The densities of protein crystals and the hydration of proteins

Abstract

The apparent densities of proteins in aqueous solutions have been studied by a number of investigators, including Chick and Martin (1913), and, more recently, by Svedberg and his colleagues, but comparatively few observations of the densities of protein crystals have been published. Eyer (1932) gave the value 1·315 for insulin, Crowfoot (1935) 1·306; Adair and Adair (1934) found 1·296 for haemoglobin; Bernal and Crowfoot (1934) gave 1·28 for pepsin. The media used for the measurements on pepsin and insulin were not recorded. A method for measuring the densities of small crystals by means of observing their flotation in media of different densities was described by Retgers (1889). The present communication records a modification of his method adapted for the investigation of the density of crystalline and of denatured proteins. The first problem studied was the choice of suitable media. The mixtures of organic liquids which are generally used for density determinations of inorganic crystals are open to objections on the grounds that protein crystals cannot be freed from films of their mother liquor without grave risks of irreversible changes. In working with aqueous media this difficulty need not arise, but the media chosen must fulfil the somewhat exacting requirements of a high density with freedom from the capacity to denature or to dissolve the protein.