X. TMV STUDIES IN GENETIC CODING 511 



authors suggested that of the 64 triplets some "made sense/' i.e., they 

 represented an amino acid, while others did not. As an example, if A-G-U'* 

 stood for leucine and C-A-U^ for valine, then G-U-C and U-C-A would be 

 condemned to represent nothing ("nonsense"), so that . . . A-G-U-C-A-U 

 . . . anywhere along the chain could be read only as Leu.\'al. What made 

 this theoiy most appealing was the fact that the 64 possible triplets were 

 thus automatically reduced to 20 sense-making ones, in agreement with 

 the number of common building blocks in proteins. The resulting code 

 was obviously non-overlapping, and it was comma-free, since each 

 nucleotide along the chain pertained to only one sense-making and thus 

 coding triplet. It should be noted that in a commaless code triplets 

 composed of the same nucleotide (i.e., A-A-A) are excluded, since a 

 sequence such as -A^-A--A^-A*-A^- does not give a unique sense; for it 

 can be read starting at position 1, 2, or 3. 



More recently. Crick et al. (1961) presented the first experimental 

 evidence, based on genetic studies of T4 phage and acridine mutants 

 derived from it, that the code for the DNA of bacterial genes was 

 actually triplet in nature. They also arrived at the conclusion that the 

 reading of the code must be from one end, and thus does not require 

 nonsense triplets as commas. Their new data seemed to rule out the 

 existence of meaningless triplets. Only between cistrons was there evi- 

 dence for a break in the continuity of decipherable nucleotide sequences. 



In addition to the theoretical and genetic approaches to the coding 

 problem, an analytical approach was also attempted. Thus, Yeas pro- 

 posed a code based on correlations between published analyses of the 

 nucleotide and amino acid compositions of simple viruses. Woese 

 (1961a,b) followed this path further, additionally using amino acid 

 exchange data from several kinds of j)roteins in related species. 



A great advance in our understanding of the coding jiroblem has 

 recently come about when Nirenberg and ]\Iatthaei reported on their 

 use of an excellent system to attack this problem in direct and experi- 

 mental manner (1961). We have previously described the cell-free 

 protein synthesizing system of E. coli (containing S-RNA, ribosomes, an 

 ATP-generating system, and C"-amino acids) which these authors 

 reported to be greatly stimulated by the addition of any RNA, and par- 

 ticularly by TMV-RNA. Surprisingly, polyuridylic acid (poly-U) was 

 found to be particularly effective in this regard, and Nirenberg and 

 Matthaei were soon able to demonstrate that this additive led to the 

 exclusive synthesis of polyphenylalanine. 



This finding attracted the deserved attention, and othei- l;d)oratories 



^A represents a nucleotide residue, the base of which is adenine; G represents 

 guanine; U, uracil; and C, cystosine. 



