Cryofixation of thick (500 μm) biological specimens by high-pressure freezing

Abstract

Cryofixation-based preparation techniques are capable of portraying the biological ultrastructure more closely related to the living state than conventional procedures employing chemical fixation and dehydration (for review: Sitte 1987). Currently used rapid freezing procedures e.g. spray freezing, propane-jet freezing, plunge freezing, slam freezing, yield adequately frozen specimens with no visible ice crystal induced segregation patterns in freeze-fractured or freeze-substituted samples (for review: Steinbrecht and Müller 1987) with high reproducibility, provided that they are properly applied to thin samples (approx. 10μm), e.g. suspensions of cells, microorganisms, organelles. They are however of limited use for the cryoimmobilization of thicker samples e.g. animal or plant tissues. Adequate structural information, in a thin superficial zone at the natural or cut surface of tissue samples, is sometimes obtained by slam-freezing. The thickness of this zone, in which no segregation patterns can be observed, depends on the concentration of cellular components that exhibit cryoprotective properties and may often reach approx. 20 μm. This depth however, is generally insufficient to analyse tissue cells that have not suffered from traumatized excision or when studying more complex systems e.g. fungus/host interactions, root nodules. Thicker systems can be studied by cryofixation-based electron microscopy only if the physical properties of the cellular water are influenced in a way that adequate cryoimmobilization is achieved with much slower cooling rates. This is accomplished by freezing the samples under high hydrostatic pressure. High pressure freezing is at present the only known practical way of cryofixing larger samples (200 - 500 μm). Its development was initiated approx. 20 years ago by Moor and co-workers (for review see Moor 1987). Adequate instrumentation became commercially available only recently. The commercial high pressure freezer works well with respect to the physical performance and reliability. It provides high cooling rates at the surface of the sample reaching 2500 bar within approx. 20 ms with precise coordination of the rise in pressure with the drop in temperature. Despite the high instrumental reliability, the yield in adequately cryofixed biological samples was only marginal. Major problems seem to arise from the way pressure and cold are transferred to the sample. The yield in well cryofixed specimens could be slightly improved (10 - 30 %) if the sample exactly fitted the cavities of the metal specimen supports between which it was sandwiched for high pressure freezing (Müller and Moor 1984). A high yield in adequately cryofixed samples, however, is of primary importance if one wishes to correlate structure and function in practice.An 80 % yield, in well frozen samples (plant and animal tissues, suspensions), was achieved by immersion of the excised tissue blocks into 1-Hexadecene prior to high pressure freezing. 1-Hexadecene is insoluble in water, osmotically inactive and replaces the free water surrounding the tissue blocks or cells. It facilitates the transfer of pressure and cold to the specimen. In addition, it may reduce the danger of ice crystal nucleation outside the specimen. In contrast to the established rapid freezing techniques relatively slow cooling rates (approx. 500 Ksec−1 are achieved in the center of high pressure frozen samples. These might be too slow e.g. to catch dynamic events at membranes or to prevent structural alterations due to lipid phase transition and seggregation phenomena. Little is known about the effects of the high pressure, which lasts for about 15-20 msec on the sample before freezing (Müller and Moor 1984). The achieved high yield in well frozen samples by the 1-Hexadecene treatment allows us now to look carefully at the above questions and to judge the relative merits of high pressure freezing. The morphology of slam-frozen and high pressure-frozen biological specimens appears identical after freeze substitution. Differences are expected to occur at the level of the preservation of the spacial distribution of diffusible ions as well as the conformation of macromolecules.