Combined approaches to map protein-ligand interactions using NMR and X-ray crystallography

Understanding protein interactions with modulators, signalling molecules or inhibitors can give information about ligand binding sites, ligand interaction topology and allosteric regulation at an atomic level, which is key for the design of novel inhibitors and protein molecules. Many techniques in-vitro have been developed and utilised including X-ray crystallography, nuclear magnetic resonance (NMR), electron microscopy, mass spectrometry etc... Further advances in these biophysical techniques and combining the information from the different techniques are still required to provide detailed insights into the structure and function of proteins and relate these to the physiological environment.

Both X-ray crystallography and NMR techniques were explored with a number of previously characterised protein models, including Cyclophilin A (CypA) (18 kDa), the catalytic domain of Phosphodiesterase B1 from Trypanosoma brucei (TbrPDEB1) (37 kDa) and the catalytic domain of Bromodomain 4 BD1 (BRD4-BD1) (15 kDa). The results were mixed for each protein system. CypA showed successful NMR spectra but X-ray structures were less amenable. TbrPDEB1 was a good crystallographic system, where four novel liganded crystal structures were solved, but was deemed unsuitable for NMR studies. BRD4-BD1 was a successful crystallographic system, and also suitable for ligand and protein-observed NMR utilising 1H, 15N and 19F nuclei.

Seven novel liganded BRD4-BD1 crystal structures and NMR assignments are presented in this thesis. The pair of triple resonance experiments HNCA, HN(CO)CA, HNCACB, HN(CO)CACB for BRD4-BD1 led to the sequential backbone assignment of 1HN, 15N, 13Cα and 13Cβ nuclei. Major chemical shift differences ranging above 0.3 ppm were observed for the residues W81, V87, D88, A89, N93, L94, I146, A150, L158 and K160 with the ligand BTB 07004. This chemical shift mapping was in agreement with the binding site in the

19

BRD4-BD1 crystal structures, particularly for the residues W81, V87, L94, Y97, C136, Y139, N140 and I146.

The thesis confirms that combining X-ray crystallography and NMR leads to an advanced understanding of ligand interaction sites in proteins. This could be used for generating accurate topology maps of ligand-binding sites of any proteins with specific ligands, known as the "Group Epitope Mapping" (GEM) in the future and be extended to novel proteins.