Unnatural selection: using an expanded genetic code to enable site-specific drug discovery

Encoded library technologies have revolutionised early drug discovery, enabling up to 10^13 compounds to be screened simultaneously against a target protein. This pooled approach is facilitated by the attachment of DNA barcodes to every library member, allowing for the deconvolution of hits by next generation sequencing. These methods have been exceptionally useful for the discovery of compounds for a wide range of proteins, however there are currently no methods to allow screening to be performed at a specific site on a target protein. This would be particularly beneficial for screening allosteric binding pockets and for the disruption of selected protein-protein interactions. This project will develop an encoded library platform that directs screening to a desired site on a target protein, facilitating the discovery of compounds that modulate specific functions. Using genetic code expansion, reactive amino acids will be incorporated into the target protein to covalently capture library members bound at that site. The platform will be demonstrated by screening libraries against an allosteric site on Aurora kinase A, an important cancer target, to identify novel tool compounds. This technology will be broadly applicable to any site of interest, thereby accelerating the discovery of useful tool compounds to probe biological functions.