Subsite Mapping of Enzymes

Abstract

Abstract The product distributions of Bacillus amyloliquefaciens amylase acting on homologous maltodextrins have been measured and used to map the active site of the enzyme. These data can be used to estimate the energetics of subsite-monomer interactions on the enzyme. The resulting energy contour supplied the information needed to compute the population distribution of productive and nonproductive positional isomers for any set of enzyme-substrate complexes. These calculations lead to a theoretical dependence of the apparent Michaelis parameters, Ki, Km, and Vmax on chain length. The theoretical dependence of the enzyme parameters upon chain length predicted from the representational subsite model is in rather poor agreement with the experimental data. However, modification of the model by introduction of substrate-induced strain brings the computed and experimental data into exceptionally good accord for chain lengths 1 to 12. Thus, the subsite model quantitatively accounts for many important thermodynamic and kinetic facets of enzyme-substrate interactions. Our results reveal that the specific glycosidic monomer undergoing hydrolysis is probably highly strained and that long chains are hydrolyzed more rapidly than shorter ones primarily because of increased strain. To our knowledge this is the first instance in which Michaelis parameters have been quantitatively correlated with product patterns over a wide range of chain lengths.