Diversity of oligosaccharide structures on the envelope glycoprotein gp 120 of human immunodeficiency virus 1 from the lymphoblastoid cell line H9. Presence of complex-type oligosaccharides with bisecting N-acetylglucosamine residues.

The N-linked oligosaccharide structures on the envelope glycoprotein gp120 of human immunodeficiency virus 1 derived from chronically infected lymphoblastoid (H9) cells have been investigated by enzymatic microsequencing after release from protein by hydrazinolysis, labeling with NaB3H4, and chromatography on adsorbent columns of Phaseolus vulgaris erythrophytohemagglutinin and Ricinus communis agglutinin (Mr 120,000) and on Bio-Gel P-4. A substantially greater diversity of oligosaccharide structures was detected than among those released by hydrazinolysis from recombinant gp120 produced in Chinese hamster ovary cells and investigated by similar procedures (Mizuochi, T., Spellman, M.W., Larkin, M., Solomon, J., Basa, L.J., and Feizi, T. (1988) Biochem J. 254, 599-603) and among those released by endoglycosidases from virus-derived gp120 isolated from infected H9 cells after metabolic labeling with D-[2-3H]mannose or D-[6-3H]glucosamine (Geyer, H., Holschbach, L., Hunsmann, G., and Schneider, J. (1988) J. Biol. Chem. 263, 11760-11767). In this study, 16% of the oligosaccharides were identified as complex-type bi-, tri-, and tetraantennary sialo-oligosaccharides with bisecting N-acetylglucosamine residues. Such structures were lacking on recombinant gp120 and could not be detected on the metabolically labeled, virus-derived glycoprotein. As in the earlier investigations, complex-type chains lacking bisecting N-acetylglucosamine residues, hybrid-type chains, and a series of high mannose-type structures with 5-9 mannose residues were identified. In addition, an array of complex-type chains having one or more outer chains with beta-galactosyl residues were detected in this study, but with additional substitutions that require further investigation. The number of potential N-glycosylation sites on gp120 is on the order of 20, but the oligosaccharide structures are far more numerous. Thus, the salient conclusion from this and earlier investigations is that alternative structures occur on at least some of the glycosylation sites and that numerous glycosylation variants of this glycoprotein are produced even within a single cell line. Since the glycosylation is the product of host cell glycosyltransferases, an even greater number of glycosylation variants of gp120 are predicted to arise from the heterogeneous cell populations harboring the virus in in vivo infection.