The Formation of Natural Membranes 441 



branes, and second, how such membranes may couple the organism with certain 

 environmental components while isolating it from others, 



Pringle [13] has discussed the different hypotheses that have been advanced 

 to explain the appearance of heterogeneity which must have been a precondition 

 for the independent development of Hving organisms. Of these, only the approach 

 adopted by Oparin [14] seems to be consistent with the general concepts of 

 molecular architecture developed in the paper. Oparin pointed out that ' ... as 

 soon as organic substance became spatially concentrated into coacervate droplets 

 or bits of semi-liquid colloidal gels ; as soon as these droplets became separated 

 from the surrounding medium by a more or less definite border, they at once 

 acquired a certain degree of individuahty'. The work on coacervate systems, 

 largely carried out by Bungenberg de Jong and his collaborators [15], shows that 

 the separation of coacervates is due to an association between substances (usually 

 of large molecular weight) which carry opposite ionic charges. This association 

 is not as regular or as close packed as in a crystal, but it occurs spontaneously 

 because the approximation of dipoles or of charges of opposite sign (and in some 

 cases also the packing together of hydrophobic groups) results in a lowering of 

 the free energy of the system. As Oparin pointed out, the coacervates may 

 separate as droplets in equihbrium with very low bulk concentrations of their 

 components. The point wliich is of particular interest to us here is that coacer- 

 vates containing lipids are covered by membranes which are probably composed 

 of two monolayers, and that droplets of medium within a coacervate drop may 

 also be covered and prevented from coalescing by similar Upid membranes [15]. 

 These spontaneously formed membranes resemble the Hpid plasma-membranes 

 which cover the protoplasm of all the present-day organisms that have been studied 

 [16, 17], and it is therefore reasonable to suggest that both the natural and the 

 coacervate membranes may owe their formation and stabiHty to the same simple 

 laws of approximation of charges and separate close packing of hydrophobic 

 and hydrophilic parts. 



Five years ago, at a symposium on active transport and secretion, I pointed 

 out [18] that 'the view expressed nicely by Rosenberg [19] that by virtue of 

 specific permeability properties, the natural membranes act as connecting links 

 between particular components of the phases which they separate has its counter- 

 part in the view of the enzymes as couplers of reactions which can proceed only 

 on or in the enzyme molecules. Rosenberg's treatment shows, in fact, that the 

 energetics of the reactions in two phases connected by a membrane can be 

 described in the same terms as 'homogeneous' enzyme-linked reactions; the 

 important impHcation being that the efficiency (or reversibility) of transport 

 reactions is determined by the specificity of membrane permeabihty, exactly as 

 the efficiency of coupled enzyme reactions is determined by the enzyme-substrate 

 and enzyme-carrier specificities. In complex biochemical systems such as those 

 carrying out oxidative phosphorylation [20], the osmotic and enzymic specifi- 

 cities appear to be equally important and may be practically synonymous'. Recent 

 work [21-23] has provided additional support for the view that osmotic and 

 enzymic coupling may be regarded as analogous mechanisms which depend 

 upon the properties of specific proteins in present-day living organisms. Further, 



