the sense that they reinforce and support the 

 interpretations of each other. 



Similar results have been obtained from 

 the purine nucleosides, especially the adenine 

 nucleosides series. In the case of 2'-0 methyl- 

 adenosine, 2'-deoxyadenosine, and 6-methyl- 

 adenosine, the concentration dependence of the 

 chemical shifts is even larger than that of the 

 purine. In all these cases, the H-2 proton of the 

 6 member ring of the adenine is shifted to the 

 higher field than the H-8 proton in the 5 member 

 ring. This indicates that the 6 member ring of 

 the adenine does participate to a greater extent 

 in the stacks than the 5 member ring of the 

 adenine nucleosides. The pentose protons of 

 H-l' are also shifted considerably to higher 

 fields when concentration is increased while 

 the pentose protons of the H-5' are hardly 

 affected. As one proceeds around the pentose 

 ring from the C-1' to the C-5', there is a 

 progressive drop or decrease in the magni- 

 tude of these concentration-dependent chemical 

 shifts. This indicates that adenine nucleoside 

 interaction is preferentially localized at the 

 purine base of the nucleoside so that the ring 

 current magnetic aniosotropy is principally felt 

 by the base protons. From this type of study, 

 therefore, not only can we obtain the general 

 picture about the mode of association, we can 

 even get down to the detailed molecular struc- 

 ture of the stacks. 



Currently, we are also working on the as- 

 sociation of the nucleotides by vapor pressure 

 osmometry as well as by nuclear magnetic 

 resonance. In this case we have relaxed our 

 restriction on the electrostatic effect of the 

 phosphate group and have included this effect 

 as a part of our model system with increasing 

 complexity. Very interesting observations have 

 been made. For instance, preferential inter- 

 actions of the phosphate group with certain base 

 protons of the nucleotides have been observed 

 which have never been suspected before. Asso- 

 ciations of base nucleosides and nucleotides 

 have also been independently studied by 

 Jardetzky (4). 



The experiments detailed above concern 

 solutions containing only one kind of solute. 

 They concern, then, the interactions of the 

 purine or nucleosides with themselves. What 

 are the interactions between different com- 

 pounds, for example, between a purine and a 

 pyrimidine? The increase in solubility of the 

 sparingly soluble adenine and thymine caused 

 by the presence of highly soluble purine and 

 nucleosides was adapted as the method for 



investigation of this type of interaction. As 

 shown in Table IV the solubilities of adenine 

 are much enhanced by the presence of purine. 

 The enhancement is moderate in the presence 

 of cytidine, uridine or pyrimidine and is prac- 

 tically nil in the presence of cyclohexanol, 

 adonitol and urea. Similarly, the solubility of 

 thymine (Table V) is enhanced by the purine 

 and to a less extent by uridine and cytidine. 

 These data were also analyzed by the treatment 

 of multiple equilibria. The assumption in the 

 treatment is that the bases interact to the 

 same extent with the free and the associated 

 forms of the interactants. Equilibrium constants 



TABLE IV 



Solubility of Adenine in the Presence of Interacting 

 Compounds * 



Data from Ts'o, Melvin and Olson, /. Am. Ckem. Soc. 85, 

 1289 (1963); reproduced with permission of the American 

 Chemical Society. 

 '' Standard deviation. 



187 



