Electron cryotomography studies of membrane-associated influenza haemagglutinin

Abstract

The lipid-enveloped Influenza A virus (IAV) is responsible for yearly epidemics and sporadic pandemics. Important steps for virus entry into the host cell are mediated by the haemagglutinin (HA) glycoprotein on the virus surface. Following binding of cell surface receptors, the virus enters cells by receptor-mediated endocytosis. Acidification of the endosome triggers a low pH-induced conformational change in the HA that brings about fusion of the virus and endosome membrane. Studies of the HA by X-ray crystallography as well as electron microscopy have revealed structures of the protein ectodomain at neutral pH and at the pH of membrane fusion. However, to fully understand biological functions of the HA, it is important to study structures and structural changes in the context of membranes. Cryotomography has previously been applied to image virus fusion with liposomes which suggests that in addition to structural changes in the HA, ultrastructural changes in the virus may be important to membrane fusion. The aim of this thesis is to understand the structure of HA and its role in membrane fusion in the absence of other viral components to provide a comparison to protein and membrane structure observed in studies on virus.

As the main approach, I use cryo-electron microscopy to study full-length influenza HA reconstituted in liposomes (virosomes). I perform cryotomography and subtomogram averaging to obtain a structure for the ectodomain on virosomes at sub-nanometer resolution, showing that the HA ectodomain has the same structure as the isolated ectodomain from X-ray studies. I also show that the HA is flexibly linked to the membrane, develop a method to measure the angle of the HA with respect to the membrane by the relative orientation assignment of subtomograms, and discuss the role of flexibility in membrane fusion. I go on to image the association and fusion of virosomes with liposomes at the pH of fusion and compare these to images and tomograms of influenza virus fusion with liposomes with and without the presence of an inhibitor of the M2 proton. While initial stages in membrane fusion of virosomes resemble that of virus fusion, later steps show differences which identify potential roles of the viral matrix layer in membrane fusion. Virosomes exposed to the pH of fusion also, on occasion, presented hexameric lattices of HA proteins on their surface, a feature not observed on virus. I then go on to apply subtomogram averaging to HA on virus and virosomes following incubation at low pH and proteolysis, revealing a structure on the membrane similar to the structure of the fragment of the HA generated at membrane fusion pH and determined by X-ray crystallography. This structure supports previous studies that show the transmembrane and fusion peptide regions colocalise at the membrane when the HA protein has refolded into its post-fusion form.