In order to accelerate the hydrolysis of glycosidic bonds by factors approaching 10(17)-fold, glycosidases have evolved finely tuned active sites optimally configured for transition-state stabilization. Structural analyses of various enzyme complexes representing stable intermediates along the reaction coordinate, in conjunction with detailed mechanistic studies on wild-type and mutant enzymes, have delineated the contributions of nucleophilic and general acid/base catalysis, as well as the roles of noncovalent interactions, to these impressive rate enhancements.

Hen egg-white lysozyme (HEWL) was the first enzyme to have its three-dimensional structure determined by X-ray diffraction techniques. A catalytic mechanism, featuring a long-lived oxocarbenium-ion intermediate, was proposed on the basis of model-building studies. The 'Phillips' mechanism is widely held as the paradigm for the catalytic mechanism of beta-glycosidases that cleave glycosidic linkages with net retention of configuration of the anomeric centre. Studies with other retaining beta-glycosidases, however, provide strong evidence pointing to a common mechanism for these enzymes that involves a covalent glycosyl-enzyme intermediate, as previously postulated. Here we show, in three different cases using electrospray ionization mass spectrometry, a catalytically competent covalent glycosyl-enzyme intermediate during the catalytic cycle of HEWL. We also show the three-dimensional structure of this intermediate as determined by X-ray diffraction. We formulate a general catalytic mechanism for all retaining beta-glycosidases that includes substrate distortion, formation of a covalent intermediate, and the electrophilic migration of C1 along the reaction coordinate.

No abstract

The structure of human pancreatic a-amylase has been determined to 1.8 A resolution using X-ray diffraction techniques. This enzyme is found to be composed of three structural domains. The largest is Domain A (residues 1-99, 169-404), which forms a central eight-stranded parallel @-barrel, to one end of which are located the active site residues Asp 197, Glu 233, and Asp 300. Also found in this vicinity is a bound chloride ion that forms ligand interactions to Arg 195, Asn 298, and Arg 337. Domain B is the smallest (residues 100-168) and serves to form a calcium binding site against the wall of the @-barrel of Domain A. Protein groups making ligand interactions to this calcium include Asn 100, Arg 158, Asp 167, and His 201. Domain C (residues 405-496) is made up of antiparallel @-structure and is only loosely associated with Domains A and B. It is notable that the N-terminal glutamine residue of human pancreatic a-amylase undergoes a posttranslational modification to form a stable pyrrolidone derivative that may provide protection against other digestive enzymes. Structure-based comparisons of human pancreatic a-amylase with functionally related enzymes serve to emphasize three points. Firstly, despite this approach facilitating primary sequence alignments with respect to the numerous insertions and deletions present, overall there is only -15% sequence homology between the mammalian and fungal a-amylases. Secondly, in contrast, these same studies indicate that significant structural homology is present and of the order of -70%. Thirdly, the positioning of Domain C can vary considerably between a-amylases. In terms of the more closely related porcine enzyme, there are four regions of polypeptide chain (residues 237-250, 304-310, 346-354, and 458-461) with significantly different conformations from those in human pancreatic a-amylase. At least two of these could play a role in observed differential substrate and cleavage pattern specificities between these enzymes. Similarly, amino acid differences between human pancreatic and salivary a-amylases have been localized and a number of these occur in the vicinity of the active site.Keywords: amylase; crystallography; enzymology; glycogen; pancreatic; sequences; starch; structure a-Amylases catalyze the hydrolysis of a-1,4 glucan linkages in starch and are widely distributed in nature, being found in bacteria, plants and animals. In humans, a-amylase is composed of 496 amino acids in a single polypeptide chain, which is encoded on chromosome 1 as part of a multigene family (Gumucio et al., 1988). These genes are regulated so that different isozymes are synthesized in either salivary glands or the pancreas. The salivary and pancreatic a-amylases are highly homologous in terms of primary sequence (Nishide et al., 1986) but d o exhibit somewhat different cleavage patterns (Minamiura, 1988). The functional differences observed undoubtedly arise from the 15 amino Reprint requests to: Gary D. Brayer, Department of Biochemistry and Molecular Biology, University of B...

Insights into glycosidase mechanisms have come from X-ray crystallographic studies on complexes with substrate analogs and inhibitors, representing all the intermediate species along the reaction coordinate. Site-directed mutagenesis continues to play a significant role in understanding mechanisms, but is also proving important in generating glycosidases of modified mechanism or specificity.