Relative susceptibilities of the glucosamine-glucuronic acid and N-acetylglucosamine-glucuronic acid linkages to heparin lyase III.

Heparin lyases are valuable tools for generating oligosaccharide fragments and in sequence determination of heparan sulfate (HS). Heparin lyase III is known to cleave the linkages between N-acetylglucosamine (GlcNAc) or N-sulfated glucosamine (GlcNS) and glucuronic acid (GlcA) as the primary sites and the linkages between GlcNAc, GlcNAc(6S), or GlcNS and iduronic acid as secondary sites. N-Unsubstituted glucosamine (GlcN) occurs as a minor component in HS, and it has been associated with various bioactivities. Here we investigate the specificity of heparin lyase III toward the GlcN-GlcA linkage using a recombinant enzyme of high purity and as substrates the partially de-N-acetylated polysaccharide of Escherichia coli K5 strain and derived hexasaccharides. The specificity of lyase III toward the GlcN-GlcA linkage is deduced by sequencing of the oligosaccharide products using electrospray mass spectrometry with collision-induced dissociation and MS/MS scanning. The results demonstrate that under controlled conditions for partial digestion, lyase III does not act at the GlcN-GlcA linkage, whereas GlcNAc-GlcA is cleaved. Even under forced conditions for exhaustive digestion, the GlcN-GlcA linkage is only partly cleaved. It is this property of lyase III that has enabled the isolation of a unique, nonsulfated antigenic determinant DeltaUA-GlcN-UA-GlcNAc from HS and from partially de-N-acetylated K5 polysaccharide. It was unexpected that pentasaccharide fragments were also detected among the digestion products of the K5 polysaccharide used. It is possible that these are products of an additional glycosidase activity of lyase III, although other mechanisms cannot be completely ruled out.

On-line overpressure thin-layer chromatographic separation and electrospray mass spectrometric detection of glycolipids.

On-line thin-layer chromatographic separation and electrospray mass spectrometry (TLC/ESI-MS) has been accomplished by direct linking of a commercial overpressure TLC instrument, OPLC 50, and a Q-TOF mass spectrometer. Mass spectrometric detection sensitivity and chromatographic resolution achieved by this configuration were assessed using acidic glycolipids as examples. Under the optimized conditions, a sensitivity of 5 pmol of glycosphingolipid was readily demonstrated for TLC/ESI-MS and 20 pmol for TLC/ESI-MS/MS production scanning to derive the saccharide sequence and long chain base/fatty acid composition of the ceramide. Initial preconditioning of TLC plates is necessary to achieve high sensitivity detection by reducing chemical background noise. Plates can be used repeatedly (at least 10 times) for analysis, although this may result in a minor reduction in TLC resolution. Following solvent development, separated components on the TLC plates can be detected in the conventional way by nondestructive staining or UV absorption or fluorescence and can be stored for on-line TLC/ESI-MS analysis at a later stage without reduction in mass spectrometric detection sensitivity and chromatographic resolution. Aspects for further improvement of OPLC instrumentation include use of narrower TLC plate dimensions and refined design of the eluate exit system.

An investigation of the interactions of E-selectin with fuco-oligosaccharides of the blood group family.

This investigation is concerned with assignments of Lewis(a) (Le(a)) and Le(x) analogs on linear and branched di- to hexasaccharide backbones as components of the recognition motifs for E-selectin. The influence of the location of fucose residue(s) was investigated using 14 structurally defined and variously fucosylated oligosaccharides in biotinylated form or as neoglycolipids in static binding assays, in microwells, and on thin-layer chromatograms. Results of the two assay systems were in agreement overall and showed that the recognition motifs for E-selectin include 4-fucosyl-lacto (Le(a)) and 3-fucosyl-neo-lacto (Le(x)) sequences strictly at capping positions and not Le(x) at an internal position as a part of VIM-2 antigen sequence. There is greater potency of the Le(a) over the Le(x) series. Additional fucose residues alpha1-2-linked to neighboring galactoses or alpha1-3-linked to inner N-acetyglucosamines or to reducing-terminal glucose residues of the tetrasaccharide backbone had little or no effect on the selectin binding. E-selectin binding to the Le(a) or Le(x )capping motif on a 3-linked branch was equivalent to the binding on the corresponding linear backbone. A lack of E-selectin binding to the Le(x) motif capping a 6-linked branch and to the Le(x) trisaccharide linked to biotin via a nine-carbon spacer indicates that the -GlcNAcbeta1-3Gal- sequence on the oligosaccharide backbone adjoining the Le(x) is a part of recognition motif for E-selectin. These findings contribute to understanding the molecular basis of E-selectin recognition and could influence future designs of selectin antagonists as possible therapeutic substances.

Chemically synthesized solid phase oligosaccharide probes for carbohydrate-binding receptors. Interactions of the E-, L- and P-selectins with sialyl-Le(x) and O-sulphated forms linked to biotin or to polyacrylamide.

There is a growing interest in chemically defined oligosaccharide reagents for identifying proteins that bind carbohydrates and determining the specificities of carbohydrate-binding proteins. Here, we compare three sets of chemically synthesized commercially available oligosaccharide conjugates as immobilized probes, for the binding signals that they elicit with known carbohydrate-binding receptors of the immune system, the E-, P- and L-selectins. The first set of conjugates is of oligosaccharides linked to biotin via a nine-carbon spacer. The second and third sets are multivalent derivatives in which the oligosaccharides are linked, via a three-carbon spacer to poly[N-(2-hydroxyethyl)acrylamide] (PAA) or to biotinylated PAA with an average of 20% substitution of the hydroxyethyl-amide groups by carbohydrate. The conjugates were immobilized on streptavidin-coated microwells if biotinylated, otherwise by drying in uncoated wells. The most robust binding curves, overall, were with the biotinylated PAA derivatives of the ligands immobilized on streptavidin wells. These reagents have permitted a reevaluation of selectin binding signals elicited by sialyl-Lewis(x) (SLe(x)) analogues having sulphate at position 6 of the galactose (6'SuSLe(x)) or of the N-acetylglucosamine (6SuSLe(x)). The results clarify the role of 6SuSLe(x), rather then 6'SuSLe(x), as a ligand for the selectins.