GrASP Project

Viruses replicate into the host cells and interact with multiple host proteins, also called viral interacting proteins. These viral interacting proteins come in two flavors: (i) the cellular co-factors, which viruses usurp to complete their replication cycle, and (ii) the cellular immune antiviral proteins, which directly recognize the virus and/or are being counteracted by the pathogens. Therefore, host proteins and pathogenic viruses are engaged in a genetic conflict that follows the Red Queen model. In particular, at the site of virus-host protein interactions, the host genes evolve very rapidly and signatures of episodic positive selection are left over evolutionary time. These episodes are driven by physicochemical adaptation to the matching viral protein site. We propose to use this peculiar evolutive scheme to identify among the hundreds of interferon-stimulated genes (ISGs), those that are in natura in direct relevant contact with a pathogenic virus and/or potent relevant antiviral effectors.
Hence, we will build a new approach to detect site-specific positive selection, and that considers explicitly the physicochemical properties of the amino acids. In this approach, on a classic modeling based on a given codon model – of any kind, two robust methods will be added, stochastic mapping and multivariate analysis. This will allow us to better detect rapid evolution that implies important changes in interaction sites and may therefore be more reflective of an arms-race between two directly interacting adversarial proteins. Furthermore, from a short list of ISGs that target HIV in vitro, we will discover the ones that are relevant “antiviral genes” in primates. To do so, we will use the herein developed model and other evolutionary-based methods in combination with functional assays of virus-host interactions (i.e. we will notably test the capacity of the host protein to restrict viral replication in various in vitro infectivity assays). For the proteins that have a confirmed anti-retroviral phenotype, we will further use the information provided by the evolutionary-guided approach to dissect the exact virus-host interface and the antiviral mechanism.
Finally, we will address a major question the field: which of the hundreds of ISGs are encoding for direct viral effectors and viral interacting proteins. To do so, we will upscale and automate the evolutionary screen to identify which of the hundreds of ISGs have been under adaptive selection in primates. We predict that these signatures are the result of a genetic conflict with pathogenic viruses during primate evolution and that the encoded proteins are bona fide novel viral interacting proteins. We will further test this in virus-host in vitro assays during the grant period and in the mid-term.
This project blends evolutionary analyses and functional assays of virus-host interactions to answer major questions in the field of infectious diseases, in particular in cellular immunity and viral infections, as well as to build new methodological evolutionary models that will have broad scientific interest. The research program will bring discoveries that can ultimately lead to novel antiviral strategies, as well as innovative ways to prevent emergences (i.e. antiviral proteins may be species barriers) of pathogenic viruses, such as HIV.

Co-leaders : Lucie Etienne, team "Interactions hôte-pathogène pendant l'infection lentivirale" (CIRI) & Laurent Guéguen, team "Bio-informatique, Phylogénie et Génomique Evolutive" (LBBE).
Collaborative teams: Julien Dutheil (Max Planck Institute for Evolutive Biology, Plön, Germany).
Project duration : 3 years
Financing : PhD and engineer fellowships & consumable money
PhD fellow: Léa Picard
Engineer: François Gindraud