MAR 1952 Reprinted from PROCEEDINGS OF THE SECOND NATIONAL CANCER CONFERENCE American Cancer Society, Ine. ° National Cancer [nstitute of the U. S. Public Health Service « American Association for Cancer Research SOME BIOLOGICAL ASPECTS OF BACTERIAL GENETICS (Original title: Microbial genetics in relation to cancer) Josuua LEpERBERG University of Wisconsin Madison Bacterial genetics has grown up so rapidly’ ** that it may be difficult to say what it is about and what it may have to do with the genetics or bi- ology of higher organisms, and their cells and tissues. Many kinds of problems and organisms have been studied, but there are some general principles that can be fairly inferred. Perhaps the most fundamental cleavage in the thinking of microbiologists has had to do with the impressive adaptive plasticity of bac- 1148 PROCEEDINGS OF THE SECOND NATIONAL CANCER CONFERENCE teria. I need only cite the rapidity with which some pathogens have caught up to our chemotherapeutic progress by the development of “drug-resistance’’*. Some workers, of whom the physical chemist, Sir Cyril Hinshelwood, is per- haps the most articulate spokesman have interpreted such adaptations as a direct chemical reaction of the bacteria to the drug. To the challenge that such adaptations pose a genetic, as well as a physiological problem, a new theory of heredity has been offered by which bacteria are sharply set off from other organisms, for which Lamarckian theories have long since been abandoned?®. Largely as a result of the biometric analysis of bacterial resistance to bac- teriophage by Luria and Delbruck® most biologists have reached a rather dif- ferent conclusion: that these adaptations result from a Darwinian struggle for survival, more precisely that a few individuals in the large populations tested are marked by random, pre-adaptive resistance. The spontancous variations or mutations occur sporadically and are inherited by their descendants in the absence of the drug. The mutants hold no significance for the natural history of the population unless and until the drug is encountered, whereupon the mutants will overgrow until some other ecological factor becomes limiting. By this theory, the most significant action of the drug is selective, not inductive. The selective theory has successfully accounted for every inherited adapta- tion? ®*, (This discussion is not applicable to the numerous instances of enzymatic adaptation that are characterized by a rapid attentuation after removal of the evoking substrate.) Unfortunately, the only means by which the supposedly spontaneous mutants could be detected has been by the intro- duction of the drug, so that a satisfactory (or at least a simple) experimental disqualification of inductive effects has been very difficult. This difficulty has been circumvented more recently by the development of a technique of in- direct selection. As the details are about to appear in print elsewhere’, they will be summarized only briefly. The development of resistance to streptomycin in Bacterium (Escherichia)coli will serve as an example. Indirect selection depends in turn on replica plating, which is a means of producing accurate replicas of the bacterial growth pattern on an initial agar plate. This is done simply by transferring the initial pattern to a sheet of vel- veteen fabric, which can then be used to imprint a series of fresh agar plates. If the imprints are made to streptomycin-agar, the development of resistant colonies will register the locations of resistant cells growing on an original, plain agar plate. That these cells are present in clones prior to exposure to the drug is signified by the recurrence of resistants at superimposable sites in a serics of replica plates. These recurrent sites, in turn, reveal the locations of the resistant clones on the original plate. Fortunately, the replica procedure does not dam- age the source plate, and part of the clone is left in its initial location, thus revealed. In practice, to find any mutants the initial plate must be so crowded with cells that a single stage of indirect selection does not permit the precise GENETICS PANEL 1149 localization of single, pre-adapted, resistant cells. However, the experimenter can locate these cells to the extent that an inoculum taken from the approxi- mate site is considerably enriched in the proportion of resistants (perhaps by a hundred-fold) . Then, by a reiteration of indirect selection, the resistant mutants can be isolated in pure culture. In this experiment, the role of the streptomycin cannot have been inductive, for the resistant cells isolated by indirect selection have never been exposed to streptomycin. The selective action of the drug has, however, permitted us to record the locations of resistant clones that could not otherwise be detected. This approach is, in principle and in practice, applicable to many other systems of bacterial adaptation. As already mentioned, the con- clusion is not new. The didactic value of the present argument is its principal justification. But the main purpose of this introductory divertissement is its illumination of the underlying philosophy of genetic microbiology, with its focus on the individual cell and its social behavior. A recurrent theme at this conference has been the desirability and achievement of successes in the treat- ment of tissues and tumors as populations of individual cells for physiological, cytological, virological and genetic analysis. The orientation to a microbiologi- cal approach is bound to be the more fruitful as methodological as well as analogical concepts are transferred. The selective isolation of pre-determined types is a fundamental element not only of mutation study, but in the investigation of genetic recombination and bacterial life cycles. The results have been strikingly different in different bac- teria. The most extensive experiments have been conducted with B. coli?» ™. We have not yet succeeded in detecting sexual processes in bacteria by direct microscopic observation, chiefly because the frequency with which it occurs 1s usually too low to encourage such an approach. We may hope that this barrier will soon be breached. Instead, we set up the question in genetic terms. Can we, in mixed bacterial cultures, detect the exchange of genetic factors that is the fundamental meaning of sexuality? To answer this question conclusively, we use different growth-factor-dependent or auxotrophic mutants, which are unable to form colonies on a basal medium owing to the lack of the specific growth factors required. If genetic recombination takes place, it should result in prototrophs, cells which have reacquired the wild type combination of genetic factors regulating nutrition. Since prototrophs can be isolated selectively by platings on minimal medium, this method is very powerful: a single proto- troph is readily detected in the presence of a billion auxotrophs. Somewhat to the surprise of E. L. Tatum and myself, this simple experiment worked the first time it was tried on B. coli, strain K-12. The appearance of prototrophs is the beginning, not the culmination of the analysis. From subsequent cxperiments, we have convinced ourselves! that the prototrophs are not artefacts resulting from spontaneous reverse-mutation of the parent mutants; (2) that the agent of recombination is the intact cell, and 1150 PROCEEDINGS OF THE SECOND NATIONAL CANCER CONFERENCE cannot be replaced with smaller units obtained by the extraction, disruption, or sedimentation of bacterial cultures; (3) that all, or almost all, of the charac- ters of the organism are involved in the recombination system; and (4) that the genes or factors controlling these characters are organized in a very definite way, entirely consistent with a linear-linkage chromosome system. The genetic evidence points to a life cycle similar to that of Neurospora, with haploid vege- tative cells and a transient diploid zygote. Exceptional cultures which persist in a diploid, heterozygous condition provide some of the strongest confirmation of this picture. These cultures can be subjected to rigorous single-cell pedigree isolations, following which the hybrid character of the single cells is confirmed by the segregation of the parental components. In its fundamental organiza- tion, therefore, this bacterial species closely resembles such higher forms as yeast, molds, algae and man. ° The next question that any naturalist would ask is certainly: “How about other strains and species?” After a time, methods were developed for rapidly screening new strains of B. coli for recombination. About two or three percent of some 2,000 strains tested have proved to be fertile; probably many others were not detected. As far as can be seen at present, these new fertile strains, which cross with the original and with each other, are constructed along the same lines as K-12. They do, however, differ in a number of characteristics of which the cellular antigens are perhaps the most promising for further research. For the question of the genetics of related species, the Salmonella group appeared to be a likely choice. Salmonella looks enough like B. coli that one might expect a similar story. In addition, we could depend on the same general methods for cultivation, obtaining and characterizing mutants, and so on. Technically, Salmonella typhimurium has, indeed, behaved very much like B. coli. It also has a recombination mechanism, but its details are almost diametrically opposed to those of B. coli! 8. In this species, genetic exchange is mediated by an agent much smaller than the cell, readily passing through bacteria-tight filters, and for this reason non- committally nicknamed “FA”—for filtrable agent. FA is most readily demon- strated in cultures that have been exposed to weakly lytic phages, but since every culture of S. typhimurium carries latent phages (lysogenicity) this is not a restrictive requirement. FA is probably associated with granules about 0.1 micron in diameter, casily seen in electron micrographs and barcly visible with dark-field microscopy. It has not yet been chemically characterized, but is re- markably resistant to heat, antiseptics, and a number of enzymes of which ribo- and desoxyribo-nuclease are the most significant. The genctic activity of Sal- monella FA also contrasts with recombination in B. colt. In a number of tests, involving altogether thirty or forty characters (mostly biochemical mutants, as in B. coli), no instance of associated exchange or linkage has been found. In fact, within the range of our experiments, each exchange has involved no more GENETICS PANEL 1151 than one character per bacterium. For this kind of transmission of hereditary fragments from one cell to another we have applied the term transduction. The limited scope of transduction does not prevent the synthesis of new bac- terial forms, and we have, for example, obtained serotypic “hybrids” of $. typhi with S. typhimurium, carrying the somatic antigen of the former with the flagellar characteristic of the latter. Many others of the ever-increasing number of Salmonella serotypes can be readily visualized as recombinants of existing forms. In Salmonella transduction, we are dealing with an agent squarcly athwart what we now classify as genes and viruses, respectively, and it would be tempt- ing to speculate further in this vein. In some other bacteria, genetic transduc- tion goes under the name of type transformation, for example, in the pneu- mococcus and the influenza bacterium. The present work has led to the further conclusion that, as far as could be told, the entire heredity of Salmonella typhi- murium is involved in transduction, if only one fragment at a time. It is dif- ficult to reconcile the experimental facts of transduction, with its fragmenta- tion of the genotype, to any plausible picture of genetic stability which seems to me to require something very similar to an integrated chromosomal system. Perhaps the FA particles correspond to fragmented pieces of chromosomes, as suggested for the pneumococcus by H. J. Muller**. Whatever its form, the resolution of this paradox cannot help but challenge our thinking about genes and our insight into their behavior in all organisms. REFERENCES 1, “Heredity and Variation in Microorganisms.” Cold Spring Harbor Symposia, Volume XI, 1946. 2. “Genes and Mutations.” Coid Spring Harbor Symposia, Volume XVI, 1951. 3. ‘Papers in Microbial Genetics: bacteria and bacterial viruses.” (Selected by J. Lederberg. ) 1951. University of Wisconsin Press, Madison. 4. Mirier, C, P. ann BouNnuorr, M.: 1950. The development of bacterial resistance to chemo- therapeutic agents. dun. Rev. Microbiol. 4: 201-222, 5. HINSHELWoop, C. N.: 1946. The chemical kinetics of the bacteria cell. Oxford University Press. 6. Luria, 5. E., anp DELBRUCK, M.: 1943. Mutations of bacteria from virus sensitivity to virus resistance. Genetics 28: 491-511 [Also in 3]. 7. Braun, W.: 1947. Bacterial dissociation. Bact. Rev. 11: 75-114. 8. Luria, S. E.: 1947, Recent advances in bacterial genetics. Bact. Rev. 11: 1-40. 9. STANiER, R. Y.: 1951. Enzymatic adaptation in bacteria. Ann. Rev. Microbiol. 5: 35-56. 10. Leperperc, J. AND LepERperc, E. M.: 1932. Replica plating and indirect selection of bac- terial mutants, J. Bact. 63: 399-406. lt. Tatum, E, L. anp Leprrserc, J.: 1947. Gene recombination in the bacterium Escherichia coli. J. Bact. 53: 673-684, 12. LEDERBERG, J., LepeRserc, E, M., Zinper, N. D. anp Livezy, E. R.: 1951. Recombination analysis of bacterial heredity. In 2. 13. Zinper, N. D. ano Leperwerc, J.: 1952. Genetic exchange in Salmonella. J. Bact. 64: 679-699. 14. Mutter, H. J.: 1947. The Gene. Proc. Roy. Soc. London, B, 134: 1-37.