Carbohydrate--protein interactions in capillary morphogenesis?
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to T Feizi.
Explore the source record for details and available documents.
The binding of the leucocyte adhesion molecule L-selectin has been investigated toward several structurally defined lipid-linked oligosaccharides immobilized on silica gel chromatograms or plastic wells. In both assay systems the 3'-sulphated Le(a)/Le(x) type tetrasaccharides [formula: see text] were more strongly bound than 3'-sialyl analogues. A considerable binding was observed to the 3'-sulphated oligosaccharide backbone in the absence of fucose but not to a 3'-sialyl analogue or fuco-oligosaccharide analogues lacking sulphate or sialic acid. Affinity for other sulphated saccharides: 3'-sulphoglucuronyl neolactotetraosyl ceramide and glycolipids with sulphate 3'-linked to terminal or sub-terminal galactose or N-acetylgalactosamine was detected in the chromatogram assay only. These studies, together with earlier reports that L-selectin binding to endothelium is inhibited by sulphatide, highlight the relative importance of sulphate in the adhesive specificity of this protein.
E-selectin is the inducible adhesion protein on the surface of endothelial cells which has a crucial role in the initial stages of recruitment of leucocytes to sites of inflammation. In addition, it is almost certainly involved in tumor cell adhesion and metastasis. This report is concerned with identification of a new class of oligosaccharide ligand--sulfate-containing--for the human E-selectin molecule from among oligosaccharides on an ovarian cystadenoma glycoprotein. This has been achieved by application of the neoglycolipid technology to oligosaccharides released from the glycoprotein by mild alkaline beta-elimination. Oligosaccharides were conjugated to lipid, resolved by thin-layer chromatography, and tested for binding by Chinese hamster ovary cells which had been transfected to express the full-length E-selectin molecule. Several components with strong E-selectin binding activity were revealed among acidic oligosaccharides. The smallest among these was identified by liquid secondary ion mass spectrometric analysis of the neoglycolipid, in conjunction with methylation analysis of the purified oligosaccharide preparation as an equimolar mixture of the Le(a)- and Le(x)/SSEA-1-type fucotetrasaccharides sulfated at position 3 of outer galactose: [formula: see text] To our knowledge this is the first report of a sulfofucooligosaccharide ligand for E-selectin. The binding activity is substantially greater than those of lipid-linked Le(a) and Le(x)/SSEA-1 sequences and is at least equal to that of the 3'-sialyl-Le(x)/SSEA-1 glycolipid analogue.
The soluble dimeric beta-galactoside-binding lectin (subunit molecular mass, approximately 14 kDa) of bovine heart muscle, in common with the 14-kDa lectins of several other animal species, displays carbohydrate-binding activity when it is in the reduced state, but the purified lectin loses this activity upon oxidation. In the present study, the presence of any post-translational modification and the mechanism of the oxidative inactivation have been investigated by analyses of the reduced and oxidized forms of the purified bovine lectin by electrospray ionization-mass spectrometry (ESI-MS) and by liquid secondary ion mass spectrometry (LSIMS) of tryptic and peptic peptides. By ESI-MS, the molecular mass of the reduced lectin is determined to be 14,654.6 +/- 0.9 Da, and that of the oxidized lectin is 14,649.3 +/- 1.1 Da. These masses correspond to the amino acid sequence of the protein with the cysteines having free sulfhydryl groups in the reduced state and forming disulfide bonds in the oxidized state. There is no evidence of post-translational modification in either lectin form except for monoacetylation already predicted for alanine at the blocked N-terminal end. Pronounced differences in charge distribution in the electrospray ionization mass spectra of the reduced and oxidized lectin, reflecting a change in the number of accessible protonation sites in the oxidized protein, are consistent with the protein being held in an altered conformation by covalent bonding. The results of LSIMS analyses of tryptic and peptic peptides in conjunction with Edman sequencing indicate that disulfide bonding occurs predominantly between Cys2 and Cys130, Cys16 and Cys88, and Cys42 and Cys60. There is no evidence of oxidation of Trp68. These results, taken together with observations that almost the complete polypeptide chain is necessary for the functional integrity of the carbohydrate recognition domain (Abbott, W. M., and Feizi, T. (1991) J. Biol. Chem. 266, 5552-5557) point to intramolecular disulfide bonding with a change in protein folding and conformation as the mechanism of oxidative inactivation of the purified bovine lectin.
Binding specificity of the major surfactant protein SP-A from human and dog lung has been investigated. Radiobinding experiments have shown that both proteins bind in a Ca(2+)-dependent manner to galactose, mannose, fucose, and glucose linked to bovine serum albumin. These results are in accord with a previous study in which monosaccharides were linked to agarose (Haagsman, H. P., Hawgood, S., Sargeant, T., Buckley, D., White, R. T., Drickamer, K., and Benson, B. J. (1987) J. Biol. Chem. 262, 13877-13880). Chromatogram overlays in conjunction with in situ liquid secondary ion mass spectrometry (TLC-LSIMS) of several purified glycosphingolipids and neoglycolipids as well as binding assays with glycolipids immobilized on plastic wells, demonstrate recognition of galactose (human and dog SP-A), glucose, and lactose (human SP-A) in association with specific lipids. In addition, the occurrence of several neutral and acidic glycosphingolipids in human and rat extracellular surfactants and rat alveolar type II cells is described. Selected components among the neutral glycolipids are bound by radiolabeled human SP-A; these are identified by TLC-LSIMS as predominantly ceramide mono- and disaccharides (human surfactant) and ceramide tri- and tetrasaccharides (rat surfactant and type II cells). A recombinant carbohydrate recognition domain (CRD) of human SP-A inhibits the binding of human SP-A to galactosyl ceramide and to galactose- and mannose-bovine serum albumin, indicating that the CRD is directly involved in the binding of SP-A to these ligands. These results provide evidence for a novel type of binding specificity for proteins that have Ca(2+)-dependent CRDs and raise the possibility that glycosphingolipids are endogenous ligands for SP-A.
In earlier studies, the minimum structure which inhibited the binding of anti-i to an i-active glycoprotein was the linear trisaccharide, beta-D-Galp-(1----4)-beta-D-GlcpNAc-(1----3)-D-Gal. There was an increasing hierarchy of inhibitory activities in the linear tetrasaccharide, beta-D-Galp-(1----4)-beta-D-GlcpNAc-(1----3)-beta-D-Galp-(1----4)-beta-D -GlcNAc , its methyl beta-glycoside, and in the methyl beta-glycoside of the hexasaccharide. The linear octasaccharide methyl beta-glycoside in this series is approximately only half as active as the hexasaccharide methyl beta-glycoside. Analyses by high resolution 1H-n.m.r. of these two oligosaccharides indicated that they have similar conformations in solution, and there is no evidence for the occurrence of inter-molecular interactions which might partially hinder the binding of anti-i to the octasaccharide methyl beta-glycoside. These results are consistent with the size of the i antigen being in the region of a hexasaccharide. It is proposed that the methyl aglycon group of the hexasaccharide methyl beta-glycoside confers an above normal activity by presenting a hydrophobic area for additional contact in the vicinity of the antibody-combining site.
Twenty-two neutral O-linked oligosaccharides ranging from monosaccharides to octasaccharides were identified in bovine submaxillary-gland-mucin glycoprotein by a combination of liquid secondary-ion mass spectrometry, methylation analysis and 1H-NMR. Only five of these have been previously detected in bovine submaxillary-gland mucin although several have been described from other sources of mucin. The structures include short linear sequences 3-linked to N-acetylgalactosaminitol (GalNAcol) and branched structures based on either a GlcNAc(beta 1-6) [Gal(beta 1-3)]GalNAcol or GlcNAc(beta 1-6)[GlcNAc(beta 1-3)]GalNAcol core region. Oligosaccharides not previously characterised from any source were the disaccharide GalNAc alpha 1-6GalNAcol (GalNAc, N-acetylgalactosamine and the hexasaccharide GlcNAc(beta 1-6) [GalNAc(alpha 1-3)( Fuc (alpha 1-2)]Gal(beta 1-4)GlcNAc(beta 1-3)]GalNAcol (Fuc, L-fucose). Oligosaccharides of the blood-group-A type have not been detected previously in bovine submaxillary-gland mucin although their occurrence on bovine gastric-mucosal glycoproteins has been established by classical immunochemical studies.
Carbohydrate recognition by amyloid P component from human serum has been investigated by binding experiments using several glycosaminoglycans, polysaccharides and a series of structurally defined neoglycolipids and natural glycolipids. Two novel classes of carbohydrate ligands have been identified. The first is 6-phosphorylated mannose as found on lysosomal hydrolases, and the second is the 3-sulphated saccharides galactose, N-acetyl-galactosamine and glucuronic acid as found on sulphatide and other acidic glycolipids that occur in neural or kidney tissues or on subpopulations of lymphocytes. Binding to mannose-6-phosphate containing molecules and inhibition of binding by free mannose-6-phosphate and fructose-1-phosphate are features shared with mannose-6-phosphate receptors involved in trafficking of lysosomal enzymes. However, only amyloid P binding is inhibited by galactose-6-phosphate, mannose-1-phosphate and glucose-6-phosphate. These findings strengthen the possibility that amyloid P protein has a central role in amyloidogenic processes: first in formation of focal concentrations of lysosomal enzymes including proteases that generate fibril-forming peptides from amyloidogenic proteins, and second in formation of multicomponent complexes that include sulphoglycolipids as well as glycosaminoglycans. The evidence that binding to all of the acidic ligands involves the same polypeptide domain on amyloid P protein, and inhibition data using diffusible, phosphorylated monosaccharides, is potentially important leads to novel drug designs aimed at preventing or even reversing amyloid deposition processes without interference with essential lysosomal trafficking pathways.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Long-chain sialo-oligosaccharides with poly-N-acetyllactosamine backbones (Ii antigen type) are major host cell receptors for the human pathogen Mycoplasma pneumoniae. Previous immunofluorescence studies of the human bronchial epithelium, using sequence-specific monoclonal antibodies to the branched I-type and linear i-type backbones, have indicated that sialylated and nonsialylated long-chain sequences of both types are richly expressed on the ciliated cells, where they are polarized at the apical aspects. These sequences are lacking in the goblet cells. In the present study, the display of these oligosaccharides has been investigated by electron microscopy (immunogold labelling) in the human bronchial epithelium and in that of the hamster, an animal model commonly used for M. pneumoniae infection. In the human bronchial epithelium, the long-chain branched sequences have been detected along the entire length of the cilia and on microvilli, whereas the linear sequences are confined to the microvilli and the basal aspects of the cilia. On the ciliated epithelial cells of the hamster, by contrast, the branched and linear sequences (sialo- and asialo-) have been detected exclusively on microvilli. A further striking difference is that in the hamster these structures are expressed in abundance on the goblet cells and in the intracellular globules. We suggest that the latter finding may partly explain the relatively large doses of M. pneumoniae required to establish experimental infection in the hamster, as the receptor-bearing secreted mucus may have a protective role in binding to the microorganisms, leading to their clearance by bronchociliary action.
Membrane glycolipids contain the lactose sequence (galactose linked to glucose), and the oligosaccharide is variously extended such that there is a cell-type-specific repertoire. In this study, binding of Pseudomonas aeruginosa M35 to lipid-linked lactose (Gal beta 1-4Glc [structure 1]), lacto-N-neotetraose (Gal beta 1-4GlcNAc beta 1-3Gal beta 1-4Glc [structure 2]), lacto-N-tetraose (Gal beta 1-3GlcNAc beta 1-3Gal beta 1-4Glc [structure 3]), and asialo GM1 (Gal beta 1-3GalNAc beta 1-4Gal beta 1-4Glc [structure 4]) was evaluated and compared with binding of Escherichia coli C600 to these compounds. Oligosaccharides were linked to the lipid phosphatidylethanolamine dipalmitoate, and the resulting neoglycolipids were resolved on thin-layer chromatograms or coated onto plastic microtiter wells. Lipid-linked structures 1 to 4 were bound by P. aeruginosa and E. coli in the chromatogram assay, but only structure 4 was bound in the microtiter well assay. As shown previously for E. coli binding to lipid-linked structures 1 to 3, binding to lipid-linked structure 4 was not inhibited with oligosaccharide, indicating a requirement for lipid and oligosaccharide. With few exceptions, sialylation and fucosylation of structures 1 to 4 resulted in impaired or abolished binding. Comparisons of binding intensities in the chromatogram assay indicated that recognition by P. aeruginosa and recognition by E. coli are not identical. Presence of the additional disaccharide unit, as in structure 2, resulted in enhanced binding of P. aeruginosa but diminished binding of E. coli relative to lactose binding; fucosylation at galactose of lactose resulted in markedly diminished binding of P. aeruginosa only. In the microtiter well assay, binding of E. coli to asialo GM1 was much weaker than P. aeruginosa binding. The saccharide-plus-lipid-dependent adhesion may be an important factor in increased susceptibility to infection of epithelia already damaged by microbial and chemical agents; the differing strengths of adhesion to the structural variants may relate to tissue tropism.
As part of the establishment of a data base for core and backbone sequences of O-linked oligosaccharides of human meconium glycoproteins, the minor tetra- to hepta-saccharides released from mild-acid treated blood group [corrected] H-active glycoproteins have been studied. These oligosaccharides are heterogeneous and difficult to isolate, and a t.l.c.-m.s. microsequencing procedure has been applied to the neoglycolipid derivatives, in conjunction with 1H-n.m.r. spectroscopy, methylation analysis, and mass spectrometry (m.s.) of native and methylated oligosaccharides. Among an array of oligosaccharides characterised are those having the branched beta-GlcNAc-(1----6)[beta-Gal- (1----3)]-GalNAcol core, and others with the following linear sequences not characterised previously from this source: beta-Gal-(1----3)-beta-GlcNAc-(1----3)-beta-Gal-(1----3)-GalNAcol, beta-Gal-(1----4)-beta-GlcNAc-(1----3)-beta-Gal-(1----3)-GalNAcol, alpha-GalNAc- (1----3)-beta-Gal-(1----3/4)-beta-GlcNAc-(1----3)-GalNAcol, beta-GlcNAc- (1----3)-beta-Gal-(1----3/4)-beta-GlcNAc-(1----3)- GalNAcol, beta-Gal-(1----3/4)-beta-GlcNAc-(1----3)-beta-Gal-(1----3/4)-beta- GlcNAc-(1----3)-beta-Gal-(1----3)-GalNAcol, and beta-GlcNAc-(1----3)-beta-Gal-(1----3/4)-beta-GlcNAc-(1----3)-beta-Gal- (1----3/4)-beta-GlcNAc-(1----3)-GalNAcol.
The aim of the present study has been to investigate the binding specificity of the soluble 14,000-dalton lectin of bovine heart muscle towards immobilised oligosaccharides in clustered form. To this end, chromatogram overlay assays and quantitative plastic-microwell-binding assays have been performed using several natural glycolipids and neoglycolipids containing one or more of the disaccharide units, beta-D-Galp-(1----4 or 3)-D-GlcNAc or beta-D-Galp-(1----4)-D-Glc and related structures. The microwell assay gave the most consistent results. It was observed that for binding by the soluble lectin the optimal sequence, which is beta-D-Galp-(1----4 or 3)-D-GlcNAc, must occur at the nonreducing end of longer oligosaccharides when linked to lipid. These oligosaccharides may be of poly(N-acetyllactosamine) type or they may be mono- or multi-antennary, complex-type chains in which the disaccharide is joined directly to a trimannosyl core. The lectin bound to such immobilised lipid-linked oligosaccharides on which the terminal D-galactosyl groups are substituted with alpha-L-Fucp-(1----2), alpha-D-Galp-(1----3), or alpha-NeuAc-(2----3) groups. However, no binding was detected if the terminal D-galactosyl groups were substituted with an alpha-NeuAc-(2----6) group or the subterminal N-acetylglucosamine units with an alpha-L-Fucp-(1----3 or -4) group. Internally located N-acetyllactosamine units where the D-galactose units are disubstituted by beta-D-GlcNacp-(1----3) and -(1----6) units, as in branched poly(N-acetyllactosamine) backbones were not bound by the bovine lectin. These results are in accord with previous observations on the bovine lectin and the corresponding human and rat lectins, using structurally defined oligosaccharides as inhibitors of binding. The results of comparative binding experiments using paragloboside and ceramide hexasaccharide which contain one and two N-acetyllactosamine units, respectively, joined in linear sequence to the lactosylceramide core, were equivocal with respect to the availability of the internal N-acetyllactosamine units for binding by the bovine lectin.
The soluble beta-galactoside-specific bovine lectin of subunit 14 kDa has been expressed in vitro by transcription and then translation in a rabbit reticulocyte lysate. The protein thus expressed shows the predicted binding to lactose coupled to Sepharose. Several mutants of the 134 amino acid protein have been expressed and insight gained into (a) the polypeptide length required to form the carbohydrate recognition domain and (b) the functional importance of some of the highly conserved amino acids. The following amino acids have been deleted: 1-9, 1-23, 88-122, 88-134, 107-134, or 124-134. In addition, a frame-shift mutant has been made in which the 23 amino acids at the C-terminal end were completely changed. Among these seven mutants only mutant 1-9 shows carbohydrate binding but with congruent to 30% of the activity of the wild-type protein (as assessed by the percentage of the protein bound to lactose-Sepharose). On the other hand, carbohydrate binding is relatively well preserved (75-90%) in mutant proteins where the C-terminal octapeptide sequence of the bovine lectin has been changed to sequences that resemble those in the chick 14-kDa lectin. When the single tryptophan at position 68 is changed by point mutagenesis to phenylalanine or to a leucine residue, a weak binding activity (congruent to 20%) is retained only with the former. When either of the cysteines 2 or 60 is changed to serine, binding activity is reduced to congruent to 60%, and when both are changed, to congruent to 20% of that for the wild-type protein. The susceptibility of the lectin to oxidative inactivation is unaffected when these 2 cysteines and cysteine 130 are changed to serine individually or in tandem (cysteines 2 and 60). In a second approach we show that the natural protein isolated from bovine heart is protected from proteolysis by trypsin and V8-protease in the presence of saccharide ligand. Although further work is required to identify residues which come into contact with the carbohydrate ligand, these results indicate that almost the complete polypeptide chain is necessary for the integrity of the carbohydrate recognition domain.
Explore the source record for details and available documents.
Explore the source record for details and available documents.