Electron cryotomography (cryoET) is capable of yielding 3D reconstructions of cells and large-macromolecular machines. It does not depend on fixing, staining, or embedding, so the contrast is related to the mass density of the specimen. The 3D reconstruction itself does not require that the specimen consist of identical, conformationally homogeneous units in random orientations, as is the ideal case for single-particle reconstruction from 2D images. However, if the specimen contains multiple copies of a macromolecular assembly, these copies can be extracted as 3D subvolumes from the tomographic reconstruction, aligned to each other, and averaged to achieve higher signal-to-noise (S/N) ratios and higher resolution. If conformational variability is present, it is more straightforward to separate the conformational heterogeneity from the orientation of the particles using the 3D information from the subvolumes than it is for single-particle reconstructions. This chapter covers the techniques of detecting, classifying, aligning, and averaging subvolumes (subtomograms) extracted from cryoET reconstructions. It considers methods for dealing with the unique problems encountered in tomographic analysis, such as the absence of data in the "missing wedge," and the overall extremely low S/N ratio inherent in cryoET. It also reviews applications of the inverse problem, that of orienting a template back into a tomogram, to determine the position of a molecule in the context of a whole cell.
Copyright © 2011 Elsevier Inc. All rights reserved.