Antithrombin−Heparin Affinity Reduced by Fucosylation of Carbohydrate at Asparagine 155

• Published in: Biochemistry (Easton) Vol. 35; no. 27; pp. 8881 - 8889
• Main Authors: Garone, Louise, Edmunds, Tim, Hanson, Eric, Bernasconi, Richard, Huntington, James A, Meagher, Jennifer L, Fan, Bingqi, Gettins, Peter G. W
• Format: Journal Article
• Language: English
• Published: United States American Chemical Society 09-07-1996
• Subjects: Amino Acid Sequence; Antithrombin III - analogs & derivatives; Antithrombin III - metabolism; Antithrombin III - pharmacology; Asparagine - analogs & derivatives; Base Sequence; Fucose - analogs & derivatives; Glycoproteins - metabolism; Glycoproteins - pharmacology; Glycosylation; Heparin - metabolism; Humans; Mass Spectrometry; Molecular Sequence Data; Monosaccharides - analysis; Peptide Fragments - chemistry; Protein Binding; Protein Processing, Post-Translational; Recombinant Proteins - metabolism; Recombinant Proteins - pharmacology
• ISSN: 0006-2960; 1520-4995
• DOI: 10.1021/bi960542m

The two human plasma antithrombin isoforms, α and β, differ in glycosylation at asparagine 135. Only the α form carries carbohydrate at this position and has lower affinity for heparin than the β form. We previously found additional heterogeneity in a recombinant N135Q antithrombin variant, evidenced by two isoforms with a 2-fold difference in heparin affinity [Turko, I. V., Fan, B., & Gettins, P. G. W. (1993) FEBS Lett. 335, 9−12]. To test whether this heterogeneity of heparin affinity results from specific glycosylation differences, we have determined the carbohydrate composition at the three remaining glycosylation sites, asparagine residues 96, 155, and 192, in each of the two N135Q isoforms, by a combination of peptide fragmentation and electrospray mass spectrometry. Patterns of glycosylation at residues 96 and 192 were similar for each isoform and showed the presence of mono-, bi-, and triantennary complex carbohydrate, as well as fucosylation of all types of chains. At position 155, however, there was a marked difference between the isoforms, with the form with lower heparin affinity being 97% fucosylated at this position, whereas the form with higher affinity for heparin was not fucosylated. Other differences in carbohydrate type showed no strong correlation between the two isoforms. We conclude that formation of the two heparin-affinity isoforms of N135Q antithrombin results from the specific difference in fucosylation at residue 155, which may result in different structural properties of the carbohydrate. Consistent with these findings was the elimination of heparin-affinity heterogeneity in a double N135Q-N155Q variant antithrombin. It is possible that fucosylation of antithrombin may occur in vivo as a means of modifying the physiological properties of the antithrombin through alteration of the amount of antithrombin bound to surface heparin-like species.