Bluetongue virus (BTV) is an arbovirus (transmitted by Culicoides midges) remarkably variable in its host range (cattle, sheep, deer…) and clinical manifestations ranging from asymptomatic infection to lethal haemorrhagic fever. This variability is due to several factors related both to the infected host and the viral serotypes/strains. My project aims to develop high-throughput screening approaches, notably based on the yeast two-hybrid (Y2H) system, to map interactions between viral and cellular proteins to provide a systematic picture of cellular machineries that are essential targets of BTV. Viral proteins from two unusual serotypes of BTV were used as baits to screen two cDNA libraries originating from hosts naturally infected by BTV: Culicoides and cattle. These libraries have been chosen to provide a comparative view of virus-host interactions at multiple levels: viral (between serotypes) and host (mammalian host versus arthropod vector, which is quite unique in the field of arboviruses). Identification of cellular components that are specifically targeted by one or several serotype(s) explains the virulence/pathogenesis and/or cross species transmissions associated with its infection. On the contrary, by targeting conserved protein functions, it should be possible to design generic drugs against multiple serotypes, in particular by identifying peptides that are able to disrupt the interactions previously identified. This functional interactomic approach has now been extended to 12 other pathogens with known or potential risk of zoonotic transmission to human, representing more than 200 Y2H screens that led, up to date, to the identification of about 4 000 pathogen-host protein interactions.