AbstractHydrogen deuterium exchange coupled to mass spectrometry (HDX-MS) is a powerful and sensitive technique for the analysis of protein dynamics in solution. HDX-MS reports on the transient uptake of deuterium by a protein in bulk deuterated solvent due to the phenomenon of hydro-gen exchange, whereby labile hydrogen atoms can spontaneously exchange for other hydrogens or deuterium in solution. HDX-MS is typically used to inform about the location of binding interfaces between proteins and other proteins/small molecules by comparing exchange profiles of a protein in its bound and unbound states, where regions of comparatively reduced deuterium uptake are indicative of a binding surface. HDX-MS is however not typically used to inform on the structure of proteins by itself due to a lack of understanding about the factors that cause changes in exchange rate.
In this thesis, we set out to develop a method that could accurately discriminate between native and non-native protein structures using nothing but the protein’s experimental HDX-MS profiles. This was achieved by coupling state of the art mass spectrometry to computational chemistry techniques in order to develop a method that could accurately calculate HDX-MS profiles from in silico three-dimensional structures and compare these calculated profiles to experimental data in order to classify the structures as being either native or non-native. We achieved reasonable classification accuracy using this method over the course of this thesis with the potential for much greater ac-curacy with subsequent research and development. We also took the first steps towards modifying this methodology to work on classifying binary complex structures as well as individual protein structures.
In addition to our primary focus on structure classification, we also undertook a more traditional HDX-MS side project involving the determination of the location of the binding interface between the enzyme dUTPase and its inhibitor Stl. We successfully characterised this interaction and helped develop a model of the mechanism of inhibition based on our data.
The work presented in this thesis is extensive in its breadth and variety, incorporating a diverse range of different techniques spanning multiple scientific disciplines. From classical biochemical approaches such as the manipulation of DNA, cell culture and the production of proteins to analytical chemistry in the form of HDX and native mass spectrometry and computational chemistry and computer science techniques such as Molecular Dynamics, protein docking and Python programming.
|Date of Award||1 Jul 2021|
|Supervisor||Antoni Borysik (Supervisor) & Rivka Isaacson (Supervisor)|