In silico characterization of substrate and inhibitor specificity of histone deacetylases

Abstract

The field of epigenetics has received considerable attention in recent years due to its impact on genetics, developmental biology, cancer biology, and medicinal chemistry. This research project focused on histone deacetylases (HDAC), a family of epigenetic regulator enzymes that contribute to the DNA expression regulation, through hydrolysis of the ε-N-acylated lysine residues of histone and non-histone protein side chains. The first chapter reports the in silico study performed to evaluate the binding mode and affinity of a collection of known macrocyclic HDAC inhibitors and their analogous, towards class I HDACs. To the present date, a few HDAC inhibitors have been approved for treatment of various cancers. One particularly interesting class of HDAC inhibitors is the macrocyclic peptides and depsipeptides, which are highly potent and moderate selective, and can be found in nature. My results confirmed the higher potency of hydroxamate analogues in comparison with their norvaline counterparts, as well as the crucial role of an aspartate in achieving the optimal binding position. Furthermore, the unexpected interaction between aromatic side chains of the inhibitors and the catalytic zinc ion opened a new line of investigation for possible HDAC inhibitors. Chapters 3 and 4 focuses on class III HDACs, also known as sirtuins, a class of NAD+- dependent enzymes with homology to the silent information regulator 2 yeast enzymes. There are seven proteins in the sirtuin family and they all share a conserved 270 amino acid catalytic domain, with variable N- and C- termini. These enzymes have been suggested as therapeutic targets for diabetes, cancer, neurodegenerative diseases and inflammation, which has led to the investigation of regulatory molecules. In Chapter 3, the NADH inhibition of SIRT1, SIRT3 and SIRT5 was investigated using long Molecular Dynamics simulations. The results showed that the addition of a single proton in the nicotinamide ring can induce an important conformational change on NADH, causing misalignment of the nicotinamide amide to the key residues in the C pocket and ultimately resulting in a significant loss of binding affinity. Chapter 4 reports the generation of the first homology model of SIRT7. With reasonable (39%) to high (100%) identity templates, this model includes the secondary structure of both N- and C- termini. The catalytic domain, similar to SIRT6, depicts a Rossmann-fold and a Zn2+-binding domains. The termini show a well-defined a-helix organization, essential for the DNA binding. The model is further supported by phylogenetic and structural analysis, and secondary structure prediction.