Science Wiki

DNA polymerase I (frequently abbreviated as Pol I, EC 2.7.7.7) is a monomeric, three-domain enzyme encoded by the *polA* gene in prokaryotes and initially isolated from the Gram-negative bacterium *Escherichia coli* by Arthur Kornberg in 1956. As the first discovered member of the DNA polymerase superfamily and the structural prototype of Family A DNA polymerases, Pol I established the biochemical and biophysical principles of template-directed nucleic acid synthesis. While originally hypothesized to be the primary chromosomal replicase, the isolation of viable *polA1* nonsense mutants by Paula DeLucia and John Cairns in 1969 demonstrated that DNA polymerase III holoenzyme executes high-speed chromosomal DNA replication, whereas Pol I (~400 molecules per *E. coli* cell; polymerization rate ~10–20 nucleotides/s at 37 °C; processivity ~15–50 nt) is dedicated to Okazaki fragment maturation via nick translation and gap-filling in DNA repair (base excision repair and nucleotide excision repair).

The Three Globular Domains of the 103 kDa Holoenzyme

In *E. coli*, the intact Pol I holoenzyme is a single polypeptide chain of 928 amino acids with a molecular mass of approximately 103 kilodaltons (kDa). Rather than operating as a single catalytic unit, the polypeptide folds into three globular, functionally coupled structural domains connected by flexible peptide linkers, enabling the enzyme to coordinate three distinct catalytic activities on the double helix:

• N-Terminal 5′ to 3′ Nuclease (Flap Endonuclease) Domain (Residues 1 to 323, ~35 kDa): Located at the extreme N-terminus, this domain is structurally and evolutionarily homologous to bacteriophage T5 5′-exonuclease and eukaryotic Flap Endonuclease 1 (FEN1). Although historically termed a "5′ to 3′ exonuclease," biochemical characterization demonstrates that it operates primarily as a structure-specific 5′ flap endonuclease. When the advancing polymerase domain displaces a downstream RNA primer or damaged DNA strand into a single-stranded 5′ flap, the nascent flap threads through a helical arch in the N-terminal domain and is cleaved endonucleolytically at the junction between single-stranded and double-stranded DNA (typically one nucleotide into the duplex), releasing mono- or short oligonucleotides bearing 5′-monophosphate termini. This 5′ nuclease activity is unique to Pol I among the five known *E. coli* DNA polymerases (Pol I through Pol V).
• Central 3′ to 5′ Exonuclease (Proofreading) Domain (Residues 324 to 517, ~20 kDa): Positioned between the N-terminal and C-terminal domains and separated from the polymerase catalytic site by approximately 30 to 35 angstroms (Å), this domain executes exonucleolytic proofreading. The active site contains three conserved sequence motifs (ExoI, ExoII, and ExoIII) whose acidic carboxylate side chains (Asp355, Glu357, Asp424, and Asp501 in *E. coli*) coordinate two divalent metal ions that hydrolytically excise mispaired 3′-terminal nucleotides in the 3′ to 5′ direction.
• C-Terminal 5′ to 3′ Polymerase Domain (Residues 518 to 928, ~48 kDa): Catalyzes the template-directed nucleophilic addition of deoxynucleoside triphosphates (dNTPs) onto the 3′-hydroxyl (3′-OH) terminus of a base-paired primer strand, extending the nascent polynucleotide chain exclusively in the 5′ to 3′ direction.

The Right-Hand Polymerase Fold (Palm, Fingers, and Thumb Subdomains)

High-resolution X-ray crystallography of the C-terminal polymerase domain (pioneered by Thomas A. Steitz in 1985) revealed that the catalytic core folds into a structure resembling a half-opened human right hand, an architectural fold conserved across nearly all replicative DNA and RNA polymerases in the biosphere. The polymerase domain is partitioned into three functional subdomains that enclose a deep DNA-binding cleft:

1. The Palm Subdomain: Forms the structural floor of the catalytic cleft, constructed from a conserved antiparallel beta sheet. The Palm houses the invariant catalytic carboxylate residues (Asp705, Asp882, and Glu883 in *E. coli* Pol I) that coordinate the catalytic divalent magnesium ions (Mg²⁺) and position the primer 3′-OH adjacent to the α-phosphate of the incoming dNTP.
2. The Fingers Subdomain: A predominantly α-helical domain (centered around the mobile O-helix, containing conserved residues Arg754, Lys758, Phe762, and Tyr766) that binds the incoming dNTP substrate. Upon binding a complementary dNTP that forms proper Watson-Crick base pairing with the template base, the Fingers subdomain undergoes a ~40° inward rotational conformational transition from an open binary state (E•DNA) to a closed, catalytically competent ternary state (E•DNA•dNTP). In the closed conformation, Tyr766 and Phe762 stack directly against the nascent base pair to enforce strict planar Watson-Crick geometry, while conserved arginine and glutamine side chains probe the minor groove for the universal hydrogen-bond acceptors (N3 of purines and O2 of pyrimidines).
3. The Thumb Subdomain: Composed of two antiparallel α-helices connected by a flexible tip that curls over the minor groove of the newly synthesized upstream duplex DNA. The Thumb stabilizes the primer-template complex against premature dissociation and guides the mechanical translocation of the duplex by one base pair (~3.4 Å) following each catalytic cycle.

Two-Metal-Ion Phosphoryl Transfer and Thermodynamics

The chemical reaction catalyzed by the 5′ to 3′ polymerase active site is the nucleophilic substitution of the primer 3′-OH group onto the innermost (α) phosphorus atom of an incoming deoxynucleoside 5′-triphosphate (dNTP), liberating inorganic pyrophosphate (PP_i, or P₂O₇⁴⁻):

DNA_*n* + dNTP ⇌ DNA_*n*+1 + PP_i

Because a single phosphodiester bond is formed in the DNA backbone while a single phosphoanhydride bond (between the α- and β-phosphates of the dNTP) is cleaved, the intrinsic biochemical standard Gibbs free energy change of the isolated polymerization step is only modestly exergonic (ΔG°′ ≈ −15 to −20 kJ/mol, with an equilibrium constant *K*_eq ≈ 500 to 1,000, permitting the reverse reaction known as pyrophosphorolysis at high PP_i concentrations). In vivo, polymerization is driven irreversibly forward by the subsequent ubiquitous action of cytosolic inorganic pyrophosphatase, which hydrolyzes the released pyrophosphate into two molecules of orthophosphate (PP_i + H₂O → 2 P_i, ΔG°′ ≈ −19 to −33 kJ/mol), yielding a net coupled free energy change of ΔG°′_net ≈ −40 to −50 kJ/mol.

At the atomic scale, Pol I executes phosphoryl transfer via the universal two-metal-ion catalytic mechanism situated within the Palm subdomain:

• Metal Ion A (Catalytic Cation, Mg²⁺_A): Octahedrally coordinated by the carboxylate oxygens of Asp705 and Asp882, the pro-*R*_p oxygen of the dNTP α-phosphate, and the 3′-OH of the primer terminus. Acting as a Lewis acid, Mg²⁺_A polarizes the terminal 3′-O–H bond and lowers the p*K*_a of the 3′-hydroxyl group from ~13 to ~7, facilitating proton abstraction (assisted by a general base or water network) to generate a potent 3′-alkoxide nucleophile (3′-O⁻) positioned at an inline attack angle (~165° to 180°) relative to the scissile P_α–O_αβ bond.
• Metal Ion B (Nucleotide-Binding Cation, Mg²⁺_B): Coordinated by Asp705, Asp882, the backbone carbonyl oxygen of residue 883, and the non-bridging oxygens of the α-, β-, and γ-phosphates of the incoming dNTP (alongside positively charged Arg754 and Lys758 on the O-helix). Mg²⁺_B neutralizes the dense negative electrostatic charge of the triphosphate tail, chelates the β- and γ-phosphates into a reactive conformation, stabilizes the buildup of negative charge on the pentacoordinate trigonal-bipyramidal phosphorane transition state, and facilitates the departure of the MgPP_i²⁻ leaving group. Time-resolved X-ray crystallographic studies have additionally resolved a transient third divalent metal ion (Metal C) that binds between the α- and β-phosphates during bond formation to neutralize product charge and suppress reverse pyrophosphorolysis.

Stereochemical analysis utilizing chiral phosphorothioate analogs (where a non-bridging oxygen on the α-phosphate is replaced by sulfur, dNTPαS) demonstrates complete inversion of stereochemical configuration at the α-phosphorus atom, proving that phosphoryl transfer proceeds via a single, concerted inline S_N2 nucleophilic displacement without a covalent enzyme-DNA intermediate.

The Glu710 Steric Gate and Three-Stage Fidelity Partitioning

Inside a proliferating bacterial cell, the intracellular concentrations of ribonucleoside triphosphates (rNTPs, such as ATP at ~3 to 5 mM) exceed those of deoxynucleoside triphosphates (dNTPs, at ~50 to 300 μM) by 10- to 100-fold. To prevent the erroneous incorporation of rNTPs into genomic DNA (which destabilizes the backbone via 2′-OH intramolecular transesterification), Pol I employs a conserved steric gate residue within the Palm/Fingers binding pocket: Glu710. The bulky aliphatic and carboxylate side chain of Glu710 (backed by Tyr762) packs directly beneath the C2′ position of the incoming nucleotide's deoxyribose ring, creating a severe van der Waals steric clash with the 2′-hydroxyl group of any incoming ribonucleotide and reducing rNTP incorporation efficiency by 10³- to 10⁴-fold.

Pre-steady-state transient kinetics reveal that Pol I achieves an overall replication error frequency of only ~10⁻⁶ to 10⁻⁷ errors per base pair polymerized through a three-stage kinetic partitioning mechanism:

Because the cost of proofreading depends on the kinetic competition between extending a primer (*k*_ext, which is proportional to [dNTP]) and excising the 3′ terminus (*k*_exo), artificially elevating dNTP concentrations in vitro forces the polymerase to extend mismatches before they can partition into the exonuclease site, known as the next-nucleotide effect.

Coupled Strand Displacement and 5′ Flap Cleavage (Nick Translation)

During chromosomal DNA replication, the antiparallel architecture of the DNA duplex (coupled with the universal requirement that all DNA polymerases synthesize exclusively in the 5′ to 3′ direction) forces the lagging strand to be synthesized discontinuously as a series of 1,000- to 2,000-nucleotide Okazaki fragments. Because no replicative DNA polymerase can initiate synthesis de novo without a pre-existing 3′-OH group, *E. coli* primase (DnaG) synthesizes a 10- to 12-nucleotide RNA primer to initiate each Okazaki fragment, which is then rapidly elongated by the DNA polymerase III holoenzyme.

When the lagging-strand core of Pol III finishes synthesizing an Okazaki fragment and collides with the 5′-triphosphate end of the RNA primer belonging to the preceding Okazaki fragment, Pol III disengages from the DNA, leaving a nicked RNA-DNA junction tethered by a ring-shaped β-sliding clamp (DnaN). To convert these discontinuous RNA-primed fragments into a continuous, stable genomic DNA strand, the cell executes Okazaki fragment maturation:

1. RNase HI and Pol I Recruitment: While RNase HI endonucleolytically degrades the bulk of the RNA moiety within the RNA:DNA hybrid (though it cannot cleave the final ribonucleotide covalently linked via a 3′–5′ phosphodiester bond to the 5′ deoxyribonucleotide junction), Pol I is recruited directly to the nick through a conserved peptide motif that binds the hydrophobic cleft of the resident β-sliding clamp.
2. Coupled Strand Displacement and 5′ Flap Cleavage (Nick translation): Binding to the 3′-OH terminus of the upstream Okazaki fragment, the C-terminal polymerase domain of Pol I initiates 5′ to 3′ DNA synthesis. As Pol I advances into the downstream fragment, it unwinds and displaces the 5′-terminal RNA primer (or damaged DNA) from the template strand, extruding a short 5′ single-stranded flap. This displaced 5′ flap is guided directly into the N-terminal 5′ nuclease domain, which cleaves the phosphodiester backbone at the base of the flap to leave a base-paired 5′-monophosphate (5′-P) deoxyribonucleotide terminus.
3. Pol I Dissociation and Ligation: Because Pol I couples every nucleotide addition to the downstream movement of the single-strand nick along the duplex without creating a widened gap, this coordinated reaction is termed nick translation. Once the RNA primer is completely excised and Pol I has polymerized ~10 to 50 deoxyribonucleotides into the downstream DNA, the enzyme dissociates from the duplex. This exposes a clean, ligatable nick (a 3′-OH immediately adjacent to a 5′-P) that is covalently sealed by *E. coli* DNA ligase (LigA, which uses NAD⁺ as an adenylate cofactor to form an enzyme-AMP intermediate, transfers AMP to the 5′-P to generate a pyrophosphate linkage, and catalyzes attack by the 3′-OH to seal the backbone and release AMP and nicotinamide mononucleotide).

Gap-Filling Synthesis in Base and Nucleotide Excision Repair

Beyond lagging-strand processing, Pol I serves as the primary gap-filling polymerase across bacterial excision repair pathways:

• Base excision repair (BER): Following the removal of oxidized, alkylated, or deaminated bases (such as uracil or 8-oxoguanine) by specific DNA glycosylases and incision of the resulting abasic (AP) site by AP endonucleases (Exonuclease III / XthA or Endonuclease IV / Nfo), Pol I binds the nick, utilizes its 5′ nuclease / 5′-deoxyribophosphodiesterase (dRPase) activity to excise the residual 5′-deoxyribose-5-phosphate moiety, and fills the single-nucleotide (short-patch) or 2- to 10-nucleotide (long-patch) gap prior to LigA ligation.
• Nucleotide excision repair (NER): When bulky helix-distorting lesions (such as ultraviolet-induced cyclobutane pyrimidine dimers or 6-4 photoproducts) are detected by the UvrAB complex, the UvrABC excinuclease makes dual incisions 8 phosphodiester bonds 5′ and 4 to 5 phosphodiester bonds 3′ to the lesion. After DNA helicase II (UvrD) unwinds and displaces the damaged 12- to 13-nucleotide oligomer, Pol I synthesizes across the resulting single-stranded gap using the undamaged complementary strand as a template.

Subtilisin Proteolysis and the Klenow / Exo⁻ Klenow Fragments

In 1970, Danish biochemist Hans Klenow and colleagues (alongside Peter Setlow, Douglas Brutlag, and Arthur Kornberg) demonstrated that mild, limited proteolysis of native *E. coli* DNA polymerase I with the bacterial serine protease subtilisin selectively cleaves the exposed interdomain peptide linker between residues Thr323 and Val324, partitioning the 103 kDa holoenzyme into two discrete, enzymatically active fragments:

• The Small Fragment (~35 kDa, Residues 1 to 323): Contains the isolated N-terminal 5′ to 3′ flap nuclease domain.
• The Large Fragment or Klenow fragment (~68 kDa, Residues 324 to 928): Retains both the C-terminal 5′ to 3′ polymerase domain and the central 3′ to 5′ proofreading exonuclease domain, while completely lacking 5′ to 3′ nuclease activity. In modern recombinant DNA production, the Klenow fragment is expressed directly from a genetically truncated *polA* reading frame lacking the first 323 codons to eliminate trace contamination by intact holoenzyme.

Because the Klenow fragment cannot degrade downstream 5′ strands ahead of the advancing replication fork, it executes pure template-directed polymerization and strand displacement without destroying existing 5′-labeled probes or primers. Furthermore, site-directed mutagenesis of two conserved catalytic carboxylate residues within the ExoI proofreading motif (specifically the D355A / E357A double mutation) abolishes the 3′ to 5′ exonuclease activity while preserving 5′ to 3′ polymerization. This engineered derivative, termed the exo⁻ Klenow fragment, exhibits enhanced strand-displacement synthesis and catalyzes the template-independent addition of a single 3′-deoxyadenosine monophosphate (dAMP) overhang to the 3′ ends of blunt double-stranded DNA fragments.

Cloning, Labeling, and TaqMan Hydrolysis Probe Applications

The distinct catalytic profiles of intact Pol I, the standard Klenow fragment, the exo⁻ Klenow fragment, and thermostable Family A homologs (most notably Taq polymerase from the thermophile *Thermus aquaticus*, which retains an active N-terminal 5′ flap nuclease domain and C-terminal polymerase domain but possesses a naturally vestigial, inactive central 3′ to 5′ exonuclease domain) underpin several core methodologies in molecular biology and biotechnology:

• Second-Strand cDNA Synthesis and Sanger Sequencing: Prior to the widespread adoption of engineered bacteriophage T7 DNA polymerase (Sequenase) and thermostable *Taq* variants, the Klenow fragment served as the primary enzyme for synthesizing double-stranded complementary DNA (cDNA) from single-stranded reverse-transcribed templates and for executing dideoxy chain-termination DNA sequencing developed by Frederick Sanger.
• End-Modification (Blunting) of Restriction Fragments: In molecular cloning, incompatible ends generated by restriction endonucleases are converted into blunt-ended duplexes using the Klenow fragment:

1. Filling in 5′ Overhangs: In the presence of all four dNTPs, the 5′ to 3′ polymerase activity extends recessed 3′-OH termini using the protruding 5′ single-stranded overhang as a template until a flush blunt end is formed.
2. Polishing 3′ Overhangs: Because single-stranded 3′ overhangs cannot serve as primers for 5′ to 3′ synthesis, the 3′ to 5′ proofreading exonuclease activity of the Klenow fragment digests the protruding single-stranded 3′ tail backward until it reaches the base-paired duplex, where the presence of dNTPs and rapid polymerase re-synthesis prevent further inward degradation.

• Random-Primed DNA Labeling and 3′ dA-Tailing: Using the Feinberg-Vogelstein method, denatured template DNA is hybridized with a mixture of random hexamer oligonucleotides; the Klenow fragment extends these primers using radiolabeled (such as [α-³²P]dCTP) or fluorophore/biotin-conjugated dNTPs without degrading downstream newly synthesized probes. Conversely, the exo⁻ Klenow fragment is utilized in next-generation sequencing library preparation and TA cloning to append single 3′-dA overhangs onto sheared, end-repaired genomic DNA fragments, enabling efficient ligation to 3′-dT-tailed sequencing adapters.
• Nick Translation Labeling and Quantitative PCR Hydrolysis Probes: Intact Pol I holoenzyme paired with trace quantities of pancreatic deoxyribonuclease I (DNase I, which introduces random single-strand nicks across duplex DNA) is utilized to incorporate labeled nucleotides uniformly across double-stranded DNA probes for fluorescence in situ hybridization (FISH). Analogously, in real-time quantitative polymerase chain reaction (RT-qPCR), the homologous N-terminal 5′ to 3′ flap nuclease domain of *Taq* polymerase cleaves hybridized dual-labeled fluorogenic TaqMan oligonucleotides during primer extension, separating the 5′ reporter fluorophore from the 3′ quencher to generate a quantitative fluorescence signal proportional to target amplicon accumulation.

Why did the 1969 *polA1* mutant of *Escherichia coli* prove that DNA polymerase I is not the primary replicative enzyme?

When Paula DeLucia and John Cairns isolated the *E. coli* *polA1* amber nonsense mutant in 1969, extracts from the mutant exhibited less than 1% of normal DNA polymerase I activity, yet the bacteria still replicated their chromosomes and divided at near-normal rates (while showing extreme sensitivity to ultraviolet and alkylating DNA damage). This unexpected viability proved that another enzyme—subsequently purified by Thomas Kornberg and Malcolm Gefter in 1970 as DNA polymerase III—catalyzes bulk chromosomal replication, while Pol I specializes in primer removal and repair. However, complete deletion of the N-terminal 5′→3′ flap nuclease domain of *polA* is lethal in rich media because the cell cannot excise RNA primers from Okazaki fragments.

What is the functional difference between the 3′→5′ exonuclease and 5′→3′ nuclease domains of DNA polymerase I?

The central 3′→5′ exonuclease domain moves backward relative to DNA synthesis, hydrolyzing a newly added, mispaired single-stranded nucleotide from the 3′-OH primer terminus to perform real-time proofreading and boost replication fidelity by ~100-fold. In contrast, the N-terminal 5′→3′ nuclease (flap endonuclease) domain acts ahead of the advancing polymerase in the direction of synthesis, cleaving displaced 5′ RNA primers or damaged DNA flaps at the duplex junction during nick translation and Okazaki fragment maturation.

Why is the Klenow fragment used instead of intact DNA polymerase I for filling in 5′ restriction overhangs and Sanger sequencing?

Intact DNA polymerase I possesses N-terminal 5′→3′ nuclease activity that degrades exposed 5′ single-stranded overhangs and digests downstream 5′-labeled sequencing primers or newly synthesized DNA strands ahead of the polymerase. By removing the N-terminal 323 amino acids to create the Klenow fragment, the enzyme retains full 5′→3′ polymerase and 3′→5′ proofreading activities without any 5′→3′ degradation, allowing clean fill-in of 5′ overhangs and uniform extension of sequencing primers.