Frangois Jacob and Elie Wollman 



number of Hfr. This number is 100 times greater than the number of 

 infective centers found in the control. After the 60th minute, the number 

 of infective centers suddenly increases and at about the 150th minute 

 reaches a value 50 to 100 times greater than that of the first plateau. 



The difference between the mixture and the control Hfr is further ac- 

 centuated if anti-phage serum and streptomycin are added to the two cul- 

 tures at the beginning of the experiment. For, the number of infective 

 centers in the control falls rapidly to zero while the number of infective 

 centers in the mixture is only reduced by one-half. 



The development of phage X has thus been induced by putting Hfr ly+ 

 in contact with F~ ly~. This development can only take place in F~ 

 bacteria (or in the zygotes), since the Hfr bacteria were sterilized by 

 streptomycin. If the same cross is repeated with an F~ S^ strain, addition 

 of streptomycin during the latent period arrests the development of the 

 phage. 



An even more direct demonstration of the passage of the prophage from 

 Hfr to F" bacteria (or to the zygote) and of its development in the latter 

 is furnished by the following experiment. It is known that certain bacterial 

 strains confer a phenotypic modification to the bacteriophages which they 

 produce.'* Thus bacteriophage X(K12) produced by E. coli K12(X), 

 whose efficiency of plating is identical on E. coli K12 S and on E. coli C, 

 has a greatly reduced efficiency of plating (10~' to 10~^) on K12 S after 

 passage on E. coli C. If development of the prophage actually takes place 

 in the non-lysogenic F~ bacteria, then in a cross of K12 Hfr(X) S^ with 

 C/X F~ S'' the bacteriophages produced should be X(C) and not X(K12). 

 Experiments show this to be the case. 



When F~ bacteria perpetuate a prophage, normal or defective, the 

 immunity which the prophage confers on them* protects them equally 

 against the development of a prophage coming from the Hfr parent. 



If one crosses Hfr bacteria superinfected with virulent mutants of the 

 type Xv,* one observes little or no passage of the superinfecting phages into 

 the F~ bacteria. This result indicates that the transfer of phage material 

 from Hfr to F~ takes place in the prophage state and not in the vegetative 

 state. It also indicates that the induction of phage development in the 

 course of recombination takes place in the F~ bacterium (or in the zygote) 

 and not in the Hfr bacterium. 



It is evident that the effect described here must introduce important 

 distortions in the genetic analysis of E. coli K12: in effect, only those zygotes 

 give rise to recombinants in which no multiplication of the phage has 

 taken place. The proportion of these survivors will vary considerably 

 depending on the nature of the cross. Thus in the cross of Hfr ly"^ Bj S^ 

 by W678 T~ L~ B~ ly~/XS'^ the number of recombinants of the type 



*G. Bertani and J. J. Weigle, J. Bad., 65, 113, (1953). 



T. Jacob and E. L. Wollman, Cold Spring Harb. Symp., 18, 101, (1953). 



337 



