What is the difference between DNA polymerase I and DNA polymerase III?

Both are bacterial enzymes that synthesize DNA 5′ to 3′ and proofread with a 3′ to 5′ exonuclease. DNA polymerase III is the main replicative polymerase and copies both the leading and lagging strands. DNA polymerase I has an extra 5′ to 3′ exonuclease that removes the RNA primers from Okazaki fragments, and it then fills each gap with DNA. Pol I exists only in bacteria; in human cells, RNase H and the flap endonuclease FEN1 remove primers and Pol δ fills the gaps.

What enzyme joins Okazaki fragments together?

DNA ligase joins Okazaki fragments. After the RNA primer is removed and the gap is filled with DNA, a nick remains in the sugar-phosphate backbone. Ligase forms the final phosphodiester bond between the 3′-OH of one fragment and the 5′ end of the next. In human cells, DNA ligase I, an ATP-dependent enzyme, does this job, and inherited LIG1 deficiency causes a rare immunodeficiency with increased sensitivity to DNA-damaging agents.

Why are Okazaki fragments made only on the lagging strand?

DNA polymerases can only add nucleotides to a 3′-OH, so all synthesis runs 5′ to 3′. On the leading strand, the polymerase can follow the fork continuously from a single primer. On the other template, synthesis must run away from the fork, so the polymerase backstitches in short pieces, each started by its own RNA primer. These Okazaki fragments are 1,000-2,000 nucleotides long in bacteria and much shorter, about 100-200, in eukaryotes. Both strands are still made 5′ to 3′.

What do DNA polymerases alpha, delta, epsilon and gamma do?

In eukaryotes, polymerase alpha works in a complex with primase to make the short RNA-DNA primers that start each stretch of DNA. Polymerase epsilon extends and proofreads the leading strand, and polymerase delta extends and proofreads the lagging strand and fills the gaps left after primer removal. Polymerase gamma is the only polymerase that replicates mitochondrial DNA; variants in its gene, POLG, cause mitochondrial disorders such as Alpers-Huttenlocher syndrome and progressive external ophthalmoplegia.

Which drugs target DNA replication enzymes?

Fluoroquinolones such as ciprofloxacin inhibit bacterial DNA gyrase and topoisomerase IV, both type II topoisomerases, and the newer gepotidacin and zoliflodacin, both first approved by the FDA in 2025, inhibit the same two enzymes. Etoposide inhibits topoisomerase II and irinotecan inhibits topoisomerase I. Cytarabine, a pyrimidine analog, inhibits DNA polymerase, and acyclovir, once phosphorylated by viral thymidine kinase, terminates viral DNA chains. Rifampin is a trap: it blocks bacterial RNA polymerase, so it inhibits transcription, not replication.

What does telomerase do, and why don't bacteria need it?

Linear chromosomes cannot be fully copied at their very ends, because the lagging strand cannot be finished at the tip, so they would shorten with every division. Telomerase is a protein-RNA reverse transcriptase that carries its own RNA template and adds telomere repeats (TTAGGG in humans) to chromosome ends. It is switched off in many somatic tissues, such as skin. Bacteria have circular chromosomes with no ends, so they need no telomeres or telomerase.

DNA Replication Machinery: Enzymes, Pol I vs III & Drugs

DNA replication is semiconservative and bidirectional, and every DNA polymerase builds a new strand 5′→3′ by adding nucleotides to an existing 3′-OH, so each new stretch of DNA must start from a short RNA primer made by primase. In bacteria, DNA polymerase III does the bulk copying, and DNA polymerase I removes the primers with its 5′→3′ exonuclease and fills the gaps with DNA; in eukaryotes, the Pol α-primase complex starts each stretch, Pol ε copies the leading strand and Pol δ the lagging strand. Helicase unwinds, single-strand binding proteins hold the strands apart, topoisomerases relieve supercoiling, and DNA ligase seals the last nick between Okazaki fragments. Drugs exploit the same machinery: fluoroquinolones block bacterial DNA gyrase and topoisomerase IV, etoposide inhibits topoisomerase II, irinotecan inhibits topoisomerase I, and cytarabine and acyclovir jam DNA polymerases.

The fork in six verbs: unwind, hold, prime, extend, erase, seal

Unwind: helicase. It burns ATP to run along one strand and pry the helix apart ahead of the fork. Topoisomerases work ahead of the fork as swivels, so the twisted parental DNA can rotate instead of jamming.
Hold: single-strand DNA-binding proteins (SSB in bacteria; replication factor A, now usually called RPA, in eukaryotes) coat the exposed template so it cannot re-anneal or fold into hairpins. They hold strands apart; they do not unwind them.
Prime: primase is an RNA polymerase, so it can start a chain from nothing. It lays down a short RNA primer (4-15 nucleotides in bacteria, 8-12 in eukaryotes) that supplies the 3′-OH no DNA polymerase can make for itself.
Extend: DNA polymerase, held on the template by a ring-shaped sliding clamp (β clamp in bacteria, PCNA in eukaryotes) that a clamp loader snaps on using ATP. The clamp keeps the polymerase from falling off.
Erase: primer removal. In bacteria, DNA polymerase I's 5′→3′ exonuclease, helped by RNase H, strips out the RNA and refills the gap with DNA. Eukaryotes have no Pol I: RNase H and the flap endonuclease FEN1 remove the primer, and Pol δ fills in.
Seal: DNA ligase forms the missing phosphodiester bond between the 3′-OH of one fragment and the 5′ end of the next. In human cells, DNA ligase I, an ATP-dependent enzyme, joins Okazaki fragments.

Why the primer is RNA: the classic explanation is that an enzyme that starts chains from scratch cannot proofread its own start, so building primers from ribonucleotides flags them as temporary; they are always removed and replaced with DNA by polymerases that do proofread.

Bacteria vs eukaryotes: who does each job

Job	Bacteria (E. coli)	Eukaryotes (human cells)
Start sites	A single origin	Many origins, 30,000-300,000 bp apart, each bound by the origin recognition complex (ORC)
Hold strands apart	SSB	Replication factor A (RPA)
Relieve supercoils	DNA gyrase, a type II topoisomerase that introduces negative supercoils	Topoisomerases I and II
Prime	Primase	Pol α-primase complex makes short RNA-DNA primers
Clamp and clamp loader	β clamp; γ complex	PCNA; replication factor C (RFC)
Main copying	Pol III on both strands: 5′→3′ synthesis, 3′→5′ proofreading	Pol ε (leading) and Pol δ (lagging); both proofread
Remove primers	Pol I 5′→3′ exonuclease plus RNase H	RNase H and FEN1; Pol δ fills the gap
Seal	DNA ligase	DNA ligase I (ATP-dependent)
Okazaki fragments	1,000-2,000 nucleotides	Much shorter: about 100-200 nucleotides
Fork speed	Roughly 10 times faster than in eukaryotes	About 50 nucleotides per second
Chromosome ends	Circular chromosome, no telomeres	Telomerase extends telomeres
Mitochondrial DNA	Not applicable	Pol γ, the only polymerase that copies mtDNA

Leading vs lagging strand

	Leading strand	Lagging strand
Synthesis	Continuous, in the direction the fork moves	Discontinuous, "backstitching" away from the fork
Primers	One, at the start	One for every Okazaki fragment
Eukaryotic polymerase	Pol ε	Pol δ
Direction of synthesis	5′→3′	Also 5′→3′

Trap: the lagging strand is not made 3′→5′; no polymerase can do that. Building 5′→3′ is what makes proofreading work: a mismatched base at the growing 3′ end can be clipped off, and synthesis simply resumes.

Pol I vs Pol III: the classic trap

Activity	Pol III	Pol I
5′→3′ polymerase	Yes: the main replicative enzyme for both strands	Yes: refills the gap left by the primer
3′→5′ exonuclease (proofreading)	Yes	Yes
5′→3′ exonuclease	No	Yes: removes RNA primers
Stem clue	"Main bacterial replicase," "elongation stops"	"RNA primers persist," "primer removal"

Pol I exists only in bacteria. If the stem says human or eukaryotic cells, primer removal belongs to RNase H and FEN1, with Pol δ filling the gap.

Fidelity comes in layers. Proofreading improves accuracy 100- to 1,000-fold, and post-replication mismatch repair removes most of the errors that slip through, for an overall rate of about 1 error per 10⁹-10¹⁰ nucleotides copied. Mismatch repair is the system that fails in Lynch syndrome; the other pathways are in the DNA repair cheat sheet.

Telomeres and telomerase

End-replication problem: on a linear chromosome, the lagging strand cannot be finished at the very tip, so the ends would shorten with every division. Circular bacterial chromosomes have no ends to lose.
Telomerase is a protein-RNA complex that works as a reverse transcriptase: it carries its own RNA template and adds copies of the telomere repeat (TTAGGG in humans) to chromosome ends.
Telomerase is switched off in tissues such as skin, which caps how many times those cells can divide and may guard against uncontrolled proliferation.
Reverse transcriptase is also the enzyme NRTIs block in HIV; see HIV antiretroviral drugs by life cycle.

Drugs that target replication

Drug	Target	What to remember
Fluoroquinolones (ciprofloxacin, levofloxacin, moxifloxacin)	Bacterial DNA gyrase and topoisomerase IV (both type II topoisomerases)	Bactericidal. Boxed warning: tendinitis and tendon rupture, peripheral neuropathy, CNS effects, worsening myasthenia gravis. Also QT prolongation, aortic aneurysm or dissection, dysglycemia
Gepotidacin (Blujepa) and zoliflodacin (Nuzolvence), both first FDA-approved in 2025	Bacterial type II topoisomerases: DNA gyrase and topoisomerase IV	New non-fluoroquinolone classes. Gepotidacin: uncomplicated UTI in females 12 and older (at least 40 kg), and uncomplicated urogenital gonorrhea (12 and older, at least 45 kg) when there are limited or no alternative options. Zoliflodacin: a single 3 g oral dose for uncomplicated urogenital gonorrhea (12 and older, at least 35 kg)
Etoposide	Topoisomerase II (cuts both strands)	Solid tumors, leukemias and lymphomas
Irinotecan	Topoisomerase I (cuts one strand)	The topo I half of the classic "topo I vs topo II" drug pair
Cytarabine	DNA polymerase	Pyrimidine analog used for leukemia
Acyclovir	Herpes virus DNA polymerase	Guanosine analog first phosphorylated by viral thymidine kinase; the triphosphate terminates the viral DNA chain
5-FU, trimethoprim, sulfonamides	Nucleotide supply	Starve the fork of building blocks; see de novo vs salvage synthesis
Rifampin (trap)	Bacterial DNA-dependent RNA polymerase	Blocks transcription, not replication; orange body fluids, CYP450 inducer

Per its label, zoliflodacin binds inside the cleaved DNA-gyrase complex and blocks re-ligation, so the cut DNA cannot be resealed. For coverage of each antibiotic class, see antibiotic mechanisms and coverage.

The Q-bank case

A researcher studies a temperature-sensitive mutant strain of Escherichia coli. At the restrictive temperature, the cells keep synthesizing DNA on both template strands, but new lagging-strand DNA accumulates as pieces of about 1,000-2,000 nucleotides. The pieces contain no ribonucleotides, and the 3′-OH end of each piece sits directly against the 5′ end of the next, with no missing nucleotides between them. Which enzyme is most likely defective?

A. DNA gyrase B. DNA ligase C. DNA polymerase I D. DNA polymerase III E. Helicase F. Primase

Answer: B. DNA ligase. Every step before sealing worked: the fragments were primed, extended, stripped of RNA and gap-filled. All that is missing is the final phosphodiester bond across each nick, and only ligase makes it.

A. DNA gyrase: relieves the positive supercoils that build up ahead of the fork. Without it, the fork stalls and synthesis slows on both strands; gyrase does not join fragments. It is a fluoroquinolone target, along with topoisomerase IV.
C. DNA polymerase I: a Pol I defect leaves RNA primers on the 5′ ends of Okazaki fragments (no 5′→3′ exonuclease) or leaves gaps unfilled. Here there is no RNA and no gap.
D. DNA polymerase III: the main replicative polymerase. Losing it stops elongation on both strands instead of producing many full-length, unjoined fragments.
E. Helicase: without unwinding, the fork never opens and neither strand is copied.
F. Primase: without primers, new Okazaki fragments cannot start, so lagging-strand synthesis stalls instead of producing a pile of finished fragments.

Human version: the same job belongs to DNA ligase I. Inherited LIG1 deficiency is a rare immunodeficiency with increased sensitivity to DNA-damaging agents.

How USMLE tests it: match the clue to the enzyme

Stem clue	Answer
Okazaki fragments pile up with no RNA and no gaps	DNA ligase
RNA stays attached to Okazaki fragments (bacteria)	DNA polymerase I (5′→3′ exonuclease)
More point mutations; mismatched bases not removed during synthesis	3′→5′ exonuclease (proofreading)
No new fragments can start; no primers	Primase
Strands never separate	Helicase
Exposed single strands re-anneal or form hairpins	SSB or RPA
Polymerase falls off the template after a few nucleotides	Sliding clamp (β clamp, PCNA)
Overwinding ahead of the fork; fluoroquinolone target	DNA gyrase
Chromosome ends shorten with each division	Telomerase
Copies mitochondrial DNA	Pol γ

Topoisomerase vs ligase: topoisomerases reseal their own breaks. Topo I cuts one strand and reseals it without extra energy, because the covalent enzyme-DNA link stores the bond energy; topo II cuts both strands and uses ATP. Ligase seals the nicks left by replication and repair.
Replication vs transcription: rifampin blocks bacterial RNA polymerase, the transcription enzyme, so it belongs in transcription questions. For the next step after transcription, see the translation acronym trick.
Pol γ in the clinic: POLG variants cause mitochondrial disorders such as Alpers-Huttenlocher syndrome and progressive external ophthalmoplegia.

Key takeaways

All DNA synthesis runs 5′→3′ from a 3′-OH, so primase must supply an RNA primer first.
Pol III copies bacterial DNA; Pol I removes primers with its 5′→3′ exonuclease, a bacterial-only job.
In eukaryotes, Pol α-primase starts, Pol ε leads, Pol δ lags, PCNA clamps, DNA ligase I seals and Pol γ copies mtDNA.
Leading = continuous; lagging = Okazaki fragments (1,000-2,000 nucleotides in bacteria, 100-200 in eukaryotes), and both are made 5′→3′.
Proofreading = 3′→5′ exonuclease; mismatch repair is the post-replication backup.
Telomerase is a reverse transcriptase with its own RNA template that adds TTAGGG repeats.
Fluoroquinolones → DNA gyrase and topo IV; etoposide → topo II; irinotecan → topo I; cytarabine and acyclovir → DNA polymerase; rifampin → RNA polymerase.