Kinetic Isotope Effects as Transition State Probes for Glycosidic Reactions

Abstract

The interpretation of experimental kinetic data pertaining to enzyme-catalysed reactions has allowed for the development of some of the tightest binding, non-covalent enzyme inhibitors known to mankind. The use of kinetic isotope effects allows rare insight into the nature of the chemistry occurring in the protein active site. Kinetic isotope effect data has guided transition state analogue design, which has used computational tools to predict structures for the transition state of the enzyme-catalysed reaction. These computational studies, however, have been limited to QM vacuo treatments of the substrate(s) with no inclusion of the wider protein environment.

It is the aim of this thesis to demonstrate that a more rigorous treatment of the protein environment, including active site residues in the QM region through a QM/MM approach, alongside a broader interpretation of the constituent isotope effects that make up the measured V⁄K isotope effects, can provide chemical structures that better reproduce both the chemistry occurring in the protein and the measured isotope effects.

This thesis details the QM/MM modelling of a non-enzymic model system for glycoside hydrolysis in Chapter 11, and four enzyme-catalysed reactions: the hydrolysis of 5’-methylthioadenosine by S. pneumoniae and E. coli 5’-methylthioadenosine/S-adenosylhomocysteine nucleosidase (MTAN) in Chapter 12 and the phosphorolysis of inosine by Homo sapiens and Bos taurus purine nucleoside phosphorylase (PNP) in Chapter 13.

The study of the non-enzymic hydrolysis of 2-(p-nitrophenoxy)-tetrahydropyran sets the scene for the application of QM/MM methods to the study of enzyme-catalysed glycosidic bond cleavage. This chapter details the application of higher-level QM corrections to semi-empirical QM/MM free energy surfaces for the glycosidic cleavage and water nucleophile attack steps and discusses the effect of applying these higher-level corrections to the overall interpretation of the mechanism. This chapter also highlights the potential pitfalls of relying on one-dimensional reaction coordinates to describe the true intrinsic reaction coordinate, even in “simple”, non-enzymic systems.

The studies of the mechanisms for both bacterial MTANs are comprised of semi-empirical and higher-level QM corrected potential/free energy surfaces and representative stationary point structures for states along the mechanistic pathways. Isotope effects have been calculated for these stationary point structures, which has allowed for location of the true contributing state to the experimentally determined isotope effects. An interpretation of how to better interpret enzyme isotope effects is offered alongside these results to aide future validation of computationally determined enzyme mechanisms.

The study of mammalian PNPs is aimed at determining the origin of the difference in experimentally determined KIEs for arsenolysis of inosine catalysed by human and bovine PNP. Initial work discusses the similarity in free energy surfaces when semi-empirical methods are applied to a QM region consisting of solely the reacting substrates. This is followed by a study into the mechanism of the bovine isozyme using a larger QM region and application of higher-level QM corrections. A hypothesis for the mechanism is outlined, alongside attempts to corroborate calculated isotope effects with the experimentally determined effects.

Both chapters on enzyme-catalysed reactions contain discussion pertaining to the nature of enzyme catalysis. The origin of the relative stabilisation of transition state over substrate is outlined and comparisons are drawn between MTAN and PNP. Both classes of enzyme employ similar schemes of active site pincer residues, positioned at either end of the substrate binding site, to provide stronger hydrogen bonding interactions as the C_α-N bond length increases, to compensate for the endergonic glycosidic bond cleavage step.

Date of Award	18 Jan 2023
Original language	English
Awarding Institution	University of Bath
Supervisor	Ian Williams (Supervisor) & Christopher Pudney (Supervisor)