Technical Highlight - September 2011
Short description: A new approach promises to eliminate a major hurdle to structure determination of membrane proteins using electron crystallography.
Generating atomic-resolution structures of membrane proteins is a notoriously difficult process, and electron crystallography has been one of the most successful approaches used in this field since its initial application by Henderson and Unwin to bacteriorhodopsin in 1975. By using two-dimensional crystals of the target membrane protein embedded in a lipid bilayer, researchers can generate both image and diffraction data that can then be combined to produce a high-resolution protein structure.
However, as useful as it has been, electron crystallography is not without pitfalls. One of the biggest hurdles to high-resolution structure determination using this technique is the image-acquisition step. Although it is comparatively easy to generate low-resolution (6 Å) image data from two-dimensional crystals, technical limitations make acquisition of higher-resolution image data challenging. Unfortunately, the electron diffraction data cannot provide the necessary phasing information that is needed for high-resolution structures, so researchers have to either use molecular replacement or tackle the imaging challenges head-on.
Now, Wisedchaisri and Gonen (PSI TEMIMPS) provide an alternative approach for membrane proteins based on phase extension of low-resolution image data combined with high-resolution diffraction data, which are both relatively easy to obtain from two-dimensional crystals. In their approach, α-helix fragments are placed into the phases from the image data, refined with the diffraction data and then used to generate probability-based high-resolution phases that are combined with the low-resolution image data. These phases are then subjected to successive iterations of density modification, fragment expansion and structure refinement until a high-resolution structure model is produced. The authors demonstrate the validity of their approach by generating atomic-level structures of aquaporin-0, aquaporin-4 and bacteriorhodopsin that compared favorably to structures for those proteins determined using a traditional electron crystallography approach. Each structure achieved a resolution of 3.2 Å or lower, with R-factors and B-factors well within acceptable ranges.
To enhance the general accessibility of their new approach, the authors used only publicly available software. Thus, in addition to saving time and labor by bypassing the need for high-resolution image data, this phase-extension technique carries no additional costs to users and will undoubtedly be of great benefit to researchers studying membrane proteins in various systems.
G. Wisedchaisri & T. Gonen Fragment-based phase extension for three-dimensional structure determination of membrane proteins by electron crystallography.
Structure 19, 976-987 (2011). doi:10.1016/j.str.2011.04.008