Human/bovine chimeric MxA-like GTPases reveal a contribution of N-terminal domains to the magnitude of anti-influenza A activity

Type I interferons (IFN-) provide powerful and universal innate intracellular defence mechanisms against viruses. Among the antiviral effectors induced by IFN-, Mx proteins of some species appear as key components of defence against influenza A viruses. The body of work published to date suggests that to exert anti-influenza activity, an Mx protein must possess a GTP-binding site, structural bases allowing multimerisation, and a specific C-terminal GTPase effector domain (GED). The human MxA and bovine Mx1 proteins both meet these minimal requirements, but the bovine protein is more active against influenza viruses. Here we measured the anti-influenza activity exerted by two human/bovine chimeric Mx proteins. We show that substituting the bovine GED for the human one in human MxA does not affect the magnitude of anti-influenza activity. Strikingly, however, substituting the human GED for the bovine one in bovine Mx1 yields a chimeric protein with much higher anti-influenza activity than the human protein. We conclude, in contradiction to the hypothesis currently in vogue in the literature, that the GED is not the sole determinant controlling the magnitude of the anti-influenza activity exercised by an Mx protein that can bind GTP and multimerise. Our results suggest that one or several motifs that remain to be discovered, located N-terminally with respect to the GED, may interact with a viral component or a cellular factor so as to alter the viral cycle. Identifying, in the N-terminal portion of bovine Mx1, the motif(s) responsible for its higher anti-influenza activity could contribute to the development of new anti-influenza molecules.