What history tells us XXX. The emergence of the fluid mosaic model of membranes

Abstract

In 1968, Walther Stoeckenius summarized in Science the conclusions of a meeting held the previous year at Frascati near Rome on ‘membrane modelling and membrane formation’ (Stoeckenius 1968). The diversity of the issues that were raised, and of the methods and models that were used, was such that no consensus could be reached. At that time, even the existence of a cell membrane was not unanimously accepted! Three years later, S Jonathan Singer (1971) proposed the mosaic model of the structure of cell membranes to replace the previously dominant Danielli–Robertson unit membrane model, and one year later he and Garth L Nicolson added the word ‘fluid’ to ‘mosaic’ (Singer and Nicolson 1972). The new model was rapidly and unanimously accepted, and it remained unaltered during the next forty years. This rapid evolution of models of the membrane structure has not been explored by historians. I will show that the emergence of the fluid mosaic model was not as smooth as indicated in the description given by one of its authors (Singer 2004). What happened in the mid-1960s was an unsuccessful attempt to replace the lipid bilayer model, also more precisely named the ‘bimolecular lipid leaflet’ model, by a new model giving proteins a major role in the organization of membranes. The unexpected resistance of the lipid bilayer model, as well as new data demonstrating the rapid displacement of proteins within the membrane plane, led to a progressively elaborated synthesis that was brilliantly outlined by Singer and Nicolson in their famous Science article. 2. The challenges to the lipid bilayer model of membrane structure in the 1960s