Rapid freezing techniques and cryoprotection of biomedical specimens.

It has been realized almost from the start of biological electron microscopy that the use of low temperature in specimen preparation might alleviate the problems associated with conventional chemical fixation and plastic embedding. In practice it has been very difficult to realize any benefits, owing to the problems encountered in trying to prepare samples with the water transformed to ice, yet without the sample itself hopelessly distorted by ice crystals. The difficulty is caused by the poor diffusibility of heat through water, the main constituent of almost all biological samples. Even with a maximal temperature gradient between the sample surface and the coolant, and a perfect contact between the two, heat diffusibility through the sample quickly becomes the limiting parameter affecting cooling rate. Unless the cooling rate is as high as 10(2) or even 10(3) degrees C.msec.-1, the ice formed will be crystalline. Such high cooling rates can only be attained in the first few micrometers of even very small samples: therefore, for bulk samples one must either accept the use of a 'suitable' cryoprotectant, or else accept the presence of ice crystals. Theoretical and practical considerations for freezing samples using liquid and solid coolants will be reviewed. Special consideration will be given to the present status of cryoprotection and the indications and contra-indications for its use.