POCKETSPPI · Discovering and exploiting hidden pockets at protein-protein interfaces

Discovering and exploiting hidden pockets at protein-protein interfaces

The objectives of this project were to develop, validate and apply new computational methodologies that, starting from a protein structure solved by experimental methods, identify and characterize alternative conformational states that are deemed more “druggable” by small molecules. In selected cases, predictions from computational modelling were married with in vitro experiments in collaboration with other research groups. Diverse sub-goals were set and completed during the project. In the first aim of the work, initial efforts were focussed on the elucidation of the mechanisms of small molecule binding to the intrinsically disordered protein c-Myc. Our simulation studies have ruled out the presence of stable hidden pockets in this target protein and suggest instead that current inhibitors bind to this target protein through weak, non-specific interactions. Additional work used molecular dynamics simulations to detect transient conformations that are ‘’druggable’’ by small molecule. This has led to the development of the JEDI algorithm to predict the druggability of a protein structure. A unique aspect of the JEDI approach is that the druggability descriptor is fully compatible with ‘’on the fly’’ molecular dynamics simulations. This enables us to perform biased molecular dynamics simulations to steer protein conformations towards a priori-unknown, yet druggable, conformational states. Promising results have been obtained and published in leading chemistry journals, and work on further developments of the JEDI tool continues beyond the funding period of this project. We have also initiated a collaborative computational/experimental effort to characterise the binding of macrocyclic ligands and small fragments to selected cyclophilin proteins. Molecular simulations, X-ray analyses and in vitro assays have identified fragments that bind weakly to different regions of diverse cyclophilin proteins and via distinct molecular recognition mechanisms. The results of this work have not yet been published, by funding has been secured to develop further the results via follow-up experiments aimed at improving binding affinity and selectivity of the cyclophilin fragments. Overall the primary outputs of the project were: 1) improved computational methods to guide the rational structure-based design of protein ligands; 2) new insights into biomolecular recognition.

Текст от CORDIS, на английски · Данни: CORDIS, © Европейски съюз