494 A. TSUGITA AND H. FRAENKEL-CONRAT 



viruses, which was found to vaiy widely (Knight, 1952, 1959). It appears 

 to be a general phenomenon that within a family of virus strains the 

 RNA composition seems alike, but it must be remembered that each base 

 occurs about 1500 times per molecule and that the error of the analytical 

 method is at best ±30 residues. 



In contrast to the uniformity in RNA composition, the amino acid 

 make-up of most TMV strains was found to vary. Table I summarizes 

 the results of recent analyses (Tsugita, 19f)2a; Knight et al., 1962) 

 which in general bear out the earlier data. These analyses show 

 that TjMV strains can conveniently be classified in the following analyti- 

 cal groups: A. One group of many strains which are either identical with 

 TMV, or quite similar to it, showing 1-3 amino acid residues replaced 

 by others, and always lacking histidine and methionine. To this group 

 belong most of the seemingly one-step mutants detected under green- 

 house conditions. B. Another group, exemplified by Y-TAMV, showing 

 a large change in composition including one methionine and C-terminal 

 serine (Knight et al., 1962). Apparently the Dahlemense strain analyzed 

 by Wittmann (1960b) belongs to this group. C. G-TAMV (again quite 

 different, with two methionines). D. The Holmes ribgrass strain [again 

 different, including three methionines and 1 histidine (Tsugita, 1962a)]. 



All strains, even those showing a very different composition, appear 

 to contain the same number of residues, namely, 158 per peptide chain. 

 Almost all strains studied in our laboratoiy which differ in composition 

 from T]\IV give local lesions on A^ sylvestris, rather than the systemic 

 disease, but no other correlations between composition and symptoma- 

 tology were detected. Thus, strains of the same group gave no symptoms 

 on A'', tabacum, or killed the plant; and some of the symptoms due to 

 strains of all groups could resemble one another on one or the other host. 



Recent studies of representatives of these classes of TMV strains 

 have not confined themselves to amino acid composition. For the pur- 

 pose of locating the amino acid exchanges on the peptide map, analyses 

 of the amino acid composition of the peptides derived from trypsin 

 digest of the protein, as well as preliminarj^ studies of their amino acid 

 sequences, have been initiated. Figure 4 shows comparative aspects of 

 the tentative structure of parts of these proteins (Tsugita, 1962a). These 

 data suggest some interesting generalizations. Although the difference in 

 net composition between two strains may appear relatively small (e.g., 

 8 exchanges for Y-TAMV as compared to TMV), the difference in 

 peptide composition is much greater, involving in the pair used as 

 example more than 30 exchanges, i.e., 20% of the total amino acids. On 

 the other hand, some particular peptide sequences such as Arg.Asp-NH2. 

 Arg (90-921. Val.Tyr.Arg (69-71), and Arg.Arg.Val.Asp.Asp.Ala.Thr. 



