THE CELL MEMBKANE AS A SITE FOR MERCURIAL ACTION 905 



membrane to cause lysis and yet have no inhibitory effect on glucose uptake. 

 One might conclude that Hg++ penetrates into the erythrocytes more readi- 

 ly than 2?-MB, which is undoubtedly the case, but it is also possible to spec- 

 ulate that the bifunctional Hg++ can distort the membrane pores by form- 

 ing S — Hg — S bridges in such a way that glucose penetration is slowed 

 while K+ permeability is increased, as in the concept of critical pore sizes 

 formulated by Mullins (1960). Another problem is how these results can 

 be reconciled with those of LeFevre (1948), who showed that glucose utiliz- 

 ation by human erythrocytes is inhibited by 0.002 mM p-MB and abolished 

 by 0.01 mM. 



The very rapid loss of K+ from human erythrocytes observed by Weed 

 et al. (1962) — maximally 50% of the total cell K+ in 3 min at 7.5 X 10-i« 

 mole Hg/cell — is not seen in rabbit erythrocytes, from which there is a 

 slow loss of K+ in the presence of Hg++, a rapid loss occurring only upon 

 hemolysis, a result which is quite reasonable (Joyce et al., 1954). It is dif- 

 ficult to compare the results of these two groups of investigators because 

 the Hg++ concentrations are expressed differently. However, it is possible 

 to estimate that when 7.5 X 10~^^ mole Hg/cell is bound, the free Hg++ 

 concentration is roughly 0.01 mM (see Fig. 2 of Weed et al.). In the work 

 of Joyce et al., 0.032 mM Hg++ caused a 30% loss of total K in 4 hr, so 

 that apparently there is a very marked difference in the response of rabbit 

 and human erythrocytes to Hg++. 



The question as to the relation of glucose metabolism to hemolysis is 

 still unanswered. There is one observation which suggests a relation, the 

 finding by Moore (1959) that 10-100 roM glucose inhibits the hemolysis 

 induced by p-MB, this being manifest mainly in a lengthening of the lag 

 period. No reaction between j^-MB and glucose can be detected spectrosco- 

 pically and it is not an osmotic effect. Certain other sugars, e.g. fructose 

 and sorbose, are also effective. It was postulated that glucose may combine 

 with some component of the rat erythrocyte and protect it from the mer- 

 curial; if so, this would probably be the transport system for glucose, which 

 is inhibited readily by p-MB, the situation being similar to the protection 

 of enzymes by substrates. Sheets et al. (1956 a) had found that glucose exerts 

 no protection against hemolysis of human erythrocytes by p-MB, but the 

 glucose was added at various times after the p-MB and was only 3.3 mM. 



Membrane or transport ATPase is inhibited by the mercurials but the 

 Hg++ or chlormerodrin which is initially bound is without effect (Roth- 

 stein, 1964). Chlormerodrin can bind to about 3% of the total SH groups 

 without inhibition of ATPase, but by the time of maximal binding with 

 25% of the SH groups the ATPase is inactivated, possibly leading to the 

 loss of K+, although some increase in permeability may also play a role. 



The possibility that mercurial hemolysis is related to the reaction of 

 erythrocytic glutathione with SH reagents was considered by Tsen and 



