When we determine a protein’s structure by single-particle cryo-EM, we use image processing tools to combine information from many individual copies of the same protein complex, generating a three-dimensional map of the complex, which we use to build an atomic model. Although these reconstructions can now reach high resolutions, allowing us to confidently build many parts of complexes including the polypeptide and polynucleotide scaffold, we often have difficulty unambiguously identifying other densities in our maps including bound metals and other ions, lipids, substrates and inhibitors. Because these species often play out-sized roles in the activity and regulation of complexes, the uncertainty in assigning these densities has large consequences for the accuracy of the atomic models, and of future work carried out based on these models. Therefore, we see the need for a technique that would allow us to map elements in the 3D particle space. Several techniques already exist for elemental mapping in the electron microscope, but these require much higher doses than our biological samples will tolerate. We are working to combine techniques from analytical electron microscopy with the many ’tricks’ of biological cryo-EM: low-dose acquisition, cryogenic temperatures, automated data acquisition, and image processing techniques that allow a large total dose to be spread across many identical copies of a given object.
This project is carried out in collaboration with CEOS GmbH, and Holger Stark and Dietmar Riedel of the Max Planck Institute for Multidisciplinary Sciences.