Introduction

The idea of preserving biological material is probably as old as civilization itself. It appears in several different con texts in history: in ceremonial ways, such as burials and memorials, as a means to preserve specimens collected on natural history expeditions, and to preserve foodstuffs for later consumption. The techniques worked out over centuries for these uses have been adopted and modified by generations of cytologists working with light and electron microscopy. So, what is considered state-of-the-art today may well have its roots in some earlier methods for some possibly unrelated application. For example, today's bio logical electron microscopists have discovered that ultrarapid freezing is the best way to preserve the native structure of cells and molecules (1,2,3), but the first application of freezing techniques in biological cytology can be traced back to Altmann (4), and the idea of storing food by freezing goes back far beyond that. In this article we are going to discuss how one type of cryotechnique, high-pressure freezing (HPF) is being used in conjunction with electron tomography to reveal unprecedented three-dimensional (3-D) details in cells and tissues.

Sample Preservation for Biological Electron Microscopy

The in vivo study of cells, tissues, and entire animals by noninvasive light and fluorescence microscopy has revolutionized our understanding about their dynamic behavior. Light microscopy is inherently limited, in terms of resolution, to about 200–400 nm. However, a mechanistic understanding of cellular processes requires the structural analysis of individual proteins and their assembly into molecular machines (on the order of tens of nanometers) which cooperate to sustain life.

The transmission electron microscope remains the principal biophysical instrument for the study of the ultrastructure of cells, tissues, and whole organisms, because it allows a direct visualization of the molecules and molecular complexes within cells. However, because samples have to be observed under high vacuum and because the electron beam leads to the destruction of native biological structure, methods were devised to preserve ultrastructure while minimizing the effects of vacuum and radiation. Further more, the electron beam cannot usually penetrate an entire cell, so cells and tissues have to be cut into thin (typically 50–200 nm) sections for observation. Classically, the solution to these problems relies on embedding of the biological specimen into a plastic resin, which can be hardened by heat polymerization and then sectioned very thinly using an ultramicrotome. The different cell components are decorated with heavy metal molecules (usually osmium, lead, and uranium) that resist the effects of the electron beam and also give contrast for image formation (5).

The Problem with Conventional Electron Microscopy Methods

Conventional methods of electron microscopy (EM) specimen preparation for biological samples often involve processing in a series of chemicals that penetrate the sample by diffusion. For very thin samples (less than a few micrometers in the shortest dimension) diffusion is fast enough to give rapid fixation. However, for larger pieces of tissue or for organisms that have natural diffusion barriers, such as plant cell walls, animal exoskeletons, or special layers around embryos, diffusion throughout all cells is relatively slow, up to hours in some cases. When the fixative reaches cells in minutes or hours after immersion, the cell structure is likely to suffer from autolysis and other consequences of a slow death. Even when the fixative penetrates cells in less than a second, the cross-linking reactions are selective (i.e., glutaraldehyde will only react with certain amino acid residues on proteins and will not cross-link nucleic acids or carbohydrate molecules at all). The consequence of this selective fixation shows up in later processing when these unlinked molecules get extracted from the cell during rinse steps and especially during dehydration at room temperature (6). It is also during dehydration that many molecules are moved around from their original positions in the cell and give rise to what are usually called ‘fixation artifacts’ (Figure 1A).

Figure 1. (Click to enlarge)

Why Ultrarapid Freezing?

In contrast to these conventional methods, ultrarapid freezing immobilizes all molecules in a cell within milliseconds. Hence, ultrarapid freezing is highly desirable, because it allows the instantaneous fixation of all molecules in their current position, and, as long as subsequent processing steps do not alter the cellular architecture, we obtain a true snapshot of the cell at the moment of freezing. We prefer to use the term cryo-immobilization to describe this event rather than cryofixation, because there is no fixation taking place in the usual cytological sense of the term. The actual fixation (i.e., formation of chemical cross-links) takes place during a process called freeze substitution. Freeze substitution involves substituting organic solvent for the cellular water at low temperatures, typically −78° to −90°C. The most common solvent is acetone, though methanol is used instead by some investigators. These compounds are liquid at these low temperatures, and over a period of hours to days (depending on the size of the sample), they will dissolve out and replace the frozen water in each cell. Although not always necessary, it is usual to include fixatives such as osmium tetroxide or glutaraldehyde in the acetone to provide actual fixation chemistry. At −90°C there is probably very little chemical cross-linking taking place, but as the temperature is slowly raised (typically 5°–10°C/h), there will come a point when the temperature is high enough for the chemical reactions to proceed. Though there is little experimental evidence addressing the actual temperatures of fixation during freeze substitution, it is generally believed that glutaraldehyde begins fixing around −50°C and osmium tetroxide at about −30°C (7). At these low temperatures, the protein and lipid molecules do not have sufficient thermal energy to move around, and this is why the distortions seen at room temperature processing are avoided (2). By the time movement is possible at higher temperatures, the cells are already fixed.