Mihaela Smilova Abstract

Fragment Hotspot Mapping to Drive the Semi-Automated Elaboration of Fragment Screening Hits

Mihaela Smilova1, Peter Curran2,3, Chris Radoux4, Will Pitt5, Jason Cole3, Anthony Bradley6, Brian Marsden1

1Nuffield Department of Medicine, University of Oxford, Oxford, OX3 7DQ, United Kingdom
2Department of Chemistry, University of Cambridge, Lensfield Rd, Cambridge, CB2 1EW, United Kingdom
3Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge, CB2 1EZ, United Kingdom
4European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, CB10 1SD, United Kingdom
5UCB, 208 Bath Road, Slough, West Berkshire SL1 3WE, United Kingdom
6Exscientia Ltd, 36 St. Giles’, Oxford, OX1 3LD, United Kingdom
The past decade has seen an explosion in the availability of genomic and structural data for a great number of biomolecular disease targets. Rational drug discovery aims to use this knowledge to assist in designing chemical and biological agents that modulate the activity of these targets. Despite recent technological advances, the development of these agents is exceedingly expensive, is not routine, and carries the risk of the compound failing after many years of development. This means that there is great need for methods that can streamline and automate the development process, both for drugs, as well as for small molecule chemical probes, the highly characterized, high quality agents which can be used to investigate and dissect the underlying biology of disease. Fragment based drug discovery (FBDD) has established itself as a powerful tool for developing drug and probe candidates by rationally elaborating small chemical fragment hits into larger, optimised lead compounds. Recent technological developments have enabled the use of X-ray crystallography as a medium-throughput screening tool for FBDD, resulting in a wealth of structural data on low molecular weight molecules in complex with a protein target. Interpreting this data and distilling it into prioritised suggestions for the elaboration of fragment hits into leads with increased potency and selectivity for the target protein is currently a significant challenge to using this technique. Thus, computational methods to address this problem are in demand.
Introduced in 2016 by Radoux et al., fragment hotspot mapping is a new and promising method that highlights the specific interactions a protein makes with a fragment to drive its binding[1]. The resulting contoured score maps calculated for a single protein/fragment structure provide an intuitive visualisation of key features of the binding site. The highest scoring interactions predicted are often those that drive binding of the initial fragment, and moderately scoring areas correspond to areas of the binding site into which initial fragment hits can be elaborated. This makes the method a promising tool within structure-based drug discovery campaigns. However, crystallographic FBDD campaigns result in multiple structures of the same protein, often with evidence of intrinsic protein flexibility. To accommodate this variability across a FBDD dataset, we combined fragment hotspot maps calculated for each protein/fragment structure into an “ensemble hotspot map” for the protein target. We also investigated ways in which the ensemble maps of a target and a related off-target protein from the same protein family can be compared, with the aim of developing a method that can be used to inform the elaboration of fragment hits into lead compounds with a specific selectivity profile within a target protein family. Current work uses the ensemble and difference maps as a starting point for the semi-automated detection of areas in which a fragment can be preferentially elaborated, and especially areas denoting interactions that could confer selectivity over an off-target protein. A set of diagnostics for assessing the quality of the method predictions has been developed, based on ensembles of apo, fragment-bound, and selective inhibitor bound structures of proteins in two well-researched families: protein kinases and bromodomains.

References
1. Radoux, C. J., Olsson, T. S. G., Pitt, W. R., Groom, C. R. & Blundell, T. L. Identifying Interactions that Determine Fragment Binding at Protein Hotspots. J. Med. Chem. 59, 4314–4325 (2016).