Two linkage-region fragments isolated from skeletal keratan sulphate contain a sulphated N-acetylglucosamine residue.

Peptido-keratan sulphate fragments were isolated from the nucleus pulposus of bovine intervertebral discs (6-year-old animals) after chondroitin ABC lyase digestion followed by digestion of A1D1 proteoglycans by diphenylcarbamoyl chloride-treated trypsin and gel-permeation chromatography on Sepharose CL-6B. Treatment of these peptido-keratan sulphate fragments with alkaline NaB3H4 yielded keratan sulphate chains with [3H]galactosaminitol end-labels, and these chains were further purified by gel-permeation chromatography on Sephadex G-50 and ion-exchange chromatography on a Pharmacia Mono-Q column in order to exclude any contamination with O-linked oligosaccharides. The chains were then treated with keratanase, and the digest was chromatographed on a Bio-Gel P-4 column followed by anion-exchange chromatography on a Nucleosil 5 SB column. Two oligosaccharides, each representing 18% of the recovered radiolabel, were examined by 500 MHz 1H-n.m.r. spectroscopy, and shown to have the following structures: [formula: see text] The structure of oligosaccharide (I) confirms the N-acetylneuraminylgalactose substitution at position 3 of N-acetylgalactosamine in the keratan sulphate-protein linkage region found by Hopwood & Robinson [(1974) Biochem. J. 141, 57-69] but additionally shows the presence of a 6-sulphated N-acetylglucosamine. Electron micro-probe analysis specifically confirmed the presence of sulphur in this sample. This sulphate ester group differentiates the keratan sulphate linkage region from similar structures derived from O-linked oligosaccharides [Lohmander, De Luca, Nilsson, Hascall, Caputo, Kimura & Heinegård (1980) J. Biol. Chem. 255, 6084-6091].     

Studies of a monoclonal antibody to skeletal keratan sulphate. Importance of antibody valency.

A mouse monoclonal antibody (AN9P1) to keratan sulphate is described. In a competitive-inhibition solution-phase radioimmunoassay employing 125I-labelled intact proteoglycan, it reacts preferentially with keratan sulphate bound to the core protein of adult human articular-cartilage proteoglycan and to a much lesser degree with keratan sulphate purified from this proteoglycan. Proteolytic cleavage of the proteoglycan by pepsin and trypsin has little effect on antibody binding, but treatment with papain decreases binding considerably and more than does treatment with keratanase. An even greater decrease in binding is observed after treatment with alkaline borohydride. A comparison of binding of antibody AN9P1 with that of another previously described monoclonal antibody, 1/20/5-D-4, to keratan sulphate [Caterson, Christner & Baker (1983) J. Biol. Chem. 258, 8848-8854] revealed similar binding characteristics, both showing much diminished binding after papain digestion of proteoglycan and even less with purified skeletal keratan sulphate. Removal of the Fc piece of antibody AN9P1 had no significant effect on the differential binding of divalent F(ab')2 fragment to proteoglycan, to papain-digested proteoglycan and to keratan sulphate, although there was a small decrease in binding to papain-digested proteoglycan. Conversion of the antibody into univalent Fab fragment with removal of the Fc piece resulted in diminished binding to proteoglycan, compared with that observed with IgG, and in enhanced binding to free keratan sulphate and to papain-digested proteoglycan. These results suggest that close proximity of keratan sulphate chains on the core protein of proteoglycans favours preferential reactivity of bivalent antibody with these species through cross-bridging of chains by antibody. Conversely, much decreased binding to keratan sulphate on proteoglycan core-protein fragments and to free keratan sulphate results from a lack of close proximity of keratan sulphate. By using univalent Fab fragment in these assays these differences in binding are minimized by preventing cross-bridging and thereby enhancing detection of smaller fragments without sacrificing too much sensitivity of detection of larger proteoglycan species. The persistent preferential binding of Fab fragment to proteoglycan is probably in part the result of the increased epitope density in the intact molecule compared with keratan sulphate in a more disperse form.     

Conformation of keratan sulphate

X-ray diffraction patterns from stretched films of keratan sulphate, isolated from bovine cornea, indicate that the molecules are twofold helices with an axial rise per disaccharide residue of 0.945 nm. These helices are oriented with their twofold screw axes parallel and about equally spaced, but are not further organized into regular crystalline arrays. Computer methods were used to construct a molecular model with the observed symmetry and axial rise per disaccharide residue, with standard bond lengths, bond angles and pyranose ring conformations and with a hydrogen bond of length 0.270 nm between O(3) of N-acetylglucosamine and O(5) of galactose. This model has no unacceptably short non-bonded interatomic distances. The intensity distribution of the diffraction pattern calculated from the co-ordinates of the model is in reasonable agreement with the observed intensity distribution. This keratan sulphate model is an extended polysaccharide chain fringed with charged sulphate side groups, and is similar to those that have already been reported for chondroitin sulphates and dermatan sulphate, paralleling the similarities in covalent structure and biological occurrence among these substances.     

Characterization of the keratan sulphate proteoglycans from bovine corneal stroma.

The keratan sulphate proteoglycans that can be prepared from bovine corneal stroma [Axelsson & Heineg?rd (1975) Biochem. J. 145, 491-500] were characterized by gel chromatography, gel electrophoresis and analytical ultracentrifugation in associative (0.6 M-NaCl) and dissociative (6M-guanidinum chloride) solvents. The proteoglycans aggreagated at low salt concentrations and pH. The weight-average molecular weight of the monomer proteoglycans was established. Keratan sulphate peptides and oligosaccharide peptides were isolated after proteolysis. Their composition indicated that both are linked to protein via asparagine residues. A tentative model for corneal keratan sulphate proteoglycans is suggested.     

Isolation and partial characterization of keratan sulphate from human brain

A sulphated glycoconjugate was isolated from adult human brain from a glycosaminoglycan fraction which was not precipitated with 1% cetylpyridinium chloride or ethanol below 50% concentration. It appeared heterogeneous on gel filtration, exhibiting a molecular weight range from about 7000 to over 10 000. Its main covalent structure was shown to contain sulphated, repeating disaccharide units of (beta-D-galactose-(1----4)-N-acetyl-D-glucosamine-(1----3)). In addition, it was susceptible to degradation by keratan sulphate endo-beta-galactosidase and thus was assumed to be keratan sulphate.     

Sulphate release from keratan sulphate and heparan sulphate by bovine N-acetylglucosamine-6-sulphate sulphatase

The functions of sulphated monosaccharides within glycosaminoglycans(GAGs) and glycoproteins are being studied intensely, but progressis hindered by an inability to selectively desulphate glycoconjugates.We recently identified an N-acetylglucosamine-6-sulphate sulphatase(NG6SS) from bovine kidney that can remove sulphate from N-acetylglucosamine-6-sulphate(GlcNAc-6-SO₄) within oligosaccharides and glycoproteins. However,the potential ‘endosulphatase’ activity of the NG6SStoward GAGs is not known. To test for this possibility, [³H]glucosamine-,[³H]galactose- and ³⁵SO_4- labelled keratan sulphate (KS) wereseparately prepared by metabolic radiolabelling of bovine cornea.NG6SS quantitatively removed sulphate from KS without releaseof sugar fragments. The enzyme had a K_m of 4.7 mM toward freeGlcNAc-6-SO₄, but its K_m for commercially available bovine cornealKS was found to be 9.1 µM. Analyses of both KS and heparansulphate after treatment with NG6SS demonstrated significantloss of sulphate from GlcNAc-6-SO₄ in both GAGs. These findingsmay be relevant for future studies aimed at defining the function(s)of GlcNAc-6-SO₄ residues in GAGs and understanding the catabolismof GAGs, especially in regard to sulphatidoses, such as SanfilippoD syndrome in humans, which involves a deficiency of NG6SS activity catabolism endosulphatase glycosaminoglycans sulphation     

Skeletal keratan sulfate from different tissues. Characterization and alkaline degradation.

Keratan sulfate-rich peptides were isolated after digestion of proteoglycans from bovine nasal cartilage and bovine nucleus pulposus with chondroitinase ABC, trypsin and chymotrypsin. The keratan sulfate enriched peptides from nucleus pulposus were larger than those from nasal cartilage. Keratan sulfate chains were isolated after treatment of the keratan sulfate-rich peptides under alkaline, reductive conditions. Proteoglycans from nucleus pulposus contain longer keratan sulfate chains, as is shown primarily by gel chromatography of the keratan sulfate-rich peptides and the keratan sulfate chains, but also from end-group analyses of the keratan sulfate chains.     

Degradation of keratan sulphate by beta-N-acetylhexosaminidases A and B.

Enzymic cleavage of beta-N-acetylglucosamine residues of keratan sulphate was studied in vitro by using substrate a [3H]glucosamine-labelled desulphated keratan sulphate with N-acetylglucosamine residues at the non-reducing end. Both lysosomal beta-N-acetylhexosaminidases A and B are proposed to participate in the degradation of keratan sulphate on the basis of the following observations. Homogenates of fibroblasts from patients with Sandhoff disease, but not those from patients with Tay--Sachs disease, were unable to release significant amounts of N-acetyl[3H]glucosamine. On isoelectric focusing of beta-N-acetylhexosaminidase from human liver the peaks of keratan sulphate-degrading activity coincided with the activity towards p-nitrophenyl beta-N-acetylglucosaminide. A monospecific antibody against the human enzyme reacted with both enzyme forms and precipitated the keratan sulphate-degrading activity. Both isoenzymes had the same apparent Km of 4mM, but the B form was approximately twice as active as the A form when compared with the activity towards a chromogenic substrate. Differences were noted in the pH--activity profiles of both isoenzymes. Thermal inactivation of isoenzyme B was less pronounced towards the polymeric substrate than towards the p-nitrophenyl derivative.     

Menstrual-cycle-dependent expression of keratan sulphate in human endometrium.

Immunochemical methods have been used to detect and characterize two classes of polypeptide-associated keratan sulphate (KS) in epithelial secretions from human endometrium. Monoclonal antibody D9B1 binds to a hormonally regulated sialylated epitope associated with KS in a high relative molecular mass (250,000-350,000) component that bands as a doublet in SDS/PAGE. These KS chain(s) are sensitive to keratanase, endo-beta-galactosidase and N-glycanase. A second, more highly sulphated, type of KS is also present, that is resistant to all three enzymes. This can be detected using monoclonal antibody 5D4. It is present throughout the menstrual cycle and is associated principally with a component of Mr 140,000. Thus secretory KS contributes to the environment of the implanting embryo, may be used as a molecular index of endometrial function and could be important in the establishment of pregnancy.     

Immunohistochemical staining for chondroitin sulphate and keratan sulphate. An evaluation of two monoclonal antibodies

Immunohistochemical staining with commercially available antibodies against chondroitin sulphate (clone CS-56) and keratan sulphate (clone 1/20/5-D-4) was compared with two conventional histochemical methods for the demonstration of glycosaminoglycans, namely Alcian Blue with varying pH and critical electrolyte concentrations, and a modified PAS stain. The antibodies were tested on sections from both frozen and fixed, paraffin embedded human material from umbilical cord, skin, and bronchus. The results showed immunostaining to function equally well on frozen and routine sections, and to be superior to Alcian Blue and PAS with regard to morphological detail. Thus, reactivity with anti-chondroitin sulphate was demonstrated in vessel walls, in small nerves, in the basal membrane zone of the skin, in perichondrium, and in and around chondrocytes. Reactivity with anti-keratan sulphate occurred in chondroid matrix and in perichondrial tissue; however, some cells of the bronchial epithelium and mucous glands also exhibited positivity.     

Immunohistochemical staining for chondroitin sulphate and keratan sulphate. An evaluation of two monoclonal antibodies

Summary Immunohistochemical staining with commercially available antibodies against chondroitin sulphate (clone CS-56) and keratan sulphate (clone 1/20/5-D-4) was compared with two conventional histochemical methods for the demonstration of glycosaminoglycans, namely Alcian Blue with varying pH and critical electrolyte concentrations, and a modified PAS stain. The antibodies were tested on sections from both frozen and fixed, paraffin embedded human material from umbilical cord, skin, and bronchus. The results showed immunostaining to function equally well on frozen and routine sections, and to be superior to Alcian Blue and PAS with regard to morphological detail. Thus, reactivity with anti-chondroitin sulphate was demonstrated in vessel walls, in small nerves, in the basal membrane zone of the skin, in perichondrium, and in and around chondrocytes. Reactivity with anti-keratan sulphate occurred in chondroid matrix and in perichondrial tissue; however, some cells of the bronchial epithelium and mucous glands also exhibited positivity.     

The structure and composition of cartilage keratan sulphate

Keratan sulphate was isolated from bovine intervertebral disc and bovine nasal septum after hydrolysis with proteinases and treatment with dilute alkali. Each preparation was found to contain, per keratan sulphate chain: (a) 1 residue of mannose; (b) 3 residues of N-acetylneuraminic acid (2 residues after alkali treatment); (c) 1 residue of N-acetylgalactosamine (lost after alkali treatment); (d) 1 residue or less of fucose. N-Acetyl-neuraminic acid residues were at non-reducing termini and were bonded to keratan sulphate through galactose residues. Evidence is presented for two different types of linkage between skeletal keratan sulphate and protein. Consideration of molecular parameters and compositions leads to a proposed structure for keratan sulphate-protein as found in skeletal proteoglycans.     

Impaired degradation of keratan sulphate by Morquio A fibroblasts.

Upon incubation of keratan [35S]sulphate with normal fibroblasts both [35S]sulphate and N-acetylglucosamine 6-[35S]sulphate are liberated. From the products obtained after digestion with various mutant fibroblasts and with purified N-acetylgalactosamine 6-sulphate sulphatase we suggest that (i) [35S]sulphate is released almost exclusively from galactose 6-sulphate residues; (ii) N-acetylgalactosamine 6-sulphate sulphatase exhibits galactose 6-sulphate sulphatase activity; (iii) both sulphatase activities are deficient in Morquio disease type A.     

Sequential distribution of keratan sulphate and chondroitin sulphate epitopes during ameloblast differentiation

Proteoglycans are complex macromolecules containing one or more glycosaminoglycan chains and exhibiting a variety of biological functions in connective tissues. The aim of the present study was to immunolocalize the distribution of keratan sulphate and chondroitin sulphate epitopes during initial enamel formation in order to study temporo-spatial expression patterns of these macromolecules. Third molars of four-months-old pigs were used for immunolocalization of keratan sulphate and chondroitin sulphate epitopes in the developing enamel layer. Tooth organs were prepared for paraffin sections in order to perform indirect immunohistochemistry. The results demonstrated a mutually exclusive positioning between these two epitopes. Keratan sulphate epitopes were observed in pre-secretory pre-ameloblasts and adjacent stratum intermedium while chondroitin sulphate epitopes were demonstrated in secretory ameloblasts and adjacent stratum intermedium. Our findings suggest that proteoglycans containing glycosaminoglycan chains may play a regulatory role during enamel mineralization.     

The alkali-labile linkage between keratan sulphate and protein

Keratan sulphate was isolated from adult intervertebral disc in 90% yield by sequential digestion of the whole tissue with papain, Pronase and Proteus vulgaris chondroitin sulphate lyase. Treatment of this preparation with alkali cleaved a glycosidic bond between N-acetylgalactosamine and threonine and produced, by an alkali-catalysed ;peeling' reaction, an unsaturated derivative of N-acetylgalactosamine which reacted as a chromogen in the Morgan-Elson reaction, but remained covalently bonded to the keratan sulphate chain. This derivative was reduced and labelled by alkaline NaB(3)H(4). The substituent at position 3 of N-acetylgalactosamine in the keratan sulphate-protein linkage was identified as a disaccharide, N-acetylneuraminylgalactose, which was isolated from the reaction mixture after alkali treatment.     

Expression of matrilins during maturation of mouse skeletal tissues.

The matrilins are a recently discovered family of non-collagenous extracellular matrix proteins. During embryogenesis, all matrilins are expressed in skeletal tissues. Additionally, matrilin-2 and -4 are expressed in the dermis and in connective tissues of internal organs, e.g. of the lung and kidney. After birth, the expression of matrilin-1 and -3 remains specific for cartilage and bone whereas matrilin-2 and -4 display a broader tissue distribution and could be detected in epithelial, muscle, and nervous tissue as well as in loose and dense connective tissue. In epiphyseal cartilage of growing long bones, matrilin-1 and -3 are present in all cartilage regions, in contrast to matrilin-2, which is expressed in the proliferative and the upper hypertrophic zones. Similarly matrilin-4 was detected all over the epiphyseal cartilage, with the weakest expression in the hypertrophic zone. Although it was shown that matrilin-1 and -3 can form hetero-oligomers and are often co-localized in tissue, clear differences in their spatial distribution could be demonstrated by double-immunolabelling. During joint development matrilin-2 and matrilin-4 are present at the developing joint surface, while in articular cartilage of 6-week-old mice all matrilins are only weakly expressed.