Epigenetic mechanisms in silico : understanding demethylation and rational design of bromodomain inhibitors

Histone octamer proteins are crucial for DNA packaging and storage in the confined space of the cell nucleus. Epigenetic post-translational modifications to specific histone monomers facilitate changes in chromatin flexibility that are necessary for access by transcription machinery, and therefore have control over gene expression. Targeting enzymes that regulate these powerful modifications has been established as a promising strategy in the treatments of some diseases, such as cancer, male infertility and adult obesity. The work presented in this thesis aims to shed light on the mode of action of epigenetic proteins, and in so doing, aid in the design of more potent and selective small molecules that target them (e.g. Jumonji C (JmjC) and bromodomain containing-proteins). We have used a diverse array of computational techniques such as quantum mechanics (QM), classical Molecular Dynamics (MD) simulations, and hybrid Newtonian molecular mechanics/quantum mechanics (QM/MM) approaches. Following an introduction to these computational chemistry techniques (Chapter 1) and a description of the methodology used in this thesis (Chapter 2), we then present our studies on: (a) improving Bromodomaincontaining proteins inhibition (Chapters 3, 4) and (b) providing novel insights into the lysine demethylation catalysed by JmjC proteins (Chapters 5, 6). In Chapter 3, we have investigated a series of dihydroquinoxalinone (DHQ) derivatives, analogues of (R)-2 (Figure A.1), previously proven inactivators of the epigenetic molecular target CREBBP bromodomain, resulting in the first potent inhibitors of a bromodomain outside the Bromomodomain and extra-terminal (BET) family. Through QM and MD calculations, I established the significance of cation-π interactions for the binding of CREBBP inhibitors. In Chapter 4, we then developed a novel electrostatic model for the quantification of cation-π interactions in biological environments, inspired by a good fit between experimental binding affinity data of a series of fifteen 5-isoxazolylbenzimidazole derivatives and electrostatic potential (ESP) surfaces. This model has been prospectively applied to newly synthesised DHQ derivatives. Our approach could be used for the development of force fields and docking scoring functions with increased reliabilities to reproduce cation-π interactions. Together with the understanding of DHQ binding in CREBBP, I have investigated one of the most common but poorly understood epigenetic processes: histone lysine demethylation by a JmjC protein, JMJD2A (Figure A.2). In Chapter 5, in the first QM/MM studies reported for this system, we describe the importance of the protein environment on the binding of molecular O₂, while in Chapter 6 we analyse the PES of the elementary steps associated with enzymatic demethylation: cofactor/ O₂ activation, C-H abstraction and hydroxyl-rebound. Insights into this reaction mechanism and energetic contributions have been used to the role of specific JMJD2A residues in this process, which can be used to accelerate the design of effective drug molecules.