Does DNA Replication Occur in Mitosis?

Bulk DNA replication does not occur during mitosis. The copying of a cell’s nuclear genome takes place earlier, during a dedicated window called S phase, and is normally finished well before a cell enters mitosis to divide. However, cells sometimes arrive at mitosis with small stretches of DNA still uncopied, and researchers have discovered that human cells possess an emergency backup system, called mitotic DNA synthesis or MiDAS, that finishes the job during mitosis itself. This exception turns what seems like a straightforward textbook answer into a more interesting story about how cells cope when their replication schedule runs long.

Why the Cell Cycle Separates Copying From Dividing

A dividing cell moves through a fixed sequence of stages, always in the same order: G1, S, G2, and M (mitosis).¹ During G1, the cell grows and prepares the molecular machinery it will need to copy its DNA. S phase is where the actual genome duplication happens, with thousands of replication origins firing across chromosomes to produce two complete copies of every DNA molecule. G2 serves as a gap between copying and dividing, during which the cell checks for errors and repairs damage. Only then does the cell enter M phase and physically split its duplicated chromosomes into two daughter cells.

This separation makes biological sense. Duplicating roughly six billion base pairs of DNA is an enormous, error-prone task. The cell invests heavily in keeping the copying machinery and the dividing machinery apart in time. Once mitosis begins, chromosomes condense into tightly packed structures, most transcription factors are kicked off the DNA, and gene activity drops dramatically.² Histone modifications change, and the whole chromatin landscape reorganizes.³ This highly condensed, transcriptionally silent environment is not designed for large-scale DNA synthesis. It is designed for one thing: accurately pulling chromosomes apart.

When the Copying Isn’t Done in Time

In an ideal cell cycle, every last stretch of DNA is fully copied before mitosis starts. In reality, cells frequently enter mitosis with replication not quite finished.⁴ Certain parts of the genome are inherently harder to replicate. Some regions contain repetitive sequences that cause the replication machinery to stall or slow down. Others are so large and gene-poor that the replication forks traveling through them simply run out of time. These trouble spots are known as common fragile sites, and they are among the most frequent places where the genome is left under-replicated when mitosis arrives.⁵

Sometimes the problem isn’t the DNA sequence itself but rather stress on the replication process. Cells can experience what researchers call replication stress, where the normal pace of copying is disrupted by things like limited building blocks for new DNA, collisions between the replication machinery and other processes on the chromosome, or exposure to certain chemicals that slow replication forks. Under these conditions, even regions that normally replicate without trouble can be left unfinished.

The cell faces a dilemma. Checkpoint mechanisms exist to delay mitosis until replication is complete, but these checkpoints are not perfect. They can be overwhelmed, or the amount of remaining unreplicated DNA can be small enough to slip past the surveillance system. When a cell commits to mitosis with gaps in its genome, it needs a plan B.

MiDAS, the Emergency Replication Program

That plan B is mitotic DNA synthesis. MiDAS was first characterized as an unusual form of DNA replication that takes place during mitosis, initially linked to common fragile sites in cells under replication stress.⁶ Over time, researchers recognized it as a more widespread salvage mechanism that can be called upon to finish duplicating any under-replicated region of the genome.⁷

Genome-wide mapping studies have provided strong confirmation that MiDAS targets the same regions that failed to finish replicating during S phase. When researchers tracked which parts of the genome were still being copied very late in S phase, those regions matched precisely with the spots where MiDAS was detected once cells entered mitosis.⁸ The system is not random. It zeroes in on whatever was left unfinished.

Importantly, MiDAS is not simply a continuation of normal S-phase replication using the same tools. It relies on a distinct set of proteins. The process requires an enzyme called MUS81-EME1, which cuts DNA structures at stalled replication sites, and a subunit of a specialized DNA-copying complex called POLD3.⁹ This dependency on POLD3 and other factors associated with a repair-like copying process called break-induced replication suggests that MiDAS is mechanistically closer to a DNA repair pathway than to conventional replication. The cell is essentially using its damage-response toolkit to complete a task that its normal replication machinery failed to finish on schedule.

Different Flavors of Mitotic DNA Repair

Research over the past few years has revealed that the molecular details of MiDAS are not identical in every cell type. In normal, untransformed human cells, MiDAS at common fragile sites is driven primarily by a protein called FANCD2, which is part of the Fanconi anemia DNA repair pathway. This version of MiDAS appears to preserve chromosomal stability at fragile sites quite effectively.¹⁰

Cancer cells, by contrast, rely on a different protein, RAD52, to drive their mitotic DNA synthesis. This alternative version avoids lethal chromosome segregation errors but comes at a cost: it sacrifices the stability of common fragile sites. In other words, cancer cells keep themselves alive through mitosis but accumulate more chromosomal damage in the process.¹⁰ This distinction between the FANCD2-driven pathway in healthy cells and the RAD52-driven pathway in cancer cells is a relatively recent finding and has generated considerable interest among researchers studying tumor biology.

Beyond these two main routes, a newer pathway has entered the picture. A repair process called theta-mediated end joining (TMEJ), carried out by the DNA polymerase theta (Pol theta), also operates during mitosis to handle replication-stress-induced DNA damage. In experiments where researchers depleted Pol theta, cells showed significantly more nuclear envelope defects after dividing, roughly 39% compared to about 22% in control cells under the same replication stress conditions.¹¹ Strikingly, this TMEJ pathway appeared to play a more important role in preventing nuclear envelope problems than the classic MUS81-POLD3 pathway, suggesting that cells have layered multiple independent repair strategies on top of each other to handle unfinished business during mitosis.

What Happens When the Safety Nets Fail

If MiDAS and related mitotic repair pathways cannot resolve under-replicated regions, the consequences can be severe. When chromosomes that are still physically connected by unreplicated DNA try to separate during anaphase, they can form ultrafine bridges: thin threads of DNA stretched between the two daughter cells. If these bridges break, the resulting DNA damage shows up as lesions in the next cell cycle. Breakage of these bridges has been shown to cause chromosome fusions through an error-prone repair process, contributing to the kind of genome reshuffling seen in many cancers.¹²

Another potential outcome involves micronuclei, small packages of DNA that get left outside the main nucleus after a problematic cell division. DNA trapped inside micronuclei does not replicate or repair properly because the micronuclear environment is defective. This can lead to massive, localized DNA shattering, a phenomenon called chromothripsis, where a chromosome is fragmented and reassembled in a scrambled order.¹³ Chromothripsis can also be triggered more directly when cells with still-replicating DNA are forced into premature chromosome condensation.¹⁴ These dramatic rearrangements are increasingly recognized as a major driver of cancer evolution, capable of reshaping a genome in a single catastrophic event rather than through the slow accumulation of individual mutations.

The picture that emerges is that MiDAS is not just a biochemical curiosity. It is a front-line defense against genome instability. Without it, the relatively minor problem of a few unreplicated stretches of DNA can cascade into chromosome breakage, abnormal segregation, and the kind of genetic chaos that fuels tumor development.

Why Cancer Researchers Care About Mitotic DNA Synthesis

Cancer cells experience far more replication stress than normal cells. Oncogenes that drive uncontrolled proliferation also push the replication machinery harder and faster, generating more stalled forks and more under-replicated DNA. This means cancer cells depend on MiDAS more heavily than healthy cells do, which creates a potential vulnerability.

Researchers have noted that RAD52, the protein that cancer cells lean on for MiDAS, could be a drug target. Because RAD52-driven MiDAS helps cancer cells survive the replication stress caused by their own oncogenes, blocking RAD52 might selectively kill cancer cells while sparing normal cells that use the FANCD2-driven pathway instead.¹⁵ This idea fits into a broader strategy in cancer therapy: rather than attacking cancer cells directly, you attack the repair pathways they depend on to survive their own genetic instability.

More recently, researchers have identified another angle. The CIP2A-TOPBP1 axis, a signaling system that helps coordinate mitotic DNA repair, turns out to be important for the survival of cells that lack the BRCA1 or BRCA2 tumor-suppressor genes.¹⁶ BRCA-deficient cancers already have compromised DNA repair, and their reliance on mitotic repair pathways could represent another exploitable weakness. These findings are still in the early stages of translational research, but they highlight how a basic cell biology question about when DNA gets replicated has direct implications for cancer treatment strategies.

An Ancient Problem With an Ancient Solution

MiDAS is not unique to human cells. Budding yeast, one of the simplest organisms used in genetics research, routinely finish replicating their DNA during anaphase, after chromosome segregation has already begun. Researchers have noted striking parallels between yeast anaphase replication and human MiDAS. Both depend on the same family of proteins (the Pol32/POLD3 subunit of DNA polymerase delta), and neither requires the Rad51 recombination protein, suggesting both use a form of break-induced replication.¹⁷

The fact that organisms separated by over a billion years of evolution use mechanistically similar strategies for finishing replication during cell division suggests this is not a fluke or a sign of dysfunction. It may reflect a deeply conserved solution to an inherent challenge: some parts of any genome are just hard to copy on time, and cells have always needed a last-resort copying mechanism during division. Rather than evolving ever more stringent checkpoints that would halt division and slow down growth, cells evolved a way to patch things up on the fly.

Mitochondrial DNA Doesn’t Follow the Rules

Everything discussed so far applies to nuclear DNA, the chromosomes inside the cell’s nucleus. Mitochondria, the energy-producing compartments scattered throughout the cell, carry their own small circular genome, and its replication follows completely different timing rules. Mitochondrial DNA replication is not coupled to the cell cycle the way nuclear DNA replication is. It happens throughout G1, S, G2, and mitosis, and continues even in cells that have stopped dividing entirely.¹⁸

This independence makes sense given that each cell contains hundreds or thousands of mitochondria, each with its own copies of mtDNA. Mitochondrial DNA copy number scales with cell size and energy demand rather than with the need to divide, so the replication machinery in mitochondria runs on its own schedule, responding to metabolic signals rather than cell-cycle cues.¹⁹ So in a literal sense, yes, some DNA replication does occur during mitosis in every dividing cell, just not the nuclear genome replication that biology courses are usually talking about when they ask the question.

How Researchers Actually See MiDAS Happening

Detecting DNA synthesis during mitosis is technically challenging because the amount of copying is tiny compared to S-phase replication. Researchers use a labeled nucleotide analog, typically EdU, that gets incorporated into newly made DNA and can then be visualized as bright spots on chromosomes using fluorescent chemistry. By adding EdU only during mitosis and then examining metaphase chromosome spreads, scientists can pinpoint exactly where on the genome MiDAS is occurring.²⁰ These EdU foci often co-localize with markers of DNA damage on mitotic chromosomes, confirming that the synthesis is happening at sites of unresolved replication problems.¹¹

Most experiments that study MiDAS use a drug called aphidicolin to deliberately slow replication forks, which amplifies the number of under-replicated regions entering mitosis and makes MiDAS events frequent enough to study systematically. Under normal conditions in healthy cells, MiDAS events are rarer and harder to detect. This has led to some debate about how often MiDAS occurs in truly unperturbed cells versus how much of the published data reflects artificially elevated replication stress. The consensus, though, is that even without drug treatment, some level of under-replication at difficult loci is a routine feature of normal cell division, and MiDAS is a physiologically relevant process rather than just a laboratory artifact.

1
PubMed Central. A Journey through Time on the Discovery of Cell Cycle Regulation
2
PubMed Central. Transcriptional repression across mitosis: mechanisms and functions
3
PubMed Central. Mitotic chromosome condensation resets chromatin to safeguard transcriptional homeostasis during interphase
4
PubMed. TTF2 processes sites of incomplete DNA replication during mitosis via sister-chromatid exchanges
5
Molecular Cell. Mitotic DNA synthesis and its roles in genome stability – Section: Regions of the genome most dependent on MiDAS
6
Molecular Cell. Mitotic DNA synthesis and its roles in genome stability
7
Nature Communications. Mitotic DNA synthesis in response to replication stress requires the sequential action of DNA polymerases zeta and delta in human cells
8
Cell Research. High-resolution mapping of mitotic DNA synthesis regions and common fragile sites in the human genome through direct sequencing
9
PubMed. RAD52 Facilitates Mitotic DNA Synthesis Following Replication Stress
10
PubMed Central. Mitotic DNA synthesis in untransformed human cells preserves common fragile site stability via a FANCD2-driven mechanism that requires HELQ
11
Nature Communications. Mitotic DNA repair by TMEJ suppresses replication stress-induced nuclear envelope reassembly defect
12
PubMed Central. A new class of ultrafine anaphase bridges generated by homologous recombination
13
PubMed. Impaired nuclear functions in micronuclei results in genome instability and chromothripsis
14
PubMed. Stress induced by premature chromatin condensation triggers chromosome shattering and chromothripsis at DNA sites still replicating in micronuclei or multinucleate cells when primary nuclei enter mitosis
15
Molecular Cell. RAD52 Promotes Mitotic DNA Synthesis in Human Cells – Section: Discussion
16
Nature Communications. The CIP2A-TOPBP1 axis facilitates mitotic DNA repair via MiDAS and MMEJ
17
Nature Communications. Budding yeast complete DNA synthesis after chromosome segregation begins
18
PubMed. Replication of mitochondrial DNA occurs throughout the mitochondria of cultured human cells
19
Nature Structural & Molecular Biology. Regulation with cell size ensures mitochondrial DNA homeostasis during cell growth
20
PubMed. Inducing and Detecting Mitotic DNA Synthesis at Difficult-to-Replicate Loci