The replication of numerous procaryotes was studied in recent years. The mechanisms of the replication of phages T4, T7 and λ are now well known, while those of B. subtilis are beginning to be understood. However, the most extensively studied system remains Escherichia coli, thanks to the combination of various types of approaches.
Experimental Approaches to the Study of Replication:
In a first step, the isolation and then enzymological study of DNA polymerase I gave a better understanding of the mechanism of polymerization of nucleotides. In a second step, the isolation of thermosensitive conditional mutants, i.e. incapable of replicating their DNA in certain temperature conditions, enabled the identification of the proteins which play a role of primary importance in replication. It could thus be shown that the “replicative” enzyme is not DNA polymerase I, but DNA polymerase III.
A third type of approach consists in studying the replication in vitro of DNA phages whose genome is circular, single-stranded and relatively short (5 to 6 000 nucleotides). For their replication, these phages (M13, ϕX174, G4) use mainly the proteins of the E.coli replication machinery.
It was thus possible to show, for example, the existence of the primase, responsible for the synthesis of not only the RNA primer allowing the replication of phage G4, but also the RNA primers of Okazaki fragments in the bacterium.
Lastly, more recently, using genetic engineering techniques it was possible to obtain large quantities of enzymes relatively scarce in the cell, or to create template deoxyribonucleic acids having the desired sizes or sequences.
Major advances could be made thanks to the construction of a plasmid carrying the origin of the replication of the chromosome of E.coli (oriC), and then the isolation of a semi-purified cell-free fraction capable of replicating this DNA in vitro. Presently, there are at least 25 known proteins which participate in the replication of the chromosome in E.coli.
Initiation at the Origin of Replication:
There is an important point to note: the fact that E.coli has only one origin of replication provides the bacterium an easier control of the replication of the chromosome. However, practically nothing is known at present, of the processes responsible for the origin of replication being active or at rest.
On the contrary, knowledge is being acquired on the proteins involved in the initiation at the origin of replication.
In fact, these can be divided in two classes:
(a) Proteins involved in the initiation;
(b) “Specificity proteins” which prevent the initiation process at sites other than oriC.
The specific binding of the product of gene dnaA to oriC represents the first step of initiation. Complexed with ATP, the protein dnaA will open the double helix, permitting the formation of the initiation complex. The separation of the two strands of the DNA molecule by the product of dnaB will introduce new constraints which will cause twists of the whole DNA molecule upon itself (supercoils). These supercoils are eliminated by DNA gyrase.
This enzyme belongs to the category of topo-isomerases which are enzymes capable of causing a transitory and reversible rupture of the phosphodiester bond. One may imagine a situation where the two DNA fragments rotate round one another. The DNA is thus “ready”, the ribonucleotide primer can then be synthesized, probably by the primase.
All the proteins playing a role in this step are sometimes defined as a complex called orisome. Some of them specifically recognize oriC (gyrase, dnaA) or interact with the other proteins of the orisome (dnaB, gyrase).
Lastly, other proteins called “specificity proteins” act in preventing initiation at sites other than oriC (protein HU) or eliminating RNA primers made at illegitimate sites (RNase H, an enzyme which specifically degrades the RNA strand of a RNA/DNA hybrid).
Progression of the Replication Fork:
Before considering the events which permit the progression of the replication fork, it would be useful to briefly describe here, the principal agent of replication in E.coli: the DNA polymerase III holoenzyme.
As mentioned above, there are three DNA polymerases in this bacterium. These three enzymes differ by a few biochemical criteria which permit a distinction between them in semi-purified extracts or during the purification of the enzymes. The DNA polymerase III holoenzyme is responsible for the synthesis of the two DNA strands at the replication fork.
This enzyme is composed of at least 7 sub-units: three of them form the core enzyme; the 4 others are auxiliary sub-units. The core enzyme comprises sub-unit a responsible for the polymerization, sub-unite responsible for the 3′ —> 5′ ex- onuclease activity (“Fidelity of replication”) and a third, θ, whose role is still unknown. The four other sub-units (τ, γ, β and δ) confer on the core enzyme a greater efficiency during the polymerization and control of fidelity.
The mechanisms of the elongation step are schematically represented in fig. 6.32. We mentioned above that replication is continuous on one strand (see fig. 6.32 part e) and discontinuous on the other. One must continue to unwind the two strands of the DNA from one another and hence the presence of a DNA helicase is necessary (see fig. 6.32, part a).
It is not yet known which helicase is really implied in this process (product of the gene dnaB or gene rep?). As described in the initiation step, one must eliminate the supercoils thus formed, with the help of a topoisomerase, probably DNA gyrase (see fig. 6.32, part b). The unwinding of the two strands of the DNA molecule brings about the formation of single-stranded zones.
These single-stranded zones must be stabilized (prevention of the re-association of the two strands of the DNA molecule) and protected from the action of enzymes capable of degrading them (DNase). These functions are carried out by a protein called single- stranded DNA binding protein, generally known by its abbreviation SSB (see fig. 6.32, part c).
The ribonucleotidic primers of Okazaki fragments are synthesized by the primase (see fig. 6.32, part d). This enzyme may be distinguished from RNA polymerase by its resistance to rifampicin. It synthesizes short RNA fragments (about 10 nucleotides) without any particular sequence specificity. These primers are then extended by the DNA polymerase III holoenzyme.
This therefore leads to the formation of a chain which contains at its 5′ end, an RNA fragment covalently bonded to the DNA fragment (see fig. 6.32, part f). But we know that the chromosome contains no ribonucleotide when the replication is completed. It is therefore necessary to eliminate the ribonucleotide primers and then fill the gaps thus formed.
These two functions are carried out by DNA polymerase I, which eliminates the ribonucleotide primers thanks to its 5′ → 3′ exonuclease activity (which is not specific for DNA) and simultaneously performs the polymerization of the missing DNA fragment using as a primer the 3′ end of the next Okazaki fragment, i.e. the nearest on the side of the replication fork (see fig. 6.32, part g).
The synthesis of this strand is therefore discontinuous, and it is then in the form of juxtaposed DNA fragments. It is therefore necessary to bind these fragments together covalently. This is accomplished by DNA ligase, which is capable of forming a phosphodiester bond between a 3′-OH and a 5′- monophosphate, provided that these groups are maintained near one another thanks to hydrogen bonds exchanged between the newly synthesized DNA strand and the complementary parental strand.
The DNA ligase cannot achieve the closure if the 5′-monophosphate belongs to a ribonucleotide which compulsorily implies the removal of RNA primers from the Okazaki fragments.
It must be emphasized that the situation described here is still a model, very probable, but not proved. More recently, other hypotheses — complementary — were advanced, in order to take into consideration, particularly the fact that, in vivo, the replication fork progresses relatively fast (500 to 1 000 nucleotides per second).
For a better coordination of the action of the various enzymes, the latter could be physically associated in a multi-enzymatic complex: the replisome (see fig. 6.33). This hypothetical structure could contain 2 molecules of DNA polymerase III, one or several DNA helicases and the primosome (the set of proteins permitting the initiation of chains).
To enable this structure to replicate the 2 strands concomitantly, one must imagine that the strand which will be copied discontinuously winds on itself at 180°; this allows the two parental strands to have the same polarity locally, at the fork.
By such a mechanism, not only a molecule of polymerase (or a dimer of polymerase) would copy the two chains simultaneously, but it would also remain attached to the primosome which progresses in a direction opposite to that of the elongation on the discontinuous synthesis strand.
As far as termination is concerned, it is known that in E.coli, there is a genetically silent termination region where the two replication forks slow down. Going over this region could be the signal of cell division.
New Hypotheses on the Replication in E.coli:
As regards initiation of replication, it appears that the bacterial membrane plays an important role, on one hand because the protein dnaA is associated with this membrane, and on the other hand, because the activity of the complex dnaA-ATP is modified by a membrane phospholipid: cardiolipin.
At the level of elongation, according to a recent hypothesis, the DNA polymerase III holoenzyme is in the form of an asymmetric dimer. This asymmetry would be due to the mutually exclusive presence of two partly identical sub-units. The half of the dimer comprising the γ sub-unit would perform continuous synthesis and the other half comprising the τ sub-unit would carry out discontinuous synthesis.