Understanding vaccinia virus entry by super-resolution microscopy and particle averaging

While viral structure and function are closely linked, the small size of viruses renders their visualisation by light microscopy difficult, often making investigation of this relationship challenging. The advent of super-resolution microscopy now allows for component-specific imaging of viruses at resolutions that allow for the investigation of virus architecture. In addition, the ability to image many copies of a virus opens up the possibility for structural mapping of individual virion proteins. In this thesis, a new analytical approach, VirusMapper, is developed to facilitate the application of single-particle analysis to super-resolution images of viruses and other macromolecular complexes. Detailed validation of VirusMapper demonstrates that it allows for the generation of high-confidence, robust 2D structural models of stable multi-protein complexes. VirusMapper is then applied to model the substructure of the prototype poxvirus, vaccinia, one of the most complex mammalian viruses that exists. Models of the substructure of vaccinia with high precision and accuracy are produced, both with isolated viral particles and during the early stages of infection. Furthermore, through these studies a novel structural feature of the membrane protein architecture of this virus is identified: the vaccinia fusion machinery is polarised. Further investigation demonstrates that fusion protein polarisation depends on fusion machinery intactness and the presence of a non-fusion protein, A27. Functional characterisation of a mutant virus lacking A27, in which the fusion machinery is relocalised but retains its membrane fusogenic activity, reveals a severe defect in fusion pore formation. The power of super-resolution microscopy and averaging are thus harnessed to reveal a structure-function relationship between virus fusion machinery localisation and fusion efficiency