Résumé
Myxovirus resistance 1 (MX1) proteins are potent viral restriction factors that are induced by interferon (IFN) and are most famously known for their inhibition of influenza A virus (IAV). These antiviral dynamin-like GTPases have been long-studied and yet the molecular mechanisms of how they inhibit IAV are not fully understood. They are nevertheless known to restrict early stages of IAV lifecycle at the primary transcription and/or replication steps. Furthermore, several intrinsic determinants are required for this antiviral activity, namely an intact GTPase domain, oligomerization through the stalk domain and an intact Bundle Signaling Element (BSE), L2 and L4 loop. In this thesis, I firstly extend this list by showing that the N-terminal domain and more precisely an evolutionarily conserved leucine is essential for the antiviral activity against IAV and other RNA viruses as well as being important for correct subcellular localization of MX1 proteins. Secondly, I redefine human MX1 (HsMX1) and mouse Mx1 (MmMx1) as restriction factors of early and late stages of IAV infection. Indeed, I show that MmMx1 and to a lesser extent, HsMX1, inhibit the splicing of IAV M and NS segments resulting in the nuclear localization of NP at late timepoints post-infection, potentially due to an inability to form viral nuclear export complexes. In addition to this, HsMX1 is able to induce an inhibition of cytoplasmic viral trafficking complexes that accumulate at the perinuclear region into a condensed endocytic recycling compartment (ERC). Furthermore, HsMX1 is able to impact virus-induced Rab11a inclusion morphology and induce their trafficking back towards this condensed ERC, providing the first evidence of a restriction factor able to act on virus-induced inclusions. These new findings provide new pieces of the puzzle that aims to understand the mechanisms of action of these IFN-induced Swiss knife-like GTPases.