512 A. TSUGITA AND II. FRAENKEL-CONRAT 



(juickly initiated research projects along similar lines and reported 

 preliminary results. Thus Lengyel et al. (1961), using essentially the 

 same system, coniirmed that poly-U led to the synthesis of polyjihenyl- 

 alauine. These workers obtained a molar ratio of 3.25 between the 

 amount of nucleotide used and the newly incor])orated amino acid, 

 compared to 2.4 reported by Nirenberg and Alatthaei (1961j. This figure 

 may represent a first a])proximation by biochemical nu'thods to the 

 coding ratio, and it is heartening that it is so close to the tri])let recently 

 postulated by Crick et al. (1961) on genetic grounds. 



Further amino acid incorporation studies in the laboratories of both 

 Nirenberg and Ochoa have centered on the use of various nucleotide 

 copolymers of different composition (Martin et al., 1962; iSpeyer et al., 

 1962a,b; Lengyel et al., 1962). These favored in more or less specific 

 manner the incorporation of other amino acids, but, as one might expect, 

 these random copolymers were cjuantitatively much less effective than 

 poly-U. All polymers that have been reported on contained uridylic acid 

 and their effectiveness in regard to the incorporation of various amino 

 acids was generally expressed in terms of the ratio to phenylalanine 

 incorporation. Although the use of random copolymers did not enable 

 the authors to determine the sequential structure of any coding unit, 

 except for poly-U, nor whether the unit was definitely a triplet, they 

 were able to suggest probable coding combinations for 17 out of the 20 

 naturally occurring amino acids, with only a few instances of disagree- 

 ment in the results of the two laboratories participating in the race. In 

 addition, Nirenberg et al. (1962) have shown that phenylalanine linked 

 to soluble RNA was an intermediate in the process, using polyuridylic 

 acid. 



The finding that poly-U carries the message for phenylalanine poly- 

 merization is in disagreement with Crick's "commaless triplet code" 

 as described. As mentioned before, these facts and their own data have 

 led Crick et al. (1961) to modify their concepts and to abandon the 

 postulation of nonsense triplets. It thus becomes ever more likely that 

 the code is a degenerate one. 



Two years preceding these exciting demonstrations, we reported the 

 amino acid exchanges in a chemically evoked nnitant of TJNIV (Tsugita 

 and Fraenkel-Conrat, 1960), and Wittmann (1959b) observed no amino 

 acid exchanges in another chemical mutant. Since then both groups have 

 concentrated their efforts on analyzing many mutant proteins, and these 

 observations have supplied important experimental data pertaining to 

 the coding problem. 



Wittmann analyzed 120 nitrous acid mutants (1961) and we analyzed 

 56 chemically induced mutants (Tsugita and Fraenkel-Conrat, 1960, 



