Search

During aging and degeneration, many changes occur in the structure and composition of human cartilaginous tissues, which include the accumulation of the AGE (advanced glycation end-product), pentosidine, in long-lived proteins. In the present study, we investigated the accumulation of pentosidine in constituents of the human IVD (intervertebral disc), i.e. collagen, aggrecan-derived PG (proteoglycan) (A1) and its fractions (A1D1-A1D6) in health and pathology. We found that, after maturity, pentosidine accumulates with age. Over the age range studied, a linear 6-fold increase was observed in pentosidine accumulation for A1 and collagen with respective rates of 0.12 and 0.66 nmol · (g of protein)-1 · year-1. Using previously reported protein turnover rate constants (kT) obtained from measurements of the D-isomer of aspartic residue in collagen and aggrecan of human IVD, we could calculate the pentosidine formation rate constants (kF) for these constituents [Sivan, Tsitron, Wachtel, Roughley, Sakkee, van der Ham, DeGroot, Roberts and Maroudas (2006) J. Biol. Chem. 281, 13009-13014; Tsitron (2006) MSc Thesis, Technion-Israel Institute of Technology, Haifa, Israel]. In spite of the comparable formation rate constants obtained for A1D1 and collagen [1.81 ± 0.25 compared with 3.71 ± 0.26 μmol of pentosidine · (mol of lysine)-1 · year -1 respectively], the higher pentosidine accumulation in collagen is consistent with its slower turnover (0.005 year-1 compared with 0.134 year-1 for A1D1). Pentosidine accumulation increased with decreasing buoyant density and decreasing turnover of the proteins from the most glycosaminoglycan-rich PG components (A1D1) to the least (A1D6), with respective kF values of 1.81 ± 0.25 and 3.18 ± 0.37 μmol of pentosidine · (mol of lysine)-1 · year-1. We concluded that protein turnover is an important determinant of pentosidine accumulation in aggrecan and collagen of human IVD, as was found for articular cartilage. Correlation of pentosidine accumulation with protein half-life in both normal and degenerate discs further supports this finding. © 2006 Biochemical Society. Chemicals / CAS: collagen, 9007-34-5; pentosidine, 124505-87-9; Aggrecans; Arginine, 74-79-3; Collagen, 9007-34-5; Extracellular Matrix Proteins; Lectins, C-Type; Lysine, 56-87-1; pentosidine, 124505-87-9; Proteochondroitin Sulfates; Proteoglycans

We screened more than 60000 RIKEN mouse cDNAs for novel ECM (extracellular matrix) proteins by extensive computational screening followed by recombinant expression and immunohistochemical characterization. We identified two novel olfactomedin-family proteins characterized by the presence of tandem CXCXCX9C motifs in the N-terminal region, a coiled-coil domain and an olfactomedin domain in the C-terminal region. These proteins, named photomedin-1 and photomedin-2, were secreted as disulphide-bonded dimers (photomedin-1) or oligomers/multimers (photomedin-2) with O-linked carbohydrate chains, although photomedin-1 was proteolytically processed in the middle of the molecule after secretion. In the retina, photomedin-1 was selectively expressed in the outer segment of photoreceptor cells and photomedin-2 was expressed in all retinal neurons. Among a panel of ECM components, including glycosaminoglycans, photomedins preferentially bound to chondroitin sulphate-E and heparin. These results, together, indicate that photomedins are novel olfactomedin-domain-containing extracellular proteins capable of binding to proteoglycans containing these glycosaminoglycan chains.

The keratan sulphate proteoglycan lumican regulates collagen fibrillogenesis to maintain the integrity and function of connective tissues such as cornea. We examined the role of a highly conserved cysteine-containing domain proximal to the N-terminus of lumican in collagen fibrillogenesis using site-specific mutagenesis to prepare plasmid DNA encoding wild-type murine lumican (Cys(37)-Xaa(3)-Cys(41)-Xaa-Cys-Xaa(9)-Cys) and a Cys-->Ser (C/S) mutant (Cys(37)-Xaa(3)-Ser(41)-Xaa-Cys-Xaa(9)-Cys). cDNAs were cloned into the pSecTag2A vector, and cultures of MK/T-1 cells (an immortalized cell line from mouse keratocytes) were transfected with the cDNAs. Stable transformants were selected and cloned in the presence of Zeocin. All stable transformants maintained a dendritic morphology and growth rate similar to those of parental MK/T-1 cells. Western blot analysis with anti-lumican antibody detected a 42 kDa lumican protein secreted into the culture medium of both wild-type and C/S mutant lumican cell lines. Ultrastructural analyses by transmission electron microscopy showed both cell lines to form a multi-layered stroma ex vivo, but the matrix assembled by the two cell lines differed. Compared with the mutant cell line, the wild-type cells assembled a more organized matrix with regions containing orthogonal collagen fibrils. In addition, the fibrils in the extracellular matrix formed by the mutant cell line exhibited alterations in fibril packing and structure. Immunostaining analysed by confocal microscopy showed a further difference in this matrix, with the marked occurrence of lumican and collagen I co-localization in the lumican wild-type cells, but a lack thereof in the lumican C/S mutant cells. The results indicate that the cysteine-rich domain of lumican is important in collagen fibrillogenesis and stromal matrix assembly.

Previously, only matrix metalloproteinases were believed capable of cleaving the cartilage proteoglycan, aggrecan, between Asn341 and Phe342, to yield a small G1 fragment terminating in the residues VDIPEN. We show that the combined endo- and exopeptidase activities of the cysteine protease, cathepsin B, also generate this epitope, suggesting that it should no longer be considered as an exclusive marker of metalloproteinase activity.

A rat chondrosarcoma cell line and bovine cartilage explants have been used to study the control of aggrecan degradation by chondrocytes treated with interleukin-1 (IL-1) or retinoic acid (RA). Aggrecan fragment analysis with anti-neo-epitope antibodies suggests that aggrecanase (an as yet unidentified enzyme) is the only aggrecan-degrading proteinase active in these cultures. With rat cells, aggrecanase converts the aggrecan core protein into two major G1-domain-bearing products (60 kDa with a C-terminal Glu-373, and 220 kDa with a C-terminal Glu-1459). Both products were quantified on a standardized Western analysis system with a G1-specific antibody. Immunoblots were analysed by scanning densitometry and the sensitivity, linearity and reproducibility of the assay were established. With rat cells the aggrecanase response to IL-1 was optimal at about 2 mM glutamine, but was progressively inhibited at higher concentrations, with about 90% inhibition at 10 mM glutamine. Such inhibition by glutamine was not, however, observed with bovine explants. On the other hand, marked inhibition of aggrecanase-dependent cleavage was observed with both rat cells and bovine explants when d(+)-glucosamine was included at concentrations above 2 mM. Inhibition was apparently not due to cytotoxicity or interference with IL-1 signalling, since biosynthetic activity was not inhibited and inhibition of the aggrecanase response was also obtained when RA was used as the catabolic stimulator. Possible mechanisms for the inhibition of the aggrecanase response by glucosamine in chondrocytes treated with IL-1 or RA are discussed.

N-terminal analysis of aggrecan fragments lost from bovine nasal cartilage cultured in the presence of recombinant human interleukin 1alpha revealed a predominant ARGSVIL sequence with an additional ADLEX sequence. Production of the ARGSVIL-containing fragments has been attributed to the action of a putative proteinase, aggrecanase. The minor sequence (ADLEX) corresponds to a new reported cleavage product; comparison of this sequence with the available partial sequence of bovine aggrecan indicates that this is the product of a cleavage occurring towards the C-terminus of the protein. Matrix metalloproteinase (MMP) inhibitors inhibited aggrecan loss from bovine nasal explants incubated in the presence of recombinant human interleukin 1alpha. A strong correlation between inhibition of aggrecan metabolism and inhibition of stromelysin 1 (MMP 3) (r=0.93) suggests a role for stromelysin or a stromelysin-like enzyme in cartilage aggrecan metabolism. However, the compounds were approx. 1/1000 as potent in inhibiting aggrecan loss from the cartilage explants as they were in inhibiting stromelysin. There was little or no correlation between inhibition of aggrecan metabolism and inhibition of gelatinase B (MMP 9) or inhibition of collagenase 1 (MMP 1). Studies with collagenase inhibitors with a range of potencies showed a correlation between inhibition of collagenase activity and inhibition of collagen degradation in the cartilage explant assay. This indicates that in interleukin 1alpha-driven bovine nasal cartilage destruction, stromelysin (or a closely related enzyme) is involved in aggrecan metabolism, whereas collagenase is principally responsible for collagen degradation. [on SciFinder (R)]

Profound changes occur in the uterine cervix during pregnancy. In particular, the extracellular matrix of the connective tissue is remodelled extensively. To elucidate the mechanisms involved in this process, we have analysed the proteoglycan pattern in the human cervix from pregnant and non-pregnant women. Proteoglycans of the cervix tissue specimen were extracted with 4 M guanidine hydrochloride and precipitated with 80% ethanol. Purification of proteoglycans was performed by several chromatographic steps. Characterization of proteoglycans was done by SDS/PAGE before and after digestion with glycosaminoglycan-specific enzymes. Proteoglycans were detected by combined Alcian Blue/silver staining or, after blotting of biotin-labelled proteoglycans on to poly(vinylidene difluoride) membrane, with peroxidase-conjugated avidin or by the use of keratan sulphate- or decorin-specific monoclonal antibodies. In contrast with previous reports, where only chondroitin/dermatan sulphate proteoglycans have been found in the uterine cervix, we have shown in the present study the existence of a large keratan sulphate proteoglycan with an Mr > 220000 in cervix samples from non-pregnant and pregnant women. This proteoglycan showed a strong reaction with the keratan sulphate-specific monoclonal antibody 5D4 and could be degraded by keratanases. The size of the core protein of this keratan sulphate proteoglycan was estimated to be about Mr 220000.

To elucidate the correlation between vascular cholesterol metabolism and proteoglycan (PrGl) biosynthesis, we investigated PrGl synthesis in human aortic smooth muscle cells (SMCs) after cholesterol enrichment with cationized low-density lipoproteins (LDL). Compared with normal SMCs, total PrGl synthesis by cholesterol-enriched cells decreased 2.4-fold (11874 +/- 530 d.p.m. per 10(5) cells compared with 4890 +/- 385 d.p.m. per 10(5) cells). This was the net result of a 6.9-fold reduction in medium PrGl (11000 +/- 490 d.p.m. per 10(5) cells compared with 1580 +/- 246 d.p.m. per 10(5) cells) and a 3.8-fold increase in cellular PrGl over controls (874 +/- 27 d.p.m. per 10(5) cells compared with 3310 +/- 193 d.p.m. per 10(5) cells). Prior incubation of SMCs with native LDL had no effect on PrGl synthesis by these cells. The decrease in PrGl synthesis in cholesterol-enriched cells correlated with a 90% and 20% reduction in the steady-state level of mRNA for biglycan and decorin respectively, and a virtual elimination of the steady-state level of mRNA for versican over controls. Despite the down-regulation of PrGl synthesis, cholesterol-loaded cells produced a 2-fold increase in a PrGl subfraction with high affinity for LDL. Compared with the corresponding PrGl subfraction from normal cells, that from the cholesterol-enriched cells exhibited increased charge density and a higher molecular mass and contained relatively larger proportions of chondroitin 6-sulphate and dermatan sulphate. These results show that PrGl metabolism is dramatically altered in cholesterol-enriched human SMCs.

Glial hyaluronate-binding protein (GHAP) is a 60 kDa glycoprotein with an amino acid sequence identical to that of the hyaluronate-binding region of versican, a large fibroblast aggregating proteoglycan found in the brain. Both GHAP and versican were identified by immunoblot in bovine brain extracts prepared only minutes after death. Human recombinant collagenase, stromelysin, mouse gelatinase and gelatinases isolated from human brain by affinity chromatography digest versican and give rise to a polypeptide with electrophoretic mobility identical to GHAP. Immunoblot analysis, peptide mapping and C-terminal amino acid sequencing indicate that the polypeptide generated by digestion with human brain gelatinases is identical to GHAP. We suggest that GHAP is a naturally occurring versican degradation product.

The complete exon-intron organization of the human aggrecan gene has been defined, and the exon organization has been compared with the individual domains of the protein core. A yeast artificial chromosome containing the aggrecan gene was selected from the Centre d'Etude du Polymorphisme Humaine yeast artificial chromosome library. A cosmid sulibrary was created from this, and direct sequencing of individual cosmids was used to provide the exon-intron organization. The human aggrecan gene was found to be composed of 19 exons ranging in size from 77 to 4224 bp. Exon 1 is non-coding, whereas exons 2-19 code for a protein core of 2454 amino acids with a calculated mass of 254379 Da. Intron 1 of the gene is at least 13 kb. Overall, the sizes of the 18 introns range from 0.5 to greater than 13 kb. Each intron begins with a GT and ends with an AG, thus obeying the GT/AG rule of splice-junction sequences. The entire coding region is contained in 39.4 kb of the gene. The organization of exons is strongly related to the specific domains of the protein core. The A loop of G1 and the interglobular domain are encoded by exons 3 and 7 respectively. The B and B' loops of G1 are encoded by exons 4-6, and those of G2 are encoded by exons 8-10. These sets of exons, coding for the B and B' loops, are identical in size and organization. This is supported by the intron classes associated with these exons. Exon 11 codes for the 5′ half of the keratan sulphate-rich region, and exon 12 codes for the 3′ half of the keratan sulphate-rich region as well as the entire chondroitin sulphate-rich region. G3 is encoded by exons 13-18, including the alternatively spliced epidermal growth factor-like and complement regulatory protein-like domains. The correspondence between the exon organization and the protein domains argues strongly for modular assembly of the aggrecan gene.

Recent studies have shown that mesangial cells derived from human adult glomeruli synthesize a number of 35S-labelled proteoglycans including a large chondroitin sulphate proteoglycan (CSPG), two dermatan sulphate proteoglycans (biglycan and decorin) and two heparan sulphate proteoglycans [Thomas, Mason and Davies (1991) Biochem. J. 277, 81-88]. In the present study we have examined the interaction of these proteoglycans with hyaluronan (HA) using associative gel chromatography. Only the large CSPG bound to HA, with 60% of those molecules in the medium and 80% of those in the cell layer being able to interact. Reduction and alkylation, or treatment of the monomer CSPG with proteinases, prevented the formation of aggregates, suggesting that the core protein was involved. The aggregates formed between purified CSPG and HA could be dissociated in the presence of HA-oligosaccharides of at least 10 monosaccharides in length. The inclusion of link protein with CSPG and HA promoted the formation of aggregates. Experiments with 3H-labelled mesangial-cell proteoglycans confirmed that only the large CSPG, with core protein molecular masses of 400 kDa and 500 kDa, interacted with HA. After chondroitin ABC lyase treatment of CSPG isolated from conditioned culture medium, several bands similar to those observed with 3H-labelled core proteins were identified using a polyclonal antiserum that recognizes versican. A monoclonal antibody recognizing the 1-C-6 epitope in the G1 and G2 globular regions of aggrecan did not recognize either mesangial-cell CSPG or bovine aortic versican. Northern-blot analysis confirmed that human mesangial cells express versican. Thus human mesangial large CSPG is a member of the versican family of proteoglycans. The interaction of CSPG and HA within the glomerulus may be important in glomerular cell migration and proliferation.

To determine the effects of lipid accumulation on proteoglycan synthesis, we studied proteoglycan biosynthesis in rabbit aortic smooth muscle cells in culture. Cholesterol-enrichment was accomplished by incubating confluent smooth muscle cells with cationized low-density lipoprotein. Control and cholesterol-enriched cells were incubated with [35S]sulphate, [3H]glucosamine, or [3H]serine. Metabolically labelled proteoglycans in the cell layer and medium were quantified. During a 20 h incubation period, proteoglycan synthesis in cholesterol-enriched cells increased by 40-50% above that in control cells. A similar increase in precursor incorporation into proteoglycans was also noted following a short 15 min pulse. The cholesterol-enriched cells also showed a 45-50% increase over control rates in the intralysosomal accumulation of a large chondroitin sulphate proteoglycan and a small dermatan sulphate proteoglycan. The enhanced synthesis of proteoglycans in cholesterol-enriched cultures was inhibited by cycloheximide and actinomycin D, which are inhibitors of protein synthesis and transcription respectively. Proteoglycan turnover was investigated by pulse-chase analysis. Following a 2-h pulse, intracellular proteoglycans in cholesterol-enriched cells disappeared, having a half-life of 26.5 h compared with 2.8 h for those in the control cells. The amount of trypsin-releasable proteoglycan was significantly reduced in cholesterol-enriched cells. In addition, the degradation of proteoglycans was severely retarded in cholesterol-enriched cultures. The activities of three acid hydrolases, N-acetyl-beta-hexosaminidase, beta-glucuronidase and cathepsin C, were significantly reduced in cholesterol-enriched cells compared with activities in control cells. The results indicate that proteoglycan metabolism is altered in cholesterol-enriched smooth muscle cells.

The proteoglycans synthesized by an osteoblast-like cell line of rat origin (UMR 106-01) were defined after biosynthetic labelling with [35S]sulphate and [3H]glucosamine. Newly synthesized labelled proteoglycans were characterized by differential enzymic digestion in combination with analytical gel filtration and SDS/PAGE. UMR 106-01 cells were found to synthesize three major species of proteoglycan: a large chondroitin sulphate proteoglycan of Mr approximately 1 x 10(6), with a core protein of Mr approximately 350,000-400,000; a small chondroitin sulphate-containing species of Mr approximately 120,000 with a core protein of Mr 43,000; and a heparan sulphate proteoglycan of Mr approximately 150,000, with a core protein of Mr approximately 80,000. Over 70% of the newly synthesized intact proteoglycan species are associated with the cell layer of near-confluent cells; however, accessibility to trypsin digestion suggests an extracellular location. Chemical characteristics of the proteoglycans and preliminary mRNA hybridization indicate that the small chondroitin sulphate proteoglycan is probably PG II (decorin). The large chondroitin sulphate proteoglycan is most likely related to a hyaluronate-aggregating species from fibroblasts (versican), and the heparan sulphate proteoglycan bears striking similarities to cell-membrane-intercalated species described for a number of cell types.

Incubation of cultured fibroblasts with p-nitrophenyl beta-D-xyloside resulted in a concentration-dependent increase in galactosaminoglycan synthesis. At low concentration of added xyloside large and small radiolabelled proteoglycans and xyloside-bound polysaccharides were recovered from the medium, whereas at high concentrations only xyloside-bound polysaccharides were found. In the cell layer proteoglycans and xyloside-bound polysaccharides were found at all concentrations tested. Only galactosaminoglycan chains were polymerized on the xyloside primer. At low concentrations of added xyloside the structure of the galactosaminoglycans formed on the xyloside was similar to that of the small dermatan sulphate proteoglycan, i.e. mainly composed of L-iduronic acid-containing 4-sulphated disaccharides. With increasing concentration of added xyloside the co-polymeric structure of the small dermatan sulphate proteoglycan and the xyloside-bound polysaccharide was changed to contain a larger proportion of D-glucuronosyl residues with only slight changes in the sulphation pattern. No structural change in the polysaccharide chains of the large glucuronic acid-rich proteoglycans occurred. At 1 mM-xyloside, where no proteoglycans were formed, the polysaccharide was shorter and composed mainly of D-glucuronosyl-containing disaccharides with a ratio of 4-sulphate to 6-sulphate substituents of 1:2. This is similar to the structure of the large glucuronic acid-rich proteoglycan synthesized by these cells. Thus the main difference induced by the xyloside treatment was changed polymer modification at high xyloside concentrations. The specific activities of the polymer-modifying enzymes, uronosyl C-5-epimerase and 4-sulphotransferase, were therefore measured and found to be decreased by 30-50% in fibroblasts treated with high xyloside concentrations. It is suggested that the protein core is of importance for regulating the activity of the polymer-modifying enzymes.

The structure of human skin fibroblast heparan sulphate has been examined by depolymerization with heparinase, which specifically cleaves highly sulphated disaccharides of structure GlcNSO3 (+/-6S)-alpha 1,4IdoA(2S) [N-sulphated glucosamine (6-sulphate)-alpha 1,4-iduronic acid 2-sulphate]. Heparan sulphate contained only a small proportion (approximately 10%) of linkages susceptible to this enzyme. The major products of depolymerization with heparinase were large oligosaccharides with an average molecular mass of 10 kDa (dp approximately 40, where dp is degree of polymerization; for disaccharides, dp = 2 etc.) as assessed by gel filtration on Sepharose CL-6B, compared with a molecular mass of 45 kDa (dp approximately 200) for the intact chains. The large heparinase-resistant oligosaccharides were highly susceptible to depolymerization with the enzyme heparitinase, which cleaves heparan sulphate in areas of low sulphation, where N-acetylated disaccharides [GlcNAc-alpha 1,4GlcA (N-acetylglucosaminyl-alpha 1,4-glucuronic acid)] are the predominant structural unit. Further analysis of the location of the heparinase cleavage sites indicated that they were predominantly found in a central position in GlcNSO3-alpha 1,4IdoA repeat sequences of average length four to seven disaccharides (dp 8-14). These results indicate that heparinase cleaves heparan sulphate in approximately four or five N-sulphated domains, each domain containing a cluster of two or three susceptible disaccharides; the domains are separated by long N-acetyl-rich sequences that are markedly deficient in sulphate groups. On the basis of these findings a model is proposed which depicts heparan sulphate as an ordered polymeric structure composed of an alternate arrangement of sulphate-rich and sulphate-poor regions. The sulphate-rich regions are likely to be flexible areas of the chain because of their high content of the conformationally versatile IdoA and IdoA(2S) residues. The model has important implications for the biosynthesis and functions of heparan sulphate.

Heparan sulphate proteoglycans were solubilized from whole rat livers by homogenization and dissociative extraction with 4 M-guanidinium chloride containing Triton X-100 and proteinase inhibitors. The extract was subjected to trichloroacetic acid precipitation and the proteoglycan remained soluble. This was then purified to apparent homogeneity by a combination of (a) DEAE-Sephacel chromatography, (b) digestion with chondroitinase ABC followed by f.p.l.c. Mono Q ion-exchange chromatography, and (c) density-gradient centrifugation in CsCl and 4 M-guanidinium chloride. Approx. 1.5 mg of proteoglycan was obtained from 30 livers with an estimated recovery of 25%. The purified proteoglycan was eluted from Sepharose CL6B as an apparently single polydisperse population with a Kav. of 0.19 and displayed a molecular mass of greater than or equal to 200 kDa (relative to protein standards) by SDS/PAGE. Its heparan sulphate chains were eluted with a Kav. of 0.44 and have an estimated molecular mass of 25 kDa. Digestion of the proteoglycan with a combination of heparinases yielded core proteins of 77, 49 and 44 kDa. Deglycosylation using trifluoromethanesulphonic acid, though slightly decreasing the sizes, gave an identical pattern of core proteins. Electrophoretic detergent blotting demonstrated that all of the core proteins were hydrophobic and are probably integral plasma membrane molecules. The peptide maps generated by V8 proteinase digestion of the two major core proteins (77 and 49 kDa) were very similar, suggesting that these two core proteins are structurally related.

Incorporation of [35S]sulphate by cultures of matrix-free cells from chick embryo sterna in the presence of the glutamine analogue 6-diazo-5-oxo-L-norleucine (0.58 mM) was inhibited in a time-dependent manner to less than 15% of that in control cultures after 2 h. Characterization of the major cartilage proteoglycan synthesized under these conditions showed that it contained few, if any, normal-sized chondroitin sulphate chains and only about half of the normal complement of substituted serine residues. Subsequent addition of D-glucosamine hydrochloride (final concn. 2 mM) resulted in a time-dependent recovery of [35S]sulphate incorporation to 90% of control cultures after 2 h, but restored the chondroitin sulphate chains to normal size within 15 min. On the basis of these results, it is concluded that a 2 h preincubation is necessary to deplete the chondrocytes of the endogenous supply of UDP-N-acetyl-D-glucosamine required for optimal glycoconjugate synthesis, and that this situation results in the synthesis of a chondroitin sulphate proteoglycan with significantly altered properties, owing to the paucity of glycosaminoglycan chains; however, this condition is completely reversible if the D-glucosamine pool is repleted.

Small dermatan sulphate proteoglycan II from cultured human skin fibroblasts interacts with type I collagen in vitro and in vivo. When fibroblasts are maintained in a type I collagen lattice the proteoglycan remains exclusively within the lattice, and its association with fibrils can be demonstrated immunocytochemically. On the basis of [35S]sulphate incorporation, small proteoglycan II comprises about 80% of total proteoglycans secreted by cells in monolayer culture. In a collagen lattice, fibroblasts down-regulate its synthesis to the level of large chondroitin sulphate/dermatan sulphate and of heparan sulphate proteoglycans, the synthesis of which remains unaffected. Compared with the product from monolayer cultures, small proteoglycan II from collagen gels contained a longer polysaccharide chain which is characterized by a larger proportion of disulphated and a smaller proportion of monosulphated glucuronic acid-containing disaccharides. The half-life varied between 60 and 110 h. It is suggested that the compositional differences between the proteoglycan from monolayer cultures and from cells in a collagen lattice are related to the slower intracellular trafficking of the proteoglycan under the latter culture conditions.

Proteoglycans in mineralized (0.5 M-EDTA/4 M-guanidinium chloride-extractable) and non-mineralized (4 M-guanidinium chloride-extractable) matrices synthesized by a mouse osteoblastic-cell line MC3T3-E1 were characterized at different phases of mineralization in vitro. Cell cultures were labelled with [35S]sulphate and either [3H]glucosamine or 3H-labelled amino acids. At the mineralization phase a large majority of proteoglycans were extracted with 4 M-guanidinium chloride (G extract), and at least five species of labelled proteoglycans were identified; dermatan sulphate proteoglycans (DSPG), apparent Mr approx. 120,000 and 70,000), heparan sulphate proteoglycans (HSPG, apparent Mr approx. 200,000 and 120,000) and DS chains with very little core protein. DSPGs weakly bound to an octyl-Sepharose CL-4B column and HSPGs bound more tightly, whereas DS chains did not bind to the column. Amounts of labelled proteoglycans extracted with 0.5 M-EDTA/4 M-guanidinium chloride (EDTA extract) were much less than those in G extract. Although the predominant species in the EDTA extract were comparable with the DS or DSPGs in the G extract, none of them bound to octyl-Sepharose CL-4B, indicating their lack of hydrophobicity. At the nonmineralizing phase a large chondroitin sulphate proteoglycan (Mr greater than 600,000) was found in the matrix in addition to the five proteoglycan species similar to those at the mineralization phase. Although DS chains at the early phase were similar in size to those at the mineralization phase, the ratio of 2-acetamido-2-deoxy-3-O-(beta-D-gluco-4-enepyranosyluronic acid)-4-O-sulpho-D-galactose to 2-acetamido-2-deoxy-3-O-(beta-D-gluculo-4-enepyranosyluronic acid)-6-O-sulpho-D-galactose was less than that at the mineralization phase. These results agree with those of previous studies performed in vivo and suggest that alteration in the synthesis of proteoglycans is involved in the mineralization process. They also suggest that at the osteoblastic mineralization front proteoglycans undergo partial degradation and lose their hydrophobicity.

Confluent cultures of mouse aortic endothelial (END-D) were incubated with either [35S]methionine or 35SO4 2-, and the radiolabelled proteoglycans in media and cell layers were analysed for their hyaluronate-binding activity. The proteoglycan subfraction which bound to hyaluronate accounted for about 18% (media) and 10% (cell layers) of the total 35S radioactivity of each proteoglycan fraction. The bound proteoglycan molecules could be dissociated from the aggregates either by digestion with hyaluronate lyase or by treatment with hyaluronate decasaccharides. Digestion of [methionine-35S]proteoglycans with chondroitinase and/or heparitinase, followed by SDS/polyacrylamide-gel electrophoresis, indicated that the medium and cell layer contain at least three chondroitin sulphate proteoglycans, one dermatan sulphate proteoglycan, and two heparan sulphate proteoglycans which differ from one another in the size of core molecules. Among these, only the hydrodynamically large chondroitin sulphate species with an Mr 550,000 core molecule was shown to bind to hyaluronate. A very similar chondroitin sulphate proteoglycan capable of binding to hyaluronate was also found in cultures of calf pulmonary arterial endothelial cells (A.T.C.C. CCL 209). These observations, together with the known effects of hyaluronate on various cellular activities, suggest the existence of possible specialized functions of this proteoglycan subspecies in cellular processes characteristic of vascular development and diseases.

The hyaluronan-binding region (HABR) was prepared from pig laryngeal cartilage aggrecan and the amino acid sequence was determined. The HABR had two N-termini: one N-terminal sequence was Val-Glu-Val-Ser-Glu-Pro (367 amino acids in total), and a second N-terminal sequence (Ala-Ile-Ser-Val-Glu-Val; 370 amino acids in total) was found to arise due to alternate cleavage by the signal peptidase. The N-linked oligosaccharides were analysed by examining their reactivity with a series of lectins. It was found that the N-linked oligosaccharide on loop A was of the mannose type, while that on loop B was of the complex type. No reactivity was detected between the N-linked oligosaccharide on loop B' and any of the lectins. The location of keratan sulphate (KS) in the HABR was determined by Edman degradation of the immobilized KS-containing peptide. The released amino acid derivatives were collected and tested for the presence of epitope to antibody 5-D-4. On the basis of 5-D-4 reactivity and sequencing yields, the KS chains are attached to threonine residues 352 and 357. There is no KS at threonine-355. This site is not in fact in G1, but about 16 amino acid residues into the interglobular domain. Comparison of the structure of the KS chain from the HABR and from the KS domain of pig laryngeal cartilage aggrecan was made by separation on polyacrylamide gels of the oligosaccharides arising from digestion with keratanase. Comparison of the oligosaccharide maps suggests that the KS chains from both parts of the aggrecan molecule have the same structure.

A heparin-binding protein was isolated from bovine uteri and purified to homogeneity. This protein appears as a double band of approx. 78 kDa in SDS/polyacrylamide-gel electrophoresis and has an isoelectric point of 5.2. The binding of heparin to this protein is saturable. No other glycosaminoglycan from mammalian tissue, such as hyaluronic acid, chondroitin sulphate, dermatan sulphate or keratan sulphate, binds to the 78 kDa protein. Dextran sulphate binds in a non-saturable fashion. Certain heparan sulphate polysaccharide structures are required for binding to the 78 kDa protein. Some proteoheparan sulphates, such as endothelial cell-surface proteoheparan sulphate, show only weak interaction with the 78 kDa protein in contrast with a basement-membrane proteoheparan sulphate from HR-9 cells. Antibodies against the 78 kDa protein inhibit binding of proteoheparan [35S]sulphate from basement membranes to smooth-muscle cells. Conventional antibodies, Fab fragments and some monoclonal antibodies, inhibit smooth-muscle cell proliferation in a similar range as that observed for heparin. The protein was detected in a variety of tissues and cells but not in blood cells. A possible role of this protein as a receptor for heparin or heparan sulphate and its function in the control of the arterial wall structure are discussed.

The degradation of intravenously administered chondroitin sulphate-peptide, obtained by trypsin digestion of rat cartilage preparations labelled in vitro with 35S (and, in some cases, with 3H), was studied in rats. As with free chains of chondroitin sulphate, the major site of accumulation and degradation in the body was the liver, although peptide-linked chains were taken up more rapidly than free chains. In the first 2h after intravenous injection of a chondroitin sulphate-peptide fraction, labelled macromolecular components were excreted in the urine. These were shown to be chondroitin sulphate-peptide of the same degree of sulphation but of smaller average size than the injected material. A similar observation was made when free chains of chondroitin sulphate from the same source were administered intravenously. An isolated perfused rat kidney failed to de-sulphate circulating chondroitin sulphate-peptide, but a component of lower average molecular weight was excreted in the urine. When a chondroitin sulphate-peptide fraction of relatively larger hydrodynamic volume was administered, very little chondroitin sulphate appeared in the urine in the first 2h. It was concluded that, depending on size and/or peptide content, the chondroitin sulphate-peptide released from connective tissues into the circulation would probably be subjected to one of two alternative fates. The smaller fragments are more likely to be excreted in the urine, whereas the larger ones are taken up by the liver and there degraded to inorganic sulphate and undefined carbohydrate components.

Proteoglycans were isolated from cartilage by extraction with 4M-guanidinium chloride followed by direct centrifugation in 4M-guanidinium chloride/CsCl at a low starting density, 1.34 g/ml. N-Ethylmaleimide was included in the extraction solvent as a precaution against contamination of proteoglycans with unrelated proteins mediated by disulphide exchange. A novel, discrete, low-buoyant-density proteoglycan (1.40-1.35 g/ml) was demonstrated by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. Its proteoglycan nature was revealed by the shift in the molecular size observed on gel electrophoresis after treatment with chondroitinase ABC. The core protein was monodisperse. The proteoglycan was further purified by gel chromatography with and without addition of hyaluronate. The proteoglycan constitutes less than 2% (by weight) of the total extracted proteoglycans and is not capable of interacting with hyaluronate. The same proteoglycan was purified in larger quantities by sequential associative and dissociative CsCl-density-gradient centrifugation, zonal rate sedimentation in a sucrose gradient and gel chromatography on Sepharose CL-4B. The pure proteoglycan had a molecular weight of 76 300 determined by sedimentation-equilibrium centrifugation and an apparent partial specific volume of 0.59 ml/g. It contained about 25% protein (of dry weight) and had remarkably high contents of leucine and cysteine as compared with other proteoglycans. The proteoglycan contained two to three large chondroitin sulphate chains and some oligosaccharides.

Newly synthesized rat glomerular [35S]proteoglycans were labelled in vivo after injecting Na2[35S]SO4 intraperitoneally. At the end of the labelling period (7 h) the kidneys were perfused in situ with 0.01% (w/v) cetylpyridinium chloride. This fixed proteoglycans in the tissue and increased their recovery 2-3-fold during subsequent isolation of glomeruli from the renal cortex. The glomeruli were fractionated by a modified osmotic lysis and detergent extraction procedure [Meezan, Brendel, Hjelle & Carlson (1978) in The Biology and Chemistry of Basement Membranes (Kefalides, N.A., ed.), Academic Press, New York; Kanwar & Farquhar (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4493-4497] to obtain a basement membrane preparation. The proteoglycans released at each stage of the procedure were characterized using DEAE-Sephacel ion-exchange chromatography, chondroitinase ABC and HNO2 digestion and Sepharose CL-4B gel-permeation chromatography. About 85% of the [35S]proteoglycans synthesized were of the heparan sulphate variety, the remainder being chondroitin sulphate proteoglycans. Three sizes of heparan sulphate proteoglycans were identified. The largest (HS1, Kav. 0.47) accounts for 44% of the total extractable heparan sulphates. About one third of HS1 were extracted from the glomerular basement-membrane fraction with 8 M-urea and 4 M-guanidine hydrochloride but the remainder were released from the glomerulus during preparation of the fraction. The two smaller molecules (HS2, Kav. 0.56 and HS3, Kav. 0.68) accounted for 27% and 28% of the extractable heparan sulphate respectively and were not associated with the basement membrane fraction. HS1, HS2 and HS3 were also isolated from non-fixed glomeruli labelled in vivo but with much lower recovery. In glomeruli labelled in vitro, heparan sulphate accounted for only 35% of the proteoglycans, the remainder being of the chondroitin sulphate type. Proteoglycans similar to HS1, HS2 and HS3 were present in glomeruli labelled in vitro but, in addition, a large, highly charged heparan sulphate (HS1a) was extracted from the glomerular basement-membrane fraction of these glomeruli. It accounted for 6% of the total heparan sulphate.

1. Proteoglycans were extracted from sclera with 4 M-guanidine hydrochloride in the presence of proteinase inhibitors and purified by ion-exchange chromatography and density-gradient centrifugation. 2. The entire proteoglycan pool was characterized by compositional analyses and by specific chemical (periodate oxidation) and enzymic (chondroitinases) degradations. The glycan moieties of the molecules were exclusively galactosaminoglycans (dermatan sulphate-chondroitin sulphate co-polymers). In addition, the preparations contained small amounts of oligosaccharides. 3. The scleral proteodermatan sulphates were fractionated into one larger (I) and one smaller (II) component by gel chromatography. Proteoglycan I was eluted in a more excluded position on gel chromatography in 0.5 M-sodium acetate than in 4.0 M-guanidine hydrochloride. Reduced and alkylated proteoglycan I was eluted in the same position (in 0.5 M-sodium acetate) as was the starting material (in 4.0 M-guanidine hydrochloride). The elution position of proteoglycan II was the same in both solvents. Proteoglycans I and II had s0 20,w values of 2.8 × 10(-13) and 2.2 × 10(-13) s respectively in 6.0 M-guanidine hydrochloride. 4. The two proteoglycans differed with respect to the nature of the protein core and the co-polymeric structure of their side chains. Also proteoglycan I contained more side chains than did proteoglycan II. The dermatan sulphate side chains of proteoglycan I were D-glucuronic acid-rich (80%), whereas those of proteoglycan II contained equal amounts of D-glucuronic acid and L-iduronic acid. Furthermore, the co-polymeric features of the side chains of proteoglycans I and II were different. The protein core of proteoglycan I was of larger size than that of proteoglycan II. The latter had an apparent molecular weight of 46 000 (estimated by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis), whereas the former was greater than 100 000. In addition, the amino-acid composition of the two core preparations was different. 5. As proteoglycan I altered its elution position on gel chromatography in 4 M-guanidine hydrochloride compared with 0.5 M-sodium acetate it is proposed that a change in conformation or a disaggregation took place. If the latter hypothesis is favoured, aggregation may be due to self-association or mediated by an extrinsic molecule, e.g. hyaluronic acid.

Afferent lymph vessels entering the popliteal lymph nodes of sheep were infused with [3H]acetyl-labelled hyaluronan [HA; Mr of (0.85-1.2) x 10(5)] for up to 4 h at a rate of 17.4-23.1 micrograms/h. As much as 22.8 micrograms (99%) of infused [3H]HA was taken up by the node per h and degraded. During this interval it was observed that infused HA polymers of higher Mr were absorbed by the node to a greater degree than those of lower Mr. When proteoglycan monomer (PG; Mr 5 x 10(5); 400 micrograms of hexuronic acid/h) was infused concurrently with [3H]HA, the absolute amount of radioactivity appearing in efferent lymph (i.e. labelled material not absorbed by the node) increased, whereas the amount of labelled metabolites of low Mr was reduced considerably. During this period the Mr distribution of labelled HA in efferent outflow reverted to that of the infused material within 30-60 min. Our findings suggest that PG subunits and their chondroitin sulphate chains compete with HA for uptake into the peripheral lymph node of sheep. This indicates that PG, chondroitin sulphate and HA share the same pathway of elimination in this tissue, and is consistent with the view that the lymph node is involved in the metabolic turnover of normal intracellular matrix.

Low molecular mass proteoglycans (PG) were isolated from human articular cartilage and from pig laryngeal cartilage, which contained protein cores of similar size (Mr 40-44 kDa). However, the PG from human articular cartilage contained dermatan sulphate (DS) chains (50% chondroitinase AC resistant), whereas chains from pig laryngeal PG were longer and contained only chondroitin sulphate (CS). Disaccharide analysis after chondroitinase ABC digestion showed that the human DS-PG contained more 6-sulphated residues (34%) than the pig CS-PG (6%) and both contained fewer 6-sulphated residues than the corresponding high Mr aggregating CS-PGs from these tissues (86% and 20% from human and pig respectively). Cross-reaction of both proteoglycans with antibodies to bovine bone and skin DS-PG-II and human fibroblasts DS-PG suggested that the isolated proteoglycans were the humans DS-PG-II and pigs CS-PG-II homologues of the cloned and sequenced bovine proteoglycan. Polyclonal antibodies raised against the pig CS-PG-II were shown to cross-react with human DS-PG-II. SDS/polyacrylamide-gel analysis and immunoblotting of pig and human cartilage extracts showed that some free core protein was present in the tissues in addition to the intact proteoglycan. The antibodies were used in a competitive radioimmunoassay to determine the content of this low Mr proteoglycan in human cartilage extracts. Analysis of samples from 5-80 year-old humans showed highest content (approximately 4 mg/g wet wt.) in those from 15-25 year-olds and lower content (approximately 1 mg/g wet wt.) in older tissue (greater than 55 years). These changes in content may be related to the deposition and maintenance of the collagen fibre network with which this class of small proteoglycan has been shown to interact.

Dermatan sulphate proteoglycans were purified from juvenile human articular cartilage, with a yield of about 2 mg/g wet wt. of cartilage. Both dermatan sulphate proteoglycan I (DS-PGI) and dermatan sulphate proteoglycan II (DS-PGII) were identified and the former was present in greater abundance. The two proteoglycans could not be resolved by agarose/polyacrylamide-gel electrophoresis, but could be resolved by SDS/polyacrylamide-gel electrophoresis, which indicated average Mr values of 200,000 and 98,000 for DS-PGI and DS-PGII respectively. After digestion with chondroitin ABC lyase the Mr values of the core proteins were 44,000 for DS-PGI and 43,000 and 47,000 for DS-PGII, with the smaller core protein being predominant in DS-PGII. Sequence analysis of the N-terminal 20 amino acid residues reveals the presence of a single site for the potential substitution of dermatan sulphate at residue 4 of DS-PGII and two such sites at residues 5 and 10 for DS-PGI.

The N-terminal fragment (G1-G2) of cartilage proteoglycan protein core contains two globular domains, binding region (G1) and a second globular domain (G2), G1-G2 was isolated after mild trypsin digestion of purified proteoglycan aggregates followed by chromatography first on Sepharose CL-2B under associative conditions and then on a TSK-4000 column in 4 M-guanidinium chloride. It migrated as a single band (apparent Mr 150,000) on SDS/polyacrylamide-gel electrophoresis. G2 was isolated by V8-proteinase digestion of G1-G2 followed by aggregation of the G1-containing fragments with hyaluronate and chromatography on TSK-4000. It migrated as a single band on SDS/polyacrylamide-gel electrophoresis of apparent Mr 66,000 after digestion with keratanase. G2 did not interact with proteoglycan monomer, hyaluronate, link protein or other extractable cartilage matrix proteins. A polyclonal antibody raised against G2 did not cross-react with G1 or link protein. These data show that, despite a high degree of sequence similarity, G1 and G2 do not share any functional properties nor have major antigenic sites in common.

We have isolated from the conditioned medium of an established endothelial cell line a heparan sulphate proteoglycan whose involvement in the inhibition of the extrinsic coagulation pathway was reported in previous studies [Colburn & Buonassisi (1982) Biochem. Biophys. Res. Commun. 104, 220-227]. The proteoglycan was purified by gel filtration and ion-exchange chromatography, and appears to be free of contaminating proteins as determined by polyacrylamide-gel electrophoresis of the radioiodinated protein core before and after removal of the glycosaminoglycan chains by treatment with heparitinase. By this procedure the Mr of the protein core was estimated to be 22000. The N-terminal end was sequenced up to amino acid 25. The 21st residue is likely to be glycosylated. Analysis of the purified proteoglycan by gel-filtration chromatography yielded Kd values of 0.2 for the whole molecule and 0.35 for the glycosaminoglycan chains. The structure that emerges from these data is that of a heparan sulphate proteoglycan characterized by a relatively small protein core and few glycosaminoglycan chains.

Total RNA was extracted from the cartilage tissues rat Swarm chondrosarcoma, neonatal-rat breastplate and embryonic-chicken sterna and translated in wheat-germ cell-free reactions. The core protein of the chondroitin sulphate proteoglycan subunit was identified among translation products of rat mRNA by its apparent Mr of 330 000 and by its immunoprecipitation with specific antisera prepared against rat or chicken proteoglycan antigens. The apparent Mr of the rat proteoglycan core protein is 8000-10000 less than that of the equivalent chicken cartilage core-protein product.

Heparan sulphate proteoglycan was solubilized from human glomerular basement membranes by guanidine extraction and purified by ion-exchange chromatography and gel filtration. The yield of proteoglycan was approx. 2 mg/g of basement membrane. The glycoconjugate had an apparent molecular mass of 200-400 kDa and consisted of about 75% protein and 25% heparan sulphate. The amino acid composition was characterized by a high content of glycine, proline, alanine and glutamic acid. Hydrolysis with trifluoromethanesulphonic acid yielded core proteins of 160 and 110 kDa (and minor bands of 90 and 60 kDa). Alkaline NaBH4 treatment of the proteoglycan released heparan sulphate chains with an average molecular mass of 18 kDa. HNO2 oxidation of these chains yielded oligosaccharides of about 5 kDa, whereas heparitinase digestion resulted in a more complete degradation. The data suggest a clustering of N-sulphate groups in the peripheral regions of the glycosaminoglycan chains. A polyclonal antiserum raised against the intact proteoglycan showed reactivity against the core protein. It stained all basement membranes in an intense linear fashion in immunohistochemical studies on frozen kidney sections from man and various mammalian species.

Two dermatan sulphate-containing proteoglycans from bovine sclera were examined by rotary shadowing and electron microscopy, and the results were compared with previous biochemical findings. Both the large iduronate-poor proteoglycan (PGI) and the small iduronate-rich proteoglycan (PGII) possessed a globular proteinaceous region. Whereas PGI had a branched extension from the globular region, with five to eight side chains attached to it, PGII had only a single tail, which was of glycosaminoglycuronan. PGII aggregated via globular-region interactions, which were much diminished by reduction and alkylation. PGI aggregated via side chains and globular-region interactions. Although a few PGI aggregates were large, and similar to the hyaluronan-cartilage proteoglycan aggregates [Weidemann, Paulsson, Timpl, Engel & Heinegård (1984) Biochem. J. 224, 331-333], hyaluronan did not cause enhanced aggregation. PGII is very similar in shape to the small cartilage chondroitin sulphate proteoglycan, whereas PGI somewhat resembles the large cartilage chondroitin sulphate proteoglycan, although with many fewer glycosaminoglycan side chains, and probably only one globular region as opposed to two in the cartilage proteoglycan.

Fibroblasts in culture were incubated with [3H]leucine and [35S]sulphate for 1-24 h. A large glucuronic acid-rich and a small iduronic acid-rich dermatan sulphate proteoglycan were isolated with the use of isopycnic density-gradient centrifugation, ion-exchange and gel chromatography. After 3 h the accumulation in the cell layer of the small proteoglycan reached a steady state, whereas the large one continued to increase, albeit more slowly. In the medium both proteoglycans accumulated ‘linearly’, although the large one appeared somewhat later than the small one. The composition of the polysaccharide chains and the size of the protein cores did not vary during the experiment. The two proteoglycans were synthesized at approximately similar rates, but were distributed differently in the culture. The small proteoglycan was mainly confined to the medium, whereas the large one was found in the medium as well as in a cell-associated pool. There was an intracellular accumulation of iduronic acid-rich dermatan sulphate as free polysaccharides.

Proteoglycans on the cell surface play critical roles in the adhesion of fibroblasts to a fibronectin-containing extracellular matrix, including the model mouse cell line Balb/c 3T3. In order to evaluate the biochemistry of these processes, long-term [35S]sulphate-labelled proteoglycans were extracted quantitatively from the adhesion sites of 3T3 cells, after their EGTA-mediated detachment from the substratum, by using an extractant containing 1% octyl glucoside, 1 M-NaCl and 0.5 M-guanidinium chloride (GdnHCl) in buffer with many proteinase inhibitors. Greater than 90% of the material was identified as a large chondroitin sulphate proteoglycan (Kav. = 0.4 on a Sepharose CL2B column), and the remainder was identified as a smaller heparan sulphate proteoglycan; only small amounts of free chains of glycosaminoglycan were observed in these sites. These extracts were fractionated on DEAE-Sepharose columns under two different sets of elution conditions: with acetate buffer (termed DEAE-I) or with acetate buffer supplemented with 8 M-urea (termed DEAE-II). Under DEAE-I conditions about one-half of the material was eluted as a single peak and the remainder required 4 M-GdnHCl in order to recover it from the column; in contrast, greater than 90% of the material was eluted as a single peak from DEAE-II columns. Comparison of the elution of [35S]sulphate-labelled proteoglycan with that of 3H-labelled proteins from these two columns, as well as mixing experiments, indicated that the GdnHCl-sensitive proteoglycans were trapped at the top of columns, partially as a consequence of their association with proteins in these adhesion-site extracts. Affinity chromatography of these proteoglycans on columns of either immobilized platelet factor 4 or immobilized plasma fibronectin revealed that most of the chondroitin sulphate proteoglycan and the heparan sulphate proteoglycan bound to platelet factor 4 but that only the heparan sulphate proteoglycan bound to fibronectin, providing a ready means of separating the two proteoglycan classes. Affinity chromatography on octyl-Sepharose columns to test for hydrophobic domains in their core proteins demonstrated that a high proportion of the heparan sulphate proteoglycan but none of the chondroitin sulphate proteoglycan bound to the hydrophobic matrix. These results are discussed in light of the possible functional importance of the chondroitin sulphate proteoglycan in the detachment of cells from extracellular matrix and in light of previous affinity fractionations of proteoglycans from the substratum-adhesion sites of simian-virus-40-transformed 3T3 cells.

A chondroitin sulphate proteoglycan capable of forming large aggregates with hyaluronic acid was identified in cultures of human glial and glioma cells. The glial- cell- and glioma-cell-derived products were mutually indistinguishable and had some basic properties in common with the analogous chondroitin sulphate proteoglycan of cartilage: hydrodynamic size, dependence on a minimal size of hyaluronic acid for recognition, stabilization of aggregates by link protein, and precipitability with antibodies raised against bovine cartilage chondroitin sulphate proteoglycan. However, they differed in some aspects: lower buoyant density, larger, but fewer, chondroitin sulphate side chains, presence of iduronic acid-containing repeating units, and absence (less than 1%) of keratan sulphate. Apparently the major difference between glial/glioma and cartilage chondroitin sulphate proteoglycans relates to the glycan rather than to the protein moiety of the molecule.

Fibroblast proteoheparan sulphate has a disulphide-bonded subunit structure. The core protein appears to consist of two polypeptides each of Mr 80 000-100 000. As shown elsewhere [Carlstedt, Cöster, Malmström & Fransson (1983) J. Biol. Chem. in the press], both polypeptide molecules carry four to six heparan sulphate side chains (approx. Mr 20 000) and an unknown number of oligosaccharide units, giving the whole macromolecule an Mr in the range 300 000-400 000.