150 ELECTROLYTES IN BIOLOGICAL SYSTEMS 



markedly as the concentration of KCl and KNO3 in the two solutions was 

 raised equally. The plot of exchange rate as a function of concentration was 

 concave upward, but approximated linearity over the concentration range 

 from 10 to 100 mM/1. Dr. M. Gottlieb has informed me that his recent meas- 

 urements of K-NH4 exchange across an anion permeable membrane yields 

 similar results. The situation is not precisely analagous to our previous ap- 

 praisal of the effect of varying external K concentration on the diffusion of K 

 into red cells. In the case of the ion exchanger membranes, changes in the 

 critical ion concentration certainly altered the electrical potential distribution 

 at the two membrane-solution interfaces. In the red cell experiments, the con- 

 centration of chloride (critical ion) was held constant. Nevertheless, these 

 results on charged membranes show that diffusion of the non-critical ion does 

 not show saturation kinetics. They are in general agreement with our theo- 

 retical treatment of cation diffusion in the red cell membrane. 



We now turn to a brief consideration of model membranes in which selective 

 transport of ions of identical charge and similar chemical properties (e.g. Na 

 and K) might occur. Such models may be somewhat arbitrarily classified into 

 two groups: 7) those in which the selectivity is accomplished by the greater 

 affinity of a membrane reactant for either Na or K; and 2) those in which the 

 selectivity depends on differences in the behavior of the membrane toward 

 complexes between an organic compound and Na or K which are equally 

 stable. In order for models of either type to perform the desired transport 

 work it is necessary that a source and sink of the carrier molecule be present 

 and appropriately located. For example, the source and sink might be the cell 

 interior and the external medium, or the inner face and outer face of the 

 membrane itself. 



Considering first models of the first type, Bregman (3) has recently dis- 

 cussed cation selectivity of cation exchangers. From the theoretical and ex- 

 perimental findings there reviewed, we may expect strong acid groups (SO:r, 

 H2P04~) to prefer K to Na whereas the converse will obtain for weak acid 

 groups (HCOs", the second dissociable proton of H3PO4). Quite a few organic 

 compounds, some of biochemical interest, have been shown to have selective 

 affinity for K or Na (89). 



Models of the second type may accomplish selectivity by differential solu- 

 bility of the complex of K or Na in the membrane phase. The guaiacol models 

 of Osterhout (71, 73) depend on this property. Also of interest in this regard 

 are the experiments of Rosenberg with the salts of oxaloacetate and aceto- 

 acetate (89). Unfortunately, diethyloxaloacetate was found to have no effect 

 on cation transport in red cells. In addition, models of the second type can 

 gain selectivity from differences in the reactivity of the complexes of the ions. 

 Melchior (61) has recently proposed such a model involving the K and Na 

 complexes of adenosine triphosphate (ATP). Several enzymes which catalyze 



