Okazaki Fragment Maturation and DNA Ligase
Every time a human cell divides, it must stitch together roughly 50 million short DNA stitches on its lagging strands. Each stitch begins life as an RNA-tipped fragment that is wrong — chemically and informationally. It must be corrected before the genome is complete. This correction process, called Okazaki fragment maturation, is one of the most elegant quality-control operations in molecular biology. It converts a strand built of contaminated pieces into one continuous, error-checked, entirely DNA polymer. Understanding it matters because failure to mature fragments properly causes chromosome breaks, replication stress, and genome instability. That same instability underlies cancer and premature aging phenotypes. The problem exists because of an asymmetry built into Leading and Lagging Strand Synthesis. The leading strand is synthesized continuously in the same direction as the replication fork moves. But the lagging strand must be built discontinuously, backward relative to fork movement, in short pieces called Okazaki fragments. Each fragment starts with an RNA primer laid down by Primase and the RNA Primer. This happens because no DNA polymerase can begin synthesis de novo. That primer is essential for initiation but toxic if left in the final product. RNA is chemically less stable than DNA and does not belong in a permanent genome. This article explains how cells solve that problem. It covers how RNA primers are recognized and removed, and how the resulting gap is filled with DNA. It also explains how the last phosphodiester bond is sealed by DNA ligase to produce one unbroken strand. The mechanism involves a precise handoff between three replicative enzymes. These are DNA polymerase δ, flap endonuclease 1 (FEN1), and DNA ligase I in eukaryotes. Bacteria use an analogous but mechanistically distinct system (Pol I and DNA ligase). The story is also a triumph of classical biochemistry. The discovery of Okazaki fragments themselves, the isolation of DNA ligase, and the genetic dissection of maturation pathways in yeast reveal something important. Together they show how experiments at the bench converted a theoretical necessity into a fully resolved molecular pathway. This sets up the fuller picture of The Replisome: An Integrated Replication Machine.
- Lagging-strand DNA is made as short Okazaki fragments, each starting with an RNA primer that must be removed before ligation.
- In eukaryotes, Pol δ displaces the primer into an RNA/DNA flap, which FEN1 cleaves; DNA ligase I seals the final nick.
- In bacteria, DNA polymerase I removes RNA via 5′→3′ exonuclease activity while simultaneously filling the gap with DNA.
- DNA ligase catalyzes phosphodiester bond formation using ATP (eukaryotes/archaea) or NAD⁺ (most bacteria) as the energy source.
- Okazaki's original pulse-chase experiments in the 1960s proved discontinuous synthesis by detecting short, radioactively labeled DNA fragments.
- Defects in FEN1 or ligase I cause fragment accumulation, replication stress, and genome instability linked to cancer predisposition.
Every Okazaki fragment is deliberately built wrong — primed with RNA — so that a dedicated maturation pathway of strand displacement, flap cleavage, and ligation can proofread and seal the lagging strand into continuous, error-free DNA.
The core idea
Okazaki fragment maturation is the multistep process that converts a discontinuous lagging strand — a series of RNA-primed DNA fragments — into one continuous, ligated DNA polymer. It has three obligatory phases: primer removal, in which the RNA primer (and sometimes a short stretch of the first DNA synthesized) is excised; gap filling, in which DNA polymerase extends the upstream fragment to replace the excised material with proper DNA; and ligation, in which DNA ligase forms the final phosphodiester bond that joins the newly filled gap to the downstream fragment, eliminating the last nick in the sugar-phosphate backbone. The central molecular trick that makes this efficient, rather than requiring a separate enzyme to nibble away RNA base by base, is strand displacement synthesis. As the polymerase synthesizing a new Okazaki fragment reaches the 5′ end of the previous fragment's RNA primer, it does not stop — it continues synthesizing DNA and physically displaces the old primer, pushing it up into a single-stranded flap. This flap-generating mechanism is shared, in modified forms, across bacteria, archaea, and eukaryotes, though the enzymes that then cleave the flap differ. Terminologically, students should distinguish: a primer (RNA, ~8–12 nucleotides in eukaryotes, made by primase/Pol α); a flap (the displaced single-stranded RNA or RNA-DNA segment); an endonuclease (an enzyme that cleaves within a strand, as opposed to an exonuclease that trims from an end); and a nick (a break in one strand's phosphodiester backbone with no missing nucleotides, as opposed to a gap, which is missing one or more nucleotides). Ligase seals nicks, not gaps — a subtlety that matters because ligase cannot act until the gap-filling polymerase has completed the strand right up to the adjacent fragment.
Mechanism
Step 1 — Priming and fragment synthesis
On the lagging strand, primase (as part of the Pol α-primase complex in eukaryotes) synthesizes a short RNA primer roughly 8–12 nucleotides long. Pol α extends this with about 20–30 nucleotides of DNA, and then the sliding clamp (PCNA in eukaryotes, the β clamp in bacteria) is loaded, allowing the more processive Pol δ (eukaryotes) or Pol III (bacteria) to take over and extend the fragment. In humans, Okazaki fragments are roughly 100–200 nucleotides long; in E. coli, they are longer, roughly 1,000–2,000 nucleotides, reflecting differences in the priming frequency of the two replisomes.
Step 2 — Strand displacement and flap creation
Synthesis continues until the elongating polymerase reaches the 5′ end of the downstream Okazaki fragment (the previously made fragment, whose RNA primer sits at its 5′-most end). Rather than halting, the polymerase's strand-displacement activity peels the old primer and often a short piece of old DNA off the template, creating a single-stranded 5′ flap that projects outward from the duplex.
Step 3 — Flap cleavage
This is where eukaryotes and bacteria diverge mechanistically.
In eukaryotes, the flap is cleaved by flap endonuclease 1 (FEN1), a structure-specific nuclease that recognizes the single-stranded 5′ flap (not its sequence) and cleaves it at its base, leaving a nick. FEN1 works iteratively with Pol δ: short flaps (a few nucleotides) are cleaved directly by FEN1; if a longer flap forms, it can be coated by the single-stranded DNA-binding protein RPA, which blocks FEN1 and instead recruits the helicase/nuclease Dna2, which trims the flap down to a short stub that FEN1 then finishes cleaving. This two-nuclease "long-flap pathway" is a backup that prevents flaps from folding into secondary structures that would otherwise resist cleavage.
In bacteria, DNA polymerase I performs both the strand-displacement synthesis and the removal in one enzyme: its 5′→3′ exonuclease domain degrades the RNA primer nucleotide by nucleotide just ahead of the polymerase domain's synthesis, a process sometimes called nick translation because the nick appears to move along the DNA as RNA is degraded and replaced with DNA. Pol I is uniquely suited for this dual role; no separate flap endonuclease is generally required, though bacteria do possess RNase H, which can degrade RNA within RNA-DNA hybrids and assist primer removal.
Step 4 — Ligation
Once FEN1 (or Pol I's exonuclease activity) leaves a clean nick — a 3′-OH on the upstream fragment abutting a 5′-phosphate on the downstream fragment, with no missing bases — DNA ligase seals it. Ligase catalyzes formation of the phosphodiester bond in three steps: 1. Ligase reacts with its cofactor (ATP in eukaryotes and archaea; NAD⁺ in most bacteria) to form a covalent ligase-AMP intermediate, releasing pyrophosphate (from ATP) or nicotinamide mononucleotide (from NAD⁺). 2. The AMP group is transferred to the 5′-phosphate at the nick, forming an activated DNA-adenylate (AppDNA) intermediate. 3. The 3′-OH of the adjacent fragment attacks this activated phosphate, forming the new phosphodiester bond and releasing AMP. In human cells this final sealing step is carried out by DNA ligase I, which travels with the replisome and physically interacts with PCNA, ensuring it acts immediately once FEN1 has cleaved the flap — coupling cleavage and ligation so that a nick is never left exposed for long.
Step 5 — Regulation and coordination with the replisome
The lagging strand is not matured in isolation; it operates as one arm of the replisome. PCNA, loaded by the clamp loader RFC, remains on the DNA after Pol δ finishes a fragment and serves as a mobile platform that recruits FEN1 and then ligase I in turn, physically choreographing the handoff so each enzyme acts precisely when needed and then departs. This clamp-centered coordination — sometimes summarized as a "toolbelt" model — minimizes the time a nick or gap is exposed, which matters because unsealed nicks are substrates for unwanted recombination or breakage if the fork encounters them before ligation.
The evidence
Discovery of discontinuous replication — Okazaki, 1968. Before the mechanism could even be studied, its existence had to be proven. Watson and Crick's semiconservative model implied continuous synthesis on both strands, but DNA polymerases can only extend in the 5′→3′ direction, creating a paradox for the strand running antiparallel to fork movement. Reiji Okazaki, Tsuneko Okazaki, and coworkers resolved this in the mid-1960s using E. coli and bacteriophage-infected cells. They performed brief "pulse" labeling with radioactive ³H-thymidine (as short as a few seconds), followed by rapid extraction and alkaline sucrose gradient sedimentation to separate DNA by size. If replication were continuous, even a very short pulse would appear only in long DNA molecules; instead, they found the newly labeled DNA sedimented at a small size, consistent with fragments of roughly 1,000–2,000 nucleotides in bacteria (now called Okazaki fragments). When they extended the labeling period with a "chase" of unlabeled thymidine, the label progressively shifted into larger, high-molecular-weight DNA — showing that the short fragments were genuine replication intermediates, not degradation artifacts, and that they were subsequently joined into long strands. This pulse-chase logic directly proved discontinuous synthesis and gave the fragments their name.
Evidence for RNA primers. Subsequent work showed the 5′ ends of nascent Okazaki fragments contained ribonucleotides. Using RNase-sensitive labeling and characterizing the fragments' 5′ termini biochemically, researchers (including further work from the Okazaki laboratory and others in the 1970s) demonstrated that each fragment begins with a short RNA stretch, consistent with primase activity, and that mature, ligated DNA in genomic strands lacks this RNA — implying an active removal step must exist between synthesis and ligation.
Purification and mechanism of DNA ligase. DNA ligase activity was independently identified in 1967 by several laboratories (including those of I. Robert Lehman, Charles Richardson, Jerard Hurwitz, and Nicholas Cozzarelli), each showing that extracts of bacteria or bacteriophage-infected cells could seal nicks in duplex DNA in vitro. Lehman's group in particular characterized the E. coli enzyme's cofactor requirement for NAD⁺, distinguishing it from mammalian ligases requiring ATP, and demonstrated the ligase-adenylate covalent intermediate biochemically, confirming the three-step mechanism described above. These purified enzymes made in vitro DNA joining possible and were essential tools for the subsequent development of recombinant DNA technology.
Genetic proof that FEN1 and ligase are essential for maturation — yeast studies. The clearest causal evidence that specific nucleases and ligases are required for fragment maturation, rather than merely convenient in vitro activities, came from yeast genetics. Mutants of RAD27 (the yeast gene encoding FEN1) accumulate unligated, duplex DNA fragments with unremoved 5′ flaps, show elevated rates of duplications in repetitive sequences, and are synthetically lethal when combined with mutations affecting Dna2 or ligase, directly implicating FEN1 in the essential removal of RNA-DNA flaps during replication in living cells. Similarly, temperature-sensitive alleles of CDC9 (yeast DNA ligase I) cause an accumulation of unjoined Okazaki fragments at the restrictive temperature, detectable by pulse-chase labeling and alkaline gel electrophoresis, directly linking ligase activity to the final sealing step in vivo — an elegant genetic parallel to Okazaki's original biochemical pulse-chase experiment, now performed with a defined mutant instead of wild-type cells.
Reconstitution in vitro. Perhaps the most rigorous confirmation came from reconstituting the entire maturation pathway with purified components — Pol δ, PCNA, RFC, FEN1, and DNA ligase I — on model substrates containing a nicked or flapped primer-template junction. These experiments, done across several laboratories from the 1990s onward, showed that this minimal set of purified proteins is sufficient to convert a synthetic RNA-primed nicked substrate into fully ligated duplex DNA, confirming that no additional unknown factors are required for the core reaction.
Worked examples
Example 1 — Counting fragments and estimating flux. A cultured human cell has a genome of about 3.2 × 10⁹ base pairs and replicates it with, on average, roughly 200,000 replication origins firing per S phase (in practice cells license many more origins than they use, but assume this active number for this estimate). Because only the lagging strand is discontinuous, roughly half the newly synthesized DNA is made as Okazaki fragments. If each fragment averages 165 nucleotides: Total lagging-strand DNA ≈ 3.2 × 10⁹ / 2 = 1.6 × 10⁹ nucleotides. Number of fragments ≈ 1.6 × 10⁹ / 165 ≈ 9.7 × 10⁶ fragments per genome duplication. This means nearly ten million individual primer-removal-and-ligation events must occur correctly, every S phase, in every dividing human cell — a striking illustration of why FEN1 and ligase I must act with very high fidelity and speed; even a modest per-fragment failure rate would translate into thousands of unrepaired nicks per genome.
Example 2 — Diagnosing a ligase-deficient phenotype. Suppose a research team engineers a yeast strain with a hypomorphic (partially functional) cdc9 (ligase I) allele and grows cells at a semi-permissive temperature. They pulse-label cells for 30 seconds with ³H-thymidine, then chase with unlabeled thymidine for 2, 10, and 30 minutes, and run alkaline gel electrophoresis to track fragment size over time — precisely mirroring Okazaki's classic experiment. In wild-type cells, label shifts almost completely from short fragments (~150 nt peak) into high-molecular-weight DNA within 10 minutes. In the cdc9 mutant, the team observes that even after 30 minutes of chase, a substantial fraction of the label remains at the short-fragment size. What can be concluded? Because the fragments are being synthesized (label incorporation is normal at the pulse stage) but not converted into long DNA, the defect must lie downstream of primer synthesis and gap-filling — that is, specifically in the ligation step, exactly as expected for a ligase mutant. If instead FEN1 were defective, one would predict a related but distinguishable phenotype: fragments that fail to lose their RNA/flap character (detectable by RNase sensitivity of their 5′ ends) and pile up as unjoined but flap-retaining species, sometimes with characteristic small duplications from flap mis-processing. This reasoning — matching a specific biochemical block to a specific gel/kinetic signature — is exactly how the original researchers assigned distinct genes to distinct steps of the maturation pathway.
Common misconceptions & traps
| Misconception | Why it's wrong |
|---|---|
| "Ligase joins the RNA primer directly to the DNA fragment." | Ligase seals only DNA-DNA nicks with matching sugar-phosphate chemistry; RNA must first be entirely removed and replaced with DNA by Pol δ/Pol I before ligation can occur. |
| "FEN1 degrades RNA the same way RNase H does, nucleotide by nucleotide from the end." | FEN1 is a structure-specific endonuclease that recognizes flap geometry and typically cleaves at the flap's base in one or a few cuts, not a processive exonuclease. |
| "Okazaki fragments are a bacterial phenomenon; eukaryotic replication is continuous on both strands." | Discontinuous lagging-strand synthesis with Okazaki fragments occurs in all cellular life; eukaryotic fragments are simply shorter (~100–200 nt vs ~1,000–2,000 nt in bacteria). |
| "Ligase is only needed at the end of replication, to join the first and last fragment of the chromosome." | Ligase acts continuously throughout S phase, sealing millions of nicks — one after essentially every Okazaki fragment matures, not just at the terminus. |
| "The leading strand needs no primer at all." | The leading strand also requires one RNA primer to initiate synthesis at the origin; it simply does not require repeated re-priming, unlike the lagging strand. |
The scientist behind it
Reiji Okazaki (1930–1975) was a Japanese molecular biologist at Nagoya University who, together with his wife and close scientific collaborator Tsuneko Okazaki, designed the pulse-chase labeling experiments that first revealed discontinuous DNA replication. Working with E. coli in the mid-1960s, the Okazakis combined careful radiochemistry with sedimentation analysis to detect replication intermediates invisible to prior, cruder assays. Reiji Okazaki died in 1975 from leukemia attributed to radiation exposure from the atomic bombing of Hiroshima, which he survived as a child; Tsuneko Okazaki continued the research program for decades afterward, extending the analysis of fragment size, primer structure, and maturation enzymes. Their surname became permanently attached to the fragments they discovered, an unusually direct form of scientific memorial.
Origins · 1 minReiji Okazaki and how this was discovered
Where this matters today
Cancer biology and genome instability. Cells with reduced FEN1 activity accumulate unresolved flaps and double-strand breaks, and FEN1 variants have been linked to increased cancer susceptibility in mouse models; conversely, some tumor cells become dependent on backup maturation pathways, making FEN1 and ligase I candidates for synthetic-lethal cancer drug targeting, especially in tumors with other replication-stress vulnerabilities.
Antibiotic and antiviral development. Because bacterial DNA ligase (NAD⁺-dependent) is structurally distinct from human ligases (ATP-dependent), it is an attractive selective antibacterial target; several experimental compounds have been designed to inhibit bacterial ligase specifically without affecting human ligase I or III.
Biotechnology and molecular cloning. Purified DNA ligase, first isolated in these very 1967 studies, became one of the indispensable enzymatic tools of recombinant DNA technology, used routinely to join restriction fragments, construct plasmids, and, more recently, to assemble synthetic genes and CRISPR donor constructs.
Diagnostics. Ligase-based assays (ligase chain reaction, ligation-dependent probe amplification) exploit the exquisite specificity of ligase for perfectly matched, nick-adjacent ends to detect single-nucleotide mutations in clinical genotyping, directly repurposing the enzyme's natural substrate requirement for medical diagnosis.
Aging and replicative stress research. Because millions of maturation events must succeed every cell cycle, accumulating defects in this pathway are studied as contributors to replicative senescence and age-related genome instability, linking a seemingly narrow biochemical step to organismal aging phenotypes.
Ask the tutor: How does the replisome physically coordinate leading- and lagging-strand synthesis so that fragment maturation keeps pace with the moving fork?
Why can't DNA polymerase simply start each Okazaki fragment without an RNA primer?
All known DNA polymerases can only add nucleotides to an existing 3′-OH group; none can initiate synthesis on a bare single-stranded template. Primase, a specialized RNA polymerase, can start a chain de novo because RNA polymerases lack this restriction. So every Okazaki fragment — and the leading strand at the origin — needs a short RNA primer to create the initial 3′-OH that DNA polymerase then extends. This is a direct consequence of DNA polymerase's proofreading mechanism, which depends on having a correctly paired existing base to check before adding the next one; starting from nothing would remove that safeguard.
What would happen if FEN1 were completely nonfunctional in a human cell?
Without FEN1, displaced 5′ flaps generated during strand-displacement synthesis would never be cleaved, so nicks would remain unsealed and unligated across the genome. Persistent flaps can also fold into secondary structures or be extended further, sometimes causing small duplications or triplications when repetitive sequence is present. Because ligase requires a clean nick, accumulated flaps also block ligation entirely at those sites. In practice, complete loss of FEN1 is lethal in mammalian cells; partial loss-of-function is tolerated but linked to genome instability and increased cancer risk, since the backup Dna2-dependent long-flap pathway can only partially compensate.
Why are eukaryotic Okazaki fragments shorter than bacterial ones?
Eukaryotic fragments average about 100–200 nucleotides, roughly matching the length of DNA wrapped around one nucleosome, whereas bacterial fragments run about 1,000–2,000 nucleotides. This difference largely reflects priming frequency: eukaryotic replisomes reprime the lagging strand far more often, partly because chromatin structure and nucleosome spacing appear to influence where primase acts efficiently. The shorter fragment length in eukaryotes also means proportionally more RNA primers and more maturation events per genome, which is one reason FEN1 and ligase I activity must be extremely efficient and tightly coupled to Pol δ in eukaryotic cells.
Is Okazaki fragment maturation the same in archaea as in bacteria and eukaryotes?
Archaea generally use a eukaryotic-like maturation pathway rather than a bacterial one. Archaeal genomes typically encode a FEN1 homolog and an ATP-dependent DNA ligase, and archaeal replication machinery (including PCNA-like clamps) is evolutionarily closer to the eukaryotic replisome than to the bacterial one. This supports the broader phylogenetic picture that eukaryotic replication proteins arose from an archaeal ancestor, with bacteria representing an independently evolved, though functionally convergent, replication system — a point often illustrated using replisome-protein phylogenetic trees.
How is DNA ligase distinct from DNA polymerase, mechanistically?
DNA polymerase catalyzes formation of new phosphodiester bonds by adding nucleoside triphosphates to a growing 3′-OH end, using the energy of pyrophosphate release, and requires a template to select the correct incoming base. DNA ligase does not add any new nucleotides at all; it joins two pre-existing DNA ends across an already-correct, fully base-paired nick, using ATP or NAD⁺ hydrolysis to activate the 5′-phosphate before bond formation. In short, polymerase builds the chain forward one base at a time, while ligase performs a single, one-time sealing reaction between two already-adjacent ends.
Why does the cell tolerate the inefficiency of making, then destroying, RNA primers rather than starting each fragment directly with DNA?
No polymerase-catalyzed synthesis can initiate a new chain de novo; every known DNA polymerase needs a pre-existing 3′-OH to extend. Using RNA primase for initiation exploits a different enzyme class capable of de novo synthesis, then leverages the requirement for a temporary marker that is easy to recognize and remove later — RNA is chemically distinguishable from DNA (it contains ribose and uracil rather than deoxyribose and thymine), so downstream nucleases and polymerases can specifically target and replace it. This built-in "self-destruct tag" system is arguably more robust than any hypothetical DNA-only initiation mechanism would be, because errors made during initiation are automatically discarded when the primer is excised.
What experimental technique first made it possible to detect Okazaki fragments, and why was it necessary?
Alkaline sucrose gradient centrifugation was essential because Okazaki fragments are held together by hydrogen bonding to the template strand and by proximity to neighboring fragments in the double helix, so a normal (neutral) gradient would not separate them from larger DNA. Under alkaline conditions, DNA strands are denatured and hydrogen bonds disrupted, allowing individual single strands — including short, newly synthesized fragments — to be separated purely by size via sedimentation rate. Combined with brief radioactive pulse-labeling, this technique let the Okazakis directly visualize a population of small, newly made DNA molecules, providing the first physical evidence for discontinuous replication.