Metal ions play essential roles in many aspects of biological chemistry, including oxygen transport, enzyme catalysis, and maintenance of biopolymer integrity. Cryo-electron microscopy is sensitive to metal ions because of their strong Coulomb potential relative to surrounding light elements. However, the conventional mode of defocus phase contrast in wide-field cryo-TEM is not well suited for the identification of single metal atoms against the protein and water background. In practice, in order to resolve single metal atoms, atomic resolution of the whole protein complex is needed and at lower resolution even small gold nanoparticles can be difficult to identify.
Annular dark-field STEM imaging provides quantitative contrast based on atomic scattering. We show that this alternative modality, recently applied to cryo-tomography, can detect the presence and location of isolated metal ions in frozen-hydrated protein complexes. Image simulations were used to optimize experimental conditions to provide the highest SNR for the metals compared to protein and water. In experiment, we imaged the 24-subunit ferritin complex bound to stoichiometric amounts of Zn or a limited number of Fe atoms. Cryo-STEM micrographs of metal-loaded ferritin were processed by conventional 3D single-particle alignment and averaging. Even in very small datasets (hundreds of particles), clear density peaks for isolated Zn and Fe atoms were observed. The density peaks localize with angstrom precision to the predicted binding sites of Zn and Fe on the ferritin shell. Our results offer a new understanding and a straightforward technique for atomic detection of metals in macromolecular and cellular contexts, as well as to use of synthetic metal tags as specific molecular labels.
