X. TMV STUDIES IN GENETIC CODING 



505 



thus i)i'oducc either no virus progeny, or too little matci'ial for chemical 

 analysis. 



We have discussed one class of mutants showing the same amino acid 

 composition as connnon T]\IV, and another group showing a limited 

 number of alterations. Finally, another class was observed at our lab- 

 oratory wliicii was veiy markedly different from the parental strain, 

 differing by 8 or by 16-17 net amino acid exchanges, and probably in 

 many more sites on the molecule. Representatives of this group closely 

 resembled two natural strains previously described (groups B and C in 



CH^CONH 



273 



282 -BS- 



171 -BB- 



332 -B- 



DEAMINATION 220 



329 



237 



32IB 



_284 



218 -B- 

 233 -fl- 

 207 -9- 

 235 -B- 



187 



326 



D ODD 



-B— e- 



-B— &■ 



BROMINATION 



D ODD 



DDD 



I 214 — B- 

 METHYLATION ^Z? "f" 



-B- 



i 



176 

 215 



5 7 



t^?^^4i(^6^W8-^ 9 



10 



ee- 



fiB- 



I1K-I2^C00H 



HUBU 



Fig. 7. Location of amino acid exchanges in chemicalb' evoked mutants. Each 

 horizontal line represents the peptide chain of a mutant, subdivided into tryptic 

 peptides as in Fig. 4. SoHd rectangles indicate definite location of a change, open 

 rectangles possible alternate sites of a single change in a specific sequential peptide 

 (#1-12). Only in strain 326 are 2 of the 4 potential aspartic acids of peptide 

 #1 transformed to .serines. 



Table I), and concern arose that such virus isolates were actually derived 

 from contaminating lesions. The finding that three of the five members 

 of this group clearly differed from the natural strains by the replacement 

 of 1 or 2 amino acids, in one instance involving an increase in glycine 

 content by 25% seemed to rule out this interpretation. Concerning the 

 origin of these strains, they were derived from random lesions on iV. 

 sylvestris resulting from the same reaction mixtures (deamination. 

 bromination, and alkylation), which also yielded strains showing from 

 to 2 amino acid replacements. The average number of hits per 

 particle, as judged from the level of inactivation was 3-6. Thus, no 

 reason becomes evident for the appearance of a new family of mutants 



