352 



CHAPTER 38 



trophs for only one nutrient, is proof that 

 these results are not due to spontaneous 

 mutation. They must, therefore, be attrib- 

 uted to some type of genetic recombination. 



Could the genetic recombination observed 

 be the result of transformation? This possi- 

 bility is unlikely since almost all transforma- 

 tions involve single loci. Recall that the 

 frequency with which two loci are transformed 

 in the same bacterium is much lower than 

 that for single transformations, and is pre- 

 sumably dependent upon the close proximity 

 of two loci on the bacterial chromosome. 

 Yet, in the present experiment, double re- 

 combinants are common. In fact, certain 

 recombinations for two loci occurred more 

 frequently than did the recombination of 

 these loci singly. Transformation would 

 even less readily explain the large number of 

 triple recombinants, the complete proto- 

 trophs. 



Nevertheless, specific tests may be made to 

 rule out transformation as an explanation. 

 It is found that the number of prototrophs 

 obtained by recombination is uninfluenced 

 when DNAase is added to the mixing and 

 plating media. No transforming activity is 

 demonstrated when one culture is exposed to 

 filtrates or autolysates of the other. Finally, 

 a U-tube can be constructed with a filter of 

 sintered glass separating the arms. Broth can 

 be added and the two strains placed in differ- 

 ent arms. The medium plus soluble sub- 

 stances and small particles (including viruses) 

 can be flushed back and forth through the 

 filter. Yet no recombinants are found in 

 platings made from either arm. We may 

 conclude, therefore, that the genetic recombi- 

 nation detected in E. coli is not due to trans- 

 formation. It is also not dependent upon a 

 virus. It can be shown, moreover, by plating 

 a mixture of three lines of K12 diff"ering in 

 the mutants which they carry, that the large 

 number of different recombinants obtained 

 can all be explained as the result of recombi- 

 nation between any two lines; no individuals 



are obtained recombinant with respect to 

 markers in all three lines. Apparently, this 

 type of genetic recombination depends upon 

 actual cell-to-cell contact between pairs of 

 bacteria, and therefore involves a sexual proc- 

 ess. 



The frequency of sexual recombination in 

 the first experiment is only about one per 

 million cells (100 sites of recombination per 

 100 million bacteria placed on the master 

 plate). The rarity of the event makes it 

 fruitless, at this point in the investigation, to 

 search microscopically for evidences of bac- 

 terial mating. (You should, however, recog- 

 nize that the importance of a new phenom- 

 enon should not be judged by the frequency 

 with which it occurs in experiments that first 

 detect it. Recall, for example, that the 

 initial quantity of DNA first synthesized in 

 vitro was infinitesimal in comparison with the 

 amount synthesized in later work, and that 

 the rate of transformation observed initially 

 was very much smaller than the 10-25% rate 

 which can be obtained today with modified 

 techniques.) 



It should be noted that we have used a 

 medium that selects certain recombinants for 

 detection and not others. In the first experi- 

 ment discussed, only recombinants possessing 

 certain markers for nutritional independence 

 were selected. These are called selective 

 markers. Thus, while the prototroph r+L+ 

 Bi+B^Pa^C^ was detectable, it was not pos- 

 sible to test for the occurrence of the comple- 

 mentary polyauxotroph T~L~BrBPa^C~. 

 Since no test has been made for the multiple 

 auxotroph one could doubt its occurrence. 

 It is entirely reasonable that the immediate 

 result of mating is a zygote which contains a 

 combination of part or all of the genotypes 

 of the two parental cells. Although we have 

 assumed that integration could take place if 

 DNA passed from one cell in contact with 

 another, that is, between two bacteria in 

 conjugation, we have so far no evidence that 

 it does. In other words, the possibility re- 



