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



nucleotides, i.e., the tendency of phosphates of /3-aldehydo alcohols to un- 

 dergo ehmination reactions (cf. p. 437). The initial action of acid must lead 

 to products of the apurinic acid type. If we consider a section (XXXV; 

 R = next residue in the chain) of such a molecule, ready elimination should 

 occur at Cs- of the deoxyribose residue (here written in the aldehyde form) 

 as indicated by the broken line a a. 



CHO Py-CH- 



T o 





CH r°\ OH CH- 



"^0^ CHOH " p1 / CH— I 



^CH2 ,' CH2 







XXXV 



The alternative fission i.e., at the point indicated by the broken line h 

 h in XXXV, which would result from intervention of a cyclization be- 

 tween C3- and C4' in the deoxyribose residue, analogous to the mechanism 

 involved in ribonucleic acid hydrolysis, ^^ is less likely; if it occurred, one 

 would expect nucleosides or nucleotidic materials deficient in phosphorus 

 to be produced. If the next residue (R) ni the formula (XXXV) represents 

 a pyrimidine nucleoside residue, a product larger than a mononucleotide 

 will be first formed, but in the event of R being another deoxyribose residue 

 (i.e., the original site of a purine nucleoside residue) production of a pyrimi- 

 dine nucleoside diphosphate might be expected. It is clear that further de- 

 tailed investigation of the products of acid hydrolysis is desirable. 



Tamm et a/.'^^ have found that apurinic acids are rapidly degraded in 

 alkaline solution to give a mixture of dialyzable and nondialyzable prod- 

 ucts. The nondialyzable fraction contains ca. 85 % of the pyrimidine nucleo- 

 tide residues but only some 40% of the deoxyribose phosphate residues 

 present in the starting material. They interpret this degradation by invok- 

 ing the cyclization mechanism mentioned above in discussing the break- 

 down of XXXV, and from their findings they have deduced that calf thy- 

 mus deoxyribonucleic acid consists of a chain in which sections of the chain 

 containing principally pyrimidine nucleotides are followed by stretches in 

 which purine nucleotides predominate. There appears to be some discrep- 

 ancy between this conclusion and the isolation of large amounts of pyrimi- 

 dine nucleoside diphosphates from acid hydrolysates if these latter sub- 

 stances arise by the mechanism discussed above. The appearance of the 

 diphosphates suggests a fairly even or random distribution of purine and 

 pyrimidine nucleotide residues. A likelier mechanism for the alkaline degra- 



1^3 C. Tamm, H. S. Shapiro, R. Lipshitz, and E. Chargaff, J. Biol. Chem. 203, 673 

 (1953). 



