Disrupted type II collagenolysis impairs angiogenesis, delays endochondral ossification and initiates aberrant ossification in mouse limbs.

Cartilage remodelling and chondrocyte differentiation are tightly linked to angiogenesis during bone development and endochondral ossification. To investigate whether collagenase-mediated cleavage of the major cartilage collagen (collagen II) plays a role in this process, we generated a knockin mouse in which the mandatory collagenase cleavage site at PQG775↓776LAG, was mutated to PPG775↓776MPG (Col2a1Bailey). This approach blocked collagen II cleavage, and the production of putative collagen II matrikines derived from this site, without modifying matrix metalloproteinase expression or activity. We report here that this mouse (Bailey) is viable. It has a significantly expanded growth plate and exhibits delayed and abnormal angiogenic invasion into the growth plate. Deeper electron microscopy analyses revealed that, at around five weeks of age, a small number of blood vessel(s) penetrate into the growth plate, leading to its abrupt shrinking and the formation of a bony bridge. Our results from in vitro and ex vivo studies suggest that collagen II matrikines stimulate the normal branching of endothelial cells and promote blood vessel invasion at the chondro-osseous junction. The results further suggest that failed collagenolysis in Bailey leads to expansion of the hypertrophic zone and formation of a unique post-hypertrophic zone populated with chondrocytes that re-enter the cell cycle and proliferate. The biological rescue of this in vivo phenotype features the loss of a substantial portion of the growth plate through aberrant ossification, and narrowing of the remaining portion that leads to limb deformation. Together, these data suggest that collagen II matrikines stimulate angiogenesis in skeletal growth and development, revealing novel strategies for stimulating angiogenesis in other contexts such as fracture healing and surgical applications.

Abundant LacZ activity in the absence of Cre expression in the normal and inflamed synovium of adult Col2a1-Cre; ROSA26RLacZ reporter mice.

Recent analyses of Col2a1-Cre; ROSA26R reporter mice showed that synovial fibroblasts in 7-day mice were LacZ positive, due to a history of Col2a1-Cre expression conferred by their origin in the interzone of the developing joint. We have examined LacZ staining in adult Col2a1-Cre(+/0); ROSA26R(LacZ) mice, with and without inflammatory arthritis, and found that synovial fibroblasts in normal and inflamed synovium are LacZ positive, but Cre negative. Our results suggest that Cre-mediated recombination in joint interzone cells during development endure in adult synovial cells despite the absence of ongoing Cre expression. These findings have important implications and applications for the study of synovial inflammation in models of experimental arthritis.

Identifying the human aggrecanase.

It is clear that A Disintegrin And Metalloproteinase with ThromboSpondin motif (ADAMTS)-5 is the major aggrecanase in mouse cartilage, however it is not at all clear which enzyme is the major aggrecanase in human cartilage. Identifying the human aggrecanase is difficult because multiple, independent, molecular processes determine the final level of enzyme activity. As investigators, we have good methods for measuring changes in the expression of ADAMTS mRNA, and good methods for detecting aggrecanase activity, but no methods that distinguish the source of the activity. In between gene expression and enzyme action there are many processes that can potentially enhance or inhibit the final level of activity. In this editorial we discuss how each of these processes affects ADAMTS activity and argue that measuring any one process in isolation has little value in predicting overall ADAMTS activity in vivo.

Matrix metalloproteinase 13-deficient mice are resistant to osteoarthritic cartilage erosion but not chondrocyte hypertrophy or osteophyte development.

OBJECTIVE: To investigate the role of matrix metalloproteinase 13 (MMP-13; collagenase 3) in osteoarthritis (OA). METHODS: OA was surgically induced in the knees of MMP-13-knockout mice and wild-type mice, and mice were compared. Histologic scoring of femoral and tibial cartilage aggrecan loss (0-3 scale), erosion (0-7 scale), and chondrocyte hypertrophy (0-1 scale), as well as osteophyte size (0-3 scale) and maturity (0-3 scale) was performed. Serial sections were stained for type X collagen and the MMP-generated aggrecan neoepitope DIPEN. RESULTS: Following surgery, aggrecan loss and cartilage erosion were more severe in the tibia than femur (P<0.01) and tibial cartilage erosion increased with time (P<0.05) in wild-type mice. Cartilaginous osteophytes were present at 4 weeks and underwent ossification, with size and maturity increasing by 8 weeks (P<0.01). There was no difference between genotypes in aggrecan loss or cartilage erosion at 4 weeks. There was less tibial cartilage erosion in knockout mice than in wild-type mice at 8 weeks (P<0.02). Cartilaginous osteophytes were larger in knockout mice at 4 weeks (P<0.01), but by 8 weeks osteophyte maturity and size were no different from those in wild-type mice. Articular chondrocyte hypertrophy with positive type X collagen and DIPEN staining occurred in both wild-type and knockout mouse joints. CONCLUSION: Our findings indicate that structural cartilage damage in a mouse model of OA is dependent on MMP-13 activity. Chondrocyte hypertrophy is not regulated by MMP-13 activity in this model and does not in itself lead to cartilage erosion. MMP-13 deficiency can inhibit cartilage erosion in the presence of aggrecan depletion, supporting the potential for therapeutic intervention in established OA with MMP-13 inhibitors.

Keratan sulphate in the interglobular domain has a microstructure that is distinct from keratan sulphate elsewhere on pig aggrecan.

The microstructure of keratan sulphate purified from the interglobular domain, the keratan sulphate-rich region and total aggrecan was compared using fluorophore-assisted-carbohydrate-electrophoresis. Keratan sulphate in the interglobular domain was substantially less sulphated than keratan sulphate elsewhere on aggrecan, based on the ratio of unsulphated: monosulphated disaccharides generated by endo-beta-galactosidase digestion, and the ratio of monosulphated: disulphated disaccharides generated by keratanase II digestion. The ratio of unsulphated: monosulphated: disulphated disaccharides was 1:4:5 for keratan sulphate from total aggrecan and the keratan sulphate-rich region, but only 1:0.9:0.8 for the interglobular domain. These results show that keratan sulphate in the interglobular domain of pig aggrecan has a microstructure that is distinct from keratan sulphate in the keratan sulphate-rich region.

ADAMTS-5: the story so far.

The recent discovery of ADAMTS-5 as the major aggrecanase in mouse cartilage came as a surprise. A great deal of research had focused on ADAMTS-4 and much less was known about the regulation, expression and activity of ADAMTS-5. Two years on, it is still not clear whether ADAMTS-4 or ADAMTS-5 is the major aggrecanase in human cartilage. On the one hand there are in vitro studies using siRNA, neutralising antibodies and immunoprecipitation with anti-ADAMTS antibodies that suggest a significant role for ADAMTS-4 in aggrecanolysis. On the other hand, ADAMTS-5 (but not ADAMTS-4)-deficient mice are protected from cartilage erosion in models of experimental arthritis, and recombinant human ADAMTS-5 is substantially more active than ADAMTS-4. The activity of both enzymes is modulated by C-terminal processing, which occurs naturally in vivo. The most interesting finding to emerge from our comparison of ADAMTS-5 and ADAMTS-4 is that in terms of gene regulation, these two enzymes are the antitheses of each other. In most cases, ADAMTS-5 is constitutively expressed in human chondrocytes and synovial fibroblasts, whereas ADAMTS-4 expression is induced by proinflammatory cytokines. This paper reviews the data on ADAMTS-5 so far. It represents a snapshot in time of a field that is fast-moving and very exciting.

Hyaluronan synthesis and degradation in cartilage and bone.

Hyaluronan (HA) is a large but simple glycosaminoglycan composed of repeating D-glucuronic acid, beta1-3 linked to N-acetyl-D-glucosamine beta1-4, found in body fluids and tissues, in both intra- and extracellular compartments. Despite its structural simplicity, HA has diverse functions in skeletal biology. In development, HA-rich matrices facilitate migration and condensation of mesenchymal cells, and HA participates in joint cavity formation and longitudinal bone growth. In adult cartilage, HA binding to aggrecan immobilises aggrecan, retaining it at the high concentrations required for compressive resilience. HA also appears to regulate bone remodelling by controlling osteoclast, osteoblast and osteocyte behaviour. The functions of HA depend on its intrinsic properties, which in turn rely on the degree of polymerisation by HA synthases, depolymerisation by hyaluronidases, and interactions with HA-binding proteins. HA synthesis and degradation are closely regulated in skeletal tissues and aberrant synthetic or degradative activity causes disease. The role and regulation of HA synthesis and degradation in cartilage, bone and skeletal development is discussed.

MMP and non-MMP-mediated release of aggrecan and its fragments from articular cartilage: a comparative study of three different aggrecan and glycosaminoglycan assays.

OBJECTIVE: Aggrecan is the major proteoglycan in articular cartilage and is known to be degraded by various proteases, including matrix metalloproteinases (MMPs). The present study was undertaken to develop immunoassays detecting aggrecan and its fragments generated by MMP and non-MMP-mediated proteolysis. METHODS: Two immunoassays were developed: (1) the G1/G2 sandwich assay employing a monoclonal antibody (F-78) both as a capturing and a detecting antibody, and (2) the 342-G2 sandwich assay substituting the capturing antibody in the G1/G2 test with a monoclonal antibody, AF-28 recognizing the 342FFGVG neo-epitope generated by MMP cleavage. These assays were compared to the commercially available glycosaminoglycan (GAG) assay. RESULTS: In supernatants of Oncostatin M and Tumor Necrosis Factor alpha (OSM/TNFalpha) stimulated explants, high levels of G1/G2 fragments and GAGs were released in the initial phase (days 2-5), followed by low levels in the intermediate (days 9-12) and late phase (days 12-21). MMP-generated fragments were detected in the late phase only. In the presence of the general MMP inhibitor GM6001, 342-G2 was not detected, whereas the G1/G2 profile remained virtually unchanged. In patients with rheumatoid arthritis (RA), the release of G1/G2 molecules was decreased (27.3%), and that of the 342-G2 fragments increased compared to healthy controls (33.3%). CONCLUSION: The stimulation of bovine articular cartilage explants with OSM/TNFalpha released aggrecan fragments both in an MMP and non-MMP-mediated route. These immunoassays carry a potential as diagnostic tools for the quantitative assessment of the cartilage turnover in RA patients in addition to their utility in ex vivo explant cultures.

ADAMTS5-mediated aggrecanolysis in murine epiphyseal chondrocyte cultures.

OBJECTIVE: Aggrecan degradation by aggrecanases [a disintegrin and metalloproteinase with thrombospondin-like motifs (ADAMTS) 1, 4, 5, 8, 9, 15] is considered to initiate much of the cartilage pathology seen in human arthritis, however, the proteinase responsible and its mode of control is unclear. The present work was done to examine mechanisms of aggrecanase control in a novel murine epiphyseal cell system and to determine whether ADAMTS5 alone is responsible for aggrecanolysis by these cells. METHODS: Epiphyseal cells from 4-day-old mice (wild type, TS-5 (-/-), CD44(-/-), syndecan-1(-/-), membrane type-4 matrix metalloproteinase [MT4MMP(-/-)]) were maintained in non-adherent aggregate cultures and aggrecanolysis studied by biochemical and histochemical methods. Confocal immunolocalization analyses were done with specific probes for ADAMTS5, hyaluronan (HA) and aggrecanase-generated fragments of aggrecan. RESULTS: Aggrecanolysis by these cells was specifically aggrecanase-mediated and it occurred spontaneously without the need for addition of catabolic stimulators. Chondrocytes from ADAMTS5-null mice were aggrecanase-inactive whereas all other mutant cells behaved as wild type in this regard suggesting that ADAMTS5 activity is not controlled by CD44, syndecan-1 or MT4MMP in this system. Immunohistochemical analysis supported the central role for ADAMTS5 in the degradative pathway and indicated that aggrecanolysis occurs primarily in the HA-poor pericellular region in these cultures. CONCLUSION: These findings are consistent with published in vivo studies showing that single-gene ADAMTS5 ablation confers significant protection on cartilage in murine arthritis. We propose that this culture system and the analytical approaches described provide a valuable framework to further delineate the expression, activity and control of ADAMTS-mediated aggrecanolysis in human arthritis.

Perlecan from human epithelial cells is a hybrid heparan/chondroitin/keratan sulfate proteoglycan.

Perlecan is a multidomain proteoglycan, usually substituted with heparan sulphate (HS), and sometimes substituted with both HS and chondroitin sulphate (CS). In this paper, we describe perlecan purified from HEK-293 cells substituted with HS, CS and keratan sulphate (KS). KS substitution was confirmed by immunoreactivity with antibody 5D4, sensitivity to keratanase treatment, and fluorophore-assisted carbohydrate electrophoresis. HEK-293 perlecan failed to promote FGF-dependent cell growth in an in vitro assay. This study is the first to report perlecan containing KS, and makes perlecan one of only a very few proteoglycans substituted with three distinct types of glycosaminoglycan chains.

Neoepitopes as biomarkers of cartilage catabolism.

Progressive degradation of articular cartilage is a central feature of arthritis and a major determinant of long term joint dysfunction. There are no treatments able to halt the progression of cartilage destruction presently available, and monitoring the benefit of potential therapies is hampered by our inability to measure the "health" of articular cartilage. Serial radiographic assessment of joint space narrowing, the current gold standard, requires measurements over a prolonged time (1-5 years) and is prone to technical difficulties. Other strategies for evaluating cartilage degradation are needed to enable both short and long term monitoring of disease progression and response to therapy. One avenue that holds promise is the use of biomarkers that accurately reflect the degradative state of the articular cartilage. Antibodies that recognise terminal amino acid sequences generated by proteolysis at specific sites in the core protein of both aggrecan and type II collagen (neoepitope antibodies) have become available in recent years. These antibodies have been invaluable for identifying the proteinases responsible for cartilage breakdown both in vitro and in vivo. The presence of neoepitope sequences generated by specific metalloenzyme cleavage of aggrecan and type II collagen correlates well with the progression of cartilage degeneration, both in vitro and in mouse models of arthritis. Preliminary results with quantitative assays of type II collagen neoepitopes suggest that they may be useful markers of joint disease in humans. Long term studies correlating neoepitope concentration with clinical and radiographic disease are now required to validate the utility of neoepitopes as surrogate markers of cartilage degeneration and joint disease.

Immunolocalization of matrix metalloproteinases in partial-thickness defects in pig articular cartilage. A preliminary report.

BACKGROUND: Partial-thickness defects in mature articular cartilage do not heal spontaneously. Attempts at repair often result in limited integration between the repair tissue and the surrounding cartilage, with formation of chondrocyte clusters adjacent to a zone of cartilage necrosis. In wound repair, spatially and temporally controlled expression of matrix metalloproteinases and their inhibitors have been implicated in proteolytic degradation of damaged extracellular matrix components, but the sequence of events following damage to cartilage is unknown. To determine this sequence, we studied the distribution of matrix metalloproteinases and their inhibitors during early in vivo repair of partial-thickness defects in pig articular cartilage. METHODS: With use of a model that elicits the ingrowth of mesenchymal cells into partial-thickness defects, partial-thickness defects were created in knee joint cartilage. The distributions of matrix metalloproteinase-1, 2, 3, 9, 13, and 14; tissue inhibitors of metalloproteinase-1 and 2; and the neoepitope DIPEN341 specifically generated following matrix metalloproteinase cleavage of aggrecan were determined by immunolocalization of repair tissue and surrounding cartilage excised from immature pigs during the first eight weeks of repair and from adult minipigs at eight days and three weeks. RESULTS: Synthesis of matrix metalloproteinase-13 was usually confined to hypertrophic chondrocytes in immature cartilage and to the radial zone in adult cartilage. Following injury, strong induction of matrix metalloproteinase-13 synthesis was observed in chondrocyte clusters surrounding lesions in all of the animals. The migration of macrophages into defects was prominent at two and eight days, with synthesis and deposition of matrix metalloproteinase-9 onto damaged cartilage matrix and newly synthesized matrix in the defect. The DIPEN341 neoepitope was localized to damaged cartilage matrix at eight days and six weeks, indicating partial degradation of aggrecan. Focal synthesis of matrix metalloproteinase-1, 3, and 14 and of tissue inhibitor of metalloproteinase-1 occurred at later times, suggesting continuous remodeling of the increasingly compact repair tissue. CONCLUSIONS: The expression of matrix metalloproteinase-13 by normal hypertrophic chondrocytes and the induction of synthesis in chondrocyte clusters adjacent to the zone of cartilage necrosis suggest that this enzyme participates in the pericellular proteolysis required for lacunar expansion. The localization of matrix metalloproteinase-9 to damaged cartilage matrix suggested matrix proteolysis, which was confirmed with DIPEN341 localization. Reduced matrix metachromasia persisted and was colocalized with DIPEN341 at six weeks. However, under the conditions investigated, there was only limited proteolytic degradation in the zone of cartilage necrosis. This may render the zone mechanically weakened, thereby contributing to subsequent instability of the region, and may form a barrier to integration of repair tissue with viable cartilage. CLINICAL RELEVANCE: Osteoarthritis initially involves the superficial layers of cartilage. The development of procedures to promote the healing or repair of early defects will have major advantages in terms of disease alleviation as well as economic importance. Identification of the enzymes involved in the early repair of partial-thickness defects in articular cartilage is clinically relevant because proteolysis of damaged matrix has to take place in order for repair tissue to integrate with surrounding healthy cartilage.

Mutations in the interglobular domain of aggrecan alter matrix metalloproteinase and aggrecanase cleavage patterns. Evidence that matrix metalloproteinase cleavage interferes with aggrecanase activity.

We have expressed G1-G2 mutants with amino acid changes at the DIPEN(341) downward arrow(342)FFGVG and ITEGE(373) downward arrow(374)ARGSV cleavage sites, in order to investigate the relationship between matrix metalloproteinase (MMP) and aggrecanase activities in the interglobular domain (IGD) of aggrecan. The mutation DIPEN(341) to DIGSA(341) partially blocked cleavage by MMP-13 and MMP-8 at the MMP site, while the mutation (342)FFGVG to (342)GTRVG completely blocked cleavage at this site by MMP-1, -2, -3, -7, -8, -9, -13, -14. Each of the MMP cleavage site mutants, including a four-amino acid deletion mutant lacking residues ENFF(343), were efficiently cleaved by aggrecanase, suggesting that the primary sequence at the MMP site had no effect on aggrecanase activity in the IGD. The mutation (374)ARGSV to (374)NVYSV completely blocked cleavage at the aggrecanase site by aggrecanase, MMP-8 and atrolysin C but had no effect on the ability of MMP-8 and MMP-13 to cleave at the Asn(341) downward arrowPhe bond. Susceptibility to atrolysin C cleavage at the MMP site was conferred in the DIGSA(341) mutant but absent in the wild-type, (342)GTRVG, (374)NVYSV, and deletion mutants. To further explore the relationship between MMP and aggrecanase activities, sequential digest experiments were done in which MMP degradation products were subsequently digested with aggrecanase and vice versa. Aggrecanase-derived G1 domains with ITEGE(373) C termini were viable substrates for MMPs; however, MMP-derived G2 fragments were resistant to cleavage by aggrecanase. A 10-mer peptide FVDIPENFFG, which is a substrate analogue for the MMP cleavage site, inhibited aggrecanase cleavage at the Glu(373) downward arrowAla bond. This study demonstrates that MMPs and aggrecanase have unique substrate recognition in the IGD of aggrecan and suggests that sequences at the C terminus of the DIPEN(341) G1 domain may be important for regulating aggrecanase cleavage.

Matrix metalloproteinases 19 and 20 cleave aggrecan and cartilage oligomeric matrix protein (COMP).

Matrix metalloproteinase (MMP)-19 and MMP-20 (enamelysin) are two recently discovered members of the MMP family. These enzymes are involved in the degradation of the various components of the extracellular matrix (ECM) during development, haemostasis and pathological conditions. Whereas MMP-19 mRNA is found widely expressed in body tissues, including the synovium of normal and rheumatoid arthritic patients, MMP-20 expression is restricted to the enamel organ. In this study we investigated the ability of MMP-19 and MMP-20 to cleave two of the macromolecules characterising the cartilage ECM, namely aggrecan and the cartilage oligomeric matrix protein (COMP). Both MMPs hydrolysed aggrecan efficiently at the well-described MMP cleavage site between residues Asn(341) and Phe(342), as shown by Western blotting using neo-epitope antibodies. Furthermore, the two enzymes cleaved COMP in a distinctive manner, generating a major proteolytic product of 60 kDa. Our results suggest that MMP-19 may participate in the degradation of aggrecan and COMP in arthritic disease, whereas MMP-20, due to its unique expression pattern, may primarily be involved in the turnover of these molecules during tooth development.

Generation and novel distribution of matrix metalloproteinase-derived aggrecan fragments in porcine cartilage explants.

We have studied aggrecan catabolism mediated by matrix metalloproteinases (MMPs) in a porcine cartilage culture system. Using antibodies specific for DIPEN(341) and (342)FFGVG neoepitopes, we have detected MMP-derived fragments in conditioned medium and cultured cartilage, by radioimmunoassay, Western blotting, and immunolocalization. Radioimmunoassay revealed that the amount (pmol of epitope/mg of total glycosaminoglycan) of (342)FFGVG epitope released from cartilage remained constant over a 5-day culture period and was not increased by IL-1alpha or retinoate. However, the proportion (pmol of epitope/mg of released glycosaminoglycan) of (342)FFGVG epitope released was decreased upon stimulation, consistent with the involvement of a non-MMP proteinase, such as aggrecanase. The data suggest that in vitro MMPs may be involved in the base-line catabolism of aggrecan. Immunolocalization experiments showed that DIPEN(341) and ITEGE(373) epitopes were increased by treatment with IL-1alpha and retinoate. Confocal microscopy revealed that ITEGE(373) epitope was largely intracellular but with matrix staining in the superficial zone, whereas DIPEN(341) epitope was cell-associated and widely distributed in the matrix. Surprisingly, the majority of (342)FFGVG epitope, determined by radioimmunoassay and Western blotting, was retained in the tissue despite the absence of a G1 domain anchor. Interleukin-1alpha stimulation caused a marked increase in tissue DIPEN(341) and (342)FFGVG epitope, and the (342)FFGVG fragments retained in the tissue were larger than those released into the medium. Active porcine aggrecanase was unable to cleave (342)FFGVG fragments at the downward arrowGlu(373) downward arrowAla(374) bond but cleaved intact aggrecan at this site, suggesting that (342)FFGVG fragments are not substrates for aggrecanase. The apparent retention of large (342)FFGVG fragments within cartilage, and their resistance to N-terminal cleavage by aggrecanase suggests that (342)FF6V6 fragments may have a role in cartilage homeostasis.

Recombinant human aggrecan G1-G2 exhibits native binding properties and substrate specificity for matrix metalloproteinases and aggrecanase.

A recombinant human aggrecan G1-G2 fragment comprising amino acids Val(1)-Arg(656) has been expressed in Sf21 cells using a baculovirus expression system. The recombinant G1-G2 (rG1-G2) was purified to homogeneity by hyaluronan-Sepharose affinity chromatography followed by high performance liquid chromatography gel filtration, and gave a single band of M(r) 90,000-95,000 by silver stain or immunoblotting with monoclonal antibody 1-C-6. The expressed G1-G2 bound to both hyaluronan and link protein indicating that the immunoglobulin-fold motif and proteoglycan tandem repeat loops of the G1 domain were correctly folded. Further analysis of secondary structure by rotary shadowing electron microscopy confirmed a double globe appearance, but revealed that the rG1-G2 was more compact than its native counterpart. The size of rG1-G2 by SDS-polyacrylamide gel electorphoresis was unchanged following digestion with keratanase and keratanase II and reduced by only 2-5 kDa following digestion with either O-glycosidase or N-glycosidase F. Recombinant G1-G2 was digested with purified matrix metalloproteinases (MMP), isolated aggrecanase, purified atrolysin C, or proteinases present in conditioned medium from cartilage explant cultures, and the products analyzed on SDS gels by silver stain and immunoblotting. Neoepitope antibodies recognizing the N-terminal F(342)FGVG or C-terminal DIPEN(341) sequences were used to confirm MMP cleavage at the Asn(341) downward arrow Phe bond, while neoepitope antibodies recognizing the N-terminal A(374)RGSV or C-terminal ITEGE(373) sequences were used to confirm aggrecanase cleavage at the Glu(373) downward arrow Ala bond. Cleavage at the authentic MMP and aggrecanase sites revealed that these proteinases have the same specificity for rG1-G2 as for native aggrecan. Incubation of rG1-G2 with conditioned medium from porcine cartilage cultures revealed that active soluble aggrecanase but no active MMPs, was released following stimulation with interleukin-1alpha or retinoic acid. Atrolysin C, which cleaves native bovine aggrecan at both the aggrecanase and MMP sites, efficiently cleaved rG1-G2 at the aggrecanase site but failed to cleave at the MMP site. In contrast, native glycosylated G1-G2 with or without keratanase treatment was cleaved by atrolysin C at both the aggrecanase and MMP sites. The results suggest that the presence or absence per se of keratan sulfate on native G1-G2 does not affect the activity of atrolysin C toward the two sites.