422 D. M. BROWN AND A. R. TODD 



further studies using enzymes. Evidence from chemical hydrolysis to mono- 

 nucleotides does not permit any answer to the question whether ribonucleic 

 acids are straight-chain or branched-chain polynucleotides. It does, how- 

 ever, clearly define the types of branching which can be considered. Given 

 that the polynucleotide chain in each branch follows the standard 3',5'- 

 linkage pattern, then two types of branching can be envisaged: (a) branch- 

 ing on phosphorus, i.e., through alkali-labile phosphotriester groupings, 

 and (b) branching at C2' in one of the nucleoside residues in the main chain, 

 attachment being through the usual phosphodiester linkage to C3. (or C2') 

 in the first nucleoside residue of the branch; linkage to Cs- would give an 

 alkali-stable, and therefore inadmissible, structure. The question of chain- 

 branching in ribonucleic acids will be discussed later; it is mentioned at 

 this point because the evidence of chemical hydrolysis clearly defines the 

 only types of branching which need be considered. 



4. Evidence for Nucleoside-5 '-Linkages 



Although evidence has gradually accumulated, mainly as a result of en- 

 zymic studies, that 5'-phosphoester linkages occur in nucleic acids, the 

 complete failure to isolate nucleoside-5 '-phosphates from chemical hydroly- 

 sates for long prevented the general acceptance of their presence. Follow- 

 ing early studies by Takahashi** and Klein and Rossi,^^ Gulland and 

 Jackson^^ examined the action of Russell's viper venom. This venom had, 

 in addition to 5 '-nucleotidase and phosphodiesterase, a weak nonspecific 

 monoesterase activity; however, they showed that when it acted on yeast 

 ribonucleic acid it liberated 25% of the bound phosphorus as inorganic 

 phosphate, and this was increased to 75% by the addition of bone phos- 

 phomonoesterase. These early indications of 5'-phosphoester linkages have 

 been confirmed in recent years by the work of Cohn and his co-workers. 

 Thus Cohn and Volkin^" treated ribonucleic acid with ribonuclease, followed 

 by intestinal phosphatase in the presence of arsenate, to inhibit phospho- 

 monoesterase.^^ Analysis of the hydrolysate by ion-exchange chromato- 

 graphy yielded, in addition to unidentified products, the 5 '-phosphates of 

 adenosine, guanosine, uridine, and cytidine. These were characterized by 

 comparison with the synthetic nucleoside-5 '-phosphates prepared by 

 Michelson and Todd.^'^ Using venom diesterase virtually free from mono- 

 esterase (Hurst, Little, and Butler^^), Cohn andVolkin^* obtained a yield of 



88 H. Takahashi, /. Biochem. Japan 16, 463 (1932). 



89 W. Klein and A. Rossi, Z. physiol. Chem. 231, 104 (1935). 



90 W. E. Cohn and E. Volkin, Nature 167, 483 (1951). 



91 cf. W. Klein, Z. physiol. Chem. 218, 164 (1933). 



92 A. M. Michelson and A. R. Todd, J. Chem. Soc. 1949, 2476. 



93 R. O. Hurst, J. A. Little, and G. C. Butler, J. Biol. Chem. 188, 705 (1951). 



94 W. E. Cohn and E. Volkin, Arch. Biochem. and Biophys. 35, 465 (1952). 



