Attempted isolation of a heparin proteoglycan from bovine liver capsule

(1) Polysaccharides were isolated from bovine liver capsule by extraction with 2m-potassium chloride followed by precipitation from 0.8m-potassium chloride with cetylpyridinium chloride. Chondroitin sulphate was eliminated by digestion with hyaluronidase. The yield of heparin was approx. 40% of that obtained after extraction of the papain-digested tissue. (2) The macromolecular properties of the hyaluronidase-digested polysaccharide were studied by gel chromatography on Sephadex G-200 of the intact, as well as of the alkali-degraded, material. The results suggested the presence of single heparin chains in addition to a dermatan sulphate proteoglycan. (3) A purified heparin preparation was analysed for amino acids and neutral sugars. Xylose, galactose and serine were found in amounts corresponding to 0.1, 0.2, and 0.4 residue/polysaccharide chain (mol.wt. 7400), respectively. It is suggested that the isolated material had been degraded by a polysaccharidase with endo-enzyme properties.     

Evidence for the existence of a multichain proteoglycan of heparan sulphate

1. Glycosaminoglycans were extracted with 2m-potassium chloride from bovine aorta and purified by precipitation with cetylpyridinium chloride from 0.5m-potassium chloride. The yield amounted to 24% of the total glycosaminoglycan content of the tissue. 2. After removal of chondroitin sulphate by digestion with testicular hyaluronidase, the residual glycosaminoglycan material (11% of the extracted polysaccharide) was fractionated by gel chromatography on Sephadex G-200. Two peaks (I and II) were obtained, the more retarded of which (II) corresponded to single polysaccharide chains. 3. The macromolecular properties of fraction I were investigated by repeated gel chromatography, after treatment of the fraction with alkali or digestion with papain. In both cases the elution position of fraction I was shifted towards that of the single polysaccharide chains. 4. Analysis of fraction I showed approximately equal amounts of heparan sulphate and dermatan sulphate. It is concluded that these glycosaminoglycans both occur in the aortic wall as multichain proteoglycans.     

Purification and characterisation of a minor low-sulphated dermatan sulphate-proteoglycan from ray skin.

A minor low-sulphated dermatan sulphate proteoglycan was isolated from ray skin by extraction with 2% sodium dodecyl sulphate, followed with ion-exchange chromatography, gel chromatography and density gradient centrifugation. The proteoglycan with a relative molecular mass (Mr) ranging from 70 to 120 kDa is composed of about two dermatan sulphate chains (Mr 33 kDa) bound on a protein core of Mr 27 kDa, and oligosaccharides consisting of uronic acids, hexosamines and neutral sugars. The major amino acids of the protein core were glycine (corresponding to about one-fourth of the total amino acids), serine, threonine, glutamic acid/glutamine, leucine and cysteine, together amounting to 56% of the total. The isolated proteoglycan does not interact with hyaluronic acid and does not form self-aggregates. Dermatan sulphate was rich in iduronic acid (62% of total uronic acid) and composed of non-sulphated (44%), and mono-sulphated disaccharides bearing esterified sulphate groups at positions C-4 (53%) or C-6 (3%) of the N-acetyl galactosamine. HPLC analysis of a pure preparation of dermatan sulphate, showed the presence of galactose and glucose possibly as branches on the dermatan sulphate chain.     

Isolation of a chondroitin sulfate--dermatan sulfate proteoglycan from bovine aorta.

Material containing proteoglycans was extracted from bovine aorta by the dissociative solvent 3.0 m MgCl². The proteoglycan that remained in solution at low ionic strength was purified by isopycnic CsCl centrifugation (?, 1.75 – 1.89 g/ml). From the lower third of the gradient a proteoglycan was isolated which behaved as a homogeneous material when analyzed by the ultracentrifuge and by electrophoresis on cellulose acetate. The proteoglycan contained 12% protein, 21% uronic acid, and 28% hexosamine. Analyses by hyaluronidase digestion and gas-liquid chromatography of the polysaccharide moieties of the proteoglycan showed a composition of 56% chondroitin 6-sulfate, 20% chondroitin 4-sulfate, and 7% dermatan sulfate. A copolymeric structure for the polysaccharide of the proteoglycan is proposed.     

Antithrombin activity and disaccharide composition of dermatan sulfate from different bovine tissues

Dermatan sulfate is a glycosaminoglycan that selectively inhibits the action of thrombin through interaction with heparin cofactor II. Unlike heparin it does not interact with other coagulation factors and is able to inhibit thrombin associated with clots. This property has made dermatan sulfate an attractive candidate as an antithrombotic drug. Previous studies have showed that dermatan sulfate derived from porcine/bovine intestinal mucosa/skin or marine invertebrates is capable of stimulating heparin cofactor II-mediated thrombin inhibition in vitro. This biological activity is reported for the first time in this study using dermatan sulfate derived from mammalian tissues other than intestinal mucosa or skin. Ten different bovine tissues including the aorta, diaphragm, eyes, large and small intestine, esophagus, skin, tendon, tongue, and tongue skin were used to prepare dermatan sulfate-enriched fractions by anion exchange chromatography and acetone precipitation. Heparin cofactor II/dermatan sulfate-mediated thrombin inhibition measured in vitro revealed activity comparable to or higher than the commercial standard with 2-fold differences observed between some tissues. Analysis of the extracted dermatan sulfate using fluorophore-assisted carbohydrate electrophoresis revealed significant differences in the relative percentage of all the mono-sulfated disaccharides, in particular the predominant mammalian disaccharide uronic acid-->N-acetyl-D-galactosamine-4-O-sulfate, confirming previous reports regarding variations in sulfation in dermatan sulfate from different tissues. Overall, these findings demonstrate that dermatan sulfate extracted from a range of bovine tissues exhibits in vitro antithrombin activity equivalent to or higher than that observed for porcine intestinal mucosa, identifying additional sources of dermatan sulfate as potential antithrombotic agents.     

The synthesis of dermatan sulphate proteoglycans by fetal and adult human articular cartilage.

Non-aggregating dermatan sulphate proteoglycans can be extracted from both fetal and adult human articular cartilage. The dermatan sulphate proteoglycans appear to be smaller in the adult, this presumably being due to shorter glycosaminoglycan chains, and these chains contain a greater proportion of their uronic acid residues as iduronate. Both the adult and fetal dermatan sulphate proteoglycans contain a greater amount of 4-sulphation than 6-sulphation of the N-acetylgalactosamine residues, in contrast with the aggregating proteoglycans, which always show more 6-sulphation on their chondroitin sulphate chains. In the fetus the major dermatan sulphate proteoglycan to be synthesized is DS-PGI, though DS-PGII is synthesized in reasonable amounts. In the adult, however, DS-PGI synthesis is barely detectable relative to DS-PGII, which is still synthesized in substantial amounts. Purification of the dermatan sulphate proteoglycans from adult cartilage is hampered by the presence of degradation products derived from the large aggregating proteoglycans, which possess similar charge, size and density properties, but which can be distinguished by their ability to interact with hyaluronic acid.     

Isolation and some structural analyses of a proteodermatan sulphate from calf skin.

A proteodermatan sulphate was isolated from 0.15 M-NaCl and 0.45 M-NaCl extracts of newborn-calf skin. The proteoglycan was separated from collagen and hyaluronic acid by precipitation with cetylpyridinium chloride and CsCl-density-gradient centrifugation. Further purification was performed by ion-exchange, affinity and molecular-sieve chromatography. The proteoglycan bound to concanavalin A-Sepharose in 1 M-NaCl. It gave a positive reaction with periodic acid/Schiff reagent and contained 8.3% of uronic acid. The dermatan sulphate, the only glycosaminoglycan component, was composed of 74% iduronosylhexosamine units and 26% glucuronosylhexosamine units. The Mr was assessed to be 15000-20000 by gel chromatography. The core protein was found to be a sialoglycoprotein that had O-glycosidic oligosaccharides with N-acetylgalactosamine at the reducing termini. The molar ratio of oligosaccharide chains to dermatan sulphate was approx. 3:1. From these results the proposed structure of proteodermatan sulphate is: one dermatan sulphate chain (average Mr 17500), three O-glycosidic oligosaccharide chains and probably N-glycosidic oligosaccharide chain(s) bound to one core-protein molecule (Mr 55000).     

A glycomic approach to proteoglycan with a two-dimensional polysaccharide chain map

Glycosaminoglycan chains were liberated from proteoglycans (bovine lung, tracheal cartilage, and cerebrum) by successive digestion with actinase and with cellulase from Aspergillus niger, which has endo-beta-xylosidase activity. The glycosaminoglycan chains were fluorescence-labeled with 2-aminopyridine after digestion with Streptomyces hyaluronidase. The resulting pyridylamino-glycosaminoglycans, including heparan sulfate, chondroitin sulfate/dermatan sulfate, and heparin, were separated by high-performance liquid chromatography. Each separated fraction was analyzed by two types of high-performance liquid chromatography: gel-filtration chromatography and anion-exchange chromatography. The correlation between molecular weight and degree of sulfation could be shown on the two-dimensional polysaccharide chain map. Use of a commonly available cellulase with endo-beta-xylosidase activity together with the two-dimensional polysaccharide chain map allows easy analysis of various glycosaminoglycan chains and comprehensive comparison among the structures. These techniques will become useful tools in the further development of glycotechnology and glycome analysis.     

Distribution of glucuronic and iduronic acid units in heparin chains

The distribution of glucuronic and iduronic acid within the chains of anticoagulantly active and inactive beef lung heparin was investigated. A fraction with an average molecular weight of 19,500 was isolated from the heterodisperse mixture and then separated into active and inactive components by affinity chromatography. Each sample was linked through its reducing terminus to tyramine, reduced with sodium borotritide, and bound covalently to Sepharose via an azo bridge. The bound reduced heparin was treated with a limited amount of HNO2 and the degraded fragments were removed. The sections of the chain contiguous with the original reducing terminus were then detached from the insoluble matrix by reaction with sodium dithionite. The recovered polysaccharide was fractionated according to size on Sephadex G-200 and the amount of each uronic acid in the individual fractions was determined. Inactive heparin showed a constant percentage of glucuronic acid in all fragments, i.e. about 8.9% of the total uronic acid. With active heparin the percentage of glucuronic acid increased with the distance from the reducing terminus of the polysaccharide chain, ranging from 9.5 to 20% of the uronic acids. These results suggest that the biosynthesis of active heparin involves unique reactions or specific processing of the macromolecule.     

Isolation and characterization of a collagen fibril-associated dermatan sulphate proteoglycan from bovine lung

Dermatan sulphate proteoglycans have been extracted from bovine lung with 2.0 M CaCl2 and isolated using CsCl density gradient centrifugation, DEAE ion-exchange chromatography, gel chromatography and preparative sodium dodecyl sulphate-polyacrylamide gel electrophoresis. Ultrastructurally these proteoglycans are specifically associated with collagen fibrils. Dermatan sulphate (Mr 15.10(3)-35.10(3), with a strong prevalence for the higher Mr) is link via an O-glycosidic bond to a protein core, which is rich in Asx, Glx and Leu. Of the total uronic acid, 91% is iduronic acid. A part of the glucuronic acid residues is located near the protein core and a large cluster of disaccharides is devoid of glucuronic acid residues. An inhibition enzyme immunoassay has been developed to quantitate the proteoglycan. A model for the interaction between dermatan sulphate proteoglycans and collagen fibrils is proposed.     

Determination of the distribution of D-glucuronic acid units within the chain of pig-skin dermatan sulfate near the linkage region

A method for analyzing the distribution of D-glucuronic acid units within the chain and near the linkage region of dermatan sulfate has been developed. The method consists of a chemical modification of the reducing terminal residue in the polysaccharide by reductive amination with excess 1,2-diaminoethane in the presence of sodium cyanoborohydride, desulfative fragmentation of the polysaccharide, labeled with 2-aminoethylamino (AEA) groups, in hot dimethyl sulfoxide containing 10% of water followed by 2,4-dinitrophenylation of the 2-aminoethylamino group, separation of the 2-(2,4-dinitrophenylamino)ethylamino labeled dermatan fragments from nonlabeled fragments on Octyl-Sepharose CL-4B gel, and determination of the uronic acid composition of the labeled fragments having various chain-length. A preparation of pig-skin dermatan sulfate (Mr 21,000, ratio of GlcA to total uronic acid, 93:500) showed an average distribution pattern of D-glucuronic acid residues near the linkage region of one N-acetylchondrosine unit in the disaccharide sequence 1-5(6) linked to the Xyl----Gal----Gal----GlcA residue, a cluster of 6-8 N-acetyldermosine units in the sequence 6(7)-12(13), and four separate N-acetylchondrosine units between the sequence adjacent to the N-acetyldermosine cluster and the sequence 23 or higher.     

Fractionation of proteoglycans from bovine corneal stroma.

Proteoglycans were extracted from bovine corneal stroma with 4M-guanidinum chloride, purified by DEAE-dellulose chromatography (Antonopoulos et al., 1974) and fractionated by precipitation with ethanol into three fractions of approximately equal weight. One of these fractions consisted of a proteoglycan that contained keratan sulphate as the only glycosaminoglycan. In the othertwo fractions proteoglycans that contained chondroitin sulphate, dermatan sulphate and keratan sulphate were present. Proteoglycans which had a more than tenfold excess of galactosaminoglycans over keratan sulphate could be obtianed by further subfractionation. The gel-chromatographic patterns of the glucosaminoglycans before and after digestion with chondroitinase AC differed for the fractions. The individual chondroitin sulphate chains seemed to be larger in cornea than in cartilage. Oligosaccharides, possibly covalently linked to the protein core of the proteoglycans, could be isolated from all fractions. The corneal proteoglycans were shown to have higher protein contents and to be of smaller molecular size than cartilage proteoglycans.     

Human liver iduronate-2-sulphatase. Purification, characterization and catalytic properties.

Human iduronate-2-sulphatase (EC 3.1.6.13), which is involved in the lysosomal degradation of the glycosaminoglycans heparan sulphate and dermatan sulphate, was purified more than 500,000-fold in 5% yield from liver with a six-step column procedure, which consisted of a concanavalin A-Sepharose-Blue A-agarose coupled step, chromatofocusing, gel filtration on TSK HW 50S-Fractogel, hydrophobic separation on phenyl-Sepharose CL-4B and size separation on TSK G3000SW Ultrapac. Two major forms were identified. Form A and form B, with pI values of 4.5 and less than 4.0 respectively, separated at the chromatofocusing step in approximately equal amounts of recovered enzyme activity. By gel-filtration methods form A had a native molecular mass in the range 42-65 kDa. When analysed by SDS/PAGE, dithioerythritol-reduced and non-reduced form A and form B consistently contained polypeptides of molecular masses 42 kDa and 14 kDa. Iduronate-2-sulphatase was purified from human kidney, placenta and lung, and form A was shown to have similar native molecular mass and subunit components to those observed for liver enzyme. Both forms of liver iduronate-2-sulphatase were active towards a variety of substrates derived from heparin and dermatan sulphate. Kinetic parameters (Km and Kcat) of form A were determined with a variety of substrates matching structural aspects of the physiological substrates in vivo, namely heparan sulphate, heparin and dermatan sulphate. Substrate with 6-sulphate esters on the aglycone residue adjacent to the iduronic acid 2-sulphate residue being attack were hydrolysed with catalytic efficiencies up to 200 times above that observed for the simplest disaccharide substrate without a 6-sulphated aglycone residue. The effect of incubation pH on enzyme activity towards the variety of substrates evaluated was complex and dependent on substrate aglycone structure, substrate concentration, buffer type and the presence of other proteins. Sulphate and phosphate ions and a number of substrate and product analogues were potent inhibitor of form A and form B enzyme activities.     

The effect of preganglionic stimulation on the incorporation of l-[U-14C]valine into the protein of the superior cervical ganglion of the guinea pig

The acid glycosaminoglycans were extracted from the skins of young rats less than 1 day post partum. The isolated products were fractionated by a cetylpyridinium chloride-cellulose column technique and identified by chemical analysis, electrophoretic mobility and susceptibility to testicular hyaluronidase digestion. Hyaluronic acid (56%) dermatan sulphate (15.6%) and chondroitin 6-sulphate (9.1%) were the major components, but chondroitin 4-sulphate, heparan sulphate and heparin were also present, together with two further fractions tentatively suggested to be a heparan sulphate-like fraction and a dermatan sulphate fraction, both of short chain length or low degree of sulphation.     

Proteoglycans of bovine periodontal ligament and skin. Occurrence of different hybrid-sulphated galactosaminoglycans in distinct proteoglycans.

A proteoglycan purified from 4 M-guanidinium chloride extracts of bovine periodontal ligament closely resembled that of bovine skin, except for a rather lower protein content and a higher molecular weight (120 000 compared with about 90 000) by sodium dodecyl sulphate/polyacrylamide-gel electrophoresis. The latter difference was explained by the molecular weights (29 000 and 16 000) of the respective dermatan sulphate components, each of which was rich in L-iduronate (about 75% of the total hexuronate). Significant amounts of other glycosaminoglycans did not occur in these proteoglycans, which were homogenous on gel chromatography and agarose/polyacrylamide-gel electrophoresis. Polydispersity was observed in sedimentation equilibrium experiments, but proteolysis or self-association of the proteodermatan sulphates may have affected these results. Ligament proteoglycans that were almost completely extracted with 0.1 M-NaCl contained less protein of a completely different amino acid composition than the proteodermatan sulphates. They were heterogeneous in size but generally smaller than cartilage proteoglycans and L-iduronate was a component, comprising about 7% of the total hexuronate of the sulphated galactosaminoglycan chains. The latter consisted of two fractions differing in molecular weight, but a dermatan sulphate with a high L-iduronate content was not present. These proteoglycans had some resemblance to D-glucuronate-rich proteoglycans of other non-cartilaginous tissues. Such compounds, however, are difficult to categorize at present.     

1H-n.m.r. investigation of naturally occurring and chemically oversulphated dermatan sulphates. Identification of minor monosaccharide residues.

The 1H-n.m.r. spectra of various dermatan sulphate preparations present, besides the major signals of the basic disaccharide unit, several other minor signals. We have assigned most of them by n.m.r., using two-dimensional proton-proton double-quantum-correlation and nuclear-Overhauser-effect spectroscopy experiments. This allowed us to identify 2-O-sulphated L-iduronic acid and D-glucuronic acid residues as well as 6-sulphated N-acetylgalactosamine (presumably 4-O-sulphated as well). 2-O-Sulphated iduronic acid was present to similar extents (6-10% of total uronic acids) in pig skin dermatan sulphate and pig intestine dermatan sulphate, whereas glucuronic acid represented 17% of the uronic acid of pig skin dermatan sulphate and was virtually absent (1%) from the other preparation. 6-O-Sulphated N-acetylgalactosamine was present in minor amounts in pig intestine dermatan sulphate only. The influence of sulphation of iduronic acid units on their conformation was assessed by using chemically oversulphated pig intestine dermatan sulphate. Introduction of sulphate groups in this unit in dermatan sulphate tends to shift the conformational equilibrium towards the 1C4 conformer.     

A heparin-binding protein involved in inhibition of smooth-muscle cell proliferation.

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 nature of the protein moieties of cartilage proteoglycans of pig and ox.

Proteoglycans extracted with 4M-guanidinium chloride from pig laryngeal cartilage and bovine nasal septum were purified by density-gradient centrifugation in CsCl under 'associative' followed by 'dissociative' conditions [Hascall & Sajdera (1969) J. Biol. Chem. 244, 2384-2396]. Proteoglycans were then digested exhaustively with testicular hyaluronidase, which removed about 80% of the chondroitin sulphate. The hyaluronidase was purified until no proteolytic activity was detectable under the conditions used for digestion. The resulting 'core' proteins of both species were fractionated by a sequence of gel-chromatographic procedures which gave four major fractions of decreasing hydrodynamic size. Those that on electrophoresis penetrated 5.6% (w/v) polyacrylamide gels migrated as discrete bands whose mobility increased with decreasing hydrodynamic size. The unfractionated 'core' proteins had the same N-terminal amino acids as the intact proteoglycan, suggesting that no peptide bonds had been cleaved during hyaluronidase digestion. Alanine predominated as the N-terminal residue in all the fractions of both species. Fractions were analysed for amino acid, amino sugar, uronic acid and neutral sugar compositions. In pig 'core' proteins, the glutamic acid content increased significantly with hydrodynamic size, but in bovine 'core' proteins this trend was less marked. Significant differences in amino acid composition between fractions suggested that in each species there was more than one variety of proteoglycan. The molar proportions of xylose to serine destroyed on alkaline beta-elimination were equivalent in most fractions, indicating that the serine residues destroyed were attached to the terminal xylose of chondroitin sulphate chains. The ratio of serine residues to threonine residues destroyed on beta-elimination, was similar in all fractions of both species. Since the fractions of smallest hydrodynamic size contained less keratan sulphate than those of larger size, it implies that in the former the keratan sulphate chains were shorter than in the latter.     

Molecular distinctions between heparan sulphate and heparin. Analysis of sulphation patterns indicates that heparan sulphate and heparin are separate families of N-sulphated polysaccharides.

Heparan sulphate and heparin are chemically related alpha beta-linked glycosaminoglycans composed of alternating sequences of glucosamine and uronic acid. The amino sugars may be N-acetylated or N-sulphated, and the latter substituent is unique to these two polysaccharides. Although there is general agreement that heparan sulphate is usually less sulphated than heparin, reproducible differences in their molecular structure have been difficult to identify. We suggest that this is because most of the analytical data have been obtained with degraded materials that are not necessarily representative of complete polysaccharide chains. In the present study intact heparan sulphates, labelled biosynthetically with [3H]glucosamine and Na2(35)SO4, were isolated from the surface membranes of several types of cells in culture. The polysaccharide structure was analysed by complete HNO2 hydrolysis followed by fractionation of the products by gel filtration and high-voltage electrophoresis. Results showed that in all heparan sulphates there were approximately equal numbers of N-sulpho and N-acetyl substituents, arranged in a similar, predominantly segregated, manner along the polysaccharide chain. O-Sulphate groups were in close proximity to the N-sulphate groups but, unlike the latter, the number of O-sulphate groups could vary considerably in heparan sulphates of different cellular origins ranging from 20 to 75 O-sulphate groups per 100 disaccharide units. Inspection of the published data on heparin showed that the N-sulphate frequency was very high (greater than 80% of the glucosamine residues are N-sulphated) and the concentration of O-sulphate groups exceeded that of the N-sulphate groups. We conclude from these and other observations that heparan sulphate and heparin are separate families of N-sulphated glycosaminoglycans.     

Studies on the incorporation of [U-14C]glucose and [35S]sulphate into the acid glycosaminoglycans of neonatal rat skin

1. The incorporation of [(35)S]sulphate in vivo into the acid-soluble intermediates extracted from young rat skin showed three sulphated hexosamine-containing components. 2. The rates of synthesis of these components were determined in vivo by measuring the incorporation of radioactivity from [U-(14)C]glucose into their isolated hexosamine moieties. 3. The incorporation of radioactivity from [U-(14)C]glucose into the isolated hexosamine and uronic acid moieties of the acid glycosaminoglycans was also measured. These results, combined with those obtained on the intermediary pathways of hexosamine and uronic acid biosynthesis previously determined in this tissue, indicated that the acid-soluble sulphated hexosamine-containing components were not precursors of the sulphated hexosamine found in the acid glycosaminoglycans. 4. The rates of synthesis of the acid glycosaminoglycan fractions were calculated from the incorporation of radioactivity from [U-(14)C]glucose into the hexosamine moiety. The sulphated components containing principally dermatan sulphate, chondroitin 6-sulphate and in smaller amounts, chondroitin 4-sulphate, heparan sulphate and heparin appeared to be turning over about twice as rapidly as hyaluronic acid and about four times as rapidly as the small keratan sulphate fraction. The relative rates of synthesis of the sulphated glycosaminoglycans were calculated from the incorporation of [(35)S]sulphate and were in agreement with those from (14)C-labelling studies.