Study of autophagy-mediated restriction of avian influenza virus replication in mammalian cells

Abstract

Host barriers restrict cross species transmission of animal influenza viruses to humans. However, in recent decades, some H5 and H7 subtype avian influenza viruses with adaptive mutations in the viral genome have caused limited human infections. There are concerns that such adaptive strategies may allow avian influenza viruses to gain further mutations that enable efficient transmission in humans, leading to a new pandemic. Adaptive mutations in the PB2 proteins of influenza A viruses are common and well-studied, but details of the mechanism underlying PB2 host adaption remain unclear. My study revealed distinct differences between vRNA trafficking patterns in mammalian cells infected with H7N9 viruses isolated from human cases and avian hosts. Through analysis of a series of reassortant viruses, I confirmed that avian-like PB2 causes the formation of vRNA aggregates in infected mammalian cells, whereas acquisition of adaptive mutations in the PB2 subunit subverts accumulation of vRNA aggregates during virus replication. A detailed mechanism was characterized using a pair of PB2-627K (human-like) and PB2-627E (avian- like) A/WSN/33 background viruses to delineate host restriction associated with the PB2 protein. While Rab11-positive recycling endosomes play a key role in viral ribonucleoprotein (vRNP) trafficking during influenza virus infection, my results showed that Rab11 is not responsible for the formation of vRNA aggregates in cells infected with WSN-627E virus. Further characterization of the vRNA aggregates in WSN-627E infected human cells showed that they are vRNP aggresomes, as evidenced by their features, including proximity to the MTOC (microtubule organizing centre), formation via the microtubule network, altering the distribution of the Golgi complex and changing the structure of the intermediate filament (IF) protein vimentin. Importantly, LC3, the hallmark of autophagy, was found to co-localize with these vRNP aggresomes. Therefore, it seems likely that replication of avian PB2-627E virus is limited by the aggresome- autophagy pathway in mammalian cells. I further characterized the interaction between influenza virus vRNPs and the autophagy cargo receptor p62, and identified co-localization of p62, Rab11, ubiquitin, and LAMP1 (a lysosomal marker) with vRNP in WSN-627E virus infected cells. This suggests that p62 targets vRNP to autophagosomes. The influenza A M2 protein is known to block fusion of autophagosomes and lysosomes to facilitate virus replication. I showed that overexpression of M2 protein or treatment with BafA1 (an inhibitor of autophagic flux and V-ATPase) disrupts the formation of vRNP aggresomes in mammalian cells. Moreover, inhibition of autophagic flux by the influenza virus M2 protein is known to enhance the efficiency of vRNP trafficking and virus replication. The study shows that autophagic flux efficiency in WSN-627E virus infected cells is higher than that in cells infected with WSN-627K virus. Therefore, inefficient blocking of fusion between autophagosomes and lysosomes leads to the formation of vRNP aggresomes when influenza viruses with unadapted PB2 infect mammalian cells. This study shows for the first time a mechanism by which mammalian cells restrict influenza virus intracellular trafficking through targeting avian-type PB2 subunits of newly synthesized vRNP to the autophagy pathway, and details adaptive strategies that alter PB2 to evade such restriction.