VARIATION AND ITS CHEMICAL CORRELATES 143 



BS3, BS9, and BSIO, and their protein components were analyzed by the 

 Moore and Stein technique of ion exchange column chromatography supple- 

 mented by colorimetric procedures (de Fremery and Kjiight, 1955). No 

 decisive differences were foimd and it was concluded that the amino acid 

 compositions of the three strauis are very similar, if not identical. A com- 

 position which could pass for all three of the strains is given in Table II. 



From the above, it can be concluded that the situation with respect to the 

 protein compositions of these particular BSV strains is lilce that of only the 

 most closely related TMV strains in which also no differences in amino acid 

 content were observed. 



c. Summary of Compositio7is of Strain Proteins. From data of the sort in- 

 dicated in Table III certain conclusions can be drawn: 



1. Strains often differ from one another in protein composition. These diff- 

 erences usually are manifested by different proportions of constituent amino 

 acids. In some cases amino acids may be present in one stram and entirely 

 lacking in another. The HE. strain exemplifies this situation (see Table III). 



2. In general, there is little evidence for pronomiced change in protein 

 composition during individual mutational steps, for the closely related 

 strains are found to be most similar in composition. In fact, in some cases, 

 as with TMV and M, and with three strams of BSV, it has not been possible 

 as yet to distinguish between the proteins on the basis of composition. 



3. Differences in strain proteins seem to involve almost any of the various 

 amino acids. Only three, cysteine, leucine, and proline, appear thus far to be 

 constant in the thirteen strains of TMV analyzed. 



2. Structural Features of Strain Proteins 



A logical extension of the compositional studies on virus strains is the 

 investigation of structural features. Of course, the size and shape of virus 

 particles are based on their chemical structure, but thus far no significant 

 differences in morphology between virus strains have been demonstrated. 

 One can, therefore, tentatively conclude that the gross structures of the 

 strams are very similar, if not identical. On the molecular level, however, 

 there are many possibilities for differences between strains. In the case of the 

 protein, there are the questions of how^ many peptide chains there are, how 

 they terminate, how they are arranged in space, and how they are linked to 

 the nucleic acid. Within peptide chains, there is the question of sequence of 

 amino acid residues. These and similar problems are under investigation and 

 some of the residts obtained will be described. 



a. Some StruMural Features of TMV and Strains. Early studies of the low- 

 angle scattering of X-rays by TMV and some of its strains (Bernal and 

 Fankuchen, 1941) revealed a regularity of structure so great that it was 



