GENE RECOMBINATION IN THE BACTERIUM ESCHERICHIA COLI E L TATUM AND JOSHUA LEDERBERGi Department of Botany and Microbiology Yale University New Haven Connecticut Received for publication March 10 1947 The study of inheritance in bacteria has for the most part been confined to the investigation of mutational changes in the course of clonal reproduction With the exception of experiments on pneumococcus type transformations there have been few studies on the direct hereditary interaction of one bacterial type with another The conception that bacteria have no sexual mode of reproduction is widely entertained This paper will be devoted to the presentation of evidence for the occurrence in a bacterium of a process of gene recombination from which the existence of a sexual stage may be inferred On the basis of mutation studies many investigators have concluded that the hereditary properties of bacteria are based on the existence of genes Luria and Delbruck 1943 Roepke et al 1944 Lwoff 1941 Demerec and Fano 1945 Gray and Tatum 1944 although it is not clear whether these genes should be homologized with the Mendelian factors of higher organisms or with the extranuclear factors which have been demonstrated in some microorganisms and higher plants Sonneborn 1943 Spiegelman et al 1945 Rhoades 1943 The genic basis of microbial inheritance does not depend on the demonstrability of a sexual phase in bacteria However more powerful genetic methods paralleling classical Mendelian analysis would be available if it were possible to follow the inheritance of characters in the products of a sexual fusion The few examples of this approach thus far reported have provided no incontrovertible evidence for sexual reproduction in bacteria The phenomenon of paragglutination in the colon typhoid dysentery group might be regarded as an instance of bacterial hybridization and was so interpreted by Almquist 1924 As reported by numerous authors paragglutination refers to the development of new types which react with antisera for each of two distinct strains when these are grown together in mixed culture Kuhn and Ebeling 1916 Salus 1917 Wollman and Wollman 1925 The significance of these observations has been attacked by several authors Breinl 1921 Arkwright 1930 Kauffmann 1941 chiefly on the grounds that the paragglutination represents a nonspecific cross reactivity characteristic of rougher phases of these organisms Hansen 1929 failed to obtain paragglutination in her experiments In the light of more detailed recent information on the antigenic structure of this group this problem certainly deserves a critical reinvestigation 1 Fellow of the Jane Coffin Childs Memorial Fund for Medical Research This work was supported in part by a grant from the Jane Coffin Childs Fund and will be submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at Yale University 673 674 E L TATUM AND JOSHUA LEDERBERG VOL 53 Sherman and Wing 1937 have described experiments designed to detect recombinations of fermentative characters in mixed cultures of various Escherichia coli and Aerobacter aerogenes strains Although new combinations of biochemical characters were found similar types were found to ad equal extent in pure cultures so that these authors could not infer the occurrence of a sexual fusion Their experiments are of the greatest interest however since they represent the first attempt to study this problem in bacteria by genetic methods using clear cut characters Gowen and Lincoln 1942 later performed similar experiments with strains of Phytomonas 8tewartii using cultures differing in morphological and pigment characteristics As in Sherman and Wing s studies these authors were unable to differentiate the new types they found in their mixed cultures from types which arose spontaneously in single cultures For this reason a definite conclusion could not be drawn from their results A discussion of hereditary processes in bacteria must take into account the extensive work on transformation of pneumococcal types first described by Griffith 1928 and culminating in the isolation of the transforming principle in chemically characterizable form by Avery MacLeod and McCarty 1944 These studies have revealed that under special experimental conditions a product isolated from a serologically specific smooth pneumococcus culture will convert cells of a nonspecific rough culture to the smooth type characteristic of the source of the transforming prineiple So far as is known such transformations can be performed in only one direction rough to smooth and only under very special conditions Boivin and Vendrely 1946 have reported a similar tramnformation involving the capsular polysaccharide of a strain of E coli There have been reported other instances of varying credibility Kasarnowsky 1926 Lommel 1926 Legroux and Genevray 1933 Frobisher and Brown 1927 Burnet 1925 Holtman 1939 Cantacuzene and Bonciu 1926 These studies have a direct bearing on recombination experiments since transformations of this sort might be responsible for the occurrence in mixed cultures of some new types which are interpretable as recombination types This will be discused in more detail below Morphologically unusual forms of various bacteria have been described by Mellon 1925 as zygospores and been taken to imply a sexual fusion It has been suggested by Dienes and Smith 1944 and bySmith 1944 that the Large Bodies which appear under certain conditions in cultures of Bacteroides may represent a sexual phase There is as yet no evidence that the recombination phenomena in E coli which will be discussed in this paper are related to any special structural form such as those which have been described by these authors MATERIAL AND METHODS Except as otherwise stated the experiments reported on in this paper have all been performed with mutants of a single strain of Escherichia coli K 12 This is a typical coliform bacterium originally isolated from human feces a gram negative rod motile lactose fermenting producing indole and susceptible to each of the coli phages in the series Ti to T7 of Demerec and Fano 1945 1947 675 GENE RECOMBINATION IN BACTERIUM E COLI It has been used at Stanford University as a student laboratory strain for a number of years Mutant strains of E coli characterized by specific growth factor requirements have been obtained after treatment with X rays ultraviolet light and nitrogen mustard A single nutritional requirement is