Dale_Molecular Genetics of Bacteria 4th ed

upstream from the insertion site, so each of the gene cassettes is transcribed from the integron promoter. Integrons are thus a naturally-occurring analogy to the expression vectors that will be discussed in Chapter 8, for obtaining expression of foreign genes by inserting them into a vector adjacent to a promoter.

Integrons are not only found in the Tn21 family of transposons, several classes of recognizably distinct integrons have been found in other transposons as well as in plasmids which do not carry functional transposons. They therefore represent a significant additional mechanism for the evolution of bacterial plasmids and the spread of antibiotic resistance.

7.3 Mechanisms of transposition

7.3.1 Replicative transposition

In considering mechanisms of transposition, it is not necessary to distinguish between insertion sequences and transposons as the same mechanisms apply to both types of transposable elements. Transposons such as Tn3 transpose by a replicative mechanism: a copy of the element is inserted at a different site (on the chromosome or on a plasmid) while the original copy is retained. The stages of replicative transposition are outlined in Figure 7.8 which depicts transposition from one plasmid (A) to a second plasmid (B). The transposase mediates a form of recombination between plasmid A carrying the transposon and the target plasmid B. With some transposons, this target site appears to be more or less random, i.e. there is no requirement for a specific sequence, while other transposons do have a degree of specificity in their target sequences. An extreme example is Tn7 which has only one insertion site in the E. coli chromosome.

The outcome of this stage is the formation of a larger plasmid known as a cointegrate, which consists of the complete sequence of both plasmids fused together, but now with two copies of the transposon, in the same orientation. With some naturally-occurring plasmids, cointegrate molecules such as this can be readily isolated; in other cases, the intermediate is rapidly resolved into two separate plasmids, each of which contains a copy of the transposon, with the target sequence on the recipient plasmid being duplicated on either side of the inserted transposon.

Since the two copies of the transposon are in the same orientation (i.e. it is a direct repeat), resolution of the cointegrate can occur by recombination between the two copies as described in Chapter 6 (see Figure 6.11). This can be achieved by general recombination using host recombination systems. However, some transposons (including Tn3) encode their own resolution system. The tnpR gene of Tn3 (see Figure 7.4) codes for a resolvase which mediates a site-specific recombination at the resolution site within the transposon, thus ensuring an efficient resolution of the cointegrate, independent of host recombination activity.

A

Cointegrate

Transposon

Target

B

Resolution

A

Plasmid B with a copy of the transposon flanked by the duplicated target sequence

Figure 7.8 Schematic diagram of replicative transposition. The first step shows formation of a cointegrate structure, carrying two copies of the transposon, with duplication of the target sequence. Recombination between the two transposon copies leads to separation into two plasmids, each having a copy of the transposon. The direct repeats flanking the original copy of the transposon are not shown

Molecular basis of replicative transposition

The molecular basis of replicative transposition can be considered in two stages: formation of the cointegrate, followed by resolution. A simplified model of cointegrate formation is shown in Figure 7.9 in which the transposon is on one plasmid (A) and the target sequence on a second plasmid (B) as in Figure 7.8. It is important to remember that the linear structures shown are parts of circular molecules and of course the strands are twisted around each other rather than lying side by side. The steps shown are as follows.

(a) Single strand breaks (nicks) in the DNA are produced at each 30 end of the transposon (in opposite strands). The recipient plasmid is also nicked on either

(c) DNA synthesis duplicates transposon and target, leading to a cointegrate

X M

Transposon

Target

Target

Transposon

L Y

Figure 7.9 Model of the mechanism of replicative transposition: formation of a cointegrate (see the text for details)

side of a short target sequence. The staggered nature of the nicks is the ultimate cause of the duplication of the target sequence.

(b) The free ends of the transposon are joined to the free ends generated by the nicks in the recipient plasmid. At no stage in this process is the transposon itself liberated as an independent molecule.

(c) The free 30 ends of the recipient plasmid sequence act as primers for the synthesis of DNA strands using host DNA polymerase. This synthesis will proceed through the transposon, separating the two strands until it reaches the existing complementary strand to which the new strand will be joined by the action of DNA ligase. This replicative step is responsible for the duplication of the transposon itself and of the target sequence. This produces the cointegrate structure (see Figure 7.8), i.e. the structure in (c) is now one large circle).

The cointegrate plasmid (Figure 7.10) contains two copies of the transposon in the same orientation (direct repeat). Recombination between the two copies gives rise to the end products of transposition: two plasmids each containing a copy of the transposon. Plasmid A is as it started while plasmid B has acquired a copy of the transposon flanked by a direct repeat of the target sequence.

7.3.2 Non-replicative (conservative) transposition

Not all transposons show exactly the same behaviour. In particular, some transposons and some insertion sequences, do not replicate when they transpose, exhibiting a mode of transposition known as conservative (or non-replicative) transposition. This occurs with the insertion sequence IS10 and the related transposon Tn10.

It is possible to use a modified version of the model in Figure 7.9 for nonreplicative transposition. If, following the joining of the free ends of the transposon to the nicked recipient plasmid (step b), the previously unbroken donor strands are cut, this will release the recipient replicon which now contains the transposon flanked by a single-stranded target sequence. These single-stranded regions are filled in by repair synthesis while the donor replicon, now linearized by excision of the transposon, is degraded.

An alternative model of non-replicative transposition is the ‘cut and paste’ process presented in Figure 7.11. In this case, the transposon is completely excised from the donor molecule before being attached to the target site. With some transposons, there is evidence for the existence of small amounts of free transposon DNA, often in a circular form.

The difference between these models is not as great as appears at first, since it consists mainly in the relative timing of the events concerned. Many transposons appear to use both replicative and non-replicative transposition; indeed, this may be the norm rather than the exception.

Although, with conservative transposition the plasmid molecule that acted as the transposon donor is degraded in the process, this does not necessarily mean the complete loss of that plasmid from the cell. Since there may be a number of copies of the plasmid, the loss of one copy is easily rectified by replication. This can make it very difficult to determine with certainty which type of transposition has occurred.

direction of transcription

Figure 7.13 Activation of chromosomal genes from an insertion sequence

inactivation of the gene into which the element is inserted. With some transposable elements it is known that the converse effect can occur, i.e. insertion of the element actually promotes the expression of genes adjacent to the site of insertion. The reason for this is that some insertion sequences (such as IS10) contain a promoter (identified as pOUT in Figure 7.13) which is directed outwards, i.e. away from the transposase gene and towards any genes that may be found in the flanking chromosomal DNA. These genes, if in the correct orientation, will therefore be turned on by the presence of IS10.

7.3.5 Mu: a transposable bacteriophage

The name Mu is short for mutator, which is derived from the fact that E. coli cells which carry this phage show an abnormally high rate of mutation.

The Mu phage particle contains about 38 kb of DNA as a linear structure which carries variable ends that are derived from bacterial DNA. After injection into an E. coli cell, the Mu DNA becomes inserted into the bacterial DNA at a random position. In common with transposons and insertion sequences, this insertion involves the duplication of a short region (5 bp in this case) of host sequence at the target site. Thereafter replication occurs by repeated replicative transposition events with the transposed copies being inserted at different sites around the chromosome. The insertion of copies of Mu DNA at various positions on the E. coli chromosome causes a loss of function of those genes, hence the high rate of mutation. Eventually, the productive phase of phage infection involves the excision of Mu DNA copies from the chromosome starting from a fixed point to the left of the Mu DNA insert (thus including some bacterial DNA). Packaging proceeds until the phage head is full (cf. bacteriophage T4; Chapter 6); since this requires more DNA than the length of Mu itself, some bacterial DNA is also included at the right-hand end.

7.3.6Conjugative transposons and other transposable elements

In Chapter 6, conjugative transposons were described, which are not only able to transpose from one site to another within a cell, but can also transpose to other cells by a conjugative mechanism. Transposition of these elements differs radically from the models described above since it involves excision of the transposon to generate a covalently closed circular intermediate. This structure is unable to replicate but is a substrate for subsequent integration into a different site, as well as being able to be transferred by conjugation to a recipient cell.

Some other mobile elements also transpose via circular intermediate structures. These elements are noticeably different from typical transposons and insertion sequences in lacking inverted repeat ends and in not generating direct repeats of the target sequence at the site of insertion.

7.4 Phase variation

In addition to the transcriptional and translational regulatory mechanisms described in Chapter 3, many bacteria have evolved additional adaptive mechanisms to enable them to respond to changing environments. In these bacteria a genetically diverse population is generated by reversible genetic changes known as phase variation. The principle is illustrated in Figure 7.14 which envisages a reversible (but inherited) switch in the state of a specific gene. In phase 1, the gene is not expressed, but when a cell switches to phase 2, the gene is expressed. Initially, the population is mainly in phase 1 (expression off) with a small minority of cells in the alternative state (expression on). When this population is subjected to changed conditions (such as the immune response of the host) to which phase 1 cells are susceptible, all the cells in phase 1 will be killed, but the phase 2 cells will survive and multiply. This gives rise to a population that is expressing this gene (phase 2). Since the change is reversible, when conditions change again, the population will become diverse again or even switch back to being predominantly phase 1. Some examples of phase variation, and the mechanisms involved, are considered below.

Reversible variation offers considerable advantages over conventional (irreversible) mutation. For example, a pathogen may be killed either by the immune response of the host or by antibiotic treatment. With conventional mutations, a tiny minority of the original population might escape from the immune response, but would still be sensitive to the antibiotic and a different minority would become resistant to the antibiotic but would be killed by the immune response. However, if the antigenic variation is due to a reversible phase variation, then this is a characteristic of all the cells in the population (even though it is only expressed

Population mostly in phase 1: expression off

Growth under phase 2-selective conditions

Population in phase 2:

expression on

Growth under non-selective conditions

Reverts to a mixed population

Figure 7.14 Phase variation. In the original state, the population is mainly in phase 1 (not expressing a specific gene), but a few cells are in phase 2 (expression on) due to a reversible genetic change. Exposure to conditions under which only phase 2 cells will survive will produce an homogenous phase 2 population. When changed to non-selective conditions, the population will gradually revert to a mixture

by a minority at any one time), so the cells that survive the antibiotic will still be capable of expressing different antigens. Furthermore, reversible phase variation enables the population to retain potential genetic characteristics that are actually disadvantageous under certain circumstances – such as expression of an antigen that is necessary for adhesion and thus for colonization, but also provides a target for the immune response. In these respects, evolutionary pressures will operate at the population level rather than on individual cells.