480 D. O. JORDAN 



the disappearance of streaming birefringence.^"^ ■""•^^ This decrease in the 

 viscosity on the addition of acid and alkaU was considered by Vilbrandt 

 and Tennent"^ to be caused by depolymerization, which slowly reversed 

 when the solution was returned to neutrality. Gulland et al.,'^^ however, 

 have shown that this depolymerization cannot involve the rupture of the 

 internucleotide phosphoester linkages since there is no increase in the 

 amount of secondary phosphoric acid dissociation on back-titration, and 

 these authors suggested that some of the amino groups of the pyrimidines 

 and purines are linked by hydrogen bonds to the — NH — CO — dissocia- 

 tions; the ionization of either group would then bring about the liberation 

 of both as shown by the position of the back-titration curve (Figs. 15 and 

 16). The liberation of the groups at pH values more alkaline or acid than 

 would be expected from the normal pK'a values of the — NH^ — ^CO — and 

 amino groups, respectively, will be due, most probably, to the stability of 

 the nucleic acid hydrogen-bonded structure and the need to break several 

 of the hydrogen bonds simultaneously. The ionization would thus be 

 shifted in the direction of higher hydrogen and hydroxyl ion concentrations 

 as is shown by the titration curve. 



From the titration data alone it is not possible to decide whether the 

 bonds are intra- or intermolecular. The marked decrease in viscosity which 

 has been observed by Creeth et al,^^'^ at almost the same pH values as the 

 ionization of the groups would serve to indicate that the bonds join smaller 

 molecular units, which in solution show lower viscosity than the hydrogen- 

 bonded macromolecules. The sedimentation data,^" although not conclusive 

 on this point, tend to confirm this conclusion. 



These conclusions, drawn from titration results, are in remarkable agree- 

 ment with the structure proposed by Watson and Crick. ^^ The hydrogen 

 bonds between the single nucleic acid chains involve the amino groups of 

 adenine and cytosine and the — NH — CO— groups of guanine and thymine. 

 Furthermore, more careful examination of the titration curve^^ shows that 

 the anomalous behavior ceases at approximately pH 3.5 after the titration 

 of 1.8-2.0 equivalents of amino group, i.e., the back- and forvvard-titration 

 curves are coincident at pH values more acid than 3.5, which indicates that 

 the most acid amino group, viz., that of guanine, does not partake in the 

 formation of hydrogen bonds as suggested by Watson and Crick. 



b. Pentose Nucleic Acids 



The titration of pentose nucleic acids has yielded important information 

 concerning the nature of the ionizable groups and of the internucleotide 

 linkage. No information concerning the macromolecular structure has been 

 obtained, however, since the samples studied have all been isolated by use 



111 C. F. Vilbrandt and H. G. Tennent, J. Am. Chem. Soc. 63, 1806 (1943). 



