INFECTIVITY OF TOBACCO MOSAIC VIRUS 441 



265 niju,) of an 0.1 % solution of TMV. The amount of pelletable material 

 found in a reaction mixture is generally quite similar to the yield of reconsti- 

 tuted virus, as calculated from the infectivity of that reaction mixture. 



When poorly reconstitutmg viral nucleic acid (see later) or other nucleic 

 acids, e.g., yeast nucleic acid, are used, very little material is sedimented by 

 ultracentrifugation, and the character of its spectrum approaches that of the 

 small amount of aggregated protein which is separated by the same technique 

 from nucleic acid-free protein-buffer solutions. A second cycle of sedimenta- 

 tion from pH 7.0, 0.001 M tris-buffer produces essentially no sediment from 

 TMV protein alone or from mixtures of protein with inactive (e.g., yeast) 

 nucleic acid, while virus reconstituted by the use of highly active TMV-RNA, 

 like midegraded virus, is recovered in at least 70 % yield. These findings show 

 that there is limited significance to the electron microscopic observation of 

 Hart and Smith (1956) that reconstitution of virus-like rods occurs with any 

 RNA and even with polyribonucleotides. Apparently, these rods, the relative 

 abmidance of which was not determined, are of considerably lesser stability 

 than is TMV or virus reconstituted from active TMV nucleic acid and protein. 

 The latter material, after one or two sedimentations, shows an infectivity 

 ranging from 50 to 100 % of that of the same amount of standard virus and 

 proves, upon electron microscopic survey, to be composed, to at least 50 %, 

 of its mass of rods of close to 300 m^ length (Fig. 1, electron micrograph). 

 It is thus by all criteria almost indistinguishable from intact TMV, 



The fact that the infectivity of TMV-RNA can be greatly enhanced by the 

 reconstitution reaction has been confirmed in several laboratories (Siegel et 

 al., 1957; Commoner et al., 1956; Commoner, 1957; Bawden, 1957; Bawden 

 and Pirie, 1957). The use of pyrophosphate has given good reconstitution 

 even with the phenol type of RNA (Fraenkel-Conrat, 1957a), which pre- 

 viously had been found not to reconstitute, in both Tiibingen and Berkeley. 



V. Infectivity of Viral RNA 

 A. TMV 



The primary step toward identification of a viral component as the infec- 

 tious moiety consists in its chemical separation from other components. 

 Methods for the separation of nucleic acid and protein in comparatively pure 

 forms from TMV and TYMV have been described. Differences in the efficacy 

 of those methods for different strains of TMV have been recorded (Fraenkel- 

 Conrat and Singer, 1957; Siegel et al, 1956, 1957), and it is to be expected 

 that for other viruses these procedures for splitting will have to be modified 

 and additional ones developed. 



The noninfectivity of the resulting protein fractions can generally be de- 

 monstrated without difficulty. Since proteins may inhibit viral activity, it is 



