Evidence for multilayer nanoscale enzyme active sites

Abstract

One of the most fundamental questions in biochemistry today is how enzymes work. Most often this discussion focuses on the amino acid residues in direct contact with the reactive metal or reacting substrate within the three-dimensional (3D) structure of the protein. What is very rarely mentioned is the influence that remote residues have on enzyme catalysis. Remote residues refer to those residues which are, or are farther from, second-nearest neighbors to the reactive metal or reacting substrate molecule. The literature has scarce information pertaining to the importance of these second- and third-shell residues in enzyme catalysis. The idea of the involvement of residues located in outer coordination spheres in catalysis was first introduced by Leatherbarrow, Fersht and Winter when the concept of site-directed mutagenesis was introduced. It was discovered that mutations made to residues located far from the reaction site resulted in proteins with reduced catalytic rate. In some cases; however, these mutations resulted in proteins whose catalytic rate was increased. It was at this time that the term `protein engineering' was coined. While limited studies have been performed to understand the role of remote residues in enzyme catalysis, a thorough investigation of the importance of second- and third-shell residues has not been performed. In this thesis, two different computational methods, THEMATICS and Evolutionary Trace (ET), based on two very different types of input, are used to identify functionally important residues in the first-, second- and third-shells of an enzyme. It is shown that both of these methods predict residues in the second- and third- shells to be important. Once the concept of remote residue involvement in enzyme catalysis has been established theoretically, the focus shifts to one particular enzyme, Co-type nitrile hydratase from Pseudomonas putida, for which both THEMATICS and ET predict a multilayer active site. First, the x-ray crystal structure of the wild type enzyme and its kinetic properties are reported. A kinetic analysis of single point mutations is presented for five second- and third-shell residues that were predicted computationally to be functionally important. Additionally, crystal structures are presented for four of the mutants. It is shown that for some of the mutants there are small, local structural differences which may explain the effects on catalytic rate, however, for others, no structural differences are observed compared to wild type. For these examples, it is proposed that the differences are due primarily to electrostatic effects. While no unequivocal explanation emerges at this stage for why these residues in the outer coordination spheres influence catalysis, this work makes a strong case for the concept that enzyme active sites are built in multiple layers. It is suggested that computational approaches, and the concept of multilayer active sites introduced herein, can help to guide protein engineering efforts.