The three main steps of DNA replication are initiation, elongation, and termination. During initiation, the double helix unwinds and separates. During elongation, enzymes build new strands of DNA using each original strand as a template. During termination, the process wraps up and two complete DNA molecules are released. Each step involves a different cast of molecular machinery, and the whole process happens with remarkable speed and accuracy.
Step 1: Initiation
Before DNA can be copied, the two intertwined strands of the double helix have to be pulled apart. Special initiator proteins bind to a specific starting point on the DNA, called an origin of replication, and pry the two strands open by breaking the weak bonds holding the paired bases together. This creates a small bubble of exposed, single-stranded DNA.
Once that bubble forms, an enzyme called helicase is loaded onto the exposed strand. Helicase travels along the DNA, continuing to unzip the double helix ahead of it and creating a Y-shaped structure known as the replication fork. As the strands separate, single-strand binding proteins coat the exposed DNA to keep it from snapping back together or getting damaged. Meanwhile, another enzyme relieves the tension that builds up ahead of the fork as the helix unwinds, preventing the DNA from getting tangled.
In bacteria like E. coli, there is a single origin of replication on the circular chromosome. Human cells, which have much longer linear chromosomes, fire thousands of origins simultaneously so the job can be finished in a reasonable timeframe.
Step 2: Elongation
Elongation is where the actual copying happens. DNA polymerase, the central enzyme of replication, reads each exposed template strand and adds matching nucleotides (the building blocks of DNA) one at a time. There’s an important constraint: DNA polymerase can only build a new strand in one direction, from the 5′ end to the 3′ end. This creates an asymmetry at the replication fork that makes one strand much simpler to copy than the other.
The “leading strand” points in the easy direction. DNA polymerase latches on and synthesizes it continuously, racing along behind the helicase as the fork opens. The “lagging strand” faces the opposite way, so it has to be built in short, backward segments. Each segment starts with a small RNA primer laid down by an enzyme called primase, which gives DNA polymerase a starting grip. The polymerase then extends the primer with DNA until it bumps into the previous segment.
These short segments on the lagging strand are called Okazaki fragments. In eukaryotes, each one begins with a primer of roughly 7 to 10 RNA nucleotides extended to about 20 to 30 bases total. Once a fragment is complete, the RNA primer from the previous fragment is displaced and cut away, replaced with DNA, and the remaining gap is sealed by an enzyme called DNA ligase. This stitching process repeats hundreds of thousands of times across the genome during a single round of replication.
The speed difference between organisms is striking. In E. coli, DNA polymerase adds about 1,000 nucleotides per second. In human cells, the rate is closer to 50 nucleotides per second, but because human cells use many replication forks working in parallel, the entire genome is still copied within hours.
Step 3: Termination
Termination is the least dramatic step, but it poses different challenges depending on whether the DNA is circular (as in bacteria) or linear (as in human chromosomes).
In E. coli, two replication forks travel in opposite directions around the circular chromosome from a single origin. They meet in a designated termination zone on the far side of the circle, which contains special DNA sequences called ter sites. These sites act as one-way gates: a fork can enter the zone but cannot pass through and leave. A protein binds to these sites and stalls each fork, ensuring they converge in a controlled manner rather than overshooting and re-replicating DNA that’s already been copied. Once the two forks meet, the final stretch of DNA between them is filled in, the strands are sealed, and the two resulting circular chromosomes are separated.
In human cells with linear chromosomes, termination happens when a replication fork simply reaches the end of the chromosome. The leading strand polymerase appears to extend to within a few nucleotides of the chromosome tip before the replication machinery slides off. The real challenge here involves the very ends of chromosomes, called telomeres. Because the lagging strand always needs a primer upstream to get started, there’s a small stretch at the tip that can’t be fully copied. Telomeres, which are long repeating sequences of non-coding DNA, act as a disposable buffer so that no important genetic information is lost. Over many rounds of cell division, telomeres gradually shorten.
How Errors Are Caught
DNA replication is fast, but it’s also astonishingly accurate. DNA polymerase has a built-in proofreading ability: after adding each nucleotide, it checks whether the base pair is correct. If a mismatch is detected, the enzyme reverses, removes the wrong nucleotide, and tries again. Without proofreading, the polymerase would make roughly one error for every 10,000 to 100,000 nucleotides it adds. Proofreading improves accuracy by 10- to 100-fold.
After replication is finished, a second safety net called mismatch repair scans the newly synthesized DNA for errors that slipped past proofreading. Together, these systems reduce the final error rate to an extraordinarily low level. In E. coli, strains that lack proofreading have a mutation rate about 4,000 times higher than normal, which illustrates how critical these quality-control mechanisms are.
Key Enzymes at a Glance
- Helicase: unzips the double helix by traveling along one strand and prying the two strands apart
- Primase: lays down short RNA primers that give DNA polymerase a starting point, especially important on the lagging strand
- DNA polymerase: reads the template strand and adds matching nucleotides to build the new strand, also proofreads its own work
- DNA ligase: seals the gaps between Okazaki fragments on the lagging strand, joining them into one continuous strand
What Happens When Replication Goes Wrong
Defects in the proteins that carry out replication can cause serious health problems. Cancer is the most familiar consequence of unchecked replication errors, but several rare congenital conditions are directly tied to faulty replication machinery.
Mutations in the proteins responsible for initiation cause Meier-Gorlin syndrome, a condition characterized by proportional dwarfism and skeletal abnormalities. Other initiation defects have been linked to a form of immune deficiency involving natural killer cells, which are part of the body’s frontline defense against viruses and tumors.
Problems with elongation enzymes lead to a different set of disorders. Mutations in the gene for one of the main copying polymerases can cause a syndrome involving growth restriction, adrenal gland underdevelopment, and abnormal facial features. Defects in another polymerase involved in lagging-strand synthesis cause a condition with premature aging features, hearing loss, and jaw underdevelopment. Notably, no known congenital diseases have been traced to defects in replication termination, likely because termination relies on fewer specialized proteins.