DNA replication

Which proteins the replisome includes and what each does; how DNA replication is initiated; how the new chains grow, leading and lagging; how primers are removed; how DNA packaging is maintained after replication; how chromosomal telomeres replicate; replication on the rolling circle model; and how mitochondrial DNA is replicated.

1. For replication, the two DNA strands first separate.
2. This allows single complementary nucleotides to join by forming hydrogen bonds with the exposed DNA bases.
3. Then DNA polymerase forms covalent bonds between the newly joined nucleotides, connecting them into a new DNA strand.

This way the initial DNA molecule is replaced by two identical double helices. Because each of them is composed of one old and one new strand, the mechanism is called semiconservative.

The semiconservative model was proven by Matthew Meselson and Franklin Stahl in 1958.

• First, they grew bacteria for a long time in the presence of the rare, heavy, non-radioactive ¹⁵N.
• Then the bacteria were transferred to a medium containing the common, light ¹⁴N.
• The ratio of the two isotopes in DNA was estimated by density gradient centrifugation, with samples taken at 0 minutes, at 20 minutes (one round of replication) and at 40 minutes (two rounds).

The result: one generation after the transfer, the DNA was all intermediate, half heavy and half light. After another generation, half was intermediate and half was purely light, and there was no heavy-only DNA.

This pattern could only have been observed if each DNA molecule contains a template strand from the parental DNA. The conservative model would have kept a heavy-only band; a fully dispersive model would not have produced a clean light band at generation two.

All the enzymes and other molecules involved in replication move along DNA as one big complex. The whole complex is called the replisome - the machinery responsible for DNA replication.

Its members are:

1. the replication fork;
2. the leading strand;
3. the lagging strand;
4. the Okazaki fragment;
5. helicase;
6. topoisomerase;
7. single-strand binding proteins;
8. primase;
9. the RNA primer;
10. DNA polymerase III;
11. the sliding clamp;
12. DNA polymerase I;
13. ligase.

Replication origins are often designated ori.

• They are symmetrical sequences (palindromes), rich in AT to allow easier strand separation.
• Circular prokaryotic chromosomes have only one ori.
• In eukaryotes, each chromosome has multiple ori, which give rise to replication eyes, that is two symmetrical replication forks. During replication, neighbouring forks grow towards each other.

The initiator protein. The origin sequence is specifically recognised by a replication initiator protein. In prokaryotes and viruses it is a single molecule; in eukaryotes it is multi-subunit and is called the origin recognition complex (ORC).

The initiator protein binds to the origin, marking it and recruiting other proteins to assemble on it. The origin also contains a DNA unwinding element.

A replicon is a segment of a chromosome which is replicated starting from one ori.

• Ori are not distributed evenly throughout the chromosome, so replicons are of different lengths.
• Replicons are organised in clusters, usually 20 to 80 replicons per cluster.
• All the ori in the same cluster are activated at the same time during S phase.
• Replicons in heterochromatin tend to be replicated in late S phase, since they are more difficult to access than euchromatin.

DNA synthesis always proceeds in the 5' to 3' direction. The template strand is of course read in the opposite, 3' to 5', direction. Nucleotides are used in the form of triphosphates.

The two steps are worth keeping separate in your mind, because only one of them is enzymatic.

1. The nucleotide to join is selected by complementary binding. Its base forms hydrogen bonds with the exposed base on the template DNA. This is spontaneous, requires no enzyme and no energy.
2. DNA polymerase then connects the nucleotide to the growing strand, forming a covalent phosphodiester bond.

The polymerase, in other words, does not choose the nucleotide. Complementarity chooses it; the enzyme only makes the choice permanent.

The hydrogen bonds in the DNA double helix are weak but numerous. Strand separation therefore requires energy and a special protein called helicase.

The cell has several helicases; one of them is for replication. Another protein, the helicase loader, is needed to place the helicase on the DNA.

The helicase is a hexameric complex which encloses the lagging DNA strand as a ring and uses ATP energy to proceed forward. Its action resembles unzipping. It is a motor protein, and it deserves the title of enzyme just for its ability to consume ATP, so it is an ATPase.

The structure where DNA is being synthesised is fork-shaped and is called the replication fork. It moves gradually during elongation.

Because DNA synthesis can proceed only in the 5' to 3' direction, it can conveniently follow the replication fork for only one of the two new strands. That one is the leading strand, and once started from a primer it progresses continuously by the addition of nucleotides.

The other new strand, the lagging strand, is synthesised in a more complex way.

DNA polymerase is a very precise enzyme: it checks for mispaired bases. Every time it adds a nucleotide it looks back to check whether the pairing seems correct, and removes it if it does not. This is called proofreading activity.

The consequence is unexpected and it is the reason primers exist. The first nucleotides of a nascent chain do not seem correct even if they are perfect, because there is nothing behind them to check against. The proofreading activity therefore does not allow DNA polymerase to initiate a new chain. It needs help to start.

The enzyme which starts the new strand works not with deoxyribonucleotides but with ribonucleotides. This enzyme is called primase, and it synthesises a short RNA chain called the primer.

After that, DNA polymerase steps in and starts to synthesise the new DNA strand as an extension of the primer.

Unlike the "true" RNA polymerase responsible for transcription, primase can synthesise only short chains: it has low processivity.

In eukaryotes, primase is in a complex with a DNA polymerase that has no proofreading activity and low processivity: DNA polymerase α. That is why the primer is not pure RNA - it consists of a mixture of ribonucleotides and deoxyribonucleotides.

To put the arrangement in order:

• Because DNA synthesis proceeds 5' to 3', the replication fork is asymmetrical, with continuous synthesis on the leading strand and discontinuous synthesis, via Okazaki fragments, on the lagging strand.
• The Polα/primase complex assembles RNA-DNA primers. It is composed of several subunits, some working as primase and one being DNA polymerase α. The subunit numbers are not to be learned.
• The replicase then elongates the primer with DNA.

The most important enzyme in replication is the DNA-dependent DNA polymerase, for short DNA polymerase. It connects deoxyribonucleotides by phosphodiester bonds after their complementary non-covalent binding to the template.

• Like helicases, DNA polymerases are several in the cell. The one responsible for replication is called the replicative polymerase, or replicase.
• The DNA polymerase is a multi-subunit complex. The catalytic subunits, having polymerase and proofreading activities, form the core DNA polymerase.
• The two cores working at the two DNA strands are connected by a pair of subunits (τ).

The identities differ by kingdom:

• Eukaryotic replicase is a complex of pol ε (epsilon) for the leading strand and pol δ (delta) for the lagging strand.
• Bacterial replicase is called DNA polymerase III (pol III), and it has identical cores for both strands.

During replication, parts of the lagging strand template remain unpaired for a long time. These "naked" DNA parts are covered by special single strand binding (SSB) proteins, to protect them from unwanted interactions. The SSB proteins leave when the primase or the replicase comes.

On the leading strand, SSBs are not needed, because the DNA polymerase comes immediately after the helicase.

In eukaryotes the single-stranded DNA-binding protein is RPA, replication protein A.

Replication requires separation of the two parental strands. However, attempted separation is hindered by growing torsional strain. Moreover, replication of circular DNA produces two interlocked rings called catenanes.

This requires enzymes called topoisomerases, which change the degree of DNA supercoiling. They cut one or both DNA strands, rotate the molecule to change its supercoiling, and then reseal it. In this way they relieve torsional strain and separate catenanes.

• Topoisomerase I breaks one polynucleotide chain.
• Topoisomerase II breaks both.

Each adheres to the double helix through temporary covalent bonds between the enzyme and the DNA until it relaxes and sutures the torn ends again. Topoisomerases act at a certain distance on both sides of the replication fork.

The lagging strand is produced discontinuously, as short fragments called Okazaki fragments after their discoverer, Tsuneko Okazaki.

Their length is:

• 1000 to 2000 nucleotides in prokaryotes;
• 100 to 200 nucleotides in eukaryotes.

The synthesis of an Okazaki fragment begins at the replication fork and proceeds backwards until it reaches the previous primer. Then a nuclease digests the primer and the DNA polymerase fills its space. When ready, the two fragments are joined together by DNA ligase.

Unlike the primase, replicases have very high processivity: they can synthesise DNA chains of any length.

This property is based on a subunit called the sliding clamp. Changing conformation, it can open to settle on DNA and then close around it as a ring. After that, the core cannot dissociate from DNA until the clamp is opened again.

The official name of the sliding clamp is Proliferating Cell Nuclear Antigen (PCNA), and it is met again as a marker of proliferating tissue.

How the clamps increase processivity. Clamps slide freely along DNA to tether the polymerases to the templates being copied, while still allowing the polymerases to move at a rate limited only by the rate of nucleotide incorporation.

Both strands share one clamp loader, but they use it very differently: the leading strand needs it only twice, to begin and to end replication, while the lagging strand needs it for each Okazaki fragment.

1. An Okazaki fragment grows until it meets and is joined with the previous fragment.
2. It is then released from the sliding clamp.
3. A new primer starts the next Okazaki fragment.

The trombone model. This describes how a loop forms on the lagging strand during replication, resembling a trombone's slide, and it is what allows both leading and lagging strands to be synthesised simultaneously by the one replisome. The sliding clamp encircles the DNA, keeping the polymerase attached to the template; as the lagging strand polymerase extends an Okazaki fragment it forms this loop, and upon completion it detaches from the sliding clamp and moves to a new primer to start the next fragment.

After the primer has fulfilled its task it must be removed, and the two kingdoms do it differently.

In bacteria: the replicase stops and dissociates when it reaches the previous primer. Another DNA polymerase, pol I, takes its place. It has 5' to 3' exonuclease activity and, while elongating the DNA strand, digests the primer.

In eukaryotes: they have a separate 5' to 3' exonuclease called RNase H, from "hybrid", which destroys RNA primers.

In fact eukaryotic replication does not absolutely require such an RNase. The eukaryotic replicase has very high processivity, so when it reaches the previous Okazaki fragment it does not stop but pushes the primer off the template. It displaces not only the RNA primer, if it is still there, but also the short imperfect DNA part synthesised by pol α - which is exactly what you would want, since that stretch was made by a polymerase with no proofreading. The resulting single-stranded flap is then cut away by an endonuclease called FEN1.

DNA ligase recognises nicks, that is single-strand cuts in DNA, and stitches the fragments together. It needs energy from ATP for its work.

Eukaryotic cells have several DNA ligases, of which only one is used in replication - the same pattern as with the helicases and the DNA polymerases.

Putting the parts in their places: the helicase opens the fork; the topoisomerase works upstream of it, relieving the strain; single strand binding proteins cover the exposed lagging template; primase lays down primers; DNA polymerase extends them, continuously on the leading strand and as Okazaki fragments on the lagging one; and DNA ligase seals the fragments together.

The close physical connection between the replication proteins in the replisome is what makes replication both efficient and accurate. They are not a series of enzymes that happen to act in sequence; they are one machine.

During replication, chromatin decondenses to the 10 nm nucleosomal "beads on a string", and heterochromatin becomes indistinguishable from euchromatin.

• Nucleosome core histones stay on the DNA to the end and leave only when the replication fork arrives.
• After that, the old histones are randomly distributed to the two daughter DNA helices, and an equal quantity of new histones is added to finish the packing, with the help of chromatin assembly factors.

The random distribution matters: each daughter molecule inherits a mixture of old and new histones, which is how epigenetic marks are diluted but not lost at each division.

Circular DNA molecules replicate their entire sequence without complications.

Linear DNAs, such as eukaryotic chromosomes, have problems with their 5' ends. Even if the primase begins from the very end, the 5' terminal primers cannot be replaced with DNA, because replacement requires a free 3'-OH upstream and at the end of the molecule there is none.

Hence replication will produce strands shorter than the templates.

Germline cells have an enzyme called telomerase which adds the missing flap.

• It has an RNA cofactor and can synthesise one copy of the telomere repeat without a template - more precisely, with only the cofactor as template.
• The telomerase can be regarded as a special reverse transcriptase.

Chromosome telomeres are multiple tandem repeats of a short sequence. Before replication, the 3' end has one copy more, protruding as a single-strand flap. After replication the flap is lost, and this is why the chromosomes of somatic cells shorten slightly with each cell division.

How telomerase works, step by step. Telomerase has an associated RNA that complements the 3' overhang at the end of the chromosome. The RNA template is used to synthesise the complementary strand. The telomerase then shifts along and the process is repeated. Finally, primase and DNA polymerase synthesise the complementary strand.

We distinguish three modes of replication - bubble, rolling circle and D-loop - based on how the replication fork interacts with the DNA template.

The bubble mode, also called the eye or theta mode, is the one described in everything above. In it, both strands of the parent DNA serve as templates simultaneously. Replication begins with separation of the two strands and then proceeds in the shape of a widening bubble.

The vast majority of DNAs, both linear and circular, replicate using the bubble mode.

Some small circular DNAs, such as plasmids and viral genomes, reproduce as rolling circles.

1. First, an endonuclease cleaves a phosphodiester bond in one of the strands, producing a nick, a single-stranded cut in DNA.
2. Then the DNA polymerase binds to the free 3' end, as if it were a primer, and starts to elongate it, pushing aside the 5' end.
3. The lagging strand is then synthesised, using one or more primers.
4. If the polymerase makes several full circles, it produces a long, tandemly repeated leading strand.
5. Then a specific endonuclease cuts the product into single copies, and they are closed into rings by a ligase.

There is a bacterial plasmid called the F-factor, or fertility factor, which has a strategy for spreading to other cells.

It usually replicates by the bubble mode. However, sometimes it forces its cell to form a bridge connecting it to another cell. Then the F-factor replicates as a rolling circle and sends a copy of itself to the other bacterium, turning it from F- to F+.

When the F-factor integrates into the chromosome. Sometimes the F-factor integrates itself into the bacterial chromosome, and the resulting cell is called Hfr, for high-frequency recombination.

When an Hfr cell attempts the same manoeuvre, cellular genes are transferred through the bridge along with the genes of the F-factor. The phenomenon is called bacterial conjugation and is as close to a sexual process as bacteria can get.

In theory the entire chromosome of the Hfr cell can be copied to the F- cell, but usually the transfer is interrupted at some point.

In specific conditions, some phages do rolling-circle replication, as in the bacteriophage lambda replication cycle. The product is a long chain of repeated phage genomes, called concatemers. A phage protein cuts the chain into single phage genomes ready for packaging.

Herpesvirus uses both rolling-circle and theta replication in its reproduction cycle. The rolling circle is used for multiplication of the viral genome when there are enough proteins to arrange capsids for packing the new virions - the same accounting logic seen in other viruses, where genome production waits until there are shells to put the genomes in.

D-loop replication is used by mitochondrial and plastid genomes.

Their two strands have different and non-equal replication origins.

1. Replication begins on the strand carrying the more "active" origin (OH).
2. The newly synthesised DNA displaces the old complementary strand. Unlike the rolling circle, here the displaced strand remains whole.
3. As the D-loop widens, it reaches the second replication origin (OL). Then replication is initiated also on the displaced strand.
4. Both strands are synthesised as a whole from one primer.

Hence in D-loop replication there is no lagging strand - which means no Okazaki fragments, no repeated clamp loading, and none of the machinery that exists only to solve the antiparallel problem.

• Replication is semiconservative, proved by Meselson and Stahl (1958) with ¹⁵N and ¹⁴N: one generation all intermediate, two generations half intermediate and half light, never any heavy-only band.
• Origins are AT-rich palindromes. Prokaryotes have one ori; eukaryotes have many, grouped into replicons organised in clusters that fire together, with heterochromatin replicating late.
• Complementary pairing selects the nucleotide for free; the polymerase only makes it permanent.
• DNA polymerase cannot start a chain, because its own proofreading rejects the first nucleotides. That is why primase exists, and why the eukaryotic primer is a mixed RNA-DNA stretch laid down by Polα/primase.
• Eukaryotic replicase: pol ε leading, pol δ lagging. Bacterial replicase: pol III, identical cores.
• Okazaki fragments are 1000 to 2000 nt in prokaryotes, 100 to 200 nt in eukaryotes, and they are made backwards, away from the fork.
• The sliding clamp is PCNA. The leading strand loads it twice; the lagging strand loads it once per fragment.
• Primer removal: bacteria use pol I's 5' to 3' exonuclease; eukaryotes displace the primer into a flap and cut it with FEN1, removing pol α's unproofread DNA along with it.
• Topoisomerase I cuts one strand, topoisomerase II cuts two, and both relieve torsion and untangle catenanes.
• Old histones are randomly split between the two daughter helices, and new ones make up the difference.
• Linear chromosomes lose their 5' terminal primer at every round. Telomerase, a reverse transcriptase with its own RNA cofactor, restores it in germline cells.
• Three modes: bubble/theta (almost everything), rolling circle (plasmids, phages, herpesvirus, and the F-factor in conjugation), D-loop (mitochondria and plastids, with no lagging strand).