The bacteriophage Mu is a transposable element (1). It uses transposition to insert into the E. coli chromosome after infection and form Mu lysogens. The low target-site selectivity of insertion results in insertion mutations at many different locations, reflecting the disruption of many different genes, leading to a "mutator" ("Mu") phenotype. Mu also uses transposition to replicate its DNA during lytic growth; multiple rounds of transposition result in the formation of multiple Mu-specific replication forks, and DNA replication from these forks produces multiple copies of Mu. Understanding Mu has played a key role in the dissection of the control and mechanism of transposition. Mu has also been key in studying the mechanism of transposition, because Mu transposition occurs at a high frequency during lytic growth, rather than the low frequency observed for most other elements. Indeed, the first in vitro transposition system was established in 1983 by Mizuuchi using Mu (2).
The DNA of Mu is about 40 kbp in length and, in addition to a number of viral proteins, encodes two transposition proteins: One is MuA, the transposase that binds specifically to the ends of Mu DNA and executes the DNA breakage and joining reactions; the other is MuB, an ATP-binding protein that interacts with the target DNA and also regulates the activity of MuA. There are special transposition sequences at each end of the element, including multiple MuA binding sites near each terminus and, at about 1.2 kbp internal from the end, an enhancer site that can strongly stimulate recombination from different positions in either orientation (see Fig. 1 in Transposable Elements). The enhancer also contains several MuA sites flanking a binding site for a sequence-specific DN-bending protein called integration host factor. Another nonspecific DNA-bending protein, called HU, also participates in recombination. Both of these proteins probably promote a particular architecture of the transposase-end complex. A key early step in recombination is the three-way synapsis of the MuA-bound transposon ends and enhancer. This step is critical in promoting a conformational change in MuA that converts it from a simple DNA-binding protein to an active recombinase that can execute DNA breakage and joining. Although not essential, MuB bound to target DNA can also play a role in regulating end synapsis. Thus, by influencing the assembly of the synaptic structure, MuB, via its interaction with target DNA, can play a key role in controlling transposition.
The first chemical step in Mu transposition is the introduction by transposase of a nick at the 3′ ends of the transposon; no double-strand breaks are made (3) (Fig. 1). These exposed 3′OH transposon ends then attack the target DNA at staggered positions. The product of these reactions contains the transposon covalently linked to the target DNA through its 3′ ends and also linked to the donor DNA by its 5′ ends; it is variably called a fusion product, a strand transfer product, or a Shapiro intermediate [Shapiro first proposed this intermediate in a model for Mu transposition (4)]. This intermediate can undergo either of two alternative processing reactions. In one pathway, an endonuclease activity (possibly provided by the host) clips the 5′ Mu ends, disconnecting the target DNA from the transposon inserted into the target DNA; thus, Mu forms a simple insertion product by a two-step mechanism. In the other processing reaction, replication across both strands of the phage initiates from the two 3′OH target ends that flank the transposon in the fusion product intermediate. Once the replication of both strands is completed, there are now two copies of the transposon linked by the donor DNA and the target DNA, in a structure called a cointegrate. Thus, a transposition event can provide two replication forks that mediate Mu replication; that is, two copies of transposon are derived from the transposition product formed by a single Mu substrate DNA. Formation of a cointegrate is an example of replicative transposition.
Figure 1. Mu transposition pathways. Mu transposition initiates with cleavages that expose the 3′OH ends of the element; these exposed ends are then joined to target DNA to make a fusion product that contains the transposon attached to the target DNA via its 3′ ends and still attached to the donor DNA via its 5′ ends. In replicative transposition, replication initiates from 3′OHs in the flanking target DNA at both ends of Mu. This replication results in a cointegrate, a single DNA molecule containing two copies of the transposon, the donor backbone, and the target DNA. In the nonreplicative pathway, another set of cleavages occur that disconnect the 5′ ends of the transposon from the flanking donor DNA. A simple insertion results after host repair of the gaps that flank the newly inserted element.
The key chemical steps of Mu transposition are cleavages to expose the 3′ ends of the elements and the joining of the 3′OH transposon ends to the target DNA. This polarity was first described with Mu, and we now know it is true of all other transposition reactions examined that involve processing of a DNA intermediate. The X-ray crystallography structures of the MuA transposase (5) and human immunodeficiency virus (HIV) retroviral integrase (6) demonstrate that the catalytic centers of these transposases are very similar in structure, although there is little amino acid sequence homology between these proteins. A notable feature of these structures is a cluster of acidic amino acids that are not adjacent in the primary sequence. These acidic residues form a binding site for Mg , which is an essential cofactor in all known transposition reactions and probably places a key role in the actual execution of the DNA breakage and joining steps. Many other transposable elements in bacteria and eukaryotes execute the same 3′OH steps in transposition, and it is likely that their catalytic regions will be related to the common form found in MuA transposase and HIV integrase (7, 8).
