Wait, What? DNA Polymerase Can Proofread and Still Leave the Wrong Base Behind
Replication is extraordinarily accurate, but not perfect.
Polymerase proofreading removes many newly inserted errors. A small fraction survive. Those residual mismatches are dangerous because after the next round of replication they can become ordinary, permanent base pairs.
Eukaryotic mismatch repair, or MMR, is the post-replication system that finds many of these remaining errors and removes the newly synthesised strand around them.
The central problem is not only finding the mismatch. It is knowing which strand is wrong.
The One-Sentence Answer
Learn eukaryotic mismatch repair by following MutSα or MutSβ from mismatch recognition to ATP-driven signalling, MutLα activation, PCNA-directed strand incision, excision, DNA resynthesis and ligation—then ask how failure produces microsatellite instability and characteristic cancer risk.
Stage 1: Start With the Error Types
Replication can create:
- base–base mismatches, such as G paired with T;
- small insertion/deletion loops when polymerase slips;
- repeat-length errors in microsatellites.
These are structurally different from many bulky chemical lesions. MMR therefore has its own recognition machinery.
Stage 2: MutSα Recognises Most Base Mismatches
Human MutSα is a heterodimer of MSH2 and MSH6. It recognises many base–base mismatches and small insertion/deletion loops.
Structural studies show that the complex bends distorted DNA and makes asymmetric contacts with the mismatch.
Stage 3: MutSβ Extends the Detection Range
MutSβ contains MSH2 and MSH3. It is especially important for larger insertion/deletion loops.
The use of overlapping sensors expands the error space that MMR can inspect.
Stage 4: Recognition Is Not Simple Sequence Reading
MMR proteins detect abnormal DNA geometry and dynamics rather than looking for a universal “wrong-letter sequence”.
The same base can be correct in one pairing and wrong in another.
Stage 5: MutS Changes State After Finding a Mismatch
Mismatch recognition promotes nucleotide exchange in MutS complexes. ATP binding drives a conformational transition that can create a mobile sliding-clamp-like state on DNA.
This separates:
error recognition from downstream signalling.
Stage 6: One-Dimensional Motion Helps Connect the Error to Repair Machinery
Single-molecule studies support models in which MutS complexes diffuse along DNA after mismatch recognition.
The precise choreography remains an active area of study, but the important idea is that the detector can convert into a signalling platform rather than remaining locked at one base pair.
Stage 7: MutLα Is a Major Downstream Effector
MutLα is a heterodimer of MLH1 and PMS2. It interacts with mismatch-recognition complexes and replication-associated factors.
PMS2 contains a latent endonuclease activity that becomes crucial for strand-directed repair.
Stage 8: The Hardest Question Is Strand Discrimination
After replication, one strand contains the original template sequence and the other contains the new error.
If the system removes the old correct strand instead, repair can convert a temporary mismatch into a permanent mutation.
MMR therefore needs evidence of nascent-strand identity.
Stage 9: PCNA Provides Orientation Information
PCNA is loaded onto newly synthesised DNA by RFC with defined orientation at primer–template junctions and nicks.
MutLα can interact with PCNA and introduce additional strand-directed incisions.
This helps convert replication architecture into repair directionality.
Stage 10: The Strand Signal Is Not the Bacterial Methylation Rule
Classic bacterial MutH-dependent MMR uses methylation state in organisms such as E. coli.
Human cells do not use that same mechanism.
This is a major misconception created by transferring one textbook model across domains of life.
Stage 11: MutLα Endonuclease Activity Creates Entry Points
Activated MutLα can nick the discontinuous nascent strand. These additional incisions make error-containing DNA accessible to excision machinery.
Repair direction is therefore partly generated rather than merely read from one permanent marker.
Stage 12: EXO1 Can Remove the Error-Containing Tract
Exonuclease 1 can excise DNA from a nick past the mismatch, creating a single-stranded gap.
MMR can also proceed through partially EXO1-independent routes, so no single excision enzyme should be treated as the universal mechanism in every context.
Stage 13: Replication Protein A Protects Single-Stranded DNA
RPA binds exposed single-stranded DNA during repair.
This prevents inappropriate secondary structures and helps organise the repair intermediate.
Stage 14: DNA Polymerase Resynthesises the Missing Segment
DNA polymerase δ and associated replication factors fill the gap using the intact strand as template.
DNA ligase seals the final nick.
The completed route is:
recognise → signal → identify nascent strand → incise → excise → resynthesise → ligate
Stage 15: MMR Works Closely With the Replication Machinery
PCNA, RFC and newly created discontinuities make MMR tightly coupled to replication.
This is why “post-replication repair” does not mean “long after replication is finished”. Much of the useful strand information is most available near the replication fork.
Stage 16: Chromatin Changes the Search Problem
Human DNA is wrapped in nucleosomes. MMR proteins therefore do not scan naked DNA in vivo.
Histone marks, nucleosome assembly and replication-coupled chromatin dynamics influence where repair proteins are recruited and how accessible mismatches remain.
Stage 17: H3K36me3 Can Help Recruit MutSα
The MSH6 subunit contains a PWWP-related region that can interact with H3K36me3-associated chromatin.
This provides one route linking epigenetic state to mismatch surveillance.
Stage 18: Proofreading and MMR Are Different Layers
Polymerase proofreading acts at the polymerase active site during synthesis.
MMR acts after a mismatch escapes proofreading.
Removing one layer increases mutation rate even if the other remains intact.
Stage 19: Microsatellites Are Especially Sensitive to MMR Loss
Short tandem repeats are prone to polymerase slippage.
Without effective MMR, repeat lengths change frequently.
This produces microsatellite instability, or MSI.
Stage 20: MSI Is a Molecular Phenotype, Not a Gene Name
A tumour may be MSI-high because MMR function is deficient, but the upstream cause can differ:
- germline pathogenic variants;
- somatic mutations;
- epigenetic silencing such as MLH1-promoter methylation.
Same downstream phenotype. Different causal routes.
Stage 21: Lynch Syndrome Reveals the Cancer-Suppression Role
Inherited pathogenic variants in MLH1, MSH2, MSH6 or PMS2 can predispose to Lynch syndrome.
Loss of the remaining functional copy in susceptible cells can create MMR deficiency and accelerate mutation accumulation.
This is an educational explanation of mechanism, not personal genetic or medical advice.
Stage 22: MMR Deficiency Changes the Mutation Spectrum
When MMR fails, tumours often accumulate characteristic base substitutions and insertion/deletion patterns, especially at repetitive sequences.
Mutation signatures therefore provide a genomic record of past repair failure.
Stage 23: MMR Deficiency Can Increase Neoantigen Burden
High mutation and frameshift rates can generate abnormal peptides. In some cancers this increases immune recognition and helps explain why MSI-high/MMR-deficient status can predict response to immune-checkpoint therapy.
A repair defect can therefore alter the tumour–immune relationship.
Stage 24: MMR Can Also Respond to Certain DNA Lesions
MMR proteins recognise some chemically modified base pairs, including lesions involving O6-methylguanine.
Repeated attempts to process a lesion that is not removed by MMR can trigger checkpoint and cell-death pathways.
Repair machinery can become damage signalling.
Stage 25: Repeat Expansion Diseases Reveal a Different Side of MMR
MutSβ and other MMR components can promote somatic expansion of certain trinucleotide repeats in disorders such as Huntington’s disease.
A genome-maintenance system can therefore become maladaptive when the substrate is a pathological repeat.
Stage 26: Immunohistochemistry Measures Protein Presence
Clinical pathology often examines MLH1, PMS2, MSH2 and MSH6 protein expression.
Loss patterns can suggest which complex is disrupted.
But protein presence does not directly measure every catalytic step of repair.
Stage 27: MSI Testing Measures the Genomic Receiver
PCR-based or sequencing-based methods measure instability at repeat loci.
This captures the consequence of failed MMR rather than merely the abundance of repair proteins.
Stage 28: Single-Molecule Studies Reveal Search Dynamics
Fluorescently labelled MutS/MutL homologues can reveal scanning, clamp formation and diffusion on DNA.
These experiments expose kinetic states that static structures cannot show.
Stage 29: Structural Biology Reveals Recognition Geometry
Crystal and cryo-EM structures show how MutSα contacts mismatched DNA and how ATP-dependent conformations change the complex.
A structure explains what is possible; kinetics and perturbation show what is used.
Stage 30: Professional MMR Biology Is a Directionality Problem
The advanced question becomes:
How does the repair system couple mismatch recognition to a temporary mark of the newly synthesised strand quickly enough to remove the error before that distinction disappears?
Evidence: How We Know
- Biochemical reconstitution demonstrates nick-directed human MMR.
- MutSα structures show mismatch-specific DNA deformation and recognition.
- MutLα experiments identify PMS2-dependent endonuclease activity.
- PCNA/RFC studies explain how replication architecture can bias strand incision.
- Lynch syndrome, MSI and mutation signatures reveal the organism-level consequences of MMR loss.
Misconceptions Worth Hunting
- “Proofreading and mismatch repair are the same.” They are separate fidelity layers.
- “Human MMR uses DNA methylation exactly like the classic bacterial model.” It does not.
- “MutS simply cuts out the wrong base.” Recognition, incision, excision and resynthesis are distributed across several complexes.
- “MSI is synonymous with Lynch syndrome.” Sporadic tumours can become MSI-high through other mechanisms.
- “MMR always protects against repeat expansion.” Some MMR components can promote pathological repeat growth.
Transfer Check
Case 1: A polymerase inserts G opposite T and proofreading misses it. Which system gets the next chance? MMR.
Case 2: MSH2–MSH6 recognises the mismatch but the cell cannot distinguish the nascent strand. Is recognition alone enough? No.
Case 3: A tumour lacks MLH1 protein and is MSI-high. Does that prove an inherited MLH1 variant? No. Epigenetic silencing and somatic events are alternatives.
Model Limits
Eukaryotic MMR is mechanistically complex and not every step is identical across organisms. Reconstituted naked-DNA systems simplify chromatin. MSI assays sample selected loci or sequence patterns rather than observing repair directly. Cancer phenotypes reflect many pathways beyond MMR alone.
Professional MMR reasoning keeps:
mismatch class + MutS complex + ATP state + nascent-strand signal + MutLα incision + excision + resynthesis + genomic outcome
visible together.
Connections Across the eduKate Science Estate
- Eukaryotic Base Excision Repair — a different repair pathway for damaged bases.
- RAD51–BRCA Homologous Recombination — high-fidelity double-strand-break repair.
- Eukaryotic Translesion DNA Synthesis — damage tolerance rather than ordinary mismatch correction.
Research Foundations
Foundational evidence comes from human MMR biochemical reconstitution, MutSα structural work, MutLα endonuclease studies, PCNA-directed strand-discrimination research, and modern genomic analysis of MSI/MMR-deficient tumours.
The Quiet Ending
The beginner asks, “How does the cell spot a wrong base pair?”
The developing molecular biologist asks, “How does it know which strand to remove?”
And the professional asks:
Which transient replication signal lets a mismatch-recognition event become accurate directional repair before the error is fixed into the genome?
