Influenza haemagglutinin: fuse or break trying?
To replicate, viruses must deliver their genome into a host cell. However, all enveloped viruses must first overcome a cellular entry barrier: the plasma membrane. To traverse this barrier, viruses exploit their fusion proteins, which have been perfected for this task throughout evolution. In this fascinating process, the influenza A virus uses its envelope protein, haemagglutinin, to fuse viral and endosomal membranes. Consequently, the viral genome is released into the cell and replication may begin. Understanding structural details of the influenza fusion machinery may reveal new ways to curb the flu.
Cryo-electron tomography allows the study of viral structural details in three-dimensions and in a frozen-hydrated state. Recently, cryo-electron microscopy was revolutionized by the direct detection of electrons and Volta phase plate (VPP), to improve contrast to the point that in-focus imaging is approximated. We have been studying the structural details of the influenza virus fusion machinery in vitro for using influenza A virus-like particles (VLP). VLP are structurally similar to the influenza A virus but unlike the virus, they can carry mutations in the haemagglutinin fusion peptide to arrest membrane fusion at intermediate stages. During our studies on the viral fusion process, an opportunity arose to use a transmission electron microscope equipped with both a direct electron detector and a VPP, which impressively improved the signal-to-noise ratio of cryo-electron tomograms.
Cryo-electron tomography using Volta phase plate allows imaging in 3D with high signal-to-noise ration and close to focus. At low cholesterol concentration the liposomal membrane (cyan) ruptures and is inserted in the influenza A virus-like particle (yellow) forming a hemifusion diaphragm with large diameter (~20nm). Scale bar: 50 nm.
It was exciting to use such a microscope located at Rocky Mountain Labs (RML) in Hamilton, Montana. The RML is famous for tremendous achievements in microbiology, including the discovery of the causative agents for the tick-borne diseases known as Rocky Mountain spotted fever and Lyme disease.
Rocky Mountain Laboratories established in 1928 are now part of the National Institute of Allergy and Infectious Diseases (photo kindly provided by RML).
The first step was to prepare enough samples to study under the microscope. Towards this goal, my colleague Elena Mekhedov and I spent five weeks producing VLPs carrying different mutations in the haemaglutinin gene. This would allow us to study different arrested stages of the membrane fusion process. We next vitrified mixtures of the VLP and liposomal artificial membranes mimicking the endosomal membrane subjected to low pH, triggering fusion. The samples were finally shipped in liquid nitrogen from our home laboratory at the Eunice Kennedy Shriver National Institute of Child Health and Human Development in Bethesda, Maryland.
Our collaborators at RML, Elizabeth Fischer and Cindi Schwartz trained me to use the microscope and together we collected data, which kept us all busy analyzing for more than half a year! To our surprise, regardless of the mutation in haemagglutinin, we found that haemagglutinin is able to rupture and stabilize edges on liposomal membranes. In addition, we detected around 50% of the fusion products being arrested in a stage called “hemifusion diaphragm,” which has never before been directly observed during viral fusion.
Cholesterol is known to affect the physical properties of lipid bilayers i.e. curvature and elasticity. We uncovered two independent pathways leading to hemifusion, depending on the concentration of cholesterol in the target membrane. Using mathematical models, our collaborators Fred Cohen (Rush University) and Rolf Ryham (Fordham University) calculated that below cholesterol concentrations of 31 mol% membrane appositional energy predominates, leading to rupture and abortive fusion.Read the paper: The hemifusion structure induced by influenza virus haemagglutinin is determined by physical properties of the target membranes