STRUCTURE OF NERVE CELL MEMBRANES 157 



Discussion 



Led by P. W. Davies 



Dr. Mullins' idea of a protein-column cell membrane is a most intriguing 

 one, and it fits in particularly well with the helical models of protein molecules 

 that have recently gained prominence. It is also encouraging that there are 

 some who are taking the stereochemical or lock-and-key ideas of molecular 

 interaction seriously enough to work out actual and very subtle correlations, 

 which will be essential in the further development of membrane theory. How- 

 ever, before we abandon the lipid-film theories of membrane structure that 

 have been with us for so long, starting back in the days of the Overton hy- 

 pothesis and continued by such people as Danielli, it might be well to recall 

 some of the evidence that remains in favor of a laminar structure for the cell 

 membrane. 



I might start off by recalling the evidence from physical chemistry about 

 soap molecules, and I wish to say here merely that it has been shown by x-ray 

 studies (McBain, 1950) that in concentrated soap solutions — sodium oleate, 

 for example — the molecules tend to form bilaminar aggregates with the polar 

 heads facing the aqueous medium and the non-polar tails inside. Thus they 

 aggregate in such a way that their nonpolar tails are held together and are 

 separated from the medium by their polar heads. The tails may intertwine in 

 irregular fashion, but are held together strongly enough by van der Waals 

 forces to overcome the attraction of the medium for the polar heads. 



Secondly, there is the x-ray evidence of Schmitt, Bear and Palmer (1941) on 

 myelin which indicates strongly the existence of a certain spacing. The myelin 

 substance, which has water, neurokeratinogens and lipids in it, apparently 

 goes into a laminar structure in which the polar groups of the lipid layers are 

 in aqueous interfaces and attracted weakly to thin protein layers which lie 

 between the lipid laminae. 



A third, more recent, line of evidence in favor of the laminar structure comes 

 from the work of Sjostrand (1953) and others with the electron microscope. 

 By his technique for working with ultra-thin sections, Sjostrand has found that 

 in nuclear membranes and mitochondria there is a tendency to find a three 

 layered structure. In one case he found, in frozen-dried ultra-thin sections, two 

 outer layers 160 A thick and an inner layer between them 210 A thick. Sjos- 

 trand further finds that if he fixes his materials with osmium, which is a good 

 electron stain and also a good protein stain, that the two outer layers take the 

 stain nicely, whereas the inner layer does not. The presumption, according to 

 Sjostrand, is that this central layer contains some lipidlike material. This is 

 direct evidence for a laminar structure of certain kinds of living membrane. 

 A similar structure for what may be the functional axon membrane has been 

 found by Geren (1954) and by Gasser (1955). The axon membranes shown in 



