NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Griffiths AJF, Miller JH, Suzuki DT, et al. An Introduction to Genetic Analysis. 7th edition. New York: W. H. Freeman; 2000.

  • By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.
Cover of An Introduction to Genetic Analysis

An Introduction to Genetic Analysis. 7th edition.

Show details


Some phages are able to “mobilize” bacterial genes and carry them from one bacterial cell to another through the process of transduction. Thus, transduction joins the battery of modes of genetic transfer in bacteria—along with conjugation, infectious transfer of episomes, and transformation.

Discovery of transduction

In 1951, Joshua Lederberg and Norton Zinder were testing for recombination in the bacterium Salmonella typhimurium by using the techniques that had been successful with E. coli. The researchers used two different strains: one was phetrptyr, and the other was methis. (We won’t worry about the nature of these markers except to note that the mutant alleles confer nutritional requirements.) When either strain was plated on a minimal medium, no wild-type cells were observed. However, after the two strains were mixed, wild-type cells appeared at a frequency of about 1 in 105. Thus far, the situation seems similar to that for recombination in E. coli.

However, in this case, the researchers also recovered recombinants from a U-tube experiment, in which cell contact (conjugation) was prevented by a filter separating the two arms. By varying the size of the pores in the filter, they found that the agent responsible for recombination was about the size of the virus P22, a known temperate phage of Salmonella. Further studies supported the suggestion that the vector of recombination is indeed P22. The filterable agent and P22 are identical in properties of size, sensitivity to antiserum, and immunity to hydrolytic enzymes. Thus, Lederberg and Zinder, instead of confirming conjugation in Salmonella, had discovered a new type of gene transfer mediated by a virus. They called this process transduction. In the lytic cycle, some virus particles somehow pick up bacterial genes that are then transferred to another host, where the virus inserts its contents. Transduction has subsequently been shown to be quite common among both temperate and virulent phages.

There are two kinds of transduction: generalized and specialized. Generalized transducing phages can carry any part of the chromosome, whereas specialized transducing phages carry only restricted parts of the bacterial chromosome.

Transducing phages and generalized transduction

How are transducing phages produced? In 1965, K. Ikeda and J. Tomizawa threw light on this question in some experiments on the temperate E. coli phage P1. They found that, when a donor cell is lysed by P1, the bacterial chromosome is broken up into small pieces. Occasionally, the forming phage particles mistakenly incorporate a piece of the bacterial DNA into a phage head in place of phage DNA. This event is the origin of the transducing phage.

Because the phage coat proteins determine a phage’s ability to attack a cell, transducing phages can bind to a bacterial cell and inject their contents, which now happen to be donor bacterial genes. When a transducing phage injects its contents into a recipient cell, a merodiploid situation is created in which the transduced bacterial genes can be incorporated by recombination (Figure 7-26). Because any of the host markers can be transduced, this type of transduction is termed generalized transduction.

Figure 7-26. The mechanism of generalized transduction.

Figure 7-26

The mechanism of generalized transduction. In reality, only a very small minority of phage progeny (1 in 10,000) carries donor genes.

Phages P1 and P22 both belong to a phage group that shows generalized transduction (that is, they transfer virtually any gene of the host chromosome). During their cycles, P22 probably inserts into the host chromosome, whereas P1 remains free, like a large plasmid. But both transduce by faulty head stuffing in lysis.

Linkage data from transduction

Generalized transduction allows us to derive linkage information about bacterial genes when markers are close enough that the phage can pick them up and transduce them in a single piece of DNA. For example, suppose that we wanted to find the linkage between met and arg in E. coli. We might set up a cross of a met+arg+ strain with a metarg strain. We could grow phage P1 on the donor met+arg+ strain, allow P1 to infect the metarg strain, and select for met+ colonies. Then, we could note the percentage of met+ colonies that became arg+. Strains transduced to both met+ and arg+ are called cotransductants.

Linkage values are usually expressed as cotransduction frequencies (Figure 7-27). The greater the cotransduction frequency, the closer two genetic markers are.

Figure 7-27. Genetic map of the purB- to-cysB region of E.

Figure 7-27

Genetic map of the purB- to-cysB region of E. coli determined by P1 cotransduction. The numbers given are the averages in percent for cotransduction frequencies obtained in several experiments. Where transduction crosses were performed in both directions, (more...)

Using an extension of this approach, we can estimate the size of the piece of host chromosome that a phage can pick up. The following type of experiment uses P1 phage:

Image ch7e7.jpg

We can select for one or more donor markers in the recipient and then (in true merozygous genetics style) look for the presence of the other unselected markers, as outlined in Table 7-5. Experiment 1 in Table 7-5 tells us that leu is relatively close to azi and distant from thr, leaving us with two possibilities:

Image ch7e8.jpg

Table 7-5. Accompanying Markers in Specific P1 Transductions.

Table 7-5

Accompanying Markers in Specific P1 Transductions.

Experiment 2 tells us that leu is closer to thr than azi is, so the map must be:

Image ch7e9.jpg

By selecting for thr+ and leu+ in the transducing phages in experiment 3, we see that the transduced piece of genetic material never includes the azi locus.

If enough markers were studied to produce a more complete linkage map, we could estimate the size of a transduced segment. Such experiments indicate that P1 cotransduction occurs within approximately 1.5 minutes of the E. coli chromosome map (1 minute equals the length of chromosome transferred by an Hfr in 1 minute’s time at 37°C).


In the 1920s, long before E. coli became the favorite organism of microbial geneticists, some interesting results were obtained in the study of phage infections of E. coli. Some bacterial strains were found to be resistant to infection by certain phages, but these resistant bacteria caused lysis of nonresistant bacteria when the two bacterial strains were mixed together. The resistant bacteria that induced lysis in other cells were said to be lysogenic bacteria or lyso-gens. When non-lysogenic bacteria were infected with phages derived from a lysogenic strain, a small fraction of the infected cells did not lyse but instead became lysogenic themselves.

Apparently, the lysogenic bacteria could somehow “carry” the phages while remaining immune to their lysing action. Initially, little attention was paid to this phenomenon after some studies seemed to show that the lysogenic bacteria were simply contaminated with external phages that could be removed by careful purification. However, in the mid-1940s, André Lwoff examined lysogenic strains of Bacillus megaterium and followed the behavior of a lysogenic strain through many cell divisions. Carefully observing his culture, he separated each pair of daughter cells immediately after division. One cell was put into a culture; the other was observed until it divided. In this way, Lwoff obtained 19 cultures representing 19 generations (19 consecutive cell divisions). All 19 cultures were lysogenic, but tests of the medium showed no free phages at any time during these divisions, thereby confirming that lysogenic behavior is a character that persists through reproduction in the absence of free phages.

On rare occasions, Lwoff observed spontaneous lysis in his cultures. When the medium was spread on a lawn of nonlysogenic cells after one of these spontaneous lyses, plaques appeared, showing that free phages had been released in the lysis. Lwoff was able to propose a hypothesis to explain all his observations: each bacterium of the lysogenic strain contains a noninfective factor that is passed from bacterial generation to generation, but this factor occasionally gives rise to the production of infective phages (without the presence of free phages in the medium). Lwoff called this factor the prophage because it somehow seemed to be able to induce the formation of a “litter” of infective phages. Later studies showed that a variety of agents, such as ultraviolet light or certain chemicals, could activate the prophage, inducing lysis and infective phage release in a large fraction of a population of lysogenic bacteria.

We now know exactly how Lwoff’s observations occur. A lysogenic bacterium contains a prophage, which somehow protects the cell against additional infection, or superinfection, from free phages and which is duplicated and passed on to daughter cells in division. In a small fraction of the lysogenic cells, the prophage is induced, or activated, producing infective phages. This process robs the cell of its protection against the phage; it lyses and releases infective phages into the medium, thus infecting any nonlysogenic cells present in the culture.

Phages can be categorized into two types. Virulent phages have an infectious cycle that is always lytic—for these phages, there are no lysogenic bacteria. (Resistant bacterial mutants may exist for virulent phages, but their resistance is not due to lysogeny.) Temperate phages follow a lytic cycle under some circumstances, but they usually initiate a lysogenic cycle, in which the phage exists as a prophage within the bacterial cell. In this case, the lysogenic bacterium becomes resistant to superinfection, an “immunity” conferred by the presence of the prophage, which is transmitted genetically through many bacterial generations. Temperate phages also cause lysis when the prophage is induced, or activated. Figure 7-28 diagrams the lytic and lysogenic infectious cycles of a typical temperate phage.

Figure 7-28. Alternative cell cycles of a temperate phage and its host.

Figure 7-28

Alternative cell cycles of a temperate phage and its host. (After A. Lwoff, Bacteriological Reviews 17, 1953, 269.)


Virulent phages cannot become prophages; they are always lytic. Temperate phages can exist within the bacterial cell as prophages, allowing their hosts to survive as lysogenic bacteria; they are also capable of direct bacterial lysis.

Genetic basis of lysogeny

What is the nature of the prophage? On induction, the prophage is capable of directing the production of a complete mature phage, so all of the phage genome must be present in the prophage. But is the prophage a small particle free in the bacterial cytoplasm—a plasmid—or is it somehow associated with the bacterial genome? Fortuitously, the original strain of E. coli used by Lederberg and Tatum (page 209) proved to be lysogenic for a temperate phage called lambda (λ). Phage λ has become the most intensively studied and best-characterized phage. Crosses between F+ and F cells have yielded interesting results. It turns out that F+ × F(λ) crosses yield recombinant lysogenic recipients, whereas the reciprocal cross F+(λ) × F almost never gives lysogenic recombinants.

These results became more understandable when Hfr strains were discovered. In the cross Hfr × F(λ), lysogenic F exconjugants with Hfr genes are readily recovered. However, in the reciprocal cross Hfr(λ) × F, the early genes from the Hfr chromosome are recovered among the exconjugants, but recombinants for late markers (those expected to transfer after a certain time in mating) are not recovered. Furthermore, lysogenic exconjugants are almost never recovered from this reciprocal cross. What is the explanation? The observations make sense if the λ prophage is behaving like a bacterial gene locus (that is, like part of the bacterial chromosome). In interrupted-mating experiments, the λ prophage always enters the F cell at a specific time, closely linked to the gal locus. Thus, we can assign the λ prophage to a specific locus next to the gal region.

In the cross of a lysogenic Hfr with a nonlysogenic (nonimmune) F recipient, the entry of the λ prophage into the nonimmune cell immediately triggers the prophage into a lytic cycle; this process is called zygotic induction. But, in the cross Hfr(λ) × F(λ), any recombinants are readily recovered; that is, no induction of the prophage, and consequently lysis, occurs (Figure 7-29). It would seem that the cytoplasm of the F cell must exist in two different states (depending on whether the cell contains a λ prophage), so contact between an entering prophage and the cytoplasm of a nonimmune cell immediately induces the lytic cycle. We now know that a cytoplasmic factor specified by the prophage represses the multiplication of the virus. Entry of the prophage into a nonlysogenic environment immediately dilutes this repressing factor, and therefore the virus reproduces. But, if the virus specifies the repressing factor, then why doesn’t the virus shut itself off again? Clearly it does, because a fraction of infected cells do become lysogenic. There is a race between the λ gene signals for reproduction and those specifying a shutdown. The model of a phagedirected cytoplasmic repressor nicely explains the immunity of the lysogenic bacteria, because any superinfecting phage would immediately encounter a repressor and be inactivated. We present this model in more detail in Chapter 11.

Figure 7-29. Zygotic induction.

Figure 7-29

Zygotic induction.

Prophage attachment

How is the prophage attached to the bacterial genome? Allan Campbell proposed in 1962 that λ attaches to the bacterial chromosome by a reciprocal crossover between the circular λ chromosome and the circular E. coli chromosome, as shown in Figure 7-30. The crossover point would be between a specific site in λ, the λ attachment site, and a site in the bacterial chromosome located between the genes gal and bio, because λ integrates at that position in the E. coli chromosome.

Figure 7-30. Campbell’s model for the integration of phage λ into the E.

Figure 7-30

Campbell’s model for the integration of phage λ into the E. coli chromosome. Reciprocal recombination takes place between a specific attachment site on the circular λ DNA and a specific region on the bacterial chromosome between (more...)

One attraction of Campbell’s proposal is that it allows predictions that geneticists can test by using phage λ:


Integration of the prophage into the E. coli chromosome should increase the genetic distance between flanking bacterial markers, as can be seen in Figure 7-30 for gal and bio. In fact, studies show that time-of-entry or recombination distances between the bacterial genes are increased by lysogeny.


Deleting bacterial segments adjacent to the prophage site should delete phage genes at least some of the time. Experimental studies also confirm this prediction.

Specialized transduction

We can now understand the process of specialized transduction, in which only certain host markers can be transduced.

Lambda is a good example of a specialized transducing phage. As a prophage, λ always inserts between the gal region and the bio region of the host chromosome (see Figure 7-30). In transduction experiments, λ can transduce only the gal and bio genes. Let’s visualize the mechanism of λ transduction.

The recombination between regions of λ and the bacterial chromosome is catalyzed by a specific enzyme system. This system normally ensures that λ integrates at the same point in the chromosome and, when the lytic cycle is induced (for instance, by ultraviolet light), it ensures that the λ prophage excises at precisely the correct point to produce a normal circular λ chromosome. Very rarely, excision is abnormal and can result in phage particles that now carry a nearby gene and leave behind some phage genes (Figure 7-31a). In λ, the nearby genes are gal on one side and bio on the other. The resulting particles are defective due to the genes left behind and are referred to as λdgal (λ-defective gal), or λdbio. These defective particles carrying nearby genes can be packaged into phage heads and can infect other bacteria. In the presence of a second, normal phage particle in a double infection, the λdgal can integrate into the chromosome at the λ-attachment site (Figure 7-31b). In this manner, the gal genes in this case are transduced into the second host. Because this transduction mechanism is limited to genes very near the original integrated prophage, it is called specialized transduction.

Figure 7-31. Specialized transduction mechanism in phage λ.

Figure 7-31

Specialized transduction mechanism in phage λ. (a) A lysogenic bacterial culture can produce normal λ or, rarely, an abnormal particle, λdgal, which is the transducing particle. (b) Transducing by the mixed lysate can produce (more...)


Transduction occurs when newly forming phages acquire host genes and transfer them to other bacterial cells. Generalized transduction can transfer any host gene. It occurs when phage packaging accidentally incorporates bacterial DNA instead of phage DNA. Specialized transduction is due to faulty separation of the prophage from the bacterial chromosome, so the new phage includes both phage and bacterial genes. The transducing phage can transfer only specific host genes.

By agreement with the publisher, this book is accessible by the search feature, but cannot be browsed.

Copyright © 2000, W. H. Freeman and Company.
Bookshelf ID: NBK21760


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...