Gold specimen supports improve cryo-EM image quality. Credit: Christopher J. Russo, MRC Laboratory of Molecular Biology

The field of single-particle cryo-electron microscopy (cryo-EM) has made amazing technical strides in the past few years. In particular, the recent development of highly sensitive direct electron detectors are allowing cryo-EM to reach what was previously the realm of crystallography: atomic-resolution structure determination, but without the requirement of protein crystallization.

But even researchers using direct electron detectors still encounter the pesky problem of the specimen support moving in the electron beam. In a typical cryo-EM experiment, protein particles are frozen in ice suspended across tiny holes in an amorphous carbon film supported by a metal grid. For reasons that are not fully understood, but likely because of certain instabilities in the carbon film, the specimen support undergoes deformation when irradiated by the electron beam. This causes the captured images to blur and limits researchers' ability to align the particle images to achieve high-resolution structure determination.

Lori Passmore and her postdoctoral fellow Christopher Russo of the Medical Research Council (MRC) Laboratory of Molecular Biology have been interested in addressing cryo-EM specimen motion at the root of the problem: the support itself. In the June 2014 issue of Nature Methods, they reported that a hydrogen plasma–treated graphene support was subject to substantially less electron beam–induced motion than the conventional amorphous carbon support. Moreover, they found that a continuous layer of graphene allowed much more optimal control of protein distribution on the support surface than a continuous layer of amorphous carbon.

Now, reporting in Science, Passmore and Russo show that gold is even better than graphene at reducing specimen motion, to the point that such movement is nearly eradicated. Compared to a standard amorphous carbon support, their use of gold reduced motion of the specimen by 50-fold, which translates to a twofold reduction in image blurring. The gold support allowed them to solve the structure of the iron-storage protein apoferritin, a small, spherical protein that has presented challenges for cryo-EM structure determination, at 4.7-angstrom resolution.

The gold support consists of an array of micrometer-sized holes in a thin gold foil, which is suspended across a mesh grid also made of gold. “Chris chose gold because it is compatible with biological specimens, it is conductive, stable (non-oxidizing) and radiation hard,” says Passmore. Although gold supports should be relatively straightforward for many labs to manufacture themselves, Passmore and Russo are also working with a company to commercialize them. Passmore further points out that gold supports have the potential to improve images from any electron microscope, not only those equipped with the latest-generation detectors.

Passmore and Russo are interested to next explore whether combining the advantages of a continuous graphene layer with the gold support will even further improve protein structure determination by cryo-EM. “Graphene addresses the problem of providing a reproducible, invisible and tunable surface for the adsorption of proteins. Gold supports virtually eliminate the motion of the specimen support during irradiation,” Russo explains. “Taken together, these address two major problems in cryo-EM.”