REACTIONS WITH PROTEINS 759 



1954). The disulfide dimer, alb — S — S — alb, however, is dissociated by cys- 

 teine very slowly. The dimerization is accomparied by an increase in levoro- 

 tation, and this implies that the mercaptalbumin molecule undergoes some 

 unfolding in the region of the reactive SH groups as a necessary prelude 

 to dimerization; this may be thought of as a partial denaturation, adding 

 one more item of evidence for configurational changes induced by mercuri- 

 als (Kay and Marsh, 1959). 



Organic mercurials such as p-MB, PM, and MM react in a 1 : 1 ratio with 

 mercaptalbumin and, of course, no dimer is formed. The equilibrium con- 

 stant for the reaction: 



alb— SH + CH3— Hgl ±5 alb— S— Hg— CH3 + H+ + Cl- 



has been found by \V. L. Hughes (1950) to be 3.5 x lO^^ (p^ = 4 45)^ 

 from which the value for the dissociation constant of the mercaptide in 

 Table 7-6 was calculated. On the other hand, bifunctional organic mercuri- 

 als, such as: 



CH-O 



XHg-CH2— HC CH— CH2— HgX 



O— CH, 



can link two mercaptalbumin molecules together (Straessle, 1951; Edsall 

 et al., 1954). The pA" for the equilibrium (dimer )/(monomer+) (alb") is 18.2 

 at pH 4.75 and 25°, the corresponding p^ for the Hg++ dimer equilibrium 

 being 13.5.* This difference of some 4.7 p/C units between the two dimers 

 is undoubtedly due to the fact that the mercaptalbumin molecules must 

 approach about 10 A closer in the Hg++ dimer than in the bifunctional 

 mercurial dimer, and the steric and electrostatic factors could easily ac- 

 count for the some 6.7 kcal/mole difference. 



The importance of these results with mercaptalbumin for inhibition stud- 

 ies with the mercurials is clear. First, the possibility of the dimerization 

 of certain enzymes by Hg++ leads to the concept that inhibition may oc- 

 casionally result through steric sequestration of the active sites and not 

 necessarily through the reaction of SH groups at the active sites. It is also 

 possible that occlusion of active sites could occur by linking a nonenzyme 

 protein to the protein through a Hg bridge. It is known that Hg++ often 

 produces an increase in turbidity of enzyme solutions and even precipita- 

 tion, although this can be due to other factors as well. Second, this reason- 

 ably well understood and quantitatively investigated system provides a 

 model on which the effects of ligands, pH, temperature, and other factors 



* Gurd and Wilcox (1956) give values of 17.2 and 12.6 for the p/C's, respectively, 

 due to different assumptions regarding the ligand constants. 



