Procedure for freeze-drying molecules adsorbed to mica flakes.

The quick-freeze, deep-etch, rotary-replication technique is useful for visualizing cells and cell fractions but does not work with suspensions of macromolecules. These inevitably clump or collapse during deep-etching, presumably due to surface tension forces that develop during their transfer from ice to vacuum. Previous protocols have attempted to overcome such forces by attaching macromolecules to freshly cleaved mica before drying and replication. I describe here an adaptation of this procedure to the deep-etch technique as otherwise practiced. My innovation is to mix the molecules with an aqueous suspension of tiny flakes of mica and then to quick-freeze and freeze-fracture the suspension exactly as if one were dealing with cells. The fracture inevitably strikes the surfaces of many mica flakes and thereby cleaves the adsorbed macromolecules cleanly enough to reveal interesting substructure within them. The subsequent step of deep-etching exposes large expanses of unfractured mica and thus reveals intact macromolecules. These macromolecules are not obscured by salt deposits, even if they were frozen in hypertonic solutions, apparently because the fracturing step removes nearly all of the overlying electrolyte. Moreover, these macromolecules are minimally freeze-dried (since exposure is sufficient after only 3 min of etching at -102 degrees C) so they retain their three-dimensional topology. I show that molluscan hemocyanin is a good internal standard for this new technique. It is available commercially in stable solutions, mixes well with all sizes of macromolecules, and consists of particles that display distinct five-start surface helices, which have been measured carefully in the past and which possess a known handedness, useful for determining the orientation of micrographs when examining the various helical patterns possessed by most types of extended macromolecules. The fractured hemocyanin particles also display characteristic internal structures, which permit determination of the elevation of the "molecular cleavage" described above. Finally, molluscan hemocyanin is delicate enough to reflect bad freezing or poor replication, if these steps become a problem. A survey of several macromolecules is presented, including soluble enzymes, antibodies, filamentous proteins and nucleoproteins. These images, for the most part, correspond to those previously obtained by negative staining. New details of their structures are noted, and the images are used to illustrate both the advantages and drawbacks of the new procedure.