Structural analysis of herpes simplex virus by optical super-resolution imaging

Laine, Romain F.; Albecka, Anna; van de Linde, Sebastian; Rees, Eric J.; Crump, Colin M.; Kaminski, Clemens F.

Images of individual virus particles (as those shown in Figs 2a and 3) were typically reconstructed from 100 to 300 localizations, with a typical localization precision of 5-10 nm (1 s.d. as described by Mortensen et al. 38 , see Supplementary Note 3). By aligning the obtained particles, the localizations can be combined to reconstruct a high-density super-resolution image of a specific protein layer. Furthermore, with the large number of localizations obtained in the aligned data set, the radial distribution of the aligned localizations can be accurately retrieved and analysed to determine the position of this individual layer. For this, the radial distribution of localizations was fitted to a Monte-Carlo model of virus localization data set (Monte-Carlo-based virus (MCV) model, presented in Fig. 2b and Supplementary Note 2), developed assuming spherically symmetric virus particles. In this model, each viral protein is assumed to lie within a spherical shell described by two parameters: the shell diameter and the shell thickness. The model takes both the linker size and the localization error into account. The shell diameter and thickness are the structural parameters of interest and are both obtained by fitting, whereas the linker size and localization precision are first estimated and then fixed during the analysis (Supplementary Note 3). The shell diameter represents the average radial position of the protein of interest in a population of virus particles. On the other hand, the shell thickness obtained by our method is a result of several contributions: the actual thickness of the protein layer (the thickness of the shell shown in Fig. 2b ), the variability of the diameter from particle to particle in a population of virus particles and any deviation from the spherical symmetry.