BARBARA W. LOW AND JOHN T. EDSALL 



possible that other bonds, as yet unidentified, are more important 

 in maintaining the molecule in the native configuration, and 

 that the breakage of the links involving the tyrosine hydroxyls 

 may be a secondary result of the breakage of these other linkages. 

 However, it seems likely that the bonds involving the tyrosine 

 hydroxyls are important in maintaining the native structure. 

 In any case we know that there are other proteins in which this 

 is not true ; Crammer and Neuberger found the tyrosine groups 

 in insulin to react as if they were part of a free tyrosine peptide, 

 Tanford and Roberts (103) have found an intermediate state 

 of affairs in serum albumin; the tyrosine hydroxyls appear to 

 be involved in some sort of hydrogen bonding, as judged from 

 their free energies and entropies of ionization, but the breakage 

 of the bonds is readily reversible on lowering the pH, and in- 

 volves no such drastic alteration of the native structure of the 

 protein as in ovalbumin. 



Laskowski and Scheraga (60) have set forth a detailed 

 analysis of the formation and breakage of various types of 

 hydrogen-bonded structures which may occur in proteins, and 

 have pointed out that some of the observed peculiarities of protein 

 titration curves might be readily explained in terms of such 

 bonding. It is difficult as yet to obtain proof or disproof of the 

 validity of their explanation, for any actual protein, but they have 

 pointed out important new possibilities, which deserve careful 

 examination. In any case, it appears that many side-chain 

 ionic and polar groups in many proteins — perhaps most such 

 groups in most proteins — are not tied down but are reactive 

 like similar groups in smaller molecules (52). 



The cohesion between nonpolar side chains is probably 

 often important as a stabilizing factor. As Kauzmann (53) has 

 pointed out, these side chains tend to hold together with con- 

 siderable force. The energy required to separate them is 

 decidedly greater than would be calculated from van der Waals 

 attraction alone, for the separation of two nonpolar side chains 

 involves bringing each of them in contact with additional water 

 molecules, and therefore requires the breaking of some hydrogen 



392 



