Between 1947 and the mid-1950s, Joshua Lederberg and his collaborators in the Department of Genetics at the University of Wisconsin described a steady stream of important experimental techniques and results which transformed the science of bacterial genetics and helped define the classical era of molecular biology. Their most important discoveries were that of transduction--the transfer of genetic fragments from one cell to another by a virus--and of extra-chromosomal genetic particles called plasmids. Lederberg has invoked the excitement and enduring mysteries of the field at the time: "We were exploring a completely new territory that we only dimly understood. We weren't looking for transduction--we bumped into it. We weren't looking for plasmids--we bumped into them. Every time we turned around we found something unexpected." Yet, these serendipitous discoveries confirmed that bacterial cells differ in fundamental ways from cells in higher organisms, and laid the foundation for genetic engineering and modern biotechnology.

After Lederberg took up his first faculty position at the University of Wisconsin in 1947, he and the other members of his small laboratory, in particular his wife, microbiologist Esther Zimmer, and graduate students Norton Zinder and Larry Morse, expanded the systematic search for genetic recombination in bacteria by looking for it in Salmonella, a cousin of Escherichia coli which was of considerable medical interest because of its virulence. By the early 1950s they had pioneered methods for using penicillin and streptomycin to select for antibiotic resistance as an additional genetic marker in nutritional mutants. Streptomycin-resistance proved especially important because Lederberg was able to use it to quickly identify strains that were fertile and able to mate, until then a laborious procedure. Another important genetic marker isolated by Lederberg was that for Beta-galactosidase, a group of enzymes that enable bacteria to ferment the sugar lactose. This work presaged Jacque Monod's use of Beta-galactosidase some years later in formulating his theories on the mechanism of genetic expression and control in E. coli.

During the same period in the early 1950s, Lederberg and Zimmer developed a procedure, replica plating, which made possible the selection of mutants that were resistant to antibiotics or to bacterial viruses (called bacteriophages, or phages for short) without exposing them to the selective agent, the drug or the phage. They transferred impressions of a large number of bacterial colonies from a master plate, a circular glass dish where they had been cultured on growth media, to other, sterile plates (sterile meaning they contained not bacterial colonies of their own). For this they at first used blotting paper (numerous pieces of which can be found taped into his laboratory notebooks), then a beaker full of toothpicks, and finally velveteen cloth, which was pressed onto the colonies on the master plate, picking up members of colonies in their original spatial layout, which was then pressed onto the secondary plates, producing a copy, or replica, of the lawn of bacterial colonies on the master plate. The selective agent was spread on the secondary plates, which meant that only resistant colonies were able to grow on them. Their locations on the master plate could be inferred from their congruent location on the secondary plate. More colonies could be transferred from their locations of origin on the master plate, and the process repeated, until all resistant colonies on the master plate had been isolated.

Replica plating confirmed that drug or phage resistance in bacteria was the result of a genetic mutation, and not a physiological adaptation to the presence of the selective agent. It discredited Lamarckian interpretations, still held by some scientists including Nobel laureate Sir Cyril Hinshelwood, that characteristics in bacteria such as drug resistance were acquired under adverse environmental influences, for instance the presence of antibiotics, and were then imprinted into the bacterial genome. As a laboratory technique, replica plating became widely used for genetic studies of large populations of bacteria.

Most notably, Lederberg and Zinder in 1951 uncovered a third mechanism of genetic transfer in bacteria, in addition to the mechanisms of transformation, discovered by Oswald Avery, and of mating, discovered by Lederberg himself. Lederberg named it transduction, from the Latin transducere, to lead across. Lederberg and Zinder had observed conjugation in several nutritional and drug-resistant mutants of Salmonella. Now they wanted to undertake a reverse test to ascertain that their results indeed reflected genetic recombination. They used a glass tube, bent into a U-shape and fitted with an extremely fine glass filter at the crook, and filled it with broth containing none of the nutrients their Salmonella mutants needed to grow. They then added one mutant parent strain to each arm, and pumped the broth back and forth through the filter. They expected to find no recombinants, because as Bernard Davis had recently shown the year before by using the same U-shaped device, the bacterial cells in Lederberg's original crossing experiments had to be in direct contact with one another for conjugation to take place.

To their surprise, several bacteria nonetheless appeared and multiplied, an indication that recombination had occurred and had enabled each mutant strain to compensate for the nutritional deficiencies of the other. Obviously, a biological agent small enough to pass through the filter was at work. When they purified the agent, they found that it did not consist of pure deoxyribonucleic acid (DNA), Avery's transforming principle, or ribonucleic acid (RNA), because it was unaffected by enzymes that cleave these two nucleic acids. Instead, it proved to be a bacteriophage. The Salmonella Lederberg and Zinder had used turned out to be lysogenic: it had been infected many generations ago by a phage, whose viral DNA had been integrated into the chromosome of its bacterial host and had replicated along with the host chromosome. During the experiment the viral DNA had induced the synthesis of new phage within some of the Salmonella hosts, followed by the bursting, or lysis, of the host cell and the release of phage into the broth. Prior to lysis, the phage particles had picked up random snippets of host DNA, which they now carried across the filter (the phage were small enough to pass through) to the other arm of the test tube, there to infect other Salmonella bacteria and insert the fragments of bacterial DNA, along with their own viral genes, into the chromosomes of their new hosts. The phage had acted as a vector for bacterial genes. In this manner, Lederberg and Zinder concluded, the newly-infected mutant bacteria had acquired the genes that made up for their mutual genetic (nutritional) defects.

The discovery of transduction was of fundamental scientific significance. It foretold William Hayes' discovery in 1952 that recombination in prokaryotes, cells without a clearly defined nucleus and with only a short segment of double-stranded DNA as their genetic material (such as bacteria), was a fragmentary process, meaning that these cells did not merge their complete genomes but only fragments thereof. This was in contrast to eukaryotes, the more complex cells of higher organisms, which have a nucleus encapsulated by a membrane as well as chromosomes on which DNA is associated with various proteins in an intricate conformation. Eukaryotic cells reproduce by endowing their two daughter cells with full complements of the chromosomes of their parents. The differentiation between prokaryotes and eukaryotes, to which Lederberg's transduction experiment contributed in crucial ways, became one of the founding concepts of molecular biology.

Transduction was of equal consequence in medicine and biotechnology. It broadened conceptions about the composition, action, and biological significance of viruses. As an experimental technique, it helped researchers map bacterial chromosomes in fine detail. It explained how bacteria acquired genes that made them resistant to drugs much more quickly than the natural rate of genetic mutation and natural selection would allow: they simply picked up such genes from another strain by way of a viral medium.

Transduction also played an essential role in the development of genetic engineering. Although the term was unknown at the time, the fact that the Genetics Department in which Lederberg held tenure was part of the University of Wisconsin's College of Agriculture made him well aware of the many practical applications of genetic research. That transduction could be similarly used to manipulate genes did not escape him. During the 1970s scientists learned to direct bacterial viruses to pick up pieces of DNA with a known sequence of nucleotides (and thus with a known function) and insert them into a bacterial chromosome, where they functioned in the same manner as the host's genes. During the 1990s, the use of viral vectors to deliver healthy genes to cells lacking such genes became the basis of gene therapy. Lederberg himself advanced the development of the biotechnology industry in the United States from its beginning by serving as a consultant to the Syntex Corporation and the Cetus Corporations, commercial pioneers in the field, between 1961 and 1978.

A second important factor in the growth of biotechnology was Lederberg's discovery that bacteria contain ring-shaped, extrachromosomal pieces of DNA, which he named plasmids. Plasmids replicate independently of the chromosomes and transmit genes that specify functions not essential for cell growth. One such function is conjugation, which takes place only in bacterial strains carrying the so-called F plasmid. (The strain Lederberg first used in his mating experiments in 1945 lacked this plasmid.) Another is the production of enzymes, a property that would make plasmids an important tool for genetic engineering. During the 1970s, scientists succeeded in cleaving plasmids at specific locations with the help of so-called restriction enzymes, and in splicing into them foreign pieces of DNA, such as the gene for human insulin. Once closed up and reinserted into the bacterium, the altered plasmid directed the bacterium and its progeny (which inherited the engineered plasmid) to produce insulin for medical treatment, the first such drug created through genetic engineering.