Base Excision Repair

Base excision repair (BER) is an evolutionarily conserved cellular DNA repair pathway responsible for detecting, excising, and replacing single chemically modified or inappropriate nucleobases that cause minimal structural distortion to the B-form DNA double helix. Discovered in 1974 by Swedish biochemist Tomas Lindahl—who isolated uracil-DNA glycosylase in Escherichia coli after calculating that spontaneous hydrolytic decay of DNA in aqueous physiological solution would degrade the genome within days without dedicated enzymatic surveillance, an achievement for which he shared the 2015 Nobel Prize in Chemistry—BER serves as the primary frontline defense against endogenous metabolic and hydrolytic DNA damage. Every human diploid cell experiences approximately 10,000 to 30,000 endogenous single-base lesions per day generated by spontaneous hydrolytic deamination (such as cytosine converting to uracil), spontaneous N-glycosidic bond hydrolysis (depurination and depyrimidination), reactive oxygen species (ROS), and endogenous methylating metabolites (such as S-adenosylmethionine). Through the sequential action of lesion-specific DNA glycosylases, AP endonuclease 1 (APE1), DNA polymerase β (Pol β), and DNA ligase IIIα–XRCC1 (or DNA Ligase I), BER restores native sequence integrity and prevents mutagenic base mispairing that would otherwise fix permanent genetic mutations during replication.

Spontaneous Hydrolytic Deamination and Depurination

Unlike bulky, helix-distorting lesions such as cyclobutane pyrimidine dimers induced by exogenous ultraviolet radiation (which are removed by Nucleotide Excision Repair, NER), the substrates of BER arise continually from the chemical reactivity of DNA in water at 37 °C and pH 7.4:

• Hydrolytic Base Deamination: The exocyclic amino groups of cytosine, 5-methylcytosine (5mC), adenine, and guanine undergo spontaneous nucleophilic attack by water. Approximately 100 to 500 cytosine residues per human cell per day deaminate into uracil (U), generating promutagenic U:G mismatches; if left unrepaired prior to the next round of DNA replication, DNA polymerases pair U with adenine (A), fixing a permanent C:G → T:A transition mutation. Similarly, 5-methylcytosine at methylated CpG dinucleotides deaminates at a 2- to 4-fold higher rate directly into thymine (T), creating a T:G mismatch that accounts for roughly one-third of all somatic and germline point mutations in human genetic disease and oncology. Adenine deaminates to hypoxanthine (Hx) (driving A:T → G:C transitions), and guanine deaminates to xanthine (X).
• Spontaneous Depurination and Depyrimidination: The N-glycosidic bond linking purine bases (adenine and guanine) to the C1' carbon of 2'-deoxyribose undergoes spontaneous hydrolytic cleavage at a rate of ~9,000 to 10,000 purines lost per mammalian cell per day (alongside ~500 pyrimidines lost per day), leaving an apurinic/apyrimidinic (AP or abasic) site that stalls replicative polymerases.

Oxidative Base Damage and Non-Enzymatic Alkylation

Cellular aerobic respiration in mitochondria, inflammatory burst by NADPH oxidases, and ionizing radiation generate reactive oxygen species—including superoxide (O2•−), hydrogen peroxide (H2O2), and the hydroxyl radical (•OH)—that oxidize nucleobases at thousands of sites per cell daily. Because guanine possesses the lowest oxidation potential among the four canonical DNA bases, its C8 position is oxidized to form 7,8-dihydro-8-oxoguanine (8-oxoG or GO) (~1,000 to 2,000 lesions/cell/day) or ring-opened FapyG. During DNA replication, 8-oxoG rotates from the normal anti conformation into the syn conformation to form a Hoogsteen base pair with adenine (8-oxoG:A), driving G:C → T:A transversion mutations. Pyrimidines are oxidized into thymine glycol (Tg) and 5-hydroxyuracil, while endogenous methyl donors (S-adenosylmethionine) alkylate ring nitrogens to produce cytotoxic 3-methyladenine (3meA) and 7-methylguanine (7meG).

Step 1: Base Recognition and Base Flipping by DNA Glycosylases

Base excision repair is initiated by 11 mammalian DNA glycosylases, each evolved with a specialized active-site pocket tailored to recognize a specific subset of damaged or mismatched bases without requiring ATP. Scanning duplex DNA via facilitated diffusion, a DNA glycosylase bends the DNA backbone at a lesion and executes nucleotide flipping (base flipping): the target nucleoside is rotated ~180° out of the double-helical base stack into the enzyme's extrahelical catalytic pocket, where the enzyme cleaves the C1'–N9 or C1'–N1 N-glycosidic bond to release the free damaged base:

1. Monofunctional DNA Glycosylases (UNG, SMUG1, TDG, MBD4, MPG/AAG, MUTYH): Use an activated water molecule as a nucleophile to hydrolyze the N-glycosidic bond directly, leaving an intact sugar-phosphate backbone containing an abasic (AP) site. Notably, MUTYH (MutY homolog) recognizes post-replicative 8-oxoG:A mispairs and excises the undamaged adenine base incorporated opposite 8-oxoG, giving Pol λ/Pol β a second opportunity to insert cytosine (8-oxoG:C) for subsequent repair by OGG1.
2. Bifunctional DNA Glycosylases / AP Lyases (OGG1, NTHL1, NEIL1, NEIL2, NEIL3): Specific for oxidative lesions, these enzymes use an active-site amine (Lys ε-NH₂ in OGG1/NTHL1, or N-terminal Pro1 α-NH₂ in NEIL1–3) to form a covalent Schiff base intermediate with the C1' carbon and catalyze an intrinsic AP lyase cleavage: β-elimination (OGG1, NTHL1) cleaves 3' to the AP site to leave a 3'-α,β-unsaturated aldehyde (3'-PUA) and 5'-phosphate, whereas β,δ-elimination (NEIL1, NEIL2) cleaves both 3' and 5' bonds to leave a 3'-phosphate and 5'-phosphate gap.

Step 2–4: APE1 Incision and Short-Patch (1 nt) vs. Long-Patch (2–12 nt) Completion

Following base excision by a monofunctional glycosylase, AP endonuclease 1 (APE1) hydrolyzes the phosphodiester backbone 5' to the AP site, generating a single-nucleotide nick with a 3'-OH primer and a 5'-deoxyribose phosphate (5'-dRP) moiety (whereas after bifunctional lyase action, APE1 removes 3'-PUA or polynucleotide kinase 3'-phosphatase [PNKP] removes 3'-phosphate to restore a clean 3'-OH). Repair then bifurcates into two sub-pathways:

1. Short-Patch BER (1-nt replacement, ~80%–90% of events): DNA Polymerase β (Pol β) uses its 31-kDa polymerase domain to insert a single nucleotide onto the 3'-OH and its 8-kDa N-terminal dRP lyase domain (Lys72) to remove the 5'-dRP group via β-elimination. The scaffold protein XRCC1 (recruited by PARP1 poly(ADP-ribosyl)ation) then positions DNA Ligase IIIα to seal the nick.
2. Long-Patch BER (2 to 12 nt strand displacement, ~10%–20% of events): When the 5'-dRP sugar is chemically oxidized or reduced so Pol β lyase cannot cleave it, Pol β or replicative DNA Polymerase δ / ε (with PCNA) synthesizes 2 to 12 nucleotides, displacing a 5' single-stranded flap that is cleaved by Flap Endonuclease 1 (FEN1) and sealed by DNA Ligase I.

Pathway Integration, MUTYH Polyposis, and PARP Inhibitor Synthetic Lethality

While BER repairs small base lesions, bulky UV photoproducts are excised by Nucleotide Excision Repair (NER), replication errors by DNA mismatch repair, and double-strand breaks by non-homologous end joining or homologous recombination. Germline biallelic loss of MUTYH causes autosomal recessive MUTYH-Associated Polyposis (MAP), driving colorectal adenomas and carcinomas via G:C → T:A transversions in APC and KRAS. In oncology, PARP inhibitors (olaparib, niraparib, rucaparib, talazoparib) inhibit PARP1 catalytic activity and trap PARP1 on BER single-strand break intermediates; during S-phase, colliding replication forks convert trapped PARP complexes into double-strand breaks that selectively kill BRCA1/2-deficient breast, ovarian, pancreatic, and prostate tumor cells via synthetic lethality.

Why does DNA contain thymine (5-methyluracil) instead of uracil as its canonical genetic base?

As Tomas Lindahl demonstrated, cytosine in cellular DNA spontaneously undergoes hydrolytic deamination to form uracil (U) hundreds of times per human cell every day (C → U). If uracil were a standard, natural base of DNA (as it is in RNA, where U pairs with A), repair enzymes would have no chemical way to distinguish a newly deaminated mutagenic U arising from a former C:G pair from a legitimate ancestral U:A pair, causing rapid, irreversible loss of C:G base pairs across evolution. By methylating the C5 carbon of uracil to use thymine (5-methyluracil, T) as the canonical adenine-pairing base in DNA, any unmethylated uracil base detected in duplex DNA is recognized as foreign damage by uracil-DNA glycosylase (UNG) and excised via BER.

What determines whether a cell uses Short-Patch versus Long-Patch Base Excision Repair?

Approximately 80% to 90% of cellular BER proceeds through the Short-Patch (1-nt) pathway because unmodified abasic sites incised by APE1 carry a standard 5'-deoxyribose phosphate (5'-dRP) group whose open-chain aldehyde readily forms a covalent Schiff base with Lys72 in the 8-kDa N-terminal lyase domain of Pol β, enabling β-elimination and XRCC1–Ligase IIIα sealing. However, if the C1' carbon of the abasic sugar has been chemically oxidized (forming 2-deoxyribonolactone) or reduced, it cannot undergo β-elimination, forcing the repair machinery to switch to Long-Patch (2–12 nt) strand-displacement synthesis followed by FEN1 flap cleavage and Ligase I sealing.

How does the glycosylase MUTYH work together with OGG1 to prevent oxidative mutations?

The dual glycosylase system (OGG1 plus MUTYH, corresponding to bacterial MutM and MutY) executes a two-tier defense against 8-oxoguanine (8-oxoG). Prior to DNA replication, OGG1 recognizes 8-oxoG paired opposite cytosine (8-oxoG:C) and excises the oxidized 8-oxoG base directly so Pol β can restore a normal G:C pair. However, if a replication fork encounters an unrepaired 8-oxoG before OGG1 arrives, the replicative polymerase frequently misincorporates adenine opposite 8-oxoG (8-oxoG:A). If OGG1 were to excise 8-oxoG from an 8-oxoG:A mispair, the polymerase would insert a thymine opposite the remaining A, fixing a G:C → T:A mutation. Instead, MUTYH recognizes the 8-oxoG:A mispair and excises the undamaged adenine, allowing Pol λ/Pol β to insert a cytosine (8-oxoG:C) and handing the substrate back to OGG1.

1. Lindahl, T. (1974). "An N-glycosidase from Escherichia coli that releases free uracil from DNA containing deaminated cytosine residues." Proceedings of the National Academy of Sciences, 71(9), 3649–3653. https://doi.org/10.1073/pnas.71.9.3649
2. Lindahl, T. (1993). "Instability and decay of the primary structure of DNA." Nature, 362(6422), 709–715. https://doi.org/10.1038/362709a0
3. Krokan, H. E., & Bjørås, M. (2013). "Base excision repair." Cold Spring Harbor Perspectives in Biology, 5(4), a012583. https://doi.org/10.1101/cshperspect.a012583
4. Beard, W. A., Horton, J. K., Prasad, R., & Wilson, S. H. (2019). "Eukaryotic base excision repair: new approaches shine light on structures and mechanisms." Annual Review of Biochemistry, 88, 137–162. https://doi.org/10.1146/annurev-biochem-013118-111315