Replisome

DNA replication begins at sites called origins of replication. In organisms with small genomes and simple chromosome structure, such as bacteria, there may be only a few origins of replication on each chromosome. Organisms with large genomes and complex chromosome structure, such as humans, may have hundreds, or even thousands, of origins of replication spread across multiple chromosomes.

DNA structure varies with time, space, and sequence, and it is thought that these variations, in addition to their role in gene expression, also play active roles in replisome assembly during DNA synthesis. Replisome assembly at an origin of replication is roughly divided into three phases.

For prokaryotes:

• Formation of pre-replication complex. DnaA binds to the origin recognition complex and separates the duplex. This attracts DnaB helicase and DnaC, which maintain the replication bubble.
• Formation of pre-initiation complex. SSB binds to the single strand and then gamma (clamp loading factor) binds to SSB.
• Formation of initiation complex. Gamma deposits the sliding clamp (beta) and attracts DNA polymerase III.

For eukaryotes:

• Formation of pre-replication complex. MCM factors bind to the origin recognition complex and separate the duplex, forming a replication bubble.
• Formation of pre-initiation complex. Replication protein A (RPA) binds to the single stranded DNA and then RFC (clamp loading factor) binds to RPA.
• Formation of initiation complex. RFC deposits the sliding clamp (PCNA) and attracts DNA polymerases such as alpha (α), delta (δ), epsilon (ε).

For both prokaryotes and eukaryotes, the next stage is generally referred to as 'elongation', and it is during this phase that the majority of DNA synthesis occurs.

DNA is a duplex formed by two anti-parallel strands. Following Meselson-Stahl, the process of DNA replication is semi-conservative, whereby during replication the original DNA duplex is separated into two daughter strands (referred to as the leading and lagging strand templates). Each daughter strand becomes part of a new DNA duplex. Factors generically referred to as helicases unwind the duplex.

Helicase is an enzyme which breaks hydrogen bonds between the base pairs in the middle of the DNA duplex. Its doughnut like structure wraps around DNA and separates the strands ahead of DNA synthesis. In eukaryotes, the Mcm2-7 complex acts as a helicase, though which subunits are required for helicase activity is not entirely clear.[2] This helicase translocates in the same direction as the DNA polymerase (3' to 5' with respect to the template strand). In prokaryotic organisms, the helicases are better identified and include dnaB, which moves 5' to 3' on the strand opposite the DNA polymerase.

Examples of topological coils introduced during duplex unwinding and strand separation.

As helicase unwinds the double helix, topological changes induced by the rotational motion of the helicase lead to supercoil formation ahead of the helicase (similar to what happens when you twist a piece of thread).

Gyrase (a form of topoisomerase) relaxes and undoes the supercoiling caused by helicase. It does this by cutting the DNA strands, allowing it to rotate and release the supercoil, and then rejoining the strands. Gyrase is most commonly found upstream of the replication fork, where the supercoils form.

A rendering of the 70kd sub-unit of Replication protein A

Single-stranded DNA is highly unstable and can form hydrogen bonds with itself that are referred to as 'hairpins' (or the single strand can improperly bond to the other single strand). To counteract this instability, single-strand binding proteins (SSB in prokaryotes and Replication protein A in eukaryotes) bind to the exposed bases to prevent improper ligation.

If you consider each strand as a "dynamic, stretchy string", the structural potential for improper ligation should be obvious.

An expanded schematic reveals the underlying chemistry of the problem: the potential for hydrogen bond formation between unrelated base pairs.

Binding proteins stabilise the single strand and protected the strand from damage caused by unlicensed chemical reactions.

The combination of a single strand and its binding proteins serves as a better substrate for replicative polymerases than a naked single strand (binding proteins provide extra thermodynamic driving force for the polymerisation reaction). Strand binding proteins are removed by replicative polymerases.

From both a structural and chemical perspective, a single strand of DNA by itself (and the associated single strand binding proteins) is not suitable for polymerisation. This is because the chemical reactions catalysed by replicative polymerases require a free 3' OH in order to initiate nucleotide chain elongation. In terms of structure, the conformation of replicative polymerase active sites (which is highly related to the inherent accuracy of replicative polymerases) means these factors cannot start chain elongation without a pre-existing chain of nucleotides, because no known replicative polymerase can start chain elongation de novo.

Priming enzymes, (which are DNA-dependent RNA polymerases), solve this problem by creating an RNA primer on the leading and lagging strands. The leading strand is primed once, and the lagging strand is primed approximately every 1000 (+/- 200) base pairs (one primer for each Okazaki fragment on the lagging strand). Each RNA primer is approximately 10 bases long.

The interface at (A*) contains a free 3' OH that is chemically suitable for the reaction catalysed by replicative polymerases, and the "overhang" configuration is structurally suitable for chain elongation by a replicative polymerase. Thus, replicative polymerases can begin chain elongation at (A*).

In prokaryotes, the primase creates an RNA primer at the beginning of the newly separated leading and lagging strands.

In eukaryotes, DNA polymerase alpha creates an RNA primer at the beginning of the newly separated leading and lagging strands, and, unlike primase, DNA polymerase alpha also synthesizes a short chain of deoxynucleotides after creating the primer.

Processivity refers to both speed and continuity of DNA replication, and high processivity is a requirement for timely replication. High processivity is in part ensured by ring-shaped proteins referred to as 'clamps' that help replicative polymerases stay associated with the leading and lagging strands. There are other variables as well: from a chemical perspective, strand binding proteins stimulate polymerisation and provide extra thermodynamic energy for the reaction. From a systems perspective, the structure and chemistry of many replisome factors (such as the AAA+ ATPase features of the individual clamp loading sub-units, along with the helical conformation they adopt), and the associations between clamp loading factors and other accessory factors, also increases processivity.

To this point, according to research by Kuriyan et al.,[3] due to their role in recruiting and binding other factors such as priming enzymes and replicative polymerases, clamp loaders and sliding clamps are at the heart of the replisome machinery. Research has found that clamp loading and sliding clamp factors are absolutely essential to replication, which explains the high degree of structural conservation observed for clamp loading and sliding clamp factors. This architectural and structural conservation is seen in organisms as diverse as bacteria, phages, yeast, and humans. That such a significant degree of structural conservation is observed without sequence homology further underpins the significance of these structural solutions to replication challenges.

Clamp loader is a generic term that refers to replication factors called gamma (prokaryotes) or RFC (eukaryotes). The combination of template DNA and primer RNA is referred to as 'A-form DNA' and it is thought that clamp loading replication proteins (helical heteropentamers) want to associate with A-form DNA because of its shape (the structure of the major/minor groove) and chemistry (patterns of hydrogen bond donors and acceptors).[3][4] Thus, clamp loading proteins associate with the primed region of the strand which causes hydrolysis of ATP and provides energy to open the clamp and attach it to the strand.[3][4]

A rendering of a fully assembled, trimeric PCNA clamp

Sliding clamp is a generic term that refers to ring-shaped replication factors called beta (prokaryotes) or PCNA (eukaryotes). Clamp proteins attract and tether replicative polymerases, such as DNA polymerase III, in order to extend the amount of time that a replicative polymerase stays associated with the strand. From a chemical perspective, the clamp has a slightly positive charge at its centre that is a near perfect match for the slightly negative charge of the DNA strand.

In some organisms, the clamp is a dimer, and in other organisms the clamp is a trimer. Regardless, the conserved ring architecture allows the clamp to enclose the strand.

Replicative polymerases form an asymmetric dimer at the replication fork by binding to sub-units of the clamp loading factor. This asymmetric conformation is capable of simultaneously replicating the leading and lagging strands, and the collection of factors that includes the replicative polymerases is generally referred to as a holoenzyme. However, significant challenges remain: the leading and lagging strands are anti-parallel. This means that nucleotide synthesis on the leading strand naturally occurs in the 5' to 3' direction. However, the lagging strand runs in the opposite direction and this presents quite a challenge since no known replicative polymerases can synthesise DNA in the 3' to 5' direction.

The dimerisation of the replicative polymerases solves the problems related to efficient synchronisation of leading and lagging strand synthesis at the replication fork, but the tight spatial-structural coupling of the replicative polymerases, while solving the difficult issue of synchronisation, creates another challenge: dimerisation of the replicative polymerases at the replication fork means that nucleotide synthesis for both strands must take place at the same spatial location, despite the fact that the lagging strand must be synthesised backwards relative to the leading strand. Lagging strand synthesis takes place after the helicase has unwound a sufficient quantity of the lagging strand, and this "sufficient quantity of the lagging strand" is polymerised in discrete nucleotide chains called Okazaki fragments.

Consider the following: the helicase continuously unwinds the parental duplex, but the lagging strand must be polymerised in the opposite direction. This means that, while polymerisation of the leading strand proceeds, polymerisation of the lagging strand only occurs after enough of the lagging strand has been unwound by the helicase. At this point, the lagging strand replicative polymerase associates with the clamp and primer in order to start polymerisation. During lagging strand synthesis, the replicative polymerase sends the lagging strand back toward the replication fork. The replicative polymerase disassociates when it reaches an RNA primer. Helicase continues to unwind the parental duplex, the priming enzyme affixes another primer, and the replicative polymerase reassociates with the clamp and primer when a sufficient quantity of the lagging strand has unwound.

Collectively, leading and lagging strand synthesis is referred to as being 'semidiscontinuous'.

Prokaryotic and eukaryotic organisms use a variety of replicative polymerases, some of which are well-characterised:

• DNA polymerase III
• DNA polymerase delta
• DNA polymerase epsilon

This polymerase synthesizes leading and lagging strand DNA in prokaryotes.

This polymerase synthesizes lagging strand DNA in eukaryotes.[5] (Thought to form an asymmetric dimer with DNA polymerase epsilon.)[6]

This polymerase synthesizes leading strand DNA in eukaryotes.[7] (Thought to form an asymmetric dimer with DNA polymerase delta.)[5]

Although rare, incorrect base pairing polymerisation does occur during chain elongation. (The structure and chemistry of replicative polymerases mean that errors are unlikely, but they do occur.) Many replicative polymerases contain a "error correction" mechanism in the form of a 3' to 5' exonuclease domain that is capable of removing base pairs from the exposed 3' end of the growing chain. Error correction is possible because base pair errors distort the position of the magnesium ions in the polymerisation sub-unit, and the structural-chemical distortion of the polymerisation unit effectively stalls the polymerisation process by slowing the reaction.[8] Subsequently, the chemical reaction in the exonuclease unit takes over and removes nucleotides from the exposed 3' end of the growing chain.[9] Once an error is removed, the structure and chemistry of the polymerisation unit returns to normal and DNA replication continues. Working collectively in this fashion, the polymerisation active site can be thought of as the "proof-reader", since it senses mismatches, and the exonuclease is the "editor", since it corrects the errors.

Base pair errors distort the polymerase active site for between 4-6 nucleotides, which means, depending on the type of mismatch, there are up to six chances for error correction.[8] The error sensing and error correction features, combined with the inherent accuracy that arises from the structure and chemistry of replicative polymerases, contributes to an error rate of approximately 1 base pair mismatch in 10⁸ to 10¹⁰ base pairs.

Errors can be classified in three categories: purine-purine mismatches, pyrimidine-pyrimidine mismatches, and pyrimidine-purine mismatches. The chemistry of each mismatch varies, and so does the behaviour of the replicative polymerase with respect to its mismatch sensing activity.

The replication of bacteriophage T4 DNA upon infection of E. coli is a well-studied DNA replication system. During the period of exponential DNA increase at 37°C, the rate of elongation is 749 nucleotides per second.[11] The mutation rate during replication is 1.7 mutations per 10⁸ base pairs.[12] Thus DNA replication in this system is both very rapid and highly accurate.

There are two problems after leading and lagging strand synthesis: RNA remains in the duplex and there are nicks between each Okazaki fragment in the lagging duplex. These problems are solved by a variety of DNA repair enzymes that vary by organism, including: DNA polymerase I, DNA polymerase beta, RNAse H, ligase, and DNA2. This process is well-characterised in prokaryotes and much less well-characterised in many eukaryotes.

In general, DNA repair enzymes complete the Okazaki fragments through a variety of means, including: base pair excision and 5' to 3' exonuclease activity that removes the chemically unstable ribonucleotides from the lagging duplex and replaces them with stable deoxynucleotides. This process is referred to as 'maturation of Okazaki fragments', and ligase (see below) completes the final step in the maturation process.

Primer removal and nick ligation can be thought of as DNA repair processes that produce a chemically-stable, error-free duplex. To this point, with respect to the chemistry of an RNA-DNA duplex, in addition to the presence of uracil in the duplex, the presence of ribose (which has a reactive 2' OH) tends to make the duplex much less chemically-stable than a duplex containing only deoxyribose (which has a non-reactive 2' H).

DNA polymerase I is an enzyme that repairs DNA.

RNAse H is an enzyme that removes RNA from an RNA-DNA duplex.

After DNA repair factors replace the ribonucleotides of the primer with deoxynucleotides, a single gap remains in the sugar-phosphate backbone between each Okazaki fragment in the lagging duplex. An enzyme called DNA ligase connects the gap in the backbone by forming a phosphodiester bond between each gap that separates the Okazaki fragments. The structural and chemical aspects of this process, generally referred to as 'nick translation', exceed the scope of this article.