INFECTIVITY OF TOBACCO MOSAIC VIRUS 433 



the crucial aspect of the observations is the nonuifectivity of these mock- 

 viruses. If one relates the mode of infection by bacteriophages, as elucidated 

 by Hershey and Chase (1952) and Hershey (1957), with the occurrence of 

 these inert proteins, then the hypothesis suggests itself that in aU viruses it 

 is the nucleic acid that carries the infectivity. This concept was not generally 

 favored for plant viruses, because the analyses of different strains of TMV 

 had revealed definite differences in amino acid composition, but not in 

 nucleotide composition (Knight, 1947, 1952; Black and Knight, 1953; 

 Markham and Smith, 1950; Cooper and Loring, 1954). However, since 1955 

 our knowledge about the nature of the infectivity of TMV has advanced 

 beyond the stage of reasoning by analogy or inference. For we now have ex- 

 perimental proof that it is the nucleic acid that carries the infectivity of TMV. 

 And the remaining inference that the nucleic acid is the infective component 

 of all viruses is being experimentally supported at ever more frequent 

 instances. The subsequent sections of this chapter will deal with those 

 experimental results, largely concerned with TMV, which represent both 

 the basis and the consequences of this new concept of the nature of viral 

 infectivity. 



II. Degradation of TMV 



Tobacco mosaic virus and its strains have probably been more intensively 

 studied than all other purified plant viruses together. Separate chapters will 

 be devoted to the chemical structure of TMV protein and nucleic acid, and 

 particularly to the structural differences between strains. Much of the rest of 

 this chapter wiU be concerned with the degradation of TMV into its two 

 components, with their functional and some of their macromolecular chem- 

 ical properties, and with the mechanism of reconstitution of complete virus 

 from the two components. While none of these steps has been clearly dupli- 

 cated with other plant viruses, there are suflSicient hints and indications to 

 establish confidence that the primary functional properties and relationships 

 are the same for aU viruses. This will surely become evident in the next few 

 years, when techniques will be developed for the separation of other viruses 

 into their native components. 



A. Preparation of Nucleic Acid 



A great variety of agents, or conditions, almost aU protein denaturants, 

 cause the breakdown of the architecture of the viruses. There is probably no 

 chemical agent which has not been so employed in early studies on TMV by 

 Stanley (1935b) and by Bawden and Pirie (1937b). However, only a small 

 number of these agents have been studied in detail, and have yielded practical 



