Retinoic acid treatment elevates matrix metalloproteinase-2 protein and mRNA levels in avian growth plate chondrocyte cultures

Matrix metalloproteinases (MMPs) play a crucial role in tissue remodeling. In growth plate (GP) cartilage, extensive remodeling occurs at the calcification front. To study the potential involvement of MMPs in retinoic acid (RA) regulation of skeletal development, we studied the effect of all-trans-RA on MMPs levels in mineralizing chicken epiphyseal chondrocyte primary cultures. When treated for 4 day periods on days 10 and 17, RA increased levels of an ∼70 kDa gelatinase activity. The N-terminal sequence of the first 20 amino acid residues of the purified enzyme was identical to that deduced from chicken MMP-2 cDNA. Time-course studies indicated that RA elevated MMP-2 activity levels in the cultures within 16 h. This increase was inhibited by cycloheximide and was enhanced by forskolin. The increase in MMP-2 activity induced by RA was accompanied by an increase in MMP-2 mRNA levels and was abolished by treatment with cycloheximide. This upregulation of MMP levels by RA in GP chondrocytes is consistent with its effects on osteoblasts and osteosarcoma cells and opposite its inhibitory effects on fibroblasts and endothelial cells. It may well be related to the breakdown of the extracellular matrix in the GP and would be governed by the availability of RA at the calcification front where extensive vascularization also occurs. J. Cell. Biochem. 68:90–99, 1998. © 1998 Wiley-Liss, Inc.     

Age-related changes in the expression of gelatinase and tissue inhibitor of metalloproteinase genes in mandibular condylar,growth plate,and articular cartilage in rats

Summary Mandibular condylar cartilage acts as both articular and growth plate cartilage during growth, and then becomes articular cartilage after growth is complete. Cartilaginous extracellular matrix is remodeled continuously via a combination of production, degradation by matrix metalloproteinases (MMPs), and inhibition of MMP activity by tissue inhibitors of metalloproteinases (TIMPs). This study attempted to clarify the age-related changes in the mRNA expression patterns of MMP-2, MMP-9, TIMP-1, TIMP-2, and TIMP-3 in mandibular condylar cartilage in comparison to tibial growth plate and articular cartilage using an in situ hybridization method in growing and adult rats. MMP-2 and MMP-9 were expressed in a wide range of condylar cartilage cells during growth, and their expression domains became limited to mature chondrocytes in adults. The patterns of TIMP-1 and TIMP-2 expression were similar to those of MMP-2 and MMP-9 during growth, and were maintained until adulthood. TIMP-3 was localized to hypertrophic chondrocytes throughout the growth stage. Therefore, we concluded that TIMP-1 and TIMP-2 were general inhibitors of MMP-2 and MMP-9 in condylar cartilage, while TIMP-3 regulates the collagenolytic degradation of the hypertrophic cartilage matrix.     

Co-localization of integrins and matrix metalloproteinases in the extracellular matrix of chondrocyte cultures

Beta1-integrins were found in the cartilage matrix, suggesting their implication in the assembly of its architectural scaffold, but the mechanism for this event is not yet clear. Matrix metalloproteinases (MMPs) may be involved in an integrin-shedding mechanism and matrix beta1-integrins may act to alter MMP activity. To begin to address this question, this study was designed to determine whether beta1-integrins and MMPs are colocalized in the chondrocytes or in the extracellular matrix of cartilage. We investigated high-density cultures of limb buds of 12-day-old mouse embryos by double immunofluorescence, immunoelectron microscopy and by coimmunoprecipitation assays in order to examine the localization of beta1-integrins and matrix metalloproteinases (MMP-1, MMP-3 and MMP-9) in cartilage. It was found, that all investigated MMPs and beta1-integrins were specifically co-localized in high-density cartilage cultures. Immunogold and immunofluorescence labelling of both beta1-integrins and MMPs were observed not only at the surface of chondrocytes but mainly also in the pericellular space and distributed between collagen fibrils in the extracellular matrix (ECM) as well. Results of immunoprecipitation experiments suggest a functional association of MMPs and beta1-integrins in chondrocytes as already described for other cell types. Further investigations are needed to elucidate the functional association between beta1-integrins and MMPs in chondrocytes.     

Recombinant-derived chicken growth hormone used for radioimmunoassay

The use of recombinant-derived chicken growth hormone (rcGH) in an avian growth hormone (GH) radioimmunoassay (RIA) procedure is described. Antiserum to turkey GH bound 125I-labeled rcGH, and unlabeled rcGH or turkey GH displaced binding in a dose-related manner. The dose-response curves of sera and pituitary extract from chickens and turkeys were parallel to the rcGH standard curve. Sera from hypophysectomized (hypox) chickens and turkeys produced no dose-response and did not inhibit binding of labeled rcGH. Recovery of rcGH added to hypox sera was quantitative. Modification of the homologous turkey GH RIA protocol of Proudman and Wentworth (1) to use rcGH made possible either an increase in assay sensitivity or a 3-day reduction in incubation time.     

Contributions of matrix metalloproteinases toward Meckel’s cartilage resorption in mice: immunohistochemical studies,including comparisons with developing endochondral bones

The middle portion of Meckel’s cartilage (one of four portions that disappear with unique fate) degrades via hypertrophy and the cell death of chondrocytes and via the resorption of cartilage by chondroclasts. We have examined the immunolocalization of matrix metalloproteinase-2 (MMP-2), MMP-9, MMP-13, and MMP-14 (members of the MMP activation cascade) and galectin-3 (an endogenous substrate for MMP-9 and an anti-apoptotic factor) during resorption of Meckel’s cartilage in embryonic mice and have compared the results with those of developing endochondral bones in hind limbs. MMP immunoreactivity, except for MMP-2, is present in nearly all chondrocytes in the middle portion of Meckel’s cartilage. On embryonic day 15 (E15), faint MMP-2-immunoreactive and intense MMP-13-immunoreactive signals occur in the periosteal bone matrix deposited by periosteal osteoblasts on the lateral surface, whereas MMP-9 and MMP-14 are immunolocalized in the peripheral chondrocytes of Meckel’s cartilage. The activation cascade of MMPs by face-to-face cross-talk between cells may thus contribute to the initiation of Meckel’s cartilage degradation. On E16, immunopositive signaling for MMP-13 is detectable in the ruffled border of chondroclasts at the resorption front, whereas immunostaining for galectin-3 is present at all stages of chondrocyte differentiation, especially in hypertrophic chondrocytes adjacent to chondroclasts. Galectin-3-positive hypertrophic chondrocytes may therefore coordinate the resorption of calcified cartilage through cell-to-cell contact with chondroclasts. In metatarsal specimens from E16, MMPs are detected in osteoblasts, young osteocytes, and the bone matrix of the periosteal envelope, whereas galectin-3 immunoreactivity is intense in young periosteal osteocytes. In addition, intense MMP-9 and MMP-14 immunostaining has been preferentially found in pre-hypertrophic chondrocytes, although galectin-3 immunoreactivity markedly decreases in hypertrophic chondrocytes. These results indicate that the degradation of Meckel’s cartilage involves an activation cascade of MMPs that differs from that in endochondral bone formation.     

Comparison of hepatic drug metabolizing enzyme activities in several agricultural species

1. Several pathways of drug metabolizing enzyme activity were measured in hepatic fractions of cattle, sheep, goats, chickens, turkeys, ducks, rabbits and rats. The pathways examined included the O-demethylation of p-nitrophenol, microsomal ester hydrolysis of procaine and glucuronidation of p-nitrophenol, and the cytosolic acetylation of sulfamethazine and sulfation of 2-naphthol. 2. For most enzymatic pathways measured, goats were more similar to sheep (wether) than to cattle (steers). The exception was UDP-glucuronyltransferase activity, which was significantly higher for the goat than for any other species studied. 3. Within the avian subset, the chicken and turkey were usually the most similar species. 4. The activities of arylsulfotransferase isozymes III and IV were particularly low for the duck compared to the chicken and turkey. 5. N-acetyltransferase activity was very high for rabbits and very low for sheep and goats.     

Galectin-3 is a downstream regulator of matrix metalloproteinase-9 function during endochondral bone formation

Endochondral bone formation is characterized by the progressive replacement of a cartilage anlagen by bone at the growth plate with a tight balance between the rates of chondrocyte proliferation, differentiation, and cell death. Deficiency of matrix metalloproteinase-9 (MMP-9) leads to an accumulation of late hypertrophic chondrocytes. We found that galectin-3, an in vitro substrate of MMP-9, accumulates in the late hypertrophic chondrocytes and their surrounding extracellular matrix in the expanded hypertrophic cartilage zone. Treatment of wild-type embryonic metatarsals in culture with full-length galectin-3, but not galectin-3 cleaved by MMP-9, mimicked the embryonic phenotype of Mmp-9 null mice, with an increased hypertrophic zone and decreased osteoclast recruitment. These results indicate that extracellular galectin-3 could be an endogenous substrate of MMP-9 that acts downstream to regulate hypertrophic chondrocyte death and osteoclast recruitment during endochondral bone formation. Thus, the disruption of growth plate homeostasis in Mmp-9 null mice links galectin-3 and MMP-9 in the regulation of the clearance of late chondrocytes through regulation of their terminal differentiation.     

Allantoin,the oxidation product of uric acid is present in chicken and turkey plasma

Urate oxidase is not present in birds yet allantoin, a product of this enzyme, has been measured in birds. Studies were designed to compare the concentrations of plasma purine derivatives in chickens and turkeys fed inosine-supplemented diets. The first study consisted of 12 male chicks that were fed diets supplemented with 0.6 mol inosine or hypoxanthine per kilogram diet from 3- to 6-week-old. Study 2 consisted of 12 turkey poults (toms) fed inosine-supplemented diets (0.7 mol/kg) from 6- to 8-week-old. Plasma allantoin and oxypurines concentrations were measured weekly using high performance liquid chromatography. Plasma uric acid (PUA) in chickens fed inosine-supplemented diets increased from 0.31 to 1.34 mM (P<0.05) at the end of week 2. In turkeys, those fed control diet had 0.17 mM PUA concentration compared to 0.3 mM in those fed the inosine diet at week 2 (P<0.05). Allantoin concentration increased in chickens from week 1 to 2 while a decrease was observed in turkeys (P<0.005) for both treatments. These data show that allantoin is present in turkey and chicken plasma. The presence of allantoin in avian plasma is consistent with uric acid acting as an antioxidant in these species.     

Biliary bile acid profiles of domestic fowl as determined by high performance liquid chromatography and fast atom bombardment mass spectrometry

1. The biliary bile acid profiles of domestic chickens (Gallus domesticus), turkeys (Meleagris gallopavo), and ducks (Anas platyrhynchos) were determined by high performance liquid chromatography (HPLC) and fast atom bombardment mass spectrometry (FAB-MS). 2. Chenodeoxycholyltaurine and cholyltaurine were the predominant bile acids in chicken and turkey bile, whereas duck bile contained primarily chenodeoxycholyltaurine and phocaecholyltaurine. 3. Allocholyltaurine was also detected in chicken and turkey bile, but not in duck bile. 4. FAB-MS analyses of individual HPLC peak fractions from chicken and duck bile extracts confirmed the presence of either taurine-conjugated dihydroxy- or trihydroxycholanoates. 5. Direct FAB-MS analyses of avian bile extracts not subjected to HPLC permitted a rapid assessment of the relative proportion of taurine-conjugated dihydroxy- to trihydroxycholanoates.     

Interleukin-1beta-induced expression of the urokinase-type plasminogen activator receptor and its co-localization with MMPs in human articular chondrocytes

The urokinase-type plasminogen activator receptor (uPAR) plays a critical role in cartilage degradation during osteoarthritis as it regulates pericellular proteolysis mediated by serine proteinases. Another important family of proteinases responsible for ECM destruction in arthritis are the matrix metalloproteinases (MMPs). MMPs are regulated by IL-1beta, a cytokine that plays a pivotal role in pathogenesis of osteoarthritis. This study was undertaken to address two questions: 1. Is uPAR-expression regulated by proinflammatory cytokines such as IL-1beta? 2. Does a functional co-localization exist between uPAR and MMPs? By immunohistochemical analysis we observed enhanced expression of uPAR on chondrocytes derived from osteoarthritic human cartilage compared to non-osteoarthritic controls. We found an IL-1beta-mediated expression of uPAR by immunoelectron microscopy. Western blot analysis demonstrated that IL-1beta-stimulated expression of uPAR on chondrocytes in vitro increased in a dose-dependent manner. Furthermore, we found a functional co-localization between uPAR and MMP-9 on IL-1beta-stimulated chondrocytes by means of a co-immunoprecipitation assay. Expression of uPAR in osteoarthritic cartilage but not in healthy cartilage suggests that uPAR plays a role in cartilage breakdown. We propose that uPAR-mediated effects e.g. pericellular proteolysis are one of other cytokine (IL-1beta)-mediated events that contribute to the pathogenesis of osteoarthritis. Furthermore, we found that MMPs and uPAR were part of the same cell surface complexes in chondrocytes. This finding underlines a functional interaction between MMPs and the serine proteinase system in the fine regulation of pericellular proteolysis. Interfering with uPAR signaling may present a novel target in arthritis therapy to prevent excessive proteolytic degradation.     

Pro-MMP-9 is a specific macrophage product and is activated by osteoarthritic chondrocytes via MMP-3 or a MT1-MMP/MMP-13 cascade

In joint diseases of both the inflammatory (rheumatoid arthritis, or RA) or the degenerative variety (osteoarthritis, or OA), matrix metalloproteinases (MMPs) are essential mediators of irreversible tissue destruction. MMP-9 is secreted as a stable, inactive zymogen and is proteolytically converted to the active enzyme. To understand the activation mechanism of MMP-9 in joint diseases, the process was investigated in serum-free cocultures of human articular chondrocytes and macrophages. Macrophages extensively expressed and secreted pro-MMP-9 whereas chondrocytes failed to produce the enzyme. However, efficient activation of pro-MMP-9 required soluble and membrane-associated chondrocyte proteinases. Two alternative activation pathways mainly involved MMPs and, marginally, serine or cysteine proteinases. MT1-MMP (MMP-14), the only MT-MMP expressed in chondrocytes, converted pro-MMP-13 which, in turn, cleaved pro-MMP-9. Alternatively, pro-MMP-9 was activated less efficiently by MMP-3, which was converted by autocatalysis or by serine or cysteine proteinases. Both pathways were triggered by chondrocytes from OA, but not normal joints. Therefore, articular chondrocytes are not innocent bystanders in joint diseases. They not only produce destructive enzymes guided by environmental cues but also they can instruct inflammatory cells or cells from surrounding tissues to do so by converting in several ways zymogens produced but not activated by these cells themselves.     

The effects of high magnitude cyclic tensile load on cartilage matrix metabolism in cultured chondrocytes

Excessive mechanical load is thought to be responsible for the onset of osteoarthrosis (OA), but the mechanisms of cartilage destruction caused by mechanical loads remain unknown. In this study we applied a high magnitude cyclic tensile load to cultured chondrocytes using a Flexercell strain unit, which produces a change in cell morphology from a polygonal to spindle-like shape, and examined the protein level of cartilage matrixes and the gene expression of matrix metalloproteinases (MMPs), tissue inhibitors of matrix metalloproteinases (TIMPs) and proinflammatory cytokines such as IL-1beta and TNF-alpha. Toluidine blue staining, type II collagen immunostaining, and an assay of the incorporation of [35S]sulfate into proteoglycans revealed a decrease in the level of cartilage-specific matrixes in chondrocyte cultures subjected to high magnitude cyclic tensile load. PCR-Southern blot analysis showed that the high magnitude cyclic tensile load increased the mRNA level of MMP-1, MMP-3, MMP-9, IL-1beta, TNF-alpha and TIMP-1 in the cultured chondrocytes, while the mRNA level of MMP-2 and TIMP-2 was unchanged. Moreover, the induction of MMP-1, MMP-3 and MMP-9 mRNA expression was observed in the presence of cycloheximide, an inhibitor of protein synthesis. These findings suggest that excessive mechanical load directly changes the metabolism of cartilage by reducing the matrix components and causing a quantitative imbalance between MMPs and TIMPs.     

Authentic matrix vesicles contain active metalloproteases (MMP). a role for matrix vesicle-associated MMP-13 in activation of transforming growth factor-beta

Matrix vesicles (MV) play a key role in the initiation of cartilage mineralization. Although many components in these microstructures have been identified, the specific function of each component is still poorly understood. In this study, we show that metalloproteases (MMP), MMP-2, -9, and -13 are associated with MV isolated from growth plate cartilage. In addition, we provide evidence that MV contain transforming growth factor-beta (TGF-beta) and that MV-associated MMP-13 is capable of activating latent TGF-beta. To determine whether MMPs are associated directly with MV, vesicles isolated from growth plate cartilage were sequentially treated with hyaluronidase, NaCl, and bacterial collagenase to remove matrix proteins and other components attached to their outer surface. Finally, the vesicles were incubated with detergent to rupture the MV membrane and expose components that are inside the vesicles. Each treated MV fraction was subjected to substrate zymography, immunoblotting, and substrate activity assay. Whereas active MMP-13 was lost after combined treatment with hyaluronidase and NaCl, MMP-2 and -9 activities were still retained in the pellet fraction even after detergent treatment, suggesting that the gelatinases, MMP-2 and -9, are integral components of MV. In addition, MV contain TGF-beta in the small latent complex, and MMP-13 associated with the MV surface was responsible for activation of TGF-beta. Since the amount of TGF-beta activated by hypertrophic chondrocytes increased with mineral appearance in serum-free chondrocyte cultures, a role for active MV-associated MMPs is suggested in activation of TGF-beta seen during late chondrocyte hypertrophy and mineralization of growth plate cartilage.     

Interleukin-4 antagonizes oncostatin M and transforming growth factor beta-induced responses in articular chondrocytes

Oncostatin M (OSM) stimulates cartilage degradation in rheumatoid arthritis (RA) by inducing matrix metalloproteinases (MMPs) and aggrecanases (ADAMTS; a disintegrin and metalloproteinase with thrombospondin motif). Transforming growth factor beta (TGF-beta1) induces cartilage repair in joints but in excessive amounts, promotes inflammation. OSM and TGF-beta1 also induce tissue inhibitor of metalloproteinase-3 (TIMP-3), an important natural inhibitor of MMPs, aggrecanases, and tumor necrosis factor alpha converting enzyme (TACE), the principal proteases involved in arthritic inflammation and cartilage degradation. We studied cartilage protective mechanisms of the antiinflammatory cytokine, interleukin-4 (IL-4). IL-4 strongly (MMP-13 and TIMP-3) or minimally (ADAMTS-4) suppressed OSM-induced gene expression in chondrocytes. IL-4 did not affect OSM-stimulated phosphorylation of extracellular signal-regulated kinases (ERKs), protein 38 (p38), c-Jun N-terminal kinase (JNK) and Stat1. Lack of additional suppression with their inhibitors suggested that MMP-13, ADAMTS-4, and TIMP-3 inhibition was independent of these mediators. IL-4 also downregulated TGF-beta1-induced TIMP-3 gene expression, Smad2, and JNK phosphorylation. Additional suppression of TIMP-3 RNA by JNK inhibitor suggests JNK implication. The cartilage protective effects of IL-4 in animal models of arthritis may be due to its inhibition of MMPs and ADAMTS-4 expression. However, suppression of TIMP-3 suggests caution for using IL-4 as a cartilage protective therapy.     

Chondrocytes isolated from tibial dyschondroplasia lesions and articular cartilage revert to a growth plate-like phenotype when cultured in vitro

We report here a comparative study of the development and behavior of chondrocytes isolated from normal growth plate tissue, tibial dyschondroplasic lesions, and from articular cartilage. The objective of these studies was to determine whether the properties exhibited by chondrocytes in dysplasic lesions or in articular cartilage were due to their cellular phenotype, their environment, or both. We had previously analyzed the electrolytes and amino acid levels in the extracellular fluid of avian growth plate chondrocytes. Using these data, we constructed a culture medium (DATP5) in which growth plate cells essentially recapitulate their normal behavior in vivo. Here, we used DATP5 to examine the behavior of chondrocytes isolated from lesions of tibial dyschondroplasia (TD). We found that once isolated from lesion and grown in this supportive medium, dysplasic chondrocytes behaved essentially like normal growth plate cells. These findings suggest that the cause of TD is local factors operating in vivo to prevent these cells from developing normally. With respect to articular chondrocytes, our data indicate that they more closely retain normal protein and proteoglycan synthesis when grown in serum-free media. These cells readily induced mineral formation in vitro, both in the presence and absence of serum. However, in serum-containing media, mineralization was significantly enhanced when the cells were exposed to retinoic acid (RA) or osteogenic protein-1 (OP-1). Our studies support previous work indicating the presence of autocrine factors produced by articular chondrocytes in vivo that prevent mineralization and preserve matrix integrity. The lack of inhibitory factors and the presence of supporting factors are likely reasons for the induction of mineralization by articular chondrocytes in vitro.     

Contrasting expression of membrane metalloproteinases, MT1-MMP and MT3-MMP, suggests distinct functions in skeletal development

Membrane-type 1 matrix metalloproteinase (MT1-MMP) is the most ubiquitous and widely studied of the membrane-type metalloproteinases (MT-MMPs). It was thus surprising to find no published data on chicken MT1-MMP. We report here the characterization of the chicken gene. Its low sequence identity with the MT1-MMP genes of other species, high GC content, and divergent catalytic domain explains the absence of data and our difficulties in characterizing the gene. The absence of structural features in the chicken gene that have been suggested to be critical for the activation of MMP-2 by MT1-MMP; for the effect of MT1-MMP on cell migration and for the recycling of MT1-MMP suggest these features are either not essential or that MT1-MMP does not perform these functions in chickens. Comparison of the expression of chicken MT1-MMP with MT3-MMP and with MMP-2 and MMP-13 has confirmed the previously recognized co-expression of MT1-MMP with MMP-2 and MMP-13 in fibrous and vascular tissues, particularly those surrounding the developing long bones in other species. By contrast, MT3-MMP expression differs markedly from that of MT1-MMP and of both MMP-2 and MMP-13. MT3-MMP is expressed by chondrocytes of the developing articular surface. Similar expression patterns of this group of MT-MMPs and MMPs have been observed in mouse embryos and suggest distinct and specific functions for MT1-MMP and MT3-MMP in skeletal development.     

Induction of matrix metalloproteinase MMP-9 (92-kDa gelatinase) by retinoic acid in human neuroblastoma SKNBE cells: relevance to neuronal differentiation

Retinoic acid (RA) has been shown to induce human neuroblastoma SKNBE cell differentiation into a neuronal phenotype. Whether this neuronal differentiation is associated with modulation of matrix gelatinase [matrix metalloproteinase (MMP)-2 and MMP-9] expression was investigated in SKNBE cell cultures exposed to RA for 14 days. Their differentiation into a neuronal phenotype was typified by neural cell adhesion molecule and growth-associated protein-43 expression. Gelatinase expression was assessed by gel zymography, quantitative RT-PCR, and immunocytochemistry. Neuronal markers were located in neurites and ganglion-like clusters of neuronal cells induced upon RA exposure. MMP-2 expression was constitutive and remained unchanged at both the mRNA and protein levels in response to RA, tumor necrosis factor-alpha (TNFalpha), or phorbol 12-myristate 13-acetate (PMA) treatment. In contrast, MMP-9 was inducible by RA, TNFalpha, or PMA. MMP-9 was progressively enhanced by RA as a function of time exposure until day 14. The addition of TNFalpha or PMA potentiated RA-induced MMP-9 expression with a synergic maximal effect at day 14 of RA exposure. Immunoreactive MMP-9 was located early in outgrowing neurites, but only at day 14 of RA exposure in extensive neuritic networks. Taken together, the correlation between the MMP-9 expression by SKNBE cells and the time scale of their differentiation into a neuronal phenotype allowed us to propose that MMP-9 could participate in the neurite growth process and cell migration and organization into ganglion-like clusters.     

Induction of CD44 and MMP expression by hyaluronidase treatment of articular chondrocytes

In this study, the effects of fragmentation of the glycosoaminoglycans of the cell-associated matrix by hyaluronidase (HAase) on the expression of CD44 receptor and matrix metalloproteinase (MMP) mRNAs in cultured articular chondrocytes were examined. Chondrocytes, isolated from rabbit and bovine articular cartilage, were treated with bovine testicular HAase (0-200 units/ml) in the presence or absence of an antibody for CD44. The mRNA levels of CD44, CD44 variant (CD44v), MMPs (MMP-1, -3 and -9), and tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2) were determined by RT-PCR. The treatment of cultured chondrocytes with HAase resulted in the production of low molecular weight fragments of hyaluronan (HA). The expression of CD44, CD44v and MMP (MMP-1, -3 and -9) mRNAs, but not TIMP-1 or TIMP-2 mRNA, was up-regulated in the cultures treated with HAase, whereas this expression was not affected by treatment with purified HA of 1.0 x 10(5) Da. Furthermore, the induction of CD44 and MMPs on treatment with HAase was suppressed by an anti-CD44 antibody. The results suggest that the fragmentation of HA may lead to cartilage destruction in terms of the enhanced expression of MMPs as well as the upregulation of CD44.