Cytokine-induced cartilage proteoglycan degradation is mediated by aggrecanase

Osteoarthritis and Cartilage (1998) 6, 214–228 7 1998 Osteoarthritis Research Society

1063–4584/98/030214 + 15 $12.00/0

Cytokine-induced cartilage proteoglycan degradation is mediated by aggrecanase By Elizabeth C. Arner*, Clare E. Hughes‡, Carl P. Decicco†, Bruce Caterson‡ and Micky D. Tortorella* *Inflammatory Diseases Research and †Medicinal Chemistry, The DuPont Merck Pharmaceutical Company, Experimental Station, Wilmington, Delaware 19880-0400, U.S.A.; and ‡Connective Tissue Biology Laboratory, School of Molecular and Medical Biosciences, University of Wales, Cardiff, CF1 3US, U.K. Summary Objective: To evaluate the relationship between specific cleavage of aggrecan at the Glu373–Ala374 ‘aggrecanase’ site and degradation and release of proteoglycan catabolites from cartilage in explant cultures. Design: The monoclonal antibody, BC-3, which specifically recognizes the new N-terminus, ARGSVIL, generated by cleavage of aggrecan at the Glu373–Ala374 ‘aggrecanase’ site, was used to follow the generation of fragments produced by cleavage at this site as compared to degradation of proteoglycan as assessed by glycosaminoglycan (GAG) release from cartilage in response to cytokines and the ability of inhibitors to block this cleavage. Results: (1) There was a strong correlation between specific cleavage at the Glu373–Ala374 bond and the release of aggrecan catabolites in response to interleukin-1 (IL-1) or tumour necrosis factor (TNF) stimulation. (2) This cleavage in the interglobular domain of aggrecan was inhibited by the inclusion of cycloheximide, thus indicating a requirement for de novo protein synthesis in the induction of ‘aggrecanase’ activity. (3) The inhibitors, indomethacin, naproxen, tenidap, dexamethasone and doxycycline were ineffective in blocking either specific cleavage at the ‘aggrecanase’ site or aggrecan degradation as measured by GAG release from cartilage. (4) In contrast, compounds which act through two different mechanisms to inhibit MMPs were effective in blocking both specific cleavage at the ‘aggrecanase’ site and proteoglycan degradation. Conclusions: Our data suggest that ‘aggrecanase’ is primarily responsible for proteoglycan cleavage in these experimental systems and that this protease has properties in common with metalloproteases including members of the MMP and ADAM family. Inhibition of ‘aggrecanase’ may have utility in preventing cartilage loss in arthritis. Key words: Cartilage, Proteoglycan degradation, Aggrecanase, Matrix metalloproteinase, Cytokines.

(MMPs), synthesized and secreted by connective tissue cells, have been shown to cleave aggrecan in this IGD region at a specific site between amino acid residues Asn 341–Phe342 [2, 3] and G1 fragments with this cleavage site have been identified within articular cartilage bound to hyaluronic acid [4]. However, C-terminal fragments with the new N-terminus, ARGSVIL . . ., formed by cleavage between the amino acid residues Glu373–Ala374 have been identified in synovial fluid of patients with osteoarthritis [5], inflammatory joint disease [6] and in the media from cartilage explant and chondroctye cultures stimulated with interleukin1 or retinoic acid [2, 7–10], suggesting that cleavage at this site may also play an important role in cartilage degradation. Many matrix metalloproteinases (MMP-1, -2, -3, -7, -8, -9 and -13) have been shown to cleave aggrecan in vitro at the Asn341–Phe342 site [3, 4, 11– 12]. In contrast, attempts to generate cleavage at the Glu373–Ala374 site with a number of purified

Introduction In arthritic conditions there is a loss of cartilage matrix components which in normal joints provide cartilage with its mechanical properties of reducing load during joint articulation. One of the earliest changes observed in cartilage morphology is depletion of proteoglycan (aggrecan) [1] which, due to its high negative charge and water-binding capacity, provides cartilage with its properties of compressibility and resilience. This loss appears to be due to an increased rate of aggrecan degradation which can be attributed to proteolytic cleavage within the interglobular domain (IGD) located between the G1 and G2 globular domains of aggrecan. Several matrix metalloproteinases Received 18 September 1997; accepted 20 Janaury 1998. Address correspondence to: Inflammatory Diseases Research, The DuPont Merck Pharmaceutical Company, Experimental Station E400/4239, Box 80400, Wilmington, DE 19880-0400, U.S.A. Tel: 302-695-7078; Fax: 302-695-7873.

214

Osteoarthritis and Cartilage Vol. 6 No. 3

proteinases have been less successful [3, 4, 13]. Therefore, this cleavage is thought to be the result of a novel, as yet unidentified proteinase, given the name ‘aggrecanase’ based on its ability to cleave the aggrecan core protein. However, recent in vitro studies, using a purified G1–G2 substrate or native bovine cartilage aggrecan substrate, demonstrated the ability of native and recombinant neutrophil collagenase (MMP-8) to cleave at the Glu373–Ala374 ‘aggrecanase’ site, although the preferential cleavage site was the Asn341–Phe342 bond [14, 15] and articular chondrocytes have been shown to be capable of synthesizing MMP-8 [16]. These data demonstrate that an MMP can cleave at this site under specific conditions and open the possibility that the cartilage ‘aggrecanase’ may be a matrix metalloproteinase. In the studies reported herein, we use an antibody, BC-3, which specifically recognizes the new N-terminus, ARGSVIL, generated by cleavage of aggrecan at the Glu373–Ala374 ‘aggrecanase’ site [17] to follow the generation of proteoglycan catabolites produced by cleavage at this site in response to cytokines in explant cultures of both bovine nasal and bovine articular cartilage cultures and also the ability of inhibitors to block this cleavage. These studies indicate that a variety of anti-arthritic compounds which are ineffective in blocking cleavage at the ‘aggrecanase’ site do not block proteoglycan release from the tissue in response to interleukin-1 (IL-1) stimulation of cartilage in culture. In contrast, we show that two compounds which broadly inhibit MMP activity by different mechanisms (an active-site MMP inhibitor and an inhibitor of activation of the MMP zymogens), block specific cleavage at the ‘aggrecanase’ site and the release of cartilage aggrecan catabolites. The data reported herein suggest that ‘aggrecanase’ is primarily responsible for proteoglycan cleavage in these experimental systems and that this protease may be activated by or have properties in common with metalloproteases including members of the MMP and ADAMD (a disintegrin and metalloprotease domain) family of proteases.

Materials and Methods materials Dulbecco’s modified Eagle’s medium (DMEM), fetal calf serum, penicillin/streptomycin/ amphotericin B, and neomycin were from Gibco (Grand Island, NY, U.S.A.). The IL-1 used was a soluble, fully-active recombinant human IL-1b produced as described previously [18]. The specific

215

biological activity was 1 × 107 units/mg of protein, with 1 unit being defined as the amount of IL-1 that generated half-maximal activity in the thymocyte proliferation assay. Keratanase, keratanase II and chondroitinase ABC were from Seikuguku (Kogyo, Japan). Monoclonal antibody BC-3 was prepared as ascitic fluid and was shown to be specific for the amino acid sequence ARGSVI . . . at the N-terminus of aggrecan fragments [17, 19]. Western blot detection used the Promega AP system from Promega (Madison, WI, U.S.A.). The hydroxamic acid active-site proMMP inhibitor, BB-16 ((2S, 3R)-2-methyl-3-(2-methylpropyl)-1-(N-hydroxy)-4(o-methyl)-L-tyrosine-N-methyl amide) and the isothiazolone inhibitor of proMMP activation, XG076 (7-aza-2-phenylbenzisothiazol-3-one), were synthesized at DuPont Merck as previously described [20, 21]. Indomethacin, naproxen, tenidap, doxycycline and dexamethasone were from Sigma (St. Louis, MO, U.S.A.). All other reagents used were of analytical grade. tissue preparation Bovine nasal septa were obtained fresh at the time of slaughter. Uniform cartilage discs (1 mm thick, 8 mm in diameter, 080 mg) were prepared under sterile conditions as described by Steinberg et al. [22]. Bovine articular cartilage was obtained from the metacarpophangeal joints of 2–3-year-old cows. Cartilage was allowed to equilibrate for at least 3 days in culture with DMEM supplemented with 5% heat-inactivated fetal calf serum, penicillin, streptomycin, amphotericin B, and neomycin (100 IU/ml, 100 mg/ml, 0.25 mg/ml and 50 mg/ml, respectively) prior to treatment as described below. organ culture Cartilage pieces were cultured as described previously [23]. For bovine nasal cartilage, discs were either weighed and incubated two per well in a 12-well culture dish with 1 ml media or were cut into eighths and each eighth of a cartilage disc was weighed, then placed into a well of a 96-well culture dish containing 180 ml of serum-free DMEM containing penicillin, streptomycin, amphotericin B, and neomycin (100 IU/ml, 100 mg/ml, 0.25 mg/ml and 50 mg/ml, respectively). Paired explants from the same disc were used to compare the effects of various inhibitor treatments. Bovine articular cartilage slices were cut into pieces of approximately 10 mg, weighed and cultured as described for nasal cartilage. Eight replicates per treatment group were run for each

216

Arner et al.: Aggrecanase-mediated cartilage degradation

experiment. Cultures were incubated at 37°C in an atmosphere of 95% air/5% CO2.

inhibitor studies Compounds were dissolved in DMSO (10 − 2 m) and further diluted with DMEM, supplemented with antibiotics, to the required concentrations. DMSO concentrations in the culture media never exceeded 1%; this concentration of DMSO had no effect on alterations in cartilage proteoglycan metabolism in response to IL-1. Cartilage was incubated in the absence or presence of IL-1 (10–50 ng/ml) with or without inhibitor for 48 h. For timecourse studies, cultures were incubated in the absence or presence of IL-1 with or without inhibitor for time periods up to 72 h. For cycloheximide studies, cultures were incubated in control media or in IL-1 (50 ng/ml) with and without cycloheximide (5 mg/ml). At the end of the incubation, media and cartilage slices were harvested and frozen for further analysis. For proteoglycan synthesis studies, at the end of the incubation, the media were replaced with Ham’s F-12 media, containing 20 mCi/ml of [35S]-sulfate and cultures were incubated for a 2-h labeling period and assayed for incorporation as previously described [23]. Briefly, the media were then removed, the cartilage was digested overnight with papain, the proteoglycan in the digest precipitated with cetylpyridinium chloride, and the precipitates counted for 35S on a Packard Matrix 96 counter. [35S]sulfate incorporation was determined as disintegrations per minute (dpm) per milligram of wet weight cartilage.

extraction and purification of aggrecan and aggrecan catabolites Bovine nasal cartilage slices were sectioned (0.2 mm) and then the proteoglycans were extracted for 48–72 h at 4°C with 4 m guanidinium hydrochloride, 50 mm sodium acetate buffer, pH 6.8, containing the proteinase inhibitors disodium EDTA, 6-aminohexanoic acid, benzamidine hydrochloride and phenylmethanesulfonyl fluoride. Aggrecan was further purified on a CsCl density gradient under dissociative conditions [24].

glycosaminoglycan assay Glycosaminoglycan (GAG) levels in the culture media or cartilage were determined by the the

amount of polyanionic material reacting with 1,9-dimethylmethylene blue [25], using shark chondroitin sulfate as a standard. Results were reported as micrograms of GAG per milligram of wet weight cartilage or as percent total GAG released.

deglycosylation of aggrecan and aggrecan catabolites Proteoglycans and proteoglycan metabolites from media or cartilage were digested with chondroitinase ABC (0.01–0.1 units/10 mg GAG), keratanase (0.01–0.1 units/10 mg GAG and keratanase II (0.001–0.002 units/10 mg GAG) in buffer containing 50 mm sodium acetate, 0.1 m Tris-HCl, pH 6.5, at 37°C for 2 h or overnight at 4°C [17]. The digests were monitored by measuring the decrease in GAG content. After digestion, the samples were either dialyzed exhaustively against dH2O and dried using a speed vac or the proteoglycans were precipitated with five volumes of acetone and reconstituted in an appropriate volume of SDS-PAGE sample buffer.

analysis of aggrecan catabolites Aggrecan catabolites were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS/PAGE) under non-reducing conditions on 4–12% or 4–15% gradient polyacrylamide gels [26], followed by electroblotting onto nitrocellulose membranes. The membranes were probed with monoclonal antibody BC-3 (which recognizes the new N-terminal neoepitope 373ARGSVI . . . generated by cleavage of the aggrecan core protein by the unknown enzyme activity ‘aggrecanase’) [17] at a dilution of 1:1000. The BC-3 antibody has been shown to specific for this sequence of amino acids at the N-teminus of aggrecan fragments and does not recognize this same sequence when present within the peptide spanning the cleavage site [17, 19]. Subsequently, membranes were incubated with an alkaline phosphataseconjugated goat anti-mouse IgG and aggrecan catabolites visualized by incubation with the appropriate substrate for 10–30 min to achieve maximal signal intensity [27]. Sum density of all BC-3-reactive aggrecan fragments were then quantified by scanning densitometry using a UVP Image Store 7500 system (UVP, Upland, CA, U.S.A.) and IP Lab Gel software (Signal Analysis Corporation, Vienna, VA, U.S.A.) for density analysis and quantitation. Overnight transfer

Osteoarthritis and Cartilage Vol. 6 No. 3

resulted in complete transfer of low and high molecular weight standards and the densitometric response was found to be linear over the density range required for the blots, as determined by loading increasing amounts of BC-3-reactive material.

Results analysis of proteoglycan catabolites from bovine nasal cartilage exposed to either il-1bor tnfa Media samples containing aggrecan catabolites from bovine nasal cartilage exposed to either IL-1b (Fig. 1) or tumour necrosis factor (TNF) (Fig. 2) were subjected to SDS/PAGE (4–12%) and Western blot analysis with monoclonal antibody BC-3 that specifically recognizes the N-terminal neoepitope sequence ARGSVI . . . on aggrecan catabolites [Figs 1(a) and 2(a)] or analyzed for the release of proteoglycan metabolites by monitoring sulfated GAG using the dimethylmethylene blue assay [Figs 1(b) and 2(b)]. Western blot analysis

statistical analysis Significant differences in GAG release between groups were tested by Duncan’s multiple range test when analysis of variance was significant (P Q 0.05). Values are shown as mean 2 s.e.m. The r 2 value for GAG release versus BC-3 epitope generation in response to IL-1 or TNF was determined by regression analysis. (a)

(a)

IL-1 (ng/ml) MW

C 1000 100

20

10

217

1

TNF (ng/ml)

0.1

MW

250 kDa

250 kDa

98 kDa

98 kDa

64 kDa

64 kDa

C 10 000 1000 100 10

1

(b) 25

(b)

GAG release (µg/mg)

GAG release (µg/mg)

40

30

20

10

0

20

15

10

5

0 C

1000

100

20

10

1

0.1

IL-1 (ng/ml)

Fig. 1. IL-1 stimulation of bovine nasal cartilage aggrecan degradation. Bovine nasal cartilage was incubated in control media or in media containing varying concentrations of IL-1 (0.1–1000 ng/ml). At the end of the culture period, media were analyzed for (a) aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western blot analysis on 4–12% gels with loading on an equal GAG basis (20 mg/lane) or (b) for sulfated GAG concentrations by the dimethylmethylene blue dye assay.

C

10 000

1000

100

10

1

TNF (ng/ml)

Fig. 2. TNF stimulation of bovine nasal cartilage aggrecan degradation. Bovine nasal cartilage was incubated in control media or in media containing varying concentrations of TNF (1–10 000 ng/ml). At the end of the culture period, media were analyzed for (a) aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western blot analysis on 4–12% gels with loading on an equal GAG basis (20 mg/lane) or (b) for sulfated GAG concentrations by the dimethylmethylene blue dye assay.

Arner et al.: Aggrecanase-mediated cartilage degradation

indicated that exposure to IL-1b at a concentration of 10 ng/ml was sufficient to generate the release of BC-3-positive aggrecan degradation products into the media of the explant cultures [Fig. 1(a)]. In contrast, the minimum concentration of TNF required to generate aggrecan metabolites with the BC-3 epitope [Fig. 2(a)] was found to be 10-times that of the observed IL-1b concentration (100 ng/ml vs 10 ng/ml, respectively). No ‘aggrecanase’-generated product was evident in control cultures, without cytokine, or at cytokine concentrations which did not induce GAG release [Fig. 1(b) and 2(b)]. Densitometric analysis of the immunopositive staining for BC-3 and calculation of BC-3 epitope released (measured as arbitrary units of epitope/ml), at each concentration of IL-1b or TNF, demonstrated that there was a concentration-dependent increase in ‘aggrecanase’ activity with both cytokines. Comparison of these data with total proteoglycan metabolites (measured as micrograms of GAG per milliliter) (Fig. 3) indicated a significant positive correlation (r 2 = 0.99 and P = 0.0001) between GAG released into the media and ‘aggrecanase’-mediated fragment generation which suggests, but does not prove, a commonality of their origin.

(a)

80

60 % Total GAG released

218

40

20

0

10

50

4

Time (h)

8

16

24

48 200 97 68 45

–

+

–

+

–

+

–

+

–

+

Fig. 4. Timecourse of the release of aggrecan catabolites from control and IL-1 stimulated bovine nasal cartilage. Bovine nasal cartilage slices were cultured in the presence (R) or absence (W) of 50 ng/ml IL-1 for different time periods. At the end of the culture period, the media and cartilage slices were analyzed for GAG content by the dimethylmethylene blue dye assay and the percentage of GAG released was plotted vs time (a). To determine ‘aggrecanase’-generated cleavage, media were evaluated for fragments containing the N-terminus, ARGSVIL, by BC-3 Western analysis on 4–15% gels with loading on an equal GAG basis (20 mg/lane) (b).

6000

5000

4000 GAG (µg/ml)

40

(b)

IL-1 (50 ng/ml)

3000

2000

2

r = 0.98

time course of the generation of aggrecan catabolites from control and il-1b treated bovine explant cultures

1000

0

20 30 Time in culture (h)

2

4

6 8 10 BC-3 (epitope/ml)

12

14

16

Fig. 3. Correlation between IL-1- and TNFa-induced generation of BC-3 reactive aggrecan fragments and GAG release from bovine nasal cartilage. Sum density of all BC-3 reactive bands was quantitated by scanning densitometry for each treatment and BC-3 epitope contentration per milliliter of media plotted versus GAG concentration per milliliter from the experiments in Fig. 1 and 2 and the r 2 value determined by regression analysis.

Bovine nasal and bovine articular cartilage slices were cultured in the presence or absence of IL-1b (50 ng/ml) for different time periods over 48 h. Analysis of the release of proteoglycan metabolites into the media by GAG analysis showed that there was a difference in the percent release of total glycosaminoglycan from the bovine nasal compared to articular cartilage slices that were treated with a single dose of IL-1b [Figs 4(a) and 5(a)]. At 48 h the percentage of GAG released from

Osteoarthritis and Cartilage Vol. 6 No. 3

IL-1-treated bovine nasal cartilage was 72% compared to 8% release in control cultures or a ninefold increase, whilst in bovine articular cartilage there was 15% GAG released in IL-1-treated cultures, compared with 2% release in control cultures, or approximately an eightfold increase. Thus, although there was a difference in total percent GAG released in nasal vs articular cartilage, the fold increases were similar. BC-3 Western blot analysis of media detected ‘aggrecanase’-generated fragments from 8 to 48 h in IL-1b cultures of nasal cartilage [Fig. 4(b)] and articular cartilage [Fig. 5(b)]; no immunoreactive bands were detected at earlier timeperiods. A weak BC-3 positive immunostaining was seen in control bovine articular cartiage cultures;

(a)

219

1

2

3

250 kDa

98 kDa

64 kDa

16 14

% Total GAG released

12 10

Fig. 6. Effect of cycloheximide on ‘aggrecanase’-generated fragments produced in response to IL-1. Bovine nasal cartilage was incubated for 48 h in control media (lane 1) or 50 ng/ml IL-1 in the absence (lane 2) or presence (lane 3) of cycloheximide. At the end of the culture period, media were analyzed for aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western blot analysis on 4–12% gels with loading on an equal GAG basis (20 mg/lane).

8 6 4 2

0

10

20 30 Time in culture (h)

40

50

(b) Time (h)

4

8

16

24

48

200 97 68 IL-1 (50 ng/ml) –

+

–

+

– +

–

+

–

+

Fig. 5. Timecourse of the release of aggrecan catabolites from control and IL-1 stimulated bovine articular cartilage. Bovine articular cartilage slices were cultured in the presence (R) or absence (W) of 50 ng/ml IL-1 for different time periods. At the end of the culture period, the media and cartilage slices were analyzed for GAG content by the dimethylmethylene blue dye assay and the percentage of GAG released was plotted vs time (a). To determine ‘aggrecanase’-generated cleavage, media were evaluated for fragments containing the N-terminus, ARGSVIL, by BC-3 Western analysis on 4–15% gels with loading on an equal GAG basis (20 mg/lane) (b).

however, this was much less than that seen in the IL-1b treated cultures. In bovine nasal cartilage control cultures this was only weakly discernible at 24–48 h [Fig. 4(b)]. Loading of blots in these experiments was done on an equal GAG basis and essentially equal amounts of total epitope per GAG were detected from IL-1-stimulated cultures between 8–48 h. In other words, the time-dependent increase in GAG concentration was accompanied by an equivalent increase in the concentration of BC-3 reactivity. On the other hand, the BC-3 epitope per GAG in control samples was very low or undetectable in most cases. A BC-3 positive band migrating at 080 kDa was seen in the IL-1-stimulated cultures in this experiment which was absent in media from experiments shown in Fig. 1 and 2, and a band at 050 kDa varied in intensity in different samples in this experiment. These differences may represent animal to animal variations in GAG substitution,

220

Arner et al.: Aggrecanase-mediated cartilage degradation

and thereby susceptibility of C-terminal ‘aggrecanase’ sites to cleavage, due to age or other factors.

their effects on IL-1-stimulated GAG release [28, 29 and E.C. Arner and M. A. Pratta, unpublished data].

analysis of media from explant cultures treated with and without il-1b, and with il-1b in the presence of cycloheximide

inhibition of ‘aggrecanase’ cleavage by an active-site matrix metalloproteinase inhibitor, bb-16

Bovine nasal cartilage slices were cultured in the presence of IL-1b and cycloheximide for 48 h to investigate the requirement of de novo protein synthesis for the generation of ‘aggrecanase’ activity. Western blot analysis of aggrecan degradation products released into the medium of the explant cultures treated with IL-1b identified the expected BC-3 immunopositive bands (Fig. 6, lane 2). In contrast, the presence of ‘aggrecanase’-generated (BC-3-positive) proteoglycan catabolites was not observed in explant cultures containing IL-1b plus cycloheximide or in control cultures (Fig. 6, lanes 3 and 1, respectively). This result indicates that de novo protein synthesis is required for ‘aggrecanase’-mediated cleavage to occur.

In order to determine whether inhibitors of MMP activity would block ‘aggrecanase’-mediated

(a) IL-1 + drug MW

C IL-1

1

2

3

4

5

250 kDa

98 kDa 64 kDa 50 kDa

We investigated the ability of a variety of anti-arthritic compounds with differing mechanisms of action to inhibit specific cleavage of aggrecan at the Glu373–Ala374 ‘aggrecanase’ site in an IL-1-stimulated bovine nasal cartilage organ culture system following a 48 h incubation [Fig. 7(a)]. Media from control cultures did not exhibit any detectable BC-3-reactive fragments, whereas media from cartilage incubated with IL-1 (50 ng/ml) showed intense bands representing fragments generated by cleavage at the ‘aggrecanase’ site. Inclusion of indomethacin, tenidap, naproxen or doxycycline at 10 mm or dexamethasone at 1 mm did not inhibit the generation of BC-3-reactive fragments in response to IL-1. This lack of inhibition of ‘aggrecanase’-mediated cleavage was reflected in a lack of significant inhibition of aggrecan degradation by these compounds as monitored by release of GAG from the cartilage into the culture media [Fig. 7(b)]. Although there appears to be somewhat less GAG release in the presence of tenidap and somewhat more in the presence of doxycycline, these differences were not statistically significant and this is consistent with results seen in previous studies examining

(b)

50

40 GAG (µg/mg)

effect of anti-arthritic drugs on ‘aggrecanase’mediated cleavage of aggrecan in cartilage explant cultures exposed to il-1

30

20

10

0

C

IL-1

1

2

3

4

5

IL-1 + drug

Fig. 7. Effect of anti-arthritic drugs on cartilage aggrecan degradation. Bovine nasal cartilage was incubated for 48 h in control media or 50 ng/ml of IL-1 in the absence or presence of drug (1 = 10 mm indomethacin; 2 = 10 mm tenidap; 3 = 10 mm naproxen; 4 = 10 mm doxycycline; 5 = 1 mm dexamethasone). At the end of the culture period, media were analyzed for aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western blot analysis on 4–12% gels with loading on an equal volume basis (a) and for total aggrecan degradation by measuring the glycosaminoglycan (GAG) released per milligram of cartilage wet weight (b).

Osteoarthritis and Cartilage Vol. 6 No. 3 (a)

+ BB-16 (µM) MW

IL-1

C

0.1

1

10

100

250 kDa

98 kDa

(b) 30 25

GAG (µg/mg)

20 15 10 5 0

Control

IL-1

0.1

1

10

100

+ BB-16 (µM) (c) 100

% Inhibition

80

60

221

cleavage in cartilage explants, we used the peptidic hydroxamate, BB-16 [20], which is a potent, broad inhibitor of MMPs (MMP-3 Ki = 1.0 nm; MMP-1 Ki = 0.05 nm; MMP-8 Ki = 0.7 nm). Western blot analysis using monoclonal antibody BC-3 [Fig. 8(a)] showed that inclusion of BB-16 during IL-1-stimulated cartilage degradation resulted in a concentration-dependent inhibition of the generation of BC-3-reactive fragments. This inhibition of specific cleavage at the ‘aggrecanase’ site was reflected in a concentration-dependent decrease in GAG release [Fig. 8(b)] as monitored by dimethylmethylene blue dye assay of the culture media. Significantly, both BC-3 fragment generation and GAG release were inhibited by greater than 90% with this inhibitor. Although 30–100 mm BB-16 was required to achieve complete inhibition of GAG release, no effect on proteoglycan synthesis was seen at concentrations up to 100 mm (control = 3678 2 315 dpm/mg; IL-1 = 577 2 74 dpm/mg; IL-1 + BB-16 (100 mm) = 540 2 69 dpm/mg), indicating that inhibition was not due to a cytotoxic effect on chondrocytes or to interference with IL-1 signalling. The amount of compound required to cause half-maximal inhibition (IC50) of GAG release was estimated from the concentration–response curve to be 6.3 mm and the IC50 for inhibition of specific cleavage at the ‘aggrecanase’ site was estimated to be 2.0 mm [Fig. 8(c)]. The similar IC50s for inhibition of GAG release and inhibition of specific ‘aggrecanase’-mediated proteolysis are consistent with cleavage at the Glu373–Ala374 bond representing the primary proteolytic event in the catabolism of aggrecan in this explant system.

40

timecourse of il-1-induced cleavage in the absence and presence of bb-16

20

The timecourse of IL-1-induced aggrecan cleavage was investigated in the absence and presence of the MMP inhibitor, BB-16. Evaluation of specific cleavage at the Glu373–Ala374 ‘aggrecanase’ site by BC-3 Western blot analysis [Fig. 9(a)] showed an increase in the fragments formed by cleavage at this site with time in both the absence and presence of BB-16 with the rate of cleavage decreased in the presence of BB-16. Complete inhibition of BC-3 reactive fragment generation was achieved at 16 h with 30 mm BB-16. Although in the presence of inhibitor a faint band was detectable by 24 h which increased by 48 and 72 h, significant inhibition (81%) was maintained through 72 h. Fig. 9(b) shows that cummulative GAG release from the cartilage increased in a time-dependent manner in both the absence and

0 0.1

1

10

100

BB-16 (µM)

Fig. 8. Inhibition of ‘aggrecanase’-mediated cleavage by the hydroxamate matrix metalloproteinase inhibitor BB-16. Bovine nasal cartilage was incubated for 48 h in control media or 50 ng/ml of IL-1 in the absence or presence of BB-16. At the end of the culture period media were analyzed for aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western analysis on 4–12% gels with loading on an equal volume basis (a) and for aggrecan breakdown by measuring the glycosaminoglycan (GAG) released per milligram of cartilage wet weight (b). Percent inhibition of IL-1stimulated BC-3 epitope generation (Q) and GAG release (W) were plotted vs inhibitor concentration (c) to determine IC50 values.

222

Arner et al.: Aggrecanase-mediated cartilage degradation (a) IL-1 16 h

24 h

IL-1 + BB-16

48 h

72 h

16 h

24 h

48 h

72 h

250 kDa

98 kDa

(b) 18 16

GAG release (µg/mg)

14 12 10 8 6 4 2 0

8

16

24

32 40 Time (h)

48

56

64

72

Fig. 9. Timecourse of IL-1-induced aggrecan cleavage in the absence and presence of BB-16. Bovine nasal cartilage was incubated in IL-1 (10 ng/ml) in the absence (W) or presence (Q) of 30 mm BB-16 for various times. The media were then analyzed for BC-3 fragment generation by BC-3 Western blot analysis on 4–15% gels with loading on an equal GAG basis (a) and for GAG released into the media per milligram wet weight cartilage (b).

presence of inhibitor; however, the rate of release was significantly lower in the presence of BB-16. inhibition of aggrecan cleavage by an inhibitor of pro-mmp activation To determine whether interferring with normal MMP zymogen activation would also result in inhibition of ‘aggrecanase’-mediated cleavage, the isothiazolone, XG076, which interfers with activation of proMMPs [30] was included with IL-1 during culture. XG076 caused a concentration-

dependent inhibition of specific cleavage at the ‘aggrecanase’ site as measured by BC-3 Western blot analysis [Fig. 10(a)]. This again corresponded with inhibition of total aggrecan degradation as monitored by GAG release from the cartilage [Fig. 10(b)]. The IC50s for inhibition of GAG release and BC-3 fragment generation were estimated from the concentration–response curves [Fig. 10(c)] to be 3.0 mm and 0.5 mm, respectively. As was found with BB-16, the IC50s are similar, again suggesting that cleavage at the ‘aggrecanase’ site is the primary proteolytic event in the catabolism of aggrecan in this explant system.

Osteoarthritis and Cartilage Vol. 6 No. 3 (a)

223

Discussion + XG076 (µM) MW

C

IL-1

10

1

0.1

0.01

250 kDa

98 kDa

(b) 30

GAG (µg/mg)

25 20 15 10 5 0

Control

IL-1

10

1

0.1

0.01

XG076 (µM) (c) 100

% Inhibition

80

60

40

20

0 0.001

0.01

0.1 XG076 (µM)

1

10

Fig. 10. Inhibition of aggrecan degradation by an inhibitor of matrix metalloproteinase activation. Bovine nasal cartilage was incubated for 48 h in control media or 10 ng/ml of IL-1 in the absence or presence of the isothiazolone, XG076. At the end of the culture period, media were analyzed for aggrecan fragments containing the N-terminus, ARGSVIL, by BC-3 Western analysis on 4–12% gels with loading on an equal volume basis (a) and for aggrecan breakdown by measuring the glycosaminoglycan (GAG) released per milligram of cartilage wet weight (b). Percent inhibition of IL-1stimulated BC-3 epitope generation (Q) and GAG release (W) were plotted versus inhibitor concentration (c) to determine IC50 values.

We have demonstrated that proteoglycan degradation in bovine cartilage in response to either IL-1 or TNF stimulation is associated with the release of aggrecan fragments generated by cleavage at the Glu373–Ala374 ‘aggrecanase’ site. Cleavage at this site is observed during stimulated proteoglycan catabolism in articular as well as nasal cartilage. There was both a temporal and concentration-related association between the appearance of BC-3 immunoreactive fragments (indicating specific cleavage at the ‘aggrecanase’ site) and the release of GAG from cartilage (as a measure of total proteoglycan degradation). The strong positive correlation which we observed between GAG release and specific cleavage at the ‘aggrecanase’ site in concentration–response studies with IL-1 or TNF suggests that the degradation of aggrecan in response to these cytokines is associated with cleavage by ‘aggrecanase’. Although we have found that MMP production is stimulated in response to IL-1 and TNF in these cultures, including proMMP-1, -2, -3, and -9 (E. C. Arner and M. D. Tortorella, unpublished observations), Western blot analyses using antibody BC-14 (which recognizes the catabolic neoepitope FFGVG . . . on aggrecan catabolites produced by proteolysis of the interglobular domain of aggrecan with MMPs) were negative (data not shown), thus suggesting that the aggrecan catabolites were not generated through the action of MMPs. Our data are consistent with studies in rat chondrosarcoma cell lines and primary bovine chondrocytes stimulated with retinoic acid or IL-1, where aggrecan catabolic products contained a single N-terminal sequence initiating at Ala374 and a single G1 fragment with the C-terminus Glu373 indicating that aggrecan catabolism in these systems primarily involves cleavage at the Glu373–Ala374 bond by ‘aggrecanase’ [10]. The work of Lark et al. [31] demonstrate that both ‘aggrecanase’ and MMP activities are involved in cleavage of aggrecan in normal and arthritic human cartilage. However, as the authors point out, since these studies evaluated N-terminal aggrecan cleavage fragments remaining within the cartilage, they only show that cleavage at both sites had occurred, but do not provide any information on the relative contribution of these two activities or whether the fragments present were generated during normal turnover or during the arthritic process. Clearly in the current studies where aggressive degradation

224

Arner et al.: Aggrecanase-mediated cartilage degradation

of cartilage aggrecan is induced by IL-1, ‘aggrecanase’-mediated cleavage appears to predominate. Furthermore, we have demonstrated that specific cleavage of aggrecan at the Glu373–Ala374 ‘aggrecanase’ site is blocked by MMP inhibitors. This inhibition is concentration related and corresponds with inhibition of aggrecan catabolism as monitored by release of GAG from the cartilage. Similar IC50s obtained with BB-16 for inhibition of GAG release and inhibition of the generation of fragments containing the new N-terminus ARGSVIL, produced by ‘aggrecanase’mediated proteolysis, are consistent with cleavage at the Glu373–Ala374 bond representing the primary proteolytic event in the catabolism of aggrecan. The possibility that there is a slight difference in sensitivity of GAG release and BC-3 fragment generation to the inhibitors tested could not be ruled out. This could potentially be due to cleavage by ‘aggrecanase’ within the C-terminal region of aggrecan, where three additional ‘aggrecanase’ sites have been proposed [32], that would not be detected with the BC-3 antibody. Although the tissue content of BC-3 was not evaluated in these studies to confirm that fragments containing this epitope are eluting out of the cartilage as would be expected, in ongoing studies, data on cartilage extracts indicate that there is no buildup of BC-3 reactive fragments within the cartilage of inhibitor-treated cultures, but rather that these fragments are not produced in the presence of inhibitor (E. C. Arner and M. A. Pratta, unpublished data). The data reported herein for inhibition with BB-16 are supported by studies demonstrating a significant positive correlation (r 2 = 0.84; P = 0.0001) between percent inhibition of GAG release and percent inhibition of BC-3-reactive fragment generation for a series of MMP inhibitors (E. C. Arner, C. P. Decicco and M. D. Tortorella, unpublished data). Consistent with this suggestion is the strong positive correlation that we observed between GAG release and specific cleavage at the ‘aggrecanase’ site in response to IL-1b or TNFa stimulation of bovine nasal cartilage. In contrast, three different nonsteroidal anti-inflammatory drugs, indomethacin, tenidap and naproxen, were ineffective in blocking generation of BC-3-reactive fragments as were the steroid, dexamethasone, and the tetracycline analog, doxycycline. The compounds used in these experiments were previously tested in concentration–response studies for their ability to inhibit GAG release from bovine nasal cartilage in response to IL-1 [28, 29, and E. C. Arner and M. A.

Pratta, unpublished observations] and shown to be ineffective over a range of concentrations. The concentrations used in this study were generally chosen to be the highest concentration which did not cause a decrease in proteoglycan synthesis. Our goal in the current studies was to determine whether any of these compounds which do not block GAG release from cartilage would, in contrast, result in an inhibition of cleavage at the ‘aggrecanase’ site. If this were the case, it might suggest that cleavage by ‘aggrecanase’ is not critical for proteoglycan degradation. However, none of these compounds affected specific cleavage at the ‘aggrecanase’ site as detected by BC-3 Western analysis. These results are consistent with the hypothesis that release of GAG is correlated with aggrecanase-mediated cleavage. The nonsteroidal anti-inflammatory drugs, indomethacin and naproxen, have been reported by a number of investigators to have either no effect or to negatively effect aggrecan catabolism in vitro [28, 29, 33–39]. Although there are some reports of naproxen inhibiting cartilage proteoglycan breakdown [40, 41], neither indomethacin nor naproxen have been shown to be effective clinically in blocking cartilage degradation [42]. Tenidap has been reported to be capable of blocking cartilage degradation via several mechanisms including inhibition of IL-1 production, inhibition of MMP synthesis and activity, and reduction of IL-1 receptor levels [43–46]. However, in the current studies no significant inhibition of aggrecan loss was observed with 10 mm tenidap, and this correlated with a lack of inhibition of specific cleavage at the ‘aggrecanase’ site. Interestingly, however, the tenidap-treated cultures which showed a nonsignificant 31% inhibition of GAG release, also showed a 28% decrease in BC-3 sum density. When the concentration of tenidap was increased to 100 mm, decreased loss of GAG was achieved; however, this was paralleled by a significant inhibition of aggrecan synthesis (E. C. Arner, unpublished data). Since tenidap has been reported to decrease chondrocyte cell viability and decrease DNA content in chondrocyte cultures [47] at similar concentrations, this suppression of aggrecan synthesis suggests that a general cytotoxic effect of tenidap may be responsible for the decrease in aggrecan catabolism at these high drug concentrations in our studies. Tetracyclines have been reported to inhibit MMP-1, -8, and -9, but not MMP-3 [48–50] and thus may be more limited inhibitors, not providing broad inhibition of the entire MMP family. The

Osteoarthritis and Cartilage Vol. 6 No. 3

lack of inhibition of BC-3-reactive fragment generation by the tetracycline analog, doxycycline, suggests that these compounds are not effective against ‘aggrecanase’-mediated cartilage degradation. Dexamethasone was also ineffective in blocking specific cleavage at the ‘aggrecanase’ site or general aggrecan degradation as monitored by GAG release. These results are consistent with the lack of clinical efficacy of steroids in preventing cartilage loss in man. Timecourse studies demonstrated that BC-3-reactive aggrecan catabolites were not detectable in the media or cartilage until 8 h of stimulation with 50 ng/ml IL-1 or 16 h of stimulation with 10 ng/ml IL-1, indicating that time is required for the generation of ‘aggrecanase’-mediated cleavage. This observation together with the cycloheximide data indicates that the degradative mechanisms leading to the induction of ‘aggrecanase’ activity in cartilage requires de novo protein synthesis. BC-3 reactive bands were generally not detectable in media from control cultures with equal loading on a GAG basis, indicating that aggrecanase is not responsible for proteoglycan release under unstimulated conditions. Interestingly, in the timecourse studies carried out in the absence and presence of BB-16 (Fig. 9), although samples were loaded on an equal GAG basis, different amounts of BC-3 epitope were observed in the 16 and 24 h as compared with the 48 and 72 h cultures. This indicates that the BC-3 epitope per GAG is different for the early versus later timepoints, suggesting that at the early times a lower proportion of the released GAG-containing fragments had been cleaved at the ‘aggrecanase’ site. Therefore, it is likely that at the early timeperiods, when a low amount of stimulated GAG release (over that seen in control cultures) is observed, only a small portion of the GAG release is due to ‘aggrecanase’, while the remainder occurs via the mechanism responsible for unstimulated turnover. At later timepoints, the majority of the GAG release would be due to stimulated degradation, which our data indicates is ‘aggrecanase’ mediated. A lower amount of BC-3 epitope per GAG was also seen in the inhibitor-treated cultures. This is to be expected if the compound is inhibiting the ‘aggrecanase’ activity responsible for stimulated degradation, but not effecting the factor responsible for control release. Although MMP-1, -2, -3, -7, -8, -9, and -13 all cleave at the Asn341–Phe342 bond on incubation with aggrecan in solution [3, 4, 11, 12], the report that the Glu373–Ala374 bond can be cleaved by incubation

225

with high concentrations of MMP-8 [14] opened the possibility that ‘aggrecanase’ might be a new member of the MMP family which may directly degrade aggrecan at the Glu373–Ala374 bond or be part of a proteolytic cascade which results in this cleavage. The data presented herein showing that cleavage at the Glu373–Ala374 bond in response to IL-1 stimulation can be inhibited by a hydroxamate active-site inhibitor with a broad range of selectivity within the MMP family is consistent with this hypothesis. We have also demonstrated inhibition of ‘aggrecanase’-mediated cleavage by an isothiazalone compound that acts through a completely different mechanism to inhibit MMPs by interfering with the activation of the latent MMP zymogen [30]. XG076 has no effect on proteoglycan synthesis in control cartilage cultures or on protein synthesis in chondrocyte cultures (unpublished data) and does not abrogate the inhibition of proteoglycan synthesis in response to IL-1 [30] indicating that it does not interfere with IL-1 signalling, thus supporting the contention that this compound is specifically inhibiting the activation of a pro-MMP in this system. Data with this inhibitor provides further support to suggest that an enzyme with characteristics in common with the known MMPs is involved in degradation of proteoglycan at the ‘aggrecanase’ site. We have recently demonstrated that BB-16 is also capable of inhibiting the proteolytic activity of soluble active ‘aggrecanase’ in conditioned media from IL-1-stimulated bovine nasal cartilage cultures [51]. In contrast, not all MMP inhibitors are effective in blocking ‘aggrecanase’ activity. Although BB-16 is ineffective against serine, cysteine or aspartyl proteases (E.C. Arner, unpublished data), hydroxamates have been shown to inhibit other related metalloproteases, such as those of the ADAMs (a disintegrin and metalloprotease domain) family, like TNF convertase [52, 53]. Thus, although inhibition by BB-16 supports the possibility that ‘aggrecanase’ may be a member of the MMP family, it does not rule out the alternative possibility that ‘aggrecanase’ may be a member of a related family of metalloproteases. The Ki’s for the active-site inhibitor, BB-16, against MMP-1, -3, and -8 in enzymatic assays using a peptide substrate, are in the nanomolar range while the IC50 for inhibition of GAG release and specific ‘aggrecanase’-mediated cleavage at the Glu373–Ala374 bond are in the micromolar range, suggesting that none of these known MMPs represent ‘aggrecanase’. However, the possibility

226

Arner et al.: Aggrecanase-mediated cartilage degradation

exists that these differences may reflect differences in substrate concentration or presentation within the cartilage matrix. Interestingly, the IC50 for inhibition of normal pro-MMP activation by the isothiazolone, XG076, using proMMP-3 as a model matrix metalloproteinase zymogen, is in the micromolar range (E. C. Arner and M. D Tortorella, unpublished data) similar to the value for inhibition of cartilage aggrecan cleavage. MMP-3 does not cleave aggrecan in solution at the Glu373–Ala374 bond. However, since the cysteine switch mechanism of activation is conserved across the MMP family, inhibition of activation would be expected to be similar for all MMPs and again this is consistent with ‘aggrecanase’ representing a protease having properties in common with members of the MMP family. Since the ADAM proteins also possess a propeptide with a conserved cysteine similar to the MMPs, it is conceivable that the isothiazolone can also block this activation, although this is speculative at this point. Aggrecan catabolites identified in synovial fluid of patients with inflammatory joint diseases and osteoarthritis exhibit N-terminal amino acid sequences indicating formation by cleavage at the Glu373–Ala374 ‘aggrecanase’ cleavage site [5, 6]. Our studies using in vitro cartilage model systems give further information indicating a selective up-regulation of specific cleavage at the ‘aggrecanase’ site in response to the cytokines IL-1 and TNF and demonstrate that MMP inhibitors, as opposed to other anti-arthritic compounds, are effective in blocking cleavage at the ‘aggrecanase’ site. These data suggest that ‘aggrecanase’ may be a member of the MMP family or closely related protease family, and that inhibiting this enzyme may have utility in preventing cartilage loss in arthritis.

4.

5.

6.

7.

8.

9. 10.

11.

12.

Acknowledgments We thank Robert Copeland for MMP Ki values and Robert Newton for critical reading of the manuscript. This work was supported in part by the Arthritis and Rheumatism Council of Great Britain.

References 1. Mankin HJ, Lippiello L. Biochemical and metabolic abnormalities inarticular cartilage from osteo-arthritic human hips. J Bone Joint Surg[Am] 1970;52:424–34. 2. Sandy JD, Boynton RE, Flannery CR. Analysis of the catabolism ofaggrecan in cartilage explants by quantitation of peptides from the three globular domains. J Biol Chem 1991;266:8198–205. 3. Fosang AJ, Neame PJ, Last K, Hardingham TE, Murphy G, Hamilton JA. The interglobular

13.

14.

15.

16.

domain of cartilage aggrecan is cleaved by PUMP, gelatinases, and cathepsin B. J Biol Chem 1992;267:19470–4. Flannery CR, Lark MW, Sandy JD. Identification of a stromelysin cleavage site within the interglobular domain of human aggrecan. Evidence for proteolysis at this site in vivo in human articular cartilage. J Biol Chem 1992; 267:1008–14. Sandy JD, Flannery CR, Neame PJ, Lohmander LS. The structure of aggrecan fragments in human synovial fluid. Evidence for the involvement in osteoarthritis of a novel proteinase which cleaves the Glu373–Ala374 bond of the interglobular domain. J Clin Invest 1992;89:1512–6. Lohmander LS, Neame PJ, Sandy JD. The structure of aggrecan fragments in human synovial fluid. Evidence that aggrecanase mediates cartilage degradation in inflammatory joint disease, joint injury, andosteoarthritis. Arthritis Rheum 1993;36:1214–22. Sandy JD, Neame 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: 8683–5. Loulakis P, Shrikhande A, Davis G, Maniglia CA. N-terminal sequence of proteoglycan fragments isolated from medium of interleukin-1- treated articular-cartilage cultures. Putative site(s) of enzymatic cleavage. Biochem J 1992; 284(Pt 2): 589–93. Ilic MZ, Handley CJ, Robinson HC, Mok MT. Mechanism of catabolism of aggrecan by articular cartilage. Arch Biochem Biophys 1992;294:115–22. Lark MW, Gordy JT, Weidner JR, Ayala J, Kimura JH, Williams HR et al. 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;270:2550–6. Fosang AJ, Last K, Knauper V, Neame PJ, Murphy G, Hardingham TE et al. Fibroblast and neutrophil collagenases cleave at two sites in the cartilage aggrecan interglobular domain. Biochem J 1993;295:273–6. Fosang AJ, Last K, Knauper V, Murphy G, Neame PJ. Degradation of cartilage aggrecan by collagenase-3 (MMP-13). FEBS Lett 1996;380: 17–20. Flannery CR, Sandy JD. Aggrecan catabolism in cartilage: studies on the nature of a novel proteinase (aggrecanase) which cleaves the Glu373–Ala374 bond of the interglobular domain. Trans Orthop Res Soc 1993;18:190. Fosang AJ, Last K, Neame PJ, Murphy G, Knauper V, Tschesche H et al. Neutrophil collagenase (MMP-8) cleaves at the aggrecanase site E373-A374 in the interglobular domain of cartilage aggrecan. Biochem J 1994;304:347–51. Arner EC, Decicco CP, Cherney R, Tortorella MD. Cleavage of native cartilage aggrecan by neutrophil collagenase (MMP-8) is distinct from endogenous cleavage by aggrecanase. J Biol Chem 1997;272:9294–9. Chubinskaya S, Huch K, Mikecz K, Cs-Szabo G, Hasty KA, Kuettner KE,Cole AA. Chondrocyte

Osteoarthritis and Cartilage Vol. 6 No. 3

17.

18.

19.

20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

30.

matrix metalloproteinase-8: up-regulation ofneutrophil collagenase by interleukin-1 beta in human cartilage from knee and ankle joints. Lab Invest 1996;74:232–40. Hughes CE, Caterson B, Fosang AJ, Roughley PJ, Mort JS. Monoclonal antibodies that specifically recognize neoepitope sequences generated by ‘aggrecanase’ and matrix metalloproteinase cleavage of aggrecan: application to catabolism in situ and in vitro. Biochem J 1995;305:799–804. Huang JJ, Newton RC, Pezzella K, Covington M, Tamblyn T, Rutlege SJ et al. High-level expression in Escherichia coli of a soluble and fully active recombinant interleukin-1 beta. Mol Biol Med 1987;4:169–81. Hughes CE, Buttner FH, Eidenmuller B, Caterson B, Bartnik. Utiliation of a recombinant substrate (rAgg1) to study the biochemical properties of ‘aggrecanase’ in cell culture systems. J Biol Chem 1997;272:20269–72. Campion C, Davidson AH, Dickens JP, Crimmins MJ. Preparation of N-(4-(N-hydroxyamino) succinyl)amino acid amides as collagenase inhibitors. November 23, 1989;PCT/GB89/01399. Wright SW, Petraitis JJ, Abelman MM, Batt DG, Bostrom LL, Corbett RL, Decicco CP et al. Heteroaryl-fused 2-phenylisothiazolone inhibitors of cartilage breakdown. J Med Chem 1994;37: 3071–8. Steinberg J, Sledge CB, Noble J, Stirrat CR. A tissue-culture model of cartilage breakdown in rheumatoid arthritis. Quantitative aspects of proteoglycan release. Biochem J 1979;180:403–12. Arner EC, Pratta MA. Modulation of interleukin-1-induced alterations in cartilage proteoglycan metabolism by activation of protein kinase C. Arthritis Rheum 1991;34:1006–13. Hascall VC, Sajdera SW. Proteinpolysaccharide complex from bovine nasal cartilage. The function of glycoprotein in the formation of aggregates. J Biol Chem 1969;244:2384–96. Farndale RW, Sayers CA, Barrett AJ. A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures. Connect Tissue Res 1982;9:247–8. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. Hughes CE, Caterson B, White RJ, Roughley PJ, Mort JS. Monoclonal antibodies recognizing protease-generated neoepitopes from cartilage proteoglycan degradation. Application to studies of human link protein cleavage by stromelysin. J Biol Chem 1992;267:16011–4. Arner EC, Pratta MA. Independent effects of interleukin-1 on proteoglycan breakdown, proteoglycan synthesis, and prostaglandin E2 release from cartilage in organ culture. Arthritis Rheum 1989;32:288–97. Arner EC, Darnell LR, Pratta MA, Newton RC, Ackerman NR, Galbraith W. Effect of antiinflammatory drugs on human interleukin-1induced cartilage degradation. Agents Actions 1987;21:334–6. Arner EC, Pratta MA, Freimark B, Lischwe M, Trzaskos JM, Magolda RL, Wright SW. Isothiazolones intefere with normal matrix

31.

32.

33.

34.

35. 36.

37.

38. 39.

40.

41.

42.

43. 44.

227

metalloproteinase activation and inhibit cartilage proteoglycan degradation. Biochem J 1996;318: 417–24. Lark, MW, Bayne EK, Flanagan J, Harper CF, Hoerner LA, Hutchinson NI et al. Aggrecan degradation in human cartilage: evidence for both matrix metalloproteinase and aggrecanase activity in normal, osteoarthritic and rheumatoid joints. J Clin Invest 1997;100:93–106. Loulakis P, Shrikhande A, Davis G, Maniglia. N-terminal sequence of proteoglycan fragments isolated from medium of interleukin-1-treated articular-cartilage cultures: putative site(s) of enzymatic cleavage. Biochem J 1992;284:589–93. Brandt KD, Albrecht M. Effect of naproxen sodium on the net synthesis of glycosaminoglycans and protein by normal canine articular cartilage in vitro. J Pharm Pharmacol 1990;42: 738–40. Gentry C. The effects of the active metabolite of nabumetone (6MNA) on the synthesis of normal canine articular cartilage in vitro: comparison with other non-steroidal anti-inflamatory drugs (NSAIDS). Br J Rheumatol 1992;31(Suppl. 2):168. Rainsford KD. Scand J Rheumatol 1994;(Suppl.) 98:138. Lane NE, Williams RJ III, Schurman DJ, Smith RL. Inhibition of interleukin 1 induced chondrocyte protease activity by a corticosteroid and a nonsteroidal anti-inflammatory drug. J Rheumatol 1992;19:135–9. Dingle JT, Shield MJ. The interactions of cytokines, NSAIDs and prostaglandins in cartilage destruction and repair. Adv Prostaglandin Thromboxane Leukot Res 1991;21B:955–66. Shield MJ. Anti-inflammatory drugs and their effects on cartilage synthesis and renal function. Eur J Rheumatol Inflamm 1993;13:7–16. Steinmeyer J, Daufeldt S, Kalbhen DA. Pharmacologic influence on the activity of stromelysin from bovine articular cartilage. Ann NY Acad Sci 1994;732:482–3. Glazer PA, Rosenwasser MP, Ratcliffe A. The effect of naproxen and interleukin-1 on proteoglycan catabolism and on neutral metalloproteinase activity in normal articular cartilage in vitro. J Clin Pharmacol 1993;33:109–14. Ratcliffe A, Azzo W, Saed-Nejad F, Lane N, Rosenwasser MP, Mow VC. In vivo effects of naproxen on composition, proteoglycan metabolism, and matrix metalloproteinase activities in canine articular cartilage. J Orthop Res 1993;11:163–71. Cawston TE, Curry V, Ramsey S, Clark IM, Kyle VA, Adebajo A, Silverman B et al. The levels of collagenase, tissue inhibitor of metalloproteinases-1 (TIMP-1), collagenaseTIMP-1 complexes and glycosaminoglycan (GAG) in sequential samples of synovial fluid aspirated from patients with osteoarthritis. Clin Exp Rheumatol 1995;13:431–7. Dingle JT, Leeming MR, Martindale JJ. Effect of tenidap on cartilage integrity in vitro. Ann Rheum Dis 1993;52:292–9. Pelletier JP, McCollum R, DiBattista J, Loose LD, Cloutier JM, Martel-Pelletier J. Regulation of human normal and osteoarthritic chondrocyte

228

45.

46.

47.

48.

Arner et al.: Aggrecanase-mediated cartilage degradation interleukin-1 receptor by antirheumatic drugs. Arthritis Rheum 1993;36:1517–27. Fernandes JC, Martel-Pelletier J, Otterness IG, Lopez-Anaya A, Mineau F, Tardif G, Pelletier JP. Effects of tenidap on canine experimental osteoarthritis. I. Morphologic and metalloprotease analysis. Arthritis Rheum 1995;38:1290–303. Fernandes JC, Caron JP, Martel-Pelletier J, Jovanovic D, Mineau F, Tardif G et al. Effects of tenidap on the progression of osteoarthritic lesions in a canine experimental model. Suppression of metalloprotease and interleukin-1 activity. Arthritis Rheum 1997;40:284–94. Srinivas GR, Chichester CO, Barrach HJ, Matoney AL. Effects of certain antiarthritic agents on the synthesis of type II collagen and glycosaminoglycans in rat chondrosarcoma cultures. Agents Actions 1994;41:193–9. Greenwald RA. Tetracyclines, derivatives as potential inhibitors of connective tissue degradation. Drug News & Perspect 1990;3:161–6.

49. Greenwald RA, Golub LM, Lavietes B, Ramamurthy NS, Gruber B, Laskin RS, McNamara TF. Tetracyclines inhibit human synovial collagenase in vivo and in vitro. J Rheumatol 1987;14:28–32. 50. Arsenis C, Greenwald RA, Moak SA, Laskin RS. Inhibition of neutral metalloproteinase activities in epiphyseal and articular cartilages by tetracyclines both in vivo and in vitro. Trans Orthop Res Soc 1990;15:268. 51. Arner EC, Decicco CP, Pratta MA, Copeland RA, Trzaskos DM, Magolda RL, Tortorella MD. Characterization of soluble bovine cartilage ‘aggrecanase’. Trans Orthop Res Soc 1997;22:103. 52. Black RA, Rauch CT, Koziosky CJ, Peschon JJ, Slack JL, Wolfson MF et al. A metalloproteinase disintegrin that releases tumour-necrosis factor-a from cells. Nature 1997;385:729–33. 53. Moss ML, Jin S-L C, Millia ME, Burkhart W, Carter HL, Chen W-J et al. Cloning of a disintegrin metalloproteinase that processes precursor tumour-necrosis factor-a. Nature 1997;385:733–6.