INFECTIVITY OF TOBACCO MOSAIC VIRUS 443 



detectable traces of DNA, while 0.1 % of added DNA could be quantitatively 

 recovered. Also microbiologically no DNA could be detected in sterile 

 TMV (0.001 %, Hoff-J©rgensen, personal communication). 



Of particular importance naturally is the extent of contamination witli 

 any virus protein, and much work has been expended on this question 

 (Fraenkel-Conrat and Williams, 1955; Fraenkel-Conrat et al., 1957a; Gierer 

 and Schramm, 1956; Gierer, 1957). Colorimetric tests, using a sensitive 

 biuret test, have indicated the presence of less than 0.4-0.5 % of protein, 

 Apphcation of a microadaptation of the Folin-Lowry test has shown a 

 definite color, indicative of the presence of 0.1 to 0.5 % of protein in different 

 preparations (Ramachandran and Fraenkel-Conrat, 1958); since no type of 

 nucleic acid was available which gave less than this amount of color, and 

 since guanylic acid also gave a certain amount of color, which in TMV-RNA 

 would contribute about 0.1 % of spurious protein, the significance of the 

 observed, slight chromogenicity is doubtful. Were it undegraded TMV, 

 it would be 1-2 orders of magnitude below that required to account 

 for the infectivity of the nucleic acid, and 4-5 orders below that of the 

 reconstituted infectivity obtainable from it. Also, the variations in 

 chromogenicity are in no way correlated with the infectiousness of different 

 preparations. 



Of other chemical tests for protein, amino acid analyses have been most 

 intensively used (Gierer and Schramm, 1956; Fraenkel-Conrat et al., 1957a). 

 The fact that nucleic acid yields considerable amounts of glycine, and possibly 

 of other amino acid-like products under the conditions of jDrotehi hydrolysis 

 (16 hours, 108°C., sealed evacuated tubes, 6 N HCl) complicates this apj^roach, 

 or rather its interpretation. A technique which circumvents this difficulty is 

 to degrade the nucleic acid by alkali (A'^ KOH, 23°C., 18 hours), precipitatmg 

 any undegraded protein by TCA, and analyzing this precipitate after 

 hydrolysis by the FDNB method. Both this and the direct method have 

 been repeatedly applied. At times only traces of glutamic and aspartic acid 

 were detected, while other preparations have yielded more complex patterns 

 of amino acids, not usually in the proportions characteristic of the virus 

 protein, and ranging in total amount from 0.02 to 0.04 % of the amoimt of 

 nucleic acid used (Fraenkel-Com-at, mipublished). Again, such findings can- 

 not be interpreted in terms of viral contamination. If the results were more 

 consistent, then one might conclude that some peptide-lil^e mciterial was 

 associated with the active RNA and might possibly play a functional role in 

 the genetic process. 



Serological tests have also been used to search for the presence of native 

 virus protein in TMV-RNA, and these have given a figure of less than 0.02 % 

 (Gierer and Schramm, 1956). The use of S^^ to detect the presence of virus 

 protein was advocated by Knight (1957b). 



