CHIPSeT

Protein complexes are at the heart of most biological functions. The CHIPSeT project aims at a deeper understanding of interaction surfaces and will integrate large-scale experimental and computational strategies to disentangle the complexity underlying protein-protein interaction (PPI) networks. It will first focus on newly discovered physical interactions ensuring the cross-talk between chromatin remodelers (INO80) and proteostasis pathways (Cdc48/VCP/p97). Combining large scale proteomic screens and genome-wide CHIP-seq technologies, partner 2 found that both machineries act in a concerted manner for the disassembly and subsequent proteolysis of large complexes involved in DNA-metabolic processes. A key limitation of the current approach is that the PPI-network of both INO80 and Cdc48 is highly intricate. To tackle these challenging systems and reveal the molecular logic associated with PPI networks, we propose to develop two connected strategies that exploit co-evolution information and rely on computational and experimental settings. First, partner 1 showed that the use of multiple sequence alignments to reveal co-evolutionary constraints acting at PPI interfaces could critically increase the reliability of structural models in docking simulations. The main future challenge will be to integrate co-evolutionary constraints not only at the coarse-grained step of the docking process but also at the fine-grained, during optimization of models. The second strategy has a highly innovative potential and consists in producing experimentally and artificially co-evolution events at high rates at protein complex interfaces. Next generation sequencing methods are providing us with the possibility to reach a very rich landscape of compensatory mutants which can be used with great potency in the modeling process. Generating this artificial coevolution at high rate is a significant challenge. In our project, we devised a strategy that takes the best from the expertise of partner 3 and 4 to propose a robust pipeline likely to fulfil our goals. On the one hand, we set up a two-hybrid system running with a variety of fluorescent probes to report for gain or loss of interactions in a relatively quantitative manner (partner 3). Next, compensatory mutations can be screened and prepared for sequencing very efficiently thanks to the cutting-edge droplet-based microfluidic technology (partner 4). Both co-evolution-based strategies should provide unprecedented power to explore and tune the properties of complex interfaces, in vivo. This work will provide invaluable tools for modeling protein interaction networks at large scale and analyzing the effect of mutations in pathologies connected to Cdc48/P97/VCP system or to specific subunits of the INO80 machinery.