444 H. FRAENKEL-CONRAT 



Electron microscopy has been used as a specific means of searching for 

 TMV particles in nucleic acid preparations. Control experiments showed that 

 added traces of TMV (1 %) can be sedimented almost quantitatively from 

 nucleic acid in the microtubes (2 ml.) available for use with the 40-rotor of 

 the Spinco preparative ultracentrifuge (Model L), Yet no, or very few, par- 

 ticles were detected in many infectious detergent preparations of nucleic 

 acid (Fraenkel-Conrat et al., 1957a). With a somewhat different technique, 

 Mcliaren and Takahashi (1957) failed to find TMV particles in the phenol 

 type of preparations. 



The positive aspect of this search for active contaminants is the actual 

 identification of the infectivity with the main component of the preparation, 

 i.e., with the RNA. The foremost tool for such identification has been the 

 enzyme, ribonuclease (RNAase). Thus, extremely low concentrations of 

 pancreatic RNAase were found to inactivate, while other pancreatic enzymes, 

 including deoxyribonuclease (DNAase), depressed the infectivity only at 

 much higher concentrations (Fraenkel-Conrat et al., 1957a). The infectivity 

 of the intact virus was not affected by the same or by several higher orders 

 of magnitude RNAase concentrations. Evidence was also adduced that the 

 physicochemical parameters of the RNA were affected by enzyme action that 

 caused incipient loss of infectivity (Gierer, 1957). The sensitivity to ultra- 

 violet light (2537 A) shown by the infectivity is also a strong indication 

 that it is a property of nucleic acid, and in some strains differentiates the 

 nucleic acids from the corresponding virus (Siegel et al., 1956; McLaren and 

 Takahashi, 1957). The finding that in various types of fractionation experi- 

 ments the infectivity parallels the RNA concentration is additional evidence. 

 Thus, the sedimentation experiments mentioned earlier not only failed to 

 reveal the presence of virus particles but showed that infectivity and RNA 

 concentration were proportional at various levels of the tube. Similar results 

 were obtained in sucrose gradient centrifugation (unpubhshed experiments). 

 Anti-TMV sera were used, as previously mentioned, to search for con- 

 taminating virus protein. But they were also employed to differentiate viral 

 from nucleic acid infectivity. When the latter was found not to be inhibited 

 by such sera or the corresponding y-globulin fraction at levels that inhibited 

 the intact virus, the nonprotein nature of this infectivity was further 

 substantiated (Gierer and Schramm, 1956; Fraenkel-Conrat et al., 

 1957a). 



Additional criteria for the differentiation of the tw^o types of infectivity 

 are based on the greater lability of the nucleic acid infectivity. Thus, exposure 

 at room or incubator temperatures to 0.02 to 0.1 M salts generally caused 

 rapid loss of nucleic acid infectivity, but not that of TMV (Gierer and 

 Schramm, 1956; Fraenkel-Conrat et al., 1957b). The mechanism of this un- 

 expected sensitivity of the RNA to salts will be discussed below. The shift 



