Reprinted from Scimyce, July 20, 1951, Vol. 114, No. 2951, pages 68-69. Prevalence of Escherichia coli Strains Exhibiting Genetic Recombination Joshua Lederberg Department of Genetics,’ College of Agriculture, University of Wisconsin, Madison The first bacterium to be tested by an efficient selec- tive method for the occurrence of genetie recombina- tion was strain K-12 of “scherichia coli, Experiments with auxotroph mutants of this strain promptly gave conclusive, positive evidence of genetic exchanges be- tween different mutant cells in mixed cultures (1, 2). However, subsequent attempts to obtain comparable results with a number of other strains used for genetie work were fruitless. Cavalli and Heslot (3) examined a number of auxo- troph strains from the National Type Culture Collec- tion (England) and found one that could be crossed with K-12. Unfortunately, this isolate has a complex nutrition, so far unanalyzed, which greatly hinders further work. In other characteristics it closely re- sembles K-12, - It would be surprising if K-12, the first EH. coli strain examined, should prove to be uniquely suitable for crossing experiments. Unfortunately, the method for testing fertility involved a good deal of work: it was necessary to prepare at least two nonoverlapping, double nutritional mutants from each strain. Despite improved techniques (4), such a procedure is almost prohibitive for routine survey of new strains. The fol- lowing procedure was therefore put into: effect for preliminary screening. A multiple marker strain, W-1177 (= 677-sr in [3]) has been developed from K-12 by a long sequence of mutational steps. This strain differs from the wild- type strain K-12 in these markers: polyauxotrophy for threonine, leucine, thiamin; resistance to strepto- myein and to bacteriophage Tl; failure to ferment lactose, maltose, mannitol, xylose, galactose, or L- arabinose. These may be symbolized as: T- L—B,- 8 V\’ Lae-Mal~-ete. Typical wild-type EH. coli strains are T+ L+B,+8". These four markers are useful in detecting reeombination between W-1177 and new strains to be sereened. Heavy inocula of W-1177 and of the propositus are mixed in a complete broth tube, and incubated for 6-24 hr. The mixed culture is then harvested, and the washed cells are plated on a mini- mal agar medium containing 100-1,000 ug/ml strep- tomycin. The minimal agar selects prototroph cells; the streptomycin selects 8’. The minimal streptomycin agar thus permits the growth only of T7+2+B,+S8" colonies and suppresses the two parents. This assort- ment of characters can arise either by recombination, or by mutation of the prapositus from S* to 8’. Fortunately, this mutation occurs at an extremely low TPaper No. 451. This work was supported in part by a grant from the Research Committee, Graduate School, Uni- versity of Wisconsin, with funds made available by the Wis- consin Alumni Research Foundation. rate, about once per 10° cell divisions (6), and there- fore confusion between recombinants and mutants is minimized. On the other hand, the improbable coinci- dence of three reverse mutations needed to produce a prototroph from W-1177 has never been observed in extensive controls (1,2). The principal function of the screening procedure is the rational selection of cultures appropriate for more detailed analysis by the development of auxo- troph mutants. Even in this preliminary test, however, recombination of unselected markers (V,, Lac, Mal, ete.) among the 8” prototroph selections usually veri- fied the occurrence of genetic interchange. Two groups of cultures have been screened so far for cross-fertility with W-1177 (ie., K-12). About 40 cultures from chicken cecal flora (supplied by courtesy of §. Shapiro) yielded one isolate that crosses, but very poorly. About 100 isolations fiom human urine cultures (secured through courtesy of the Wiseonsin State Laboratory of Hygiene) have given 8 that cross with about the same facility as K-12, and an equal number that appear to be less fertile (if fertile at all), so that the evidence for recombination in the latter is still inconclusive. The possibility that some ecotypie differentiation is revealed by the breed- ing test deserves further study when it is recalled that K-12 is also of human origin. Nutritional mutants are being prepared in the new isolates. The three cultures so far tested cross freely with each other, as well as with K-12 and within each strain. The new strains differ in a number of characteris- tics, including fermentation patterns (3 are sucrose- positive; 6, sucrose negative; one is a lactose-negative “paracolon” type), colony morphology (R, S, and in- termediates by the acriflavine test), and patterns of resistance to and production of colicins (7) and phages. Preliminary serological studies are under way, in addition to experiments to uncover cryptic genetic differentiation. There is a strong suggestion that coli- cin and lysogenicity interactions may act as genetic isolation mechanisms. Unfortunately, the survey method does not reveal other intrafertile, intersterile breeding groups, nor, owing to the dominance of S* (8), can it reveal unre- duced diploid hybrids between the different strains. Despite these shortcomings, however, the strepto- mycin-prototrophy selection method has succeeded in displacing strain K-12 from its position as the only “sexual” bacterium. References 1. TaruM, BE. ., und LEDERBERG, J. J. Bact., 53, 673 (1947). 2. LEDERBERG, J. Genetics, 32, 505 (1047). 3. CAVALLI, L. L., and HESLOo?T, H. Nature, 164, 1057 (1949). 4. Turia, S. E. Wethods Aled. Research, 3, 1 (1950), 5, LEDERBERG, J. J. Bact., 59, 211 (1950). 6. NEWCUMBE, H. B., and HAawirky, R. bid., 57, 565 (1949). 7. FREDERICQ, P. Rev. med. Liége, 4, 193 (1949). 8. LEDERBERG, J. J. Bact., 61, 549 (1951). 14