Some early history of membrane molecular biology.

This article is mostly about the beginnings of the molecular biology of membranes, covering the decade 1964-1974. It is difficult to read (or write) this article because of a sense of deja vu. Most of the material in it is considered commonplace today, having been established experimentally since then. But at the time this work was begun, practically nothing was known about the molecular structure and the mechanisms of the functions of membranes. This situation existed because no membrane proteins of the kind I called integral had as yet been isolated in a pure state, and therefore none had had their amino acid sequence determined. The first integral membrane protein to be so characterized was human erythrocyte glycophorin, in 1978. It was the use of the thermodynamic reasoning that had been developed for the study of water-soluble proteins, together with the information from several key experiments carried out in a number of laboratories during the early decade, that led us to the fluid mosaic model of membrane structure in 1972. Without direct evidence to confirm the model in 1971-1972, my colleagues and I nevertheless had the confidence in it to pursue some of the consequences of the model for a new understanding of many membrane functions, which I present here in some detail. Finally, I discuss two recent high-resolution X-ray crystallographic studies of integral proteins to ask how well the structural and functional proposals that we derived from the fluid mosaic model fit these remarkably detailed X-ray results.