Polysaccharides show a huge degree of diversity owing to their variety of sugar components (aldoses, ketoses and their stereoisomers), the variety of glycosidic bonds (e.g., α 1,4 , α 1,6 , β 1,4 , β 1,3 , etc.), different degrees of polymerization, and different degrees of branching. Therefore, they can be involved in many types of biological function with distinct physicochemical characteristics. It is especially intriguing that α and β glucans, such as amylose (starch) and cellulose, are built from the same glucose units, differing only in the anomeric configurations of glycosidic bonds. Enzymes that hydrolyze these glucans, amylases and cellulases, are the two largest groups of glycoside hydrolases, and have been studied for a long time. A framework to discuss a whole variety of glycoside hydrolases with their three dimensional structural aspects has been established within the last 15 years. In the late 1980s, Henrissat et al. began classification of the amino acid sequences of enzymes, focusing on cellulases and hemicellulases. 1) At the earliest stage of their classification, the α amylase family was treated as belonging to a separate system.2) In 1991, the naming system of the families changed from alphabetical to numerical, and the system expanded to 35 "glycosyl hydrolase" families, incorporating three well known amylase families, GH13 (retaining α amylase family), GH14 (inverting β amylase family), and GH15 (inverting glucoamylase family), as well as many other enzymes.3) All enzymes cleaving, modifying, and creating glycosidic bonds, and modules binding to carbohydrates have been incorporated into the CAZy database (http: www.cazy.org CAZY ). Glycoside hydrolase (GH) families now include enzymes that hydrolyze the glycosidic bonds between two or more carbohydrates or between a carbohydrate and a non carbohydrate moiety. At present, there are 101 families in the GH class ( Fig. 1; note, five have been deleted or reassigned to other classes), and it will continue to grow.The first 3 D structure of a glycoside hydrolase to be determined was that of hen egg white lysozyme (classified into GH22) in the mid 1960s.4) Even in the early 1990s, however, only a limited number of GH enzyme structures were available. Henrissat and colleagues took amino acid sequence information as the primary basis for classification, but they also used hydrophobic cluster analysis from the earliest stage, trying to incorporate as much 3 D structure based information as possible.5) The significant advances in structural studies of GH enzymes over the last decade have indicated that these enzymes show a vast array of tertiary scaffolds, described as "a marketplace of different structures".
6)Our group has been trying to solve crystal structures of useful enzymes for use in bioindustry. Two of these were the first crystal structures within the GH42 and GH57 families, 7,8) and we then began to focus on CAZy enzymes belonging to structurally unknown families. We have since solved two more GH families, and provided the first crystal s...