A procedure for obtaining complementary replicas of ultra-rapidly frozen sandwiched samples.

Complementary replicas of samples prepared for electron microscopy by the freeze-fracture/etch technique are extremely valuable in the interpretation of the exposed surfaces, the nature and location of the membrane fracture plane, and as an aid in the recognition of the potential artefacts of this technique. This paper describes a procedure for the preparation of complementary replicas of thin samples sandwiched between copper foil strips and frozen ultra-rapidly in the absence of chemical pretreatments. In this procedure, the copper foil support bearing the replica is floated on the surface of a chromic acid solution, resulting in the controlled dissolution of the copper metal. The replica which remains at the surface of the chromic acid is then stabilized against fragmentation during subsequent cleaning and rinsing steps by placing a 50 mesh gold grid on top of the replica. To minimize agitation of the replica/grid, all cleaning steps are performed in a single depression plate well. The clean replica/grid is picked up from below on a thin Formvar film, dried, and then separated from the extra film. Careful placement of the gold grid on the replicas and low magnification electron micrograph montages of the complementary grids facilitate the location of complementary regions and simplify examination of complementary specimen areas at higher magnification.

Actin-myosin interactions visualized by the quick-freeze, deep-etch replica technique.

A new method of preparing biological samples for electron microscopy has been used to re-examine the structure of actin filaments, actin filaments decorated by myosin subfragment-1 (S1), and insect flight muscles. Samples were quick-frozen by contact with a block of copper cooled to approximately 4 K; then were freeze-fractured, deep-etched, rotary-replicated with platinum, and viewed in a transmission electron microscope. By this approach, actin filaments display prominent transverse bands whose repeat (approximately 5.5 nm) and pitch (approximately 15 to 20 degrees) fit with the expected left-handed "genetic" helix. Freeze-etched actin filaments do not, however, display the usual two-start helix as prominently as is seen after negative staining, and they also appear substantially thicker than after negative staining (9 to 10 nm versus 8 nm). The latter two-start helix appears very clearly after S1 decoration. Nevertheless, freeze-etched acto-S1 does not display the "arrowheads" that are seen after negative staining. Instead it displays the outer envelope of the helically deployed S1, and as would be expected from current models derived from optical reconstruction of negatively stained samples, this surface view looks only slightly polarized. Finally, the quick-freeze, deep-etch approach provides particularly distinct images of the crossbridges in insect flight muscles. These are plentiful and regularly arranged in rigor muscles, but rare in muscles relaxed with ATP before freezing. In rigor muscles fixed with aldehydes, these crossbridges assume a broad distribution of inclination, ranging from 45 degrees to 90 degrees with a mean of approximately 80 degrees, which is less tilt than has been seen before in thin-sectioned muscles. However, when aldehyde fixation is followed by exposure to tannic acid with or without uranyl acetate block-staining, crossbridges assume a more acute angle with respect to the fiber axis, centering around 45 degrees. This is associated with a commensurate reduction in interfilament spacing within the muscle fibers, such that tilted crossbridges are not any longer than untilted ones (both measuring approximately 15 nm). At the opposite extreme, crossbridges often become stretched in unfixed muscles, owing to an unnatural increase in interfilament spacing that occurs during sample preparation; in such regions, crossbridges display narrow "stalks", which invariably emerge from the thick filaments at close to 90 degrees. We conclude that crossbridge shape and orientation is strongly affected by different methods of sample preparation, and this will make it difficult to visualize natural crossbridge movements by electron microscopy.