Dale_Molecular Genetics of Bacteria 4th ed

Plasmids like R100 also contain a region known as the par locus (for partitioning) that is necessary for accurate partitioning of plasmid copies at cell division. This sequence can only act in cis, i.e. it must be present on the plasmid itself and not on some other plasmid. (Genes that are able to have an effect on other DNA molecules are said to be active in trans, while those that can only affect the same DNA molecule are cis-acting). It is likely that all low copy number plasmids have a par sequence (or some other mechanism for ensuring accurate partitioning), while plasmids such as ColE1 rely primarily on high copy number to ensure that each daughter cell receives a copy of the plasmid.

Control of plasmid replication by DNA repeats (iterons)

Regulation of plasmid replication through control of the expression of RepA, as described above, is adequate for many low-copy plasmids. But the absence of a direct connection between the number of plasmids and their ability to replicate means that we would expect a distribution of copy number in a narrow range either side of a mean value. For example, most cells might have five copies, but some would have six and others would have four. In the presence of an active and effective partitioning system, this is adequate for stable inheritance of such a plasmid. But some plasmids, such as F, are present as only a single copy, after cell division, replicating once per cycle so that at cell division there are two copies. There is no room for a statistical distribution of plasmid copy number. Stable maintenance of such a plasmid requires a much tighter control of replication that is directly linked to the number of copies in the cell.

This is provided by the presence of repeated DNA sequences, 17–22 bp long, known as iterons, which are found in the replication initiation regions. For example, the F plasmid has nine repeats of a 17-bp sequence. The RepA protein binds to these iteron sequences. When there is more than one copy of the plasmid, the RepA protein can bind to iterons on both copies, coupling them together (Figure 5.7). This prevents further replication of the plasmid. This ‘coupling’ or ‘handcuffing’ model provides a mechanism for extremely tight control of the number of copies of the plasmid.

It also provides another mechanism for plasmid incompatibility. If two different plasmids carry related iteron sequences (so that RepA can bind to both), the plasmids will be coupled together and replication will be prevented. The two plasmids cannot therefore be stably maintained in the same cell.

Plasmid replication via single-stranded forms

Replication of plasmids in E. coli usually seems to follow the route described earlier in this chapter, that is by copying both strands as part of the same process.

RepA binds to iterons

Single copy of the plasmid

Replication is allowed

Plasmid replication

Plasmids coupled by RepA bound to iterons

Further replication is prevented

Figure 5.7 Coupling model for the control of iteron-containing plasmids

This model is not universally applicable. Many plasmids, especially in Grampositive bacteria, replicate via a single-stranded intermediate (Figure 5.8). Although such plasmids are often referred to as single-stranded plasmids, it should be noted that the single-stranded form is only a replication intermediate and that normally the majority of plasmid molecules are double stranded.

A similar mode of replication has already been described, namely that of the single-stranded bacteriophages fX174 and M13 (Chapter 4). There is indeed a considerable similarity between these systems, not only in the general concept but also in the mechanisms involved. The general mechanism of replication of all of these elements, although differing in some details, follows the outline shown in Figure 5.8. A specific site (the plus origin) on the plus strand of the plasmid is first cut by a plasmid-encoded protein (Rep). The nicked DNA provides a site for initiation of DNA synthesis using host enzymes which displaces the old plus strand. While this is happening, the Rep protein remains attached to the free 50 end of the nicked plus strand. When this process has travelled right round the

Nick at plus origin extended by DNA polymerase

Synthesis of new plus strand

Synthesis of complementary minus strand

Rep releases old plus strand

Figure 5.8 Replication of single-stranded plasmids

plasmid and returned to the plus origin, the Rep protein makes another nick to release the old plus strand and ligates the ends of this molecule to produce an intact single-stranded circular structure. This is then converted to the doublestranded form by synthesis of the complementary (minus) strand starting from a separate origin (the minus origin).

The same model can be applied to the single-stranded phages (see Figure 4.5). The important difference is, that with these phages, the single-stranded form is

not merely a replicative intermediate but is incorporated into the phage particle. In the early stages of replication, the displaced single-stranded form is stabilized by coating with a host protein known as SSB (single-stranded DNA binding protein). Later in the cycle, the progeny viral strands are packaged into phage coats rather than producing more RF DNA. In the case of fX174, complementary strand synthesis is prevented because the viral strand is packaged directly into phage pro-heads as it is produced. With M13, a specific phage protein (gene V protein) accumulates and replaces SSB on the viral DNA, thus preventing complementary strand synthesis. The gene V protein is replaced by the proteins of the phage particle during the assembly of the phage at the cell membrane.

The sequences of the plus origins of a number of these elements have been determined and fall into several groups within which there is a high degree of similarity. For example, bacteriophage fX174 and the S. aureus plasmids pC194 and pUB110 all have the same short sequence at the origin. This is the recognition site for the Rep proteins of these elements, which also show sequence similarity. This indicates an evolutionary connection between bacteriophage fX174 and plasmids from Gram-positive bacteria.

The minus origins, on the other hand, are more diverse. Since this origin requires the action of host proteins, it must have evolved to fit in with the specificity of the current host. One consequence is that if such a plasmid is put into a different bacterial species, the plus origin may function quite well, but the minus origin, which requires host proteins, may be comparatively ineffective. This will result not only in frequent loss of the plasmid by segregational instability, but also in the generation of a variety of DNA rearrangements, since the singlestranded DNA forms that will accumulate will stimulate recombination.

Replication of linear plasmids

Another type of plasmid represents even more of a challenge to common ideas of bacterial DNA structure and replication. Linear DNA plasmids have been characterized from several bacterial genera including Borrelia and Streptomyces. In these bacterial species (and some others), the chromosome is also linear. The replication of these linear molecules poses something of a problem. As previously described (Chapter 1), one new DNA strand (the leading strand) is made continuously, while the other strand (the lagging strand) is made discontinuously – that is, it is produced as short fragments that are subsequently joined together. This is necessary as nucleic acids are always made in the 50 to 30 direction. Remember also that DNA synthesis requires a primer, which is normally generated by an RNA polymerase. This priming sequence is removed and replaced as the DNA polymerase reaches it when synthesizing the next fragment.

Now consider what happens as the replication fork approaches the end of a linear DNA molecule. The leading strand will continue right to the end, but what

Initiation of replication

L R

Complete replication produces a dimeric circular structure

L R

Figure 5.11 Model of replication of linear plasmids in Borrelia

In Streptomyces, the problem of replicating the ends of the linear DNA is solved in a different way. A key feature is the presence of a protein (terminal protein, TP), covalently attached to the 50 ends of the DNA. The simplest model (Figure 5.12) is that this protein acts as a primer for DNA synthesis, allowing replication of the ends of the linear DNA – and indeed the replication of some linear DNA molecules (such as that of the Bacillus subtilis bacteriophage f29) occurs in this way. But it is not an adequate explanation for the replication of

Figure 5.12 Replication of linear DNA primed by covalently attached protein

Streptomyces plasmids (and even less so for the replication of the linear chromosome). These have an origin of replication in an internal position in the molecule, from which replication occurs bidirectionally in a conventional manner. While the leading strand (made in the 50 ! 30 direction) can be synthesized right to the end of the template, the lagging strand will stop short, leaving a short portion of DNA unreplicated (Figure 5.13). The role of the terminal protein in this case lies in patching this unreplicated portion, assisted by the formation of secondary structures in the unreplicated strand due to the presence of inverted repeat sequences.

5.3 Plasmid stability

One of the characteristic features of plasmids is their instability. Plasmid-borne features are often lost from a population at a higher frequency than would be expected for the normal processes of mutation. The extent of this instability varies enormously from one plasmid to another. Naturally-occurring plasmids are usually (but not always) reasonably stable: selection will tend to operate in that way, and in addition in isolating the strain and looking for the plasmid the more stable plasmids will tend to be selected. Unstable ones will be harder to find.

Artificially constructed plasmids on the other hand are often markedly unstable. This is usually merely a nuisance on the laboratory scale, but can become a very expensive problem in the industrial use of strains carrying such plasmids.

There are three quite distinct phenomena associated with the concept of plasmid stability: (1) plasmid integrity, (2) partitioning at cell division and (3) differential growth rates.