THE GENETICS OF SALMONELLA RG=1445 (2) National Institute of Health, U.S. Public Health Service, Bethesda, Md. Third Annual Progress Report, submitted April 20, 1951. Joshua Lederberg Associate Professor of Genetics University of Wisconsin, Madison. SUMMARY Current work suggests that, under certain conditions, Salmonella cul= tures form reduced cells whieh a) readily pess filters retaining the usual bacteria; b) are more resistant to disinfection (heat, alcohol, chloroforn); end ¢) remain dormant in the absence of a stimulus from living cells. If verified, these findings, might require a reexamination of concepts of bacteriological sterility. This result wes unexpectedly encountered in connection with work on the mechanism of genetic recombination, Such a process, analogous to that established in Escherichia coli K-12, seems to occur in some strains of S. typhimurium, Whether there is a specific connection between the filtrable agent and genetic recombinetion, as appears possible, is the subject of current experiments, a THE GENETICS OF SALMONELLA RG=1445 (c2) Third Annual Progress Report, submitted April 20, 1951. Recapitulation and Introduction This research project on genetic aspects of the biology of Salmonella is directly related to previous and concurrent studies on genetic recombi~ nation in the releted group of coliform bacteria, It will help to out-— line progress and aims of the Salmonella project if the results from Escher ichiea coli are briefly recapitulated. About five years ago, E. Le Tatum and the undersigned (at Yale University), discovered thet genetic recombination occurs among cells of E. coli strain K-12 (1,2). This was demonstrated by mixing cells from different nutrition ally execting or auxotrophic mutant cultures on a synthetic agar mediun, By suppressing the auxotrophic parents, this medium selects for a small pro- portion of cells which have become prototrophic, i.e., able to form colonies in the absence of supplementery growth factors, With appropriately devel~ oped mutant stocks it was possible to rule out the possibility thet these prototrophs erise from intrinsic instability of the mtant parents; they could erise only from the interaction of the distinct mutants: e.g., Ab with aB give rise to AB. This type of recombinetion process immedictely suggested the possibility of a sexual process in this bacterium, This pos— sibility has been reinforced by a number of experiments which showed the following: , 1) Recombination was not confined to nutritional factors. Many other genetic differences introduced with the parents (e.g., fermentetion characters, virus resistance, drug resistance) reassort in all possible combinetions among the prototronhs. 2) Recombination occurs pairwise. In mixtures of three kinds of mtants, only those recombinants occur which could arise from pairwise exchanges, whereas uniquely tri-partite exchanges are not found. 3) In certain crosses, cultures ("heterozygous diploids") have been isoleted which cerry the genetic fectors from both perents, which may later segregete during the further proliferation of the culture (3). Single cell studies showed conclusively thet this sesregetion involves the separation of intra-celluler units, not entire single cells (4). 4) iumberous attempts to effect genetic exchanges by means of culture filtretes, cell extracts, or other preparations not containing normal viable cells of both nerents have failed completely. This requirement for intact cells from both parents supports the concept that the fusion of ordinary vegetative bacterial cells is the besis of genic conjunction (as in many other microbes) end thet no special gametic forms need be invoked, However, the actual fusion of cells has not been observed in this material, so that our eonclusions on this detail of the "sexual" process are entirely infer=— ential, -~2- Until recently, evidence for gene recombination in bacteria was confined to strain K-12 of BE, cold. Cavalli, working et Cambridge, Ingland, has discovered e strain there which can be crossed with K-12 (5) and more re~ cently a considereble number of strains heve been isolated (about 3% of a series of 500 tested) which also show this phenomenon (6). Many of these new "crossable" strains are quite different from K-12, some being classi- fiable as peracolon or as coliform intermedietes. This result makes it all the more necessary and hopeful to scrutinize other bacteria for similar genetic processes. Two objectives ere preeminent: 2) to provide the besis for genetic analysis of problems uniqué to other bacterial groups, 6.Z-— antigenic and pethogenic variation in Salmonella, and b) to find better material for the determination of the mechanisms, scope, end ecological role of recombination. Experimental Results with Salmonella Considerable time was spent during the first two years in collecting suitable cultures, developing techniques for producing biochemical muta- tions in Salmonella (7), and in perfecting the training of the research assistant assigned to this problem (Mr. Norton D. Zinder). After an inter- val during wiich mitants were induced in a diversity of types, including S. typhimurium, poona, madelia, "coli", end others, it wes decided to con= centrate on a coherent set of cultures of S, typhimuriun, Such a set was provided by Dr. S, Lilleeungen (Stockholm) who had worked out 2 procedure for bacteriophage tyning of this species. He kindly placed at our disposal representative cultures of each of his 22 types of world- wide origin. In this way we could be feirly sure of covering a compre— hensive sample of S. typhimurium cultures without wsanecessarily reduplicating our worl, Auxotrophic mutants heve been induced in 20 of these types. The present meterial has allowed crossing tests (like those in Z coli K-12) to be mede in ebout half of the 200 possible combinations (inter— and intra=strein crosses); the mutants required to complete all of the possible combinetions ere being produced from day to day. It should be pointed out thet, despite technical advances, the production and cheracterizaticn of at least tio double auxotrophic miutahts in eech of 20 strains represents 2 considerable maltiplication of effort and hes occupied the larger pert of the time svent on this project to date, Of the 99 combinetions so fer tested, 9 heve more or less consistently given prototrophs on minimal agar medium, while the parents seperetely do not form colonies under these conditions. This is preliminary presumptive evidence for recombination in Selmonella typhimurium, Some of the combi- nations heve given very low yields of prototrophs, a result which hinders the further study of the mechanisms of the anperent recombinetion. Our effort has been focussed on one perticuler combinetion which proved to be excentionally fertile, Culture "A" is a mutant derived in two stens from Lilleeungen's type 2, "B" from type 22. "A" requires histidine end methionine "Bl phenylalanine "plus* tyrosine, and tryptophane. Cultures of A or of B by themselves have never been found to produce prototrophs, even when very dense suspensions were pleted on minimal ager. However, mixtures of A with -~3- B heve given yields of prototrophs of at least 1075 of the perental inoculun, (considerably higher than has been found in He coli), Since this combina~ tion gives the highest yield of presumably recombinant prototrophs, it was selected for further study. Two directions are being followed a) the genetic rules of recombination, and b) its biological mechanism, With respect to a) certain peculiarities have been noted already.