INFECTIVITY OF TOBACCO MOSAIC VIRUS 437 



cause a marked drop in the ])H of the solution. Quantitati^'e analyses of this 

 reaction have indicated the dissociation of about two protons per subunit 

 (Fraenkel-Conrat, 1957a; Koshland et al., 1958). This finding is the most 

 definite indication of the nature of some inter-subimit bonds, and has been 

 interpreted (Fraenkel-Conrat, 1957a) as evidence that two carboxyl groups 

 are involved in such bonding and thereby prevented from dissociating over 

 the entire stability range of the virus, i.e., up to pH 9 for short time periods 

 at low temperature. Further evidence that carboxyl groups are involved is 

 the finding that divalent metals, particularly lead, are able to replace two 

 protons at pH 6. Presimiably the same carboxyl groups are involved; how- 

 ever, the metal in these cases does not disrupt the structure but actually 

 seems to stabiHze it when it replaces the not normally dissociatable protons 

 (Fraenkel-Conrat and Narita, 1958). Much further work is required before 

 this intriguing structural relationship will have been clarified. 



The native protein gives clear solutions in water except over the range of 

 pH 3.5 to 6.5, where it at,-gregates. Its characteristic ultraviolet absorption 

 curve, with a sharp maximum and minimum at 280 and 250 mfx [ratio of 

 A max/min = 2.4 for TMV; A (max) for 0.1 % solution (pH 8) is 1.3] can 

 be used both as a measure of its concentration, and as an indication of its 

 freedom from nucleic acid. Repeated freezing and thawing of the protein 

 solutions, which are usually stored frozen to prevent spoilage, causes gradual 

 denaturation, as indicated by decreases in the masked — SH titer and in 

 solubility (Fraenkel-Conrat and Singer, 1957). 



Considerable work has been done in several laboratories concerning the 

 effect of chemical modification of the protein on the infectivity of the virus. 

 These studies have been repeatedly reviewed, most recently by Knight (1954) 

 and by Stanley and associates (1958). No additional work has been done in 

 this field, and therefore it will not be discussed here. In the Hght of present 

 knowledge, it is not surprismg that extensive protein modifications often had 

 little or no effect on the viral infectivity. 



III. Assay of Infectivity of TMV and TMV-RNA 



Several varieties of tobacco give local lesions upon inoculation with TJ\IV 

 and can thus be used for quantitative evaluation. The preferred mutants in 

 this laboratory at present are Nicotiana gluiinosa and the Xanthi variety of 

 N. tabaccum.. The former host has the advantage of permitting the differentia- 

 tion of one particular strain (HR) by the size and appearance of the lesions 

 it produces (Fraenkel-Conrat and Smger, 1957). The latter host is the most 

 sensitive variety of those compared by us. 



In aU tests for the activity of an miknown virus preparation, a standard 

 solution is included for reference. This is obtained by diluting an aliquot of a 



