Electron Micrographs of Myelin Figures of Phospholipide Simulating Intracellular Membranes

In an earlier study (3) it was observed that if cells isolated from guinea pig testis were allowed to stand in Tyrode solution, the endoplasmic reticulum gave rise to concentric systems of membranes (Fig. 1) strongly reminiscent of the myelin figures studied in great detail by Nageotte (6). These findings suggested that an electron microscope study of phospholipide films formed in vitro might shed some light on the behavior of cytoplasmic membranes observed in vivo, and might also provide information of value in interpreting the chemical nature of the membrane components responsible for the dense lines seen in electron micrographs of cells. In thin sections of osmiumfixed tissue, the individual membranes (~90 A) exhibit a characteristic fine structure consisting of two dense layers (~30 A) separated by a region of lower density (~30 A). Owing to our poor understanding of the mechanism of fixation and staining with osmium tetroxide, this doublecontoured appearance has been variously interpreted. Some investigators, adhering to the traditional view that osmium reacts primarily with the double bonds of unsaturated fatty acids, have considered the dense components of the membranes to be lipide. Other workers, taking cognizance of recent studies (I) showing that osmium also reacts with certain amino acids, have interpreted the two dark lines as representing mainly protein, while the light interspace is believed to consist of a bimolecular leaflet of lipide (8-10). The lack of staining of the latter with osmium has usually been attributed to a high degree of saturation of the hydrocarbon chains or to extraction of lipide in the course of specimen preparation. The model systems of membranes described in this preliminary report were produced in vitro by hydrating phospholipides which, although inhomogeneous in lipide content, were essentially