WALDO E. COHN 



residue, this remaining link can only be through the 5' hydroxyl 

 group of the terminal ribose. Again, barring transphosphoryl- 

 ation, this was evidence for the existence of linkages other than 

 2' or 3' in at least 40% of the pyrimidine nucleotides. 



There were other, less direct warnings that all was not well 

 with the 2 ',3' hypothesis. To explain the enormous difference 

 between RNA and DNA (deoxyribonucleic acid), the latter of 

 3', 5' structure (Figure 1^), in susceptibility to alkaline hydrolysis, 

 it was necessary to assume a special alkali sensitivity on the part 

 of nucleoside 2 '-phosphates. However, the presumed synthetic 

 2' nucleotides showed no special lability to alkali. (These 

 substances are now known to have been 5' nucleotides.) Then, 

 too, the first x-ray diffraction studies (2), coupled with considera- 

 tion of bond angles and distances, gave support to a 3 ',5' (or 

 2 ',5') linkage and not to the 2 ',3'. 



The strength of the 2',3' hypothesis, which successfully 

 withstood the challenge of the foregoing contradictory evidence, 

 derived largely from the following arguments : ( 7) the (seeming) 

 recovery of only one nucleotide of each base from alkaline 

 hydrolyzates ; (2) the (apparent) identification of the sugar 

 phosphate residue as ribose 3-phosphate; (3) the absence of 

 the acid-stable ribose 5-phosphate (or its pyrimidine derivatives) 

 in acid hydrolyzates ; (4) the alkali stability of the necessarily 

 3 ',5' linked DNA. Yet each of these bulwarks was faulty. It 

 is now known that two isomers of each nucleotide (the 2' and 

 the 3') are produced by alkaline hydrolysis (16,17) and that the 

 earlier identification of ribose 3-phosphate as the sole sugar 

 phosphate was faulty (39,40). Ribose 5-phosphate nucleotides 

 can be produced from RNA by enzymic hydrolysis (19,20) and 

 the alkali lability of RNA, in contrast to DNA, stems from the 

 presence of an unesterified 2' OH group, not from a different 

 internucleotide linkage (8,10,29,63,64). These findings, which 

 were made in rapid succession in 1949-1952, fulfilled the 

 neglected prophecies and warnings of earlier date that all was 

 not well with the 2 ',3' hypothesis of RNA structure (25,55). 



The application of a new analytical technique, ion-exchange 

 chromatography, to the analysis of the conventional alkali digest 



462 



