Micro-organisms such as bacteria and viruses frequently have a multi-layered protein shell, often containing 50 or more distinct proteins in a shell less than 100 nm thick. Distinguishing the order of protein layers can reveal the morphogenetic plan of the microbe, and hence shed light on the function of different proteins - for example, which proteins form the outermost layers that protect a bacterial spore from lytic enzymes in the environment, and which hold the structure together? The only practical method for non-invasively identifying specific proteins in these specimens is to use fluorescent fusion proteins and fluorescence microscopy. However conventional fluorescence microscopy lacks the resolution to resolve adjacent protein layers. Eric Rees and Graham Christie, have pioneered a method of Ellipsoid Localisation Microscopy (ELM) which is able to measure protein layer separation. The principle of the method is to derive a mathematical model for the image of a spherical (or ellipsoidal) fluorescent shell, and fit its parameters to fluorescent micrograph data, enabling shell size to be inferred very precisely. The fitted parameters can also be fed back into the image model to generate a reconstructed image of the spores.
The method itself has wide applications, and is the first example of a set of Fluorescent Shell Localisation methods being developed at CEB. In future work, Fluorescent Shell Localisation will be used to optimise the structure of bacteria strains being developed for therapeutic drug delivery, in a joint project with MedImmune. Nonlinear optimisation techniques, also a key interest at CEB, may also be developed to speed up the image analysis.
Biophysical Journal paper doi: 10.1016/j.bpj.2015.09.023