Proteoglycan-degrading enzymes of rabbit fibroblasts and granulocytes.

A neutral proteinase secreted by rabbit synovial fibroblasts in parallel with specific collagenase was partially purified by ion-exchange chromatography. At pH 7.6 this proteinase degraded 35S-labelled bovine nasal proteoglycan and azo-casein. The enzymic activity was inhibited by EDTA, 1,10-phenanthroline and serum, whereas di-isopropyl phosphorofluoridate and soya-bean trypsin inhibitor had little effect. By gel filtration the apparent mol.wt. of the enzyme was 25000. The fibroblast neutral proteinase was compared with the proteoglycan-degrading neutral proteinases of rabbit polymorphonuclear-leucocyte granules. Two distinct activities were found in the granules: one was inhibited by soya-bean trypsin inhibitor and the other by EDTA. The proteoglycan-degrading proteinases of rabbit fibroblasts and polymorphonuclear leucocytes at acid pH also were examined. Both cathepsin D and a thiol-dependent proteinase contributed to the degradation of proteoglycan at pH 4.5.     

Proteoglycan- and collagen-degrading enzymes from human interleukin 1-stimulated chondrocytes from several species: proteoglycanase and collagenase inhibitors as potentially new disease-modifying antiarthritic agents

Human IL-1-stimulated chondrocytes derived from rabbit, bovine, and human articular cartilage produce proteoglycan- and collagen-degrading enzymes. These studies demonstrate that the biological activity of IL-1 is not species specific. Several thiol, carboxyalkyl, and hydroxamic acid peptide inhibitors showed differential effects. The thiols were equipotent inhibitors of both the collagen- and proteoglycan-degrading enzymes whereas the carboxyalkyls appear to inhibit solely the proteoglycan-degrading enzyme(s). The hydroxamic acid peptides, the most potent inhibitors, appear to be more active against the proteoglycan-degrading enzymes. These synthetic inhibitors of proteoglycan- and/or collagen-degrading enzymes may represent a new class of disease-modifying antiarthritic agents.     

Evidence for polymorphonuclear-leucocyte-derived proteinases in arthritic cartilage.

1. An enzyme that degrades proteoglycan at neutral pH was extracted with 4 M-guanidine hydrochloride from the articular cartilage of rabbits with antigen-induced arthritis. 2. The enzyme had an apparent molecular weight on Ultrogel AcA 54 of about 8000 and was optimally active at pH 7.5 in Tris/HCl buffer containing 0.2 M-NaCl. The partially purified preparation was totally inhibited by 0.01 mM-N-acetyldialanylprolylvalylchloromethane, severely inhibited by 2 mM-phenylmethanesulphonyl fluoride and soya-bean trypsin inhibitor (200 microgram/ml) and slightly inhibited by 10 mM-EDTA. Marked inhibition was also obtained with a cytosolic fraction prepared from rabbit polymorphonuclear leucocytes. 3. All properties of the enzyme were virtually identical with those of an 'elastase-like' proteinase that was isolated from rabbit polymorphonuclear-leucocyte granules. 4. The results are consistent with the idea that cartilage proteoglycan degradation in acute joint inflammation is due at least partly to the diffusion into the cartilage of proteinases derived from synovial-fluid polymorphonuclear leucocytes.     

Two latent metalloproteases of human articular cartilage that digest proteoglycan

Human articular cartilage contains very low levels of metalloprotease activity; the activity in 1 g of cartilage is approximately equivalent to the activity of 1 microgram of trypsin. Development of a sensitive assay, based on the digestion of radioactive proteoglycan, has made it possible to study protease activity in 1-2-g specimens of cartilage. Cartilage was extracted with Tris buffer in the cold and with Tris buffer containing 10 mM CaCl2 at 60 degrees C. The extracts were passed through Sepharose 6B; two major and two minor metalloprotease activities were detected. A neutral metalloprotease activity, pH optimum 7.4, was found as a latent form of Mr = 56,000. It could be activated with aminophenylmercuric acetate or trypsin with a resultant decrease of Mr to 40,000. An acid metalloprotease, pH optimum 5.3, also occurred as a latent form of Mr = 50,000. Activation converted this to Mr = 35,000. Removal of calcium ions by dialysis reduced the activity of the neutral enzyme by 80-85% and of the acid enzyme by 100%. Both activities were restored by 10 mM Ca2+. Both enzymes were completely inhibited by 1 mM o-phenanthroline in the presence of excess calcium. This inhibition was overcome by 1 mM Zn2+ and, to a lesser extent, by Co2+. These proteases may be important in the metabolism of the cartilage matrix and in its destruction in osteoarthritis.     

Coincubation of bovine synovial or capsular tissue with cartilage generates a soluble "Aggrecanase" activity

The culture of bovine synovial or capsular tissue generated proteoglycan-degrading activity. When these tissues were incubated with living or dead bovine articular cartilage significantly more proteoglycan-degrading activity was revealed. The activity was present in a soluble form and required protein synthesis for its generation. The conditioned medium did not contain matrixin activity, although experiments with proteinase inhibitors suggested that the activity was due to a metalloproteinase. Western blotting of the aggrecan fragments suggested cleavage of aggrecan within the interglobular domain at the "aggrecanase" site, but not at the major matrixin site. N-terminal sequencing confirmed cleavage of aggrecan at a number of glutamyl bonds, including the aggrecanase site in the interglobular domain. We conclude that cultured synovial or capsular tissue produces soluble aggrecanase and an enzyme which releases aggrecanase from cartilage, possibly by cleavage of a chondrocyte membrane-bound form of aggrecanase.     

Purification and characterization of an acid metalloproteinase from human articular cartilage

A metalloprotease that digests cartilage proteoglycan optimally at pH 5.3 has been purified (4400-fold) to homogeneity from 20-g samples of human articular cartilage containing about 100 micrograms of enzyme. This enzyme was cleanly separated from a related neutral metalloprotease with an optimum pH of 7.2. The acid metalloprotease displays 40% of its maximum activity at pH 7.2 and so has significant activity at physiological pH. The protease is calcium-dependent and indirect evidence suggests that it may contain zinc at its active center. It occurs largely in a latent form that can be activated by aminophenylmercuric acetate. The apparent Mr of the latent form is 55,000 and of the active form, 35,000. The isoelectric point is at pH 4.9. The protease activity is inhibited by chelators, Z-phenylalanine, ovostatin, and tissue inhibitor of metalloproteinase from human articular cartilage. It differs from metalloproteinases such as enkephalinase and kidney brush-border protease in its failure to be strongly inhibited by phosphoramidon and Zincov. It cleaves the proteoglycan monomer of bovine nasal cartilage to fragments of approximately 140,000 Da. It cleaves the B chain of insulin at Ala14-Leu15 and Tyr16-Leu17. A survey of 26 cartilage extracts indicates this enzyme is elevated to about 3 times the normal level in human osteoarthritic cartilage and that the tissue inhibitor of metalloproteinase is only slightly diminished. Preliminary evidence points to the presence of a similar acid metalloprotease activity in human polymorphonuclear leukocytes.     

Degradation of cartilage proteoglycan and collagen by synovial cells. Stimulation by macrophages under activation by phagocytosis, lymphocyte factors, bacterial products or other inflammatory stimuli

When cultured together with dead 35S-labelled cartilage discs or at the surface of [3H]proteoglycan/[14C]collagen-coated plates, synovial cells from either arthritic or normal rabbit joints digested both the proteoglycan and the collagen of the substrates after a lag-period of 1-2 days. These digestions were inversely related to the age (number of subculture passages) of the synovial cells and they could be modulated by serum components that were either inhibitory or stimulatory. They were dependent on a protein synthesis by the cells and were paralleled, in young cultures, by the release of collagenase and of a proteoglycan-degrading neutral proteinase. The co-culture of synovial cells with macrophages or their culture with macrophage-conditioned culture media caused a more rapid and more extensive degradation of collagen and proteoglycan due to the stimulation of the synovial cells by a nondialysable macrophage factor. The production of this synovial cell-activating 'matrix regulatory monokine' by the macrophage was enhanced by several immunological or inflammatory stimuli such as lymphocyte factors, phagocytosis, asbestos fibres, endotoxin, adjuvant muramyl dipeptide or chemotactic formyl-methionyl peptide, as well as by other membrane-active agents (phorbol myristate acetate, concanavalin A). It is presumed that these interactions are of importance in the development of cartilage destruction in rheumatoid and other chronic inflammatory arthritis.     

Characterization of cathepsins in cartilage

The presence of a cathepsin B-like enzyme in rabbit ear cartilage was established by the use of the synthetic substrates benzoyl-l-arginine amide and benzoyl-dl-arginine 2-naphthylamide. This was facilitated by using a technique that permits the incubation of a fixed weight of thin (18mu) cartilage sections with an appropriate exogenous substrate. The enzymic properties of cathepsin B in cartilage have been compared with an endogenous enzyme that liberates chondromucopeptide by degrading the cartilage matrix autocatalytically at pH5. Besides being maximally active at pH4.7, these cartilage enzymes are enhanced in activity by cysteine and inhibited by arginine analogues, iodoacetamide, chloroquine and mercuric chloride. They are not inhibited by EDTA, di-isopropyl phosphorofluoridate and diethyl p-nitrophenyl phosphate. When inhibiting the release of chondromucopeptide from cartilage at pH5, the arginine-containing synthetic substrates are hydrolysed simultaneously. These enzymes also share the same heat-inactivation characteristics at various pH values, being stable at acid pH and unstable at neutral and alkaline pH. The experimental evidence indicates that a cathepsin B-like enzyme may be partly responsible for the autolytic degradation of cartilage matrix at pH5.     

The electrophoretic heterogeneity of bovine nasal cartilage proteoglycans.

1. Proteoglycans were extracted from bovine nasal cartilage with 2.0M-CaC2 or with 0.15M-KCl followed by 2.0M-CaC2.. Proteoglycan fractions were prepared from the extracts by density-gradient centrifugation in CsCl under 'associative' and 'dissociative' conditions. 2. The heterogeneity of the proteoglycan fractions was investigated by large-pore-gel electrophoresis. It was concluded that extracts made with 2.0M-CaCl2 or sequential 2.0M-CaCl2 contain two major species of proteoglycan 'subunit' of different hydrodynamic size, together with proteoglycan aggregates. Both 'subunits' have mobilities that are greater than those of proteoglycans obtained from pig articular cartilage McDevitt & Muir (1971) Anal. Biochem. 44, 612-622] and are therefore probably smaller in size than the latter. 3. Proteoglycan fractions isolated from cartilage extracted lith 0.15M-KCl separated into two main components on large-pore-gel electrophoresis with mobilities greater than those of proteoglycans extracted with 2.0M-CaCl2. Proteoglycans extracted at low ionic strength from bovine nasal cartilage are of similar hydrodynamic size to those extracted from pig articular cartilage under the same conditions [McDevitt & Muir (1971) Anal. Biochem. 44, 612-622]. 4. The role of endogenous proteolytic enzymes in producing proteoglycan heterogeneity, particularly in low-ionic-strength cartilage extracts is discussed. 5. Hyaluronic acid and 'link proteins' were present in the proteoglycan fraction separated from KCl extracts as well as in the fraction separated from CaCl2 extracts. Hyaluronic acid can only be identified in proteoglycan fractions by large-pore-gel electrophoresis after proteolysis and further purification of the fraction. 6. Collagen was extracted by both salt solutions and was tentatively identified as type II. Small amounts of collagen appear to be associated with the proteoglycan-aggregate fraction from the high-ionic-strength extract but not with the corresponding fraction from the KCl extract.     

Hyaluronic acid in cartilage and proteoglycan aggregation

1. Dissociation of purified proteoglycan aggregates was shown to release an interacting component of buoyant density higher than that of the glycoprotein-link fraction of Hascall & Sajdera (1969). 2. This component, which produced an increase in hydrodynamic size of proteoglycans on gel chromatography, was isolated by ECTEOLA-cellulose ion-exchange chromatography and identified as hyaluronic acid. 3. The effect of pH of extraction showed that the proportion of proteoglycan aggregates isolated from cartilage was greatest at pH4.5. 4. The proportion of proteoglycans able to interact with hyaluronic acid decreased when extracted above or below pH4.5, whereas the amount of hyaluronic acid extracted appeared constant from pH3.0 to 8.5. 5. Sequential extraction of cartilage with 0.15m-NaCl at neutral pH followed by 4m-guanidinium chloride at pH4.5 was shown to yield predominantly non-aggregated and aggregated proteoglycans respectively. 6. Most of the hyaluronic acid in cartilage, representing about 0.7% of the total uronic acid, was associated with proteoglycan aggregates. 7. The non-aggregated proteoglycans were unable to interact with hyaluronic acid and were of smaller size, lower protein content and lower keratan sulphate content than the disaggregated proteoglycans. Together with differences in amino acid composition this suggested that each type of proteoglycan contained different protein cores.     

The secretion of enzymes into the pericellular environment.

Connective tissue cells are capable of both synthesizing and degrading the macromolecular components of the extracellular matrix. The degradation of proteoglycan and collagen has been shown to be associated with the extracellular release of proteolytic enzymes, some of which are of lysosomal origin. The identity in carilage of two previously unrecognized proteases, capable of proteoglycan breakdown (CPGases), has recently been achieved by the use of a new assay for proteoglycan degradation. These enzymes have been shown to be synthesized and released in response to vitamin A. The third proteoglycan degrading enzyme of connective tissue cells, cathepsin D, has been located in the pericellular environment by trapping with specific antibody and the pattern of release studied in organ culture, experimental arthritis and in human rheumatoid tissues. The secretion of this enzyme and possibly also of the other CPGases is thought to be of importance in the local (pericellular) turnover of matrix macromolecules and, in association with collagenase, to be the cause of the excessive degradation in the pannus erosion of articular cartilage in rheumatoid arthritis.     

Metalloproteinases in endochondral bone formation: appearance of tissue inhibitor-resistant metalloproteinases

Dissected embryonic chick limbs release neutral metalloproteinases during endochondral bone development. These enzymes degrade cartilage proteoglycan and gelatin in culture medium. We found the enzymes active in the medium conditioned by explants of the region adjacent to the bone marrow cavity (cavity-surround). These enzymes degrade proteoglycan (PG) and/or gelatin. These spontaneously active enzymes are resistant to serum and tissue proteinase inhibitors, alpha 2-macroglobulin, and cartilage metalloproteinase inhibitor (TIMP). The other enzymes secreted from tarsus and bone marrow explants are mostly latent in the culture medium. Activated tarsus enzymes (PG degrading and gelatinolytic) are blocked by the above inhibitors. Activated marrow enzyme does not degrade PG but is resistant to those inhibitors. Cavity-surround enzymes may play an important role in embryonic osteogenesis of long bones because of their resistance to tissue and serum inhibitors. The in vivo mechanisms by which cavity-surround enzymes are activated are yet to be determined.     

Cleavage of cartilage proteoglycan between G1 and G2 domains by stromelysins

Normal and pathological turnover of proteoglycans in articular cartilage involves its cleavage close to the N-terminal G1 domain responsible for aggregation. A fragment containing G1 and G2 N-terminal domains of pig cartilage proteoglycans was therefore used as a substrate to investigate its degradation by the metalloproteinase stromelysin and related recombinant stromelysin enzymes. The stromelysins produced an apparent single cleavage yielding a G1 fragment of 56 kDa and a G2 fragment of 110 kDa. Rabbit bone stromelysin was much more active against the G1-G2 fragment and against proteoglycan aggregates than recombinant human stromelysin-1 and stromelysin-2. All metalloproteinase preparations were active against proteoglycan and the G1-G2 fragment at acid (pH 5.5) and neutral pH (7.4). N-terminal sequencing of the G2 fragment derived from the action of recombinant human stromelysin-1 revealed that cleavage between G1 and G2 occurred at the N-terminal end of the interglobular domain, close to the last cysteine in G1. The specific cleavage site was between an asparagine and a pair of phenylalanine residues, where the asparagine corresponds to residue 341 in human and rat mature core protein sequence.     

The action of human articular-cartilage metalloproteinase on proteoglycan and link protein. Similarities between products of degradation in situ and in vitro.

Interleukin 1 stimulation of human articular cartilage in organ culture produced the concomitant release of proteoglycan fragments and latent metalloproteinase. The released fragments ranged in size from that of almost intact proteoglycan subunits to the product of limiting digestion generated by the activated metalloproteinase. None of the fragments possessed the ability to interact with hyaluronic acid. Analysis of proteoglycan aggregate digested with the activated metalloproteinase showed that isolated hyaluronic acid-binding regions were produced from the proteoglycan subunits, and that the two higher-Mr link-protein components (Mr 48,000 and 44,000) were converted into the lowest-Mr component (Mr 41,000). Link protein extracted from cartilage under stimulation with interleukin 1 showed a similar conversion. These results suggest that interleukin 1 stimulates the release of latent metalloproteinase from chondrocytes and that a proportion of the enzyme is activated in situ in the cartilage matrix. The mode of action of the activated enzyme is compatible with a role in the changes in proteoglycan structure seen in aging.     

Interaction between proteoglycan subunit and type II collagen from bovine nasal cartilage, and the preferential binding of proteoglycan subunit to type I collagen.

We studied the interaction of proteoglycan subunit with both types I and II collagen. All three molecular species were isolated from the ox. Type II collagen, prepared from papain-digested bovine nasal cartilage, was characterized by gel electrophoresis, amino acid analysis and CM-cellulose chromatography. By comparison of type I collagen, prepared from papain-digested calf skin, with native calf skin acid-soluble tropocollagen, we concluded that the papain treatment left the collagen molecules intact. Interactions were carried out at 4 degrees C in 0.06 M-sodium acetate, pH 4.8, and the results were studied by two slightly different methods involving CM-cellulose chromatography and polyacrylamide-gel electrophoresis. It was demonstrated that proteoglycan subunit, from bovine nasal cartilage, bound to cartilage collagen. Competitive-interaction experiments showed that, in the presence of equal amounts of calf skin acid-soluble tropocollagen (type I) and bovine nasal cartilage collagen (type II), proteoglycan subunit bound preferentially to the type I collagen. We suggest from these results that, although not measured under physiological conditions, it is unlikely that the binding in vivo between type II collagen and proteoglycan is appreciably stronger than that between type I collagen and proteoglycan.     

Proteoglycans of hyaline cartilage: Electron-microscopic studies on isolated molecules.

Proteoglycan monomers from guinea-pig costal cartilage, bovine nasal and bovine tracheal cartilage were observed in the electron microscope after being spread in a monomolecular layer with cytochrome c. The proteoglycan molecule appeared as an extended central core filament to which side-chain filaments were attached at various intervals. The molecules from the three sources displayed great ultrastructural similarities. On average, the core filament was about 290 nm long, there were about 25 side-chain filaments per core filament, the side-chain filaments were about 45 nm long, and the distance between the attachment points of the side-chain filaments to the core filament was about 11 nm. With regard to the overall size of the molecules, no evidence of distinct subpopulations was obtained. Good correlation was found between ultrastructural data for the proteoglycan molecules and chemical data obtained by enzyme digestions and gel chromatography. Together these data strongly support the interpretation of the electron-microscopic pictures as indicating a central filament corresponding to the protein core and side-chain filaments corresponding to the chondroitin sulphate chain clusters of the proteoglycan monomers.     

Superficial and deeper layers of dog normal articular cartilage. Role of hyaluronate and link protein in determining the sedimentation coefficients distribution of the nondissociatively extracted proteoglycans

Proteoglycans were extracted under nondissociative conditions from superficial and deeper layers of dog normal articular cartilage. The purified a-A1 preparations were characterized by velocity gradient centrifugation. Superficial specimens exhibited an abundant population of slow sedimenting aggregates whereas the aggregates of deeper preparations sedimented as two well-defined families of molecules. These dissimilarities in the size distribution of the aggregates observed between superficial and deeper a-A1 preparations derived most of all from differences in their content of hyaluronate and link proteins: (a) superficial preparations contained twice as much hyaluronate as deeper specimens; (b) superficial aggregates were link-free and unstable at pH 5.0 whereas deeper preparations contained link-proteins and their faster sedimenting aggregates were stabilized against dissociation at pH 5.0. In these proteoglycan preparations from different cartilage layers, the monomers exhibited an identical capacity for aggregation and the hyaluronate molecules displayed quite similar molecular weight (Mr = 5 x 10(5] and aggregating capacity. These observations as well as aggregating studies conducted with highly purified link protein and purified hyaluronate specimens of different molecular weights support the following conclusions: (a) link protein not only stabilizes proteoglycan aggregates but also enhances the aggregating capacity of hyaluronate; (b) for all practical purposes, the slow sedimenting aggregates represent a secondary complex of hyaluronate and proteoglycan monomers whereas the fast sedimenting aggregates may be considered as a ternary complex wherein link protein stabilizes the hyaluronate-proteoglycans interaction; (c) the distinctive heterogeneity of articular cartilage can be related to structurally different proteoglycan aggregates. The structural dissimilarities observed between superficial and deeper aggregates could reflect the different macromolecular organization of the proteoglycan molecules in the territorial and interterritorial matrices, respectively.     

Effects of tissue compression on the hyaluronate-binding properties of newly synthesized proteoglycans in cartilage explants.

The effects of tissue compression on the hyaluronate-binding properties of newly synthesized proteoglycans in calf cartilage explants were examined. Pulse-chase experiments showed that conversion of low-affinity monomers to the high-affinity form (that is, to a form capable of forming aggregates with 1.6% hyaluronate on Sephacryl S-1000) occurred with a t1/2 of about 5.7 h in free-swelling discs at pH 7.45. Static compression during chase (in pH 7.45 medium) slowed the conversion, as did incubation in acidic medium (without compression). Both effects were dose-dependent. For example, the t1/2 for conversion was increased to about 11 h by either (1) compression from a thickness of 1.25 mm to 0.5 mm or (2) medium acidification from pH 7.45 to 6.99. Oscillatory compression of 2% amplitude at 0.001, 0.01, or 0.1 cycles/s during chase did not, however, affect the conversion. Changes in the hyaluronate-binding affinity of [35S]proteoglycans in these experiments were accompanied by no marked change in the high percentage (approximately 80%) of monomers which could form aggregates with excess hyaluronate and link protein. Since static tissue compression would result in an increased matrix proteoglycan concentration and thereby a lower intra-tissue pH [Gray, Pizzanelli, Grodzinsky & Lee (1988) J. Orthop. Res. 6, 777-792], it seems likely that matrix pH may influence proteoglycan aggregate assembly by an effect on the hyaluronate-binding affinity of proteoglycan monomer. Such a pH mechanism might have a physiological role, promoting proteoglycan deposition in regions of low proteoglycan concentration.     

Synthesis and structure of proteoglycan core protein

Studies of the structure and synthesis of cartilage proteoglycan core protein have been carried out. Deglycosylation of completed, secreted proteoglycan by HF-pyridine treatment yielded an intact homogeneous core protein of approximately 210,000 daltons, with a blocked amino-terminus. Greater than 95% of chondroitin sulfate chains and 80% of N- and O-linked oligosaccharides were removed by the procedure, which made the product an excellent xylosyltransferase acceptor. Little alteration of core protein structure occurred during the HF-pyridine treatment as shown by complete immunoreactivity with antiserums prepared against hyaluronidase-digested proteoglycan. In other studies, the initially synthesized precursor for proteoglycan core protein was found to be approximately 376,000 daltons and localized to the rough membrane fractions. This precursor already contained N-linked oligosaccharides, and was also able to accept xylose, thereby initiating chondroitin sulfate chains. The precursor was translocated intact in an energy-dependent manner to smooth membrane-Golgi fractions where further processing of high mannose type of oligosaccharides and addition of glycosaminoglycan chains occurred. The subcellular distribution pattern of the chondroitin sulfate-synthesizing enzymes corroborated the proposed topological modifications of the proteoglycan core protein precursor.     

The composition of cartilage proteoglycans. An investigation using high-and low-ionic-strength extraction procedures

1. A proteoglycan fraction (the proteoglycan subunit fraction) was prepared from extracts, with 0.15m-KCl (low-ionic-strength) and 0.5m-LaCl(3), 2.0m-CaCl(2) and 4.0m-guanidinium chloride (high-ionic-strength), of bovine nasal cartilage by equilibrium-density-gradient centrifugation, essentially as described by Hascall & Sajdera (1969). 2. The use of different centrifugation times showed that near-equilibrium conditions were reached by 48h for the fractions prepared from the high-ionic-strength extracts. The fraction isolated from the low-ionic-strength extract required a longer centrifugation time to reach equilibrium conditions. 3. The composition of the proteoglycan fractions from the various extracts was compared by analyses of their carbohydrate and amino acid contents. Difference indices were calculated from the amino acid analysis to compare the degree of compositional relationship between the protein components of the proteoglycans. 4. Small compositional differences were found between the proteoglycans isolated from the various high-ionic-strength extracts. The protein content of the fractions from the CaCl(2) extract and the guanidinium chloride extract showed the greatest difference in this respect, although their amino acid analysis was similar. 5. The proteoglycan fraction isolated from the low-ionic-strength extract shows marked differences in composition from the fractions isolated from the high-ionic-strength extracts. Its protein and glucosamine contents were lower whereas its hexuronic acid and galactosamine contents were higher than those of the latter. It also exhibits major differences in its amino acid composition. The glucosamine:galactosamine ratio of the fraction from the low-ionic-strength extract indicates that it may be an almost exclusively chondroitin sulphate-proteoglycan. Its analysis correlates closely with that of a low-molecular-weight proteoglycan isolated from pig laryngeal cartilage by Tsiganos & Muir (1969). 6. The proteoglycan fractions from both the low- and high-ionic-strength extracts migrate as a single band in zone electrophoresis carried out in a sucrose-density gradient at both pH3.0 and pH7.0, although each showed evidence of band widening during the electrophoresis. All the proteoglycan fractions migrated with the same electrophoretic mobility at pH3.0, irrespective of the differences in composition between them. 7. The differences between the proteoglycans from the low- and high-ionic-strength extracts are discussed and the view is advanced that they may be due to association between predominantly chondroitin sulphate-proteoglycans and a keratan sulphate-enriched proteoglycan species.