Correlation of surface topography of metal-shadowed specimens with their negatively stained reconstructions

We present a comparison of surface reconstructions from three different freeze-dried and unidirectionally metal-shadowed specimens (i.e. bacteriophage T4 polyheads, crystalline actin sheets and nuclear pore complexes) with two- or three-dimensional reconstructions of the same specimens when prepared by negative staining. Based on these and many published results, the following conclusions have been reached: With a "cooperative" specimen (e.g. the polyheads), the surface reconstruction computed from a metal-shadowed replica compares favorably with two- or three-dimensional reconstructions obtained from the same specimen after negative staining at the 3-4 nm resolution level. This relatively "poor" level at which the surface topographies of the two preparations can be compared appears to be set by a "practical" resolution limit (i.e. of distinct and reproducible structural detail) of metal replicas of biological specimens, despite the appearance of weak higher-order diffraction spots (i.e. corresponding to 2-3 nm). While in some cases (e.g. the crystalline actin sheets) the surface reliefs of metal replicas may bear little resemblance to the actual structure under investigation, the replicas may still contain sufficient features to establish the polarity or handedness of the structure (i.e., the "top" and "bottom" surfaces of a crystalline sheet). Information from negatively stained specimens is usually complementary with information from freeze-dried and metal-shadowed specimens. However, there are artifacts in both techniques, and we present an example with the nuclear pore complex, where these techniques yield confusing results.