The Development of Bacterial Genetics

Between 1947 and 1952, Joshua Lederberg and his small lab group at the University of Wisconsin significantly reshaped the field of bacterial genetics. By showing that certain strains of bacteria reproduce by mating--combining their genetic material--he overturned prevailing assumptions among scientists that bacteria were primitive organisms not suitable for genetic analysis. Rather, he demonstrated that bacteria could serve as a powerful experimental system, with widespread application in genetic research. His laboratory findings gave crucial insights into the chemical mechanisms of gene action and helped explain the evolution and adaptation of microorganisms.

Lederberg was inspired to study bacterial genetics by Oswald Avery, Colin MacLeod, and Maclyn McCarty's seminal 1944 paper identifying deoxyribonucleic acid (DNA) as the "transforming principle," the genetic material, in Pneumococcus bacteria. Avery had set out to examine the phenomenon that non-infectious strains of the pneumonia-causing bacteria became virulent--were transformed--when mixed with heat-killed and ground-up infectious strains. His painstaking investigation revealed that the chemical substance, or "principle," responsible for the transformation from non-virulence to virulence was DNA, which had been released by the killed virulent Pneumococcus bacteria and which had been picked up by the non-infectious ones. After they were integrated into the genome of their new hosts, these segments of DNA directed the synthesis of a smooth outer polysaccharide shell that made the previously non-virulent bacteria infectious, a trait they passed on to their progeny.

Lederberg thought Avery's findings "the most exciting key" to uncovering the chemical nature of the gene. He promptly decided to look for DNA-mediated transformation in Neurospora, the red bread mold that his mentor, the biochemist Francis J. Ryan, had brought with him to Columbia from his postdoctoral studies at Stanford with George W. Beadle and Edward L. Tatum. Beadle and Tatum had used nutritional mutants of Neurospora, mutants that were genetically blocked in the synthesis of growth factors such as a particular amino acid or vitamin, to reach their famous "one gene, one enzyme" hypothesis. Their hypothesis stated that genes were specific sections of the genetic material that direct the synthesis of particular proteins. Adopting their method for his own purpose, Lederberg used Neurospora mutants unable to synthesize, or make, the amino acid leucine in an attempt to induce transformation.

His experiment failed because the mutants, at first unable to grow unless leucine was provided in the jelly in which they were cultured, spontaneously reverted to their ancestral, or wild-type, condition. They could once again grow on culture medium containing no leucine because they had regained the ability to synthesize their own. Lederberg was able to show that this phenomenon was caused by a reverse mutation, and that the mutation and the reversion were allelic, meaning that they occurred in the same place on the Neurospora chromosome. These results led to Lederberg's first published article, written with Ryan. In it Lederberg gave revertants that had the same nutritional requirements as the ancestral strain the name prototroph, now widely used in bacterial genetics. More importantly, their paper introduced the prototrophic recovery technique, the method by which prototrophs--nutritional mutants that had regained their ability to make particular amino acids or vitamins themselves--could be picked out, or recovered, from among billions of other mutants because only they were able to grow on the culture medium provided. The prototrophic recovery technique was an important innovation because it allowed detection of even extremely rare genetic changes in microorganisms through the experimentally straightforward use of selective growth media. It was to become crucial to Lederberg's subsequent research, as well as to bacterial genetics generally.

Far from being discouraged by his failure to find transformation in Neurospora, Lederberg pursued a new ambition. He conceived of the possibility of using the prototrophic recovery technique to look for genetic recombination in bacteria, to test whether bacteria mated and were thus susceptible to genetic study. The conventional wisdom, expounded in Lederberg's medical school courses, held that since bacteria were primitive organisms which reproduced by dividing into two genetically identical daughter cells, it was impossible to study them with the crossing experiments and comparative analysis of classical genetics. Yet, this assumption had never been subjected to rigorous testing in the laboratory, as René J. Dubos reported in his comprehensive survey of bacteriology, published in 1945. Moreover, Avery's research had suggested that bacteria were genetically more complex than scientists had previously believed, that they had clearly defined genes and chromosomes not unlike those of higher organisms, such as Neurospora or the malaria parasites Lederberg had studied during his wartime service as a hospital corpsman. Lederberg thus set out to determine whether bacteria display sexual behavior, that is, reproduce by a recombination of their genes.

To try out his ideas, Lederberg isolated two nutritional mutants of Escherichia coli, a generally benign bacterium that resides in the human colon. The mutants were incapable of synthesizing the amino acids methionine and proline, respectively, which meant that they could not grow unless these were supplied in the medium. Lederberg carried out crossing experiments in search of recombinants that were able to grow on a medium containing neither of the two amino acids. Finding even one such recombinant among billions of cells would be evidence of mating, and would therefore show that bacteria were susceptible to genetic analysis.

As it turned out, the strain of E. coli from which Lederberg obtained the mutants was sterile, and his crossing experiments were unsuccessful. To help his student break the impasse, Ryan suggested a collaboration with his own former mentor, Edward Tatum, who had recently moved from Stanford to Yale University, so that Lederberg could take advantage of Tatum's expertise as a bacteriologist and of his stock of E. coli mutants.

Once Lederberg joined Tatum in late March 1946, he quickly resumed what he only later realized was a "long-shot experiment" to find a sexual stage in the life cycle of bacteria. This time he used Tatum's K12 strain of E. coli, a strain different from the one he had used previously. The choice would prove fortunate. He took two of Tatum's double mutants of the K12 strain, each unable to synthesize two different nutrients (the vitamin biotin and the amino acid methionine in the first mutant, the amino acids threonine and proline in the second), to minimize the possibility of reversion to prototrophy, which in double mutants is extremely rare. When he crossed the two double mutants, Lederberg discovered that some of their progeny regained the ability to synthesize the two respective nutrients which previously had to be supplied in the broth for them to grow, and that this ability was inherited by succeeding generations. Such prototrophs were obtained only when the two mutants were mixed, not when single strains were incubated separately. Moreover, all individual bacteria within a colony of recombinants had the same genotype, or genetic constitution.

Furthermore, from among the double mutants Lederberg isolated several that in addition to having the nutritional requirements described above were resistant to a bacteriophage, a bacterial virus, that infected E. coli. Resistance to bacteriophage T1 provided a second, so-called unselected genetic marker in this sub-group of nutritional mutants. When Lederberg tested prototrophs for T1 resistance, he found that some were resistant while others remained sensitive to the virus. Most importantly, he was able to demonstrate that the ratio of resistance to sensitivity depended on which parent carried the resistance marker, and that this ratio was reversed in reciprocal crosses (i.e. when instead of parent strain A, parent strain B carried this marker).

Taken together, these findings were evidence that the two parent mutants had mated and exchanged genes, a process Lederberg called conjugation. Moreover, his experiments demonstrated that E. coli was haploid, meaning that it possessed only one complete set of chromosomes. In fact, Lederberg concluded that E. coli carried only a single chromosome. This was in contrast to the cells of higher organisms, which are diploid and possess two complete complements of chromosomes. Lederberg completed these crossing experiments in about six weeks, in time to present the results to a gathering of the world's leading microbial geneticists at the Cold Spring Harbor Conference on the Genetics of Microorganisms in July 1946.

The very low incidence of mating, together with his technique of studying the genetic products of bacterial conjugation (namely, inherited nutritional requirements and drug resistance) in culture medium, prevented Lederberg from examining the physiology and kinetics, the stages and dynamics, of the mating process in detail. What actually happened on the cellular level during mating remained unknown. When new ways of photographing the stages of the mating process with the electron microscope were developed in the 1950s, they revealed that during conjugation the two bacteria of the mating pair lay side by side and formed a connecting bridge through which DNA is passed. This confirmed the finding by Bernard Davis in 1950 that for recombination to occur, the two parent bacteria had to have direct cell-to-cell contact. (Davis' finding also ruled out the possibility that conjugation in E. coli was a form of transformation by lose strands of DNA akin to that described by Avery in Pneumococcus, which did not require direct contact between cells.)

Lederberg's experimental protocol, which prescribed the use of nutritional requirements both as genetic markers and to select mutants against parental types, was as simple as it was innovative. It also involved an element of luck, since it depended on the use of a bacterial strain that was fertile. With the benefit of hindsight, Lederberg estimated that the chance that mutants of the E. coli K12 strain he used were fertile was less than one in twenty. Yet, within a very short period of time he was able to disprove long-held assumptions about the absence of sex in bacteria. He thus made an experimental system available to geneticists that, compared to established experimental organisms in genetics like the fruit fly and maize, had the advantage of physiological simplicity and of rapid growth.

Between 1947 and 1952, Lederberg continued to work on a linkage map the E. coli K12 chromosome, exploiting the fact that in recombination, unrelated genes are functionally linked to show where these genes were located on the chromosome in relative distance to one another. His efforts grew to "rabbinical complexity," as James Watson wrote in The Double Helix (1968), for the very reason that Lederberg continued to adhere to his analogy between bacterial cells and the cells of higher organisms. This analogy suggested that in the process of mating, the two parental bacteria merged their entire genetic resources, as is the case in higher organisms. The complexities in Lederberg's linkage map could not be resolved until the Irish bacteriologist William Hayes reported in 1952 that during conjugation, one partner of the mating pair, the male, acted as a donor, and that the other, the female, acted as recipient. Furthermore, Hayes showed that the male donor transferred only part of its genes to the female recipient; complete genomes were not merged in the same cell. Both results ran counter to Lederberg's assumptions about the nature of the bacterial cell, and both proved essential to understanding the mechanism of sexual recombination in bacteria in molecular detail. More generally, Hayes' discovery created an understanding that with regard to the mating process and many structural and physiological characteristics, bacterial cells are fundamentally different from the cells of higher organisms. Since 1962, these differences have been denoted by the terms prokaryotic and eukaryotic cells.