Classification of glycosyltransferases

Whereas originally grouped by linkage and substrate specificity, glycosyltransferases are nowadays rather grouped by sequence similarity. This move is the logical consequence of the identification of thousands of potential glycosyltransferases across genomic databases. Because 95% of these potential glycosyltransferases have not been assigned any clear substrate specificities yet, they are best maintained in groups of structurally similar proteins. Nevertheless, several examples have shown that structural similarity does not imply similarity in substrate specificity.

All enzymes active on carbohydrates, including potential glycosylhydrolases and glycosyltransferases, have been annotated in a database called CAZY (Carbohydrate-Active enzymes, see www.cazy.org), which is curated by Bernard Henrissat and his team at the University of Marseille. CAZY classifies enzymes in families based on sequence similarity, protein fold prediction and reaction mechanisms. To date, 95 families of glycosyltransferases have been defined, which encompass more than 30’000 putative glycosyltransferases. All proteins of a same CAZY family have likely evolved from the same progenitor sequence, present conserved mechanisms of reaction, and share similar structural folds.

For example, the CAZY glycosyltransferase family #25 currently includes 1263 proteins, the majority coming from bacterial genomes (1145 proteins), but also features 85 eukaryotic proteins and even 33 virally encoded glycosyltransferases. The catalytically known glycosyltransferases from this family transfer Glc and Gal in β-linkage mainly in the context of bacterial LPS biosynthesis. By contrast, the eukaryotic glycosyltransferases in this family transfer Gal to collagen, thereby demonstrating that structural similarity does not relate to substrate specificity. Despite the limited usefulness of the CAZY database to predict glycosyltransferase activities, the direct comparison of protein sequences within a CAZY family outlines strongly conserved amino acid, which are likely to be involved in the catalytic activity or in maintaining protein stability.