Science Portal

DNA replication step by step: how cells copy their DNA

October 3, 2026approx. 15 minutesby mybody®x Editorial & Expert Team

DNA-Replikation Schritt für Schritt: wie die Zelle ihre DNA kopiert

Key points

During DNA replication, the cell copies its DNA before dividing. An enzyme called helicase opens the double helix, or the two twisted DNA strands. A short RNA starting piece, the primer, gives DNA polymerases a starting point for building new strands; DNA ligase 1 joins the pieces at the end (Burgers and Kunkel, 2017). Proofreading removes incorrectly inserted building blocks.

Of the two new strands, the leading strand is considered to be built continuously, while the lagging strand is formed in short pieces called Okazaki fragments (Burgers and Kunkel, 2017).

Half of the genome, or the entire genetic material, is copied in pieces on the lagging strand (Smith and Whitehouse, 2012).

Proofreading is estimated to improve accuracy by two to three orders of magnitude (Bębenek and Ziuzia-Graczyk, 2018).

The process at a glance

01

Start at the origin of replication

At thousands of starting points, called origins of replication, preparatory complexes called pre-replication complexes prepare for copying (Fragkos et al., 2015).

02

Helicase opens the DNA

Active helicase unwinds the double helix using energy from the breakdown of the ATP molecule, creating the replication fork where the strands separate (Burgers and Kunkel, 2017).

03

Primase sets the starting point

An RNA primer of about 10 nucleotides, the individual building blocks, provides the starting point for each new strand (Burgers and Kunkel, 2017).

04

Leading and lagging strands

The leading strand is considered to be built continuously, while the lagging strand is formed in Okazaki fragments (Burgers and Kunkel, 2017).

05

Ligase seals the gaps

Once the RNA primer has been broken down, DNA ligase 1 joins the pieces (Burgers and Kunkel, 2017).

06

Proofreading

Incorrectly inserted building blocks are removed from the end of the growing strand before it is extended further (Bębenek and Ziuzia-Graczyk, 2018).

Also in this article

Where does DNA replication begin?

At thousands of starting points, called replication origins: pre-replication complexes form there before copying to prepare for the start of DNA synthesis (Fragkos et al., 2015).

The article DNA Structure Explained Simply explains that copying is semiconservative, meaning each new double helix retains one template strand and gains one new strand, using the image of a zipper - there, the principle; here, the stages and their enzymes.

Cells must copy DNA because each new cell needs an exact copy of the DNA from the old cell when it divides (MedlinePlus Genetics, NLM, 2021). The scale is considerable. Human DNA contains about 3 billion bases, the chemical letters in which its information is stored (MedlinePlus Genetics, NLM, 2021). Most DNA is located in the cell nucleus (MedlinePlus Genetics, NLM, 2021); the article Where Is DNA Located? explains how it is housed there.

Cited source

“Each strand of DNA in the double helix can serve as a template for duplicating the sequence of bases.”

National Library of Medicine (NLM), MedlinePlus Genetics
What is DNA?, 2021, translated from English

Initiation follows the cell cycle, the sequence of phases between one cell division and the next. During the G1 phase, which precedes copying, pre-replication complexes form at replication origins (Fragkos et al., 2015). At the transition to the S phase, the stage during which DNA synthesis takes place, these become initiation complexes, and the helicase is activated (Fragkos et al., 2015).

Burgers and Kunkel (2017) divide this initiation into four phases. First, the MCM helicase is loaded as a double ring at sites marked by the Origin Recognition Complex (ORC), a complex that recognizes origins (Burgers and Kunkel, 2017). Then additional helper factors bind. In the third step, the helicase is converted from its inactive to its active form; in the fourth, the first primer is created (Burgers and Kunkel, 2017).

Not every origin is used: only some origins are used during each S phase (Fragkos et al., 2015).

How does helicase open the DNA double helix?

The active helicase encircles a single strand and unwinds the double helix using energy from the breakdown of adenosine triphosphate (ATP) (Burgers and Kunkel, 2017). This creates the replication fork.

Before initiation, the Mcm2-7 ring is still inactive around the double-stranded DNA; only after it is remodeled into an active helicase does it encircle a single strand (Burgers and Kunkel, 2017). The active helicase is a complex of three factors (Burgers and Kunkel, 2017). Mcm2-7 forms the core, joined by Cdc45 and GINS; together they are called the CMG complex (Burgers and Kunkel, 2017).

Where the helicase has separated the strands, they lie exposed as templates. This site is called the replication fork because the double helix splits here into its two single strands. The fork continues moving during replication, and the next steps take place there (Burgers and Kunkel, 2017).

The helicase thus works like the advance team of a construction crew: it clears the way before building begins. The enzyme names in this article refer to eukaryotes, organisms whose cells have a nucleus; some of the findings come from baker’s yeast (Burgers and Kunkel, 2017).

What does primase do during DNA replication?

It provides the starting piece: The complex of polymerase α and primase begins both new strands with an RNA primer about 10 nucleotides long (Burgers and Kunkel, 2017).

A primer is a short starting piece of RNA that is then extended with DNA. Primase is the enzyme that makes this primer. Polymerase α (Pol α) is a DNA polymerase, an enzyme that links nucleotides together to form a new strand. Nucleotides are the individual building blocks of a DNA strand (MedlinePlus Genetics, NLM, 2021).

The primer does not grow longer than about 10 nucleotides, because as it lengthens, it runs into spatial constraints within the complex (Burgers and Kunkel, 2017). On the lagging strand, Pol α then adds about 20–30 DNA nucleotides and stops, allowing DNA polymerase δ (Pol δ) to take over and continue building the strand (Burgers and Kunkel, 2017). This estimate is based on classic studies of the SV40 virus (Burgers and Kunkel, 2017).

The complex is particularly hard at work on the lagging strand. Flexible connectors allow it to make large movements as it produces the RNA primers for the millions of Okazaki fragments needed to replicate eukaryotic chromosomes (Burgers and Kunkel, 2017). Chromosomes are the units in which DNA is found in the cell nucleus. The primase thus works like a surveyor, setting out the starting points one stake at a time.

Another copying process, rewriting a gene into mRNA, is explained step by step in the article Rewriting DNA into mRNA and translating it.

What distinguishes the leading and lagging strands of DNA?

Structure: The leading strand is generally considered to be synthesized continuously, that is, in one piece, while the lagging strand is formed discontinuously in short Okazaki fragments, each beginning with its own RNA primer (Burgers and Kunkel, 2017).

“Why is one strand continuous and the other not?” It is a fair question. What can be shown here is how the two strands differ at the start. On the lagging strand, the Pol α-primase complex lays down a new primer for each Okazaki fragment; Pol δ extends it until it reaches the next Okazaki fragment (Burgers and Kunkel, 2017). On the leading strand, by contrast, synthesis is generally assumed to proceed without such restarts (Burgers and Kunkel, 2017).

Which polymerase builds which strand has been studied in yeast. For undamaged yeast cells, the findings strongly suggest that DNA polymerase ε (Pol ε), another DNA polymerase, synthesizes primarily the leading strand, while Pol α and Pol δ synthesize primarily the lagging strand (Burgers and Kunkel, 2017).

The short pieces are called Okazaki fragments, named after the Okazaki couple, who first proposed a model for their synthesis in 1968 (Burgers and Kunkel, 2017). Millions of them are needed for eukaryotic chromosomes (Burgers and Kunkel, 2017). In other words, half of the genome is replicated discontinuously as Okazaki fragments on the lagging strand (Smith and Whitehouse, 2012).

However, the leading strand is not entirely seamless either. When polymerases incorporate unusual building blocks, such as uracil, which is not one of DNA's bases, or RNA building blocks, and these are later removed and repaired, it is discontinuous in places (Burgers and Kunkel, 2017).

Leading and lagging strands compared

Feature Leading strand Lagging strand
Structure continuous, that is, in one piece; this is the general assumption discontinuous, in Okazaki fragments
Start RNA primer from the Pol α-primase complex a separate RNA primer for each fragment
Polymerase in yeast primarily Pol ε primarily Pol α and Pol δ
Accuracy Pol ε proofreads its own errors lower, offset by more active mismatch repair, which corrects mismatched bases after copying

Source: Burgers and Kunkel, Annual Review of Biochemistry, 2017; the information about the polymerases is based on findings in yeast cells.

How does ligase seal the gaps in DNA?

It joins the pieces: Once the RNA of all primers has been degraded, DNA ligase 1, an enzyme that joins DNA pieces, seals the remaining nick (Burgers and Kunkel, 2017).

Woman climbs a white wooden staircase up to the dunes

The order of events on the lagging strand is clearly described: Pol δ extends each primer until it reaches the next Okazaki fragment (Burgers and Kunkel, 2017). The primer RNA is then degraded. What remains is a nick, a place where DNA adjoins DNA without the two ends being joined.

Ligase works like a welder on a construction site: The components are already aligned; it just adds the seam. Because the lagging strand is made up of millions of fragments (Burgers and Kunkel, 2017), each of these seams is a separate step.

Only then have the fragments become a continuous strand.

Chapter at a glance

On the lagging strand, Pol δ extends each primer to the next Okazaki fragment (Burgers and Kunkel, 2017). The primer RNA is then degraded. DNA ligase 1 then joins the adjacent DNA pieces (Burgers and Kunkel, 2017). This turns the fragments into a continuous strand.

How accurate is the copy during DNA replication?

Accuracy relies on several safeguards; proofreading alone is estimated to improve it by two to three orders of magnitude, or about 100- to 1,000-fold (Bębenek and Ziuzia-Graczyk, 2018).

One safeguard is built into the polymerase itself: It selects the nucleotides it incorporates (Bębenek and Ziuzia-Graczyk, 2018).

Proofreading comes into play next: The polymerase removes incorrectly incorporated nucleotides from the end of the growing strand before extending it further (Bębenek and Ziuzia-Graczyk, 2018). Most polymerases that copy DNA during replication have this ability, but not all (Bębenek and Ziuzia-Graczyk, 2018). The term comes from the workplace, and it fits: As in copy editing, the fresh text is checked while it is still open on the desk.

After replication, mismatch repair also takes effect, a repair system for mismatched bases that works only after copying is complete (Bębenek and Ziuzia-Graczyk, 2018).

There are differences between the polymerases. Pol ε proofreads its own errors with a 3′ exonuclease, a built-in cutting function, making it more accurate than Pol δ and considerably more accurate than Pol α, which has no proofreading function (Burgers and Kunkel, 2017). Its lower accuracy on the lagging strand is offset by more active mismatch repair there (Burgers and Kunkel, 2017).

Still, the copying process is fast. Supported by the replication clamp PCNA, an auxiliary protein that increases its speed, Pol δ can add about 250 nucleotides per second under saturating conditions, that is, when building blocks are present in excess (Burgers and Kunkel, 2017). At the concentrations of building blocks found in the cell, and with competing RNA building blocks, the rate is about 50 nucleotides per second, which matches the speed at which the fork moves in the cell (Burgers and Kunkel, 2017).

Speed and accuracy in numbers

about 10

Nucleotides long is the RNA primer that starts each new strand (Burgers and Kunkel, 2017).

up to about 250

Nucleotides per second is what Pol δ achieves under saturating conditions, and about 50 under physiological conditions (Burgers and Kunkel, 2017).

2 to 3

Proofreading is estimated to improve accuracy by several orders of magnitude (Bębenek and Ziuzia-Graczyk, 2018).

Source: Burgers and Kunkel, Annual Review of Biochemistry, 2017; Bębenek and Ziuzia-Graczyk, Current Genetics, 2018

The article DNA metabolism and genetic influences explains how DNA repair and metabolism are related to genetic factors.

What does folate have to do with DNA replication?

Folate, also called vitamin B9 (MedlinePlus Genetics, NLM, 2019), is part of one-carbon metabolism, a metabolic cycle whose functions Raghubeer and Matsha (2021) also include DNA synthesis.

The authors also include folate and methionine metabolism in this cycle; a key enzyme in it is MTHFR (Raghubeer and Matsha, 2021). The MTHFR gene of the same name contains the instructions for making the enzyme methylenetetrahydrofolate reductase (MedlinePlus Genetics, NLM, 2019). This enzyme converts one form of folate, 5,10-methylenetetrahydrofolate, into another, 5-methyltetrahydrofolate, the main form of folate in the blood (MedlinePlus Genetics, NLM, 2019).

The body needs this form for the multistep conversion of the amino acid homocysteine into the amino acid methionine, which it uses for proteins and other important compounds (MedlinePlus Genetics, NLM, 2019).

This brings us back to cell division: every new cell needs an exact copy of the DNA (MedlinePlus Genetics, NLM, 2021), and the metabolism in which MTHFR works is also involved in DNA synthesis (Raghubeer and Matsha, 2021). No conclusion about your own folate needs follows from this; that belongs in a conversation with your doctor.

Reading a saliva sample is clearly distinct from replication in every dividing cell, and that is what the next chapter is about.

What does a DNA test read from your saliva?

Predefined individual positions rather than the entire genome: The NutriCare | The INFINITY DNA test by mybody®x (MYBODY Lab GmbH) uses SNP genotyping with over 700,000 SNPs and is not whole-genome sequencing (information current as of: 03.10.2026).

Whole-genome sequencing reads the entire genome. By contrast, SNP genotyping checks individual positions selected in advance; an SNP is a variant at a single base position where people may differ. The result is your genotype, meaning the variants you carry at these positions.

Do not confuse the readout with the copying process described in this article. The test does not measure how quickly or accurately your cells copy their DNA; it reads variants at predefined positions and, according to the product information, covers 140 genetic variations (information current as of: 03.10.2026).

Your genetic makeup is not set in stone. The test is not a diagnostic procedure and does not replace medical advice.

NutriCare | INFINITY DNA test by mybody®x

From the mybody®x range · DNA test

NutriCare | INFINITY DNA test

Reads predefined positions in your DNA from a saliva sample using SNP genotyping, analyzed by a specialist laboratory in Germany. It is not whole-genome sequencing or a diagnostic procedure; it does not measure the replication of your cells.

Price €269.00 As of 03.10.2026, subject to change
Sample type Saliva sample
Processing time Kit shipping 1–3 business days
Lab analysis 15–25 business days after sample receipt

Information current as of: 03.10.2026

Go to NutriCare | INFINITY DNA test

Limitations: What don’t the DNA replication figures show?

They apply to specific experimental systems, not automatically to every human cell: the approximately 20–30 DNA nucleotides after the primer are based on studies of the SV40 virus, while the approximately 250 nucleotides per second figure comes from saturating laboratory conditions (Burgers and Kunkel, 2017).

The division of labor between Pol ε and Pol δ comes from experiments with baker’s yeast (Saccharomyces cerevisiae), so it should be understood as a finding in yeast, not as a measurement in humans (Burgers and Kunkel, 2017). The same review also describes an alternative model in which Pol δ synthesizes both strands.

The two to three orders of magnitude are an estimate of how much proofreading improves accuracy (Bębenek and Ziuzia-Graczyk, 2018). They describe how many times more accurate the copy is with proofreading than without it.

The saliva test has its own limitations: It is not a diagnostic procedure and does not predict disease. If you have questions about a result, discuss them with your doctor.

How do you remember DNA replication?

With a sequence of verbs: open, start, build, join, check.

Each verb represents a step. Helicase opens, primase starts, polymerases build, ligase joins, and during proofreading, the polymerase checks its own work. If you can confidently recite this sequence, you have the process in mind and can match each verb to its enzyme.

A mnemonic can help distinguish the two strands: The leading strand runs continuously, while the lagging strand is built in pieces and stitched together. To check your understanding, tell the story of the lagging strand from the first primer to the final join. If you can do it without missing a step, you’ve got the difference from the leading strand down too.

This article began with a single word: copy. Now you can see the sequence of steps behind it, each carried out by an enzyme you can name.

Frequently asked questions

How can DNA replication be explained simply?

Think of it as copying from a template: Each strand of the double helix serves as a pattern for the sequence of bases, with A pairing with T and C with G (MedlinePlus Genetics, NLM, 2021). The cell opens the DNA at starting points and lays down short RNA primers, from which new strands grow. This is necessary because each new cell needs an exact copy of the DNA when it divides (MedlinePlus Genetics, NLM, 2021).

Which enzymes are needed for DNA replication?

Helicase gets things started. In eukaryotes, this is the CMG complex, made up of three factors, which unwinds the double helix using energy from ATP (Burgers and Kunkel, 2017). The Pol α-primase complex lays down the primer. In yeast cells, Pol ε builds mainly the leading strand, while Pol α and Pol δ build the lagging strand (Burgers and Kunkel, 2017). Finally, DNA ligase 1 joins the Okazaki fragments (Burgers and Kunkel, 2017).

How fast does DNA replication occur?

Measurements are available for the lagging-strand polymerase: under saturating conditions, that is, with an excess of building blocks, Pol δ manages about 250 nucleotides per second (Burgers and Kunkel, 2017). Under conditions like those in the cell, with fewer building blocks and competing RNA building blocks, it manages about 50 nucleotides per second (Burgers and Kunkel, 2017). This value matches the speed at which the replication fork moves in the cell.

When in the cell cycle does DNA replication take place?

During the S phase. Preparation already takes place in the preceding G1 phase, when pre-replication complexes form at thousands of replication origins (Fragkos et al., 2015). At the transition from the G1 to the S phase, the helicase is activated and DNA synthesis begins (Fragkos et al., 2015). However, only some of the origins are used in each S phase (Fragkos et al., 2015).

What are Okazaki fragments?

Okazaki fragments are the short DNA segments that make up the lagging strand; each begins with its own RNA primer (Burgers and Kunkel, 2017). They are named after the Okazaki couple, who proposed the model for their synthesis in 1968 (Burgers and Kunkel, 2017). Millions are needed for eukaryotic chromosomes (Burgers and Kunkel, 2017). Half of the genome is replicated discontinuously in this way (Smith and Whitehouse, 2012).

Next step

Your DNA variants from a saliva sample

The NutriCare | The INFINITY DNA test reads specific positions in your DNA using SNP genotyping. What it cannot do is covered in the chapter on limitations.

Go to NutriCare | INFINITY DNA test How a DNA test reads variants from a saliva sample From sample to result, step by step

Read more

Sources

  1. Burgers P. M. J., Kunkel T. A.: Eukaryotic DNA Replication Fork, Annual Review of Biochemistry (as of 2017) – europepmc.org
  2. Fragkos M. et al.: DNA replication origin activation in space and time, Nature Reviews Molecular Cell Biology (as of 2015) – europepmc.org
  3. Bębenek A., Ziuzia-Graczyk I.: Fidelity of DNA replication – a matter of proofreading, Current Genetics (as of 2018) · europepmc.org
  4. National Library of Medicine, MedlinePlus Genetics: What is DNA? (as of 2021) – medlineplus.gov

[1] documents the process at the replication fork, including the figures for the primer and Pol δ. Replication origins and the cell cycle are based on [2]. Proofreading and orders of magnitude come from [3], while base pairing and the purpose of copying come from [4]. Smith and Whitehouse (Nature, 2012) report the proportion of the lagging strand in the genome; the information on MTHFR comes from MedlinePlus Genetics (NLM, 2019) and Raghubeer and Matsha (Nutrients, 2021). Product information reflects the status as of 03.10.2026; all sources were accessed on 03.10.2026.

mybody®x (MYBODY Lab GmbH) Certificate / quality seal

mybody®x Editorial and Expert Team

DNA replication Molecular biology

The mybody®x editorial and expert team explains the basics of genetics based on reviews and official sources; every figure in this article is accompanied by a source and year. Learn more about the team on the author page.

Published on 03.10.2026 · Last updated on 03.10.2026

The content is for general information and does not replace medical advice, diagnosis, or treatment.

Our products are not intended to diagnose, treat, cure, or prevent diseases.

DNA analysis is intended to provide nutrition and lifestyle guidance. It is not a diagnostic procedure, does not predict disease, and does not replace a medical examination or consultation. Genetic variants describe probabilities in population groups, not certainties for individuals.

The images in this article were created using AI; the people depicted are not real.

mybody®x (MYBODY Lab GmbH) Certificate / quality seal

Read more

Latest posts

View all