312 C, LEVINTHAL 



B. Host-controlled Variation 



The third exception to the rule that a phage always reproduces its own 

 kind is the phenomenon of host-controlled variation. It is not known whether 

 this phenomenon is directly connected with genetic processes or not. The 

 phenomenon itself, however, is very striking and seems to occur in all the 

 well-studied phage systems. The first clearly recognized examples were 

 reported by Luria and Human (1952), Ralston and Krueger (1952), and 

 Anderson and Felix (1952). In each of these cases the ability of the phage to 

 grow on one or more host strains is altered by a single passage in a particular 

 cell type. The remarkable characteristics of this alteration which distinguish 

 it from the selection of a spontaneously occurring mutation are that it can 

 occur in all the infected cells and that a normal burst of altered phage is 

 produced from each cell. Even more important is the fact that only the last 

 cell through which a phage was grown but not any of its previous history 

 determines its host range. The host-induced modifications produce a non- 

 heritable change in the ability of a phage to grow in some particular hosts. 

 The phage P22, for example, which was studied by Garen and Zinder (1955), 

 plates with the same efficiency on the cells Salmonella typhimuriuni and S. 

 gallinarum if it has been grown on S. typhiinurium. However, the phage 

 which is liberated by S. gallinarum is modified and it plates with an effi- 

 ciency of 10~^ on S. typliimurium as compared to plating on S. gallinarum. 

 The situation here is adaptive in the sense that the modification induced by 

 S. typhimurium results in an increased ability of the phage to grow on that 

 cell while at the same time it may reduce the ability of the phage to grow on 

 some other cell. 



Although most of the host-induced modifications are adaptive (Luria, 

 1953), the one first described by Luria and Human is not. The phage T2 

 gives a normal yield when grown in E. coli B/4o, which is a one-step mutation 

 from E. coli B. However, the phage particles which are produced, designated 

 T*2, are unable to undergo further growth on either B or B/4o except for 

 about 10~^ of the infected cells, which do produce phage. A related cell, 

 Shigella dysenteriae, will yield progeny with either T2 or T*2. In addition, 

 the fraction of cells which yield progeny when infected with modified phage 

 depends very much on the growth conditions of the cells and on the medium 

 in which the cells are infected. 



In order to determine whether the host-induced modification is an altera- 

 tion of the genetic material or a phenotypic phenomenon, it is necessary to 

 know if the DNA of the modified phage is injected normally into the cell in 

 which it will not produce progeny. The direct experiment to determine this 

 with radioactive isotopes has not as yet been carried out because of various 

 technical difficulties. Garen and Zinder (1955) showed that a modified phage 

 which also contains a plaque-type mutation could contribute its genetic 



