Library Design for Tackling Protein-protein Interactions: The 2P2I Approach
Conference Recording Jun 10, 2013
About the SpeakerDr Philippe Roche is a senior scientist at the CNRS, Marseille, France. He started his career in Toulouse (France) where he got its PhD in structural biochemistry, working on the structure of carbohydrates involved in Plant-Microbe interactions (Rhizobium Legume-Symbiosis). This research led to the development of the patented LCO Promoter Technology used as nitrogen inoculants in the USA and South America by Novozymes . Since 2012, he has moved to the Laboratory of integrative Structural & Chemical Biology (iSCB), at the Cancer Research Center in Marseille where he is in charge of chemoinformatics and molecular modeling projects.
In the last decade, the inhibition of protein-protein interactions (PPI) has emerged from both academic and private research as a new way to modulate protein networks. Due to the implication of PPI in numerous diseases, inhibitors of these original interactions are certainly the next generation of highly innovative drugs that will reach the market. However, in silico design of such compounds still remains challenging. We have developed 2P2IDB, a hand-curated structural database dedicated to PPI with known orthosteric inhibitors. Using structural knowledge from the recent success stories our goal is to derive some common principles to help future target selection by assessing the druggability of PPI and to accelerate the process of drug discovery by improving the quality of chemical libraries dedicated to PPI. Analysis of the small molecule inhibitors present in 2P2IDB led us to propose the ‘rule-of-four’ as a guideline to characterize PPI inhibitors. Using dedicated support vector machine approaches, we have developed 2P2IHUNTER, a tool for filtering potential orthosteric PPI modulators from large collections of compounds (article under review). This innovative tool has been applied to a set of 8.3 million compounds from the “big vendors” to design several in silico PPI focused chemical libraries. Compounds corresponding to medicinally important privileged structures, identified as core structures in numerous therapeutics, were prioritized in a medicinal oriented version of the library. The library was filtered with carbon bond saturation index (Fsp3) to escape from flatland, which resulted in a structurally-diverse chemical library of 1,683 compounds. The molecules have been purchased from the providers, stored in 364-well plates and are currently tested against a standard PPI target (MDM3/P53) to evaluate their ability to enhance hit rates in general screening campaigns. The design and molecular properties of the different in silico chemical libraries and preliminary results from the experimental screening will be discussed during the presentation.