W. KAUZMANN 



15 



used in studies of denaturation at the present time.- It is therefore worth 

 devoting some space to a discussion of the structural significance of 

 changes in these two properties. What can we infer if the viscosity and 

 the optical rotation of a protein solution change their values under the 

 influence of some denaturing agent? 



INTERPRETATION OF CHANGES IN SOLUTION VISCOSITY 

 DURING PROTEIN DENATURATION 



It has been usual in the i)ast, and it continues to be the fashion, to in- 

 terpret the viscosity increment of protein solutions in terms of models 

 that are rigid ellipsoids (see, for instance, Scheraga and Mandelkern, 30). 

 This seems to be a reasonable procedure when it is applied to a protein in 

 a relatively compact folded state, such as that assumed by the native 

 protein, and also by the denatured protein in a poor solvent. On the other 

 hand, if the denatured protein is dissolved in a good solvent which inter- 

 acts strongly with a large fraction of the groups that occur along the 

 polypeptide chain, then very viscous solutions are obtained, indicating 

 that the molecule has unfolded to a much more extended form. (This is 

 probably the case for strong urea solutions of most proteins, for instance.) 



Fig. 1 



In such solvents a more reasonable model than the rigid ellipsoid is pro- 

 vided by the randomly coiled chain that is so familiar to the polymer 

 chemist (chap. 10 of Flory, 6). If the secondary forces (van der Waals 

 forces, hydrogen bonds, etc.) that operate between different segments of 

 the polymer chain are small compared with the thermal energy, so that 



-The measurement of the effect of a protein on the viscosity of a sohition at 

 elevated temperatures is not too easy to accomplish because of the need for fairly 

 good temperature control. This need can be avoided by using reasonably well stirred 

 baths containing two viscosimeters, one containing pure solvent and the other con- 

 taining the protein solution; outflow times are measured concurrently on the two 

 viscosimeters. 



