ANNUAL REPORT 1951 Project No. 742. The nature and action of the gene in bacteria. Lederberg (Mra. E. M. Lederberg, Project Associate, Chemical Corps; Palmer D. Skaar, Project Associate, Research Committee and Rockefeller Foundation. Flise Cahn, Phyllis Fried, Ethelyn Lively and Norton D. Zinder, Research Assistants.) State Funds, W.A.R.F., Rockefeller Foundation, National Institutes of Health, Chemical Corps, and Atomic Energy Commission. 1. Formal genetics of Escherichia coli; effects of irradiation. 2. A new mechanism of genetic transmission in Salmonella typhimurium. 3. Genetic basis of drug resistance: replica-plating and indirect selection techniques. 4. Lysogenicity (confidential). AAS , 742 (Lederberg) 1 ae Pe. i The analysis of formal genetics of Escherichia coli is being continued. Special emphasis is being given to discrepancies between experimental data and the theory of a linear chromosome as the basis of heredity in this bacterium (as it is in higher forms). Single cell studies on segregating diploids suggest that some discrepancies may not be inherent in the actual genetic ratios, but in differential growth of different types. This work is necessarily slow, and definite contlusicns «iil not be available until it is completed. K study of the genetic effects of radiations on diploid E. coli has been initiated, with contract support from tne atomic Energy Com— mission. At this point, the project is very modest. The most signifi- cant finding to date is that single celis treated with UV give vrogenies with a variety of effects: either the cell (probably the nucleus as well) is not a unit target, or the effects of radiations persist as a biclogical disturbance despite the apparent growth-recovery of the cell. These effcts then later give rise to the variety of changes in different offspring. Skt Asolation of new strains of F. coli that can be crossed with existing fertile strains was a major subject this year. Narlier estinates were optimistic: of about 1500 straina tested, only 30, or 2% could be @rossed. These 30 strains are being studied closely for 2} their linkage patterns, b) the possibility of compatibility groups and c} characters of interest for genetic study. So far as can be determined so far, the new strains resemble the original FE. coli K-12 in their fundamental biology and life cycle. However, many of them are eublarally and serclogically distinct. We are now planning studies on the genetic basis of the natural (as opposed to laboratory-created) differences between the strains. The serological differences, in particular, permit of an immunogenetic analysis /4,2-Lederberg -3- L For a variety of reasons, bacteria provide excellent experimental material to study the biological bases of gene-controlled antigenic differences such as are the foundation of blood-typing in man, cattle, and other mammals. In addition, although FE. coli is "not pathogenic", the genetig basis of serotypes is of intrinsic interest to the student of infectious akasaase: The Researeh Committee provided funds (W.A.R.F.) to initiate this program ntal definitive support was secured from the Rockefeller Foundation, Dr. P. D. skaar has under- taken its immediate direction, and has completed preliminary experiments and preparation of antiseral reagents. 74,2- Lederberg -4- 2 Salmonella typhimurium (mouse-typhoid; enteric fever group) has been the subject of genetic study in this laboratory since 1947. During the past year, a reasonably clear picture of its behavior has developed for the first time. The Salmonella bacteria were chosen for this study because they provide good experimental material for studies on virulence, because they are readily cultured in the laboratory and are related to the BE, coli familiar to us, and because thelr serological characters (presumably under genic control) are impor- tant in public health bacteriology. Mr. Norton Zinder undertook this problem in 1948, and is using the experi- mental data for his Ph. D. dissertatdon (officially in Medical Microbiology). The plan was to parallel the work with E. coli. Biochemical mutants were made in a large number of strains, and attempts were made to detect crossing by plating mixed cultures on a selective, minimal agar medium. The results obtained were very confusing, as illustrated in previous reports, until it was realized that the recombination mechanism here is quite different from that in E. coli. In #. coli, all the evidence points to a typical sexual process, wherein recom- bination results from the fusion, and later segregation, of two intact nuclei. In Salmonella, under certain conditions, the cells release fragments of genetic material into the medium. These fragments may be either single "genes" or small aggregates, but much less than an entire nucleus. Other cells may absorb these fragments. The next steps we can only guess at, but an end result is that a "recombinant" cell is sometimes formed in which the absorbed fragment becomes a part of the genetic mechanism of its new host@cell. To distinguish this process from fertilization, which is the union of two essentially equal genotypes, a we have designated tnhts-process as genetic transduction. 7L2~Lederberg -5— 2 Transduction is an infectious process, if regarded in a certain light, and one may ask whether it does not merely depend on the transmission of a virus "disease" from one cell to another. However, a great many characters have been examined, and every one of them is transductb1 ¢fin the same way. These include many different nutritional requirements, fermentation differences, resistance Lo streptomycin, and type-diagnostic anthgans, the same groups of markers that are inherited as if carried on a chromosome in E. coli. The converse possibility, that genkc transduction may throw light on the origin of viruses, should be given close consideration. From the point of view cf adaptive plasticity, transduction is not so efficient as fertilization. In tale experiments, only a single factor is transducible to a ceil at one time, so that"crosses" of cells differing in many cactors result in only a smali iraction of ait the possibie gene com binations. The Saluojjel iss hit been subjucted to very close serological study, and a very large nuabar of antigenic types or “species*aks recognized in the diagnostic scheme. These types represent different combinations of somatic antigens (designated oy roman numerals) and flagellar antigens (arabic numerals and lower case letters). For example, S. tyvhimurium is desionated as IV,V,it1; 1; 1,2,3., while S. typhi is given as IX, XTI; d—. Recteriologists have often spedulated on the evolution of the Salmonells grown, and the origin of the different antigenic combinations, but owing to the lack of con- vincing precedents, recombination was not implicated. By transduction, however, a hybrid of 3. typhi x typhimurium has been obtained, with the antigenic forma Ix, X11; i —. This hybrid has not previously been described as a Salmonella type. If it had been isolated (and it might well be anticipated to occur in a patient suffering from a double infection) from a carrier or patient, it would certainly have been recognized as a new species. On this precedent, we may pre- 742~-Lederberg 2 -6- dict that! waaay Salmonella types have arisen, and will arise again, from the recombination of factors of previously established forms. Transduction is ao different from the hereditary patterns familiar to geneticists that it is difficult to reconcile it with the cytological observa- tions that suggest the presence of similar nuclear structures in E. coli and in Salmonella. We will have to learn a great deal more about genetic transd duction before we can evaluate its significance for our concepts of the nature of the gene and its relationship to the cell. 142-Lederberg 3 The newspapers rdeently have granhically described the continued race that we must anticipate in the development of new antibiotics for the chemo- therapy of infections caused by bacteria that have adapted to the old. The medical usefulness of the sulfonamides, penicillin, and streptomycin is al- ready limited, to varying degrees, by the prevalence of pathogens resistant to these agants. Tho developinent of drug-rasiatances can be demonatarzted in laboratory cuitures of bacteria exposed to a particular antibiotic. The mechantsa: of this alaptation will not immediately affect the outcome of this race, but is bound to enter into any longterm considerations. Gwing to expasinentel difficulties, a rather unegual contsoversy on the mechaniaa of this adaptation has persisted. Most workers, sspecially those with venetic interests, Lave concluded that sacil nuubexs of drug=resistant mutants ccctr soonlanevusiy. The vecurrence of these mutants has no relation to the presence of the drug. In lis absencs, however, they remain undetected and they come te visu only when the preponderant sensitive esils are eliminated by the chomothergpuutlc agent. this view 13 precisely anaiLogous to the Darwinian theory cf evolution by natural seiesction of rundos ciurnges. dhe alterhative opinion has held that the drug-resistant forms do not occur, even in small numbers. in the absence of the drug. the Llatie: aust liave seas direct effect on the mutation. Experiments to settle this question have been difficult to carry out beaaugs of the scarcity of the mutants. on unly ous call among its billion neighbours is resistant, the only means of detecting lt had been to expose the entire population to the drug, and this does not help to decide the controversy. However, a considerable body of indirect evidence has accumulated in favor of the "Darwinian" theory. : 742-Lederberg ~8- 3 In the course of studies unrelated to this problem, a method was devloped for handladg large numbers of bacterial colonies with a minimim of effort, and this later provided a decisive answer to the problem of drug-resistance. Much of the laboratory work in microbial genetics (and other aspects of microbiology) involves "picking" each of a great many bacterial colonies grewing on an agar medium, and transferring them with a platinum wire needle to several other types of medium to see how they will perform on these. In our laboratory, the number of times this operation is carried out is numbered in the hundreds or thousands per day. A method was dewhoped so that an entire plate, containing up to several hundred colonies, could be handled as a unit. A dise of a pile fabric, such as velvet or velveteen was used as a sort of template or stencil. When the fabric is gently pressed onto an agar plate, the colonies become entangled in the hairs of the nap. When the template is then pressed on one ore more fresh plates, a few cells are depseited from the velvet, in positions correspon- ding to the colonies on the original plate. In effect, it becomes possible to "copy" a pattern of growth on one agar plate to a series of others. With care, the copies are quite faithful (Figures 1 to 3). This technique, which 4s called replica-plating, has been quite useful in the detection and caassi- fication of mutants of bacteria and of actinomycetes, also. It should save a great deal of work in similar operations, such as testing groups of bacteria for their reactions to various antibiotics. It was only after the replica method was well developed that we realized its applicability to the problem of adaptation to antihisbics. Two lines of pertinent evidence arose. If a suspension of bacteria not previously exposed to a drug (streptomycin was used in our experiments) is spread on a plain agar plate, we should expect that a few of the cells placed on the plate are already drug-resistant, if we reason from the "Darwinian" hypothesis. If the plate is 742-Lederberg 3 - incubated for a few hours, so that the bacteria grow undisturbed, the original resistant cells will have mitiplied several times to give a clone or asexually produced family of resistant cells at that place. This was verified by making copies of such a plate to a series of other plates containing the drug. In almost every case, the copies showed resistant cells developing at corresponding places, showing that the original plate mist have had the resistant cells in families. The resistant cells must have developed the families on the plain agar before they were exposed to the drug on the copy plates. This is contradictory to the theory of a direct effect of the drug. An even more convincing result, however, is a procedure that allowa us to isolate the resistant cells without ever exposing them directly to the drug. In the previous experiment, families of resistant cells are located on the plain agar at places corresponding to the resistant colonies developing on the copies. When the velvet template is prepared, it does not remove most of the cells on the original plate. The drug-containing copy plates tell us where to look for resistant mitants on the original plate. This plate is too crowded to allow us to tell exactly which cells correspond to the resistant families, but the method narrows the range: instead of looking for over an area of about five square inches (a whole plate), we can lpcate the clone in an area of about .05 square inches. By using these cells as an inoculum for a fresh plate, we will have concentrated the resistant cells about one-hundred fold, say from a proportion of 1 : 1,000,000 to 1: 10,000. This "enrichment" can be repeated several times, until the resistant cells are well separated on the plate, and can be picked easily. In all of this, the actual selection line remains on plain agar, and never is exposed to the drug. The sibs #Hat are carried over to make the replica plates 74,2-Lederberg 3 -1LQ— serve to tell us where to look for the resistant mutants. Thus we can practise our selection indirectly, and aan isolate the resistant mutahts without ever exposing them to the drug they are adapted to. This is rather like pedigree—selection of roosters for hreeding stock. We can use information on egg—production records of the sib hens of a rooster flock to select the birds with the best genotypes, without ever demanding that the rooster prove his potentialities directly. This work was done in collaboration with Dr, E. M. Lederberg, and will appear in the January 1952 issuenof the Journal of Bacteriology.