The ability to accurately measure bilayer structure is crucial for unraveling membrane structure/function relationships that can lead to a better understanding of biological phenomena and improved therapeutics. Fluid membranes often lack long-range structure within the plane of the bilayer but possess a well-defined layered structure in the normal direction that can be elucidated through the analysis of neutron and/or X-ray scattering data. Although these techniques have proven highly successful when applied to relatively simple model membranes, heterogeneous membranes pose significantly greater challenges due to ensemble-averaging of the scattering signal. Cryo-EM imaging offers an alternative approach for revealing nanoscopic structural features that benefits from the ability to visually isolate and analyze specific membranes or membrane domains within complex samples. However, unlike neutron and X-ray scattering, electron scattering is influenced by the membrane dipole potential and further subjected to microscope focusing optics that are absent from traditional scattering instruments. Here, we present a model for analyzing cryo-EM images which properly accounts for these features. We validate the model using synthetic ground truth image data sets derived from atomistic simulations of lipid bilayers. We then apply the model to experimental data to recover structural parameters including bilayer thickness, area per lipid, and dipole potential.
Heberle et al. (Sun,) studied this question.