• Internal search
  • Xavier Morelli
    +33 (0)

    Protein-Protein Recognition : Specificity and Inhibition

    The 2P2I Approach

    This work was initiated in 2004 for the design of inhibitors targeting the Nef protein from the HIV virus, in collaboration with Dr. Yves Colette (INSERM-Marseille) and with a strong financial support from the National Agency of Research against AIDS (ANRS, 2 research grants and one PhD grant for the 2006-2010 period).

    The Inhibition of protein-protein complexes is limited by several parameters such as biophysical parameters (size & depth of the interface), chemical parameters (chemical space available which does not correspond to what could inhibit a protein-protein interaction) and conceptual parameters (existence of ‘reference’ compounds to help in the discovery of new compounds).

    The principal handicap relates to the latter where the absence of ‘reference’ molecules is a real impediment in the optimization of new lead compounds as biological probes/tools or potential drugs. To circumvent this problem we thus proposed a methodology in three stages which was called 2P2I (for Protein-Protein Interaction Inhibition) (Betzi et al., PNAS, 2007). We proposed to identify the first ‘hits’ using a combination of an original in silico screening protocol (Betzi et al., J. Chem. Inf. Model, 2006) coupled to NMR validation (Morelli & Rigby, Curr. Computer-Aided Drug Disc., 2007). The goal of this approach is to test a minimum number of molecules from a diverse database of selected chemical compounds.

    Thanks to this approach we could highlight the first two inhibiting molecules of a complex involving the Nef protein from HIV (Betzi et al., PNAS, 2007 and Patent Publication n°: WO/2009/063412 International Application n°: PCT/IB2008/054741). This approach is now applied within our in silico screening platform (INT-3D) with several protein-protein complexes targeting the development of antivirals (SARS, HCV, measles virus, dengue virus…) and of anti-cancer compounds (prostate, brain and severe septic shock) (see § below for more details).

    Specificity in Protein-Protein Complexes. 

    This research project is in collaboration with Dr. Y. Collette, Dr. ST. Arold (University of Texas, USA) and Pr. M. Huang from Harvard University & Fuzhou University, China. Two PhD students: Adrien Lugari (PhD grant from the National Agency of Research against AIDS) and Xiaoli Shi (PhD grant from the French Embassy in China) are involved in this project. ANRS also financed this project through a research grant (2008-2010).

    Fyn SH3 domain interacts with the viral protein Nef but is also able to recognize other cellular proteins specifically, through the same three-dimensional mode of interaction. We identified, using double hybrid methodology (collaboration Dr. Y. Collette), multiple human cellular proteins which interact with Fyn SH3 domain via a specific recognition of the RT loop. We would like now to understand how this SH3 domain is able to recognize specifically these proteins by solving the 3D structure (x-ray diffraction) of these complexes and to compare the molecular basis in the understanding of the specific 3D molecular recognition of SH3 domains with their cellular or viral partners. This project is strongly supported by the ANRS which has financed a PhD grant (Mr. Adrien Lugari, third year PhD student) and a research project (2008-2010) as by the French Embassy in China which also granted a PhD grant on this project (Mrs. Xiaoli Shi, third year PhD student). We also obtained an Advanced Research program (PRA) that we got in collaboration with the laboratory of Professor Huang (Harvard Medical School, US & Fuzhou University, China) on this project. One research article has been already published on ALIX protein (Shi et al., Biochem. J., 2010).

    The INT-3D Platform

    In collaboration with Dr. Philippe Roche

    As mentioned above, the success of the Nef-Fyn project has permitted us to present our 2P2I approach in numerous international conferences (26 invited conferences since 2006, 11 international – 15 national) and thus to extend the approach to other protein-protein complexes targeting viruses such as the Measles virus (Bernard et al., FEBS Letter, 2009), HCV virus (Betzi et al., antiviral research, 2009), SARS virus (Lugari et al., J. Biol. Chem, 2010), Dengue virus etc.. or cancer (Salmi-Smail et al., J. Med. Chem., 2010) such as prostate cancer (PAIR prostate grant just obtained for the next 3 years incollaboration with Dr. Palma Rocchi, INSERM-Marseille), brain cancer or severe septic shock (El Turk et al., Bioorg. Med. Chem., 2010). The increasing demand in anti-viral and anti-cancer research projects led Xavier Morelli & Philippe Roche to create a platform (INT-3D, director: X. Morelli; which handle these ‘side projects’. This platform has been already validated by the “cancéropôle PACA” and is a GIS-IBiSA validated platform through a strong collaboration with the PCML platform in Marseille-Luminy University (director: JC. Guillemot). We recently obtained a specific grant for the screening of the French National Chemical Library on the SARS coronavirus replication complex (Program Innovation Recherche PIR Thérapeutique, CNRS 2010) and also signed a collaborative project with a pharmaceutical company (2010-2011). We are now also part of the GDR “Chemoinformatics”.

    The 2P2I Database

    In collaboration with Dr. Philippe Roche

    A detailed description of the chemical and topological spaces of protein-protein interfaces that can be disrupted by small chemical compounds was not available. Such a description would lead to a better definition of a potentially successful PPI target. To define the general parameters that characterize known PPIs targets, we analyzed 17 protein-protein hetero-complexes that have been structurally characterized by X-ray crystallography or NMR.

    For each complex family, a three dimensional structure of one of the two partners bound to a small molecule inhibitor was also available, representing a total of 56 different small molecules inhibitors. We have then analyzed the protein/protein and protein/inhibitor interfaces in terms of geometrical parameters, atom and residue properties, and buried accessible surface (Bourgeas et al., PLoS ONE, 2010). The result of our study has permitted to expand the current knowledge with new data, focusing at the interface of protein/protein complexes (refer to review Betzi et al. Comb. Chem & HTS, 2009 or book chapter by Roche and Morelli, ‘In silico lead discovery’, Ed. Bentham, 2010). We have developed the 2P2I database (2P2IDB) containing the dataset used in this survey. This interacting structural database is publicly available and can be accessed and queried online at The general biophysical, biochemical and structural properties of the interfaces found in 2P2IDB were compared to those of representative datasets of heterodimers complexes in their free, bound or ‘inhibited’ biological conformation. We have then established a characteristic profile for “druggable” protein-protein complexes (presence of minimum 2 pockets at the interface, structural rearrangement <2A, #segment at the interface, # of hydrogen bond per 100A² etc…). This characteristic profile will be used to screen the Protein Data Bank (PDB) in order to select the most promising target complexes that will be screened in the second phase of our program with our biological partners from neighbor institutes and in partnership with chemical providers (private or academic) and with pharmaceutical companies (already identified).