Chemical Evolution of Enzyme-Catalyzed Glycosylation
Chang-Cheng Liu,Jinfeng Ye,Hongzhi Cao
DOI: https://doi.org/10.1021/acs.accounts.3c00754
2024-01-29
Abstract:ConspectusThe limited availability of structurally well-defined diverse glycans remains a major obstacle for deciphering biological functions as well as biomedical applications of carbohydrates. Despite tremendous progress that has been made in past decades, the synthesis of structurally well-defined complex glycans still represents one of the most challenging topics in synthetic chemistry. Chemical synthesis of glycans is a time-consuming and labor-intensive process that requires elaborate planning and skilled personnel. In contrast, glycosyltransferase-catalyzed enzymatic synthesis provides a more efficient, convenient, low-cost, and sustainable alternative to affording diverse and complex glycans. However, the existing methods are still insufficient to fulfill the increasing demand for specific synthetic glycan libraries necessary for functional glycomics research. This is mainly attributed to the inherent character of the glycan biosynthetic pathway. In nature, there are too many glycosyltransferases involved in the in vivo glycan synthesis, but only a small number of them are available for in vitro enzymatic synthesis. For instance, humans have over 200 glycosyltransferases, but only a few of them could be produced from the conventional bacterial expression system, and most of these membrane-associated enzymes could be overexpressed only in eukaryotic cells. Moreover, the glycan biosynthetic pathway is a nontemplate-driven process, which eventually ends up with heterogeneous glycan product mixtures. Therefore, it is not a practical solution for the in vitro enzymatic synthesis of complex glycans by simply copying the glycan biosynthetic pathway.In the past decade, we have tried to develop a simplified and transformable approach to the enzymatic modular assembly of a human glycan library. Despite the structural complexity of human glycans, the glycoinformatic analysis based on the known glycan structure database and the human glycosyltransferase database indicates that there are approximately 56 disaccharide patterns present in the human glycome and only 16 disaccharide linkages are required to account for over 80% of the total disaccharide fragments, while 35 disaccharide linkages are sufficient to cover over 95% of all disaccharide fragments of human glycome. Regardless of the substrate specificity, if one glycosyltransferase could be used for the synthesis of all of the same glycosidic linkages in human glycome, it will require only a few dozen glycosyltransferases for the assembly of entire human glycans. According to the glycobioinformatics analysis results, we rationally designed about two dozen enzyme modules for the synthesis of over 20 common glycosidic linkages in human glycome, in which each enzyme module contains a glycosyltransferase and a group of enzymes for the in situ generation of a nucleotide-activated sugar donor. By sequential glycosylation using orchestrated enzyme modules, we have completed the synthesis of over 200 structurally well-defined complex human glycans including blood group antigens, O-mannosyl glycans, human milk oligosaccharides, and others. To overcome the product microheterogeneity problem of enzymatic synthesis in the nontemplate-driven glycan biosynthetic pathway, we developed several substrate engineering strategies to control or manipulate the outcome of glycosyltransferase-catalyzed reactions for the precise synthesis of structurally well-defined isomeric complex glycans.