DNA replication: when and where it happens, and how

A human cell copies about 6 billion base pairs in each S phase, according to OpenStax, and it does so with remarkably few mistakes. A team of enzymes unzips the double helix and builds a new partner for each strand, using the base pairing rules: A with T, and C with G.

When does DNA replication occur?

The cell cycle has two main parts: interphase, when the cell grows and prepares, and the M phase, when it divides. Interphase is split into G1, S and G2, and DNA replication happens in S phase. By the time mitosis begins, every chromosome has been copied into two identical sister chromatids.

• Before mitosis: so each of the two daughter cells gets an identical full set of chromosomes.
• Before meiosis: once only, before meiosis I. There is no replication between meiosis I and meiosis II.
• In bacteria: before binary fission, and in fast-growing bacteria a new round can start before the last one has finished.

Replication happens exactly once per cell cycle. Cells have checks that stop any stretch of DNA being copied twice, because copying part of the genome twice would damage it.

Where does DNA replication occur?

An origin of replication is a particular sequence where copying starts. From each origin, copying runs in both directions, opening a bubble with a Y-shaped replication fork at each side.

Semi-conservative replication

Each new DNA molecule is made of one old strand and one new strand. The two original strands come apart, and the cell builds a fresh partner along each one, using it as a pattern. This is called semi-conservative replication, because half of each original molecule is kept.

In 1958, Matthew Meselson and Franklin Stahl proved it. They grew bacteria on food containing heavy nitrogen (¹⁵N), so all their DNA was heavy, then moved them to ordinary nitrogen (¹⁴N) and spun samples of DNA in a centrifuge, where DNA settles by density.

After one generation, every molecule was exactly halfway between heavy and light. That ruled out conservative replication, where the original molecule stays intact and a fully new one is made, which would have given one heavy band and one light band. After two generations, half was hybrid and half light, just as semi-conservative replication predicts.

What is the role of enzymes in the DNA replication process?

Enzymes do every step: they unwind the DNA, start each new strand, build it, check it and join the pieces. These are the main ones, with names from bacteria, where replication was first worked out.

What does DNA polymerase do?

DNA polymerase builds the new strand. It reads the template, picks the free nucleotide that pairs with each base, and bonds it onto the end of the growing strand, extending the sugar-phosphate backbone by one unit. It has three limits that explain much of how replication works:

• It can only add to an existing strand, so it needs a primer to start.
• It only builds in one direction, 5′ to 3′, adding to the 3′ end of the new strand.
• It checks as it goes. If a wrong base is added, the polymerase backs up, removes it and tries again. This proofreading corrects most mistakes, and repair enzymes fix most of the rest after copying, a process called mismatch repair.

Eukaryotes use several polymerases. OpenStax describes pol α starting each strand after the primer, pol ε building the leading strand and pol δ building the lagging strand.

The steps of DNA replication

1. Initiation: proteins recognise an origin of replication and helicase starts unwinding the DNA, opening a bubble with a fork at each end.
2. Priming: primase lays down short RNA primers on the exposed template strands.
3. Elongation: DNA polymerase adds nucleotides to each primer, matching A with T and C with G.
4. Primer removal: the RNA primers are removed and replaced with DNA.
5. Joining: DNA ligase seals the remaining gaps in the backbone.
6. Termination: forks meet, or reach the end of the DNA, and the two new molecules separate.

Try putting the steps in order yourself, then press Mark to check them in our Genetics and DNA course.

Leading and lagging strands

The two strands of DNA run in opposite directions, but DNA polymerase can only build 5′ to 3′. At each fork, that creates a problem with one of the two strands.

• The leading strand is built continuously, in one long piece, because the polymerase can follow the fork as it opens.
• The lagging strand is built in short pieces called Okazaki fragments, working away from the fork. Each fragment needs its own primer, and ligase joins the fragments afterwards.

The fragments are named after Reiji and Tsuneko Okazaki, who discovered them in the 1960s. Think of painting a stripe down a corridor while walking backwards: on one side you can keep going, but on the other you have to keep turning round and painting short sections towards where you have been.

Worked examples

Example 1: the new strand

A template strand reads 3′ TACGGA 5′. What is the new strand? Pair each base, A with T and C with G, keeping the strands antiparallel: 5′ ATGCCT 3′. The new strand is built 5′ to 3′, so it grows from left to right here.

Example 2: copying a bacterial chromosome

The chromosome of the bacterium E. coli has about 4.6 million base pairs. Copying starts at one origin and runs both ways, with each fork copying about 1,000 nucleotides a second, the prokaryote rate OpenStax gives. How long does it take?

Two forks together copy 2 × 1,000 = 2,000 base pairs a second. Time = 4,600,000 ÷ 2,000 = 2,300 seconds, about 38 minutes.

Example 3: why we need many origins

Human chromosomes hold about 3.1 billion base pairs per set, and eukaryotic forks are slower, about 100 nucleotides a second. If copying started from a single origin with two forks, it would take 3,100,000,000 ÷ 200 = 15.5 million seconds, about 179 days. With thousands of origins working at once, the whole genome is copied within the hours of one S phase.

Example 4: Meselson and Stahl

After 3 generations in light nitrogen, what share of DNA molecules is hybrid? There are always just 2 strands of original heavy DNA, each in a separate molecule, and the number of molecules doubles each generation, to 2 × 2 × 2 = 8 for each original molecule. So 2 out of 8 molecules are hybrid: 25%. The rest are fully light.

Common mistakes

• Saying replication happens during mitosis. It happens in S phase of interphase, before mitosis starts.
• Saying helicase breaks the backbone. Helicase breaks the hydrogen bonds between bases. The backbone stays intact.
• Saying DNA polymerase builds 3′ to 5′. It adds to the 3′ end, so the new strand grows 5′ to 3′.
• Mixing up ligase and polymerase. Polymerase adds nucleotides. Ligase joins fragments together.
• Thinking the whole molecule is new. Each new DNA molecule keeps one original strand: that is what semi-conservative means.
• Using U in DNA. Uracil appears only in the RNA primers, which are removed.

Read next

Biology · Genetics and inheritanceWhat does DNA stand for? DNA meaning, structure and functionDNA stands for deoxyribonucleic acid. Learn what DNA is made of, the sugar in it, what it does, who discovered it and why it matters, with examples.