In transposition, a transposable element, or transposon, moves from one DNA address to another. Barbara McClintock discovered transposons in the 1940s in her studies on the genetics of maize. Since then, transposons have been found in all kinds of organisms, from bacteria to humans. We will begin with a discussion of the bacterial transposons.
Discovery of Bacterial TransposonsJames Shapiro and others laid the groundwork for the discovery of bacterial transposons with their discovery in the late 1960s of phage mutations that did not behave normally. For example, they did not revert readily the way point mutations do, and the mutant genes contained long stretches of extra DNA. Shapiro demonstrated this by taking advantage of the fact that a l phage will sometimes pick up a piece of host DNA during lytic infection of E. coli cells, incorporating the “passenger” DNA into its own genome. He allowed l phages to pick up either a wild-type E. coli galactose utilization gene (gal1) or its mutant counterpart (gal2), then measured the sizes of the recombinant DNAs, which contained l DNA plus host DNA. He measured the DNA sizes by measuring the densities of the two types of phage using cesium chloride gradient centrifugation (Chapter 20). Because the phage coat is made of protein and always has the same volume, and because DNA is much denser than protein, the more DNA the phage contains the denser it will be. It turned out that the phages harboring the gal2 gene were denser than the phages with the wild-type gene and therefore held more DNA. The simplest explanation is that foreign DNA had inserted into the gal gene and thereby inactivated it. Indeed, later experiments revealed 800–1400-bp inserts in the mutant gal gene, which were not found in the wild-type gene. In the rare cases when such mutants did revert, they lost the extra DNA. These extra DNAs that could inactivate a gene by inserting into it were the first transposons discovered in bacteria. They are called insertion sequences (ISs).
The Simplest Bacterial TransposonsBacterial insertion sequences contain only the elements necessary for transposition. The first of these elements is a set of special sequences at a transposon’s ends, one of which is the inverted repeat of the other. The second element is the set of genes that code for the enzymes that catalyze transposition. Because the ends of an insertion sequence are inverted repeats, if one end of an insertion sequence is 59-ACCGTAG, the other end of that strand will be the reverse complement: CTACGGT-39. The inverted repeats given here are hypothetical and are presented to illustrate the point. Typical insertion sequences have somewhat longer inverted repeats, from 15 to 25 bp long. IS1, for example, has inverted repeats 23 bp long. Larger transposons can have inverted repeats hundreds of base pairs long. Stanley Cohen provided one graphic demonstration of inverted repeats at the ends of a transposon with the experiment illustrated in figure. He started with a plasmid containing a transposon with the structure shown on the figure. The original plasmid was linked to the ends of the transposon, which were inverted repeats. Cohen reasoned that if the transposon really had inverted repeats at its ends, he could separate the two strands of the recombinant plasmid, and get the inverted repeats on one strand to base-pair with each other, forming a stem-loop structure as shown on the right in figure. The stems would be double-stranded DNA composed of the two inverted repeats: the loops would be the rest of the DNA in single-stranded form. The electron micrograph in figure shows the expected stem-loop structure. The main body of an insertion sequence codes for at least two proteins that catalyze transposition. These proteins are collectively known as transposase; we will discuss their mechanism of action later in this chapter. We know that these proteins are necessary for transposition because mutations in the body of an insertion sequence can render that transposon immobile. One other feature of an insertion sequence, shared with more complex transposons, is found just outside the transposon itself. This is a pair of short direct repeats in the DNA immediately surrounding the transposon. These repeats did not exist before the transposon inserted; they result from the insertion process itself and tell us that the transposase cuts the target DNA in a staggered fashion rather than with two cuts right across from each other. Figure shows how staggered cuts in the two strands of the target DNA at the site of insertion lead automatically to direct repeats. The length of these direct repeats depends on the distance between the two cuts in the target DNA strands. This distance depends in turn on the nature of the insertion sequence. The transposase of IS1 makes cuts 9 bp apart and therefore generates direct repeats that are 9 bp long.
More Complex TransposonsInsertion sequences and other transposons are sometimes called “selfish DNA,” implying that they replicate at the expense of their hosts and apparently provide nothing useful in return. However, some transposons do carry genes that are valuable to their hosts, the most familiar being genes for antibiotic resistance. Not only is this a clear benefit to the bacterial host, it is also valuable to molecular biologists, because it makes the transposon much easier to track. For example, consider the situation in figure, in which we start with a donor plasmid containing a gene for kanamycin resistance (Kanr) and harboring a transposon (Tn3) with a gene for ampicillin resistance (Ampr); in addition, we have a target plasmid with a gene for tetracycline resistance (Tetr). After transposition, Tn3 has replicated and a copy has moved to the target plasmid. Now the target plasmid confers both tetracycline and ampicillin resistance, properties that we can easily monitor by transforming antibiotic-sensitive bacteria with the target plasmid and growing these host bacteria in medium containing both antibiotics. If the bacteria survive, they must have taken up both antibiotic resistance genes; therefore, Tn3 must have transposed to the target plasmid.
Mechanisms of TranspositionBecause of their ability to move from one place to another, transposons are sometimes called “jumping genes.” However, the term is a little misleading because it implies that the DNA always leaves one place and jumps to the other. This mode of transposition does occur and is called non replicative transposition (or “cut and paste”) because both strands of the original DNA move together from one place to the other without replicating. However, transposition frequently involves DNA replication, so one copy of the transposon remains at its original site as another copy inserts at the new site. This is called replicative transposition (or “copy and paste”) because a transposon moving by this route also replicates itself. Let us discuss how both kinds of transposition take place.
Replicative Transposition of Tn3Tn3, whose structure is shown in figure, illustrates one well-studied mechanism of transposition. In addition to the bla gene, which encodes ampicillin-inactivating b-lactamase, Tn3 contains two genes that are instrumental in transposition. Tn3 transposes by a two-step process, each step of which requires one of the Tn3 gene products. Figure shows a simplified version of the sequence of events. We begin with two plasmids; the donor, which harbors Tn3, and the target. In the first step, the two plasmids fuse, with Tn3 replication, to form a cointegrate in which they are coupled through a pair of Tn3 copies. This step requires recombination between the two plasmids, which is catalyzed by the product of the Tn3 transposase gene tnpA. Figure 23.6 shows a detailed picture of how all four DNA strands involved in transposition might interact to form the cointegrate. Figures illustrate transposition between two plasmids, but the donor and target DNAs can be other kinds of DNA, including phage DNAs or the bacterial chromosome itself. The second step in Tn3 transposition is a resolution of the cointegrate, in which the cointegrate breaks down into two independent plasmids, each bearing one copy of Tn3. This step, catalyzed by the product of the resolvase gene tnpR, is a recombination between homologous sites on Tn3 itself, called res sites. Several lines of evidence show that Tn3 transposition is a two-step process. First, mutants in the tnpR gene cannot resolve cointegrates, so they cause formation of cointegrates as the final product of transposition. This demonstrates that the cointegrate is normally an intermediate in the reaction. Second, even if the tnpR gene is defective, cointegrates can be resolved if a functional tnpR gene is provided by another DNA molecule the host chromosome or another plasmid.
Nonreplicative TranspositionFigures illustrate the replicative transposition mechanism, but transposition does not always work this way. Some transposons (e.g., Tn10) move without replicating, leaving the donor DNA and appearing in the target DNA. How does this occur? It may be that non replicative transposition starts out in the same way as replicative transposition, by nicking and joining strands of the donor and target DNAs, but then something different happens. Instead of replication occurring through the transposon, new nicks appear in the donor DNA on either side of the transposon. This releases the gapped donor DNA but leaves the transposon still bound to the target DNA. The remaining nicks in the target DNA can be sealed, yielding a recombinant DNA with the transposon integrated into the target DNA. The donor DNA has a double-stranded gap, so it may be lost or, as shown in figure, here, the gap may be repaired.
Transposons contain inverted terminal repeats