Development and Application of Multidimensional HPLC Mapping Method for O-linked Oligosaccharides

Abstract

Glycosylation improves the solubility and stability of proteins, contributes to the structural integrity of protein functional sites, and mediates biomolecular recognition events involved in cell-cell communications and viral infections. The first step toward understanding the molecular mechanisms underlying these carbohydrate functionalities is a detailed characterization of glycan structures. Recently developed glycomic approaches have enabled comprehensive analyses of N-glycosylation profiles in a quantitative manner. However, there are only a few reports describing detailed O-glycosylation profiles primarily because of the lack of a widespread standard method to identify O-glycan structures. Here, we developed an HPLC mapping method for detailed identification of O-glycans including neutral, sialylated, and sulfated oligosaccharides. Furthermore, using this method, we were able to quantitatively identify isomeric products from an in vitro reaction catalyzed by N-acetylglucosamine-6O-sulfotransferases and obtain O-glycosylation profiles of serum IgA as a model glycoprotein.

Figure 1

Isolation and identification of an O-glycan derived from porcine stomach mucin. (a)Chromatogram of PA-glycans derived from mucin on an amide column; (b) Chromatograms of the PA-glycans corresponding to fraction A on the C30 column; (c) MALDI-QIT-TOF-MS/MS spectra of the PA-glycan corresponding to fraction B. Precursor ion was m/z 666 as protonated ion; (d) Chromatograms of the PA-glycan corresponding to fraction B on the C30 column (upper) before and (lower) after β1,3-galactosidase treatment. The asterisk indicates the fractions containing no detectable PA-saccharide.

Figure 1

Isolation and identification of an O-glycan derived from porcine stomach mucin. (a)Chromatogram of PA-glycans derived from mucin on an amide column; (b) Chromatograms of the PA-glycans corresponding to fraction A on the C30 column; (c) MALDI-QIT-TOF-MS/MS spectra of the PA-glycan corresponding to fraction B. Precursor ion was m/z 666 as protonated ion; (d) Chromatograms of the PA-glycan corresponding to fraction B on the C30 column (upper) before and (lower) after β1,3-galactosidase treatment. The asterisk indicates the fractions containing no detectable PA-saccharide.

Figure 2

Identification of the disialyl PA-saccharide produced by the reaction catalyzed by α2,6-sialyltransferase. Chromatograms of (a) the precursor Galβ1-3(Neu5Acα2-6)GalNAc-PA; and (b) the reaction product Neu5Acα2-6Galβ1-3(Neu5Acα2-6)GalNAc-PA on the C30 column.

Figure 3

HPLC map of O-glycans. •, neutral; ▪, sialylated; ♦, sulfated glycans. Key: Gal, galactose; Glc, Glucose; GlcNAc, N-acetylglucosamine; GalN, N-acetylgalactosamine; Man, mannose; Fuc, fucose; S, sulfate; Neu, N-acetylneuraminic acid.

Figure 4

HPLC-based characterization of branch specificity of GlcNAc6ST-1. (a) Time-dependent change of the HPLC profiles on the ODS column for products resulting from glycopeptides possessing two terminal GlcNAc residues during the sulfation reaction catalyzed by the recombinant GlcNAc6ST-1. The asterisk indicates the fractions containing the substrate O-glycosylated peptide: GlcNAcβ1-6(GlcNAcβ1-3Galβ1-3)GalNAcα1-peptide-FAM; (b) Time course of the amounts of the resultant glycopeptides corresponding to fraction A (solid line) and B (dashed line). The O-glycan structures of fractions A and B were identified as GlcNAcβ1-6(HSO₃-GlcNAcβ1-3Galβ1-3)GalNAc and HSO₃-GlcNAcβ1-6(GlcNAcβ1-3Galβ1-3)GalNAc, respectively.

Figure 5

HPLC profiles of PA-O-glycans derived from human serum IgA. (a) Chromatogram of PA-O-glycans derived from serum IgA on a Mono-Q column. The PA-glycan mixture was separated according to sialic acid contents. N, neutral; MS, mono-sialyl; DS, di-sialyl; (b) Chromatograms of the neutral, mono-sialyl, and di-sialyl fractions on a C30 column. The structures of PA-O-glycans in each fraction were identified on the basis of the HPLC map. Molar percent of the O-glycan content in the IgA sample was calculated based on the peak areas. The structure and incidence of major O-glycans are shown on the profiles. Asterisks indicate the peaks derived from melobiose used for IgA purification.