490 



A. TSUGITA AND H. FRAKXKEL-CONRAT 



for its own structure as well as for the structure of its homologous 

 protein, undisturbed by the jircscncc of heterologous protein in the 

 infecting particle (see Fig. 5). 



ORIGINAL 

 STRAINS 



DEGRADATION 



REPLICATION PROGENY 



Fig. 5. Mi.xed rcconstitution. The identity of tlu^ progeny protein, as compared 

 to the two parental strains, symbolized as A and B, has been established bj^ sero- 



lofiiial tests and amino acid analyses. 



D. STRUCTURAL DIFFERENCES IN THE RXA OF DIFFERENT STRAINS 



While differences in the amino acid composition of viral strains have 

 long been known to exist, attempts to detect differences in nucleotide 

 composition have been unsuccessful. However, as previously pointed out, 

 an analytical error as low as ztO.5% would give the composition of 

 TMV-RNA with an accuracy of only ±6 to 9 residues for each base. 

 Therefore, one cannot conclude from the available data that the strains 

 of a given virus are identical in composition, but only that they 

 resemble one another. Actually, jiresent knowledge of the genetic cai)a- 

 bility of the RNA strongly suggests that there must exist structural 

 differences between the RNA of viral strains. Such differences could 

 conceivably be due to base interchange and thus become evident only 

 from sequential analyses rather than from the composition of the RNA. 



Experimental searches for structural difference of the RNA from 

 strains have only recently been begun. End group analyses of four 

 strains have shown these to have the same tenninal adenine residues as 



