with MT1-MMP immunoreactivity is found by simple linear regression (r2 = 0.680; n = 18).

appeared to be higher in the OA samples than in the normal cartilage samples. Importantly, the active MMP-2 species of Mr 62,000 was detected in all of the samples from OA cartilage, whereas normal samples produced a negligible amount of active MMP-2 (Figure 5A). In fact, the activation ratio of pro-MMP-2 (the ratio of the active MMP-2 to proMMP-2 and active forms) was significantly higher in the OA cartilage samples (41 ± 16%) than in the normal samples (13 ± 8%; P < 0.01). The proportion of the activation ratio demonstrated a linear dependence on the MT1-MMP immunoreactivities of chondrocytes (r2 = 0.680; Figure 5B). There was also a linear correlation between the activation ratio and the

Mankin scores (p = 0.664; data not shown). ProMMP-9 of Mr 92,000 and the processed form of Mr 83,000 were detectable in 28% of the OA samples (5/18 samples) but not in the normal (0/7 samples). As gelatin zymography demonstrated enhanced production and activation of pro-MMP-2 in the OA cartilage, we further quantified the amounts of proMMP-2 protein secreted in the culture media by a sandwich enzyme immunoassay for pro-MMP-2 and gelatinolytic activities in a solution assay using 14C_ gelatin. As shown in Figure 6A, the amount of proMMP-2 in the media from OA cartilage (10 + 7 ng/mg tissue dry weight) was significantly higher than that from normal cartilage (2 ± 1 ng/mg tissue dry weight). Gelatinolytic activities assayed in the absence of APMA were also remarkably higher in the OA samples (38 + 30 U/g tissue dry weight) than in the normal samples (3 ± 4 U/g tissue dry weight; Figure 6B). The presence of APMA in the assay significantly increased the activity in the normal cartilage samples (7 + 3 U/g dry weight; P < 0.05) without a definite change of activity observed in the OA samples, although the level (42 ± 26 U/g dry weight) was remarkably higher than that of the APMA-treated normal samples (Figure 6B). These data are consistent with the finding of the gelatin zymography demonstrating that the OA cartilage has an enhanced production of MMP-2 with a large proportion of the enzyme activated, in contrast to the limited production and activation in normal cartilage. An increase in the pro-MMP-2 activation ratio did not occur with pro-MMP-2 secretion (data not shown), suggesting that the activation and the production of pro-MMP-2 are two independent events.

Imai et al

252

AJPJuly 1997, Vol. 151, No. 1

I

kDa 94-

2

3

1

C

B

A

1)

4 5

kDa

kDa

94-

94-

I

2

A

3 4 MT1-MMP

77-

68-

7-7 68-

55-

18S

68-

rRNA

55-

55-

43-

43-

43-

B_

~~~29 Figure 7. Expression ofMT1-MMP and pro-MMP-2 activation by chondrocyte membranes. A: Cell membranes (3.2 gg) prepared from MDAMB-231 cells (lane 1), cultured normal chondrocytes (lane 2), and OA chondrocytes (lanes 3 to 5) were subjected to immunoblotting for MT1-MMP as described in Materials and Methods. B: Pro-MMP-2 was incubated for 8 hours at 3 70C with buffer alone (lane 1) or cell membranes prepared from MDA-MB-231 cells (lane 2), cultured nor-

mal chondrocytes (lane 3), and OA chondrocytes (lanes 4 to 6), and the generation of active MMP-2 utas demonstrated by gelatin zymography (8.5% total acrylamide). C: Pro-MMP-2 activation with cell membranes directly isolatedcfrom normal (lane 2) and OA (lanes 3 and 4) cartilage u'as examined by a 10-houir incubation at 37°C as described above. Lane 1 is pro-MMP-2 incuibated utith buffer alone. Note that cell membranes from both normal and OA chondrocytes in cuilture contain MT1-MMP and activate pro-MMP-2, whereas activation is seen only in chondrocyte membranes isolated directly from OA cartilages.

Activation of Pro-MMP-2 by Chondrocyte Membranes Immunoblotting analyses demonstrated that, under the culture conditions, chondrocytes from one normal and three OA cartilage samples expressed MT1MMP protein of Mr 60,000, which corresponds to activated MT1-MMP on the cell membranes23 (Figure 7A). RT-PCR also demonstrated that MT1-MMP mRNA is unambiguously expressed in those chondrocytes (data not shown). When cell membrane fractions were prepared from these cultured chondrocytes and incubated with pro-MMP-2 of Mr 68,000, the zymogen was processed into the active species of Mr 62,000 (Figure 7B). Similar pro-MMP-2 activation was obtained with plasma-membrane-enriched fractions of the chondrocytes directly isolated from OA cartilage with MT1-MMP expression, but such activation was not observed with the chondrocyte membranes from the normal cartilage (Figure 7C).

Effects of IL- 1 a and/or TNF-a on MT1 -MMP Gene Expression and Pro-MMP-2 Activation in Chondrocytes As IL-la and TNF-a are commonly present in OA cartilage, their effects on MT1-MMP mRNA expression were examined by Northern blot analysis. Although chondrocytes without cytokine treatment ex-

Figure 8. M71-MMP mRNA expression (A) andpro-MMP-2 activation (B) in cultuired OA chondrocytes treated with IL-la and/or TNF-a. A: Chondrocytes in culture were treated with serum-free medium alone (lane 1), 100 Ulml IL-la (lane 2), 500 Ulml TNF-a (lane 3), or 100 Ulml IL-la and 500 Ulml TNF-a (lane 4) as described in Matenials and Methods. Total RNA ( 15 jig) from eacb sample separated on a formamidelformaldehyde agarose gel was blotted onto nylon membranes that were probed with MT1-MMP cDNA. The 18 S rRNA was stained with ethidium bromide for comparison of the total RNA loaded. B: Culture media (30 gIl/lane) were harvested from the chondrocytes treated as described above and subjected to gelatin zymography (8% total acrylamide). Small and large arrows indicate pro-MMP-2 and active MMP-2, respectively. The arrowhead shows pro-MMP-9.

pressed a low level of MTl-MMP mRNA, both IL-la and TNF-a tremendously enhanced the expression of MTl-MMP mRNA (Figure 8A). The stimulatory effect of IL-la was 2.5- to 8-fold higher than TNF-a. In addition, the stimulation with both IL-la and TNF-a was not synergistic. Culture media were harvested from the chondrocytes and subjected to gelatin zymography. There was no definite difference in the secreted pro-MMP-2 levels among the chondrocyte cultures with or without the cytokines. However, it is notable that activation of pro-MMP-2 was enhanced by IL-la and TNF-a, the latter of which also upregulated the pro-MMP-9 production as shown with the HT1080 human fibrosarcoma cell line.24

Discussion Our studies demonstrate for the first time that MT1MMP is highly expressed in the chondrocytes of OA articular cartilage. Immunolocalization studies demonstrated that approximately 50% of the chondrocytes located in the superficial and transitional zones stained for MTl-MMP in all of the OA samples examined with only a few chondrocytes in the superficial zone of normal cartilage positive for MTl-MMP. RT-PCR confirmed the predominant MTl-MMP expression in OA cartilage. Also, in situ hybridization with RNA probes demonstrated that

MT1-MMP Expression in OA Cartilage

253 AJP July 1997, Vol. 151, No. 1

OA chondrocytes are the cells expressing MT1MMP mRNA. Previous studies have reported that OA chondrocytes express various MMPs including MMP-1, MMP-2, MMP-3, MMP-8 (neutrophil collagenase), MMP-9, and MMP-13 (collagenase 3).32i27 The present results add MT1-MMP to the list of MMPs produced in the OA cartilage. The previous studies by Mohtai et al11 reported that expression of both MMP-2 and MMP-9 is enhanced in OA cartilage, although they emphasized the role of MMP-9 in the progressive cartilage degradation in OA. In the present studies, increased production of pro-MMP-2 from the OA cartilage explants was ascertained by gelatin zymography. A sandwich enzyme immunoassay specific to proMMP-2 also indicated that the production level of MMP-2 is approximately fourfold higher in OA cartilage samples compared with normal cartilage. Furthermore, we showed that activation of pro-MMP-2 in the cartilage is accelerated by the OA changes, demonstrating a linear correlation between the activation ratio and Mankin score. Recent molecular cloning and experimental studies of MT1-MMP have contributed to the understanding of the activation mechanism of pro-MMP-2. The activation is considered to be initiated by direct cleavage of the Asn37Leu38 bond of pro-MMP-2 by MT1-MMP and completed by intermolecular autocatalysis of MMP2.28 30 It is also established that, once MT1-MMP is synthesized within the cells, it is expressed in the active form of Mr 60,000 on the cell surface after being processed by furin.2331 Involvement of MTlMMP in the in vivo activation of pro-MMP-2 has been described in various human cancer tissues such as stomach and lung carcinoma.16'32 The present studies demonstrated a linear correlation between the pro-MMP-2 activation and MT1-MMP immunostaining in the OA cartilages. In addition, the chondrocyte membranes isolated directly from the OA cartilage activated pro-MMP-2, whereas no such activation was found in the normal cartilage membrane. Under the present culture conditions, the normal chondrocytes also expressed MT1-MMP. This is not surprising as chondrocytes cultivated on the flasks are modulated to change their phenotype,33 resulting in the MT1-MMP expression. Thus, these data suggest that MT1-MMP expression in the chondrocytes may be involved in the activation of pro-MMP-2 in OA cartilage. In the OA cartilage, degenerative changes start from the superficial layer and progress into the deeper zones.34'35 Previous immunolocalization studies on MMP-336 and MMP-837 demonstrated that the expression levels correlate with the severity of

OA changes in the cartilage. An identical correlation is seen in this study with MTl-MMP expression in the cartilage. As gene expression of many MMPs such as MMP-1 and MMP-3 is regulated by various cytokines and growth factors,4 enhanced MT1-MMP expression in the OA cartilage may be due to stimulation by such factors. MTl-MMP mRNA expression is known to be up-regulated by concanavalin A, phorbol myristate acetate, basic fibroblast growth factor, and TNF-a and down-regulated by glucocorticoids in various cells.28'38-41 However, the effects of these cytokines and growth factors on the expression of MMPs depend on the cell type used for the experiments, and information is not available on factors modulating MTl-MMP expression in human chondrocytes. Lohi et al40 have recently reported that MTl-MMP mRNA expression is enhanced by TNF-a in HT1080 fibrosarcoma cells but not in embryonic lung fibroblasts, and IL-1,B has no effect on the expression of MTl-MMP in these cells. In the present study, we demonstrate that both TNF-a and IL-la stimulate OA chondrocytes to express the MTl-MMP gene. Gelatin zymography further showed that MT1MMP up-regulated by the cytokines functions to activate pro-MMP-2. As both cytokines are major mediators of joint damage in OA cartilage,42 it seems likely that MTl-MMP expression in OA cartilage is controlled by these cytokines, which also up-regulate expression of MMP-1 and MMP-3. The regulation of MMP-2 gene expression is different from other MMPs in that MMP-2 is not inducible by many cytokines and growth factors such as IL-la and TNF-a 3 but is expressed constitutively in many cells in culture. In both normal and OA cartilage, MMP-2 was immunolocalized in many more chondrocytes than MTl-MMP (K. Imai and Y. Okada, manuscript in preparation). Our preliminary studies also show that in OA cartilage chondrocytes express MT3-MMP, but the expression is found only in the advanced disease cartilage and the mRNA is quite low (K. Imai and Y. Okada, unpublished data). It might be possible to speculate that MT3-MMP also accelerates the cartilage degradation through pro-MMP-2 activation in advanced OA cartilage but plays a minor role when compared with MT1-MMP. Expression of MT2MMP was not detectable in the OA cartilage examined (K. Imai and Y. Okada, unpublished data). Thus, the data suggest that MTl-MMP expression is a key determinant for MMP-2 activity in the cartilage through activation of the zymogen by MTl-MMP. Hollander et al34 showed by immunohistochemistry that type 11 collagen breakdown in OA cartilage is first observed around the chondrocytes. Similar degradation of type 11 collagen on the cell surface is also

254

Imai et al

AJPJuly 1997, Vol. 151, No. 1

described when the normal articular cartilage is cultured in the presence of IL-1.4344 In addition to collagen degradation, aggrecan breakdown in cartilage is also thought to be a cell-membrane-dependent event.45 These data suggest that pericellular ECM breakdown is a key process that is occurring in OA cartilage. Recent experimental studies demonstrate that when MT1-MMP is expressed on the surface of cells pro-MMP-2 is captured and activated on the cell surface.29 As MT1-MMP and MMP-2 co-localized in the OA chondrocytes, MMP-2 activated by MT1-MMP may attack various ECM macromolecules such as aggrecan around the chondrocytes. In addition, recent biochemical studies on MMP-13 (collagenase 3) suggests that MT1-MMP also activates pro-MMP-13.46 As MMP-13 is expressed in OA cartilage and readily digests type 11 collagen,25'27 intermolecular activation of pro-MMP-13 by MT1-MMP may be important in the ECM breakdown in OA cartilage. On the other hand, it has been reported that MT1-MMP per se is an ECM-degrading enzyme that digests laminin, fibronectin, vitronectin, dermatan sulfate proteoglycan, and gelatin.31 We have also recently demonstrated that MT1-MMP is active against aggrecan and interstitial collagens, including type II collagen, cleaving them at the triple helical region similar to the cleavage by interstitial collagenases.15 Aggrecan fragments released from the OA cartilage matrix demonstrate the special NH2 terminus beginning at Ala374, which is thought to be generated by the action of so-called aggrecanase.47 It has been reported that the aggrecanase activity is inducible by IL-l a48 and inhibited by metalloproteinase inhibitors.49 Among the MMPs, only MMP-8 at high concentrations cleaves the Glu373-Ala374 bond of aggrecan, 5 but its expression level in the OA cartilage is low.25 Therefore, the possibility that MMP-8 is an aggrecanase is still speculative. On the other hand, as MT1-MMP is expressed on the cell surfaces and the gene expression is stimulated by treatment of the chondrocytes with IL-la, it is plausible to speculate that MT1-MMP may be responsible for the so-called aggrecanase activity.

Acknowledgments We are grateful to Dr. G. 1. Goldberg (Division of Dermatology, Washington University, School of Medicine, St. Louis, MO), a visiting professor at Kanazawa University, for helpful discussion and to Dr. J. D'Armiento (Department of Medicine, Columbia University, New York, NY) for critical reading of the manuscript. We also thank M. Takegami and S. Makino for their technical assistance.

References 1. Felson DT: Epidemiology of hip and knee osteoarthritis. Epidemiol Rev 1988, 10:1-28 2. Houg AJ, Sokoloff L: Pathology of osteoarthritis. Arthritis and Allied Conditions. Edited by DJ McCarty. Philadelphia, Lea & Febiger, 1989, pp 1571-1594 3. Nagase H, Woessner JF Jr: Role of endogenous proteinases in the degradation of cartilage matrix. Joint Cartilage Degradation. Edited by JF Woessner Jr, DS Howell. New York, Marcel Dekker, 1993, pp 159-185 4. Nagase H, Okada Y: Proteinases and matrix degradation. Textbook of Rheumatology. Edited by WN Kelly, ED Harris Jr, S Ruddy, CB Sledge. Philadelphia, Saunders Co, 1997, pp 323-341 5. Fosang AJ, Neame PJ, Last K, Hardingham TE, Murphy G, Hamilton JA: The interglobular domain of cartilage aggrecan is cleaved by PUMP, gelatinases and cathepsin B. J Biol Chem 1992, 270:19470-19474 6. Nguyen Q, Murphy G, Hughes C, Mort JS, Roughley PJ: Matrix metalloproteinases cleave at two distinct sites on human cartilage link protein. Biochem J 1993, 295:595-598 7. Imai K, Hiramatsu A, Fukushima D, Pierschbacher MD, Okada Y: Degradation of decorin by matrix metalloproteinases: identification of the cleavage sites, kinetic analyses and transforming growth factor-,1 release. Biochem J 1997, 322:809-814 8. Okada Y, Morodomi T, Enghild JJ, Suzuki K, Yasui A, Nakanishi I, Salvesen G, Nagase H: Matrix metalloproteinase 2 from human rheumatoid synovial fibroblasts: purification and activation of the precursor and enzymic properties. Eur J Biochem 1990, 194:721-730 9. Welgus HG, Fliszar CJ, Seltzer JL, Schmid TM, Jeffrey JJ: Differential susceptibility of type X collagen to cleavage by two mammalian interstitial collagenases and 72-kDa type IV collagenase. J Biol Chem 1990, 265:13521-13527 10. Murphy G, Cawston TE, Galloway WA, Barnes MJ, Bunnig RAD, Mercer E, Reynolds JJ, Burgeson RE: Metalloproteinases from rabbit bone culture medium degrade types IV and V collagens, laminin and fibronectin. Biochem J 1981, 199:807-811 11. Mohtai M, Smith RL, Schurman DJ, Tsuji Y, Torti FT, Hutchinson NI, Stetler-Stevenson WG, Goldberg GI: Expression of 92-kD type IV collagenase/gelatinase (gelatinase B) in osteoarthritic cartilage and its induction in normal human articular cartilage by interleukin 1. J Clin Invest 1993, 92:179-185 12. Ward RV, Atkinson SJ, Reynolds JJ, Murphy G: Cell surface activation of progelatinase A: demonstration of the involvement of the C-terminal domain of progelatinase A in cell surface binding and activation of progelatinase A by primary fibroblasts. Biochem J 1994, 304:263-269 13. Sato H, Takino T, Okada Y, Cao J, Shinagawa A, Yamamoto E, Seiki M: A matrix metalloproteinase ex-

MT1-MMP Expression in OA Cartilage

255

AJP July 1997, Vol. 151, No. 1

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

pressed on the surface of invasive tumour cells. Nature 1994, 370:61-65 Mankin HJ, Dorfman H, Lippiello L, Zarins A: Biochemical and metabolic abnormalities in human articular cartilage from osteoarthritic human hips. ll. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971, 53A:523-537 Ohuchi E, Imai K, Fujii Y, Sato H, Seiki M, Okada Y: Membrane-type 1-matrix metalloproteinase digests interstitial collagens and other extracellular matrix macromolecules. J Biol Chem 1997, 272:2446-2451 Nomura H, Sato H, Seiki M, Mai M, Okada Y: Expression of membrane-type matrix metalloproteinase in human gastric carcinomas. Cancer Res 1995, 55:32633266 Fujimoto N, Mouri N, Iwata K, Ohuchi E, Okada Y, Hayakawa T: A one-step sandwich enzyme immunoassay for human matrix metalloproteinase 2 (72-kDa gelatinase/type IV collagenase) using monoclonal antibodies. Clin Chim Acta 1993, 221:91-103 Okada Y, Naka K, Kawamura K, Matsumoto T, Nakanishi 1, Fujimoto N, Sato H, Seiki M: Localization of matrix metalloproteinase 9 (92-kilodalton gelatinase/type IV collagenase = gelatinase B) in osteoclasts: implications for bone resorption. Lab Invest 1995, 72:311-322 Davies B, Miles DW, Happerfield LC, Naylor MS, Bobrow LG, Rubens RD, Balkwill FR: Activity of type IV collagenase in benign and malignant breast disease. Br J Cancer 1993, 67:1126-1131 Aydelotte MB, Kuettner KE: Differences between subpopulations of cultured bovine articular chondrocytes. I. Morphology and cartilage matrix production. Connect Tissue Res 1988, 18:205-222 Chomczynski P, Sacchi N: Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987, 162:156-159 Sato H, Kida Y, Mai M, Endo Y, Sasaki T, Tanaka J, Seiki M: Expression of genes encoding type IV collagen-degrading metalloproteinase and tissue inhibitors of metalloproteinases in various human tumor cells. Oncogene 1992, 7:77-83 Imai K, Ohuchi E, Aoki T, Nomura H, Fujii Y, Sato H, Seiki M, Okada Y: Membrane-type matrix metalloproteinase 1 is a gelatinolytic enzyme and secreted in a complex with tissue inhibitor of metalloproteinases 2. Cancer Res 1996, 56:2707-2710 Okada Y, Tsuchiya H, Shimizu H, Tomita K, Nakanishi I, Sato H, Seiki M, Yamashita K, Hayakawa T: Induction and stimulation of 92-kDa gelatinase/type IV collagenase production in osteosarcoma and fibrosarcoma cell lines by tumor necrosis factor a. Biochem Biophys Res Commun 1990, 171 :610-617 Cole AA, Chubinskaya S, Schumacher B, Huch BK, Cs-Szabo G, Yao J, Mikecz K, Hasty KA, Kuettner KE: Chondrocyte matrix metalloproteinase-8: human articular chondrocytes express neutrophil collagenase. J Biol Chem 1996, 271:11023-11026 Mitchell PG, Magna HA, Reeves LM, Lopresti-Morow

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

LL, Yocum SA, Rosner PJ, Geoghegan KF, Hambor JE: Cloning, expression, and type 11 collagenolytic activity of matrix metalloproteinase-13 from human osteoarthritic cartilage. J Clin Invest 1996, 97:761-768 Reboul P, Pelletier JP, Tardif G, Cloutier JM, MartelPelletier J: The new collagenase, collagenase-3, is expressed and synthesized by human chondrocytes but not by synoviocytes. J Clin Invest 1996, 97:2011-2019 Atkinson SJ, Crabbe T, Cowell S, Ward RV, Butler MJ, Sato H, Seiki M, Reynolds JJ, Murphy G: Intermolecular autolytic cleavage can contribute to the activation of progelatinase A by cell membrane. J Biol Chem 1995, 270:30479-30485 Sato H, Takino T, Kinoshita T, Imai K, Okada Y, StetlerStevenson WG, Seiki M: Cell surface binding and activation of gelatinase A induced by expression of membrane-type-i-matrix metalloproteinase (MT1-MMP). FEBS Lett 1996, 385:238-240 Kinoshita T, Sato H, Takino T, Itoh M, Akizawa T, Seiki M: Processing and a precursor of 72-kilodalton type IV collagenase/gelatinase A by a recombinant membrane-type 1 matrix metalloproteinase. Cancer Res 1996, 56:2535-2538 Pei D, Weiss SJ: Transmembrane-deletion mutants of the membrane-type matrix metalloproteinase-1 process progelatinase A and express intrinsic matrix-degrading activity. J Biol Chem 1996, 271:9135-9140 Tokuraku M, Sato H, Murakami S, Okada Y, Watanabe Y, Seiki M: Activation of the precursor of gelatinase A/72 kDa collagenase/MMP-2 in lung carcinomas correlates with the expression of membrane-type matrix metalloproteinase (MT-MMP) and with lymph node metastasis. Int J Cancer 1995, 64:355-359 Hauselman HJ, Fernandes J, Mok SS, Schmid TM, Block JA, Aydelotte MB, Kuettner KE, Thonar EJA: Phenotypic stability of bovine articular chondrocytes after long-term culture in alginate beads. J Cell Sci 1994, 107:17-27 Hollander AP, Heathfield TF, Webber C, Iwata Y, Bourne R, Rorabeck C, Poole AR: Increased damage to type 11 collagen in osteoarthritic articular cartilage detected by a new immunoassay. J Clin Invest 1994, 93:1722-1732 Hollander AP, Pidoux I, Reiner A, Roranbeck C, Bourne R, Poole AR: Damage of type 11 collagen in aging and osteoarthritis starts at the articular surface, originates around chondrocytes, and extends into the cartilage with progressive degeneration. J Clin Invest 1995, 96: 2859-2869 Okada Y, Shinmei M, Tanaka 0, Naka K, Kimura A, Nakanishi I, Bayliss MT, Iwata K, Nagase H: Localization of matrix metalloproteinase 3 (stromelysin) in osteoarthritic cartilage and synovium. Lab Invest 1992,

66:680-690 37. Chubinskaya S, Huch K, Mikecz K, Cs-Szabo G, Hasty KA, Kuettner KE, Cole AA: Chondrocyte matrix metalloproteinase-8: up-regulation of neutrophil colla-

256 Imai et al AJPJuly 1997, Vol. 151, No. 1

38.

39.

40.

41.

42.

43.

genase by interleukin-1 3 in human cartilage from knee and ankle joints. Lab Invest 1996, 74:232-240 Lewalle JM, Munaut C, Pichot B, Cataldo D, Baramova E, Foidart JM: Plasma membrane-dependent activation of gelatinase A in human vascular endothelial cells. J Cell Physiol 1995, 165:475-483 Lohi J, Keski-Oja J: Calcium ionophores decrease pericellular gelatinolytic activity via inhibition of 92-kDa gelatinase expression and decrease of 72-kDa gelatinase activation. J Biol Chem 1995, 270:17602-17609 Lohi J, Lehti K, Westermarck J, Kahari VM, Keski-Oja J: Regulation of membrane-type matrix metalloproteinase-1 expression by growth factors and phorbol 12myristate 13-acetate. Eur J Biochem 1996, 239:239247 Okada A, Bellocq JP, Rouyer N, Chenard MP, Rio MC, Chambon P, Basset P: Membrane-type matrix metalloproteinase (MT-MMP) gene is expressed in stromal cells of human colon, breast, and head and neck carcinomas. Proc Natl Acad Sci USA 1995, 92:2730-2743 Westacott Cl, Sharif M: Cytokines in osteoarthritis: mediators or markers of joint destruction. Semin Arthritis Rheum 1996, 25:254-272 Dodge GR, Poole AR: Immunohistochemical detection and immunohistochemical analysis of type 11 collagen degradation in human normal, rheumatoid, and osteoarthritic articular cartilages and in explants of bovine articular cartilage cultured with interleukin 1. J Clin Invest 1989, 83:647-661

44. Mort JS, Dodge GR, Roughley PJ, Liu J, Dipasquale G, Poole AR: Direct evidence for active metalloproteinases mediating matrix degradation in interleukin 1-stimulated human articular cartilage. Matrix 1993, 13:95-102 45. Lark MW, Gordy JT, Weidner JR, Ayala J, Kimura JH, Williams HR, Mumford RA, Flannery CR, Carlson SS, Iwata M, Sandy JD: Cell-mediated catabolism of aggrecan: evidence that cleavage at the "aggrecanase" site (Glu373-Ala374) is a primary event in proteolysis of the interglobular domain. J Biol Chem 1995, 270:2550-2556 46. Knauper V, Will H, L6pez-Otfn C, Smith B, Atkinson SJ, Stanton H, Hembry RA, Murphy G: Cellular mechanism for human procollagenase-3 (MMP-13) activation. J Biol Chem 1996, 271:17124-17131 47. Sandy JD, Plaas AHK, Koob TJ: Pathways of aggrecan processing in joint tissues: implications for disease mechanism and monitoring. Acta Orthop Scand 1996, 66:26-32 48. Sandy JD, Neam PJ, Boynton RE, Flannery CR: Catabolism of aggrecan in cartilage explants: identification of a major cleavage site within the interglobular domain. J Biol Chem 1991, 266:8683-8685 49. Buttle DJ, Handley CJ, Llic MZ, Saklatvala J, Murata M, Barret AH: Inhibition of cartilage proteoglycan release by a specific inactivator of cathepsin B and an inhibitor of matrix metalloproteinases. Arthritis Rheum 1993, 36: 1709-1717