RECOMBINATION IN VIRUSES AND BACTERIA 



135 



where they can come in contact. Of course, it could be suggested that a 

 filterable substance is still the agent of gene exchange but that it is ad- 

 sorbed on the filter, or has a very short half-life. The assemblage of facts 

 to be presented precludes this interpretation. 



Actually, the bacteria can be observed to join with one another in 

 pairs. In this behavior there is a system of polarity; not every strain of 

 bacteria will mate with every other strain. Those that are referred to as 

 F~ will not mate with each other, but they will mate with certain other 

 strains called F"*" and Hfr. The Hfr (high frequency of recombination) 

 strain is by far the most active. When Hfr bacteria are genetically marked 

 by factors for motility and cell shape, they can be distinguished micro- 

 scopically in the same suspension from F~ bacteria possessing the 

 opposite characters. In such mixed cultures it is observed that members 

 of the two strains quickly pair, and join together across a narrow cyto- 

 plasmic bridge (Figure 5.14). This attachment is temporary, lasting 

 about an hour. The paired members then separate and each exconjugant 

 divides to form a clone. But onlv the F~ clone gives recombinant 

 progeny; the Hfr parent continues to produce its original type ex- 

 clusively. This suggests first that there has been a one-way transfer of 

 genetic material from Hfr to F and, second, that since the Hfr parent 

 survives, the transfer does not include all of its genes. But we know that 

 the cells of E. coli usually have more than one set of genes, in the form 

 of several haploid DNA-containing bodies, so it might be asked whether 

 a whole or an incomplete set is transferred across the cytoplasmic bridge. 



Lederberg, in his original work on this subject, found that certain 

 genes of the Hfr strain were very frequently absent from the recom- 

 binants; in the strain he was using these included the loci for strepto- 

 mycin resistance and maltose utilization. Evidence on this point came 

 from his studies on nondisjunctional strains, which were produced as 

 progeny of crosses in which a gene called het is present. In this case. 



FIGURE 5.14. A conjugating pair of E. coli formed be- 

 tween morphologically different members of an Hfr and an 

 F~ strain (after Lederberg, 1956, J. Bad., 71:497). 



