Control of IgG/Fc glycosylation: a comparison of oligosaccharides from chimeric human/mouse and mouse subclass immunoglobulin Gs

Oligosaccharide profiles were obtained for chimeric mouse-human antibodies corresponding to each of the human IgG subclasses 1-4, and mouse IgG2b antibodies each expressed in the mouse J558L cell line. These antibodies have specificity for the NIP hapten and form a matched set of IgGs. An IgG4 chimeric antibody (B72.3) produced in the chinese hamster ovary (CHO-K1) cell line was also analysed for carbohydrate. Additionally aglycosylated mutants of this IgG4 (B72.3) and anti-NIP mouse IgG2b were analysed. The total lack of carbohydrate found in the aglycosylated site-directed mutants human chimeric IgG4 B72.3 (Asn 297-->Gln) and mouse IgG2b (Asn 297-->Ala) demonstrates that there are no N-glycosylation sites other than Asn 297. Therefore glycosylation profiles for all the IgGs analysed reflect carbohydrate attached to this site. Factors such as cell type (A), template direction by the IgG heavy chains (B) and culture conditions (C) are shown to influence IgG glycosylation profiles. (A) The anti-NIP IgG antibodies expressed by the J558L cell line may have one or two Gal (alpha 1-->3) Gal residues per oligosaccharide unit, indicative of the presence of (alpha 1-->3) galactosyl transferase in the J558L mouse cell line. (B) The galactosylation profiles obtained for the IgG heavy chains, in particular the preference for galactosylation of the Man (alpha 1-->6) arm rather than the Man (alpha 1-->3) arm, contrary to the beta-galactosyltransferase specificity, suggest that the polypeptide chain may act as a template to influence the extent of galactosylation and hence the proportions of each oligosaccharide incorporated. The IgG2 antibody does not display this galactosylation preference. (C) The extent of galactosylation appears to be influenced by the growth conditions, with the highest levels of galactosylation being found for IgG produced by cells grown in still cultures, rather than cells grown as ascites or in hollow fibre bioreactors. It is concluded that though the profile of glycosylation is controlled predominantly by the glycosylation activity of the cell in which the IgG is expressed, differences between the IgG heavy chain templates of the various subclasses and culture conditions can also influence glycosylation.