JOSHUA LEDERBERG 



to conclude that recombination in E. coli, strain K-12, results from a sexual 

 process (13). This conclusion is based upon evidence which is, to date, entirely 

 negative, indirect, or genetic. It is substantiated, however, by single-cell studies 

 on heterozygous diploid cultures (14, 142), and by a direct demonstration of 

 the nonfiltrability of the agents of recombination (40). A variety of selective 

 techniques can be used for the isolation of recombinants from strain K-12 of 

 E. coli. In addition to the selection of prototrophs from mixtures of auxotrophs 

 which is most commonly used, one may select for dually resistant recombinants 

 from mixtures of cells each resistant to one inhibitor ( 83 ) . A combination of 

 these methods has been profitably employed to detect new strains crossable to 

 K-12; a number we're found, some with many properties differentiating them 

 from each other. Developments in the author's laboratory to the date of this 

 writing have been summarized (47-Lederberg) . 



Recombination in strain K-12 has been applied by a number of other 

 workers for various problems ( 102) . The genetics of resistance to Chloromycetin 

 (6) and to streptomycin have been contrasted (46, 101). The very occurrence 

 of factorial segregation shows, of course, that bacterial enzymes are not to be 

 identified with the genes required for their formation (98). In fact, it is 

 controversial whether the patterns of gene-enzyme interrelationships in bacteria 

 support the "one gene-one enzyme" proposed as a generalization from the 

 properties of auxotrophs in Neurospora (69, 47-Lederberg, Horowitz, Bonner). 

 Transformations. — The history of "type transformation" is often regarded as 

 dating from Griffith's classical experiment on pneumococcal types in 1928 (64), 

 but this work is antedated by a confusing array of studies which date at least 

 as far back as the controversy over the etiology of typhus fever. When the 

 rickettsial etiology of this disease was finally established, many workers sug- 

 gested that the serological reaction of Proteus OX- 19 with rickettsial antibodies 

 (Weil-Felix reaction) was due to a transformation of non-reactive Proteus 

 by rickettsial products, and gave the name "paragglutination" to induced 

 serological variations such as were supposed to be involved here. In support 

 of this hypothesis, (to which no credit is now given) many workers reported 

 that the serological reaction of E. coli and other enteric bacteria could be 

 modified by cultivation in filtrates, extracts, or lysates of serologically distinct 

 cultures (reviewed in 81 ) . Griffith's demonstration was, however, the first which 

 proved to be readily reproducible by other workers, and remains a proper 

 starting point for modern discussions of transformation. His paper is unfortun- 

 ately too lengthy to be appropriate for reprinting here, but his observations, and 

 those of other workers leading to the isolation and characterization of the 

 active principle responsible for the pneumococcus transformation are sum- 

 marized in the paper reprinted here ( 16) . 



The genetic interpretations of this transformation are currently a subject of 

 lively discussion. Its description as a specific induced mutation is probably less 

 fruitful than as a transfer of hereditary material from one cell to another 

 (47-Taylor) . The main questions which still have to be answered include 



