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

Bacteriophage genetics

In this section, we shall describe how crosses can actually be done with viruses (phages) that infect bacteria and the experiments that dissect the fine structure of the gene.

Infection of bacteria by phages

Most bacteria are susceptible to attack by bacteriophages, which literally means “eaters of bacteria.” A phage consists of a nucleic acid “chromosome” (DNA or RNA) surrounded by a coat of protein molecules. One well-studied set of phage strains are identified as T1, T2, and so forth. Figures 7-17 and 7-18 show the structure of a T-even phage (T2, T4, and so forth).

Figure 7-17. Phage T4, shown in its free state and in the process of infecting an E.

Figure 7-17

Phage T4, shown in its free state and in the process of infecting an E. coli cell. The infecting phage injects DNA through its core structure into the cell. On the right, a phage has been diagrammatically exploded to show its highly ordered three-dimensional (more...)

Figure 7-18. Enlargement of the E.

Figure 7-18

Enlargement of the E. coli phage T4 showing details of structure: note head, tail, and tail fibers. The T4 phage was used by Benzer in his experiments on the nature of the rII (rapid lysis) gene. (Photograph from Jack D. Griffith.)

During infection, a phage attaches to a bacterium and injects its genetic material into the bacterial cytoplasm (Figure 7-19). The phage genetic information then takes over the machinery of the bacterial cell by turning off the synthesis of bacterial components and redirecting the bacterial synthetic material to make more phage components (Figure 7-20). (The use of the word information is interesting in this connection; it literally means “to give form,” which is precisely the role of the genetic material: to provide blueprints for the construction of form. In the present discussion, the form is the elegantly symmetrical structure of the new phages.) Ultimately, many phage descendants are released when the bacterial cell wall breaks open. This breaking-open process is called lysis.

Figure 7-19. Micrograph of a bacteriophage attaching to a bacterium and injecting its DNA.

Figure 7-19

Micrograph of a bacteriophage attaching to a bacterium and injecting its DNA. (Dr. L. Caro/Science Photo Library, Photo Researchers.)

Figure 7-20. Progeny particles of phage λ maturing inside an E.

Figure 7-20

Progeny particles of phage λ maturing inside an E. coli cell. (Jack D. Griffith.)

How can we study inheritance in phages when they are so small that they are visible only under the electron microscope? In this case, we cannot produce a visible colony by plating, but we can produce a visible manifestation of an infected bacterium by taking advantage of several phage characters. Let’s look at the consequences of a phage’s infecting a single bacterial cell. Figure 7-21 shows the sequence of events in the infectious cycle that leads to the release of progeny phages from the lysed cell. After lysis, the progeny phages infect neighboring bacteria. This is an exponentially explosive phenomenon (it causes an exponential increase in the number of lysed cells). Within 15 hours after the start of an experiment of this type, the effects are visible to the naked eye: a clear area, or plaque, is present on the opaque lawn of bacteria on the surface of a plate of solid medium (Figure 7-22). Such plaques can be large or small, fuzzy or sharp, and so forth, depending on the phage genome. Thus, plaque morphology is a phage character that can be analyzed. Another phage phenotype that we can analyze genetically is host range, because phages may differ in the spectra of bacterial strains that they can infect and lyse. For example, certain strains of bacteria are immune to adsorption (attachment) or injection by phages.

Figure 7-21. A generalized bacteriophage lytic cycle.

Figure 7-21

A generalized bacteriophage lytic cycle. (After J. Darnell, H. Lodish, and D. Baltimore, Molecular Cell Biology. Copyright © 1986 by W. H. Freeman and Company.)

Figure 7-22. The appearance of phage plaques.

Figure 7-22

The appearance of phage plaques. Individual phages are spread on an agar medium that contains a fully grown “lawn” of E. coli. Each phage infects one bacterial cell, producing 100 or more progeny phages that burst the E. coli cell and (more...)

Phage cross

Can we cross two phages in the same way that we cross two bacterial strains? A phage cross can be illustrated by a cross of T2 phages originally studied by Alfred Hershey. The genotypes of the two parental strains of T2 phage in Hershey’s cross were hr+ × h+r. The alleles are identified by the following characters: h can infect two different E. coli strains (which we can call strains 1 and 2); h+ can infect only strain 1; r rapidly lyses cells, thereby producing large plaques; and r+ slowly lyses cells, thus producing small plaques.

In the cross, E. coli strain 1 is infected with both parental T2 phage genotypes at a phage:bacteria concentration (called multiplicity of infection) that is high enough to ensure that a large percentage of cells are simultaneously infected by both phage types. This kind of infection (Figure 7-23 is called a mixed infection or a double infection. The phage lysate (the progeny phage) is then analyzed by spreading it onto a bacterial lawn composed of a mixture of E. coli strains 1 and 2. Four plaque types are then distinguishable (Figure 7-24 and Table 7-3). These four genotypes can be scored easily as parental (hr+ and h+r) and recombinant (h+r+ and hr), and a recombinant frequency can be calculated as follows:

Image ch7e6.jpg

Figure 7-23. A double infection of E.

Figure 7-23

A double infection of E. coli by two phages.

Figure 7-24. Plaque phenotypes produced by progeny of the cross h− r+ × h+ r−.

Figure 7-24

Plaque phenotypes produced by progeny of the cross hr+ × h+r. Enough phages of each genotype are added to ensure that most bacterial cells are infected with at least one phage of each genotype. (more...)

Table 7-3. Progeny Phage Plaque Types from Cross h− r+ × h+ r−.

Table 7-3

Progeny Phage Plaque Types from Cross h r+ × h+ r.

If we assume that entire phage genomes recombine, then single exchanges can occur and produce viable reciprocal products, unlike bacterial crosses where two crossover events are required. Nevertheless, phage crosses are subject to complications. First, several rounds of exchange can potentially occur within the host: a recombinant produced shortly after infection may undergo further exchange at later times. Second, recombination can occur between genetically similar phages as well as between different types. Thus, P1 × P1 and P2 × P2 occur in addition to P1 × P2 (P1 and P2 refer to phage 1 and phage 2, respectively). For both of these reasons, recombinants from phage crosses are a consequence of a population of events rather than defined, single-step exchange events. Nevertheless, all other things being equal, the RF calculation does represent a valid index of map distance in phages.


Recombination between phage chromosomes can be studied by bringing the parental chromosomes together in one host cell through mixed infection. Progeny phages can be examined for parental versus recombinant genotypes.

rII system

Seymour Benzer’s work in the 1950s refined the phage cross to the point where extremely small levels of recombination could be detected. This work led to a greater understanding of the nature of the fine structure of the gene, which we consider in detail in Chapter 9. The key to this work was the development of a system that allowed the selection of rare recombinants. This system used the rII genes of phage T4.

One type of mutant T4 phage produced larger, ragged plaques: these were r (rapid lysis) mutants. Benzer mapped the mutations responsible for the r phenotype to two loci: rI and rII. He then studied the rII mutants intensively.

One extraordinary property of rII mutants made all of Benzer’s work possible: rII mutants have a different host range from that of wild-type phages. Two related but different strains of E. coli, termed B and K(λ), can be used as different hosts for phage T4. Both bacterial strains can distinguish rII mutants from wild-type phages. E. coli B allows both to grow, but plaques of different sizes result: wild-type phages produce small plaques, and rII mutants produce large plaques. E. coli K, an abbreviation for E. coli K(λ), does not permit the growth of rII mutants, but it does allow wild-type phages to grow. The rII mutants are then conditional mutants—namely, mutants that can grow under one set of conditions but not another. E. coli B is said to be permissive for rII mutants, because it allows phage growth, whereas E. coli K is said to be nonpermissive for rII mutants, because it does not allow phage growth. Table 7-4 shows the growth characteristics and plaque morphology of these phages on each host strain.

Table 7-4. Plaque Phenotypes Produced by Different Combinations of E. coli and Phage Strains.

Table 7-4

Plaque Phenotypes Produced by Different Combinations of E. coli and Phage Strains.

Selection in genetic crosses of bacteriophages

Benzer crossed various rII mutants of the T4 phage and obtained recombination frequencies, which he then used to map mutations within the rII gene region.

Let’s see how this works. Suppose that we wish to cross two rII mutants and recover wild-type recombinants. Because wild-type and rII mutants make plaques that can be distinguished from each other, we could cross two different rII mutants in E. coli B and examine the progeny on E. coli B (Figure 7-25, top photograph at lower right), hoping to find small wild-type plaques among the large parental rII plaques. If the recombination frequency is high enough to yield from 2 to 3 percent or more wild-type plaques, then this method would suffice. However, for recombination that is less frequent than 1 percent, a lot of work would be required to generate a map of numerous rII mutations.

Figure 7-25. The process of recombination permits parts of the DNA of two different phage mutants to be reassembled in a new DNA molecule that may contain both mutations or neither of them.

Figure 7-25

The process of recombination permits parts of the DNA of two different phage mutants to be reassembled in a new DNA molecule that may contain both mutations or neither of them. Mutants obtained from two different cultures are introduced into a broth of (more...)

Instead of plating the progeny phages from the cross on E. coli B, however, we could plate the progeny on E. coli K (Figure 7-25, bottom photograph at lower right), so that only the wild-type recombinant phages could grow. Even if the recombination frequency is very low (say, 0.01 percent), we could easily detect the recombinant wild-type phages. Why? Because a typical phage lysate (the phage mixture released after lysis of the bacteria) from such an infection (whether it includes a cross or not) contains in excess of 109 phages per milliliter. If we mix 0.1 ml of such a phage lysate with 0.1 ml of E. coli K bacteria, then we will have more than 105 (100,000) wild-type recombinant phage infecting the bacteria when the recombination frequency is 0.01 percent. (In practice, increasing dilutions of the phage lysate are used until one yields a countable number of plaques.) Now we can see the power of Benzer’s rIIE. coli B/K system. In a single milliliter, it can find one recombinant or revertant per 109 organisms. Contrast this with trying to find one recombinant in 109 Drosophila or 109 mice!

After we have made our cross, we need to determine the recombinant frequency. First, we count the number of active virus particles, or plaque-forming units (pfu), that grew on E. coli K (these plaque-forming units, remember, are only wild-type recombinant phages) and the number that grew on E. coli B (which represent the total progeny phages, because all of the virus particles can grow on strain B). The recombinant frequency can then be calculated as twice the number of pfu on E. coli K divided by the number of pfu on E. coli B. Why do we use twice the pfu frequency for E. coli K? To account for the recombinants that are double mutants and that we cannot detect; such mutants should be present at the same frequency as the wild-type recombinants.

Finally, in any cross of this type, we need to plate each parental lysate on E. coli K to see how many revertants to wild type there were in the population. Back (reverse) mutations occur at some very low but real frequency. It is important to monitor this frequency and to compare it with our calculated frequency of recombination to be sure that recombination—not back reversion of the parental types—has occurred.

In summary, Benzer’s use of the rII system and two different bacterial hosts provided him with a method for selecting for rare crossover events within the gene without having to screen large numbers of plaques.


Benzer capitalized on the fantastic resolving power made possible by a system that selects for rare events in rapidly multiplying phages; this system allowed him to map a gene in molecular detail.

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: NBK21824


  • Cite this Page
  • Disable Glossary Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...