CHEMICAL INTERFERENCE WITH PHAGE GROWTH 283 



rapidly in simple media (Doermann, 1952), or very slowly in 

 nutrient broth (Benzer and Jacob, 1953), although injection is 

 probably slowed or prevented by cyanide in both instances 

 (Benzer, personal communication). Similarly, bacteria first 

 starved and then infected with T2 in salt solution lose productive 

 potential by a process that can be partly prevented by feeding 

 amino acids but not by feeding glucose (Gross, 1954b). In this 

 instance, too, there is indirect evidence for delayed injection; the 

 phage-producing potential of the rescuable infected cells remains 

 sensitive to antiphage serum for variable times up to 10 minutes 

 after infection (Gross, 1954a). 



For experimental purposes, it is often desirable to be able to 

 infect bacteria under conditions in which abortive infection does 

 not occur, and yet phage growth is arrested at an early stage. 

 The use of cyanide for this purpose, mentioned above, is not 

 always successful, probably because injection is interfered with. 

 Chloramphenicol added before infection may prove more useful. 

 It arrests phage growth at an early stage (Tomizawa and 

 Sunakawa, 1956) but does not interfere with injection of T2 

 (Melechen and Hershey, personal communication). Low 

 temperatures, on the other hand, are not suitable (Adams 1954, 

 1955). 



5. Noninfective Phage Particles Containing Structural Analogues 



In the case of abortive infection, phage development is halted 

 before the completion of a mature phage particle. With other 

 chemical manipulations it is possible to allow phage development 

 to proceed to completion with the formation of structurally 

 mature phage which, however, are noninfective. In effect, 

 still-birth rather than abortion is brought about. This results 

 from the incorporation of an unnatural amino acid or pyrimidine 

 into the protein or nucleic acid of the phage. 



In the presence of sulfanilamide, E. coli requires thymine and 

 other metabolites for growth. Utilizing this fact, Dunn and 

 Smith (1954) were able to replace thymine in the DNA of T2 

 and T5 by analogues of thymine in which the 5-methyl group 



