Dissecting the catalytic mechanism of a plant beta-D-glucan glucohydrolase through structural biology using inhibitors and substrate analogues

Abstract

Higher plant, family GH3 β-d-glucan glucohydrolases exhibit exo-hydrolytic and retaining (e→e) mechanisms of action and catalyze the removal of single glucosyl residues from the non-reducing termini of β-d-linked glucosidic substrates, with retention of anomeric configuration. The broad specificity β-d-glucan glucohydrolases are likely to play roles in cell wall re-modelling, turn-over of cell wall components and possibly in plant defence reactions against pathogens. Crystal structures of the barley β-d-glucan glucohydrolase, obtained from both native enzyme and from the enzyme in complex with a substrate analogues and mechanism-based inhibitors, have enabled the basis of substrate specificity, the mechanism of catalysis, and the role of domain movements during the catalytic cycle to be defined in precise molecular terms. The active site of the enzyme forms a shallow ‘pocket’ that is located at the interface of two domains of the enzyme and accommodates two glucosyl residues. The propensity of the enzyme to hydrolyze a broad range of substrates with (1→2)-, (1→3)-, (1→4)- and (1→6)-β-d-glucosidic linkages is explained from crystal structures of the enzyme in complex with non-hydrolysable S-glycoside substrate analogues, and from molecular modelling. During binding of gluco-oligosaccharides, the glucosyl residue at subsite −1 is locked in a highly constrained position, but the glucosyl residue at the +1 subsite is free to adjust its position between two tryptophan residues positioned at the entry of the active site pocket. The flexibility at subsite +1 and the projection of the remainder of the substrate away from the pocket provide a structural rationale for the capacity of the enzyme to accommodate and hydrolyze glucosides with different linkage positions and hence different overall conformations. While mechanism-based inhibitors with micromolar Ki constants bind in the active site of the enzyme and form esters with the catalytic nucleophile, transition-state mimics bind with their ‘glucose’ moieties distorted into the 4E conformation, which is critical for the nanomolar binding of these inhibitors to the enzyme. The glucose product of the reaction, which is released from the non-reducing termini of substrates, remains bound to the β-d-glucan glucohydrolase in the −1 subsite of the active site, until a new substrate molecule approaches the enzyme. If dissociation of the glucose from the enzyme active site could be synchronized throughout the crystal, time-resolved Laue X-ray crystallography could be used to follow the conformational changes that occur as the glucose product diffuses away and the incoming substrate is bound by the enzyme.