X. TMV STUDIES IN GENETIC CODING 513 



1962; Tsugita, 1962a,b). In both laboratories at least half of the observed 

 mutants showed no amino acid exchange. We have ])reviousIy discussed 

 the grounds for accepting this negative finding as valid. This finding 

 means that a good part of the RNA docs not find expression in the 

 composition of the virus coat protein. Thus, coding schemes based on the 

 direct com])arison between the composition of the RXA and the protein 

 of the viruses, such as those proposed by Yeas and Woese, seem to have 

 lost any basis in reality. The finding that not all RXA alterations lead to 

 coat protein alterations is in line with expectation, if one considers the 

 numbers of residues involved. Assuming a triplet code, the 158 amino 

 acids would be expected to be governed by only 474 nucleotides, which 

 corresponds to only 7.4% of the nucleotides in TMV-RNA. Most of the 

 RNA would thus be assumed to carry information not pertaining to the 

 coat protein, which alone has been analyzed. Actually one might expect 

 that only about 1 out of 10 mutants should show protein changes, but it 

 is ciuite possible that a high proportion of the particles altered in seg- 

 ments pertaining to properties other than protein coat structure are not 

 viable. 



The chemical mutant (#171) first analyzed by us (Tsugita and 

 Fraenkel-Conrat, 1960) had 3 amino acid exchanges and the location of 

 these changes was scattered throughout the sequence. This observation, 

 as well as the detection of mutants with single exchanges, disproves any 

 overlapping code. 



Investigations of the deamination reaction with DNA have led Viel- 

 metter and Schuster (1960) to the conclusion that only the deamination 

 of cytosine and adenine was mutagenic. The resulting uracil and hy- 

 poxanthine would on chemical grounds be expected to pair preferentially 

 as if they were thymine and guanine, respectively. In contrast, the 

 deamination product of guanine, xanthine, still resembles guanine. Based 

 on such consideration and an assumed triplet code, Gierer (1962) 

 arranged all possible permutations of the 64 triplets in a scheme gov- 

 erned by the result of their deamination (C^'U, A— > (hypoxanthine) 

 — > G) in the form of 8 octets as shown in the diagram. 



CAC 



./ 



\ 



UAC CGC CAU 



UGC UAU CGU 



\ 1 ^ 

 UGU 



