Exploiting sequence variation to guide drug design

Structure-guided design of small molecule inhibitors relies on interactive cycle of in-silico design, synthesis of newmolecule, determination of inhibitory activity and/or binding, determination of crystal structure of target-inhibitor complex and use of that structure to guide the next round of design. This is process can be powerful and result in high affinity inhibitors, but at the same time it is far from perfect. In each round of design process a number of molecules are synthesised and tested, but only few of them are better than the starting molecule, even if the design suggest they should all be improved versions of the original molecule. This breakdown between prediction and realityis a result of the still limiting accuracy of in-silico design and lack of appreciation of subtle features (sequence, mobility, water structure) in the target and in its interactions with the inhibitor.
In this project we aim to make more efficient use of the designed inhibitors by using variations in homologous proteins, paralogues and orthologues, and evaluate how the design process can be used in parallel to target multiple proteins. As the proof-of-concept example, we will use an exisiting collection of inhibitors we have designed against a human protein and test how they can inhibit orthologous proteins from pathogenic parasites. We have a large dataset of SAR data and crystal structures of inhibitors against human recombinase RAD51 which we will use as test bed for targeting parasite recombinases. These inhibitors will be first analysed in silico against homology models of parasite RAD51 proteins and then selected inhibitors will be tested against parasite RAD51 to evaluate the predictive power of in silico analysis. We are expecting to see clear differences SAR between human and pathogen RAD51s and hopefully identify chemical series that are sufficiently different between the two targets and thus enables the development of new series of inibitors that distiguish between the two, taking the structural diversity into account.
In a second part of the project we will select a suitable family of human and parasite, or possibly microbial, drug targets and design new inhibitors simultaneously against this family of proteins, targeting conserved regions and seeking differentiating SAR from more diverse regions of the proteins. The two possible, and perhaps equally desireable, outcomes of this are i) inhibtors that act on targets from multiple organisms or ii) multiple, different chemical series that each inhibit the target from a different species, thus maximising the benefit of synthesing number of molecules at each round.