Molecular Biology

Okazaki Fragments

100-200 nt RNA-primed pieces that build the lagging strand of DNA

Okazaki fragments are short, RNA-primed DNA pieces synthesized on the lagging strand during DNA replication. In E. coli they are 1000-2000 nucleotides long; in eukaryotes 100-200 nt. They exist because DNA polymerase only adds nucleotides 5' to 3', so the strand running antiparallel to the fork direction must be built backward in pieces. Each fragment begins with a ~10 nt RNA primer laid by primase (DnaG in bacteria, Pol α-primase in eukaryotes), is extended by the replicative polymerase, has its primer removed by FEN1 or RNase H, and is sealed by DNA ligase. Reiji and Tsuneko Okazaki demonstrated their existence in 1968 via pulse-chase experiments with tritiated thymidine in E. coli.

  • Bacterial length1000-2000 nt
  • Eukaryotic length100-200 nt
  • Primer~10 nt RNA
  • Fork speed (E. coli)~1000 nt/s
  • DiscoveredOkazaki 1968
  • Sealed byDNA ligase I (eukaryotes)

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

Why Okazaki fragments matter

  • They make replication possible at all. Without a discontinuous strategy on the lagging strand, the antiparallel double helix could not be copied by 5' to 3'-only polymerases. The whole architecture of DNA replication — primase, sliding clamp, single-strand binding protein, helicase coupling — exists to manage these short pieces.
  • They mark the lagging strand for repair. In eukaryotes, the ribonucleotides at the 5' end of each fragment are recognized by RNase H2 and used as strand-discrimination signals during mismatch repair. Mutations in RNase H2 cause Aicardi-Goutières syndrome, a neuroinflammatory disease driven by accumulated RNA-DNA hybrids.
  • Their length is set by chromatin. Smith and Whitehouse (Nature 2012) showed in S. cerevisiae that the median Okazaki fragment is 165 nt — exactly the nucleosomal repeat. CAF-1-mediated histone deposition behind the fork dictates when primase fires.
  • Defects cause cancer. Loss of FEN1 or DNA ligase I produces persistent single-strand nicks; in mouse models, hypomorphic FEN1 alleles cause a 10-fold increase in lymphoma. Cells from Werner syndrome and Bloom syndrome patients accumulate unligated fragments and replicate ~3x slower.
  • They are the key to PCR primers and Sanger sequencing. The molecular logic — short oligonucleotide primer + 3' extension — was directly inspired by lagging-strand biochemistry. Every PCR cycle creates artificial Okazaki-like products before the strands are dissociated.
  • Telomere shortening starts here. The lagging strand at chromosome ends cannot be fully replicated because the final RNA primer has nothing to extend from. This is the end-replication problem, solved in germ cells and stem cells by telomerase, and it loses ~50-100 bp per division in somatic cells.
  • They are the staging ground for SV40 and viral replication. Many tumor virus replication systems (SV40 large T antigen, HSV UL30) reproduce the eukaryotic lagging-strand machinery in vitro, making them the workhorse system used by Bruce Stillman, Tom Kelly, and others to dissect the replisome biochemically since the 1980s.

Common misconceptions

  • Both strands have Okazaki fragments. Only the lagging strand. The leading strand is synthesized as one continuous run from a single primer at the origin. There is, however, occasional re-priming on the leading strand after stalling — modern measurements suggest this happens once per ~50 kb in mammalian cells.
  • The primer is DNA. It is RNA, ~10 nucleotides long, laid by primase. Primase is the only polymerase in the cell that can synthesize DNA-template-directed nucleic acid de novo without a primer of its own — every other polymerase needs a 3'-OH to extend.
  • Length is the same in all organisms. Bacterial fragments (E. coli 1000-2000 nt) are roughly 10x longer than eukaryotic fragments (100-200 nt). Archaea fall in between. The difference reflects fork speed, primase concentration, and chromatin context, not any inherent property of the chemistry.
  • The lagging strand polymerase moves backward. The polymerase moves forward (5' to 3') on its template, but its template runs opposite to the fork direction, so each fragment is built moving away from the fork. The replisome handles this by looping the lagging strand back through the polymerase — the trombone model.
  • Ligase joins fragments at random. Ligase I (eukaryotic) is recruited by PCNA, the sliding clamp left behind on each completed fragment. The ATP-dependent reaction proceeds in three steps — adenylylation of the ligase, transfer of AMP to the nick 5'-phosphate, and nucleophilic attack by the 3'-OH — sealing in roughly 1 ms per nick.
  • Okazaki fragments are only made during S-phase. Mitochondrial DNA replication, plasmid replication, and DNA repair synthesis (e.g., gap-filling after nucleotide excision repair) all produce short primed fragments by the same general logic, even outside the nuclear S-phase context.

How a single Okazaki fragment is born and joined

The cycle starts when the replicative helicase (DnaB in E. coli; CMG = Cdc45-MCM-GINS in eukaryotes) unwinds 1000-2000 bp ahead of the fork, exposing single-stranded lagging-strand template. SSB (bacteria) or RPA (eukaryotes) coats the single strand, preventing reannealing and secondary structure. Primase — DnaG in E. coli or the Pol α-primase complex (POLA1, POLA2, PRIM1, PRIM2) in eukaryotes — docks onto the helicase and synthesizes a short RNA primer of 8-12 nucleotides. In eukaryotes Pol α immediately extends this with another ~20 deoxynucleotides at low fidelity (no proofreading), creating an RNA-DNA hybrid primer of ~30 nt total.

The replicative polymerase then takes over. In E. coli that is Pol III holoenzyme, with its β sliding clamp loaded by the γ clamp loader, extending at ~1000 nt/s. In eukaryotes the lagging-strand polymerase is Pol δ (POLD1 catalytic subunit, PCNA clamp loaded by RFC), extending at ~50 nt/s. Synthesis proceeds until the polymerase strand-displaces into the next downstream Okazaki fragment, peeling its primer up into a single-stranded 5' flap. FEN1 (or RNase H2 followed by FEN1 for short flaps) cleaves the flap exactly at the RNA-DNA junction. DNA ligase I, recruited by PCNA, then seals the final phosphodiester bond using ATP and producing AMP + PPi. The whole cycle — prime, extend, displace, cleave, seal — takes 1-3 seconds per fragment in eukaryotic cells, with ~10 million fragments produced per S-phase in a diploid human cell.

Bacterial vs eukaryotic Okazaki fragment biology

PropertyE. coliEukaryotes (yeast / mammal)
Fragment length1000-2000 nt100-200 nt
Fork speed~1000 nt/s~50 nt/s (mammalian); ~25 nt/s (yeast)
PrimaseDnaG (60 kDa, 1 subunit)Pol α-primase (4 subunits, ~340 kDa total)
Primer length~11 nt RNA~10 nt RNA + ~20 nt DNA (Pol α)
Replicative pol on lagging strandPol III holoenzymePol δ (POLD1)
Sliding clampβ clamp (homodimer)PCNA (homotrimer)
Primer removalPol I 5'-3' exonuclease (nick translation)FEN1 flap endonuclease (+ RNase H2, Dna2)
LigaseNAD+-dependent LigAATP-dependent ligase I
Length determinantPrimase frequencyNucleosome spacing (CAF-1)

Famous experiments

  • Reiji and Tsuneko Okazaki, 1968 (PNAS). Pulse-chased E. coli with [³H]thymidine for 5-30 seconds, then ran the DNA on alkaline sucrose gradients. Found a 7-11S peak (1000-2000 nt) that disappeared into the bulk DNA peak with longer chases — direct evidence of discontinuous synthesis. The same paper showed the short peak persisted in DNA ligase mutants, proving ligase joins the fragments.
  • Sugino, Hirose & Okazaki, 1972. Showed that the 5' end of each nascent fragment is alkali-sensitive — i.e., contains RNA. This was the first direct evidence for RNA primers, before primase was even purified.
  • Rowen & Kornberg, 1978. Purified DnaG primase from E. coli and reconstituted RNA primer synthesis on phi-X174 DNA in vitro. Established that primase is a distinct enzyme, not a side activity of polymerase.
  • Smith & Whitehouse, 2012 (Nature). Used Okazaki fragment sequencing (OK-seq) in S. cerevisiae to map every fragment genome-wide. Median length 165 nt, periodicity matches nucleosome positioning — proved chromatin sets fragment length in eukaryotes.
  • Yeeles et al., 2017 (Cell). Reconstituted the entire eukaryotic replisome from purified yeast components, observed Okazaki fragment synthesis in single-molecule fluorescence at ~150 nt average length, matching in vivo measurements.

Frequently asked questions

Why do Okazaki fragments exist at all?

Every known DNA polymerase adds nucleotides only to a free 3'-OH, so synthesis proceeds strictly 5' to 3'. The two strands of the parental duplex are antiparallel, but the replication fork unwinds in one direction. On the leading strand, the polymerase can chase the fork continuously. On the lagging strand the template runs the wrong way, so the polymerase must repeatedly hop back toward the fork, prime, extend backward away from the fork, then hop again. Each backward run is one Okazaki fragment. The fragments are not a flaw — they are the inevitable consequence of asymmetric polymerase chemistry combined with antiparallel double helices, and ligase glues them together afterward.

How long are Okazaki fragments?

Length is set by how often primase fires relative to fork speed. In E. coli, where the fork advances at roughly 1000 nt/s, each fragment averages 1000-2000 nt — primase initiates roughly once per second per lagging-strand polymerase. In eukaryotes the replication fork moves at only ~50 nt/s but primase fires far more frequently, producing fragments of 100-200 nt that match the spacing of nucleosomes (one fragment per nucleosomal repeat of ~165 bp). The yeast S. cerevisiae and mammalian cells both fall in this range. Archaea sit between the two: ~100-300 nt fragments at fork speeds of a few hundred nt/s.

Who discovered Okazaki fragments?

Reiji Okazaki and Tsuneko Okazaki at Nagoya University, in a series of papers culminating in 1968. They pulse-labeled E. coli with tritiated thymidine for very short intervals (5-30 seconds), then sedimented the DNA on alkaline sucrose gradients and saw two populations: a heavy peak corresponding to long parental-length DNA, and a light peak around 1000-2000 nt that represented the freshly made short pieces. With longer chase times the short peak disappeared into the heavy peak as ligase joined the pieces. Reiji Okazaki died of leukemia in 1975 at age 44 — likely a delayed consequence of radiation exposure from the 1945 Hiroshima bombing — and Tsuneko continued the work, identifying primase function in 1980.

What removes the RNA primer from an Okazaki fragment?

Two pathways exist. In E. coli, DNA polymerase I uses its 5' to 3' exonuclease (the Klenow fragment is what's left when this domain is cleaved off) to chew through the RNA primer of the downstream fragment as it extends forward — nick translation. Then DNA ligase seals the final phosphodiester bond. In eukaryotes the polymerase delta cannot degrade RNA but it can displace the primer into a single-stranded flap, which the structure-specific endonuclease FEN1 cleaves off. The cleaved nick is then sealed by DNA ligase I. A backup pathway uses the 5' to 3' exonuclease activity of Dna2 helicase-nuclease for unusually long flaps.

Why do eukaryotic fragments match nucleosome spacing?

Behind the eukaryotic fork, the leading-strand and lagging-strand polymerases collaborate with chromatin assembly factor CAF-1 to deposit fresh histone octamers onto newly replicated DNA. Each octamer wraps ~147 bp of DNA, with linker DNA bringing the repeat to ~165 bp. The lagging-strand machinery appears to use this nucleosomal periodicity as a positional cue: primase fires once per nucleosome, producing ~165 nt Okazaki fragments that each carry exactly one new nucleosome. Smith and Whitehouse (2012) showed that disrupting CAF-1 lengthens fragments while overexpressing chromatin assembly factors shortens them — direct evidence that chromatin spacing controls primer-firing frequency.

What happens if Okazaki fragments are not joined?

Unligated fragments leave single-strand nicks every 100-2000 nt across the lagging strand. Most are sealed within seconds by ligase I, but failure causes persistent nicks that collapse into double-strand breaks when the next fork arrives, triggering ATR-Chk1 checkpoint activation. Mice null for FEN1 or ligase I die as embryos. Hypomorphic FEN1 mutants accumulate single-strand breaks and develop lymphomas at 10-fold higher rates. Bloom syndrome and Werner syndrome both involve accessory helicases that help resolve stalled lagging-strand intermediates; loss of either causes hyper-recombination, premature aging, and cancer predisposition. The lagging strand is, in this sense, the most fragile and most checkpoint-monitored part of every replication fork.